arxiv: 2605.13205 · v1 · submitted 2026-05-13 · 🧮 math.OC

Recognition: no theorem link

Voltage-Aware Grid Aggregation: Expanding the European High-Voltage Network

Benjamin St\"ockl , Marco Anarmo , Sonja Wogrin , Yannick Werner

Authors on Pith no claims yet

Pith reviewed 2026-05-14 17:54 UTC · model grok-4.3

classification 🧮 math.OC

keywords voltage-aware aggregationpower grid aggregationtransformer expansion planningEuropean electricity networknetwork expansion optimizationspatial aggregationPyPSA

0 comments

The pith

A voltage-aware aggregation method for European power grids preserves up to 70 percent of transformer expansion costs in simplified models.

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

The paper proposes a network partitioning and aggregation technique that keeps separate voltage levels and the transformers linking them instead of collapsing everything to a single level. Existing aggregation approaches lose the costs and decisions tied to transformers, which matter for planning how much to upgrade the grid during the energy transition. The new method is tested on a European case study solved with PyPSA, where it retains a much larger share of the transformer investments that appear in the full-resolution model. Readers would care because the approach lets large-scale optimization models stay computationally tractable while still producing credible investment signals for transformers.

Core claim

We propose a novel voltage-aware network partitioning and aggregation methodology that preserves individual voltage levels and transformers. When this method is applied to a European network expansion problem, the aggregated model preserves up to 70 percent of the transformer expansion costs obtained from the full grid model, thereby improving the accuracy of investment decisions for transformers.

What carries the argument

voltage-aware network partitioning and aggregation methodology that keeps distinct voltage levels and the transformers connecting them

If this is right

Aggregated models can now produce investment decisions for transformers that are much closer to those of the full network.
The computational cost of European-scale expansion planning drops while still retaining key voltage-specific constraints.
Planning studies that previously ignored transformers because of aggregation errors can now include them without restoring full network detail.

Where Pith is reading between the lines

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

The same partitioning logic could be tested on grids outside Europe that also operate at multiple transmission voltages.
Extending the method to include other voltage-sensitive components such as reactive power devices would be a direct next step.
If the 70 percent preservation holds under different cost assumptions or renewable penetration levels, the technique could become a standard preprocessing step for any large transmission model.

Load-bearing premise

The voltage-aware partitioning and aggregation accurately captures the essential physical flows, costs, and expansion decisions of the original network, especially for transformers.

What would settle it

Running the same European expansion problem on the full network and on the voltage-aware aggregated network and finding that transformer expansion costs preserved in the aggregated model fall substantially below 70 percent of the full-model costs.

Figures

Figures reproduced from arXiv: 2605.13205 by Benjamin St\"ockl, Marco Anarmo, Sonja Wogrin, Yannick Werner.

**Figure 2.** Figure 2: Representation of the European high-voltage network taken from [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Figure 3: Network partitioning and aggregation results obtained with NPAP [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: Network expansion results shown for Sicily for the models representing the full grid and the voltage-unaware and voltage-aware aggregations, [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

read the original abstract

Energy system optimization models are indispensable for planning the European energy transition. Yet their applicability is constrained by the fundamental trade-off between spatial detail and computational tractability. Modelers often tackle this by spatially aggregating electricity networks. Existing methods, however, neglect differences in voltage levels, reducing them to a single level and thereby overlooking the critical role of transformers in expansion planning. Therefore, we propose a novel voltage-aware network partitioning and aggregation methodology that preserves individual voltage levels and transformers. We demonstrate the effectiveness of this approach and compare it against a voltage-unaware grid aggregation by solving a network expansion problem for a European case study using PyPSA. Our findings show that the proposed methodology preserves up to 70% of the transformer expansion costs in the aggregated model compared to the full grid model, thereby significantly improving the accuracy of investment decisions for transformers in the aggregated grid.

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.

Desk Editor's Note private letter to a colleague

The paper gives a voltage-aware aggregation method that keeps transformers and voltage levels separate, with a 70% cost preservation result in one PyPSA run, but the validation details are too thin to judge how general the improvement is.

read the letter

The new piece is a partitioning and aggregation scheme that treats voltage levels as distinct and keeps the transformers between them, rather than collapsing the whole network to a single voltage. They run a network expansion problem on a European PyPSA model and report that the aggregated version captures up to 70% of the transformer expansion costs that appear in the full-resolution run, beating a voltage-unaware baseline. That is a concrete, practical step for anyone who needs to run continental-scale capacity expansion without losing the transformer decisions that matter for real investment planning. The comparison to the full grid is direct and the target problem is relevant, so the work earns credit for addressing a modeling gap that existing single-level methods ignore. The main limitation is that the abstract and available description give the 70% figure without showing whether the aggregated model reproduces the same optimal flows, nodal balances, line loadings, or marginal costs on voltage constraints. If those quantities diverge, the cost preservation could be an artifact of the particular topology or solver path rather than a reliable property of the method. No sensitivity checks or dual-value comparisons are mentioned, which leaves the central claim under-supported for now. This is aimed at energy-system modelers who already use PyPSA or similar tools for European studies and who need better spatial reduction without throwing away transformer economics. It is not a foundational theoretical advance, but it is a usable incremental fix. I would send it to peer review so referees can check the flow equivalence and ask for the missing validation metrics.

