Time Varying Load Models Application for Optimal Allocation and Sizing of PV Systems in Distribution Networks

Abstract

Distributed Generation (DG) is widely used in distribution systems to enhance the dynamic characteristics of the network such as improving the voltage profile, minimizing active and reactive power losses. A variable voltage load model is assumed in many distribution systems. Therefore, the results are more reliable and accurate. This thesis presents an optimal method for selection of size and site of DG unit based photovoltaic system to minimize a multi objective function that depends on active and reactive power losses and voltage deviation. This is possible by giving appropriate weight for each one of the above variables towards the design or operational limits of the PV system taking into consideration the variable load model and mixed load between residential, commercial and industrial. Optimization is made by using a new analytical method and also by using a new optimization technique which is called Big Bang-Big Crunch optimization technique which provides an effective approach for computing the sensitivity of power loss and voltage profile for active and reactive power insertion in the radial distribution system. A 33-bus test distribution system and 69-bus test distribution system have been applied to demonstrate the effect of DG on the distribution network. The Forward/Backward Sweep Load Flow technique is used in this thesis.