Automatic History Matching for One of the Iranian Fractured Oil Reservoirs

Production history matching is necessary for achieving validate prediction regarding reservoir future performance under different scenarios. Reservoir comprehensive model consider the best estimation for reservoir parameters but when simulator is run, reservoir model cant reproduce all observation data. Considering this point confirm necessity of production history matching. Manual and automatic history matching are two current methods in production history matching process. If fracture study has not been done for a fractured reservoir, manual history matching cant carefully predict reservoir future performance because unique quantity is allocated for fracture properties in all grids and as a result, reservoir fractures heterogeneity is ignored. Therefore the model in which history matching has been done using conventional method cant be representative for heterogeneous nature of reservoir fractures. In this research, automatic history matching for dynamic model of fractured oil reservoir has been done using Simopt Software. Simopt Software using optimization algorithm not only decrease required time for achieving a satisfactory match in contrast to conventional method, but also increase accuracy of parameters allocated to model. As well if property map is not present for a specific parameter, Simopt can achieve an accurate match by allocating different quantities for different grids in three different modes: evaluation, gradient and regression. Therefore we can access to a more validate model for prediction of reservoir future performance under different scenarios.