ASHP 101, how does a reversible heat pump work?

[Dreadnaught]

Beginner's question. How does a heat pump reverse to both heat and cool?

 

Is the following correct? Simply put, a split heat pump comprises:

 

1. outside: a heat-exchanger coil and a fan, which is connected to…
2. gas piping, with a compressor, which is connected to…
3. Inside:
   1. if it is air-to-air, another heat-exchanger coil and fan
   2. if it is air-to-water, a plate heat exchanger
4. which is connected to liquid piping and a one-way valve, which is connected to  (1) above to close-the-loop

 

(Circulating from outside to inside is not water but refrigerant.)

 

From Wikipedia:

 

 

So far so good. For cooling it makes sense. But how does the system go in reverse to heat the inside? It would seem to me you would need to physically swap the circuit around. I am sure this is just a problem of over simplification but is there someone who can explain it in simple terms?

[Jeremy Harris]

All heating ASHPs reverse, they have a four port reversing valve that switches the relative positions of the condenser and the evaporator.  The reason they have to have this is to defrost the evaporator.  By switching the four port reversing valve the heat pump can draw heat from the water side and use it to warm up what was the evaporator (but is now the condenser) on the air side.  This melts any ice build up, at the cost of sucking some heat energy out of the house whilst it's doing it. 

 

This is why a badly set up ASHP can have a really bad COP if it ends up defrosting fairly regularly, as defrosting not only means it's not heating for the duration of the defrost cycle but it's also extracting heat from the house, which then needs to be put back the next time the unit switches to heating mode.

[Dreadnaught]

Thanks @JSHarris. Four-port reversing valve the magic words that makes it all make sense.

 

Does anyone know of the relative efficiency in each of the two directions, heating and cooling. I have read somewhere that heat pumps are optimised for one direction, either heating or cooling, and quite majorly so.

[jack]

There's a good explanation here.

[Dreadnaught]

@jack brilliant link. Thank you. What a handsome object:

 

[SteamyTea]

1 hour ago, Dreadnaught said:

Beginner's question. How does a heat pump reverse to both heat and cool?

PV/T = C

Where:

P = Pressure

V = Volume

T = Temperature

C = Constant

 

(this is for an 'ideal gas', but good enough for this)

 

So separate the the working of the Heat Pump (Carnot Cycle, you can't get better) and the Plumbing, which is just where you want the energy delivered to.

 

[MrMagic]

2 hours ago, Dreadnaught said:

Thanks @JSHarris. Four-port reversing valve the magic words that makes it all make sense.

 

Does anyone know of the relative efficiency in each of the two directions, heating and cooling. I have read somewhere that heat pumps are optimised for one direction, either heating or cooling, and quite majorly so.

 

For A2A heat pumps I've certainly noticed that there has always been a bit of a discrepancy between the heating and cooling power & efficiency - annoyingly HP manufactures seem to like to measure heating and cooling performance differently to make it even harder to compare. I'm sure someone smarter than me can explain.

 

e.g. https://www.orionairsales.co.uk/daikin-air-conditioning-ftx20kv-wall-mounted-inverter-heat-pump-20kw7000btu-a-240v50hz-6533-p.asp

 

Cooling capacity (Min./Nom./Max): 1.3 / 2.0 / 2.6 kW
Heating capacity (Min./Nom./Max): 1.3 / 2.5 / 3.5 kW
Consumption power - cooling (Min./Nom./Max): 0.31 / 0.503 / 0.72 kW
Consumption power - heating (Min./Nom./Max): 0.25 / 0.524 / 0.95 kW

 

...so this system has up to 1Kw of additional capacity when in heating mode (max).

 

Using my very basic maths to get a rough CoP - 

Cooling CoP (Nominal) - 2 / 0.503 = 3.97

Heating CoP (Nominal) - 2.5 / 0.524 = 4.77

 

[SteamyTea]

Te reason that there is a difference in name place capacity is because of the relative differences in temperatures and entropy.

