Boiler keeps short cycling with only UFH is on

[windsor-tg]

Good morning, wondered if anyone can advise please on a resolution?

My set up at home is that I have a 4-bed detached with 21 radiators (includes 4 towel rails), a 6m x 4m orangery/conservatory that was built last year with under-floor heating, and a Megaflo for storing hot water. The UFH contains two loops in a screeded floor, each tube loop in the floor to the manifold is approx 60m long.

Last year, I had all my radiators changed (unfortunately I did not have the system fully flushed) and the main pump upgraded to a Grundfos UPS2 25-80 pump. I control the hot water, central heating, and UFH through the Heatmiser app. The hot water, CH and UFH each have their own zone valve. I have attached a photo (UFH manifold) of my UFH manifold taken last October and a photo ('Zone valves labelled') of my airing cupboard where the megaflo, main pump and zone valves are located. Zone valves marked: 1 = HW, 2 = UFH, 3 = CH.

I currently have a Vaillant ecoTEC Plus 438 38kW system boiler which is 13 years old located in the garage. A few years ago, I sused to get an S.53 message displayed on the boiler. To eliminate this display message, I had to turn down the power of the boiler to 24kW (d0 = 24) so I have been running the heating & HW system with boiler capped at 24kW.

The problem I am having is that when I only have the UFH turned on, the boiler is continously/short cycling. When I turn on the central heating, the issue goes away. In discussion with the Vaillant technical team, their view is the reason why my boiler is cycling is because the flow rate through the boiler & system is not enough when only the UFH is operating. In other words, they are saying the boiler continues to cycle because the boiler is not able to modulate down far enough. For a 24kW boiler, the Vaillant technical person said the min flow rate needs to be 17.2L/min, and for a 30kW boiler, the flow rate needs to be minimum 21.5L/min. I have no idea to know what the flow rate is for my UFH.

There is mention from a couple of heating engineers about introducing a Low Loss Header.

Recently, my boiler is making the 'kettling' sound so I am happy to change the boiler but I also want to eliminate this short cycling boiler issue when I only have the UFH turned on.

Can anyone advise here?

 

[Timmy D]

Low loss header will help.

 

[Chuck]

A low-loss header (LLH) is the solution if the root cause is low flow rate through the UFH. If this is the cause the UFH won't be putting out enough heat to maintain the room temperature and this will get progressively more noticeable as the weather gets colder.

A LLH may or may not help if the root cause is low demand for heat from the UFH, i.e. it's below the boiler minimum because the weather is relatively warm or the house insulation is good. In this case, the cycling will get less frequent as the weather gets colder and the UFH will work as designed.

In the second case, the LLH will help to some extent by increasing the effective volume of water in the boiler-UFH loop, which will lengthen the cycling period. If the boiler's "anti-short cycle lock-out" state is being triggered an LLH may be enough to fix it. If not, the dead-cert cure in this case is to add a 'buffer store' to the heating circuit. People don't often do this because (a) cost, (b) space, and (c) modern boilers are designed to cycle (the amount of fuel it wastes is a tiny fraction of the annual total) when heat demand is low, and (d) it only happens in the autumn and spring when you barely need the heating on anyway.

Changing the maximum power setting on the boiler won't affect this aspect of the behaviour, which is related to the minimum power the boiler can deliver while running continuously. (This is governed by the burner design.) If you do change the boiler go for one, e.g. Vessmann, with a low minimum power.

 

[windsor-tg]

A low-loss header (LLH) is the solution if the root cause is low flow rate through the UFH. If this is the cause the UFH won't be putting out enough heat to maintain the room temperature and this will get progressively more noticeable as the weather gets colder.

A LLH may or may not help if the root cause is low demand for heat from the UFH, i.e. it's below the boiler minimum because the weather is relatively warm or the house insulation is good. In this case, the cycling will get less frequent as the weather gets colder and the UFH will work as designed.

In the second case, the LLH will help to some extent by increasing the effective volume of water in the boiler-UFH loop, which will lengthen the cycling period. If the boiler's "anti-short cycle lock-out" state is being triggered an LLH may be enough to fix it. If not, the dead-cert cure in this case is to add a 'buffer store' to the heating circuit. People don't often do this because (a) cost, (b) space, and (c) modern boilers are designed to cycle (the amount of fuel it wastes is a tiny fraction of the annual total) when heat demand is low, and (d) it only happens in the autumn and spring when you barely need the heating on anyway.

Changing the maximum power setting on the boiler won't affect this aspect of the behaviour, which is related to the minimum power the boiler can deliver while running continuously. (This is governed by the burner design.) If you do change the boiler go for one, e.g. Vessmann, with a low minimum power.

A low-loss header (LLH) is the solution if the root cause is low flow rate through the UFH. If this is the cause the UFH won't be putting out enough heat to maintain the room temperature and this will get progressively more noticeable as the weather gets colder.

A LLH may or may not help if the root cause is low demand for heat from the UFH, i.e. it's below the boiler minimum because the weather is relatively warm or the house insulation is good. In this case, the cycling will get less frequent as the weather gets colder and the UFH will work as designed.

In the second case, the LLH will help to some extent by increasing the effective volume of water in the boiler-UFH loop, which will lengthen the cycling period. If the boiler's "anti-short cycle lock-out" state is being triggered an LLH may be enough to fix it. If not, the dead-cert cure in this case is to add a 'buffer store' to the heating circuit. People don't often do this because (a) cost, (b) space, and (c) modern boilers are designed to cycle (the amount of fuel it wastes is a tiny fraction of the annual total) when heat demand is low, and (d) it only happens in the autumn and spring when you barely need the heating on anyway.

Changing the maximum power setting on the boiler won't affect this aspect of the behaviour, which is related to the minimum power the boiler can deliver while running continuously. (This is governed by the burner design.) If you do change the boiler go for one, e.g. Vessmann, with a low minimum power.

thanks @Chuck. As requested by another person, they wanted to see what the flow & return temps were showing on the boiler as well as the flow meters on the UFH manifold. Below are the readings (I also attached a file showing these readings incase the image below did not come through) . As well as taking the flow & return readings from the boiler, I also took readings from the UFH manifold. The blending valve on the UFH is set around 47 degrees.

I took the readings across 5 minute intervals. What I found is that when I had both, CH and UFH turned on, the boiler was happily firing (without any sound of kettling which there was earlier in the week). The flow meters on the UFH was showing 2.5 (range 0-5). When I turned off the UFH, the flow meters on the UFH went to 5 and these flow meters stayed on 5 when I turned on the UFH and turned off the CH.

After about 10 minutes of the UFH only being on, I could see the boiler short cyling (every 2 mins) and the boiler flow temperature would rocket from 45 to 72 degrees.

When I provided this information to the other person, their response was "Simple fix a 2 port buffer of around 25 to 50 L capacity (closer to 50L is better), across the flow and return loop to the boiler or across the flow and return from the UFH, where ever is easiest to install. No need for low loss headers or 4 port buffers and additional pump (s).

Parts list

• Two Tee pieces,
• A buffer or volumiser - these are cheap enough, you need to insulate yourself.
• Two isolation valves"

Would this potential setup work?

 

[fowlerboi]

Your UFH loops are running at 5LPM of flow each?

 

[windsor-tg]

Your UFH loops are running at 5LPM of flow each?

The flow meters have been tinkered about of late by some heating engineers. You will see from the readings that when both, the CH and UFH are on, the UFH loops are each running at 2.5LPM. When the UFH then runs on its own, the flow bumps up to 5LPM. I do not know why.

Would this possibly make the boiler short cycle when only the UFH is on?

