Short Cycling, any advice?

[Super_Paulie]

hello friends. Yet another "short cycling" thread sadly that id love some advice on. Not sure if its a) my plumbing b) my electrics or c) my lack of knowledge on it.

System

My system is a gas combi boiler Baxi 830 (or 800, cant recall) (30kw (or 24kw, id need to check the model), modulation 5:1, way oversized by plumber when i didnt know better), UFH downstairs and rads upstairs, on separate circuits via 2-ports etc. 4 loops, roughly 400m of 16mm, 2 zones (running as 1 zone currently)100/120mm centers in biscuit mix, 170mm PIR, flow rate via Salus auto-balance 2.5l-ish. On the advice of you wonderful gents i also fitted a buffer tank (50l) at the time of install with help from @Nickfromwales and @JohnMo, at the time i didnt have a clue about having to do this, i was uneducated on it (still am to be fair)
 

Buffer tank saga

 

I run the tank plumbed into the return as a 2-port and its only in play when the UFH valve is open. When the rad valve opens the tank is bypassed via a 3-port valve wired into the rad valve, everything's golden. I also run a small nearby bathroom rad as a bypass which runs independently as it is plumbed in before the 2x2-port valves.

 

Ok, so i can run my system at the lowest setting on the mixer of 38 and have it run 24/7 keeping the house a constant 20/21. Or i can run it as a timed system with a mixer at 40/45 to make the house 21 in the morning and then 22 for the evening with a setback of 18. I have tried both ways and the consumption is basically the same, when its very cold outside it costs about a fiver to do this, other times 3 or 4 quid, great (i guess).

 

Short Cycling
So, i assumed this was all working correctly, the boiler is in the loft so out of sight, out of mind. I went up there on a whim to time my cycle and found out it was short cycling. What it does is the boiler gets up to its target temp, flames off (pump still running/over-run?) for exactly 3 minutes while the temp drops, then flames on and repeat. It seems to do this no matter what i set the boiler flow to. When it fires up with a cold buffer it takes a good half an hour for the return temp to come up before the above happens, but when warm it cycles as below:

 

UFH mixer	Boiler flow	Return	Flame on	Flame off
		/buffer temp
38	40 (45)	34	0.47	3.00
38	45 (50)	35	1.40	3.00
38	60 (65)	49	4.20	3.00

 

after i ran it at 60 for a while i then turned the boiler flow down to 40 and it took a good half an hour for the buffer to empty while being pumped around the system by the boiler, satisfying the UFH demand. This i what i thought i was building (what id like to have?), the boiler fills up the buffer over 10 minutes then it takes 30 minutes to empty, and repeat. Obviously it doesnt, the boiler is always trying to satisfy the heat it is set at, duh, its a volume buffer not a storage tank.

 

So my question is, is this running as its meant to? i thought the addition of the buffer was meant to alleviate the short cycling, which it does but only if running from cold so its short lived. Im not sure what i was expecting but as the 3 minute flame off is always exactly 3 minutes thats obviously an anti-short cycling function of the boiler kicking in. I mean its only costing a few quid each day but it cant be good and is obviously a failing somewhere on my behalf. I have attached a simplified diagram below to maybe help explain my system.

 

What have i done wrong guys, do i just need more volume? run the mixer hotter to bring my return down? Unfortunately i am going through a redundancy process (sob story) so my build is on hold so id really like to see what i can about this if anything. Thank you people.

 

Edited 19 hours ago by Super_Paulie

[Marvin]

Where in the building is the buffer tank? In the loft?

 

[Marvin]

What causes the water to go through the Bypass radiator? temperature or pressure or ?

 

Where is the bypass radiator? in the bathroom or ?

Edited 19 hours ago by Marvin
clarification

[MikeGrahamT21]

Some boiler controls have a hardwired minimum burn period, I know I’ve read it for Tado which I have, and I think this was set at 3mins, so maybe worth checking which the manufacturer of your controls.

 

another thing you could try is seeing if you can turn down the pump speed on your boiler, I read about this whilst reading the Tado stuff and it can help with short cycles, with a few additional benefits, less pump noise, power consumption and presumably it’ll help it last longer. Most boiler pumps are preset for max pump speed from the factory, and with your oversized boiler, turning it a lot lower may be a really good idea

[Super_Paulie]

17 minutes ago, Marvin said:

What causes the water to go through the Bypass radiator? temperature or pressure or ?

