Simplified modeling techniques for predicting outputs of grid-connected photovoltaic power systems

Abstract

In any grid-connected photovoltaic (GCPV) system, yield is amongst the most important aspect that the client looks at, as it has direct impact when it involves a Feed-In-Tariff (FiT) scheme. A good technique is needed for this, but, at the same time, should be simple yet precise. Depending on the cost, time and complexity, a balance is needed between simplicity and accuracy. In this work, a simplified modeling technique is proposed to predict the yield and check the optimum power sizing of a GCPV system. These models not only predict reasonably accurate yield, but also detects the optimal matching of PV to inverter power capacities. The user can then examine the GCPV system in a simple and relatively fast, yet reasonably accurate manner. The models used are steady-state equations that take into consideration the major factors affecting yield, such as: array power; irradiation; mismatch; temperature; dirt; cable efficiency; and inverter efficiency. This technique was then tested on two separate GCPV systems installed in Malaysia. One month datasets were used to investigate and validate the models. The results show good agreement between the predicted against the actual values measured. These models not only predict reasonably well results but also enables the user to detect the matching of optimum sizing of the GCPV system. The owner can then examine the system in a simple and relatively fast manner.