On the search for representative characteristics of PV systems: Data collection and analysis of PV system azimuth, tilt, capacity, yield and shading

• Published in: Solar energy 2018-10, Vol.173, p.1087-1106
• Main Authors: Killinger, Sven, Lingfors, David, Saint-Drenan, Yves-Marie, Moraitis, Panagiotis, van Sark, Wilfried, Taylor, Jamie, Engerer, Nicholas A., Bright, Jamie M.
• Format: Article
• Language: English
• Subjects: Azimuth; Clustering; Data analysis; Data collection; Data processing; Distribution functions; Electric power; Elevation angle; Empirical equations; Engineering Science; Engineering Sciences; Equator; Geographical locations; Irradiance; Metadata; Parameters; Photovoltaic cells; PV system characteristics; Roofs; Shading; Solar energy; Teknisk fysik; Yield
• ISSN: 0038-092X; 1471-1257; 1471-1257
• DOI: 10.1016/j.solener.2018.08.051

•We aim to derive representative PV characteristics for regional power modelling.•Tilt, azimuth, capacity and specific annual yield are analysed from 2.8 m systems.•Strong interdependencies motivated a clustering by countries and systems sizes.•For eased utilisation, each parameter was approximated by a distribution function.•Recommendations for a consideration of shading influences on PV power are derived. Knowledge of PV system characteristics is needed in different regional PV modelling approaches. It is the aim of this paper to provide that knowledge by a twofold method that focuses on (1) metadata (tilt and azimuth of modules, installed capacity and specific annual yield) as well as (2) the impact of shading. Metadata from 2,802,797 PV systems located in Europe, USA, Japan and Australia, representing a total capacity of 59 GWp (14.8% of installed capacity worldwide), is analysed. Visually striking interdependencies of the installed capacity and the geographic location to the other parameters tilt, azimuth and specific annual yield motivated a clustering on a country level and between systems sizes. For an eased future utilisation of the analysed metadata, each parameter in a cluster was approximated by a distribution function. Results show strong characteristics unique to each cluster, however, there are some commonalities across all clusters. Mean tilt values were reported in a range between 16.1° (Australia) and 35.6° (Belgium), average specific annual yield values occur between 786 kWh/kWp (Denmark) and 1426 kWh/kWp (USA South). The region with smallest median capacity was the UK (2.94 kWp) and the largest was Germany (8.96 kWp). Almost all countries had a mean azimuth angle facing the equator. PV system shading was considered by deriving viewsheds for ≈48,000 buildings in Uppsala, Sweden (all ranges of solar angles were explored). From these viewsheds, two empirical equations were derived related to irradiance losses on roofs due to shading. The first expresses the loss of beam irradiance as a function of the solar elevation angle. The second determines the view factor as a function of the roof tilt including the impact from shading and can be used to estimate the losses of diffuse and reflected irradiance.