Professional Documents
Culture Documents
Model-Based Pore-Pressure Prediction in Shales: An Example From The Gulf of Mexico, North America
Model-Based Pore-Pressure Prediction in Shales: An Example From The Gulf of Mexico, North America
Model-Based Pore-Pressure Prediction in Shales: An Example From The Gulf of Mexico, North America
net/publication/314264461
CITATIONS READS
4 233
4 authors:
Some of the authors of this publication are also working on these related projects:
IGCCS-Induced-seismicity Geomechanics for Controlled CO2 Storage in the North Sea View project
All content following this page was uploaded by Tongcheng Han on 21 October 2020.
Tongcheng Han1, Marina Pervukhina1, Michael Ben Clennell1, and David Neil Dewhurst1
Manuscript received by the Editor 26 September 2016; revised manuscript received 6 December 2016; published online 06 March 2017.
1
CSIRO Energy Flagship, Kensington, Australia. E-mail: tongcheng.han@csiro.au; marina.pervukhina@csiro.au; ben.clennell@csiro.au; david.dewhurst@
csiro.au.
© 2017 Society of Exploration Geophysicists. All rights reserved.
M37
M38 Han et al.
sured well drilled in the Gulf of Mexico, North America (Figure 1). CPS MODEL
The velocities are simulated using the clay-plus-silt (CPS) model
proposed by Pervukhina et al. (2015) from independent wireline The CPS model is based on an experimental observation of a first-
measurements, such as the density and the neutron porosity logs. order linear correlation between the water fraction in the wet clay
The modeled shale velocities are compared with the velocities pack and the elastic coefficients of clays. The wet clay pack is a hypo-
measured by the sonic tool. The abnormal pore pressure is then pre- thetical composite material that includes all clay minerals and water
Downloaded 03/08/17 to 134.7.153.132. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
dicted based on the Eaton’s (1975) equation by using the simulated but without silt inclusions, and the linear correlation is independent
shale velocities as the apparent NVT. of the clay mineralogy and microstructure (Pervukhina et al., 2008,
2015). The anisotropic differential effective
5200 medium (DEM) model (Nishizawa, 1982) is then
used to add a silt (quartz) component into the wet
clay pack to simulate the CPS nature of shales (the
5400 composite of wet clay and silt). It is assumed that
pores in shales exist in the wet clay pack only, and
the silt particles are isolated within the clay ma-
5600
trix. Porosity and clay contents of the shales are
initially estimated from standard well-logging
Depth (m)
φcn − φcd
l 1 0% Cla
% Cla
ay
0%
Density porosity
al
0.6 where φn is the neutron porosity and φcn and φcd are the neutron
ica al 5
00
tic
re
i c
t
re
Th
eo
Figure 2.
0.4
Th
f c ¼ f 0c ð1 − φÞ and f s ¼ 1 − f c − φ: (3)
0.2
4) Computation of wet clay porosity (WCP) κ as a ratio of the vol-
ume of pores to the total volume of the wet clay constituent,
0 φ
0 0.2 0.4 0.6 0.8 1 κ¼ : (4)
fc þ φ
Neutron porosity
Figure 2. Neutron-density crossplot for the example well. A point 5) Calculation of wet clay elastic moduli ccij based on the exper-
with neutron porosity of 0.37 and density porosity of 0.07 is chosen imentally observed first-order linear dependency of clay elastic
as a 100% clay point, as indicated by the gray-filled circle. moduli on WCP (Pervukhina et al., 2008, 2015),
Shale overpressure prediction M39
3
pect ratio pores caused by the increasing pore pressure (decreasing
effective stress) can reduce the elastic moduli of a rock significantly,
resulting in a profound decrease in velocity (Eberhart-Phillips et al.,
2.5
1989), and therefore the modeled shale velocity should be higher be-
cause it does not take into account this low aspect ratio pore effect.
On the other hand, the total porosity may either be enhanced or un-
changed by overpressure, and the modeled shale velocity based on 2
this total porosity is therefore expected to decrease slightly or stay the
same, respectively, on top of the increased modeled velocity due to
the low-aspect-ratio pores. This not only explains the higher than 1.5
VP
measured CPS velocity in overpressured wells, but it also implies
VS
that the modeled velocity tends to be what the shale velocity should
be if the abnormal pore pressure has not changed the total porosity. 1
1 1.5 2 2.5 3 3.5 4
The above analysis paves the way for the employment of the CPS
Well-log determined velocity (km/s)
modeled velocity as the apparent NVT (see the “Discussion” section
for the difference between the apparent NVT and the NVT) for the Figure 3. A comparison between the CPS modeled and the well-log
pore-pressure prediction. determined P- and S-wave velocities in the example well.
