The long and short of it is that if the heat pump works below -20F, then the boiling point of the refrigerant must be below -20F. This, in turn, implies a higher pressurization (as per the Clausius-Clapeyron eq) required in order to achieve a T_hot of 80F (or whatever output temperature you want. The higher pressurizations require more expensive components and compressors.
Scroll compressors have come down in cost a bunch in the last few years. There's really very few reasons not to get a heat pump, now, for new construction.
The main reason in my area is that a.) these fancy new efficient ones aren't available easily and b.) no contractors in the area to install/service them. I just got a new gas furnace and really wanted to add a heat pump to the mix but after several quotes all of them either said no or strongly avoided the question. I couldn't get a straight answer.
Maybe I’m dumb, but aren’t compressors pressure-agnostic? They just add pressure, right?
Other than needing to be a bit stronger to keep from bursting against the higher delta with the ambient atmospheric pressure like the rest of the components.
In this case there is a higher delta between the system and the atmosphere as well as a higher delta between the hot and the cold side, which is what requires a stronger compressor.
The ideal model of everything is everything-agnostic.
Refrigeration stuff is generally soft copper, and modern refrigerants are already working up in the hundreds of PSI. So I can see that getting expensive or requiring a sea change in materials.
Even if you do need to make thicker pipes, strength in a metal generally increases with the square of the thickness, so it doesn't cost twice as much in materials to double the strength.
On top of that, the maximum theoretical COP for a cold source at -30c and a hot side at 35c is 4.7. Real world heat pumps usually get half of that, possibly even less for this extreme delta t, so around 2. Natural gas tends to cost less than half per unit energy compare to electricity, even less typically, so you'll never make back the capital costs.
Of course if this technology matures to the extend it's much, much cheaper and you tend to have milder winters and hot summers it's good enough and possibly worth it, but we haven't gotten to maintenance and installation costs yet, as the 35c hot temp assumes underfloor heating or tons of fancoils which have their own issues, the fact that the heat pump must be massively overspectd to output enough heat in peak conditions etc.
> then the boiling point of the refrigerant must be below -20F
This isn't the limiting factor for choice of refrigerant... There is always a low enough pressure that anything boils.
The problem is that at very low pressures (think a few millibars), gasses need huge diameter pipes and huge pumps to move even a small number of kilowatts of heat.
Another issue is that at pressures below one atmosphere, the system will suck in ambient air, causing condensation to collect inside your refrigeration circuit, which creates ice crystals, etc… nobody does that.
> “additional efforts are needed to address common technical and market barriers to wider adoption by consumers – which include performance at temperatures of 5F and below, installation challenges, and electricity grid impacts during peak demand periods.”
There are definite market barriers at play. In my house in New England, I tried to replace my aged boiler with an air-to-water heat pump (after carefully verifying, via experiments during a cold week in February, that my heat distribution would indeed work fine at a supply of 130°F). Only one company was even willing to come out and provide a quote and their quote was around 2.5x the costs of "put another boiler in", such that the payback period would be "literally never".
If, after doing the research to find out about them and specifically seeking one out, I couldn't manage to make an air-to-water heat pump make sense, I doubt that very many of them are being sold. I suspect it's one of those items that, if more were sold, more firms would sell/install them, bringing the costs into the realm of economically reasonable (and lowering the risk of having a difficult-to-support heating plant in the decades to come).
If it's not a company that solely does heat pumps, I have heard a lot of contractors will give outrageous estimates because gas is simpler for them and they don't want to do it without the huge markup.
I had ground source heat pump installed with vertical wells in a city by a dedicated geo installer. The cost with tax credits came out not much more than a high end gas furnace and water heater. Going airsource would have been even more cost competitive, especially with the federal tax credits in place starting in 2023.
They did. To (only partially) guard against that, I made it clear that I wouldn’t be hiring them to hang a gas boiler. (That still allows them to prefer someone else’s gas boiler job over my air-source job, but at least prevents me bidding against myself.)
And there definitely would be more labor, more piping, and more electrical work to switch to an ASHP; that’s part of the market forces problem that is hard to overcome with anything other than large price increases for gas or larger direct subsidies for switching.
I watch the PBS show This Old House which takes place in the Boston area. The plumber / HVAC guy Richard Trethewey is a fan both ground ("geothermal") and air source heat pumps including the air source heat pumps that work at 0F and heat water for radiant floor heating. I'm surprised that more companies aren't doing it there.
The 'cast' of 'This Old House' is mostly MIT MechE grads who figured out they could make more money doing renovations for Route 128 techies than burning themselves out working at their companies.
> their quote was around 2.5x the costs of "put another boiler in", such that the payback period would be "literally never".
Doesn't that depend on the costs of both energy sources?
Last year I made a similar choice, albeit at smaller scale, just for one water heater. Picked an electric heater with heat pump, also cost 2.5x more than plain electric, but 1/3 the energy cost. It will take a few years to pay back ...
> Doesn't that depend on the costs of both energy sources?
Yes, it depends on the costs and efficiencies of the competing energy sources, the difference in capex, the annual building heat load, the projected lifespan of each source, annual maintenance costs, and the interest rate.
I recently spoke to a well driller that also does geothermal well drilling. He was telling me that the systems are more or less obsolete at this point, as the air-to-air units have such a high SEER rating that the ground loops really will never pay off.
I have been doing a bunch of research into geothermal recently and this doesn't seem to match with what I have seen. But it probably also varies a lot by location and such.
They appear to still be about twice as efficient as an air to air one, and have less parts and maintenance (no outside unit to deal with). Yes the initial install is a lot higher because of the drilling but that should last for decades. In the US there is also 30% rebate at tax time which helps cut the costs down a bunch too.
What attracted me to geothermal was the year round availability of 50F water - easier to extract winter heat at that temperature, and you could create a simple fan / radiator ducted air cooling system for the summer. The latter would actually store summer heat in the ground, and could be powered purely by solar.
I think this is most of the way to what the future looks like: a high efficiency heat pump with cogeneration, so you can still burn propane or natural gas for heat and limited power to drive the heating system in an emergency (and if you don't have gas service to the premises, you have either a propane tank [1] or an exterior outlet to hook up outside). You must have a solution for when utility power is down for substantial periods of time.
During the winter storm that just passed, a friend in the Midwest called that their house had no power and was rapidly cooling. The utility could provide no ETA to resolution. I walked them through (over the phone) safely enough backfeeding enough power from a gasoline generator (outside, with the extension cord run through a basement escape window) into their furnace circuit to bootstrap the furnace (and run the blower fan) to keep the house warm so that the pipes didn't freeze and burst. If the HVAC system had had a small battery and some way to generate power from the heat it was burning, the gasoline generator would've been unnecessary. Perhaps an integrated thermoelectric generator [2]? A standby generator isn't financially practical for most folks ($6k + install).
(EDIT: to the safety folks out there possibly concerned, the furnace breaker and main breaker were tripped, and the meter was pulled to prevent any chance of harm to electrical linemen from inadvertently energizing the utility line; take no chances with safety, do not attempt this at home)
My house uses a NatGas boiler and in-floor radiant heat. It also has wood stoves for those -20F (or worse) nights when I need a little supplemental heat (or the power is out and the main heating system is offline).
A heat pump would be a waste up here in Alaska esp given I don't need A/C. Just opening the windows and running some fans in the summer tends to do the trick for cooling.
My use case is similar. The oil furnace feeds hot water into a network of radiators. My design goal is to maximize the electricity outage that we can make it thru without frozen pipes, and it has to be on a budget (so no generator or giant UPS).
The house is mostly unoccupied (a second home). Operation in an outage has to be fully automatic - it needs an automatic transfer switch I guess. The typical indoor temperature is just 5.5C (42F) - this does not leave much margin for the house to cool down during an outage when outside is like -5C (or -10 or -20).
But it turn out that an oil pump and a water circulation pump do not draw THAT much power, so if I can run them off backup power (say, a new car battery plus an inverter), it should last for some time before ice has any chance to form.
I’ve got the same thing. I did a bunch of tests last year to figure out relative performance and costs of heating my house at various outside temperatures. This included doing fine grained electricity monitoring with an iotawatt and rough fuel oil usage using an ultrasonic sensor.
Of course, now I’ve got a ton of data that I haven’t finished building my automation with. And, unfortunately, I’ve got an annoyingly bad Lennox iComfort for my heat pump thermostat and may have troubles building the automation anyway.
Air to water is pretty new technology for the U.S. As you mentioned most oil boiler hydronic systems are spec’d for a much higher temp (like 160 to 180F). I’ve been curious myself if you can salvage any of the existing baseboards with a 130F supply.
I have a 5 head mitsubishi cold weather mini split installed last year but I am still keeping my oil boiler for now for domestic hot water and supplemental heat. If I could switch that to an air to water heat pump for a reasonable cost that would be nice.
We need heat pump that work below -20F like we need an electric car that can go 1000 miles in a single charge - which is we really don't for the 99% of the use case. What's needed is a heat pump that's cheaper to install than gas furnace or oil boiler for the 80% of the population. On few days of the year when it's -20F or below, it's ok to use resistive heater as a back up.
Yup. And at least here in Finland the pumps come with the resistive heater included so it automatically switches over to it once it gets too cold.
Which where most people live here averages to less then 1 day per year so not a big deal.
Way bigger issue here is some fall storm destroying the power lines and being without electricity for multiple days when it already is cold enough they one needs heating.
Isn't the primary issue not the cost of installation, but rather the cost of the fuel? Natural gas in most of the US is far cheaper than electricity, and even if heat pumps are theoretically more efficient than natural gas (energy in:heat out), if the fuel is two or three times cheaper none of that matters.
Unless gas is subsidised I don't really understand how gas can be cheaper. A modern heat pump can give 3-5 kWh of heat using 1 kWh of electricity. And a modern gas generator is roughly 50% effecient at creating electricity (according to the Internet). So by using electric and a heat pump you should be able to get 1.5-2.5 times more heat from the same gas by making electricity of it first compared to burning it directly.
In most places in the US, in normal working conditions, Heat Pumps are 30%+ cheaper than a gas furnace to run.
However, in some really cold places like Chicago, Minneapolis, etc - the days where current heat pumps are inefficient might be enough to make it cheaper to always run gas.
OP is proposing to have both systems and only run the gas furnace on extreme days - which would lead to a ~30% reduction in running costs.
I suspect the CapEx of having two heaters wouldn't make sense, though.
It'd be better to just have a hear pump that can run efficiently at colder temps.
Maybe the gas is cheap, but the monthly and up-front connection costs mess with the economics.
And that cost will only go up as people cut the gas cord.
Depends on how your utility bills out it’s infrastructure: some charge minimal monthly connection fees, others a lot.
I honestly wouldn’t want to own a residential-focussed gas distribution company unless someone revolutionizes stirling engines or micro cogen systems and people start cutting their electric cord.
Interesting thing here is that natural gas is about 1/3 the cost of electricity per energy unit, so when the COP is above 3, it's cheaper to heat home using heat pump.
The reason why natural gas is about 1/3 of the cost of electriciy is because most natural gas power plants run at thermal efficiency of about 30%.
The thing with heat pumps is that they can be as much as 300% efficient (effectively) or more. This is because they can get "free" heat from the environment.
So even if per unit of energy gas is way cheaper, heat pumps can still come out ahead.
Sort of, but really we need both. Yes cheaper for most people. But as someone who lives in Minnesota, we definitely get below those temps. By code we require full backup resistive heaters for a heat pump at such a rating which increases the full cost and installation cost. Plus, it is much less efficient than a heat pump (though efficiency lowers as the temp gets colder due to defrost cycles).
I'm not an expert on the topic, but I imagine that a heat pump that can handle -20F is also much more energy efficient when the temp is 0F (in comparison to a heat pump that was rated only for -5F, operating at 0F).
Eh, for the first time I got an alert message on my cell phone asking people to reduce their electricity usage during a blizzard here in rural Minnesota.
If everyone is suddenly using electric heat when it’s -20 in an area there might be load issues.
The number of electric devices added to the grid is likely a predictable percent increase every day/month/year, and likewise, the acceleration of that change is probably somewhat predictable, too.
This is not to say we shouldn't be concerned, but denying yourself the most energy efficient technologies available (EVs, heat pumps, etc) because you're afraid of power outages 5-10 years from now seems like overkill.
If everyone uses resistive heat during that 1% of the time that it's -20F or below, the electric grid goes down and then no one gets heat. Consider what just happened in the southeast with TVA and rolling blackouts. That was precisely because it was too cold for heat pumps and so everyone's resistive heat engaged at the same time. I don't think your EV range comparison is a particularly good one. You can control your stops on a road trip, you can't control when it's colder than -20F outside.
