Hydrotest vs. Testing with Air 10

[clooney]

Dear gents,
I have not come across any work instructions or specs in our office establishing a guideline on "deciding when one should consider testing a piping system or pipe line by Hydrotest vs. Pneumatic Air Test?" or any other method for that matter. Can anyone direct me in finding the answer? I have been checking with the more experienced people in our office but, despite all their past achievements, none of the seniors were able to give me a straight answer or at least guide me in my search . Thanking you in advance.

Farzad

 

[zdas04]

I've got a document under "samples" on my web page where several of us tried to provide guidance on choice of test media and other topics related to static testing.

Some of the guys working on it felt that the accumulated energy in a pneumatic test created an unreasonable risk. The ASME (in B31.8) disagreed. If the hoop stresses allow, I will always do an air test preferentially a water test because: (1) I live in the Rockies and a 1,000 ft elevation change results in a 431 psi difference in test pressure between the bottom and the top of the hill, for a 900 psi test that puts me in the position of either not really testing the top of the hill or getting really high hoop stresses in the bottom; and (2) disposal of test water is becoming a big issue (you have to treat the water for microbes, and the EPA is cranky about putting biocides into ditches-washes-rivers.

B31.8 has some pretty extensive information about when you can use air, nitrogen, or natural gas (as long as your hoop stress is less than 30% of SMYS, sometimes you can buy your test media from a nearby mainline pipeline and get the gas already at test pressure and you don't have to throw it away, you can sell it back).

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The Plural of "anecdote" is not "data"

 

[1969grad]

I prefer to use a pneumatic test for equipment fuel gas lines and lube oil lines because of the difficulty of getting the line dry after testing.

 

[danthemeterman]

I refuse to use pneumatic, same reason as "Z"'s coworkers gave. I have witnessed the destruction a vessel can cause when it belches.If you havent seen it, don't wish for it. Some automitive shops have used 125 gallon discarded propane(no longer refillable, working press. was 350#)for extra capacity. One in particular had approx. 145#'s of air, in the middle of the night(thank God) the tank let go at the seams. this shop "WAS about 75' x 100", the property taxes have declined considerably, the insurance company refused to pay since they were using a non spec tank for thier backup air.


A word to the wise is not wasted

 

[clooney]

Thank you guys. At least now I have an idea of what the advantages and pitfalls are in using either of these methods. And thank you Dave Simpson for putting up your site. I'll definitely look it up.
much obliged

 

[zdas04]

Danthemeterman,
From what I can tell from your post, the propane bottle explosion was a failure in service of a vessel being used for an air receiver for a period of time. One has to assume that the failure mode was fatigue.

I wonder what the hoop stressesn the vessel were? ASME has some very specific restrictions on hoop stress during a pneumatic test, and I'll bet your ad hoc reciever did not conform to those restrictions.

Auto shops are not the only people who follow this dangerous practice, I recently did some flow testing at the engineering department of a major US university and their air reciever was a 500 gallon propane tank. I couldn't find a code stamp on it and asked when was the last time the tank had been certified as fit for service. The tech gave me a blank stare and the department head said he would take care of it. I haven't been back to see if he did.

I have never found the fact that air-receivers fail in prolonged or inappropriate service to be a reason not to use pneumatic tests.

David

 

[danthemeterman]

Zdas
I fully agree there is no reason to assume that beacuse of a catastrophic failure in an inappropriate installation would be reason to cancel a proven procedure. however i maintain that due to the possilbility of even a new vessel (tank,pipe etc) failing it's integrity the hydrotest method is manyfold safer than pneumatic. When I test a flex coupling to 415 psi. I put it in a cage I built to somewhat contain the possibility of schrapnel.
The failure of the propane tank was most likely due to the moisture introduced from the compressor and the climate in the shop causing humidity?? Just a thought. Since air tanks are built with coating.

 

[danthemeterman]

Zdas, What is hoop (stressing)? Are there any good sights that can get into detail? Thanks

 

[zdas04]

That was supposed to be "hoop stress in ...". ASME B31.8 has a good explaination of hoop stress. It is also discussed in the document on my web page that I referenced above.

So if the propane tank failure was due to the introduction of water, was it a corrosion failure? Did anyone get pieces of the shrapnel to verify that the failure was corrosion related? Most corrosion failures I've seen have been more weeping than tearing failures.

David

 

[danthemeterman]

zdas
Sorry , so long ago I cnnot remember the stats, just remeber the occurrence.
Side Bar: Did you hear about the 30,000 gallon propane tank that blew it's end off in Gwinner N. D. approx. 1988. the tank was full of propane. The end blew off, sent it approx. 100 yds. into a block wall and flew another 200'? hit another block wall and came to rest. No one hurt, happened about 1/2 hr. before maintance arrived. NO FIRE!!!
I witnesed a 100# l.p. tank some years after it blevied, some one put it in the basement full of gas, gass heated faster than the furnace used it. The tank was found splaid out flat with the ends missing.

 

[SteveK4]

The choise of Hydro versus Pneumatic really depends on the specific piping system being tested. If there is a big elevation change, the possibility of contaminating the system, disposal considerations, or if your are testing a low pressure system such as service air; you would most likely opt for pneumatic. However for most systems a hydro-test would be correct choice. Several of the piping codes give general guidlines on testing. On a side note, "Hoop Stress" is the tensile stress applied to a pipe or vessel caused by the pressure contained by them. You calculate it by P*D/(2t) where P=pressure D=outside diameter and 2t is twice the wall thickness. the result is the stress applied to your pipe or vessel wall. it is important to keep this number below the allowable strees of the material used in construction.

