Energy is always a political hot topic. Affordability is in focus and I expect how to achieve long term affordability to be a clear priority topic for the coming years. Unfortunately, it is a complex problem.
Yesterday, former Energy Secretary and current Shadow Secretary of State Claire Coutinho launched a report with a deliberately provocative claim: Britain could save £320 billion over twenty years by placing greater emphasis on firm power, particularly nuclear and gas, and building less wind, solar and network infrastructure. The political message was equally direct: if we want households and businesses to electrify, electricity must be affordable.
The underlying work by Transira Energy is not a slogan dressed up as analysis. It uses comprehensive modelling including PLEXOS, and is unusually candid about important limitations. Read the Transira report and Claire Coutinho’s launch post.
But models do not discover a single future. They compare internally consistent futures built from assumptions. The right question is not whether £320bn is a ‘real’ number. It is what must be true for that difference to arise, whether those policy choices are politically and internationally deliverable, and what costs sit outside the modelled power-system account.
What the £320bn actually compares
The report compares two pathways from 2030 to 2050. Its Business as Usual case is already more cautious than Government ambition. It does not reach a clean power system until 2045. Its Alternative Policy Pathway removes the statutory net zero framework after 2029, takes the GB power sector out of the UK Emissions Trading Scheme (UK ETS), ends new support for most renewables and flexible low-carbon technologies, terminates Renewables Obligation (RO) payments for wind and solar from early, slows policy-led electrification, builds less network, and prioritises more gas and nuclear capacity.
On those assumptions, the alternative has £320bn lower cumulative modelled power-system costs in real 2025 money. It also produces 524 million tonnes more direct GB power-sector carbon dioxide and uses 575 TWh less electricity cumulatively. The report pro-rates costs to account for that lower consumption.
This is an important distinction. The £320bn is not a forecast of household bill savings and it is not a pile of avoidable expenditure sitting inside today’s policy. It is the modelled difference between two systems pursuing different outcomes. Supplier costs, margins and VAT are outside the cost index. The modelling notes also exclude effects in heat, transport, industry and the wider economy, investor returns, economy-wide emissions, network losses and full lifecycle emissions. The impact of this outcome therefore would be different to £320bn as costs change elsewhere.
The useful way to read the headline is: £320bn is a quantified trade-off inside a defined power-system boundary. It is not a complete balance sheet for UK PLC.
The biggest caveat for me: difficult policy reversals
Removing the UK ETS contributes to about £94bn
The report’s own cost breakdown attributes £94bn of cumulative saving by 2050 to lower wholesale electricity expenditure, principally following removal of the power-sector UK ETS. That is about 29% of the £320bn total. In the earlier period it matters even more: wholesale savings are approximately £20bn of the £44bn accumulated by 2035, or about 45%. The other 2050 savings are £137bn from networks, £67bn from balancing and ancillary services and £22bn from generation subsidies net of additional capacity-market costs. See the report.
The £94bn should be treated as an attribution, not a clean ETS-only sensitivity. The report does not show a third model run that removes the ETS while holding demand, generation build, renewable support and network investment constant. Wholesale prices, dispatch, CfD top-ups, interconnector flows and investment all respond together.
There is also a whole-economy accounting issue. Buying an allowance is a cost to a generator, but auction proceeds are revenue to government. ONS recorded total UK ETS receipts of £4.1bn in 2024, although only part came from power generation. Removing the carbon price can lower the wholesale market clearing price, but some of the apparent saving is a transfer that disappears from public revenue rather than a resource cost removed from the economy. Lower wholesale prices also increase top-up payments to generators with fixed-price CfDs. ONS UK ETS receipts.
The international constraint is substantial
As of August 2026, the UK and EU are negotiating linkage of their emissions trading systems. The EU’s negotiating mandate explicitly includes electricity generation and requires the UK cap and reduction pathway to be at least as ambitious as the EU’s. It also envisages dynamic alignment and the conditions for mutual exemption from UK and EU Carbon Border Adjustment Mechanisms. Removing GB electricity generation would therefore cut directly across the proposed agreement. EU negotiating mandate.