Referee Report

2 major / 2 minor

Summary. The paper proposes a novel voltage-aware network partitioning and aggregation methodology for high-voltage electricity grids that preserves distinct voltage levels and transformer representations, unlike conventional single-level reductions. In a European case study solved with PyPSA, the approach is compared to a voltage-unaware aggregation and is reported to preserve up to 70% of transformer expansion costs relative to the full grid model, thereby improving the accuracy of investment decisions in the reduced network.

Significance. If the equivalence of flows and expansion signals holds, the method would meaningfully advance spatial aggregation techniques in energy system optimization by retaining critical transformer cost structures and voltage constraints that are typically lost. This could support more reliable large-scale planning for the European energy transition without sacrificing as much fidelity as standard aggregations.

major comments (2)

[Abstract / Results] Abstract and results section: the central quantitative claim of preserving 'up to 70% of the transformer expansion costs' is presented without any reported comparison of optimal power flows, nodal injections, line loadings, or dual values on voltage bounds between the full and aggregated models. This leaves open whether the percentage reflects preserved physical equivalence or case-specific solver behavior.
[Methodology] Methodology section: the voltage-aware partitioning must demonstrably maintain equivalent power balance, voltage constraints, and transformer cost structures; no explicit verification (e.g., via constraint residual checks or marginal price comparisons) is provided to confirm that the reduced network reproduces the same expansion signals for transformers as the original.

minor comments (2)

[Abstract] The abstract would benefit from stating the number of nodes, lines, and voltage levels in the European PyPSA case study to allow readers to gauge the scale of the aggregation.
[Notation / Introduction] Notation for voltage levels and transformer variables could be introduced with a small schematic diagram for clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thoughtful review and constructive feedback on our manuscript. We address each major comment below, clarifying the basis of our claims and outlining revisions to strengthen the presentation of physical equivalence and verification.

read point-by-point responses

Referee: [Abstract / Results] Abstract and results section: the central quantitative claim of preserving 'up to 70% of the transformer expansion costs' is presented without any reported comparison of optimal power flows, nodal injections, line loadings, or dual values on voltage bounds between the full and aggregated models. This leaves open whether the percentage reflects preserved physical equivalence or case-specific solver behavior.

Authors: The 70% figure is obtained directly from the objective values of the transmission expansion optimization (transformer investment costs) solved on the full-resolution versus aggregated networks using identical PyPSA formulations and solver settings. We agree that explicit side-by-side verification of physical quantities would better demonstrate that the cost preservation arises from retained voltage and transformer structure rather than numerical artifacts. In the revised manuscript we will add a dedicated results subsection reporting (i) maximum line-loading deviations, (ii) nodal price differences, and (iii) voltage-bound dual comparisons between the full and reduced models for the same scenario set. revision: yes
Referee: [Methodology] Methodology section: the voltage-aware partitioning must demonstrably maintain equivalent power balance, voltage constraints, and transformer cost structures; no explicit verification (e.g., via constraint residual checks or marginal price comparisons) is provided to confirm that the reduced network reproduces the same expansion signals for transformers as the original.

Authors: The aggregation procedure preserves power balance by construction: each super-node inherits the net injection of its constituent buses, and inter-voltage transformers remain as explicit edges with their original cost and capacity parameters. Voltage constraints are retained at every voltage level because the partitioning never merges nodes of different nominal voltages. Nevertheless, we accept that the current text lacks explicit post-aggregation verification. We will insert constraint-residual tables and marginal-price correlation plots in the methodology/results section to quantify how closely the reduced model reproduces the original expansion signals. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected in voltage-aware aggregation claims

full rationale

The paper proposes a voltage-aware partitioning and aggregation method, then validates it by solving an expansion problem on a European network using the external PyPSA solver and directly comparing results to the full grid model. The 70% transformer cost preservation figure is presented as an empirical outcome of this case-study comparison rather than a quantity derived from fitted parameters, self-definitions, or load-bearing self-citations. No equations or steps in the provided abstract reduce the central result to its own inputs by construction; the methodology is tested against an independent benchmark (the unaggregated model), satisfying the criteria for a self-contained, non-circular derivation.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only view shows no explicit free parameters, axioms, or invented entities; the contribution is a new algorithmic partitioning approach relying on standard network modeling assumptions.

