OTC 13103 Stress-Controlled Porosity in Overpressured Sands at Bullwinkle (GC65), Deepwater Gulf of Mexico P.B. Flemings, J. Comisky, X. Liu, and J.A. Lupa, Pennsylvania State University Copyright 2001, Offshore Technology Conference This paper was prepared for presentation at the 2001 Offshore Technology Conference held in Houston, Texas, 30 April 3 May 2001. This paper was selected for presentation by the OTC Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Offshore Technology Conference or its officers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Offshore Technology Conference is 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 acknowledgment of where and by whom the paper was presented. Abstract Porosities, calculated from wireline density logs, decrease from 33% to 30% over a vertical depth range of 1200 feet in the J3 sand of the Bullwinkle Field (GC65), Deepwater Gulf of Mexico. This 10% decrease in porosity results in a 58% decline in permeability (3.3 to 1.4 darcies). The spatial variation in porosity results in a 75% decrease in acoustic reflectivity over a depth range of only 500 feet. The decrease in porosity with depth is interpreted to result from compaction over geological time scales, which results from higher in-situ stresses at the reservoir low points than at the highpoints. The study provides an approach to predict the spatial variation of porosity and permeability expected in a homogenous, isotropic, and compressible reservoir. These rock properties are important parameters for reservoir simulation. The acoustic results imply that the reservoir will be imaged differently due to in-situ stress conditions alone and independent of changes in lithology or pore fluid. Introduction The Bullwinkle oil field is located on the western flank of a circular salt-withdrawal mini-basin on the slope of the Gulf of Mexico, approximately 150 miles to the southwest of New Orleans, Louisiana (Figure 1). The mini-basin is primarily located in Green Canyon blocks 65 and 109, in approximately 1350 ft. water depth (412 m). Holman and Roberston 1 and O Connell et al. 2 describe the Bullwinkle field; Kikani and Smith 3 describe the Rocky field which lies on the southeastern margin of the mini-basin. The J sands are of early Nebraskan (3.35 Ma) age and host the majority of the reserves at Bullwinkle. 1 The 5 sands (J0-J4) are bowl-shaped interconnected channel and sheet turbidite sands that are interbedded with debris flow deposits and shales, and overlain by a thick section (500 ft.) of bathyal shales. The top of the section is characterized by multiple erosional unconformities, suggesting a period of sediment bypass in the basin. 1 Pressure drawdown at each well followed the same depletion curve, indicating that all the sands are in pressure communication. 1,3 The J3 sand is located on the western flank of the minibasin (Figure 2). It is very fine- to fine-grained and has a blocky log character (Figure 3). The lithology and the grain size are relatively homogenous across the field. It is interpreted to be a ponded, internally amalgamated, sheet sand. 1 There is a small oil and gas pool at its crest, while the majority of it is brine saturated (Figure 2). It has been penetrated by 23 wells. Porosity Analyis in the J3 Sand Whole core was taken in the J3 sand at the A-32-BP, which is beneath the J3 oil-water contact (Figure 2). In this well, density porosities (DPHI) correlate with whole core porosities when a fluid density of 1.05 g/cc and a grain density of 2.65 g/cc are assumed (Figure 4). We interpret that 1.05 g/cc is the density of the pore fluid measured by the bulk density tool in the water-saturated portion of the J3. This is less than the brine density that is estimated from the measured salinity (200,000 ppm) to be 1.18 g/cc at reservoir conditions (160 F and 8400 psi). The lower density for the formation fluid is most likely due to invasion of the filtrate into the formation. Representative values for J3 sand wireline neutron porosity (NPHI) and wireline density porosity (DPHI) were determined by taking average values wherever the shale content was less than 10%. The neutron porosity declines linearly with depth from 0.36 at the highest penetration to as low as 0.30 at the lowest penetration (Figure 5). Beneath the oil-water contact, the neutron porosity is generally one or two porosity units greater than the density porosity calculated with a fluid density of 1.05g/cc. We infer that this results from some clay being present in the sand. Above the oil-water contact, the density porosity (assuming ρ f = 1.05 g/cc) jumps abruptly to values that approximately equal the neutron porosity (Figure 5). We attribute the high values of density porosity above the oil-
