Heatpump output vs efficiency dilemma

[Archer]

Hi All, 

 

We're in the process of specifying a ducted A2A system for our house, for heating only (hot water supplied via a separate system). System setup is a little unusual and restricted because we are using the existing ground floor ductwork from a gas warm air system, with the upstairs/1st floor supply coming courtesy of new insulated ducts in our loft (ie. with ceiling diffusers). 

 

Our engineer has specced a system from Mitsubishi Heavy Industries which he is happy will work in the existing GF ductwork, but has said that we need to split the system with 2x indoor units (one for each set of ducts/each floor). He also thinks - and I am not 100% sure about this, but it's the advice we've been given - that it's not possible to use a single outdoor compressor unit to supply both indoor ducted fan coil units. A google search sort of backs this up - the MHI ducted A2A units come packaged essentially with a specified indoor unit. 

 

Ok so anyway - I have now completed a heat loss survey - through Heat Engineer and also through the Heat Punk app - they've come back surprisingly low and suggest whole house heat loss around 7.8kw @ -2.5 degrees OAT (and recommended a 8 or 12 kw pump for the whole house). As I said in a different thread, our engineer originally specified a whopping 26kw in total (1x 14 and 1x12); I queried this and he has revised to 17kw (1x7 and 1x10). 

 

Looking for any advice on here  that people can give; I'll put the data sheets below, but the MHI pumps appear to have really excellent modulation outputs; the 10 can operate down to 1.7kw and the 7 to 1.1kw. So unless there are any "gotchas" to be aware of they both should be fine to operate without excessive short-cycling even if they are significantly oversized? 

 

The other main difference between them is the efficiency, with the 7kw unit looking quite poor with COP @ 7degrees  of just 3.76 (only 2.73 for cooling). The larger pump is more respectable - around 9-15% more efficient for heating/cooling using the same parameters (COP 4.08). Confusingly the SCOP for both models is almost identical (4.13) which doesn't make much sense... Sorry about the long read - my main question is should I be asking for the the larger output unit(s) to be installed for both floors, even though they are significantly oversized because of the efficiency increase? Or is it bad to have the pump operating towards the bottom of its output and 2x7's might give better overall performance?  I trust the engineer and want to use them and there isn't a big cost difference between the difference size pumps, just thinking about what will be best in day to day use (ie. more affordable, more comfortable etc). 

 

Any advice massively appreciated. Cheers - 

Edited August 11, 2023 by Archer

[Archer]

10kw Option

 

 

7kw Option

 

 

[DanDee]

24 minutes ago, Archer said:

but has said that we need to split the system with 2x indoor units (one for each set of ducts/each floor). He also thinks - and I am not 100% sure about this, but it's the advice we've been given - that it's not possible to use a single outdoor compressor unit to supply both indoor ducted fan coil units. A google search sort of backs this up - the MHI ducted A2A units come packaged essentially with a specified indoor unit. 

In the brochure that I shared the link to, you have the outdoor unit with either 1 single indoor unit, 2 indoor units or 3 indoor units.

The system with 2 indoor units having a better COP than the one with 1 indoor unit, but this only when they work in tandem.

 

P.S. Yet it will be more efficient to run 2 separate systems, than one outdoor with 2 indoor(multisplit) if you need to run the upstairs and downstairs separate. 

 

Edited August 11, 2023 by DanDee

[Archer]

18 minutes ago, DanDee said:

In the brochure that I shared the link to, you have the outdoor unit with either 1 single indoor unit, 2 indoor units or 3 indoor units.

The system with 2 indoor units having a better COP than the one with 1 indoor unit, but this only when they work in tandem.

 

P.S. Yet it will be more efficient to run 2 separate systems, than one outdoor with 2 indoor(multisplit) if you need to run the upstairs and downstairs separate. 

 

Thanks Dan, appreciated. 

