Improving Jeju Island's Power Demand Forecasting Using Electric Vehicle Power Usage Pattern Analysis; A Case Study of Korea

Long-term power demand forecasting is essential for ensuring reliable electricity supply by securing appropriate generation capacity. However, the global shift toward carbon neutrality and the growing electrification of energy use—especially electric vehicles (EVs)—are introducing significant uncertainty and new challenges.

Jeju Island, a major tourist destination in Korea, is experiencing rapid EV adoption, particularly in rental car fleets. Although it is connected to the mainland grid, the region faces severe transmission constraints and a rising share of intermittent renewable energy.

To maintain reliability, KPX(Korea Power Exchange; Power System Operator of South Korea) has traditionally forecast peak demand using historical data such as temperature, GDP, and GRDP via functional coefficient models. However, Jeju’s accelerated EV growth raises concerns about increased forecast errors.

This study analyzes the limitations of the current model and proposes improvements by incorporating Jeju’s EV expansion. Specifically, it examines the island's unique hourly EV power usage patterns and presents a methodology to integrate these into demand forecasting models.