Daikin Altherma 3 M water pump speed

Hardware Heatpump
1

Hello.

We’ve had an Octopus installed Daikin now for about a month and so far, it hasn’t had to much as it’s been relatively mild. But all is well generally speaking.

Last night was the first cold snap and at some point today, I realised the hum of the water pump had stopped. After checking the MMI panel and an energy clamp meter, I was reassured that it was still running but instead of pushing 30lpm like it has been doing since install, it had dropped down to around 10lpm. Blissfully quiet and the rads were all a nice toasty 35c.

It seems like it drops to this lower throughput after an amount of constant compressor runtime - I haven’t determined what that is.

But it raised a question in my mind… if it can operate at that lower lpm, can it do that when it’s milder too? If so, do you know what setting I might need to tweak?

I noticed in the last few weeks while it was mild but still calling for occasional heat that much of the time, the water pump was running whilst the compressor was stopped, so consuming 180w or so just moving water for very little benefit. Do I need to enable sampling mode to optimise that?

We’re running in weather comp mode, of course, 45c at -8c and 25c at 15c if it matters.

thx
Chris

2

Hi Chris,

Welcome, I think you’ll find the forum useful.

I assume you have a 9/11/14/16kW model from the pump consumption.

The circulation pump is an electricity guzzling monster!

I doubt you can hear the pump running indoors, you will be hearing the water in the pipes at 30lpm.

The pump is outside in the heat pump.

All current Daikin monobloc models work the same way

They try to maintain the set dT between flow and return temperatures and they do this by varying the pump speed. They actually use the return temperature as the controlling part.

They aim to maintain a return temperature. If you have a flow temperature set of say 40c and a dT of 5c then the heat pump will aim to maintain a return temperature of 35c using the pump speed. The return temperature is set by the flow temperature set (either fixed lwt or wdc in your case) minus the dT requested.

If the dT is too small it will reduce the pump speed and vice versa.

The flow temperature may vary based on overshoot and modulation settings.

At the beginning of every heating cycle the pump will run at maximum speed for around 20 minutes and then ramp down over a further 10 minutes or so.

10lpm is the minimum pump speed on the 9 to 16kW models so your flow rate has reduced so that the heat pump can get to, or as close as possible to your requested return temperature.

You can adjust the maximum pump speed and set how the pump operates between cycles i using the installer settings in the MMI

I suspect you have been experiencing short cycles up until now and that is why you have not noticed a drop in pump speed, sounds like it has never made it past the first 20 minutes due to warm weather.

25c is too low for these heat pumps, they don’t like it below 30c and often not much below 35c.

It depends on the rest of your installation.

Tell us more.

These are the basics and I would suggest looking at the data on heatpumpmonitor.org to see how these heat pumps behave in detail.

I assume you have no monitoring installed?

3

Hi, thanks for the reply! Yes, we have the 11kw model, EDLA11DA3V3.

I can hear what sounds like the pump, it makes a sorta whining noise at full chat and I can hear it vary a little during the first 20 minutes which sounds about right. Looking at Home Assistant and the power consumption data from the ct clamp, I can see the compressor was running for no more than 20 minutes in cycles last week. So that explains the change in behaviour today.

Am I right in thinking that if it can maintain temp in the rads at 10lpm, that reducing the maximum down a bit should be ok? If so, do you know where in the installer settings I should be looking? I’ve got the installer reference guide and it looks like “Pump limitation” in section [4.7]. I looked at that earlier and it had values like 3.0, 2.5, 2.0 and it didn’t quite make sense to me or if I was in the right area on the MMI.

What’s the issue with 25c vs 30c? Is it short cycles on the compressor resulting in wear?

4

It is not pump limitation. Don’t touch that.

I think it is in pump speed and I think you can set various options like sampling and constant running. I can’t remember as I haven’t looked at it for some time.

The initial start on these Daikins is very aggressive, they don’t work like many others.

They get to the set flow temperature very quickly.

They run at maximum pump speed for about 20 minutes so you get a load of heat straight away, a lot!

The return temperature will rise very quickly although it depends on the ability of your emitters to deliver the heat produced at the time.

When it is warmer this often results in the return temperature rising too much, overshoot helps with this.

The return temperature being too high results in the heating cycle stopping.

Then it restarts about 5 minutes later, but not where it left off, it starts again from scratch at full flow rate.

It ends up with short cycling that you cannot get out off until everything cools down enough.

I limit my pump speed to 60% to make sure get no short cycling but I wouldn’t advise that unless you know what you are doing.

You will see longer cycles by limiting the pump speed, it only applies to space heating and it is the first thing I would do with any Daikin. The first 45 minutes on my 9kW was way too aggressive for my house even with huge radiators.

