The Hum You Won't Miss: Inside Britain's Quiet Experiment With Magnetic Cooling

Stand in any British kitchen at 2am and you'll hear it: a low, patient drone from under the counter. That's a compressor, and it has been the beating heart of refrigeration for about a century. It works brilliantly, it costs very little, and it is extraordinarily hard to beat. Which is why the idea of replacing it with a lump of metal and a magnetic field is both thrilling and maddening in equal measure.

The idea has a name — magnetocaloric cooling — and it has been kicking around laboratories since the 1970s, with a burst of interest every decade or so. Right now, with refrigerant gases under regulatory pressure and energy bills still making people wince, it's having another moment. So what is it, who in Britain is working on it, and why isn't there one in your utility room?

The physics, in one paragraph and then a bit more

Some materials get warmer when you put them in a magnetic field, and cooler when you take it away. That's the magnetocaloric effect, and it isn't magic — it's thermodynamics doing its usual bookkeeping.

Inside a magnetic material, unpaired electrons behave like tiny compass needles. Normally they point in all directions, a state of high disorder — high magnetic entropy. Apply a strong field and they snap into line, which is suddenly very orderly. Nature dislikes order appearing from nowhere, so the material compensates: the atoms in the crystal lattice start vibrating harder, and the temperature rises. Switch the field off, the spins scramble again, the lattice gives up its energy to make that happen, and the material cools.

Build a machine that does this in a loop — magnetise, dump the heat to the room, demagnetise, soak up heat from the food compartment — and you have a fridge with no compressor, no refrigerant gas, and almost nothing that moves except a bit of fluid and a rotating magnet assembly.

The textbook material is gadolinium, a rare-earth metal whose magnetic ordering temperature sits conveniently near room temperature. It's also expensive and awkward to work with, so much of the modern research has moved to alloys: lanthanum-iron-silicon compounds tuned with cobalt or hydrogen, manganese-iron-phosphorus materials, and various Heusler alloys with a habit of producing sharp, useful transitions. The goal in each case is the same: a large temperature change from a modest field, using elements you can actually afford.

Why anyone bothers

The environmental argument is the easy one. Conventional fridges and air conditioners rely on hydrofluorocarbons, and while modern HFCs are far better than the CFCs they replaced, many still have global warming potentials in the thousands. UK F-gas rules and the international Kigali timetable are steadily squeezing them out, and the replacements — propane, carbon dioxide — bring their own headaches, from flammability to very high operating pressures.

A magnetic fridge sidesteps all of that. Its 'refrigerant' is a solid slab of alloy, and the heat-transfer fluid can be water with a bit of antifreeze in it. Nothing to leak, nothing to top up, nothing to phase out.

The efficiency argument is the more interesting one, and also the more slippery. Laboratory prototypes have reported efficiency gains of 20 to 30 per cent over vapour compression in like-for-like tests. Real machines, with real pumps and real heat exchangers and real pressure drops, tend to eat into that margin. And a modern A-rated fridge is already a remarkably efficient device, so the bar is not low — it's a moving target that improves every time the Ecodesign rules tighten.

The British angle

Britain has form here. Camfridge, a Cambridge spin-out, spent well over a decade building magnetocaloric demonstrators, winning awards and attracting serious industrial partners before quietly winding down — a familiar arc for deep-tech hardware in the UK. Its work fed into the wider research base rather than a product on a shelf.

That base is still there. Groups at Cambridge and Imperial have published extensively on magnetocaloric and the closely related barocaloric materials, where pressure rather than a field drives the effect. Materials characterisation, magnet design and thermal engineering are all areas where UK departments punch above their weight, and there is a steady trickle of EPSRC-funded projects looking at the next generation of alloys.

What Britain does not have, at least not in any volume, is a domestic supply chain for the powerful permanent magnets such a machine needs. That's a global problem, and a pointed one, given how concentrated rare-earth processing is.

The awkward questions

Here's where the enthusiasm meets a spreadsheet.

You need a real field. To get a useful temperature lift you need something in the region of one to one and a half tesla across a decent volume. Permanent magnet arrays can do it, but they're bulky, heavy, and stuffed with neodymium and dysprosium. Electromagnets are simpler but burn power. Superconducting magnets are magnificent and completely impractical for a £400 appliance.

It's a batch process, not a continuous one. A fridge needs steady cooling, so a real machine runs several beds in sequence with a regenerator and a heat-transfer fluid shuttling between them. Every pump, valve and pressure drop is a small tax on your efficiency gain.

Materials get tired. Millions of cycles, thermal expansion, brittleness, hysteresis losses. An alloy that performs beautifully for a thousand cycles in a paper may look considerably less impressive after a million.

The incumbent is cheap. A domestic fridge is a low-margin box built at enormous scale. A magnetic equivalent has to be manufactured, certified and serviced at a price that competes with something already sold in every supermarket in the country.

Where it might actually land first

Probably not in your kitchen. The likelier beachheads are places where the economics already work: supermarket chillers, where a few per cent of efficiency is real money and refrigerant leaks are a genuine compliance headache; medical cold chain and laboratory equipment, where quiet and precise matters; data centre cooling and air conditioning, where the load is steady and the scale is industrial.

Get those right, build a supply chain, drive down the magnet cost, and a domestic unit becomes plausible. That's a decade-scale project, not an autumn one. Anyone promising a magnetocaloric fridge in a John Lewis showroom next year is selling you something other than physics.

Which leaves us where we started, listening to the hum. It's a good hum, in fairness — the sound of a technology so refined that it has become invisible. The quiet race is to make it disappear entirely, and the most interesting part is that the physics to do it has been sitting in the textbooks for fifty years, waiting for the manufacturing to catch up.