CCTs, separation, bypass, and all that guff

[HughF]

I’m adding a ufh mixer to my rad system. If I use Close Coupled Tees (ccts) just before the input the manifold, do I need to add a lock shield valve in the middle of the t’s so that I don’t make a short circuit and then end up with no flow through my rads?

 

My head says yes….

[Lofty718]

You shouldn't use CCT's at the manifold. Seperation should be under the boiler not at the manifold

[HughF]

Caleffi make a manifold that has exactly that, separation at the manifold….

 

why shouldn’t I put separation at the manifold? Surely it doesn’t matter where it is? And the boiler (ashp actually) is at the bottom of the garden so that’s not particularly practical.

[Lofty718]

 

[HughF]

Yep, I’ve seen that. Can’t see a way around providing hydraulic separation between the two pumps without it though….

 

I don’t have any zone valves, just a stat to turn the ufh pump on/off. Rads are on all the time on a time clock with weather compensation.

Edited October 29, 2023 by HughF

[Lofty718]

Can you not run the UFH on pure weather compensation rather than using a stat? it doesn't sound like good design

 

 

Ideally the UFH and rads should all run off one pump at the same flow temp

Edited October 29, 2023 by Lofty718

[HughF]

17 minutes ago, Lofty718 said:

Can you not run the UFH on pure weather compensation rather than using a stat? it doesn't sound like good design

 

 

Ideally the UFH and rads should all run off one pump at the same flow temp

Rads are sized for 45 deg at -4, can’t run the ufh at that temp so I need a blending valve/mixing station.

 

Ideally they should, yes…. But we’re dealing with what we’ve got and have to make it work. Or we put rads in the extension and forget about the underfloor, I don’t really want to do that.

Edited October 29, 2023 by HughF

[Lofty718]

The best way to control UFH is with an electronic mixing valve rather than one on the manifold, they are crap

 

Look into esbe valves, I have one controlling my UFH.

 

 

 

[HughF]

I’m using an ivar unimix 3, the rolls Royce of blending valves. @Nickfromwalesdo you have anything to add?

[Lofty718]

Is the ivar unimix an electronic mixing valve? I can't find much info on it but looks to be a standard mixing valve

[HughF]

13 minutes ago, Lofty718 said:

Is the ivar unimix an electronic mixing valve? I can't find much info on it but looks to be a standard mixing valve

Yep, standard blending valve…. But it’s supposed to be a really good one

[Lofty718]

Using a high temperature UFH blending valve with a heat pump is insane if you can avoid it.

 

@JohnMo what do you think?

 

 

[HughF]

The unimix 3 is one of the few valves that will mix down to 20 degrees or so, it’s a good valve, or so I’ve been told.

 

Plenty of people still say you should have a blending valve with a heat pump, for safety. I know @Nickfromwales is an advocate of this strategy.

 

The situation is what it is, I’ve got rads in the rest of the house sized for 45 flow at -4, and I can’t feed that temperature into the ufh…

[JohnMo]

5 hours ago, HughF said:

If I use ccts just before the input the manifold, do I need to add a lock shield valve in the middle of the t’s so that I don’t make a short circuit and then end up with no flow through my rads?

The tees will be at least one size bigger than the pipe they are installed in, the drop in velocity will create a nul point in the system. The UFH pump will only draw enough water through the CCT to match what is required to satisfy the UFH flow rate - possibly less then 3L/min. out of 20 plus.

 

 

5 hours ago, Lofty718 said:

You shouldn't use CCT's at the manifold. Seperation should be under the boiler not at the manifold

 

Not sure that makes sense, a CCT can be anywhere in a circuit, its a pressure modifier due to the step change in piping cross section. Its function doesn't care where its in the system

 

4 hours ago, HughF said:

Yep, I’ve seen that. Can’t see a way around providing hydraulic separation between the two pumps without it though….

You do not do a CCT like the video shows it, he has the primary loop and secondary loops mixed up, assuming he is showing the boiler at the top. The boiler is the primary loop the secondary loop tees out and back in again, the low velocity zone means the primary flow doesn't notice fluid coming and going.

 

I would just plumb as detailed on the Bosch Tech bulletin (attached). A CCT with an additional 3 port valve to give the return back to the heat pump if only the UFH is on. But not sure if only 2 loops of UFH will be practical on it own without lots of short cycling and the UFH will have a very different time delay heating/cooling than the rest of the system.  As the rest of the system will be on long and slow, I would be tempted to do the following.

 

Batch heat the floor (storage heater mode), running a slightly elevated temperature, so it only needs to run 4 to 5 hours a day (when the radiators are on), and use the thermostat for that room to switch on/off the UFH pump. If you do that you would need a zone valve at all.

 

4 hours ago, HughF said:

I’m using an ivar unimix 3

 

I used the Ivar manifold mixer for last years heating season (no longer installed as I run direct from the heat pump, with no radiators), they work well but be prepared to do some adjustments and fine tuning to get the best out of them.  

