Sizing your ASHP

[Stones]

Run a smaller ASHP harder or a larger ASHP more gently for a given heating requirement?

 

I've opted to do the latter, oversize so that I have spare capacity and so the heat pump only really has to tick over.  Just wondering what others have done or plan to do?

[Barney12]

From what I've read (and it's a subject that really makes my brain hurt!) your chosen path is the best option. Along that is with keeping the temp at an optimum for the COP.

[ragg987]

I was told that an inverter ASHP can only modulate down to 30% of capacity (unlike a gas boiler which can go down much lower). So when the heat load is lower (e.g. mild spring day) it might end up cycling more often if lowest modulation is still too much heat input. And to amplify this effect, ASHP COP will be higher so your 30% is worth more than the 30% on a really cold day.

 

In my case we had space heating demand of 3kW but went for a 7-8kW ASHP (the calculator said oversized by 263%) so as to improve DHW recovery times - household of 6 did not want to risk running out and having to wait forever to heat.

[Stones]

I suppose there is always going to be a point where the ASHP lowest modulation point is greater than your heating requirement and you end up with a degree of short cycling. 

[le-cerveau]

When I spoke to the Panasonic team they said that the lowest modulation was 30% so that agrees.  So there is always going to be a point where the heat demand is les than the 30% so put in a large enough thermal store/buffer to reduce the short cycling to an acceptable level yet to have sufficient spare capacity.

[ragg987]

We seem to be saying the same thing - my point was that "ticking over" for ASHP is lowest 30%, so no good trying to put in a 10kW ASHP when your heat load is 2kW. IMO.

 

It is a compromise - of course cycling is inevitable so put in a buffer, point would be to reduce cycling by combination of sizing and buffer. If yours is a very low energy house then perhaps DHW becomes the key factor. And if you use PV with solar diverter, or PT, then you can reduce ASHP for space heating in the shoulder months.

[Stones]

2 hours ago, Barney12 said:

From what I've read (and it's a subject that really makes my brain hurt!) your chosen path is the best option. Along that is with keeping the temp at an optimum for the COP.

 

Agreed, although interestingly we have experience of both routes on the forum, Jeremy has a larger ASHP and Jack a smaller 5kW.  Both seem to do the job from what I can gather.

 

39 minutes ago, ragg987 said:

We seem to be saying the same thing - my point was that "ticking over" for ASHP is lowest 30%, so no good trying to put in a 10kW ASHP when your heat load is 2kW. IMO.

 

It is a compromise - of course cycling is inevitable so put in a buffer, point would be to reduce cycling by combination of sizing and buffer. If yours is a very low energy house then perhaps DHW becomes the key factor. And if you use PV with solar diverter, or PT, then you can reduce ASHP for space heating in the shoulder months.

 

No PV in my case so totally reliant on the ASHP for heating and DHW.  I had at one point considered splitting heating and DHW, in which case a 5kW ASHP would have been the go to choice for heating, but I wasn't convinced a 5kW was sufficient for both.

 

The other issue is longevity, if you run a smaller heat pump for longer continuous periods compared to a larger heat pump where you have shorter run periods (lets say 2 hours on tickover vs 1 hour), does the extra compressor run time outweigh or equal the effect of shorter periods of operation on the lifespan of the compressor?

[Nickfromwales]

My knee jerk thinking is leaning towards running a bigger unit at a relaxed pace. 

Its not the same discipline across the board though, as with oil your told to size match as close as possible and run them hard for best efficiency. 

I think it's down to how much dhw you need and how often, and size from there. 

[ragg987]

My understanding is that ASHP COP will improve when modulating compared to at full capacity. However I am not sure this would reflect overall system efficiency as power consumption of ancillary devices like pumps, valves etc will become proportionately higher when pump is modulating.

 

Another factor could well be impact of ASHP start-up on COP. I have not seen any published data on this, however I would expect that maximum pump efficiency will not be achieved immediately on start-up - it might take a few minutes of operation to get there.

 

11 hours ago, Nickfromwales said:

I think it's down to how much dhw you need and how often, and size from there. 

+1

When your DHW is cool, would the ASHP not be running at maximum speed? I would only expect it to modulate when the return temp rises close to flow temp. So in this case the COP improvement from modulation has little benefit.