 

Where is the bypass radiator? in the bathroom or ?

 

the boiler and the tank are in the loft. The tank is jacketed, plus i have it basically encased in a 75mm PIR tomb. The bypass rad is in the bathroom, its tiny but will be replaced on bathroom refurb time. Its operated from the flow and return before the 2 ports so its running through whenever the boiler is running.

 

Buffer
 

Edited 19 hours ago by Super_Paulie

[JohnMo]

What house thermostat are you using. If it runs on TPI (time proportional integral) that maybe driving the cycle. So basis of that type of thermostat is up to a couple of degrees of target runs boiler flat out, then it cycles to blip heat in to house in short bursts.

 

What UFH mixer did you use?

[Super_Paulie]

29 minutes ago, JohnMo said:

What house thermostat are you using. If it runs on TPI (time proportional integral) that maybe driving the cycle. So basis of that type of thermostat is up to a couple of degrees of target runs boiler flat out, then it cycles to blip heat in to house in short bursts.

 

What UFH mixer did you use?

 

Thats interesting, i'll google that now. I use Nest 3rd Gen for the stat, ( Nest stat )but when im running it low and slow i set the stat so its basically sat at a cut out temp of 25. It never reaches that so just chugs along and equals out at the temp i want, usually about 21. When im doing the timed heat i set it at what i want, 23 usually.

The mixer, im not sure what it is. Might be Reliance but id imagine its a cheap knock off version. It goes down to 38 and seems pretty accurate to that.

Edited 17 hours ago by Super_Paulie

[John Carroll]

 Would this work?, I'm assuming the boiler fires flat out in one case(s) and at its minimum output of 6kw? in the other case(s), depends on how fast the boiler ramps down once the target temp is reached after firing up. When/if the heating demand exceeds the minimum boiler output, then the boiler should fire continuously.

External Flowside Buffer Extract Rev0.xlsx

Edited 17 hours ago by John Carroll

[JohnMo]

Other things I found with my boiler.

 

A thing called gradient, manual had a default and just said nothing about it, had to read a few other manuals to find out what it meant. It is basically the acceleration rate, so how many degrees it adds to the flow temp per minute.  The default was 5 Deg/min. Moved to 3, and ramp up was way slower and boiler was able to balance the flow and modulation better. Not sure if you have this feature?

 

The other thing, is the bypass rad open to much and allowing too much hot water back to the boiler.

[Super_Paulie]

1 hour ago, John Carroll said:

 Would this work?, I'm assuming the boiler fires flat out in one case(s) and at its minimum output of 6kw? in the other case(s), depends on how fast the boiler ramps down once the target temp is reached after firing up. When/if the heating demand exceeds the minimum boiler output, then the boiler should fire continuously.

External Flowside Buffer Extract Rev0.xlsx 10.46 kB · 4 downloads

Once the target on the boiler goes below its temp setting it fires up, goes over temp and then the short cycle delay kicks in for 3 minutes. 

 

I just tried setting my mixer up 45 and my boiler to 35. Buffer is basically hanging at around 34. What happens is the boiler gets to 34.9, fires up to around 45, flames off and then cuts out, cycling for 3 mins then repeat.

 

1 hour ago, JohnMo said:

Other things I found with my boiler.

 

A thing called gradient, manual had a default and just said nothing about it, had to read a few other manuals to find out what it meant. It is basically the acceleration rate, so how many degrees it adds to the flow temp per minute.  The default was 5 Deg/min. Moved to 3, and ramp up was way slower and boiler was able to balance the flow and modulation better. Not sure if you have this feature?

 

The other thing, is the bypass rad open to much and allowing too much hot water back to the boiler.

Can't find anything in relation to gradient, but I'll keep looking. I've closed the bypass rad so it's barely on, hasn't helped sadly.

[JohnMo]

Other thing to try is, let the thermostat(s) start controlling and see if things are better or worse

[Super_Paulie]

11 minutes ago, JohnMo said:

Other thing to try is, let the thermostat(s) start controlling and see if things are better or worse

I previously had it as I guess timed, where the stat has set times and it just reached those set points and pulled the plug, keeping it going on and off to keep that temp. This actually works well but still cycles as it does so. 

I'm disappointed as I thought the buffer tank would stop that, or at least help extend the times. So having the boiler fire for like a minute then cycle out is sad and seemingly hasn't been worth the effort so far. I just end up with a tank full of hot water. I mean my bills are small(ish) but this is obviously out of control compared to what I had in mind.

