Professional Documents
Culture Documents
SPE 59398 Influence of Geological Features On Well Test Behavior
SPE 59398 Influence of Geological Features On Well Test Behavior
SPE 59398 Influence of Geological Features On Well Test Behavior
Copyright 2000, Society of Petroleum Engineers Inc. well test is evaluated with the help of a numerical well test
This paper was prepared for presentation at the 2000 SPE Asia Pacific Conference on simulator. The systematic evaluation of the well testing
Integrated Modelling for Asset Management held in Yokohama, Japan, 25–26 April 2000.
responses of various geological objects yields a new source of
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
reliable information for reservoir characterization.
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at Introduction
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
Reservoir characterization includes the description of the
for commercial purposes without the written consent of the Society of Petroleum Engineers is geology and the dynamic behavior of a reservoir. Geological
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous description involves a detailed knowledge of the hierarchical
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. sequence of geological heterogeneities influencing reservoir
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
properties and fluid flow characteristics in the reservoir.
Engineering characterization, on the other hand, is concerned
Abstract with understanding and being able to predict the dynamic
Early recognition of porosity and permeability contrasts in a behavior of these reservoir heterogeneities under specific
reservoir has economic implications in reservoir management. primary, secondary and tertiary oil field development plans.
Reservoir characterization aims at developing a detailed Much research is being carried out on both subjects because
knowledge of the geological heterogeneities that may affect there is still uncertainty about many aspects of reservoir
production rates and recovery factors. In the area of reservoir heterogeneities and the role that they play in oil recovery
characterization, geological and engineering data have been (Weber4 ).
traditionally inverted separately: a 3-D static reservoir model Some heterogeneity can be adequately quantified by direct
is established from static data (geology, geophysics and measurements on cores or in boreholes. However, evaluation
petrophysics) and the parameter distributions in that model are of the overall effects of heterogeneities on fluid behavior must
adjusted by history matching with production and other be achieved with the aid of field experiments at a large enough
dynamic data. The underlying assumption in this process is scale or the use of models. Geological reservoir modeling can
that production data correspond to the same model as static be based on deterministic or stochastic techniques, with the
data (Floris et al.1 ; Landa et al.2 ). This may or may not be the latter becoming widely accepted as a basis for the analysis and
case. Hence, the requirement to first identify separately a development of oil reservoirs.
model for each type of measurement and to verify that each The application of stochastic modeling techniques to
model is consistent with all the others. Because identification reservoir characterization is often based on theoretical
is an inverse problem, with a non-unique solution, it is also sedimentological parameters. Improvement in these modeling
desirable to confirm each feature of the reservoir model with techniques depends on the input of reliable static and dynamic
different types of measurements 3 . For some features this is data. No single measurement, however, has enough resolution,
possible with today’s technology. For instance, well test accuracy and volume of investigation to provide all the
analysis can confirm the location of a fault identified from required data for stochastic modeling. Pressure transient tests
geophysical measurements and vice-versa. Other features, on may come close, depending on the test duration. Hence, the
the other hand, require additional development. need to research the extent at which well-test analysis could
The present work looks at the medium-scale architectural provide further reliable information for reservoir heterogeneity
elements that have been formed in different depositional characterization. The present paper reports results of well-test
environments,, with the objective of reconciling geological numerical simulations performed in order to capture the
and well testing modeling in order to increase the confidence impact of reservoir heterogeneities on well-test behavior
in the evaluation of fault sealing, sandstone bodies geometry beyond what has been available to-date. The focus of the
and lithofacies distribution. In this approach, the present study is on the external geometry of specific genetic
heterogeneities are regarded as geological objects units in various depositional environments.
characterized by geometrical parameters as well as by
petrophysical properties. Synthetic genetic-type objects are
generated by stochastically modeling their external geometries
and internal rock-property distribution and their response in a
2 J. A. ZAMBRANO, R.W. ZIMMERMAN AND A.C. GRINGARTEN SPE 59398
channel curvature (point A and B’ in Figure 1-a) and thinner at responses generated in such way are then analyzed and
its inner part (point A’ and B in Figure 1-a). A symmetric verified by analytical methods.
