Research on the application of artificial intelligence tools to predict the operating temperature and pressure of the pipeline transportation system at the Hai Thach - Moc Tinh field
- Tác giả: Tuan Thanh Nguyen*, Thinh Van Nguyen, Truong Hung Trieu
Cơ quan:
Hanoi University of Mining and Geology, Hanoi, Vietnam
- *Tác giả liên hệ:This email address is being protected from spambots. You need JavaScript enabled to view it.
- Từ khóa: Artificial intelligence (AI), Hai Thach-Moc Tinh, Machine learning, Pipeline transportation system.
- Nhận bài: 15-07-2023
- Lĩnh vực: Dầu khí và năng lượng
Tóm tắt:
The observation of operating parameters for the single-phase or multi-phase flows of oil and gas pipelines always plays an important role and prerequisite in daily operation. The inlet and outlet parameters of the pipeline such as pressure and temperature will greatly affect the efficiency of the production and transportation process of fields. These parameters are very important. It must be required to observe carefully and strictly. In fact, pressure and temperature signal gauges will always be set up at the inlet and outlet of the pipeline for continuously variable transmission of the signal to the central control room, then the operator can observe regularly. Alarms will be triggered when the signal goes out of the setting operation area. However, the operational parameters transmitted from this gauge are only available after the actual fluid flow passes through the pipeline. This will limit when the operator needs to apply calculation methods to pre-predict the flow regime, and operating parameters of the fluid in the pipeline when fluid flows has not passed through the pipeline. From that approach, the paper presents the result of research on the application of machine learning algorithms (ML) to build models to predict operating conditions of three-phase flows in the pipeline at Hai Thach- Moc Tinh field based on input parameters such as the open of wellhead flow control valve of wells located in wellhead platform WHP-MT1, WHP-HT1, and the commercial gas volume. The results of the research show that the ML calculation method gives the result which is only +/-2 barg/20C difference compared to the field data obtained from pressure (HT1-PT-0911) and temperature (HT1-TI-0911) signals set up at the outlet of the transportation pipeline at the Hai Thach-Moc Tinh field.
McNeil, R.,G. and Lillico, D. R. (2005). Gas Gathering System Modelling the Pipeline Pressure Loss Match, Paper presented at the Canadian International Petroleum Conference, Calgary, Alberta. https://doi.org/10.2118/2005-230.
Jane, O., Tobinson., Briggs, A. K. (2020). Prediction of Pressure Drop in Subsea Pipeline Using Pipesim Software. Industrial Engineering Letters, vol 10, pp 62-81. doi: 10.7176/iel/10-2-06.
John, M., Campbell,. (1992). Gas conditioning and processing. Vol. 2 , Campbell Petroleum Series.
Montgomery, D. C., A. P. lizabeth, G. G. Vining., (2021). Introduction to Linear Regression Analysis, 6th Edition.
John Wiley and Sons.McDonald, G. C. (2009). Ridge regression. WIREs Computational Statistics, vol. 1, no. 1, pp. 93-100. doi: https://doi.org/10.1002/wics.14.
Zhao, P and Yu, B. (2006). On model selection consistency of LASSO. Journal of Machine Learning Research, vol. 7, pp. 2541-2563.
Jia, J and Yu, B. (2010). On Model Selection Consistency of the Elastic Net When p>>n. Statistica Sinica 20, pp. 1-19.
Nguyen-Sy, T., To, Q.d., Ngoc-Vu, M., Duong-Nguyen. T., and Trung- Nguyen, T. (2021). Predicting the electrical conductivity of brine-saturated rocks using machine learning methods. Journal of Applied Geophysics, vol. 184, p. 104238. https://doi.org/10.1016/j.jappgeo.2020.104238.T.
Nguyen-Sy, T., Wakim,J., Dong-To,Q., Ngoc- Vu, M., Duong-Nguyen,T., Trung-Nguyen, T. (2020). Predicting the compressive strength of concrete from its compositions and age using the extreme gradient boosting method. Construction and Building Materials, vol. 260, p119757. https://doi.org/10.1016/j.conbuildmat.2020.119757.
Các bài báo khác