M40 Han et al.
in the interbedded sands and were assumed to be in pressure equi- with the conventional Eaton (1975) model. The Eaton model de-
librium with the shales. pends on knowing the normal compaction trend for shale velocity.
To correctly determine the normal compaction trend, a large number
of wells with normally compacted sediments in the nearby area need
DISCUSSION
to be analyzed (Athy, 1930; van Ruth et al., 2004; Tingay et al.,
The proposed pore-pressure prediction method based on the 2009). For new exploration regions without many wells drilled, it
Downloaded 03/08/17 to 134.7.153.132. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
CPS-modeled shale velocity offers several advantages compared is difficult to predict such a normal compaction trend, and in this case
the proposed approach will be more appropriate to
a) 5200 b) predict pore pressure once the exponent n is cali-
brated against a well with known pore pressure.
Another advantage of the new method is that
5400 the CPS model on which the method is based cal-
culates the shale velocity using porosity and the
gross mineralogy difference of clay and quartz of
5600 a specific well. This makes the modeled shale
velocity more specific to the well of interest than
Depth (m)
In the formation investigated, depths are approximately 3300– almost constant (i.e., approximately 1.45). This justifies the single
4500 m below the seafloor with temperatures of 75°C–105°C, con- apparent Eaton exponent used for the prediction of pore pressure for
ditions, which usually result in a 50%–80% illitization of smectite the entire depth range.
(Hower et al., 1976). This generally validates the empirically se- In this work, the modeled shale velocity was integrated into Ea-
lected 100% clay point used in the neutron-density porosity cross- ton’s (1975) equation to predict pore pressure. There may be other
plot. For wells with ECS information measured or X-ray diffraction alternative models (Bowers, 1995; Dutta, 2002) that can be used for
Downloaded 03/08/17 to 134.7.153.132. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
data available on core samples, the clay point neutron and density the pore-pressure prediction. However, most of those models will
porosity can be better calibrated and more fully constrained. need more unknown parameters than does the Eaton equation,
As mentioned previously, the CPS modeled velocity is consid- which requires only one, i.e., the apparent Eaton exponent. Indeed,
ered to be what the velocity should be if the total porosity has not the Eaton equation is empirical and does not reveal the physical
been changed by overpressure, and it is used as the apparent NVT. relationship between the acoustic velocity and pressure (Eberhart-
This apparent NVT differs from the true NVT depending on the over- Phillips et al., 1989; Jones, 1995; Khaksar et al., 1999). Thus, the
pressure mechanisms that have distinct effects on the total porosity best way to accurately estimate pore pressure from the difference of
(Tingay et al., 2009; Zhang, 2011; Suwannasri et al., 2014). Because velocity in shales might be to conduct an experimental measure-
the CPS model calculates shale velocity through its total porosity, for ment of velocity at varying pore pressures. However, this is a time-
overpressured shales with no abnormal porosity (e.g., postdeposi- consuming experiment due to the extremely low permeability of
tional fluid expansion), the modeled velocity is greater than the mea- shales (Domnesteanu et al., 2002). The obtained relationships may
sured velocity and in this case, the apparent NVT is expected to be the not be universal, especially because shales can show frequency-de-
true NVT. On the other hand, for overpressures generated by mech- pendent behavior (Duranti et al., 2005; Fjaer et al., 2013) meaning
anisms with higher porosity (e.g., under-compaction), the higher that laboratory, wireline log, and seismic velocities may be different
porosity will make the difference between the modeled velocity for the same shale because they are typically obtained at ultrasonic
and the log measurement smaller, and therefore the apparent NVT (approximately 0.5–1.0 MHz), sonic (approximately 1–10 kHz),
is slower than the true NVT. This is confirmed by the results given and seismic (approximately 1–50 Hz) frequencies, respectively.
in Figure 5a for the example well with overpressure generated by Nevertheless, given the reasonable agreement between the predicted
and the measured pore pressure presented in this context, the Eaton
under-compaction, which is indicated by fitting the measured pore
equation based on the CPS modeled velocity can be used as a prac-
pressure using the Eaton equation with an exponent of 3.0 (Fig-
tical method for accurate pore-pressure prediction.
