«Automation and Informatization of the fuel and energy complex»

Development and testing of methodological approaches and digital technologies of neural network proxy modeling of steady-state two-phase flow of a multicomponent mixture in gas collection and field treatment systems (using the example of the Chayandinskoye oil and gas condensate field)

UDC: 519.673
DOI: -

Authors:

¹ NIIgazekonomika, Moscow, Russia
² Gazprom dobycha Noyabrsk, Noyabrsk, Russia
³ National University of Oil and Gas "Gubkin University", Moscow, Russia

Keywords: gas collection and treatment, proxy modeling, surrogate models, machine learning, neural networks

Annotation:

Mathematical modeling is one of the main objective instruments for preparing justified decisions when operating and developing hydrocarbon deposits. In this field of activity, lately, there observed a qualitative transition from using local models of individual technological objects to creating integrated digital models of "reservoir–well–gas collection network–integrated gas treatment unit–booster compressor station". They are based on the models that describe the physics of filtration processes and gas flow. At the same time, the analysis of modern modeling packages shows that such physical models are built in different software systems and their use requires large computational and time costs to carry out calculations. In this regard, in practice, attempts are increasingly being made to develop and implement their analogues – based on the data (for example, the results of full-scale and/or computational experiments) of proxy models (also called surrogate), which simulate technological processes with acceptable accuracy. At the same time, the speed of calculations using proxy models is significantly higher than that of traditional calculations using physical models. The authors of the article propose methodological approaches to building proxy models of gas collection and field treatment systems. The architecture of the computer environment for training and deploying proxy models is considered. The results of testing the proposed approaches are presented using the example of building a neural network proxy model of a steady two-phase flow of a multicomponent mixture in one of the gas collection reservoirs of the Chayandinskoye oil and gas condensate field. The resulting model demonstrated rather high accuracy of the "physical" model approximation, while the speed of calculations for the selected assumptions significantly exceeds classical models. It allows using such models when solving actual practical problems, primarily system optimization problems. It is noted that the proposed approaches can be used not only to solve operational problems of analysis, control and regulation of a field development, but also to design development and assess geological and technical activities. The features of the proposed approaches are discussed and ways of their further development are outlined.

Bibliography:

