An Efficient Method for Estimating Inertia of a PV-Integrated Power Grid for Enhanced Security

Abstract

The decommissioning of traditional thermal power plants with high levels of renewable energy resources (RES) always causes a reduction in system inertia. A hybrid power grid must maintain stable frequency during severe disturbances. Therefore, this study provides a generalized approach for estimating the online grid inertia under normal and disturbed operating conditions. Large-scale deployment of photovoltaic (PV) power generation is considered to replicate the rapid growth of RES and its impact on actual power grids. The individual inertia of existing conventional synchronous generators in the hybrid system is quickly estimated using the online recursive least square (RLS) method. A Kalman filter is used to reject any noise contained in the measured signals. The noise-free rate of change of frequency (RoCoF) and electrical power are further used to refine the estimation of inertia, damping coefficient, and mechanical power to enhance the overall accuracy of the proposed tracking scheme. Next, the virtual inertia of the existing PV system is derived mathematically from the identified model parameters, which are further verified using simulation results, thereby demonstrating the robustness of the proposed approach. Finally, the spectral cluster-based method is used to determine the sets of coherent machines and provide online tracking for each coherent group’s inertia, as well as the total grid inertia in response to different disturbances and intermittency of RES power generation. The proposed inertia estimation method is proven to be very fast, allowing preventative action to be taken before any instability occurs.The effectiveness of the proposed method is tested and verified using IEEE 39-bus, 68-bus, an islanded AC microgrid, as well as a modified IEEE 39-bus based MTDC test system.