Ecodan standby power consumption

[PhilT]

On 29/06/2022 at 08:10, JamesPa said:

Your measurements are very plausibly consistent with the data Mitsubishi gave me and which I posted above. 

 

1kWh/day is 40W.  Mitsubishi says its 15W for the outdoor unit, leaving 25W unaccounted for.  I cant find a spec for the controller, but Mitsubishi do state that the standby power of their cylinders (which include the controller) as 26-45W.  Apart from the controller, the cylinders are mostly plumbing, so you would expect the standby power to be mostly the controller plus perhaps one or two solenoid valves thrown in (or something similar).  Thus 25W for the controller seems a very likely (and anyway quite plausible) figure, albeit one that would probably make it feel quite warm to the touch.  The controller is never really on 'standby' I suspect, its sitting there 'doing its thing' all the time.

 

All in all I think your measurement plus the Mitsubishi product fiche convinces me that the PUZ-WM112VAA R32 is very likely ~15W, with the controller very likely ~25W.  These are, to my mind, both reasonable figures and nowhere near the 200W that started this thread.  There does seem to be evidence (from Mitsubishi and others) that some Mitsubishi models, particularly the PUHZ (R410) 3-phase models, have a significantly higher consumption.  if anyone has any further figures for specified models then it would be interesting to aggregate them and check out further the various data Mitsubishi have provided.

Seems reasonable. I plead guilty for my early contribution to the apparently high power draw. My house was using around 7kWh per day back in April while I was away with no heating or DHW left on. One reason was a very inefficient 30 year old fridge/freezer which has been replaced saving around 2kWh a day. The other reason is the 1.4kWh per day of power consumed by all the "idling" devices in my house - TV/box standby, internet router, land line phones, power line adaptors etc, some of which will be culled as a result. All in all a useful learning curve!

"Idling" Appliances		W	hrs/day	kWh/day
Router		12	24	0.288
Power line adaptor		3	24	0.072
Phone/base station		2	24	0.048
Night light		9	8	0.072
TV standby		11	24	0.264
BT box		7	24	0.168
Power line adaptor		3	24	0.072
Kitchen phone		2	24	0.048
Study phone/printer		2	24	0.048
Bedroom phone		2	14	0.028
Landing night light		12	8	0.096
Spare PLA (now disconnected)		3	24	0.072
Aerial amplifier (loft)		5	24	0.12
				1.396
Fridge/Freezer (saved 2kWh/day)		35	24	0.85
				2.246
Heat Pump Idle	max	45	24	1.08
(PDF data guess)	min	37	24	0.888
TOTAL IDLE	max			3.326
	min			3.134

[SteamyTea]

26 minutes ago, PhilT said:

will be culled as a result. All in all a useful learning curve

Yes, took me a while to find out what was gobbling power and what I could get rid of.

 

It was a sudden increase in daily usage that made me realise that something was wrong.  Fridge had gone faulty.

Edited June 30, 2022 by SteamyTea

[BEJB]

I've come to this discussion very late. The hot weather has meant that my ASHPs have not been delivering heat for a month or so, so I've ended up looking at Melcloud reports that only show 'standby' figures (accepting these are just 'indicative'). And I've been quite shocked.

 

Our house used to be two small cottages, so it has two systems. Cottage 1 was installed in 2015 and the outside unit is a PUHZ-HW140VHA2-BS. This is a larger system, and has 2 zones (rads and UFH for the kitchen) and DHW with a pre-packaged cylinder (for both cottages). There are solar panels and a solar diverter (which has allowed me to turn off the DHW recently, so I can see the standby for this system.

 

Cottage 2 has a much smaller system that was installed in 2020 with a PUHZ-W85VAA outside unit. This just heats a few rads - nothing complicated.

 

In recent days, when I have been able to have both systems 'on standby' the Melcloud reports show the larger Cottage 1 system (PUHZ-HW140VHA2-BS) has a daily energy consumption of 0.45kWh (equiv average power 19W). The smaller Cottage 2 system (PUHZ-W85VAA) has a daily energy draw of 2.5kWh (equiv average power - 104W). 

