Interrogating the Installation Gap and Potential of Solar Photovoltaic Systems Using GIS and Deep Learning

Sumit Kalyan and Qian (Chayn) Sun
Additional contact information
Sumit Kalyan: Geospatial Sciences, School of Science, RMIT University, Melbourne 3000, Australia
Qian (Chayn) Sun: Geospatial Sciences, School of Science, RMIT University, Melbourne 3000, Australia

Energies, 2022, vol. 15, issue 10, 1-21

Abstract: Non-renewable-resource consumption and global greenhouse-gas (GHG) emissions are critical issues that pose a significant threat to sustainable development. Solar energy is a promising source to generate renewable energy and an appealing alternative electricity source for households. The primary goal of this research is to detect the rooftops that have no solar photovoltaic (PV) system deployed on them but that receive moderate to high solar-energy radiation using the Geographic Information System (GIS) and deep-learning techniques. Although various studies have been conducted on this subject, not many addressed these two issues simultaneously at a residential level. Identifying the installed solar PV systems in a large area can be expensive and time-consuming work if performed manually. Therefore, the deep-learning algorithm is an emerging alternative method to detect objects using aerial images. We employed the Single-Shot-Detector (SSD) model with the backbone of residual neural network 34 (ResNet34) to detect the solar PV systems and used GIS software to compute solar isolation and calculate the electricity production estimate (EPE) of each rooftop. Our results show that the SSD model detected 6010 solar panels on 4150 properties with an accuracy of 78% and observed that there were 176 Statistical Area 1s (SA1s) that had no rooftops with solar PV systems installed. Moreover, the total electricity production from the suitable area was estimated at over 929.8 Giga Watt-hours (GWhs) annually. Finally, the relation between solar-PV-system density and EPE was also identified using the bivariant correlation technique. Detecting the existing solar PV systems is useful in a broad range of applications including electricity-generation prediction, power-plant-production management, uncovering patterns between regions, etc. Examination of the spatial distribution of solar-energy potential in a region and performing an overlay analysis with socio-economic factors can help policymakers to understand the explanation behind the pattern and strategize the incentives accordingly.

Keywords: renewable energy; solar PV systems; sustainable development; GIS; deep learning; Ballarat (search for similar items in EconPapers)
JEL-codes: Q Q0 Q4 Q40 Q41 Q42 Q43 Q47 Q48 Q49 (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
https://www.mdpi.com/1996-1073/15/10/3740/pdf (application/pdf)
https://www.mdpi.com/1996-1073/15/10/3740/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jeners:v:15:y:2022:i:10:p:3740-:d:819447

Access Statistics for this article

Energies is currently edited by Ms. Agatha Cao

More articles in Energies from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().