2 P.B. FLEMINGS, J. COMISKY, X. LIU, J.A. LUPA OTC 13103 water contact to result from an overestimate of fluid density. To correct for this effect, we developed a correlation between the neutron porosity and the density porosity beneath the oilwater contact (Figure 6, dashed line). We then modified the fluid density above the oil water contact to match this trend (Figure 6, solid line). A value of ρ f = 0.94g/cc successfully reproduced the NPHI-DPHI correlation observed beneath the oil water contact (Figure 6). The true pore fluid density calculated from known oil density (0.68 g/cc) and brine density (1.28 g/cc) and water saturation (0.12) is 0.74 g/cc. The higher density of 0.94 g/cc recorded by the density log is most likely due to invasion of the filtrate into the formation. Lighter oils and gas have an effect on the NPHI reading because they have a lower hydrogen concentration than water. This effect can be considered using empirical corrections that adjust the NPHI reading by an amount NPHI 4. We have calculated an upper bound of NPHI = 0.5 porosity units for neutron porosity measurements in the oil leg. We did not use this relatively insignificant correction to the NPHI data in Figure 6. Accounting for this effect would make the change in porosity with depth even greater than what we estimate here. The final porosity predictions are illustrated in Figure 5 and Table 1. The porosity of the J3 sand declines from approximately 33% to 30% over a vertical depth of 1200 feet. This remarkable change in porosity with depth has significant implications for a range of processes. Porosity and Permeability Distribution in the J3 Sand The porosity distribution is contoured across the J3 sand (Figure 7). As expected, high porosities correlate with structural highs (compare Figure 7 and Figure 2). Measurements of permeability vs. porosity in one of the J sands (Figure 8) indicate an exponential relationship between porosity and permeability: log K = 12.8Φ 0.7 Eq. 1 Thus, while porosity decreases from 33% to 30% (10%), permeability decreases from 3.3 to 1.4 darcies (58%) (Figure 9). Ostermeier 5,6,7 documented that the reduction in permeability is typically 4 to 5 times greater than that in porosity in deepwater turbidites of the Gulf of Mexico. Compressibility Derived From Deformation Experiments and Wireline Logs Pore pressures in the J3 sand are overpressured and reach 80% of the overburden stress (S v ) (Figure 10). The vertical effective stress (σ v =S v P) is a measure of the stress the rock frame is under. This stress increases with depth in the reservoir because the fluid pressure gradient is lower than the overburden gradient (Figure 10). At the crest of the J3 sand, σ v is 1155 psi less than at the low point of the sand (Figure 10). a pore compressibility, C p : p = V V 1 Φ = σ Φ(1 Φ) σ 1 p C Eq. 2 and a bulk compressibility, C b : b V p V 1 Φ = σ (1 Φ) σ 1 b C Eq. 3 b are calculated from the log-based porosities (Figure 11, Table 1) and the vertical effective stress in the sand (Figure 10). The resulting value of C b and C p are 85 x 10-6 psi -1 and 283 x 10-6 psi -1, respectively. Pore compressibilities of the J sand that are derived from laboratory deformation vary as a function of stress state and range from 30 x 10-6 psi -1 to 75 x 10-6 psi -1 (Figure 12). Thus, the lab-derived compressibilities are one order of magnitude less than the log-derived compressibilities. It is not uncommon for lab-derived compressibilities performed over short time-scales to be significantly smaller than observed porosity-stress relationships observed in sedimentary basins. 8 Most likely there is time-dependent creep that occurs over geologic timescales that results in these different apparent compressibilities 9. The in-situ vertical effective stress for this sample is 2310 psi, which is 500 to 1000 psi lower than the isostatic stresses where the sample exhibits strain-softening 5,6 (increasing compressibility) (Figure 12). Ostermeier presented a possible mechanism for the stress-dependent compressibilities that are observed. Significance We propose that the decline in porosity with depth is due to the increase in effective stress that occurs with depth in the J3 sand (Figure 13). In this simple mechanical model, the bulk rock is envisioned as a spring that is more compressed at its deeper end where the overburden and the effective stress are greatest. In contrast, at the crest of the reservoir, the overburden and effective stresses are lowest and the rock is least compacted. As in any dipping permeable body, this results from the fact that the pore fluid pressure gradient is less than the overburden gradient (Figure 13). The decline in porosity and permeability is interpreted to result from the changing stress state in the reservoir and not because of lithologic change (e.g. composition, size, or sorting). The model can be used to predict the spatial variation of porosity and permeability in highly compressible deepwater turbidite sands where limited data are available. In general, the porosity and permeability will be highest at the reservoir crests and that these properties will decline in deeper parts of the sand. A better understanding of these reservoir properties may allow more accurate predictions of reserves, strengthen reservoir models, and assist exploitation decisions.
OTC 13103 STRESS-CONTROLLED POROSITY IN OVERPRESSURED SANDS AT BULLWINKLE (GC65), DEEPWATER GOM 3 Ultimately, it may be possible to relate wireline estimates of compressibility to the compressibility that will be present during reservoir production. This is potentially important because compaction is an important control on reservoir drive and a better understanding of this parameter may allow better exploitation decisions. The observation that experimental compressibilities are one order of magnitude less than the compressibilities observed in wireline data is not uncommon 8. We interpret that over geologic timescales there is greater strain than over experimental timescales. 5 One of the most striking features of the change in porosity with depth is the degree to which acoustic reflectivity is affected (Figure 14). Beneath the oil-water contact, the impedance in the J3 sand declines 12% and the reflectivity declines 84% over a depth range of only 500 ft. This remarkable decline in reflectivity implies that seismic images of dipping sand bodies will show sharp changes that parallel structure independent of changes in lithology or pore fluid composition. Conclusions Porosities decrease from 33% to 30% over a vertical depth range of 1200 feet in the J3 sand of the Bullwinkle Field (GC 65). This 10% decrease in porosity results in a 58% decline in permeability (3.3 to 1.4 darcies). The decrease in porosity results in a 75% decrease in acoustic reflectivity. The decrease in porosity with depth is interpreted to result from compaction over geological time scales resulting from higher in-situ stresses at the reservoir low points than at the highpoints. The study provides a methodology to predict the spatial variation of porosity and permeability in a dipping, isotropic, and compressible reservoir. Nomenclature C b = bulk compressibility (M/LT 2 ) C p = pore compressibility(m/lt 2 ) DPHI = density porosity (-) K = permeability (L 2 ) NPHI = neutron porosity (-) P = fluid pressure (M/LT 2 ) vel bsd = bulk velocity of sand (L/T) vel bsh = bulk velocity of shale (L/T) Vb = bulk volume (L 3 ) Vp = pore volume (L 3 ) φ = porosity(-) ρ f = fluid density (M/L 3 ) ρ b = bulk density (M/L 3 ) ρ bsh = bulk density of shale (M/L 3 ) ρ bsd = bulk density of sand (M/L 3 ) ρ g = solid grain density (M/L 3 ) S v = overburden stress (M/LT 2 ) σ v = vertical effective stress (M/LT 2 ) σ = effective stress (M/LT 2 ) Acknowledgements This project is part of the Penn State GeoSystems Initiative and was also supported by the Penn State GeoFluids Consortium (an industry-academic research consortium of 10 petroleum companies). All data were provided by Shell Exploration and Production Company. We thank John Dacy (Core Laboratories) for providing insight into correlating neutron and density data. We thank R.M. Ostermeier for providing a thoughtful review of this manuscript. We thank H. Johnson for assistance in manuscript presentation References 1. Holman, W. E., and Robertson, S. S., 1994. Field Development, Depositional Model, and Production Performance of the Turbiditic J Sands at Prospect Bullwinkle, Green Canyon 65 Field, Outer-Shelf Gulf of Mexico, GCSSEPM Foundation 15th Annual Research Conference, Submarine Fans and Turbidite Systems, December 4-7, p. 139-150. 2. O Connell, J. K., Kohli, M., and Amos, S., 1993. Bullwinkle: A unique 3-D experiment, Geophysics, v. 58, No. 1, p. 167-176. 3. Kikani, J. and Smith, T., 1996. Recovery Optimization and Modeling Depletion and Fault Block Differential at Green Canyon 110, Society of Petroleum Engineers, 36693, p. 157-170. 4. Gaymard, R., and Poupon, A., 1968, Response of neutron and formation density logs in hydrocarbon bearing formations, The Log Analyst, v.9, No.5, p. 3-12. 5. Ostermeier, R.M., 1995, Deepwater Gulf of Mexico Turbidites Compaction Effects on Porosity and Permeability. SPE Formation Evaluation, v. 10, No. 2, pp. 79-85. 6. Ostermeier, R. M., in press, Compaction Effects on Porosity and Permeability Deepwater Gulf of Mexico Turbidites, Journal of Petroleum Technology. 7. Ostermeier, R.M., 1996, Stressed Oil Permeability of Deepwater Gulf of Mexico Turbidite Sands: Measurements and Theory. SPE Formation Evaluation, 229. 8. Stump, B.B. and Flemings, P.B., in press. Consolidation State, Permeability, and Stress Ratio as Determined from Uniaxial Strain Experiments on Mud Samples from the Eugene Island 330 area, Offshore Louisiana: Alan R. Huffman, ed. AAPG Special Volume. 9. Dudley, J.W., Myers, M.T., Shew, R.D., and Arasteh, M.M., 1998, Measuring Compaction and Compressibilities in Unconsolidated Reservoir Materials by Time-Scaling Creep, SPE Reservoir Evaluation & Engineering, 430-434.
4 P.B. FLEMINGS, J. COMISKY, X. LIU, J.A. LUPA OTC 13103 Table 1: J3 Sand Petrophysical Data (GC 109 & 65) Well MD TVDSS RHOB NPHI DPHI S v P s v feet feet g/cc psi psi psi A-41 14480 11532 2.093 0.342 0.326 10078 8261 1817 A-11 BP 14367 11553 2.100 0.334 0.322 10097 8268 1829 A-3 BP 14141 11558 2.082 0.358 0.332 10102 8269 1833 A-37 13639 11667 2.085 0.347 0.330 10202 8303 1899 A-1 13618 11776 2.089 0.310 0.328 10302 8337 1965 A-34 15400 11913 2.132 0.337 0.324 10427 8389 2038 A-2 BP 13828 12025 2.131 0.322 0.324 10529 8441 2088 109-1 12198 12124 2.135 0.330 0.322 10620 8487 2133 A-32 BP 13034 12126 2.145 0.322 0.316 10622 8488 2134 A-5 BP 15286 12228 2.140 0.330 0.319 10715 8535 2180 A-38 12758 12314 2.152 0.300 0.311 10794 8576 2218 A-60 14472 12329 2.166 0.311 0.303 10807 8582 2225 A-4 BP 12780 12427 2.113 0.312 0.336 10898 8628 2270 65-1 ST 13165 12541 2.154 0.320 0.310 11002 8681 2321 A-36 14422 12728 2.168 0.299 0.301 11173 8768 2405 Figure 1: The Bullwinkle Basin (Green Canyon 65 and 109) is on the upper continental slope in approximately 1300 ft. water depth (400m) approximately 150 miles southwest of New Orleans, LA. Figure 2: Map of the depth of the J3 sand in feet beneath the sea surface. The polygon marks the boundary of the J3 sand. The structural crest of the sand is along the western margin and the oil-water contact is at 11,850 feet (dashed line).
