42mm piping?

[anonymous]

So I'm in a position where I have to replace my copper piping regardless because they're microbore

 

Should I then just go straight to 42mm pipework? I ask after seeing this 

 

Surely I could get away with maybe a bit less if I insulate the house well first? Or is the 42mm piping effectively a fail safe? Ie: insulation does eventually need replacing 

[PeterW]

I’ve fitted copper of that size once in 30 years and it was copper waste.. never even seen it used in a domestic situation, even with ASHP 28mm is max needed. Decent flow and return backbone using 22mm Hep2O, tee off it into copper at 15mm for the rad tails if you really want to but unless it’s a 20 bed mansion and you’ve got a lot of rads, you don’t need anything none standard here.

 

[sharpener]

Advice towards the end of that thread looks more sensible e.g. this

 

<The higher flow rates requirements comes from the smaller dT across the radiators that is sought - gas/oil systems typically specify 20°K across the rads, whereas heat pump wants 5. All else being equal, the flow rate is quadrupled. In round numbers, 1m3 per hour for each 6kW of output. Typically we want the water velocity less than 1m per second in the pipes, so more about 6kW then best to have 28mm flow from the heat pump and branch to 2x 22mm circuits to the radiators (e.g. upstairs/downstairs might be convenient).>

 

I will probably bring 28mm from the HP into the house up to the point the various utilisation circuits split off, after that I am expecting 22mm to be adequate. There is a useful flow rate calculator here. I am a bit surprised @PeterWsuggests 22mm Hep20 as this shows its cross-section is only 2/3 that of the equivalent copper but I might use it for the 28mm run as above, I hope it will be sufficiently flexible to avoid a lot of small joggles done with elbows.

 

Edited March 20, 2023 by sharpener

[PeterW]

4 minutes ago, sharpener said:

22mm Hep20 as this shows its cross-section is only 2/3 that of the equivalent copper

Less restrictions due to less need for elbows etc means the difference is negligible. On a pair of supply backbones upstairs and downstairs in a 5 bed house you won’t see a difference between the two other than time and cost to install. 

[sharpener]

13 minutes ago, PeterW said:

Less restrictions due to less need for elbows etc means the difference is negligible.

 

I can believe that, the ability to create swept bends at will is a definite plus.

 

As posted elsethread the new ch in our house was plumbed in plastic but using swaged elbows at every bend so they have got neither the penny nor the bun!

[anonymous]

32 minutes ago, sharpener said:

Advice towards the end of that thread looks more sensible e.g. this

 

<The higher flow rates requirements comes from the smaller dT across the radiators that is sought - gas/oil systems typically specify 20°K across the rads, whereas heat pump wants 5. All else being equal, the flow rate is quadrupled. In round numbers, 1m3 per hour for each 6kW of output. Typically we want the water velocity less than 1m per second in the pipes, so more about 6kW then best to have 28mm flow from the heat pump and branch to 2x 22mm circuits to the radiators (e.g. upstairs/downstairs might be convenient).>

 

I will probably bring 28mm from the HP into the house up to the point the various utilisation circuits split off, after that I am expecting 22mm to be adequate. There is a useful flow rate calculator here. I am a bit surprised @PeterWsuggests 22mm Hep20 as this shows its cross-section is only 2/3 that of the equivalent copper but I might use it for the 28mm run as above, I hope it will be sufficiently flexible to avoid a lot of small joggles done with elbows.

 

Thanks. I did think 42mm seems a bit unnecessary 

 

I won't however be getting a heat pump itself, until after I've done a load of other efficiency upgrades 

 

I do still have gas, but I intend on making sure it's effectively heat pump ready, but in a more "modular" way 

 

The other efficiency upgrades have an ROI. This doesn't have a monetary ROI, but long term it is better for the environment. I kinda need that saved up cash from returns from other stuff first though! 

 

[Temp]

42mm is massive. Not a hotel is it 🙂

 

We have some 28mm copper from a thermal store to a manifold between two bathrooms. From there we have 22mm plastic to high flow rate showers. Works well.

 

Everything else is 15mm plastic. Even the pipes  to UFH manifolds are only 22mm copper.

 

If you have any very long runs consider a secondary loop. However this needs to go right close to the tap. Even a 6ft branch from the loop to the tap can negate the value of having a loop. Loop must be well insulated.

 

 

[anonymous]

26 minutes ago, Temp said:

42mm is massive. Not a hotel is it 🙂

 

We have some 28mm copper from a thermal store to a manifold between two bathrooms. From there we have 22mm plastic to high flow rate showers. Works well.

 

Everything else is 15mm plastic. Even the pipes  to UFH manifolds are only 22mm copper.

 

If you have any very long runs consider a secondary loop. However this needs to go right close to the tap. Even a 6ft branch from the loop to the tap can negate the value of having a loop. Loop must be well insulated.

 

 

Not at all. Fairly low ceilings too. Not as low as new houses ofc but not exactly as high as a 1930s house. I can touch the ceiling on my tip toes, and I'm like 5 ft 7, so 42mm did seem a bit excessive to me 

 

 

[markocosic]

Pay £300 for somebody to do the math and advise? (heat geek will do it off visits or plans)

 

You'll not need larger than 35 mm on a unit < 16 kW. You'll not need larger than 28 mm on a unit <8 kW.

