1
1
Well Control Well Performance
CONTENTS
1
1.2
INTRODUCTION
SYSTEMS ANALYSIS OF THE PRODUCTION
SYSTEM
1.3 HYDROCARBON PHASE BEHAVIOUR
1.4 RESERVOIR INFLOW PERFORMANCE
1.4.1 Liquid Inflow
1.4.2 Gas Inflow
1.4.3 2 Phase (Gas-Liquid) Inflow
1.4.4 Examples of IPRs
1.5 TUBING (OUTFLOW) PERFORMANCE
1.5.1 Tubing Pressure Traverse
1.5.2 The Tubing Friction Term
1.5.3 Introduction to Multiphase Flow in Vertical
Tubing
1.5.4 Prediction of Multiphase Fluid Properties
1.6 “GRADIENT” OR PRESSURE TRAVERSE
CURVES
1.7 FLOW MAPS AND CORRELATIONS
1.7.1 Duns and Ros
1.7.2 Hagedoorn and Brown
1.7.3 Beggs and Brill
1.7.4 Gray
1.8 TEMPERATURE MODELLING
1.9 SURFACE PRESSURE LOSSES
1.9.1 Surface Components
1.9.2 Flow Through Chokes
1.9.2.1Single Phase Liquid Flow
1.9.2.2Single Phase Gas Flow
1.9.2.3Multiphase Flow
1.9.3 Gathering System Layout
1.10 COMPLETIONS INFLOW PERFORMANCE
1.10.1 Perforated Completions
1.10.1.1 Perforation Charge Performance
1.10.1.2 Perforation Gun Selection
1.10.2 Gravel Packed Completion
1.10.2.1 Non-Darcy Turbulence Pressure Losses
1.10.2.2 Restriction of Gravel Pack Drawdown
1.11 COMPUTERISED WELL PERFORMANCE
PREDICTION PROGRAMS
1.12 WELL PERFORMANCE SENSITIVITY
STUDY EXERCISE
1.12.1 Reservoir Inflow and Tubing Outflow
Restrictions
1.12.2 Tubing Size and Liquid Loading
1.12.3 Effect of Water Cut and Depletion
1.12.4 Opportunities for Skin Removal by
Stimulation
1.12.5 Completion Design
1.12.6 Well Head Pressure
1.13 FURTHER READING
1
LEARNING OUTCOMES
Having worked through this chapter the Student will be able to:
• Describe Well Inflow and tubing vertical lift performance.
• Discuss the implementation of these concepts in computerised well completion
design programs.
• Discuss the need for artificial lift, i.e. the addition of external energy when the
natural reservoir energy is insufficient to continue economic production.
Later modules will extend these concepts by:
• Discussing the many artificial lift techniques and develops selection criteria.
• Describing the design process for a gas lift and electric Submersible Pump system.
2
1
Well Performance
1 INTRODUCTION
A simple producing system is illustrated in figure 1.
Gas
Pwh
Horizontal
Flow Line
Liquid
Psep
Separator
Surface
Vertical Tubing
Casing Annulus Isolation Packer
Figure 1
Simplified hydrocarbon
production system
Radial Flow in
Porous Media
Hyrodocarbon Reservoir
Pwf
PR
The hydrocarbon fluid flows from the reservoir into the well, up the tubing, along the
horizontal flow line and into the oil storage tank. During this process the fluid’s
pressure is reduced from the reservoir pressure to atmosphere pressure in a series of
pressure loss processes (Figure 2):
(1) across the reservoir
(2) across the completion (perforation/gravel pack etc.)
(3) across the tubing and any restrictions
(4) across the sub surface safety valve
(5) across the surface choke
(6) across flowline
Institute of Petroleum Engineering, Heriot-Watt University
3
1
These pressure losses can be grouped into three main components:
(7) summarises the total pressure losses in the reservoir and completion
(8) summarises the total pressure losses in the tubing
(9) summarises the total pressure losses at the surface
A pump or compressor are often used to aid evacuation of fluids (gas/water/oil) from
the separator. The separator is operated under gas pressure control and liquid (oil and
water) level control. Hence it normally acts as the end point of the flowing system
since a pump is necessary to aid evacuation of the liquids from the separator.
∆P9=(Pwh-Psep)
Gas
∆P6=(PDSC-Psep)
PWH
Psep
PDSC
Separator
Surface Choke
∆P5=(Pwh-PDSC)
Safety Valve
∆P4=(PUSV-PDSV)
PDSV
∆P8=(Pwf-Pwh)
PUSV
Bottomhole Restriction
∆P3=(PUR-PDR)
∆P1=(PR-Pwfs)
= Loss in Hydrocarbon Reservoir (Porous Medium)
= Loss Across Completion
∆P2=(Pwfs-Pwf)
= Loss Across Tubing and any Restrictions
∆P3=(PUR-PDR)
= Loss Across Safety Valve
∆P4=(PUSV-PDSV)
= Loss Across Surface Choke
∆P5=(Pwh-PDSC)
= Loss or Downstream
∆P6=(PDSC-Psep)
N.B. U refers to Upstream and D to Discharge or Downstream
SUMMARY PRESSURE LOSSES
= Total Loss in Reservoir and Completion
∆P7=(Pwf-PR)
= Total Loss in Tubing
∆P8=(Pwf-Pwh)
= Total Loss at the Surface
∆P9=(Pwh-Psep)
PDR
PUR
Pwf
∆P2=(Pwfs-Pwf)
PR
Pwfs
Pwf
PUR
PDR
PUSV
PDSV
PWH
PDSC
Psep
4
Liquid
Pwfs
PR
∆P1=(PR-Pwfs)
Reservoir Pressure
Flowing sand face Pressure
Flowing Bottom Hole Pressure
Upstream Restriction Pressure
Downstream Restriction Pressure
Upstream Safety Valve Pressure
Downstream Safety Valve Pressure
Well Head Pressure
Downstream surface Choke Pressure
Separator Pressure
Figure 2
Pressure losses during
production
1
Well Performance
The magnitude of these individual pressure losses depend on the reservoir properties
and pressures; fluid being produced and the well design. Production Technologists/
Engineers need to understand the interplay of these various factors so as to design
completions which maximise profitability from the oil or gas production. There are
no standard “rules of thumb” which can be used. Figure 3 schematically represents
the pressure distribution across the production system shown in Figure 2. It identifies
the most significant components, flowline, tubing and the reservoir and completion
where pressure losses occur.
Table 1 was developed by Duns and Ross (“Vertical flow of gas and liquid mixtures
in wells”, Proc Sixth World Petroleum Conference, Frankfurt, Vol 2, paper 22, 1963)
to illustrate one possible distribution in a conventional land oil field developed with
vertical wells.
Well Productivity Index
(bopd/psi)
Table 1
Pressure Loss Distribution
as a Function of Well
Productivity Index
Production Rate
(bopd)
2.5
5.0
10.0
15.0
Pressure Loss Distribution (%)
Across Reservoir
and Completion
(∆ P7)
Across Tubing
(∆ P8)
Across Flowline
(∆ P9)
36
25
15
11
57
68
78
82
7
7
7
7
2700
3700
4500
4800
Reservoir
Drainage
Boundary
Sand Face
PR
Wellbore
Pwf
Tubing Restriction
Pressure
Safety Valve
Wellhead
Reservoir
Choke
Pwh
Completion
Psep
Figure 3
Pressure across production
system
Inflow ∆P7
Well ∆P8
Surface (∆P9)
Position
Institute of Petroleum Engineering, Heriot-Watt University
5
1
The corresponding figures for a field developed with horizontal wells (much greater
well productivity indices), subsea wells (long flowlines, possibly over hilly terrain)
and pipelines would have a very different distribution. This is due to the flow
occurring mainly in a horizontal direction rather than vertical orientation associated
with wells. The balance between the “Elevation terms” and the “Friction terms”
across the pipe (i.e. contribution to ∆ P8) change drastically as shown in table 2.
Pressure Loss Term
(vertical) well
Well
Orientation (horizontal) pipeline/well
Elevation
Friction
85–98%
2–15%
0–30%
70–100%
Table 2
Typical pressure loss
Distributions
1.2 Systems Analysis of the Production System
The use of systems analysis to design a hydrocarbon production system was first
suggested by Gilbert (“Flowing and Gas Lift Performance”, API Drilling and
Production Practices, 1954”). Systems analysis, which has been applied to many
types of systems of interacting components, consists of selecting a point or node
within the producing system (well and surface facilities). Equations for the relationship
between flow rate and pressure drop are then developed for the well components both
upstream of the node (inflow) and downstream (outflow). The flow rate and pressure
at the node can be calculated since:
(i)
Flow into the node equals flow out of the node.
(ii) Only one pressure can exist at the node.
Further, at any time, the pressure at the end points of the system {separator (Psep) and
reservoir pressure (PR)} are both fixed. Thus:
PR
- (Pressure loss upstream components) = Pnode
(1)
Psep
+ (Pressure loss downstream components) = Pnode
(2)
Pressure
Node Outflow
Pressure at Node
Operating Point
Node Inflow
Flow Rate Through Node
Flow Rate
6
Figure 4
Node flow rate and
pressure
1
Well Performance
Typical results of such an analysis is shown in Figure 4 where the pressure-rate
relationship has been plotted for both the inflow (Equation 1) and outflow (Equation 2)
at the node. The intersection of these two lines is the (normally unique) operating
point. This defines the pressure and rate at the node. This approach forms the basis
of all hand and computerised flow calculation procedures. It is frequently referred to
as “nodal analysis”. (This name is also a trademark of Schlumberger Technology
Corporation for this process).
Having established the concept of nodal analysis, the following sections will discuss
how the hydrocarbon phase behaviour (Section 1.3) effects the reservoir fluid Inflow
Performance into the well (Section 1.4). The outflow, or tubing performance will be
reviewed (Section 1.5 et seq.) and the interaction between the in- and out-flow
discussed in Section 1.12 which gives a number of well performance sensitivity studies.
1.3 Hydrocarbon Phase Behaviour
Hydrocarbon reservoir fluids are a complex mixture of hydrocarbon molecules, the
composition of which is dependent on the source rock, degree of maturation etc. Phase
changes occur when this complex hydrocarbon fluid flows from the (high temperature
and pressure) reservoir environment to the (cool, low pressure) separator conditions.
Such changes are sketched for an undersaturated oil in Figure 5. Here it can be seen
that the fluid:
(a) Reservoir
(PR,TR)
(Pwf,Twf) (b) Wellbore
Critical Point
(c)
Pressure
Liquid Phase Only
le
bb
Bu
in
Po
100% Liquid
e
80%
Two
Phase
Region
in
tL
(d)
60%
40%
(e)
20%
(Psep,Tsep)
5% 0% Liquid
(f)
Separator
w
De
Figure 5
Schematic phase diagram
for an undersaturated oil
in
Po
tL
in e
Gas Phase Only
Temperature
(i)
is present as a single phase liquid in the reservoir {point (a)}
Institute of Petroleum Engineering, Heriot-Watt University
7
1
(ii) remains a single phase liquid at the wellbore (significant reduction in pressure
and small change in temperature during flow in reservoir) {point (b)}
(iii) starts to evolve gas {point (c)} as temperature and pressure are reduced during
flow up the tubing
(iv) evolves increasing amounts of gas {points (d) and (e)} until the separator {point
(f)} is reached.
Some or all of the flow regimes illustrated in figure 6 may occur.
The phase behaviour of the hydrocarbon fluid controls the fluid’s gas/liquid ratio as
a function of bottom hole pressure. This, in turn, will effect flow rate, i.e. the Inflow
Performance Relationship (IPR) discussed in section 1.4 and the outflow tubing
performance.
1.4 Reservoir Inflow Performance
The Inflow Performance Relationship (IPR)is routinely measured using bottomhole
pressure gauges at regular intervals as part of the field monitoring programme. This
relationship between flow rate (q) and wellbore pressure (Pwf) is one of the major
building blocks for a nodal-type analysis of well performance.
8
1
Well Performance
(f)
Psep,Tsep
Mist Flow
Separator
(e)
Annular Flow
Churn Flow
Wellhead
(d)
Slug Flow
(e)
(d)
Bubble Flow
(c)
(c)
Figure 6
Schematic view of possible
phase changes in tubing
Pbubble Tbubble
Pbubble, Tbubble (Bubble Point)
Single Phase Flow
(b)
Reservoir
(Pwf)
PRTR (a)
1.4.1 Liquid Inflow
Field measurements have shown that wells producing undersaturated oil (no gas at the
wellbore) or water have a straight line IPR (Figure 7).
q = PI (PR - Pwf)
(3)
where q is the flow rate and PI the Productivity Index, i.e. the well inflow rate per unit
of well drawdown.
Institute of Petroleum Engineering, Heriot-Watt University
9
1
Well Bore Flowing Pressure (Pwf)
PR (Reservoir Pressure)
Well
Drawdown
Pwf
q
Well Production
AOF
Liquid Flow Rate (q)
Figure 7
Staightline IPR (for an
incompressible liquid)
q (max)
A theoretical basis for the straight line IPR can be derived using Darcy’s Law, radial
inflow into the well along with other assumptions about rock and fluid properties. PI
is a useful tool for comparing wells since it combines all the relevant rock, fluid and
geometrical properties into a single value to describe (relative) inflow performance.
The Absolute Openhole Factor (AOF or qmax), is the flowrate at zero (bottomhole),
wellbore flowing pressure. AOF, although often representing unrealistic conditions,
is a useful parameter when comparing wells within a field since it combines PI and
reservoir pressure in one number representative of well inflow potential.
A straight line IPR can be determined from two field measurements:
(i)
the stabilised bottomhole pressure with the well shut in {reservoir pressure of (PR)}
(ii) the flowing, bottom hole, wellbore pressure (Pwf) at one production rate
The well’s inflow potential can then be calculated at any draw-down (or Pwf)
1.4.2 Gas Inflow
The compressible nature of gas results in the IPR no longer being a straight line.
However, the extension of this steady state relationship derived from Darcy’s Law,
using an average value for the properties of the gas between the reservoir and wellbore,
leads to
q = C (PR2 - Pwf2)
(4)
where C is a constant
This relationship is valid at low flow rates, but becomes invalid at higher flow rates
since non-Darcy (or turbulent) flow effects begin to be observed. This can be
accounted for by use of the “Bureau of Mines” equation that was developed from field
observations:
10
1
Well Performance
q = C (PR2 - Pwf2)n
(5)
where 0.5 <n <1.0
A log-log plot of q versus (PR2 - Pwf2) yields a straight line of slope n and intercept C.
Standard practice for testing gas wells is to measure the bottom hole, flowing,
wellbore pressure (Pwf) at four production rates. Figure 8 shows that the change of
slope from the initial value of 1 (Darcy flow, equation [4]). Non-Darcy flow effects
(equation [5], n<1) are observed at the two higher rates.
low
Both equations [4] and [5] are illustrated in Figure 8 which shows the >50% reduction
in AOF (from 1.4 to 0.9) due to these non-Darcy flow effects.
AOF
(PR2-Pwf2 )
Non
-Da
rcy
F
PR2
Da
rcy
F
low
Reduction Due to
Non-Darcy Flow
Figure 8
Gas well deliverability
reduced by non-Darcy flow
pressure losses
0.1
1.0
Test Rates
10
Gas Flow Rate (qG)
(logarithmic scale).
1.4.3 2 Phase (Gas-Liquid) Inflow
Straight line IPR (Section 1.4.1) are also not applicable to when two phase inflow is
taking place, e.g. when saturated oil is being produced. Vogel (“Inflow Performance
Relationships in Solution-Gas Drive Wells”, J Pet Tech, 1968, 83-92) proposed the
following equation based on a large number of well performance simulations:
Institute of Petroleum Engineering, Heriot-Watt University
11
1
q
q max
= 1 − 0.2( Pwf / PR ) − 0.8( Pwf / PR )
2
(6)
where qmax is the AOF, i.e. q when Pwf = 0
Vogel’s key contribution was the introduction of the concept of normalising the
production rate to the AOF value (qmax). Rewriting equation [6] in this manner gives:
q
q max
{
= 1 − ( Pwf / PR )
}
2 n
(7)
which is virtually equivalent to Vogel’s equation when n = 1 (Fetkovitch, “Isochronal
testing of oil wells”, SPE 4529, Las Vegas, Sept 1973). i.e:
q
q max
= 1 − ( Pwf / PR )
2
(8)
When multirate test data is available then equation [7] is preferred since it includes
high rate (non-Darcy or turbulent) effects. This is best done by plotting the data in a
similar manner to Figure 8, the resulting staight line has a slope of 1/n.
Figure 9 compares the production rate as a function of drawdown for an undersaturated
oil (straight line IPR, line A) and a saturated oil showing the two phase flow effects
discussed above (curve B). The figure also shows the special case (curve C) when the
wellbore pressure is below the bubble point while the reservoir pressure is above, i.e.
(incompressible) liquid flow is occurring in the bulk of the reservoir.
Normalised Wellbore Flowing Pressure (Pwf/PR)
1.0
Reservoir
Pressure (PR)
A Straight line IPR (undersaturated oil)
B Vogel or Curved IPR (saturated oil)
C Combination of A and B when reservoir
pressure was above the bubble point.
IPR becomes curved at the bubble point
Bubble Point
A
C
B
qb max
Oil Flow Rate
12
qc max
qa max
Figure 9
Inflow performance
relationships
1
Well Performance
1.4.4 Examples of IPRs
(i)
Reservoir Depletion
0.
2%
1%
ul
m
Cu
4%
a ti
6%
ve
8%
Oi
10%
lR
eco
ver
12%
14%
Wellbore Flowing Pressure
The previous section discussed the value of normalising the IPR. This is illustrated
with data on the IPR of a saturated (or solution gas drive) oil reservoir. Figure 10a
shows that the IPR rapidly decreases with increasing cumulative oil recovery. This
is not only due to reservoir pressure depletion; but is also related to the increasing
gas saturation which is making oil flow progressively more difficult.
y
Production Rate
Figure 10a
IPR curves
Normalised Wellbore Flowing Pressure (Pwf/PR)
A plot of the same data in a normalised manner (Figure 10b) shows that the curves
are quite similar, the increasing gas saturation being responsible for the (relatively)
greater drawdown for similar normalised production rate.
1.0
Cumulative Recovery
0.8
= 0.1%, 2%,4%,
6%, 8%
0.6
10%
12%
0.4
14%
0.2
0
0
0.2
0.4
0.6
0.8
1.0
Normalised Production Rate (q/q max)
Figure 10b
Normalised IPR curve
(ii) Crude Oil Properties
Institute of Petroleum Engineering, Heriot-Watt University
13
1
Consider the production of crude oil A, which is significantly more viscous than
crude oil B. Providing all other factors were kept constant, the IPR curves would
show substantial differences (Figure 11a) while the normalised IPR curves are
essentially the same (Figure 11b).
Wellbore Flowing Pressure (Pwf)
Oil A is More Viscous than Oil B
B
A
Production Rate (q)
Figure 11a
Actual IPR's
Normalised Wellbore Flowing Pressure (Pwf/PR)
1.0
0.8
B
0.6
A
0.4
0.2
0
0
0.2
0.4
0.6
0.8
1.0
Normalised Production Rate (q/q max)
Figure 11b
Dimensionless IPR
Figure 11b is a simple method to estimate the future well IPR as the reservoir
undergoes pressure depletion.
(iii) Heterogeneous Formations
All the above refers to a well producing from a homogeneous reservoir. Frequently,
wells are completed on heterogeneous formations where production from several
different zones is commingled. Reservoirs with differing permeabilities will be
depleted at different rates - the resulting composite IPR being the sum of the separate
individual IPRs (Figure 12). It will change as the well depth, fluid type, production
rate etc. alter.
14
1
Wellbore Flowing Pressure (Pwf)
Well Performance
Figure 12
Composite IPR for
heterogenuous formation
Composite IPR for
all three Zones
1-md
zone
0
10-md
zone
100-md
zone
Production Rate
1.5 Tubing (Outflow) Performance
Chapter 1.4 discussed the inflow of reservoir fluids into the wellbore and the pressure
drop required to achieve this. The outflow pressure drop required to lift a fluid from
the perforations to the wellhead and then the separator is the second factor which
determines the well production. This outflow performance will now be discussed.
Starting from the top of the well, the parameters which contribute to the pressure at
the bottom of the well are:
(i)
the (back) pressure at the wellhead
(ii) the hydrostatic head between the wellbore and wellhead. This is a function of
the change in elevation between the wellhead and the wellbore and the average
density of the fluid in tubing all multiplied by the acceleration due to gravity.
(iii) the pressure loss required to overcome friction losses due to viscous drag. This
depends on the fluid’s flow rate, flow regime and its viscous properties as well as the
length, diameter and roughness of the tubing.
NB - Pressure losses due to acceleration of the expanding fluid are normally low
and can be neglected.
Figure 13 illustrates the relative importance of these factors and their interaction of
these components for a given well depth as a function of flow rate and fluid type.
(i) Figure 13a is for a single (incompressible) liquid production. Being dense, the
hydrostatic head component is relatively large and constant (the density of water and
the heavier crudes, e.g. 20˚API, shows only minor variations with pressure and
temperature changes typically found in producing oil wells). The friction component increases rapidly, once turbulent flow is achieved, after the erratic behaviour
when the transition region between laminar and turbulent flow has been passed.
Institute of Petroleum Engineering, Heriot-Watt University
15
1
(ii) Figure 13b is for a gas well. The hydrostatic head component is now much
smaller, but increases with depth and rate since gas density is very pressure
dependent. Frictional pressure losses are normally the most important component,
turbulent flow being encountered even at low flow rates.
(iii) Multiphase (gas/liquid) production is illustrated in Figure 13c. The variation
of friction and hydrostatic pressure losses with production rate is complicated; their
relative importance may change, depending on the exact conditions.
NB It has been assumed that the reader is familiar with the basic fluid mechanics
equations that describe flow in pipes through the laws of:
(i)
conservation of mass and momentum (for calculating pressure changes) and;
Gas Rate
(b)
Liquid
Gas
Hydrostatic Head
Wellhead Pressure
Liquid Rate
(c)
ultihase Liquid
Gas iture
Figure 13
Components of tubing
pressure loss for different
fluids
n
ctio
Fri
Hydrostatic Head
Depth
Wellhead Pressure
Depth
Friction
ion
Frict
16
Liquid Rate
(a)
Wellhead
Wellhead Pressure
Depth
Wellhead Pressure
tion
Fric
Hydrostatic Head
Wellhead Pressure
Friction
Wellhead Pressure
Hydrostatic Head
ull
Turbulent
Bottomhole Pressure
Friction
Bottomhole Pressure
Turbulent
Transition
Laminar
Hydrostatic Head
Bottomhole Pressure
(ii) conservation of mass and energy (for calculating enthalpy, and hence temperature
changes).
Pressure
(a)
Pressure
(b)
Pressure
(c)
Liquid
Gas
Multiphase Liquid /
Gas Mixture
Figure 14
Tubing pressure traverse
1
Well Performance
1.5.1 Tubing Pressure Traverse
These differences between the three fluid systems are also apparent when the change
in pressure as a function of depth at a constant well production rate is considered
(Figure 14). This plot is known as a “Tubing Pressure Traverse” and the change in
total pressure with depth is known as a “gradient” curve. The behaviour of the
individual pressure components which make up the final gradient curve is summarised below.
(i) Figure 14a is for a single (incompressible) liquid. The hydrostatic head and
friction components, are straight lines. This is because the fluid density and the
friction loss per unit tubing length remain constant over the complete tubing length
(the latter assumes no restriction in the tubing).
(ii) Figure 14b is for a gas well. The (relative) contribution of the hydrostatic head
component increases with depth since the gas density increases with the total
pressure (i.e. well depth). The friction component shows the reverse behaviour - the
gas velocity is greatest at shallow depth since the pressure is lowest at this point and
the same amount of gas is entering and exiting the tubing. Thus the ratio:
(friction pressure/hydrostatic head) per unit length of tubing
increases as the well depth decreases i.e. the importance of the friction component is
less at greater well depths for gas flow in a constant diameter tubing.
The opposing behaviour of the friction and hydrostatic head components with depth
results in the (total) pressure traverse approximating a straight line.
This conclusion is only true for a single tubing diameter. Restrictions in gas wells can
often lead to unacceptably high pressure losses due to the consequent large increase
in fluid velocity.
(iii) A simple description of multiphase flow (Figure 14c) does not exist since
simple analytical equations etc are not available for this complex flow regime; but
the overall shape of Figure 14c is between the two earlier figures.
NB. In all the above cases the only parameter that is under the operational control of
the production engineer is the wellhead pressure or system “back pressure”. The
remainder of the completion can only be influenced by the engineer at the design stage.
This is thus the time when a wide range of sensitivity analyses should be performed
in order to ensure that the installed well will be “fit for purpose” during its lifetime.
Use of a standardised well design in a field can bring significant cost savings.
However, these have to be balanced against the costs e.g. foregone production, extra
workovers to change tubing size, etc, that a non-optimum well completion will bring.
Reduction in total, lifetime unit costs of constructing and operating the well is the aim
of the production engineer, optimising the profitability of field development.
Section 1.5.2 will discuss calculation of pressure losses in pipes. Section 1.5.3
discusses the importance of the phase behaviour of the hydrocarbon fluid followed by
the introduction of “Gradient Curves” as a simple means of describing multiphase
(outflow) tubing performance (Section 1.6).
Institute of Petroleum Engineering, Heriot-Watt University
17
1
1.0
Laminar,
Highly
Non-Newtonian
V / V max
Turbulant,
Newtonian
Laminar,
Newtonian
Pipe Wall
Pipe Centre
1.5.2 The Tubing Friction Term
Figs 13 and 14 schematically indicated the importance of the friction component when
predicting the pressure at any point in the wellbore. This frictional pressure will be
a function of the fluid characteristics (Newtonian or non-Newtonian fluid viscosities),
fluid flow conditions (velocity and laminar or turbulent flow) and the properties of the
tubing (diameter and roughness).
A full fluid mechanical description of all the situations that are encountered in
Production Engineering is beyond the scope of this text - however, this section is
designed to introduce some basic concepts for the simplest case - single phase flow
of an incompressible, Newtonian fluid.
(i) Reynolds Number
The Reynolds Number (Re) is the ratio of the inertial forces to the viscous forces for
fluid (density, ρ, and viscosity, µ) flowing in a circular pipe (diameter, D)
Re =
inertial forces  
Dvρ  ρv 2
or 

≡
µ  µv / D  viscous forces  
where v is the average fluid velocity.
The velocity profile of a Newtonian fluid flowing in LAMINAR flow is shown in
figure 15. Laminar flow is characterised by the individual fluid particles moving
ONLY in the flow direction with no fluid movement across the pipe, i.e. the fluid can
be pictured as flowing in a series of concentric tubes with the maximum velocity at the
pipe centre and a minimum velocity at the pipe wall. On the other hand, TURBULENT flow is characterised by rapidly fluctuating flow velocity components in
random directions.
18
Figure 15
Velocity profiles in laminar
and turbulent pipe flow
1
Well Performance
Newtonian fluids are defined as having a viscosity that is independent of shear rate.
Non-Newtonian fluids, by contrast, have a shear rate dependent viscosity, e.g. a shear
thinning fluid has an apparent viscosity which decreases as the shear rate increases.
They also have a very different velocity profile across the tubing. The third profile
in Figure 15 is for a highly shear thinning, non-Newtonian fluid. This type of fluid
is encountered in hydraulic fracturing stimulations and gravel pack operations. Their
unusual behaviour allows high concentrations of solid material (gravel pack sand or
proppant) to remain in suspension at low shear rates while the, apparently highly
viscous, fluid can be pumped through the tubing with lower pressure drops than would
be expected if pure water was being pumped.
Not only will the frictional pressure drop across a length of tubing be different between
the laminar and turbulent cases, but the flow velocity profiles will have significant
consequences for several production operations (Figure 16). For example:
Direction of Flow
Fluid 2
Fluid 2
Fluid 2
Extensive contamination
of fluid1 by fluid 2, in a limited
mixing zone
but bulk of
fluids still
Mixture
remain
Fluid 1 & 2
separate
Inter penetration of
fluid 1 by fluid 2
occurs over
a large
tubing
length
Limited
Mixing
Zone
Fluid 1
Figure 16
Mixing of fluid in pipe flow
as a function of flow regime
Fluid 1
Cross Sections
Turbulent
Fluid 1
Both Fluids
Laminar
Cross Sections
(i) When pumping a series of fluids into a well which should not mix e.g. for a sand
control treatment; the mixing zone between the fluids will be small if both fluids are
in turbulent flow. On the other hand, laminar flow allows the centre portion of the
trailing fluid to penetrate a long way into the leading fluid. This results in the two
fluids arriving simultaneously at the bottom of the tubing and being mixed during
injection into the perforation.
(ii) Mixing of two fluid streams being combined in a T-piece will only occur
rapidly if the flow is turbulent. Laminar flow will result in concentration gradients
occurring in the transverse direction across the pipe for a substantial distance
downstream of the T-piece.
Institute of Petroleum Engineering, Heriot-Watt University
19
1
10000
Reynolds Number, vs Flow Rate
For 1.0 gm/cc Fluid
4
3
Turbulent
Flow
5
1 Centpoise
8
(in
Transition
Zone
R
3
)
5
(in
s
6
pe
Pi
1000
4
i
ad
u
di
8
a
pe
R
10 Centpoises
Pi
5
6
(in
4
3
)
s
iu
8
Reynolds
Number,
Reynolds
Number
(Re) RNe
6
)
us
ad
pe
R
Laminar
Flow
Pi
100
100 Centpoises
10
10
100
1000
Flow
FlowRate
Rate(bbl/d)
(bbl/day)
Laminar flow is characterised by low Reynolds Numbers. Turbulent flow first
becomes apparent at a Reynolds Number of 2100 with fully turbulent flow being
observed at about 3500 and higher. Figure 17 plots the Reynolds Number as a function
of fluid viscosity, pump rate and pipe radius. It can be seen that, at the typical flow
rates encountered in petroleum engineering, fluids with a water-like viscosity are
normally in turbulent flow while viscous oil is in laminar flow.
(iii) Frictional Pressure Drop
Experiments have been made to measure the pressure drop (per unit length of pipe)
of a liquid flowing through a pipes of known diameter. The measurements were
repeated with pipes of differing materials and also with smooth wall pipes which had
been deliberately treated to create a surface of known roughness. All the possible
combinations of variables have been studied by varying the flow rate and fluid
viscosity as well.
These experiments showed that the frictional pressure drop (∆P) may be calculated
from the Fanning equation:
fmρv 2 L
∆P =
2D
where fm is the Moody friction factor, ρ the density of the fluid flowing at a velocity
v in a pipe of length L and Diameter D.
(a) Laminar Flow (Re <2000): The frictional pressure drop is independent of
tubing roughness and is proportional to the fluid velocity. The friction factor (fm) is
inversely proportional to the Reynolds Number
20
Figure 17
Reynolds number with
volumetric flow rate,
viscosity, and pipe size
1
Well Performance
fm = 64/Re
(b) Turbulent Flow (Re >2000): The frictional pressure drop is very sensitive to the
exact nature of the inner pipe wall as well as to the fluid flow conditions (Reynolds
Number). Experiments showed that the important parameter was the relative pipe
roughness (ε/D) where ε refers to the absolute height of roughness features that
protrude from the pipe surface into the flow stream and D the pipe diameter.
The Chen equation (Chen, “An explicit equation for friction factors in pipes”, Ind.
Eng. Chem. Fund., 18, p296, 1979) is probably the easiest equation for calculating the
friction factor:
0.8981
 ε
 
 ε1.098
 7.149 
1
5.0452
log 
= − 4 log 
−
+


fm
Re
 
 2.8257  R e 
 3.7065
0.1
TZ*
LZ*
.09
Turbulent Zone
.08
Complete Turbulence, Rough Pipes
.05
.07
.04
.06
.03
.05
.02
.015
∫=6
.01
.008
4/R e
Moody .04
Friction
Factor
(f)
Relative
Roughness
.006
.03
.004
ε=
K
D
.025
.002
.02
.001
.0008
.0006
.0004
Smooth
Pipes
.015
.0002
.01
Figure 18
Moody friction factor
diagram
*LZ = Laminar Zone
.0001
*TZ = Transition Zone between
Laminar Flow and Turbulent Flow.
.00005
.009
.008
.00001
3
10
2
3
4
4
5 6 8 10
2
3
4
5
5 6 8 10
2
3
4
6
5 6 8 10
Re = Reynolds Number =
2
3
4
7
5 6 8 10
2
3
4
8
5 6 8 10
Dvp
µ
It has a similar accuracy to the more normally quoted Colebrook-White equation,
which was used to generate Figure 18 (a plot of the Moody friction factor as a function
of Reynolds Number and turbulence).
The (absolute) pipe roughness depends on many factors; the bulk of which the
engineer has little control over. These include:
• Pipe metallurgy and any coating materials applied.
• Fluid velocity (erosion at high rates) and fluid corrosivity (pH, the presence of
solids, CO2, H2S etc).
• Deposits (hydrates, paraffins, asphaltenes).
• Years in service.
Institute of Petroleum Engineering, Heriot-Watt University
21
1
It comes as no surprise to learn that, in turbulent flow where it is an important
parameter, roughness is normally treated as an empirical parameter which is used as
a fitting parameter to match calculated results to actual pressure drop measurements.
However, the roughness value used must be realistic. Table 3 quotes typical values
for use in these calculations:
Material
Roughness
Plastic Pipe or Coating
New Tubing
Dirty Well Tubing
0.0
0.00005
0.00075
Table 3
Typical pipe roughness
values
1.5.3 Introduction to Multiphase Flow in Vertical Tubing
Multiphase flow would be greatly simplified if the two phases behaved as a homogeneous mixture whose properties were an appropriately averaged value of the individual phase properties. However, experiments have shown that this is not the case
and that one fundamental phenomenon occurring in vertical multiphase (oil-gas,
water-oil, etc) flow is the concept of SLIP and HOLD UP. These phenomenon are
most important for the gas/liquid case since the density differences are greatest:
(i) SLIP refers to the ability of the less dense (“lighter”) phase to flow at a greater
velocity than the denser (“heavier”) phase.
(ii) HOLD UP is a consequence of slip - the volume fraction of the pipe occupied
by the denser phase is greater than would be expected from the (relative) in - and
outflow of the two phases - since its flow velocity is slower than that for the light
phase.
NB. This accumulation of the denser phase in the pipe is an equilibrium phenomenon
i.e. the in- and out-let flow rates of a particular phase flowing in the pipe are the same.
In/Outflow
VL=VSL, HL=λL, VG=VSG
Insitu
VG
λL
VL
HL
Phase Ratio
NO SLIP
Phase Ratio
In/Outflow
VL<VSL, HL>λL, VG>VSG
Insitu
VG
λL
VL
HL
Phase Ratio
SLIP
Phase Ratio
These concepts can be best understood with the help of the following mathematical
description and Figure 19.
22
Figure 19
Volume fraction changes
when slip occurs during
flow
1
Well Performance
NB. Subscripts G, L, O and W refer to the gas, liquid, oil and water phases
respectively, while q is the phase volume flow rate, V the velocity and Ap the cross
sectional area of the pipe.
Superficial phase velocities (VSL and VSG) are given by:
VSL = qL/Ap and VSG = qg/Ap
In situ (or actual) velocities (VL and VG) are given by:
VL = qL/AL and VG = qG/AG
where AL and AG are the actual areas of the pipe occupied by that liquid and phases
respectively.
AL and AG under NO SLIP conditions can be calculated from the in- and out-flow
phase rates
AL = qL/(qL + qG) and AG = qG/( qL + qG)
The slip condition can be quantified by the liquid Holdup (HL) defined as the Fraction
of the pipe filled with liquid:
HL = AL/Ap and HG = AG/AP = 1 - HL
and the No Slip Holdup:
λL = qL/(qL + qG)
where λL is the input liquid volume fraction.
The relationship between all these variables is illustrated in figure 19.
Hence if slip occurs, then the slip velocity, Vs, is given by:
Vs = VG – VL = VG/HG - VL/HL
1.5.4 Prediction of Multiphase Fluid Properties
The above concepts can be used to predict some of the properties of the multi phase
mixture using a phase averaging mixing rule e.g. the density of a liquid/gas mixture
(ρm) is:
ρm = ρLfL + ρG(1 – fL)
where fL is the liquid volume fraction. This is equal to:
ρm = ρLλL + ρG{1 – λL}
or
NO SLIP
ρsm = ρLHL + ρG{1 – HL } SLIP
Institute of Petroleum Engineering, Heriot-Watt University
23
1
other properties, such as viscosity, cannot be predicted by this averaging technique
and require the use of special correlations.
(i)
Liquid/Liquid Flow
Downhole sampling and video has shown that many light oils are flowing as two
separate phases which only form an emulsion once the fluid is subjected to a high
shear rate e.g. in the surface choke. The oil and water are flowing as separated phases
with one phase will form the continuous phase with the second phase being dispersed
as small droplets within this continuous phase. The (oil) volume fraction at which
the continuous phase will change from being oil to water depends on the oil
properties, in particular the surface tension and the amount and types of any surface
active chemicals that the oil contains. The size of the discontinuous phase droplets
will depend on the above oil density, the oil and water fluid properties as well as the
type and amount of shear that the mixture is subjected to.
Most properties of the liquid/liquid mixture can be calculated with the phase averaging
mixing rule discussed at the beginning of this chapter. One exception is the viscosity
- the emulsion formed when mixing oil and water have been experimentally observed
with viscosities up to 50 times greater than that predicted from the averaging rule
(Figure 20). The height of this viscosity maximum is dependent on:
(a) the extent and type of shear imposed on the oil-water mixture and
(b) the type of emulsion formed; loose, (i.e. easy to separate), medium or
tight (i.e. difficult to separate).
Viscosity
N.B. The increased viscosity of low API gravity crude oils often leads to greater
separation problems in the surface facilities.
µo
µm
=µ
o fo
0
24
+µ
ω (1
-f
o)
Water Fraction (fω)
µω
1.0
Figure 20
Schematic representation of
the viscosity of water / oil
mixtures
1
Well Performance
Studies have also shown that the oil concentration at which the emulsion changes from
being oil phase continuous to being water phase continuous is related to the oil
viscosity (and hence density). It can thus be seen that the emulsion properties are quite
specific to the system being considered e.g. in one field use of an electric submersible
pump to artificial lift a viscous oil/water mixture can result in breakage of the pump
drive shaft at oil/water ratios near the inversion point (or zone of maximum viscosity).
These problems are often not observed during earlier production experience using a
Rod Pump - this is because of the greater shear being imposed by the rapidly rotating,
centrifugal impeller of the electric submersible pump was required to form the high
viscosity emulsion. Specific laboratory tests can be very useful in ensuring a proper
design is made if the emulsion viscosity is potentially a critical parameter in the
production system.
Liquid/liquid flow in tubings shows many of the flow phenomena which are discussed
in the following section on gas/liquid flow; though the range of flow regimes
observable during gas/liquid flow is much greater due to the greater density contrast
between the phases and the greater velocity associated with gas flow.
(ii) Gas/Liquid (Oil) Flow
Figs 5 and 6 introduced the concept of phase changes to the hydrocarbon fluid as it
travels up the tubing. The following is an elaboration of these ideas and introduces
the concept of a flow map to describe multiphase tubing flow which will be combined
with the multiphase flow concepts described in the previous chapter.
Figs 5 and 6 relate to the favourable production case when the wellbore flowing
pressure (Pwf) is greater than the bubble point i.e. single phase oil is entering the well.
Single phase flow in the tubing will continue until the pressure (and temperature)
reduce sufficiently that the bubble point is reached (point c).
The flow patterns in the tubing that will result from this gas bubble formation is a
function of:
(a) gas and liquid flow rates
(b) pipe angle of inclination
(c) pipe diameter
(d) phase densities
(A) FLOW IN VERTICAL TUBING (Figure 21)
(i) at the point (c) that the first gas bubbles appear the fluid mixture’s velocity in
the tubing will increase and the average fluid density decrease.
(ii) The initially formed bubbles will be widely dispersed within the liquid.
Continued flow up the tubing results in a further pressure reduction, increasing the
number of bubbles - which still remain widely dispersed in a continuous liquid phase.
This is called “bubble flow” regime.
Institute of Petroleum Engineering, Heriot-Watt University
25
1
(f)
Psep,Tsep
Mist Flow
Separator
(e)
Annular Flow
Churn Flow
Wellhead
(d)
Slug Flow
(e)
(d)
Bubble Flow
(c)
(c)
Pbubble Tbubble
Pbubble, Tbubble (Bubble Point)
Single Phase Flow
(b)
Reservoir
(Pwf)
PRTR (a)
(iii) Further upward movement of the produced fluid generates an increasing volume
(and mass) of gas phase; with a corresponding reduction in the mass (and volume) of
the liquid phase. Intense mixing will ensure the gas and liquid phases remain in
equilibrium as the pressure reduces i.e. the composition of the gas phase will change
with the evaporation of progressively higher molecular weight, hydrocarbon molecules. Availability of an “equation of state” for describing the reservoir fluids PVT
properties allows “flash” calculations to be carried out, i.e. to calculate the composition of the liquid and gas phases at any required temperature and pressure.
26
Figure 21
Schematic view of phase
changes in tubing
1
Well Performance
(iv) The increasing tubing volume fraction occupied by the gas allows bubble
coalescence to occur to such an extent that they fill the entire pipe cross section and
form a slug i.e. a series of very large gas bubbles of roughly constant size separated
from each other by areas of liquid containing smaller gas bubbles. This is called the
“slug flow regime”.
Gas slugs can act as an efficient mechanism to lift liquid to the surface.
(v) Velocity increases, associated with continued expansion of the available gas
and further volume increase of the gas phase, eventually result in the large gas slug
breaking up into a wider range of gas bubble sizes. This is called “churn flow”, a
highly turbulent flow pattern associated with oscillating liquid flows. This trend may
continue so that the phases become dispersed within one another, i.e. neither is
continuous. This has been called “froth flow” (not illustrated in Figure 6).
(vi) Further upward fluid flow continues the gas liberation and expansion processes
so that the phases separate into a central, high velocity core of gas with a continuous
film of liquid on the tubing wall - the “Annular flow” regime.
(vii) Shear at the gas/liquid interface resulting from continually increasing gas
velocities will eventually destroy the annular ring of liquid on the tubing wall and
disperse it as a “mist” of small droplets - the “Mist flow” regime.
NB. The high velocities experienced near the surface can result in the increase in the
frictional pressure gradient exceeding the decrease in the hydrostatic head pressure
gradient, so that the pressure in the tubing may increase as the depth (and rate)
decreases (e.g. Figure 13c).
Dimensionless Liquid Velocity
These flow regime transitions have been studied both theoretically and experimentally
(by visually observing the flow regime as a function of gas and liquid velocity in a
transparent, vertical column). Correlations are generated so that the boundaries
between the various flow regimes can be plotted on a Flow Map (Figure 22).
Bubble
Mist
Churn
Slug
Annular
Dimensionless Gas Velocity
Figure 22
Example flow map
Institute of Petroleum Engineering, Heriot-Watt University
27
1
(B) FLOW IN INCLINED TUBING
The complex description of gas/liquid multiphase flow in vertical pipes is simplified
by the fact that the low density (gas) phase is tending to rise (due to its low density)
in the same direction as the overall flow. This is not the case for inclined or horizontal
flow - under these conditions it is much easier for the gas to separate from the liquid
and for the difference between the actual and superficial phase velocities to become
much greater than for the corresponding vertical flow conditions. This naturally alters
the flow regime as the pipe’s angle of inclination (θ) increases from the vertical. A
second effect is that the tubing length (L) becomes greater than H (the vertical depth)
as θ increases.
L = H/cosθ
The hydrostatic head component of the total downhole will also tend to increase with
increasing deviation angle, θ, since, under most conditions, the average fluid density
will increase due to an increase in the liquid hold up (HL).
(C) FLOW IN HORIZONTAL WELLS AND FLOW LINES (Figure 23)
Smooth
Stratified
Wavy
Annular
Intermittent
Elongated
Bubble
Slug
Bubble
Distributed
Mist
28
Figure 23
Horizontal pipe liquid / gas
flow patterns
1
Well Performance
All the above trends (observed in inclined tubings) become progressively more
extreme as the angle θ increases to 90º - a horizontal flow line. Here the hydrostatic
head component is of minor importance while the tendency for phase separation due
to density difference is at its greatest. Experiments have been carried out in
transparent pipes and have identified the following regimes.
Stratified
Intermittent
Annular
- smooth
- wavy
- elongated bubble
- slug
- annular mist
- annular wavy
Dispersed bubble
increasing
velocity
increasing
liquid
fraction
Large charges are observed in the flow pattern when the pipe inclination angle
changes from +1˚ to -1˚ under stratified or (relatively) low velocity flow conditions.
This can be particularly important when considering the flow regimes in horizontal
wells (which are never exactly horizontal and whose liner/casing has a greater
diameter than normal production tubings).
The above, empirical description of multiphase flow has discussed the phenomena
identified by experimental studies. The object of these studies was:
(i) to produce a flow map which delineated the boundaries between the different
flow regimes and
(ii) to develop a correlation between pressure drop and liquid and gas phase
properties and velocities, as a function of tubing diameter, within that flow regime.
The combination of these two factors allow the calculation of the tubing outflow
performance.
Different investigators have published flow maps (which are associated with their
names) and experimental correlations which, unfortunately, can be contradictory i.e.
under certain flow conditions they predict different flow regimes (and pressure
drops). Flow maps and correlations (along with recommendations) will be discussed
in greater detail in chapter 1.7. These pressure drop calculations were initially
implemented using hand calculation procedures. Nowadays, several (commercial)
computer programs are available to rapidly and easily complete these complicated
calculations (chapter 1.11). Prior to the widespread availability of computers or
electronic calculators, use was made of “gradient” curves which greatly simplified the
calculation process.
1.6 “Gradient” or Pressure Traverse Curves
Gradient curves were originally proposed by Gilbert (“Flowing and Gas Lift Well
Performance”, Drilling and Production Practice, API, 1954). Gradient curves
correlate pressure drop as a function of tubing length (Figure 24). Field experience
lead Gilbert to identify that the main factors in controlling vertical multiphase flow
Institute of Petroleum Engineering, Heriot-Watt University
29
1
were tubing diameter, oil rate and gas/liquid ratio. His curves were developed using
field data. However, later curves published by other investigators are based on
laboratory experimental data and flow maps.
00
4
8
12 16
Pressure, 100 PSIG
20 24 28 32 36
40 44
48 52
56
1
Tubing Size, In. : 1.995
2
Liquid Rate, STBL/D : 500
3
Water Fraction
4
:
0
5
6
id
qu
Li
8
tio
Ra
Depth, 1000 Ft.
as
G
7
9
10
11
12
13
14
0
0
10
15
0
20 0
30 0
40
16
17
0
50 0
60
0
80
00
10
00
12
00
20
15
19
0
0
200 300
18
Figure 24 shows that a different gradient curve is required for each tubing size, liquid
rate and water fraction. Each graph consists of a series of lines referring to a different
gas/liquid ratio. Figure 25 explains how the gradient curve can be applied to calculate
the flowing bottom hole (Pwf) or, more precisely, the flowing tubing intake pressure
30
Figure 24
Example "Gradient" or
Pressure Traverse Curve
1
Well Performance
for a fixed wellhead pressure. This is then repeated for a series of production rates to
derive the Tubing Performance Relationship (TPR).
Pressure
q0
1
Pressure
Pwh
Pwf 2
q0
GLR
d2
H
Pwf 3
Pwh
q0
2
3
d3
d1
Pressure
Pwh
Depth
Pwf 1
Depth
Depth
Pwh
Depth
Pressure
GLR
q0
4
d4
GLR
Pwf 4
GLR
H
H
H
Figure 25
Contruction of the tubing
performance relation (TPR)
using gradient curves
Required Tubing
Intake Pressure
Pwf 4
Pwf 1
Pwf 2
q0
1
q0
2
Pwf 3
q0
3
q0
4
Oil Flow Rate, q0
The gradient curve is used as follows:
(i) Select the gradient curve appropriate for the specified oil rate (qo1), tubing size,
gas/liquid ratio and water fraction.
(ii) find the point on the x axis at which the pressure equals the wellhead pressure.
Move vertically downwards to find the depth (d1) on the appropriate gas/liquid ratio
line that corresponds to this wellhead pressure.
(iii) Move downwards by a distance (H), equivalent to the tubing length.
(iv) Moving horizontally and then vertically, identify the pressure on the same gas/
liquid ratio line as was used in (ii) corresponding to this new depth (d1 + H). This
is the required tubing intake pressure (Pwf1).
The process may now be repeated at other oil rates (qo2, qo3 and qo4). Each rate requires
use of a different gradient curve appropriate to these higher rates (tubing size, water
fraction and gas/liquid ratio are constants!). The tubing intake pressures (Pwf2, Pwf3,
and Pwf4) may now be plotted as a function of oil rate(qo1, qo2, qo3 and qo4).
This is the OUTFLOW curve specific to the set of conditions that were used to
generate it (wellhead pressure, tubing size, liquid rate). It will be combined in chapter
1.12 with the Inflow Performance Relationship to estimate the well production rate.
Institute of Petroleum Engineering, Heriot-Watt University
31
1
However, first we need to amplify our earlier discussion or flow maps (chapter 1.7)
then look at how they can be implemented in well performance computer programs
(chapter 1.11).
1.7 Flow Maps and Correlations
Many authors have studied the phenomena of 2- and 3-phase flow in vertical, inclined
and horizontal pipes. They have proposed a large number of flow maps and
correlations based on the large (field and laboratory generated) databases generated
during their studies e.g. the Duns and Ros study contained 4000 separate laboratory
data sets. In this case the database analysis procedure consisted of:
• Expressing the data (e.g. gas and liquid velocities, pipe diameter and liquid
viscosity) in a dimensionless form.
• Using these dimensionless numbers to draw a flow map indicating the boundaries
between the various flow regimes (some studies concentrated on only one flow
regime) equations were developed to describe these boundary lines.
Dimensionless Liquid Velocity Number N
LV
• Correlations for each flow regime allow the calculation of slippage, hold up and
friction factor and hence pressure drop in the pipe.
Region 1
10
Froth
Flow
Region 2
Region 3
1
Bubble
Flow
10-1
10-1
Mist
Flow
Slug
Flow
Plug
Flow
1
102
10
103
Dimensionless Gas Velocity Number N
GV
Alternatively, the models can be based on a mechanistic description of the underlying
fluid mechanics. Typical examples of two types of flow maps - one for vertical flow
based on experiments (Duns and Ros, “Vertical Flow of Gas and Liquid Mixtures in
Wells” Proc Sixth World Petroleum Congress, Vol 2, paper 22, 1963) and the other
from a theoretical analysis of horizontal flow regimes (Taitel and Dukler, 1976) are
presented as figs 26 and 27 Many of the methods were subsequently modified to
extend their range of application - an example being Figure 27b (Taitel, Barnea and
Dukler, “Modelling Flow Pattern Transitions for Steady Upward Gas-Liquid Flow in
Vertical Tubes”, AIChE, J, 26, 345-354, May 1980) which extends the horizontal flow
work summarised in Figure 27a.
32
Figure 26
Duns and Ros flow pattern
map
1
Well Performance
We discussed earlier that many different studies have been published and that none
of them are universally applicable - some being more limited than others. Application
to oil and gas production requires that they should include:
75.0
Bubbly
10.0
Superficial Liquid Velocity (ft/sec)
Intermittent
Annular
1.00
0.10
Stratified Smooth
Stratified Wavy
0.01
0.1
1
10.0
100.0
Figure 27 (a)
Taitel-Dukler horizontal
flow map
900.0
Superficial Liquid Velocity (ft/sec)
Superficial Gas Velocity (ft/sec)
30
Dispersed Bubble
3
Barnea
Transition
Bubbly
Annular
0.3
Slug or Churn
0.03
Figure 27 (b)
Taitel-Barnea-Dukler flow
map
0.001
0.01
0.3
3.0
30
300
Superficial Gas Velocity (ft/sec)
Institute of Petroleum Engineering, Heriot-Watt University
33
1
(i)
Phase Slippage and
(ii)
Flow Regime under
(iii)
Vertical/horizontal and inclined flow condition.
A few of the studies that will be encountered by the practising Production Engineer
are listed in table 4:
Reference
Data Source
Fluids
Comments
Gilbert
Field data
G, O, W
Introduced vertical,
multiphase gradient curves
Duns and Ros
Field and Lab. data
(air, oil & water flow
in 11/4 - 31/8 in. pipes)
G, O, W
Vertical flow over wide flow
rate range
Griffith and Wallis1
Laboratory data
(air & water flow in
narrow pipes)
G, W
Good slug flow correlation
used by later investigators
Hagedoorn and
Brown2
Field experiment
(gas, oil & water
flow in 1 - 4in. pipes)
G, O, W
Aziz and Govier3
Field & Lab. data
(air, oil & water flow
in a wide range of pipes)
G, W
Correlations developed by
mechanistic fluid mechanical
study tested against field data
Beggs and Brill4
Laboratory data
(air & water flow in
1-11/2 in. pipes)
G, W
Correlations useable at all
inclination angles
Forms basis for widely
used correlation
1. Griffith, P. and Wallis, G.B., “Two-Phase Slug Flow,” J. Heat Transfer, Trans.
ASME, Ser. C, 83, 307-320, August 1961.
2. Hagedorn, A.R. and Brown, K.E., “Experimental Study of Pressure Gradients
Occurring During Continuous Two-Phase Flow in Small-Diameter Vertical Conduits,”
JPT, 475-484, April, 1965.
3. Govier, G.W. and Aziz, K., The Flow of Complex Mixtures in Pipes, Robert E.
Drieger Publishing Co., Huntington, NY, 1977.
4. Beggs, H.D. and Brill, J.P., “A Study of Two-Phase Flow in Inclined Pipes,” JPT,
607-617, May 1973.
Complex calculation procedures are required to calculate the tubing performance
relationships based on the multiphase flow correlation methods developed by the
various investigators. This will not be discussed in detail here, but can be found in the
34
Table 4
Flow Correlations
1
Well Performance
original papers and text books. The practising petroleum engineer will most often
obtain access to these techniques in two manners:
(i) Use of gradient curves e.g. Figure 24 was calculated using the Hagedoorn and
Brown correlation
(ii) Direct application of the correlations calculation procedure in one of the many
(commercial) computerised well performance prediction packages available on the
open market (chapter 1.11).
Hence it is sufficient for our purposes to mention just a few of the key points for four
of the main correlations.
1.7.1 Duns and Ros
Duns and Ros defined a flow map (Figure 26) together with a series of correlations for
calculating the boundaries between the flow regimes as well as the slip velocity (Vs).
The Friction factor is calculated from the liquid Reynolds Number when flow is in the
Bubble or Slug regions; while the gas Reynolds number is used in the Mist region.
Finally, calculation of the pressure drop is completed by adding an acceleration term
for flow in the Mist region only.
Many well flow simulation computer programs include modifications of the original
Duns and Ros correlation. These include some or all of:
(i)
use of a different flow map (by Gould et al)
(ii) addition of the Beggs and Brill correction to modify the hold up correlation to
allow for well deviation
(iii) use a modified friction factor (Kleyweg et al, “Gas Lift Optimisation in the
Claymore Field,” Offshore Europe Conference, 1983).
1.7.2 Hagedorn and Brown
Hagedorn and Brown developed a simple flow map and a liquid hold up correction for
the slug flow regime; the Griffith correlation being used for the bubble flow regime.
The friction factor is calculated using a two-phase Reynolds number.
This original work has been modified to include points (ii) and (iii) discussed above
- viz. the “Beggs and Brill” correction to modify the hold up for all angles of deviation
and the Kleyweg friction factor for single phase flow.
The modified Hagedorn and Brown correlation is probably the most widely used
correlation for well performance calculations.
1.7.3 Beggs and Brill
The “Beggs and Brill” method is based on a study of the flow regimes that occur in
horizontal pipes. The flow regime and hold up are calculated as though the pipe was
horizontal and a correction made to account for the change in hold up due to the angle
of inclination when the pipe is not perfectly horizontal. (NB. The flow regime
Institute of Petroleum Engineering, Heriot-Watt University
35
1
calculated is the one that would have occurred if the pipe was horizontal). The
Kleyweg single phase friction factor approach can also be used.
Given the above, it is probably not surprising that this is often the preferred correlation
for simulating the flow in the horizontal and highly deviated portions of wells as well
as in flowlines and pipelines. The correlation is particularly suitable for simulating
pipelines in hilly terrain since it can cope with both upward and downward flow.
1.7.4 Gray
The Gray correlation was specifically developed for gas wells producing small
amounts of liquid - either water or condensate. Experience has shown that this is
normally the best correlation for these conditions.
1.8 Temperature Modelling
ugh
Thro
sfer
n
a
r
eat T
s
Wall
Fluid and Heat Out
POut,TOut
H
,T Avg
P Avg
Figure 28
Temperature modelling /
Calculation proceedure
Angle
PIn,TIn
Fluid and Heat In
Surface
Fluid Temperature Profile (q1)
Fluid Temperature Profile (q2)
>q1
q2
al
m
t
en
di
ra
G
Depth
er
th
eo
G
Temperature
The law of conservation of energy dictates that all enthalpy changes e.g. phase
changes driven by pressure changes, work done overcoming frictional forces etc. are
reflected by a corresponding temperature change. Further, large scale heat loss from
the (hot) produced fluid produced from the (hot) reservoir will occur as it flows
upwards to the (cool) surface. Figure 28 and 29 illustrate the calculation procedure and
36
Figure 29
The average tubing
temperature increases as
the production rate
increases
1
Well Performance
compares the formation (geothermal) temperature with the fluid temperature during
production. In general, the higher the production rate, the hotter the fluid will be at any
given depth (since the increase in the (rate of) supply of energy (heat) is proportional
to the production increase while the heat losses from the wellbore by thermal
conductivity etc. are a only function of the temperature difference between the well
and the surroundings i.e. independent of the production rate.
This temperature change will effect the average fluid properties - which in turn will
alter the pressure drop calculation (and hence the temperature change). A full
simulation of flow in a well thus requires a coupling of the fluid temperature
prediction model with the pressure calculation. This temperature model may range
from a simple analytical equation to a rigorous numerical description of heat flows.
The coupling of temperature and pressure requires an iterative procedure for their
calculation. Figure 30 charts the pressure (inner) and temperature (outer) loops.
START
Given P1,T1,H1,DL,q
Estimate ∆T & ∆P
Outer Loop
T=T1+∆T/2
Calculate ∆HEST
P2 = P1+∆PEST
T2 = T1+∆T
P=P1+∆P/2
Calculate H2
∆H = H2_H1
Inner Loop
∆P = ∆PEST
NO
Figure 30
Pressure and temperature
calculation
Calculate Fluid Props.
Calculate ∆PEST
| ∆PEST _ ∆P |<ε∆P
| ∆HEST _ ∆H |< ε∆H
YES
NO
∆Tnew = ∆Told.∆HEST/∆Hold
∆P = ∆PEST
YES
STOP
ε∆P and ε∆H are the allowable differences in the calculations between successive iterations
1.9 SURFACE PRESSURE LOSSES
1.9.1 Surface Components
The principal surface system pressure loss is often the surface choke. This is an
“optional” pressure loss in the sense that it is designed into the well completion in
order to control the well flow rate and the pressures to which the surface equipment
is exposed. The choke can be eliminated completely when the wellhead pressure has
been depleted to such an extent that economic flow rates can only be achieved by
lowering the wellhead pressure to its minimum value.
A second source of pressure losses in the surface system is the flow line. It should be
remembered that flow line pressure losses are not only related to the length, diameter
and wall roughness of this pipe; but that additional pressure losses will occur in pipe
fittings (T-pieces, elbows, etc) and valves. These additional pressure losses are
Institute of Petroleum Engineering, Heriot-Watt University
37
1
accounted for as an increase in the effective length of the pipeline. These increases
can be quite substantial e.g. the pressure losses in some types of valve - especially
when only partly open - can be up to several hundred times the pipe inner diameter.
Excessive flow line pressure losses can be reduced by installing a parallel, or looped,
flow line in order to reduce the fluid’s flowing velocity. (This is an option not available
within the well!) Modeling of looped pipelines is relatively simple for single-phase
flow; since the flow will divide itself between the two branches so that there is an equal
pressure drop along the two pipelines. The looped pipeline constructed from individual
flow lines of diameters d1 and d2 will behave in a similar manner to a single pipe with
an effective diameter of (d12.5 + d22.5)0.4 (this assumes that the same fluid properties and
friction factors apply to both branches).
The above concept is not appropriate for multiphase flow since the liquid and gas
phases are unlikely to split equally between the two branches - often most of the liquid
will go into one branch while most of the gas will be diverted to the second one. The
mass split ratio between the two lines is difficult to predict - it will depend on the exact
arrangement of the T piece and the Reynolds number associated with the flow in each
line. Thus a lower flow velocity and a T junction design where the loop flow lines are
not at the same elevation, will result in most of the denser, liquid flow going into the
downward pointing line.
1.9.2 Flow Through Chokes
Flow from a well often has to be controlled for reasons such as:
(i) limitation of the drawdown to prevent water coning, gas cusping or sand
production;
(ii) dissipation of well energy to meet pressure limitations of the downstream surface
production equipment etc;
(iii) control of well production rates to meet regulatory, reservoir management or
production equipment constraints.
Chokes differ from other completion equipment in that they are designed to produce
a pressure loss while other components, such as subsurface safety valves, are designed
so that their presence has a minimal effect on the total system pressure losses.
Commonly employed chokes (Figure 31) disturb the fluid flow pattern by use of a
fixed bean, an adjustable rod (which (partly) blocks an orifice) and a rotating disc.
Chokes achieve the desired pressure loss by restricting the flow diameter and
acceleration of the flowing fluid. The phenomenon of critical flow occurs once this
acceleration in the throat of the choke is sufficient that the flowing fluid’s sonic
velocity is exceeded.
38
1
Well Performance
Replacable Fixed
Restriction or Bean
Direction of Flow
Fixed Choke
Direction of Flow
Adjustable
Rod
Adjustable
Rod
Direction of Flow
Direction of Flow
Replaceable
Orifice
Replaceable
Orifice
Adjustable Choke
Fully Open
Choked Flow
Rotating Discs
Fully Opened
Figure 31
Example of surface choke
designs
Throttling
Fully Closed
Rotating Disc Choke
Critical flow prevents a pressure disturbance downstream of the choke from being
propagated upstream, since a pressure wave can not travel faster than the speed of
sound. The well’s performance (upstream of the choke) can thus be decoupled from
events occurring in the downstream flow line and separation system. This has obvious
advantages when trying to control the well’s performance.
Institute of Petroleum Engineering, Heriot-Watt University
39
1
Eddy Currents
lead to irreversible
pressure losses
Upstream
Pressure
(Pu)
D2
Downstream
Pressure
(Pd)
D1
Gas
Flow
Rate
(Q)
Flow
Critical
Pu
w
2.0
ic
al
F lo
Pressure ratio (Pu / Pd)
Abrubt Restriction
Vena Contracta
Abrubt Enlargement
Leads to
Leads to Low Velocity Flow
High Velocity Flow and Low Pressure
and (incomplete) Pressure Recovery
1.0
Sub
-C
rit
Mass Flow Rate of Gas
Figure 32 sketches the fluid flow pattern through the choke and the resulting flow
behaviour. The top part of the figure shows how the choke represents an abrupt
restriction in the fluid flow in the pipe. This restriction results in an area of high
velocity and decreased pressure in the centre of the choke. This is known as the “Vena
Contracta”. As shown, it forces the liquid to flow through an even smaller diameter
than that of inner diameter of the choke. The fluid flow expands again to its original
diameter at the (abrupt) end of the choke. The decrease in velocity results in recovery
of (some) of the pressure that had been lost during passage through the choke. Full
pressure recovery is not normally experienced since there are irreversible pressure
losses due to eddy currents which create disengagement and reattachment of the
flowlines to the pipeline wall.
The bottom section of this figure shows how the flow rate through the choke is related
to the ratio: {(upstream pressure (Pu) / downstream pressure (Pd)}. For sub critical flow
conditions, the flow rate will increase with decreasing downstream pressure {or
increasing (Pu/Pd) ratio} until this ratio is sufficiently large that critical flow occurs.
The ratio (Pu/Pd) normally has a value of the about 2 at this point - the exact figure will
depend on the properties of the flowing fluid, as discussed below - with critical flow
continues to occur for all higher (Pu/Pd) ratios.
40
Figure 32
Critical and sub-critical
flow through a choke
1
Well Performance
The mass flow rate under critical flow conditions is independent of the upstream
pressure for incompressible (liquid) flow. By contrast, it will depend on the upstream
pressure when a compressible fluid (gas) is flowing. Single-phase critical flow
typically occurs when the (P /P ) ratio is greater than 1.5. Critical flow in multi-phase
u d
mixtures requires a somewhat greater pressure ratio; (P /P ) typically having a value
u d
of between 2.0 and 3.0.
1.9.2.1 Single Phase Subcritical Liquid Flow
Single phase, liquid flow is described by:
Q = C * CD * D2
PU − PD
ρ
where:
PU = Upstream Pressure
PD = Downstream Pressure
D2 = Choke Diameter
CD = the flow discharge coefficient through the choke,
C = a constant depending on the units employed and
ρ = the density.
The choke manufacturer normally supplies a choke performance chart or correlation
that relates the discharge coefficient (CD) to the diameter of the choke (D2) and the
Reynolds Number.
1.9.2.2 Single Phase Gas Flow
It is relatively simple to derive equations describing the isentropic flow of an ideal gas
through a choke. These can be found in the standard text books on the subject. It can
be shown that:
 γ +1 
 PU   γ + 1  γ 
  =
2 
 PD c
where:
{Pu/Pd}c = the ratio of the up and downstream pressures at which critical flow occurs,
and
γ = the ratio of the gas heat capacities at constant pressure and constant volume
{or (Cp/Cv)}.
γ has a value of approximately 1.4 for diatomic gasses such as air. Hence the critical
pressure ratio is 1.89, confirming the values quoted above and illustrated in Figure 32.
1.9.2.3 Multiphase Flow Critical Flow Rate
Multiphase (gas-liquid) flow is not easily described theoretically - empirical correlations
have been developed by a number of investigators which are all of the form:
Pu = b * Q L * R
Da 64
c
Institute of Petroleum Engineering, Heriot-Watt University
41
1
where:
Pu = the upstream pressure (psig, except for Ros who uses the unit psia)
QL = the liquid critical flow rate (Stb/d)
D64 = the choke diameter (64th of an inch)
R = the gas / liquid ratio (scf/STB) and
a,b, & c = are constants given in Table 5.
Correlation
a
b
c
Ros
Gilbert
Achong
Ausseens
Baxendell
2.00
1.89
1.88
1.97
1.93
17.40
10.00
3.82
3.86
9.56
0.500
0.546
0.650
0.680
0.546
Table 5
Flow through chokes empirical choke
correlations
The effect of changing the choke size on well production and flow system pressure
losses can be studied using the Nodal analysis technique (see section 1.12). The
upstream side of the choke is normally chosen as the Node. The Wellhead
Performance (or combination of the well’s Inflow Performance and Tubing
Performance Relationships) is the upstream, inflow component to the node and the
choke, flowlines and separator are the downsteam, outflow component. A typical
result from such a calculation is shown in Figure 33. This shows an operating point
of 460 b/d for a 16/64 in choke increasing to 1370 b/d for a 40/64 in. choke.
Choke size
16/64 in.
Node pressure (psi)
1200
24/64 in.
800
Well head
performance
32/64 in.
400
40/64 in.
0
0
sub critical flow
750
Well production (bfd)
1500
1.9.3 Gathering System Layout
The layout of the surface facilities and flow lines is important for land fields where,
typically (near) vertical wells are drilled at a relatively small interwell spacing. The
wellhead locations thus reflect the subsurface locations - the resulting grid pattern
being illustrated in Figure 34 and 35. Connection of each wellhead directly to a local
gathering stations with primary separation facilities (Figure 34) results in each well
being connected by a short length of flowline directly to the primary separator. This
allows lower wellhead pressures than the alternative (Figure 35) where the wells are
tied into a common pipeline.
42
Figure 33
Choke performance curves
1
Well Performance
Export pipeline
Well head
Gathering station
Main processing facility
Figure 34
Oil field developed with
local gathering station
Trunk line
Well head to gathering
station flow line
g
g
y
y
Export pipeline
Well head
Main processing facility
Figure 35
Oil field developed with
single processing facility
only
Trunk line
Local gathering line to
trunk line
Flowline pressure drops are thus much larger in the Figure 35 case, unless wide
diameter pipes are installed e.g. a 50% increase in flow rate can sometimes lead to a
300% increase in the frictional pressure loss across a section of pipe. The increase in
flow rate in the gathering system as one gets nearer the separator means that wells with
a more direct connection to the separator can be produced at a lower wellhead
pressure. The performance of each well thus has a much greater impact on its
neighbours compared to the installation of a local gathering station (Figure 34).
1.10 Completions Inflow Performance
Well performance prediction programs require that the Inflow Performance Relationship is specified. This is normally in the form of a Productivity Index value. For
existing wells this value can be obtained from analysis of a Well test, Production
Logging Survey etc. A model of the completion, in conjunction with either a “straight
Institute of Petroleum Engineering, Heriot-Watt University
43
1
line PI” or Vogel type inflow relationship, is required in the absence of these field
measurements or when designing a new well.
1.10.1 Perforated Completions
Crushed Zone where
permeability of crushed
sand grains has lower
permeability than the
undamaged formation
Casing
Cement
Perforation
Diameter
Penetration Length
Undamaged formation
Formation
Damage Zone
The Open Perforation model (Figure 36) can be used to predict well performance. The
inflow performance is affected by the:
(i) Perforation length (L) - longer perforations are more productive
(ii) Perforation diameter (Dperf) - wider perforations will show a reduced frictional
pressure loss
(iii) Perforation density (n) - reducing the distance between perforations will
increase the well productivity
(iv) Perforation phasing - reducing the angle between adjacent perforations will
increase the well productivity
(v) Depth and Permeability reduction caused by Formation Damage - formation
Damage has limited effect on well productivity provided it is penetrated by the
perforation.
(vi) Permeability and depth of crushed zone around the perforation - perforation
clean up procedures should be designed to remove this impaired crushed zone prior
to production.
(vii) Formation vertical and horizontal permeability - reduced vertical permeability
impedes well production when the perforations are far apart (low shot densities).
44
Figure 36
Flow through completions perforated completion
model
1
Well Performance
(viii) Drawdown and properties of the produced fluids - high gas and very high oil
flow rates through the perforation lead to extra pressure losses from non-Darcy flow
effects.
The relationship between some of the factors discussed above is illustrated in Figures
37 - 40 in which the productivity of an example completion is compared with that of
the equivalent open hole completion.
1.2
Perforation Density (shots/ft)
16
8 90° Phasing
1.1
1.0
Productivity Ratio
4
16
8 180° Phasing
4
Open Hole
16
8 0° Phasing
4
0.9
0.8
0.7
•Crushed and Formation
Damage zones omitted and
•Turbulence factor not included
0.6
Figure 37
Influence of perforation
density and phasing
0.5
0
5
10
Perforation Penetration Length (in.)
15
Figure 37 compares the effect of perforation density and phasing. For this particular
example, avoiding tortuous flow paths associated with 0º phasing perforations will
force the fluids to flow “round the casing” (Figure 38), has a greater influence on the
well productivity than the perforation density.
Institute of Petroleum Engineering, Heriot-Watt University
45
1
y
Casing
Tortuous fluid flow
path to perforation
Cement
Zero phased
(in - line)
perforation
Figure 38
0° phased perforation
reduce well productivity
Open Hole
1.0
Productivity Ratio
1
0.1 Kv
K
0.01 h
1
0.8
Kv
Kh
0.6
0.1
0.01
0.4
0
4 and
12 Shots/ft. at 120° phasing
5
10
Penetration Length (in.)
15
Figure 39 illustrates how the effect of an unfavourable vertical permeability can be
overcome by placing the perforations closer together. Also, it can be seen how a well
completed in a formation with a vertical permeability similar to the horizontal
permeability can have a productivity approaching that of an open hole completion,
even when there is a low perforation density.
46
Figure 39
Influence of vertical and
horizontal formation
permeability and
perforation density on well
production
1
Well Performance
1.0
Open Hole
0.9
Crushed Zone
Removed or
0.8
K formation
Productivity Ratio
0.7
K crushed zone
0.6
K formation
K crushed zone
0.5
=4
0.4
K formation
0.3
K crushed zone
Figure 40
Crushed zone around
perforation and high rate
gas flow reduce well
productivity
=1
= 10
0.2
Darcy Flow (low rate gas flow)
with Non - Darcy Effect (high rate flow)
4 Shots/Foot at 0° Phasing
0.1
0
0
5
10
Penetration Length (in.)
15
Figure 40 illustrates how the cylindrical crushed zone reduces the well performance.
Further, high velocity (non-Darcy or turbulent) flow effects often associated with
gas wells will be accentuated, further reducing the inflow performance. This
explains why so much care needs to be taken when designing completion procedures,
Modern perforating technology has increased the options available to the completion
engineer e.g.
(i) perforation charges with a 1.37 m penetration depth for the standard API target,
(ii) ability to perforate > 2600 m of casing in one run by simultaneous detonation of
25,000 perforation charges and
(iii) creation of a large inflow area equivalent to 400 cm2/m length for a 9.875 in
casing.
1.10.1.1 Perforation Charge Performance
More quantitative calculations on the parameters that affect the performance of
perforated completions can be made using the correlations supplied in paper SPE
18247, “Semi-analytical Production Models for Perforated Completions” by M.
Karaka and S. Tariq. Experiments under realistic downhole conditions have shown
that the (downhole) performance of (shaped) perforating charges depends on the:
(i)
weight of explosive,
(ii) whether it is designed to produce a wide diameter or deeply penetrating hole,
(iii) type of charge liner,
(iv) perforation gun design and stand-off from the casing wall,
Institute of Petroleum Engineering, Heriot-Watt University
47
1
(v) thickness and type of the casing, rock strength and insitu stresses, formation
pressure etc.
Prediction of the Well’s Inflow Performance requires that this downhole perforator
performance must be estimated through knowledge of the perforation length, width
and properties of the crushed zone (if left in place). Measurements for the above
factors, preferrably made under simulated downhole conditions, are normally supplied by the service company providing the perforating system. These will either be
based on specific physical experiments carried out for the specified well situation or
by use of a computer program containing correlations developed from extensive
experimentation. For example, Schlumberger’s SPAN™ (Schlumberger Perforating
Analysis) software will predict the:
(i)
downhole gun performance,
(ii) optimal under balance required to remove the perforating debris, crushed zone,
etc. and
(iii) well productivity.
The impact on well inflow efficiency for different gun types, perforation charge
designs, perforation phasing can be estimated, compared and an optimum selection
made.
The standard against which the productivity of the different perforating systems are
compared needs to be evaluated carefully. One option, preferred by the author and
used here, is the productivity equivalent to that of the open hole originally drilled in
the absence of formation damage.
1.10.1.2 Perforation Gun Selection
So far we have discussed perforating system choice in terms of well productivity only.
There are many other practical considerations to be borne in mind when selecting a
perforating system. These include:
(i) compatibility of the physical dimensions of the perforating gun and the completion.
(ii) the completion technique e.g. perforating prior to or after running tubing.
Selection of “Through Tubing”, Casing or Tubing Conveyed Perforating guns.
(iii) casing damage/perforating debris.
(iv) management of sand production. The phasing, orientation and design of the
perforating pattern can impact on the severity of sand related production problems
when the well is placed on production.
(a) A change in the perforation phasing from 60° to 90° in BP’s Magnus Field, while
maintaining the same perforation density and charge type, gave a substantial
reduction in production problems attributed to sand production. This was
ascribed to a new gun design that gave a 56% increase in the minimum spacing
48
1
Well Performance
between perforations. This improvement is based on the concept that sand
production problems are accentuated by the failure of individual perforations to
the extent that they grow together and form a (relatively) large cavity. Maximising
this separation minimises the chance of perforation collapse and amalgamation
without compromising the well’s inflow performance.
(b) Sand “run in” into the casing between the perforating operation and placing of
the gravel pack screen can be an operational problem when installing a gravel
pack completion in deviated wells in some unconsolidated formations. Field
experience has shown that this “run in” can be minimised by omitting the
upward facing perforations from the perforation gun (this requires that the
perforation gun has to be orientated before firing). This improvement was
attributed to the ease with which the unconsolidated sand fell vertically downwards (under the influence of gravity) during (weak) pressure surges
caused by running the completion equipment.
(c) Experience has shown that long term sand production can be minimised by
orientating the perforations in the direction of the maximum insitu rock stress.
1.10.2 Gravel Packed Completions
Perforating strategy in natural flow completions i.e. those which require neither sand
control nor hydraulic fracture stimulation, is aimed at delivering sufficient perforations
open to flow so that the overall well productivity is not reduced by the presence of the
perforations. In the previous section we discussed underbalance perforating techniques
used to reduce formation damage by cleaning out the perforation change debris and
formation crushed zone. This formation / perforation damage removal process may
continue when the well is placed on production since there are open perforations
through which the debris can flow into the well. However, the installation of a gravel
pack traps any remaining debris in the perforation tunnel behind a sheath of gravel.
Gravel packed well completion strategy has the same objective to that for natural flow
completions - ensuring that the perforations do not limit well production in any way.
The restriction on the ability of the flow from the well to remove damage in the longer
term implies that the perforating process has to be designed to minimise the damage
creation. In addition, there is the new factor of minimising formation damage from
the gravel packing operation itself (Figure 41a).
N.B. Production zones to be gravel packed often have a high permeability and a
reservoir pressure depleted below the hydrostatic value. They are particularly prone
to formation damage due to fluid loss and/or exposure to Lost Circulation Material.
Institute of Petroleum Engineering, Heriot-Watt University
49
1
Casing
Formation damage from drilling
Perforated tubing
Wire wrapped
sand screen
Gravel
Well center line
Cement
Undamaged formation
Formation damage
from the perforating
process, perforation
gun debris and
gravel packing
Tunnel
length
Formation
Perforation Diameter
Radius Bit
Flow
Figure 41 a
Flow through a gravel pack
completion in a semi
competent sand
Weak sands are often unable to support a defined perforation tunnel. The resulting
gravel packed completion is then best represented by Figure 41b. In practice, it is often
unclear which of the above models is most suited to a particular situation. Further, the
depth and extent of permeability damage due to fluid loss etc. is normally not known.
Casing
Formation damage from drilling
Perforated tubing
Wire wrapped
sand screen
Gravel
Well center line
Cement
Formation damage
from gravel packing
Tunnel
length
Undamaged
Formation
Perforation Diameter
Radius Bit
Flow
One practical solution to the above is to use a composite model of the inflow process
similar to that sketched in Figure 42. This combines a gravel filled tunnel of length
equivalent to the distance from the wire wrapped screen to the edge of the cement
sheath together with the porous media, radial inflow equation.
50
Figure 41b
Flow through gravel pack
completion in a weak sand
1
Well Performance
Skin? Casing Skin?
Cement
Porous media
radial inflow
equation
Perforated tubing
Figure 42
Flow through completions a composite gravel pack
completion models
Well center line
Wire wrapped
sand screen
Perforation Diameter
Tunnel
length
Undamaged
Formation
Gravel
Radius Bit
Flow
Impairment to the gravel pack sand can then be represented:
(i) by reducing the number of perforations that are open to flow,
(ii) as a “skin” value at the wire wrapped screen or formation interface or
(iii) by the gravel permeability being reduced below its unimpaired value e.g. 20/40
US Mesh gravel typically has an unimpaired permeability of about 180D.
Formation damage can be simulated by use of a Vogel “Flow Efficiency” factor less
than 100% or by inclusion of a “skin” between the gravel and the formation.
N.B. This skin would have a negative value if it was believed that unimpaired gravel
had been displaced past the cement sheath and that permeability damage to the
formation was absent!
This model can be used to design the gravel pack completion by ensuring that the
pressure drop across the gravel filled tunnel has a minimal impact on the well
production.
(i) As large as possible gravel pack sand size is selected that is capable of controlling
the formation sand and a choice made as how to represent gravel pack sand
impairment (discussed above).
(ii) The perforation density and diameter are varied until an acceptable pressure drop
is predicted. The relationship between these two parameters is discussed in chapter
1.12.5 on completion design.
N.B. Remember that the perforations are now filled with gravel. Also, that gravel
filled perforations will exhibit extra turbulent flow induced pressure losses (nonDarcy flow effects) at much lower flow velocities than open perforations.
1.10.2.1 Non-Darcy Turbulence Pressure Losses
The Darcy contribution to the pressure drop through the perforations can be represented
by a gravel-pack skin factor, sg, and the non-Darcy flow coefficient for the gravelfilled perforation , Dg. The latter term is normally calculated using the Forcheimer
Institute of Petroleum Engineering, Heriot-Watt University
51
1
equation. Different equations are derived for gas (Dgg) and oil (Dgo) wells. Golan and
Whitson (see reference 4, section 1.13) quote the following equations for the
calculation of pressure losses across the gravel-filled perforations of an inside-casing
gravel packs:
Sg =
96 * kh * L
and
k g * d 2 perf * n
For gas wells:
D gg =
2.45 * 10 −10 * γ g * kh * L * β g
µ * d 4 perf * n 2
For oil wells:
D go =
1.8 * 10 −11 * Bo * kh * L * β g
µ * d 4 perf * n 2
In these equations:
kh = the formation permeability-thickness product (md:ft),
L = the gravel-packed tunnel length (in),
kg = the permeability of the gravel (md),
dperf = the perforation diameter (in),
γ = the gas relative gravity,
µ = viscosity (cp),
n = the number of perforations and
βg = the gravel turbulence factor
Βο = the oil volume factor (bbl/stb).
These terms must be added to other sources of “skin” in the conventional radial inflow
equation e.g. for outflow:
Productivity Index =
Q
P reservoir − P sandface
=
kh
141.2 * Bo * µ *{In( re / rw ) + S + Sg + D g * Q}
Where S represents all skin factors apart from that due to the gravel pack and Q is the
flow rate. The term Dg*Q is often referred to as the rate dependent skin.
The turbulence factor (βg) is a rock property of the gravel pack sand - its numerical
value is related to the permeability and sand grain size. Formation damage, which
reduces the (Darcy) permeability of the gravel pack sand, will also increase the value
of βg; leading to even greater pressure losses. The turbulence factor is correlated with
the gravel permeability (Cooke, “Conductivity of Fracture Proppants in Multiple
layers”, J.P.T., 1101-1104, September 1973) as:
β g = bk g − a
Values for the constants a and b for common gravel sizes are suggested in Table 6.
52
1
Well Performance
U.S. Mesh Average
Size
diameter
(in)
40/60
0.014
20/40
0.025
10/20
0.056
Table 6
Gravel pack sand properties
Permeabiliy
(md)
1.2 x 105
1.8 x 105
5.6 x 105
Turbulence factor
βg = bkg -a
a
b
1.6
1.54
1.34
2.12 x 1012
3.37 x 1012
8.4 x 1011
1.10.2.2 Restriction of Gravel Pack Drawdown
Some operators restrict the pressure drop allowed across the gravel pack completion.
This arose because they observed an increased sand control failure rate when
producing at a high well drawdown - a figure of 300 psi has been used in the USA Gulf
Coast fields (though other operators have had good experience at much higher
drawdowns). The impact on the production rate of varying this allowable drawdown
can be evaluated by carrying out a nodal analysis calculation which places the node
at the sand face and omits the presence of the gravel pack i.e.
Inflow
Average reservoir pressure - drawdown across formation - sandface pressure
Outflow
Sandface pressure = pressure drop across (tubing + flowline) + separator pressure
Average reservoir pressure
Inflo
Figure 43
Gravel pack pressure
analysis
Sand Face Pressure
w
∆Pgp1
∆Pgp2
∆Pgp3
∆Pgp=0
∆Pgp4
Out
Q1
Q2
Q3
Q4
flow
Q max
Production Rate
The intersection point (Figure 43) represents the maximum flow rate (Qmax) when the
pressure drop is often across the gravel pack (∆Pgp) is zero. In practice, the pressure
drop across the gravel pack is often greater than zero - it can be estimated from the
difference between the inflow and outflow curves at the production rate actually
achieved by the well. This value can be compared with that predicted for the
unimpaired completion by calculating the pressure losses through the gravel filled
tunnels, as described above.
1.11 Computerised Well Performance Prediction Programs
The basic concepts required for the calculation of the pressure difference between two
points that are connected by a pipe have been discussed in the preceding chapters.
Section 1.6 introduced the graphical pressure traverse method for estimating pressure
drops; while the subsequent chapters detailed the requirements for a numerical
description of these calculations. The complicated nature of these calculations
Institute of Petroleum Engineering, Heriot-Watt University
53
1
resulting from a full description of the flow processes require a computer to carry out
the calculations in a reasonable period of time.
The calculation procedure employed consists of splitting the pipe into a number of
segments and calculating the pressure drop across each segment. Figure 44 schematically
illustrates the procedure. Steady state flow requires that mass, momentum and energy
be conserved between the inlet and outlet of a given volume, or segment (figure 44a).
Steady-State Flow
(1) [ Mass flowIn = Mass flowOut ],
(2) [ MomentumIn - MomentumOut ] = Sum of momentum changes ] and
(3) [ EnergyIn + Work + Heat = EnergyOut ]
Out
dL
In
dZ
θ
V
.
dX
Figure 44a
Conservation of mass,
momentum and energy
This conservation of mass, momentum and energy is applied to the pressure drop
calculation procedure across a segment of pipe (figure 44b)
1
VG
2
VL
ρG ρL
L
P1 > P2
L
P1 > P2 = ∆P1-2 = ∫ 2 dP(P,T) dL
L1 dL
Figure 44b
Calculation procedure
And is then extended to a series of pipe segments (figure 44c).
1
2
3
i=1
∆P1-2 ≈
Σ
i=n
n
i=1
∆L i
dP
dL
i
Suitable arrangements must be made to ensure that the TOTAL mass, momentum and
energy are conserved when two separate flows join together
54
Figure 44c
Pressure drop calculation
across a series of segments
1
Well Performance
Wellhead
Large Change in
Temperature and Pressure
Surface
Restriction
Sub-surface
Safety Valve
(Normally) Small Changes in
Temperature and Pressure
Segments
q2
q1
Figure 45
Pressure and temperature
calculation
q3
q4
q6
q7
Small Diameter Tubing
Hydrocarbon Reserve
q1-q7 Inflow at Particular Locations
Large Diameter Casing
Figure 45 illustrates the segmentation process extended to simulate a complete well.
It is a “horizontal” well producing with hydrocarbon inflow at 7 specific locations.
The pressure (and temperature) changes along the “horizontal” section are much
smaller than occur in the vertical part of the wellbore. The size of the segment is
related to the magnitude of the pressure or temperature change that occurs across the
segment:
• Larger segments decrease calculation time;
• Small segments maintain accuracy when pressure and temperature are varying
rapidly.
• Fluid properties are calculated at the average of the inlet and out let conditions
(temperature and pressure).
Essentially the same process is followed when the calculations are being performed
by hand or by by use of a computerised simulation program (the size of the segments
is much smaller in the later case!). There are several commercial software packages
which can be used to estimate well performance, e.g Wellflo TM, Prosper TM, Pipesim
TM
etc.
1.12 Well Performance Sensitivity Study Exercise
Early chapters in this module studied the Well Inflow (1.4 and 1.5) and Well Outflow.
Combinations of these two parameters along with the concept of systems analysis of
production systems (nodal analysis) (chapter 1.2) allow us to estimate the well
productivity under today’s actual or future expected producing conditions. The
sensitivity of the well design (or its robustness) to the many factors which effect well
production as the well ages can then be tested.
Institute of Petroleum Engineering, Heriot-Watt University
55
1
I nf
low
To
No
d
Pressure at Node
e
w
tflo
Ou
System Operating
Pressure
System
m
Fro
No
de
Capacity
Flow Rate
Figure 46 illustrates the systems analysis concept. The point (or node) at which the
analysis is carried out can be chosen to be anywhere in the producing system - the
inflow and outflow being calculated for the complete system up- and down-stream
from the chosen node. The sensitivity of the production rate to changes in the
dimensions of a particular component (situated next to the node) can then be
evaluated. This allows the performance of each individual well component to be
isolated in turn. Typical examples of node selection are:
• Wellhead: evaluate the effect of flow line size
• Safety Valve: evaluate the effect of the reduced flow caused by the diameter of the
safety valve being smaller than that of the tubing (important in high rate gas wells)
• Sandface: select the optimum tubing size or evaluate well inflow performance (is
there a requirement for reperforation, stimulation {to remove a positive skin
(acidisation) or create negative skin (hydraulic fracturing)}?
Other possible node points can be seen in Figure 1 which analyses pressure losses in
the complete production system. Some of the more frequently encountered sensitivity
analyses are described below.
1.12.1 Reservoir Inflow and Tubing Outflow Restrictions
The impact of (relatively) inadequate reservoir inflow (case 1) with a (larger than
necessary) tubing is illustrated in Figure 47a. The opposite case, production
restriction by a too small tubing (case 2) is shown in Figure 47b for the same reservoir
inflow performance. It comes as no surprise to see that:
q1 >> q2 and preservoir ≈ psandface (2) >> psandface (1) ≈ pseparator
56
Figure 46
Systems or nodal analysis
1
Well Performance
flow
ir In
rvo
se
Re
Pressure at Sandface
Reservoir Pressure
ing
Tub
1
P1
Separator Pressure
q1
Production Rate
Figure 47a
Reservoir outflow restricts
production
Tu
b
in g
2
Reservoir Pressure
ir
rvo
se
Re low
Inf
Pressure at Sandface
P2
Separator Pressure
Figure 47b
Small tubing restricts
production
q2
Production Rate
1.12.2 Tubing Size and Liquid Loading
The well production will normally increase as the tubing size increases. (The pressure
drop in the tubing decreases so that a greater well drawdown is possible for the same
reservoir and separator pressure). However, at a certain point the upward (gas) flow
velocity has decreased so much (due to the tubing diameter increase) that it is no longer
sufficient to efficiently lift the liquid to the surface i.e. slip phenomena commence and
liquid holdup (or liquid loading) begins (figure 48a).
Institute of Petroleum Engineering, Heriot-Watt University
57
1
g
bin
Tu
w
tflo
Ou
Re
se
rv
oi
rI
nf
lo
w
Pressure at Sandface
Reservoir Pressure
Increasing
Tubing
Diameter
Separator Pressure
Production Rate
Figure 48a
Liquid loading analysis
Production Rate
Maximum Rate
Unstable
Production
Tubing Diameter
Eventually, the increased hydrostatic head (due to the liquid loading) will be greater
than the reduced friction pressure losses as the tubing diameter increases further. This
leads to a maximum production rate (figure 48b) at a certain tubing diameter. Unstable
flow is encountered with even larger tubing diameters - it is not recommended to
operate in this region since liquid loading will eventually progress to the stage that the
well ceases to flow. The underlying cause for the above is a change in flow regimes
as the flow velocity decreases. This allows liquid holdup (slip) to occur, which
becomes progressively more important as velocities decrease further.
58
Figure 48b
Systems analysis for
increasing tubing diameter
1
Well Performance
WC=100%
1.12.3 Effect of Water Cut and Depletion
=2
C
Sand Face Pressure
W
C=
5%
50
%
P1 Reservoir
W
at
W
tF
lo w
ser
v o ir
Ou
Re
g
Tu bin
er
%
=0
C)
(W
t
Cu
I n fl o w
Figure 49a
Effect of wate cut on
production
Production Rate
P1 Reservoir Pressure at t1
P3 Reservoir Pressure at t3
%
=0
C)
(W
Re
ser
v o ir
Sand Face Pressure
P2 Reservoir Pressure at t2
I n fl o
Tubing
Inflow
w
Figure 49b
Effect of depletion on
production rate
25
C=
W
Sand Face Pressure
w
flo
ut
O
%
0
Tu C=
W
g
n
bi
Re
ser
v o ir
I n fl o
w
Figure 49c
Sensitivity of production
rate to pressure depletion
and water cut development
%
WC
=5
0%
q3
q2
q1
Production Rate
qt3
qt2
qt1
Production Rate
Institute of Petroleum Engineering, Heriot-Watt University
59
1
An increasing water cut reduces the gas liquid ratio as well as increasing the
hydrostatic head between the reservoir and the surface. This is illustrated in Figure
49a for a slightly over pressured reservoir. Reservoir simulation can be used to predict
the reservoir pressure depletion with time along with any increase in water cut. Such
a simulation is illustrated in table 7.
Time
t1
t2
t3
Reservoir Pressure
pres1
pres2
pres3
Water Cut
0%
25%
50%
Table 7
Reservoir Simulator
Predictions
The effect of this pressure depletion on the production rate is summarised in Figure
49b and the two are combined in figure 49c. The production rate at time t3 is only 25%
of the initial production, while a small further reduction in reservoir pressure or
increase in water cut beyond 50% will cause the well to cease production altogether.
1.12.4 Opportunities for Skin Removal by Stimulation
Well testing frequently identifies that a positive skin effect is restricting well
production. The economic incentive for removing this skin (or even inducing a
negative skin) can be evaluated with the help of nodal analysis. Figure 50 shows the
current well inflow (skin = +8) together with its partial (skin = +2) and complete (skin
= 0) removal. The carrying out of a hydraulic fracture (skin = -3) is also illustrated.
Tubing 2 Outflow
S ki
Sk
in =
n =-
3
Tubing 1 Outflow
0
Re
+2
=+8
in =
n
Ski
se
rv
oi
rI
nf
lo
w
Pressure at Sandface
Sk
Reservoir Pressure
q1+8
q2,+8 q20
q2+2
q2,-3
q1,+2
q1,0
q1,-3
Separator Pressure
Production Rate
The flat outflow profile of tubing results in large gains in production that might allow
these treatments to be carried out. Tubing 2 (with a more vertical outflow profile) is
already restricting production with the impaired (skin = +8), while only minor
(probably uneconomic) production gains are recorded when the skin is removed - the
60
Figure 50
Opportunities for increased
production by skin removal
1
Well Performance
most favourable production (-3) being still less than that achieved with the larger
tubing and the high (+8) reservoir skin.
Appropriate remedial action can only be taken - and economically justified - when the
pressure losses within the complete well system are understood.
1.12.5 Completion Design
Reservoir Pressure
Pressure at Sandface
N1<N2<N3<N4
In
a
cre
s
N
in g
um
be
w
tflo
Ou
N4D1
N1D1
Figure 51a
Effect of number of
perforations on production
rate
ro
s
ion
rat
rfo
e
fP
N2D1
N3D1
Separator Pressure
Production Rate
The high skin discussed in the previous case could have many causes e.g. formation
damage, partial completion etc. One factor under the control of the production
engineer is the number and type of perforations. Figure 51a illustrates an increase in
the numbers of perforations {N1<N2<N3<N4, all of diameter D1} while Figure 51b
shows that effect of a restricted number of perforations (N1) can be (partially)
compensated for by an increase in the diameter from D1 to D3 (i.e. reduction in the
frictional pressure loss in the perforation tunnel itself). Theoretically increasing the
number of perforations is more beneficial since it improves the inflow from the
reservoir as well as decreasing the (average) frictional pressure drop in the perforation
tunnel (see Figure 51c for comparison). The cost of the perforation operation will
increase as the number and diameter of the perforations are increased - an economic
optimum will be found when both factors are varied simultaneously.
These theoretical calculations are a useful but not complete guide - it is frequently
observed in the field that not all perforations are effective e.g. in gravel packed wells
it is standard practice to assume only 33%–50% of the perforations are effective (i.e.
open to flow).
Institute of Petroleum Engineering, Heriot-Watt University
61
1
Reservoir Pressure
Pressure at Sandface
Re
se
r vo
I
ir I
n fl
ea
n cr
D1<D2<D3
ow
g
sin
ra
rfo
Pe
r
ete
iam
D
n
t io
w
tflo
Ou
g
in
Tub
N1D1
N1D3
N1D2
Separator Pressure
Figure 51b
Effect of perforation
diameter on production rate
Production Rate
Production Rate
Perforations all Diameter D1
N3
N4
N2
D3
D2
N1 Perforations
N1
D1
Number of Perforations (N)
Diameter of Perforations (D)
1.12.6 Well Head Pressure
The separator pressure is often the main component in the surface pressure losses. It
exerts a restrictive “back pressure” on the well production which limits the total
pressure drop available for fluid inflow from the reservoir and onward transportation
to the surface. This effect is illustrated in figure 52 - where the wellhead was chosen
as the node about which the analysis was carried out. Reducing the separator pressure
is often an effective way of increasing the well production.
62
Figure 51c
Optimisation of perforating
schedule
1
Well Performance
Reservoir Pressure
Separator Pressure
500 psi
T u bing O
ut
Flo
w
Flo
Tubing Out
Flo
w
rv o
ir I
O ut
w
se
Tubing
50 psi
Re
Pressure at Wellhead
200 psi
n flo
w
Figure 52
Effects of separator
pressure on production rate
q500 q200 q50
Production Rate
This type of “backpressure” on the wells is often encountered in more subtle ways e.g.
the gas collecting in the tubing/casing annulus of a well equipped with an artificial lift
pump can limit the maximum available drawdown by acting as a back pressure.
Venting this casing gas increases the drawdown with a corresponding production rate
improvement.
Institute of Petroleum Engineering, Heriot-Watt University
63
1
1.13 FURTHER READING
(1) Beggs H. D.
“Production Optimisation using Nodal Analysis”
ISBN 0-930972-14-7
published by Oil and Gas Consultants Inc., 1991.
(2) Economides M., Hill A. & Economides C.
“Petroleum production Systems”
ISBN 0-13-658683-X
published by Prentice Hall, 1994.
(3) Economides M. J., Watters L. and Dunn-Norman S.
“Petroleum Well Construction”
ISBN 0-471-96938-9
Published by Wiley, 1998.
(4) Golan M. & Whitson C.
“Well Performance” 2nd edition
ISBN 0-13-946609-6
published by the Norwegian University of Science and Technology (NTNU), 1996.
(5) Mian M. A.
“Petroleum Engineering Handbook for the Practicing Engineer”, Volume 2
ISBN 0-87814-379-3
Published by PennWell Books, 1992.
64
2
Selection of Artificial Lift Types
CONTENTS
1.
2.
3.
4.
5.
6.
7.
ARTIFICIAL LIFT METHODS:
INTRODUCTION AND SELECTION CRITERIA
THE NEED FOR ARTIFICIAL LIFT
REVIEW OF ARTIFICIAL LIFT TECHNIQUES
CURRENT STATUS OF THE ROLE OF
ARTIFICIAL LIFT IN FIELD DEVELOPMENT
SELECTION OF ARTIFICIAL LIFT CRITERIA
5.1.
Well and Reservoir Characteristics
5.2.
Field Location
5.3.
Operational Problems
5.4.
Economics
5.5.
Implementation of Artificial lift
Selection Techniques
5.6.
Long Term Reservoir Performance and
Facility Constants
ROD PUMPS
6.1.
Introduction
6.2.
The Pumping Unit
6.3.
The Sucker Rods
6.4.
The Pump
6.5.
Rod Pump Operation
6.5.1. Pump-Off Control
6.5.2. Gas Influx
2.6.5.3 Centralisers
2.6.5.4 Solids
2.6.5.5 Pump Diagnosis
2.6.6
Pump Design
2.6.6.1 Pump Rate
2.6.6.2 Rod Stretch and Over Travel
2.6.6.3 Pumping Limit Load Calculations
ELECTRIC SUBMERSIBLE PUMPS (ESPs)
7.1.
Introduction
7.2.
Well Completion Design with ESP's
7.2.1. Applications of ESP's
7.2.2. Horizontal Wells
7.2.3. “Y” Tool
7.3.
Basic Pump Selection
7.4.
Advantages and Disadvantages of
Electric Submersible Pumps
7.5.
Monitoring the Performance of
Electric Submersible Pumps
7.6.
New Technology
7.6.1. Coiled Tubing Deployed ESP's
7.6.2. Auto "Y" Tool
7.6.3. Dual Pump Installations
7.6.4. Reducing Water Production
7.7.
Electric Submersible Pump Performance
7.7.1. Simplified Electric Submersible Pump Design
8. HYDRAULIC PUMPS
8.1.
Advantages of Hydraulic Pumps
8.2.
Disadvantages of Hydraulic Pumps
8.3.
New Technology (Weir Pumps)
9. PROGRESSING CAVITY PUMPS
9.1.
Progressive Cavity Moyon Pump Principle
9.2.
Progressing Cavity Pump Power Supply
9.3.
New Technology
9.3.1. The Progressing Cavity Electric
Submersible Pump (PCESP)
9.3.2. Wireline retrievable PCESP
10. HYBRID SYSTEMS
11. ARTIFICIAL LIFT TUTORIAL
12. FURTHER READING
LEARNING OBJECTIVES:
Having worked through this chapter the student will be able to:
INTRODUCTION
• Explain the importance of Artificial Lift (AL) for world oil production.
• List the different types of AL.
SELECTION
• Select appropriate type AL based on ranking criteria.
ELECTRIC SUBMERSIBLE PUMP
• Identify the components of an Electric Submersible Pump.
• Describe the preferred applications and the mode of operation of the Electric
Submersible Pump (ESP).
• Select well conditions suitable for ESP installation as preferred Artificial Lift option.
• Identify the application areas where an ESP is NOT suitable.
• Evaluate the advantages of an instrumented ESP completion.
BEAM PUMP
• Describe the concept and component parts of a Beam Pump.
• Select well conditions suitable for beam pump installation.
• Explain the beam pump design methodology
• State the background to the use of the Dynanometer card for troubleshooting.
FLUID DRIVEN HYDRAULIC PUMPS
• Describe the concept, implications for the well completion and advantages of using
high pressure fluid as a power source.
• Explain the mode of operation of the:
(i)Jet pump;
(ii)Weir Multiphase pump;
(iii)Hydraulic pump.
• Identify their advantages and disadvantages.
PROGRESSIVE CAVITY PUMP
• Describe the concept and area of application of the Progressive Cavity Pump (PCP).
• Compare rod and electric motor driven PCP’s.
• Discuss advantages of wireline retrievable PCESP.
2
2
Selection of Artificial Lift Types
ARTIFICIAL LIFT METHODS:
1. INTRODUCTION AND SELECTION CRITERIA
This module will introduce the topic of artificial lift - a production engineering topic
of increasing importance in field development. The reasons leading to this increasing
importance in the field development process will be reviewed. The main factors
influencing the selection of the most important artificial lift techniques will be
highlighted.
A brief description will then be given of all the common artificial lift techniques (rod
pumps, electric submersible pumps, progressive cavity pumps and hydraulic pumps)
apart from gas lift.
Hydrocarbons will normally flow to the surface under natural flow when the discovery
well is completed in a virgin reservoir. The fluid production resulting from reservoir
development will normally lead to a reduction in the reservoir pressure, increase in the
fraction of water being produced together with a corresponding decrease in the
produced gas fraction. All these factors reduce, or may even stop, the flow of fluids
from the well. The remedy is to include within the well completion some form of
artificial lift. Artificial lift adds energy to the well fluid which, when added to the
available energy provided “for free” by the reservoir itself, allows the well to flow at
a (hopefully economic) production rate. It has been estimated that in 1994 there was
a world inventory of more than 900,000 producing wells. Only 7% of these flowed
naturally while the remaining 93% required some form of artificial lift. The average
production per well was less than 70 bpd.
2. THE NEED FOR ARTIFICIAL LIFT
Artificial lift is required when a well will no longer flow or when the production rate
is too low to be economic. Figure 1(a) illustrates such a situation - the reservoir
pressure is so low that the static fluid level is below the wellhead. Question: Is it
possible for this well to flow naturally under and conditions.
Zero pressure at wellhead
Q=0
Static Fliud Level
ic
at
St
re
su
es
Pr
Depth
d
iu
Fl
Figure 1
Artificial lift fundamentals
t
en
di
ra
G
Preservoir
Pr reservoir pressure
Preservoir = Pr
1 (a) The well is unable to start flowing flow
naturally due to low reservoir pressure
i
p
Production Rate
Pressure
Figure 1a
The well is unable to
initiatenatural flow.
e
Inflow P
Department of Petroleum Engineering, Heriot-Watt University
rform
anc
eR
t
ela
ion
sh
3
Answer: Yes: If the well productivity Index is sufficiently high and the produced fluid
contains enough gas that the flowing fluid pressure gradient gives a positive wellhead
pressure. But, tthe well has to be "kicked off" (started flowing) by swobbing or other
techniques.
Figure 1(b) shows how installation of a pump a small distance below the static fluid
level allows a limited drawdown (∆p') to be created. The well now starts to flow at rate
q. N.B. the static and flowing pressure gradients in figures 1(a) & 1(b) are similar since
frictional pressure losses in the tubing are small at this low flow rate.
Pressure Ps now measured on
wellhead pressure gauge
Flowing Tubing Head Pressure
(FTHP)
Q=q
Fluid Level in Annulus
∆ P’pump
Depth
Pump
Flow Fluid
Pressure
Gradients at
Rate q
P’flowing Pr
∆p’
P’flowing
Pr reservoir pressure
Pressure
Drawdown
1 (b) Pump creates a small drawdown
and flowrate
Production Rate
P’
q
qmax or Absolute Open Flow (AOF)
It can be readily seen that the same production rate will occur when the pump is
relocated to the bottom of the tubing, provided the pressure drop across the pump, and
hence the drawdown, remains the same. The advantage of placing the pump near the
perforations is that the maximum potential production can now be achieved {figure
1(c)} by imposing a large drawdown (∆P") on the formation and “pumping the well
off” by producing the well at q2 is slightly smaller than the AOF.
4
Figure 1b
Pump creates a small
drawdown and flow rate
2
Selection of Artificial Lift Types
Q = q2
Flowing Tubing Head Pressure Ps
Depth
Flowing Fluid Pressure
Gradients at
Rate q2
P"flowing
Fliud Level in Annulus
Pr reservoir pressure
Drawdown ( ∆P" )
P"flowing
P"flowing
Pr
Pressure
Pump
Production Rate
Figure 1c
Installation of pump
suction below the
perforations maximises
potential drawdown and
production rates
∆ P"pump
Installation of pump suction below
the perforations maximises potential
drawdown and production rates
q2
Artificial lift design requires that the pump to be installed is matched to the well inflow
and outflow performance
3. REVIEW OF ARTIFICIAL LIFT TECHNIQUES
The most popular forms of artificial lift are illustrated in figure 2. They are:
Gas Flow Rate
Control Valve
Production
High
Pressure
Power
Fluid
Gas Flow
Meter
Electric Motor
Production
Production
Production
Rod
Production
Gas Flow
Rod
Power
Fluid
Electric
Power
Cable
Rod
Pump
Gas Lift Valve
(unloading)
Pump
Tubing
Anchor
Fluid
Driven
Pump
Operating Valve
Stator
Motor
Figure 2
The most popular types of
artificial lift
(i) Rod Pump (ii) Hydraulic Pump
(iii) Submersible
Electric
Pump
Department of Petroleum Engineering, Heriot-Watt University
(iv) Gas lift
(v) Progressing Cavity Pump
(May also be driven by
electric submersible motor)
5
(i) Rod Pumps - A downhole plunger is moved up and down by a rod connected
to an engine at the surface. The plunger movement displaces produced fluid
into the tubing via a pump consisting of suitably arranged travelling and
standing valves mounted in a pump barrel.
(ii) Hydraulic Pumps use a high pressure power fluid to:
(a)
drive a downhole turbine pump or
(b)
flow through a venturi or jet, creating a low pressure area which produces
an increased drawdown and inflow from the reservoir.
(iii) Electric Submersible Pump (ESP) employs a downhole centrifugal pump
driven by a three phase, electric motor supplied with electric power via a cable
run from the surface on the outside of the tubing.
(iv) Gas Lift involves the supply of high pressure gas to the casing/tubing annulus
and its injection into the tubing deep in the well. The increased gas content
of the produced fluid reduces the average flowing density of the fluids in the
tubing, hence increasing the formation drawdown and the well inflow rate.
(v) Progressing Cavity Pump (PCP) employs a helical, metal rotor rotating inside
an elastomeric, double helical stator. The rotating action is supplied by
downhole electric motor or by rotating rods.
In fact, nearly all the major classes of pumps are employed in the various forms of
artificial lift (figure 3).
Other
Jet or Venturi Pump
Centrifugal
Dynamic
Rotodynamic
Pumps
Rotary
Positive
Displacement
Reciprocating
Axial
Electric
Submersible
Pump
?
Progressing
Cavity
Pump
Rod Pump
4. CURRENT STATUS OF THE ROLE OF ARTIFICIAL LIFT IN
FIELD DEVELOPMENT
Figure 4 shows relative frequency of the different types of artificial lift installed in the
USA in 1992. The predominance of rod pumps indicates the vast majority of wells
are on land locations in mature fields with low well production.
6
Figure 3
Pump classification
2
Selection of Artificial Lift Types
Hydraulic Pump < 2%
Electric Submersible Pump 4%
Progressing Cavity Pump < 1%
Others < 1%
Rod Pump – 85%
Gas Lift 10%
Figure 5
Production from artificial
lift wells for a major
international oil company
Figure 5 is a corresponding breakdown for a major international oil company’s more
than two and a half million barrels a day of gross fluid production (which yields more
than one million barrels a day of oil) which is lifted by the various types of artificial
lift. The larger contribution from Gas Lift and ESPs reflects the greater contribution
of high rate and offshore wells compared to the figures for the USA.
Figure 4
Relative frequency of
different types of artificiallift methods installed in the
USA in 1992 (data from
J.Clegg, S.Buchan and
N.Heln, JPT, December
1993, p1128)
Beam Pumping 17%
Electric Submersible Pump 29%
Gas lift 51%
Progressing Cavity Pump 2%
Hydraulic 1%
Venturi 1%
Artificial lift is being more widely applied in field development than ever before due
to:
(i)
Field development status - oil producing provinces such as the North Sea
have become mature with the consequent reductions in flowing bottom hole
pressure (depletion) and increasing water cuts.
(ii) Absence of Pressure maintenance. The development plans for many of the
early, giant North Sea fields employed early water injection to maintain the
reservoir pressure above the hydrocarbon fluid’s bubble point, even after a
significant fraction of the hydrocarbon reserves had been produced. This meant
that the high water cut wells still continued to flow at high production rates.
Many of the current crop of smaller fields currently being developed do not
employ any form of pressure maintenance, resulting in a early need for artificial
lift.
(iii) Satellite or Subsea Wells. These wells are often positioned a considerable
distance from the host platform. The extra pressure drop caused by flow
through these long, subsea pipelines needs to be overcome by some form of
pressure boosting. This could either be an increased pressure boost from an
ESP installed downhole or by a multiphase pump mounted on the sea bed.
Department of Petroleum Engineering, Heriot-Watt University
7
(iv) Business drivers. Profitable field development requires that the average well
production rate exceed a minimum value with higher values being more
profitable. Well design can increase the well flow rate of return. Recent well
design innovations include:
(a)
Advanced well design: drilling of long (horizontal) exposures to the
producing formation.
(b)
Large diameter tubing to decrease the frictional pressure losses
e.g. Norske Shell’s Draugan field achieved flow rates of over 76,000
bopd with a 9 in. production tubing.
(c)
Early installation of artificial lift to increase the flow rate.
Marginal (possibly subsea) field development requires reservoir development
plans with a minimum number of wells where there is little or no need for well
intervention to repair or modify the downhole installation. High equipment
reliability is thus a “must”. Developments in the application of downhole
electronics and measurement sensors means that monitoring of the performance
of the artificial lift equipment of the operating conditions results in improved
performance. One operator found that implementation of a real time,
Supervisory, Control and Data Analysis (SCADA) system in a large artificial
lift project consisting of more than 500 rod pumped wells resulted in a:
(a)
6% production increase (lift conditions optimised and immediate alarm
given when wells ceased producing).
(b)
50% reduction in well entries (earlier recognition of developing problems
allowed preventive maintenance).
(c)
5% reduction in energy consumption (wasteful “over-lifting” operating
conditions recognised immediately).
(v) Integration of artificial lift software into field development planning and
operation. Design procedures for artificial lift techniques has undergone a
tremendous development:
8
(a)
1975: nomograms, slide rules and early calculators allowed single well,
“snap shot” optimisation for the current production conditions.
(b)
1985: early, software based, field wide surveillance systems became
available.
(c)
1995: “Thinking in systems” became a reality. The first field/production
system models became available. These contain (simple) versions of
models describing the performance of the reservoir, well (including any
installed form of artificial lift), manifolds, facilities and pipeline (Figure 6).
This integrated system has the ability to:
2
Selection of Artificial Lift Types
•
Calibrate (or automatically history match) the various model
elements against actual measured data.
•
Compare the measured field data with the production plan or
forecast.
•
Optimise (normally maximise) the production within any
(permanent or temporary) constraints.
•
Recognise discrepancies between forecast and actual production.
The identified wells/production systems are thus listed for
engineering investigation and possible remedial activities.
Integrated Production Model
Figure 6
An integrated production
model
Simple
Simple
Simple
Simple
Reservoir
Model
Well
Model
Manifold
and
Pipeline
Model
Process
Model
Compare to Forecast
Auto History Match
Field Data
for Model
calibration
and
comparison
with
production
plan/forecast
Results
Remedial
activities
and
Archiving
Building and operation of such a model requires input from geoscientists, production
technologists, pipeline and process engineers as well as production operations staff.
The availability of such an integrated models allows the “total cost of ownership”, or
“total lifetime cost” of a particular form of artificial lift, to be evaluated when selecting
the preferred artificial lift technique for a particular field development. For example,
two or more lift methods may be technically capable of producing a well at the
designated production rate. It is relatively easy to obtain figures for the initial, capital
costs of installation. However, the equally (or even more important) costs due to
(un)reliability, energy consumption, maintenance, manpower etc. figures have to be
obtained from field operational data.
(vi) Technical innovation has increased the scope of artificial lift. One example is
the development of multiphase pumps which are now available for both subsea
and downhole application. For example, the B area of Captain field in the North
Sea contains a viscous (50-150 cp) crude oil bearing zone overlain by a gas cap.
Production of these reserves was only possible due to development of a
pump capable of operating with free gas fractions in excess of the 30% vol (the
normal operating limit of a conventional ESP). Innovation has also resulted in:
(a)
the development of hybrid technologies such as downhole separation
and pressure boosting for oil production and water injection.
(b)
a step change in artificial lift reliability (see Table 1) through:
Department of Petroleum Engineering, Heriot-Watt University
9
•
improved engineering design
•
the ability to monitor downhole conditions from the surface.
•
better materials selection and, most importantly,
•
better training of wellsite personnel who install and operate the
equipment.
Date
1970 1990
Lift Type
Sucker Rods
1990 1993 1997
Mean Time Before Failure (months)
20
75
Rod Pump
20
40
ESP
15
48
Data Source
1983 1990 1997
Amoco*
10
20
37
♦
Thums , California
12
30
North Sea
56
†
* Private communication J. Lea, Amoco.
♦
T. Lutz, presented at Artificial Lift 1997, Dubai.
†
Presented at 1997 SPE Artificial Lift Workshop.
5. SELECTION OF ARTIFICIAL LIFT CRITERIA
There are many factors that influence which is the preferred form of artificial lift.
Some of the factors to be considered are:
5.1. Well and Reservoir Characteristics
(i)
Production casing size.
(ii) Maximum size of production tubing and required (gross) production rates.
(iii) Annular and tubing safety systems.
(iv) Producing formation depth and deviation (including doglegs, both planned and
unplanned).
(v) Nature of the produced fluids (gas fraction and sand/wax/asphaltene production).
(vi) Well inflow characteristics. A “straight line” inflow performance relationship
associated with a dead oil is more favourable than the curved “Vogel” relationship
found when well inflow takes place below the fluid’s bubble point. Figure 7
shows that reducing the flowing bottomhole pressure from 2500 to 500 psi
increases the well production rate by 125% for the dead oil. This is more than
double the 60% increase expected for the same reduction in bottom hole pressure
if a “Vogel” type inflow relationship is followed with a well producing below
the bubble point.
10
Table 1
Artificial Lift Reliability
2
Selection of Artificial Lift Types
5000
Figure 7
Influence of fluid in flow
performance on production
increase achieved when
well drawdown is increased
Flowing Bottomhole Pressure (PSI)
4500
4000
3500
3000
125% Increase in Production
2500
"Straight Line" IPR
2000
1500
1000
Reduction
in FBHP due to
artificial lift
installation
Vogel Curve
(Below Bubble Point)
60
180
500
~ 60% Increase in Production
0
0
120
240
300
360
420
480
540
600
660
720
780
Total Liquid Production, (BFPD)
5.2. Field Location
(i)
Offshore platform design dictates the maximum physical size and weight of
artificial lift equipment that can be installed.
(ii) The on-shore environment can also strongly influence the artificial lift selection
made. For example:
(a) an urban location requiring a maximum of visual and acoustic impact or
(b) a remote location with minimal availability of support infrastructure
can lead to different artificial lift types being selected for wells of similar design and
producing characteristics.
(iii) Climatic extremes e.g. arctic operations will also limit the practical choices.
(iv) The distance from the wellhead to the processing facilities will determine the
minimum wellhead flowing pressure (required for a give production rate). This
may, for example, make the choice of an ESP more attractive than Gas Lift. This
is because the extra pressure drop in the flowline, due to the injected gas, makes
Gas Lift an unsuitable option for producing satellite hydrocarbon accumulations
isolated from the main field.
(v) The power source (natural gas, mains electricity, diesel, etc) available for the
prime mover will impact the detailed equipment design and may effect reliability
e.g. the voltage spikes often associated with local electrical power generation
have been frequently shown to reduce the lifetime of the electrical motors for ESP’s.
5.3. Operational Problems
(i)
Some forms of artificial lift e.g. gas lift are intrinsically more tolerant to solids
production (sand and/or formation fines) than other forms e.g. centrifugal pumps.
Department of Petroleum Engineering, Heriot-Watt University
11
(ii) The formation of massive organic and inorganic deposits - paraffins, asphaltenes,
inorganic scales and hydrates - are often preventable by treatment with suitable
inhibitors. However, additional equipment and a more complicated downhole
completion are required unless, for example, the inhibitor can be carried in the
power fluid for a hydraulic pump or can be dispersed in the lift gas.
N.B. The physics and chemistry of these processes was discussed in Chapter 4
entitled “Formation Damage”.
(iii) The choice of materials used to manufacture the equipment installed within the
well will depend on the:
(a)
Bottom Hole Temperatures.
(b)
Corrosive Conditions e.g. partial pressure of any hydrogen sulphide and
carbon dioxide, composition of the formation water etc.
(c)
Extent of Solids Production (erosion).
(d)
Producing Velocities (erosion/corrosion).
5.4. Economics
(i)
A lot of attention is often paid to the initial capital investment required to install
artificial lift. However, the operating costs are normally much more important
than the capital cost when a full life cycle economic analysis is carried out. This
was illustrated by J Clegg et al (JPT, December 1993, p 1128). His data has been
used to prepare Figure 8. It can be seen that, for this well, the capital cost
represents a small proportion of the total project costs. Thus it is often viable to
invest extra to ensure the best equipment is installed in the well if this will result
in increased revenue (production) and/or reduced operating costs.
Land Well
Major Contributing Factors
to Project Net Value
Net Production Rate (BOPD)
200
Maintainence
and Operators 6%
Nominal Decline
= 25% / Year
100
9
8
7
Energy 6%
Capital Cost 1%
Net Revenue 87%
6
Reserves
290 x 103 bbl oil
4
3
3.5
3
Economic Limit
0
5
10
15
Time - (Years)
20
(ii) Good operating cost data for the different artificial lift methods in different
locations is difficult to find. Reliability (discussed earlier) is one key issue
while the second is energy efficiency (and hence energy costs). This latter is
12
Figure 8
Example full life cycle
eceonomics
2
Selection of Artificial Lift Types
more tractable since it can be calculated from first principles. There is a wide
variation - see Figure 9. Only rod pumps, ESP’s and PCP’s show values >50%
while gas lift, particularly of the intermittent variety, is inefficient in energy
terms. Changing energy costs can alter the ranking order of the various artificial
lift methods.
80
Figure 9
Comparison of the energy
efficiency of the major
artificial-lift methods
adapted from J.Clegg et al
(JPT, December 1993,
p1128)
Energy Efficiency (%)
70
60
50
40
30
20
10
0
Rod
Pump
PCP
ESP
Turbine Venturi
Hydraulic Hydraulic
Pump
Pump
Gas
Lift
Intermittent
Gas Lift
(iii) Maintenance costs will vary between operating locations depending on the
state of the local, service company infrastructure. It can be costly in remote
locations.
(iv) The number of wells in the field with that particular form of artificial lift
(economy of scale) will influence the operating costs.
(v) Similarly, the desirability and/or need for automation (how many operators are
to be employed) and the decision as to whether or not to install centralised
facilities will also influence the operating costs.
(vi) The speed with which the “learning curve” is climbed for the more sophisticated
forms of artificial lift will depend on the training provided and the skill base of
the operations staff.
5.5. Implementation of Artificial lift Selection Techniques
As discussed the artificial lift design engineer is faced with matching facility
constraints, artificial lift capabilities and the well productivity so that an efficient lift
installation results. Frequently, the type of lift has already been determined and the
engineer has the problem of applying that system to the particular well. A more
fundamental question is how to determine the optimum type of artificial lift to apply
in a given field.
There are certain environmental and geographical considerations that may be overriding.
For example, sucker rod pumping is by far the most widely used artificial lift method
in North America. However, sucker rod pumping may be eliminated as a suitable form
Department of Petroleum Engineering, Heriot-Watt University
13
of artificial lift if production is required from the middle of a densely populated city
or on an offshore platform with it’s limited deck area. There are also practical
limitations - deep wells producing several thousands of barrels per day cannot be lifted
by rod pumps. Thus, geographic and environmental considerations may make the
decision. However, there are many considerations that need to be taken into account
when such conditions are not controlling.
Some types of artificial lift are able to reduce the sand face producing pressure to a
lower value than others. The characteristics of the reservoir fluids must also be
considered. Wax & formation solids present greater difficulties to some forms of
artificial lift than others. The producing gas-liquid ratio is key parameter to be
considered by the artificial lift designer. Gas represents a significant problem to all
of the pumping methods; while gas lift, on the other hand, utilizes the energy contained
in the produced gas and supplements this with injected gas as a source of energy.
The “Advantages and Disadvantages of the Major Artificial Lift Methods” are listed
and compared in Tables 2 & 3.
Rod Pumps
Electric
Submersible Pump
Venturi Hydraulic
Pump
Simple, basic design
Extermely high volume
lift using up to1,000 kw
motors
High volumes
Solids tolerant
Can use water as
power fluid
Large volumes in high
PI wells
Unobtrusive surface
location
Remote power source
Simple maintenance
Downhole telemetry
available
Tolerant high well
deviation / doglegs
Unobtrusive surface
location / remote power
source
Unit easily changed
Simple to operate
Gas Lift
Can achieve low BHFP
Can lift high temperature,
viscous oils
Pump off control
Tolerant high well
elevation / doglegs
Progressing Cavity
Pump
Solids and viscous
crude tolerant
Energy efficient
Unobtrusive surface
location with downhole
motor
Tolerant high well
deviation / doglegs
Corrosion / scale
treatments possible
Table 2
Advantages of major
artificial lift methods
Tolerant high GOR
reservoir fluids
Wirleine maintenance
Rod Pumps
Electric
Submersible Pump
Venturi Hydraulic
Pump
Gas Lift
Progressing Cavity
Pump
Friction in crooked / holes
Not suitable for shallow,
low volume wells
High surface pressures
Lift gas may not be
available
Elastanes swell in some
crude oils
Not suitable for viscous
crude oil or emulsions
Pump off control
difficult
Susceptible to gas
freezing / hydrates at
low temperatures
Problems with rotating
rods (windup and after
spin) increase with
depth
Pump wear with solids
production (sand, wax etc.)
Free gas reduces pump
efficiency
Full workover required
to change pump
Obtrusive in urban areas
Cable susceptible to
damage during
installation with tubing
Downhole corrosion
inhibition difficult
Cable deteriorates at
high temperatures
Heavy equipment for
offshore use
Gas and solids intolerant
Increased production
casing size often
required
14
Sensitive to change in
surface flowline
pressure
Free gas reduces pump
efficency
Power oil systems
hazardous
High minimum FBHP.
Abandonment pressure
may not be reached
High minimum FBHP.
Abandonment pressure
may not be reached
Casing must withstand
lift gas pressure
Table 3
Disadvantages of major
artificial lift methods
2
Selection of Artificial Lift Types
5.6. Long Term Reservoir Performance and Facility Constraints
Another factor that needs to be considered is long term reservoir performance. Some
years ago Neely indicated that two approaches, both of which have disadvantages, are
frequently used to solve the problem of artificial lift selection and sizing.
(i)
A prediction of long term reservoir performance is made and artificial lift
equipment installed that can handle the well’s production and producing
conditions over its entire life. This frequently leads to the installation of oversized
equipment in the anticipation of ultimately producing large quantities of water.
As a result, the equipment may have operated at poor efficiency due to underloading over a significant portion of its total life.
(ii) The other extreme is to design for what the well is producing today and not
worry about tomorrow. This can lead to many changes in the type of lift
equipment installed during the well’s producing life. Low cost operations may
result in the short term, but large sums of money will have to be spent later on
to change the artificial lift equipment and/or the completion.
Likewise, in a new field development, the fluid handling requirement from some
artificial lift types can significantly increase the size and cost of the facilities required.
Only the produced fluid is handled through the facilities with rod pumps and ESPs.
However, gas lift requires injection gas compression and distribution facilities and the
additional, produced gas increases the size of the production facilities required.
Similarly, the use of Hydraulic pumps can result in the additional power fluid volumes
being many times that of the produced oil volume. This results in high fluid handling
costs as well as difficulties in accounting for the oil produced (when oil is used as a
power fluid).
The selection of the artificial lift for a particular well must meet the physical
constraints of the well. Once a particular type of lift is selected for use, consideration
should be given to the size of the well bore required to obtain the desired production
rate. It can happen that the desired production can not be obtained because the casing
programme was designed to minimize well cost, resulting in a size limitation on the
artificial lift equipment that can be installed. Even if production rates can be achieved,
smaller casing sizes can lead to higher, long term production costs due to well
servicing problems, gas separation problems etc.
Figure 10 is offered as a screening selection tool in which areas where particular
artificial lift methods have been frequently applied are compared as a function of
depth and well rate. It must be realised that there are many proven applications where
a particular form of artificial lift has been installed in a well at greater depths or
produced at higher rates than is indicated in this figure.
Department of Petroleum Engineering, Heriot-Watt University
15
Gross Production Rate (bpd)
0
1
10
100
1000
10000
100000
Intermittent
Plunger
Gas Lift
2000
4000
Depth, Ft.
6000
8000
10000
12000
Continuous
Gas Lift
Sucker
Rod
14000
ESP
16000
18000
20000
Gross Production Rate (bpd)
0
2000
4000
1
10
100
1000
10000
100000
Hydraulic
Venturi
Depth, Ft.
6000
8000
PCP
10000
12000
14000
16000
18000
Hydraulic
Reciprocating
6. ROD PUMPS
6.1. Introduction
(Sucker) rod or beam pump was the first type of artificial lift to be introduced to the
oil field. It is also the most widely used in terms of the number of installations world
wide. In 1993, some 85% of the USA population of artificially lifted wells was
produced by rod pumps and more than 70% of these produced less than 10 barrels of
oil per day. The low cost, mechanical simplicity and the ease with which efficient
operation can be achieved makes rod pumps suitable for such low volume operations.
Rod pumps can lift moderate volumes (1,000 bfpd) from shallow depths (7,000 ft) or
small volumes (200 bfpd) from greater depths (14,000 ft). They are normally
manufactured to standards set by the American Petroleum Institute (API). This means
that, unlike other artificial lift methods, the equipment manufactured by the various
supplies is fully interchangeable.
6.2. The Pumping Unit
The surface equipment for a rod pump is illustrated in Figure 11. The prime mover,
normally an electric motor or gas engine, drives a speed reducing set of gears so that
its fast rotation, of say 600 revolutions per minute, is reduced to as low as 20 strokes
16
Figure 10
Typical application areas of
artificial lift techniques
2
Selection of Artificial Lift Types
per minute or less. The connection between the surface pumping unit and the
downhole pump is the polished rod and the sucker rods. The polished rod moves up
and down through a stuffing box mounted on top of the wellhead. This stuffing box
seals against the polished rod and prevents surface leaks of the liquids and gasses
being produced by the well.
Bridle
Horse Head
Beam
Prime Mover
Gear Reducer
V Belt
Clamp
Polished Rod
Counter
Weight
Stuffing Box
Tubing Head
Figure 11
The surface equipment for
a rod pump
Flow line
Casing Head
PUMPING UNIT
Casings
Tubing
Sucker Rod
6.3. The Sucker Rods
The sucker rods, typically 25 ft long, are circular steel rods with diameters between
0.5 in and 1.125 in, in increments of 0.125 in. A threaded male connection or pin is
machined at each end of the rod. The two rods can be joined together by use of a double
box coupling (Figure 12). Square flats are machined near the pins and at the centre
of the coupling to provide a grip for a wrench to allow the rods and couplings to be
screwed together. The sucker rods are subjected to continuous fatigue when the pump
is in operation. The weight of the rod string is one component of this fatigue load it can be minimised by using a tapered sucker rod string. This involves installing
lighter, smaller diameter rods lower down in the well where the load they have to
support (weight of rods and fluid in the tubing string) is less than at the top of the well.
Double Box Coupling
Pin
Figure 12
Sucker rods are joined
together by a coupling
Circular Sucker Rod
Pin
Square Flat for Wrench
Circular Sucker Rod
6.4. The Pump
The pump is located near the perforations at the bottom of the string of sucker rods.
Figure 13 shows that it consists of a hollow plunger with circular sealing rings
mounted on the outside circumference moving inside a pump barrel which is either
inserted into the tubing or is part of the tubing itself. A standing valve is mounted at
Department of Petroleum Engineering, Heriot-Watt University
17
the bottom of the pump barrel while the travelling valve is installed at the top of the
plunger. The standing and travelling valves consist of a ball which seats (closes off)
an opening.
Downward
Rod Movement
Up-ward
Rod Movement
Tubing
Sucker
Rods
Travelling
Valve
Off Seat
Polished
Pump
Barrel
Fluid
being
Lifted to
Surface
Fluid Flow
from Pump
Barrel to
Tubing
Fliud
Flow
Standing
Valve
On Seat
Sealing
Rings
Around
Plunger
Circumference
Traveling
Valve
Off Seat
Traveling
Valve
On Seat
Plunger
Standing
Valve
On Seat
Standing
Valve
Off Seat
Fluid
Inflow From
Perforations
The “UP” and “DOWN” movement of the pump barrel allows the fluid flow to open
and shut these valves as shown in Figure 13. The left hand schematic shows the
plunger status at the end of the “DOWN” stroke. The "Upward" rod movement
reduces the pressure within the pump barrel and the upward flow of fluid from below
the pump lifts the standing valve’s ball off its seat. The pressure due to the fluid
column above the plunger keeps the travelling valve ball on its seat. The situation is
reversed during the “DOWN” stroke - compression of fluid within the pump barrel
forces it to flow through the hollow plunger and to lift the travelling valve off its seat;
while ensuring that the standing valve remains closed.
6.5. Rod Pump Operation
This chapter describing of rod pump operation deals with some of their more
important operational aspects.
6.5.1. Pump-Off Control
The well will produce the maximum gross volume of fluid when the drawdown is
maximised i.e. the fluid level in the well is maintained at a limited distance above the
pump. The pump capacity will often be greater than the well inflow capacity - the
pump motor must be stopped at regular intervals when the fluid level is reduced to a
specified, minimum safety level above the pump. This monitoring is often performed
with an “Echometer”. Figure 13a. This tool is attached to the wellhead in a pressure
tight housing. It consists of a firing mechanism, a microphone and an amplifier
recorder. Typically, a gas actuated device generates an accoustic pulse at the wellhead.
18
Figure 13
Rod pump operation
2
Selection of Artificial Lift Types
Rods
Tubing
Sonolog
equipment
Casing
collars
Reflected wave amplified
and arrival time recorded.
0
Charge ignited
Accoustic pulse
Reflection from
tubing collars
Reflection
time
Fluid level
Fluid level
Figure 13a
"Echometer" or "Sonolog"
fluid level survey.
Pump
This is reflected back by subsurface items such as tubing collars, but the main
reflection is from the fluid level at the bottom of the casing. The depth of the fluid level
can now be found by multiplying this time by half the velocity of sound in the casing/
tubing annulus.
The pump can now be restarted once the casing fluid level has risen sufficiently due
to inflow from the reservoir.
Once calibrated, the pump unit can be put on timer control, i.e. the “Echometer” need
not be used continuously since well inflow performance normally shows a steady,
predictable decline rate with time. Further, the performance of the pump itself can be
checked by measuring the dynanometer card at regular intervals (see section 2.6.5.5
on Pump Diagnosis).
6.5.2. Gas Influx
(Free) gas sucked into the pump will reduce the pump efficiency due to its compressible
nature. Placing the pump below the perforations maximises the use of the (limited)
gas separation capacity of the casing. Minimising the volume between the travelling
valve at the bottom and the downstroke and the standing valve helps ensure the gas
is pushed out of the pump during each down stroke.
Placing the pump below the perforations is advantageous since it increases the
maximum possible drawdown. This is not always possible in practice. Many types of
“gas anchors” have been tried in order to overcome the resulting difficulties. They all
aim to separate the “free” gas from the liquid prior to the liquid entering the pump. The
gas flows into the tubing/casing annulus where it is vented or gathered at as low a
pressure as possible (to allow a minimum bottom hole pressure to be reached). An
example of an effective gas anchor is shown in Figure 14. Here a packer and crossover
direct the multiphase flow above the pump, ensuring that only liquid is sucked into the
pump barrel. (The extra packer also allows the tubing to be anchored - minimising the
fatigue loads to which the rod string is exposed.
Department of Petroleum Engineering, Heriot-Watt University
19
Rods
Gas
Tubing
Casing
Liquid Flow
Bypass
Tube
Pump Barrel
and Valves
Produced Gas
and Liquid
Pump Intake
(Perforated Pipe)
Crossover
Packer
Producing Zone
Figure 14
Packer type gas anchor
6.5.3. Centralisers
The sucker rods may require centralisers or protectors in deviated wells to reduce wear
on the tubing and rods (Figure 15). This requirement becomes more extreme in
crooked or highly deviated wells, sometimes to the extent that rod pumps can not be
used.
Rod Movement
Centraliser / Protector
Tubing
Sucker Rod
Extra centraliser / protector required here
to prevent rod / tubing wear
Centraliser / Protector
Figure 15
Centralising rod reduces
friction and wear
20
2
Selection of Artificial Lift Types
6.5.4. Solids
Rod pumps also have a very limited ability to lift sand due to the low fluid velocity
in the production tubing plus wear on the pump valves, seats and plunger. The latter
can be overcome by suitable pump design and choice of construction materials.
Wax and inorganic scale deposition also interfere with efficient rod pump operation.
Continuous injection of an inhibitor below the pump to ensure protection of the
complete downhole equipment is complicated unless there is a (normally unacceptable)
complication of the completion design. Removal of wax by hot oil/solvent circulation
or injection of a scale inhibitor into the formation are also possible (see Chapter 4).
Further, recovery of the pump and rods using a well pulling hoist is often a relatively
low cost, simple operation.
6.5.5. Pump Diagnosis
The condition of the pump can be evaluated by measuring the load at the top of the
polished rod as a function of its position i.e. as it moves up and down during the stroke
length. This is recorded in the form of a dynamometer card. Examples of theoretical
dynanometer cards are shown in Figure 16.
Polished Rod Up
Maximum Load
Load
"Up" Stroke
Minimum Load
"Down" Stroke
Polished Rod Down
Stroke Position
(a) Inelastic Rods
Polished Rod Up
"Down" Stroke
es
Rela
x
Stre
Rod
Load
Minimum Load
"Up" Stroke
Rod
tche
s
Maximum Load
Polished Rod Down
Stroke Position
(b) Elastic Rods
Maximum Load
Travelling Valve
Closes
Walking Beam
Decelerating
Rods and Fluid
Being lifted
Polished Rod Up
(Maximum Stroke Position)
Recoil
Polished Rod down
(Minimum Stroke Position)
Load
Figure 16
(Theoretical) dynamometer
cards for:
(a) inelastic rods
(b) elastic rods
(c) elastic rods with rod,
fluid and surface pump unit
Standing Valve
Closes
Walking Beam
Decelerating
Rods and Plunger
Falling Through Fluid
Minimum Load
Stroke Position
(c) Full Simulation
Department of Petroleum Engineering, Heriot-Watt University
21
(i)
Figure 16 (a) records the variation in load for inelastic rods. The load is either
high or low depending on whether the polished rod is moving up or down.
(ii) Figure 16 (b) adds the elasticity of the rods - the full increase in load is no longer
instantaneous when the polished rod starts moving in a particular direction.
(iii) Figure 16 (c) adds the further dimension of rod - fluid and surface pump unit
dynamics. The times at which the various processes become controlling during
the pump cycle are indicated.
These theoretical calculations have been made for a perfectly operating pump unit
pumping liquid only. Practical problems such as:
(i)
excessive rod or pump friction
(ii) restriction in the flow-path
(iii) vibrations
(iv) sticking plunger
(v) gas lock etc.
will all alter the shape of the dynamometer trace in a distinctive manner allowing the
source of the problem to be diagnosed and then rectified.
6.6. Pump Design
API Recommended Practice 11L, published by the American Petroleum Institute,
describes a field proven method for designing all elements of a Rod Pump. We will
only discuss selected elements here.
6.6.1. Pump Rate
The pump rate (Q) is related to the volume displaced (V) by each pump stroke and the
speed rate or number of strokes per minute (N). Thus:
Q = K*V*N*φ = K*A*S*N*φ
Where:
A is the area of the pump barrel
S is the length of the pump stroke
φ is the efficiency factor
K is a constant to convert the above units to barrels per day
The maximum speed (N) of the pump unit is determined by the speed at which the
sucker rods fall downward in the “DOWN” stroke. (Early pump rod failure occurs due
to metal fatigue if they are placed under compression due to them being forced
downwards by the pump unit exceeding this maximum speed.) As would be expected,
this maximum speed decreases as the length of the pump stroke increases. Typical
maximum values are quoted in Table 4.
22
2
Selection of Artificial Lift Types
Maximum Allowable Pump Speed†
Stroke Length (in.)
Table 4
Maximum allowable pump
speed
30
60
90
120
180
240
300
Maximum Pump Speed (SPM*) 34
24
19
17
14.5
11.5
10.5
For a Conventional Pump Unit
* Strokes Per Minute
N.B. Low pump speeds and large diameter plunges lead to the greatest energy
efficiency, but also the largest equipment loads. It is common practice to put a few
larger rods capable of carrying any compression loads due to buckling at the bottom
of the rod string. Also, the addition of sinker bars will increase the rate of rod fall, but
also increase the load on the rods.
Two factors which reduce the efficiency of the pump are gas influx (see section
2.6.5.2) and rod stretch.
6.6.2. Rod Stretch and Over Travel
During the “UP” stroke the rods support their own weight and that of the fluid in the
tubing. Hooke’s law of elasticity dictates that the rods will increase in length in
response to this load. This decreases the effective travel of the plunger downhole
compared to the distance moved at the surface. For example, 0.875 in sucker rods
driving a 2 .25 in plunger pump set at 6,000 ft will show a stretch of 29 in when lifting
fresh water.
The load imposed on the rods by the fluid is related to the area of the plunger times
the hydrostatic head. Rod stretch thus increases when a larger diameter pump is
selected. There is a similar effect of tubing stretch. This also reduces the effective
stroke length, if the tubing is not anchored.
Plunger Overtravel”, by contrast, increases the effective stroke length. This effect is
caused by elongation of the rods due to dynamic forces generated by the weight of the
rods reversing direction during the pumping cycle. At this point, the weight of a 6,000
ft string of 0.875 in rods (weight 2.25 lb/ft or 13,500 lb total weight) will be brought
to a halt and reverse direction over a time period of less than one second. The rod string
velocity will be approximately 6 ft/second when the plunger nears the bottom of its
stroke if the pump unit is operating with a 64 in stroke length at a pump speed of 15
strokes per minute. The resulting increase in length of the rods generated by halting
and reversing this momentum is called “Plunger Overtravel” (approximately 14"
according to Marsh and Coberly’s 1931 method for this example).
We can now estimate that:
Effective stroke length = surface stroke length - rod stretch + plunger overtravel
≈ 64" - 29" + 14" ≈ 49"
6.6.3. Pumping Limit Load Calculations
API RP 11L can be used to calculate the maximum and minimum polished rod loads,
the peak torque and the theoretical horsepower required once the pump speed, stroke
length, plunger diameter and rod sizes have been chosen. Typically, a motor power
of twice this theoretical value should be installed to allow for surface and downhole
energy losses.
Department of Petroleum Engineering, Heriot-Watt University
23
It was mentioned previously that the rod string was continuously subjected to fatigue.
Goodman showed how the “Maximum Allowable Stress” on the sucker rods was a
function of the grade of the sucker rods and the minimum polished rod load/cross
sectional area of the top rod. This “Maximum Allowable Stress” has to be decreased
by a service factor related to the operating conditions e.g. the presence of corrosive salt
water or, more importantly, hydrogen sulphide. This explains why these aggressive
fluids limit the application (either maximum depth of a pump installation or maximum
fluid volume which can be lifted) of rod pumps unless more expensive, speciality
grade sucker rods are used.
7. ELECTRIC SUBMERSIBLE PUMPS (ESPs)
7.1. Introduction
Electric Submersible Pumps (ESP’s) are a versatile form of artificial lift with pumps
ranging from 150 to 60,000 bfpd in operation. A typical low pressure well that is being
artificially lifted using an ESP system is illustrated in figure 17. The functions of the
various components are summarised as follows:
Switchboard
and Motor
Controls
Vent Box
Cable Penetrator
Through Wellhead
High Voltage Electrical Supply
Surface Cable
Cable
Cable Banded
to Tubing
Cable Protector
Tubing
Pump Discharge
Head
Flat Cable
Centrifugal Pump
Pump Intake with
Optional Gas
Separator
Seal or Protector
Casing
Pothead Connects
Cable to Motor
Electric Motor
Downhole Sensor
Package
Centraliser
24
Fluid
Figure 17
A well completed with
artificial lift using an
electric submersible
centrifugal pump
2
Selection of Artificial Lift Types
Frequency Drive (VFD). A VFD allows the speed of the electric motor to be
altered e.g. starting the pump using the “nameplate” design frequency of 50Hz
(Europe) or 60Hz (North America) results in high instantaneous electric motor
currents since the power developed by the pump is proportional to the
frequency. These can be reduced by supplying the electric power at lower
frequencies. It also allows the pump flow rate to be adjusted to the well inflow
conditions since flow rate is also proportional to frequency.
Practical experience shows that a 60Hz motor can be operated between
35Hz and 80Hz. VFD installation increases the surface energy losses from
some 3% to 5-15% of total power supplied.
(ii) The vent box separates the surface cable from the downhole cable. This ensures
that any gas, which travels up the downhole cable, does not reach the electrical
switchgear.
(iii) The downhole cable penetrates the wellhead. It is banded to the tubing at
regular intervals. Additional protection is supplied by cable protectors which
are installed at critical points to prevent damage while the completion is being
run into the hole. A “flat pack” cable shape is employed across the larger
diameter completion components to minimise total width. The cable enters the
electric motor housing at the Pothead. It not only carries the electrical power
supply for the motor (up to 750HP motors are being routinely installed), but
also carries the measurement signal from the downhole sensor package
installed underneath the motor.
(iv) The pump unit consists of a stacked series of rotating centrifugal impellers
running on a central drive shaft inside a stack of stationary diffusers, i.e. it is
essentially a series of small turbines. The pressure increase is proportional to
the number of stages while the pump capacity (volume) increases as the
diameter of the impeller increases. Rotation of the impeller accelerates the
liquid to be pumped which is then discharged into the diffuser where this kinetic
energy is transformed into potential energy i.e. a pressure increase. The
impeller/diffuser pairs are arranged in series with the discharge of one unit
being the suction of the next one. The number of pump stages (impeller/
diffuser) pairs may range between 10 and more than 100, depending on the
pressure increase required. Abrasion resistance to produced solids is very
dependent on the detailed design and materials selection employed during
pump design.
However, as discussed at the beginning of this chapter, ESP’s with their rapidly
rotating internals are not really compatible with large quantities of produced sand
even when hardened, wear resistant materials are used. The option of using other
forms of artificial lift - gas lift and PCPs - should be considered.
Standard pump impellers are very sensitive to gas fractions greater than 20% vol in the
produced fluid. Alternatively, changes in pump design such as altering the design
of the impeller from pure radial flow to mixed (i.e. a combination of both radial and
axial) flow can double the gas/fluid ratio to 40% vol.
Department of Petroleum Engineering, Heriot-Watt University
25
Motor Housing
Impeller
Drive Shaft
A tapered pump design using mixed flow impellers in the lower pressure stages
(with the higher gas volume fractions) and radial flow impellers in the upper stages
can prove to be effective.
(v) The pump intake may include a rotary gas separator if gas fractions higher than
20%. This consists of a centrifugal device, which separates the lower density,
gaseous phase from the denser liquid phase. The latter is concentrated at the
centre of the device and enters the pump suction while the lighter, gas phase is
directed towards the casing/tubing annulus where gas is vented/gathered at
surface.
A reduction in casing/tubing annulus pressure increases the maximum achievable
drawdown at the formation face. A single rotary gas separator can increase the
ESP’s gas handling capabilities up to 80% vol. Two separators, arranged in
tandem, are even more efficient, increasing the pumpable gas fraction to >90%
vol. However, the addition of extra equipment always comes with the cost of
greater operational problems e.g. produced formation solids can damage the
rotary separator, scale formation can unbalance rapidly rotating equipment.
Some operators will not use them due to these problems which have resulted
in rotary separators having a poor reputation for reliability .
26
Figure 17(a)
Cutaway schematic
drawing of Electric
Submersible pump
2
Selection of Artificial Lift Types
Alternative completion strategies where the pump is placed below the
perforations or some forms of gas anchor (see also section 2.4) may be
employed to limit the gas influx. Figure 18(a) illustrates the use of a shroud to
make use of the casings ability to separate produced (free) gas from the liquid.
Note that, if it is decided to mount the pump below the perforations {figure
18(b)}, then a shroud is also required to initially direct the production flow
below the electric motor where it provides the necessary cooling. The normal
completion design of mounting the motor at the bottom of the ESP system,
which is then placed above the perforation, provides this cooling automatically.
Liquid
Pump
Gas
Liquid
Gas
Liquid
Pump
Gas
Liquid
Gas
Perforations
Motor
Liquid
Figure 18
Two ESP completion
designs to aid gas
separation in the casing
Liquid
Motor
Protector
Protector
Shroud
Perforations
Downhole
Sensor
(a)
(b)
It should be noted that:
(a)
Liquid/gas separation in the casing will only work when the upward
velocity of the gas bubbles is greater than downward fluid velocity, i.e.
it will work better for lower rate wells with larger annular clearances
and where the gas is produced as large bubbles e.g. from a separate
zone or different perforations to those producing the liquid.
(b)
The fitting of a shroud increases the maximum ESP diameter (requiring
a large casing for installation of the same motor/pump combination or
requires that a smaller diameter pump/motor be chosen. Practical
experience has shown that, for a given power requirement, smaller
diameter equipment is often less reliable Further, high power equipment
is not available in the smaller sizes.
Department of Petroleum Engineering, Heriot-Watt University
27
(vi) The Protector or Seal unit connects the drive shaft of the electric motor to the
pump or gas separator shaft. It also performs as:
(a)
an isolation barrier between the clean motor oil and the well fluids;
(b)
an expansion buffer for the motor oil when it reaches operating
temperature;
(c)
equalises internal motor pressure with the well annular pressure
and
(d)
absorbs any thrust generated by the pump.
(vii) The electric motor is powered by three phase alternating current supplied by the
cable connected to the motor at the pothead. They are available in sizes
between 15 and 900 HP in the manufacturer’s catalogue. Two or even three
motors may be placed in series if high pump power requirements exist.
The motor is filled with oil which insulates the electrical winding. {Ingress of
reservoir fluids (water) is a common cause of motor failure}. A second, less
obvious cause of failure is power surges/voltage spikes/harmonics on the
power supply. These are more prevalent when the power is generated locally
rather than supplied by the (electrical) utility grid. The requirements for the use
of a completion design which directs the fluid flow along the motor to
provide sufficient cooling was discussed above in section (v) on pump intake
design.
When the motor is switched off the head of fluid present in the tubing will
reverse the flow direction through the pump as it flows back into the reservoir.
This will cause the motor to spin backwards. Trying to restart the motor while
it is rotating backwards will lead to the motor burning out very quickly. This
can be avoided by
(a)
installing a check valve in the tubing to prevent fluid backflow
(however this results in a “wet string” when the tubing/ESP is
recovered
(b)
electronically preventing motor restart for a specified time after it
has been shut down or
(c)
using a sensor to detect backspin and preventing motor restart
(see the next section on downhole sensors).
(viii) A downhole sensor package may be mounted underneath the motor.
Measurements can include:
28
(a)
Pump suction and discharge pressures and temperatures.
(b)
Fluid intake temperature.
2
Selection of Artificial Lift Types
(c)
Electric motor temperature.
(d)
Vibration.
(e)
Current leakage.
A downhole flow meter and/or phase cut can be added to the above and all the above
data transmitted to surface via the power cable. The above can be combined with
measurement of the power supply frequency and surface current/voltage as well as
wellhead temperature, pressure and surface flow rate so as to be able to present a
complete picture of well performance. The data can be:
(a)
stored at the well site and downloaded to a (hand held) data log
at regular intervals for later analysis
(b)
used to trigger on-site alarms which shut the ESP unit down e.g. if the
pump suction pressure falls below a preset value indicating that the
fluid level in the well is reducing and the well is being “pumped off”
(c)
transmitted continuously to the operations office where more
sophisticated monitoring analysis can be carried out
(d)
replace non-routine well surveillance operations e.g. the sensors may be
sufficiently accurate to obviate the need for running memory gauges
into the well when preforming flowing bottom hole pressure surveys,
build up tests or reservoir pressure monitoring.
Such sensor packages are now being applied to other forms of artificial lift e.g.
directly to EPCPs (see Section 2.9.3) but also to PCP (Section 2.9) rod pumps
(Section 2.6) and gas lift (Chapter 3), where similar measurements will enable
corrective action to be taken to maximise the efficiency of the lift operation.
At the beginning of this section we stated that Figure 17 was for a low pressure well
since a packer had not been installed in the well, i.e. the well will probably not flow
without artificial lift. Inclusion of a packer in the completion design, as is often
required by the regulatory authorities in live and many offshore wells, precludes
venting the gas to the surface via the casing/tubing annulus unless a dual packer
arrangement is employed with a safety valve installed on at least the main production
tubing and (possibly) on the gas vent line as well. One example of a possible
completion design is illustrated in Figure 19.
Department of Petroleum Engineering, Heriot-Watt University
29
Wellhead
Wellhead Penetrations
Gas Vent
1/4" Hydraulic
Control Lines
Surface Controlled,
Subsurface
Safety Valves
Penetrator
Dual Packer
Tubing
Cable, Banded
to Tubing
Sliding Circulation
Sleeve for Well Killing
Pump
Pump Intake
Seal Section
Motor
Production Casing
9 5/8" Casing
Shoe
Sensor Package
Gravel Pack Screen
Gravel Pack
7.2. Well Completion Employing Electric Submersible Pumps (ESPs)
7.2.1. Typical ESP Applications
The ESP application illustrated in figure 17 is the standard application where it is used
as to lift production from a single zone through a single tubing. Many other application
configurations are possible. Examples are given in Figure 20:
(i)
Figure 20 (a) shows aquifer water being lifted from supply zone and pumped
directly to an injection well.
(ii) Figure 20 (b) illustrates a dump flood powered by an ESP where the water
supply well and injection well are combined into one. Note that the ESP is
inverted with the pump at the bottom. The ESP is being used here to replace
the conventional surface mounted transfer pump. Measurement of the
injection flow rate and pressure can be made by inclusion of a sensor package
(not shown) in the well completion design.
30
Figure 19
ESP completion
incorporating packer and
surface controlled sub
surface safety valve
2
Selection of Artificial Lift Types
(iii) Figure 20 (c) shows an ESP placed in a shallow well being used to boost
pressure in a surface flow line (note the shroud installed to ensure adequate
motor cooling).
Low Pressure
Flow Line
High Pressure
Flow Line
Pump
Cable
Protector
Motor
Protector
Protector
Shroud
Pump
Motor
Motor
Aquifer
Supply
Zone
Figure 20
ESP applications
(a) Direct water injection
(b) Powered dumpflood
with ESP
(c) Pressure boosting
surface pipelines with ESP
(d) Horizontally mounted
ESP surface pump
Pump
Aquifer
Supply
Zone
(a)
(c)
Injection
Zone
Injection
Zone
(b)
Low Pressure Line
High Pressure Line
(d)
Motor
Protector
Pump
The pump and motor may also be mounted at the surface - either horizontally (along
the ground) or placed vertically (reduces required platform area for offshore application).
In these cases a conventional, air cooled electric motor is used. Pressure increases of
up to 3000 psi at 7000 bfpd flow rates have been achieved. Higher volume
applications can be catered for by manifolding a series of pump units in parallel.
7.2.2. Horizontal Wells
The ability of ESP to pump large volumes of produced fluid coupled with the
flexibility of pump design and operation makes them very suitable for the large
volume production associated with horizontal wells. Experience has shown that the
pump can be placed anywhere within the well at angles up to 80° providing the dogleg
severity is not too great (<6º/100ft). Placing the pump near the bottom of the well not
only maximises the potential drawdown which can be created at the formation while
the (near) horizontal section will enhance the separation of the gas to the upper portion
of the wellbore due to its lower density.
Department of Petroleum Engineering, Heriot-Watt University
31
7.2.3. “Y” Tool
The “Y” tool is a device to allow wireline or coiled tubing access below the ESP. It
is illustrated in Figure 21. The bypass tubing should be at least 2.375” OD (allowing
1.6875"/16" logging tools to pass), but 2.875” OD tubing is preferable since this
allows the larger sizes of coiled tubing to pass. However, the larger diameter tubing
does reduce the maximum diameter of ESP that can be installed. Larger diameter
motors are more efficient, tend to have longer run lives and are shorter for a given
power requirement.
Wellhead Penetrator
Power Cable
TRSSSV Control Lines
Tubing Retrievable Subsurface Safety Valve (TRSSV)
Centraliser
Packer Penetrators
Production Casing
Pump
Pump Intake
Seal / Protector
Motor
Packer
Fishing Neck for Plug
Wireline / Coiled
Tubing Retrievable
Plug in Nipple
(Optional) 1/4" Chemical
Injection Tubing
Downhole Sensor Package
Packer
Production Liner
Bypass Tubing
Producing Formation
Installation of “Y” tools allows all the normal wireline and coiled tubing conveyed
operations to be carried out below the ESP. These include:
(i)
cased hole logging
(ii) well stimulation
(iii) perforating
(iv) setting bridge plugs for water shut off
(v) installation and recovery of pressure memory gauges
(vi) running and retrieval of plugs
(vii) downhole sampling.
Omission of the “Y” tool from the downhole completion design implies that these
operations are only possible when tubing and ESP are recovered to the surface.
32
Figure 21
The "Y"tool"
2
Selection of Artificial Lift Types
7.3. Basic Pump Selection
The pressure increase that the pump is required to deliver, also called the “Total
Dynamic Head (TDH)” or difference between the pump discharge and suction
pressure, is the sum of three components. Figure 22 pictures the various components:
Multi Well Manifold
Separator
(iii) Surface flow line
back pressure Psurf
(ii)
Hydrostatic head due to fluid column from ESP
depth to surface (p * g * h)
(i)
Friction loss of production tubing (∆Pfric)
(Minimum) Total Dynamic Head (TDH) to be supplied
by pump is the sum of (i), (ii) and (iii)
(Assuming flowing wellbore pressure at pump inlet is zero)
Figure 22
Pump duty requirements
(i)
The hydrostatic head from the ESP pump to the surface. This is equal to the
(average) density of the produced fluid in the tubing (ρ) multiplied by the True
Vertical Depth at which the ESP is installed (h) and the acceleration due to
gravity (g).
(ii) Friction pressure loss in the tubing (∆Pfric)
(iii) The surface pressure (Psurf) required to overcome flowline back pressure and
flow the produced fluid to the separator at the required production rate. This
can have a high value if the completion is a satellite well situated some distance
(up to 50 miles) from the host platform.
Thus:
TDH = p*g*h + ∆Pfric + Psurf
Assuming the flowing wellbore pressure at the pump inlet is essentially zero i.e. well
is "pumped off" and is producing with a maximum drawdown.
Department of Petroleum Engineering, Heriot-Watt University
33
The data describing the performance (see Section 2.7.A) of ESP’s provided by the
manufacture is normally measured with water. They also supply a correction factor,
based on the actual density and viscosity, when other fluids are being pumped. Further
correction is required if significant volumes of free gas are being injected by the pump
- not only will its volume decrease as the pressure increases, but it may also dissolve
completely in the oil. One popular application area of ESP’s is the production of
viscous crude oils at high water cuts. The design process can be simplified here since
the density of the crude oil is similar to that of water, there is little gas and the produced
fluid stream has an external water phase, i.e. the manufacturer’s performance curves
based on pumping water can only be applied directly.
Once the pump has been chosen, optimum motor and seal section can be identified,
along with the electric cable, variable speed drive, etc. It also needs to be checked that
the chosen combination will operate efficiently for a variety of well conditions
(higher/lower well PI, greater water cut, lower reservoir pressure etc).
The choice of correct ESP design, along with the actual, operational installation of
ESP’s, is thus a complex task. This results in the average run lifetime, or “mean time
before failure (MTBF)” often being very low initially when EPS's are first introduced
into a field / producing area (Especially when experienced staff are not available). The
MTBF then increases as the “learning curve” is climbed. This is illustrated in Figure
23 prepared from data presented at IIR’s conference on Artificial Lift Equipment,
Dubai, 1997, and IBC’s Artificial Lift Workshop, Aberdeen, 1997. This figure shows
how the THUMS project at Long Beach, California, has been employing ESP’s since
1965 and how the average run lifetime gradually increased throughout the 16 year
history. During this period, the numbers of operational ESP’s remained constant at
about 600. ESP’s were only introduced into the North Sea in the late 1980’s. The
initial run times were low compared to those achieved at THUMS, but a steep learning
curve developed and by 1996 the average run lives had become similar.
1200
1100
MTBF North Sea, Days
1000
MTBF at THUMS, Days
Long term trends
900
800
a
Se
r th
s
um
600
Th
No
MTBF Days
700
500
400
300
200
100
0
34
1983 1984
1985 1986
1987
1988
1989 1990 1991
YEAR
1992 1993
1994
1995
1996
1997
Figure 23
Average electric
submersible pump lifetimes
2
Selection of Artificial Lift Types
It need to be mentioned here that average statistics need to be treated with caution the actual run times will be very dependent on the aggressiveness of the producing
conditions e.g. temperature, concentration of sand produced, corrosivity of the
produced fluid (H2S, CO2, etc), skill of the rig crew, manufacturing quality control etc.
This is illustrated by a more detailed study of the THUMS data:
(i)
Oil is produced from three horizons where the average run lifetimes are 300,
700 and 800 days - indicating the importance of the role played by the actual
producing conditions.
(ii) 10% of the ESP’s fail within 30 days and 32% within 180 days, confirming the
potential for damage during installation and the need for the highest manufacturing
quality control standards.
7.4. Advantages and Disadvantages of Electric Submersible Pumps
Tables 5 and 6 respectively (Advantages and Disadvantages of ESP’s) have been put
together based on the discussion in Sections 2.7.1 and 2.7.2. The points discussed in
these tables should be self-explanatory when read in conjunction with previous
sections.
Advantages of Electric Submersible Pumps
Can be installed in deviated wells (<80…)
High production rates
Suitable for high water cut wells
Controllable production rate
Efficient Energy usage (>50% possible)
Access below ESP via "Y" tool
Comprehensive downhole measurements available
Can pump against high Flowing-Tubing Head Pressure
No extra flow lines required
Minimum surface footprint - 6ft well spacing
Low surface profile for Urban and offshore environments
Quick restart after shut down
Table 5
Advantages of electric
submersible pumps
Concurrent drilling and production safer compared to gas lift
(high pressure gas not present in annulus)
Long run pump life possible
Disadvantages of Electric Submersible Pumps
Susceptible to damage during completion installation
Tubing has to be pulled to replace pump
Not suitable for low volume wells (<150bpd)
Pump susceptible to damage by produced solids (sand / scale / asphaltene)
Table 6
Disadvantages of electric
submersible pumps
High GOR’s presents gas handling problems
Power cable requires penetration of well head and packer integrity
Viscous crude reduces pump efficiency
(Viscous) emulsions form over a range of water / oil ratios
High temperatures can degrade the electrical motors
Department of Petroleum Engineering, Heriot-Watt University
35
7.5. Monitoring the Performance of Electric Submersible Pumps
Prior to the implementation of automated SCADA (Supervisory Control and Data
Acquisition) systems, the monitoring of ESP performance was limited to a surface
measurement of the current supplied to the pump along with an infrequent (possibly
monthly) well test. However, considerable information can be derived on well
performance - Figure 24(a) is a schematic example of a 24 hour chart recording of
electrical current consumed by an electric submersible pump during normal operation.
The current taken is very constant.
5PM
6PM
4PM
7P
M
70
M
3P
2P
M
40
50
40
2
1 0
30
1 0P
1PM
50
30
20
10
30
40
2
10 0
M
DING AM 0
COR
ME
RE
T
ER
50
40
70
60
30
20
10
7.30 AM
40
2
10 0
30
20
10
60
30
40
50
40
70
60
M
70
2A
50
M
3A
M
9A
50
M
1 0A
30
1AM
2
10 0
7.30 AM
1 MIDNIGHT
70
60
TIME
DATE ON Day two TIME
50
DATE ON Day one
40
30
20
10
MIDDAY
M
9P
60
70
M
70
50
60
11AM
8P
60
60
M
4A
8A
70
5A M
M
7AM
6AM
By contrast, the current taken when the well is being pumped off shows a much more
erratic behaviour (Figure 24b).
36
Figure 24(a)
Ammeter chart monitors
electric submersible pump
performance. Normal
operation.
2
Selection of Artificial Lift Types
5PM
6PM
4PM
7P
M
70
M
3P
2P
M
40
50
40
2
10 0
30
10P
1PM
50
30
20
10
30
40
2
10 0
M
DING AM
COR
ME
RE
T
ER
50
40
70
60
30
20
10
7.30 AM
40
2
10 0
30
20
10
30
40
50
60
40
70
70
M
50
M
3A
M
9A
60
2A
60
M
4A
8A
70
5A M
M
7AM
6AM
Figure 24(b)
Ammeter chart monitors
electric submersible pump
performance. Well pumped
off.
50
M
10A
30
1AM
2
10 0
7.30 AM
1 MIDNIGHT
70
TIME
60
Day one
DATE ON Day two TIME
50
DATE ON
40
30
20
10
MIDDAY
M
9P
60
70
M
70
50
60
11AM
8P
60
The chart was installed at 07.30 am and the pump started at 08.15 am - note the large,
initial surge in current while the motor is getting “up to speed”. A steady current is
then drawn for the next 3 hours - decreasing slightly as the fluid head above the pump
decreases. At 11.10 the current begins to oscillate rapidly - the size of these
oscillations increases until 1.15 pm when the pump was shut down. It was suspected
that the problem was due to gas being formed when the flowing bottom hole pressure
was reduced to below the bubble point, leading to gas locking and the pump ceasing
to pump. This was confirmed by leaving the fluid level in the well to build up for 100
minutes and restarting the pump at 3.05 pm. The same cycle repeats itself, however
this time the problems appear after some 2.5 hours steady production. The pump was
shut down a second time at 6.15 pm. A third cycle was started at 8.20 pm after a second
100 minute shut in - current oscillation starting again after 2 hours production. The
well was shut in just after midnight.
The basic problem is that the pump is pumping faster than fluid is flowing into the well.
Continual stopping and restarting the ESP motor is not recommended due to excessive
wear and tear followed by motor burnout. The options are to:
• Install a lower capacity (smaller) pump section.
• Operate the pump at a lower speed.
• Stimulate the well to improve the inflow.
Department of Petroleum Engineering, Heriot-Watt University
37
Modern SCADA systems allow a much more complete picture to be built up.
Installation of a downhole monitoring package (as shown in Figure 17) allow the
motor/pump conditions to be monitored closely. Figure 25 is an example of a normal
start up. It has been analysed as follows:
A
Pump Shut
Down
B
Normal Steady Operation
Surface Choke
Adjustment
Pump Discharge Pressure
Pump Suction Pressure
Motor Temperature
Pump
Start Up
Vibration
Fluid Temperature at Pump Intake
Current Leakage to Earth
Time
•
•
Initially, prior to energising the pump, the pump intake and discharge pressures
have the same value; as does the motor and pump intake pressures. The pump
starts up at point A, as shown by:
(i)
the pump discharge pressure increasing,
(ii)
the motor temperature becoming warmer than the fluid entering the
pump,
(iii)
a limited amount of vibration,
There then follows a period of surface choke adjustment, as shown by
fluctuations in the pump discharge pressure and increased vibration. However,
after point B, steady operating conditions are achieved and a slow decline in
pump suction and discharge pressure are observed as the well is pumped “off”.
Protection of the ESP can now be achieved by monitoring the pump’s condition and
shutting it down when problems develop before physical damage to the pump results.
Thus “pump off” control can be implemented by stopping the pump when the intake
pressure drops below a preset value. The pump is then restarted once the well pressure
builds up to a second, higher predetermined value. This type of monitoring is very
diagnostic when problems develop, but can have much more “added value” when
combined with surface flow measurements (gross flow rates and water cut). However,
continually restarting pump motors reduces their operational life. Installing a correctly
sized pump unit is the preferred solution.
38
Figure 25
Electric submersible pump
and motor condition
monitoring
2
Selection of Artificial Lift Types
7.6. New Technology
7.6.1. Coiled Tubing Deployed ESP’s
The completion designs discussed so far within this section on ESP’s all employ the
pump installed as part of a conventional completion string with the power cable
attached to the outside of the tubing. Replacement of any part of the ESP following
a failure requires a workover. The process of ESP installation and recovery can be
speeded up and made more efficient by installing the ESP at the end of a coiled tubing
(see Figure 26). The set up is “conventional” in the sense that the cable is mounted
on the outside of the coiled tubing while the produced fluids flow to the surface via
the inside of the coiled tubing. A dual packer arrangement is required with ESP
arrangement shown. The produced fluid is pumped into the casing via the annulus,
flows passed the seal and electric motor sections (cooling) and then back into the
coiled tubing via a cross-over mounted below the upper packer.
Coil Tubing Reel
Power Cable
(Mounted on Reel)
Power Cable Clamped
to Coiled Tubing
(Cable Cannot Support it’s Own Weight)
Production Via Coiled Tubing
Upper Packer
Crossover
Motor
Seal
Pump Discharge
Figure 26
Schematic view showing
installation of a coiled
tubing deployed ESP
Sensor Package
Pump Intake
Lower Packer
Fluid Inflow
Replacement of an ESP producing a shallow, depleted horizon in a land well requires
a light workover hoist. A coiled tubing deployed ESP can speed up the process, but
the advantages become much greater when access to the well site is limited e.g. when
for an offshore well located in a small platform. The (limited) weight requirements
of a coiled tubing package often allow its installation on the platform using the
platform crane; while making a conventional (jackup) rig available is a much more
time consuming, expensive operation.
An alternative, simpler arrangement is illustrated in figure 27 in which the ESP cable
is installed within the coiled tubing and the production travels to surface via the
annulus. This arrangement has the advantages that:
Department of Petroleum Engineering, Heriot-Watt University
39
(i)
Reduced frictional pressure losses lead to higher flow rate or reduced power
requirements.
(ii) Faster running can be achieved with the cable inside the protected environment
of the coiled tubing.
(iii) It opens up the possibility of installing the ESP in a live well (well killing is a
major source of production loss due to formation impairment).
Coiled Tubing
Internal Power Cable
Shear Sub
Electrical Penetrator
Electric Motor
Seal or Protector
Pump Discharge
(Production to Surface
Via Annulus)
Pump
Pump Intake Stinger
Packer
Production Casing / liner
or Production Tubing
Fluid Flow
The disadvantage to the system is that the production takes place via the Production
Casing/Coiled Tubing Annulus, which raises a number of safety issues concerning
barrier policy and corrosion. Alternatively, a wider diameter well can be drilled and
a (large diameter) production tubing installed.
7.6.2. Auto “Y” Tool
The advantage of installing a “Y” tool to allow access below the pump was described
earlier. However, wireline/coiled tubing recovery and replacement of the plug in the
bypass tube are required each time the conventional “Y” tool is used. The cost and risk
associated with these two wireline operations can be avoided by use of the Auto “Y”
tool developed by Phoenix Petroleum Services of Aberdeen. It’s operation is
illustrated by figure 28:
40
Figure 27
Alternative design for
coiled tubing deployed ESP
featuring annular
production
2
Selection of Artificial Lift Types
(a) A spring holds the diverter plate across the pump leg when the ESP is switched
off - full access is now provided to the bypass.
(b) Flow generated by the ESP starting up moves the diverter plate and opens the
pump leg.
(c) The diverter plate is seated onto the bypass leg by the high pressure in the
production tubing during normal ESP operation.
(d) The diverter plate can be held in the mid position if it is desired to operate the
ESP during a logging run.
Coiled Tubing
or wireline
Figure 28
Auto "Y" tool removes need
for pulling / rerunning
plugs
Pump
Leg
Bypass
Leg
High
Pressure
(a) Diverter blocks
pump leg when
ESP shut down
full access
to bypass
Flow
(b) Diverter moves
across due to
fluid circulation
via bypass when
ESP starts up
Flow
Lower
pressure
in bypass
leg
(c) Diverter continues to
close bypass leg when
pump operational due
to greater pressure
in production tubing
Flow
(d) Diverter held open
during CT / wireline
logging run with pump
operating
7.6.3. Dual Pump Installations
More than one Electric Submersible Pump can be installed for a number of reasons:
(i)
greater power installed downhole than can be (economically) achieved with a
single ESP. The ESPs are placed such that the discharge of the lower ESP forms
the suction of the upper ESP.
(ii) Dual Zone Completion. Figure 29 shows a dual completion with each zone
having its own ESP and production tubing. The production tubing for the upper
zone is installed concentrically within that for the Lower Zone.
Department of Petroleum Engineering, Heriot-Watt University
41
Production Tubing
(Upper Zone)
Dual Cable Clamp
Production Tubing
(Lower Zone)
Install lower
zone downhole
flow meter here
Crossover
Pump Discharge
Bypass Tubing
Upper ESP
Sensor Package
Pump Support
Production Tubing
Production Tubing
Cable penetrator
Install lower
zone downhole
flow meter here
Pump Discharge
Lower ESP
Sensor Package
Production Tubing
Stinger
Upper Production Zone
Packer
Polished Bore
Casing or Liner
Perforations
Lower Production Zone
(iii) the completion could have been simplified by use of a single production tubing
and allocating production between the Upper and Lower zones by installing
downhole flow meters above the pumps (see figure 29).
(iv) remote, offshore locations have very high cost associated with the frequent
replacement of ESPs if the run lives are short e.g. due to excessive sand
production. The concentration of sand in the produced fluid is often particularly
high after a workover. Completion designs have been developed with a
“sacrificial” lower ESP - see figure 30. The operational concept is as follows:
42
Figure 29
Dual completion with
separate ESP's for each
production string. ESP's
discharge into dual
concentric production
tubing
2
Selection of Artificial Lift Types
Dual Cable Clamp
Auto "Y" Tool
Upper ESP
Sensor Package
ESP Support
Auto "Y" Tool
Lower ESP
Figure 30
Dual pump installation with
"sacrificial" lower ESP
Sensor Package
(a) The lower ESP is started and the well produced until the sand production
decreases to an acceptable level. The upper ESP may now be started to further
increase the well production. The upper ESP can be operated independently of
the lower ESP in the case that the lower unit fails due to erosion by produced
sand or other reasons.
(b) The auto “Y” tool allows production to be switched between the two ESPs
without wireline or coiled tubing intervention. However, it must be checked
that fluid is not being circulated around the pump, since this will lead to rapid
pump failure (motor burnout due to overheating).
7.6.4. Reducing Water Production
The level of water production is an ever increasing problem as reservoirs mature.
Department of Petroleum Engineering, Heriot-Watt University
43
Hydrocyclones (see section 9.25) have become the preferred technique to separate the
produced oil and water. Their dimensions and lack of moving parts make them highly
suitable for installing downhole (see also chapter 10). Figure 31 illustrates one
equipment design. This shows a single electric motor powering an upper and lower
pump unit. The equipment is installed below the producing zone. The lower pump
unit supplies sufficient power to operate the hydrocyclone and to inject the underflow
(water containing approximately 100ppm oil) into the water injection zone. The
hydrocyclone overflow (typically 50% oil) is transferred via a bypass tube to the upper
pump which pumps it to surface. The maximum production rate depends on the casing
size and the number of hydrocyclones installed in parallel (maximum capacity of a
hydrocyclone is 2,500 bfpd). Rates up to 20,000 bfd for a 9.675 in. casing are feasible.
Oil Concentrate
to Surface (50% Water)
Upper Pump Unit
Oil Concentrate Bypass
Producing Zone
(Water Cut > 85%)
Electric Motor
Pump Inlet
Lower Pump Unit
Hydrocyclone Separator
Water (± 1000 ppm oil)
Water Injection
Zone
This technology has been extended to meet the challenges of:
(i)
Installing two separators in series. This 2 stage separation allows downhole
water separation to be started at the lower water cut from the production zone
of 65%. A minimum produced water cut of 80% is required to achieve efficient
operation (acceptably low oil concentration in the rejected water stream) with
a single hydrocyclone stage.
(ii) Produced Sand. Any sand particles produced will be separated with the water
flow to be injected due to its greater density. Blockage of the injection zone will
occur rapidly if significant volumes of sand are being produced. This can be
avoided by treating the injection water stream with a hydrocyclone designed to
concentrate the produced sand particles in the underflow (see figure 32). This
(small) underflow stream is added to the oil concentrate and produced to
surface while the bulk of the water stream is injected as normal.
44
Figure 31
Downhole water separation
driven by an electric
submersible pump
2
Selection of Artificial Lift Types
Oil Concentrate
and Sand Particles
to Surface
OVERFLOW:
Oil Concentrate
High Water
Cut Production
From Production Zone
Hydrocyclone
(Oil
Concentrator)
Upper Pump Unit
Lower Pump Unit
To Injection Zone
UNDERFLOW:
Water, ±1000 ppm Oil
and Sand Particles
Figure 32
Flow diagram for downhole
separation with sand
rejection hydrocyclone
OVERFLOW:
Water, 100 ppm Oil
Hydrocyclone
(Sand
Rejection)
UNDERFLOW:
Some Water and
Sand Particles
(iii) Coning Suppression. This concept is illustrated in figure 33. Production from
the oil zone perforations results in a water cone being formed once the critical
oil production rate has been exceeded. This critical rate can be increased by
producing water from below the oil/water contact via “coning suppression
perforations” and injecting the water into an injection zone using an inverted ESP.
Oil production to surface
Oil Zone Perforations
Oil / Water Contact
"Anti Coning"
perforations
"Normal" Water Cone
Due to high water production
at "Oil Zone" perforations
Supressed Water Cone
Due to water production
at "Anti Coning" perforations
Electric Submersible
Pump
Figure 33
Suppresed coning and in
situ water disposal
Water Injection Zone
This concept has been shown to work - but alternative technologies such as
horizontal production wells will often be more attractive economically.
(iv) Managed Water Injection. The water injection zones illustrated above were
simple perforated completions. Advanced well concepts in which the water is
injected into a long (near) horizontal lateral split into a number of zones, where
the volume of injected water will be regulated by an electrically adjustable
choke, will become feasible in the next few years.
Department of Petroleum Engineering, Heriot-Watt University
45
7.7. Electric Submersible Pump Performance
The centrifugal pump unit employed in ESP’s is a dynamic-displacement pump in
which the pump rate depends in the pressure head generated - the pump rate is low
when the pressure head is high and vice versa. This is different from the positive
displacement pumps discussed earlier in which the pump rate and discharge pressure
are independent of one another.
The relationship between pump rate and pressure generated for dynamic displacement
pumps is called the pump characteristic (see Figure 34). It is measured by the pump
manufacturer in laboratory tests using a standard fluid (water) with the pump running
at 3500 rpm (60 Hz electrical supply) or 2915 rpm (50 Hz supply).
84
Maximum Efficiency
72
Pump Head (ft. water)
Pump Head
60
Pump Efficiency
48
70
60
50
40
36
24
.60
30
.40
20
.20
10
Pump Only Efficiency
Hydraulic Horse Power / Stage
Recommended Pump
Operating Range
Motor Power
12
300
600
900
1200
1500
1800
2100
2400
2700
Pump Rate (B/day)
3
Measured at 50 Hz, motor speed of 2915 rpm and a fluid viscosity of 1 cp and density 1g/cm
The “pump head”, or increase in pressure per stage (∆P), is expressed in terms of the
pressure generated by an equivalent column of water (Hwater). It decreases as the pump
rate increases:
∆P = ρ*g*h = 0.433*γ* Hwater
The discharge pressure is proportional to the specific gravity (γ) for other liquids with
the same viscosity, i.e. the pump head - pump rate relationship can be used for all
liquids, but only requiring correction for changes on viscosity.
Pump power is the workdone per unit time which equals the pump rate multiplied by
the pump head (q*∆P).
Power is normally expressed in terms of Kilowatt (KW) or Horse Power (HP); where
1 HP = 0.746 KW. The pump or hydraulic power is the (useful) work done by the pump
46
Figure 34
A typical pump
characteristic curve for a
centrifugal pump
2
Selection of Artificial Lift Types
while the mechanical power is the work done by the electrical motor which is required
to drive the pump. The pump efficiency (E) is thus:
E = hydraulic power/mechanical power
Pump Efficiency is also recorded in the pump characteristic curve and a recommended
pump operating range indicated based on ±10% of the maximum efficiency point.
7.7.1. Simplified Electric Submersible Pump Design
ESP design is available as an option in many of the commercially available well
performance programs, e.g. WellfloTM. The simplified, manual procedure outlined
below to evaluate the installation of an ESP into the vertical well Edinburgh-1 follows
the same basic steps as the more complex, computerised, design procedures.
Table 7 summarises the Edinburgh-1 well conditions.
Well Edinburgh 1
Table 7
Data for Well Edinburgh 1
Depth (h)
7000ft
Reservoir pressure (Pr)
1700psi
Well productivity Index (PI)
2 STB / day / psi
Tubing Internal Diameter (d)
2.26 in or 0.188ft
Surface manifold Pressure (Pm)
50 psi
Design Well Production (Q)
1400 STB / day
Produced Fluid properties
Water
Fluid Density
0.433 psi / ft
Viscosity
1cp
Pump set at same depth as perforations
(i)
The pipe friction loss (∆Pf) at the desired well production is given by:
L   v2 

