Behaviour of Gaseous Chlorine and Alkali Metals During Biomass Thermal Utilisation

The behaviour of gaseous chlorine and alkali metals of three sorts of biomass (Danish straw, Swedish wood, and sewage sludge) in combustion or gasification is investigated by the chemical equilibrium calculating tool. The ranges of temperature, air-to-fuel ratio, and pressure are varied widely in the calculations (T=400-1800 K, gimel=0-1.8, and P=0.1-2.0 MPa). Results show that the air excess coefficient only has less significant influence on the release of gaseous chlorine and potassium or sodium during combustion. However, in biomass gasification, the influence of the air excess coefficient is very significant. Increasing air excess coefficient enhances the release of HCl(g), KOH(g), or NaOH(g) as well as it reduces the formation of KCl(g), NaCl(g), K(g), or Na(g). In biomass combustion or straw and sludge gasification, increasing pressure enhances the release of HCl(g) and reduces the amount of KCI(g), NaCl(g), KCI(g), or NaOH(g) at high temperatures. However, during wood gasification, the pressure enhances the formation of KOH(g) and KCI(g) and reduces the release of K(g) and HCl(g) at high temperatures. During wood and sewage sludge pyrolysis, nitrogen addition enhances the formation of KCN(g) and NaCN(g) and reduces the release of K(g) and Na(g). Kaolin addition in straw combustion may enhance the formation of potassium aluminosilicate in ash and significantly reduces the release of KCl(g) and KOH(g) and increases the formation of HCl(g).

内循环流化床生物质气化过程的神经网络模型

基于BP人工神经网络原理，利用MATLAB神经网络工具箱，以实验得到的57组气化实验数据作为样本，建立了一个以加料量和送风量为输入变量，以燃气热值、产气率、碳转化率和气化效率为输出变量，用于描述连续稳定气化过程的内循环流化床生物质气化模型。对模型的隐层节点数和训练周期改变对模拟结果的影响进行了分析，发现当隐层节点数为20，训练步骤为50步，模型的4个输出变量的模拟结果与实验结果相关系数均超过0．95；同时对该模型的预测能力进行了考察，模型预测结果与实验结果吻合良好，证明了该模型具有较强的泛化能力，为生物质内循环流化床气化系统的优化设计和自动控制提供新思路。