@lickofpaint  You will find that there are at least two schools of thought about the way that heat pumps should be run. There is the low and slow school which advocates pure weather compensation, minimal thermostat interaction and running the heat pump continuously over 24 hours. The gas boiler mentality is anathema. Then there is another school that has found, perhaps to its surprise that running the heat pump on weather compensation for long periods, but with setbacks when the heat pump is off, particularly overnight, is quite cost effective. This is especially so if the heat pump is oversized, a common finding due to the difficulties of assessing heat loss and the MCS insistence on guaranteeing warmth when the outside temperature drops to -3 or -5 C. This method requires more interaction from a thermostat, but works best if the latter isn't the sort that is constantly turning on and off with minor changes in room temperature.

We leave the bathroom window open after the evening bath for a few hours in winter when the heating is off and this deals with the condensation and doesn't stop the bathroom being warm by the time we get up next morning.

Posted by: @mike-h

↑

 

 

Posted by: @bontwoody

↑

Posted by: @bontwoody

↑

Cost wise there is no difference between running at a COP of 6 for 24 hours or a COP of 4 for 18 hours.

 

Posted by: @jamespa

↑

Sorry but that is simply not true and worse still is quite seriously misleading.  

Sorry, but it is true! If 24 kWh are required to supply the energy that the house loses for the whole 24 hrs, then 18 hours at a COP of 4 requires a consumption of 4 kWh and so does 24 hours with a COP of 6. That is without taking into account the possible benefits of a setback on reducing total 24 hour heat loss - a more controversial topic.

Er no, you are confusing power and energy.  You pay for energy not power.

 If 24kWh is required then the input energy is 24/cop.  So if the cop is 6 that means that you have to pay for 4kWh.  If the cop is 4 you have to pay for 6kWh. So unless your tou tarrif makes 6kWh cheaper than 4kWh then they are certainly not the same.

Posted by: @bontwoody

↑

@jamespa I think you have misinterpreted what I was saying. If the amount of heat energy delivered in both cases is the same, then as long as the electricity unit cost is constant then the cost for both cases is exactly the same. 

I am not making any claims about how the temperature of the house responds. 

In my own home I have experimented with running the heat pump at night on cheap electricity but found that the room temperature dropped back to default between the period when the cheap tariff ended and when I got up. (Yes, I do get up quite late 😁)

If the energy is delivered at cop4 is xkWh then you must supply x/4kWh. If the same amount of energy is delivered at cop 6 you must supply x/6kWh.

If you heat only for 18 hours you will have to deliver to the house a bit less energy than if you heat for 24 hrs.  But not by a factor of 4/6.

Posted by: @mike-h

↑

@lickofpaint  You will find that there are at least two schools of thought about the way that heat pumps should be run. There is the low and slow school which advocates pure weather compensation, minimal thermostat interaction and running the heat pump continuously over 24hours. The gas boiler mentality is anathema. Then there is another school that has found, perhaps to its surprise that running the heat pump on weather compensation for long periods, but with setbacks when the heat pump is off, particularly overnight, is quite cost effective.

 

This is especially so if the heat pump is oversized,

If oversized then it is indeed quite possible that an overnight setback will save energy, because you are reducing the inherent inefficiency of cycling.

@jamespa ok you got me checking my maths. 

if the HP was drawing 0.5 kW for 24 hours with a COP of 6 then it would deliver 72 kWh and use 12kWh

if the HP was drawing 1kW for 18 hours with a COP of 4 then it would also deliver 72 kWh and use 18 kWh

For it to work as I intended the COP in the second instance would need to be over 4.5 not 4   

Apologises, my bad  

Referring to the above Im going to try to clear up some basics, with apologies to any physicists or engineers reading!

Firstly energy is measured (for the purposes of heating) in kWh and its energy that you pay for.  Think of it like miles.  1kWh used to be called a 'unit' although that terminology has fallen out of fashion

Power is measured in kW, it is the rate at which energy is consumed/supplied.  Think of it like miles per hour.  You don't pay for kW (power), its kW*h (energy) that you pay for.

To keep your house warm, the heating system needs to supply to the house exactly the same amount of heat (energy) as the house loses to the outside.  

The amount of heat lost to the outside is proportional to the difference between the outside temperature and the inside temperature.

The house doesn't stop losing heat when you switch the heating off.  What happens is that the fabric cools, generally quite slowly, but heat is still lost (albeit at a lower rate because IAT-OAT is smaller).  This must be replaced.  

