Advancing Predictive Analytics for Gas Sweetening Plants Through Machine Learning and Feature Selection

Document Type : Research Paper

Authors

1 Caspian Faculty of Engineering, College of Engineering, University of Tehran, Iran

2 Digital Transformation Center, Research Institute of Petroleum Industry (RIPI), Tehran, Iran

3 School of Chemical Engineering, College of Engineering, University of Tehran, Tehran, Iran

4 Department of Chemical Engineering, Faculty of Engineering, Ferdowsi University of Mashhad, Iran

Abstract

Predictive models employing random forest regression and support vector machines (SVMs) were developed to predict output parameters in an industrial natural gas sweetening plant. Extensive data comprising 550 input/output variables from a gas processing facility in western Iran was leveraged to construct and evaluate the models. The key output forecast was rich amine loading (mole of acid gas per mole of amine). The dataset was partitioned into training (80%), optimization (10%), and testing (10%) subsets after normalization. An R-squared value of 0.97 and a Mean Absolute Error (MAE) of 0.008 were achieved by the random forest regression, outperforming SVM’s R-squared score of 0.91 with an associated MAE of 0.012. Furthermore, the random forest model was optimized using particle swarm optimization (PSO), a metaheuristic technique. The pivotal innovation entails exploiting comprehensive empirical data with hundreds of variables to build data-driven models capable of exceptional predictive fidelity exceeding 0.9 R-squared. This research establishes random forest regression, especially after optimization with PSO, as a highly efficacious and robust methodology for the simulation and optimization of natural gas treating plants