∆Pf = (f ) * (ρf ) *
*
 d   2g 
where fm is the moody friction factor, v is the fluid velocity and g the
acceleration due to gravity {32.173 (ft/s2) (lbm/lbf)}. Now:
v=
=
Q(STB / day) * 5.615(ft 3 / bbl)
86, 400(s / day) * ( πd 2 / 4)(ft 2 )
1400 * 5.615 * 4
= 3.28ft / s
86, 400 * 3.14 * 0.1882
As discussed previously, the value of fm, a function of Reynolds Number, pipe
roughness and the fluids’ properties, can be found from a Moody Diagram. It has
a value of 0.03 for this calculation.
Department of Petroleum Engineering, Heriot-Watt University
47
7000  
∆Pf = (0.03) * (0.433) * 
*
 0.188  
3.282 
 = 81psi
2* 32.2 
(ii) Pd = Ps + ∆Pf + ∆PHH
where Pd is the required pump discharge pressure and ∆PHH is the hydrostatic
head due to the 7000 ft column of fluid. Ps is the wellhead pressure required to
transfer the fluid to the surface facilities (50 psi).
Pd = 50 + 81 + (0.433 psi/ft) * (7000 ft) = 3162 psi
(iii) The Flowing Bottom Hole Pressure and the pump intake pressure (PIn) are the
same and can be calculated from:
PIn = Pr - Q/PI = 1700 - 1400/2 = 1000 psi
N.B. It is essential that {P - PIN} > 50 psi to ensure select there is a minimum height
of fluid above the pump section so that it doesn't "run dry".
(iv) Using the pump performance chart shown in Figure 34, the head per stage (H)
at 1400 b/d is 58 ft and the hydraulic horsepower per stage (HHP) is 0.52.
The number of pump stages (N) and the minimum electric motor power (HHP)
required can now be calculated for a pump running at 2915 rpm.
N=
and
(Pd − Pin )
( psi)
H(ft ) * γ * 0.433 ( psi / ft )
=
(3162 − 1000)
58 * γ* 0.433
= 86 stages
HHP = 86 (stages)*0.52 (HHP/stage)*(γf) = 45 HP
where γf is the specific gravity of the fluid (unity in our case)
(v) An electric motor to power the pump may now be chosen (minimum 205 HP
and 50Hz).
N.B. Choosing a pump speed other than 2915 rpm introduces extra complications
since the pump rate of an ESP is proportional to the speed
i.e.
pump rate2 pump speed2
=
pump rate1 pump speed1
where (1) denotes the initial rate (2915 rpm) and (2) refers to the new speed of
the motor (and pump, since ESP’s do not have a gearbox). Further, the motor
speed also controls the hydrostatic head produced.
hydrostatic head2  pump speed2 
=

hydrostatic head1  pump speed1 
48
2
2
Selection of Artificial Lift Types
The power required may now be calculated
motor power 2  pump speed 2 
=

motor power 1  pump speed 1 
3
Variable Frequency Drive (VFD) provides the ability to change the pump and
electric motor speed by altering the frequency of the electricity supply. The pump
characteristic performance curves are also measured by the manufacturer for a range
of conditions and are reported in their data books - Figure 35 shows the format of a
typical example (not the same one as discussed above).
21
90 Hertz
Pump Head (m. of water)
18
15
3.07
12
70 Hertz
2.16
9
60 Hertz
6
Figure 35
Typical changes in pump /
motor characteristic
performance as a function
of electric supply frequency
(Hydraulic Power) / Stage at
Maximum Efficiency
80 Hertz
1.45
50 Hertz
.91
.53
40 Hertz
3
.27
30 Hertz
Operating Range
15
30
.11
45
60
75
90
105
120
135
3
Pump Rate (m /day)
(vi) The final stage in this simplified design procedure is to evaluate the robustness
of the design for a series of well inflow conditions i.e. changes in well productivity
index or reservoir pressure. These are performed by carrying out a nodal
analysis on the ESP pump. Figure 36 is a typical example of such an analysis
(also not the same one as discussed above).
Department of Petroleum Engineering, Heriot-Watt University
49
1700
1500
PI = 4 STB/psi
C ur
p In
t ake
1000
Pum
Pressure at ESP Inlet (psi)
ve
1250
750
PI = 2 STB/psi
PI = 0.5 STB/psi
500
250
PI = 1.0 STB/psi
0
500
750
1000
1250
1500
1750
2000
Well Production Rate (STB / day)
This figure shows that:
(a) The well fluid level above the pump (1000 psi or 2310 ft TVD) is high if the
well’s PI was 2 STB/d/psi and 1400 STB/day were being produced i.e. the well
is not being “pumped off” and a larger pump could have been installed. The
production rises to 1540 STB/d (and the fluid level to 1315 psi or 3035 ft
TVD) if the well productivity index increased to 4 STB/d/psi.
(b) The well production reduced to 1190 STB/d for the lower well productivity
index of 1 STB/day. The lowest well PI plotted (0.5 STB/d/psi) results in a
negative well inflow pressure. The well has now been “pumped off” - an
unacceptable situation which would be corrected by restricting the tubing
outflow with a choke. The minimum well inflow pressure will be dictated by
the minimum pump charging pressure required (depends on pump design), gas
interference e.g. bubble point etc.
(vii) Cable selection - which depends on pump power, voltage selected and downhole
temperature, may now be made.
(viii)Further details can be found in API RP 1154 - “Recommended Practice for
Sizing and Selection of Electric Submersible Pump Installations”.
The ESP manufacturers can supply software to carry out a more sophisticated design
analysis than that described here. Further, many of the well design of nodal analysis
packages included data from the pump manufacturers so that the well analysis and
selection process can be automated.
50
Figure 36
Well flow / pump out flow
(nodal) performance curve
2
Selection of Artificial Lift Types
8. HYDRAULIC PUMPS
Hydraulic pumps use a high pressure power fluid pumped from the surface (Figure 37)
which:
Manifold Values
Controls power fluid supply
rate to each downhole engine.
Pressurising
Pump
Desanding
Hydrocyclone
Filter
To Well, 2, 3 etc.
Power Fluid
(Usually Produced
Water)
Sand
Commingled Power Fluid
and Produced Fluid to
Separation
Power Fluid
Commingled
Exhaust Power
Fluid and
Produced Fluid
Downhole Engine
(Turbine, Positive
Displacement Type
or Venturi)
Pump Discharge
Pump
Figure 37
Principals of hydraulic lift
operation. Turbine pump is
illustrated.
Produced Fluid
(i)
drives a downhole, positive displacement pump. Figure 38 shown how the flow
of power fluid through the upper engine unit is translated into a flow of high
pressure produced fluid during both the “UP” and “DOWN” strokes.
Department of Petroleum Engineering, Heriot-Watt University
51
Engine and pump pistons
move downwards during
Down Stroke
Engine and pump pistons
move upwards during
Up Stroke
Power Fluid
Power Fluid
Engine Exhaust
Engine Exhaust
Engine Unit
Engine Piston
Piston Connecting
Rod
High Pressure
Produced Fluid
Valve Closed
Valve open
Valve Open
(Produced Fluid
Intake)
Pump Piston
High Pressure
Produced Fluid
Pump Unit
Valve Open
Valve Closed
Produced Fluid
High Pressure
Power Fluid
Produced Fluid
Exhaust
Power Fluid
Produced
Fluid
High Pressure
Produced Fluid
(ii) powers a centrifugal or turbine pump (see Section 2.8.3).
(iii) creates a reduced pressure by passage through a venturi or nozzle (Figure 39)
where pressure energy is converted into velocity. This high velocity/low
pressure flow of the power fluid commingles with the production flow in the
throat of the pump. A diffuser then reduces the velocity, increasing the fluid
pressure and allowing the combined fluids to flow to surface.
52
Figure 38
Operation of positive
displacement hydraulic
pump
2
Selection of Artificial Lift Types
Power Fluid
Supplied
by Inner
Concentric
Tubing
Inner Tubing
Fishing Neck
for Pump Recovery
Vented Gas
Spring Holds
Nozzle in Place
Venturi
Nozzle
Produced Fluid
Throat
Well Casing
Diffuser
Commingled
Produced Fluid
and Exhaust
Power Fluid
Outer Tubing
Vented Gas
Standing Valve
(Ball and Seat Valve)
Tubing Packer
Figure 39
Jet or Venturi pump
operation
Produced Fluid
High pressure
Power Fluid
Produced Fluid
Commingled
Power and
Produced Fluid
Vented Gas
The power fluid consists of oil or production water (the large oil inventory in the
surface power fluid system makes oil accounting difficult once high water cuts are
being produced). The power fluid is supplied to the downhole equipment via a
separate injection tubing. The majority of installations commingle the exhaust fluid
with the production fluid {an “open system” Figure 40(a)}. If difficulties or high costs
are encountered in preparing power fluid of the required quality from the production
fluid, then a “closed system” may be installed in which the power fluid returns to the
surface via a (third) separate tubing {Figure 40(b)}. This option is not available with
a venturi pump. The completion design may also allow gas to be vented to surface via
the casing/tubing annulus.
Department of Petroleum Engineering, Heriot-Watt University
53
Produced Fluid
Power Fluid
Power Fluid
Gas
Gas Vent
to Annulus
Produced
Fluid Level
Produced Fluid
Engine
High Pressure
Power Fluid
Commingled
Power and
Produced Fluid
Pump
Commingled
Produced
and
Power Fluid
High Pressure
Produced Fluid
Standing
Valve
Exhaust
Power Fluid
Produced Fluid
Produced Fluid
(a) Open Power Fluid System
(b) Closed Power Fluid System
A typical power fluid supply pressure of between 1,500 and 4,000 psi. is provided by
a pressurising pump (Figure 37). This may be a reciprocating plunger (triplex) pump
or a multi-stage, centrifugal pump. This pressure determines the pressure increase
achievable by the downhole (positive displacement or centrifugal) pump. The pump
rate (and the rate at which power fluid has to be supplied) is determined by the diameter
and speed of the downhole pump.
“Clean” power fluid is required to avoid erosion of the downhole pump components.
The power fluid is often drawn from a settling tank where the larger solids are
removed. It is then pumped via a desanding hydrocyclone and a guard filter before
having its pressure raised to the operating pressure by the charge pump. The power
fluid from the pressurising pump may supply one or more wells (Figure 37).
8.1. Advantages of Hydraulic Pumps
Hydraulic pumps have the following advantages:
(i)
Suitable for crooked and deviated wells.
(ii) Reciprocating and turbine pumps can work at great depths (up to 17,000 ft),
while jet pumps are restricted to about half that value.
54
Figure 40
Types of hydraulic pump
installation
2
Selection of Artificial Lift Types
(iii) Very flexible speed control by the (surface) supply of power fluid. Turndown
to <20% of design maximum speed can be achieved.
(iv) Jet pumps, with no moving parts can handle solids. Weir pumps are
claimed to be manufactured from erosion resistant materials which aim to
give a 5 year lifetime with “reasonable” solids production associated with
a prepacked screen completion for installation in a soft formation.
(v) The power source is remote from the wellhead giving a low wellhead profile,
attractive for offshore and urban locations.
(vi) The power fluid can carry corrosion or other inhibitors downhole, providing
continuous inhibition when the well is producing.
(vii) The pump unit can be designed as a “free” pump; the pump unit having the
capability of being pumped through the power fluid tubing from the surface to
its downhole location (figure 41). It can then be recovered by reversing the
flow direction. The ability to recover the pump without the need to move a rig/
workover hoist to the wellsite is attractive for offshore platforms as well as
remote and urban locations.
Install Pump
Power
Fluid
Production
Production
Flow Line
Power
Fluid
Pump Recovery
Commingled
Power and
Produced Fluid
Power
Fluid
"Free" Hydraulic Pump
Figure 41
Installation and recovery of
a "free" hydraulic pump
Standing Valve
Closed
Standing Valve
Open
Department of Petroleum Engineering, Heriot-Watt University
Standing Valve
Closed
55
8.2. Disadvantages of Hydraulic Pumps
(i)
Pumps with moving parts have a short run life when supplied with poor quality
(solids containing) power fluid. Jet pumps can have a long run life under similar
conditions.
(ii) Positive displacement and centrifugal pumps can achieve very low flowing
bottom hole pressures in the absence of a gas effect. The lowest pressure
achievable by jet pumps are much higher, being comparable to gas lift.
8.3. New Technology (Weir Pumps)
A recent innovation that has been field tested over the last few years is the Weir Pump
- a hydraulically driven engine coupled to a turbine pump with improved gas handling
abilities. It uses an open power fluid circuit - the power fluid is returned to the surface
commingled with the production. Its performance characteristics are very similar to
that of an ESP - but it also has the ability to achieve stable operation over a wide range
of flow rates with gas fractions of at least 80%. In addition, large gas slugs associated
with surging flow can be handled without mechanical damage to the pump or its motor.
Hydraulic (and Weir) pumps have several intrinsic advantages over ESP’s:
(i)
The continuous supply of cool power fluid increases the maximum allowable
bottom hole formation temperature at which the pump can be installed.
(ii) Solids free power fluid lubricates the pump bearings, enhancing their produced
solids handling capabilities (typical power fluid specifications are 100 ppm
solids with a maximum diameter of 0.1 mm). The need (and cost) of continual
cleaning of the power fluid in a (multi-well) subsea development could be
minimised by use of a “closed” power loop.
(iii) There is no requirement for a mechanical seal to be installed between the motor
and the pump, as is the case for an ESP.
(iv) Variation in the power fluid flow rate provides pump speed control as well as
a “soft start” capability.
(v) The pump set operates at high speeds. This results in a short (3-4 m) pump unit
since:
Number of pump stages required α {1 / (pump speed)2}
This short length (and weights typically less than 500 kg) allow wireline retrieval of
the pump unit in wells at deviation angles of up to 55º. Coiled tubing installation is
required for higher deviation angles (up to 80º).
Figure 42 illustrates the completion design used in the Texaco field trial in their North
Sea “Captain” field. This included:
56
2
Selection of Artificial Lift Types
Actuated Choke
Produced Fluid
Actuated Choke
Power Fluid
Power Fluid Temperature and Pressure Measurements
Flow Rate (Venturi), and Outlet Temperature and Pressure
Measurement
Pump Unit Speed and Vibration Measurement
Inlet Temperature and Pressure Measurement
7" Tubing
Upper "Y" Tool
Packer With Polished Bore
Figure 42
A hydraulic (Weir) pump
completion with down hole
monitoring
2 3/8" By-pass Tube
6 5/8" Prepacked Wirewrapped Screens
10,000 BPD Hydraulic Submersible Pump Assembly
Lower "Y" Tool
9 5/8" Casing Shoe
5 1/2" Surface Controlled Subsurface Safety Valve
(i)
Fluid Loss Control Valve
Installation of an upper and lower “Y” tool connected by a 2.375" by-pass
tubing. This tubing allows access (e.g. for production logging purposes) to the
producing interval without having to recover the pump. A Surface Controlled,
Sub Surface Safety Valve has been installed below the lower Y tool. Also, a
“Fluid Loss Control” valve is installed at the bottom of the tubing to prevent
injection of workover fluid into the completion during pump recovery or other
operations being carried out above the packer.
(ii) Comprehensive flow rate, temperature and pressure measurement as well as
pump performance monitoring, such as pump speed and vibration measurements,
has also be installed.
(iii) As discussed previously, the development of the viscous oil rim in Texaco’s
“Captain” field depended on developing a reliable artificial lift system having
a combination of effective gas handling and viscous fluid pumping capabilities.
It was concluded after a one year’s field trial that:
(a)
The viscous crude oil, and any resulting high viscosity emulsions, was
pumped at flow rates ranging from 20% - 120% of design with water
cuts varying from 0 - 100%.
(b)
Fluid gas fractions of 30% to 75% at the pump suction were routinely
pumped while exceptional, slug flow conditions of greater than 90%
gas volume fraction, were managed by the fluid driven turbine pump.
(c)
The pump and turbine design was robust. It withstood the imposed loads
and was resistant to the typical solids production encountered in
completions requiring sand control.
Department of Petroleum Engineering, Heriot-Watt University
57
9. Progressing Cavity PUMPS
Progressing Cavity (or Moyno) Pumps are becoming increasingly popular for the
production of viscous crude oils. Figure 10 summarises the application area (well
rates & depths) where Progressing Cavity Pumps (PCP) are typically employed. A
typical completion is illustrated in Figure 43 where a prime mover (in this case an
electric motor) is shown rotating a sucker rod string and driving the PCP. This
section will describe the principle on which the pump operates, the resulting
advantages and disadvantages and, finally, takes a look at new technology.
Belt Driven
Speed Reducer
Electric Motor
Coupling connects
rods to drive shaft
Casing Vent
Wellhead
Rods
Tubing
Centraliser
Casing
Progressing Cavity Pump
Torque Anchor
Production Zone
58
Figure 43
A well completed with
artificial lift using a
progressing cavity pump
2
Selection of Artificial Lift Types
9.1. Progressing Cavity (or Moyno) Pump Principle
Figure 44 illustrates the main components of a PCP. A steel shaft rotor of diameter
d has been formed into a helix {Figure 44(a)}. The rotor is rotated inside an
elastomeric pump body or stator, which has been molded in the form of a double helix
with a pitch of the same diameter and exactly twice the length of the pitch given to the
rotor {Figure 44(b)}. Figure 44(c) shows that, when assembled, the centre line of the
rotor and the stator are slightly offset, creating a series if fluid filled cavities along the
length of the pump. Figure 45 is a perspective view of Figure 44(c), which helps
explain how the interference fit between the rotor and stator creates two chains of
spiral (fluid filled) cavities.
Centreline of
Rotor
d
Centreline of
Stator
Steel Rotor
Elastomeric Stator
Centreline of Stator
Rotor contacts
stator here
creating a
sealed cavity
Figure 44
Cross section progressing
cavity pump and its
components.
Fluid
Filled
Cavities
Pitch Stator
Pitch Rotor
Centreline of Rotor
ecc
d+2ecc
d+4ecc
(a) Assembled Pump
(b) Rotor Geometry
(c) Stator Geometry
d is minor diameter of rotor and slator, ecc is rotor eccentricity
Department of Petroleum Engineering, Heriot-Watt University
59
ecc
d
d+4ecc
Pitch of Stator
Elastometric Stator
Rotor contacts
stator here,
creating
sealed cavity
Pitch of Rotor
Fluid Filled Cavities
Steel Rotor
The rotor within the stator operates as a pump. This causes the fluid, trapped in the
sealed cavities, to progress along the length of the pump from the suction to the pump
discharge. These cavities change neither size nor shape during this progression.
Figure 46 (a-e) shows how, as one cavity diminishes, the next one increases at exactly
the same rate; giving a constant, non-pulsating flow. It acts as a positive displacement
pump. The pressure increase that can be achieved by the pump depends on the number
of “seal-lines” formed along the pump body by the rotor and stator. Typically, this is
found to be 300-200 kPa pressure increase per stage. It is found that fluid will “slip”
backwards if a greater pressure increase is demanded from the pump. This can be
avoided by increasing the number of pump stages. Wear of either the stator of rotor
will decrease this value since the the maximum pressure increase depends on this
interference fit. However, the construction of the stator body from an elastomer makes
this pump design relatively tolerant to produced solids - particularly since they are
often used to pump viscous oils which provides a lubrication film to protect the rotor
and stator from wear.
60
Figure 45
Perspective view of
operating Progressing
Cavity Pump
2
Selection of Artificial Lift Types
Centre Line
Cross Section at
Centre Line
Rotor at top side of cavity
Stator Body
90… Rotation
180… Rotation
Rotor is at bottom of cavity
270… Rotation
Figure 46
Operating principle of
Progressing Cavity Pump
360… Rotation
Rotor returned
to topside of cavity
The flow rate achieved by a liquid filled (no gas) PCP pump is directly proportional
to the speed of rotation of the rotor (N):
Flow rate = k*Pitch of Stator*4*eccentricity*Stator minor diameter*N
Where k is a constant.
The presence of gas reduces the efficiency and a gas anchor is frequently included in
PCP completions (see also Section 2.6.5.2 for discussion of gas anchors). Figure 14
illustrates one of the many available forms of gas anchors that can be used.
The advantages and disadvantages of a PCP are summarised in Tables 8 and 9. It can
be seen that the pump’s characteristics make it very suitable for artificial lifting wells
producing medium to high viscosity crude oil reserves. These crude oils often have
a tendency to form highly viscous emulsions when mixed under high shear with the
produced water (as occurs in a centrifugal pump). They are often found in (relatively)
shallow, young (geologically speaking), soft formations where the inclusion of sand
control in the completion design is a necessity. Finally, elastomer selection problems
are minimised because these crude oils tend to have a low GOR as well as a low
aromatic content and their shallow location results in a cool Bottom Hole Temperature.
Department of Petroleum Engineering, Heriot-Watt University
61
Advantages of PCP
Comment
Simple design
Quick pump unit repaired by replacing
High volumetric efficiency
In the absence of gas
Efficient design for gas
Tolerant of produced solids at reasonable
anchors available
levels
High energy efficiency
PCP is a Positive Displacement Pump
rotor and stator as a complete unit
Emulsions not formed due to
ESPs and Weir pumps promote emulsion
low shear pumping action
formation due to high pump speeds
Capable of pumping viscous
(1) Diluent mixed as required with crude
crude oils
oil if extreme viscosities to be pumped
(2) "Water-like" behaviour observed at high
water cuts when oil becomes the
internal phase
Disadvantages of PCP
Table 8
Advantages of a PCP
Comment
High Starting Torque
Fluid compatibility problems with
Carry out tests prior to producing a new
elastomers in direct contact with
crude oil
aromatic crude oils
Gas dissolves in the elastomers,
at high bottom hole pressure
Avoid rapid depressurisation of the pump
destructive bubbles formed when
pressure is lowered rapidly
9.2. Progressing Cavity Pump Power Supply
Traditionally, PCP’s have been powered by an electric motor and gearbox mounted
above the wellhead and turning a string of sucker rods connected to the PCP pump;
i.e. the rods are rotated rather than reciprocated (Figure 43). As discussed in section
2.6 on Rod Pumps, this string of sucker rods is susceptible to failure - especially in
crooked or deviated wells or when formation sand is being produced. There is a
similar tendency to a higher frequency of tubing failures, since the rods are rotating
inside the tubing. Installation of centralisers on the sucker rod string can mitigate this
problem (Figure 15). This rod/tubing frictional contact, even when reduced by
centralising the sucker rod string, leads to a large loss of starting torque as well as
wastage of power when the pump is operating.
Further, the tubing has to be pulled and then rerun when the pump unit requires repair.
This can normally be done by a light workover hoist, since the wells are not normally
capable of natural flow.
62
Table 9
Disadvantages of a PCP
2
Selection of Artificial Lift Types
9.3. New Technology
The resulting high torque and friction losses, as well as the tubing and rod failure
discussed above, can be reduced by placing the motor downhole - this is known as a
Progressing Cavity Electric Submersible Pump. Secondly, low cost replacement of
the PCP unit can be achieved by making it wireline retrievable. These new
developments will be discussed in the following two sections.
9.3.1. The Progressing Cavity Electric Submersible Pump (PCESP)
The PCESP share the same electric motor, seal, cable and control technology as the
conventional Electric Submersible Pumps (ESP) discussed in Section 2.7. The major
difference is that a gearbox is required to reduce the speed of rotation since the
centrifugal pump employed with a conventional ESP is a high-speed device; while a
PCP is a low speed device. The layout of the two pump types is compared in Figure
47. Some typical dimensions have been included to illustrate the length of the
complete PCESP unit. In the example shown, which can be run inside a 7" casing, it
ranges from 34 ft for a low power unit to 81 ft for a medium power device.
Electric Submersible Pump (ESP)
Progressing Cavity
Electric Submersible Pump (PCESP)
Cable
Cable
Tubing
Progressing
Cavity
Pump
Centrifugal Pump
*9 ft.- 40 ft.
Pump Intake or
Gas Separator
Seal or Protector
Gearbox
Electric Motor
Figure 47
Comparison between layout
of an ESP and a PCESP
9 ft.
11 ft.
*10 HP - 5 ft.
60 HP- 21 ft.
*Depends on power required
Department of Petroleum Engineering, Heriot-Watt University
63
Practical experience has shown that the expected gains in reduced pump power
requirements and tubing failures are achieved on changing from a rod driven PCP to
a PCESP. Electric submersible motors have become very reliable - providing proper
design and operational procedures are in place. The most frequently replaced item in
many PCESP projects is the PCP unit itself. These costs could be reduced by
redesigning the wellhead, completion and the PCP unit so that it becomes retrievable
by wireline. This is discussed in the next section.
9.3.2. Wireline Retrievable PCESP
The wellhead and the tubing are enlarged so that they are both large enough to pass
the PCP motor and stator unit, which is modified as follows (Figure 48):
Cable
2 7/8" Tubing
4 1/2" Tubing
(for 7" Casing)
Fishing Neck
Pump Pack Off
Wireline Retievable
Progressing Cavity
Pump
Pump Body
Rotor
Stator
Progressing Cavity
Pump
Seal Unit
Seating nipple
on Top of Seal Unit
Re-latchable Shaft
Connection
Gearbox
Drive Shaft
Electric Motor
(i)
A (re)latchable drive shaft connection is made at the bottom of the unit so that
it can be connected and disconnected to the motor unit and gearbox.
(ii) The PCP Unit is seated on a nipple placed on top of the seal unit.
(iii) The PCP unit body is sealed against a pump pack-off element.
(iv) A fishing neck is provided for wireline recovery.
64
Figure 48
Modification of a PCESP to
allow wireline retrieval
2
Selection of Artificial Lift Types
The motor, gearbox and seal unit are all installed as per the conventional PCESP
design at the bottom of the production tubing.
Extensive testing of this concept at the Thums field, Long Beach, California, provided
the operational data summarised in Table 10. This shows that, despite the considerable
increase in capital cost, the much reduced pump charge out costs resulted in a very
significant saving in production lifting costs. (This cost saving is very dependent on
the MTBF or “Mean Time Before Failure” of the pump and other downhole
components). This is a good example of the concept discussed earlier (Section 2.5.4)
when extra up-front capital expenditure can result in reduced “Total cost of ownership”
through more efficient operations. Also, that these decisions can only be made when
a comprehensive cost database is available.
Economics of wire line retrievable PCESP*
Table 10
Economics of wireline
retrievable PCESP
Initial Installation Cost
Pump Change Cost
180% of ESP
18% of ESP
125% of ESPCP
13% of ESPCP
Savings on "total cost of ownership"
15%
*Data provided by Thomas Lutz to the conference on Artificial Lift Equipment '97,
Dubai, December 1997
10. HYBRID SYSTEMS
The combination of two forms of artificial lift has been used in a few cases. Most
hybrid systems combine gas lift with one of the other types of artificial lift. For
example it has been shown that injection of gas above an ESP or PCP can reduce the
hydrostatic head against which the positive displacement pump (e.g. ESP, PCP etc.)
is pumping against by upto 40%. This can lead to a significant increase in pump
performance. It may also allow the well to continue to produce even when the pump
is experiencing mechanical problems e.g. ESPs are susceptible to high levels of sand
production and may show frequent failure. Gas lift is tolerant of sand production and
will often be capable of drawing the well down sufficiently that production continues,
although at a lower rate compared to when the pump is operational.
Table 11
Benefits of combining gas
lift with a positive
displacement artificial lift
method e.g. ESP, PCP
sucker rod etc.
•
•
•
•
•
•
•
•
Increased volumetric efficiency - higher liquid volumes.
Decreased injection gas requirements compared to gas lift alone.
Increased reservoir drawdawn and production.
Increase pump installation depth - allows greater reservoir drawdown.
Reduction in pump and motor power requirements.
Lower electrical energy consumption compared to pump alone.
Reduces electrical conduit requirements.
Gas lift provides backup in case of pump failure.
Department of Petroleum Engineering, Heriot-Watt University
65
11. ARTIFICIAL LIFT METHODS TUTORIAL
Question 1.
Hardly any artificial lift equipment was installed during the first 10 years of Oil
Production in the North Sea. Suggest four factors that could explain this.
Answer 1.
Artificial lift was not required due to properties of developments typical in that period:
• Light oil being produced from high permeability reservoirs i.e. (relatively low
drawdowns),
• Reasonable high GOR aided natural flow,
• Water Injection supported reservoir pressure above the bubble point pressure at
near hyrostatic or greater pressures, allows the production wells to continue to flow
under natural flow
• Large reservoirs and distant well spacing delayed water breakthrough
• High water cut wells were shut-in. Other, lower watercut wells produced in
preference to watered out wells since production was facility or pipeline constrained.
Question 2.
Two types of artificial lift were installed once it became apparent that production
pressure boosting would be required. Which types were these and what were the
reasons that they were chosen?
Answer 2.
The two artificial lift types were:
Gas Lift
• High GOR water drive reservoirs
• High reservoirs permeabilities
• No need for very low FBHPs (Water Injection, drawdown was relatively low).
• High Pressure gas availability, often surplus to power or export requirements
• (reasonably) deviated holes
• Wireline equipment maintenance - no extra demand placed on drilling rig which
could continue drilling development wells.
• Tolerate some sand production as was experienced in completions without sand
control
• No extra space required at production wellhead
• Full-access to oil producing formation
ESPs
• High production rates even at high water cuts
• Suitable for highly deviated wells
• Logging / coiled tubing access to formation via Y tool
• Flexible - flow rate controllable over a wide range
- downhole flow measurement and pump condition monitoring available
66
2
Selection of Artificial Lift Types
• Efficient use of Energy
• Can pump against high wellhead pressures e.g. for satellite wells
• No extra space required at production wellhead
Question 3.
New completion technology has contributed to reduction of the Operating costs of
Artificial Lift Equipment. Name two significant developments.
Answer 3.
Examples of new technology:
• Coiled tubing for:
- Insert strings
- Conveyed pumps
• Wireline maintainable pumps.
• Variable speed drives for ESPs
• Downhole measurement & control for ESPs
Question 4.
Which IPR Curve (AL1 or AL2) is more beneficial for Artificial Lift?
Bottom Hole Pressure (psi)
Inflow Performance Relationship (AL1)
2500
2250
2000
1750
1500
1250
1000
750
500
250
0 0
x
x
x
x
x
70
140 210 280 350 420 480 560
Total Liquid Production bbl/day
630 700
Bottom Hole Pressure (psi)
Inflow Performance Relationship (AL2)
2500
2250
2000
1750
1500
1250
1000
750
500
250
0
x
x
x
x
x
0
60
120 180 240 300 360 420 480
Total Liquid Production bbl/d
Department of Petroleum Engineering, Heriot-Watt University
540 600
67
Answer 4.
A “straight line” inflow performance relationship associated with a dead oil is more
favourable than the curved “Vogel” relationship found when well inflow takes place
below the fluid’s bubble point. This is because for a “straight line” inflow performance
relationship the % Increase in production is directly related to the % increase in
drawdown achieved by the introduction of the form of artificial lift. The increase in
production is considerably less for a curved “Vogel” relationship.
Question 5.
i) Why are these curves different?
ii) What impact might this have on the selection of the Artificial Lift type?
Answer 5.
i) A straight-line IPR assumes that oil is undersaturated, that is, only slightly
compressible. This condition does not apply to gases or saturated oil wells, both of
which are highly compressible. The effect of compressible gas and two-phase flow
on IPR results in “larger-than-linear” pressure drops being required to increase the
production rate i.e. a curved IPR is observed in this case. The rate - pressure relation
tends to show a more pronounced curvature at higher production rates.
ii) By applying the same drawdown, i.e. producing under similar flowing bottomhole
pressures, wells with the ‘Straight line” IPR (dead or undersaturated oils) would
yield higher production rate than wells with the curved “Vogel” IPR. The increasing
production of associated gas due to producing below the bubble point pressure in the
latter case would tend to favour the installation of Gas lift while, for example, Rod
Pumps can be applied to the dead oil or undersaturated oil wells.
Question 6.
List up to 6 key features for both Rod Pumps and Gas Lift that form the basis of the
following statement:
“Worldwide, 85% of Artificial Lift equipment installed is rod pumps. This is mainly
in low production rate wells while gas lift is the most popular artificial lift technique
for medium rate wells”.
Answer 6.
Rod Pump - Main features
• The vast majority of wells produce at low rates (generally less than 100 bpd) and
moderate depths
• Relatively cheap, so their use can be justified on such low rate wells
• Rod pumps are mechanically simple to operate and easy to repair/maintain/replace.
Can be operated by inexperienced personnel
68
2
Selection of Artificial Lift Types
• Sensitive to gas and solids (wax/scale/sand) - Solids can cause wear as well as
damage moving parts which then need to be replaced
• Not suitable for (highly) deviated wells (most land wells are near vertical)
• Obtrusive in urban locations. Equipment too heavy for offshore use
• Pump can be easily changed and performance monitored using relatively simple
and inexpensive techniques
• Viscous oil can be pumped
Gas Lift - Main features
• Suitable for medium to high rates
• Suitable for water drive reservoirs with a high bottomhole pressure
• High well PIs and high permeabilities mean FBHP can be excessively high, limiting
production
• High GOR => advantage rather than a drawback
• Gas has to be available
• Wireline serviceable at deviation up to 65˚. Coiled tubing can service more highly
deviated / horizontal wells
• Limited surface requirements once gas available
- can be used off-shore or in urban locations
• Fully open tubing giving access for production logging
• Subsurface tubing, and annular, safety valve
• Flexible - gas lift string design can be adjusted as well conditions change
• Forgiving of poor design & operation, but difficult to run efficiently
• Can handle (tolerate) produced solids
Question 7.
What considerations are important when choosing an Artificial Lift Method for
subsea wells for a satellite development at a distance 30 Km from a host platform?
Answer 7.
ESPs are normally the preferred Artificial lift method for the following reasons:
• Can generate high pump pressures to overcome extra friction from 30 km
pipeline
• Does NOT use power fluid (gas for gas lift, liquid for hydraulic pumps) which
would lead to extra friction in pipeline.
- Hence electricity probably preferred source of power
- Remote control capability at long distances
• Advantageous to place pump so as to minimise length of flowline with multi phase
flow
• System design (artificial lift type/tubing/flowline/reception facilities) should be
suitable for complete life of oil field.
• Long pump lifetime (reliability to complement lifetime design)
Multiphase pumping means Subsea pressure boosting can also be considered as an
option as well as ESPs
Department of Petroleum Engineering, Heriot-Watt University
69
FURTHER READING
(1) CHOLET, C
Progressing Cavity Pumps
Paris: Edition Technip 1997
ISBN 2-7108-0724-6
(2) BROWN, K E
The Technology of Artificial Lift Methods. Volumes 1-4
Pen Well Books, 1980
(3)
70
PERRIN, D.
Well Completion and Servicing - Oil and Gas Field Development Techniques
Paris: Editions Technip, 1999
ISBN 2710807653
2
Selection of Artificial Lift Types
Department of Petroleum Engineering, Heriot-Watt University
71
Gas Lift
3
CONTENTS
1.
2.
3.
4.
5.
6.
7.
8.
9.
INTRODUCTION
GAS LIFT INTRODUCTION
GAS LIFT APPLICATION
3.1. Gas Lift Advantages and Limitations
3.2. Review Example Gas Lift Completion
Designs
GAS LIFT DESIGN OBJECTIVES
4.1. Gas Lift Design Constraints
4.2. Gas Lift Design Parameters
4.3. The Surface Gas Network
THE UNLOADING PROCESS DESCRIBED
5.1. Safety Factors
5.2. Gaslift Valve Spacing Criteria
Summarised
SIDE POCKET MANDRELS
3.6.1 Other Uses of Side Pocket Mandrels
GAS LIFT VALVE MECHANICS
7.1. Casing or Inflow Pressure Operated (IPO)
Valves
7.2. Dome Pressure Calibration
7.2.1. Temperature Correction
7.3. Valve Performance
7.3.1. Dynamic Valve Performance
7.3.2. Valve Performance Flow Model
7.4. Proportional Response Valves
7.5. Dynamic Valve Response and Gas Lift
Completion Modeling
7.6. Well Stability
GAS LIFT DESIGN PROCEDURES
8.1. An Example Design - Optimising The
Performance of a Gas Lifted Well
8.2. An Example Design - Gas Lift Unloading
Calculations
8.3. Further Gas Lift System Considerations
8.4. Further Gas Lift System Calculations
OPERATIONAL PROBLEMS
9.1. Gas Quality
9.2. Solids
9.3. Changes in Reservoir Performance
9.4. Gas Supply Problems
9.5. Well Start - Up (Unloading)
9.6. Well Stability
9.7. Dual Gas Lift
9.8. Trouble Shooting
9.9. Trouble Shooting Techniques
9.10. Some Field Examples of Operational
Problems
10.
11.
12.
13.
14.
FIELD PRODUCTION OPTIMISATION
NEW TECHNOLOGY FOR CONTINUOUS
FLOW GAS LIFT
INTERMITTENT GAS LIFT
GRAPHICAL GAS LIFT DESIGN EXERCISE
FOR WELL EDINBURGH - 2
13.1. Introduction
13.2. Initial Condition - The "Dead" Well
13.3. Construction of The "Equilibrium Curve"
13.4. The Unloading Process
13.5. Gas Lift Optimisation Exercise
FURTHER READING
LEARNING OBJECTIVES:
Having worked through this chapter the Student will be able to:
•
Describe the gas lift process.
•
Explain the impact of the key gas lift process variables.
•
Identify application areas/advantages for gas lift.
•
Discuss the limitations of the gas lift process.
•
Describe the well unloading process.
•
Identify and explain the action of gas lift hardware components.
•
Design a gas lift completion.
•
Identify reasons why efficient gas lift depends on availability of high quality
data.
•
Construct a methodology for revenue optimisation with limited gas availability.
•
Describe the intermittent gas lift and plunger lift processes.
2
Gas Lift
3
1. INTRODUCTION
Chapter 2 introduced the concept of artificial lift and discussed the different types of
equipment that a Production Technologist can choose from. It was complete apart
from gas lift, the subject of this chapter. The objective of installing gas lift in a
completion is to increase the drawdown on the producing formation by injecting gas
into the lower part of the tubing string and consequently reducing the flowing gradient
in the production string. The concepts of multiphase flow and well performance
discussed in Chapter 1 are obviously very important here.
We will first introduce the basics of gas lift and discuss its advantages and disadvantages.
The design, operation and maintenance of the gas lift valves, which control the gas
injection from the annulus into the tubing, will then be described. The procedure to
design a gas lift completion string using one of the commercially available computer
programs will be discussed. A manual design exercise will illustrate the design
process. Typical gas lift operational problems and their solution will then be dealt with
and the need for continual optimisation of the gas lift reviewed. Finally, some of the
most recent developments in gas lift technology will be discussed.
2. GAS LIFT INTRODUCTION
A continuous flow gas lifted well completion has been sketched in figure 1. The
completion differs from the natural flow completions discussed earlier in that:
Produced Fluid
and Injected Gas
to Separator
Injected Gas
(Control and Metering)
(c) Large Gas Bubble
Displaces Liquid Slug
Gas
Gas
(b) Gas Bubble
Expands as
the Hydrostatic
Pressure Reduces
Liquid
(c)
Displacement of
Liquid Slugs
by Gas Bubbles
Liquid
Gas Lift Valves
Liquid
(b)
Expansion of Gas
Bubbles
(a) Injected Gas Reduces
Average Fluid Density
Gas Injected at
"Operating Valve"
Gas
(a)
Reduction of
Fluid Density
Producing Formation
Figure 1
Gaslifted Well Completion
Perforations
Department of Petroleum Engineering, Heriot-Watt University
3
(i) Gas, at a controlled volume and pressure, is injected into the tubing/casing
annulus.
(ii) The tubing string has been fitted with a number of gas lift valves. These valves
are installed at carefully spaced intervals so that any liquid present above them in the
casing/tubing annulus (e.g. due to killing of the well) can be removed by injection
of gas at the top of the well annulus leading to the liquid U-tubing into the tubing and
its subsequent ejection from the well. The gas injection point into the tubing is then
transferred to successively deeper gas lift valves (see section 3.5 for details).
(iii) The gas is injected into the tubing through the “operating valve”. The injected
gas enables the well to resume production by :
(a)
the injected gas reducing the average fluid density above the injection
point.
(b)
some of the injected gas dissolving into in the produced fluids, providing
they are undersaturated with respect to the gas solubility. The remainder,
in the form of bubbles, will expand due to reductions in the hydrostatic
pressure as the fluids rise up the tubing.
(c)
the coalescence of these gas bubbles into larger bubbles occupying the
full width of the tubing. These bubbles are separated by liquid slugs,
which the gas bubbles displace to surface. This is called slug flow.
The design of a gas lift completion thus consists of two separate distinct parts:
(i) Choice of the installation depth, type and design of the gas lift valves placed
above the operating valve so that any liquid in the tubing and casing/tubing annulus
can be unloaded via the wellhead (see section 3.5).
(ii) Optimisation of the flowing gas lifted well. The well essentially behaves as a
conventional flowing well, except that the gas/liquid ratio (GLR) suddenly increases
at the operating valve depth (see section 3.8).
The wellbore opposite the perforations is treated as the node pressure when the system
is analysed using the “nodal analysis” process discussed in chapter 1.12. The analysis
equates the following at any given flow rate:
Inflow to Node (the perforations):
Preservoir - Pdrawdown = Pperforations
Outflow from Node (the perforations):
Pseparator + ∆Pflowline + ∆Pchoke + ∆P(tubing above operating valve) + ∆Ptubing below operating valve = Pperforations
The pressure drop across the tubing below the gas injection valve is estimated with
using multiphase flow correlations (chapter 1.1.7) or pressure traverse curves (chapter
4
Gas Lift
3
1.1.6) using the “natural” gas liquid ratio. The pressure drop between the gas injection
valve and the surface is calculated using the “enhanced” gas liquid ratio calculated
from the sum of the {lift + produced} gas rate divided by the liquid production rate.
Figure 2 illustrates a pressure traverse across the well when it has reached steady state
operation. The gas is being injected at the wellhead at a pressure of 1100 psi. The
pressure of the gas in the annulus increases with depth due to its density (typically at
the rate of 30 psi/1000 ft). The gas is initially being injected at the valve 4 at 3800 ft.
The well is producing with a 500 psi drawdown. The flowing pressure gradient from
the producing perforations to the operating gas lift valve is equal to 0.44 psi/ft. There
is a 250 psi pressure drop across the gas lift valve and the average fluid gradient above
the injection valve has been reduced 0.27 psi/ft by the injected gas. The situation for
deeper gas injection is also sketched in which the gas is being injected through valve
7 at 5000 ft. The gas lift pressure is now just sufficient to allow injection to occur if
the pressure drop across the gas lift valve is restricted to 50 psi. It can also be seen that
the deeper injection allows the drawdown to increase to 850 psi.
Produced Fluid
and Injected Gas
to Separator
Injected Gas
(Control and Metering)
Wellhead Annular Gas Injection Pressure (psi)
500
1000
1500
2000
2500
0
3000
2
4
2000
3000
Casing (Gas) Pressure Gradient
nt
3
Depth (Ft. TVD)
1
ie
rad
eG
sur ction
res
g P f Inje
ubin oint o
gT
win bove P
A
Flo
1000
4000
Gas Injection
at Valve 4
5
Pressure Drop Across Valves
Gas Injection at Valve 7
Operating Gas Lift Valve
ts
ien
ad
Gr n
re tio
su jec
es In
Pr of
g int
bin Po
Tuow
ing el
ow B
5000
Fl
6
7
Flowing Bottom Hole
Pressure, Valve 4
6000
Producing Formation
Figure 2
Pressure Traverse Through
a Well
7000
Flowing Bottom Hole
Pressure, Valve 7
Perforations
Reservoir
Pressure
Drawdown, Valve 4 gas injection
Drawdown, Valve 7 gas injection
It can be appreciated from this diagram that the gas injection pressure is the main
control on the depth of gas injection while the gas injection rate also contributes to the
extent of the reduction in the flowing pressure gradient. These parameters can be
adjusted as required on a day-to-day basis. The pressure settings of the gas lift valves
Department of Petroleum Engineering, Heriot-Watt University
5
(which control the pressure levels at which the valve opens and closes - see section
7.3) can be adjusted when required using wireline techniques - see section 6). The
depths at which the valves are set can only be altered by pulling the tubing and
recompleting the well with a tubing string in which the spacing between the side
pocket mandrels has been altered.
Increases in the gas injection rate through a gas lift valve set at a given depth will
increase the fluid production rate until a maximum is reached (figure 3). At this point
the “reduction in average fluid density in the tubing due to a slight increase in the gas
injection rate” is being exactly counterbalanced by the “increased frictional pressure
losses due to the greater mass of fluid flowing in the tubing”. Further increases in the
gas flow rate will result in the friction term increasing relatively faster than the
hydrostatic head reduction term. This is the “technical optimum gas injection rate”
at which the well production is maximised.
Production Rate
Economic optimum gas injection rate
where marginal extra gas injection
cost balances marginal extra
production revenue.
Maximum liquid production
or technical optimum gas
injection rate
Unstable flow below this
rate due to too low
gas injection rate
Some wells flow
"naturally" without
gas lift.
Others require
"Kick off" gas to
initiate production
Gas Injection Rate
“The maximum economic gas injection rate” will be somewhat lower - this is the gas
injection rate at which the marginal cost of providing extra injection gas is equal to the
marginal revenue from the extra well production.
Figure 3 also illustrates that gas lift may be applied to increase the production from
wells in which will flow naturally at a low(er) rate. The second case illustrated is for
a well which is “dead” and does not produce without some form of artificial lift. Gas
then has to be injected at a certain rate (“kick-off” gas) before any well production is
possible.
6
Figure 3
Effect of gas rate on well
production
Gas Lift
3
An efficient gas lift system depends on a continuous supply of gas at the specified
pressure. A considerable infrastructure is required for gas lift. This is normally only
installed when there are a number of wells in the area using gas lift as the preferred
form of artificial lift. A typical gas lift system arrangement is shown in figure 4. This
figure shows several wells producing into a production manifold. The gas is then
separated, compressed and dried in a dehydration unit. Any excess gas may be sold
or make up gas imported, as required by the demand of the gas lift system. The lift
gas is supplied to the gas lift manifold, after which the injection gas flow rate and
casing head pressure are adjusted before injection into the individual wells.
Import
make up
gas
Surplus
sales gas
Dehydration
unit
Compressor
Gas
3 Phase
Separator
Oil to
storage
Water to
disposal
Injection Gas manifold
Injection gas
presssure and flow
rate measurement
P
F
Production pressure
and flow rate
measurement
P
F
Production manifold
Figure 4
Gas lift system
The metering and control equipment for a gas lifted well that is being individually
tested is illustrated in figure 5. Both manual and automatic lift gas control are
illustrated.
Department of Petroleum Engineering, Heriot-Watt University
7
Wing Valve
Choke Box
Safety Valve
Inlet Valve
(Orifice) Gas
Flow Meter
Master Valve
Data Logger
Gas
Oil
Echometer
(Measures fluid
level in annulus)
Water
Oil Pipe Line
Gas Lift Pipe Line
Casing
Unloading
Gas Lift Valves
Tubing
Operating Gas Lift Valve
Packer
Data Logger
Production test Separator
(Orifice)
Flow Meter
Needle Valve
to Control Lift Gas
Main
Valve
Manual control
Gas Lift Manifold
Producing Formation
OR
Flow meter that
automatically adjusts
choke setting
Perforations
Main
Valve
Automatic control
Gas Lift Manifold
3. GAS LIFT APPLICATIONS
The process described above is called “continuous flow gas lift”. “Intermittent gas
lift” is used in low rate production wells. This approach involves switching off the
injection gas at regular intervals so as to allow the fluid level in the well to build up.
The gas injection is recommenced, and the fluid in the tubing lifted to surface, when
a sufficient depth of produced fluid is present in the well. The cycle is then repeated.
Intermittent gas lift is thus used for cases when the outflow capacity of the gas lifted
tubing is greater than the formation’s capacity to produce fluid into the well.
The module on multiphase flow in vertical tubing explained how the flowing gas will
by-pass some of the liquid in the tubing (the slip phenomenon). This liquid will fall
back down the well each time the gas lift is switched off. Fall back can be avoided by
installing a plunger at the bottom of the well. Gas injection now occurs underneath
this plunger, which rises upwards, displacing the liquid above it to the surface. The
plunger falls to the bottom of the well when the gas is switched off. The downhole
completion is arranged so that inflowing fluid can collect above the plunger while a
check valve ensures that the injected gas can not be injected into the formation. The
cycle can now be repeated at a regular time interval. This will depending on the well
8
Figure 5
Metering and control of a
gas lifted well.
Gas Lift
3
productivity and the volume of liquid displaced to the surface by the plunger. This
method is described in greater detail in chapter 3.12.
Gas lift has been applied to a wide range of production scenarios - as can be seen from
Table 1. In fact, gas lift is the only artificial lift method that actually works better in
a well that is producing at a significant gas/liquid ratio. Gas lift is often the preferred
artificial lift method for wells with a:
(i)
high gas-oil ratio;
(ii) high productivity index;
(iii) (relatively) high bottom hole pressure due to reservoir pressure support being
provided by a natural or artificial water drive.
Table 1
Continuous flow gas lift
applications
1
2
Production wells which will not flow naturally.
Increase production rate in flowing wells.
3
4
5
6
Unload liquid from wells that will flow naturally once on production.
Unload liquid in wet gas wells which would otherwise cease to flow.
Back flow injection wells.
Lift aquifer wells.
The key point of gas lift is that a reliable, adequate (in terms of pressure and flow rate)
gas supply has to be available at all times. The proviso at the end of the sentence is
the key one. When the field/wells are operating normally the (lift) gas system (figure
3) will be fully charged with gas. This gas will be recovered and recirculated many
times. Extra volumes of “make-up” gas associated with the current oil production will
only be required to make good any losses from the system, as well as any gas used for
compression or other power requirements. When planning a gas lift installation for
a field one should specifically allow for the:
(i) decrease in (fresh or make-up) gas supply as the field reserves are depleted and
the well water cut increases. This can result in gas being imported during the late
project life, particularly for offshore developments when the produced gas is also
used to generate the platform’s electrical power.
(ii) case when none of the wells flow naturally. An external gas source is then
required to bring the (first) well(s) onto production after a facility shutdown.
(Vapourised, liquid) Nitrogen can be used for this purpose if there is no provision to
import natural gas.
(iii) fact that, if only a low rate gas supply is available, it will take a long time to
return all the wells to production after a shutdown.
(iv) choice of lift gas injection pressure has to be made at an early stage in the project
lifetime when the gas compressor specifications are drawn up and little information
may be available about actual well performance.
Department of Petroleum Engineering, Heriot-Watt University
9
3.1 Gas Lift Advantages and Limitations
These are summarised in tables 2 and 3. They are self explanatory if read in
conjunction with the above discussion.
Operation of gas lift valves is unaffected by produced solids (sand etc.)
Gas lift operation is unaffected by deviated or crooked holes.
Use of side pocket mandrels allows easy wireline replacements of (inexpensive)
gas lift valves when deviation <60 .
Provides full bore tubing access for coiled tubing or other well service work.
High fluid gas oil ratio improves lift performance rather than presenting problems
as with other artificial lift methods.
Flexible - can produce from a wide range depths & flow rates
- uses the same well equipment from 100-10,000bpd production rates
- copes with uncertainties and changes in reservoir performance,
reservoir pressure, water cut & production index over the well life.
Low surface profile important for offshore & urban locations.
Tubing & annular subsurface safety valves available when required by
safety regulations.
Gas lift tolerates "bad" design - though "good" design is more difficult.
Gas lift has a low initial (downhole) equipment cost.
Gas lift has a low operational and maintenance costs. Major workovers are
infrequent when wireline servicing is possible.
Well completions are relatively simple. This can be important in remote areas.
Gas lift operation independent of bottom hole temperature.
High back pressure on sandface due to fluid in the tubing restricting production.
- e.g. lifting a well with a Productivity Index of 1 bpd/psi from 10,000 ft with a
static bottom hole pressure of 1000 psi is difficult.
- Flowing bottom hole pressure is greater than with e.g. Electric Submersible
Pumps. This leads to potential loss of reserves.
Gas lift is inefficient in energy terms (typically 15-20%).
Gas compressors have a high capital cost. They require expensive maintenance &
require skilled operations staff. However, they may already be required for gas sales.
Annulus full of high pressure gas represents a safety hazard.
High installation cost can result from top sides modifications to existing platforms
e.g. Compressor installation.
Adequate gas supply required throughout project life
- Decreasing BHP, increasing water cut etc.
- Sufficient gas to start up FIRST well
- Slow start up after facility shut down
- Increased gas handling requirements in facilities.
Gas lifting of viscous crude (<15 API) is difficult and less efficient.
Wax precipitation problems may increase due to cooling from (cold) gas injection &
subsequent expansion.
Hydrate blocking of surface gas injection lines can occur during cold weather if gas
inadequately dried.
Lifting of low fluid volumes is inefficient due to gas slippage.
Good data management and complete network modelling required for efficient /
maximum profitability operation.
10
Table 2
Gas lift advantages
Table 3
Gas lift limitations
Gas Lift
3
3.2 Review Example Gas Lift Completion Designs
Question
Figure 6 shows types of completion designs for gas lifted wells. You should:
(i) identify the type of completion;
(ii) describe the completion’s advantages and disadvantages.
Each completion is discussed in turn below:
Gas
Production
Gas
Production
SCSSSV
SCSSSV
Production
Gas
Figure 6 (a) to (c)
Gas lift completions designs
(a)
(b)
(c)
Answers
Figure 6 (a) Single String Continuous Gas Lift Completion
This is the standard completion design. Gas is injected in the annulus and the produced
fluids are lifted to the surface through the tubing. The well may be completed on a
single or on multiple formation zones. In the latter case, the separate zones may be:
(i) produced together (commingled) or
(ii) isolated from one another by packers. The required zone can then be produced
selectively by opening and closing the appropriate sliding side doors.
Figure 6 (b) Annular Flow Gas Lift Completion
The gas is injected down the tubing and the production flows up the annulus. This well
Department of Petroleum Engineering, Heriot-Watt University
11
design can be found onshore in the Middle East. Higher production rates are achieved
compared to the conventional production configuration where the produced fluid
flows up the tubing. This is due to the reduced (frictional) pressure drop in the annulus
compared to the tubing due to its annulus’s larger flow area. The disadvantages are
that corrosion of the casing by the produced fluids will lead to a loss in well integrity
(see section 3.9.1). Also, a decline in the well production rate will lead to severe
slugging earlier than for tubing flow.
Figure 6 (c) Continuous Gas Lift without the surface section of the Casing/Tubing
Annulus being Live (filled with Gas)
This completion features the gas being injected into a separate injection string with its
own Surface Controlled Sub Surface Safety Valve (SCSSSV) installed below a dual
packer. The gas is then injected into a single tubing designed for conventional,
continuous gas lift. This production string also has a SCSSSSV installed below the
upper, (dual) packer. This type of well design has been installed in the North Sea.
Production
Production
Long string
production
Short string
production
Gas
Gas
Gas
Plunger
One way or
check valve
(d)
(e)
(f)
Figure 6 (d) Dual, Gas Lifted Completion
This completion allows two zones to be independently produced by gas lift through
separate production strings. It is quite difficult to achieve optimum lift on both strings
since the action of the gas lift valves on the different strings interfere with each other
- see section 3.9.7.
12
Figure 6 (d) to (f)
Gas lift completions designs
Gas Lift
3
Figure 6 (e) Intermittent, Plunger Lift
This is installed in low rate wells, particularly in the USA, where the inflow rate from
the formation is low and smaller than the outflow capacity of the gas lifted tubing. The
plunger prevents fallback of the liquid when the gas is switched off. This liquid would
normally have been bypassed by the gas flowing up the tubing (slip) on its way to
surface.
Figure 6 (f) Single Valve (Subsea) Completion
This completion is used when intervention (change of gas lift valve settings etc) is
difficult and/or expensive. The single (orifice) operating valve minimises operational
problems. However:
(i)
the depth of lift gas injection is restricted since unloading valves are not used.
(ii) this injection depth is often maximised by increasing the gas injection pressure
above the normal 1,000-1,200 psi. A compressor capable of delivering gas at such
a higher pressure can only be provided at a substantial extra cost. Compressor
pressures over 3,000 psi have been used during the unloading process so that a
substantially greater depth of injection can be achieved. Conventional, much lower
pressures will be required once gas lift has been initiated and the well is flowing
steadily.
4. GAS LIFT DESIGN OBJECTIVES
The gas lift system designed for installation in a specific well should meet the
following objectives:
(i)
Maximise the (net) value of oil produced. This normally implies that the:
(a)
operating valve, through which the gas will be continuously injected,
should be situated as deep as possible and
(b)
gas injection rate should equal the economic limit at which the marginal
value of the extra oil produced equals the marginal cost of providing this
extra gas (figure 3).
Further optimisation is required when more than one well is being produced and
there is insufficient lift gas available to meet this economic criteria in all wells (see
section 3.10)
(ii) Maximise design flexibility. The gas lift design should be capable of coping
with the expected changes in the well producing conditions during its lifetime, as
well as the “unplanned” uncertainties in reservoir properties and performance.
These changes normally involve deterioration, from a well productivity point of
view, due to decreases in the Reservoir Pressure and Well productivity Index and
increases in the Water Cut.
Department of Petroleum Engineering, Heriot-Watt University
13
(iii) Minimise well intervention. This is particularly important in subsea or other
wells where wireline access is difficult or impossible.
Well completions with a “dry” tree and deviations less than 60o allow the option to
replace the gas lift valve by a relatively quick, wireline operation. The operating
parameters (or valve performance) of the gas lift valves installed in the side pocket
mandrels can thus be adjusted at any time in the well’s life i.e. the tubing production
conditions can be adapted to take into account changes in the reservoir conditions and
the well performance. These operating parameters include the:
(a)
tubing or casing pressures (depending on the type of side pocket
mandrel installed) at which gas flow through the gas lift valve starts
and stops and
(b)
port (or choke) size, which controls the maximum volume of gas that
can be injected as well as the associated pressure drop due to the gas flow
through the valve.
This ability to modify the valve performance when required leads to great flexibility
in the choice of gas lift operating parameters, despite the fact that the installation
depth of the gas lift valves is fixed. {The (side pocket) gas lift mandrels within which
the valve is placed are permanent fixtures in the completion string, having been
installed during the well completion process}.
The flexibility of a particular gas lift design is further increased by installing one or
two extra gas lift valves as possible both above and below the chosen depth of the
operating valve. They should be placed as close together as possible, but sufficiently
far apart that they do not interfere with each other’s operation (typically 150 m
vertical depth apart). This is known as the “bracketing envelope”. The inclusion of
the bracketing envelope mandrels will allow the operating valve to be moved to a
slightly higher or lower depth, as dictated by the well & reservoir performance
changes during the well life. This procedure maximises the (liquid) production by,
for a given gas injection rate, allowing the well to be lifted from as deep as possible
commensurate with the current producing conditions.
(iv) Stable Well Operation. Well “heading”, in which the Tubing Head or Casing
Head Pressure Shows regular changes (see figure 40 and section 9.6) should be
avoided. Stable operation - with a constant value for the casing and tubing head
pressures - should be aimed for. This is because stable well operation will always
produce more oil and, often, require less lift gas than unstable gas lifted well
operation.
Casing Head pressure excursions of as little as 5 psi can indicate valve multipointing
{the gas injection point changing from one valve to another or a second valve cycling
(opening and closing) in addition to the operating valve}.
4.1 Gas lift design constraints
There are three different sets of circumstances in which a gas lift design has to be made
and the above gas lift design objectives need to be met:
14
Gas Lift
3
(i) The valves are to be installed as an integral part of the tubing i.e. side pocket
mandrels and retrievable gas lift valves are not used. The valve spacing and
operating parameters is then fixed until the tubing is pulled and the well recompleted.
Such completions are usually used in shallow, relatively depleted land wells where
a low cost hoist can quickly carry out the operation.
(ii) Side pocket mandrels are included in the completion string. Dummy valves will
be installed in these mandrels initially if the well is to be produced for a period under
natural flow. Gas lift valves will only be installed at a later date when they are
required to maintain the production rate. Producing the well under natural flow will
provide the information to remove much of the uncertainty in the well and reservoir
performance. This production experience can then be used to choose the valve
settings when the time comes to replace the dummy valves with real valves.
(iii) A gas lift design is to be installed in a well that was completed sometime ago.
The valves are to be run into the existing side pocket mandrel locations. The well
conditions (Productivity Index, Water Cut, Reservoir Pressure etc.) may have
changed considerably compared to those when the gas lift was first designed and /
or first installed. Further, these conditions may have changed in a manner which was
not anticipated when the original design calculations were made.
The constraints, which limit the design options, increase from (i) to (iii).
4.2 Gas lift design parameters
The gas lift design process has to answer the following questions to meet the above
objectives:
(i)
How many unloading valves are required and at what depths should they be
placed?
(ii) What are the required settings for the Unloading Valves?
(iii) What is the depth of the operating valve where the gas is continuously injected?
(iv) What is the gas injection (or casing head) pressure?
(v) How much lift gas should be injected?
(vi) What is the tubing head pressure for the target flow rate?
This is translated into practice by ensuring that the gas lift valve spacing and pressure
setting are such that:
(i)
the operating valve should have adequate flow capacity and be placed as deep
as possible,
(ii) the available lift gas pressure must be able to displace the fluid in the casing to
the operating valve depth,
Department of Petroleum Engineering, Heriot-Watt University
15
(iii) all valves can be opened by the appropriate producing pressure gradient, while
the other valves above it are closed.
4.3 The Surface Gas Network
Figure 4 illustrated the complete gas lift system. It will have become apparent from
the above and the following sections that the performance of the gas lift system will
depend on the pressure and flow capacity of the lift gas available at the well head. The
surface piping network should thus be designed to:
(i)
have minimal (< 100psi) pressure loss between the compressor and the most
distant wellhead,
(ii) prevent one well from interfering with a second well by having sufficient pipe
volume to dampen pressure surges and
(iii) provide individual gas measurement and flow control for each well
Large diameter piping encourages all the above - typically 4 in OD piping is used for
the main backbone of the system with individual 2 in OD flow lines installed to each
well. A ring main system is an option for large systems employing more than one
compressor - the gas lift manifold for a group of wells and the compressors being
attached to the gas supply ring main as appropriate.
5. THE UNLOADING PROCESS DESCRIBED
To Separator / Storage Tank
Injection Gas
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Open
re
su
es
e
Pr d sur
an res
P
g
in
6000
Producing Formation
Figure 7
The "dead" well
7000
Tubing Pressure
16
g
4000
5000
Fourth Valve Open
Perforations
sin
3000
b
Tu
Third Valve Open
2000
Ca
Second Valve Open
True Vertical Depth (Ft. TVD)
1000
Casing Pressure
Reservoir pressure
Gas Lift
3
Figure 7 shows the situation when a well planned for gas lift has just been (re)completed.
The fluid level in the casing and the tubing is just below the surface and balances the
reservoir pressure. The well is dead - no fluids are being produced. The hydrostatic
head of the fluid column will equal the reservoir pressure, the actual fluid height will
depend on the liquid density - a column of water will have a lower height than an oil
column. No gas is being injected into the casing - both the tubing and casing have been
depressurised at surface to atmospheric pressure. All the gas lift valves are open due
to the hydrostatic head of the fluid.
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Open
Second Valve Open
True Vertical Depth (Ft. TVD)
1000
2000
3000
4000
Third Valve Open
5000
Fourth Valve Open
Figure 8
Gas lifted well unloading,
stage 1
6000
Producing Formation
7000
Injection pressure
Tubing Pressure
Perforations
Casing Pressure
Reservoir pressure
Gas injection into the casing / tubing annulus has been started in Figure 8. The fluid
is being U-tubed from the casing into the tubing through all the open gas lift valves.
The gas lift pressure is sufficient to increase the fluid level in the tubing to the surface
so that it flows via the surface flowlines into the separator. The pressure in the
wellbore at perforation depth is greater than the reservoir pressure i.e. some of the
liquid originally present in the well is being injected into the formation. This injection
of contaminated, potentially formation damaging, fluid can be prevented by installing
a one way flow valve or check valve at the bottom of the tubing. It is important that
the unloading process should occur at a controlled rate - the gas injection rate is
carefully controlled through the partially opened injection gas choke (see section
3.9.5). This will prevent damage to the gas lift valves as the fluid flows from the casing
and into the tubing via the open gas lift valves.
Department of Petroleum Engineering, Heriot-Watt University
17
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Open
Second Valve Open
True Vertical Depth (Ft. TVD)
1000
Third Valve Open
2000
3000
4000
5000
6000
Fourth Valve Open
Producing Formation
7000
Tubing Pressure
Perforations
Casing Pressure
Reservoir pressure
Figure 9 shows the situation when the unloading process has lowered the fluid level
in the casing annulus to the top gas lift valve. Gas injection into the tubing has now
commenced. The injected gas partially evacuates the liquid in the tubing above the
top gas lift valve into the separator under multi-phase flow conditions. This partial
evacuation reduces the fluid density in the tubing above the top gas lift valve and
ensures that further casing fluid to be unloaded through valves No. 2, 3 and 4; since
the pressure in the tubing at these points is lower than the pressure in the casing. The
well will also start to produce formation fluid if this reduction in pressure is sufficient
to give a drawdown at the perforations.
N.B. Any fluid lost to the formation earlier on in the unloading process will be
produced back first. This will be “dead” i.e. not “live” formation fluid whose intrinsic
gas content would reduce the hydrostatic head in the tubing and help bring the well into
production.
18
Figure 9
Gas lifted well unloading,
stage 2
Gas Lift
3
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Open
Second Valve Open
True Vertical Depth (Ft. TVD)
1000
Third Valve Open
3000
4000
5000
6000
Fourth Valve Open
Figure 10
Gas lifted well unloading,
stage 3
2000
Producing Formation
7000
Drawdown
Tubing Pressure
Perforations
Casing Pressure
Flowing Bottom Reservoir pressure
Hole Pressure
In Figure 10 the fluid level in the casing has now been lowered sufficiently to expose
gas lift valve No. 2. The top two gas lift valves are open and gas is being injected
through both valves. All valves below also remain open and continue to pass casing
fluid into the tubing. The tubing has now been unloaded sufficiently to reduce the
bottom hole pressure below that of the reservoir pressure (SIBHP). This pressure
difference, or drawdown, induces flow of formation fluid from the reservoir into the
wellbore i.e. the well is starting to produce.
Department of Petroleum Engineering, Heriot-Watt University
19
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Closed
Second Valve Open
True Vertical Depth (Ft. TVD)
1000
2000
3000
4000
Third Valve Open
5000
6000
Fourth Valve Open
Producing Formation
7000
Drawdown
Tubing Pressure
Flowing Bottom Reservoir pressure
Perforations
Casing Pressure Hole Pressure
The process continues in Figure 11. The top gas lift valve has now closed due to the
reduced pressure at this point. All the gas is being injected through valve No. 2.
Unloading the well continues with valves 2, 3 and 4 open and casing liquid flowing
into the tubing via valves 3 and 4.
There are two basic types of gas lift valves - the valve open and closing action being
in response to either the tubing or the casing pressure. Thus the closure of the top gas
lift valve was triggered by the reduction in the casing pressure (for casing pressure
operated valves) or tubing pressure (for fluid operated and proportional response
valves) after gas lift had been established through valve number two. {See section 7
on “Gas Lift Valve Mechanics” for a description of the construction and operation of
these two types of valve.}
20
Figure 11
Gas lifted well unloading,
stage 4
Gas Lift
3
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Closed
Second Valve Open
Third Valve Open
True Vertical Depth (Ft. TVD)
1000
3000
4000
5000
6000
Fourth Valve Open
Figure 12
Gas lifted well unloading,
stage 5
2000
Producing Formation
7000
Drawdown
Tubing Pressure
Perforations
Casing Pressure
Flowing Bottom
Hole Pressure
Reservoir pressure
Figure 12 shows valve No. 3 having just been uncovered so that both the No. 2 and
3 valves are passing gas. The bottom valve below the liquid level is also open and
liquid unloading from the casing / tubing annulus into the tubing continues.
NB. a deeper point of injection lowers the Flowing Bottom Hole Pressure; creating a
greater drawdown and hence increasing the production rate.
Department of Petroleum Engineering, Heriot-Watt University
21
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Closed
Second Valve Closed
True Vertical Depth (Ft. TVD)
1000
Third Valve Open
2000
3000
4000
5000
6000
Fourth Valve Open
Producing Formation
7000
Drawdown
Tubing Pressure
Perforations
Casing Pressure
Flowing Bottom Reservoir pressure
Hole Pressure
Figure 13 shows that, similar to the chain of events that lead to the closure of valve No.
1, the reduction in casing or tubing pressure once valve No. 3 starts to flow gas will
result in valve No. 2 closing. All the gas is now being injected through valve No. 3.
22
Figure 13
Gas lifted well unloading,
stage 6
Gas Lift
3
To Separator / Storage Tank
Injection Gas
Choke Partially
Open
0
500
1000
Pressure (psi)
1500
2000
2500
3000
Top Valve Closed
Second Valve Closed
True Vertical Depth (Ft. TVD)
1000
Third Valve Closed
3000
4000
5000
6000
Fourth Valve Open
Producing Formation
Figure 14
The producing gas lifted
well
2000
7000
Drawdown
Tubing Pressure
Flowing Bottom
Perforations
Casing Pressure Hole Pressure
Reservoir pressure
The process has continued to its logical conclusion in Figure 14. Valve No. 4 has been
exposed to gas flow and valve No. 3 has shut. All the gas is being injected through
valve No. 4 - this is the operating valve. An operating valve can be either a gas lift
valve or a simple orifice.
Figure 15 is an ideal illustration of the development of the tubing head and casing head
(or gas lift) pressure with time during unloading process described in Figures 7 to 14.
The sequential reduction in casing head pressure as gas is successively injected
through the lower gas lift valves is shown. This is not always so clearly observed in
practice. The annulus pressure is not only controlled be the settings of the gas lift
valve(s) referred to above, but also by the balance between the casing head (surface)
gas injection rate and the (total) gas passage rate through the valves. Further, it can
be seen that the erratic behavior of the tubing head pressure as the well is being
unloaded is replaced by a steady value as stable production is achieved when the well
is lifted from valve No. 4.
Department of Petroleum Engineering, Heriot-Watt University
23
First gas to surface.
M
M
M
Re
c
300
200
100
500
400
800
700
600
1000
900
Twin Pen
6AM
7AM
10
0
90 0
0
80
0
70
0
60
0
50
0
400
300
200
100
8A
50
0
40
0
30
0
20
10 0
0
600
500
400
0
60
700
Stable production
via 4th valve
800
M
9P
4
4th valve
uncovered
8P
900
10
90 00
0
80
70 0
0
2A
1000
10P
M
11PM
M
3A
300
200
100
MIDNIGHT
AM
7PM
10
90 00
0
80
70 0
0
0
60
0
50
Tubing head
pressure
6PM
3rd valve
uncovered
5P M
AM
900
AM
TIME
1000
TIME
DATE ON
M
DATE ON
4P
2nd valve
uncovered
800
700
600
500
f Casing and
go
Tu
din
bi
or
3P
0
80
0
70
0
60
0
50
0
40
0
300
200
100
500
400
300
200
100
00
90
40
0
30
0
20
10 0
0
Casing head or
lift gas pressure
5A M
Top valve
uncovered
800
700
600
M
10
M
9A
2P
1000
900
ssure.
Pre
ng
Switch on
gas lift.
1PM
400
300
200
100
Open choke
bleed off tubing
pressure.
MIDDAY
M
11A
Liquid only being
expelled from
AM
10
tubing.
M
1AM
5.1 Safety Factors
Several safety factors are normally introduced when preparing gas lift designs for real
wells. This is to account for:
(i) errors in the valve’s pressure settings,
(ii) errors and fluctuations in the well data, lift gas injection pressure and the estimate
of the valve temperatures under flowing conditions,
(iii) the pressure drop across the valve’s choke and that required to obtain sufficient
movement of the stem.
One method by which safety factors can be built into the design is illustrated in section
3.13.4
5.2 Gaslift Valve Spacing Criteria summarised
The well unloading process was described in the previous section. The chosen
installation depths of valves Nos. 1 - 4 will have been based on the following criteria:
(i) Specify a minimum number of gas lift valves. This will not only reduce the cost
but also the number of potential leak paths.
(ii) Gas lift valves to be installed sufficiently far apart that they do not interfere with
each other’s operation (150m suggested minimum spacing).
24
Figure 15
Typical casing / tubing
pressure and production
rate measurement during
unloading of a gas lifted
well.
Gas Lift
3
(iii) (continuous) Gas injection through the operating valve occurs as deep as
possible based on current producing conditions.
6. SIDE POCKET MANDRELS
Completion equipment is available so that gas lift valves can be permanently installed
as part of the tubing. Such wells require a workover if any repairs or changes to the
settings of the gas lift system need to be made. However, most completions employ
side pocket mandrels (figure 16) installed at appropriate depths in the tubing string as
part of the permanent completion. Side pocket mandrels allow gas lift valves to be
installed (and recovered) in a live well using wireline techniques. They are oval
shaped accessories with an outside diameter greater than that of the tubing {figure 16
(c)}. This shape allows the gas lift valve to be installed in the pocket placed to one
side of the tubing conduit, thus maintaining fullbore access throughout the complete
tubing length.
Lift Gas Injection
Lift Gas Injection
Produced Fluids
and
Lift Gas
Produced Fluids
and
Lift Gas
Lift Gas Injected
Via Port in Tubing
Gas Lift Valve
Pressure Sensitive
Element (Bellows)
Pocket for Gas
Lift Valve
Lift Gas Injected
Via Port in Casing
Seal Element
Pocket for Gas
Lift Valve
Polished Bore
Lift Gas Injected
Via Port in Casing
Produced Fluids
(a)
Figure 16
Schematic view of side
pocket mandrel showing
comparison of injection and
tubing pressure operated
valves
Produced Fluids
Side pocket mandrel for injection
pressure operated valve
(b)
Side pocket mandrel for tubing
pressure operated valve
Full Bore
Access
Through
Tubing
Seal Element
Polished bore
Gas Lift Valve Body
(c)
Cross section of side pocket mandrel
with gas lift valve installed in pocket
Department of Petroleum Engineering, Heriot-Watt University
25
Figure 16(a) is a schematic illustration of a side pocket mandrel intended for use with
an Injection (or Casing) Pressure Operated (IPO) valve; while figure 16(b) shows the
equivalent mandrel construction details for a Tubing Pressure Operated (TPO) valve.
An IPO valve uses the injection or casing pressure acting on a pressure element
(bellows) mounted in the upper portion of the valve to open and close the valve. A TPO
valve uses the tubing pressure for the same purpose. It can be seen that a different,
internal arrangement of the mandrel is required for the two types of gas lift valves.
N.B. The mechanical design and operation of the gas lift valve are discussed in section 3.7.
Orienting Sleeve
Side Pocket Mandrel Body
Tool Discriminator to ensure gas lift
valve is inserted in side pocket mandrel
Latch Lug to hold gas lift valve in place
Pocket for gas lift valve
Port
Polished Bore
Figure 17
A side pocket mandrel
Figure 17 is a more detailed view of a side pocket mandrel showing some of the design
features which are necessary for valve installation and retrieval by wireline (see also
figure 18). These include:
26
Gas Lift
Figure 18
Wireline installation of a
gas lift valve in a side
pocket mandrel
(a)
(b)
(c)
(a) Gas lift valve being
run into well on wireline
tool sling.
(b) Wireline "kickover" tool
placing gas lift valve into
side pocket mandrel.
(c) Recovering wireline tool
string after latching gas lift
valve in place in side pocket
mandrel.
3
(i) An orientating sleeve. This contains a device (e.g. a vertical slot) which fits
into its counterpart (finger) on the kickover tool so that gas lift valve and knuckle of
the kickover tool are correctly aligned with the pocket orientation. This allows the
tool to insert a gas lift valve into the mandrel or to recover the gas lift valve, as
appropiate.
(ii) A tool discriminator which guides the gas lift valve into the pocket while
deflecting larger diameter tools back into the main tubing.
Department of Petroleum Engineering, Heriot-Watt University
27
(iii) A latch ring on the valve locks underneath the mandrel’s latch lug, securing the
valve in place. This prevents the valve becoming detached from the mandrel once
the well is placed on production.
(iv) Fluid tight seals are created between the gas lift valve seal elements and the
pocket’s polished bore. These are situated both above and below the casing port for
IPO valves or the tubing port (TPO valves).
6.1 Other Uses of Side Pocket Mandrels
The gas lift valve Side Pocket Mandrels can also be replaced by tools with the
following functions:
(i)
Chemical injection valves;
(ii) Differential dump/kill valves;
(iii) Circulating valves;
(iv) Circulating sleeves;
(v) Dummy valves;
(vi) Water injection control valves.
7. GAS LIFT VALVE MECHANICS
The mode of action of the upper gas lift valves, in which the top valves are designed
to open and close to allow the fluid in the casing/tubing annulus to be unloaded so that
deep gas injection can be achieved, was described in the previous section. The
operating valve is different, being designed to allow for a continuous flow of gas. The
upper gas lift valves have ports sized to pass only the required volume of gas, limiting
the rate at which the unloading takes place. A larger port is often installed in the
operating valve so that gas injection can be increased, if dictated by future well or
reservoir conditions. Dummy valves are installed in the “Bracketing Envelope”
where “live” valves are currently not required.
The following sections discusses the valve’s mechanical construction.
7.1 Casing or Inflow Pressure Operated (IPO) Valves
A schematic diagram of a casing pressure operated valve is shown in figure 19. The:
28
Gas Lift
3
Plug (Removed to charge dome with nitrogen)
Nitrogen Charged Dome
Pdome
Bellows
Abellows
Stem
Spring
(To prevent
excessive
bellows collapse)
Pcasing
Ball or Stem Tip
Square Edged Seat
Port
Choke
Chevron seals
(Forms seal against polished
bore in body of side pocket mandrel)
Figure 19
Schematic diagram of a
casing pressure operated
valve
Aport
Check Valve
(Prevents back flow of produced
fluids from tubing into valve)
Ptubing
(i) dome or top section of the valve is charged, via the plug, with nitrogen to the
required pressure (Pdome).
(ii) nitrogen charge acts on the bellows; exerting a force pushing the ball against the
choke or port and halting the flow of lift gas into the tubing.
(iii) spring prevents damage to the bellows due to excessive collapse when exposed
to forces much greater than those generated by the nitrogen charge. Such excessive
collapse would result in the bellows loosing their elastic response to pressure
changes. If this occurs the gas lift valve needs to be changed.
(iv) choke or port prevents excessive (gas) flow rates.
(v) check valve prevents formation fluids flowing from the tubing into the annulus.
Simple, mechanical considerations allow us to derive the following equations:
The Closing Force, Fc, tending to seat the ball is given by:
FC = Pdome * Abellows
Where Ax refers to the Area of component x and Py refers to the Pressure at point y.
See figure 19 for further explanation of nomenclature.
The Opening Force (FO) is made up of two components: FO1, and FO2, where
FO1 = Ptubing * Aport
FO2 = Pcasing * (Abellows - Aport)
Department of Petroleum Engineering, Heriot-Watt University
29
and
FO = FO1 + FO2
The Opening and Closing forces are equal just before the valve opens.
Pdome * Abellows = Pcasing * (Abellows - Aport) + Ptubing * Aport
Pcasing =
or
Pdome − Ptubing (A port / A bellows )
1 − (A port / A bellows )
Thus the key factors controlling the gas lift pressure required to open the valve are the
dome and tubing pressures and the ratio of the bellows and port areas. IPO valves
normally have the ratio (Aport/Abellows) set as small as practical.
7.2 Dome Pressure Calibration
The dome is charged with nitrogen and the gas lift valve installed in a Test Rack placed
in a temperature controlled enclosure set at 60ºF (see Figure 20). The simulated Ptubing
is at atmospheric pressure (or O psig). A gradually increasing gas pressure is applied
to simulate Pcasing. The valve opening pressure (Popening) is recorded as the pressure
when gas begins to flow.
Pdome
Abellows
Gas Pressure
(Simulates Pcasing)
Aport
Simulated Ptubing
Application of the above force balance equation gives:
Popening = Pdome/(1 - Aport/Abellows)
This can be used to confirm that the manufacturing of the gas lift valve is on
specification and that the dome nitrogen pressure, Pdome, has been set correctly.
The valve closing pressure (Pclosing) may be measured by pressurising both the injection
and tubing sides so that the valve is fully open. The pressure on the tubing side is
reduced and valve closure recognised as the pressure at which Pcasing no longer
30
Figure 20
Test rack for gas lift valve
Gas Lift
3
decreases in line with Ptubing. The design of the valve dictates that Pclosing will be lower
than Popening, since starting from the open valve situation means that the injection
pressure acts on the complete bellows area.
The valve spread is defined as the difference between the valve test rack opening and
closing pressures i.e. (Popening - Pclosing). It is a measure of the difference between the
effective area of the bellows and the port. Figure 21 illustrates a typical valve spread
and its variation with changes in the casing pressure.
Gas injection rate (MMscf/d)
Critical flow through a square edged orifice
(same diameter as choke installed in gas lift valve)
Sub - critical
flow region
Gas lift valve performance with
casing pressures (b), (c), (e) and (d)
Throttling flow region
(e)
Pressure (psi)
P casing (d)
(d)(c)(b) Pcasing
(a)
P casing (c)
Figure 21
Valve performance
P casing (b)
Valve spread at various values of Pcasing
7.2.1 Temperature Correction
A compressible fluid (gas) is required for charging the dome (to avoid valve rupture
as the fluid heats up when it is run into the well or the well heats up when it is placed
on production). Nitrogen is used for this purpose since it is non-corrosive, inflammable
and the temperature effect on the pressure is well known.
P2 = P1 * Tc
and
where
Tc =
{1 + 0.00215 * ( T2 − 60)}
{1 + 0.00215 * ( T1 − 60)}
P1 = gas Pressure at Temperature T1 (oR),
P2 = gas Pressure at Temperature T2 (oR),
Tc = Temperature Correction factor
In practice, part of the dome volume is filled with a silicone liquid to dampen
vibrations created by the gas flow through the valve. This silicone liquid occupies part
of the volume of the dome, reducing the effective volume of the nitrogen charge.
However, it will also expand as the valve heats up, giving an additional pressure
increase to that calculated above. This secondary correction becomes less important
for the larger (1.5") valves where the volume of liquid compared to the gas volume is
relatively less important.
Department of Petroleum Engineering, Heriot-Watt University
31
7.3 Valve Flow Performance
The operating valve at the bottom of the gas lift string, through which gas will be
continually passed, is normally equipped with square edged, orifice choke. The
resulting flow performance is shown in figure 21, curve (a). The gas flow rate passing
through the choke increases with increasing pressure difference between the casing
and tubing until a maximum value is reached when the critical flow rate is achieved.
At this point the gas flow velocity has become supersonic and the volume of gas passed
does not increase further with increasing pressure differential. The valve is often
referred to as being “choked” at this point. Orifice flow is described by the “ThornhillCraver equation” which relates flow capacity to the difference between injection gas
and tubing pressure at the valve depth and the port (orifice) size (see chapter 3.13). The
orifice size determines the maximum or “choked” volume of gas that can be injected
to aid lifting the well fluids to the surface.
This equation is often (incorrectly) used to describe the gas flow through an IPO or
TPO gas lift valve. The Thornhill-Craver equation assumes that the flow through the
valve’s port is unimpeded by the presence of the ball, i.e. it assumes the valve is fully
open and the ball does not impede the passage of gas. (Dynamic) Valve Response tests
(see below) have shown that the Thornhill-Craver equation can overpredict the actual
flow rate by more than 200%.
Further, simple gas lift design procedures assume that the valve will shut immediately
the casing (IPO valve) or tubing (TPO valve) pressure drops below the preset value.
This “instantaneous-closure” assumption is not correct, creating an even greater error
when Proportional Response Valves (see section 3.7.4) are used.
7.3.1 Dynamic Valve Performance
Dynamic testing of the valve, in which the injection gas flow rate is measured for a
fixed casing pressure and a range of tubing pressures, is required to properly
understand the valve’s performance. These dynamic valve performance characteristics
describe the:
(i) valve’s ability to start passing gas when the ball first lifts from its seat,
(ii) rate of increase in the gas flow rate as this clearance increases (flow capacity),
(iii) corresponding decrease in gas flow rate as the valve closes and
(iv) resistance to vibration under flowing conditions.
The valve's load rate (measured in units of psi/inch) is the pressure difference
required to move the stem a given distance; while the valve spread is the difference
between the opening and closing pressures. The shape of the performance curve
represents the valve’s sensitivity to pressure.
The American Petroleum Institute has published a Recommended Practice (Number
IIV2 or API RP IIV2) which describes a standardised testing procedure to measure a
particular gas lift valves dynamic performance. The standard is summarised as follows:
32
Gas Lift
3
The valve is installed in the test equipment after being set up with a suitable dome
pressure using the manufacturer’s recommended procedure. Measurements are made
of the:
(i) valve opening pressure,
(ii) pressure increase required to move the stem to its value of maximum travel and
(iii) position of the stem for various intermediate pressures.
The maximum effective travel distance influences the volume of gas that the valve can
pass. The maximum effective stem travel (test one) needs to be measured only once
for each valve design.
The flow capacity of the valve (as a function of stem position) is tested in a separate
test designed to measure the valve’s flow coefficient (or Cv). In this test the stem is
adjusted to various positions between 5% and 100% of its maximum value and the
flow rate measured at 5 differential pressures - one of which should display choking
flow (see above). This flow coefficient test and the dynamic test (see below) need to
be carried out for each size of choke (or port) that are planned for installation in the
valve.
Gas injection rate (MMscf/d)
Critical flow through a square edged orifice
(same diameter as choke installed in gas lift valve)
Throttling flow region
(e)
Pressure (psi)
Figure 21
Valve performance
Sub - critical
flow region
Gas lift valve performance with
casing pressures (b), (c), (e) and (d)
P casing (d)
(d)(c)(b) Pcasing
(a)
P casing (c)
P casing (b)
Valve spread at various values of Pcasing
A final test is performed for several set pressures in which the injection (or tubing)
pressure is slowly decreased (or increased). The flow rate is measured at a number
of pressures until the valve closes. Either procedure should give the same result. The
results of such a testing scheme is illustrated in figure 21. Here, the valve’s flow rate
has been measured at four casing pressures {Pcasing (b), (c), (d) and (e)} and a variety of tubing
pressures. N.B. A fixed port size and dome pressure was used for all these tests.
Curve (b) shows the performance with the casing pressure adjusted to Pcasing(b). The gas
flow rate is initially zero, even though the valve is open, since the tubing pressure is
set to the same value. Gas starts to flow into the tubing when the tubing pressure is
reduced, though the flow rate will not increases so rapidly as recorded for curve (a)
- the equivalent test with a square edged orifice.
Department of Petroleum Engineering, Heriot-Watt University
33
N.B. Pcasing(a) and Pcasing(b) have the same numerical value in figurer 21. It can be seen
that the gas flow rate passed by the valve {curve(b)} initially increases at a slightly
lower rate than that for the orifice {curve (a)} of the same diameter as the gas lift
valve’s port. This difference is due to the ball, positioned slightly downstream of the
valve’s port (see figure 19), interfering with the gas flow pattern through the valve.
Further reductions in tubing pressure will increase the gas flow rate until a maximum
is reached. After this point, the valve’s closing force (FC) becomes relatively greater
than the opening force (FO); allowing the ball to move closer to its seat and reduce the
gas flow. This gas flow reduction occurs in a (relatively) linear manner (the “throttling
region”) until it has decreased to zero at Ptubing(b). The slope of the gas flow rate / tubing
pressure plot in the throttling region is a function of the gas lift valve’s mechanical
design.
The dynamic valve performance at a slightly lower casing pressure {Pcasing(c)} has a
similar shape (determined by the gas lift valve’s mechanical design). The maximum
gas flow rate and the corresponding valve “spread” are now smaller since the valve
opening force (FO) has been reduced by the lower value of the casing pressure.
This process continues further when the casing pressure is reduced to Pcasing(d). Here
only a small gas flow rate is recorded over a small range of tubing pressures. Once
the casing pressure is set equal to, or lower than, Pcasing(e), the valve will no longer open
and gas can not flow into the tubing.
N.B. It will be discussed later (13.4) that each gas lift valves in a completion string
should be set up with an opening pressure slightly lower (typically 50 psi) than the
setting of the valve above it. This progressive lowering of the setting pressure for the
deeper valves aids the closure of the upper valves so that the gas injection is
progressively transferred to lower valves during the unloading process. It should be
noted here that IPO valves do not control the annulus pressure. A drop in the annular
pressure can only occur when the gas injection rate, possibly through both valves, is
greater than gas flow rate into the casing annulus at the surface.
7.3.2 Valve Performance Flow Model
Sophisticated, computerised gas lift design programs require quantitative models
which describe the above flow tests. The Thornhill-Craver model {figure 21, curve
a}, the previous industry standard, describes the flow through orifices reasonable well
but does not capture the throttling action of “real” gas lift valves. (The original
publication proposed the equation to describe flow through surface chokes with fixed
beans!) An extension of this model, by Winkler and Eads, uses the same formula but
substitutes the actual choke or port area open to flow rather than the fully open valve
used by Thornhill-Craver. Despite using a constant discharge coefficient for all stem
positions, the flow curves generated by the Winkler-Eads model have the correct
shape {figure 21(b) et seq.}. It works well (accuracy 20-30% over the full pressure
range) for valves with a port size of less than 0.25", though the errors rise to >100%
when ports >0.25" are installed.
Both the Thornhill-Craver and the Winkler-Eads equations have an idealised, mechanistic basis - the equations are not “tuned” using the actual flow test data measured in
34
Gas Lift
3
the API test procedure described above. These gas flow rate measurements can be
represented by:
(i) a Flow Coefficient (Cv). This is a measure of the gas flow rate as a function of
pressure differential across the valve
(ii) the distance of the stem from its seat.
The Flow Coefficient (Cv) also determines the pressure rates at which choke flow
occurs across the valve as a function of the stem position. This is an important factor
since most valves show choke flow when the gas injection is being transferred to the
next lower valve. A higher value of Cv implies a greater valve flow capacity. It’s value
is determined by the:
(i) valve inlet port location with respect to the seat,
(ii) actual profile of the seat or valve,
(iii) design of the cross over ports and
(iv) any flow restrictions downstream of the port e.g. the check valve.
The measured Cv values are useable over a wide range of temperature and pressure
conditions, as well as for both liquid and gas flow. They can be used for all valves of
the same type and port size. However, any design change which alters the flow path
through the valve will effect the value of Cv and require that the flow test be repeated.
API RP IIV2 recommends a simplified approach to analysing this flow data using Cv
values. Alternative approaches have been proposed by:
(i) Tulsa University Artificial Lift Project (TUALP) using a statistical approach to
data analysis. This can only be applied over the pressure range used for the flow tests.
(ii) Valve Performance Cleaning-house (VPC). The VPC method uses a mechanistic
model based on the force balance equation and load rate to calculate a value of Cv;
which is then adjusted so as to be able to reproduce the dynamic test data.
Department of Petroleum Engineering, Heriot-Watt University
35
7.4 Proportional Response Valves
The design of a Proportional Response Valve (see figure 22) involves:
Plug (Removed to charge dome with nitrogen)
Pdome
Nitrogen Charged Dome
Spring
(To provide
proportional
response)
"Stop" to prevent excessive
bellow collapse
Bellows
Abellows
Stem
Pcasing
Large Ball
Tapered
Seat
Port
Choke
Chevron seals
(form seal against polished bore
in body of side pocket mandrel)
Aport
Check Valve
(prevents back flow of produced
fluids from tubing into valve)
(i)
Ptubing
The addition of a spring to counterbalance the force trying to expand the bellows,
(ii) A “stop” to prevent excessive bellows collapse,
(iii) A larger ball at the tip of the stem and
(iv) A tapered seat instead of a square edged seat.
The proportional response valve has more parts than the “traditional” valve. It uses
more “O” rings, which are not only expensive but also prone to failure i.e. Proportional
Response Valves should only be specified when definite operational advantages have
been identified.
7.5 Dynamic Valve Response and Gas Lift Completion Modeling
Simple (manual) gas lift completion design programs assume that the valves will open
fully at (casing or tubing) pressures slightly greater than the set value and will
completely close when the pressure drops below this value. They assume that the well
“jumps” immediately from steady state operation when lifting through one (fully
open) valve to a second, steady state condition when lifting through the next deeper
valve. In practice:
(i)
the volume of gas injected through a given valve will vary with time.
(ii) the well will only reach steady state inflow a considerable period of time after
the unloading process has finished.
36
Figure 22
Proportional response gas
lift valve
Gas Lift
3
(iii) the combination of changes in (i) and (ii) will result in continuous changes in
the well’s Gas Liquid Ratio (GLR) during the unloading process. This is especially
true when any fluid lost to the formation during the workover or the early stages of
well unloading has to be produced first before live formation fluids can enter the well
and improve the well outfllow.
(iv) there is a significant volume of gas in the casing/tubing annulus. Pressure
changes will not occur instantaneously due to the limited gas supply rate at the
surface or the finite gas flow rate through a particular valve.
(v) the valves can open and close a number of times during the unloading of a multivalve completion string e.g. as well temperature changes.
(vi) the temperature changes during the unloading process (being neither the
geothermal nor the steady state flowing temperature gradient).
It can be appreciated that a transient rather than a steady state computer program is
required to simulate the unloading process. Such programs must accurately predict
the development of the above parameters as a function of time. The output of the
program will include, as a function of time, the:
(i) tubing and annulus pressure at valve depths as well as surface casing (annulus)
and tubing head pressure,
(ii) flow rate of liquid and gas through open valves,
(iii) depth of liquid levels in tubing and annulus,
(iv) volume of fluids (gas and liquid) produced to the surface and
(v) inflow from the reservoir (after making allowance for any liquid lost to the
formation, as discussed above).
One commercially available program is DynaliftTM marketed by Edinburgh Petroleum
Systems. It has been shown to be useful for troubleshooting problematic wells e.g. the
prediction of when the following difficulties can be expected during well unloading
- erosion of (gas lift) valves, severe slugging, excessive well stabilisation times etc.
7.6 Well Stability
Dynamic Well Simulation programs can also be used to evaluate potential causes of
lift problems once the well has been placed on production. These valve problems
manifest themselves by the:
(i)
well not producing or producing at lower than expected rate,
(ii) unstable well behaviour (casing and/or tubing head pressures oscillate regularly)
and
(iii) excessive unloading and well stabilisation times.
Department of Petroleum Engineering, Heriot-Watt University
37
The quickest method to correct the problem is to identify which valve is the cause of
the production difficulty and then replace it with a correctly calibrated replacement.
The alternative, to replace the gas-lift valves in a random, ad hoc manner is often much
more time consuming and may never solve the problem.
Comparison of the actual and the predicted well behavior allows identification of the
type of problem being experienced as well as which gas lift valves is the cause of the
improper operation. Typical problems which have to be diagnosed include:
(i) valves that have become either:
(a) (partially) plugged or
(b) enlarged (cut out)
(ii) valves set to:
(a) incorrect dome pressures or
(b) operating incorrectly due to mechanical problems and
(iii) multi-pointing (injection through more than one valve at the same time). This
is often due to an upper valve opening and closing.
An example of a specific operational problem that could be diagnosed is the case when
the surface gas injection rate is greater than the single valve flow rate calculated from
the annular and tubing pressures. Possible causes include:
(a) gas injection through two or more valves,
(b) the valve choke (or port) of the operating valve was enlarged (cut out) during
unloading or
(c) an upper valve is opening and closing.
A quantitative evaluation of the flow rates and the tubing and casing pressure values
and their stability are required to differentiate between the above causes, as well as to
identify the specific problem valve.
8. GAS LIFT DESIGN PROCEDURES
This chapter will discuss the procedure used to design the gas lifted well. It will
combine the flow performance of operating valves discussed above with the concepts
discussed in Section 1 on multiphase flow and Nodal Analysis. Table 4 lists the well
data required to initiate this design process which will enable us to specify the
optimum:
38
Gas Lift
Table 4
Data set for gas lift design
Tubing Outside Diameter
4.5 in
Tubing weight
12.6 lb/ft
Deviation Survey
Vertical Well
Target Production Rate
5000 bfpd
Watercut
60% vol.
Produced Water Density
1.06 g/cm3
Gas Relative Density
0.65
Packer Setting Depth
9,500ft.
Mid Perforation Depth
10,000ft.
Wellhead Flowing Pressure
100 psig
Shut In Bottom Hole Pressure
3,700 psig
Productivity Index
0.50 STB/d/psi
Oil Gravity
37 API
Bubble Point
1200 psig
Gas Oil Ratio
400
Gas Injection (Casing) Pressure
1500 psig
Gas Available For Injection
3 MM scf/d
Ambient Temperature at Surface
70 F
Flowing Temperature at Surface
140 F
Temperature at Mid Perforation
240 F
Kill Fluid Gradient
0.45 psi/ft. *
3
* In Annulus / Tubing for Unloading Calculations
(i)
Tubing size,
(ii) Injection gas supply parameters (pressure and volume),
(iii) Installation depth of operating valve and
(iv) Separator pressure.
The first item is to check that gas lift valves can be placed at suitable distances above
the operating valve to ensure that the well can be unloaded using the gas lift parameters
chosen (see section 3.5). The gas lift design will have to be changed if it is found that
the well cannot be unloaded. Further, the design will be influenced by the gas lift
design philosophy. For example:
(i)
A Maximum Production / Expected Case Design would:
(a) maximise production by injecting gas as deep as possible,
(b) install extra gas lift valves around the operating valve to allow the depth
of the operating valve to be adjusted and
(c) budget for (some) well entries to be required since the well may stop
producing if well inflow conditions change.
(ii) Worst Case / Robust Design would be designed to:
Department of Petroleum Engineering, Heriot-Watt University
39
(a) always work by placing the operating valve at a relatively shallow depth
compared to that normally achieved with to the available gas lift pressure.
(b) results in a lower production rate. Deferred production probably also
implies a reduced reserve recovery.
The philosophy chosen for a particular well / field will depend on the operational
conditions and the cost scenario.
8.1 An Example Design - optimising the performance of a gas lifted well
Sandface flowing pressure
The first requirement is to study the well’s in- and out-flow and confirm that artificial
lift is required. This is illustrated in figure 23. It can be seen that the well production
rate will decrease rapidly once well Edinburgh-1 begins to produce water. It will stop
flowing when the water cut reaches 30%. Any reduction in the flowing sand face
pressure due to depletion of the reservoir pressure and / or the development of a near
wellbore “skin” due to Formation Damage will reduce the well production rate. This
will also reduce the critical water cut at which natural flow ceases. It should be
remembered that the form of the Inflow Performance Relationship used to describe
fluid inflow from the formation (e.g. Vogel or “straight line PI”) will have an impact
on these calculations.
Tubing outflow at various water cuts 80%
60%
40%
20%
0%
0%
Reservoir inflow at various water cuts
80%
Well production rate
We will assume that Gas Lift has already been identified as the most suitable form of
artificial lift for this well and an operating valve has been installed at a suitable depth.
The impact of the injection of varying amounts of gas on the well production rate is
evaluated in Figure 24. Figure 25, a plot of production performance against lift gas
injection rate can now be derived. It can be seen that the injection of the first 1.25
MMscf/d of lift gas leads to a rapid increase in production rate. Further increases in
the rate of gas injection yields ever reducing benefits until a maximum production rate
is reached at 2.5 MMscf/d. The same data are plotted in Figure 26. Here the lift gas
injection rate is represented as a Gas-Liquid Ratio - the sum of the Produced fluid gas
liquid ratio and the gas injection rate divided by the (gross) liquid production rate. As
expected, the graph shows a maximum production at the 2.5 MMscf/d gas injection rate.
40
Figure 23
Performance well
Edinburgh - 1 at varying
water cuts
Gas Lift
All at
40%
Watercut
0.5
1.25
3.75
2.5
Reservoir
Inflow
Lift gas injection rate (MM scf/d)
Well production rate
Well production rate
Figure 24
Sensitivity study for tubing
outflow to lift gas injection
rate at 40% water cut
Sandface flowing pressure
Natural flow
3
Figure 25
Production performance
and gas lift injection rate
0
1.25
2.5
3.75
Lift gas injection rate (MMscf/d)
g
Gas - Liquid Ratio
3.75 MM scf/d
Figure 26
Production performance
against gas liquid ratio
2.5 MM scf/d
(Technical) Optimum gas
injection rate
1.25 MM scf/d
0.5 MMM scf/d
Maximum production rate
Well production rate
Formation Damage - or the presence of a near wellbore skin - will reduce the well
productivity. Figure 27 is similar to Figure 24, except that the reservoir inflow from
a damaged well (skin = +6) has also been plotted. Figure 28 not only illustrates this
loss in well productivity due to the presence of the skin, but also indicate that the
conclusion drawn as to the optimum gas injection rate is unchanged.
Department of Petroleum Engineering, Heriot-Watt University
41
j
Natural flow
Sandface flowing pressure
x
40%
Watercut
0.5
x
1.25
3.75
x
2.5
x
x
x
Reservoir
inflow
skin = +6
Reservoir
Inflow
skin = 0
Figure 27
Sensitivity study for tubing
outflow to lift gas injection
rate at 40% water cut for
an impaired well
Lift gas injection rate (MM scf/d)
Well production rate
Well production rate
Skin = 0
Skin = +6
0
1.25
2.5
Figure 28
Production performance
and lift gas injection rate
for an impaired well
3.75
Lift gas injection rate (MMscf/d)
Well production rate
Increasing depth of injection
Figure 29 examines the effect of changing the depth of lift gas injection. A deep
injection point creates a greater drawdown leading to an increased production rate and
a more efficient GUF (Gas Utilisation Factor or Gross production rate / Gas Injection
Rate).
0
1.25
2.5
3.75
Lift gas injection rate (MMscf/d)
Figures 30 and 31 examine the sensitivity of the well production rate to an increase in
the diameter of the production tubing. The production increases rapidly as the tubing
OD increases to 4.5 in, but then decreases for the largest tubing (5.5 in). Excessive
slip between the gas and liquid phases occurs at the lower flow velocity associated
with the larger tubing.
42
Figure 29
Production performance
and lift gas injection depth
Gas Lift
3
2.875" Tubing OD
Sandface flowingpressure
Figure 30
Sensitivity study of tubing
outflow at various to tubing
diameters
3.5"
4.5"
5.5"
Well in
fl
40% W ow at
ater c
ut
Well production rate
Figure 31
Production performance
and tubing diameter
Well production rate
4.5"OD
5.5"OD
3.5"OD
2.875"OD
2
2.5
3
3.5
4
4.5
5
Tubing internal diameter (in)
We showed in Chapter 1 (Well Performance) that separator pressure affected the
outflow performance of naturally flowing wells. The same is true for gas lifted wells
- Figure 32 show how the well outflow performance improves as the separator
pressure is decreased. This is shown explicitly in Figure 33, where the production rate
is plotted as a function of separator pressure. In general, the lower the separator
pressure, the higher the production rate.
Figure 32
Sensitivity study of tubing
outflow to separator
pressure
Sandface flowing pressure
Separator pressure (psi)
250
200
150
100
75
50
25
We
l
40% l inflo
Wa w at
ter
cut
Well production rate
Department of Petroleum Engineering, Heriot-Watt University
43
Well production rate
0
50
100
150
200
250
Separator pressure (psi)
Figure 33
Production performance
and separator pressure
It should be emphasised that, in addition to the separator pressure, a back pressure on
the well performance is created by pressure losses across:
(i)
Valves (should be fully opening type to minimise any pressure losses)
(ii) Choke body and/or choke nipples, if installed
(iii) The flowline (particularly pipe elbows) and long narrow diameter, tortuous
flow lines with the separator mounted at a higher elevation than the wellhead
cause an increased back pressure.
(iv) The wellhead itself (a Y block design will have a reduced pressure loss
compared to the standard T block for a high rate well)
A further parameter that needs to be optimised is the lift gas injection pressure - Figure
34 illustrates the effect this has on the well outflow performance while Figure 35
records the production rate with changes in injection gas pressure.
400
Sandface flowing pressure
600
44
1200
1500
800
1000
Gas injection lift pressure
W
40 ell in
% flo
W w
at at
er
cu
t
Well production rate
Figure 34
Sensitivity study of tubing
outflow curves to gas
injection pressure
3
Well production rate
Gas Lift
Figure 35
Production performance
and gas injection pressure
500
750
1000
1250
1500
Gas injection pressure (psi)
N.B. This well is capable of production with low injection gas pressures because the
depth of the operating valve was not too deep and the well was almost capable of
producing under natural flow.
The conclusions reached from the above sensitivity studies are valid for the Edinburgh1 well producing under the chosen combination of Reservoir Pressure and water
production studied. The conclusions need to be tested for robustness by repeating the
sensitivity analysis for combinations of water cut and reservoir pressures that are
expected to be encountered during the well’s lifetime. Planned recompletions e.g.
extension of the perforated interval, should also be included in this. Figure 36 is a
typical example of the deterioration of the Reservoir Inflow Performance at various
times (T1, T2 and T3) under the twin influences of the decrease of reservoir pressure
and increasing water cut. The tubing (outflow) performance is plotted for the
optimum gas injection rate that maximises the well production for each reservoir
inflow curve i.e. the optimum gas injection rate changes with time. The GUF will
decrease with time due to the:
Gas injection rate adjusted
for each reservoir inflow curve
in order to maximise production
depletion
x
T3
x
x
oir
rv
se
low
inf
Figure 36
Performance of well
Edinburgh - 1 at various
times
T2 Reservoir
Re
Sandface flowing pressure
T1
Well production rate
(i)
Increasing average liquid density (increasing water cut).
(ii) Decreasing (own) gas production rate (decreasing net oil production and some
reservoir drive mechanisms e.g. solution gas) and
(iii) Reducing reservoir pressure.
Department of Petroleum Engineering, Heriot-Watt University
45
Figure 37 records the decreasing well production resulting from the changing well
inflow and tubing outflow performance curves, being derived from the data presented
in Figure 36.
Well production rate
x
x
x
T1
T2
Time
T3
One further variable to be discussed is the choice of the multiphase flow correlation
used to calculate the well outflow performance. This can have a great influence on the
results, as discussed in Chapter 1 (Well Performance). The flow correlation can be
chosen on the:
(i)
basis of production experience in other, similar wells in the same field,
(ii) theoretical considerations based on a comparison of the well type with the
parameters originally used to generate the correlation or
(iii) comparison of predicted and actual well performance.
The latter is the best choice when (field) data is available. The actual and predicted
gas lifted well performance need to be compared at a range of gas lift injection rates.
{Chapter 1 discussed how the flow correlation for a naturally flowing well could be
chosen by comparing the predicted and the measured pressure traverse (the pressuredepth profile).} Figure 38 illustrates how the use of an inappropriate correlation can
lead to false expectations as far as the well production rate is concerned. Flow
correlation number 4 refers to “Hagedorn and Brown”, probably the one most
frequently used to describe the performance of gas lifted wells. The preferred flow
correlation for the gas lifted well may be different from that previously used to
describe the same well’s performance under natural flow.
46
Figure 37
Performance of well
Edinburgh - 1 as the
reservoir pressure depletes
Gas Lift
3
Flow correlation
1
Table 5
Summary of chosen gas lift
parameters
Well production rate
Figure 38
Sensitivity of production
performance under gas
injection to choice of
multiphase flow correlation
2
3
4,5
6
Actual well data
0
1.5
3
4.5
Lift gas injection rate (MM scf/d)
Tubing Size
4.5"(O.D)
Separator Pressure
As low As Possibleφ
Gas Injection Pressure
1000 or 1200 psi*
Gas Injection Rate
1MM scf/d
Flow Correlation
Hagedorn and Brown
* Depends on detailed cost calculations
φ
Some operators have profitably reduced back pressure on the
well by "twinning" the flow line to the separator
Table 5 summarises the conclusions from the above gas lift design exercise. Chapter
10 of this module discusses the need to develop a well / field model to model current
and predict future production performance as well as to systematically monitor any
differences between this prediction and the actual measured values. Such a monitoring
process will gather the basic data needed to test whether the chosen inflow and outflow
correlations are still appropriate. For example, table 6 illustrates the correlations used
by one operator to achieve the most accurate predictions based on such a monitoring
programme.
Table 6
Preferred correlations
Inflow
< 50% Water Cut
Vogel
>50% Water Cut
"Straight Line" PI
Outflow
< 10,000 bfpd
Hagedorn and Brown
> 10,000 bfpd
Duns and Ross
8.2 An Example Design - Gas Lift Unloading Calculations
The section above describes how the operating valve location and the design
objectives for the gas lift system can be chosen. An illustrative description of the
Unloading Process was given in section 3.5 and Figures 7 to 15. Concerns, to be
avoided by careful design of the unloading process to be followed in a particular well
are that the:
(i) Well does not unload resulting in the gas being injected at a shallow depth. This
can be overcome by employing a conservative design philosophy to valve spacing.
(ii) Well does not lift at an optimal rate since the above conservative design
philosophy has resulted in the operating valve being situated at too shallow a depth.
Department of Petroleum Engineering, Heriot-Watt University
47
This can be overcome by carrying out the gas lift design using the expected well
properties. However, if an incorrect choice of well conditions are made, this may
result in the well being unable to unload).
(iii) The well is stymied i.e. there is a combination of conditions under which all the
gas lift valves close and no more gas can be injected.
These two approaches are compared in the worked example summarised in section 3.13.
8.3 Further Gas Lift System Considerations
Two further points need to be evaluated.
(i) Flow Velocity. Corrosion / erosion of the tubing can become excessive if a
critical flow velocity. Exceeding this limit will lead to excessive maintenance costs
e.g. early replacement of damaged tubing or accessories. This is particularly true if
significant quantities of formation solids e.g. sand are being co-produced with the
formation fluids. One operator sets a maximum velocity of 12 ft/s in 3.5 in tubing
for these conditions. The American Petroleum Institute published recommendations
in this area tend to be somewhat conservative, one needs to compare them with local
experience. The critical value for the onset of corrosion / erosion can be very
sensitive to the composition of the produced fluid; in particular the concentrations
of carbon dioxide and hydrogen sulphide in the gas phase.
(ii) Shear Rates. Exposing water and oil mixtures to high shear rates can lead to
the formation of emulsions. The produced fluid experiences increased velocities
when gas lift is introduced. This is due to the improved production rate and the
increased gas oil ratio. The increased shear rates can result in the formation of
viscous emulsions, especially when a choke or bean is included in the surface flow
lines, though it is difficult to predict the precise conditions which will lead to
emulsion formation. The well model will need to be adapted to the field data by
“tuning” the fluid properties if emulsion formation is observed.
8.4 Further Gas Lift System Calculations
(i)
Annulus Gas Pressure with Depth
The maximum depth at which gas can be injected into the tubing is controlled by the
surface (casing) gas pressure, the pressure drop across the gas lift valve and the tubing
pressure gradient. Most lift gas systems operate at between 1000 and 1200 psi and are
filled with a natural gas with a high methane content. A useful “rule of thumb” is that
the gas pressure gradient can be approximated to 30 psi/1000 ft. A more exact value
can be calculated from the equation:
 RD gas * L 
PL = PS exp 

 53.34 * Tav * Z 
where:
48
PL
PS
L
= gas pressure at depth L (psia)
= surface gas pressure (psia)
= true vertical depth (ft)
Gas Lift
RDgas
Tav
Z
Pav
3
= gas relative density (to air)
= average temperature of gas column (°R)
= gas compressibility factor at Tav and Pav
= {PS + PL}/2
(ii) Stored Gas Volume
A considerable volume of gas is stored in the casing/tubing annulus. It can be
estimated from the gas equation:
 P *T 
V = Vannulus *  av

 Z * P * Tav 
= the gas volume at standard conditions (ft3)
= total annular volume (ft3)
= average gas pressure in annulus (psia)
= pressure at standard condition (14.3 psia)
= temperature at standard condition (460 °R)
= average temperature in annulus (°R)
where
V
Vannulus
Pav
P
T
Tav
and
Vannulus = 0.022 * L * {dic2 - det2}
where
L
dic
det
= (along hole) length between well head and tubing packer (ft)
= internal casing diameter (in)
= external tubing diameter (in)
(iii) Gas Compressor Power Requirements
The power required by the gas compressor which will supply gas to the lift system can
be estimated from
 P  0.2 
HHP = 2.23*10 * Q  out  − 1
 Pin 

−4
Where
HHP
Q
Pout
Pin
= power required (hydraulic horse power)
= gas injection rate (scf/d)
= compressor outlet pressure (psia)
= compressor inlet pressure (psia)
9. OPERATIONAL PROBLEMS
Typical problems associated with the operation of gas lifted wells include:
9.1 Gas Quality
This can lead to:
Department of Petroleum Engineering, Heriot-Watt University
49
(i) Blockage of gas injection lines due to (solid) hydrate formation during periods of
low ambient temperatures. Such blockages are most easily removed by depressurising the line since hydrates are unstable at low pressures, decomposing into their
constituent components of gas and water. The problem can be avoided by proper
drying of the gas so that water dew point is below the lowest expected ambient
temperature.
(ii) Loss of casing integrity due to use of corrosive lift gas. This is associated with
the presence of the acid gases (carbon dioxide and hydrogen sulphide) in the natural
gas. Guidelines are available to determine at which concentrations corrosion can be
expected. Often part of the produced gas will be sold, only sufficient gas being
retained in the lift gas system to “top-up” losses. In this case, all the gas is normally
treated to sales gas specification. Such specifications will typically limit the carbon
dioxide concentration to 2 or 3%, while hydrogen sulphide levels above 4 ppm are
unacceptable.
In the case that none of the produced gas is sold, it may be practical to tolerate higher
levels of these acid gases. The resulting problems, from these more relaxed, gas
quality specifications which will need to be resolved include:
(i) Natural gas containing carbon dioxide is only corrosive in the presence of liquid
water rather than gaseous water. Corrosion should be looked for in those points
where liquid water can collect (the bottom of pipelines, low points in undulating
pipelines etc.) when the partial carbon dioxide pressure exceeds 30 psi.
N.B. The partial pressure of a gas corresponds to its concentration (expressed as a
molar fraction) multiplied by the total system pressure.
Higher carbon dioxide partial pressures require either the specification of special
metallurgy e.g. 13 chrome stainless steel, tubing or the continuous injection of a
corrosion inhibitor. This corrosion inhibitor will passivate metal surfaces by
forming a protective layer. High gas flow velocities (typically above 30-50 ft/s) will
strip the inhibitor from the metal surfaces and remove this protection.
(ii) Hydrogen sulphide partial pressures of greater than 0.05 psia lead to sulphide
stress cracking of many of the steels used in completion systems and pipelines (in
particular, those with a Rockwell C hardness of greater than 22). A partial pressure
of 0.05 psia hydrogen sulphide corresponds to approximately 50 ppm hydrogen
sulphide in a system operating at 1000 psi. Higher hydrogen sulphide levels can be
coped with by specifying the appropriate metallurgy. Injection of corrosion
inhibitors to passivate metal surfaces is possible on a temporary basis to protect
against excess hydrogen sulphide levels. It is not normally recommended on a
permanent basis since the sulphide stress cracking proceeds rapidly if inhibitor
addition is accidentally halted.
9.2 Solids
These may either be produced from the formation {e.g. sand - see Chapter 7 (Sand
Control) or formed from the produced fluids {inorganic and organic scales - see
Chapter 4 (Formation Damage)}. Wax is the most frequently encountered solid in gas
lift operations}. Whatever the cause, the presence of solids within the production
50
Gas Lift
3
tubing can be expected to result in operational difficulties when carrying out wireline
operations such as changing gas lift valves.
Excessive sand production can only be prevented by a suitable completion design e.g.
installing a gravel pack. It should be repeated here that gas lift is relatively tolerant
towards sand production compared to other forms of artificial lift, such as electric
submersible pumps.
Recent developments in Production Chemistry has resulted in the development of
scale inhibitors which can be injected into the lift gas at the surface and transported
down hole to be injected, via the operating valve, with the lift gas and subsequently
mixed with the stream of produced fluids.
N.B. Scale inhibitors are chemicals which, at low concentrations, inhibit (delay) the
formation of massive deposits of the material in question from the minute “seed”
crystals. These seed crystals are formed when the concentration of the chemical
exceeds the chemical’s solubility limit. This occurs because of temperature and
pressure reductions undergone by the formation fluids as it is produced to the surface.
9.3 Changes in Reservoir Performance
Reservoir Performance will change as the well/field matures. Typically, the water cut
(and, consequently, the hydrostatic head of the fluid in the tubing) will increase and
the reservoir pressure will decrease. These factors, which reduce the well outflow
performance, may also be accompanied by a deterioration of the well inflow
performance due to the relative permeability effect brought about by the increasing
produced water fraction. Since the production target will be to maintain the same net
oil (plateau) production for as long as possible, the well will be produced at a higher
drawdown with an increased gross fluid volume produced.
All these changes will require adjustment of the gas injection system parameters e.g.
increase in the gas injection rate, moving the operating valve to a different depth
(bracketing envelope concept - see section 13) etc.
9.4 Gas Supply Problems
These usually manifest themselves in one of two ways:
(i) The volume of lift gas available is less than that required to produce every gas
lifted well at its (individual) optimum rate. It then becomes necessary to allocate the
gas so that the total field production is maximised. This is discussed in section 10
and an illustrative example studied in section 13.
(ii) Fluctuating compressor suction and discharge pressures. This is due to unstable
operation of the process plant, which may in turn be caused by unstable producing
well operation. For example, tubing “heading” leads to the production of a large
“slug” of liquid followed by a period of gas only or no production at all (see Chapter
9.6 for further details). The surface gas distribution network should also be designed
to minimise the propagation of such pressure transients (see Chapter 4.3).
Profitability will be maximised by stabilising the entire production system (wells and
surface facilities) since this will maximise the reservoir fluid production and minimise
the maintenance costs associated with these fluctuating pressures.
Department of Petroleum Engineering, Heriot-Watt University
51
9.5 Well Start-Up (Unloading)
As mentioned previously, good practices are required to prevent fluid damage to the
lower gas lift valves as the liquid in the casing / tubing annulus is unloaded when
starting up a gas lifted. This is achieved by avoiding excessive fluid flow rates. A
suitable Start-Up procedure, as recommended by the American Petroleum Institute,
is summarised as follows:
(i) Provision should be made to monitor the tubing and casing pressures and
production rate (both liquid and gas) during the unloading process.
(ii) The tubing head pressure should be blown down to the separator flowline prior
to starting the unloading process. Wells that produce under natural flow will be fitted
with a choke. This should be fully opened while undergoing the “kick-off” process.
Continuous gas lifted wells which do not flow naturally normally do not have a choke
installed.
(iii) The Unloading Process is commenced under casing head pressure control (either
manual or automatic). The casing gas injection rate is adjusted so that the pressure
increases at a rate of 5 psi/min until a pressure of 400 psi is reached. An increased
pressure rise rate of 10 psi/min can then be used until the system pressure is reached.
In addition, the maximum, lift gas injection rate should be less than 50% of the
planned design rate. The well’s production rate can only be used to measure the
unloading rate once the liquid level in the tubing reaches the surface. The progress
of the unloading process can not be followed prior to this point in time, unless
arrangements are made to measure the fluid level in the annulus e.g. with an
echometer.
(iv) There is often a drop in casing head pressure when gas is first injected into the
upper gas lift valve and subsequent, lower lift valves. There will be a change in the
gas injection rate, even if this drop in pressure is not observed.
(v) There is often a period of instability during well start up. Apart from the need to
establish a steady state flow regime within the well, there are considerable longer
term changes occurring during well unloading e.g. due to the well heating up. Thus
the pressure exerted by the gas lift valve’s nitrogen charge will increase as the
temperature increases. This can result in a gas lift valve opening and closing more
than once during well unloading.
The gas lift valve settings are normally designed using the operating conditions and
well temperature profile for the final, planned steady production conditions. There are
several published correlations and many commercial programs available to predict
this equilibrium temperature profile. By contrast, there are few programs capable of
making meaningful dynamic predictions during the unloading process, particularly
since the well conditions during unloading tend to be relatively ill-defined.
It is good practice to check that the well unloading process can still proceed with valve
pressure settings based on a cooler temperature profile than that predicted for the
equilibrium producing conditions. Figure 39 illustrates this situation. It indicates how
the unloading calculations should be carried out using a progressively smaller
52
Gas Lift
3
difference between the geothermal and the equilibrium producing temperature profile
for the deeper gas lift valves. The suggested profile for a gas lift string with 8 valves
would show a temperature profile of {geothermal plus 12%, 25%, 37%, 50%, 62%,
75%, 87% and 100% of the "equilibrium temperature value"} respectively.
Surface
equilibrium flowing temperature for valve
setting
estimate of valve temperature for
unloading calculations
1st valve depth
25%
Depth
equilibrium flowing
temperature profile
50%
2nd valve depth
75%
3rd valve depth
4th Operating valve depth
Figure 39
Temperature profiles
during steady state flow and
well unloading
100%
Perforations
Geothermal
temperature profile
Temperature
It is clear that use of the reduced temperature profile is “safer” i.e. that the additional
safety introduced by these measures increases the chance that the well will unload
successfully. It is less likely to become “hung up” or “stymied” {see section 3.9.7 in
this chapter on trouble shooting}. This design can thus be viewed as a more robust
design case. However, its use will normally result in somewhat lower well production
rate as the final lift configuration will be less optimum.
9.6 Well Stability
Well instability can take several forms.
(i) The Casing Head pressure remains constant and the Tubing Head pressure shows
significant, but regular, fluctuations. The size and frequency of these pressure
fluctuations will depend on the particular combination of factors that are leading to
the well instability. An example of tubing heading is illustrated in figure 40. The
casing head pressure is constant at 1000 psi while the tubing head pressure fluctuates
regularly between 60 and 180 psi with a cycle frequency time of 90-120 minutes.
Department of Petroleum Engineering, Heriot-Watt University
53
5P M
6PM
4PM
7P
M
M
3P
8P
M
2P
M
Casing Head Pressure : 1020psi
Twin Pen
MIDDAY
11A M
AM
AM
1A
M
10A
M
TIME
MIDNI
GHT
TIME
11PM
DATE ON
Tu
bi
M
DATE ON
10P
Re
c
1PM
M
9P
f Casing and
ssure.
Pre
ng
go
din
or
2A
M
9A
Tubing Head Pressure : 60 - 180 psi
Well details
M
Liquid flow rate
1300 stb/d
Flow rate gas
630 Mscf/d
Water cut
98%
Cycle frequency
90 - 120 min
Gas injection rate 350 Mscf/d
Well depth
3500ft
Vertical well
8A
M
3A
M
7AM
4A M
5A M
6AM
(ii) Both the Tubing Head and the Casing Head Pressure show regularly fluctuating
values.
Professor Asheim of Trondheim University defined two parameters (F1 and F2)
which can be used to determine whether a stable flow condition exists in either the
tubing or casing / tubing annulus respectively. He showed that unstable flow
conditions have to exist in both flow conduits before the overall well behavior is
unstable.
The tubing (F1) parameter considers whether the changes in tubing inflow and gas
injection pressure resulting from chance fluctuations in the tubing head pressure will
tend to die away and return the well to the current producing conditions (the definition
of stable flow). Thus a stable well condition (F1 > 1) requires that a momentary
decrease in the downhole pressure will result in a greater influx of (high density)
formation fluids relative to that of (low density) injection gas. The combination of
these extra inflows result in a relative increase in the average tubing density,
counteracting the original, momentary decrease in downhole pressure. The well is
thus flowing with a stable condition since it tends to return to the original operating
condition. By contrast, unstable flow (F1 < 1) would have resulted in the well moving
to a different operating point. This stability criterion is equivalent to that discussed
in in the 'Well Performance' module for naturally flowing wells.
The casing / tubing annulus (F2) parameter reflects the response of the flow of casing
gas to momentary fluctuations in the Tubing Pressure opposite the operating valve
when the tubing flow is potentially unstable (F1 < 1). Thus the well is unstable (F2 < 1)
if a momentary decrease in downhole tubing pressure causes the casing pressure to
decrease more slowly than the tubing pressure. The resulting increased pressure drop
across the gas lift valve will lead to a greater inflow of lift gas, providing the valve is
not “choked”. This would tend to decrease the bottom hole pressure further - moving
the well even further from its previous operating point.
54
Figure 40
Example of tubing heading
Gas Lift
3
Well stability can be improved by instituting choke control of the casing gas injection
pressure. It is more difficult for the well to show unstable behavior when the pressure
drop across the surface choke controlling the gas injection rate is greater than the
pressure drop across the downhole operating valve. This arrangement prevents the
gas flowing out of the casing annulus and into the tubing at a faster rate than it is being
replaced by the lift gas supply to the well. {The inclusion of any extra pressure drop
in the gas lift supply system will often decrease the maximum well production rate.}
Practical considerations will often result that, if a well is suffering from unstable
behavior, an operator will find it easier to adjust the surface choke rather than the
change the size of the operating valve’s downhole orifice. The pressure drop across
the surface choke may thus become the greater pressure drop; resulting in greater
tendency for annulus heading (the reverse of what was being attempted). A
momentary drop in tubing press can then result in an excessive gas flow rate through
the downhole orifice. There will be a delay in the pressure reduction being felt at the
surface, due to the large annular volume and high gas compressibility. The lift gas
supply will not increase quickly enough to compensate for this pressure reduction
sufficiently, particularly when its supply is being impeded by a small surface choke.
The pressure at the bottom of the annulus will continue to drop - eventually the falling
casing pressure will reduce the gas supply to the tubing. Once again, there will be a
delay in the response of the surface choke and it will continue to supply excess gas,
leading to a build up in the annulus pressure. These pressure excursions can be
repeated regularly - the size and frequency of the fluctuations depending on the
particular well circumstances.
One further cause of casing and tubing head pressure fluctuations is valve multipointing. This occurs when more than one valve is passing gas i.e. gas is being injected
into the tubing at several different levels. This can represent an apparently stable
situation from the point of view of the Tubing and Casing Head pressures. However
fluctuations may also be induced by one valve repetitively opening and closing due
to the temperature changes induced by the changing flow conditions.
9.7 Dual Gas Lift
The most common dual gas lift configurations are two strings of 2.375in or 3.5 in OD
tubing run inside 7.0 or 9.635 in OD casing respectively. Combinations of other sizes
or even concentric strings are feasible, but rarely used.
The first problem with Dual Gas Lift completions is the well completion operation
itself - there is insufficient width for the gas lift mandrels in the two strings to pass one
another. This is overcome by designing the spacing of the mandrels in the two tubing
strings and the dual tubing running procedure so that the gas lift mandrels do not need
to pass one another during the well completion operation.
A well with a Dual gas Lift Completion can be treated as a normal single completion
if only one tubing string requires gas lift. The difficulty arises in ensuring a proper
distribution of the lift gas between the strings if both strings are to be gas lifted i.e. one
string may take most of the gas while the second tubing is starved of gas.
One solution to prevent the valves in the different strings interfering with the operation
of the other one is to install valves with significantly different operating characteristics
Department of Petroleum Engineering, Heriot-Watt University
55
in the different strings. e.g. One string uses IPO valves while TPO valves are installed
in the second string. The gas allocation between the strings is then controlled by the
selection of the chokes installed in the operating valves. An alternative solution is to
install TPO valves in both strings, since they do not respond strongly to the annulus
pressure. However, practical experience shows that multipointing is a common
occurrence for this design.
9.8 Trouble Shooting
Inefficient gas lift operation can be caused by:
(i)
Inflow Problems due to:
(a) Incorrect size of choke (too large or too small),
(b) Incorrect lift gas (casing) pressure (too high or too low),
(c) Fluctuating compressor lift gas pressures and
(d) Incorrect lift gas flow rate (too high or too low).
(ii) Outlet Problems can include:
(a) Inadequate or restricted (due to partial plugging) gas flow through the
operating valve,
(b) Increased hydrostatic gradient across tubing due to increasing water cut and
consequent reduction in the produced fluid’s “natural” gross Gas Oil Ratio and
(c) The wellhead choke not having been fully opened or the separator pressure
being set too high.
(iii) Downhole Problems. These include:
(a) A hole in the tubing and the
(b) Incorrect operation of valves due to either:
(i)
Mechanical problems with the valve operation itself,
(ii)
Wrong pressure setting of the gas lift valve’s nitrogen charge or
(iii)
Incorrect gas lift valve spacing in the tubing string.
9.9 Trouble Shooting Techniques
Operational problems will be resolved most quickly by carrying out a systematic
investigation to identify the fundamental cause of sub-optimum production from a gas
lifted well. The analysis techniques used and the data gathered should include the:
(i)
56
(Computer) modeling of annulus and tubing pressure profile.
Gas Lift
3
(iii) Comparison with a flowing temperature and pressure survey.
(iv) Carrying out a production test with multi gas injection rates.
(v) Performing of an analysis of the historical well test and production data to
identify trends.
(vi) Carrying out an echometer survey to identify the fluid level in the casing /
tubing annulus. The echometer is an acoustic device which measures the transit time
for a signal generated at the surface to travel down the annulus and be reflected back
to the surface from the liquid level in the casing / tubing annulus. This time
measurement can be translated into a depth if the acoustic velocity is known.
9.10 Some Field Examples of operational problems:
(i) Injection pressure rises to its maximum value while the injection rate drops to
zero. This indicates that an upper value closed before the lower valve opened i.e. the
gas injection point has not being passed to successively deeper gas lift valves. The
well is “stymied” or “hung up” and the setting on one or more gas injection valves
needs to be changed.
(ii) Casing head pressure fluctuates by 5 to 10 psi but the gas injection rate is
constant. The injection point is being transferred between two valves (one form of
valve multi- pointing).
(iii) Casing head pressure fluctuates by 10-20 psi while the injection rate is also
changing. This suggests that a lower valve is opening and closing while a second,
higher valve is continually open.
It should be remembered that it is always possible that the problem is more apparent
than real i.e. the accuracy of the flow meters (in particular) and pressure gauges needs
to be verified. Once it has been confirmed that the flow and pressure measurements
are correct, then one can adjust the lift gas injection rate to the casing or increase the
flowing tubing head pressure by closing the surface choke. These changes can rectify
the problem, though frequently the cause is a valve problem (mechanical damage or
incorrect setting) which can only be solved by its replacement i.e. a wireline well
intervention is required.
10. FIELD PRODUCTION OPTIMISATION
We have already discussed how to maximise the production of gas lifted for a single
well. This involved the development of (theoretical) (computer) model of the single
well’s performance and its calibration against the actual well data via carrying out
multi-rate well tests. A calibrated model will allow those parameters controlled from
the surface {injection gas choke setting, injection gas pressure and separator pressure}
to be set to their optimum values. This process requires the systematic collection,
verification, storage and analysis of the field data since the (computer) model of the
well can never be more accurate than the data against which the model was calibrated.
Department of Petroleum Engineering, Heriot-Watt University
57
The individual well performance models for the wells in the field should now be
combined to give a total field model so that the field performance can be maximised
against the relevant constraints. These constraints can either be:
(i) Physical constraints imposed by the overall system e.g. the total, available lift gas
injection rate, the volume of water that can be handled by the separation system, the
(net) oil production from wells producing from different reservoirs may have to be
mixed in a certain ratio to ensure that the specified crude oil quality is maintained.
(ii) Individual wells or reservoir specific constraints e.g. certain wells may have to
be produced at a specified minimum rate for reservoir management purposes.
The task is then to maximise the NPV operating cash income over the life of the wells.
This normally equates to maximising the oil production by ranking each individual
well’s optimum production response to incremental gas injection. This concept is
explained in figure 41, where the net oil production (or revenue) is plotted against the
lift gas injection rate (or production cost). A minimum (or “kick-off”) lift gas injection
rate, (V1) is required to bring the well on production at the initial production rate Q1.
The well performance curve is then plotted so that the incremental production
increase, (∆Qn) can be evaluated for equal increments in the lift gas injection rate,
(∆xn). It can be seen that the incremental GUF {Gas Utilisation Factor or (∆Qn / ∆xn)}
decreases as the lift gas injection rate increases. In fact, the GUF represents the slope
of the well performance curve (see Figure 41). This slope is also related to the factor:
{incremental (net oil) production revenue}
{incremental (gas lift & other production) costs}
Maximum
production
Net oil production or revenue
GUF7
Q1
Economic limit
*
GUF4
∆x4
∆Q3
∆x7
∆x8
∆x9 ∆x10
∆x6
∆x5
∆Q4
∆x3
∆Q2
GUF1
∆x1
V1
∆x2
∆Q1
*
Economic limit where marginal
(net production) revenue equals
marginal lifting costs.
Lift - gas injection rate
V1
"Kick - off" lift gas injection rate
Q1
Initial oil at "kick - off"
∆xn
Equal lift gas rate increments
∆Qn
Corresponding incremental oil
production or revenue.
GUFn Gas Utilisation Factor
This factor is plotted against the gas injection rate in figure 42. The economic limit
and the technical maximum oil production are the points at which this factor equals
1.0 and 0 respectively.
58
Figure 41
The incremental gas
utilisation factor decreases
with increasing gas
injection rate
Gas Lift
3
2.5
Incremental (Revenue / Costs)
2.0
1.5
Economic limit where incremental
costs = incremental revenue
1.0
0.5
Technical maximum production
0
Lift gas
∆ x 1 ∆ x 2 ∆ x 3 ∆ x 4 ∆ x 5 ∆ x 6 ∆ x 7 ∆ x 8 ∆ x 9 ∆ x 1 0 ∆ x 1 1 ∆ x 1 2 increments
Figure 42
Gas lift incremental cost
"Kick - off" gas
injection rate
Lift gas injection rate
curve.
Allocation of the available lift gas between wells is performed by comparing the GUF
values. This calculation is complicated by the:
(i) Varying well “kick-off” gas requirements (figure 43 shows three wells exhibiting
different “kick-off” behavior).
(ii) Reservoir and well specific constraints, as well as any other production system
constraints which also have to be honored.
(iii) The numbers of wells involved (large gathering systems can connect more than
1000 production wells).
Figure 43
Well performance curves
for three wells
Well a
Well b
Natural flow
(well a )
Net oil production rate
KOGb, KOGc = "Kick - off" lift - gas
required by wells b and c.
Initial Gas Utilisation Factors
Well c
KOGb
KOGc
Lift - gas injection rate
Specialist software packages e.g. FieldfloTM, NetOptTM etc. are available from a
number of vendors which are capable of optimising such large systems. However, it
must always be borne in mind that the optimised production recommendation results
will only be as good as the input field data. Once again, this points to the need to ensure
that a systematic data collection, validation and management system is installed at the
same time as such optimisation packages are purchased.
Department of Petroleum Engineering, Heriot-Watt University
59
Q oil
Experience in many fields has shown that a sustained 2-4% increase in net oil
production can be achieved by a dedicated engineering effort to provide continuous
optimisation of the gas lift systems performance. Large fields often require employing
an engineer solely to manage gas lift and to obtain the commitment required from the
operations staff with respect to data gathering etc. This provides the information
required to overcome the typical gas lift operational problems summarised in figure
44 i.e. so that deviations from optimum operation can be identified and rectified as
quickly as possible.
(8) Unknown well
performance
curve
Stable flow
(1) Oscillating
control valve
Q gas
(9) Excessive surface pressure losses
Q gas
(2) Hydrate
blockage
(3) Unstable injection
gas system
pressure
GAS
(4) Leaking
unloading valves
OIL
OIL
WATER
(5) Hole in tubing
(6) Valve multipointing
(7) Incorrect orifice choke
Q oil
Q water
(10) Unreliable test separator data
(11) Low well test frequency
11. NEW TECHNOLOGY FOR CONTINUOUS FLOW GAS LIFT
It will have become apparent from the above that optimising the gas injection rate at
the operating valve is a key parameter in the control of a gas lifted well. Two new
developments in this area:
(a) A surface controlled gas lift valve where the required choke settings are
transmitted from surface via a cable or wireless transmission system. Two-way
information exchange can be implemented. Thus, annulus and tubing pressure and
temperature measurements at the gas lift valve depth as well as flow measurements
can be incorporated in the same instrumentation package which transmits data to the
surface. This allows the optimum choke setting to be specified.
(b) A redesigned choke that develops critical flow when the tubing/casing pressure
ratio is as high as 90% compared to the normal ratio of 55% (figure 45). This critical
flow condition decouples the pressure behavior in the annulus from that in the tubing
as well as giving a constant gas injection rate for most producing conditions (this
certainly simplifies modeling of the process of gas lift and operation!)
60
Figure 44
Problems for gas lift system
management
Gas Lift
3
Critical flow (Nova “) valves
Figure 45
Orifice valve performances
compared
Gas injection rate (MMscf/d)
Critical flow region
Sub - critical
flow region
Conventional port valve
P tubing≈55% P casing
P tubing≈90% P casing
P casing
Tubing pressure (psi)
(c) Coiled Tubing based Gas Lift Completions in which gas lift valves and mandrels
can be incorporated as an integral part of a coiled tubing string. These can be
employed in two manners:
(a) To replace the conventional production tubing (Figure 46). This particular
completion was designed to test an exploration or appraisal well.
Production to separator
Lift gas
Surface controlled subsurface
safety valve (SCSSSV)
30" casing shoe
18.625" casing shoe
1" gas lift valve in 2.375" mandrel
1.75" coiled tubing
Orifice valve in 2.375" mandrel
9.625" casing shoe
Locator / seal assembly
Polished bore receptacle
Figure 46
Gas lift completion where
coiled tubing has replaced
the tubing string
7" liner shoe
Producing formation
Perforations
Department of Petroleum Engineering, Heriot-Watt University
61
(b) As a second production string within an existing tubing (Figure 47). This allows
gas lift to be introduced to aid well production without having to mobilise a drilling
rig to recomplete the well. This does however require that a polished bore receptacle,
into which the coiled tubing seal assembly can be located, has included in the original
tubing string. The advantage of this approach is that it is probably the cheapest
methods of installing gas lift in a well where the casing integrity has been lost.
Production to separator
Control line hanger
Coiled tubing hanger
Gas injection to tubing/
coiled tubing annuls
Control line for SCSSSV
2" coiled tubing
Surface Controlled Subsurface
Safety Valve ( SCSSSV)
Gas lift valve mounted
in mandrel
Orifice
Locator seal assembly
Polished bore receptacle
Producing Formation
Perforations
(c) Alternatively, the orifice valve can be omitted with the coiled tubing suspended
from its hanger - the gas injection taking place via the coiled tubing’s open end.
12. INTERMITTENT GAS LIFT
All the above has discussed continuous flow gas lift. The intrinsic flexibility of the
gas lift concept allows it to be adapted to a wide range of situations; however it does
become inefficient at low formation fluid inflow rates (<150 bf/d with a 2.375 in OD
tubing rising to < 300 bf/d for a 3.5 in OD tubing). A partial answer to the problem
is intermittent lift where the gas is switched on for a short period of time at regular
intervals :
62
Figure 47
A coiled tubing inner
production string
Gas Lift
3
(i) The formation fluid level in the tubing increases during the periods that the gas
lift is switched off.
(ii) The formation fluid that has collected above the valve is lifted out during a period
of lift gas injection.
(iii) The flow of lift gas to the well is halted and the cycle is repeated.
A time cycle controller opens the surface lift-gas valve for a predetermined time at
regular intervals and then shuts it. The process typically produces some 2-5 bbls liquid
per cycle with a frequency of 1-3 cycles per hour (Figure 48). The valve spread (or
difference between pressure needed to open and close the valve - see section 3.7.2)
controls the minimum amount of gas used during each intermittent gas lift cycle.
Casing Pressure
Wellhead Pressure
M
11A
2P
1000
900
M
800
700
600
M
8A
Re
c
Twin Pen
80
0
70
0
60
0
50
0
400
300
200
100
90
0
10
00
300
200
100
00
500
400
00
00
6PM
0
Well details
Cycle time
Injection time
Production rate
Injection gas rate
30 min
2 min
144 bfpd
350 M scf/d
5
00
5P M
1000
7AM
M
800
900
AM
AM
500
TIME
700
600
TIME
40
0
30
0
20
10 0
0
DATE ON
400
300
200
100
f Casing and
go
Tu
din
bi
or
3P
0
80
0
70
0
60
0
50
0
40
0
300
200
100
4P
500
400
300
200
100
6AM
00
90
DATE ON
5A M
M
10
ssure.
Pre
ng
Figure 48
Intermittent gas lift
1PM
M
9A
1
M
0A
MIDDAY
This approach is more efficient in terms of the volume of gas required to lift a given
volume of liquid than continuous gas lift for these low rate wells. However, it still
requires relatively high gas volumes due to the fall back of the liquid (of the order of
10% of the tubing volume) present in the tubing at the time when the gas is switched
off. This inefficiency can be removed by incorporating a plunger to displace all the
liquid to the surface, the plunger itself having been displaced up the tubing by injecting
lift gas underneath {see figure 6(e)}. The process is as follows:
(i) The plunger displaces the fluids that have entered the well during the period that
the lift gas was shut off.
(ii) The flow of lift gas is halted once the plunger reaches the wellhead and the cycle
repeated.
This process is also known as Plunger Assisted Intermittent Lift (PAIL) or chamber
lift. PAIL has particular advantages when the well is suffering from severe wax
deposition since the regular passage of the plunger ensures the tubing remains clear
of wax.
Department of Petroleum Engineering, Heriot-Watt University
63
13. GRAPHICAL GAS LIFT DESIGN EXERCISE FOR WELL
EDINBURGH-2
13.1 Introduction
The earlier section in this chapter have showed how the well production was
influenced by parameters such as the gas injection rate, tubing size, depth of gas
injection, and flowing tubing head pressure etc. This sensitivity study can be used to
determine the optimum gas lift completion design parameters. It is now necessary to
space out the unloading valves so that the gas can reach the operating valve at the depth
shown above. The procedure used will be illustrated by using the Edinburgh-2 well
as an example (see Table 7 for well design conditions).
Gas injection pressure
1250 psig
Gas injection rate
500 M scf/d
"Kill" brine density
0.465 psi / ft
Depth mid perforations
10,000 ft
Reservoir pressure
3600 psi
(Gross fluid) Production index
0.5 bf/d/psi
Water cut
65% vol.
Oil density
35 API or 0.37 psi/ft
Production water density
1.05 g/cm3 or 0.455 psi/ft
( Average) produced fluid density
0.426 psi/ft
(relative) gas gravity
0.6
Flowing wellhead pressure
100 psi
Average flowing gradients: For producing well
250 bf/d at GLR of 2105 scf/b
0.07 psi/ft
500 bf/d at GLR of 1105 scf/b
0.1psi/ft
750 bf/d at GLR of 770 scf/b
0.157 psi/ft
800 bf/d at GLR of 730 scf/b
0.18 psi/ft
64
Table 7
Well Edinburgh - 2 gas lift
design conditions
Gas Lift
3
0.08
Gas Gradient, psi/ft.
0.9 Gas S.G.
0.07
0.8 Gas S.G.
0.06
0.05
0.7 Gas S.G.
0.04
0.6 Gas S.G.
0.03
T = 70ºF at Surface
190ºF at 10000 ft TVD
0.02
Figure 49a
Injection gas gradients for
well Edinburgh 2
0.01
600
700
800
900
1000
1100
1200
1300
1400
1500
Surface Gas Injection Pressure (psig)
Pressure (psi)
500
2500
3000
n pressure
5000
Gas SG = 0.6
Gas SG = 0.8
6000
e
in
br
ill) ft
(k i /
ic ps
at
St 465
0.
7000
ad
gr
ie
8000
nt
Reservoir
pressure
3600 psi
9000
Figure 49b
Initial conditions for
"killed" well
3500
ce gas injectio
Depth (ft)
2000
1600 psi surfa
sure
4000
sure
jection pres
3000
e psi gas in
2260 ft
1500
1200 surfac
Equilibrium
level
kill brine
800 psi surface gas injection pres
2000
500 psi surface gas injection pressure
1000
1000
10,000
500
1000
1500
2000
2500
3000
3500
Pressure (psi)
Department of Petroleum Engineering, Heriot-Watt University
65
13.2 Initial Conditions - the “Dead” Well
The well has been completed and circulated to a “kill” brine of density 0.465 psi/ft.
Figure 49 shows that the reservoir pressure is only sufficient to support the fluid level
in tubing and tubing/casing annulus to a depth of 2258 ft. Gas pressure gradients for
various gas injection pressures have been drawn. The effect of using gases of different
relative densities (0.6 and 0.8) for the 1200 psi surface pressure case has also been
included.
13.3 Construction of the “Equilibrium Curve”
The range of production rates achievable using gas lift and the required gas injection
depths must be calculated first. A nodal analysis calculation is carried out for a
production rate of 500 b/d using a gas lift injection point at depth D ft as the node and
the data from Table 7.
Calculation from Surface
Pressure at gas injection point = separator pressure + 0.1 D
Calculation from Reservoir:
Pressure at gas injection point = Reservoir Pressure - drawdown - 0.426* (10000 - D)
* We have assumed that the frictional pressure losses below the gas lift valve are
negligible at this low flow rate.
These two pressures are equal and thus :
D = 5400 ft and the gas injection pressure at this depth is 640 psi.
The latter corresponds to a surface gas pressure of 560 psi. The same calculation can
be carried out for gross production rates ranging from 250 to 800 b/d - see Table 8 and
figure 50. (Remember that the flowing pressure gradient has to be changed as well as
the drawdown for each calculation.) This table shows that gas injection pressures
ranging from 350 psi to 1825 psi are required. It would not be realistic to use a constant
gas density for calculating the surface pressure for this wide range of conditions. The
injection gas pressure gradient is also dependent on the average pressure. Table 8
includes an approximate value which allows us to estimate the surface gas injection
pressure required. A more accurate value could have been calculated using the
equation given in section 8.4.
Production rate (bf/d)
Gas injection depth (ft)
Gas injection pressure at gas lift valve (psi)
Average gas gradient at this pressure (psi/ft)
Required surface gas injection pressure (psi)
66
250
3540
348
0.01
313
500
5400
640
0.015
560
750
8400
1419
0.03
1167
800
9594
1827
0.045
1395
Table 8
Calculation of "The
Equilibrium Curve"
Gas Lift
3
Pressure (psi)
500
1000
560
1500
2500
3000
3500
1462
Surface gas injection pressures (psi)
1167
313
2000
1000
2000
Gas gradients SG = 0.6
3000
250 bf/d
5000
500 bf/d
um
bri
uili
6000
Eq
e"
Th
Depth (ft)
4000
Inf
low
7000
low
utf
/O
Cu
8000
"
rve
750 bf/d
9000
Figure 50
Construction of "The
equilibrium curve"
Reservoir
pressure
800 bf/d
10,000
500
1000
1500
2000
2500
3000
3500
Pressure (psi)
Figure 50 summarises the producing conditions which simultaneously satisfy the
reservoir inflow and tubing outflow conditions. This line is called the “Equilibrium
Curve”. It allows a quick estimate of the depth, and associated pressures, at which the
operating valve must be installed to achieve a given production rate. As expected, the
highest production rates are achieved with the deepest gas injection. The available gas
lift pressure is 1250 psi (Table 7) - hence we will set a target production rate of 750
bf/d for the Edinburgh -2 gas lift completion by installing the operating valve at 8400
ft. It now needs to be confirmed that the well can be unloaded when completion brine
(of density 0.465 psi/ft) is present in the well.
Department of Petroleum Engineering, Heriot-Watt University
67
Pressure (psi)
100 psi 400
800
1200
2000
2400
2800
K
0. ill b
46 rin
5 e
ps gr
i / ad
ft ien
t
Valve
number
1
2530 ft
750
3000
5000
adient
Injection gas gr
scfd
MM
00
+ 5 i / ft
fluid .15 ps
ed
ent"
duc as = 0 Gradi
pro
g
e
bpd jection bjectiv
in
eO
"Th
4000
Depth (ft)
2000
Surface gas injection pressure (1250 psi)
50 psi safety margin
Flowing wellhead pressure
1000
1600
6000
7000
8000
Operating valve at 8400 ft
=
9000
10,000
400
800
1200
P
(7 rod
50 uc
0. bp ed
42 d flu
6 ) g id
ps ra
i / die
ft
nt
1600
2000
1500 psi drawdown
required to produce
750 bf/d
2400
2800
Pressure (psi)
13.4 The Unloading Process
The unloading calculation is begun by calculating the flowing bottom hole pressure
(2100 psi) required to produce 750 bf/d. The pressure profile for the flowing well is
then constructed from this point to the surface wellhead pressure (100 psi) using the
flowing fluid gradients before (0.426 psi/ft) and after gas injection (0.15 psi/ft). This
line is known as the “Objective Gradient” (Figure 51). The flowing gradients have
been represented here as straight lines to simplify the calculations. The intersection
of these two lines is the depth at which the operating valve will be installed (8400 ft).
Also shown is the intersection between a (static) brine gradient drawn from the
(flowing) wellhead pressure of 100 psi to the gas gradient line for a surface pressure
of 1200 psi. This pressure is 50 psi less than the nominal gas lift system operating
pressure of 1250 psi. The lower value is as an additional safety margin to ensure that
the gas lift completion will still unload despite fluctuations in the compressor output
68
Figure 51
Well unloading stage 1,
maximum depth of first gas
lift valve
Gas Lift
3
pressure. It also compensates for any frictional pressure loss due to gas flow across
the valve.
The intersection between the gas and brine gradient lines is at 2530 ft - this is the depth
at which the first unloading valve should be placed (Figure 51). The surface gas
pressure is sufficient to displace the brine from the annulus into the tubing and out of
the well.
Pressure (psi)
400
100
800
1200
Flowing wellhead pressure
Br
1000
in
e
2000
2400
2800
Surface gas injection pressure (1250 psi)
50 psi safety margin
r
de
un
g
e
cin
nc
du
re nflue tion
i
jec
the as in
g
lift
gr
ad
1600
ie
2000
Valve
number
1
nt
2530 ft
Br
3000
in
e
gr
ad
Gas injection reduces density
of brine above first gas lift valve.
ie
nt
number
2
4485 ft
Injection gas gra
"Th
5000
nt"
ie
grad
6000
dient
ive
ject
e ob
Depth (ft)
4000 Valve
7000
8000
Operating valve at 8400 ft
9000
Figure 52
Well unloading stage 2,
depth selection for 2nd gas
lift valve
10,000
400
800
1200
P
(7 rod
5
= 0 uce
0. bp d
42 d flu
6 )
id
ps
i/
ft
1600
2000
1500 psi drawdown
required to produce
750 bf/d
2400
2800
Pressure (psi)
The gas will now enter the tubing via the gas lift valve and reduce the pressure exerted
by the fluid in the tubing above the valve. This reduced pressure also ensures that the
Department of Petroleum Engineering, Heriot-Watt University
69
pressure in the annulus remains greater than that in the tubing at all depths. Hence fluid
continues to be displaced from the annulus into the tubing through the second and
lower valves. This process is illustrated in figure 52 where it is shown how the brine
density is reduced under the influence of lift gas injection until it reaches the
“Objective Gradient”. The depth (4485 ft) at which the completion brine gradient
intersects the gas gradient line is the maximum depth at which the second unloading
valve can be placed.
Pressure (psi)
400
100
800
1200
1600
Flowing wellhead pressure
Injection gas gra
1000
2400
2800
Surface gas injection pressure (1250 psi)
50 psi safety margin
Valve spacing
2530 ft
dient
2000
2000
Valve
number
1
2530 ft
3000
1750 ft
4000
5500 ft
4
6250 ft
5
6
6650 ft
6850 ft
100 psi safety margin
750 ft
nt
7000
3
1200 ft
adie
e gr
ctiv
5000
6000
50 psi safety margin
obje
Depth (ft)
4300 ft
The
2
150 psi safety margin
400 ft
200 psi safety margin
250 psi safety margin
} 200 ft
Operating valve
at 6650 ft
Pr
od
gr uce
ad d
ie flu
nt id
8000
1070 psi drawdown
required to produce
535 bf/d
9000
10,000
400
800
1200
1600
2000
2400 2530 2800
Pressure (psi)
Classical design procedures assume that the upper valve shuts immediately the gas
reaches a lower valve. Section 7.3, which described valve performance, showed that
70
Figure 53
Well unloading stage 3,
inclusion of safety margins
Gas Lift
3
this is not realistic. Practical experience has shown that closure of the first valve can
be made more certain by designing the second valve to accept gas at lower pressure
than the valve above it. Safety margins of up to 50 psi/valve are often employed.
Comparison of figures 52 and 53 show how this 50 psi safety margin can be included,
as well as why it reduces the setting depth of valve No. 2 from 4485 ft to 4300 ft.
The design process is continued in the same manner (see figure 53 and Table 9) with
valve setting depths of 5500 ft, 6250 ft, 6650 ft and 6850 ft being identified. The latter
valve spacing of only 200 ft is obviously impractical - in fact a minimum spacing
distance of 450 ft or 150 m is normally recommended. Hence an operating valve depth
of 6650 ft is chosen. Placing the operating valve depth at 6650 ft implies, to a first
approximation, that a flowing bottom hole pressure of 2530 psi and associated
reservoir productivity of 535 bf/d is achievable, rather than the target value of 750 bf/d.
Worst case conditions
(with safety margins)
Table 9
Unloading valves for well
Edinburgh - 2
Expected rate conditions
(no safety margins)
Initial design
Valve Depth (ft)
No.
1
2550
2
4300
Final design
Valve Depth (ft)
No.
1
2530
2
4300
3
4
5
3
4
5
6
5500
6200
6650 *
7100 φ
7
7550 φ
8
5500
6250
6650*
Initial design
Valve Depth (ft)
No.
1
2650
Final design
Valve Depth (ft)
No.
1
2650
2
4690
2
4690
3
6190
3
6190
4
7050
4
7250
5
7500
8000 φ
5
7970
6
7950
9
8450 φ
6
8400*
7
8400 *
10
8900 φ
8
8850 φ
11
9350 φ
9
9300 φ
12
9800 φ
10
9750 φ
*
φ
Operating valve
Bracketing envelope
Dummy valves
Further it is normal to install at least one gas lift at the minimum spacing both above
and below the planned depth for the operating valve. This is done because a higher
or lower operating point may be required to adapt the well’s operation to conditions
somewhat different from the assumptions made during the design process and/or
changes that have occurred during the lifetime of the well. This zone with minimum
valve spacing is called the Bracketing Envelope. Some of the valve setting depths may
be modified when developing the bracketing envelope - as shown in Figure 54 and
Table 9, columns 3 and 4.
Department of Petroleum Engineering, Heriot-Watt University
71
Pressure (psi)
100
400
800
1200
1600
2400
2800
Surface gas injection pressure (1250 psi)
50 psi safety margin
Flowing wellhead pressure
die
Injection gas gra
1000
Valve
number
1
nts
2000
2000
2530 ft
3000
4300 ft
3
5500 ft
4
nt
adie
e gr
ctiv
5000
6000
50 psi safety margin
obje
2
The
Depth (ft)
4000
5
Actual operating valve
100 psi safety margin
150 psi safety margin
6200 ft
6650 ft
7000 6
The bracketing envelope
7100 ft
7
8000
8
9
9000
10
Target operating valve depth
Extended
bracketing
envelope.
Pr
od
uc
ed
11
flu
id
10,000
12
400
800
1200
1600
1500 psi drawdown
required to produce 1070 psi drawdown
required to produce
750 bf/d
535 bf/d
2000
2400 2530 2800
Pressure (psi)
In addition, it is shown that the:
(i) bracketing envelope has been extended in the final design as far as the bottom
of the well to allow deep gas lift if the reservoir pressure / well productivity decrease
sufficiently. This worst case design requires 12 gas lift valves to be installed with
the operating valve placed at 6650 ft and dummy valves installed in the bottom 7
mandrels.
(ii) 50 psi safety margin used between each succeeding valve has created a robust
deign with high confidence that the well will unload to a depth of 6650 ft.
(iii) depth for the operating valve is shallower than the target depth of 8400 ft. This
reduces the well production to 535 bf/d, lower than the target value of 750 bf/d.
72
Figure 54
Final design: worst case
conditions
Gas Lift
3
By contrast, the target fluid production can be achieved if an “expected rate
conditions” design is made (Figure 55 and Table 9). The final design shown
incorporates the modifications made due to introduction of the “Bracketing Envelope” concept. It:
(i) uses two fewer gas lift valves.
(ii) achieves the target production rate of 750 bf/d.
(iii) depends for its success on the well conditions being exactly as prognosed. This
is required to ensure that the unloading process works efficiently.
Pressure (psi)
400
800
1200
Flowing wellhead pressure
Injection gas gradie
1000
Final
design
Valve
number
1
Depth (ft)
4000
2
5000
2800
Surface gas injection
pressure (1250 psi)
Valve Spacing
2650 ft
2040 ft
4690 ft
1500 ft
6000
3
6190 ft
1060 ft
7000
- 7050 ft
720 ft
8000
5
Operating
valve
- 8400 ft
8400 ft
The bracketing
envelope
- 8850 ft
9300 ft
9 9250 ft
10,000
430 ft
7
8
7500 ft
- 7950 ft
7970 ft
6
9000
Recommended spacing
7250 ft
4
Figure 55
Final design: expected rate
conditions
2400
2650 ft
r
t" fo
dien
Gra f/d gas
tive
c
bjec 00 Ms
5
eO
"Th bpd +
750
3000
2000
nt
2000
1600
1500 psi drawdown
required to produce
750 bf/d
Extended
bracketing
envelope.
400
800
1200
1600
2000
2400
2800
Pressure (psi)
Department of Petroleum Engineering, Heriot-Watt University
73
13.5 Gas Lift Optimisation Exercise
Question
Figures 56 to 58 show the (gross) fluid production rate from three gas lifted wells. The
following two exercises illustrate a procedure to manually allocate a limited volume
of gas (2 and 5 MM scf/d) so as to produce the maximum volume of oil.
6000
Water cut = 25%
5600
(Gross fluid) Production rate (bf/d)
5200
4800
4400
4000
3600
3200
2800
1
2
3
4
Gas injection rate (MM scf/d)
74
5
6
Figure 56
Gas lift performance - well 1
Gas Lift
3
3200
Water cut = 40%
2800
(Gross fluid) Production rate (bf/d)
2400
2000
1600
1200
800
400
Figure 57
Gas lift performance - well 2
0
1
2
3
4
5
6
5
6
Gas injection rate (MM scf/d)
Gas Lift Performance - Well 3
3200
Water cut = 45%
2800
(Gross fluid) Production rate (bf/d)
2400
2000
1600
1200
800
400
Figure 58
Gas lift performance - well 3
0
1
2
3
4
Gas injection rate (MM scf/d)
Department of Petroleum Engineering, Heriot-Watt University
75
Answer
Figures 56 to 58 allow the incremental net oil production to be calculated for as the
lift gas injection rate is increased in increments of 0.5 MM scf/d (see Figure 59 and
Table 10). The simplest method is to allocate each (0.5 MM scf/d) increment of lift
gas to the well which shows the highest net oil production. The process is thus repeated
for the second and subsequent increments until all the available gas has been allocated.
Table 10 shows the case when 5 MM scf/d of lift gas (ten increments of 0.5 MM scf/
d) was available. 2.5 MM scf/d of lift gas has been allocated to well 2 to produce an
additional 1074 b/d of net oil production. A similar gas volume was allocated to Well
3 and an additional 1188 bopd produced.
Lift Gas Allocation To Maximise Oil Production.
0.5 MM scf/d
Well 1
Gas increment
1
Incremental oil production (bo/d)
228
228
2
3
4
5
Well 2
-
6
7
-
228
222
-
8
9
-
-
10
-
Net oil production
0
Well 3
Total incremental
oil production (bo/d)
228
456
194
684
906
1104
324
252
1428
1680
1922
2098
168
242
176
-
2266
1074
1188
2266
Table 10
Lift gas allocation to
maximise oil production
6000
5600
)
5%
=2
te
5200
(Gross fluid) Production rate (bf/d)
ut
rc
ll 1
We
(wa
120 bfpd = 90 bopd
4800
110 bfpd = 82.5 bopd
110 bfpd = 82.5 bopd
130 bfpd = 97.5 bopd
4400
140 bfpd = 105 bopd
120 bfpd = 90 bopd
4000
3600
3200
2800
76
1
2
3
4
Gas injection rate (MM scf/d)
5
6
Figure 59a
Incremental net oil
production with increasing
lift gas volumes
Gas Lift
3
(w
at
er
cu
t=
40
%
)
3200
W
el
l2
2800
2400
(Gross fluid) Production rate (bf/d)
240 bfpd = 144 bopd
2000
280 bfpd = 168 bopd
370 bfpd = 222 bopd
1600
380 bfpd = 228 bopd
1200
380 bfpd = 228 bopd
800
380 bfpd = 228 bopd
400
Figure 59b
Incremental net oil
production with increasing
lift gas volumes
0
1
2
3
4
Gas injection rate (MM scf/d)
3200
5
6
5
6
)
%
=
ut
45
rc
ate
2800
ell
W
(Gross fluid) Production rate (bf/d)
2400
w
3(
280 bfpd = 154 bopd
320 bfpd = 176 bopd
2000
440 bfpd = 242 bopd
1600
1200
560 bfpd = 252 bopd
800
590 bfpd = 324.5 bopd
Figure 59c
Incremental net oil
production with increasing
lift gas volumes
400
360 bfpd = 198 bopd
0
1
2
3
4
Gas injection rate (MM scf/d)
Department of Petroleum Engineering, Heriot-Watt University
77
Table 10 shows that this calculation is not quite as simple as it appears due to the need
for a small volume of "Kick - off" gas to bring well 3 into production. This is illustrated
when only 2 MM scf/d of lift gas is available. The simple approach used above
suggests that it should all be allocated to Well 2 where it will recover an extra 906 bopd.
However, allocating the same volume of gas to Well 3 will produce 1016 b/d of extra
net oil (Table 11). This calculation illustrates why the slope of the curve (the Gas
Utilisation Factor) is a better optimisation factor rather than the simplistic incremental
production approach suggested above.
0.5 MMscfd
Gas increment
1
2
3
4
Total
Incremental oil production (bopd)
Well 1
Well 2
Well 3
90
228
194
105
228
324
97.5
228
252
82.5
222
242
375
906
1012
Commercial network simulation computer programs are essential to carry out this
type of optimisation calculation. There are many types of constraints that the
simulator has to honour while at the same time maximising net revenue. These can
include:
(i) well constraints e.g. sand production,
(ii) gas/water coning or other reservoir constraints,
(iii) gas/water separation, compression or disposal facility constraints or bottlenecks,
(iv) production constraints, such as mechanical equipment failure, and
(v) export quality requirements.
78
Table 11
Allocation of 2MM scf/d of
lift gas
Gas Lift
3
14. FURTHER READING
(1) Beggs H. D.
“Production Optimisation using Nodal Analysis”
ISBN 0-930972-14-7
published by Oil and Gas Consultants Inc., 1991.
(3) Economides M., Hill A. & Economides C.
“Petroleum production Systems”
ISBN 0-13-658683-X
published by Prentice Hall, 1994.
(4) Economides M. J., Watters L. and Dunn-Norman S.
“Petroleum Well Construction”
ISBN 0-471-96938-9
Published by Wiley, 1998.
(5) Golan M. & Whitson C.
“Well Performance” 2nd edition
ISBN 0-13-946609-6
published by the Norwegian University of Science and Technology (NTNU), 1996.
(6) Mian M. A.
“Petroleum Engineering Handbook for the Practicing Engineer”, Volume 2
ISBN 0-87814-379-3
Published by PennWell Books, 1992.
(7)
API - EXPLORATION AND PRODUCTION DEPARTMENT
API Gas Lift Manual. Book 6 of the Vocational Training Series, 3rd edition
API, 1994
Department of Petroleum Engineering, Heriot-Watt University
79
80
U
V WXYZ[\ZW]^]_]`a[b]^W[cdeYf
g
n
S
t
hiZj `W]ik[l`j `mY
o `WZ]p[\_]q]f]km
012131415131416
7899
899 899 9 87899 9 9!" #
9 8899 $%&9'%&9!($#
9 8899 )*+(%&9'%&,9!$*+(%&#
-8999 .9/ 0 1 910 19
2899 9/9/ 0 1 910 1
9 28789 / 3%&9 3 3"49 %%&"9
99 9 9 $3%&,
9 2878789 / 3%&9 3 3"90 (53&%, ,9
99 9 9 +%&"9%%&"
9 28789 /+%*96,,
9 2878879 3 95%,
9 28789 /%73'9/3%&
8 9:;:<8 =>?@ABC>8DECFGH
9 2899 /3%&933"9+%&"9&%&"
9 28899 / 3%&9 3 3"9 +%&"9
99 9 9 )I3%&"
9 28-899 /3%&933"9+%&"9)*+(%&
9 28-8789 +(,9I9/3%&933"979/%&,99
9 9 9 0 J&
9 28-889 +(,9I9/3%&933"9799 9
9 9 9 $%&93&*9 'J9/+%*,
9 28-889 +(,9I9/3%&933"979(3
9 28-887899 &"3&%(9(3,
9 28-88899"3&%(9(3
9 28-8-899 +(,9I9/ 3%&9 3 3"79
9 99 9 K3(%3
9 28-82899 +(,9I9/ 3%&9 3 3"79
99 9 9 ),,+9 *+(%&
9 28-8L899 +(,9I9/3%&933"9799 9
9 9 9 %+3%&
9 28-8M89 +(,9I9/3%&933"979 39
9 9 9
&N(%&9 9
9
L899 / 0 1 910 19 99
9
9 9) 1 9 L8789 'J9/+%*9O+3% 9+%*%&,999
9 9 9 0 %&%%,9/3%&933"
9 L8899 'J9 (5&%P+,990 %&%%,999
9 9 9 /3%&933"
9 L8899
("&%%&9I959),,+9I9
99 9 9 /3%&933"
M899 / . 9 1 Q899 1)) R91
r Yaa[sYZ^iZj `k_Y
udqZ`va]_[hZ`_WvZ]km
wkfW`xaY[hiZj `W]ikf[`kq[y`k
{
V ]a[`kq[| `f[sZi_Yff]km
}
r `WYZ[u`kqa]km
S
V ]a[`kq[| `f[h]Yaq[V eYZ`W]ikf
1234565789
2
62
!"#! $$%%$&
'()*+,-./(012/(-/3(140(5467-8.4/(3-7-92(:4/86.;<82=(84(+446(02,,(+265467-/:2>
'(?32/8.5@(812(:6<:.-,(/2-6(02,,;462(-62-(=<=:2+8.;,2(84(5467-8.4/(3-7-92>
'(A-,:<,-82(812(:4=8(45(5467-8.4/(3-7-92(B./(8267=(45(,4=8(+643<:8.4/C>
'(?32/8.5@(812(7-D46(=4<6:2=(45(5467-8.4/(3-7-92>
'(E2,2:8(-++64+6.-82(62723.-,(862-872/8=>
'(A-,:<,-82(812(.7 +-:8(45(36.,,./9(-/3(:47 +,28.4/(5467 -8.4/(3-7-92(B32+81(-/3((
( 2*82/8(45(.7+-.672/8C>
'(F.=:<==(812(:-<=2=(45(+643<:8.4/(62,-823(5467-8.4/(3-7-92>
'(?32/8.5@(812(8@+2=(45(=:-,2(2/:4</82623(./(02,,(4+26-8.4/=(-/3(812(G-6.-;,2=(
(( 01.:1(2552:8(812(=2G26.8@(45(812(+64;,27>
'()*+,-./(140(=:-,2(./1.;.846=(-62(<=23>
'(E8-82(812(=:4+2H(</326,@./9(:-<=2(-/3(62723.-,(-:8.4/(62I<.623(84(32-,(0.81(-(0-*(
(( 32+4=.8.4/(+64;,27>
'(E8-82(812(=:4+2H(</326,@./9(:-<=2(-/3(627 23.-,(-:8.4/(62I<.623(84(32-,(0.81(
((-/(-=+1-,82/2(32+4=.8.4/(+64;,27>
'(E8-82(812(=:4+2H(</326,@./9(:-<=2(-/3(627 23.-,(-:8.4/(62I<.623(84(32-,(0.81(
(
( +26546-8./9(3-7-92>
'((J64G.32(9<.32,./2=(546(7 ./.7 .=./9(5467 -8.4/(3-7 -92(3<6./9(046K4G26(
(( 4+26-8.4/=>
LMNOPQRSTUMVWXMTVUMYZU[[\ZUMW]M]WZ^STQWOMPS^S_UMRSOM`UMQPUOTQaUPMQOMSMYZWP\RTQWO(
(( 46(./D2:8.4/(02,,>
0
!"#"$"%&'"#()*+
,
3
7
:
-./ %".0 1%/ %2
4 %"5 !$"6"+"02
DEFGHIJKLMNIGKH
8 &&9#./ %0$
OPQRSQTQUVWXURYZ[XZQQUXZ[R\W]^_QR]XT`^TTQ]RPWaRbPQRPX[PQTbRcUQTT^UQR]UWcTRdUQR
QecQUXQZ`Q]RXZRbPQRZQdURaQ__fWUQRUQ[XWZTR]^QRbWRbPQRTQ\Xg_W[dUXbP\X`RZdb^UQRWhRbPQR
ijkllmjknoplqrstkujkvrqpwslxpiluwyuqxkuzmpoupszw{uk|mrqpwsl}uu~xkutjpqptrvuirjrkqkju
%<&
"$RaQ_
-
%$
"02
]QbQU\XZXZ[RaQ__RcUW]^`bXVXb€RXTRbP^TRbPQR`WZ]Xb;)6
XWZRWhRbPQR
ZQdU
_fWU
QRU<
Q[XWZ
RR‚Z€R
`PdZ[QTRhUW\RbPQRWUX[XZd_RhWU\dbXWZRcQU\QdfX_Xb€RdUQR`PdUd`bQUXTQ]Rf€RbPQRƒT„XZ R
†rvmk‡upsqjwomtkouˆ‰uŠrsu‹†kjopsŒksursoumjlquŽlkkuklkj†wpjuszw{u‘ womvk’‡uru
opkslpwsvkllusmˆkjurqxkrqptrvv‰ursrvwŒwml
uqwuqxku
“vu
q-.
jrslykj
utwky
“tp
ksqup%06
su ?%
=0+
%>&
/
%
"
.0+
PQdbRbUdZThQU
A
”WU\dbXWZR•d\d[QR`WZ`QUZTRbPQRhWU\dbXWZRWhRdRVW_^\QRWhRUW`„RaXbPRdRUQ]^`Q]R
cQU\QdfX_Xb€RXZRbPQRZQdURaQ__fWUQR–WZQRROPXTRcQU
_Xb€RUQ]^`b
`dZR
fQR]^QRbWR
dR
\QdfX
"&%06
B %+XWZR
9
.$++"
02
\^_bXb^]QRWhR̀d^TQTRgRf^bRXZRd__R̀dTQTRXbRaX__RUQ]^`QRbPQRƒZdb^Ud_ RcUW]^`bXVXb€R]^QRbWRbPQR
piwlpqpwsuwyursuk—qjruijkllmjkuojwiurluqxkuzmpouzw{luqwuqxku{kvvˆwjk}uu~xplutxriqkju
aX__RfQ[XZRaXbPRdR]XT`^TTXWZRWhRbPQRTW^U`QT˜RQebQZbRdZ]RbPQR\W]Q__XZ[RWhRhWU\dbXWZR
%Zdb
;%06&
"PQR\Q`PdZX
02
]d\d[QRR™bRaX__R`WZbXZ^QRaXbPRdR\WUQR]QbdX_8
Q]RQed\X
XWZRWhRb
T\TR
_Qd]XZ[RbWR•UX__XZ[RdZ]RšUW]^`bXWZRXZ]^`Q]RhWU\dbXWZR]d\d[Q›RhW__WaQ]Rf€RdRUQVXQaR
WhRbQ`PZXœ^QTRaPX`PR`dZR_Qd]RbWRXbTRdVWX]dZ`Q
C
"b&UQdb
%06
-"
\QZb
*
OPQRhW__WaXZ[R\ W]^_QTRa X__R]XT`^TTRa Q__RTbX\^_dbX
WZR
\QZbTBR%+
bX\^_
dbX&
WZR6
bUQdb
TR%".0+
dUQR]QTX[ZQ]RbWRXZ`UQdTQRbPQRaQ__RcUW]^`bXVXb€˜RQXbPQURf€ž
ŸXR ¡¢£¤¥¦§¨RWUR`W\c_QbQ_€R©¢ª «¬¦§¨RbPQRhWU\dbXWZR]d\d[QRf€RdR`PQ\X`d_R
RR \dbUXeRbUQdb\QZbRŸQ[Rd`X]XTXZ[ RWU
ŸXXR ­®¯°±±¦§¨RbPQRhWU\dbXWZR]d\d[QRf€R`UQdbXWZRWhRdRPX[PRcQU\QdfX_Xb€R`PdZZQ_R
RR f€RdRP€]Ud^_X`RhUd`b^UXZ[RbUQdb\QZbRWUR]QQcRcQZQbUdbXWZRcQUhWUdbXWZR[^ZT
²EFK³´J³G´µ
OPQRcWbQZbXd_RhWURbPQRcQU\QdfX_Xb€RWhRbPQRZQdURaQ__fWUQRhWU\dbXWZRfQXZ[RUQ]^`Q]R
Ÿ]d\d[Q] RQeXTbTRhUW\RbPQR\W\QZbRbPdbRbPQR]UX__fXbRQZbQUTRbPQRhWU\dbXWZR^ZbX_RbPQR
{kvvuplu“srvv‰urˆrsowsko}uu¶jwtkllklu{xptxuvkrouqwuywjrqpwsuorrŒkuq‰iptrvv‰urtqu
qxjwmŒxurujklqjptqpwsuwyuzw{uŽwsuruiwjkuqxjwrqultrvk’uomkuqwukpqxkj·
ŸXR šP€TX`d_Rf_W`„d[QRWURUQ]^`bXWZRXZRTX–QRWhRbPQRcWUQRbPUWdb
ŸXXR SQ]^`bXWZRXZRbPQRUQ_dbXVQRŸWX_RcQU\QdfX_Xb€RQ[R]^QRbWRŸd]VQUTQ RhWU\dbXWZR
uu {kqqrˆpvpq‰utxrsŒkl‡uixrlkutxrsŒklupsuqxkuijwomtpsŒuzmpol}
Žppp’u~xkurˆw†kukyyktqlutrsuˆkurttksqmrqkourquxpŒxuzw{ujrqku{xksuqmjˆmvksquzw{lR
RR \d€RRW``^URgR_Qd]XZ[RbWR\^`PR[UQdbQURcUQTT^UQR_WTTQTRbPdZRW``^UUQ]RhWURbPQRTd\QR
uu zw{ujrqkuˆkywjkuqxkuywjrqpwsuorrŒkuqww¸uivrtk}
”WU\dbXWZR]d\d[QR`dZRUQT^_bRhUW\R\dZ€R]XhhQUQZbRTW^U`QTRgR]UX__XZ[˜R`Q\QZbXZ[˜R
cQUhWUdbXZ[˜R̀W\c_QbXWZ¹[UdVQ_Rcd`„XZ[˜RcUW]^`bXWZ˜RXZºQ`bXWZ˜Ra WU„WVQU˜RTbX\^_dbXWZ˜R
kqt}uu~xklkukyyktqlu{pvvurvvuˆkuopltmllkoupsuŒjkrqkjuokqrpvuvrqkj}uuuw{k†kj‡u“jlqu{kuskkou
qwu|mrsqpy‰uqxkul¸psutwstkiqujkykjjkouqwukrjvpkjursouqwuk†rvmrqkupqlupirtquwsu{kvvupszw{}
01234567 28609 0 2659 2 7 0882258002596 4660825600
1234567
89 9 8 9 !
9 9 "# ! $ $ ! 9
% 9 $ $! $ %$!%$ $! $! "
89 & '
(
) ! ) ) 8! 9 9 $ $ &"89 ! 9
*+,-..-/+,*012+3+*+4516078*/,2+-9+6:2+;054+;-<+1672*.=5/21+/-7.*=+6-+6:2+<2==>
?@A@BC DEEBFDGH DIJKBLMNDGH DIJKO
P:2+<2==+Q2-.267R+1S5/+72;2,61+Q2-.2675,*=+,-/15427*65-/1+<:5,:+*=627+6:2+1S5/+T*=02+
9 & " 89 '
U ! V& 9 9 9 W
& ! 9 X
9
9 $ !
89 $ & ! V & !"
89 & 99 9 V & &9 9 Y 9 9& & !
9 & "
?@ZB[GH \EDI]G^BM_]^BLM`GH \O
abcdecfghfijkhlmdniodpcdklmqgrskclj "" & !% ++ 1:-76+-7+6--+/*77-<+1:*t2u+5/,-772,6+t:*15/Q+=2*45/Q+6-+;-<+,-/T27Q2/,2+-7
++ 42T5*65-/+97-.+6:2+/-7.*=+;-<+=5/21+26,>v+*/4+5.t242+;-<+97-.+72127T-57
& "w9 %9 9 & $!
++ x-/2+*7-0/4+6:2+t279-7*65-/+< :5,:u+59+/-6+72.-T24u+7210=61+5/+*+7240,24+;-<+7*62>
+++ P:2+5/;-<+5/6-+*+t279-7*624+,-.t=265-/+51+/-7.*==R+=211+6:*/+6:2+y6:2-7265,*=v
++ 5/;-<+6-+6:2+y0/5.t*5724v+-t2/+:-=2+-75Q5/*==R+475==24u+5>2>+6:272+51+*+t-1565T2++
+ 1S5/>+z -<2T27u+*+:5Q:+42/156R+-9+=-/Q{<542+t279-7*65-/1+,*/+7210=6+5/+*+109|,52/6
++ 5/,72*12+5/+6:2+5/;-<+6:*6+*+/2Q*65T2+1S5/+7210=61>++P:2+5/627*,65-/+826<22/
9 ""9 %" "
9 9 9 9 9 ! & $ & 9 & & !"89&$ 9 9 "
}fi~cd€iskl‚ V &9 9 9 9 & $ &9 99 $! % ƒ ! & "
0
!"#"$"%&'"#()*+
,
3
7
:
-./ %".0 1%/ %2
4 %"5 !$"6"+"02
DEEEFGHIJKLMINOPQRSTUVSWRXYUZWY[[\R]^^ZWWTX_R]WR`YWUTa^TY[[\bR^WSYUScRdW]ddScG
RR V\cWYZ[T^ReWY^UZWSfRQRgT[[R[SYcRU]RTX^WSYfScR
XS_YU
jTXfRk\Rd[Y^TX_G
8TXh]gR
&&YXcR
9
#.TiSR
/ f%0$
GG lGmEnmGopqrplsEtEuvGolumwlvGxqyrGzppoGE{GumpGxyqrluEy{GuyGumpGwpttsyqp|G
GGG }ql~ptGol€E{nGl{zGxqluqE{nGl{GspGyrsE{pzGE{GumpG‚ƒqlGl{zG„l€ Goqyp††GuyGG
G vEptzGyrotpuEy{†GwEumGpxxpuE~pG†l{zGy{uqytGl{zG{plqG‡pqyG†€E{†|
;)6%<&"$ -%$<"02
ˆ‰ˆŠ‹ŒŽ‘’“Š”•’“–Š—”˜ŒŽ‘’“™
šwyGp›lrotp†GyxGoqyzuEy{G†€E{†Glqpœ
G-.
%~pq
".0+
%06
DEFG GqlupGzpop{zp{uG†€E{GE†Gyxup{Gys†pq~pzGE=0+
{GmEnmGq%>&
lupGnl†
wptt†/
GDl{zG
vGmEnmG
G ?%
R WYUSR]T[RgS[[fbžRRŸVTfRTfRcZSRU]RX]XQ YW^\R]WRUZWkZ[SXURh]gžRR¡UfRdWSfSX^SR^YXG
GG spGlG†pxtGE{zEluEy{GumluGumpGwpttGE†GlGoyup{uEltG†uErtluEy{Gl{zEzlup|
C
R
`YbRRR¢Rh]gTX_Rk]UU]£RV][SRdWSffZWSRkS[]gRUVSRkZkk[SRd]TXUR`dWSfSX^SR]eR_YfbR]W¤
R
GG
GG
G
G
GG
`kbRR¢RWSUW]_WYcSR^]XcSXfYUSRhZTcRf]RUVYURUVSWSRTfRUg]RdVYfSRWS_T]XRYURUVSG
G opqxyqluEy{†|Gšmp†pGl{GsyumGtplzGuyGoqp††qpGzpop{zp{u¥GqptluE~pG
G opqr plsEtEuvGpxxpu†GDl{Gloolqp{u
{q
pl†pGE%+
{G†€E
{Gyq
ywpq
"GE
&l†G
%06
-"
6GlGt
pl†pzF
*|GG%".0+
G uml{Gp›opupzGE{qpl†pGE{Goqyzu
Ey{G
umpGzqB
lwzyw{G
E&
†GE{q
G šmpGoqp†p{pGyxGumpGp›uqlGoml†pGqpzp†GumpGpxxpuE~pGopqrplsEtEuvGuyGumpG
G rl¦yqGoml†p|GGšmE†GwytzG{yqrlttvGspGE{upqoqpupzGl†Goy†EuE~pG†€E{|
8 %;%06&"02
§¨©©ª«¬©­®¯° ±ª²®³©´±° ±µ¬©¶·²³
G¸ y†uGxyqr†GyxGxyqrluEy{GzlrlnpGqpzpGumpGqy€GopqrplsEtEuvGuyGlGpqulE{GzpoumG
YgY\ReW]£RUVSRgS[[žRR¹T_ZWSRºRT[[ZfUWYUSfRUVSRWSfZ[UTX_RdW]cZ^TX_RdWSffZWSRdW]a[SR
YXcR^]£dYWSfRTURgTUVRUVSRS»ZTiY[SXURdWSffZWSRdW]a[SRe]WRYXRZXcY£Y_ScRgS[[žRRŸVSR
WSfZ[UTX_RS¼UWYRdWSffZWSRcW]dR`½¾zFGml†GuyGspGyrop{†lupzGxyqGpEumpqGsvGlGqpzpzG
oqp††qpGzqyoGlqy††GumpGmy€pGyqGsvGlG†rlttpqGoqyzuEy{Gqlup|GƒEnqpG¿|G†myw†GumluG
umpGoqp††qpGzqyoGyxGumpG{plqGwpttsyqpG‡y{pGE†Gy{tvGy{pGyroy{p{uGyxGumpGqp†pq~yEqÀ
U]QfU]^jRUYXjRh]gRf\fUS£ž
×ÄÑ ÄÒÃÂÆ
çËÐÃ
ë ÃÅÅìËÈÃ
íÃÐÖÈÃÅÍÐÃ
åÃÉÃÈæËÍÈ
ØÂ
Ø
ÇÈ
ÁÂÃÄÅÆÇÈÃÉÉÊÈÃÆÇÈËÌÍÅÃ
ÎÏÐÂÄÑ ÄÒÃÂÓ
ÇÝ
ÔÕÖÊÄÅÆÇÈÃÉÉÊÈÃÆÇÈËÌÍÅÃ
Î×ÄÑ ÄÒÃÂÓÆÎØÂÆÙÆØÓ
ÚÛÜ
HîïMINPð
ñòNPNóóNKLPôóPOõîöPôöP÷NøøP
îöùô÷PúINOOMINPúIôûøNOü
A
"&%06 B %+ 9.$++"02
DEEFG „qyzE{nGl{GyEtGwpttGwEumœ
ÇÞ
Èß
È
Â
ÚÛÜ àÆáâÖÈÄÆãÈÃÉÉÊÈÃÆÂÈËãÆ
ÂÊÃÆÖËÆäËÈÑ ÄÖÍËÐÆ×ÄÑ ÄÒÃ
ÈÃ
GGGGGGGGGGGGGGGGGGGGG
äÍÒÆèéê
01234567 28609 0 2659 2 7 0882258002596 4660825600
3443
3
$ 7
$%
,73
95
("79357)5
! 3 5
3 ! 3 4 *
&4575' 9
4795 77! 57435 38 3
555359
555"3 759
55#359
2 34456789
3
+33 7
+33 7
3 ! 3 4
(*
ª«¬­®¯°Ã
Ä¿¯°¹´¾­®´¶°Å»µ°Æ®»Â­Ç«¹¬°
ÈÉÈÊËÌ
1111111
/01121131451623178492314011:75;18818140941<1931451=57>94?52139>9@1A192314011540171
BCDECFGFHIJHKLHJDLMGJNEJHKGJHCHLOPJQIJQOONIHRLHGSJQFJTUNRGJVJWKQBKJBCDELRGIJHKGJ
5-.
LXGRLUGJWGOOJIMQFJYRCDJIGXGRLOJZCRHKJ[GLJTGOSI\JJ]KG^JWGRGJLOOJSRQOOGSJLHJLRCNFSJHKGJ
ILDGJHQDGJLFSJICDGJWGOOIJQFJGLBKJTGOSJWGRGJSRQOOGSJWQHKJCQOJ_LIGSJDNSJ`abc dJLFSJ
CHKGRIJWQHKJWLHGRJ_LIGSJDNSJ`e bc dJSRQOOQFUJDNSI\JJfHJQIJBOGLRJHKLHJHKGJTGOSJLXGRLUGJ
WGOOJIMQFJXLONGJQIJXLRQL_OGJ_GHWGGFJTGOSI\JJgNRHKGRPJQHJQIJBCFIQIHGFHO^JOCWGRJWKGFJCQOJ
_LIGSJDNSJQIJNIGSJLIJLJSRQOOQFUJhNQSJQFJERGYGRGFBGJHCJWLHGRJ_LIGSJDNS\JJiCWGXGRPJQHJ
?8162;<1971j014017140119k5l11l9<6181=5715?<1k98131>63197115:4?>6>1571j0140171401m1
;56<31k11=67401717136;13n
u|
‚ƒ„ qƒr} ƒ††r‡ˆp‰
{ru|
{ry|
w|
v|
|
v
w
} ~
€~
x
u
opƒ†Š
x
y
z
ª«¬­®¯°±
²³¯®´¬¯°µ¯¶¶°·¸«¹°³´¶­¯·°º»®°
·»¼¯°½»®¾¿°À¯´°Á¯¶Â·
1111111111111111111
opqrstu
/0?812113145131417>?211j 014017140115:4?>6>18‹?21l9<6110981k11219;0?1l1311?819337188131
km14011Œ9j‹?281Ž694?52A140?81;9<;6<941814011=57>94?52139>9@118‹?2139>9@1‘17186<4?2@1
=75>191;m<?231715=17136;131:17>19k?<?4m197562314011j1<<k571’
1 “”•–—–˜™”š””Ÿ”””””•”¡¢£¤
”””””””””š””””£”£”””›œ””ž”§§•1 1
¥¦
•
¨
0
1
1
1
©‘
!"#"$"%&'"#()*+
,
3
7
:
-./ %".0 1%/ %2
4 %"5 !$"6"+"02
DEFGHIJIKFLFJMGIJFGNFOPFNGQPGORSJFMGTGIPNGUGIPNGQPGINNQLQVPW
8 &&9#./ %0$
XYZY[\]^_`a\bYXcadX`
eGfGgSQNGgVhGJILF
iGfGgSQNGjQMkVMQLl
;)6%<&"$ -%$<"02
mno_`a\bYpqrYsX\tsY`X_`Y`XcY[\]^_`a\bYu_^_dcYsvabYabw]c_scsY_sY`XcYxc]^c_yaza`{Y
u_^_dcY]_`a\Y|v}vu~Yabw]c_scsY\]ỲXcY]_uaosY\[Yu_^_dcY|]u~Yabw]c_scsYY€XasYasYazzos`]_`cuY
QPGORSJFG‚GGDESMGQƒGLEFGƒVJKILQVPGHFJKFI„QQLlGQMGJFNSkFNGLVGT†‡ GVƒGQLMGVJQRQPI GjISFG
=0+
%>&M‹Q
-.
".0+
%06 ?%
|v}vuGfGT†ˆGVSLGLVGIGJINQSMGVƒG‰†GkKŠGLEFPGLEFGƒVJKIL
QVPGNIKIRFG
PGŒu/
ˆGQM%
GFeSI
GLVGT†‚
—
³•´‘¯³‘—©® ©ž²
œš
äåæçèéêë
ìíåîêïðñòóèêåîñèéðôéôêõåòöê
ðê÷ðèæéèêøðùðæéêèðøåçôêðîøê
øðùðæéêèðòåóêíúíø
”š
•–•— ˜
 žŸŸ ¡¢ž
™š
C
8 %;%06&
"02
›š
¤š
•–•— ˜ œ
"&%06
B %+ -"&6 *%".0+
•–• ˜ ™
›š
š
—
š
š“›œ
YYYYYYYYYYYYYYYYYYYYYYYYY
š“¤š
š“”œ
š“£š
š“¥œ
š“¦š
›“šœ
¢—§‘¨©—ª«‘¡¬‘­©® ©ž‘¯° ž±ž¢«²
›“™š
›“¤œ
›“œš
¾¹¹¹
Ž‘’“”
ÔÑÅÂÃÕÓ¶Å˸ÀÁÓÉÒ¸ÐÐÏÖÂÁÍ
¾¹¹
äåæçèéêû
üïïéñòêóïêïóèùðòåóîêøðùðæéê
óîêõé÷÷êýèóøçñòåóî
×Â¸Ø¸Ù¹È Â
׸ظ¾¹È Â
¾¹
׸ظ¾È Â
¾
ÔÉÑÈ ÉÁж϶Ó͸ÅƸÃËÂÁÈ Á·É¸ÑÉÄÑÉÑÚŶѸظ¾¹¹¸È Â
µÅÑÈ ÁÓ¶Å˸ÓÛ¶Õ×ËÉÄĸ¾¹ÆÓ¿
Î ÉÏÏÐÅÑɸÑÁ¶Ãĸ¹¿ºÙÆÓ¿¸
ÇÑÁ¶ËÁ·É¸ÑÁ¶Ãĸٹ¹ÆÓ¿
ܶϸڶÄÕÅĶÓ͸¹¿Ù¸ÕÝ
ÇÑÁÞÂÅÞ˸Ùß¼¸ÝĶ
¹¿¾
¹¿¹¾
Y YYYYYYYY
A
"&%06
B %+ 9.$++"02
™š
•–• ˜
£š
¹
º
»
¼
½
¾¹ ¾º ¾» ¾¼ ¾½ º¹
ÀÁ¶ÃĸÅƸÇÁÈ Á·É¸ÊÅËɸÌÉÍÅ˸ΠÉÏÏÐÅÑÉÒ¸ÆÓ
ºº
º»
àÆÆÉÕÓ¸ÅƸµÅÑÈ ÁÓ¶Å˸ÇÁÈ Á·É¸Å˸ΠÉÏϸÔÑÅÂÃÕÓ¶ÅË
µ¶·¸µ
DEFGJFMS LMGVƒGáINQI GQPgVhGkI kS ILQVPMGSMQPRGLEFGI„VjFGLlHQkI ŠGHFJKFI„QQLlGNIKIRFG
JILQVMGIJFGƒVSPNGQPGORSJFGâ‚GGã F GHJVNSkLQVPGQMGJIHQN lGJFNSkFNGIMGƒVJKILQVPGNIKIRFG
QPkJFIMFMŠGhQLEGLEFGNIKIRFGQPGLEFGjFJlGPFIJGhF „VJFGJFRQVPGŒOJMLGƒFhGƒFFLˆG„FQPRG
`XcY^\s`Ya^x\]`_b`
01234567 28609 0 2659 2 7 0882258002596 4660825600
123456789 4 4254 68 5 48683454546 49 353 4373 34 64234
! !"!#$%&! !$'(!"!)$!*+%!,"-$ .!/*,01
=>? @? BCDCEC F 23345676894
4 ICJ>K BCLC233<;
CCCCCCCCCCCCCC9CCCCACCCCCCCCCCC:C596;
CCCCCCCLC 4
<;=>?9@?A BCDCECGHG6;F
:596;
4
4
MNO
P2334Q5384R4Q 84S4Q8683T45T3T4Q53849 85 484549 5UV 5 4WXWYZ[\Y]^ 4 68 5 4
8683484_`538484`8683a48342349 3 7 547 V9 5 48 3 T
n†‡ˆq‰ŠŠo‹o„‹Œq€
!b)(!$+c ()!d$%$$ !e%!$"-$ .!c !(f !(!$!c(!-g$!h1!!b)$!'c!
!)$!%$++! !"!)$!c'cg$!(f !"!ij&!c++c$! !)$!$kc'+$!((($!
$c+$&!$$(! !! +.!ljm !"'!%)c!!%+!)c#$!d$$ ! !)$!c($!)$!
68 5 486834P8 48U 3 T44
yv
rv
zv
{v
tv
|v
xv
wv
uv
v
u
t
uv
ut
wv
wt
xv
}~oq€}q‚ƒ ‚p„
µ¶·¸¹º»¼
µ½¾¿»ºÀÁ¶ºÃÂĻź¹ºÆǺǻ
¿¶ÈÉ»¶Ã¹ºÆǶ÷»À¾¹ÊÆȶ¾Ã»
ÅÆÊÆ·º»Ç˶Ã
tv uvv uvvv
44444444444444444444
¦£³£¨¨£ š´¨© ª«¬˜©Ÿš¡Ÿ š
¨£ ª«£ š´¨©­£«¬š²¡¥ª£š¨£¯ª¥¬š
³¨©¢ š³©¨¢ ¡¬˜©Ÿš ¡¢ ¡™£
¤ £¥¥š§¨© ª«¬˜©Ÿ
Ž)!$ (! !e%!$"-$ .!%++!$$!)$!%$++(! !!$+c. g!‘$!
c.dcf!c(!%$++!c(!$ g!’$!“$($ !”c+$!-cd+.!/-g$!•01!!*no)$
)$!
pqrs&
t!
$kc!$(($!+(($(! !)$! $c!%$++d$!c$c!/*g$!–0!'$c !)c!c-c+!+"!%++!
2834 4U345 8343853454 34 43374234P34 47 V9 5 T41254P545938 34
234P34 738 549 T
žŸ ¡¢ ¡™£ š¤ £¥¥
¦¡¢ ¡™£ š¤ £¥¥
§¨©­£«¬š®£¬š
§¨£¯£Ÿ¬š°¡¥ª£
±˜¢ £
§©¯˜¬˜²£
žŸ ¡¢ ¡™£ š¤ £¥¥
¦¡¢ ¡™£ š¤ £¥¥
®£™¡¬˜²£
µ¶·¸¹º»Ì
µ¾¹ÊÆȶ¾Ã»ÅÆÊÆ·ºÍ
ι¾Å¸Âȶ¾Ã»ÆÃŻϹ¾ÁÈÆж½¶ÈÄ
±˜¢ £
0
—˜™š›œ
!"#"$"%&'"#()*+
,
3
7
:
-./ %".0 1%/ %2
4 %"5 !$"6"+"02
DEFGHIJKLGFHMFMHJN OPQHRFSON OTL
8 &&9#./ %0$
UVWXYZ[V\]^YXY_`]aY\]VaabW]ZcWVb_cVbZ]Zc`]d[e`]Ve]Zc`]f`dd]eWVX]Zc`]XVX`\Z]ZcYZ]
ghijklmnnjomgjplqgjrisigltgiqjghijuvlwtgmvsxjjynnjzinnjt{gm|mgmiqjsiikjgvjoiji|tn}tgikj
eVW]Zc`[W]~VZ`\Z[Yd]eVW]aYb[\_]eVWXYZ[V\]^YXY_`€]]c`‚][\adb^`ƒ
„] W[dd[\_
„]†`X`\Z[\_
;)6%<&"$ -%$<"02
=0+%>& -./ %".0+ %06 ?%
A
„]‡`WeVWYZ[\_
„]†VX~d`Z[V\ˆ‰WYŠ`d]‡Ya‹[\_
„]‡WV^baZ[V\
„]Œ\`aZ[V\
„]Ž VW‹VŠ`W
„]Z[XbdYZ[V\
"&%06 B %+ 9.$++"02
C
8 %;%06&"02
"&%06 B %+ -"&6 *%".0+
hijtmwjvujghij‘lvk}{gmvsji{hsvnv’mqgjmqjgvjt{hmi|ijtsjmsmgmtnj“vzjiup{mis{”jvuj•––— j
Y\^]ZV]XY[\ZY[\][Z]YZ]Zc[]d`Š`d]^bW[\_]Zc`]d[e`]Ve]Zc`]f`dd]Zc[][]Yac[`Š`^]˜‚]`\bW[\_]
ZcYZ]~WV~`W]V~`WYZ[V\]~WYaZ[a`]Y\^]~WVa`^bW`]YW`][X~d`X`\Z`^]^bW[\_]\VWXYd]f`dd]
V~`WYZ[V\]Y]f`dd]Y]^bW[\_]Zc`]^W[dd[\_]Y\^]fVW‹VŠ`W]aYX~Y[_\€]]VX`]Ve]Zc``]
VbWa`]Ve]eVWXYZ[V\]^YXY_`]f[dd]˜`]^[ab`^][\]Zc`]eVddVf[\_]`aZ[V\€
™š›šœžŸ ¡¢£ž¤œ¥¡ ¡¦§¨œ¥Ÿ£©©£¤¦œª«§Ÿ¡¢£ž¤¬
­® vqg¯jklmnnms’j“}mkqjtlijw tkij}rjulvwjrtlgm{}ntgijw tgilmtnqjix’xj{nt”q°jz im’hgms’jt’isgqj
`Za€±]Y]f`dd]Y]aV\ZY[\[\_]^W[dd]abZZ[\_€]]²\d`]Zc`]f`dd][]^W[dd`^][\]Y\]b\^`W˜YdY\a`]
{vskmgmvs³jghijrliqq}lij­v|il¯jotnts{i°jli´}mlikjoigziisjghijklmnnms’j“}mkjtskjghij
W``WŠV[W]~W`bW`]ZV]‹``~]Zc`]f`dd]b\^`W]aV\ZWVd±]f[dd]W`bdZ][\]Zc``]Xb^]~YWZ[abdYZ`]
oims’juvl{ikjmsgvjghijuvlwtgmvsxjhijrt”jµvsijzmnnjgh}qjoijrlv’liqqm|in”jpnnikjtskjtj
pngilj{t¶ijzmnnjoijuvlwikjvsjghijq}lut{ijvujghijzinnovlixjjhiqijrtlgm{}ntgijqvnmkqjzmnnj
svgjitqmn”j“vzjot{¶jmsgvjghijzinnovlijzhisjghijrliqq}lij’ltkmisgjmqjli|ilqikj­mxixjghij
VW]f`dd][]~dYa`^]V\]~WV^baZ[V\·€]UVWXYZ[V\]^YXY_`]cY]Zcb]˜``\]aW`YZ`^€
‘lmvljgvjisgilms’jghijuvlwtgmvs°jghijklmnnms’j“}mkjqhv}nkjoij{hvqisjqvjtqjgvjlik}{ij
Zc`]ZVZYd]^W[dd[\_]aVZ]¸Zc[]\VWXYdd‚][X~d[`]eYZ]^W[dd[\_·€]]¹\a`]Zc`]~Y‚]ºV\`]cY]
˜``\]~`\`ZWYZ`^±]XY»[X[[\_]Zc`]f`dd]~WV^baZ[Š[Z‚]˜`aVX`]Zc`]‹`‚]aW[Z`W[Y±]`Š`\]YZ]
Zc`]`»~`\`]Ve]dVf`W]^W[dd[\_€]]c`]Vd[^]~YWZ[abdYZ`][\]Zc`]^W[dd[\_]Xb^]cVbd^]˜`]
{hvqisjgvjht|ijtjq}mgtonijqmµijqvjtqjgvjuvlwjtjpngilj{t¶ijvsjghijovlihvnijztnnxjj”rm{tnj
lintgmvsqhmrqjoigziisjklmnnms’j“}mkjg”ri°j{vqgjtskjghijlmq¶jvuj¼vlwtgmvsj½twt’ijtlij
qhvzsjmsjp’}lij¾x
01234567 28609 0 2659 2 7 0882258002596 4660825600
#45
9
6
6
22222222
‡ˆ‰Š‹ŒŽ
‘ˆ’“”‹Œ”“•ˆ–—˜™ˆ‘˜
šŒ•›ŒŒ—œŠ•‘Œž’–˜•“—
‹ˆ˜Ÿ– –‹œ“•ˆ–—“œ“‰Œ
–‹œ“•ˆ–—“œ“‰Œ
36 6
$
$
4 6!6" 5
6%6 6 &6 '&
9 6%6946 &6 '&
#45
56
78
7 FD2
()*+,*-.*/012134+56+053)+3)17+823*-+9/:*+/;<+3)*+65-./315;+1;=>*;9*+3)*+?@A34
B2
@A
2CDE
3)*+<-1221;G+.><+823-/3*+1;H/<*7+3)*+65-./315;I++()17+-/3*+56+1;H/715;+./4+0*+9/29>2/3*<+
6-5.+3)*+-/<1/2+=5J+*K>/315;7+J13)+/+25J+,*-.*/012134+L5;*M3)*+.><+9/:*N+,2/9*<+
;*O3+35+3)*+J*2205-*I++()*+-*7>237+/-*+122>73-/3*<+1;+8G>-*+PI
XYY
gbh_iRabTj_[\TkllZimn[mo]p
XY
X
YVX
YVYX
YVYYX
Qa]r _[RabT
q\]r \_lRZR[sTkr tp
XYYYY
XYYY
aZ
XYY
]
[
ab
XY
^
T
`\
Qa]r _[RabTq\]r \_lRZR[sT^ab[]aZ
X
_
T^
YVX
[\]
Z
R
Q
YVYYYX
YVYYYYX
XcdYe
XcdYf
YVYYYX YVYYX
YVYX
YVX
X
XY
‡ˆ‰Š‹Œ¡
‡–‹œ“•ˆ–—“œ“‰Œ¢”•Œ‹
’“ŸŒ’–—•‹–”
XYY
QRZ[\]T^_`\Tq\]r \_lRZR[sTkr tp
2222222222222
()17+8G>-*+7)5J7+3J5+-*G15;7u
QRSTUVW
M1N+ v )*;+3)*+823*-+9/:*+)/7+/+.>9)+25J*-+,*-.*/012134+3)/;+3)*+65-./315;I+()*++
+ 1;H/715;+-/3*+17+<*79-10*<+/7+0*1;G+>;<*-+823*-+9/:*+95;3-52I
wEEx2 yDB2Dz?E{z|A@}2~z?AEz|2z?2B@FD2z?€@AEz|2~B?€B@E}EA‚2E|ƒEF@AB„2ADB2{z|B2
22 E|2CDEFD2ADB2z?€@AEz|2~B?€B@E}EA‚2D@„2BFz€B2ADB2ƒz€E|@|A2@FAz?2E|2Fz|A?z}2
++ 21;G+3)*+2*/:+566+-/3*I++()17+599>-7+/7+3)*+,*-.*/012134+56+3)*+823*-+9/:*+1;9-*/7*72
22 ?B}@AE B2Az2AD@A2z2ADB2z?€@AEz|†
01
!"#"$"%&'"#()*+
,
3
7
:
-./ %".0 1%/ %2
4 %"5 !$"6"+"02
DEFGHIJHFIKLFMGNHFIOFPQLRLKSGPTUGKELVWTFMMXGJYGKEFGOZUVPWFGPIFGUFHFTUFTKGJTGKEFG
KSHFGJYGUILRRLT[GOZUGLTGZMFGPMG\FRRGPMGJTGKEFGUI
T[G
POFK
FIMG/
PTUG
KEFG]ZPRLKSGJYG
8LRRL&
&HPI
9
#.
%0$
KEFGOZUGFT[LTFFILT[GF^[^
_G`MGKEFGUILRRGMJRLUMGQZLRUGZHGQFLT[GVJTKIJRRFUGPKGPGRJ\GaPRZFb
;)6%<&"$ -%$<"02
_GcIFGKEFGOZUGHIJHFIKLFMGQFLT[GIFdPUeZMKFUGKJGKEFLIGVJIIFVKGaPRZFMb
DSHLVPRGFYYFVKMGJYGKEFGOZUGHIJHFIKLFMGJTGKEFGLTaPMLJTGUFHKEGPTUGYJIOPKLJTGUPOP[FG
%>&
-.
/ %
".0+
%06 ?%
fghijklmgklnohpqlrqlistgkluvwllx tylzl{hg|nlml}=0+
or~€lt}
l|hgklfkg
|kmj
k€lij
}kgl~mkl
por~olmjjhpnl‚tryljkmlh{{€lmqylmqƒl„n|mjjlnhjrylygrjjrqsl|tylfmg}r~jknl}om}lfmnnl
}oghtsol}oklij}kgl~mk€l}hlrq†mykl}okl{hg|m}rhql}hlml~hqnrykgmjklykf}owll‡oklgkyt~}rhql
LTGHFIOFPQLRLKSGJYGKEFGYJIOPKLJTGdGJIGYJIOPKLJTG
dGGUZFG
KJGOL9
ˆKZI
FGJYGUILRRLT[G
UPOP[FG
"&%06
B %+
.$++"
02
|tylij}gm}klrnlorsolmqy€l|hgklr|fhg}mq}jƒ€lhqjƒlyk~gkmnknlnjhpjƒlpr}olsgkm}kglykf}ol
h{lrq†mnrhqwllx tyl‰€lƒl~hq}gmn}€l{hg|nlml}orq€lorsojƒlr|fkg|kmjk€lij}kgl~mkwll‡ornl
IFMZRKMGLTGFMMFTKLPRRSGKJKPRGYJIOPKLJTGUPOP[FGLOOFULPKFRSGPUePVFTKGKJGKEFG\FRRQJIFŠG
QZKGKELMGUFVIFPMFMGIPHLURSGKJGPGRJ\GRFaFRGPMGKEFGUFHK
EGJY
LJTGLTVI
FPMFM^
8 %
GLTaPM
;%06&
"02
A
C
“’“
§ ¨¦Ž ª
“
§ ¨¦Ž ª
“”
­¬°̄›ŒŽ®Ÿ
®˜±̄¡¥¤™Œ
Ž®˜³¡²¦¥̈
•“ ´¦¥̈Ž˜¢¡±¯Œœ
‹
Ž
£˜