So to take a concrete example, if a house heated to 20 C loses heat at the rate of 4kW when the OAT is 10C, then over 24 hours it will lose 4*24=96kWh.  Thus you must supply 96kWh to the house.

If it is allowed to cool linearly to 15C over 6 hours, and then heats up back to 20C linearly over 3 hrs, then it will lose a bit less than 96kWh over a period of 24hrs.  The amount lost over 24hrs will be 96- (0.5 * 5/10 * 9/24) *96 = 96-9 = 87kWh.  

So instead of having to supply 96kWh we have to supply 87kWh.

This means that we break even if the COP for the 'setback' situation is 87/96 times the COP for the 24x7 situation.  So if the COP for the 24x7 situation is 6, we need a cop of at least 5.4 for the 6 hrs setback to break even.  Similarly if the 24x7 COP is 3, we need a COP of at least 2.7 in the setback situation for the 6 hrs setback to break even.  Neither of these are likely to be the case if its necessary to increase the flow temperature by more than a degree or so to recover sufficiently quickly for comfort.

This is the basic calculation from the thermodynamics.  However in the real world there are some other effects

• Power consumption that does not go towards heating the house, mostly the water pump(s) and the electronics.  This is likely to be ~100W, so 2.4kWh over 24 hrs.  This is likely to be insignificant as a proportion at low OATs, but can be quite significant as a proportion when the OAT is mild, ie either end of the season
• Heat (or efficiency) lost due to cycling, which may be ocurring in the 24*7 case but (because the power requirement is larger) may not be occurring in the setback case.  This is difficult to quantify but is more likely with an oversised heatpump or at either end of the season, or if the WC curve is set higher than optimum so that IAT is controlled by thermostat not WC.

The upshot of these real world effects is that setback may be productive in a greater range of cases than the basic thermodynamics would indicate.  However, in a well designed system these should only be significant at the ends of the heating season, ie when its pretty mild outside.

Unfortunately the precise numbers depend on the house and heat pump and indeed the diurnal variation of temperature.  As a very general guide however.

With a correctly sized heat pump:

• it is unlikely that, at the coldest part of the season, a night time setback of ~6hrs will save money (in the UK in the coldest part of the season the temp is cold day and night)
• in the mid season a 6 hr nighttime setback may possibly save money, principally because its generally warmer during the day than at night
• at the ends of the season then a 6 hr night time setback is also likely to save money (at the end of the season daytime is significantly warmer than night and there is a greater effect from the standing consumption due to pumps etc).
• in any of the above, if you have to raise the flow temperature (and thus reducing COP) in order for the recovery to be sufficiently fast, then you may well be better off reducing the setback time instead!

With an oversized heat pump

• its difficult to predict because unquantified engineering effects may dominate the basic thermodynamics.

Posted by: @bontwoody

↑

 

@jamespa ok you got me checking my maths. 

if the HP was drawing 0.5 kW for 24 hours with a COP of 6 then it would deliver 72 kWh and use 12kWh

if the HP was drawing 1kW for 18 hours with a COP of 4 then it would also draw 72 kWh and use 18 kWh

 

Agree (assuming that 'draw' in the second sentence should read 'deliver').

Posted by: @bontwoody

↑

For it to work as I intended the COP in the second instance would need to be over 4.5 not 4   

So in the second instance at COP=4.5, 81kWh would be delivered for 12kWh consumed.  To break even (ignoring ToU tarifs) you would need to consume 12kWh leading to 54kWh delivered at COP 4.5. 

To reduce the demand to 54kWh over the 24 hour period the house would need to cool instantly to OAT when you switch off (so it stops losing heat immediately).  If it did this then it would also (helpfully) recover instantly.  Most houses don't do that, or anything like that.

The fundamental here is that houses still continue to lose heat even if the heating is switched off, and that heat must be replaced when the heating is on.

@jamespa

I must thank you for clearing up a misconception I was labouring under, the COP differential is as you say and much closer than I intuitively thought.

Whilst I agree with the vast majority of what you have written, there are a few points that I think you may have oversimplified (probably for clarity of explanation).

In your example you talk about linear cooling of the house, which as you allude to earlier in your post isn't the case, its more like exponential with the greatest rate of cooling when the house is warmest. The house will also stop losing heat when its temperature equals the outside temperature, so you cant really just multiple the heat loss by the number of hours if the house reaches that point in the meantime.

