TL;DR? Bouncing off a literally unreadable schizopost? Just read the graph that should be displayed above to see the prices of various heating options (colored lines or dots) compared to heat pumps (somewhere in the blue area). Better units or milder climates move right on the X-axis, shitty units or cold climates move left. The blue area shows the realistic range of costs with electricity between 12 and 24 cents per kilowatt-hour, and poor to very good efficiency. I'm lucky to be at the bottom right where it's cheaper than gas even at last decade's low prices.
Are you in the market for a new heating or cooling system? Or buying or renting new construction that came with a mini-split system? Or have you just heard the endless advertising and propaganda about MINI SPLIT HEAT PUMPS SAVING THE WORLD and wondered if you were being sold a bill of goods?
The New York Times tells us they are The Future. They will Save The Planet while saving you money for being green! The price might seem high, but this listicle of intangible benefits will convince you it’s the right thing to do (or else)!
If you’re as cynical as you should be, one of your hands is now guarding your wallet while the other flips the safety on your Browning. Don’t shoot just yet, I'm only the messenger!
This post is an effort to make some use of the autistic research I did before buying and installing my own, now that I’ve confirmed its performance and the rebate check is safely cashed. It's a bit of a rough draft that I hope can be improved with feedback, but I wanted to get it out as a response to Haroldbkny's question from a few months ago.
We can skip the basics: you’re here, you know how refrigerators and air conditioning move heat using the latent heat of phase transitions between liquid and gas. It’s literally just that but flipped to heat up the inside and cool down the outside instead. Simple as. Let's get right down to the costs.
Coefficient of Performance (COP) and Running Cost
COP is your heat/work ratio. Electric heating elements give you 1 watt of heat per watt of electricity, making it the shittiest way to produce heat short of burning charcoal in your van with the windows taped up. A decent heat pump can move 4 or more watts of heat for every watt of electricity in the right conditions, for an efficiency of 400% or COP=4.
(The commonly stated HSPF efficiency rating is an American measurement that’s basically just estimated seasonal average COP multiplied by the 3.412 BTU (British Thermal Units)/watt ratio.)
Because electricity is almost always more expensive per watt than other fuels, you need this over-100% efficiency to be competitive. See the above chart for average current prices, or play around with cost calculations on this page. (Maybe 10milBTU would be more readable?)
Short version: fuel prices are currently high enough that heat pumps are almost always the cheapest option unless your electricity is very expensive, but this is likely to change in the next decade (right before posting I got an email from my utility announcing a 6% rate increase every year from now on). They are the cheapest option in a mild climate with low electricity costs, but in a harsh climate with high power costs they can be 40% more expensive than gas even at current record gas prices. However, they handily beat oil and propane at any kWh price. Electricity price matters far more than efficiency of the unit or the climate.
Let me know if you’re interested in European prices, which are literally off the scale of this graph (€73/Mcf vs $18/Mcf for gas, with heat pumps’ cost zone well above the cost of US fuel oil).
While this naive cost comparison leaves heat pumps looking very good, especially with CurrentYear fuel prices, there’s a big problem the NYT doesn’t like to mention: both the output and COP of heat pumps drops with air temperature due to physical law.
High end cold climate mini splits are now misleadingly advertised as “maintaining 100% capacity down to -5F,” but this is using rated rather than maximum capacity; it still loses output, it simply started with almost twice what it’s rated for (21kBTU vs 12kBTU, for a Mitsubishi FS12). It also doesn’t mean that the efficiency stays high, and everyone pushing cold climate heat pumps talks about “capacity” in a way that deceptively insinuates that cold temperatures have no effect on the system's efficiency.
Heat pumps are just inverse heat engines that use work to move heat from a cold temperature source (Tcold) to a warm temperature sink (Thot), rather than generating work by moving heat from hot to cold like a steam engine. It takes more work to move heat from a 0F source to a 100F sink than it does from 47F to 100F. Think of it lifting the heat like a bucket of water from a well: the deeper the well, the more work you do pulling the rope. The colder the outdoor air, the more work the heat pump has to do for every ounce of heat.
The theoretical maximum (Carnot) efficiencies for those temperature gradients are COPs of 5.6 and 10.5 respectively, and you can’t do better than that without the thermodynamics police arresting you for building a perpetual motion machine (a heat pump powered by the steam engine it makes heat for). So ignore anyone who says that new heat pumps are "just as good" in cold weather. (Edit: holy fuck fix all three of these paragraphs)
The formula is Carnot COP=Thot/(Thot-Tcold), so efficiency drops as the weather gets colder or the outlet temperature goes up. Thot has to be much higher than room temperature for comfort, airflow limits, or radiator sizing for liquid systems (the radiators on air to water systems get ridiculous to bring the hot side down from the 160F used by normal central heating to the 120F of heat pumps). Still, Thot policing is important for heat pump efficiency: radiant floor heating can go down to 95F if you have money to burn on the install, but on an air unit turning the fan up and keeping coils clean to reduce condenser temperature helps too.
