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

doi:10.11949/0438-1157.20240647

摘要/Abstract

摘要：

针对生物质气化制备绿氢过程中气化效率低、氢气选择性差的挑战，提出了一种热解串联挥发分化学链重整制氢的解耦工艺。在对上述过程理论分析时发现，热解挥发分的产量和组成与生物质性质和热解操作条件之间的复杂关系难以通过传统模型化方法被准确关联，从而制约了上述工艺的精确分析调控。因此本文基于机器学习方法建立了生物质快速热解过程的产物分布预测的神经网络模型，并结合粒子群优化算法确定最佳热解条件，使热解挥发分的氢原子比和高位热值最大化，氧原子比最小化。随后，基于流程模拟对挥发分化学链重整制氢工艺进行了分析和优化。研究结果显示，所建立的神经网络模型能够准确预测热解三相产物的产率、热解气的详细组成、热解油的元素分布及高位热值等。在上述输出参数的综合测试集中，模型的平均决定系数为0.821，平均均方根误差为2.00。优化后，草本生物质（小麦秸秆、玉米秸秆）和木本生物质（榕树、松木）的热解挥发分产率为64.49%~78.62%，氢原子占比在3.77%~4.39%之间。在重整温度700 ^oC，蒸汽/生物质质量比0.71~0.88的优化工况下，小麦秸秆的氢气产量和CO₂负排放能力最高，分别为0.60 m³/kg与-1.74 kg /m³。采用生物质挥发分化学链重整制氢工艺，四种生物质的氢气产量相较常规气化分别增加了61%、35%、16%和34%。研究结果为生物质制备绿氢提供了有效的基础支撑。

关键词: 机器学习, 化学链重整, 气化, 生物质, 优化

Abstract:

To address the challenges of low gasification efficiency and poor hydrogen selectivity in the production of green hydrogen from biomass gasification, a decoupling process involving pyrolysis followed by chemical looping reforming of volatiles for hydrogen production is proposed. In the theoretical analysis of the above process, it was found that the complex relationship between the yield and composition of pyrolysis volatiles and the properties of biomass and pyrolysis operating conditions is difficult to be accurately associated with the traditional modeling method, which restricts the precise analysis and regulation of the above process. Therefore, this paper establishes a neural network model for the product distribution of the biomass fast pyrolysis process using machine learning methods and determines the optimal pyrolysis conditions using the particle swarm optimization algorithm. The goal is to maximize the hydrogen atom ratio and heating value of the pyrolysis volatiles while minimizing the oxygen atom ratio. Subsequently, the process of hydrogen production from chemical looping reforming of volatiles was analyzed and optimized through process simulation. The results show that the established neural network model can accurately predict the yield of the three-phase pyrolysis products, the detailed composition of the pyrolysis gas, the elemental distribution of the pyrolysis oil, and the higher heating value, etc. The average coefficient of determination of the predictions is 0.821, and the average root mean square error is 2.00, in the test set of the above output parameters. After optimization, the pyrolysis volatiles yield for herbaceous biomass (wheat straw, corn stover) and woody biomass (ficus, pine wood) ranged from 64.49% to 78.62%, with a hydrogen atom ratio between 3.77% and 4.39%. Under optimal conditions at a reforming temperature of 700 ℃ and a steam-to-biomass mass ratio of 0.71 to 0.88, wheat straw showed the highest hydrogen yield and CO₂ negative emission capability, with values of 0.60 m³/kg and -1.74 kg /m³, respectively. Using chemical looping reforming of biomass volatiles for hydrogen production, the hydrogen yield from the four types of biomass increased by 61%, 35%, 16%, and 34% respectively compared to conventional gasification. The research results provide effective foundational support for the production of green hydrogen from biomass.

Key words: machine learning, chemical looping reforming, gasification, biomass, optimization

中图分类号:

TQ 051

刘根, 孙仲顺, 张博, 张榕江, 吴志强, 杨伯伦. 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.

Gen LIU, Zhongshun SUN, Bo ZHANG, Rongjiang ZHANG, Zhiqiang WU, Bolun YANG. 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.