1. Model Order Reduction and Applications / M. Hinze, J.N. Kutz, O. Mula, K. Urban. – Cham: Springer, 2021. – XIV, 230 p. – (Lecture Notes in Mathematics (LNM, volume 2328)). – DOI: 10.1007/978-3-031-29563-8
2. Sardanashvili S.A. Raschetnye metody i algoritmy (truboprovodnyy transport gaza). – M.: RGU nefti i gaza im. I.M. Gubkina, 2005. – 577 s.
3. Ping Jiang, Qi Zhou, Xinyu Shao. Surrogate Model-Based Engineering Design and Optimization. – Singapore: Springer, 2020. – IX, 240 p. – (Springer Tracts in Mechanical Engineering). – DOI: 10.1007/978-981-15-0731-1
4. Kuleshov A.P. Kognitivnye tekhnologii v adaptivnykh modelyakh slozhnykh ob"ektov // Inform. tekhnologii i vychisl. sistemy. – 2008. – № 1. – S. 18–29.
5. Ivlev D.A. Rekonstruktsiya kart glubin granits stratigraficheskikh gorizontov s pomoshch'yu generativno-sostyazatel'nykh- setey // PROneft'. Professional'no o nefti. – 2023. – T. 8, № 1(27). – S. 188–197. – DOI: 10.51890/2587-7399-2023-8-1-188-197
6. Metodologiya postroeniya metamodeley i perspektivy ikh primeneniya dlya resheniya aktual'nykh zadach neftyanogo inzhiniringa / M.V. Simonov, A.V. Penigin, A.S. Margarit [i dr.] // ProNeft'. Professional'no o nefti. – 2019. – № 2(12). – S. 48–53. – DOI: 10.24887/2587-7399-2019-2-48-53
7. Integratsiya metodov mashinnogo obucheniya i geologo-gidrodinamicheskogo modelirovaniya pri proektirovanii razrabotki mestorozhdeniy / L.S. Brilliant, A.S. Zav'yalov, M.Yu. Dan'ko [i dr.] // Neft. khoz-vo. – 2022. – № 3. – S. 48–53. – DOI: 10.24887/0028-2448-2022-3-48-53
8. Mikhaylovskiy A.A. Primenenie uproshchennykh gazogidrodinamicheskikh proksi-modeley dlya operativnykh tekhnologicheskikh raschetov gazovykh promyslov i podzemnykh khranilishch // Nauch.-tekhn. sb. Vesti gazovoy nauki. – 2018. – № 1(33). – S. 193–202.
9. Ob obosnovannosti primeneniya, sovremennom sostoyanii i nekotorykh perspektivakh razvitiya neyrosetevykh modeley Edinoy sistemy gazosnabzheniya Rossii / N.A. Kislenko, A.V. Belinskiy, A.S. Kazak, O.I. Belinskaya // Avtomatizatsiya i informatizatsiya TEK. – 2022. – № 5(586). – S. 6–17. – DOI: 10.33285/2782-604X-2022-5(586)-6-17
10. Nikulin A.S., Stepnov D.A., Kurkin D.S. Razrabotka surrogatnoy modeli kolonny stabilizatsii gazovogo kondensata // Gazovaya prom-st'. – 2022. – № 11(840). – S. 42–45.
11. Motaei E., Ganat T. Smart proxy models art and future directions in the oil and gas industry: A review // Geoenergy Science and Engineering. – 2023. – Vol. 227. – P. 211918. – DOI: 10.1016/j.geoen.2023.211918
12. Opyt razrabotki programmno-vychislitel'nogo komp'yuternogo kompleksa modelirovaniya dvukhfaznogo techeniya mnogokomponentnoy smesi dlya usloviy Chayandinskogo NGKM / V.V. Samsonova, S.K. Mitichkin, V.A. Marishkin, V.I. Churin // Komp'yuternye tekhnologii podderzhki prinyatiya resheniy v dispetcherskom upravlenii gazotransportnymi i gazodobyvayushchimi sistemami: tez. dokl. VII Mezhdunar. nauch.-tekhn. konf., M., 14–15 okt. 2021 g. – M.: Gazprom VNIIGAZ, 2021. – S. 43.
13. Kamath C. Intelligent sampling for surrogate modeling, hyperparameter optimization, and data analysis. – 2023. – DOI: 10.48550/arXiv.2306.04066
14. Alizadeh R., Allen J.K., Mistree F. Managing computational complexity using surrogate models: a critical review // Research in Engineering Design. – 2020. – Vol. 31, Issue 3. – P. 275–298. – DOI: 10.1007/s00163-020-00336-7
15. Kumar P., Gupta A. Active Learning Query Strategies for Classification, Regression, and Clustering: A Survey // J. of Computer Science and Technology. – 2020. – Vol. 35, No. 4. – P. 913–945. – DOI: 10.1007/s11390-020-9487-4
16. Kislenko N.A., Belinskiy A.V., Kazak A.S. Metody, algoritmy i instrumenty modelirovaniya i optimizatsii rezhimov raboty Edinoy sistemy gazosnabzheniya Rossii na osnove tekhnologiy iskusstvennogo intellekta. Chast' 1 // Gazovaya prom-st'. – 2021. – № 9(821). – S. 88–96.
17. Svidetel'stvo o gos. registratsii programmy dlya EVM 2023664462 Ros. Federatsiya. Bazovye tsifrovye platformennye resheniya gazovoy promyshlennosti (Arifma-gaz) / A.V. Belinskiy, V.S. Makhina, A.A. Plachinda [i dr.]; pravoobladatel' OOO "NIIgazekonomika". – № 2023663726; zayavl. 30.06.2023; opubl. 05.07.2023.
18. Svidetel'stvo o gos. registratsii programmy dlya EVM 2015616283 Ros. Federatsiya. Komp'yuternyy dispetcherskiy trenazher gazovogo promysla (skvazhina-GSS-DKS-UKPG-MG) OOO "Gazprom dobycha Noyabr'sk" / V.A. Marishkin, A.V. Kononov, V.P. Smirnov [i dr.]; pravoobladatel' OOO "Gazprom dobycha Noyabr'sk". – № 2015612717; zayavl. 07.04.2015; opubl. 20.07.2015.
19. R Gazprom 2-3.5-1037-2016. Modelirovanie tekhnologicheskikh rezhimov ekspluatatsii sistem sbora i vnutripromyslovogo transporta gaza senomanskikh zalezhey / Gazprom VNIIGAZ. – Vved. 2016–01–15. – SPb.: Gazprom ekspo, 2019. – IV, 20 s.
20. Vyakhirev R.I., Korotaev Yu.P., Kabanov N.I. Teoriya i opyt dobychi gaza. – M.: Nedra, 1998. – 479 s.
21. Ueyles S.M. Fazovye ravnovesiya v khimicheskoy tekhnologii: v 2 ch. Ch. 1 / per. s angl. A.V. Bespalova, A.P. Zhukova, V.V. Paukova pod red. V.S. Beskova. – M.: Mir, 1989. – 304 s.
22. Kingma D.P., Ba J.L. Adam: A Method for Stochastic Optimization // ICLR. – 2015. – DOI: 10.48550/arXiv.1412.6980