Central heating advice needed!

[Mandrake]

Hi All,

Needing a bit of advice about some problems with the central heating in our "new" 1930's bungalow before the winter really sets in... its an old system that has a few problems and needs to be nursed along to next summer on a shoestring.

I think I have a pump problem and posted a long post (yes who would have guessed! ) in one of the diy forums without a single reply! So I thought I'd try here as you're such a knowledgeable bunch.

Rather than reposting the entire long post I'll just post a link to it here, but please reply here on frenchcarforum if you have any ideas...

http://www.diynot.com/diy/threads/ch-pu ... es.443608/

I think I might have actually just figured out the issue (a combination of a leaky o-ring on the pump bleed screw, and the pump being incorrectly mounted at an angle) but it would be great to get a 2nd opinion!

PS if anyone has any recommendations for the best (most technical and helpful) forum for central heating issues that would be great too.

[Stickyfinger]

Hi, I'll try

I have only ever see these pumps mounted vertically, flow Up. Thinking about it, I have never seen them horizontal like yours....., it seems to me you have a particulate trap like that.

Have you drained down and refilled with clean ?
How often do you need to bleed the rads ?, when you do, how much air ? (oxygenated water will massively increase corrosion, do you have inhibitor in in ? (Fernox F1)

Look at Fernox pump care additive (ONLY after a couple of drains and refills, same with the F1, add together to a clean water fill)

[elma]

I'm no plumber but this interested me so I got googling. The 10th post here may give some explanation, clogged up feed pipe.

Are theses pumps strippable in situ? I wonder if you're pulling in air/steam or if it's just worn out.

[elma]

What Sticky suggests was my first thought as well. When I read the pump instructions it seems that they can be installed with the pipes in any orientation as long as the impeller shaft is horizontal. Still looks impossible to bleed properly though.

[Stickyfinger]

And what you say regarding checking the feed is as important.

If the pump is over 10years, well who knows on that.

Check, >clean, >fill with clean, >run, >if the same, try an additive, >if no improvement switch pump out.

[elma]

Agreed, plus change that o ring and bleed valve.

[Mandrake]

Hi, I'll try

Have you drained down and refilled with clean ?
How often do you need to bleed the rads ?, when you do, how much air ? (oxygenated water will massively increase corrosion, do you have inhibitor in in ? (Fernox F1)

When the work described (new valves, TRV's etc) was done back in February the plumber drained the entire system down, power flushed it and refilled with inhibitor. They were there all day from about 8:30 in the morning to about 4:30 in the afternoon (there was a lot of pipe work that needed doing too) so it was thoroughly flushed back in February.

Last week when I did some work I drained the system right down thoroughly (I didn't flush it though) and added inhibitor when I refilled it.

The radiators don't seem to need much bleeding, I check them maybe every 2 or 3 months and there are only two that have a little bit of air.

I'm no plumber but this interested me so I got googling. The 10th post here may give some explanation, clogged up feed pipe.

Interesting suggestion there but I don't think that's it. When I refilled the system I didn't notice any flow problems with the refilling - given that you're filling about 100 litres worth of water through what is basically a toilet cistern it didn't take that long, and during the filling process the flow through the ball cock couldn't keep up with the demand from the filling radiators - if the pipe was a bit blocked I'd expect the header tank to fill up and close the ball cock.

Are theses pumps strippable in situ? I wonder if you're pulling in air/steam or if it's just worn out.

If the isolation valves on either side of the pump are working properly (which mine aren't - they're seized open) you can actually remove the whole head of the pump with 4 allen screws without touching the pipe work - the heads are actually designed to be replaced without replacing the base with the pipes if you wanted to.

What Sticky suggests was my first thought as well. When I read the pump instructions it seems that they can be installed with the pipes in any orientation as long as the impeller shaft is horizontal. Still looks impossible to bleed properly though.

Unfortunately with seized isolation valves a full drain of the system would be necessary.

What Sticky suggests was my first thought as well. When I read the pump instructions it seems that they can be installed with the pipes in any orientation as long as the impeller shaft is horizontal. Still looks impossible to bleed properly though.

Yes the shaft is apparently supposed to be horizontal otherwise one of the end bushings can run dry - I think that's my main remaining issue so I'm going to try to loosen the nuts just enough to turn the pump to bring the shaft horizontal without draining the system down and hopefully without causing a new leak by disturbing the seal of the olives on the pipes...

Bleeding is no trouble when it is horizontal - remember you have a 5 metre head of water above you trying to force out any air as soon as you open the screw. It only dribbles out of the bleed screw on the pump though because it has to work it's way through the gaps in the motor armature. As long as it isn't completely air locked any bubbles of air will quickly find their way through the system and settle in one of the radiators. The bleed screw is there mainly to release a total air lock that would prevent the pump even being able to start to pump.

[Stickyfinger]

As they have power flushed then all OK. Same with the inhibitor as as you do not bleed often again OK.

Best change/look at the pump then.
http://www.judgeelectrical.co.uk/centra ... -pump.html

[xantia_v6]

I have used diynot a few times, sometimes with good responses, sometimes none.
Those pumps are designed so that the bearings run in water, and in normal operation there is no metal to metal contact, hence they "never" wear out. On our underfloor heating system in NZ, the pump (I think the same model as yours) ran continuously 24/7 (even in summer!) from 1995 until 2011 (the Christchurch earthquake).

When we moved back in after the earthquake repairs, it had been switched off for 18 months, and was seized. I gave the armature a knock and a spin with a screwdriver and it was back in business, but (in the interests of saving power) I added some timers and relays to switch the pump off when not required, but run it for 5 minutes a day in summer to prevent it from gumming up.

BTW, our pump is mounted with the pipes vertical.

If, due to the mounting orientation, yours has run with a dry bearing, it may have worn the shaft and bearing so that it has excessive clearance and won't run true any more. I that is the case, then a replacement pump would be required.

[Mandrake]

I have used diynot a few times, sometimes with good responses, sometimes none.
Those pumps are designed so that the bearings run in water, and in normal operation there is no metal to metal contact, hence they "never" wear out.

[...]
BTW, our pump is mounted with the pipes vertical.

If, due to the mounting orientation, yours has run with a dry bearing, it may have worn the shaft and bearing so that it has excessive clearance and won't run true any more. I that is the case, then a replacement pump would be required.

I see - so in many ways it is similar to a Xantia hydraulic pump where you just have bushes at the ends which rely entirely on the oil as a fluid bearing - let the oil run dry and the pump will make a horrible noise (I've heard it) and wear out very quickly.

This makes a lot of sense. With the piping being horizontal and the pump angled up about 45 degrees any air ingress through the leaking bleed screw o-ring or anywhere in the system for that matter would tend to collect in the "top" of the pump around the top bearing and bleed screw, when enough collected the top bearing would basically be running dry, metal to metal with no lubrication or immersion cooling for the bearing surface. No wonder it was squealing like a stuck pig! It wouldn't take much of an air pocket with the pump at a steep angle for the bearing to be totally exposed.

As well as air there is the matter of steam - a non pressurised gravity fed system like this is going to tend to generate a bit of internal steam at the top end of its temperature range since the boiling point of water is a lot lower than that in the pressurised systems. After a few hours running near the top end of its range the steam may build up and accumulate in the top bearing of the pump and then the bearing will run dry again. The pump doesn't care if its air, steam or hydrogen from corrosion, they would all cause the bearing to run dry in sufficient quantity.

With the pump shaft orientated horizontally the bleed screw end of the shaft is no longer higher than the feed pipes so air/steam won't tend to accumulate there, and even if a bit of air does build up inside the pump the air pocket will be to one side of the housing (up) so that unless a huge amount of air got in both ends of the shaft would remain under water with the air pocket harmlessly sitting at the outer edge of the armature.

I've just finished adjusting the angle of the pump - I was able to loosen the large gland nuts directly on the side of the pump without too much trouble and turn the pump around without loosing more than a small splash of water - I've set it exactly level with a spirit level on the housing. The before and after photos speak volumes about how far from level it was, it was actually closer to 45 degrees than 30 degrees:





I've put it back on test, so far it's completely silent, lets see how it goes over the next few hours!

Another thing I did last night is rewire the boiler so it runs from the wall thermostat together with the pump instead of directly from the timer - on the down side this means the boiler won't fire up to heat the hot water cylinder unless the wall stat is calling for heat, but in the winter that happens often enough that it makes no difference really, and in the summer we don't use the boiler for hot water anyway - we just manually switch the immerser on for washing dishes etc. (The shower is electric, so not needed there)

The up side is that when the wall stat cuts out the boiler will go off with the pump rather than continuing to heat until it reaches its cutout temperature, then boiling the water into steam soon after cutoff due to heat soak from the cast iron heat exchanger. Previously there was a good couple of minutes or more of vapour lock with the pump running screaming and dry every time the wall stat came back on due to the stationary water in the boiler heat exchanger boiling into vapour while the pump was off.

Later on I'm going to take the further step of fitting a pump overrun thermostat - all you need is a ~60 degree bi-metallic pipe thermostat with changeover contacts - wire the moving contact to the pump, the normally closed to the wall stat feed and the normally open right back to the main boiler switch before the timer. That way whenever the flow temperature is above 60 degrees the pump will keep running until it drops below 60, even if both the wall stat and the timer go off. Vapour lock should be eliminated completely then unless the flow temperature is set excessively high.

[wurlycorner]

Havent read all of the text, rather speed read it all, so apologies if what I have to say below is irrelevant/mis-read something. However, some random ramblings below...

I had exactly the same pump in the old back boiler system that was in my house (I've since fitted a condensing boiler). It was mounted vertically rather than horizontally.
I think the angle yours was at is probably not ideal, so having rotated it level is probably a good thing. That's just a hunch though - you could probably find the data sheet on the internet to confirm that?

I don't like the idea of the thermostat shutting off the pump and leaving the boiler running, at all! That sounds totally wrong.

My system had a 2 way valve that switched the flow between the hot water tank coil, or radiators. Does yours not have this? Presumably you are able to tell it to heat water rather than the rooms? Which would suggest it does???

If it does have said valve, is it sticking? Is there a thermostat on the hot water tank? Is that working correctly, or demanding heat all the time?

I had the 2 way valve stick on mine and the boiler just ran and ran (room stat was demanding heat, but the valve was sending it all to the tank). Pump got a bit noisy, which was the first clue (i think cavitation rather than vapour locking). Boiling hot water coming out of the cold tap was the second (hot tank vented into the cold tank, caught that just in time before it split and sent 700l of water through the ceiling, although running it back to cold again took some time, I can tell you...) However, it never ran the boiler without the pump operating.

Mine isn't a microbore system though, FWIW, standard 22mm main loop/15mm rad feed.

I also found that the system was quieter with an inhibitor in it, rather than just plain water.


I have the old pump sat spare in the shed btw, should you need one?

[Deanxm]

Havnt fully read so excuse any repetition, I am a plumber or was.
The pumps can be mounted at any angle but head up like yours was, the shaft loads the inner bush and air gathers at the bush in the head wearing it out, the bleed screw you mention controls end float of the impella shaft but also allows you to bleed air off the end bearing, do not opoen it while the pump is running as in some situations it can draw air in, although no big issue as it will collect at the rads.
Fitting trvs is all good but the pump is a bit big for your size system, I personally would go with a 15-50 but reduced floiw could be causing cavitation as the trvs close down, if the room stat is still calling for heat though the system will be trying to run to just one rad which can cause cavitation.

Your pump sounds like the bearing is a little warn and its spinning in air, this could well be due to residual air in the system.
Best way to be sure its not is shut ALL rads apart from the end of circuit rad on each floor, purge the pump of air by slackening a valve nut either side (system cold) of the pump and wait till no bubbles imerge, then run the pump and make sure the lockshields and trvs are full open on the two end rads and that you can hear water flowing through them.

This is typical of a badly designed micro bore system, they can be a pig to bleed.

Alternative is the pump has picked up some crud from the power flushing which is common or the diverter valve sticking.

D

[DHallworth]

Simon, if you're not having any luck getting it sorted, my brother is a heating engineer. I can put you in touch with him if you like.

David.

[RichardW]

Been away over the weekend.... my 2ps worth.

1. The pump is almost certainly knackered, but if it runs quietly now you've rotated it then I'd ignore it till it gets worse. New ones are not dear though. Unlikely it is sucking air in - there's will be best part of 0.5 bar in the pump even when it is running.
2. Wiring the pump through the wall stat and the boiler through the timer was std practice for a gravity primary / pumped secondary system. You don't need to worry about heat soak - the gravity primary will take care of that. If you want to improve the efficiency then you can fit a tank stat - wire the boiler off the common and the HW to the 'calling for heat' side, and the CH to the 'satisfied side'. That way the boiler will be powered when the tank tank is calling for heat, and when the CH is running, regardless of the state of the tank.
3. By the sound of it the boiler stat is on the way out too if you are getting 85C at the pump with it turned down - do you get kettling in the boiler when you turn it up? It's unlikely you are getting vapour in the pump - it wouldn't just whine it would cavitate and sound horrible before it got to that stage! More likely it is suffering a partial seizure (see 1 above)...!

Had the same sort of system in this house when we moved in, with a Baxi back boiler - was going to fit a tank stat but never got round to, and junked the system for a thermal store system about 8 years ago. Never had any problem with the CH pump on that, but the boiler has not been a paragon of reliability!

[Mandrake]

Hi Chaps, thanks for the feedback.

I had exactly the same pump in the old back boiler system that was in my house (I've since fitted a condensing boiler). It was mounted vertically rather than horizontally.
I think the angle yours was at is probably not ideal, so having rotated it level is probably a good thing. That's just a hunch though - you could probably find the data sheet on the internet to confirm that?

The installation instructions for the Grundfos pumps I have found (many seem to be those silly international picture only manuals with ticks and crosses...) clearly show that if the pipe work is horizontal the pump must be on it's side with the shaft horizontal, and definitely not with the shaft facing up vertical.

I don't like the idea of the thermostat shutting off the pump and leaving the boiler running, at all! That sounds totally wrong.

I agree, neither do I - whether it was "normal" back in the day (which Richard suggests below) I still don't like the idea, and I don't see why I can't adapt some more "modern" ideas to the control of the boiler, even if the boiler itself is old.

My system had a 2 way valve that switched the flow between the hot water tank coil, or radiators. Does yours not have this? Presumably you are able to tell it to heat water rather than the rooms? Which would suggest it does???

If it does have said valve, is it sticking? Is there a thermostat on the hot water tank? Is that working correctly, or demanding heat all the time?

I'm afraid this system is a lot simpler/cruder than your system was - there is no 2 way valve at all, the indirect loop of the hot water cylinder is not pumped, it's only thermo-syphon. The feed for the central heating goes from the front top right port on the heat exchanger to the pump, with the return coming to the front right bottom. Meanwhile the loop to the cylinder is pipes to the rear left hand side bottom and top.

I suspect that when the central heating pump is running that the vertical flow of water in the heat exchanger helps "encourage" the flow in the un-pumped circuit that shares the same heat exchanger, but with the pump off it will be entirely thermo-syphon, eg very slow and ineffective.

There is no stat on the hot water cylinder - at least for the indirect heating part. The immerser element has it's own thermostat but that is independent of heating the hot water from the boiler where you just get hot water at whatever temperature the boiler is set to!

The pumps can be mounted at any angle but head up like yours was, the shaft loads the inner bush and air gathers at the bush in the head wearing it out, the bleed screw you mention controls end float of the impella shaft but also allows you to bleed air off the end bearing, do not opoen it while the pump is running as in some situations it can draw air in, although no big issue as it will collect at the rads.

I think that's exactly what has happened - for various reasons (air, vapour etc) an air pocket was forming at the bleed screw end bearing causing it to run dry. Also because the heat exchanger was vapour locking when the stat turned the pump off (but left the boiler running, trying to maintain the flow temperature in the heat exchanger without any flow) the pump was suffering from 2-3 minutes worth of vapour lock and dry bearings when the stat next came on, it was taking that long for it to clear itself and get pumping again. During that time the bearing would have been overheating and wearing out. This could have been going on for years for all I know as it was already doing that when we moved in.

Since I turned the pump around to exactly level it has been behaving much better - just before I turned it even with at least one radiator open all the time (bypass radiator) and the boiler stat set to about 65 degrees it would run OK for a couple of hours then suddenly burst into a loud squeal - that has not happened since I turned the pump around. Also now that the boiler is wired to go off with the pump the vapour lock after the stat cycles off and on is greatly mitigated. Previously it would sit there vapour locked for several minutes sometimes, now it whines for maybe 5-10 seconds then clears the lock and gets on with things. Still not ideal but a big improvement. Adding pump overrun should eliminate what's left of that problem completely I think.

Fitting trvs is all good but the pump is a bit big for your size system, I personally would go with a 15-50 but reduced floiw could be causing cavitation as the trvs close down, if the room stat is still calling for heat though the system will be trying to run to just one rad which can cause cavitation.

The pump is actually a 3 speed pump. I don't know what speed it was set to when we first moved in as I didn't even realise the pump had a speed switch at the time - the guys who came in to do the TRV's either left it at full speed or set it at full speed, which was causing quite a bit of noise around the system, and I think was causing a bit of cavitation.

I think like you say, the full speed (6 metre head ?) is a bit much for the system, it's a bungalow with a converted loft, but it is quite a high roof with the cistern right at the very apex of the roof, so there is probably a 5 metre head of water in the system. The lowest speed is too low to remove heat from the boiler fast enough while the highest speed does cavitate a bit, it seems fairly happy at the middle speed, which is where I have left it the last few months.

Alternative is the pump has picked up some crud from the power flushing which is common or the diverter valve sticking.

By diverter valve are you referring to a valve to divert the pump between the heating loop and the hot water cylinder ? If so it doesn't have one, the pump only feeds the central heating with the cylinder being only thermo-syphon. Or do you mean the automatic bypass valve ?

In which case I think it's possible it may have got a bit of crud in it when I drained the system down because I have noticed since then that adjusting it doesn't really do anything any more. Previously if you deliberately closed all the radiators and adjusted the bypass pressure (0.1 - 0.6 on this valve) you could clearly hear a difference in the noise of the pump as the differential pressure changed, and you could also judge by turning the screw where it started to get tight, how high the pressure was. Now it doesn't seem to make any difference which makes me think it might be stuck partially open.

That might also contribute to the problems as it would mean even with the bypass radiator open some of the return flow would be short circuited back to the boiler without any heat being removed...If I change the pump I think I'll change the ABV as well or at least remove it and try to blow it clear.

1. The pump is almost certainly knackered, but if it runs quietly now you've rotated it then I'd ignore it till it gets worse. New ones are not dear though. Unlikely it is sucking air in - there's will be best part of 0.5 bar in the pump even when it is running.

As above, the pump is much better now that it's running horizontal, but I think the damage has already been done to the bearings - it runs near silent when cool but once it gets to about 60 degrees it starts to get audibly noisier and noisier with a dry bearing sound, even if it's not the loud scream from when it was at 45 degrees. Not only that, I'm only running the boiler stat at the minimum at the moment while I get to the bottom of the various issues - it won't cope with running at the higher temperatures a cold winter will need without making nasty noises, so I'm thinking seriously of replacing the pump at the end of the month - the bearing has been put under strain for probably years with a combination of poor mounting angle and wall stat induced vapour locking...

2. Wiring the pump through the wall stat and the boiler through the timer was std practice for a gravity primary / pumped secondary system. You don't need to worry about heat soak - the gravity primary will take care of that.

But it doesn't - if by gravity primary you mean the thermo-syphon hot water cylinder, it does nothing to prevent vapour lock - until I rewired the boiler to go off with the pump a couple of days ago every time the stat went off if it came back on in 10-60 minutes or so afterwards you'd get a massive vapour lock of the pump with it screaming out with no water in the bearings and no ability to pump. It could take many minutes to clear this vapour lock. There just isn't enough flow in the thermo-syphon system to remove the excess heat from the heat exchanger after the pump goes off to prevent the water boiling in the heat exchanger, causing a vapour pocket. Maybe the pipes to the cylinder are blocked up, but it seems unlikely when they are 28mm and the 8mm microbore around the house is not suffering from any blockages that I'm aware of.

If you want to improve the efficiency then you can fit a tank stat - wire the boiler off the common and the HW to the 'calling for heat' side, and the CH to the 'satisfied side'. That way the boiler will be powered when the tank tank is calling for heat, and when the CH is running, regardless of the state of the tank.

I could do that but when the pump is not running the boiler is far to slow to heat the hot water. The immerser can heat the tank from cold in about 45 minutes, the boiler with the CH pump running can do it in about an hour but when the pump is not running it was taking a couple of hours at least (with the boiler flame off most of that time due to the cut out temperature being reached) to heat the water. Presumably due to insufficient flow. If the loop to the hot water cylinder was pumped I'm sure it would have been much more effective.

Been away over the weekend.... my 2ps worth.
3. By the sound of it the boiler stat is on the way out too if you are getting 85C at the pump with it turned down - do you get kettling in the boiler when you turn it up? It's unlikely you are getting vapour in the pump - it wouldn't just whine it would cavitate and sound horrible before it got to that stage! More likely it is suffering a partial seizure (see 1 above)...!

With the stat set right down the average flow temperature measured at the pump (IR gun pointed at black tape) is about 65 degrees, which seems a bit high to me for the minimum setting. Even on this minimum setting it sometimes peaks as high as 75 degrees. So yes, I think the stat is starting to misbehave and getting out of calibration. It has a sensing capsule pushed down a long hole (about 3-4") into the top of the heat exchanger, thinking it might not be making good thermal contact I pulled it out and put a little bit of copper grease on it, that seems to have made it a bit worse for some reason. Grr....

If I turn the boiler up I can get to about 80 degrees before I start to hear any noise from the heat exchanger, it starts to sound a bit like the hiss of steam, I wouldn't call it kettling though which I presume would sound like the early stages of bubbling a kettle. Above about 85 degrees (measured externally on the pipe, so probably a bit hotter inside the heat exchanger) you can hear it start to bubble and the pump will then suddenly vapour lock and start screaming. So the boiler stat is definitely letting it go too hot if you turn it up.

As for cavitation, I have observed that too at considerably lower temperatures, possibly as a result of air getting in (from where ?) where it will sound like cavitation from the pump, but no noise at all from the heat exchanger. Not sure whether that is air or vapour that is causing that, it only seems to happen after the system has been running for several hours.