Optimization of Surge Arrester Locations in Overhead Distribution Networks

Abstract

Lightning is considered one of the main causes of faults in overhead distribution networks. Direct strokes usually lead to flashovers due to the insulation levels that are used. Induced overvoltages caused by indirect lightning are usually lower and can be efficiently reduced by metaloxide surge arresters. Hence, its associated flashover rate can be reduced. In this paper, a heuristic method is proposed to optimize the number of surge arresters as well as their locations. The method presented is based on genetic algorithms and an economic approach is taken into account by means of evaluating the cost of insulation flashover. © 19862012 IEEE. Abstract-Lightning is considered one of the main causes of faults in overhead distribution networks. Direct strokes lead usually to flashovers due to the used insulation levels in those networks. Induced overvoltages caused by indirect lightning are usually lower and can be efficiently reduced by metal oxide surge arresters, hence its associated flashover rate can be reduced. In this paper an heuristic method is proposed to optimize the number of surge arrester as well as their location. The presented method is based on genetic algorithms (GA) and an economic approach is taken into account by means of evaluation the cost of insulation flashover. Author keywords

Optimized Surge Arrester Allocation Based on Genetic Algorithm and ATP Simulation in Electric Distribution Systems

“…The 10 kA magnitude was chosen for having a significant incidence in typical urban distribution grids and also as being a standard lightning discharge to test surge arresters in distribution grids with nominal voltage of 13.8 kV and 34.5 kV [11,12]. Induced lightning overvoltage in distribution feeders can cause significant impacts, and are widely used in the literature to evaluate the allocation of lightning arresters and other protection devices [4,6,[13][14][15]. Due to the lower voltage class of these feeders, the induced overvoltage caused by lightning discharge also tends to be smaller.…”

“…For that, repeatedly two chromosomes are selected and a random number is generated using a uniform distribution. If it is verified that this number satisfies the stipulated crossover rate, a random number α is generated which will serve to create weights by performing the weighted average of the parents chromosomes genes, generating new genes for each child chromosome, as in Equations (5) and (6). Otherwise, two new chromosomes with the same parents' genes are generated.…”

“…Many works can be found in the literature whose focus is the allocation of surge arresters using metaheuristics (e.g., [2][3][4][5][6][7][8]). Several intelligent techniques may be used for the allocation procedure, for example, the use of Fuzzy Logic technique [2], the Discrete Imperialist Competition Algorithm [7], and the Genetic Algorithm, which is one of the computational intelligence techniques that seems to be quite consolidated in the optimal solution of this type of problem [3][4][5][6]8].…”

“…It is common to adopt a statistical approach using the Monte Carlo's method for the exhaustive generation of lightning discharges of different magnitudes [2][3][4]7,8], to analyze the induced overvoltage along the electrical grid [2][3][4][5][6][7][8] in order to perform a risk analysis based on the probability of flashover occurrence [2,3,7]. In some cases, this assessment is made based on the critical flashover voltage (V CFO ) [5,6,8].…”

Methodology for Analysis of Electric Distribution Network Criticality Due to Direct Lightning Discharges

“…This procedure would make the electric network project unfeasible due to the high costs of purchasing and installing lightning arresters, insulators, and cables in view of the electric grid extension, number of poles, and complexity of these electric networks. However, given the economic infeasibility of this solution and seeking an optimal point of using these measures, several technical papers have addressed the problem associated with lightning discharges in distribution networks, either in the optimization process for determining the best location for the insertion of lightning arresters [7][8][9] in addition to disruptive tests in high voltage laboratories [10] or computational simulations that consider the statistical nature of the incidence of lightning discharges [11,12].…”