MISOCP-BASED OPTIMAL CAPACITOR ALLOCATION IN POWER DISTRIBUTION SYSTEMS CONSIDERING ZIP LOAD MODEL

Abstract

The problem of power loss reduction and voltage profile improvement is of great importance in the planning and operation of distribution systems when the
electricity demand shows a substantial increase. This paper presents a Mixed-Integer Second-Order Cone Programming model (MISOCP) to optimally determine the
location and size of shunt capacitor banks as a way of improving the operational efficiency of distribution networks. The optimization problem aims at minimizing the
total cost associated with capacitors, power loss and energy loss while complying with power flow equations, voltage constraints, branch power flow constraints,
power factor limits and constraints pertaining to the capacitor banks installed. At the same time, this optimization problem considers a model of the voltagedependent load (ZIP load model). The MISOCP model is transformed from a Mixed-Integer Nonlinear Programming model (MINLP) by exploiting a second-order cone
model of power flow equations and the equivalent ZP load model. The globally optimal solution of the proposed MISOCP model is achieved by using a commercial
solver such as CPLEX with the GAMS programming language. The proposed MISOCP formulation is evaluated using an IEEE 15-node distribution system. The
calculation results show that optimization of capacitor placement contributes significantly to reducing the total cost and improving the voltage quality in the
distribution networks.