A Study on the Enhancing Uncertainty Management Strategies in Day-Ahead UC in Korean Power System

Abstract

In contemporary society, stable power supply interruptions can cause significant economic losses and social disruptions. Preventing blackouts is the main mission for Independent System Operators (ISOs), but preventing emergency stages of power supply and demand due to uncertainties is also an important task for ISOs, especially when there is enough capacity reserve. Recently, due to the dramatic increa se in solar generation, the uncertainty of power supply and demand has spiked. Nowadays, additional static reserve margins are considered when planning day-ahead unit commitment (UC), but given the current conditions of the electric power industry, more efficient methods are needed for dealing with uncertainty. This paper compares the effectiveness of current deterministic operational strategies with probabilistic approaches to address future power supply uncertainties. Using actual data from Korea's large-scale power system, several methods for incorporating real-time power supply uncertainties a day in advance are explored. A methodology is presented for comparing different response strategies, including the current approach and more advanced probabilistic methods. Simulations were conducted using actual power demand data and hypothetical weather forecast uncertainties to predict net demand and measurable solar power generation. The results show that incorporating Dynamic Reserve in day-ahead planning with Deterministic Unit Commitment (DUC) can lead to cost savings in power generation. However, securing both upward and downward reserves effectively requires Stochastic Unit Commitment (SUC) with Multi-Scenario, which proves to be the most effective strategy and offers a viable alternative for addressing the major issue of securing downward reserves in recent power systems. It is important to note that both SUC with Multi-Scenario and Dynamic Reserve approaches assume the ability to know the uncertainty of day-ahead weather forecasts. Until this assumption can be reliably integrated into power demand-supply operations, the current uncertainty management approach remains the most practical solution. This research contributes to the ongoing effort by providing a comparative analysis of different approaches and highlighting the potential for improved power supply management in Korea's large-scale power systems.