pith-pipeline@v0.9.0 · 5448 in / 989 out tokens · 33892 ms · 2026-05-14T17:54:43.857934+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages · 1 internal anchor

[1]

Statistical factsheet 2024

ENTSO-E, “Statistical factsheet 2024”, 2025. [Online]. Available: https://www.entsoe.eu/data/power-stats/

work page 2024
[2]

A comparison of electricity transmission technologies: Costs and char- acteristics

“A comparison of electricity transmission technologies: Costs and char- acteristics”, The Institution of Engineering and Technology, 2025. [On- line]. Available: https://www.theiet.org/media/axwkktkb/100110238 001 - rev - j - electricity - transmission - costs - and - characteristicsfinal - full.pdf

work page arXiv 2025
[3]

Netzentwicklungsplan strom 2035, version 2021

NEP, “Netzentwicklungsplan strom 2035, version 2021”, Netzen- twicklungsplan Strom, 2021. [Online]. Available: https : / / www . netzentwicklungsplan.de/archiv/netzentwicklungsplan-2035-2021

work page 2035
[4]

Oeding and B

D. Oeding and B. R. Oswald,Elektrische Kraftwerke und Netze, 8th ed. Berlin, Heidelberg: Springer Vieweg, 1107 pp

work page
[5]

Advanced spatial and technological aggregation scheme for energy system models

S. Patil et al., “Advanced spatial and technological aggregation scheme for energy system models”,Energies, vol. 15, no. 24, p. 9517, 15, 2022

work page 2022
[6]

A review of mixed-integer linear formulations for framework-based energy system models

M. Hoffmann et al., “A review of mixed-integer linear formulations for framework-based energy system models”,Advances in Applied Energy, vol. 16, p. 100 190, 2024

work page 2024
[7]

A graph-based formulation for modeling macro-energy systems

L. G ¨oke, “A graph-based formulation for modeling macro-energy systems”,Applied Energy, vol. 301, p. 117 377, 1, 2021

work page 2021
[8]

EMPIRE: An open-source model based on multi- horizon programming for energy transition analyses

S. Backe et al., “EMPIRE: An open-source model based on multi- horizon programming for energy transition analyses”,SoftwareX, vol. 17, p. 100 877, 2022

work page 2022
[9]

ETHOS.FINE: A framework for integrated energy system assessment

T. Kl ¨utz et al., “ETHOS.FINE: A framework for integrated energy system assessment”,Journal of Open Source Software, vol. 10, no. 105, p. 6274, 20, 2025

work page 2025
[10]

Switch 2.0: A modern platform for planning high- renewable power systems

J. Johnston et al., “Switch 2.0: A modern platform for planning high- renewable power systems”,SoftwareX, vol. 10, p. 100 251, 2019

work page 2019
[11]

Bonaldo et al.,Genxproject/genx.jl: Genx, 2024

L. Bonaldo et al.,Genxproject/genx.jl: Genx, 2024. [Online]. Available: https://doi.org/10.5281/zenodo.15865702

work page doi:10.5281/zenodo.15865702 2024
[12]

LEGO: The open-source Low-carbon Expansion Generation Optimization model

S. Wogrin et al., “LEGO: The open-source Low-carbon Expansion Generation Optimization model”,SoftwareX, vol. 19, 2022

work page 2022
[13]

PyPSA: Python for power system analysis

T. Brown et al., “PyPSA: Python for power system analysis”,Journal of the Operational Research Society, vol. 6, no. 1, p. 4, 16, 2018

work page 2018
[14]

Siqueira et al.,Tulipa energy model

S. Siqueira et al.,Tulipa energy model. [Online]. Available: https:// github.com/TulipaEnergy/TulipaEnergyModel.jl

work page
[15]

A Network Aggregation Tool for the Energy System Modelling Framework Spine

I. Kouveliotis-Lysikatos et al., “A Network Aggregation Tool for the Energy System Modelling Framework Spine”, en, in2020 International Conference on Smart Energy Systems and Technologies, Istanbul, Turkey: IEEE, 2020, pp. 1–6

work page 2020
[16]

Unraveling the spatial complexity of national energy system models: A systematic review

K. Javanmardi et al., “Unraveling the spatial complexity of national energy system models: A systematic review”, en,Renewable and Sustainable Energy Reviews, vol. 213, p. 115 470, 2025

work page 2025
[17]

Spatial representation of renewable technologies in generation expansion planning models

K. Phillips et al., “Spatial representation of renewable technologies in generation expansion planning models”,Applied Energy, vol. 342, p. 121 092, 2023

work page 2023
[18]

An open-source framework for balancing computational speed and fidelity in production cost models

K. Z. Akdemir et al., “An open-source framework for balancing computational speed and fidelity in production cost models”,Environ. Res.: Energy, vol. 1, no. 1, p. 015 003, 1, 2024

work page 2024
[19]

