Weird result on LPG tightness test

[CND1962]

New LPG installation with 2x 19kg cylinders outside, plus change-over regulator with isolator valve on output, then 10 metres of 15mm copper pipe (3 soldered joints only) running into house to the kitchen, to gas hob (with another isolator just prior to hob).
System Failing tightness test. Dropping approx 0.2 mBar per minute. Hence drops 1 mBar over 4-5 minutes.
No leaks seen anywhere with leak detection fluid. No smell. New install so want to see "no perceptible drop"

During efforts to isolate the fault section and narrow down where it might be, the copper pipe that heads into the house was disconnected just after the regulator and before it goes into the house. This short section still fails TT!! So now we are TT on just a 200mm stub of 15mm copper pipe coming out of the bottom of the change-over regulator and its butterfly isolator valve on its output. See photo and white ringed area. This butterfly isolator also has the test access point. Isolator valve is closed during test. No leaks on the threaded joints.

Swapped digital tester for another and also tried with a mano….. see same dropping with these 2 other testers. Used another new rubber hose to tester too.
So…Swapped the isolator valve…… see same dropping.
remade the joint to copper stub… still drops.
Tried extensive spraying with leak detector all over this tiny area. No leak visible.

Any leak must surely be escaping either.. from the copper stub (nothing visible after extensive spray testing) and remaking the joint and the cap, to be 110% sure…. Or from the test access point on the side of the isolator valve , or the rubber tube to tester (have swapped it for new tube)….. or at some other point on the isolator valve itself, eg leaking up the shaft (nothing seen w spray on the shaft. It cannot surely be leaking back upstream because pressure is higher upstream of the isolator valve..and we can prove this by opening the butterfly on the isolation valve and test device sees pressure rise as expected. Plus we have swapped the isolator valve for new replacement twice now.

Starting to think that there is actually no leak…. Just something weird happening. Being heavier than air..Sinking?..into the stub.. Mixing with air, not fully purge?

can anyone explain this weird dropping. Its a tiny test volume so not ideal. There seems no visible contamination or flux etc, that might be generating a chemical reaction, inside the copper pipe section.

i have attached photo. The area ringed with the white mark around is the only section under test… eg half of an isolator valve and a short copper stub!

any ideas?

 

[ShaunCorbs]

Remove the yellow handle and check behind eg spindle

The female thread isn’t cracked bottom side ?

Also weather will effect you on a new install eg if just soldered the heat left in the tube will cause it to drop when testing

 

[CND1962]

Remove the yellow handle and check behind eg spindle

The female thread isn’t cracked bottom side ?

Also weather will effect you on a new install eg if just soldered the heat left in the tube will cause it to drop when testing

Thx, but tried all that. And extensively sprayed. No visible leak.
so then replaced the valve. Still dropping. so replaced again w different brand. Still same. yet cannot see a leak anywhere. No smell.

no temperature change over testing either.

 

[ShaunCorbs]

Have you just tested the pipework?

 

[CND1962]

Have you just tested the pipework?

Started off w TT on the whole system after install, hence right up to the hob. Observed the dropping. Despite spraying 2 tins of leak detector on everything right up to the hob we could not find a leak anywhere so we began to isolate sections working back to the cylinders. As stated above, we ended up disconnecting the copper pipe BEFORE it goes into house ie. just after this yellow isolator which is on the output of the regulator. So now the only section under TT is just the isolator and a short stub of copper. Still drops.

 

[ShaunCorbs]

yes as above but have you just tested the pipework no regulator ?

Could be the vent is passing

 

[CND1962]

yes as above but have you just tested the pipework no regulator ?

Could be the vent is passing

fuel is allowed through to fill, then the yellow isolator butterfly is CLOSED, so everything prior to the isolator valve (such as the regulator) is isolated from the section under test. So cannot be the cause of the (apparent) leak.

we have reduced the circuit to the point where all we have left under test is just an isolator valve and a stub. Nothing else. And we changed out the valve twice.

i do not think it is leaking. I think this is something to do with the behaviour of lpg.

 

[ShaunCorbs]

Doesn’t keep dropping don’t forget your short run won’t take much to drop it a mbar

Would air test it upto 40 mbar (the pipework) and see if it passes

Then air test the valve

 

[CND1962]

Doesn’t keep dropping don’t forget your short run won’t take much to drop it a mbar

Would air test it upto 40 mbar (the pipework) and see if it passes

Then air test the valve

hi again Shaun.
yes we have almost reduced the circuit to the point of being ridiculously small. That does bother me.

ALL the pipework passes air pressure testing. We separately tested the 10 metre section going into the house up to the hob isolator AND also this short section of isolator valve and 200mm section. BOTH were separately air pressure tested, at 1.0 bar for 48 hrs. Right up to the isolator valves at the far ends of both. No movement whatsoever over 48 hrs. Joints are just small number id end feed solder so I dont see scope where is could pass high pressure air test yet fail low pressure lpg tt

Despite this very high confidence in all pipework, still not comfortable walking away while we see a noticeable drop on lpg tt.

do you think that the small volume (after all its just an isolator valve plus about 6” of 15mm copper) is just too ridiculously small to perform a meaningful TT. Is it a pointless exercise on such a tiny volume?? What would be the effect if this tiny volume is not air purged and we are thus mixing in some fuel. As it mixes would you expect a pressure drop? As you say, on such a small volume it cannot take much to cause such a drop rate?

thx for your time

 

[ShaunCorbs]

So pipework fine did / can you air test the regulator ecv ?

No but to find a fault on yes as your such a small volume it can change a lot eg small escape could reduce to 0