¡
¥
¦
Ž
Ÿ
œ
Œ
š
¤
ܼݸǹÀÞß
àáá¹â»À½áÀã¸äÀ»åæ¹À½¾À»ç¹À
¼¾è麼½¾Àä¹æ»çÀé¾äÀ»ç¹À
á½Çã黼½¾ÀäéãéݹÀæǽêë¹
"&%06 B %+
6 *%".0+
§ ¨¦Ž -"
© &
žŸ¡¢ŒœŽ‹¤œŸš
§ ¨¦Ž ©
¦¥̈Ž˜±̄¡¥¤™œ
´¦¥̈Ž˜¢¡±¤š¶
“–
­¬¤˜³²Œ›
š±̈Ž®
®˜³¡²
«˜¤œ¡šŒŸŽ—˜™š›
“
µ ˜¤œ±̄¯
—˜™š›ŽœžŸ¡¢Œœ
GGGGGGGGGGGGGGGG
·¸¹º»¼½¾¿ÀÁ ELVEGOZUG[LaFMGQFKKFIGHIJUZVKLJTb
ŒŽ‘ PVKLJTGPMG
DEFGHIJHFIKLFMGJYGKEFGUILRRLT[GOZUGPTUGKEFGUILRRLT[GOZUGÂGYJIOPKLJTGLTK‹FI
pkjjlmnl}okl~h|fjk}rhql}ƒfkllprjjl~hq}ghjl}oklpkjjÃnlfghyt~}r†r}ƒlm{}kglr}lomnlkkql
VJOHRFKFUÄ
žºÆ¹Ç¿
NLXG ÈJIGPGÉÊËÌÍÎÊÏÍÎÐÌÑÒÓÑÊÔÌÍÎÉÓÕ ÐÖÌÔ×ÓϊGØ ZUGÙGLMGFˆHFVKFUGKJG[LaFGKEFGEL[EFIG
ll pkjjlfghyt~}r†r}ƒlnrq~kl}okljkqs}olh{l}oklfkg{hgm}rhqlrnlnt{i~rkq}l}hlfkqk}gm}kl
G
G VJOHRFKFRSGKEIJZ[EGKEFGLOHPLIFUGÚJTFŠGVJTTFVKLT[G\LKEGYJIOPKLJTGMKLRRGIFKPLTLT[G
GG LKMGJIL[LTPRGHFIOFPQLRLKS^
LLXG ÈJIGPTGÓÐÌÏÎÛÓÖÌÎÉÓÕ ÐÖÌÔ×ÓÏGKEFGVEJLVFG\LRRGUFHFTUGJTGKEFGFˆKFTKGKJG\ELVEG
ll }okl|tyl~mklmqylij}gm}klmgklgk|h†kyl{gh|l}oklpmjjlh{l}oklpkjjhgklhq~kl}okG
GG \FRRGLMGHRPVFUGJTGHIJUZVKLJT^
01234567 28609 0 2659 2 7 0882258002596 4660825600
2234 5674869 26424 !" !#$!%&'())(*+,-(&(%./!*4
44 0 1428 42 436794 26414945 484463 46 4 42747 968 1467467 4 14
44 674 09 2542717682 672844821464290 4 42 49 :
;:<:44224246 7470 492 42 44821424 48 1468 742 4869 264 5 2=42: :464
24 42 44821486 48 44 42 4 73674 2644147 968 42 436794 26414945 437694
8 774 73674 26:44
> #*?/#".'%#$#'/'.@ !$'.$!'(!%./ !&'())(*+,-(&".*'.) !&!$ .A (" 36794 26414945 484468807:
B234 CDEFGHIDJKLEDLMEHIMNO#*(*"'!#%!&&!$ ./&'()),-(&(*P#%(.*(%./!*4
4 66 78 1492 444475 74 67 Q2R 4Q46742252 74 79 462274Q436794 264246 254
&'())!&&-!.'!&-"!&!/S"(!*"?./S)!'"#T!/.'@ #(.*UV*/#" (+ )?4
(@$#('(*+).%%!%./A .)!@-&W(U!U*.S)!'"#T!/.'@#(.*X"#*Y!.Y%!'P!&4
44 68 4 2 436794 264 79 46227424Z4[4787467457 4 74B1 125464 2 4
44 1 42 14 67 45 69 773=40 4 824 430214 6486 76 4411228 4
WY#%("#))?)#'+!'$#'(")!%.$'.@.!S)!'"#T!/.'@#(.*X#'!#&&!&. !4
&'())(*+,-(&U
B2234 \LMJK]D^MKHIFMO,-(&).%%(%#".*(*-.-%$'."!%%O !).*+!' !.$!* .)!4
44 24_ 6 14642 4172 254901=42 457 4 742 42842641 2:44̀22429 2 424 4
!'!(%#+'!#!''(%T./(@$#('@!*/.' #$#'./ !/.'@#(.*&'())!&S'%4
44 869 47 14642 436794 264172 144 746:44`22424 47280 477429 674 43674
).*+ .'(a.*#)A!))%A !'!".@$)!(.*(*!'P#)%+'!#!' #*bccc@#'!4
44 6 4086996:
B22234defghijiklfO,-(&).%%m(*P#%(.*&!$ (*"'!#%!%#% !&(//!'!*"!(*4
44 7 07 46 9 42 4172 2549 0144142 47 786274 7 07 4B68 764 48 34
44 287 4 :4n1 7464 48 4172 254414869 264 822o0 46946 25
44 0 1464172 49 46444644862142247 463436794 26414945 :
B2834pqgfrqjfstukivwlxA()) #P!#)#'+!!//!".* !,-(&).%%
4
4
44
44
B434 y475 4B 65441z674921 41249 734<6 69 4{6 4| 9 6 2 4
4
:5:44622 744142 48477725460 463494747601472 4 :5:4648245 4
4
2 4172 462467441}0 42 4127 8 264 4172 2544 9 6 7492 47 0 424
!/'!~-!*%"'#$(*+./ !@-&"#T!U€'#,-(&).%%A())'!%-)#%4
4
2 4901484 4247 67 146424 o022672094228 :
4 B634 {252482780 4 26474 492 4287 4 42 4174928468 764 48 4922 42 4
(+ ,-(&P!)."((!%A())!'.&! !@-&"#T!U‚ ! (**!'@-&"#T!4
#*& !+'!#!'.P!'Y#)#*"!A())Y. '!%-)(*(*"'!#%!&,-(&)!#T.//U
‚ !'!)#(P!,-(&).%%EGHMN4446824 1492242 4847260424 46342 4172 254768 4
#'!())-%'#!&(*S+-'!bbU‚ !".''!%$.*&(*+'!)#(P!ƒD^„FMN43674 4824 24 447 4
#)%.% .A*U‚ !'!)#(.*% ($Y!A!!*&!$ ./(*P#%(.*#*&,-(&)!#T.//P.)-@!
@#?Y!~-#*(S!&Y?#%(@$)!P.)-@!'(""#)"-)#(.*#%% .A*(*‚#Y)!bU‚ !#Y)!
% .A% #)#'+!P.)-@!%./,-(& #P!.Y!).%.#" (!P!%-Y%#*(#)&!$ %./
284264247 4646 4 6760436794 26:4
01
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EVSFbHVWXXHTYU]HiVWGHX\YV]j
9 ''!: $/ 0 &1%
K
K QRSTUHVWXX
YUHZFUH[Y\]
YXH[T]X^
[WT_ YUFW`H
]aRWX]b
O P
N
<*7 &='#%!. &%= #13
M
P EVSFbHVWXXH
WRRWXFU]
XUYZFVFe]TX
O c]YTHZFU
USTZSV]`UH[VWd
[VSFbHVWXX
M EVSFbHVWXXHd^FV]HbTFVVF`G
WRRWXFU]HbTFVVHRFR]
EVSFbHVWXXHfWVS_ ]
>1,&?'!./ 0 &#/1,!&17!@&
B
N QUYUF\H[FVUTYUFW`Hd^FV]
UTFRRF`GHbTFVVHRFR]
#'!&17!C &,!: /%,,#13
!4566
3 53 ""5 ! 545
!33 5! #4""
PLHX]\
gH_ F`
KH^T
KbYh
kF_ ]
D
PHbYhX
9 & !<&17'#13
OLHbYhX
#'!&17!C &,!.#'7! + &
#/1,
EFGHIJ
KL
01234565
741859 5 3 451545
“{”Ž•yy
uˆ‰ m mmmmmmmmnsˆ‰ m mmmmmmmmuoˆ‰ m mmmmmmmmnsoˆ‰ mmmmmmmmmmuooˆ‰ mmmmmmmmvooˆ‰
–—˜|Œ—y‹‡
yz{|}y~y€‚ƒ„ {y„ †‡m
Š‹Œ
mŽy‘’ ‡
owopm
ptm
owrtm
owtvm
num
nom
nqpm
owosm
vvm
owutm
nwovm
nqm
npm
rstm
owov
qsm
nwutm
owpqm
rum
nq
uunm
owoq
nosm
owspm
nwsxm
rqm
rrm
posm
owoxm
nnpm
owvum
nwtsm
unm
rpm
ppr
ll
™ š››œžšlžŸ ¡¢£l¤ž¥ ¦§¨§©ª«¬­®¯§°±²³´²µ¶±·§¸¹¶±²ºµ»§¼½¦§¨§¾ª«¿®
ÀšŸÁÃÃlĝ›¢ÅšlÆlÇÈÉlʤž¥ lËlžÌÍlÂlž¥ ÍÎl l
l
l
l
¤ÏÌ
ÐÑÒÑÒÑÓÔÕÖ× ØÙÚÕÛÓÜØ× ØÝÞÓß ÞàáØÛÚâ× âÓÜãÖÚÛÝÓÜÖÚääÚÛÝ
ÐÑÒÑåÑÓÔäãÚæÓçÕââ
l
è½±§®±­½µ¹¸é®é§êë§ì½¸­½§¶½±§·²¸»»¸¹í§®î·§ï»¶²µ¶±§¶½µ¶§»±µð駴³³§¶½²´îí½§¶½±§ï»¶±²§
ñŸÁšl¡òólóɚlݞşó¡òl¡òñ›¢ šô
¤¡Ìl õòñžšŸ£¡òölóɚl¥Ÿóšžl£Ÿó¢žŸó¡òl¡òlóɚlòšŸžl¥š››œžšlŸžšŸl¤žš ¢ñš£lóɚlžš›Ÿó¡÷šl
§§ ø±²®±µê¸»¸¶ë§¶´§´¸»¦§¼ïíî²±§©¾¦«§è½¸é§¸é§øµ²¶¸­î»µ²»ë§¸®ø´²¶µ¹¶§³´²§»´ì§ø±²®±µê¸»¸¶ël
ll žñÁ£l¥ÉšžšlóɚlžšÅ÷Ÿ›lÃlóɚlùšúóžŸûl¥Ÿóšžl£Ÿó¢žŸó¡òlşülóŸÁšlŸl›òöl
ll ó¡Åšýl¡þšþl¥š››lùñ›šŸòl¢ÿûlşülóŸÁšlşòülŝòóÉ£þllõòlßñóýl¡ólşülòólÿž÷šl
ll ÿ££¡œ›šlól¡ò¡ó¡Ÿóšlÿž ¢ñó¡òl¡òlŸlžšŸ£òŸœ›šló¡Åšl¡Ãlóɚl¥š››l žŸ¥ ¥òl¡£l
§§ ¹´¶§éî³ï­¸±¹¶§é¸¹­±§¶½±§¼²±»µ¶¸º±¦§ø±²®±µê¸»¸¶ë§´³§¶½±§½ë·²´­µ²ê´¹§ø½µé±§½µé§§
l œšñÅšl£l›¥þ
01234567 28609 0 2659 2 7 0882258002596 4660825600
23345 6758598 5
35 5833755852345358 55 58 85 33795
!"##$%&!#"' ()"#"*+,-%./01
DEFGHEIJKL KME
NFOEHIJKL KME
23456473789:6;
:6;
2344<5=9> <5=?
555
2;<Y9Z<?56W;=8
23<59[<4B9\?<64
555555555
[<4B9\?<648
@4A<B=B9C6;5?<5=
]7<?5^93?9
:5_=?92=S =459
23483;6B<5=89\?<648
z{|}~{€‚ ƒ{
:6;9U?3T;=5
„|{~€‚ ƒ{
jabdefghd‘u tnwnw
`abcdefghdgidjab
ˆalwdjparlb
‰yapmŠdgpd
qm‹npdknu nlmd
kglbgfrwamnbdjparlb
qrfdspgtfnm
jab
233345’“ 8”87 5 75375 5 33795“35• 85”879 5 5 • 8375
55 – 8—333 5235 83 5 • 8—33535– 537587535– 5 •8375875375
55 85–8 5– 57 4˜55™“ š5 8”3755 5 “ 8”875 75375 5•“5
wwrmrgldgidm‹rpwdt‹abndpnwy‡nbdŒabdifgh erŒdŽa
-#›11œ+("'+"#(%#"+(%$ž
(*&%!"++œ5
'"&+%Ÿ"Ÿ+"#(%#"+ "!œ+("+,-%. 01
01
¡¢£¤¥¦§¨©§
¡ª¥«¬­¢ª®§¢«¯¬¢¥«¦®­§°¤¦§
­ª§±¬­¦¥§§²³ª´µ
C6P9QR2
S <56349@S T<6?S =459U7=9
539
> <5
9V;
kgllamndoC3?
amnp
qrfdsp
gtf
nm=?
dep
gu3WX
d
qrfdvabnwdx yw
kgllamndo amnp
kfacd†apmr‡fnb
kgamdˆalwdjparl
\<8
¡¢£¤¥¦§¨¶§
·°°¢­¢ª®§ª¸§­¹¢¥°§¯¹¬º¦§
¥¦°¤´¦º§£¬º§»ª±
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
XYZ[\Y]^_` _aY
<*7
&='#%!. &%= #13
hJLiGKLMNJOMjMTHMR PIQSMkl KOLJGH
FGHIJHKGKLMNJO
FGHHPIQMR PIQS
EEE
FOPmMnPSIJiOQL
FGPIMoPHVMpSPJH
bZcY\]^_` _aY
9 ''!: $/ 0 &1%
oPHVMpSPJHL
WJO#
IS/1,!
PIQ
>1,&?'!./ THUPVQVM
0 &
&17!@&
B
pPL
NJOC
MtSGuO
QI : /%,,#
#'!&17!
&,!
13
qKPSIrMGSM
NIsQSMFQl QHIM
FGHLGOJVPIQLMpSPJHL
D
vwxyz{|}z~€€‚}zƒ‚}„z€ z‚|}z†‡ zˆ‰‚„‚}z„9
zw€Š&
}~‚
w€z
z†Œz}z
w€~†Ž‚w‰}E
!‹‡w
<&17'
#13
EE ‘’“”•–—˜E–—E™‘š–™–•›•‘’Eœ‘–—˜EžŸ ‘ ¡E¢“š£E™‘š–™–•›•‘’E¤–
””E‘ “š‘E
WJdMefg
EE •£‘E™‘Ÿ‘›œ–”–•¥EžE•£‘E—‘›E¤‘””œž‘EžŸ›•–ž—¡EE¢–Ÿ–”›”¥¦E™ž”¥Ÿ‘’E–’’ž”§‘ E
zz w€z‚|}z†‡ zˆ‰‚„‚}z†Œz¨„z€z‚|}z©„†‚w€z¨‡„©~}¨z„}¨‚„w~‚w€ªz‚|}z„}E
#'!&17!C &,!.#'7! + &#/1,
zz Ž}€z‚z‹ƒzvˆª‡„}z«¬y­
01234567 28609 0 2659 2 7 0882258002596 4660825600
!"#$%"&'() (*"
+#,"%&'() (*"
95 5 3
9 3953
53
9 5 5 3
9 5 5- 3
5- 3
523 358 9
. /59 5
0 5 5
993 5- 3
- §¡¨©ªž«¬­
§®ª¯ ¡®°«¡±¯¡ªž° «²©ž«
®«Ÿ³¯´«µ¶ž³³¡°¨
123456789:;<34=>4:234?3@536A9B9:C4@3<;D:9=84>@=54A=:24:C?3E4=>4>=@56:9=8FG;9<4
HIJKLMNJHOIPQHRRPSKTKISPOIPJUKV
WHXP YRMZPNUK[H\JLZPWJUKLKPMLKP[MIZPJZTK\PO]PNRMZX
395 3 58595 59 2345678
29
WHHXP YRMZP[OLTUORO^Z
WHHHXP_[O`IJPO]PNRMZPMIS
a9bc4d2359E:@C4=>4:2345;<4eB:@6:34f4?g468<4E6B989:C46@3495?=@:68:423@3
hUKPSH]]KLKIJPJZTK\PO]PNRMZPNMIPiKPLKNO^IH\KSPiZPJUKHLPMTTKMLMINKjPPkH^`LK\PlmPnPolP
\NUK[MJHNMRRZPLKTLK\KIJ\PJUKP]OL[\PQUHNUPSH]]KLKIJPNRMZPJZTK\PMLKP]LKp`KIJRZP]O`ISV
qPrstuvwvxy4>=@5E4A==zB3:E4{29D24A@9<734?=@34:2@=6:E4=@4eBB4?=@34{6BBE4ae7;@34|}c
qP~uuvxy4>=@5E4>@679B34{9E?C4eB6538:E4{29D24B9834?=@34{6BBE468<4A@9<734:234?=@3EP
4 4:235E3Bb3E4a97;@34|€c
‚ƒ„ †‡†ˆ‰ŠŠ‹Œ†‡ƒˆŽŠŒ‚‡†‘
’Š †ˆ“‚‚”ˆ•–ŒŠ ‚‡†‘
§¡¨©ªž«¬·«¸³ž¹º
»¡µŸªž ž«¼®®½³ž µ«
¸¾¯®³¡°¡žº
‚ƒ„ †‡†ˆ‰ŠŠ‹Œ†‡ƒˆŽŠŒ‚‡†‘
PPPPPPPPPPPP
‰ ‚Š—”‚
”ˆ˜ŒŒ‚‡†‘
’Š †ˆ“‚‚’Š
”ˆ†ˆ
•–Œ
‚‡†‘
qPœžŸ ¡ž426E4642=83CD=5A4E:@;D:;@34a97;@34¢£c
qP¤¥ut¦vxyPH\PMIPHLOIPNOIJMHIHI^PNRMZPQUHNUPRHIK\PJUKPTOLKPQMRRPWkH^`LKPolX
01
’Š †ˆ‰ ‚—”‚”ˆ˜ŒŒ‚‡†‘
™ ‚ˆ•Š‡‚”ˆš› †„‡‚‡†‘
§¡¨©ªž«¬¿«¸ª¡¨À º
Á®ªž«¼ª¡²¨¡°¨«¸Â³³¡žº
™ ‚ˆ•Š‡‚”ˆš› †„‡‚‡†‘
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
9 ''!: $/ 0 &1%
<*7 &='#%!. &%= #13
>1,&?'!./ 0 &#/1,!&17!@&
B
#'!&17!C &,!: /%,,#13
D
9 & !<&17'#13
#'!&17!C &,!.#'7! + &#/1,
01234567 28609 0 2659 2 7 0882258002596 4660825600
2345678998 8 88356 78 8 7 87 8 8776 8768 78388678
!"#$%&'()*!+)&,,-.(,/&$! !%-'!+ !"+!*/!+01 !2"&2!( !'&$ !'!-
7856 78878678 84567899 3843877 8 878553867378 878
+#'8&)!%'-%!'(!+-8 !%&'()*!+&,$ !%-'!.&**+098 !%&'()*!+$-,-&$ !'!-
!%-'!.&**+ !2)&,)-,(,#!-&,(,!',&*%-'! '-&. !'!:*-);(,<"&2,-.
5684567899388=8 8 787 3783 43 78 88674 3 3 48786343 8767 3378
67>53678676 8858 83 76 8768 7?8 8358673
@!'2+"&**$(&"!!'%&'()*!+.(**%&++ '-#< !%-'! '-&+.( -#:*-);(,<0
A () -8 !&:-/!%'-)!++!+-))#'+$!%!,$+"&(,*2-, !B%&'()*!C8-'"&(-,D+(E!
'&(-&,$ !%&'()*!)-,)!,'&(-,0F-'! '-&+&'!-8!,$!+)'(:!$&+:!(,<"()'-G
7868 68778 7898 8456789H86837 3 I
J KLMNOPNMQORSMNTRUOVOWXYOZO WXYOZOVOJ QUNOPNMQORSMNTRUOVO[XYZ J TLMNOPNMQORSMNTRUO\O[XYZ
‡‚p ‚sqt^ˆn`]on‰‚n xu`„Š †
a||
c defg
c hig
ÇÈÉÊËÌÍÌ
ÎÏÐÑÒÓÔÕÑÒÖÔ×ÑØÙÔÏØÙ
c jefg
k^_l`
mnop nqr^s^tu`
vwxy
zw{
mnop nqr^s^tu`
vwxy
|
|
|}~
|}
a
a}

mwon`€lowqt`vq^‚ƒ`„p ^xow ƒ†
a|
~|
ÚÛÜÝÞËÌÍß
àáÞËÌâãÞáÈâÌäÛåËÌáæÌãÛÜãÌÈçèÌ
ÊáéÌêËÞëËÈÉÛÊÛâìÌÞáíîä
478768 786 3 86783856877 ?86884367 86 ]^_`3ab ?8 878
676378878633 48553388478 683 567863 3 8578873 78
3‹78878 637887878878633 48553?85788Œ 8Œ 6 ?88385 6378
3 84 78H88776 8768 78 8 78678 78 63787>536 7 88 78368
8786786 83 776?866378778 678677 83 8537 8 7 6 3 38
Ž7678 8 6378 768 8‘ 88786786 83 7768 8 86548
!'!+'()(-,+%'!+!,!$:2 !%-'! '-&09,!'"!$(&!+(E!$%&'()*!+.(**'!+#*
3 8 83 76 8768 73823456789H84 6787873 86 3 86786 83‹78
378456789’84367878667 3 48456786878 6378 788633 48
5538 877 3
“”•–—˜™š”›
¿«¦³¯±±
¹¯¶ ¦º°¥
01
À¦©¯
œžŸ œ¡š
žŸ —œ¡š—œ—žŸ¢—œ¡š
¤¦«¶ ±¨µ²¥©¯«¨³°¸¯¨°©¨ ¤¦«¶ ±¨²´©¯«´°¥¨µ²¥©¯«¨³°¸¯¨°©¨
µ¦«¶ °©²¦´¨±¬«µ°³¯ ·¯®©ª¨²´¨©ª¯¨µ¦«¶ °©²¦´
¹¯¶ ¦º°¥¨»¼¨½°³²·¾¨ ¨Æ²µµ²³¬¥©¨©¦¨«¯¶ ¦º¯¨¨
§°±ª
œ£žŸ¢—œ¡š
¤¥¦§¨©ª«¦¬­ª¨®¦«¯¨
©ª«¦°©±¨§²©ª¦¬©¨
³°¬±²´­¨µ¦«¶ °©²¦´¨
·°¶ °­¯
À¦©¨´¯³¯±±°«¼¨Å¨
´¦¨·°¶ °­¯¨
Á²­ª¨®°«©²³¥¯¨³¦´³¯´©«°©²¦´±¨½ÂÃÄ ¾¨«¯·¬³¯±¨©ª¯¨°»¦º¯¨º°¥¬¯±
ÇÈÉÊËÌßÌ
Ì
ÚÛÊâËÞÌïÈîËÌæáÞëÈâÛáçÌ
êÞáíËääÌèËêËçèäÌáçÌêÈÞâÛíÊËÌ
äÛåËÌðÌêáÞËÌâãÞáÈâÌäÛåËÌÞÈâÛáÌ
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
E FGHEIEFJKLMHMNEGONMEHPNOKHEEEEFEIEFJKLMHMNEGKNHJQRM
9 ''!: $/ 0 &1%
EEE
{jVe abfVjTwab
Xcd
e abf
Wgh
e abf
jTZT`[VSZkl^
YZ[\V][^_T`ZabV
hV
>1,&?'!./ 0 &#/1,!&17!@&
B
Yat am_V][^_T`Za
z
!"#$%&'#(&
ih
e abf= #13
<*7 &='#%!. &%
ch
rh
uh
ih
#'!&17!
C &,!: /%,,#13
Wrh Wuh Wih chh
Whh Wch
e T`^kmVj`[Za
D
9 & !<&17'#13
EEEEEEEE
|}~EPME€LMP‚EJƒMENM„MNNMFEHOENMRKHMEHOEHPME †EHK‡FKNFEJMˆMEJƒM‰EKE„NMŠ‹M‡HRŒE
t[_
TkmV
v[t
[UaV
jTwaV
[_THkmbfT
yb
‹MFEGKNSk^
HJQRME
JƒME
LMK
‹NME
J‡EHPME
OJRxaZ
EJ‡F‹
NŒEM~~
E„ONENKˆMREGKQŽEK‡F~
nWohVTm`fVpVcdqrhhVe T`^kmbs
‘’“”•–—˜™š•“›œšžŸ
#'!&17!C &,!.#'7! + &#/1,
PMEQOL J‡MFEJ‡HMNKQHJO‡E MH¡MM‡EHPMELK‡ŒE„KQHONE¡PJQPEM„„MQHEHPMEORJFEK‡FE
¢£¤¥¦¤§¨©ª«¦¬©­ª¨®¥­¢£§¨©ª¤­¨¤¯§¨°­¥±¦¤©­ª¨¦¥§¨¬²±±¦¥©¬§³¨©ª¨¢´²¥§¨µ¶·¨¸¯©¹¯¨¬¯­²£³¨ STUVWX
MEMR„EMºGRK‡KHONŒ~
ã¼ÇÉÂӼƼÏÄÀÕÂÞÂӾƽ ¾ÈÔÄÅÀÕ
ëìæåíîïðïäåæçèéçê
â¾ÆÕÂÙÄÖÌÂ
ëìæåíî
äåæçèéçê
ÙÄÖÌÂ
äåæçèéçê
ñ ¾ÎÄͽ Â
äåæçèéçê
)
*+,-./
ò¼ËÂ
ó¼ÅÄÎÏ
ßÄÅÀÆÈÀ¾
ô¿Ä¿ÒÈξÎÂƼÇÉ
»¼½ ½ ¾¿À
Á¼ÂÃÄÅÀ¾ÆÂÇÈɾÂüƽ ÈÀļ¿Ê
Ë̼ž½ ÍÎÂżÏÏÂÄ¿À¼Â
üƽ ÈÀļ¿ÊÐÂξ¾ÑÂÄ¿ÒÈÏļ¿
ÓÈÆÀÄÈÅÂÃÄÅÀ¾ÆÂÇÈɾÂǼ¿ÀƼÅÊ
Ä¿ÒÈÏļ¿ÂξÑÀÌÂƾÎÍǾÎÂÔÕ
ÌÄÖÌÂѼƼÏÄÀÕ
×¼¼ÎÂÃÄÅÀ¾ÆÂÇÈɾÂüƽ ÈÀļ¿Ê
Ǽ¿ÀƼÅÏÂÃÅÍÄÎÂżÏÏ
×¼¼ÎÂÃÄÅÀ¾ÆÂÇÈɾÂüƽ ÈÀļ¿Ø
ÙÄÖÌÂξÑÀ̼ÃÂÄ¿ÒÈÏļ¿ÂÎ;ÂÀ¼Â
żËÂüƽ ÈÀļ¿ÂѼƼÏÄÀÕ
ÐÂÚ Ì¼Å¾Â½ ÍÎÂÇÈ¿ÂÔ¾ÂżÏÀÂÄ¿Â
üƽ ÈÀļ¿ÏÂËÄÀÌÂÉÂÛÂÜÝÊ
ξѾ¿ÎÄ¿Ö¼¿Â½ ÍÎÂÞÂѼƾÂ
ÀÌƼÈÀÂÎÄȽ ¾À¾ÆÂÎÄÏÀÆÄÔÍÀļ¿
E
õö÷öøùúûü ýþÿú0ø1ýü ý23ø14ûÿ02ø53ü 30þÿ02
PME‹QQMEO„EEKEQKJ‡EONERJ‡MNEQMLM‡HKHJO‡EJEFMGM‡FM‡H‰EKLO‡EßÄOH
PMNEHPJ‡‰EO‡EHPME
ÖÂàØÜá
NMLOˆKREO„EHPMEL‹FEQKŽM~EPJEJEO„HM‡EKQPJMˆMFE ŒEG‹LGJ‡EˆKNJO‹
E¡KPMEK‡F6ONE
GKQMNEQO‡HKJ‡J‡EFJGMNK‡HEKFFJHJˆME7‹N„KQHK‡H8~EEPMENMLOˆKREO„EHPMEL‹FEQKŽME
¤¥©´´§¥¬¨©ª¹¥§¦¬§³¨9²©³¨£­¬¬¨¸¯©¹¯¨±¦ ¨ §¨²®¨¤­¨¦¨³§®¤¯¨­°¨¬§«§¥¦£¨¹§ª¤©±§¤¥§¬
01234567 28609 0 2659 2 7 0882258002596 4660825600
2345675894 596 9798 85946693934968753895 979 6 753 947 9895 9359
34974743893997568995974 97 97939 856753 9 9743 9753 9
397948795 968335753 9 9743 976 7589743 96 38943 89 69
!"#!$%&'($#$!")*+'',%-).*-/'!*&*0)'(#%--$*('."1*'"'1*-2'.$3.'0")%-"#'4%$5'#6(('%0#*(('
!60)-6##*5'72'(%$)"7#*'"55$)$1*(+'8-69*-'4%$5'#6(('!60)-6#'$('"'0*!*(($)2'($0!*'*:!*(($1*'
86753 9 95894669344948795 9734689 95898 859; <958946698 39
5 93 95 9 95 393478 859 95898 859359 89 935968 7339
$0').*'!"($03'$)(*#=+''>.%(/'5*(9$)*').*'5"&"3$03'0")%-*'6=').*'!*&*0)'4%$5'#6((/'$)'$('
06)'06-&"##2'"'9-67#*&'($0!*').*'?#)-")*'@ABCDE979859 937 3 93943 3589
7979897 89958986 6753 >.*'*==*!)'6='5-$##$03'"0'61*-'3"%3*'.6#*'60').*'9*-=6-")$03'*=?!$*0!2'$('6='&%!.'
3-*")*-'!60!*-0+''>.$('$('$##%()-")*5'$0'?3%-*'FG'H.*-*').*'96H*-'6=').*'9*-=6-")$03'
3%0'H"('(%=?!$*0)'=6-').*'1*-)$!"#'9*-=6-")$60(')6'9*0*)-")*').*'!*&*0)'(.*").+''I)'H"('
$0(%=?!$*0)#2'96H*-=%#')6'9*0*)-")*').*'!*&*0)'(.*").'$0').*'.6-$J60)"#'5$-*!)$60+''K6)'
497 9 898897 5895 9 78958 89L86 6753 MN959 6893 657549589
84496 5335934498968893895898 359 986 6753 9344989 49749
5898 39748926586958968 733986 6753 93449 575948 9 6 753 9579
$0').*'$5*"#'!"(*+''O0P3"%3*'5-$##$03'6=').*'9"2'J60*'!"0'"16$5').$('9-67#*&'Q?3%-*'FGR+
hVVUWXYZU[\U]V^]_XY^T`
\UTUX]_XUc[XeU[WUf UTX
ceU_Xe[a_Tk[V^]f _XY^T[k_f _jUtg
STUVVUWXYZU[\U]V^]_XY^T`
ab^Uc[T^X[dUTUX]_XU[
X^[XeU[V^]f _XY^Tg
vTwxy_jU[l^iU
b]Yii[zYX[kY_f UXU]
u^]f _XY^T
mno npq
99999999999999999999
rXUUi[sYTU]
hTi_]jUk[l^iU
{98 9 6 9 9 6 753 97 789 6 9898 853979 6 753 93939
0")%-"##2'=-"!)%-*5+''>.*'4%$5'#6(('!60)-6#'"55$)$1*'"55*5')6').*'!*&*0)'(#%--2'H$##'7*'
$0*==*!)$1*'$0'(%!.'H$5*-'5$"&*)*-'=*")%-*('|'(6'!*&*0)'H$##'46H'$0)6').*'=-"!)%-*(/'
5853744986 785494 3958 99}8986 6758984496 533593449 989
94 8695795898~3748597489 6979 89 489 4853 9389 853 9
34497898894 5935958967568 939789 563589589 7; 6359 9589
H*##'$046H+
789 9589 4853 98 395 9 89 48939589 3 48 59795 97 39539
6 48 99}893586359 9589 489 79896 58589979 4 558943869 69 6889
}3939 6 7449 3489389589 6 753 9 598956 98 95 9 659 89
67568 9869589687343968 86 369568 9 353 01
€‚ƒ„ †‡ˆ
‰Š ‹Œ ŠŽ† „‹„‘ŽŠ‚†Š†‘†
 ’ ŠŽ “†’” ŽŠ†“ƒ †Ž†
‘Š†• „‚‘ƒ‚ †–”
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFGFHIJKL MNOJPHQML MRSHQTKOPRHUSKVJKMNOPR
WXYZ[Y\]^\_`abcZ[\^dYeeZ\Yefg`eZabZ_bZ^[YbZ[Y\
]^\'
_`
a^bZYbdg
`XabZ_Zg^iZ
9 '
!:
$/^e0YhZia
&1%
jklmknopqprstuvwktvxtxlnyrzlk{t|lnpw}tnw{tlkmwnwr}tvxtr~ktjklxvlnrpw|ty~nl|kt€|zlkt
‚ƒ„tt ~p}t†{nmn|k{‡tnlkntlk}zqr}tpwtntlk{zyk{tpwˆv‰tpwrvtr~ktjklxvlnrpvw„ttŠv‰k‹klŒt
r~ktmkn}zlk{t‰kqqtlv{zyrp‹prstŽw{ktp„k„tˆzp{tjlv{zyrpvwtjkltzwprt‰kqqt{ln‰{v‰wƒt
e^Y`aYeZabd\Y_eYeZia`XZ`aYZ_eZ`XYZiYggZaeZ[\<*7
^hfdYh‘&='
ZZWXa#
e%!
Z’dg
bcZf[“Z
^dYeeZ
.Y_ba
&%
= #[\
13
aeZ^]`YbZ_ed\a”YhZ`^Z`XYZ\Y^•_gZ^]Z[Y\]^\_`abcZhY”\aeZ–dX_\cYZhY”\ae—Z\^d˜Z]\_cYb`eZ
nw{tr~ktqv‰tjklmknopqprstylz}~k{tuvwkƒ™tnqqtvxt‰~py~tlk{zyktr~kt‰kqqtpwˆv‰„tt ~p}t
\Y^•_gZabd\Y_eYeZ`XYZ`\_beaeea”aga`šZ”Y`iYYbZ`XYZiYggZ_bhZ`XYZ]^\_`a^b‘Z›f\`XY\—Z
&?'!./
0 &tr
#/1,!
}vmktvlp|pwnqqstyvmjqkrkqstoqvyœk{tjklxvlnrp>1,
vw}tmnstvjkwtr
vtˆv‰tn}
~kt‰kq&17!
qt @&
{ln‰{v‰wtpwylkn}k}„tt ~p}t†yqknwtzj‡tjlvyk}}tynwtoktnyykqklnrk{tost}jkyp€yt‰kqqt
`\Y_`Yb`eZefdXZ_e
B
#'!&17!C &,!: /%,,#13
pƒt žnyœtˆv‰pw|tjlv{zyrpvwtzw{klt~p|~t{ln‰{v‰wƒ
D
–aaŸZ dahaeabcZ–`^Zhaee^g•YZ`XYZa[_a\abcZ_`Y\a_gŸ
9 & !<&17'#13
–aaaŸZ¡Y\]^\_`abcZia`XZ̀XYZiYggZfbhY\”_g_bdYZ–`XYZd\feXYhZ_bhZa[_a\YhZ[Y\Y_”aga`šZ
ZZ \^d˜Z_eZiYggZ_eZ`XYZdX_\cYZhY”\aeZ_\YZ\Y^•YhZ_eZe^^bZ_eZa`ZaeZ]^\YhŸ
¢£¤ ¥¦§¨©ª«¬­®¯°©±²³©ª¬´£µ©
¶·£¸¬­©±¤ ©¦¹²¯²¨º»
#'!&17!C &,!.#'7! + &#/1,
­©±Â£±¦¨²£µ¬¢Ã¦±¾©¬Ä©¹±²Å
Áµª¦¤ ¦¾©ª¬¿£§À
ÙÚÛÜÝÞßàáß
âãäÞåæçÚãßèéßêæåæÛÞêß
ÝÞÛÚèëßæÝèÜëêßìÞÝéèÝæçÚèëß
éèÝåÞêßêÜÝÚëÛßçäÞß
ìÞÝéèÝæçÚèëßìÝèãÞíí
Ç¥©µ¬­©±Â£±¦¨²£µ¬È®µµ©¯
ZZZZZZZZZZZZZZZZZZ
¼±¦²µ¬½±¦§¨®±²µ¾
¢¦Å²µ¾¬Æ¬¢©¤ ©µ¨ ¿©ª®§©ª¬­©±¤ ©¦¹²¯²¨º¬¿£§À ¶·£¸¬­©±¤ ©¦¹²¯²¨º»
EFÉFHIJKL MNOJPHQML MRSHQTKOPRHUKJÊTËNOJP
WXYZg_e`ZeYd`a^bZhaedfeeYhZ_Zd_eYZabZiXadXZ`XYZiYggZ[\^hfd`a•a`šZabhYÌZabd\Y_eYhZia`XZ
`aYZhf\abcZb^\_gZiYggZ[\^hfd`a^b‘ZZÍb]^\`fb_`Ygš—Z`XaeZ]_•^f\_”gYZe`_`YZ^]Z_]]_a\eZ
]\YÎfYb`gšZh^YeZb^`Z^ddf\ZÏZ\Yhfd`a^beZabZiYggZ[\^hfd`a•a`šZabhYÌZ^]`YbZ^ddf\ZiXYbZ
_ZiYggZaeZ^bZ[\^hfd`a^b‘ZZWš[ad_gZ[\^hfd`a^bZ]^\_`a^bZh__cYZ[XYb^Yb_Z`X_`ZgY_hZ
`^ZefdXZ\Yhfd`a^beZabZiYggZ[\^hfd`a•a`šZ_\Y
БZ›abYeZ^•YYb`
„tÑ}ktvxtpwyvmjnrpoqkt‰vlœv‹kltˆzp{}
ґZÓb^\c_badZ_bhZ^\c_badZed_gYZ]^\_`a^b
ԑZÕ_d`Y\a_
֑Z¡\Yeef\YZ\Yhfd`a^b
בZØ`afg_`a^b
01234567 28609 0 2659 2 7 0882258002596 4660825600
12234567638695 7873 223 363 797763 36 46364 2
!"# $%&'($# $)*"&' + ,# '%
-./0123/450670489::08;3/79:0<97=;>:/40?>:9@4A0B/:C4<9740/=>D0E.;>.097/0B6F3C0630=./0
7G 963G345637 3 73 5 9538 H63I345639H3G 935 J639863 436K 283
E;=.0=./07/4/7L6;70MF;C40CF7;3N0N/6:6N;>9:0=;8/O00P68/=;8/4A0=./0>63C;=;6340E/7/04F>.0
=.9=0=./023/407/89;3/C0940C;4>7/=/0<97=;>:/4079=./70=.930Q/;3N0>/8/3=/C0=6N/=./7093C0=60
=./0493C0N79;3076>R0B9Q7;>O00-./0MF;C40;30=./0<67/04<9>/40CF7;3N0=./4/0N/6:6N;>9:0=;8/0
</7;6C4097/0/;=./704=9=;>0670M6E09=0L/7@04:6E0L/:6>;=;/4O00S6E/L/7A0=.;40/TF;:;Q7;F80
4=9=/0;40C;4=F7Q/C063>/090E/::0;40C7;::/C093C0<:9>/C0630<76CF>=;630U0.;N.0MF;C0M6E0
L/:6>;=;/4097/0/V</7;/3>/C0;30=./03/970E/::Q67/07/N;63O00W=0E9406Q4/7L/C0;30=./02/:C0
45 4X3I 492 2Y3G3 6227329 463 3Y X3 9 72 46376 86 47X3 3 8 493
694 3 3 6223I94J 4Y3923993 56 34563 6223I94 3 463Z36 3
I3 46[3 73 96 76343K 9H2Y
oahg_fghi gjk]^ld
\]^_`^ab_\c^d_e_fghi gjk]^]ld_mgi j]cn_oacnljcl
\]^_v_bjlgh_s^ab
mgu|pgu_gssgpl]tg_}ghi gjk]^]ld
u|g_la_hg^jl]tg_}ghi gjk]^]ld_gssgplny
ohl]pj^_hjlg_bq]pq_
lh]rrghn_s]cgn_i atgi gcl
sah_pai k]cgu_a]^_v_bjlgh_s^ab
`^ab_wg^ap]ld
3333333333333333333333
~I686 473 6639 6343 345632 4Y3434Y3436I96345 73I56 86 3
`]r_xyz{ 63 2
139637 8I263G834563676J 3 737€6946343 96 7 3 23 398
E9=/70M6EO00W=0E9406Q4/7L/C0=.9=0=./0</78/9Q;:;=@0=606;:0M6E07/89;3/C0>634=93=A0/L/30
9=07/94639Q:/0M6E0L/:6>;=;/4A0E.;:/0=./0</78/9Q;:;=@0Q/N930=60Q/07/CF>/C09=090>7;=;>9:0
M6E0L/:6>;=@0E./30=./0>67/0E9404FQ‚/>=/C0=6090>68Q;3/C06;:ƒE9=/70M6E07/N;8/0?2NF7/0
„ DO00†F7=./70/V</7;8/3=40E/7/0</7B678/C0;30E.;>.0=./09>=F9:0M6E0=.76FN.0=./0<67/0
4<9>/40>6F:C0Q/06Q4/7L/C0L;4F9::@0U02NF7/0„‡904.6E40=.9=0=./0489::A0:664/0<97=;>:/40
<7/4/3=0630=./0493C0N79;304F7B9>/07/89;30;30<:9>/0CF7;3N06;:0M6EO00S6E/L/7A0=./4/0
23/40Q/>68/086Q;:/063>/0=./0E/==;3N0<.94/0?E9=/7A0;30=.;40>94/D0;40M6E;3N09=09079=/0
N7/9=/70=.930=.;40>7;=;>9:0L9:F/0?2NF7/0„‡QDO00ˆ 6Q;:;49=;6306B0=./023/409::6E40=./80=60
8J63434563I6345 473 566X3G395 963 94 467345 43 3 863G3456763I 492673
63I676 434345637 863486X329H 639 39933‰5 7399 473G34563I686 24Y3
3 6223Š94J 4Y3‹ 63694 7376J6373I94 34863 96 7673 34563
:9Q679=67@0>67/0=/4=4093C0=./02/:C08/94F7/8/3=4O
00
ŒŽ‘’“”
•––‘—˜’™–’š™›’œ‘™—˜ž’™Ÿ’
¡¢£¡¤¥¦§¦¨©
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
IJKKLMNOP LMNQ
RST T JUGHNV
FGH
XHJZGK[OFGH
ÐÑÒÓÔÕÖ×ØÙ
Ú ÙÛÕÔÖÜÕÛÖÝÞÕßÖÙÔÕÖ
ÑààáâÑãÕÖäÓÔÑÞÒÖáÑãÖ
åÔáäÓæÛÑáÞÖáÞãç
_LK`OaQLGK
9 ''!: $/ 0 &1%
<*7 &='#%!. &%= #13
>1,&?'!./ 0 &#/1,!&17!@&
B
ST T JUGHNOP LMNQOWOP NMOXGKNY
EEEEEEEEEEEEEEEEEEEEE
#'!&17!
: /%,,#13
XG[O\]^\L C &,!
D
bcdefghij klmn
ÐÑÒÓÔÕÖ×Øâ
Ú ÙÛÕÔÖÜÕÛÖÝÞÕßÖâÕæáàÕÖ
àáâÑãÕÖÜÑÛèÖÜÙÛÕÔÖ
åÔáäÓæÛÑáÞéÖãÕÙäÑÞÒÖÛáÖåáÔÕÖ
ÛèÔáÙÛÖâãáæêÙÒÕë
EEEEEEEEEEEEEEEEEEEEEEEEEEE
rstuvwxy
zvs
okgpiqnkfg
9 & !<&17'#13
‹Œ Œ d~fcm€i
j klmnij mlibfgm
bcdefghidfc
bcdefghij klmn
#'!&17!C &,!.#'7! + &#/1,
 d~fcm€ij klmnij mlibfgmi‚eƒf„ƒi ~m† mglc‡icmkpildiˆdnmilƒndkli~cd„‰fghŠ
Ž‘E’“”’•EŽ‘E–“•—“˜˜™Eš›œ
bfhi{|}{~
žŸ ¡¢£¤¥£¦§¨¦ž¢£ ¨£© £¨¦ª§¥ ¢«¦¬¥ ­£¥®¯§¥®¥£¦¢£ ¦¬¥ ®¨£© °§¨ž£ ®ª§«¨¤¥®¨£©
E ±Ž•²ŽE•²E³“—‘
žžŸ ´µ¢¶ ·¥µ¢¤ž¦¸ ¢«¦¬¥ ¹ªž© ¶¥¦¦ž£° ¯¬¨®¥
žžžŸ º ¥¦¦¨»žµž¦¸ ®¦¨¦¥ ¢«¦¬¥ ¼­£¥®½
¾’™Ž¿˜E’¿‘À—‘E±²Ž–²E²¿”EšEÁ“À™E•“EšEÁÁ–•Ž”E•“E—™À–E•²‘EÁÁ–•‘E
Ž–˜À™œ
ŽÃE ē•—“˜˜™ÅE‘˜“±Eš¿EÀ³E“ÁE•²E±˜˜ÆEEDzŽ‘EŽ‘Eš–¿À‘E‘Ž’Ž˜¿—EÈ³—Ž’•‘E•“E
¦¬¢®¥ ©¥®¤§ž»¥© ¨»¢·¥ ¬¨·¥ ®¬¢¶£ ¦¬¨¦¨ ®É ¨µµ¥§¤¬¨£°¥ ž£ ¹¢¶ ·¥µ¢¤ž¦¸E
EE —™À–‘E•²EÀ’š—E“ÁE³¿—•Ž–˜‘E•²¿•E¿—E’“šŽ˜Ž‘™E¿•E¿›E“E•Ž’ÅE˜¿™ŽÊE
EE •“E¿E˜“±—E–²¿–E“ÁEš˜“–˿ʐE¿•E•²E³“—E•²—“¿•‘ÆEE̒¿˜˜EŽ–—¿‘‘EŽE
¯§¢©ª¤¦ž¢£ ¨µµ¢¶ ¦¬¥ ­£¥® ¦¢ »¥ ¼¤µ¥¨£¥© ¢ª¦½ ¢«¦¬¥ «¢§É ¨¦ž¢£ ¨¦µ¢¶E
EE —¿•‘E¿™E–“–•—¿•Ž“‘ÃÍE‘“E•²¿•E•²E™‘Ž—™E±˜˜E³—“™À–•Ž“EE—¿•E¿™E
EE ™—¿±™“±ÃE–¿EšE¿–²Ž”™E±²Ž˜E’ ¿Ž•¿ŽŽÊE•²E¿—α˜˜š“—E—“–ËE
EE ³—’¿šŽ˜Ž•›Eš›E³—”•ŽÊE³“—E•²—“¿•Eš—Ž™ÊŽÊÆ
ŽŽÃE Ž‘E–“•—“˜E•—¿•’ •‘Eš¿‘™E“E–“‘“˜Ž™¿•ŽÊE•²E‘’ ¿˜˜E³¿—•Ž–˜‘E•“E
EE •²E—‘—”“Ž—E—“–ËE±Ž•²E–“‘ÏÀ•E—™À–•Ž“EŽEÁ“—’¿•Ž“E³—’¿šŽ˜Ž•›ÃE“—E
EE •²EÀ‘E“ÁE‘À—Á¿–•¿•‘E•“E–²¿ÊE•²EÁ“—’¿•Ž“E±••¿šŽ˜Ž•›
01234567 28609 0 2659 2 7 0882258002596 4660825600
2345657895 575 5934999755 96599 59559 8939
!"#$"#% &'!%$"(%%!)$*$%+%$*(%!"*%$",-".'/'"%
0%% 1(&$*2
345464789:;<=>79?7@9;A BCD9E7FBA BG=7H7I9A JK=CD9E7BEL7M 9;N9O=;7@K:DL>
P 1$0%( $ % &'!%$"$QR00$"#% $#$"!*'R*%$"!"0
S5T5 9437U365793 573 9 55939S5935 975V9395 75995V9759
3 S59W5W939 397594 959373 W9XY 93 S V93 5 949
735957579Z953 9[W\W]W]^W
_559V9973SSV93 593959 39756595 9 5 593 9 9755763 7W9
`% (% ($0%!"#&!%$"($% $*-*0)$"aP $$$**%!%0$"-#
bc( % R'!)$*$%+0#!0!%$"&!'%$%*!+"%!$"$"#!"0%"
0%R!!#&!$&)$"$ ("aP -# (% !%d
e!f gh (%!''"$' *$0eij [k*f)$"1(#$/% $# %R'!)$*$%+
9 5755 e)f k !"#$"#%!lh (%0$' *$0ei!k*f)$"#$/!00R'!)$*$%+
X^9 Y 9 57 5S V9753 995 5945 9593 937 59
""%!%$"$00%gh (%i!k*
e0f m e0$%$**0f(!%1(#!/% *(%R'!)$*$%+&!**% !)/
ef m*(&!-"!*1 &gh (%ij [n8959S7 599o6595937o 89 o59
99 575S V89 9 S59397536579 937o 9X o57^9 575S VW9Y59
99 373 9o595939595939 3 S59S7 59935 9 57 5 xy
–ˆ—
xŽ s{‘’
|Ž sˆ‘’
xŽ sˆ‘’
p€“”s• ˆŠ€
~€ ˆ‰qƒqŠ‹Œs 
–‰—
–˜—
y
y
vyy
xyy
zyy
{yy
~€s‚ƒ„ spƒ†‡
–‡—
–—
|yy
}yy
999999999999
Y 9 575S V9753 9959SV959S37 3 939V579394579S5455 9
59V9 55W99nV9759 7V9 5o65V97o589 993 5759SV9 9
37S593 9X3 89 39379 9 9 95^W99Y59V945
pq
rstuo9
vw
X S57939V579394579S37S5^993 7359SV9593 9S37S59 959
3 33 93959S7 5W99™ o759š]934993 95594593759 9
01
›œžŸ ¡¢£¡
¤¥¦œ§¨©¡¦ Ÿª ¨«œ©œ¬¥¡
­ Ÿ¨­¨¬œ®¯¡®°¡¨¡±ª §¬œ¬ ¡
§®Ÿ ¡²³ ¯¡´®²œ¯¡µ¨Ÿœ®ž±¡
¶·¸¹º»¼
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFGHIJKILMNHOPHQOEJRRLMNHIKJSRTHUFLKGHEFGHRUGKKLVWHXOPHJKKHEFPGGHIKJSRHLRHPGYMIGYHJEH
EFGHFLWFGPHZPLVGH[\J]K^HIOVIGVEPJELOVR_
9 ''!: $/ 0 &1%
\`_HaUGKKLVWHOXHEFGHIKJSRHVOEHOVKSHNGJVRHEFJEHEFGSH[QJPELJKKS^HZKOIbHEFGHQOPGHRQJIGcH
defgfhijgklmngkllnogfhigplkjgqkrfsplimgfngdriktguriigkfglnoirgvnogwilnpsfsimxglikysz{gfng
&='
#%!
. HL&%
#13KLES_
ZKOIbJWGHJEHEFGHQOPGHEFPOJER_HH`OEFHNGIFJVLRNR<*7
HKGJYHEOH
PGYMIE
LOVR
VHQGP=
NGJZL
|LWMPGH}~HLNQKLGRHEFJEHbGGQLVWHEFGHIKJSHLVHEFGHQOEJRRLMNHXOPNHLRHQPGXGPPGYHIONQJPGYHEOH
EFGHIJKILMNHOPHROYLMNH[EFGHUOPRE^HJPLGELGR_H€ JVSHEGRERHJVYHREMYLGRHOXHEFGHMVYGPKSLVWH
&?'
!./
0 f&
#/1,!
qhjmspmghkwigmhnozgfhkfgg‚ gofgqnfkmmseƒgphlnr>1,
syig„ †l
‡g{swim
gfhiglikm
gmoil
lsz{ˆgg‰&17!
fg @&
LRHEFMRHEFGHQPGXGPPGYHZPLVGHIONQORLELOVHXOPHVOVŠYJNJWLVWHUOPbOGPHJVYHIONQKGELOVH
vesymˆgg‹hsmgiuŒpkpjgnug †lgsmgyeigfngfhigmsignugfhigqnfkmmseƒgpkfsnzgohsphgkllnomgsfg
EOHGJRLKSHLVRGPEHLERGKXHLVEOHEFGHIKJSHKJEELIG_HHŽFGHJNNOVL
MNHIJE
LOVH
LRH:
OXHJH
RLNLKJPHRLGH
#
'
!
&17!
C
&,!
/%,,#
13
ghizpigsfmgkdslsfjgfng{swigfhighs{himfgqirƒikdslsfjgƒikmeriƒizfgszgŒ{erig‘ˆ
HHHHHHH
ǻ
’“”•–—˜™““š›œ
ž”•™Ÿ –¡¢–£ ”¤™Ÿ¥
D
9 & !<&17'#13
«¬
Ö×ØÙÚÛÜÝÞÜ
ßàÛáá×âØÜãÙÛÜäåÜàæäÛÚÜ
æçèåÚéä×åâÜ×èÜêåâäÚåááÛãÜ
çëÜêáæëÜäëéÛÜæâãÜçÚ×âÛÜ
êåìéåè×ä×åâ
B
—¡¯š´µ –—µ ™¶¤š¤™
·¡¤” š´µ –—µ ™¶¤š¤™
’”“¶š´µ –—µ ™¶¤š¤™
#'!&17!C &,!.#'7! + &#/1,
©¬
©ª
­¬
­ª
¬
ª
®š ¤š““™¯–£ ”¤™Ÿ
ª¨°± –²”’³
§¨¬± –²”’³
HHHHHHHHHH
¸NNOVLMNHIFKOPLYGHLRHMRGYHJRHJHVOVŠYJNJWLVWHZPLVGHXOPHRQGILJKHRLEMJELOVRTHRMIFHJRH
YMPLVWHJILYHRELNMKJELOVRTHUFGVHLERHRMZRELEMELOVHZSHQOEJRRLMNHIFKOPLYGHUOMKYHKGJYH¦šEOH
œ–§¨©ª
EFGHXOPNJELOVHOXHYJNJWLVWHQPGILQLEJEGR_
¹º»º¼º½¾¿ÀÁÂÃĽ¿Å½Æ¿ÁÇ ÈÉʿ˽ÌÈÇ ÈÍýν¾ÂÈÏÃ
aIJKGHPGXGPRHEOHEFGHQPGILQLEJELOVTHLVHZOEFHEFGHUGKKHLERGKXHOPHEFGHVGJPHUGKKZOPGHXOPNJELOVTH
OXHOPWJVLIHOPHLVOPWJVLIHNJEGPLJK_HHŽFLRHRIJKGHNJSHZKOIbHEFGHQOPGRHLVHEFGHXOPNJELOVH
nrgiwizgfhigqirunrkfsnzmg„riyepsz{goillgszvno‡gnrgdlnptgfhigfedsz{g„riyepsz{goillg
nefvno‡gOPHLNQGYGHEFGHOQGPJELOVHOXHUGKKHJIIGRROPLGRHG_W_HQPGGVEHEFGHOQGPJELOVHOXH
EFGHaMPXJIGH]OVEPOKKGYHaMZHaMPXJIGHaJXGESHÐJKGH[a]aaaÐ^_HHÑVHEFGHKJEEGPHIJRGHEFGH
RJXGHOQGPJELOVHOXHEFGHUGKKHUOMKYHZGHIONQPONLRGY_HH|LWMPGH}ÒHRFOURHJHEMZLVWHUFLIFTH
nzgripnwirjgurnƒgfhigqrnyepsz{goillÓgokmgunezygfnghkwigfhigkrikgkwkslkdligfngvnog
riyepiygdjgÔՂ gyeigfngkzgsznr{kzspgmpkliˆ
01234567 28609 0 2659 2 7 0882258002596 4660825600
4 9
9
222222222222222222222222222222222222222
34567
89 97 75
7 599 7
6
2!!2" #$#2%&$2'($")')% %)*2)#2%()++$($,2-.22"&$/)" !2)#% -)!)%.2%& %2& #2-$$2"($ %$,2
0123245367829:29582:;<7<63=28>?<=<0;<?@24:6A<9<:6B2345<8C8A20129582D:;@39<:62E?<A2
*F$(2+$*!*+)" !2%)/$#G2H&)#2" 2-$2,I$2%*J
9 K<L2 M2A84;83B82<6298@N8;39?;82A?;<672E:O2?N295829?0<67
K<<L2 M2A84;83B82<62N;8BB?;82A?;<6729582;8B8;C:<;2A8N=89<:62:;2E:O2?N2<6295829?0<67
K<<<L2P <Q<672O<95232458@<43==12<64:@N39<0=82E?<A28R7R24:663982O398;2O<952<6S849<:62
22 T %$(2*(2T %$(2U(*/2 2#$' ( %$2($#$(F*)(
HT*2%.'$#2*U2*(+ )"2#" !$#2V2T W2,2#'& !%$$#2V2($2$"*I%$($,2T&)!$22/I!%)%I,$2
*U2)*(+ )"2#" !$#2& F$2-$$2*-#$(F$,G22
XYZY[Y\Y]^_`abcd]_e]f_ag hij_k]lhg hmc]n]ok_amhkjb]^bhpcd
q2T),$2( +$2*U2)*(+ )"2#" !$#2& F$2-$$2$"*I%$($,2,I()+2T*(!,2T),$2'(*,I"%)*2
*'$( %)*#G22H&$#$2)"!I,$J
K<L2 r;84<N<939<:62:D2s3t=2B3=92A?829:24::=<672:D2O8==2E?<AB236Au:;28C3N:;39<:62:D29582
22 T %$(2)%*2%&$2+ #2'& #$2,I()+2'(*,I"%)*2*U2# %I( %$,2-()$#2%*2%&$2#I(U "$G
2 v w!2x#*!y2z 2v w!{2x#*!),y
w !")I/2#I!'& %$2" 2-$2'($")')% %$,2,I$2%*2"**!)+G2H&)#2)#2,I$2%*2%&$2U "%2%& %2-$!*T2
2"()%)" !2%$/'$( %I($2%&$2!$##2#*!I-!$2)#*/*('&|2&.,()%$|2-$"*/$#2%&$2$}I)!)-()I/2
U*(/2%& %2)#2'($")')% %$,G
2 w ~~2~2€ VV2z 2w € {2x#*!),y
x))y2 ‚($")')% %)*2*U2" !")I/2" (-* %$2,I$2%*2'($##I($2($,I"%)*ƒ2%&*I+&2*"$2
22 9582E?<AB23;82<6295829?0<67295812?6A8;7:2324:@0<68A28DD8492:D2N;8BB?;8236A2
22 %$/'$( %I($2($,I"%)*G
2 w x„w€ y†2z 2w w€ {2~2w€ †‡2~2„ †€
01
ˆ‰Š‹ŒŽŽ
‘’“”•‰–ŠŽ“—Ž˜‹™‹’šŒŽ›“œŽ
šŒšŽ‹Ž˜“Žž”š’Ž”“–˜Œ“’Ž
ŸŒ”‰Ÿ‰˜š˜‰“–
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFFFGHIJKLKMNHLOMNPLHOMJHQPHPJORKJSPTPUHPVWVHLKNXYKTHXTPOFXFSMSFRJHZTR[H\PT]HLRKTH
HH WMLH^PNNLHYMLHQPPJHRQLPT\PUHFJHLP\PTMNHXT
MTPML
9RUKOF
''JWH
!:
$/VHH 0 &1%
_`VHaTPLLKTPHSTFWWPTPUHLOMNPHZRT[MSFRJHFLHLPNZHMWWTM\MSFJWHFVPVHMJHPbSTMHXTPLLKTPHUTRXH
FLHOTPMSPUHMSHSYPHRJLPSHRZHLOMNPHZRT[MSFRJVHHcYFLHNPMULHSRHMHZKTSYPTHUTRXHFJHXTPLLKTPHMLH
<*7
&='
#%!
. LH&%
= HXT
#13
SYPH^PNNHRXPTMSRTHSTFPLHSRH[MFJSMFJHSYPHSMTWPSH^PN
NHXTRUKOS
FRJV
HHcYF
ZKTSYPT
PLLKTPH
UTRXHOTPMSPLHPbSTMHLOMNFJWHXTRQNP[LV
EF\GHd FbFJWHRZHS^RHTPLPT\RFTHQTFJPLHEJRT[MNN]HFJHSYPHXTRUKOSFRJHSKQFJWGHRTHQPS^PPJH
>1,
&?'
!{ul
./xeo0hpqp
&
#/1,!
ee feghihgjklgemglnheohpqpeglrselnetfgluvwefnxe
fnelny
hrzhxe
eihfe
|fzhg&17!
wH @&
HH OMJHNPMUHSRHXTPOFXFSMSFRJV
B
H `M}}eoghihgjklge{ulxwe~e€ ‚‚EFJƒPOSPUHLPMH^MS
H`M †C‡H
ELRN:
FUG/%,,#13
PT#G'H„!&17!
&,!
D
†SYPTHFJLRNKQNPH[MSPTFMNLHPVWVH T † HMTPHZRT[PUHQ]HMHLF[FNMTHXTROPLLVHH KOYHLOMNPLHMTPH
ZTPˆKPJSN]HZRKJUHSRHQPHORJSM[FJMSPUHQ]HNR^HNP\PNHTMUFRMOSF\FS]VHcYFLHURPLHJRSHJRT[MNN]H
9 PH
&
#13
TPXTPLPJSHMHYPMNSYHYM‰MTUHKJNPLLHMHUT]HLOMNPHLM[XN
FLH
UTF!
NN<&17'
PUHEKJXTRS
POSPUHQTPMSYFJWH
RZHSYPHUKLSHFLHYM‰MTURKLHUKPHSRHTMUFRMOSF\PHXMTSFONPLHNRUWFJWHFJHSYPHNKJWLGV
Š PNNHXTRUKOSF\FS]HOMJHRJN]HQPHTPPLSMQNFLYPUHQ]HTP[R\FJWHSYPHLOMNPH[MSPTFMNVHcYFLHFLH
#'!&17!C &,!.#'7! + &#/1,
MOYFP\PUHQ]‹
EFGH ŒŽŽ‘’“HSYPHLOMNPHZTR[HFSLHXTPOFXFSMSFRJHXRFJSHEFJHSYPHZRT[MSFRJ”H^PNNHRTH
HH ZMOFNFSFPLGHKLFJWHMHLKFSMQNPHLRN\PJS”HPVW‹
H
EMGH _M•NHLMNSHOMJHQPHUFLLRN\PUHQ]H^MSPT
H
EQGH •MNOFK[H•MTQRJMSPHE•M•† –GHOMJHQPHUFLLRN\PUHQ]HY]UTROYNRTFOHMOFU
H EOGH `MTFK[H KNXYMSPHE`M † weliejhg—exl˜™rušzezkexliikšjhpe›kvvhgrlfšš—H
HH H M\MFNMQNPHLRN\PJSLHSPJUHSRHQPHPbXPJLF\PHMJUHOMJHRJN]HUFLLRN\PHMHNF[FSPUH
H H M[RKJSHRZH[MSPTFMNV
EFFGH Œœ’“ekuzezshexl˜™rušzezkexliikšjhžeirfšhierkuŸšhxe|lzs el˜enhrhiifg—H
HH TPXPTZRTMSFJWHSRHQ]XMLLHXPTZRTMSFRJLHQNRO¡PUHQ]HLOMNP¢HYMLHQPPJHZRKJUHSRHQPHPZZPOSF\PV
EFFFGHcYPHXTPZPTTPUHSPOYJFˆKPHFLHSRH£¤¥£¦£§¨SYPHZRT[MSFRJHRZHSYPHLOMNPHQ]HOFTOKNMSFRJH
HH EFJHSYPH^PNNGHRTHFJƒPOSFRJHFJSRHSYPHZRT[MSFRJHRZHMHLOMNPHFJYFQFSRTVHH©JHSYPHNMSSPTH
HH OMLPHSYPHFJYFQFSRTHMQLRTQLHRJSRHSYPHZRT[MSFRJHLR[PHUFLSMJOPHZTR[HSYPH^PNNQRTPH
ee fnxelieŸgkxurhxeišk|š—e|lzsezsheŸgkxurhxe{ulxiekjhgefeŸhglkxek˜evfn—evknzsiH
HH ‚HXTRSPOSFJWHSYPHZRT[MSFRJ”H^ PNNHMJUHZMOFNFSFPLHZTR[HUM[MWPHQ]HLOMNPHXTPOFXFSMSFRJVH
H cYPHFJYFQFSRTH^RT¡LHQ]HXTP\PJSFJWHSYPHFJFSFMNN]HZRT[PU”HL[MNNHLOMNPHXMTSFONPLH
HH EJKONPFGHZTR[HMWWNR[PTMSFJWHMJUHZRT[FJWH[MLLF\PHLOMNPHOT]LSMNLHFVPVHSYPH[FJKSPH
HH XMTSFONPLHRZHSYPHLOMNFJWH[FJPTMNHMTPHTPSMFJPUHFJHLKLXPJLFRJVHH©SHLYRKNUHQPH
HH P[XYMLFLPUHSYMSHSYPHFJYFQFSRTHFLHLNR^FJWHUR^JHSYPH¡FJPSFOLHRZHLOMNPHMWWNR[PTMSFRJ”H
HH ¤ª§HSYPHSYPT[RU]JM[FOLHRZHLOMNPHZRT[MSFRJV
01234567 28609 0 2659 2 7 0882258002596 4660825600
2343536789
7979 97 77 78 7 77
!"#$%&'(#)*+,#-*++#.)%/##,)+*'#0%($*0*1 1(#-2(!#12("#%(#$))+('3##42*,#,)+*'5#6!)-!#
,#- 75#8 %*(,#*!#.)%/#.%)/# #,).1#1)# #9%*11+(#,)+*'3##:1#*,#/ '(#&0#).#+)!;5#+*!( %5#
<=>?@=>ABCD=EFGH=>FIJCHKB@LI=@MLNCOLEAI?EA<CPF>HC=CIH=FNCEANQ>HCLRCSTC>LCUVCI=@MLNC
1)/,3##42(#W,)+*'X#- 7#*,#'*,,)+8('#*!#12(#$%&'(#)*+# 1#%(,(%8)*%#1(/0(% 1&%(,# !'#
.)%/,# #$%",1 ++*!(#0%($*0*1 1(#-2(!#12(#1(/0(% 1&%(#%('&$(,#9(+)-#12(#$+)&'#0)*!1#
D>HAC>AOGA@=>?@AC=>CPHFIHC>HACY@<>CDOFI@LJC<AABCI@K<>=E<C=GGA=@JZCC[HAC>AOGA@=>?@AC
'*..(%(!$(#9(1-((!#12(#%(,(%8)*%#1(/0(% 1&%(#!'#12(#$+)&'#0)*!1#% !;(,#.%)/#)!+"##
.(-#'(;%((,#$(!1*;% '(#1)#/ !"#1(!,#).#'(;%((,3##\%(,,&%(#$2 !;(,#)!+"#2 8(##/*!)%#
(..($1#)!#12(#8 +&(#).#12(#$+)&'#0)*!1#1(/0(% 1&%(3
42(#/)&!1#).#- 7#'*,,)+8('#*!#12(#$%&'(#)*+#*,#+,)#2*;2+"#8 %* 9+(#]#.%)/#+(,,#12 !#
S^ CP>C>LC<?IHCHFQHC_=E?A<CDI=NCMAC=ML_A`CaV^ CP>JC>H=>C>HACILOGEA>ACI@?BAC<=OGEAC
1&%!,#*!1)# #,(/*],)+*'#&!0&/0 9+(#/ ,,3
42(#- 7#*,#!)%/ ++"#6(01#*!#,)+&1*)!#*!#12(#$%&'(#)*+#9"#(!,&%*!;#12 1#12(#1(/0(% 1&%(#
*,#/ *!1 *!('# 9)8(#12(#$+)&'#0)*!1# 1# ++#1*/(,3##42*,#*!8)+8(,#12(#*!,&+ 1*)!#).#
bLPCEFNA<`C>=Nc<`CA>IC=<CPAEEC=<C>HAC?<ACLRCHA=>A@<CPHA@AC=GG@LG@F=>AZCCdA=>ABCI@?BAC
)*+#1 !6(%,# %(#&,('#.)%#,( #)%#%) '#1% !,0)%13##42(#% 1(#).# ;;+)/(% 1*)!#).#12(#- 7#
$%",1 +,#$ !#+,)#9(#%('&$('#9"#12(#''*1*)!#).#!#*!2*9*1)%#]#)!$(#; *!#6*!(1*$,#).#12(#
G@AIFGF>=>FLNCG@LIA<<CI=NCMACFNb?ANIABeCNL>C>HAC>HA@OLBKN=OFI<ZCCf@LGA@COLBAEEFNQC
).#12(#1(/0(% 1&%(#'*,1%*9&1*)!#*!#12(#-(++,# !'#. $*+*1*(,#!((',#1)#9(#0(%.)%/('#1)#
(!,&%(#12 1#12*,#/*!*/&/#1(/0(% 1&%(#*,#/ *!1 *!('# 1# ++#1*/(,3##
42(#/ *!1(! !$(#).#12*,#/*!*/&/#1(/0(% 1&%(#*,#)98*)&,+"#( ,*(%#-2*+(#12(#-(++,#
=@ACbLPFNQCgC<FNIACHA=>CF<CMAFNQCILN>FN?=EEKC<?GGEFABC>LC>HAC<K<>AOZCDhLNQC>A@OJC<H?>C
BLPN<C@AG@A<AN>C=CG@LMEAOCgCA<GAIF=EEKCRL@CELNQCbLPCEFNA<CL@CGFGAEFNA<ZCC[HACG@LMEAOC
*,#$$(!1& 1('#.)%#,&9,( #-(++,5#,*!$(#- 1(%#*,#!#(..($1*8(#/('*&/#.)%#2( 1#1% !,.(%3#
[HACELNQCbLPCEFNA<C=NBCELPCP=>A@C>AOGA@=>?@A<CDijklJC=<<LIF=>ABCPF>HCOLBA@NC
, 1(++*1(#'(8(+)0/(!1,#*!#'((0#- 1(%,5#%(0%(,(!1# #$2 ++(!;(#12 1#*,#$&%%(!1+"#9(*!;#
%(,( %$2('#(3;3#'(8(+)0/(!1#).#/)%(#(..($1*8(#*!,&+ 1*)!#/ 1(%* +,3##:!#12(#/( !1*/(5#
,&*1 9+(#)0(% 1*!;#0%)$('&%(,#!(('#1)#9(#'(8(+)0('#1)#(!,&%(#12 1#12(#+*!(#')(,#!)1#
9($)/(#$)/0+(1(+"#9+)$6('#!'#2 ,#1)#9(#9 !')!('#W(3;3#0&%;(#12(#+*!(#).#$%&'(#)*+#
0%*)%#1)#,2&1')-!X3##42*,#1"0(#).#0*0(+*!(#9+)$6 ;(#2 ,# +%( '"#2 00(!('#)!#/)%(#
>H=NCLNACLII=<FLNZCd=@BEKC=NKCLG>FLN<CAmF<>CLNIAC>HACG@A<<?@AC@An?F@ABC>LC@A<>=@>CbLPC
%(#;%( 1(%#12 !#12(# 8 *+ 9+(#0&/0,#$ !#;(!(% 1(#)%5#/)%(#0 %1*$&+ %+"5#(7$((',#12(#
0*0(+*!(#1(,1#0%(,,&%(3#:/0%)8('##W0*0(]*!]0*0(X#*!,&+ 1*)!#,",1(/,# ,#-(++# ,#12(#
HA=>FNQCLRCbLPCEFNA<CMKC=CP=>A@Co=IcA>CL@CAEAI>@FI=EC>@=IFNQC=@ACMAFNQCBA_AELGABZCp =mC
PFEEC>KGFI=EEKC<>=@>C>LCRL@OCLNC>HACFNNA@C<?@R=IACLRC>HAC>?MFNQCL@C>HACbLPCEFNAeC<FNIAC
12*,#*,#12(#$)+'(,1#0)*!13##q!$(#.)%/('5#12(#- 7#$ !#9(#%(/)8('#9"r
W*X# ($2 !*$ +# $1*)!#W,$% 0*!;X
W**X# s*,,)+8*!;#*!# ]#
# ##
#
]#
# ##
#
#
01
2)1#$%&'(#)*+#W$2( 05#%( '*+"# 8 *+ 9+(X
,)+8(!1,#W/)%(#(70(!,*8(5#9&1#$ !#).#'*,,)+8(#2*;2(%#
- 7#$)!$(!1%1*)!3# "#9(#2"'%)$ %9)!#)%#- 1(%#9 ,('3X
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFFGHIJKLMNOPQPJ
RSTUVWXTUYZ[\SW]^TVT[_U`TaT`[U^bUWVUWXb`\Sbc[
ef^'
cXf
Vbc[
dU[b]
XW^T`fW]UhSfiSU
9dU'
!:
$/
0fgU&1%
\`Tif\f^W^T[Ua`bXU[bXTUi`cgT[jUk^Ua`TlcTV^]mUibV^WfV[UWUSfnSUXc]^foW]TV^UXT^W]U
ibViTV^`W^fbVdUUU[ciSUW[UpWVWgfcXjUUk^Uf[UXWgTUc\UbaUWUibX\]TqUXfq^c`TUbaUW[\SW]^TVT[dU
`T[fV[UWVgUXW]^TVT[UhSfiSUhT`TUb`fnfVW]]mU\`T[TV^UfVU^STUi`cgTUbf]UcVgT`Ub`fnfVW]U
`T[T`obf`UibVgf^fbV[UW[UWUXT^W[^We]TUib]]bfgW]Ugf[<*7
\T`[fbVj
UURSTU
\`Tif
^W^fbVU
&='
#%!
.\f&%
=\`#biT[
13[Uf[U
rstuuvsvwxyzx{sv||}svxsvw}~rt€|xx‚|{ƒ‚„rv€vx{sv~t{tr‚rt€xt|x rv€x†s|rxy|vs‡vwx€v‚sx
^STUecee]TU\bfV^UfjTjU^STUiSWVnTUfVUi`cgTUbf]UibX\b[f^fbVUgcTU^bU^STU`TXboW]UbaU[bXTU
baU^STU]bhT`UXb]Tic]W`UhTfnS^U[\TifT[Ua`bXU^STUi`cgTUbf]UgT[^Wef]f[T[U^STUib]]bfgW]U
>1,
!f./
gf[\T`[fbVU^SW^UXWfV^WfVTgU^STUW[\SW]^TVfiUXW^T`
fW]UfVU[&?'
c[\TV[
bVjUU0 &#/1,!&17!@&
B
ˆWeb`W^b`mUTq\T`fXTV^[UfVUh SfiSU^STUTaaTi^UbaU\`T[[c`TUbVU^STU[^Wef]f^mUbaUWVUW[\SW]^TVfiU
i`cgTUhW[U[^cgfTgUSWoTU[SbhVU^SW^U^ST`TUW`TU^hbU
bV[UbaC
U[^&,!
Wef]f^:
mjUU/%,,#
‰VTUbaU^ST[13
TU
#`'Tnf
!&17!
f[UVb`XW]]mUWeboTU^STUecee]TU\bfV^U\`T[[c`TUŠW[Ugf[ic[[TgUTW`]fT`‹UhSf]TU^ST`TUf[UWU
[TibVgU`TnfbVUbaU[^Wef]f^mUW^U]bhU\`T[[c`T[jUUZ[\SW]^TVT[UW`TU^Sc[UXWfV]mU\`Tif\f^W^TgU
W^UfV^T`XTgfW^TU\`T[[c`T[jUURSTUaWi^U^SW^U^ST`TUf[UWU[TibVgUW`TWUbaUY[b]cef]f^m_Uf[UbaU
]fXf^TgU\`Wi^fiW]Uc[TUeTiWc[TUW[\SW]^TVT[UgbUVb^9
UTW[
f]mU
[[b]oT_j#
U13
URSf[Uf[UeTiWc[TU
&
Y`Tgf
!<&17'
^STmUhT`TUb`fnfVW]]mU\`T[TV^UfVU^STUi`cgTUbf]UW[UWUib]]bfgW]Ugf[\T`[fbVdU`W^ST`U^SWVUfVU
[b]c^fbVUŠW[Uf[U^STUiW[TUbaUhWq‹j
D
#t'v€~vx
!&17!
C‚x&,!
.#'
w|+
Œƒvx„„t€ux|}ŽŽ‚st|v|x{vs‚rt€‚„x†v„wxv{vs
sŽx
€}Žyvs
x7!
x†v„
 &#/1,
Šf‹U RSTU[ToT`f^mUbaUb\T`W^fbVW]U\`be]TX[Uf[UVb^ÙT]W^TgU^bU^STUi`cgTUbf][UW[\SW]^TVTU
UU ibV^TV^jU‘`cgTUhf^SU`T]W^foT]mU]bhUW[\SW]^TVTUibV^TV^[UiWVUnfoTU^STUX b[^U
UU fV^`Wi^We]TU\`be]TX[
’tt“x ”t€~vx‚|{ƒ‚„rv€vx{sv~t{tr‚rt€x~~}s|x‚rx‚x|{v~t†~x{sv||}sv•xtrxt|x€„zxy|vs‡vwU
UU fVUW`TW[UbaU^STU\`bgci^fbVU[m[^TXUhST`TU^Sf[Ub`UWU]bhT`U\`T[[c`TUf[UTq\T`fTViTgU
UU TjnjUW[\SW]^TVTU\`Tif\f^W^fbVUXWmUeTUfVf^fW]]mUbe[T`oTgUfVU^STUaWif]f^fT[jUUZ[U
U ^STU`T[T`obf`UgT\]T^T[dU^STU]bhT[^U\bfV^UbaU\`Tif\f^W^fbVUhf]]UXboTU^bU^STU^b\UbaU
UU ^STU^cefVnUWVgU^STVUgbhVU^STUhT]]UcV^f]UW[\SW]^TVTU\`Tif\f^W^fbVUiWVUW][bUbiic`U
UU fVU^STU`T[T`obf`
’ttt“x–v|{trvxrƒt|•x†v„wxsv{sr|x xsŽ‚rt€xw‚Ž‚uvxw}vxrx‚|{ƒ‚„rv€vx{sv~t{tr‚rt€U
UU fVU^STU`T[T`obf`UW`TU`W`T
Šfo‹U—b]fgUW[\SW]^TVTUf[UVb`XW]]mU`TXboTgUXTiSWVfiW]]mUa`bXU^STUhT]]UWVgUaWif]f^fT[
Šo‹U ˜ TiSWVfiW]U`TXboW]Uf[UVTiT[[W`mUeTiWc[TUW[\SW]^TVT[UW`TU\bb`]mU[b]ce]TUfVUU
U [b]oTV^[
Šof‹UˆW`nTU[iW]TUW[\SW]^TVTU\`Tif\f^W^fbVUbiic`[UhSTVU^STUW[\SW]^TVfiUi`cgTUbf]Uf[U
UU ibV^Wi^TgUemUWifgU™Ueb^SUa`T[SUŠšc[^U\cX\TgUfV^bU^STUhT]]‹Ub`U[\TV^UWifgU
UU ŠWa^T`UibV^Wi^Uhf^SU^STUab`XW^fbV‹jUURSf[UTaaTi^Uf[U\W`^fic]W`]mU[ToT`TUhSTVU^STU
U
U WifgUibV^WfV[UaT``fiUiW^fbV[UŠ›Tœœœ‹UTjnjUa`bXU`TWi^fbVUbaU^STUWifgUhf^SU`c[^jU
UU RSTUfX\Wi^UbaU^Sf[UbVUWifgU[T]Ti^fbVUhf]]UeTUgf[ic[[TgU]W^T`j
žŸžŸžH ¡¢£¤PJH¡¥H¦¡£§ MOF¡QH¨M§ M©PHªH«M¤OP£FM
RSTU\`T[TViTUbaUeWi^T`fW]Uib]bVfT[UWVgdUXb`TUfX\b`^WV^]mU^STU`T[fgcT[U^STmU\`bgciTdU
iWVU]TWgU^bUWUVcXeT`UbaUfV^`Wi^We]TUb\T`W^fbVW]U\`be]TX[jUUZ]^SbcnSU^STmUW`TUXb[^U
01234567 28609 0 2659 2 7 0882258002596 4660825600
234564789 8487487499 7974 44 843 843 797
!"# $%!#&!!'#()*'+#,'#$--#."/!0#)(#1!--0#$'+#($2,-,.,!0#3#($2,-,.,!0#$4!#)/!'!+#.)#. !#
$.5)0/ !4!#()4#,'0/!2.,)'#$'+#&). #1!--0#$'+#($2,-,.,!0#$4!#$.#)22$0,)'$--"#!6/)0!+#
8 6 8 8 34674 237 84 632 486 49 634 7 94 8 78 77 87
$'+#!0.$&-,0 5!'.#)(#. !#&$2.!4,$-#2)-)'":## !#()--)1,';#0*55$4,0!0#0)5!#)(#. !#
5)0.#1,+!-"#()*'+#/4)&-!50<
=,># ?@ABCDE@FGHFEI CGJGACFFKF@DGDCI L@LBM#!.2#1,--#/ "0,2$--"#/-*;#. !#()45$.,)'#
## /)4!0#,(#,'N!2.!+#,'.)#. !#1!--:# !#0)-*.,)'#.)#/4!%!'.#. ,0#,0#.)#!'0*4!#. $.#
## . !#&$2.!4,$-#2)-)',!0#+)#').#&!2)5!#!0.$&-,0 !+#,'#. !#1!--0#$'+#($2,-,.,!0:#
O7 P 843 34 4787 9984 6 674 78 84 499 69 P4 834 84#
8 62Q47874783 87 2 P 8434 8 4R4 ST99U 94R49#
# V)1!%!4W#,.#)'-"#4!X*,4!0#)'!#&$2.!4,$-#0/)4!#.)#&!#5,00!+#1 ,2 #2$'#. !'#
8 38 74 97 Y
=,,># ZD[LG?@ABCDE@#$4!#$!4)&,2#&$2.!4,$#1 ,2 #2$'#!0.$&-,0 #. !50!-%!0#*'+!4#
## $'#,4)'#+!/)0,.#. $.#. !"#()45#. !50!-%!0:## !,4#;4)1. #/4)2!00!0#. !'#4!0*-.#
## ,'#. !#!0.$&-,0 5 !'.#)(#$'#$'$!4)&,2#&$2.!4,$-#2)-)'"#!:;:#)(#0*-/ $.!#
## 4!+*2,';#&$2.!4,$W#-!$+,';#.)#$#/,..,';#."/!#2)44)0,)':## !"#$4!#."/,2$--"#()*'+#
7 7487 974 434 84 6 7842Q4789 49 47 84 \4 #
34 2Q6988 T998 P 8434 74 84 34 38484 P 84 632 4#
## ,4)'#+!/)0,.:##]!;*-$4#5!2 $',2$-#4!5)%$-W#!:;:#&"#/,;;,';#. !#-,'!W#,0#. !#5)0.#
## !((!2.,%!#.4!$.5!'.:
=,,,>#HKF^_@BCG`CaKAELbG?@ABCDE@#$4!#$'$!4)&,2#&$2.!4,$#1 ,2 #$4!#2$/$&-!#)(#
## /4)+*2,';#-$4;!#X*$'.,.,!0#)(# "+4);!'#0*-/ ,+!#=V cd>#1 !'#&). #. !#
## 0*-/ $.!#$',)'#=*0*$--"#(4)5#0!$#1$.!4>#$'+#$#2$4&)'#0)*42!#=24*+!#),->#$4!#
## /4!0!'.#,'#. !#$&0!'2!#)(#)6";!':##V cd#/4!0!'.0#$#5$N)4#0$(!."#$'+#2)44)0,)'0#
## ,00*!#0,'2!<
=$># e6/)0*4!#.)#V cd#$.#$#2)'2!'.4$.,)'#)(#$#(!1#.!'0#)(#//5#,0#($.$-#.)# *5$'#-,(!
=&># f !--#$'+#($2,-,."#2)5/)'!'.0#5$+!#(4)5#2)'%!'.,)'$-#0.!!-#$--)"0#1,--#($,-#
47 49 4 8 6 8 9 38 934 634 2 cd:
g'N!2.,)'#1$.!40#2)'.$,',';#!%!'#-)1#-!%!-0#)(#. !#(!44,2#2$.,)'#=h!iii>#1,--#/4!2,/,.$.!#
. !#!6.4!5!-"#,'0)-*&-!#0)-,+#(!44,2#0*-/ $.!#=h!cdj>#1 !'#!6/)0!+#.)#V cd:#h!44,2#2$.,)'0#
$4!#')45$--"#()*'+#,'#,'N!2.,)'#1$.!4#=0,'2!#,.#,0#/4)2!00!+#,'#0.!!-#%!00!-0#$'+#)6";!'#
$0#&!!'#/4!0!'.#,'#. !#0"0.!5#!.2>:## !#0)-,+#h!cdj#1,--#/-*;#. !#()45$.,)':
klmlklGH[KDACMG[nGo[DI @BE[LGp@I @bCGqGrDCMMKDCG`CaKABE[L
!#4!0!4%),4#/4!00*4!#)(#. !#%$0.#5$N)4,."#)(# "+4)2$4&)'#4!0!4%),40#1,--#&!#4!+*2!+#
+*4,';#,.0#/4)+*2,';#-,(!.,5!:## ,0#2$'#.4,;;!4#$#'*5&!4#)(#)/!4$.,)'$-#/4)&-!50#
1 ,2 #4!0*-.#,'#()45$.,)'#+$5$;!:
=,># !#/4!2,/,.$.,)'#)(#,')4;$',2#$'+#$0/ $-.!',2#02$-!0#1$0#+,02*00!+#!$4-,!4:
=,,># !#4!+*2.,)'#,'#4!0!4%),4#/4!00*4!#,'24!$0!0#. !#()45$.,)'#!((!2.,%!#0.4!00:###
# !#,5/$2.#)(#. ,0#,'24!$0!+#0.4!00#2$'#-!$+#.)#()45$.,)'#($,-*4!#4!0*-.,';#,'<
01
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFGH IFJKHLMNKOPQRNJ
9 ''!: $/ 0 &1%
ESGH TNULFPQRNJHNVHQWXHMXYXMZNRMHLMNKOPRJ[HRJQXMZF\]H^WRYHMXYXMZNRMHKMRZXHU XPWFJRYUH
__ `a_bcdcef`gh_a`dfc_`i_j`hh_hcgk_il_gd_`dfmcgack_mcflncmo_lp_iqc_l`h_lm`r`dghhoH
HH RJHL\FPXHRJHQWXHMXYXMZNRMsHSOQHRQHUFtHF\YNHSMRJ[HXuQMFHNLXMFQRNJF\HPNYQYHKOXHQNH
HH YOMVFPXHYOSYRKXJPXHFJKHKFUF[XHQNHQWXHvX\
\YHQWXUY
X\ZXY
HwE.
PFY&%
RJ[xQOSR
SXRJ[H
<*7
&='
#%!
= J[H
#13
HH YyOXXzXKHRJQNHFJHNZF\HYWFLXHMFQWXMHQWFJHSXRJ[HMNOJK{HNMHXZXJELFMQRJ[HKOXHQNH
HH YWXFMRJ[G|]
>1,
&?'!G./
#/1,!
EPGH }FO\QHPMXFQRNJHNMHMXFPQRZFQRNJHwR]X]HEURJR
GHXFMQWyOF~XY
]HHTM0
XFQ&
RNJHNV
HXuQM&17!
FH @&
__ pghia_€go_hcgk_il_`€mlnck_`d‚lj_jq`hc_pghi_mcgfi`ngi`ld_€go_mcahi_`dH
HH LMXZRNOY\tHYXF\RJ[HVFO\QYHSXPNURJ[Hƒ\XF~t]„
B
'!&17!
13
EKGH XKOPXKHVNMUFQRNJHLXMUXFSR\RQt]HH^WRYH
RYH#
JNM
UF\\tHC
UR&,!
JNM{H:
SOQH/%,,#
PFQFYQMNLWR
PH
HH LNMXHPN\\FLYXHQNHzXMNHLXMUXFSR\RQtHWFYHSXXJHNSYXMZXKHVNMHYNUXHPWF\~YHvWXJHFH
HH PXMQFRJHMNP~HXVVXPQRZXHYQMXYYHRYHXuPXXKXK]
D
9 & !<&17'#13
†```‡_ˆcacmnl`m_€hi`qgac_‚lj_cppcfia_iqgi_€go_bc_pghacho_giim`bick_il_plm€gi`ld__
_ kg€grc‰__Š_apef`cdi_mckfi`ld_`d_mcaamc_al_iqgi_iqc_mcacmnl`m_‚`k_gaacaH
HH QWXHSOSS\XHNMHKXvHLNRJQHMXYO\QYHRJHQWXHLMXYXJPXHNVHFJHXuQMFHLWFYXHRJHQWXH
#'€cgb`
!&17!
+ &#/1,
__ mcacmnl`m_jq`fq_j`hh_mckfc_iqc_†mchgi`nc‡_cm
h`iC
o_i&,!
l_l`h_.#
lm_'
rga7!
_‚lj
‹ŒŒŽŒ‘’“”•–‘—˜‘“™ š›œ‘žš™ šŸ•›œ™ ’¡š›œ‘
¢ chh_ai`€hgi`ld_`a_mlbgbho_iqc_€lai_pmc£cdi_lcmgi`ld_jqcd_‚`ka_gmc_€ck_
`dil_g_jchh‰__¤qc_emai_iqmcc_pgfilma_k`afaack_bchlj_gmc_ldho_gh`fgbhc_il_iqc_
€`dr_lp_q`rq_mcgfi`n`io_‚`ka_afq_ga_gf`k‰__¥ljcncm¦_iqc_mc€g`d`dr_cppcfia_fgd_
bc_cdfldicmck_j`iq_gdo_‚`k_iqgi_`a_€ck_`dil_iqc_jchh¦_dhcaa_mlcm_iqlrqi_`a_
[RZXJHQNHRQYHYX\XPQRNJ]
ERGH XFPQRNJHLMNKOPQYH[XJXMFQXKHStHQWXHMXFPQRNJHSXQvXXJHQWXHRJ§XPQXKHFPRKHFJKH
HH QWXHVNMUFQRNJHMNP~HUFtHLMXPRLRQFQX{HPFOYRJ[HFHMXKOPXKHLXMUXFSR\RQtHEVNMUFQRNJHKFUF[XG
ERRGH ^WXHFPRKHUFtHvXF~XJHEKXPNJYN\RKFQXGHQWXHMNP~{HStHFQQFP~RJ[HQWXHRJQXM[MFRJH
HH PXUXJQHYNHQWFQHEJNMUF\\tHQXULNMFMtGHYFJKHLMNKOPQRNJHRYHNSYXMZXKHvWXJHQWXH
HH vX\\HRYHMXQOMJXKHQNHLMNKOPQRNJHRJPNJYRYQXJQ
†```‡_¤qc_gblnc_kcfldalh`kgi`ld_mlfcaa_€go_rcdcmgic_¨edca©_jq`fq_fgd_€`rmgic__
_ gdk_bhlfª_lmc_iqmlgia‰__¤q`a_mlfcaa_fgd_lffm_plm_liqcm_‚`ka_†acc_acfi`ld__
_ «‰¬‰«‰­_ld_‚`k®mlfª_fl€gi`b`h`io‡
ERZGH¯PRKHRYHNVQXJHRJPNULFQRS\XHvRQWHPMOKXHNR\H\XFKRJ[HQNHVNMUFQRNJHNVHFHYN\RKH
HH ƒY\OK[X„HvWRPWHPFJHS\NP~HLNMXYHEP]V]HYXPQRNJHKXYPMRSXKHQWXHLMXPRLRQFQRNJHNVH
HH FYLWF\QXJXYHStHFPRKGHNMHFHZRYPNOYHFPRKHxHNR\HXUO\YRNJHVNMUFQRNJ]
EZGH ¯HVOMQWXMHVNMUHNVHFPRKxPMOKXHNR\HRJPNULFQRSR\RQtHRYHQWXHVNMUFQRNJHNVHFHWR[W\tH
HH ZRYPNOYHvFQXMxNR\HXUO\YRNJ]HH^WRYHXUO\YRNJHYQFSR\RQtHRYHNVQXJHRJPMXFYXKHStHQWXHH
_ edc_gmi`fhca_kcafm`bck_`d_†```‡_gblnc
01234567 28609 0 2659 2 7 0882258002596 4660825600
234567896 9 6 9 96 4 6 9 446 6968 966466 9 6
!"#$%&' " ((#('!)*"(#+,#- (.'6
/0 1'.1!#-2+'#"3 '2!#"241!#"24"5#
23445678964 9 696 9 66 7 4 64 6896 9 6 9 96 9 69674 66896
28!29'#,250 1
78969 4396:9 9 44668968 6:8 6966 46969796684668486
9 66 7 4 ;66< 6=9 67:6496>?64;9;68964969@749669796896 9 6
6 7 4 66896: 9A647 4 63 79?66 866 6939 6 9 68 39699 669 396
6 6:9 9 446 4 6;
&').# 1 24#(('"+-(,""++"(!29'#! 1'50 1$
4;9;646687696 7 9666 4 6 96966 67 ;
BCDCECFGHIJKLMFHNFOHJP QRSHTFUQP QVLFWFX QRLJFYTZLKRSHT
< 96379666 9 6 964 [9964 646 969 3466 6: 9667 96 4 9 96
234 96 9: 99 56 664: 396896 939 6 66=6 99:4 >689646
9,'28!"'+10!"%&'!#+#! .9#228!2((2#4 2\(1
466 6]66^649668 66 6: 74 6 9;66789696 96 96379664:?6
29.",#((!2#, 2#2(("9'28!150 1_'##+ 1+(044 249
'+9#2"!#2!!0%&' ""`,+(\12\40aa'"#(28!2
9 668 39699 6:7:968 7866 96 689639 96:9 9 4469 67 9666
67 4 668964 [94 6379;66789696966668 68 b
2456 78967 9 9667 96 9 6:766896 96 66494 9;6678466466 66
6 67 : 464 664 96896 9 6 6639 67 4?694 6 c9 66 6967 $ d (#29' ""09#!# '#ef2, 2#(\(1!#""' "#9
6 4: 49 u”z{xkz“l„z{Œ zx•j~jy–l— zx˜ {zlƒ€{zl™ “
$g"9#22\2\(_hf_1!#""' "290'
666666666
tqq
sr
ulvlw xyz{l|}lij~yz{zly€lol
rq
‚lvlw xyz{l|}lij~yz{zly€lrl
ƒlvl„{€ †zlw xyz{l‡ˆy{zxyz
|lvl‡ˆy{zxyzl‰ {Šx†zlw xyz{
or
q
q
snr
tr
oonr
‹€~ Œ zlˆŽz†yzljy{z‘’z{Š“
pq
psnr
š›œžŸ ¡¡
¢£¤¥¦§Ÿ ¨©žŸ ªŸ«ªžŸ«§ª«ª©
ž¤¬­ ›¬®Ÿ¨ª›©¬ ¯¤ªŸž°¤§›ª±
01
ijklmnop
!"#"$"%&'"#()*+
,
3
7
:
-./ %".0 1%/ %2
4 %"5 !$"6"+"02
DEFGHGIJKLMMKNOGKLPEHGKHEMQRGSKETKF LNGIKUVWGXNGYKLIGKSRLMMKXERZLIGYKNEKNOESGKIG[QUIGYK
UVKLXNQLMKUVWGXNUEVKZIEWGXNS\K]VKZILXNUXG^
8 &&9#./ %0$
_U`K aOGKUVWGXNUEVKFLNGIKUSKETNGVKRQXOKXEMYGIKNOLVKNOGKIGSGIHEUI\KKaOUSKXEEMUVbKETKNOGK
KK UVWGXNUEVKUVNGIHLMKIGYQXGSKNOGKIGSGIHEUIKSNIGSSKUVKNOGKVGLIKFGMMPEIGKLIGLKLVYK
KK LMMEFSKTILXNQIGKXIGLNUEVKLNKLKRQXOKMEFGI
KZIGSSQI
GKNOLVK
FEQM
YKVEI
McKPGK
;)6
%<&
"$ -
%$
<RLM
"02
KK GdZGXNGYK_NOGIRLMKTILXNQIUVb`\KKaOUSKTILXNQIGKGSSGVNULMMcKUVXIGLSGSKNOGKTEIRLNUEVK
KK SQITLXGKLIGLKEZGVKNEKNOGKFGMMJKUVXIGLSUVbKNOGKFGMMKUVWGXNUHUNcKLVYKIGYQXUVbKUNSK
KK SGVSUNUHUNcKNEKTEIRLNUEVKYLRLbG
=0+%>& -./ %".0+ %06 ?%
effghfiijkljimnojkpfjqhkrmksltjrfuvsrfwftkijgfxftgjltj
A
_L`K aOGK[QLMUNcKETKNOGKLHLUMLPMGKFLNGIKSEQIXG "&%06 B %+ 9.$++"02
_P`K y OGNOGIKNOGIRLMKTILXNQIUVbKEXXQISKLVY
_X`K aOGK[QLMUNcKETKNOGKTEIRLNUEV
C
8 %;%06&"02
zkjiplvhgj{fjtlkfgjkpmkjkpfjqtfjqhkrmksltjfuvsxwftkjpmij{fftjrfwl|fgj}rlwjkpfjst~fksltj
€mkfrjkrfmkwftkjkrmstjl}jwlikjelrkpjfmjqfhgij{fmvifjl}jkpfj‚llgj€mkfrjuvmhsknjmtgj
TEI
"&RK
%06
BK%+
&
NOGKVLNQILMKNGVYGVXcKTEIKNOGKNLIbGNKTEIRLNUEVSKNEK
NOGIRLM
TILXN-"
QIGS
\6 *%".0+
ƒss„j pfj}lrwmksltjswxmsrwftkjgvfjkljilhsgijxrfiftkjstjkpfj}ffgj€mkfrjƒq‚vrfj††„jl}K
KK IGLMKUVWGXNUEVKFGMMSKUSKSQZZMGRGVNGYKPcKRLVcKGdNILKSEQIXGSKETKYLRLbUVbK
KK ZLINUXMGSKG\b\KXEIIESUEVKZIEYQXNSJKEUMJKPLXNGIULMKIGSUYQGSKLVYKENOGIKSEMUYSK_SQXOK
KK LSKSXLMGKZIGXUZUNLNGS`KZUX‡GYKQZKYQIUVbKUNSKZLSSLbGKNOIEQbOKNOGKTLXUMUNUGS\KKaOGK
KK MGHGMKETKXEIIESUEVKZIEYQXNSKLVYKPLXNGIULMKXEVNLRUVLNUEVKLIGKIGYQXGYKPcKNOGK
KK UVWGXNUEVKETKXEIIESUEVKUVOUPUNEIJKYGEdcbGVLNUEVKETKNOGKUVWGXNUEVKFLNGIKLVYKUNSK
KK XEVNUVQEQSKNIGLNRGVNKFUNOKPLXNGIUXUYG\
K
ˆQPSNUNQNUVbKZIEYQXGYKFLNGIKTEIKTIGSOKFLNGIKPIUVbSKLKVGFKILVbGKETKZIEPMGRSKYQGKNEK
NOGKZIGSGVXGKETK_MEFKMGHGMS`KETKEUMKLVYKSRLMMKSEMUYKZLINUXMGS\KKaOGK_IGMLNUHG`KFLNGIK
uvmhsknjmtj{fj~vg‚fgj}rlwjhstfj‰jstjq‚vrfj††Šjj‹vrkpfroj{fst‚jplkojkpfrwmhj}rmkvrst‚j
USKQVMU‡GMcKNEKEXXQI\
ŒŽ‘’ “”•–Ž—“’ “˜™Ž—š‘•–˜Ž› ‘œ™‘Žž™‘“”•–Ÿ
aOGKZIGHUEQSKSGXNUEVSKETKNOUSKXOLZNGIKOLHGKYUSXQSSGYKNOGKbGVGILMKXEVXGZNKETKTEIRLNUEVK
YLRLbGKLVYKOLSKbUHGVKSGHGILMKGdLRZMGSKFOGIGKUNKEXXQIS\KK VGKETKNOGKLIGLSKFOGIGK
TEIRLNUEVKYLRLbGKTIG[QGVNMcKEXXQISKYQIUVbKNOGKMUTGKETKLKZIEYQXUVbKFGMMKUSKYQIUVbK
FEI‡EHGIKEZGILNUEVS\KKaOUSKUSKYQGKNEKURZLUIRGVNKETKNOGKZIEYQXUVbKTEIRLNUEVKPcKSEMUYK
ZLINUXMGS\
¡tnjilhsgijxrfiftkjstjkpfj€lr¢l|frj£vsgoj€shhj{fjst~fkfgjstkljkpfjrfifr|lsrjgvrst‚j
€fhhj¢shhst‚jlxfrmksltiŠjj psijsijipl€tjstjq‚vrfj†¤j€pfrfjkpfj€fhhjxrlgvks|sknj€mij
IGYQXGYKPcKVGLIMcKLKNOUIYKFOGVKNOGKFGMMKFLSK‡UMMGYKFUNOKYIUMMUVbKRQY\
01234567 28609 0 2659 2 7 0882258002596 4660825600
!46"!6#$
6%&
' ((6)4((6' 4!*6+ 9
9
4564 6
6%&
6
22222222222
¡¢£¤¥¦§¨
©¤ª«£¬­¡ª®¦¯¡°­ª¤±¦°¯ª²°¦
³ª¤´µ­¡ª®¦«µ´µ¢¥¦¬¤¥µ­¥«¦
¶±¦²¥··¦¸¡··¡®¢
,-./.2/012324567281./295:251/02;.2<.=.657.32;:2362112=<28.9.=7>22/34
8656542
=<285/2=<>2
789
?@AA@BCDEFGHDDI@CJKEDLMDNOPQNDOPQDQPKRPSEKDTJ@UDAPNNEND@BFPDFOEDVPK
?WF
@PBDQEKED
6.1572X.1:280=/75=72Y062Z235:/2[-.=252458\.62[5/2;.2=<29211.3]22^=8.27-.2[06\0X.62
P_EKWF@PBD_KPCKENNEUDFPD?@AA@BCDPBDWD`NODFOWFDQWNDAPUCEUDWGKPNNDFOED_EKVPKWF@PBNaDFOED
APNNENDUEGKEWNEUDbcDMde DVPKDFOEDBEfFDFQPDUWcNgDUKP__@BCDS@KFJWAAcDFPDhEKPDWVFEKDUWcD
ij]22k722/2X.6:212\.1:27-5727-.2062<2=511:204.=24.6Y065720=/2-532;.809.280941.7.1:2
;108\.32;:27-.2.=320Y27-.204.65720=]22l26.3m8.32[.1124603m8720=285=2;.2.n4.87.32
[-.=27-.2[.112[5/26.7m6=.327024603m8720=]
oDNEGPBUDEfW?_AEDPVDUW?WCEDCEBEKWF@PBD@ND@AAJNFKWFEUD@BD̀CJKEDLpDqDFOED@B@F@WADQ PKRPSEKD
TJ@UDAPNNDKWFEDUEGKEWNENDbcDrMe DQOEBDFOEDGE?EBFD@NDUK@AAEUDWBUDNFP_NDWAFPCEFOEKDQOEBD
7-.285/2=<22/2/8654.32=.5627-.24.6Y065720=/]
s PKRPSEKDTJ@UDtJWA@FcDCJ@UEA@BENDWBUDKEGP? ? EBUEUDP_EKWF@PBWADFEGOB@tJEND
6./4.872X.1:256.25X5215;1.270292=292/.27-2/286.57.32Y0695720=23595<.]
„ ††ƒ‡ƒ€š“ŒŽ
€ †ƒ’ Ž“”ŒŽƒ‰†• –
—‚‚Œ‚ƒ˜‹ w™ƒ
„ ††ƒ‰ ‚Š
ŒŽŽ€‘‡‚ƒ›‘††ƒœŒ‡ƒž
u
2
v
w
x
y
z
{
|
}
€‚ƒ
u~ uu uv uw ux uy uz u{ u| u}
„ ††‡ˆƒ‰ ‚Šƒ€‹ €ˆŒ‚ƒŒŽŽ€‘‡‚ƒ
‰ ˆƒ|Ÿvy
01
¡¢£¤¥¦§¹
º ª¤¸ª»¥¤¦¯¡°­ª¤±¦°¯ª²°¦
²¯¥®¦³ª¤´µ­¡ª®¦«µ´µ¢¥¦
ª¬¬£¤¤¥«
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
9 ''!: $/ 0 &1%
FRHIdJeYTTLTJfggITh
FGHIIJKLM LNO
×ØÙÚÛÜÝÞß
à áÛâáãÜÛÝäØåæáÛçÝØèÜéæØêÜåÝ
ëìÚåÜåÝáíÝíáÛîìæØáéÝ
èìîìÙÜ
]
EEEEEEEEEEEEEEEEEEEEE
<*7 &='#%!. &%= #13
PQGRSLJKRTHNUJJ
VLRGJWLGXYGROHYNT
WLGXYGROHYNT
>1,
&?'
![./ 0 &#/1,!&17!@&
PO
HIIJZSLNJ
\
^
_
FRdT
`
a
b
B
c
#'!&17!C &,!: /%,,#13
D
lmnmop qrsqturovwxyzo{x|wy}~oxyzuwy€uo9
}qo‚&y
€yƒ!<&17'
yuovqrƒ#13
|}yq€o
„|ƒ | u
iHUJjk^b
†‡ˆ‰ˆEŠ‹ŒˆŽŒˆ‰E‘ˆEEˆ’“‘ˆ‰‰Œ”E”•E–‡ˆEE•‹—ˆ–Ž‰E”•E•”‘—–Œ”E—ŠˆE–‡–E
#'!&17!C &,!.#'7! + &#/1,
‡˜ˆE™ˆˆEŒ‰š‹‰‰ˆEˆ‘ŽŒˆ‘›
œžŸ ¡¢£¤Ÿ¥¦£Ÿ¦§¨©¦Ÿª¤££«Ÿ¬¨¡£¤£©Ÿ­¤®£¦Ÿ”‘«ŸªŸ¡¢£Ÿ¤£¦¥¨¡®¯Ÿ°¥©Ÿ¨§¦¦Ÿ¤±¡£Ÿ¦Ÿ¡§§E
EE Š‘ˆ–²EEE³”‰–E´Œ‘š‹Ž–Œ”Eµ –ˆ‘ŒŽE¶³´µ ·E¸‡Œš‡EŒ‰Eš”‘‘ˆš–Ž¹E‰ŒºˆE–”E
ŸŸ ª§¤»Ÿ±®Ÿ£¼¡£¤®±¨Ÿ¬¨¡£¤Ÿ½±¾£Ÿ§®Ÿ¡¢£Ÿª§¤»±¡§®Ÿª±½£Ÿ§¤Ÿ¿£¤ª§¤±¡§®¦ÀŸŸÁŸ¦§¨¥­¨£E
EE ³´µ EŒ‰E–‡ˆE“‘ˆ•ˆ‘‘ˆEš‡”ŒšˆEˆ›Š›E¸’²E‰”Œ‹—Eš‡Ž”‘ŒˆE”‘EšŽšŒ‹—Eš‘™”–ˆE
ŸŸ ¿±¤¡½¨£¦Ÿœ¤£»§Â±¨Ÿ§ªŸ±¨¨Ÿ§ªŸ¡¢£¦£Ÿ¤£Ã¥¤£Ÿ±Ÿ¢Ä©¤§½¢¨§¤½Ÿ±½©Ÿ°¥¦¢Ÿ¦®½£Ÿ¡¢£E
EE ‰”ŽŒE“‘–ŒšŽˆ‰E™ˆš”—ˆEš”–ˆEE¸Œ–‡EE“”Ž¹—ˆ‘E‹‰ˆE–”E˜Œ‰š”‰Œ•¹E–‡ˆ—E¸‡ˆE
EE ‹‰ˆE‰E³´µ ·›EEŋˆE–”EŒ–‰E‰”Ž‹™ŒŽŒ–¹EE¸ˆŒŠ‡–ˆE™‘ŒˆEŒ‰E‘ˆÆ‹Œ‘ˆE¸‡ˆE
EE ‹‰ŒŠE‰”Œ‹— Eš‡Ž”‘ŒˆE“‘–ŒšŽˆ‰›
E
œžŸ Ç ®» ¦£Ÿ½§®¡±» ®±¡§®Ÿ§ªŸ¡¢£Ÿ¬¨¡£¤£©Ÿ­¤®£ÀŸŸÈ¿£¤±¡§®±¨Ÿ¡£½¢®Ã¥£¦E
EE –”Eš‡Œˆ˜ˆE–‡Œ‰EŒšŽ‹ˆÉ
EE ¶·EE
EEE E
´ŽˆEŒ‘–E¶—ŒŽŽE‰šŽˆ·²Ešˆ—ˆ–EE—‹E‘ˆ‰Œ‹ˆE•‘”—E–‡ˆE–Ê‰²E
ŽŒˆ‰²E“‹—“‰²E–‹™‹Ž‘‰Eˆ–š›E“‘Œ”‘E–”E‹‰ˆ
EE ¶™·E
µ ŒŒ—Œ‰ˆE‹‰ˆE”•E“Œ“ˆE”“ˆE¸‡ˆE‘‹ŒŠE–‹™ŒŠ
EE ¶š·E
ŸŸ Ÿ
EE E
ŸŸŸ Ÿ
ˉˆEˆŒš–ˆE–Ê‰EÌE“‘–Œš‹Ž‘Ž¹E•”‘EŠ‘˜ˆŽE“šÊEEšŒŒ‰ŒŠE
§¿£¤±¡§®¦ŸÍ ¢£¤£Ÿ¡¢£Ÿ°¥©Ÿ¨§¦¦Ÿ¤±¡£¦Ÿ¡§Ÿ¡¢£Ÿª§¤»±¡§®Ÿ¦Ÿ¢¯¢ÀŸÎ ¦£Ÿ§ªŸ¡¢£Ÿ¦±»£E
ˆÆ‹Œ“—ˆ–E•”‘Ešˆ—ˆ–ŒŠE”“ˆ‘–Œ”‰E‘ˆ‰‹Ž–‰EŒEš”–—Œ–Œ”E–‡–EŒ‰E
£¤ÄŸ©ª¬½¥¨¡Ÿ¡§Ÿ¤£»§Â£
œžŸÏЧ¨©¦Ÿª¤££ÑŸ­¤®£Ÿ¦Ÿ®§¤»±¨¨ÄŸ©£¬®£©Ÿ±¦Ÿ½§®¡±®®¯Ÿ¨£¦¦Ÿ¡¢±®ŸÒÓӟ¿¿»ŸÔ§¡±¨E
ŸŸ Ð¥¦¿£®©£©ŸÐ§¨©¦ÀŸŸÕÓÖ Ÿ§ªŸ¡¢£¦£Ÿ¦§¨©¦Ÿ¦¢§¥¨©Ÿ­£Ÿ¦»±¨¨£¤Ÿ¡¢±®Ÿ§®£Ÿ¡£®¡¢Ÿ§ªE
EE –‡ˆE˜ˆ‘ŠˆE“”‘ˆE‰Œºˆ›EE
01234567 28609 0 2659 2 7 0882258002596 4660825600
22
3456789 5 79 7 77 7 7745 5577 47 5 7572
77 97757 9 79 77 9 7 7 4735 6797 5 77 2
22 !"#2$%&'(#"2$!")*+&#,2)-!)2-!.#2/%)20*,,%&.#02%"2-'0"!)#02$"%$#"&'21%2)#/2
22 03#2)%2*/!0#43!)#2(*5*/6782292+-#(*+!&2:"#!;#"2,-%3&02!&,%2:#2!00#02)%2)-#2
22 .*,+%3,2,%&3)*%/2)%2#/,3"#2)-!)2)-#2.*,+%,*)'20#6"!0#,2$"%$#"&'<2*8#82,%2)-!)2
77 5 75997 775 777= 7 745 7 5 797 72
22 >#&&:%"#2!)2)-#2#/02%22)-#2%$#"!)*%/8
3467 ? 55577 47 5 79 777555577 7 52
22 1%.#":!&!/+#72:#)>##/2)-#2"#,#".%*"2$"#,,3"#2!/02)-#2-'0"%,)!)*+2-#!02%22)-#2
77 47 5 77@749 7 7ABB7 57 77 75757 7 45 72
77
5775757 7 CDEDFG HIJHKLIFMLNOPQRSLTFUHFV QPQW QTLFXHIW YUQHPFZYW Y[L
\)2*,2+&#!"22"%(2)-#2!:%.#2)-!)2(%,)2%22)-#22%"(!)*%/20!(!6#2%++3",2>-#/2)-#2>#&&2*,2
;*&&#02$"*%"2)%2)-#2,)!")2%22)-#2>%";%.#"822]"%$#"2>%";%.#"2$&!//*/62+!/2#/,3"#2)-!)2
(!/'2)'$#,2%22>%";%.#",2+!/2:#2+!""*#02%3)2>*)-%3)2-!.*/62)%2;*&&2)-#2>#&&822^%(#2
5 77 757 7_`77a7 7 a75997 77 5 inmdgmfcdivnilmhnewfvin{
rl{ ludfqhmtfknwm
ghigjklmhnef{ lmdihlk
nifwlerfckju
hefglwheu
~ tdefqnipheuflsnodfmtdfclgpdi
wmncfvkjhrwfsdheufhedgmdrfhemnf
mtdfvni{ lmhnefs€fcklgheufckjufhef
mjsheufmlhkfchcd
2222222222222222222222222
bcdefghigjklmhnefolkodfmn
phkkfqdkkflonhrfsjkktdlrheuf
vkjhrfhemnfmtdfcdivnilmhnew
‘’“”•–—
˜”™š›œ’”•žŸ“• ¡Ÿ¢›‘•
žŸ“£ ¤Ÿ›•¢ £ ‘”•¢’“›‘•
¥Ÿ“¦Ÿ¡”“
xwdfgkdlefyfenezrl{ luheu
phkkfchkkw|f}dwhuefvnifgn{ ckdmdf
id{ nolkflvmdifqnipnodifvhehwtdr
1*72 † #+-!/*+!&&'2*,%&!)#2)-#2:%))%( 2%22)-#2)3:*/62>*)-2!2$&362*22)-#2>%";%.#"2
22 +%/+#"/,2#43*$(#/)2$&!+#02!:%.#2)-#2$!+;#"8
1**72 ‡,#2>*"#&*/#2>%";%.#"2(#)-%0,2%/2!2ˆ&*.#‰2>#&&2#868
‚hufƒ|„
22 1!72 ˆŠ3($2:!*&#",‰2)%2$&!+#2(!)#"*!&2>-*+-2>*&&2*,%&!)#21$"%)#+)72)-#2
222 2 $#"2%"!)*%/,822‹"!.#&2*,2%2)#/23,#0822Œ-*,2-!,2)%2:#2"#(%.#02!2)#"2)-#2
777 7 47 7 5 7 77799777 7 7 5 22 1:72 † !/'2>#&&2,#".*+*/62%$#"!)*%/,2,3+-2!,2$#"2%"!)*/6<2$&!+#(#/)2%22:"*06#2
222 2 $&36,2#)+82+!/2:#2+!""*#02%3)2:'2>*"#&*/#8
1***72 *"#&*/#2%$#"!)*%/,2!"#2/%)2)-#2%/&'2)'$#2%22%$#"!)*%/2>-*+-2+!/2:#2+!""*#02%3)2
22 %/2!2$"%03+)*%/2>#&&2Ž2+%*&#02)3:*/62!/02,/3::*/623/*),2+!/2:%)-2%$#"!)#2>*)-2
22 !2-*6-2>#&&-#!02$"#,,3"#8
01
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFGHIJKLMFNOIPKFQRISTULIVWXIRKTSYZXIVUIVWXIMXTSUTYVFUNRIVWTUKOWIVK[FNOIVWYVIWYRII
\ ]^^_\`_abcdd^e\fgh\c\_ij]^h\gf\k^cha\lh^cb
\bmg\
nhg]d$
^ja
9^a'
'!:
/ o0 &1%
EYHI pWXIVK[FNOIFRIZUUqXQI[rIVWXIZUqQIPKFQIYNQIZUNVTYZVRsIItSIYIVK[FNOIYNZWUTIWYRI[XXN\
\\ ia^e\`u^u\bv^\lgj nd^b`g_\eg^a\wxy\lg_bc`_\c_k\j gz`_{\a^cda|\bv`a\
<*7^\&='
#%!
13cnchb
\\ lg_bhclb`g_\gf\bv^\bi]`_{\jck\i_a^cb\bv^\egm_vgd
ncl}^h
\gh.
\nid&%
d\bv^\=
bi]`#
_{\
E[HI pWXIMKLMXQIPKFQI~FqqIZqXYNIVWXIFNRFQXIUSIVWXIUqQIVK[FNOIYNQIVWXITXRKqVFNOIQX[TFR\
\\ m`dd\]^\`_^lb^e\bg\bv^\n^hfghcb`g_au
>1,&?'!./ 0 &#/1,!&17!@&
€a^\gf\c\abh`_{\gf\lg`d^e\bi]`_{\gh\a_i]]`_{\i_`b\bi]`_{‚\mv^h^\bv^\`_a`e^\gf\bv^\
bi]`_{\vca\]^^_\ld^c_^e\]^fgh^\ia^\m`dd\czg`e\]gbv\nhg]d^jau\\yv^\dcbb^h\nhg]d^j\
jck\]^\h^eil^e\]k\c\nhgn^h\nijn`_{\alv^eid^\
`f#
\'
bv^\
m^dd\C
c_e\
nh^a
aih/%,,#
^\lg_e`b`g_a
\
!
&17!
&,!
:
13
ch^\cnnhgnh`cb^u\\ƒ\e`dib^|\`_v`]`b^e\cl`e\`a\nijn^e\`_bg\bv^\bi]`_{\c_e\e`andcl^e\bg\
YIRYSXIQFRVYNZXIY[UGXIVWXIMXTSUTYVFUNRIERUIVWYVINUNXIUSIVWXIPKFQIZULXRIFNVUIZUNVYZVI
m`bv\bv^\n^hfghcb`g_a‚u\\„b\`a\bv^_\l`hlidcb^e\]cl}\bg\bv^\aihfcl^\^u{u\]k\gn^_`_{\c\
!\<&17'
ad`e`_{\a`e^\eggh\`_abcdd^e\`_\bv^\bi]`_{\c]gz^\b9
v^\&
ncl}^h
c_e\h^z^ha#13
^\l`hlidcb`_{\]k\
nijn`_{\`_bg\bv^\c__idiau
B
D
ƒ\a`j`dch\nhgl^eih^\m`bvgib\bv^\_^l^aa`bk\bg\gn^_\ad`e`_{\a`e^\eggha‚\jck\]^\ia^e\
i_`
#'!&17!
&,!
.#'
+
bg\ld^c_\bv^\`_a`e^\gf\bv^\lg`d^e\bi]`_{ a_i]]`_{\
b\g_l^\`C
b\vca
\]^^_\
`_a7!
^hb
^e\
`_bg\&#/1,
bv^\m^ddu
†‡ˆ‡‰Š‹ŒŽ‘‰’‰‘“‹‰”•‹–‹Œ‹‰’‰—•˜ ™‘‰š™˜ ™Ž‹
›IMTUQKZVFUNIqUOI~WFZWILXYRKTXRIVWXIPU~ITYVXIYVIYNrIMUFNVIFNIVWXI~XqqIFRIUNXI
j^bvge\bg\e^b^hj`_^\`f\fghjcb`g_\ecjc{^\vca\bc}^_\ndcl^u\\œ`{ih^\ž\\alv^jcb`lcddk\
TXqYVXRIVWXIPU~ITYVXIYVIGYTFUKRIQXMVWRIVUIVWXIŸKYqFVrIUSIVWXIEMUVXNVFYqqrIWrQTUZYT[UNI
nhgeil`_{‚\ac_e|\ca\j^caih^e\]k\bv^\{cjjc\hck\dg{u\\„b\`a\ld^ch\bvcb\bv^\]gbbgj\ac_e\
vca\ngb^_b`cddk‚\]^^_\`jnc`h^e\a`_l^\`b\`a\_gb\lg_bh`]ib`_{\bg\bv^\nhgeilb`g_u
ÅÆÇÈÉÊËÌÍ
ÎÉÏÐÈÑÒÆÏÓËÔÏÇËÕÊÖÇÖË
×ØÆÓÓÊÉËÏÉËÙÆÐÊÏÚËÆÐÊÓÒÆÛÊ×Ë
ÜÏÓÊ×ËÊÝÞÆßÆÒÆÓÇËØ××ÆßÔÊËË
àÏÉáâÒÆÏÓËÐâáâÇÊ
\\\\\\\\\\\\\\\\\
¦§¨¥©¢ª«
¡¢£
¦§¨¥©¢ª«
¡¢£
¦§¨¥©¢ª«
¤¢¥ ¡¢£
¬§­ ­ §¥®§¨
¯ «§°±²«³¥ ¡¢£¥®§´«
yv^\mghe\»¼½¾¿½ÀÁÂÂÃ\`a\ia^e\v^h^\a`_l^\bv^h^\ch^\gbv^h\ngaa`]d^\h^cag_a|\cnchb\fhgj\
fghjcb`g_\ecjc{^|\bvcb\lgide\cdag\cllgi_b\fgh\bv`a\dcl}\gf\m^dd\nhgeilb`g_|\^u{u\bv^\
]gbbgj\ac_e\jck\vcz^\cdh^cek\]^^_\e^nd^b^e\c_e\bv^\h^a^hzg`h\nh^aaih^\`a\bgg\dgm\
µ¶¥·¸¹º§
VUIYqqU~IVWXIÄUNXIVUIPU~s
01234567 28609 0 2659 2 7 0882258002596 4660825600
23456789 5555345!"5###453$%&5'%5'5
'54545%454&4(344)552345!"*'5'545+$55
3$%&5'%54455543&5#45,5344555&$544$55
-(535'5455#445(34545%35544)55.(4/405
3456789 5'5'545+$55534&544541%4535345
234567896:;<=5>6?@A6B8C?DE8@68F6DGH6F89I?DE8@6A?I?JH6I?K6LH6EAH@DEMHA6N7O9675
EDP69HI8Q?B6EP6?DDHIRDHAS662GEP6EAH@DEMC?DE8@6R98CHPP68FDH@6E@Q8BQHPT
5 U5345(4&&54554'$5#4535353/454%&455
55 '54
7EE>6 :?99KE@J68VD6PRHCEMC6B?L89?D89K6DHPDE@JW6PVCG6?P6?69HPH9Q8E96C89H6XVPGE@JW6D85
66 AHDH9IE@H6EF6DGH6EAH@DEMHA68RH9?DE8@P6AEA6E@AHHA6BH?A6D86C89H6A?I?JH6F896DGH5
66 R?9DECVB?96C8ILE@?DE8@68F6DGH6XVEAP6E@6YVHPDE8@6?@A6DGH69HPH9Q8E96F89I?DE8@
5U53454545(35#345&$&5431%45%35
66 ?P6DGH6PC?@@E@J6HBHCD98@6IEC98PC8RH6D86C8@M9I6DGH6A?I?JH6DKRH6?@A6DGH5
55 45&5534454/4&#54553(554/45)
2345$5#+&45$#45'5'545%455353#4545(3$5
'5345%453/455+4545%555$454)5Z5(&&5&5
LH6PHH@6DG?D6H?CG6P8V9CH68F6F89I?DE8@6A?I?JH6G?P6EDP68[@W6PRHCEMC6CV9HS6
01
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
EFGHIIJKLMNGHG
9 ''!: $/ 0 &1%
OPPQRSTUVOVTWVXVWYZZX[\V]^V_]ZZ]RV`\PQWV]^V^][ZX`T]RVSXZXaQbVSTXaR]W`T_V_cYQWV
XRSV[QZQSTX`T]RV`Q_dRTeYQWV^[]ZVfORVTR`[]SY_`T]RV`]V`dQVgXWTWV]^VhQccV_]ZPcQ`T]RW
ijklmnojkpqiroqsrtpuvpwxuiryzqs{rxskjkpq|}r~pmujxirp€rxpu‚xrƒkq‚ry„mnio}r ††‡|{
<*7 &='#%!. &%= #13
ˆ‰‰Š‹ŒŽˆ
‘’ ’ ‘‹“”’ ”•Š–—‰Š˜
 ‘‹Œ‘‹‘Ÿ–—‰Š‘ž“ ”’ ”•Š
½¼»°¶­¬µ±º®¹¬¸
>1,&?'!./ 0 &#/1,!&17!@&
“ ”•‹‘˜œ ™›Š˜
š Š’ ŠŒ”™
·¶µ± ´ ¬³³²± ®«­¬°¯ ®­­¬«¢¤ ª£© ¨¤ §¦¡¥¤ ¢£¢¡ ·³¼¹¿³±¿°±¶³¹¿½¿³±®«Ä¬¹­®¿«±Á¬³¿´ ±½¼»°¶­¬
µ½¾­¿ÀÀ²±À¿³³¿´¬»±Á¼±°¬­¾°«±­¿±À³¿´±®«±¶ À¿°¯ ¶­®¿«±Ã¿®«­Å±±Æ«µ¾³¶­¬»±°®µ¬°µ±¿°±­¾Á®«·Å
À¬´±¯ ®«¾­¬µÂ±Ã°¿»¾¹¬µ±¶±µ¯ ¶³³±¶¯ ¿¾«­±¿À
´ ¶­¬°
¿®³±´ ¬³³²±¾µ¾¶³³¼±¿«³¼±À¿°¯ µ±«¬¶°±¯ ¾»±³®«¬ ®«µ¾³¶­¬»±°®µ¬°µ
®«±¶°¹­®¹±°¬·®¿«µÂ±¯ ¶¼±¶³µ¿±À¿°¯±®«±»°®³³®«·
¯ ¾»
ÔÓÎËÊÒÊÉÈÇÎÏÇÑÊÐÏÎÍÌÇÊËËÊÉÈǹ³¶¼±¿°±µ¿¯ ¬ ÚËÓÛÇÚÎÒÉÏÎËÇÑÎÏÇØÏÈßÈÒÉÊÎÒáÇÏÈÉÓÏ×È×ÇÓÚÊ×
ÑÈË×ÌØÓÏÌÇ ÙÒÎÒÇ ÚËÓÛÌÜÝÇ ÐÏÊÒÈÇ ÌÓËÊÒÊÉÛ ÑÎÏÇÏÈÖ ÎßÓËÝÇÖÓÛÇÒÈÈ×ÇÉÎÇËÊÖ ÊÉÇÏÓÉÈÇÊÒ
ÚÞÓÒÕÈÌÇ Ö ÓÛÇ ÉÏÊÕÕÈÏÇ ÑÊÒÈÌÇ Ö ÎßÈÖÈÒÉÝ ÈâÉÏÈÖÈÇÚÓÌÈ
ÌØÎÏÓ×ÊÚÇÏÈ×ÍÚÉÊÎÒÌÇÊÒÇÑËÎàÇÏÓÉÈáÇßÓÏÊÓÐËÈ
ØÏÎ×ÍÚÉÊÎÒÇ ÏÓÉÈÇ ÉÈÌÉÌÝÇ ÑÊÒÈÌÇ ÊÒÇ ØÏÎ×á
ÈÖ ÍËÌÊÎÒÌÇÏÓÏÈÇÐÍÉÇØÎÌÌÊÐËÈ
æä ãìíîíãç ìëêéèç æåäãÉÊÖ ÈÌÇ ÊËËÊÉÈÇ ÓÒ× êìîôîùãçñîíøçúûüýú þçîÿçôêä êïãçîæçæøêëëåñ
æä ãìíîíãçîðíãòóãôôãôçìëêéõçöãòäãêóîëîíéçîæ ú éôòê0ëîìçþòêìí0òãçîÿçôêä êïãçîæçôããöãò
æãðæîíî÷ãçíåçìøêðïãçîðçñêíãòçæêëîðîíéçåò íøêðç123
óòîðãçíéöã
0æ0êëëéç ïêæç ñ ãëëèç æä êëëç öåòãç íøòåêíæè ä êíòî çíòãêíçñîíøçêëìåøåëçåòçæ0òÿêìãçíãðæîåð
0ðíòãêíãôçñêíãòèçëåñçöòãææ0òãçÿåòä êíîåðõçîÿ ëåñ ãòîðïç æ0òÿêìíêðíõç îð ãìíçïêæç îðç ïêæ
÷ãòéçëåñçöòãææ0òãç56çêóå0íç782çöæîýÿí9èçöåòã òãæãò÷åîòçíåçôîæíêðìãçåÿç17çÿí
íøòåêíçæîùãçôåãæð íçäêííãò
0æ0êëëéç ÿåòäêíîåðç ôãöãðôãðíõç äêé òã÷ãðíîåðç çê÷åîôçåòçä îðîä îùãçñ êíãò
ôãìòãêæãçöòåô0ìíîåðçòêíãçîðçæã÷ãòãçìêæãæçåò ìåðíêìíçêðôçëåñ ãòçæ0òÿêìãçíãðæîåðçåÿçñ êíãò8
ìê0æãçëåðïçìëãêð0öçíîä ã8çç åæíçìåä ä åðçîð ãä å÷êëççêëìåøåëæçêðôçæåä ãçæ0òÿêìíêðíæ
ÿåòäêíîåðæç ñîíøç æä êëëç öåòãæç êðô
ä îìòåöåòå0æçìëêéæ
øîïøçæ4îðèçìêðçîð ãìíçîðíåçñãëëçêíç÷ãòéçëåñ ìëãêðå0íçåÿçòãöãòÿåòêíã
òêíãæõç òãöãòÿåòêíîðïç ä êéç æøåñ ç æøêòö
îðìòãêæãæ
ä öåææîóëãçíåçîð ãìí #)##!# !!
"#$%!!" " !
&%&'#%&#!!##
!#$( "!#&!" ##
32-0/10/.*-,+*%#'#$ C DE/1:/CDEFC G/68-+/-0/7 69:-H/68-+/I1.
317 49-743/3<1=3/47*,3-103/-0/1-,/=4,,3> 31,J409/=63</?1,,1=/=-9</68-+/-0/=4,,3
;11:/7 *+/810+-9-10-05/ .4?1:4/847409> = -9</ 1-,/ .634/ 7 *+>/ ?167 /1:/ I499-05
?:4@*409,A/ 4081*094:4+/ -0/ 1;40/ <1,4 8,460*;3/6:4/1?940/*34?*,
817 ;,49-103/60+/<1:-B1096,/=4,,3
7 6A/6,31/.4/4081*094:4+/?1,,1= -05/68-+/1: ?-,94:/9:4697 409/?,*-+3/K/*34/8,460/96023
?:68/39-7*,69-103/1:/=1:21J4:/?,*-+3/=<4:4
6/+-:9A/=694:/= 63/*34+/1:/9<4/=694:/=63
<6*,4+/1:/391:4+/-0/6/+-:9A/9602
860/:4+*84/-0I489-10/-0/?,11+/;:1I4893/K .4994:/8,460-05/1?/=694:
.682?,1=/3<1= 3/;6:9-8,43/60+/1-,/86::AK
1J4:/:49*:0
,6:54/ = <1,4/ 7 *+/ ,13343/ -0/06 9*:6,,A 68-+/*34?*,/-?/+67 654/-3/3<6,,1=>/?:68/-?
?:689*:4+/ ?1:7 69-103>/ -094:7 -99409 +44;>/91/;:4J409N/-7 ;:1J4/31,-+3/:481J4:A
;:1+*89-10/69/,1= /60+/ 7 1+4:694/:6943>
-0?:4@*409/:481J4:A/1?/=<1,4/7 *+/60+
7 *+/ ?-043L/M174/47 *,3-103N/43;48-6,,A
6?94:/68-+/9:4697 4093>/7 6A/6,31/188*:/-?
<1,4/-3/;11:,A/8,4604+/+*:-05/+:-,,-05
B
#'!&17!C &,!: /%,,#13
D
Ö ÊÕÏÓÉÊÒÕÇÑÊÒÈÌ
9 & !<&17'#13
#'!&17!C &,!.#'7! + &#/1,
æñ ãëëîðïçìëêé
ñ êíãòçóëåì4æ
öååòçëåêôçÿë0îôçòãì8
ÿîëëçîðçöãòÿæ
!
;6:9-8,4/+67 654
?:689*:4/;,*55-05
01234567 28609 0 2659 2 7 0882258002596 4660825600
4357 97 53878 2 5 47 9538736534232 57 8 537 37 583 5 3457 8 2 5 47 9 8 5 47 7 4 7 385 53 8 8737397 57 5387 458 38 3 57 88387 427 99 345 8 2 57 8374 57397 5 574 58 385 7 38 75 7 7 53875 7 3 8338 87 8838
5 82728 2 4
4327 53878 2 5 4
4 5 78 2 5 47 97 5387 57 7 4 4 7583 5 34578753 8 875735
8 87 5783 58 7479 27 538
574 83 37479 2
5389 27 83 58 7 58 7 36 437 8 2 327 5387 4 7 42 738 3 57 437 273732
383 57 43
837 579 27 538
2337277 !"798 78 2 387 4274#3 5 8
347 783 38798 7 4377 353
8 7 8 5
3277 379 8 7 97 7 7 42 4 7583 5 345743 3 8
5387 427 37 7$783 2 47 47 ,3 21&0)(
2(&%+.1,',3 21&0)(&/'.-,+*&')('&% 42 445 ()'7)<(';)1,'%-,+%3 ,(%'&244,&%,75'%-,+%')(
6-,+%,7'0('%2.0(4'+%'8-,&&2-,'7-)8'8)0(%&'0( &2-:+6,'0:'&%-0(4'-,7,&04('0&'03 8-+6%06+1
8080(4'&9&%,3 ':-)3 '823 8&'%)'6;)*,
&%+.010>,7',3 21&0)(
=,-9'&%+.1,',3 21&0)(/'8+-%0+119'< ,%%,7':0(,& %-,+%'<0%;'32%2+1'&)1=,(%'+(7'+6075'-,3)=,
+%'0(%,-:+6,5'6)33 )('+:%,-'7-0110(4'327 7)<(;)1,'&)2-6,'):'&)107&'0:'8)&&0.1,
70&8,-&+1')-'61,+(28'):'3 27')-'6,3,(%
:0(,&'.9'+6075'3 +9'+1&)')662-')('8)193 ,61,+(285' 6)3 3 )(' 0(' 8-)726%0)(' :)-3
2(6)(&)107+%,7':)-3 +%0)(&/',&8,60+119'+:%,+607')-'4-+=,1'8+6*0(4')8,-+%0)(&
)01'.+&,'3 27',3 21&0)('7+3+4,
=,-9'6)3 3)('0('<,11&'7-011,7'<0%;')01'.+&, < +&;'< 0%;'+-)3 +%06'&)1=,(%':)11)<,7'.9
3 275'0:'%-,+%,7'<0%;'+607')-'.-0(,'.,:)-, 3 2%2+1'&)1=,(%' +(7'+6075'3+9' -,?20-,
%;,'&21:)(+%,',3 21&0:0,-&'+-,'<+&;,7'):: &,=,-+1'%-,+%3 ,(%&5'62%%0(4&' -,3 )=+1'0&
%;,'62%%0(4&' .9' 8-)726%0)(')-' &)1=,(% 03 8)-%+(%
%-,+%3 ,(%&/'+(',3 21&0)(':-)3 '%;+%'6+('1)6*
28'%;,'<,115'%;,':0-&%')(,')-'%<)'%-,+%3 ,(%&
3 +9'.,'&;)-%'10=,7'+&'3 )-,'3 27'+(7'3 27
:01%-+%,'3 )=,&'.+6*'%)< +-7'%;,'<,11.)-,/
%;0&'0&'):%,('%;,'6+&,'0('(+%2-+119':-+6%2-,7
:)-3+%0)(&
4 4797 538
@ABCDEFDGFH@FDIJKJLMNHMO AK@JPG
@KANZM
VDBDCCJGH]HW DX
01
@FDIKMHFPHQMBRH@FDIKMHMOAK@JPGSHEPO OPGHFP FBMDFHPGH@ABCDEMHJCHFMO VPBDBRSHA@MHO AFADK
@MQMBMHDCFMBHDEJNHFBMDFO MGFSHO DRHIMHDIKMHFP @PKQMGFHPBH@ABCDEFDGFHFPHVBMQMGFHMO AK@JPG
@MMHFTMH@FDIJKJLMNH@UJGHDFHFTMHNBPVHJGFMBCDEM W JFTHGMXFHDEJNHYPI
@KANZMHJ@HDGHMO AK@JPGHFTDFHJ@HGMDBKRH@PKJNS VBMQMGFJPGHJ@HIM@FHEABMSHA@MHGPGH@KANZJGZ
JFHODRHIMHFBJZZMBMNHIRHDEJN[HPJKHID@MNHOAN[ DEJNH @R@FMO @[H FM@FH W JFTH JBPGH EPGFMGF
D@VTDKFMGM@[HPBHJBPGHEPO VPAGN@SHNJ@VMB@M MXVMEFMNHJGHW MKK
FTMH@KANZMHJGHXRKMGMHDGNHDGDKRLMH CPB
EPO VPGMGF@[HVDBFJEAKDBKRHJBPG
\PBO @HJGHFAIJGZHGMDBH@ABCDEMHD@HPJKHEPPK@ @EBDVJGZ[HEAFFJGZHCPBHO METDGJEDKHBMOPQDKS
DGNHEKPANHVPJGFHJ@HBMDETMNSHO DRHIMEPO M TPFHPJKHA@MCAKHJCHNMVP@JFHJ@H^H_aaHCFHCBPO
JGEBMD@JGZKRHDHVBPIKMOHD@HCJMKNHDZM@SHO P@F @ABCDEMSH@PKQMGFH@PDU@HPGHNMMVMBHNMVP@JF@S
VDBDCCJGHNMVP@JF@HO MKFHDFH^H_`ab\cHd CFMG @PO MHJGTJIJFPB@HDQDJKDIKMHCPBHVJVMKJGM@HDGN
CPBOHDFHVBM@@ABMHBM@FBJEFJPGHJGHFAIAKDB@S CMWHVBPIKMOHW MKK@SH@PO MHWMKK@HBMeAJBM
@PCFHFPHTDBNHOD@@HCPAGNHDFHVBM@@HNBPV@S EPGFJGAPA@H NPW GTPKMH FBMDFOMGFHFTBPAZT
EPKPBHBMNNJ@THIBPWGHFPHIKDEUSHMQMGHWTJFM O DEDBPGJH@FBJGZSH@VMEJDKHIDEFMBJDHDBMHDK@P
A@MNHFPHVBMQMGFHVBMEJVJFDFJPG
PBHZBDRHVP@@JIKMS
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
SJLJRRPGHPGHRFLK JEPFG
SJLJRRPGHJREOLHNEPK XVJEPFG
tjtccfuhfuhcdmz hdfuki
‰ˆ‘‰Ž‡†‡ˆ
NOOGHJNHNQPGHFGHEONEMHK JIHFLHK JIHGFE ELOJEH UPEWH TFUGWFVOH WOJEH \OGOLJEPG\
TPNJSSOJLHPRHUOVVHPNHNWXEPGHRFLHNOYOLJVHTJINZ SLF[ONNONHPRHUOVVHPNH\FFTHSLFTX[OLZHNFVYOGE
[VFXTHSFPGEHFRHFPVHPNHGOJLHEWOHLONOLYFPL NFJQNHJVNFHXNOTZHNFK OHPGWP]PEFLNHJYJPVJ]VO
EOK SOLJEXLOZH SLONNXLOH TLFSH K JIHELP\\OL EFH ]OH XNOTH UPEWHJHW ITLJXVP[H RLJ[EXLO
ELOJEK OGE
SJLJRRPGHTLFSHFXE
PG^O[EPFGHFRHJH[FFVHNEPK XVJEPFGHRVXPTHKJI tddmzhw{khz kddhwmhldktuhexhmuhfwihmzusheik
SLO[PSPEJEOHSJLJRRPGHPGHLONOLYFPLHFGH[FGEJ[EZ |rdkukhxjkcdei{ht{ktghmchtlfghz{kuhw{k
U OVVHK JIH]OH[VOJGHXSHNVFUVIH_`HEFHa xjmydko hfihqumz uhwmhmllej
U OOQNbHJREOLHNEPK XVJEPFGHOYOGHEWFX\WHVFJT
cdefghfihjklmnkjkghomjkhpeflqdrshotrhikkht
gkljktifuvhiqfuhfchoedwfxdkhyefdgexhwkiwi
tjkhjeu
imcwhwmh{tjghgkxmifwih~umwhiltdk}hfuhiejctlk  ‡‘‰† ‰„‚„ˆ‚Ÿ‡†Ž„‡‚Ÿ‰„‚„ ““ †“
€‚ƒ„†‡ˆ„‰†Š„‡‹Œ  ‡†Ž„‰‰€€†„ƒ  †‘ — Ž‚ˆ„‰†„—‡„Œˆ‡Š
 ‡Ž„ ƒ‘‡†„ ‡’‚ˆ‡Š„ Ž‚„ ‡†‚Œ“‘„ ‘‡‰Ž
”ŒˆŒ‰•–„‰—‚ŒŽ„˜™š›œ„ˆ„ˆŒ€€ ‡†Žž
ˆ‚€Ž„Ž‚„‘‰Š„—‰ „ ‰ˆˆ„Ž‘‰Ž„ ‰•„‚Œ„‰ˆ ­µ®¨­· ®¯­§ ²³­«§ ¨µ°· ¨­³´§¹´Ã´¬³´§ µ³¯¶À
€‰ ‡ˆ–„ˆŒŠ“‡–„ ‰—‡¡ˆ ¢‡„—‰ˆ„‰†Š„‰ˆ„‰ ©°¬Ä®¯­©§©¸´«§¨©§Å쮯®§°µ§­° ¬¸®¯®¦§©°· ®
ˆŽ „—Œ ŠŒ„‚€Ž‡†„‚Œˆ„ƒ Ž‘„‰‰€€† ©¸µ±¨´­¨¯­©§¨µ®§¨¬©°§¸©®±¸¬§±°µ§Â³©ª®µ©³°¯§°±
‡  Ž‰Ž‚†„ˆ„Ž““‡‡Š„—•„Š‡ˆŽ‰— ¢‰Ž‚† ¨©ª«¨¬­³´§· ¨©©¦§¸©®§¨¯­³Æ©¬¸Â¶®§¨Â³­³Ä®§°µ
‚€„‰Ž‡†‡„‡ˆ†ˆ„‰Œˆ‡Š„—•„‰ Š„‚†Ž‰Ž– Å쮯®§²³­«§¨´³Â§³¯§µ®©®µÄ°³µ©§²³­«§· °µ®
‚ŒŽ“‰ˆˆ†“–„ˆ‘‡‰„ †„Œˆ–„‡‡Ž‰• ­«¨¯§½Ç¼È §¨©ª«¨¬­§­°§ªµ®Ä®¯­§©¬¸Â¶®©
‘‰“‡Š„  ‡Ž‰„ ˆŒ€‰‡–„ Ž‡ ‡‰ŽŒ‡
‡ŠŒŽ‚†„ ‰†Š„ £¤ ¥¦§ ¨©ª«¨¬­®¯®©§ ©°±­®¯
² ³­«§³¯´µ®¨©³¯¶§­®·ª®µ¨­¸µ®§¹­§º§»¼½¾¿À
Á¸­§Â°§¯°­§· ®¬­
±¬°² ©§Ä®µÃ§©¬°² ¬Ã§³¯­°§ª®µ±°µ¨­³°¯©§Â¸µ³¯¶ ©°¬Ä®¯­§©°¨Ì§¨©§¯®®Â®Â¦§­®©­§©°¬Ä®¯­§²³­«
ªµ°Â¸´­³°¯§°±§°³¬¦§¸©¸¨¬¬Ã§¨©©°´³¨­®Â§²³­« ©¨· ª¬®§°±§­¨µ§Á®±°µ®§Î°Á¦§«®¨­§°±­®¯§«®¬ª©
­«®§ªµ®©®¯´®§°±§¨§­¨µ§Â®ª°©³­§¯®¨µ§ª¨ÃÉ
±µ®Ê¸®¯­¬Ã§«³¶«¬Ã§¨©ª«¨¬­³´¦§·¨Ã§´°¯­¨³¯
©°· ®§²¨­®µ§­«¨­§³©§­³®Â§¸ª§¨©§Âµ°ª¬®­©§°µ
˪°´Ì®­©Í§³¯§­«®§«³¶«§Ä³©´°©³­Ã§· ¨©©
· ¨Ã§±°µ· §¨­§¨¯Ã§ªµ®©©¸µ®§Âµ°ªÉ§®³­«®µ§³¯ Ð ÑÒ§­°§µ®·°Ä®§¨¯Â§³¯«³Á³­°µ§­°§ªµ®Ä®¯­¦
­«®§±°µ·¨­³°¯§°µ§³¯§­«®§­¸Á¸¬¨µ©¦§· ¨Ã§±°µ· ³¯«³Á³­°µ§ · ¨Ã§ Á®§ ©Ê¸®®Ó®Â§ ³¯­°§ ­«®
Ä®µÃ§±¨©­§¨¯Â§´¨¯§©«¨µª¬Ã§¬³·³­§ªµ°Â¸´­³°¯ ±°µ· ¨­³°¯§ ±°µ§ ¬°¯¶®µ§ ¬³Ä®Â§ªµ°­®´­³°¯¦
®©ª®´³¨¬¬Ã§¨­§¶µ¨Ä®¬§ª¨´Ì§³̄­®µ±¨´®©§°µ§¯®¨µ ©°· ®§ ÐÑÒ§ ΰÁ©§ · ¨Ã§ ­µ³¶¶®µ§ ´¨¬´³¸·
ª®µ±°µ¨­³°¯©§³¯§²®¬¬©§²³­«§«³¶«§Âµ¨²§Â°² ¯ ´¨µÁ°¯¨­®§©´¨¬®§³¯§µ¨µ®§´¨©®©¦§³¯«³Á³­§¨´³Â
¨´µ°©©§­«®§ª®µ±°µ¨­³°¯©¦§· ¨Ã§°´´¸µ§· °µ® °µ§­µ®¨­§²³­«§ÔÕÖק³±§­«³©§³©§¨§ªµ°Á¬®·
±µ®Ê¸®¯­¬Ã§³¯§®¨µ¬³®µ§©­¨¶®©§³¯§©°· ®§±³®¬Â©
² «®¯§ ªµ®©©¸µ®§ µ°ª§ ³©§ ·°µ®§ ©®Ä®µ®Ç
Ï ©¸¨¬¬Ã§¨· °µª«°¸©É§¯°¯Æ´µÃ©­¨¬¬³¯®§±°µ·
¸©¸¨¬¬Ã§ ±°µ· ©§ ¨­§ªµ ®©©¸µ®§ µ°ª§ ² «®µ® ´«®· ³´¨¬§´°¯Ä®µ­®µ§°µ§Â³©©°¬Ä®µ§±°¬¬°²®Â
³¯Â¸´®Â§ Áç ­¸µÁ¸¬®¯´®¦§ · °µ®§ ±µ®Ê¸®¯­ Á稴³Â¦§ ¹Â°§ ¯°­§ ´°¯­¨´­§´°¯Ä®µ­®µ§ °µ
² «®µ®§«³¶«§©¸¬±¨­®§²¨­®µ©§´°¯­¨´­§«³¶« ³©©°¬Ä®µ§²³­«§¨´³ÂÀ¦§¨´³Â§³©§¯°­§¸©®±¸¬§ÁÃ
´¨¬´³¸· §²¨­®µ©§¨¯Â§³¯§ÑØ Ù ±¬°°Â©¦§©´¨¬®§³© ³­©®¬±¦§ ³¯«³Á³­°µ©§ ª¬¨´®Â§ Áç ©Ê¸®®Ó®
¯°­§¨´³Â§©°¬¸Á¬®¦§· ¨Ã§Á®§±°¸¯Â§°¯§°¸­©³Â® ­µ®¨­· ®¯­©§¨µ®§¸©®±¸¬§­°§ªµ®Ä®¯­
°±§ª¸· ª©§¨¯Â§¨­§³¯­¨Ì®©§¨¯Â§¶¨©§®Åª¸¬©³°¯
ª°µ­©§¨¯Â§Ä¨¬Ä®©§³¯§Â°² ¯«°¬®¦§´µÃ©­¨¬©§¨µ®
´«¨µ¨´­®µ³©­³´§±°µ§­«³©§©´¨¬®
9 ''!: $/ 0 &1%
<*7 &='#%!. &%= #13
>1,&?'!./ 0 &#/1,!&17!@&
B
#'!&17!C &,!: /%,,#13
D
9 & !<&17'#13
­¨µ
#'!&17!C &,!.#'7! + &#/1,
´¨¬´³¸· §´¨µÁ°¯¨­®§©´¨¬®
´¨¬´³¸· §©¸¬±¨­®§©´¨¬®
01234567 28609 0 2659 2 7 0882258002596 4660825600
234567 896 3 893
3255842 7 9
4 3 5 8 47 842 7
4 443 8 93
625842 7
93
01
84 3 5 893 8 3 8 47 983 84 9964 94358 3 48 2 3958 48 4
43 9848 4 86 39958649 7 353 84 7 3 8 7 5384 37 859
84 358334 8493 83 48587 3 6963 8 89952 85 893 8649839
95987386483984353 5 8 348648 8!"# $8 958483985%8
93 858 97 834 398 4 84353 5 &'()8*7 7$8 9598585 988+382
59 9847 858 83 5 88 893 9 4 828552549
34 8 32 85837 7 38438 5
9
4 -6 864983 4838358,28 4 4 3 8587 6 63 89 8 398 48 844958552548398 87 5. 85 3859 3 83893%
8358,6982 4 85, 58 7 6 959
38 4 6 8 8 34 8 7 7
3456 3483 48552548 9984858239
7 68 3 8 7 38 2 8 7 4348 4
47 3 826 8798 38689 85
57 838 83 7 9884 49 83 643
325586963 834 8948 5 83 643
32558598 47 5 828 83 643
964 33 9858 846 85 9
7 38648 8585985, 85 83839 4 3 85858012843589 896839
/ 8483985985, 85 83858/ 83 93 848. 88 3985 $
59832 8 82622 85
9-6 / 83846 823%
389 53 839 8 8 4 3 5 8 47 3255 4 8 4398 38 4 4 9964
838 93 85-65839 $8 3 4 9 45484 958 8 6258 7 38 2
4798 47 83845839858 8 93 4 -654 479858 86258483958 3587 3
648 58 93 8 4798 58 473 58384 3 58 83 4839 85
4 6 8 47 3255883986963 8649
348 8 24 8 4 8587 35 99859
39838 . 4384 9964 84
84 69 98 8 83 684 9964 859 4 958945838 5894583 85
96 55 858 8 . 8399593 58595 8 .5958 6258 7 38 8 5
47 8599 8398 8738 38385 6258598834
485-65896987382 846 858945859
834 843 85984 945 82845 5823%
4 9964 8586258598897 3
4 983 4848 3%8245 8 39
93 84 553 9839838 5 87399858 626 3498 48 58 8 473 58 6963 39953 8 58 38 58 8 96 4
93 643 83 4826 87 383 982 845 4 28384 9964 8487382 89 8 3485
85 98 897 8 98382 7 8 998
3 42 78398 3 486 854 39 98%8 4
42 7 98 58 47 3 598 5846 3 4893 5583 8598 34893 643 5
!" #$#%#&'!(#$!)*+,
4
8
;
./ 0 &#/1!2&0 &3
5 & #6!"%#7#,#13
LSOHJMPSKFPTQJUHSHPGT
SFRPQQPONKHFEMGLKJI FHGFE MJRPOHVFPHFPUFI GMJ HMJSHFXPHYFLKJSOYFQGKKGXJRFL[F^efFgRGFTGH
OGI IGTFPTFPTWJOHPGTFXJKKUFXYJMJFUNMQSOJ SKKGXFOGTHSOHFGQFLKJSOYFSTRF^efhVFIGMJ
GMFEMGRNOJRFXSHJMFPUFPTWJOHJRZFPQFHYJFOGKGT[ ijklm nlomipokiq olim qkrmsomlotouukprv
PUF JUHSLKPUYJRF PTF HYJF XSHJMF YSTRKPT\ kwioplkixyomipokiq oliumkpomtjwnpxlomzxn{xzo
U[UHJI ZFPHFOSTFGOONMFXPHYFPTWJOHPGTFGQFST[ klzmsktiopxtxzomuw|}u
X SHJMUVFXYJTFHNLPT\FPUFENKKJRZFLMGX TFHG
LKSO]FUKPI [FISUUJUFGMFSTF^ _`FGRGMFSMJ
PTRPOSHPGTUVF LSOHJMPSF IS[F OSNUJF UKGX
MJRNOHPGTUFPTFHYJFPTWJOHPaPH[FGQFSFX YGKJ
QPJKRVFOGI EKJHJFMJIGaSKFGQFHYJFLSOHJMPSFPU
MSMJK[FEGUUPLKJFbFHMJSHI JTHFPUFNUNSKK[FGTFS
MJI JRPSKF LSUPUVF NTHMJSHJRF X SHJMF PT
HMJSHI JTHUVFRMPKKPT\FQKNPRFGMFPTWJOHPGTFXSHJM
OSTFUGNMFMJUJMaGPMUFX PHYF`cdFLSOHJMPS
ujn um|~mkumzkq k}omnlmkms|xwzm|~miouims|i „ † „‡ˆ„‰Š
tkllnimsomt|pozmxijmktxzmnpmq otjklxtkw
tjkl}oumxlmijom owwvm~pnswoqumu|tjmku
utpooln|iumn€mjrzpk|wxtm€pkti|pomnsu‚mjx}j
xlotixnlm ~pouu|pou‚m uklzm ~pnz|tixnl‚
zn ljnwom utkwom ntt|ppolto‚m |luikswo
oq |wuxnlu‚mklzm znljnwom~kpk€€xlmklz
ku~jkwiolomzo~nuxiumkpomtnq qnlvm~pnswoq u
 xijmkmowwmijkimtkllnimsomspnƒolmznlmnp
oyolm ~|q ~ozm xlinm ujn|wzmkwkrum so
k~~pnktjozmsrmpo~op€npkixl}mijomoww
‹ ‰“Žˆ””’„ŽŒŽ‰Œ“Ž‡ˆ„‹ ‰ŠˆŒŽŠ•‰ŠŽ‘Ž†ˆˆ„“ š„‰œ Ž†‰”Œš™ †‰‘Š”Ž”ˆŒ‘ˆ–‰ŠˆŒ™Žˆ„
”ˆŒ‘ˆ–‰Š –Ž ˆ„Ž Š•‰ŠŽ —Ž ‡‰Ž ’Œ– „ „‰Š Ž‹ Š‘
†„ ‘‘’„ ˜Ž‹ ‰“Žˆ””’„Ž—Š•ŽˆŒ‘ ŠŽˆ‡Ž—‰Š „
†„ˆ–’”ŠˆŒŽˆ„Ž ˆ‘‘Ž ˆ‡Ž †„ ‘‘’„ Ž ‡„ˆ‹
– † ŠˆŒ˜Ž†„ˆ ‹ ‘ŽŒ”’– Ž ‹  –‹ ŒŠŽˆ‡
†„ˆ††‰ŒŠ™Ž”ˆ‘ŒšŽˆ‡Ž‰”–› –Ž”•‰ŒŒ ‘ŽŒ
‰”–Ž‡„‰”Š’„ ‘™Ž‘†‰ŒšŽˆ‡Ž‡ˆ„‹ ‰ŠˆŒŽŒŠˆ
† „‡‘Žˆ„Ž— ˆ„ Žˆ„Ž†„ˆ–’”ŠˆŒŽˆ‡Ž‘ˆ–‘
‹ ‰¥¦ Ÿ§£¨¦ © ¦¡Ÿ¦ªž «©¤ž«¦ ª¡¬ž¦ £ª ¼¯¦¤¡©Ÿž«¦¶¥¦ž¡ª¬§¦Ÿ§­¯¬¦¯£ª¤žŸ¹¦©Ÿž¦§ž¡¸­žª
«žŸ¬ª©¤¬­£®¦£¯¦¢­¯¬¦ž°©­ ± ž®¬²¦¦³§ž¤´¦¨ ­¬§ ­ ž¦£ª¦± ©¢¬­ ¢ž¦Ÿ¬ª­®µŸ²¦¦½­®žªŸ¹¦¤ž± ž®¬¹
¡ µ¡©µž¦ª­®µ¦£®¦¨­ªž¢­®ž¦£ª¦¬©¶­®µ²¦¦·¡Ÿ¦¡ ¡®«¦ ¡¬¤§žŸ¦©Ÿž«¦¯£ª¦ªž ¡­ª
¸¡ª­ž¬¥¦ £¯¦ ¤¡©ŸžŸ¦ ­®¤¢©«­®µ¦ Ÿž¸žªž
¤£ªª£Ÿ­£®¹¦ ± ¡¢¯©®¤¬­£®­®µ¦ žª¯£ª¡¬­®µ
µ©®Ÿ¹¦ ­ ž¦¯¢¡¨Ÿ¦¡®«¦¨ž¡ª¦£¯¦¬©¶©¢¡ªŸ¦¯ª£±
«ª­¢¢­®µ¦£ª¦¢­¯¬¦Ÿ¥Ÿ¬ž± ²¦¦º £Ÿ¬¦¤£± ± £®
¤¡©ŸžŸ¦ ¡ªž¦ž¡ª¬§¦Ÿ§­¯¬¦¢£¡«Ÿ¦¤¡©Ÿž«¦¶¥
Ÿ©¶Ÿ­«ž®¤ž¦£¯¦ ª£«©¤­®µ¦¯£ª± ¡¬­£®Ÿ¦¨ ­¬§
¯¢©­«¦¡®«¦Ÿ¡®«¦¨­¬§«ª¡¨¡¢Ÿ¹¦¡¤¬­¸ž¦¯¡©¢¬Ÿ¹
¡®«¦¯£ª± ¡¬­£®¦± £¸ž± ž®¬Ÿ¦®ž¡ª¦Ÿ¡¢¬¦»£®žŸ
Ÿ©««ž®¦¤§¡®µžŸ¦­®¦¾¿À¹¦Á ¿À¹¦ ªžŸŸ©ªž¹ ªž ¡­ªÂ¦¤£®Ÿ­«žª¦¤£ªª£Ÿ­£®¦¤£®¬ª£¢¦ª£µª¡±
£ª¦¤§ž±­¤¡¢¦¡®¡¢¥Ÿ­Ÿ¦£¯¦¨¡¬žª
9 ''!: $/ 0 &1%
<*7 &='#%!. &%= #13
~nnpm~op€npkixnlu
>1,&?'!./ 0 &#/1,!&17!@&
B
#'!&17!C &,!: /%,,#13
’Œ‘Š‰ Ž‡ˆ„‹ ‰ŠˆŒ
D
9 & !<&17'#13
#'!&17!C &,!.#'7! + &#/1,
¤£¢¢¡ Ÿž–Ž††
¢ž¡´Ÿ
01234567 28609 0 2659 2 7 0882258002596 4660825600
1234 567289 54 5
96 62375
!"#$%&'()*(+,)$-.&/*/)*..$!"#-).&%&'(."&#-(01
D
G
28 ;B F37 COI;<= 48
E38 H 49
XYZ[\\Y]^_`\[a[bY^cYdYeffY_gYh^i_daj^cYZjakYdYZ[\\Yl`d]jcmY^hYnfYd]i[lYoipqrs
%.tu *
*3,&'-"
(.4'v)*$qqwxyuz."${*##-.&{'%&'..{+|&# ".
D
G
8 .{'%&'."&#/'0$}-{.&%{#.'."&#&%
.'&)0
2;B.F),"7#0
CO0)#
I;<= '
3
4
~r"#uz".E"$-/"
8 --
9 .&*u
H .&/$
-'"**"#0()-)$-.&-'"**. **/&#.'&*$.€.#.&%{'(,"*".+"({"'(#.
?&#L€.#-$,+&#-. **,&'u%&**& "#0()-
OR
JPQQ
L;><=J'0"
#-.
&%C-(0-
2 B-{.
O
.+{$'K
}"
? M
OQQ*,*1M KQSTU
2;B RSV
?


