Multiobjective planning of power distribution networks with facility location for distributed generation

A multiobjective algorithm dedicated to simultaneously plan power distribution network and facility location, with focus on substation and distributed generation placement, is proposed in this work. This tool can perform the following operations: to plan the network topology, to assign the conductor capacities and types, to locate new generation units, and to analyse the robustness of the final network in order to help on decision making. In the design procedure, the minimization of both the monetary cost and the fault cost of the network for the “most likely” peak-load scenario are considered, for a future time horizon. The optimization of those different objective functions is performed in a multiobjective setting, leading to the determination of a Pareto-optimal solution set that describes the trade-offs involved in designer choices. The optimization algorithm is composed by a multiobjective genetic algorithm, deterministic local search operators, a procedure to locate new generation units, and a Monte Carlo simulator for evaluating system robustness. Uncertainties are considered in the load growth, energy price, and power supplied by the distributed generation units. The proposed tool allows scenario analyzes that go far beyond the simple cost per kilowatt or the availability rate figures.