Uncertainty analysis of annual daylight metrics and energy yield weighting factors in optimising photovoltaic window orientation

Abstract

Application of semi-transparent photovoltaic (PV) windows in buildings has now become an emerging trend towards sustainable design practice. Such application requires a careful thought on finding the optimum window orientation. From the perspective of energy generation, equator-facing orientation is the best option. However, such orientation may lead to glare and visual discomfort. To solve this problem, optimisation is required, but the outcome often depends on the weighting factors of each metric considered. This study therefore aims to quantify the uncertainty of weighting factors in optimising PV window orientation, by considering various annual daylight metrics and energy yield. Computational modelling and simulation using Rhinoceros and Grasshopper were performed for a building prototype with a single, semi-transparent monocrystalline PV window, under the climate of Gowa, Indonesia (5.2°S). Weighting factor for each metric was randomised using Monte Carlo simulation for 1000 scenarios, and the uncertainty were analysed. When all metrics are considered, on average, the South orientation appears as the best orientation in 73% of the cases, followed by East (24%) and North (3%). When a single metric is assigned to each metric group, the South orientation still appears as the best orientation in 62% of the cases.