If they can manage the grid granularly enough and isolate critical environments like hospitals, nursing homes, etc., I really don't see a good reason for us to overbuild to handle the third standard deviation of electricity demand. It makes a lot more sense to set the expectation that on the coldest days of the year, your house may spend a few hours disconnected from the grid in order to shed load. Not only does the avoid overbuilding, but it also contributes to less overall fragility in society, because black swan events that happen once every 50 or 100 years are the sorts of things we can't build for anyway, so it's better if people are prepared to endure the unexpected from time to time anyway.
AFAICT one of the main difficulties with heat pumps is that they want to use low temperature heat emitters, similar to condensing boilers. This is a general thermodynamic rule, but hits sources aiming for high efficiency extra hard, since they've been designed around exploiting it.
So you can't just take a decades old system with oil/gas using finned radiators, just replace the boiler, and have it supply enough heat on the coldest day ("design day"). Rather you'd at least need to add some additional emitters, greatly increasing the scope of the project for a professional installer.
What I haven't been able to find an answer to is that everybody says hydronic heat pumps need low delta T of 5-10 degree F (implying high flow rate for given heat transfer). But I would think the real constraint would be just on their leaving water temperature, and a heat pump (load side) that took in 100F and put out 120F (at say 5GPM) would be happier and more efficient than one that took in 110F and put out 120F (at 10GPM). But I've yet to find anything that confirms this.
Your intuition is correct on the last point. Where the "everybody says" side is coming from is the average water temperature in the emitters (and therefore heat flux from emitters to the building) will be higher if the flow (supply) is 120°F and return is 110°F than if the flow is 120°F and return is 100°F.
That's why designers are often specifying lower delta-Ts for low-temperature emitters: to allow the flow temperatures to remain as low as possible [for efficiency] at a given average water temperature [for effective heating].
> What I haven't been able to find an answer to is that everybody says heat pumps need low delta T of 5-10 degree F (implying high flow rate for given heat transfer)
I don’t see why this would matter at all. Maybe the heat exchanger would need to be sized differently for a different flow rate, but in general a lower entering water temperature on the hot side seems preferable.
Having heat pumps that can operate during arctic freezes is of course, and pardon my pun, pretty cool, but I wonder which percentage of the consumer market actually requires this? Especially keeping in mind that your pump not actively heating its internal storage for a few hours every day is not a huge issue: it only becomes problematic when the external heat exchange is unavailable for 6-8 hours or so.
Another comment in the article, regarding electricity grid impacts during peak demand periods, is more interesting to me. Currently, there is no mechanism whatsoever for heat pumps to automatically shift their grid draw (or re-delivery) to certain time slots, and/or to coordinate those slots with other units nearby. Both of these would greatly help to balance the grid, but won't be available until standardization gets off the ground and expensive retrofits are done. That's a shame, really...
There's quite a big market for this, such as a lot of the Nordics, where -15C for a few days is common enough that not having heating then would rapidly become an issue as homes freeze quickly. At those temperatures you have heating on 24/7 to keep the cold at bay. Electricity costs on those cold days are indeed high due to the constant demand of heating.
There may be no existing, comprehensive central management of electrical demand from heat pumps but that doesn't mean there are no mechanisms whatsoever, because there are systems already for regulating AC demand. In our region of the US, residential customers can opt in to a system where the electric utility can remotely disable their air conditioning compressor during times of peak load, in exchange for discounts. There are agreements on how long that can remain in effect, which would probably have to be stricter for heating systems, and probably have other safeguards added. There are also non-centralized AC mechanisms through Nest thermostats, where they can pre-cool buildings early in the morning to spread load out more evenly during the day.
Maybe just as important, whether or not there is a centralized mechanism for dealing with it, utilities already have exactly the same load problem with AC in the summer and have more or less figured out how to deal with that. It may be different in places where AC is less widespread, but in most of the US I wouldn't expect heat pump load issues to be much of a problem in the winter.
> House heating tends to have controls from thermostats or other devices; I'd tend to look to them for grid-related time-shifting
Not for heat pumps. In a typical configuration (large heat reservoirs combined with under-floor heating in a well-insulated house), temperature-setpoint changes take 12-24 hours to propagate.
