On Detecting Internal Coating Defects in Pipes using EMAT

Abstract

Pipeline integrity is a must to maintain to ensure safe and smooth operation. Therefore, periodic inspection is required to make timely decisions for replacement or repair. One of the problems faced in a pipeline network, and targeted by this work is internal coating defect detection, whereby it is required to locate and characterize damage in the internal coating of the pipe. Current industry practice is to replace an entire set of pipelines or its large segment due to unknown defect location, implying significant wastage. A widely used NDT technology, electromagnetic acoustic transducer (EMAT) is employed in this laboratory-based work, however, it is a challenging task to extract internal coating damage information from the received acoustic signal as it has low SNR and is highly non-linear and non-stationary, thus warranting the use of sophisticated algorithms for data analysis. Artificial defects in the form of 100% tape-wrap coating removal are introduced with dimensions 100 x 100 mm2, 30 x 30 mm2, and 10 x 10 mm2 (square-shaped). The Carbon-Steel pipe sample has a 12” diameter and 10 mm wall thickness. The objectives are to externally inspect the pipe sample using commercially available hand-held EMAT sensors with Lamb modes, and develop and employ Singular Spectrum Analysis (SSA) algorithm to characterize the internal coating defects. Only the largest defect is detected, whereby A1 Lamb mode shows precise location and shape, and acceptable sizing accuracy compared with A0, S0, and S1. Moreover, the results are repeatable over several scans, thus showing the effectiveness of the adopted methodology. The main novelty lies in the fact that the methodology does not require any modifications to the hardware of the EMAT setup, thus increasing the capability of the available hand-held equipment. Current and future work is focused on the detection of smaller defects.