pith. sign in

arxiv: 2604.06582 · v1 · submitted 2026-04-08 · 📡 eess.SY · cs.SY

DAE Index Reduction for Electromagnetic Transient Models

Fiona Majeau , Jose Daniel Lara , Eduardo Corona , Bri-Mathias Hodge This is my paper

Pith reviewed 2026-05-10 18:34 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords electromagnetic transientDAE index reductionmodular subsystemstransformer modelspower system simulationnumerical integrationmodel constructionmodified nodal analysis
0
0 comments X

The pith

Two modular index-reduced subsystem models let electromagnetic transient models integrate with standard solvers without symbolic algorithms or approximations.

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

Electromagnetic transient models formulated with modified nodal analysis often become index-2 differential-algebraic equations that resist direct numerical integration. The paper derives two modular index-reduced models, one for an isolated transformer and one for a machine-coupled transformer, that convert these systems into forms standard solvers can handle. These models avoid both model approximations and the symbolic index-reduction steps that become prohibitive for large networks. Construction tests on models reaching 1152 buses show orders-of-magnitude drops in memory use and runtime compared with general symbolic methods, shifting the main cost to the integration phase. If the subsystems work as derived, detailed EMT analysis becomes feasible for networks that previously exceeded practical limits.

Core claim

The paper derives and presents two modular index-reduced subsystem models, one isolated transformer and one machine-coupled, that allow EMT models to be integrated with standard solvers without approximations or symbolic algorithms.

What carries the argument

Modular index-reduced subsystem models for isolated and machine-coupled transformers that lower the DAE index for direct use with ordinary numerical integrators.

If this is right

  • EMT models with up to 1152 buses can be constructed with far lower memory and runtime than symbolic index-reduction methods allow.
  • The computational bottleneck moves from model construction to the numerical integration step itself.
  • Standard solvers can integrate the models directly without requiring symbolic representation of the full network.
  • The modular subsystems can be placed inside larger arbitrary topologies while preserving the original EMT behavior.

Where Pith is reading between the lines

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

  • Similar pre-derived index reductions could be developed for other frequent power-system components to extend the approach beyond transformers.
  • The construction-time savings may support embedding detailed EMT models inside real-time or optimization-based grid studies.
  • Tools lacking advanced symbolic engines could now incorporate accurate EMT subsystems that were previously inaccessible.

Load-bearing premise

The index-reduced forms for the isolated and machine-coupled transformer subsystems stay accurate and stable when embedded in arbitrary larger network topologies without needing further adjustments.

What would settle it

A side-by-side run of a 1000-plus-bus EMT network built with the modular subsystems versus the same network built with full symbolic index reduction; any large deviation in transient waveforms or solver failure would show the subsystems do not generalize.

Figures

Figures reproduced from arXiv: 2604.06582 by Bri-Mathias Hodge, Eduardo Corona, Fiona Majeau, Jose Daniel Lara.

Figure 2
Figure 2. Figure 2: Example of a loop of ca￾pacitors and voltage sources. KVL introduces constraint equations. C. Constraint equations from KCL Consider a network that includes an instance of [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The typical algorithmic workflow in software tools for performing [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Schematic of a minimal grid featuring an inverter connected to a [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Schematic of a minimal grid featuring a SG connected to a transformer, [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Memory allocated during build of ODEProblem assuming the initial condition is already known. 10 20 40 60 80 100 200 400 600 800 1000 100 101 102 Number of Buses Memory [GiB] Maximum Resident Set Size (Log-Log Scale) Model Build - General IR Model Build - Custom IR 8 GiB RAM 16 GiB RAM [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of the GIR and CIR index-reduced 9-bus models for a [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 9
Figure 9. Figure 9: Wall clock time to build an ODEProblem assuming the initial condition is already known. The simulation runtime is included for reference. 3) Discussion: These experiments measure the computa￾tional complexity of building an index-reduced EMT-dq model that can be integrated by standard solvers. With GIR, the starting point is a higher-index DAE in symbolic form. With CIR, the starting point is a set of modu… view at source ↗
read the original abstract

