Project schedules for hydrogen are usually built outward from the electrolyser. Stack lead time, EPC, commissioning, first production. It is a natural way to plan, because the electrolyser is the thing being bought and the thing the sponsor controls.
The evidence from the network operators says the schedule is being set somewhere else entirely, by a queue the project joins alongside applicants that have nothing to do with hydrogen.
Germany: the rule changed because the queue broke it
Figure 1
What arrived at four German transmission operators
Grid connection applications held by the four German TSOs at the end of the third quarter of 2025.
717 applications. Types named by the operators: large battery storage, data centres, electrolysers and other large consumers.
545 applications, against a projected 2037 need of 41 to 94 GW.
717 applications in total, 545 of them for large battery storage. For context the operators give themselves: the scenario framework for the 2037/2045 network development plan expects a need for large battery storage of between 41 and 94 GW in 2037. The applications are not a forecast of what will be built — that is the operators' stated reason for changing the rule.
Source: Amprion, on behalf of the four German transmission system operators · Data as of 2026-02-11
Show the data ▸Hide the data ▾
| Measure | Applications | Capacity |
|---|---|---|
| All connection applications, end Q3 2025 | 717 | around 270 GW |
| Large battery storage | 545 | 211 GW |
| Projected battery need in 2037 (scenario framework) | — | 41–94 GW |
On 11 February 2026 the four German transmission system operators announced a Reifegradverfahren — a maturity procedure — for connection applications from storage and large consumers. The reason they give is arithmetic:[1]
Bei den vier ÜNB lagen zum Ende des dritten Quartals 2025 insgesamt 717 Netzanschlussanträge mit einer kumulierten Gesamtleistung von rund 270 Gigawatt (GW) vor, davon allein 545 Anträge für den Anschluss von Großbatteriespeichern mit 211 GW Leistung.
717 applications, around 270 GW, of which 545 applications and 211 GW are large battery storage. Against that, the operators' own scenario framework for the 2037/2045 network development plan expects a need for large battery storage of between 41 and 94 GW in 2037.[1]
Their diagnosis of the cause is unusually direct for a network operator:[1]
Das Windhundprinzip („First come, first served") führte in den vergangenen zwei Jahren zu einer unverhältnismäßigen Flut an Anträgen, die die verfügbaren Netzanschlusskapazitäten und Ressourcen in den Umspannwerken sowie den im Rahmen der Netzentwicklungsplanung ermittelten Bedarf bei Weitem übersteigen.
A disproportionate flood of applications, far exceeding both the available connection capacity and the need identified in network planning.
And here is the sentence that puts a hydrogen project inside this story rather than beside it:[1]
Anträge auf Anschluss an das Höchstspannungsnetz, z. B. für Großbatteriespeicher, Rechenzentren, Elektrolyseure und andere Großverbraucher, werden künftig anhand einer Reihe nachprüfbarer Kriterien und ihrer Realisierungswahrscheinlichkeit bewertet.
Elektrolyseure. Electrolysers, named between data centres and other large consumers, as one of the categories whose applications will now be assessed against verifiable criteria and probability of realisation.
Britain: the queue that matters tripled in seven months
Figure 2
The queue that an electrolyser actually stands in
Total contracted offers in the Great Britain demand connections queue, split between transmission and distribution.
- Transmission
- Distribution
17 GW transmission, 24 GW distribution.
97 GW transmission, 29 GW distribution. Transmission grew by a factor of nearly six in seven months.
An electrolyser connects as demand, not as generation, so this is the relevant queue — and it roughly tripled in seven months, with almost all of the growth on the transmission side. The document does not mention hydrogen or electrolysers anywhere; the queue is being filled by data centres and other large loads, and hydrogen projects are competing inside it without being counted separately.
Source: Ofgem, Demand connections update, citing Energy Networks Association data · Data as of 2025-11-06
Show the data ▸Hide the data ▾
| Date | Transmission | Distribution | Total |
|---|---|---|---|
| November 2024 | 17 GW | 24 GW | 41 GW |
| June 2025 | 97 GW | 29 GW | 125 GW |
There is a common framing error here worth naming, because it is easy to make and it points at the wrong number. Britain's famous connection queue — the hundreds of gigawatts — is a generation queue. An electrolyser is not generation. It is a large load, and it joins the demand queue.
That queue is doing something dramatic. Ofgem, citing Energy Networks Association data:[2]
Since November 2024, there has been a significant increase in total contracted offers in the demand queue. Recent Energy Networks Association data illustrates the scale of this increase: • November 2024: 41 GW (17 GW Transmission, 24 GW Distribution) • June 2025: 125 GW (97 GW Transmission, 29 GW Distribution)
Three times as much in seven months, and the transmission component up by nearly a factor of six.
The document does not mention hydrogen or electrolysers anywhere. That absence is the point rather than a gap: a hydrogen project competes for transmission-connected demand capacity against data centres and other industrial loads that are not in the hydrogen conversation at all, are not tracked in any hydrogen pipeline database, and in several cases have deeper balance sheets and shorter build times.
What the timeline is, according to the institution proposing to shorten it
Figure 3
How long this takes, according to the institution proposing to fix it
Timelines stated by the European Commission in the explanatory memorandum of its December 2025 proposal to accelerate permit-granting.
