Probability-based approach for parametrisation of traditional underfrequency load-shedding schemes

Underfrequency load shedding (UFLS) is one of the measures performed when an emergency situation occurs in the power system in order to prevent system collapse. Owing to its importance, a lot of effort is dedicated to research and development of the optimal UFLS. Therefore, several different approaches are available to improve the so-called traditional UFLS schemes into more sophisticated adaptive schemes. On the other hand, the literature lacks a systematic analysis of traditional UFLS schemes, which still present the majority within all UFLS schemes actually used in real power systems. For that reason, the probabilistic power flow is used to quantify the level of performance of many different configurations of traditional UFLS schemes (varied both in the number of shedding steps as well as their size) in the whole variety of operating conditions. With the use of probability power flow the probability density function of active-power imbalance is calculated. The obtained results enable to extract those configurations that yield the best results in most probable events and at the same time avoid giving too much attention on poor performance during less probable system conditions. The results were obtained by performing dynamic simulations on the north-western part of a Slovenian power system.