 

Any ideas why the SCOP for the two sized units is so similar consider the COP figures are quite different? In my mind it seems preferable to go for 2x 10's because the efficiency is so much better, but it's quite oversized and specifying 2x7 or 1x7 and 1x10 gives a more appropriate overall output for the house. Is SCOP for A2A calculated differently from A2W (because they allow for a cooling season / no DHW load)?

[DanDee]

1 minute ago, Archer said:

Thanks Dan, appreciated. 

 

Any ideas why the SCOP for the two sized units is so similar consider the COP figures are quite different? In my mind it seems preferable to go for 2x 10's because the efficiency is so much better, but it's quite oversized and specifying 2x7 or 1x7 and 1x10 gives a more appropriate overall output for the house. Is SCOP for A2A calculated differently from A2W (because they allow for a cooling season / no DHW load)?

I do not have an answer for that.

[SteamyTea]

40 minutes ago, Archer said:

Confusingly the SCOP for both models is almost identical (4.13) which doesn't make much sense

 

3 minutes ago, Archer said:

Any ideas why the SCOP for the two sized units is so similar consider the COP figures are quite different

Because they are using statistical averages.

 

Take a simple journey in two very different cars. A small diesel and a heavy petrol. The journey takes the same time, but the small diesel uses less fuel.

But we can also measure it another way, the speed variations. Say the small diesel never goes above 50 MPH, but never below 20 MPH, but the the large petrol does 70 MPH but also 0 MPH. Mean speed is the same, 35 MPH.

Then we can measure how the speed us distributed, this may swing the the MPG figure in favour of one or the other vehicles i.e. lots of town driving or lots of open road driving.

 

[Archer]

4 minutes ago, SteamyTea said:

 

Because they are using statistical averages.

 

Take a simple journey in two very different cars. A small diesel and a heavy petrol. The journey takes the same time, but the small diesel uses less fuel.

But we can also measure it another way, the speed variations. Say the small diesel never goes above 50 MPH, but never below 20 MPH, but the the large petrol does 70 MPH but also 0 MPH. Mean speed is the same, 35 MPH.

Then we can measure how the speed us distributed, this may swing the the MPG figure in favour of one or the other vehicles i.e. lots of town driving or lots of open road driving.

 

But following that analogy through for the heatpumps  is harder because they are both achieving the same outputs - ie we know that one is 10% more efficient at 7 degrees outside temperature so why is there a 0% difference across the whole seasonal average?

 

Or to put it another way, which one should I trust and piu my money in - the COP or the SCOP?

[SteamyTea]

36 minutes ago, Archer said:

ie we know that one is 10% more efficient at 7 degrees outside temperature so why is there a 0% difference across the whole seasonal average?

Because 7⁰C does not happen very often.

 

Go for the larger one, less chance of frosting up.

[markocosic]

Go for two far smaller ones. Oversizing not a good idea for cycling, life expectancy, comfort etc. COP often drops at minimum output too.

 

One thing to note is distribution of work. Upstairs unit will do almost nothing in heating season. Downstairs unit will no almost nothing in cooling session. Because cold air falls. On this basis a multi split is perhaps not quite so silly in efficiency terms 

 

Consider a single 10kW unit with a 7kW head downstairs and a 5 kW head upstairs perhaps?

 

Or a pair on single splits at say 5 kW and 3.5 kW (cooling rating; just outputs will be higher). Might even get away with 3..5 kW and 2.5 kW on the cooling rating.

 

They will then modulate down beautifully in all conditions and operate near to their efficient mid load operating point most of the time.

[Archer]

I've been doing some further reading online about this and I'm sharing here in case it's of use to anyone else in a similar position. 

 

Just to be absolutely clear I am in no way expert - what I've read and am posting below may not hold true and goes against conventional wisdom on sizing so happy to be challenged/ corrected by other more knowledgeable forum members...

 

So to recap, the proposed system for our house is 2x outdoor and indoor A2A units with a total output of 17kw (1x10 serving ground floor, 1x7 kw serving first floor). The total estimated demand based on Heat loss is a little under 8kw @ -2.5 degrees OAT, our MCS design temp. So significantly oversized. 

