Renewable Energy Challenges: Expanding Power Grids Alone Is Not Enough

The expansion of renewable energy is increasing the importance of efficient utilization of existing power grids through generation forecasting and system management. Since renewable energy sources like solar and wind vary based on weather conditions, it is suggested that not only should the power grid be expanded, but also that the operational system of the power grid should be restructured using energy storage systems (ESS) and artificial intelligence (AI).


On November 17, Pohang University of Science and Technology's Open Innovation Big Data Center (OIBC) and the renewable energy platform company H-Energy held a seminar at the Ministry of Climate, Energy, and Environment's conference room on the topic of 'The Role of AI in the Era of Renewable Energy and the Restructuring of Distributed Power Systems.'


During the seminar, it was suggested that to reduce the burden on the power grid due to the expansion of renewable energy, generation forecasting technology should be advanced, and economic compensation should be provided for resources that contribute to power supply and demand adjustments.


Solar and wind energy generation varies with weather conditions, and the power demand and transmission and distribution network capacity differ by region. If generation exceeds consumption in a specific area, reverse flow or system congestion can occur, necessitating output control. Conversely, if actual generation is less than predicted, adjustments to other generation sources or the deployment of reserve power may be required.


To address these issues, it is necessary to not only expand the power grid but also to utilize ESS to store surplus power and adjust electricity usage times through demand response (DR), thereby enhancing the utilization of the existing power grid.


Professor Choi Dong-gu of Pohang University emphasized the need to improve both the capacity and control ability of the power grid and the compensation system in the electricity market. In a situation where distributed power sources are increasing, the power grid must be able to accommodate additional generation resources and adjust output according to system conditions, while the electricity market should provide economic compensation for this supply and demand adjustment capability.


To achieve this, it is necessary to restructure the operational technologies of the power grid, such as distribution network ESS, microgrids, and automated distribution management systems (ADMS), along with the market participation and settlement systems for distributed resources.


AI technology has been proposed as a means to solve the complexities of operating the electricity market due to the increase in distributed resources. According to the Korea Energy Agency's renewable energy cloud platform, as of November 9, there were approximately 216,000 permits for solar power projects under 100 kW in South Korea, with about 149,000 of those projects actually starting operations.


Each power plant has different locations, equipment configurations, and weather conditions, making it challenging to accurately predict generation and assess operational status. In this regard, H-Energy's Chief Technology Officer Im Seong-bin proposed a method that combines machine learning and deep learning-based forecasting technology with existing statistical and optimization algorithms.


AI analyzes weather forecasts, satellite images, and past generation performance to predict generation and market prices. Based on this, virtual power plant (VPP) operators can determine the bidding volume in the electricity market considering expected generation and output control possibilities. When operating ESS, the timing for storing or supplying electricity can be optimized by reflecting battery charge status, storage capacity, and charge/discharge efficiency.


The impact of restructuring the electricity market system on the profitability of power producers was also highlighted as a major issue. Choi In-sun, vice president of the Korea RE100 Council, emphasized that power producers must manage actual generation, electricity selling prices, contract conditions, and system constraints together.


Generation forecasting errors and compliance with dispatch instructions can affect settlements according to market rules. If a direct power purchase agreement (PPA) is made with an electricity purchasing company, the difference between the power plant's supply and the company's actual electricity usage must also be managed.


Accordingly, it has been pointed out that private operators should enhance data accessibility to utilize electricity prices and system information and establish criteria for compensating flexible resources such as ESS and DR based on their actual supply and demand adjustment performance.





* This article has been translated by AI.