Yet Another Cycling Forum
General Category => The Knowledge => OT Knowledge => Topic started by: Canardly on 22 December, 2018, 10:28:14 am
-
Any heating pros out there?
CH fully pumped with F and E and cylinder, is pumping over and driving me bonkers. New rads throughout last year, F and E tank cleaned out twice, new Alpha 2L 15/60 pump last year, cleaned system using Fernox F3, plenty of inhibitor, and Sentinel boiler silencer (was kettling). Pump speed 2 constant or lower variable constant pressure (running on 1 constant produces kettling). All Rads 10 rads of 14 panels amd two towel rails are fully hot, boiler is now quiet, water feed and air separator are clear and use off strong neodyme magnet produces no surprises in the cylinder cupboard, but system is pumping over significantly. Is not related to water cylinder as occurs on heating cycle and water is clear. Ball cock seems to be ok. Flow temp seems to be ok using food thermometer.
Is it time for DS40?
Hard water area.
Any ideas?
-
sounds like there is air (or gas anyway) in the system. This expands when it gets hot and pushes water out. When the pump runs if there is a significant pressure change (due to pressure drops in the pipe runs) in the air-containing part of the system then again the level in the tank can go up an down and/or you get pumping over.
So have someone eyeball the tank whilst the pump is cycled on and off (system cold) in the first instance. Then do the same thing with the boiler running.
If the level changes enough, and there isn't enough static head in the system, and the cold feed is badly placed, you can end up in a situation where the pump is drawing air in (eg through the cold feed) when it is running.
There may well be radiators that need bleeding out (and NB there is nearly always some air rather than no air trapped in the top of a radiator, so bleeding when the pump is running is different to bleeding when it is not). If there is active corrosion in the system then gas will be generated continuously; test the inhibition by using a nail in a jam jar full of progressively diluted sample fluid.
By shutting off both valves on individual radiators, you can identify which ones contain air and which ones don't, even if you can't bleed the radiator properly for some reason.
hth
cheers
-
Air is the most likely culprit, can you hear any air moving or bangs from the pump?
I've been fighting our system to get the air out and learned two lessons.
1) Only bleed the rads when the system is not running. Bleeding with the pump running can result in air pulled into the rad instead of expelled. A length of clear pipe with a few inches of water in it pressed to the vent revealed the water moving the wrong way.
2) If you have an auto vent keep it screwed firmly shut and only open it to let air out when the system is not running. Since I had air trouble I left mine open thinking it would help. In fact the auto vent was the source of the air in the first place (https://yacf.co.uk/forum/index.php?topic=46587.msg2353245#msg2353245).
Ours wasn't expelling much water but the expansion tank was full of warm water to the overflow so we must have been losing some.
-
Thanks both have bled the rads when cold not much came out. We do have an auto vent will look at that.
-
Have just read that AAVs can draw air in when pumping so have closed the cap on ours which is on the cylinder coil circuit (and which replaced a previous broken manual model). Thanks. Be interesting to see if that is the culprit.
-
Mmm still pumpiing over when boiler gets to temp. I am stumped.
-
Splurts, stream or gusher?
-
Dripping but significantly so at high temp. Boiler is set to 65 deg for efficient condensating and legionella. Don't think its a pinhole in the cylinder as water is clear and pumping over occurs when water heating is switched off. Dont think its the boiler stat as flow temps fine. Wonder if there is an air lock somewhere, but if so, lord knows where.
-
Please explain what pumping over is, cheers.
-
fluid coming out of the vent into the header tank = 'pumping over'
one of the things I have done (for diagnostic purposes) is to fit a blanking cap over the vent pipe, which I have previously modified to have a presta valve in it.
With the fill valve and air bleeds closed also, a good system can be subjected to either vacuum or (modest) pressure. This allows leaks to be found more easily and it also allows the volume of gas in the system to be estimated; the way pressure builds in the system (i.e. how many pump strokes) when it is pumped up using a bicycle pump can be compared with a bicycle tyre of known volume.
cheers
-
Thanks.
-
An airlock is still the most likely explanation. The restriction on regular flow resulting in back pressure looking for somewhere to go. Most of the system is in good, recent fettle so my money is still on that AAV letting air in which still needs to be cleared. What is the layout? For example mine is:
Floor mounted boiler on ground floor.
Single set of horizontals under 1st floor floor.
Pump upstairs almost at the highest point [1]
Drops and risers from the horizontals mostly to single rad up & down.
One instance of three rads off the same drop.
Three port valve.
