Pith. sign in

REVIEW 2 major objections 1 minor 16 references

Reviewed by Pith at T0; open to challenge.

T0 means a machine referee read the full paper against a public rubric. The mark states how deep the mechanical check went, never who wrote it. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.3

The PROTECT-90 dataset supplies 9022 open EMT simulations of high-voltage faults to standardize protection algorithm benchmarking.

2026-06-25 22:43 UTC pith:2KDXGH4D

load-bearing objection The paper's contribution is a documented release of 9022 EMT-simulated fault episodes on one fixed 90 kV double-line topology, with metadata and open access. the 2 major comments →

arxiv 2606.24298 v1 pith:2KDXGH4D submitted 2026-06-23 eess.SP cs.LG

PROTECT-90: A Fault Dataset for Power System Protection

Julian Oelhaf , Georg Kordowich , Christian Bergler , Andreas Maier , Johann J\"ager , Siming Bayer This is my paper
classification eess.SP cs.LG
keywords power system protectionfault datasetelectromagnetic transient simulationhigh-voltage networksbenchmarking datasetshort-circuit analysisdigital fault recorderdata-driven protection
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Power system protection studies increasingly rely on data-driven signal processing and learning methods, yet lack shared public waveform collections that support direct comparisons. This paper releases the PROTECT-90 collection of 9022 electromagnetic transient simulations performed on a fixed 90 kV double-line network. Every episode records three-phase voltages and currents at eight synchronized locations together with machine-readable labels for fault type, location, inception time, and operating state. Parameter ranges for grid conditions, line impedances, and fault scenarios are fully documented so that later studies can regenerate or extend the same distribution. Adoption of this resource would let separate research groups evaluate their algorithms on identical, physically consistent data without each team building its own simulation setup.

Core claim

PROTECT-90 is an open electromagnetic transient simulated reference benchmark consisting of 9022 short-circuit episodes on a standardized 90 kV double-line topology. It provides synchronized voltage and current waveforms at multiple measurement points together with machine-readable metadata for each fault scenario. All modeling choices and randomization ranges are documented to support reproducible evaluation of protection methods.

What carries the argument

The 90 kV double-line EMT simulation model with domain randomization of operating points, line parameters, and fault conditions, generating consistent digital-fault-recorder-like measurements.

Load-bearing premise

The chosen 90 kV double-line topology together with the documented ranges for domain randomization of operating points, line parameters, and fault conditions are representative of real high-voltage systems for the purpose of general benchmarking.

What would settle it

A side-by-side comparison showing that key waveform features like fault inception transients or impedance trajectories in the dataset differ systematically from those measured in actual 90 kV or similar high-voltage networks.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Researchers can benchmark their protection algorithms against a common set of physically consistent fault waveforms.
  • Comparisons between signal processing techniques and learning-based methods become possible on identical data.
  • New methods can be developed and validated without needing to generate custom simulations each time.
  • The documented parameter ranges allow controlled testing of robustness to variations in grid conditions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The dataset could serve as a starting point for extending to other voltage levels or network topologies if similar documentation is followed.
  • Integration with real-world fault records might reveal how well the simulations capture actual system behaviors.
  • Standardized metadata could enable automated evaluation pipelines for protection studies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 1 minor

Summary. The paper introduces PROTECT-90, an open dataset of 9,022 EMT-simulated short-circuit episodes generated on a standardized 90 kV double-line topology. It includes synchronized three-phase voltage and current waveforms at eight measurement locations, structured metadata on fault type/location/inception/operating conditions, and explicit documentation of domain randomization over grid operating points, line parameters, and fault conditions. The central claim is that this combination of physically grounded simulation, balanced scenario coverage, and public release establishes a standardized benchmark for reproducible evaluation of protection-oriented signal processing and learning-based methods.