 

I have been in email contact with Mitsubishi and they have directed me to the Melcloud developer, to see if the reports are wrong. I think it may be possible that the PUHZ-HW140VHA2-BS report is wrong because Melcloud doesn't recognise that model number, whereas it does recognise the newer smaller system.

 

I'll try and use my old 'smart meter' and get a clamp on some power feeds to measure more accurately.

 

 

[JamesPa]

There is some evidence that the PUHZ (R410 - older) models perform differently to the PUZ (R32 - newer) models.  Attached a document posted by another contributor with PUHZ data, purportedly from Mitsubishi (although Mitsubishi have subsequently denied it) and an extract from the Mitsubishi R32 ATW datasheet (for the 11.2kW model and others) which (for the 11.2kW model at least) is pretty consistent with measurements reported on this forum.

 

Basically I think the overall conclusion is that its model and generation (in reality - presumably compressor and FTC model) dependent.  

Standby Power Consumption.pdf wm112xx_001_001_uk .pdf

[BEJB]

Many thanks for those comments. 

 

I have been in discussion with the Melcloud team over the reports for my PUHZ-W85VAA system. The consumed estimate report shows a 'standby' estimate (ie no heat being generated) of just over 2.5kWh/day which is shocking, particularly at the cost per unit now.

 

I have been able to turn off all other circuits in the house, so the only thing running is the heat pump (and FTC etc). Then I watched the power meter for the house (a bit like watching paint dry, and the meter kept ticking over overnight when I wasn't looking, so it has some inaccuracies). It took between 45 and 37 hours to consume 1kWh. I calculate that a steady standby usage during that time would be between 0.65kWh/day (27W) and 0.53kWh/day (22W)? This is very similar to the figures in the Standby document for a 'similar' ASHP.

 

Interestingly, I also have my own clamp ‘smart meter’ attached and that shows a much higher figure of about 160-450w or a total of 7-11kWh/day – again as the Standby document says it would.

 

So my conclusion s that, as it stands neither clamp meters or the Melcloud reports are accurate ways to measure the standby power usage.

 

I confess that I had missed the fact that the estimate is actually made in the heatpump itself and just collected regularly by Melcloud. I had rather assumed that the Melcloud system itself calculated the estimate. That feels like it will make it more difficult to alter the estimation routine to make it more accurate.

 

I'm hoping they will take this information and come out with an improved estimate. 

Edited July 27, 2022 by BEJB

[SteamyTea]

15 minutes ago, BEJB said:

I'm hoping they will take this information and come out with an improved estimate

A few of us are mucking about with wires and code, hoping to make a cheap and accurate energy meter.

May be worth spending 25 quid and knocking one up.

 

https://forum.buildhub.org.uk/blogs/entry/946-the-energy-meter-experiment/page/2/?tab=comments#comment-5986

[AdrianP]

I found this website after the antique NPower supplied clamp on power meter showed that we had about 240W usage, with hot water and heating off, even with all of the breakers switched except for the ASHP.

We have a PUZ-HWM140VHA i.e. 14kW R32 single phase.

According to the meter (originally installed with the unit and looking very similar to the picture on a previous page) it went from xxx87.57 kWh at 1025hrs to xxx88.09 kWh at 2028hrs. That is 0.52kWh over 10 hours or 52W.

I also found a document online, called Ecodan ATW data book V5.3 (72 Mb so cannot upload) which gives all of the tech data for current machines and it lists a hermetic scroll compressor for the 11.2 and 14kW units, but hermetic twin rotary for the 5.0 and 8.5 kW units, which may explain the differences between unit types.

I may, if I can persuade my wife to go out for a few hours due to lack of internet, repeat this for three hours tomorrow and check the meter reading outside.

[Radian]

1 hour ago, AdrianP said:

antique NPower supplied clamp on power meter

You may be aware that it almost certainly doesn't take into account Power Factor, so may over-report. You have the right approach in using the utility meter to check consumption as these do account for PF and truly represent your billed usage.

[PhilT]

11 hours ago, AdrianP said:

That is 0.52kWh over 10 hours or 52W.

