A dynamic probabilistic method to determine the reserve forsecondary frequency regulation in electric power systems with highintegration of wind and solar generation

Abstract

The rapid growth of wind and solar generation has introduced significant variability and uncertainty into the operation of electric power systems (SEP), challenging traditional methods for sizing operating reserves. This article presents a dynamic probabilistic method based on the convolution of probability distribution functions, aimed at accurately estimating the reserve required for Secondary Frequency Regulation (RSF) in contexts with high penetration of intermittent renewable sources. The proposed approach employs historical time series data with halfhour resolution for electricity demand, wind generation, and solar generation, applying a discrete convolution technique to combine their respective
probability distributions. Unlike conventional methods, this methodology allows for differentiated reserve estimation according to the hour of the day
and the type of day (weekday or non-weekday), enhancing the precision and operational efficiency of the system. The validity of the method was assessed using data from the Peruvian power system, demonstrating its scalability and adaptability to other systems with increasing renewable energy participation.
The results show significant improvements in the sizing and allocation of SFR, contributing to a more secure, efficient, and reliable operation of modern
power systems.