ure 5b), as demonstrated by Ebrom et al. (2003) for most under-com-
The proposed method predicts pore pressure through sonic logs
paction wells in the Gulf of Mexico. The difference between the
and the modeled shale velocity from density and neutron porosity
modeled shale velocity (i.e., the apparent NVT) and the true NVT
logs. All of these measurements are available from logging-while-
can be used to distinguish under-compaction from other overpressure
drilling (LWD) tools. Therefore, it is possible that the pore pressure
mechanisms that do not cause abnormal porosity.
can be predicted in real-time based on LWD data, provided that the
The apparent Eaton exponent n was determined to be 5.0 and 2.3
grain density and 100% clay neutron and density porosity for the
for P- and S-wave velocities, respectively, whereas the original Eaton
calculation of the shale velocity and the Eaton exponent for the es-
exponent χ used 3.0 for the P-wave velocity to fit the measured pore
timation of pore pressure are fully calibrated. Application of the
pressure. The difference between the two exponents arises due to the
current approach to real-time pore-pressure prediction based on
different NVTs, which in turn is a result of the under-compaction LWD data will be a topic for future study.
mechanism that enhances the shale porosity. As discussed above,
the enhanced porosity caused by under-compaction reconciles part
of the difference between the expected (i.e., the true NVT) and CONCLUSION
the measured velocity, making the apparent NVT slower than the true We have shown that the CPS-modeled velocities are systemati-
NVT. Hence, a higher exponent is needed to amplify the reduced cally higher than the log-measured velocities for 1200 m of over-
velocity difference. This confirms that a larger value of the exponent pressured shales in a deepwater well, Gulf of Mexico. The CPS
indicates the insensitivity of the velocities to changes in pore pressure model works through total porosity and gross mineralogy variables
(Ebrom et al., 2003). In this sense, the S-wave velocity is more sen- of clay- and silt-sized quartz of a shale, and it does not take into
sitive to the variation in pore pressure because the apparent Eaton account the pore-pressure effects; the modeled velocities are there-
exponent obtained for the S-wave is smaller than that for the P-wave. fore considered to be what the velocities should be if no abnormal
For overpressures with no abnormal porosity, the apparent NVT porosity is associated with the overpressure. Using the CPS-mod-
should coincide with the true NVT, and therefore the apparent Eaton eled velocities as the apparent normal compaction trend for shale
exponent n is expected to be equal to the original Eaton exponent χ. velocity, the abnormal pore pressure is ultimately predicted based
In addition to the different overpressure mechanisms, the degree on an exponential relationship between the pore pressure and the
of overpressure may also affects the apparent Eaton exponent. It is ratio of measured to modeled velocity. The pore pressure predicted
understandable that varying degrees of overpressure will lead to a using this new approach shows good agreement with direct pore-
difference between the measured and modeled velocities, resulting pressure measurements in the example well.
in changing apparent Eaton exponents. Therefore, a single value of
the exponent is insufficient to predict the pore pressure for overpres-
ACKNOWLEDGMENT
sures generated by the same mechanisms or for the entire section of
a single well (Sarker and Batzle, 2008). Fortunately, as shown in The authors would like to thank J. Krushin for kindly providing
Figure 4, the overpressure coefficient (defined as the ratio between the log data presented in this work and for his helpful comments on
the actual and hydrostatic pore pressure) for our example well is the manuscript.
M42 Han et al.
REFERENCES Krushin, J. T., 2014, Quantifying shale pore pressure by modeling the con-
trols on compaction and porosity: Interpretation, 2, no. 1, SB79–SB88,
doi: 10.1190/INT-2013-0110.1.
Athy, L. F., 1930, Density, porosity, and compaction of sedimentary rocks: La Vigne, J., M. Herron, and R. Hertzog, 1994, Density-neutron interpre-
AAPG Bulletin, 14, 1–24. tation in shaly sands: Presented at the SPWLA 35th Annual Logging
Bauluz, B., 2007, Illitization processes: Series of dioctahedral clays and Symposium, paper EEE.
mechanisms of formation, in F. Nieto, and J. Millan, eds., Diagenesis and Mavko, G., T. Mukerji, and J. Dvorkin, 2009, The rock physics handbook:
low-temperature metamorphism. Theory, methods and regional aspects: Cambridge University Press.
