A contingency severity screening methodology based on static calculations as proxies for dynamic behaviour

This paper presents a system-wide approach for contingency severity screening intended as a practical and scalable precursor to full dynamic security assessment (DSA). DSA is the standard practice used by Transmission System Operators (TSOs) to ensure that power systems remain stable following disturbances, with key variables—such as voltage, frequency, and thermal loading—kept within defined dynamic performance limits. However, applying DSA comprehensively across all credible contingencies and operating points is prohibitive, making a pre-screening stage indispensable. Commonly used static-security filters offer limited insight to transient dynamics, motivating the development of alternative screening strategies. To support early-stage analysis and prioritization, three static indicators are devised as proxies for dynamic behaviour, grounded in the power–angle characteristics determining transient stability. For eachn−1 contingency, these indicators provide a severity estimate and serve as inputs to a supervised learning model trained on labels derived from time-domain simulation. This approach aims to reflect how each disturbance might impact system stability, consistent with the objectives of DSA. Contingency rankings derived from these estimates serve to guide further analysis, enabling TSOs to focus dynamic simulations on the most critical scenarios.