DGS Luncheon - Rock Physics Diagnostics and Lithology Substitution for Seismic Reservoir Characterization; South TX and GOM

                  Gary Mavko, Stanford University

 

Well logs are commonly used in seismic interpretation, often to tie key lithologies or rock boundaries to amplitude events in the seismic section.  However, since the introduction of amplitude versus offset methods (AVO) to geophysics and dipole shear wave tools to well logging in the 1980s, well logs are also commonly used to estimate the expected amplitude change of the seismic events with offset. This is done by computing several synthetic seismograms with increasing source-receiver offset, typically by a ray-tracing method. In most cases AVO interpretation requires that multiple cases be modeled to represent different fluid saturation conditions in the reservoir (eg water, oil or gas filled pores).  These “fluid substitution” calculations are well established and reliable thanks to the versatility of the commonly used Biot-Gassmann fluid saturation equations which can be used to compute the changes in P-wave and S-wave velocity caused by changes in Sw.

 

In many cases however, it is not sufficient to just change the fluid saturation. Often we also need to model the seismic response to changing mineralogy or porosity of the reservoir.  In this case, Biot-Gassmann is no longer appropriate and some other set of governing equations must be used.  In general, for this “lithology substitution” problem we need an effective medium model that reasonably fits the data from the reservoir.

 

We present a method called Rock Physics Diagnostics (RPD).  RPD is the process of comparing data from the wells under investigation to a range of published or proprietary rock physics effective medium models.  After a suitable model is found, then we use that model to compute the change in Vp and Vs expected from a change in lithology or porosity.  If shear wave velocity is not available from the log suite, then RPD can also help determine the best method to predict Vs.

 

In this paper we describe a process to investigate the synthetic seismic response resulting from changes in porosity, minerals and fluid saturation of the reservoir. The workflow is configured to model changes in clay, quartz, calcite, and dolomite mineral volumes. The process computes ray-traced offset synthetics and half-space AVO models for several porosity-mineralogy cases, as well as the in-situ case. We demonstrate the method on well log data from the Gulf of Mexico and South Texas.