Our heat pump is oversized by a factor of 2. It is capable of 10Kw at temperature of -5, the most we have used in one day in the last 2 and a half years is 144Kw (according to the machine) this equates to 6Kw per hours. It was -6 outside, windy and wet so the absolute worst conditions for a heat pump. In the last 2 and a half years we have only had 5 days when we needed more than 120kw. It is a Veissmann and seems to modulate down well, but it has no chance in spring and autumn when the load is less than 25% of its capacity.

For information we are in North Yorkshire, we have a barn conversion, single storey, well insulated, underfloor throughout, high thermal mass, 195 sqm.

The best way I have found to reduce cycling and have a good COP it to run with a long overnight set back of 8 hours,(21.00 to 0500) in this time the room temperature drops by 1.5 to 2 degrees. In the colder months Dec to March it then runs all day and once the room temperature returns to 21 by mid-morning ,I allow the room stat to regulate the call for heat, we only have one in the living areas as it is all open plan, the bedrooms are always off . In the spring and autumn I introduce a second set back from 11.00 until 16.00, effectively letting the slab cool so that the pump can run to catch up when it comes back on. Weather compensation is set to +10=26, +5=28.5, 0=31, -5=33.5.

If I could automate the adjustments I think a 2 hours on 2 hours off would give a bit better performance, but I am looking for something simple that I don’t have to keep adjusting.

For the people new to heat pumps the good news is, the heat pump OCD of continually checking settings and COP does eventually wear off.(a bit)

Posted by: @mikef

↑

If I could automate the adjustments I think a 2 hours on 2 hours off

Can't a room stat with timer achieve that? It would need enough on/off times per 24 hours, then just set the desired temp very high/very low.

HDHD (Heatpump Deficit Hyperactivity Disorder) is a well known condition that I have myself had to struggle with.   

Posted by: @cathoderay

↑

Posted by: @mikef

↑

If I could automate the adjustments I think a 2 hours on 2 hours off

Can't a room stat with timer achieve that? It would need enough on/off times per 24 hours, then just set the desired temp very high/very low.

HDHD (Heatpump Deficit Hyperactivity Disorder) is a well known condition that I have myself had to struggle with.   

Yes, as can more capable heat pump controllers

Posted by: @jamespa

↑

Yes, as can more capable heat pump controllers

Not to mention python scripts, but a room stat / timer solution is the sort of thing people can set up without complex knowledge of heat pump controller manuals or python scripts. 

Two hours on / two hours off controlled this way (room stat / timer) is self induced cycling, albeit low frequency. 

Posted by: @cathoderay

↑

Two hours on / two hours off controlled this way (room stat / timer) is self induced cycling, albeit low frequency

Indeed.  In this case I would personally leave it to the heat pump.  However if you can align the cycling to cheap periods that's a wholly different scenario.

@jamespa It depends on what "bad" cycling is. In spring it would run for a total of 4 hours in the 9 hour period that i had it set for, it would cycle 16 times, is this classed as 4 per hour (bad) or less than 2 per hour (good).

i am in the North so the smart meter fitted a year ago is actually a dumb meter with no signal. If a smart meter would work i would consider running the large pump a bit harder when electric was cheap and use the expensive time to set back.

I am not sure we fully understand either the supposedly harmful or indeed the potentially beneficial effects of cycling, eg when a heat pump uses it to lower output in low demand situations when it cannot use any other means to lower output, or we ourselves cause the heat pump to cycle, perhaps to take advantage of times when our tariff is lower. Nor is it clear when tolerable cycling becomes bad cycling. I have heard a suggestion that anything more than six cycles an hour is bad, but why six? Why not four or eight cycles per hour? What happens at six cycles that doesn't happen at five cycles per hour? Do we even know for certain that cycling is bad, and if so for what? Efficiency? Wear and tear? If so, where is the evidence? 

