Data-Center Synchrophasor Frequency-Set Benchmark: ParaEMT Waveforms and Estimator Replay

Version 1.0.0 of the Data-Center Synchrophasor Frequency-Set Benchmark accompanies the manuscript Estimator-Specific Closed-Catalogue Frequency Sets from Reporting-Rate-Aliased Synchrophasor Measurements: A Data-Center Case Study (under review). The dataset contains 104 source-bus three-phase voltage records from nonlinear ParaEMT simulations on two separately AC-solved IEEE 39-bus conditions, 384 original source-bus PMU-derived streams (one- and two-cycle filters at 30, 60 and 120 frames/s), eight 20-kHz terminal histories, forcing definitions, retained results, and SHA-256 manifests. Portable Python code replays the research synchrophasor estimators, frequency-set inference, noise studies, empirical calibration and off-catalogue stress tests. Data include 64 original trials and 40 additional nominal, frequency-offset and frequency-drift trials. The forcing catalogue is 14.7, 29, 31 and 45.3 Hz. The 76,800 channel-noise trials are dependent perturbations of the original events, not independent physical simulations. Negative and unresolved outcomes are retained. Extract the code ZIP, then extract the data ZIP into the code directory; follow README.md and REPRODUCIBILITY.md. The release reproduces measurement analysis from saved numerical waveforms. It does not include a complete EMT solver environment, commercial PMU validation, confidential data-center measurements, third-party field recordings, or reviewer correspondence. Resampling saved 1-kHz waveforms to 4.8 kHz provides estimator compatibility without recovering lost bandwidth. Simulation data and documentation are licensed CC BY 4.0. Own code is BSD-3-Clause; the relevant license and ParaEMT upstream notice are included. Code repository: https://github.com/mshams8/datacenter-pmu-frequency-sets.