A Decision-Support Framework for Determining Minimum Synchronous Generation Requirements in Low-Inertia Isolated Power Systems

High renewable penetration in isolated power systems introduces stability and security challenges, primarily due to reduced synchronous inertia and limited fault-current contributions from inverter-based resources (IBRs). This paper presents a data-driven industrial decision-support framework that identifies secure unit commitment (UC) scenarios by integrating data analytics, generation clustering, scenario reduction, and multi-criteria security assessment. Historical measurements are processed, validated, and clustered to extract representative operating points. Candidate UC scenarios are then filtered through a six-stage procedure covering: (1) credible UC combinations; (2) frequency stability constraints; (3) minimum stable generation level; (4) fault level and system strength requirements; (5) static N-k contingency analysis; and (6) transient stability and low-voltage ride-through (LVRT) assessment via time-domain dynamic simulations. Compared to static or rule-based must-run practices, the framework provides transparent, explainable, and reproducible operational intelligence and quantifies trade-offs between security margins and renewable curtailment. A Cyprus transmission system case study shows that the proposed informatics-driven approach sharply reduces the feasible solution space and reveals dominant operational bottlenecks across operating conditions and prospective network developments, thereby supporting robust minimum synchronous generation requirements.