PDA

View Full Version : What happens when a tank fails a hydro test?



winchman
07-15-2008, 02:09 AM
I know exploding tanks can be really violent and dangerous events when they're filled with gas, but what happens when a tank fails a hydro test?

Does it just split open to let the water run out? Or, is it just less violent than it would be with gas in it?

I'm talking about something like a 100-gallon pressure tank for a shop air compressor or a propane storage tank, but any info would be interesting.

Roger

doctor demo
07-15-2008, 02:25 AM
I know exploding tanks can be really violent and dangerous events when they're filled with gas, but what happens when a tank fails a hydro test?

Roger
When a tank fails a hydro test it is removed from service and sold for scrap:D .
I couldn't resist:) .
It is almost as exciting as watching paint dry.
Steve

Doc Nickel
07-15-2008, 02:37 AM
Small tanks, like for oxy/acetylene or SCUBA, are hydroed inside a water chamber. The expansion of the (water filled) test tank displaces some of the water in the outer tank, which gives the amount of expansion. Simple math then tells whether that's too much or not enough expansion.

When the inner pressure is released, the amount that's not drawn back into the test chamber shows the amount of permanent expansion of the tank- if any. Over a certain amount and the tank is known to be weak, and is removed from service.

It's virtually impossible for a tank to "explode" under these conditions. If the tank manages to rupture somehow, pressure is vented from the inner tank to the outer- but since it's pressurized fluid, not gas, the pressures equalize easily and there's little fanfare.

This works for larger tanks too- they're tested filled with water, so there's little energy stored in the form of compressed gas. If a pinhole or even full rupture occurs, pressure is lost rapidly with little or no chance of flying parts or other danger.

Doc.

macona
07-15-2008, 04:06 AM
Acet tanks are not hydroed. They do a ultrasonic I believe.

When a tank is bad it just leaks. They are lowered into pits in the ground for the test though.

dp
07-15-2008, 04:26 AM
It becomes a planter.

Davek0974
07-15-2008, 04:53 AM
It ends up in China

bollie7
07-15-2008, 05:49 AM
To get back to the original question, I would think that it would depend on how high the test pressure was and how fast the pressure was being applied.
Whilst I have only ever tested a couple of my own air receivers, years ago we used to repair and reset our own boiler safety valves using a hydrostatic tester.
Basically a bucket of water with a hand operated pump piston pump, a master gauge and the safety v/v on test, bolted to a flange. Was easy to get quite high pressures and when the valve lifted the pressure dropped off immediately and water ran out of the valve discharge.
I have tested my own air receivers using the same principle. fill it with water to the very top making sure there was no air space at all, and then putting some pressure on using a hand operated grease gun. easy to put 150psi on a receiver this way. When the relief valve opens the pressure drops off immediately and a bit of water leaks out.
Before anyone jumps on me, this is for my own personal gear, not commercial stuff. I set my compressor cut out at 100psi and the receiver relief valve at 105 psi. When testing I run the test pressure to 150psi. I figure if the receiver doesn't open up at 150 it will safe at 105. If it does open up, well, its time for a new one.

I think a tank failing would be pretty much the same as a relief valve lifting. the pressure would drop off pretty quick.

regards
bollie7

Chipslinger
07-15-2008, 06:15 AM
Acet tanks are not hydroed. They do a ultrasonic I believe.

When a tank is bad it just leaks. They are lowered into pits in the ground for the test though.


You are correct the stuff in a acetylene shouldnt get wet. it is what the acet . absorbs into and when it stops absorbing it is no good anymore.

So never buy a acet. bottle that has been left open, and always close an empty bottle.

oldtiffie
07-15-2008, 06:31 AM
Most receivers have both a working pressure (WP) and a test pressure (TP) test to pass hydro-statically. They are held for a specified time.

The principle is as said previously.

The relief and safety valve settings are applied later.

Here in OZ, all design, construction, testing and use are prescribed in the "Pressure Vessels Code". I'd imagine that it would be similar in other countries.

The Code also specifies the installation, draining, inspection openings as well as the methods and time intervals of (re)testing as well as repair if/as required.

Inspections and repairs etc. are to be carried out by "Licenced Persons".

