Online First

[ Article ]
New & Renewable Energy
Abbreviation: New. Renew. Energy
ISSN: 1738-3935 (Print) 2713-9999 (Online)
Online publication date 13 Jan 2025
Received 04 Feb 2025 Revised 28 Feb 2025 Accepted 29 Apr 2025
DOI: https://doi.org/10.7849/ksnre.2025.0004
비계량 태양광 발전량을 고려한 제주도의 단기 전력수요 예측
김현구1), * ; 김창기2) ; 오명찬3) ; 김대진4) ; 김병기5)
Short-term Load Forecasting of Jeju Island Considering the Behind-the-Meter Solar Power
Hyun-Goo Kim1), * ; Chang Ki Kim2) ; Myeongchan Oh3) ; Dae-Jin Kim4) ; Byungki Kim5)
1)Director, Renewable Energy Institute, Korea Institute of Energy Research
2)Chief, Renewable Energy Big Data Laboratory, Korea Institute of Energy Research
3)Senior Researcher, Renewable Energy Big Data Laboratory, Korea Institute of Energy Research
4)Chief, Electric Power System Laboratory, Korea Institute of Energy Research
5)Principal Researcher, Electric Power System Laboratory, Korea Institute of Energy Research
Correspondence to : *hyungoo@kier.re.kr Tel: +82-42-860-3376
© 2025 by the New & Renewable Energy This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

In 2023, renewable energy accounted for 18.2% of Jeju Island’s electricity demand; however, the continuous expansion of solar and wind power has also resulted in a rapid increase in power curtailment. Expanding renewable energy dissemination and achieving a “Carbon-Free Juju Island” by 2035 requires distributed energy management based on virtual power plants (VPPs) and improved accuracy of short-term load forecasting (STLF), which is the core technology of VPPs. This study conducted a 24-h advanced electricity load forecasting for 13 years (2010–2022) to identify the correction effect of behind-the-meter (BTM) solar power, a major factor in power demand distortion on Jeju Island. The generalized additive model (GAM), which is an extension of a regression model that considers nonlinear factors, was used as the prediction model to distinguish the temperature effect and weekday –weekend characteristics. Because BTM solar power generation had to be estimated using limited data, we assumed that the solar power generation pattern is the same across Jeju Island, further confirming that the prediction accuracy improved when the BTM solar power was corrected. The mean absolute percentage error (MAPE) decreased from 4.4% to 3.9%, and root mean-square error (RMSE) decreased from 37.9 to 35.2 MW.