Cold Is Just Absent Heat: Inside the Vapour-Compression Cycle
written by Stefan Christoph
A quick note: this is part two of Lunch Break Physics, the little series that started with a suction cup and the realisation that “suction” is a misnomer. “Cold” is the same kind of misnomer, and the fridge humming in your kitchen is a more interesting machine than it lets on. Standard physics below; sources at the bottom; corrections welcome in the comments.
You can’t make cold, you can only move heat
Heat flows from hot to cold on its own, and never the reverse. That one-way street has a name: it is the second law of thermodynamics, and in particular the Clausius statement, that heat will not pass from a colder body to a warmer one without something else driving it. So chilling your food is not a matter of pouring “cold” into the box; it is a matter of forcing heat to travel the wrong way down that street, out of the cold interior and into the warmer room around it.
Forcing anything uphill costs something, and here the currency is work. This is not an engineering limitation that a cleverer fridge could someday dodge — it is a statement about entropy. Pulling heat out of the cold box lowers that box’s entropy, and the second law only permits a local decrease if a larger increase happens somewhere else to cover it. The compressor’s electrical work is what covers it: that work ends up dumped into the kitchen as additional heat, and the entropy it creates there more than pays for the order restored inside the fridge. So the electricity does not manufacture cold. It buys the right to move heat against its natural direction, with the ledger balanced by your kitchen ending up slightly warmer than the food ever was cold.
Plug it in
Watch the refrigerant loop carry heat around like a conveyor belt. Plug it in and watch what it is, not just how warm it is: small dots are vapour, big droplets are liquid. It boils on the left, condenses on the right, and its pressure changes at exactly two places.
The loop never runs out of heat to carry because it keeps going back for more. What changes around the loop is the refrigerant’s pressure, and with it, crucially, its boiling temperature.
The four stops on the heat conveyor
Follow one parcel of refrigerant around the circuit.
At the evaporator, the coil inside the cold box, the refrigerant arrives as a cold liquid at low pressure and boils. Boiling absorbs a lot of energy (latent heat), and it steals that energy from the food and air around the coil. For a typical domestic refrigerant like R600a (isobutane), the low-pressure side sits below normal atmospheric pressure, roughly 0.5 to 1 bar, so the liquid boils somewhere around −24 °C inside the box. It leaves as a cool gas carrying the stolen heat.
At the compressor, the electric heart of the machine, that gas gets squeezed to high pressure. Compressing a gas also heats it, so the vapour comes out hot, hotter than your kitchen. That “hotter than the room” part is not a bug; it’s the whole point of the next stop.
At the condenser, the warm coils on the back, the hot high-pressure gas meets kitchen-temperature air. Because the gas is now hotter than the room, heat flows out of it naturally, into your kitchen. As it loses heat the gas condenses back to a liquid, releasing everything it picked up inside plus the compressor’s work. For R600a this high side runs around 5 to 8 bar, condensing near +45 °C.
At the expansion valve, a tiny nozzle, the warm high-pressure liquid squirts into the low-pressure side. Pressure plummets, its boiling point plummets with it, part of it flashes to vapour, and the mixture drops back to that bitterly cold evaporator temperature. Then it’s back to stop one, forever.
The one trick that makes it work: pressure sets the boiling point
Here’s the fact the whole machine leans on. Water boils at 100 °C at sea level, but at only about 70 °C on top of a tall mountain, because the air pressure up there is lower. Refrigerants take that to an extreme. R600a boils at −11.7 °C at ordinary atmospheric pressure [1]; drop the pressure below atmospheric and it boils even colder, raise it to several bar and it won’t condense until it’s quite warm.
So the compressor isn’t changing what the fluid is. It’s changing the pressure, and therefore choosing where boiling happens (in the cold box, where you want heat absorbed) and where condensing happens (on the back, where you can dump heat). And it uses a phase change on purpose: boiling and condensing move far more heat per kilogram of fluid than simply warming or cooling that fluid would [3]. That’s why fridges shuttle a boiling-and-condensing refrigerant around instead of just circulating something cold.
The surprisingly good deal
Because the machine moves heat rather than conjuring a temperature difference from scratch, it can shift more heat energy than the electrical energy it draws:
COP = heat removed ÷ electricity used ≈ 2 to 3
100 W of electricity → 200 to 300 W of heat pumped out of the box
That ratio is the coefficient of performance, and for a domestic fridge it sits around two to three [2]. A figure above one looks suspicious only until you remember the electricity is not the source of the heat, merely the fee for relocating it; the heat itself arrives free, out of your food and your kitchen air.
There is, however, a hard ceiling on how good the deal can get, and thermodynamics sets it precisely. For an ideal, perfectly reversible machine the best achievable COP depends on nothing but the two temperatures it works between: COP_max = T_cold ÷ (T_hot − T_cold), with the temperatures measured in kelvin [2]. Feed in the temperatures of the room and the box — a 4 °C interior in a 20 °C kitchen — and that ceiling comes out enormous, around 17. So why does a real fridge manage only two or three? Because it cannot pump heat directly between the box and the room. It has to pump between its own coils, and those deliberately run well past both temperatures so that heat will actually flow across them. Put the evaporator’s −24 °C and the condenser’s +45 °C into the same formula and the ideal ceiling collapses to roughly 3.6. A real fridge landing at 2 to 3 is therefore not being wasteful; it is running close to the best its own operating temperatures permit, with friction, imperfect heat exchange and a non-ideal refrigerant accounting for the modest shortfall. It also explains an everyday intuition made rigorous: the colder you ask the box to get, the wider that temperature gap grows, the lower the ceiling falls, and the more electricity each unit of heat removed costs you. A heat pump warming a house in winter is this same machine with its two sides swapped, bound by the very same ceiling.
The same trick, elsewhere
“Move heat with a phase change” is one of the most reused ideas in your home:
- Sweating. Your body’s built-in evaporator. Water on your skin absorbs your heat to evaporate and carries it away: the same latent-heat trick as the fridge’s coil, just without the compressor to recycle the fluid [3]. It’s also why you feel chilly stepping out of a warm shower into warm air: the water is busy evaporating, on your heat budget.
- A pressure cooker. The same “pressure sets the boiling point” dial, turned the other way. Seal the pot, let steam raise the pressure, and water now boils around 120 °C instead of 100 °C, so dinner cooks faster because the boiling point moved up, exactly the way the fridge moved it down [4].
- A deodorant spray turning cold. Let a compressed gas expand and evaporate suddenly and it soaks up heat: a hand-held, one-shot version of what happens at the fridge’s expansion valve.
- Dehumidifiers. The identical cycle, doing a different job: room air passes over the cold coil, its moisture condenses and drips into the tank, and the heat goes right back into the room.
False friends
Some things cool you without any of this machinery, and get mistaken for it:
- A fan. Cools you, not the room. A fan doesn’t lower the air temperature by a single degree; it speeds up evaporation and convection from your skin. Point it at a wall and you’re actually heating the room slightly with the motor. Nothing is being pumped anywhere.
- A USB mini-fridge (Peltier cooler). No refrigerant, no phase change, no boiling. Thermoelectric elements shove heat across a junction using electric current directly [5]. It’s genuinely a heat pump, but a far less efficient one — which is why your kitchen fridge hums with a compressor instead.
- The camping fridge that runs on a gas flame. An absorption fridge cools with heat as its energy input — refrigerant driven around by a heat source and an absorber fluid, no compressor at all. Cooling a box by burning propane sounds like a contradiction; it’s a different loop entirely, and a rabbit hole for another Tuesday.
- An ice pack. No pump, no cycle. It’s just a parcel of “missing heat” you charged up in the freezer, which means your freezer’s compressor did the real work, earlier, while you weren’t looking.
Fun consequences
| Observation | Why |
|---|---|
| The back of the fridge is warm | That’s your food’s heat, plus the compressor’s work, being dumped into the kitchen. |
| Leaving the door open won’t cool the room | It moves heat from the front to the back and adds the compressor’s work on top. Net effect: the room gets warmer. |
| Putting hot soup in is a bad idea | The pump has to move all that extra heat out, so the compressor runs for ages and everything else warms up while it does. |
| The fridge hums, then goes quiet, then hums | A thermostat runs the compressor only when the inside drifts above the target temperature. |
| Frost forms on the cold coil | The evaporator runs below 0 °C, so humidity from the air freezes onto it. |
| Your air conditioner is the same machine | Evaporator indoors, condenser outdoors. A heat pump is the same machine again, run in reverse. |
So the next time the fridge kicks on with that low hum, picture the conveyor belt starting up: heat being escorted, parcel by parcel, out of the box and into the room, with a little extra added on top to keep the second law satisfied. Nothing is being made. Something is just being moved, at a price physics has already fixed.
Lunch Break Physics runs Tuesdays at noon. Last week: why suction cups stick. Next Tuesday: why the sky is blue (and sunsets red). Got an everyday-physics puzzle you’d like poked at? The comments are open.
Sources
- [1] R-600a (isobutane) refrigerant properties — normal boiling point ≈ −11.7 °C at atmospheric pressure; used in most modern domestic refrigerators.
- [2] Coefficient of performance — domestic refrigerators typically operate around COP 2–3.
- [3] Latent heat and the vapour-compression cycle — phase change moves far more heat than sensible heating of the same fluid.
- [4] Pressure cooking — at roughly 2 atm, water boils near 120 °C.
- [5] Thermoelectric (Peltier) cooling — solid-state heat pumping, typically far lower COP than vapour-compression.
About the Author
Stefan Christoph is a Principal Solutions Architect at AWS, focused on agentic AI, media & entertainment, and helping builders move from demo to production. He writes about AI architecture, developer productivity, and the future of software.
This is a personal blog. Opinions expressed here are my own and do not represent the views or positions of my employer.
❤️ Created with the support of AI (Kiro)