PDA

View Full Version : looking for a tiny gas leak



wigs
03-07-2008, 09:21 PM
Got a gas leak on an install, it appears to be losing about g/g every 6 months (system is only holding g). After recovering gas and leaving on 600psi pressure test for hrs there was no pressure drop I striped down this unit inside and out and could see no signs of any leaks. Customer cannot not have this system out of action so I will be adding the UV die and go and check over the system with the lamp in about 2 months. Just wanted to get people opinions on how they go about finding leaks that are very small like the one above.

wigs
03-07-2008, 09:22 PM
just noticed that spell check removed the amount of gas lost. it was 300g/500g over 6 months.

GXMPLX
03-07-2008, 09:35 PM
Armed with patience! ... it seems that ammount could be sniffed.

Turn off or close anything that may cause air movement inside.

Try to cover condenser and evaporator with plastic and sniff at the bottom, the hardest to find are there!

Take into account that pipes may also corrode so look everywhere.

Be carefull with Dye near food.

Good luck!

Makeit go Right
03-07-2008, 10:29 PM
Got a gas leak on an install. After recovering gas and leaving on 600psi pressure test for hrs there was no pressure drop.

If you find no leaks, after you remove the gauges and just before you pack your tools away, test the shrader valves with electronic tester. Schrader valves sometimes catch you out....you test the rest of the system and it's all sound. But the shrader valves might have a poor seating etc.

Worth doing this quick test after commissioning an install to save a call-back (and save the planet etc).

Like the plastic sheeting idea.

GXMPLX
04-07-2008, 01:23 AM
Yup, and replacing the cap of the shrader would also help!

chabby
04-07-2008, 02:23 AM
I have encountered similar problems like this , I figure out if it is a slow leak , its on the low side and if it is a fast leak , it is on the high side ..Isolate high and low side and try re testing..

GXMPLX
04-07-2008, 02:52 AM
I have encountered similar problems like this , I figure out if it is a slow leak , its on the low side and if it is a fast leak , it is on the high side ..Isolate high and low side and try re testing..

Not if it is a large leak in the low side (it would leak fast) or small one in the high side (that would leak slow).

refmech
04-07-2008, 03:43 PM
I have had several calls lately that I found to be scheader cores, with poor sealing caps! mostly brass caps with out deadsoft copper seats! man can this make a tech look bad.ran every one of these as call backs behind other techs.Also running a lot of cell cabinet self-contained wall hangers. Seems if there out of gas the leak is in sharp tubing bends, strees cracked at manufactures stamp. Very hard to find even with ultrasonic detector. good luck-take your time:)

wigs
07-07-2008, 07:36 PM
Thanks guys

Some good advise for where to look, but I was interested on your methods for looking for tiny leaks and patients seems to be the best method so far. I do like the idea of covering the condenser and evaporator with plastic to seal it so that it might be possilbe to be sniffed. I still think adding dye and leaving it for a month or two could be my best chance.

Peter_1
07-07-2008, 09:35 PM
Try to divide the setup in different sections where you then can pressurize the different sections. You will then notice faster which section drops in pressure.
You may no longer add refrigerant (EN378/2008) as long as you haven't found the leak.
So, your customer must allow you to trace the leak, even if it takes a whole day.

philfridge
07-07-2008, 09:55 PM
Hi this is quite a large leak 1/2 kg over few months. An electronic leak detector is your best bet as they can detect a 3g per year leak. As i see its a new install so you need to check all your joints first as the leak is likely to be on something you have done on the interconnecting pipework. This is because the condenser and evaporator come pressure tested as you probably know. As peter says adding gas to a system with a known leak is now illegal.

Peter_1
07-07-2008, 10:02 PM
This is indeed a big leak which can be found with a good electronic leak detector (sniffer, not an ultrasonic).