Without linkage, UK exporters remain exposed to EU CBAM reporting and to any top-up where the EU carbon cost exceeds the price already paid in the UK. Traded electricity is within scope. In August 2026, Ofgem approved provisions for a potential CBAM flow tariff on the North Sea Link from January 2027 if linkage has not been completed. This is tangible evidence that carbon-border costs can alter whether interconnector exports are economic. Commons Library CBAM briefing; Ofgem decision.
The report proposes a bespoke Carbon Reference Price on GB electricity exports. That is an interesting attempt to address leakage, but a unilateral UK mechanism would not automatically secure EU acceptance, CBAM exemption or access to a linked market. It could leave cheaper electricity inside GB while reintroducing an EU-linked carbon cost at the border.
There is a potential economic opportunity cost too. HM Treasury’s preliminary modelling estimated that linking could increase the UK’s long-run GDP level by around 0.1%, approximately £2.7bn in 2023 prices, before adding any further benefit from CBAM exemption. The estimate is uncertain and should not be read as the exact loss from withdrawal, but it demonstrates that the wider economic relationship is material. HM Treasury linkage analysis.
Ending RO payments early adds an investor-confidence cost
The alternative pathway also ends Renewables Obligation payments for wind and solar from 2033. These are legacy assets whose investment, debt and subsequent ownership were priced on the policy intention that accredited generators would receive a set level of support for their eligibility period. Three years’ notice may help operational planning, but it does not preserve the lifetime economics on which financing decisions were made.
The Government’s own 2026 response on changing RO inflation indexation is instructive. Even that narrower intervention produced warnings about retrospective policy, legal challenge, refinancing stress, higher risk premia and increased costs in future CfD rounds. The UK and devolved governments acknowledged that a stable and predictable policy framework is critical to maintaining investment appetite. Early termination of payments would be a far larger intervention than changing indexation. Government RO response.
This matters beyond the projects directly affected. Investors finance UK energy infrastructure across renewables, nuclear, storage, networks and emerging technologies. If they conclude that long-dated policy-backed revenues can be withdrawn retrospectively, they will demand a higher return, shorten the tenor of debt, reduce leverage or invest elsewhere. Any rise in the cost of capital feeds back into future CfD bids, RAB allowances and consumer bills. It may also affect the value of existing assets held by pension funds and infrastructure investors.
The Climate Change Committee has similarly stressed that long-term policy certainty is needed to rebuild investor confidence and unlock private capital. The precise financing penalty is not quantified in the Transira report, so it cannot simply be deducted from £320bn. Omitting it still makes the headline an incomplete measure of the effect on UK PLC. CCC progress report.
Why £320bn is not a UK PLC saving
None of these points invalidates the PLEXOS modelling. They show that it answers a narrower question: what happens to selected GB power-system expenditures under a different policy pathway? A UK-wide economic assessment would also need to consider lost ETS receipts, carbon-border costs, the value of UK-EU market linkage, higher financing risk, legal and compensation exposure, industrial and supply-chain effects, additional fossil-fuel imports, the economic activity enabled by electrification, and the environmental impact of higher emissions.
The network saving illustrates the same boundary problem. £137bn is avoided partly because the alternative builds around 80 GW less capacity and uses less electricity. That is a saving if the only objective is to minimise the power-system account. It may be a lost enabling investment if transport, heating, industry and data centres need more electricity to grow and decarbonise.
The phrase ‘£320bn saving’ therefore carries more certainty than the analysis can support. It combines a series of contingent policy decisions, some of which would be difficult to implement internationally, while excluding several effects on the wider economy. The useful contribution is not the single number. It is the challenge to identify which system costs are necessary, which can genuinely be reduced, and where apparent savings move risk or cost somewhere else.
Four other sensitivities behind the headline
1 | Gas-price risk
The model applies the same gas-price path to both scenarios, blending ICE futures and DESNZ values to 2030 and using DESNZ assumptions thereafter. The pathway falls below £25/MWh thermal for much of the long term. That is a legitimate modelling input, but it is not a neutral one: the alternative pathway burns materially more gas, so a higher gas price affects it more heavily.
DESNZ’s own 2025 fossil-fuel assumptions make the uncertainty explicit. By 2050 its three gas cases span approximately £11, £23 and £37/MWh. DESNZ warns that these are assumptions rather than forecasts and encourages users to test the full range. Near-term NBP market indicators at the time of writing were also materially above the report’s long-run central path. A current spot or front-month price is not a 2050 forecast, but the gap is a useful reminder that fuel-price sensitivity matters. DESNZ assumptions; ICE UK NBP market data.