This whole debate reminds me of the old nut about whether its better to run a gas boiler constantly or not, but in the case of heat pumps the differential COP magnifies the issue.

Clearly it is a nonsense to keep your heat pump running for 24/7 if you are going to be away from the house for a prolonged period, to make the point obvious lets say a year. The issue then becomes how much can you shorten the prolonged period before it starts to make sense to do it.

I think this depends on a number of factors, such as the differential between the temperature of your house and the outside, the level of insulation and the minimum flow temperature you can run at (related to the emitter size and the minimum output of your heat pump).

If the insulation level is poor, then the house will likely lose the vast majority of its heat pretty quickly, lets say within 3 hours for example. Does it then make sense to heat the house for another 5 hours (even at a better COP) or just start heating it from scratch again in time for when you wake up. Obviously this will depend on how quickly the house warms. In my house, last year, heating it continually seemed to use more electricity, but I have increased the insulation levels this year so I will see what happens when it gets colder.

I once corresponded with a gentleman who was lamenting his very high running costs compared to oil and the fact that his energy usage was the same in electricity and when he used oil. The only conclusion I could come to was that the amount of heat he was losing at night was was wiping out any gains he was making in the day (as he only used his oil boiler in the day). He did admit his house was very poorly insulated.

We are probably in danger of hijacking this thread, but hopefully the OP might find this stuff interesting (I do) 🙂

@bontwoody

In truth it is a very widely held misconception, whether we are talking about boilers or heat pumps, that turning off the heating for (say) 25% of the time saves 25% of the energy.  In reality the saving is far, far less than this because, as I say, the house continues to lose energy even though the heating is off, and that energy must be replaced when its switched back on again.

Posted by: @bontwoody

↑

there are a few points that I think you may have oversimplified (probably for clarity of explanation).

In your example you talk about linear cooling of the house, which as you allude to earlier in your post isn't the case, its more like exponential with the greatest rate of cooling when the house is warmest. The house will also stop losing heat when its temperature equals the outside temperature, so you cant really just multiple the heat loss by the number of hours if the house reaches that point in the meantime.

I have simplified for clarity, but the principles still apply.

You are correct that the house cools exponentially not linearly.  It also recovers exponentially not linearly and the two curves are inverted relative to each other.  A linear approximation for the cooling/recovery cycle is thus better than it is individually for cooling or recovery, particularly in the most common case where the temperature loss is a fraction of the OAT-IAT difference (eg just a few degrees).

As you say once the house reaches OAT then you can do the simple calculation.  That is rarely the case for an overnight setback but of course if you go away for a few days will be the case.

Posted by: @bontwoody

↑

Clearly it is a nonsense to keep your heat pump running for 24/7 if you are going to be away from the house for a prolonged period, to make the point obvious lets say a year. The issue then becomes how much can you shorten the prolonged period before it starts to make sense to do it.

I think this depends on a number of factors....

Agree which is why its near-impossible to give definitive rules. 

However I think there are two basic cases: (a) regular (generally night time) setback, recurring every 24 hours, and (b) longer turn off during an extended absence. 

I think we can say with some certainty that turning off during an extended absence (>48hrs) makes sense.  To maximise the saving the heating should ideally be turned on again before you return so that it can recover slowly at a good COP and is not forced to recover quickly at a poor COP.  Preferably perform the recovery during the daytime when its warmer outside.

Giving guidance on the regular night-time setback is more problematic and does depend on all of the factors you list. 

My strong suspicion (actually I'm nigh-on certain) is that many who claim large savings (eg >10%) are either measuring it incorrectly, or have a heat pump which is either oversized or poorly set up.  They would be well advised to first check that they have optimised their weather compensation curve so that the heat pump is running at its most efficient when it is on, before experimenting with setback.  That said it may not be possible to optimise their set up (eg because the heat pump is oversized) and it may well be that the setback does help more than the pure thermodynamics would indicate.

For a correctly-sized heat pump with optimally adjusted weather compensation thermostats, if any, set above the desired room temperature, a good test would be to run it 24x7 for a few days, then run it with setback for a few days.  If you can do the latter without having to raise the WC curve (or manually adjust the flow temperature) to get fast enough recovery, then setback is saving you money.  If you find you have to raise the WC curve to get recovery fast enough then its probably better to cut back the setback time until this is not the case.  For this test to work you might need to turn off any 'auto adaption' mode that some heat pumps have, otherwise it will adjust the WC curve without you knowing!

Unfortunately the maths to model this is challenging (at least for me) and the measurement of the result is equally challenging, because of the fact that every day is different so you cant generally run the control experiment. 

Finally, for those who dont want to know the detail, the basic point here is that its a widely held misconception, whether we are talking about boilers or heat pumps, that turning off the heating for (say) 25% of the time saves 25% of the energy.  In reality the saving is far, far less than this because the house continues to lose energy even though the heating is off, and that energy must be replaced when its switched back on again.  With a heat pump the modest saving may be eliminated entirely if it has to work harder to recover, so care must be taken before assuming that (eg) a night time setback will save money.

I know we talk about setbacks being used in an attempt to save money, @jamespa, and you're certainly not the first to be raising the practice for discussion and debunking. However, there is an obvious alternative reason for not heating low and constantly - because many find their ideal comfortable temperature varies during the day, particularly wanting a cooler house whilst asleep - and I've yet to see any evidence that money rather than comfort is the primary driver for setbacks being used.

I do get concerned that the constant drip-feed message on this forum is "setbacks are bad because they are inefficient" if (and here I simply don't know) most people's priority is comfort and they're being put off achieving that. Do you have any evidence or insights that might help identify the split between savings-seekers and comfort-seekers?

Posted by: @majordennisbloodnok

↑

people's priority is comfort

Get that. But everyone of our relatives has a 13 tog duvet on the bed, we used to use low tog for summer and high for winter. We now use a low tog all year round. In fact trying to buy a low tog duvet in shops isn't that easy and you don't get much choice, but an abundance of high togs are freely available.

It's back you need to flicking a mental  switch, when changing to low temp heating, it's not the same as 70 Deg heating.

Posted by: @majordennisbloodnok

↑

I know we talk about setbacks being used in an attempt to save money, @jamespa, and you're certainly not the first to be raising the practice for discussion and debunking. However, there is an obvious alternative reason for not heating low and constantly - because many find their ideal comfortable temperature varies during the day, particularly wanting a cooler house whilst asleep - and I've yet to see any evidence that money rather than comfort is the primary driver for setbacks being used..... Do you have any evidence or insights that might help identify the split between savings-seekers and comfort-seekers?

I agree with the first point entirely.

As to 'do I have any evidence that money rather than comfort is the primary driver' then I would say yes some, for some people.  The evidence (in as much as it exists) is that some people come onto this forum (and others) and make extravagant claims about how much money setback is saving, sometimes going so far as to say that they don't mind feeling a bit cold in the morning in order to make the saving.  This suggests that their primary motive is cost saving.  The extravagant claims generally exceed the saving which the thermodynamics tells us is possible with a well adjusted system (indicating that either their system is not well adjusted in the first place, or the measurement underpinning the claims is dodgy).

Although its almost impossible to give general guidance I think this is one of those lucky (and rare) cases where comfort and cost probably more or less coincide for many practical cases.  If we divide the heating season into very cold (mid season), moderate (most of the season in Southern England, less so in east coast Scotland)), and end of season (both start and finish) then reasonable qualitative arguments can be made for the following giving both a good level of comfort and a modest saving (<10%):

• Mid season (very cold)  - operate 24x7.  
• Moderate (most of the season in Southern England, probably less so in east coast Scotland!)  Set back/turn off for a few hours overnight - perhaps 5-6, allowing 3 hours for recovery).  Don't overdo the setback!
• Ends of season - turn off overnight and during daytime, late afternoon/evening heating only (maybe with a midday boost).  

The qualitative arguments, which apply to a reasonably well adjusted correctly sized system run on WC, consider comfort, solar gain, variation in COP due to outdoor temperature variations, the change in the pattern of diurnal temperature variation at different times of the season, the thermodynamics and the fixed running energy due to pumps etc.  For brevity I will avoid going through all the arguments here, but can do so if people want.  As I say in an earlier post a good test is that, if you need to increase the flow temperature for the house to recover sufficiently quickly, you are probably better off reducing the setback time instead.

I stress this is very qualitative, very rough, and may well not apply to any individual case particularly to very well insulated/passive houses which are almost certainly better sticking with slow/low.

Finally just to emphasise, the only debunking I wish to do is to debunk the idea that turning off heating (whether heat pump or boiler) for say 25% of the time saves 25% of the cost.  It doesn't, the saving is much less because the house continues to lose heat even though the heating is off.  With a heat pump any modest saving may be wiped out entirely because it has to work harder to recover, so care is needed.