Of which permitting accounts for more than half of the time needed.
Depending on the Member State and the technology involved.
Depending on the technology.
The most useful sentence is the qualifier on the first bar: of the roughly ten years an electricity transmission grid takes to complete, permitting accounts for more than half. These are the Commission's own figures, published in support of new legislation — which is itself the clearest available evidence that the 2023 directive did not solve this.
Source: European Commission, COM(2025) 1007 final · Data as of 2025-12-10
Show the data ▸Hide the data ▾
| Asset | Time stated | Qualifier |
|---|---|---|
| Electricity transmission grid | around 10 years to completion | permitting is more than half of it |
| Renewable energy project permitting | up to 9 years | by Member State and technology |
| Storage installation | 1 to 7 years | by technology |
In December 2025 the European Commission proposed a new directive to accelerate permit-granting for infrastructure projects. The explanatory memorandum states the problem in its own numbers:[3]
Electricity transmission grids take around 10 years for completion, of which permitting accounts for more than half of the time needed.
Similarly, permitting of renewable energy projects can still take up to nine years depending on the Member State and the technology involved.
Storage installations can take from one to seven years depending on the technology.
Two things are worth extracting. The first is the qualifier: of ten years for a transmission grid, more than five are permitting. That is not engineering time or supply chain time; it is administrative time, and it cannot be compressed by ordering earlier.
The second is what the existence of this proposal implies. RED III already set permitting deadlines in 2023. The Commission proposing a further acceleration directive in December 2025 is the clearest available evidence, from the body best placed to know, that the 2023 deadlines did not deliver the outcome.
The law already knows grid connection is part of permitting
RED III caps the permit-granting procedure at twelve months inside renewables acceleration areas, and at two years for offshore renewable projects.[4]
The provision that matters more for a hydrogen project is the definition of what the procedure covers:[4]
The permit-granting procedure shall cover all relevant administrative permits to build, repower and operate renewable energy plants […] as well as assets necessary for the connection of such plants, heat pumps and storage to the grid, and to integrate renewable energy into heating and cooling networks, including grid-connection permits…
Grid-connection permits are inside the capped procedure. The European legislator did not treat connection as a separate downstream administrative matter; it treated it as part of the same bottleneck. Anyone building a schedule that puts "permitting" and "grid connection" in two different phases, running in sequence, is modelling something the directive itself does not describe.
The one named case, and the one this page cannot show
The clearest documented instance of a grid constraint binding on a real hydrogen project comes from the network operator, not the developer. TenneT, on Shell's 200 MW Holland Hydrogen 1 at the Maasvlakte, quoting its own chief operating officer Maarten Abbenhuis:[5]
TenneT heeft op de bestaande hoogspanningsstations onvoldoende ruimte om de waterstoffabriek aan te sluiten
TenneT does not have enough room at the existing high-voltage substations to connect the hydrogen plant. The resolution was a temporary and limited connection at the Maasvlakte 380 kV substation, pending the Amaliahaven substation, which TenneT expected to be ready at the end of 2026.[5]
Read honestly, that is a constraint that bound and was then mitigated by a contractual arrangement rather than by steel. Europe's largest such project needed a workaround to get connected at all.
Spain: capacity granted is not capacity in service
One more national observation, useful because it separates two things that are often conflated. Red Eléctrica, publishing demand access capacities:[6]
since 2022, a total of 11.8 GW of access capacity for demand has already been granted on the transmission grid alone, though none of it is yet in service.
Eleven point eight gigawatts granted. None in service. The same release records permits granted for 129 GW of wind and solar, 16 GW of storage and 19 GW of demand facilities, and notes that the transmission grid still has capacity for new demand at 25% of its nodes.[6]
An access permit is a right to connect at a point, not a connection. A model that treats "we have access capacity" as "we can draw power on this date" has skipped the step where the physical works, the substation bay and the commissioning sequence happen — and that step, on the Commission's own figures, is measured in years.
What to do with this before your next schedule
- Put the connection application at the front of the critical path, not the electrolyser order. The queue cannot be expedited by paying a premium, and in Germany it can no longer be held speculatively either.
- Find out which queue you are in. Generation queue figures do not describe your project. An electrolyser is a large load, and the demand queue is a different, faster-growing and much less publicised number.
- Know who you are queuing against. Data centres and battery storage are the volume in these queues. They are not in any hydrogen database, they are not slowed by RFNBO compliance, and several of them can sign a connection agreement faster than a hydrogen project can reach FID.
- Prepare for a maturity test, not a ticket. Site control, planning consent, offtake and financing evidence are becoming the price of a queue position. A project that has them should be advancing them explicitly for this purpose.
- Do not model an access permit as a connection date. Spain has 11.8 GW of granted demand access with none in service. The gap between the two is the number that belongs in the schedule.
- Carry the permitting period as its own line, with its own contingency. More than five of the ten years a transmission grid takes are administrative. Whatever share of that lands on the project, it is not a line that responds to procurement.
The uncomfortable summary is that the part of a hydrogen project a sponsor controls best — buying and installing the plant — is not the part that decides when it produces. That is set by a queue, filled largely by other industries, and increasingly rationed by a test of how real each applicant is.