[Super_Paulie]

Currently I've got my mixer at 45.

I've got the boiler at 55.

The return temp from the buffer is 41.

 

Flame on takes it to 59 and it messes about a bit up and down at around that, slowly creeps to 60 then it cycles, dropping to around 45 as it does so. Repeat. 

 

This feels like a problem I could solve, but I've got redundancy brain... 😕

[John Carroll]

44 minutes ago, Super_Paulie said:

Once the target on the boiler goes below its temp setting it fires up, goes over temp and then the short cycle delay kicks in for 3 minutes. 

 

I just tried setting my mixer up 45 and my boiler to 35. Buffer is basically hanging at around 34. What happens is the boiler gets to 34.9, fires up to around 45, flames off and then cuts out, cycling for 3 mins then repeat.

 

Can't find anything in relation to gradient, but I'll keep looking. I've closed the bypass rad so it's barely on, hasn't helped sadly.

All gas boilers fire up at ~ 65% of max output before modulating down, if you have a 30kw boiler then it will fire up at 30*65%, 19.5kw, if its set to say a target temp of 45C, then the burner will trip at target temp+5C, 50C, its very unlikely that the flow temperature on fire up will be less than 30C so the flowrate must be sufficiently high enough to maintain a dT of 50-30, 20C, (until it modulates down to its min output  or whatever) which requires a flowrate of 19.5*860/60/20, 13.98LPM, if the flowrate isn't at least at this level then it doesn't matter  what capacity buffer/volumiser you have installed, if you set the target temp to a more "normal" say 65C, then the flowrate only has to be 19.5*860/60/(70-30), 6.98LPM to exceed a burner trip temperature of 70C, ideally the flow temperature should just about reach target temperature so a flowrate of 8.0LPM required.  At a target temp of 60C, these two flowrates are 8.0 and 9.3LPM, at a target temp of 50C are 11.18 and 14.0LPM, etc. Also if the UFH manifold TMV is sluggish in operation, it might be returning a flowrate (to the boiler) well less than the above. Suggest, initially setting the target temp as high as is required to get the boiler away without cycling, then you can gradually reduce it to its optimum setting, obviously the boiler will still cycle when/if the UFH demand is less than the boiler's minimum output but this is quite normal and unavoidable.

 

Edited 14 hours ago by John Carroll

[marshian]

I might not be following but you've only got the volumiser (Buffer) in the underfloor circuit and one rad - not the rest of the rads??

 

I had huge issues with my glow worm boiler cycling (24kW - min output 10kW * see note) house heat loss after improvements 4.2 kW at -2.4 Deg C any warmer than that and it was going to cycle - shoulder season it was 3 mins on 3 mins off

 

What I found helped was the anti cycle setting (on glow worm it was also related to flow temp) - does your boiler have that parameter - because if it does you can basically force the boiler to wait longer between burns so the circuit cools more.

 

Obviously if the house isn't getting warm because of the anti cycle parameter then you need to reduce it but I found it very useful in managing the cycling issue

 

Note * on initial fire it was throwing 85% of the max at the circuit so if the circuit was small it was overshooting and shutting down

[MikeGrahamT21]

Consider how your thermostat is connected too. The gradient (unless I’ve got the wrong gradient meaning) which @JohnMo speaks of is relevant when the boiler is connected to thermostat via an analog method (on/off), but once you connect via digital (opentherm/bus) the thermostat can then take over, and it’s basically weather compensation, a heating curve which tells the boiler the flow temperature for a given external temperature.

 

how is your stat connected?

[JohnMo]

My boiler currently thinks it's running WC (fixed resistor) gives a solid flow temp. My thermostat just tells the boiler normal flow or setback. 

 

This a run the other day, after reducing flow temp a little. It was manually started and stopped when I got fed up.

 

 

The boiler is putting out about (min modulation) 6-7kW against a demand of about 1-2kW, but floor is buffering the energy. Running the same mode at 32 flow runs 6 mins on 6 mins off.

 

Will be moving to WC via Opentherm in a few days (driven by a new UFH wiring centre), if it doesn't do as good a job I will move back to simulated WC via on/off control.

Edited 12 hours ago by JohnMo

[Super_Paulie]

1 hour ago, John Carroll said:

Suggest, initially setting the target temp as high as is required to get the boiler away without cycling, then you can gradually reduce it to its optimum setting, obviously the boiler will still cycle when/if the UFH demand is less than the boiler's minimum output but this is quite normal and unavoidable.

