ANALYSIS OF API 5C3 FAILURE PREDECTION FORMULAE FOR
RIALS FOR
CASING & TUBING
EP OIL AND GAS WELL
NSTRUCTION
S. Staelens1, T. Galle1, W. De Waele1, P. De Baets1
1
Ghent University, Laboratory Soete, Belgium
Abstract Due to the increasing demand for oil and gas, coupled with the fact that oil reserves are becoming
rather scarce, the petroleum industry is pushed to drill and complete deeper wells. Threaded connections
are often the weakest link in this process and are therefore the subject of research and optimization.
terials for deep well
At first, this paper presents a brief overview of the design characteristics of today’s premium connections.
n must haveSecondly,
high the failure mechanisms of Oil Country Tubular Goods (OCTG) are discussed. At last, an in-depth
analysis of theatAPI 5C3 formulae is given. Four formulas for collapse are given in API 5C3, each valid for a
d corrosion resistance
specific D/t range. With increasing yield strength of the steel, the difference between the yield strength
atures.
collapse and the plastic collapse gets larger. Also, the elastic collapse zone gets bigger, so stronger
*
materials with relatively large D/t ratios will collapse in the elastic zone instead of the plastic or transition
zone. These four formulas can be approached by a third-order polynomial equation that is valid for all D/t
ratios.
o
Keywords: threaded and coupled; OCTG; buttress thread; premium connection; failure mechanisms
strates some of these successful
ns, in addition to more recent
Texas and the Gulf of Mexico.
as had to pursue ever more hosthan the original HPHT limits
o keep up with demand. In this
ns exceeding 400°F (204°C) and
MPa) at bottomhole have been
sly as Extreme HPHT (xHPHT)
HT. (The term HPHT is used
s article to include both extreme
Reservoir temperature, °F (°C)
1 demand
INTRODUCTION
asing worldwide
for oil
s coupled with the fact that the
and
completion of High Pressure/High Temperature (HPHT) wells are common practice. Some
oil productionDrilling
has been
reached
conditions
of
field examples are illustrated in Figure 1. Well depths up to 12000m are no longer unusual.
will be, has pushed the petroleum
Because
of
this
increasing depth, the pressure and temperature also increases. Clearly, the applied casing
drilling ever deeper wells. Well
and
tubing
have
0 ft (7620 m) and greater are noto be able to withstand extreme conditions. Failure of these OCTG can cost millions of
dollars, can
cause
al, and even deeper
wells
are environmental disasters and hence are to be avoided.
Because of the harsh conditions associated with HPHT wells, traditional OCTG threaded connections can
creasing depthno
means
increasing
longer
guarantee safety and premium connections are required [1]. Such connections have a better leak
mperature. High-pressure/highresistance and an overall
improved
integrity.
Drilling
rig at sunset.
PHT) wells have generally been
s in which temperatures and presInitial reservoir pressure, MPa
om of the well exceed 300 to 350°F
69
83
97
110
124
138
525
and 10,000 psi (69 MPa), respec(274)
HPHT wells have already been
pleted in this category with great
475
(246) Madden
unusual requirements for special
East Texas
Deep
425
(219)
!
!
!
Mobile Bay
Mary Ann
375
(190)
325
(163)
10
Erskine
!
!
Embla
12
!
Shearwater
!
Franklin
Thomasville
! !
! Elgin
Villa/Trecate
!
New Gulf of
Mexico
Fields
! Malossa
14
16
18
Initial reservoir pressure, ksi
20
Fig.
1 — Some
of the prominent
HPHT fields
in thewells
world.in the world [2]
Figure
1: Some
field examples
of HPHT
e challenges will be for both the Updated from Ref. 1.
The design
of premium
connections
differs
from the standard API connections in various ways. At first,
Liners,
tubing, and
strings
selves and materials
engineers
metal-to-metal
(MTM) seals
are
used to
guarantee
sealability.forIn addition, a torque shoulder can be
Figure
2 shows
a typical
wellthe
completion
m industry for the
future. Drilling
used to withstand
high unfamiliar
torques required
in HPHT
wells.
At last, the standard hooked or buttress thread
with the
industry
terminology.
ells requires specialized
methodsthethose
can be have
modified
in numerous
withstand
higher loads and conditions of HPHT wells [3].
Surface
casingways
and to
some
of thethe
intermediate
e planning, butshape
the materials
ed steel drill pipe and steel com- casing strings are not affected by HPHT condigh other alloys such as titanium tions, and thus standard steel tubulars function
well.
dered.
The major components that require greater ate real materials challenges are in
d producing the wells after they tention and represent materials challenges are the
s is the reason this article focuses liner at the bottom of the well, the tieback casing
string, the tubing (and associated jewelry), and
ction rather than drilling.