composite profile is typical of a straight channel, in which Straight composite-thickness pseudo-channels (Figs 2-3)
thickness follows a parabolic profile along the body width. Two 11-km straight channels were simulated. These
According to Table 2, combination of these profiles models consist of a main zone with net pay, h, of 56.64 ft and
generates thirty-six (36) different types of geological bodies, a second zone (three in the symmetric case) with reduced
but only twelve can occur in recent/ancient sedimentary thickness, h L2 . Three values of h L2 were compared, i.e., 0.25h,
environment. If we grouped these combinations according to 0.50h and 0.75h. A well was 82.5 ft away from the boundary,
their analogous genetic-type sandstone body, we would obtain BL1 , which separates one zone from the other. The position of
six pseudo-channels, three pseudo-bars and three pseudo-fans. the closest no-flow boundary, BR , was varied (300 ft, 600 ft,
The term pseudo-channel was used for the first time by Zheng 1200 ft and 2400 ft, respectively).
et al11 . These twelve genetic type sandstone bodies are defined
as genetic-type objects. Table 3 summaries the two- Sinuous constant pseudo-channel (Figure 5)
dimensional cross-sections of each genetic-type object. Four no-flow boundaries were arranged in two parallel
wedge systems. As shown in Figure 1-b, one system is
Methodology intersecting at 120° and the other at the complementary angle
A methodology to investigate the well test behavior of each (240°). Each system is parallel to the other separated by the
genetic-type object is presented. Each genetic-type object has channel’s width. Such a configuration can also be visualized
a set of features whose effect on pressure response is as two intersecting linear channels (i.e., ch1 and ch2; see
unknown. By following a two-step process, it is possible to legend in Figure 6), which mimic a single-bend meander
understand the pressure behavior of a well tested within a channel. The channel wavelength and amplitude are 26,400
genetic-type object and to evaluate in a systematic way the and 7,621 ft, respectively.
kind of information obtainable from well test analysis. A well was located 51 ft from fault 1 and at 200-ft from the
In the first step, simple simulation models are constructed 120°-apex formed by intersecting faults 1 and 2. Four different
and the pressure response numerically generated (direct values of channel width were used, 330, 660, 1320 and 2640
problem). Subsequently, the pressure response is analyzed by ft.
applying a well test pattern recognition process (inverse
problem)12 using the knowledge base (model catalog) Results
developed in the first step. This section presents some preliminary results. Description of
Once the effect of each individual feature has been the well test behavior observed in constant-rate drawdown
determined, the next step consists in combining features in simulations is presented. In each case, a brief discussion of
more complex simulation models and repeating the process results is included.
depicted above. This approach makes possible to separate the
effect of each individual element of heterogeneity and to Asymmetric composite-thickness pseudo-channel
identify them in an actual test. Figure 2 shows the log-log plot of pressure response of a
In this paper, we report some preliminary results that have well within a straight channel with an asymmetric composite
been obtained for two features, i.e., symmetric/asymmetric thickness profile. In Figure 2, linear borders were used to
composite thickness-profile and sinuous plan view. model a thickness-profile with h L2 =0.75h.
Pressure derivative log-log behavior: At late times, before the
Numerical Models linear flow half-slope line is seen, the pressure derivative
Constant rate drawdown (CRD) behavior of a fully vertical exhibits two stabilizations above the middle time radial
well perforated at different locations within a genetic-type stabilization. If we define a dimensionless pressure such that
object was determined by single-phase numerical simulation. radial flow stabilization occurs at a dimensionless value
A constant compressibility (1.31E-5 1/psi), 35 API degrees p’D =0.514 , the first stabilization is at p’D = 0.57 and the second
undersaturated oil was assumed as well as a viscosity of 5.59 at p’D = 1.33. The first stabilization is 1.14 times the 0.5-radial
cp and formation volume factor of 1.28 rb/stb in all the cases. flow stabilization and the second one is 2.66 times. The
Three 3-D homogeneous simulation models were second stabilization is followed by a log-log half-unit straight
constructed, i.e., two straight composite-thickness pseudo- line corresponding to the conventional linear flow channel
channel and one sinuous constant-thickness pseudo-channel. behavior of corresponding width.
