Tolerance in ASHP sizing? Drawbacks of oversizing?

[embra]

Our heating engineer has prepared an MCS heat loss calculation to size an ASHP. There are some factors in there which I feel are conservative (such as ventilation losses which don't consider air tightness improvements), and when comparing to my own best case calculations I've found we could potentially drop to a heat pump with a lower output (e.g. 7 kW to 5 kW), saving a bit of money on equipment costs.

 

I have tendency to over analyse things, so I'm wondering if there's a general rule of thumb around the tolerance in heat pump sizing, or if there are any specific drawbacks or watch outs when it comes to potentially oversizing a system due to conservative assumptions in the heat loss calculations... any guidance most appreciated!

 

edit: if it helps, we're looking at a Vaillant Arotherm Plus unit

Edited July 24, 2023 by embra

[sharpener]

54 minutes ago, embra said:

if it helps, we're looking at a Vaillant Arotherm Plus unit

 

There is full performance data here, it seems not to be available on English-language web sites but is easy enough to work out.

 

Your experience is similar to mine, installers play safe and make conservative assumptions as do Vaillant tech support.

 

Watch out for lower than necessary OAT (Outside Air Temperature) assumptions, they should vary by region, the table is on p10 here.

 

Also if you have additional heat sources e.g. WBS (Wood Burning Stove), AGA, MVHR (Mechanical Ventilation and Heat Recovery) make sure they know these are available to help out below the design OAT. Equally supplementary water heating by solar PV. These considerations are not officially allowed by MCS but might tip the balance of thinking.

 

Arotherm + specs seem to be quite generous so you may well be right and could get away with a smaller unit. I was orginally told their 12kW was too small but after arguing that the proposed 15kW Stiebel Eltron was physically too large they reworked the figures based on 18C everywhere (?) and conceded the Vaillant would do the job. They also proposed to put lever valves in so an inline heater could if necessary be installed in the light of experience over the first winter.

[embra]

9 minutes ago, sharpener said:

 

There is full performance data here, it seems not to be available on English-language web sites but is easy enough to work out.

 

Your experience is similar to mine, installers play safe and make conservative assumptions as do Vaillant tech support.

 

Watch out for lower than necessary OAT assumptions, they should vary by region, the table is on p10 here.

 

Also if you have additional heat sources e.g. WBS, AGA, MVHR make sure they know these are available to help out below the design OAT. Equally supplementary water heating by solar PV. These considerations are not officially allowed by MCS but might tip the balance of thinking.

 

Arotherm + specs seem to be quite generous so you may well be right and could get away with a smaller unit. I was orginally told their 12kW was too small but after arguing that the proposed 15kW Stiebel Eltron was physically too large they reworked the figures based on 18C everywhere (?) and conceded the Vaillant would do the job. They also proposed to put lever valves in so an inline heater could if necessary be installed in the light of experience over the first winter.

 

Thanks for sharing, strange that this information is not available in all regions. 

 

Do you know what "rps" means in the below tables?

 

[MikeSharp01]

12 minutes ago, sharpener said:

water heating by solar PV.

Did you mean PV (Photo Voltaic) or could this be Solar Thermal (ST)!

[Beelbeebub]

I beleve it"s "revolutions per second" of the compressor.

 

Basically the modulation of the HP from full power (120rps) down.

 

Note the most efficient speed is not full speed and varies with external temp and flow temp.

 

Also note the max power delivered can increase with falling external

Edited July 24, 2023 by Beelbeebub

[sharpener]

9 minutes ago, MikeSharp01 said:

Did you mean PV (Photo Voltaic) or could this be Solar Thermal (ST)!

 

PV since that is what I have but could of course be either like @embra in the other thread

[JohnMo]

Drawbacks are

Not enough modulation on the shoulder months, so could short cycle during these months. Which will impact CoP.

May need a buffer to compensate, which if not really careful will impact CoP.

[JamesPa]

5 hours ago, JohnMo said:

Drawbacks are

Not enough modulation on the shoulder months, so could short cycle during these months. Which will impact CoP.

May need a buffer to compensate, which if not really careful will impact CoP.

Agree.  Remember that most of the time the unit will be running at perhaps 60pc of peak capacity and either end of the season one quarter or less.

 

The downside of under sizing is that you may need some source of secondary heating if it's persistently very cold.  Depending on where in the country you live, this may happen very rarely.

 

Good luck though trying to persuade an installer to downsize. The advantage to the installer of oversizing is that they get fewer call outs and definitely meet the BUS requirement to deliver 100% of the load.  From an installers pov it's almost a no brainer.

 

I'm in exactly the same situation.  In the depths of last winter 7.5kW 24*7 did the job, based on smart meter recordings.  The calcs say 10-11kW.  A '7kW' vaillant is really about 8.5kW at my design FT and the -2 design temp.  It's a nice unit. Do I go for it or something 'safe'.  I may even be forced to a spreadsheet to work out the penalty for a top up at COP 1 for the few days per year it might happen.

 

Still TBD, hopefully in the next couple of weeks.

Edited July 24, 2023 by JamesPa

[embra]

1 hour ago, JamesPa said:

Agree.  Remember that most of the time the unit will be running at perhaps 60pc of peak capacity and either end of the season one quarter or less.

