1. Assessing the safe transportation of multiphase pipeline by integrating mechanism and Machine learning models.
- Author
-
Li, Zhuochao, Wang, Bohong, Yan, Fengyuan, Xu, Ning, Yan, Dongyin, Qiu, Rui, and Liang, Yongtu
- Subjects
- *
MACHINE learning , *PIPELINE transportation , *OIL fields , *SYSTEM safety , *HEAT transfer - Abstract
• The hybrid model has a multiphase flow mechanism and a machine learning model. • The model calculates multiphase transport boundaries to ensure pipeline safety. • The Maximum transportable Gas-Oil Ratio is a boundary under different conditions. • Multiphase transport boundaries cover the entire lifecycle of oilfields. • The average relative error of the hybrid model is less than 4.7%. Developing a safe Multiphase Transport Boundary (MTB) is crucial due to the complex flow and heat transfer laws of multiphase transportation of fluids in pipelines. However, there is no efficient, accurate, and widely adaptable method for MTB formulation. This paper proposes a hybrid method based on traditional multiphase flow mechanisms and machine learning models to determine the safe operational MTB for multiphase pipeline transportation processes. The interactive logic enables the efficient and accurate calculation of MTB under real-time measurement data-driven for the proposed new hybrid model. Using a northwest oil field in China as an example, the results indicate that the hybrid model has an average relative error of less than 4.7% compared with simulated data, showing high accuracy. The calculation of a million operating conditions only takes 232 s. The research results can guide promoting the design, operation, and modification of pipeline transportation systems to ensure safety. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF