Assessment of Heat Pump heating water to 50 °C and 70 °C

Friends, our Air Source Heat Pump (ASHP) (a 5 kW Vaillant Arotherm Plus) has been installed for over a month now and I am beginning to get a feel for how it is working.

At this time of year (early September) we have no space heating requirements so the work load for the heat pump is low.

Most of the day it sits in the garden admiring itself, and consuming 12 W of electrical power (0.29 kWh/day)

Our heat pump idling away the late summer days in the back garden. It only works for an hour a day!

Each night it wakes itself at 3:00 a.m. and if the hot water tank requires a top up, it operates for about an hour, heating the tank to roughly 50 °C.

  • Typically it uses ~1 kWh of electricity and delivers ~3 kWh of heat.

On Wednesday mornings it additionally heats the water in the tank to 70 °C in a so-called Anti-Legionella cycle.

  • This typically uses ~3 kWh of electricity and delivers ~7 kWh of heat.

Later in the year I expect that the heat pump will begin to be required to heat the house, and I’ll write about that in a little while.

But for now let me just describe how the system is working at present.

A Normal Cycle

Two typical water-heating cycles from the 5th and 6th September are shown below. The external air temperature in each case was about 15 °C.

Click for a larger version. Typical performance of the heat pump when heating domestic hot water. The two upper panels show data from 5th September and two lower panels show data from 6th September. In each case the left-hand panel shows electrical power consumed (watts), the heat delivered to the cylinder (watts) and the water temperature (°C). The right-hand panel shows instantaneous COP and dotted lines show two estimates of the average COP. 

The key measure of how well a heat pump works is its coefficient of performance (COP) which measures the ratio of thermal energy delivered, to electrical energy consumed.

The graphs on the right above show how the COP varies from minute to minute through the heating cycle.

Also shown as dotted lines are two estimates of the average COP.

  • The blue estimate includes all the electrical energy which the heat pump uses during the 23 hours when it is not ‘working’.
  • The purple estimate includes only the electrical energy which the heat pump uses during the heating cycle’.

Depending on which measure one uses, the COP is between 2.5 and 3 i.e. the heat pump delivers between 2.5 and 3 times as much as heat as the electrical energy it uses

An Anti-Legionella Cycle

Legionella bacteria, which can cause Legionnaires Disease, are capable of lurking in hot water systems at temperatures below 60 °C.

To counteract this, every Wednesday morning the heat pump system additionally executes an Anti-Legionella cycle which heats the water to 70 °C. It should be noted that it is very unusual for heat pumps to operate at all at such high temperatures.

Click for a larger version. Typical performance of the heat pump during an anti-legionella heating cycle on 1st September. The left-hand panel shows electrical power consumed (watts), the heat delivered to the cylinder (watts) and the water temperature (°C). The right-hand panel shows instantaneous COP and dotted lines show two estimates of the average COP. 

From the graphs above one can see that heating to higher temperatures is hard work for the heat pump and the average COP falls from the range 2.5 to 3.0 when heating to 50 °C, to just around 2.1 when heating to 70 °C.

Hot Water Temperatures

Click for a larger version. The measured temperature of hot water at three hand-basins in the house over a period of 20 days. After the anti-legionella cycle in the early hours of Wednesday morning, the flow temperature of water at the taps can reach almost 70 °C, a potential scalding hazard. At other times, the hot water is delivered at just under 50 °C

One unanticipated feature of the Anti-Legionella cycle is that on Wednesday mornings, the temperature of water delivered from the hot water taps is very high – almost 70 °C.

With our level of water use, the system typically skips the Thursday heating cycle because the water is still hot from Wednesday’s ‘super’ heating. Indeed, the water does not return to ‘normal’ temperatures until Saturday!

Delivering water at almost 70 °C is a significant hazard and so I will shortly have anti-scalding valves fitted to the outlets which will limit the maximum temperature of hot water to about 45 °C.

Once I have finished with my tests, I will also reduce the normal hot water temperature by a few degrees.


Overall the system is doing well.

Click for a larger version. COP performance of the heat pump during normal heating cycles and during anti-legionella heating cycles. Heating the water to 70 °C degrades the performance of the heat pump.

Looking at the performance during normal heating cycles, the heat pump heats water from around 15 °C to 50 °C with a COP of typically 3.4

Looking at the performance during anti-legionella heating cycles it heats water from around 15 °C to 70 °C with a COP of typically 2.4

These COPs do not include the electrical energy consumed during the 23 hours when the heat pump is on ‘stand by’. This better indicates the operating performance of the pump, but of course this ‘stand by’ energy still has to be paid for.

Overall (including the ‘stand by’ consumption) the heat pump is delivering on average 4.5 kWh/day of hot water heating at the expense of about 1.77 kWh of electricity/day.

At this time of year, all this electricity comes from solar energy stored in the battery and so costs nothing.

But as the winter season draws in, we will eventually operate this using mains electricity on the Octopus Go tariff. This provides electricity at 5p per kWh between 00:30 and 4:30 a.m. each day.