So, governing the (in-room) target temperature settings is unreliable/unpredictable. Whereas the in-pump storage temperature is a whole lot more manageable.
The old advice "turn your thermostat down at night" therefore also doesn't apply to most heat pump installations -- in fact, it might be disadvantageous. "Select the average temperature you need and don't touch it" is much better advice. Need localized heat/cold? Use another solution for that...
The standard large buffer tank is 120 gallons, which stores about 20kBTU at 20F dT. So to last 4 hours you'd need your design heating load to be 5kBTU/hr, which I think is out of the realm of even new construction.
You could raise the max water temp and install a few tanks to get in the ballpark, but that's an assumption that most systems won't have. Also I wouldn't be surprised if "operates down to -20F" includes reduced efficiency/output that already relies on the buffer/storage to compensate for.
It's usually about 35F to 22F daytime when I turn on the heater, so efficiency at low temperatures would be the main thing that matters. It doesn't get to -20F that often most years, but it does hover around -5F for days at a time on occasion.
Current generation heat pumps are probably still worth it but ultra low temp performance would be nice.
I'd like to see an efficiency curve of this heat pump at -20F, 0F, +20F, +40F
I have a heat pump that can be used for both cooling an heat along with a natural gas burner. The installer has set the system to use the heat pump at 40F and above and switch to natural gas at below 40F below based on the efficiency of the heat pump dropping at low temperature.
My heat pump is a SEER 18 unit primarily for cooling in the US south so I'm sure a heat pump designed specifically for northern cold climates will be more efficient than mine at low temps but I'd like to see how much.
Unfortunately only the very top-of-the-line (very expensive) variable speed compressor units are rated for anything close to 24 SEER / -5F (unless that's what you meant by modern systems).
> Well that installer cutoff was likely waaaaay too high. Your heat pump should be fine to 30F, maybe even 25F before needing gas
One possibility might be that the unit lacks de-icing circuitry. If that were cut for cost optimization, the unit would still work fine for cooling and for moderate-temperature heating, but anywhere near 30F it would ice up and stop working.
The needs of the homeowner and the needs of the grid are at odds with one another here.
The electricity grid wants the highest possible efficiency on the coldest days, so that they can serve as many users as possible without building more infrastructure.
The homeowner wants the average efficiency to be as high as possible over the whole season, to reduce heating/cooling costs. They don't care if one or two really cold days have bad efficiency, as long as the system has sufficient output to keep the house comfortable.
Someone needs to use laws or incentives to align those two - because if every home owner used one of todays heat pump systems, then the electricity grid would fail on the coldest days of the year.
> Someone needs to use laws or incentives to align those two - because if every home owner used one of todays heat pump systems, then the electricity grid would fail on the coldest days of the year.
Or a technology where grid maintainers can tune down your heat pump, EV charger, hot water cylinder, etc all the way to minimums.
I believe this is already required in Australia for some tech.
If my power goes out (due to ice storms or hurricanes) and I need to rely on a local power supply (battery, solar panels, gas generator), I would like my AC to be as efficient as possible.
I didn't explain clearly... AC systems don't have a single efficiency number - they have different efficiencies depending on the indoor temperature, the outdoor temperature, and the number of kilowatts you want delivered.
You can generally design any AC system to work efficiently at any specific combination of those variables - but if any variable deviates far from the optimum design point, efficiency will drop.
So the real question is, not "I want an efficient system", but "I want an efficient system when it is 20F outdoors, because that's the temperature most of the year".
Yes. If the cold side gets too hot, the pressure on the compressor input gets too high, which means the compressor is doing far more work with every stroke - and input electrical power goes up massively. The motor overheats and the thermal cutout stops it. When the motor is cutting in and out efficiency goes way down.
This is an issue with fridges. When you buy a new fridge and first turn it on, it's called a 'pulldown'. The compressor gets far hotter than it ever normally gets in normal operation. Most fridges are only rated for 3 pulldowns in their lifespan - and if you do more than that and the fridge fails, they'll claim it isn't in warranty anymore. And in modern fridges, the software keeps track of how often so they can deny the warranty claim too...
I think the ultimate combo will be heat pumps + wood pellet stove as supplemental heat.
