An approach to predict output of PV panels using weather corrected global irradiance

Abstract

Predicting the power output of a Photovoltaic (PV) installation, whether small roof-top or utility scale, is important from resource planning perspective. For a home user, knowledge of solar power availability in the next 12 hours helps in optimal load scheduling. For a utility operator with a mix of power sources, this predicted PV output (along with a knowledge of consumer load) helps in controlling optimal mix of generation and enabling a reliable sustained power with better resource utilization. PV output estimation models for clear sky are mature and well researched, however, the effect of weather-related factors such as cloud cover, smog, fog etc. are difficult to model and include in the PV output calculations. The aim of this innovative study is to forecast the power a PV system may generate in the next 24 hours using the appropriate theoretical models. The estimated data is compared to the actual PV system readings installed at the test site. The paper presents the results in two stages, firstly theoretical predicted ground level irradiance is calculated without any attenuation factor and in second stage the irradiance is adjusted using the cloud attenuation model (C.IGH) proposed in this paper. The theoretical dataset acquired shows a reasonable accuracy when compared to the subjected PV system installed at the test site.