Large-scale Interconnection Power System Model Sanity Check, Tuning, and Validation for Frequency Response Study

TL;DR

This paper demonstrates a method for validating and tuning large-scale power system models using high-frequency measurement data from WAMS, improving model accuracy for frequency response studies.

Contribution

It introduces a systematic approach to check, tune, and validate large-scale power system models with real measurement data, enhancing their reliability.

Findings

Model tuning improves frequency response accuracy.

Quantitative metrics effectively validate model performance.

Validated model closely matches actual system measurements.

Abstract

The quality and accuracy of power system models is critical for simulation-based studies, especially for studying actual stability issues in large-scale systems. With the deployment of wide-area monitoring systems (WAMSs), the high-reporting-rate frequency measurement provides a trustworthy ground truth for validating models in frequency response studies. This paper documented an effort to check, tune, and validate the U.S. power system model based on a WAMS called FNET/GridEye. Four metrics are used to quantitatively compare the simulation results and the actual measurement, including frequency nadir, RoCoF, settling frequency and settling time. After tuning governor deadband and the governor ratio, the model frequency response shows significant improvement and matches well with the event measurement data. This work serves as an example for tuning and validating large-scale power system models.