A review of 700 heat pump hot water runs in 2024

A review of 700 heat pump hot water runs in 2024

Throughout the whole of 2024 I ran my 5kW Vaillant Arotherm heat pump in Eco hot water mode, heating our 250L Mixergy cylinder in Sheffield. 

This article will be a deep dive into almost 700 hot water runs across the whole year.

The Arotherm heat pump has three hot water modes; Eco, Normal and Balance.

You can read about the differences here:

Arotherm Plus Hot Water Modes

Essentially, Eco is the most efficient option leading to better COP, which is why I chose it for this year.

Through all of 2024 we have done two hot water reheats, one at 1am and the second at 1pm.

We are a family of four including two teenagers, so we have a reasonable hot water demand that can’t be satisfied by a single cylinder reheat.

To try and be as efficient as possible we chose the following target temperatures for the cylinder.

• January to end April: 48C
• May to end December: 45C

By the month of May I realised I was ending the day with a surplus of water when heating to 48C, so dropped the target temperature down a little to 45C for the rest of the year.

You can read in this article how you might be able to calculate your own hot water demand and target temperature.

Hot Water Cylinder Sizing and Mixing

Collecting all the data

Using a combination of Open Energy Monitor and the data from the Mixergy IO platform I have been able to pull the following data elements together for each individual hot water reheat.

• AM or PM
• Average Outside Temperature during hot water reheat
• Electrical Input (kWh)
• Heat Output (kWh)
• COP of Run (output divided by input)
• Starting % State of Charge (SOC)

This led to me transposing figures from around 700 hot water runs into Excel for analysis.

Note: not EVERY hot water reheat is included from 2024.  

I think I removed about 35 runs through the year to tidy up the data.

For example, some reheats that were done in Normal mode, some days where I’d only taken one run (so no AM and PM for comparison), some runs where I’d heated to 60C coming back off holiday for legionella protection, huge chunks of water drawn in the middle of a reheat that skewed figures etc etc.

You can download the raw data here if you want to take a look:

https://energy-stats.uk/wp-content/uploads/2025/01/energy-stats-Mixergy-2024.xlsx

Headline Performance Figures

Across the whole of 2024 running Eco mode we achieved a hot water COP of 3.84

• Heating output: 7877 kWh (68% of usage) – Heating only SCOP: 4.45
• Hot Water output: 3682 kWh (32% of usage) – Hot Water only SCOP: 3.84

You can find a whole house breakdown for all 2024 in this article

Whole House Running Costs and Stats 2024

This article will show you how to measure your own Arotherm COP / performance values

How to Measure Vaillant Arotherm COP

Here is a graph showing the outside temperature and COP from every recorded Eco mode run in 2024.

As you can see from the graph, the single best performing reheat came when it was 12.1C outside.  Curiously, this was on Christmas Eve with a COP of 4.83

The Mixergy cylinder started from 16%, so not even fully empty.

It was a surprise that one of the summer reheats when it was over 20C outside didn’t have the best performance.

I wonder what circumstances or enviromentals caused that one in December to be the best?

One of my thoughts is because the incoming cold water / bottom of the tank was much lower in winter than in summer?  The bottom of the tank was only 8.1C on Christmas Eve.

On the reheat of 28th July, with a COP 4.72, despite the cylinder emptier at 13% full, the bottom of the cylinder was 17.8C.

This is because the incoming cold water is always colder in winter.  I’ve seen as low as 5C for us and as high as 20C in summer.

Because of this, the Christmas Eve run started at a much cooler point, meaning better early COP in the run, aided by the fact it was 12C outside.

You can view the run yourself here: (put in COP lines?)

https://emoncms.org/energystatsuk?mode=power&start=1735038420&end=1735046860

If we remember back to the “Smaller heat pumps will never reach top of the scops” article.  There seems to be a performance ceiling around 15C outside.

Smaller Heat Pumps will never reach Top of the SCOPS

So that would make for a good “high SCOP” experiment.  15C outside and really cold incoming water.

Overnight versus Afternoon Temperatures

One of the golden rules for efficient hot water production with a heat pump is to schedule reheats at the highest outside temperature.

Here’s a recap of the four golden rules.