‚ ƒ„ 


‰Š‹Œ ŠŽ‘’“„Œ ”Š„•–“Š—Œ ˆ˜
99ˆŠ‘ž–Ÿ“ ¡–“—Ÿ‘4˜  
 
™šš 
™¢£ 
† 
›š 
š¢¤ 
‡ 
œšš 
›¢š 
ˆ
›šš
›¢œ
<= : 4
56
9
;;;()-
; ;$B+$.(
4
¥ "/
-&
57 <= : +&)'/&((#-%'&( **{'&-)/."}".+{&"#.&%}" %&'
N 2&*/&({*."&#¦
49 -
#&{#&*#-
/$
§¨©D
7?C OG
@;;<
?
4:=B48WVP@;<=;4: A49 B WWS
284;9B;BF3QSTU
I
38 !"#$
49 H%&'()*
ª{{*+"#0E
D
G
7CO;<
WWS
=?48
28;BF356
I
;;;E;38 B H 49
D37 WWS
G 8?
2;B F57
CO?I;<=N4WO
56
3 ?H 49
;D;;3;E 8BGQVR
48 f±²¬³haYdcbY´ljcmYµ ´bYXY`did_[a[il¶Yj±[±Y· Y°Y®ffY_gYdcbY
q 57
'2;p
Bq
F3wx7Y«CYoO¬­®¬¯
I;<=Y4°Y
'-Y°Y®±E¸Y
¹Y
f±
²¬³Y
°Y
®±
º²¬³Y
8
9
H
? L
2 B JKPQQ COLM;><=JK OR
QSTU
OQQ
? M
½¾¾
¿ÀÁÂÃ
È
J
L
J
L
»2J;BPQQ
CL¿MżJÆÇKOR
?
RSV
B
?
L
K
ÈM M
2 B K ¿¾¾COM;><=K ¾ÁÂÃ
?
OQQ
QSTU
È
»ÅB Ä¿
2;B RSV
?
<=4: 4
56
#-
B "}".+9Ê."&ËpÌ"({"'-wÌ&'"0"#*t "**,1
;;.;
; É';&-)/.
57 <=44::
<= 449 N 2
56
;;;; ;B 4 9
57 <= :
49 ;B QSTU
? @49;4:NB2WVP@;<=;4: A49 B WWS
?
49 ;B QSTU
? @;4:?B WVP@;<=;4: A49 B WWS
?
;;;;56 B WWS
? N WO
?
57 WWS
56
WWS
?
56
00
;;;;;;;; B B QVR
?
5757 WWS
? N WO
?
56
!" #$#%#&'!(#$!)*+,
MNHJ H
EF
LLLL LR H K
EG MN J
HK S I
4
8
;
./ 0 &#/1!2&0 &3
HK LR TUVW
O PLHJ R XYZPLMNLHJ QHK R XXU
O
O
LLLLEF R XXU
O S X[
O
EG XXU
LLLLEF R TY\
O
EG
5 & #6!"%#7#,#13
9 ''!: $/ 0 &1%
<*7 &='#%!. &%= #13
]^_^`abcc_de`fcghbae]i]ej`kdh_l`]m`nbme`opq `gr`es_`]d]e]tu`itub_^`vs_`rguugw]dx`etyu_`
z{||}~z€z~€z{‚ƒz„~ |ƒ†€ ƒ†€~‡ {ˆƒ‰Š€z‹
>1,&?'!./ 0 &#/1,!&17!@&

Œ Ž‘’“
™

£
¦
”
š›œ
¡¢›Ÿ
œ›¤
¤›œ
•–—•˜
›žŸ
›¡œ
›¥Ÿ
›šœ
B
#'!&17!C &,!: /%,,#13
D
§~ |}¨ |Š‚€ƒ ©ª~ ˆ{«ƒ¬ƒ‰Š ©ƒ„¬€­®¯~
‚&
‚©°
‡!{ˆ
ƒ‰Š€¨z13
~€« ||€©ˆ€ˆ
­†€©Š‚}©©©°}©Š€©† ‚€« |Š‚€ƒ ©®z©«€ƒ«9~€}ƒ
€z
ƒ†€
|<&17'
©|{|#
ˆ}|}°€±²³´‹µ¶®
c_hba]dx`es_`fcghbae]i]ej`kdh_l`eg`·pq `gr`es_`bdht¸tx_h`itub_^
‡ {ˆƒ‰Š€¹z~€« ||€©ˆ€ˆ„}«}z€ˆ† ‚€« |Š‚
©zŠ‚C
}©©€ˆ‹
Š~ˆ{«€z
#€ƒ
'!&17!
&,!‡.#{ˆ
'¹
7!
+ &#/1,
}†°†‚‰|Š}~€ˆ„ ~|}ƒ ©ˆ}|}°€º ©€­†«††}zƒ†€†°†€zƒz»©}©ˆƒ†€‚ ­€zƒ
Š~ ˆ{«ƒ¬ƒ‰~}ƒ „}‚‚ƒ†€ˆ~‚‚©°| {ˆz‹¼ ­€¬€~®ƒz}‚z ƒ†€ƒ†©©€zƒ‹½†€Š‚}©©€ˆ
¾_crgcte]dx`h_¾es`¿Ào`]d^Á`]m`o^o`]d`xc_te_c`estd`es_`rgc¸te]gd`ht¸tx_`¿Â^p`]d^ÁÃ`s_da_`
ƒ†€Š€~„ ~}ƒ ©z­‚‚Ä€}Àƒ Ä‰Š}zzƒ†€„ ~|}ƒ ©ˆ}|}°€‹ÅŠ€~„ ~}ƒ ©Š‚}©
w]es`mbrÆa]_de`¾_crgcte]dx`h_dm]ej`tdh`mb]etyu_`¾stm]dx`msgbuh`y_`tyu_`eg`h_u]i_c`td`
{©ˆ}|}°€ˆ« |Š‚€ƒ ©±ÇÈÇ ³´¶‹
01234567 28609 0 2659 2 7 0882258002596 4660825600
23456789 747
!" !!# $" !%""&
"! "!'!'(
)*+,--./0+1*+2/3+456.7890+,*0+:);;<=*++>753?@8A5/+BC.728A5/90+D5-?E.+F+:G8H+I3A8A5/=*+
>?6-A9H.3+6J+BKLM*+MNOP+QR;<Q;SFRFQR)*
F*+I@5/5EA3.90+T *0+U A--0+, *+V *+2/3+IH-AWRI@5/5EA3.9+L*0+:);;<=*+>.875-.?E+>753?@8A5/+
NJ98.E9*++>?6-A9H.3+6J+>7./8A@.+U2--*++MNOP+QR)<RXYZXZ<R[*
<*+I@5/5EA3.90+T *+2/3+P5-8.0+\*0+:);Z;=*++4.9.7]5A7+N8AE?-28A5/*+>?6-A9H.3+6J+
01
Acidising and Other Matrix Treatments
CONTENTS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
1.1
12.
13.
14.
INTRODUCTION
WELL INFLOW AND STIMULATION
AN INTRODUCTION TO WELL STIMULATION
ECONOMICS
CANDIDATE SELECTION
4.1.
Treatment Timing
4.2.
Further Treatment Selection Criteria &
"The Stimulation Cycle"
SELECTION OF (CHEMICAL) TREATMENT
TYPE
POTENTIAL FORMATION DAMAGE
CAUSED BY MATRIX STIMULATION FLUIDS
MATRIX STIMULATION FLUID SELECTION
TYPICAL ACID FORMULATIONS USED FOR
MATRIX ACIDISING
8.1.
Hydrochloric Acid (HCl)
8.2.
Organic Acids
8.3.
Mud Acid
8.4.
Selection of Acid Composition
8.5.
Selection of Treatment Volume
8.6.
Selection of Injection Rate
8.7.
Selection of Additives
8.8.
Selection of Treatment Type
8.9.
Selection of Diversion Technique
MATRIX STIMULATION FIELD CAMPAIGNS
STIMULATIONOFCARBONATEFORMATIONS
10.1. Acid Composition Selection
10.2. Treatment Types for Carbonate Rock
Acidising
10.2.1. Matrix Treatments
10.2.2. Acid Wash or Acid Soak Type Treatments
ACIDISING OF SPECIAL WELL TYPES
11.1. Gravel Packed Wells
11.2. Horizontal Wells
11.3. Naturally Fractured Formations
ALTERNATIVE ACID FORMULATION
APPENDIX A
FURTHER READING
Revised 26/07/05
5
1
LEARNING OBJECTIVES:
Having worked through this chapter the student will be able to:
• Explain the importance of matrix stimulation in Production Engineering.
• Identify and contrast the application areas of the various types of matrix stimulation
techniques.
• Generically list the primary chemical reactions in sandstone and carbonate acidising.
• Explain the potential negative impacts of “matrix stimulation” and identify migration
strategies.
• Select acid formulation on the basis of source of formation damage and rock
composition.
• Identify and discuss the role of acid additives.
• Discuss placement and diversion techniques.
• Design a matrix acidising treatment (acid volume and injection rate).
2
Acidising and Other Matrix Treatments
5
1. INTRODUCTION
Chapter 4 described the many sources of formation damage which can lead to a
reduction in the permeability of the near wellbore area and creation of an extra,
positive skin, as measured by a well test. Matrix (stimulation) treatments are a
common form of well intervention aimed at removing this formation damage and
restoring the well to its natural, undamaged inflow performance. An alternative
stimulation technique - propped hydraulic fracturing - will be covered in a later
chapter. This latter well treatment can bypass this damage and/or increase the
effective wellbore radius. Either of these stimulation treatments may be carried out
immediately after drilling the well is completed or at any time in the well’s producing
lifetime when they can be economically justified.
The next section will discuss the parameters which control well inflow and describe
how this can be improved.
2. WELL INFLOW AND STIMULATION
The well known, steady state, radial-flow equation describes the well inflow:
Qo =
Where
Ko h (Pe - Pwf)
141.2µBo{In(re/rw)+S}
Qo
Ko
h
Pe
Pwf
µ
Bo
re
rw
S
=
=
=
=
=
=
=
=
=
=
(1)
well oil production rate (STB / day)
formation permeability to oil (mD)
reservoir thickness (ft)
reservoir pressure (psi)
wellbore flowing bottom hole pressure (psi)
oil viscosity (cp)
oil formation volume factor (reservoir bbl / STB)
well drainage radius (ft)
well radius (ft)
skin (dimensionless)
An increased well inflow (Q), or well stimulation, can be achieved by:
(i) increasing the factor {k.h} or
(ii) decreasing one of the factors S or re/rw or µ.
The frequently used stimulation techniques are catalogued in table1. This table also
identifies the parameter which this type of well treatment targets in order to increase
the well’s fluid inflow.
N.B. Mechanical and combined mechanical/chemical methods will be discussed in
the Hydraulic Fracturing chapter.
Department of Petroleum Engineering, Heriot-Watt University
3
1
Technique
Objective
Mechanical Methods
• Propped Hydraulic Fracturing
Increase rw
• Explosive Fracturing
• Underreaming
Increase rw and k
Increase rw
• Re - and Additional Perforating
Increase h
Chemical Methods
• Matrix Acidising
• Tubing Acid Washes
Decrease S
Improve Well Outflow by
• Other Chemical Matrix Treatments (Surfacants, Solvents, Mutual Solvents Etc.)
Biological Methods
• Microbial Stimulation
Removing Tubing Deposits
Increase k
Mechanism Uncertain
Combined Mechanical / Chemical Methods
• Acid - Fracturing Including Propped Acid - Fracturing
Increase re
• Closed Fracture Acidising
Increase re
Thermal Methods
• Steam Soak
Decrease µ
• Heat / Gas Generation From Injected Chemicals
Decrease µ and Improve Well
Outflow by Increasing GOR
• Electrical Heating
Decrease µ
Table 1
Available stimulation
techniques
The choice of which technique is the most appropriate for a particular well can be made
with the help of table 2.
Clastic Reservoirs
Treatment Type
Propped Hydraulic Fracture
Propped Hydraulic Fracture
Matrix*
Treatment Probably Not Required
Skin Permeability
Low
Low
High
High
Low
High
High
Low
* Frac and Pack Bypasses Formation Damage in Medium Permeability
Formations
Carbonate Reservoirs
Treatment Type
Propped Hydraulic Fracture
Acid Matrix Treatments
As For Clastic Reservoirs
Widen Natural Fractures
Acid Etch (Short) Channels
(Wormholes)
This chapter, entitled “Acidising and other Matrix Treatments”, discusses the chemical
methods of well stimulation. The common factor among these treatments is that they
are carried out under matrix conditions i.e. the injected fluids flow radially away from
the wellbore since the treatment fluid is injected into the well at rates and pressures
below that required for creation of a hydraulic fracture. This chapter mainly concentrates
on the injection of acid (“Acidising”), the most frequently employed of the chemical
treatments.
4
Table 2
Stimulation treatment
selection
Acidising and Other Matrix Treatments
5
As noted in Table 1, tubing washes are a related technique which share a similar
technology, in terms of fluid and additive selection, with matrix treatments. They are
used to improve the well outflow by removing deposits which have formed in the
tubing i.e. by increasing the (effective) tubing radius. The treatment fluid is normally
retained with in the tubing during the such a tubing wash treatment and is NOT
injected into the formation. This procedure avoids impairing the formation with any
undissolved particles which have become dispersed in the treatment fluid. Tubing
washes will not be discussed any further - but the requirements e.g. for corrosion
inhibition (discussed in section 5.8.7) if an acid wash is selected (e.g. because the
deposit in the tubing is acid soluble) is very similar to matrix acidising.
Matrix treatments, and acidising in particular, aim to remove the excess flowing
pressure drop (∆Pd) created by the presence of a volume rock which has suffered
formation damage (i.e. has a lower than original permeability) in the near wellbore
area (figure 1).
Idealised shape of
formation damage
(modeled below)
Formation (k)
undamage
Actual shape of
formation damage
kd
rw
rd
Zone With Formation Damage /
Reduced Permeability
Damaged
Zone
Wellbore
Centreline
Reservoir
kd
k
Pr
Ideal Pressure Profile
(Undamaged)
P2
Figure 1
The effect of a near
wellbore damaged zone on
the well inflow pressure
profile
Actual Pressure Profile
(Damaged) (kd < k)
∆ Pd
P3
rw
rd
∆Pd = Extra pressure drop due to Formation Damage
re
The removal of this formation damage will restore the “natural” well productivity.
The Hawkins formula:
Department of Petroleum Engineering, Heriot-Watt University
5
1
S=
where
S
K
Kd
rd
rw
r
K
-1 In d
rw
Kd
=
=
=
=
=
(2)
skin
formation (original or undamaged) permeability
(near wellbore) damaged formation permeability
radius of formation damage
wellbore radius
is a convenient tool for analysing the influence of varying levels and depths of
formation damage.
Matrix stimulation treatments increase well productivity by pumping a specially
formulated treatment fluid (frequently, but not always, an acid) which is designed to
remove (normally dissolve) the formation damage. However, the keys to successful
treatments are:
(1) the identification of a suitable candidate well which is capable of a greater
hydrocarbon production rate,
(2) the selection of the optimum type of treatment fluid for the removal of the
formation damage
(3) the design of the operational aspects of the treatment
such that the required economic criteria are successfully achieved.
The next section of this chapter will discuss some simple, economic concepts which
help identify whether a stimulation treatment could be economically viable.
3. AN INTRODUCTION TO WELL STIMULATION ECONOMICS
Well stimulation is only justified when the net (discounted) monetary benefit of the
resulting extra oil or gas production is greater than the cost of the stimulation
treatment. Previous field experience from the stimulation of similar wells is often a
good guide when predicting the expected gain from the well stimulation treatment. A
second simple, but approximate, method to estimate the potential benefits from a
stimulation treatment is to use the fact that the production wells typically show a
constant, long term, annual percentage decline in net hydrocarbon production. This
decline can be expressed in the form of a straight line when the logarithm of the net
hydrocarbon production is plotted against time (figure 2).
6
Acidising and Other Matrix Treatments
5
Expected production
rate gain due to
stimulation treatment
Net Oil
Production
Rate
('logarithmic'
scale)
Increased decline rate
of stimulated well
Actual oil production
rate of unstimulated well
Extrapolated long term
production decline rate of
unstimulated well
The stimulated well
production may revert to
original decline rate or even lower
Extra net oil produced
by stimulation
Figure 2
Improved oil production
from a stimulated well
Time
N. B. A straight line with a steeper slope in figure 2 corresponds to a greater, annual,
net hydrocarbon production decline rate. Note the extra oil reserves created by
producing the well above the economic limit for a longer period of time.
The expected, net hydrocarbon production gain from the stimulation treatment needs
to be estimated as the first step in carrying out the economic evaluation. This can be
estimated if:
1.
the well’s skin value is known (equation 1) or
2.
by use of the Hawkins formula (equation 2) if the extent and depth of formation
damage are known or can be guessed with reasonable confidence.
Field experience has shown that this gain in production will normally be followed by
an increased production decline rate. In time, the well’s production rate will often
revert to its predicted, original (unstimulated) value or even drop below this extrapolated
value. The latter occurs if the well’s reserves have undergone an accelerated depletion
resulting from the increased well production following the well stimulation.
Thus an estimate of the length of time that the well stimulation treatment will increase
the well production is also required. Remember that these prognoses must not only
consider:
(1) the well inflow i.e. whether the well has sufficient inflow capacity and
remaining reserves;
(2) the well (tubing) outflow capacity and
(3) whether the production facilities have sufficient capacity to process the extra
fluid volumes.
The above, or other techniques, allow the net hydrocarbon production gain from the
stimulation to be estimated. These volumes should be reduced by the:
Department of Petroleum Engineering, Heriot-Watt University
7
1
(i)
appropriate discount rate (since hydrocarbons produced today are more valuable
than later hydrocarbon production). N.B. Any reductions in facility capacity
while the spent stimulation fluids are being treated must also be included
in the economic evaluation.
(ii) hydrocarbon production lost while the well was taken off production to carry
out the well stimulation treatment,
(iii) expected chance that the stimulation will be successful (this is often significantly
lower than 100%).
The resulting increased revenue can be calculated from this discounted increase in net
hydrocarbon production multiplied by the net revenue per unit of production (sales
price minus marginal operating expenditure and taxes plus royalties). This has to be
compared with the cost of the stimulation which should include the:
(i)
cost of mobilisation and rental of equipment (pumps, tanks etc.) and personnel
employed for the well stimulation treatment. Also the cost of returning the well
to production e.g. initiate production by lifting with nitrogen gas. These costs
are related to the type of stimulation chosen and the stimulation treatment size.
(ii) cost of consumables e.g. chemicals etc. used for the well stimulation treatment.
This cost is related to the size of the stimulation treatment, while the earlier ones
are related to the type of stimulation treatment chosen.
From the point of view of stimulation candidate well selection, the well stimulation
treatment yielding the highest prognosed (discounted) rate of return is the treatment
which, in priciple, should be carried out first.
A somewhat simpler calculation method is to calculate the payback time i.e. the
production time required for the increased, net hydrocarbon production to pay back
the costs of the well stimulation treatment. The most profitable well stimulation
candidate is the stimulation treatment yielding the most rapid pay back. Most
production companies require a very high rate of return from this type of well
treatment, leading to pay back times of between 6 and 12 months from this type of well
treatment.
N.B. These economic concepts are treated in greater detail in the economics module
of this Petroleum Engineering course.
4. CANDIDATE SELECTION
The selection of stimulation candidates that potentially meet the economic screening
criteria discussed in the previous section is the key to a successful stimulation
campaign. This involves two stages:
(i)
8
the identification and (accurate) quantification of those parameters which
control the productivity of the specific well and
Acidising and Other Matrix Treatments
5
(ii) an analysis to determine whether the well stimulation treatment would actually
improve the well production.
As mentioned earlier, it is important here to distinguish between well (inflow)
productivity and well production. This is because an improvement in the well
“inflow” performance (from the reservoir to the well) can have only a limited effect
on the daily well production if the well “outflow” is limited by tubing / artificial lift
/ facilities restrictions. The emphasise the need to consider the flow path from the
reservoir boundary to the production storage facilities as a complete production
system.
Gas
Pwellhead
Pseparator
GAS
Choke
Skin
(Zone of
damaged
permeability)
Oil to Tank
Flowing bottom hole pressure
P1 = -ve Skin
P2 = Zero Skin
P3 = +ve Skin
Figure 3
The Producing System
re
Reservoir
Reservoir
Permeability
(K)
Well Boundary
Kd
Pr
P1
P2
P3
Table 3 summarises typical values for the minimum requirements for a successful
(matrix) stimulation treatment. The criteria in this table can be used for the preliminary
screening of well stimulation candidates. These figures are based on experience from
a number of fields. They will be modified when studying specific conditions
pertaining to a given field.
Parameter
Hydrocarbon Saturation
Water Cut
Table 3
Minimum Screening
Matrix Treatment
Candidate Well Selection
Criteria
Permeability †
Reservoir Pressure
Production System
*
ø
†
**
Oil Reservoir
Gas Reservoir
>40%
<30%
>50%
<200 bbls/MMscf**
>20 mD *
<70% depleted
>1 mD
twice abandonment pressure
20% spare capacity ø
higher value required for viscous oil production
lower value if several wells are manifolded together
wells with lower permeabilities are potential hydraulic fracturing candidates
evaluate potential for well killing itself due to liquid loading in the tubing
Department of Petroleum Engineering, Heriot-Watt University
9
1
The above criteria evaluate whether the well has a sufficient minimum of:
(i) remaining reserves to justify carrying out the remedial stimulation technique,
(ii) well inflow productivity and
(iii) capacity in the facilities to process the extra fluid production.
The well skin (S) is not mentioned in the table 3 criteria - this is because the skin can
not be considered apart from the other parameters that control the well inflow {the
reservoir permeability thickness (k.h) and the potential well drawdown (Pe - Pwf)}.
Thus, a small reduction in skin value from a well with a large reservoir permeability
thickness can yield a much larger production increase than removal of a high skin
value from a well with a small reservoir permeability thickness. Alternatively,
installation of an artificial lift method which allows an increased drawdown or
re-perforation of the producing interval may be the more effective methods of
increasing production.
Each candidate well needs to be evaluated on its own merits.
4.1. Treatment Timing
Well stimulations may be carried out immediately after the initial drilling/completion
programme has been finalised e.g. to correct formation permeability impairment
caused by the drilling mud. The well would normally meet the criteria set out in table 3.
Alternatively, the stimulation candidate may be identified as a result of routine, field
production surveillance e.g. the well is identified as producing less than the surrounding
wells with comparable reservoir quality or reservoir permeability thickness (kh).
Figure 4 schematically illustrates the typical output from a modern, production
surveillance, computer package. This type of graphical display helps with the easy
recognition of potential well stimulation candidates. Once a particular well has been
identified, its attributes must be checked against the criteria in table 2.
Well location on reservoir map
Bubble
Possible
Stimulation
Candidate
Key:
Low
Medium
High
Reservoir Premeabilty Thickness (kh)
Low
Medium
High
Relative Bubble Size Reflects Cumulative Well Production
10
Figure 4
Visualising field production
data using Bubble Maps
Acidising and Other Matrix Treatments
5
4.2. Further Treatment Selection Criteria & "The Stimulation Cycle"
Further selection criteria which should be considered include:
(i)
Is Sdam>30% of Stotal (the total well skin ) ?
i.e. could other inflow improving measures (e.g. reperforation ) be a more
economical approach to increasing well production? Table 4 gives examples
of poor well productivity which can not be “stimulated away”.
Examples of Possible Poor Well Productivity that can not be
'Stimulated away'
Observation
Table 4
Stimulation Treatments
Selection
Possible cause
1. Gas well with > 200 bbls liquid/MMscf
2. Three phase production
poor tubing
outflow
1. Gas well - high drawdown
2. Oil well - > 20 bbl/ft
non-Darcy
(turbulence)
effects
1. Oil wells > 5 bbl/day/perf
2. Partial penetration resevoir
Completion
Geometrical
skin
(ii) Does the well show sand production?
Are sand control measures in place? (matrix stimulation treatments of gravel
packed completions have historically shown a lower success rate than when
perforated completions are treated).
(iii) Is the cause of formation damage known (or at least suspected)?
Identification of the cause of the formation damage greatly increases the chance
of matrix treatment success since a treatment fluid which efficiently removes
that specific form of formation damage can be selected.
(iv) Is the stimulation feasible?
The final stage of stimulation candidate selection is to evaluate the practical
aspects of the stimulation (e.g. what is the mechanical condition of the well?
Are there any logistical, scheduling, or other overriding considerations
which prevent the well being taken out of production?).
Once the above questions have been answered the following choices can be made:
(i)
The composition of the pre and post-flushes and any additives are determined
and the volume of all flushes chosen. Remember that the post flush has to
be displaced to the perforations by a compatible brine or hydrocarbon fluid.
(ii) The detailed treatment design (including injection rates and pressures) can now
be made and the strategy for returning the well to production chosen.
(iii) The treatment is now carried out and, eventually, evaluated.
The complete stimulation cycle described above is captured in Figure 5.
Department of Petroleum Engineering, Heriot-Watt University
11
1
Identify
Potential
Stimulation
Candidate
Modify
Stimulation
Guidlines
Evaluate
Stimulation
Economics
Evaluate
Treatment
Success
Return
Well to
Production
Matrix
Stimulation
Suitable?
The Stimulation Cycle
Identify
Formation
Damage
(type + location)
TREATMENT
EXECUTION
Select
Treatment Fluid
Compositions,
Additives + Volumes
Site and Job
Preparation
Sc
hed
lo g
is t i
u l in
cs
g
Operational
Stimulation
Program
(including well
clean-up strategy)
Operational
constraints
Select Treatment
Type, Diversion
Techniques and
Specify Injection
Rates and Pressure
Although well stimulation is often a high reward well activity, it must be reiterated
here that, for both clastic formations and many carbonate reservoirs:
• Prevention of formation damage is nearly always better than the cure (a remedial
well stimulation).
• Acidisation and other stimulation techniques can create formation damage (see
section 5.6).
5 SELECTION OF (CHEMICAL) TREATMENT TYPE
The chosen chemical treatment fluid should be targeted at the particular type and
location of the formation damage to be removed or treated. As discussed in chapter
4 (Formation Damage), the formation damage/impairment may be related to:
(i)
drilling, completion or workover operations,
(ii) produced or (continually) injected fluids,
(iii) injected fluids during specific well operations e.g. well killing.
Table 5 can be used to select the optimum type of chemical treatment once the type
and location of the formation damage/impairment has been identified.
12
Figure 5
The Stimulation Cycle
Acidising and Other Matrix Treatments
5
yp
Treatment Type
Objective
Location
Comments
Acid (inorganic) and
solvent (organic)
washes
Remove flow restrictions
eg. inorganic / wax scale,
kill pills, etc
Tubing, perforation
and completion
Circulated in wellnot injected into
formation
Matrix acidising
Dissolve rock components
(drill solids, precipitates, clays etc)
and formation damage to
improve well / formation
connectivity
Near wellbore
(< 1m depth from
sand face)
Other matrix
treatments
Solvents and surfactants remove
emulsions, wax etc. from oily wells
and inhibitor residues from gas wells.
Near wellbore
(< 1m)
Chemical inhibitor squeezes
prevent scaling etc. during
well production
at depth (< 10m)
from wellbore
Table 5
Chemical treatment types
Large volume treatment
Inhibitor returns with
produced fluid
6. POTENTIAL FORMATION DAMAGE CAUSED BY MATRIX
STIMULATION FLUIDS
The reaction of the formation rock / insitu (formation/injected) fluids with an
incorrectly chosen stimulation fluid may generate further formation damage/
impairment. Such sources of formation damage include:
(i)
deconsolidation of the rock matrix due to the acid dissolving the cementing
material that holds the sand grains together. (Temporary) sand production is
often observed when a well is returned to production after and acid stimulation.
(ii) generation of migrating, small diameter particles (“fines”) which can block
the pore throats. These particles result from the acid only partly dissolving
the formation minerals present between the grains. This allows insoluble,
small diameter, particles to be created and injected into the formation pore
throats where bridging and blockage can occur.
(iii) the reaction products created by the chemical reaction between the acid and
the formation rock can be insoluble in the spent stimulation fluid. This is
called secondary precipitation. This precipitation process leads to blockage
of the pores and pore throats (impairment). This precipitation often does not
occur immediately - this implies that the options are to either:
(a) immediately produce the (spent) acid (i.e. return the well to production)or
(b) inject the (spent) acid deep into the formation where any precipitation
will have limited effect on the well productivity.
(iv) fluid incompatibilities. A matrix acidising treatment consists of sequentially
injecting a series of fluids. It must be checked that these fluids are compatible
with each other and with the formation fluids i.e. to not form a (solid) precipitate
at the prevailing, downhole temperature when mixed in any proportion.
Department of Petroleum Engineering, Heriot-Watt University
13
1
Further, for the acid fluids which react with the formation; both the “fresh”
(unreacted) and the “spent” (reacted) acids need to be tested in this way.
(v) acid precipitation of an insoluble sludge when mixed with the crude oil. This
is particularly true for asphaltenic crudes and for acids containing ferric cations
(rust - or ferric oxide - is a corrosion product produced by steel surfaces e.g. the
tubing internals which is dissolved in the acid as it is injected into the well).
Such sludge precipitation can be avoided by injecting a compatible hydrocar
bon based preflush to displace the crude oil away from the wellbore and the
following acid.
(vi) surfactants. Surfactants added to the treatment fluids may create a (highly)
viscous emulsion with the crude oil leading to blockage of the pores by this low
mobility fluid. Prescreening laboratory tests in which a sample of the treatment
fluid and the crude oil are shaken together and then the mixture examined for
emulsion formation. A series of such trials can be used to select a suitable, nonemulsifying, surfactant.
(vii) wettability changes. Surfactants can change the wettability of the pore surfaces.
An oil wet formation has a lower permeability to oil than the equivalent water
wet formation. Once again, any surfactants that are planned to be used should
be tested as described above.
(viii)“water blocks”. Significant volumes of water are injected into the formation
during the stimulation treatment. This may lead to an increased water
saturation (water block) in the near wellbore area and can take a long time to
disappear (months or even years) from low permeability formations. This can
be minimised by the addition of gas to the stimulation fluid or the (partial)
replacement of water by more volatile solvents.
All these forms of formation damage were previously discussed. Non-acid, matrix
stimulation treatments also show these effects - apart from (i) and (iii). Surfactants are
capable of mobilising loosely bound, small particles (such as clay particles or other
“fines”) which can then reduce the permeability by blocking the pore throats, a source
of formation damage similar to (ii).
It is clear from above that formulation of the acid and the other fluid flushes employed
is the key to ensuring that formation damage due to the stimulation treatment does not
occur or, if it is unavoidable, is at least minimised.
7. MATRIX STIMULATION FLUID SELECTION
The mineralogy and chemistry of the formation, together with the chemistry of the
formation fluid and that of the formation damage, combine to give the stimulation
fluid selection criteria. This can be summarised as a balance between the:
(i)
14
positive effects e.g. solubility of the formation and formation damage in the
selected fluid, and the
Acidising and Other Matrix Treatments
5
(ii) negative effects e.g. deconsolidation, “fines” generation, secondary precipitates etc.
The type of formation damage present in the well has a large influence on this
selection. Acid is not always the most appropriate fluid - this can be seen from tables
A-1 and A-2 at the end of this chapter which give examples of fluids for various types
of formation damage.
Another factor which needs to be taken into account when designing the stimulation
treatment is the formation and well temperatures both during and after the well
treatment. The speed of reaction of an acid with the formation rock and / or formation
damage will be greater at elevated temperatures (typically doubling for every 10˚C
temperature rise). Acidic fluids are highly corrosive to the steel surfaces that make
up the completion (tubing / casing / packers etc.). The type and concentration of
corrosion inhibitor required to inhibit (i.e. limit) this corrosion reaction depends on the
treatment temperature and the treatment time during which the treatment fluids are
pumped.
Type and Location of Formation Damage
Inorganic scale
in perforation
Organic or Inorganic
scale in tubing
Emulsion or
Water Block
/Wettability
Change
Fractures
Plugged with
drilling mud
or cement
Figure 6
Type and location of
formation damage for cased
and openhole completions
Drilling mud
filtration
Mud Filter
Cake
Open Hole
Completion
Types
Cased, Cemented,
and Perforated
The types and location of typical forms of formation damage are depicted in figure 6.
This, together with knowledge of the completion (perforated) length and associated
formation inhomogeneities (variation in permeability and pressures across the complete
length), will determine if special “diversion” arrangements need to be made. “Diversion”
techniques ensure that the treatment fluids are evenly “placed” across the formation
i.e. that at least the required minimum volume of fluid is injected into each perforation
open to flow.
8. TYPICAL ACID FORMULATIONS USED FOR MATRIX ACIDISING
The behaviour and chemistry of the three most frequently used acids for well
stimulation treatments are discussed in the following sections.
Department of Petroleum Engineering, Heriot-Watt University
15
1
8.1. Hydrochloric Acid (HCl)
Hydrochloric acid is widely available commercially at concentrations up to 28% wt.
Its main reaction is to dissolve carbonate minerals (or scale) present in the formation
and the well itself e.g. calcite (chalk or limestone), dolomite, siderite etc. The amount
of the mineral dissolved is a function of the volume and concentration of acid used:
2HCl + CaCO3→ CaCl2 + CO2 + H2O
or 1m3 of 15% wt HCl dissolves 220kg or 0.09m3 of limestone.
Hydrochloric acid is also capable of dissolving chlorite (an iron containing clay).
Hydrochloric acid only shows secondary precipitation reactions when it reacts with
iron containing minerals. Highly impairing iron hydroxide is precipitated if the acid
become spent (i.e. the pH reduces towards the neutral value of 7, see section 5.8.7 Sequestering Agent).
It is essential to inhibit the acid with a corrosion inhibitor since hydrochloric acid is
highly corrosive to the conventional mild steels used in many well completions.
Corrosion inhibition is even more difficult when treating wells completed with (high
cost) 13 %wt chromium (stainless steel) alloys or one of the special duplex steels. It
must also be remembered that hydrochloric acid will dissolve any rust (ferric oxide)
present on the tubing wall - even in the presence of a corrosion inhibitor. The potential
negative effects of ferric cations were discussed in section 5.6.
8.2. Organic Acids
Acetic acid (CH3COOH) is sometimes employed when it is desirable to use a weaker
or slower reacting acid than hydrochloric acid e.g. the reaction with dolomite is:
4CH3COOH +Mg.Ca(CO3)2 → Mg(CH3COO)2 + Ca(CH3COO)2 + 2H2O + 2CO2
Calcium acetate has only a limited solubility - this means that 15%wt acetic acid is the
maximum concentration that should be used. Acetic acid has two major advantages:
(i)
It is non corrosive to aluminium and (chrome) steel alloys at temperatures
below 90oC. Therefore a corrosion inhibitor is not required for these lower
temperature treatments.
(ii) It retains ferric iron in solution as the acid is neutralised (“spends”) by reaction
with the formation. The chemical name for this effect is sequestration - it
prevents the formation of ferric hydroxide precipitates from the depleted acid.
This avoids the highly permeability impairing form of formation damage
discussed above, that can occur with hydrochloric acid based stimulation fluids.
Formic acid (HCOOH) behaves in a similar manner to acetic acid, but is even weaker
(more slowly reacting) than acetic acid.
Organic acids are often used to replace hydrochloric acid for treatments of higher
temperature wells e.g. for formations with a temperature greater than 120oC. Many of
the available corrosion inhibitors are less effective at this temperature. Organic acids
still require a corrosion inhibitor at this temperature, but the corrosivity is less than that
of a hydrochloric acid based of the acid at the same temperature.
16
Acidising and Other Matrix Treatments
5
8.3. Mud Acid
The majority of acid stimulations of clastic reservoirs are carried out with “Mud
Acid”. This acid is a mixture of hydrochloric (HCl) and Hydrofluoric (HF) acids. This
very aggressive acid is capable of dissolving minerals such as quartz, clays, micas etc.
(hence the name “Mud Acid”). These minerals are inert to hydrochloric acid alone.
NB. The aggressive nature of mud acid also means that it is a major safety hazard to
the wellsite personnel involved in the treatment. Exposure to the acid results in almost
instantaneous blistering of the skin and permanent scarring of the cornea (leading to
sight loss).
Mud Acid is made by adding the appropriate amount of solid ammonium bifluoride
to the hydrochloric acid solution. Typical formulations used in the field are “Full
Strength Mud Acid” (12%wt HCl and 3%wt HF) and “Half Strength Mud Acid”
(6%wt HCl and 1.5%wt HF). Acid formulations with lower fluoride concentrations
and higher chloride : fluoride ratio {e.g. 0.5% wt HF and 6% wt HCl} have become
more popular in recent years, as discussed at the bottom of this section in the
paragraphs concerning reprecipitation reactions.
Virtually all the chloride salts formed by the reaction of the formation (damage) with
hydrochloric acid have a high solubility. In contrast, some fluoride salts {the simple
fluorides (F-) or the fluoro silicates (SiF6- -)} of sodium (Na), potassium (K), calcium
(Ca) etc. are very insoluble e.g.
CaCO3 + 2HF → CaF2 ↓ + CO2 + H2O
SiO2 + 6HF → H2SiF6 + 2H2O
H2SiF6 + 2K+ → K2SiF6 ↓ + 2H+
The insolubility of these salts implies that mud acid should:
(i)
never be diluted with sea water (since it contains calcium and sodium ions).
The hydrochloric acid, which forms the basis of the mud acid, is normally
delivered to the wellsite in a concentrated form and diluted on site to the
designated concentration.
(ii) never be used to acidise a carbonate formation (they contain calcium),
(iii) always be used with a preflush of hydrochloric acid. This preflush should be
of sufficient size to dissolve these cations so that they are removed before the
formation is contacted by the mud acid. Typically, the volume of the preflush
is half that of the mud acid flush. For convenience, the preflush normally uses
the same hydrochloric acid concentration as the main, mud acid flush.
(iv) always be overflushed with a dilute (3%wt or less) solution of hydrochloric
acid (HCl) or ammonium chloride (NH4Cl) (ammonium salts have a high
solubility).
The clays, micas etc that are dissolved by the mud acid undergo a series of reactions
that result in precipitation of silica gel (Si(OH4) - a hydrated form of silica). This
reaction can be described in simplified form as:
Department of Petroleum Engineering, Heriot-Watt University
17
1
slow
HF + SiO2 (silica or quartz) → H2SiF6 (dissolved)
fast
HF + clay → Si, Al in solution
slow
H2SiF6 + clay → Al (dissolved) + Si(OH)4 ↓
NB: SiO2 represents quartz or other forms of silica; while Al and Si represents
aluminium and silicon respectively. They are the constituents of clays and micas.
These reprecipitation reactions can not be avoided but their effect can be minimised:
(i)
For some formations, an increase in the HCl : HF ratio of the mud acid
formulation will reduce the amount of (re) precipitation.
(ii) Since the (re)precipitation reaction occurs slowly, its damaging effects can be
avoided by producing the spent acid back rapidly (returning the well to
production immediately after the treatment has been finished) or overflushing
the mud acid deep into the formation where the effects of the reprecipitation is
minimised.
This latter option and the effects of treatment size are illustrated in figure 7.
1 The Damaged Well
kd
rd
k
kd
k
rd
Damaged Zone Radius
Virgin Rock Permeabilty
Damaged Zone Permeabilty
k > kd
2 A Small Acid Treatment
kacid kprecip
kd
k
rd
kacid
Fully Acidised Zone, All
Soluble Minerals Dissolved
kprecip
Partically Acidised Zone
with Precipitated Acid / Rock
Reaction Products
3 A Large Acid Treatment
kacid
kprecip
k
kprecip < k > kacid
rd
4 A Large Acid Treatment Followed by an Overflush
kacid
18
k
kprecip
k
Figure 7
Permeability changes
during a mud acid
treatment
Acidising and Other Matrix Treatments
5
8.4. Selection of Acid Composition
Impairment
NATIVE
ROCK
Cement (carbonate)
Pore lining
clay
Quartz
x x xxxxx
x xxx
xx
xx x x
AC
ID
Pore filling
clay
(H
CL
/H
F)
Increased porosity
and permeability and
impairment removal
Remaining
pore space
Secondary
reaction
products
Figure 8
A mud acid treatment
ACIDISED
ROCK
The chemistry of a mud acid treatment is pictured in figure 8. It illustrates how the
impairment, formation clays and inter-granular cements are removed by the mud acid
and partially replaced by secondary reaction products. However, there is an overall
increase in porosity and permeability, leading to stimulation of the well.
Formation Characteristics
Acid Recommendation
Comment
HCl solubility > 15%
HCl only
HF causes CaF2 precipitation.
Organic acids or EDTA/ surfactant
treatments an option.
10% < HCl solubility > 15%
Increase HCl preflush
volume
Ensure all carbonate minerals dissolved
prior to mud acid
High quartz (80%); low clay (5%)
12% HCl - 3% HF
Regular mud acid
High feld spar/illite (> 15%)
13.5% HCl - 1.5% HF
Avoid fluosilicate precipitation
High clay (> 10%)
6.5% HCl - 1% HF
Reduce potential migratory fines due
to partial dissolution
High (Fe) chlorite clay
3% HCl - 0.5% HF
Avoid iron dissolution. HCl preflush with
Fe sequestrant an option
High quartz (80%); low clay (5%)
6% HCl - 1.5%HF
Reduce migrating fines potential
High feld spar/illite
6% HCl - 0.5% HF
Avoid fluosilicate precipitation
High clay (> 10%)
6% HCl - 0.5% HF
Reduce migrating fines potential
High (Fe) chlorite clay
3% HCl - 0.5% HF
Avoid iron dissolution. HCl preflush with
Fe sequestrant an option
HIGH PERMEABILITY (> 100 MD)
LOW PERMEABILITY (< 100 MD)
Table 6
Acid selection guidelines
Replace at least half of the HCl by HCOOH in the region 120…C - 170…C bottom hole temperature
Replace all the HCl by HCOOH for temperatures greater than 170…C
NB 1% wt HCl = 1.3% wt HCOOH
Department of Petroleum Engineering, Heriot-Watt University
19
1
The ideas presented in the previous section, combined with laboratory core flooding
and field experience, have shown the need to adjust the mud acid formulation to the
type of formation mineralogy and damage that is being treated. This experience is
consolidated in table 6. Salient points of note are:
(i)
Formations with a high (>10%wt), solubility in hydrochloric acid require a
larger (hydrochloric acid) preflush. Mud acid should not be used once this
solubility level increases to 15%wt or higher.
(ii) Higher permeability formations with a low clay content can be stimulated with
fluids containing a higher HF concentration. Increasing clay contents and the
presence of certain minerals require lower HF acid concentrations and higher
HCl : HF ratios.
(iii) A similar approach is followed for lower permeability formations - except that
lower acid concentrations are employed even for the “cleanest” (low clay
content) formations.
This approach can be made even more specific to the particular formation by using a
chemical thermodynamic simulator to calculate the amounts of rock minerals dissolved
and secondary precipitates formed when a given volume of acid is injected into a
specified formation volume. Coupling of the thermodynamic simulator to a simplified
reservoir simulation of the injection process allows the effect of the injection of further
quantities of fresh acid into the partially acidised formation to be simulated. Changes
in formation porosity can then be equated with changes in permeability using the
Kozeny Carmen relationship and the increase in production due to the acid treatment
be calculated. This process can be carried out for many acid formulations - allowing
the identification of the optimum mud acid formulation of 10%wt HCL and 0.7%wt
HF for that particular formation mineralogy (Figure 9).
Contours With Constant Production Increase (Q acid / Q original)
5
Insufficient HCI
4
Concentration HF (% Wt)
Acidized Porosity
3
φ < 35 %
φ > 40 %
1.45
Optimum
Acid
Formations
(10% wt HCl
0.7% wt HF)
2
1
Composition of Rock
35 % < φ < 40 %
Quartz
74.9 % wt
K-Feldspar 7.0 % wt
Illite
6.0 % wt
Kaolinite
10.0 % wt
Dolomite
0.1 % wt
Anhydrite
0.5 % wt
Albite
1.0 % wt
Siderite
0.5 % wt
Porosity (%) : 20
No Drilling Impairment
1.45
1.40
1.35
0
0
5
10
15
20
25
Concentration HCI (% Wt)
This Figure illustrates many of the key aspects of acid formulation selection:
20
Figure 9
Selection of optimum acid
formulation
Acidising and Other Matrix Treatments
(i)
5
obtain the highest possible Production Increase (Qacid / Qoriginal) consistent
with:
(ii) minimisation of the increase in porosity (i.e. chance of sand production by
deconsolidation of the formation).
It can be seen from figure 9 that this calculation did not include any formation damage.
An alternative formulation, 9%wt HCL and 1%wt HF, also meets the above criteria.
This second formulation has a higher HF acid content, which would be capable of
removing greater amounts of, for example, clays from drilling fluids than the 0.7%wt
HF formulation. Both these formulations have a high HCL : HF ratio, which is suitable
for treating formations with a significant clay content.
8.5. Selection of Treatment Volume
The next stage in treatment design is the selection of the treatment volume. This can
be based on:
Formation Temperature
Permeability
Table 7
Typical Mud Acid treatment
volume guidelines
(i)
<150ºF
150-250ºF
>250ºF
Volume of mud acid (US gal/ft perforations)
< 20 mD
100
50
50
20-100 mD
150
100
100
>100 mD
200
150
100
Field experience when treating wells in the same or similar fields. Often
between 50% and 100% of the volumes suggested in Table 7 are used.
(ii) Practical considerations such as logistics e.g.
(a)
how much acid can be delivered to the wellsite? or
(b)
how large an acid the volume pumpable during daylight hours?
(iii) Economics (how much acid can we afford based on the expected gain in
hydrocarbon production ?)
(iv) Laboratory core flow testing (acid volume required to increase the permeability
by a target amount). The core may be pre-treated to include damage to the
core inlet face by the suspected form of formation damage.
Laboratory testing in support of matrix stimulation campaigns can also include:
(a)
Identification of the formation mineralogy by use of thin sections,
X-ray diffraction or Scanning Electron Microscope studies.
(b)
Petrographic analysis of formation samples yields information on the
type and location of the minerals, porosity, cementation and clays
Department of Petroleum Engineering, Heriot-Watt University
21
1
(c)
Fluid-Fluid compatibility tests can be carried out using API RP
(Recommended Practice) 42 test methodology. Carrying out these tests
will help ensure that all the fluids used are compatible with each other
and the insitu crude oil. Typical tests that can be carried out include
checks to see if:
(i)
a sludge is formed by the crude oil* on contacting with acid,
(ii)
a “fines” stabilised acid/crude oil* emulsion can be formed,
*these two tests require a fresh (non-aged) crude oil sample.
(iii)
a viscous oil based mud / acid emulsion is formed.
(v) Theoretical calculation based on the formation mineralogy and formation
damage as used in Figure 9 and described at the end of section 5.8.3. This
technique has been applied to a well in which formation damage was present
to a depth of 30 cm. The results summarised in Figure 10. This Figure indicates
that an acid volume of 16 US gal/ft of perforations is required to achieve the
optimum Productivity Index (PI) for both unimpaired (no formation damage)
and impaired (formation damage present) wells. However, this calculation
assumes perfect placement of the acid (see section 5.8.8). In practice, a larger
acid volume has to be injected to ensure that each perforation receives the
required treatment volume of 16 US gal/ft of perforations.
1.6
1.4
PI Acidised / PI Unimpaired
1.2
Unimpaired Well
Increasing
Well Impairment
1.0
0.8
0.6
0.4
0.2
Steep
Increase
Productivity Index (PI) Similar
for Impaired and Unimpaired Well
0
3
(m /m) 0.0
(US Gal/ft.) 0
0.2
16
0.4
32
0.6
48
0.8
64
1.0
80
1.2
96
1.4
112
1.6
128
1.8
144
Volume of Acid
Notes:
(i)
Preflush: 40% of mud acid volume rising to 100% as carbonate content
increases.
(ii) Main flush: mud acid volume between 50% and 100% of Table 7 values.
22
Figure 10
Selection of optimum acid
volume
Acidising and Other Matrix Treatments
5
(iii) Post flush: 10% of mud acid volume if production resumed immediately
increasing to 100%-200% if production can not be resumed within 4 hours.
8.6. Selection of Injection Rate
Matrix treatment fluids have to be injected below the Fracture Propagation Pressure
(FPP) (this term is explained in Chapter 6 on Hydraulic Fracturing) to ensure that the
fluids are injected radially from the wellbore through the matrix. The maximum
allowable injection rate can be calculated from the equation:
141.2*10-6 Kav.h (FG.d - ∆ps - pe)
qmax =
µ {In (rf / rw) + S}
qmax
h
d
µ
pe
rw
Kav
FG
∆ps
rf
S
=
=
=
=
=
=
=
=
=
=
=
maximum injection rate (bpm)
net treated height (ft)
well depth (ft)
viscosity of injected fluid (cP)
reservoir pressure (psi)
wellbore radius (ft)
(average) undamaged permeability (mD)
fracture gradient (psi/ft)
safety margin* (500 psi)
radius of injected fluid (ft)
skin factor
* this saftey margin should be larger if the fracture gradient is not well known.
This version of the radial inflow equation assumes the injected and reservoir fluids do
not differ greatly in viscosity and ignores transient flow effects. This equation allows
the maximum injection rate to be calculated for a constant bottom hole pressure as a
function of skin and injection rate. The Bottom Hole Flowing Pressures can be
translated to Wellhead Pressure by estimation of the hydrostatic head due to the
density of the fluid in the tubing and any frictional pressure drop. This requires
knowledge of the treatment flow rate & fluid density, depth of the top perforation and
internal diameter roughness of the tubing etc. It can be used to prepare a graph of
Wellhead Pressure against the treatment injection rate for a number of different values
of the well skin can then be prepared Figure 11.
N.B. only injection rates below the Fracture Propagation Pressure need to be
considered.
Department of Petroleum Engineering, Heriot-Watt University
23
1
Well Head Pressure (psi)
2000
ressure
ogation P
rate
Fracture Prop
mp
le pu
b
a
w
lo
um al
Maxim
}
300 psi
safety margin
1500
2
3
S=340 S=70
1000
5
4
S=30
7
6
9
S=15
8
S=7
S=3
1
500
S=0
10
0
0
2
4
6
8
10
Treatment Injection Rate (bpm)
= Pressure and Pumprate measurements taken at various times
kw = 75 mD
ρ = 8.92 lb/gal
h = 154 ft
FG = 0.74 psi/ft
rw = 0.35 ft
d = 7490 ft
Pe = 3020 psi
Tubing I.D. = 4.56 in
This type of plot was first suggested by Paccaloni. He advocated plotting the treatment
pump rate against wellhead pressure so that the reduction in well skin value could be
monitored in real time as the treatment proceeded. The pump rate could also be
continually maximised (which he believed to improve diversion - see section 5.8.9).
Figure 11 is used to estimate the current skin from the pump rate / wellhead pressure
data over the wide range of pump rates experienced during the acid treatment. For this
particular example there is little change in the apparent skin value for the last three
measurements (points 8, 9 and 10). Paccaloni further proposed that the treatment
should be stopped once it was observed that the apparent skin was no longer
decreasing.
This Paccaloni approach to injection rate control can not be used if the reservoir is
depleted and the wellhead pressure is zero during the treatment (“the well goes on a
vacuum”); unless some other means of real-time, bottom hole pressure measurement
is available.
8.7. Selection of Additives
A range of additives to the treatment formulation have been developed to combat one
or more of the forms of formation damage associated with stimulation treatments.
They can be expensive, especially those added to acid treatments (e.g. the cost of the
corrosion inhibitor required when acidising a high temperature well can often be
greater than that of the acid). Further, many of the additives are incompatible with
each other and may themselves cause formation damage. The use of each additive has
to be justified separately - it should not be just chosen because of the claimed
advantages in the service company sales catalogue!
The more important, frequently used additives are summarised below:
24
Figure 11
Example of Paccaloni plot
for control of treatment
injection rate
Acidising and Other Matrix Treatments
5
(i) Corrosion Inhibitors
This additive type is almost always required for acid treatments due to the corroding
reaction of acid on steel:
Fe + 2H+ → Fe++ (dissolved)+H2(gas)
The (acid) corrosion rate increases rapidly with increase in temperature. Further, the
corrosion inhibitor looses its effectiveness as the temperature and treatment time
increase (due to degredation). A field proven “rule-of-thumb” is that a maximum
weight loss of 0.05 lb/ft2 of tubing area (equivalent to the removal of 0.001 in of the
tubing wall thickness) during the treatment duration is acceptable . This implies that
the allowable corrosion rate decreases as the acid treatment time increases. Further,
the corrosion should be in the form of a general weight loss type rather than pitting or
stress corrosion. The above allows a specification to be developed for the corrosion
inhibition of the acid. The type and concentration of corrosion inhibitor chosen will
depend on the acid type, bottom hole temperature, the type of steel contacted and the
expected treatment duration.
N.B. The “spent” acid produced back after the treatment when the well is returned to
production is often highly acidic and will corrode the tubing at a similar rate to the
fresh acid. This is because the corrosion inhibitor has been depleted by reaction with
the acid and absorption on the formation. Injection of extra corrosion inhibitor may
be considered during this phase if, for example, the well is being returned to
production by nitrogen lifting with a coiled tubing unit.
(ii) Sequestering Agents
Both the ferrous (Fe++) and ferric (Fe+++) forms of iron will precipitate as the acid
“spends” (pH increases). They both form an amorphous, high volume iron hydroxide
precipitate which is highly efficient at creating formation damage. Fe+++ is by far the
most insoluble form (see Table 8). Ferric hydroxide already has a low solubility at pH
values greater than 2, though this value is increased to 6 when mud acid is being used
since sequestration takes place due to the presence of the fluoride ion (F ). By contrast,
ferrous hydroxide precipitation is delayed until the pH rises to values greater than 6.
Table 8
Solubility of ferrous and
ferric ions as a function of
pH
pH
Ferrous Iron
1
sol
60000 ppm
5
2
sol
60 ppm
6
3
sol
insol
sol
insol
4
Ferric Iron
pH
Ferrous Iron
Ferric Iron
sol
insol
90000 ppm
insol
900 ppm
insol
7
8
insol
insol
NB Iron Sulphides precipitate at any pH when hydrogen sulphide is present
The main source of Fe+++ is the acid reacting with rust in the surface tanks, flowlines
and millscale on the tubing. The Fe++ is mainly (>80%) derived from the formation
minerals e.g. chlorite siderite, pyrite etc.
Department of Petroleum Engineering, Heriot-Watt University
25
1
A number of “sequestering” - or solubilising - agents are available to increase the
solubility of iron by forming soluble complexes. The concentration of “sequestrant”
required to prevent iron hydroxide precipitation depends on the expected ferric ion
(Fe+++) concentration. This is because it is unusual for the pH of the “spent” acid to
increase to a value of 6 when treating clastic formations unless they contain a high
percentage of carbonates i.e. ferrous hydroxide does not normally precipitate under
these circumstances.
The cheapest sequestering agent is citric acid. This has the disadvantage that the
maximum concentration allowable - and the maximum amount of iron cations that can
be sequestered - is limited by the solubility of calcium citrate. A more expensive
alternative, which can be used at higher concentrations, is EDTA (Ethylene Diamine
Tetracetic Acid). An alternative approach to preventing ferric hydroxide precipitation
is to reduce the Fe+++ to Fe++ by Erythorbic acid or ascorbic acid (vitamin C).
N.B. Fe+++ can catalyse the formation of an asphaltenic sludge when the acid contacts
some crude oils.
(iii) Solvents / Mutual Solvents / Surfactants
Use of these materials may reduce emulsion formation but can also be the cause of very
stable emulsion formation. Further, they may render the corrosion inhibitor ineffective
by preventing the absorbtion of the inhibitor onto the steel surface. They can be useful
in some circumstances - but their employment needs to be properly justified and a full
range of compatibility tests carried out (API RP 42 referred to earlier).
(iv) Nitrogen
Nitrogen gas may be added to the treatment fluid to assist flow back of the spent acid
and hence a rapid clean up when treating gas wells or depleted zones.
8.8. Selection of Treatment Type
The manner of execution of most matrix treatments falls into one of two classes:
(i) “Bullheading” - this term describes a treatment which is pumped down the
production tubing. The treatment - particularly if it employs acids - can displace rust,
scale, pipe dope, etc. present on the tubing’s inner wall into the formation; leading
to formation damage. Sometimes the matrix treatment is carried out prior to running
the tubing. This can be even worse since dried mud, cement, etc. which are present
on the casing wall can now also be dislodged as well as the damaging materials
referred to above.
If the formation pressure is sufficiently high, or artificial lift is installed, it is possible
to clean the tubing with a pre treatment (known as “pickling”). This involves injecting
into the tubing a volume of acid - usually equal to 20% of the tubing contents - and
displacing it until the leading front of the acid is a safe distance above the top
perforation. The acid, along with the dislodged, potentially impairing particles, is then
produced back to the surface.
Bullheading of (cold) fluids from the surface also causes the tubing to contract in
length. If this contraction generates too large tensile stresses in the tubing, the tubing
may part or cause unseating of the packer.
26
Acidising and Other Matrix Treatments
5
Both theses problems may be avoided by:
(ii) Pumping the treatment through a coiled tubing (CT) the end of which is
positioned opposite the perforations. Prior to the treatment being carried out the
inside of the CT often needs to be cleaned e.g. by “pickling” with acid.
However, this can relatively easily be carried out at the surface, if necessary.
Coiled tubing has a smaller diameter than production tubing - so the maximum pump
rate is limited (due to friction) and the use of ball sealers for diversion (see Section
5.8.9) is only practical if an unusually large diameter CT is employed.
8.9. Selection of Diversion Technique
Most formations are not homogeneous - in practice the perforated interval will contain
a number of formation layers with a range of permeabilities and, most likely, differing
levels of skin damage. The treatment fluid injection rate into each layer will be
governed by the radial flow equation since the pressure in the wellbore is in all
probability very similar for all the layers. This situation is depicted in figure 12 which
shows that by far the highest proportion of the acid is injected into the middle, high
permeability layer (Zone B). The natural tendency for the acid to be injected into this
layer was accentuated because this layer also has the lowest skin value. This was due
to the depth of penetration of the formation damage being the least. The Figure shows
that the formation damage in zone B alone was removed by the treatment employing
V1, a small treatment volume. Even doubling the treatment volume to V2 does not
allow the acid to successfully remove the formation damage in Zone A.
Formation Damage
V1
V2
Unaltered Reservoir
Acid
Penetration
Fronts
Zone A
K = 100 mD s = 1.3
V1
V2
Figure 12
Matrix (acid) treatment
profiles for a
heterogenuous formation
with varying levels of
formation damage
Acid
Penetration
Fronts
V1
V2
Acid
Penetration
Fronts
Zone B
K = 200 mD s = 0.2
Zone C
K = 50 mD s = 22
Acid thus takes the path of least resistance
Treatment one (V1) employs a smaller volume than treatment two (V2)
N.B. The injection rate into a particular layer, relative to the other layers, increases
rapidly once the formation damage has been removed by the treatment. Obtaining an
even distribution of the (acid) treatment is further complicated if the layers have
differing pore pressures.
Department of Petroleum Engineering, Heriot-Watt University
27
1
Flow Meter Surveys
Pre
Stimulation
Post
Stimulation
Depth
Perforated
Intervals
Formation
Damage
Remains
Flow Rate
The Flowing bottomhole pressure was the same for both flow meter surveys
An extra factor to be taken into account when considering acid placement is that, in
a bullhead treatment, the treatment fluid reaches the top perforation first i.e. the first
opportunity for the acid to flow into the formation and remove of formation damage
will occur in this top layer. The effect of this process is illustrated in Figure 13. This
illustrates a flow meter survey made in a vertical gas well before and after a “bullhead”
matrix acid treatment. The flow rates were adjusted so that the flowing bottom hole
pressure i.e. the drawdown, was the same in both cases. The post stimulation survey
shows that the acid job was successful from an economic point of view (50% increase
in production at the same drawdown); but that the acid mainly stimulated the top
perforations. The reserves from the bottom three quarters of the top perforated interval
are possibly not being produced, unless crossflow is occurring within the reservoir.
This differential pressure depletion also could lead to large pressure differences
developing between the various formation layers, causing drilling problems for future
wells.
In practice, many matrix treatments are “bullheaded” into the well but employ one of
the diversion techniques which have been developed to aid the more even fluid
distribution between the various formation layers. The more frequently used diversion
techniques are described below:
28
Figure 13
Flow meter surveys of a
vertical gas well made
before and after a
"Bullhead" acid treatment
Acidising and Other Matrix Treatments
5
ZONE A
ZONE B
ZONE C
(A) Zones A and B to
be selectively treated,
without treating zone C.
(B) Install packer between
Zones B and C using wireline
(C) Isolate zones A and B from
the rest of the well using a
retrievable bridge plug
(D) Run work string
and set inflatable packer
between zones A and B.
Treat zone B by pumping
down work string
(E) Treat zone A by
pumping down workstring/
casing annulus
(F) Unset packer and recover
workstring. Recover retrievable
bridge plug. Return well to production
ZONE A
ZONE B
ZONE C
Figure 14
Use of retreivable packers
and bridge plugs to
separately treat the upper
two zones of a multi zone
well
(i) Mechanical separation using conventional techniques.
These techniques include control of the point of fluid injection by use of retrievable
bridge plugs placed in packers set between completion zones (see Figure 14), dual
packers on a work string (equivalent to the Selective Placement Tool (SPT), Figure 15),
sequential perforation etc.
(ii) Coiled Tubing (CT)
Conventional CT may be used with the fluid exit ports at right angles to the tubing so
that the perforations are sprayed with a high pressure jet of treating fluid. Alternatively,
the CT may be modified with a single or dual packers to form the SPT (Figure 15 shows
a dual packer SPT). These packers are capable of being expanded and deflated many
times so that all the zones can be treated during one run into the well.
N.B. The spacing between the packers is constant while the CT is in the well.
Department of Petroleum Engineering, Heriot-Watt University
29
1
Coiled Tubing
Treatment Fluid
Tubing End Locater
Nozzle
Zone A
being
treated
ZONE A
ZONE A
Expanded
Packers
Unexpanded
Packers
Zone B
to be
treated
next
ZONE B
ZONE B
(A) Running In
(B) Performing Treatment
Packers not expanded
Packers expanded
Figure 15
Selective placement tool
(iii) Ball Sealers
Ball sealers are nylon covered balls sized so that they can seal off the perforations
Different sized balls are required depending on the perforation diameter e.g. when
“big hole” or “deep penetrating” (narrow hole) perforating charges are used. The balls
are pumped whenever it is desired to change fluid injection from one zone to another.
Their mode of action is illustrated in Figure 16. It is found in practice necessary to
pump a 30% - 100% excess of balls above the number of perforations to be sealed. The
density of the ball may be chosen so that it:
(i)
is buoyant (floats upwards in the treatment fluid due to having a slightly lower
density) or
(ii) sinks into the rat hole (i.e. denser than the treatment fluid). The ball density
relative to the produced fluid controls whether the ball is produced back to the
surface after the treatment is finished.
A
Treatment fluid
being injected
into bottom zone
30
B
Ball sealers
being carried
by the treatment
fluid to the bottom
perforators
C
D
Bottom zone
blocked by last
ball sealers,
fluid injection
diverted to
top zone
Fluid
injection
into
top zone
E
Welll returned to
production-pressure
differencial into the well
releases ball sealers which
are caught at the surface or
drop into rat hole (if density
ball > density fluid)
Figure 16
Ball sealer diversion
Acidising and Other Matrix Treatments
5
(iv) Viscous Fluids (gels and foams)
The idea behind the use of viscous fluids is that they increase the flow resistance in
the layer taking excessive amounts of treatment fluid so that the fluid is diverted into
a new layer. The skin (Svis) due to this viscous fluid can be estimated by an equation
similar to the Hawkins formula discussed earlier:
rvis
visc
Svis = µ
µ -1 In r
form
µvisc
µform
rvisc
re
where
e
=
=
=
=
viscosity of the viscous fluid
viscosity of the formation fluid
depth of the invasion of the viscous fluid
well drainage radius
The diversion process is only effective if the viscous fluid is highly shear thinning i.e.
its viscosity increases rapidly as its flow velocity decreases at greater depths of
injection. This allows it to form a viscous “plug”.
(v) Pack the perforation tunnel with a granular particulate (typically ± 200 mD)
(vi) Form a low permeability film, on the wall of the perforation. (typically < 1 mD)
Both techniques are illustrated in Figure 17. The permeability of the diverting agent
is mainly dictated by its particle size. Its presence in a perforation will reduce the
injectivity into that perforation, reducing the rate of treatment fluid injection and
increasing the flowing wellbore injection pressure. This will divert the injection
stream into a new zone. Further quantities of the diverter material may be added
continuously or intermittently in batches.
Casing
Cement
Thin, impermeable (< 1mD)
film on perforation wall
Formation
Figure 17
Diversion with granular
particulates or film formers
Diversion with
"Film Forming"
chemicals
Granular Diverter
(± 200mD)
packs perforation
Formation
Diversion with
Granular particulates
The concept behind the choice of materials used for these two diverting techniques is
that they must be capable of:
(a)
being prepared in the required range of particle sizes (hence filter
cake permeability),
(b)
be stable in the treatment fluid,
Department of Petroleum Engineering, Heriot-Watt University
31
1
(c)
disappear (dissolve or sublime in the produced fluids) from the
perforation so that it becomes fully open to flow once the well
is returned to production and
(d)
be non-toxic, cheaply and readily available.
Clearly, a higher concentration of the granular particulate material is required than
when a film forming chemical is used . Treatment design is difficult when using either
of these concepts - the injection profile can be made more uneven (“anti-diversion”)
if an inappropriate treatment design is used. In the extreme case, the excessive use of
the film forming chemicals can result in a complete loss in well injectivity before all
the treatment has been pumped (very embarrassing when acid is in the tubing!).
Typical chemicals used for both particulate and film forming diversion agents, by
preparing them in the correct particle size ranges, are:
(a) Benzoic acid (water / oil / gas),
(b) Sodium Chloride (NaCl or rock salt) crystals (water) and
(c) Oil soluble resin (oil) particles.
N.B. These particles are soluble, or sublime in the case of gas, in the phases indicated
in brackets after the name of the chemical.
9. MATRIX STIMULATION FIELD CAMPAIGNS
From a technical point of view, the preferred way to carry out matrix treatments in an
oil or gas field is to treat a number of wells in the form of an organised campaign, rather
than carrying out the work on an ad hoc basis. This allows the design methodology
introduced in Figure 5 to be applied, i.e. that experienced gained in EVALUATING
early treatments is used to improve the efficiency of later treatments.
Figure 18 compares the pre and post stimulation well productivity of a number of wells
in a field both before and after matrix stimulation treatments were carried out. The
decrease in average well skin achieved by the matrix stimulation is clear. In fact, these
gains show that matrix stimulation is often a high reward activity. In this case, despite
suffering a 20% treatment failure rate, the stimulation campaign delivered an
increased production capacity at 10% of the cost of generating the same capacity
through infill drilling.
32
Acidising and Other Matrix Treatments
Well Productivity
Distribution
Prior to Matrix
Stimulation
15
12
5
Average
Pre / Post Matrix Stimulation
No. of Wells
9
6
3
0
0
0.35 0.4
0.2
Figure 18
A stimulation campaign
0.6 0.65
0.8
1
PI actual / PI undamaged
The operator followed the Matrix Stimulation Design Methodology set out in figure
5 and by evaluating the treatment's performance and was able to halve the acid volume
(per net unit perforated interval length) during the course of the campaign; while still
maintaining the same, favourable, well response to stimulation.
A second example of the results of following this stimulation cycle methodology is
shown in Figure 19. Careful candidate selection ensured that the wells selected early
in the campaign showed higher levels of formation damage - hence higher production
gains were achieved compared to later in the campaign. However, cost saving
measures were so effective that, despite the average production gain nearly halving
during this five year period, the cost per unit of production gain (bopd or barrel oil per
day) was also halved, i.e. later treatments were more economic.
Example Stimulation Campaign
1200
1000
Cost
(US$/bopd)
800
600
400
Production
after Matrix
Acid Stimulation
(bopd)
150
80
200
Figure 19
A second example of a
stimulation campaign
0
'87
'88
'89
'90
'91
'92
Time
In a third case history and operator achieved and average technical cost of US$ 0.40/
bbl oil gained from their stimulation activities. This cost represents a high return on
Department of Petroleum Engineering, Heriot-Watt University
33
1
the capital employed. It is obviously greatly influenced by the candidate selection
criteria used.
10. STIMULATION OF CARBONATE FORMATIONS
Acidising of carbonate formations is fundamentally different from the acidising of
clastic formations. This is due to their differing physical nature and chemistry:
(i)
Carbonates consist of very fine grains exhibiting a vugular or fracture porosity
rather than the intergranular porosity shown by sandstones.
(ii) Carbonates react much more rapidly with hydrochloric acid than sandstones,
for the same formation temperature. Also, the use of mud acid is prohibited due
to the limited solubility of calcium fluoride.
Carbonates are normally found as massive deposits of chalk, limestone or dolomite.
Their constituent particles are much smaller than the typical sand grains found in
clastic formations. They will have undergone large porosity and permeability
reductions during burial and diagenesis. Although they are often pure (>95% wt
carbonate), they can also include iron minerals, clays and silicaceous materials giving
them a very variable composition.
The many possible diagenetic processes can lead to formations with similar chemical
compositions having a strength that varies from very strong to behaving similar to
toothpaste. Strong and weak layers can be present a small distance apart. This
complicates the planning of well completion - and stimulation - procedures.
10.1. Acid Composition Selection
Hydrochloric acid is used to:
(i)
bypass drilling or completion damage by dissolving the rock matrix;
(ii) widen natural fractures or secondary porosity so as to improve fluid conductivity
to the wellbore (see section 5.11.3);
(iii) increase the effective wellbore radius by wormhole formation.
Dolomite reacts much more slowly with Hydrochloric Acid than chalk or limestone
- the optimum reaction rate is achieved with a concentration of 28% wt HCl acid for
all dolomitic reservoirs. 15% wt HCl is used with the other carbonate formation types.
The amount of rock dissolved by the acid is determined by:
constant {volume acid * concentration acid * reaction stochiometry}
The constant depends on the units employed e.g. 1m3 of 14% wt HCl will dissolve 206
Kg of limestone with a volume of 0.073m3 assuming a porosity of 5% vol. The
corresponding amounts dissolved when dolomite is treated are some 7.5% smaller.
34
Acidising and Other Matrix Treatments
5
The injected acid does not dissolve the rock uniformly, instead it forms “wormholes”
- see Figure 20a and 20b. “Wormholes” consist of a main channel from which many
highly branched structures are formed. The number and extent of the wormholes
depend on:
(i)
the carbonate formation’s reactivity (high reaction rates promote few, long
wormholes);
(ii) The acid leak-off rate into the matrix (controlled by formation permeability,
acid and formation fluid viscosities and the injection pressure overbalance);
yyyy
;;;;
;;;;
yyyy
;;;;
yyyy
;;;;;
yyyyy
;;;;;
yyyyy
;;;;;
yyyyy
yyyyy
;;;;;
yyyyy
;;;;;
(iii) The presence of higher permeability streaks, fractures, vugs etc will determine
the preferred direction of wormhole growth.
Casing
Carbonate Formation
Cement
Figure 20a
Wormhole formation from
a perforation during matrix
acidising
Borehole Wall
Figure 20b
Wormhole formation from
an open hole completion in
a carbonate formation
Pre-Perforated Liner
Department of Petroleum Engineering, Heriot-Watt University
35
1
10.2. Treatment Types for Carbonate Rock Acidising
10.2.1. Matrix Treatments
Wormhole formation during matrix treatments improves the well inflow performance
by providing a high conductivity channel at depth from the wellbore. They are created
using either a:
(a) low rate, low volume, low acid concentration treatment.
(Typical values are 0.004m3/min/m, 0.3m3/m and 14% wt HCl for the injection
rate, injection concentration and acid concentration respectively). The low rate
and long contact time encourages wormhole formation and the bypassing of
shallow formation damage. This type of treatment is most suitable for short
intervals (< 12m).
(b) high(er) rate, large volume, high concentration treatment.
(Typical values are 0.025m3/min/m, 1.6m3/m and 14-28% wt HCl respectively).
The larger acid volume compensates for the reduced wormhole formation
caused by the use of the higher pump rate. Ball sealers are more effective in high
rate treatments - making them more suitable for treating longer perforated zones.
Both types of treatments have been applied with success - the preferred method
probably depends on the local situation with regard to formation damage, presence of
natural fractures & vugs etc.
10.2.2. Acid Wash (or Soak) Type Treatments
Wormhole formation is undesirable if the treatment objective is to remove near well
bore damage (e.g. perforations plugged with drilling mud, cement etc) present in a new
completion or after a workover. This is because forming the wormhole will consume
a large part of the available acid. Wormhole formation is avoided by keeping the
injection rate very low (<0.0002m3/min/m). The treatment is now called an acid wash
or soak. Treatment guidelines are:
(a) use the highest possible HCl acid concentration (max 28%) permitted by
corrosion considerations for the planned treatment time (10 hours or longer);
(b) use a coiled tubing (CT) to place a volume of acid, equal to the casing volume
of the perforated interval. If CT is not available the tubing should be “pickled”
(see section 5.8.7);
(c) The acid should contain an iron sequestrant (see section 5.8.6) and, if an oil
based mud cake is to be removed, a mutual solvent or dispersing surfactant (see
also section 5.11.3).
36
Acidising and Other Matrix Treatments
5
11. ACIDISING OF SPECIAL WELL TYPES
Some examples of well types requiring special treatment techniques.
11.1. Gravel Packed Wells
This well type poses a number of areas where fluid inflow can be impeded viz the
screen, perforations, gravel, grave/sand interface and the near wellbore formation.
Dedicated treatments require identification of the type and location of the formation
damage eg:
(i)
the screen: removal of material plugging the screen requires placing (or
“spotting”) the acid across the entire screen length while minimising inflow
into the gravel pack. This is best done by use of a CT which is moved across the
completion while pumping acid. A small acid volume is used e.g. 120% of the
wellbore completion volume.
(ii) the gravel: removal of residue left from the viscosifier. Viscosifier residues
are normally best removed with solutions of enzymes (for formation tempera
ture <65˚C), hypochlorite (a bleach) or 2% wt hydrochloric acid (at higher
temperatures). A treatment volume equal to 120% of the gravel pack volume
should be used together with a diverting agent which passes through the gravel
pack sand and filters out on the formation.
(iii) near wellbore formation damage: this should be treated in a similar manner
to perforated completions, except that the available diversion techniques are
much more limited due to the presence of the gravel pack.
11.2. Horizontal Wells
Treatment of horizontal wells is no different from conventional wells with respect to
the candidate and fluid selection criteria concerned. The most common, horizontal
well completion techniques use a slotted or perforated liner or a wire wrapped screen.
Cased, cemented and perforated completions are less frequent because of the greater
costs. Many of the conventional diversion treatments (ball sealers / chemical
diverters) would not be successful in the horizontal orientation since the treatment
fluid / diverter density differences will result in shutting off either the top or bottom
of the completion section. The SPT (Figure 15) is not effective in the uncemented liner
or screen completions because of the presence of the open annulus. Diversion with
a viscous fluid (foams or gels) is a possibility, but the very large volumes required may
make them operationally impractical.
Department of Petroleum Engineering, Heriot-Watt University
37
Po
Sp lish
ac ed
ed B
Fa ore
rA s
M
u
pa
Re dc
rt
m ake
ov
B
ed ei
n
by g
Po
Tr
Sp lish
ea
tm
ac ed
ed B
en
o
t
Cl re
Fl
os s
uid
eT
og
et
he
r
M
on udc
Op ake
en Re
Ho ma
le ini
Se ng
cti
on
Ex
te
rn
al
Ca
sin
Sl
g
Pa
Lin otte
ck
W er d or
er
ra or
P
pp W er
ed ire fo
r
at
Sc
e
d
re
en
Po
l
i
Sp sh
ac ed
ed B
Be ore
tw s R
ee eg
n
th ular
e
EC ly
P'
s
Ru
bb
er
Se
als
W
Co orks
ile trin
dT g
ub or
ing
Ce
m
Lin ent
er
1
yyyyyyyyyyy
;;;;;;;;;;;
;;;;;;;;;;;
yyyyyyyyyyy
;;;;;;;;;;;
yyyyyyyyyyy
;;;;;;;;;;;
yyyyyyyyyyy
;;;;;;;;;;;
yyyyyyyyyyy
FORMATION
Fluid Returns to
Surface
Injected
Treatment Fluid
FORMATION
Mechanical diversion, using the set up shown in Figure 21, is a possibility. The
horizontal well has been divided into a number of sections using external casing
packers. Rubber sealing elements installed on the outside of the CT seal against a
polished bore installed at regular intervals inside the liner or screen. This arrangement
allows for the completion zone between the external casing packers (ECP) to be
selectively treated in two manners:
(i)
circulation of the treatment fluid passed the mud cake, when the flow returns
to the surface via the CT / tubing annulus. This will aid removal of the mud cake;
(ii) injection of the treatment fluid into the formation by closing the wellhead valve
which allowed this annular flow.
The regular arrangement of ECPs and polished bores allow the treatment to be
repeated along the length of the completion interval.
This method requires the installation of a large number of ECPs and polished bores
during the initial completion of the well. This is expensive (increasing the well cost)
and also presents an increased risk of failure during the initial completion. It is also
often not known in advance whether - and at which point in the horizonal wellbore such treatments will be required.
38
Figure 21
Removal of mudcake from
behind slotted liner using
washpipe and seals
5
Acidising and Other Matrix Treatments
Open Fracture
Blocked Channel
Figure 22
(Acid) insolubles remaining
after the acid frac of impure
carbonate formation will
block channel if not
removed
11.3. Naturally Fractured Formations
(Natural) fractures often have a conductivity many times that of the formation. In fact,
in many carbonate formations the matrix has a very low permeability and all the
production comes from the fracture. Stimulation may have two possible objectives:
(i)
removal of damage to the fracture conductivity by drilling mud or cement;
(iii) enhancing the (natural) fracture conductivity by dissolving the cementing
materials deposited in the fracture.
As usual, fluid acid selection depends on the treatment objectives. A key point in the
fluid selection procedure is whether the fracture contains materials insoluble in
hydrochloric acid - if so these insoluble materials will remain in the fracture and block
the created permeability. This is illustrated in Figure 22. The top picture shows a clean
open channel created by acidising a pure, fractured limestone. It appeance is
contrasted with that of a similar experiment carried out on an impure, silty limestone
(lower picture). In the latter case the channel has been blocked by insoluble particles
remaining after the acidisation is completed. These particles can be mobilised by the
use of a silt suspending agent (a surfactant) so that they are flushed out of the fracture
when the well is returned to production.
Similar effects are observed when creating wormholes during matrix acidising
treatments of impure carbonate formations (see point c at the end of section 5.10.2).
If more than one fracture is connected to the wellbore, then ball sealers can be used
to divert the acid treatment from one fracture to another. In addition, the pump rate
can also be used to aid acid placement within the fracture.
(i)
low pump rates (and hence long contact times) enhance removal of mud cake/
cement etc. in the near wellbore region;
(ii) high pump rates - even exceeding the FPP - encourage removal of fracture
filling materials from at depth from the wellbore.
Department of Petroleum Engineering, Heriot-Watt University
39
1
12. ALTERNATIVE ACID FORMULATIONS
Proprietary acids based on other chemistry than that described here has been found to
be useful in specific circumstances e.g. fluoroboric acid has been shown to be useful
in gravel packed, gas production wells producing from clay rich formation prone to
permeability damage due to fines movement. Conventional mud acid was shown to
be effective at removing this damage (figure 23), but a rapid decrease in well
productivity over the next year resulted from further formation damage as shown by
the increase in turbulence factor during this period (figure24).
14
Trunkline
Restriction
Gas Production, MMscfd
16
Decline
after
Mud Acid
2,300
FTP, psi
2,800
18
1,800
12
10
8
6
4
2
0
Figure 23
Production history for a gas
well prone to formation
damage due to fines
movement
Proprietary
Acid Treatment
1
2
Time, Years
3
1.4
(P2r-Pwf2)/q, psi2/scfd
1.2
1.0
One year after
Mud acid treatment
0.8
0.6
0.4
Squares measured
> four years after
proprietary acid
Immediately after
mud acid
Line measured
immediately following
proprietary acid
0.2
0.0
0.0
5
Rate, MMscfd
10
Use of the proprietary acid, by contrast, lead to a sustained increase in well production
with no increase in turbulence over a four year production period. N.B. careful
selection of the mud acid compaction (table 6) might have achieved a similar long term
beneficial effect to the Proprietary acid.
40
Figure 24
Turbulence analysis plot
for the gas well shown in
figure 23
Acidising and Other Matrix Treatments
5
13. Appendix A :
Matrix fluid selection chart depending formation damage type to be
removed
Matrix Fluid Selection
Type
Damage
Symptom
Cause
Solution
Oil wet formation
Reduced oil
production
Corrosion inhibitors.
Oil based mud.
Mutual solvent* / water
wetting surface
Water block
Reduced gas
production in low
permeability or
depleted formations
Increased near wellbore,
water saturation
Fluid loss during
drilling / work over
Surfactant / alcohol
solution injected
with nitrogen
Emulsion
Viscous emulsion
near wellbore
Emulsion stabilised by
surfactant or oil wet
solids e.g. asphaltenes
Inject mutual
solvent / demulsifier
Wax
Deposit on tubing
or in formation
Oil cooled below
cloud point
Dissolve in heated oil /
solvent or disperse in
surfactant solution
Mechanical removed (scraping)
Crystal growth inhibitors
Asphaltene
Deposit in facilities,
tubing or in
formation
Pressure reduction
(precipitation often
starts near bubble point)
Mechanical removal
Dissolution and inhibition
not very effective
Bacterial
Slime or
Deposit
Deposits (mainly
in well) and corrosion
Bacterial Infestation
Bactericide prevents infestation
Mechanical removed and
oxidising agents (e.g. bleach,
chlorine dioxide removes
deposits)
Silts and Clays
Reduced well
inflow
Clay swelling and
fines migration
Clastics - dissolve damage
with HCI/HF
Carbonates - HCI to dissolve
formation / bypass damage
Inorganic
Scale
Deposit on well
equipment or in
formation
Formation brines
becoming super saturated
due to T and P reduction
or mixing of incompatible
brines
Mechanical removal
in tubing
Inhibition (inject inhibitor
in tubing or squeeze
into formation)
Dissolution (see table A-2)
Table A-1
Matrix fluid selection
* Mutual solvents dissolve both oil and water phases
Department of Petroleum Engineering, Heriot-Watt University
41
1
Fluid Selection for removal of Inorganic Scale
Type of Scale
Usual Manifestation
Treating Fluid
Comments
Carbonates
CaCO3
HCl
very common treatment
Sulphates
CaSO4 / CaSO4.2H2O
"Conversion Treatment"
(Na2CO3 solution followed
by HCl) or EDTA
common treatment
BaSO4 / SrSO4
EDTA + Synergist
very expensive
Chlorides
NaCl
H2O
gas well only
Sulphur
S
liquid at most downhole
temperatures
gas well only
Iron
FeS / Fe2O3
HCl + chelating agent,
also sometimes
reducing agent
encountered in an
acidic environment
(CO2 / H2S)
Silica /
Aluminosilicate
SiO2 / NaAlSiO4
HCl / HF
these scales cause
Formation Damage as
well as deposit in tubing
NB Scales are inorganic deposits formed by supersaturation of formation brines or mixing of two incompatible brines.
Mechanical removal is always an option in the tubing.
MATRIX STIMULATION TUTORIAL
Question 1
1. List three important factors each from drilling and production operations that
reduce well inflow that can be changed by matrix stimulation. Comment on
considerations that will guide the choice of the stimulation fluid.
Answer 1
Drilling and Completion Fluids
1. Invasion of solid particles from the drilling mud lead to blocked pore throats and
hence reduced permeability
2. These solids have many sources e.g. drilled solids, insoluble contaminants within
the base chemical components that were used to make up the drilling or completion
fluid, corrosion products eroded from the walls of the casing or work string etc.
3. Fluid loss from drilling and completion fluids into the formation increases the
near wellbore liquid saturation and hence reduces the (relative) permeability to oil
flow. This is often called a “Water block” when water based fluids are lost to the
formation.
4. Polymers absorbing from the drilling & completion fluid reduce the (relative)
permeability to oil flow.
5. Fluid loss into an formation incompatible with the this fluid leading to clay
swelling and reduced permeability
42
Table A-2
Fluid selection for removal
of inorganic scale
Acidising and Other Matrix Treatments
5
Production Operations:
1. Inorganic scales formed by mixing of incompatible brines within the well or due
to precipitation induced by temperature and pressure changes as the brine is
produced up the well.
2. Wax precipitated from cooling of the oil as it is produced up the well.
3. Asphaltene solids precipitated the oil phase due to pressure decreases undergone
by the oil as it passes through the production system.
4. “Fines” (very small diameter particles) that are dislodged by the flow of the
produced fluid followed by their migration through the pore spaces. They tend to
lodge in the pore throats in the near wellbore area. This is because their concentration
is highest here due to the nature of the radial flow process towards the well.
Matrix stimulation fluid selection
The Matrix stimulation fluid should remove (dissolve, solubilise or at least mobilise
the source of the formation damage, e.g.
wax
calcium carbonate scale
clay particles
“water block” in a oil well
-
hot, organic fluids
hydrochloric acid
mud acid
mutual solvent that solubilises both oil &
water phases
Bonus marks: choice of fluid has to maintain the materials in solution (avoid
reprecipitation of the dissolved solids) and be compatible with the formation.
Question 2.
You are required to rank the 5 wells in the PetEng field in the order that they should
be acidised during the planned well stimulation campaign. You are required to justify
your selection in terms of the potential production increase and treatment cost
efficiency (US$ per additional BPD oil production capacity). The artificial lift
installations in each well and the surface production facilities have sufficient capacity
to cope with the production increases.
The completions installed for all the 5 wells is similar with a wellbore radius (rw) of
0.3542 ft and a drainage radius (re) of 800ft. The full formation height is fully
perforated with an interval height (h) of 100ft. The reservoir oil has a viscosity (Uo)
of 0.92cp under downhole conditions and the volume shrinkage value during production
(Bo) is 1.35 bbl/STB. The formation porosity is 33% and is independent of permeability.
The wells were all damaged during the completion phase due to lack of fluid loss
control - greater volumes of (damaging) completion fluid were lost into the higher
permeability wells. The fluid loss caused a smaller (percentage) damage in the higher
permeability, more productive wells. One pore volume of acid to be injected into each
well to remove this formation damage.
Department of Petroleum Engineering, Heriot-Watt University
43
1
Parameter
Undamaged permeability (k)
Depth of damage
Percentage permeabilty damage
Producing drawdown "DP"
Well No.
Units 1
2
3
4
5
mD
10 50 100 250 500
in
4
5
8
10 12
%
80 75 50
30 25
psi
120 120 120 120 120
You have negotiated a stimulation contract with the service company in which the cost
is related to the volume of acid pumped. The charge is 1500 US$/ft3 of stimulation
fluid pumped.
Two useful equations are the Hawkins Formula to calculate the skin (S) and that for
calculating the well’s productivity ratio without and with formation impairment (Jo/Ji):
 k
 r 
S =  − 1 In d 
 k d   rw 
r 
In e 
 rw 
Jo
=
Ji
r 
In e  + S
 rw 
while the oil production rate can be calculated from the radial inflow equation:
qo =
k * h * ( Pe − Pwf )
r
141.2 * µ o * Bo * In e r  + S
w