So the cost of 4.5 kWh of hot water in winter will be about 1.77 kWh x 5 p/kWh = 8.85 p per day.

This is equivalent to just under 2p/kWh (thermal) – which is about 40% cheaper than gas heating which costs about 3.3p /kWh (thermal)

Things will be a little harder in winter as the average external temperature falls, but I am very curious to see how the Vaillant ASHP performs.



6 Responses to “Assessment of Heat Pump heating water to 50 °C and 70 °C”

  1. Joe Hc Wentworth Says:

    Really interesting.

    I have a new 7kw arotherm. I wondered why you chose morning for your water cycle. If you have powerwall I would think you would get a better cop running it in the generally warmer afternoon. But using night tariff elec via the power wall.

    I only have a independant power meter for the pump. I find the vaillant controller live info under reports power by about 50watts. Perhaps it’s not counting the fan or something. But using my independent ashp meter readings and vaillants yeild reading I have estimated cop of 4.5 for 47 degree tank temp. This seems very good but the last two months have been very warm….

    Interested to know if you trust the vaillant ‘working figure’ the controller gives and if this has any bearing on the cop you get with your independent Metering?

    Thanks for this blog is been very useful.


    • protonsforbreakfast Says:

      Joe, Good Evening.

      In the summer we run the hot water cycle in the afternoon to take advantage of the warmer temperatures, and at that point the battery is generally fully charged and so the excess solar PV would otherwise go to the grid.

      In winter we run the hot water cycle at night during the Octopus Go hours. This is because peak winter daily demand is around 25 kWh but we only have 13.5 kWh of storage. This means we run out of battery on the coldest evenings and have to buy some peak-time electricity. By running the hot water cycle at night we eke out the time we can run on the battery a little longer and minimise the number of units we have to buy at 43 p/kWh later.

      Regarding the Vaillant control software and documentation, I think it is *appalling*. I wrote about it recently in this article

      An average COP of 4.5 is outstanding – I am getting about 3.7 averaged over a month – that includes DHW and a little heating. Congratulations!

      I am glad you found the blog useful.

      Best wishes


      • Joe Wentworth Says:

        Thank you.
        Makes a lot of sense.

        I will see what the year brings. And how much the cop dips with the cold..

        It’s very true that the vaillant interface is unfriendly. I get the impression it’s intentional to stop users fiddling. Which might be OK if engineers tuned it well before leaving. But in my case there was alot of improvement to the setting from how it was left. It seems just as with combis the goal for installer is no call back and that means ere on the side of hotter is better. Not for efficency though….

        As you say cop is a key factor and perhaps installs of the future might come with a guarantee minimum cop. Much like broadband speeds do.

        How do you find vaillant ‘working figure’ compare to your independent measurements?

        Great work.

      • protonsforbreakfast Says:

        The good news is the the Vaillant appears o be beautifully engineered, so I am sure you will get a good COP.

        And I agree that the unfriendliness is at least in part deliberate. But some features are also just because they don’t care.

        Yes, it would be great to have someone guarantee a COP figure, but at the moment it is hard to find anyone to install heat pumps. Or to get people to quote for them even. And the gas industry is determined to sell as many boilers as they can while they are still legal.

        As I mentioned in the article, the electrical measure is in good agreement with the independent monitoring, but the thermal measurements are just meaningless.

        All the best


      • Joe Wentworth Says:

        Thanks Michael, when you say the thermal measurements are meaningless, what do you mean? a percentage off? out right wrong?

        I don’t have the app or senso unit. but the weather comp unit bar chart for ‘environmental yield’ graph rounds up to round numbers and does feel a bit indicative rather than measured.

        But by looking in the wiring centre menus it does give ‘energy yield’ to one decimal place for the day, month and total, for DWH and heating (and cooling if i had it). These feel like more meaningful figures.

        (as you say it is odd not to have a weekly total. )

        As i have no alternate heat measuring I’m stuck using their figures, but if you could shed any light on the error margin I might be working with it would be really useful.

        Perhaps I could do some calculations with spec heat of water and as i have a temp logger on the HWC and compare theoretical heat needed to raise the water temp and the stated delivered KWh. maybe a day with a legonella cycle and no one drawing off water that day… at least i can see if their number are in the ball park.

        thanks for your insights. Joe

      • protonsforbreakfast Says:


        What do I mean by ‘Meaningless’.
        1. The data on the Senso App is simply missing.
        2. The Senso ‘Unit’ – the wireless controller – does not have this data.
        3. The heat pump control unit has data available by pushing buttons. It’s at the back of the cupboard contains the hot water cylinder and doesn’t give data by (say) date.

        Perhaps I should have used the word ‘unavailable’. It’s just appalling software.

        So I’m afraid I can’t tell you if the Vaillant figures make sense because they are not available to for me to check.

        Good luck with you calculations, but I am afraid I can’t quite see how they will work.

        It’s crazy isn’t it: they could have made this so easy, and they chose not to.

        Best wishes: M

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: