Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

UT TEST RESULTS AND FLAWS 1

Status
Not open for further replies.

ali513704

Civil/Environmental
Nov 26, 2014
8
I AM WORKING ON PIPE LINE PROJECT 600 mm dia carbon steel pipe is being but welded .my problem is that we have to carry out 10% ut of weld .results show that 6 out of 10 welds are flawed our specifications suggest repair of flawed welds.i m insisting that test must be repeated with increased sample size however contractor is arguing that specification or any code(i-e api 1104) require only repair of joints
please guide me what is best for quality and safety
 
Replies continue below

Recommended for you

You are certainly within owner's rights.

B31.4 or or B31.8?

B31.4 says,
434.8.5 Required Inspection and Acceptance Criteria
(a) Required Inspection
(1) The quality of welding shall be checked by
visual inspection and supplemental nondestructive
methods
or by removing completed welds as selected
and designated by the inspector
for destructive testing.
(2) All welds shall be visually inspected.
(3) When the pipeline is to be operated at a hoop
stress of more than 20% of the specified minimum yield
strength of the pipe, the welds shall be inspected. A
minimum of 10% of the girth welds and 10% of the other
welds completed each day shall be randomly selected by
the operating company and nondestructively inspected.
The inspection of girth welds shall be by radiographic
or other accepted volumetric methods.
Nonvolumetric
methods, such as dye penetrant or magnetic particle,
may be used for other welds. Each weld inspected shall
be inspected completely with the selected method. In
the following locations or conditions, all welds in the
pipe shall be completely inspected; however, if some of
the welds are inaccessible, a minimum of 90% of the
welds are to be inspected ....

You can find similar wordings in B31.8


you must get smarter than the software you're using.
 
Tell the contractor to "Put up or Shut Up". 10% UT was not "chiseled in stone", it is a starting point. With a Reject rate of 60% of course the contractor is scared to have the remainder of his [pathetic] welds tested. And of course this outrageously high Reject rate justifies and requires increased inspection!

My favorite scheme of 'percentage testing' is what ASME B31.3 uses: Every welder's welds get tested, at an average rate of 5%. Any Reject requires two 'tracers' - testing of similar size welds, made during a similar time frame, by that same welder. If either of those tracers fail, two more tracers are taken [same method of selecting the welds]. If, out of those four 'tracer' tests two or more welds fail, then ALL that welder's welds are required to be either tested or completely cut out and replaced. Thus, the reason that your 'local contractor' is terrified to have any more welds tested. At a 60% Reject rate, I'd be scared too.

One of the most expensive things on a pie job are 'cheap' welders. Hire the best ones you can find, and don't allow them to have very many Rejects. Historically, the big [huge] contractors will allow one Repair per month and ZERO cut-outs* during the life of the job/contract.

*cut-out means a weld with sufficiently big or numerous defects that cutting it out instead of repairing the defect.
 
You should shout the name of this "contractor" from every rooftop.

you must get smarter than the software you're using.
 
thank you Duwe6 and BigInch i will get ASME B31.3
 
At a 6 to 10 rate, you should be considering a different contractor.

Mike McCann, PE, SE (WA)


 
Do you have any project or company specifications for construction, welding or testing? Often these will specify additional tests required when failure rates apply and can go to 100% of welds until the failure rate starts to significantly decrease.

I agree with the posts above.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way
 
thank you littleinch
problem is according to our project specification contractor has to perform ut test of 10% of joints and 100 of tie joints ,bends,golden joints and if flaw is detected he will rectify the flaw .my opinion is what will be the condition of remaining 90% if flaw rate is such high as 60% .in this regard our design engineer who set the specifications is also supporting contractors opinion
 
That engineer is dead wrong.

Go past him, go to the next higher levels.
 
What a joke. As owner you can specify whatever inspection rate you want. If you are finding that rate of weld failure, why wouldn't you go to 100% inspection? Time to re-qualify the welders, this rate of weld rejection is ridiculous. I would stop the job and re-qualify all welders, assuming your NDT inspection and criteria is correct...
 
If your contractor doesn't like the increased NDT he can take a hike. It doesn't seem like you would be loosing much to me. If you have to keep the contractor, get one of the rejected welds cut out and sent to a lab to verify the NDT and shut him up.
 
brimmer i not the owner of project i m employee of the organization my organization hired an design engineer who has designed the project and also the specifications
 
You have been strangled by a poor specification from the 'design engineer.' If the contractor is also responsible for the NDT, then expect it to play out something like this: you want increased NDT, Contractor will say OK but that is a variation to contract so you will pay, your management will ponder a contract variation whilst the Contractor continues the current debacle, by the time management are convinced that there is a potentially unacceptable risk exposure the pipeline will be buried.

Steve Jones
Corrosion Management Consultant


All answers are personal opinions only and are in no way connected with any employer.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor