Integrating Data of Different Types and Different Supports into Reservoir Models

In this paper, we focus on the joint integration of production and 4-D inverted seismic data into reservoir models. These data correspond to different types and different scales. Therefore, we developed two-scale simulation workflows making it possible to incorporate data at the right scale. This issue also emphasized the need for adapting traditional history-matching methodologies. For instance, the formulation of the objective function and the development of customized parameterization techniques turned out to be two key factors controlling the efficiency of the matching process. Two application examples are presented. The first one is a small-size synthetic field case. It aims to build a set of reservoir models respecting either production data only or both production and 4-D seismic-related data. It is shown that the incorporation of 4-D seismic-related data in addition to production data into reservoir models contributes to reduce the uncertainty in production forecasts. The second example is a field in the North Sea offshore Norway operated by Statoil. It stresses difficulties in conditioning reservoir models to both real production and 4-D inverted seismic data among the very large number of uncertain parameters to handle and the comparison of real noisy data with numerical responses