Bypass valve at the tank. (on instruction from the plumber kept almost closed but I found opening it with the system stopped helped clear air from the return side)
Two downstairs rads permanently on to provide base load, the rest on TRVs
AAV on the HW side only 2
I would set the controls to HW only and then use the tank stat to (10/20 sec mark space) start / stop / start / stop / start the system. Open the AAV while stopped and repeat until you are reasonably confident the primary loop is clear of air. Next shut off all the upstairs rads that have TRVs and run the CH until all the downstairs rads are cooking. If any don't get close to boiler flow temp shut off the ones that do using the lock shields. Be mindful of where is the air you are clearing going, is it reaching the primary loop and getting to the AAV? Once downstairs is all flowing well set the TRVs to normal values and let the rooms warm up. Next open one upstairs rad TRV and lock shield fully and ensure the feeds to / from that are clear, stop the system and bleed that rad, bleed the primary loop. Repeat.
This assumes a system basically like mine. If you have horizontals below floors on 2 levels you need to treat both floors as upstairs above.
I've had a lot of practice clearing airlocks recently. Only since the penny dropped about the damn AAV has that done any good.
1 No longer true. Our upstairs pump being almost at the highest point was filling with air and the pump would gradually stall. I had to stop the pump, fiddle the 3way over to HW and then bang the pipes to get some air out. Then, start/stop the pump umpteen times until a reasonable volume of air had been expelled. I recently had a second pump fitted downstairs just above the boiler the theory being when the original pump stalled due to being full of air the new pump would keep things moving enough to clear it. This worked. I straight away got air in the only upstairs rad taller than the top of the pump & 3way arrangement. I had never previously got any air in any rad. The downside of the new pump was that even more air was getting into the system. Setting the TRV full on to the rad now collecting air made it collect lots of air. As soon as I set the TRV back to normal the air would gather in the pump again. That led me to concentrate on finding where the air was getting in. Previously I'd assumed the warm water getting pushed up to the header tank was becoming aerated. With the only upstairs tall rad now a player in the game I looked for pinholes in that. Eventually I started experimenting with the clear tube and found that rads suck if they are bled with the pump running and that the AAV sucks in more ways than one.
2 Got that part already.
-
Well........Still driving me batty. Closing all downstairs rads and maxing out the pump yesterday produced no pumping over initially and this for quite some time i.e. a couple of hours. All rads really hot. Eureka he thought but ........no. I opened a towel rail upstairs in order to ensure dry towels this pm and then the overflow started cascading like a good un, I mean lots. So something is happening. Debris in the sytem moving around perhaps? I turned pump down a notch and it reduced to constantly dripping slowly. Pump is on constant as against variable speed constant pressure. Boiler started kettling a bit this am with upstairs rads still closed.
System was well dosed with inhibitor so annoying to see it diluting daily. Incidentally, auto bypass seems to be operating properly.
This am. Have now closed all downstairs rads, whacked the room stat up a bit, all upper trvs are maxed out radiators nice and hot and no pumping over. Very very odd.
HW circuit seems fine and is set to operate once only a day.
Loath to drain down as it always causes lock out problems but may have to clean system again and put up with it.
Have opened and closed the AAV each time. Currently closed.
-
the pumping over you saw when the towel rail was opened is 100% consistent with said towel rail being full of air. When the valve is opened the trapped air gets hot and then expands, pushing water out of the system.
It might be that the towel rail (and/or other radiators) 'gather' trapped air/corrosion gas product that gets pumped round the system. This can happen more easily (even when said radiators are 'off') if the valves are plumbed in wrongly, i.e. the used valve/TRV is on the exit rather than entry to any given radiator. If the valves are correctly positioned the collection of gas (circulating in the pipework rather than being generated in that particular radiator) is restricted to the spur in the pipe run between the main feed pipe and that particular radiator.
Note also that under certain conditions the position and nature of 'T' pieces in combination with the flow rates can mean that you get a negative pressure on some radiators when the inlet valve is closed and the system is pumping. This can suck air in through open bleed valves and (of course) automatic bleeds can also suck air in.
cheers
-
Have bled all radiators when heating off and all are fine with none giving evidence of trapped air. Ho hum.
-
if there is trapped air somewhere in the system (and/or corrosion generates gas and/or air is drawn in somewhere...) then it will get pumped round the system until it finds a way to escape or it 'gathers' somewhere. It sounds quite likely that it 'gathered' in the towel rail for some reason, then expanded when it got hot.
The only other possibility is that once there was flow through the towel rail, air was being drawn into the system somewhere.
cheers
-
Well for the life of me I can't find any air trapped in system. have bled rads until I am blue in the face and rads glow in the dark. Boiler was a bit clanky until recently but boiler silencer has stopped that (it was repositioned last year and had a new pcb). but it does whistle a bit and we did power flush it last year and the condensate pipe split ( a known fault with Vaillants). Have ordered some x400 will try this for a month and then x800 if nesc. If I have a sub floor leak it is not showing itself. as yet. Ground floor circs are buried in the screed but are protected. I do not really want to go for a sealed system.
-
I've got a sealed system and it's never been a problem. Why don't you want one?
-
if (for example) the system is gradually drawing air in when running, and it is gathered in an unused/occasionally used radiator or pipe spur, the symptoms you would see could well be
a) when the system is running 'normally' there might be a slow dribble through the overflow ( corresponding roughly to the rate at which cold air is drawn into the system; there would only be a few hot bubbles at any one time)
b) when you open the valves to normally unused radiators (which may have gathered gas in the meantime, either inside or in the pipe spur leading to each one), there will be a sudden expansion in the gas (it will go from cold to hot) and this will push water out of the overflow.
If the system is merrily corroding internally, gas will be generated continuously and may give similar symptoms. Corrosion may produce hydrogen in which case the bleed gas from the radiators will certainly smell different and may be inflammable.
The system will draw air in if the pressure is low enough at the pump entry. If you can do, fit a 'T' near the pump entry and (temporarily) fit a clear plastic hose to this, configured as a second overflow. This will also operate as a pressure indicator (manometer) at the pump entry. If there is enough resistance in the pipework (which will vary with the number of radiators that are working at any one time) you may find that the pressure reduction at the pump entry exceeds the static head of water that is pressurising the system. This is quite likely to occur with an upstairs pump, and if it does the system may be drawing in air eg via the cold feed, auto bleeder, etc.
cheers
-
I've got a sealed system and it's never been a problem. Why don't you want one?
I meant a combi. A sealed system may be the answer.
Brucey there are a couple of redundant spurs where rads have been removed and sealed. Ho hum.
-
À la Brucey,When did you last dose the system with corrosion inhibitor ?
'Scuse the crap typing, I've consumed snaps.
-
Recently. It
has had 7 or so litres of Fernox MB1 in there.
-
7 litres, how big was the system.
-
MB1 is not concentrated and takes a 4 litre container for a typical system of 8-10 rads. I have 10 rads with 14 panels (one double rad being 2m long) and a couple of towel rails and of course system is pumping over. The idea of converting to a sealed system is becoming more attractive but even these, it seems, can have problems on older systems where an increased circulation pressure of 1.5 bar can find weaknesses.
-
Made an interesting discovery a few mins ago. Upstairs towel rad valve on side where lock shield would normally go starting coming off and letting by when when opening by hand. removing the flat cap resulted in finding that the valve is of two part design with a valve stem the inner nut of which was loose and adrift. Have nipped both inner and outer nuts up. Could this be the source of air getting in I wonder? We will see. Stand by to be disappointed. ;)
-
Fingers crossed.
-
ah, dodgy gland then....!
cheers
-
The saga continues. Still pumping over but possibly less amount spilling over. Yesterday, checked tightness of bleed points. Not all of them as I need a 22mm box spanner for some that have outer cases and I dont have one atm and cant find socket set :facepalm: . Turned all rads bar upper towel rail off initially, whacked up the room stat and ran up to boiler temp no pumping over. Boiler kettles, opens one more upper rad. No pumping over boiler quiet. Open another rad run heating no pumping over. I carried on with this running each rad up to temp and then moved on to ground floor. No pumping over. Whole operation say two hours. No pumping over bar a couple of drips. I checked the vent pipe in cyclinder cupboard each time I switched on another rad, was warm but not really that hot and cooler further up pipe. Cold feed was ok.
Heating comes on last night and yes, pumps over. What the ???
This morning significant water on flat roof below outlet.
Could the boiler stat (thermistors Vaillant eco max pro 28) be playing up? Digital boiler panel read out never reads more than 66 deg.
Could there still be some sludge in system? Rads really hot though but.
By pass is working fine.Air separator was removed last year for inspection no blockage to water feed apparent. AAV on HW loop after Motorised valves currently open. I will close it again.
Pump is currently set at lower variable speed and is only 18 months old.
No wonder heating engineers run a mile when told is pumping over.
-
to me what you describe is more suggestive of something (corrosion say) producing gas in the system over time rather than the system drawing air in when it is running (unless when running normally things are somehow different than in your test; Q. is there a motorised valve in the system that might be sticking?).
Does it pump over if you stop and restart the system shortly after your (successful) test run?
A simple suggestion; it is almost certain that there is gas in the system if you fire it up after a few hours and it almost immediately pumps over, where it didn't when you restarted it before. If you try to set light to the gas that bleeds out of radiators this may tell you if it is trapped/entrained air or gas evolved by corrosion.
However the test need not be conclusive; if the gas burns (BTW NB hydrogen flames do not look like other types of flame) it is definitely corrosion. However if it doesn't burn then there still might be corrosion.
cheers
-
That thought regarding a sticky motorised valve occured to me. I replaced a syncro motor to the HW water circ MV last year as it was buzzing, the other seemed ok. Moving to manual seems ok and they both close when required but as to wether they are both fully opening/closing I dunno. I could try replacing the other motor as I have a spare? Venting rads does not appear to result in any 'air' being released at all. I shall run both open on manual for a time tonight and see what happens.
Note although the replacment motors are marked as genuine 'Syncro' they are not the Honeywell versions which Honeywell state have a higher torque.
-
Two hours with both Auto valves on manual and not a drip (Well actually one drip from main tank overflow which is interesting in itself!). Cylinder stat is set to 60 Deg C so will be keeping overall temps down a bit. Boiler stat is set at 65 Deg C. I am now about to turn over to auto.....
-
A Vaillant boiler and pouring over to me sounds like a failed expansion chamber diaphragm. they are renowned for this failing and the cure is to fit a new boiler made by Worcester Bosch.
Been there, got the new boiler.
You may be able to fit a new external expansion chamber but basically vaillant are not good boilers.
-
It's a trad boiler so no expansion vessel Chris but it is a bloody mystery. IME Vaillants are pretty reliable but they have their foibles like most others. I wonder if there could be a pin hole in the cylinder coil as nothing seems to happen until max heat? Hot water is clear though but. We seem to be narrowing down to control gear and cylinder.
-
F and E overflow started dripping a few moments ago at 10:38 or so. Heating has been on all evening and appears to have stopped by 10:51.
-
It's a trad boiler so no expansion vessel Chris but it is a bloody mystery. IME Vaillants are pretty reliable but they have their foibles like most others. I wonder if there could be a pin hole in the cylinder coil as nothing seems to happen until max heat? Hot water is clear though but. We seem to be narrowing down to control gear and cylinder.
Sorry, so used to sealed systems these days.
-
River of Babylon apparent outside this morning. Sigh.........
-
apologies if this is b-obvious but is it possible that your header tank is incorrectly configured? Specifically
a) the ballcock has absolutely not to leak. Check the level in the tank (when the system is cold) and make sure that it doesn't slowly go up by itself. If in doubt shut off the cold feed to the header tank and compare.
b) there has to be an appreciable difference in the height at which the cold feed is shut off and that at which the water starts spilling into the overflow.
Even if there is no air in the system, the water expands slightly as it is heated. The difference between the fill and overflow levels in the header tank is there to allow for such expansion. If there is a little air/gas in the system too, the expansion will be more, obviously.
When you say the system is 'pumping over' I would normally assume that this means you see water burping out of the vent pipe into the header tank. However it now seems clear that you mean that you see water coming out of the header tank overflow, which can be caused by all kinds of things, not just a fault in the heating system pipework per se. So there are some things that perhaps have not been checked yet.
It isn't at all unusual for the system to lose water by evaporation during the summer, for the ballcock to crack open, but then leak afterwards when it should be closed, eg because there is a tiny piece of limescale or something on the valve seat. The flow rate through the ballcock never gets high enough for the valve to open fully and for the limescale to be dislodged, so the leak persists.
cheers
-
For what they cost I will swop out the ballcock and see if it has been letting by. Doubt this as I have measured water levels without heating on for several days and they have not moved. Outlet is well above normal water level.
-
If the Primary ( heating ) header tank is overflowing, then the water is coming from somewhere.
If the ballcock is not leaking, then the water may be entering the Primary system through a pinhole leak in the indirect coil in the hot water cylinder.
This will cause the water level in the small primary header tank to slowly equalise to the level in the larger Domestic Hot Water header tank, which will usually be higher.
-
If the Primary ( heating ) header tank is overflowing, then the water is coming from somewhere.
If the ballcock is not leaking, then the water may be entering the Primary system through a pinhole leak in the indirect coil in the hot water cylinder.
This will cause the water level in the small primary header tank to slowly equalise to the level in the larger Domestic Hot Water header tank, which will usually be higher.
Yep, this was the first thing our plumber said when we got an overflow from the CH header. But if it were the case it would be a continuous leak, not intermittent, no?
