Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.733250
Title: Current deflection NDE for pipe inspection and monitoring
Author: Jarvis, Rollo
ISNI:       0000 0004 6496 989X
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
The detection of corrosion on insulated and/or coated pipes in the oil and gas industry remains a challenge. Routine inspection, which is commonly achieved with in-line tools known as "pigs", is not possible where there is any risk of the pig becoming stuck. There are thousands of kilometers of pipe worldwide deemed ''unpiggable'' whose safety must be ensured using Non-Destructive Evaluation (NDE) external to the pipe if potentially catastrophic failure is to be avoided. Many NDE techniques lack sufficient sensitivity due to the coating thickness producing a high standoff distance between the pipe and the sensor and therefore require costly and time-consuming removal of the coating. A method capable of detecting and/or monitoring of defects (e.g. one-third-wall depth corrosion) while leaving the insulation/coating intact would be highly attractive. This thesis documents the development of a technique in which a low-frequency AC current is directly injected into the pipe at distant locations, and perturbations in the magnetic field caused by "current deflection" around defects are measured using solid-state magnetic sensors. Two methods of applying this novel technique were investigated. Firstly, scanning the sensors to measure perturbations in the field and screen for defects, and secondly, permanently installing sensors outside the pipe for Structural Health Monitoring (SHM). A Finite Element (FE) model has been developed and used to investigate the practical challenges that are faced by the technique and how these may be overcome. The sensitivity of the technique for defect detection by external pipe scanning in a practical scenario has then been evaluated using a model-assisted Probability of Detection (POD) framework that combines the measurements of the signal from an undamaged pipe with synthetic damage profiles and contributions from general corrosion and sensor misalignment. The results indicate that good performance is expected for damage detection by scanning above a typical insulation thickness with just a few amps of injected current. A similar framework has then been used to evaluate the sensitivity of the technique as an SHM solution which suggests excellent corrosion detection performance with the permanent installation of inexpensive magnetic sensors. The technique has potential advantages over competing methods in both scanning and monitoring modes and there are many opportunities for future development.
Supervisor: Cawley, Peter ; Nagy, Peter B. Sponsor: Engineering and Physical Sciences Research Council ; British Petroleum Company
Qualification Name: Thesis (D.Eng.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.733250  DOI:
Share: