ASHP vs Oil in 2022

[PhilT]

On 09/10/2022 at 08:45, JohnMo said:

Gas at 10p is £1650, energy in to house 13200kWh

 

HP at 3.5 CoP 1320/3.5 =3771kWh

 

3771kWh at 34p is £1282.

 

Best part of £400 a year saving

 

 

I reckon a SCOP of 4 to 5 is achievable for the latest models, well installed and set up. I just re-worked my numbers based on these latest costs (including daily fixed rates) and my ASHP is saving me £1,148 pa with no heat loss saving measures. Yes granted the old boiler was probably less than 80% (perhaps as low as 70%) efficient, but that is probably no different from the average uk household, still operating at crazy flow temps up in the 70s degC (the legal max is now 55degC). Even allowing for 70% efficiency I'm getting an implied SCOP of around 5.

[MortarThePoint]

2 hours ago, PhilT said:

I reckon a SCOP of 4 to 5 is achievable for the latest models, well installed and set up. I just re-worked my numbers based on these latest costs (including daily fixed rates) and my ASHP is saving me £1,148 pa with no heat loss saving measures. Yes granted the old boiler was probably less than 80% (perhaps as low as 70%) efficient, but that is probably no different from the average uk household, still operating at crazy flow temps up in the 70s degC (the legal max is now 55degC). Even allowing for 70% efficiency I'm getting an implied SCOP of around 5.

 

What model ASHP do you have?

[PhilT]

25 minutes ago, MortarThePoint said:

 

What model ASHP do you have?

Mitsubishi Ecodan R32 (PUZ-WM112A). It's similar to most other latest brands/models using scroll compressors with R32 or R290 refrigerant. I was lucky to find an installer who really knew their stuff and did a fantastic job.

[MortarThePoint]

12 hours ago, PhilT said:

Mitsubishi Ecodan R32 (PUZ-WM112A). It's similar to most other latest brands/models using scroll compressors with R32 or R290 refrigerant.

 

Sounds great. That's the unit I'm considering, but I'm tossing up between 3phase and single phase. 3phase is £500 extra.

 

12 hours ago, PhilT said:

I was lucky to find an installer who really knew their stuff and did a fantastic job.

Programming the controller scheme or were there other aspects to look out for?

 

[MortarThePoint]

I have found a databook from Mitsubishi Ecodan and here is some performance data for their 11.2kW unit:

https://library.mitsubishielectric.co.uk/pdf/book/ATW_Databook_R32_2020#page-60-61

 

Based on my SAP and that data (nominal), I should achieve (my model) an average COP of 4.34 with a flow temp of 35C or 3.45 with a flow temp of 45C. Having put in UFH upstairs as well as downstairs, I will hopefully be able to use the lower flow temperature (UFH power output W/m2 data below).

 

This all ignores domestic hot water, but I'm OK with working out the space heating needs as a priority.

 

Modelling as linear above 2C ambient would have been fine.

 

 

 

Here's UFH heat output vs flow temperature:

data from link below except red which is from a good linear fit.

https://www.tradingdepot.co.uk/info/plumbing/polypipe/underfloor-heating-heat-output-tables/

[MortarThePoint]

A reasonable formula for the COP of the Ecodan 11.2kW is

COP=(0.521 - 0.0068E*FT)*AT + (4.44 - 0.0485*FT)

where AT is the ambient temperature and FT is the flow temperature, both in C.

 

 

 

 

Using that I get my expected average COP based on the flow temperature with the assumption that the flow temperature stays constant throughout the heating season:

COP = 7.47 - 0.088*FT

That assumption feels reasonable given most of the demand is across 3 months with an ambient temperature within 1C (DEC-FEB) and the rest is mostly still within an ambient range of 2.8C (NOV - MAR).

Edited October 13, 2022 by MortarThePoint

[MortarThePoint]

Samsung have come out with an interesting looking new ASHP, EHS Mono HT Quiet, R32, Heat Pump, 12.0kW. Also available 8kW and 14kW. Available single and 3-phase. Unlike with Ecodan/Zubadan the 14kW unit is the more standard 1m height which could be a big win for people who need a little extra power. Prices look similar. Black in colour.

They have a headline COP figure, but then give power out and power in data at varying FT and AT:

 

On first inspection the COP looks worse, as at most ambient temperatures the resulting values are lower. However, at the key temperatures in the South of England, they have done a better job of optimising the design and get better numbers (suspicious?).

