Pressure test

[RDB]

Does anyone know of a good method for testing pressure vessels and/or fuel tanks?

[Dutch Peter]

In my line of work we tend to pressure test fuel tank with air, at 0.2 bar and a lot of soapwater to find the leacks. By attaching a hose to the tank making a double loop and filling one side with water (2 meters) you'll get 0.2 bars.

Pressure vessels have, most of the time, a pressure test value on the manufacturer's label (it's something between 1.1 and 1.5 times the design pressure, depends on the standard used). Pressure vessels are most times tested with water, it expands less then air and therefor you have less of a bomb.
I've pressure tested a boiler for a steam tractor once with a high pressure cleaning device, it came up to 22 bars. For higher pressures go to a diving shop, they have to test breathing gear at 200-300 bars.

Succes,


Peter

[Thunderhead19]

A good place to start is recommended in ABYC 33.17. Are you still involved with Jeff and the boys over at Coastal? Almost everyone pumps their new tanks up with air (say 5 psi) with the ports sealed off. The tank is monitored for 24hrs for pressure loss. Installed tanks must be vented. There must not be blowback when filling the tanks and no spilled fuel should leak into the boat. You get to decide the pressure. Purpose specific pressure vessels are covered by CSA and you should consult them if your're manufacturing pressure vessels. Things like extinguishers, scuba tanks, propane bottles etc.. require hydrostatic testing to their designed pressure. I don't know if you have to go to the CSA lab to do this, but I think there are a few facilities in the lower mainland that will do it.

[mmd]

DoT regs for small craft require that a pressure test of 20kPa or 1.5 x maximum head, whichever is greater. The tank is to be charged (they do not specify with what, so air would be fine, but Dutch Peter's comments about building bombs should be paid attention to, especially if the tank is big) and fitted with a valve with a measurement range of not more than 3x the applied pressure. There is to be no pressure drop in 24 hours. They also want you to check seams for leaks (this seems counter-productive, but it is the government, right?) by applying a soapy compound over the seams while the tank is under pressure and checking for bubbles.

[RDB]

Ya, I'm working for jeff at Coastal. We are looking for a better method of checking for leaks in our fuel tanks. Often times the leaks are too small to register on our pressure gage. Every time somebody checks a tank over, they find a leak the last guy missed. They're small, so its tough to get them all. We use the soap and water test, but the welder asked me to hunt around for new ideas. Thanks everyone for your imput.

[DGreenwood]

I've come to extending a hose up from the fill inlet a min 8 feet usually more like 12 feet and filliling with water to the top. 24 to 48 hours will tell the story. I believe the CE regs now require this and it is not a bad method.

[Thunderhead19]

I have been toying with the idea of making a fuel tank with only one welded seam. This would reduce some of the potential for leaks. It's easy to make. Other than that, I'd get a more sensitive pressure transducer.