Prediction of final product yield in the process of plant biomass pyrolysis using machine learning methods

About the authors:

Roman M. Bachurin, Postgraduate Student, Department of Software, School of Computer Sciences, University of Tyumen, Tyumen, Russia; stud0000107305@utmn.ru, https://orcid.org/0009-0001-2177-3024

Irina G. Zakharova, Cand. Sci. (Phys.-Math.), Professor, Department of Software, School of Computer Science, University of Tyumen, Tyumen, Russia; i.g.zakharova@utmn.ru, https://orcid.org/0000-0002-4211-7675

Abstract:

The article is devoted to the development of methods for mathematical modeling of final product yields in plant biomass pyrolysis using machine learning models.

The main objective of the study is to develop and analyze predictive models. For training the models the data obtained from full-scale experiments are used. Chemical and physical characteristics of biomass composition, parameters of the pyrolysis process temperature regime act as the main attributes. The predicted final yields of pyrolysis products are liquid pyrolysis products, including tar; solid pyrolysis products and pyrolysis gas in percentage ratio.

Based on these data, various machine learning models are developed and analyzed for the possibility of integration into an intelligent research support system. As a result, a model has been selected that allows predicting the ratios of different fractions of the output product during pyrolysis with sufficiently high accuracy, which opens up new opportunities for optimizing pyrolysis plants and improving their energy efficiency.

References:

Le, Thi, Bondarev, A. V., Bondareva, L. I., Monakova, A. S., & Barshin, A. V. (2022). Application of machine learning algorithms in predicting pyrolytic analysis result. Proceedings of Higher Educational Establishments. Geology and Exploration, 63, 8–19. https://doi.org/10.32454/0016-7762-2020-63-6-8-19 [In Russian]

Balsora, H. K., S, K., Joshi, J.B., Sharma, A., & Chakinala, A. G. (2023). Artificial neural network-based models for the prediction of biomass pyrolysis products from preliminary analysis. Industrial & Engineering Chemistry Research, 62(36), 14311–14319. https://doi.org/10.1021/acs.iecr.3c01662

Cheng, Y., Ekici, E., Yildiz, G., Yang, Y., Coward, B., & Wang, J. (2023). Applied machine learning for prediction of waste plastic pyrolysis towards valuable fuel and chemicals production. Journal of Analytical and Applied Pyrolysis, 169, 105857. https://doi.org/10.1016/j.jaap.2023.105857

Fatehi, H., & Bai, X. S. (2014). A comprehensive mathematical model for biomass combustion. Combustion Science and Technology, 186(4-5), 574–593. http://dx.doi.org/10.1080/00102202.2014.883255

Fernando, N., Amin, M., Narayana, M., Jayawickrama, T., Asadumllah, & Jayasena, S. (2015). A mathematical model for pyrolysis of biomass. In Proceedings of the Moratuwa Engineering Research Conference (MERCon) (pp. 6–11). IEEE. https://doi.org/10.1109/MERCon.2015.7112311

Hasse, Ch., Debiagi, P., Wen, X., Hildebrandt, K., Vascellari, M., & Faravelli, T. (2021). Advanced modeling approaches for CFD simulations of coal combustion and gasification. Progress in Energy and Combustion Science, 86, 100938. https://doi.org/10.1016/j.pecs.2021.100938

İşçen, A., Öznacar, K., Tunç, K. M. M., & Günay, M. E. (2023). Exploring the critical factors of biomass pyrolysis for sustainable fuel production by machine learning. Sustainability, 15(20), 14884. https://doi.org/10.3390/su152014884

Kaczor, Z., Buliński, Z., & Werle, S. (2020). Modelling approaches to waste biomass pyrolysis: A review. Renewable Energy, 159, 427–443. https://doi.org/10.1016/j.renene.2020.05.110

Klason, T., & Bai, X.-S. (2003). FG-DVC modeling of biomass combustion in fixed bed. In Proceedings of the Third Mediterranean Combustion Symposium (pp. 188–194).

Lee, H., Choi, I.-H., & Hwang, K.-R. (2024). Machine learning prediction of bio-oil production from the pyrolysis of lignocellulosic biomass: Recent advances and future perspectives. Journal of Analytical and Applied Pyrolysis, 179, 106486. https://doi.org/10.1016/j.jaap.2024.106486

Leng, L., Yang, L., Lei, X., et al. (2022). Machine learning predicting and engineering the yield, N content, and specific surface area of biochar derived from pyrolysis of biomass. Biochar, 4(63). https://doi.org/10.1007/s42773-022-00183-w

Mighani, M., Shahi, A., & Antonioni, G. (2017). Catalytic pyrolysis of plastic waste products: Time series modeling using least square support vector machine and artificial neural network. In Proceedings of the 16th International Conference on Sustainable Energy Technologies — SET 2017.

