|Publisher||Society of Petroleum Engineers||Language||English|
|Content Type||Journal Paper|
The Use of Finite-Element-Calculated Magnetic-Flux-Leakage Signals To Study the Characterization of Defects in Steel Coiled Tubing
Chase Breidenthal, Gabriel E. Larín, Steven M. Tipton, and Roderic K. Stanley, The University of Tulsa Coiled Tubing Mechanics Research Consortium
|Volume||Volume 14, Number 1||Pages||pp. 120-127|
2009. Society of Petroleum Engineers
1.2 Drilling Design and Analysis
A primary consideration with coiled tubing (CT) is that it is consumed by fatigue loading during routine operations. Also, rugged oilfield conditions routinely lead to corrosion and other mechanical surface damage. Since fatigue is a surface phenomenon, the presence of a surface imperfection has a significant influence on fatigue-damage mechanisms. This paper describes the study of magnetic-flux-leakage (MFL) inspection signals caused by surface defects in the form of milled circumferential grooves in steel CT. The focus of the investigation is to identify and estimate the size of surface defects on the basis of characteristic MFL signal features. It is demonstrated that this effort is greatly enhanced by finite-element analysis (FEA). The ultimate objective is to extract surface-flaw dimensions accurately from conventional MFL signals. These dimensions are used in computer CT life-prediction models.
An axisymmetric FEA model is developed and used to calculate leakage flux density solutions for milled circular and rectangular shaped grooves in 1.75-in.0outside0diameter (OD), 0.156-in.-wall-thickness (WT), 90-ksi CT samples. FEA results are compared to axial and radial MFL signals measured with an experimental inspection unit. Favorable agreement is observed between experimental and FEA data. Furthermore, signal features are correlated with the known slot geometries to identify basic geometry-recognition patterns for different circumferential grooves. Signal features reveal qualitative and quantitative trends relative to surface-flaw dimensional characteristics.
The need persists to make the operator’s string-management decision-making process more reliable and automatic with respect to determining fatigue life expectancy. The obstacle here is that because of the inherent inaccuracies in commonly used MFL inspection techniques, reliable real-time flaw-evaluation and characterization capability is limited.
End users of CT are keenly aware--and much has been published concerning this matter--of the fact that bending fatigue is one of the primary threats to in-service CT integrity and has tended to impede the progress of CT usage. Fatigue problems are further intensified by rugged oilfield conditions that routinely cause corrosion and other surface damage such as scratches, nicks, gouges, dents, and impressions. Since fatigue is a surface phenomenon, the presence of a surface imperfection can accelerate fatigue-damage mechanisms and reduce the useful life of CT significantly.
Current CT integrity-assurance and string-management needs are being driven by usage demands energized in part by reliable analytical fatigue life-prediction models (Tipton et al. 2002), the more severe service conditions accompanying the progression into deeper and higher-pressure wells (McCoy et al. 2002; Stanley 2005), and the strong outlook for existing and future conventional well applications for which CT intervention can provide unrivalled solution (Adam 2003). Consequently, the number of CT strings in service and user inventories is growing rapidly. The implication is that the major need of end users is increased confidence and reliability of CT, whether new or used.
The natural responses to the emergence of the CT industry have been the development of technologies to identify problematic conditions and to make the operator’s integrity-assessment process and string-management decisions more efficient and automatic. The detection and characterization of metal loss represents an essential part of a CT integrity-monitoring system (Stanley 1996). Inspection techniques have been adapted from the pipeline and oilfield-tubulars sectors, where MFL has become the most popular means of detecting flaws in CT. MFL has been selected over ultrasound because it is a technique that is not affected seriously by the state of the pipe surface. However, string-management decisions based on the indications from MFL inspection still require manual proving to characterize the defect that caused the MFL response (Stanley 2005; Moran et al. 2002; Stanley 2004a; Stanley 2004b).
Sources such as Moran et al. (2002), Stanley (2005), and others (Stanley and Varner 1998; Stanley 1996; Stanley 2004b; Rosen 1997; Rosen 1998; Stanley 2001) document and describe the challenges facing the development of CT-inspection technology. To address these challenges, experimental MFL inspection of CT specimens with flaws of known geometries permits systematic examination and mapping of the signals to determine how much geometric information can be obtained from magnetic nondestructive testing. At the University of Tulsa, this process is performed by means of a simple and compact bench-top MFL inspection unit with the benefit of repeatability.
This paper describes the measured MFL inspection signals obtained from the bench-top inspection unit for circumferential flaws of known geometry. Furthermore, a 2D axisymmetric FEA model is described that is used to conveniently simulate this technique and reproduce the axial and radial MFL signals. The FEA predictions are compared to the experimental data.
|File Size||1,197 KB||8|