Multiobjective optimization-based design and dispatch of islanded, hybrid microgrids for remote, off-grid communities in sub-Saharan Africa

A new AI-powered optimization framework, developed by researcher Vineet Jagadeesan Nair, provides a blueprint for designing affordable and reliable hybrid microgrids for remote, off-grid communities. The model tackles a complex, non-convex problem by using particle swarm optimization to determine the optimal sizing of solar PV, wind generation, lithium-ion battery storage, and diesel backup generators. It evaluates system performance over a full year at hourly resolution, simultaneously minimizing four key objectives: the lifecycle levelized cost of energy, carbon emissions, lost load (blackouts), and dumped (wasted) renewable energy. This multi-period, multiobjective approach identifies Pareto-optimal solutions that reveal critical trade-offs, demonstrating that a cost-only design leads to poorer environmental and reliability outcomes.

The study's results consistently show that a solar-wind-battery-diesel hybrid configuration outperforms alternatives. A key finding is that cost considerations primarily dictate the allocation between solar and wind, while the required size of renewables and storage is heavily influenced by the rating of the standby diesel generator due to reliability constraints. Sensitivity analyses highlight how fluctuating fuel prices and falling battery costs shift the optimal design. By accurately sizing components, the framework can reduce the capital cost oversizing typically used as a safety margin in off-grid systems by 20-30%, making clean electricity more affordable. The accompanying dispatch model creates day-ahead schedules for operating the microgrid, which are generally robust, though short-term weather uncertainty increases reliance on the fossil fuel backup, underscoring the critical role of smart battery and generator dispatch.