[weldin mike 27]

Hi everyone,

Has anyone had to locate leaks in a mig gas shielding manifold system? This is a big system that serves about 50 outlets and runs high in beams of the building. They believe its leaking because the bank of cylinders seems to be running out quicker than normal.

They have walked around the whole shop testing outlets with soapy water (no joy) but I wonder if there is a more technical solution out there.

BTW this system has been installed for about 6 years but seems to be leaking now, (damaged somewhere?)

Hoping for info,

Mick

[Otto Nobedder]

Hi, Mick,

Unless you have a helium leak detector, the soapy water is the way to go. (More on soap solutions later...) I may have a few suggestions, however.

Is the cylinder bank you mentioned a rack(s) of High-pressure cylinders that get swapped out regularly, or a bank of liquid Dewers that get recharged? Or a bank of high-pressure that gets recharged rather than swapped out? I ask because cylinders of all types have relief valves built into them. If you're recharging existing bottles, rather than swapping out, you may have a faulty relief valve. It only takes one; With multiple bottles open, in the absence of independent check valves between the bottles and the inlet manifold, all bottles will back-feed to the faulty relief valve. The bonnet seals on each cylinder valve are also potential failure points. Are you certain the person responsible for the bottles has all supply valves fully open (for high pressure especially, they must be open tightly against the stop)?

Every joint, welded, threaded, compression, Swegelock, etc. is a potential suspect. Long, straight, tightly secured runs are subject to thermal stress. When many of those outlets are in use, the pipe will be cooler than the structure supporting it up in the rafters for a different expansion rate. (I realize it's winter in Oz, but it's had six years of thermal cycles.) Has anyone been on the iron to soap-test those connections?

Do you operate three shifts, or is there idle time when the system is shut down? If there's idle time, you may be able to narrow this down. I assume there is a master pressure gauge? By valving the supply bottles off, you can, with a stopwatch, establish a leakdown rate. Then, by valving off individual sections of the system, individual bottles, or whatever is appropriate for your setup and comparing the leakdown rate, you may be able to point to a particular section of the system.

While I have access to "Snoop", a Swegelock product for soap bubble tests, you can make it yourself. To one gallon of distilled water, add one tablespoon (15ml) liquid dish soap (I like Palmolive for this), and 1 teaspoon (5ml) glycerin (available at the drugstore). The glycerin makes the bubbles hold better, and slows the drying of the solution.

Good luck,

Steve

[weldin mike 27]

Thanks Steve,

I knew you would have some good imput. Im not in charge of the system, just trying to save the day. Its a high pressure system made of copper with soldered joins. The plumbers have checked the connections with the soapy water. However they haven't gotten up to check the lines. I think the people in charge need to start thinking out of the box. Only one 12 hour shift so plenty of downtime.
I'll keep you all posted.

Mick

[weldin mike 27]

Hey,

They got some plumbers to go along with an argon cylinder and pressurise each individual welding station, that way they found if any leaks were down stream of the wall mounted valves. They found a few leaks and replaced those hoses. also put bungs in the outlets that arnt being used, to stop accidental oppening. Forward progress i spose.

Mick