Then you can cutoff the gas grid connection and its associated standby/account/blah blah charges. It’s a big sunk cost in a lot of places that messes with the economics of switching to heat pump as primary heat.
Pellet stoves are semi-automated. Around 90% efficient. If you already have central heat pumps, you can install one and let your HVAC circulate the heat around. Can stockpile as much fuel as you want. Cheaper than oil or propane and not much more expensive than firewood once accounting for improved burn efficiency. Just need to empty the ash gray once a week or so, and dump a nice smelling bag in for every ~24h of operation.
Relatively straightforward install: just need a wall to punch through and a standard power outlet. Minimal clearance requirements. Fun to watch the fire tornado.
Big downside is they need some electricity (mainly for for the powered vent). Hit or miss when it comes to insurance companies that think explosive gas systems or high current electric devices are safer.
Pellet stoves require electricity to run the auger. Also- nobody gives away free pellets but it's pretty easy to come by free wood people are giving away.
Readit News
The long and short of it is that if the heat pump works below -20F, then the boiling point of the refrigerant must be below -20F. This, in turn, implies a higher pressurization (as per the Clausius-Clapeyron eq) required in order to achieve a T_hot of 80F (or whatever output temperature you want. The higher pressurizations require more expensive components and compressors.
Other than needing to be a bit stronger to keep from bursting against the higher delta with the ambient atmospheric pressure like the rest of the components.
Refrigeration stuff is generally soft copper, and modern refrigerants are already working up in the hundreds of PSI. So I can see that getting expensive or requiring a sea change in materials.
These all change based on the pressure delta.
Sure, that sounds like an acceptable compromise for those who need the lower operating temperatures
And maybe technology can get those compressors at the same price point of current compressors
This isn't the limiting factor for choice of refrigerant... There is always a low enough pressure that anything boils.
The problem is that at very low pressures (think a few millibars), gasses need huge diameter pipes and huge pumps to move even a small number of kilowatts of heat.
There are definite market barriers at play. In my house in New England, I tried to replace my aged boiler with an air-to-water heat pump (after carefully verifying, via experiments during a cold week in February, that my heat distribution would indeed work fine at a supply of 130°F). Only one company was even willing to come out and provide a quote and their quote was around 2.5x the costs of "put another boiler in", such that the payback period would be "literally never".
If, after doing the research to find out about them and specifically seeking one out, I couldn't manage to make an air-to-water heat pump make sense, I doubt that very many of them are being sold. I suspect it's one of those items that, if more were sold, more firms would sell/install them, bringing the costs into the realm of economically reasonable (and lowering the risk of having a difficult-to-support heating plant in the decades to come).
If it's not a company that solely does heat pumps, I have heard a lot of contractors will give outrageous estimates because gas is simpler for them and they don't want to do it without the huge markup.
I had ground source heat pump installed with vertical wells in a city by a dedicated geo installer. The cost with tax credits came out not much more than a high end gas furnace and water heater. Going airsource would have been even more cost competitive, especially with the federal tax credits in place starting in 2023.
And there definitely would be more labor, more piping, and more electrical work to switch to an ASHP; that’s part of the market forces problem that is hard to overcome with anything other than large price increases for gas or larger direct subsidies for switching.
Doesn't that depend on the costs of both energy sources?
Last year I made a similar choice, albeit at smaller scale, just for one water heater. Picked an electric heater with heat pump, also cost 2.5x more than plain electric, but 1/3 the energy cost. It will take a few years to pay back ...
Yes, it depends on the costs and efficiencies of the competing energy sources, the difference in capex, the annual building heat load, the projected lifespan of each source, annual maintenance costs, and the interest rate.
They appear to still be about twice as efficient as an air to air one, and have less parts and maintenance (no outside unit to deal with). Yes the initial install is a lot higher because of the drilling but that should last for decades. In the US there is also 30% rebate at tax time which helps cut the costs down a bunch too.
During the winter storm that just passed, a friend in the Midwest called that their house had no power and was rapidly cooling. The utility could provide no ETA to resolution. I walked them through (over the phone) safely enough backfeeding enough power from a gasoline generator (outside, with the extension cord run through a basement escape window) into their furnace circuit to bootstrap the furnace (and run the blower fan) to keep the house warm so that the pipes didn't freeze and burst. If the HVAC system had had a small battery and some way to generate power from the heat it was burning, the gasoline generator would've been unnecessary. Perhaps an integrated thermoelectric generator [2]? A standby generator isn't financially practical for most folks ($6k + install).