Significance. If the dataset's representativeness and physical consistency hold, the work would provide a valuable public resource addressing the documented scarcity of standardized high-voltage waveform datasets, enabling transparent cross-study comparisons in power system protection research. The emphasis on machine-readable metadata and fully documented generation procedures is a clear strength for reproducibility.

major comments (2)
  1. [Abstract] Abstract: the claim that the 90 kV double-line topology plus documented domain randomization 'establishes a standardized foundation' for general high-voltage protection benchmarking is load-bearing but unsupported. A single fixed topology omits structural variations (bus arrangements, line lengths, meshed vs. radial configurations, transformer interactions) that materially alter fault current paths, zero-sequence components, and transient waveforms; no multi-topology coverage or calibration against other standard test systems is described.
  2. [Abstract] Abstract: the assertion of 'physically consistent' simulations and 'balanced scenario coverage' lacks any validation against measured field data, error analysis of EMT model fidelity, or quantitative checks on randomization coverage (e.g., distribution statistics or sensitivity to parameter ranges). This prevents full assessment of whether the 9,022 episodes suffice for general benchmarking.
minor comments (1)
  1. Ensure consistent formatting of the episode count (9,022) and parameter ranges across all sections and the metadata description.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on PROTECT-90. We respond point-by-point to the major comments below, indicating revisions where the manuscript will be updated for clarity and precision.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that the 90 kV double-line topology plus documented domain randomization 'establishes a standardized foundation' for general high-voltage protection benchmarking is load-bearing but unsupported. A single fixed topology omits structural variations (bus arrangements, line lengths, meshed vs. radial configurations, transformer interactions) that materially alter fault current paths, zero-sequence components, and transient waveforms; no multi-topology coverage or calibration against other standard test systems is described.

    Authors: We agree that the original abstract wording implies broader generality than the work supports. PROTECT-90 is explicitly constructed on one standardized 90 kV double-line topology with domain randomization over operating points, parameters, and faults to enable reproducible comparisons within that setting. We will revise the abstract to state that the dataset provides a standardized benchmark for this topology class rather than claiming a foundation for general high-voltage protection benchmarking. A limitations paragraph will be added noting the lack of multi-topology coverage and calibration to other test systems. revision: yes

  2. Referee: [Abstract] Abstract: the assertion of 'physically consistent' simulations and 'balanced scenario coverage' lacks any validation against measured field data, error analysis of EMT model fidelity, or quantitative checks on randomization coverage (e.g., distribution statistics or sensitivity to parameter ranges). This prevents full assessment of whether the 9,022 episodes suffice for general benchmarking.

    Authors: The dataset is generated entirely from EMT simulation; direct validation against field measurements is not included because the work prioritizes controlled, reproducible, and publicly documented scenarios rather than proprietary utility recordings. All EMT modeling assumptions and parameter ranges are documented in the manuscript. We will add quantitative distribution statistics for the randomized parameters and a brief discussion of EMT fidelity assumptions based on standard practices to allow readers to assess coverage. The 9,022 episodes were sized to achieve balanced coverage across the documented metadata categories. revision: partial

Circularity Check

0 steps flagged

No circularity; dataset release has no derivation chain or self-referential predictions

full rationale

The paper presents a new open EMT-simulated fault dataset on a fixed 90 kV topology with documented randomization. No equations, fitted parameters, predictions, or uniqueness theorems appear. The contribution is the generation procedure and public release itself; the reader's circularity score of 0.0 is confirmed by inspection of the abstract and described content. No load-bearing step reduces to its own inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The paper's value rests on the domain assumption that EMT simulation produces usable benchmark data and on the authors' choices of topology and randomization ranges; these choices are stated to be documented but remain free parameters of the generation process.

free parameters (1)
  • ranges for grid operating points, line parameters, and fault conditions
    These ranges are selected to generate the 9,022 episodes and are part of the documented domain randomization; their specific bounds constitute modeling choices.
axioms (1)
  • domain assumption Electromagnetic transient simulation produces physically consistent short-circuit waveforms suitable for protection studies
    The dataset is generated via EMT simulation and the abstract claims physical consistency on this basis.

pith-pipeline@v0.9.1-grok · 5734 in / 1280 out tokens · 25118 ms · 2026-06-25T22:43:24.945364+00:00 · methodology

