Contracting Strategies for Electrolyzers to Secure Grid Connection: The Dutch Case

Arjan van Voorden; Han Slootweg; Jalal Kazempour; Johan Morren; Thomas Swarts; Wouter van den Akker

REVIEW 3 major objections 5 minor 13 references

Contracting Strategies for Electrolyzers to Secure Grid Connection: The Dutch Case

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Dutch electrolyzer grid contracts flip at a €10/MW curtailment price, a bilevel model shows.

desk verdict Well-structured bilevel model of Dutch CTA/CRC contracts, but the central 10 €/MW threshold result is unsupported because Eq. (2k) forces the key curtailment variable to zero. read the letter →

arxiv 2502.09748 v1 pith:VD4TZ7DY submitted 2025-02-13 math.OC cs.SYeess.SY

Thomas Swarts , Jalal Kazempour , Wouter van den Akker , Johan Morren , Arjan van Voorden , Han Slootweg This is my paper · ORCID

classification math.OCcs.SYeess.SY MSC 90C1191A6590C90

keywords congestionmanagementelectrolyzernon-firmconnectionandtransportagreementcapacityrestrictioncontractbilevelprogrammingleader-followergamegridDutchelectricitymarket

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks how an electrolyzer project in the Netherlands should contract for grid capacity when the grid is congested, and how the network operator should respond. It tries to establish that the optimal contract is not a single best type but depends on the price paid for curtailment: below roughly €10/MW, voluntary curtailment contracts raise the electrolyzer's profit, while above that price they lower it. It also claims that who leads the negotiation matters, and that a naive optimization that ignores the other party's response overstates profits on both sides. A sympathetic reader would care because the answer bears directly on whether stalled green hydrogen projects can get bankable grid connections under the new Dutch rules.

What carries the argument

The central mechanism is a pair of leader-follower games written as bilevel optimization problems: in the first game the electrolyzer owner leads and the network operator follows, and in the second game the roles are reversed. Each lower-level problem is reformulated using optimality conditions, with complementarity handled through special-ordered-set constraints and strong duality added to preserve optimality, yielding mixed-integer linear programs. The network operator's decision variables—connection capacity, NFA85 and NFA curtailments, and CRC/CRC+ activations—are what transmit the leader's choice into the follower's profit, so comparing the two hierarchies isolates who captures the surplus from congestion.

What would settle it

Re-run the numerical study with Eq. (2k) replaced by a budget inequality such as $\sum_{t\in T}\lambda^{crc+}s_t^+ \leq \theta B_{CM}$ for $\theta>0$, and inspect the electrolyzer profit curve around €10/MW; if the crossover and the profit reversal disappear or change sign, the paper's central threshold is an artifact of the zero-budget setup. Alternatively, compare realized contract choices of Dutch electrolyzer projects with CRC prices above and below €10/MW.

Watch

Extended reading notes

Core claim

The paper claims that the optimal grid contract for a Dutch electrolyzer is not fixed but flips at a curtailment-compensation price of about €10/MW. Below that price, the network operator's voluntary congestion management through capacity restriction contracts raises the electrolyzer's profit; above it, the operator curtails less and the electrolyzer is pushed into higher-tariff contracts, so its profit falls. The same regime change decides which player benefits from leading: above €10/MW, the network operator earns more when it reacts to the electrolyzer owner's connection choices than when it sets the connection capacity first. The paper also claims that dropping the game-theoretic structure, by optimizing only one side's objective over the other side's feasible region, overestimates profits for both parties.

Load-bearing premise

The reported sensitivity to the extra curtailment price assumes the operator has a budget to pay those activations, but the model as written sets that budget term to zero, so the threshold results depend on an uncorrected equation.

