Seasonal photovoltaic soiling: analysis of size and composition of deposited particulate matter

Abstract

Photovoltaic soiling - energy loss due to dust and particulate matter (PM) deposition - remains a top concern for global renewable energy generation. To minimize energy losses, the seasonality of soiling needs to be understood. This comprehensive study took place in Gandhinagar India and combined soiling monitoring (using a Campbell Scientific soiling station and a new, low-cost sensor called the Low-cost Alternative to Monitoring Photovoltaic Soiling, or LAMPS station) for all of 2019 with 3-week sampling cycles to analyze size distribution and composition. The LAMPS station has been found to monitor soiling to an accuracy of within 1.5% soiling. Soiling rates were found to be 0.45±0.10% day−1 during dry periods and negligible during the monsoon. The monsoon rains dramatically shifted size distribution with a 90% reduction of deposited > PM10 mass, but more than twice the mass of deposited PM2.5 particles. Rain and humidity also lead to non-uniformity in mass loading over the scale of millimeters, deemed milli-scale non-uniformity (MSNU). Composition was found to be >95% crustal dust and there was no difference in composition seasonally. There was also no compositional difference between sizes of particles larger than PM2.5. Moisture led to a variety of cementation products deriving from reactions with water soluble gases as well as precipitation reactions all taking place within droplets on the surface. The most prevalent cementation effect was caking masses high in carbon and salts. Fungal growth was seen to be growing, and spore producing, after just 3-weeks.