(EDIT: to the safety folks out there possibly concerned, the furnace breaker and main breaker were tripped, and the meter was pulled to prevent any chance of harm to electrical linemen from inadvertently energizing the utility line; take no chances with safety, do not attempt this at home)
[1] https://www.amerigas.com/about-propane/propane-tank-sizes
[2] https://patents.google.com/patent/US5427086A/en
A heat pump would be a waste up here in Alaska esp given I don't need A/C. Just opening the windows and running some fans in the summer tends to do the trick for cooling.
The house is mostly unoccupied (a second home). Operation in an outage has to be fully automatic - it needs an automatic transfer switch I guess. The typical indoor temperature is just 5.5C (42F) - this does not leave much margin for the house to cool down during an outage when outside is like -5C (or -10 or -20).
But it turn out that an oil pump and a water circulation pump do not draw THAT much power, so if I can run them off backup power (say, a new car battery plus an inverter), it should last for some time before ice has any chance to form.
Of course, now I’ve got a ton of data that I haven’t finished building my automation with. And, unfortunately, I’ve got an annoyingly bad Lennox iComfort for my heat pump thermostat and may have troubles building the automation anyway.
https://youtu.be/TlX5z32T1J4
Air to water is pretty new technology for the U.S. As you mentioned most oil boiler hydronic systems are spec’d for a much higher temp (like 160 to 180F). I’ve been curious myself if you can salvage any of the existing baseboards with a 130F supply.
I have a 5 head mitsubishi cold weather mini split installed last year but I am still keeping my oil boiler for now for domestic hot water and supplemental heat. If I could switch that to an air to water heat pump for a reasonable cost that would be nice.
Which where most people live here averages to less then 1 day per year so not a big deal.
Way bigger issue here is some fall storm destroying the power lines and being without electricity for multiple days when it already is cold enough they one needs heating.
However, in some really cold places like Chicago, Minneapolis, etc - the days where current heat pumps are inefficient might be enough to make it cheaper to always run gas.
OP is proposing to have both systems and only run the gas furnace on extreme days - which would lead to a ~30% reduction in running costs.
I suspect the CapEx of having two heaters wouldn't make sense, though.
It'd be better to just have a hear pump that can run efficiently at colder temps.
And that cost will only go up as people cut the gas cord.
Depends on how your utility bills out it’s infrastructure: some charge minimal monthly connection fees, others a lot.
I honestly wouldn’t want to own a residential-focussed gas distribution company unless someone revolutionizes stirling engines or micro cogen systems and people start cutting their electric cord.
The reason why natural gas is about 1/3 of the cost of electriciy is because most natural gas power plants run at thermal efficiency of about 30%.
So even if per unit of energy gas is way cheaper, heat pumps can still come out ahead.
If everyone is suddenly using electric heat when it’s -20 in an area there might be load issues.
This is not to say we shouldn't be concerned, but denying yourself the most energy efficient technologies available (EVs, heat pumps, etc) because you're afraid of power outages 5-10 years from now seems like overkill.
So you can't just take a decades old system with oil/gas using finned radiators, just replace the boiler, and have it supply enough heat on the coldest day ("design day"). Rather you'd at least need to add some additional emitters, greatly increasing the scope of the project for a professional installer.
What I haven't been able to find an answer to is that everybody says hydronic heat pumps need low delta T of 5-10 degree F (implying high flow rate for given heat transfer). But I would think the real constraint would be just on their leaving water temperature, and a heat pump (load side) that took in 100F and put out 120F (at say 5GPM) would be happier and more efficient than one that took in 110F and put out 120F (at 10GPM). But I've yet to find anything that confirms this.
That's why designers are often specifying lower delta-Ts for low-temperature emitters: to allow the flow temperatures to remain as low as possible [for efficiency] at a given average water temperature [for effective heating].
I don’t see why this would matter at all. Maybe the heat exchanger would need to be sized differently for a different flow rate, but in general a lower entering water temperature on the hot side seems preferable.
