OT: Tank testing for leaks?

[Krunch]

If this is off-topic, I apologize. I wasn't sure where to put it, but I figured one of you boys would know the answer here.

I have an old galvanized 150-gal tank that I want to set up to use to have kero delivered to the house. Tank has been sitting here empty for 10 years. It looks in good shape. Only minor surface rust on outside in some places, hardly any rust inside, and dry inside (I looked with a flashlight).

At the bottom of tank, there was a female threaded piece welded on where I guess a nipple would go, but it was pretty wiped out with rust. I got a NPT nipple in there by a few threads, but it's kind of sketchy, so I ran a bead of 6011 around it to try to seal it up in case it might leak. I then put a brass ball valve with teflon tape onto the nipple.

Before I have 100 gal of kero delivered, I want to test for leaks, since if I discover it leaks after the kero is delivered, I'll have a big mess on my hands.

How would you boys test for leaks? I was thinking of somehow trying to plug up the fill nozzle as best I can with rags/visqueen/duct tape and blowing air in there with a compressor, and then putting soapy water on the nipple joint and anywhere else that looks like could be a problem...is this the best way, and are there any other tricks you could recommend?

Thanks for any tips.

Jeff

[ehgauss]

You should be very cautious leak testing a tank with air pressure versus water. The reason is that air, being compressible, stores a great deal of energy as pressure is increased. If your tank is not really sound, a failure could result it a major damage and injury. Water, being virtually incompressible stores very little energy when compressed and a failure, if it occurs, would be much less threatening. If you don't want to use water, which would give you a good test without pressurizing, keep the air pressure very low, and use a soap bubble check in the suspect areas. (Where I worked, a pipe fitter was killed about 45 years ago when a valve exploded while he was checking for leaks with air pressure.)

[enlpck]

If you really want to use air pressure, limit yourself to a few PSI at most for a cylindrical tank, and less than 1PSI for non-cylinder, and use soap solution.

I, personally, would use water at hydrostatic head. Dye the water with radiator leak test dye (fluorescent), let it sit, and look for leaks after a few hours with a blacklight.

Could also apply dye penetrant to the inside and look for signs on the outside.

[Jim-Tx]

I've tested a couple of fuel tanks that I built by using air and soapy water. Like others have said, use LOW pressure. What I did was just plug all of the threaded openings and use one fitting for the air line. I was able to find a small leak or two, but it might not detect a slow seep type leak. That one showed up after the tank had diesel in it and it wasn't too much fun welding on a full tank.

[Pumpkinhead]

if you can access the area(s) you want to test from the inside, mix some corn starch with water to make a slurry (kinda like latex paint) and paint the outside of these areas with it. when it's throughly dried, wet the inside of these areas with kerosene w/ a little clean motor oil mixed in, leakers will spot the outside 'paint' immediately, pinholes a lil' later, given enough time microscopic pores will eventually spot the 'paint'.
using hyrostatic testing is only gonna wet your nice dry tank, which you'll need to re-dry before filling.