Tracing tiny leaks in sealed system

Feb 16, 2007 32 Replies

Okay, several months down the line I'm getting sick of filling up and diluting the additive in my combi system. New combi boiler in the summer, changed from vented to sealed, fair bit of old pipework and rads. All the rad valves are new at various times within the last 8 years. I'm guessing there are leaks at the new pressure though nothing is visible and upstairs pipework is concealed. I suppose the leaks are so tiny that the water evaporates on the surface making it harder to trace. Maybe when the rads aren't on come the spring it will be more obvious. Any one know how I can track down the source?


Just a thought - is the pressure much higher than an open vented system with a tank in the loft?

Thinking of changing mine and wondered if some of the radiators would be at risk.

|!> -- |!> Mike W |!>

|! |!Just a thought - is the pressure much higher than an open vented system with |!a tank in the loft?

Yes pressure goes up from less than a bar to 1-2.5 Bar

|!Thinking of changing mine and wondered if some of the radiators would be at |!risk.

Only if they are very corroded. I used inhibitor in my radiators with the vented system and despite having to flush out a lot of black crud, the old radiators are fine.

If this was a high vacuum system, I'd pump it down, connect a mass-spectrometer, then spray helium round any suspect pipe areas. When you get a spike of helium coming through the mass spec, you've found a leak.

Unfortunately, this is probably not practical, unless you happen to have a source of high vacuum, a mass spectrometer, and some way to bake out the existing water in the piping.

In my previous career as an analytical chemist I could of at least borrowed a pump and sampled a makeshift port and taken in the sample to work. As a programmer I can't think of any software I could write to help me!

If it was originally an open vented system you could run it at similar pressure /with/ /the/ /same/ /boiler/.

If the boiler is designed to run at higher pressures and is run at maximum temperature kettling noises will be greater than when run at the higher pressure. This may be less true with condensing boilers due to the lower operating temperatures.

Auto air release valves may also leak, due to less sealing pressure.

Try adding a UV dye to the water, turn the lights out and go hunting with a UV torch.

I wonder if the stuff sold for PC modders would be OK

-- Adrian C

Or something really stinky. Ex-chemists should be good at this.

In message , visionset writes

Some kind of UV sensitive dye? Buy a forged note detector light, wait until dark then follow pipework looking for glowing patches.

In message , visionset writes

Not surprised, it's a hardware problem. For a change.

My bet would be, as someone said above, the pressure relief valve which normally vents outside the house. In my experience once they have lifted in anger they seldom re-seat. Not expensive or difficult to replace, just make sure you get the right gender on the inlet and outlet.

Drain the system then connect a car footpump and check with soapy water at all the joints?

Is it possible to put a food dye, or even better a fluorescent dye into the system, turn down the temperature and leave it for a few hours pumping water throughout the system by making heating and water demands?

My thoughts are that by reducing the temperature, you will loosen any slack joint and enable it to leak slightly.

Then go over the pipe work looking for the dye/fluorescent marker. (A fluorescent black light does not cost all that much. See Maplins site.) After you find/don't find any leak, flush the system out and use as normal.

I could be wrong, but I hope this helps.

Dave

Tie a clear plastic sandwich bag round the outlet and see. If there is a leak the bag will accumulate water.

If you've eliminated the PRD and are sure it really is a leak in the pipework then go round all the rad joints feeling for damp. If you're having to top up several times a year you may well feel dampness from the leak even if it tends to evaporate before it can accumulate and drip off visibly. Obviously test joints to the boiler and elsewhere also. Compression joints are the main culprits here.

As you suggest, doing it when system is cold makes it much more likely you will see the leak. Trouble is the pressure will also be lower, leak may reduce or stop, so increase the pressure up to the normal pressure when hot (system all turned off, obviously). Then go around a few hours later checking all the pipework, rads, etc for leaks. I installed a new system 5 years ago, and I've had two of the rubber O-rings on radiator blanking plugs fail, which I tracked this way. The leak wasn't enough that I needed to top it up, but it annoyed me there was any leak at all. Other places to check are the pipework inside the boiler casing, and hardest of all, a leak in a condensing heat exchanger, because you can't usually get to it. Check PRV isn't leaking.

To let the pressure down to normal afterwards, draw off water from a radiator bleed valve or drain valve. Don't use the pressure relief value as you can end up with debris stuck in it which makes it leak.

Remember to take the plastic tops of rad valves & TRVs and check for signs of leakage around the spindle , seepage can evaporate away before showing up on the visible body of the valve. Robert

To check rad valve spindles for leakage wrap a tissue around each one. The tissue may not be damp at time of checking but will generally show a telltale stain, left behind from water that has evaporated.

And if the valve is wet, a light blue tissue will show it up better than a white one.

Dave

Toilet paper - especially cheap stuff, will shrink and crumple when it gets a spot of water on it.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required