Prediction of Wax Deposition Using Machine Learning Algorithms: Multilayer Perceptron and Random Forest

Document Type : Original Article

Authors
Milad Shadbakht 1
Eghbal Sahraei 2
Ahmad Hosseinzadegan 3
1 M.Sc. Student in Petroleum Engineering, Faculty of Petroleum and Gas Engineering, Sahand University of Technology, Tabriz, Iran
2 Associate Professor, Faculty of Petroleum and Gas Engineering, Sahand University of Technology, Tabriz, Iran
3 Ph.D. Petroleum Engineering, Khazar Exploration & Production Company (KEPCO), Tehran, Iran
Abstract
The phenomenon of paraffin deposition is one of the major problems that has been permanently present in the oil and gas industry. One of these deposits is wax deposition, which is problematic in all stages of oil extraction, transportation, and processing, and has caused blockages in wells, and equipment. wells, oil transmission lines and exploitation units, and refining sites. Many of these researchers presented materials about wax apparent temperature (WAT) and thermodynamic investigation of wax deposition, which were widely used in the oil industry to predict initial deposition. In this work, models with multi-layer perceptron algorithm and random forest were created for wax deposition in the pipeline, and these models indicate that increasing the length of the pipeline can cause more wax deposition, as well as models for predicting these deposits relative to the length of the line. The tube was provided.

Keywords

Subjects

  1. Jafari Behbahani T. Fundamentals of Hydrocarbon Deposition in the Oil Industry (Asphaltenes, Waxes). Tehran: Farahmand Publishing; 2015. [In Persian].
  2. G. W. Gilby, Chemicals in the oil industry (1983). 108-113.
  3. M. S. Wertheim, Fluids with Highly Directional Attractive Forces. IV. Equilibrium Polymerization May 31, 1985.
  4. Mozes, G., Freund, M., Csikos, R. Keszthelyi, S. “Chemical crystallographical, and Physical Properties of liquid Paraffins and Paraffins Waxes, in Paraffin Product, Properties, Technologies, Applications”, Mozes, G, ED, Elsevier Scientific, New York, 1982.
  5. Wax Precipitation from North Sea Crude Oils. 1. Crystallization and Dissolution Temperatures, and Newtonian and Non-Newtonian Flow Properties. Hans Petter Ranningsen and Brit Bjamdal, June 11, 1991.
  6. Prediction of Wax Appearance Temperature Using Artificial Intelligent Techniques, Chahrazed Benamara, Kheira Gharbi, 2019.
  7. Wax deposition and prediction in petroleum pipelines, Alnaimat F, Ziauddin M, 2019.
  8. A review of wax formation/mitigation technologies in the petroleum industry, Meagan white, Kelly Pierce, 2018.
  9. A review. (PDF) Preventing and removing wax deposition inside vertical wells: a review, A. L. Sousa· H. A. Matos · L.  P. Guerreiro, 2019.
  10. Wax Deposition in Crude Oil Transport Lines and Wax Estimation Methods, Fadi Alnaimat, Mohammed Ziauddin and Bobby Mathew, 2020.
  11. Predicting wax deposition using robust machine learning techniques Menad Nait Amar, Ashkan Jahanbani Ghahfarokhi, 2021.
  12. Prediction of wax disappearance temperature using artificial neural networks, Gholamreza Moradi a, Majid Mohadesi, Mohammad Reza Moradi., 2013.
  13. Artificial Intelligence: A Modern Approach, Russell و Peter Norvig.
  14. Thermodynamics of Wax Precipitation in Petroleum Mixtures, C. Lira-Galeana and A. Firoozabadi, john M. Prausnitz,1996.

  • Receive Date 02 October 2024
  • Revise Date 31 October 2025
  • Accept Date 10 May 2025