A multi-scale approach to integrate dynamic data in reservoir models

Reliable predictions of reservoir flow response require a realistic geological model of heterogeneity and an understanding of its relationship to flow performance of the reservoir. This paper presents a novel approach for integrating dynamic data in reservoir models that utilizes the probability perturbation approach for the simultaneous calibration of geological models at field scale and multiphase flow functions associated with porelevel spatial representations of the porous media. In this probabilistic approach, a stochastic simulator is used to model the spatial distribution of a discrete number of rock types identified by rock/connectivity indexes (CIs). Each CI corresponds to a particular pore network structure with a characteristic connectivity. Primary drainage and imbibition displacements are modeled on the 3D pore networks to generate multiphase flow functions, including effective permeability and porosity of the rock, the relative permeabilities and capillary pressure, linked to the CIs. During the assisted history matching, the stochastic simulator perturbs the spatial distribution of the CIs to match the simulated pressures and flow rates to historic data. Perturbation of the CIs in turn results in the update of all the flow functions. The results from the integrated history matching procedure are presented for a synthetic field example first. The convergence rate of the proposed method is comparable to other current techniques with the distinction of enabling consistent updates to all the flow functions while at the same time honoring the geological/ sedimentary model for the distribution of petrophysical properties. Consequently, the reservoir model and its predictions are consistent with realistic geological processes. The paper concludes with results for a realistic field example.