Pipe Pressure Test Failure 7

[dik]

A portion of 8" dia pipe failed a pressure test. Failure was along the longitudinal welded seam. It is my understanding that ASTM does not permit welding within an 1" of the seam. Is this correct? and is it possible under some other regulation that the seam, with a proper welding procedure, can be rewelded to repair the damage.

Dik

 

[dik]

Is there any non-destructive testing that can be done? Can they be proofed to the capacity of any of the fittings?

There are several runs of several hundred feet of the material already installed. The work was almost finished and failed the pressure testing part of the specification. The structure is enclosed and it would be difficult (not impossible) for the pipes to be removed and replaced.

 

[Ron]

dik...if the thickness is 1/4 or greater, you could use ultrasonic flaw detection. Magnetic particle would work as well, unless the defect is on the inside of the pipe.

Radiography would work, but would be expensive.

 

[dik]

Thanks, Ron, I hadn't considered ultrasonic for this app... Because there's so much of it, I was hoping for an easier method.

Dik

 

[brimmer]

dik,

you might be able to test with internal ultrasonic tool, should detect these types of defects, I don't know what your scenario is

 

[racookpe1978]

You don't want ANY of that manufactor's pipe in your project carrying anything more important than gutter water as a downspout.

Rip it out - DEMAND your money back.

Look - pipe that is failing at 80 psi is tissue paper.

On an offshore rig, are YOU going to live UNDER it 24/7 for 30 years with anything "important" or flamable inside? Tell that rig's crew what happened: See how of them stay on board.

 

[dik]

racookpe1978:

That was my original thought, but there is a lot of it installed and it is difficult to remove and re-install and I was wondering if there was another solution.

Dik

 

[BrianCNTFM]

Dik,I have heard of this type of failure but never witnessed it. From recollection it was attributed to welding process start up. This might allay some of your doublts but it would be worthwhile completing some random UT and possibly RT if the UT techniques available are limited.

Regards BrianC

 

[dik]

by UT and RT you are referring to Ultrasonic and Radiographic?

Dik

 

[racookpe1978]

No - R/T won't be an effective answer.

Consider that a single R/T of a circum. weld takes 1 hr setup (in a powerplant at least) and an exclusion area of around 100 ft radius. 45-75 some-odd minutes per weld (36 inch circle) then 1/2 hour take down time. Move to the next weld, repeat.

Now image that much time for a linear distance of "several hundred feet". The R/T might work if it (the souorce) were dragged down the inside of the pipe with the film tapes continuously to the outside over the weld. But you'd need calibration points on the pipe and film to track down defect locations. Add thousands more dollars for exam time, rip out and replacement of the bad/suspect pipe anyway. Scaffolding. Insulation takeut, hangers in the way of the exams.

Worse: he has had catastrophic failure at low pressure in one pipe. How many other pipes from that supplier has he got installed? The supplier is overseas, behind a protective (and corrupt) local and national government. Replacement pipe is expensive, but imagine trying to collect billions to pay for the burned out rig PLUS medical liability claims for the killed workers.

Faced with that bill, the Chinese company will likely simply "disappear" - nobody will be be there to be liable.

1) UT/MT might show extent of what pipe needs to be replaced.

2) Get the repair records of what was re-welded. DON'T trust the re-welded pipe.

 

[BrianCNTFM]

Dik/Racookpe I did mean Ultrasonics and radiography though I would limit it to % inspection.

Whatever you decide to do further inspection will be required and the results past up the chain to ensure root cause and corrective action.

Regards Brian

 

[dik]

In view of

ASTM A53, states:

12.2.2 Type S pipe and the parent metal of Type E pipe, except within 1?2 in. [13 mm] of the fusion line of the electric-resistance-weld seam, are permitted to be repaired in accordance with the welding provisions of 12.5. Repair welding of Type F pipe and the weld seam of Type E pipe is prohibited.

is there any means of re-welding to effect a repair? 'Prohibited' seems to be a pretty strong word.

Dik

 

[AllHandlesTaken]

You could take the view/argue that the A53 specification does not apply to you because you are not a manufacturer and have already purchased the specified pipe. Then you could work with, perhaps, ASME B31.3, PCC-2, API 570 ...., all of which (MUST) allow welding near the fusion line or you would never even weld on a flange.

 

[BrianCNTFM]

May I suggest you pose your question to a metallugist/welding engineer, or technical area of supplier (actual or otherwise). This might give you confidence in pursuing your line of action.

It might well be that a repair can be justified pwht.


I take it you have found no more defective pipe on your or other projects and the supplier has had no other reports of similar failures?