Incorporating power transmission bottlenecks into aggregated energy system models

K.-K. Cao et al., “Incorporating power transmission bottlenecks into aggregated energy system models”,Sustainability, vol. 10, no. 6, p. 1916, 7, 2018

work page 1916
[20]

Overview of the clustering algorithms for the formation of the bidding zones

G. Chicco et al., “Overview of the clustering algorithms for the formation of the bidding zones”, in2019 54th Int. Univ. Power Eng. Conf. (UPEC), Bucharest, Romania: IEEE, 2019, pp. 1–6

work page 2019
[21]

Congestion-sensitive grid aggregation for DC optimal power flow

B. St ¨ockl et al., “Congestion-sensitive grid aggregation for DC optimal power flow”, in2025 IEEE Kiel PowerTech, 2025, pp. 1–7

work page 2025
[22]

A comparison of clustering methods for the spatial reduction of renewable electricity optimisation models of europe

M. M. Frysztacki et al., “A comparison of clustering methods for the spatial reduction of renewable electricity optimisation models of europe”,Energy Inform, vol. 5, no. 1, p. 4, 4, 2022

work page 2022
[23]

Pache et al.,E-highway 2050 WP8 : Enhanced pan-european transmission planning methodology, 30, 2015

C. Pache et al.,E-highway 2050 WP8 : Enhanced pan-european transmission planning methodology, 30, 2015. [Online]. Available: https://zenodo.org/records/8232533

work page arXiv 2050
[24]

NPAP: Network Partitioning and Aggregation Package for Python

M. Anarmo et al.,NPAP: Network partitioning and aggregation pack- age for python, 12, 2026. arXiv: 2605.12137

work page internal anchor Pith review Pith/arXiv arXiv 2026
[25]

Modelling the high-voltage grid using open data for europe and beyond

B. Xiong et al., “Modelling the high-voltage grid using open data for europe and beyond”,Scientific Data, vol. 12, no. 1, 2025

work page 2025
[26]

Xiong et al.,Prebuilt electricity network for PyPSA-Eur based on OpenStreetMap data, en, 2026

B. Xiong et al.,Prebuilt electricity network for PyPSA-Eur based on OpenStreetMap data, en, 2026. [Online]. Available: https://doi.org/10. 5281/zenodo.18619025

work page 2026
[27]

Anarmo,Network partitioning and aggregation package, Python Package index

M. Anarmo,Network partitioning and aggregation package, Python Package index. [Online]. Available: https://pypi.org/project/npap/

work page
[28]

[Online]

Joint Allocation Office,Static grid model, version September 2025 release. [Online]. Available: https://www.jao.eu/static-grid-model

work page 2025
[29]

[Online]

Nominatim,Nominatim, version 5.2.0. [Online]. Available: https : / / nominatim.org/

work page
[30]

[Online]

Overpass API,Overpass API. [Online]. Available: https : / / overpass - api.de/

work page
[31]

Unit investment costs indicators for energy infrastructure categories

“Unit investment costs indicators for energy infrastructure categories”, Energy Community Regulatory Board, 24, 2025. [Online]. Available: https://www.energy-community.org/news/Energy-Community-News/ 2025/04/24.html

work page 2025
[32]

Unit investment costs indicators for energy infrastructure categories

“Unit investment costs indicators for energy infrastructure categories”, ACER, 4, 2023. [Online]. Available: https : / / www . acer . europa . eu / electricity / infrastructure / network - development / transmission - infrastructure-reference-costs

work page 2023
[33]

St ¨ockl,Voltage-aware grid aggregation, 2026

B. St ¨ockl,Voltage-aware grid aggregation, 2026. [Online]. Available: https : / / github . com / BenjaminStoeckl / EEMV oltage - AwareGrid - Aggregation

work page 2026
[34]

H ¨orsch et al.,PyPSA-Eur: An open optimisation model of the european transmission system (dataset), 16, 2023

J. H ¨orsch et al.,PyPSA-Eur: An open optimisation model of the european transmission system (dataset), 16, 2023

work page 2023
[35]

The pareto-optimal temporal aggregation of energy system models

M. Hoffmann et al., “The pareto-optimal temporal aggregation of energy system models”,Applied Energy, vol. 315, p. 119 029, 1, 2022

work page 2022
[36]

Anarmo,Integrate NPAP: Network partitioning and aggregation package for PyPSA spatial network clustering #1568

M. Anarmo,Integrate NPAP: Network partitioning and aggregation package for PyPSA spatial network clustering #1568. [Online]. Avail- able: https://github.com/PyPSA/PyPSA/pull/1568 [37]GUROBI optimizer, version 13.0.0. [Online]. Available: https://www. gurobi.com/

work page