The oil and gas industry in North America operates with an ageing infrastructure of pipelines, 70% of which were installed prior to 1980, and with 50% of those installed during the 1950s and 1960s. While there is a growing demand for knowledge of pipe properties in order to determine safe operating pressure, records are incomplete in a significant number of cases.
Current in-line inspection (ILI) technologies focus on defect detection and characterisation for pipeline testing
, such as corrosion, cracking and the achieved probability of detection (POD). As a part of the process of assessing defect significance it is necessary to know the pipe properties
in order to determine potential failure limits. The mechanical properties (yield strength, tensile strength and fracture toughness) of steel pipe must be either known or conservatively estimated if there is to be an appropriate and safe response to the presence of detected defects, and in setting operating pressure. Material property measurements
such as hardness, chemical content, grain size, and microstructure can likely be used to estimate the mechanical properties of steel pipe without requiring cut-outs to be taken from pipes for destructive tests.
In-ditch methods of inspection are available, or in development, that can potentially be used to determine many of the material characteristics and some of the mechanical properties. Potential ILI data may also be used to obtain certain information. Advances in ILI technologies for this purpose are being explored by several interested parties, with ILI companies specifically focusing on relating magnetic measurements from eddy current and magnetic flux leakage measurements to mechanical properties. ILI also regularly uses ultrasound measurements to determine wall thickness.
Potential applications of advances in ultrasound measurements for grain size and other properties are also being explored. Commonly used non-destructive methods of inspection do not typically allow for the determination of either the material or mechanical properties, making destructive testing the only viable option in most cases. This limitation is costly, time-consuming and often impractical for pipes in use.
This paper reviews ILI and in-situ techniques and presents an analysis of a sample set of data. It explores the possibility of obtaining mechanical property data from data potentially measurable by ILI and in-situ measurements. Ideally, the results would allow mechanical property measurements to assess pipelines, thereby ensuring the proper response, at a specific operating pressure, to anomalies that might pose a significant threat.
The use of a multivariate regression analysis yielded better results than the traditional two-variable regression plots, and may be key in determining the properties necessary to provide the best results to reliably estimate the mechanical properties of pipe. However, there is still much work to done in understanding and accounting for the many sources of variability within the pipe material, and its impact on the relationships between mechanical and material properties.