It Sounds Easy Enough --- Heat Your House with a Backwards Air Conditioner
But as with a lot of Decarbonization Strategies, the Devil is in the Details
By Matthew L. Wald
The “electrify everything” movement is coming out of the garage and targeting the basement, aiming for the home heating system. This might be a good idea, but it’s hard for a consumer to tell. And the government incentives can sometimes seem poorly thought-out.
Heat pumps to replace fossil boilers are a keystone of some state decarbonization strategies, but success relies on something that isn’t happening yet: The addition of vast amounts of 24/7 zero-carbon energy to the electric grid. Much of the progress so far comes from substituting gas for coal, which is helpful in the short term but can’t take us to zero. We can’t fix our energy and climate problems only on the downstream end; it will take careful attention to where the electricity comes from.
New technology, however, is changing the math of home heating, in a way that might lower costs and environmental impact.
Unfortunately, the math is exceptionally hard to follow for someone shopping for a replacement heating system. (For a new system, it would be easier, but most home heating systems go into houses that have been around for years, or decades.) The buyer is making a bet on the future costs of fossil fuel and of electricity, and even the future climate. If you like to gamble, you can get a shorter term, lower stakes commitment in Las Vegas.
Re-writing the Economics of Home Heating
The big change is to overturn an axiom of energy, which is that a kilowatt-hour of electricity supplies a fixed amount of heat when it is used, 3,413 BTUs. No matter what kind of space heater, electric radiator or baseboard heater you buy, and no matter what the guy at the hardware store tells you, that’s how much heat it will give you from one kilowatt-hour.
But heat pumps throw that figure out the window. A heat pump resembles an air conditioner. It pulls in air and chills it, and sends the heat out in the other direction. But with a heat pump, the side that spits out the heat can be inside the house, not outside. More realistically, consumers can buy a device that is reversible, so it will pull heat out of a room in the summer and send it outside, and in winter it will do the opposite.
If you like your living room at 68 degrees, a heat pump can take in air from outside at 60 degrees, or 50, or even at 30 degrees, pull the heat out of it, and then return the air to the outdoors even colder. In effect you are chilling the great outdoors, to heat your home.
But it works the other way too. A heat pump is a little like a car with an engine that turns in one direction, but gears that allow it to go backwards or forwards. Thus with the flip of a valve, it can cool the inside and heat the outside, providing air conditioning, which is in increasing use around the country.
Why bother? Because there are tens of millions of houses with boiler exhausts, using natural gas, oil or propane, and we’ll never be able to control their carbon emissions; the solution is replacement. There are other options for multi-family apartment buildings, but most Americans live in single-family houses.
Heat pumps have been around for a while, but until recently, they weren’t feasible as a replacement in most parts of North America because they didn’t work well below about 30 degrees. Now, some are certified to run below zero degrees Fahrenheit.
The easiest version to install is called an air-source heat pump, because it’s pulling warmth out of air. A variant, highly effective but more complicated to install, is a ground-source heat pump, which circulates water in pipes through the ground. The water comes to the surface at somewhere between 40 degrees and 70 degrees, depending on where you are. In summer, the heat pump can pull heat out of the house and put it in the cool water, which is then pumped back into the ground to give off its heat to the soil. In winter, it makes cool water even colder, and sends the heat to the interior of the house.
The hard part is burying the pipes.
But an air-source heat pump can be installed as easily as a simple whole-house air conditioner. And according to the Department of Energy, using a fixed amount of electricity, it can deliver up to three times as much energy as would result from running the electricity through a space heater.
The Energy Department cites a study that found that if you are replacing resistance heat, the heat pump can save you 3,000 kilowatt-hours over a winter. In most of the country, a kilowatt-hour retails for twelve or 13 cents, so you could save in the range of $350 to $400 per winter.
Perhaps more importantly, in some places it is heating that produces the highest level of demand for a year. Cutting peak demand means the utility needs fewer generators and less transmission and distribution, cutting costs for everybody.
But is this a good deal for consumers?
Here’s the Math, but Some of it is Fuzzy
Here is a non-random case study: My cabin in the woods in upstate New York. It’s heated with propane, delivered a few times a year by a tanker truck.
The furnace that warms the air has an efficiency of roughly 90 percent, meaning that 90 percent of the energy content ends up heating the home, and 10 percent goes up the stack. A gallon of propane has 91,500 BTUs, so a top-notch furnace would deliver about 82,350 BTUs. If propane sells for $2.25 a gallon at retail, then 1 million BTU (A standard industry measure) delivered to inside the house would cost about $27.32.
I pay about 13 cents for a kilowatt-hour. With electric resistance heat, I would need about 293 KWH to make that much heat, so it would cost me about $38 per million BTU. (In contrast, the space heater or baseboard electric heater is 100 percent efficient. Even if it gives off a tiny bit of energy as a red glow, that, too, converts to heat.) But with a heat pump, if weather conditions are favorable (i.e., temperatures in the 50s or the low 60s), I could get the same heat for somewhere in the neighborhood of $12 or $13.
The sales pitch for a heat pump (and in some parts of the country, the sales pitch seems to be everywhere) is that the homeowner is getting off fossil fuels. But to produce that electricity for me, somebody somewhere is going to burn extra natural gas. At the wholesale level, which is the way power plants buy their gas, the price lately has been between $3 and $6 per million BTU, not the $27 or so that I am paying for delivered heat at my house. But with electricity from a utility company, between the generator and the transmission system, between one third and one half of the energy value would be lost. In carbon terms, the electric system is going to produce a little less than 1 pound of carbon dioxide per kilowatt-hour. If I need 42.2 million BTU from resistance heat, at 293 KWH per million BTU, that’s about 12,300 KWH a year, or around 12,000 pounds a year for my heating needs. If it’s through a heat pump, I could get by on perhaps half or one-third that amount, say 4,000 to 6,000 pounds of carbon dioxide. The 725 gallons of propane I’m using would produce about 12.65 pounds CO2 per gallon, or about 7,300 pounds.
(As they say in the car ads, your actual mileage may vary. In this analysis, I’m ignoring the propane I use for heating tap water, and using the stove, but those are small.)
Right now, the heat pump business is hot. One recent evening I googled “heat pump” and up popped a form asking for my address and phone number, to get a price quote. I filled it out and a contractor called me in less than five minutes; while I was on the line with him, a competitor called.
One reason for the enthusiasm is state and federal tax breaks and incentive payments. Some of these may be structured in a counterproductive way. New York State provides double the incentive if you rip out the fossil fuel system. One problem with this radical approach is that when the outdoor temperature is cold enough, then burning natural gas, propane or oil is more efficient than using electricity (which will probably come from natural gas burned elsewhere). And in rural areas, prolonged power outages are an anticipated risk. Lots of house-owners find it prudent to keep a back-up generator, even if the machine is dirtier and more expensive than grid power. But heat pumps don’t get along well with back-up generators because the electric current the generators produce tends to have an unsteady tempo, and that can destroy the heat pump’s circuit board. So keeping the old oil- or gas-fired furnace is a good idea.
The tax benefits are complicated enough that one contractor I spoke to, who seemed to know his business very well, told me I should talk to whoever prepares my taxes to figure out what my after-tax cost would actually be.
The Power of Inertia
And the contractors may be chasing a limited market. Replacing my air conditioner with a heat pump would save me hundreds of dollars per year, and if the price of propane goes high enough, maybe well over $1,000 per year, but the installation would be in the range of $20,000 to $30,000. This would make perfect sense if my air conditioner needed replacement, but for now, it doesn’t. (It might be a good idea, because a top-of-the-line air conditioner in 2024 is about 40 percent more efficient than the six-year-old one I have now, but like most consumers, I’ll probably delay replacement until the current one needs an expensive repair. And if I replace it with a heat pump, or just with a newer air conditioner, I will be installing something that will cool the house with fewer kilowatt-hours, because newer equipment is more efficient.)
Another complication: heat pumps can function as the source of warmth in a house that has a forced hot-air system, because the output of the heat pump, around 114 degrees F, is high enough to heat the air. But in a house with radiant heat—that is, pipes under the floor that carry hot water—the heat pump’s temperature is too low, the contractors tell me.
It would have been more sensible to build the house with a heat pump in the first place. But this was not possible. The house was built by a smart, capable couple; they even wired the garage for 220 volts, in anticipation of an electric car. But when they planned the house, about six years ago, heat pumps capable of handling an upstate New York winter were not generally available.
So the inertia of the built environment may triumph, at least until the equipment I have now wears out. That’s generally the way efficiency gains take hold. With lightbulbs, the time is counted in months, but with big ticket items like heating systems, it’s likely to be more than a decade.
If I’m considering the leap to a heat pump, figuring out the payback period is tough. The system’s electricity consumption is weather-dependent, and it’s hard to pin down in advance what the efficiency of the system will be. The efficiency ratings provided by the Department of Energy are good for comparing one heat pump to another, but aren’t as helpful in telling a consumer in one specific location just what the operating costs will be.
One technician, who worked for a big firm and came into the house with a tablet that he used to scan all the walls and windows and a sophisticated system for measuring its heating and cooling needs, said that yes, a heat pump would save money, but no, he couldn’t say how much. The heat pump would work best when winter temperatures were mild, and would have to consume more and more electricity to maintain a comfortable inside temperature as the outside temperature fell, but, he said, “no one has dared to produce a kilowatt-hour chart,” to describe heat pump electricity consumption measured against outdoor temperature.
In the old days, before heat pumps, the choice was simpler; when your furnace died, you compared the efficiency of the replacement models available, figured the fuel savings in gallons or therms, and multiplied times the price to figure out how much you would save, and whether the higher cost of a high-efficiency unit would be justified. Some people switched fuels, but slowly. The oil shocks of the 1970s took decades to play out in the home heating market, as homeowners slowly switched from oil to gas.
Getting off oil, which in those days was more of a price and national security imperative than a climate strategy, was generally only possible if your house had gas service. Now, people who are still on oil and far from a gas pipeline can opt for heat pumps run by electricity, and the calculation is easier, especially since heating oil, like diesel fuel, became so expensive.
But for many people the default setting is to make no decision, at least until the old heating system dies. This is a struggle fought in one basement at a time.
But the “electrify everything” movement does point to another imperative: no matter what the cost of switching from a fossil-fired system to an electric heat pump, or from a gasoline car to an electric one, a big factor is the carbon intensity of the electric system itself. A variety of new nuclear reactors, some big workhorses and some smaller, flexible models that can pair well with variable generators like wind and solar, would make every investment in electric end-use equipment more valuable. This goes for everything from home heating to municipal buses to electric mills that recycle steel.
Cleaner electricity makes heat pumps a better deal for the environment. Whether it’s better for the homeowner’s checkbook is a little harder to figure. But cheap, abundant electricity from a new generation of nuclear reactors would help that calculation, too.
Setting goals is easy, but the trip to zero emissions will be a long, hard slog.
A few quibbles from a non-MIT grad (only a Professional Engineer with ~40 years experience in all types of power generation and transmission technologies) - with some engineering thoughts so that policy makers can make informed decisions to achieve their intended objectives):
1) Wind and Solar are NOT "variable" energy resources, which implies that operators can take action to vary their output up or down as desired to cooperate with other generation resources. They are "Intermittent" generation resources, whose output varies as a result of environmental conditions such as wind speed & direction, air qualities such as humidity and particulates, and the time of day - including the daily phenomenon called "night".