Editorial extensions

If this is right

At CRC+ prices below €10/MW, voluntary congestion management can make an electrolyzer project more profitable and more likely to secure a grid connection.
At CRC+ prices above €10/MW, the same mechanism lowers electrolyzer profit, so high curtailment compensation can backfire for the load.
The network operator earns more by responding to the electrolyzer's contract choices than by leading them when CRC+ prices exceed €10/MW, so leadership positions change the value split.
Ignoring the other party's optimization, as the high-point relaxations do, overstates profits for both sides, so coordinated game-theoretic treatment is needed to evaluate contracts.
Lower hydrogen prices push the electrolyzer toward low-tariff NFAs rather than NFA85, making projects viable only with the tariff discount.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the €10/MW crossover is robust, regulators could use the CRC+ price as a policy lever: set it below the threshold to improve electrolyzer bankability, or above it to curb speculative congestion revenue, though the paper does not test this.
The alternating-leader setup suggests the contract outcome depends on bargaining power; an auction or Nash bargaining model would be needed to say which hierarchy actually prevails in negotiations.
The same two-level contracting logic could be applied to other curtailable industrial loads, where the threshold price would depend on their tariffs and grid residual capacity.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Two things to know. First, the paper does something genuinely useful: it models the interplay between non-firm CTAs and CRC contracts as two alternating-leader bilevel games, grounded in the actual Dutch grid code. The comparison between EL-NO, NO-EL, and the high-point relaxations makes a plausible qualitative point—ignoring the other party's optimization overestimates profits for both sides. That framing is worth keeping. Second, the headline quantitative result is not supported by the model as written. Equation (2k) sets λ^{crc+} s_t^+ = θ B_CM for every hour, and the paper sets θ = 0 in the main case. With λ^{crc+} > 0, that forces s_t^+ = 0 in all hours. So λ^{crc+} cannot affect the objective, the constraints, or any solution. The sensitivity sweeps in Figures 2–4 and the 'below/above 10 €/MW' conclusion are therefore unreproducible from the stated formulation. This is not a modeling disagreement; it is an internal inconsistency that removes the paper's main contribution. The fix might be as simple as changing (2k) to an annual budget inequality, but then the results depend entirely on that correction, and the threshold claim has no current support.

The secondary issues are less severe but real. The relaxation of the binary variable b_t in Game I is justified only by tractability, not by any argument that it preserves the NFA85 activation logic. And no data or code are provided, so the case-study results are hard to verify independently. The citations look fair; the use of a co-author's earlier paper is minor.

Who is this paper for? Researchers working on congestion management contracts, especially in the Dutch context, and anyone modeling Stackelberg games between grid operators and flexible consumers. The framework and the qualitative insight about coordination are worth serious attention. But the current version cannot be accepted as is. I would send it to peer review—the problem is important and the flaw is fixable—but the reviewers must be pointed directly at Eq. (2k) and asked to re-run the case study with a corrected budget constraint. Without that correction, the central claim should not be cited.

Referee Report

3 major / 5 minor

Summary. The paper models the contracting decision of an electrolyzer owner facing Dutch FA, NFA85, NFA, CRC, and CRC+ contracts as a Stackelberg game with two possible leadership orders. It formulates both games as bilevel programs, reformulates them via KKT conditions with SOS1/strong-duality treatment, and conducts a case study inspired by the GROWH project. The headline result is that voluntary congestion management (CRC+) increases electrolyzer profitability for CRC+ prices below €10/MW and decreases it above that threshold, and that the network operator prefers a reactive role at high CRC+ prices. The paper also compares the bilevel solutions with high-point relaxations to argue that ignoring the other party's optimization overestimates profits for both players.

Significance. If the model were correct, the paper would provide useful policy-relevant insight into the interaction between non-firm connection agreements and congestion-management contracts in the Netherlands, a topic with little existing quantitative work. The regulatory detail is translated carefully into a mathematical framework, and the use of KKT-based reformulations is methodologically appropriate. The model is not circular: parameters are taken from stated regulatory sources and are not fit to reproduce a target result. However, the principal quantitative claim is currently not derivable from the model as written because Eq. (2k) with θ=0 eliminates the CRC+ variable entirely, and the Game I relaxation of the binary variable is not justified. The central contribution is therefore unestablished.

major comments (3)

