Re-designing my heating system

[jayc89]

Current setup;

• 32kw Baxi Platinum+ System Boiler, with weather comp, fitted upstairs, believe max flow rate is 20 LPM.
• 250L UVC (same cupboard as the boiler) assuming 3kw requirement. 
• Upstairs rads, heat loss approx. 6.2kw. Rads spec'd for DT25 (lowest I could reasonably go)
• Downstairs UFH, heat loss approx. 10kw. 150mm centres. All single zone. 
• Automatic bypass valve fitted before any zone valves/pumps

 

Boiler was spec'd by plumber before I had even heard of BuildHub and therefore pre-renovation. So it's massively oversized for what we now need. 

 

Currently all 3 are separate zones with their own zone valves and pumps. The boiler, rads and UVC were fitted by a plumber, the UFH came as a kit, including zone, pump and mixer, that I fitted myself, the plumber just left me the flow/return pipes to connect the UFH to. I've never been convinced the system as a whole has been running efficiently. 

 

When the UVC is calling for heat, the boiler flow rate goes to max (approx 80c), otherwise it runs on a weather comp curve. 

 

If my calcs are correct, for a system running at DT25, I'm assuming the UVC will require a max flow rate of approx 1.8 LPM, the rads will require a max flow rate of 3.54 LPM and the UFH requires a max flow rate of 5.7 LPM, resulting in a total max flow rate of 11.04 LPM. Which appears to be well within the capacity of the boiler pump. 

 

Long term my aim is for an ASHP to be, near enough, a drop in replacement when the boiler eventually gives up the ghost. 

 

I've been geeking out of Heat Geek videos recently, especially with regards to hydraulic separation and the dos/don't of close coupled tees on UFH systems. 

 

So my questions are;

- Should I remove the UFH and/or UVC pumps and let the boiler pump do its thing? Feels like they're all currently competing against each other, and just feels wrong?

- Should I consider fitting a close coupled tee, and lock shield valve balanced to 25DT, just before the UFH mixer?

Videos that got me thinking, in case anyone's interested;

https://www.youtube.com/watch?v=mNcRx45DQ8M
https://www.youtube.com/watch?v=K_OWE2dJht0

https://www.youtube.com/watch?v=Gk4Mzvcaen0

 

Edited March 5 by jayc89

[OwenF]

1) if you introduce hydraulic separation, you’ll need the UFH/mixer pump in addition to the boiler pump.

 

they might be fighting against each other now, but not once separated.

 

2) CCTs at the mixer are a no-no. This creates a bypass. CCTs or other separation is done at the boiler

 

look up ‘Heating system design’ group on Facebook. Very good resource for the type of system design you’re looking at.

[JohnMo]

Assume the UVC is via a diverter valve and it will only have when a call for heat. So DHW or heating never both together.

 

Your LPM flow rates sound very low especially for the UVC and your UFH. If you need 10kW into the floor with dT of 7 the flow is 20l/min.

 

So then you only need to look at rads and UFH. The issue is likely you will have two distinct flow temps for your heating, UFH and radiators.  So you need a mixer on the UFH, so you also need a pump on the UFH.

 

15 minutes ago, OwenF said:

) CCTs at the mixer are a no-no. This creates a bypass. CCTs or other separation is done at the boiler

Why? All depends how you plump as long as the return to the boiler always exists there is zero issue. Just follow the attached.

 

 

CCT WORCESTER BOSCH.pdf

[SimonD]

When running a mixed circuit with an unmixed circuit (UFH and Rads) you really should have hydraulic separation. Does the system currently have hydraulic separation with all the pumps - how many do you actually have?

 

With your UFH figures, they seem low given the 10kW load. At between 5-7 delta T your flow rate is going to be more like 20-28lpm.

 

Your flow rates with a delta 20 across the boiler will be about:

 

12lpm at the boiler for 16kW

4.3lpm through the rads (Delta 20)

 

Ignore the flow rates for DHW as part of the heating load as you really should be running priority hot water.

 

As mentioned above, you definitely don't want a CCT just before the mixer.

 

So design for a flow of 55 from the boiler, return of 35, then a 40 flow ufh (unless there's  specific need to run hotter due to floor coverings) and 35 return.

