UK Capacity Auction Results
By Paul Homewood
Timera analyse the latest capacity auction:
Implications of a surge in UK capacity prices
The UK capacity market auction process normally takes two days of bidding rounds before a result. This year’s auction for 2026-27 capacity was over almost before it started.
The auction cleared at a record 63 £/kW/yr, around 3 times the average clearing price of previous T-4 auctions. Successful new build projects were dominated by batteries, engines, DSR, a CCGT & an interconnector. These will benefit from a substantial portion of required revenue being underwritten by multi-year fixed price capacity agreements (15 years in most cases).
Let’s take a look at what caused this surge in capacity prices and what its implications are for flexible asset investment, both in the UK and more broadly across Europe.
Auction deconstruction
It was clear before the event that this was going to be a tight auction. There was a structural deficit of existing capacity (38.48 GW) vs the government set auction demand target (43.00 GW at the clearing price). This was exacerbated by an older CCGT (0.75 GW) exiting the auction early (South Humber 1).
Chart 1: Deconstruction of 2026-27 T-4 capacity auction result
Successful new capacity was dominated by a new CCGT (EPH’s Eggborough 1.55 GW project), batteries (1.25 GW derated), the Viking Interconnector (1.04 GW), DSR (0.79 GW) and gas engines (0.51 GW). The high auction price saw 352 MW of 4-hour duration batteries successful, which have less penal derating factors than shorter duration BESS.
Much tougher battery derating factors also contributed to the capacity deficit in this auction. For example 2-hour BESS projects were derated at 23.63% (vs 39.73% in last year’s auction). Tougher derating factors saw 5.0 GW of nominal BESS capacity contribute only 1.2GW of derated capacity. BESS derating factors are likely to decline further in future as the marginal capacity contribution of batteries erodes with scaling.
The big surprise in the auction was how high offer price levels were for new build capacity. Historically there has been a strong overhang of new build gas engine & CCGT projects above 30 £/kW/yr that have acted as strong price resistance. The increase in cost structure of this thermal capacity is playing an important role in pulling up capacity prices.
5 factors driving the capacity price surge
At the simplest level, the reasons for a high clearing price were:
- A relatively aggressive government demand target
- An increase in the cost structure of new build capacity to meet that target.
Digging a level deeper reveals a set of drivers that are relevant for capacity prices & flexible asset investment across all power markets in Europe.
For example the impact of the current energy crisis is likely to see governments providing more proactive capacity payment support to reduce the risk of extreme market tightness as experienced across 2022 (note this will likely weigh on wholesale power prices). A strong policy push towards decarbonisation is also acting to drive up the cost of the marginal sources of flexible thermal capacity.
Previous auctions have been running at a cost of about a billion a year, but this one, which comes into effect in 2026/27, will cost £2.7 billion, equivalent to £100 per household.
The main reason for the jump in cost is much simpler than Timera think; it is the gradual disappearance of existing capacity which previously won contracts, notably coal and nuclear. This has brought new build on to the scene, but this needs much higher capacity support to justify the capital costs involved. In contrast, for existing capacity capacity payments are just bunce.
It is also worth noting Timera’s comments about the high risks faced by new CCGTs, including the risk of being excluded from the market after 2035, and ESG mandates. Investors will be reluctant to put money into such risky projects. Also the comments on interconnectors are interesting – much of the business case so far has been to take advantage of price spreads, buying cheap from Europe and selling high in the UK, and often the other way round. But more significant is the concern about the security of supply in a tight market; in short, the firm operating the interconnector may end up losing money to fulfil its Capacity Market obligations.
The graph below from the EMR shows how tight the market is becoming. Raising demand from 43 to 46 GW would push the price up from £60 to £75/KW/yr
https://www.emrdeliverybody.com/CM/Capacity%20Auction%20Information.aspx
More alarmingly though is the lack of new generation being brought forward. While gas and interconnectors still dominate the overall mix, there is only one new CCGT, Eggborough, rated at 1.5 GW. As existing gas plants age and shut down, this lack of new capacity will come back to haunt us.
Comments are closed.
NOT A LOT OF PEOPLE KNOW THAT 
Does anyone know how much dispatchable capacity (i.e. excluding the wind, solar and interconnectors) the Island of Great Britain has + how peak demand declined from 55-60 GW to about 45 GW in the last 10 or so year? Is it diesel generators being used for triad avoidance?
BM Reports might be your friend:
https://www.bmreports.com/bmrs/?q=foregeneration/capacityaggregated
59GW to answer your first question, if you include biomass, pumped storage and ‘other’. Batteries and pumped storage are dispatchable to deal with peaks, I suppose.
