Digital Twin for Operations - Present Applications and Future Digital Thread
Mat Podskarbi,David J Knezevic
DOI: https://doi.org/10.4043/30553-ms
2020-05-04
Abstract:Abstract We present a structural digital twin system based on component-based parametric reduced order modeling using Reduced Basis Finite Element Analysis (RB-FEA) framework. This enables fast, detailed, and holistic modeling of large-scale structures, such as floating and fixed production and drilling platforms, and crucially, the digital twin's parametric nature and fast solve time enables the system to be connected seamlessly with inspection and sensor data. We demonstrate RB-FEA based structural digital twins, and capabilities for connecting them to sensor data and data science methods to increase predictive power and engineering insight. We show how the digital twin is a key enabler for a Digital Thread – a data-driven architecture that links together information generated from across the product lifecycle. Inspection data, such as thickness measurements, can be directly mapped into the structural digital twin. Sensor data, such as motion and strain can be transformed into loading data, e.g. to estimate sea-states and calculate stress and fatigue accumulation based on actual conditions, or to "tune" model parameters to match data-based measurements, such as mode shapes and frequencies. This enables continuous tracking of the asset and assessment of its current state, based on quantitative and detailed structural analysis, at any time. Ultimately the structural digital twin provides an interpretation layer for integration of monitoring and inspection data, leading to a fully automated digital thread. We present a framework, based on a cloud-based simulation platform and Python-based scripting, to automate the workflows listed above. Digital twin provides a single source of truth about asset condition and allows data driven communication and decision taking between multiple stakeholders – operators, contractors, regulators, etc. We will show examples of real-world applications of RB-FEA digital twin for FPSO and Jacket platforms and development work for digital thread with emphasis on data driven value creation along the entire life cycle of the asset. The Reduced Basis FEA based structural digital twin is a key enabler of data driven digital thread in addition to sensor frameworks and data science systems. RB-FEA provides fast, parametric, and physics-based analysis providing a transparent interpretation layer for data obtained from other systems. It provides a key element allowing knowledge management along asset lifecycle – from concept to decommissioning. High fidelity, physics based global simulations, such as RB-FEA digital twins, are a key enabler for effective digital transformation of industrial operations. They enable real-time insights into asset performance that are otherwise unattainable, providing tools for Asset Integrity Management, Risk Based Inspection, Automation and Optimization of Operations as well as enabling full digital thread enhancing Safety and CAPEX and OPEX.