 

Then ideally choose a new boiler capable of priority hot water, or is Opentherm compatible and use something like the EPH priority hot water Opentherm Pack. I personally prefer systems that have load compensation and weather compensation, as I think it is always good to have room temp data for the system, but just weather comp and priority hot water work fine.

[jayc89]

Ah I used DT20 for the UFH flow calcs. I was using Flow Rate (LPS) = KW / (DHC * DT), so

 

Rads 6.2kw / (4.2 * 25) = 0.059 LPS (or 3.54 LPM)
UFH 10kw / 4.2 * 7) = 0.34 LPS (or 20.4 LPM)

 

Total requirement, minus DWH approx 24LPM which is more than the boiler pump alone can supply.

 

Currently looks something like this;

 

DWH and heating can be on at the same time, the boiler doesn't have priority hot water, nor differing temps for water/heating. I'm looking to make the best out of the boiler we have rather than considering replacing it - when I do, it'll be for a heat pump. The way the Baxi Weather Comp works is by using a curve if it's only rads/UFH calling for heat, but boosts to max flow temp as soon as the DWH is calling for heat, which is why I need to keep there UFH mixer has a fail safe. 

It was this clip, at 11:55 that suggested the CCT at the UFH end - https://www.youtube.com/watch?v=mNcRx45DQ8M

 

[jayc89]

10 hours ago, JohnMo said:

Assume the UVC is via a diverter valve and it will only have when a call for heat. So DHW or heating never both together.

 

Your LPM flow rates sound very low especially for the UVC and your UFH. If you need 10kW into the floor with dT of 7 the flow is 20l/min.

 

So then you only need to look at rads and UFH. The issue is likely you will have two distinct flow temps for your heating, UFH and radiators.  So you need a mixer on the UFH, so you also need a pump on the UFH.

 

Why? All depends how you plump as long as the return to the boiler always exists there is zero issue. Just follow the attached.

 

 

CCT WORCESTER BOSCH.pdf 114.59 kB · 4 downloads

 

Looks similar to what I had seen in the Heat Geek videos. To ensure I'm understanding that diagram correctly;

 

- Close Couple Tees on the flow pipe

- 2x gate valves for isolation one each leg

- 3-port valve to replace the existing 2 port valve, and connect the two legs ahead of the UFH pump. 

 

Which would allow the UFH pump to run as fast as it needs to, assuming it's capable of doing so (Grundfos UPS3),  and would look something like this;

 

 

The biggest differences between this and the Heat Geek videos, that I can see, is;

1) they suggest a gate/lockshield valve to connect the two legs, opposed to a 3-port valve, and to balance that to the same DT as the rest of your system (DT20 for me) - still balancing the UFH loops down correctly further downstream. 

2) The UFH would be "returning" into the flow pipe, so anything downstream to it would receive a lower flow temp (in my case that's probably not an issue, as the rads and DWH are teed off earlier).

3) When the UFH mixer closes, whilst the 3P Valve is still open, the flow has no where to go. I assume that would just trigger the bypass at the boiler earlier, so is acceptable? 

 

I don't quite understand what the close couple tees do opposed to just fitting the 3P valve alone, is it so if the UFH pump does start to suck, it will most likely suck from the UFH return and not the Rads/DWH? 

That doc also states;

Quote

A 3-port motorised valve with spring return must be used to ensure primary water circulation is possible when there is only a demand on the underfloor heating circuit. It should be powered open to the radiator circuit which it is controlling by the thermostat for that zone. The switched live heating demand for that zone should be via the end switch on the 3-port valve

 

Why does the 3-port valve need to be open for the rads? I assumed, in my set up, it would only need to be open when the UFH is calling for heat and the rad circuit would behave as standard. 

Edited March 6 by jayc89

[JohnMo]

And what benefits are the changes making?

 

If you are flowing a high temp into UFH most of the flow going through the pump is recycling with some hot water being introduced.

38 minutes ago, jayc89 said:

don't quite understand what the close couple tees do opposed

They provide a hydraulic null point in the system and hydraulic separation so pumps do not see each other. For UFH the mixer does that already, so adding a cct to an already working system adds no value.