 

My concern was mainly that oversizing would lead to excessive cycling and inefficiency, but both HP's have a low minimum output - 2.8kw in total; so the question in my mind was why is oversizing a problem if the pumps can modulate down? 

 

A few considerations from the info and chart below:

 

1. PDesign/ output @ -10 degrees;

Heatpumps lose capacity, quite significantly towards the lower end of their temperature range. In my case the total output drops from 17kw to 12.1kw @-10 degrees (presumably an exceptionally rare temperature north of London)

 

2. Inverter HP's run less efficiently at the higher end of their range (80-100% output capacity); they also seem to run more efficiently at the lower- to mid point of range (30-60% capacity). These differences can be quite considerable (the research paper below suggests from 10% to 40% efficiency increase in tests). 

 

3. Manufacturers set a cut-off for the lowest output a compressor will operate at based on it's efficiency, but this tends to drop significantly at the very bottom of the range (less than 15-20% of max output). This concept is shown in the image below from the John Cantor presentation. I've found other similar ones online showing similar findings. The efficiency peak seems to be around 40% of peak capacity. 

 

So what does this suggest (in my mind)?

 

- Pumps should be sized at least 25% more than the design load to ensure that inverter efficiency is maintained and to provide a safety factor for extreme weather; 

 

- Use the minimum output as the key figure not just the maximum output; Heat Pump's with a high turndown ratio may be more suitable (for example, larger HP's can have better COP efficiencies than equivalent smaller pumps in the same range; lots of technical chat about this on Green building Forum - see links below. 

 

Size to avoid or minimize excessive cycling using the minimum output figure - ie. What heat loss is there at 12 degrees OAT?

 

- The minimum output should be significantly less than 50% of MCS design heatloss (ie. 8kw heat loss, minimum output of less than 3 kw). This should avoid excessive cycling in most cases and ensure that peak inverter efficiency is achieved in most outdoor conditions. 

 

https://aceee.org/files/proceedings/2014/data/papers/1-371.pdf

 

https://www.greenbuildingadvisor.com/question/how-does-efficiency-in-a-mini-split-vary-over-various-operating-levels

 

https://www.greenbuildingadvisor.com/question/oversizing-a-multi-zone-inverter-heat-pump-does-it-matter

 

Any thoughts...?

[TonyT]

Would 2 units heat the remote corners of the house?

 

I plan on fitting one in my family room 60m2 to provide heating/cooling  and another at a later date in the ground floor hall to heat the hall/first floor hall and any rooms off it

 

plan will be to utilise off peak electricity to preheat the home and if required top up with the combi boiler if required.

 

[markocosic]

3 hours ago, Archer said:

Pumps should be sized at least 25% more than the design load to ensure that inverter efficiency is maintained

 

I don't think so.

 

The time spent at full load is very small. (plot the temperature difference each hour of the year as a load duration curve)

 

Lots will be spent in the happy 25-75% range if your unit is sized at 100% of design load.

 

E.g.

 

 

From:

 

https://www.researchgate.net/publication/291352700_Stochastic_Modelling_and_Simulation_of_Energy_Flows_for_Residential_Areas

 

With VDI reference method from:

 

 VDI, “4655 Referenzlastprofile von Ein-und Mehr-
familienhäusern für den Einsatz von KWK-Anlagen,”
Verein Deutscher Ingenieure (VDI), pp. 3–18, 2008.

 

You can use hourly temperature data and a "balance point" to get similar curves for your choice of climate and building fabric standard.

[JamesPa]

6 hours ago, markocosic said:

I don't think so.

 

The time spent at full load is very small. (plot the temperature difference each hour of the year as a load duration curve)

 

Lots will be spent in the happy 25-75% range if your unit is sized at 100% of design load.

I'm beginning to wonder if there is a good argument for undersizing a (A2W) heat pump, particularly one with poor modulation range, or if not for deliberately undersizing, at least for erring towards undersizing, rather than the 'received wisdom' of erring towards oversizing then adding 20% for (insert your reason here).