These design features and their effects are briefly discussed in the following section. More detailed
information can be found in [4, 5]. In the last section of this paper, the failure mechanisms of traditional
OCTG are summarized and an analysis of the formulae in API 5C3 is conducted.
2
DESIGN FEATURES OF PREMIUM CONNECTIONS
A typical premium connection is illustrated in Figure 2. It shows the three main features that can be
changed for obtaining an optimal design: the thread profile, the torque shoulder and the MTM-sealing area.
Though only these are discussed in the following paragraphs, it should be mentioned that the design of
premium connections is not limited to the above-mentioned three aspects. Various other improvements can
and are made to the premium connections of today such as applying a lubricant relief groove and/or stress
relief grooves, the use of “dopeless technology”, ….
SPE
97585
SPE
97585
Wedge effect
Figure 2: Typical premium connection (HES Seal-Lock Apex) [6]
Flank
toto
Flank
toto
Root
Flank
Flank Crest
Crest
Root
SPE
97585
SPE
97585
Figure
1. 1.
Thread
iInterferences
Figure
Thread
iInterferences
2.1
Thread design
The commonly used thread profile of premium connections is trapezoidal. Changes can be made to the
shape of the threads, the pitch (e.g. FOX, JFE [7]) and the interference between the thread of the pin and
the box (e.g. VAM21, VAM Services [8]). Each has its own effects.
Figure
5. 5.
Res
Figure
Re
Sample
Sample
For instance, when changing the shape of the threads, the following geometries are possible (Figure 3):
-
1 1
90°90°
90°90°
Positive load flank + positive stab flank (PLPS)
2 2
Neutral load flank + positive stab flank (0LPS)
Flank
to
Flank
Crest
to
Root
Crest
to Root
Negative load Flank
flankto+Flank
positive
stab
flank (NLPS)
Figure
1. 1.
Thread
iInterferences
Figure
Thread
iInterferences
Figure
5. 5.
Resilient
seal
in in
threaded
and
coupled
connection.
Figure
Resilient
seal
threaded
and
coupled
connection.
Negative load
flank
+ negative
stab flank (NLNS) Positive
Figure
6. 6.
ISO
Figure
IS
Load
Flank
Zero
Load
Flank
Positive
Load
Flank
Zero
Load
Flank
Load flank
Stab flank
90°90°
90°90°
Positive
Load
Flank
Positive
Load
Flank
PLPS
Zero
Load
Flank
Zero
Load
Flank
0LPS
90°90°
7.97"
1.123"-wall
P110
VAM
PRO
7.97"
1.123"-wall
P110
VAM
PRORSRS
E
IP IP
Seal
Seal
Sample
A Series
AS
C Series
Sample
A Series B Series
B Series
A
C Series
Tested
Tested
(psi)
(psi)
(psi)
(psi)
(psi)
(psi) (deg-F)
(deg-F) (p
MTM
- 25000
425
15I
MTM 27500
27500
25000
425 Inte
1 1
25000
425
15
RSRS
25000
- 25000
425
25000
90°90°
25000
425
15
MTM
- 25000 Figure
425
MTM 27500
27500
7. 7.
Typ
Figure
Ty
2 2
RSRS
25000
- 25000
425
15
25000
25000
425
Figure
6.Load
ISO
testing
with
metal
and
resilient
seals.
Negative
Load
Flank
and
Stab
Flanks
Negative
Load
Flank Negative
Negative
Load
and
Stab
Flanks
Figure
6.
ISO
testing
with
metal
and
resilient
seals.
NLNS
Figure
2. NLPS
forms
Figure
2.Thread
Thread
forms
Figure 3: Thread shape design possibilities [9]
Expanded
BoxBox
Expanded
External
Metal
SealSeal
External
Metal
Figure
8. 8.
VAM
Figure
VA
connection.
connection
easy to fabricate and repair, but isn’t often used in
Typical
SemiFlush
(SF)
Connection
Typical
SemiFlush
(SF)
Connection
with positive stab
flank
(0LPS)
diminishes
hoop
Figure
7. 7.
Typical
expanded
box
semi-flush
connection.
Figure
Typical
expanded
box
semi-flush
connection.
Internal
Metal
SealSeal
Internal
Metal
A combination of positive
load and stab flank (PLPS) is
90°90°
90°90°
casing and tubing applications. A neutral
load flank
stresses in the connection and is used for instance in the Hunting Energy Services (HES) FJ-150
connection [10]. Next, a negative load flank with positive stab flank (NLPS) reduces outward radial forces.