 

 

I've had to call it a night up there, but before I did I set the boiler at 60, the mixer at 45. This takes a while to fill the buffer and the boiler creeps to 65.1 then flames out for 3 minutes, the buffer is at 51 at this point. After the 3 minutes the boiler is now at 54 and it starts up again. Runs back up to 65.1 in around 30 minutes. 

 

Is this still short cycling even with the long burn time and then 3 minute flame out? 

 

1 hour ago, marshian said:

I might not be following but you've only got the volumiser (Buffer) in the underfloor circuit and one rad - not the rest of the rads??

What I found helped was the anti cycle setting (on glow worm it was also related to flow temp) - does your boiler have that parameter - because if it does you can basically force the boiler to wait longer between burns so the circuit cools more.

 

Obviously if the house isn't getting warm because of the anti cycle parameter then you need to reduce it but I found it very useful in managing the cycling issue.

 

Correct, just on the UFH circuit. Rads are rarely used. 

 

I have no means to control that, I wonder if Baxi themselves can advise, I'll email them. 

 

House is toasty and not expensive to get it that way. Just the boiler is probably getting upset.

51 minutes ago, MikeGrahamT21 said:

how is your stat connected?

 

It's just on/off, coming from a Heatmiser wiring center. Not sure OpenTherm works with that, I'll have to check online.

[marshian]

8 minutes ago, Super_Paulie said:

I've had to call it a night up there, but before I did I set the boiler at 60, the mixer at 45. This takes a while to fill the buffer and the boiler creeps to 65.1 then flames out for 3 minutes, the buffer is at 51 at this point. After the 3 minutes the boiler is now at 54 and it starts up again. Runs back up to 65.1 in around 30 minutes. 

 

Is this still short cycling even with the long burn time and then 3 minute flame out?

 

If in 60 mins you have 30 mins of burn, 3 mins of flame out and then another 27 mins of burn (in the hour) I would not call that short cycling - but that's just me............

 

If you had 3mins of burn and 3 mins of flame out and that process is repeated throughout 60 mins so that it's 10 burns in that time I would call that short cycling and badly short cycling at that.

 

8 minutes ago, Super_Paulie said:

Correct, just on the UFH circuit. Rads are rarely used.

 

Ahhh explains it

 

My question is if the rads are rarely used why are you using an elevated flow temp much higher than the floor needs and mixing it down - why not set the flow temp much lower (or as low as you can get it - like 45) it makes no sense to me to heat the circuit to a higher temp than the circuit is set too

 

If the rads are required due to OAT being low then just increase the flow temp when the rads are required

 

8 minutes ago, Super_Paulie said:

I have no means to control that, I wonder if Baxi themselves can advise, I'll email them. 

 

House is toasty and not expensive to get it that way. Just the boiler is probably getting upset.

 

I'm not sure you need to do that based on my comments above

 

Incidentally I'm all rads no underfloor and my boiler is currently at a flow temp set point of 29.4 deg C, since midnight last night (or this morning depending on your viewpoint) with the heating demand being 24/7 with no setbacks the boiler has been on for 780 mins - it's fired 21 times today with an avg burn time 37 mins.

 

I have no volumiser or buffer as the circuit size is 130 Litres with 100 Litres being the volume in the rads and the boiler will modulate down to 3.2 kWh

[JohnMo]

Just having the UFH on only maybe the issue. Your boiler just has so little to work with to absorb the energy. Think you have two options 

 

Run everything on including all rads. This will give the boiler more surface area to work with, it will also up the flow rate. It will be a compromise to get flow temp low as you can so you don't overheat. That's where WC would be handy.

 

The other way is just on/off control. This utilises the floor as a buffer also. Basics are you allow the room temp to drop, floor core drops also, so your boiler has plenty of stuff to absorb energy. Then it's a matter of matching energy take by floor with delivery. This is done by slowly reducing flow temp until its cycling then increase temperature a bit. 

 

Your thermostats are maybe a little smart to do something as primitive as hysterisis mode.

[Super_Paulie]

20 minutes ago, marshian said:

My question is if the rads are rarely used why are you using an elevated flow temp much higher than the floor needs and mixing it down - why not set the flow temp much lower (or as low as you can get it - like 45) it makes no sense to me to heat the circuit to a higher temp than the circuit is set too

 

 

I did originally. Having the boiler at 45 and my mixer at 38 I'd get a cycle of 49 seconds flame, then 3 minute off. That's my problem really! I now know (from John C above) that I need to have my flow high enough to beat the boiler trip and modulate down. Currently at 60 with my mixer at 45 but that's just my first test. It's all confusing for myself as a mere mortal so far.

