Assessing solar potential of commercial and residential buildings in Indianapolis using LiDAR and GIS modeling

Abstract

Renewable energy systems (RES) have become a vital part of energy use due to the fact that fossil energy declines and demand for energy keeps growing. Among all types of renewable energies, solar energy is the most important resource for the urban areas, since it is easy to obtain, renewable, and produces very little waste. In dense urban areas, building roofs have been utilized or are considered as the locations for Photovoltaic system (PV) or solar panel installation. This research aimed at extracting building roofs from LiDAR data using an object-based segmentation method in City of Indianapolis, USA, and calculating annual solar energy yield for each extracted roof. LiDAR Digital Elevation Model (DEM) was subtracted from LiDAR Digital Surface Model (DSM) to produce the Normalized Height Model (NHM), which represented the absolute heights of objects on the ground. The building extraction was implemented by using a fuzzy rule-based classification method to filter out commonly seen features in the urban environment such as trees, lawns, and roads. Annual solar radiation yield for each roof plane was calculated based on a hemispherical viewed algorithm. Shadowing effect caused by surrounding trees and solid objects were also taken into account to obtain more accurate solar energy values. Finally, suitability of PV installation for individual roofs was rated based on their annual solar energy output. All extracted roof planes were saved as polygons containing the information of size, azimuth, slope, and exposure for future use.