Downloaded 03/08/17 to 134.7.153.132. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
Seminarios de la Sociedad Española de Mineralogía, Sociedad Española McPherson, B., and G. Garven, 1999, Hydrodynamics and overpressure
de Mineralogía, 31–39. mechanisms in the Sacramento basin, California: American Journal of
Bowers, G. L., 1995, Pore pressure estimation from velocity data: Account- Science, 299, 429–466, doi: 10.2475/ajs.299.6.429.
ing for overpressure mechanisms besides undercompaction: SPE Drilling Meissl, S., J. Behrmann, and J. H. Behrmann, 2010, Data report: Preliminary
& Completion, 10, 89–95, doi: 10.2118/27488-PA. assessment of Pleistocene sediment strength in the Ursa Basin (Gulf of
Burst, J. F., 1969, Diagenesis of Gulf Coast clayey sediments and its possible Mexico continental slope) from triaxial and ring shear test data, in P.
relation to petroleum migration: AAPG Bulletin, 53, 73–93. B. Flemings, J. H. Behrmann, and C. M. John, and , the Expedition 308
Chopra, S., and A. Huffman, 2006, Velocity determination for pore-pressure Scientists, eds., Proceedings of IODP, 308: Integrated Ocean Drilling Pro-
prediction: The Leading Edge, 25, 1502–1515, doi: 10.1190/1.2405336. gram Management International Inc., 1–18.
Domnesteanu, P., C. McCann, and J. Sothcott, 2002, Velocity anisotropy and Mouchet, J. P., and A. Mitchell, 1989, Abnormal pressures while drilling:
attenuation of shale in under- and overpressured conditions: Geophysical Editions Technip.
Prospecting, 50, 487–503. Muggeridge, A., and H. Mahmode, 2012, Hydrodynamic aquifer or reser-
Duranti, L., R. Ewy, and R. Hofmann, 2005, Dispersive and attenuative voir compartmentalization?: AAPG Bulletin, 96, 315–336, doi: 10.1306/
nature of shales: Multiscale and multifrequency observations: 75th An- 06141110169.
nual International Meeting, SEG, Expanded Abstracts, 1577–1580. Nishizawa, O., 1982, Seismic velocity anisotropy in a medium containing
Dutta, N. C., 2002, Geopressure prediction using seismic data: Current status oriented cracks: Transversely isotropic case: Journal of Physics of the
and the road ahead: Geophysics, 67, 2012–2041, doi: 10.1190/1.1527101. Earth, 30, 331–347, doi: 10.4294/jpe1952.30.331.
Dvorkin, J., G. Mavko, and A. Nur, 1999, Overpressure detection from com- Pervukhina, M., D. N. Dewhurst, B. Gurevich, U. Kuila, A. F. Siggins, M.
pressional- and shear-wave data: Geophysical Research Letters, 26, 3417– Raven, and H. M. Hordgård-Bolås, 2008, Stress-dependent elastic proper-
3420, doi: 10.1029/1999GL008382. ties of shales: Measurement and modeling: The Leading Edge, 27, 772–
Eaton, B. A., 1975, The equation for geopressure prediction from well logs: 779, doi: 10.1190/1.2944162.
SPE of AIME, Paper 5544. Pervukhina, M., P. Golodoniuc, B. Gurevich, M. B. Clennell, D. N.
Eberhart-Phillips, D., D. H. Han, and M. D. Zoback, 1989, Empirical rela- Dewhurst, and H. M. Hordgård-Bolås, 2015, Prediction of sonic velocities
tionships among seismic velocity, effective pressure, porosity, and clay in shale from porosity and clay fraction obtained from logs: A North Sea
content in sandstone: Geophysics, 54, 82–89, doi: 10.1190/1.1442580. well case study: Geophysics, 80, no. 1, D1–D10, doi: 10.1190/
Ebrom, D., P. Heppard, M. Mueller, and L. Thomsen, 2003, Pore pressure geo2014-0044.1.
prediction from S-wave, C-wave, and P-wave velocities: 73rd Annual Inter- Sarker, R., and M. Batzle, 2008, Effective stress coefficient in shales and its
national Meeting, SEG, Expanded Abstracts, 1370–1373. applicability to Eaton’s equation: The Leading Edge, 27, 798–804, doi: 10
Elliott, W. C., and G. Matisoff, 1996, Evaluation of kinetic models for the .1190/1.2944165.
smectite to illite transformation: Clay and Clay Minerals, 44, 77–87, doi: Sayers, C. M., G. M. Johnson, and G. Denyer, 2002, Predrill pore pressure
10.1346/CCMN. prediction using seismic data: Geophysics, 67, 1286–1292, doi: 10.1190/
Ellis, D. V., and J. M. Singer, 2007, Well logging for earth scientists: 1.1500391.
