Everyone’s talking about heat pumps these days as a better way to warm your home. They even pull double duty, replacing your air conditioner in summer. The great thing is that they’re generally more energy-efficient than old-school heaters, and they don’t burn any fuel, so replacing that old gas furnace can really shrink your carbon footprint.
But what the heck is a heat pump, and how does it work? If I tell you that it transfers heat from outside on a chilly winter day to keep you toasty, you’re probably going to have even more questions—like, how can you use the cold outside to increase the temperature inside? Don’t worry, I’ll explain. There’s some cool physics involved here, so let’s get started.
Temperature Is Weird
What’s the difference between warm air and cool air? In fact, what does “temperature” even mean? It’s actually stranger than you think. At a basic level, if you increase the temperature of air, you increase the kinetic energy of the air molecules. That’s the energy an object has due to its motion. So the molecules in 75-degree air are moving faster than the molecules in 40-degree air.
This means that if you want to increase the temperature of air, you need to have some type of energy input. Really, it’s all about energy.
Now, you might reasonably think a “heat pump” pumps heat around, like a water pump moves water. But heat isn’t really a thing; it’s an energy transfer from one object to another due to a temperature difference. To be honest, even physicists misuse the term sometimes as a shorthand, but properly speaking, an object can’t “have heat” or “lose heat.” Think of it as a verb, not a noun.
Just like kinetic energy is a measure of the motion of an object, thermal energy is the energy it has due to its temperature. The amount of thermal energy also depends on the object’s mass and its “specific heat capacity.” That’s a measure of how much energy is needed to increase the temperature of the object. Metal has a low specific heat, so it heats up fast. That’s why frying pans often have wooden or plastic handles.
And here’s a crucial fact: According to the second law of thermodynamics, thermal energy only transfers from hot things to cold things; it doesn’t go the other way. So despite what you might think, the ice cubes in your drink aren’t “cooling” the liquid; rather the liquid is heating the ice cubes—which means it’s losing energy and getting colder, while the ice gains energy and melts. Bringing the Heat
OK, when you have two objects at different temperatures and place them in contact, there is a thermal interaction. If you set a cup of coffee on your desk, both the cup and the desk will change temperature. In this interaction, the coffee loses thermal energy and the table (and surrounding air) gains it. This interaction will continue until everything reaches the same temperature. Blah. Lukewarm coffee.
But we can use this idea to increase the temperature of air in a house—it’s pretty simple. Just go outside on a warm summer day and grab a big rock that’s been sitting in the sun. (A small rock won’t work—it’s low mass means it doesn’t have much thermal energy—so remember to lift with your legs and not your back!) When you bring this hot rock inside, there will be a thermal interaction with the cooler air, and the air temperature will rise. Boom. You just made a heater.
Wait. This requires the outside temperature to be warm. What if you want to heat your house in the winter when it’s cold outside? Well, it’s still not that difficult. There are lots of ways to increase the temperature of something other than putting it next to something hotter. If you rub two pieces of wood together, they both get hot. Or how about this homemade electric heater:
A battery with copper wire attached to both ends and a piece of blue tape over its surface Photograph: Rhett Allain
Yes, it’s just a wire connected to a battery. When electric current runs through the wire, the wire gets warm. Now you have an object that is warmer than the air and can increase the temperature in your house. This is how an electric heater works.
Or you could start a fire. When you burn stuff, there’s a chemical reaction between the material and oxygen in the air, releasing energy that can be used to increase air temperature. That’s just like a gas furnace. This type of heating is simple. Unfortunately, when you’re burning carbon-based materials like wood or natural gas, it also produces carbon dioxide.
Making Things Colder
OK, now let’s cool the air in a house. Yes, you could get a cold rock to drop in your room, but it’s probably hard to find a cold rock on a hot summer day. Now we have a problem. There’s no simple way to make things colder like there is to make things warmer.
