Machine learning-based surrogate correlation for prediction of wax deposition rate in oil and gas pipelines

Abstract

This study developed a data-driven surrogate-box correlation to predict wax deposition rates in crude-oil pipelines and help address the limited generalizability of empirical and mechanistic models under variable flow and thermal conditions. Literature-derived dataset of 215 experimental and field observations was compiled with inputs as oil temperature T_o, wall temperature T_w, dynamic viscosityμ, wall shear stressσ, flow velocity v_f, wall temperature gradient ΔT, and wall concentration gradient ΔC, and target as wax deposition rateδ _W (g·m^-2.h^-1). After outlier control, standardization, and ANOVA-based feature engineering, Five supervised black-box models; SVR, Random Forest, MLP, Gradient Boosting, and KNN, models were trained with an 80/20 train–test split, 5-fold cross-validated, hyperparameter-tuned and evaluated using MSE, RMSE, MAE,R², and AAPRE. The top performing model, KNN, was modeled into an interpretable surrogate via elastic net (benchmarked against polynomial Ridge degree-2/3) to yield a closed-form correlation. KNN achieved excellent test performance (R² = 0.985, RMSE ~ 0.101, AAPRE ~ 5.09%), outperforming alternative models. The Elastic Net surrogate preserved predictive ability with balanced generalization (R² ≈ 0.61–0.65 for train/validation/test) while exposing parameter level effects and nonlinear interactions among temperature, viscosity, shear stress, velocity, and wall-scale gradients. When compared to a physics-based correlation, the surrogate exhibited tighter clustering to measurements, which indicates improved field relevance. The principal contribution of this study is a hybrid workflow that couples high-accuracy black-box learning with transparent surrogate modeling to help enable real-time monitoring and control, proactive pigging scheduling, and optimized chemical/thermal treatments of oil flow in pipelines. The resulting correlation offers a deployable, interpretable alternative to nontransparent machine learning models or assumption-heavy empirical relations, with clear value for flow-assurance planning and operational reliability.

Keywords

Wax deposition rate / Flow assurance / Surrogate box modeling / Black box models

CRediT authorship contribution statement

Kwame Sarkodie: Conceptualization. Joseph Agyepong: Methodology. Emmanuel Agyei: Software, Investigation. Godfred Arkoh: Visualization, Data curation. Rhoda Arhin: Writing – original draft, Validation. Franklin Ankomah: Writing – original draft. Caspar Daniel Adenutsi: Writing – review & editing, Investigation. Samuel Erzuah: Methodology, Formal analysis.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Funding

The authors declare that no financial support, grants, or funding was received for the research, authorship, and/or publication of this article.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

The authors acknowledge the support of the Reservoir simulation laboratory at the Department of Petroleum Engineering for provided facilities and software for coding.

References

[1] B.M. Lonje, G. Liu, Review of wax sedimentations prediction models for crude-oil transportation pipelines, Petrol. Res. 7 (2) (2022) 220-235, https://doi.org/10.1016/j.ptlrs.2021.09.005.

[2] S. Banerjee, S. Chakrabortty, J. Nayak, S.K. Tripahy, M. Shah, Environmental Friendly Green Technologies for Improvement of Heavy Crude Oil Flow Assurance, 2025. https://books.google.com.eg/books?id=9NheEQAAQBAJ.

[3] S. Misra, S. Baruah, K. Singh, Paraffin problems in crude oil production and transportation: a review, SPE Prod. Facil. 10 (1) (1995) 50-54, https://doi.org/10.2118/28181-pa.

[4] A. Gholami, H.R. Ansari, S. Ahmadi, Combining of intelligent models through committee machine for estimation of wax deposition, J. Chin. Chem. Soc. 65 (8) (2018) 925-931, https://doi.org/10.1002/jccs.201700329.

[5] A. Japper-Jaafar, P.T. Bhaskoro, Z.S. Mior, A new perspective on the measurements of wax appearance temperature: Comparison between DSC, thermomicroscopy and rheometry and the cooling rate effects, J. Petrol. Sci. Eng. 147 (September) (2016) 672-681, https://doi.org/10.1016/j.petrol.2016.09.041.

[6] M.O. Akinyede , Development of a Thermodynamic Model for Wax Precipitation in Produced Crude oil-case Study of Hydrocarbon Fluid from Niger-Delta, 2019. Nigeria.

[7] K. Fan, Q. Huang, S. Li, W. Yu, The wax deposition rate of water-in-crude oil emulsions based on the laboratory flow loop experiment, J. Dispersion Sci. Technol. 38 (1) (2017) 8-18, https://doi.org/10.1080/01932691.2015.1050729.

