It has been well documented that reservoirs show sometimes staggering amounts of compaction over years of production. The aquifer under Las Vegas, the Ecofisk development, and the Lake Maracaibo area are all examples where the subsidence of porous reservoirs have undergone a compaction observable to the naked eye.
Studies of the elastic deformation of the surrounding country rock have shown the distribution of failure types to be associated with the volumetric collapse of a reservoir structure Segall (1998). However, the prediction of the nature and timing of the compaction of the reservoir itself need also be understood. Zoback (2001) has made inquiries into the nature of the compaction of granites, Ottawa sand, and the Adamswiller sandstone. This work focuses on an elastic deformation model that does not take into account time progression and creep behavior.
The motivation for the study of the creep behavior of reservoir rocks is a fundamental problem of available time. This uncertainty manifests itself in two questions: Over what time scale do reservoirs compact? Can lab measurements capture meaningful parameters to understand the nature and development of reservoir compaction? Simply put, are compaction studies measured in the lab valid at all? After these difficult questions are answered, then next logical step is to ask what impact this can have on repeat seismic experiments that are all the rage in the new millennium.
With the publishing of lab compaction data in Dudley and Myers (1994), the case is clear that meaningful measurements can be made despite implementation on time scales orders of magnitude shorter than significant to either geologic time or reservoir life. This results in our ability to make and trust compaction measurements on unconsolidated sand material. While this may solve engineering problems such as pore volume compressability and drive, it introduces a major wrinkle into the 4D seismic experiment that must be recognized. As a reservoir continues to compact over a decade of production, the best models for fluid substitution or pressure dependence of a seismic attribute are meaningless if the rock frame has changed significantly.