Sizing Crack Indications from Ultrasonic ILI: Challenges and Options

Proceedings Publication Date:

01 Aug 2017
Presenter
Herbert Willems
Presenter
Company
Author
Herbert Willems, Rogelio Guajardo, Gerhard Kopp
Part of the proceedings of
Abstract
Ultrasonic in-line inspection (ILI) tools for crack inspection using piezoelectric transducers are commercially available for more than two decades. These tools have proven very successful for the detection of various types of cracks or crack-like anomalies present in many pipelines worldwide. While the probability of detection (POD) of these tools is considered to be generally sufficient the sizing of crack indications has some inherent limitations related to the inspection technology. The inspection itself is performed by using 45° shear waves in the pulse-echo mode and recording the reflections obtained from cracks or crack-like reflectors. In particular, the depth sizing of cracks is thus primarily based on the measurement of signal amplitudes. The signal amplitude, however, depends not only on the crack depth but also on additional parameters that need to be taken into account. Furthermore, the reflection amplitude is saturated for crack depth above a certain, probe dependent value that limits the sizing capabilities especially for thicker pipe wall. The objective of the paper is to explain to the users of ultrasonic ILI on a physical basis the potential and limits of crack sizing when using amplitude based crack inspection technologies. In order to do so, the principle of the inspection technology and the procedures used for length and depth sizing are explained in detail. The influence of a variety of variables (crack depth, inspection angle, tilt angle, skew angle etc.) on the reflection signal is examined by using experimental results as well as modelling results. Several options that may be helpful to improve depth sizing for ultrasonic ILI are discussed.

To view the video or download the paper please register here for free

You already have access? Sign in now.