Heat pump vs Gas

The results are in, here’s the verdict

So it’s been just over a year since we had our Daikin Altherma high-temperature air source heat pump installed. Being a high-temperature system, this unit directly replaced a normal gas condenser boiler with no modifications required to our existing pipework and radiators. I now have a years-worth of data running this system, versus the gas boiler we had before, so it’s time to look at how they both performed.

First, the systems. The gas boiler was an 8-year-old De Dietrich Vivadens MCR with instant(ish) water heating. I can’t remember the exact model and I’ve since binned the manual, but I think it was the 28 kW. The heat pump is a single-phase Daikin Altherma R HT 16 kW with a 300 litre hot water tank attached.

The period measured is from 1st April 2019 to 31st March 2021. The heat pump was installed 17th March 2020. The temperatures over this period here in the Gard region of France went from between -3°C and +40°C, which is quite a range. It’s worth noting heat pumps are most efficient in “moderate” temperatures, between +10°C and about +25°C for our model, as far as I can tell (based on checking how much it consumes on our real time monitor).

Edit: I’m adding in some important information about the internal temperature of the house and heating times. When we had the gas boiler, we heated the house up to 20.5°C mornings and evenings and let it cool off to no lower than 18°C during the middle of the day when the kids are at school and at night. To top up the temperature in the kitchen and office on cold days we used a couple of little electric convector heaters, usually on the 750 W setting.

With the heat pump the house is never cooler than 20.5°C day or night, and heats up to 22.5°C on evenings and first thing in the morning, so basically with the heat pump we’re keeping the house more or less 2°C warmer at all times. I haven’t factored that in to any of the calculations that follow, but bear in mind our heat pump efficiency and operating costs would be even better if we kept the settings we had with the old gas boiler.

We also need to talk energy prices, at the point of both these measurements we were on a green energy contract with EDF costing 0.1257€ per kWh off peak and weekends and 0.2125€ at peak times, with a monthly standing charge of €16.79. The gas prices fluctuated, but I’ve settled on a value of €2.20 for a kilo of gas as a reasonable average over our 8 year contract.

Servicing costs are very similar, we paid and pay about €300 a year for someone to come and check the system and carry out routine maintenance.

Cost of energy

I’m going to start with the simplest measure, money. Every year has a period when it’s coldest and, ergo, we tend to use the most energy (we don’t tend to use air conditioning, even though it gets pretty hot here). With the gas boiler our coldest and most expensive period cost approximately €28 per day. I say approximately, because I checked the gauge on the gas tank frequently and averaged the consumption, but it was not an accurate gauge. Also, this is a value combined with our electricity consumption, also averaged as we didn’t yet have a smart-meter for real-time daily readings.

With the heat pump it was precisely €20.29, for 97 kWh of electricity consumed that day. I should probably note we have an electric car and have had since December 2019, so these values, recorded in January 2020 and January 2021 respectively, include “fuelling” the car.

More importantly as an indicator, perhaps, is our average cost. January is consistently our coldest month and our average cost of energy per day in January 2020 was €20.83 versus €11.95 in January 2021. This is clearly a huge difference, the heat pump was an enormous magnitude cheaper to run on average. This is a consistent pattern, €10.11 vs €5.49 in April, €8.77 vs €5.00 in May, etc. Indeed, I calculated our household energy costs to be €4,383 for the year with a gas boiler, against €2,313 for the heat pump year — almost half! (Note, we also had a 3kW solar array since 11th August 2020, which helped some. But even so, the heat pump is a lot more cost effective.)

The heat pump was a lot more expensive to install than the gas boiler. But at those running costs we’ve got the difference back in 3 years, even before considering the tax credits we’ll get for replacing an old gas boiler.

Household efficiency

If the price difference is impressive, the difference in efficiency is even more so. The number of kilowatts consumed by our household in a year dramatically dropped after the heat pump was installed.

Again, there are a few notes to make before I continue. Firstly, when I say “household” I do include the car. Between April 2019 and November 2019 the vast majority of our driving was done in a 10-year-old 1.2l Fiat Panda. From December 2019 onward it has been a Renault Zoé EV. I kept detailed records of our mileage, so to calculate the kW per day consumed by the Panda I took the round average litres of fuel per month (70 litres) multiplied by 8.9 kW per litre of fuel (taken from this website), multiplied by 12 months and divided by 365 days. It comes to 20.48 kW per day for the car.

For calculating the kW used by the gas boiler I used the figure of 13.4 kWh of energy per kilogram of gas burned. This is a debatable figure, the top source in Google cites 14.019 kWh but I also found reputable sources stating sub-13 kWh values, so I opted for something in the middle.

In January 2020 the kWh generated with propane gas peaked at an average of 134 kWh per day over a 5-day period. If you add in the parallel electricity use (the car is electric at this point) in the high-20s of kWh, we’re around the 160 kWh per day mark. I’ve already stated our heaviest day since was 97 kWh.

And again, when you look at averages the heat pump (mostly, but also the electric car) is way more efficient. 133.69 kWh per day was the average in January 2020 versus 69.74 kWh in January 2021. And if we go back to April 2019, when we still had the Fiat Panda, the average kWh per day was 66.81 versus 29.92 with the heat pump and the EV in April 2020, significantly less than half the energy consumption year on year.

CO2 emissions

The only measure that really matters, when it comes to the environment, is how much CO2 is being emitted. This study clearly excludes the CO2 impact of other factors, such as our food choices and other lifestyle choices like clothes and holidays. It looks solely at household energy, heating and local transport.

It also only deals with the point of consumption and not the manufacture of the plant, the latter point being up for debate. I couldn’t find any sources raising concern about the environmental impact of manufacturing heat pumps, but on the EV side, for example, cynics will point at the amount of water consumed by lithium production for batteries. (Though most of them will have smartphones, tablets, laptop computers, and so on, so they tend to be selective about when they choose to criticise lithium batteries — and they also ignore car battery re-use programs in industrial storage and re-use in forklift trucks, golf carts, and so on.) In any case, I don’t want to debate that here. I simply want to be clear this is not “full life-cycle” CO2 emissions, rather “in service” CO2 emissions.

How you measure the CO2 emitted by different fuels and by the electricity grid is also contentious. I’ll explain what I’ve done, you can either agree with me this is reasonable or apply your own maths to my measurements.

For propane gas I settled on 2.99 kg of CO2 per kg of propane burned, taken from this web page, which seems to me to be a reasonable source with no reason to have any particular bias. For petrol (for the period we were running the Fiat Panda) I used 2.3 kg of CO2 per litre of petrol burned, which Google indicates is a widely accepted figure.

Electricity is less clear-cut. According to Climate Transparency, CO2 emissions per kWh of French electricity is already a scant 0.048 kg, thanks to the huge investment in nuclear energy in the 1960s. Perhaps controversially, I’ve gone in harder than that. Although clearly we cannot specify only green energy comes “down the pipe”, so to speak, we have a contract that states any money we pay goes 100% back into renewable energy providers and sites, a mix of solar, hydro and wind power. Good luck finding an agreed-upon figure, but Carbon Brief claim these energy sources have CO2 emissions of just 0.005 kg of CO2 per kWh generated. So that’s the number I’m using for my electricity. It might be naive, it might be accurate, I have no way of knowing!

So with the base numbers explained, let’s look at the CO2 emissions chart:

Wow. In January 2020, even with an EV already and a green energy contract, we were averaging 24.66 kg of CO2 emitted per day. In stark contrast, and based on the figures above, in January 2021 the average CO2 emitted per day was only 0.59 kg. That is an astounding drop in CO2 emissions!

Extrapolating that out to annual figures, in the last year we have emitted around 153 kg of CO2 as a household, compared with 5,051 kg of CO2 — over 5 tons! — the year before. That is a massive win.

Learning points

The main point is “going green” with your household energy is definitely worthwhile. Even if you don’t really care about the environmental aspects, which you probably do or you wouldn’t even be reading this, this is viable technology that is categorically cheaper, in the long-run, than the “dirtier” equivalents, even though the up-front costs can sting a little.

We have a single-phase electricity contract and a 12 kW single-phase connection, which is the maximum possible with single-phase in France. (Anywhere?) With the benefit of hindsight, especially in these days where we’re turning to electricity far more as a cleaner way of generating energy of all sorts, I wish we’d gone for a triple-phase connection and a triple-phase heat pump. On a cold winter’s morning, when it’s just started up and the heating loop is cold, the heat pump can consume 8 or 9 kW. If you’re charging an EV (minimum 2 kW) and somebody boils the kettle (also 2 kW) you just went over your 12 kW and tripped the power out. Doh!

With a triple-phase system and heat pump it would’ve cost a bit more per month, but you can go up to 15 kW, 18 kW or higher. You have to make this decision before you order your heat pump, as a single-phase heat pump will probably be unusable in a triple-phase context, and vice versa. But I’d recommend it, if you had triple-phase and 18 kW (so 3 times 6 kW) then your heat pump, when cold on a cold day, would draw 3 kW off each phase leaving 3 kW for other things. So if your garage was on one phase, where your car was charging, and your kitchen was on another phase, where you were boiling your kettle, you’d have no problem. You could also fit a triple-phase EV charging point for rapid charging of an EV. In the coming years I believe triple-phase electricity supply is likely to become the norm rather than the exception — good news for electricians!

Hot water via a heat pump if, like us, you use quite a bit of hot water (we’re a family that likes a hot bath) is not as cost-effective as a gas combi-boiler. It’s efficient if you disable on demand mode and make do with a 300 litre tank per day, which will do a lot of people, frankly — and you can always give it a burst manually if you run out — but if you use a lot of hot water you’ll want to consider a secondary means of heating your water. I believe Daikin do a hybrid vacuum tube/heat pump hot water system which looks attractive. But it all depends on your usage patterns. If you’re fine with a quick shower of a morning and so’s everyone else in your family, then 300 litres of hot water produced once a day at the most economically advantageous time will do!