OTC 13103 STRESS-CONTROLLED POROSITY IN OVERPRESSURED SANDS AT BULLWINKLE (GC65), DEEPWATER GOM 5 Figure 3: Whole cores were taken within the J sands at the A-32- BP (located in Figure 2). The J3 sand has a blocky log character. Figure 5: Log-based porosity measurements in the J3 sand vs. depth beneath the sea surface (TVDSS) for 14 wells across the Bullwinkle field. Density porosity (DPHI) and neutron porosity (NPHI) are shown. Diamonds illustrate the density porosity assuming r f = 1.05 g/cc and r s = 2.65 g/cc. Circles (plotted above the oil water contact) show the density porosity assuming r f = 0.94 g/cc and r s = 2.65 g/cc. Figure 4: Cross plot of measured porosity from the A-32-BP core (Core Porosity) vs the porosity predicted by the density log assuming a pore fluid density of 1.05g/cc (DPHI) for the J3 sand. The straight line is a one to one correlation. Core porosities were measured under isostatic loading conditions at 2100 psi. Figure 6: Log-derived density porosity (DPHI) vs. log-derived neutron porosity (NPHI) for each penetration of the J3 sand. Diamonds illustrate the density porosity assuming r f = 1.05 g/cc and r s = 2.65 g/cc. Circles (plotted for density measurements made above the oil-water contact) show the density porosity assuming r f = 0.94 g/cc and r s = 2.65 g/cc. Dashed line is a leastsquares linear regression of the DPHI vs. NPHI measurements beneath the oil water contact. The data point at NPHI=0.31, DPHI=0.328 was excluded in the regression. The solid line is the regression of the circles (r f = 0.94 g/cc). It was found that a fluid density of 0.94 g/cc resulted in values that most closely matched the trend shown with the dashed line.
6 P.B. FLEMINGS, J. COMISKY, X. LIU, J.A. LUPA OTC 13103 Figure 7: Porosity distribution in the J3 sand based on correlation developed in Figure 6. The map is contoured to exactly match the data at the well penetrations (Table 1). Away from the well penetrations, the map is contoured to match the trend of decreasing porosity with depth that is observed (Figure 5). Figure 9: Permeability distribution in the J sand in units of millidarcies. Permeability decreases from 3600 millidarcies to 800 darcies across the J3 sand. The map is contoured to exactly match the data at the well penetrations. Away from the well penetrations, the map is contoured to match the trend of decreasing permeability resulting from the declining porosity (Figure 7). Figure 8: Brine permeability vs. porosity derived from deformation experiments on a single sample from the J sands. The solid line is Eq. 1 from the text. Figure 10: Pressure vs. depth in the J3 sand at GC 65. The hydrostatic pressure ( Hydrostat ) is calculated assuming a pressure gradient of 0.465 psi/ft. The overburden stress ( Lithostat (Sv) ) is calculated assuming an overburden gradient (pressure divided by depth beneath sea level) of 0.83 psi/ft. At the structural crest (11,000 ft), the vertical effective stress (s v=s v-p) is 1373 psi (9.5MPa) whereas at the bottom of the structure (13,000 feet) the vertical effective stress is 2528 psi (17.4 Mpa).
OTC 13103 STRESS-CONTROLLED POROSITY IN OVERPRESSURED SANDS AT BULLWINKLE (GC65), DEEPWATER GOM 7 Figure 13: Diagram illustrating geomechanical model to describe the observed porosity distribution in the J3 sand. Figure 11: Porosity vs vertical effective stress (s v) in the J3 reservoir. Porosity declines linearly with increasing effective stress. On the right axis is the depth in the reservoir. This is not linear because above the oil water contact the fluid pressure increases along the oil gradient (0.3 psi/ft) whereas below the oil water contact the fluid pressure increases along the water gradient (0.465 psi/ft). The linear regression to the data does not include the outlying A-4-BP, which has an anonymously high porosity for its depth. Figure 12: Compressibility vs. effective stress in one J sand sample derived experimentally under isostatic stress. Figure 14: Top: Impedance (r b*vel) vs. depth for the J3 sand (solid line) and the overlying shale (dashed line). The shale impedance is found to be approximately constant at each penetration whereas the impedance in the sand increases with depth due to the decline in porosity. Bottom: Reflectivity [(r bsd*vel bsd-r bsh*vel bsh)/(r bsd*vel bsd+r bsh*vel sh)] vs. depth in the J3 sand. The reflectivity decreases six-fold in the water leg of the J3 sand.