Streamline-based history matching of bottomhole pressure and three-phase production data using a multiscale approach

Reconciling geological models to the available dynamic information, commonly known as history matching, is an essential step for optimizing reservoir management and field development strategies, including improved recovery methods. There are several challenges in the current history matching workflow, particularly for high resolution geologic models with multimillion cells and complex geologic architecture. Streamline-based inverse modeling has shown great promise in this respect because of computational efficiency and analytic calculation of sensitivity of production response to reservoir properties. However, the current streamline-based approach has mostly been applied to history matching water-cut and tracer response in two-phase flow. More recently streamline-based sensitivity calculations have been extended to bottomhole pressure and gas oil ratio. This allows for generalization of streamline-based history matching to the three-phase. In this paper we present an integrated application of the streamline-based three-phase history matching by incorporating water-cut, gas-oil ratio and bottomhole pressure data while updating high resolution geologic models. The crux of our approach lies in the analytic computation of well response sensitivities based on streamline which allows for efficient inversion of production and pressure data. We also validate the accuracy and efficiency of the streamline-based bottomhole pressure sensitivities by comparison with an adjoint-based method using a finite difference simulator. In the history matching workflow, we incorporate the streamline-based method within a multiscale framework to account for the disparity in resolution of different types of history data. This integrated method proposed in this paper leads to capturing of the large- and fine-scale heterogeneity while matching the pressure and production responses efficiently. We demonstrate the power and utility of the streamline-based approach using synthetic and field applications with three phases. The synthetic example involves the SPE9 benchmark field case with water flooding and aquifer drive. The field example involves full-field history matching of the Norne Field in the North Sea using water-cut, gas-oil ratio and bottomhole pressure data. The proposed multiscale workflow clearly demonstrates the power and advantage of our approach in achieving geologically consistent history matching at the full-field level.