Weird airlocks in HW cylinder/loose Sacraficial Anode?

[PeteH]

Hello all,
I have an ongoing problem with recurrent airlocks of all HW in traditional gravity fed indirect/open vented cylinder system.

My previous posts below refer (hope these show correctly):
//www.diynot.com/forums/viewtopic.php?t=22341
//www.diynot.com/forums/viewtopic.php?t=22196

Plumber called twice last week to look at airlock problem and can't understand it. He checked system - pipe outlet from top of HW cylinder full of air, hence no water flow. He drained system down/refilled it. It filled quickly and seemed OK. He checked loft tank - filled OK. Checked flow through pipe/gate valve from loft tank to HW cylinder - OK. Said pipes from cylinder to hot taps must be OK as good flow. Worked OK 4 hours then airlocked again!! He came back next morning - pipe at top of cylinder full of air again!! Drained down, checked same things again and both the bottom of cylinder and vent pipe into loft - not blocked. The cylinder is only 14 months old, but he thinks maybe the sacraficial anode (?) has dissolved/it's rubber fixing is rising to the top of the tank and blocking outlet pipe when hot. Some little black bits did come out of hot tap this weekend. Has anyone heard of this problem? Please help, we are just waiting for it to go wrong again/don't know what to do!!

He has diagnosed other (banging) problem as a faulty mid position valve, we are trying to get plumber that fitted that back to replace it.

Thanks for any advice you can give (apart from moving house!).[/quote]

 

[felix]

One rather obvious possibility is a blocked vent pipe. The trouble is, it's so obvious that I can't understand how your plumber could have missed it. Then there's the problem of how he ever got the air out of it. Unless you have a rogue plumber who keeps putting a cork in the end I think we'll have to abandon that one.

My next theory is that you have a small uphill section in your cold feed pipe. That's the one with the gate valve in it. I know this can cause trouble because an incompetent plumber left us with no hot water. The cold tank was full but the water simply didn't want to flow. It couldn't because there was a column of air trapped in the cold feed. Once again there is a flaw in the theory. How could you have been getting hot water for four hours?

I'm not really sure that I believe that sacrificial anode idea. I can't quite see a piece of rubber floating up and exactly blocking the outlet, not once but twice! And why doesn't it float in cold water which is denser?

There's one fairly conclusive test you can do but it's not for the faint hearted. You'll need to get your mouth around the open end of that vent pipe (or extend it with a piece of hose) and suck! Do this with all hot taps turned off and don't have the cylinder water too hot. If you can get water out you can at least eliminate the piece of rubber theory.

 

[felix]

I had a re-think last night and realized that the airlocked cold feed theory might just make sense. Let's face it, not much else does right now! The clue was in "filled quickly".

Here's what happened to us. Everything was plumbed correctly and with no blockages - except for this slight upward slope in the cold feed. It took me a while to figure it out because the pipe was hidden. The cold tank had filled and water would have run down the cold feed into the hot cylinder. Once the cylinder level rose above the entry point there was no escape for air in the feed pipe. It was trapped at the top. The tank filled up and the extra pressure squeezed the air in the pipe allowing water to get over the 'brow of the hill' and down.

So far so good but here's the snag. In a normal system the cylinder and vent fill until the water level matches that in the tank. In ours the cylinder/vent level fell short by the height of that trapped air column. It never even reached the top of the cylinder.

Now for your problem and a rather tentative theory. Your cylinder filled FAST. This means that some - maybe most - of the air could be swept out of the cold feed by the initial rush of water. (The ball valve would need a good flow to keep the water level above the tank outlet.) Result - the water level in the cylinder/vent is high enough to reach the taps.

Now what happens over time is that air, dissolved under pressure in your mains water, comes out and if it comes out inside the cold feed it has no escape. The trapped column gets bigger and the cylinder level falls until your taps run dry. The only problem I have with this theory is that four hours doesn't sound like long enough. But what else have we got?

If - AND ONLY IF - you can prove beyond doubt that your vent is not blocked (use the suck test) then you can try clearing the (presumed) cold feed airlock the same way we did. You cross connect a hot tap to a mains fed cold tap (don't tell the water board) and carefully open both taps. You really need somebody watching the cold tank while you do this. If they report an eruption of air from the tank outlet then QED.

 

[PeteH]

Thanks for your advice Felix.

I wondered about the vent pipe being blocked. However, the second time the plumber called I mentioned this and he said he had just blown through the vent pipe (?) and checked it wasn't blocked. Having looked in the loft before, I can't remember any raised sections of cold feed pipe between the loft tank and cylinder, but I will check this again, as what you are saying makes sense. What I can't understand is why we are having these problems now - the system worked fine from December 2003 when we had a new cylinder fitted, until last October when a new mid position valve was fitted (this being on the separate primary circuit). The hot water does seem to still be working since his second call (although a bit gurgly at times!), could this be because since then we make sure to leave the tank to cool slightly for 2/3 hours after water heated?

The only other theory I have, is that until now we have been using a plastic shower mixer that pushes over bath taps, but ensuring that we run the bath hot tap only (to avoid pushing mains cold water from the cold tap back into system). This fits quite tightly over tap outlet - could this have been drawing air into system somehow? The plumber didn't think so.

Another thought. We had previously cleared airlock by using this mixer to push mains cold back up HW pipes as you describe - between October/December 04, about 5 times in total. Last 2 times this has happened since new year this method hasn't worked - won't clear airlock. Does this give any clue as to cause of problem?

Any ideas much appreciated. If problem happens again I may consider having a mains pressure HW unvented system such as Megaflo fitted (b expensive though!). Are these any good and would plumber be able to use existing HW pipework to taps as he has suggested, to avoid ripping all the upstairs floors up?

 

[felix]

1) A warning about that suck test. It occurs to me that if your cold feed is blocked and you suck too vigorously you might just collapse the tank. My idea was that you would know immediately if you were sucking on a blocked pipe but this is a rather bold assumption on my part. Have somebody below watching carefully and if there is any sign of the tank sides moving inwards stop sucking.

2) When you try to blow out the airlock with mains water, the stuff must be going somewhere. Perhaps it's coming up the vent pipe, which poses a new question. Why can't it get up the feed pipe?

3) Just a thought but do you have a separate header tank for your heating?

I can't imagine how fitting that valve could have anything to do with it UNLESS something else was moved in the process. You might only need to push that cold feed pipe one inch up through the ceiling to create an uphill section in the loft. Check all your horizontal pipes with a spirit level.

 

[croydoncorgi]

How about something really silly - like a gate valve on the outlet from the cold tank in the loft supplying the cylinder which SEEMS to work but does not actually open.

 

[felix]

Yes, I thought of that too but it had to be open to get the cylinder filled - and it did fill fast. This does not rule out the possibility that somebody is fiddling with it. (Rogue plumber leaves gate valve almost shut perhaps.) Are there any inquisitive little fingers in the house?

 

[felix]

Just as I was dozing off last night I thought of this one (funny how often that happens):

1) You need to change a valve so you have to drain the heating system.

2) You go into the loft to tie up the header tank ballcock.

3) You accidentally stand on a horizontal bit of your cold feed.

4) It looks alright - no harm done - so you breathe a sigh of relief.

5) That pipe now has a barely visible dip in the middle.

6) Air bubbles forming in the cold feed are trapped at the dip.

7) Your hot water stops running.

It's a possibility.

 

[PeteH]

Thanks for the advice so far guys.

There is a separate header tank for the heating side of system, above the cold water tank in loft - that seems OK. I have checked that the gate valve on cold feed to HW cylinder is fully open and the plumber did say last week that this valve was working OK/not blocked. There aren't any inquisitive fingers playing with the gate valve (that I know of!), but there certainly do seem to be some gremlins in our heating system!

Having checked the cold feed pipe in loft, there are no obvious uphill sections. When the (previous) plumber fitted the new boiler (Sep 04) I think he did go up into loft/use a sling on tank ball valves, because after this he also fitted us a new internal mains water stopcock in the kitchen - the old one didn't work. I will check for any dents in the feed pipe to cylinder!! He didn't have to go into the loft when he replaced the mid position valve though (when problems started), as he used a gate valve that he had already fitted between the CH header tank and primary pipework - apparently because he thought the mid position valve may have been going wrong (how did he know that?).

The hot water does still seem to be working - nearly a week now since last plumbers visit - the only thing we have done differently is not used the plastic shower attachment on the bath taps. Is there any way this could have been the problem?

Any more ideas much appreciated. I will check that cold feed for dents and post another reply if any found!

 

[croydoncorgi]

Felix is perhaps in Australia! How can bubbles get 'trapped in a dip'??? You'd need to go some with very big feet to push a pipe DOWN so far as to create a 'high spot' somewhere else - where a bubble COULD form.

But if using a hand shower with push-on fixings, you MIGHT be pushing air mixed with cold water back up the hot pipe and causing an airlock that way.

 

[felix]

On re-reading that post I see I didn't explain it very well. Of course bubbles can't get trapped in a dip but they can get trapped in the pipe on the cylinder side of the dip. Their only way out is through the tank but to get there they would have to run downhill into the dip - and they won't do that. As more bubbles collect, the water level in the cold feed drops and so does the level in the vent until water no longer runs from the taps.