TB_0137_b_-_Hydraulic_separation_of_UFH_via_closely_spaced_tees (1).pdf

[HughF]

26 minutes ago, JohnMo said:

The tees will be at least one size bigger than the pipe they are installed in, the drop in velocity will create a nul point in the system. The UFH pump will only draw enough water through the CCT to match what is required to satisfy the UFH flow rate - possibly less then 3L/min. out of 20 plus.

 

 

 

Not sure that makes sense, a CCT can be anywhere in a circuit, its a pressure modifier due to the step change in piping cross section. Its function doesn't care where its in the system

 

You do not do a CCT like the video shows it, he has the primary loop and secondary loops mixed up, assuming he is showing the boiler at the top. The boiler is the primary loop the secondary loop tees out and back in again, the low velocity zone means the primary flow doesn't notice fluid coming and going.

 

I would just plumb as detailed on the Bosch Tech bulletin (attached). A CCT with an additional 3 port valve to give the return back to the heat pump if only the UFH is on. But not sure if only 2 loops of UFH will be practical on it own without lots of short cycling and the UFH will have a very different time delay heating/cooling than the rest of the system.  As the rest of the system will be on long and slow, I would be tempted to do the following.

 

Batch heat the floor (storage heater mode), running a slightly elevated temperature, so it only needs to run 4 to 5 hours a day (when the radiators are on), and use the thermostat for that room to switch on/off the UFH pump. If you do that you would need a zone valve at all.

 

 

I used the Ivar manifold mixer for last years heating season (no longer installed as I run direct from the heat pump, with no radiators), they work well but be prepared to do some adjustments and fine tuning to get the best out of them.  

TB_0137_b_-_Hydraulic_separation_of_UFH_via_closely_spaced_tees (1).pdf 114.59 kB · 0 downloads

Thanks for the input…. The rads are on from 0600-2300 at the moment, and that might change to 24hrs a day with the weather comp bumped down a bit as we head into winter.

 

I wasn’t intending any zone valves at all, just use a floor sensing stat to switch the pump on/off.

 

It was more the hydraulic separation issue I wanted to get clarified. I’ve already got short circuiting issues through my rads (see my other thread) and didn’t want to exacerbate the issue with a close coupled T arrangement between flow/return. Hence why I assumed it would need some resistance in the separation.

[JohnMo]

22 minutes ago, HughF said:

didn’t want to exacerbate the issue with a close coupled T arrangement between flow/return. Hence why I assumed it would need some resistance in the separation

A cct doesn't go to the flow and return, both tees are in the flow line next to each other. Your just doing what's on the Bosch sheet without any zone valves (2 way or 3 way) to complicate, your UFH will never run on its own, as it's just to small for the ASHP.

 

Just add two tees in the pipe going to radiators one pipe size bigger, than the pipe. The flow going to the UFH will come from the upstream tee, the flow going back to the system will go into the downstream tee.  This will cool the radiator flow temp slightly but not much as the flow rate being taken by the UFH is so low.

[HughF]

I’ve brought the flow and return down from the ceiling void to where my mixer is. So I would put the Ts in the flow line, come off, through the mixer then back again to the flow.

 

Joining the flow and return together after the Ts to complete the circulation. 
 

What stops the now joined flow and return acting as a short circuit and my rads all staying cold?

[JohnMo]

The return line isn't involved, if you connect it in to there it will just short circuit, not what you want.

 

The flow comes down, add the cct's and the flow continues to the radiators. The return comes from the radiators back to the ASHP.

 

Basically just do it like this (below) - you just borrowing small amounts of water from the primary flow and giving back a little cooler.

[HughF]

Ok, gotcha. I’ll need to drain the system down then to add those t’s before the feed to the last rads. I’ve got isolation valves on the flow/return after the last of the rads at the moment, so I could plumb this all up with the rest of the system live, but the installation of the Ts will require a drain down.
 

Any benefit in T’ing into the return leg instead? Might be able to do away with the mixer if I can get a low enough temperature.

[JohnMo]

Could just as easily go in the return line, would be cooler flow. 

[Lofty718]

@JohnMo CCT's on my boiler are installed like in the video.

[HughF]

Just realised that where I have easy access to the flow/return in the ceiling void, before the last rads come off, those rads have TRVs on as they are the upstairs bedrooms...

 

If the TRVs shut down the bedroom rads, the flow will stop, the UFH will be drawing on a pipe that has gone cold, as there's no flow.

 

I'll need to try and come off before the drops for the non-TRVd rads... 

[dpmiller]

auto bypass valve across the rads?

[HughF]

Pipe in the corner is the flow, one out from the corner is return. I’m thinking close coupled Ts on the flow, loop flow and return with a gate valve, adjust to get delta-t 5 at this point. No different to if there was a rad on the end.

 

 

[HughF]

Ceiling being boarded tomorrow by the builders, and I don’t have a chance to drain down, so it’s plan B…. Lockshield gate valve between flow and return, CCTs on the flow line to give a couple of 22mm runs to the manifold.