Session: 01-01-03 Offshore Platforms-3

Submission Number: 183657

Experimental Characterization and Monitoring of Leaks in Dense-Gas/Liquid Two-Phase Pipe Flow Using Artificial Intelligence 

Leak detection remains a critical challenge due to the need of ensuring safe and efficient operation of offshore production systems, where failures in subsea lines or flexible risers can lead to severe environmental and economic consequences. This study presents the development, validation, and application of an experimental and artificial intelligence–assisted methodology for detecting, characterizing and monitoring leaks in flow conditions similar to those observed in ultra deepwater production, i.e., dense-gas/liquid flow with low liquid-gas density ratios. The experimental campaign was conducted using sulfur hexafluoride (SF₆) as the dense gas phase and mineral oil as the liquid phase. The test facility was designed to reproduce operating conditions representative of offshore pipelines, enabling controlled leak generation at different circumferential positions (top, side, and bottom), and continuous acquisition of pressure, differential pressure, flow rate (volumetric and mass), and temperature signals under a wide range of flow conditions and leak sizes. A gamma-ray densitometer was employed to determine the liquid holdup distribution along the test section, and a high-speed camera system was used to visualize the two-phase flow patterns. From these experiments, a comprehensive database was constructed, integrating hydrodynamic and thermodynamic variables, visual data, and system responses under both intact and leaking conditions. This dataset supports the development of artificial intelligence models for identifying and classifying leak scenarios based on transient flow behavior. The proposed framework aims to integrate data-driven learning techniques to detect the onset and intensity of leaks in real time. The approach combines advanced signal monitoring, multiphase diagnostics, and data analytics to enhance the reliability of leak detection in multiphase flow transport systems. Its outcomes are expected to contribute significantly to flow assurance, integrity management, and environmental protection in offshore oil and gas production, thereby providing a foundation for next-generation intelligent monitoring systems for flexible pipelines and subsea infrastructure.

Presenting Author: Carlos Mauricio Ruiz Diaz University of São Paulo

Presenting Author Biography: Carlos Mauricio Ruiz Díaz is a Ph.D. candidate in Mechanical Engineering at the São Carlos School of Engineering (EESC) of the University of São Paulo (USP), Brazil. He holds a Master’s degree in Mechanical Engineering from the Industrial University of Santander (UIS), Colombia, where he received an Honorable Mention for the excellence of his research project.

He works as a researcher at the Industrial Multiphase Flow Laboratory (LEMI), where he serves as principal investigator in two projects supported by Petróleo Brasileiro S.A. (Petrobras). His main expertise lies in the analysis and characterization of dense gas/liquid and liquid–liquid two-phase flows, as well as in flow assurance studies, applying artificial intelligence and data-driven modeling to experimental data.

Carlos has participated in research projects funded by both public and private institutions in Colombia, Brazil, and Norway, including his recent collaboration as a visiting researcher at the Norwegian University of Science and Technology (NTNU) and SINTEF Energy Research, where he supported two experimental projects on multiphase flow monitoring and cold-flow systems.

He has presented his work at more than 20 national and international conferences, and his publication record includes 1 book, 1 book chapter, 14 papers in indexed journals, and 2 registered software systems.

He is an active member of the Energy and Environment Research Group (GIEMA) and the Research Group on New Technologies, Sustainability, and Innovation (GINSTI). Currently, he serves as Membership Chair of the Society of Petroleum Engineers (SPE) Student Chapter at USP and lectures in the Postgraduate Program in Industrial Automation at Francisco de Paula Santander University – Ocaña Campus (UFPSO).

He received the Best Paper Award at the SPE Brazil Flow Assurance Technology Congress (Rio de Janeiro, 2024) and at the International Mechanical, Manufacturing and Process Technology Conference (IMRMPT 2021 – Medellín, Colombia). Additionally, he is the co-founder of the Artificial Intelligence Subgroup (SIA) within the GIEMA research group.

Authors:

Carlos Mauricio Ruiz Diaz University of São Paulo
Johann E. Castro-Bolivar University of São Paulo
João M. A. M. Fróes University of São Paulo
Gustavo Bochio University of São Paulo
Rodrigo Galvão D'império Teixeira Petrobras
Oscar M. H. Rodriguez University of São Paulo