UT Thickness Results are inaccurate for an on-line pipe? 3

[MeltedSnow]

I had some UT thickness results back claiming a pipe was significantly thicker than its nominal thickness! We believe that this is because the pipe was on-line at the time, so the temperature of metal altered the speed of sound within the material, making the UT device think that the pipe was thicker than it was. Is there any calculation I can do that will allow me to convert this hot UT reading into something more sensible?

Many Thanks

 

[BigInch]

Highly improbable.

I'd bet simply that someone put the wrong pipe there.

 

[MeltedSnow]

Thanks for the response! It's consistent all across previous years. The pipe has a nominal of 9.52 (let's say) and 6 or 7 thickness readings (from separate occasions) can be above 10! This could be because of a tolerance, but the pipe switches from 9.6mm in 2008, to 5.8mm in 2009, to 10mm in 2011! Is it possible that the 5.11 is an outlier?

 

[BigInch]

At the same points being measured?
Highly unusual.
Instrument damage?
What's it carrying. Is there any chance of fluid solidification against the wall?

 

[jte]

Figure out how the pipe added thickness, patent the method, sell it, retire wealthy!

Are the readings taken at the exact same thickness measurement location? If at the same pup piece, but different orientations then the readings are within mill tolerance. One side could be a bit thin, while the other is a bit thick. This is more likely with seamless pipe, of course. The 5.8mm reading is an outlier, it is possible you have localized corrosion / pitting.

As BigInch says, not likely that the pipe temperature is causing this amount of change. What material, and what temperature?

 

[SJones]

Clearly, there is a need to improve the reliability of data collection at the facility. Start with developing a detailed inspection procedure, including equipment selection for elevated temperature, and qualify it to a level of acceptable reliability. Then qualify the technicians and ensure appropriate supervision. Finally, ensure that inspection drawings are accurate and up to date. You should be able to go back to the 5.8 mm location and perform a verification, rather than ask on a forum if it is an outlier.

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/pub/8/83b/b04

All answers are personal opinions only and are in no way connected with any employer.

 

[MeltedSnow]

Thanks for all the helpful responses! The pipeline is carrying gas, so no fluid solidification. I did ask about a double check before we started making progress towards replacing the pipe - but I'm only here until September and the earliest another UT check can be performed is December!

 

[JR97]

MeltedSnow,

Elevated temperature is NOT the source of your error. As temperature increases so does the velocity of sound which would decrease the apparent thickness of the pipe. A more likely cause is the instrument was calibrated on a material with a different velocity. If the results are "too thick" that would indicate the material used for calibration had a faster material velocity. For example, the velocity of sound in wrought iron is about 20% faster than in cast iron. IF you can determine what material was used to calibrate the instrument (and it is in fact different from the pipe material) you can calculate the error in thickness as the ratio of the two velocities.

JR97

 

[SectionIX]

The drawing could be wrong, the instrument/transducer could be faulty, or, with all due respect, your technique (calibration and inspection) may have contributed to the erroneous results. And as Big Inch has mentioned, your data points may not have been the same from year to year, contributing to inconsistent readings. The 5.8mm/5.11 mm readings could have been internal corrosion, as has been pointed out, or possible lamination.

 

[MeltedSnow]

Thanks for all the helpful information! Did some more digging and it looks like the readings are taken from random points across the line - the reason the pipe looked as if it was getting thicker or thinner each year was because the *minimum* thicknesses were tabulated over the years - regardless of where in the line these readings were taken! After finding the original documents the numbers were compared and it all appears to line up. Good to know that heat can't affect UT thickness readings!

 

[jte]

MeltedSnow-

Thanks for getting back to us to close the loop. Always nice to know what the solution to the mystery is!