 

[fowlerboi]

John.g is the pump expert so hopefully he can chirp in on what pump setting you should be on to keep it more consistant

 

[windsor-tg]

thanks @fowlerboi. @John.g - what do you think?

 

[ShaunCorbs]

I suspect your boiler pump is interfering with the ufh pump hence why your getting more flow as you have a short / low demand system

You need to reset the flow meters back to spec / Heatloss as you could be damaging the floor / having higher return temps should be 7-10 dc difference between flow and return

 

[windsor-tg]

I suspect your boiler pump is interfering with the ufh pump hence why your getting more flow as you have a short / low demand system

You need to reset the flow meters back to spec / Heatloss as you could be damaging the floor / having higher return temps should be 7-10 dc difference between flow and return

Unfortunately, I don’t have any spec as the UFH was installed by the builder’s heating guy and there was no design.

Do I need to turn down the flow meters to 2.5L when I only have the UFH on? Will that reduce the boiler cycling?

 

[ShaunCorbs]

Unfortunately, I don’t have any spec as the UFH was installed by the builder’s heating guy and there was no design.

Do I need to turn down the flow meters to 2.5L when I only have the UFH on? Will that reduce the boiler cycling?

Need to get some temp clamps and get the individual flow and return temp differential on each set of white pipes to 5-7 dc on start up before the slab is hot eg when it’s had a day or two to cool

Won’t stop the boiler cycling due to ufh being let’s say 2kw and the min boiler kw to 6 etc

 

[Kurt Stone]

I had this same problem after installing my UFH. With radiator circuit on no problem but with ufh only the boiler was short cycling. I opened the manual bypass valve about 1 turn and the problem has gone away.

 

[Knappers]

Could you not just fit an auto bypass across flow and return to UFH manifold (or wherever there is most pipework/water available to increase themal mass cycling through boiler)?
This boiler has a very effective one fitted internally which is prob what is contributing to the kettling.
An external bypass at a lower resistance would bring the thermal mass of water in the flow and return pipes into play and automatically compensate for different heating demands?

If its not too difficult you could maybe even consider increasing size of primary pipework?

 

[Ben-gee]

If you’re considering replacing the boiler anyway, I would just do that.
38kW is way too big - get it sized to your actual heat loss and choose one which can modulate down to the size of your underfloor loss and there will be no problems…. if such a beast exists.

 

[windsor-tg]

I had this same problem after installing my UFH. With radiator circuit on no problem but with ufh only the boiler was short cycling. I opened the manual bypass valve about 1 turn and the problem has gone away.

@Kurt Stone - thanks. I have an automatic bypass value fitted along with a gate value (photo attached with yellow line showing bypass valve). If I were to close the gate valve and then open it one turn, would that be similar to what you did?

 

[windsor-tg]

If you’re considering replacing the boiler anyway, I would just do that.
38kW is way too big - get it sized to your actual heat loss and choose one which can modulate down to the size of your underfloor loss and there will be no problems…. if such a beast exists.

Even though I have a 38kW boiler, it has been capped at 24kW as I was getting an S.53 display message on the boiler (flow & return temp differential too large) so reducing the power of the boiler (d.0) to 24kW eliminated the message. A 30kW boiler is what is being suggested but I’m going to find it hard to find a boiler that can modulate down to around 1.5kW when I just have the UFH on.

 

[John.g]

@Kurt Stone - thanks. I have an automatic bypass value fitted along with a gate value (photo attached with yellow line showing bypass valve). If I were to close the gate valve and then open it one turn, would that be similar to what you did?

Where is the photo of the above?.

Re your tests:
At T+15, with both CH&UFH on, the UFH shows flow/returns of 48C/27C with a combined flowrate of 5LPM, this is a UFH output of 7.33kw, at T+40, UFH only on, UFH flow/returns of 49C/34C, combined flowrate 10LPM = UFH output of 10.5kw, in both cases, dts of 21C & 15C seem way higher than normal of ~ 7C to 10C.
Also, at T+20 & T+25, CH only on, the UFH still shows a combined flow of 10LPM? can you check your labelling again?

The UPS2 8M pump at any "fixed speed" setting will pump at a head of 6M at even the lowest setting., might be able to suggest a more suitable setting even though there is no constant pressure mode on that pump.
THe ABV at any index setting will have a very high flowrate at speeds 2&3 (7M &M) and will cause rapid cycling and might even still pass at speed 1 (6M).

So,
Check the pump setting.
post a photo of the bypass with the manual valve & the ABV and its make and setting.
what is/was the boiler flow target temperature at the above tests?

 

[windsor-tg]

Thanks @John.g
I will look to take a fresh set of readings over the weekend in case I made a mistake.
I have attached photos of the ABV. Hopefully this has come through.

Where you say ‘check the pump settings’ - do you mean the main pump or the UFH pump?
The boiler flow temp is set to 65 but I will double check before I take the new set

 

[John.g]

No, check the main circ pump near the zone valves or just take a photo of the LED display while running, if the manual by pass valve isn't well throttled then the ABV at that setting could well be by passing ~ 8 LPM or more as the minimum fixed speed pump head is 6M.

 

[ShaunCorbs]

The red pump

 

[Ben-gee]

Even though I have a 38kW boiler, it has been capped at 24kW as I was getting an S.53 display message on the boiler (flow & return temp differential too large) so reducing the power of the boiler (d.0) to 24kW eliminated the message. A 30kW boiler is what is being suggested but I’m going to find it hard to find a boiler that can modulate down to around 1.5kW when I just have the UFH on.

boilers that can modulate 10:1 , I doubt you need 30 kw.
18kW is probably enough if you just time the hot water to heat at a time you don’t require heating- then that gets you down to 1.8kw for when you only want ufh.

 

[windsor-tg]

@John.g - I just ran a fresh test and the readings are shown below. The boiler temp is set at 65 degrees. When the UFH is off, the flow meters are 0 so I must have taken a wrong reading yesterday. Also, on T+40, the flow temp on UFH was either 40 or 50. I wrote down 50 at the time but I think that was wrong and it should have possibly been 40.

Are the readings still strange? No kettling sound from the boiler when I was running these test readings.

I have attached a photo of the main pump as well as the UFH pump so that you can see what lights are lit up on each of the pumps. I will send a short video on the main pump in a few mins so that you can possibly hear the sound of the pump when it is running with both, CH and UFH on.

Let me know if more tests and/or photos/videos will be helpful.

 

[windsor-tg]

video of main pump running with both, CH and UFH on

 

[windsor-tg]

also attached photo of UFH pump when UFH is running

 

[John.g]

First I would check the setting of the manual by pass valve with the red handwheel, put a mark on the top of the handwheel at "12 oclock", then turn the hand wheel very slowly clockwise and carefully note the portion of a turn or the full turns required to close it fully, (should be ~ 3 to 4 full turns between fully open and fully closed) then return it initially to the same setting, you may find a 1/2 turn or so of play before the valve starts to reopen so allow for this, (of course if the valve has been left fully (unlikely) open then you will get 3 or 4 full turns to close it)
The main pump is running on full speed (3) which is 8M, I would change that immediately to speed 1 which is 6M and still more than adequate IMO, the 8M setting is probably the reason for the UFH flowrates doubling when UFH running alone. To change the speed, just press the setting button briefly a few times to highlight I (solid green), that may still be too powerful at 6M but we'll see.
The manifold pump is running on minimum speed in fixed speed (CC) mode so leave that as is even though constant pressure (CP) mode is recommended for UFH.

THe UFH seems to be emitting a fair amount of heat, at T+15, CH+UFH on, a total flow of 5LPM with flow/return of 44C/28C gives a output of 5.59kw and at T+50, UFH on its own,, a total flow of 10LPM with flow/return of 44C/34C gives a output of 6.98kw.
You might run a few more tests after changing the pump speed and checking the bypass opening. Maybe just run the UFH only for 15/20 minutes or so to speed things up., you can then run on both once you post the UFH findings.

 

[windsor-tg]

First I would check the setting of the manual by pass valve with the red handwheel, put a mark on the top of the handwheel at "12 oclock", then turn the hand wheel very slowly clockwise and carefully note the portion of a turn or the full turns required to close it fully, (should be ~ 3 to 4 full turns between fully open and fully closed) then return it initially to the same setting, you may find a 1/2 turn or so of play before the valve starts to reopen so allow for this, (of course if the valve has been left fully (unlikely) open then you will get 3 or 4 full turns to close it)
The main pump is running on full speed (3) which is 8M, I would change that immediately to speed 1 which is 6M and still more than adequate IMO, the 8M setting is probably the reason for the UFH flowrates doubling when UFH running alone. To change the speed, just press the setting button briefly a few times to highlight I (solid green), that may still be too powerful at 6M but we'll see.
The manifold pump is running on minimum speed in fixed speed (CC) mode so leave that as is even though constant pressure (CP) mode is recommended for UFH.

THe UFH seems to be emitting a fair amount of heat, at T+15, CH+UFH on, a total flow of 5LPM with flow/return of 44C/28C gives a output of 5.59kw and at T+50, UFH on its own,, a total flow of 10LPM with flow/return of 44C/34C gives a output of 6.98kw.
You might run a few more tests after changing the pump speed and checking the bypass opening. Maybe just run the UFH only for 15/20 minutes or so to speed things up., you can then run on both once you post the UFH findings.

View attachment 84853

Thanks @John.g, I will look to do this tomorrow. Just to clarify, for the ABV, are you just wanting to know how many turns it will take to close the valve? I will see how many turns it will take to close it before applying the same number of turns anti-clockwise to return it back to its same position. Unfortunately I do not have the manual for the ABV.

On the main pump, how do I know if the pump is running at fixed speed operation vs proportional pressure control? I’m pretty certain that I have to keep the button pressed before I can then change the speed setting. By doing this, I’m not sure if I am changing the pump from fixed speed to proportional pressure, or vice-versa. Do I need to have it at fixed speed or proportional pressure, whilst being on speed (1)?

At the time I had the UFH installed last year and adding a couple of more radiators, I had the main pump changed to this one. I do recall the installer setting it to speed (3) to ensure the radiators furthest away from the boiler heated up sufficiently.

Once I carry out the UFH tests for 15/20 mins after changing the main pump speed, is it worth then changing the UFH pump to CP mode and noting down the results for a further 15/20 mins?

 

[John.g]

Don't touch the ABV, I'm refering to the manual valve with the red handwheel
As per pump instructions, the green LED is solid on any fixed speed mode and a flashing green in any PP mode.
Will get back to you re UFH manifold pump setting.

 

[windsor-tg]

@John.g - I got a chance to do this. On the manual valve, I placed a mark on the 12 o'clock position on the valve and turned it to 5 o'clock position before the valve closed. I returned it back to the 12 o'clock position.

i changed the main pump speed to speed (1) - photo attached.

I took the readings below with just the UFH on. The boiler was pretty much cycling from the start.

 

[John.g]

Well, the average flow/return of those 4 readings is 43.25C/29C which gives a output of 5.06kw at a flowrate of 5LPM, the boiler's minimum output is, I think, 6.5kw so it has to cycle.
Before you run it on CH+UFH will you just check that manual valve again, just shut it fully and then confirm that it is fully shut by fully opening it by counting ~ 3 to 4 full turns then close it again and reopen it to 12 oclock.

 

[windsor-tg]

Thanks @John.g - one thing I forgot to do was to close the manual value and re-opening it before running the last set of tests. So I did this afterwards.

