‘War is peace. Freedom is slavery. Ignorance is strength. Rationing is flexibility.’ — George Orwell, Nineteen Eighty-Four (revised and updated by NESO)
FOR YEARS the working assumption of the electricity system was that supply follows demand: you flick the switch, the grid responds. This week the paperwork for the opposite arrangement arrived. When generation follows the weather, demand becomes the variable that adjusts — and in seven days the state published the machinery for adjusting it, stripped a procedural right that stood in the way of the build-out, and had its favourite cost claims corrected twice by its own side’s house journal.
Start with the pair that frames the week. On Thursday NESO expanded its Demand Flexibility Service — the scheme that pays households to shift electricity use away from peaks — reporting £23.3m of ‘flexibility value’ between April and July. A day later DESNZ published draft load-control licence regulations: a new Ofgem-administered licence for organisations that remotely manage the demand of domestic heat pumps, EV chargers and batteries, with applications opening in March 2027 and the requirement binding a year later. It is sold as ‘consumer-led flexibility’ that helps consumers ‘make energy bill savings’. Read it plainly: you are offered a small payment to let a licensed third party turn your appliances off, because supply can no longer be guaranteed to follow your demand. Flexibility is rationing; the payment is the anaesthetic. It is the demand-side half of the firming problem The Energy Trap walks through.
The build-out side made the same admission from the other end. DESNZ opened a consultation to let ground investigations for grid projects proceed without a planning application, the stated aim being to ‘de-risk electricity network delivery’ for Clean Power 2030. Programmes remove procedural steps when the schedule binds on the physical side. If the constraint were ambition or money, you would not need to touch planning law; the constraint is how fast the wires can actually be surveyed, consented and built — the one variable the timetable assumed away.
On cost, the sharpest number of the week came from Professor Gordon Hughes and Dr Lee Moroney of the Renewable Energy Foundation, who total UK renewable subsidies at £274bn over 2005–2025 — around £890 per household per year recently, and, by an independent route, above my own Subsidy Clock figure of £228bn. The total will be contested; the unit price is harder to wave away. They put the spend at £486 per tonne of CO₂ abated in the most recent period (2021–25), against social-cost-of-carbon estimates below £50. The subsidy debate keeps stalling on ‘compared to what’. Cost per tonne abated ends that argument, because it prices the programme in the one currency it was bought to deliver — and finds a roughly ten-to-one overpay before reliability enters the room.
Carbon Brief, meanwhile, spent the week conceding things. Its Q&A on carbon capture lays out the record: 75 operational projects worldwide capturing under 0.2% of fossil-fuel emissions, against an IEA pathway assuming nearly thirty times that by 2035; real plants capturing 17–80% of their CO₂ against a 90–95% design target; and an Oxford analysis finding a low-CCS route to net zero around $1tn a year cheaper. CCS is the assumption that lets the models call net zero affordable. The technology most conspicuously failing to arrive is the one the bill depends on — and its absence would make the plan cheaper, not dearer.
Its heatwave factcheck conceded something subtler. The piece argues every generation type derates in heat — nuclear loses about 6% of output as cooling water warms, gas turbines about 10% per 10°C — so singling out renewables is unfair. Then it reports that during June’s heatwave UK wind fell to roughly 15% of the mix from a norm near 30%, at the exact hours cooling demand peaked, and still files wind under ‘resilient’. That is answering an adequacy question with an efficiency answer. The derating that matters is not a reactor losing 6%; it is a third of the grid becoming a sixth when the system needs it most.
China supplied the week’s perspective. John Kemp reports that Chinese policymakers now describe coal as a ‘ballast stone’ and a ‘safety net’ — held in reserve to steady a system dominated by renewables — and the new five-year climate plan contains no new ambition: emissions peak ‘before 2030’, and by 2035 fall only 7–10% below that peak. So in the week Britain rewrote planning law and licensed the control of household appliances to decarbonise faster, the world’s largest emitter confirmed it will keep growing emissions for years yet. The tonne we pay roughly £480 to avoid is a tonne China remains free to add.
And for anyone consoled by the claim that Britain will become ‘Europe’s energy powerhouse’, Paul Homewood ran the arithmetic. Exports fetch around £40/MWh; the renewable generators supplying them are paid around £167 under their contracts; the £127 gap — some £8bn a year on projected volumes — lands on domestic bills. And the deeper problem is meteorological: north-west Europe shares one weather system, so our exportable surplus arrives in the same hour as our neighbours’. You cannot sell a glut that is synchronised across the market you would sell it into.
Seven days, one direction of travel: the plan is no longer bending the system to serve demand — it is bending demand, planning law and the ledger to serve the plan.
Subscribe and the watching comes to you, every Saturday.
The Energy Trap: Why the Renewable Energy Transition Can’t Work — And What Can is published by Swift Press in September.*
"Rationing is flexibility" made me laugh, though it's the kind of laugh that dies off fairly quickly when you realise they mean it. Being paid a few quid to let someone else switch off my appliances is not a saving, it's a polite way of saying the lights might not stay on.
It still surprises me why there is such a huge demand in public organisations to ever increase the complexity of the systems by which services to the public are delivered. Increasingly the drive to centralisation and computer-dominated controls just increases the risk of failure, and to an ever expanding impact on populations. All in the name of saving costs by reduction of persons employed and in the redundancy of delivery systems. I would not be surprised if those nominal cost savings have been long eradicated and even become more expensive due to the much higher demand for maintenance and security measures to keep the systems working.