Nechelik Field Reservoir Simulator — 54×22 implicit finite-difference grid, sparse matrix solver, Peaceman well model, Ei-function analytical validation, 3D pressure visualization
Production forecasting workflows — Rate-decline analysis, cumulative production curves, well interference diagnostics
Training modules — Interactive scripts for teaching transmissibility assembly, well index calculations, and timestep stability analysis
Multiphase Waterflood Simulator — 3D finite-difference black-oil model, coupled oil-water flow, Newton-Raphson with analytical Jacobian, 5-case scenario optimization, material-balance validation
PTA/RTA Automation Platform — Automatic flow regime identification, type-curve matching, parameter estimation, report generation
DCA Production Forecasting Engine — Duong, LGM, SEPD, T-model implementations, probabilistic forecasting, EUR estimation
Reservoir Data Analytics Workbench — Well log integration, feature engineering, statistical analysis, ML-ready data pipelines
Finite-difference method (FDM). Our core discretization approach for reservoir flow equations. Cell-centered structured grids in 2D and 3D with harmonic mean permeability averaging for transmissibility calculations.
Implicit formulation. Unconditionally stable timestepping via sparse matrix inversion (MATLAB) and Newton-Raphson iteration with Jacobian assembly (Python). Eliminates stability constraints on timestep size.
Newton-Raphson nonlinear solver. Analytical Jacobian construction for coupled pressure-saturation systems. Robust convergence under strong nonlinearities via line search and damping strategies.
Analytical validation. Cross-validation against Theis/Ei-function (radial flow), Buckley-Leverett (1D displacement), and material-balance checks ensures code correctness before field application.
| Component | MATLAB Stack | Python Stack |
|---|---|---|
| Grid generation | Native matrix operations | NumPy ndarrays |
| Transmissibility | Sparse matrix assembly | SciPy sparse |
| Linear solver | Backslash (mldivide) | SciPy spsolve / GMRES |
| Nonlinear solver | Custom implicit iteration | Newton-Raphson + Jacobian |
| PVT modeling | Tabulated / analytical | NumPy interpolation |
| Rel perm / cap pressure | Functional forms | Corey / Brooks-Corey |
| Well model | Peaceman | Peaceman + dynamic PI |
| Visualization | surf, contour, plot | Matplotlib + mplot3d |
| Data I/O | load, xlsread | pandas, LAS, CSV |
| Validation | Ei-function analytical | Material-balance error |
No licensing barriers. Commercial reservoir simulators (ECLIPSE, INTERSECT, CMG) cost tens of thousands of dollars per license. MATLAB academic licenses and Python eliminate this barrier entirely.
Institutional sustainability. African universities can teach, research, and deploy these tools indefinitely without vendor dependency or renewal cycles.
Skill portability. Python and MATLAB skills are globally recognized. GSERC trainees gain competencies that transfer directly to international operators, service companies, and graduate programs.
Customizability. Open-source code can be modified for specific fields, unconventional plays, or emerging challenges (CCUS, geothermal, hydrogen storage) without waiting for vendor updates.
Community & ecosystem. Python’s global developer community produces continuous improvements in numerical methods, visualization, and machine learning — all accessible to GSERC projects immediately.
