Heat Pump Sizing?

NedS

michaels said:

So I need a heat pump that can actually output 10kw at -3.

If the heat loss calculation for the property (or better, the actual measured heat loss) is 10kW at -3C, then yes on the coldest days (when it is -3C outside), you will need a heat pump that can actually deliver at least 10kW of heat to keep the house warm (at design temp).

I say at least 10kW, as that would require the heat pump to be running constantly to match the heat loss. But you should also factor in time out of heating for the DHW cycle(s), and also any defrost cycles, so most system designers would factor in an extra percentage to allow for this, and I would expect them to specify an 11or 12kW model as a result. The rated value (e.g, 10kW) is normally the output for a flow temp of 35C at an ambient 7C. Any given heat pump may put out more or less than it's rated nominal output at your design temp of -3C, so also check the manufactures output tables to see that it is actually capable of meeting your needs at those temps.

On the other side of that coin, it is unlikely to be -3C all day. It may be -3C over night, but may be warmer during the daytime, so less heat output will be required than the full 10kW that would be required at -3C. This will offset slightly against DHW and defrost cycles.

As you can see, it all starts with an accurate heat loss figure. Get that wrong and you either have a house you cannot adequately heat on the coldest days in winter, or a heat pump that is over-sized for the property and will be less efficient in milder autumn and spring months.

matt_drummer

There are some people, me included, who think that a slightly undersized heat pump is desirable.

We don't get many days at design temperature where I live and as Ned S says, it is rarely that cold all day.

It's also generally colder at night when maybe not so much heat is required.

Some think that is easier to supplement the heat a little when its really colder than suffer a poor performing heat pump for most of the time, I agree having suffered a grossly oversized/unsuitable heat pump

Installers tend to oversize just to be safe making the problem even worse.

michaels

NedS said:

michaels said:

So I need a heat pump that can actually output 10kw at -3.

If the heat loss calculation for the property (or better, the actual measured heat loss) is 10kW at -3C, then yes on the coldest days (when it is -3C outside), you will need a heat pump that can actually deliver at least 10kW of heat to keep the house warm (at design temp).

I say at least 10kW, as that would require the heat pump to be running constantly to match the heat loss. But you should also factor in time out of heating for the DHW cycle(s), and also any defrost cycles, so most system designers would factor in an extra percentage to allow for this, and I would expect them to specify an 11or 12kW model as a result. The rated value (e.g, 10kW) is normally the output for a flow temp of 35C at an ambient 7C. Any given heat pump may put out more or less than it's rated nominal output at your design temp of -3C, so also check the manufactures output tables to see that it is actually capable of meeting your needs at those temps.

On the other side of that coin, it is unlikely to be -3C all day. It may be -3C over night, but may be warmer during the daytime, so less heat output will be required than the full 10kW that would be required at -3C. This will offset slightly against DHW and defrost cycles.

As you can see, it all starts with an accurate heat loss figure. Get that wrong and you either have a house you cannot adequately heat on the coldest days in winter, or a heat pump that is over-sized for the property and will be less efficient in milder autumn and spring months.

And indeed I do have a heat loss calc from Octopus of 9,5kw that while not perfect ties in with my max gas use.

Overnight with the temp set back the heat loss and therefore required input is potentially lower giving some scope for water heating then too.

But as you say the actual 'continuous' output of heat pumps is impacted by defrost cycles so is harder to fathom and then there is still the fact that most manufacturers have their 'small' heat pump and 'large' heat pump that come in different max outputs but have the same weight, dimensions and refrigerant volume and likely the same minimum modulation output so if you are at the bottom of the 'large' range then you get the lowest amount of modulation

Reed_Richards

Defrost cycles only happen over quite a limited range of temperatures.  If it's really cold out then there isn't enough moisture in the air to give rise to significant "frosting". And if it's a bit warmer out than 0 C out then "frost" does not form; that depends a lot on the air humidity.  The weird thing is that the Weather Compensations curves for heat pumps ought to have a kink, making the water a bit hotter than it otherwise would have been to compensate for the time spent defrosting.  But AFAIK that never happens.  Maybe it's built-in and not revealed to the end user?          

NedS

michaels said:

NedS said:

michaels said:

So I need a heat pump that can actually output 10kw at -3.

If the heat loss calculation for the property (or better, the actual measured heat loss) is 10kW at -3C, then yes on the coldest days (when it is -3C outside), you will need a heat pump that can actually deliver at least 10kW of heat to keep the house warm (at design temp).

I say at least 10kW, as that would require the heat pump to be running constantly to match the heat loss. But you should also factor in time out of heating for the DHW cycle(s), and also any defrost cycles, so most system designers would factor in an extra percentage to allow for this, and I would expect them to specify an 11or 12kW model as a result. The rated value (e.g, 10kW) is normally the output for a flow temp of 35C at an ambient 7C. Any given heat pump may put out more or less than it's rated nominal output at your design temp of -3C, so also check the manufactures output tables to see that it is actually capable of meeting your needs at those temps.

On the other side of that coin, it is unlikely to be -3C all day. It may be -3C over night, but may be warmer during the daytime, so less heat output will be required than the full 10kW that would be required at -3C. This will offset slightly against DHW and defrost cycles.

As you can see, it all starts with an accurate heat loss figure. Get that wrong and you either have a house you cannot adequately heat on the coldest days in winter, or a heat pump that is over-sized for the property and will be less efficient in milder autumn and spring months.

And indeed I do have a heat loss calc from Octopus of 9,5kw that while not perfect ties in with my max gas use.

So you are in the right ball park, and I wouldn't over-think it beyond that.

Compare with my situation - I have a heat loss calculation of 7.4kW, which I believe is on the high side. How high, I do not know, as since measuring my actual heat loss with our old oil boiler, we've since had substantial insulation upgrades at the same time as the ASHP install. We also keep our house closer to 19C than the 21C MCS calls for (and 23C in bathrooms!). Our actual heat loss, therefore, may be as low as 5kW, but likely somewhere in between (we will know in a year).

The range offers 5kW, 8kW or 12kW (which is actually a de-rated 16kW) models. On paper, the 8kW could barely deliver the "required" 7.4kW at the design temp of -2C, so due to the safety margin and factors discussed above (DHW and defrost cycles), they specified the 12kW model, which they admitted was not ideal. So we likely have an ASHP which is twice the size we may actually need, yet although far from ideal, is still kind of workable because we also have suitably sized radiators to allow running at flow temps of 35C. (I pushed like hell to get the 8kW model, offering to sign disclaimers etc, but they were bound by the confines of the scheme and would not budge).

If your heat loss is 9-10kW, and you are considering 9-11kW ASHPs, you are in the right ball park. If they were pushing a 16kW unit on you, and your actual heat loss turns out to be only 8kW, then you would be in a similar situation to me and it would be of more concern, especially if you don't upgrade your radiators and are forced to run high flow temps too.

Having determined that you are in the right ball park, I would focus on either reducing the heat loss with insulation upgrades and/or reducing flow temps by installing more / larger radiators.

Reed_Richards

NedS said:

Compare with my situation - I have a heat loss calculation of 7.4kW, which I believe is on the high side. 

That beats the calculated 8.95 kW for my 4-bedroom bungalow.  But I don't need my water to be as hot as calculated so that heat loss must have been an over-estimate.  

michaels

So how about an estimate of the trade off between scop and flow temp so I can think about rad investment?

Currently looking at 7 rads costing say £1500 (not including fitting) to get the flow temp at -3 down from 52 to 45.  Is it worth doubling this spend to get the temp down from 45 to 40?

Not forgetting we need about 18kwh of heat and 8kwh of hot water (where the scop won't change however many rads we change) and I am working on an electricity cost of 9p per unit.

[Maths would centre I guess on the 18kwh for heating, if the scop was 3 then this could cost 6000 x 3p = £540, with a scop of 3.5 it would cost £462 so assuming £1.5k extra on rads the payback time would be 19 years s probably not worth it.  Problem is I don't know what the 3 and 3.5 example SCOPS might actually be to inform the decision]

Thanks for all your thoughts.

Reed_Richards

The maximum theoretical efficiency of a heat pump is given by the formula

Th/(Th-Tc)

In the context of an air-to-water heat pump then Th is the leaving water temperature and Tc is the air temperature.  And Th has to be expressed in Kelvin.

So for Th = (273+52) = 325 and suppose it's 10 C outside then Th/(Th-Tc) = 7.7
      for Th = (273+45) = 318 and suppose it's 10 C outside then Th/(Th-Tc) = 9.1
      for Th = (273+40) = 313 and suppose it's 10 C outside then Th/(Th-Tc) = 10.4

You'll never get close to these theoretical maximum values but they might give a guide.  Suppose the average outdoor temperature when you want heating is 10 C (or supply your own estimate).  Then maybe the improvement in SCOP going from 52 C to 40 C would be 10.4/7.7 = 1.35 (35% higher)

Quite possibly there is a better way to so this but this was the best I could think of.  

   

NedS

Or you can use the manufacturers own tables. Using the published data for my own heat pump, for flow temps of 40, 45 and 50C at an ambient outside temp of 10C, I see quoted COPs of 4.53, 4.10 and 3.46 which should give an indication as to the extra efficiencies achievable when reducing flow temp (around 31% more efficient going from 50C to 40C).

Based on this, you can calculate the pay back time required for an investment in larger radiators in your house.

Reed_Richards

NedS said:

Or you can use the manufacturers own tables. Using the published data for my own heat pump, for flow temps of 40, 45 and 50C at an ambient outside temp of 10C, I see quoted COPs of 4.53, 4.10 and 3.46 which should give an indication as to the extra efficiencies achievable when reducing flow temp (around 31% more efficient going from 50C to 40C).

Based on this, you can calculate the pay back time required for an investment in larger radiators in your house.

So a 31% improvement going from 50 C to 40 C is actually quite close to my estimate of a 35% improvement going from 52 C to 40 C.  I'm quite pleased that I got so close.