Notifications

Clear all

Heat pumps and on/off working - starting to quantify it

JamesPa · 2023-09-13T17:14:43Z

There has been much discussion on on/off working/nighttime setback in relation to heat pumps, but not much that is quantitative. @fazel and @derek-m have made some helpful comments in another thread and I have now started to think about it more quantitatively. I present an example of how one might assess the effect, it needs development and discussion but is a 'starter for 10' Worked example: Lets consider a 10kW heat pump fitted to a house which consumes 8kW when the OAT is 0C and the target IDT is 20C. Lets also assume that the emitters are radiators with a FT of 45C at 0C, and furthermore that we wish save energy by setting back/switching off for a maximum period of 9 hours (overnight) - eg from 10pm-7am. Lets consider what happens when the OAT is 10C. Lets further assume that the temperature drops by 5C during an 'off' period of 4.5 hours, when we switch it back on to get the house warmed up in time for 7am. Once we have worked through this example it will be relatively easy to change the assumptions. At OAT of 10C and IDT of 20C the load will be 4kW and the flow temperature required to deliver this will be 34.7C (applying the rule that emission from radiators is proportional to (deltaT rad-room)^1.3. At this FT and OAT we can typically expect to achieve a COP of 4 (as will become evident later it doesn't matter what the exact COP is for the calculation, but it does matter roughly what it is). During the on/off cycle the average IDT will be approximately 17.5C, so instead of losing 4kW the house will lose (on average) 3kW. Thus over the 9 hour cycle the house will lose 9x3=27kWh instead of 9*4=36kWh had we left the heat pump on. More importantly during the 4.5hour cool down period we will lose approximately half of this amount, 13.5kWh , and this is lost from the fabric so will have to be replaced to return to the state we started with. We will return to this shortly. However we don't pay for energy lost from the house, we pay for energy consumed by the heat pump. This is where it gets interesting as the amount of energy consumed by the heat pump depends on the control strategy during the recovery period. Control strategy 1 - forced recovery In this strategy we attempt, during the recovery period, to get the house back to the state it was when we first switched off the heat pump. This means bringing the house back up to 20C and replacing the 12.5kW we have lost from the fabric. To do this we will have to supply to the house, in addition to the heat loss during the recovery period (average 3kW), an additional 13.5kWh to replace the heat lost from the fabric. thus we need to supply on average 3+13.5/4.5 = 6kW to the house. The heat pump is perfectly capable of doing this, but the only way the radiators can emit that amount of energy is by raising the flow temperature during the recovery period to about 40C (40.04C to be precise). A heat pump that has a fixed WC curve wont do this but one that has room compensation might. This causes the efficiency of the heat pump to drop by about 10% (2% per degree is a good rule of thumb). So a heat pump which was operating at a COP of 4 will instead operate, during the recovery period, at a COP of 3.6. If we work through all of these numbers we find that the heat pump would have consumed 9kWh to deliver the 36kWh that would have been lost had we left the heating on, and 7.5kWh to deliver the 27kWh that is lost if we switch the heat pump off, a saving of only 16% in energy consumed as opposed to the saving of 25% in energy delivered/lost. With these figures its still a saving so still worthwhile, but with other figures, particularly as the OAT drops, the saving becomes much smaller. Some exploration of the 'space' with an excel model is indicated (my next step). Control Strategy 2 - natural recovery In this strategy we don't alter the flow temperature during the recovery, it remains at 34.7C. Heat pumps operating pure WC will do this. However the rads will still deliver to the house more energy than that needed to maintain the steady state and this the temperature will still recover, albeit more slowly because a) the rad-room delta T is higher (but gets smaller as the house starts to heat up) and b) the house to OAT delta T is lower (but gets larger as the house starts to heat up) I dont have a formula for this (I think its logarithmic/exponential) but its easy to work out in a spreadsheet, which I have done. Putting all of the above assumptions in the house will reach about 18.9C after 4.5 hours, and will get to 19.8 after 10 hrs. The saving in energy is 27/4 vs 36/4, 25% but at the expense of a bit of a loss in comfort, because the house is roughly 1C under temp at the end of the recovery period. This may be tolerable or alternatively after experiencing it for some days the owner might nudge the WC curve up to compensate, increasing FT by about 1-1.5C, causing a 2-3% penalty on efficiency for the entirety of the day. However thats a ...

Page 8 / 15 Prev Next

Air Source Heat Pumps (ASHPs)

Last Post by Derek M 2 years ago

170 Posts

12 Users

31 Reactions

27.1 K Views

RSS

Derek M

(@derek-m)

Illustrious Member Member

Joined: 5 years ago

Posts: 4427

14/10/2023 8:10 pm

Posted by: @jamespa

↑

Posted by: @derek-m

↑

I have now completed some further improvements to the modelling tool.

It is now possible to adjust the desired IAT setpoint and also the OAT value to one decimal place. This therefore provides the capability to ramp up and down the setpoint and more closely match variations in OAT.

Have fun in the 'Playroom'.

I downloaded your modelling tool but there are lots of '#value' in the playroom. Is there something I need to dial in to start it working?

