Machine learning-based prediction and experimental validation of the pyrolysis product distribution in tar-rich coals

Authors

• DONG Shuai, China Huaneng Group Yimin Coal and Power Co., Ltd., Hulunbuir 021130, ChinaFollow
• LI Hongqiang, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• WU Zhiqiang, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, ChinaFollow
• YU Zunyi, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• LYU Zitao, China Huaneng Group Yimin Coal and Power Co., Ltd., Hulunbuir 021130, China
• GUO Wei, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• YANG Panxi, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• FU Keming, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• HAO Xuanzhi, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• LIU Gen, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China
• YANG Bolun, School of Chemical Engineering and Technology, Xi’an Jiaotong University, Xi’an 710049, China; Shaanxi Key Laboratory of Energy Chemical Process Intensification, Xi’an 710049, China

Abstract

Background Tar-rich coals serve as an important coal-based oil and gas resource in China, while their pyrolysis product distribution is governed by the coupling effects of coal properties and reaction conditions. Therefore, rapidly identifying the pyrolysis product distribution patterns holds great significance for the resource evaluation and experimental design of tar-rich coals.Methods Existing studies on tar-rich coals suffer from the insufficient integration of exclusive data and limited synergistic prediction capacities for multiple products. To address these issues, this study constructed a dedicated dataset involving proximate analysis, ultimate analysis, elemental molar ratios, maceral composition, and pyrolysis conditions by systematically collecting experimental data from associated literature. Subsequently, a feedforward neural network model for multi-target regression was established for the basic prediction of pyrolysis product yields. Accordingly, this study identified primary factors controlling the pyrolysis product distribution of tar-rich coals by combining Van Krevelen diagrams, Spearman correlation analysis, and feature contribution analysis based on SHapley Additive exPlanations (SHAP) values. Finally, through pyrolysis experiments under 400‒600 ℃, this study validated the model’s capability to reproduce the major temperature response patterns of the pyrolysis products of the tar-rich coals.Results and Conclusions The established feedforward neural network model effectively reproduced the typical temperature response patterns of the pyrolysis products of tar-rich coals. Specifically, with an increase in the pyrolysis temperature, the char yield decreased, the gas yield increased, the water yield increased overall, and the tar yield increased initially and then decreased. This model exhibited an average coefficient of determination (R2) of 0.89 and an average root mean square error (RMSE) of 1.53 on the testing set. Among primary influencing factors, coal properties, particularly carbon content, vitrinite content, and volatile constituent content, generally produced more significant impacts on the product distribution than external operating parameters like heating rate and particle size. Experimental validation demonstrates that good agreement existed between predicted and experimental gas and char yields, while the tar yield was slightly overestimated under high-temperature conditions. Therefore, future improvement should focus on sample coverage and secondary pyrolysis under high-temperature conditions. The results of this study reveal the basic distribution patterns of the pyrolysis products of tar-rich coals and verify the capability of the established feedforward neural network model to reproduce the major temperature response trends of the pyrolysis products of tar-rich coals. These results will provide references for the resource evaluation, pyrolysis experimental design, and product control research of tar-rich coals.

Keywords

tar-rich coal, coal pyrolysis, product distribution, machine learning, experimental verification

DOI

10.12363/issn.1001-1986.25.07.0536

Recommended Citation

DONG Shuai, LI Hongqiang, WU Zhiqiang, et al. (2026) "Machine learning-based prediction and experimental validation of the pyrolysis product distribution in tar-rich coals," Coal Geology & Exploration: Vol. 54: Iss. 4, Article 5.
DOI: 10.12363/issn.1001-1986.25.07.0536
Available at: https://cge.researchcommons.org/journal/vol54/iss4/5

Reference

[1] WEI Yufan,WU Ying,DU Zhenyi,et al. Interaction mechanisms of coal macerals during pyrolysis:Insights from TG and Py–GC/MS experiments combined with ReaxFF MD simulations[J]. Chemical Engineering Journal,2025,504:158768.

[2] SHI Shuai,HE Jinxian,ZHANG Xiaoli,et al. Pore structure evolution of tar–rich coal with temperature–pressure controlled simulation experiments[J]. Fuel,2023,354:129298.

[3] 葛世荣,刘淑琴,樊静丽,等. 低碳化现代煤基能源开发关键技术体系[J]. 煤炭学报,2024,49(7):2949−2972

GE Shirong,LIU Shuqin,FAN Jingli,et al. Key technologies for low–carbon modern coal–based energy[J]. Journal of China Coal Society,2024,49(7):2949−2972

[4] 《矿产资源工业要求手册》编委会. 矿产资源工业要求手册[M]. 北京：地质出版社，2014.

[5] 王双明,师庆民,王生全,等. 富油煤的油气资源属性与绿色低碳开发[J]. 煤炭学报,2021,46(5):1365−1377

WANG Shuangming,SHI Qingmin,WANG Shengquan,et al. Resource property and exploitation concepts with green and low–carbon of tar–rich coal as coal–based oil and gas[J]. Journal of China Coal Society,2021,46(5):1365−1377

[6] 王双明,王虹,任世华,等. 西部地区富油煤开发利用潜力分析和技术体系构想[J]. 中国工程科学,2022,24(3):49−57

WANG Shuangming,WANG Hong,REN Shihua,et al. Potential analysis and technical conception of exploitation and utilization of tar–rich coal in western China[J]. Strategic Study of CAE,2022,24(3):49−57

[7] 王双明,鲍园,郝永辉,等. 富油煤研究进展与趋势[J]. 煤田地质与勘探,2024,52(4):1−11

WANG Shuangming,BAO Yuan,HAO Yonghui,et al. Research on tar–rich coals:Progress and prospects[J]. Coal Geology & Exploration,2024,52(4):1−11

[8] 吴志强,杨盼曦,郭伟,等. 富油煤绿色低碳转化技术研究进展[J]. 绿色矿山,2023,1(1):138−165

WU Zhiqiang,YANG Panxi,GUO Wei,et al. Research progress on green and low–carbon conversion technology of oil–rich coal[J]. Journal of Green Mine,2023,1(1):138−165

[9] 王双明,师庆民,孙强,等. 富油煤原位热解技术战略价值与科学探索[J]. 煤田地质与勘探,2024,52(7):1−13

WANG Shuangming,SHI Qingmin,SUN Qiang,et al. Strategic value and scientific exploration of in–situ pyrolysis of tar–rich coals[J]. Coal Geology & Exploration,2024,52(7):1−13

[10] 王敏欣,师印光,刘长波,等. 机器学习模型预测煤热解产物分布研究[J]. 煤炭转化,2024,47(4):11−22

WANG Minxin,SHI Yinguang,LIU Changbo,et al. Prediction of coal pyrolysis product distribution using machine learning model[J]. Coal Conversion,2024,47(4):11−22

[11] NIKSA S. FLASHCHAIN theory for rapid coal devolatilization kinetics. 3. Modeling the behavior of various coals[J]. Energy & Fuels,1991,5(5):673−683.

[12] SERIO M A,HAMBLEN D G,MARKHAM J R,et al. Kinetics of volatile product evolution in coal pyrolysis:Experiment and theory[J]. Energy & Fuels,1987,1(2):138−152.

[13] JIANG Peng,FAN Jing,LI Lin,et al. A hybrid approach combining mechanism–guided data augmentation and machine learning for biomass pyrolysis[J]. Chemical Engineering Science,2024,296:120227.

[14] MITTAL S,PATHAK S,DHAWAN H,et al. A machine learning approach to improve ignition properties of high–ash Indian coals by solvent extraction and coal blending[J]. Chemical Engineering Journal,2021,413:127385.

[15] ZHANG Tianli,ZHANG Chenxu,REN Hai,et al. Co–pyrolysis of coal–derived sludge and low–rank coal:Thermal behaviour and char yield prediction[J]. Fuel Processing Technology,2025,267:108165.

