Developing an SCR Digital Twin Using Machine Learning
Machine learning is an enabling technology that can be used in the structural design, integrity management and life extension of riser systems, to better predict in-service response faster and with less input data streams. While in-field riser strain or motion monitoring provides direct indication of riser fatigue and strength performance, direct structural monitoring of these systems has been limited in offshore deployments and is not often used for every riser system on a production vessel. This is due to the high perceived capital or operational costs of direct subsea monitoring when deployed on every riser to reliably monitor over the entire service life. A lack of known structural response in service can result in qualitative inspection planning, conservative fatigue predictions and reduced asset utilization. Machine learning provides an alternative, low cost approach for life-of-field riser monitoring and predictive inspection planning.
1 Nov 2019
Author

Shankar Sundararaman
Senior Principal Engineer

About
Shankar Sundararaman is a Senior Principal Engineer based in 2H’s Houston office. He has over 17 years' experience in the oil and gas industry, including design and analysis of riser systems (drilling and completion risers, top tensioned risers, steel catenary risers, steel lazy wave risers, flexible risers), tendons, jumpers (rigid and flexible), machine learning (including digital twins) and cloud computing, structural health monitoring, drill string systems, drilling engineering analysis, and vibration analysis. He has over 20 years’ engineering and R&D experience with numerical analysis, scientific computing, and finite element analysis.
Shankar has an undergraduate degree in Naval Architecture from IIT Madras, masters and PhD degrees in Mechanical Engineering from Purdue University and is a Licensed Professional Engineer (Naval Architecture) in the State of Texas. He has 5 granted patents, over 30 publications in international journals, conferences, and book articles, and has presented at several international conferences.
Insights
During this webinar, we will describe the use of machine learning to develop a digital replica of real-life riser systems to aid in operational decision-making. A steel catenary riser (SCR) structural digital twin is developed to calculate stresses in the time domain at the fatigue hot spots and high stress locations along the riser, including the touchdown zone (TDZ), based on the platform motions and environmental loads that are typically observed in the field.
Learning Outcomes:
How the application of machine learning can help to provide a more accurate assessment of the remnant life of risers and assist with life extension and operational planning
How to develop an SCR structural digital twin at "hot spot" locations along the riser using simulation (FEA) data, measured platform motions and environmental loads
How to assess the accuracy of fatigue life predictions from a machine learning model
This webinar is ideally suited for riser engineers, asset integrity engineers/managers and those wanting to increase their knowledge in how to approach life extension of dynamic offshore structures.
Author

Shankar Sundararaman
Senior Principal Engineer

About
Shankar Sundararaman is a Senior Principal Engineer based in 2H’s Houston office. He has over 17 years' experience in the oil and gas industry, including design and analysis of riser systems (drilling and completion risers, top tensioned risers, steel catenary risers, steel lazy wave risers, flexible risers), tendons, jumpers (rigid and flexible), machine learning (including digital twins) and cloud computing, structural health monitoring, drill string systems, drilling engineering analysis, and vibration analysis. He has over 20 years’ engineering and R&D experience with numerical analysis, scientific computing, and finite element analysis.
Shankar has an undergraduate degree in Naval Architecture from IIT Madras, masters and PhD degrees in Mechanical Engineering from Purdue University and is a Licensed Professional Engineer (Naval Architecture) in the State of Texas. He has 5 granted patents, over 30 publications in international journals, conferences, and book articles, and has presented at several international conferences.