Numerical investigation of soiling of multi-row rooftop solar PV arrays

Abstract

The soiling behaviour of multiple solar PV arrays on multi-storey building rooftop was explored using Computational Fluid Dynamics (CFD). The CFD simulation study employed the SST k–ω turbulence model together with the discrete phase model. A grid independency analysis was done to determine the mesh size that is adequate for the simulation study. Three main parameters were investigated in this study on five rows of solar PV arrays and these were wind speed, dust particle size and tilt angle. Airflow characteristics on the solar PV arrays were evaluated and analysed on their influence on soiling. The study revealed that the PV arrays located at the front experience less soiling compared to those located at the rear although there was a small difference of 0.16% in the soiling rate. An average soiling rate of 15.77% was experienced on all the five PV arrays used in the simulation studies. The overall percentage soiling from the front row to the last row was found to be, respectively, 15.85%, 15.82%, 15.77%, 15.74% and 15.69%. A predictive model was developed and had a coefficient of determination, R2 of 96.82% which could accurately predict overall soiling on the PV arrays.