The net revenue from the produced oil is $10 /bbl and your companies economic
criteria demands a 4 month pay back line.
Answer 2.
The first requirement is to calculate the Skin effect resulting from the formation
damage.
Well No.
Damaged permeability (kd)
Permeabilty ratio (k/kd)
Damage radius (rd)
Skin (S)
Units
mD
ft
Rank Order
1
2
3
4
5
2 12.5 50 175 375
5.00 4.00 2.00 1.43 1.33
0.69 0.77 1.02 1.19 1.35
2.66 2.33 1.06 0.52 0.45
1
2
3
4
5
Well No. 1 is the well with the highest skin (most heavily damaged). The apparent
stimulation candidate ranking using the Skin value as the ranking criteria is Well No.
1, 2, 3, 4, & 5.
44
Acidising and Other Matrix Treatments
5
However, this raking by skin factor does not take into account that the wells show a
wide formation permeability range. The potential production increase from the wells
achieved by removing the damage represents the payback for the stimulation treatment cost:
Well No.
Units 1
2
3
4
5
Skin (S)
2.65 2.33 1.06 0.52 0.45
Productivity ratio (Jo/Ji)
0.74 0.77 0.88 0.94 0.95
Production rate with damage (Qi) bpd
66 340 779 2,076 4,188
Undamaged production rate (Qo) bpd
89 443 886 2,215 4,430
Increase in production (Qo-Qi)
bpd
23 103 107 139 242
Rank Order
5
4
3
2
1
It can be seen that the stimulation candidate ranking order is reversed once we base
the ranking criteria on the potential gain in production capacity.
A different picture results gain if we base the ranking criteria on economics, based on
the fact that we know the depth of the formation damage and can thus alter the acid
treatment size depending on the volume of rock to be treated.
Well No.
Units
Increase in production (Qo-Qi)
bpd
Formation Pore Volume
ft3
from which damage removed
Cost treatment
US$
or US$ per BPD additional
US$/bpd
production
Rank Order
1
23
36
54,545
2,407
2
103
49
3
107
96
4
139
135
5
242
179
73,635 143,997 201,812 268,355
716
1,348
1,448
1,107
5
1
3
4
2
Well No. 2 is now the most attractive stimulation candidate since it minimises the
investment required for the gain in production capacity. This comes about because
well No. 2 had a shallow depth of impairment combined with a reasonable production
rate response to the removal of the formation damage. The ranking order is quite
different from either of the other two quoted above. This criteria is often the best one
since the alternative investment is often to drill a new well or to sidetrack an existing well.
The four month pay back whether time criteria can now be used to determine if the
treatments are economic. The preferred order in which the wells can be treated is also
shown
Well No.
Units
1
2
3
4
5
US$ per BPD additional production
Pay back time
Economic to carry out treatment?
Treatment order
US$/bpd
days
2,407
241
No
-
716
72
yes
1
1,348
135
No
-
1,448
145
No
-
1,107
111
Yes
2
However, note that well 5 is producing more than 60% of the fields total production.
It would be unwise to place this well at risk by carrying out an acidisation treatment
for a potential 5% increase in production. Thus well 2 is the only stimulation
candidate.
Department of Petroleum Engineering, Heriot-Watt University
45
1
14. FURTHER READING
“Production Operations” Volume 2 (4th edition)
by T. Allan and A. Roberts
published by Oil and Gas Consultants Inc
ISBN 0-930972-18-X
“Petroleum Production Systems”
by M.J. Economides, A.D. Hill and C. Ehlig-Economides
published by Prentice Hall
ISBN 0-13-628683-X
“Well Performance” (2nd edition)
by M. Golan and C. Whitson
published by Tapir
ISBN 0-13-9046609-6
“Reservoir Stimulation” (2nd edition)
edited by M.J. Economides and K.G. Notle
published by Schlumberger Educational Services
ISBN 0-13-775115-X
“Acidising Fundamentals”
by B.B. Williams, J.L. Gidley and R.S. Schecter
published by The Society of Petroleum Engineers
ISBN 0-89520-205-0
46
6
Hydraulic Fracturing
CONTENTS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
INTRODUCTION
HYDRAULIC FRACTURE TREATMENT
SELECTION GUIDELINES
FRACTURE STIMULATED WELL INFLOW
PERFORMANCE
A PROPPED HYDRAULIC FRACTURING
TREATMENT
TYPES WELL FRACTURING STIMULATION
TREATMENTS
A BRIEF REVIEW OF ROCK MECHANICAL
ISSUES RELATED TO HYDRAULIC
FRACTURING
6.1.
Insitu Sress
6.1.1. Effective Stresses
6.1.2. Fracture Initiation and Perforation
Programme
6.1.3. Data Gathering
6.2.
Fracture Size
6.2.1. Fracture Containment
6.2.2. High Insitu Sress Contrasts
6.2.3. Fracture Growth into Boundaries
6.2.4. Fracture Height Measurement
6.2.5. Fracture Azimuth and Well Deviation
Orientation
6.3.
Modelling the Shape of the Induced
Fracture
6.3.1 2D Fracture Models
6.3.2. Fracture Width
CREATION OF A PROPPED HYDRAULIC
FRACTURE
HYDRAULIC FRACTURING IN PRACTICE
OPTIMISATION OF HYDRAULIC
FRACTURE DIMENSIONS
PROPPED FRACTURE CONDUCTIVITY
THE INFLUENCE OF FRACTURING FLUID
AND THE FRACTURE CONDUCTIVITY
FRACTURING FLUID
TIP SCREEN OUT FRACTURING
13.1. Applications of TSO Fracturing
FURTHER READING
LEARNING OBJECTIVES:
• List the nomenclature of propped Hydraulic Fracturing (HF)
• Describe the factors which control the Productivity Increase Factor (PIF) achievable by HF
• Relate PIF to Net Present Value economics as a function of treatment size so as to
optimise HF treatment design
• Explain the role of Rock Mechanics in supplying basic design data for an HF
treatment
• Identify the key elements of the Rock Mechanics of Fracture Initiation and
Propagation
• Analyse Fracture Propagation Pressure Record to derive basic design data
• Discuss the importance of the perforation programme design to the success of an
HF treatment
• Distinguish between the different Fracture Propagation Models
• Explain how to select fracturing materials (fluids/proppants) for an HF treatment
• Discuss the factors influencing Hydraulic Fracture geometry (fracture shape and length)
• Critically describe the Hydraulic Fracture Treatment Design Procedure
• Describe the stages of a Hydraulic Fracturing Treatment operation
2
6
Hydraulic Fracturing
1. INTRODUCTION
Propped Hydraulic Fracturing consists of pumping a viscous fluid at a sufficiently
high pressure into the completion interval so that a two winged, hydraulic fracture is
formed. This fracture is then filled with a high conductivity, proppant which holds the
fracture open (maintains a high conductivity path to the wellbore) after the treatment
is finished (Figure 1). The propped fracture can have a width between 5mm and 35mm
and a length of 100m or more, depending on the design technique employed and the
size of the treatment.
Propped Hydraulic Fracture
Carbonate or
Sandstone
Formation
Side View
Cross Section
Confining Boundary
Confining Boundary
An Acid Fracture
Side View
A strong, inhomogeneous
carbonate formation
Figure 1
Fracturing Concepts
Acid Etched
Channels
Cross Section
Confining
Shale
Top Boundary
Confining Shale
Both types of fracturing treatments create highly conductive paths from deep in the
reservoir to the wellbore
Propped hydraulic fracturing is aimed at raising the well productivity by increasing
the effective wellbore radius for wells completed in low permeability carbonate or
clastic formations. The radial well inflow equation:
Q=
kh(Pe - Pwf)
r
141.2 µB0 In e +S
rw
=
kh(Pe - Pwf)
r
141.2 µB0 In ' e
rw
shows that the well production rate (Q) can be increased by:
Department of Petroleum Engineering, Heriot-Watt University
3
(i) increasing the formation flow capacity (k.h) {the fracture may increase the
effective formation height (h) or connect with a formation zone with a higher
permeability (k)};
(ii) bypassing flow effects that increase the skin (s) e.g. near wellbore formation
damage;
(iii) increasing the wellbore radius (rw) to an effective wellbore radius (r'w) where
r'w is a function of the conductive fracture length Lf (see Figure 2a).
Impermeable
Formation
Boundary
2rw
w
h
kf
Lf
Impermeable
Formation
Boundary
Fracture
Wellbore
Lf
r'w
Fracture
w
2rw
pe
Productivity hydraulically fractured well
Productivity of unstimulated well.
re
100
Formation Permeability (k)
k=
k=
10
0.1
1.0
md
md
k=1
0 md
k = 100 m
d
k = 1,000 md
1
k = 10,000 md
0.01
0.1
1.0
Time
4
Figure 2a
Propped hydraulic fracture
geometry
10
100 months
Figure 2b
Production increase due to
150 ft long hydraulic
fracture with a flow
conductivity of 8,000 mD ft
6
Hydraulic Fracturing
If the hydraulic fracture has infinite conductivity i.e. the pressure drop along its length
due to flow is negligible, then:
rw' = Lf/2
Thus high conductivity fractures allow fluids to flow to the well whose effective
radius has been enlarged to a value equal to half the single wing fracture length.
Alternatively, if the actual wellbore radius is used, this improved inflow can be
expressed as a negative skin.
The relative increase in production achievable by placement of a hydraulic fracture
is much greater in the case of low permeability formations (see figure 2b).
2. HYDRAULIC FRACTURE TREATMENT SELECTION GUIDELINES
Hydraulic fracture stimulation is required for the economic development of low
permeability reservoirs. This is because a highly conductive fracture results in a
negative skin. The wellbore flowing pressure (P1) has been increased, at a given flow
rate, compared to an unimpaired (P2) or impaired (P3) well (see Figure 3):
Gas
Pwellhead
Pseparator
GAS
Choke
Skin
(Zone of
damaged
permeability)
Oil to Tank
Flowing bottom hole pressure
P1 = -ve Skin
P2 = Zero Skin
P3 = +ve Skin
Figure 3
The Producing System
(i)
re
Reservoir
Reservoir
Permeability
(K)
Well Boundary
Kd
Pr
P1
P2
P3
the pressure observed (P2) for the same flow rate for a well with an ideal
(S = 0) completion or
(ii) the even lower pressure (P3) measured for the equivalent well showing a
positive skin due to formation damage.
The hydraulically fractured well with the negative skin will have the greatest
production rate. Propped hydraulic fracture well stimulation should only be considered
when the:
Department of Petroleum Engineering, Heriot-Watt University
5
(i)
well is connected to adequate produceable reserves;
(ii) reservoir pressure is high enough to maintain flow when producing these
reserves (or it is economically justifiable to install artificial lift);
(iii) production system can process the extra production.
These minimum criteria are equivalent to those used for matrix treatments and are
summarised in table 1. There is, however, one extra, unique requirement for propped
hydraulic fracturing:
Parameter
Oil Reservoir
Hydrocarbon Saturation
Water Cut
Permeability †
Reservoir Pressure
>40%
<30%
1-50 mD *
<70% depleted
>50%
<200 bbls/MMscf**
0.01-10 mD
twice abandonment pressure
Gross Reservoir Height
>10 m
>10 m
Production System
Gas Reservoir
20% spare capacity ø
* Frac and Pacs may be applied to higher pereability formations
ø Lower value if several wells are manifolded together
† see also criteria for matrix treatments (Table 3 of Chapter 5)
** evaluate potential for well killing itself due to liquid loading in the tubing
(iv) professional, experienced personnel are available for treatment design,
execution and supervision along together with high quality pumping, mixing
and blending equipment.
This latter requirement arises because a propped hydraulic fracturing treatment has a
complexity and difficulty an order of magnitude greater than that associated with
matrix or acid fracturing treatments. This arises because the ability to complete the
treatment to the specified design requires numerous, on-site adjustments during the
treatment. The first hydraulic fracturing treatments in a new area often experience
early screen out (see section 6.11) resulting in premature stopping of the treatment i.e.
the "learning curve" has to be climbed. The steepness of this "learning curve" can be
increased by employing personnel who have gained experience in successful fracturing
treatments in another area.
3. FRACTURE STIMULATED WELL INFLOW PERFORMANCE
The Inflow Performance of a Fracture Stimulated well is controlled by the dimensionless
Fracture Conductivity (Fcd):
Fcd =
6
kf*w
k*Lf
Table 1
Minimum hydraulic
fracturing candidate well
selection screening criteria
6
Hydraulic Fracturing
where:
kf*w
= fracture permeability (kf) * conductive fracture width (w)
≈ ability of the hydraulic FRACTURE to conduct fluid to the wellbore
= the fracture conductivity.
k*Lf
= formation permeability (k) * conductive fracture single wing length (Lf)
≈ ability of the FORMATION to deliver fluid to the hydraulic fracture.
These parameters are illustrated in Figure 4. It can be easily visualised that the
objective of the propped hydraulic fracture treatment design process is to ensure that
the pressure drop down the length of the fracture is low compared to the pressure drop
across the formation. Thus, as much as possible of the well drawdown should be taken
across the reservoir with the pressure drop within the fracture making a negligible
contribution to the total value of the well drawdown. We will now consider these two
factors in turn:
Cement
yyyyyyy
,,,,,,,
,,,,,,,
yyyyyyy
Formation Permeability- k
Casing
w
Fracture Permeability-kf
Figure 4
Factors contributions to the
dimensionless fracture
conductivity
Lf
the fracture conductivity is increased by:
(a) an increased fracture width (w),
(b) an increased proppant permeability (large, more spherical, proppant grains
have a higher permeability) and
(c) minimising the permeability damage to the proppant pack from the fracturing
fluid.
Frequently the increased production achieved by carrying out a hydraulic fracturing
treatment is represented by the "Folds of Increase" or FOI:
FOI = Qf/Qo
where
Qo =
k.h (Pe - Pwf)
µ.Bo.In(re/rw)
and
Qf =
k.h (Pe - Pwf)
µ.Bo.In(re/r'w)
Department of Petroleum Engineering, Heriot-Watt University
7
Qo and Qf are the well production under equivalent conditions before and after carrying
out the hydraulic fracturing treatment. Thus:
FOI =
In(re/rw)
In(re/r'w)
There have been several studies of the composite effect of fracture length, fracture
conductivity and formation permeability on the well inflow performance (see CincoLey, SPE 10043, 1982 for a review). A widely used correlation is that published by
Cinco-Ley and Samaniego (JPT, 1981, 1749-1766) in which {r'w/Lf}, or effective
wellbore radius divided by conductive fracture length, is plotted against the
dimensionless fracture conductivity (FCD). This is illustrated in Figure 5. This Figure
shows that a FCD value of 15 is required to ensure that the well inflow is not being
limited by the fracture conductivity. An alternative presentation (Figure 6) allows the
(negative) skin effect due to the propped hydraulic fracture (Sf) to be calculated from
the dimensionless fracture conductivity.
1
r’w / Lf
Maximum value for Fcd > 15
.1
.01
.1
1
10
Dimensionless Fracture Conductivity: Fcd = kfw / (kLf)
8
100
Figure 5
Cinco-Ley and Samaniego's
1981 correlation between
effective wellbore radius
and fracture conductivity
6
Sf + In ( xf / rw )
Hydraulic Fracturing
2
1
Constant value for Fcd > 15
0
0.1
Figure 6
Fracture skin effect varies
with fracture conductivity
1
10
100
1000
FCD
Dimensionless Fracture Conductivity: Fcd = kfw / (kLf)
The above correlations and equations can be used to quantify the relationship between
the increased production (FOI) as a function of the fracture length (Lf), formation
permeability (k) and the fracture conductivity (kf*w) - see Figure 7. Figure 7 shows
that for wells in low permeability (0.1mD) formations:
10
Folds Increase in Production
Fracture Conductivity(kf.w, mD.ft)
8
100
Formation
Permeability
1000
= 0.1mD
10000
6
4
Formation
Permeability
= 10 mD
2
0
Figure 7
Well productivity response
to hydraulic fracturing
0
200
400
600
800
1000
Fracture Half Length (Lf, ft)
(i)
high values of the FOI are possible;
(ii) FOI is related to fracture half length, while the fracture conductivity has a
limited effect, providing its value is greater than a certain minimum.
Department of Petroleum Engineering, Heriot-Watt University
9
The (low) formation permeability is controlling the well inflow and increased fracture
conductivity does not improve well performance.
An increase in the formation permeability to 10 mD results in a different picture:
(i)
a fracture with a low conductivity (100 mD.ft) has essentially no effect on the
well production;
(ii) increasing the fracture conductivity by a factor 10 (to 1,000 mD.ft) increases the
well production (or FOI); but the FOI is still independent of fracture length for
values greater than 100 ft;
(iii) a further increase to 10,000 mD.ft is required before the inflow performance
becomes sensitive to created fracture length i.e. the fracture conductivity is no
longer the only limiting factor in well inflow.
Inflow from the formation into the fracture is no longer the controlling factor for this
higher permeability reservoir. The above considerations will control the hydraulic
fracture treatment design process since long and highly conductive fractures are more
difficult to make and have a greater cost.
4. A PROPPED HYDRAULIC FRACTURING TREATMENT
The major steps needed to carry out a propped, hydraulic fracture treatment are
summarised as follows:
(i)
Pumping the fracturing fluid at a sufficiently high pressure to overcome the
rock stresses i.e. initiate and propagate a fracture.
(ii) The fluid properties are adjusted to ensure efficient fracture creation - low fluid
loss and tubing head pressure values are frequently achieved by use of a
viscous, shear thinning, water based, cross-linked gel.
(iii) The created fracture is then filled with proppant to "hold it open" or provide
conductivity for fluid flow when fluid pumping is halted.
(iv) The viscous fracturing fluid is degraded after the treatment to a viscosity
similar to that of water by incorporation of a chemical breaker into the
fracturing fluid formulation. This will allow it to be produced back after the
treatment, followed by the initiation of hydrocarbon production.
The surface and well set up required to achieve the above is schematically illustrated
in Figure 8.
10
6
Hydraulic Fracturing
Proppant
(Quality Control)
Pump
(Pressure Rating)
Spare Pump Units Available
in Case of Breakdown
Blender
Viscous
Fracturing Fluid
(Quality Control)
Wellhead
Pressure casing
/Tubing Annulus to Reduce
Tubing Burst Stress
(Pressure Rating,
Tree Saver Required?
If Yes , Erosion?)
Tubing
(Tensile Strength
Burst Pressure, condition?)
Packer
Casing
(Unseating Forces)
(Integrity)
Fracture
(Fracture
Length and Width)
Proppant
(Fracture
Containment ?)
Figure 8
Practical issues during a
propped hydraulic
fracturing treatment
Proppant Transport
and Conductivity)
Pay Zone
Fracturing Fluid
(Viscosity Degradation After Completion of Treatment)
Figure 8 shows the (viscous) fracturing fluid being combined in a blender with
proppant (e.g. sand grains) used to keep (or prop) the fracture open once the treatment
has been completed. The quality of both the fluid and the proppant need to be
subjected to proper quality control measures.
The proppant/fluid slurry is then passed to a high pressure pump where the fluid
pressure is increased to a value that a hydraulic fracture can be created in the pay zone.
Hydraulic fracturing has to be done as a continuous process - spare pumps have to be
IMMEDIATELY available if pump breakdowns are experienced. The well head must
have a sufficiently high pressure rating. This pressure rating of the wellhead can be
temporarily increased by the installation of a "tree saver" at the wellhead. This is
essentially a length of smaller diameter, thick walled tubing installed inside the
christmas tree. It has seals installed at the top and bottom to ensure that the wellhead
components with a lower pressure rating are protected from the high pressures
experienced during the hydraulic fracturing treatment. The tree saver's smaller
diameter leads to increased pressure losses and to the possibility of erosion of the
tubing if the proppant slurry exits the tree saver at too high a velocity.
The production tubing will be subject to burst forces due to the high pressures required
for fracturing. They can be reduced somewhat by pressurising the casing/tubing
annulus. The tubing will also contract due to the pumping of the cold fracturing fluids.
This may lead to tensile failure (tubing parting) or the packer unseating. In practice,
the design value for the tubing strength should be reduced appropriately to allow for
any corrosion if the well has been on production for a number of years. Field
experience indicates that it is often impossible to mechanically carry out a hydraulic
fracturing treatment in a well unless this was included in the original well design
specifications.
Department of Petroleum Engineering, Heriot-Watt University
11
The created length of the fracture is considerably longer than the propped length since
the fracturing treatment is still in progress. Issues to be evaluated during the design
include:
(i)
transport of the proppant to the fracture tip;
(ii) settling of proppant due to inadequate fracturing fluid viscosity;
(iii) creation of the required proppant pack width and degradation of the fracturing
fluid to minimise permeability damage to the proppant pack and formation and;
(iv) containment of the hydraulic fracture to the pay zone.
These will all be discussed in the following sections.
5. TYPES WELL FRACTURING STIMULATION TREATMENTS
Some of the main variants in fracturing technology are depicted for vertical wells in
Figure 9. Their areas of application have been summarised in table 2.
Candidate Well
Reservoir Type
Sandstone or
Carbonate
(Figure 9 a)
Treatment
Skin
(S)
Permeability
(k)
Conventional Proped Fracture
High / Low
Low
Skinfrac*
High
Medium
Matrix Acid
High
High
None
Very Low
High
Naturally Fractured
Reservoirs filled with
Calcite Cement
(Figure 9 b)
HCl Acid pumped
near or above Fracture
Propogation Pressure (FPP)
Low
Low
(Matrix)
Inhomogenous
Carbonates
(Figure 9 c)
HCl Acid pumped
above FPP
High / Low
Low
WISPER+- Pump Visious
Pad follow by HCl Acid.
BOTH above FPP
High / Low
Low
Homogenous
Carbonates
(Figure 9 d)
CFA∅- Pump Visious Pad
above FPP, follow by
HCl Acid just BELOW
FPP.
High / Low
Low
* Skinfrac - creation of a short, highly conductive fracture for treating high permeabilty formations
+
WISPER - Wide Spaced Etched Ridges
∅
CFA
- Closed Fracture Acidising
12
Table 2
Treatment selection
guidelines
6
Hydraulic Fracturing
Sand
Laden
Fluid
Figure 9a
Propped hydraulic
fracturing for sandstones
and carbonates
h
Hydraulic Fracture
Containment?
(i)
Proppant
(settling due to
insufficient
viscosity)
Figure 9a illustrates a propped hydraulic fracture treatment. Propped hydraulic
fracturing is applicable to both sandstone and carbonate formations. Proppant
transport and hydraulic fracture containment within the pay zone are key issues
to be addressed during the treatment design. The proppant will form the
required highly conductive channel from at depth in the reservoir to the
wellbore. Skin fracing (the creation of short but highly conductive fractures)
is applied to medium permeability reservoirs (k >100 mD).
Calcite cemented
natural fracture
Low
matrix
permeabilty
Calcite
cement
filling
natural
fracture
Figure 9b
Acid treatment of a low
permeability fomation with
natural fractures filled with
a calcite cementing material
Width of
natural fracture
increased
for carbonate
formations
Limited
open
channels,
probably
not
connected
Acid insoluble
residue
Open channel
created by
acid dissolving
calcite cement
Prior to Acidisation
After Acidisation
(Limited fracture conductivity)
(Acidised channel is highly
conductive)
Figure 9b illustrates the increased conductivity achieved by pumping hydrochloric
acid into a naturally fractured carbonate formation. see also chapter 5.11.3
Department of Petroleum Engineering, Heriot-Watt University
13
Lf
HCl Acid Etched
Fracture Walls
w
h
deeper channel etched
where rock is more reactive
Figure 9c
Acid fracturing for
inhomogenous carbonates
Figure 9c illustrates an (HCl) acid fracture treatment of an inhomogeneous, carbonate
formation. These naturally occurring inhomogeneities will ensure that some parts of
rock will react more quickly with the acid than others - resulting in a deeper etching
of the fracture wall at this point. Providing the formation is strong and inhomogeneous
enough, the (deeper) etched channels will remain open after the treatment is finished,
forming a conductive flow path to the wellbore.
N.B. the carbonate acid formulation selection guidelines (chapter 5 on "Acidising"
and other Matrix treatments, section 5.10,1) should be followed for both (9b) and (9c);
Lf
}
h
w
Widely spaced
of Groups of
Perforations
Viscous
Preflush
HCl Acid
Fingers
CFA (Closed Fracture Acidising) gives the similar result,
and allows higher perforation density
Figure 9d illustrates the results of a treatment designed to create a number of separate,
acid etched fingers. Soft, homogeneous carbonates (e.g. chalks) require the artificial
creation of the necessary inhomogeneities. This is done either through:
(i)
the low viscosity acid fingering through the high viscosity gel preflush at
pressures greater that the Fracture Propagation Pressure (FPP) (the WISPER or
Widely SPaced acid Etched Ridges process) or
(ii) by pumping the acid at just below the FPP (CFA or Closed Fracture Acidising
process). The special perforating procedure (widely spaced groups of
perforations) illustrated for WISPER is not required for CFA.
14
Figure 9d
WISPER (Wide Spread
Etched Ridges) process for
acidising homogenous
chalks
6
Hydraulic Fracturing
Table 2 shows that, with the exception of "skinfrac", all the fracturing based
technologies deliver an increased well inflow both by bypassing any near wellbore
formation damage (skin) and increasing the effective wellbore radius. "Skinfrac" type
treatments, as their name implies, are intended to bypass skin due to formation
damage. The increased well inflow due to an increased effective wellbore radius is not
so important because they are carried out in formations with a higher formation
permeability.
6. A BRIEF REVIEW OF ROCK MECHANICAL ISSUES RELATED TO
HYDRAULIC FRACTURING
6.1 Insitu Stress
It is well known (see Rock Mechanics appendix of the Petroleum Geoscience Module)
that there are three principle earth stresses oriented at right angles to one another
(Figure 10). Below about 500 m, in a relaxed tectonic environment, the vertical stress
(σv) is normally the greatest. It can be quantified by integrating the density log from
the point of measurement to the surface. An average value of 1.0 to 1.1 psi/ft is
measured for wells at reasonable depth - though lower values are encountered in
shallow, particularly offshore environments subject to rapid deposition.
zzzz
,,,,
yyyy
zzzz
||||
,,,,
yyyy
{{{{
z
y
,
zzzz
||||
,,,,
yyyy
{{{{
||||
zy,|{y,{
{{{{
Typically
at shallow
depths
(<500m)
σv > σH >σh
σH > σv >σh
or
σH > σh >σv
Overburden Stress (σv)
Figure 10
Insitu stresses in the
subsurface
Maximum
Horizontal
Stress (σH)
Department of Petroleum Engineering, Heriot-Watt University
Minimum
Horizontal
Stress (σh)
15
The other two stresses - the maximum and minimum horizontal stresses are oriented
at right angles to each other and at right angles to the vertical stress. The maximum
horizontal stress is also referred to as the intermediate stress.
N.B. In tectonically active areas e.g. in the foothills of mountain chains where
mountain building is occurring; the maximum insitu stress may no longer be vertical.
The orientation between the three insitu stresses remains the same - hence the
minimum and intermediate stresses are also no longer horizontal.
Propagation of a Hydraulic Fracture, involving the actual splitting apart the fabric of
the formation. The split will propogate in the direction of least resistance i.e. will occur
most easily in the direction perpendicular to the minimum insitu stress. Thus, in
tectonically relaxed environments, we can assume that a hydraulic fracture will have
a vertical orientation and will grow (propagate) in the direction intermediate (or
maximum horizontal) of the insitu stress {at right angles to the minimum insitu
stress}.
As discussed earlier, the vertical stress (σv) can be measured or assumed with
reasonable accuracy. The important rock property for predicting the other two stresses
from the vertical stress is called Poisson's Ratio (v), the ratio between Lateral Strain
(εy) and the Longitudinal Strain (εx)
i.e.
ε
Lateral Strain
v = εy =
Longitudinal Strain
x
{the negative sign is included because (convention) states expansion should be treated
as negative and we wish Poisson's Ratio to be a positive number}.
Figure 11 explains how subjecting a rock sample to a vertical (the overburden) stress
results in it shortening in the vertical direction and expansion in the horizontal
direction. A similar effect occurs in a reservoir rock deposited in a sedimentary basin.
The magnitude of the vertical stress at any depth is related to the weight of the
overlying rock mass. This can often be estimated by integrating the density log from
the depth under study to the surface (a default value of 1.0 - 1.1 psi/ft can be used if
this log is not available).
16
6
Hydraulic Fracturing
Overburden Stress (σv)
d
Longitudinal Strain:
dx
Original
Undeformed
Geometry
l
l
Lateral Strain:
εy =
d y /2
Figure 11
Measurement of Rock
properties:
Poissons's Ratio (v)
εx = dx
d
dy
d
Deformed Geometry
Due to Overburden Stress
If we assume that the horizontal stress depends only on the elastic behaviour of the
rock, the overburden load can be related to the horizontal stresses via the above
Poisson effect together with assumptions about the lateral boundary conditions. In a
tectonically relaxed area, the two horizontal stress components will have the same
value within a specific lithology:
1
σh = σH = v σv ≈ σv if v = 0.25
3
1−v
Variations in lithology, and hence variations in Poisson's Ratio, thus lead to abrupt
changes in horizontal stresses with depth.
The above is a simplified picture of the behaviour of rock formations but it does allow
the derivation of a first estimate of the potential horizontal stress changes between
layers. Figure 12 illustrates the resulting stresses from deposition of a formation with
constant lithology. Here tectonic forces resulting from movement of the earth's crust
have induced a stress component (σtec) so that the two horizontal components are no
longer equal.
Department of Petroleum Engineering, Heriot-Watt University
17
Insitu Stress Magnitude
Original Level Ground Surface
Current Ground Level
(dcrit )
σH > σh >σv
Depth
σH > σv > σh
Horizontal
Induced
Fractures
σtec
σv > σH > σh
Vertical
Induced
Fracture
Overburden Stress Gradient
σv = ∫ρ.g.h
≈ 1 - 1.1 psi/ft
σH, max= σh + σtec
σh= υ σv ( ≈ 0.33 σv )
1−υ
Original Stress
Today's Stress after erosion lowers ground level
Surface elevations change over geologic time. Figure 12 shows how surface erosion
has changed the vertical stress profile while the horizontal stresses have remained
"locked-in" due to, for example, inelastic rock deformation. This explains why the
vertical stress is normally not the largest one at shallow depths. Other geological
processes, such as burial and uplift, also lead to similar anomalous stress patterns.
The presence of a minimum horizontal stress implies that there will be a preferred
direction of fracture propagation. The fracture will grow in "easiest" direction
(requires the least amount of energy) i.e. as a vertical fracture perpendicular to the
minimum horizontal stress below the critical depth (dcrit in Figure 12) and as a
horizontal fracture at shallower depths.
6.1.1 Effective Stresses
The pore fluids present within the rock matrix will support a proportion of the total
applied stress. This means that effective stress (σ') carried by the rock matrix grains
is smaller than the total stress. This was quantified by Terzaghi as:
σ' = σ − P
where σ is the total stress, p is the pore pressure and σ' the effective stress which will
govern the failure of the material. This is illustrated in figure 13.
18
Figure 12
Magnitude of insitu stress
6
Hydraulic Fracturing
,,
yy
yy
,,
y
,
yy
,,
yy
,
yy
,,
,
,,
yy
yy,,
,,
,
yy
yy,
y
,,
,
yy
,,
yy
y
,
y
,
y
,
y
Pores
Insitu
Stress
Pore
Pressure
Figure 13
Effective stress acting on a
propped fracture.
Insitu
Pore
Pressure
Proppant or
Grains
Pore
Pressure
Stress
It was later recognised by Handin that the intergrain cementation does not allow the
pore pressure to completely counteract the applied load. A correction factor, the poroelastic constant α, was introduced:
σ' = σ α − P
where α can vary between 0 and 1 but has a typical value of 0.7 for petroleum
reservoirs.
One important conclusion from these equations is that the values of the stresses which
control fracture propagation can change as the reservoir pressure depletes during the
life of a petroleum reservoir. Hence the stress profile measured early in a field's
lifetime may become invalid as the field matures. E.g. A hydraulic fracture created
later in the life of a field will tend to be more confined to the pay zone than a similar
treatment carried out early in field life. This occurs because the pay zone reservoir
pressure will have decreased due to oil or gas production, while the pressure and hence
the insitu stress in the bounding shale will be unchanged.
6.1.2 Fracture Initiation and Perforation Programme
As discussed, the induced hydraulic fracture propagates at right angles to the
minimum insitu stress. From a conceptual point of view, it can be seen that the:
Fracture Propagation Pressure (FPP) ≈
minimum rock effective stress + pore pressure + fracture toughness
since the propagating fracture has to overcome the forces that are preventing further
splitting of the rock i.e. the minimum rock insitu stress and its fracture toughness (the
cohesion between the formation grains which needs to be overcome to allow the
fracture to propagate).
The Fracture Initiation Pressure (FIP) i.e. the pressure needed to start the fracture
propagating from the perforation will normally be greater (see section 6.6.4) then the
FPP. This is because fracture initiation requires additional energy to overcome the
tensile stresses present around the borehole plus any extra pressure required to
overcome the fact that the perforation is not oriented in the preferred direction for
Department of Petroleum Engineering, Heriot-Watt University
19
fracture propagation. Such a case of inefficient perforating leading to an increase in
the FIP is illustrated in Figure 14. This illustrates how it is unlikely that inline (00
phasing) perforations will be aligned with this preferred direction of fracture propagation
(In the case illustrated the perforation is oriented at right angles to the preferred
fracture criteria).
Formation
Perforation Tunnel
Restriction
Restriction
Maximum Width
Liner
Cement
σh
σH
Hydraulic Fracture
0
0 Phasing
(or in line) perforations
0
60 Phasing
Figure 14 shows how the fracture has to initiate from the perforation and then travel
around the well until it achieves the preferred orientation, after which it will propagate
away from the well. The width of the induced fracture is related to the difference
between the fluid pressure in the fracture and the insitu stress. This means that, in this
case, the induced fracture will be much narrower at the point of initiation since the
maximum rather than the minimum, horizontal stress is acting on the fracture. This
area of restricted width may not be wide enough to allow the passage of proppant
during the later stages of the fracturing treatment - resulting in a premature screen out.
(The relative importance of this restriction will depend on the difference between the
two horizontal stresses).
Field experience (backed up by simple geometric considerations) indicates that the
FIP can be minimised by perforating the well with 600 phasing - the maximum
difference between the perforation and induced fracture orientation will now be
reduced to 300.
20
Figure 14
Fracture initiation
6
Hydraulic Fracturing
6.1.3 Data Gathering
Prior to carrying out the minimum hydraulic fracturing treatment it is advisable to
carry out a smaller data gathering fracture treatment to measure the formation and
fluid properties. Depending on the volumes pumped, this is called a microfrac (<5 m3)
or a mini frac (<50 m3). Figure 15 illustrates the set up for such a treatment and some
of the important pressure reference points. The fluid is pumped at a constant rate for
the required time and the treatment pressure measured. Bottom Hole Pressure
measurements are much more accurate than surface measurements, since the hydrostatic
head and friction pressure drop components no longer have to be estimated.
Tubing head
treatment
pressure
(THTP)
X-Mas Tree
Fluid
Storage
Tanks
High
Pressure Pumps
Ground Surface
SC-SSSV
Tubulars
∆Pfriction
∆Phead
Packer
Minimum in-situ stress acting on fracture walls
Figure 15
Arrangement for a mini or
microfrac treatment
Downhole
pressure guage
Bottom
hole treating
pressure
(BHTP)
∆Pperforation
∆Pfracture propagation
∆P fracture friction
N.B. Proppant is not used during these treatments. Also only a short interval is often
perforated.
An example pressure/time recording is shown in Figure 16 and is described as follows:
Department of Petroleum Engineering, Heriot-Watt University
21
Pressure
Fracture
Initiation
Pressure (FIP)
Fracture
Reopening
Pressure (FRP)
Fracture
Propogation
Pressure (FPP)
˘PPerforation
˘P Propagation
+
˘PFracture Friction
Instantaneous Shut - In
Pressure (ISIP)
Fracture
Closure
Pressure (FCP)
0
Pump
Rate
Pump Cycle 1
Pump Cycle 2
Time
(i)
The bottom hole pressure begins to rise as soon as the pumps are started. This
increase continues until the Fracture Initiation Pressure (FIP) is reached, after
which it drops rapidly to the Fracture Propagation Pressure (FPP).
(ii) The pumps are stopped when the desired fluid volume has been pumped and
fracture propagation ceases. This pressure drops rapidly to the Instantaneous
Shut-In Pressure (ISIP):
ISIP = FPP - ∆Pperforations - ∆Pfracture friction
Where
∆Pperforation = Pressure drop across the perforations and
∆Pfracture friction = Frictional pressure drop from the perforations to the tip
of the fracture
(iii) The fracture is still open at the ISIP. Leak off continues at a high rate from the
open fracture. The pressure drops until the Fracture Closure Pressure (FCP), equal
to the minimum insitu stress (σh), is reached.
ISIP = FCP + ∆Pfracture propagation
Where ∆Pfracture propagation is the pressure required to overcome the fracture toughness
The FCP is recognised as a change in slope in the pressure decay curve. Fluid is
leaking off into the formation from the whole fracture surface when the fracture is
open.The fluid loss rate decreases to a low value after the fracture closes, since the area
available for fluid loss (perforations rather than an open fracture) is so much smaller.
N.B. If observed long enough, the pressure will eventually equalise at the reservoir
pressure. This may take a long time due to the low reservoir permeability.
22
Figure 16
Bottom hole pressure
recorded during micro or
mini frac
6
Hydraulic Fracturing
(iv) Reopening of the fracture during a second pump cycle will normally occur at
a lower value than the FIP - often the pressure “hump” does not occur and
the FPP is observed immediately.
Table 3 records typical values for the various parameters discussed above for a mini/
microfrac treatment in a well perforated at 6000 ft. These values are specific to the well
to be treated. They are required when designing the full scale, hydraulic fracturing
treatment. In addition one can derive further, valuable information.
Example
Tubulars
psi
THTP
∆Pfriction
∆Phead
BHTP
-6200
+4500
6400
4700
Fracture
Table 3
Typical pressure values for
a mini/micro frac treatment
at 6,000ft
(i)
∆Pperforation
∆Pfriction
∆Ppropagation
-100
-100
-300
Minimum in-situ stress
4200
The longer the fracture takes to close after the cessation of pumping, the lower
the leak off coefficient and the greater the fracture volume created (higher
fracture fluid efficiency). A volume balance can be performed to quantify this
fluid loss coefficient which can then be used as input to fracture treatment
design programs.
(ii) The fracture height can be determined using a temperature log. A temperature
log is run across the perforated interval before and immediately after comple
tion of the minifrac treatment. This production log, which incorporates a high
resolution thermometer, will record a (cool/lower temperature) zone
across the created fracture due to the injection of the cold fracturing fluid.
Figure 17 is a schematic example of such a temperature log, the fracture being
observed to have grown upwards from the perforated interval.
Department of Petroleum Engineering, Heriot-Watt University
23
Temperature Log
Ge
Before mini frac
ot h
After mini frac
er m
al G
i
rad
Upward Growth
of Fracture
en t
e
ord
Depth
(r e c
)
frac
ini
re m
efo
db
Zone of cooling
recorded
immediately after
mini frac
Created
Fracture
Height
Perforated
Interval
Figure 17
Fracture height
Temperature
measurement using a
temperature log
6.2 Fracture Size
Greater volumes of fracturing fluid will create larger fractures - with higher treatment
costs but also potentially more productive. However, often uncontrolled growth of
fractures is not desirable from a production point of view e.g. when the target oil zone
is overlain by gas with water underneath. Figure 18 shows how the maximum fracture
size is limited for this situation. It assumes that:
yyyy
,,,,
,,,
yyy
,,,,
yyyy
,,,,
yyyy
yyyy
yyy
,,,,
,,,
,,,,
,,,,
yyyy
yyyy yyyy
yyy
,,,,
,,,
GAS
OIL
Lf
WATER
Initial fracture geometry
(i)
Maximum size fracture allowable
assuming radial propagation
the fracture is initiated from perforations at the mid point of the oil zone.
(ii) the fracture propagates radially (in practice, this implies that the formation is
homogenous with a stress gradient equal to the hydrostatic head of the
fracturing fluid - see section 6.6).
The resulting maximum allowable fracture half length (Lf) is slightly less than half the
height of the oil column.
24
Figure 18
Fracture size limited by
geometry and fluid contacts
6
Hydraulic Fracturing
6.2.1 Fracture Containment
The hydraulic fracture should thus be designed so that it does not contact unwanted
fluids within a single formation layer. It must also be consider whether the hydraulic
fracture is contained within the pay zone i.e. whether upward and/or downward
fracture growth is retarded by changes in the formation property contrast between the
two layers. Important formation properties include:
Formation
Shale
Sand
Shale
Figure 19
Insitu stress contrasts
(i)
yyyyyy
,,,,,,
,,,,,,
yyyyyy
yyyyyy
,,,,,,
yyyyyy
,,,,,,
Poisson's
Ratio (υ)
Minimum insitu stress (σh)
0.25
0.15
Pay Zone
0.25
(Minimum) insitu stress: permeable formations e.g. sand typically have a lower
Poisson’s Ratio compared to the bounding shale layers. The resulting lower
insitu stress will aid hydraulic fracture containment (see section 6.1 and Figure 19).
(ii) Fracture toughness: increased values of the fracture toughness imply that it is
more difficult for the fracture to propagate in that zone.
(iii) Leak off: high fluid loss rates will retard fracture propagation through the zone.
3-D fracture propagation models are available to predict the simultaneous lateral and
vertical growth of the hydraulic fracture. These complex numerical codes have been
used to generate the following examples which illustrate some of the important factors
concerning fracture containment.
Department of Petroleum Engineering, Heriot-Watt University
25
6.2.2 High Insitu Stress Contrasts and Fracture Shapes
Pay Zone
(ii) Hydrostatic Stress Gradient (0.45 psi/ft)
in pay zone
Perforations
(Sand)
(iii) Extensional Stress Gradient (0.7 psi/ft)
in pay zone
Perforations
Frac length
(Sand)
10 MPa
Pay Zone
Frac length
(Salt)
Cap Rock
Minimum insitu
Stress
Depth
(Salt)
Cap Rock
Depth
Pay Zone
10 MPa
Frac length
10 MPa
(i) Constant stress in pay zone
Minimum insitu
Stress
(Salt)
Cap Rock
Frac length
Perforations
(Sand)
Perforations
(Sand)
Pay Zone
10 MPa
Minimum insitu
Stress
Depth
(Salt)
Depth
Cap Rock
Minimum insitu
Stress
(ii) Overburden (or Maximum) Stress Gradient
(1.1 psi/ft) in pay zone
N.B. Fracture shape indicated by shading, volume fracture fluid pumped is the same in all cases
Figure 20 shows a case where a massive, homogeneous, gas bearing sandstone is
overlain by a (sodium chloride) salt zone (the cap rock). The well is perforated near
the top of the pay zone. Due to the plastic nature of salt:
σh = σH = σV
There is thus a very large insitu stress contrast (estimated as 10 MPa) at the salt/
sandstone boundary resulting in upward fracture growth being immediately halted.
Due to the homogeneous nature of the sandstone, it is expected that there is a constant
stress gradient. The effect on fracture shape and containment of differing values of
the stress gradient is illustrated in figure 20
(i)
constant stress (or zero stress gradient): the fracture grows downwards due to
the density of the fracturing fluid giving rise to an increased pressure at the
bottom surface of the fracture compared to the top surface.
(ii) hydrostatic stress gradient (0.45 psi/ft): the stress gradient in the formation is
now the same as that of the fracturing fluid - resulting in radial propagation of the
fracture, apart from the top surface where upward growth is stopped due to the large
(10 MPa) stress contrast at the salt/sandstone boundary.
(iii) extensional stress gradient (0.7 psi/ft) is commonly found in relaxed tectonic
environments. A longer fracture results since the fracture downward growth
becoming limited due to the minimum insitu stress with increasing at a greater
rate (0.7 psi/ft) than the hydrostatic head of the fracturing fluid (0.45 psi/ft)
.
26
Figure 20
Effect of stress gradient on
a fracture shape and
containment
6
Hydraulic Fracturing
(iv) overburden (or maximum) stress gradient (1.1 psi/ft) . This scenario yields the
longest fracture with the least downward growth. In fact, the fracture is trying
to grow upwards, the reverse of scenario (i), but is constrained by the high stress
contrast of the salt/sand boundary.
6.2.3 Fracture Growth into Boundaries
Whether a pay zone boundary is capable of containing a fracture will depend on the
magnitude of the fracture containment mechanism e.g. minimum insitu stress contrast
and the thickness of the boundary. Figure 21 schematically illustrates fracture
containment for 3 different values of the stress contrast. Initially the fracture
propagates radially in the pay zone until the boundary layer is reached; after which is
becomes more elongated - with greater stress contrasts giving rise to the more
elongated shapes.
Depth
N.B. The consecutive lines growing from the left hand side refer to the fracture shape
at increasing times/volume of hydraulic fracturing fluid pumped.
Formation B
500 psi
Stress Contrast
5,590'
6,000'
Formation A
6,050'
Fracture length
Depth
0
120
240
360
Upward fracture growth stopped at formation boundary
Formation B
200 psi
Stress Contrast
5,590'
6,000'
Formation A
6,050'
Fracture length
Depth
0
120
240
360
Limited upward fracture growth at formation boundary
Formation B
50 psi
Stress Contrast
5,590'
6,000'
Formation A
Figure 21
Vertical growth of fracture
for various stress contrasts
6,050'
Fracture length
0
120
240
360
Almost unimpeded upward fracture growth
Department of Petroleum Engineering, Heriot-Watt University
27
Figure 22 schematically illustrates what occurs when fracture containment is no
longer effective due to the height of the upper barrier and the available insitu stress
contrast being insufficient to prevent the fracture breaking through into a shallower
(water bearing) zone.
(a) The fracture is initiated at the centre of the pay zone (time 1) and initially grows
radially (time 2 and 3) since a hydrostatic stress gradient is present. This results
in the FPP decreasing as the fracture becomes longer.
(b) The stress contrast at the upper and lower boundary cause upward and
downward fracture growth to be retarded. Upward fracture growth is some
what easier as the stress contrast in this direction is less. The effect of this partial
fracture containment is to increase the FPP (times 4, 5 and 6).
7
Water Zone
(Sand)
Depth
6
5
Barrier
(Shale)
4
3
2
1
Pay Zone
(Sand)
∆σh
Barrier
(Shale)
Fracture Propagation Pressure
Insitu Stress
1
2
4
5
3
7
Time 7 - Pressure drop shows fracture breakthough
into the low stress water sand
Time
28
6
Figure 22
Fracture containment no
longer achieved
6
Hydraulic Fracturing
(c) At time 7 the upward fracture growth allows the fracture to break through into
the upper water bearing zone. Since this has a very low, constant fracture
gradient, the fracture will grow rapidly upwards, resulting in a drop in the FPP.
Note that the fracture length in the payzone has actually decreased due to the
fracture's rapid expansion into the upper, water sand. Real time measurement of the
FPP thus allows monitoring of fracture containment.
6.2.4 Fracture Height Measurement
Knowledge of the fracture height is clearly important when designing and executing
a hydraulic fracturing treatment. There are a number of possible measurement
techniques which can be used to measure this:
(i)
run a temperature log immediately after the fracture treatment to measure
cooled zone denoting fracture fluid entry {see figure 7}.
(ii) the depth at which fluid is entering into the well from the fracture can be
measured by running a production log across the perforated interval to measure the
flow profile (spinner a flow meter log) or the flow induced noise (noise log) or
temperature changes.
(iii) the proppant can be given a lightly radioactive coating. Running a gamma ray
log after the excess proppant has been cleaned out of the well will measure the
propped fracture height.
(iv) the fracture can be physically observed in open hole completions using :
(a) a formation Microscanner (a resistivity log with many, closely spaced,
measurement pads which can differentiate between the fracture and formation) or:
(b) a borehole camera (these cameras generate a picture of the borehole
wall using video or acoustic signals).
(v) making passive seismic measurements. This involves triangulation of
seismic events emitted from the propagating fracture tip. These seismic
events are measured with geophones installed at the surface or in the well.
(vi) use of a tiltmeter at the surface (for land wells only). This very sensitive device
measures (very small) changes in the surface topography due to propagation of
the hydraulic fracture. These changes indicate the length and orientation of the
hydraulic fracture.
It is apparent that each of the above techniques {apart from (v)} makes measurements
at the wellbore only. Further, they do not measure the same fracture attribute since
the created {(i) and (v)}, propped {iii)} and permeable {(i), (ii) and (iii)} heights, even
at the wellbore, are not necessarily the same.
Department of Petroleum Engineering, Heriot-Watt University
29
6.2.5 Fracture Azimuth and Well Deviation and Orientation
Earlier discussions (6.6.1) discussed how the induced hydraulic fracture would
orientate itself at right angles to the direction of the minimum insitu stress. This
discussion implicitly assumed that the well was vertical. A more complicated
situation arises when the well is deviated from the vertical - since the fracture and well
azimuth are unlikely to be orientated in the same direction.
σH
σh
(a) Well deviation in preferred fracture plane.,
Good communication
(i)
σH
σh
(b) Well deviation not in preferred fracture plane.,
Poor communication
Figure 23a illustrates the case when the well was deviated in the plane of the
fracture. Good fracture/well communication results once the well has been
efficiently perforated {Section 6.6.1.2 and figure 14}.
(ii) Figure 23b shows how the fracture only has a short length of contact with the
well when the fracture and well orientations are not coincident. This can
(potentially) lead to poor fracture/well communications and reduced well
productivity.
The propped, hydraulic fracturing of deviated wells can require special measures,
especially when the two insitu stresses have very different magnitudes. The drilling
of S shaped wells which are vertical across the pay zone is one option to be considered.
6.3 Modelling the Shape of the Induced Fracture
Several (commercial) programmes are available to predict the shape (height, length
and width) of the induced hydraulic fracture. They fall into one of three classes:
(i)
2D: these models use two dimensional, analytical equations where the fracture
height is required to be input.
This class is used extensively in engineering programs, requiring a reduced data set
to the others, as well as limited computing requirements.
(ii) P3D: or pseudo three dimensional programs. These combination of analytical
and numerical routines will predict the fracture height and width to vary
somewhat independently.
30
Figure 23
Hydraulic fracturing of
deviated wells
6
Hydraulic Fracturing
This class is used by fracturing specialists and has moderate computing requirements.
(iii) Fully 3D: these fully three dimensional programs are complex numerical
modelling programs with extensive input data and high end computing
requirements. The fracture height, width, length and shape can all vary
completely independently.
This class of program is normally only found in the research laboratories. Realistic
predictions form any of these programs depends on accurate input data being
available. This includes:
(i)
Geological data (formation boundaries etc.)
(ii) Rock mechanical properties from sonic and density logs and core measure
ments (Young’s modulus, Poisson’s Ratio etc.)
The extent of data provision required for specific case studies often limits the
application of fully 3D programs to more generic studies evaluating the effects of
changes in a given treatment or formation parameter.
6.3.1 2D Fracture Models
The complexity of hydraulic fracturing models derives from the need to simultaneously
satisfy two sets of laws:
(i)
conservation of momentum, mass and energy;
(ii) a fracture propagation criteria that controls the advance of the fracture tip.
Department of Petroleum Engineering, Heriot-Watt University
31
Elliptical cross section
Area of largest
flow resistance
Rectangular cross section
H
(a) Kristanovic, Geertsma
xf = L/2 and De Klerk (KGD)
Assumes: H > L
xf = L/2
H
Elliptical cross section
(b) Perkins,Kern and Nodgren
(PKN) Assumes: L > H
R
2R
(c) Radial Model
Assumes: L = H = 2R
Key
H (or 2R) is fracture height at wellbore.
xf (or L/2 or R) is fracture half length.
(A i
df
llb )
We will look at the class (i) or 2D models in greater detail:
(i)
The simplest solution available is when the fracture height is greater than the
fracture length and free slippage occurs at the upper and lower boundary giving
a rectangular shape at the wellbore. The fracture shape will not depend on the
vertical position in the fracture (see Figure 24a). This model was introduced
by Kristianovic, Geerstma and de Klerk (KGD).
(ii) A second situation is when the fracture is confined by boundary layers-it has
an elliptical shape at the wellbore and the length is much greater than the height.
32
Figure 24
Fracture shapes predicted
by two dimensional
hydraulic fracturing models
6
Hydraulic Fracturing
Figure 24b sketches this model. It was introduced by Perkins, Kern and
Nordgren. The FPP decreases as the fracture grows longer.
(iii) A limiting case between the two is the radial one when the fracture height equals
the fracture length (Figure 24c).
As mentioned earlier, they require a value for the fracture height to be input. Within
this constraint and when used in the correct application area (i.e. L/H ratio), they give
results which agree well with the more complex models.
6.3.2 Fracture Width
The preceding sections introduced the concept of fracture containment. However, to
be of practical use, the created fracture also has to be wide enough to admit proppant
that will provide the permeable flow path from the reservoir to the wellbore after the
fracturing treatment has finished. Pumping of the proppant when the fracture is not
wide enough will lead to an immediate premature screen out (bridging of the proppant
particles at the mouth of the fracture). The 2D fracture models discussed above are
used to predict this width value:
ωmax ≈
1
4
viscosity*pumprate
(ISIP -FCP)
Youngs modulus
where ωmax is the maximum fracture width at the wellbore.
The quarter power means that changes in the viscosity or treatment pump rates over
practical ranges have limited effect on the fracture width e.g. increasing the viscosity
or treatment pump rate by a factor three only increases the width by 32%. On the other
hand, rock mechanical properties of the formation have a much larger impact:
(i)
The Young’s modulus can vary by a factor 100, from 1 x 105 psi in soft
formations (soft chalks, diatomite, coals etc) to 1 x 107 psi in strong sandstones
as found in deep, low permeability formations.
(ii) The fracture toughness - which controls the ease of (or excess pressure required
for) fracture propagation and hence the value of the term (ISIP - FCP).
Department of Petroleum Engineering, Heriot-Watt University
33
7. CREATION OF A PROPPED HYDRAULIC FRACTURE
Fracture Leak-off
(a) Pad creates fracture. Fluid loss
limits rate of fracture length creation.
Fracture Leak-off
(b) Inject Proppant fluid slurry. Prop
displaced towards fracture tip. Fract
length growth continues.
Fracture Leak-off
(c) Stop injection when proppant reaches
fracture tip. Fluid leak-off continues.
(b) Fracture closed on proppant.
Well ready for production after
fracturing fluid's viscosity degrades
Figure 25 summarises the main stages in the process involved in creation of a propped
hydraulic fracture:
(i)
An initial fracture of appropriate length and width is created by pumping
fracture fluid called the pad. The most common fracturing fluids are water
based, cross-linked, polymer solutions (or gels) - which exhibit highly nonNewtonian rheological properties (see figure 40 chapter 7 and figure 15,
chapter 1) and appropriate fluid loss properties. This is often preceded by a
sacrificial pre-pad (a low viscosity fluid which satisfied part of the fluid loss
from the fracture at a reduced cost). Typically, about 50-80% of the total
fluid volume pumped leaks off to the formation while only 20-50% creates
useful fracture volume.
(ii) Proppant and Gravel Pack Sand are of similar size and the same material may
be used for both applications (see the section 7.5.4. on "Gravel Pack Sand
Selection"). Proppant particles are added at low concentration to the
fracturing fluid once:
(a) the fracture width is sufficient to admit the proppant without causing a
screen out and
(b)
34
the created fracture is nearing its design length.
Figure 25
Creation of a propped
hydraulic facture
6
Hydraulic Fracturing
Proppant Concentration (ppg)
9
Figure 26(a)
Hydraulic fracturing
proppant concentration
schedule during an
hydraulic fracturing
treatment
8
7
6
5
4
3
2
1
Pad Creates Fracture
Slurry Stage
0
Start
Treatment Time
End
Fracture containing boundary
1 lb/gal
proppant
concentration
First proppant
stage reaches
perforation
Pad
(i)
4 lb/gal
2 lb/gal
to 3 lb/gal
1 lb/gal
concentrated
to 4 lb/gal due
to leak off
Proppant concertration schedule
increases proppant concentration
injection into fracture at perforation
injected Proppant concentration in
fracture increases during job due to
leak off.
Pad
Proppant settle downwards due
to particle density greater than
that of fracturing fluid.
(ii)
Settling of dense proppant
continues until fracture
closure complete.
Figure 26(b)
Proppant profile
development during a
hydraulic fracture
treatment.
6 lb/gal
8
lb/gal to 8 lb/gal
4 lb/gal
to 8 lb/gal
2 lb/gal
to 8 lb/gal
1 lb/gal
concentration
to 8 lb/gal
Pad
Proppant concentration
profile at end of treatment
(iii)
Department of Petroleum Engineering, Heriot-Watt University
35
The concentration of proppant is increased towards the end of the job with proppant
concentrations as high as 40% vol being pumped {(Figure (26a)}. This ensures that
a more uniform, final proppant concentration in the fracture is achieved. This occurs
because the low proppant concentration in the slurry pumped initially will become
more concentrated as it is displaced towards the fracture tip as the fracturing fluid leaks
off into the formation. "Slumping" of the denser proppant will also occur due to the
influence of gravity. This process will continue until fracture closure is complete.
These processes are illustrated in figure 26(b).
(iii) The proppant slurry in the wellbore is displaced to the perforations and fluid
injection halted. This normally occurs at about the same time as the first
proppant reaches the fracture tip. There will be minimal stimulation from the
fracturing treatment if the final proppant slurry volumes are overdisplaced
away from the wellbore i.e. the well looses (direct) communication with the
hydraulic fracture. However, excessive under displacement of the slurry will
leave large amounts of proppant in the wellbore at the end of the treatment. This
would then have to be removed by a special cleanout trip made with a coiled
tubing or work over unit prior to returning the well to production.
(iv) Leak-off continues and the fracture closes on the proppant. Viscosity degradation
from the action of the chemical breaker (see section 6.12) added to the
fracturing fluid aids in the back production of the degraded fracturing fluid
followed by hydrocarbon production.
A complex computer program is required for treatment design since, in addition to the
complexities of fracture shape prediction described earlier, it must also optimise the
transport of the proppant within the fracture.
8. HYDRAULIC FRACTURING IN PRACTICE
The equipment required for a Massive Hydraulic Fracturing (MHF) treatment on land
is schematically illustrated in Figure 27.
36
6
Hydraulic Fracturing
Well
Low Pressure Pump
Production Flow Line
Base
Fluid Tanks
Clean Up Pit
Base
Fluid Tanks
Base
Fluid Tanks
Pump
Pump
Fuel
Low Pressure Pump
Proppant Bins
Base Fluid
Suction Manifold
High Pressure
(Spare)
Control Vans
High Pressure Pumps
Base
Fluid Tanks
Base
Fluid Tanks
Discharge Manifold
Blender 2
Conveyer
Spare
Conveyer
Blender 2
Base
Fluid Tanks
Gel Tanks
(i)
Additives e.g.
Buffer,
Cross Linker,
Breaker
Blender 1
Gel Tanks
Sand Bins
}
Figure 27
Equipment for a massive
hydraulic fracturing
treatment on land
Conveyer
Blender
Spare
Liquid
Gel Concentrate
Base
Fluid Tanks
The base fracturing fluid (a brine compatible with the formation e.g. 1% wt KCl
solution) is placed in a series of clean tanks which are manifolded together.
(ii) A low pressure pump transfers the fluid to blender No 1 where it is mixed with
a concentrated polymer gel solution to give a typical polymer concentration of
40-100 lbs/1000 gal. Sufficient residence time combined with intense agitation
within the blender ensure the gel is properly hydrated (dissolved) in the base
fluid.
Department of Petroleum Engineering, Heriot-Watt University
37
(iii) The base gel is transferred to blender No 2 where additives such as crosslinking
agents, buffer, breaker surfactants etc are added. Quality control checks are
carried out to ensure the (cross-linked) gel has the desired properties i.e. the
additives are being added in the correct amounts and no contaminants are
present. Proppant is also added in blender No 2.
(iv) The proppant was loaded into bottom discharge bins prior to the commence
ment of the fracturing treatment. It is transported to the blender using a moving
band conveyor. Measurement of the proppant slurry density ensures that the
correct addition rate of proppant is being achieved.
(v) Blender No 2 transfers the prepared fracturing fluid or slurry to a number of
high pressure pumps manifolded in parallel. The pumps must be
sufficiently powerful to be able to pump to the well itself the fluids/slurries at
the required pressure and rate. A dual, high pressure flow line is often
provided between the pumps and the well to minimise frictional pressure losses.
(vi) The instrumented control van allows the “frac master” to monitor the treatment
progress, to order adjustments to be made as required and to cope with
operational problems e.g. faulty pumps, conveyors etc. as they occur.
(vii) Spare or standby pumps, blenders, conveyors, etc should be hooked up and
operational before the treatment commences. This ensures failures have a
minimal impact on the treatment progress.
(viii)A fuel tanker is required to ensure that all the engines driving pumps, blenders
etc have sufficient fuel for the treatment duration.
(ix) The proppant slurry is displaced with base gel to just above the perforations and
the pumps stopped. The well is shut in for sufficient time to allow the fracture
to close and the fracturing fluid. viscosity to degrade The well may now be
placed on production. Initially only degraded fracturing fluid will be produced
- this can be directed to a clean up pit. Once hydrocarbons are being produced
the well production should be directed to the facilities via the production flow line.
The process for fracturing an offshore well is essentially the same as the above,
although all the equipment is installed in a specially designed stimulation boat (figure
28). Sea water is often used as the base fluid, reducing the fluid storage requirements.
Carrying out a hydraulic fracturing treatment is an expensive, complex undertaking.
The costs, volumes of material used and the dimensions for a typical MHF treatment
are summarised in table 4.
38
6
Hydraulic Fracturing
Crew's
Quarters
and
Wheelhouse
N.B. Heavy loads (base fluids, proppant etc.)
stored on the lower boat decks to
maximise vessel stability
Observation Deck
and
Control Room
Pumps
High Pressure
Pumps
High Pressure
Pumps
High Pressure
Pumps
High Pressure
Pumps
High Pressure
Proppant holding
tank,main proppant
storage in
lower decks
Platform Arrangements
Pumps
High Pressure
Pumps
High Pressure
Pumps
High Pressure
Proportioning
Blender
Well
Valve
Electronic
Pressure
Transducers
Flow Meter
Base fluid, gel and
additive
feed lines from
below deck tanks
Safety, high pressure
pop-off valve
Radioactive
Densometer
Check Valve
Hyraulic quick
disconnect
Figure 28
Layout of an offshore
stimulation vessel
Flexible high-pressure line to platform
Department of Petroleum Engineering, Heriot-Watt University
39
Base Fluid
6000 bbl
Proppant
1,000,000 lb
Pump Rate
30 - 50 bbl / min
Wellhead Treating Pressure
10,000 psi
Cost
1,000,000 US$ (offshore, UK Southern North Sea)
350,000 US$ (land, USA)
Fracture
Length
500 ft
Height
200 ft
Width
0.5 - 2 inch
Dimensionless Fracture
Conductivity
Production Improvement
Factor
3 - 100
2-5
9. OPTIMISATION OF HYDRAULIC FRACTURE DIMENSIONS
The expense of carrying out a hydraulic fracturing treatment dictates that the propped
hydraulic fracture dimensions are carefully optimised to generate the maximum
discounted net present value (NPV) for the stimulation project. Part of this process
is illustrated in Figure 29. It consists of two strands, (A) and (B), which are combined
to generate a plot of NPV against fracture length.
A well performance prediction computer program is used to generate a plot of
well production rate against time for various fracture lengths (Lf) (see Figure 29a).
We assume a high {CFD > 15} dimensionless fracture conductivity at this stage. As
discussed in section (6.3) longer fractures will give higher production rates. These
figures can be translated into a plot of cumulative production against time, for each
fracture length (Figure 29b), and then further into a plot of discounted revenue against
fracture length (Figure 29c). Longer fractures generate more revenue, but the rate of
revenue increase decreases as the fracture length increases.
The second strand of the process (B) uses a computer program for hydraulic fracture
treatment design to generate estimates of the fluid volume required for a number of
fracture lengths (Figure 29d). These volumes may be equated with treatment cost
(Figure 29e) - as expected, longer fractures require more volume and are more costly
to produce. Further the rate of cost increase becomes larger as the fracture length
increases.
Finally, the data from Figures 29c and 29e are combined in the fracture optimisation
step (c) to give a plot of project net present value. The trends in the later two figures
ensure that there is a fracture length which gives the maximum project profitability
(Figure 29f).
40
Table 4
Typical massive hydraulic
fracture treatment statistics
6
Hydraulic Fracturing
The above only describes part of the treatment optimisation process since the fracture
conductivity (a combination of proppant type, size and impairment from residues left
behind by the fracturing fluid) and fracturing fluid selection all impact on the
treatment cost. Some of the considerations involved in selecting these aspects are
discussed below.
A
Well Performance
Computer Program
B
Hydraulic Fracture
Design
Computer Program
(a)
Lf = 3000 (ft)
Production Rate
1000
Treatment Volume
(d)
Lf = 500 (ft)
100
Lf = 0 (ft)
10
1
1
10
100 1000 10000
Time (days)
(b)
Treatment Cost($)
(e)
Lf = 3000 (ft)
Cumulative
Production
0
1000 2000 3000
Fracture Length, Lf (ft)
Lf = 1000 (ft)
Lf = 500 (ft)
Lf = 0 (ft)
Time (years)
0
1000 2000 3000
Fracture Length, Lf (ft)
C
Fracture Value
Optimisation
Discounted Revenue
Net Present Value ($)
(c)
0
1000
2000
3000
Fracture Length, Lf (ft)
Figure 29
part of the process to
optimise the dimensions of
an hydraulic fracture
Net Present
Value ($)
(f)
0
1000 2000 3000
Fracture Length, Lf (ft)
Department of Petroleum Engineering, Heriot-Watt University
41
10. PROPPED FRACTURE CONDUCTIVITY
The proppant placed within the fracture is stressed as the fracturing fluid leaks away
and the fracture closes. This fracture closing stress (FCS) is equal to the minimum
insitu effective stress (σ'h)where:
FCS = σ'h= σh - Pfrac
where Pfrac is the fluid pressure within the fracture.
Sand Grains
Soft Proppants
Hard Proppants
Crushing
Deformation
Little Embedment
Sand Grains
Soft Proppants
Hard Proppants
Crushing
and Embedment
Deformation
Much Embedment
and Embedment
No Closure
Pressure
Hard
Rock
Closure Pressure
Applied
No Closure
Pressure
Soft
Rock
Closure Pressure
Applied
Depending on the properties of the proppant and the strength of the formation, figure
30 illustrates how this fracture closing stress will result in:
(i)
crushing of the proppant grains leading to reduced proppant permeability, and
hence reduced fracture conductivity;
(ii) deformation of (soft) proppants which leads to reduced fracture width, and
hence reduced fracture conductivity;
(iii) embedment of the proppant in the fracture wall, leading to a further reduction
in fracture conductivity.
The size range for commercial proppants are specified in a similar manner as gravel
pack sand (section 7.5.4). Figure 30 shows how the resulting fracture conductivity is
a result of a combination of the proppant type (quality) and the formation properties.
The available proppant types are listed in Table 5 while Figure 31 schematically
illustrates the variation in permeability as a function of closure stress (steel platens
were substituted for the fracture wall so that negligible proppant embedment occurred).
It can be seen that the more well rounded a proppant is the higher its
42
Figure 30
Behaviour of proppant
under stress
6
Hydraulic Fracturing
Proppant Type
Resistance
to Crushing
Cost
Low
Low
High
High
Low Quality Sand
High Quality (Ottawa) Sand
Resin Coated Sand
Intermediate Strength Proppant (Ceramic)
Table 5
Proppant types
High Strength Proppant (Bauxite)
permeability will be for the same proppant size and the greater its strength since the
fracture closing stress will be spread more evenly over the proppant grain's surface.
Thus.
(i)
the low quality sand {with its multi-crystalline, angular (or not well-rounded)
grains} began to crush at low stresses (< 200 psi).
(ii) the high quality sand (well rounded, mono-crystalline grains) shows a much
greater stress resistance.
(iii) bauxite, a high strength proppant, showed negligible grain crushing and a small
amount of deformation; even at the highest closure stresses (10,000 psi).
% Fracture Conductivity Remaining
NB. It must be remembered that the highest Fracture Closure Stress will be experienced
by the proppant in the fracture when the reservoir is depleted and the well is under
production i.e. equate Pfrac to the minimum expected flowing bottomhole pressure
when calculating the FCS to be used for proppant selection.
Figure 31
Fracture conductivity
variation with stress
100
High Strength Proppant
(e.g Bauxite)
80
High Quality Sand
60
40
20
Low Quality Sand
0
0
2000
4000
6000
8000
10000
Fracture Closing Stress (psi)
Department of Petroleum Engineering, Heriot-Watt University
43
11 . THE FRACTURING FLUID AND THE FRACTURE CONDUCTIVITY
There is another process that reduces the fracture conductivity. Most modern
fracturing fluid consist of a low concentration of a polymer dissolved in a brine. The
dilute polymer solution's viscosity is increased by joining the polymer molecules
together with a crosslinking agent. During the fracture treatment this dilute polymer
solution is pumped into the fracture at a pressure much greater than the reservoir
pressure - resulting in a high percentage of the base brine “leaking off” into the
formation. The large polymer molecules are too large to be able to flow through the
pore throat and hence form an external filter cake on the fracture surface. This is
particularly true for the low permeability formations where fracturing is often applied.
The chemical breaker is normally dissolved in the fracturing fluid. There is sufficient
chemical breaker present in fracturing fluid within the closed fracture to degrade the
fluid viscosity. However, it is not capable of destroying the filter cake, which remains
in the fracture. This further degrades the fracture conductivity (see Figure 32).
The efficiency of the removal of the remnants of the fracturing fluid from the proppant
pack itself is measured by the Retained fracture conductivity where:
Retained fracture conductivity =
(conductivity after exposure to fracturing fluid x 100%)
conductivity prior exposure to fracturing fluid
Fracture closure stress
Embedment in
formation
(Soft) Formation Fracture Face
Filter cake Original propped
width fracture
Effective fracture width for flow
(Soft) Formation Fracture Face
N.B. Proppant grain crushing and deformation not illustrated (see Fig 30)
It can be imagined that the type and concentration of polymer used to prepare the
fracturing fluid effects the fracture conductivity. This is illustrated in Figure 33 for
a number of different fracturing fluids - retained fracture conductivities of between
50% and 100% are observed.
N.B. The skin effect due to the filter cake or fracturing fluid - reservoir rock
incompatibility has little effect on the reservoir inflow performance providing the
permeability damage is not excessive e.g. (kdamage / koriginal < 0.5). This is due to the large
inflow area of the hydraulic fracture surface. The effective fracture width open to flow
and the proppant permeability are the key parameters controlling the fracture flow
capacity.
44
Figure 32
Actual hydraulic fracture
conductivity
6
Hydraulic Fracturing
Fracturing Fluid Type
Key to Polymer Types
CMHPG with
Aluminium Crosslinker
HPG
- Hydroxy Propyl Guar
HEC
- Hydroxy Ethyl Cellulose
HPG Solution with
Titanium Crosslinker
CMHPG - Carboxy Methyl Hydroxy
Propyl Guar
Emulsion (67% Diesel,
33% HEC Solution)
HPG Solution with
Borate Crosslinker
Oil Gel
70% vol Nitrogen Gas
Foam / 30% HEC Gel
Figure 33
Typical values for fluid type
and retained fracture
conductivity
0
10
20
30
40
50
60
70
80
90
100
Retained Conductivity (%)
N.B. Proppant Loading - 2 lbs/ft2
Polymer Concentration - 40 lb/1000 gal
Ammonium Persulphate Breaker - 2 lb/1000 gal
It can be surmised from Figure 32 that the values measured for this retained
conductivity are dependent on the fracture width or the number of layers of proppant
(there are three proppant layers in Figure 32). This is because embedment of the
proppant layers in the fracturing fluid filter cake and in the fracture face itself becomes
progressively more important to the final fracture conductivity as the number of layers
of proppant grains decreases.
The number of proppant layers can also be expressed as a proppant loading (weight
of proppant per unit fracture area). Figure 34 summarises experiments performed at
varying proppant loadings. They clearly show that the filter cake effectively destroys
the fracture conductivity at low proppant concentrations.
Retained Fracture Conductivity (%)
0
10
20
30
40
50
60
Proppant Concentration (lb / sq ft)
Experimental Conditions
Figure 34
Typical values for proppant
loading and retained
fracture conductivity
Fracturing Fluid
0.5
1
Cross Linker
40lb / 1000 gal HPG
Titanium
Chemical Breaker
2lb / 1000 gal
(ammonium persulphate)
Proppant Loading
0.5 - 2 lb / ft2
Proppant Type
Ottawa Sand
2
Department of Petroleum Engineering, Heriot-Watt University
45
The chemical breaker added to degrade the fluid viscosity is capable of partially
destroying the filter cake. This is made clearer in Figure 35 where an increase in the
breaker concentration, in this case ammonium persulphate, results in greatly increased
retained fracture conductivity. However, addition of large concentrations of breaker
to the fracturing fluid is not a viable approach since this will result in the fluid viscosity
being degraded during, rather than after, the hydraulic fracturing treatment. This
early decrease in fracturing fluid viscosity will prevent the (denser) proppant being
transported to the tip of the fracture. Instead, the proppant will sink to the bottom of
the fracture under the influence of gravity and a premature screen out can result.
Retained Fracture Permeability (%)
50
0
0
1
2
3
4
5
Ammonium Persulphate Breaker Concentration (lb / 1000 gal)
Fracture Fluid Composition 40 lb / 1000 gal HPG Polymer Solution with Borate Cross Linker
PROPPANT
FRACTURING FLUID
FRACTURE GEOMETRY
& PRODUCED FLUID
Size
(Average Grain Size and
Grain Distribution)
Polymer Type and
Concentration
Fracture Width
(especially at wellbore)
Grain Roundness
& Sphericity
Fluid Loss Additive
Proppant Crush Resistance
&
Fracture Closure Stress
Crosslinker Type
Time
Breaker Type &
Concentration
Temperature
Temperature
Reduced PermeabilityNon-Darcy or turbuent
flow effects at high flow
rates and multiphase
flow
The factors which influence the effective proppant permeability when placed in the
fracture are summarised in table 6. These include, in addition to the above, the:
(i)
fracture width - particularly at the wellbore since:
fracture conductivity = constant*(fracture width)*(slurry proppant concentration).Kprop
46
Figure 35
Chemical breaker
concentration and retained
fracture conductivity
Table 6
A summary of factors
affecting proppant
conductivity
6
Hydraulic Fracturing
where Kprop is the permeability of the proppant. Larger proppant grains will have a
greater permeabiltiy {see chapter7 (Sand Control) section 7.5.4}.
(ii) reduced effective permeability due to:
(a)
high rate (non-Darcy or turbulent), flow effects. These can be particularly
important when stimulating gas wells.
(b)
multiphase flow effects (as observed during oil production).
(iii) greater activity from the chemical breaker at higher bottom hole temperatures
leading to greater retained fracture conductivity.
(iv) gradual “clean up” (reduction in water saturation from the fracturing fluid in the
near-fracture formation). This leads to a long term increase in the effective
permeability of this formation situated next to the hydraulic fracture. This
process can take many months after hydraulically fracturing a low permeability
formation.
12. FRACTURING FLUID
The properties and function of the fracturing fluid have been discussed in many of the
previous sections. These are all summarised in this section and a more in-depth
discussion provided on the aspects of fluid loss.
Fracturing Fluid Functions
(1) Initiate and propogate the fracture
(2) Develop fracture width
(3) Transport proppant thoughout the length of the fracture
(4) Easily produced back to the surface after the fracture treatment is finished,
leaving a fracture with the maximum permeability
Characteristics (Required to achieve the above)
Table 7
Fracturing fluid summary
(a) Stable, predictable rheology under surface and downhole treating
conditions and treatment duration
(b) Low friction pressures drop at high pump rates in tubing and flow lines
(c) Provide fluid loss control
(d) Clean and easily degradable to minimise formation damage to propped fracture
(e) Compatible with reservoir formation and fluids
(f) Economical / low cost
The functions of a fracturing fluid are listed in table 7. The fluid initiates and then
propagates the hydraulic fracture; creating the required fracture width so that proppant
can be admitted during the slurry stage. Gravity settling of the denser proppant in the
less dense fracturing fluid will result in a higher concentration of proppant at the
bottom of the fracture than at the top. This settling has to be controlled so that proppant
since we require that it be transported to the tip of the fracture. This is to ensure that
the required (but not necessarily complete) fracture height is propped. (Remember
that massive hydraulic fracturing treatments can take many hours). Finally, the
Department of Petroleum Engineering, Heriot-Watt University
47
viscosity of the fluid has to reduce to a value similar to that of water so that the base
fluid can be easily produced back to surface leaving a fracture with the maximum
possible retained conductivity.
These fracturing fluid functions can be translated into a series of characteristics which
are needed to achieve the above.
(a) Points (1) to (3) in table 7 require that the fluid has a stable and predictable
rheology under both surface and downhole (treating) conditions. Thus the
action of any cross linker used has to be stable at high temperature and high
shear conditions, while the chemical breaker should not become effective
during the time taken to pump the treatment.
(b) Fluids with a highly shear thinning rheology (a low power law n value - see
figure 40, chapter 7 and figure 15, chapter 1) combine the requirements of:
(i)
low frictional pressure drop in the surface flow lines and tubing and
(ii)
good proppant suspension properties (limited settling of the denser
proppant) in the fracture.
This is because the fluid’s rheological properties give it a:
(i)
low apparent viscosity in the high shear rate regime present in the tubing and
(ii)
an high apparent viscosity in the low shear rate regime present in the
fracture.
Obtaining a low pressure loss in the tubing is aided by a second effect due to the fluid’s
low power law "n" value. This increases the pump rate at which the transition to
turbulent flow occurs - this transition is associated with a large increase in the pressure
drop. In fact, the wellhead treating pressure often decreases when changing, at
constant pump rate, from pumping water (power law n = 1, viscosity = 1 cp) to
pumping a cross-linked fracturing fluid which has the consistency of a very thick, hair
gel when placed in a bottle .
(c) The fracture volume is created by that proportion of the fracturing fluid (pad)
which does NOT leak-off into the formation (see Figure 36). In practice, this
“useful” portion of the pad, or its efficiency (see below), corresponds to 20%60% of the total volume pumped. Figure 36 also shows that the fluid leak-off
occurs in a linear manner from the fracture face and can be described by a
fluid loss coefficient with the units:
volume / (area * time) ≡ distance / time.
48
6
Hydraulic Fracturing
Fluid leakoff
from fracture face
Well
c
Fra
ctu
Fra
Figure 36
Fluid leak-off during an
hydraulic fracturing
treatment
eig
e
Fac
ht
Fra
Fracture Fluid Volume Pumped (1000 gal)
Figure 37
The fluid loss coefficient
controls the created
fracture length
re H
ture
re
ctu
Len
gth
Fluid Efficiency
15%
400
Fluid Loss
Coefficient
30%
200
0.005 ft/min
0.004 ft/min
0.003 ft/min
0.002 ft/min
0.001 ft/min
45%
70%
0
0
500
1000
1500
Created Fracture Length (ft)
2000
Figure 37 schematically illustrates the effect of the value of this fluid loss coefficient
on the created fracture length as a function of fluid volume pumped. This fluid loss
coefficient determines the fluid efficiency, where:
fluid efficiency =
volume of fracture created
total fracture fluid volume pumped
The best estimate of the fluid efficiency when designing a specific well hydraulic
fracturing candidate is to measure the fluid efficiency during a mini frac treatment.
fluid efficiency =
constant*(fluid loss between ISIP and FCP)
total fluid loss to FCP or total volume pumped
Department of Petroleum Engineering, Heriot-Watt University
49
This field calibration method of measuring the fluid loss coefficient for use as input
into the fracture treatment design program produces value which takes into account
variations in the formation geology (the fracture may contact formation zones of
higher and lower permeability compared to that observed at the wellbore).
(d) The need for a fracturing fluid that was easy to degrade to leave a maximum
permeability propped fracture was discussed in detail in Section 6.11 (the
fracturing fluid and fracture conductivity).
(e) Choosing a fracturing fluid which is compatible with the reservoir formation
and fluids ensures that the leaked off fracturing fluid does not produce a zone
of formation damage on the fracture face. Due to the large inflow area of the
fracture face, such a damage zone will have a much more limited impact on the
formation inflow than similar skin values in the radial flow case (see chapter
4 (Formation Damage), section 4.4).
The preferred option is to choose a compatible fluid and to ensure such formation
damage does not occur.
(f)
Low cost solutions are always preferred! As shown in Figure 27, the fracturing
fluid consists of a base gel prepared by diluting a concentrate polymer solution
with the base fluid (normally water) and then adding one or more additives
selected from:
(i)
cross linking agents
(ii)
temperature stabilisers
(iii)
viscosity breakers
(iv)
clay stabilisers
(v)
surfactants e.g. foaming or anti foaming agents
(vi)
fluid loss additives
(vii) bactericide
(vii) buffers etc.
The Service Companies market the many, commercially available fracturing fluid
systems together with a wide choice of additives. In practice, an engineering
compromise has to be made when selecting a particular fracturing fluid and judging
its performance against the criteria listed above. A preferred choice for use when
fracturing medium temperature wells is a fluid based on a low concentration (40 lb/
1000 gal) of Hydroxy Propyl Guar (HPG) polymer dissolved in a dilute brine with the
addition of a borate ion cross linking agent. This fluid meets the table 7 criteria while
the low polymer concentration minimises the cost and maximises the retained fracture
conductivity.
50
6
Hydraulic Fracturing
13. TIP SCREEN OUT FRACTURING
Chapter 6 Described the carrying out of a conventional hydraulic fracturing treatment
designed to stimulate wells completed in low permeability formations with long, thin
fractures. The created fracture width at the wellbore depends on the fracture length
and the rock compliance (longer fractures are wider) while the propped fracture width
depends on the slurry proppant concentration (higher proppant concentrations will
lead to greater propped width).
The capabilities of hydraulic fracturing technology have been extended into softer
formations (which often have a medium to high-permeability). This involves the
hydraulic fracture treatment design to deliberately create an early Tip Screen Out
(TSO) This process involves the pumping of the proppant slurry early during the
fracturing treatment. When the first proppant arrives at the fracture tip it forms a
“bridge”, or proppant plug at the tip (the TSO).
A. Fracture intiaties
B First proppant enters
C. Tip screenout
Figure 38(a)
Fracture width inflation
with the tip-screenout
technique.
Proppant reaches fracture
tip and screens out
Increasing pressure in fracture
D. Fracture inflation by continued
pumping proppant slurry
E. Packed fracture
Fluid leak off
The proppant plug has a relatively high pressure drop across it compared to the
fracture. This reduces the fluid pressure acting on the fracture tip, so that further
fracture length growth is halted and fracture width growth can begin to provide the
extra fracture volume to accommodate any further volumes of proppant slurry
injected into the fracture. This results in an increase of the FPP since the fracture is no
longer increasing in length.
A typical treatment history is schematically illustrated in Figure 38b and table 8. The
treatment proceeded as follows:
(i)
The fracturing fluid used was a low concentration (40 lb/1000 gal) HPG
polymer solution with a borate cross linker (as recommended above).
(ii) The treatment was started with a pump rate of 5 bbl/min. Fracture re-opening
was confirmed by a peak in the downhole pressure, followed by a slow decrease
as the fracture extended in length.
(iii) After 13 minutes the pump rate was increased to 25 bbl/min - with a corresponding
rise in treating pressures. A much greater increase was observed in the surface
measurement since the increased pump rate gave rise to some 3000 psi extra
frictional pressure drop across the tubing.
Department of Petroleum Engineering, Heriot-Watt University
51
(iv) Pumping of proppant slurry was begun after 23 minutes with the first proppant
arriving at the perforations some 3 minutes later. The bottom hole pressure
began to increase after 28 minutes. The tubing head pressure drops between 28
and 30 minutes because the proppant containing slurry is denser than the
fracturing gel above.
(v) Pumping of the higher proppant concentrations - stage 3 began at 30 minutes.
The bottom hole pressure continued to rise smoothly while the tubing head
pressure shows a much steeper increase in pressure after 33 minutes - the denser
proppant slurry has an increased viscosity compared to the cross-linked gel alone.
(vi) The final proppant stage was pumped (stage 4) and displaced to the perforations
(stage 5).
Figure 38b and table 8 show that the crucial stage of this fracturing treatment occupies
a very short time period - 67% of the proppant is pumped in a only a 5 minute time
period. Also, the total treatment took less than 40 minutes.
6000
Measured Bottom
Hole Pressure
5000
Pump
Rate
Pressure (psi)
4000
Tubing Head
Pressure
3000
Surface Proppant
Concentration
2000
Start
Pumps
Pump
Rate
First Proppant
arrives Downhole
1000
Tubing Head
Pressure
0
0
10
Downhole
Proppant
Concentration
Start Pumping
Proppant at
Surface
20
30
40
Time (min)
Stage No.
Fracture initiation
52
1
2
Fracture creation
3
4
5
Pack fracture
Overflow
with the increasing
proppant concentrations
Figure 38(b)
Treatment record for tip
screen out hydraulic
fracture stimulation
6
Hydraulic Fracturing
Table 8
Pumping schedule for tip
screen out hydraulic
fracture stimulation
treatment
Stage
Fluid
Volume
(gals)
Proppant
Consentration
(lb/gal)
Cumulative
Proppant
( pumped lb)
1
Fracturing Gel
Creates Fracture
15,000
0
2
Dilute Slurry
7,000
0-4 (increasing)
16,000
3
Concentrated
Slurry
3,500
4-12 (increasing)
36,000
4
Very
Consentrated
Slurry
1,000
12
48,000
5
Base Fluid
Overflush
1,800
0
These figures should also be compared with the equivalent ones for a massive
hydraulic fracture stimulation treatment (table 4). The TSO fracture uses only 5% of
the amount of proppant and 10% of the fluid volume. These differences between the
two types of treatment are summarised in table 9 and Figure 39.
Table 9
Tip screen out and
conventional fracturing
compared
Fracture Type
Conventional
Tip Screen Out
Description
Long and Thin,
(Lower Conductivity)
Short and Fat,
(Higher Conductivity)
Width (inches)
> 0.25
0.25 - 1.5
Length (ft)
500 - 1500
50 - 500
0.5 - 2.0
4 - 12
Proppant
Concentration
( lb / ft2 )
Department of Petroleum Engineering, Heriot-Watt University
53
0.25
-1
500
500
0.2
in.
ntio
nve
Co ture
c
Fra
Pr
nal
opp
ed
ft.
1.
5-
5 in
50 ft.
500
.
-O
Tip een
r ed
c
S pp
ut
o
Pr cture
a
r
F
Figure 39
Tip screen out and
conventional fracturing
compared
13.1. Applications of TSO Fracturing
TSO fracturing is now a field proven technology for use in a number of areas.
(1) Sand control: gravel packing and fracturing are combined into a single
operation. In addition, a number of successful field trials in which fracturing
alone was used have been reported (successful sand control was attributed to the
reduced flow velocities associated with the TSO fracture compared to a
conventional perforated completion).
(2) Alternative to matrix acidising: using highly conductive fractures to bypass
near wellbore formation damage. Successful matrix acidising often requires
that the source of formation damage be identified in order to select the optimum
acid formulation {see chapter 5 (Acidising and other matrix treatments) section
5.2}. In contrast, TSO fracturing has the advantage that it is independent of the
type of formation damage. This type of treatment is generally called a "skinfrac".
Figure 40(a) shows how the well productivity increases slowly once the highly
conductive hydraulic fracture has penetrated the damage zone.
(3) Reserve increase in laminated sands: well completion in formations consisting
of finely laminated sands is problematic since there is a high chance that the
perforation density will not be sufficient for the well will to make contact with
all the hydrocarbon bearing zones (Figure 40b). Production via a TSO fracture
will ensure that recovery is achieved from all the zones (Figure 40c).
54
6
Hydraulic Fracturing
Pay Zone
Pay Zone
Shale Permeability
Barrier
Pay Zone
Unconnected Zone
Pay Zone
Figure 40a
Lost reserves in sands not
connected by perforations
Pay Zone
Pay Zone
(a)
TSO Fracture
Pay Zone
Pay Zone
Figure 40b
Tip screen out connects
extra reserves. All
productive sand bodies
being produced after
connection via hydraulic
fracture
Pay Zone
Pay Zone
Pay Zone
Pay Zone
(b)
Figure 40c
Cumulative production for
different fracture lengths
piercing a 10 ft radius
damage zone around the
wellbore.
Cumulative production
10
6
105
k = 100 md
rd = 10 ft
kd = 0.05 * k
S = 65
ks.w = 8,000 md-ft
Production from undamaged well
Fracture partially
penetrates
damage zone
No fracture (damaged well)
Lf = 5 ft
Lf = 15 ft
104
Lf = 40 ft
Lf = 80 ft
Lf = 150 ft
103
0.01
0.1
1.0
10
100 Months
Time
(c)
Department of Petroleum Engineering, Heriot-Watt University
55
(4) Production of medium viscosity oil from lower permeability formations: the
increased production rates achievable using TSO fracturing can sufficiently
improve the project cash flows that economic development of fields becomes
practical. A possible alternative development scenario is to use horizontal
wells to increase the well rates or even combine the two technologies. This is
schematically illustrated in Figure 41 where the large number of closely spaced
propped TSO hydraulic fracturing treatments have to be placed along a 1200m
horizontal well. This type of completion has been used on a wide scale for the
oil bearing chalk fields in the Danish sector of the North Sea.
Figure 42 shows that the potential production rate increases as the number of hydraulic
fractures increases, but that the rate of increase for each extra fracture is continually
decreasing because they interfere with each other in the terms of the reservoir inflow
as well as frictional pressure losses along the length of the horizontal well becoming
progressively more important.
(5) Fines migration: is triggered by high flow velocities due to radial inflow into
a perforated well. The flow velocities are reduced, and this form of formation
damage avoided, by the linear flow associated with a fractural well. Similarly,
sources of formation damage which are caused by pressure reductions may be
avoided since the improved well inflow will increase the flowing bottom hole
pressures (at a given production rate).
yyyyyyyyyy
,,,,,,,,,,
,
y
,,,,,,,,,,
yyyyyyyyyy
, ,,
y
yy,,
y,y,y,y,yy
,
y
,
y
,y
y
,
,,,,,,,,,,
yyyyyyyyyy
,,,,,,,,,,
yyyyyyyyyy
Horizontal
Well
140m
145m
GAS
Massive
Chalk
Formation
OIL
7th
frac
6th
frac
5th
frac
4th
frac
3rd
frac
2nd
frac
1st
frac
WATER
Figure 41
Completion employing
combined horizontal well
and multiple fracturing
technologies
Flow Rate (1000 m3 / day )
40
30
20
10
0
0
2
4
6
Number of Fractures
56
8
10
Figure 42
Example of increased gas
production achieved from a
multiply fractured
horizontal well in a low
permeability formation
6
Hydraulic Fracturing
HYDRAULIC FRACTURING TUTORIAL
Question 1
Contrast the application areas of Matrix Stimulation and Propped Hydraulic Fracturing
Answer 1
Matrix (often Acidising)
- removal of near wellbore damage
Propped Hydraulic Fracturing - improving well inflow performance by creating a
high permeability channel with a large surface area
for fluid inflow
Stimulation treatments are designed to increase the well’s Productivity Index (PI).
PI =
Kh
r
Bµ ln e r  + S


