Automatic Flaw Detection and Identification for Coiled Tubing
The purpose of this project is to develop an automatic detection algorithm to correlate between signal data acquired during coiled tubing (CT) inspections and physical flaws in coiled tubing samples. The automatic detection algorithm will be integrated with a previously developed lifetime prediction model to provide essentially real-time assessment of coiled tubing(CT) reliability.
Idaho National Laboratory (INL)
Idaho Falls, ID
University of Tulsa (TU)
Signal analysis was completed in FY2003. The project developed CT inspection software. The project identified initial correlation between magnetic flux leakage (MFL) signal features and manufactured flaws. Due to the signal analysis completion, a redesign of the sensor head began in FY2005.
The redesign of the sensor head includes more circumferential sensors. This will allow for better circumferential location of detected flaws in coiled tubing.
Current CT field inspection technology is relatively crude, consisting of rolling friction wheels to monitor depth and limited systems that monitor diameter and ovalit and MFL systems to identify defects but having little reliability in identifying the type of defect present (cracks, corrosion, etc.) or any dimensional information. CT drilling is operated in an extremely severe mechanical environment, where large bending strains are combined with significant internal pressure. This can cause diametrical growth, wall thinning, ovality, elongation, residual stresses, and low-cycle fatigue cracks. Surface defects can shorten operation life spans significantly.
The reliability of CT has been enhanced by extensive refinements to its manufacturing process. CT strings are now making as many as 80 trips into boreholes, resulting in a wide variety of service-induced defects. Additionally, there are often long periods of inactivity (storage) during which time corrosion can enhance the damage to the tube surfaces. This extended operational life of the tubing can be partly attributed to laboratory experimentation and theoretical work involving fatigue modeling. Sophisticated plasticity and fatigue-damage models predict life for discrete sections along the entire string throughout its service history. However, used CT tends to fail sooner than predicted, due to the presence of defects incurred through mechanical damage and/or corrosion.
The project tasks break out as:
- Task 1, algorithmic developments (FY2005-2006). This entails processing a CT MFL signal library and developing a flaw detection algorithm.
- Task 2, flaw identification engine (FY 2005). This task calls for developing a flaw characterization algorithm and the initial release of an automatic flaw detection, characterization, and acquisition software system.
- Task 3, evaluation (FY 2006). Here the researchers are to conduct blind validation studies of flawed CT strings.
- Task 4, integration (FY 2006-2007). The project performers are to report results from blind validation studies of flawed CT strings, develop a method to map the state of the CT string for TU's lifetime prediction model, and integrate a lifetime prediction model into data-acquisition/analysis system software.
- Task 5, project management (ongoing). This involves general project management, e.g., monthly reports, annual reports, etc.
Data acquisition was accomplished in FY2002. Researchers acquired a data system for inspection of CT and installed the INL-designed and -built interface on commercially available inspection heads. They acquired inspection data on virgin and flawed CT.
Current Status (July 2006)
The project was felt to contain patentable material, and no public information will be released until a disposition on the patent can be made.
Project Start: March 27, 2002
Project End: May 9, 2006
Anticipated DOE Contribution: $238,000
Performer Contribution: $165,000 (63% of total)
NETL - Virginia Weyland (firstname.lastname@example.org or 918-699-2041)
INL - Charles Tolle (email@example.com or 208-526-1895)
INL - David Weinberg (firstname.lastname@example.org or 208-526-9822)