[Nickfromwales]

Ashp's switch into dhw 'mode' upon demand. Upon doing so they revert to a predefined or max flow temp output to produce dhw as fast as possible. Modulation won't come into that factor and CoP goes south until the cylinder is satisfied. 

Solar Pv is your friend when you have a heat pump. 

[TerryE]

Dave, IIRC you have a pretty thermally efficient house so the number of heating days (and even modest heating days) should be small  -- even where you live. 

 

You need to do the maths on the sustained heating demands but reckon on roughly 7W/Km² where the temperature is the delta between the slab and the room temperature (RT).  We estimate that we'll only rarely want the slab more than a couple of degrees warmer than RT.  This is why JSH went for the approach of using a buffer tank at ~35°C and a mix-down to circulate through the slab at 25°C.  This is the approach the we are going to adopt as well.  This means that having an AHSP minimum O/P of 30% isn't that much of the issue.  You need some control logic, but your ASHP will maintain the buffer tank in the 25-35°C range, and this will trickle feed the slab circuit.  The "mark" length will be dictated by the heat required to raise buffer tank capacity by ~10°C in temp, and the "space" dependent on insulation-dependent heat losses around the tank and the draw-down into the slab. 

 

I need to play with the heating algos, but my current thinking is to prime the buffer tank at or near the end of cheap rate so the slab is up to RT+ 2 and the buffer tank at temperature by 08:00, then only do buffer tank reheats during the day if needed.

 

As far as DHW goes, we are planning on using SunAmps heated by E7.  The SunAmps are raised to full ~85°C using its internal heater overnight, and these are then used to via a heat exchanger and mixer to raise the CW feed from a year-round average mains feed of say 15°C to 45°C or whatever for DHW.  

 

In order to use the ASHP in this scenario I would have to preheat the CW via heat exchanger from the buffer tank, but this assumes that we are using the ASHP to maintain the buffer tank at 35°C 24×7 year-round.  However we won't be using the buffer tank for slab heating ~10 months a year, so I suspect that the overall heat losses from maintaining the buffer tank at 35° will exceed any gains here. 

 

So the bottom line is that we've decided to use E7 only to heat the SunAmps.  This keeps the design KISS and has minimal heat losses.  Hence the price of the DHW will be the cost of heating actual DHW used by ~30°C (year round average) at E7 rates, except on the very odd occasion where we have guests and a high demand and therefore need a daytime boost.

 

PS, the bloody planners stopped us installing PV

Edited September 6, 2016 by TerryE

[jack]

We have a 5kW Panasonic Aquarea for our close-to-Passive-House-standard 289m2 house.  As I've mentioned elsewhere, we have quite a bit of shading and nothing like Jeremy's warm microclimate, so the ASHP spent a lot of winter ticking over gently keeping the insulated slab at around 21-22 degrees.  

 

As for DHW, over winter we have the ASHP heat the tank to 55 degrees between 4 and 6 o'clock each morning.  Looking at the temperature plots, that seemed to be about the maximum time it took to get the whole 250L tank up to temperature (and it did get to pretty-well bang on 55 degrees, despite the ASHP being located 9 or 10m away from the tank).  The top 30% of the tank is then heated to 85 degrees by the upper immersion heater that switches on at 6am.  

 

I don't believe we've ever run out of water with our family of four.

 

Also, we have 8.5kW of PV.  It occurred to me the other day that the ASHP probably hasn't had cause to turn on for several months due to the amount of energy being pumped into the tank every day.  I wonder whether not being used in this way could negatively impact longevity?

[Stones]

 

7 hours ago, jack said:

We have a 5kW Panasonic Aquarea for our close-to-Passive-House-standard 289m2 house.  As I've mentioned elsewhere, we have quite a bit of shading and nothing like Jeremy's warm microclimate, so the ASHP spent a lot of winter ticking over gently keeping the insulated slab at around 21-22 degrees.  

 

As for DHW, over winter we have the ASHP heat the tank to 55 degrees between 4 and 6 o'clock each morning.  Looking at the temperature plots, that seemed to be about the maximum time it took to get the whole 250L tank up to temperature (and it did get to pretty-well bang on 55 degrees, despite the ASHP being located 9 or 10m away from the tank).  The top 30% of the tank is then heated to 85 degrees by the upper immersion heater that switches on at 6am.  