w
Matrix acidising aims to increase PI by reducing S by dissolving formation damage
and rock in the near wellbore region.
Fracturing increases PI by increasing the effective wellbore radius. Creates a high
conductivity fracture or channel from the wellbore. This extends deeper into reservoir
than can be achieved by acidising. It can also bypass the near-wellbore formation
damage zone and is occasionally used for this purpose alone.
Skin
High
High
High
Permeability
High
Medium
Low
Low
Low
Low
High
Treatment
Matrix treatment
Matrix/"Frac and Pack"
Fracturing (matrix may be
possible if perm. not too low)
Hydraulic fracturing
Treatment economic?
Question 2.
You are a service company representative who has designed the following fracturing
treatment:
Wellbore radius (rw):
Reservoir height:
Reservoir Permeability (k):
Proppant to be used:
Design Fracture Conductivity (kf*w):
0.328 ft
200 ft; bounded by competent shales
1.0 mD
500,000 lb
1,500 mD*ft at 4 lb/ft2 proppant loading
• Use the Cinco-Ley and Samaniego type graph (below) to advise your client on the
expected well negative skin (the fracture treatment covers the full formation height)
• Do you think this is the optimum fracturing treatment for this well?
Department of Petroleum Engineering, Heriot-Watt University
57
Sf + In(Lf/rw)
2
1
0
0.1
1
FCD = kf.w
k.(Lf)
10
100
1000
Lf is fracture half length
Answer 2.
Fracture length (both wings) =
Proppant
loading * frac. height
=500,000 / (4*200)=625 ft
Fracture half length (Lf) = 312.5 ft
FCD = kf*w/k*Lf = 1,500 / (1*312.5) = 4.8
From graph
In (Lf/rw) + Sf = 0.91
∴ Sf = 0.91 - In (312.5/0.328)
Sf = - 5.9
∴ Fracturing treatment will achieve a negative skin of - 5.9.
Fracture treatment design will produce a well with a negative skin of -5.9, a reasonable
value, However, a propped fracture with a higher FCD might be even more beneficial
to ensure that well productivity will not be impeded by lack of fracture conductivity.
An FCD of 10 - 15 is the standard recommendation, but the optimum in a particular
case depends on a detailed well production / fracturing treatment cost optimisation for
the case under study.
58
6
Hydraulic Fracturing
Question 3
The operator you are working for notices the above and queries whether this is an
optimum fracturing treatment design. However, he says that he is already over budget
so you are not allowed to increase the size of the hydraulic fracturing treatment but
commissions laboratory tests that show a fracture of Fracture Conductivity (kf*w) of
2
2,500 mD*ft can be expected at proppant loading of 5 lb/ft and 3,500 mD*ft at
2
proppant loading of 6 lb/ft . You have to advise the operator as to which fracture
design is optimum and why.
Answer 3
The limitation on the cost of the fracturing treatment means that the only way to
increase the well productivity is to alter the manner in which we deploy the proppant.
We need to increase the dimensionless fracture conductivity (FCD), implying that we
need to produce a shorter, more conductive (fatter) fracture.
Repeating the above calculation scheme gives:
Fracture length (both wings)
Fracture half length (Lf)
FCD = kf*w/k*Lf
From graph
In (Lf/rw) + Sf
∴ Sf
Sf
= 500,000 / (5*200) or 500,000 / (6*200)
= 500 ft
j = 416 ft
= 250 ft
= 208 ft
= 2,500 / (1*250) = 3500 / (1*208)
= 10 or
= 16.8
= 0.81 or
= 0.81 - In(250/0.328)
= - 5.8
= 0.76
= 0.76 - In(208/0.328)
= -5.6
The original fracture design thus appears to be optimum, giving the most negative
Skin value. This arises from the possible combinations of proppant loading (which
controls fracture length) and the created fracture conductivity.
Department of Petroleum Engineering, Heriot-Watt University
59
14. FURTHER READING
“Production Operations” Volume 2 (4th edition)
by T. Allan and A. Roberts
published by Oil and Gas Consultants Inc
ISBN 0-930972-18-X
“Petroleum Production Systems”
by M.J. Economides, A.D. Hill and C. Ehlig-Economides
published by Prentice Hall
ISBN 0-13-628683-X
“Well Performance” (2nd edition)
by M. Golan and C. Whitson
published by Tapir
ISBN 0-13-9046609-6
“Reservoir Stimulation” (2nd edition)
edited by M.J. Economides and K.G. Notle
published by Schlumberger Educational Services
ISBN 0-13-775115-X
ELY J W
Stimulation Treatment Handbook: An Engineer's Guild to Quality Control
Penn Well 1985
ISBN 0-87814-284-3
60
1
7
UnstableWell
Formations
Control and Sand Control
CONTENTS
1.
2.
3.
4.
5.
INTRODUCTION
TYPES OF SAND PRODUCTION
2.1.
A Southern North Sea Case History
2.1.1.
"Continuous" Sand Production
2.1.2.
“Clean-up” sand production
2.1.3.
“Fines” Production
2.2.
Sand Cementation
2.3.
When is Sand Production a Problem?
2.3.1.
The Consequences of Sand Production
2.3.2.
Living With Sand Production
2.3.3.
Monitoring of Sand Production
PREDICTION OF SAND FAILURE
3.1.
Field Experience
3.2.
Petrophysical Analysis
3.3.
Rock Strength Measurement
3.3.1.
On Site Strength Estimation
3.3.2.
Rock Mechanical Strength
Measurements
3.3.2.1. Unconfined Compressive Strength
3.3.2.2. Brinell Hardness Number
3.3.2.3. Thick Wall Cylinder Collapse Strength
3.3.2.4. Trixial Rock Strength Measurement
3.4.
Prediction of Downhole Rock Failure
3.5.
A Final Word
COST OF SAND CONTROL
SAND EXCLUSION
5.1.
Introduction
5.2.
Different Types of Mechanical
SandExclusion
5.2.1.
Open Hole Completions
5.2.1.1. Slotted Pipe
5.2.1.2. A Wire Wrapped Screen
5.2.1.3. Resin Coated Sand Pre-Packed Screen
5.2.1.4. Application of Open Hole Completions
5.2.1.5. Enhanced Drilling Fluid Requirements
for Open Hole Completion’s
5.2.2.
External Gravel Packs
5.2.3.
Internal Gravel Pack
5.2.4.
Special Gravel Packs
5.3.
Advantages / Disadvantages of
GravelPacking
5.4.
Gravel Pack Sand Selection
5.4.1.
Operational Considerations
5.5.
Gravel Packing - Surface Operations
5.6.
Fluids for Gravel Packing
5.6.1.
Properties of Viscous Gravel Packing
Fluids
5.6.2.
5.7.
5.8.
5.9.
6.
2
Other Base Brines
Fluid Loss Control
The Gravel Pack Operation
Gravel Placement with Low Viscosity
Fluids
5.10.
Gravel Placement with High Viscosity
Fluids
5.11.
New Technology
5.11.1. Gravel Pack Evaluation
5.11.2. "Frac and Pack"
5.11.3. New Screen Technology
5.12.
Chemical Sand Consolidation
FURTHER READING
1
LEARNING OBJECTIVES:
Having worked through this chapter the student will be able to:
• Describe the impact of an incorrect decision of whether or not to allow for the
installation of sand control during the original completion design
• Explain the reasons for the lack of definition of “what constitutes a sand problem
in the field”
• State measures taken to modify completion/surface facilities to “live with” sand
production
• Construct a methodology for predicting sand failure
• Discuss the various rock mechanical measurement that may help with this prediction
• Identify and contrast the advantages/disadvantages of the various sand control
options (none, sand consolidation, bare screen, internal and external gravel packs,
“Frac and Pack”)
• Identify the key elements of liner/screen design
• Discuss openhole completions design and drill-in fluid selection
• Discuss the characteristics and functions of the fluid and proppant used in a gravel
pack
• Design a gravel pack be selecting the optimum gravel pack size and fluid for a
gravel pack operation
• Calculate the pressure drops associated with flow through a gravel pack completion
• Describe fluid and proppant placement for the various types of gravel pack options
2
72
Unstable Formations and Sand Control
1. INTRODUCTION
Sand production from (relatively) unconsolidated reservoirs is a frequently encountered,
costly operational problem which has a significant impact on the case of well
operation and the economics of oil or gas production. It is usually associated with
shallow, young formations, but has also been encountered at depths greater than
4000m. Other types of unstable formations also show similar production problems i.e. they allow “pieces” of the formation to break away and enter the perforation or
wellbore. Soft chalks, shales, siltstones and rubble zones can all flow particles,
undergo plastic failure or slough particles due to mechanical formation failure the
stresses imposed by well production.
This chapter will:
(i)
describe the rock mechanical process that lead to sand production,
(ii) discuss what institutes a sand production problem from a production operators
point of view,
(iii) indicate the operational measures taken to identify a sand problem well,
(iv) describe the various types of sand control.
These include sand consolidation and the mechanical methods:
screens / prepacks / gravel packing / frac packing and finally:
(v) discuss the completion operations required for the installation of sand control.
Identification of which wells require the inclusion of sand control equipment,
especially when during (initial) phase of a new field development, is a difficult but
key decision.
Incorrect omission of sand control measures during the completion causes Loss of
Production due to:
• Sand bridges forming in the tubing, sand filling up the casing so that the perforations
are covered and production from the lower intervals is lost
• Damage to the well’s integrity e.g. casing/liner collapse due to loss of lateral
support in the areas where a cavity has formed due to sand production
• Valves and other downhole equipment become stuck due to the presence of solid
particles preventing their mechanism working properly
• Erosion (sand blasting) of holes in the surface pipework leading to the pollution
associated with an oil or gas spill and the consequent loss of well control.
• The produced sand needs to be removed from production equipment and disposed
of in an environmentally acceptable manner.
Department of Petroleum Engineering, Heriot-Watt University
3
1
The unnecessary inclusion of sand control measures results in:
• Extra completion cost (higher initial CAPEX) and loss of future workover flexibility
for the well
• Reduced well production from the impairment brought about by gravel packing
• Increased artificial lift costs due to this reduction in well Productivity Index
• Loss of reserves since the well is taken off production due to the Well’s Minimum
Economic Production Rate, at which the well operating expenses equal income
from the oil production, being reached after less total cumulative production.
The technology developed by the industry for efficient sand exclusion along with a
more detailed discussion of the above will be covered in the following chapters.
2. TYPES OF SAND PRODUCTION
The determination of the sand production potential of a given completion is difficult
because there is no clear picture of the underlying (scientific) mechanism(s) that result
in formation failure and sand production. Formulation of the problem is difficult. The
pay zone or sand body is heterogeneous, displaying a wide range of mechanical and
chemical properties. Drilling of the well and shooting of the perforations will alter the
natural stress regime. Naturally strong formations will accommodate this redistributed
stress; while for weaker sand this stress may exceed the formation strength. To further
complicate the issue, the increased stresses due to fluid flow towards the wellbore,
(frictional forces) as well as pore pressure changes (draw down and/or depletion) may
all contribute to this exceeding the formation strength. Sand production can commence
once the sum of all these forces exceeds the formation strength (figure 1).
the problem area
sand control required
sand control may be required
Low
sand production
not expected
High
Rock Strength
The fact that the forces initiating sand production increase with a rising well
production rate, implies that there can be a critical flow rate below which sand
production will not occur. This has been observed in the field - beaning the well back
4
Figure 1
Is/Will sand control be a
problem ?
72
Unstable Formations and Sand Control
sand production
to reduce the production rate is a field proven method to temporarily alleviate
problems caused by wells suddenly producing sand (figure 2).
maximum
sand-free
rate
background level
0
0
Figure 2
Sand production triggered
by high production rates or
water production
W
at
simultaneous
sand & water
production
er
water cut
sand production
production rate
background level
0
0
time
2.1 A Southern North Sea Case History
Gas production was initiated from this Rotligindes gas reservoir in 1971. All the wells
on the platform started producing sand in 1976 at a FBHP of around 2,600 psi.
However, it was noted that the sand production diminished with time at a constant
FBHP and stopped altogether when the FBHP was increased. The following sand
control strategy was developed over the next 8 years:
• Avoid perforating the weakest rock
Department of Petroleum Engineering, Heriot-Watt University
5
1
• Ensure wells are beaned up slowly
• “Pre-condition” all sand prone wells prior to the (high rate) winter production
season by producing them to an acceptably low sand production rate at a lower
FBHP than is expected to be reached during the forthcoming winter season.
• The sand production will then stop once the production rate is decreased (FBHP
is increased) below the “Pre-conditioning” level.
Field experience showed that routine gas production has been virtually sand free for
many years since 1984. This strategy has the advantage of being cheap does not
require the installation of downhole sand exclusion and has allowed an economic well
production rate. All wells have been produced at the maximum rate during the winter
period and the absence of formation damage associated with types of sand exclusion
has maximised recoverable reserves.
N.B. In common with many Rotligendes gas fields; there is no, or only very limited,
water production at the perforations. The question that the production engineer often
has to answer is: "when is water production expected and will it trigger sand
production?"
2.1.1 "Continuous" Sand Production
Very soft formations may exhibit sand production at the time of initial discovery - the
so called “sloughing” sands can lead to great difficulties during completion
operations when hole collapse can become a major problem. Many other, normally
somewhat more consolidated formations, begin to show sand production after a
considerable period of production - due to reservoir pressure depletion (decrease in
minimum in-situ stress), water production, increased fluid velocities etc. Proper
completion practices are often critical in marginal situations where sand production
can be created by poor completion practices e.g.
increased drawdown + plugged perforations
↓
increased local fluid velocities + potential for water coning
One possible solution is to repeat the perforation process so as to minimise the well
producing drawdown.
Continuous sand production - as discussed above - has to be distinguished from:
2.1.2 “Clean-up” sand production:
This refers to (relatively minor amounts) of sand that are produced from a new well
during its first few days of production. This form of sand production is attributed to
the production of material (partially) de-consolidated by the shock wave produced
during the perforating process. Further, local rock failure may alter the shape of the
initially formed perforation tunnel so that it changes to a more stable shape as a
response to the local stresses. In extreme cases, the change in the perforation shape
may take place to such an extent that the individual perforations merge together to
form a “cavity”. Such “cavity” formation results in the casing becoming unsupported,
leading to casing failure in some cases.
6
72
Unstable Formations and Sand Control
The key point about transient sand production is that it is temporary i.e. it reduces to
a low level after a certain period of time when the well is kept on production at a
constant rate. The time required for stabilisation to occur is dependent on the type of
well - gas wells responding more quickly than light oil wells, which, in turn, respond
faster than heavy oil wells (figure 3).
Production Rate
Sand Production (counts)
a. Gas well
Sand volume ~ 1litre
Interval 20m
0
Time (hrs)
30
30
Sand Cut (pptb)
b. Light oil well
sand volume ~ 35 litre
Interval 10m
20
Bean-up
10
0
10
40
50
Time since startup (days)
500
c. Heavy oil well
sand volume ~ 200litre
Interval 20m after
beam pump artificial
installed
Figure 3
Typical transient sand
production for different
types of wells
Sand Cut (g/m3)
400
300
200
100
0
Time since startup (days)
200
A new burst of sand production is observed each time the well's production is
increased (beaned up). This is thought to be due to formation of a stable arch around
the entrance to a perforation cavity (figure 4). This arch remains stable as long as flow
rate and drawdown are constant. If these are altered, the arch may collapse and a new
one forms once flow stabilises again. To encourage such stable arch formation and
hence prevent sand production, sand-prone wells should be opened up slowly
(possibly over a period of hours or days). Well production conditions should be
maintained as constant as possible.
Department of Petroleum Engineering, Heriot-Watt University
7
1
I I
I
I I
I
Fluid Flow
I
I I
I I
I
I
I
I
I
Formation
I
I I
I
I
I
I
I I
I
I I
I
I
I
I I
I
I
I
I
Casing
I
I
I
I
I
I
I
I
I
I I
I
I
I
I I
I
I
I
I I
I
I
Cement
Fluid Flow
Perforation Tunnel
Fluid Flow
Sand Grain
I I
I
I I
I
I
I I
I I
I
I
I I
I
I
I
Fluid Flow
I
I I
I
I
I
I
I
I
I I
I
I
I I
I
I
I
I
Casing
I
I
I
I
I
I
I
I
I
II I
I
I
I I
I
I
I
I I
I
I
Figure 4
Cavity stabilisation by Arch
Formation
Enlarged perforations
Enlarged perforations
merge to form cavity
increased wellbore radius
Cavity
Fluid Flow
Fluid Flow
Fluid Flow
Fluid Flow
Fluid Flow
Fluid Flow
Debris
The individual perforation cavities may merge into a larger cavity, which may also
stabilise (figure 5).The formation of a large cavity means that the well is no longer
supported by the formation. Complete loss of the well due to casing failure then
becomes a very real possibility. A residual, constant level of solids production may
continue once a stable arch has been formed. This is discussed in the next section.
8
Figure 5
Sand production can lead
to large cavities
72
Unstable Formations and Sand Control
2.1.3 “Fines” Production
Sand production relates to the production of load bearing formation solids while
"fines" production involves the production of mobile, very small, solids which are not
part of the mechanical structure of the formation. These clay, feldspar or silica "fines"
are much smaller than the formation sand grains since they can pass through the pore
throats. Attempting to stop production of these “fines” would drastically impair the
well’s production. Allowing them to pass into the well with the producing fluids
prevents them causing permeability damage at the interface of the gravel pack and the
formation. As a practical “rule of thumb”, the smallest 5% of solids found during
sieve analysis of (soft) cores are probably interstitial solids and can be considered to
be potentially produceable “fines” in the above context.
2.2 Sand Cementation
Sandstone formations were originally laid down as a bed of loose sand grains at the
bottom of a river, or as a beach at the sea shore. Over geological time these individual,
loose grains became cemented or consolidated together - a process which resists sand
production. The individual sand grains making up most sandstone formations are
bonded together by clay, quartz, calcite, mineral growth or precipitate bonding. The
“overburden load” (or weight of all the sediments on top of the formation) is resisted
by the strength of the individual sand grains, the pressure of the fluids within the pore
spaces and the strength imparted to the consolidated formation by these intergranular
cements (figure 6).
Sand Deposit
Burial + Pressure Soloution
Unconsolidated
Cementation
Dolomite
Calcite
Overgrowth: Quartz, Feldspar
0
Pyrite
Figure 6
Sand burial and
cementation process
Grain Coating Clays: Chlorite, Smectite
Ilite, Siderite, etc.
0 0
0
0
0 00
Pore filing
cements
0
0 0 0 0 0 0 00 0 0 0
0
0 0
0
0 0 0
0 0
0 0 00 00 000 00 0 00 00 0 000 0
0000
00
000 0 0 0 000 00 00 0
0 0 00 00000 0
0 0 00
0
0
0
0
00 0
0
0 00 0 0 0 0 0 0 0
0
0
0
0
0
0
0
0
00
00
0 0 000 00 0
0
0
0
0 0 00 0 0
0
00 0
0
0
0
Siderite
Kaolinite
Weak
Department of Petroleum Engineering, Heriot-Watt University
Strong Consolidated
9
1
Cementation is developed in several ways but is most commonly due to precipitation
of minerals from the water phase within the pore space between the sand grains. The
localisation of the cementing precipitate at the grain-to-grain contact is caused by a
pressure dissolution and precipitation cycle. This is brought on by the application of
(high) overburden load on the small area of the point-to-point contacts of the
individual sand grains. At the high pressures associated with the point-to-point
loading, dissolution of the contacting surfaces of the sand grains takes place since the
sand (quartz) is more soluble at the point of loading. The water surrounding the active
sites becomes over-saturated in quartz, compared to the unloaded part of the sand
grain. As time passes, the contact points of the grains become flatter, spreading the
load over a wider area of total contact i.e. the local point loading decreases.
Eventually, the increased contact area created by this method is sufficient to withstand
the overburden and the form of the individual sand grains is stabilised. Precipitation
of the super- saturated matrix minerals in the water, occurs on the sand grain surfaces.
The result is that the grains are “cemented” together. This pressure solution driven
"dissolution and precipitation" process at the grain contact is only one method of sand
grained cementation. Additional amounts of the same or different cements may be
precipitated from flowing hydro thermal fluids. Mineral cements may be precipitated
as the water in the pore space comes to equilibrium with the local conditions of the
temperature, pressure and mineral composition. Quartz overgrowth around calcium
cementation or clay development at the grain boundaries is a sign of secondary
cementation.
Formations may not consolidate for a number of reasons, e.g. reduced compaction
loading caused by shallow burial or very large, load supporting arches above the pay
(e.g. grabens). Faults may block the necessary stresses and leave high porosity and
poor intergrain bonding. If the sealing structure is filled with hydrocarbon shortly
after formation, mineral solution or precipitation reactions from water cannot take
place. High pore pressure (geopressures) also reduces the compaction loading; in this
case the pore fluid carries a greater percentage of the overburden supporting load.
Other types of chemical reactions such as ground water leaching of the matrix grains
or dolomitization, may heavily modify the cement or the size of the sand grain.
2.3 When is Sand Production a Problem?
The well’s sand production tendency may change during the lifetime of the well. In
particular, what was a negligible or marginal problem may become much more severe
late in field life due to:
(i)
Water Production. This results in:
(a) higher drawdowns (increased rock stress) for some production rate due to
relative permeability effects.
(b) loss of capillary pressure between the sand grains. It can be shown theoretically,
or from experimental measurements, that this does not significantly reduce the
rock strength. However, it does allow the already failed (loose) rock particles
to be produced by the forces associated with fluid flow.
10
72
Unstable Formations and Sand Control
(ii) Production Plans that call for maintaining the net oil production. This results
in a continually increasing gross fluid production as the water cut increases. i.e.
production occurs at an ever increasing drawdown and fluid flow velocities.
(iii) Reservoir Pressure Depletion results in further increases in stress on the rock.
Avoiding perforating the weakest rock may reduce the well’s sanding tendency.
However, a decision still has to be made whether to control sand production (the safe
option) or live with it and risk the (later) consequences of sand production, as
described in the next chapter.
2.3.1 The Consequences of Sand Production
Sand production has numerous technical, environmental, operational and economic
consequences:
The operationally acceptable level of sand production will vary with the location, the
well and production facility design and local conditions. The effects of sand on
production operations and the potential safety, financial and environmental
consequences will influence whether sand production limits are set and their level.
Potential Consequences of Sand Production
Location
Problem
Wellbore fill from bottom
Completion interval
• Restricted access to production
interval
Cavity formation
• Reduced work over success since
zonal isolation no longer practical
• Increased well productivity Index
Casing buckling
• Loss of full diameter access to
completion interval
• Casing failure in extreme case
Tubing
Sand bridge
Subsurface equipment
Sand deposition
in tubing and accessories
• Sub surface, surface controlled,
safety valve (SSSV) not operating a safety issue
• Difficult Wireline operations
Erosion
• Equipment replacement and failure
Sand accumulation
Surface installations
Table 1
Potential Consequences of
Sand Production
Effect
Erosion
Department of Petroleum Engineering, Heriot-Watt University
• Well production halted
• Loss of productivity and reserves
• Control equipment malfunction
• Reduction in separator residence time
• Unscheduled shut-downs/deferred
production
• Sand separation from hydrocarbons
and disposal
• Failed equipment replacement
• Oil/gas spill (environmental/safety
issue)
11
1
Unconsolidated formations are rarely homogeneous and apparent formation strength
is affected by changes in depletion, water saturation and pressures over the life of the
field. This results in one of the biggest problems for predicting sand movement and
for designing control methods for new field developments The flow rate that will
produce sand in one interval may cause no damage only a foot away and the control
process that will halt the flow of sand in a zone with large formation sand grains may
not work in zones made up of smaller grains. The whole formation must often be given
the same treatment as the worst-case zone. Remedial Sand control measures are
required in formations where large sections of the hydrocarbon bearing sand are weak.
However, in formations where only a small percentage of the zone presents a risk of
sand production, selective perforating can often avoid the need for installation of sand
control. The answer is to determine which sections of the pay zone sand will fail.
2.3.2 Living With Sand Production
A sand management system has to be installed if the “living with sand” option is
chosen. Further, the production system may be changed so that it becomes more
tolerant of the volumes of sand that are produced. The measures taken could include:
(i)
measurement of the amounts of sand produced
(ii) installation of hard faced chokes in the “bean” box
(iii) installation of appropriate artificial lift methods e.g. gas lift with no moving
parts through which the oil flows - is more sand tolerant than electric submersible
pumps (esp) with their rapidly rotating impellers.
(iv) monitor flow line wall thickness. An X-ray or sonic measurement device can
be used for this purpose. The same techniques may be used to monitor sand build
up in surface vessels (figure 7)
Ray Path
Radioactive
source
Movable
Detector
SAND
high velocity water jets
fluidises sand at
bottom of vessel
(v) design flow line so that large changes in velocity and direction do not occur e.g.
remove 90° pipe bends.
(vi) install surface sand collection and disposal facilities e.g. water jets installed at
the bottom of separators. The high velocity water stream fluidises the sand and
transports it to a collection and disposal vessel. Here, the adhering oil may be
removed by vigorous agitation with water (and surfactants) to allow easier subsequent
clearing and disposal.
12
Figure 7
Monitoring sand level in
vessel through increased x-ray
absorbtion and removal of
settled sand.
72
Unstable Formations and Sand Control
(vii) increase the well production in a number of steps (slow bean up) and avoiding
cycling the downhole pressure.
(viii) ensure that fluid velocities are sufficient to transport sand to the surface and
transport it to an operationally convenient collection point. This is done by
calculating the terminal settling velocity (Stokes law) and comparing this with the
minimum (upward) flow velocity. Sand particles will concentrate at points at which
the settling velocity is greater than the flow velocity. The settling velocity increases
with increasing particle size and decreases with increasing fluid viscosity (figure 8).
N.B. The settling velocity will change as the fluid composition changes e.g. it will
decrease when (viscous) emulsions are formed or increase in the presence of water
continuous oil-in-water dispersions formed at high (~>50%) water cuts.
0.25
0
Terminal Settling Velocity (m/s)
50 API
0.2
Suspending
Fluid
(Crude Oil)
0.15
35 API
0
0.1
0
0.05
Figure 8
Terminal settling velocity of
sand particles in different
crude oils
0
20 API
0.1
0.2
0.3 0.4
0.5 0.6
0.7
0.8 0.9
1
Sand Particle Size (mm)
It is not practical to live with unlimited amounts of sand production. Typical
operationally allowable levels are summarised in table 2.
Typical Allowable Sand Production Levels
Table 2
Typical Allowable Sand
Production Levels
Produced Fluid
Production Rate
Allowable Sand Level
Gas
(<50 Mscf/d)
(>50 Mscf/d)
1 lbs/MMscf
0.5 lbs/MMscf
Light Oil
<5000 bopd
5000-15000
>15000
30 lbs/1000 bbls*
10 lbs/1000 bbls
5 lbs/1000 bbls
Heavy Oil
Department of Petroleum Engineering, Heriot-Watt University
* reduced by 50% for high
GOR
200 lbs/1000 bbls or even
higher
13
1
It can be seen that the allowable sand content is very dependent on the fluid velocity
(production rate) and fluid viscosity.
2.3.3 Monitoring of Sand Production
The various techniques used for monitoring sand production are summarised in table
3. This table indicates the measurement principle employed, its key points and the
associated drawbacks.
MONITORING OF SAND PRODUCTION
Technique
Principle
Measurement
Drawbacks
Wellhead Sampling
Centrifuge wellhead
produced oil sample
Simple
Not continuous;
Non-representative
Manual
In-Line Sand Trap
Sand settling into pots
located in flow lines
Simple
Allows some estimation of sand
at flowline production rate
Manual
Erosion Probes
Hollow probe inserted into
flowline: penetration of probe wall
by sand erosion activates alarm
since flow line pressure appears
inside probe
Erosion measurement.
Depends on sand production
rate,flow rate, type of
produced fluids,
probe position etc.
Alarm system only; install where
severe erosion expected
Sonic Detectors
Acoustic noise of solids impact
measured by a piezo electric crystal
which is mounted on a
solid probe which is mounted in
the flowline
Detects mass rate of sand
produced i.e. larger particles
make bigger impact
Interference from flow noise gas
bubbles and small solid particles
(fines)
No erosion measurement
Flowline Erosion
Measure wall thickness by ultra
Sonic / x-ray techniques
Erosion detection only
Not on-line or continuous
Expensive
3. PREDICTION OF SAND FAILURE
A number of techniques have been employed when deciding whether to install sand
exclusion techniques. These include:
(i)
Field experience
(ii) Petrophysical log and core analysis
(iii) Wellsite rock strength estimation
(iv) Rock mechanical measurements and calculation
Unfortunately, none of these techniques yields a perfect answer, as will become
apparent in the following sections.
3.1 Field Experience
The production history of wells producing from the same formation in the field where
the new well is planned is the best guide as to whether sand control should be installed.
Alternatively the same formation may be found in a nearby field with the same
geological history.
14
Table 3
Monitoring of sand
production
72
Unstable Formations and Sand Control
3.2 Petrophysical Analysis
Laboratory measurements on core material from the same geological section typically
show that:
xx
x xx
x x
x
x xx
x x
x x
xx
Cloud of experimental
measurements
x
x
x
x
xx
x xx
x
x x
x xx
x x
x x
xx
thick wall cylinder strength (bar)
or unconfined compressive strength (bar)
(i) The sonic travel time (∆T, µs / ft) is proportional to the porosity i.e. the higher
the porosity, the longer the travel time (figure 9)
Average correlation line
porosity
xx
x xx
xx
xx
x xx
x xx
x
x xx
x xx
x x
x x
xx
sonic travel time ( µs )
x
x
xx
x xx
x xx
x xx
x x
x x
xx
Cloud of experimental
measurements
Average correlation line
Figure 9
Porosity is linearly related
to sonic travel time and
thick wall cylinder strength
porosity
(ii) The rock strength (as measured by an unconfined compressive strength test or
thick wall cylinder collapse test - see rock mechanical tests, section 7.3.3) is
inversely related to porosity (Figure 9), i.e. the lower the porosity, the greater the
rock strength.
Department of Petroleum Engineering, Heriot-Watt University
15
1
N.B. These correlations are in the form of a trend line drawn through a cloud of data
points. This inexactness is due to the heterogeneity of the formation properties and
the errors in the measurements themselves.
Total well pressure drawdown since discovery
A,B producing conditions for
wells A and B on initial
completion
A’, Abandonment conditions for
well A, NO sand failure predicted
B’, Abandonment conditions for
well B, sand failure possible
A’
B’
A
B
SAFE
RISK
Sonic travel time (µs/ft)
Data points observed from field experience
sand failure.
no failure
NB Position of boundary between "safe"
and "risk" regions varies from field to field
The well's sonic travel time log can thus be processed to derive a continuous estimate
of the formation strength. This allows the identification of the weakest sandstone
which can be left (selectively) unperforated. The method can be extended using the
field observation that the vertical stress near the wellbore i.e. overburden minus
flowing total drawdown (reservoir depletion + near wellbore pressure drop due to well
production) at which continuous sand production is first observed is related to the
acoustic travel time (Figure 10). It uses the acoustic velocity as a strength indicator
coupled with the laboratory finding that the increase in insitu rock stress due to
pressure depletion of the reservoir as a whole and due to the near wellbore producing
drawdown are equivalent in terms of contributing to sand failure.
The boundary between the “safe region” (where sand production is not expected)
and the (risk region) where it may occur - is based on field experience. It varies from
one field to another. It can also be used to predict:
16
Figure 10
Field experience rock based
approach to sand failure
prediction.
72
Unstable Formations and Sand Control
(i)
whether sand production is expected at initial completion
(ii) the earliest time that sand failure can be expected
i.e. when a well's position crosses into the “risk” region. This uses a combination of
the results from the field reservoir engineering model together with production inflow
modelling of the well itself.
It can be seen that this approach requires a lengthy production history together with
actual sand failure in one or more wells. If this is not available; a high drawdown/high
production rate test can be designed to simulate future producing conditions.
3.3 Rock Strength Measurement
3.3.1 On Site Strength Estimation
The simplest approach is attributed to D. Sparlin who stated that “a potential sand
problem can be expected if the core is friable (finger nail makes groove) or weaker”.
This judgement can be extended using table 4 which relates the rock classification to
simple strength observations made on the core to the approximate sonic travel time
and the core recovery. The two strength measurements also included in the table will
be discussed in the next section.
3.3.2 Rock Mechanical Strength Measurements
A number of rock mechanical tests are used to characterise the rock. They include:
3.3.2.1 Unconfined Compressive Strength
An unsupported cylinder of rock with a length to diameter ratio of at least 2:1 and a
plan parallel end is loaded axially at a steady rate of between 0.5 - 5 MPa/min to failure
(Figure 11). The compressive failure is in the form of diagonal fractures across the
body of the sample. The maximum stress reached is called the Unconfined or Uniaxial
Compressive Strength (UCS).
σA
σA
σUCS
od
e
Load Plattern
Fa
ilu
re
M
Failure
Figure 11
Unconfined Compressive
Strength measurement
End Platten
Unsupported cylindrical
rock sample
Department of Petroleum Engineering, Heriot-Watt University
Time
Load sample at 0.5-5MPa/min
17
1
3.3.2.2 Brinell Hardness Number
The Brinell Hardness Number (BHN) is the load required to press a standard spherical
indenter a constant distance into a slabbed core face (Figure 12). The stronger the rock,
the greater the load required and the higher the BHN. The Brinell Harness Number
(BHN) measurement is a straightforward method of classifying the strength of
sandstone as shown in table 4.
σBHN
Slabbed Core
Figure 12
Brinell hardness
measurement
ONSITE ROCK STRENGTH ESTIMATION
Uniaxial
Compressive
Strength (psi)
Brinnel
Hardness
Number
2
(kg / mm )
Rock Classification
Core Observation
Approximate
Sonic Travel
Time
(µs / ft)
Core Recovery
0
0
Quicksand
Hole Slumps
>150
zero
<1000
<2
Unconsolidated
No apparent cement
between sand grains
>145
sleeve (poor )
2
conv. (Zero)
1000-2500
2-5
Semi-Consolidated
Easily Crushed
>130
sleeve (good)
2
3
conv. (poor )
2500-3500
5-10
Friable
Rub-off Grains
105-130
conv. (good)
3500-7500
10-30
Consolidated
Crushable with forceps
105-175
excellent
7500-12000
30-50
Moderate Hard
Cannot Crush
65-75
excellent
12000-20000
50-125
Hard
Cannot Crush
40 / 65
excellent
1
3
1
2
1
sleeve refers to use of rubber sleeve core barrel to support cored material
2
conv. refers to use of conventional steel core barrel (no support for core)
3
great care required during coring - prevent core barrel jamming, avoid excessive mud velocities etc.
Apart from the unconsolidated sands (BHN < 2kg/mm2), all other sands in the
classification may exhibit post-failure stabilisation, following the onset of initial sand
production. The phenomenon will be dependent on the frictional strength characteristics
as well as the degree of cementation. In this context, it is important to note that the
BHN is dominated by the degree of grain-to-grain cementation, and by the intergrain
frictional strength to a much lesser extent.
BHN is used because it is linearly related to, but much quicker and easier to measure
than the Thick Wall Cylinder collapse strength (TWC) discussed in the next section.
Thus rapid, point measurements can be made at many depths on a slabbed core and
a hardness (strength) profile created.
18
Table 4
On site rock strength
estimation
72
Unstable Formations and Sand Control
3.3.2.3 Thick Wall Cylinder Collapse Strength
A hollow, thick wall cylinder is created by drilling a narrow hole in the middle of a
rock cylinder similar to that described in the unconfined compressive strength test. A
rubber sleve is mounted around the outside of the cylinder which is then mounted
between two end caps (figure 13). The hole in the centre of the cylinder is maintained
at atmospheric pressure. The stress applied to the solid end cap and the outside of the
sleeve are kept the same and increased at a steady rate of 0.5 - 5 MPa/min. The stress
at which the cylinder collapses (implodes) is the Thick Wall Cylinder Collapse
Strength (σtwc).
The TWC test was developed because it replicates of the geometry of perforation
failure (see figure 14)
End Cap
Pressure (P)
Pressure (P)
Hollow Cylinder Core Sample
Sleeve
End Cap
Pressure (P)
Pressure (P)
σtwc
Failure
Increase
load at
0.5 - 5 MPa/min
Figure 13
Hollow (thick wall) cylinder
collapse
Time
Department of Petroleum Engineering, Heriot-Watt University
19
Wellbore
1
Pe r f o
ration
Pe r
on
fo r a t i
Laboratory measurements, confirmed by field experience, have shown that INITIAL
sand failure occurs when the near wellbore vertical effective stress at the perforation
depth equals the TWC strength.
The frictional strength characteristics of the rock will depend on grain size distribution,
porosity and degree of roundness, which for a specific field, will be related to the
depositional environment. The frictional characteristics will have a major impact on
the TWC strength. In general, sands with a low degree of sorting, and sub-angular
grain geometrics, will exhibit higher strengths than sandstones composed of wellrounded grains, for similar degrees of cementation.
3.3.2.4 Triaxial Rock Strength Measurement
This is the most sophisticated of the rock mechanical tests that will be discussed here.
A cylinder of rock is mounted in a sleeve. The axial stress (σa) - imposed by the end
pieces - and the radial stress (σr) are controlled separately. The axial strain is measured
by strain gauges. Figure 15 shows that the maximum axial stress reached prior to
failure increases as the radial stress is increased i.e. the rock sample shows stronger
behaviour as the confinement stress increases. Much more information can be derived
about the strength properties of the rock from this test compared to those described
earlier - this is illustrated in figure 16, where the various phases of rock failure under
influence of the imposed stresses are illustrated.
20
Figure 14
Crushing of thick walled
cylinder simulates
perforation performance
72
Unstable Formations and Sand Control
r = 20 MPa
r = 10 MPa
Axial
stress
r = 5.0 MPa
A
r = 2.5 MPa
a
r=0
r
Axial strain
Figure 15
Triaxial test strength
measurement
Traxial stresses
imposed on core
C
well developed
fracture system
(sample fails)
Axial stress
B
Figure 16
Elasto-Plastic material
response involves the
formation of fractures
during sample testing to
failure
onset formation of
small fractures
A
B
B
C
C
Elastic
Elastic-plastic hardening
Elastic-plastic softening
A
Axial strain
3.4 Prediction of Downhole Rock Failure
(Sand) failure of rock in a producing borehole is a complex process. Characterisation
of the downhole properties of the rock is a complex subject that was briefly touched
on in the paragraphs above. Similarly, (sand) failure of rock in a producing borehole
Department of Petroleum Engineering, Heriot-Watt University
21
1
is complex - a number of possible failure modes are sketched in figure 17. Simple,
analytical descriptions of some of the processes are available, but tend to be (highly)
conservative. This is because they represent initial failure of the perforation tunnel but
do not capture the re-stabilisation processes such as cavity enlargement which gives
reduced flow velocities and drawdowns, which delay the onset of unacceptable sand
production. Development of complete descriptions of these processes require the use
of complex numerical computer code. The actual results are dependant on:
Failure
mode :
Shear or
Compressive
Tensile
Flow
Erosion
Flow
Flow
Flow
Flow
Cause of
Failure :
(i)
Far field stress
+ drawdown
Drawdown
Flow
the rock failure model used.
(ii) the type of laboratory rock strength measurements made.
(iii) the accurate forecasting of the producing well conditions.
(iv) the availability of core material representative of the failed section of formation.
This latter point is often problematic - the weakest formation is likely to fail first.
However, this is the least likely to be recovered by coring!
3.5 A Final Word
The prediction of sand failure from basic principles i.e. when field history data is not
available is a difficult process. The incentive for doing this is considerable, due to the
high cost of unnecessary installation of sand control (see next section). Management
of this process by use of a risk based approach to the timing of sand control installation
as described is the preferred route - however, when estimating the (Net Present Value)
economics of possible schemes, it should include the fact that field experience shows
that the installation of Remedial Sand Control in a well that has already undergone
sand failure has a lower chance of success than the equivalent completion operation
in a new well.
N.B. Proper estimation of the chance of success of both operations depends on local
field experience!
22
Figure 17
Possible rock failure modes
for a producing borehole
72
Unstable Formations and Sand Control
4. THE COST OF SAND CONTROL
High well productivity losses - typically 60% - are observed after well killing followed
by gravel packing operations - see figure 18. This figure summarises the result of a
field test in which the well’s productivity index was measured at various stages in the
completion process. The final well productivity, despite a 6-month clean up period,
was only 18% of that achieved prior to gravel packing. This high impairment has been
noted in many gravel packed wells completed in different fields.
40
31
PI = Productivity Index (bbl/psi day)
30
20
10
10
4.4
2.6
0.7
Figure 18
Field measured changes in
well productivity
0
Stage:
perforate
%
original 100
PI
+ve
kill well
gravel
pack
produce
45 days
produce
180 days
32
2
8
14
reduced (Present Value)
due to cost of
remedial sand control installation
delay in remedial sand
exclusion installation
No sand failure
Sand failure
Discounted
cumulative
cash surplus
Sand failure and
remedial exclusion
Sand control
installed during
initial completion
0
Low well productivity
Figure 19
Manage risk of sand failure
by delaying instalation of
sand exclusion
extra cost
installation
sand control
-ve
Time
well
construction
starts
well
production
started
Department of Petroleum Engineering, Heriot-Watt University
23
1
It implies that it can often be advantageous to delay installation of sand control i.e. on
a project cumulative cash surplus basis the reduced cost of completion plus the higher
initial well productivity more than compensates for the later lost production and
remedial costs to remedy sand failure (figure 19). The "time effect of money" reduces
the later cost of the installation of sand control measures on a discounted cash flow
basis.
There is a varying efficiency for the different types of sand control that can be
installed. Typical well productivity and the implications for the various sand control
options for a West African field study are:
Sand
Exclusion
Type
Average Well
Productivity
Index (b/d/psi)
No. of Wells
Required For
Field
Development
Gaslift
Requirements
(MMscf/d)
Oil Recovery at
Field
Abandonment
Water Cut
20% 50% 80%
Internal Gravel
Pack
5
60
0
5
15
External Gravel
Pack
15
55
0
0
5
None or
Chemical
Consolidation
30
30
0
0
5
Lowest
Highest
These different types are illustrated in figure 20. They will be explained in the
remaining chapters of this module. However, as can be seen from the table above,
installation of sand control:
(i)
increases the number of development wells
(ii) increases the capital costs per well
(ii) requires earlier installation of gas lift
(iv) results in higher abandonment pressures and reduced oil recovery
5. SAND EXCLUSION
5.1 Introduction
There are essentially two types of sand exclusion:
(I) Mechanical techniques where “gravel” particles, a few times larger than the
formation sand grains, are used to retain the formation in place by forming a filter
through which the formation sand cannot pass. The gravel is itself held in place by
a screen which has been sized so that it in turn can not pass through the gaps (figure
20). In its simplest form, the gravel is omitted and the screen alone “holds back” the
formation.
24
Table 5
Example: Impact of
completion efficiency on
Field Development
72
Unstable Formations and Sand Control
(ii) Chemical Techniques where a chemical cement increases the strength of the
formation while retaining a permeable pore structure.
The various types of sand control are illustrated in Figure 20. These will all be
described in detail in the next chapters. However, one can deduce from this figure that
their installation is a complex process, which would only be entered into if the
alternative, allowing the tubing or casing to fill with sand were not a practical
proposition. The following exercise shows that this is not a practical option; even if
the sand fill in the tubing is only 1m deep:
Strengthened Zone
SAND
CONSOLIDATION
Producing Interval
Cement
Packer
Screen
"FRAC PACK"
Gravel
Perforation
INTERNAL
GRAVEL
PACK
Casing
Screen
Producing Interval
Figure 20
Sand control variations
(open hole completions not
illustrated).
Department of Petroleum Engineering, Heriot-Watt University
Gravel
EXTERNAL
GRAVEL
PACK
25
1
EXERCISE
Estimate the oil flow through a 100 cm long sand bridge in a 3.5 in OD tubing when
there is a 100 bar pressure drop across it.
This exercise involves the application of Darcy’s Law for single phase, incompressible
flow in porous media. If the gravitational effects are neglected:
Q = k A ∆P
µ
L
(1)
where :
Q = flow rate of fluid (cm3 / sec)
A = Tubing Cross Sectional Area (cm2)
µ = fluid viscosity (cp)
L = Length of Sand bridge (cm)
k
= Permeability of sand in the bridge (cm2)
∆P = Pressure Drop across the sand bridge (atm)
The permeability of most formations are less than 3 Darcies, so it is reasonable to
assume the sand bridge permeability of 1 & 10 Darcies. The tubing ID is typically
2.6 in or 6.6 cm and we will assume that the oil has the same viscosity as water (1
cp). Substituting in Eqn 1 gives:
k / (Darcy)
Q / (cm3 / sec)
Q / (m3 / day)
1
34
3
10
342
30
Even though all the assumptions made have been optimistic in terms of maximising
the flow through the sand bridge (high permeability and available pressure drop (100
bar, low fluid viscosity); the resulting oil production is low. Thus as soon as either
a:
(i)
sand bridge is formed or
(ii) a perforation is covered by sand fill
then production is essentially lost from below that point in the well. i.e. Fluid can flow
radially into the well through formation sand; but pressure drops immediately become
unacceptably high when linear flow through sand within the well takes place.
5.2 Different Types of Mechanical Sand exclusion
5.2.1 Open Hole Completions
The basic configurations for mechanical sand control are shown in Figure 20. The
sand exclusion system is either installed below the bottom of the lowermost casing or
liner (see Figure 20 bottom for an external gravel pack). The advantages and
disadvantages of the various sand exclusion systems are summarised in table 6. The
systems are described in detail as follows:
26
72
Unstable Formations and Sand Control
5.2.1.1 Slotted Pipe
This consists of steel pipe (e.g. tubing) where a series of parallel slots have been cut
through the metal (figure 21). The width of these slots are normally made as small as
mechanically practical so that they will retain as large a fraction of the formation sand
as possible. The inflow area is low (2-3% of pipe surface area). Sand grains which
are small enough to pass through the slot can still form a stable arch around the slot
in a similar manner to that described in section 2.1.3 where arch formation around a
perforation was discussed. It is mainly used as a low price option to reinforce an open
borehole and to retain a coarse grained formation; although narrower, laser cut slotted
pipe is now becoming available (expensive).
Figure 21
Slotted pipe used for Sand
Control
Table 6
Comparison of Liner and
Screen Characteristics
ITEM
SLOTTED LINER
(Mild Steel)
WIRE WRAPPED SCREEN
(Stainless Steel)
PRE-PACKED SCREEN
Resin Coated Sand
Description
rectilinear slots /
machined in pipe
wire wielded to longitudinal rods
gravel sandwiched between two
wire wrapped screens
Concept
wellbore reinforcement;
sand bridges around slot
formation sand exclusion or gravel
retention
gravel provides sand exclusion
Material
mild steel
stainless steel on mild steel base
pipe
stainless steel on mild steel base
pipe
Sand exclusion
poor: 0.012” slot width
minimum
better than Slotted Liner since
slot width 0.006” - 0.040”
excellent: as with gravel pack
Works with
gravel pack
yes
yes
yes, but should not be necessary
Flow restriction
high
low, ≈ 10 times flow area of
slotted liner
high, as for wire wrapped screen
Mechanical
resistance
good
poor to collapse / tension if base
pipe omitted. Also susceptible to
erosion
fair: base pipe reinforces structure
Plugging
tendency
low (too wide to retain
formation sand)
moderate
high: fines + mud cake. Also
impairment while running in hole.
Cost
cheapest
2-3 x slotted liner
2-3 x wire wrapped screen, but
often less than gravel pack
Application
borehole reinforcement
coarse grained formation
higher productivity wells medium
grained formation.
allows fines production
retains sand grains of all sizes
5.2.1.2 A Wire Wrapped Screen
This consists of a triangular shaped wire which is carefully wound so that there is a
constant gap between successive turns (figure 22). It is held in place by spot welding
the wire to vertical formers placed at 1cm intervals around the internal diameter of the
screen. Wire wrapped screens have the advantage over a slotted liner that the gap
between the wires can be made smaller and be held to the target value with a much
greater accuracy; allowing the screen to retain finer grained formations than the
slotted liner.
Department of Petroleum Engineering, Heriot-Watt University
27
1
Vertical former
(for strength)
Stainless steel wire
in triangular shape
Gap kept constant
Key form allows any
sand grain that passes
narrowest point to be
flushed from slot
Spot weld
Figure 22
Wire wrapped screen
Wire wrapped screens also have a much greater inflow area - making them more
suitable for higher productivity wells with a greater inflow rate per unit completion
length. Wire wrapped screens have to be handled carefully at the rig site - their
strength is much lower than slotted pipe (figure 22 omitted the perforated base pipe
which is often used to increase the screen’s strength. This base pipe has been included
in the cross sectional view in figure 23).
They are ideally suited for use in higher productivity wells where it is required to retain
a medium grain sized formation.
triangular
wire
perforated
base pipe
connection
28
Figure 23
Cross section of wire
wrapped screen showing
perforated base pipe inner
support
72
Unstable Formations and Sand Control
5.2.1.3 Resin Coated Sand Pre-Packed Screen
Pre-packed screens are constructed from two concentric screens with a layer of gravel
placed in between them (figure 24). The gravel had been coated with a layer of
thermosetting resin. The construction process is as follows:
(i)
the dual concentric screens have been welded onto the base pipe
(ii) the gap between them is filled with the resin coated sand and the final welds
made
(iii) the completed screen is placed in an oven where the thermosetting resin,
hardens creating a strong ring of gravel
The pore throats of the consolidated gravel provide a series of narrow openings which
provide the sand exclusion and retain the formation in place. The presence of the
gravel with its narrow pore throat diameter provides a greater flow restriction than the
wire wrapped screen alone; as well making the screen susceptible to plugging by
formation fines etc. The greater complexity of the prepacked screen increases the
cost.
Outer
Screen
Gravel Consolidated in Place
by Chemical Cement
(Large enough that
it cannot pass through
the gaps in the screen)
Inner Screen
(wider gap than outer screen)
Perforated Base
Pipe for Strength
Connection
Figure 24
Pre-packed screen
yyyy
,,,,
yyyy
,,,,
yyyy
,,,,
yyyy
,,,,
yyyy
,,,,
yyyy
,,,,
,,,
,,,
,,
,,,
,,,
,,
,,,
,,,
,,
,,,,
,,,,
,,,,,
,,,
,,,,
,,,,
,,,
,,,
,,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,
,,,
,,,
,,,,
,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,
,,,
,,,,
,,,,
,,,,,,
,,,
,,,,
,,,
,,,
,,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,
,,,
,,,,
,,
,,
,,,
,,
,,,,
,,
,,,
,,,
,,
,,
,,,
,,
,,,
,,,,
,,,,
,,,,,,
,,,
,,,,
,,,
,,,
,,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,
,,,
,,,
,,,,
,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,
,,,
,,,,
,,,,
,,,,,,
,,,
,,,,
,,,
,,,
,,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,
,,,,
,,
,,,,
Formation Sand too small
to pass through pores
in gravel)
5.2.1.4 Application of Open Hole Completions
Open hole completions have become much more popular since horizontal wells
became widespread. Their use avoids the high cost and technical difficulties in
cementing and perforating long (up to 2km) horizontal liners and casings. Some of
the problems - and opportunities - presented to the completion design engineer when
designing horizontal well completions are illustrated in figures 25 - 28.
Department of Petroleum Engineering, Heriot-Watt University
29
1
13.375"
9.625"
7"Pre-perforated liner
Figure 25
Conventional completion,
no Sand Control
13.375"
9.625"
5.5"wire-wrapped screens
Figure 26
Conventional completion
with wire wrapped screen
for Sand Control
9.625"
ECP
7"Slotted liner
stage cementing collar
Figure 25 shows a “conventional” well design with the production casing cemented
in place just above or just into producing formation. The 7” perforated liner provides
support to the open hole but does not provide sand control. Sand control is provided
in the completion shown in figure 26 in which the 7” pre-perforated liner is replaced
by a (smaller) 5.5” diameter wire wrapped screen. This change could increase the well
cost by typically 10 - 20%.
Formations consisting of coarse sand grains allow this well cost to be decreased by
substituting a 7” slotted liner for the wire wrapped screen (figure 27). The well cost
can be further decreased if the use of an external casing packer possibly backed up by
cement placed via a stage cementing collar can give sufficient (zonal) isolation of the
producing interval from shallower formations. This low cost option gives a typical
well cost of only 60% of that for above wire wrapped screen case.
30
Figure 27
Low cost option
incorporating sand
exclusion
72
Unstable Formations and Sand Control
Tubing retrievable
Surface controlled subsurface safety valve
5 1 / 2 " Tubing
Gas lift valve in side pocket mandrel
4 1 / 2 " Tubing
Sliding side door (allows circulation)
Permanent guages with surface read out
9 5 / 8 " Packer (ca 60˚)
ECP
ECP
Wireline
nipples
Polished bore and
elastomer seals
Figure 28
Horizontal well completion
9 5 / 8 " Shoe (90˚)
8 1 / 2 " Horizontal hole
7" Slotted
Liner
A full Horizontal Well Completion is pictured in figure 28.
It incorporates (from the top):
• tubing retrievable surface controlled, sub surface safety valve (SCSSSV) so as to
be able to close the well in an emergency
• 51/2" tubing reducing to 41/2" tubing in the bottom section - where the accessories
(gas lift valves etc.) are installed. The use of 51/2" tubing maximises the flow
capacity while a 51/2" accessory would have too great a diameter for the casing.
Hence 41/2" accessories (and tubing just above the packer) are used.
The accessories consist of:
(i)
gas lift valve installed in a side pocket mandrel
(ii) sliding side door which can be opened to allow fluid circulation between the
tubing and the tubing / casing annulus
(iii) permanent gauges connected to surface for continuous monitoring of downhole
producing conditions
(iv) wireline nipples where plugs or other devices can be placed
(v) polished bore and elastomer seals. This allows tubing expansion or contraction
due to the temperature changes to be taken up. The tubing can also be (partially)
retrieved to replace the SCSSSV as required without removing the packer.
The open hole section is completed with a 7" slotted liner in a 8 1/2" drilled hole.
External casing packers are included at regular intervals so that sections of the
completion interval can be isolated at a later date to stop extraneous gas or water
production. Removal of the ECPs allow a gravel pack to be placed (Figure 28a).
Gravel packing of long horizontal wells is a technology that has now matured.
Department of Petroleum Engineering, Heriot-Watt University
31
1
5.2.1.5 Enhanced Drilling Fluid Requirements for Open Hole Completion’s
Open hole completion’s with sand exclusion place special requirements on the drilling
fluid properties. This is illustrated in figure 29a which is a schematic illustration of
the results of large scale experiments in which a mud cake was deposited in a simulated
section of a horizontal well. An acidisation treatment removed the mud cake covering
the upper part of the hole but left extensive mud cake remnants on the lower sections.
Simulated production from the reservoir into the well (figure 29b) resulted in more
than half the screen area being covered by mud cake, reducing fluid flow.
Remedial treatments to dissolve mud cakes are a poor option since:
Mudcake
Formation
Screen
Mudcake fallen on
top of screen
Screen
(Partially dissolved)
mudcake deposited at bottom
of horizontal hole after becoming detached
from formation face following breaker treatment
(i)
they are expensive
(ii) they tend to destabilise an already weakly / un-consolidated formation resulting
in permeability impairment
(iii) It is difficult to ensure that the complete length of the open hole interval is
treated with the dissolving fluid.
Remedial treatments can be avoided and higher productivity completions achieved by
selecting the drilling fluid so that:
(i)
32
the mud cake is easily and evenly lifted from the borehole when the well is
Figure 29a
Mud/mud cake deposits on
lower side of horizontal well
section observed after
displacement/breaker
treatment
Figure 29b
Screen appearance after
breaker treatment to remove
mud cake
72
Unstable Formations and Sand Control
placed on production. Drawdown of a horizontal well is often low and decreases
from the heel of the well towards the toe. ONLY if the “filter-cake-lift-off” pressure
is low will it have a chance of being removed from the complete length of the well.
(ii) the mud cake should break up and flow easily between the narrowest gaps in
the installed sand exclusion equipment.
This process can be accelerated by choosing (some of) the mud cake constituents to
be soluble in the produced fluids (water or oil). Specialist drilling fluids are now
available from the service companies that meet these requirements.
5.2.2 External Gravel Packs
All the options described above just used only the screen or liner as the basis of the
completion. It was inserted into the open hole and the gap between the screen or liner
and the borehole wall remain empty. Once the well was placed on production the mud
cake should be produced through the screen as described above. The behaviour of the
formation will depend on its strength:
(i)
strong formations: borehole wall remains intact and gap between the liner and
the screen remains empty
(ii) weak formations: borehole collapses and the original liner/borehole gap
becomes filled with failed formation material.
An alternative is the under reamed, external gravel pack introduced earlier (bottom
Figure 20 and Figure 30). This involves enlarging (by typically 10 -15 cm) the gap
between the sandface and the screen using an under reamer. The under reamer
replaces the drill bit at the bottom of the drill string and consists of a series of arms with
cutters at the end which expanded so as to enlarge the hole. When the drill string is
rotated the hole enlargement is carried out with a non damaging fluid i.e. the filter cake
will decay or not prevent the flow of oil or gas from the enlarged borehole diameter . The
enlarged liner borehole gap is completely filled with gravel, a process known as gravel
packing. This is described in the remainder of this module.
Department of Petroleum Engineering, Heriot-Watt University
33
1
production
casing
production
tubing
cement
packer
casing shoe
gravel
Formation sand
Formation sand
slotted liner or
wire wrap screen
retains the gravel
under reamed
hole section
Figure 30
Detail of an external
gravelpack
5.2.3 Internal Gravel Pack
The screen or liner is placed inside a cased hole for an internal gravel pack with gravel
being placed in the screen / casing annulus and in the perforations
t
Pe Lar
rfo ge
W rati Dia
ith on m
Gr pa ete
av ck r
el ed
Fo
rm
at
io
n
sa
nd
m
en
|||
zzz
,,
yy
{{
zzz
|||
,,
yy
{{
zzz
|||
,,
yy
{{
Ce
g
sin
Ca
Gr
a
an velnu fill
lu ed
s
Sc
re
lin en
er or
An internal gravel pack is illustrated in middle of Figure 20 (right hand side) and
enlarged as Figure 31
,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
,,
,,
,,
,,
,,
,,
,,
,,
,,,,
,,,,
,,,
,,
,,
,,
,,,
,,
,,
,,,
,,,
,,
,,,,
,,,,
,,,
,,
,,
,,
,,,
,,
,,
,,,
,,,
,,
,,,
,,
,,,,
,,,
,,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,
,,
,,
,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,
,,
,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,
,,
,,
,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,,,,
,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,
,,
,,,
,,
,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,
,,,
,,
,,,
,,,
,,
,,,,
,,,
,,,
,,,
,,
,,
,,
,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,
,,,,
,,,
,,,
,,,
,,
,,,
,,
,,
,,
,,
,,,
,,,
,,,
,,
,,,,
,,,
,,
,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,
,,,,
,,,
,,,
,,
,,
,,
,,
,,
,,,
,,
,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,
,,
,,
,,,
,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,,
,,,
,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,
,,
,,,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,
,,,
,,
,,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,,
,,,
,,
,,,
,,,
,,
,,,,
,,,
,,
,,
,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,,
,,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,
,,,,
,,,
,,
,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,,
,,,
,,
,,
,,,
,,
,,
,,,,
,,
,,
,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,
,,,
,,
,,,
,,,
,,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,
,,,,,
,,,
,,
,,,
,,,,,
,,,
,,,,
,,,
,,,
,,,
,,,,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,
,,,
,,
,,,
,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,
,,,,
,,
,,,
,,,
,,
,,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,
,,,,
,,
,,,
,,,
,,,
,,,,
,,
,,,
,,,,
,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,,
,,
,,,,
,,
,,,,,,
,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,
,,,
,,,
,,,
,,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,,,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
,,,
34
Figure 31
Detail of internal gravel
pack
72
Unstable Formations and Sand Control
5.2.4 Special Gravel Packs
The combination of an internal gravel pack and a short, highly conductive, hydraulic
fracture stimulation has become very popular in the Gulf of Mexico (USA). This is
known as “Frac and Pack” -an example is illustrated in the middle of Figure 20 (Left
Hand Side).
A second specialist option is the use of “Resin Coated Gravel”. This product is similar
to that used for making the pre-packed screens - The resin coated gravel sets into a
consolidated mass under the influence of the Well’s Bottom Hole Temperature.
5.3 Advantages / Disadvantages of Gravel Packing
These have been summarised as table 7 which is self explanatory when read in
conjunction with the remainder of chapter 5.
Advantages
Comment
Effective over long intervals
Copes with varying rock properties
>100 m operationally possible
Clay / Silt particles can pass through
gravel pack if large enough
Gravel placed with water based fluids
Employs simple, non-toxic materials
Disadvantages
Mechanical restriction in wellbore
Workover difficult
Expensive for multiple intervals
Diminished workover options
Table 7
Gravel Packing
Considerations
Sensitive to poor completion practices
e.g. Production logging not possible
Requires removal (fishing) of packers /
screens etc.
Consider co-mingled production
Difficult to identify source as well
as shut off undesirable water and gas
Many points where permeability damage
can be created
5.4 Gravel Pack Sand Selection
Gravel pack sand is a well rounded, clean (minimum fines and acid soluble material)
that has been sieved so that its size falls between carefully selected size ranges. If sized
correctly, it acts as a high permeability (i.e. not restricting flow of oil or gas) filter
where the pore throats between the gravel grains are small enough to restrict the
passage of the formation sand grains. This is illustrated in figure 32.
Department of Petroleum Engineering, Heriot-Watt University
35
1
gravel
formation
sand
Too Large Gravel
Too Small Gravel
• Formation sand produced
• Gravel pack failed
• Formation sand retained, but
• low permeability of gravel
creates unnecessary extra
loss of well productivity
Figure 32
Effect of incorrect gravel
pack sand selection
The formation sand is characterised by breaking down any consolidated, multi-grain
particles into the individual sand grains. These are then passed through a series of preweighed sieves of progressively smaller mesh size and the weight of sand trapped on
each sieve recorded. The cumulative distribution is then plotted in the manner shown
in figure 33 using log-linear graph paper. The size corresponding to a cumulative
weight percentage of 10%, 40% and 90% is known as D10, D40, and D90 respectively.
100
D90
90
Cumulative Weight Percentage
80
70
60
D50
50
D40
40
30
20
D10
10
0
1.0
0.5
0.1
0.05
0.01
0.05
Grain Diameter (inches)
Coarse
Fine
The sand grain size distribution is characterised by the Uniformity Co-efficient (C)
which is defined as:
C = D40 / D90
36
Figure 33
Sieve analysis of typical
formation sand
72
Unstable Formations and Sand Control
with formation sands being classified as:
C<3
3<C<5
5<C<10
C>10
well sorted, highly uniform sand
uniform sand
moderate/poorly sorted sand
poorly sorted highly non-uniform sand
Examples of a well sorted (C≈2) and poorly sorted sand (C≈8) are compared in figure 34.
100
90
80
Weight (Percent)
70
1
60
50
40
2
30
20
10
0
0.01
Sand Grain Diameter (cm)
100
1
90
Cumulative Weight (Percent)
80
2
70
Sample 1 is well
sorted (C=2)
60
50
Sample 2 is poorly
sorted (C=8)
40
30
20
10
0
Figure 34
Example of well and poorly
sorted sand distribution
0.1
0.01
Sand Grain Diameter (cm)
Department of Petroleum Engineering, Heriot-Watt University
37
1
The gravel pack sand may be selected once the formation sand has been characterised.
There are a number of criteria used - the most common being the Saucier criteria.
D50 {Gravel} = 6 * D50 {Formation Sand}
However this relationship makes no allowance for the sand uniformity. For poorly
sorted sand, a second criteria attributed to Schwartz should also be examined:
D10 {Gravel} = 6 * D10 {Formation Sand} for C<5
D40 {Gravel} = 6 * D40 {Formation Sand} for C<5
D70 {Gravel} = 6 * D70 {Formation Sand} for C<10
However only a limited range of gravel packsand sizes are available commercially (table 8)
N.B. Tight (cubic) packing of the gravel is required for the Schwartz or Saucier
selection criteria to be effective (figure 35).
gravel
d
d
D
D
Figure 35
Gravel selection criteria
only works if a tight pack
can be created
formation
sand
d = 15% D
d = 42% D
Cubic packing required
Rectangular packing
undesirable
In all cases the wire wrapped screen is selected so that the gap is equal to 0.5 times the
smallest gravel size. This criteria is often not practical for slotted liners - in which case
the smallest available possible slot size should be chosen.
100
90
range for
producing
sand formation
Cumulative Weight Percentage
80
70
fine grained
laminae
(low permeability;
non-productive)
60
50
d50{gravel} = 6*d50{formation}
40
range of total completion interval
30
20
10
0
Grain size
Coarse
38
Fine
Figure 36
Typical gravel / formation
sand size disrtibution
72
Unstable Formations and Sand Control
Frequently, several sieve analyses are available from formation sand samples taken
from different depths in a particular completion interval. Engineering judgement
needs to be exercised - a typical scenario that may be encountered is sketched in
figure 36. Possible changes in the permeability of the gravel / sand combination once
flow has commenced are shown in figure 37. Smaller gravel forms an effective filter
for the sand - but has a low permeability due to its small grain size. Invasion of the
gravel by the formation sand begins to occur once the D50 size is greater than 7. The
permeability of the sand / gravel mixture can actually become less than that of the
formation sand if large scale mixing takes place.
1.0
Formation
well productivity
reduced by using
too small gravel
sand produced
completely
through gravel
0.6
Formation Sand invades gravel;
50:50 mixture
has lower permeability
than formation sand alone
IDEAL
Permeability (final)
Permeability (initial)
0.8
0.4
0.2
Figure 37
Schematic diagram of sand
/ gravel interaction
0
Formation sand
permeability
0
2
4
6
8
10
12
14
16
18
20
D 50 {gravel}
D 50 {sand}
gravel acts as
effective filter
As the gravel increases in size (D50 > 17), the gravel pore throats become sufficiently
large that the formation sand can pass through with minimal impact on the gravel
packs permeability.
The properties of the standard gravel pack sands are listed in table 8
Properties of Standard Gravel Pack Sands
US Mesh
Table 8
Properties of standard
Gravel Pack sands
40 / 60
20 / 40
12 / 20
Range
Gravel Size
(µm)
Median Gravel
Diameter
(µm)
Typical Permeability
425 - 250
850 - 425
1700 - 850
340
640
1275
55
170
600
(D)
N.B. 16/30, 30/50 and 50/70 gravel can be made available to special order (higher cost)
5.4.1 Operational Considerations
The choice of gravel size influences the points in the operation at which the
permeability of the pack can be damaged e.g. smaller gravel is more prone to
Department of Petroleum Engineering, Heriot-Watt University
39
1
plugging by dirty completion fluids but is less prone to plugging by fines during
production.
The specification for the gravel used for gravel packing operations is laid down in a
"Recommended Practice" by a working committee of the American Petroleum
Institute (API). It must not only be carefully sieved so that 98% out of the sample falls
between the maximum and minimum specified sieves, but also the source of the gravel
is selected to meet minimum roundness, sphericity, grain strength criteria along with
a maximum acid solubility level and percentage clay. Synthetic “gravel” is also
available at a premium price - the grains are stronger (less permeability impairing fines
produced during pumping due to grain breakage) and are more spherical (higher
permeabilitys - typically 25% greater for the same nominal grain size).
Pressure drop across perforation (psi/perf)
800
Perforation filled
with 0.5 Darcy
formation sand
Cross sectional area
of perforations
3"
8
700
1"
2
600
500
2
Diameter, in.
3
8
1
2
3
4
7
8
Area, in.
1
0.785
0.110
0.196
0.442
0.301
Perforation filled
with 180 Darcy gravel
400
300
Perforation
3"
4
200
2 in long
fluid data
µ = 1 cp
sg = 0.8
4
β = 2.3 x 10
100
0
0
25
50
75
100
Flows Rate (bbl/d/perf)
Gravel permeability is a key issue for internal gravel packs due to the high pressure
drops potentially associated with flow through the gravel filled perforation penetrating
the casing and cement. This is illustrated in figure 38 where the pressure drop
calculated by Darcy’s law is plotted as a function of flow rate for a cp oil flowing
through different size perforations filled with 180 Darcy gravel. Minimal flow rates
are achieved through the perforation if formation sand replaces the gravel.
The completion engineer is in charge of specifying the perforation density (number
of perforations per metre) and the perforation diameter (design and weight of
explosive for the perforating charges). The perforation programme should be chosen
so that the gravel filled perforations do not limit the oil/gas inflow into the well.
5.5 Gravel Packing - Surface Operations
The surface operations occuring during a gravel packing treatment are summarised in
Figure 39
40
Figure 38
Importance of perforationtunnel size and filling
material
72
Unstable Formations and Sand Control
A 1% wt potassium chloride (KCL) brine base fluid is used “as is” or is viscosified
by the addition of a polymer e.g. hydroxyethyl cellulose. This base fluid is filtered to
remove impairing contaminants and gravel added to the required concentration (low
concentrations of gravel are added continuously (“on the fly”) while pumping; high
concentrations can be pre-mixed in the stirred vessel when viscous fluids are used
since they have sufficient viscosity to retain the gravel in suspension). The slurry is
then pumped into the well using the high pressure pump.
Filter unit
Polymer
10µ
2µ
Paddle
mixer
Pump
Viscous or
non-viscous fluid
base fluid (1% wt KCl brine)
Gravel
Figure 39
Surface equipment for
batch gravel packing
operation
Filtered fluid
High
pressure
pump
Pump to well
5.6 Fluids for Gravel Packing
The properties of the gravel packing fluid can range from being similar to water to
highly viscous. In both cases the fluid performs functions described in table 9.
Table 9
Gravel Packing Fluid
Functions
Function
Comment
Control formation pressures
Similar to drilling fluids
Transport gravel to completion
interval
BOTH at surface and downhole
Return (reverse) excess gravel
from completion interval
to surface
As above
“Clean” fluid essential
Filter solid particles in water and (undissolved)
polymer residue
Compatible with clay particles
in formation
Add 1% wt KCl
Viscosity of viscous fluids must
Add chemical breaker (acid or oxidising agent)
degrade to allow production to surface
Department of Petroleum Engineering, Heriot-Watt University
41
1
5.6.1 Properties of Viscous Gravel Packing Fluids
The viscosity behaviour of a typical viscous gravel packing fluid {a dilute (80 lbs
polymer / 1000 gal water) solution of hydroxy ethyl cellulose polymer in brine} is
shown in Figure 40 where the viscosity is plotted as a function of shear rate. Two types
of behaviour are shown:
100
Viscosity [Pa.s]
10
Newtonian
behaviour at
low shear rate
1
"Po
wer
" la
0.1
wr
egi
me
Shear rate [s-1]
0.01
(i)
10
100
1000
10000
Figure 40
Hydro Ethyl Cellulose
(HEC) polymer viscosity as
a function of shear rate
a settling particle
casing
tubing
annulus
{
{
{
1
{
shear rate
regime for:
0.1
{
0.01
drill pipe
formation
A low shear rate “Newtonian” region where viscosity is independent of shear rate
i.e. µ = τ / δ
(2)
where µ = coefficient of viscosity, τ = shear stress and δ = shear rate
(ii) A higher shear rate region where “power law” behaviour is shown i.e.
kδ η
= Kδ η-1
δ
where η = power index and k = consistency index
∴ µ =
(3)
Newtonian fluids have a power index of 1, hence equations (2) and (3) become
equivalent.
The slope of the log viscosity against log shear rate plot shown in figure 40 is the power
index η. The typical shear rate region which is encountered in various parts of the well
are also indicated in the figure.
42
72
Unstable Formations and Sand Control
5.6.2 Other Base Brines
1% wt KCl solution is the standard brine since it is compatible with the majority of
formations {i.e. no permeability damage}. Geopressured formations require the use
of higher density brines - see table 10
Completion Brines for Geopressured Formations
Chemical
Formula
Maximum
specific
gravity
(g.cm3)
Ammonium Chloride
Table 10
Completion Brines for
Geopressured Formations
Potasium Chloride
Sodium Chloride
Potasium Bromide
Sodium Formate
Calcium Chloride
NH4Cl
KCl
NaCl
KBr
HC00Na
CaCl2
Sodium Bromide
Potasium Formate
Mixed Calcium Bromide/ Calcium Chloride
NaBr
HC00K
CaBr/CaCl2
Calcium Bromide
Mixed Zinc Bromide/ Calcium Bromide
CaBr
ZnBr2/CaBr
1.13
1.17
1.2
1.31
1.33
1.42
1.54
1.6
1.7
1.87
2.4
The brine chosen will depend on the density required, cost, compatibility with the
formation and well equipment, crystallisation temperature (for North Sea or other
lower temperature operations) as well as health and safety issues for the wellsite staff.
N.B. A range of formate salts is being introduced to cover the same density range.
Formate salts have an improved environmental profile compared to the higher
density, bromide based brines.
5.7 Fluid Loss Control
Great care is taken to prepare the well in an unimpaired state prior to the start of the
gravel pack operation. This leads to the result that even a small overbalance of the
completion fluid’s hydrostatic head compared to the reservoir pressure leads to large
scale losses of completion fluid into the formation. This is undesirable since it often
leads to extensive formation permeability impairment. Conventional lost circulation
fluids, as employed during drilling, also lead to formation impairment. Four options
are available:
(i) placing a small volume of a very high viscosity fluid across the perforated zone.
This either contains a viscosity breaker (which will degrade the viscosity after a
specified time/temperature exposure period) or it is removed by pumping acid
immediately before (preferred) or after the gravel pack has been completed.
(ii) as above with filter cake forming solids (oil soluble resin or water soluble
sodium chloride crystals) which will dissolve once production commences.
(iii) mechanical solutions such as large bore flapper valves (figure 41).
Department of Petroleum Engineering, Heriot-Watt University
43
1
The valve is run into the well in the open position - being held open by, for example,
the wash pipe of the gravel pack assembly (see section 5.8). On withdrawal of the
washpipe, the spring holds the flapper valve closed against the seal; thus preventing
further brine losses from the tubing to the formation. The valve is constructed from
glass or ceramic material so that it can be broken by pressure or mechanical impact.
Top View
Spring
Flapper
Connection
Spring
Seal
Figure 41
Large - bore flapper valve
Tubing
Side View
(iv) Formation Inflow Valve (FIV) figure 42. This is a large bore adaptation of a drill
stem test valve - it can be open and shut by either annulus or tubing pressure. The FIV
is placed near the bottom of the tubing string. A seal assembly is placed at the bottom
of the tubing string. The FIV is large enough for perforating guns mounted at the end
of a work string to be run through it. The sequence of operations is as follows:
packer
(not yet set)
formation
inflow
valve
seals on
tubing
polished bore
receptical at
top of liner
cement
cement
cemented line
formation perforated
through tubing
(a) Open FIV by pressurising annulus
(b) Perforate completion interval with tubing retrievable perforating guns.
(c) Recover perforating guns to above FIV.
(d) Close FIV by pressurising annulus to stop completion fluid losses.
44
Figure 42
Horizontal well completion
schematic
72
Unstable Formations and Sand Control
5.8 The Gravel Pack Operation
Figure 43 schematically illustrates the main stages of a gravel pack operation. This
figure illustrates the process for an EGP; the process being the same for an IGP. The
wire wrapped screen or liner is placed across the completion interval with a length of
normal tubing above it. A specialised gravel pack tool, called the “cross-over tool”,
and a packer are mounted above the tubing. The “cross-over tool” allows various
circulation paths from the tubing to the annulus to be selected. The operation is as
follows:
,
y
,
y
,
y
gravel slurry
"cross-over"
tool
"cross-over"
tool
production
casing
screen
wash
pipe
Figure 43
The gravel pack operation
using a crossover tool
gravel
pack
(a) Circulating gravel
gravel
pack
(b) Screen Out
(c) Reverse Out Excess Gravel
(i) Gravel slurry is pumped down the tubing and “crosses-over” into the liner /
casing annulus. The gravel falls to the bottom of the hole where it builds upwards.
The fluid flows through the liner, up the wash pipe and “crosses-over” so that it can
return to the surface via the casing / tubing annulus (figure 43a).
(ii) The pressure rises rapidly once the (de-hydrated) gravel level has covered the
top of the liner since flow through the gravel leading to a much higher pressure drop.
This is called a screen out (figure 43b).
(iii) The gravel pack tool is raised to allow circulation directly from the tubing to the
tubing / casing annulus (the flow paths to the gravel pack itself are disconnected).
Reverse circulation (DOWN casing/tubing annulus and UP the tubing) allows the
excess gravel slurry to be recovered at the surface. (figure 43c).
Department of Petroleum Engineering, Heriot-Watt University
45
1
The “cross-over” tool, packer and work string are recovered so that the final well
completion tubing can be run.
5.9 Gravel Placement with Low Viscosity Fluids
The manner in which the gravel pack is formed depends on the viscosity of the fluid
and the deviation angle that the well is drilled through the completion interval. The
key points concerning low viscosity fluids are summarised as follows:
(i) Gravel placement with low viscosity fluids is mainly applied to external gravel
packs or shallow formations - The low gravel concentrations (2lb gravel / gal fluid)
and the low frictional pressure losses minimise the chance of fracturing these weak
formations.
(ii) the time required to pump all the gravel and complete the gravel pack is long
due to this low gravel concentration, despite the high pump rate of 5-10 bbl/min.
(iii) This high pump rate can cause permeability impairment due to gravel/sand
intermixing (see section 5.4).
(iv) the large volumes of fluid pumped during the operation through the screen/liner
increases the opportunity for liner or screen plugging.
Large scale laboratory tests have shown that the gravel pack is formed from the bottom
upwards in vertical or low deviation (<45º) wells - see figure 44.
yy
,,
yy
,,
yy
,,
yy
,,
yy
,,
yy
,,
7
Crossover
Packer
Upper
Tell Tale
6
Wire-wrapped
Screen
Packing Sequence
5
4
3
Wash Pipe
2
1
The process that occurs in perforations - dune formation - is more complicated.
(Figure 45.) The gravel is deposited at the mouth of the perforation since it cannot be
held in suspension by the slow moving fluid (leak off rate per perforation is low). The
dune builds up until the gap between it and the top of perforation is so narrow that the
fluid velocity has increased sufficiently to carry the gravel further down the perforation.
46
Figure 44
The gravel packing process
with low viscosity fluids
72
Unstable Formations and Sand Control
The dune can then develop down the perforation until the perforation tip is reached
as shown. The final packing stage is then backwards from the tip to the mouth of the
perforation.
Cement
11
Flow
Figure 45
The perforation packing
process with low viscosity
fluids
1
2
3
10
4
5
6
7
8
9
Casing
5.10 Gravel Placement with High Viscosity Fluids
Internal Gravel Packs are usually placed with high viscosity fluids. The high viscosity
(up to 300 cp) carrier fluids allows the gravel to be suspended at high gravel
concentrations (up to 15lb gravel/gal carrier fluid or 50% vol). Also the resulting
slurry can be pumped at low rates (0.5 - 1.5 bbl/min). This gives considerable
advantages which should lead to higher productivity completions:
(i)
limited fluid volume
(ii) low pump rate
→
reduced screen plugging
→
no gravel/sand mixing
The gravel packing process is quite different with high viscosity fluids:
(i) gravel slurry dehydration is initiated at the perforation tip followed by formation
of a node once the perforation is filled (figure 46). The process is driven by fluid leak
off into the formation.
Cement
Formation
Flow
7
6
5
4
3
2
1
Gravel Node Forms
at Perforation Mouth
Figure 46
The perforation packing
process with viscous fluids
Casing
(ii) simultaneously, a sheath of dehydrated gravel is formed around the screen - the
process being driven by fluid flowing circulating through the screen and returning
to the surface via the washpipe (figure 47).
Department of Petroleum Engineering, Heriot-Watt University
47
1
Crossover
Packer
8
9
6
7
Wire-wrapped
Screen
Gravel Nodes at
Perforation Mouth
Wash Pipe
4
5
Gravel Sheath
Around Screen
3
2
1
(iii) The remainder of the gravel pack is then completed from the bottom upwards
(figure 47). Once the gravel pack is completed as far as the top of the wire wrapped
screen, the pressure increases rapidly a screen out having been achieved. Production
may be established once the gravel pack operation is completed.
(iv) A chemical breaker has to be mixed with the viscous fluid to degrade the
viscosity so that production may be established once the gravel pack operation is
complete.
(v) more difficulty is experienced in obtaining complete gravel packs when:
(a)
the completion interval contains large permeability contrasts
(b)
the completion zone is long (>15m)
(c)
the deviation angles are high (>50º)
Use of special equipment changes - e.g. of a large diameter wash pipe - which help
distribute the slurry across the complete completion interval can improve the gravel
pack quality.
5.11 New Technology
5.11.1 Gravel Pack Evaluation
Several of the (nuclear) density logs can be used for gravel pack evaluation. Intervals
with an incomplete gravel pack give a lower shallow density reading than intervals
with a good pack. However, the readings are also influenced by changes in the
completion equipment e.g. connections, packers, transition from wire wrapped screen
to tubing etc. These effects are schematically illustrated in figure 48.
48
Figure 47
The gravel packing process
with viscous fluids
72
Unstable Formations and Sand Control
y
,
y
,
y
,
y
,
y
,
y
,
y
,
Density Log Reading
Top of Gravel
5000
Gravel
Casing
Collar
Log
Figure 48
Gravel pack evaluation with
a nuclear density log
100% Complete Gravel
Pack
Screen
Perforations
Inomplete Gravel
Pack
100% Complete Gravel
Pack
5.11.2 "Frac and Pack"
"Frac and Pack" was introduced in section 5.2.4. Field experience, particularly in the
U.S. Gulf Coast, has shown that this combination of hydraulic fracturing and gravel
packing leads to completions with a lower “skin” and hence much higher well
productivities (figure 49). The hydraulic fracture pierces the ring of formation
permeability impairment around the well by providing a high conductivity channel for
the flow of oil or gas figure 50. The (vertical) hydraulic fracture also ensures that the
well is connected to all thin, laminated sands, some of which might be missed by a
conventional gravel pack.
Gulf Coast, USA, Frac - Pack Field Results Based on Well Build Up Tests
65
Skin Value
55
45
35
25
15
5
-5
0
10
20
(36 Frac - Pack, 30 Gravel Pack)
•
30
40
50
60
Wells
Gravel Pack Skins Range - 1-300+
Average Gravel Pack Production Efficiency < 25%
•
Figure 49
"Frac Pack" results
Frac - Pack Skins Range - 4 to 27
Average Frac - Pack Production Efficiency < 95%
Department of Petroleum Engineering, Heriot-Watt University
Gravel Packs
Frac - Packs
49
1
Damaged
Zone
Screen
Damaged
Zone
10mD
Propped Fracture
Screen
100D
Gravel Pack
0.1mD
"Frac and Pack" fractures are short and fat, compared to the long thin fractures used
for stimulating low permeability formations. This is because formations requiring
sand control normally have a high permeability. The background to these statements
will be explained in the hydraulic fracturing module.
5.11.3 New Screen Technology
It was discussed earlier that high flow rates have often been observed to trigger sand
inflow problems (figure 2). This has lead to a large effort has been made to develop
screens with an increased area open to inflow from the formation i.e. screens that will
minimise the inflow velocity. Figure 51 compares the inflow area for a perforated
casing with that of various types of sand control screens and the ideal completion (an
open hole).
50
Figure 50
Conceptual view of frac
packing
72
Inflow Area as % of Openhole Flow Area
Unstable Formations and Sand Control
Figure 51
Inflow area as a function of
completion type
100
10
1
0.1
6 SPF
12 SPF
24 SPF
Casing with 0.75in. diameter
perforations
(SPF = Shots per foot)
Wrapped
Screen
Slotted
Liner
Openhole
Uniform
Membrane
(High Cost)
The inflow area of the wirewrapped screen is reduced to 3% once a gravel pack is
placed in front of it. High cost membranes with a size small enough to control sand
particle movement without need for a gravel pack have been developed with a very
high inflow area. The membrane is run in the hole during the completion process and
assumes that the formation sand will collapse around it.
An alternative approach being developed is the use of an expandable sand screen. Here
the sand screens diameter is expanded by 33% to 50% by pumping, pushing, pulling
or rotating an expansion tool through the screen once it has been placed across the
completion interval (figure 52).
Expansion Cone
Formation requiring support
by sand screen
Unexpanded Pipe
Expanded Pipe
Figure 52
The expandable screen
Department of Petroleum Engineering, Heriot-Watt University
51
1
5.12 Chemical Sand Consolidation
Chemical sand consolidation - the artificial strengthening of the oil/gas producing
formation - is the least frequently used of the sand control. This is despite its two
major advantages:
(i) unlike mechanical forms of sand exclusion, it leaves the wellbore completely
unobstructed without imposing any restrictions as far as future work-overs are
concerned (compared to a standard perforated completion)
(ii) the consolidation treatment can also be carried out through tubing i.e. a drilling
or work-over rig does not have to be moved onto the well and the tubing etc. does
not have to be pulled as the first stage of the well repair. Instead a pump truck and
a series of chemical storage tanks with suitable manifolding has to be provided (see
figure 54). This can often be mobilised more quickly and at a lower cost than a rig.
The chemical glue must:
(i)
wet and adhere to the sand grains
(ii) leave the pore spaces open so that the permeability is retained.
This is illustrated in figure 53 which shows an enlarged view of a number of sand
grains which have been covered with chemical cement. Organic resins (epoxy, furan
or phenolic) are the most widely used chemical cements -though other materials have
been used in the past -solder, water glass, alumina, nickel plating. The typical target
formation for chemical consolidation are thin ( < 3m), high permeability, clean (low
clay content) sands. This restriction on the formation type to be treated and the
operational complexity explains the lack of popularity of chemical consolidation as:
Layer of chemical cement
Sand grain
Sand grain
Open
pore
space
Sand grain
Cement concentrated
at intergrain
contact point
Sand grain
(i) the consolidation treatment is operationally complex (figure 54). This is due
to the need to prepare the sand grain surface so that the chemical cement will adhere
to it followed by the need to re-establish the formation permeability e.g. by
overflushing excess chemical cement away from the near wellbore area and
displaceing it deeper into the formation. The complex series of fluids to be pumped
52
Figure 53
An enlarged view of a
chemically consolidated
formation
72
Unstable Formations and Sand Control
could include some, or all, of: acid, neutraliser, preflush, spacer, resin, overflush,
displacement fluid. Each fluid requires a separate tank which have to be manifolded
together so they can be pumped in the correct sequence in one smooth, continuous
operation:
Closed Tank
For Returns
Wellhead
High
Pressure
Pump
To Mud Pits
(truck)
Booster/Charge Pump
Figure 54
Site layout for chemical
consolidation N.B. May
also require acid and
neutraliser tanks if preconsolidation matrix
stimulation treatment
required
Filtration
Unit
Mixer
Pre-flush
Spacer
Brine
A
B
C
D
Storage
Tanks
Diesel/
Kerosene
Small Tanks With
Chemical Cement
Properties
(ii) many of the chemicals used have aggressive properties and due care and
attention has to be given to operator safety and potential environmental impact.
(iii) the formation composition can be detrimental to the strength development of
the chemical cement e.g. A chemical formulation capable of developing a strength
of 150 bar unconfined compressive strength in a formation containing clean sand
grains, will typically show a negligible strength development when the clay content
rises to 10% wt (see figure 55). The need to treat clean sand indicates that high
permeability formations are the main target for chemical consolidation.
Department of Petroleum Engineering, Heriot-Watt University
53
1
Unconfined compressive strength
of consolidated formation (bar)
150
100
0
0%
5%
Clay content 10%
100
95%
Sand content
90
(iv) All the chemical flushes have to be applied to all the perforations in the correct
order since one or more incorrectly treated perforations could lead to future sand
production. Field experience has shown that this requirement for correct placement
means that only thin zones (maximum length 3m) should be treated at one time.
Longer zones require a series of consecutive treatments.
54
Figure 55
Typical strength
development of a
consolidation system with
increasing clay content
72
Unstable Formations and Sand Control
SAND CONTROL TUTORIAL
“Installing sand control equipment is sometimes a necessary evil”
Question 1.
Discuss this statement with respect to
• Well productivity
• Easy of well operation
• Easy of future workovers
Rank the various types of control for their impact on well productivity
Answer 1.
Well productivity
High well productivity losses (typically 60%) are observed after well killing
followed by sand control equipment implementation such as gravel packing. In
some cases the final well productivity was only around 20% of that achieved prior
to gravel packing. Such high impairment has been noted in many gravel packed
wells.
Easy of well operation
Installation of gravel packing creates mechanical restriction in wellbore e.g.
Production logging is normally no longer possible.
Easy of future workovers
Installation of gravel packing also makes future workover operations difficult.
Requires removal (fishing) of packers / screens, etc. Workover options are
diminished, making difficult to identify the source as well as making it more
difficult to shut-off undesirable water and gas.
The various types of sand control can be ranked by their impact on the above
parameters as follows:
Sand control type
Well
production
2
Easy of well
operation
1
(after sand failure)
(after sand failure)
1
5
4
3
1
4
4
2
Perforated completion
(living with sand)
(before sand failure) (before sand failure)
5
5
Screen/Slotted lines
Internal gravel pack
External gravel pack
Consolidation
Easy of future
workovers
1
3
4
5
1
Question 2.
Pressure drops as a function of flow rate has been measured through a number of
simulated perforations filled with gravel and gravel/sand mixtures. The results are
shown in graphs a to d. These graphs illustrate many of the aspects of flow through
Department of Petroleum Engineering, Heriot-Watt University
55
1
porous media. Identify those aspects which can be derived from the graphs and
comment on how these would influence your design of an internal gravel pack.
Answer 2.
Pressure loss per inch of perforation
tunnel length psi
Graph a
2000
Legend:
3/8" perforation
0.017" - 0.033" gravel
ø = gravel porosity
1800
1600
1400
Darcy law
computation
K - 1 Darcy
1200
1000
ø = 35%
ø = 40%
L
TA
EN
M
RI
800
PE
EX
600
400
Darcy law computation
200
0
0
25
50
75
100 125 150
Flow rate per perforation (bbls/day)
175
a) Turbulent flow through gravel packed perforations greatly increases the pressure
drop with increasing flow rate compared to the pressure drop value calculated from
Darcy’s Law (which is independent of flow rate).
b) Increased pack porosity gives increased permeability (reduced pressure drops)
Graph b
D
Pressure drop (psi)
200
Perforation
Diameter
150
C
100
D 3/8 inch
C 1/2 inch
B 3/4 inch
A 1 inch
50
B
A
0
0
5
10
15
20
25
Flow rate (bbl/d per perforation)
a) Pressure drop across the gravel filled perforation decreases as the perforation
diameter increases
b) Larger perforations with lower flow velocities show reduced turbulent effect
56
72
Unstable Formations and Sand Control
Graph c
200
Pressure drop (psi)
A
1 Darcy formation sand
150
A
B
C
D
100
B
50
1 Darcy
Formation sand
40 -60 gravel
20 - 40 gravel
12 - 20 gravel
C
D
0
0
5
10
15
20
25
Flow rate (bbl/d per perforation)
a) Even the smallest (40/60) gravel pack sand shows a greatly reduced pressure drop
compared to a good formation sand (1 Darcy) when packed in the formation tunnels.
b) “Extra” pressure drops due to turbulence (or non-Darcy flow) decreases as the
gravel size increases.
Graph d
Permeability (Darclea)
70
60
50
40
30
20
10
0
0
20
40
60
80
100
% sand mixture
a) Permeability of gravel/formation sand mixtures decreases as the concentration of
formation sand increases.
b) Some mixtures show a permeability even lower than the formation sand alone.
Department of Petroleum Engineering, Heriot-Watt University
57
1
6. FURTHER READING
(1) Golan M. & Whitson C.
“Well Performance” 2nd edition
published by Norwegian University of Science and Technology
(2) Allen T. & Roberts A.
Volume 2 “Production Operations” 4th edition
published by OGCI
(3) Economides M., Hill A. & Economides C.
“Petroleum production Systems”
published by Prentice Hall
(4) PENBERTHY, W.L. & SHAUGHNESSY, C.M.
Sand Control. SPE Series on Special Topics. Volume 1.
Richardson: SPE, 1992
ISBN 1555630413
58
72
Unstable Formations and Sand Control
Department of Petroleum Engineering, Heriot-Watt University
59
1
8
Oil and Gas Processing
Well Control
CONTENTS
1.
2.
INTRODUCTION
FIELD DEVELOPMENT
2.1.
Oil Fields
2.2.
Gas Fields
3.
FIELD DEVELOPMENT EXAMPLES
4.
PRODUCTION PLATFORM FUNCTIONS
5.
PRODUCTION PHILOSOPHIES
5.1.
Separation Objectives
5.2.
Oilfields
5.3.
Gas Fields
6.
BA SI C PR OC ES S ING SCHEME
6.1.
Processing Conditions
6.2.
Process Plant Metallurgy
6.3.
Separators
6.3.1. Horizontal Separators
6.3.2. Vertical Separators
6.4.
Operational Production Problems with
Gravity Separators
6.4.1. Foaming
6.4.2. Solids
6.4.3. Emulsion
6.4.4. Surging Flow
6.4.5. Production Chemicals
6.4.6. Miscellaneous Processes and Comments
7.
SEPARATOR SIZING BASICS
7.1.
Gas Capacity
7.2.
Liquid Capacity
8. TEST SEPARATOR
9.
COMPRESSORS
10. OIL EXPORT
11. GAS HANDLING
11.1.
Increasing NGL Recovery
11.2.
NGL Stabilisation
11.3.
Gas Dehydration
11.3.1. Dew Point Depression
11.3.2. Hydrates
11.4.
Continuous Dehydration Process
11.5.
Batch Dehydration Process
11.6.
Acid Gas Treating
12. FURTHER READING
1
LEARNING OUTCOMES:
Having worked through this chapter the student will be able to:
• Discuss the impact of the field location, well rate, produced fluid and secondary
processing on the design and operation of the production facilities
• Relate the required Production Facility services to the field’s oil recovery mechanism
• Draw a basic, outline production process scheme
• Describe the components and discuss the operation of a 3 phase separator
• Identify the advantages of horizontal and vertical separators
• Describe the operational problems associated with these separators
• Quantify the sizing (gas and liquid capacity) of a 3 phase separator
• Discuss fiscal measurement of produced cride oil
• Describe a pipeline “pigging” operation
• Describe the components of a gas handling facility viz NGL separation and
stabilisation, gas dehydration and sweetenting.
2
8
Oil and Gas Processing
1. INTRODUCTION
This section covers the treatment of the produced fluid from and during its passage
from the well-head, through the facility to the point of sale. The nature, scope and
geographical spread of these facilities will vary greatly depending on the:
(i)
Location: On or off-shore
(ii) Well and Field Production Rate
(iii) Oil, gas or condensate Field
(iv) Central gathering station or local (well-head) facilities
(v) Secondary processing requirements e.g. removal of contaminants such as H2S
However, in all cases the facility is designed to separate the (multi) well stream into
the three basic components (oil, gas and water) and to process the phases into:
(i)
marketable products (i.e. to sales specification) - or
(ii) a form that they can be disposed of in an environmentally acceptable manner.
A simplified production scheme for an oil or gas field is given as Figure 1.
Gas
Compressor
Manifold
(Multi stage)
Three phase
separators
Lower water and
hydrocarbon
dewpoints.
Remove contaminents
as required
Sales
pipeline
Oil
Storage
Tank
Sales
Water
Choke
Figure 1
Simplified processing oil
facility scheme
Reuse or disposal
Simplified processing oil facility scheme
The primary separation process takes place in gravity separators. The process is
driven by the density difference between the gas, oil and water phases and the high
fluid pressure frequently available at the well head. Gas “flashes” from the well-head
fluid as the pressure is reduced, while at the same time “free water” is separated from
the oil.
Department of Petroleum Engineering, Heriot-Watt University
3
1
The gas must be compressed to the export pipeline pressure and treated to remove
water vapour and heavier hydrocarbons. More complicated processes may be
required to “sweeten” the gas by removing other contaminants such as carbon
dioxide, hydrogen sulphide etc.
Typically, the oil and water will only show partial separation, and an “emulsion”, or
intimate mixture of oil or water droplets in the other phase, will require more intensive
processing to achieve a suitable level of separation. Any produced solid material is
also normally found in this layer.
2. FIELD DEVELOPMENT
One of the fundamental decisions to be made during field development is whether to
develop the discovered field as an oil or gas field. This has a large impact on:
(i)
the sales contract
(ii) the field development philosophy and any measures taken for improving the
reservoir recovery efficiency and
(iii) the production facilities.
The design and operation of the production facilities influences the (relative) recovery
of oil and gas. In virtually all cases, the production facilities will be designed to
maximise the recovery of (hydrocarbon) liquids since their sale is normally more
profitable than gas. However, more complex facilities are usually required to recover
a higher proportion of liquids; resulting in a trade off between increased capital
expenditure, (possibly) reduced operating cost and increased revenue.
2.1 Oil Fields
The production from oil fields build up rapidly as more wells are drilled and brought
onto production. The oil field is then produced at a maximum (plateau) for the next
few years. This rate is determined by the capacity of the production facility. The oil
production then gradually decreases until the income from the oil production no longer
pays for the field operating expenses (Figure 2). The field is then ready for abandonment.
The plateau oil production rate will be determined by the individual well rates, and the
numbers of wells to be drilled; not to mention the geology, reservoir fluid properties
etc. The length of the production plateau and the decline will also be a function of the
reservoir size and the recovery mechanism.
Oil fields will typically produce a much larger volume of water than oil. Hence, oil/
water separation typically occurs near to the production well to minimise unnecessary
expense in the pumping of large volumes of water over large distances. In particular,
offshore separation will be employed in the case of offshore fields; while wellhead
separation (at the sea floor, and disposal into an injection well) is now being developed
for subsea wells. Offshore operations are thus becoming more similar to land
operations, where piping and pipeline costs are minimised by:
4
8
Oil and Gas Processing
(i)
local separation and disposal of the bulk of the produced water
(ii) oil (with a relatively low water content) being sent to a central gathering station
for final water separation treatment.
Plateau
ti o
De
uctio
n Buil
n
Prod
First
Oil
uc
d-Up
od
Discovery
Pr
Oil Production Rate
Production
Field Abandonment
c li
ne
Operating Expenses (Oil Equivalent)
Time
Figure 2
Oil field production
Large volumes of sales quality crude oil have to be piped to the coastal export terminal
via a pipeline where it can become part of the world crude oil trade. This is particularly
true for onshore oil fields where the only economic means of transporting significant
volumes is by pipeline, though smaller volumes are sometimes transported by train/
barge or road. Offshore, the choice between a pipeline and local storage near the
platform together with a shuttle tanker to transport the crude oil directly to the refinery
or to the export terminal will be determined by the economics of the two scenarios.
If a dedicated pipeline to the coast can not be justified, then the presence of existing
infrastructure - i.e. a nearby pipeline - with sufficient available capacity to transport
that projected volumes is the key factor. The increasing density of the pipeline
transport network in the North Sea is illustrated in Figure 3.
Gas export is normally only possible when the volumes are sufficient to make building
of a dedicated pipeline economic. Thus, the available pipeline infrastructure is even
more important in the case of produced gas since there are no storage alternatives - it
has to be exported, used or otherwise disposed of at the same time as it is produced.
Many oil developments will use (part of) the gas as a fuel to power the platform and
facilities. However, the changes in the volumes of produced fluids from the wells
(table 3), imply that there may be excess gas early in the life of the project available
for export, while the platform becomes gas-short later on in the project lifetime. The
installation of a gas pipeline allows the field to export gas in the early years while
importing gas to power platform equipment in late project life. This is particularly
attractive if the operators own a nearby field with surplus gas. All these factors have
a large impact on the design of the platform facilities.
Department of Petroleum Engineering, Heriot-Watt University
5
1
Magnus
"A"
Eider
Don
Thistle "A"
Tern
North
Cormorant
South
Cormorant "A"
Ospery
Murchison
Central
Cormorant
UMC
D
N.W.Hutton
Heather "A"
C
Hutton
Stratfjord
B
Ninian Nth
A
BRENT
Ninian Sth
Central
North
Alwyn
Any remaining gas has to be disposed of by flaring, venting (releasing to the
atmosphere without combustion) or underground disposal. Reduction in the emitted
(flared and vented) gas has become a governmental objective with the imposition of
a carbon tax in some countries. Further, company management has encouraged the
development of many innovative facility designs/field development options to
achieve this.
As mentioned earlier, large volumes of water are “co-produced” with the oil.
Frequently large volumes of water are injected into the reservoir to increase the oil
recovery by “sweeping” the oil from the injection to the production well, and by
maintaining the down-hole reservoir pressure.
This requires facilities to:
(i) extract large volumes of water from a surface (sea or river) or under ground
source (dedicated water production wells),
(ii) remove solids (filtration) and (corrosive) oxygen to the specified levels and
(iii) inject the water into a dedicated injection well.
2.2 Gas Fields
The development of a hydrocarbon resource such as a gas field implies that, unlike
oilfields, the produced fluid will be mainly gas accompanied by small volumes of
water and condensate {sometimes called “natural gasoline” or natural gas liquids
6
Figure 3
Northern North Sea
Pipeline Systems
8
Oil and Gas Processing
(NGL)}. Offshore separation is normally employed with the (small volumes of) water
being disposed of to the sea and the gas/condensate being exported to the coast via
multiphase pipelines. Advances in control engineering (automation, data transmission, infra red gas detectors etc.) mean that (small and mature) fields can now be run
on a “not-normally-manned” basis with the consequent reduction in operating
expense and manpower employed.
A new type of field that is being developed in the North Sea is the High Temperature,
High Pressure Gas Condensate field. These fields produce (relatively) much larger
volumes of condensate compared to gas fields. They also exhibit unusually large
changes in wellhead pressures and fluid compositions as the reservoir depletes.
3. FIELD DEVELOPMENT EXAMPLES
The majority of offshore production platforms are supported by steel jackets which
are held in place on the sea bottom by steel piles (see Figure 4 and 5, a large, integrated
steel jacket platform). These welded pipe structures provide a support for various
prefabricated modules, e.g. accommodation, power facilities etc. {see section 5 ‘Platform Functions’ for a more detailed discussion}.
0
50
Meters
Mono Pod Platform
(Usually Unmanned)
100
Jack - Up Production Platform
With Subsea Storage
150
200
Concrete Gravity Platform
Figure 4
Types of offshore platforms
Steel Jacket Platform
Floating Production Platform
With Subsea Collection Manifold
250
Tension Leg Platform
300
Figure 5
A large steel jacket
platform
Department of Petroleum Engineering, Heriot-Watt University
7
1
Concrete gravity platforms depend on their large weight to hold them in place on the
sea bed. The Brent D platform weighs over 200,000 tons, but is also capable of storing
1,000,000 barrels of oil; so as well as avoiding the need for piling, it provides local
crude storage capacity for transfer at regular intervals to the shuttle tankers.
An offshore storage buoy or storage tanker would be required if this integral storage
was not available within the platform construction. Offshore loading is illustrated in
Figure 6.
Figure 6
Offshore loading
Fixed platforms become uneconomic as the water depth increases and floating
platforms are attached - either tethered (anchored) over subsea production wells or
more rigidly held in place by tensioned cables. The most recent developments to
exploit the smaller hydrocarbon reservoirs currently being developed, can be found
in the combination of subsea wells drilled from one or more drilling centres, with a
FPSO (Floating Production Storage and Offtake) vessel. The FPSO hosts all the
normal platform functions such as production separation and water injection equipment. As discussed earlier, gas export requires a pipeline while oil export can be by
pipeline or shuttle tanker. Figure 7 illustrates a typical example of this type of
development. An FPSO is not the only option for this type of development - depending
on the reserves, location etc., a fixed platform can play host to the production facilities
(Figure 8) for the subsea wells
4. PRODUCTION PLATFORM FUNCTIONS
Production platforms perform a multitude of functions as listed below. All functions
may be carried out in one large platform - as in the case of integrated platforms
illustrated in Figure 5. Alternatively they may be split into a series of separate
platforms - each with one or more of the following functions - connected by a bridge
so that personnel and equipment can move easily between them. Figure 9 illustrates
a typical organisation of the following modules.
8
8
Oil and Gas Processing
H
Oil Line
Water Injection Line
Producer
FPSO
Gas export
Injection
Field A
Producers
P3
P2
P1
Field B
Production
Manifold
P4
Figure 7
A subsea FPSO
development
Injector
Field C
Water Injection
Manifold
Field 1
To
Field 4
2 x 6" Gas Pipelines
Diverter
Diverter
Crossover
Field 2
Subsea
Isolation Valve
20" Gas
Pipeline
Field 4
Drill Centre A
16" Oil
Pipeline
Drill Centre B
Drill Centre C
Figure 8
A host platform connected
Field 3
3-5km to Host Platform
Drill Centre D
To Field 1
Existing 20"
Gas Pipeline
Riser Platform
> 100km
to several subsea fields
Department of Petroleum Engineering, Heriot-Watt University
9
1
1.
2.
3.
Drilling derrick with wellheads situated directly underneath
Drilling support module such as drilling mud preparation
Process equipment where oil, gas and water are separated and treated to sales
or end-user specifications
4. Export and (sometimes import of) sales quality crude oil and gas
4a. Compression module where gas is compressed to the required pressure for
export or own use (gas lift, gas injection or power generation). Excess gas is
flared via the flare boom
4b. Oil export module which house the pumps for exporting the crude to pipeline,
local (floating) storage or shuttle tanker
4c. Water Injection module consisting of filtration, deoxygenation, chlorination,
and high pressure injection pumps
5. Platform utilities such as power generation (frequently gas turbines but standby diesel power in the case that the gas supply is unavailable)
6. Accommodation / life support for the personnel manning the platform. For
safety reasons, this is normally situated as far as possible from the process/
compression/wellhead areas.
7. Control room, the hub of the platform’s operation. Safety and loss control
systems are monitored here, as is the fiscal metering that accurately measures
the volumes of crude oil exported.
8. Maintenance workshops, communications, cranes to lift supplies and equip
ment onto the platform from supply vessels, transport (helicopters) etc.
The chosen Reservoir Oil Recovery Mechanism and associated artificial lift used to
increase the energy available to lift fluids from the downhole to the surface will have
a significant impact on the required (platform) services (Table 1).
Helideck
Drilling Derrick &
Substructure
Flareboom
Control
Room
Accomodation
Process
Module
Utility
Module
Power
Generation
Gas
Sea Water
Injection
Sea Level
10
Oil Export
Oil
Gas Lift
Crude Oil,Gas
and Water from
Reservoir
Gas
For
Fuel
Gas Export
Wellhead
Module
Drilling
Module
Gas
Turbines
Gas to Fuel
Cellar
Decks
Figure 9
An integrated steel jacket
platform
8
Oil and Gas Processing
Recovery Mechanism
Primary Recovery
Process
Natural Flow
Artificial Lift
Secondary Recovery
Tertiary Recovery
-
No extra platform services
Gas Lift
Extra Gas Compression
ESP*
Extra Electrical Power Generation
Water Flood
Water Injection Facilities
Pressure
Water injection
Water Injection Facilities
Maintenance
Gas injection
(High Pressure) Gas Compression
Steam,
Carbon Dioxide,
Table 1
Scale and type of platform
services depend on recovery
mechanism.
Impact
Dedicated, Specialised Facilities
Miscible Gas,
Chemical etc.
* ESP - Electric Submersible Pump
5. PRODUCTION PHILOSOPHIES
The process plant capacity is set by the number of production wells and the well
production rates (determined by reservoir permeability, fluid properties, and pressure
thickness of the pay zone, well design, individual well reserves etc.) and the field
production philosophy. These differ between oil and gas fields:
5.1 Separation Objectives
The complete separation process system is designed to produce on specification
export fluids. The actual values will depend on the specific (crude oil or gas)
properties and the transport route (Table 2).
5.2 Oilfields
Oilfields are produced at maximum plateau rates dictated by the facility design for as
long as possible. The (oil) production rate will then decline usually accompanied by
increasing water production until the minimum economic rate is reached. This is the
point at which operating expenses are no longer paid for by the oil income (Figure 2
and Table 3).
Department of Petroleum Engineering, Heriot-Watt University
11
1
PRODUCT
PROPERTY
CRUDE OIL
SPECIFICATION
Vapour pressure
< 16 psi at 25º C *
Water content
< 0.5% wt
Salt content
< 70 g/m3
Temperature
< 40º C
Pressure
~ atmospheric for tanker export
or pipeline operating pressure
* Higher values allowed pipeline export
GAS
Calorific value
~ specified limits
Liquids
- none
Hydrocarbon dewpoint
< -3º C
Water dewpoint
< -8º C in NW Europe
Carbon dioxide
< 3% wt
Hydrogen sulphide
< 4 ppm
Temperature
< 40º C
Pressure
~ pipeline operating value
* varies with ambient conditions e.g. 0º C in sub-tropical areas, -20º C in Canada
WATER
Dispersed oil
< 40 ppm for marine discharge in N.W. Europe,
< 32 ppm in Gulf of Mexico
Table 2
Typical sales/disposal
specifications
Phase
Fluid
Early
Mid-Life
Mature
Abandonment
100k
100k
30k
10k
Gas (MM sft /d)
100
100
30
10
Table 3
Water (bwpd)
<500
30k
100k
>100k
Water cut
<0.5%
30%
77%
>90%
Typical changes in
production rates during the
lifetime of a 100,000 bopd
oil field
Oil (bopd)
3
The gas associated with the oil production must be used at the time of production whether for export, fuel, lift gas, re-injection or flared. If the opportunity for pipeline
export does not exist, then projects frequently have excess gas early in their early/mid
lifes, while they are short of gas in the later, decline phase.
5.3 Gas Fields
The pattern of gas usage tends to be seasonal i.e. with periods of high and low demand:
(i) Winter peak in N.W. Europe, Northern USA and Canada, here space heating
uses a large proportion of the gas.
12
8
Oil and Gas Processing
(ii) Summer peak in Southern U.S.A. where air conditioning is the major user.
(iii) Constant demand when industry is the main user (e.g. Aluminium Smelting) or
when supplying a Liquefied Natural Gas Plant.
A typical nomination contract in N.W. Europe will specify both a total yearly
production and a minimum production rate at the end of the winter period.
The composition of any produced water will change significantly during the field life
time. Initially it will be (fresh) condensed water vapour that was originally present in
the gas phase in the reservoir. Later on, (liquid) formation water will start to be
produced from the reservoir - resulting in a gradual increase in salinity until the
formation water value is reached if large scale water production occurs. Frequently,
the easiest export route for any liquid hydrocarbons (condensate) recovered during the
separation process is to spike (inject) them back into the gas export pipeline for later
recovery at a central, onshore facility.
A typical production profile is illustrated in Figure 10 and Table 4.
Annual Peak e.g. Winter in NW Europe
Discovery
Production
Decline
Abandonment
Plateau
First Gas
Operating Expenses (Gas Equivalent)
Figure 10
A typical gas field
production profile
1
2
3
4
5
6
7
8
Years
Phase
Table 4
Typical changes in
production rates during the
lifetime of a typical
southern North Sea gas
field
Fluid
Early
Mid-Life
Declining
Abandonment
Gas (MM sft /d)
500
500
50
20
Condensate (bpd)
5000
5000
500
200
Water (bwpd)
250
500
500
200
3
Department of Petroleum Engineering, Heriot-Watt University
13
1
6.BASIC PROCESSING SCHEME
The building blocks of a simplified processing scheme are shown in Figure 11. The
reservoir fluids are produced to surface at the wellhead followed by separation into the
oil, gas and water phases. Each phase is then treated individually to reach the sales
specification.
(i) The oil content of the produced water is reduced to a level suitable for disposal.
Any recovered oil is passed to the final stages of oil treatment.
(ii) The separated gas is treated to sales specification by reducing its water and
liquid hydrocarbon content. The recovered liquid hydrocarbons are “spiked” into
the crude oil stream if it is being transported by pipeline. Alternatively, it may be
added to the sales gas pipeline, followed by onshore separation. The gas may be sold,
used for power generation, gas lift, injection back into the reservoir with any excess
being flared.
(iii) The crude oil is transported by pipeline or tanker (sea, road, or rail) after being
treated to the appropriate specifications. The latter option requires larger, local
storage than when transport is by pipeline.
Gas Injection
Gas Lift
Possible Own Use
Flare
Gas
Treatment
Gas
Pipeline
if Present
Gas
Fuel
Gas
Recovered
Liquid
Hydrocarbons
Manifold
Primary Oil/Gas/Water
Separation
Final Oil
Treatment
Oil
Pipeline
Recovered
Oil
Water
Treatment
or
Local
Storage
Tanker
Disposal
6.1 Processing Conditions
The process equipment design and the materials of construction are determined by:
14
Figure 11
A basic process scheme
8
Oil and Gas Processing
Fluid Quality
Wellhead pressure and temperature of the produced reservoir fluid. These change
dramatically over the lifetime of the field - see table 3 and module 10, section 10.2.
Fluid Properties
The properties of the crude oil are analysed in the laboratory, and phase equilibria
(“PVT”) properties are determined. This aspect is discussed in detail in the Reservoir
Engineering part of this course. Hydrocarbon mixtures can be classified into five main
types based on these phase equilibria properties as summarised in Table 5.
Hydrocarbon type
Low
Shrinkage Oil
High
Shrinkage Oil
Retrograde
Condensate
Wet Gas
Dry Gas
Field Development
type
Oil reservoir
Oil reservoir
Oil reservoir
Gas reservoir
Gas reservoir
Stabilised crude
density
< 30°API
30° < API < 50° API < 60°
API > 50°
API > 50°
Gas Oil Ratio
(sft3/bbl)
< 500
< 100,000
> 100,000
Comments
Table 5
Types of Hydrocarbons
produced
500 < GOR >
8,000
Broad phase
Narrow phase
envelope with
envelope
high proportion with few heavy
of heavy
hydrocarbons
hydrocarbons
< 70,000
Even more
lighter and
fewer heavier
hydrocarbons
The hydrocarbon fluid properties along with the wellhead pressures determine the
number of separation stages and the process conditions under which the oil/gas/water
separation is carried out. Many of the stabilised crude oils produced by the different
North Sea fields have similar properties. These are summarised in Table 6.
Property
Specific Gravity
Value
0.84 g/cm
36° API in oilfield units
Viscosity
4 cSt at 50° C
Sulphur content
< 1.0 % wt
Gas Oil Ratio
1000 scf/stb
Crude Oil Composition
50% wt paraffins
30% wt napthenes
20% wt aromatics
Wax
Asphaltenes
Table 6
Typical North Sea crude oil
properties
Gas Composition
variable, but often low (<5%)
variable, but usually low (<2%)
> 80% methane
Carbon Dioxide and Hydrogen Sulphide in the gas phase are also
normally low.
Department of Petroleum Engineering, Heriot-Watt University
15
1
The similarity of the crudes developed from the many fields means that they can be
piped to the coastal terminal in a “common carrier” pipeline. The lower the specific
gravity (the higher the API gravity), the greater the market value of the crude due to
the increased yield in high value, refinery products. Economics thus dictate that
denser, more viscous, lower API gravity crudes - such as that produced by the Alba
field - are transferred to shore via a shuttle tanker, even when a “common carrier”
pipeline with sufficient free capacity is available. This is to protect the value of the
other crude streams using the pipeline.
6.2 Process Plant Metallurgy
Maximum producing wellhead temperature and pressures along with the presence of
corrosive components (e.g. carbon dioxide, hydrogen sulphide, saline formation
waters) determine the specification (metallurgy) of the process equipment. Process
conditions can be modified so that a more economical construction metallurgy can be
chosen viz:
(i) Inlet temperatures can be reduced by cooling in a heat exchanger (seawater is
available in large quantities as a cheap, cooling medium for offshore operations)
(ii) Pressures can be reduced by use of a choke
(iii) Corrosivity can be reduced by the injection of corrosion inhibitors. These are
a mixture of chemicals injected into the process stream at a low dosage level which
coat the metal surfaces, reducing their susceptibility to corrosive attack.
6.3 Separators
Separators form the heart of the production process. There are two basic types:
(i)
Gravity separators which depend on the density difference between the phases
to be separated
(ii) Centrifugal separators in which the effect of gravity is enhanced by spinning the
fluids at a high velocity
Gravity separators are essentially large cylindrical pressure vessels up to 5m in
diameter and 20m long. They are used in either 2-phase (liquid/gas) separation, or 3phase (water/oil/gas separation). They are normally mounted in a series of 2, 3 or even
4 separators (figure 12) with inlet and outlet under pressure control. They be mounted
either vertically or horizontally.
16
8
Oil and Gas Processing
Pressure Control
at 100 bar
From Well
Manifold
PC
Choke
High Pressure
Separator
High
Pressure
Gas
LC
Pressure Control
at 20 bar
PC
Intermediate Pressure
Separator
Level
Control Valve
Figure 12
Simplified 3-stage, two
phase (gas / liquid)
separation
Intermediate
Pressure
Gas
LC
Pressure Control
at 4 bar
Low
Pressure
Gas
PC
Low Pressure
Separator
Level
Control Valve
LC
Vent Gas
(Pressure
=1.5 bar)
Crude Oil
Storage
Tank
Gravity separators (see figure 14) consist of an:
(i)
Inlet section with momentum breaker/inlet deflector to rapidly change the inlet
liquid velocity; hence helping disengage free gas.
(ii) Gravity settling section, typically sized so that:
(a)
2-phase separators: sufficient gas phase residence time such that liquid
droplets of 100 µm will separate from the gas.
(b)
3-phase separators: as for 2-phase separators plus sufficient oil phase
residence time such that 500 µm water droplets will settle into the water
phase. The separated water phase will typically contain 500 ppm
dispersed oil, hence requiring further treatment before disposal.
The above criteria are based on settling theory (see section 8.7); alternatively,
experiment may be carried out to measure the rate of oil/water separation. Typical
results for such a settling experiment are shown in Figure 13 - the initially produced
intimate oil and water mixture (emulsion) separates into a lower, "clean" (or low oil
content) water layer and an upper, "dry" (or low water content) oil layer with a (more
persistent) emulsion layer in-between.
(iii) Gas outlet with mist extractor. One design of which is a wire pad of finely
woven stainless steel wire wrapped in a cylinder. It is designed to remove liquid
droplets between 10mm and 100µm. These droplets impinge on the wire, coalesce
and flow down in to the liquid phase. Efficient operation depends on operating with
the correct gas velocity:
(a) too high gas velocity - liquid drops “eroded” from wire and is re-entrained.
(b) too low liquid velocity - liquid drops drift past mesh without impinging.
A gas scrubber is employed if larger droplets (up to 500µm) have to be removed since
a mist eliminator would flood with this level of liquid loading.
(iv) Liquid outlet under level control to evacuate liquid or separate oil and water
phase (2/3-phase separation operation, respectively). The outlet is usually equipped
with a vortex breaker to prevent re-entrainment of gas.
Department of Petroleum Engineering, Heriot-Watt University
17
"CLEAN" WATER
"DRY" OIL
hw/h
ho
EMULSION
he
h
"DRY" OIL
hw
1
EMULSION
"CLEAN" WATER
Figure 13
Separation of oil / water /
emulsion liquid with time
Time
Appearance Oil/Water Emulsion Sample
Behaviour of Oil/Water Emulsion With Time
6.3.1 Horizontal Separators (Figure 14)
These are most suited to separation of large volumes of gas from liquid. Their
advantages/disadvantages compared to vertical separators are:
(i)
Larger interface area gives better foam/emulsion handling characterisation
(ii) Can be modularised, but require larger surface area
(iii) Solids removal is less efficient and requires a more complex jet wash system
(iv) Lower surge capacity i.e. reduced ability to deal with uneven, inlet flow
Pressure Control Valve
Inlet Deflector/
Momentum
Breaker
PC
Gas
Mist Eliminator
Weir
GAS
Inlet
From
Production
Manifold
OIL and EMULSION
OIL
WATER
To Produced
Water Treatment
Oil Outlet and
Level Control
Water Outlet and
Level Control
To Oil Export
6.3.2 Vertical Separators (Figure 15)
These are most suited for separation of gas from large volumes of liquid. Compared
to horizontal separators they are:
(i)
Good for uneven, surging inlet flow due to their greater height
(ii) The oil/gas and oil/water interface level control is less critical for the same reason
(iii) However, they do tend to be larger than their horizontal equivalent for the same
separation capacity
18
Figure 14
Three phase horizontal
separator
8
Oil and Gas Processing
Pressure Control Valve
PC
Inlet Deflector/
Momentum
Breaker
Gas
Mist Eliminator
Inlet
Oil Outlet and
Level Control
To Oil Export
OIL
WATER
To Water
Disposal
Figure 15
Vertical three phase
separator
Water Outlet and
Level Control
6.4 Operational Production Problems with Gravity Separators
The process equipment used in platform operations is designed to achieve the required
product specifications using standard chemical engineering design procedures. However, a number of operational problems referred to above can be encountered if
allowance is not made for the fact that we are not dealing with “pure” fluids. The
complex mixture that makes up the produced hydrocarbon contains many minor
components that cause:
6.4.1 Foaming:
Presence of semi stable gas bubbles at the oil/gas surface that prevent the gas
disengaging quickly and cleanly from the liquid surface. A de-foaming chemical
(surfactant) is injected at a low concentration to overcome this problem; coupled with
the use of a mist eliminator which will remove liquid droplets from in the size range
10-100µm.
6.4.2 Solids
Low concentrations of solids are frequently produced to surface with the well fluids.
They may be present in a low concentration but represent a large absolute solid
volumes given the large volumes of produced fluid These solids range from “fines”
- (micron) sized clay particles which can flow through the formation pore throat
structure - to individual sand grains / larger “lumps” of failed formation which can
be produced from weak or unconsolidated formations. These solids collect in the base
Department of Petroleum Engineering, Heriot-Watt University
19
1
of the separators where they reduce the separator performance (shorter residence time)
and can lead to corrosion since the fluid is stagnent. They have to be regularly
removed. Water jets are mounted at the bottom of the tank - the jetting action resuspends the solids followed by transport to a holding tank for ultimate disposal.
Alternatively, manual removal of the solid deposit is required.
Several other solid phases can form in the gravity separators and the other components
of the production system. These include:
(i) Scale
By the mixing of incompatible waters {e.g. barium sulphate from mixing sea water
(sulphate source) and produced water (source of barium), or due to pressure/
temperature changes (e.g. to calcium carbonate scale)}. The point of deposition can
be controlled by the addition of scale ‘inhibitors’.
(ii) Wax
Cooling of the crude oil can result in its paraffin content precipitating as a solid wax.
This is avoided by preventing the crude cooling below the wax cloudpoint temperature; while the temperature at which deposition occurs can be controlled to some
extent by the use of wax inhibitors.
(iii) Asphaltenes
Pressure reduction below the bubble point and the consequent loss of the more
volatile compenents can lead to the precipetation of asphaltenes from some crude oils
6.4.3 Emulsion
Momentum breakers and corrugated plate settling packs are installed internally in the
separator to promote oil droplet coalescence and separation. However, emulsion
separation may not be sufficiently complete within the residence time available in the
primary separator. This depends on the chemical properties of the crude oil/water
system (see section 6.3) and the time/rate of shear that the fluids have been subjected
to during the production process. Frequently used, practical solutions include the
addition of heat (viscosity reduction) or emulsion breaking chemicals together with
acceptance of reduced oil and water quality being passed to the subsequent separation
stages. Reverse emulsion breakers can be employed for treating water-in-oil emulsions.
6.4.4 Surging Flow
The primary separator provides pressure control on all three (gas/oil/water) outlets. It
controls residence time via the oil/water level controls. Production wells frequently
do not produce fluid at a constant even flow rate. This is due to the length and topology
of the production tubing and flow lines. These problems are accentuated by
incorrectly set gas lift (well heading) as well as the long flow lines associated with
subsea wells. The separation system must be capable of dealing with the resulting
high, instantaneous feed rates which can cause the levels to increase above their
normal, operating values. The control system needs to be suitably adjusted to deal
with this.
20
8
Oil and Gas Processing
6.4.5 Production Chemicals
In addition to those already discussed, other production chemicals used include
biocides, corrosion inhibitors. Fig 16 is a typical example of where they might be used
in a oil/gas/water separation system.
Gas Dehydration
Sour Gas
Gas
Sweetening
Fuel Gas
Gas
Injection
Water
Gas
Cooler
Glycol Regeneration
NGL
NGL
Recovery
Cooler
Defoamer
Compressor
Oil Cooler
Cooler
Knock Out
Compressor
Drum
HP
Sep
Corrosion
Inhibitor
NGL
Metering
MP
Sep
Compressor
Knock Out
Drum
Booster
Pump
LP
Sep
Export
Pump
Crude Oil
Metering
Demulsifier
Deoiling
Chemical
Gas
Oily Water Treatment
Flash
Drum
Scale
Inhibitor
Figure 16
Typical Oil / Gas / Water
process scheme
Oil Pipeline
Cooler
Wax
Inhibitor
Corrosion
Inhibitor
Well Fluids
Oil (& NGL)
Gas
Water
Glycol
Scale Corrosion
Inhibitor Inhibitor
From
Closed Drains
Slop
Oil
Produced Water
Disposal
Caisson
Corrosion
Inhibitor
Wax
Inhibitor
From
Open Drains
*: Deoiling Chemicals
To LP
Separator
6.4.6 Miscellaneous Processes and Comments
A fresh water wash may be required to reduce the salt content in the crude oil to the
specification value if the formation water has a high salt content. This involves mixing
the crude with the water followed by subsequent separation.
The main separation train may be twinned (i.e. two separation trains of equal capacity)
to increase the reliability (uptime) of the separation facility. In addition a smaller test
separator will almost always be installed to allow individual well rates to be measured
as part of the field monitoring programme.
7. HORIZONTAL SEPARATOR SIZING BASICS
The liquid droplets will settle at a velocity determined by equating the gravity force
on the drop with the drag force caused by its relative motion to the fluid continuous
phase. Typical tragectories of the settling liquid droplets are shown in Figure 17. This
assumes that the internals of the separator have been designed so that turbulence
effects are minimised e.g. by use of inlet deflectors/momentum breakers, vortex
breakers on the outlet.
Department of Petroleum Engineering, Heriot-Watt University
21
1
GAS
GAS
Vf
Vf
Vw
OIL
OIL
Vw
WATER
Vo
WATER
Vo
Plan View
Axial View
Vw = Water droplet settling velocity in oil
Vo = Oil droplet rising velocity in water
Vf = Liquid droplet settling velocity in gas
7.1 Gas Capacity
Separators are typically designed to allow liquid droplets larger than 100µ to settle
from the gas phase to the liquid interface. The maximum allowable gas velocity (Vg)
which achieves this separation may be calculated by the Souder-Brown equation:
Vg = k
where
pl
pg
k
pl - pg
pg
= liquid density
= gas density
= constant
The constant k is related to the diameter of the droplet to be separated, the gas viscosity
as well as the densities of the liquid and gas phases. Calculation of the constant k is
complicated by the fact that the settling flow regime is not laminar, and qualitative
calculations are beyond the scope of this text. However, suitable correlation’s can be
found in the Chemical Engineering literature (e.g. Arnold and Stewart, in section 8.11).
The minimum area required for gas flow (Ag) is then:
Ag = Q g / Vg
where Qg is the specified maximum gas flow.
7.2 Liquid Capacity
The separation velocity (V) of one liquid from another is described by Stokes Law:
22
Figure 17
Setting droplet trajectories
8
Oil and Gas Processing
v=
Kd2 (ρd -ρc)
µc
d
ρd
ρc
µc
K
=
=
=
=
-
where
Also
minimum specified droplet size to be separated
density discontinuous phase
density continuous phase
viscosity of continuous phase
a constant
Required droplet separation time = vertical height of continuous phase
separation velocity
The separation of water droplets from oil is normally more difficult than oil from water
since oil is normally more viscous than water. If no other information is available,
field experience indicates that if 500µm water droplets are removed from the oil the
resulting oil-in-water content will typically be 2000 ppm or less. Similar arguments
apply to the calculation of the constant K as was discussed for gas capacity. These
calculations, along with the specified liquid flow rates, allow the area required for
liquid flow to be specified.
An alternative approach is to base the separation requirements on emulsion separation
tests which specify a required separation time (Section 6.3 and Figure 13). Separators
are typically run 50-75% liquid full, so knowledge of the design liquid flow rates and
the required residence times allow the separator length and diameter to be calculated.
An increased separator diameter is required to compensate for surging flow as well as
allowing sufficient volume to give an adequate response time between high/low level
alarms before the high/low level trip leads to a shut down. An additional 30-60
seconds hold-up volume is typically specified. The resulting performance of a typical
three stage separation process is summarised in Table 7
Separator
Stage
Table 7
Performance targets for
typical 3-stage separation
system
Water
Residence Time
(min)
Oil
Residence Time
(min)
Target Water
in Oil
(%vol)
Target Oil
in Water
(ppm)
1st
4
2
2
2000
2nd
>6
4
1
400
3rd
>8
6
0.25
100
Department of Petroleum Engineering, Heriot-Watt University
23
1
8. TEST SEPARATOR
The production from each well has to be regularly measured accurately as part of the
field's well surveillence programme. Accurate three phase flow meters have only
recently become available, so traditionally a small three phase separator (the test
separator) has been used for this purpose. Here, the well production is split into its
constituent phases (oil/water/gas); the flow rates of each measured separately with
conventional orifice meters. The oil and water flows recombined for further processing.
Orifice Box
Gas Out
Gas Measurement
Inlet
from Well
Water Flow Meters
Interface
Controllers
BSW
Oil Flow Meters
Recombined
Flow
Gas
Oil
Water
Multi phase flow meters are now becoming reliable. This is reducing the need to install
a test separator (platform space and weight saving) as well as the need to install a
second (test) pipeline in subsea field developments between the manifold and the host
facility (module 10, figure 7).
9. COMPRESSORS
The gas liberated from the produced fluids must be delivered to the export pipeline at
the specified pressure. It is flashed from the produced liquid at a variety of pressures
during the production process - so economically achieving this export pressure
requires a number of gas compressors rather than one large one. Minimising the
compression power requirements while maximising the liquid recovery (determined
by the thermodynamics of the produced fluids), will dictate that the gas from each oil
separation stage is compressed to the operating pressure of the previous stage. This
is schematically illustrated in Figure 19.
24
Figure 18
A test separator
8
Oil and Gas Processing
(Pressure
Gas = 100 bar)
3rd Stage Compressor
(Pressure
= 20 bar)
2nd Stage Compressor
(Pressure
= 4 bar)
1st Stage Compressor
Out
LKO
Gas from high
pressure separator
LKO
LKO
Gas from intermediate
pressure separator
Gas from low
pressure separator
LKO
Vent gas from
storage tanks etc.
Liquid out
KEY
Figure 19
Schematic 3-stage
(Pressure
= 1.5 bar)
Liquid Knock Out vessel
Cooler
Compressor
compression scheme
The action of the compressor performing work on the gas being compressed raises its
temperature considerably (the “bicycle pump effect”). Minimisation of the compressor power requirements together with maximising of the liquid recovery (which
condenses on cooling the gas) dictates the use of interstage coolers, as shown. Offshore, sea water is used for this cooling duty.
An example of the combined result of the above effects is illustrated in Table 8.
Separator Pressure(bar)
Table 8
Effect of multiple separator/
compressor stages
3
Oil Production(m )
Compression Power *(kw)
80, 5
1300
640
80, 30, 5
1315
370
80, 30, 12, 5
1321
295
*Omission of compressor interstage coolers increases this by 50%
The prime mover or engine chosen to power the compressor depends on the
compressor’s location and the power required. Gas turbines, diesel engines and
electric motors are all frequently employed.
Centrifugal and positive displacement reciprocating compressors are both commonly
used in oil field applications. Both compressor types are susceptible to damage by
liquid droplets, hence the presence of the liquid knockout vessels prior to each
compressor.
Department of Petroleum Engineering, Heriot-Watt University
25
1
Centrifugal compressors are large, complex machines that contain internals that rotate
at high velocities. They are much more difficult to install and maintain than crude oil
or water pumps. One of the critical areas in compressor design is to ensure that a recycle valve is included which opens to prevent compressor “surge”. This occurs when
the feed rate is insufficient to allow the compressor to reach its design discharge
pressure.
The subsequent treatment dehydration etc. of the gas to achieve export specifications
once it has been compressed is similar to that described in gas field operations.(Section
8.10).
10. OIL EXPORT
The volume of oil being exported has to be measured to the highest accuracy - since
not only does it define the project product and cashflow, but also involves both fiscal
(tax/royalty etc.) and intercompany transfer (e.g. to a “common carrier” pipeline)
aspects.
Turbine flow meters - which involve the measurement of the number of times that the
flowing oil revolves a paddle or turbine placed in the oil flow path - have a high,
intrinsic accuracy and are normally used for this purpose. However, they have to be
calibrated - or “proved” at regular intervals. A bank - say 5 to 10 - of smaller capacity
turbine flow meters are used rather than a single, large meter. The calibration process
is illustrated in Figure 20. The flow rate as measured by the turbine flow meter is
compared with that calculated from the time for a sphere to be displaced between two
detectors and the volume between these detectors.
26
8
Oil and Gas Processing
Inlet Header
Open
Open
Turbine
Flow Meter in Nomal Operation
Turbine
Flow Meter Being Calibrated
Sphere
Turbine
Flow
Meter
Sphere
Deflectors
Open
Shut
Reversing Valve
Shut
Open
Export Header
Figure 20
Meter calibration system for
oil export pipe line
Oil export depends on efficient pipeline operation. The pipeline requires regular
cleaning by a “pig”. This removes settled sand, stagnant water collected at low points
(for corrosion prevention), wax deposits etc. The “pig” may be in the form of a sphere
to displace fluids or a cylinder with brushes to scrape the inside surface of the line.
Alternatively it may have “intelligence” in that it can inspect the pipeline condition
Department of Petroleum Engineering, Heriot-Watt University
27
1
and record the results for later replay. The “intelligent” pig uses similar techniques
as employed for well (tubing) condition monitoring {mechanical arms to “feel” for
corrosion pits and grooves, acoustic wall thickness measurement devices etc.}. As an
aid to recovery, they often carry a transmitter so that their position can be pinpointed
if they become stuck in the pipeline.
The pig is “launched” into the pipeline using a pig launcher (see Figure 21). This is
an oversize barrel with end closure, pressure gauge and venting/purging system to
allow any hydrocarbons present in the “launcher” to be depressurised/disposed of in
a safe manner. Opening of any pressurised lines which could potentially contain
hydrocarbons is a hazardous operation which requires proper attention to safety and
environmental aspects.
A pig “signaller” is built into the launcher.This confirms the successful launch of the
pig into the pipeline. A duplicate pig “signaller” and “launcher” is installed at the
receiving end of the pipeline to record the arrival of the pig and to allow its safe
removal from the pipeline.
Vent
Pressure Guage
Pig Passage Signaler
Pig
Launcher
End
Closure
Pig
C
Pipeline
Purge
A
B
Drain
Pipeline
Status
Valve A
Valve B
Valve C
Normal Pipeline Operation
Closed
Open
Closed
Displace Pig in Launcher
Open
Closed
Open
11. GAS HANDLING
Similar process units are used in oil field gas handling as are used in gas fields. In both
cases the objective is to maximise the recovery of liquid Natural Gas Liquids or NGL's
(mainly hydrocarbons with a chain of four or five carbon atoms). This is achieved in
oil fields by multi-stage separation and (gas) compression - as discussed in the
previous chapter. Single stage separation, with optional gas compression and
processing as dictated by the producing wellhead pressure and gas composition
respectively, is the norm for gas fields. Figure 22 illustrates typical gas field facilities.
11.1 Increasing NGL Recovery
The increased NGL yield, and reduction in the water content, is achieved by cooling
the gas. The equipment used is:
28
Figure 21
Schematic of pig launch
system
8
Oil and Gas Processing
(i) Refrigeration to between -30ºC and -40ºC using a conventional refrigeration
plant. (Freon and Propane as refrigeration agents, respectively).
(ii) Joule-Thomson expansion - the gas will be cooled when its pressure is
reduced by flowing through a throttle or choke (isenthalpic conditions). This process
is favoured when there is a large pressure drop between wellhead and pipeline
operating pressures. The gas may require compression back to the export pipeline
operating pressure subsequent to the separation of the condensed liquids.
(iii) Turbo Expansion - this involves extracting energy from the gas by getting it
to do work e.g. powering a compressor or generating electricity while it’s pressure
is being reduced. This achives a lower temperature than for Joule Thomson
expansion e.g. Shell’s St. Fergus gas plant cools the gas to -99ºC using two-stage
turbo expansion.
(iv) Processing of Chilled Liquids
Glycol is added prior to cooling the gas so that any condensed water does not form
(solid) ice or hydrates (see Section 10.3.2). The cool water/glycol/liquid hydrocarbon
mixture is lead to a three-phase separator where the water/glycol mixture is removed
from the bottom while the hydrocarbon liquid (middle layer) is pumped to a
distillation column {see Section 10.2 on NGL Stabilisation for further details}.
Adjustment of the distillation conditions (number of trays, bottom temperature,
pressure etc.) allows ethane, propane and butane to be recovered separately e.g. for
use as a chemical feed stock. The (denser) NGL’s are recovered from the column
bottom. Cryogenic distillation conditions (<-50ºC) are necessary if recovery of
substantial volumes of ethane is desired.
Liquid
Separation
NGL
Gas
Gas
Processing
Dehydration
Gas
Sales
Water
Cooling
Water
Gas
Wells
Produced High Pressure
Fluids
Separation
Compressor
NGL
Water
NGL
Stabilisation
Column
Figure 22
Schematic gas field facility
system
NGL
Sales
Oily Water
Disposal
Department of Petroleum Engineering, Heriot-Watt University
29
1
11.2 NGL Stabilisation
The separated NGL has a high vapour pressure due to dissolved, volatile gasses
(methane, ethane etc.). It is not suitable for storage in a tank or onward transport to
customers. The vapour pressure is reduced by heating the NGL to progressively
higher temperature and allowing the gas to “flash-off” at constant pressure (Figure
23) {See section on PVT properties in Reservoir Engineering module} The same
process can be carried out more efficiently in a distillation column. The column
consists of a number of trays containing bubble caps (Figure 24) which forces the gas
rising upwards into intimate contact with the liquid on the tray, ensuring they come
to equilibrium. The higher trays are operated at a progressively lower temperature
than the lower trays. Hence the liquid undergoes a series of multiple flashes at
increasing temperature and constant pressure as it trickles down the column from the
top tray to the bottom tray. Any water that is present will tend to concentrate at the
tray operating at just below its boiling point under the column operating pressure. A
separator may be installed on this tray and the water drawn off for disposal.
PC
Pressure Control
Unstabilised
NGL
Low
Temperature
Separator
Gas
Gas
Intermediate
Temperature
Separator
Heater
High
Temperature
Separator
Heater
Stabilised NGL
Storage
Cooler
Packing material is frequently used as an alternative to distillation trays. These ensure
that the liquid and gas come into equilibrium by providing a large surface area with
which the upcoming gas and the downward flowing liquid can brought into equilibrium. There are many proprietary types of packing - some examples are illustrated in
Figure 25.
.
11.3 Gas Dehydration
Produced Natural gas is saturated with water vapour under the prevailing downhole,
reservoir conditions. Production to the surface and subsequent processing normally
involves a progressive reduction in the gas temperature (and pressure). The equilibrium
water content of the gas phase decreases as the temperature drops or the pressure
increases, resulting in the separation of the liquid water. (Figure 26). Water presents
as vapour in the gas phase does not present a problem to the gas facility operation unlike liquid or solid water (ice):
30
Figure 23
Process scheme for NGL
stabilisation via multiple
flashes at constant pressure
and increasing temperature
8
Oil and Gas Processing
Gas
Gas
Distillation
Column
Produced Fluid
Top Tray
From Wells
Bubble Caps
Separator
Tray
Separator
Gas
Water to Oily
Water Treatment
Gas
NGL
Water to Oily
Water Treatment
Gas
Heat
Reboiler
Gas
Vent
Cooler
Figure 24
NGL separation and
stabilization
Cold
Water
NGL
Storage
Customers
Liquid water accumulates at pipeline low points, reducing well capacity and accelerating corrosion. It also forms solid hydrates (see Section 10.3.2) which, similar to ice,
can plug lines, valves etc.
Figure 25
Examples of column
packing
Department of Petroleum Engineering, Heriot-Watt University
31
1
The saturated water content of natural gas under various conditions of temperature and
pressure can be derived from Figure 26 The Dew Point is the temperature at a given
pressure, at which the gas is saturated with water vapour, and any further reductions
in temperature will cause the water to begin to condense. Thus, if the water content
of a gas stream is known, its Dew Point can be derived from Figure 26.
Water Content of Hydrocarbon Gas
80000
80000
Correction Factor for Gas Gravity
60000
1.0
40000
60000
50 ºF
100 ºF
150 ºF
200 ºF
CG
0.9
40000
250 ºF
0.8
300 ºF
20000
20000
Gas Relative Density
0.7
0.6
1.0
25
1.2
30
1.4
35
1.6
40
45
1.8
50
10000
Molecular Weight
8000
8000
6000
6000
Correction Factor for Salinity
0.98
ps
ia
0.96
0
40
0
60
0
00 10
00
30
20
0
50 0
00
80 00 40
00
6 00
10 000
00
0
0
0
15
600
80
4
30
3
Total Solids in Brine. %
800
50
2
10
1
20
0
0
0.90
0
50
0.92
1000
2000
.7
0.94
14
2000
4000
1.00
25
Cs = H20 From Brine
H20 From Water
4000
400
100
80
60
40
800
600
400
100
80
60
40
20
20
50
6
Position of this line
is a function of gas
composition
10
8
6
4
20
0
30
0
50
0 40
80
0 60 0
15
0
00 10
00
20
00
8
10
0
10
4
2
Water contents of natural gasses with
corrections for salinity and relative density.
After McKetta and Wehe, Hydrocarbon
Processing. August 1958
2
1
1
-60
-40
-20
0
20
40
60
Temperature ºF
32
1000
200
HYDRATE FORMATION LINE
200
14
.7
ps
25
ia
Water Content of Natural Gas (lb water/million cu ft. of wet gas at 60º F and 14.7 psia)
0.8
20
10000
80
100 120 140 160 180 200
240
280
Figure 26
Dew point of natural gas.
(adapted from Gas
Processors Suppliers
Association, Engineering
Data Book, 10th. Edition)
8
Oil and Gas Processing
11.3.1 Dew Point Depression
Dew Point Depression is the number of degrees that the dew point (or water
condensation temperature) has to be lowered by the dehydration process e.g. a 80ºF
Dew Point Depression for a gas at 100ºF and 1000 psi would require the removal
{60 - 4.5 =} 55.5 lbs water/MM ft3 gas and implies that liquid water condensation
would not occur above 20ºF (at 1000 psi) or even -10ºF if the pressure was reduced
to 200 psi.
11.3.2 Hydrates
Hydrates are solid materials which form from light hydrocarbons and liquid water.
The light, hydrocarbon molecule is embedded in the solid water crystalline lattice,
which makes up the majority of the compound’s composition. Viz:
Methane
Ethane
CH4 • 7H2O
C2H6• 8H2O
Figure 26 shows that they are formed above the normal freezing point of water and
indicates the maximum temperature at which they are encountered.
One solution to preventing operational problems from hydrate formation is to add an
(expensive) inhibitor e.g. methanol to the gas stream or to reduce the water content
of the gas such that liquid water cannot form under the full range of operating
conditions. Water vapour removal is achieved by:
(I)
Refrigeration or cooling (see section 10.1)
(ii) Continuous absorption in a liquid desiccant (e.g. glycol)
(iii) Batch absorption by a solid desiccant (e.g. silica gel)
11.4 Continuous Dehydration Process
The process scheme for counter current absorption in a (packed) column using a liquid
desiccant such as (triethylene) glycol is shown in Figure 27. The wet gas enters the
contactor column at the bottom while the “lean” (low water content) glycol is pumped
in at the top. The glycol trickling down the column is brought into intimate contact
with the gas bubbling upwards. The water vapour is absorbed by the glycol since the
partial pressure of water in the gas phase is greater than its partial pressure in the liquid
(glycol) phase. The lower the water content of the “lean” glycol, the lower the
resulting dew point temperature and water content of the gas phase.
Department of Petroleum Engineering, Heriot-Watt University
33
1
Dry Gas Outlet
Water Vapour
Reflux Condenser
"Rich" Glycol / Water Mixture
Column
Glycol Boiler
Glycol Stripping
Column
Wet Gas Inlet
Glycol / Gas
Column
Low Pressure
"Lean"
Glycol
Storage
Gas Dehydration
Glycol Regeneration
Stripping Gas
(Optional)
The “rich” glycol/ water mixture is drawn off from the bottom of the contactor tower.
The absorbed water vapour dramatically lowers the mixtures boiling point, so that the
glycol may be regenerated by boiling in a distillation column. It works in a similar
manner to those described earlier in section 10.2; except that a condenser (cooler) is
placed at the top of the columns. The temperature of the condenser is adjusted so that
the glycol condenses and is returned as column reflux, while the water vapour passes
through it as a gas. The “lean” glycol is drawn from the bottom of the still and can
now be used to treat further more wet gas. The water content of the “lean” glycol can
be further reduced by gas stripping. This involves bubbling a gas through the lean
glycol - either in the regeneration column itself or, if higher purity glycol is required,
in a separate small column prior to the gas being passed to the main, regeneration
column. Low pressure (say 50 psi at 60ºF) gas is used since increasing its temperature
to column operating temperatures will increase its water carrying capacity from, say,
250 to 100,000 lbs/MMscf.
Triethylene glycol can typically achieve a dew point depression of 55ºC by operating
the glycol reboiler at 205ºC. This can be increased to 85ºC by use of gas stripping.
Further reductions in the dewpoint e.g. to -100ºC for gas destined for a cryogenic
liquefied natural gas plant, require use of a solid desiccant.
11.5 Batch Dehydration Process
Very low dew points, such as those required for cryogenic LNG plants where a water
vapour concentration of less than 1ppm is required, can only be achieved with solid
bed dehydration systems such as illustrated in Figure 28. It consists of:
(i)
two or more desiccant filled contactor vessels,
(ii) a heater to supply hot, regeneration gas,
(ii) a cooler, to condense water from the used regeneration gas, together with a separator.
The solid desiccant (e.g. silica gel, alumina gel, molecular sieve etc.) has a very large
specific surface area. This allows it to absorb water from the (wet) gas passing
34
Figure 27
Continuous gas
dehydration
8
Oil and Gas Processing
downwards through the contactor at low near ambient temperatures (30º - 40ºC).
Once the entire bed has become saturated with water, the inlet (wet) gas is switched
to a second tower containing fresh absorbent. The saturated (spent) tower is then
regenerated by passing heated (260º - 350ºC) gas through it. The water absorbed by
the solid dessicant is vaporised into the hot gas stream. Cooling of this hot, wet gas
reduces the water saturation level and liquid water is recovered from the separator. A
hot tower will not work efficiently. The dessicant bed needs to be cooled - by passing
cool inlet gas - to its operating temperature (30º - 40ºC) before it will work efficiently.
The “fresh” tower is then available for operation when the tower currently in use
becomes saturated with water vapour. The towers are typically sized so that they work
on an 8-hour adsorption cycle followed by regeneration through six hours heating and
two hours cooling.
Regeneration
and Cooling Gas
to Compressor
Separator
Inlet, Wet Gas
Cooler
Water
Absorbing
Tower
in Operation
Figure 28
Schematic flow diagram for
solid bed dehydration
Dehydrated
Sales
Gas
Tower
Being
Regenerated
Regeneration
Gas
Heater
Cooling
Gas
11.6 Acid Gas Treating
Natural gas frequently contains other contaminants than those discussed to date.
These include carbon dioxide (CO2), hydrogen sulphide (H2S) and other sulphur
compounds such as mercaptans. Since these materials form acidic solutions when
dissolved in water, they are known as acid gasses. These compounds are undesirable
since they:
(i) Cause corrosion
e.g. CO2 is corrosive in the presence of liquid water at a partial pressure (=total
pressure * mole% of CO2) of 30 psi. H2S can lead to sulphide stress cracking and
hydrogen embrittlement of many metals at partial pressures as low as 0.05 psi}
Department of Petroleum Engineering, Heriot-Watt University
35
1
(ii) reduce the heating value of the gas and,
(iii) in the case of H2S, can be poisonous in quite low concentrations
(H2S has a distinct odour at 0.15 ppm, exposure to 100 ppm H2S leads to drowsiness
after 15 minutes while 500 ppm causes sufficiently severe breathing problems after
5 minutes that prompt the requirement of artificial respiration. Unfortunately, H2S
cannot be smelt at these lethal concentrations.
The maximum allowable CO2 and H2S concentrations are normally specified in the
gas sales contract - typical values are 3% and 4 ppm respectively. Low concentrations
of H2S may be removed by solid absorbents (e.g. iron oxide, zinc oxide) which are
replaced when the bed is spent. Higher concentrations of H2S as well as CO2 are
removed in solvent extraction process similar to a glycol dehydration unit (section
10.4 and figure 26).
The extraction unit consists of two parts - an adsorption column in which the acidic
gas to be treated is fed in at the bottom and the liquid solvent is added at the top. The
column internals - trays or packing - ensure that there is intimate contact between the
two phases and the process conditions adjusted so that the CO2 / H2S concentrations
are reduced to specification levels.
One class of solvents used for treating acid gasses are alkaline liquids that reversibly
react with the acidic gases H2S and CO2 e.g. mono- ethanolamine.
2 RNH2 + H2S
heat
(RNH3)2 S
(RNH3)2 S + H2S
heat 2(RNH3)HS
2 RNH2 + CO2
heat RNHCOONH3 R
An alternative is a physical solvent e.g. Sulfinol in which one can dissolve large
quantities of CO2 and H2S at low temperature; while releasing them again at high
temperatures. Thus, in both cases, a solvent “rich” in absorbed CO2 and H2S is
recovered from the bottom of the adsorption column and transferred to the regeneration column, where heat liberates the absorbed CO2 or H2S. The regenerated solvent
is used again. The gasses are vented or flared {H2S being converted to sulphur dioxide
(SO2)}. However, there are normally strict environmental constraints on the quantities of H2S and SO2 that can be released into the environment; while the emission
of CO2 attracts a tax penalty in some countries. Alternatives are:
(i)
Injection into an underground disposal reservoir
(ii) Conversion of the H2S into solid sulphur which can be sold to the chemical
industry. The Claus process is frequently used - this involves the oxidation of part
of the H2S to SO2; followed by conversion to sulphur by reaction with further H2S.
2 H2 S + 3 O2 → 2 SO2 + 2H2 O
SO2 + 2H2S → 3 S + 2H2O
36
8
Oil and Gas Processing
12. FURTHER READING
Arnold K and Stewart M. "Surface Production Operations"
Volume 1 (2nd Edition, 1998) and Volume 2 (1st Edition, 1989)
Published by Gulf publishing Company
ISBN 0-88415-821-7 and 0-87201-175-5
Kennedy J "Oil and Gas Pipeline Fundamentals"
Published by Penwell Books, 1993
ISBN 0-87814-390-4
.
Department of Petroleum Engineering, Heriot-Watt University
37
1
Well Control
CONTENTS
1.
CHEMICAL PROPERTIES OF FORMATION
/ PRODUCED WATER.
1.1.
Mineral Scale
1.2.
Scale Inhibiton
1.3.
Corrosion
1.3.1.
Carbon Dioxide Corrosion
1.3.2.
Oxygen Corrosion
1.3.3. Hydrogen Sulphide
2. PRODUCED WATER TREATMENT
2.1.
Introduction
2.2.
(Corrugated) Plate Interceptors
2.3.
Flocculation / Coagulation
2.4.
Flotation
2.5.
Hydrocyclones
2.6.
Coalescer Units
2.7.
Centrifuges
3. OTHER SOURCES OF (PLATFORM)
WATER REQUIRING TREATMENT
4. DISPOSAL OF PRODUCED WATER
4.1.
Marine Discharge
4.1.1. Production Chemicals
4.2.
Reducing Environmental Impact of
Produced Water Discharge
5. WATER INJECTION
5.1.
Water Sources
5.2.
Water Quality
5.3.
Injection Water Treatment
5.3.1. Example Process Schemes Employing
Coarse and Fine Filtration
5.3.2. Suspended Solids Removal
5.3.3. Filter Systems
5.3.4. De-Oxygenation
5.3.5. Hydrogen Sulphide Removal
5.3.6. Chemical Treatments
6. FURTHER READING
9
Water Handling
1
LEARNING OBJECTIVES:
Having worked through this chapter the student will be able to:
• Discuss the (highly) variable chemical composition of formation water
• Describe operational problems associated with production water (scale, corrosion, etc)
• Identify equipment used to treat oily water separated by the primary gravity
separation equipment (corrugated plate interceptor, flocculation, coagulation,
(gas) flotation and hydrocyclones and coalascers)
• Discuss the disposal options for produced water
• Identify the sources of water to be used for water injection
• Outline the processes required to “clean up” water for water injection (filtration,
deoxygenation, chlorination, desulphation etc)
2
9
Water Handling
1. CHEMICAL PROPERTIES OF FORMATION / PRODUCED WATER
Formation water, and hence produced water, show a wide range of chemical
compositions. The formation water and their formation rocks are in equilibrium with
one another under the reservoir conditions of temperature and chemical pressure,
where they have normally been in contact with one another for a long time. The range
of formation water chemistry encountered in the North sea is illustrated in Table 1.
The first three fields can be regarded as normal with the relatively low salinity {Total
Dissolved Solids (TDS) 25,000 - 60,000 ppm}. The latter three fields illustrate the
more extreme values encountered in high salinity brines with high cation levels in
general, in barium cation (Ba++) concentration and in sulphate anion (SO4) levels
respectively.
1.1 Mineral Scale
The production process involves bringing formation fluids from their original hot,
high pressure reservoir conditions to the surface. This temperature and pressure
reduction can result in the solubility of one or more minerals in the water phase
becoming less than the amount present in the water. The mineral precipitates as a solid
phase (scale) which separates from the water.
A second source of scale is mixing of:
(i)
Incompatible formation waters (if more than one reservoir zone is being
produced at the same time), or
(ii) (Incompatible) formation water and injection water (frequently sea water) are
being produced from different perforations producing from the same formation
Frequently, the amounts of scale that can be produced are apparently small - in the
order of mg solid per litre of water. However, a large number of litres are produced
from a well in a short period of time. Thus the potential rate of deposition {or volume
of solid scale produced in unit time} is large. These solid precipitates may either
remain as a suspension in water or form a layer adhering to the pipe wall. Such layers
restricts fluid flow in the tubing, as well as plugging restrictions such as (narrow)
orifices etc.
(Calcium) Carbonate (CaCO3) scales are normally caused by decreases in the calcium
bicarbonate containing fluids temperature and pressure.
Ca (HCO3)2
CaCO3 (dissolved) + H2O + CO2 (dissolved)
→ Ca CO3 (solid) + H2O + ↑ CO2 (gas)
The partial pressure of Carbon Dioxide (CO2) above the liquid phase, and hence the
concentration of CO2 that will dissolve in the liquid, will decrease when the pressure
is reduced. CO2 will be liberated from the liquid to the gas phase; driving the above
equilibrium to the right and precipitating part of the CaCO3. The presence of other
salts e.g. a sodium chloride concentration to 200,000 ppm will significantly increase
the solubility of CaCO3 {up to 250% for this example}.
Department of Petroleum Engineering, Heriot-Watt University
3
4
125
438
200
1,500
1,820
341
462
19,000
18,469
9,000
69,200
28,800
23,800
10,890
Miller
Piper
(Carboniferous)
Na
K+
1366
667
115
3,700
50
408
250
Mg++
428
3,480
1,060
25,500
250
1,856
1,900
Ca++
8
156
110
1,100
25
324
165
Sr++
0
1
1,030
15
60
14
12
Ba++
extracted from : ‘North Sea Formation Water Atlas’ edited by E.A. Warren and P.C. Smalley
Geological Society Memoir No. 15
Published by The Geological Society, 1994
* TDS - Total Dissolved Solids (ppm or mg / l)
Sea water
Ravenspurn
Brent
Troll
Beatrice
Field
+
Ion (ppm)
Examples of the range of North Sea Oil Field Formation Water Chemistry
Si
-
-
-
55
-
-
-
Fe
-
-
10
160
-
-
0.1
Cl'
19699
41,890
47,680
142,200
14,300
33,288
35,530
SO"4
2962
1,845
7
260
10
-
-
HCO'3
123
691
2,070
1,050
-
465
pH
6.5
-
6.7
3.7
-
6.4
7.8
TDS*
-
71,827
82,690
265,770
24,945
54,989
57,440
Note sulphate
level
High Sulphate
Extreme Barium
and Cations
High TDS
Comments
1
Table 1
Examples of the range of
North Sea oil field
formation water chemistry
9
Water Handling
By contrast, chemical incompatibility between injected sea water and formation water
is the cause of the deposition of sulphate mineral scales. This occurs because sea water
contains reasonable concentrations of Sulphate anions (up to 2,800 ppm) but is low
in divalent cations {420 ppm Calcium (Ca++), trace Ba++ and Strontium (Sr++)}. By
contrast, many formation waters contain significant concentrations of barium - from
tens of parts per million in the Brent field formation waters to thousands in the Miller
field. The solubility of Barium Sulphate (BaSO4or Barite) is very low, being only 4%
of that of calcium carbonate. Barite is precipitated by the reaction:
Ba++ + SO"4 →
Ba SO4 ↓
BaSO4 is one of the most insoluble of the scaling minerals. A similar problem is
encountered with Strontium Sulphate (celestite or SrSO4) in some fields.
A less frequently encountered scale is Calcium Sulphate (CaSO4). This is due to the
unusual solubility behaviour of Gypsum (CaSO4. 2H2O), the most commonly encountered form of calcium sulphate. Gypsum has a solubility maximum at 40°C (i.e. it
shows reduced solubility at both higher and lower temperatures). The issue is
complicated by the fact that the equilibrium form above 40ºC is Anhydrite (Ca SO4);
which is even less soluble. A further complication is that this transition temperature
is itself dependent on the salinity. Further information on this complicated system can
be found in chemical textbooks.
The key facts summarising formation of the above and other scales, are summarised
in Table 2.
Typical Causes and Types of Inorganic Scales Encountered in Oilfield Operations
Table 2
Typical causes and types of
scales encountered in
oilfield operations
Scale
Formula
Principal Cause of Scale Formation
Calcium Carbonate (Calcite)
Ca CO3
Reduced partial pressure of CO2
due to pressure reduction
Mixing of formation water
and Sea Water
Barium Sulphate (barite)
Ba SO4
Strontium Sulphate (celestite)
Sr SO4
Calcium Sulphate (gypsum)
Ca SO4.2H2O
Temperature change - maximum solubility at 40ºC
Ferrous Sulphide
FeS
Corrosion followed by contact with Hydrogen Sulphide
Salt
NaCl
Temperature and pressure reduction
{some (liquid) water also vaporises into gas phase}.
Gas wells only
Sulphur
S
Temperature and pressure reduction.
Sour gas wells only
Department of Petroleum Engineering, Heriot-Watt University
5
1
1.2 Scale Inhibiton
The exact location at which scale will actually form is a function of many factors temperature and pressure changes, mixing patterns, water chemistry, precipitation
kinetics etc. Scale formation is observed in the reservoir formation, the perforations,
producing well tubing, pumps and topside facilities. Formation damage (permeability
reduction) in the near wellbore area of production wells has also been observed.
An effective scale management programme consists of:
(i)
Predicting whether scale formation is likely to occur in the lifetime of the
producing asset based on the Formation and Injection Water Chemistry and
predicted producing conditions in the reservoir, well and the facilities
(ii) Identification of when significant changes in producing conditions have
occurred that could initiate scale formation e.g. sea water breakthrough into a
producing well
(iii) Instituting a scale inhibitor injection programme.
A scale inhibitor is a chemical e.g. a polyphosphonate, which has the ability to
prevent the (tiny) barium sulphate seed crystals from growing large enough that
they can form bulk precipitates; i.e. scale inhibitors reduce the speed (kinetics)
of solid scale formation, but do not alter the final equilibrium (thermodynamics).
They achieve this by absorbing on the active sites of the growing barium
sulphate crystal. This (temporarily) inhibits the accretion process which leads
to crystal enlargement. The (minute) seed crystals are thus unable to stick
together and are transported from the well with the produced fluids.
The scale inhibitor can be injected:
(a) Downhole: the inhibitor is injected (squeezed) into the formation where it is
absorbed onto the formation rock. It is then re-dissolved in the produced fluid
at low concentration over a period of many months. Scale inhibitors show a
“threshold” activity level i.e. if they are present above this level in the aqueous
phase they will inhibit scale formation. Thus a monitoring programme needs
to be instituted to measure the inhibitor level in the produced water. Providing
the inhibitor concentration remains above a threshold value, it will protect the
near wellbore formation and the production tubing from scale formation. A
further inhibitor treatment is required when the concentration in the produced
water becomes too low (approaches the threshold (or minimum) inhibitor
activity level.
(b) Inhibitor may be continuously injected into the tubing at a low rate through
an injection valve situated near the bottom of the well. The inhibitor is carried
to the injection valve via the casing / tubing annulus. This, obviously, does not
provide protection to the (near wellbore) formation, perforations or to the
casing / tubing below the injection point.
(c) At the wellhead to provide further protection to the facilities.
6
9
Water Handling
1.3 Corrosion
Corrosion of production facilities costs operators many millions of pounds every year.
Corroded downhole equipment - tubulars or valves etc. are difficult and expensive to
replace while corrosion of the well’s casing can threaten the integrity of the well itself.
Corrosion of surface facilities is in principle easier to repair - since access is simpler
- but can (potentially) cause the loss of production (income) from all wells tied into
that facility; as well as resulting in significant environmental damage if the operator
was not aware that he had a significant corrosion problem.
Iron bacteria deposit a sheath of iron hydroxide around them as they grow. This is
obtained from soluble iron ions in water. Once established, an iron bacteria colony
can precipitate large quantities of plugging iron hydroxide as well as provide a host
environment for other bacteria, e.g. sulphate reducers, that lead to localised pitting
corrosion.
1.3.1 Carbon Dioxide Corrosion
Carbon dioxide (CO2) dissolved in water can form a sufficiently strong acidic solution
to enable steel to dissolve in water if the CO2 partial pressure is greater than 30 psi.
CO2 + H2O → H2CO3 (carbonic acid)
H2 CO3 + Fe → FeCO3 + H2
CO2 corrosion is characterised by a uniform attack which gradually reduces the metal
thickness.
1.3.2 Oxygen Corrosion
Oxygen (O2) dissolved in water, as may occur in water injection projects or in process
equipment with operating pressures below one atmosphere, is highly corrosive:
4 Fe + 6H2O + 302 → 4 Fe(OH)3
Oxygen corrosion is more aggressive than CO2 corrosion since it is not uniform - it
leads to pits in the metal surface which can rapidly grow to form holes which pierce
the metal itself.
1.3.3 Hydrogen Sulphide
Hydrogen Sulphide (H2S) partial pressures greater than 0.5 psi, can lead to both metal
loss corrosion and to hydrogen embrittlement:
H2S + Fe → FeS + 2H
•
2H → H2
•
Hydrogen embrittlement involves the steel loosing its ductability. Soft steels, which
are capable of relieving the internal stresses caused by atomic hydrogen penetrating
the (crystal) structure and subsequently forming molecular hydrogen, are less and
susceptible to this form of attack.
Any form of dissolved iron present in the water phase due to corrosion processes will
be precipitated by contact with hydrogen sulphide:
Department of Petroleum Engineering, Heriot-Watt University
7
1
Fe++ + H2S →
2H+ + FeS ↓
+++
2Fe + 3H2S → 6H+ + Fe2 S3 ↓
The iron sulphides have a very low solubility in water. These solids can form copious
plugging precipitates which create extensive formation damage.
2. PRODUCED WATER TREATMENT
The large changes in the volumes of produced water which are typically encountered
during the life of a production project are summarised in table 3. The facilities have
to be designed to operate efficiently over these wide range of conditions.
Phase
Fluid
Early
Mid-Life
Mature
Abandonment
Oil (bopd)
100k
100k
30k
10k
Gas (MM sft3 / d)
100
100
30
10
Water (bwpd)
<500
30k
100k
>100k
Water cut
<0.5%
30%
77%
>90%
2.1 Introduction
Gravity separators - the normal method for separating water from the co-produced oil
and gas phases - was discussed in section 6 of chapter 8 describing the "Fundamentals
of Oil and Gas Processing" The basic mechanism of the separator operation was
described. We also discussed that this primary separation may be enhanced by:
(i)
Heating of the crude oil to enhance water separation via viscosity reduction,
(ii) Addition of demulsification chemicals which alter the interfacial tension
between the oil droplets and the water; so allowing the emulsion to break and the
water droplets to settle. These emulsions can have a much greater (up to 100 times)
viscosity than that of either the water or oil phases. This partly accounts for their
stability.
(iii) Electrostatic separation may be used to further reduce the water content of
relatively dry oil. The water droplets suspended in the oil carry a small electrical
charge. Their rate of settling can be increased by imposing the appropriate electrical
field across (part) of the settling region inside the separator. The gravity separator
has now become an electrostatic separator - it is not widely used but is occasionally
employed in conjunction with the more difficult to separate, typically denser, crude
oils. Electrical short circuiting occurs when the water content becomes too high, say
above 10% volume.
Even with the above enhancements; the first stage gravity separator leaves typically
500 - 2000 ppm oil still present in the water. A further treatment / processing step(s)
are required before the water is fit for disposal e.g. the oil content has to be reduced
to the 40 ppm average that is required by legislation in N.W. Europe. The following
6 different process schemes (sections 9.2.2 to 9.2.7) that have been developed to
reduce this oil content of this oily-water.
8
Table 3
Typical changes in
production rates during the
lifetime of a 100,000 bopd
oil field
9
Water Handling
2.2 (Corrugated) Plate Interceptors
Plate interceptors work by reducing the distance required for a droplet to migrate
before it comes into contact with other oil droplets and coalesces. The interceptor pack
consists of a series of closely spaced, parallel plates mounted at an angle of 45º (Figure 1).
These plates have a corrugated profile (similar to the well known roofing material figure 2). This profile allows the oil to collect at the high point, where the individual
droplets coalesce and the resulting larger droplet moves upwards, under the influence
of gravity and against the direction of water flow.
Oil Outlet
Oil Weir
Oil Droplet
Water
Outlet
Weir
Oil Layer
Inlet Weir
Water
Outlet
Pipe
Figure 1
(Corrugated) Plate
Interceptor
(Corrugated)
Plate Interceptor
Large oil droplet formed by
coalescence of many small
droplets on top surface of
corrugated plate
Oily Water
Inlet Pipe
Small oil droplets rising
under gravity
Figure 2
Oil droplets coalescing in a
corrugated plate
The corrugated plate interceptor works off because of the short separation distance
between the plates. The ability of the oil droplets to move upwards through the water
phase increases rapidly (with the square of the oil particle diameter) as the droplet size
increases. Hence the overall rate of oil-water separation will be increased if the oil
droplets can be encouraged to grow larger in a shorter time. The oil layer collected
at the surface, is skimmed off by an oil outlet weir.
The configuration of the water inlet and outlet pipes is so arranged that any turbulence
that might lead to re-mixing of the small oil droplets is minimised. Removal of oil
droplets as small as 60 µm is frequently achieved. These plate separation packs may
also be installed inside gravity separators so as to enhance their performance.
Department of Petroleum Engineering, Heriot-Watt University
9
1
2.3 Flocculation / Coagulation
The separation of suspended liquids (oil) and solid material can be enhanced by
artificially increasing their size and their ability to coalesce. This is done by chemical
flocculation. This process uses a chemical (such as Ferrous Sulphate) which forms a
voluminous precipitate in contact with water. This precipitate has the ability to
coagulate into large flocs and, in the process, destabilise any suspended particles at the
same time. A typical flocculation unit is shown in Figure 3 - it consists of a mixing
chamber and a floc growth chamber. This growth chamber gives the microflocs,
formed in the mixing chamber, the opportunity to collide, grow and form large flocs.
This process can be speeded up by the addition of floc growth catalyst. These are
proprietary chemicals available from service companies.
Flocculant
(e.g. Fe SO4
Water
Inlet
Floc
Growth
Catalyst
Paddle Mixer
Floc Formation
Paddle Mixer
Outlet
Floc Growth
The flocs are subsequently removed from the water stream by settling / sedimentation
(gravity separation) or by flotation (see section 9.2.4). Organic polymers with
polyectrolyte properties work efficiently (only 1-10 ppm addition required) in saline
waters. These polymers can be cationic, anionic or amphoteric - the optimum choice
depending on the water salinity and the nature of the solid particles. This type of unit
is mainly used for cleaning water prior to re-injection.
2.4 Flotation
Gas flotation uses accelerated separation via the buoyancy (density reduction) effect
obtained from rising gas bubbles attaching themselves to the oil droplets. Once they
have been carried to the surface they can coalesce, followed by separation of the oil
using a weir. The gas can be added to the oily water in two manners:
(i) Injected into the water and dispersed by a rapidly rotating impeller (dispersed
gas flotation - see Figure 4). It is normal practice to arrange a number (e.g. four) of
these units in series. A typical design criteria is that they should be capable of
removing oil droplets larger than 15 µm.
(ii) Dissolved in the water under high pressure. Oily water is mixed with the gas or
air in the pump suction. It travels via a pressure vessel where a limited residence time
10
Figure 3
Flocculation unit
9
Water Handling
allows the gas to dissolve. When the pressure is rapidly reduced - by passage of the
water through a throttling valve - the gas comes out of solution in the form of many
small bubbles (the champagne bottle effect) - see Figure 5.
Gas Injection
Oil Froth
Perforated Hood
Oil Outlet
Gas / Oil
Bubbles
Rapid Rotation of Paddle
Produces Small Gas Bubble
Figure 4
Dispered gas flotation
Clean
Water Outlet
Oily Water Inlet
In both cases the oil froth is removed at the top and denser sediment / solids from the
bottom. The water is re-cycled a number of times to improve the unit’s effectiveness,
with cleaned water being drawn from the bottom of the flotation unit.
Gas Vent
Oil Froth
Clean Water
Water Recycle
Oil
Gas Bubbles
Pressure
Reducing
Valve
Gas or Air
Oily Water
Figure 5
Dissolved gas flotation unit
Sediment
Solids
Removal
Gas or Air
PRESSURE VESSEL
(provides residence time
for gas adsorption)
PUMP
2.5 Hydrocyclones
Hydrocyclones have become the standard device for cleaning oily water. Developed
in the early 1990s, they work by using centrifugal force to increase the effect of
gravity. They have proven to be compact, efficient, and - with the increasing use of
Floating Production and Storage Vessels to develop marginal oil fields - operate
independently of platform motion.
Department of Petroleum Engineering, Heriot-Watt University
11
1
A schematic diagram of a hydrocyclone is included as Figure 6. This long, thin device
has an upper inlet, where oil-water feed flows tangentially into the upper, swirl
section. The oil- water swirls around in a circular path, and is then accelerated as the
flow enters the taper section with it’s reducing diameter. This swirling action leads
to an increase in the effective gravitational component, in the basic separation
equation. The lighter hydrocarbon phase migrates to the central core which flows
upwards and exits the device via the upper outlet. The (cleaned) water phase passes
into the tail section equipped with a back pressure valve to control the pressure drop
across the device, and then exits the hydrocyclone.
Hydrocarbon (Lightphase) Outlet
Oily-Water Inlet
SWIRL SECTION
(with circular flow)
Reverse Flowing
Central Core
TAPER SECTION
Lighter Phase
Migrates to Central Core
TAIL SECTION
Back Pressure
Device
Figure 6
Schematic of a
hydrocyclone
Water Outlet (Underflow)
12
9
Water Handling
Hydrocyclone performance is sensitive to:
(i)
Feed rate - excessively high flow rates reduces performance by causing turbulence.
(ii) Pressure drop - this is controlled by the back pressure device mounted in the
tail section. Proper adjustment of the pressure drop maintains the efficiency of the
device’s performance. This results in a large turn down ratio i.e. the hydrocyclone
will separate oil from water efficiently over a wide range of flow rates.
(iii) High oil content in the oily-water feed - this reduces the separation efficiency.
In practice, a number of cyclones are mounted in parallel in a pressure vessel with
common inlet and outlets that are suitable for modular construction.
A typical modern flow scheme for produced water is shown in Figure 7 - the oily-water
phases recovered from the first and second stage separators are treated in separate
banks of de-oiling Hydrocyclones. The addition of a de-oiling chemical to the
hydrocyclone feed is optional and based on the actual producing circumstances. The
oil recovered from the Hydrocyclones is passed to the low pressure separator while
cleaned water is ready for disposal after a further de-gassing stage.
It has become the “equipment-of-choice” replacing gas flotation for cleaning
produced water. This is due to its compact size, efficient performance and it’s
operational robustness in the presence of solids.
Gas
Gas
(Optional)
Demulsifier
Chemical
From Wells
1st Stage
Gravity Separator
Oil plus
some Water
Oily Water
2nd Stage
Gravity Separator
Export
Crude Oil
Oily Water
De-oiling
Hydrocyclones
(Optional)
De-oiling
Chemical
Oil
De-oiling
Hydrocyclones
(Optional)
De-oiling
Chemical
Oil
Gas
Figure 7
Modern scheme for clean
Produced Water
Oil Recovery
Produced Water
De-gasser
Disposal Water
Disposal Water
2.6 Coalescer Units
There are a number of designs of coalescer units. The basic concept is to provide a
(usually oleophilic) surface on which the small droplets of oil can collect, grow and
eventually break free and be removed for subsequent separation. They are capable of
producing the lowest oil concentrations (5 ppm oil in water has been achieved in ideal
Department of Petroleum Engineering, Heriot-Watt University
13
1
circumstances). However, they are very susceptible to plugging by organic (wax,
asphaltenes etc.) or inorganic solids (scale, formation material, corrosion products
etc.). They can be (partially) regenerated by backwashing - though disposal of the
backwash water may present a problem.
One design is to pack the coalescer vessel with granular, fibrous material. The (small)
input oil droplets coalesce on the granular fibres. The resulting, larger oil droplets are
separated downstream of the coalescer. A second type of coalescer is where the oily
- water feed is passed down a filter tube. The properties of the semi-porous material
are adjusted so that water can pass through it, concentrating the remaining oil /
emulsion phases which can then be more efficiently processed by other means, such
as gravity separators.
2.7 Centrifuges
Hydrocyclones are typically capable of removing 50% of the 10 µm oil droplets rising
to 100% of the 20 µm ones. The principle of enhanced gravitational force employed
by Hydrocyclones can be further extended by use of centrifuges where an external
electric motor is used to spin the fluid at high velocity together with a suitably designed
internals to promote oil/water separation. Experience has showed that centrifuges
have the following advantages:
(i)
more effective than Hydrocyclones (complete removal of all oil droplets larger
than 5 µm is claimed),
(ii) their operation is not dependent on the available fluid pressure drop,
(iii) the above has to be weighed against their increased space, weight, power and
maintenance requirements.
3. OTHER SOURCES OF (PLATFORM) WATER REQUIRING
TREATMENT
Produced water makes up the bulk of the water that has to be processed for disposal.
Particularly in offshore operations, there are other sources of water that require
treatment before disposal. These include:
(i)
Water used for washing / cleaning of equipment,
(ii) Sea spray and rain water,
(iii) Utility water previously used for heating and cooling duty,
(iv) Displacement water from crude oil storage systems and shuttle tankers. (The
storage cells are kept full of water which is displaced by the produced oil. The
process is reversed when the oil is pumped out for onward transport.)
Care needs to be taken with respect to compatibility problems when mixing waters of
different chemical compositions (see section 1.2 on scaling). Scale inhibitors may be
required if severe problems are encountered.
14
9
Water Handling
4. DISPOSAL OF PRODUCED WATER
4.1 Marine Discharge
Historically, the vast majority of offshore produced water and platform water has
been disposed of to the marine environment after reducing the oil content to at least
that specified by the national legislation e.g. (40 ppm oil in water in N.W. Europe with
a voluntary company limit of 30 ppm agreed for the UK Continental Shelf for 1999).
Onshore, disposal has been by underground injection, either returned to the producing
zone as part of a water flood / pressure maintainance scheme or injected into a
dedicated disposal well or interval. Although the contribution of offshore operations
to the amount of oil reaching the North Sea is low (table 4); the amount contributed
by produced water is rising since the maturity (increasing number of years in
production and average water cut) is continually increasing. E.g. In 1985, produced
water contributed about 10% of the oil entering the North Sea due to Drilling and
Production Operations. This had increased to more than 50% by 1995. (see chapter
10.4 - Environmental Impact of production Operations).
Source
Table 4
Oil in the marine
environment
Input
(million tonnes per annum)
Marine Transportation
1.13 - 2.13
Offshore Production
0.08 - 0.2
Coastal Oil Refineries
0.2 - 0.3
Industrial Waste
0.3 - 1.98
Municipal Waste
0.3 - 0.45
River Runoff
0.3
Natural Seeps
1.6
Atmospheric Rainout
0.6 - 9.0
Total Pollutants
2.06 - 4.91
This rise in the importance of produced water as a source of oil emission to the
environment has lead to the desire to implement produced water re-injection in
offshore operations. This is an active topic of research and development at the present.
The water will mainly be as a (partial) replacement for the large volumes of sea water
injected for the purpose of maintenance and improved reservoir sweep (see section 5.3).
4.1.1 Production Chemicals
Many production chemicals that are injected into the produced fluids, both downhole
and at the surface into the production facilities. Most of these chemicals are
surfactants and are partially or fully soluble in the water phase. This implies that they
(or their residues) may also require final disposal with the produced water. There are
strict rules and regulations with respect to environmental acceptability testing of the
(fresh) production chemicals - but there are currently no standards with respect to the
testing of the produced water. In fact, while figures for the tonnage of chemicals used
are available (table 5), there are no generally accepted accounting standards to
measure the fraction of this figure that is disposed of to the environment.
Department of Petroleum Engineering, Heriot-Watt University
15
1
4.2 Reducing Environmental Impact of Produced Water Discharge
The environmental impact of produced water can be minimised by:
(i)
Replacing the use of fresh injection water (see section 5) by produced water
(ii) Reduce the oil content of produced water. The relatively recent development
of the hydrocyclone has contributed to increasing the ability of the operators to
meet the discharge regulations. Further reductions of this figure (20 ppm oilin-water monthly average with a maximum of 100 ppm) are currently under
discussion
(iii) Reduce the volume of produced Water which has to be disposed of
Production Chemical
Usage (in Tonnes)
Norway 1988
UK in 1987
Biocides
1372
2316
Corrosion Inhibitors
1062
1100
Oxygen Scavengers
997
4348
Scale Inhibitors
766
4025
Emulsion Breakers
174
2132
Oil Removers
8
293
Defoamers
128
34
Flocculating Agents
166
108
Surfactants
0
53
Cleaning Agents
13
552
Gas Treatments - Glycol, Methanol etc.
2234
9237
C.M. Hudgins “Chemical Use in North Sea Oil and Gas E & P” J. Petroleum Technology, SPE, Jan 1994
The major International Oil companies are currently producing, treating and disposing
of a similar volume of produced water as their oil production. More mature
companies, with mainly USA production, may have average water cuts of 75% or
more. Total produced water costs are typically US $ 0.20 - 0.75 / bbl for lifting,
treating and disposal. This is a very significant economic incentive for reducing its
volume. This is being achieved in a number of different ways:
(i)
Water shut off in the well (cement squeezes, scab liners, opening / closing
downhole flow control valves etc).
(ii) Water shut off in the near wellbore formation or at greater depth from the well
by injecting chemical treatments (e.g. total blocking of the pores with an
impermeable polymer gel).
Both these methods require identification of the completion zone that is producing
excessive amounts of water. Further technologies that can be employed are:
16
Table 5
Chemicals in the
environment
9
Water Handling
(iii) Re-design of the wells so that they produce at a greater oil / water ratio e.g. use
of horizontal wells which can be placed at the maximum possible distance
above the oil / water contact while the resulting reduced drawdown achieved
by the long exposure to the reservoir minimises water coning effects.
(iv) Downhole water separation of the bulk of the “free” water using hydrocyclone
technology followed by injection in a separate lateral (Figure 8). The disposal
zone may be above or below the production zone. One or two electric
submersible pumps may be used too depending on the depths of the production
and disposal zones, the formation pressure, and permeability’s as well as the
flow rates. Over 100 installations of this type had been installed by 1999.
Casing
Tubing
Oil Rich Stream
(75% Water)
Produced Oil (4%) and
Water(96%)
Producing Zone
Separated
"Free" Water
Downhole Hydrocyclone Separator and One or Two
Submersible Pumps Run from a Single Motor
Disposal Zone
(Open Hole)
Figure 8
Downhole Separation /
Injection
5. WATER INJECTION
Disposal of produced water was discussed in the previous section. However, large
volumes of water are also injected into the sub-surface for pressure maintenance and
improved oil recovery. North Sea light oil reservoirs, which might recover 20 - 30%
of the original oil in place under primary (pressure depletion) can have their recovery
doubled to 40 - 50% by use of water injection. The processes involved are:
(i)
Sweep of mobile oil from injection well to production well.
(ii) No loss of oil by maintaining formation pressure above the bubble point.
(iii) Improved cash flow / profitability by increasing early oil production and
reducing the project lifetime.
(iv) Reduced / no requirement for artificial lift.
Department of Petroleum Engineering, Heriot-Watt University
17
1
The reservoir engineering study carried out as part of the field development plan will
dictate where, how much, and at what rate the water should be injected. The task of
the Production Technologist is to write the injection water specifications so that it will
not be:
(i)
Corrosive to the injection well facilities and
(ii) Compatible with the injection formation so that the injectivity is maintained
over the injection project lifetime. Large volumes (30,000 bwpd) may be injected
into each well while the total field requirements can be up to 1,000,000 bwpd in the
case of the giant, Middle East oil fields.
5.1 Water Sources
There are often several potential water supply sources that are frequently used. The
water from the different sources will have different properties and will impose
different treatment costs to make it suitable for injection. When a source is chosen,
studies are required so that there is sufficient water available for the project lifetime.
Sources include:
(i)
Sea water
(ii) Fresh surface water
(iii) Aquifer water (not oil producing zone)
(iv) Produced water (from oil producing zone)
Surface waters will normally show seasonal variations in their quality. Hence
analyses need to be made over a complete 12-month cycle. Careful positioning of the
water inlet will help minimise water treatment costs, e.g. a seawater intake is best
sighted upstream of platform discharge points, e.g. of drill cuttings and drilling mud,
and at below the depth to which summer algal blooms occur; while being sufficiently
above the sea bed to be unaffected by bottom sediment stirred up by currents, storms etc.
5.2 Water Quality
The principal factors which define the water quality are:
(i)
solids - dissolved or suspended
(ii) dispersed oil
(iii) dissolved gasses
(iv) bacteria
Their likely effect and solution of some quality issues are is summarised in table 6. In
addition, some receiving formations contain clay particles which swell and disperse
in the presence of fresh water. This leads to blockage of pore throats and reduced
permeability. Hence, prior to the implementation of a water injection project, the
18
9
Water Handling
compatibility of the injection water and formation should be tested by pumping
injection water through a formation core sample and measuring any permeability changes.
Issue
Suspended solids
Suspended oil
Effect
Treatment
Plugging of Injection formation
Filtration
Plugging of Injection formation
Hydrocyclones /
(particularly in presence of solids)
Flotation / Filtration
Dissolved Gases
Corrosion of well and facilities.
Degasification
{O2 / CO2 / H2 S}
Plugging of formation by corrosion products
Corrosion inhibitor
Injection
Table 6
Quality issues in Water
Injection
Formation of Solids
Equipment and formation plugging
Scale inhibitor
{CaCO3 / Ba SO4 / CaSO4 / FeS}
by scale
Injection
Bacteria
Formation plugging by bacterial
Biocides
{Aerobic / Anaerobic (sulphate reducing)}
residues or corrosion products
Water incompatible with formation
Loss of permeability of injection formation
- Pretreat formation
(clay stabilisers)
- Alter injection water chemistry
5.3 Injection Water Treatment
Dump flooding is the only field proven case in which the injection water does not
require treatment. This is because the water is never produced to the surface (Figure 9).
Specific geological conditions required for dump flooding - a prolific, high pressure
aquifer has to under-or-over-lay the (lower pressure) oil producing zone. Injected
water volumes are unknown, unless a downhole flow meter is positioned between the
two intervals. The water injection rate can be increased by positioning an electric
submersible pump between water production and injection internals (powered dump
flooding).
Casing
Aquifer
Water
Supply
Prolific (High kh), High Pressure Aquifer
Seat for Plug to Isolate Oil Zone or for
Electric Submersible Pump
(for Powered Dump Flood)
Injection
Interval
Low Pressure
Oil Producing Zone
Figure 9
Dump water flooding
Department of Petroleum Engineering, Heriot-Watt University
19
1
Research projects are being conducted on “Raw (sea) water injection”. These are
examining the conditions in which untreated (sea) water could be injected. This would
be a valuable technology for the development of marginal oil fields using minimum
facilities and subsea wells.
5.3.1 Example Process Schemes Employing Coarse and Fine Filtration
Figure 10 is an example of a typical process scheme - it employs coarse and fine
filtration prior to de-oxygenation together with the addition of several inhibitors. The
process scheme operates as follows:
Stripping Gas Out
De-Oxygenating Tower
Chlorine Injection
Polyelectrolyte
Biocide
Scale Inhibitor
Coarse Filter
(Sea) Water
Lift Pump
Fine Filter
Corrosion Inhibitor
Injection
Pump
Back Flush Solids
Foam Inhibitor
To
Water Injection
Well
Intake With Fish Barrier
Stripping Gas Inlet
Oxygen Scavenger
The (sea) water lift pumps are protected from large solids (e.g. fish) by barrier filters.
Chlorine is injected into feed water in the suction of the lift pumps. It is then pumped,
after the addition of polyelectrolyte coagulant and scale inhibitor, to the coarse and
then to the fine filters. The filters retain the solids while the chlorine kills the bacteria
which could have lead to microbial growth on the filters. The solids removed from
the water will eventually block filters - they are removed by back-flushing (reversing
the flow of process water) so that their solids removal capacity is regenerated. The
filtered water, after the addition of a defoaming chemical if necessary, is pumped to
the top of the de-oxygenating tower, stripping gas is fed in at the bottom of the tower.
A packed or trayed tower (section 5.3.4 and figure 11) is used to bring the water and
gas into intimate contact to reduce the oxygen to the required level. The deoxygenated water is drawn off at the bottom of the tower. An alternative to gas
stripping is to reduce the pressure to a low value and run the tower as a vacuum dearator
(figure 11). The treatments do not completely reduce the oxygen to the very low levels
required - this is achieved by the addition of an oxygen scavenger.
The water is now ready for pumping into the injection wells after the addition of further
scale inhibitor and biocide to ensure the injection water sterility.
20
Figure 10
Injection process scheme
with fine filtration
9
Water Handling
High Vacuum
Vent Gas
Raw Water Inlet
Raw Water Inlet
Bubble Caps
Gas
Packing
Gas
Figure 11
Gas stripping to
deoxygenate water (left)
Vacuum dearator to
deoxygenate water (right)
Stripping Gas
Gas
Deoxygenated Water
Deoxygenated Water
5.3.2 Suspended Solids Removal
(i) Sedimentation - sometimes in land operations the water source may contain a
high concentration of coarse particles. These can be removed by allowing the water
to stay stagnant for a period of time and allowing the solids to settle. Stokes Law
describes the settling process; it is a function of the particle diameter and density hence in a continuous process the required residence time to clarify the liquid at any
depth can be calculated from the time required for particles originally at the top to
settle to this depth.
Hydrocyclones can be used to speed up the removal of these solid particles. The
Hydrocyclones principle of operation was described in section 9.2.5. For this duty,
the centrifugal force ensures that the denser solid particles are concentrated in the
underflow while the (cleaner) water phase exits at the top.
(ii) Coagulation / Flocculation - This was described in section 9.2.3.
5.3.3 Filter Systems
(i) Single and Dual Media Filters
These filters consist of a pressure vessel filled with a filter medium. The flow
direction may be either up or down. Single media filters are filled with homogenuously
mixed sand particles or with layers made up of differently sized particles - the sand
particles graded from coarser to finer size. This arrangement by which the coarser
particles are contacted prior to the finer particles is called “deep bed” filtration since the whole bed takes part in the filtration operation. Dual media filters, e.g.
garnet / anthracite, are built up of layers of the different media arranged so that the
least dense medium also has the largest particle size.
These filters are cleaned by backwashing - removing the solids by flowing in the
reverse direction at a high velocity. They can be fouled by oil since it will also be
Department of Petroleum Engineering, Heriot-Watt University
21
1
retained in the filter media. Washing with detergent may help with oil removal. Few
problems are encountered with the operation of these filters.
(ii) Diatomaceous Earth Filters
The advantage of Diatomaceous Earth (DE) filters is that they are lighter than single
or mixed media filters. DE is nearly pure silica material originally formed by a small,
single cell, marine organism. It is slurried with water and fed to the filter where it
forms a porous bed on a filter support (the “precoat”). The action of this effective
filter bed is enhanced by continually adding a small concentration of DE to the water
being filtered (“body feed”). This helps maintain the porosity and permeability of
the filter cake, resulting in a longer life between removal of the filter cake by
backwashing.
The size and weight of DE filters is counter-balanced by their tendency to be operator
and maintenance intensive, their susceptibility to plugging by oil (>10 ppm) and high
(>20 ppm) suspended solids concentration. Also there is a possibility of rupture of
the filter support; resulting in injection of DE (an excellent formation plugging
material) into the injection well. This is prevented by employing a “guard” cartridge
filter.
(iii) Cartridge Filters
Cartridge filter units consist of one or more cartridges (similar to a sock) mounted
over a perforated pipe support. The flow direction is from the outside (greatest
surface area, where the filtering action takes place) to the inside. A wide range of
particle sizes can be removed - the cartridge construction material can be chosen so
that it can remove particles from 100 to 2 µm. The cartridges have a relatively limited
solids removal capacity - they are most frequently used as a final (“polishing”)
filtration stage, where they act as a safety check. In this duty, they would only
normally deal with low solids concentrations e.g. intercepting solids introduced as
a result of filter backwashing operations.
These units are relatively light in weight compared to the alternative filter types
discussed earlier. They are often used for pilot test applications and for filtering
completion fluids during work-over operations.
5.3.4 De-Oxygenation
The presence of Oxygen in concentrations greater than 5 x 10-3 g/m3 (5 ppb) in water
flood operations can cause severe corrosion and plugging of the formation by
corrosion products.
(i) Gas Stripping
Removal of oxygen by gas stripping is based on lowering of the solubility of oxygen
in water by reducing the oxygen partial vapour pressure. Henry’s Law states that
gas’s solubility is proportional to the vapour pressure of the gas over water. Oxygen
from the water may be stripped by passing a (low oxygen content) stripping gas
through the water in co-current or counter-current flow.
Gas Stripping is normally performed in towers containing packing or perforated
trays. The water runs into the top of the tower and the stripping gas is fed in at the
22
9
Water Handling
bottom. It bubbles up through the water; the trays or packing provide good contact
between the water and the gas (Figure 11).
The primary requirement of the gas is that it be oxygen and hydrogen sulphide free.
Nitrogen natural gas or the exhaust gas from engines are commonly used. The
principle of removal is to reduce the concentration of oxygen in the gas coming in
with the water by dilution with the stripping gas. This reduces the partial pressure
of oxygen in the gas mixture and lowers the concentration of oxygen dissolved in the
water. N.B. Where oxygenated water must be artificially lifted to charge the water
supply pumps, the use of gas lift has been found to remove a large portion of the
oxygen.
The lowest residual oxygen values in the injection water require the use of oxygen
free stripping gas. Cryogenic separation of air can provide nitrogen with a very low
oxygen content - however this is often not practical offshore, particularly in floating
production facilities. This can be overcome by, for example, the Minox process
which the oxygen content of the air is removed in a catalytic converter where the
oxygen reacts with methanol. The nitrogen stripping gas is re-circulated. Once the
stripping plant has been charged with oxygen free nitrogen, the methanol based
catalytic oxidiser is only used to remove the oxygen stripped out from the treated
injection water.
(ii) Vacuum De-aeration
The principle of vacuum de-aeration is to reduce the partial pressure of oxygen by
boiling the water. At a temperature of 15ºC, water boils at a pressure of about 0.017
atm and the residual water oxygen content is reduced to 150 ppb.
(iii) Chemical Treatment With Oxygen Scavengers
Oxygen removal to the required 5 ppb level is rarely possible. Oxygen scavengers
are used to achieve this very low value. Oxygen scavengers remove oxygen from
water by chemical reaction. A large number of chemical compounds can be used for
this purpose. Selection of appropriate compounds should be based on cost,
compatibility of these compounds or their reaction products with other additives
used (bactericides, corrosion inhibitors, etc.) and ease of handling.
Sodium sulphite is the scavenger most frequently used in water flooding Its
scavenging effect is explained by the following chemical reaction:
2Na2SO3 + O2 → Na2SO4
Theoretically, an addition of 7.9 ppm sodium sulphite is required per 1 ppm oxygen
to be removed. Usually an additional 10-15 ppm is added to ensure that there is no
possibility of un-reacted oxygen being carried into a well. As the reaction proceeds
slowly at room temperature a catalyst, usually 1 ppm cobalt ion, is added. Other
bivalent metals such as copper, nickel, iron and manganese ions will also catalyse the
reaction. The sulphite reaction is also sensitive to pH - a minimum pH value of 6.0
being required.
Department of Petroleum Engineering, Heriot-Watt University
23
1
5.3.5 Hydrogen Sulphide Removal
As discussed in section 1.3.3, Hydrogen sulphide is highly corrosive as well as
representing a safety hazard. it can be removed in a similar manner to oxygen:
(i)
Mechanical removal by vacuum or counter current gas stripping,
(ii) Chemical removal by the addition of oxidising agents e.g. chlorine, potassium
permanganate etc.
5.3.6 Chemical Treatments
(i) Chlorination
Chlorination is a widely used, inexpensive, effective biocide. Chlorine hydrolyses to
form hydrochloric and hypochlorous acid with water:
+
+
Cl2 + H2O = H + Cl' + HOCl → 2H + Cl' + OCl'
The degree of ionisation is dependent on the pH and the higher the pH, the less
effective is a given quantity of chlorine (Table 7):
pH
6-8
8-9
9-10
Chlorine residual
required for 10 minute
bacterial "kill” (g/m3)
0.2
0.4
0.8
The chlorine residual or ‘free’ chlorine is the total sum of Cl2, HOCI and OCI'.
Chlorine, reacts with many materials; being a strong oxidising agent. The quantity
required is termed the ‘chlorine demand’. This includes the chlorine used up by:
(a)
(b)
(c)
(d)
oxidising ferrous to ferric ion
reacting with H2S to form H2SO4
reacting with some organic corrosion inhibitors and scale inhibitors
reacting with sulphite oxygen scavengers.
A residual chlorine of 0.4 g/m3 is recommended to ensure bacterial control (table 7).
(ii) Polyelectrolytes
Polyelectrolytes are large water soluble organic molecules made up of small building
blocks called monomers, repeated in a long chain. They usually incorporate in their
structures ion exchange sites which give the molecule an ionic charge. Those having
a positive charge are cationic, and those with a negative charge are anionic. These
molecules react with colloidal solids in the water by neutralising any electrical
charge or by bridging (tying together) individual particles to form a visible, insoluble
precipitate or floc. These chemicals therefore greatly enhance the efficiency of
filtration systems.
Ferric chloride is also used as a coagulant prior to the water filtration treatment step.
It's action is, however, very sensitive to pH.
24
Table 7
Minimum residual chlorine
content required for 10
minute bacterial "kill"
9
Water Handling
(iii) Oxygen Scavengers
Oxygen scavengers were discussed in section 5.3.4.
(iv) Corrosion Inhibitors
Corrosion will occur if oxygen is present in a treatment system. Treatment options are:
(i) Remove the oxygen mechanically or chemically,
(ii) Prevent oxygen ingress into the system,
(iii) Coat or line the system or use corrosion resistant materials,
(iv) Use an organic corrosion inhibitor which will form a protective film at the
metal/water interface, thus reducing the corrosion rate.
There are a large number of chemicals available. Inhibitor selection must be tailored
to each specific system.
(v) Scale Inhibitors
Most scale inhibitors in use are organic. Examples are:
(i) Organic phosphate esters. Not effective above 100ºC
(ii) Organic phosphonates. Effective up to about 175ºC
(iii) Organic polymers. High temperature application.
Their use was discussed in section 9.1.2
(vi) Biocides
Additional bacterial protection can be provided by injection of Biocides. Microbiology is a very complex subject. The introduction of sulphate reducing bacteria into
a reservoir may cause souring (generation of H2S) of the reservoir. The implications
of this can obviously be severe in terms of potential corrosive attack of the producing
wells and facilities.
(vii) Antifoam Agent
This is occasionally used to break the foam on deaerator tower trays.
(viii) Desulphation of Seawater
Effective inhibition of barium sulphate scale may not be practical for the very high
barium content formation waters (table 6) when using sea water as the injection water
source. This can be overcome by (partial) sea water desulphation by employing the
reverse osmosis process. In this process, (figure12) (Sea) Water is pumped through
a semi-permeable membrane under a high pressure. The membrane is chosen so that
it restricts the flow of sulphate ions - they are retained on the high pressure side at
a higher concentration than is found in sea water - while the water passed to the low
pressure side has a much reduced sulphate concentration. The water with the
enhanced sulphate concentration is returned to the sea as waste.
The costs of this desulphation process has to be balanced against the costs of
developing an alternative water source e.g. an aquifer.
Department of Petroleum Engineering, Heriot-Watt University
25
1
Sea Water
Feed
Semi-Permeable
Membrane
High
Pressure
High Sulphate
Fluid to Waste
26
Low
Pressure
Low Sulphate
Concentration Fluid
for Water Injection
Figure 12
Desulphation of seawater
by reverse osmosis
9
Water Handling
6. FURTHER READING
Arnold K. andStewart M.
"Surface Production Operations" Volume (2nd. Edition)
Published By Gulf Publishing company
ISBN 0-88415-821-7
Department of Petroleum Engineering, Heriot-Watt University
27
1
10
Oil
and
Gas Field Operations
Well
Control
CONTENTS
1.
2.
3.
4.
INTRODUCTION
MANAGEMENT OF UNCERTAINTY IN
PRODUCTION OPERATIONS
2.1.
Dissolved Gas Drive
2.2.
Gas Cap Drive
2.3.
Water Drive
2.4.
The “Real World”
2.5.
Overall Impact
2.6.
Resulting Production Constraints
MANAGEMENT AND OPTIMISATION OF
PRODUCTION OPERATIONS
3.1.
Introduction
3.2.
Production Target Setting
3.3.
Data Gathering
3.4.
Use of Model
3.5.
Production Planning / Scheduling and
Forecasting
THE ENVIRONMENTAL IMPACT OF
PRODUCTION OPERATIONS
4.1.
Overview of Discharges From The Oil
and Gas Industry
4.2.
Potential Environmental Impacts of
Production Operations
4.2.1.
Produced Water
4.2.2.
Other Aqueous Discharges
4.2.3.
Solids Disposal
4.2.4.
Atmospheric Emissions
4.2.5.
Accidental Spills and Flaring
4.2.6.
Other Forms of Environmental Impact
4.3.
A Waste Management Strategy
4.4.
Produced Water
4.4.1.
Regulatory Requirements
4.4.2.
Minimisation of Produced Water
Volumes
4.4.3.
Alternatives to Surface Disposal
4.5.
Atmospheric Emissions
2
1
LEARNING OBJECTIVES:
Having worked through this chapter the student will be able to:
• Discuss the management of Uncertainty in Production Operations
• Relate this uncertainty management to the various drive mechanisms and the extra
complexity of the “real world”
• Identify the activities required to manage and optimise production operations
• Discuss the need for a realistic well gathering system model complemented by upto-date, realistic field data
• Describe the use of the model for preparation of a production forecast and the
identification of profitable workover opportunities
• Identify the main elements of the environmental impact of production operations
• Discuss the elements of a waste management strategy
• Propose options to reduce the marine discharge of produced water
• Discuss the importance of atmospheric emissions
2
102
Oil and Gas Field Operations
1. INTRODUCTION
This chapter deals with three aspects of oil and gas field operations:
(i)
Management of the uncertainty involved in facility design and operations,
(ii) Management and optimisation of production,
(iii) The environmental impact of production operations.
2. MANAGEMENT OF UNCERTAINTY IN PRODUCTION OPERATIONS
It must be recognised that designing and operating an oil or gas field is very different
from designing and operating a refinery or chemical plant. Refineries and chemical
plants are much more complex, but the composition of the inlet flows are well defined
at the design stage. Also the quantities and qualities of products remain unchanged
during the lifetime of the project. In contrast, oil and gas production involves simpler
processing; but has to deal with the large uncertainty represented by our lack of a full
description of the initial state of the reservoir as well as our inability to predict how
it will behave once reservoir fluids are being produced. Alteration of the production
facilities is frequently very expensive due to their remote locations e.g. offshore. In
fact, addition of extra equipment may be impractical due to space or weight
limitations. Cost effective management of this factor is becoming more important in
the current, stringent economic climate of low oil prices. “Slim” platform construction techniques mounting minimum facilities are being inevitably favoured in the race
to profitably develop marginal prospects.
Typical examples of uncertainty that need to be taken into account within the Field
Development Plan and the facility design include:
(i) Wellhead conditions are subject to uncertainty until the formation is actually
drilled e.g. if the fluid in an isolated fault block was found to be gas rather than oil
bearing, the required wellhead completion could change from the planned 5000 psi
oil producer to a 10,000 psi gas producer. This has a large impact on the specification
for wellhead equipment, surface flow lines etc.
(ii) Field (potential) plateau production and reserves may be (much) larger or a
(small) fraction of the facility design values. It is normally a very profitable project
to de-bottleneck the oil production facilities to take advantage of any greater-thanexpected well / reservoir production potential.
(iii) The oil and gas production rates will decrease and the water production rate
increase as the field matures.
The rate at which these fluids change, and the long term production performance will
depend on the reservoir production mechanism e.g. the drive mechanism which forces
oil from a distant location in the reservoir into the well, and subsequently to the
surface. Typical examples are discussed below. Note that a homogeneous reservoir
is assumed.
Department of Petroleum Engineering, Heriot-Watt University
3
1
Tubing Head Pressure
2.1. Dissolved Gas Drive
Gas Bubbles Will
Eventually Form Gas Cap
Bubble Point
Well Dies
OIL
Time
Constant GOR
Reducing BHP
Gas - Oil Ratio
Reservoir Pressure
Pressure depletes to
bubble point and gas
cap formed
Rate of reduction in BHP slows
GOR reduces
N.B. Assumes Constant
Production Rate
Bubble Point
Depletion
Gas Expansion
Time
Time
During the initial, depletion stage of reservoir production, the producing gas-oil ratio
will remain constant so long as the field produces above the bubble point; the
reservoir pressure decreasing as fluid is removed from the reservoir. The Gas-Oil
Ratio (GOR) will then decrease once the bubble point pressure has been passed - the
gas will have started to separate from the oil prior to it reaching the production well.
It will eventually form a gas cap whose expansion will “cushion” or reduce the rate
of reservoir pressure decrease due to further fluid production. The producing Tubing
Head Pressure (THP) will reduce steadily until the bubble pressure is reached in the
reservoir. The THP will then reduce more rapidly as the GOR decreases since this
will increase the hydrostatic head of the fluid column within the well itself.
4
Figure 1
Production mechanism and
development of GOR and
Reservoir Pressure with
time for a dissolved gas
drive mechanism
102
Oil and Gas Field Operations
Tubing Head Pressure
10.2.2. Gas Cap Drive
Gas Breakthrough
Gas Cap
Well Dies
Time
OIL
GOR
decreases further
as gas comes out
of soloution
Time
Reservoir Pressure
Figure 2
Production mechanism and
development of GOR and
Reservoir Pressure with
time for a gas cap drive
mechanism
Gas - Oil Ratio (GOR)
Reducing GOR
Limited change
to BHP
GOR increases
Reduced average
fluid density
Increased tubing friction
from extra gas overriding
the reduction in hydrostatic
head
Gas Expansion "cushions"
pressure depletion
Gas cap reaches
top perforation
Time
N.B. Assumes Constant
Production Rate
This is similar to the latter stages of the dissolved gas drive. The decrease in reservoir
pressure due to fluid production from the reservoir is “cushioned” by expansion of
the gas cap. The gas-oil ratio continuously reduces as the reservoir pressure
decreases due to well production. This extra, liberated gas will further decrease the
rate in reservoir pressure reduction, but the (surface) gas-oil ratio of the produced
fluid will decrease, resulting in greater pressure drops across the production tubing.
Eventually, the gas will expand to such an extent that the gas-oil contact will reach
the top perforation, and a dramatic increase in gas-oil ratio will occur.
The THP will show a continuous decrease, driven by the reducing GOR and reservoir
pressure. It will initially increase on gas cap breakthrough, due to the increased GOR
reducing the average hydrostatic head. However, it will then decrease, as the ever
increasing gas rate leads to a more rapidly increasing frictional pressure drop in the
tubing than the corresponding reduction in the fluid column’s hydrostatic head.
Department of Petroleum Engineering, Heriot-Watt University
5
1
OIL
S tr
o ng W
aterdrive
We
Water Breakthrough
ak Wat
erd
riv
e
Tubing Head Pressure
2.3. Water Drive
Well Dies
Aquifer response begins
strong waterdrive
Water Cut
Reservoir Pressure
Time
weak waterdrive
water breakthrough at
bottom perforation
time
time
N.B. Assumes Constant
Production Rate
With this production mechanism, the rate of the reservoir pressure reduction due to
well production is reduced due to expansion of a (large) aquifer that is in good
pressure communication with the oil reservoir. This expansion can only begin to
happen once the reservoir pressure has reduced, e.g. there is a lag between initiation
of production and the pressure support from the aquifer becoming apparent. This
aquifer support manifests itself as a reduction in the rate of pressure drop with
continuing well production. Strong aquifer support (large aquifer with good
connections to the oil reservoir) will lead to limited further pressure decline, while
a weak aquifer will have proportionally less effect. The rising oil/water contact will
eventually reach the bottom perforation and the water cut will rapidly increase. The
tubing head pressure shows a similar story - an initial drop prior to initiation of
aquifer support, followed by a period when the rate of pressure reduction depends on
the strength of the water drive, i.e. the THP is controlled by the changes in the
reservoir pressure. Once water break-through occurs, the fluid column density (and
hydrostatic head) increases rapidly, and the well ceases production shortly thereafter.
These (simplified) examples illustrate how the reservoir drive mechanism affects the
reservoir wellhead and pressures, and the composition of the gas/oil mixture which
is to be processed in the surface facilities. It further defines the need for artificial lift,
e.g. if gas lift is to be employed, then gas compression facilities will be required
unless a separate source of high pressure gas is available. The use of electric
submersible pumps will increase electrical generation requirements.
6
Figure 3
Production mechanism and
development of Water Cut
and Reservior Pressure
with time for a water drive
mechanism
102
Oil and Gas Field Operations
2.4. The “Real World”
This is much more complex than the simplified examples discussed above.
(i) The reservoir will be heterogeneous - leading to rapid reserve from high
permeability streaks, followed by early water or gas break-through.
(ii) The reservoir may be compartmentalised by sealing faults which reduce the
reserves connected to the well. This would lead to an (unexpected) rapid drop in
bottom hole pressure. Possible early installation of artificial lift or the need to sidetrack the well to a new location will then be required.
(iii) A strong natural water drive may have been assumed. Provision has to be made
e.g. space for water injection plant and spare conductors in the case of an offshore
platform so that water injection can be implemented before the project economics are
put in jeopardy.
These factors all point to the need to develop a systematic approach to the collection
and analysis of production data, so that changes in the pattern of well production can
be recognised, and remedial measures initiated. This is discussed in greater detail in
section 3. The carrying out of special tests, e.g. well interference and build up tests,
cased hole production logs etc. in addition to the routine work, will add to the
understanding of the reservoir performance.
2.5. Overall Impact
The combination of the above - and many more - factors result in large changes in the
rate at which oil, gas and water are produced during the project’s lifetime. A typical
example of these changes during the early, mid-life and mature phases of an oil and
gas field’s production history is summarised in tables 1 & 2 respectively:
Typical changes in Production Rates during the
Lifetime of a 100,000 bopd oil field
Phase
Fluid
Table 1
Typical changes in
Production Rates during
the Lifetime of a
100,000 bopd oil field
Oil (bopd)
Early
Mid-Life
Mature
Abandonment
100k
100k
30k
10k
3
Gas (MM sft /d)
100
100
30
10
Water (bwpd)
<500
30k
100K
>100k
Water cut
<0.5%
30%
77%
>90%
Department of Petroleum Engineering, Heriot-Watt University
7
1
Typical Changes in Produced Gas Volumes From a
‘Typical’ Southern North Sea Gas Field
Phase
Fluid
Early
Mid-Life
Mature
Abandonment
Gas (MM sft3/d)
500
500
50
20
Condensate (bpd) 5000
5000
500
200
Water (bwpd)
5000
500
200
250
Table 2
Typical Changes in
Produced Gas Volumes
From a ‘Typical’ Southern
North Sea Gas Field
2.6. Resulting Production Constraints
First Oil
Plateau
De
ild-
Appraisal
Well
Up
Discovery
Well
cli
Bu
Oil Production Rate
The production phases are described in table 4. The production philosophy for oil
fields is often to build up to plateau production as quickly as possible after "first oil"
and to produce at this maximum (plateau) rate for as long as possible, giving the
production and operational cost profile illustrated in figure 4. Wells are often predrilled prior to installation of the platform to maximise the early production. Workover
of existing wells sidetracking of wells from depleted zones and infill drilling are all
used to ensure that the decline rate is minimised. In the final phase of field life the
operating costs are lowered as much as possible to delay field abandonment. This
occurs when the net revenue is less than the operating expenses. Thus operational cost
reduction is the main driver in late field life.
Cost
Time
Platform&
Facilities
Abandonment
Economic Limit
Operations
Pipeline
Drilling
8
ne
Figure 4
An oil production scenario
102
Oil and Gas Field Operations
Production
Phase
Constraint
Discovery/
• Exploration successful. Minimise time to "First Oil"
Appraisal
• Drill appraisal wells to provide further information as required
First Oil
• Time between discovery and first oil should be minimised.
• Fast track development reduces (discounted) exploration costs.
Build-Up
• Objective is to reach plateau production as quickly as possible.
• Well production potential limits (number of wells) production
• Build up phase can be reduced by:
- Pre-drilling wells,
- Using more than one drilling rig,
- Drill most prolific well first.
Plateau
• Production reaches maximum and remains constant
• One factor in the well/facility/pipeline system will be constraining production
• Excess well deliverability often available in early plateau period
• Consider de-bottlenecking facility or increasing pipeline capacity e.g. by use
of drag reducer to raise plateau production rate
• Drilling extra wells extends plateau period
• Well draw down increased to maintain oil production as water cut increases in
later plateau phase
• Eventually oil production will be constrained by, for example, the facility water
handling capacity e.g. Maintain plateau by shutting in high water cut production wells
Decline
• Maximise oil production within production constraints
• Identify candidate wells for work-over/re-completion/stimulation/artificial lift to
produce remaining reserves
• Examine operating cost structure e.g. alliances, outsourcing etc. to maintain
income above expenses for as long as possible. This maximises reserves.
Table 4
The Production Phases
Abandonment
• Plug wells and remove platform / facilities / pipelines etc.
3. MANAGEMENT AND OPTIMISATION OF PRODUCTION
OPERATIONS
3.1. Introduction
Production Management is based on controlling all aspects (technical and organisational) of the production process with each well in a field being produced to plan. This
plan’s objective is to maximise return on investment by maximising oil production
within constraints. Constraints can occur at the reservoir, well, gathering network,
process and marketing levels. Incentive schemes will frequently reward the field
manager based on his ability to achieve the economic optimum production that
maximises the field profitability within this constraint framework.
Department of Petroleum Engineering, Heriot-Watt University
9
1
Reservoir and well conditions change throughout the life of a field. Production
management is a continuous process of surveillance and reaction to these changes.
Systematic comparison of forecast and actual production will allow predictions to be
improved, and production maximised. Production Optimisation aims to maximise
return of investment by:
(i) Maximising Economic Production Rates - the impact of temporary operational
difficulties can be minimised by re-routing wells, adjustment of producing conditions etc.
(ii) Minimising Production Downtime - this requires an effective surveillance
program - to rapidly identify well problems and programme a repair by providing an
operating baseline against which actual measured rates can be compared.
(iii) Identifying Production Restrictions - locate those factors limiting production.
e.g. compressor capacity, pipeline restrictions, process limits, well productivity etc.
The economic benefit resulting from capital expenditure to remove the restriction
can then be assessed.
(iv) Maintenance Planning - Field lift gas allocation, production rate setting etc.
need to be reset in order to maximise production when equipment has to be taken out
of service for maintenance.
(v) Planning for the future - It is common practice to operate the platform and/or
pipeline as an independent profit centre. Production forecasts can be made by
simulating expected future production conditions. The optimum time for the
installation of the need for artificial lift, additional compression capacity, drilling of
new wells or development of near field potential etc. can then be identified. The
availability of a realistic, up-to-date, production optimisation model allows preparation of ‘what-if’ contingency scenarios with minimal delay.
3.2. Production Target Setting
The conventional practice is to base future production forecasts on an extrapolation
of past history. This produces an achievable forecast based on the current operating
practices. It does not necessarily take into account the field’s (untapped) production
potential. This can be done with a (calibrated) production system mode. This
estimates the field’s production with optimum operating practices (lift gas allocation,
well choke settings, separator pressures etc.). The (current) operational efficiency can
then be estimated by comparing how close the actual “oil in the tank” is to this
production potential.
Operation at maximum well potential may not be practical for a number of reasons e.g.
process limitations, sand production, contract or quota restrictions, gathering system
or pipeline limitations etc. Proper production management calls for the effect of each
restriction on the overall production rates to be quantified. One can then check whether
retention of this restriction is (still) economically justified - as well as to identify the
cost (capital expenditure) of its removal.
The production management system consists of two parts:
10
102
Oil and Gas Field Operations
(i) Technical
This consists of four component parts;
(a) a reservoir fluid description (matched PVT model)
(b) a systems analysis models and software used to match the production perform
ance of each well
(c) a gathering system model (pipeline geometry, length, elevation, diameter,
roughness) used to perform production optimisation calculations
(d) a set of system constraints at a well, gathering system and process level. Typical
well constraints are water coning, sand production, reservoir depletion. Gathering
system constraints will include maximum velocities (corrosion/erosion
effects), maximum allowable export rate and pressure etc
(ii) Field Data
Accurate and up-to-date data from the field are required to provide realistic
predictions from the technical model. Well performance is continuously
changing as the reserves are recovered, reservoir pressure depletes and the
well's produced fluid ratio changes. This is shown in figure 5
Wellbore Flowing
Pressure
ure
ss
pre
IPR @ time-0
-1
R
IP
-2
R
IP
-3
R
IP
PR
se
Re
ir
rvo
-4
R
IP
5
RIP
6
RIP
Figure 5
Deterioration of IPR with
reservoir recovery
e
Tim
lo
A b s olute O p e n F
w
Prod
u
Rate cing
3.3. Data Gathering
The implementation of modern measurement and control systems at the wellhead
allow the well's production to be controlled, while the well production performance
and that of the facilities are continuously measured and recorded. The situation is
illustrated in figure 6, which illustrates the process for a remote oil field operated with
flowing wells in addition to many types of artificial lift (Beam pumping / gas lift /
electric submersible pump). The production performance of the well and the
efficiency of the artificial lift process (if any) are measured and transmitted; along
with data from the gathering station / well test facility, to the (central) Field Operations
Office (FOO). Monitoring of the (automated) Distributed Control System (DCS)
which controls the field’s operation also takes place at the FOO. Two-way communication has been installed, so personnel at the field operations office can intervene
when required and adjust producing conditions. Digital microwave communications
between the FOO and the headquarters office ensures that the data can be shared with
all the petroleum engineers monitoring the field.
Department of Petroleum Engineering, Heriot-Watt University
11
1
Main office
Digital Microwave
Subsciber Radio
UHF Radio
Remote Surveillance
Data Transfer
Field Overview
Well Test
Gas Lift well
RTU
Solar Power RTU
Field Operations Office
Rod Pump
ESP Well
RPC
Individual Well Optimisation
RTU
Key: RTU - Remote Transmitter Unit
DSC - Distributed Control System
DCS
Figure 6
An integrated well
management system for a
land field
The problem presented by a remote land location is similar to that found offshore. This
is illustrated by the development where a minimum facilities platform acts as the FOO
for a number of platform wells shown in figure 7 (A field) as well as associated subsea
developments (B, C and D Fields). Aberdeen, a distance of some 125km, is the
headquarters office.
KEY
To
S
hor
D Field
Pipelines
Bundles
Umbilicals
Tender
Assisted
Drilling
e E
xi s
tin
g
oil
ex
B Field
po
Minimum Process Facilities
Minimum Manning
rt l
ine
SSSV
Diverter
Pre-installed
Flowline Bundles
Satellite Fields
Producing to
Mother Platform
Platform
A Field
C
ld
Fie
Early production
by Pre-drilling
Satellite Wells
Key SSV - Subsea Safety Valve
12
Underwater Completions
for Satellite Fields
Sp
ur
ga to tie
sp
ipe into
line ex
to isting
sho
re
Figure 7
A subsea development
102
Oil and Gas Field Operations
For both the land and offshore cases, a production systems model is required to turn
the data recorded by the various measuring instruments into “information” which can
be used to control and optimise the field’s operation. The procedures used to acquire
this data, carry out quality control, and display the way it is presented into well test
and down-hole pressure survey data. This is used to maintain the model by carrying
a systematic review of modelled and measured performances. Success depends on the
availability of good quality data and trained personnel to analyse it.
The key to the process is the provision of quality data. Operational personnel are most
motivated to obtain this once they can see a use for the data - this is achieved in the
set up described above since the system model is made available to them to help them
with their duties in the FOO.
3.4. Use of Model
Once the field model has been set up and calibrated, the next step is to run an
optimisation to determine whether additional production can be obtained by reallocating lift gas or adjusting well chokes. Constraints must be carefully set to
correctly model system limitations. Obvious constraints are compressor and separation capacity. Well performance constraints may also be required to deal with well
and pipeline stability issues. For example, minimum lift gas allocation rates can be
used to ensure that the operator does not attempt to run wells at an unstable rate.
Reservoir level constraints also need to be included. e.g. if gas flaring is a production
constraint, then high GOR producers should be beaned back or even shut in. If
reservoir considerations require that this well continue producing at a certain minimum rate, it can be kept on production. However, this will incur a production penalty,
as lower GOR producers must be then choked back to enable the higher priority wells
to produce.
These calculations can be done automatically by setting up and calibrating the model
to the well test rates. An automatic, optimisation algorithm can be used by setting the
separator pressures and lift gas availability and starting the calculations. Results must
be carefully checked to determine if the resulting recommended well rates are
acceptable for the prevailing reservoir and production policy.
As mentioned above, each active constraint in the system will result in less
production. By comparing optimisation run results, the cost to production (and cash
flow, profitability etc.) of reservoir policies can be quantified.
Optimisation of a large field is a complex task. All aspects of building, calibrating and
maintaining the model must be properly organised if accurate results are to be
obtained and manpower requirements minimised. A first step is to systematically
organise the PVT and well completion data. This ensures that the current well
configuration information can be readily incorporated into the well models. An
effective well file system ensures accuracy and minimises the time required to build
a well model.
A large volume of routine production test data is generated during field operation.
Each individual well test carried out with the test seporator provides valuable
information on reservoir and well performance for use in history matching purposes.
Department of Petroleum Engineering, Heriot-Watt University
13
1
Such well tests are used in production allocation (allocation of the produced oil to the
variuous wells being produced); hence they directly affect reservoir material balance
calculations.
Each test must be carefully screened before being used for production allocation. Such
systems analysis models are a convenient way to identify poor tests and identify
changes in reservoir or well conditions by comparing the well model and the actual
test rates.
Flowing bottom-hole pressure surveys enable well models to be calibrated. The
quality of the field data can be evaluated by comparison with the well model results.
The reason for any significant difference should be investigated and the well model
updated, if appropriate. This ensures that the well model is always current, so well
problems can be quickly identified.
Platform Activities
platform
shutdown
sidetrack
new well
tie in satelite field
compression upgrade
Jan Feb Mar Apr
Individual Well Performance
Individual Well Performance
May
Process and Pipeline Model
ut
rC
te
Wa
Produc
tion
GOR
Flow Constraints +
Equipment Uptime
Time
Production
Production Forecast
Time
Field Production (or Cash) Forecast
14
Figure 8
Development of a
production forecast
102
Oil and Gas Field Operations
3.5. Production Planning / Scheduling and Forecasting
A Field production forecast is normally prepared every month. The production
conditions can be re-optimised as part of the process or earlier if significant changes
in well or facility operation occur. It is essential that the field optimisation model has
been updated to represent the current well status. This will ensure that the results are
valid and represent current, actual field performance. The interplay of the various
parameters is shown in figure 8. A field-wide production allocation should be done
to check for discrepancies in measurement and allocation. Differences should be
identified and eliminated as necessary.
With the model properly calibrated, an optimisation is run to determine the production
targets, operating pressures and lift gas allocation for the next month. These results
form the basis of the field operating plan. Additional optimisation runs can be made
to provide contingency operating conditions for planned maintenance or equipment
breakdowns (e.g. compressor trips). The production plan can also include a priority
list of wells to adjust in the case of e.g. lift gas shortage, water disposal limitations,
gas export limitations etc. The production plan also includes reservoir pressure and
injection targets as appropriate.
The final part of the planning process is a summary of well repair and work-over
opportunities. Problem wells are identified as part of the model calibration process.
Likely repair costs and post-work-over production rates need to be prepared in order
to provide data for allocating capital and scheduling work-overs. Analysis of the
competing production improvement opportunities is the basis of the medium term
well operations workover plan.
Pipelines and platforms are frequently treated as independent profit centres. Spare
capacity can be used to exploit company owned near field potential {extended reach
wells, satellite, subsea completion’s employing natural flow, down-hole artificial lift
or subsea (surface) pressure boosting}. Alternatively, this capacity may be used for
third party business - some platforms in the North Sea meet more than 50% of their
operating costs by processing / pumping production from other companies. Similary,
the Shell/BP central North Sea ETAP development was only economically possible
because both companies developed a cluster of fields as a single project. Here, a BP
platform based development (Marnoch) provides host to a series of Shell subsea fields
(Heron, Skua etc). Neither company could have developed their own fields as a "stand
alone" project.
4. THE ENVIRONMENTAL IMPACT OF PRODUCTION
OPERATIONS
Production Operations take place within a regulatory framework set by international
agreements or by the government of the nation where the oil or gas field is situated.
In addition to meeting these legal requirements, the production operation has to be
conducted in such a manner that they meet the expectations of the public at large, as
influenced by publicity appearing in the media and by overview by non-governmental
environmental organisations such as “Green Peace”, “Friends of the Earth” etc. This
combination of legal and social issues is summed up by the need for the operator to
conduct his operations in such a manner that he earns a “licence to operate”.
Department of Petroleum Engineering, Heriot-Watt University
15
1
This chapter summarises the various routine discharges to the environment that occur
during routine operations and unplanned emissions (oil spills etc.).
4.1. Overview of Discharges From The Oil and Gas Industry
The North Sea is relatively enclosed and bordered by highly populated and industrialised nations. In addition to effluent discharged or dumped directly into the sea;
several major rivers such as the Rhine, Thames, Elbe etc. flow into the sea carrying
a high contaminant load - making it one of the worst polluted seas.
Relatively speaking, the pollution from the North Sea oil and gas installations are
modest. Hydrocarbon discharges provide the greatest input, accounting by weight for
over 97% of material discharged into the marine environment (i.e. excluding atmospheric emissions). Hydrocarbons from the North Sea oil and gas installations account
for about 18% of the total, or 24% if inputs from shore based installations are included
(figure 9).
Inshore Oil (6*109.kg)
Shipping (7*109.kg)
Coastal Discharges (10*109.kg)
Waste/Dredge Dump (13*109.kg)
Atmospheric (13*109.kg)
Offshore Oil (18*109.kg)
River/Run-off/Sewage (50*109.kg)
Natural Seepage (260*109.kg)
Whilst the oil and gas industry makes a significant contribution to the total hydrocarbon input into the North Sea, the same is not true for other contaminants. Emissions
to the atmosphere are much greater - figures 10 & 11 were taken from a major
operators 1996 Annual Health Safety and Environmental Report on their Exploration
and Production Operations. These large values contrast with the some 3 *109kg tonnes
of oil that were discharged to the marine environment with produced water.
16
Figure 9
Estimated inputs of
Hydrocarbons to the North
Sea.
102
1996
1995
1993
1993
1994
1992
Figure 10
Mass hydrocarbons flared
during a major operators E
and P operations.
16.0
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0.0
1992
Hydrocarbons flared (109kg/gr)
Oil and Gas Field Operations
50.0
1996
0
1995
Figure 11
Total Carbon Dioxide
Emissions by a major
operators E and P
operations.
100.0
1994
Carbon Dioxide
emissions (109kg/gr)
150.0
4.2. Potential Environmental Impacts of Production Operations
The sources of potential Environmental Impact of Production Operations are summarised below:
Table 5
Sources of potential
environmental impact from
Exploration and
Production operations
Oil Spillage
Cooling Water
Production Water
contains oil
Noise
Cuttings Discharge
Thermal Discharge
Pipeline Discharges
Sewage
Atmospheric
Emissions
(flares/vents/turbine
exhausts etc.)
Physical
Disturbances
Ballast/Crude
storage displacement
water
Washing / cleaning /
& main drainage
water
The relative proportion that each of the sources contributes to oil discharge to the
marine environment is summarised in figure 12. Some discharges e.g. pipeline
discharges when very large volumes of treated water are discharged over a short
period of times e.g. after pressure testing the pipeline during initial commisioning.
Historically, by far the greatest input of hydrocarbons comes from the discharge of
drilling cutting; while the impact of oil in produced water is now dominant.
Department of Petroleum Engineering, Heriot-Watt University
17
1
94.3%
76.3%
1982
1987
4.2%
1.5%
3.2%
20.5%
39.6%
86.9%
1992
1997
1.8%
58.6%
Figure 12(a)
Changes in relative
quantities of discharges to
the sea from offshore oil
operations 1982-1997
13.1%
Cuttings
Produced Water
Spills
40
Oil content (ppm)
35
30
25
20
15
10
5
0
1990
18
1991
1992
1993
1994
1995
1996
1997
1998
1999
Figure 12(b)
Long term trend for oil
content in discharged
producted water for the UK
continental shelf.
102
Oil and Gas Field Operations
4.2.1. Produced Water
Produced water consists of a mixture of formation water (found with the oil) and
breakthrough injection sea water (used to push the oil through the rock strata). The
water discharged contains oil, the statutory limit being set at a monthly average of 40
mg/l, with no sample having a value greater than 100 mg/l. During 1998, the concept
of a lower (currently voluntary) company wide allowable average discharge limit of
30 mg/l was introduced within U.K. waters; rather than the above (higher) individual
field value. Different limits apply elsewhere e.g. 32 ppm maximum average oil
concentration in the Gulf of Mexico (USA).
As a well ages, there is greater (injection) water breakthrough resulting in greater
quantities of produced water being discharged. This trend, combined with the
increase in number of wells drilled, has led to a substantial rise in the quantity of
hydrocarbons released into the North Sea from produced water {Figure 12(a)}. The
quantity of oil discharged with production water is now dominant especially since the
discharge of oil containing cuttings has now been banned. The discharge of drill
cuttings from water based muds is still allowed; as is the discharge of cuttings from
muds based on organic esters, ethers etc. if suitable permits are obtained. These special
chemicals are allowed as a replacement for oil in some circumstances. Cuttings
discharge from these latter mud types have been recently disallowed in the North Sea
but are still allowed in other offshore drilling areas.
Large effects are being made to reduce the oil content in the discharged produced
water. Figure 12(b) records the progress made in the United Kingdom Continental
Shelf). Produced water re-injection, where the produced water, is substituted for
specially prepared injection water is also being pursued as a means to reduce the long
term environmental impact of production operations.
Production chemicals are also used on a large scale - e.g. more than 25,000 tonnes in
the UK sector during 1988. It is unknown how much of this is eventually discharged
with the production water since the (relative) oil/water/gas phase solubility of all the
chemical and their subsequent reaction products is unknown - but it is suspected that
20-30% is discharged via the produced water. The quantity of production chemicals
reaching the environment will be reduced as re-injection of produced water into a
formation requiring water injection or into a dedicated injection formation becomes
more common.
4.2.2. Other Aqueous Discharges
Produced water is not the only source of aqueous discharges. This can come from
many sources - process cooling water, excess treated (sea) water for injection, water
used for pipeline testing, drainage water etc. Some of these sources have a (relatively)
small volume while others e.g. water used for pipeline testing, (treated with at least
a bactericide and corrosion inhibitor) may have a large volume but only occur once
or twice in the life of the project. Others, such as cooling water (treated by filtration
and chlorination coupled with a discharge temperature 25° - 30° C above sea
temperature) have large continuous flows (up to 500,000 bwpd or 3,300 m3 / hr).
Other well treatment fluids are emitted as temporary, one-off events e.g. acidic brines
produced after a well stimulation with acid, may also be discharged with the produced
water.
Department of Petroleum Engineering, Heriot-Watt University
19
1
4.2.3. Solids Disposal
Two organic solid types - wax and ashphaltenes - may separate from the crude oil
during the production process. The point at which they form - in the well or the
facilities - and the quantities of solid, will depend on the properties of the crude oil and
the production conditions. They have to be removed in an appropriate manner along
with any separated inorganic solids e.g. clay fines, formation sand grains etc. The
amount deposited would build up in the separator if it were not regularly removed.
This would reduce residence time and oil/water separation efficiency, as well as block
flow lines.
Formation sand particles and "fines" are also produced from the wells along with the
oil and gas. As discussed in the Sand Control module , these are collected in the
separators et. and may, after suitable washing procedures to remove any adhering oil,
be discharged. Alternatively, they may be transported to a licensed disposal site.
A special class of solids are inorganic scales e.g. CaCO3, BaSO4, formed by pressure
/ temperature changes during the production process or the mixing of incompatible
waters. Carbonate scales are normally removed from tubing by acid but sulphate
scales requre removal by drilling them out. Care should be taken if dry drilling of scale
from tubing recovered from the well is practised. Barium and strontium scales often
contain contaminants giving them a low level of radioactivity (which can be a health
hazard if the dust, from the drilling out of dry samples of these scales, is deposited in
the lungs).
4.2.4. Atmospheric Emissions
Production stations and (offshore) platforms are often sited in remote locations.
However, the environmental concerns (acid rain, ozone, greenhouse gasses etc.) are
similar to other industrial activities. Significant amounts of electrical power (50 - 100
kW) may be generated for use at the production location if electrical supply from a
National Grid is not feasible or economic. Gas turbines are frequently employed.
Often, heat recovery is not employed since fuel gas is available at relatively low cost.
VOC emission (109kg/gr)
Large quantities of volatile organic compounds are emitted to the atmosphere during
Exploration and Production Operations. (See figure 13 for figures from Shell’s
Exploration and Production Operations). This emission of Volatile Organic Compounds (VOC) is due to deliberate venting (65%), un-burnt gas during flaring (25%)
and fugitives from tanker loading operations, valves, tank roof vents etc. (10%). They
are equivalent to:
3000
2500
2000
1500
1000
500
0
1992
20
1993
1994
1995
1996
Figure 13
Volatile Organic
Compound emitted by a
major E&P operator
102
Oil and Gas Field Operations
(i)
20% of the hydrocarbon mass flared or
(ii) 1000 times (three orders of magnitude) greater than that of the volume of oil
discharged with produced water or
(iii) 200 times the volume of oil lost due to oil spills.
Apart from carbon dioxide, flare gasses may contain nitrogen oxides, carbon monoxide, sulphur dioxide, particulate carbon as well as un-burnt hydrocarbons.
The large scale hydrocarbon losses e.g. venting, tanker loading operations etc. are
easiest to identify and can be controlled with appropriate technology. The larger
numbers of small scale emissions e.g. diffusion / leakage through valves, gaskets,
flexible hoses and connections, isolation of process equipment during maintenance
etc. are more difficult to manage. The following statistics put these emissions in
context - The offshore oil industry contributes less than 4% of the UK greenhouse gas
emissions, while for Norway the corresponding figure is 20%.
4.2.5. Accidental Spills and Flaring
Oil spills tend to have a high public profile. This is because they are immediately
visible and the damage they cause to wildlife such as birds is obvious and distressing
to witness. The amount of oil that enters the environment through accidental spillage
is small (figure 12). Oil also enters the sea from flaring due to incomplete hydrocarbons combustion. Studies have shown that as much as 30% of the hydrocarbons are
being flared remain un-burnt. A new generation of flare tips (burners) with designs
that give more complete combustion have been developed to address this problem.
4.2.6. Other Forms of Environmental Impact
The environment around offshore developments may also be disturbed by noise and
vibration and by thermal pollution. Thermal pollution results from the discharge of
cooling water and by local warming around risers and oil storage cells. The effect is
not necessarily deleterious, the localised warming encourages fish, although it may
also enhance growth of fouling organisms.
Physical disturbance of the seabed by oil and gas explorations is largely confined to
the area immediately around fixed installations and pipelines. Installations and their
exclusion zones cover around 0.1% of the area of the North Sea, accounting for twice
the area closed off by ship wrecks. By far the greatest disturbance of the seabed is
caused by fishing activity (54%).
4.3. A Waste Management Strategy
A Waste Management strategy in any business area, not just Exploration and
Production, will address:
(i)
REDUCTION
generate less waste through more efficient practices
(ii)
REUSE
reuse material in original form
Department of Petroleum Engineering, Heriot-Watt University
21
1
(iii)
RECYCLE
convert waste back to usable material
(iv)
RECOVER
extract material or energy from waste for other uses
(v)
DISPOSAL
dispose of final residue in most environmentally
benign manner possible.
Many companies are developing strategies that reduce the environmental load below
the legal requirement and are developing “zero discharge options”. In every case,
when developing a Waste Management Strategy, it is important to look at the total
impact of the complete system rather than concentrating on a particular pollutant e.g.
is it always sensible to gas strip hydrocarbons out of produced water if the net effect
is that the pollution load becomes air - rather than water-borne and CO2 emissions are
increased?
4.4. Produced Water
4.4.1. Regulatory Requirements
Dry oil (no water production) is produced when an oil field is initially brought into
production. However, the level of water production rapidly increases as the field ages;
with many wells only being abandoned when the water cut reaches 95% or even
higher. In 1995 the Major Oil Companies typically produced as much water as oil in
their world-wide operations. Their water production is expected to increase rapidly;
a doubling within 5 years not being unusual in their corporate projections. This water
production entails a high economic penalty - it comes with a typical lifting, treating
and disposal cost of some 25-75 US cents/bbl water.
The requirement for produced water discharges to have a (monthly average) maximum of 40 ppm oil was set by the Paris Commission (PARCOM) in 1978 and hasn’t
been significantly revised since then. The analytical technique specified essentially
measures the aliphatic hydrocarbons, ensuring that dispersed rather than total oil is
measured. This standard is not based on an underlying environmental consideration;
but is an equipment based standard chosen on the basis of “what is achievable” with
“best available technology” driven by a belief that “less oil is better”.
The regulatory value of 40 ppm, which has been in existence for nearly 20 years, has
been discussed at regular intervals during this period. A general reduction to say, 30
ppm, would increase the fraction (maybe 25-30%) of the North Sea fields that already
have difficulty in reaching the regulation criteria. A voluntary, company wide target
for an average discharge level of 30 ppm oil has now been accepted for UK continental
shelf waters.
Although the quantities of oil discharged into the surface water environment are
significant, they are a small proportion of the total oil produced, e.g. a figure of 10g
oil discharged per tonne of product for Shell’s Exploration and Production Operations. For comparison, these 3,000 tonnes are 25% of the amount of oil accidentally
lost by Shell during oil spills.
In Shell’s case, more than 50% of the produced water is re-injected into underground
reservoirs while the average concentration of oil in discharged production water was
28 ppm.
22
102
Oil and Gas Field Operations
4.4.2. Minimisation of Produced Water Volumes
It must be remembered that large volumes of water, several times that of the oil,
frequently have to be produced during a fields lifetime due to the physics of the
production process associated with flow in porous media. Advanced Well Design,
such as horizontal wells, minimise the draw-down and hence extend the production
period prior to water breakthrough. Such well designs ensure that the volume of
“extraneous” water due to coning, high permeability (or thief) zones showing early
water breakthrough etc. are minimised.
Scab
Liner
Stops
Water
Inflow
,,,
yyy
yyy
,,,
,,,
,,,
yyy
yyy
,,,yyy
,,,
yyy
Water Producing
Zone
Polymer Blocking
Gel
e.g. Maraseal TM
Pumped Into
Formation
Water
Producing
Zone
Cement
Squeezed
Into
Perforations
Cement
Plug
Figure 14
Water shut off options
Water Producing
Zone
Bridge Plug
Improved techniques for selective Down-hole Water Shut-Off of water producing
zones minimises this “extraneous” water production. Not only does this reduce the
volumes of water (and its entrained oil) discharged, but also reduces the (artificial)
lifting costs, potentially increases the well’s (oil) production capacity and extends the
well’s producing life above the “economic limit”. Figure 14 illustrates mechanical
shut-off options (e.g. scab liners, bridge plugs etc.) which place a mechanical
restriction across the perforation of the wellbore to stop the water flow have been
employed for many years. Their effective application requires identification of the
water producing zone. Similar knowledge is required when placing a cement plug in
the well bore to stop water production . An alternative is to perform a chemical
polymer (gel) shut off operations where the gel can be pumped into the formation.
Both of these can be carried out, rapidly, (relatively) cheaply and easily via coiled
tubing. These latter, chemical options are being intensively researched to develop
improved, more environmentally friendly, materials and better placement techniques.
Department of Petroleum Engineering, Heriot-Watt University
23
1
Development of reliable procedures to selectively reduce water production via
bullhead techniques using the so-called “relative permeability modifiers” would find
wide application.
Equipment for Down-hole Separation and Injection has been developed (figure 15).
The concept is based on hydrocyclone technology for the separation of “free water”
followed by down-hole injection into a disposal zone, typically with the aid of an
electric submersible pump. The oil and remaining water is lifted to surface for
conventional separation and treatment. Apart from reducing lift energy and water
separation and disposal costs; Down-hole Separation and Injection provides new
production options such as dynamic control of the oil-water contact. Practical
problems such as monitoring, control and maintenance need to be resolved prior to
widespread application.
A final option is Wellhead (rather than down-hole) Separation and Injection at a
subsea production manifold followed by re-injection.
Production to
Surface Reduced Watercut
(Factor 10 Less Water)
Concentrated "Emulsion" Pump
Upstream Full Flow Pump
± 975 m KB
Submersible Motor
Producing Interval
± 10 m
Hydro-Cyclone
Packers
Shale Streak
24
± 25 m
Figure 15
Downhole separation and
injection: Oil/water
102
Oil and Gas Field Operations
4.4.3. Alternatives to Surface Disposal
Although water treatment techniques have improved significantly, there are no
available methods for the complete removal of oil. Furthermore, with increasing
concern over the environmental effects of the dissolved, chemical components of
produced water; operators are being challenged to develop alternatives by their own
management (drivers: company reputation, “license to operate”), as well as by the
regulators,.
Gross Fluids
Produced Water Re-Injection has been trialed in several fields with variable success.
It has usually proved to be more difficult (and expensive) to retain high, long term well
injectivity with produced water than with sea or aquifer water. Well injectivity is
effected by produced water ‘quality’ (composition, oil and solid content, temperature), formation permeability, fracture gradient and down-hole scaling potential.
Further, corrosion and souring problems have been observed. However, it can
normally be successfully introduced once sufficient experience has been gained and
the influence of the above factors controlling the process has been understood.
1 2002
0 200
99 200
004
2003 2
Water
B
Water
A
Oil A
1998
Oil B
2004
2010
Field A and B
2016
9
1998 1
m
5-50 k
500m -1500m
Water Depth
Field B
Start up 2004
10 km
Figure 16
Development of a satellite
field using subsea
separation and injection
Separation and
Water Re-injection
Install in 2004
Field A
Start up 1998
Water Injection
Production
Seabed separation and reinjection is being developed as a means of increasing
capacity of Floating Production, Storage and Offtake (FPSO) vessels by reducing the
volume of produced water to be processed. This is illustrated in figure 16 where
production of a satellite field is accelerated through use of the FPSO facilities once the
production from the main (field A) field has peaked. The technology required
separation , wellheads and power distribution and pumping and control are all field
proven but require "repackaging" for seabed operation.
Produced Water should ideally be used to replace “fresh” injection water during
mature water floods e.g. BP’s Ula field currently meets 95% of its water injection
Department of Petroleum Engineering, Heriot-Watt University
25
1
requirements with Produced Water. However if it is being injected into a disposal
zone, the re-injection could result in an increase of energy consumption equivalent to
250 - 500 tonnes of carbon dioxide emissions for every tonne of oil prevented from
being discharged into the sea. The total impact on the environment must be considered
when considering such options.
It should be noted that surface discharge of produced water is not an option for onshore
operation where contamination of surface water by discharged oil is not acceptable.
Further, selection of disposal zones must also ensure that (potentially) potable
aquifers are not contaminated.
4.5. Atmospheric Emissions
Apart from CO2 (the major component by weight - see Figure 11) this covers Volatile
Organic Compounds (figure 11) that are emitted to the atmosphere due to deliberate
venting, un-burnt gas during flaring and fugitives from tanker loading operations,
leaking valves and pump seals, tank roof vents etc. Reduction in hydrocarbon
emissions via reduced flaring is now receiving the full management attention that it
deserves. The scale of the problem - and the opportunity with respect to improving
recovery - was summarised in Figure 10.
Being able to usefully use the gas co-produced with the oil depends on the availability
of a suitable infrastructure (transport pipeline and end user) to allow the economic
evacuation of the gas. Gas fields remote from existing infrastructure, or those
containing only small quantities of gas, make the finding of economic solution for the
so called "stranded gas" problem particularly challenging. This is particularly true
now that large scale continuous flaring of excess gas is now seen as unacceptable by
both the regulatory authorities and many company managements. Solutions employed
to date have included:
(i)
injecting the gas back into the reservoir to increase liquid recovery (either by
initiating a gas drive or condensate recycling mechanism);
(ii) temporarily storing the gas in an underground formation while awaiting the
commissioning of the sales pipeline;
(iii) developing a market for the gas e.g. Liquefied Natural Gas scheme, Synfuel/
methanol synthesis plants, power generation (though offshore location of the
power plant apart from local platform requirements has not proved economic
to date);
(iv) use of the gas for artificial lift (gas lift);
(v) disposal of the gas in a suitable reservoir to avoid the “carbon tax” penalty
levied in some countries.
Other innovative solutions will no doubt be developed in the coming years since
several research projects to solve the “stranded gas” problem are now in progress.
26
102
Oil and Gas Field Operations
OIL AND GAS FIELD OPERATIONS TUTORIAL 1
List 3 key uncertainties faced by Production Technology/Operations/Engineering.
SOLUTION
(a) Reserves.
Surface facilities design, in terms of handling the production rates, will depend
largely on the size of reserves associated with that particular field. Real field’s
(potential) plateau production and reserves may be much larger or smaller than the
facility design values.
(b) Reservoir Description.
Wellhead conditions are subject to uncertainty until the formation is actually drilled.
Reservoir pressure conditions and fluid type characteristics will affect, for example,
the design and specifications for wellhead equipment and surface flow lines.
(c) Reservoir Drive Mechanism.
The production rates of the individual fluids will change and then decline as the field
matures. The rate at which these fluids change, and the long term production
performance will depend on the reservoir production mechanism i.e. the drive
mechanism which forces oil from a distant location in the reservoir into the well and
subsequently to the surface.
OIL AND GAS FIELD OPERATIONS TUTORIAL 2
Explain the statement “Production Planning and Optimisation occurs at different time
scales”.
SOLUTION
Reservoir and well conditions change throughout the life of a field. Production
management is a continuous process of surveillance and reaction to these changes.
The processes designed to meet the objectives of the Production optimisation model
and the contingency operations are executed as follows:
(a) “Maximise Economic Production Rate” => Immediate.
The impact of temporary operational difficulties can be minimised by immediate
action (re-routing wells, adjustment of producing conditions, etc).
(b) “Minimise Production Downtime” => Hours.
This requires an effective surveillance program to rapidly identify and repair well
problems.
(c) “Identify Production Restrictions” => Short/medium term.
Identify those factors limiting production (e.g. compressor capacity, pipeline restrictions, well productivity, etc) and perform an economical analysis to asses remedial
actions.
Department of Petroleum Engineering, Heriot-Watt University
27
1
(d) “Maintenance Planning” => Short/medium term.
Operating conditions such as field lift gas allocation, production rate setting, etc.
need to be reset in order to maximise production when equipment has to be taken out
of service for maintenance.
(e) “Planning for the future” => Long term.
Production forecasts can be made by simulating expected future production conditions.
The optimum time for the installation/implementation of artificial lift, additional
compression capacity, drilling of new wells or development of near field potential,
etc, can then be identified.
28
The Haggis Field Exercise
For
Distance Learning
Combined Question & Blank Answer
Form1
VERSION 2.2
1
The % mark noted at the end of each question refer to its relative weighting with respect to the complete
exercise
Figures, which are referred to in the text, are placed at the end of the exercise.
DECLARATION
I …………………………………confirm that this work submitted for
assessment is my own and expressed in my own words. Any uses made within it of
the works of other authors in any form (e.g. ideas, equations, figures, text, tables,
programs) are properly acknowledged at the point of their use. (A list of the
references employed should be included.)
Signed : …………………………………………..
………………
Date :
Name in Capitals……………………………………………………..
-----------------------------------------------------------------------------------------------------------------
Guidance Notes for Students
Heriot-Watt University Ordinance 9, para 2.4 (c) states:
A student may be deemed to be in breach of discipline if he should:
During an examination or other test copy from or communicate with another
person or be found in possession of books or any printed or written papers or
any other material containing information relevant to the subject of the
examination other than those allowed in the examination or use any other
unfair means.
20.1.3 The words highlighted in this extract refer to “other test” and “other
unfair means”. These guidance notes are intended to draw the attention of
students to the importance of avoiding inadvertent use of what might be
regarded as unfair means.
“Other test” is construed in the widest possible sense of any assessed work
that contributes to the award of a degree or other qualification of the
University.
“Unfair means” is construed in the widest possible sense of any practice that is
intended to gain a dishonest advantage.
20.1.4 The particular unfair means noted in this paper are copying, plagiarism
and collusion, although this is by no means an exhaustive list.
20.2 COPYING
20.2.1 Copying the work of others, including that of other students in the class
or group, is an indication of unfair means, whereby one person gains credit
for work undertaken by another.
The Haggis Field
20.2.3 Recommendation to students Make clear in your submission any permitted
reproduction that you have carried out. Check the rest of the work to ensure that
it is your own. Working with other students in informal study groups is a
desirable part of the academic experience, but be sure that the work that you
eventually submit is yours and not that of others. Keep copies of material such as
working notes, or sketches of diagrams, or drafts of essays that show that the
work is your own effort.
20.3 PLAGIARISM
20.3.1 Plagiarism is the act of stealing from the writings or ideas of another. This is
regarded as a very serious offence in academic works. It is generally accepted
that such “stealing” occurs where there is no acknowledgement that the
writings or ideas belong to someone else. Most academic scholarship involves
building on the work of others, while acknowledging their contribution. There
are accepted conventions for making that acknowledgement, although the
conventions may vary marginally from one subject to another.
20.3.3 Recommendation to students When you undertake an assessed work which
involves drawing on the writings or ideas of others, make sure that you
acknowledge each contribution. Use a style of acknowledgement that is good
practice in the academic discipline. If you are not sure what is good practice,
read the guidance provided by your School or seek advice from academic
staff. When the work is completed, check carefully that you have not
overlooked acknowledgement of any source used.
20.4 COLLUSION
20.4.1 Collusion involves a secret agreement to deceive. This means that more than
one person is involved in the deception. Where an accusation of collusion is
added to an accusation of, for example, copying, there must be clear evidence
of the involvement of each party. One person may copy the work of another
without the knowledge or approval of that person.
20.4.3 Recommendation to students Be careful about lending your completed work to
other persons. You may think that you are helping them to meet a deadline,
but it may result in problems for you if they copy your work without telling
you. What started out as a friendly action might risk an accusation of
collusion.
The Haggis Field
The Haggis Field
The Haggis field was discovered in June 1994 in the UK sector of the North Sea in a water depth of
300 ft. The field was developed using 5 wells and reached peak production in 1996. Since then, oil
production has decreased rapidly due to an increase in water production.
Reservoir Properties:
The Haggis sand was deposited in a turbidite environment. It is quite homogeneous with an average
porosity and permeability of 22% and 200 mD, respectively. The reservoir sand, however, is
isotropic with a Kv/Kh ratio of 0.1. The top of the sand was encountered at 6400 ft TVDSS, and the
oil-water contact is at 6500 ft TVDSS.
The reservoir is normally pressured with an initial reservoir pressure of 3300 psia and little or no
aquifer support. Reservoir pressure has declined with production to 2800 psia at present. Pressure
maintenance was not considered when the field was being developed.
For information and for those wishing to run this exercise using a completion design software
package, Table 1 lists the PVT data for the Haggis fluids at current reservoir conditions.
150° F
Reservoir Temp.
40° API
Oil API Gravity
0.80
Gas Relative Density
550 scf/STB
G.O.R.
2030 psia
Pb
1.27
Bo
0.66 cp
Oil Viscosity
0.0046
Bg
0.022 cp
Gas Viscosity
1.023
Bw
0.73
Gas Z-Factor
200000 ppm
Water Salinity
0.67 cp
Water Viscosity
Table 1: Haggis PVT Data
Haggis wells:
The Haggis field wells have an economical limit of 1500 STB Oil/d/well; i.e. producing at rates lower
than that is not feasible.
Haggis-3 was drilled in May 1995. It is taken to be the case study for this field as it has average
parameters for Haggis wells. Figure 1 is the completion diagram for Haggis-3. Above the wellhead,
the well was completed with the same 5 1/2" OD production tubing encased in a mud line (no
insulation). The mud line connects the wellhead (on the sea-bed) to the Xmas tree on the platform.
For information, results of a recent pressure survey from Haggis-3 are listed in Table 2.
Depth
Pressure
650
525
1605
735
2590
3600
4590
990
1292
1629
Table 2: Haggis-3 Pressure Survey
5587
1920
6490
2266
ft TVD
psia
The Haggis Field
Haggis-3's well parameters, and results from both well testing and production logging are summarised
in Table 3. Table 3 also contains data on Haggis-1. Haggis-3 and Haggis-1 are essentially twin
wells, except that completion damage has resulted in Haggis-1 being considered the worst well in the
field.
Oil Production Rate
Water Cut
WH Flowing Temperature
Pressure at Xmas tree
Skin (Well Test)
P.I. (J) (Well Test)
Damaged Zone Relative
Permeability
Damage Zone Thickness
Crushed Zone Skin
Drainage Radius
Haggis-3
4730
30
65
Haggis-1
3930
28
61
445
2.92
12.36
7.17
9.11
50
25
12
0.100
4000
Table 3: Well Data
STB/d
%
°F
psia
STB/d/psi
%
in
ft
The Scenario:
The rate of oil production decline in the Haggis field is alarming, and if no action is taken, Haggis
will become uneconomical by the end of this year. The Operator of the field, Big Kahuna Oil Inc.,
does not accept this situation and has fired the field's former team leader for improper management of
the field. Big Kahuna has hired you to improve production from the Haggis field.
Your Mission:
Big Kahuna Oil Inc. has asked you to study the field’s potential. A model has been created using the
company approved software (EPS's FloSystem). A variety of production proposals have been
modelled and the outcome of these simulations has been provided in graphical form. It is your job to
evaluate these proposals using the all the data provided.
The report should outline:
A) the model used in the study,
B) the potential of the base case scenario,
C) your assessment of production enhancement proposals from the Haggis engineers and
D) your recommendation for a project which will enhance production from Haggis.
Note: Wellflo graphs show gross production rates e.g. total production rates. Net production rates are
required for the well analysis.
The Haggis Field
A) Develop a Well Model for Haggis-3:
a) Using Haggis-3 as your case study, complete the missing data in Figure 2.
b) The well Haggis-3 is used as the base case well for the Haggis field throughout this exercise. To
minimise the computer time involved in simulations, the model contains only those components
that contribute significantly to the pressure drop along Haggis-3. These are shown on Table 4.
Node
No.
1
2
3
4
5
6
7
8
Component Name
Outlet node / Xmas tree
Riser
Wellhead
5.5” Tubing
S.C.S.S.S.V
5.5” Tubing
5” Tubing
7” Liner
Table 3: Well Data
Measured
Depth (ft)
0
350
350
850
850
4000
5600
6530.5
Compare this model to the completion design in Figure 1.
When would it be necessary to include the nipples and tubing constrictions in the model design?
_____________________________________________________________________________
_____________________________________________________________________________
[2.5%]
c) The Outflow curve for Haggis-3 was modelled using a number of well-known correlations. The
sensitivity analysis is shown in Figure 3. The correlation that is most appropriate is:
__________________________________________________________________________
[2%]
There are two main reasons why this correlation was chosen:
1) __________________________________________________________________________
__________________________________________________________________________
2) __________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
[2.5%]
d) The Big Kahuna Oil Inc. is unfamiliar with the concept of nodal analysis used in this flow
simulation. You are asked to explain the process and the conditions for flow stability.
The inflow-outflow diagram below is for Haggis-3 at 30% water cut and utilising an appropriate
flow correlation.
The Haggis Field
Point A
Point E
Curve A
Point D
Point B
Curve B
Point C
Diagram 1: Haggis-3 base case
ƒ
ƒ
ƒ
Curve A represents the _________________________________________________.
Curve B represents the _________________________________________________.
The reservoir pressure is _____________psia (Point______).
ƒ
The_________________ IPR model was used to produce this curve. Why was this chosen?
______________________________________________________________________________
_____________________________________________________.
ƒ
Point ______ represents the operating point. The operating point is [stable/unstable]. Explain:
______________________________________________________________________________
______________________________________________________________________________
_________________________________________.
Nodal analysis may be carried out at any point in the producing system. In this report, the analysis
was carried out to find the operating point at the sandface (see above diagram).
Other typical examples of nodes selected during completion design are:
1)__________________________________________________________to evaluate
_________________________________________________________________________________
2)__________________________________________________________to evaluate
_________________________________________________________________________________
[11.5%]
The Haggis Field
B) Base Case Analysis:
As a good manager, the first thing you have to do is evaluate the potential of what you have at the
moment. To achieve that, you must determine what effect the decline in reservoir pressure and the
increase in water cut will have on Haggis-3's production if nothing is done to improve its production.
In other words, determine the reservoir pressure and the water cut at which Haggis-3 will becomes
uneconomical to produce under the current production scenario.
a) In order to evaluate the viability of producing Haggis-3 as water-cut increases and reservoir
pressures reduces, a number of sensitivity analyses were performed on the production from
Haggis-3 using WellFlo. These sensitivities are illustrated by Figures 4a, b, c and d, and their
results should be summarised in Table 5:
PRes.
WC
30%
35%
40%
45%
2800
2700
2600
2500
psia
4770
Table 5: Haggis-3 Production Forecast
[4%]
b) You discuss these figures with the field’s engineers. Since artificial lift can not be supported by
the production facilities on the Haggis platform, you agree with your engineers that you have to
start a water injection scheme to maintain the reservoir pressure at 2800 psia. Figure 5 shows the
well sensitivity to water cut. Under these circumstances, Haggis-3 will produce economically
(1500+ BOPD) at maximum water cut of ____________, and at an oil production rate of
____________ BOPD. This is considered to be the Base Case scenario, against which all other
schemes in section C will be compared.
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Base Case
c) With a water injection scheme in place, you expect to face even more severe water production
from Haggis-3. One way of dealing with such a problem is to plug-off “watered-out”
perforations. List two advantages and two disadvantages of this scheme.
Advantages:
1)____________________________________________________________________________
_______________________________________________.
2)____________________________________________________________________________
_______________________________________________.
Disadvantages:
1)____________________________________________________________________________
_______________________________________________.
2)____________________________________________________________________________
_______________________________________________.
[5%]
The Haggis Field
d) A sensitivity study is outlined in Figures 6a, b and c for three plugging-off policies. Based on
the advantages and disadvantages you identified above, the most advantageous plugging off
policy is:____________________________________________________.
The minimum open interval at which Haggis-3 will produce economically under this policy is
________________________ft.
[2.5%]
C) Further Projects:
You phone management and convince them to fund the pressure maintenance scheme. In addition,
they agree to provide a budget for an extra project to improve Haggis’s production. In this section
you must select the best project from your engineers’ suggestions below.
Since production, and thus revenue, from Haggis is greatly affected by water production form the
field, Big Kahuna Oil Inc. defines the best production enhancement project for the Haggis field as the
one that sustains economical production form Haggis-3 at the highest water-cut. Therefore, this
should be the criterion you use to select the project you recommend to management.
C.1) Production Technology Solutions:
You walk into the team’s senior production technologist office, and give him the good news from
management. He congratulates you, and immediately suggests that the project should be a production
technology project. He goes on to explain that as the Haggis wells are extremely damaged, acidising
all the wells make the field very profitable.
a) Determine the benefit from acidising Haggis-3 in terms of the maximum water cut at which the
acidised Haggis-3 will sustain economic production, if acidising restores the original rock
permeability. Figure 7 shows the sensitivity of Haggis-3 to water cut after acidising. The
maximum water cut at which the well can produce economically is ________%.
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Acidising
(Use this result to compare the benefits from this project to the other proposed projects, and as a
basis for your recommendations in section D).
b) A production model of well Haggis-1 is already available and shows that Haggis-1 will produce
economically at maximum water cuts of 50 % after acidising. What are the implications of this if
it is decided to carry out a campaign in which all the Haggis wells are to be acidised?
______________________________________________________________________________
______________________________________________________________________________
______________________________________________________________________________
[3.5%]
The Haggis Field
c) The diagram below shows the inflow-outflow curves for Haggis-3 at the maximum economic
water cut after acidising. Assuming that acidising succeeds in restoring the original rock
permeability in both wells, sketch on the diagram the inflow-outflow curves for Haggis-1.
Diagram 2: Haggis-3 after acidising
[2%]
C.2) Suggestions from the Drilling Engineer:
The drilling engineer walks into your office and says that he heard that you were going around the
different departments asking for ideas on how to improve the field's production. He says that he
could have saved you all the trouble because he has the perfect solution. "Side-track the well", he
says. He goes on to explain that, he would have gone for a horizontal well. However, since Big
Kahuna have recently had bad luck with drilling horizontal sections longer than 400 ft, a 75° deviated
well through the reservoir is probably the better option from a drilling point of view.
a) Given the drilling department’s recent experience with horizontal wells, should you side-track
Haggis-3 to a 400 ft horizontal well running through the middle of the reservoir or propose a 75°
deviated well penetrating the whole of the reservoir height?
The production prognosis from Haggis-3 sidetrack as a 400ft horizontal well is shown in Figure
8 while a 75° deviated well is illustrated in Figure 9. Summarise the results in the table below:
Option
Scenario
C2.a1
C2.a2
400ft Horizontal
75° Deviated
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Option C2.a2 produces the better results because:
______________________________________________________________________________
______________________________________________________________________________
[3%]
The Haggis Field
b) Given the engineering and economic factors and assumptions below, design the optimum
horizontal well (to the nearest 500-ft) for the Haggis-3 side-track:
Length of build-up section is:
Pay-back Time:
Side-track cost:
Price of Oil:
Cost of Processing and Shipping:
Assumptions:
2000 ft-MD
6 months
US$850 /ft-MD
US$15/bbl
US$8 / bbl
1) Economics based on un-dicounted
cash flows
2) Production rate is constant for the first
6 months
3) Use Profit to Investment Ratio as
measure of value, see below.
(Profit to Investment (PI) Ratio = TCS/MCO where TCS is Terminal Cash Surplus and MCO is
Maximum Capital Outlay. The PI ratio is a measure of the cash surplus or profit generated for every
unit of currency invested.)
Table 6 below can be completed using the results from Figure 10 and taking in account the
above engineering and economic factors for drilling horizontal wells.
Horizontal length (ft)
Parameter
and unit
500
1000
1500
2000
2500
3000
3500
Formula
Table 6:Economic analysis of sidetracking Haggis-3.
(In the space above, enter the parameter, formulas used and intermediate working)
Table 6 indicates that the optimum horizontal section length for Haggis-3 is _____ ft. However,
by carefully analysing the economic outcome and technical aspects of the design, what other
conclusion could be drawn?
_____________________________________________________________________________
_____________________________________________________________________________
_____________________________________________________________________________
4000
The Haggis Field
c) Determine the benefit from side-tracking Haggis-3 in terms of the maximum water cut you can
economically produce the well with if it was side-tracked to the optimal horizontal length
determined above.
If Haggis-3 is sidetracked to the optimal horizontal length, then the maximum water-cut at which
the well will produce economically is ____%. (Figure 11)
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Horizontal Well
(Use this result to compare the benefits from this project to the other proposed projects, and
as a basis for your recommendations in section D).
[15%]
C.3) Another Production Technology Suggestion: Artificial Lift
After lunch, Haggis’s senior production technologist steps into your office. He says that while he
was having lunch, he remembered that Big Kahuna has been quite successful using Electric
Submersible Pumps (ESP’s) in the nearby Tatties field. Additionally, he says that gas lift may be
an option as the well has a high PI and the Big Kahuna has sufficient gas supply from another
nearby field. He says that he is aware of the drilling department suggesting a slanted well, but
installing artificial lift will be quicker and cheaper than side tracking the well. He suggests
investigating the installation of an artificial lift scheme with a target liquid rate of 9000 STB/d.
C.3.1: Electrical Submersible Pumps
a) The Big Kahuna has had success with the following Centrilift pumps in the North Sea.
Pump
GC 8200
HC 7000
HC 9000
KC 12000
Motor
562 Series
562 Series
562 Series
562 Series
Cable Size
#1
#1
#2
#2
Given the current conditions choose the optimum pump for Haggis-3 given the details and
engineering assumptions below,
Pump Types:
Centrilift
Setting Depth:
5000ft
Minimum Equipment OD:
5”
Maximum Equipment OD:
6.8”
Platform Electricity Supply Frequency: 60Hz
Assumptions:
1) Production tubing is unaltered.
2) No wear on Pump or motor i.e. wear
factor is 1.
3) Efficiency of the gas separator is 100%
i.e. separator efficiency is 1.
4) Viscosity and gassiness corrections are
used.
Figures 12a & b show the performance plots of the pumps at current conditions. The most
suitable pump of Haggis-3 given the present conditions is ______________.
The Haggis Field
This pump is the optimum choice because
___________________________________________________________________________
___________________________________________________________________________
__________________________________________________________________________.
[3%]
b) As water production is the limiting factor, the production technologist suggests halting water
injection and allowing the reservoir pressure to drop. Determine which pump would be
suitable in these conditions. Figures 13 a-d and Table C.3 are designed to assist you with that
determination.
PRes.
Psia
2800
2600
2400
2200
Pump
GC 8200
HC 7000
HC 9000
KC 12000
Table C.3: Haggis-3 Production Forecast with ESP installed,
** denotes rate outwith the operating range of the pump.
According to Table C.3 the optimum ESP for Haggis-3 for declining reservoir pressure is
________________, at a water cut of 30%.
This pump is the optimum choice because
___________________________________________________________________________
___________________________________________________________________________
__________________________________________________________________________.
[5%]
c) Determine the benefit from installing an ESP in Haggis-3 using Figure 14 in terms of the
maximum water-cut at which the optimised pump will sustain economic production prior to
suspending water injection (i.e. no depletion, reservoir pressure 2800psia).
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Optimised ESP
(Use this result to compare the benefits from this project to the other proposed projects, and
as a basis for your recommendations in section D).
[1.5%]
The Haggis Field
C.3.2: Gas Lift Design
a) A gas lift design for Haggis-3 is undertaken based on current conditions and engineering
assumptions/details outlined below,
Max casing head pressure:
1200 psi
Gas available for injection:
20MM scf/d
Injected gas gravity:
0.6
Setting depth:
3900 ft
Valve differential pressure:
100 psia
Minimum spacing:
450 ft
“Kill” brine density:
0.465 psi / ft
Minimum safety margin (see A on diagram)
50 psi
Assumptions:
1) Production tubing is unaltered.
2) Unload the tubing full of static fluid
against the well head pressure (i.e. static
fluid to 0ft MD).
3) No transfer margin is required.
The gas lift design is shown below (Diagram 3). Note briefly on the diagram the roles of the
different valves in the design. The upper valves should be OPEN/CLOSED when assessing gas
lift capabilities during field life.
A
Diagram 3: Gas lift design for Haggis-3
[3%]
The Haggis Field
b) Determine the optimum injection rate as the reservoir pressure declines from Figure 15 and
summarise the results in Table C.4.
PRes.
2800
2600
2400
2200
2000
Psia
Optimum
injection rate
MMscf/day
Table C.4: Optimum gas injection rate for Haggis-3.
The required gas injection rate for the design production rate is __________________________.
The criteria used to choose the optimum injection rate are
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
______________________________________________________________________________.
Higher injection rates do not improve production as the reservoir declines because
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
[8%]
c) The gas lift scheme is redesigned using the optimum gas injection rate for a reservoir pressure
of 2800 psia. Determine the benefit from installing Gas Lift in Haggis-3 in terms of the
maximum water-cut at which the optimised injection rate will sustain economic production
using Figure 16.
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Optimised Gas Lift
(Use this result to compare the benefits from this project to the other proposed projects, and
as a basis for your recommendations in section D).
[2%]
C.4) Suggestions from the Facilities Engineer:
The gentleman behind you introduces himself as the Haggis field's facilities engineer when you are
walking to your car after a long day at the office. He shares with you a very interesting discovery he
made today. He explains that if you agree to provide a dedicated line from the Haggis wells to the
low-pressure separator, you can lower the average Xmas tree pressure to only 100 psia. He points out
that suggestions from the other department have large uncertainties associated with them since they
deal with the subsurface. His suggestion is simple, neat and will solve the field's problems.
The Haggis Field
a) Consider the base case inflow-outflow diagram for Haggis-3 below. Sketch on the diagram how
the base case curves would alter if the well head pressure were decreased.
Diagram 4: Haggis-3 base case
[2%]
b) Evaluate the benefits of lowering the Xmas tree pressure, in terms of the maximum water-cut you
can economically produce Haggis-3 with after the Xmas tree pressure is lowered to 100 psia.
If the production facilities on the Haggis platform can be modified to allow the Xmas tree
pressure of Haggis-3 to be lowered to 100 Psia, then the maximum water-cut at which the well
will produce economically will become _____%. (See Figure 17)
Scenario
Maximum Economic
Water Cut
Production Rate @ 30%
Water Cut
Lowering Xmas
Tree Press
(Use this result to compare the benefits from this project to the other proposed projects, and as a
basis for your recommendations in section D).
c) List 2 advantages and 2 disadvantages of this scheme.
Advantages
1)____________________________________________________________________________
______________________________________________________________________________
2)____________________________________________________________________________
______________________________________________________________________________
and disadvantages
3)____________________________________________________________________________
______________________________________________________________________________
4)____________________________________________________________________________
_____________________________________________________________________________
[5%]
The Haggis Field
D) Recommendations to Management:
a) Assess the production enhancement projects proposed by the Haggis field engineers in section C
above and compare them to one another and to the base case scenario. Bear in mind that Big
Kahuna Inc. has set the ranking criteria for these projects to be the maximum water-cut at which
Haggis-3 can sustain economic production (i.e. > 1500 STB oil/d).
Table 8 below summarises the results from the various simulations carried out on the Haggis-3
well.
Scenario
Maximum Economic
Water Cut
Production Rate @ 30%
Water Cut
Base Case
Acidising
75° Deviated Well
Optimum Horizontal
Well
Optimum ESP
Optimum Gas Lift
Lowering Xmas Tree
Pressure to 100 psia
Table 8: Haggis-3 Production Forecast
[2.5%]
b) Recommend to management a plan of action which either recommends maintaining the base case
scenario or executes one of the proposed projects.
Based on the WellFlo simulations and my assessment of them, I recommend that Big Kahuna Inc.
invest in a water injection scheme to maintain Haggis’s reservoir pressure at 2800 psia. In
addition, I recommend that Big Kahuna adopts [the _______________________ project /none of
the projects investigated above] because:
_______________________________________________________________________________
______________________________________________________________________________.
[5%]
3) A number of risks have been overlooked in this assessment since the maximum water cut at which
the wells will flow at an economic rate has been used as the ranking criteria for the above projects.
These risks add to the uncertainty of achieving the results on which your recommendation was
based. As the Haggis field team leader it is your duty to report and account for these risks to
management.
Complete Table 9 below which should identify three major risks that have been overlooked by
this assessment. Briefly explain how each one could add to the uncertainty of the assessment and
prescribe steps that need to be taken to account for their effects.
The Haggis Field
Risk/Uncertainty
How this adds uncertainty to
above assessment?
Steps that can be taken to account for
/ minimise this uncertainty
1)
2)
3)
Table 9: Haggis-3 Production Forecast
[9%]
The Haggis Field
Figures:
The Haggis Field
The Haggis Field
The Haggis Field
Figure 3: Sensitivity to flow correlation.
The Haggis Field
Figure 4a: Layer pressure =2800 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 4b: Layer pressure = 2700 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 4c: Layer pressure = 2600 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 4d: Layer pressure = 2500 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 5: Layer pressure = 2800 psia, maximum economic water cut.
Operating point region
The Haggis Field
Plugging off policies
Figure 6a: Plug off at 30% water cut, sensitivity to open interval.
Operating point region
The Haggis Field
Figure 6b: Plug off at 40% water cut, sensitivity to open interval.
Operating point region
The Haggis Field
Figure 6c: Plug off at 48% water cut, sensitivity to open interval.
Operating point region
The Haggis Field
Figure 7: Haggis-3 after acidising (e.g. damage zone permeability 200mD),
sensitivity to water cut.
Operating point region
The Haggis Field
Figure 8: 400ft horizontal well, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 9: 75° deviated well, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 10: Horizontal well, sensitivity to effective length.
Operating point region
The Haggis Field
Figure 11: Optimum horizontal well, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 12a: Performance curves for ESP GC8200 and HC7000
Figure 12b: Performance curves for ESP HC9000 and KC12000
The Haggis Field
Figure 13a: GC 8200, sensitivity to reservoir pressure
Operating point region
The Haggis Field
Figure 13b: HC 7000, sensitivity to reservoir pressure.
Operating point region
The Haggis Field
Figure 13c: HC 9000, sensitivity to reservoir pressure.
Operating point region
The Haggis Field
Figure 13d: KC 12000, sensitivity to reservoir pressure.
Operating point region
The Haggis Field
Figure 14: Optimum ESP, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 15: Performance analysis of the gas lift design, sensitivity to reservoir pressure and gas
injection rate.
The Haggis Field
Figure 16: Optimum gas lift design, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 17: 100psia well head pressure, sensitivity to water cut.
Operating point region
The Haggis Field Exercise
For
Distance Learning
Combined Question & Blank Answer
Form1
VERSION 2.2
1
The % mark noted at the end of each question refer to its relative weighting with respect to the complete
exercise
Figures, which are referred to in the text, are placed at the end of the exercise.
DECLARATION
I …………………………………confirm that this work submitted for
assessment is my own and expressed in my own words. Any uses made within it of
the works of other authors in any form (e.g. ideas, equations, figures, text, tables,
programs) are properly acknowledged at the point of their use. (A list of the
references employed should be included.)
Signed : …………………………………………..
………………
Date :
Name in Capitals……………………………………………………..
-----------------------------------------------------------------------------------------------------------------
Guidance Notes for Students
Heriot-Watt University Ordinance 9, para 2.4 (c) states:
A student may be deemed to be in breach of discipline if he should:
During an examination or other test copy from or communicate with another
person or be found in possession of books or any printed or written papers or
any other material containing information relevant to the subject of the
examination other than those allowed in the examination or use any other
unfair means.
20.1.3 The words highlighted in this extract refer to “other test” and “other
unfair means”. These guidance notes are intended to draw the attention of
students to the importance of avoiding inadvertent use of what might be
regarded as unfair means.
“Other test” is construed in the widest possible sense of any assessed work
that contributes to the award of a degree or other qualification of the
University.
“Unfair means” is construed in the widest possible sense of any practice that is
intended to gain a dishonest advantage.
20.1.4 The particular unfair means noted in this paper are copying, plagiarism
and collusion, although this is by no means an exhaustive list.
20.2 COPYING
20.2.1 Copying the work of others, including that of other students in the class
or group, is an indication of unfair means, whereby one person gains credit
for work undertaken by another.
The Haggis Field
20.2.3 Recommendation to students Make clear in your submission any permitted
reproduction that you have carried out. Check the rest of the work to ensure that
it is your own. Working with other students in informal study groups is a
desirable part of the academic experience, but be sure that the work that you
eventually submit is yours and not that of others. Keep copies of material such as
working notes, or sketches of diagrams, or drafts of essays that show that the
work is your own effort.
20.3 PLAGIARISM
20.3.1 Plagiarism is the act of stealing from the writings or ideas of another. This is
regarded as a very serious offence in academic works. It is generally accepted
that such “stealing” occurs where there is no acknowledgement that the
writings or ideas belong to someone else. Most academic scholarship involves
building on the work of others, while acknowledging their contribution. There
are accepted conventions for making that acknowledgement, although the
conventions may vary marginally from one subject to another.
20.3.3 Recommendation to students When you undertake an assessed work which
involves drawing on the writings or ideas of others, make sure that you
acknowledge each contribution. Use a style of acknowledgement that is good
practice in the academic discipline. If you are not sure what is good practice,
read the guidance provided by your School or seek advice from academic
staff. When the work is completed, check carefully that you have not
overlooked acknowledgement of any source used.
20.4 COLLUSION
20.4.1 Collusion involves a secret agreement to deceive. This means that more than
one person is involved in the deception. Where an accusation of collusion is
added to an accusation of, for example, copying, there must be clear evidence
of the involvement of each party. One person may copy the work of another
without the knowledge or approval of that person.
20.4.3 Recommendation to students Be careful about lending your completed work to
other persons. You may think that you are helping them to meet a deadline,
but it may result in problems for you if they copy your work without telling
you. What started out as a friendly action might risk an accusation of
collusion.
The Haggis Field
The Haggis Field
The Haggis field was discovered in June 1994 in the UK sector of the North Sea in a water depth of
300 ft. The field was developed using 5 wells and reached peak production in 1996. Since then, oil
production has decreased rapidly due to an increase in water production.
Reservoir Properties:
The Haggis sand was deposited in a turbidite environment. It is quite homogeneous with an average
porosity and permeability of 22% and 200 mD, respectively. The reservoir sand, however, is
isotropic with a Kv/Kh ratio of 0.1. The top of the sand was encountered at 6400 ft TVDSS, and the
oil-water contact is at 6500 ft TVDSS.
The reservoir is normally pressured with an initial reservoir pressure of 3300 psia and little or no
aquifer support. Reservoir pressure has declined with production to 2800 psia at present. Pressure
maintenance was not considered when the field was being developed.
For information and for those wishing to run this exercise using a completion design software
package, Table 1 lists the PVT data for the Haggis fluids at current reservoir conditions.
150° F
Reservoir Temp.
40° API
Oil API Gravity
0.80
Gas Relative Density
550 scf/STB
G.O.R.
2030 psia
Pb
1.27
Bo
0.66 cp
Oil Viscosity
0.0046
Bg
0.022 cp
Gas Viscosity
1.023
Bw
0.73
Gas Z-Factor
200000 ppm
Water Salinity
0.67 cp
Water Viscosity
Table 1: Haggis PVT Data
Haggis wells:
The Haggis field wells have an economical limit of 1500 STB Oil/d/well; i.e. producing at rates lower
than that is not feasible.
Haggis-3 was drilled in May 1995. It is taken to be the case study for this field as it has average
parameters for Haggis wells. Figure 1 is the completion diagram for Haggis-3. Above the wellhead,
the well was completed with the same 5 1/2" OD production tubing encased in a mud line (no
insulation). The mud line connects the wellhead (on the sea-bed) to the Xmas tree on the platform.
For information, results of a recent pressure survey from Haggis-3 are listed in Table 2.
Depth
Pressure
650
525
1605
735
2590
3600
4590
990
1292
1629
Table 2: Haggis-3 Pressure Survey
5587
1920
6490
2266
ft TVD
psia
The Haggis Field
Haggis-3's well parameters, and results from both well testing and production logging are summarised
in Table 3. Table 3 also contains data on Haggis-1. Haggis-3 and Haggis-1 are essentially twin
wells, except that completion damage has resulted in Haggis-1 being considered the worst well in the
field.
Oil Production Rate
Water Cut
WH Flowing Temperature
Pressure at Xmas tree
Skin (Well Test)
P.I. (J) (Well Test)
Damaged Zone Relative
Permeability
Damage Zone Thickness
Crushed Zone Skin
Drainage Radius
Haggis-3
4730
30
65
Haggis-1
3930
28
61
445
2.92
12.36
7.17
9.11
50
25
12
0.100
4000
Table 3: Well Data
STB/d
%
°F
psia
STB/d/psi
%
in
ft
The Scenario:
The rate of oil production decline in the Haggis field is alarming, and if no action is taken, Haggis
will become uneconomical by the end of this year. The Operator of the field, Big Kahuna Oil Inc.,
does not accept this situation and has fired the field's former team leader for improper management of
the field. Big Kahuna has hired you to improve production from the Haggis field.
Your Mission:
Big Kahuna Oil Inc. has asked you to study the field’s potential. A model has been created using the
company approved software (EPS's FloSystem). A variety of production proposals have been
modelled and the outcome of these simulations has been provided in graphical form. It is your job to
evaluate these proposals using the all the data provided.
The report should outline:
A) the model used in the study,
B) the potential of the base case scenario,
C) your assessment of production enhancement proposals from the Haggis engineers and
D) your recommendation for a project which will enhance production from Haggis.
Note: Wellflo graphs show gross production rates e.g. total production rates. Net production rates are
required for the well analysis.
The Haggis Field
A) Develop a Well Model for Haggis-3:
a) Using Haggis-3 as your case study, complete the missing data in Figure 2.
b) The well Haggis-3 is used as the base case well for the Haggis field throughout this exercise. To
minimise the computer time involved in simulations, the model contains only those components
that contribute significantly to the pressure drop along Haggis-3. These are shown on Table 4.
Node
No.
1
2
3
4
5
6
7
8
Component Name
Outlet node / Xmas tree
Riser
Wellhead
5.5” Tubing
S.C.S.S.S.V
5.5” Tubing
5” Tubing
7” Liner
Table 3: Well Data
Measured
Depth (ft)
0
350
350
850
850
4000
5600
6530.5
Compare this model to the completion design in Figure 1.
When would it be necessary to include the nipples and tubing constrictions in the model design?
_____________________________________________________________________________
_____________________________________________________________________________
[2.5%]
c) The Outflow curve for Haggis-3 was modelled using a number of well-known correlations. The
sensitivity analysis is shown in Figure 3. The correlation that is most appropriate is:
__________________________________________________________________________
[2%]
There are two main reasons why this correlation was chosen:
1) __________________________________________________________________________
__________________________________________________________________________
2) __________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
__________________________________________________________________________
[2.5%]
d) The Big Kahuna Oil Inc. is unfamiliar with the concept of nodal analysis used in this flow
simulation. You are asked to explain the process and the conditions for flow stability.
The inflow-outflow diagram below is for Haggis-3 at 30% water cut and utilising an appropriate
flow correlation.
The Haggis Field
Point A
Point E
Curve A
Point D
Point B
Curve B
Point C
Diagram 1: Haggis-3 base case
ƒ
ƒ
ƒ
Curve A represents the _________________________________________________.
Curve B represents the _________________________________________________.
The reservoir pressure is _____________psia (Point______).
ƒ
The_________________ IPR model was used to produce this curve. Why was this chosen?
______________________________________________________________________________
_____________________________________________________.
ƒ
Point ______ represents the operating point. The operating point is [stable/unstable]. Explain:
______________________________________________________________________________
______________________________________________________________________________
_________________________________________.
Nodal analysis may be carried out at any point in the producing system. In this report, the analysis
was carried out to find the operating point at the sandface (see above diagram).
Other typical examples of nodes selected during completion design are:
1)__________________________________________________________to evaluate
_________________________________________________________________________________
2)__________________________________________________________to evaluate
_________________________________________________________________________________
[11.5%]
The Haggis Field
B) Base Case Analysis:
As a good manager, the first thing you have to do is evaluate the potential of what you have at the
moment. To achieve that, you must determine what effect the decline in reservoir pressure and the
increase in water cut will have on Haggis-3's production if nothing is done to improve its production.
In other words, determine the reservoir pressure and the water cut at which Haggis-3 will becomes
uneconomical to produce under the current production scenario.
a) In order to evaluate the viability of producing Haggis-3 as water-cut increases and reservoir
pressures reduces, a number of sensitivity analyses were performed on the production from
Haggis-3 using WellFlo. These sensitivities are illustrated by Figures 4a, b, c and d, and their
results should be summarised in Table 5:
PRes.
WC
30%
35%
40%
45%
2800
2700
2600
2500
psia
4770
Table 5: Haggis-3 Production Forecast
[4%]
b) You discuss these figures with the field’s engineers. Since artificial lift can not be supported by
the production facilities on the Haggis platform, you agree with your engineers that you have to
start a water injection scheme to maintain the reservoir pressure at 2800 psia. Figure 5 shows the
well sensitivity to water cut. Under these circumstances, Haggis-3 will produce economically
(1500+ BOPD) at maximum water cut of ____________, and at an oil production rate of
____________ BOPD. This is considered to be the Base Case scenario, against which all other
schemes in section C will be compared.
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Base Case
c) With a water injection scheme in place, you expect to face even more severe water production
from Haggis-3. One way of dealing with such a problem is to plug-off “watered-out”
perforations. List two advantages and two disadvantages of this scheme.
Advantages:
1)____________________________________________________________________________
_______________________________________________.
2)____________________________________________________________________________
_______________________________________________.
Disadvantages:
1)____________________________________________________________________________
_______________________________________________.
2)____________________________________________________________________________
_______________________________________________.
[5%]
The Haggis Field
d) A sensitivity study is outlined in Figures 6a, b and c for three plugging-off policies. Based on
the advantages and disadvantages you identified above, the most advantageous plugging off
policy is:____________________________________________________.
The minimum open interval at which Haggis-3 will produce economically under this policy is
________________________ft.
[2.5%]
C) Further Projects:
You phone management and convince them to fund the pressure maintenance scheme. In addition,
they agree to provide a budget for an extra project to improve Haggis’s production. In this section
you must select the best project from your engineers’ suggestions below.
Since production, and thus revenue, from Haggis is greatly affected by water production form the
field, Big Kahuna Oil Inc. defines the best production enhancement project for the Haggis field as the
one that sustains economical production form Haggis-3 at the highest water-cut. Therefore, this
should be the criterion you use to select the project you recommend to management.
C.1) Production Technology Solutions:
You walk into the team’s senior production technologist office, and give him the good news from
management. He congratulates you, and immediately suggests that the project should be a production
technology project. He goes on to explain that as the Haggis wells are extremely damaged, acidising
all the wells make the field very profitable.
a) Determine the benefit from acidising Haggis-3 in terms of the maximum water cut at which the
acidised Haggis-3 will sustain economic production, if acidising restores the original rock
permeability. Figure 7 shows the sensitivity of Haggis-3 to water cut after acidising. The
maximum water cut at which the well can produce economically is ________%.
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Acidising
(Use this result to compare the benefits from this project to the other proposed projects, and as a
basis for your recommendations in section D).
b) A production model of well Haggis-1 is already available and shows that Haggis-1 will produce
economically at maximum water cuts of 50 % after acidising. What are the implications of this if
it is decided to carry out a campaign in which all the Haggis wells are to be acidised?
______________________________________________________________________________
______________________________________________________________________________
______________________________________________________________________________
[3.5%]
The Haggis Field
c) The diagram below shows the inflow-outflow curves for Haggis-3 at the maximum economic
water cut after acidising. Assuming that acidising succeeds in restoring the original rock
permeability in both wells, sketch on the diagram the inflow-outflow curves for Haggis-1.
Diagram 2: Haggis-3 after acidising
[2%]
C.2) Suggestions from the Drilling Engineer:
The drilling engineer walks into your office and says that he heard that you were going around the
different departments asking for ideas on how to improve the field's production. He says that he
could have saved you all the trouble because he has the perfect solution. "Side-track the well", he
says. He goes on to explain that, he would have gone for a horizontal well. However, since Big
Kahuna have recently had bad luck with drilling horizontal sections longer than 400 ft, a 75° deviated
well through the reservoir is probably the better option from a drilling point of view.
a) Given the drilling department’s recent experience with horizontal wells, should you side-track
Haggis-3 to a 400 ft horizontal well running through the middle of the reservoir or propose a 75°
deviated well penetrating the whole of the reservoir height?
The production prognosis from Haggis-3 sidetrack as a 400ft horizontal well is shown in Figure
8 while a 75° deviated well is illustrated in Figure 9. Summarise the results in the table below:
Option
Scenario
C2.a1
C2.a2
400ft Horizontal
75° Deviated
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Option C2.a2 produces the better results because:
______________________________________________________________________________
______________________________________________________________________________
[3%]
The Haggis Field
b) Given the engineering and economic factors and assumptions below, design the optimum
horizontal well (to the nearest 500-ft) for the Haggis-3 side-track:
Length of build-up section is:
Pay-back Time:
Side-track cost:
Price of Oil:
Cost of Processing and Shipping:
Assumptions:
2000 ft-MD
6 months
US$850 /ft-MD
US$15/bbl
US$8 / bbl
1) Economics based on un-dicounted
cash flows
2) Production rate is constant for the first
6 months
3) Use Profit to Investment Ratio as
measure of value, see below.
(Profit to Investment (PI) Ratio = TCS/MCO where TCS is Terminal Cash Surplus and MCO is
Maximum Capital Outlay. The PI ratio is a measure of the cash surplus or profit generated for every
unit of currency invested.)
Table 6 below can be completed using the results from Figure 10 and taking in account the
above engineering and economic factors for drilling horizontal wells.
Horizontal length (ft)
Parameter
and unit
500
1000
1500
2000
2500
3000
3500
Formula
Table 6:Economic analysis of sidetracking Haggis-3.
(In the space above, enter the parameter, formulas used and intermediate working)
Table 6 indicates that the optimum horizontal section length for Haggis-3 is _____ ft. However,
by carefully analysing the economic outcome and technical aspects of the design, what other
conclusion could be drawn?
_____________________________________________________________________________
_____________________________________________________________________________
_____________________________________________________________________________
4000
The Haggis Field
c) Determine the benefit from side-tracking Haggis-3 in terms of the maximum water cut you can
economically produce the well with if it was side-tracked to the optimal horizontal length
determined above.
If Haggis-3 is sidetracked to the optimal horizontal length, then the maximum water-cut at which
the well will produce economically is ____%. (Figure 11)
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Horizontal Well
(Use this result to compare the benefits from this project to the other proposed projects, and
as a basis for your recommendations in section D).
[15%]
C.3) Another Production Technology Suggestion: Artificial Lift
After lunch, Haggis’s senior production technologist steps into your office. He says that while he
was having lunch, he remembered that Big Kahuna has been quite successful using Electric
Submersible Pumps (ESP’s) in the nearby Tatties field. Additionally, he says that gas lift may be
an option as the well has a high PI and the Big Kahuna has sufficient gas supply from another
nearby field. He says that he is aware of the drilling department suggesting a slanted well, but
installing artificial lift will be quicker and cheaper than side tracking the well. He suggests
investigating the installation of an artificial lift scheme with a target liquid rate of 9000 STB/d.
C.3.1: Electrical Submersible Pumps
a) The Big Kahuna has had success with the following Centrilift pumps in the North Sea.
Pump
GC 8200
HC 7000
HC 9000
KC 12000
Motor
562 Series
562 Series
562 Series
562 Series
Cable Size
#1
#1
#2
#2
Given the current conditions choose the optimum pump for Haggis-3 given the details and
engineering assumptions below,
Pump Types:
Centrilift
Setting Depth:
5000ft
Minimum Equipment OD:
5”
Maximum Equipment OD:
6.8”
Platform Electricity Supply Frequency: 60Hz
Assumptions:
1) Production tubing is unaltered.
2) No wear on Pump or motor i.e. wear
factor is 1.
3) Efficiency of the gas separator is 100%
i.e. separator efficiency is 1.
4) Viscosity and gassiness corrections are
used.
Figures 12a & b show the performance plots of the pumps at current conditions. The most
suitable pump of Haggis-3 given the present conditions is ______________.
The Haggis Field
This pump is the optimum choice because
___________________________________________________________________________
___________________________________________________________________________
__________________________________________________________________________.
[3%]
b) As water production is the limiting factor, the production technologist suggests halting water
injection and allowing the reservoir pressure to drop. Determine which pump would be
suitable in these conditions. Figures 13 a-d and Table C.3 are designed to assist you with that
determination.
PRes.
Psia
2800
2600
2400
2200
Pump
GC 8200
HC 7000
HC 9000
KC 12000
Table C.3: Haggis-3 Production Forecast with ESP installed,
** denotes rate outwith the operating range of the pump.
According to Table C.3 the optimum ESP for Haggis-3 for declining reservoir pressure is
________________, at a water cut of 30%.
This pump is the optimum choice because
___________________________________________________________________________
___________________________________________________________________________
__________________________________________________________________________.
[5%]
c) Determine the benefit from installing an ESP in Haggis-3 using Figure 14 in terms of the
maximum water-cut at which the optimised pump will sustain economic production prior to
suspending water injection (i.e. no depletion, reservoir pressure 2800psia).
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Optimised ESP
(Use this result to compare the benefits from this project to the other proposed projects, and
as a basis for your recommendations in section D).
[1.5%]
The Haggis Field
C.3.2: Gas Lift Design
a) A gas lift design for Haggis-3 is undertaken based on current conditions and engineering
assumptions/details outlined below,
Max casing head pressure:
1200 psi
Gas available for injection:
20MM scf/d
Injected gas gravity:
0.6
Setting depth:
3900 ft
Valve differential pressure:
100 psia
Minimum spacing:
450 ft
“Kill” brine density:
0.465 psi / ft
Minimum safety margin (see A on diagram)
50 psi
Assumptions:
1) Production tubing is unaltered.
2) Unload the tubing full of static fluid
against the well head pressure (i.e. static
fluid to 0ft MD).
3) No transfer margin is required.
The gas lift design is shown below (Diagram 3). Note briefly on the diagram the roles of the
different valves in the design. The upper valves should be OPEN/CLOSED when assessing gas
lift capabilities during field life.
A
Diagram 3: Gas lift design for Haggis-3
[3%]
The Haggis Field
b) Determine the optimum injection rate as the reservoir pressure declines from Figure 15 and
summarise the results in Table C.4.
PRes.
2800
2600
2400
2200
2000
Psia
Optimum
injection rate
MMscf/day
Table C.4: Optimum gas injection rate for Haggis-3.
The required gas injection rate for the design production rate is __________________________.
The criteria used to choose the optimum injection rate are
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
______________________________________________________________________________.
Higher injection rates do not improve production as the reservoir declines because
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
_______________________________________________________________________________
[8%]
c) The gas lift scheme is redesigned using the optimum gas injection rate for a reservoir pressure
of 2800 psia. Determine the benefit from installing Gas Lift in Haggis-3 in terms of the
maximum water-cut at which the optimised injection rate will sustain economic production
using Figure 16.
Scenario
Maximum Economic
Water Cut
Production Rate @
30% Water Cut
Optimised Gas Lift
(Use this result to compare the benefits from this project to the other proposed projects, and
as a basis for your recommendations in section D).
[2%]
C.4) Suggestions from the Facilities Engineer:
The gentleman behind you introduces himself as the Haggis field's facilities engineer when you are
walking to your car after a long day at the office. He shares with you a very interesting discovery he
made today. He explains that if you agree to provide a dedicated line from the Haggis wells to the
low-pressure separator, you can lower the average Xmas tree pressure to only 100 psia. He points out
that suggestions from the other department have large uncertainties associated with them since they
deal with the subsurface. His suggestion is simple, neat and will solve the field's problems.
The Haggis Field
a) Consider the base case inflow-outflow diagram for Haggis-3 below. Sketch on the diagram how
the base case curves would alter if the well head pressure were decreased.
Diagram 4: Haggis-3 base case
[2%]
b) Evaluate the benefits of lowering the Xmas tree pressure, in terms of the maximum water-cut you
can economically produce Haggis-3 with after the Xmas tree pressure is lowered to 100 psia.
If the production facilities on the Haggis platform can be modified to allow the Xmas tree
pressure of Haggis-3 to be lowered to 100 Psia, then the maximum water-cut at which the well
will produce economically will become _____%. (See Figure 17)
Scenario
Maximum Economic
Water Cut
Production Rate @ 30%
Water Cut
Lowering Xmas
Tree Press
(Use this result to compare the benefits from this project to the other proposed projects, and as a
basis for your recommendations in section D).
c) List 2 advantages and 2 disadvantages of this scheme.
Advantages
1)____________________________________________________________________________
______________________________________________________________________________
2)____________________________________________________________________________
______________________________________________________________________________
and disadvantages
3)____________________________________________________________________________
______________________________________________________________________________
4)____________________________________________________________________________
_____________________________________________________________________________
[5%]
The Haggis Field
D) Recommendations to Management:
a) Assess the production enhancement projects proposed by the Haggis field engineers in section C
above and compare them to one another and to the base case scenario. Bear in mind that Big
Kahuna Inc. has set the ranking criteria for these projects to be the maximum water-cut at which
Haggis-3 can sustain economic production (i.e. > 1500 STB oil/d).
Table 8 below summarises the results from the various simulations carried out on the Haggis-3
well.
Scenario
Maximum Economic
Water Cut
Production Rate @ 30%
Water Cut
Base Case
Acidising
75° Deviated Well
Optimum Horizontal
Well
Optimum ESP
Optimum Gas Lift
Lowering Xmas Tree
Pressure to 100 psia
Table 8: Haggis-3 Production Forecast
[2.5%]
b) Recommend to management a plan of action which either recommends maintaining the base case
scenario or executes one of the proposed projects.
Based on the WellFlo simulations and my assessment of them, I recommend that Big Kahuna Inc.
invest in a water injection scheme to maintain Haggis’s reservoir pressure at 2800 psia. In
addition, I recommend that Big Kahuna adopts [the _______________________ project /none of
the projects investigated above] because:
_______________________________________________________________________________
______________________________________________________________________________.
[5%]
3) A number of risks have been overlooked in this assessment since the maximum water cut at which
the wells will flow at an economic rate has been used as the ranking criteria for the above projects.
These risks add to the uncertainty of achieving the results on which your recommendation was
based. As the Haggis field team leader it is your duty to report and account for these risks to
management.
Complete Table 9 below which should identify three major risks that have been overlooked by
this assessment. Briefly explain how each one could add to the uncertainty of the assessment and
prescribe steps that need to be taken to account for their effects.
The Haggis Field
Risk/Uncertainty
How this adds uncertainty to
above assessment?
Steps that can be taken to account for
/ minimise this uncertainty
1)
2)
3)
Table 9: Haggis-3 Production Forecast
[9%]
The Haggis Field
Figures:
The Haggis Field
The Haggis Field
The Haggis Field
Figure 3: Sensitivity to flow correlation.
The Haggis Field
Figure 4a: Layer pressure =2800 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 4b: Layer pressure = 2700 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 4c: Layer pressure = 2600 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 4d: Layer pressure = 2500 psia, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 5: Layer pressure = 2800 psia, maximum economic water cut.
Operating point region
The Haggis Field
Plugging off policies
Figure 6a: Plug off at 30% water cut, sensitivity to open interval.
Operating point region
The Haggis Field
Figure 6b: Plug off at 40% water cut, sensitivity to open interval.
Operating point region
The Haggis Field
Figure 6c: Plug off at 48% water cut, sensitivity to open interval.
Operating point region
The Haggis Field
Figure 7: Haggis-3 after acidising (e.g. damage zone permeability 200mD),
sensitivity to water cut.
Operating point region
The Haggis Field
Figure 8: 400ft horizontal well, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 9: 75° deviated well, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 10: Horizontal well, sensitivity to effective length.
Operating point region
The Haggis Field
Figure 11: Optimum horizontal well, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 12a: Performance curves for ESP GC8200 and HC7000
Figure 12b: Performance curves for ESP HC9000 and KC12000
The Haggis Field
Figure 13a: GC 8200, sensitivity to reservoir pressure
Operating point region
The Haggis Field
Figure 13b: HC 7000, sensitivity to reservoir pressure.
Operating point region
The Haggis Field
Figure 13c: HC 9000, sensitivity to reservoir pressure.
Operating point region
The Haggis Field
Figure 13d: KC 12000, sensitivity to reservoir pressure.
Operating point region
The Haggis Field
Figure 14: Optimum ESP, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 15: Performance analysis of the gas lift design, sensitivity to reservoir pressure and gas
injection rate.
The Haggis Field
Figure 16: Optimum gas lift design, sensitivity to water cut.
Operating point region
The Haggis Field
Figure 17: 100psia well head pressure, sensitivity to water cut.
Operating point region
Course:- 28117
Class:- 289033b
HERIOT-WATT UNIVERSITY
DEPARTMENT OF PETROLEUM ENGINEERING
Examination for the Degree of
MEng in Petroleum Engineering
Production Technology 2
Thursday 2X April 200X
09.30 - 11.45
NOTES FOR CANDIDATES
1.
This is a Closed Book Examination.
2.
15 minutes reading time is provided from 09.15 - 09.30.
3.
Examination Papers will be marked anonymously. See separate instructions for
completion of Script Book front covers and attachment of loose pages. Do not write your
name on any loose pages which are submitted as part of your answer.
4.
This Paper consists of 1 Section:Attempt 3 numbered Questions from 4
5.
Marks for Questions and parts are indicated in brackets
6.
This Examination represents 70% of the Class assessment.
7.
State clearly any assumptions used and intermediate calculations made in numerical
questions. No marks can be given for an incorrect answer if the method of calculation is
not presented.
1.
(a) Sketch the main components of a 3 phase (gas/oil/water) horizontal separator and briefly
(one sentence) explain the function of each of the main components.
[8]
(b) Indicate how the export of the oil/water/gas flows are controlled and why the outlets are
situated at your indicated locations.
[3]
(c) Stokes Law (below) describes the velocity of separation (v) of one liquid from another
kD 2 (ρd − ρc )
V=
µc
Where D is the droplet size, ρ the density, µ the viscosity and c and d refer to the continuous and
discontinuous phases respectively. An oil/water 2-phase separator has been in use in a field for
many years. The main producing zone (35˚ API, saline formation water) is now depleted and it
is proposed to produce a shallower, subsidiary zone (17˚ API oil, fresh formation water). The
required data are given in Table 1.
You are required to advise management as to whether the existing separator capacity is sufficient
when the subsidiary zone is producing at 1% and 75% water cut.
Table 1.
Fluid properties at
separator conditions
Oil
viscosity cp
density/g.cm-3
Water viscosity cp
density/g.cm-3
Production zone
Main Sand Subsidiary Sand
2.5
40
0.85
0.95
0.9
0.7
1.1
0.99
N.B. The production rate from the subsidiary zone is only 10% of that achieved from the main
zone.
[9]
(d) An assumption has to be made in the above calculations. Indicate its impact on the
conclusion reached in the unfavourable (separator capacity insufficient) case and indicate
two remedial actions that could be taken.
[5]
2.
(a) You are the Production Technologist responsible for completion of a well in a new field.
Briefly list what techniques you would use to help you in the decision as to whether sand
control measures need to be installed.
N.B. A core has been taken across the pay zone.
[8]
(b) This field has been declared marginal and can only be economically developed with
subsea wells. Briefly describe how this will affect your decision:
(i) on the need for the installation of sand control measures and
(ii) type of sand control measures installed.
[5]
(c) The field is developed with an oil well producing through a gravel pack. The (Darcy)
skin due to presence of the gravel pack and the resulting pressure drop (∆Ps) may be
calculated from:
S=
(
)
96 k / k g L
2
d n
and
∆Ps =
Dq
141.2 qBµ 
S + 4 2  or

