Great Britain · long term

Published views put 2030 between £49 and £80 a megawatt hour in today’s money. On the forward curve, summer 2028 trades at £68 and the winter after it at £75.

Power has averaged £102 a megawatt hour so far in 2026, every half hour to September, against £76 across 2024 in today’s money. Vitreous has no long term price of its own yet. Until it does, this page carries every public long term view on one money basis, with the outturn and the forward curve on the same axis, so a reader can see who was right.

Outturn, monthlyForward curve, one settlement dayDESNZ reference projectionDESNZ low to high fuel priceLCP Delta, one scenario

Pounds per megawatt hour in 2026 money; the axis starts at £30 because nothing published sits below £40. Outturn is the Elexon market index price, monthly mean, June 2025 to September 2026, and the 2024 point is the hourly LCCC reference price for the calendar year. Forward prices are ICE baseload season contracts as reported by Catalyst Commercial, settled on 9 September 2026, nominal, deflated here, each drawn at the middle of its delivery period. DESNZ is a policy conditioned projection and LCP Delta a single named scenario; neither is a probability, and the span between the lowest and highest view is the span of published opinion, nothing more. DESNZ is the Energy and Emissions Projections 2024 to 2050, annex M, wholesale baseload, six scenarios of which three are drawn. LCP Delta is the accelerated build scenario read from the chart in its June 2026 outlook. Cornwall Insight’s April 2024 curve is in the table and not on the chart, because its money basis is not stated. Every value and its source is in the table.

Table view of this chart, with sources

Published view	Vintage	Basis as stated	2026	2027	2028	2030	2035	2040	2050
DESNZ EEP reference	Feb 2026	real 2024	81	68	64	63	68	83	115
DESNZ EEP low fuel prices	Feb 2026	real 2024	62	53	49	49	54	70	102
DESNZ EEP high fuel prices	Feb 2026	real 2024	112	93	84	80	82	95	112
DESNZ EEP reference, previous edition	Dec 2024	real 2023	74	67	64	49	63	76	85
LCP Delta, accelerated build	Jun 2026	real 2025, read from chart	98	72	80	73	72
Cornwall Insight benchmark curve, not drawn	Apr 2024	not stated, fiscal years, as printed	82	83	79	72

£ per MWh in 2026 money, converted with the OBR GDP deflator DESNZ publishes with its fuel price assumptions. DESNZ figures are policy conditioned projections; the LCP Delta line is one named scenario. Cornwall Insight is kept here and not drawn, because its money basis is not stated and its years are fiscal years, so it cannot be put on the same footing. Aurora Energy Research publishes only a change, minus £19/MWh (20%) from 2024 to 2028, 17 June 2024, so it cannot be drawn.

The forward curve, contract by contract

Season contract	Delivery from	Delivery to	Settled £/MWh, nominal	In 2026 money	Settlement date
Winter 2026	1 Oct 2026	31 Mar 2027	158.61	156	9 Sep 2026
Summer 2027	1 Apr 2027	30 Sep 2027	103.92	102	9 Sep 2026
Winter 2027	1 Oct 2027	31 Mar 2028	105.47	101	9 Sep 2026
Summer 2028	1 Apr 2028	30 Sep 2028	70.48	68	9 Sep 2026
Winter 2028	1 Oct 2028	31 Mar 2029	79.49	75	9 Sep 2026

ICE GB baseload season contracts as reported by Catalyst Commercial, every one from the same settlement day. A forward price is what the season traded at on that day, not a forecast of the spot price; on the chart each season sits at the middle of its delivery period. Quoted with attribution as the publisher permits; not a price feed.

Batteries

If every battery project at gate 2 connected on its contracted date, Britain would have 21.2 GW by the end of 2030. NESO’s ten year path has 21.7 GW.

The blue line is the contracted book carried forward, not a delivery forecast: some of it will slip and some projects not yet at gate 2 will build. The 0.5 GW between the two lines is the difference between a register and a pathway, and the two count the fleet on different bases, set out under the chart.

Built, then the gate 2 book carried forwardNESO Ten Year Outlook 2026Range of the four FES 2025 pathwaysGate 1 added as well

Grid scale battery power in gigawatts. History is DUKES table 5.16, year end, United Kingdom. The operating figure is the Vitreous fleet register, 7,194 MW on 27 September 2026. The gate 2 book is the connection register of 29 September 2026: each project’s contracted capacity increase added in the year of its contracted connection date, with the 380 MW dated before 2026 counted in 2026. Every series is a year end figure and sits at the end of its year. The operating figure sums the fleet register’s capacities, which mix registered and auction bases and have not been reconciled unit by unit; DUKES counts the United Kingdom and the register Great Britain; NESO figures exclude domestic and micro batteries but include distribution connected batteries that may never appear in this register. A connection date is not a commissioning date. These are the reasons the gap is a difference between datasets rather than a measured shortfall. The FES 2025 pathways run to 2050, where they span 26 to 32 GW.

