My hydrostatic tank test rig & result

I figured I would share this in the hopes it will help others. I found a lot of discussion on hydrotesting compressor tanks that was somewhat lacking information or simply referred to using a grease gun to accomplish the pressurization.

I have an old horizontal 80gal tank that was an IR T30 compressor, mid 80's vintage. I figured I was going to build a new compressor around it and should test the tank before proceeding.

For my test I left it horizontal, but lifted one end with a stacked pair of 4x4 scraps so that the 3/4" NPT hole from the check valve was the highest point. I filled the tank with a hose until this was full.

First attempt to pressurize was done using a cheapo grease gun (typical task-force type deal) which would not seal the plunger to the gun to even hold water. Second attempt was with a nicer grease gun which kind of worked with water vertically but didn't really work that great.

So I changed direction and built a pump out of a cheap 2 ton bottle jack I bought at Northern Tool for $13.

I used a 3/4" NPT to 1/2" NPT bushing, a 1/2" NPT to 1/4" NPT bushing, and then a 1/4" NPT check valve from McMaster to fill the hole from the original tank check valve.

From this I used two 18" grease gun hoses (Northern Tool) which happen to have male 1/4" NPT ends, and a 1/4" NPT coupler.

The bottle jack was modified thusly:
The screw to release the piston was removed, and a small 1/4" NPT ****** and coupler were used to attach it to the grease gun hose. The pump handle and pump fitting were removed, and replaced with an 8" 1/2" NPT ****** and 1/2" NPT ball valve. Then from the top of the valve I used a barb fitting and some tube which I kept topped off on water. I then placed the bottle jack in my shop press, and used this new 1/2" valve to fill the jack with water.



Basically I did the following:
Fill tubing with water
Close hand valve on bottle jack
Extend (pull) piston out all the way
Open hand valve on bottle jack - water level drops as jack is filled with water.
Close hand valve on bottle jack - allow to pressurize water out of old screw hole
Push piston back into jack.

Rinse and repeat

I was able to pump the tank up to about 70 PSI with water just by hand operating the piston in the bottle jack. Past that I started having to use the shop press to compress the piston. I took it up to about 190 PSI.



At about 180 PSI the vent/relief on the bottle jack popped, and I had to close the hole up with an improvisation:


Overall this worked really well.

Unfortunantley, the tank failed the hydro test. A small dripping leak appeared at the bottom of the tank between a leg. This did not appear until about 180 PSI. The tank was ASME rated and had a tag stating it was 200 PSI working pressure.


I'm glad I tested it, and since it failed it will now either be scrapped or turned into some form of interested metal sculpture. Maybe a bomb or airplane??

Nice! I like the creativity there.

Yes, I too like someone who thinks outside of the box. One time long ago I took a condemned tank and with an oxy/acet. Torch I cut one head along the weld seam you see in your picture. Only cut half way around the tank. (90*) then make a right angle from your cut around the tank, cut about 12" up toward the other end of the tank. Then cut around the tank to the other side. The head should fall off. Now to imagine what you have is think heavy duty BBQ grill. Turn it so your ball valve is down and weld three legs on and a cover over the top. Get a wire grill from the BBQ that rusted out on you and now you have one that will last a hundred years

Nice work, first real hydro testing I've seen.

A failed tank is to good to scrap out. May be you can find an idea or two here:

http://www.colinselig.com/home.html

why were you trying to pump water with your grease gun. i never understood why people try that. i know its possible and all, with a good grease gun.

but my point is, why not just pump grease into the damn tank? you are only going to pump a small amount in, and it wont hurt anything what so ever. not to mention how much better a grease gun works when its pumping grease instead of water. you could even use high viscosity hydraulic oil.

a little grease or oil residue wont hurt your tank. plenty ends up in there from the pump anyway. the very little that would remain will be gone after a couple water drains after the compressor is up and running too.

i wont even go there about running water through your hydraulic jack...

the OP did a great job with his rig to test for leaks - only. True hyrdostatic is a different animal all together. I don't want other people getting the idea that this method is a replacement for the industrial hydrostatic testing done on cylinders and vessels to qualify them for safe use.

Hydrostatic_test

read the parts about test pressures and vessel expansion. that is a critical part of the test as well as leak detection.
please be careful if you choose to do this.

I have a scheme with my pressure washer

Good job, but look at the trail from where the drip came from. From the picture, it looks like the underside of the tank is wet from much higher to the right and the water ran down to the drop you see hanging. I wonder if the tank was leaking or that came from a hose connection on the other end.

Either way, glad you showed us a good way to test tanks, hope your tank works out for you!

I have heard about the pressure washer idea.
But never any details.
How do you hook it up?

Pressure washer is what we use but you have to be very careful with them. You can shoot the pressure thru the roof (along with you and every thing else in your shop) very quickly. Depending on the size of your tank, your pressure washer and the proper conditions you can shoot the pressure over 1000 psi in about 2 seconds. I wouldn't recommend this unless you know exactly what you are doing.

Many moons ago, I worked for a control valve manufacturer and they did their hydro tests on the shop floor, using a sealed valve body, a water fill, and hydraulic pressure applied to the inlet of the valve, with the appropriate indicator gauge. They had various adaptors on hand, that they made to seal the inlet/outlet and stuffing box connections of many valve configurations.

Typical valve body tested was this one..

http://www.documentation.emersonpro...documents/instruction_manuals/d100386x012.pdf

And, the hyd. pressure was supplied by some like this..

http://www.enerpac.com/en/industrial-tools/hydraulic-pumps-and-valves/hydraulic-manual-pumps

Only witnessed a few of these tests, as I was always curious as to how it was done. Never saw a failure.

Be advised working pressure is not the same as test pressure and I think the tanks when testing occurs is in a vat of water, but I could be wrong.

Obviously this was done with a cheapo chinese bottle jack. I wouldn't recommend using one that you like. And yes of course this was a simple leakdown test, nothing more. However as I saw this was able to detect an issue.

Unfortunately the leak was an actual leak and not something that was running down the tank. I wiped the tank down good and was able to observe the droplets building there from the tank, and definitely not some residual water.

Had the tank not started leaking I was planning on conducting a test to 200 PSI and then a second test to higher pressure. All I had on hand at the time was that 200 PSI gauge to start with, but with the leak it was bunk.

Blow it up! See when it gives!

I put a pressure relief valve in that gizmo I have in the pic and the little ball valve is to close it off and regulate it.

You guys are going to kill yourselves doing this stuff... you've been warned