Negative
Load
Flank
and
Stab
Also, the connection’s
yield
strength
will Load
be
equal
toFlanks
or
greater than that of the pipe body [9]. However, the
Negative
Load
Flank Negative
Negative
Load
and
Stab
Flanks
pipe
axis
Parallel
toto
thread
cone
Parallel
to
pipe
axis
Parallel
thread
cone
Figure
2. 2.Thread
forms
Figure
Thread
forms the thread, so it isParallel
negative load flank
will
“undermine”
less3.tosuited
for
applications
where
high axial forces
Figure
– root
designs
Figure
3.Crest
Crest
– root
designs
Optional
R
Optional
Seal
Seal
Seal
Metal
occur. Examples of connections with such thread are numerous [8, 10, 11]: HES Seal-lock Apex, KSBEAR, Metal
9. 9.
VAM
Figure
VA
VAM21, TenarisHydril HW… At last, a negative stab flank can also be used. When combined with a Figure
connection
w
connection
negative load flank (NLNS or “dovetail”), the connection can withstand high
torque
loads
[9].
Examples
are
Figure
8. 8.
VAM
PRO
- Pin
nose
toto
pin
nose
with
radial
metal
seal
Figure
VAM
PRO
- Pin
nose
pin
nose
with
radial
metal
sea
connection.
connection.
the Tenaris Wedge 553 and the XL Systems – National Oilwell Varco XLF connections [10].
Secondary
Metal
Seal
Secondary
Metal
Seal
Parallel
toto
pipe
axis
Parallel
toto
thread
cone
Parallel
pipe
axis
Parallel
thread
cone
Figure
3. 3.Crest
– root
designs
Optional
Resilient
Figure
Crest
– root
designs
Optional
Resilient
Figure
4. 4.10!”
x 101
lb/ft
VM110SS
VAM
HPHP
SC80
Threaded
and
Figure
10!”
x 101
lb/ft
VM110SS
VAM
SC80
Threaded
and
2.2
Shoulder design
When applying a torque shoulder (Figure 2), a higher torque load can be applied and the total applied
torque load can be controlled better. Therefore, the risk of galling and over-torque reduces. In addition,
make-up gets easier and the sealability of the connection can be better guaranteed [4, 12].
A connection with a “reversed torque shoulder angle”, as seen in Figure 2, causes a “wedge effect”. This
improves the self-alignment, radial stability and structural strength of the connection [4].
2.3
Metal-to-metal sealing design
The last and essential design characteristic of premium connections is the radial metal-to-metal seal. They
are currently the most effective pressure seal and are located above the torque shoulder. The following two
types exist (Figure 4):
Cone-to-cone MTM seal: both MTM surfaces of pin and box have a conical shape;
Sphere-to-cone MTM seal: the MTM surface of the pin has a toroidal shape, whilst the MTM
surface of the box is conical.
-
Structural Analysis
Figure 4: Schematic diagram of a cone-to-cone MTM seal (left) and a sphere-to-cone MTM seal (right)
'%###
:1.3204/-.0-2;61,,<6127=>.9
If a cone-to-cone
MTM seal is Connections
used, overall, a smaller seal contact pressure will be applied =?2@A4/1B
than when a
OCTG Premium
&%$##
sphere-to-cone MTM seal is used. On the other hand, the
actual seal length will be larger. This!##C2>02@A4/1B
is visualized
!DEC2>02@A4/1B
in Figure 5, were the seal contact pressure in function&%###
of the distance along the seal area is measured
Cone-cone
sealconnections at make-up, 100% collapse pressure and 197%
through numerical analysis for
OCTG premium
!%$##
collapse pressure. The large peaks in the contact pressure at the beginning and the end of the seal area for
!%### rather than being physical truths. They occur
a cone-to-cone MTM seal are due to numerical errors
because of geometric transitions in the connection and$##will enlarge the risk of galling. Based on these
results, it cannot be concluded which MTM seal gives the best results. Additional experimental testing
#
should be conducted.
)
$
(
'
&
!
#
*+, -./012.34/52, 1.32.61.278 8 9
ections
=?2@:;F161B
=?2@A4/1B
!##C2>02@:;F161B
&%$##
!##C2>02@A4/1B
&%$##
:1.3204/-.0-2;61,,<6127=>.9
ne seal
'%###
'%###
:1.3204/-.0-2;61,,<6127=;.9
nalysis
!DEC2>02@A4/1B
&%###
!%$##
!%###
!DEC2>02@:;F161B
&%###
!%$##
!%###
$##
$##
Sphere-cone seal
#
)
$
(
'
&
!
#
#
)
$
(
'
&
!
#
*+, -./012.34/52, 1.32.61.278 8 9
*+, -./012.34/52, 1.32.61.278 8 9
'%###
:1.3204/-.0-2;61,,<6127=;.9
=?2@:;F161B
Figurewww.simytec.com
5: Seal contact pressure and distance
along seal area of a cone-to-cone (left) vs. a sphere-to-cone
!##C2>02@:;F161B
&%$##
G/2-F1284H13+/524I2+/H<,-6+.32;6401,,1,
!"
!DEC2>02@:;F161B
MTM seal (right) [13]
3
one seal
&%###
FAILURE MECHANISMS OF THREADED CONNECTIONS IN OCTG
!%$##
3.1!%### Structural failure
$##
When
a connection fails structurally, an ultimate limit state is crossed: the connection will completely or
partially
“break down”. The following failures comply with this definition:
#
)
$
(
'
&
!
#
Jump-out When
a connection
is axially loaded, excess radial displacement can occur. Space is
*+, -./012.34/52,
1.32.61.278 8 9
created between the thread profiles of pin and box. When this space becomes large enough, the
pin will jump out the box. This is illustrated in Figure 6 for a standard buttress API connection;
G/2-F1284H13+/524I2+/H<,-6+.32;6401,,1,
!"
- Rupture of the pipe or box at the most critical
section without the development of large plastic
strains. This can occur due to an excessive and/or cyclic load (fatigue).
-
-
1278
Collapse failure Pipelines under excessive axial load, external pressure and bending can fail by
global (helical buckling of the pipe) or localized (sectional loss of round shape) collapse. To
quantify the collapse pressure, it is important to take manufacturing imperfections such as (initial)
ovality, eccentricity and residual stress into account. When the imposed bending curvature
increases, the effect of the initial ovality on the collapse pressure diminishes due to the “Brazier
effect” [14].
Y. Guangjie et al. / Engineering Failure Analysis 13 (2006) 1275–1284
the pipe is evidently higher than that of the coupling. The strain at the first engaged thread of the two ends of
the pin is very large even to 6000 le. At make-up about two turns there is a generalized plastification
of the pin
/012#3454
ᑨ࡯
nose.
6#7$8
3.2. Distribution of normal contact force
The normal contact force distribution of the tubing thread flank along the axial direction is shown in Fig. 4.
The shape of the normal contact force distribution is like a saddle. This is in accordance with the phenomenon
that thread galling failure easily appears at the two ends of tubing threaded connections in !"#$%
oil and gas fields. It
was further found that make-up or break out on a tubing connection will affect the taper of the pin and reduce
the pitch diameter. Ovality of tubing or coupling does not affect the contact stress distribution pattern in the
circumferential direction; however, it affects the actual value of the contact stresses.
) W
) W
) W impact
) W
)normal
W contact
W
) W
The tensile load
has a definite
on the !"#$%&,(-.
force )distribution.
The normal contact force of
!"#$%&'()
!"#$%&'*(
!"#$%&+)'
!"#$%&,(-!"#$%&,('( !"#$%&,,*(
the lead flank of the
thread
diminishes
under
the
action
of
tensile
load,
and
is
shown
in Fig. 5. Furthermore, it
6,8
6)8
698
6.8
6*8
6-8
6:8
has a strong influence on the equivalent stress distribution, and the maximal equivalent stress!appears at the
first engaged thread
of the
end of thephenomenon
male thread. for
Thea average
pressure
at the stab flank decreases
Figure
6: Jump-out
standardcontact
buttress
API connection
/tension. The leak resistance
with
increasing
according
to
the
API
(1987)
formula
does
not depend on applied
(P-110
244.47mm
x
11.99mm)
[15]
%&'()'!*+,-'+,#*&'.
0'12,'3"4&4'45%&440'6&789*+45":';2,&)'<$=982$5'!>'4"=$*+5"2,'2
tension. However, the test results and FEA calculations indicate that tension will reduce the sealability of an
$55%&44':2,,&:5"2,'
API- 10-R
tubing
connection.
Galling
When
two metal surfaces are in relative converging contact, for instance during make-up,
there of
is the
a chance
of adhesive
and transport
of material
between of
thenormal
two metallic
The taper
pipe thread
is also anwear
important
factor affecting
the distribution
contactsurfaces.
force of
This leads
directlyThe
to the
deterioration
of sealcontact
performance
connection
[16].
risk of
the oil casing
and tubing.
distribution
of normal
force isand
shown
in Fig. 6strength
under the
APIThe
ultimate
galling
is higher
at the
the pin
and
box.
This
is also
indicated
by the
saddle-shaped
normal
conjugate
of 0.067708
taper
pinend
and of
0.0599
taper
box.
The
work
situation
is named
as small
head and big
hat
contact
forceFig.
distribution
of the
tubing thread
flank
alongforce
the axial
anultimate
API 8-round
in oil and
gas fields.
7 shows the
distribution
of normal
contact
underdirection
the other of
API
conconnection
(Figure
7).0.067708 taper box. This is named as big head and small hat in oil and gas fields.
jugate of
0.0599 taper
pin and
/012#3454
ᑨ࡯
6#7$8
Fig. 4. The normal contact force distribution of tubing thread flank after making-up two turns.
Figure 7: The normal contact force distribution of
tubing thread flank after making-up two turns (API 8-round) [16]
!"#$%
3.2
Functional failure
) W
) W
) W
) W
In addition to structural
failure,
there
is
functional
failure.
Functional
failure
of a threaded connection occurs
6B0D[
6B0D[ 6B0D[ 6B0D[ when the connection
no
longer
complies
with
the
demands
of
serviceability
and sustainability, for instance
6,8
6)8
698
6.8
when a connection fails in preventing a gas or liquid from leaking. This can be due to! misuse of the
connection or poor design. Numerical analysis can estimate the sealability of a connection relatively, but
only laboratory /
or full scale testing can fully quantify the exact sealability [17].
%&'B)'!*+,-'+,#*&'C 0'12,'3"4&4'45%&440'6&789*+45":';2,&)'<$=982$5'!>'4"=$*+5"2,'2
$55%&44':2,,&:5"2,'
Fig. 5. The normal contact force distribution of tubing thread flank after making-up two turns and applying 120 kN tensile load.
3.3
In-depth analysis of API 5C3 formulae
Some of the above-mentioned structural failures can be calculated by the formulae and methods presented
in API 5C3 [18]. The collapse pressure, internal pressure resistance (also known as burst resistance) and
joint strength of threaded connections have been analysed and are discussed in the following paragraphs.
The analysis is limited to standard API buttress threaded connections and forms a first step towards the
research of premium connections.
3.3.1
Collapse pressure
The collapse pressure in the absence of axial loading and internal pressure can be calculated by means of
four formulas (see below). Depending on the type of collapse, which translates itselves to a certain
applicable D/t range, and the minimum yield strength of the pipe body (Yp), another formula governs, rather
than the collapse formula that gives the lowest collapse pressure.
For the smallest D/t ratios (D/t < ∼ 11-16 depending on steel grade) or for thick wall pipe, the yield strength
collapse pressure is decisive. It is derived based on yield at the inner wall using Lamé thick wall elastic
solution. It is thus not a true collapse pressure, but rather the external pressure PYp that generates hoop
stress exceeding the yield strength of the material before a collapse instability failure occurs [19].
!/! !!
!!" = 2!!
(1)
!/! !
Next, for increasing D/t ratios, the plastic collapse pressure governs. It has been empirically derived based
on 2488 tests of K-55, N-80 and P-110 seamless casing. Through regression analysis, the following
formula for the minimum collapse pressure for the plastic range of collapse was found (A, B and C are
curve fit factors depending on the steel grade):
!! = !!
!
(!/!)
−! −!
(2)
A plastic/elastic transition collapse pressure formula was determined on an arbitrary basis to connect the
plastic and elastic regimes with each other (F and G are curve fit factors depending on the steel grade):
!! = !!
!
(!/!)
−!
(3)
At last, for thin-wall pipe or D/t ratios larger than ∼ 20-40 depending on steel grade, the elastic collapse
pressure is conclusive. It is based on a theoretical elastic instability failure and is independent of the yield
strength. It can be calculated by means of the following formula:
!! =
!",!" × !"!
!/!
(4)
!/! !! !
The above-mentioned formulas and their representative zones of validity are graphically presented for steel
tubes of grade N-80 with various D/t ratios in Figure 8 (left). Next, a comparison is made of the collapse
pressures depending on the grade used (J-55, N-80 or P-110) (see Figure 8, right). The vertical lines in the
graphs represent the boundaries of elastic collapse zones. P_YP represents the yield strength collapse
pressure, P_P the plastic collapse pressure, P_T the transition collapse pressure, P_E the elastic collapse
pressure and P_c the effective collapse pressure depending on the type of collapse or the representative
D/t ratio.
%#!"
*.,/0"
1+2,34+5"
%!!"
(/61+.7"
+26318
.+.93""
,/61+.7"
!"##$%&'(%)'&&*)'(+,-$.(
!"##$%&'(%)'&&*)'(+,-$.(
%#!"
$#!"
;316:,"
$!!"
#!"
16:,"
$!"
()*("
02.-345"
$#!"
1*-./0"
$!!"
#!"
-./0"
!"
!"
%!!"
%!"
()("
/01(+2.(
()+"
&!"
'!"
(),"
#!"
()-"
!"
!"
$!"
()##"
%!"
/01(+2.(
*)+!"
&!"
'!"
#!"
,)$$!"
Figure 8: Grade N-80 collapse pressures (left) and grade J-55 vs. N-80 vs. P-110 collapses pressure (right)
As stated above, depending on the type of collapse or D/t range and the minimum yield strength of the pipe
body, another formula governs, rather than the collapse formula that gives the lowest collapse pressure.
This can be seen in Figure 8 (right). The transition collapse pressure formula gives a smaller value in the
first two zones, but there, the yield strength collapse pressure formula or the plastic collapse pressure
formula should be used.
When the grade of the material increases, so does the collapse pressure (Figure 8, right). However, only
the yield strength collapse and plastic collapse show a significant difference between the collapse
pressures of the different grades. When entering the transition and elastic collapse zone, the collapse
pressures of the different grades do not differ (much) from each other.
Another trend that can be seen when the material gets stronger (e.g. N-80 or P-110) is that the yield
strength collapse, plastic strength collapse and transition collapse zone gets smaller while the elastic zone
gets relatively larger (see Figure 8, right). This means that when stronger materials are used, the pipe will
rather fail due to elastic collapse than due to plastic or transition collapse.
!"#$#+,(,,)'./012)#3#,(444%./012%#+#-5(-66./012#3#%)%(6&#
!"##$%&'(%)'&&*)'(+,-$.(
%7,#
!"#$#%&'()*+,(-,-./012#
%,,#
EF8G/#
HIJ8KLIM#
-7,#
-,,#
!GNHIFO#
IJNKH+
FIFPK##
8GNHIFO#
RKHNQ8#
7,#
HNQ8#
,#
,#
-,#
%,#
89:;<*<=>?#1@*<A#?B<*#
/01(+2.(
),#
4,#
!;?C<;DB>?#1@*<A#?B<*#
7,#
!"#$%&'&()&"*+,-(+.$/,"0'&(1",$(&"234"
30/(1/'1"
Through curve fitting, a third-order polynomial trend line and an exponential trend line are found to match
the effective collapse pressures, independent of the sort of collapse or D/t ratio (see Figure 9, left). The
differences (in percentage) between the polynomial trend line and the formulas given in API 5C3 are shown
in Figure 9 (right) for steel tubes of grade N-80. Similar trends and differences were found for the
exponential trend line.
"#$%&
"#$"&
"#'%&
"#'"&
"#"%&
"#""&
!"#"%&
!"#'"&
!"#'%&
,0.12&
3-4.56-7&
!"#$"&
*183-09& -4853!
0-0:5&&
.183-09&
!"#$%&
"&
*+,*&
'"&
$"&
*+*&
#50"678"
*+-&
("&
)"&
*+.&
%"&
*+/&
Figure 9: Polynomial and exponential trend lines of the effective API 5C3 collapse pressure (left) and
Differences (%) between the polynomial trend line and the API 5C3 collapse pressures (right)
When a next type of collapse zone is entered, for instance going from yield strength collapse to plastic
collapse, it can be seen that the difference between the trend lines and the API 5C3 formulas reaches a
peak. This means that API 5C3 changes formulas because the calculated collapse pressure differs too
much from the “real collapse pressure”, calculated with a polynomial or exponential trend line valid for all
D/t ratios.
However, when going from the transition collapse zone to the elastic collapse zone, this peak doesn’t
occur. This can be explained as follows. The plastic/elastic transition zone is derived on an arbitrary basis
to connect the plastic collapse pressures with the elastic collapse pressures for given D/t ratios. As stated
above, depending on the grade of the material, the yield strength collapse and plastic collapse pressures
differ, whilst the elastic collapse pressures remain constant. A peak is therefore expected when going from
the plastic collapse to the transition collapse zone, but not when going from the transition collapse to the
elastic collapse zone.
Figure 10 (left) graphically presents a comparison between the differences of the polynomial and
exponential trend lines (Grade N-80) versus the API 5C3 collapse pressures. It can be seen that a thirdorder polynomial equation gives an overall better result. However, the differences (in percentage) are very
small (between -0,025% and 0,075%), so a single equation to calculate the collapse pressures could be
appropriate. Figure 10 (right) compares the percentual differences between polynomial trend lines and API
5C3 predictions for tubes with J-55, N-80 and P-110 steel grade. When increasing the steel grade, the
difference between the trend line and the in API 5C3 calculated collapse pressures also increases, but
overall the same conclusion can be drawn: the differences (in percentage) are very small compared with
the in API 5C3 calculated collapse pressures. Therefore, it should be possible to derive a single (third
degree polynomial) formula for each steel grade to calculate the collapse pressures as a function of D/t
ratio.
