System Frequency Nadir and Trajectory Prediction With Discontinuity Constraints in Governor Dynamics

Accurately predicting the frequency nadir and estimating the overall frequency trajectory are crucial analytical tasks in power system planning. Given the large number of operating scenarios and contingency events that must be evaluated, low-order frequency nadir prediction (FNP) models have been recently developed to avoid the computational burden of full dynamic simulations in large, complex systems. However, a major technical limitation of existing FNP models is their inability to capture inherent discontinuities such as limits, piecewise functions, and deadbands that strongly influence the actual frequency dynamics. To overcome these challenges, this paper proposes a discontinuity-aware frequency nadir prediction (DA-FNP) model that explicitly implements discontinuity constraints into the frequency response estimation. By implementing these discontinuities, the model not only predicts the system frequency trajectory with high fidelity but also identifies which generators are subject to enforced constraints. This capability provides new insights for system planners, enabling a more realistic evaluation of frequency security margins and resource adequacy in future power systems with high renewable penetration. The methodology is validated against detailed dynamic simulations on both small- and large-scale synthetic grids. The case study demonstrates significant enhancement on the accuracy of system configuration and system frequency trajectory, while retaining computational efficiency of low-order models. Furthermore, the approach offers a practical and scalable tool for planning studies in large, complex power systems.