Tcold is the outside air temperature, and this being so variable is the big downside of using an air source rather than a ground source with a stable temperature. The colder it gets the more work is needed for every BTU moved, and at the same time your house needs more BTUs to stay warm.
Here are two top of the line 12kBTU mini splits for comparison: one a regular Panasonic, the other a Mitsubishi cold climate hyperheat. Both have much higher maximum BTU output than their rated capacity: 21k and 18k max vs 12k rated. You can see that both lose maximum capacity at lower temperatures, but the Mitsubishi retains its rated output. Both lose COP as temperatures drop, although the Mitsubishi’s magic Japanese regenerative vapor injection compressor holds on better (but sacrifices efficiency at low output, making it a poor choice for mild climates).
Modern heat pumps already work at almost half of theoretical efficiency, which is pretty good, but they’re not magic.
COP and cost aren’t linear: doubling the COP halves the cost. If resistance heat would cost $100, a COP 2 heat pump will cost $50. Doubling again to COP 4 will give $25, or twice the efficiency gain for half the savings. An impossible COP of 8 would save only another $12.5. Most of the savings are gained early on the curve, which is something to keep in mind weighing the cost/benefit ratio of a more efficient but more expensive unit. This explains the shape of the curve in the pricing graph. The latest Mitsubishi FS Hyperheats maintain a COP above 2 down to -13F, so consider this an absolute floor when comparing running costs.
Note that at a COP below 2.5 heat pumps have little advantage over gas from an energy conservation standpoint: a 40% thermal efficient gas power plant sending power to a 2.6 COP heat pump with 10% transmission loss uses exactly as much gas per BTU as a 95% efficient gas furnace. In cold weather events such simple cycle peaking turbines will be generating the marginal watts for all this new electrical demand. (The math for a 30%ηth german brown coal-burning power plant is left as an exercise for the reader, but needless to say there is a significant toll paid.)
2.6 is around the COP of a Mitsubishi unit at 17F, which is a newsworthy cold snap for me but Tuesday for much of the country. This will probably still work out in your favor due to pricing, but it’s a thumb in the eye for the “ending fossil fuels” justification.
The drop in COP during extreme weather poses a problem for the grid similar to being overwhelmed by summer air conditioning, which I will discuss in the culture war sibling post. Here I’ll just say that it’s a very good reason to have a backup system that doesn’t rely on the grid, because it will cause problems in future. Being slapped with “peak demand” fees on a cold winter night requires an alternative even if the power stays on.
Finally, how often your home is occupied makes or breaks heat pump cost-effectiveness. Their efficiency depends on modulating to the load, and a system that can rapidly ramp up the heat right before you get home would be too large and expensive. If you work from home or are retired, they’re a great choice. If you only come back to your house to eat and sleep for 9hrs a day, they will waste an enormous amount of energy heating your home while you’re away, or working in inefficient turbo mode to raise the temperature when you come home. A furnace will be much more efficient in that use case.
This is almost never mentioned, as most of the people pushing heat pumps are either from the laptop class or rich retirees building custom “zero net energy” mansions. The current move to mandate heat pumps rather than gas for the working poors is likely going to hurt them for no benefit, but what else is new.
Equipment cost, installation cost, and rebates:
So far I’ve only talked about high end $2400+ units, but I bought the cheapest $1100 white label chinesium piece of shit made by Midea that had an inverter and a decent US warranty (22 seer/11.5 hspf 5yr parts). Part of this was a rigorous cost-benefit calculation: the extra $1300 for a Mitsubishi would save only 20 cents a day on average, for a silly 18 year payback time. But the bigger part was that I am a money-grubbing bastard and wanted to make a profit on my power company’s crazy $1500 rebate. That tiny bit of free money gives me a warmer glow than the actual heat output every time I look at it.
If you also live in a very mild climate or are only going to use a mini split’s heating mode for the shoulder season, I’d strongly recommend at least looking at cheaper, non-hyperheat units, which home depot will deliver for free.
If you get one of these after this December, you’ll also be able to get a 30% federal tax credit thanks to the inflation protection act. Right now it’s only $300. Many states and local utilities also have their own rebate programs with various requirements, so check before buying a unit. Some aren’t even worth it, especially if they’re limited to specific models or contractors. The rebate I got was an exceptionally good deal.
If you don’t DIY it and get a contractor to install it for you, expect to pay like $2500-5000 just for them to bash a hole in your wall and connect the indoor and outdoor units. It’s ridiculous and there is no excuse: HVAC in this country is just piracy. If it makes you feel better, at least you’re not in Britain (unless you are, god help your soul), where you’d be paying higher rates for work that looks like it was done by literal monkeys, according to nonsensical gold-plated government standards that were also written by monkeys who have a banana stake in the companies they’re subsidizing.