As to your second – 10 years ago we made more of our own steel, aluminium, cement and things. These days we import a lot from China.
E.g. “In 1970 the UK produced 28.3 million tonnes of steel, compared with 17.8 million in 1990, and 9.7 million in 2010.” (House of Commons library) Today it’s just over 7m tons.
“BM Reports might be your friend”
Unless you believe its claim that Britain has 5,603MW of Hydro Pumped Storage for 2023!
Up from the 1,928MW it claimed we had for 2022.
https://www.bmreports.com/bmrs/?q=foregeneration/capacityaggregated
Yet for 2021 it claimed we had 4,309MW!
https://www.bmreports.com/bmrs/?q=foregeneration/capacityaggregated
The late Prof Sir David MacKay pointed out we had 2,860MW
https://www.withouthotair.com/c26/page_191.shtml
🤔🤔🤔
Oh dear, you’ve dashed my illusions now, Joe. Since the likes of Gridwatch.templar use it, I’d assumed that site to be reliable!
I don’t believe the BM Reports include all the remaining coal (as some of it being run as some kind of strange expensive reserve – where the coal still is being burn to have it on hot standby) and I don’t buy we can run 37 GW of gas generation at the same time especially during a cold spell with high heating demand (I don’t think I have seen gas generation above 27 GW but I will look at Gridwatch templar).
Then some of the BM Reports generation categories are clearly political; they are detailed enough to add Fossil to Hard coal & gas but dump 3000 MW as other that I suspect is mainly Oil fired OCGT & diesel generators + the “Other renewable” is probably waste to energy i.e. plastic so I would put it under oil. Can anyone figure what else this could be as geothermal is the only other so-called “renewable” that would not fit into hydro, solar or biomass. My guess is someone added batteries to the Hydro Pumped Storage but hasn’t bothered to rename that category.
So a better question is how many GW of load could the GB national grid realistically continuously meet (so ignoring the storage and much of the hydro) during an extended cold spell with high gas demand for heating without wind, solar or interconnectors?
Google the “Digest of UK Energy Statistics” or just “DUKES 5.1” All will be revealed.
https://joannenova.com.au/2023/02/china-approves-two-new-coal-plants-every-week-to-ensure-stability-when-renewables-fail/
I am not even going to discuss the minutiae involved here. The entire system is a sick joke and should be abolished in its entirety immediately.
More Sturm und Drang about nothing. Batteries–in a pig’s eye. Gas generation–perish the thought. We want windmills, very expensive, very unreliable, windmills. The item of importance I did not see in the graph, was the cost of “spinning inertia via synchronous capacitors. The power derived from the windmills plays havoc with the connected grid. Once upon a time, when the grid was powered by dinosaur fueled generation, and green was but 2-5 % it was inconsequential. But scaling up gigwatts of flimsy, choppy, wind is a whole nother mater. Some one has to pay for the synchronous converters, so just jack up the price. Lest you doubt the veracity of my claim, see ABB Journal–Synchronous Capacitors to Shore up the Grid for wind-connected generation.
I keep banging on about the economics of such points – consumers pay for the entire costs of the system, not the costs of a windfarm. Politicians I suspect have been duped (as they are easily duped) by civil servants and advisors who gave shown them simple costs of an hypothetical windfarm versus CCGT – at a forecast gas price that makes CCGT expensive. But what matters interesting of our bills is how much the whole system costs, taking into account back-up generation, grid balancing and all the rest. Modelling that would be far more informative than squabbling about CfD bid numbers.
Why has the Govt demanded an increase from 38 to 43GW. Sureley the UK demand is going to go down; I can’t afford electricity, business will go bust throughout 2023, EV and ASHP sales have stalled, so who wants this extra electricity? Why would anybody invest in the UK energy business. If Sir Jim Ratcliffe can say sod it, and go to Texas, then that is a bad sign.
Government has its targets and those targets drive forecast demand. And when government targets are not met, out goes the carrot and in comes the stick. We will buy what we have been told to buy.
‘The UK capacity market auction’
Would one of you please be so kind to explain WTF this is? It’s completely alien to Americans.
Also from America (actually Washington State).
I think it means the UK government is buying little piggies in little pokes and expecting future hams and bacon.
“Would one of you please be so kind to explain WTF this is?”
Energy has two main values to (almost) all consumers. One is the usage value of the energy itself. The other is the reliability/security of supply.
The capacity market gives us the “flick of the switch” level of service from the electricity supply industry, because there is a huge amount of investment needed in fixed assets (including reserve) to provide it.
Capacity markets are not alien in the US:
https://www.pjm.com/markets-and-operations/rpm/
It would take pages and pages to explain but in a nutshell it can be summed up by the terms “Cartel” and “Price Rigging” – all you really need to know!
