Impacts of Dynamic Probabilistic Reserve Sizing Techniques on Reserve Requirements and System Costs

Abstract

The increasing share of intermittent renewable energy sources (RES) potentially endangers a reliable power system operation. Due to the variability and unpredictability of added RES, reserve requirements will increase in the future. To counter this, adequate reserve sizing techniques are of major importance. While most system operators apply simple deterministic or probabilistic models assuming RES forecast errors to follow a Gaussian distribution, we propose an improved dynamic reserve sizing method using nonparametric distributions as a forecast error description. The added value of the presented methodology is the use of a conditional kernel-based estimator in combination with a clustering approach to derive dynamic reserves through a convolution of conditional load, wind and solar forecast errors, and plant outage distributions. For comparability, traditional static and deterministic approaches are applied. Based on recent historical data for the German power system, we quantify reserves and demonstrate the feasibility and economic benefits of this improved approach using a scheduling model.