[§III-B, Eq. (2k); §IV-A] With θ=0 set in Section IV, Eq. (2k), λ^{crc+} s_t^+ = θ B_CM, forces s_t^+ = 0 for every hour t whenever λ^{crc+} > 0. It follows that the CRC+ variables are identically zero and λ^{crc+} does not enter the effective objective or constraints of either bilevel program. The sensitivity sweeps in Figures 2 and 3, the comparison of EL-NO and NO-EL for CRC+ prices above €10/MW, and the abstract's regime-shift claim therefore cannot be reproduced from the model as documented. If the authors intended a budget constraint such as λ^{crc+} Σ_t s_t^+ ≤ θ B_CM, that correction is not stated and the numerical results would depend entirely on its specific form. This is a load-bearing internal inconsistency, not a modeling assumption.
[§III-D(i) (solution of Game I)] The binary variable b_t in the lower-level problem of Game I is relaxed to [0,1] with the statement that this makes the lower-level problem linear and convex. No argument is given that the relaxation is exact: the original lower level is a mixed-integer linear program, and curtailments under NFA85 are tied to the number of activation days through constraints (2f)-(2h). Replacing b_t with a continuous variable enlarges the feasible set, so the KKT-reformulated solution may not be an equilibrium of the original game. This does not by itself invalidate the paper's central mechanism, but it means the EL-NO results in Figures 2-4, including the profit comparisons, rest on an unproven relaxation.
[§IV-A, Figures 2-3] The narrative in Section IV-A repeatedly attributes the low-price behavior to 'curtailed via CRC+ at near-zero costs' and the high-price behavior to reduced CRC+ activations. These statements presuppose that s_t^+ can be positive, which is impossible under the stated parameterization θ=0. The interpretation of the threshold at €10/MW is therefore not supported by the model as written, independent of whether the first major comment's correction is adopted.

minor comments (5)

[§III (end)] The sentence 'Note that all variable sets Θ(.) in both bilevel programs (3) and (5) belong to R+' is inaccurate because b_t is defined as binary in (2h); please clarify the exception or the intended relaxation.
[§III-B, Eq. (2a)] The lower-level objective uses λ^{crc}_t and λ^{crc+}_t with a time index, while the parameters in Table II and Eq. (1a) are written without a time index; please make the notation consistent.
[Figures 2 and 4] The axis labels in the provided manuscript are garbled (for example 'C¯ntraﬁt ﬁapaﬁities' and 'eur¯s'); please ensure the final PDF displays the labels correctly.
[Throughout] The paper alternates between 'C10/MW' and '€/MW' for the same quantity; please use a single currency symbol consistently.
[Eq. (1c) and (4e)] The upper-level constraints include lower-level curtailment variables such as s_t and r_{2,t}; as written this is a coupling through the follower's response, which is acceptable in a bilevel formulation, but the text should state this explicitly to avoid confusion.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity; the bilevel model is constructed from regulatory inputs, with only a minor non-load-bearing self-citation.

full rationale

The paper's derivation chain is a bilevel optimization model whose inputs are regulatory parameters (tariffs, budgets, capacities, prices) and stated player objectives, and whose outputs — including the C10/MW CRC price threshold — are obtained by solving the model across a range of lambda_CRC+ values. No parameter is fitted to reproduce a target result, and no result is defined in terms of another result by construction. The only self-citation is [5] (shared co-author J. Morren), cited in the introduction as background on budget allocation for redispatch and CRCs; it is not invoked to justify the model, the reformulation, or the threshold claim, so it is not load-bearing. References [10] and [11] are external methodological and data sources. A serious internal inconsistency exists but is not circularity: Eq. (2k) states lambda_crc+ s_t^+ = theta B_CM for every t, and Section IV sets theta = 0, which forces s_t^+ = 0 for every hour; the reported sensitivity to lambda_CRC+ and the associated C10/MW regime claim therefore cannot be reproduced from the model as documented. That is a correctness or modeling defect, not a definitional equivalence between inputs and outputs. Similarly, absence of external benchmark validation is a validation gap, not circular reasoning.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

No new physical entities are introduced. The only novel construction is the assumed new NFA contract design, which is a regulatory assumption rather than a physical entity, so it is captured under free parameters and axioms instead.

free parameters (5)

λ_CRC (mandatory CRC price) = 40 EUR/MW
Fixed compensation per MW curtailed under the mandatory CRC; chosen without a cited source, affects the trade-off between curtailment income and tariffs.
B_NFA (NFA energy budget ratio) = 1
Assumed daily energy budget of the new NFA contract, which is stated to be under development; directly bounds NFA curtailments.
B_NFA85 (NFA85 time budget) = 0.15
Fraction of hours per year that the network operator may curtail under NFA85, from the contract design.
θ (budget share for the electrolyzer) = 0 and 0.2
Proportion of the mandatory congestion-management budget spent on the electrolyzer; set to 0 in the main results, which forces s_t^+=0 under Eq. (2k).
λ_H2 (hydrogen price) = 10 and 5 EUR/kg
Assumed hydrogen selling price; key driver of project profitability and contract choice.

