Research Progress of Edge Intelligent Class Imbalanced Learning for Gas Well Fluid Accumulation Prediction

DOI: https://doi.org/10.62517/jes.202602135

Author(s)

Jingguo Du1, Yan He2, ManMan Feng3, Wei Qin4, Li Wang5, Qi Dai6,*

Affiliation(s)

1College of Mining Engineering, North China University of Science and Technology, Tangshan, China 2Oil and Gas Field Surface Engineering, Sichuan Changning Natural Gas Development Co., Ltd., Yibin, China 3Sichuan Jiacheng petroleum and natural gas pipeline quality inspection testing Co., LTD, Chengdu, China 4Chongqing Gas Field, PetroChina Southwest Oil & Gas Field Company, Chongqing, China 5Institute of Exploration and Development, Southwest Oil and Gas Field Company, PetroChina, Chengdu, China 6College of Sciences, North China University of Science and Technology, Tangshan, China *Corresponding Author

Abstract

Liquid accumulation in gas wells is one of the most common and destructive wellbore flow problems in natural gas development. Since gas cannot be effectively discharged, liquid accumulates in the well, leading to increased bottomhole back pressure, decreased gas production, and increased drainage costs. With the widespread adoption of wellhead sensors, SCADA systems, and digital gas field platforms, edge intelligence is becoming a crucial support for real-time gas well monitoring. However, gas-well liquid accumulation prediction data suffers from problems such as scarce liquid accumulation events, significant operational condition shifts, and obvious label lag, making imbalanced learning a critical step for the successful implementation of liquid accumulation prediction. This paper systematically reviews the gas well liquid accumulation mechanism and prediction task modeling, data-driven modeling, imbalanced learning methods, and edge intelligence system architecture. Analysis results show that considering the imbalanced data distribution problem can improve the accuracy of gas well liquid accumulation analysis and early warning.

Keywords

Gas Well Fluid Accumulation; Anomaly Detection; Edge Intelligence; Class Imbalance Learning; Cloud Collaboration

