Economic Analysis and NPV Optimization of Control Parameters for Gas Injection and Polymer Flooding, with Low Salinity Post Water Flush for Heavy Oil Recovery in Niger-Delta Reservoir

Abstract

Heavy oil reservoirs in the Niger Delta present significant development challenges due to unfavorable mobility ratios, reservoir heterogeneity, and rapid water breakthrough. This study presents a comprehensive technical and economic evaluation of a hybrid enhanced oil recovery strategy integrating gas injection, polymer flooding, and low salinity post-water flush in a Niger Delta heavy oil reservoir. A fully compositional CMG-based simulation model was constructed incorporating 131×81×54 grid blocks (573,084 total grids), average porosity of 0.34, permeability ranging from 1500 to 2400 mD, initial water saturation of 0.27, bubble point pressure of 1441 psi, and oil viscosity of 8.5 cP. The reservoir was produced under primary depletion from 2006 to 2012, followed by gas injection (CO2 and N2 at 2.0–2.5 MMSCFD per well) from 2012 to 2014, polymer flooding (2000 bbl/day) from 2014 to 2017, and low salinity post-water flush (15000 ppm, 1000 bbl/day) from 2017 to 2020. Optimization was conducted using CMOST AI with Particle Swarm Optimization, where operational parameters including gas injection rates, gas injection pressures (1500–2500 psi), polymer injection rates, and water injection rates were tuned to maximize Field Net Present Value as the primary objective function. Sensitivity analysis using a Reduced Linear surrogate model (R2 = 0.899) and Sobol variance decomposition quantified parameter influence and uncertainty contributions. The optimized hybrid EOR case achieved a cumulative oil recovery factor of approximately 11.1% by 2020, compared to 9.2% in the base case, representing an absolute incremental recovery of 1.9 percentage points. While seemingly modest, this increment is economically substantial at field scale, corresponding to a Field NPV spread of approximately 4.52 × 108 USD across the design space. The optimized case maintained higher average reservoir pressure (1460 psi versus 1385 psi in the base case) and accelerated oil production, generating earlier cash flow and improved discounted economic returns. Water cut in the optimized case reached 52–53% by 2020, compared to 36–37% in the base case, indicating improved volumetric sweep efficiency and accelerated depletion of movable oil rather than premature breakthrough. The surrogate model equation (Field NPV = 2.57425×109 – 39711.7×gas_inj_press_B + 197.848×gas_inj_rate_A + 470.511×gas_inj_rate_B + 188144×polymer_inj_rate_A + 133462×polymer_inj_rate_B + 178316×water_inj_rate_A + 109386×water_inj_rate_B) revealed that gas injection rate at Injector B exhibited the highest positive economic coefficient, indicating dominant sensitivity. Sobol variance decomposition confirmed that gas injection rate at Injector B alone accounted for approximately 61% of total Field NPV variance, making it overwhelmingly the most influential control parameter. Polymer injection rates contributed secondary influence (approximately 10% combined), confirming that mobility control enhances but does not independently dominate economic outcome. Low salinity water injection contributed moderate economic leverage (approximately 9%), while injection pressures exhibited negligible sensitivity (≤0.009% contribution each) within the operational bounds of 1500–2500 psi, indicating that the system behaves in a rate-controlled rather than pressure-controlled regime. Interaction effects among parameters were minimal (≤0.01%), confirming near-linear additive economic behavior of the hybrid system. This study establishes a structured optimization hierarchy for hybrid EOR implementation in heterogeneous Niger Delta heavy oil reservoirs. The findings demonstrate that economic performance is primarily governed by gas volumetric injection control (rate-dominated), polymer flooding enhances sweep efficiency but remains secondary to pressure support, low salinity post-flush provides incremental improvement but limited standalone economic leverage, and injection pressure optimization plays a minor role compared to rate control. The integration of full-physics simulation with AI-assisted optimization successfully identified economically optimal control parameters, demonstrating that significant economic value can be unlocked through systematic control parameter tuning in mature heavy oil fields. Future work should extend the framework to include uncertainty quantification in oil price scenarios, chemical degradation effects, and pilot-scale validation to further strengthen deployment confidence.