assumptions (4)

domain assumption The interaction is a deterministic Stackelberg game with perfect information of the other player's optimization problem.
Used throughout Section III; ignores uncertainty in prices, residual capacity, and regulatory changes.
domain assumption The proposed NFA contract with a daily energy budget B_NFA exists and is correctly parameterized.
Section II-A states this contract is 'being developed' and is assumed; Table I lists B_NFA as a model input.
standard math The network operator's lower-level problem is convex after relaxing the binary variable b_t; strong duality and KKT conditions are applicable.
Section III-D(i) relaxes b_t to [0,1] and uses strong duality [10]. The relaxation is not proven exact.
domain assumption The residual capacity time series S_t and electricity price time series λ_t^e are known and represent a Dutch congested grid.
Used in the case study; no source or generation method is given for these time series.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Contracting Strategies for Electrolyzers to Secure Grid Connection: The Dutch Case." pith.science (2026). https://pith.science/paper/VD4TZ7DY

@misc{pith2026250209748,
  author       = {Pith},
  title        = {Pith review of: Contracting Strategies for Electrolyzers to Secure Grid Connection: The Dutch Case},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VD4TZ7DY}},
  note         = {Machine review of arXiv:2502.09748}
}

read the original abstract

In response to increasing grid congestion in the Netherlands, non-firm connection and transport agreements (CTAs) and capacity restriction contracts (CRCs) have been introduced, allowing consumer curtailment in exchange for grid tariff discounts or per-MW compensations. This study examines the interaction between an electrolyzer project, facing sizing and contracting decisions, and a network operator, responsible for contract activations and determining grid connection capacity, under the new Dutch regulations. The interaction is modeled using two bilevel optimization problems with alternating leader-follower roles. Results highlight a trade-off between CRC income and non-firm CTA tariff discounts, showing that voluntary congestion management by the network operator increases electrolyzer profitability at CRC prices below 10 euro per MW but reduces it at higher prices. Furthermore, the network operator benefits more from reacting to the electrolyzer owner's CTA decisions than from leading the interaction at CRC prices above 10 euro per MW. Ignoring the other party's optimization problem overestimates profits for both the network operator and the electrolyzer owner, emphasizing the importance of coordinated decision-making.

Figures

Figures reproduced from arXiv: 2502.09748 by the authors.

**Figure 2.** Optimal grid connection and CTA capacities are shown against CRC+ prices. The profits of the electrolyzer owner (Ely) and network operator (NO) [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** Profit of the electrolyzer owner in both EL-NO and NO-EL games [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Game II (NO-EL): Contracting decisions at low hydrogen price (left) [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

[1]

capaciteitkaart

Netbeheer Nederland, “capaciteitkaart.” Available: https: //capaciteitskaart.netbeheernederland.nl/. Accessed January 25, 2025

work page 2025
[2]

The role of standalone hydro- gen areas in decentral hydrogen infrastructure development,

R. van Zoelen, N. Dooley, and D. Boer, “The role of standalone hydro- gen areas in decentral hydrogen infrastructure development,” HyDelta Technical Report, pp. 1–81, 2024. Available: https://zenodo.org/records/ 10868624. Accessed January 25, 2025

work page 2024
[3]

Het energiesysteem van de toekomst: de II3050-scenario’s,

“Het energiesysteem van de toekomst: de II3050-scenario’s,”

work page
[4]

Codebesluit congestiemanagement | ACM.nl

“Codebesluit congestiemanagement | ACM.nl.” Available: https://www. acm.nl/nl/publicaties/codebesluit-congestiemanagement. Accessed Jan- uary 25, 2025

work page 2025
[5]

The Activation of congestion service contracts for budget-constrained congestion management,

B. van der Holst, G. Verhoeven, P. H. Nguyen, J. Morren, and K. Kok, “The Activation of congestion service contracts for budget-constrained congestion management,” Electric Power Systems Research , vol. 235, pp. 110664, 2024

work page 2024
[6]

Comprehensive congestion management for distribution networks based on dynamic tariff, reconfiguration, and re-profiling product,

F. Shen, S. Huang, Q. Wu, S. Repo, Y . Xu, and J. Østergaard, “Comprehensive congestion management for distribution networks based on dynamic tariff, reconfiguration, and re-profiling product,” IEEE Transactions on Smart Grid , vol. 10, no. 5, pp. 4795–4805, 2018

work page 2018
[7]

Non-Firm grid connections for low-voltage generators: A case study in France,

L. Muller and F. Cadoux, “Non-Firm grid connections for low-voltage generators: A case study in France,” in the Proceedings of the 27th International Conference on Electricity Distribution (CIRED) , Rome, Italy, 2023

work page 2023
[8]

