Defrost hell

Hardware Heatpump
1

Hi all,

Thank for letting me join.
I should confess that I don’t have any OEM equipment with my installation, please don’t hurt me. However, I could really do with some help/advice with my ASHP.

My setup is:
7kW Vaillant Arotherm+
UFH downstairs (still has pumpset, blending valve set to 40deg as limiter)
Rads upstairs
approx 250m2
Volumiser on return
no zoning, all TRVs fully open
WC 0.3
Design temps 35/-3
Newish property
6.8kW calculated heatloss
Target room temp 20.5deg
Max/min temps 20/40

The system was retrofitted in April’24 and after a little tweaking, it’s been running really well and efficiently.
It was off (for heating) over summer and early autumn, and since turning the heating back on in October, I’ve been really happy with how it keeps the house a constant temperature, regardless of what’s going on outside.
I was running in “active” early autumn, to help manage solar gain, but have since switched to “inactive” and the WC seems pretty much nailed on.
Most of the time, the HP trundles along at around 5-600W, taking the occasional break when it can’t lose the heat into the building. CoP has been between 5-6 the whole time (according to the Vaillant app, if you believe that).
All good so far.

Now the bad bit.
Towards the end of November, we had 2 days then 1 day where the outside temperature hovered around 0deg, and this is where I’ve found the fatal flaw in my setup. Up until this point, I foolishly thought I understood my setup quite well.
At around about 3deg outside, the HP starts doing defrosts with varying frequency. However, as the temp drops to 2 or 1 deg, the power consumption goes up exponentially and it gets into a cycle of defrosting every 30mins any using max power in-between for little or no benefit, very much like this (apologies to the person whose data I’ve borrowed)…

image
image1411×835 181 KB

The only means of escape appears to be to hope the external temp goes up, or possibly down, and hope the defrost frequency declines.

I’ve read a few relevant posts on here, namely

and

…but am non the wiser as to what I could/should do.

I’m hoping someone may be able to offer me some advice/tips/tricks that may alleviate this situation, otherwise I’m staring down the barrel of quite a disappointing (and expensive) heating season.

TIA

2

Hello @Barrie and welcome

First and most important is the heat pump keeping your house at the desired temperature / comfort level? If temperatures are falling to 17-18C and you have it on all of the time then that might suggest an undersized unit.

While the poor performance the coldest days can be a bit worrying, the average temperature at which most heat is delivered in winter is usually around 6-7C and so in terms of overall annual bill impacts a couple of weeks of poor performance may not have a large impact in the grand scheme of things.

You dont really want any blending. UFH is not my expertise but the standard UFH manifolds with a pump and blending is not usually a good idea with a heat pump, you can usually simplify that out and remove / disable the pump, close off any potential for blending… that’s probably not your issue here though given high COPs for most of the season so far…

2 Likes
3

We’ve been having an interesting discussion about the maximum output capabilities of vaillant heat pumps here Vaillant maximum output capacity testing. It looks like your heat loss and vaillant capacity should be just about right even with lower than datasheet quoted outputs.

2 Likes
4

Hi Barrie

I can’t say I ‘ve solved my problems as my 10.6kw Daikin was struggling to put out 7kw on the cold day last week. A couple of months ago after talking to Daikin on their stand at Everything Electric they sent around one of their best engineers to check my unit over. He was here for most of the morning and, in those warmer conditions was able to say that it was performing as it should be. Sadly having nearly 15kw on tap when it is 14c is not a lot of use to me and it is very hard to time to get an engineer like him in when it is freezing. His best guess was moisture in the refrigerant causing the low temp problem but I cannot understand how that would be the issue as the evaporator is icing up fully before the defrost and the refrigerant is going to -15c during the problem times. (normally the refrigerant is just 4c below ambient but as the evaporator starts to block the unit drops the refrigerant temp to maintain it’s output leading to the death-by-defrost cycle). This explains why problems seem to start at 3c and below.

Just like you, when it’s warm my unit has a very good COP and it’s minimum continuous output is around 5.5 kw before cycling. So, in effect these heat pumps have a continuous operating range between 5.5kw at 10c and above and 7kw at 0c. I challenge anyone to find that, most useful, data in the tech data sheet!

I’ll send this latest data to the gent at Daikin to see if they are able to do more

Colin

PS you can see the issue clearly here;

image001.png1114×679 255 KB

5

Hi @Barrie

Based on the real word testing we’ve done, a 7kW Vaillant should be about to output about 7kW even under the worst conditions. So it should be able to cover your 6.8kW calculated heat loss, your actual heat loss is probably lower.

If your heat pump is struggling to maintain the indoor temperature then you probably need to increase your WC setting, 0.3 is very low. As a test when it’s around 0C outside try setting to WC to max i.e 4.0 and the room set point to 30C, this will make the HP work as hard as it can, if this still fails to heat your house then the heat pump is undersized.