KL
d n 
Dq
∆Ps = 0.00539 q  S + 4 2 

d n 
(see Table 2 for definition of the parameters and numerical values)
Calculate the (Darcy) skin value (S) and the resulting pressure drop for a perforation density of 4
shots/ft.
[4]
This is the target, allowable pressure drop in the well.
(d) Well testing found that the turbulent (non-Darcy) resulted in an unacceptably high
pressure drop of 374 psi. You are required to advise management as to whether the next
well should be completed with:
Case
A
B
Cost
Low
High
Shot Density Diameter
12 shots/ft
0.5 in
4 shots/ft
1.0 in
and whether it will meet the target, allowable pressure drop.
[5]
(e) Briefly comment on which case you would have expected to give the better inflow, and
why.
[3]
Table 2
Well Production (q)
Total Production Height (h)
Reservoir Permeability (k)
Oil Viscosity (µo)
Formation Volume Factor (Bo)
20-40 Mesh Gravel Permeability
Perforation Penetration (L)
Perforation Diameter (d)
Perforation Density (n)
Non-Darcy (turbulance factor) (D)
2500 STB/D
23 ft
578 mD
0.310 cp
1.636 bbl/STB
120,000 mD
6 in
0.5 in
4 shots/ft
0.01
3.
(a) List up to 6 key features for both Rod Pumps and Gas Lift that form the basis of the fol
lowing statement:
“Worldwide, 85% of Artificial Lift equipment installed is rod pumps. This is mainly in strip
per wells while gas lift is the most popular artificial lift technique for higher rate wells”.
[6]
(b) Most gas lift fields have insufficient gas to lift all the wells at their (technical) maximum
production. Briefly describe the process of optimal allocation of available lift gas;
mentioning the key economic parameters involved.
[6]
(c) Design a gas lift installation for the following conditions:
Tubing
Required Production Rate
Oil Cut
Gas Oil Ratio
Gas Specific Gravity
Average Flowing Temperature
Reservoir Productivity Index
Reservoir Depth
Reservoir Pressure
Lift Gas Injection Gradient
Minimum flowing tubing head pressure to
transfer fluids to facility
Dead Oil Density
Gas Oil Ratio
Brine Density
Lift Gas Injection Rate
3.958 in
3000 STB/day
100%
100scf/bbl
0.65
150˚F
4 bpd/psi
10,000 ft
3400 psi
20 psi/1000 ft
250 psi
35˚ API or 0.368 psi/ft
100 scf/bbl
0.44 psi/ft
3,000,000 scf/d
A pressure traverse curve is provided as Figure 1.
Tubing size, in. : 3.958
Liquid rate, STBL/D : 3000
Water fraction : 0
Gas gravity : 0.65
Oil API gravity : 35
Water specific gravity : 1.07
Average flowing temp, F. : 150
Assume that the well is closed in with dead oil in the tubing and brine in the casing/tubing
annulus.
(i) does this well require artificial lift to produce?
[2]
(ii) what depth should the gas lift valve be installed in a single valve lift installation in order
to achieve the required production?
[6]
HINT: Note that the relevant portions of the pressure traverse curve can be approximated by
straight lines.
(iii) what is the minimum surface gas injection pressure to kick the well off in the
configuration described?
[4]
(iv) how does this change if dead crude oil was present in the casing/tubing annulus instead
of brine?
[1]
4.
(a) Briefly contrast the generalised selection criteria for matrix acidising and fracturing
treatments when considering carrying out a stimulation treatment on a well.
[5]
(b) List 2 sources of formation damage encountered during drilling and completion
operations and 3 damage sources during production operations. Briefly indicate how the
fluid selection for a (matrix) removal treatment will be influenced by the damage source
(examples may clarify your answer).
[6]
(c) A well completed on 40 acre spacing (re = 745 ft) has a damaged region extending 1 ft
beyond the wellbore (rw = 0.328 ft).
The Hawkins formula may be used to calculate the skin due to formation damage:
k
r
Sd =  o − 1 d
 kd
 rw
while the productivity ratio (Ji/Jd) of the well with and without the above formation damage is
given by:
 re 
Ji In rw  + S
=
r
Jd
In e r 
w
Use the above to illustrate the statement:
{“Formation Damage reduces well productivity greatly while the stimulation effect of
increasing the near wellbore permeability above the initial value has limited effect”.}
HINT : estimate the relative well productivity with 95%, 75%, 50% formation damage and 10
times increase in near wellbore formation permeability.
[6]
(d) Your service company has designed the following fracturing treatments:
Wellbore radius (rw):
Reservoir height:
Reservoir Permeability:
Proppant available:
0.328 ft
100 ft; bounded by competent shales
0.1 mD
300,000 lb
Design Fracture Conductivity (kf*w): treatment A
- 1500 mD.ft at 4 lb/ft2 proppant loading
treatment B
- 850 mD.ft at 2 lb/ft2 proppant loading
(i) Use the accompanying graph (Figure 2) from Cinco-Ley and Samiengo to advise
management as to whether treatment A or B will give the highest well productivities.
[6]
(ii) Why would you expect one of these treatments to be preferred?
[2]
End of Paper
Model Solutions to Examination
N
M
A
U
N
O
E:
R:
N
G
SI
TI
A
TR
IS
RS
E:
G
RE
CO
EA
Y
RE
.:
O
TU
N
A
8 Pages
:
n
is
ee
tio
th
ec
il
nt
ss
hi
u
al
et
se
et
pl
m
Co
bu
t
n
io
no
at
o
td
ed
in
ish
m
fin
xa
Date:
Subject:
Production Technology 2
INSTRUCTIONS TO CANDIDATES
No. Mk.
1. Complete the sections above but do not seal until the examination is finished.
2. Insert in box on right the numbers of the questions attempted.
3. Start each question on a new page.
4. Rough working should be confined to left hand pages.
5. This book must be handed in entire with the top corner sealed.
6. Additional books must bear the name of the candidate, be sealed and be affixed to
the first book by means of a tag provided
PLEASE READ EXAMINATION REGULATIONS ON BACK COVER
1
2
Model Solutions to Examination
Answer to Q1
1(a)
(Standard) 3-phase horizontal separator diagram.
Pressure Control Valve
Inlet Deflector/
Momentum
Breaker
PC
Gas
Mist Eliminator
Weir
GAS
Inlet
From
Production
Manifold
OIL and EMULSION
OIL
WATER
To Produced
Water Treatment
Oil Outlet and
Level Control
Water Outlet and
Level Control
To Oil Export
Main components
Main function
Inlet momentum breaker =>
to reduce inlet flow velocity (hence
helping disengage free gas)
Mist eliminator/extractor =>
to remove liquid drops from gas
(by allowing liquid droplets to impinge
on the wire, coalesce and flow down in
to the liquid phase)
Weir =>
to separate oil and water phases (oil
is collected here and skimmed off
from main settling area)
Liquid outlets under level controls =>
to evacuate oil and water (usually
equipped with a vortex breaker to
prevent re-entrainment of gas)
3
Bonus marks
-
coalescer pack
-
solids wash facility
-
production chemical addition
-
fresh water wash
1(b)
In a gravity type separator => the separation process takes
place due to the action of gravitycombined with the difference in
densities of the phases/components of the produced fluid.
For a 3-phase separator (vertical/horizontal) the heavier fluids (water
and oil) go to the bottom of the vessel while the lighter phase (gas)
rises to the top. Therefore:
* Water outlet is located at the bottom of the vessel, remote from
the inlet to allow maximum residence time. It is controlled by the oil/
water interface level control.
* Oil outlet is also located at the bottom, remote from the inlet to
allow maximum residence time. It is connected to the oil level control.
* Gas will flash off and disengage from the liquid phases and exit from
the uppermost part of the vessel remore from the inlet (to allow maxi
mum residence time). Gas outlet is regulated by a pressure control
valve which controls the separator operating pressure.
4
Model Solutions to Examination
1(c)
Assume droplet size distribution the same for subsiduary and
main production zones.
V1 ( ρd 1 − ρc1 ) µ c 2
=
.
V2 ( ρd 2 − ρc 2 ) µ c1
75% water cut
• Water is the continuous phase
• Oil is the discontinuous phase => i.e. oil droplets are separating from
the water phase.
Vmain
Vsubsiduary
=
=
( ρom − ρwm ) . µ ws
( ρos − ρws ) µ wm
(0.85 − 1.1) 0.7 0.25 × 0.7
.
=
= 4.9
(0.95 − 0.99) 0.9 0.04 × 0.9
Separation of oil droplets from the water phase takes 5 times longer
for the more viscous oil. Under these conditions, an increase of
residence time of at least 5 times is required in order to achieve the
same efficiency of separation.
Since well production is reduced by factor 10, residence time is in
creased by factor 10 (i.e. 10 times). Therefore separator capacity is
sufficient and does not need upgrading.
1% water cut
• Oil is the continuous phase
• Water is the discontinuous phase => i.e. water droplets are separating
from the oil phase.
5
Vmain
Vsubsiduary
=
=
( ρwm − ρom ) . µ os
( ρws − ρos ) µ om
(1.1 − 0.85) 4.0 0.25 × 40
.
=
= 100
(0.99 − 0.95) 2.5 0.04 × 2.5
Separation of water droplets from the oil phase takes 100 times
longer, i.e would require an increase of residence time of at least 100
times to carry on with the separation process
Despite production rate being reduced by factor 10, separation
capacity is still factor 10 too small. Therefore separation capacity is
insufficient and needs upgrading.
1(d) Separation rate increases as the square of the droplet
diameter - therefore an increase in the water droplet size by a factor
3 would be sufficient to compensate for above factor 10.
Possible remedial measures are:
• installation of a coalescer pack to reduce the distance oil and water
droplets have to separate
• add demulsifier chemicals to speed up the rate of droplet
aggregation by coalescing the small oil particles to form larger
particles which will separate faster (Stokes law - the D term is
squared & so the improvement is squared).
6
Model Solutions to Examination
Answer to Q2
2(a) The main techniques used when deciding whether to install sand
control measures are:
• Field Experience
History of any sand production problems in wells producing from the
same formation in the same field, or wells in the same or similar
formation but on other nearby fields, will be the best guide for
installing sand control measures in a new well. Evaluation of the
drawdown conditions and rock stresses under which sand production
started in those wells will be very useful.
Sanding tendencies may be detected by performing high drawdown
tests in exploration wells.
• Petrophysical and core analysis (sonic travel time)
Based on laboratory measurements on core material:
i) The sonic travel time is proportional to the porosity i.e. the higher
the porosity, the longer the travel time.
ii) The rock strength is inversely related to porosity i.e. the lower the
porosity, the greater the rock strength.
Sand production tendency is dependent on rock strength, which
depends on porosity and grain cementation. A measure of porosity and
7
grain cementation can be obtained from the sonic travel time through
the rock matrix. This can be done by acquiring an acoustic wireline log.
The well sonic travel time log can thus be processed to derive a
continuous estimate of the formation strength.
Bonus
Bonus: Shell petrophysical risk approach to sand failure prediction:
• Well site strength estimation (core recovery, onsite core tests and
completion experience)
Simple tests at well site can determine if core is friable (by scratching
it or simply scraping it with the fingernails) and therefore a potential
sand problem may be expected. Other simple measures include
observation of the core recovery status and completion experience e.g.
sand run into the casing when the drill pipe or tubing is out of the hole.
• Rock mechanical measurements and calculations (unconfined
compressive strength, brinnel hardness, thick wall cylinder
strength, triaxial rock strength)
Bonus marks for extra description of the Rock mechanical tests 1
to 4 below:
8
Model Solutions to Examination
1) Unconfined or Uniaxial Compressive Strength - In this test, an
unsupported cylinder of rock is loaded axially at a steady rate to
failure. The maximum stress reached is called the Unconfined or
uniaxial compressive strength (UCS).
2) Brinell Hardness Number - is the load required to press a standard
spherical indenter (metal disk) a constant distance into a slabbed core
face. The strength of rock is proportional to the hardness number i.e.
the stronger the rock, the greater the load required and the higher
the BHN
3) Thick wall cylinder collapse strength - A hollow, thick wall cylinder
is created by drilling a narrow hole in the middle of a rock and placed
in a rubber sleeve. Then, the hollow cylinder is loaded axially and
radially until it collapses {the Thick Wall Collapse Strength ((TWC)}.
4) Triaxial test - Similar to uniaxial test but a radial stress also
applied to the core. The axial stress imposed by the end pieces and the
radial stress are controlled separately. The maximum stress prior to
failure increases as the radial stress increased i.e. the rock sample
shows stronger behaviour as the radial, confining stress increases.
This test gives most information on the rock strength properties.
2(b)
i) In a marginal field any type of workover operations is often
not economic. This means that the decision on whether or not to install
sand control has to be made correctly the first time. Further, if in
9
stalled, the sand control equipment should not impair the well
productivity below the economic limit since a corrective workover may
not be economically viable.
ii) In a Subsea well, “living with sand” is not an option since we are
unable to detect sand production can not manage its presence in subsea
flowlines.
Subsea wells have high intervention costs. Hence the need to minimise
workover requirements by correct, first time design and use of high
reliability equipment since remedial measures may not be affordable.
The two alternatives regarding sand control installation are:
- If sand control is installed but not needed it will probably cause
additional pressure drops and reduce the well Productivity Index.
Consequently well economics will be affected
- If sand control is not installed but is needed an intervention is
likely to be necessary to install sand control and/or repair blocked/
eroded equipment.
The specific case needs to be studied and a risk analysis carried out.
Sand control options include internal & external gravel packs, prepacks,
slotted liners, wire wrapped screens and chemical treatments. Gravel
packs give a higher pressure drop but also provide excellent sand
10
Model Solutions to Examination
control if designed correctly. Slotted liners are relatively cheap but
only stop coarse grained sand effectively. Wire wrapped screens are
more expensive and more effective, provide greater flow area and
lower pressure drop. Pre-packed screens are even more expensive with
good sand control but are susceptible to plugging during installation. It
is advisable to
install a completion with high productivity greater
{e.g. horizontal well with liner} than that required based on
conventional well inflow analysis to allow for lower (e.g. gravel packed)
well productivities often observed in practice.
2(c)
At a production rate of 2500 STB/D and a perforation density
of 4 shots/ft
S=
96 * (578 / 120, 000) * (6)
(0.5)2 * (4)
= 2.77
0.01 * 2500 
∆Ps = 0.00539 * (2500) *  2.77 +

0.54 * 4 2 
∆Ps = 374 psi
2(d)
Case A
S = 0.92
0.01 * 2500 
∆Ps = 0.00539 * (2500) *  0.92 +

0.54 * 12 2 
∆Ps = 49.8psi
11
Case B
S = 0..69
0.01 * 2500 
∆Ps = 0.00539 * 2500 *  0.69 +

14 * 4 2 
∆Ps = 30.4 psi
Case B perforating is recommended.
2(e)
Doubling the perforation diameter increases the inflow area by
a factor 4 while the increase in shot density increases the inflow area
by a factor 3 only. Wide diameter perforations provide a much smaller
pressure drop and also create less non-darcy flow effects (no data
given in the question).
Wide diameter perforations are normally preferred for gravel pack
completions (easier placement of gravel without a screen out).
Answer to Q3
3(a)
Rod Pump
• Typically low rates and moderate depths
• Relatively cheap to install & run
• Rod pumps are mechanically simple to operate and easy to repair/
maintain/replace. Can be operated by inexperienced personnel
• Sensitive to gas and solids (wax/scale/sand) - Solids can damage
moving parts
• Not suitable for (highly) deviated wells
12
Model Solutions to Examination
• Obtrusive in urban locations. Heavy equipment for location offshore
• Pump can be easily changed and performance monitored
• (reasonably) Viscous oil can be handled
Gas Lift
• capable of high production rates
• Suitable for water drive reservoirs with relatively high bottomhole
pressure gradients
• The above, coupled to high well Productivity Indices and high
formation permeabilities result in a high Flowing Bottom Hole Pressure,
limiting achievable reservoir depletion
• Gas has to be available
• Wireline serviceable up to 65˚ deviation
• Flexible - valve settings can be adjusted for optimum performance
based on actual, well conditions
• can be used off-shore
• Fully open tubing giving access for logging
• tubing, and annular surface controlled, subsurface safety valves
available
• Limited surface requirements once gas available
• Forgiving of poor design & operation, but difficult to run efficiently
• Can handle (tolerate) produced solids e.g. reasonable formation sand
concentrations
• High GOR => advantage rather than a drawback
N.B. Only 6 items are required from the above lists
13
3(b)
Optimal allocation of Lift Gas
• Allocate each increment of gas to that well with the highest
incremental oil production until all gas allocated.
• Implies slope of (net oil/lift gas rate) should be the same for each well.
• A brief description of the process is as follows:
- Perform a gas lift tubing performance calculation with a range of
injected gas flow rates for each well in the field. Then, plot the
results on a curve of liquid flow rate achieved versus injected gas rate.
Each well has its own characteristics (geometry and flow capacity) and
will show a different rate of oil production increase with incremental
increase on the gas lift injection rate.
- Calculate the incremental oil production rate in each well for each
extra increment of gas injected.
- Allocate each increment of gas to that well with the highest
incremental oil production until all gas allocated.
- Implies slope of (net oil/lift gas rate) should be the same for each well.
- This is a complex process that has many variables such as gas
availability, number of wells, etc, and even becomes even more
complicated by those wells that require kick-off lift gas.
- There is software available to assist with this analysis.
Bonus points
•
Check economic limit, when cost of incremental lift gas equals
income from incremental oil not exceeded
•
Chosen rates should be sufficiently far from technical maximum
production (where extra gas decreases production rate) to avoid
14
Model Solutions to Examination
unstable well operation.
3(c)
(i)
Dead oil in tubing height 3,400 / 0.368 = 9,239 ft
Fluid level = 10,000 - 9,239 = 761 ft
Well is dead with fluid level at 761 ft
(ii) GOR after gas injection = 3,000,000 / 3,000 + 100 = 1,100 SCF/STB
From pressure traverse plot:
- Average flowing gradient after lift gas injection:
between 6,000 ft and surface = 600 / 6,000 = 0.1 psi/ft
- Average flowing gradient prior to lift gas injection
from 10,000 ft to 4,000 ft = (3,040-900)/(10,000-4,000) = 0.357 psi/ft
FWP + ∆ P above valve + ∆ P below valve + Draw down = Reservoir Pressure
Drawdown = Production Rate/PI = 3,000 / 4 = 750 psi
Let depth of valve = x
250 + 0.1 * x + (10, 000 − x ) * 0.357 + 750 = 3, 400
0.1x − 0.357 x = 3, 400 − 1, 000 − 3, 570
x = 1,170 / 0.257
x = 4, 550 ft
15
Install gas lift valve at 4,550 ft.
(iii)
Minimum surface gas injection pressure (P)
P + Gas gradient = Hydrostatic head at 4,550 ft + FWHP
P + (4,550) 0.02 = (4,550) 0.44 + 250
P = 2,252 - 91
P = 2,161 psi
(iv) Required injection gas pressure decreases by the ratio of the
brine and dead oil fluid densities.
If dead crude was in the annulus
P1 + 91 = (4,550) 0.368 + 250
P1 = 1,833 psi
Answer to Q4
4a)
Matrix (Acidising) - removal of near wellbore damage
Hydraulic Fracturing - improving well inflow performance
16
Model Solutions to Examination
Skin
Permeability
Treatment
High
High
Matrix treatment
High
Medium
Matrix/frac and pack
High
Low
Fracturing (matrix possible)
Low
Low
Hydraulic fracturing
Low
High
Treatment economic?
Stimulation treatment should increase well Productivity Index:
Matrix acidising aims to increase the Productivity Index (PI) by
reducing the Skin (S) through dissolving formation damage components
and rock in the near wellbore region. It is particularly suitable for
medium/high perm reservoirs with high skin.
Fracturing increases PI by increasing effective wellbore radius throgh
the creation of high conductivity fractures from wellbore. Fractures
bypass the damaged zone and extend to a greater depth into the
reservoir than can be reached by acidising. It is suitable for low
permeability reservoirs with or without skin.
4b)
Drilling and Completion Fluids
• solid block pore throats => mud solids invasion leading to blocked pore
throats => reduce permeability
• solids have many sources => drilling mud components, drilled
formation particles etc.
• Fluid loss increases liquid saturation
17
• incompatible formation and fluid loss reduces invaded zone
permeability
• Mud filtrate invasion => clay swelling => reduced permeability
• Water block => from increasing water saturation near wellbore due to
loss of drilling / completion fluid
Production Operations:
• Scale => due to temperature and pressure changes. Salts in
formation water become over saturated and precipitate e.g. calcium
carbonate.
• Wax
• Asphaltenes => Asphaltene precipitation due to pressure drop.
• Fines migration
• Sand production due to increased effective stress
Matrix treatment fluid should dissolve (or at least mobilise) the source
of the formation damage, e.g.
wax
-
hot, organic fluids
calcium carbonate scale
-
hydrochloric acid
clay particles
-
mud acid
bonus marks
marks: choice of fluid has to maintain the materials in solution
and avoid later reprecipitation of the dissolved materials.
18
Model Solutions to Examination
4c)
 k
 r 
S=
− 1 ln d 
 kd
  rw 
rw = 0.328, rD = 1.328 ln rd / rw = 1.40
re
J i ln rw + S
=
r
Jo
ln e r
w
rw = 0.328, re = 745 ln re / rw = 7.73
J i 7.73 + S
=
Jo
7.73
Damage removal
K/kd
S
Ji/Jo
Jo/Ji
20
4
2
1
26.6
4.2
1.4
0
4.44
1.54
1.18
1.00
0.23
0.65
0.85
1.00
Stimulation
K/kd
S
Ji/Jo
Jo/Ji
0.5
0.1
-0.70
-1.29
0.91
0.84
1.10
1.20
The above tables reflect the statement “Formation damage reduces
well productivity greatly while the stimulation effect of increasing the
near wellbore permeability above the intial value has limited effect.”
19
4d) i)
Treatment A:
Fracture length (both wings) =
Proppant
loading * frac. height
= 300,000 / (4*100) = 750 ft
Fracture half length (Xf) = 375 ft
FCD = Kf*w/k*Xf = 1,500 / (0.1*375) = 40
From graph
ln (Xf/rw) + Sf = 0.75
∴
Sf = 0.75 - ln (375/0.328)
Sf = - 6.29
Treatment B:
Fracture half length = 300,000 / (2*100*2) = 750 ft
FCD = 850/(0.1*750) = 11.3
From graph
Sf
= 0.8 - ln (750/0.328)
= - 6.93
∴ Treatment B achieves a more negative skin and is recommended.
20
Model Solutions to Examination
ii) Long fractures are more effective at stimulating low permeability
wells providing the fracture conductivity is sufficient (FCD > 10 - 15).
21
22
Course:- 28117
Class:- 289033a
HERIOT-WATT UNIVERSITY
DEPARTMENT OF PETROLEUM ENGINEERING
Examination for the Degree of
MEng in Petroleum Engineering
Production Technology 1b
Friday 23rd April 1999
09.30 - 12.30
NOTES FOR CANDIDATES
1.
This is a Closed Book Examination.
2.
15 minutes reading time is provided from 09.15 - 09.30.
3.
Examination Papers will be marked anonymously. See separate instructions for completion of
Script Book front covers and attachment of loose pages. Do not write your name on any loose
pages which are submitted as part of your answer.
4.
This Paper consists of 2 Sections:- A and B.
5.
Section A & B:- Attempt 4 numbered Questions from 7 with at least 1 Question from each
Section
6.
Marks for Questions and parts are indicated in brackets
7.
This Examination represents 55% of the Class assessment.
8
State clearly any assumptions used and intermediate calculations made in numerical questions.
No marks can be given for an incorrect answer if the method of calculation is not presented.
9.
Answers must be written in separate, coloured books as follows:Section A:Section B:-
Blue
Green
SECTION A
A1. “Advanced wells and in particular horizontal and multi-lateral wells, can enhance the business
case of a field development by any of 3 primary techno-economic drivers.” What are these?
[3]
“Multi-lateral well configurations can in the main be classified as stacked, opposed or planar, but
the selection of the optimum geometry must be based on the reservoir structure and flow charac
teristics.” Discuss this statement giving examples to illustrate the application of the various
options - use sketches as appropriate.
[7]
A2. Two subsea well completion designs are shown in Figures 1 and 2 - review and compare each of
these designs for the following applications:
(a)
10,000ft T.V.D. oil producer, normally pressured with a GOR of 400scf/bbl.
[5]
(b)
Water injection completion for the same reservoir.
[5]
A3. The well shown in Figure 3 is an overpressured oil producer. If pressure buildup is experienced
in the 103/4” x 7” annulus at surface:
(a)
What are the potential sources of the pressure?
[3]
(b)
What method(s) and tools could be used to identify the cause of the leakage(s)?
[4]
(c)
What corrective measures would you propose for the causes identified in (b) above?
[3]
Typical Producer
Completion Schematic
Annulus
check
valve
Typical Injection
Completion Schematic
Annulus
check
valve
Tubing hanger
4 1/2" TRCHSV
3.812" RQ' Landing nipple
injection valve
4 1/2" 12,6 lb/ft VAM J55 Tubing
4 1/2 " 2.6lb/ft new vam tubing
4 1/2" Alloy 'MMG' SPM
4 1/2 " alloy 'MMG' SPM
4 1/2 " 12.5 lb/ftvam tubing
3.812' 'XN' No-Go lancing nipple
Seal unit
3.912" 'XN' No-Go landing nipple
Seal unit
4 1.2"PBR
4 1/2" PBR
Tubing anchor latch
9 5/8" 40lb/ft NBS vam casing
Tubing anchor latch
9 5/8' SAB-3 HYD set
permanent packer
9 5/8" SAB-3' HYD set permanent
packer O/W millout extension
Millout extension
3.588" RN' No-Go landing nipple
2.750" XN nipple
3.1 2/9' /AM perforated at
2.756" XN' No-Go landing nipple
2.313' XN' No-Go landing nipple
2.313" XN' No-Go landing nipple
Wireline re-entry guild
Wireline re-entry guide
PERFS
PERFS
PBTD
Figure 1 and 2
PBTD
7" TR SSV
10 3/4" tie -back
Liner top isolation packer
Liner hanger
13 3/8" shoe @ 5700'
9 5/8" liner
7" CRA tubing
Full bore nipple profile
A-Ryte sealassemble
w/ anochor latch
Liner hanger w/lower swal bore
9 5/8" shoe @10700'
7" CRA production liner
Perforations
Liner top isola