Integrating Hydrological Variability into Electromechanical Stability: Evidence from the Brazilian Interconnected Power System

Hydropower-dominated systems face inherent uncertainty due to inflow variability, which propagates into effective water head levels at turbines. While energy planning models treat such uncertainty stochastically, electromechanical transient studies have traditionally assumed nominal head values, potentially misrepresenting dynamic security margins. This paper introduces a scenario-driven methodology that integrates hydrological variability from planning studies into transient stability simulations. Uncertain drivers (notably water head) are injected as scenario parameters, and ensembles of time-domain runs are summarized through statistical aggregates and event-level indicators. Application to the Brazilian Interconnected Power System demonstrates three key findings: (i) frequency nadirs shift by 0.08-0.13~Hz across credible hydrological conditions - a magnitude comparable to typical under-frequency load shedding thresholds; (ii) infeasible governor initialization occurs in up to 60\% of scenarios for some plants, indicating hidden operational risks; and (iii) damping characteristics vary with hydrology, even when average mechanical power outputs remain close to nominal. These results confirm that both inter-annual and intra-annual variability materially affect dynamic performance. By bridging stochastic energy planning with transient stability analysis, the proposed framework offers a replicable pathway to more robust dynamic security assessment in hydro-dominated power systems.