This Cambridge Engineer Says We're Solving the Easy Problem Last

The Epiphany That Killed a Company

Before Caldera existed, Macnaghten spent a decade as CEO of Isentropic, a company developing large-scale thermo-mechanical electricity storage. The concept was straightforward: use a reversible heat pump cycle to heat up rock, store the energy, and recover 60-70% of it when needed. Rocks are cheap. The idea seemed sound.

Then lithium-ion batteries happened.

"When we started, I think lithium-ion batteries were over $1,000 a kilowatt hour for the packs," Macnaghten said. "When we wound up Isentropic, because it was quite clear we were too expensive, batteries were at $200 a kilowatt hour. And I think you can now buy cell packs out of China at $40 a kilowatt hour."

But the real lesson went deeper than price competition. Macnaghten arrived at what he calls a "pretty horrible epiphany" — one that reshaped his entire view of the energy landscape.

"Fundamentally, it's often just cheaper to build more renewables and throw the energy away, because that's often cheaper than storing it," he told us. "If you build more renewables, some of it you can use. That's great, you use it. The bit you can't use, you throw away. And that's cheaper than trying to create an expensive storage system to actually use it."

It is the kind of statement that sounds reckless until you sit with the math. We already pay to keep gas power stations idle for most of the year so they can run during peak demand. Curtailed renewables are not fundamentally different — just harder to accept emotionally.

Macnaghten wound down Isentropic and walked away with two hard-won engineering insights that would become the foundation of Caldera: never use air as a heat transfer medium, and insulation matters more than almost anything else.

What Caldera Actually Builds

Caldera manufactures electric Storage Boilers — large industrial units that convert renewable electricity into heat, store it, and deliver steam on demand. The storage material is a composite of volcanic rock and recycled aluminum, wrapped in proprietary vacuum insulation and housed in a cylindrical vessel roughly 10 feet in diameter and 30 feet tall, weighing about 110 tonnes.

The vacuum insulation is the key differentiator. At temperatures between 200°C and 500°C, radiative heat transfer becomes a serious engineering problem. Most industrial insulation solutions are bulky, heavy, and expensive to install. Macnaghten learned this the hard way at Isentropic.

"I looked at them one day, and I went, I think 50% of this very expensive store was insulation," he said. "You can end up with the insulation being not that far off the weight of the storage material, which is kind of crazy if you're trying to store stuff."

Caldera's vacuum insulation is thin and efficient enough to keep the unit compact and transportable. The goal — not yet achieved, but actively pursued — is to assemble a complete unit on site in a single day.

The boiler charges at a steady rate, often during the cheapest hours of the night or when on-site solar is producing, and then covers spikes in steam demand throughout the day. For a factory running gas boilers, the integration is seamless: Caldera's unit connects to the same steam header. If electricity prices spike or something goes wrong, the factory switches back to gas.

"From the factory manager's point of view, he doesn't care," Macnaghten said. "As long as the steam's there, he can always switch back to his boilers. Nobody has to change their behavior."

Solving the Easy Problem First

This is where Macnaghten's frustration with the broader energy conversation becomes pointed. That 20% emissions figure from industrial heat is not one problem — it is a spectrum.

About 40% of industrial heat demand falls below 200°C. That is roughly 8% of total global greenhouse gas emissions, and according to Macnaghten, it is the easy part. This is the temperature range used by food and beverage manufacturers, breweries, distilleries, paper mills, pharmaceutical plants, and hospitals — industries that run on steam and could switch to electric heat today.

Above 500°C, you are talking about cement, steelmaking, and glassmaking — processes that run at 1,000 to 1,300°C. That is genuinely hard.

"Everybody wants to go and decarbonize steelmaking, which is really hard," Macnaghten said. "And I go, this is stupid. Why are we doing the hard stuff when we haven't done the easy stuff?"

His argument is not that the hard problems do not matter. It is that better solutions for those problems will emerge over time — just as lithium-ion batteries blindsided every other storage technology in the span of a decade.

"If you think I couldn't buy an EV 10 years ago and now 50% of all the cars sold in China are EVs — I mean, it's just insane," he said. "If you try to fix everything 10 years ago, you're just going to waste money because you're not going to have the right solution."

The data supports his pragmatism. According to the World Business Council for Sustainable Development, commercially available solutions already exist for low and medium-temperature industrial heat — yet the share of renewable heat in industrial energy has barely moved, from 10% in 2015 to just 12% in 2023.

The Case for Hybridization

Macnaghten is emphatic that going 100% clean is, in most cases, the wrong goal right now.

"People love the idea of having something 100% clean, but to me that's almost always much, much more expensive than just being pragmatic and saying, I'll do 20%, we'll get to 20% here, we'll do another 20% then, and we'll see where we are."

This is the hybridization model: keep the gas boilers, add Caldera's electric storage alongside them, use whichever energy source is cheapest at any given hour, and ratchet down emissions incrementally. It provides resilience, too. As Macnaghten pointed out, the Ukraine war and recent tensions with Iran have demonstrated how fast energy prices can move. A factory with both electric and gas options is hedged.

His rule of thumb for when storage and flexible demand start to make economic sense is about 1,000 hours of surplus renewable generation per year. Below that threshold, the surplus is too infrequent to justify infrastructure. Above 2,000 to 3,000 hours, costs escalate, and the system needs to start absorbing that energy.

"Heat is just really unusual," he said, "because it's such a large amount of energy — this makes it a really great energy sink to use the renewable generation you'd otherwise throw away."

GEA and the Road to Scale

In March 2025, GEA — the German DAX 40-listed industrial equipment giant — invested €12 million in Caldera. Macnaghten believes Caldera is the earliest-stage company GEA has ever backed.

The strategic logic is clear. GEA already builds heat pump systems for many of the same industries Caldera targets: breweries, distilleries, dairy, food, and pharma. Heat pumps run efficiently but prefer steady loads. Factory steam demand is inherently spiky — open a valve and pressure drops, and every burner fires. Caldera's storage boiler handles those peaks, filling the exact gap that heat pumps struggle with.

"They see us very much as part of the solution," Macnaghten explained. The combined system — GEA's heat pumps for baseline efficiency, Caldera's storage for peak management — offers factories a complete electric alternative without requiring anyone to rip out existing infrastructure.

Caldera is now deploying its first commercial units, with two going into a hospital this year. The sales cycle is long — industrial hardware always is — and Macnaghten is deliberate about not rushing a product to market before it is ready.

"If you put the wrong product out, you never get over it," he said. "VC investors are particularly bad at insisting you get the first product out and then you spend the next five years trying to fix it in the field with a very unhappy customer."

Looking Out 10 Years

When we asked Macnaghten what the energy landscape looks like in a decade, he did not hedge.

"I think oil demand is going to crash," he said. He pointed out that the number of petrol cars worldwide peaked around 2022-2023, with China now putting fewer combustion vehicles on the road each year. His math suggests 20 to 30 million barrels per day of oil demand could disappear over a 10-to-15-year period.

He also offered a sharp take on AI and data center energy demand — a topic dominating energy headlines. While he acknowledges AI will consume enormous amounts of power, he questions the economics behind it.

"There are a hundred million people in well-paid jobs, and every one of those hundred million would have to pay an annual subscription of $40,000 a year to the big seven to justify their market cap," he said. "Most companies object to paying $400 an employee. So there's a factor of a hundred missing somewhere."

His prediction: a massive buildout of data centers followed by a bust, similar to the subsea fiber optic overbuild of the late 2000s. "No one's going to build data centers for five years," after that correction, he said.

The Bottom Line

James Macnaghten is not anti-fossil fuel. He is not anti-storage. He ran a storage company for a decade and shut it down when the economics no longer worked. What he is, more than anything, is a pragmatist in a space that rewards ideology over arithmetic.

His message is simple: the easy decarbonization wins are sitting right in front of us — in the steam pipes of breweries, hospitals, food plants, and pharma factories — and we keep walking past them to chase harder, more glamorous targets. Caldera's bet is that the factories will come around once the economics are undeniable. With GEA's backing and first units heading into the field, they are about to find out.

---

This article is based on the Local Energy Podcast interview with James Macnaghten, CEO of Caldera. Listen to the full episode on YouTube or wherever you get your podcasts.

Connect with James on LinkedIn or visit caldera.co.uk.