At the 2026 Energy Plus Conference held at COEX, a case study was presented showing how LLM-based forecasting and VPP optimization improved profits by 20-40%.


The expansion of renewable energy sources is necessitating changes in how power systems operate. The increase in variable resources like solar and wind power is increasing output uncertainty, and concerns are growing that maintaining grid stability solely through existing facilities is difficult.

H Energy, a renewable energy platform company, presented a strategy for operating and managing energy storage systems (ESS) using artificial intelligence (AI) as a solution to these problems during a keynote speech at the "2026 Energy Plus Conference" held at COEX in Samseong-dong, Seoul on the 4th. Hosted by the Korea Electrical Manufacturers Association, the conference was designed to discuss changes in the power system and technological challenges following the expansion of renewable energy.

Ham Il-han, CEO of H Energy, who took the stage, diagnosed that the expansion of renewable energy inevitably increases the burden on the power grid. He explained that simply increasing power generation facilities has its limitations, and that advanced operational processes, including ESS, are necessary. He also noted that the electricity market is shifting from a simple power generation-centric structure to one that prioritizes supply and demand balance.

The core of H Energy's solution is AI-based forecasting technology. The company explained that it went beyond existing statistical models and applied a Large Language Model (LLM) to a Foundation Model trained on vast amounts of data, thereby improving the accuracy of power generation and demand forecasting. Explainable AI (XAI) has been introduced here to allow operators to verify the basis for prediction results. This is a measure to reduce uncertainty in the electricity market bidding process.

This predictive technology reportedly led to improved operational performance for virtual power plants (VPPs). While individual power plants experience significant power generation variability depending on weather conditions, when these plants are integrated into a platform and operated collectively, these variances are effectively offset. H Energy has released empirical data demonstrating that applying an algorithm optimized for aggregated resources resulted in a 20-40% increase in VPP operating profits compared to previous models.

Safety issues during ESS operation were also cited as a key challenge. CEO Ham introduced the "ESS OnCare" system, which uses AI to diagnose battery status in real time and respond to abnormalities according to predefined protocols. This system prevents failures and, by assessing the residual value of ESS based on operational data, enables management of the entire asset lifecycle, including reuse and sale.

At the end of his presentation, CEO Ham Il-han assessed that ESS is transitioning beyond physical facilities to software-based assets. H Energy stated that it aims to serve as a technology partner that simultaneously supports power grid stability and profitability, leveraging AI technology and data capabilities specialized in energy assets and operations and management.