 

22 mm plastic is marginal. splitting 28 copper into 2x 22 mm plastic or 3x 22 mm plastic then running to a 22>15 plastic manifold arrangement fine.

 

Running a single 22 mm pipe for flow/return is unlikely to be ok unless <6kW. 

 

[ReedRichards]

I've never understood why heat pumps need such wide pipes and nobody here is saying.  You can achieve the same volume flow rate in a variety of pipe widths so for some reason it must be the speed of flow (metres per second) inside the pipes that is the important factor.  Perhaps the pipes inside the heat pump unit are wide and the pump isn't powerful enough to cope with a reduction in pipe width?  But if that's true then there's only one right answer, which is to match the internal pipe size, anything wider would just be spending more money than you need to.   

[dpmiller]

velocity= noise

[JohnMo]

1 hour ago, ReedRichards said:

I've never understood why heat pumps need such wide pipes and nobody here is saying.  You can achieve the same volume flow rate in a variety of pipe widths so for some reason it must be the speed of flow (metres per second) inside the pipes that is the important factor.  Perhaps the pipes inside the heat pump unit are wide and the pump isn't powerful enough to cope with a reduction in pipe width?  But if that's true then there's only one right answer, which is to match the internal pipe size, anything wider would just be spending more money than you need to.   

Generally there is a load of rubbish and generalisation spoken with respect plumbing and heat pumps.

 

For given house heat load a heat pump will be delivering lower flow temperature than a boiler for a radiator system, although this may not be true for UFH.  The lower flow temp requires more flow rate to give the kW output from the heating system.  Hence this generalisation you need big pipes.

 

But

A circulation pump will give a set output flow based on flow resistance; for a given pipe size say 22mm, the longer the pipe the higher the flow resistance for a set flow rate and if you increase flow rate for a given length the higher the resistance.

 

So to allow the circulation pump to run at the required flow rate could require bigger pipes.

 

But it all depends on the configuration of the pipes and lengths.

 

In each case the system pressure drops need to be calculated and suitable pipe sizes chosen.

[markocosic]

Heat pumps want dT 5C between supply and return.

 

Boilers are happy with dT 5-25C.

 

You need to move more water. It may be fine if pipes are oversized anyway. You may hit noise / pressure drop limits.

 

Valves are also problematic. Most resi sized trvs and lock shields are at their limits on larger rads. Will be noisy.

 

Attached was for a "7kW" unit. See how wide open all those TRVs need to be set for the larger rads there.

Knoll Pipe Calcs.xlsx

[ReedRichards]

10 hours ago, markocosic said:

Heat pumps want dT 5C between supply and return.

 

For good reason or just tradition?  dT=5 C and flow rate F is the same as dt=20 C and flow rate F/4 in terms of power supplied to the water.  Modern "Low Temperature" gas boilers operate in a similar temperature range to heat pumps but AFAIK stick to the larger dt and lower flow rates of a traditional gas boiler.  Clearly a smaller dT means the average water temperature is higher so you get more heat out of your heat sources but Low Temperature gas boilers seem to manage with a larger dT so why not heat pumps? 

[JohnMo]

8 minutes ago, ReedRichards said:

Low Temperature gas boilers seem to manage with a larger dT so why not heat pumps? 

My Atag boiler starts with a dT 20 at the highest flow temps, the dT reduced as flow temperature reduces and has dT of 4 when you get to 30 and below.

 

At low flow temps you cannot have a high dT, the return temperature will always be higher than the area you are heating.  So if the floor of UFH system is say 23 degrees, the flow temp is 30, the return temp can never be below 23, but is likely to be closer to 25 - dT 5 seems realistic if directly coupled to below pump flow rate would be modulated maintain dT4.

Edited April 2, 2023 by JohnMo
Missed context at end of final sentence

[dpmiller]

9 minutes ago, ReedRichards said:

For good reason or just tradition?

to keep the refrigerant circuit within efficiency limits

[markocosic]

16 minutes ago, ReedRichards said:

 

For good reason or just tradition? 

<snip>

Low Temperature gas boilers seem to manage with a larger dT so why not heat pumps? 

 

Good reason.

 

Heat pump efficiency is HEAVILY influenced by flow temperature.

 

Running 45/40C supply/return is significantly more efficient than say 47.5/37.5C or 52.5/32.5C. 

 

The benefits diminish must past a dT of 5C. Some systems run down to a dT of 3C (which even accounting for circulator energy can still result in a net performance improvement) but most settle on 5ish as a practical compromise. Some will go up to 8C or even 10C at peak load, but that's for overcoming lousy distribution systems not for efficiency.

 

CO2 is a different animal. That needs 60/30C to work. You won't be fitting one of those in a house though. You'll be fitting R290 (propane) or some legacy F-Gas stuff. They want a 5C dT.

