The battery we already had
In the 1970s, night storage heaters were everywhere. They charged overnight on power the system couldn't otherwise use, held it as heat, and released it when homes needed warmth.
On our own analysis, the installed fleet at its peak added up to something close to the capacity of Britain's entire nuclear fleet, and considerably more than pumped hydro. Germany's fleet was larger still.
Then they were taken out, and as heaters, they deserved to be. Economy 7 wasn't the bargain it sounded, the heat had usually gone by teatime, and nobody misses being cold in their own living room.
But we didn't just lose a bad heater. We lost several gigawatts of demand that could be moved to the hours when power was plentiful.
The gap that opened up
Today, by our estimate, Britain runs around 55 GW of inflexible nuclear, wind and solar generation against roughly 16 GW of flexible demand and storage. Germany's gap is starker still, at something like 185 GW against 18 GW.
Grid-scale batteries are growing quickly, but not as quickly as inflexible generation, so the gap keeps widening.
These are our own figures, not a published dataset. We give them as approximations because the categories are open to argument, and the conclusion doesn't rest on the second decimal place. The direction of travel isn't in doubt.
The gap has a price, and industry pays part of it
When generation can't be matched by demand, the system operator pays to fix it: wind gets curtailed while other plant is paid to fill the gap it leaves. Those balancing costs don't disappear. They end up on bills, including industrial ones.
So a site raising steam from gas pays twice: once for its own fuel, and again, through its electricity bill, for a shortage of flexibility it isn't in a position to help with.
Heat is still the cheapest flexibility
The storage heater was a bad product built on a sound principle. Heat is cheaper to store than electricity, and a store of heat is a demand that can be moved.
What was wrong with it in a house is mostly right on an industrial site. A factory's heat demand is large, runs to a schedule, and is already managed by people who think in terms of load and cost. There's no comfort penalty in charging a thermal store at three in the morning, because nobody is sitting next to it waiting to warm up.
What that looks like as a product
A Caldera Storage Boiler is an electric boiler with thermal storage and controls built in. It charges when power is cheap and plentiful, holds the energy as heat, and delivers saturated steam at up to 16 barg (204°C), over 95% dry, with 94% round-trip efficiency and a discharge response of under 60 seconds.
Because charging and discharging are decoupled, your electrical import is sized to your average load rather than your peak steam demand. That's what lets many sites electrify within the grid connection they already have, and it means the site helps the grid rather than adding another spike to it.
It integrates directly with your existing steam plant and runs alongside your current boilers, so you can turn a gas boiler down rather than switch it off. Each Heat Cell stores 5 MWh with 2 MWe of charging, and systems scale from 5 MWh to beyond 100 MWh.
In short, it gives you the flexibility of a grid battery from the electric boiler you already needed.
Test it against your own load
Whether this works on your site depends on your load shape and your tariff, so we'll model it on your numbers. Send us twelve months of half-hourly demand data and we'll simulate a Storage Boiler across all 8,760 hours of a year against real price signals, then take you through the results.
About this analysis
The capacity figures in this article are Caldera's own analysis and are given as approximations: that Britain's night storage heater fleet at its 1970s peak approached the capacity of the entire nuclear fleet and exceeded pumped hydro; that Germany's fleet was larger; that Britain today runs roughly 55 GW of inflexible nuclear, wind and solar against roughly 16 GW of flexible demand and storage; and that the German figures are roughly 185 GW against 18 GW. No external body has published these figures in this form, and we don't attribute them to anyone else.
What we say about curtailment and balancing costs reaching industrial bills is a general account of how system balancing is paid for, not a quantified claim, so we haven't given a figure per megawatt hour.
The product specifications are Caldera's published figures. Caldera's industrial-scale demonstrator at Fareham has met its targets every year since 2024.
