Estimation of snow loss for photovoltaic plants in Norway

Abstract

Large PV plants are increasingly common in locations with colder climates where snow can lead to significant PV power loss. For these locations, estimates of snow loss is necessary for accurate PV yield modeling. Robust estimation of snow loss is, however, challenging. Snow-induced loss is expected to vary with climate, weather, and PV plant design. In this work, we estimate snow loss from historical data for a set of PV plants in Norway. To extend the snow loss dataset, 12 years of weather data and a modified adaption of the Marion snow loss model are used to simulate snow loss for the analyzed PV plants over time. For the historical data, we observe variations in annual losses for the same system of more than 10 percentage points. For some of the systems, we find losses in a range from 0 to 100 % for the same month. As expected, systems with colder climates have higher loss than systems in warmer climates, and systems with higher tilt has lower loss than systems with lower tilt. With snow loss modeling we get improved understanding of typical and extreme values, and the potential inter-annual variation in monthly and annual snow loss.