Do you want me to fully shut it, fully open it and then reopen it to 12 o’clock before running tests with CH+UFH or run the tests again with UFH only?

I assume I leave the main pump at speed (1)?

 

[John.g]

Some gate valves can be a bit sticky so you want to ensure that it is ~ a 1/2 turn open, before doing any tests just shut it fully then reopen it fully, should take 3 to 4 full turns, then shut it again and reopen it to 12 oclock, then repeat the UFH only test and post findings, then do the UFH+CH tests.
Leave the pump on speed 1. It will bve interesting to see what the UFH flows are then.

Also let both UFH & CH on long enough to enable the boiler flow temperature reach its target temperature of 65C or there abouts.

 

[windsor-tg]

thanks @John.g
I did what you suggested with the manual valve. It took quite a few turns from the closed position to the open position and vice-versa. It is now back in the 12 oclock position.

Latest readings shown below. General observation is that when the UFH is only on, boiler continually cycling. When CH turned on, majority radiators luke warm with rads furthest away cold. Only when the boiler flow temp went up 60+, did the radiators start heating up. Not sure if this is due to the main pump speed being turned down to speed (1). When the main pump speed was (3), all the rads definitely get heated pretty quickly (with the furthest rads taking longer).

When both, the CH and UFH were on, and the boiler flow temp got into the 60s, I noticed the higher temps on the UFH manifold readings

 

[windsor-tg]

boilers that can modulate 10:1 , I doubt you need 30 kw.
18kW is probably enough if you just time the hot water to heat at a time you don’t require heating- then that gets you down to 1.8kw for when you only want ufh.

I’m no expert but I’ve been told that just to get a rough idea, apply 1.5kW to each rad. Since I have 21 rads, that would make the total approx 31kW excluding the demand from the UFH. I can’t see how an 18kW boiler is enough or am I missing something?

 

[ShaunCorbs]

Nope a rough ideal would be if they heat the room fine look up the size and type that will give you a rough output then you can add them all together

 

[John.g]

thanks @John.g
I did what you suggested with the manual valve. It took quite a few turns from the closed position to the open position and vice-versa. It is now back in the 12 oclock position.

Latest readings shown below. General observation is that when the UFH is only on, boiler continually cycling. When CH turned on, majority radiators luke warm with rads furthest away cold. Only when the boiler flow temp went up 60+, did the radiators start heating up. Not sure if this is due to the main pump speed being turned down to speed (1). When the main pump speed was (3), all the rads definitely get heated pretty quickly (with the furthest rads taking longer).

When both, the CH and UFH were on, and the boiler flow temp got into the 60s, I noticed the higher temps on the UFH manifold readings

View attachment 84867

THe UFH is outputting ~ 3.9kw now.
Re rapid cycling, you should be able to extend that, on most Vaillants the anticycle time depends on the target temperature and the set time, it doesn't show this table on your MIs but I'd be surprised if it isn,t.
Check d2, it defaults to 20mins which is actually a anticycle time of 4.5mins at 65C, your anticycle time is 2 mins which suggests a time setting of 10 mins or less, if its at some low setting, suggest setting it to 20mins which should give a actual anticycle time of 4.5 minutes, you can allways increase it to 25 or a max of 30 minutes which gives max actual anticycle time of 6.5 mins at 65C. You could allways consider then installing a 25L (or bigger) buffer on the UFH flow which would give a steady (theoretical) flow temp of 65C for 10 minutes at 3.9kw (1.9LPM) and 5 minutes (4.8LPM)at 10kw UFH output.

You might open the manual by pass another 1/4 to 1/2 a turn, there does seem to be some bypass as the boiler return is ~ 40C with a UFH return of 35C, a bit more might help. Also feel the bypass piping, it should be very hot (65C) if bypassing is taking place.

Also just ensure that the main (red) circ pump continues to run during the anticycle time as should the UFH manifold pump.