The science behind the heat pumps in my novels (and where I made things up)
In Humanity's Hope, one piece of technology quietly holds the whole colony together: a magnetocaloric heat pump, connected to a thermal battery, generating the electricity that keeps the colonists alive. In my next novel, Artificial Utopia, a version of the same idea powers entire desert regions of Earth. Readers sometimes ask how much of this is real. The honest answer is: more than you'd think, but not all of it — and I think the gap between the two is worth explaining.
The real science
Magnetocaloric heat pumps rely on a genuine, well-understood physical effect: certain materials change temperature when you change the magnetic field around them. Apply a strong field, the material warms slightly; remove it, and it cools. Cycle that fast enough, in the right arrangement, and you can move heat from one place to another without any of the mechanical compression or chemical refrigerants that today's heat pumps and air conditioners rely on.
This isn't a new idea — the core thermodynamic cycle behind it was patented in the early 1980s. What's new is that it's only very recently become practical. Researchers at the U.S. Department of Energy's Ames National Laboratory have built a prototype that, for the first time, matches conventional heat pumps on weight, cost, and performance all at once — using discs of gadolinium as the active material, arranged so permanent magnets can cycle the field across them while fluid carries the heat away. Earlier magnetocaloric devices could usually only hit one or two of those three targets, which is why the technology has stayed in the lab rather than reaching your home.
This isn't only lab paperwork, either. Back in 2014, GE Appliances and Oak Ridge National Laboratory built and publicly demonstrated a magnetocaloric refrigerator prototype: a "cascade" of fifty small magnetocaloric stages, each nudging the temperature down by a little over a degree, adding up to an 80°F swing overall. They famously used it to chill a beer for the cameras. It used gadolinium, the same material at the heart of the Ames Lab work, and GE predicted household versions within the decade. That didn't happen — gadolinium is expensive, the prototype was far too large to fit inside a real fridge, and it doesn't seem to have progressed to a consumer product since. More recently, a German company called Magnotherm has gone further: it now markets what it calls the world's first commercially available magnetic cooler, and has actually shipped units — including to Coca-Cola — since 2023. Rather than gadolinium, Magnotherm uses a different magnetocaloric alloy, made from lanthanum, iron, and silicon, which shows the field has more than one promising material to work with.
So: the physics is real, the recent breakthrough is real, the fridge that chilled a beer on camera is real, the commercial product actually shipping today is real, and the reason vapour-compression still dominates today is real too. What's not real yet is a magnetocaloric heat pump built at the scale, cost, and material sophistication my colonists rely on. That part is mine.
What I invented for the story
In the novel, the heat pump uses thin discs of a material I call "gadolinium ceramic," sandwiched between layers of graphene. The discs sit on a spindle, forming a cylinder, and shuttle back and forth between the graphene layers to transfer heat rapidly — each individual transfer is tiny, but it happens roughly a hundred times a second, so the cumulative effect is large. The magnetic field itself comes from superconducting solenoids rather than the rotating permanent magnets used in current real-world prototypes.
Gadolinium ceramic doesn't exist. I invented it as a plausible-sounding leap from gadolinium, the real material at the heart of today's research, on the assumption that decades of materials science might produce something with better magnetocaloric properties, more durable, and cheaper to manufacture at scale. I can't promise it's physically achievable — it might not be — but I wanted something that felt like a believable next step from where the real research already stands, not something built from nothing.
There's a nice detail buried in the plot because of this: when the colonists later find the remains of a similar technology in an alien civilization's ruins, it's the physical traces of these graphene-and-disc assemblies that the colonists identify — the design leaves a distinctive signature.
The thermal battery
The second piece is Gus Nicols's thermal battery, and this is where I extended the science furthest. He digs a pit and layers it with alternating bands of beach sand and graphene-rich sand, then connects the layers to the heat pump. The heat pump pulls heat from the air into the top layer, then relays it downward, layer by layer, each one hotter than the last, until the bottom layer is hot enough to drive a small steam turbine. It generates electricity and stores energy as heat at the same time — and the heat pump that runs the whole system can be powered by solar, or by drawing on some of the heat already stored.
Thermal storage using sand or similar cheap, abundant materials is a genuinely active area of real-world energy research — the idea of storing heat rather than electricity, because heat is so much cheaper to store at scale, is sound. What I've extrapolated is the specific graphene-layered design and the scale at which it works. In the novel, large thermal batteries on Earth, mostly constructed by machines with minimal human oversight, transform the fortunes of desert countries suffering from increasing temperatures and drought by providing affordable, plentiful energy for desalination, steel manufacturing, and data centres. That's the speculative leap: not the underlying principle, but the scale and material efficiency needed to make it work that well, that cheaply.
It also solves a real problem for the colonists specifically: interstellar travel is brutally constrained by mass, since every extra kilogram makes acceleration to the speeds required dramatically harder. A colony that can build its power infrastructure from materials already lying on the ground when it arrives, rather than hauling it from Earth, sidesteps one of the hardest physical limits of colonisation.
Where it's going next
In Artificial Utopia, I've pushed the idea further. Earth's deserts have become the site of all the planet's heaviest industries — steel, fertiliser, recycling, aviation fuel — alongside the data centres of the global AI, COGA, all running on the same abundant, clean energy. Even the androids get a miniature version: a small heat pump and steam turbine powered by a disc of radioactive material made from reprocessed nuclear waste. That last piece has a real-world cousin — radioisotope generators, which use the heat from decaying radioactive material to generate power, already exist and have powered deep-space probes for decades. Shrinking one down to android-scale, and pairing it with a magnetocaloric heat pump instead of the thermoelectric generators actually used today, is my own extension.
Why I bother with any of this
None of this research changes the plot. Nobody needs to understand the AMR cycle to follow Gus digging a pit. But I think it's part of what makes a future feel earned rather than assumed — if I can trace a reasonably honest line from a real 2026 lab prototype (and, in Magnotherm's case, a real shipping product) to a colony surviving on a technology descended from it, the world holds together a little better. That's really the same argument I made in my last post about "dystopian" as a label: I'm not interested in futures that appear from nowhere, good or bad. I want the reader to be able to see the load-bearing wall.
