- Copyright © 2001 Society of Exploration Geophysicists
Today's geophysicists work ever more closely with reservoir engineers and geologists due to the advent of geophysical techniques for reservoir surveillance. Geophysicists have transcended their traditional task of exploration in order to monitor reservoir performance. They are thus called upon to understand the nature of fluid flow within a reservoir and the relationship between reservoir/fluid properties and geophysical observations. It would seem that the physics of fluid flow within a reservoir is, for the most part, unrelated to the phenomena traditionally studied by exploration geophysicists, such as wave propagation within the earth. However, while the physics of fluid flow and elastic-wave propagation are distinct, many shared concepts are extremely useful in reservoir characterization and performance analysis.
In this paper we discuss a common framework for treating both reservoir engineering and geophysical data, based on asymptotic solutions of the governing differential equations. The asymptotic formulation leads to a practical and efficient methodology, akin to geophysical tomography, for reservoir characterization based on field production data.
Computational challenges in reservoir characterization
Reservoir characterization is an inherently multidisciplinary exercise. Building a high-resolution description of a petroleum reservoir combines the skills of geoscientists and engineers to merge the variety of data collected over its producing life. Such data can be broadly classified as either static or dynamic. Static data consist of time-invariant information such as cores, well-logs, 3-D seismic, and geologic interpretation. Dynamic data are time-varying quantities related to fluid movement within the reservoir and include such information as transient pressure observations, tracer data, multiphase production history, and time-lapse seismic measurements. Reconciling static and dynamic data from a reservoir is an important challenge for the industry. It is essential that reservoir models contain the small-scale property variations in well logs and cores while reproducing the large-scale structure and continuity in seismic data. However, a reservoir model derived from these static …