View on Low Loss Header for domestic property

[Zencap]

We are experiencing problems with our heating system which has become apparent after adding a single zone UFH to a large kitchen area.

When the UFH kicks in, the flow to the CH Rads cools off. When the UFH is not on the Rads (13 in total) all get hot (quite slowly downstairs.

UFH has own pump and CH pump is brand new 15/60 and made no difference. System was powerlfushed in last year and boiler is under a year old Viesmann-100w 26kw. When UFH was done, new piping was done to all 4 downstairs rads.

2 plumbers are stumped by the issue as set up should work. Proposals are

Change Flow and Return pipes (which are hidden in walls) from boiler to air cupboard to 28mm.

Install Low Loss Header - a destratification pump to a low loss header (LLH). Before the LLH run a 22mm f & r to the dhw cylinder. Downstream of the LLH install a primary pump to run the CH circuit and UFH.

Pressurise the system

what are peoples views on this - any other ideas?

concern on LLH is this over kill and cause issues long term?

pressurising system is risky given upstairs pipes are older and could cause issue.

help greatly received

thanks

 

[ShaunCorbs]

Is the ufh on it’s own zone eg direct from the boiler

 

[Zencap]

Is the ufh on it’s own zone eg direct from the boiler

Forgive my poor knowledge, but I think no, flow from boilers goes up to airing cupboard and S plan (I think) with UFH and Rad flows straight after pump but UFH is first drop on pipes down so seems to be take that route first when that is called, and starving CH rads of flow. Have tried restricting flow to UFH but doesn’t make material diff.

 

[exedon2]

Picture of set up in airing cupboard would be useful
13 rads in truth is a bit meaningless what's there output? You could find it out by measuring them and looking at a rad size list .
Also output of underfloor?

 

[John.g]

One might think that boiler flow rate should not be a problem with UFH as its normally only ~ 1/3rd of the UFH flow, a UFH heat demand of say 8kw might require a circ flow rate of ~ 15 LPM but the boiler circ pump portion of this might only be 3 to 5 LPM, hardy enough to trouble any boiler circ pump when combined with any existing rad demand., maybe new CH pump mode/setting problem or boiler output not big enough?

 

[Zencap]

One might think that boiler flow rate should not be a problem with UFH as its normally only ~ 1/3rd of the UFH flow, a UFH heat demand of say 8kw might require a circ flow rate of ~ 15 LPM but the boiler circ pump portion of this might only be 3 to 5 LPM, hardy enough to trouble any boiler circ pump when combined with any existing rad demand., maybe new CH pump mode/setting problem or boiler output not big enough?

Agree, the CH pump has been changed to a new one and same result. Boiler output considered high enough.
Logic is the flow return is not happening effectively - so either the return pipe is restricting flow somehow or inefficiently set, this proposal to replace with new 28mm minimum but also suggestion to add the low loss header - just seems excessive??

 

[John.g]

Any idea of the UFH demand or/and better still the boiler CH temp, the manifold flowrate from the flow meters and the manifold flow and return temps?. Is the new CH pump external to the boiler?.

 

[Zencap]

Any idea of the UFH demand or/and better still the boiler CH temp, the manifold flowrate from the flow meters and the manifold flow and return temps?. Is the new CH pump external to the boiler?.

The new CH is external in airing cupboard.

We found the differential temperature’s between the flow and return were so far apart that the boiler kept cycling when on service mode. When the boiler runs in its normal mode the boiler plateaus at 64 degrees on the flow and returns at 55 degrees.

indont have flow rates

.

 

[ShaunCorbs]

From the pump

rads
Ufh
Hot water ?

 

[Zencap]

Picture of set up in airing cupboard would be useful
13 rads in truth is a bit meaningless what's there output? You could find it out by measuring them and looking at a rad size list .
Also output of underfloor?

 

[Zencap]

From the pump

rads
Ufh
Hot water ?

Correct - closest to pump is UFH, rads and hot water at top

 

[ShaunCorbs]

You got a pic of your ufh manifold?

 

[John.g]

The new CH is external in airing cupboard.

We found the differential temperature’s between the flow and return were so far apart that the boiler kept cycling when on service mode. When the boiler runs in its normal mode the boiler plateaus at 64 degrees on the flow and returns at 55 degrees.

indont have flow rates

.View attachment 47057

When you say normal mode do you mean on rads only? and getting cycling/high temp diff on UFH only? what's service mode ?

If on UFH only then the the boiler should only cycle for two reasons, either the UFH heating demand is lower than the minimum boiler output or some boiler protection is switching it off on this unavoidable high temperature differential which hardly warrents a LLH, a ABV should do.

 

[Zencap]

You got a pic of your ufh manifold?

 

[ShaunCorbs]

Are the flow guages set correctly with the plans spec ?

 

[Zencap]

Are the flow guages set correctly with the plans spec ?