The published report acknowledges commodity-price uncertainty, but it does not show how much of the £320bn survives a sustained high-gas case. That would be one of the most useful additional results to publish, particularly because gas-price exposure is itself part of affordability and energy security.
2 | Nuclear cost, finance and delivery
Nuclear has not been ignored or left uncosted. The report includes allowed revenue during construction, assumes capital costs of £10,000-£12,500 per kW, and uses technology-specific revenue assumptions for large reactors and small modular reactors. It also assumes that an alternative government can accelerate planning, regulation, site readiness and grid connections sufficiently to build around 7 GW more nuclear capacity by 2050.
The sensitivity is less ‘has nuclear been costed?’ and more ‘how confident are we about cost, financing and delivery at that scale?’ The comparison below separates construction cost from the revenue or support measure expressed in £/MWh. These are related but not interchangeable.
| Measure | Hinkley Point C | Sizewell C |
| Capacity | 3.2 GW | 3.2 GW |
| Latest cost basis | £35bn in 2015 prices; EDF equivalent about £48bn current | £38.2bn baseline and £47.7bn higher regulatory threshold, in 2024 prices |
| Approx. capital cost | £10,940/kW in 2015 prices; £15,000/kW using EDF’s current-value conversion | £11,940/kW at baseline; £14,910/kW at the higher threshold |
| Comparable £/MWh | Fixed CfD equivalent of £129/MWh in 2025 prices | Estimated equivalent of £133-£155/MWh in 2025 prices; not a fixed strike price |
EDF’s February 2026 estimate puts Hinkley Point C at £35bn in 2015 sterling, approximately £48bn in current value, with Unit 1 targeted for 2030. A further twelve-month delay could add around £1bn in 2015 money. For Sizewell C, the NAO gives a £38.2bn baseline and a £47.7bn higher regulatory threshold in 2024 prices. EDF disclosure; NAO Sizewell assessment.
The £/MWh comparison needs particular care. Hinkley has a fixed CfD equivalent to £129/MWh in 2025 prices. Sizewell does not have a fixed strike price because its RAB allowed revenue responds to actual costs and financing. The NAO estimates an equivalent £133-£155/MWh in 2025 prices across the baseline to higher-threshold range. The Government business case also presents a net equivalent strike price of £82/MWh in 2012 prices under a moderate-outturn assumption. Transira’s own inputs of £138/MWh for Hinkley and £150/MWh for Sizewell sit within this broad evidence, but the report says they are not used for its subsidy calculation. Sizewell business case.
These figures do not make nuclear the wrong answer. Firm, low-carbon power can reduce other system costs, improve resilience and complement renewables. They do mean that construction duration, financing costs, replication benefits and cost overruns deserve explicit stress tests before nuclear-led savings are treated as bankable.
3 | Electrification and data-centre demand
The report does consider data centres, and does so aggressively. By 2050 it assigns them around 62 TWh of annual demand in the alternative pathway, about 14% of total consumption, compared with 40 TWh and 8% in the Business as Usual case. It also models electric vehicles, electric heat and electrolysis.
What changes the result is that the alternative deliberately removes mandates and support for electrification. Its 2050 electricity demand is 441 TWh, compared with 484 TWh in Business as Usual, because lower demand from transport, heating and hydrogen more than offsets additional data-centre use.
That creates a difficult boundary question. The Climate Change Committee’s Balanced Pathway envisages a much larger electricity system, with electricity use roughly doubling from today by 2040 as efficient electric technologies replace oil and gas. The figures are not directly comparable, but they illustrate the sensitivity. If more homes, vehicles, factories and data centres electrify, some network investment stops looking like an integration penalty and starts looking like productive capacity for the wider economy. Seventh Carbon Budget.
Data-centre demand also cannot be represented only by an annual TWh number. Location, coincidence with system peaks, backup generation, storage, on-site renewables and the ability to shift computing loads can make the same consumption either a network problem or a flexible grid asset. Ofgem’s current work on strategic demand connections recognises this locational and readiness challenge. Ofgem on strategic demand connections.