2) Yes, energy generators "cooperate", they do not "compete", to ensure that (A) total energy into the Grid = total energy out of the Grid Instantaneously during each of the 8,760 hours in a year (the design cycle of an electric power Grid) and (B) that the instantaneous energy in the Grid is equal to the total instantaneous amount of energy rate payers want to use from the Grid, or Demand, during every hour over those 8,760 hours.
3) Because Grid operators cannot depend on Wind and Solar - cannot send them dispatch orders to produce "Y" amount of energy at "X" point in time, they must read the Demand then subtract the amount of energy being produced by intermittent generation (which therefore act like "negative loads") and send dispatch orders to other generators to change their output to accommodate wind and solar and keep the Grid Balanced. If Wind and Solar aren't generating as projected, they must be backed up by Dispatchable power on standby.
4) These "integration costs" should be borne by wind and solar because they are directly responsible for these costs. In other words, this integration generation is required to maintain Grid stability and would not be required if Grid operators had planned to use only Dispatchable generation - especially the amount of Back up generation.
5) The carbon emissions associated with these integration costs should also be allocated back to these wind and solar generation facilities.
6) After making the adjustments for costs and carbon emissions in wind and solar noted in (4) and (5), one finds their costs now ranges between $200-400/MWh and carbon emissions are ~50% the CO2/kWh of natural gas - far from inexpensive and carbon-free, yet amazingly aligned with the overall system costs of the California and South Australia Grids, as well as their flat or only slightly declining Carbon Intensities over the last 5 years.
7) Based on (6), one wonders what the justification is for continuing the madcap deployment of wind and solar in lieu of a sustained push for both Large (Gen III+) and Small (SMRs and Advanced Reactors) Nuclear given the goal of deploying electric power Grids that are able to provide:
1 - A sufficient Amount of
2 - Reliable
3 - Resilient
4 - Affordable
5 - Safe, and
6 - Sustainable (to address the changing climate)
Electricity for everyone as fast as possible.
FYI: Criteria 1-5 fall under the statutory mandate of FERC. The 6th item, Sustainability, is not addressed by statute, but has been de facto added to the listed via policies and rules,
A few quibbles from a retired energy policy wonk and MIT grad:
1) I think your claims about flexible nuclear generators (likely SMRs?) coexisting well with intermittent sources ignore the economic costs of shutting down a capital-intensive nuclear plant when the weather-dependent generators are humming. Bad enough that SMRs are virtually certain to produce more expensive kwhs than today's big nuclear stations, but virtually all of them run whenever they are capable, and most still need subsidies (after their initial capital costs are written off) to keep going. They are a sad choice to stay idle while the wind blows and the sun shines.
2) Late in the article, you finally acknowledge the fact that any air-sourced heat pump will have a Coefficient of Performance that gradually fade down to 1.00 (the same as resistance heating) as the winter temperature dips. But the arguments you make in the early part of the article are dangerously and importantly oblivious to that fact.
And the short-term victims of that relationship are not the homeowners, unless they have foolishly eliminated all backup heaters, but their electrical utilities! Early in the article you claim that a massive switch from resistance heat to heat pumps will be a boon for utilities because their costs are powerfully linked to their PEAK load. That's true, but you have the bottom line exactly BACKWARDS!
Their peak winter load will be a near-linear function of the peak winter heating load, and that won't drop AT ALL by replacing resistance heat with heat pumps (except in Florida). What WILL drop very significantly is the heating load AWAY from the seasonal peak, and the total amount of electrical energy they will sell to the customers who've made the switch.
So the utility's costs (to meet the peak demand) will stay the same,but their total revenue (to pay those costs) will drop, by 50% or more.
The obvious remedy for the transmission and distribution utilities is to apply for rate reform to eliminate the savings their heat-pump customers have been enjoying for the near term!
My take-home slogan is "Friends don't let friends buy heat pumps!"