 

Red is the better:

 

14kW model has similar COP numbers in the 2C to 12C range (1% worse than the 12kW model):

 

The basic linear model makes for average COP values of 4.41 at FT=35C, 4.05 at FT=40C and 3.68 at FT=45C. That's about 2.4% better than the Ecodan figures, but the linear model doesn't fairly represent the improved better optimisation of the Samsung (e.g. 9% better COP at AT=7C).

 

The Ecodans have been around for longer, so perhaps there are some optimisations not reflected in their data. There's more in field experience for the Ecodan which is a plus.

Edited October 13, 2022 by MortarThePoint

[MortarThePoint]

The 11.2kW/12kW units are very similar in quietness (47 vs 46), but the Samsung 14kW units are a fair bit quieter than the Ecodan units.

 

 

[GLPinxit]

I've recently replaced a 20 year old oil boiler with an ASHP (Ecodan 11.2). The oil boiler sounded like a jet taking off, plus pump hum when not firing. The ASHP itself is as close to silent as I can imagine is possible- we were sitting 20 feet from it in the summer and plants waving in its breeze were the only indication it was running. ASHP also has pump hum but no quieter than before (Wilo pumpsj. Note, however, that it was commissioned in August so it hasn't yet run at full chat.

[Kevm]

14 hours ago, MortarThePoint said:

Samsung have come out with an interesting looking new ASHP, EHS Mono HT Quiet, R32, Heat Pump, 12.0kW. Also available 8kW and 14kW. Available single and 3-phase. Unlike with Ecodan/Zubadan the 14kW unit is the more standard 1m height which could be a big win for people who need a little extra power. Prices look similar. Black in colour.

They have a headline COP figure, but then give power out and power in data at varying FT and AT:

 

On first inspection the COP looks worse, as at most ambient temperatures the resulting values are lower. However, at the key temperatures in the South of England, they have done a better job of optimising the design and get better numbers (suspicious?).

 

Red is the better:

 

14kW model has similar COP numbers in the 2C to 12C range (1% worse than the 12kW model):

 

The basic linear model makes for average COP values of 4.41 at FT=35C, 4.05 at FT=40C and 3.68 at FT=45C. That's about 2.4% better than the Ecodan figures, but the linear model doesn't fairly represent the improved better optimisation of the Samsung (e.g. 9% better COP at AT=7C).

 

The Ecodans have been around for longer, so perhaps there are some optimisations not reflected in their data. There's more in field experience for the Ecodan which is a plus.

I've done a bit of similar modeling with my 14kW Ecodan.  Using the databook data works for comparison but the actual figures achieved are always a bit less.  The databook represents constant running and the Ecodan (and I'm sure other ASHPs) use a bit more power starting and stopping (i.e. cycling or temp set backs).  You also need to be careful with defrosting.  Some manufacturers include this in their numbers (Ecodan does but just the 2 degrees line in the table).  Also the ASHP won't be running at its 'nominal' speed all the time and you can't control that.  The Ecodan seems more efficient at its 'Mid' running and slightly less so at MIn and Max.  I don't about other manufacturers.    

[MortarThePoint]

30 minutes ago, Kevm said:

I've done a bit of similar modeling with my 14kW Ecodan.  Using the databook data works for comparison but the actual figures achieved are always a bit less.  The databook represents constant running and the Ecodan (and I'm sure other ASHPs) use a bit more power starting and stopping (i.e. cycling or temp set backs).  You also need to be careful with defrosting.  Some manufacturers include this in their numbers (Ecodan does but just the 2 degrees line in the table).  Also the ASHP won't be running at its 'nominal' speed all the time and you can't control that.  The Ecodan seems more efficient at its 'Mid' running and slightly less so at MIn and Max.  I don't about other manufacturers.    

 

I think I'm more drawn towards the Ecodan due to its proven track record. If I needed 14kW it would be a different matter.

 

My SAP shows a peak monthly heat demand of around 2700kWh in December with an average outside temp of 4.2C. that works out as about 90kWh/day so about 8 hours of 11.2kW (3.75kW average). That's with an indoor to outdoor dT=18.6-4.2=14.4C. If outside goes to -5C then dT=23.6C and that would make for 90*(23.6/14.4)=150kWh/day so about 13hours of 11.2kW (6.1kW average). There would be an additional hot water demand but pretty low (max. 300kg * 50C * 1Wh/kgC = 15kWh so about 1 - 2 hours worst case). Bear in mind that would have to be a day with average temp of -5C which where I live is unlikely.

