Detection of Tracker Failures for a 2-Axis Photovoltaic System
Abstract
Solar trackers have been widely used in photovoltaic plants due to the increase in their capacity factor. However, the operation of tracker systems still has challenges, mainly related to failures during operation. The mechanical failures on trackers significantly impact the PV generation and, consequently, have a non-negligible impact on the economic return. The data on trackers’ failures in operation situations is rare. The present paper will use a statistical method for daily failure detection from the relation between PV power output and plane-of-array (POA) irradiances. This methodology extracts features from data curves and classifies the days according to them. The accuracy obtained showed that new methods, such as those presented in this paper, may perform better in identifying tracking system failures.