Unreliable measurements of the remaining wall thickness of pipes for some forms of external corrosion when using tangential (profile) radiography

Print this pagePrint this page
Rating: 
Average: 3 (2 votes)
Alert ID: 
03667
Description of Task: 

The detailed report is available here: http://www.esrtechnology.com/centres/hois/Pages/Safety-Alert.aspx

 

When performed in-service, due to site constraints in positioning the equipment, tangential (profile) radiography failed to identify the presence of corrosion which had caused loss of containment at a trunnion pipe support. In other cases, the severity of some examples of external corrosion were significantly underestimated using this technique, even when applied with optimum equipment positioning. Other corroded areas with different characteristics were sized to within about ±0.5mm

 

Lessons Identified: 
  • The tangential radiography technique is only capable of providing measurements of the wall thickness of a pipe at positions very close to the tangent position(s), which depend on the exact positions of the source and detector. 
  • For some forms of localised external corrosion, which are misaligned from the tangent position by as little as 10-20°, the resulting measurements of wall thickness may be significantly higher than the actual minimum value, and may even miss the wall loss completely.
  • Even with optimum positioning of the source and detector, some examples of external corrosion can be significantly undersized. In one case in which the corrosion extended over 30° of the circumference of a 6" pipe, covered by a 6mm high scab, the resulting radiographs were difficult to interpret and using the location of the apparent outer edge of the pipe wall in the image, under-sizing was as much as ~1.5mm (the radiography gave a minimum ligament of 3.4mm, compared with an accurate value of only 1.9mm). Alternative interpretations of the radiographs, using features within the pipe wall may overestimate the extent of the wall loss. In another example, the external corrosion resulted in fine pitting, superimposed on larger scale wall loss. The fine pitting was not included in the resulting measurements of remaining wall thickness which again resulted in substantial (~1.3mm) under-sizing of the corrosion when using tangential radiography.
  • Careful interpretation of the radiographic images of the pipe wall under localised corrosion scabs, and those scabs that cover fine pitting, can show characteristic features that are indicative of the likelihood of complex interpretation and the risk of significant under-sizing but does not, on the basis of current knowledge, provide a means for improved sizing.
  • For more circumferentially extended and/or less deep corrosion, without any fine pitting, these issues are less significant and a typical accuracy of ±0.5mm is likely to be achievable using optimal sizing techniques applied to digital radiography images.
Recommendations: 
  • All in-service radiography of pipes should be performed in accordance with the newly published European standards EN 16407 parts 1 and 2, and Issue 2 of the HOIS recommended practice for in-service computed radiography, which can be downloaded from www.hoispublications.com.
  • If localised external corrosion is known to be present, extending over a small fraction of the pipe circumference, when performing tangential (profile) radiography, care must be taken in aligning the source and detector such that the tangent position is as close as possible to the centre of the corroded area or the highest point of the scab (i.e. the point of apparent greatest severity of the corrosion).
  • Exposures at different angles are needed either side of the apparent position of greatest severity of the corroded area to find the point of maximum wall loss/minimum thickness. The angular increment for these radiographs depends on the pipe diameter and corrosion morphology/circumferential extent. Typical values may be as small as 5° for significant corrosion extending over only a small fraction of the circumference of a 6" OD pipe, and 10-15° for more extended corrosion on small bore piping.
  • If any fitness for service (FFS) assessments are performed based on the measurements from tangential radiography of localised external corrosion, a substantial tolerance should be applied with the minimum ligament for FFS being less than that measured. There is limited information currently on the magnitude of this tolerance, but a value of about 1.5 mm is indicated from an example of a 6mm high external corrosion scab extending over ~30° of the circumference in a 6" OD pipe. A similar value is indicated for an example of corrosion that gave a finely pitted surface, superimposed on larger scale wall loss. In safety critical cases, for radiographs showing evidence for under-sizing of the corrosion, it may not be appropriate to use tangential CR measurements for FFS assessments.
  • In other cases, involving more extended and/or shallower corrosion without any fine pitting, smaller tolerances of about 0.5mm may be possible, provided the sizing method is based on careful analysis of grey level profiles taken from digital radiographic images. This smaller tolerance could only be justified if the corresponding radiographs have been assessed by a suitably trained and competent operator to be free of any of the features which can cause significant under-sizing (structure in the pipe wall image, irregular or poorly defined OD profile or evidence for fine pitting in the double-wall double image region of the image).

In any case where the measurements from tangential radiography of localised external corrosion are used to support live surface preparation by shot blasting, a substantial tolerance may need to be applied similar to that noted for FFS above.

The detailed report is available here: http://www.esrtechnology.com/centres/hois/Pages/Safety-Alert.aspx

 

Information source: 
See link above to web page
Contact Details: 
steve.burch@esrtechnology.com
Specific Equipment: 

Any pipe with external corrosion, inspected using tangential (profile) radiography, including at trunnion and other types of supports.

Incident consequence: 
HIPO
Activity: 
Inspection/testing
Location: 
Any Location Type
Cause of accident or incident: 
Other

Comments

Be-careful we are not in danger here of rubbishing an extremely useful, and sometimes the only technique for measuring wall thickness loss due to internal and external corrosion. This notice concentrates its title on a method issue, whereas the 'unreliability' was heavily due to the radiographers mis-interpretation.