Loss Minimization in High Voltage Distribution System by Capacitor Placement Using Genetic Algorithm

Abstract

Electricity distribution is the final stage in the delivery (before retail) of electricity to end users. A distribution system's network carries electricity from the transmission system and delivers it to consumers. Distribution networks are typically of two types, radial or interconnected. A radial network leaves the station and passes through the network area with no normal connection to any other supply. This is typical of long rural lines with isolated load areas. An interconnected network is generally found in more urban areas and will have multiple connections to other points of supply. These points of connection are normally open but allow various configurations by the operating utility by closing and opening switches. The benefit of the interconnected model is that in the event of a fault or required maintenance a small area of network can be isolated and the remainder kept on supply. In distribution systems, the voltages at buses reduces when moved away from the substation, also the losses are high. The reason for decrease in voltage and high losses is the insufficient amount of reactive power, which can be provided by the shunt capacitors. The work reported in this thesis is carried out with the objective of identifying the Optimal locations and sizes of shunt capacitors to be placed in radial distribution system to have overall economy considering the saving due to energy loss minimization and cost of capacitors. For the purpose two stage methodology is used. In first stage, the load flow of precompensated distribution system is carried out. On the basis of load flow solutions, loss sensitivity factors (LSF) indicating the potential locations for compensation are computed. From LSF, the candidate number of buses is identified. In the second stage, genetic algorithm is used to identify the sizes of the capacitor for minimizing the energy loss cost and capacitor cost. A coding scheme is implemented where the identification of location and size of capacitor is represented by one dimensional array. The developed algorithm is tested for distribution systems at 11 KV high voltage distribution system (HVDS) while