Abstract
Detection and sizing of surface breaking defects in pipelines is of vital importance in the petrochemical industry. This work focuses on non-contact, in-service inspections of large diameter (above 24”) pipes with ultrasonic phased arrays. The aim is to propose a data acquisition and processing technique, applicable onto oil pipeline inspection gauges (pigs) performing inline inspections. Pigs complete inspections at high velocities (1-2 m/s), resulting in limited acquisition time at a specific location. Hence, a suitable ultrasonic technique will yield precise results in immersion with few transmissions.Plane Wave Imaging (PWI) is an ultrasonic array imaging algorithm shown to meet the requirement in components with simple geometry. An improved PWI Adapted in Post-Processing (PWAPP) is developed, suitable for inspections through non-planar surfaces. It processes the captured data in two stages. The first reconstructs the surface of the component, the latter images inside of it by adapting the delays to the distortion of the plane waves upon refraction at the interface. PWAPP is found to outperform conventional PWI in terms of signal-to-noise ratio (SNR) by up to 10 dB on immersed concave samples. Importantly, even with a number of transmissions reduced to a quarter of the array elements, PWAPP maintains its high SNR.
Analytical and image-based surface reconstruction techniques are investigated. These operate on the same ultrasonic data used for the interior inspection to avoid extending the data acquisition time. The analytical algorithms, defined through a geometric envelope, produce a wider field of view and comparable or higher front surface accuracy than the equivalent, industry-accepted, image-based approaches. Reconstruction techniques allow probe misalignment and component complexity to be accounted for, increasing defect detectability in the consequent interior images. Using a number of transmissions equivalent to a quarter of the array elements, PWAPP detects all considered defects through an ultrasonically measured 2.5 mm weld cap.
Back-surface-breaking defects are depth sized directly from plane wave images. The extracted measurements are within 10% error for defects simulated with depths from 2 to 6 mm and angles from -5° to 20°. The sizing accuracy is shown stable within a reasonable range of inspection parameters such as number of transmissions and array lateral (pipeline radial) position. Notably, perpendicular to the surface defects in pipe sections above 24” diameter, can be sized from an experimental PWI formed with a single transmission. The presented results indicate plane wave data acquisition and PWI post-processing should be considered as an alternative to the methods currently utilised onto oil pigs.
| Date of Award | 2 Dec 2021 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Paul D Wilcox (Supervisor) & Alexander Velichko (Supervisor) |