图/表 15

参考文献 56

1	Osman A I, Mehta N, Elgarahy A M, et al. Hydrogen production, storage, utilisation and environmental impacts: a review[J]. Environmental Chemistry Letters, 2022, 20(1): 153-188.
2	Kuo P C, Chen J R, Wu W, et al. Hydrogen production from biomass using iron-based chemical looping technology: validation, optimization, and efficiency[J]. Chemical Engineering Journal, 2018, 337: 405-415.
3	张榕江, 张博, 刘根, 等. 化学链制化学品工艺及循环材料研究进展[J]. 化工学报, 2023, 74(10): 3979-3994.
	Zhang R J, Zhang B, Liu G, et al. Progress in chemical looping process for chemical production and looping materials research[J]. CIESC Journal, 2023, 74(10): 3979-3994.
4	Waldheim L, Nilsson T. Heating value of gases from biomass gasification[R]. Report prepared for: IEA bioenergy agreement, Task 20-thermal gasification of biomass, 2001.
5	Zhang B, Li Y C, Yang B L, et al. Controlling the reaction microenvironments through an embedding strategy to strengthen the chemical looping reforming of methane based on decoupling process[J]. Chemical Engineering Journal, 2022, 446: 137061.
6	Zhang B, Sun Z S, Li Y C, et al. Chemical looping reforming characteristics of methane and toluene from biomass pyrolysis volatiles based on decoupling strategy: embedding NiFe₂O₄ in SBA-15 as an oxygen carrier[J]. Chemical Engineering Journal, 2023, 466: 143228.
7	Liu G, Zhang R J, Sun Z S, et al. Carbon-negative syngas production: a comprehensive assessment of biomass pyrolysis coupling chemical looping reforming[J]. AIChE Journal, 2023, 69(12): e18254.
8	Zhang R J, Liu G, Huo C B, et al. Tailoring catalytic and oxygen release capability in LaFe_1- _x Ni _x O₃ to intensify chemical looping reactions at medium temperatures[J]. ACS Catalysis, 2024, 14(10): 7771-7787.
9	Liu G, Zhang R J, Zhang B, et al. Performance evaluation of torrefaction coupled with a chemical looping gasification process under autothermal conditions: flexible syngas production from biomass[J]. Energy & Fuels, 2023, 37(1): 424-438.
10	Palma C F. Modelling of tar formation and evolution for biomass gasification: a review[J]. Applied Energy, 2013, 111: 129-141.
11	Sharma A, Pareek V, Zhang D K. Biomass pyrolysis—a review of modelling, process parameters and catalytic studies[J]. Renewable and Sustainable Energy Reviews, 2015, 50: 1081-1096.
12	Li P, Shi X P, Wang X H, et al. Bio-oil from biomass fast pyrolysis: yields, related properties and energy consumption analysis of the pyrolysis system[J]. Journal of Cleaner Production, 2021, 328: 129613.
13	Hameed S, Sharma A, Pareek V, et al. A review on biomass pyrolysis models: kinetic, network and mechanistic models[J]. Biomass and Bioenergy, 2019, 123: 104-122.
14	Leng E W, He B, Chen J W, et al. Prediction of three-phase product distribution and bio-oil heating value of biomass fast pyrolysis based on machine learning[J]. Energy, 2021, 236: 121401.
15	Chong J W, Thangalazhy-Gopakumar S, Tan R R, et al. Estimation of fast pyrolysis bio-oil properties from feedstock characteristics using rough-set-based machine learning[J]. International Journal of Energy Research, 2022, 46(13): 19159-19176.
16	Ali N, Saleem M, Shahzad K, et al. Bio-oil production from fast pyrolysis of cotton stalk in fluidized bed reactor[J]. Arabian Journal for Science and Engineering, 2015, 40(11): 3019-3027.
17	Jung S H, Kang B S, Kim J S. Production of bio-oil from rice straw and bamboo sawdust under various reaction conditions in a fast pyrolysis plant equipped with a fluidized bed and a char separation system[J]. Journal of Analytical and Applied Pyrolysis, 2008, 82(2): 240-247.
18	Yang S Y, Wu C Y, Chen K H. The physical characteristics of bio-oil from fast pyrolysis of rice straw[J]. Advanced Materials Research, 2011, 328: 881-886.
19	Yang S I, Wu M S, Wu C Y. Application of biomass fast pyrolysis (part Ⅰ): Pyrolysis characteristics and products[J]. Energy, 2014, 66: 162-171.
20	Duman G, Okutucu C, Ucar S, et al. The slow and fast pyrolysis of cherry seed[J]. Bioresource Technology, 2011, 102(2): 1869-1878.
21	Kim S J, Jung S H, Kim J S. Fast pyrolysis of palm kernel shells: influence of operation parameters on the bio-oil yield and the yield of phenol and phenolic compounds[J]. Bioresource Technology, 2010, 101(23): 9294-9300.
22	Sulaiman F, Abdullah N. Optimum conditions for maximising pyrolysis liquids of oil palm empty fruit bunches[J]. Energy, 2011, 36(5): 2352-2359.
23	Heidari A, Stahl R, Younesi H, et al. Effect of process conditions on product yield and composition of fast pyrolysis of Eucalyptus grandis in fluidized bed reactor[J]. Journal of Industrial and Engineering Chemistry, 2014, 20(4): 2594-2602.
24	Koo W M, Jung S H, Kim J S. Production of bio-oil with low contents of copper and chlorine by fast pyrolysis of alkaline copper quaternary-treated wood in a fluidized bed reactor[J]. Energy, 2014, 68: 555-561.
25	Montoya J I, Valdés C, Chejne F, et al. Bio-oil production from Colombian bagasse by fast pyrolysis in a fluidized bed: an experimental study[J]. Journal of Analytical and Applied Pyrolysis, 2015, 112: 379-387.
26	Ali N, Saleem M, Shahzad K, et al. Effect of operating parameters on production of bio-oil from fast pyrolysis of maize stalk in bubbling fluidized bed reactor[J]. Polish Journal of Chemical Technology, 2016, 18(3): 88-96.
27	Salehi E, Abedi J, Harding T. Bio-oil from sawdust: effect of operating parameters on the yield and quality of pyrolysis products[J]. Energy & Fuels, 2011, 25(9): 4145-4154.
28	Kang B S, Lee K H, Park H J, et al. Fast pyrolysis of radiata pine in a bench scale plant with a fluidized bed: influence of a char separation system and reaction conditions on the production of bio-oil[J]. Journal of Analytical and Applied Pyrolysis, 2006, 76(1/2): 32-37.
29	Pattiya A, Suttibak S. Production of bio-oil via fast pyrolysis of agricultural residues from cassava plantations in a fluidised-bed reactor with a hot vapour filtration unit[J]. Journal of Analytical and Applied Pyrolysis, 2012, 95: 227-235.
30	Park H J, Park Y K, Kim J S. Influence of reaction conditions and the char separation system on the production of bio-oil from radiata pine sawdust by fast pyrolysis[J]. Fuel Processing Technology, 2008, 89(8): 797-802.
31	Park H J, Park Y K, Dong J I, et al. Pyrolysis characteristics of oriental white oak: kinetic study and fast pyrolysis in a fluidized bed with an improved reaction system[J]. Fuel Processing Technology, 2009, 90(2): 186-195.
32	Agblevor F A, Besler S, Wiselogel A E. Fast pyrolysis of stored biomass feedstocks[J]. Energy & Fuels, 1995, 9(4): 635-640.
33	Fahmi R, Bridgwater A V, Donnison I, et al. The effect of lignin and inorganic species in biomass on pyrolysis oil yields, quality and stability[J]. Fuel, 2008, 87(7): 1230-1240.
34	Lee K H, Kang B S, Park Y K, et al. Influence of reaction temperature, pretreatment, and a char removal system on the production of bio-oil from rice straw by fast pyrolysis, using a fluidized bed[J]. Energy & Fuels, 2005, 19(5): 2179-2184.
35	Eom I Y, Kim J Y, Lee S M, et al. Comparison of pyrolytic products produced from inorganic-rich and demineralized rice straw (Oryza sativa L.) by fluidized bed pyrolyzer for future biorefinery approach[J]. Bioresource Technology, 2013, 128: 664-672.
36	Kim K H, Bai X L, Rover M, et al. The effect of low-concentration oxygen in sweep gas during pyrolysis of red oak using a fluidized bed reactor[J]. Fuel, 2014, 124: 49-56.
