The queue and its projects
Of the 88.6 GW of batteries built and due to connect by 2035, just over 10 GW is built. So the 52.8 GW that would switch the commitment fee off has to come from the 78.6 GW still to connect: 67.2% of it.
NESO’s registers give each project’s connection date and its planning status, such as Awaiting Consents or Consents Approved, but not its battery revenues or whether it has accepted its offer. So the scenario takes 54 kinds of project and weights them by an assumed mix for the queue: connection dates as in the registers, battery revenues split equally across three levels and key consents across three stages, and half the offers accepted. Run with the consents split by NESO’s planning statuses instead, the engine finds nearly the same answers: 3.9 to 10.5 GW, against 4.1 to 10.4 GW.
For each kind, the engine finds what its developer would do: accept the offer or decline it, keep or give up MW at each review, build or drop the project at the build decision, or sell it to a fund. A project leaves the queue when its developer declines the offer, gives up its MW or drops the project.
Holding a single Gate 2 offer? Our research on a Gate 2 offer and the commitment fee works through what it’s worth.
- The queue88.6 GW of batteries built or with Gate 2 offers to connect by 2035, against a target of 28.7 GW for 2035 (NESO’s final modification report for CMP470). Just over 10 GW is built, and 78.6 GW is still to connect.
- The projects54 kinds, each the 200 MW, four-hour battery at 400 kV from our research on a Gate 2 offer and the commitment fee, in every combination of:
- connection date: October 2028, 2030 or 2033, with the build decision 18 months before, in April 2027, 2029 or 2032
- battery revenues: on the edge, middling or strong
- key consents in place: already, mid-way to the build decision, or at it
- the Gate 2 offer: in hand or accepted
- Key consentsIn the engine, key consents work through the security: once they’re in place, a project’s own security falls from 42% to 10% of its cancellation charge, so the commitment fee, a minimum on that security, adds more.
- The queue’s mixOctober 2028 stands for the 34% of the capacity due to connect by 2028, 2030 for the 24% due in 2029 to 2031, and 2033 for the 42% due from 2032, as in NESO’s registers. Battery revenues are split equally across their three levels, key consents across their three stages, and half the offers are accepted.
- Battery revenuesTrading revenues as a multiple of a published revenue index for two-hour batteries, plus Capacity Market payments from the year after the battery connects. On the edge, they’re just above the level at which the developer would walk away with no commitment fee: 1.65 times the index for 2028 and 2030, and 1.70 for 2033. Middling is 0.15 above that, and strong 0.45 above.
- The commitment fee£3,000 per MW from October 2027, £5,000 from April 2028, then £5,000 more at each six-monthly review to £25,000. It sets a minimum on the security a project puts up, which it gets back when the battery connects, and on the cancellation charge it pays if it pulls out.
- Ofgem’s decisionAn 80% chance that Ofgem approves the commitment fee by April 2027: our assumption, as Ofgem has said it’s minded to approve it. Once approved, the commitment fee climbs in its six-monthly steps and stays on until the battery connects, as it wouldn’t switch off before 2036. We call this the commitment fee “as it may come”, and with approval certain, “as proposed”.
- The developerIts cost of money (WACC) is 10% a year in real terms (above inflation), within the range published for merchant battery projects. It puts a weight of 0.25 on its worst 1 in 20 outcomes: a quarter of a project’s worth to it is set by them, and the rest by the average.
- The fundA buyer with a cost of money of 8% a year in real terms, about what the listed battery funds use for their operating portfolios, and the developer’s weight on the downside. A cautious fund puts a weight of 0.5 on its worst 1 in 20 outcomes.
- SellingThe fund pays the developer and takes over the offer, its security and the commitment fee. Selling costs the developer 1.5% of the price plus £50,000.
How the answers are worked out
QueueQuant’s engine answers the three questions in five steps. Steps 1 and 2 answer the first: what the commitment fee costs each kind of project, and which projects it pushes out. Steps 3 and 4 answer the second: how much of the queue leaves, and whether that’s enough for the commitment fee to switch itself off. Step 5 answers the third: what a fund would bid, and who ends up owning the queue.