I don’t know what the issue is with Daikins and low flow temperatures.

They are not happy below 30c and everybody knows this, or at least most people will confirm this. They aren’t even entirely happy below 35c really.

They use just as much electricity running at 30c or below as they do above 30c, produce less heat and just become unstable.

You shouldn’t hear the circulation pump in your house, that doesn’t sound right.

1 Like
5

I would strongly advise you to fit some sort of monitoring, either ESPAltherma or, ideally, the OEM level 3 kit.

There are a lot of adjustments available to you with Daikin and you really need to be able to see the results of any changes you make.

It’s really difficult otherwise.

6

I was wondering about the ESPAltherma but after looking at it a few weeks ago, I thought it was possibly overkill for my needs (this coming from someone who has Home Assistant for virtually all my smart home tech!). But I agree with the need for empirical data and assessing the effects of change.

7

I have a slightly different 11kW model (Daikin hydrosplit) and UFH (so may not be directly applicable) but I would suggest:

Setting pump limitation to 50/60% during sampling. There is no reason you want the pump going full chat when it is doing water pipe freeze prevention if you set back the heating overnight. This shouldn’t affect how the pump is running during normal operation.

Look at setting the overshoot to 4. I found that the Daikin struggles to start at low output and so can quickly surpass the target flow temp and shutdown. Allowing it to overshoot let’s it run for longer without cycling, and it then reduces its output allowing for longer cycles.

Consider dropping the modulation on the Madoka to 0 so it’s not impacting the target flow temps at all.

Another approach (during mild weather) is not dropping the weather compensation quite so low and allow the heat pump to overshoot the house temp and then wait for the hysteresis of the Madoka to kick in so it sits idle for a while.

The jury is probably still out on which one uses less energy in total, but other things to consider are comfort (steady heating better), compressor lifetime and pump noise…

8

I’ve recently adjusted down, as an experiment.

It was set to “80% during sampling”, which I assumed means 100% when the compressor is on, or 80% when it’s just the pump running, to test/circulate the water. I’ve changed it to just “80%” which I assume to mean 80% whether compressor is on or off.

I hadn’t twigged that it ran full speed for the first 20 mins. I see it at two speeds - 24l/m (from 30l/m before the reduction), and then 10l/m.

I may need to rebalance radiators - I had possibly done so when it was running at 30l/m and had to close them right down. But probably went too far so that the drop is too big for some of them now that it’s able to reach a steady state at 10l/m. Unfortunately, one radiator (in the hall) may be getting insufficient flow - even fully open, it drops a lot more than 5 degrees. But things do seem to be working okay, so…

I was also going to limit max power to 2kW, which I think was one of your other suggestions, but when I found a menu there were more options than I expected, so I held off on that for now.

[On a separate topic - in installer mode, there is an option to back up the settings to usb thumb drive. I managed once, but tried to do it again recently, and when I select it, it goes back to the menu very quickly, without writing anything. It’s the same drive, so don’t know what’s changed.]

9

As @matt-drummer has noted on another thread, the hysteresis is rather wide, and it heats up quite a long way before stopping, and then cools 0.5 degrees below target before starting up again. That’s very noticable.

10

I never suggested limiting the power to 2kW, do not do that!

My pump only runs with the heating active, if it is off the pump does not run.

I can’t remember what the setting is but there are a whole range of options.

I am sure somebody here will be able to tell you what the options are and the consequences of any changes.

Your pump is not running at only two speeds, it is variable and it does not just go straight from 30lpm/24lpm to 10 lpm.

Have you looked at the data from Daikin heat pumps on heatpumpmonitor.org?

If you choose some and select show flow rate you will be able to see how they behave.

Have a look at a few, my one in Ipswich is set up to run at the minimum flow rate all the time whereas most others have a variable flow rate.

You will be able to see the start ups and how the flow rate varies. You can usually tell what dT they have set and how it all behaves.

11

Sure. Without monitoring, I just see what’s shown on the ‘sensors’ on the MMI, and those are what I tend to observe. I wonder… should the flow speed tend to hover somewhere in the middle, so that it has room to modulate up and down. If I only ever see them at the bottom, I wonder if that’s a balancing thing.

(A very odd thing I have been noticing… there’s a bypass valve/pipe across the radiator flow/return. I assume it’s there in case radiators close themselves off with TRV. But when I measure the pipe temperature before and after where this meets the return flow, the temperature is usually higher after. Which made we wonder if it was allowing hot water to pass and mix with the return. I think this is better having reduced the pump speed.)

12

My flow rate is fixed firmly at the bottom! It was with my 9kW too.