37	Mullen C A, Boateng A A, Goldberg N M, et al. Bio-oil and bio-char production from corn cobs and stover by fast pyrolysis[J]. Biomass and Bioenergy, 2010, 34(1): 67-74.
38	Bok J P, Choi H S, Choi Y S, et al. Fast pyrolysis of coffee grounds: characteristics of product yields and biocrude oil quality[J]. Energy, 2012, 47(1): 17-24.
39	Kim J Y, Oh S, Hwang H, et al. Assessment of miscanthus biomass (Miscanthus sacchariflorus) for conversion and utilization of bio-oil by fluidized bed type fast pyrolysis[J]. Energy, 2014, 76: 284-291.
40	Bok J P, Choi H S, Choi J W, et al. Fast pyrolysis of Miscanthus sinensis in fluidized bed reactors: characteristics of product yields and biocrude oil quality[J]. Energy, 2013, 60: 44-52.
41	Park J W, Heo J, Ly H V, et al. Fast pyrolysis of acid-washed oil palm empty fruit bunch for bio-oil production in a bubbling fluidized-bed reactor[J]. Energy, 2019, 179: 517-527.
42	Ly H V, Lim D H, Sim J W, et al. Catalytic pyrolysis of tulip tree (Liriodendron) in bubbling fluidized-bed reactor for upgrading bio-oil using dolomite catalyst[J]. Energy, 2018, 162: 564-575.
43	Salehi E, Abedi J, Harding T G, et al. Bio-oil from sawdust: design, operation, and performance of a bench-scale fluidized-bed pyrolysis plant[J]. Energy & Fuels, 2013, 27(6): 3332-3340.
44	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 M X, Sun Q D, Fu Z, 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.
45	Maurer A, Pontil M, Romera-Paredes B. The benefit of multitask representation learning[J]. Journal of Machine Learning Research, 2016, 17(81): 1-32.
46	黄勇, 王宁波, 刘巧霞, 等. 热解条件对西湾煤与秸秆流化床加压共热解的影响[J]. 煤炭转化, 2022, 45(1): 11-19.
	Huang Y, Wang N B, Liu Q X, et al. Effects of pyrolysis conditions on pressurized co-pyrolysis of xiwan coal and wheat straw in fluidized bed[J]. Coal Conversion, 2022, 45(1): 11-19.
47	Chireshe F, Collard F X, Görgens J F. Production of low oxygen bio-oil via catalytic pyrolysis of forest residues in a kilogram-scale rotary kiln reactor[J]. Journal of Cleaner Production, 2020, 260: 120987.
48	Abhijeet P, Swagathnath G, Rangabhashiyam S, et al. Prediction of pyrolytic product composition and yield for various grass biomass feedstocks[J]. Biomass Conversion and Biorefinery, 2020, 10(3): 663-674.
49	Tsekos C, Tandurella S, Jong W D. Estimation of lignocellulosic biomass pyrolysis product yields using artificial neural networks[J]. Journal of Analytical and Applied Pyrolysis, 2021, 157: 105180.
50	Naik S, Goud V V, Rout P K, et al. Characterization of Canadian biomass for alternative renewable biofuel[J]. Renewable Energy, 2010, 35(8): 1624-1631.
51	Lu W, Ronghou L, Chen S, et al. Classification and comparison of physical and chemical properties of corn stalk from three regions in China[J]. International Journal of Agricultural and Biological Engineering, 2014, 7(6): 98-106.
52	Pérez-Arévalo J J, Velázquez-Martí B. Evaluation of pruning residues of Ficus benjamina as a primary biofuel material[J]. Biomass and Bioenergy, 2018, 108: 217-223.
53	Hu J J, Li C, Guo Q H, et al. Syngas production by chemical-looping gasification of wheat straw with Fe-based oxygen carrier[J]. Bioresource Technology, 2018, 263: 273-279.
54	Zhao Z G, Kong W G, Wu S, et al. High quality syngas production from catalytic steam gasification of biomass with calcium-rich construction waste[J]. Journal of the Energy Institute, 2023, 111: 101433.
55	Kumar P, Subbarao P M V, Kala L D, et al. Experimental assessment of producer gas generation using agricultural and forestry residues in a fixed bed downdraft gasifier[J]. Chemical Engineering Journal Advances, 2023, 13: 100431.
56	Song Y H, Tian Y, Zhou X, et al. Simulation of air-steam gasification of pine sawdust in an updraft gasification system for production of hydrogen-rich producer gas[J]. Energy, 2021, 226: 120380.