Probably find that if you plot on the Kelvin scale those difference diminish and the only thing affecting it is the entropy caused by humidity.

[Dreadnaught]

So @SteamyTea, if understand you correctly, you are suggesting that the difference in CoP in @MrMagic's example are to be explained by physical factors (humidity), not engineering. That's really interesting. That would suggest that my statement "that heat pumps are optimised for one direction, either heating or cooling, and quite majorly so" is actually not true.

 

Did I understand you correctly?

 

[SteamyTea]

Yes, basically true that a heat pump just cools one side and heats another.

If the temperature differences, within the working range of the refrigerant gas, is equal, then there should be no difference.

In practice there may well be a difference, but that could be caused by the plumbing side i.e. not capable of pumping enough because of UFH pipe restrictions, surface areas not large enough for the temperature differences, the fact that cool air falls to the floor, warm air rises which will effect efficiency.

That sort of thing.

Edited July 24, 2018 by SteamyTea

[Jeremy Harris]

That's borne out in the spec for the air-to-air heat pump that's built in to our MVHR.  The heating capacity is about 1.5 kW, the cooling capacity is just a bit over 1 kW, but the working conditions for the input air temperature and humidity are markedly different, as is the Δt

[Nickfromwales]

2 minutes ago, JSHarris said:

That's borne out in the spec for the air-to-air heat pump that's built in to our MVHR.  The heating capacity is about 1.5 kW, the cooling capacity is just a bit over 1 kW, but the working conditions for the input air temperature and humidity are markedly different, as is the Δt

 

Could you remind me which MVHR unit you went for please Jeremy?  

[Jeremy Harris]

Just now, Nickfromwales said:

Could you remind me which MVHR unit you went for please Jeremy?  

 

Ours is a Genvex Premium 1L, a bit over-sized for our house, but we went for the  Premium 1L over the Premium 1 because the ~50% great er capacity gave a similar increase in the cooling capacity from the built in air-to-air heat pump.

 

It's an expensive bit of kit though, and I'm near certain that we could have the same performance from just having a water fed duct cooler run from the MVHR.  The idea is that if you have an ASHP, then you can plumb in a duct heater/cooler to the fresh air feed and then use the ASHP to provide additional air cooling/heating.  The £150 duct heater/cooler I've just bought should have a cooling capacity of around 1.5 kW, about the same as the Genvex, but without needing an additional heat pump.  I'm part way through making the duct cooler, but have been deflected by the rush of people wanting to look at our old house.

[Nickfromwales]

Ok, so for folk without an ASHP, the 'geothermal' brine loop to duct heater / cooler is a serious contender. With no other means to cool, the Genevex also has a good bit of appeal.

Ill have a look at prices. Thanks.

[Jeremy Harris]

2 minutes ago, Nickfromwales said:

Ok, so for folk without an ASHP, the 'geothermal' brine loop to duct heater / cooler is a serious contender. With no other means to cool, the Genevex also has a good bit of appeal.

Ill have a look at prices. Thanks.

 

 

Yes, a buried pipe heat dump, capable of losing around 2 kW of so to the ground, should work,  It'd need to be sized to give a flow temp to the duct cooler of around 8 to 9 deg C and accept a return temp of around 20 deg C at a guess, so a long enough buried pipe (it's need to be about 1 to 1.5m down) that could lose that much heat to the ground would work.  It'd need a circulating pump, expansion vessel, fill loop and PRV (although technically I can't seen reason for needing a PRV, but BC might get sniffy if there isn't one).

[Nickfromwales]

1 minute ago, JSHarris said:

 

 

Yes, a buried pipe heat dump, capable of losing around 2 kW of so to the ground, should work,  It'd need to be sized to give a flow temp to the duct cooler of around 8 to 9 deg C and accept a return temp of around 20 deg C at a guess, so a long enough buried pipe (it's need to be about 1 to 1.5m down) that could lose that much heat to the ground would work.  It'd need a circulating pump, expansion vessel, fill loop and PRV (although technically I can't seen reason for needing a PRV, but BC might get sniffy if there isn't one).

