*Lots* of air, every day, in CH radiators

[Rick]

We moved into our house a few months ago. The whole wiring and central heating we think are amateur jobs.

The problem is air getting into the CH system. It's heated by an old gas boiler downstairs, the pump is upstairs under the floorboards. Overall, the CH and hot water seem to work OK, except:

Every time the CH is started up after a few hours off, it gurgles like mad from the pump and three radiators quickly fill with air - two of them over half full of air (from feeling where they are warm/cold). It takes a good couple of minutes to bleed them.

Every time, twice a day.

All the other radiators are OK and these three are fine after bleeding. Until the next cold start.

Now, the dodgy wiring I can fix, I'm an electronics engineer. But water has me stumped.

Please, any help?

Rick, Manchester, UK.

[Ian Stirling]

Water is like electrons, it can neither be created or destroyed.

As is air. The volume in the system is (more or less (expansion)) constant. Air can be introduced through leaks in the pipework.

It sounds likely you've got a fairly substantial leak. You have wandered around the house and environs looking at all the exposed pipework, and seeing if you can see any damp bits?

[Rick]

Thanks for the quick response.

Well, my first job in the house (and first foray into plumbing) was to replace the water mains due to a leak under the building in the old lead pipe. Great fun! So I've tended to watch for leaks anyway.

I've not seen any, and given the volume involved I think it would be noticeable. Also, when I first got the system working (after connecting a vacuum pump to the pilot light to clear whatever was blocking it) we ran the CH full belt for 48 hours to dry the house out. Although it needed a little extra bleeding occasionally during this, it was only a tiny fraction of what we get twice a day from a cold start.

It's a concrete floor on the ground level and the pipes disappear into the walls or behind skirting - there are no discernable damp patches anywhere I can see.

I did have to tighten the compression joint on the inlet to the pump as it had a very slow drip, but that was a week ago.

I'm still stuck!

Rick.

[mike ring]

Rick wrote in news: [e-mail]:

This sounds a bit mad, but is it possible your pump is sucking in air, but only on startup?

I must admit I can't see why, except that when water is circulating, pressure differentials would be less than when the pump's trying to get it going.

Could you turn your pump down a notch, or check the vent on startup to see if it's sucking.

If this is a load of cobblers, please don't singe me bum

mike

[Markus Splenius]

Sounds like the pump speed is too high. You are probably sucking air in down the vent when the pump is switched on. Try turning the pump speed down.

M.

[Ian Stirling]

Ok, do you know where the header tank is?

Probably the easiest way to see if it's leaking a lot, is to turn off the system, for a couple of hours, then turn off the feed valve to the header tank, and see how fast the level drops.

The other possibility is if the header tank is overflowing, to somewhere you haven't seen, and you've got oxygenated water being constantly pumped in, as the system heats and cools, with the water 'leaking' out of the overflow (I don't think this is likely).

[Rick]

Hey hey, thanks to Google I now know what a header tank is. I wondered about what happened when all that water expanded...

I've just turned down the pump as suggested by Mike and Markus (thanks) - it was on full. I'll see if there is any difference tomorrow morning. I do recall a neighbour (who seems to know a little more than me) telling me that the pump was too small for the job - don't know if this is true or makes a difference. The pump seems to be a "Bosch UPS 15-60" if that means anything to you.

If not, I'll go and investigate the header tank and try your suggestion to see if the level drops. I don't recall seeing a cutoff valve up there (I had a nosey around the tank because I didn't know what it was for), but I can tie the ballcock up.

I know where the overflow for the header tank is and have not noticed any overflow, but I'll check.

I did notice a lot of corrosion around the tap on the inlet to the pump, around the joint, where the cap that the tap's shaft goes through joins the body of the tap. No sign of any water though - dry as a bone since I tightened that compression joint.

I'll get up in the loft and have a good play with it this weekend and report back.

Many thanks, Rick.

[chris]

hi

you could have any one or more of a few faults. installation fault: the pump if turned up to too high speed will pull large gulps of air down the expansion pipe from the header tank in your loft.blocked cold feed pipe from header tank will cause same symptoms. very unlikely to be air leaks as mensioned above from joints. another favourite is a failing washer on the suction side of the ch pump (may not leak water but can suck air in).

my advice: call an expert heating engineer.

[John Rumm]

I have seen this where the vent pipe that returns to over the expansion tank was about six inches from the suction side of the pump. Every time the pump fired up it sucked a gob of air into the system.

If this is the case, see if running the pump on a slower speed (if it has one) helps. Or move the position of the vent pipe.

[ab.usenet]

We once suffered from these symptoms - the cause was as simple as the ballcock valve in the header tank was sticking due to corrosion. I bled radiators to my heart's content, but of course the air just moved around and eventually stopped due to lack of pressure.

[Aidan]

"The pump seems to be a "Bosch UPS 15-60".

Grundfos UPS 15-60

" I now know what a header tank is. I wondered about what happened when all that water expanded..."

"I do recall a neighbour (who seems to know a little more than me) telling me that the pump was too small for the job -"

An important point. The heater tank is a feed and expansion tank. When the system is cold there should be just a couple of inches of water covering the outlet and about 3 gallons worth of air space to accomodate the expansion- I thinks that's in the FAQ. When the system's hot, the tank will contain more water, but ensure it doesn't reach the overflow. This may be relevant 'cos your knowledgeable neighbour suggests the system capacity is large.

When the system is hot and operating with the pump running, make sure there's no circulation of water through the F&E tank, no pumping over.

You've got corrosion problems, so reckon on power flushing in warmer weather. Probably a new boiler when the problem(s) is/are sorted.

[Peter Ramm]

what do you get if you burn hydrogen?

[Ian Stirling]

Helium.

[Rick]

Thanks to everyone for the help - I really appreciate it.

Turning the pump down seems to have cured the problem - no more air.

I'll put some anti-corrosion stuff in tomorrow and keep the suggestion of flushing in the back of my mind for the Summer hols.

This whole house really is a botch-job. I noticed when we first moved in that using the microwave and electric oven together caused the lights to *really* dim. Checking it out, I found the whole kitchen (including electric oven, sockets and lights) is run from a single piece of 2.5mm standard ring cable straight from the fuse box. Arrgh!

Thanks again, Rick.

[Peter Scott]

Another reason might be that the pipe that feeds water down from the header tank might be blocked by sediment. This happens to me about every five years. I am in an exceptionally hard water area and there seems to be no way to stop the sedimentation.

Peter Scott

[Peter Scott]

And of course now it would be legal for you to leave this as it is, but not to fix it without getting the installation tested! Politicians are plonkers !

Peter Scott

[Peter J Hemmings]

Just in case it has not been covered, I had a problem when the pump stopped. It syphoned over into the header tank. Improved it by balancing system which enabled me to reduce pump speed significantly. It still pushed a little over into the header so I had to raise the vent pipe (top bend) another 10cms which did the trick. Apparently it is due to system design (on my system), and was caused by the inertia due the mass of water being stopped abruptly when the pump stops. When this problem occurs the header gets VERY "sludged" up and can also cause a blockage (which makes the situation worse). Sorry if terms are not technically correct! I hope this may be useful as it took quite some time to find my problem!