My pressure vessels must all be re-tested at no more than 10 year intervals as they are not in an industrial or manufacturing or critical or potentially dangerous or corrosive environment.

This applies to my compressor as well as my LPG bottles (portable) as well as the 55 litre bottle in the back of my car. The car bottle was re-tested about 3 months ago.

Its more than its worth to get my portable compressor and LPG bottles (car excepted) re-tested at 10 years, so I dispose of them at the approved site at out municipal dump.

My oxygen, acetylene and Argon-mix bottles (for my MIG welder) are re-tested by the owning company as required. We rent our oxy, acet, and argon etc. bottles as we are not allowed to own them. It works very well as I just swap them at my supplier (where I have a "Trade" account) and it is all looked after for me.

I paid my annual rental for my "D" size oxy and acet bottles (1 of each) which was AU$257.40 (~US$245) which might seem expensive to some but it suits me as all regulatory requirements are met.

If I had an accident with bottles that were out of test, my insurance company would take a very dim view of it and may walk away from it.

SGW
07-15-2008, 06:51 AM
My understanding is that not much happens because there is very little stored energy. Unlike a gas, water doesn't compress. The only stored energy is in the elastic expansion of the tank.

Mcgyver
07-15-2008, 08:26 AM
My understanding is that not much happens because there is very little stored energy. Unlike a gas, water doesn't compress. The only stored energy is in the elastic expansion of the tank.

exactly what SGW says. because (at the pressures we're talking about) the liquid is not compressed, it takes very little pumping with a hand pump to raise the pressure, therefore there is very little stored energy. make sure though that there is negligible air present or you're compressing the air, them there would be a lot of hand pumping and storing of energy. This why people don't die if a hot water tank lets go :D

if you're talking a used water tank, I don't know, how are you ever going to get comfortable with its current condition. when you are storing something under pressure that would be bad if things let go, it needs to be engineered to some safety factor (ie 8x for a boiler) yet you never test at anything close to its safety factor burst point as that could damage it...you test to something incrementally over its working pressure.

I don't know what a reasonable safety factor is for a air compressor - for the sake of argument lets say its 5x and working pressure is 100psi. The tank should be engineered for 500 psi but obviously you can't test to 450 as doing so can deform it. So you test to say 150....but may it would have let go at 175. Something that is going to burst at 175 shouldn't be used a 100 when a 5x safety factor is needed.

Bottom line is hydro testing checks for leaks and and that one didn't forget any welds or silver solder joints - its not used to determine strength. That has to be done via engineering and that's very difficult if you don;t know what you've got, material, thickness, corrosion etc

bhjones
07-15-2008, 11:26 AM
It's Acetone.



You are correct the stuff in a acetylene shouldnt get wet. it is what the acet . absorbs into and when it stops absorbing it is no good anymore.

So never buy a acet. bottle that has been left open, and always close an empty bottle.

bhjones
07-15-2008, 11:43 AM
Hydro testing is used to measure the expansion and contraction of a tank. Why it fails could be do to a leak, or metal fatigue, heat damage, chemical damage, etc.

I worked at a large dive shop for a few years and we tested aluminum tanks at 5/3 the stamped service pressure. So a 3000psi alum tanks test pressure is 5000. My recollection is a bit fuzzy, but I seem to remember steel tanks using a slightly different test pressure formula.

If a tank failed we'd pull the threads out of the neck or drill a hole in it. I always found it amusing the number of people who would plead with us to just pass the tank one last time. They're were apparently ok with the idea of wearing a failing tank with 3000psi on their back and worse yet putting it in the back of their car while the heat of the day bumps the pressure up.

Anyway, I was under the impression that only DOT stamped tanks needed to be recertified.



Bottom line is hydro testing checks for leaks and and that one didn't forget any welds or silver solder joints - its not used to determine strength. That has to be done via engineering and that's very difficult if you don;t know what you've got, material, thickness, corrosion etc

Mcgyver
07-15-2008, 01:08 PM
Brett, sounds like there's a world of hydro testing out there i know nothing about....i was referring to hydro testing on the context of home shop pressure vessel making... seemed more like what was being hinted at here.

bhjones
07-15-2008, 01:19 PM
I guess I missed that part, sorry.