 

30 minutes ago, JohnMo said:

Run everything on including all rads. This will give the boiler more surface area to work with, it will also up the flow rate. 

The other way is just on/off control. This utilises the floor as a buffer also. Basics are you allow the room temp to drop, floor core drops also, so your boiler has plenty of stuff to absorb energy.

 

This makes sense to me. I could potentially just knock the TRVS right down on the majority of the rads and only open the ones that will give me the volume I need, like in the hall for example. 

The second suggestion is what I was doing previously before I tried the really low 24/7 method.

[John Carroll]

1 hour ago, JohnMo said:

My boiler currently thinks it's running WC (fixed resistor) gives a solid flow temp. My thermostat just tells the boiler normal flow or setback. 

 

This a run the other day, after reducing flow temp a little. It was manually started and stopped when I got fed up.

 

 

The boiler is putting out about (min modulation) 6-7kW against a demand of about 1-2kW, but floor is buffering the energy. Running the same mode at 32 flow runs 6 mins on 6 mins off.

 

Will be moving to WC via Opentherm in a few days (driven by a new UFH wiring centre), if it doesn't do as good a job I will move back to simulated WC via on/off control.

 

What make/ output Boiler do you have?, its quite a achievment IMO to get any gas boiler to fire up and stay lighting in one go with such a low target temperature, some boilers, like Vaillant maintain (or used to) ignition conditions for 60 secs for some reason or other making it even more difficult to get the boiler away and required a very long anti cycle time or/and a high target temperature.

[marshian]

5 minutes ago, John Carroll said:

 

What make/ output Boiler do you have?, its quite a achievment IMO to get any gas boiler to fire up and stay lighting in one go with such a low target temperature, some boilers, like Vaillant maintain (or used to) ignition conditions for 60 secs for some reason or other making it even more difficult to get the boiler away and required a very long anti cycle time or/and a high target temperature.

 

@JohnMo has an ATAG from memory

[marshian]

6 minutes ago, John Carroll said:

 

What make/ output Boiler do you have?, its quite a achievment IMO to get any gas boiler to fire up and stay lighting in one go with such a low target temperature, some boilers, like Vaillant maintain (or used to) ignition conditions for 60 secs for some reason or other making it even more difficult to get the boiler away and required a very long anti cycle time or/and a high target temperature.

 

My Heat only 16 kW 100-W Viessmann (so 3.2 kWh min output) when OAT drops to 0 deg C will happily run for 100 mins each burn at a flow temp of between 30 to 32 deg C

 

Return temp is 24 to 26 Deg C

[John Carroll]

1 hour ago, Super_Paulie said:

 

I've had to call it a night up there, but before I did I set the boiler at 60, the mixer at 45. This takes a while to fill the buffer and the boiler creeps to 65.1 then flames out for 3 minutes, the buffer is at 51 at this point. After the 3 minutes the boiler is now at 54 and it starts up again. Runs back up to 65.1 in around 30 minutes. 

 

Is this still short cycling even with the long burn time and then 3 minute flame out? 

 

 

 

So, what is the cycle time?, now, at 60C target temp, ie, burner ON time & burner Off time.

 

 

Can you post a link to your Boiler MIs.

 

A 3 minute anticycle time is very short, but probably set in stone on your boiler, here is the look up table for a Vaillant, (think Glowworm is similar, owned by Vaillant??) which is based on a anticycle time of 20 minutes at a target temp of 20C!!, the higher the target temp then the shorter the anticycle time. A default setting of 20C set anticycle time with a target temp of 60C gives a actual anticycle time of 6 minutes.

 

The problem with UFH is that if a high target temp is required to get the boiler away then the the return flow is lower, for example, a UFH demand of say 5kw with flow return temps of 45C/37C will have a flow rate of 5*860/60/(45-37), 8.96 lpm but the flow return to the boiler (assuming 60C flow temp) will only be, 8.96*(45-37)/(60-37), 3.12LPM. (With 5.84LPM at 37C mixing with that 3.12LPM at 60C to give a UFH manifold temp of 45C).

Vaillant anticycle times general.docx

Edited 10 hours ago by John Carroll