[8] A. Harun, N.K.I.N. Ab Lah , H. Husin, Z. Hassan, An overview of wax crystallization, deposition mechanism and effect of temperature & shear, in: ICIMSA 2016 - 2016 3rd International Conference on Industrial Engineering, Management Science and Applications, 2016, pp. 1-5, https://doi.org/10.1109/ICIMSA.2016.7503992.

[9] F.O. Ochieng , M.N. Kinyanjui , J.O. Abonyo , P.R. Kiogora , Mathematical modeling of wax deposition in field-scale crude oil pipeline systems, J. Appl. Math. 2022 (8) (2022), https://doi.org/10.1155/2022/2845221.

[10] H. Zhu, Y. Lei, P. Yu, C. Li, F. Yang, B. Yao, S. Yang, H. Peng, Wax deposition during the transportation of waxy crude oil: mechanisms, influencing factors, modeling, and outlook, Energy Fuels 38 (11) (2024) 9131-9152, https://doi.org/10.1021/acs.energyfuels.3c04687.

[11] B.M. Lonje , G. Liu , Review of wax sedimentations prediction models for crude-oil transportation pipelines, Petrol. Res. 7 (2) (2022) 220-235, https://doi.org/10.1016/j.ptlrs.2021.09.005.

[12] L.F.A. Azevedo , A.M. Teixeira , A critical review of the modeling of wax deposition mechanisms, Petrol. Sci. Technol. 21 (3-4) (2003) 393-408, https://doi.org/10.1081/LFT-120018528.

[13] A. Matzain, Multiphase flow paraffin deposition modeling, in: Dissertation, 1999.

[14] A.A. Soedarmo , N. Daraboina , C. Sarica , Validation of wax deposition models with recent laboratory scale flow loop experimental data, J. Petrol. Sci. Eng. 149 (October 2016) (2017) 351-366, https://doi.org/10.1016/j.petrol.2016.10.017.

[15] E.D. Burger , T.K. Perkins , J.H. Striegler , Studies of wax deposition in the Trans Alaska pipeline, JPT, J. Petrol. Technol. 33 (6) (1981) 1075-1086, https://doi.org/10.2118/8788-PA.

[16] B. Yao, D. Zhao, Z. Zhang, C. Huang, Safety study on wax deposition in crude oil pipeline, Processes 9 (9) (2021), https://doi.org/10.3390/pr9091572.

[17] W. Wang, Q. Huang, Prediction for wax deposition in oil pipelines validated by field pigging, J. Energy Inst. 87 (3) (2014) 196-207, https://doi.org/10.1016/j.joei.2014.03.013.

[18] Z. Huang, H.S. Lee , M. Senra, H.S. Fogler , A fundamental model of wax deposition in subsea oil pipelines, AIChE J. 59 (4) (2012) 215-228, https://doi.org/10.1002/aic.

[19] F. Alnaimat, M. Ziauddin, Wax deposition and prediction in petroleum pipelines, J. Petrol. Sci. Eng. 184 (August 2019) (2020) 106385, https://doi.org/10.1016/j.petrol.2019.106385.

[20] M.K. Siljuberg , Modelling of paraffin wax in oil pipelines, Petrol. Eng. Appl. Geophys. 80 (2012).

[21] M. Stubsjøen, Analytical and Numerical Modeling of Paraffin Wax in Pipelines, 2013. June.

[22] J.R. Agarwal , C.F. Torres , S. Shah , Development of dimensionless parameters and groups of heat and mass transfer to predict wax deposition in crude oil pipelines, ACS Omega 6 (16) (2021) 10578-10591, https://doi.org/10.1021/acsomega.0c05966.

[23] M. Obaseki, P.T. Elijah, Dynamic modeling and prediction of wax deposition thickness in crude oil pipelines, J. King Saud Univ. Eng. Sci. 33 (6) (2021) 437-445, https://doi.org/10.1016/j.jksues.2020.05.003.

[24] M.M. El-Dalatony , B.H. Jeon , E.S. Salama , M. Eraky , W.B. Kim , J. Wang , T. Ahn , Occurrence and characterization of paraffin wax formed in developing wells and pipelines, Energies 12 (6) (2019) 1-23, https://doi.org/10.3390/en12060967.

[25] M. Halstensen, B.K. Arvoh, L. Amundsen, R. Hoffmann, Online estimation of wax deposition thickness in single-phase sub-sea pipelines based on acoustic chemometrics: a feasibility study, Fuel 105 (2013) 718-727, https://doi.org/10.1016/j.fuel.2012.10.004.

