Analytical Evaluation of Reservoir Simulation Model Uncertainty Quantification Using Ensemble-Based History Matching in the Bonny-Akpor Field Complex, Rivers State

Reservoir simulation model uncertainty quantification through ensemble-based history matching methods provides a probabilistic characterisation of reservoir behaviour uncertainty that is essential for informed field development decision-making, yet its application to the complex multi-reservoir Bonny-Akpor field system in Rivers State remains an analytical gap in the published Nigerian petroleum engineering literature. This study applies ensemble-based history matching to quantify uncertainty in simulation model predictions for the Bonny-Akpor Field Complex, Rivers State, South South Nigeria. An ensemble of 200 reservoir model realisations was generated through Latin Hypercube sampling of uncertain parameters including horizontal permeability multipliers, vertical-to-horizontal permeability ratios, aquifer constants, and relative permeability endpoints for three reservoir intervals. The Ensemble Smoother with Multiple Data Assimilation algorithm was applied to update ensemble parameters by assimilating thirty years of production history for sixty-eight producing wells. Posterior ensemble statistics were computed for reservoir parameters and production forecasts. Results demonstrate that production history integration reduces the porosity uncertainty range from plus or minus 3.8 percent to plus or minus 1.6 percent absolute, and reduces water cut prediction uncertainty at the ten-year forecast point from plus or minus 22 to plus or minus 9 percent. Infill drilling value of information analysis performed on the posterior ensemble demonstrates that an optimally located infill well generates an expected incremental NPV of 48 million USD, significantly higher than the mean prior ensemble prediction of 31 million USD. Keywords: ensemble history matching, uncertainty quantification, reservoir simulation, Bonny-Akpor Field, Rivers State.