Try running a diesel genny in your back garden, comply with all noise, fuel delivery and storage management, air emissions controls, noise control (all to make sure your not a damned nuisance to the whole neighbourhood), make sure you add your service contact, and buy twice as many generators than you really nee for a loss of load probability around 8 hours per year.
Then you can tell us how cheap your own supply is, and that you are being ripped off for the same service at £63/kW/yr.
In response to Jordan, what on earth has running a generator in my garden got to do with the capacity market?
Ray – you suggested that payment for capacity is due to a cartel and price rigging. For that, you put yourself in the line of fire with my response.
There are commenters here who love to talk about a genny for their back garden for security of supply. That might be you, or maybe not you. So my comment was not necessarily directed at you in particular, but you put yourself up for it.
Accepting the GB capacity auction has its flaws, I would certainly not use language like cartel or price rigging. Authorities have powers to investigate and punish that type of behaviour (but that’s regulation again, and some love to complain about regulation too).
Which comes back to complaining about £63/kW/yr. This is a historic high and another sign of shortage. At least the auction is showing some kind of supply-demand relationship.
High capacity prices should encourage new entry to provide capacity. Isn’t that what’s supposed to happen?
It leaves the question: it you think a capacity price of £63/kW/yr is such a bad deal, you can always ask what is the opportunity cost of providing security of supply for yourself. That genny in your garden, as some love to talk about here.
If you cannot show your best alternative is lower cost than the auction has provided, your attack has no merit.
As Timera comment there are several reasons behind the higher prices, notably the threat that a new CCGT would have to shut down or try to make economic sense out of adding in CCS by 2035. A normal CCGT could expect a life of 40 years plus, and as a replacement for an existing station costs would be lower than these auction prices. The EMR Delivery Body ran its calculations on the basis of a CONE at £49/kW/year, so we are already £14/kW/year above that essentially because of government threats to the economics.
Hey GC it appears my views must be shot down even on this forum. Infiltration?
No Ray. All you need to do is make better comments to uphold the quality of discussion. Talk of cartel and price rigging is fodder for your opponents.
And don’t turn to GC for comfort. S/he doesn’t even know what a capacity auction is.
WOW serious trolling on here now. No Jordan I ain’t gonna exchange a single future word with someone as clearly rabid as you. Have a nice day.
Ray – take a look at yourself. The capacity auction behaves like a market should by signalling scarcity with a price rise. Ray’s answer is to shout “cartel” and “price rigging”. It’s the type of destructive comment that politicians listen to and intervene with more regulation, snuffing out any chance of proper market behaviour.
I asked for better arguments, and you respond with “troll”. I’ll take that as the best you have.
It’s important to understand that the auction is on top of volumes already procured from earlier auctions, and it is designed to leave a gap to be filled at a future T-1 auction. The recent T-1 auction secured 5.8GW at £60/kW/year to top up the total to provide 54.7GW on a derated basis for 2023/24, as against the 55GW we had for 2022/23 which resulted in capacity shortage and emergency implementation of demand side reduction. This auction, to provide capacity for 2026/27 and a 15 year capacity agreement for new capacity, comes on top of 10.3GW from previous auctions. The 53.3GW (derated basis) still leaves space for additional capacity to be procured at the T-1 stage, which might come from e.g. mothballed CCGT that dropped out of the auction.
Even though the derating factors (the proportion of nominal capacity that is assumed to be useful and thus get paid for) for the T-4 auction are much tougher they are increasingly likely to prove too optimistic. It is one thing to rely on a battery as an alternative to a diesel generator to meet a small amount of peak demand for an hour, with all the other generation needed being fully dispatchable. It becomes quite another when much bigger amounts of demand are being covered by unreliable sources. Texas ran into the problem that it had 75GW of demand and only 69GW of capacity to try to deliver that because the wind they had assumed was not there. With no capacity margin, it ended up with a cascading trip of generating capacity and blackouts. North Carolina found it it could not rely on power from TVA when the going got tough – more blackouts.
Perhaps they will get away with it if we fail to install heat pumps and buy EVs, and reduce our demand through industrial closure and high prices. There is no real headroom in the figures. Closure risk increases: already, these numbers assume only Sizewell B nuclear and no coal, but gas assets will start closing soon: just because a plant has said it will be available doesn’t guarantee that it will be. Meanwhile windfall taxes will do nothing to speed investments to reduce the risks.
Thanks for that additional info, IDAU. 👍
The situation is potentially much worse than you depict. The general switch from HT Transmission Network connected supply to DNO connected generation can result in the national grand total demand being met in theory but not in practise. For example a wind surplus connected to SP Power Networks at 25kV is useless to London as you simply cannot get it there.