[26] S. Ito, Y. Tanaka, T. Hazuku, T. Ihara, M. Morita, I. Forsdyke, Wax thickness and distribution monitoring inside petroleum pipes based on external temperature measurements, ACS Omega 6 (8) (2021) 5310-5317, https://doi.org/10.1021/acsomega.0c05415.

[27] Y. Xie, Y. Xing, A prediction method for the wax deposition rate based on a radial basis function neural network, Petroleum 3 (2) (2017) 237-241, https://doi.org/10.1016/j.petlm.2016.08.003.

[28] B. Yao, D. Zhao, Z. Zhang, C. Huang, Safety study on wax deposition in crude oil pipeline, Processes 9 (9) (2021), https://doi.org/10.3390/pr9091572.

[29] V. Hassija, V. Chamola, A. Mahapatra, A. Singal, D. Goel, K. Huang, S. Scardapane, I. Spinelli, M. Mahmud, A. Hussain, Interpreting black-box models: a review on explainable artificial intelligence, Cogn. Comput. 16 (1) (2024) 45-74, https://doi.org/10.1007/s12559-023-10179-8.

[30] Y. Forghani, R.S. Tabrizi, H.S. Yazdi, M.R. Akbarzadeh-T, Fuzzy support vector regression, in: 2011 1st International Econference on Computer and Knowledge Engineering, ICCKE 2011, 2011, pp. 28-33, https://doi.org/10.1109/ICCKE.2011.6413319.

[31] M. Awad, R. Khanna, Efficient learning machines: theories, concepts and applications for engineers and system designers, in: Apress Open, 2015.

[32] M. Khalid, I. Ashraf, A. Mehmood, S. Ullah, M. Ahmad, G.S. Choi , GBSVM: sentiment classification from unstructured reviews using ensemble classifier, Appl. Sci. (Switzerland) 10 (8) (2020), https://doi.org/10.3390/APP10082788.

[33] S. Walczak, N. Cerpa, Artificial Neural Networks in Advanced Methodologies and Technologies in Artificial Intelligence, Computer Simulation, and human-computer Interaction, 2019, pp. 40-53. http://www.emsl.pnl.gov:2080/proj/neuron/neural/sys.

[34] Y. Song, J. Liang, J. Lu, X. Zhao, An efficient instance selection algorithm for k nearest neighbor regression, Neurocomputing 251 (2017) 26-34, https://doi.org/10.1016/j.neucom.2017.04.018.

[35] B. Williams, S. Cremaschi, Novel tool for selecting surrogate modeling techniques for surface approximation, in: M. Türkay, R.B.T.-C.A.C.E. Gani (Eds.), 31 European Symposium on Computer Aided Process Engineering, 50, Elsevier, 2021, pp. 451-456, https://doi.org/10.1016/B978-0-323-88506-5.50071-1.

[36] H. Saleh, J.A. Layous , Machine Learning - Regression by : January, 2022, https://doi.org/10.13140/RG.2.2.35768.67842.

[37] L. Boukarma, A. Aboussabek, F. El Aroussi, M. Zerbet, F. Sinan, M. Chiban, Insight into mechanism, box-Behnken design, and artificial neural network of cationic dye biosorption by marine macroalgae Fucus spiralis, Algal Res. 76 (July) (2023) 103324, https://doi.org/10.1016/j.algal.2023.103324.

[38] Datasans, Machine Learning Cheatsheet : Model Regresi, 2024, pp. 1-120.

[39] A.T. Leiroz , L.F.A. Azevedo , Paraffin deposition in a stagnant fluid layer inside a cavity subjected to a temperature gradient, Heat Transf. Eng. 28 (6) (2007) 567-575, https://doi.org/10.1080/01457630701193997.

[40] A.K. Mehrotra , S. Ehsani , S. Haj-Shafiei , A.S. Kasumu , A review of heat-transfer mechanism for solid deposition from “waxy” or paraffinic mixtures, Can. J. Chem. Eng. 98 (12) (2020) 2463-2488, https://doi.org/10.1002/cjce.23829.

[41] S. Zheng, M. Saidoun, K. Mateen, T. Palermo, Y. Ren, H.S. Fogler, Wax deposition modeling with considerations of Non-Newtonian fluid characteristics, Proc. Ann.Offshore Technol. Conf. 1 (2016) 548-565, https://doi.org/10.4043/26914-ms.