What the commitment fee costs a project
Pick a kind of project: when it’s due to connect, its battery revenues, when its key consents are in place, and whether its offer is in hand or accepted. The engine values it across 8,000 futures (200 ways its connection could turn out, each with 40 paths for the battery market) with the commitment fee as it may come, as proposed and with none, and finds what its developer would do.
For a project that stays, the commitment fee costs the money tied up in the extra security, and the bank’s charge on it, until the battery connects. For a project that leaves, it costs the project’s whole worth to its developer.
A project connecting in October 2030, with trading revenues of 1.80 times the index (middling), its key consents in place from April 2028 and its Gate 2 offer in hand.
Stays Accept all 200 MW.
An accepted project has already put up security, which it gets back when the battery connects, and its worth includes that money coming back. So its worth depends partly on how much it has put up so far, which is why it reads higher than the same project with its offer in hand.
- It would leave at a flat commitment fee of £73,190 per MW or more, from October 2027 until the battery connects.
- If a fund at 8% can buy it, it would bid £97,740 per MW. The developer would take anything above £28,020 per MW, its worth plus the costs of selling, so it would sell.
What if the developer’s cost of money is different?
At 10% a year, the scenario’s rate, the project is worth £27,350 per MW to its developer with the commitment fee as it may come, and £30,740 without it: the commitment fee costs it £3,390 per MW. From 10% to 11%, its worth with no commitment fee falls by about £20,000 per MW, more than five times what the commitment fee costs it at 10%.
At 9%, 11% and 12%, the engine values the project with its developer deciding once, at the start, rather than at every review. At 10%, the two ways agree to within £460 per MW for every kind of project.
What to notice. For a project that stays, the commitment fee as it may come costs £950 to £1,670 per MW if it connects in October 2028, £2,650 to £4,320 for October 2030 and £6,150 to £7,750 for October 2033: the later the connection, the longer the commitment fee holds the security up, and the higher it climbs. Most projects are worth far more than that, so they stay. With the commitment fee as it may come, 7 of the 54 kinds leave, all of them on the edge of walking away and connecting in 2030 or 2033. The developer’s cost of money moves a project’s worth far more than the commitment fee does.
The engine values the project with a flat commitment fee at levels from £0 to £250,000 per MW, from October 2027 until the battery connects, and finds the level at which its developer would rather leave than keep it. Where the answer flips between two levels it tried, it tries the level halfway between, again and again, until it pins the level down.
Who the commitment fee pushes out
Every project has a level of battery revenues below which its developer would walk away even with no commitment fee. The commitment fee raises that level a little, by what the extra security costs. Only a project whose revenues fall in the narrow band between the two levels leaves because of the commitment fee: one below leaves anyway, and one above stays.
The chart shows the band for each connection date, with the scenario’s three revenue levels marked. The band is wider for later connections, because the commitment fee holds the security up for longer, and climbs higher, but it’s never more than 3.73% of revenues.
Where the band is thinner than a line, it’s drawn as a line.
What to notice. The scenario puts 5% to 13% of the queue inside the band, the 4.1 to 10.4 GW the proposal pushes out, depending on how battery revenues are spread and whether Ofgem approves the commitment fee. Switching it off would need 67.2%: two-thirds of the queue within a few per cent of walking away.
How much of the queue leaves
This step adds up the 54 kinds of project across the queue, each weighted by its share of it. Each dot is 100 MW of the 78.6 GW still to connect, set out by connection date and battery revenues. Choose the commitment fee, whether funds can buy and how many offers are accepted.
Where funds can’t buy, the commitment fee as it may come clears 9.5 GW, 12.1% of the queue and 18% of the 52.8 GW that would switch it off.
Each connection date’s share of the queue comes from NESO’s registers. Supported by National Energy SO Open Data.
What to notice. Where funds can’t buy, the commitment fee as it may come clears 9.5 GW, 12.1% of the queue, 96% of it from projects on the edge of walking away. Where a fund at 8% can buy, developers sell to it rather than leave, and the commitment fee clears 0.1 GW. A cautious fund bids less than the projects are worth to their developers, so it buys none, and the queue clears as it does where funds can’t buy.
Each kind of project carries its share of the 78.6 GW: its connection date’s share from NESO’s registers, split evenly across battery revenues and key consents, with the offers split as the buttons set. For each kind, the engine works out how much of its capacity stays with the developer, is sold to the fund and kept, or leaves, on average across its 8,000 futures, with the commitment fee and without it.
Capacity can leave at once, when the developer declines the offer or gives up its MW, or later, when the project is dropped at the build decision. What leaves because of the commitment fee is what leaves with it, less what leaves without it.
Will the commitment fee switch itself off?
The commitment fee switches off once the battery capacity built and in the queue is less than 25% above the 2035 target. Today it’s 208.8% above. This step follows that figure year by year as projects leave the queue, at the rates projects of every kind have left NESO’s transmission register since 2021, as QueueWatch measures them. It assumes no new projects join, and what’s already built stays.
Applies 2021’s rate in 2026, 2022’s in 2027 and so on, starting again after 2025.
Step 3 finds how much the commitment fee pushes out of the queue. This takes it out straight away, in 2026, to see how much sooner the commitment fee would switch off.
At 5.5% a year, the average since 2021, the battery capacity falls to less than 25% above the target in 2046, and the commitment fee switches off then.
Exit rates from NESO’s TEC register, as QueueWatch reads it. Supported by National Energy SO Open Data.
What to notice. At the rates projects have left the queue since 2021, the commitment fee would switch off between 2038, at 2023’s 9.6% a year, and 2085, at 2025’s 1.9%, and in 2046 at the average. What the commitment fee pushes out as proposed, 10.4 GW, brings that forward by two to eight years. So once Ofgem approves it, the commitment fee would climb to £25,000 per MW and stay on until these batteries have connected. That’s why, in this scenario, the only uncertainty about the fee is whether Ofgem approves it.
QueueWatch follows every project in NESO’s TEC register back to January 2021. Each year’s rate is the share of the capacity still to connect at the start of the year that left during it, across every technology, counting a project that came back under a new number as staying. The rates were 8.1% in 2021, 3.1% in 2022, 9.6% in 2023, 4.8% in 2024 and 1.9% in 2025. Delays and exits on QueueWatch shows them by year and technology.
Who ends up owning the queue?
Funds often have a lower cost of money (WACC) than developers, 8% a year here against 10%, so a fund values a project’s future battery revenues more and can bid more than the project is worth to its developer. Where it does, the developer sells rather than leave, and the capacity stays in the queue with a new owner.
The chart shows, for each connection date and level of battery revenues, what the project is worth to its developer at 10% and what a fund would bid. Choose the fund’s cost of money, the commitment fee, the key consents and the Gate 2 offer, and click a row for its details.
A project connecting in October 2033, on the edge of walking away, with its key consents in place from October 2029 and its Gate 2 offer in hand. Its developer would decline the Gate 2 offer: the project is worth nothing to it with the commitment fee as it may come. A fund at 8% would bid £40,030 per MW. The developer would take anything above £250 per MW, its worth plus the costs of selling, so it would sell rather than leave, and the fund would keep 99% of the MW.
At 7% and 9%, the engine works out the fund’s bids with it deciding once, at the start, rather than at every review. At 8%, the two ways agree to within £430 per MW.
Does the commitment fee cost a fund more or less than the developer?
If the commitment fee cost funds much less than developers, it would push projects towards funds by itself. It doesn’t: on every project the developer would keep, it costs a fund at 8% within £530 per MW of what it costs the developer. What moves projects to funds is their lower cost of money.
What to notice. A fund at 8% bids £39,850 to £254,960 per MW, at least 1.7 times what a project is worth to its developer at 10%. So developers sell rather than leave, and the fund keeps 92% to 100% of the capacity, as the line under the chart shows for each row you click. The commitment fee as it may come costs the fund £150 to £480 per MW more than the developer on projects on the edge and middling ones connecting in 2028 or 2030, and £100 to £530 less on strong ones and middling ones connecting in 2033. Those differences are tiny next to the bids, so they don’t change who owns the queue.