Table view of this chart

Year	Built, DUKES	Gate 2 book, cumulative	NESO Ten Year Outlook	FES 2025 low	FES 2025 high
2017	0.1
2018	0.6
2019	0.9
2020	1.1
2021	1.5
2022	2.3
2023	3.8
2024	5.2			6.6	6.6
2025	7.5		7.9	8.2	12.5
2026		10.3	10.0	9.7	15.8
2027		16.2	13.1	11.2	18.9
2028		18.9	16.2	14.0	21.4
2029		21.1	19.0	15.9	22.6
2030		21.2	21.7	19.6	24.1
2031			23.6	20.0	26.0
2032			26.0	20.4	27.1
2033			27.8	20.7	28.4
2034			29.3	21.8	29.4
2035			30.8	22.1	30.4
2036			31.3	23.0	31.0

Gigawatts. NESO figures exclude domestic and micro batteries.

What a battery earns

Forecasters put a one hour battery at £60 a kilowatt this year and a two hour battery at £82. The three published lines paid £46 and £61. The rest is wholesale trading, which nobody publishes.

The curve is the blend of two forecasters, not named, that Gresham House Energy Storage Fund uses to value its 1,072 MW, printed in its 2025 annual report. It sits one step downstream of the price chart above: a battery model, how often it cycles, which services it sells and what spread it captures, lies between the two, and that model is theirs, not ours. The gap between the curve and the measured floor is the wholesale trading that no public file records, about £14 a kilowatt at one hour and £24 at two. The same owner reports 40% of its income from trading, which is the same size, and its own 2025 result, £69 a kilowatt on a fleet averaging 1.6 hours, sits on the curve. The same forecasters cut this curve by 14% to 21% at 2030 between the 2024 and 2025 vintages, and the owner says curve cuts have taken 66p a share off its value since December 2022. The chart does not state its money basis. If it is nominal at the 2.5% inflation the owner values with, the one hour line is flat in real terms and the two hour line rises about 5% to 2033 and then holds.

Published forecast, latest vintage, four durationsSame forecasters a year earlierActually paid, three published lines, no wholesale trading

Pounds per kilowatt a year, every revenue line together, as published: the chart in the annual report does not state whether the money is nominal or real, so the curve is drawn as printed and not converted, unlike every other line on this page. The measured points are the Vitreous measured index for the year to 27 September 2026: what the same panel of batteries was actually paid on the balancing mechanism, in the response and reserve auctions and by the capacity market, annualised, with wholesale trading left out because nobody publishes it. The cohort the index calls nearer one hour is set against the one hour curve and nearer two hours against the two hour curve; the cohorts’ average durations are not exactly one and two hours, so each gap is approximate. Both panels share one scale. Of the earlier vintage only the one hour and two hour lines are drawn; every duration of every vintage is in the table, with the source.

Table view of this chart, with sources

Published view	Vintage	Basis as stated	2026	2027	2028	2030	2035	2040	2049
GRID valuation blend, half an hour	Q4 2025	not stated on the chart; GRID values its fleet with RPI at 2.5%	40	44	42	44	45	43	43
GRID valuation blend, one hour	Q4 2025	not stated on the chart; GRID values its fleet with RPI at 2.5%	60	60	61	66	69	69	71
GRID valuation blend, 1.5 hours	Q4 2025	not stated on the chart; GRID values its fleet with RPI at 2.5%	71	74	77	83	88	89	92
GRID valuation blend, two hours	Q4 2025	not stated on the chart; GRID values its fleet with RPI at 2.5%	82	86	89	95	103	103	107
GRID valuation blend, half an hour	Q4 2024, rebased	rebased by GRID to sit beside the Q4 2025 blend; basis otherwise not stated	48	45	48	56	58	54	52
GRID valuation blend, one hour	Q4 2024, rebased	rebased by GRID to sit beside the Q4 2025 blend; basis otherwise not stated	67	66	67	80	86	93	92
GRID valuation blend, 1.5 hours	Q4 2024, rebased	rebased by GRID to sit beside the Q4 2025 blend; basis otherwise not stated	82	81	84	99	105	114	113
GRID valuation blend, two hours	Q4 2024, rebased	rebased by GRID to sit beside the Q4 2025 blend; basis otherwise not stated	92	94	98	110	118	127	125
Measured floor, nearer one hour, 50 batteries, 1,825 MW	year to 27 Sep 2026	nominal, three published lines, no wholesale	46
Measured floor, nearer two hours, 67 batteries, 2,728 MW	year to 27 Sep 2026	nominal, three published lines, no wholesale	61