Keywords


  1. Ellaf, A., Ali Ammar Taqvi, S., Zaeem, D., Siddiqui, F. U. H., Kazmi, B., Idris, A., Alshgari, R. A., & Mushab, M. S. S. (2023). Energy, exergy, economic, environment, exergo-environment based assessment of amine-based hybrid solvents for natural gas sweetening, Chemosphere, 313. doi.org/10.1016/j.chemosphere.2022.137426.##
  2. Anagnostis, A., Papageorgiou, E., & Bochtis, D. (2020). Application of artificial neural networks for natural gas consumption forecasting. Sustainability (Switzerland), 12(16), doi.org/10.3390/SU12166409. ##
  3. Rogelj, J., Den Elzen, M., Höhne, N., Fransen, T., Fekete, H., Winkler, H., Schaeffer, R., Sha, F., Riahi, K., & Meinshausen, M. (2016). Paris Agreement climate proposals need a boost to keep warming well below 2 °c, In Nature, 534, 7609, doi.org/10.1038/nature18307. ##
  4. Fortuny, M., Baeza, J. A., Gamisans, X., Casas, C., Lafuente, J., Deshusses, M. A., & Gabriel, D. (2008). Biological sweetening of energy gases mimics in biotrickling filters. Chemosphere, 71(1), doi.org/10.1016/j.chemosphere.2007.10.072. ##
  5. Azizkhani, J. S., Jazayeri-Rad, H., & Nabhani, N. (2014). Design of an ensemble neural network to improve the identification performance of a gas sweetening plant using the negative correlation learning and genetic algorithm. Journal of Natural Gas Science and Engineering, 21, 26–39, doi.org/10.1016/j.jngse.2014.07.012. ##
  6. Wang, S., Zhang, J., Zhang, Y., Wang, L., Sun, Z., & Wang, H. (2023). Review on Source Profiles of Volatile Organic Compounds (VOCs) in Typical Industries in China. In Atmosphere, 14, 5, doi.org/10.3390/atmos14050878. ##
  7. Gonzalez, K., Boyer, L., Almoucachar, D., Poulain, B., Cloarec, E., Magnon, C., & de Meyer, F. (2023). CO2 and H2S absorption in aqueous MDEA with ethylene glycol: Electrolyte NRTL, rate-based process model and pilot plant experimental validation, Chemical Engineering Journal, 451, doi.org/10.1016/j.cej.2022.138948. ##
  8. Torabi Angaji, M., Ghanbarabadi, H., & Karimi Zad Gohari, F. (2013). Optimizations of Sulfolane concentration in propose Sulfinol-M solvent instead of MDEA solvent in the refineries of Sarakhs, Journal of Natural Gas Science and Engineering, 15, doi.org/10.1016/j.jngse.2013.08.003. ##
  9. Ghanbarabadi, H., Khoshandam, B., & Wood, D. A. (2019). Simulation of CO2 removal from ethane with Sulfinol-M+AMP solvent instead of DEA solvent in the South Pars phases 9 and 10 gas processing facility, Petroleum, 5(1), doi.org/10.1016/j.petlm.2018.06.004. ##
  10. Hamedi, A. H., Abolghasemi, H., Shokri, S., Nia, H. J., & Moayedi, F. (2023). Integrating artificial immune genetic algorithm and metaheuristic ant colony optimizer with two-dose vaccination and modeling for residual fluid catalytic cracking process, Arabian Journal for Science and Engineering, 48(12), doi.org/10.1007/s13369-023-07882-x. ##
  11. Dias, T., Oliveira, R., Saraiva, P., & Reis, M. S. (2020). Predictive analytics in the petrochemical industry: Research Octane Number (RON) forecasting and analysis in an industrial catalytic reforming unit, Computers and Chemical Engineering, 139, doi.org/10.1016/j.compchemeng.2020.106912. ##
  12. Cao, J., Zhu, S., Li, C., & Han, B. (2020). Integrating support vector regression with genetic algorithm for hydrate formation condition prediction, Processes, 8(5), doi.org/10.3390/PR8050519. ##
  13. Moayedi, F., Abolghasemi, H., Shokri, S., Ganji, H., & Hamedi, A. H. (2023). Hybrid feature generation and selection with a focus on novel genetic-based generated feature method for modeling products in the sulfur recovery unit, Arabian Journal for Science and Engineering, 48(7). doi.org/10.1007/s13369-023-07609-y. ##
  14. Jiang, C., Zhong, W., Li, Z., Peng, X., & Yang, M. (2019). Real-time semisupervised predictive modeling strategy for industrial continuous catalytic reforming process with incomplete data using slow feature analysis, Industrial and Engineering Chemistry Research, 58(37), doi.org/10.1021/acs.iecr.9b03119. ##
  15. Oloso, M. A., Khoukhi, A., Abdulraheem, A., & Elshafei, M. (2009, October). Prediction of crude oil viscosity and gas/oil ratio curves using recent advances to neural networks, In SPE/EAGE Reservoir Characterization & Simulation Conference, cp-170, European Association of Geoscientists & Engineers, doi.org/10.3997/2214-4609-pdb.170.spe125360. ##
  16. Zarezadeh, F., Vatani, A., Palizdar, A., & Nargessi, Z. (2022). Simulation and optimization of sweetening and dehydration processes in the pretreatment unit of a mini-scale natural gas liquefaction plant, International Journal of Greenhouse Gas Control, 118, doi.org/10.1016/j.ijggc.2022.103669. ##
  17. Jamekhorshid, A., Karimi Davani, Z., Salehi, A., & Khosravi, A. (2021). Gas sweetening simulation and its optimization by two typical amine solutions: An industrial case study in Persian Gulf region, Natural Gas Industry B, 8(3), doi.org/10.1016/j.ngib.2021.04.006. ##
  18. Pandey, M. (2005). Process optimization in gas sweetening unit - A case study, International Petroleum Technology Conference Proceedings, doi.org/10.2523/iptc-10735-ms. ##
  19. Pellegrini, L. A., Gilardi, M., Giudici, F., & Spatolisano, E. (2021). New solvents for co2 and h2s removal from gaseous streams, In Energies, 14, 20, doi.org/10.3390/en14206687. ##
  20. Segal, M. R. (2004). Machine Learning Benchmarks and Random Forest Regression. Biostatistics. ##
  21. Rodriguez-Galiano, V., Sanchez-Castillo, M., Chica-Olmo, M., & Chica-Rivas, M. (2015). Machine learning predictive models for mineral prospectivity: An evaluation of neural networks, random forest, regression trees and support vector machines, Ore Geology Reviews, 71, doi.org/10.1016/j.oregeorev.2015.01.001. ##
  22. Smith, P. F., Ganesh, S., & Liu, P. (2013). A comparison of random forest regression and multiple linear regression for prediction in neuroscience, Journal of Neuroscience Methods, 220(1), doi.org/10.1016/j.jneumeth.2013.08.024.
  23. Cadei, L., Camarda, G., Montini, M., Rossi, G., Fier, P., Bianco, A., Lancia, L., Loffreno, D., Corneo, A., Milana, D., Carrettoni, M., & Silvestri, G. (2019). Prediction and prescription of operation upset in H2S gas sweetening unit: Implementation of an innovative big data analytics procedure, Offshore Mediterranean Conference and Exhibition, OMC 2019. ##
  24. Abdolkarimi, V., Sari, A., & Shokri, S. (2022). Robust prediction and optimization of gasoline quality using data-driven adaptive modeling for a light naphtha isomerization reactor, Fuel, 328. doi.org/10.1016/j.fuel.2022.125304. ##
  25. Abdolkarimi, V., Sari, A., & Shokri, S. (2023). A hybrid multiscale filter along with an improved adaptive SVR technique for fault diagnosis and machine learning modeling: forecasting the octane number of gasoline in isomerization reactor, Neural Computing and Applications, 35(11). doi.org/10.1007/s00521-022-08128-x. ##