 

That's about right. My 11.2 version is similar

[AdrianP]

I was unsure how accurate the digital counter would be for the ASHP so (at great personal distress due to lack or internet access over 2 hours 😞) I opened all of the fuses except for the ASHP and photographed the silly clamp on joke, the digital meter and the consumer meter outside over around two hours. Bearing in mind that the digital meter is only changing one digit every 26 minutes it was not massively accurate at this low power usage but the digital and consumer unit were similar and the clamp on device was away with the fairies. An extra (unphotographed) measurement made the digital and consumer results match.

The recorded usage was 31 vs 31 vs 235W for the consumer unit, the digital meter and the clamp on device. The digital meter is therefore accurate when measured over a reasonable time.

This was over a reasonably warm mid day period, and is lower than my results yesterday (mid-day to late evening), so it may be that the unit is trying to minimise power through some form of monitoring. Overall, the power usage is much less than I originally thought from the joke monitor.

[Kelvin]

My installer has provided me with a very reasonable quote. I checked the model they have specced: 

 

Mitsubishi Ecodan PUZ-WM85VAA-BS (R32) + FTC6

 

Typical. Although it seems to be at the lower end of the scale in terms of standby energy use. 

[BEJB]

OK, I have a bit of an update.

 

I contacted Melcloud to ask why our smaller system (PUHZ-W85VAA) had a standby figure of about 2.5kWh/day but our larger system (PUHZ-HW140VHA2-BS) had a standby usage of only 0.4-0.5 kWh/day. Apparently Melcloud just takes the estimated data from the FTC5 controller, so I checked that, which confirmed that that figure was the 2.5kWh/day.

 

So they spoke to one of their engineers, who said the figure should not be that high. They decided that they wanted to install one of their standard monitoring systems on the W85VAA to see what its actual use was. They have done that and I ran a standby test yesterday (as it was very mild). That showed that the actual standby usage for the whole system was only about 0.3kWh for the day - or between 10-15W.

 

I've reported that back to them. I'm hoping that it allows them to check their firmware and improve their standby estimates for people without monitoring systems installed.

 

So all I can suggest for the moment is that if you see very high standby usage, take it very much with a pinch of salt. 

 

I've included some pictures. The first is BEFORE – estimated with no monitoring installation – standby on 13^th July.

 

The second and third are AFTER – with monitoring installed – standby on 10^th November.

 

Cheers

Eric

[joth]

19 minutes ago, BEJB said:

So they spoke to one of their engineers, who said the figure should not be that high. They decided that they wanted to install one of their standard monitoring systems on the W85VAA to see what its actual use was. They have done that and I ran a standby test yesterday (as it was very mild). That showed that the actual standby usage for the whole system was only about 0.3kWh for the day - or between 10-15W.

 

I have the same model, and my own energy metering reported the same number you see. But useful to have it confirmed the melcloud numbers are vapourware. I was in the middle of writing sidebar details of a PH+ article based on it, I'll have to think again.

 

[SteamyTea]

24 minutes ago, BEJB said:

So they spoke to one of their engineers, who said the figure should not be that high. They decided that they wanted to install one of their standard monitoring systems on the W85VAA to see what its actual use was

 

3 minutes ago, joth said:

I have the same model, and my own energy metering reported the same number you see

Just goes to show how useful proper energy monitoring and basic analysis can be.

And it only cost a few quid.

[PhilT]

1 hour ago, BEJB said:

OK, I have a bit of an update.

Thanks for sharing. Do you know if that is just for the heat pump itself, or for both pump and controls?

[joth]

3 hours ago, SteamyTea said:

 

Just goes to show how useful proper energy monitoring and basic analysis can be.

And it only cost a few quid.

yes! And most annoyingly, I even have multiple versions of that logging here already, 7 channels on emonpi and 3 pzem_004t on ESPHome. It was just lack of care with InfluxDB instances (call me BOFH), and repurposing things to other circuits thinking the MELCloud logging already had me covered at better fidelity (DHW vs cooling vs heating) that meant I didn't leave it on the ASHP consistently 😞

 

 

[Ewan]

I’ve read all the posts here but not really sure if there’s now any ecodans to avoid. The 5/8kw ones?
 

would anyone be kind enough to summarise?