 

The downside of under sizing is that you may need some source of secondary heating if it's persistently very cold.  Depending on where in the country you live, this may happen very rarely.

 

Good luck though trying to persuade an installer to downsize. The advantage to the installer of oversizing is that they get fewer call outs and definitely meet the BUS requirement to deliver 100% of the load.  From an installers pov it's almost a no brainer.

 

I'm in exactly the same situation.  In the depths of last winter 7.5kW 24*7 did the job, based on smart meter recordings.  The calcs say 10-11kW.  A '7kW' vaillant is really about 8.5kW at my design FT and the -2 design temp.  It's a nice unit. Do I go for it or something 'safe'.  I may even be forced to a spreadsheet to work out the penalty for a top up at COP 1 for the few days per year it might happen.

 

Still TBD, hopefully in the next couple of weeks.

I'm currently crunching the numbers to see where my projected space heating load across the year sits against the Vaillant output / CoP charts shared by @sharpener. Will post back with findings... (not that I am over analysing this 🙃)

[JohnMo]

30 minutes ago, embra said:

currently crunching

A 10 deg swing between inside and outside temp halves or doubles the heat loss depending which way you look.

 

So in my case the calculated heat loss is 1.3kW at 10 while at 0 it is 2.6kW. However the generated heat from people and appliances is a bigger proportion the warmer it gets outside, so the actual heat required at 10 degs is less than calculated.

 

A heat pump with 50% turndown is therefore massive at 10 degs if it's been designed already bigger than it needs to be at -5.

 

The key is having enough system capacity so the on off cycles allow a run time of at least 10 minutes. The run cycle at warm temps could be 10 mins on 40 mins off or longer.

[Beelbeebub]

It would seem that improving the modulation is fairly crucial for heatpumps.

 

Gas boilers have good modulation. So as long as you get the heat load about right there is plenty of wiggle room. If you oversized the boiler a bit, no problem.

 

With HPs that is more difficult.  If it were easier and you could just fit a "12kw and under"cunut that would be fine down to 2kw output it would make installers life a lot easier.

 

I did notice York heatpumps seem to have a broad output range , from 6kw to 14kw in the same chassis size. 

 

That does open up the possibility of a very easy swap over of units if the power supply is rated for it.

[sharpener]

6 hours ago, Beelbeebub said:

That does open up the possibility of a very easy swap over of units if the power supply is rated for it.

 

From data sheets it seems for a given frame size they differ only in refrigerant charge, software and possibly max compressor, fan and pump revs.

Hence it would seem entirely possible to downrate a unit once installed if it proves to be too big, does anyone offer that as a feature?

[dpmiller]

Very easy to change the max speed on the CoolEnergy units.

[Beelbeebub]

32 minutes ago, sharpener said:

 

From data sheets it seems for a given frame size they differ only in refrigerant charge, software and possibly max compressor, fan and pump revs.

Hence it would seem entirely possible to downrate a unit once installed if it proves to be too big, does anyone offer that as a feature?

Or presumably up rate with a larger charge?

 

The "dream" would be to change the rating of a physical unit to match the load simply by adding or removing some charge and changing the software parameters. That would be a win for most people, the manufacturer would have simpler production, suppliers could stock one unit, installers would only need to handle one unit and could adjust "in situ" for the property.

 

I'm pretty sure gas boilers already do this. I was told my WB 18kw was identical to the 12, 15 and 22kw units. Just a different chip in the PCB (which they won't sell you!). Vaillant go even better, their units are configurable by the installer who fits one unit with a wide range and the configures it for the property.

 

Question:

 

My understanding is the upper limit of performance is related to the charge. There is only so much heat you can transfer per kg of fluid.

 

The lower limit is related to how slowly you can run the compressor before it becomes too inefficient.

 

If the York units are physically the same (IE have a big coil and compressor for the 14kw but are down charged and slowed down for the lower ratings) could you extend the modulation of a HP by having a "reservoir" that you can increase or decrease the working charge with?

 

IE, there is a gas bottle off the HP circuit and you can pump gas into it before sealing it off and then the remaing working circuit has less charge land the software rejigs). When you want more capacity you just bleed more gas from the bottle back into the system?

[DanDee]

35 minutes ago, Beelbeebub said:

 

Question:

 

My understanding is the upper limit of performance is related to the charge. There is only so much heat you can transfer per kg of fluid.

 

 

A refrigeration system power and efficiency is obtain by matching the compressor size, evaporator and condenser size, and the charge.

The heat pumps already have a charge storage device/buffer/receiver, but the actual power size is mainly due to the compressor size, the rest it's just to match the output required and the efficiency of the system.

The speed of the compressor drives the power output by varying the system pressure, a pressure too low makes the refrigeration cycle inefficient or it even stops, no pressure no flow.

In order to have lower output that the minimum of a given compressor, you need a smaller compressor. A vrf has two compressors than run as one or two.

 

[JamesPa]

The argument might be tipping in favour of oversizing!

 

https://www.theguardian.com/environment/2023/jul/25/gulf-stream-could-collapse-as-early-as-2025-study-suggests