Surrogate modelling of solar radiation potential for the design of PV module layout on entire façade of tall buildings

•The application of surrogate modelling for the simulation of solar radiation potential is investigated.•A GA-based Random Forest Machine learning technique is used.•Different scenarios for the representation of solar radiation potential are considered.•The random variation scenario is found to be the most accurate.•The surrogate modelling approach is shown to have an accuracy of up to 94%.