A morphological based PV generation and energy consumption predictive model for Singapore neighbourhood

The Singapore government emphasises the importance of sustainable development of the country and set an ambitious target for the adoption of solar power in 2014. Though solar energy is the most promising renewable energy source for Singapore, there is insufficient space for large scale PV panels deployment and hence PV systems have to be integrated into the building environment. Urban morphology determines the effectiveness of PV panels in the urban environment and hence it is important to design the new developments carefully. However, there is a lack of tool and knowledge to estimate the solar energy potential of a new planning zone, so this study developed morphological based PV output and energy consumption predictive models to identify the urban form for effective PV systems deployment by multilinear regression. Three equations are built for predicting the building cooling energy consumption, PV generation on rooftop and façade and the R 2 of these three equations are 0.33, 0.44 and 0.48 respectively. The regression equations show that the urban setting for maximizing PV generation and minimizing cooling energy consumption is different. Furthermore, the low R 2 values of the three equations reflects that building the predictive models by linear regression may not be the most suitable option and other machine learning algorithms should be explored.