Electromagnetic transient (EMT) models are index-2 differential-algebraic equations when they include certain topologies and are formulated with modified nodal analysis. Such systems are difficult to numerically integrate, a challenge that is currently addressed by applying model approximations or reformulating with index-reduction algorithms. These algorithms exist in general-purpose software tools, but their reliance on symbolic representation makes them computationally prohibitive for large network-wide EMT models. This paper derives and presents two modular index-reduced subsystem models that allow EMT models to be integrated with standard solvers, without approximations or symbolic algorithms. Both subsystems include a transformer, one isolated and one machine-coupled. We measure the computational performance of constructing EMT models with up to 1152 buses using the custom subsystem models and the symbolic algorithms. The custom approach reduces memory usage and runtime of model construction by several orders of magnitude compared to the general approach, shifting the bottleneck from construction to integration.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper derives two modular index-reduced DAE subsystem models (one isolated transformer and one machine-coupled transformer) for EMT simulations formulated via modified nodal analysis. These subsystems are claimed to be exactly index-1, dynamically equivalent to the original models, and integrable with standard solvers without approximations or symbolic index-reduction algorithms. Performance is measured on network construction for systems up to 1152 buses, showing orders-of-magnitude reductions in memory and runtime compared to general symbolic methods, shifting the bottleneck to integration.

Significance. If the derived subsystems provably remain index-1 and equivalent under arbitrary network connections, the work would enable scalable EMT modeling for large power systems by removing reliance on expensive symbolic processing. The reported construction speedups on concrete large networks (up to 1152 buses) provide concrete evidence of practical gains in model-building efficiency.

major comments (2)
  1. [Derivation of subsystems and numerical experiments] The central claim requires that the index-reduced forms remain exactly index-1 when inserted into arbitrary larger MNA networks. However, the interface conditions (KCL/KVL at connection buses) are not shown to be free of new algebraic loops that could recreate index-2 structure; the tests on specific topologies up to 1152 buses do not isolate or prove this property for general embeddings.
  2. [Numerical results section] No explicit verification (e.g., index computation, numerical stability metrics, or equivalence checks against the unreduced model) is provided for the integrated systems in the reported experiments; only construction-time metrics are given, leaving dynamic equivalence and stability unconfirmed.
minor comments (2)
  1. [Derivation sections] Notation for the retained variables after index reduction could be clarified with an explicit mapping table from original to reduced DAE variables.
  2. [Abstract and Introduction] The abstract and introduction would benefit from a brief statement of the precise index of the original EMT models and the target index after reduction.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed review of our manuscript on modular index-reduced DAE models for electromagnetic transient simulations. The comments highlight important aspects of the central claims regarding index properties under network embedding and the need for explicit verification. We address each major comment point by point below and outline the revisions we will make.

read point-by-point responses

  1. Referee: The central claim requires that the index-reduced forms remain exactly index-1 when inserted into arbitrary larger MNA networks. However, the interface conditions (KCL/KVL at connection buses) are not shown to be free of new algebraic loops that could recreate index-2 structure; the tests on specific topologies up to 1152 buses do not isolate or prove this property for general embeddings.

    Authors: We appreciate this observation on the generality of the index-1 property. The subsystem models are derived by performing local index reduction on the transformer (isolated or machine-coupled) while retaining standard MNA interface variables for voltages and currents. The KCL/KVL conditions at connection buses are handled through these interface variables without introducing additional algebraic constraints, as the internal higher-index elements have been eliminated in the reduction process. The large-scale experiments up to 1152 buses demonstrate that the models integrate successfully with standard solvers across varied topologies, providing empirical support. To address the request for a more rigorous demonstration, we will revise the manuscript to include an explicit analysis or proof sketch showing that arbitrary embeddings preserve the index-1 structure, based on the modular formulation. revision: partial

  2. Referee: No explicit verification (e.g., index computation, numerical stability metrics, or equivalence checks against the unreduced model) is provided for the integrated systems in the reported experiments; only construction-time metrics are given, leaving dynamic equivalence and stability unconfirmed.

    Authors: The numerical results focus on construction-time and memory metrics because these represent the primary computational bottleneck that our modular approach resolves, shifting the workload to integration. We agree that explicit verification of dynamic equivalence, index, and stability for the integrated systems would strengthen the presentation. In the revised manuscript, we will add targeted checks, such as output equivalence comparisons against unreduced models on representative small-to-medium networks and basic numerical stability indicators from the simulations performed. These additions will confirm the properties for the tested cases while maintaining the emphasis on scalability. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation of modular index-reduced EMT subsystems is independent and self-contained

full rationale

The paper derives two specific index-reduced subsystem models (isolated transformer and machine-coupled) from first-principles DAE analysis of EMT components under modified nodal analysis. No quoted steps reduce by construction to fitted parameters, self-citations, or prior ansatzes; the models are presented as explicit replacements for symbolic index-reduction algorithms, with performance claims limited to construction-time metrics on test networks up to 1152 buses. The central derivation chain does not invoke load-bearing self-citations or rename known results, and the index-1 property is asserted via direct elimination of algebraic variables within the subsystems rather than by tautological redefinition. This is the expected non-finding for a derivation-focused modeling paper.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard DAE theory and modified nodal analysis assumptions plus the specific derivation of the two subsystems; no free parameters, new entities, or ad-hoc axioms are introduced in the abstract.