[JohnMo]

10 hours ago, jayc89 said:

DWH and heating can be on at the same time, the boiler doesn't have priority hot water, nor differing temps for water/heating. I'm looking to make the best out of the boiler we have rather than considering replacing it - when I do, it'll be for a heat pump. The way the Baxi Weather Comp works is by using a curve if it's only rads/UFH calling for heat, but boosts to max flow temp as soon as the DWH is calling for heat, which is why I need to keep there UFH mixer has a fail safe. 

So confused - what are you doing now and trying to do. What has the heat pump got to do with your boiler?

Edited March 6 by JohnMo

[jayc89]

17 minutes ago, JohnMo said:

And what benefits are the changes making?

 

If you are flowing a high temp into UFH most of the flow going through the pump is recycling with some hot water being introduced.

They provide a hydraulic null point in the system and hydraulic separation so pumps do not see each other. For UFH the mixer does that already, so adding a cct to an already working system adds no value.

 

I've never been convinced the heating system as a whole has been working correctly, mainly because things like the UFH were "bolted on" as an after thought when the original system was designed. I want to confirm, if there are any problems with the current set up, what they are, and ensure I understand what any more suitable set up should be - if there's no problem to fix here, it's one less thing I have to worry about. 

 

13 minutes ago, JohnMo said:

So confused - what are you doing now and trying to do. What has the heat pump got to do with your boiler?

 

Nothing, it was in reply to the suggestion of replacing my boiler now - I want to make the best of what I already have. When it does finally give up the ghost is when I'll replace it, for a heat pump. 

[jayc89]

A CCT was advised alongside a mixing valve in this topic. What determines whether one is/isn't required?

Edited March 6 by jayc89

[sharpener]

15 hours ago, jayc89 said:

When the UVC is calling for heat, the boiler flow rate goes to max (approx 80c), otherwise it runs on a weather comp curve. 

 

12 hours ago, jayc89 said:

DWH and heating can be on at the same time, the boiler doesn't have priority hot water, nor differing temps for water/heating. I'm looking to make the best out of the boiler we have rather than considering replacing it - when I do, it'll be for a heat pump. The way the Baxi Weather Comp works is by using a curve if it's only rads/UFH calling for heat, but boosts to max flow temp as soon as the DWH is calling for heat, which is why I need to keep there UFH mixer has a fail safe. 

 

So you will get better efficiency out of your existing setup if you heat the DHW at times you do not need the heating on e.g. the small hours.

 

15 hours ago, jayc89 said:

Currently all 3 are separate zones with their own zone valves and pumps.

 

The diagram you drew does not show pumps on the DHW or rad circuits. The are probably not necessary anyway, and almost certainly not with a heat pump.

 

15 hours ago, jayc89 said:

If my calcs are correct, for a system running at DT25

 

Your setup is otherwise similar to standard HP schematics and therefore conversion to a HP should be quite simple, assuming the pipework will support the much higher flow rates (HP will have only 5C delta T so require 5x the flow to transfer the same amount of heat). I have a similar basic setup on a 30kW oil boiler which I am planning to replace with a 12kW Vaillant Arotherm Plus but your heat losses are too big for that.

Avoid hydraulic separation if you can (and especially CCTs), probably not required in any form and a cause of inefficiency for fundamental thermodynamic reasons. Existing mixing valve on UFH circuit is all you need to cope with the different flow temp requirements.

[SimonD]

16 hours ago, jayc89 said:

DWH and heating can be on at the same time, the boiler doesn't have priority hot water, nor differing temps for water/heating. I'm looking to make the best out of the boiler we have rather than considering replacing it - when I do, it'll be for a heat pump. The way the Baxi Weather Comp works is by using a curve if it's only rads/UFH calling for heat, but boosts to max flow temp as soon as the DWH is calling for heat, which is why I need to keep there UFH mixer has a fail safe. 

 

In that case, simply install a normally open 2-port vale on the flow to the heating system that is wired to close when the cylinder calls for heat and opens the normally closed 2-port - that way all your high temp flow will go to the cylinder and not your heating circuit. This is pretty easy to plum and wire in and gives you your priority hot water.

[jayc89]

11 minutes ago, SimonD said:

 

In that case, simply install a normally open 2-port vale on the flow to the heating system that is wired to close when the cylinder calls for heat and opens the normally closed 2-port - that way all your high temp flow will go to the cylinder and not your heating circuit. This is pretty easy to plum and wire in and gives you your priority hot water.

 

Similar dawned on me this afternoon, albeit by slightly different means; replace the UVC 2 port valve with a 3 port valve that diverts all flow to the UVC when it's calling for heat (as it'll be > 80c) and thus protect the heating zones from overshooting. 

I guess similar could be achieve by just setting the system to only heat the UVC overnight when the heating zones will never call for heat? (Save for some freak cold snap)

[SimonD]

7 hours ago, jayc89 said:

The UFH would be "returning" into the flow pipe, so anything downstream to it would receive a lower flow temp

 

Yes, it means, at least IIRC on this, there would have to be a corresponding increase to the boiler flow temp which could then cause issues in any blending at the manifold - so why do it? I don't think the WB diagram is quite correct, but then in my experience of how they've implemented their boilers with modulation, EMS and controls, it wouldn't surprise me at all if they did mean it.

[SimonD]

On the subject of Close Couple Tees and their location on the system, I thought this video provides some very sensible guidance.

 

 

 

Edited March 6 by SimonD

[OwenF]

Yes Andrew Millward (whilst a bit persistent with his views) does appear very knowledgable on hydraulic design.

 

On 05/03/2024 at 21:24, JohnMo said:

 

Why? All depends how you plump as long as the return to the boiler always exists there is zero issue. Just follow the attached

The above video explains better than I can but there seems to be plenty empirical evidence that CCTs at the mixer starves UFH of flow.

 

Heat Geek disagrees but does acknowledge a gate valve somewhere nearby (can’t remember exactly where) is necessary to ensure the UFH circuit gets required flow.

 

On 05/03/2024 at 22:20, jayc89 said:

DWH and heating can be on at the same time, the boiler doesn't have priority hot water, nor differing temps for water/heating. I'm looking to make the best out of the boiler we have rather than considering replacing it - when I do, it'll be for a heat pump. The way the Baxi Weather Comp works is by using a curve if it's only rads/UFH calling for heat, but boosts to max flow temp as soon as the DWH is calling for heat, which is why I need to keep there UFH mixer has a fail safe. 
 

I converted my previous Viessmann 100 to priority hot water by wiring a 3 port valve in reverse as an external diverter valve.

TBF the boiler did have PDHW function (I.e. different flow temps) but only using the internal diverter valve with a 4-pipe setup.

I didn’t want to run new pipes so took the LV link for PDHW from the boiler to a relay which gave 240v to power the reversed 3-port valve over to the cylinder only. Have the schematic somewhere but recall linking wires makes the 3-port normally open to heating only.

 

it’s a neat retrofit (not my idea) and sounds like your Baxi might have a similar output for the ‘boost’ mode. You just need something to trigger the relay.

Edited March 6 by OwenF

[JohnMo]

A better heat geek video on how to do your UFH connections 

 

https://www.google.co.uk/url?sa=t&source=web&rct=j&opi=89978449&url=https://m.youtube.com/watch%3Fv%3DK_OWE2dJht0&ved=2ahUKEwj57Zj31OGEAxWqQUEAHROIC2cQwqsBegQIDhAG&usg=AOvVaw3Kv9nfBIKtwsPqVzMirE55

 

[jayc89]

3 minutes ago, JohnMo said:

A better heat geek video on how to do your UFH connections 

 

https://www.google.co.uk/url?sa=t&source=web&rct=j&opi=89978449&url=https://m.youtube.com/watch%3Fv%3DK_OWE2dJht0&ved=2ahUKEwj57Zj31OGEAxWqQUEAHROIC2cQwqsBegQIDhAG&usg=AOvVaw3Kv9nfBIKtwsPqVzMirE55

 

 

A couple of options are banded about on that video;

1) his original, which Andrew responded to

2) an updated version of it in response to Andrew's comments

3) Andrew's proposed design, which he tries to debunk

4) an alternative using an electronic blending valve instead

 

If I understand them correctly, option 2 (his updated version), proposes to fit a pair of tees across the flow/return to the UFH, after the UFH pump/zone vale, with a gate/lockshield valve in-between the tees to balance at this point to the same DT as the rest of the system (so DT25 in my case). I don't think this is strictly a CCT set up, as the tees are actually bridging the flow/return pipes, not both on the flow, is that correct? 

 

Similarly, the "alternative" proposal is to fit an electronic blending valve ahead of the UFH pump, and again, bridge the flow/return to the UFH with a couple of tees and balancing valve, ahead of the UFH pump too. In this set up, isn't the purpose of the electronic blending valve to do the same job as my existing UFH mixer (albeit more accurately)? So the blending valve/existing mixer does indeed provide the hydraulic separation, what's the need for the additional tees and balancing valve? 

 

[JoeBano]

Have a look at this video, i could be wrong but looks very similar what you have now.

 

[JohnMo]

4 minutes ago, jayc89 said:

DT25 in my case

You have dT 25 radiators, that means the difference between room temp and radiator temp, flow and and return dT would be in the region of 20.

5 minutes ago, jayc89 said:

I don't think this is strictly a CCT set up

It's not

 

 

 

[SimonD]

55 minutes ago, jayc89 said:

If I understand them correctly, option 2 (his updated version), proposes to fit a pair of tees across the flow/return to the UFH, after the UFH pump/zone vale, with a gate/lockshield valve in-between the tees to balance at this point to the same DT as the rest of the system (so DT25 in my case). I don't think this is strictly a CCT set up, as the tees are actually bridging the flow/return pipes, not both on the flow, is that correct? 

 

I think it's a bit of a hybrid. What isn't entirely clear in the video, but I think nececessary is CCTs to the UFH manifold which is still Delta T 7 with the balanced 'loop' Delta T 20. So it is using CCTs, they're just plumbed in differently onto this loop. This obviously takes more work and balancing.

 

I'm still not sure, or convinced, of the overall benefit of this implementation, other than as an intellectual/academic debate between system designers, or its use under some necessary installation situations, like a retrofit/system rectification. If designing a new installation, would you do it this way as a first option? Probably not..

 

Ideally, your boiler wants to see about 20, not 25, unless like some boilers, the MIs specifically state 15 or even 11 in one model I've installed on a few occasions.

[sharpener]

2 hours ago, SimonD said:

If designing a new installation, would you do it this way as a first option? Probably not..

 

This.

[John Carroll]

On 05/03/2024 at 19:01, jayc89 said:

Current setup;

• 32kw Baxi Platinum+ System Boiler, with weather comp, fitted upstairs, believe max flow rate is 20 LPM.
• 250L UVC (same cupboard as the boiler) assuming 3kw requirement. 
• Upstairs rads, heat loss approx. 6.2kw. Rads spec'd for DT25 (lowest I could reasonably go)
• Downstairs UFH, heat loss approx. 10kw. 150mm centres. All single zone. 
• Automatic bypass valve fitted before any zone valves/pumps

 

 

 

This may (or may not) be of some help of seeing the "total" picture, I configured this when my daughter was installing UFH and just tailored it to your UFH + rad heating demands, I'm not sure when you say spec'd for DT25 if you mean rads designed as "T25" rads?, if so they will give a rad output of 40.6% of T50 sized rads and would require a oversizing factor of X2.46. say 2.5. I'm also assuming that the required boiler flow temperature required is 48C and the UFH required manifold temp is 40C. I didn't include cylinder heating.

 

 

 

[SimonD]

4 minutes ago, John Carroll said:

I'm also assuming that the required boiler flow temperature required is 48C and the UFH required manifold temp is 40C. I didn't include cylinder heating.