Regards Brian

 

[racookpe1978]

A separate problems lies in ambush for you:

One run of pipe was tested during one hydro test: That run failed at only 80 psig.

"Somebody" weld-repaired that (one run) of pipe at that failed location: so he had to heat up the failed (burst out and deformed) section and beat it back into a "sort of round" shape in order to get the original edges back near enough together to weld them manually. So now you have a heat-affected (low strength) zone around the weld, a low strength (yellow-or-dull red-heat) "manually forged" section and an out-of-round (stress -induced failure) section AND a retest-required weld that must be itself tested at full hydro pressure.

BUT, the rest of the pipe has NEVER been tested to anything past 80 psig: the original failure point of the first pressure test. Why do you assume that the rest of the pipe will even get up to 100 psig, much less 120, 140, or full hydro pressure. (As noted above it "should" have passed the 1250 pressure test at the factory - but we know that it could not have.)

 

[stanweld]

There are a number of manufacturing reasons why the ERW weld could fail at such a low pressure; however, the high pressure test at the mill should have opened it. You should assume that it was not so tested and expect that additional partially fused locations are present. You need to examine the pipe seams for their full length (all installed 8" and 10" pipe from that Chinese manufacturer) to assure freedom from future failure assuming that you think the material can be salvaged.

As a minimum, I recommend removal and replacement of the burst pipes and perform the A-53 required tests thereon. Welds should additionally be examined using magnetic flux leakage and UT methods.

 

[racookpe1978]

Try this: You're obviously facing short-sighted (economically driven) pressure to get the rig built, and to NOT replace the piping.

So,

1) Remove the pipe that has burst. Unbolt the flanges and fittings (it will be less expensive than rebiulding the pieces). Pull it out of the racks and overheads and put the pieces 9each spool) on the deck. Put blanks on each section - leave the long welded sections in the rack only if you have too. Hydro each section. As more and more break - if they break - demand your money back plus damages plus cost of testing plus delay time.

2) Use the separated spools as templates for the replacement sections: Weld brackets (dead or dummy flanges temporarily attached to the deck or brackets) at each flanged end of each spool. From each of these dead flanges, rebuild the replacement spools back towards the template at the other end - that way, the replacement spools will fit the original location with little problem.

As you find out how many spool sections are breaking, you can begin to identify the limits of your problem. Use NDE properly: to verify the fabrication welds at fittings and flanges, not to check the full length resistance welds down every length of pipe. Field NDE isn't accurate enough to do that job. Field welding isn't accurate enough (economical enough and fast enough) to rebuild pipe lengthwise welds.

Right now, all you have is a deathtrap of unknown pressure rating.

 

[dik]

This thread has whiskers... but, an update from Testing Co report (names have been changed to obscure those involved but the text is almost verbatum):

----------- Start of Quote -----------

"...Joint 001 welded by ‘Welder A’ failed the radiograph. Joints 002 and 003 the (sic) films were not able to be used. These joints are required to be re-tested.

Testing Co has reviewed all of the three pipes that, the weld seam has cracked (6" pipe that had visually bad seam - V groove; this one also offshore but from India). It is our recommendation that samples of these pipes be sent for Chemical and Mechanical testing to verify that they meet the requirements of ASTM A53 standard for pipe grade B.

The Client has requested that Testing Co perform ultrasonic inspection to the piping throughout the Project, to insure the pipe weld joint is sound. Testing Co has requested pipe samples several times from the contractor, for calibration purposes.

Once we have these samples from the different nominal pipe diameters we can begin the random spot checks of the heat numbers that have failed the hydrostatic testing as will (sic) as some additional testing of other piping that has passed the hydrostatic testing..."

----------- End of Quote -----------

It seems to be moving slowly... and will keep you posted on the end results. I haven't been involved with this, but it seems that there is some difficulty in obtaining samples from the contractor. Some of the pipes carry water at about 160F while others have cold water.

 

[itsmoked]

Wow, what a disaster... (economic so far)

Keith Cress
kcress -

http://www.flaminsystems.com

 

[racookpe1978]

I'm not convinced UT will show fundamental flaws in the longitudinal weld in the bad pipes: You're looking for flaws and cracks and lack of penetration in that weld better discovered by MT across the joint rather than UT into the joint (no penetration, partial penetration, no weld continuity, no weld at all.)

UT could show "depth" problems inside a deep piece of metal, but not across a seam that thin. (Pipe wall is less than 1/2 thick.)

 

[dik]

racookpe: By MT you mean?

Dik