0 comments
read the original abstract

The increasing interest in data-driven methods for power system protection is accompanied by a lack of standardized, publicly available high-voltage waveform datasets that enable transparent and reproducible evaluation. To address this gap, this paper introduces the PROTECT-90 dataset, an open electromagnetic transient (EMT)-simulated reference benchmark for high-voltage fault studies with consistent digital-fault-recorder-like measurements, publicly released with this work. The dataset comprises 9,022 physically consistent short-circuit simulation episodes generated on a standardized 90 kV double-line topology with systematically documented domain randomization of grid operating points, line parameters, and fault conditions. For each episode, synchronized three-phase voltage and current waveforms are recorded at eight measurement locations and released together with structured, machine-readable metadata describing fault type, fault location, inception time, and operating conditions. All modeling assumptions, parameter ranges, and data-generation procedures are explicitly documented to ensure transparency and cross-study comparability. By combining physically grounded EMT simulation, balanced scenario coverage, and open accessibility, PROTECT-90 establishes a standardized foundation for reproducible benchmarking of protection-oriented signal processing and learning-based methods.

Figures

Figures reproduced from arXiv: 2606.24298 by Andreas Maier, Christian Bergler, Georg Kordowich, Johann J\"ager, Julian Oelhaf, Siming Bayer.

Figure 1
Figure 1. Figure 1: Standardized 90 kV high-voltage double-line topology used for EMT dataset generation. Filled squares denote protection relay measurement locations. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Representative three-phase voltage (top) and current (bottom) wave [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Coverage statistics of the EMT dataset. (a) Distribution of short-circuit fault types, showing balanced sampling across SLG, LL, LLG, and LLL 1 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

16 extracted references · 10 canonical work pages · 2 internal anchors

  1. [1]

    A Practical Method for the Direct Analysis of Transient Stability,

    T. Athay, R. Podmore, and S. Virmani, “A Practical Method for the Direct Analysis of Transient Stability,”IEEE Transactions on Power Apparatus and Systems, vol. PAS-98, no. 2, pp. 573–584, Mar. 1979. [Online]. Available: http://ieeexplore.ieee.org/document/4113518/

    work page arXiv 1979

  2. [2]

    Vittal and J

    V . Vittal and J. D. McCalley,Power system control and stability. Hoboken, New Jersey: Wiley, 2020

    2020

  3. [3]

    IRTSD: Open-Source Data and Toolset for Electromagnetic Transient Analysis of Disturbances and IBR Control Malfunctions in Transmission Systems,

    Brett Ross and Kaveri Mahapatra, “IRTSD: Open-Source Data and Toolset for Electromagnetic Transient Analysis of Disturbances and IBR Control Malfunctions in Transmission Systems,” 2024. [Online]. Available: https://dx.doi.org/10.21227/mp6d-j677

    work page doi:10.21227/mp6d-j677 2024

  4. [4]

    Synchrophasors and Synchrowaveforms for the Distribution Grid: The SoCal 28-Bus Dataset

    Y . Xie, L. Werner, K. Chen, T.-L. Le, C. Ortega, and S. Low, “A Digital Twin of an Electrical Distribution Grid: SoCal 28-Bus Dataset,” Apr. 2025, arXiv:2504.06588 [eess]. [Online]. Available: http://arxiv.org/abs/2504.06588

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  5. [5]

    Database of V oltage and Current Samples Values from the French Electricity Transmission Grid, R ´eseau de Transport d’Electricit ´e (RTE), France,

    C. Presv ˆots and T. Prevost, “Database of V oltage and Current Samples Values from the French Electricity Transmission Grid, R ´eseau de Transport d’Electricit ´e (RTE), France,” 2024. [Online]. Available: https://github.com/rte-france/digital-fault-recording-database/

    2024

  6. [6]

    The Grid Event Signature Library: An Open-Access Repository of Power System Measurement Signatures,

    A. J. Wilson, A. Riza Ekti, J. Follum, S. Biswas, C. Annalicia, J.-Y . Joo, O. Aziz, and J. Lian, “The Grid Event Signature Library: An Open-Access Repository of Power System Measurement Signatures,” IEEE Access, vol. 12, pp. 76 207–76 218, 2024. [Online]. Available: https://ieeexplore.ieee.org/document/10538319/

    work page arXiv 2024

  7. [7]

    A Scoping Review of Machine Learning Applications in Power System Protection and Disturbance Management,