If you do DIY, you can just pay a few hundred to a single HVAC tech who can check your work, fix any flares you fucked up, do a proper (and vital) nitrogen purge and vac on the line set, and most importantly sign for the warranty and any rebates. I wouldn’t advise doing this part on your own with a Harbor Freight vacuum pump like some youtubers, as a fuckup will leave you both warrantyless and with many hours of work to be done by a professional who will charge you a punitive DIY-cleanup rate. There are "fully DIY" units from Mr Cool, but they are leak-prone (designed for temporary field use in Afghanistan, IIRC), older models, and the extra cost is about what a HVAC tech would charge anyway.
If you live in a mild climate and have cheap electricity, mini splits are a good buy, pretty much period.
If you have no access to gas, mini splits can be a good alternative no matter your climate. Pay attention to any peak demand surcharges your power company has, and expect them to be introduced soon if they don’t.
If you have a garage or workshop heated with electric resistance that you would like to get AC for, consider getting a mini split for it instead. Heating for less than half the cost will quickly pay for itself. There are even one-piece window units now with no install required.
If you have a long shoulder season before it gets seriously cold, mini splits can handle it incredibly cheaply and delay lighting up the furnace. Worthwhile if you do a cheap install and/or needed to add or replace AC anyway, or to hedge against fluctuating gas prices.
Always have a backup, ideally something like a fireplace, boiler, or something else with locally stored fuel. Keep old oil furnaces if you can, even if it means giving up on an enticing “fuel-switching” rebate (or fudging the application...). Propane works too, especially if you also have an emergency generator running off it.
If you’re thinking about giving up your gas connection, probably don’t unless the base charges are unsustainable. Eventually you won’t be able to get a new one, the electricity/gas price ratio will climb again soon, and I expect gas to be better managed and more reliable than the electrical grid in future.
If you’re paying over 25 cents a kWh and have access to gas at any price, for god’s sake don’t even think about it. You have no idea how many angry north-easterners complain about their bills doubling after paying $30,000 for heat pump installs because they trusted the New York Times and didn’t think their green indulgence would have ongoing costs.
Try not to go with multi-splits that have multiple heads per outdoor unit. They are notoriously shit. Seriously, one head in the living room does every room in my house, so don’t go overboard or let an HVAC contractor go nuts putting one in every closet for $3k each.
Many people repeat a mantra that mini splits “require well-insulated houses.” This is nonsense: a BTU is a BTU no matter what makes the heat. The Japanese were the first users of mini splits, and they think well-insulated means having a second layer of rice paper on the wall. The only real benefit is flattening the heating curve to reduce the range of loads the unit needs to work at. Some insulation is good, but it suffers the same diminishing returns as COP: twice as much for half the benefit with each doubling.
Also, no heat pump made recently has an electric heat strip for backup. They simply aren't used any more.
If you’re the least bit handy and drilling a hole in your wall isn’t terrifying, consider doing a DIY to save money. We can talk about it if anyone’s interested. Rest assured: if a lazy retard like me can do it, anyone can.
If you’re in a cold climate with no access to heating at a reasonable price, consider getting a shovel and digging a ground source heat pump trench; at least the exercise will warm you up, and water-to-water units are mono-blocs that don’t require refrigerant pipe connections. Downside: the guy you need to hire for hookups is worse than an HVAC tech. He is, may Allah forgive me for uttering the word, a plumber. Pump overhead is not factored into COP, and water-to-water units are not as heavily reviewed, so be careful.
Get a surge protector for all this stuff and hook it up yourself (and double check your panel and house grounding: I didn't realize how shit mine was). This shit’s expensive and sensitive to large surges. Take good care of the equipment generally. The indoor unit getting dirty and restricting airflow lets Thots get out of hand, and you saw how badly that affects efficiency in the earlier graph. Keep your filters and the outdoor coils clean.
Anything you can cost-effectively do to make yourself less reliant on the electrical grid is probably a good idea at this point. Shit’s going to get crazy, and you should expect electricity prices to rise and reliability to fall significantly to where the current overwhelming advantage of heat pumps vanishes.
I haven’t discussed solar, but it is a realistic opportunity for people in southern states. Inverters are an option, but there are mini splits that work directly from DC on the market (it just skips part of the AC-DC-3phaseAC conversion, IIUC) It requires either a battery bank or thermal storage, so wouldn’t be cheap, but there are people doing it in the Arizona desert, and 6-9kBTU mini splits are common RV retrofits.
If you live in the UK or Germany, uh… I’m sorry. If I was living there and couldn’t afford a plane ticket out, I’d get some of those charcoal burners and maybe tape for the windows.
If you’d like to hear more, head on over to the companion Culture War thread post (coming soon) where I can
rip and tear until- calmly and rationally discuss possible disadvantageous trade-offs being made in the area of energy policy regarding mandatory electrification while neglecting the consequent demand growth.
If there's any interest in additional detail about the technical side, and where the technology could go from here, I'd be happy to do a followup post with all the stuff I had to cut for clarity/sanity. I didn't even get to post a single Temperature-Enthalpy diagram in this one :(