£ per kilowatt a year, every revenue line together, as published and not converted, because the money basis is not stated. The measured floor is the Vitreous measured index: what the same panel of batteries was actually paid on the balancing mechanism, in the response and reserve auctions and by the capacity market, annualised, with wholesale trading left out because nobody publishes it. GRID valuation blend, Q4 2025: A blend of two forecasters GRID does not name, used to value its 1,072 MW operating fleet at 31 December 2025. Sampled from the PDF vector paths at each year label, rounded to £100. The printed legend labels both line sets Q4 2024; the title and the text say the solid lines are the blend used in the current valuation and the dashed lines the previous blend rebased. GRID valuation blend, Q4 2024, rebased: The same two forecasters a year earlier. Sampled from the PDF vector paths at each year label, rounded to £100.

Why there is no Vitreous line on the chart yet

A ten year price needs a model of the whole system: every gas plant’s cost, the wind and solar output hour by hour, the demand, the cables to the continent, the carbon price. The engine exists and passes its tests. The national inputs do not exist yet, there is no carbon price series in the building at all, and the government’s central gas assumption is already less than half of this month’s price. Until a historical year can be reproduced to a stated error, no Vitreous number goes on this page. What can go here is the published range, kept on one money basis and checked against the outturn as the years arrive, and the things the price depends on.

Wind was paid 91% of the average price in 2024 and gas 117%, because they generate at different hours

Each technology’s capture price is the average price in the hours it was generating, weighted by how much it generated, shown here as a share of the year’s average price. Wind sold below the average in every year and gas above it. This is the shape of the market a battery lives in, not the spread a battery can capture: what a battery actually earns depends on when it charges and discharges, its usable energy, its losses and how far ahead it can see, and that arithmetic is done from real days on the UK Live page.

Hourly LCCC Intermittent Market Reference Price weighted by NESO published generation, calendar years. Wind is transmission plus embedded; solar is the embedded estimate. A reference price weighted by output describes exposure to the wholesale price and not what any generator was paid under its contracts. A description of three past years, not a forecast.

Table view of this chart

Year	Wind	Solar	Nuclear	Gas	Average price
2022	177	205	203	226	205
2023	90	85	93	106	94
2024	66	65	72	85	73

£ per MWh, nominal.

Gas is at £65 a megawatt hour in September, above the government’s high case for 2026. The market expects it back near £28 by 2028.

Gas sets the British power price in most evening hours and will for years yet. The government’s three assumptions, published in February, put 2026 at £30 central and £51 high. Across 2026 so far the monthly average has been £42, between the central and high cases, so the year is not yet outside the government’s range; this month is. The forward curve prices this winter as a crisis and the winter of 2028 as ordinary, which is roughly where the government’s central line sits.

Pounds per megawatt hour in 2026 money. Outturn is the ONS System Average Price of gas, monthly mean, to 20 September 2026, deflated with the same OBR series DESNZ uses. Forward prices are ICE NBP settlements of 9 September 2026 as reported by Catalyst Commercial, nominal, deflated here, plotted at the middle of each season. DESNZ fossil fuel price assumptions 2025, published 3 February 2026, converted from pence per therm in 2024 money; flat after 2040 by construction.

Table view of this chart

Year	DESNZ low	DESNZ central	DESNZ high
2025	28	34	43
2026	19	30	51
2027	16	26	44
2028	15	26	44
2029	15	26	43
2030	15	25	43
2031	15	25	43
2032	14	25	42
2033	14	25	42
2034	14	25	41
2035	13	25	41
2036	13	24	40
2037	13	24	40
2038	12	24	39
2039	12	24	39
2040	11	24	39

£ per MWh in 2026 money. Forward curve settled on 9 September 2026, deflated: Winter 2026 £66, Summer 2027 £44, Winter 2027 £41, Summer 2028 £26, Winter 2028 £28.

What is on this page and what is not

On it: every public long term view of the price on one money basis, the outturn and the forward curve beside them, the fleet as built, the contracted book by year, NESO’s published pathways, every published forecast of what a battery earns with the measured floor beside it, the capture prices of the past three years, and the government’s gas assumptions against the gas market. Every number traces to a published file or to the Vitreous register, and the table under each chart carries the values and the sources.

Not on it: a Vitreous price forecast, a Vitreous battery revenue forecast, or a probability that any project builds. The delivery risk score on the Atlas reads consent stage against years to the contracted date and is not a probability.

Published views: the latest vintage from each publisher is drawn and every vintage stays in the table. When a publisher issues a new view the old line moves to the table and the new one takes its place. Contains public sector information licensed under the Open Government Licence v3.0. Supported by National Energy SO Open Data.