Just thought I’d mention that the Guardian are asking for opinions on people’s heat pump experience.
Of course I’ve provided mine.
Just thought I’d mention that the Guardian are asking for opinions on people’s heat pump experience.
Of course I’ve provided mine.
Done. Cheers for that link.
I suspect this is the outcome of the survey.
It’s a bit odd how Mitsubishi come out so well in each story. And that’s coming from me who has one.
Interesting article, I saw the call for contributions, I wonder if this is the first part of many or the final piece…
How are you getting on with your heat pump @MyForest have you done any further optimisations?
@muzza what do you have? have you been monitoring it’s performance?
Here’s a screenshot of a how water heat up cycle from mine yesterday, COP of 4.28 reaching a cylinder temperature of 53.5C with the flow temp reaching 58.8C after a slow controlled temperature ramp from cold:
I have a 5kW Mitsubushi EcoDan with a 3m2 coil in a newark cylinder, I’ve got the FTC2b flow temperature controller for the EcoDan which lets me control the flow temperature using a voltage signal created using a digital to analog converter connected to a RaspberryPi…
a 5kW Mitsubushi EcoDan
Your 5kW scales down way better than our 14kW one. It’s amazing to see yours being gentle and warming over a two hour stretch and hitting a CoP of over 4. Ours is like a tiger waiting to be unleashed and it races off to heat the water really quickly, so mostly it’s doing hot water at around a CoP of 2.5 even on a warm Summer day. I’m reluctant to tinker as much as you whilst I’m still getting the RHI because it cost us 15k for the system and I won’t save that much more by taking the bigger risk of playing inside the flow-temp controller (FTC).
I’ll talk through this example so you can feel even better about your 5kW one.
During this run we had 1.4kW solar (the yellow bar at the top) but the heat pump got as high as using 5.5kW at the end. I only have coarse-grained control using the web-based interface so I can’t tell it to operate more gently. It’s actually much worse with the hot water, when doing the space heating I have more control by fiddling with the target temperature and making it rise gently but if I set the hot water to a lower temp it gives up and turns it off.
You’re probably looking at that “desired” hot water temp of 41 °C and pondering the Legionella thing. The Ecodan is pretty blunt how it does that so we’ve overridden it and the computer watches for a time when the temp was high for a long time which would kill the Legionella. If it’s been too long since that happened it bumps up the target temp to zap 'em.
Now the fun bit. The last month has cost us 5.54GBP in hot water for a family of four including two teenagers, so quite a bit of water. The trick of course is a solar diverter. You can see the behaviour with this run:
To start with we’re getting dinged 7p for that day because the infrastructure around the heat pump uses electricity in monitoring and interacting with the internet. The hardware didn’t do anything at all - it was purposefully switched off by my program the whole time.
However, all the hot water came from the solar diverter. You can see around 15:00 that someone took a shower and it recovered pretty quickly. We’ve actually had to limit the immersion thermostat to 55 °C because it’s at the bottom of the tank and the top was getting so hot when the thermostat was at 60 °C that we couldn’t mix in enough cold to make a sensible shower. Luckily the local humans are quite good at taking showers in the middle of the day to absorb the solar.
You might also note the “flow” from the hot water was getting up to 30 °C even though the water pump wasn’t running. That’s due to the heat building up in the tank leaking out along the piping enough to warm it up where the Ecodan’s sensor is. Ho hum.
Here’s another run over a couple of days where there were five showers powered purely by the solar PV diverter:
In fact at one point we didn’t notice the circuit to the heat pump was off for a week because the solar kept it warm enough for all our showers. (there had been a power-cut which tripped it off)
I do sometimes manually nudge the heat pump on if there’s just a few hours of sunshine and I think it’s worth using the extra from the grid to multiply up by the heat pump CoP. Of course that should be automated but sadly there’s where my story is bad because something happened in March 2020 to distract me from working on the heat pump and in January 2021 I started a more serious job which is consuming all my time.
So right now it’s all humming along nicely and doing roughly the right thing.
My 16 year-old daughter said yesterday “when I move to my own house can I use your heat pump controller?” which is about the best review I can expect from someone who lives with the consequences of how it works.
Thanks for sharing @MyForest, yes I can imagine 14kW for DHW is going to push those temps up quickly!
Do you have a large surface area coil or external heat exchanger? What is the deltaT between the max flow temperature and the DHW cylinder temp that you usually set it to reach?
Here’s my flow temperature and DHW cylinder temp about 3/4 of the way up, suggesting an average DeltaT of 7.0K. The difference between the average coil temperature and the DHW cylinder temp is about 4.5K…
It’s great to see your results in your other post here showing how you have managed to improve the COP using your control algorithm, great! Heat pump experiment review after two years
Thanks for the info.
On a somewhat both related and not related topic and as currently a novice on the topic of Air Source Heat Pumps, what central heating control method(s) do they use? By this I am referring to the fact that traditional boiler heating systems originally used ‘call for heat’ control whereby a room thermostat sends a signal saying either ‘I need more heat’ or ‘satisfied’ meaning it has reached the target temperature. New boiler systems can use modulating control using either the OpenTherm or very similar eBus standard used by Vaillant and Worcester Bosch.
Would an ASHP support either call for heat or OpenTherm or both or something else?
I am considering getting an ASHP to replace a gas boiler currently feeding in to radiators and it currently uses a smart thermostat using OpenTherm. I have been looking at a Daikin Altherma and matching tank.
