Detection, characterization and sizing of hydrogen induced cracking in pressure vessels using phased array ultrasonic data processing

Pressure vessels operating in sour service conditions in refinery environments can be subject to the risk of H₂S cracking resulting from the hydrogen entering into the material. This risk, which is related to the specific working conditions and to the quality of the steel used, shall be properly managed in order to maintain the highest safety at a cost-effective level. Nowadays the typical management strategy is based on a risk based inspection (RBI) evaluation to define the inspection plan used in conjunction with a fitness for service (FFS) approach in defining if the vessel, although presenting dangerous defects such as cracks, can still be considered “fit for purpose” for a given time window based on specific fracture mechanics analysis. These vessels are periodically subject to non-destructive evaluation, typically ultrasonic testing. Phased Array (PA) ultrasonic is the latest technology more and more used for this type of application. This paper presents the design and development of an optimized Phased Array ultrasonic inspection technique for the detection and sizing of hydrogen induced cracking (HIC) type flaws used as reference for comparison. Materials used, containing natural operational defects, were inspected in “as-service” conditions. Samples have then been inspected by means of a “full matrix capture” (FMC) acquisition process followed by “total focusing method” (TFM) data post processing. FCM-TFM data have been further post-processed and then used to create a 3D geometrical reconstruction of the volume inspected. Results obtained show the significant improvement that FMC/TFM has over traditional PA inspection techniques both in terms of sensitivity and resolution for this specific type of defect. Moreover, since the FMC allows for the complete time domain signal to be captured from every element of a linear array probe, the full set of data is available for post-processing. Finally, the possibility to reconstruct the geometry of the component from the scans, including the defects present in its volume, represents the ideal solution for a reliable data transferring process to the engineering function for the subsequent FFS analysis