 

I don't believe we've ever run out of water with our family of four.

 

Also, we have 8.5kW of PV.  It occurred to me the other day that the ASHP probably hasn't had cause to turn on for several months due to the amount of energy being pumped into the tank every day.  I wonder whether not being used in this way could negatively impact longevity?

 

Interesting.  Would you say your ASHP ran pretty much 24/7 during the coldest spell (albeit ticking over)?  In terms of DHW production, was there any noticeable or measurable drop in house temperature whilst the ASHP was otherwise engaged?

[Nickfromwales]

With most ASHP's heating Ufh in a slab / screed, I doubt if you'd ever notice the absence of it's input whilst briefly topping up / heating dhw tbh. 

The entire fabric of the property wil have warmed through too, so it's not just the slab that would be retaining / regulating heat.

[TerryE]

3 hours ago, Stones said:

Interesting.  Would you say your ASHP ran pretty much 24/7 during the coldest spell (albeit ticking over)?  In terms of DHW production, was there any noticeable or measurable drop in house temperature whilst the ASHP was otherwise engaged?

 

Jason, If it did then there'd be something wrong with the thermal design of the house. Our net loss in a typical January is estimated at around 11 kWhr / day or less than 0.5 kW sustain.  OK our slab isn't going to perform as well as modelled but this isn't going to lift this by more than about 25-30%.  If the minimum output of a 5 kW ASHP is say 1.5 kW then it's mark/space ratio should still be around 30% in the depths of winter.

 

And as Nick says, the thermal capacity of the slab will totally dominate any mark/ space ripples.

 

As Jeremy kept repeating: do the calculations and have faith in them.

 

The main complication from a design PoV is whether you are going to attempt to use the ASHP to prime for DHW and to do this you need a TS or a zones PWC as in Jack's case.  Here you have to balance the gearing of the ASHP against the extra heat losses from a TS or PWC.  Your call.

Edited September 6, 2016 by TerryE

[jack]

3 hours ago, Stones said:

Interesting.  Would you say your ASHP ran pretty much 24/7 during the coldest spell (albeit ticking over)?  In terms of DHW production, was there any noticeable or measurable drop in house temperature whilst the ASHP was otherwise engaged?

 

I didn't really check it.  Sometimes it was on (generally with the fan barely moving) when I went near it, other times it was off.  I'm not presently set up to record how much energy it uses or when it's on/off - that's a plan for when everything else is finished!  

 

Certainly there was no discernible change in slab temperature.  This image shows the slab temp for one of the sensors across January (each step is 0.5 degrees - I'll eventually swap out this one-wire sensor for one with a higher resolution):

 

 

The biggest temperature difference across an entire day was about 1 degree (bearing in mind the resolution limitations) and usually it's a lot less than that. 

 

Of course, the ASHP was only in DHW mode for two hours every morning, so not a lot of time for the temperature to fall.

[Stones]

Terry, I appreciate that there probably wouldn't be a noticeable drop in temperature, but given the low hysteresis some people are employing to control their heating, there may well be a measurable drop of 0. something of a degree, which would in turn have a small knock on effect in terms of the smaller heat pump having to work a little harder catching up or running a little longer. Again, appreciate very difficult to measure this and probably not an issue in reality.

 

I've already committed and gone for a slightly larger unit than I need for a number of reasons. Whilst my heating requirement shouldn't be that great, the design of my house isn't anything near optimum for passive design. The unit I've gone for, should if the performance data is reliable, be able to heat the house, on a day without any solar gain, operating at the lowest modulation point. This gives plenty of remaining capacity for DHW production, and heating capacity if ambient temperature falls lower than average lowest temperatures or we have higher than average wind speeds. 

[le-cerveau]

I am looking at the same conundrum at the moment as I firm up the final specifications so my quandary is:

I have a heat load of 4057W (last PHPP calculation worst case) so call it 4kW in the depths of winter 5kW if we don’t get the house quite as up to spec as I put in PHPP.

