Edinburgh Time-Lapse Project
PHASE II
Reservoir Geophysics Group
Institute of Petroleum Engineering
Heriot-Watt University
Edinburgh
EH14 4AS
Colin MacBeth
2 August 2003
SUMMARY
The Edinburgh Time-Lapse Project (ETLP) is a UK-based university research
consortium specialising in the application and development of analysis tools for
engineering-consistent quantitative interpretation of 4D seismic data.
Understandably, the research programme has a strong emphasis on the
integration of seismic and reservoir engineering. The consortium is now in its
second phase of research which started on June 2003 and will last until June
2006.
Our raison d’être is to deliver accurate estimates of those reservoir parameters
and their distribution that are of most value to the reservoir engineer for
understanding the reservoir and near well-bore flow dynamics. These estimates
are designed to be of direct interpretative benefit and can be used to constrain
and modify the reservoir model for better forward predictions. Our research is
achieved through appropriate use of full-wave seismic modeling tools, reservoir
modeling, improvements to the petro-elastic model and the development of
innovative seismic processing/interpretation methods.
All ETLP research areas are driven by data donated from sponsors. Our current
work portfolio covers fields categorized as either ‘safe bets’ or a ‘good chance’ in
the northern North Sea, Southern Gas Basin, the West of Shetlands, and the
shallow water Gulf of Mexico. The next three-year phase (Phase II) of the
project will continue to concentrate on fields in the UKCS, actively pursuing 4D
seismic research that can be categorized into four general areas:




Module 1 – Quantification of fluid saturation and contacts
Module 2 – Monitoring of depleting reservoirs
Module 3 – Assessment of challenging reservoirs
Module 4 – Independent estimation of pressure and saturation
The precise emphasis and resources committed to each category was defined by
our sponsorship group at the business meeting in May 2003. ETLP is ideally
placed to pursue the above research due to its unique location within the
Institute of Petroleum Engineering at Heriot-Watt University, recognized over the
past 10 years by its 5* research rating. This provides the project with access to
well-founded Rock Physics and Rock Mechanics laboratory facilities, together
with staff in the necessary range of supportive disciplines in Reservoir
Description that include Petrophysics, Structural geology, Drilling, and Reservoir
Engineering.
The ETLP consortium started Phase II of its research activity in June 2003,
however new entrants are welcome to join the sponsorship group at any point
throughout the programme.
2
SNAPSHOT OF PHASE I ACHIEVEMENTS - June 2000 to June 2003
During Phase I of the ETLP project the consortium was sponsored by nine
companies: BP, Shell, TFE, Statoil, Enterprise Oil, Schlumberger, Landmark,
Concept Systems and Fairfield Inc.
Publications
The Heriot-Watt team wrote and submitted over 6 papers for international
journals, and presented a further 36 expanded abstracts at EAGE, SEG, PESGB,
and NPD conferences and workshops. Further details and downloads of these
publications are available from our website:
http://www.pet.hw.ac.uk/research/etlp
The ETLP toolbox
Our 4D toolbox has been assembled over the past 3 years. It now contains:
 cross-equalization and warping tools for post-stack processing
 4D cross-equalization workflows
 SVD (singular value decomposition) differencing tools
 seismic modelling codes
 Petrophysical calculator
 3D visualization software for simulator to seismic computations
 simulator-to-4D seismic link
Studies completed during Phase I
Research projects successfully completed in accordance with our Phase I
proposal fall into four originally proposed categories:
Acquisition and processing:




Investigates the influence of water velocity variations on the
interpretation of the 4D signature – with a particular emphasis on
permanent OBC installations at the Teal South and Foinaven fields
Post-stack cross-equalization of OBC and towed streamer data for
analysis of compaction and production in the Valhall field
Post-stack cross-equalization to determine whether it is possible to
obtain a 4D signature from a depleting gas reservoir in the SGB
Application of an SVD-based difference technique, selective spectral
decomposition, and a cross-equalization flow to provide an enhanced
4D signature of water sweep in low net-to-gross areas on the Nelson
field
3
4D-specific attributes



Use of a spectral decomposition technique to determine 4D timethickness variations due to production in a Gulf of Mexico reservoir
Spectral decomposition investigated to interpret shadowing effects in
producing turbidites
Robust determination of P-wave AVOA for 4D OBC at the Teal South
field, Gulf of Mexico
New-wave Petrophysics:



Development of new pressure sensitivity law for 4D feasibility studies,
upscaling of laboratory tests and subsequent categorization of North
Sea sandstones
Quantification of gas saturation from seismic amplitudes in the
Foinaven field
Feasibility study to analyse the effect of fractures on the clarity of the
4D signature for a depleting gas sand reservoir
Links and integration with the reservoir simulator:






Geological model building, reservoir and seismic modelling of low netto-gross regions on Nelson – implications for injection and production
Combined fine-scale geological and 4D seismic interpretation to resolve
well-well connectivity issues on Foinaven
Use of P-P and P-S information to jointly constrain the reservoir model
in the Teal South field
A single seismic history match using seismic anisotropy from a turbidite
sand
A single history match using the P-P and P-S amplitude response
enhanced by multi-component –related petrophysical developments
Time-lapse borehole seismic analysis for Qarn Alam (VSP) and
Steepbank (cross-well), involving the first use of spectral
decomposition and SVD for cross-equalization. 4D interpretation of
steam flooding in a heavy oil reservoir, including petrophysical
development.
ETLP 4D datasets used during Phase I studies
Foinaven (NW Shetlands), Nelson (Northern North Sea), a Southern Gas Basin
field, Valhall (Norwegian Sea), Teal South (Gulf of Mexico). Borehole seismic data
include Qarn Alam (Oman) and Steepbank (Canada).
4
PHILOSOPHY
In the past, time-lapse seismic surveys have added value by helping to infer bypassed hydrocarbons and sense significant reservoir phenomena such as gas-cap
expansion, water encroachment or remaining compartments of oil. Rudimentary
processing and interpretation of seismic amplitudes has demonstrated that in
many cases the technique does fulfill the desired objective of inferring fluid
movements associated with hydrocarbon production. This has been shown to be
achievable for a variety of reservoir types and IOR processes involving steam,
hydrocarbon gas and water injection, or more complex recovery scenarios such
as WAG. Indeed, with more than a decade of 4D examples, the scientific
rationale and business case for using this technique can be considered generally
sound. However, as 4D seismic enters a new generation of activity in which the
initial successes of the West of Shetlands and Northern North sea have now
being joined by many other successes from fields elsewhere in the world, the
challenge to industry is to make the method deliver more precise and direct
engineering nformation in those fields where 4D has already been shot and
analysed. A second priority is to open up the application of this monitoring
technique to more difficult cases, two such examples of which are depletion in
low permeability reservoirs and compacting reservoirs.
ETLP recognizes that advances in 4D interpretation require that reservoir seismic
and engineering must be fully integrated, and perhaps even combined as a
separate subject in it’s own right. In particular, to achieve a true 4D
interpretation, time-lapse seismic signatures must be interpreted by honouring
rock physics, fluid physics, seismic wave propagation, geology, and engineering
principles. Our own work achieves this by blending the understanding of wave
propagation effects with flow properties and preserving a particular focus on the
petro-elastic domain. Within this context, our aim is to increase our
understanding of what is possible and what is not, and to reduce the frequency
of ‘4D surprises’ often revealed in case studies. It is highly likely that
engineering-true interpretation might require some modifications or even a
complete re-haul to cherished seismic theories that have in many cases been
founded on naïve assumptions about the physical properties of the reservoir
fluids and geology - this remains to be seen.
ETLP provides detailed research projects on true 4D interpretation focused by
sponsors’ datasets. We seek new ways of accurately estimating reservoir
engineering parameters directly from the 4D response and evaluate how they
contribute to an improved reservoir model. The ultimate aim of our research is to
build a suite of interpretative tools or scenarios to help companies best achieve a
high-resolution dynamic reservoir characterization of their reservoirs.
5
TECHNICAL PROGRAMME
The Phase II technical programme builds on our Phase I research by identifying
several topics where substantial concentrated effort is required for progress in
quantifying the interpretation of 4D seismic. Reservoir-related problems are
broadly categorized into four modules whose central concerns are (1) saturationdominated processes, (2) pressure-related issues, (3) difficult reservoirs with
uncertain or currently un-recognizable 4D signatures, and finally (4) the
challenge of estimating pressure and saturation independently.
Each module addresses a number of specific questions for which the success of
4D relies. For example:







Can gas saturation be accurately estimated with several seismic surveys at
our disposal? Is it possible to estimate gas saturation to a fraction of a
percent?
What is the limit of the seismic method for detecting the exact vertical
and lateral distributions of the reservoir fluids? Can individual flow units be
defined?
Can pressures be estimated to within several hundred psi?
Can pressure depletion in fractured or non-fractured zones in gas
reservoirs exhibit a visible 4D signature.
What is the nature of the 4D signature in low permeability carbonates?
Can production and mechanical effects be separated in difficult reservoirs
such as the compacting chalks in the Norwegian Sea.
Can we estimate the pressure and saturation fields in the hydrocarbon
reservoir. Indeed, is there a unique answer to this, or can they never be
resolved independently? Still much debated and currently un-resolved.
These, and many more related questions are addressed in the four modules that
constitute the ETLP technical programme, laying down a series of targeted
studies that explore critical links between the seismic and engineering domains.
The work relies upon sponsor-donated data to provide a focus for extensive indepth research. Each ETLP study involves our unique blend of field data
processing, modeling, empirical measurement and theoretical developments, and
will be composed of the following basic ingredients:




full-wave seismic modeling
4D-specific processing
petro-elastic model development
reservoir modeling
Existing datasets will continue to be analysed, with additional datasets having
been set by sponsors at our business meeting in May 2003.
6
Module 1 – Quantification of fluid saturation and contacts
Changes in fluid saturation and ‘contact’ location need to be estimated with a
determined accuracy. This module aims to explore and classify the limits of
monitoring saturation changes, and to capture the uncertainty inherent in the 4D
seismic method. It will offer developments towards an improved understanding
of saturation-dominated processes for a range of production and injection
scenarios, and the possibility of new saturation equations. Innovative seismic
processing and interpretation techniques will be investigated to determine their
value in improving the overall definition of oil, water and gas movement. An
example of one possible approach (also used in Phase I) is the use of spectral
decomposition for gas saturation. If the opportunity arises we will also consider
the possibility of using multi-component seismic.
One aim is to develop alternative saturation equations beyond the patchy and
homogenous approach. For this, we intend to make use of a simple and practical
saturation model that honours the reservoir fluid and rock conditions as
measured from the wireline log (including dipole sonic) and seismic data. The
resultant law will correctly account for realistic gas and water saturation
conditions (which quite probably lies somewhere between patchy and
homogeneous as in Brie’s law), and will be based on equivalent medium theory.
This saturation law will be developed in partnership with our collaborator, Wayne
Pennington of Michigan Technological University. It is anticipated that this will
lead to a better understanding of gas coning, gas injection, and bright spots.
Seismic modelling will be used for fine-scale evaluation of dynamic features
occurring during water injection such as (a) the formation of high permeability
streaks or slumping; (b) the complex movement through regions of low net-togross leading to a characteristically patchy distribution; (c) near well-bore effects
such as coning and fingers, or the effects of thermally induced fracturing as a
consequence of the injection process. This work will help better quantify and
monitor the risk of early breakthrough, and address questions such as: Where is
the saturation ‘front’? How fast is it traveling? When and where will it break
though? Modelling will also be used for solving issues connected with gas
saturation. One such problem arises for oil reservoirs initially just above bubble
point, for which gas exsolves from the oil due to drop in pressure during
production. A critical question is whether production will lead to mobile gas that
forms a gas cap or whether the gas remains in trapped as bubbles in the pore
space. Can seismic detect the different cases? This is important for constraining
reservoir connectivity between sands or relative permeability. This work may also
help in monitoring the displacement performance in a gas injection process, or
help engineers to distinguish between viscous fingering, gravity segregation and
permeability heterogeneity.
7
Module 2 – Monitoring of depleting reservoirs
This module consists of two topics, which build on work completed in Phase I.
The first contrasts pressure depletion in sandstone and carbonate reservoirs, and
assesses the feasibility of monitoring these effects using 4D data. The work is
also relevant to other reservoirs with low porosity or high initial pressure, for
which pore pressure is a dominating influence. The following issues could be
considered in this study:






the role of reservoir heterogeneity (porosity and permeability)
the role of sedimentary architecture on monitoring reservoir
performance
the seismic influence of partial aquifer support
monitoring a waterflood, with particular interest in the lateral
displacement of the front and appropriate saturation laws
the role of stress instead of pressure
the benefits of using converted waves
Datasets for this work are from gas condensate reservoirs in the Miskar field,
Tunisia, and the Sleipner Vest field, Norway.
The second part of this module is the evaluation of stress sensitivity laws for
shaly sands, and the extension of this to treat reservoir variability. This will
catalogue existing literature and theory, and use these to make predictions of
the likely effects in a reservoir setting. Particular attention will also be placed on
clays in sandstones and the influence of the clay distribution on the stress
sensitivity. The work will draw experience from past work in the West of
Shetlands and Central North Sea. The following will be considered:








