A benchmark model and KPI framework for studying power system restoration of modern grids

Modern power system restoration is increasingly affected by high shares of inverter-based resources, reduced inertia, and more complex dynamic behaviour. Traditional restoration approaches, developed for systems dominated by synchronous machines, may not fully address these conditions, while technologies such as battery energy storage systems present new opportunities that require transparent and reproducible evaluation. This paper presents a benchmark restoration model and a structured set of key performance indicators to enable consistent assessment of restoration performance. The model reproduces essential restoration phenomena such as transformer energisation, sympathetic inrush, and cold load pick-up, and is openly available to support reproducibility. The proposed key performance indicators quantify frequency stability, voltage deviation, generator capability stress, inrush current severity, and mechanical torque loading by capturing extreme values, violation durations, and combined measures of disturbance impact. Examples from benchmark scenarios show that the key performance indicators summarise complex transient behaviour into concise metrics that reveal operational limits, equipment stress, and the influence of individual restoration steps. Current limitations include simplified component representations, lack of detailed protection modelling, and a focus on bottom-up strategies, all of which indicate clear paths for further development.