URB-Solar: An Open-Source Tool for Solar Power Prediction in Urban Areas

Abstract

The accurate prediction of the direct and diffuse solar radiation is of foremost importance for deployment of photovoltaic (PV) systems. A number of solar radiation forecasting techniques have been developed for longer and shorter forecasting times. Numerical weather prediction (NWP) models provide the best results for the longer forecasting times (4–6 h), required by utility companies. However, NWP methods are usually developed for clear-sky and open areas. These methods cannot be directly applied to urban areas with shading, trees, multisurface reflection, and other sources of solar radiation losses. To overcome these issues, improvement to the existing prediction tools are required. In this study, we develop an automated radiation forecasting tool for urban areas. This tool combines a NWP model (Weather Research and Forecasting (WRF) model) and a solar calculator (developed in the numerical toolbox OpenFOAM) to compute shading, reflection, and other losses in the urban canopy. An algorithm for extraction of building outlines and heights (if they are publicly available) is also developed as a part of the tool. Finally, the coupled solar power estimator can be applied to past, present, or future solar power predictions. Initial results obtained using the developed tool are demonstrated for an urban neighborhood in Singapore.