[MortarThePoint]

Dimensional comparison:

[JohnMo]

Reality check

Why are you sizing for an on time of 13 hrs.  If you run on WC your on time is 24 hrs, but your loading is less during the set back time so your -5 degC DT becomes 20.6degC.

 

150kWh/d becomes(150/24)*13 = 81.25 spread over 13 hours or 6.25kWh per hour.  The night time case is less than that, so no need to calculate.

 

I think you need to go back to basics and calculate the heat demand for your house, use the spreadsheet on here.  As I believe the way your are going you are talking yourself in to a much bigger HP than you need.  Looks like you need a heat pump near 8kW to me.

[SteamyTea]

12 minutes ago, JohnMo said:

sizing for an on time of 13 hrs

It is a hard habit to get out off.

People think that storage heaters are only 'on' for 7 hours at night, and this is why the house is hot in the morning, and cold in the evening.

The two dials on the old fashioned ones that I have are a complete mystery to most people.

[MortarThePoint]

4 hours ago, JohnMo said:

Reality check

Why are you sizing for an on time of 13 hrs.  If you run on WC your on time is 24 hrs, but your loading is less during the set back time so your -5 degC DT becomes 20.6degC.

 

150kWh/d becomes(150/24)*13 = 81.25 spread over 13 hours or 6.25kWh per hour.  The night time case is less than that, so no need to calculate.

 

The hour figures are simply outputs that show how long the ASHP needs to be on for at 100% to provide the required energy. It could be run at a lower setting for longer if that is desirable.

 

I have done a basic thermal time constant model that suggests the house will cool down (i.e. loose heat unheated) at a rate of 0.1C/h when it is 4.2C outside (so dT = 14.4C). The cooling rate is 0.07/h or the time constant is 144h (6days) (half life 4.2days). With 11.2kW I'd almost be able to only run the ASHP during E7 electricity and only suffer 1.4C temperature changes. I know the COP would be worse as colder at night, but the electricity is cheaper. I don't fancy making a 60kWh battery at the moment to allow the electrical demand to be spread across the day.

 

4 hours ago, JohnMo said:

I think you need to go back to basics and calculate the heat demand for your house, use the spreadsheet on here.  As I believe the way your are going you are talking yourself in to a much bigger HP than you need.  Looks like you need a heat pump near 8kW to me.

 

I know this already from the SAP: Total Fabric Heat Loss 234 W/K   (A)

Ventilation heat loss (max): 138 W/K    (B)

Heat transfer coefficient (max): 372 W/K    (A+B)

 

If AT=4.2C and indoor is 18.16C then average power requirement is 372W/K * (18.16C - 4.2C) = 5.19kW if there were no solar/metabolic/appliance/etc gains. In December the solar gains seem to average 0.47kW and the internal gains are about 1kW.

5.19kW - 1.43kW = 3.76kW

SAP has a monthly space heating demand for December of 2774kWh:

2774kWh / (24h/day * 31days) = 3.73kW

Water heating is shown in the SAP as 220kWh/mo so that averages to:

220kWh / (24h/day * 31days) = 0.3kW.

In theory I could have a smaller ASHP, but I would want extra grunt for when it is unusually cold. 8KW would suffice (down to AT = -11C).

 

Other than capital cost, what is the main downside of having too big a unit?

 

Do you mean Jeremy's Spreadsheet? 

 

Edited October 14, 2022 by MortarThePoint

[JohnMo]

Some down sides

 

Turn down issues when not at min outside temp.

 

Steady state has a higher minimum kWh loading due to larger motors.

 

Bigger starting in rush current to get motors turning.

 

Bigger cables, fuses etc all adding incremental cost increases

 

Higher minimum water flow requirements. Requiring larger diameter pipe runs, considerable cost adder.  Much more likelihood of short cycling, as min flow requirements are not met.

 

Floor hot spots, due higher than ideal water flow temps 

 

More likelihood you will need a buffer, offsetting some efficiency gains elsewhere.

 

Big ugly unit in garden instead of a smaller ugly unit in garden.