My heat pump almost always uses long cycling (one or twice an hour), apparently to control output. The only exception seems to be at an OAT of around 5 degrees, when it might managed a few hours of steady state running. I have not yet managed to get a sensible answer from Freedom Heat Pumps whether this happens by design ie is intentional or whether it is a fault (if indeed cycling is a fault, see above) in Midea heat pumps. What scant evidence there is (from others who have kindly posted their heat pump data here on the forum) suggests cycling may be the norm in Midea heat pumps, but no one has noticed before because there were few Midea installations, and far far fewer people who actually monitored their heat pump in sufficient detail to see cycling. For what it is worth, here is my heat pump's behaviour over the last 24 hours. Definite cycling, but does it matter?  

Posted by: @cathoderay

↑

I am not sure we fully understand either the supposedly harmful effects of cycling,

We don't and they may well be overstated.  Certainly there is a reduction in efficiency because the heat pump has to run at a higher ft, but this might be fairly small if the max overrun temp is low (I think there may be a trade off here between frequency of cycling and extent of FT excursion).  It used to be argued that it was bad for wear and tear but modern inverter driven pumps presumably soft start to mitigate this.

Posted by: @cathoderay

↑

Nor is it clear when tolerable cycling becomes bad cycling. I have heard a suggestion that anything more than six cycles an hour is bad, but why six?

I think it is probably because 10 mins is a nice round number for the min cycle time and seems to be the default on many pumps.  Obviously it's not a cliff edge!

Cycling is of course inevitable at higher oats because current heat pumps don't have the modulation depth, even if correctly sized, to deal with the full range of output requirements.  So heat pumps use cycling to match the output on average to the demand.  Boilers do the same.  Your plot shows cycling as you say, and the pattern is consistent with the demand being less than the minimum output.  Qualitatively my (Vaillant) heat pump does something similar when cycling, although so far Ive only witnessed it on the cusp between cycling and not cycling (simply because it hasnt been warm enough).  At this point I get about a one hour period, with a short off time and a very slow rise during the on time.  My expectation is that, at warmer OATs still, the gradient of the rise and the off time will increase.  Whether the frequency will reduce or increase I cant decide, I will observe when it happens!  

Whether it matters or not, if the demand is less than the minimum output there is no alternative, hence my question in relation to extreme cases (oversizing by a factor of saying 2 or more), where perhaps there is an opportunity to do something more advantageous than the heat pump default (which, in the normal course of events, I would stick to) using information to which the heat pump does not have access.

Does anybody have any figures re the efficiency of heat pumps when under different loads?

A car engine ticking over is more efficient than one running at full revs, so is a 10Kw pump at half capacity better than a 5kw flat out? Wear and tear wise the larger must be better, but do the extra size of all the components compromise the overall efficiency. To clarify by efficiency I really mean cost not just COP, certainly the Viessmann doesn't allow for the cost of running the pumps etc when it calculates its COP. 

Posted by: @mikef

↑

Does anybody have any figures re the efficiency of heat pumps when under different loads?

A car engine ticking over is more efficient than one running at full revs, so is a 10Kw pump at half capacity better than a 5kw flat out? Wear and tear wise the larger must be better, but do the extra size of all the components compromise the overall efficiency. To clarify by efficiency I really mean cost not just COP, certainly the Viessmann doesn't allow for the cost of running the pumps etc when it calculates its COP. 

Some manufacturers publish this data including (but not limited to) Midea, Mitsubishi, Vaillant.  Its not always easy to find, Mideas is in a password protected databook, Vaillant is only on their Czech site and Mitsubishi is the easiest to find, but it is out there.  I once took a cursory glance and decided I couldnt be bothered about it as the variations I noticed werent that great and anyway its not something you can readily control.   However I admit it was cursory so there may be something significant I missed.

Posted by: @mikef

↑

Does anybody have any figures re the efficiency of heat pumps when under different loads?

Midea and I suspect other manufacturers do have detailed tables of heat pump outputs and i suspect it should in theory at least be possible to compare a higher rated heat pump running at lower capacity to a lower one running flat out. But it is complicated by all sorts of things, for example, the Midea 12, 14 and 16kW heat pumps are the same mechanically, the rating is achieved simply by capping the maximum output. That begs the question: at the lower outputs, are the all the same? At some point i may have a look and see what I can discern, but I have to say looking at those data tables makes watching the paint dry positively thrilling!

I see @jamespa has just posted on much the same lines.