Optimal sizing of PV array for critical load with parallel redundant architecture using GA

The PV array sizing will affect the Capital Expenditure (CAPEX), Operational Expenditure (OPEX) and long term performance when feeding critical loads. In general the choice of PV array size is calculated using hourly meteorological station data (8760 steps/annum) and daily average input and output to the battery for remote or stand alone PV power plants. In this paper optimal sizing of PV array is carried out for an existing critical load consisting of constant DC load and AC load of intermittent periodic duty with starting nature, connected to single phase Domestic Supply (DS) with Diesel Electric Generator (DEG) as parallel redundant reserve source and an Uninterrupted Power Supply (UPS) in series with critical DC load. Approximate 5 minute solar radiation data (105120 steps/annum) is calculated using one hour meteorological station data and the same is used for studying the energy transactions between various devices and loads in the modified system. A typical critical load of urban roof top Base Transceiver Station (BTS) is considered as load. The optimal size of PV array is calculated for the above mentioned load for a desired Loss of Power Supply Probability (LPSP), minimum OPEX and CAPEX by considering long term performance degrading factors. The aforementioned problem is formulated and Genetic Algorithm (GA) is used to solve the optimization problem. The corresponding results are discussed.