 

[BEJB]

On 11/11/2022 at 12:32, PhilT said:

Thanks for sharing. Do you know if that is just for the heat pump itself, or for both pump and controls?

Sorry for the delay in replying. The power supply goes from the consumer unit to the outside unit, which is then T-ed off, through the wall to power the indoor unit. They just installed a single power meter, which is before the T, so it is measuring everything - heat pump, controls, circulation pump - the lot.

[BEJB]

In case it helps, I thought I'd also try to get an idea of how much the freeze-stat function uses. So on a cold night, I disabled the heating, but let the freeze-stat function operate. It was a cold night so it was operating continuously overnight - keeping the flow temperature at 20C. I took a screenshot the next morning at about twenty-five to seven. It seemed to show it was using about 0.065-0.070 kWh per hour.

 

Standby (above) seemed to take about 0.01 to 0.015 kWh/h. So the freeze-stat function seemed to take an additional 0.055 kWh/h (or 55W - or about 1 1/3 kWh/day if it's running all day).

[BEJB]

18 hours ago, Ewan said:

I’ve read all the posts here but not really sure if there’s now any ecodans to avoid. The 5/8kw ones?
 

would anyone be kind enough to summarise?

 

I'm not sure I can help much. Generally, I think Ecodans are excellent, and I'm not sure there are any to avoid (?).

 

I think the main question is whether the power measurements being discussed are taken from the Ecodan approximations, or whether they are actual measurements taken by monitoring systems - in particular for power consumption. 

 

I do think that for some models, the power consumption approximations provided by the Ecodan system are currently incorrect for standby. For my 8.5kW system, they were overestimating it by a factor of 5-6, which I can now see, having had a power consumption meter installed. The 14kW system had a much lower approximated standby power consumption. I would recommend a metering system is installed, if you can afford it - I suspect it would pay for itself quite rapidly (although I don't have one for the larger system and probably won't have one installed on the basis of cost - go figure!).

 

Hopefully, Mitsubishi will use this information to alter the firmware, to improve the approximations for all models.

[PhilT]

18 hours ago, Ewan said:

I’ve read all the posts here but not really sure if there’s now any ecodans to avoid

None, and personally I doubt that any current brand/model should be avoided on the basis of standby power consumption. For me the most important takeaway is to be very careful with one's own attempts to measure standby consumption. Things may not be what they seem! Look up the manufacturer's data sheets, they won't lie.

[Marvin]

I think most ASHP have some level of standby/frost protection power consumption of about the scale commented on.

 

Also, of course, if your running, the heating / hot water it won't be on standby!

[SteamyTea]

1 hour ago, PhilT said:

Things may not be what they seem! Look up the manufacturer's data sheets, they won't lie.

The MI may show the power of the crankcase heater, but if you don't know the conditions it is activated, it is of little use knowing.

Also, if you have a small usage i.e. 3 kWh of DHW a day, delivered, burning 1 kWh a day to warm the crankcase is going to half your CoP.

1 hour ago, PhilT said:

For me the most important takeaway is to be very careful with one's own attempts to measure standby consumption

A proper meter like this is only a few quid and easy to wire in, even as a retrofit.

[ReedRichards]

What I gathered is that some but not all Ecodans have a standby power consumption so high that you would want to consider strategies for turning them off when you don't need heating.  Some other makes of heat pump may suffer the same issue but other heat pumps have a reasonable level of standby power consumption that you wouldn't worry about.

 

  

[BEJB]

On 01/12/2022 at 09:18, ReedRichards said:

What I gathered is that some but not all Ecodans have a standby power consumption so high that you would want to consider strategies for turning them off when you don't need heating.  Some other makes of heat pump may suffer the same issue but other heat pumps have a reasonable level of standby power consumption that you wouldn't worry about.

 

  

When first installed, my Ecodan had a daily standby reported by Melcloud of about 2.5kWh. Mitsubishi then installed a meter and it dropped to about 0.5kWh.

I think some Ecodans over estimate their power usage. Hopefully Mitsubishi will correct their firmware and this problem will go away.