Improvement of Reliability Indices and Costs in Distribution Systems Considering Multiple Scenarios Through Switch Reallocation

Abstract

This paper presents a method for improving indices and costs associated with reliability in radial distribution systems through switch placement using genetic algorithms. On this paper, the switch allocation problem will be solved with a method that allows novelties that can improve the quality of the solutions obtained. First, the possibility of switch allocation on both sides of each section of a feeder is proposed. Second, a method to quantify the non-supplied energy in terms of dollars per kWh of interruption is expanded, allowing a versatile way to simulate cases in which individual customers have different economic impacts when in situation of interrupted power delivery. The proposed algorithm is also versatile when additional restrictions are applied, such as fixing the number of switches or imposing SAIDI limits. Optimal switch locations are defined to improve restoration and reduce costs associated with reliability. Simulations in test and real feeders are performed, and the presented results show significant improvements.