I will have a TS for DHW and UFH buffer, the ones I am looking at are split 60/40 (bottom/top) with a baffle plate so the lower is heated by the ASHP and the upper by a conventional gas boiler.  The TS has coils in both the lower and upper parts for DHW production.

I have 5½ bathrooms (2 baths (large) with separate showers, 3 showers + toilet) so potentially massive DHW call, however over previous discussions have whittled it down to max 3 simultaneous showers (or 2 showers and a bath).  The showers will be low flow 9-12 L/min so about 30L/min max.  (I will have a cold water accumulator to provide flow).

I intend to run the lower part of the TS at around 35^oC, good for the COP of the ASHP (a la JSH setup) and heat the top to around 50-60^oC depending on anticipated DHW load (how many people are home).

I am playing with the following variables:

ASHP power: 7, 9, 11 kW.
Boiler Power: 12-30 kW.
TS size: 500, 750, 1000L.
Number of concurrent Showers: 2, 3 (or a bath).
Follow on Showers: 0, 1, 2.

I have tried to do some calculations but TS’s are notoriously hard to model so I am working on simple tank methodology to give me some approximations. And have come up with the following:

The boiler will be a simple 12kW heat only affair, anything more is overkill because of the TS.

The TS will be 750L.

I could go for a 500L TS but could not sustain a bath and 2 showers, so would require domestic management (doable), it easily supports 5 simultaneous showers with follow-ons. To go for the 500L that I would need  to go larger on the ASHP definitely 9 and probably 11 and 30kW on the boiler to be safe for the bath load.

With the 750L TS a 12kW boiler is sufficient to do the bath and 2 showers with the TS at 35/55 (bottom/top).

The ultimate question is 7kW or 9kW ASHP, the 7kW is enough (I reckon), but the 9kW gives me more spare capacity, more power in to the lower part of the TS so less gas used for DHW, or just go with the 7kW?

[ragg987]

17 minutes ago, le-cerveau said:

 lower is heated by the ASHP and the upper by a conventional gas boiler.  

Not sure I get it - if you are paying UK prices and are on mains gas then why put in the ASHP as well? It seems you are adding complexity to your system, plus increasing capital cost quite a lot, and I doubt if the running costs of ASHP will be much lower than mains gas - at least not enough to recover the additional outlay.

[Stones]

Have to agree, if you have mains gas, I would go with that.  I certainly wish I had the option of mains gas.

 

When it came to considering the size of ASHP I thought I would need, I looked at my average heating requirement based on mean minimum temperature in the coldest month, peak heating requirement based on high (gale force) wind speed -which has the effect of increasing infiltration and a % reduction in theoretical maximum output of the ASHP if run for 24 hours, to take account of defrosting.

[joe90]

Where is Jeremy when you need him ?

[TerryE]

Jason, with a decent insulation layer and a stone skin, the decrement delay factor is so large that using minimum temperatures is a mistake, IMO.   At worst you should use daily averaged minimums, and even consider a 90 %ile or the likes.  In our case we rarely get daily averages under -5°C and if push comes to shove we can always get a 2kW electric oil heater out as an emergency boost.  But we'll never get anywhere close to that with a 5kVA ASHP.

 

OK, up where you live, it's a bit more wild and chilly during the winter, but I still think that its a mistake to compound contingencies because you still need to optimise your system for the likely load and if you have oversized 3 or 4× then this will be hard to do.

 

In our case, I ran the figure on the extra costs to use the ASHP for DHW and not and doing so didn't give sensible paybacks.   We can always rejig our design in a year or two if this assumption was incorrect.

Edited September 8, 2016 by TerryE

[Stones]

Appreciate the point about decrement delay, but it is only a delay, at some point heat is going to have to be added back into the building. Re temperatures, poor wording on my part, it is daily averaged minimum temps that I have used.

 

Contingencies (storm force wind, temperatures lower than 0C) are really the issue for me.  If I had based requirement simply on averaged minimum temperature and average wind speed factor then a 5  kW unit would be able to provide both heating and DHW, albeit it would be operating at around 75% of capacity.  Depending on the variables used, the remaining capacity would cover some contingencies, but not all.  The larger unit I have gone for would be operating at around 45% of capacity leaving a healthy margin for contingencies.