Firstly, these objects were modeled by using high contrast in The pressure stabilization at a higher level is due to a
thickness (black linear borders in Table 2); subsequently, decrease in mobility (in the present case, because of an
curve boundaries (shaded shapes in Table2) were used instead. increase in reservoir thickness), which is observed once the
A structured (PEBI – PErpendicular BIsection) type of inner boundary, BL1 is seen. The relationship expressed by
computational gridding13 was used to define radial grid Equation (1) determines the value of p’D at the first
refinement around the well and hexagonal gridding elsewhere stabilization:
within the object. Accurate simulation of constant wellbore
storage (1.44E-3 bbl/psi) and zero skin effect was achieved by
automatic calculation of gridblock-transmissibility factor for
each connection. After gridding, a black-oil simulation run
generates the well bottom hole pressure (BHP). The pressure
4 J. A. ZAMBRANO, R.W. ZIMMERMAN AND A.C. GRINGARTEN SPE 59398
h h
p 'D = × 0 .5 = × 0 .5 (1)
h − hL2 h + h L2 Sinuous constant pseudo-channel
hL2 +
2 2 Figure 5 shows the log-log plot of pressure and pressure
3 derivative responses of a well within a meander channel with
e.g., when h L 2 = h ,
4 constant thickness.
8 Pressure derivative log-log behavior: At late times, as the
p ' D = × 0 . 5 = (1 . 14 ) × 0 . 5 = 0 . 57
7 no-flow fault 1 is seen, well test behavior shows an expected
first stabilization at p’D =1.015 . When the width of the channel
It is inferred from this relationship that the level of the first (330 ft) is in the same order as the well-to-apex distance (200
stabilization is proportional to the mean thickness between h L2 ft), this stabilization is the only one exhibited on the pressure
and h.Conversely, if this behavior were observed in an actual derivative. In this case, well test behavior is similar to that
test, it could indicate that the perforated formation is thinning exhibit by a linear channel. As the channel’s width increases,
away from the well. Quantitative assessment of such a
the effect of the closest wedge (i.e., 120°) is seen on the
thickness reduction can be obtained from Equation (1). Other
pressure derivative. This yields a stabilization at p’D =360°/θ,
cases were simulated with h L2 is 0.25h, 0.50h and 1.0h, with
where θ is the intersecting angle. In the example of Figure 5,
results confirming Equation (1).
The second stabilization in Figure 2, on the other hand, is the second stabilization occurs at p’D =1.5, which is 360°
due to the presence of BR.In this case, the derivative divided by 120° times 0.5 (semi-infinite stabilization). At late
stabilization is at 1.33, obtained as follows: time, a one-half slope line develops due to the presence of two
h intersecting straight-line channels acting simultaneously,
p 'D = 2 × × 0 .5 (2 ) which is characteristic of a meander channel.
hL2
It was found that the relationship between the angle of
3
e.g., when h L 2 = h , intersection and the channel-curvature’s wavelength to
4
amplitude ratio is governed by Equation (4):
4
p ' D = 2 × × 0 . 5 = 1 . 33
3 W è
The level of the second stabilization is twice that = 2 x tan ( ) (4)
A 2
corresponding to the thickness hL2 , with the factor 2
W 26400
accounting for the boundary BR being no-flow15 . e.g., when = = 3 . 46
A 7621 . 02
120 °
Symmetric composite-thickness pseudo-channel 2 * tan ( ) = 2 x 1 . 732 = 3 . 46
2
Figures 3-4 show the well test behavior of a well within a
straight channel with symmetric composite thickness-profile.
In practice, if this peculiar combination of behaviors is
As in the asymmetric case, a high contrast in thickness was
exhibited in a well test, it could suggest the presence of a
first modeled (h L2=0.75h).
meander channel and a quantitative assessment of channel
Pressure derivative log-log behavior: At late times, the
curvature’s wavelength/amplitude ratio can be achieved by
pressure derivative exhibits two stabilizations above the radial
applying Equation (4).
flow stabilization at p’D =0.5. The first stabilization is at p’D =
0.67, which is 1.33 times the 0.5-radial flow stabilization. The
Field Data
second stabilization is at twice the level of the first one.
Figure 6 shows a log-log pressure and derivative plot of a
The stabilization of the pressure derivative at a higher level
buildup test on well W-175 in the middle Frio Formation,
than for radial flow is due to a decrease in reservoir thickness
Stratton field, Texas, during a secondary natural gas recovery
from h to h L2 , which is observed once both inner boundaries,
project16 . The well is located within one of three interfingering
BR1 and BL1 are seen. The relationship expressed by Equation
channels systems imaged with 3-D seismic dataand, based on
(3) determines the value of p’D at the first stabilization:
visual inspection of seismic interpretation, could possibly be
h located at a channel curvature. Data in Figure 6 is very similar
PD = x 0 .5 ( 3)
h2 to the numerical pressure response presented in Figure 5 for a
3 well within a sinuous channel and such a model could be used
e.g., when h2 = h,
4 for quantitative interpretation.
4
PD = x 0 . 5 = (1 . 33 ) x 0 . 5 = 0 . 67
3 Conclusions
Twelve basic genetic-type objects have been defined with
The second stabilization occurs at p’D =1.33 according to
simplified geometries for most common genetic sandstone
Equation (2).
bodies. The simplified description of these bodies is aimed at
providing models suitable for well test analysis while being
Subsequently, a more realistic parabolic surface was
geologically realistic.
substituted for the linear borders. Figure 4 compares the well
The approach presented herein has proven suitable to study
test behavior of a well within a body with square composite-
the actual well test behavior of well-known geological
thickness profile and parabolic composite thickness-profile.
features, whose well-test signatures were not available until
As it can be observed, the behavior is similar during the late
now.
time region.
SPE 59398 INFLUENCE OF GEOLOGICAL FEATURES ON WELL TEST BEHAVIOR 5
Preliminary results confirm that well-test analysis can 10. Yaxley, L.M., New Stabilized Inflow Equations for Rectangular
provide information on reservoir large-scale features, and Wedge Shaped Drainage Systems. SPE 17082, 1987.11.
specifically: 11. Zheng, S., Corbett P And Stewart G., The Impact of Variable
Formation Thickness on Pressure Transient Behavior and Well
Test Permeability in Fluvial Meander Loop Reservoirs, SPE
1) The average thickness between a thicker zone penetrated 36552, 1996.
by a well and a thinner one away from it. Examples of 12. Gringarten, A. C.: "Computer-aided Well Test Analysis," SPE
linear channels with asymmetrical composite thickness paper 14099 presented at the Second International Petroleum
profiles have been developed. Exhibition and Technical Symposium of SPE, Beijing, China,
2) The thicker-zone to thinner-zone thickness ratio in a March 17-20, 1986.
symmetric system. An example of a linear channel with a 13. GeoQuest, Weltest 200 Reference Manual, Schlumberger,
symmetrical composite thickness profile has been Abingdon, UK, 1998.
discussed. 14. Bourdet, D. P., Whittle, T. M., Douglas, A. A. and Pirard, Y. M.:
"A New Set of Type Curves Simplifies Well Test Analysis,"
3) Channel-curvature wavelength to amplitude ratio. World Oil (May, 1983) 95-106.
15. Clark, D. G. and Van Golf-Racht, T. D.: "Pressure Derivative
Further developments in these lines of research could lead Approach to Transient Test Analysis: A High-Permeability
to well-test and seismic data reconciliation regarding the North Sea Reservoir Example," J. Pet. Tech. (Nov., 1985) 2023-
geometry and dimensions of large-scale sandstone bodies, thus 2039
decreasing uncertainties in reservoir characterization.. 16. Junkin, J., Cooper, K. and Sippel, M., Pressure Transient Anaysis
Techniques Aid Secondary Recovery Method Choices,
Acknowledgement Petroleum Engineer International, March, pp. 34-41, 1996.
We would like to express our appreciation to Exploration
and Production PDV S.A. who has financially supported this
MPhil project. We also thank Schlumberger GeoQuest and Nomenclature
Baker Atlas GEOScience for providing software and continue
support. φ Average formation porosity
K Average formation permeability
References h perforated net pay thickness (measured in
core/logs)
1. Floris, F. J. T., Geostatistical Reservoir Generation Conditioned to h2 formation thickness away the well
Production Data, paper L053, Proceedings EAGE 96, BD inner boundary between adjacent thickness zones
Amsterdam, June 3-6, 1996 B2 well’s closest no-flow boundary
2. Landa, J. L. & Horne, R. N., A Procedure to Integrate Well Test W Wavelength of meander channel curvature
Data, Reservoir Performance History and 4D Seismic A Amplitude of meander channel curvature
Information into a Reservoir Description, SPE 38653. Paper µ Oil viscosity
presented at the SPE Annual Technical Conference and
Exhibition, San Antonio, Texas, October 5-8, 1997. ct Total compressibility
3. Gringarten, A. C.: “Evolution of Reservoir Management
Techniques: From Independent Methods to an Integrated
Methodology. Impact on Petroleum Engineering Curriculum,
Graduate Teaching and Competitive Advantage of Oil
Companies”, SPE 39713, presenteded at the 1998 SPE Asia
Pacific Conference on Integrated Modelling for Asset
Management held in Kuala Lumpur, Malaysia, 23–24 March
1998.
4. Weber, K. J. How Heterogeneity Affects Oil Recovery. In: Lake,
L. W. & Carroll, H.B. J. (eds) Reservoir Characterization,
Academic Press, Orlando, Florida, pp. 487-544, 1986.