;<+=>%
?@A+BC@D%
"#"'$%
8=E?@<F% @AEB?!
<@<GB%%
+=E?@<F%
"#"$%
"#"&$%
"%
!"#"&$%
!"#"$%
"%
("%
&"%
)"%
*"%
$"%
!"#$%&'&()&"*+,-(+.$/,"0'&(1",$(&2"324"
20/(1/'1"
!"#$%&'&()&"*'+*+,&-".'&(-/$(&,"0,1"
,.2(-2'-"
"#(%
"#"'$%
"#"$"%
"#"&$%
"#"""%
!"#"&$%
!"#"$"%
"%
&"%
)"%
*"%
$"%
#50"678"
#3."456"
+,-./0/123%450/6%37/0%
("%
+!$$%
8.39/.:723%450/6%37/0%
,!-"%
.!(("%
Figure 10: % Difference of grade N-80 polynomial and exponential trend lines vs. API 5C3 collapse
pressure (left) and % difference of grade J-55, N-80 and P-110 polynomial trend lines vs. API 5C3 collapse
pressures (right)
3.3.2
Internal pressure resistance
The internal pressure resistance of a threaded and coupled pipe is equal to the minimum of the internal
yield pressure for pipe or coupling and the internal pressure leak resistance at the seal plane (this is the E7
plane of perfect thread length). Formulas are given in API 5C3 and are repeated here below:
Internal yield pressure for pipe:
! = 0,875
Internal yield pressure for coupling:
! = !!
Internal pressure leak resistance at seal plane:
! = !"#$
!!! !
(5)
!
!!!!
(6)
!
! ! !!! !
(7)
!!! ! !
Where Yp and Yc are respectively the minimum yield strength of the pipe and coupling in pounds per square
inch. D and t are the nominal outside diameter and nominal wall thickness of the pipe, W is the nominal
outside diameter of the coupling and d1 is the diameter at the root of the coupling thread at the end of the
6
pipe in the powertight position, all in inches. Next, E is the modulus of elasticity (= 30 x 10 psi), T the
thread taper (in./in.), N the number of thread turns make-up and p the thread pitch (inches).
!"#$%"&'()%$**+%$(%$*,*#&"-$(./0&1(
The internal pressure resistance for buttress threaded and coupled pipe (D = 7”, grades J-55, N-80 and P110) in function of D/t is presented in Figure 11. Also, the internal pressure leak resistance at the seal plane
(IPLR), which is independent of D/t and the yield strength of the steel, is visualized.
#!!"
D = 7"
+!"
*!"
)!"
(!"
UNSAFE
'!"
&!"
%!"
$!"
SAFE
#!"
!"
!"
#!"
$!"
%!"
&!"
'!"
(!"
23#(.41(
,-''"
.-*!"
/-##!"
0/12"
Figure 11: Internal pressure resistance (Grade J55 vs. N-80 vs. P-110)
In all cases, the internal yield pressure for pipe or coupling is smaller than the internal pressure leak
resistance at the seal plane. For small D/t ratios (D/t ≤ 15,9), the internal yield pressure is no longer a
function of D/t, thus, the coupling is decisive. In all other cases, the pipe body is the weakest member in a
threaded and coupled connection.
3.3.3
Joint strength
The joint strength of a buttress thread casing joint is taken as the minimum of the pipe thread strength and
the coupling thread strength, given by the following formulas [18]:
Pipe thread strength:
!! = 0,95!! !! 1,008 − 0,0396 1,083 −
Coupling thread strength:
!! = 0,95!! !!
!!
!! !
(8)
(9)
Where Ap and Ac are respectively the cross-sectional area of the plain end pipe and the coupling in square
inches and Up and Uc are respectively the minimum ultimate strength of pipe and coupling in pounds per
square inch.
A graphical representation of the buttress casing joint strength for grades J-55, N-80 and P-110 in function
of D/t (D = 7”) is given in Figure 12.
!"#$%&'()$%*'#*+,%&-*'./01'
(!!!"
D = 7"
'!!!"
&!!!"
UNSAFE
%!!!"
$!!!"
#!!!"
SAFE
!"
!"
#!"
$!"
%!"
&!"
'!"
(!"
23*'.41'
)*''"
+*,!"
-*##!"
