Simulation of power transfer and loss allocation based on kron reduction method in deregulated power system

Abstract

Recently in the privatization era, some countries adopt deregulated power system to increase efficiency and quality of the service through competition among the power industry participants. Previously, electricity supply industry (ESI) was regulated as monopoly power market where generation, transmission and distribution were solely own by electric state company. Deregulated power system, emerged new wholesale power market, with many independent power producer (IPP). Whilst the transmission remain monopoly due to its nature offers a nondiscriminatory open access to all user. The efficient transmission usage charges which can fit all parties is difficult to implement, it must recover the transmission cost by allocating the cost to transmission users properly. The transmission usage allocation is fundamental to determined an efficient transmission usage charges. The non linier nature of flow of electricity makes it difficult to calculate a precise transmission usage allocation. The transmission usage allocation can be determined by allocating the power transfer and the power losses from individual generator to the system loads. The simulation development of fair and transparent power allocation and loss allocation with many transactions taking place at any time has been an active topic of research. This thesis suggests a simulation with an improvement of the conventional method based on proportional sharing principle to allocate the active and reactive power contribution of individual generators to the system load and losses. The simulation algorithm calculates the real and reactive power allocation based on current operating point of the system, computed through AC load flow program. The proposed active and reactive power allocation methodology adopts current tracing, instead of power tracing. Based on solved load flow and the network parameters, the method converts power injections and line flows into real and imaginary current injections and flows. These currents are then represented independently as real and imaginary current networks. In addition, the equivalent model of transmission line is improved and embedded in the tracing procedure. This study will take into account of shunt admittances that give significant effect in reactive power flow in the system. Then the method utilizes kron reduction technique to minimize the size of the admittance matrix. The reduced admittance matrix named Y-kron matrix used to calculate the real and reactive power of individual generator contribution at each buses. In order to trace the active and reactive power simultaneously in a faster manner, an intelligent approach is proposed: the hybridization of optimization technique and machine learning approach. In this research, Genetic Algorithm and Support Vector Machine (GA-SVM) are integrated into the tracing methodology to determine the power contribution from generators to line losses and loads with good accuracy, which is up to 99% The advantages of the proposed methodologies are demonstrated on commonly used test systems and actual TNB equivalent systems, i.e. southern region 25-bus systems of the peninsular of Malaysia. The proposed methodologies provide better reliability and minimize the limitations of conventional real and reactive power allocation and loss allocation methods.