Session: 07-02-03 Geohazards and Mechanical Damage
Paper Number: 87211
87211 - Machine Learning-Based Severity Assessment of Pipeline Dents
Dents are a common mechanical damage threat that could lead to delayed pipeline failures. Dents are detected through in-line inspection (Caliper) of buried oil and gas transmission pipelines. One challenge to pipeline operators is to identify potentially injurious dents among thousands of reported deformation features using limited information (e.g., reported dent’s length, width, and depth) and to prioritize the efforts and allocate the resources to obtain additional more detailed information (e.g., dent profiles) for those potentially severe dents. An innovative approach based on machine learning predictions stemming from a representative dictionary of finite element analysis (FEA) generated prototypes was developed. The proposed approach predicts multiple severity-based indicators for each dent, then combines them in an overall severity score, and finally is used to prioritize the acquisition of dent profiles. Once the dent profiles are available, detailed FEA quantitative reliability analyses, following previously developed and published methodology (QuAD), is performed allowing pipeline operators to confirm dent’s severity more accurately and drive a safe and efficient integrity program.
Four severity indicators were considered herein and intended to address both formation-induced and service-induced failure mechanisms. A formation-induced failure refers to pipeline failures due to cracks that were formed during the indentation stage. In the proposed approach, the maximum dent formation plastic strain and accumulated ductile failure damage were used for evaluating the likelihood of forming a crack during indentation. A third indicator, the volume of highly strained material, was also used to provide additional information on the dent formation cracking potential. The fourth indicator was the stress concentration factors (SCFs) to assess the potential of service-induced failure due to fatigue.
A total of ~4000 dents prototypes dictionary was generated and analyzed using FEA. These dents prototypes were specifically selected to provide good representation of the actual characteristics of real dents known throughout a gas transmission pipeline network in North America. Machine learning, as an emulator, was then trained and tested using the FEA-based dents prototypes and showed that the ML model can effectively predict the dent severity indicators previously referred to. These predicted dent severity indicators are then combined to produce an overall severity score, which was finally used to prioritize the acquisition of the detailed dent profiles. Once profiles are obtained, detailed FEA quantitative reliability assessments will ultimately confirm the severity and hence drive repair/no repair decisions, enabling in this way an efficient and effective allocation of resources.
Presenting Author: Huang Tang Enbridge GTM
Presenting Author Biography: Huang Tang received his bachelor’s degree in mechanical engineering from Tsinghua University (Beijing) and his Ph.D. degree in civil engineering from Carnegie Mellon University. He conducted postdoctoral research at MIT and Northwestern University with a focus on advanced modeling of mechanical behavior of materials. He joined ExxonMobil in 2007 and has over 13 years of experience with the company. At ExxonMobil, he worked heavily on the research and development of strain-based pipeline design technology, fitness for service assessment and hydrogen embrittlement and sour service. He led the development of ExxonMobil tensile strain capacity prediction tool for strain-based pipeline design. He provided general engineering support to ExxonMobil affiliates on pipeline integrity assessment and engineering critical assessment (ECA). He worked on multiple ExxonMobil pipeline projects conducting integrity assessment on pipeline dents, corrosion and low fracture toughness. Before joining Enbridge, he also worked at NASA Marshall Space Flight Center to support the Artemis lunar landing program on the fracture control of space launch systems.
Machine Learning-Based Severity Assessment of Pipeline Dents
Paper Type
Technical Paper Publication