Optimal placement of uPMUs to improve the reliability of distribution systems through genetic algorithm and variable neighborhood search

Abstract

Due to the dynamic nature of modern distribution systems, the deployment of micro-phasor measurement units (uPMU) is becoming increasingly common among utilities to improve system monitoring and reliability. However, given their high investment costs, deploying a large number of these devices becomes unfeasible. Hence, unlike other approaches found in the literature that focus on observability criteria, this work presents an algorithm for optimal placement of uPMUs aimed at improving distribution system reliability. The algorithm defines the optimal number and location of the uPMUs through an objective function based on the resolution of a fault location technique that works in conjunction with pseudo-measurements to successfully locate a contingency. The meta-heuristics Genetic Algorithm and Reduced Variable Neighborhood Search are employed to address this problem. The proposed method has been validated on a three-phase 39-bus distribution system and a real distribution feeder with 962 buses from an Ecuadorian electric distribution utility. The results obtained confirm the effectiveness of the method, as with the deployment of only two uPMUs, the energy not supplied decreases by 13.84 % and 24.96 % for the 39-bus and 962-bus systems, respectively. Moreover, in the 962-bus system, the System Average Interruption Duration Index (SAIDI) is reduced by 20.36 %.