【作者】 武志敏；

【导师】 罗勇；

【作者基本信息】 石家庄铁道大学 ， 供热、供燃气、通风及空调工程， 2012， 硕士

【摘要】 以白松锯末为生物质原料，借助自制的二级固定床热解装置研究了生物质催化裂解的气化特性；利用实验数据分析了实验条件对表观活化能的影响；研究了催化剂以及实验条件对气体能产率的影响；建立了人工神经网络模型，对不同条件下的生物质催化裂解气化特性进行了预测。首先以价格较为低廉的两种多孔熟料(高铝质多孔熟料和粘土质多孔熟料)为催化剂，详细研究了气化介质、催化裂解温度、生物质原料粒径和催化剂粒径等条件对生物质催化裂解气化特性的影响。结果表明：无论是否通入水蒸汽，在相同的催化裂解温度下，高铝质多孔熟料的催化效果均优于粘土质多孔熟料；水蒸汽作为气化介质较氮气的气化效果好；提高催化裂解温度有利于生物质的催化裂解气化；原料粒径与气化特性有一定的关系，在实验条件下，粒径介于0.1~1mm的小颗粒生物质比大颗粒的效果好；就催化剂的粒径而言，在实验条件下，粒径为1~3mm的催化剂的催化效果较好。其次建立了生物质气化焦油催化裂解反应动力学模型并求出动力学参数，从理论上对生物质催化裂解气化特性进行分析。结果表明：不催化时的表观活化能是三种实验条件下最高的。以多孔熟料作为催化剂时表观活化能有所减小，说明多孔熟料对焦油有较好的催化作用。以多孔熟料作为催化剂并且加入水蒸汽时，活化能进一步下降，说明水蒸汽的加入有助于焦油的催化裂解。在详细研究了生物质催化裂解气化特性的基础上，分别以特种焦、煅烧后的白云石(以下简称白云石)、高铝质多孔熟料(以下简称多孔熟料)为裂解催化剂，以获得高气体能产率为目标，研究了不同工况下的气体能产率以及焦油转化率。结果表明：特种焦作用下气体能产率最高，白云石次之，多孔熟料最差。借助扫描电镜(SEM)对特种焦、白云石、多孔熟料三种催化剂使用前后的微观形貌进行观测，分析了催化剂的微观结构与其催化特性的关联性，指出催化剂的微观结构对生物质的气体能产率有较大影响，催化剂的比表面积、孔隙率、孔径分布以及孔隙结构的无序度等因素均影响其活性。为了预测生物质催化裂解的气化特性，建立了三个BP人工神经网络模型，三个模型分别以多孔熟料的热解温度和催化裂解温度、白云石的比表面积和催化裂解温度、特种焦的S/B(Steam/Biomass)值和催化裂解温度为输入变量，以气体能产率和焦油转化率为输出变量，程序输出结果显示：模型1、模型2、模型3的最大相对误差分别为1.9601%、－8.6923%和8.0022%。表明在一定的条件下，建立神经网络模型预测生物质的催化裂解气化特性是可行的。更多还原

【Abstract】 Using sawdust as biomass raw material, the catalytic cracking characteristicswere investigated in a home-made two-stage fixed bed reactor. By means of theexperiment data, the effects of experiment conditions on the apparent activationenergy were analyzed; then the study which catalysts and experiment conditionaffected gas energy yield was also conducted; the last the Artificial Neural Network(ANN) models were established for simulation and prediction of biomass catalyticcracking.At first, using low-cost porous clinker catalysts (the porous high-clay clinker andthe porous high-alumina clinker), the catalytic cracking experiments were conductedat different conditions, such as gasificated liquids, pyrolysis temperature, particlediameter of material, catalyst size, and steam input or not, et al. The results show thatthe porous high-clay clinker has much lower effective than porous high-aluminaclinker on tar conversation on same pyrolysis temperature whether steam input or not;the gasification characteristics with steam were improved; the particle diameter ofmaterial has little influence on catalytic effect, compare with larger diameter ofbiomass material, the diameter of0.1~1mm biomass material was beneficial tocatalytic cracking; the smaller catalyst size of1~3mm has positive influence oncatalytic cracking.For analyzing biomass catalytic cracking characteristics in theory, reactionkinetic model has been established, and the kinetic parameters have been obtained.The results show that the apparent activation energy without using catalyst is largestamong the three conditions. The apparent activation energy become smaller afterusing porous high-alumina clinker as catalyst, which indicate that poroushigh-alumina clinker was beneficial to tar catalytic cracking. The apparent activationenergy is the smallest under using porous high-alumina clinker as catalyst and steaminput, it shows that this condition is more beneficial to tar catalytic cracking.The purpose of this chapter was to get high gas energy yield by catalytic cracking reaction of biomass. The tar conversion rate and gas energy yield wereinvestigated with the special coke, calcined dolomite, and porous high-aluminaclinker as catalysts. The results show that the special coke is the most effective amongthree kinds of catalysts, catalytic effect of dolomite is worse than the special coke andporous clinker is the worst. Gas energy yield is the highest when the special coke isused; and the gas energy yield using dolomite as catalyst is higher than that of porousclinker. The microstructure of catalysts was observed by scanning electronmicroscopy (SEM), and the influence of catalyst’s microstructure on gas energy yieldwas studied, it can conclude that activity of catalyst is influenced by surface area,porosity, pore size distribution and disorder degree of pore structure, etc.Microstructure of catalysts greatly influences gas energy yield of biomass.For predicting biomass catalytic cracking gasification characteristics, three kindsof BP Artificial Neural Network (ANN) models had been set up. Model1: usingpyrolysis temperature and cracking temperature as input variables under the conditionof using porous clinker as catalyst; Model2: using specific surface area and crackingtemperature as input variables under the condition of using calcined dolomite ascatalyst; Model3: using S/B value and cracking temperature as input variables underthe condition of using the special coke as catalyst. The max relative errors of Model1,Model2and Model3were1.9601%,－8.6932%and8.0022%respectively. Theresults show that the three kinds of ANN models were effective for simulation andprediction of biomass catalytic cracking gasification characteristics in certaincondition.更多还原