Hydrostatic Testing Exclusion Areas

[qualeng1]

I've been tasked with identifying exclusion zone requirements for hydrostatic testing on piping, and can't find anything formally established. I know ASME doesn't give any boundaries (unless I've missed something). Any ideas, best practices, etc out there?

 

[LittleInch]

I don't believe there is any formal requirement for an exclusion zone, however some sensible precautions need to be taken. Look at your piping item and determine what would happen if anything broke or became detached - plugs put into instrument connections are an example of something that if the plug stripped and shot off like a bullet - where would it go? Similarly flanges can suddenly fail and a high pressure jet shoot out sideways - would it hit another person or piece of equipment?

It is simply good practice to isolate the area around above ground testing with higher pressure systems needing more caution. How far? - As far as is reasonable given the pressure, size of pipe and your location. Most fab yards have an area set aside round the back with concrete walls to contain any loose bits which fly off. A/G testing often is over quite quickly so you may just need to move people out of the way for an hour or so - do it during lunch break or straight after work is finished for the day.

Take care - testing is there for a reason and every now and then something fails the test. Get it wrong and you kill people - do it right and take sensible precautions and everyone goes home at the end of the day in one piece - your pipe might break, but you won't.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[qualeng1]

Thanks for the response - We have set a standard 20 foot exclusion for hydrostatic testing, however I have a very demanding customer who wants the basis for all exclusion zone determinations. If it were pneumatic I'd understand the concern based on stored energy, but how to justify any specified exclusion zone is proving troublesome. Additionally, our organization won't allow the use of any barriers unless they're validated to be able to withstand any potential blast, pressure transient or shrapnel. We are performing tests during lunch or after hours, so I appreciate your validation of our current practices! Just wish there were something similar to the guidance in ASME PCC-2 for hydrostatic testing.

 

[LittleInch]

I had a look at some large company standards and they are even vaguer than my explanation above(!) For just pipe, you could go down the line of the incompressibility of water, but as noted above, 20 feet would not be enough for a small plug or fitting on a large pipe, but fine for just some welded pipe spools. All depends on your volume, size, fittings etc.

I'd just go for total exclusion within your facility for the hour or so you're at pressure.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[qualeng1]

Thanks for all the help and advice. My experience was similar to yours - I even looked at NASAs manual which just said keep a distance and barracade yourself if possible (!). I'd have thought NASA would at least be more specific. Either way, you've at least validated my initial assumption no concrete guidance exists.

 

[jte]

qualeng1-

Certainly the intent of the exclusion zone within PCC-2 is for high energy testing. But you could calculate the energy involved with a water based test and determine the exclusion zone based on that. It'll be pretty small. Except for the default minimum 100' distance in 5.1 App III (a)(1). Which will not be changing in the 2014 edition. But which will be looked at for potential future changes.

Just wish there were something similar to the guidance in ASME PCC-2 for hydrostatic testing.

Consider sending a note to the committee secretary (see page v in PCC-2). You'll have to be a bit patient but you will get an initial response pretty quickly if you get an inquiry in within a couple weeks, otherwise by the end of the year, and you might be happy in a few years to see your concerns addressed in the ~2016 or ~2018 edition. It doesn't have to be a lot of effort - basically what you've already stated in this thread is sufficient to get things looked at.

 

[qualeng1]

I appreciate everyone's responses. We submitted an inquiry to ASME for guidance earlier this year, although I think it was via the B31.3 committee, and not PCC. Perhaps I should resubmit to PCC?

 

[Sparweb]

A photo, published in a book informally known as "Augustine's Laws", shows an aerial view of a factory, with some surrounding buildings and parking lots.
On one side of the photo is an arrow, pointing to the site of a pressure vessel that had exploded and taken out part of the wall.
On the other side of the photo, representing about a thousand feet separation, is another arrow, pointing at the car struck by a fitting ejected from that pressure vessel.
Lesson: anything can happen. You can only prepare for what probably/possibly could happen.