But check this out. Get a rubber band (the bigger the better), and hold it between your fingers like this: Two hands stretching a rubber band Photograph: Rhett Allain
Now stretch it really hard and quickly hold it to your upper lip, which is sensitive to temperature. You will feel that it’s warmer than it was before. That’s because you’re adding energy to the rubber band, which increases its temperature.
Are you ready for the awesome part? Keep it stretched for a little while until it returns to room temperature. Now let the rubber band relax and quickly touch it to your lip again. It’s now colder than room temperature! Seriously, try this for yourself.
So if you had a big enough rubber band, could you use this to cool your house? Wait a minute, you’re gonna say: In the first stage, when we stretched the rubber band, it got hot, and then it cooled back to its original temperature—and in doing that it heated the air. You’re right. But what if we could vent that warmer air outside? Then you could keep just the cooling phase inside.
Boom. You just re-invented the air conditioner! Instead of a rubber band, an AC has a fluid called a refrigerant that circulates in a closed loop from inside to outside. This fluid has a low specific heat, so it changes temperature quickly, and a very low boiling point—turning into a gas at something like –15 Fahrenheit.
How’s it work? The gas is first compressed, causing it to heat up to like 150 degrees. The hot gas circulates in a set of copper coils outside, with a fan blowing over them, so the gas loses thermal energy to the atmosphere. (Copper also has a low specific heat.)
Then it’s pumped back inside, where the pressure is quickly reduced, causing it to expand and instantly cool down to around 40 degrees. As the now cold fluid circulates through indoor coils, a fan blows warm inside air over it, heating the fluid again and cooling the indoor air in the process. As the system circulates, it basically picks up thermal energy indoors and carries it outdoors.
By the way, this is exactly the same process that your fridge uses to keep your cheese and soda cold. In both cases, the process makes something inside cooler and something outside warmer. Put your hand behind the fridge and you’ll see what I mean. Just for kicks, here’s a guy who actually built a refrigerator that runs on rubber bands.
So Heat Pumps Aren’t New!
You thought this was going to be an article about heat pumps, right? Well guess what. We’ve been talking about heat pumps this whole time, because they run on the same principles. A heat pump cools your home just like an air conditioner, by circulating a refrigerant and varying the pressure to change its temperature, so it takes thermal energy from one place and puts it in a different place.
So back to the big mystery: How can a heat pump increase the temperature of indoor air on a cold day without actually generating any heat? Simple: Just run it in reverse! This time we let the hot compressed refrigerant cool off inside the house to raise the indoor air temperature. The low-pressure, cold gas then goes outside to warm up.
Warm up outside? Yep. Even on a freezing day, the air still has thermal energy. So long as it’s above absolute zero—which, believe me, it is, since that’s around –460 Fahrenheit—the air molecules are in motion. And since we’re cooling the refrigerant to, say, –15 degrees, which is lower than winter temperatures in most places, it will wring thermal energy out of even frigid air.
Of course, you can’t get energy for free. Heat pumps rely on electricity to drive the compressor and fans. But if you have solar panels at home, or if the electricity in your area is even partly from non-carbon sources, replacing a gas furnace with a heat pump can make a big difference in reducing greenhouse gas emissions. And it’ll probably lower your utility bills in the process.
Technology Connections has a series of videos that wonderfully explain heat pumps. https://www.youtube.com/watch?v=7J52mDjZzto
Those are great!
This is why I bought a heat pump clothes dryer. Such a great set of videos.
I’ve been looking and they don’t seem great yet, unless it’s just for 1 or 2 people. Small loads and they take a long long time.
It works fine for us, a larger than average American family. It takes basically the same time s load would to go through both a wash and a dry cycle.
which model do you have? 3 different stores told me to avoid it, with my family of 6.
GE PFVQ97HSPV0DS - the combo one. There’s six of us, 2 adults 4 kids. You basically just can’t do a “laundry day.” We’re newly on a septic system too, so we shouldn’t be doing that anyway from what I understand.
We are basically doing at least a load of laundry a day. We’ve had it since November, if you pack it more than halfway full we’ve found that running it on a timed dry cycle works better than sensor dry.
My wife hates the sensor dry cycle regardless because when stuff comes out it feels “hot and damp” but by the time it hits room temperature it just feels clean and dry to me.
You have to clean the lint trap every cycle (I didn’t know people didn’t do that anyway) and there’s a black filter that has to be washed out every 30 or so cycles, which is a few minutes, it’s not like you have to dry it out again.
Honestly, I’ll probably line dry blankets and stuff when it warms up a bit, but I did that with our old resistive one too.
Either way, this thing works just as well as any other washer and dryer I’ve used, you just have to make sure to keep the filters and seals clean so it drains and dries properly.
Wow thanks, maybe I’ll take another look.
Heat pumps are really simple tech, it’s in your fridge, freezer, dryer, car, etc. They move heat from a to b by compressing stuff.
yes that costs energy, but they can easily move 8x more than they consume
More like 4x, not 8x, but that’s still very good, especially compared to normal/resistive heating.
If I were to try to explain it to my 9 year old in the shortest way possible:
“It’s the exact same thing as an air conditioner but you play an uno reverse card on it.”
it’s incredibly neat how transferring thermal energy takes less energy than creating it.
imagine if we figured out causality violating faster than light travel because “transferring” velocity somehow took less energy than generating it. that would be fucking wild
and yet somehow that crazy bullshit really works in real life with heat.
The thermal energy that is being transferred was already created with an energy expenditure somewhere else (e.g. solar fusion), and now it’s just moving around. This isn’t true just for heat, it’s true for all energy. Kinetic energy does transfer between objects as momentum just as easily as between air molecules as heat; the little “Newton’s cradle” desktop balls that go back and forth are demonstrating this transfer of kinetic energy. The energy loss with momentum transfer is due to not being in a perfect vacuum (via drag/resistance, heat, sound), and the elasticity of the materials involved.
Mathematically, 2 perfectly friction-less, perfectly-rigid objects in a perfect vacuum, would not see any energy loss when transferring kinetic energy (ignoring the question of getting that energy transfer to occur between friction-less bodies). If they’re the same mass, the velocity will even be the exact same. Momentum transfer in that case has no “cost”/ energy loss, whereas there is always a cost to generate kinetic energy.
thank you for expounding beautifully upon something i was only wondering about idly to make this whole exchange delightfully educational :3
It’s really just a good example of how much energy phase changes take, all you’re really doing is expending work to make your refrigerant condense and evaporate where you want it to.
Heatpipes also exploit phase changes, just passively, it’s why they move so much energy.
because “transferring” velocity somehow took less energy than generating it. that would be fucking wild
Obviously not the FTL part but elastic collisions are definitely a thing. They transfer velocity with essentially no loss and take way less energy than it would to generate the same amount of force required to provide the same velocity.
… honestly i was just having a silly thought at the time but what i just realized a few minutes ago after being reminded of my post again was that basically all of our motility is velocity transfer because we achieve motion through pushing on other things. We have never generated velocity out of nothing, We’ve only ever transferred it in the first place.
“generating velocity instead of transferring it” would be entirely novel, i realize: it would constitute an essentially reactionless drive. Those don’t exist in real life as far as we know.
Presently, instead, if we want to go ←that←way←, we have to push something else really hard →that→way→.
My biggest problem with heat pumps right now is that the one contractor that will give me an estimate is about 3x above the national average for a new system installation, and I can’t get anyone else to come out and talk to me in the first place. WTF, guys? (Seriously, if the components for a 3 ton system from Mitsubishi Trane is retailing at about $9k, there’s no fucking reason to price installation of a new system at $28k.)
Every heat pump install is unique, so that cost could include weatherization, electric panel box upgrade needs, multiple heads, etc. Did you do an energy audit? Contractors jack up prices when they have full calendars, especially in shoulder seasons (transitioning in Spring or Fall).
They said that I needed to do electrical before they came in for the installation–depending on where I wanted the exterior unit–and I noted that I needed to have an energy audit and get the insulation fixed before I had a heat pump installed. That def. was not on their estimate. (For reference, it’s a 1500 ft^2 living space log cabin, and was built before there were picky little things in my county like, “residential building codes”. So it’s drafty.)
I’m trying to schedule a 2nd contractor right now. I know that I need to get multiple estimates before I open my wallet. The estimate I have doesn’t give the exact model number for any of the heat pumps they’re referencing, or the air handler, so I can’t look up the BTU rating to see if it’s in-line with very, very rough estimates of my heating needs. It lists the tonnage and the SEER/EER/HSPF ratings, but nothing past that.
The person that came out initially said that he though I’d be looking at an installation of around $9000 total, ignoring any tax credits or state rebates. His good/better/best low estimate, with credits and rebates, was a little over $13k, or about 145% of his ballpark number. The high end was 215% of his ballpark with rebates, etc., and 267% without.
Just for reference, I was tangentially involved in a heat pump focused gig for a few years, so I know a bit (relevant to NY) but my expertise is limited, and I’m not a contractor or tech. It’s so hard to give an estimate without the energy audit and a blower door test at the least because these values are put into software that installers use to right-size your system. And the audit will help narrow down the model types you should consider.
You’ve identified you need an electrical upgrade - if you’re running on a 100A panel, you may need 150A to accommodate the heat pump and any other electric appliances. You also identified that you need weatherization. That’s a huge part of preparing for a heat pump because drafty thermal bridges will ensure you’re paying more on your electric bill than you should. You can always buy weather stripping/caulk and DIY by finding the drafty areas and sealing with caulking or stripping, but without a FLIR camera, it’s hard to see all of the places you need insulation.
You are right to get multiple quotes. Not all contractors are created equally. Some certifications relevant to the work are Building Performance Institute (BPI) and EPA 608 (refrigerants). Depending on your state, there may be other recommendations.
There are only two companies that I can track down through energy.gov that are fully certified and able to do energy audits in my general area. I expect that I’m going to pay a lot to figure out just how leaky my house it… :'( Still, I really want to reduce my carbon footprint, and heating with propane is expensive when your house is leaky, even when you live in a nominally warm part of the country. A wood burning stove is enough most of the time, but still, carbon footprint.
I think that the issue with the electrical panel is that it’s full, and there’s not a great place to put the breaker. It should have sufficient amperage. There was a nasty hot tub that used to be on the porch when we moved in that’s gone; that breaker is probably sufficient for a heat pump and air handler, but the wiring would need to be re-routed.
Even with FLIR, it’s really hard to see the places that need to be sealed. I’ve rented one–back when my local Home Depot had one to rent–and there are just a ton of small nooks and crannies that it’s hard to get a read on. My guess is that it’s best to do that when there’s a really sharp temperature differential, like a cold winter day. I’m pretty sure I’m going to have to tear open the walls on the upper story to fill the voids with 2" foam board, and seal with spray foam; I looked into spray foam, and without sheathing on the upper story, it would make replacing the siding an enormous pain in the ass. Plus, it would not allow air circulation behind the clapboards, which leads to premature failure.
My state is deep red, so I’d be shocked if there was much in the way of specialty certifications beyond what’s necessary at the federal level. They’ve been talking about cash rebates on retrofits for a number of years now, but that still doesn’t seem to have actually happened yet.
Sounds like you’re well versed in navigating the switch. I’m bummed to hear some deep red states are dragging their feet on rolling out rebates. It’s up to the state agencies that manage/regulate energy to create rebate or incentive programs. Here’s hoping they get to it soon where you’re at. Good luck, friend!
