Abstract: Biomass gasification is a potential pathway for thermochemically generating renewable producer gas, which serves as a good substitute fuel in heating and electricity section and is beneficial to the reduction of greenhouse gas emission. Biomass feedstock varies significantly in its composition, especially the content of moisture, posing a challenge for biomass gasification process design and operation in practice, however few research were carried out to elucidate the gasification principles of biomass with different moisture content. In this research, the effects of moisture content and process parameters on biomass gasification characteristics including syngas quality and energy balance are investigated by adopting machine learning and process modeling approaches. The prediction accuracy of the two approaches is first validated by comparing with experimental data. The obtained results indicate that the moisture content of biomass has a great negative impact on the low heating value (LHV) of produced gas, but does not significantly affect the carbon conversion efficiency (CCE) in the downdraft gasifier. The LHV of the produced gas decreases when the air equivalence rate (ER) increases due to the increment of carbon dioxide in the producer gas, but CCE increases with the increase of ER. The energy balance analysis suggests that ER increase with the increment of moisture content in biomass in order to maintain energy balance of the system. Pretreatment of biomass by drying is favorable to maintaining the quality of syngas, but the tradeoff is to consume a certain amount of producer gas to supply heat for the drying process. The consumption rate of the producer gas increases as the moisture content of biomass goes up, which is characterized by a nearly linear increase with the moisture content in the range of 20wt%~60wt% and an exponentially increment as the moisture content goes up beyond 60wt%. The current research provides fundamental insights on gasification characteristics of biomass with different moisture contents.

Key words: Biomass gasification, Process modelling, Machine learning, drying pretreatment, Downdraft gasifier

摘要： 采用机器学习方法和Aspen过程建模分析方法系统研究了含水率对生物质气化过程的影响规律。研究结果表明生物质含水率对气化气低位热值影响较大，含水率升高，热值降低，生物质含水率对碳转化率影响较小；随着空气当量比升高，气化气低位热值降低，而气化过程碳转化率提高。气化过程能量平衡分析表明，随着生物质含水率升高，气化过程维持能量平衡所需的空气当量比升高。生物质干燥预处理可保证气化气品质，但需消耗部分气化气为干燥供能，并且随着生物质含水率升高，干燥供热消耗的气化气比例提高，消耗比例在含水率20wt%~60wt%阶段近似线性增加，但是当含水率大于60wt%时呈指数升高。

关键词: 生物质气化, 过程仿真, 机器学习, 干燥预处理, 下吸式气化炉