 

 

[sharpener]

On 01/04/2023 at 07:38, ReedRichards said:

I've never understood why heat pumps need such wide pipes and nobody here is saying.  You can achieve the same volume flow rate in a variety of pipe widths so for some reason it must be the speed of flow (metres per second) inside the pipes that is the important factor.  Perhaps the pipes inside the heat pump unit are wide and the pump isn't powerful enough to cope with a reduction in pipe width?  But if that's true then there's only one right answer, which is to match the internal pipe size, anything wider would just be spending more money than you need to.   

 

This has puzzled me too, I guess the explanation that the refrigeration cycle requires it on efficiency grounds it the closest we will get.

 

The other constraints are the available head from the pump and a velocity limit of ~ 1 m/s to avoid flow noise in the pipework. For instance here is a result I got yesterday when considering a Grant 10kW - which has a min flow rate of 10 l/s and a max pump head of 6m i.e 600 mbar - to see what happens when only one TRV is open.

 

From this ready-reckoner the pump can push 10 l/s through 44 m of 15mm copper pipe (which is about the distance to my furthest rad) with a velocity of 1.2 m/s.

 

But with 15mm plastic pipe a head of 600mbar will only cope with a 15m run and the velocity is 1.6 m/s which is too high. The cross-section area is only reduced to 3/4 but the resistance is 3x as much because it is proportional to a higher power of the velocity.

 

So this HP should happily supply my furthest rad on its own, there is about 10m of plastic but OTOH at least some of the intervening pipework is 22mm. In any case I am planning to maintain the minimum flow rate by way of an automatic bypass to the return to the buffer tank.

 

[JamesPa]

2 hours ago, sharpener said:

This has puzzled me too, I guess the explanation that the refrigeration cycle requires it on efficiency grounds it the closest we will get.

I think the argument works like this: To minimise the flow temp (and thus maximize COP) heat pumps (and CH systems designed to work with heat pumps) are designed for a deltaT across the emitter of 5-7C rather than the 20C gas and oil boilers are historically designed for.  So the flow rate needs to be correspondingly higher to achieve that relatively low deltaT for the same heat transfer from emitter to room

 

Simples really.

[markocosic]

8 hours ago, sharpener said:

In any case I am planning to maintain the minimum flow rate by way of an automatic bypass to the return to the buffer tank.

 

Nope. It doesn't work like that.

 

The minimum flow rate is for the minimum turndown (minimum power) of the compressor. 

 

It really needs that flow rate AND a dT of 5C to function.

 

You'll also find a minimum circulating water volume (10-15 litres per kW output) is needed to avoid short cycling.

 

I would start reading up on how these things work before making too many detailed plans.

 

Or pay a pro £300 to do the calcs for you...

 

The only way that you heat you get one read on a heat punk is with a large 4-pipe buffer between the two. 30 litres per kW of minimum compressor output is about right to keep cycling acceptable.

 

In reality you just don't do this 

[ReedRichards]

13 hours ago, markocosic said:

Running 45/40C supply/return is significantly more efficient than say 47.5/37.5C or 52.5/32.5C. 

 

 

Whereas for a gas boiler the opposite is true, the lower the return water temperature the better the efficiency so for a gas boiler 52.5/32.5 is more efficient than 47.5/37.5 which is more efficient than 45/40.  Correct me if I'm wrong. 

[sharpener]

54 minutes ago, markocosic said:

 

You'll also find a minimum circulating water volume (10-15 litres per kW output) is needed to avoid short cycling.

 

I would start reading up on how these things work before making too many detailed plans.

 

 

The words of mine you have signally failed to understand are <the return to the buffer tank>. After reading extensively on the subject - and benefitting from others' advice here - my intention is to have a 210 litre buffer tank. This according to the Gledhill buffer store manual is suitable for HP outputs in the range 6kW - 17.5kW and is substantially larger than the minimum buffer size for the two HPs I am considering because it is also to act as a thermal store.

 

To mention two sources of advice in particular, the system diagrams here show several 3-port buffer tank arrangements which Heat Geek says is a good compromise between a good level of engagement on the one hand and unwanted mixing on the other.

 

54 minutes ago, markocosic said:

 

The only way that you heat you get one read on a heat punk is with a large 4-pipe buffer between the two. 30 litres per kW of minimum compressor output is about right to keep cycling acceptable.

 

 

No idea what you mean, why don't you proof read your postings?

[dpmiller]

A heat pump with it's circulating pump under PWM will control flow rate to assist in setting the deltaT. It's under maximum heat demand that things get sticky as pipe restrictions might be fixed but flow though the UFH mixers is not. I'm using an auto bypass to help out in this situation.

[JohnMo]

8 hours ago, ReedRichards said:

the lower the return water temperature the better the efficiency so for a gas boiler 52.5/32.5 is more efficient than 47.5/37.5 which is more efficient than 45/40

Yes the lower the return the higher the efficiency.  But you do not control the delta T of a modern modulating boiler the boiler controller does.  It changes pump speed to have the delta T it wants to see for a given temperature. 

[ReedRichards]

If the boiler sets its own delta T it must be able to read the room temperature.  You would have difficulty matching the radiator heat output to the heating demand if delta T is not fixed so the boiler must compensate for that..