Cant help thinking that this can be cobbled together for less money. 

[PeterW]

Just found a use for that spare UFH pipe ?

 

How much do you think it would need @JSHarris in terms of a loop ..??

[Nickfromwales]

100m is the minimum according to all my research. 

[SteamyTea]

4 minutes ago, Nickfromwales said:

100m is the minimum according to all my research. 

Isn't that dependant on property size.

 

Have a hunt around on https://www.engineeringtoolbox.com/ as they have all the details.

[PeterW]

This is of use too

 

https://www.egshpa.com/ground-loop-sizing/

[Nickfromwales]

1 minute ago, SteamyTea said:

Isn't that dependant on property size.

Every single nut and bolt is, just thats the minimum that I've been "told" will be effective.  Ive considered doubling the pipework and doubling the size or number of duct heat ex too. Cooling isn't to be sniffed at so I'm giving this some serious looking at, for folk who don't have an ASHP. Just need to be mindful of over-doing it and causing condensation issues in the distribution duct-work. 

[Jeremy Harris]

32 minutes ago, Nickfromwales said:

Every single nut and bolt is, just thats the minimum that I've been "told" will be effective.  Ive considered doubling the pipework and doubling the size or number of duct heat ex too. Cooling isn't to be sniffed at so I'm giving this some serious looking at, for folk who don't have an ASHP. Just need to be mindful of over-doing it and causing condensation issues in the distribution duct-work. 

 

 

No worries about condensation inside the duct work, as the condenser will be the duct cooler, so the air will be at it's coolest as it leaves the duct cooler and will have condensed out most of the water vapour in contains on the duct cooler surfaces (bit like the condensation on a cool can of beer).  The air inside the ducts will gradually warm up a bit as it picks up heat, so will have a greater capacity to hold water vapour.  The duct sides will always be warmer than the air flowing through the duct so won't attract condensation on the inside.  There's a small risk of condensation on the outside of the ducts, but my experience with the Genvex (which chucks out cool air at around 12 deg C) is that I've never seen any condensation on the outside of the ducts.  This may be in part because the semi-rigid duct is twin wall, so inherently has a degree of insulation.

[Nickfromwales]

2 hours ago, JSHarris said:

This may be in part because the semi-rigid duct is twin wall, so inherently has a degree of insulation.

Noted. Thanks again.

[Stones]

When I looked at fitting a wet duct heater / cooler, there were two problems.  The first was the limited air flow in an MVHR system, and the second was the size of duct heater / cooler you would have to fit to get any meaningful heating / cooling.  At best I found I was only going to be able to deliver around 0.5kWh worth of cooling (or heating) through the MVHR (based on the flow temps I would be working with from the ASHP).

 

https://www.systemair.com/en-GB/UK/Products/Product-selector/Fans1/

 

Has a useful calculator that allows you to play around with input criteria - target air temp, air flow rate, flow temps etc to determine the size of unit you would need and how much kWh it will deliver. 

 

@JSHarris  What size of duct cooler have you got?

 

More discussion here:

 

 

[Jeremy Harris]

3 minutes ago, Stones said:

@JSHarris  What size of duct cooler have you got?

 

 

It's a special I bought from Alnor in Poland, and should give me around 1.5 kW of cooling with the ASHP at 9 deg C and the MVHR on speed 3.  I can up this to about 1.8 kW with the MVHR on full boost, I think.  I'm building the double row heat exchanger into a special housing, as my installation's a bit odd - I'm cooling the fresh air intake before the MVHR, and leaving the MVHR manually set to 100% bypass (which we do in the summer anyway).  The main reason for me fitting it this way is that it's a lot easier and I want to add an additional coarse intake filer any way, and it's not much more work to fit the  heat exchanger into this additional external filter box.