References

[1] Pagan E V, Waltrich P J. A simplified model to predict transient liquid loading in gas wells. Journal of Natural Gas Science and Engineering, 2016, 35: 372-381. [2] Turner R G, Hubbard M G, Dukler A E. Analysis and prediction of minimum flow rate for the continuous removal of liquids from gas wells. Journal of Petroleum technology, 1969, 21(11): 1475-1482. [3] Coleman S B, Clay H B, McCurdy D G, et al. A new look at predicting gas-well load-up. Journal of petroleum technology, 1991, 43(03): 329-333. [4] Li M, Li S L, Sun L T. New view on continuous-removal liquids from gas wells. SPE Production & Facilities, 2002, 17(01): 42-46. [5] Liu H, Lou W, Li H, et al. A modified comprehensive prediction model for wellbore temperature-pressure field and liquid loading of gas wells. Geoenergy Science and Engineering, 2024, 232: 212452. [6] Jiang J, Li K, Du J, et al. Prediction system for water-producing gas wells using edge intelligence. Expert Systems with Applications, 2024, 247: 123303. [7] Savaglio C, Mazzei P, Fortino G. Edge intelligence for industrial iot: Opportunities and limitations. Procedia Computer Science, 2024, 232: 397-405. [8] Liu X, Dong X, Jia N, et al. Federated learning-oriented edge computing framework for the IIoT. Sensors, 2024, 24(13): 4182. [9] Zhu G, Liu X, Niu J, et al. Learning by imitating the classics: Mitigating class imbalance in federated learning via simulated centralized learning. Expert Systems with Applications, 2024, 255: 124755. [10] Lin J, Xu S, Liu Y. Experimental and modeling studies on continuous liquid removal in horizontal gas wells. Frontiers in Earth Science, 2023, 11: 1288208 [11] Veeken K, Hu B, Schiferli W. Gas-well liquid-loading-field-data analysis and multiphase-flow modeling. SPE Production & Operations, 2010, 25(03): 275-284. [12] Wang W, Zhang X, Wei Y, et al. A new mechanistic model to predict liquid loading in inclined natural gas wells. Geoenergy Science and Engineering, 2024, 241: 213163. [13] Zhu Z, Han G, Liang X, et al. Rapid classification and diagnosis of gas wells driven by production data. Processes, 2024, 12(6): 1254. [14] Rezvani S, Wang X. A broad review on class imbalance learning techniques. Applied Soft Computing, 2023, 143: 110415. [15] Xia W, Liu B, Xiang H. Prediction of liquid accumulation height in gas well tubing using integration of crayfish optimization algorithm and xgboost. Processes, 2024, 12(9): 1788. [16] Zhao C, Jia Y, Qu Y, et al. Forecasting gas well classification based on a two-dimensional convolutional neural network deep learning model. Processes, 2024, 12(5): 878. [17] Zhou W, Sun W, Ding W, et al. A Deep Learning-Based Model for Real-Time Production Forecasting of Gas Wells with Intermittent Production. SPE Journal, 2025, 30(03): 1377-1395. [18]Chen Y, Huang Y, Miao B, et al. Adaptive anomaly detection-based liquid loading prediction in shale gas wells. Journal of Petroleum Science and Engineering, 2022, 214: 110522. [19]Liu Q X, Zhu J J, Wang H B, et al. Deep feature learning for anomaly detection in gas well deliquification using plunger lift: A novel CNN-based approach. Petroleum Science, 2025, 22(9): 3803-3816. [20]Chen P, Chen Y, Yang C, et al. Gas well production optimization: Classifying liquid loading severity in shale gas wells using semi-supervised learning. Gas Science and Engineering, 2024, 128: 205394. [21]Chen Y, Miao B, Wang Y, et al. A deep regression method for gas well liquid loading prediction. SPE Journal, 2024, 29(04): 1847-1861. [22]Kheirollahi H, Zayedi M, Simjoo M, et al. Machine Learning-Based Prediction of Liquid Loading in Gas Wells for Flow Assurance Management and Production Optimization[J]. Results in Engineering, 2025: 29, 108605. [23] Gao X, Xie D, Zhang Y, et al. A comprehensive survey on imbalanced data learning. Frontiers of Computer Science, 2026, 20(11): 2011622. [24] Ni Y, Zhang H, Han Y, et al. Study on gas-oil ratio prediction considering the influence of imbalance data. Petroleum Science and Technology, 2026, 44(7): 1142-1161. [25]Jamshidi Gohari M S, Niri M E, Sadeghnejad S, et al. Synthetic graphic well log generation using an enhanced deep learning workflow: imbalanced multiclass data, sample size, and scalability challenges. SPE Journal, 2024, 29(01): 1-20. [26] Yi J, Qi Z L, Li X C Z, et al. A novel class-imbalance learning framework for fluid recognition: Application to Qingshimao-Gaoshawo tight-sand gas reservoirs in the Ordos Basin, China. Computers & Geosciences, 2025: 205, 105993. [27] Wood D A, Mardanirad S, Zakeri H. Effective prediction of lost circulation from multiple drilling variables: a class imbalance problem for machine and deep learning algorithms. Journal of Petroleum Exploration and Production Technology, 2022, 12(1): 83-98. [28] Kim D, Byun J. Selection of augmented data for overcoming the imbalance problem in facies classification. IEEE Geoscience and Remote Sensing Letters, 2021, 19: 1-5. [29] Li H, Ge L, Tian L. Survey: federated learning data security and privacy-preserving in edge-Internet of Things. Artificial Intelligence Review, 2024, 57(5): 130. [30] Dritsas E, Trigka M. Federated learning for IoT: A survey of techniques, challenges, and applications. Journal of Sensor and Actuator Networks, 2025, 14(1): 9. [31] Kulkarni U, Meena S M, Gurlahosur S V, et al. Quantization friendly mobilenet (qf-mobilenet) architecture for vision based applications on embedded platforms. Neural Networks, 2021, 136: 28-39.