Niksa, S. (2020). bio-FLASHCHAIN® theory for rapid devolatilization of biomass 2. Predicting total yields for torrefied woods. Fuel, 263, 116645. https://doi.org/10.1016/j.fuel.2019.116645

Parmar, K. (2017). Biomass — An overview on composition characteristics and properties. IRAInternational Journal of Applied Sciences, 7(1), 42–51. http://dx.doi.org/10.21013/jas.v7.n1.p4

Pielsticker, S., Debiagi, P., Cerciello, F., Hasse, Ch., & Kneer, R. (2024). Comparative analysis of pyrolysis models including SFOR, CRECK, and Bio-CPD to predict reaction kinetics and products from extracted biomass components. Fuel, 371, part A, 131867. https://doi.org/10.1016/j.fuel.2024.131867

Popescu, F., Mahu, R., Ion, I., & Rusu, Eu. (2020). A mathematical model of biomass combustion physical and chemical processes. Energies, 13(23), 6232. https://doi.org/10.3390/en13236232

Ranganathan, P., & Gu, S. (2016). Computational fluid dynamics modelling of biomass fast pyrolysis in fluidised bed reactors, focusing different kinetic schemes. Bioresource Technology, 213, 333–341. https://doi.org/10.1016/j.biortech.2016.02.042

Ranzi, E., Debiagi, P., & Frassoldati, A. (2017). Mathematical modeling of fast biomass pyrolysis and bio-oil formation. Note II: Secondary gas-phase reactions and bio-oil formation. ACS Sustainable Chemistry & Engineering, 5(4), 2882–2896. https://doi.org/10.1021/acssuschemeng.6b03098

Ranzi, E., Debiagi, P., & Frassoldati, A. (2017). Mathematical modeling of fast biomass pyrolysis and bio-oil formation. Note I: Kinetic mechanism of biomass pyrolysis. ACS Sustainable Chemistry & Engineering, 5(4), 2867–2881. https://doi.org/10.1021/acssuschemeng.6b03096

Rongrong Lu, Jing Li, Xiong Zou, Anran Wang, & Hongguang Dong. (2023). Prediction of product distribution of low-medium rank coal pyrolysis using artificial neural networks model. Journal of the Energy Institute, 107, 101181. https://doi.org/10.1016/j.joei.2023.101181

Sharma, A., Pareek, V., & Zhang, D. (2015). Biomass pyrolysis — A review of modelling, process parameters and catalytic studies. Renewable and Sustainable Energy Reviews, 50, 1081–1096. https://doi.org/10.1016/j.rser.2015.04.193

Shen, J., Yan, M., Fang, M., & Gao, X. (2022). Machine learning-based modeling approaches for estimating pyrolysis products of varied biomass and operating conditions. Bioresource Technology Reports, 20, 101285. https://doi.org/10.1016/j.biteb.2022.101285

Tabakaev, R., Astafev, A., Shanenkova, Yu., Dubinin, Yu., Yazykov, N., & Yakovlev, V. (2019). Thermal effects investigation during biomass slow pyrolysis in a fixed bed reactor. Biomass and Bioenergy, 126, 26–33. https://doi.org/10.1016/j.biombioe.2019.05.010

Ting Yan Li, Huan Xiang, Yang Yang, Jiawei Wang, & Güray Yildiz. (2021). Prediction of char production from slow pyrolysis of lignocellulosic biomass using multiple nonlinear regression and artificial neural network. Journal of Analytical and Applied Pyrolysis, 159, 105286. https://doi.org/10.1016/j.jaap.2021.105286

Tsekos, C., Tandurella, S., & de Jong, W. (2021). Estimation of lignocellulosic biomass pyrolysis product yields using artificial neural networks. Journal of Analytical and Applied Pyrolysis, 157, 105180. https://doi.org/10.1016/j.jaap.2021.105180

Vikram, Sh., Rosha, P., & Kumar, S. (2021). Recent modeling approaches to biomass pyrolysis: A review. Energy & Fuels, 35(9), 7406–7433. https://doi.org/10.1021/acs.energyfuels.1c00251

Wijayanti, W., Musyaroh, Sasongko, M. N., Kusumastuti, R, & Sasmoko. (2021). Modelling analysis of pyrolysis process with thermal effects by using Comsol Multiphysics. Case Studies in Thermal Engineering, 28, 101625. https://doi.org/10.1016/j.csite.2021.101625

Willmott, C. J., & Matsuura, K. (2006). On the use of dimensioned measures of error to evaluate the performance of spatial interpolators. International Journal of Geographical Information Science, 20(1), 89–102. https://doi.org/10.1080/13658810500286976

Wooters, T., Baxter, L., & Fletcher, T. (2005). CPD model calculations of black liquor and biomass pyrolysis. Prepr. Symp. — Am. Chem. Soc., Div. Fuel Chem., 50, 108–109.

Zhu, X., Li, Y., & Wang, X. (2019). Machine learning prediction of biochar yield and carbon contents in biochar based on biomass characteristics and pyrolysis conditions. Bioresource Technology, 288, 121527. https://doi.org/10.1016/j.biortech.2019.121527