A fund that discounts a project’s future battery revenues at 8% a year, rather than 10%, values the same revenues more today. The listed battery funds use about 8% in real terms for their operating portfolios. A fund buying a project before it’s built may want more: at 9%, it values a project as a developer at 9% would. A cautious fund, with a weight of 0.5 on the worst 1 in 20 outcomes, bids less than the projects are worth to their developers, so the developers keep them.
The proposal’s commitment fee pushes out less than a fifth of what would switch it off
So once Ofgem approves it, it stays on until 2036 at the earliest, after these batteries have connected.
With the commitment fee, a project needs 0.26% to 3.73% more battery revenue to be worth keeping, so only projects that close to walking away leave because of it. The rest of the queue stays, or would leave anyway.
Where funds can’t buy, no commitment fee held at one level from October 2027, even £250,000 per MW, ten times the proposal’s top, pushes out the 52.8 GW. Above about £160,000 per MW, raising it further pushes out nothing more: the projects still in the queue would stay even at £250,000.
A fund at 8% bids more than every kind of project is worth to its developer, so developers sell instead of leaving, and the proposal’s commitment fee pushes out 0.1 GW.
Of the 18 kinds of project on the edge, all 18 stay with the commitment fee as it may come at 9% a year, and 11 at the scenario’s 10%. At 11%, 16 of them are worth nothing to their developers even with no commitment fee.
Why the results hold, and what to bear in mind
The main results don’t depend on the scenario’s assumptions. The 52.8 GW that would switch the commitment fee off comes from NESO’s own figures. However battery revenues are spread across the queue, the commitment fee only pushes out projects within a few per cent of walking away. To push out the 52.8 GW that would switch it off, two-thirds of the queue still to connect would have to be that close to walking away. When it would switch off depends only on NESO’s figures and the rates projects have left the queue since 2021. The choices below change the exact gigawatts, not these conclusions.
- The 54 kinds of project share one design, the 200 MW, four-hour battery at 400 kV from our research on a Gate 2 offer, in one charging zone, so they differ only in connection date, battery revenues, key consents and Gate 2 offer. A real project’s own site, costs and charges will change its answer, which is why we value each client’s project on its own figures.
- The registers don’t show battery revenues or which Gate 2 offers are accepted, so we’ve set these in the scenario, and Step 2 shows the main result holds whatever the spread.
- Battery revenues are held at today’s levels, though with three times the target’s capacity in the queue they would fall as more batteries connect, bringing more projects close to walking away or past it. Step 2’s slider lets you test this: the commitment fee would push out enough to switch itself off only if 67.2% of the queue still to connect were within a few per cent of walking away.
- The fund’s 8% is about the rate the listed battery funds use for their operating portfolios. A fund buying a project before it’s built may want more, but even at 9% it outbids the developer on all 54 kinds of project.
- Ofgem hasn’t decided on CMP470 yet. The 80% chance that it approves the commitment fee is our assumption; the commitment fee’s levels, thresholds and October 2027 start are the proposal’s.
- Step 4 assumes no new projects join the queue. New applications would keep the commitment fee on for longer.
What we’ll work out for you
Send us your projects or portfolio and we’ll work out the same answers for them, using your own figures and a market forecast, with the commitment fee as it may come, as proposed and without it.
For developers
For each of your projects: whether to keep it, give up MW, sell or leave, with the commitment fee and without it, and the lowest bid worth accepting.
For funds
Which projects to buy, the most to bid for each, and what the commitment fee takes off it.
For policy and network teams
What a commitment fee, or another rule, would push out of the queue, and when it would switch off, under your own assumptions about the queue’s mix, battery revenues and cost of money.
We will give you a report where every assumption is listed with its source and the switch points are given in plain words. One re-run after your comments is included. The report is decision support. It isn’t investment advice or a formal valuation.
How to cite this page
You’re welcome to quote our figures, charts and findings with this credit:
QueueQuant Research, “Will CMP470’s commitment fee clear the battery queue?”, queuequant.com/research/commitment-fee-battery-queue/ (October 2026).