Fluid substitution in laminated sands

Abstract

Petrophysicists and geophysicists do “fluid substitution” to estimate the density and sonic velocity of porous rocks at saturation conditions such as fully hydrocarbon-saturated or “fizz-bearing” with fluids such as gas, condensate, and oil. However, the Gassmann model implementation in many commercial packages sometimes yields implausible results for the compressional sonic velocity in laminated sand-shale sequences. In particular, the counterintuitive phenomenon of fluid effects increasing with computed shale fraction is often observed, as in the gas-bearing interval of a well drilled with oil-based mud shown in Figure 1. DTC FLUSUB (Gassmann fluid substitution) is the result of substituting DTC WET to the observed water saturation SW. As expected (Brie et al., 1995) the modeled fluid effect is larger than the difference between the wet and the observed log DTC OBS. The observed fluid effects reach a maximum in the cleanest sand with the highest porosity, but the model predicts a maximum in shaley sand with low porosity. Figure 1. Gassmann equation breakdown in laminated sequence. Normal practice is to specify a shale cutoff above which the computed fluid effects are ignored. This is an unsatisfactory workaround because the choice of cutoff is arbitrary, and results are counterintuitive even below the cutoff. Resolution and depth mismatches between wireline logs and the observation that the Gassmann equation does not apply in shale are frequently cited as reasons for these phenomena, but do not explain the modeled results in Figure 2. Figure 2. Model showing Gassmann equation breakdown. This model was generated from regional density, velocity, and porosity trends. A sand reservoir was defined by an abrupt transition at the base and a linear transition to shale over the top 100 ft. The clean part of the sand is not assumed totally free of clay. Sands in the study area are known to include some detrital …