 

Good possibility a small unit could run on solar in the summer, no chance of the big one doing so.

 

There may be reasons

 

 

[MortarThePoint]

30 minutes ago, JohnMo said:

Higher minimum water flow requirements. Requiring larger diameter pipe runs, considerable cost adder.  Much more likelihood of short cycling, as min flow requirements are not met.

 

I've installed UFH upstairs and downstairs so should be able to dissipate 11.2kW with a flow temp of 35C or a bit above.

 

Is short cycling when it turns on for a short duration? I imagine that is influenced by the allowed temperature swings on the thermostats. Is that correct? If so the cycle on time would be the same if I allow proportionately larger temperature drops.

 

30 minutes ago, JohnMo said:

Big ugly unit in garden instead of a smaller ugly unit in garden.

 

All my options are the same size I think. Vary between brands though.

 

[JohnMo]

I did batch floor charging end last years heating season, works fine, but not always great, as the energy you put into the floor, comes out but not always the way you want it to.  If you have under charged the floor the house temp drops a little more than you plan, if you overcharge it will be a little warmer than you plan.

 

To work well you need to constantly modify the charge time, to match the average temperature outside.  This was a pain, so this year I moved to WC with a small overnight setback to match the house/floor inertia (offset about 6 hours). 

 

My boiler is oversized, but I have a huge buffer (160l), so have room to try different schemes.  Current input into floor is varying between 0.5 and 0.8 kWh depending on outside temp.

[MortarThePoint]

Are Ecodans binary in power delivery? Full on or full off or can you get an 11.2kW to deliver 6kW of power? I presume the COP would be worse, but is there data on that?

[JohnMo]

Short cycling is where the boiler/ heat pump, starts runs for short period and trips on high return temps. It is then locked out 10 mins and then repeats.

 

Lots of energy goes into heating all the metal in the system as this sucks the heat from the water.  Not energy efficient.

[JohnMo]

You would need to find out the modulation rate, inverter heat pump do modulate.  So yes you could if the figure you quoted is in the modulation range.

 

Heat pump deliver best running long and slow.

 

My batch charge temp was 30 degs for 7 hours.  My WC flow temperature at the moment 23 deg, during the day and 21/22 at night.  Floor temp is 21-21.5, room 19-19.5, thermostat are moved to 22 to limit influenced the room heating.

 

Your issue is when the average temp drops enough to require you to put the heating on, the heat demand is really low, and the big heat pump can't cope and short cycles.

[PhilT]

34 minutes ago, MortarThePoint said:

Are Ecodans binary in power delivery? Full on or full off or can you get an 11.2kW to deliver 6kW of power? I presume the COP would be worse, but is there data on that?

The 11.2 has a scroll compressor which gives a much flatter efficiency curve (COP vs ambient) than the twin rotary compressor in the 8.5, so better efficiency at very cold temps. Power delivery is constantly variable down to around 30% of full power so it can deal with low loads really well. Also I very much agree about the desirability of having the extra headroom for the really cold nights.

Edited October 14, 2022 by PhilT

[JohnMo]

1 minute ago, PhilT said:

desirability of having the extra headroom for the really cold night

What 6kW of head room on a max 6kW heat demand? On a house with heat/time constant of 6 days?

 

Not sure I could agree.  But I will bow out, and leave you to it.

[Kevm]

4 hours ago, MortarThePoint said:

Are Ecodans binary in power delivery? Full on or full off or can you get an 11.2kW to deliver 6kW of power? I presume the COP would be worse, but is there data on that?

Look at the data book sheet you posted.  It gives output and COP for Min, Med, Nominal and Max.  Min is (I assume) the lowest continuous output.   

[Kevm]

5 hours ago, MortarThePoint said:

 

I've installed UFH upstairs and downstairs so should be able to dissipate 11.2kW with a flow temp of 35C or a bit above.

 

Is short cycling when it turns on for a short duration? I imagine that is influenced by the allowed temperature swings on the thermostats. Is that correct? If so the cycle on time would be the same if I allow proportionately larger temperature drops.

 

 

All my options are the same size I think. Vary between brands though.

 

These new Samsungs are quite chunky, 1270x1018x530 mm to be precise.  That's a volume of 0.69m3.  I don't know where you live but the limit for permitted development in England is 0.6m3.  Having to apply for planning permission would put me right off.