Models for prediction of soiling-caused photovoltaic power output degradation based on environmental variables in Doha, Qatar

In a previous study, using field measurement data from the Qatar Foundation Solar Test Facility, the daily change in Cleanness Index (CI), a measure of PV performance ratio, corrected for temperature and normalized to a clean PV module, was correlated to environmental variables including airborne particulate matter concentration (PM10), wind speed (WS), and relative humidity (RH). A linear empirical equation between daily CI change and the daily average PM10, WS, RH was developed using Microsoft Excel^®. However, the model was not extensively evaluated due to the small data set available then. In this study, a larger data set was used to fit the linear model for daily CI change and daily average values of PM10, WS, and RH. In addition, a semi-physical model was developed to take into account the non-linear mechanics of turbulent deposition, resuspension of deposited dust, and the effect of relative humidity on resuspension. The regression and solver functions of Microsoft Excel^® was employed to fit the data. The R-squared values of the linear model and the semiphysical model are 0.0949 and 0.1774, respectively. The semiphysical model predicts the daily ΔCI slight more accurately than the linear model. However, for prediction of cumulative ΔCI over longer periods of time, the two models perform roughly the same. Overall, both models are able to predict the two-month ΔCI with an uncertainty of less than 16%. The results from this study suggest that it is possible to use mathematical models to calculate PV power output degradation in the Doha, Qatar area. This may be a significant step towards development of models that can be used for economic analysis of PV solar projects and plant maintenance.