 

[bvi]

Hydrostatic is almost always preferred for a number of reasons. In piping per some codes, the hydrostatic test is conducted at a high pressure which results in "warm prestress" which increases the resistance of the tested equipment to brittle fracture. This is from crack blunting and putting crack tips, post test, in compression. The risk is also certainly less, because the consequences of a fracture in a pneumatic test can be much more severe than a hydrotest, since the stored energy of compressed gas is so much higher.

Having said all that, there are a number of good reasons for performing a pneumatic test. However, in the absence of these specific reasons, which have generally been cited by the other posts, a hydrostatic test should be performed.

 

[zdas04]

"Almost always prefered" by whom? The codes don't show a marked preference and at least 2 of us in this thread will always do a pneumatic test over a hydrostatic test if the hoop stresses are in the right place.

I don't mean to jump on you, but I keep running into folks that feel somehow superior for generating thousands of tons of waste material that is toxic to the environment and difficult to dispose of. This is simply an engineering decision, and for systems that are compatible with pneumatic tests, I find them to be quite reasonable.

David

 

[DVNMech]

In Europe, following the new European standards (EN13480-3 for piping or EN13445-3 for unfired press. vessels), the proof test shall be a hydrostatic pressure test (exceptions however are possible) The pressure equals the max of (1.43*design pressure) or {1.25*design pressure*allowstress Ttest/allowstressTdes}. In some cases, for example when a pipe or vessel operates in creep area, this leads to very high pressures.

 

[pmover]

all,

the responsible engineer is to consider the varying conditions in which the piping system (pipeline or otherwise) is to be tested and then utilize the appropriate codes & test medium for implementing the test. certainly, ambient conditions (i.e. temperature) and elevation changes need to be considered.

i have witnessed many hydro and pneumatic tests on pipelines, piping systems, and pressure vessels with the test medium decision based on sound engineering judgement and by following the applicable code(s). in some cases, agreement between the contractor and client was obtained to conduct the pneumatic test, followed by a hydro test when ambient conditions allowed for hydro test.

hope this helps...
-pmover

 

[ARenko]

zdas,

I've not read 31.8, nor do I have access to it at the moment, but I am curious - does the hoop stress you mention not apply to hydrostatic testing? In other words, does the spec consider that some additional safety is required due to the risk of damage if a failure were to occure while testing with air as opposed to water? Might this mean that if someone was intending to design a vessel for testing with air that it may be overdesigned for testing with water? I would've thought that the 30% you reference would apply to any test media.

I've mostly tested high pressure equipment (10ksi to 15 ksi rated), and therefore have done hydrostatic testing. However, I once did some testing of liquid tank containers at 20 psi. We filled those nearly full with water before testing with air in order to allow a minimal amount of air in the vessel.

 

[zdas04]

djv,
There are many hoop stress limitations throughout the testing section of B31.8. The one I refered to is the most restrictive and doesn't allow testing with natural gas to pressures that result in hoop stress over 30% of SMYS. There is another restriction that deals with personnel approaching a line when under a pneumatic test that results in hoop stress over 70% of SMYS.

All of the limitations in B31.8 are specific to a given set of conditions and location classes. It is tough to generallize it.

David

 

[moltenmetal]

It's a simple matter of comparing risk versus risk. If a hydrotest is feasible and won't produce secondary hazards, you should do a hydrotest because it's safer- the stored energy is far lower than in a pneumatic test and hence the hazard resulting from a sudden failure is FAR lower.

If introducing water into the system in question will introduce secondary hazards (i.e. corrosion, damage to equipment, or generation of an unmanageable amount of hydrotest fluid), you have to weigh that risk against the risk of the test itself. You may choose to increase other nondestructive testing on the work in question to give yourself more confidence prior to the pneumatic test.

We frequently do hydrotests for basic integrity followed by a pneumatic test for leakage. The first test is low hazard and gives confidence that the pneumatic test can be carried out safely.

 

[rmw]

Danthemeterman,

Right after reading your post, within a day or so, I was at my shop, an old building of about 2500 ft^2 and I fired up the air compressor, only a 33 gallon receiver, but it is direct connected to an older receiver with a bum compressor, about 60 gallon for extra storage, and just about the time that it pressured up to the max and shut off, I heard a sudden hiss of air that indicated something had sprung a leak. Suspecting a hose, I quickly checked out the hoses, and realized that it wasn't a hose, and that it could only be that the tank had sprung a leak (at the bottom, naturally, where the moisture collects) and all I could think of was what you described in your post, and I was torn between running out of the wooden building, and trying to open up something to reduce the air pressure before the tank burst.

It did not burst, and I got the pressure off, but what you described in your post sure played on my mind, and gave me a fright.

rmw

 

[quark]

rmw,

It was good that you took running for a precautionary measure. That could be as dangerous as you thought. There was one accident, I know, of a 0.5m[sup]3[/sup](132 gallon) receiver with a bottom leak jetting into air and landing 50feet away. This was three years back but I have yet to know whether the receiver was properly bolted to the foundation or not.

Regards,