Optimisation of Power Allocation in Nigerian Electricity Distribution Using Integer Programming

Electricity load shedding in Nigeria's distribution networks reflects a structural mismatch between generation capacity and consumer demand, compounded by technical losses and billing inefficiencies. This study applies integer linear programming to optimise daily load allocation decisions across consumer categories within the Eko Electricity Distribution Company's network covering Lagos Island, Lagos Mainland, and Lekki districts in Lagos State, South West Nigeria. The model allocates available generation capacity across residential, commercial, industrial, and critical infrastructure feeder categories to maximise social welfare, proxied by weighted consumer priority scores, subject to network capacity, voltage stability, and equity constraints. Demand data, feeder capacity ratings, and historical allocation records are obtained from the distribution company. The binary integer variables represent feeder activation decisions within defined scheduling windows. The model is solved using branch-and-bound implemented in the PuLP optimisation library in Python. Optimal allocation schedules achieve a 31 percent improvement in weighted consumer satisfaction scores compared to current manual scheduling practice, with critical infrastructure feeders receiving uninterrupted supply in all scenarios. The study identifies 14 feeder segments where network upgrades would yield the highest incremental capacity gains. Findings provide a replicable operations research framework for distribution companies operating under constrained generation environments. Keywords: integer programming, load allocation, electricity distribution, Lagos, power optimisation.