• Reheat from empty cylinder / lowest temperatures (high hysteresis) – no quick top-ups
• Choose lowest target hot water temperature you can get away with
• Use Eco mode (or some other low compressor / quiet mode)
• Schedule reheats when highest outside temperature

The elephant in the room of course is that most off-peak electricity tariffs offer cheaper rates overnight, when it’s usually colder.

So I wanted to analyse the difference between AM and PM temperatures and the effect this had on COP.

With me doing hot water reheats everyday at 1AM and 1PM I was in a perfect position to do this.

Here’s a chart showing the average temperature during every hot water reheat in the dataset across the whole year.

The blue line being the 1AM runs and the red line being the 1PM runs here in Sheffield.  Obviously, your temperatures may vary.

By looking at the daily details within the data I can work out the temperature difference between the AM and PM runs.

Here is a comparison of every run in 2024, summarised and averaged by AM and PM.

First observation is that the outdoor temperature golden rule holds true; a warmer outside temperature leads to better COP.

But the headline takeaway for me is there isn’t much to choose between the AM and PM outdoor temperatures when viewed across the whole year.

Certainly not that much to gain enough COP to outweigh the financial gain by running overnight on cheaper electricity.

For example, say we need 10 kWh of energy to heat the water (this would be approx 215 litres heated from 10C to 50C)

• AM: 10 kWh / 3.63 COP = 2.75 kWh electricity x 10p per kWh (cheap rate) = 27.5p per hot water run
• PM 10 kWh / 3.86 COP = 2.59 kWh electricity x 24p per kWh (price cap) = 62p per hot water run

Obviously, you can adjust the figures in the calculation for your own off peak / peak pricing / water demand.

You can use this calculator for water heating times

https://gettopics.com/en/calc/water-heating-calculators

Let’s look at month to month averages, how do they look?  Are there any advantages to switching approaches at different times of the year?

Again, the data says not really.  As we saw in the earlier temperature graph, the overnight and afternoon temperatures rise and fall with the seasons and the differences between AM and PM isn’t that much.  Not enough to warrant changing habits?

Reheat from as empty a cylinder as possible

One of the other golden rules of high performance hot water generation is to start the reheat from as empty/cold a cylinder as possible.

This can be achieved by either setting a high hysteresis or just letting the cylinder be used, allowing the temperature of the water to drop low and then top up on a schedule.  

I do the latter with my Mixergy cylinder.

Between our two reheats a day I let my cylinder drop to whatever level of emptiness it reaches.  Sometimes it can be 50% full when the next reheat starts.  Sometimes it can be 5% full.  All depends if we’ve all jumped in the shower or not between reheats.

Because the Mixergy works slightly differently to a conventional cylinder in that it records a percentage state of charge (SOC) I have been able to record the starting SOC for every hot water reheat alongside all the other data.

You can see in this example from the Mixergy IO platform that despite being only 8% full I still had 42.8C usable water in the top of the tank (target was 45C).

What I wanted to achieve with this analysis is to see what effect, if any, starting from an empty cylinder had on COP.  And whether there was a certain level of emptiness that you started to see better efficiency.  Basically, see if the golden rule rang true.

Having almost 700 data points across the year I have been able to group these runs into temperature bands.  Because there’s no point comparing a hot water run when it’s 2C outside to one when it’s 22C outside.

Firstly, let’s look at graphs for each temperature band starting at 4C to 4.9C and going up in 1C increments.

Note: below 4C the data is very muddy because of defrosts and to be honest, there wasn’t a great amount of usable data.

Remember, these runs are a combination of the following target temperatures across the whole year.

• January to end April: 48C
• May to end December: 45C

Now, let’s group all these data points into 3 distinct bands of outdoor temperature

• 4C to 9.9C
• 10C to 14.9C
• 15C to 25.7C

So what can we see in all those graphs?

Well I can instantly see that the golden rule holds up.

If you start the reheat from an emptier / colder state, then with everything else being equal you will see better performance / higher COP.

And if we look for a tipping points in the data and graphs, i’d say there are two main markers; 60% and 40%

• Reheats starting from higher than 60% full seem to perform pretty badly and should be avoided.
• 60% – 40% seems to be the bare minimum for reasonable performance.
• Below 40% is where you really start to see great performance.
• And below 20% is where really good things start to happen.

How to make use of this data with a conventional cylinder?

To make use of the data and ensure long hot water reheats you need to work out what hysteresis to use and that can be tricky.

The thing that makes it hard to give recommendations around hysteresis is that the location of where the temperature probe is on the cylinder can vary between cylinder manufacturers.

Some might be higher than others, some might be lower.

The temperature probe reports back to the heat pump a live report of the temperature of the hot water in the cylinder from the probe.  So probe location matters a lot.

On the Heat Geek Super Cylinder (left diagram) you can see the primary temperature probe location is around the middle of the cylinder.

Whereas on this OSO Geocoil (right diagram), it’s more like two-thirds up.

These alternate locations would report different temperatures and require slightly different setups.

But regardless of temperature probe location you first need to note at what water temperature reported by the probe you find the water unusable.  This is highly dependent on probe location and your usage.

For the sake of the forthcoming example, let’s say you are happy with a 40C bottom limit.

The second setting you need to know is what is your target temperature.

The hotter the water stored, the more usable water you will get, but the hotter you go, the less efficient it will be to produce it (back to the golden rules).

For the sake of the forthcoming example, let’s say 50C target temperature.

You would then make this simple calculation to work out your hysteresis:

Hysteresis = Target Temperature minus Bottom Limit Temperature.

In this case, 50C minus 40C equals an hysteresis of 10K 

Note: 10K is just another way of saying 10C.

Once the temperature probe detects 40C water (target minus hysteresis), assuming a hot water schedule is active, a reheat will start.

If your target temperature is 45C and you know that showers will start getting cold when the water hits 35C, then you probably want to put your hysteresis at 10K.

If your target temperature is 48C and you know that showers will start getting cold when the water hits 42C, then you probably want to put your hysteresis at 6K.

If you are heating your cylinder to a high temperature, say 55C, then your hysteresis can be much wider if you’re still happy with a 42C reheat trigger.  You’d have your hysteresis set to 13K.

So it all depends on what your target temp is and how cold you’re happy to let the cylinder get before reheating.  

Some people might let the result of the calculation go as low as 35C.  But this can sometimes depend on the location of the temperature probe within the cylinder.

All the above examples also assume you’re okay with no usable hot water once you hit the bottom limit.  As it will take some time to heat enough water to satisfy your demand.

In this case, you may need to increase your hysteresis so that you start heating water before you run out.  It’s all a balance.

Bottom line, the lower the result of “target temp minus hysteresis” calculation, the colder a starting point of your reheats and as such, better performance.

If you’re seeing poor hot water performance you may have got your hysteresis very low, say 2K.  This will lead to very short top up reheats, which we’ve shown to be inefficient.  But you’ll probably not run out of water with this approach.  If you have no desire to chase efficiency and care more about always having hot water, this is likely the best approach.

But as with all of this, “it depends”.

It depends how much hot water you need, whether you need top ups, whether you can wait for single long reheats and what size of your cylinder you have.  

All these factors will dictate the target temperature, hysteresis and scheduling.  Not forgetting whether there’s any off peak electricity in play.

You could set your hysteresis so wide that it never triggers and just use scheduled reheats at the times of your off peak electricity.

Summary and Round Up

I hope you’ve found this analysis useful.  It’s certainly been fun and eye opening pulling all the data together.

I’ve particularly enjoyed the fact that it has reinforced some of the golden rules around hot water efficiency that we’ve been preaching for some time.

• Reheat from empty cylinder / lowest temperatures (high hysteresis) – no quick top-ups
• Choose lowest target hot water temperature you can get away with
• Use Eco mode (or some other low compressor / quiet mode)
• Schedule reheats when highest outside temperature

Although on that final rule, the data  in the article does show that an overnight cheaper electricity rate likely beats the benefits from a higher afternoon outside temperature.

One point to mention is that it’s not all rosy in the garden with Eco mode though as I’ve found out through this winter, especially with my smaller 5kW heat pump.

You can read about some of my struggles here

The Problem with Eco Hot Water Mode

Finally, massive thanks to Open Energy Monitor for their awesome heat pump monitoring package.

https://openenergymonitor.org/

You can find my heat pump here: https://emoncms.org/energystatsuk

And it is listed here amongst 100’s of others on https://heatpumpmonitor.org/