Very slow leak in a pressurised heating system

[Dave]

Nearly a year ago I changed the old vented system to a sealed one, whilst also adding new rads and splitting it into zones. It all works well BUT the pressure falls VERY slowly, probably from 1 bar to 0.7 bar over about 6 weeks. I can't find any dampness around the rad valves or visible joints so have resigned myself to waiting for the leak to get worse before I can find it. Are there any suggestions on how to find the leak, or for anything I can add to seal the leak (probably not a good idea)?

The GlowWorm book suggests a minimum pressure of 1 bar, is this just to stop local boiling in the heat exchanger?

Dave

[nrh]

In news:dpjl8t$j5n$ [e-mail], Dave scribed:

Have you checked the pressure of the CH Expansion vessel? (Don't forget to depressurise the CH first) On my Vaillant, it should be between 0.7 - 0.9 bar. It should have a car-tyre type valve located on it around the base of the unit which you can check using a car tyre pressure gauge or similar.

Are you seeing any drips coming from the pressure relief valve outlet pipe, to indicate either that the valve seat is not closing tightly either due to grit or that the system doesn't have an effective 'shock absorber' because the expansion vessel has failed?

There is a good FAQ on this here:

[Fentoozler]

I take it you've got a pressure relief valve vented safely to the outside? Is this pipe dripping?

Alternitavely, I take it you've installed an expansion vessel into the system - has this got the correct charge (Around 0.7 bar cold)?

Angus

Angus

[Dave]

No drips. The expansion vessel is built-in to the boiler (30SXi).

Dave

[Fentoozler]

Auto-Air-Vent leaking (behind pump inside boiler)? You've got pressure leaving the system somewhere - check the tails on all the rads.

Angus

[Andrew Gabriel]

A leak that slow probably evaporates before a drip forms. You are more likely to find it with the heating off (cold), and the pressure increased to the hot running pressure. I suggest you leave this until summer when the heating is off.

1 bar will ensure pipework up to 30 feet above the pressure guage is pressurised, which is enough for most houses. In a 3 or more storey house with the pressure guage in the basement, you will need higher pressure if the pipework extends more than 30 feet in height above the pressure guage. (The pressure at 33 feet above the pressure guage will be 1 bar less than the guage reads.)

Also, the boiler probably has a pressure loss detector which will lockout the boiler if the pressure drops below something like 1/4 - 1/2 bar.

As nrh said, check the FAQ too.

[nrh]

In news:dpjqmn$ml3$ [e-mail], Dave scribed:

What do you mean by 'built-in'? If you mean it is an integral component of the boiler unit hung on the wall, then yes, so is mine, but it can still fail/leak. Can you not take a pressure reading of that expansion vessel alone?

As suggested in the FAQ, (and I have had to resort to this while I await the arrival on my new expansion vessel), try creating a temporary expansion 'vessel' from one of the least used rads on the circuit. You'll have to drain it of liquid first, and then shut off just one of the valves allowing one connection to operate as the pressure 'overflow' into the rad. When I did that, my pressure loss which was previously at around half a bar per day, almost stopped completely and now drops less than that in a week. ;-)

hth Nigel

[Ed Sirett]

Loads of background info in the Sealed CH FAQ. I doubt the relief valve is dripping but do check.

It will be almost impossible to find a leak that small. I suggest that you partially drain down and add some internal leak sealer. (fernox and sentinel do versions), this should seal the tiny leak up.

[Ed Sirett]

I don't think you actually have to shut off both valves. When you refill the water will enter the radiator about half way up. So you can still get some heating and you won't be sending any air round the circuit.

[Dave]

Fentoozler wrote: ... snipped

Thanks for that one, I'd forgotten about the auto vent. I'll check when I can next leave the system cold.

Dave

[Dave]

nrh wrote: ... snipped

Thanks for pointing that out (and for reminding me to look at the FAQ) but I can't see how adding additional expansion capacity can reduce leakage, do you mean that your old expansion vessel was faulty?

Dave

[nrh]

In news: [e-mail], Ed Sirett scribed:

Both? I have only turned off just the one.

Nigel

[Dave]

Andrew Gabriel wrote: ... snipped

... snipped

Sorry, I don't understand that explanation. The system is sealed, therefore it's all at the same pressure; how does atmospheric pressure (32 ft of water) affect anything?

Dave

[Dave]

Thanks Ed, it looks like the bottom line is that it's nothing too much to worry about; I'll add the sealer next time I need to drain down.

Another question: When I did the initial design I was worried that the internal expansion vessel wouldn't be enough for the size of the system (16 rads) so I was expecting to have to add an additional one. I was expecting to see a large increase in pressure from cold to hot but I've been surprised at how little it varies - maybe an increase of 0.2 at the most. Did I just overestimate the volume of water (and a small pressure increase is normal) or might there be something else going on?

Dave

[nrh]

In news:dpk980$7ks$ [e-mail], Dave scribed:

Yes, and still is as it remains in place until my new one arrives (hopefully tomorrow). It puzzled me too as to why it should improve although the 'leak' is still there, so I assumed one of two things is happening. Either the perforation in the inner airbag has been 'sealed' shut by the pressure of the surrounding water which I assume now fills the expansion vessel where the air would previously have been. Or, that the PRV (pressure relief valve) was opening under pressure when water heated up and was gently dumping water (and hence system pressure) out of the escape pipe leading to the outside. Although I could occasionally see drips hanging off the end, I was going to check that by attaching a polythene bag to the end of the pipe. But as the boiler is located on the first floor, it's a bit of a clamber up to it on a pretty long ladder. And I'm not good with heights! :-( So I just decided to live with the minimal loss of pressure now that the spare rad is taking the strain as a make-shift expansion vessel. Case of not looking the gift horse in the mouth, I guess!

Nigel

[Andrew Gabriel]

It's not at the same pressure. The weight of water increases the pressure linearly as you go lower down the system. It increases by 1 bar every 33 feet (and conversely decreases by 1 bar every 33 feet above any given point). The pressure guage only tells you what the pressure is at the height of the guage itself, but you can work out the pressure at every other point in the system by knowing its height above or below the pressure guage.

(The pump can generate a differential pressure on top of this of a couple of feet or so at different points in the system, but that's rather insignificant in the scale of things).

Well, you probably don't want the water at lower pressure than atmospheric at any point in the system.

[fred]

In article , Dave writes

Wot everyone said but also look for stains rather than drips, the leak rate you describe is very low and will likely evaporate before you see moisture on a hot system. Fernox and related corrosion inhibitors will leave a brown stain on evaporation & that is what to look for. I had a similarly slow leak for a very long time and eventually traced it to a very silly & obvious position but all I saw was the stain on a pipe, no sign of moisture at all.

HTH