Practical Gassmann fluid substitution in sand/shale sequences

When performing fluid substitution on log data Gassmann’s (1951) model is the general tool of choice. Whilst the equations and software implementations are straightforward, there are potential pitfalls. Several authors have noted the need for quality control (QC) of Gassmann input parameters (e.g., Smith et al., 2003; Han and Batzle, 2004; and Skelt, 2004 have described the problem of erroneous fluid substitution results in laminated sand/shale intervals). Despite the warnings, however, it is not clear what the interpreter should do in the application of Gassmann to mitigate the chances of erroneous results. Nor is it clear, for example, in the case of shaley sand scenarios, what the right answer should be. The approach taken here is essentially pragmatic, suggesting an adaptive Gassmann workflow that can be modified to give intuitive results. The paper illustrates the need to QC the dry rock data implicit in Gassmann’s model and to condition the dry rock data used in fluid substitution. The following discussion assumes that good quality log data, including logs of shear wave velocity Vs, are available. No discussion will be made here of issues such as log editing and Vs prediction.