Shading can be the most detrimental factor on performance for a domestic system. The impact of shading on performance varies depending on the electrical series and parallel arrangement of cells within a module and modules within an installed array. Whilst many approaches to shading analysis have been proposed, computational efficiency is not reported despite being of high importance when incorporating shading algorithms into an overall energy yield model. The lack of consideration of the non-linear impacts of shading on smaller systems for example means that the shading loss is significantly underestimated, especially from supposedly small obstacles such as antennas or chimneys. As an example, the system shown in Fig. 1 illustrates the case where the installer may have attested a shade loss factor close to unity under UK microgeneration guidelines (Microgeneration Certification Scheme, 2013), i.e. negligible, but the performance of the system is severely compromised due to the non-linear cell mismatch effects. An effective shading sub-model therefore needs to give feedback to inform decisions of array layout in the proximity of obstructions but must not rely on high power computing.