axioms (2)
  • domain assumption EMT models with certain transformer topologies formulated via modified nodal analysis are index-2 DAEs
    Stated in the abstract as the starting point for the index-reduction need.
  • domain assumption Standard numerical integrators can handle the index-reduced forms without further modification
    Implicit in the claim that the subsystems allow integration with standard solvers.

pith-pipeline@v0.9.0 · 5455 in / 1398 out tokens · 53567 ms · 2026-05-10T18:34:07.395582+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

  1. [1]

    Foundations and Challenges of Low-Inertia Systems (Invited Paper),

    F. Milano, F. D ¨orfler, G. Hug, D. J. Hill, and G. Verbi ˇc, “Foundations and Challenges of Low-Inertia Systems (Invited Paper),” in2018 Power Systems Computation Conference (PSCC). Dublin, Ireland: IEEE, Jun. 2018, pp. 1–25

    work page 2018

  2. [2]

    Understanding Small-Signal Stability of Low-Inertia Systems,

    U. Markovic, O. Stanojev, P. Aristidou, E. Vrettos, D. Callaway, and G. Hug, “Understanding Small-Signal Stability of Low-Inertia Systems,” IEEE Transactions on Power Systems, vol. 36, no. 5, pp. 3997–4017, Sep. 2021

    work page 2021

  3. [3]

    Grid Forming Inverter Small Signal Stability: Examining Role of Line and V oltage Dynamics,

    R. Henriquez-Auba, J. D. Lara, C. Roberts, and D. S. Callaway, “Grid Forming Inverter Small Signal Stability: Examining Role of Line and V oltage Dynamics,” inIECON 2020 The 46th Annual Conference of the IEEE Industrial Electronics Society. Singapore, Singapore: IEEE, Oct. 2020, pp. 4063–4068

    work page 2020

  4. [4]

    Review of Methods to Accelerate Electromag- netic Transient Simulation of Power Systems,

    S. Subedi, M. Rauniyar, S. Ishaq, T. M. Hansen, R. Tonkoski, M. Shirazi, R. Wies, and P. Cicilio, “Review of Methods to Accelerate Electromag- netic Transient Simulation of Power Systems,”IEEE Access, vol. 9, pp. 89 714–89 731, 2021

    work page 2021

  5. [5]

    Revisiting Power Systems Time-Domain Simulation Methods and Models,

    J. D. Lara, R. Henriquez-Auba, D. Ramasubramanian, S. Dhople, D. S. Callaway, and S. Sanders, “Revisiting Power Systems Time-Domain Simulation Methods and Models,”IEEE Transactions on Power Systems, vol. 39, no. 2, pp. 2421–2437, Mar. 2024. PREPRINT 11

    work page 2024

  6. [6]

    The modified nodal approach to network analysis,

    Chung-Wen Ho, A. Ruehli, and P. Brennan, “The modified nodal approach to network analysis,”IEEE Transactions on Circuits and Systems, vol. 22, no. 6, pp. 504–509, Jun. 1975

    work page 1975

  7. [7]

    Topological index-calculation of DAEs in circuit sim- ulation,

    C. Tischendorf, “Topological index-calculation of DAEs in circuit sim- ulation,”ZAMM - Journal of Applied Mathematics and Mechanics, vol. 78, no. S3, pp. 1103–1104, Jan. 1998

    work page 1998

  8. [8]

    U. M. Ascher and L. R. Petzold,Computer Methods for Ordinary Dif- ferential Equations and Differential-Algebraic Equations. Philadelphia: Society for Industrial and Applied Mathematics, 1998

    work page 1998

  9. [9]

    K. E. Brenan, S. L. Campbell, and L. R. Petzold,Numerical Solution of Initial-Value Problems in Differential-Algebraic Equations. Society for Industrial and Applied Mathematics, Jan. 1995

    work page 1995

  10. [10]

    Towards an Open-Source Solution using Modelica for Time-Domain Simulation of Power Systems,