Yea UFH has all been checked and flow adjusted, and in fact restricted right down to see if makes a difference. Is all very strange.

has anyone come across Low loss header in domestic setting?

would changing the flow and return between boiler and air cupboard pump to 28mm make a big difference?

 

[ShaunCorbs]

Yea UFH has all been checked and flow adjusted, and in fact restricted right down to see if makes a difference. Is all very strange.

has anyone come across Low loss header in domestic setting?

would changing the flow and return between boiler and air cupboard pump to 28mm make a big difference?

also I forgot to add I guess it’s all one room (the ufh is in)

 

[Zencap]

also I forgot to add I guess it’s all one room (the ufh is in)

Yes all one room and single zone so not complex and effectely just two new pipes added to existing heating system..

 

[ShaunCorbs]

On the plans does it say the kw load ?

 

[EvilDrPorkChop]

Is this Open Vented or Sealed?

And where is the return Tee'd in for the UFH? In the airing cupboard?

 

[exedon2]

On the plans does it say the kw load ?

Or the floor area in sq mts would do

 

[John.g]

Yes all one room and single zone so not complex and effectely just two new pipes added to existing heating system..

Its not quite clear (to me anyhow) if the very high boiler temperature differential problem is only happening on UFH duty only, it may be worth just turning up the cylinder stat, using the coil as a by pass to see if the boiler then remains on when on UFH "only". Also perhaps just turn down the boiler stat temporarily to the same, or below the setting as the UFH mixing temp which should increase the flow demand to the same as any heating circuit.

 

[exedon2]

Is this Open Vented or Sealed?

And where is the return Tee'd in for the UFH? In the airing cupboard?

First Post says its open vented if you look at picture of airing cupboard it looks to me that pump is connected to flow before vent and cold feed not ideal!
Looks like f&r from underfoot connect in airing cupboard at low level

 

[EvilDrPorkChop]

First Post says its open vented if you look at picture of airing cupboard it looks to me that pump is connected to flow before vent and cold feed not ideal!
Looks like f&r from underfoot connect in airing cupboard at low level

Yes thought so, was just checking as it should have been made sealed with the UFH.

Do see how any UFH runs efficiently or correctly when a system is OV, OP system needs sealing first.

 

[John.g]

Maybe not ideal, but easy to check F&E tank for circulation.

 

[Brambles]

Zencap

If we get back to the basics of your question. Can you use a low loss header on a domestic system? - Yes. However, it will be installed close to the boiler ( you need a reasonable amount of space) require all the out going circuits to be separately pumped and probably reasonably extensive pipework modifications. Ball park cost of the header, pumps and materials £1k ish plus installation ( but you know that because they will have given you a price) Dont have it installed by someone who has not done it before - feeding DHW and UFH with the same pump from a hydraulic separator will give you similar issues.

If we look at your UFH set up, you have multiple loops, none of which appear to be actuator controlled ( so why do you have an expensive controller to operate a zone valve?) in essence your UFH is either all on or all off. If you can give the floor area in m2 that it serves we can guesstimate the power consumption - it could be around 15kw - unless you have a number of very short loops. In simple terms that could be more than 50% of your net boiler output

A zone of that size should be piped directly back to the boiler - on both flow and return with correctly sized pipes.

My assumption is that the system has been installed ( plumbed in) as if it it just another radiator within the rad circuit. I often see that in DIY heating systems when a conservatory has been built with ufh. They encounter exactly the same starvation problems / symptoms that you describe. However that is normally on 1 or 2 100m loops - you have seven loops. If they have been installed properly the loop length will be marked on each circuit. It is very important to know the loop length and the floor m2 that they serve.

In any heating system the hot flow takes the path of least resistance - if you offer it a choice of radiators or a route through say 400m of 16mm ufh pipe - it will take the latter and only serve the radiators when the UFH is up to temperature.

You need a decent installer who is very familiar with the design and integration of such systems to redesign the system for you. The solution may or may not involve hydraulic separation, but in my experience it is unlikely not to require the ufh feeding and returning directly and exclusively to the boiler.

With respect to increasing the feed pipes from the boiler to 28mm, that would probably only be effective if they them terminated in a remote low loss header - feeding three separate circuits : DHW, Rads and UFH ( so three additional pumps required plus a LLH)

I have not addressed the sealed system comments, but it is a lot easier to configure a UFH into a sealed rather than vented system.

Apologies, but there is unlikely to be a quick cheap fix to your dilemma

 

[John.g]

Not familiar with UFH but would have thought that a very high boiler deltaT might be the big problem for gas fired boilers, I wouldn't have thought of a UFH system robbing the boiler flow would be a big problem as it is only supplying a thermostatic mixing valve and the manifold pump is doing all the UFH pipework circulation.

 

[Brambles]

John, nothing at all wrong with your logic.

