Corrosin defect assessments in Gas pipe lines 1

[09091960]

Hi ALL,


When analysing corrosion defects, will the Defects interaction applies to from the centre axis of the pipe to one side of the pipe or else needs to consider both sides of the pipe.

E.g. can the 3 o'clock and 9 o'clock defects interact?

 

[brimmer]

Center axis has no effect on whether corrosion features interact, so a feature at 11:55 and 12:05 can interact, or the same for near the 6 o'clock position. A quick interaction rule (there is certainly more than one way of assessing corrosion interaction or rules) is to simply use 6t both in axial or circumferential directions.

 

[09091960]

Hi Brimmer,

I'll try to be bit more specific.

The defect we got was

Defect 1
340mm(Wide)x 1.5 M (Long) at 3o'clock position

Defect 2
390mm(Wide)x 1.8 M (Long) at 9 o'clock position.

Due to the large diameter of the pipe these two defects won't interact in the circumferential direction. If the defect 2 starts at the end of defect 1 (Crossection view- due to 3o'clock and 9 o'clock positions))can the total length of the defect be 1.5+1.8 = 3.3 meters,

 

[brimmer]

Wow, those are some large corrosion areas, makes me wonder who the ILI vendor is, and how accurate the reported data is... You should field verify these.
In any event to specifically answer your question, if there is no circumferential interaction, then these are 2 seperate features that do no interact, so you have the lenghts as reported, not combined for the 3.3m. Same with all features, if they are not within 6t in any direction(or whichever method your company uses for interaction)then they do not interact. Hope this makes sense.

 

[09091960]

Hi brsrimmer,
Thanks for your reply.
Well our pipe line is over 30 years old.In past there were reports about repairs which carried out for large cluster type defects.
Resently MPI tests shows few cracks like defects.Can we assessed these by Kiefner's softwear.For e.g How can I assessed a 20mm in length crack? Do i have to make them into grids similar to river bottom profile ? If the crack length is limited to 20mm in length how will i do the calculation? Photo attached herewith.

 

[brimmer]

You can assess cracks with KAPA as a lower level assessment, remember to fill in the charpy value and click crack-like defect. You do not grid cracks as per corrosion, you simply take the entire length and depth. 20mm convert to length in imperial and use in spacing. If you buff out the cracks completely then take your depth profile as normally per corrosion. With cracking if you are completing the calculation before buffing you need to either use UT to get a depth, or visually estimate the depth for a rough guess. Better programs for crack assessments are CorLas or PAFFC. If you actually have a colony of cracks rather than a single crack, remember it is the longest interlinking crack value to enter in. In your case, I am seeing what appears to be a surface breaking lamination, it is very difficult to tell with the quality of the photo, do you have a close up? If it is a lamination I would not do anything with it, this one does not look very serious. Remember your pipe is probably full of laminations, some surface breaking and some internal only. I can only stress to use an NDE contractor who can actually classify indications and take proper measurements when dealing with cracks.
You might consider a course in pipeline defect assessment or the contactors/consultants you use have training in this.

 

[09091960]

Hi brimmer,

Thanks for the information.One last question regarding the cracks.One of the cracks i came across resently was circular in shape, in that case if i'm using KAPA the total length have to be the circumference length ?

 

[09091960]

Hi Brimmer,

Pl find the attached photo for your reference (Circumferencial crack)

 

[brimmer]

09091960,

This is a good photo. This is not a crack!, it is a lamination, much different than a crack. I would not assess these, there is no integrity threat here, if you wish you can buff these out of the pipe, but there is not much use in this as the area you buff will be larger than the original indication area. The only potential problem from laminations is if they may be through welds, or are not planar in the mid-wall, so are not consistent depth in the wall of the pipe, and usually you can get UT readings to confirm this. Why your NDE contractor is classifying these as cracks or crack like and does not know what they are is poor knowledge. If you must assess, (I am not sure how you are going to get a depth of something like this with out buffing as it is probably too shallow to be out of the dead zone with UT, in which case you are completing your standard effective area calculation after buffing), I would take the axial length of the longest part of the crack, not the entire circumference. Remember, the pipe will be full of these internally as well, you just can’t detect them with MT, but you will see them with UT.

 

[09091960]

Hi brimmer,

Many thanks for your advice.We have managed to UT this defect and it shows no indication of depth.It proved to be
lamination.Pitty that MPI contractor couldn't figure it out.
If the crack depth is less than 1mm, will the UT can detect
the depth of this defect?

 

[brimmer]

The short answer to this is yes, if a crack is under 1mm you can still detect it and measure depth with UT. Crack sizing is a speciality, not every NDT company can do this, and it often requires special probes. In general you can detect cracks of 0.5mm depth, but the shallower the crack, the more difficult to detect and determine depth. The deeper the crack, in general the more accurate the depth sizing. Creeping wave probes are often used to size cracks rather than regular shear wave probes. I could tell you much more about crack sizing, but it is too much to fit in this note, but crack sizing takes an experienced UT operator with the right probes. Some of the new phased array probes are also giving good results.