A novel approach to Solar PV cleaning frequency optimization for soiling mitigation

Soiling is a major issue that can be a drawback to the wider deployment of solar photovoltaic systems. In this study, the influence of soiling on energy loss in Muzarabani, Zimbabwe was investigated. The daily and monthly variations in energy generated and soiling were also studied. An empirical soiling loss model was developed based on the experimental studies. It was revealed that soiling is not uniform with each passing day but rather depends on the daily differences in weather conditions. Soiling was observed to be high during the period from July to November and was less in May and June. Particle Swarm Optimisation was employed to minimise the number of days between cleaning events. Cleaning was found to be necessary every 15 days to minimise the losses due to both frequent cleaning as well as losses caused by not cleaning the panels.