Hi James,

Which cells contain '=value'?

ReplyQuote

JamesPa

(@jamespa)

Illustrious Member Moderator

Joined: 3 years ago

Posts: 3098

Topic starter 15/10/2023 10:13 am

About 50% of the play sheet. I'm guessing they all rely on or derive from one 'problem' cell.

I have an old version of Excel (2003) it could be that, although unlikely because it usually issues a warning if there are unsupported features. I will try to narrow down the root cause.

4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.

ReplyQuote

newhouse87

(@newhouse87)

Reputable Member Member

Joined: 3 years ago

Posts: 231

15/10/2023 11:04 am

Had cop of 3.7 for yesterday/this morning, that seems lot better then what i had last winter, 3.04 last winter but that included dhw. Freezing at 5am this morning so running from 530-10-30 had used 27kwh, thats heavy, house now warm, prob too warm so switched off heating and will adjust curve down to 34 at- 5 and 26@20 try that tomorrow. That's a 12 degree difference to installer setting on cold side.

ReplyQuote

Derek M

(@derek-m)

Illustrious Member Member

Joined: 5 years ago

Posts: 4427

15/10/2023 4:55 pm

Posted by: @jamespa

↑

About 50% of the play sheet. I'm guessing they all rely on or derive from one 'problem' cell.

I have an old version of Excel (2003) it could be that, although unlikely because it usually issues a warning if there are unsupported features. I will try to narrow down the root cause.

Did you resolve the problem? If not, then please provide one cell location and I should be able to trace back to the cause.

I may be able to save the spreadsheet in a different format if that would help.

ReplyQuote

JamesPa

(@jamespa)

Illustrious Member Moderator

Joined: 3 years ago

Posts: 3098

Topic starter 16/10/2023 8:35 am

Unfortunately I didn't resolve it. #value cells include E1-AA1, D14, 16-18,20,23,24,31,32,36... but

opening it in Libre office calc seems to have fixed the problem. I will report back if I work out what the problem was!

4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.

ReplyQuote

Derek M

(@derek-m)

Illustrious Member Member

Joined: 5 years ago

Posts: 4427

16/10/2023 9:23 am

Posted by: @jamespa

↑

Unfortunately I didn't resolve it. #value cells include E1-AA1, D14, 16-18,20,23,24,31,32,36... but

opening it in Libre office calc seems to have fixed the problem. I will report back if I work out what the problem was!

Hi James,

As you will have noticed I carried out quite a number of modifications to create the present version, one of which was to simulate the heat pump stopping and starting. This uses the cells D1 to AA1 to set cells D14 through to AA18 to zero if the heat pump is deemed to have stopped.

The formula in cell D1 sets the cell value bases upon the values of cells D9 (Setpoint Temp.) and D11 (Actual IAT), such that if D11 > D9 + 0.1C the heat pump is switched off, cell D1 set to 0.

Try overwriting the formula in cell D1 with the actual value 1 or 0 and see what effect this may have. You could also try changing the value in cell E9, above and below that of cell E11, to see how that affects the formula in cell E1.

ReplyQuote

newhouse87

(@newhouse87)

Reputable Member Member

Joined: 3 years ago

Posts: 231

16/10/2023 6:30 pm

Hi lads,any idea why my flow temp would be this high when my curve would suggest it should be lower. I have attached pics.

ReplyQuote

JamesPa

(@jamespa)

Illustrious Member Moderator

Joined: 3 years ago

Posts: 3098

Topic starter 16/10/2023 7:59 pm

Has it just come out of a dhw cycle, which would be at a much higher ft (and thus may take a while to cool down)?

4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.

ReplyQuote

Kev M

(@kev-m)

Famed Member Member

Joined: 5 years ago

Posts: 1276

16/10/2023 8:35 pm

There might a minimum LWT setting that applies whatever the WC curve says. Maybe that is 30?

This post was modified 2 years ago by Kev M

ReplyQuote

newhouse87

(@newhouse87)

Reputable Member Member

Joined: 3 years ago

Posts: 231

16/10/2023 8:43 pm

@jamespa No DHW cycle is 2-3pm, this was much later. Judging on my curve, @7.2deg, what should flow temp be? Maybe its ok. Using lot less KWH last few days which is good as i have been lowering flow temp since and slab is probably warmed up.

ReplyQuote

newhouse87

(@newhouse87)

Reputable Member Member

Joined: 3 years ago

Posts: 231

16/10/2023 8:53 pm

Posted by: @kev-m

↑

There might a minimum LWT setting that applies whatever the WC curve says. Maybe that is 30?

Have desired temperature on front screen but think that changes with the curve.

ReplyQuote

JamesPa

(@jamespa)

Illustrious Member Moderator

Joined: 3 years ago

Posts: 3098

Topic starter 16/10/2023 9:04 pm

@newhouse87 well your WC line is 34 at -5 and 25 at 20, so at 7 you expect 25+9*(11/25) = 29. So 30 is just 1C out which is likely well within the hysteresis. Situation normal so far as I can see

4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.

👍

ReplyQuote