Bulk Modulus in Petroleum Rock Physics: Gassmann Fluid Substitution and Seismic Velocity Prediction for WCSB Montney and Cardium AVO Analysis

Key Takeaways

• Gassmann fluid substitution workflow for predicting P-wave velocity change between brine-saturated and gas-saturated WCSB Montney siltstone at reservoir conditions: The practical Gassmann fluid substitution procedure begins with in-situ log measurements of Vp (compressional wave velocity from sonic log), Vs (shear wave velocity from dipole sonic), and density (rho from density log), from which the saturated bulk modulus is computed: Ksat = rho × (Vp^2 - 4Vs^2/3). The dry frame modulus Kdry is then back-calculated by rearranging the Gassmann equation, using laboratory- or database-derived Kmineral appropriate for the mineral composition determined from XRD core analysis of WCSB Montney samples (typically quartz-dominated with Kmineral approximately 37-40 GPa). With Kdry established, the fluid is substituted by replacing Kfluid,brine with Kfluid,gas (calculated from the Batzle-Wang equations using in-situ pressure, temperature, and gas gravity) and re-applying the Gassmann equation to obtain the new Ksat,gas, from which Vp,gas = sqrt((Ksat,gas + 4G/3) / rho_gas) is predicted, where rho_gas accounts for the density change when brine is replaced by lighter gas. For a WCSB Montney siltstone at 2,800 m depth with phi = 0.10, Kmineral = 38 GPa, Kfluid,brine = 2.2 GPa, Kfluid,gas = 0.055 GPa (dry gas at 35 MPa reservoir pressure), the Gassmann substitution predicts Vp decreases from 4,280 m/s (brine-saturated) to 4,050 m/s (gas-saturated), a 5.4% reduction that translates to a measurable seismic reflection amplitude change for AVO Class II-III analysis.
• Bulk modulus of common WCSB reservoir minerals and fluids and their role in seismic velocity sensitivity to fluid changes: The sensitivity of seismic velocity to fluid substitution depends on the ratio of the pore-fluid bulk modulus to the mineral frame bulk modulus: when Kfluid is much smaller than Kmineral (as for gas), the Gassmann effect is large; when Kfluid approaches Kmineral (as for stiff brine at high pressure), the Gassmann effect is small. For WCSB reservoir rocks: quartz arenite sandstone (Kmineral 36-38 GPa; high Gassmann sensitivity to gas); dolomite (Kmineral 52-55 GPa; moderate sensitivity); calcite carbonate (Kmineral 68-77 GPa; lower sensitivity because the high mineral modulus overwhelms the fluid modulus change). Pore fluid moduli at typical WCSB reservoir conditions (35 MPa pressure, 80 degrees C): fresh water 2.5 GPa; 100,000 mg/L brine 2.9 GPa; 35-degree API oil at 100 m3/m3 GOR approximately 0.8 GPa; dry gas (methane) at 35 MPa approximately 0.055 GPa; CO2 near critical point approximately 0.01-0.10 GPa (highly pressure-temperature-sensitive). The low gas bulk modulus means that even 10-20% gas saturation (partial gas saturation) reduces Kfluid substantially from the brine value, and the Vp reduction for 20% gas saturation is approximately 70-80% of the Vp reduction for 100% gas saturation, explaining why partial gas saturation (from gas cap invasion or biogenic gas) produces amplitude anomalies that can be misidentified as full gas pay in WCSB Cretaceous bright spot interpretation.
• Dry frame bulk modulus (Kdry) estimation for WCSB tight formations from empirical correlations and laboratory ultrasonics when sonic log data are unavailable: The dry frame bulk modulus Kdry is not directly measurable from wireline logs (logs always measure the saturated rock) and must be either back-calculated from Gassmann (requiring reliable Vp, Vs, and rho logs) or estimated from empirical correlations and laboratory measurements. For WCSB Montney tight siltstone (porosity 3-10%, clay content 15-35%), Kdry is commonly estimated from the critical porosity model (Kdry = Kmineral × (1 - phi/phi_critical) for phi below phi_critical approximately 0.35 for siltstone) or from Biot's consolidation theory relating Kdry to porosity and cementation. Laboratory ultrasonic measurements on WCSB Montney core plugs under simulated reservoir stress (35-40 MPa effective stress) provide the most reliable Kdry values: typical WCSB Montney siltstone measurements show Kdry of 18-30 GPa (lower than the Kmineral of 37 GPa, reflecting the compliance of the pore space and grain-grain contacts) and G of 15-22 GPa. These laboratory values are used to calibrate the Gassmann fluid substitution for regional WCSB Montney seismic inversion workflows that must predict elastic properties in undrilled locations from surface seismic attributes alone.
• AVO classification and bulk modulus contrast at WCSB gas sand reflectors: how Gassmann predicts Class II and III AVO response: AVO (amplitude versus offset) analysis uses the fact that the reflection coefficient at a formation boundary depends on both compressional impedance (Ip = rho × Vp) and shear impedance (Is = rho × Vs) contrasts, with the P-P reflection coefficient varying with the angle of incidence according to the Shuey linearized approximation: R(theta) = R0 + G × sin^2(theta) + F × tan^2(theta) × sin^2(theta), where R0 is the zero-offset reflectivity and G is the AVO gradient. The Gassmann-predicted change in Vp upon gas substitution determines R0 (zero-offset bright spot amplitude), while the shear modulus G (unchanged by fluid substitution in the Gassmann framework, since G = rho × Vs^2 is independent of pore fluid bulk modulus) determines the amplitude variation with offset through the Vs/Vp ratio. WCSB Cretaceous Viking and Cardium gas sands with low Vp/Vs ratios (approximately 1.5-1.7 for gas-saturated porous sandstone) are Class III AVO reflectors (increasingly negative amplitude at far offset) when embedded in shale with higher Vp/Vs (approximately 2.0-2.4), enabling AVO analysis of pre-stack seismic gathers to distinguish gas-bearing from brine-bearing sandstone without drilling. The Gassmann equations quantify the expected bulk modulus and Vp change for the gas substitution, providing the theoretical prediction that is matched against the measured AVO response to assess likelihood of gas saturation versus other anomaly causes (lithology variation, tuning).
• Bulk modulus role in WCSB seismic inversion: simultaneous inversion for P-impedance, S-impedance, and density as proxy for lithology and fluid discrimination: Seismic inversion transforms the amplitude-versus-offset information in WCSB 3D seismic gathers into estimates of elastic properties (Ip, Is, and density) at each subsurface point, from which Ksat and G can be computed and fluid substitution applied to distinguish gas from brine and sand from shale. Simultaneous pre-stack inversion for WCSB Montney and Duvernay produces three elastic parameter cubes: Ip (sensitive to both lithology and fluid), Is (primarily sensitive to lithology since shear velocity is relatively fluid-independent by Gassmann), and density (sensitive to both). The lambda-rho (incompressibility times density) attribute, defined as lambda-rho = Ip^2 - 2 × Is^2 (proportional to (Ksat + G/3 - 2G/3) × rho = (Ksat - 4G/3) × rho), is particularly sensitive to fluid content because it isolates the contribution of bulk modulus to the P-wave velocity: gas-saturated rocks have low lambda-rho (low Ksat from gas), while brine-saturated rocks have high lambda-rho. WCSB Duvernay exploration programs have used lambda-rho inversion derived from Gassmann-calibrated rock physics models to rank prospects by fluid likelihood, with lambda-rho below 15 GPa × g/cc identified as prospective for gas condensate and lambda-rho above 25 GPa × g/cc interpreted as brine-saturated within the Duvernay carbonate equivalent zone.