 

It's interesting you're arriving at a boiler dT of 9.53 using a flow temp of 48C. Now, my experience playing with boilers is making me want to avoid running any boiler below a 55C flow temp. (unless it's modulating correctly through suitable controls - usually manufacturer ones). I ran a rather unscientific test on this the other day while I was sitting around waiting for a parts delivery all day long so was twiddling my thumbs. It's a system I have fully designed and installed using radiators and manifold distribution system. I tried to get the system running at 55, 50, 45, and 40C. It was a cold day so the full heat load was required for the most part. This system is modulated using Opentherm with weather and load compensation. During my test I found that as soon as I took the flow temperature below about 50, the boiler would start short-cycling. At a flow temp of 40C, I was getting cycles below 10 seconds, and 45C they were barely any better at about 15-20 seconds. When I finally brought the system back to 55C I got a consistent and constant run time of over 2 hours with a nicely consistent return temp oscillating between about 37 and 43C. Each time it got to 43C, pump/boiler would modulate to bring it back to 37C. At the lower flow temps, the boiler seemed to want a return temp of no less than about 35, which was what caused the short cycling and boiler temper tantrums.

 

But also one question - how do you achieve the 24lpm boiler flow rate if the boiler pump flow rate is max 20lpm? Surely then you need hydraulic separation to run higher boiler flow/return temps and then higher flow rates for the low temp rads/ufh?

[John Carroll]

1 hour ago, SimonD said:

 

It's interesting you're arriving at a boiler dT of 9.53 using a flow temp of 48C. Now, my experience playing with boilers is making me want to avoid running any boiler below a 55C flow temp. (unless it's modulating correctly through suitable controls - usually manufacturer ones). I ran a rather unscientific test on this the other day while I was sitting around waiting for a parts delivery all day long so was twiddling my thumbs. It's a system I have fully designed and installed using radiators and manifold distribution system. I tried to get the system running at 55, 50, 45, and 40C. It was a cold day so the full heat load was required for the most part. This system is modulated using Opentherm with weather and load compensation. During my test I found that as soon as I took the flow temperature below about 50, the boiler would start short-cycling. At a flow temp of 40C, I was getting cycles below 10 seconds, and 45C they were barely any better at about 15-20 seconds. When I finally brought the system back to 55C I got a consistent and constant run time of over 2 hours with a nicely consistent return temp oscillating between about 37 and 43C. Each time it got to 43C, pump/boiler would modulate to bring it back to 37C. At the lower flow temps, the boiler seemed to want a return temp of no less than about 35, which was what caused the short cycling and boiler temper tantrums.

 

But also one question - how do you achieve the 24lpm boiler flow rate if the boiler pump flow rate is max 20lpm? Surely then you need hydraulic separation to run higher boiler flow/return temps and then higher flow rates for the low temp rads/ufh?

The boiler dT of 9.5C is because (UFH return) 9.56LPM at 33C is mixing with (Rad return) 14.81LPM at 42C to give a mixed return of 24.37LPM at 38.5C, boiler dT, 48C-38.5C, 9.5C.

When WC keeps reducing the boiler flow temp then the rad output obviously reduces but the boiler shouldn't cycle unless its minimum output is greater than the heating demand or if, for some reason the boiler dT exceeds 30C. Also when/if the boiler does refire then it will refire at around 65% of maximum output and the flow temp will oftern exceed the target temp by 5C means burner trip and recycle  especially if the WC has driven it down to say 40/45C.

"But also one question - how do you achieve the 24lpm boiler flow rate if the boiler pump flow rate is max 20lpm? Surely then you need hydraulic separation to run higher boiler flow/return temps and then higher flow rates for the low temp rads/ufh?",                                                                                                                                                                                         A 32kw boiler with a dT of 20C, will have flowrate of (32*860/60/20), 23LPM so a flow rate of 24LPM will only result in a dP of 3.8M, a increase of 0.3M (assuming a design HEX dP of 3.5M) but if the boiler flow temp is increase by only 2C to 50C then the flowrate decreases to only 17.32LPM because the UFH return flow has fallen from 9.56LPM to 8.43LPM and the rads (because they can now be throttled in to give the same output at a flow of 50C vs 48C, rad dT is now 10C vs 6C) from 14.8LPM to 8.9LPM.

So the higher the boiler flow temperature the lower the flowrate, the UFH will automatically reduce the return flowrate by recirc more through the mixer, the rads have to be manually throttled or have TRVs.

 

Edited March 7 by John Carroll