PROTECT-90: A Fault Dataset for Power System Protection

PROTECT-90 is an open high-voltage fault waveform dataset for reproducible research in power system protection and transient analysis.

The dataset comprises 9,022 electromagnetic transient (EMT) simulation episodes generated on a standardized 90 kV double-line topology under physically constrained domain randomization. Each episode provides 1 second of synchronized three-phase voltage and current waveforms sampled at 6.4 kHz (128 samples per 50 Hz cycle), recorded at eight protection-relevant measurement locations (48 channels total).

All episodes are released as raw instantaneous waveforms (no feature engineering) and are accompanied by structured, machine-readable metadata describing fault type, faulted line segment, normalized fault location, fault resistance, inception time, randomized grid parameters, and topology switching states.

All modeling assumptions, parameter ranges, and simulation settings are fully documented to enable transparent, standardized, and cross-study comparable benchmarking.

PROTECT-90 supports evaluation of:

Fault classification

Fault localization

Line identification

Transient waveform analysis

Robustness and domain-shift studies

The dataset is released under an open license and is intended as a physically grounded reference benchmark for learning-based and signal-processing methods in high-voltage protection research.