输入变量			单位
原料性质	工业分析^①	灰分(Ash)	%（质量）
		挥发分(V)	%（质量）
		固定碳(FC)	%（质量）
	元素分析^①	碳(C)	%（质量）
		氢(H)	%（质量）
		氧(O)	%（质量）
		氮(N)	%（质量）
	组成分析	纤维素(Cel)	%（质量）
		半纤维素(Hem)	%（质量）
		木质素(Lig)	%（质量）
热解操作条件		生物质粒径(PS)	μm
		热解温度(T)	℃
		进料速度(F)	kg/h
		载气流量(G)	m³/h

输入变量			单位
原料性质	工业分析^①	灰分(Ash)	%（质量）
		挥发分(V)	%（质量）
		固定碳(FC)	%（质量）
	元素分析^①	碳(C)	%（质量）
		氢(H)	%（质量）
		氧(O)	%（质量）
		氮(N)	%（质量）
	组成分析	纤维素(Cel)	%（质量）
		半纤维素(Hem)	%（质量）
		木质素(Lig)	%（质量）
热解操作条件		生物质粒径(PS)	μm
		热解温度(T)	℃
		进料速度(F)	kg/h
		载气流量(G)	m³/h

单元	关键设备参数	模拟方法
重整反应器	流化床, 1 atm, 700~850℃	RGibbs block
一级再生反应器	流化床, 1 atm, 绝热运行	RCSTR block
二级再生反应器	流化床, 1 atm, 950℃	RStoic block
MEA CO₂ 捕集系统	CO₂ 捕集率：90%, 热量需求： 3538 kJ/kg	Sep block
氢气变压吸附分离系统	H₂分离效率：85%, H₂纯度：99.99% (mol)	Sep block

单元	关键设备参数	模拟方法
重整反应器	流化床, 1 atm, 700~850℃	RGibbs block
一级再生反应器	流化床, 1 atm, 绝热运行	RCSTR block
二级再生反应器	流化床, 1 atm, 950℃	RStoic block
MEA CO₂ 捕集系统	CO₂ 捕集率：90%, 热量需求： 3538 kJ/kg	Sep block
氢气变压吸附分离系统	H₂分离效率：85%, H₂纯度：99.99% (mol)	Sep block

原料	元素分析^①/%（质量）				工业分析^①/%（质量）			组成分析/%（质量）			文献
原料	C	H	N	O	Ash	V	FC	Cel	Hem	Lig	文献
小麦秸秆	41.03	6.02	0.14	51.38	1.38	83.30	15.32	34.60	29.30	21.30	[50]
玉米秸秆	43.56	5.85	1.18	45.99	7.13	75.59	17.28	33.46	23.77	26.19	[51]
榕树	44.40	7.21	0.91	47.48	10.90	85.13	3.97	31.01	15.69	24.92	[52]
松木	48.22	6.30	0.14	44.45	1.59	87.50	10.91	39.00	34.00	12.00	[4]