Optimal planning of power distribution networks with fault-tolerant configuration

Abstract

Power Distribution networks are essential infrastructures that should be designed by satisfying two conflicting requests: cost minimization and reliability. While traditional network planning aimed at radial configurations, which are more similar to the typical working configuration of a network but are not fault-tolerant, modern techniques seek for meshed configurations, since these architectures are more fault-tolerant. Due to the complexity of the problem and the large size of nowadays instances, most of the techniques used for planning are based on heuristic approaches. Thus, they are usually unable to guarantee optimality and not even able to provide an assessment of the distance from the optimal solution. In this work, we address the challenge of planning a fault tolerant network through an exact approach, by introducing innovative Mixed-Integer Linear Programming models designed for the planning of meshed distribution networks with loop-feeder or open-loop topology. Differently from other techniques, our approach simplifies the formulation by avoiding the need for fault scenarios, significantly reducing the computational burden of the optimization problem. The outcomes of our approach are the generation of optimal meshed network, which effectively balance cost and reliability of the electric distribution system. Comprehensive studies on realistic test instances show the advantages of the proposed formulations.