This technical report documents a parametric sensitivity study of compositional gas injection in reservoir systems using OPM Flow, an open-source finite-volume simulator. Two standard benchmark cases were executed: SPE1 (black-oil) and SPE5 (compositional with volatile oil). Injection rate parametrization was performed on the SPE5 model by modifying well control specifications (WCONINJE keyword) across three operational scenarios: baseline (12,000 rb/day), elevated (14,000 rb/day), and critical (15,000 rb/day). Results indicate that a 16.7% increase in injection rate produces a 3.5% increase in linearizations and 4.6% increase in Newton iterations, demonstrating sub-linear computational scaling within the stable operational envelope. A numerical stability boundary was identified near 15,000 rb/day, beyond which the nonlinear solver fails to converge. These findings establish quantitative relationships between injection strategy, solver robustness, and simulation tractability—metrics essential for designing CO₂-EOR conversion workflows in depleting conventional reservoirs. The study demonstrates the utility of open-source compositional simulators for pre-field feasibility assessment of injection-based storage strategies.

Software: OPM Flow 2025.10Grid: SPE1 (10×10×3), SPE5 (7×7×3)EOS: Peng-RobinsonInjection rates tested: 12,000, 14,000, 15,000 rb/day GitHub repository: https://github.com/eonwe141/SPE5-Injection-Feasibility-Study.git

OPM Flow version 2025.10; SPE1 and SPE5 benchmark cases; parametric variation of WCONINJE injection rate; Ubuntu 22.04 LTS (WSL2); automatic timestep control; BiCGSTAB linear solver with ILU(0) preconditioning.