Love this write up! Heat pumps and geothermal are here to stay. Weatherization and air sealing makes it a dream, and paired with solar panels, so cost effective (after the investment). I can’t wait to have a home to go full electrification! Probably need a panel box upgrade though.
The problem with heat pumps is that below 10 degrees you need supplemental heat unless you go with ground loop systems that are very expensive. Many people probably supplement with electric but that seems crazy. Can the grid hold up to everyone doing electric heat on those minus 30 days?
Love the idea of heat pumps but not to sure about getting rid of my gas furnace unless I went ground loop.
Not sure which model you looked at but my Panasonic 12000 BTU heat pump I installed back in 2020 can easily works down to -25c. My electric convectors haven’t been working since then and I live in Canada and it can get pretty cold usually.
The latest models are even more efficient.
My house requires about 40000btu at minus -30F. Not sure I have even seen a brand that will run below -24F and the efficiency really drops off anyway. Around here it seems that the design point is about +15F and after that supplemental heating is needed.
Edit: My current furnace is 66000 BTU/HR but I think it is way oversized.
By the way what brands are worth a look? I have looked at some and used a design tool but was underwhelmed.
If you’re in a place that experiences 4 seasons, look for a “cold climate heat pump.”
I have looked. That they will run down to -10F or a bit more does not mean they move much heat at those temperatures.
I don’t think the band matters that much. The spec sheet is much more important than the brand imo, the efficiency at low temperature in particular.
Another thing: having good insulation is really important to fully appreciate heat pumps. My house is a new construction and very well insulated and that’s why I could get away with such a small heat pump.
Yes. My house is 1948. It is well insulated for a house of that age. We had it done to the max 15 years ago. You would have to totally shell the house to do much more.
Only real opportunities we have is new Windows, a new front door, and think more about basement losses. We do not have a very clear thermal envelope in the basement direction.
I lived in Montana, and my air source was fine for -38f winters. I didn’t even realize it had a resistive heat mode until I was nearly moving out years later. God, I miss $60 summer electricity bills while running the AC in summer. Winter was like, $130. When we finally get off rentals, we want a house and a heatpump again. It felt like a physics glitch more than a machine, lol.
The first part of the article triggers me. Heat in the physical sense is thermal energy. Like with other forms of energy, you need an energy difference to actually have it perform work, or you need to invest work to create an energy difference (in a heat engine or a heat pump, respectively). Just like you would letting a weight fall to the ground and lifting it back up. And cooling is removing heat, so ice cubes are actually cooling your drink.
In a pan, low specific heat capacity is not that desirable. That’s why people use big honking chunks of cast iron to prepare food: so adding the cold food doesn’t lower the temperature too much. But the metal also gets you good heat conductivity to quickly get the heat from the stovetop to where it’s needed.
Conversely the handle is made from materials that have low conductivity so heat gets conducted more slowly towards your skin. The higher capacity helps but isn’t that crucial: air has fairly low heat capacity but you can stick your hand into an oven at 100C without getting burned. Unlike boiling water, which has quite a high heat capacity.
The refrigerant should have a high heat capacity to move as much heat as possible for a given temperature difference. Most systems employ a liquid-gas phase change somewhere in the cycle to transfer even more heat energy in the form of latent heat. R134a, a common refrigerant, has a heat capacity about 3/4 of that of water.
One more thing: even if the electrical energy is completely nonrenewable, heat pumps still offer an environmental advantage. Gas power plants are fairly efficient, around 40% of the extractable heat energy gets converted to electricity. With a COP of 2.5, a heat pump would produce as much output as burning the gas in a perfectly efficient furnace. If the COP is larger, the heat pump is more efficient than burning the gas directly, and modern heat pumps usually exceed 2.5 except in the coldest days of winter. Add to that the existence of dual-cycle power plants with 60% efficiency, and the losses of a conventional furnace, and heat pumps may win even on days where the COP is slightly less than 2.