5. Flint, S., Steward, D. J., Hyde, T., Gevers, E.C. A., Dubrule, O. R.
F., and Van Reissen, E. D., Aspects of reservoir geology and
production behavior of Sirikit oil field, Thailand: an integrated
study using well and 3-D seismic data: AAPG Bulletin, v. 72, pp.
1254-1269, 1988.
6. Reynolds, A.D., Dimensions of Paralic Sandstone Bodies, AAPG
Bulletin, v. 83, February, pp. 211-229, 1999.
7. Prasad, R. K., Pressure Transient Analysis in the Presence of Two
Intersecting Boundaries, Journal of Petroleum Technology,
January, pp. 89-96, 1975.
8. van Poollen, H. K., Drawdown Curves Give Angle Between
Intersecting Faults, The Oil and Gas Journal, December, pp. 71-
75, 1965.
9. Yaxley, L.M., Effect of a Partially Communicating Fault on
Transient Pressure Behavior, SPE Formation Evaluation,
December, pp. 590-598, 1987.
6 J. A. ZAMBRANO, R.W. ZIMMERMAN AND A.C. GRINGARTEN SPE 59398
Table 1. Summary of plan-view profiles displayed by sandstone bodies formed in principal environments of deposition.
YZ-profile
(Longitudinal)
THICKNESS Constant Increasing Composite Decreasing Composite
-PROFILES
XZ-profile
(Transverse)
Constant Symmetrical Composite Asymmetrical Composite
Table 2. Simplified geometries for sandstone bodies (blue shapes represent the actual shape of genetic type
sandstone bodies)
GENETIC-TYPE
PLAN VIEW YZ-PROFILE XZ-PROFILE
OBJECT
Straight Constant Constant
Straight Constant Symmetrical Composite
Pseudo-channel
Straight Decreasing Composite Constant
Straight Decreasing Composite Symmetrical Composite
Sinuous Constant Constant
Sinuous Constant Asymmetrical Composite
Elongate Constant Constant
Pseudo-bars
Elongate Increasing Composite Constant
Elongate Increasing Composite Symmetric Composite*
Fan/lobate Constant Constant
Pseudo-fans
Fan/lobate Decreasing Composite Constant
Fan/lobate Decreasing Composite Symmetric Composite
Table 3. Twelve genetic-type objects are defined to investigate the influence of geological features on well test
*
The symmetric composite thickness-profile of a bar compared to the one displayed by a channel is up side down.
SPE 59398 INFLUENCE OF GEOLOGICAL FEATURES ON WELL TEST BEHAVIOUR 7
behaviour.
a) b)
Figure 1. a) Meander channel displays changing thickness along its lengths (after Zheng et al.,
1996), b) Plan view of a sinuous pseudo-channel.
l
ne
an
Ch
R
B
&
BR
L
B
1.33
e
BR,A BR,B BR,C BR,D l op
2s
BL 1/
1
0.57
0.50
Figure 2. Log-log plot of drawdown test on well within a channel with asymmetric composite
thickness profile.
8 J. A. ZAMBRANO, R.W. ZIMMERMAN AND A.C. GRINGARTEN SPE 59398
1.33
pe
slo
BR,A BR,B BR,C
2
0.67
1/
0.50
Distances to:
BL2 hR2 BR,A BR,B BR,C
BR,A =300 BR,B =600 BR,C =1200 ft
hL1 BR1 hR1
Figure 3.. Log-log plot of drawdown test on well within a channel with symmetric composite
thickness profile.
1.33
pe
slo
2
1/
0.67
⊗
BL2 BR 2
BL1 B R1
Figure 4. Log-log plot of drawdown test of a symmetric composite-thickness system with a) Linear
(dotted line) and b) Parabolic (bold line) cross-sections.
SPE 59398 INFLUENCE OF GEOLOGICAL FEATURES ON WELL TEST BEHAVIOUR 9
pe
slo
1/2
w
Flo
ar
ne
Li
360 °
Channel width = 330 660 1,320 2,640 ft x 0.5 =1 .5
θ
1.0
0.5
Ch1
θo
•
Ch2
Figure 5. Log-log plot of drawdown test on well within a sinuous channel with constant thickness.
∆m(P), psi2/cp(*1E-06)
Equivalent pseudotime
Figure 6. Log-log plot of a build-up test on well No. W-175 (after Junkin et al., 1996).