Scalable methodology for the photovoltaic solar energy potential assessment based on available roof surface area: Further improvements by ortho-image analysis and application to Turin (Italy)

The ongoing rush of the UE member states to the 2020 overall targets on the national renewable energy share (see Directive 2009/28/EC), is propelling the large exploitation of the solar resource for the electricity production. However, the incentives to the large employment of PV solar modules and the relative perspective profits, are often cause of massive ground-mounted installations. These kind of installations are obviously the preferred solution by the investors for their high economic yields, but their social impact should be also considered. Over the Piedmont Region for instance, the large proliferation of PV farms is jeopardising wide agricultural terrains and turistic areas, therefore the policy of the actual administration is to encourage the use of integrated systems in place of massive installations. For these reasons, an effort to demonstrate that the distributed residential generation can play a primary role in the market is mandatory. In our previous work “Scalable methodology for the photovoltaic solar energy potential assessment based on available roof surface area: application to Piedmont Region (Italy)”, we already proposed a basic methodology for the evaluation of the roof-top PV system potential. However, despite the total roof surface has been computed on a given cartographical dataset, the real roof surface available for PV installations has been evaluated through the assumption of representative roofing typologies and empirical coefficients found via visual inspection of satellite images. In order to overcome this arbitrariness and refine our methodology, in the present paper we present a brand new algorithm to compute the available roof surface, based on the systematical analysis and processing of aerial georeferenced images (ortho-images). The algorithm, fully developed in MATLAB®, accounts for shadow, roof surface available (bright and not), roof features (i.e. chimneys or walls) and azimuthal angle of the eventual installation. Here we apply the algorithm to the whole city of Turin, and process more than 60,000 buildings. The results achieved are finally compared with our previous work and the updated PV potential assessment is consequently discussed.

► Roof-top integrated photovoltaic systems: potential estimation.
► GIS data and ortho-imagery processing and analysis.
► Roof partitioning and analysis: suitability for PV installations and theoretical potential.
► Distributed PV potential over large scale territories.