[16] 乔军伟,王昌建,赵泓超,等. 基于煤岩煤质多元指标的BP神经网络焦油产率预测方法研究[J]. 煤田地质与勘探,2024,52(7):108−118

QIAO Junwei,WANG Changjian,ZHAO Hongchao,et al. A method for predicting the tar yield of tar–rich coals based on the BP neural network using multiple indicators of coal petrography and coal quality[J]. Coal Geology & Exploration,2024,52(7):108−118

[17] WENG Junjie,WANG Jingyi,CHENG Zhanjun,et al. Machine learning and SHapley Additive exPlanations to predict product characteristics from coal and plastic co–pyrolysis[J]. Fuel,2026,404:136238.

[18] 王昌建,乔军伟. 基于BP神经网络的富油煤焦油产率预测[J]. 地质论评,2023,69(增刊1):569−572

WANG Changjian,QIAO Junwei. Rich coal tar yield prediction based on BP neural network[J]. Geological Review,2023,69(Sup.1):569−572

[19] RAHIMI M,MASHHADIMOSLEM H,VO THANH H,et al. Yield prediction and optimization of biomass–based products by multi–machine learning schemes:Neural,regression and function–based techniques[J]. Energy,2023,283:128546.

[20] 王茂先,孙启典,付哲,等. 分子水平动力学模型和机器学习方法相结合研究废弃塑料热解[J]. 化工学报,2024,75(11):4320−4332

WANG Maoxian,SUN Qidian,FU Zhe,et al. Understanding pyrolysis process of polyethylene by combined method of molecular–level kinetic model with machine learning[J]. CIESC Journal,2024,75(11):4320−4332

[21] MAURER A,PONTIL M,ROMERA–PAREDES B. The benefit of multitask representation learning[J]. Journal of Machine Learning Research,2016,17(81):1−32.

[22] YAO Qiuxiang,WANG Linyang,MA Mingming,et al. A quantitative investigation on pyrolysis behaviors of metal ion–exchanged coal macerals by interpretable machine learning algorithms[J]. Energy,2024,300:131614.

[23] 刘根,孙仲顺,张博,等. 机器学习驱动的生物质热解模型建立及挥发分化学链重整制氢工艺优化[J]. 化工学报,2024,75(11):4333−4347

LIU Gen,SUN Zhongshun,ZHANG Bo,et al. Establishment of machine learning–driven biomass pyrolysis model and optimization of volatiles chemical looping reforming hydrogen production process[J]. CIESC Journal,2024,75(11):4333−4347

[24] 李奕飞. 煤与废塑料共热解特性及焦油产率预测模型研究[D]. 南京：东南大学，2023.

LI Yifei. Study on co–pyrolysis characteristics of coal and waste plastics and prediction model of tar yield[D]. Nanjing：Southeast University，2023.

[25] SINGH Y,SINGH D,SINGH N K,et al. Production of bio–oil from lychee–based biomass through pyrolysis and maximization of bio–oil yield with statistical and machine learning techniques[J]. Journal of Cleaner Production,2023,413:137472.

[26] KANDPAL S,TAGADE A,SAWARKAR A N. Critical insights into ensemble learning with decision trees for the prediction of biochar yield and higher heating value from pyrolysis of biomass[J]. Bioresource Technology,2024,411:131321.

[27] LIU Shanke,YANG Yan,YU Lijun,et al. Predicting gas production by supercritical water gasification of coal using machine learning[J]. Fuel,2022,329:125478.

[28] TEE J X,SELVARAJOO A,ARUMUGASAMY S K. Prediction of carbon sequestration of biochar produced from biomass pyrolysis by artificial neural network[J]. Journal of Environmental Chemical Engineering,2022,10(3):107640.

[29] ZHANG Tonghuan,CAO Danyang,FENG Xin,et al. Machine learning prediction of bio–oil characteristics quantitatively relating to biomass compositions and pyrolysis conditions[J]. Fuel,2022,312:122812.

[30] LI Jie,ZHANG Weijin,LIU Tonggui,et al. Machine learning aided bio–oil production with high energy recovery and low nitrogen content from hydrothermal liquefaction of biomass with experiment verification[J]. Chemical Engineering Journal,2021,425:130649.