Your flow rate is at 10lpm because your radiators cannot get rid of the heat produced at your set dT between flow and return.

I see no need to change unless you want a higher flow rate.

To get a higher flow rate you will need to set a lower dT although you still may not get a higher flow rate, it depends on the heat demand at the time and the size of your emitters.

Octopus always fit an auto bypass valve.

Mine is firmly closed as I have no trvs.

Do you have trvs on every radiator?

Could you potentially have the heat pump running with all radiators closed?

13

Yes, Octopus installed TRVs on all radiators (MCS requirement ?) but I have all (but one) fully open. (Kitchen one is currently set to 4 since it gets very warm in there - it’s a big radiator and possibly gets more than its fair share of the water.)

14

this maybe a good stating point?

image
image769×634 35.3 KB

15

I switched to sample yesterday from request on F-0D but I’m still seeing the pump running at full speed when the compressor is not.

Screenshot 2024-11-14 082053
Screenshot 2024-11-14 0820532692×510 47.3 KB

Am I missing something?

[Edit: Figured it out]
F-0D reverted back to 1 because C-07 was set to RT mode. Since the Madoka is not hugely beneficial (to me at least) due to it’s inherent 1.5c overshoot, it seems that switching to LWT mode makes sense. That automatically selects sampling mode. So I’m going to try that with pump speed 60% during sampling and see how we go. I’ve got home assistant managing the set point to prevent overheating (due to solar gain, cooking or whatever else might be going on) anyway so on the face of it, switching to LWT mode seems like a good idea for me. Please correct me if I’m wrong.

Interestingly, Octopus left (or deliberately configured, I’m not sure which), the emitter type as ‘fancoil’. I remember a Heat Geek video saying this was useful as its sets Dt to 5c. So I assume this is correct?

16

dT is adjustable in fan coils between 1c and 10c

17

I have a 9kW Daikin and relatively small radiators. The system can’t modulate low enough for always on operation, and the Madoka has annoying behaviour, so I’m running LWT and using Home Assistant as a thermostat.

I’ll share my setup for information, although I don’t recommend making all these changes unless you are confident to do so. I am way off default now…

I am running 60% pump speed all the time (not just “during sampling” and a power limitation that is currently set to 9A, I’ve experimented with values between 8 and 10A. (Power-based limitation didn’t seem to work for my system, I think there is a minor fault as the MMI doesn’t measure power usage properly either.)

I am also running with fan coils and a delta-T of 4C as that gets my system off the pump-speed floor of 10l/min. I think the first regulation method the system tries to use to maintain temps and delta T is pump speed, so giving it some room to manoeuvre seems helpful. (Otherwise it had to adjust compressor etc.)

I prefer the lower pump speed (60% maxes out at ~18.5l/min) as I have some noisy radiators as the lock shields are almost fully closed to balance the system. Sadly the “first” radiator in the system is in a bedroom…

Note that pump can still run at full flow for DHW.

My aim with the pump and power limitations is to try and match the system power and the pump speed so that:

  1. I get smooth, even heat-up cycles without the delta-T deviating too far from the target of 4C
  2. manage the power draw so that I can stay below 5kW Powerwall power output
  3. Maybe reduce some of the defrosting in cold weather. I figure if the system can’t pull the heat out of the air so quickly, it might not frost up as much? Unproven so far.

The downsides to my approach:

  1. Slow heat-up times. It takes around 1 hour for my system to reach 38C flow from cold, overnight. I just set it to come on earlier.
  2. I have limited the power so much that on really cold days, we might not be able to maintain or raise house temperature. I will adjust if needed.

I am pleased with how my system is running, averaging a COP of over 4 for the last 30 days. (Yes, it has been mild, but my system was terrible at the start - mostly due to a tape heater that was on whenever the pump was running…)

https://emoncms.org/app?readkey=0e9ab50e73d8ae59d9e6dcc70e5dd5f7&mode=daily&start=1728950400&end=1731628800#myelectric

1 Like
18

Interesting, I never managed to get power limitation to work when I thought about using it to restrict DHW efficiency but can’t recall if I tried amps.

19

Do you use HA to modulate the flow temperature, or just as an on/off switch ?

20

No adjustment of flow temperature currently. Just on/off based on various temperature inputs and a schedule.

I’m done trying to balance the heat loss and the heat input. For a good part of the year, it just doesn’t work for our property and the 9kW. So the working model is: heat up, switch off, cool down a bit, heat up… etc.

I’m getting good duty cycles and I can better control the on/off thresholds than with the :face_with_symbols_over_mouth: Madoka. House is comfortable, COP is good, running costs are low - it’s working well for us, even if it’s not how you are “supposed” to run a heat pump.