Optimization for non-firm capacity contracts for maximized grid utilization,

K. K. Mehmood, R. A. Arruda Moura, A. Van Der Molen, R. Tonkoski, P. Van Der Wielen, and P. H. Nguyen, “Optimization for non-firm capacity contracts for maximized grid utilization,” in the Proceedings of the 7th International Conference on Electric Power and Energy Conversion Systems, Sharjah, United Arab Emirates, 2024

work page 2024

Show all 13 references

[9]

Consultatie al- ternatieve transportrechten: use it or lose it,

Autoriteit consument en markt, “Consultatie al- ternatieve transportrechten: use it or lose it,”
[10]

Why there is no need to use a big-M in linear bilevel optimization: A computational study of two ready-to-use approaches,

T. Kleinert and M. Schmidt, “Why there is no need to use a big-M in linear bilevel optimization: A computational study of two ready-to-use approaches,” Computational Management Science , vol. 20, no. 3, pp. 1–12, 2023

2023
[11]

Current status of water electrolysis for energy storage, grid balancing and sector coupling via power-to-gas and power-to-liquids: A review,

A. Buttler and H. Spliethoff, “Current status of water electrolysis for energy storage, grid balancing and sector coupling via power-to-gas and power-to-liquids: A review,” Renewable and Sustainable Energy Reviews, vol. 82, no. 3, pp. 2440–2454, 2018

2018
[2022]

Accessed January 25, 2025

Available: https://www.acm.nl/nl/publicaties/ consultatie-alternatieve-transportrechten-en-use-it-or-lose-it. Accessed January 25, 2025

2025
[2023]

Accessed January 25, 2025

Available: https://open.overheid.nl/documenten/ ronl-7219ac2558977a6050ac4db764d2ddebb156df32/pdf. Accessed January 25, 2025

2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

capaciteitkaart

Netbeheer Nederland, “capaciteitkaart.” Available: https: //capaciteitskaart.netbeheernederland.nl/. Accessed January 25, 2025

work page 2025

[2] [2]

The role of standalone hydro- gen areas in decentral hydrogen infrastructure development,

R. van Zoelen, N. Dooley, and D. Boer, “The role of standalone hydro- gen areas in decentral hydrogen infrastructure development,” HyDelta Technical Report, pp. 1–81, 2024. Available: https://zenodo.org/records/ 10868624. Accessed January 25, 2025

work page 2024

[3] [3]

Het energiesysteem van de toekomst: de II3050-scenario’s,

“Het energiesysteem van de toekomst: de II3050-scenario’s,”

work page

[4] [4]

Codebesluit congestiemanagement | ACM.nl

“Codebesluit congestiemanagement | ACM.nl.” Available: https://www. acm.nl/nl/publicaties/codebesluit-congestiemanagement. Accessed Jan- uary 25, 2025

work page 2025

[5] [5]

The Activation of congestion service contracts for budget-constrained congestion management,

B. van der Holst, G. Verhoeven, P. H. Nguyen, J. Morren, and K. Kok, “The Activation of congestion service contracts for budget-constrained congestion management,” Electric Power Systems Research , vol. 235, pp. 110664, 2024

work page 2024

[6] [6]

Comprehensive congestion management for distribution networks based on dynamic tariff, reconfiguration, and re-profiling product,

F. Shen, S. Huang, Q. Wu, S. Repo, Y . Xu, and J. Østergaard, “Comprehensive congestion management for distribution networks based on dynamic tariff, reconfiguration, and re-profiling product,” IEEE Transactions on Smart Grid , vol. 10, no. 5, pp. 4795–4805, 2018

work page 2018

[7] [7]

Non-Firm grid connections for low-voltage generators: A case study in France,

L. Muller and F. Cadoux, “Non-Firm grid connections for low-voltage generators: A case study in France,” in the Proceedings of the 27th International Conference on Electricity Distribution (CIRED) , Rome, Italy, 2023

work page 2023

[8] [8]

Optimization for non-firm capacity contracts for maximized grid utilization,

K. K. Mehmood, R. A. Arruda Moura, A. Van Der Molen, R. Tonkoski, P. Van Der Wielen, and P. H. Nguyen, “Optimization for non-firm capacity contracts for maximized grid utilization,” in the Proceedings of the 7th International Conference on Electric Power and Energy Conversion Systems, Sharjah, United Arab Emirates, 2024

work page 2024