UFH manifold problem

[JohnMo]

That makes sense

[Nickfromwales]

1 hour ago, joe90 said:

My heat pump only heats to one temp and that  I set at 45’ to cope with DHW. The buffer stat is set at 35’ (which calls fir heat from the ASHP), and the manifold blends down to 30’.

That adds some clarity.

Why have you done that? DHW running at 45oC and heating at 35oC would give you a better CoP, no?

1 hour ago, JohnMo said:

That is sort of what I was saying, could you turn down the temp, run at lower temp and get a better COP? Run all the time while there is a demand for heat from the house?

That's how I set the most recent up, with just a LLH ( low loss header ). HP is just idling most of the time and has a nice easy life vs kicking in / out t higher temps. Gives a much better operational hysteresis IMO. Was a 7kw HP servicing a ~3.5kw demand at worst case. Ran near silently and I had to keep going up to it to check it was still 'lit and burning' by putting my head in front of the fan. The minimal road noise and surrounding noise pollution ( urban setting ) was enough to nigh on cancel the HP out altogether. I would not have wanted that running harder to heat a buffer and kicking in / out / running louder etc for this instance, and I could not see a notable ( short ) cycle which would have made me do otherwise. Each instance is unique, and this was a decent 'passsive raft' slab with a huge thermal time constant, so may have been different for UFH in screed over block and beam etc.

@joe90 would you not consider getting the heating output temp down to see if that boosted CoP? With the buffer charge "not lasting long" I doubt that is doing much to stave off short cycling anyhoo?

[joe90]

1 hour ago, Nickfromwales said:

Why have you done that? DHW running at 45oC and heating at 35oC would give you a better CoP, no?

My ASHP only has one temp setting, it does not have separate DHW and heating temps.

1 hour ago, Nickfromwales said:

would you not consider getting the heating output temp down to see if that boosted CoP? With the buffer charge "not lasting long" I doubt that is doing much to stave off short cycling anyhoo?

As above only one temp setting available and any lower and DHW temps suffer. I have no idea what my COP is! My recent set up seems ok and no short cycling observed in fact in the short time since setting it up the ASHP appears to be ticking over when the heating is running so keeping up with demand rather than on off.

[Nickfromwales]

1 minute ago, joe90 said:

My ASHP only has one temp setting, it does not have separate DHW and heating temps.

Ah! OK, gotcha. ?. 
As is then seems the way. Had loads of issues when going to already installed ( by others ) UFH systems, mostly with the TMV not performing properly.

Glad you got a result at last. 

[Adam2]

4 hours ago, PeterW said:

Please go and beat your plumber to a soggy death with his blowtorch.  You can ask a lot of people on here about the difference and why you need pumps at every manifold along with a blending valve. The buffer will soon deplete and you won’t have 35°C water in it so you need to heat slightly higher and then let the valves do the work. 

To be clear pump is at each manifold. 

 

Understand the point re buffer temp needing (generally) to be higher than target flow temp but you can still do this with blending at the buffer ?

 

The issue I suppose could be that with the manifolds serving floors greatly differing in size  maybe having higher temp on larger floors would be better?

 

Also keen not to have similar issue to @joe90 as have a similar setup though not with single temp issue

[joe90]

1 minute ago, Adam2 said:

The issue I suppose could be that with the manifolds serving floors greatly differing in size  maybe having higher temp on larger floors would be better?

Would not adjusting flow rates at each loop control this?

[Adam2]

1 hour ago, joe90 said:

Would not adjusting flow rates at each loop control this?

Yes probably will be OK I guess with the auto balance actuators

[joe90]

7 minutes ago, Adam2 said:

Yes probably will be OK I guess with the auto balance actuators

I have no actuators, one zone and 5 loops, just increased flow in the lounge that we want a tad warmer .

Edited January 22, 2022 by joe90

[PeterW]

Ok so the buffer will never be at a static 36°C  as it will get cold water from every return so your idea of the feed being always at a constant temperature won’t work. This is why you need the blending valve. 

[Nickfromwales]

22 hours ago, Adam2 said:

The issue I suppose could be that with the manifolds serving floors greatly differing in size  maybe having higher temp on larger floors would be better?

Erm, no. 
The bigger rooms will already have more pipe, ergo more heat. The UFH design should mitigate through heat loss calcs for each heated space, accordingly.
If there’s any doubt, just wind the pipe around at 100mm centres and simply create more volume of heated water per m2, but at the same temp. More loops = more heated area and ( IMO ) looking to heat other areas more by opening up flow is coarse on a good day. 

[Adam2]

OK @Nickfromwales thanks for clarifying that then.

 

So given that there is generally no need to adjust flow temp per floor what is the benefit in having a blending valve to recirculate the return directly at the manifold? If I just take all returns back into the buffer and install one blending valve at the buffer I'll be pushing the same temp to all manifolds. I appreciate some small heat losses in the returns going back to the manifold but if well insulated that should have be minimal.

 

For clarity - still keeping the pump at each manifold and possibly one at buffer if also beneficial as UFH on top floor is 2 floors above the buffer tank position.

 

[Nickfromwales]

8 minutes ago, Adam2 said:

OK @Nickfromwales thanks for clarifying that then.

 

So given that there is generally no need to adjust flow temp per floor what is the benefit in having a blending valve to recirculate the return directly at the manifold? If I just take all returns back into the buffer and install one blending valve at the buffer I'll be pushing the same temp to all manifolds. I appreciate some small heat losses in the returns going back to the manifold but if well insulated that should have be minimal.

 

For clarity - still keeping the pump at each manifold and possibly one at buffer if also beneficial as UFH on top floor is 2 floors above the buffer tank position.

 

Adjusting flow temps per floor is usually dictated by which types of emitters you have, so there is no hard and fast rule there. If screed / slab ground floor, and aluminium spreader plates 1st floor, you 100% need higher flow rates for the spreader plates as the losses from what is a much poorer emitter need to be mitigated against...by increasing the flow temp. I have done many like this, and, for eg, have had to run ~30oc ground and as high as 45oC 1st. Screed over block and beam on both floors may permit a unified flow temp, but as the heat rises the 1st floor in that scenario would likely need lower temp flow compared to the ground  

 

Water is being pumped around the UFH loops and back on themselves continuously at the rate that is set by the manifold pump and the calibrated flow rates per loop. Then you have the primary pump delivering heated water to /from the manifold, therefore a 2 storey dwelling with UFH on both floors would typically be utilising 3 or 4 pumps, dependant on the location of the TMV's. These pumps will need to be operating reliably at different potentials, so the basic requirement here will be for hydraulic separation of these pumps / circuits.

 

The primary pump(s) will typically be single speed and flowing to / from the manifold(s) relieved by the calibrated or differential ( automatic ) bypass where they start the return path, but the secondary side ( manifold pumped loops ) will be constantly changing in potential, eg with fluctuations caused by changes in demand per loops / zones / by the actuators opening an closing. Also, with this approach and subsequent hydraulic separation, the flow rates through the loops cannot become influenced ( over whelmed ) by both pumps running directly in series ( the pumps never 'see' each other.

 

Also good to note that pumps need to 'suck' water through TMV's, so you cannot have one remote TMV and then tee off to 2x pumps to give flow to 2x manifolds. You would need a 6 tapping buffer with 4 of those dedicated to 2x flow and 2x return for the 2 heating zone TMV's. Just a lot of extra fuss, confusion and all can be avoided by just blending and pumping at each manifold, like it's always been been done for a VERY long time. Also, if you don't want convection heat flow to the upper floor, remember that a 2 port zone valve ( my preference ) or a single check non return aka anti-gravity valve will be required. 
 

Next; to blend ( control ) water to the loops accurately, there needs to be a guaranteed supply of water that is cooler ( TMV manufacturer will give specifics for the required % of differential necessary to promote effective operation ) than the heat source input flow temp ( TMV ) setting. That is not going to be so if you are attempting to feed back / group the returns via to the buffer as you will be causing stratification ( multiple circuits flowing in and out for feeds, plus the same from the flow and return of the heat source ) and removing that guaranteed cold(er) water 'supply'. Water flow will take the path of least resistance, so the returns will just pull that mixed up body of water and add it to the return that goes into the remote TMV at the buffer. That is why the cooled returns from each loop are usually immediately grabbed by the local TMV at the manifold for use for blending where it needs to happen.

 

It's only Nu-Heat which seem to like this option, plus a few renewable installers where TS's have been installed, and I've not seen anything resembling good results from that from type of arrangement with what I've seen ( on live installations ) to date. With UFH installs of mine coming up for 20+ years of reliable operation, I shall not be changing course with my ( standard?) method any time soon.

 

PS. For anyone building to passive / very high spec, the type of TMV required for flow temp control if sub 30oC has a capillary wire from a thermo-probe which controls a TRV like haed on the manifold blender. If you move away from this you'll lose that type / accurate type of control too. Ivar is my weapon of choice, and you can see the setup with the capillary wire / probe / TRV  here

[Bramco]

1 hour ago, Nickfromwales said:

PS. For anyone building to passive / very high spec

 

So presumably, those of us building with slab and TF from say MBC would need this type of manifold?

 

Simon

[JohnMo]

We used a normal manifold, spaced UFH pipes at 300mm centres and have the temperature for hot flow at 35 degrees.

 

Easy - way less pipe in floor, comfortable, bog standard bits.

[Nickfromwales]

5 hours ago, Bramco said:

 

So presumably, those of us building with slab and TF from say MBC would need this type of manifold?

 

Simon

Yup. Those are what I’ve installed for every one of my high spec / “passive” projects. Regular TMVs are nigh-on fully shut at their lowest setting, and you can hear them ‘whining’ too which is a further annoyance.

 

3 hours ago, JohnMo said:

We used a normal manifold, spaced UFH pipes at 300mm centres and have the temperature for hot flow at 35 degrees.

 

Easy - way less pipe in floor, comfortable, bog standard bits.

Way less pipe = way less water volume so not great at all ?. Anything less ( further apart ) than 150mm centres is also not recommended when utilising an ASHP as the primary heat source, and this is even more important in PH settings.

Less system volume means bigger buffer or loads of short cycling. Much longer times to raise temps too. 
Current project all in at 100mm centres, apart from kitchen and utility at 150 as they are ‘inboard’. 

[JohnMo]

Using thermal store as buffer, heating it to 40 degs, no issues with short cycling. Houses stays +/- 0.5 degs from thermostat setpoint.

 

So happy enough.

[Nickfromwales]

1 hour ago, JohnMo said:

Using thermal store as buffer, heating it to 40 degs, no issues with short cycling. Houses stays +/- 0.5 degs from thermostat setpoint.

 

So happy enough.

Agreed

Just my option removes the TS and uses the existing footprint ( slab ) for buffering so that is hidden in the floor and consumes no additional internal footprint.

[JohnMo]

My build was always going to have a cylinder, to use excess PV instead of export in the summer.  So no loss of space in my case.  But would not suit all builds.

[joe90]

7 minutes ago, JohnMo said:

My build was always going to have a cylinder, to use excess PV instead of export in the summer.

But would that not be the DHW tank? Heating not required in summer?

[JohnMo]

50 minutes ago, joe90 said:

But would that not be the DHW tank? Heating not required in summer?

Have got a gas combi boiler.  The thermal store DHW coil is in use all the time and has a temperature control valve prior to combi boiler.  In winter I get slightly preheated water in to the combi, to up performance, to as good or better than summer performance.

 

In summer, PV should heat water in thermal store.  This removes or reduces the need for gas for DHW production.  If the water flowing through the DHW coil is at or above the combi boiler DHW setpoint it flow straight through the boiler without additional heat being added.  Boiler is designed to take preheated water.

[joe90]

Ah, that makes sense.?

[Bramco]

On 24/01/2022 at 15:52, Nickfromwales said:

Regular TMVs

 

I'm assuming it means a drain down if you want to swap at a later date?

 

Simon

Edited January 25, 2022 by Bramco

[Adam2]

Thanks @Nickfromwales for the detailed reply re the use/position of blending valves

[Nickfromwales]

15 hours ago, Adam2 said:

Thanks @Nickfromwales for the detailed reply re the use/position of blending valves

My thumb still hurts. 

[Nickfromwales]

20 hours ago, Bramco said:

 

I'm assuming it means a drain down if you want to swap at a later date?

 

Simon

I install gate valves at regular / strategic points in my plant room installs, particularly where there’s an ASHP and antifreeze, so when needing to do routine maintenance or repairs, my clients only have to drain the bare minimum of primary water off. ?