    J. Oelhaf, G. Kordowich, M. Pashaei, C. Bergler, A. Maier, J. J ¨ager, and S. Bayer, “A Scoping Review of Machine Learning Applications in Power System Protection and Disturbance Management,”International Journal of Electrical Power & Energy Systems, vol. 172, p. 111257, Nov. 2025. [Online]. Available: 10.1016/j.ijepes.2025.111257

    work page doi:10.1016/j.ijepes.2025.111257 2025

  8. [8]

    Ziegler,Numerical distance protection: principles and applications, 4th ed

    G. Ziegler,Numerical distance protection: principles and applications, 4th ed. Erlangen: Publicis Publishing, 2011

    2011

  9. [9]

    Roeper and Mitlehner, Friedrich,Kurzschlußstr ¨ome in Drehstromnet- zen, 6th ed

    R. Roeper and Mitlehner, Friedrich,Kurzschlußstr ¨ome in Drehstromnet- zen, 6th ed. Berlin, M ¨unchen: Siemens: Publicis Corporate Publishing, 1984

    1984

  10. [10]

    Oeding and B

    D. Oeding and B. R. Oswald,Elektrische Kraftwerke und Netze, 8th ed. Berlin, Heidelberg: Springer Berlin Heidelberg, 2016. [Online]. Available: http://link.springer.com/10.1007/978-3-662-52703-0

    work page doi:10.1007/978-3-662-52703-0 2016

  11. [11]

    A Generic Data Generation Framework for Short Circuit Detection Training of Neural Networks,

    M. Wang, G. Kordowich, and J. J ¨ager, “A Generic Data Generation Framework for Short Circuit Detection Training of Neural Networks,” inPESS + PELSS 2022; Power and Energy Student Summit, 2022

    2022

  12. [12]

    PROTECT-90: A Fault Dataset for Power System Protection: Open, Standardized V oltage and Current Waveforms for Reproducible Protection and Transient Analysis,

    G. Kordowich, J. Oelhaf, C. Bergler, A. Maier, S. Bayer, and J. J ¨ager, “PROTECT-90: A Fault Dataset for Power System Protection: Open, Standardized V oltage and Current Waveforms for Reproducible Protection and Transient Analysis,” Mar. 2026. [Online]. Available: https://zenodo.org/doi/10.5281/zenodo.18418330

    work page doi:10.5281/zenodo.18418330 2026

  13. [13]

    Controlled Comparison of Machine Learning Models for Fault Classification and Localization in Power System Protection

    J. Oelhaf, G. Kordowich, C. Kim, P. A. P ´erez-Toro, C. Bergler, A. Maier, J. J ¨ager, and S. Bayer, “Controlled Comparison of Machine Learning Models for Fault Classification and Localization in Power System Protection,” 2025, version Number: 2. [Online]. Available: https://arxiv.org/abs/2510.00831

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  14. [14]

    A Systematic Evaluation of Machine Learning Methods for Fault Detection and Line Identification in Electrical Power Grids,

    J. Oelhaf, G. Kordowich, P. A. P ´erez-Toro, T. Arias-Vergara, A. Maier, J. J ¨ager, and S. Bayer, “A Systematic Evaluation of Machine Learning Methods for Fault Detection and Line Identification in Electrical Power Grids,” inICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Hyderabad, India: IEEE, Apr. ...

    work page arXiv 2025

  15. [15]

    Impact of Data Sparsity on Machine Learning for Fault Detection in Power System Protection,

    J. Oelhaf, G. Kordowich, C. Kim, P. A. P ´erez-Toro, A. Maier, J. J ¨ager, and S. Bayer, “Impact of Data Sparsity on Machine Learning for Fault Detection in Power System Protection,” in 2025 33rd European Signal Processing Conference (EUSIPCO). Palermo, Italy: IEEE, Sep. 2025, pp. 1997–2001. [Online]. Available: https://ieeexplore.ieee.org/document/11226584/

    work page arXiv 2025

  16. [16]

    Robustness Evaluation of Machine Learning Models for Fault Classification and Localization in Power System Protection,

    J. Oelhaf, M. Pashaei, G. Kordowich, C. Bergler, A. Maier, J. J ¨ager, and S. Bayer, “Robustness Evaluation of Machine Learning Models for Fault Classification and Localization in Power System Protection,” in 20th IET International Conference on Developments in Power System Protection (DPSP Global 2026). IET, 2026

    2026