How technical does your analysis have to look?
Pick any cross-section that could be fractured by any failure (fitting/flange/elbow whatever) Apply the system pressure that could cause that failure to the entire cross section. Estimate the elongation of the part that occurs when it has failed (elastic + plastic). Multiply the force times the distance, and by pi because this is scientific. Use this energy estimate, divided by the mass of the smallest component that could experience this failure, to find its ejection velocity. Estimate the distance that can be traveled by that part on a ballistic trajectory. Multiply by pi again and that's your worst-case scenario.

When it's off-the-cuff like that it probably doesn't sound like much, but the forces and potential energy involved may not be too difficult to determine for a reasonable estimate. I don't know if your customer wants to see that kind of analysis, but you certainly can't determine something like that by test! Only accident experience can point you in the right direction, whether by first-hand reports or by reading reports of investigations after the fact.

ASM publishes interesting accident/failure/fracture investigations that may speak to your problem, though it's a bit of a needle in a haystack.


STF

 

[MikeHalloran]

Small hoses used for refrigeration, hydraulics, etc., are commonly pressure tested on at least a sampling basis. Because hose walls can store some energy, the test stations are usually equipped with some sort of enclosure. Nothing that the NHRA would approve for a clutch scattershield, but something other than an exclusion distance.

Maybe you could do the calcs outlined above for fractured fittings' kinetic energy, including at least a nominal amount of trapped gas to represent what you can't bleed out without vacuum, and use that to justify or suggest use of barriers during the test. You could scale the barriers using the calculated kinetic energy and the likelihood of a failure. Maybe furniture blankets draped over the plain pipe, furniture blankets plus chains over the simple fittings, and woven wire rope ballistic blankets over the big stuff with no history.

If your customer is as nervous as suggested, their insurance carrier, contacted with the customer's permission, might have some useful input.



Mike Halloran
Pembroke Pines, FL, USA

 

[Duwe6]

Bottom Line -- how much energy is being stored?

On a pneumatic, it is huge - thus a huge exclusion zone is needed.
On a well-vented hydro, it is very small, thus the exclusion zone is tiny [1 to 10-ft]. And it is easy to tell how much energy is being stored; "How long did the hydro pump run?" On a properly vented hydro, the main mechanism is the [miniscule] elastic stretching of the steel vessel, and the pump only runs a few seconds. My personal preference for pipe spools being shop hydroed is to have the fitters use a hand pump; that forces them to get good venting so that they only have to stroke the pump 5 or 10 times to achieve 1000(+) psi. Very, very little stored energy, and that Exclusion Zone is "don't touch it".

If you have to run a 1-horsepower pump for 5 to 20 minutes, everybody needs to back waaaaay up.

The Worst-Case failure during a hydro-penumatic gives the vessel relaxing at the speed of sound in steel and water, and launching a medium-sized flanged nozzle at the speed of sound in air; about 1,125 FPS [slightly faster than a bullet from an army 45 caliber automatic]. Wow!
the probability of a nozzle failing, instead of a major seam are too small to calculate, so like most Worst Cases usually are, it is too improbable to happen during the lifespan of the planet

 

[JR97]

Found this on a Google search:

http://www.wermac.org/others/ndt_pressure_testing_safedistance.html

Previously I also found a Norwegian spec and a report from the John Glenn Flight Center that had safe distance calculations. Unfortunately I didn't save those links.

 

[Doove117]

