(675f) Enhanced Modeling, Health Monitoring and Leak Diagnosis of Hydrogen Energy Transportation Systems

Hydrogen energy as a cleaner form of energy has gained increasing attention in academia and industry over recent years [1-2]. Essentially, hydrogen energy shows great potential in reducing greenhouse gas emissions from existing carbon-based energy sectors, improving the outlooks in addressing the climate crisis, and pursuing sustainable process/energy systems engineering. As gas or liquid form, the majority of hydrogen energy transportation and distribution relies on pipelines, trucks, railways, and shipping vessels [1], where blending hydrogen into the existing natural gas pipeline networks is reported to be an effective strategy as it can be used for both local distribution and long-distance transportation. The safety, efficiency, and profitability of hydrogen pipeline transportation depend on an accurate understanding of transport dynamic modeling. To this end, this presentation will show an enhanced modeling method and novel techniques for healthy condition monitoring and leak diagnosis of hydrogen pipeline transportation systems.

In this work, we propose a discrete-time infinite-dimensional modeling method for accurately transforming the continuous-time first-principle hydrogen transport model described by hyperbolic partial differential equations (PDEs) into easily realizable computing setting while preserving essential system properties (Hamiltonian energy, input-output mapping, approximate observability, etc.), by using a bilinear transformation. Such a discrete-time infinite-dimensional model is capable of comprehensive modelling of the spatial-temporal natural gas and hydrogen dynamics within pipelines. To account for the possible model-plant mismatch in actual applications, we apply an optimization-based system identification technique to learn crucial parameters inducing the model discrepancy, remove the model-plant mismatch, and improve the modeling performance. Based on the enhanced discrete-time infinite-dimensional modeling, we propose a novel switching model for describing both leakage and normal flow dynamics in the presence of norm-bounded plant and measurement disturbances. Considering the coupling leak size and distribution terms, we apply a PDE back-stepping transformation to fully decouple them in a target system. To fully account for the constraints on state, leak, and disturbances, we propose a novel discrete-time infinite-dimensional moving horizon estimation design for simultaneously healthy condition monitoring, leak detection, localization, and size estimation, by extending the existing results [3-6]. Numerical studies will be shown to demonstrate the proposed methods. The proposed designs have the potential to be applied to real-world applications.

References

[1] Riera, Jefferson A., Ricardo M. Lima, and Omar M. Knio. "A review of hydrogen production and supply chain modeling and optimization." International Journal of Hydrogen Energy (2023).

[2] Lutostansky, Elizabeth, Leonard Creitz, Seungho Jung, Joan Schork, David Worthington, and Yongfu Xu. "Modeling of underground hydrogen pipelines." Process Safety Progress 32, no. 2 (2013): 212-216.

[3] Huang, Rui, Lorenz T. Biegler, and Sachin C. Patwardhan. "Fast offset-free nonlinear model predictive control based on moving horizon estimation." Industrial & Engineering Chemistry Research 49, no. 17 (2010): 7882-7890.

[4] Xie, Junyao, Biao Huang, and Stevan Dubljevic. "Moving Horizon Estimation for Pipeline Leak Detection, Localization, and Constrained Size Estimation." Submitted, 2024.

[5] Rao, Christopher V., James B. Rawlings, and Jay H. Lee. "Constrained linear state estimation—a moving horizon approach." Automatica 37, no. 10 (2001): 1619-1628.

[6] Xie, Junyao, Jukka-Pekka Humaloja, Charles Robert Koch, and Stevan Dubljevic. "Approximate moving horizon estimation for switching conservative linear infinite-dimensional systems." Automatica 158 (2023): 111306.