Nothing that sophisticated (as documented by others here in this forum!) When I got my Solar installed 5 years ago my journey began. Installed a Pi and connected it to the inverter to monitor the Sun. Its still running today and I have lots of stats on power, energy, temps, in total and for each string of panels (4.5 killer wasps).
After a while, I though I need more, so invested in eMonPi to see more details on house usage and exports to the grid. Now, Im Totally into all things measured.
Back then, with a very troublesome, inefficient Gas ducted heating system and with Gas prices here rising sharply, and with an increasing desire to contribute to preventing climate disaster, I moved my attention to eliminate Gas use at my home. My Gas hot water heater was approaching End of Life so I swapped over to heat pump technology. I had done some research and thought I could make better use of my day time Solar generation. I am now a huge fan (pun intended) of Heat Pumps.
When my Hot water Heat Pump was installed a few years ago, I was always wanting to check it running and to measure if my investment was good, etc … I could scan the house power and Solar generation graphs etc but some times the heat pump operation was ‘hidden’. Then I thought I have the ability to connect some temp sensors to the eMonPi, how could I make use of this.
So I have two sensors connected, one on one side of the heat pump fan, and one on the other. Whenever the unit runs I see a dramatic drop of temp on one side. And it looks like this:
The Sanden Heat Pump is programmed with a blockout timer so I have it set to only come on after 1100am to make best use of my Solar generation. From the graph it ran for about an hour today. It will also come on occasionally for defrost mode once or twice a year, so I can see that happen as well.
The temp differential today (miserable rainy winter day) is not that great, but in the middle of Summer on a 43C day the best place to walk past is the Hot Water ‘heater’. !
So just an example of a very simple way (two temp sensors floating near a fan) to monitor something. Sure it would be useful to have specific probes on water temps, power usage, etc etc, maybe one day.
PS I have eliminated all Gas and disconnected Gas meter by the way… pulled out the old furnace and rubbish ducting, and installed … more Heat Pumps, what else !
Fantastic @muzza sounds like a good system! And a nice example of simple monitoring.
I helped a friend install an EmonTx on an air-to-air heat pump a couple of weeks back, we put a CT sensor on the electricity supply and a temperature sensor either side of the indoor blower unit and that works really well to gain a lot of insight.
@jelockwood The air-source heat pumps do modulate their output, but I don’t think they use OpenTherm. My installer fitted a very simple call-for-heat thermostat. It has some delays added so it doesn’t just flick on and off when it’s hovering around the right temperature.
However, I found (with my too-big heat pump) that it was coming on and off quite often because the heat pump couldn’t modulate low enough to provide just a little bit of heat. It’s not helped by my radiators which can release some heat but aren’t as good as underfloor heating. I suspect if you get a sensible-sized heat pump it would be OK. I see people like Trystan showing his modulating down and running constantly. I prefer to turn mine off, but that’s overkill.
The heat pump is trying to maintain a difference between the flow and the return. With an Ecodan I believe it stops itself when the return gets within 3 Celsius of the flow.
So I suspect you’ll find it doesn’t need OpenTherm because it’s modulating based on the flow temperature.
In my case I actually don’t even have any room stats. Each room has a local TRV and that shuts the radiator down when that room gets too warm. Eventually most of the radiators shut down and the heat pump decides it’s not worth carrying on and stops. My system turns the heat pump back on again when the temp in the pipes gets “low” according to a stupidly complex algorithm I wrote which takes account of wind-speed, solar gain, humidity, time of day, time of year and more.
For extra kicks, my system tinkers with the “target flow temp” to try and optimise how the heat pump works to improve the CoP. It’s still not as fine-grained control as ones like Trystan’s though.
You can see in this screenshot that my system is gently increasing the “Desired” space heating (the blue line) from 30 Celsius. This stops the heat pump trying really hard which is does if I start the cycle with the desired temperature that the high point of 38 Celsius. It’s getting me a few percent improvement on my CoP, but even if yours starts at 38 Celsius it’ll be able to modulate better than mine so it won’t race up to the higher temps.
Do you have a large surface area coil or external heat exchanger?
We have a small external heat exchanger. I keep telling myself it’s a good idea because it can be replaced without the whole cylinder needing to be replaced.
What is the deltaT between the max flow temperature and the DHW cylinder temp that you usually set it to reach?
The deltaT is very high. The system runs away and I can’t stop it. I’m sure if it would just calm down that it wouldn’t need to go up to 60 °C flow in order to heat the water to the desired temp of 41 °C. The flow-return delta is about 8 °C.
Having said that, I’m OK with it costing 12p for a shower. I’ve decided to just accept it.
Also, I suppose it’s nice it can heat up so quickly because it can either do DHW or space heating so blocking the space heating for a couple of hours would be bad when it’s cold.
In the Winter it’s more sensible because it bounces between hot water and heating nicely and the over-run is used to make the radiators rather toasty. It makes the pipes tick and the humans flock to the radiator to get 1 minute of extra warmth before it goes back into sensible mode and does space heating at thirty-something, like this:
In that case someone took a shower around noon and so the algorithm decided to get the hot water sorted out because it had dropped so low and it was daytime. Then it decided that the hot water was done and switched to heating the house because the house had cooled down since the previous run ended at 11:40.
Here’s a really bad run where it did space heating, worked out the water was cold enough to heat and then used the overrun to do a few minutes space heating:
So the space heating efficiency looks great (for December) but the hot water looks terrible.