Corrosion loop development of oil and gas piping system based on machine learning and group technology method

Abstract

Purpose Corrosion loop development is an integral part of the risk-based inspection (RBI) methodology. The corrosion loop approach allows a group of piping to be analyzed simultaneously, thus reducing non-value adding activities by eliminating repetitive degradation mechanism assessment for piping with similar operational and design characteristics. However, the development of the corrosion loop requires rigorous process that involves a considerable amount of engineering man-hours. Moreover, corrosion loop development process is a type of knowledge-intensive work that involves engineering judgement and intuition, causing the output to have high variability. The purpose of this paper is to reduce the amount of time and output variability of corrosion loop development process by utilizing machine learning and group technology method. Design/methodology/approach To achieve the research objectives, k-means clustering and non-hierarchical classification model are utilized to construct an algorithm that allows automation and a more effective and efficient corrosion loop development process. A case study is provided to demonstrate the functionality and performance of the corrosion loop development algorithm on an actual piping data set. Findings The results show that corrosion loops generated by the algorithm have lower variability and higher coherence than corrosion loops produced by manual work. Additionally, the utilization of the algorithm simplifies the corrosion loop development workflow, which potentially reduces the amount of time required to complete the development. The application of corrosion loop development algorithm is expected to generate a “leaner” overall RBI assessment process. Research limitations/implications Although the algorithm allows a part of corrosion loop development workflow to be automated, it is still deemed as necessary to allow the incorporation of the engineer’s expertise, experience and intuition into the algorithm outputs in order to capture tacit knowledge and refine insights generated by the algorithm intelligence. Practical implications This study shows that the advancement of Big Data analytics and artificial intelligence can promote the substitution of machines for human labors to conduct highly complex tasks requiring high qualifications and cognitive skills, including inspection and maintenance management area. Originality/value This paper discusses the novel way of developing a corrosion loop. The development of corrosion loop is an integral part of the RBI methodology, but it has less attention among scholars in inspection and maintenance-related subjects.

Integrating inspection management (IM) in piping system of petroleum industry in Iran

“…3. Importance of piping system in oil and gas industry All oil and gas exploration and processing projects should be supported with a tailored strategy for their maintenance and inspection requirements and facilities should be started system (FMS) (Ahluwalia, 2008) Piping system of petroleum industry up with their maintenance and inspection program properly defined by function and should be ready in systems that will support the execution of that work (Rachman and Ratnayake, 2020;Keprate and Ratnayake, 2016). These are typically big support activities covering all equipment and their sub-assemblies.…”

A Survey on Industry 4.0 for the Oil and Gas Industry: Upstream Sector

“…The uses of AI has been employed to model the scour pattern around submerged pipes located in sedimentary beds [195], use of BD analysis for safety factors in the pipelines [196], use of ML to automate and reduce variability in manual output in the development of corrosion loop for pipes [197]. AI has also been applied in the prediction of optimum wellhead choke size which determines the flow rate in pipelines [198] and choke flow coefficient for both nozzle and orifice type chokes with adequate precision [199].…”

Risk-Based Approach for Improving Concrete Bridges‘ Inspection Planning

“…In order to develop optimal inspection procedures, the concept of group technology has been used, as it has the potential to increase the efficiency of the inspection procedure (Rachman and Ratnayake, 2019). Group technology is widely used in the manufacturing industry to classify the system into subsystems and subdivide them into part families, based on design attributes and manufacturing similarities.…”