Innovative renewable energy solutions for antarctic research stations: A case study on the photovoltaic systems in Horseshoe Island

Abstract

Energy generation in regions with extreme environmental conditions, such as Antarctica, poses a significant challenge regarding logistical and environmental sustainability. Traditional diesel generators are not sustainable due to high cost and refueling challenges. This study aims to investigate the performance of photovoltaic (PV) panels in Antarctic conditions with experimental and artificial intelligence-supported analyses within the scope of the 8th National Antarctic Science Expedition carried out by Turkey at Horseshoe Island. The study tested four different PV panel types: monocrystalline, polycrystalline, transparent, and flexible. The experiments were conducted under an average solar irradiance of 803.04 W/m2 and a relative humidity of 64.66 %, with a temperature range of - 0.8 degrees C to 5.8 degrees C. The maximum power generation of 21.2 W was observed for monocrystalline panels, while the average power generation was 16.9 W for monocrystalline panels. Additionally, the energy efficiency of PV panels ranged from 13 % to 20 %, while exergy efficiency varied between 1 % and 13 %. Using Pace and ElasticNet regression algorithms, the effect of environmental factors on panel performance was modeled with 95 % accuracy, and energy efficiency was optimized. During the study, 6146 data observations were collected, and 2380 data samples were analyzed using artificial intelligence models. This innovative approach provides a critical roadmap for research stations in Antarctica and increasing energy independence in extreme conditions worldwide.