Springer. Schneider, J., P. B. Flemings, B. Dugan, H. Long, and J. T. Germaine, 2009,
Fjaer, E., A. M. Stroisz, and R. M. Holt, 2013, Elastic dispersion derived from Overpressure and consolidation near the seafloor of Brazos-Trinity Basin
a combination of static and dynamic measurements: Rock Mechanics and IV, northwest deepwater Gulf of Mexico: Journal of Geophysical Re-
Rock Engineering, 46, 611–618, doi: 10.1007/s00603-013-0385-8. search, 114, B05102.
Gutierrez, M. A., N. R. Braunsdore, and B. A. Couzens, 2006, Calibration Skempton, A. W., 1970, The consolidation of clays by gravitational com-
and ranking of porepressure prediction models: The Leading Edge, 25, paction: Quarterly Journal of the Geological Society of London, 125,
1516–1523, doi: 10.1190/1.2405337. 373–411, doi: 10.1144/gsjgs.125.1.0373.
Hermanrud, C., L. Wensas, G. M. G. Teige, H. M. Nordgård Bolås, S. Han- Suwannasri, K., W. Promrak, S. Utitsan, V. Chaisomboonpan, R. J. Groot,
sen, and E. Vik, 1998, Shale porosities from well logs on Haltenbanken H. I. Sognnes, and C. K. Morley, 2014, Reducing the variation of Eaton’s
(offshore mid-Norway) show no influence of overpressuring, in B. E. exponent for overpressure prediction in a basin affected by multiple over-
Law, G. F. Ulmishek, and V. I. Slavin, eds., Abnormal pressures in hydro- pressure mechanisms: Interpretation, 2, no. 1, SB57–SB68, doi: 10.1190/
carbon environments: AAPG Memoirs, 65–85. INT-2013-0100.1.
Hower, J., E. V. Eslinger, M. Hower, and E. A. Perry, 1976, Mechanism of Swarbrick, R. E., and M. J. Osborne, 1998, Mechanisms that generate ab-
burial metamorphism of argillaceous sediment. 1: Mineralogical and chemi- normal pressures: An overview, in B. E. Law, G. F. Ulmishek, and V. I.
cal evidence: Geological Society of America Bulletin, 87, 725–737, doi: 10 Slavin, eds., Abnormal pressures in hydrocarbon environments: AAPG,
.1130/0016-7606(1976)87<725:MOBMOA>2.0.CO;2. 13–34.
Jones, S. M., 1995, Velocities and quality factors of sedimentary rocks at Tingay, M. R. P., R. R. Hillis, R. E. Swarbrick, C. K. Morley, and A. R.
low and high effective pressures: Geophysical Journal International, 123, Damit, 2009, Origin of overpressure and pore-pressure prediction in the
774–780, doi: 10.1111/j.1365-246X.1995.tb06889.x. Baram province, Brunei: AAPG Bulletin, 93, 51–74, doi: 10.1306/
Khaksar, A., C. M. Griffiths, and C. McCann, 1999, Compressional and 08080808016.
shear-wave velocities as a function of confining stress in dry sandstones: van Ruth, P., R. Hillis, and P. Tingate, 2004, The origin of overpressure in
Geophysical Prospecting, 47, 487–508, doi: 10.1046/j.1365-2478.1999 the Carnarvon Basin, Western Australia: Implications for pore pressure
.00146.x. prediction: Petroleum Geoscience, 10, 247–257, doi: 10.1144/1354-
Krushin, J. T., 2005, Quantifying shale porosity — A thermodynamically 079302-562.
based, predictive model which includes the effects of mechanical compac- Zhang, J., 2011, Pore pressure prediction from well logs: Methods, modi-
tion, temperature, mineralogy, and chemical diagenesis: Gulf Coast As- fications, and new approaches: Earth-Science Reviews, 108, 50–63, doi:
sociation of Geological Societies Transactions, 55, 401–414. 10.1016/j.earscirev.2011.06.001.