4 | Smart systems and flexible demand
It would also be unfair to say the model simply ignores flexibility. It uses hourly PLEXOS dispatch and includes batteries, pumped storage, interconnectors and demand-side response. The alternative pathway is cheaper partly because it builds fewer of these assets and faces lower balancing and network costs.
The open question is whether the published assumptions capture the full value of a more digital and responsive demand side: managed EV charging, thermal storage, automated heat-pump operation, vehicle-to-grid services, dynamic tariffs, local flexibility markets and data centres that move non-urgent computing across time or location. The appendix aggregates demand-side response, but does not make these individual sources or their adoption assumptions transparent.
The joint Government and Ofgem Smart Systems and Flexibility Plan estimated in 2021 that smarter flexibility could reduce system costs by up to £10bn a year by 2050 by avoiding generation and network capacity needed only for peaks. That estimate cannot simply be deducted from the report’s £320bn because there will be overlaps and different assumptions. It does show why demand shape matters alongside annual demand and generation mix. Smart Systems and Flexibility Plan.
Affordability and climate are not rival truths
Removing a carbon price reduces the cost visible inside the power system, but it does not remove the carbon. It changes where, or whether, the environmental cost is recognised. In the model, the alternative produces 524 Mt more power-sector CO2 by 2050, while climate damages, methane leakage, imported emissions, air quality and broader environmental effects sit outside the boundary.
This is why credible models can produce very different large headline numbers. The Climate Change Committee’s supplementary Seventh Carbon Budget analysis, using a whole-economy and social-cost framework, concludes that net zero benefits outweigh costs by 2.2 to 4.1 times and estimates avoided climate damages of £40bn-£130bn in 2050. That is not a direct refutation of Transira. It is a different question with a different boundary. The contrast is itself instructive. CCC supplementary analysis.
Globally, the prevailing direction is towards more carbon pricing rather than less. The World Bank reports that direct carbon prices now cover just over 29% of global emissions through 87 policies. Prices and designs vary greatly, but removing GB power from carbon pricing would be a notable reversal at the same time as the EU is using CBAM to extend its carbon price into trade. World Bank carbon pricing report.
The objective should be to minimise the total cost and environmental impact of the energy services society needs, not to minimise one line in the electricity account. That means considering direct and lifecycle emissions, fuel imports and price volatility, land and materials, biodiversity, resilience, health, consumer capital costs, financing conditions and the economic value enabled by new infrastructure.
The politics will become sharper
Energy affordability may now be the most important subject in the transition. High electricity prices undermine industrial competitiveness, deepen fuel poverty and make heat pumps and electric vehicles harder to justify. Questioning the pace, technology mix, market design and distributional consequences of net zero is therefore valid.
But the debate is moving rapidly from engineering into electoral politics. A single number such as £320bn is almost irresistible because it appears to turn a complex system choice into a simple saving. The danger on both sides is symmetrical: net zero advocates can underplay real delivery and integration costs, while opponents can describe a higher-carbon, lower-electrification system as though it delivered the same outcome more cheaply and without wider economic consequences.
My concern is not that large numbers should be ignored. It is that they should be opened up. In this case, almost one third of the saving sits in a wholesale category linked principally to a carbon-market withdrawal that would be politically and internationally difficult. Another policy change risks weakening confidence in long-term UK commitments. Neither consequence is captured in a single £320bn line.
The lesson from my own home
My own journey through renovating and electrifying a house has made this debate far less abstract. Removing the gas meter felt like a clear final step. Discovering later that the physical gas service still had to be found and disconnected revealed the infrastructure, cost and uncertainty hidden beneath apparently simple choices.
It has also reinforced how often the transition is described as a shopping list of new technologies. The first task is to use less: insulation, draught reduction, sensible temperatures, well-sized equipment and basic energy efficiency. The next is to use electricity more intelligently. Smart controls, batteries, hot-water storage, EV charging and flexible appliances can move demand away from expensive peaks and towards periods when power is cheaper and cleaner, often without changing the service the household receives.
None of this makes the transition costless, and not every household has the capital, tenure or confidence to participate without help. But it points towards a better organising question. Not ‘net zero at any cost’ and not ‘how little grid can we build’, but: how do we deliver affordable, resilient and progressively lower-impact energy while maintaining the confidence needed to finance it? That is a harder question than a £320bn headline. It is also the one Britain now needs to answer.