-
Have replaced the ballcock with a part two float valve as they are now known and bent the arm significantly. Float is new. Water level in tank is high atm. Reducer and washer on what I thought was CW compression feed connection had me stumped for a while until I spoke to a very helpful chap at the local plumbers merchants. The old float valve certainly looked its age.
A pinhold in the coil could be intermittent as it would not impact until the coil had heated sufficiently and then only if there were sufficent difference in head of CW tank and F and E tank. This would seem to fit the current scenario. However, whereas with a split coil water overlfow would run constantly. I have been told by a local heating firm that with a pin hole in the coil the hot water would be dirty and there is no obvious sign of this.
-
However, I have been told by a local heating firm that with a pin hole in the coil the hot water would be dirty and there is no obvious sign of this.
Shirley that depends on which way the water's flowing through the hole. Tip a pint of fluorescein in there and see what happens? :D
-
Pinhole in the coil will always result in tank (domestic hot) water flowing into the coil and raising the level of the (cold water primary) header, because the feed to the tank side is higher than the primary header. Why that would make the hot water dirty I've no idea.
-
I have been told by a local heating firm that with a pin hole in the coil the hot water would be dirty and there is no obvious sign of this.
As others have said, this is Just Wrong.
The standard mode of failure is as has been described.
The Domestic Hot Water ( secondary ) header tank is much larger then the small F+E tank for the Heating ( Primary ) system, and they are usually sat on the same platform.
The normal level is that the DHW tank water level is a couple of feet higher than the primary F+E tank.
Thus the hydrostatic head is usually greater in the secondary circuit.
This means that any flow through a pinhole leak is out of the DHW and into the dirty Primary system.
So no, you won't be seeing dirty Primary Loop water in the DHW.
-
I think it can vary with how the pumped circuit is configured and where the pressure drop is in it (in relation to any leak); the circulating pump can produce a lot more (or a lot less) pressure than 2ft of static head can. If the header tanks are about the same height you may get a situation where the thing leaks from one circuit to another, but only when the pump is running.
cheers
-
I think it can vary with how the pumped circuit is configured and where the pressure drop is in it (in relation to any leak); the circulating pump can produce a lot more (or a lot less) pressure than 2ft of static head can. If the header tanks are about the same height you may get a situation where the thing leaks from one circuit to another, but only when the pump is running.
cheers
But then there would be a nett loss to the primary, so no pumping over.
-
I think it can vary with how the pumped circuit is configured and where the pressure drop is in it (in relation to any leak); the circulating pump can produce a lot more (or a lot less) pressure than 2ft of static head can. If the header tanks are about the same height you may get a situation where the thing leaks from one circuit to another, but only when the pump is running.
cheers
But then there would be a nett loss to the primary, so no pumping over.
if the leak is in a pressurised part of the circuit, yes. But if it is the other side of a major pressure drop in the pumped circuit, the pressure could be negative vs the cylinder contents when the pump is running. In combination with a leak that changes flowrate when it is hot or cold, and/or a sticky valve in the system there is potential for all kinds of weird behaviour.
Best to assume little and keep an open mind about these things. Only one thing is fairly certain; the overflow will only run if there is more gas and/or water than there should be in the pumped circuit.
One way of checking for a leak between the pumped circuit and the cylinder is to shut off the feed from the header tank into the cylinder (there is usually a valve fitted at this point, if not the hot water header tank will have to be drained), then leave a downstairs hot tap slightly open. A bad leak will draw water from the pumped circuit and the level will go down.
Alternatively, partially drain down the pumped circuit and pressurise the pumped circuit (to low pressure with air as described upthread). An air leak (i.e. above the water level in the pumped circuit) will result in a rapid pressure drop and if the leak is in the cylinder coil t will will manifest itself as a stream of bubbles into the hot water header tank.
cheers
-
Well....Left F and E tank at its previous high level for the last few days and guess what, no overflow at all! :facepalm:
The float valve was the original suspect and the gaffa tape I had stuck to the side of the tank months ago to monitor the water level is still there, though now somewhat submerged . At that time, there was no sign of the water level rising when I switched off the heating and hot water for three days, and yet it now appears that it was the ball cock all along. Even a visit by a service engineer did not dispel my suspicion that the system was pumping over.
So a good result which has saved me considerable expense in:-
a) tracing what is actually going on
b) replacing parts that are ok ( e.g. new cylinder £600 plus)
c) not having to replace diluted inhibitor and boiler silencer.
So thanks to all for feed back and suggestions especially Brucey.
Moral of tale is should ball cock be suspect, then change it before considering other elements. Float valves only cost a few quid and are an easy DIY replacement. (make sure you have the appropriate washer for the feed connection.)
I should of course now be thinking of changing main tank valve as well.