The problem with starvation ( rather like gravity flow) is that the differnce between it working ok or not at all is a very fine margin.

In this case, in my view it is very unlikely that a seven loop system will perform in harmony with a radiator circuit unless it is correctly sized and piped back to the boiler. Unless the radiator circuit is predominantly on 22mm - the resistance to flow will be significantly higher than 16mm UFH.

Your comments on the injector / mixing valve are very valid - but bear in mind that 7 loops is a large system, that from cold could take 3 days to get up to temperature before flicking on and off.

The clue with this system is probably the controller with no loop actuators to operate - so it is unlikely that the installer who thought that was needed ( lots of terminals and no wires to connect them to!) even looked at the hydraulic aspects of the system.

 

[John.g]

Thanks for that, one/two other queries, when setting the flow rates on each loop, is this done/tuned with the whole system up to temperature? and is there any/much difference in those flow rates between a hot and cold system?.

 

[Brambles]

No, you balance the flow rates against the loop length when you commission the system. That is why knowing the loop length is important ( normally between 80 and 110m). At a 100m of 16 mm dia the flow rate shoukd be around 2.5 lpm to deliver 40 degrees C.

So 7 simultaneously controlled loops is not an insignificant on / off load

 

[John.g]

Is there anyway I can attach a file, I have a spreadsheet here with some Boiler/UFH numbers that I put together that some may find interesting.
I could attach excel files previously by zipping them but cannot do so for the past few weeks.

 

[king of pipes]

Zencap

If we get back to the basics of your question. Can you use a low loss header on a domestic system? - Yes. However, it will be installed close to the boiler ( you need a reasonable amount of space) require all the out going circuits to be separately pumped and probably reasonably extensive pipework modifications. Ball park cost of the header, pumps and materials £1k ish plus installation ( but you know that because they will have given you a price) Dont have it installed by someone who has not done it before - feeding DHW and UFH with the same pump from a hydraulic separator will give you similar issues.

If we look at your UFH set up, you have multiple loops, none of which appear to be actuator controlled ( so why do you have an expensive controller to operate a zone valve?) in essence your UFH is either all on or all off. If you can give the floor area in m2 that it serves we can guesstimate the power consumption - it could be around 15kw - unless you have a number of very short loops. In simple terms that could be more than 50% of your net boiler output

A zone of that size should be piped directly back to the boiler - on both flow and return with correctly sized pipes.

My assumption is that the system has been installed ( plumbed in) as if it it just another radiator within the rad circuit. I often see that in DIY heating systems when a conservatory has been built with ufh. They encounter exactly the same starvation problems / symptoms that you describe. However that is normally on 1 or 2 100m loops - you have seven loops. If they have been installed properly the loop length will be marked on each circuit. It is very important to know the loop length and the floor m2 that they serve.

In any heating system the hot flow takes the path of least resistance - if you offer it a choice of radiators or a route through say 400m of 16mm ufh pipe - it will take the latter and only serve the radiators when the UFH is up to temperature.

You need a decent installer who is very familiar with the design and integration of such systems to redesign the system for you. The solution may or may not involve hydraulic separation, but in my experience it is unlikely not to require the ufh feeding and returning directly and exclusively to the boiler.

With respect to increasing the feed pipes from the boiler to 28mm, that would probably only be effective if they them terminated in a remote low loss header - feeding three separate circuits : DHW, Rads and UFH ( so three additional pumps required plus a LLH)

I have not addressed the sealed system comments, but it is a lot easier to configure a UFH into a sealed rather than vented system.

Apologies, but there is unlikely to be a quick cheap fix to your dilemma

A detailed explanation Brambles

 

[John.g]

Son in law took a few measurements from his system yesterday morning for me and here are the calculated results of flow/temperatures etc, he has a sealed system with the UFH added around 6 years ago, he has no noticeable problems with rads and UFH on together, the manifold is partitioned off so he can't take any pictures, not sure but thinks that the flow and return to the UFH were just teed off the existing pipework. He also stated that the installer told him that the TMV was set to 45C.

UFH Heat Demand 10 KW
UFH Mixing (Flow) Temp 45 Deg.C
UFH Return Temp 37 Deg.C
Boiler Flow Temp 65 Deg.C

UFH Delta T 8 Deg.C
UFH Flow 17.92 LPM
Boiler Flow 5.12 LPM
Boiler DeltaT 28 Deg.C
Boiler Return Temp 37 Deg.C

 

[king of pipes]

I am going to a job tomorrow to quote a boiler upgrade it has ufh downstairs rads upstairs 20 years old approximately I will post a few pictures of the setup I have worked on other houses in the close and its pretty basic but seems to work and it's open vent. Kop

 

[king of pipes]

This one is all piped in 22mm and not much control only the one stat in the hall going to recommend a complete upgrade