What's Powering Your Hospital Today?

Introducing Grid Margin — a live, public gauge of how reliable Britain's electricity really is

ON THE evening of 23 June, with a heatwave settling over Britain and much of Europe, the National Energy System Operator did something it had never done in summer before. It issued an Electricity Margin Notice: a formal signal to the market that it might be short of power the following evening, between seven and ten o’clock, by something like 1,900 megawatts.

The lights stayed on. The notice was withdrawn by Wednesday afternoon, the market answered, and an EMN is not a warning of power cuts. NESO is right to say so, and right to say it is the system asking for more and usually getting it. But look at what “asking for more” meant that evening. To cover the gap, NESO had to ask the EU to lift the cap on how much Britain could import, and the country pulled in around 2.3 gigawatts from the continent. The lights stayed on because the neighbours had power to spare. That is luck, not a guarantee: another evening they may not, with their own demand high, their plant down for maintenance, or nothing left to sell. The price was close to £1,400 a megawatt-hour, roughly fifteen times normal. That one evening added an estimated £11 million to bills. And about 1.7 gigawatts of it came from the Netherlands, generated from coal and gas.

Sit with that last figure. This was a regional weather event: the same blocking heat and slack wind sat over most of north-west Europe at once. So when Britain went looking for power, what the neighbours had to spare was not wind or solar. They were short of those too. It was the fossil generation the whole continent is supposed to be leaving behind. We didn’t import clean power in a crisis. We bought expensive coal and gas, at a crisis price.

Kathryn Porter of Watt-Logic, whose analysis of the grid’s operational limits is essential reading, was blunt about the cause: NESO, she said, “failed to anticipate the shortfall due to poor modelling,” and was left reaching for extraordinary measures to cover it.

And NESO did it again two days later, with a second notice for the Friday evening: two summer margin notices in a single heatwave week, from an operator that had issued none in summer before. So hold on to the question of why any of this was necessary. Not a freezing January peak with the country switching on every kettle and heater at once. A warm June evening, demand well below the winter high. The margin thinned anyway, because solar fell away into the evening, wind was low, and conventional plant availability was tighter than expected, all at the same time. Those notices arrived not because demand was extraordinary, but because supply quietly wasn’t there.

That gap, between how the grid is described and how it actually behaves, is what I’ve spent the last few weeks building a tool to measure. It is called Grid Margin, it is live now at gridmargin.co.uk, and this is what it does.

The story we are told, and the one we aren’t

The official account of Britain’s electricity is a good-news story, and a familiar one. The grid is getting greener every year. Renewable power is cheap. Coal is gone, wind is up, emissions are down. Ministers say it, the regulator says it, the industry says it.

What the story leaves out is reliability. At any given moment, some of Britain’s electricity comes from firm, dispatchable plant: the gas, nuclear and biomass you can call on whatever the weather is doing. The rest relies on wind, solar and imported power across the interconnectors: sources that are cheap when they are there, and that have an awkward habit of fading away together. A cold, still, high-pressure spell does not settle over Kent and spare Calais. It becalms the whole of north-west Europe at once, which is precisely when the neighbours you were planning to import from are short themselves.

The data that would let you see this is all public. Elexon, NESO and DUKES publish generation, demand and capacity, some of it updated every few minutes, some once a year. But it arrives in separate feeds, in industry units, and it is almost always framed around carbon and cost: how clean the power is, how much it costs. Almost never around the one question a household actually cares about: can you rely on it?

I wanted a single place that reads those same feeds for reliability instead. So I built one.

What Grid Margin measures

The site does three things, all from the grid’s own numbers and nothing modelled.

It reads the live mix and splits it in two: the firm, dispatchable share against the weather-and-imports share, updated as the grid moves. That answers a question I keep coming back to. What, right now, is actually powering your hospital?

It shows the gap between installed and delivered. Britain has built an enormous amount of wind and solar capacity. The site puts that nameplate figure next to what the fleet is delivering this minute, which is very often a small fraction of it. Capacity is not the same as power, and the difference is the whole argument.

