central heating air lock

[Stewart]

For the past 4 or 5 days we have had a problem with our central heating system; it is pumped with a valve for the heating and a separate valve for the hot water. There is one pump. All the controls work properly but they fail to either bring the boiler on heat or to keep it going. It appears to me that air is getting into the system and maybe into the pump so stopping it working even although it is still running. I have bled the radiators on each of the past 3 days and get some air out each time. There is no apparent leak in the system but of course there may be under the floor. I checked the header tank in the loft and it is up to level. Can anyone think of a reason for air to be getting into the system or how it can be doing so? Thank you

[Harry Bloomfield]

Stewart explained :

Try bleeding the entire system with the pump turned off.

Pump perhaps set to too high a speed and drawing air in via the vent going up to the tank. A leaking pipe joint, or valve etc. which is under vacuum with the pump running.

[Roger Mills]

The air could be a red herring. Have you considered the possibility that the pump could be running but not actually pumping because (say) its impellor has disintegrated or come loose on the shaft?

Does the boiler fire at all? If not (assuming an S-Plan system from your description) then the pump shouldn't run either - in which case you could have a problem with the zone valves, thermostats, programmer or wiring. If the boiler runs for a minute or so and then cuts out on its own stat, that's a sure sign of lack of circulation - in which case I would suspect the pump rather than air locks unless the system has recently been drained and refilled.

[Steve Firth]

It sounds unlikely that air in the system would be stopping the pump from working. I've had several systems where radiators have had as much as 50% of their volume occupied by air, and they still worked fine as far as the boiler firing and the pump working were concerned. The only symptom of air circulating in the system as far as the pump was concerned as the occasional "gushing" noise from the pump with a change in the pump noise as it thrashed the impeller in air for a time.

I'd look elsewhere for the problem. We had symptoms like those you describe with the previous boiler. The problem turned out to be a faulty optical sensor used to monitor the boiler flame. It decided the flame was too yellow (smoky) and cut the burner out at random intervals. But that's not an attempt to diagnose your problem, just an indication that even though all thermostats, timers and valves are behaving properly that there can still be a problem elsewhere.

Since it was old and there were other problems we replaced it with a new condensing boiler.

As to where "air" comes from, there are several sources. Yes leaks can admit air but often it's dissolved gasses from the water that cause the problem. Principally CO2 from hard water. Another source is hydrogen caused by galvanic action between steel radiators and copper tube. A plumber I once knew amused himself by setting fire to the "air" when he bled radiators.

[Stewart]

Thank you both; I shall check further to-morrow but let me add that this morning I bled air from most of the radiators and the pump etc worked fine with no "knocking". When we came home at 1930 the heating was off even although the pump was running and the radiators were just luke warm. I bled all the radiators again and a significant amount of air escaped. Once this was done the heating came hack on and the house is warm again. The pump is now around 15 years old and I wonder if it is possible to get cavitation in such a small domestic pump? There is no evidence of a leak anywhere in the house but of course it could be under the ground floorboards. Unfortunately my brother, who is a Corgi registered plumber, had just left to go on a 3 week holiday to South Africa so I either struggle on until he returns or employ another plumber to change the pump.

[Stewart]

Harry, if anything I am more puzzled than ever. As you suggested I turned off the pump and then tried venting the radiators again. Those on the ground floor emmitted a little air but when I tried the upstairs radiators I got nothing at all, not even water. I have taken some of the plugs right out and yet no evidence of water or air. I shall try to-morrow when I can monitor the feed cistern.

[Roger Mills]

You said earlier that you had checked the 'header tank' in the loft. Which one? You should have a large cold water header tank for the domestic hot water system and a smaller fill and expansion tank for the primary circuit.[1] It's the small F&E tank which you need to check. Because they need topping up so infrequently, it's not uncommon for the ball valve to stick closed. Then, when the level drops due to a slight leak or evaporation, the water is not replaced and air can enter the system. If nothing - not even air - comes out when you bleed a radiator, it could well be that there's no water in the F&E tank. Also check that any taps or gate valves in the feed pipe from the F&E tank into the primary circuit are actually open, and not turned off.

[1] Unless, of course, it's a sealed (non-vented) system - in which case there will be a filling loop and pressure gauge, and a probable need to re-pressurise the system.

[Ed Sirett]

Another thing which is a fairly common fault is that joint between the fed pipe and the main circuit plugs with lime/gunk and won't let much water join the circuit on refilling.