VP/VS
Seismic anisotropy
Layered systems of sand and shale
Links between elastic and flow properties
Clay distribution in sands
Relationship to Vshale
Uniaxial compression versus pressure
Compilation and generalization of elastic properties of shales
This work will also consider methods for ground-truthing in situ pressure
dependency. In particular, it will tackle up-scaling between laboratory estimates
and the seismic, where a genuine disparity has been demonstrated with
synthetic sandstones. This will be accomplished by field study and modeling for
selected datasets where the effects for pressure can be isolated.
8
Module 3 – Assessment of challenging reservoirs
This module considers oil reservoirs that are classified as a ‘good chance’ for the
application of 4D technologies but are nevertheless at the boundaries of our
current understanding and require further specialized attention. For the purposes
of this work two categories of reservoir will be considered:
Compacting reservoirs – this particular topic is most relevant to chalk fields such
as Ekofisk and Valhall. Here, the 4D signature is easily observed and Phase I
work has demonstrated that 4D warping tools can deliver a high resolution
picture of the compaction process. One problem however is to disentangle the
seismic response due to production from that due to compaction. Our aim is to
use the data from our previous study to guide a modelling exercise to explore
the coupling between the petrophysical and geomechanical components of the
system, and investigate different pre- and post-stack seismic attributes. The
following subjects will be covered:





fluid flow versus compaction effects
database of chalk properties
development of a dynamic petro-elastic model
4D time lapse effects
possible PP versus PS analysis
Data from the Valhall field in Norway will be considered as part of this project.
Low permeability reservoirs - the seismic, rock and fluid properties of low
permeability reservoirs such as the gas-bearing Rotliegend sandstone in the
southern gas basin are strongly influenced by the presence of compliant fracture
clusters. Although forming only a small part of the available porosity, high
pressure differentials, fractures will deform these features and this could lead to
a strong stress-sensitive seismic signature. This project will involve the following:
 study of the rock and fluid physics
 stress-sensitive Petro-elastic model
 determination of the nature and magnitude of the 4D
seismic signature
 geological and reservoir modeling of the effects of
production
 field data analysis
A third Deepwater Reservoirs project is being considered, dependent upon
datasets becoming available.
9
Module 4 – Independent estimation of pressure and saturation
Separation of pressure-related and saturation-related influences on the seismic
response is the holy grail of 4D seismic. It is our belief that this procedure cannot
be uniquely accomplished without recourse to additional constraints from the
stress -dependent rock (petro-elastic) and fluid (reservoir engineering)
properties. In this module, this statement will be placed on a firmer quantitative
footing by a simulator to seismic study to examine the ultimate resolution
capability of the seismic method.
The development of a new technique will use a practical classification scheme
combined with our sandstone database to enable us to draw correlations and reorganize the reflectivity equations. Early work focusing on the West of Shetlands
has suggested some new relations for achieving pressure-saturation separability.
Future developments will expand the applicability of our results to other oil
reservoirs and test the techniques in a number of different production and
injection scenarios. Conventional and un-conventional seismic attributes
considered will include the AVO gradient, AI/GI parameters, elastic impedance,
VP/VS, extended elastic impedance, amplitude, or spectrally decomposed
amplitude. Provision of a robust method that can monitor changes in pressure
and saturation in the reservoir or near-borehole environment in the future
requires the following to be addressed:








two and three fluid phases
pressure gradients and vertical saturation distributions
calibration saturation laws using module 1
spectral decomposition
Bayesian inference methdologies
uncertainties
dynamic well ties
novel seismic attributes
10
ETLP PERSONNEL
The ETLP research team consists of allied senior staff, two postdoctoral
researchers, eight PhD students and benefits from an annual influx of summer
students on the Master’s programme. ETLP has a close association with other
active areas of research within the Reservoir Geophysics Group and the
Institute’s Reservoir Description Group.
DELIVERABLES TO SPONSORS






ETLP annual report in browsable CD form
bi-annual consortium meetings
analysis of sponsor-related data from key locations
advance viewing of submissions to geophysical and engineering journals
regular interim meetings for individual project updates
consultancy
MEETINGS
There are two meetings per year, scheduled to be in the winter and summer
months. Past meetings have typically been held in early November and late May.
SPONSORSHIP FEE
The sponsorship fee is £24k per annum, rising in £900 increments each year.
Companies are free to join at any time during our three-year phase but will be
required to contribute a penalty of 15k for doing so.
CONTACTS
Further information can be found on the ETLP web site:
http://www.pet.hw.ac.uk/research/etlp
or by contacting
Professor Colin MacBeth,
Reservoir Geophysics Group,
Institute of Petroleum Engineering,
Heriot-Watt University,
Edinburgh, UK.
Tel: 0131-451-3607; Fax: 0131-451-3127
Email: colin.macbeth@pet.hw.ac.uk
11