US Iceman
07-07-2008, 10:20 PM
Look for wet spots or areas that are darker than the surrounding area. If refrigerant is leaking then there is a good chance oil has leaked out also. If the leak has been like this for some time you might also find some areas that are fuzzy (especially around the compressor, condenser, etc.) oil tends to collect dirt from the air.

Any of these places are prime areas to investigate closely.

BigJon3475
07-07-2008, 11:57 PM
Maybe your overlooking the different types of leaks.

http://www.fileden.com/files/2008/6/18/1965139/BigBlue_Leak_Detection_Manual.pdf


If your sure about the schradders.....and all the joints....I would be looking at a temp dependent leak....most likely on the suction side since it will be the side that cools/contracts.

WINJA
09-07-2008, 07:17 AM
I found with small leaks like porosity or corosion when on systems using R134a leaks may not show up with nitrogen no matter how much preasure you use , I found the dye to be very usefull in this case eg it left specs of die on the outer radius of the evaporater bends where nitrogen was used previously and didnt show , but after seeing the dye i put big blue on the bends with r134a at 30psi and the bubbles showed as little cocoons after 5 minutes or so, be gentle when applying your big blue , apply it so it looks like a liquad not a foam

skags
10-07-2008, 09:22 AM
check around the terminals on the compressor.........

also i had a really annoying leak once and i was buying some dye from our supplier he showed me a product that you added to the system "for really hard to find leaks" and it filled them like that radiator solder you buy ,... i didn't really like the sound of the product as it didn't really sound like something i wanted to add to a system but i was wondering if anybody has tried anything like this before .. ? i am not sure of the exact name of the product but a quick search i found Cryo-Silane witch does a similar thing , has anybody tryed anything like this ?

BigJon3475
10-07-2008, 04:45 PM
check around the terminals on the compressor.........

also i had a really annoying leak once and i was buying some dye from our supplier he showed me a product that you added to the system "for really hard to find leaks" and it filled them like that radiator solder you buy ,... i didn't really like the sound of the product as it didn't really sound like something i wanted to add to a system but i was wondering if anybody has tried anything like this before .. ? i am not sure of the exact name of the product but a quick search i found Cryo-Silane witch does a similar thing , has anybody tryed anything like this ?

Super Seal....I personally wouldn't use it.

Read the manual and tell me if you would?
http://efficientcomfort.net/documents/Open_New_2006_Training_Manual_-_The_Complete_Guide_to_using_Super_Seal.html

SteveCass
11-07-2008, 11:11 PM
Don't they recommend using Helium for pressure testing on some systems now? Something to do with the smaller gas molecules on the newer blends (R410a)? The other thought I had is, do you use 6' gauge lines? Perhaps the gas you lose is from putting the guages on/ off?

Just a thought.

GXMPLX
14-07-2008, 04:07 PM
Don't they recommend using Helium for pressure testing on some systems now? Something to do with the smaller gas molecules on the newer blends (R410a)? The other thought I had is, do you use 6' gauge lines? Perhaps the gas you lose is from putting the guages on/ off?

Just a thought.

If you want to do this right, the system must fit in a vaccum chamber and you use a mass spectrograph in the outflow to the vacuum pump!

I don't see you can possibly do this in the field ... now a days!

BigJon3475
14-07-2008, 07:38 PM
A ball valve on the high side closed off and the pressure in the hose/manifold bleed back in through the low side would take care of that question.

Gapraja
14-07-2008, 08:59 PM
I agree with BigJohn regarding the sealant, had a nightmare of a job last week replacing the capilliaries due to blockage.

pendlesteve
17-07-2008, 01:45 PM
Did you comply with EN;378 when installing this?
If you did you would have pressure tested, leak tested and pulled a negative pressure. Any of which should have show a problem like this.
As has been pointed out, this a lot of refrigerant to loose. To knowingly leave a leak on a system contravenes the UK's Environmental Protection Act which carries penalties of an unlimited fine and or up to 5 years in prison.
Find the leak!