[31] 俞尊义,刘根,郭伟,等. 产物导向与机器学习驱动的富油煤热解提油热效应评估[J]. 煤炭学报,2024,49(8):3635−3646

YU Zunyi,LIU Gen,GUO Wei,et al. Thermal effect evaluation of tar–rich coal pyrolysis driven by product–oriented and machine learning[J]. Journal of China Coal Society,2024,49(8):3635−3646

[32] FAN Zhanglei,FAN Gangwei,ZHANG Dongsheng,et al. Modelling of flue gas injection and collaborative optimization of multi–injection parameters for efficient coal–based carbon sequestration combined with BP neural network parallel genetic algorithms[J]. Fuel,2024,368:131536.

[33] PATHY A,MEHER S,BALASUBRAMANIAN P. Predicting algal biochar yield using eXtreme Gradient Boosting (XGB) algorithm of machine learning methods[J]. Algal Research,2020,50:102006.

[34] SELVAM S M,BALASUBRAMANIAN P. Influence of biomass composition and microwave pyrolysis conditions on biochar yield and its properties:A machine learning approach[J]. BioEnergy Research,2023,16(1):138−150.

[35] TAO Junyu,YIN Xiaoxiao,YAO Xilei,et al. Prediction of NH₃ and HCN yield from biomass fast pyrolysis:Machine learning modeling and evaluation[J]. Science of the Total Environment,2023,885:163743.

[36] MA Jingjing,ZHANG Shuai,LIU Xiangjun,et al. Neural network–based prediction of levoglucosan yield:A novel modeling approach[J]. Energy,2025,320:135396.

[37] 俞尊义,郭伟,杨盼曦,等. 陕北富油煤热解提油产物分布特性研究[J]. 煤田地质与勘探,2024,52(7):176−188

YU Zunyi,GUO Wei,YANG Panxi,et al. Distributions of products from the pyrolysis of tar–rich coals for tar extraction in northern Shaanxi Province,China[J]. Coal Geology & Exploration,2024,52(7):176−188

[38] DONALDSON B,HUGHES B,SALINAS P,et al. Investigation of in–situ superheated steam hydrogen production based on pyrolysis of different coal types[J]. International Journal of Hydrogen Energy,2025,143:760−768.

[39] 许涛,陈杰,伍永平,等. 酸洗脱矿下富油煤热解过程中官能团演化的同步原位红外研究[J]. 煤炭学报,2025,50(增刊2):1439−1448

XU Tao,CHEN Jie,WU Yongping,et al. In–situ synchrotron ftir study on functional group evolution during pyrolysis of tar–rich coal de–mineralized by acid washing[J]. Journal of China Coal Society,2025,50(Sup.2):1439−1448

[40] 郭伟,杨盼曦,俞尊义,等. 陕北富油煤分子模型构建及其热解提油分子动力学特性[J]. 煤田地质与勘探,2024,52(7):132−143

GUO Wei,YANG Panxi,YU Zunyi,et al. Molecular modeling of tar–rich coals from northern Shaanxi and their molecular dynamic characteristics in the process of pyrolysis for tar extraction[J]. Coal Geology & Exploration,2024,52(7):132−143

[41] 黄淄博,刘倩如,刘红,等. H₂O气氛下富油煤热解过程的反应分子动力学模拟[J]. 煤田地质与勘探,2024,52(7):189−199

HUANG Zibo,LIU Qianru,LIU Hong,et al. Simulation of the pyrolysis process of tar–rich coals in an H₂O atmosphere based on the reactive force field molecular dynamics[J]. Coal Geology & Exploration,2024,52(7):189−199

[42] ZHANG Zhi,ZHOU Anning,SHI Zhiwei,et al. Explaining relationships between chemical structure and tar–rich coal pyrolysis products yield based on Pearson correlation coefficient[J]. Fuel,2025,395:135029.

[43] LI Shuai,ZENG Yongfu,QU Rui,et al. Insight into fast infrared–assisted and coal rank on pyrolysis kinetics and products distribution[J]. Journal of Analytical and Applied Pyrolysis,2025,187:106986.