And it lays out the wind record since 2016, half-hour by half-hour, because the average hides what matters. Across the decade, wind has run at a capacity factor of around 29 per cent — under a third of its nameplate. But reliability lives in the bad days, not the average. On 423 days since 2016 the whole fleet delivered less than a tenth of its capacity, and when the wind drops it stays down: the calm of June 2016 ran ten days. On the worst single day, 19 January 2016, some sixteen gigawatts of turbines averaged 340 megawatts, about two per cent of what they were built to produce.

Those are not forecasts or scenarios. They are what happened, taken from the settled record.

The number the conversation skips

There is a word for the thing Grid Margin is named after, and the grid operator uses it every day. The margin is the spare firm capacity sitting above expected demand: the headroom. When NESO plans the system, it does not count wind, solar and interconnectors at their nameplate. It de-rates them, heavily, because it knows it cannot rely on them all being there in a crunch. That is not scepticism; that is the operator’s own arithmetic, the discipline that keeps the lights on.

The public conversation rarely does the same sum. It talks about what the renewable fleet can produce on a good day, and quietly assumes the good day. Grid Margin is a citizen’s version of the operator’s margin: the de-rated, weather-aware view, kept in plain sight instead of buried in a balancing-mechanism spreadsheet.

The 24 June notice is what that view looks like when it briefly goes amber. Not a catastrophe; the system worked exactly as designed. But it was a small, honest signal that the margin is thinner and more weather-dependent than the headline story admits. These notices are worth counting and watching as a trend, not treating as alarms. A single one proves nothing; the pattern of them, over years, tells you something. When one is in force, Grid Margin puts it where you can’t miss it: a banner across the top of the page, lit only while a scarcity notice is actually live.

And the imports are not just an emergency measure. On the settled record, the interconnectors have supplied around an eighth of Britain’s electricity so far this year, a structural reliance, not a one-off. Most of the time they are cheap and convenient, which is exactly why we rely on them. The lesson of 24 June is what they cost, and what actually fuels them, in the precise moments the system is tight.

Why you can check my working

Here is the part I care about most. Every figure on the site traces to its source, and the method is written out in full on the methodology page. Where a number carries a known bias, that bias is disclosed on the page, next to the figure, rather than hidden in a footnote. The wind capacity factor, for instance, is deliberately calculated as a conservative lower bound. The data, the code and the calculations are all in a public repository. If you think a figure is wrong, you can open the workings and show me. Confirmed errors get corrected and logged.

That openness is the whole bargain. A reliability gauge written in a sceptic’s voice is going to be probed hard, and it should be. The only way numbers like these survive that scrutiny is if they are reproducible from public data by anyone who doubts them. So I have made them exactly that.

Where this fits

Grid Margin is the sibling of the Subsidy Clock, which I built to measure the cost of energy policy as this measures the risk. Both grew out of the work for my book, The Energy Trap (Swift Press, 2026), which is about the gap between what we have been promised about energy and what the system can actually deliver. The book makes the argument at length; Grid Margin lets you watch one part of it happen in real time, from the grid’s own data, whenever you like.

This is a first release, a beta, and I’m putting it out to be tested. The import story I opened with isn’t yet a live instrument on the site; building that — so you can watch imports flip from cheap convenience to expensive fossil lifeline when the weather closes in — is what comes next. After it: the ability to register for an email alert when the site exposes a reliability event, so the next margin notice or wind drought finds you, rather than you having to go looking. There’s more besides. So three things.

Use it when it counts. You’ve seen how this goes: the wind drops, someone screenshots a live grid dashboard with renewables at two per cent, and posts it under the next “bills are falling” headline. It works, but it’s fiddly, and the numbers aren’t framed for anyone who doesn’t already read the grid. So I’ve done the framing. At gridmargin.co.uk/share there are ready-made cards, sourced and refreshed daily, that you can download or paste straight into a post. One link, one picture, a current figure nobody can wave away.

Check it. If you find an error before I do, tell me. That’s rather the point of publishing the code alongside the figures: it has to survive being checked by people who’d love to catch it out.

Watch it. Go and look at the live gauge now, and subscribe so you don’t miss the import instrument when it lands. The next time the grid asks for help in a heatwave, you’ll know where to see it for yourself.

---

Grid Margin is live at gridmargin.co.uk. It is self-funded, with no advertising and no sponsors. The Energy Trap is published by Swift Press in 2026.