Brett, sounds like there's a world of hydro testing out there i know nothing about....i was referring to hydro testing on the context of home shop pressure vessel making... seemed more like what was being hinted at here.

macona
07-15-2008, 02:12 PM
It's Acetone.

Well, Acetone and a matrix of something like a ceramic. In old tanks it is a asbestos mix. Yummy...

JCD
07-15-2008, 02:47 PM
The tank I tested to failure, responded just as you described, a split and a big leak. No violent explosion. The tank contained no gas (air), zero. It was a small tank that failed at about 250 psi. If you are going to try a hydro test, remember that; most liquids are somewhat compressible, thereby capable of some energy storage. That means with enough liquid and enough pressure you will get a violent explosion.

Ernie
07-15-2008, 08:59 PM
One of the bases I was stationed at in the early 60's dealt with F106 fighter jets. To start the jet engine required 3000 psi air pressure for about 15 seconds. The air compressor operated at 4000 psi thru a regulator so the pressure would remain at 3000 for the needed time. To test the tanks, they were filled with water and a wetting agent to eliminate air bubbles. Then, using a pump like a porto power pump, the tanks were pumped to 6000 psi. They were left at that pressure for a few minutes to make sure no pressure was lost. Then, believe it or not, the air force proceedure was to beat on the tanks with a ball pien hammer. Luckily, the sargent in charge volunteered to do the beating. I was too chicken. He said he only had one ever rupture on him and all that happened was the water gushed out. Rapidly mind you but no explosion.

Ernie

h2o
07-15-2008, 10:45 PM
I work in boiler and pressure vessel safety with the ASME Code. Hydrostatic tests are used for the safety of the personnel during the pressure test. Water is virtually incompressible until you get up to around 15,000 psi I believe.

Under more reasonable pressures like 200 - 300 psi, if a vessel fails the hydro due to rupture, the pressure drops to zero almost immediately. If you were standing next to it, you may get some water on you but there would not be an explosion.

The old test pressure for air tanks and the like was 150% of the MAWP (maximum allowable working pressure) but the ASME Code lowered the test pressure a few years ago to 130% of MAWP. The safety factor used to be 4 but it is now 3.5. That means the manufacturer can now build a thinner tank for the same pressure. There is a pressure test method used by air tank manufacturers where they do a pneumatic test rather than a hydrostatic test. For safety, the tank has to be submerged in water to absorb the force of an explosion if it were to happen.

John

JRouche
07-16-2008, 12:42 AM
It ends up in China

You didnt end the story. They are given a nice coating of paint then shipped to the lowest bidder. Ahh? U.S. I would imagine LOL... J/K!! JR

oldtiffie
07-16-2008, 03:15 AM
With that sort of a sucker (tank and bidder) the bidder at least would implode - down his bidet?http://en.wikipedia.org/wiki/Bidet

Norman Warfield
07-16-2008, 06:59 AM
Think of a hydro' test as a balloon, if you blow up a balloon and stick a pin in it, it will explode with a bang. i.e. gas under pressure!

If you were to fill the balloon with water and stick a pin in it it, the water will rush out harmlessly. i.e. liquid dosen't compress!

For all practical purposes, with a Hydro' test you may get wet but you wont get hurt.

gnm109
07-16-2008, 01:00 PM
My understanding is that not much happens because there is very little stored energy. Unlike a gas, water doesn't compress. The only stored energy is in the elastic expansion of the tank.


Correct. For a while years ago I worked at the main Rocketdyne plant in Southern California doing hydrostatic testing. Units that were to be used for pressure would be tested with water pressure up to at least twice the normal pressure value. For the most part there were no leaks but from time to time, a weld would let go. It would just spray out and stop the test. We wore protective face masks and when the pressure was on we had a wooden screen with a window for the inspector to look through.

The pressure will dissapate quickly with hydraulic, unlike pneumatic pressure which will chase you around the room.

As far as the compressibility of water, this article fro Wikipedia regardint the water molecule states that in depths of 4,000 feet or so, water only compresses to a decrease in volume of about 1.8%. That's not very much.


http://en.wikipedia.org/wiki/Compressibility_of_water#Compressibility