Reservoir characterization by cross-hole seismic imaging. Final report, September 15, 1989--June 30, 1994

Abstract

Better characterization of reservoirs requires better images of those reservoirs. This report documents the research undertaken at the Massachusetts Institute of Technology`s Earth Resources Laboratory (ERL) to improve seismic tomographic images. In addition, the new imaging method was applied to a data set collected in a producing oil field. The method developed is nonlinear travel time tomography. This technique uses the travel time of the first arriving energy at a receiver and distributes that time back along realistic ray paths. This is an important distinction between this method and previous methods that used either straight ray paths from source to receiver or fixed ray paths (ray paths fixed by an a priori model). The nonlinearity arises during each iteration in the matching of observed travel times with those determined from a model. In this technique the model is updated during each iteration (the velocity structure is changed) and new ray paths are computed in that update model. Thus the resulting image is based on physically realistic ray paths. Tomography resolution is not merely a simple function of the wavelength of the seismic energy used but also involves a measure of how well a given region has been sampled by ray paths. Moreover, the ray paths must represent a wide variation in inclination as they pass through a given spatial cell. This imaging technique was applied to a compressional wave data set collected at ERL`s Michigan Test Site located in the Northern Reef Trend of MI. It consists of two deep boreholes that straddle a producing reef. Two hundred source positions and two hundred receiver positions were used to obtain 40,000 ray paths. Although ERL`s boreholes are 2,000 ft apart, kilohertz data was obtained. The resulting image of the reservoir showed a low velocity zone inside the reef and a thin layer of low velocity that intersected one of the boreholes. The presence of this thin layer was confirmed by logs and borehole engineering.