The article doesn’t go into it, but a key advantage they have is that heat pumps move heat, rather then trying to generate it. So they can move a lot more heat into your house than would be generated by running the electricity they use through a resistor. This makes them effectively more than 100% efficient (the exact amount depends on temperature) as compared with burning a fuel or resistive heat.
Does the paper have any results that say they’re still cleaner on a dirty grid? As far as I can tell it’s only cleaner in the future after at least a 50% decarbonization. Which is reasonable, even in my fairly conservative city most of our power is low on carbon.
The thing is heat from the outside gets moved inside of the house using a heat pump, and to facilitate this movement you need somewhere between 1/2 and 1/4 of the energy you end up moving. E.g. a heat pump with
(coefficient of performance) of 4 would move 4kW of heat into your house and use 1kW of electric energy to accomplish this. Gas by comparison moves 4kW of gas to your house and burns it there to get 4kW of heat.
So you could burn a bit more than 1kW of gas in a modern gas electric plant, turn it into electricity and use it to run a heat pump and you would end up emitting less CO2, the real world grid might skew that worse because generally you don’t end up burning coal to heat housing but you might still use it for electricity. So generally even though it might be unintuitive the more complicated and lossy way to heat your home (the heat pump powered by fossil powered electricity) , is the more effective one compared to burning the same fossil fuel directly because you use the heat pump to capture heat from the environment.
Maybe not in the article, but I’ve heard in other places that a carbon heavy grid still gets enough energy to the heat pump that the heat pump’s efficiency can offset that increase.
You’re also installing a system that is easier to decarbon in the future, which isn’t the case for natural gas.
It is really not hard. Heat pump coefficients of efficiency floor at 1, but typically range between ~2.5 and 7. That is, for every joule of energy they consume, they pump 2.5 to 7 joules of heat into the conditioned space.
So you have to just look at efficiencies involved.
Still, we’re just summing stuff. And while I won’t pretend any napkin math here is the same as a formal research project, we can plainly see that the HUGE energy efficiency of heat pumps can easily eclipse the inefficiency of fossil electrical production, all else being equal. Of course, whether it actually WILL be better than a fossil furnace will depend on local factors, but these places are increasingly becoming edge cases. And then, on top of that, you unlock future potential to seamlessly switch fuel sources from fossils to renewables, which becomes very important in lifecycle cost analysis.
This is the same reason electric cars beats ICE (gas driven) cars even when charged on coal power. Big coal plants and the distribution grid are more efficient than small scale car engines.
You need combustion engines to be BIG to get past 30-40% efficiency ranges, but the really big power plants can just perform efficient burns and heat water to drive efficient turbines, which is impossible in cars. And the rest of losses in electric cars are either minimal or equivalent, so you get a big net benefit.
Even better than that is an electric bus and other public transit!
I’ve done energy models for houses here in Saskatchewan (~560 tCO2e/GWh) and at the moment, they are not cleaner than heating with natural gas, which is the typical primary heat source. Obviously, it would depend on grid carbon intensity, so there is a level of grid ‘cleanness’ where heat pumps would become cleaner, but that tipping point depends on a number of factors.
You could do a rough estimation with the seasonal heating efficiency of a heat pump based on the heating-degree-days of your location versus a certain efficiency of natural gas furnace. Burning natural gas is about 0.18 kgCO2e/kWh. So, if you have a heat pump that’s 200% seasonally efficient, you’d need the grid carbon intensity to be about 0.38 kgCO2e/kWh (380 tCO2e/GWh) to be equivalent to a 95% efficient natural gas furnace.
The 200% seasonal efficiency is a bit off, Nordic models, measured with the “colder” European climate zone, get 300%+ and have guaranteed output at -25C / -13F. Example model from Mitsubishi:
It’s worse than 5.5x or 4.3x in warmer areas but the right model air source heat pumps work fine down to pretty damned cold. Norway and Sweden have a ton of them as they spend a ton of energy on heating and this saves homeowners a ton of money every year.
Best models optimized for average climate now reach 5.5x or better in the green, moderate zone, SCOP of 4.3 is actually pretty terrible but this one is built to be ice proof.
Example latest bestest heat pump with 6+ seasonal COP:
Nice. Saskatchewan is very cold though (about 6000 heating deg days at 18C where I am and can regularly go under -30C in winter), so 200% would be pretty reasonable for a typical heat pump. As a comparison, Tromsø, in very north Norway is 5600 heating deg days.
As a warm blooded, middle east dwelling humanoid, WTF is ‘6000 heating deg days’?
I have concluded that the 6000 is not days in a year or degrees of temperature.
To determine heating degree days for your area, you set a baseline temperature (18C is kind of standard in Canada) take the average temperature on each day, and sum the difference between that and the baseline temperature for every day of the year (zero if temp is above baseline). So if the average temp one day was -10C, it would be a 28 heating degree day.
It allows approximation of building heating demand. Some standards (Passive House) use heating degree hours for finer detail, which makes sense because there can be fairly significant day/night temperature swings.
Here’s a site where you can calculate what your location is. And here’s what Wikipedia says.
Huh, TIL…
Notable, but outside of very cold climates (which I think I feel safe describing Saskatchewan as being), heat pumps are a LOT more than 200% efficient. In mild climate, they can be 2-4X that.
Definitely, that’s why I say the seasonal heating efficiency is based on heating-degree-days of the location. I’m not sure they’d get to 2-4x 200% efficient, though. 350% might be more reasonable.
It gets hard to say because COP varies with climate. But even in SEER ratings, 17-20 are pretty much the norm for modern systems and I have seen as high as 23. That translates to a 4-4.5 COP in an average climate.
But those COPs get higher the more mild your climate – I am somewhere with quite mild winter where we only get a hard freeze once or maybe twice a year, and generally winter low temps are in the 40-50F range.
I believe the theoretical max efficiency for a heat pump is something like 8.8 COP. In a mild climate like mine, where most of the time if your heat is running it’s to heat to ~70ish from an ~50ish outdoor temp, you’re should be getting a lot closer to 7 than you are to 2.
I think the point is to compare the heat pump with an electricity heater, there may be other ways that generate less carbon footprint of course
The point is to compare a heat pump with a natural gas furnace.
The efficiency of getting electricity to a house may be less, but a heat pump has an energy efficiency greater than 1.