Not sure if already covered, but an important point is what the pipe material is, and determining what failures could occur and what the failure mechanism is.
e.g. plastic pipe has a Youngs Mod orders of magnitude lower than steel, i.e. it stretches more, so stores more volume, also some types of joints can slip and pop off. The point above about estimating input energy is really pertinent.
Look for the weakest points (usually fittings, tappings etc).
PVC pipe can break into dagger like shards whereas steel pipe will crack or tear. I once replaced an exploded PVC elbow off the top of a vessel (we replaced the whole line with stainless), I think it was 6" Class E (rated for 15barg), the elbow had gone, all the adjoining pipe had burst, it had gone through the slate tile roof, we found bits of slate about 15m away on the road outside, the room itself was filled with these razor pointed shards of PVC... if anyone had been in there when it went they would have been perforated.
Testing plastic pipe is, in my experience, more risky than steel pipe for material failure.
As LittleInch noted above, most shops (including ours) have a dedicated test area that is fenced off. During testing the area is cleared. After a minimum half hour of stabilisation the pipe is inspected if deemed absolutely necessary (I try and avoid this), then the area cleared for the remainder of the test (typically 1 hour).
Every time when you prep for a test be thoroughly acquainted with the setup and walk it through at least twice, thinking 'what could go wrong'. Plan accordingly.
In the past I have covered old PVC lines with heavy duty tarp, weighted down with sandbags to protect local equipment and ensured the room was fully evacuated.
I have also done testing after hours when the shop was empty (if the pieces were too big for the test bay).
Don't forget your test gear and pipe is also part of the same system. Include this in your plan too.
Hiring site fencing panels to barrier off is cheap, if you do a lot of testing then buy them, is there any reason you can't barrier off at least during stabilisation? After an hour at pressure you can be pretty certain it won't suddenly explode (not absolutely certain, but the risk is much less as theoretically the system is in equilibrium).
I'm rambling a bit. To answer your question, I would erect barriers and specify exclusion zone to barrier. If you're in a room with other equipment, consider damage to the equipment or structure of room and take whatever measures you can. If inspectors want to crawl over the pipe looking for leaks then you have every right to refuse for their own safety. At end of day I have never had an inspector pass a pipe based on visual inspection for leaks, so only thing they should be interested in is if the calibrated pressure gauge is showing no pressure loss, and that is what goes on the certificate.

 

[qualeng1]

We have hired an external company to perform the necessary calculations to determine height, width, length and material to use for barracades. I do appreciate the input as to how to calculate potential incident energy for a hydrostatic test. We will look to perform these calculations in addition to the requirements for pneumatics in accordance with ASME PCC-2. We have also submitted a request to ASME for additional hydrostatic guidance during this revision year. Fingers crossed!

 

[racookpe1978]

OK, so now that the "nearby workers" have been partitioned off and sent to the next county for storage during the hydro ... Recognize that you STILL need to walk around and closely inspect the whole piping and vessel and hose connections to validate the hydro.

You can't "barricade" that risk for those people, despite what this bureaucrat thinks when he wants you to quote "book and chapter and verse".

 

[jte]

We have also submitted a request to ASME for additional hydrostatic guidance during this revision year. Fingers crossed!

As long as you define "this revision year" as "to be published in the 2016 (or 2017?) edition". The 2013 2014 edition was frozen last summer and should be out soon... perhaps.

Glad you submitted the request. Feedback from users can result in changes. I hope to see some clarification (and reduction in conservativeness) in that area myself.

racookepe - Keep in mind that there is a difference between a "limited access area" and complete exclusion. If you have specific suggestions for wording improvements or wholesale changes, send 'em to the committee secretary. Good testing procedures can mitigate the risk to those who have to be up close and personal with the system. But I do agree with your sentiment that logic won't always work with a true bureaucrat.

 

[qualeng1]

jte - I do like the idea of a "limited access area." Up until now we were considering the whole area an exclusion area (we do not have fluid systems of sufficient height to cause notable gauge error due to static head), and all our testing equipment is maintained outside the area. However defining a limited access area and the conditions in which to enter is a fantastic idea.

 

[Duwe6]

". . .you STILL need to walk around and closely inspect the whole piping and vessel and hose connections to validate the hydro."

After the Code-mandated 10-minutes at full hydro pressure, I have them drop the pressure 10-15% and then start the hands-on inspx for weeps and seeps. Rational is that during 1000 psi hydro with everything tight [no weeps], the pressure will usually climb [water warming up]. Just because everything was fine at 1000 psi, it does not mean that at 1003 psi something won't rupture. Drop the pressure to 900 psi and inspect with confidence.