    A. Guironnet, M. Saugier, S. Petitrenaud, F. Xavier, and P. Panciatici, “Towards an Open-Source Solution using Modelica for Time-Domain Simulation of Power Systems,” in2018 IEEE PES Innovative Smart Grid Technologies Conference Europe (ISGT-Europe). Sarajevo: IEEE, Oct. 2018, pp. 1–6

    work page 2018

  11. [11]

    Interaction of transmission network and load phasor dynamics in electric power systems,

    E. Allen and M. Ilic, “Interaction of transmission network and load phasor dynamics in electric power systems,”IEEE Transactions on Circuits and Systems I: Fundamental Theory and Applications, vol. 47, no. 11, pp. 1613–1620, Nov. 2000

    work page 2000

  12. [12]

    Extending the Frequency Bandwidth of Transient Stability Simulation Using Dynamic Phasors,

    M. A. Kulasza, U. D. Annakkage, and C. Karawita, “Extending the Frequency Bandwidth of Transient Stability Simulation Using Dynamic Phasors,”IEEE Transactions on Power Systems, vol. 37, no. 1, pp. 249– 259, Jan. 2022

    work page 2022

  13. [13]

    SUNDIALS: Suite of nonlin- ear and differential/algebraic equation solvers,

    A. C. Hindmarsh, P. N. Brown, K. E. Grant, S. L. Lee, R. Serban, D. E. Shumaker, and C. S. Woodward, “SUNDIALS: Suite of nonlin- ear and differential/algebraic equation solvers,”ACM Transactions on Mathematical Software, vol. 31, no. 3, pp. 363–396, Sep. 2005

    work page 2005

  14. [14]

    EMTDC User’s Guide v4.6,

    “EMTDC User’s Guide v4.6,” Manitoba HVDC Research Centre, Win- nipeg, Manitoba, Canada, Tech. Rep., 2010

    work page 2010

  15. [15]

    F. E. Cellier and E. Kofman,Continuous System Simulation. New York: Springer, 2006

    work page 2006

  16. [16]

    Milano, I

    F. Milano, I. Dassios, M. Liu, and G. Tzounas,Eigenvalue Problems in Power Systems, 1st ed. CRC Press, Dec. 2020

    work page 2020

  17. [17]

    Milano,Power System Modelling and Scripting, ser

    F. Milano,Power System Modelling and Scripting, ser. Power Systems. Berlin, Heidelberg: Springer Berlin Heidelberg, 2010, vol. 0

    work page 2010

  18. [18]

    P. W. Sauer, M. A. Pai, and J. H. Chow,Power System Dynamics and Stability: With Synchrophasor Measurement and Power System Toolbox, second edition ed. Hoboken, NJ, USA: IEEE Press, Wiley, 2017

    work page 2017

  19. [19]

    DifferentialEquations.jl – A Performant and Feature-Rich Ecosystem for Solving Differential Equations in Julia,

    C. Rackauckas and Q. Nie, “DifferentialEquations.jl – A Performant and Feature-Rich Ecosystem for Solving Differential Equations in Julia,” Journal of Open Research Software, vol. 5, no. 1, p. 15, May 2017

    work page 2017

  20. [20]

    University of Colorado Boulder Research Computing, “Alpine,” 2023

    work page 2023

  21. [21]

    ModelingToolkit: A Composable Graph Transformation System For Equation-Based Modeling,

    Y . Ma, S. Gowda, R. Anantharaman, C. Laughman, V . Shah, and C. Rackauckas, “ModelingToolkit: A Composable Graph Transformation System For Equation-Based Modeling,” 2021

    work page 2021

  22. [22]

    PowerSimulationsDynamics.jl – An Open Source Model- ing Package for Modern Power Systems with Inverter-Based Resources,

    J. D. Lara, R. Henriquez-Auba, M. Bossart, D. S. Callaway, and C. Barrows, “PowerSimulationsDynamics.jl – An Open Source Model- ing Package for Modern Power Systems with Inverter-Based Resources,” Mar. 2024

    work page 2024

  23. [23]

    DAE-Index-Reduction-for-EMT-Models,

    F. Majeau, “DAE-Index-Reduction-for-EMT-Models,”

    work page

  24. [24]

    Available: https://github.com/fmajeau/ DAE-Index-Reduction-for-EMT-Models

    [Online]. Available: https://github.com/fmajeau/ DAE-Index-Reduction-for-EMT-Models

    work page