Addressing Glare Problematics for Photovoltaic Projects in the Immediate Proximity of Roads and Railways through the Use of Accurate Digital Surface Models

The development of large-scale photovoltaic power plants generates disturbances in the local environment, among which solar glare that may turn critical when safety is at stake. The methodology for glare studies can hardly accept approximations for PV projects in the proximity of roads or railways as ground and near shading strongly affect the results. This paper presents the interest for working with accurate digital elevation modelling (DEM) and digital surface modelling (DSM) issued from LiDAR measurement campaigns. It first shows that LiDAR data allows to reduce the errors made by global elevation databases, e.g. Google’s Elevation API which presents a root mean square error of 1.9 m and 6.6 m for two neighbouring areas located in France, respectively in lowland and valley. The sensitivity of the DSM spatial resolution on the relevance of the glare assessment is also addressed through a 27 MWp use-case. The 10-m resolution turns out to be the most appropriate one as it allows to accurately account for near shading and photovoltaic table’s configuration; selecting a coarser resolution (20 or 30 m) mistakenly leads to non-existing glare occurrences that could have compromised the photovoltaic project itself during the administrative instruction process.