Deleted Comment
Another comment in the article, regarding electricity grid impacts during peak demand periods, is more interesting to me. Currently, there is no mechanism whatsoever for heat pumps to automatically shift their grid draw (or re-delivery) to certain time slots, and/or to coordinate those slots with other units nearby. Both of these would greatly help to balance the grid, but won't be available until standardization gets off the ground and expensive retrofits are done. That's a shame, really...
I did run across a release on a new generation of a heat-pump hot-water heater, which does seem to have some kind of grid-shifting built in. A. O. Smith's Voltex AL, https://cleantechnica.com/2022/12/21/all-i-want-for-christma...
Not for heat pumps. In a typical configuration (large heat reservoirs combined with under-floor heating in a well-insulated house), temperature-setpoint changes take 12-24 hours to propagate.
So, governing the (in-room) target temperature settings is unreliable/unpredictable. Whereas the in-pump storage temperature is a whole lot more manageable.
The old advice "turn your thermostat down at night" therefore also doesn't apply to most heat pump installations -- in fact, it might be disadvantageous. "Select the average temperature you need and don't touch it" is much better advice. Need localized heat/cold? Use another solution for that...
You could raise the max water temp and install a few tanks to get in the ballpark, but that's an assumption that most systems won't have. Also I wouldn't be surprised if "operates down to -20F" includes reduced efficiency/output that already relies on the buffer/storage to compensate for.
Current generation heat pumps are probably still worth it but ultra low temp performance would be nice.
I have a heat pump that can be used for both cooling an heat along with a natural gas burner. The installer has set the system to use the heat pump at 40F and above and switch to natural gas at below 40F below based on the efficiency of the heat pump dropping at low temperature.
My heat pump is a SEER 18 unit primarily for cooling in the US south so I'm sure a heat pump designed specifically for northern cold climates will be more efficient than mine at low temps but I'd like to see how much.
Modern systems are 24 SEER and good to -5, and these research units take that to the next level.
The parent post was saying it's cheaper to use natural gas at those temps, so that's why the installer did the cutoff there.
You can look at COP numbers for heat pumps here: https://ashp.neep.org/#!/product_list/
One possibility might be that the unit lacks de-icing circuitry. If that were cut for cost optimization, the unit would still work fine for cooling and for moderate-temperature heating, but anywhere near 30F it would ice up and stop working.
The electricity grid wants the highest possible efficiency on the coldest days, so that they can serve as many users as possible without building more infrastructure.
The homeowner wants the average efficiency to be as high as possible over the whole season, to reduce heating/cooling costs. They don't care if one or two really cold days have bad efficiency, as long as the system has sufficient output to keep the house comfortable.
Someone needs to use laws or incentives to align those two - because if every home owner used one of todays heat pump systems, then the electricity grid would fail on the coldest days of the year.
Or a technology where grid maintainers can tune down your heat pump, EV charger, hot water cylinder, etc all the way to minimums.
I believe this is already required in Australia for some tech.
You can generally design any AC system to work efficiently at any specific combination of those variables - but if any variable deviates far from the optimum design point, efficiency will drop.
So the real question is, not "I want an efficient system", but "I want an efficient system when it is 20F outdoors, because that's the temperature most of the year".
This is an issue with fridges. When you buy a new fridge and first turn it on, it's called a 'pulldown'. The compressor gets far hotter than it ever normally gets in normal operation. Most fridges are only rated for 3 pulldowns in their lifespan - and if you do more than that and the fridge fails, they'll claim it isn't in warranty anymore. And in modern fridges, the software keeps track of how often so they can deny the warranty claim too...
Then you can cutoff the gas grid connection and its associated standby/account/blah blah charges. It’s a big sunk cost in a lot of places that messes with the economics of switching to heat pump as primary heat.
Pellet stoves are semi-automated. Around 90% efficient. If you already have central heat pumps, you can install one and let your HVAC circulate the heat around. Can stockpile as much fuel as you want. Cheaper than oil or propane and not much more expensive than firewood once accounting for improved burn efficiency. Just need to empty the ash gray once a week or so, and dump a nice smelling bag in for every ~24h of operation.
Relatively straightforward install: just need a wall to punch through and a standard power outlet. Minimal clearance requirements. Fun to watch the fire tornado.
Big downside is they need some electricity (mainly for for the powered vent). Hit or miss when it comes to insurance companies that think explosive gas systems or high current electric devices are safer.