Numerical studies of imaging subsurface waterfloods using CSERT with time-lapse inversion

Abstract

During the operation of the waterflooding technique, it is necessary to identify the waterflooding areas to enhance oil displacement efficiency. A casing-surface electrical resistivity tomography (CSERT) system using a well casing as a long electrode is able to detect a wide lateral scope, but its vertical resolution and ability to identify deep anomalies in the reservoir are limited, particularly for reservoirs with high-conductivity anomalies in the shallow subsurface, which disturb the response from the water floods at depth. In this study, we first simplified this kind of reservoir into a dual layered anomaly model. Then the log–inject–log method with time-lapse inversion was proposed and evaluated regarding its ability to improve the imaging of the deep waterflooding areas and the shallow anomaly. The results were compared with the commonly used static measurement with static inversion. In the static inversion results, the shallow anomaly was imaged well but the deep anomaly was unobservable. The results of the proposed log–inject–log method with time-lapse inversion showed that it is able to identify the shallow and deep anomalies better under various conditions, thus validating its ability to improve the vertical resolution of the CSERT system.