All those closed Coal, Oil, Nuclear and even Gas plants were HT Grid connected, a large part of their “replacement” is not.
Wasn’t Texas’ problem that instead of the electricity system operator having an honest discussion on the 14th Feb 2021 with the public about the lack of an operating reserve (not helped by the wind being less than forecast) and predicting demand would increase due to declining temperatures that night warm people & start rolling blackouts that evening? As it was foreseeable the weather may trip a few generating plants so when this happen it tripped the poorly planned (it included essential energy infrastructure – negligence?) automatic low frequency load sheading relays knocking out electric natural gas compressors (which shockingly don’t have backup generators) interrupted the fuel to natural gas power stations which were not equipped with onsite fuel backup (distillate) forcing the electricity system operator to engage in extreme levels of load sheading (wasn’t there over 30 GW of generation at a time with forced outage?) – basically turning off circuits that didn’t have anything essential on it but in some of the distribution systems e.g. Austin 60% of the load was on circuits considered essential so they were unable to have rolling blackout. Although it would be interested to know if it was technically possible to have avoided people being without electricity for days if rolling blackout weren’t coordinating within distribution system areas.
” North Carolina found it it could not rely on power from TVA when the going got tough – more blackouts”
Wasn’t that an issue with a lack of on site fuel storage knocking out fuel to natural gas power stations too?
I agree that ERCOT made a pig’s ear out of managing the situation and should have implemented rolling blackouts much earlier, and the cascading trip made matters much worse for gas supply. But they would never have been in that position if they had woken up to the idea that wind could not be relied on, and therefore they needed dispatchable capacity to cover demand. Now they are belatedly looking at ways of implementing a sort of capacity market, discussed here:
https://news.yahoo.com/overhaul-plan-ercot-power-market-223534635.html
1.8GW new build was awarded for 23/24 delivery year and looks like Keadby has been delivered at 0.8GW the rest was gas peakers and a EFW – how much of that has it will actually be commissioned or has even been contractually committed to?
I would be very grateful for anyone who can please supply me with a link to the current electricity prices paid to each operational wind turbine project as I understand they are not necessarily those of the strike price. Also, are the strike prices quoted in the LCCC CfD register at 2012 prices? I want to try and understand the current prices being paid. Thanks.
The CFD Register pages display the current indexed prices for all wind farms that are in operation (and the original 2012 base price is shown in the detailed record for each one). The ones that were awarded CFDs in the latest allocation round (AR 4) are still being shown at just the 2012 base price. The Low Carbon Contracts Company assure me they will update with indexed values at the end of March. Indexed prices will take an inflation related jump from 1st April, which will be of the order of 10%.
Detailed daily data is available here
https://www.lowcarboncontracts.uk/data-portal/dataset/actual-cfd-generation-and-avoided-ghg-emissions
It helps to create some additional columns to help with summarising data by month and year, and to calculate the revenue basis the strike price and basis the weighted average IMRP (intermittent market reference price, which is basically day ahead hourly pricing that is widely used in practice for wind, as on a day ahead basis they have a reasonable forecast of output at that stage). The w.a. IMRP is calculated for each wind farm over its actual hourly production, so although you don’t get to see the hourly data you can cross check with the actual CFD payment by looking at the revenue at strike price less the revenue at w.a. IMRP.
Being a dab hand at using pivot tables in spreadsheets really helps to isolate the data wind farm by wind farm, and to use it to create averages. This chart is in part derived from such analysis (the ROC subsidised wind farms draw on data from several sources).
Many thanks for all your help and particularly for the excellent chart. I shall study and try to understand. I want to keep writing to my MP that wind energy is not 9 times cheaper than any fossil fuel and that there appears to me to be no plan in the NGESO FESs for any backup when the renewables are not providing sufficient energy to meet demand.
“But more significant is the concern about the security of supply in a tight market; in short, the firm operating the interconnector may end up losing money to fulfil its Capacity Market obligations.”
Interconnector infrastructure has traditionally been justified to meet relatively short-term “economy energy” sales that typically could not be used for capacity planning purposes. In this Brave New World bidding firm capacity via interconnectors is a very risky business.
Ruinables penetration is expected to grow as a percentage of system generation resources; a highly unreliable system for meeting demand. This will lead to greater shortages of generation to meet demand. Additionally, as traditional sources of generation are planned to be phased-out, there will be less available for interconnector suppliers of contracted capacity. There may not be firm generation resources or demand side management available to the interconnector capacity contractors at any non-astronomical prices, if at all.