History Matching and Quantitative Use of 4D Seismic Data for an Improved Reservoir Characterization

Abstract

Abstract In reservoir characterization, the value of each piece of data does not lie in its isolated use, but rather in the value it adds to the analysis when integrated with other data. Earth scientists from different disciplines have made many efforts to better predict the spatial distribution of petrophysical properties. However, the interpretation of isolated piece of knowledge, without a real integration of disciplines, cannot lead to a quantitative answer in terms of reservoir characterization. In this paper we present a joint inversion scheme for estimating petrophysical properties by integrating both production and 4D seismic related data in geological modeling. Simultaneous history matching of production and 4D seismic related data leads to a better prediction of petrophysical reservoir properties and of production forecast. Introduction This paper describes an integrated methodology, based on a non-linear optimization loop, for a quantitative use of 4D seismic data for improved geological modeling and reservoir characterization. The proposed methodology is made in two sequential phases (Figure 1):Phase 1 - Pre-stack seismic data inversion: For each seismic survey, the available angle sub-stacks are jointly inverted using a 3D stratigraphic inversion methodology. This methodology allows the direct estimation of an optimal elastic model in terms of compressional and shear impedances (Ip and Is). These impedances carry information related to static (porosity and permeability) and dynamic (pressure and saturation) properties of the reservoir.Phase 2 - Production and pre-stack 4D seismic data history matching: The optimal impedance data derived from pre-stack seismic are used, in addition to production data, to constrain and update the geological model. Synthetic production data and 4D compressional/shear impedance data are computed from a multi-phase fluid flow simulator coupled with rock physics model using Gassmann's equations. The reliability of the geological model is improved through the minimization of a weighted, least-squares objective function which includes both production and 4D seismic data. We applied successfully the proposed methodology for a multi-scale reservoir characterization process using a three-dimensional synthetic case including 12 years production data and 3 seismic surveys.