 

The BRE model

 https://tools.bregroup.com/heatpumpefficiency/index.jsp and data seems to suggest their might be in at least some cases, although until I understand it I remain sceptical about some of the figures it throws up.  

 

Without more robust data on the degradation due to both short and normal cycling, it's difficult to be certain.  However it's entirely plausible that having a supplemental electric heater and suffering a cop of 1 for 20% of the load 10% of the time (when the cop would otherwise be only about 2-2.5 anyway) is a good trade off vs compromising the performance for much of the heating season.

Edited August 17, 2023 by JamesPa

[Beelbeebub]

BRE have looked at this topic. 

 

They have an online calculator/simulator you can play with. 

 

TLDR, the simulator implies slight undersizing is beneficial mainly due to lower cycling most of the time outweighing the use of booster heating occasionally. 

On 11/08/2023 at 09:20, Beelbeebub said:

Not sure if this has been posted before

 

https://tools.bregroup.com/heatpumpefficiency/index.jsp

 

Appears BRE have developed an efficency predictor tool based on more typical UK conditions than the SCOP calculation that manufacturers use in their literature. 

 

Their calculation also includes system loads like pumps and backup heating as well as hot water loads. 

 

 

One curiosity and I have to dig into the details) is that it appears the seasonal efficency (their version of SCOP) improves with undersized units using direct electric backup heaters for the few occasions needed. Their explanation is the reduction in cycling for the rest of the year more than compensates for the few days of inefficient direct heater use. 

 

[JohnMo]

3 hours ago, JamesPa said:

I'm beginning to wonder if there is a good argument for undersizing a (A2W) heat pump

I generally thought that is the assumption, even MCS use that model, you size for 99.6% of the year, not the coldest day you thought someone may have seen.

[JohnMo]

Also I know that Grant make the case for under rather than oversizing and the volumiser they sell have an immersion built in, with heat pump controller used to add immersion heat when the heat pump can no longer give enough heat to heating circuit.

[JamesPa]

24 minutes ago, JohnMo said:

I generally thought that is the assumption, even MCS use that model, you size for 99.6% of the year, not the coldest day you thought someone may have seen.

Yeah, after assuming unreasonable air change values, ignoring fabric upgrades then rounding up to the next size.

 

I don't think sizing for 99.6 could be described as undersizing.

Edited August 17, 2023 by JamesPa

[Beelbeebub]

I think Vaillant and others have similar though I think they are optional upgrades to the hydrobox or unitower indoor units.

[JamesPa]

5 minutes ago, Beelbeebub said:

I think Vaillant and others have similar though I think they are optional upgrades to the hydrobox or unitower indoor units.

They do, does anyone design to use them I wonder.

 

Interesting if as @JohnMosays, Grant advocate undersizing.

[Archer]

17 hours ago, TonyT said:

Would 2 units heat the remote corners of the house?

 

I plan on fitting one in my family room 60m2 to provide heating/cooling  and another at a later date in the ground floor hall to heat the hall/first floor hall and any rooms off

 

I think it depends on your proposed set up. You can get away with one large wall unit for a large room (I've seen this in classroom situations for example in non-resi set ups). Whether this gives completely even distribution across the room I'm not sure. 

 

In our case it's a ducted system with a set of floor vents (ground floor) and ceiling vents (1st floor). 

[JamesPa]

17 hours ago, Archer said:

My concern was mainly that oversizing would lead to excessive cycling and inefficiency, but both HP's have a low minimum output - 2.8kw in total; so the question in my mind was why is oversizing a problem if the pumps can modulate down? 

My understanding is that it isn't much of a problem for A2A.  The modulation depth of A2W is much less and this is where the problem lies AFAIK.

[DanDee]

18 hours ago, Archer said:

My concern was mainly that oversizing would lead to excessive cycling and inefficiency, but both HP's have a low minimum output - 2.8kw in total; so the question in my mind was why is oversizing a problem if the pumps can modulate down? 

 

37 minutes ago, JamesPa said:

My understanding is that it isn't much of a problem for A2A.  The modulation depth of A2W is much less and this is where the problem lies AFAIK.

 

The excessive oversizing would indeed lead to both cycling and inefficiency.

 

In heating mode a single wall split A2A will start cycling if the vane and fan speed are not set in a way that warm air output is not pushed as far as possible in the room, the air temperature around the unit rise faster that for example your case, where you will use a flow/return duct, that cycles the air around the house. Another issue with some units is the room thermistor that is seated on top or next to the condenser coil, where it reads a higher air temp than the rest of the room, triggering cycling.

The inefficiency arises from the same effect of the air warming up around the unit, which reduces the heat transfer rate vs the production. Again having a central intake with outputs in different rooms can mitigate this, to an extent.

 

In cooling mode(in countries where this is highly important) oversizing again triggers not only cycling, but the unit running less rather than continuously, it removes less humidity from the air, creating discomfort.

 

So hobbyists have to find a ways to run the unit such that it blows the air away from the unit, use a ceiling fan and move the thermistor or find a unit where the remote is the room thermistor.

 

 

Testing when the wife is not home.

↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ 

[Archer]

I love the fan strapped to the ceiling, that's absolutely brilliant!!

 

Thanks @DanDee for the detailed post as well, that's really helpful. I think from what you've said, and other comments above, oversizing shouldn't be a huge issue hopefully on our system (because of the high modulation range) but could be on others. Seems like there is a key distinction between A2A systems (which generally have a higher ratio) and A2W; even more so with non-inverter systems. In both cases it may be preferable to undersize if anything and rely on a backup heat source for extreme conditions.

 

For some A2A systems that modulate right down it's worth considering that larger output units may be more efficient than smaller, "correctly sized" ones in some circumstances. The Mitsubishi range for example have a number of units in the same series with the same minimum output but increasing max output's. The larger units have significantly higher COP values. I guess this is very situation specific though. 

 

In our system design we will have two wall thermostats, one for each floor and they'll be located away from ducts to hopefully avoid the distortion effect that you referenced. Because we're re-using the 1970's gas warm air ducts on the ground floor it's a single return duct in the hallway. Upstairs they'll be a supply and return duct for each room. 

 

The other thing I was wrestling with was optimizing the ceiling ducts for both heating and cooling which seem to be at odd's with each other. As I understand it, heated air needs to be forced directly down to reduce stratification and a warm layer at ceiling height (ie. what your ceiling fan does)... In cooling mode you want the opposite - a horizontal spread so the cool air naturally sinks and mixes. 

 

Found these Halton diffusers with a wax actuator that automatically changes the air throw and direction based on flow temperature. Seems like a neat solution, will post back once we've installed to say if they've worked out. 

 

 

Edited August 17, 2023 by Archer

[SteamyTea]

1 hour ago, Archer said:

I think from what you've said, and other comments above, oversizing shouldn't be a huge issue hopefully on our system (because of the high modulation range) but could be on others. Seems like there is a key distinction between A2A systems (which generally have a higher ratio) and A2W; even more so with non-inverter systems. In both cases it may be preferable to undersize if anything and rely on a backup heat source for extreme conditions.

Worth checking how the HP modulates, some just open or close the expansion valve, that just reduces output without saving any input.

[Beelbeebub]

The one thing I would caution about modulation range, which I noticed from A2W units and may (or may not) apply to A2A.... 

 

The modulation range is generally quoted as the max and min outputs at a given temp. 

 

So it may be a max of 12kw at - 5C and 3kw min at - 5C

 

So 4:1

 

Except you don't use the minimum modulation at - 5C. Thatvs when you need max power! 

 

You need min modulation when it's warmer, say 10C or 15C.

 

Thr minimum a HP (at least the ones I looked at) can modulate down to at the higher temps is higher than at the lower temps - which is obvious really. 

 

But that means the minimum you HP might be able to turn down to, when you actually need it to turn down, is higher than you might think. Say 5kw 

 

When I looked at the "max at minimum vs minimum at maximum" values I was seeing a modulation ratio closer to 3:1

 

Does that make sense?