Figure 12: Casing joint strength (grades J-55 vs. N-80 vs. P-110)
A similar trend as for internal yield pressure can be seen here: the casing joint strength depends for the
most part on D/t, which means the pipe thread strength is determinative. Only for D/t ratios less than ±14,5,
the coupling is the weakest part of a threaded and coupled pipe.
4
CONCLUSIONS
Due to the harsh conditions in HPHT wells, standard threaded OCTG can no longer be used and premium
connections are required. By changing the design of these premium connections, other properties can be
achieved. By modifying the thread form, structural integrity against axial loads, external and internal
pressure and bending can be acquired. The design of a torque shoulder determines how much torque load
can be applied. At last, the metal-to-metal seal provides the sealability of the connection. The design of a
premium connection is however not limited to these three parts of the connection. More information can be
found in other literature [4, 5].
When analysing the failure mechanisms of standard threaded OCTG, the following conclusions can be
drawn for the collapse pressure. When the used material gets stronger (e.g. N-80 and P-110), the
difference between the yield strength collapse and plastic collapse gets larger. Also, the elastic collapse
zone gets larger. This means that when stronger materials are used, the pipe will rather fail due to elastic
collapse than due to plastic or transition collapse. A third-order polynomial trend line can be derived that
describes the collapse pressure independent of the D/t ratio.
When analysing the internal pressure resistance and joint strength of threaded and coupled connections, it
is clear that the pipe body is the weakest link. Only for D/t ratios between smaller than ∼14-15, the coupling
is critical.
5
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
REFERENCES
J. Van Wittenberghe, et al., Fatigue investigation of threaded pipe connections. Sustainable
Construction and Design (SCAD), 2010. 1(1): p. 182-189.
Craig, B., Materials for deep oil and gas well construction, in Advanced Materials & Processes, May
2008: p. 33 - 35.
L.B. Hilbert Jr. and I.A. Kalil. Evaluation of Premium Threaded Connections Using Finite-Element
Analysis and Full-Scale Testing, SPE/IADC Drilling Conference, New Orleans, Louisiana, USA,
1992.
Galle, T., et al., Influence of design features on the structural integrity of threaded pipe connections.
Sustainable Construction and Design (SCAD), 2011. 2(2): p. 237-245.
J. Van Wittenberghe, et al., Design Characteristics that improve the fatigue life of threaded pipe
connections. Sustainable Construction and Design (SCAD), 2011. 2(2): p. 334-341.
Hunting, HES Seal lock Apex, U.S. Pat. No. 4,600,225, 2002
Yamamoto, K., et al., Stress Analysis of Premium Threaded Connection “FOX” by Finite Element
Method, in Kawasaki Steel Giho 20, 1989: p. 202-207.
Sugino, M., et al. Development of an Innovative High-performance Premium Threaded Connection
for OCTG, Offshore Technology Conference, Houston, Texas, USA, 2010.
Bradley, A.B., et al. Premium Connection Design, Testing, and Installation for HPHT Sour Wells,
SPE High Pressure-High Temperature Sour Well Design Applied Technology Workshop, The
Woodlands, Texas, U.S.A, 2005.
World Oil, 2008-09 Casing Reference Tables, November 2008
Takano, J., et al., Development of Premium Connection “KSBEAR” for Withstanding High
Compression, in Kawasaki Steel Giho 34, 2002: p. 21-28.
Santi, N.J., et al. Premium & Semi-premium Connections Design Optimization for Varied Drillingwith-Casing Applications, Offshore Technology Conference, Houston, Texas, USA, 2005.
Dvorkin, E.N. On the modeling of industrial processes, SIM&TEC. 2010.
Dvorkin, E.N. and R.G. Toscano, Finite element models in the steel industry: Part II: Analyses of
tubular products performance. Computers & Structures, 2003. 81(8-11): p. 575-594.
Lian Zhanghua, et al., Analysis of jump-out loads on connectors of buttress and round threads of
casing. China Petroleum Machinery, 2004-05.
Guangjie, Y., et al., Numerical and experimental distribution of temperature and stress fields in API
round threaded connection, in Engineering Failure Analysis 13, 2006: p. 1275 – 1284.
Cernocky, E.P., et al. A standardized approach to finite element analysis of casing-tubing
connections to establish relative sealing performance as a function of design geometry, machining
tolerances and applied loads, Fourth World Congress on Computational Mechanics, Buenos Aires,
Argentinië, 1998. p. 1167 - 1186.
API Bull 5C3, Petroleum and natural gas industries - Formulae and calculation for casing, tubing,
drill pipe and line pipe properties, 1993, American Petroleum Institute.
Mitchell, R.F., Casing Desing, in Petroleum Engineering Handbook, Volume II: Drilling Engineering,
2007.