Issue
This safety alert highlights that the non-destructive testing (NDT) process is vulnerable to tampering. HSE and other regulators have investigated several instances where NDT has been falsified. In particular, radiographic images of welds have been falsely obtained or tampered with, including:
duplicate images of welds and false identification numbers added after processing
suspected accelerated working by shortening the distance to the radiographic source
potential misrepresentation of image quality
images cropped to remove defects in adjacent welds
In these cases there is a significant risk of flaws and defects going undetected with a subsequent risk of structural failure.
Outline of problem
Metallic structures fabricated by welding are prone to defects, such as inclusions, porosities and cracking. The absence of significant defects is assured, at the construction phase, by non-destructive testing (NDT) of the welded joints. This is often achieved by radiography. The weld can be examined for internal defects using an X-ray or radioactive source and a suitable medium (plate or film).
Joints not meeting acceptance criteria are usually ground out, re-welded and examined again to ensure the repair has been effective. The quality assurance at the fabrication stage is vital to ensure the initial integrity of structures such as pressure vessels, pipework and tanks, preventing catastrophic failure when in service.
Falsification of NDT
HSE has investigated a number of instances where the NDT has been falsified, mainly at the fabrication stage. Welds have not been thoroughly tested and/or the results have been misrepresented.
Most recently, during a large-scale steam boiler construction which required hundreds of internal tube welds and thousands of radiograph images, instances of film duplication were noticed by the inspector. Further investigation revealed large-scale falsification of weld images and shortcuts taken to reduce the time spent on site:
radiographs were produced without unique identification being affixed at the time of exposure. Identification numbers were ‘flashed’ on to the film during post-exposure processing, contrary to the requirements of the standard in place at the time (BS EN ISO 17636-1:2013), allowing falsification
multiple exposures were taken of the same weld at once (by film double loading). While this is allowed in the standard, it is intended for use when establishing exposure variables and allows selection of the best image from a number of films. It may also be used for examining a larger cross section of thicknesses in one exposure. However, in this case, the spare images generated by double loading the film had false numbers added during image processing.
Other indications of gross malpractice were ignored, most notably the large number of welds radiographed in one working shift. Estimates of capability varied from between 10 welds to 20 welds per shift. But numbers regularly exceeded this, with one instance of 104 weld butts (312 images) supposedly undertaken. Neither the NDT company management, the main contractor, the customer, or the Notified Body recognised the implications of the large number of radiographs recorded.
There were other issues around competence of those verifying the work. The person appointed was qualified as a weld inspector, but not sufficiently qualified to interpret radiographs.
HSE is also aware of another example. Some film radiographs were marginally smaller than the majority submitted for approval post-fabrication. Further scrutiny revealed that:
the shorter negatives had been cropped by 10-15 mm
those negatives had captured part of the tank construction deemed to be exempt from radiography, because of access difficulties
the cropped sections contained defects that would require concession or, in some cases, cutting out and repair.