If after increasing the WC curve, the HP is able to provide sufficient heat, then the defrosts are nothing to worry about. Overall, the amount of heat lost via a defrost cycle is relatively low.

Would you mind removing the screenshot from your post? It’s quite very confusing for other forum users to see a screenshot of data which has nothing to do with your system! With a WC of 0.3 your flow temperature will only be reaching a maximum of 30C at 0C outside, the screenshot above shows a flow tempo of 45C at 10C outside.

image
image766×841 32.6 KB

1 Like
6

Hi @TrystanLea

The temp does drop away a little, maybe half a degree or so over 12hrs, but the defrosts of doom have been relatively short lived thus far as the OT has risen and they’ve ceased, after which the inside temp has eventually recovered.

I think when the heatpump was installed, there was some uncertainty regarding this and whether the onboard pump had sufficient grunt, as plans for the existing UFH design/plan/guess weren’t available.
I’ve since spoken to my installer and they’ve managed to get hold of the layout drawings from Wunda, so the plan is to remove bits when a service is due.

I think my main gripe, is probably with Vaillant for not providing useful information regarding what actually happens around defrosts. Namely, the power requirement suddenly leaps when the unit starts defrosting more than about once an hour. I had no reason to suspect the power increase wouldn’t be anything other than linear, with maybe a slight increase to account for some defrosts, until it maxed out at about -10. Whereas, what I’m actually seeing is linear to about 3deg, then a sudden step change almost doubling as frequent defrosts kick in.

Having look through most, OK a lot, of the performance chart on the monitoring site, there seems to be a corelation between what I perceive to be problem (half hour) defrosts, heatpumps closely matched to the design heatloss, and floorplan area (I’m making the leap from area to system volume). Notable examples are from Urban Plumbers, Custom Renewables and Rickman Heat (I’ve been kind of benchmarking against one of his, as it’s comparable by virtually every metric).
I thought I’d researched heatpumps to death before making the leap, but nobody warns you about this problem area.

P.S. Sorry! That sounded a bit ranty, oops! :face_with_open_eyes_and_hand_over_mouth:

7

Hi @glyn.hudson

As I said in reply to Trystan, we’re not struggling to maintain the temp, rather the vagueness of real world (UK) performance data available from Vaillant.

The huge increase in power consumption, whilst the unit is running in this defrost zone is staggering. In November, 3 days at around 0deg accounted for 1/4 of the entire months consumption, which just seems disproportionate.

I’ve changed to a more appropriate screenshot, which uses the same parameters, although it looks like they’re set to 21deg and active.

8

Hi @ColinS

I thought having previously read your topic, and seen a glimmer of hope at the end that you may have had some sort of epiphany you were going to share. Alas, not. :cry:

I was in touch with my installer last week, explain my concerns, and they’d been on to Vaillant, whom I’m waiting to hear from to come and check the compressor/refrigerant.
I’m not optimistic that will make any difference.

Maybe I’ll try shutting a load of radiators to see if less volume has any effect. :thinking:

9

For the avoidance of doubt, it’s the blending that is the bigger issue here. Blending makes sense with a boiler sending 70+C water that you don’t want in the UFH loops but any temperature coming from a heat pump is fine to send direct to UFH.

If the UFH needs a pump to help out the primary circulation pump that’s (potentially) OK - depending on how the UFH is piped with respect to the Radiator circuit (which presumably doesn’t have its own pump). It’s best if you can run everything with just the primary circulator though, so no pump and (especially) no blending valve on the UFH.

2 Likes
10

I presume that you mean energy not power.

I’m not sure that your results are hugely disproportionate.

Looking at my system (Samsung 8kW, and only just capable of meeting the heating demand in cold conditions), on the 3 days 19th-21st of November 122kWh was consumed, with a mean outside temperature of -0.4C. 571kWh was consumed for the 30 days of November, mean outside temperature 7C. 21% used on those cold days which is very similar to your result.

I’m coming to the conclusion that under sizing
a system is likely to result in excessive defrosts. The system needs to work for longer to maintain target temperatures and doesn’t have enough area to defrost quickly.

Defrosts Nov
Defrosts Nov1108×938 360 KB

1 Like
11

Heat pumps are nighly nonlinear in their electricity use with outdoor temperature. Firstly, your heating need rises linearly as outdoor temperatures decrease. Second, COP for a given flow temperature decreases as outdoor temperature decreases. Of course you need higher flow temperatures when it’s colder - bam, another efficiency hit. Then come the defrosts, which themselves eat some electricity and take heat that was already put in the building out of it (which needs to be put in again), and it’s not like the heat pump runs at full efficiency between defrosts, since the frost builds up gradually and hence efficiency is only at rated performance right after a defrost. Overall it’s entirely realistic that the electricity use is 3-4x as high at 0°C with heavy defrosting compared to +10°C.

6 Likes
12

If I can ever find the time I intend to go through all the heatpumps on heatpumpmonitor to see what their maximum output really is. Most will be oversized but you’d expect to find a few working at their limit like yours and mine.
My suspicion is that few if any will match their nameplate output on a damp freezing British day.

Also the size of the evaporator must be the limiting factor with defrosts…so plotting evaporator size vs nameplate output should identify the units best able to cope.

You might find if you close rads that the reduction in system volume has a negative effect on defrosts. A daikin engineer told me that their rule of thumb is that the flow temp should not drop below 25c during defrost.

Good luck

13

Have I got an early Christmas present for you. Check out the “Heat demand tool” in the top menu :wink:

image
image576×470 19.8 KB

14

But the blending valve is set to 45deg, which is already higher than any temperature I would expect to see through the heating circuits. That being the case, surely the return from the UFH is going straight back to the heatpump, and not being blended with the flow at all. Or have I misunderstood how it works?

Regarding the UFH pump, I’ve done some quick calcs, and have worked out the pressure drop through the longest loop is approx 25KPa. When I checked, the pump in the 7KW Vaillant had 45KPa available, so this should be more than enough, right?

1 Like
15

Well, yes and no. The energy consumption is comparatively large because of the step change in power required once the rapid fire defrosts start.

There doesn’t appear to be any smooth transition between no defrosts and all defrosts and the power requirement suddenly changes from dawdling along at 1KW at 3deg OT to 2.5KW at 2deg OT.

16

I appreciate why the energy consumption is so high for so few days. However, it’s adversely affected by the half hourly defrost whereas, if the defrosts were only at 45-60 min intervals, which had been suggested to me as being “normal”, I believe the resulting increase in consumption on colder days would be far less significant.

17

I’d be interested to see any correlations you find.

Regarding system volume, I did a rough calculation a few days ago and got to something around 195lts, which should be more than enough. Anyway, I don’t think I’ve got anything to lose by trying, it’s not like it can be much worse. :crossed_fingers:

18

If it’s set that high then - exactly as you say - the valve shouldn’t actually be blending at all, and should be sending all the Flow to the UFH (and none of the Return). There’s always a risk of if sticking though so long-term you’d be better off without it.

That’s probably OK then, but it will also depend on how your UFH is piped in relation to the radiator circuit. In order for the primary circulation pump to put enough flow through the UFH, you’ll want the radiator circuit to provide a similar ‘resistance’ so the flow gets shared between the radiators and the UFH. Is there a separate thermostat controlling the UFH pump? Does the UFH still see good flow if that thermostat is satisfied so the UFH pump isn’t running?

19

Hi Barrie,

If it’s the Rickman Stroud install you’re comparing to then that’s me. You have similar system volume, flow temp, thermal mass and heat loss to me. The worst defrosts I have seen are also half hourly, and yes it’s directly correlated to 0-5 degrees range when it’s humid (ignore my data from last few days though - my wireless temp sensor has failed so I’ve had to put the unit to fixed temp until Vaillant replace it under warranty, and in addition it is quite low humidity currently). I would say though that my unit is just the right size for my house, another few hundred watts of loss and I should be on a 7kW. OK we haven’t had a consistent period of -3 for days on end to properly test that theory, but at least at those temps the defrosts would not occur as the humidity drops out all the moisture.
As an engineer who loves his efficiency I don’t like the defrosts but they are unavoidable. Yes it costs money but the house doesn’t feel the effects and their presence as a % of total running time over the seasonal is minimal. Hope this helps alleviate your concerns.

Tom

7 Likes
20

My only reasons for removing the pump but leaving the blending valve were, firstly I’ve seen you can get a double flanged pipe section which is a straight swap out for the pump (so no real replumbing necessary), and secondly the blending valve should provide protection should the 3-port valve fail and the HP ran a DHW cycle.

The heating pipework splits soon after the 3-port valve. I did have to put quite a bit of effort in the balancing the flow between rads and UFH as we were getting too much heat upstairs.
The UFH pump is not controlled by the HP, it is off a roomstat with actuators on 4 out of 11 loops (other actuators have been removed), although its set high at 23deg (there for solar gain in Spring and early Autumn) so the pump is always on.
If the pump off, either due to no demand or I’ve cut the power, the flowrates in the sight glasses drop by about 25%. I don’t know whether this indicates the pump is still providing some resistance or the HP pump is not man enough. In all instances, the flowrate displayed in the HP interface remains at 1200ltrs/hr.