 

Like you I ran figures on the costs of using the heat pump for DHW but came to a different conclusion to you. If I have got it wrong and the ASHP I've bought is so oversized that it causes problematic short cycling issues (the biggest risk as I see it), then I can always change it for a smaller unit.  Thankfully, ASHPs do seem to hold reasonable residual value and I guess I'm going to know within a fairly short time whether I have got it right or wrong.

 

 

[le-cerveau]

Ok, to answer the question, why not just stick with gas:

doing the calculations based on average Gas and Electricity prices in the NW (0.039 £/kWh and 0.136 £/kWh)  (Gov 2015 figures), taking into account boiler efficiency (91%), boiler electricity consumption (37W), the breakeven point for an ASHP is 2.84 (Assuming the circulation pump is in the CoP calculation).  If the circulation pump is excluded (68W) then it is CoP of 3.52 (can't find a definitive answer).  There will be extensive PV on the property and in the long run gas prices will inevitably rise faster than electricity.

 

As to size, working on Stones theory of absolute worst case 4401W (PHPP largest figure) then a 9kW ASHP would be operating at 49% capacity and a 7kW one at 63% capacity (CoP 4.13 & 4.46), however looking at the performance chart (A2/W35) the 9kW machine will only produce 6.7kW and the 7kW machine still produce 6.55kW so 66% and 67% capacity with CoP's of (3.13 and 3.34).  So logic would say go for the 7kW machine as it produces almost as much heat, when you need the most, and is slightly more efficient, my head hurts.

[le-cerveau]

I have done a bit more digging into 7, 9 & 12kW ASHP performance (I am looking at the Panasonic range).

 

WHSDC07H3E5
Water Out (oC)	35			40
Outdoor Air (oC)	Capicity        (W)	Input Power (W)	CoP	Capicity        (W)	Input Power (W)	CoP
-15	4600	1980	2.32	4600	2190	2.10
-7	5150	1920	2.68	5076	2140	2.37
2	6550	1960	3.34	6575	2290	2.87
7	7000	1570	4.46	7000	1835	3.81
25	7000	970	7.22	6740	1140	5.91

 

WHSDC09H3E5
Water Out (oC)	35			40
Outdoor Air (oC)	Capicity        (W)	Input Power (W)	CoP	Capicity        (W)	Input Power (W)	CoP
-15	5900	2660	2.22	5650	2820	2.00
-7	5900	2340	2.52	5850	2610	2.24
2	6700	2140	3.13	6650	2380	2.79
7	9000	2180	4.13	9000	2485	3.62
25	9000	1260	7.14	8660	1475	5.87

 

WHSDC12F6E5
Water Out (oC)	35			40
Outdoor Air (oC)	Capicity        (W)	Input Power (W)	CoP	Capicity        (W)	Input Power (W)	CoP
-15	8900	3620	2.46	8500	3790	2.24
-7	10000	3660	2.73	9600	3950	2.43
2	11400	3310	3.44	11000	3530	3.12
7	12000	2530	4.74	12000	2960	4.05
25	12000	1660	7.23	11800	1940	6.08

 

These figures are taken from the service manual + the CoP calculation.  Assuming I take the water at 35^oC with my house load (max) of 4401W this gives me a comparison of:

Option	% required	CoP
7kW  @ -7	85.46%	2.68
9kW @ -7	74.59%	2.52
12kW @ -7	44.01%	2.73
7kW  @ 2	67.19%	3.34
9kW @ 2	65.69%	3.13
12kW @ 2	38.61%	3.44
7kW  @ 7	62.87%	4.46
9kW @ 7	48.90%	4.13
12kW @ 7	36.68%	4.74

 

This would suggest I discount the 9kW model due to poor performance at 2^oC and lower CoP, so then it is down to 7kW or 12kW.

The 7kW machine is running at 63-85% capacity from 7^oC to -7^oC, whilst the 12kW machine is at 36-44% capacity with a better CoP.

 

So in theory the12kW machine can modulate down (30%) quite happily though at 7^oC the load will start dropping.

All this is ignoring less that ideal build (higher heat load) and DHW preheat.

 

So is a 7kW machine working reasonably hard a better bet than the 12kW machine that is idling but with a greater DHW reserve?

 

The pain continues.