Uncertainty Quantification and History Matching for Naturally Fractured Carbonate Reservoirs

Summary Carbonate reservoirs host a major part of the world’s hydrocarbon reserves and over the past decade(s) have shown an increase in geothermal potential all over the world. However, naturally fractured carbonate reservoirs (NFR) contain a large uncertainty in their flow response and mechanical behavior due to the poor ability to predict the spatial distribution of discontinuity networks at reservoir-scale. In this work, we present a potential workflow for performing uncertainty quantification and data assimilation in fractured carbonate reservoirs. This workflow consists of a pre-processing step in which the original fracture network is cleaned and can be represented at the desired discretization accuracy. This method can then be used to transform a high-fidelity ensemble of models to some coarser representation. This coarser representation can be subsequently used to determine ensemble representatives. Finally, a history matching routine can be performed on each ensemble representative which characterizes the main flow patterns present in the NFR.