The proposed inversion technique of remote sensing data for reservoir monitoring.

Abstract

A map of ground surface deformations due to the geomechanical instability of the reservoir has been estimated by using a differential InSAR stacking technique. The surface deformations have valuable information about the dynamic reservoir and are most likely to occur in the production of hydrocarbon even if the reservoir is kilometers deep. In order to invert surface deformations obtained from InSAR data, we propose a two-step inversion approach based on a tensional rectangular dislocation model. The first step, we use genetic algorithms to estimate the depth and the geometry of reservoir deformation. However, this model provides a single dislocation with uniform deformation. This uniform deformation can not provide an adequate representation of the spatially varying deformation in the reservoir. Therefore, in the second step, we have applied the least square inversion with the penalty function and smoothing factor in order to efficiently invert the spatial distribution of reservoir deformations and volume changes from the surface deformation data. Through a synthetic model, we have examined our inversion approach by estimating the root mean square error and the relative error, then applied to real data. We conclude that the InSAR technology is useful to provide ground surface deformations and accurately monitor reservoir deformation using inversion techniques.