生質碳於焙燒過程之熱質傳遞模擬

Abstract

木質纖維素是地球上存量豐富的生質原料，而且藉著造林可以補充其消耗，因此是目前是一個具潛力的生質原料之一。木質生質原料經由慢速熱裂解所產生的生質碳(Biochar)，是最方便運輸來提供能源產業立即利用的能量轉換方法。但一般熱裂解投入能量多，生成物卻少，因此輕度熱裂解的焙燒(Torrefaction)製程成為一個較好的方案。木質生質原料經焙燒後，固態的產物生質碳，其體積減少可降低運輸成本，且容易破碎可以直接用於發電廠的粉煤機。本研究的目的在於以柳杉(Cryptomeria japonica)為原料，探討木質材料焙燒(Torrefaction) 的質量與能量變化，經由建立數學模式以進行模擬分析。本研究著重在於探討低溫短時間的木質原料熱裂解的焙燒程序。結合熱傳遞、質量傳遞和化學反應動力的原理，以建立數學模式進行模擬，並進行實驗結果的成份分析來驗證數學模式。模擬的結果顯示出半纖維素為焙燒反應的主要反應物，在250℃下經過了3600秒後，其濃度已減少至近乎零，而在270℃和290℃下只分別經過了1000秒、2000秒就反應耗盡。而纖維素和木質素在250℃下濃度變化的幅度很小，在270℃之後有明顯變化，所以在一般的焙燒反應中，半纖維素為主要反應物。而在重量變化方面來說，以生質碳的收率來說，以250℃、270℃、290℃、加熱3600秒，生質碳的收率分別為為77%、69%、65%。而模擬所得數值約與實驗值相差5%。而在原料化學分析方面，從化學成份百分比的曲線中可以看出模擬半纖維素、纖維素和木質素的變化曲線與實際分析出的值，整體上而言相當接近吻合。

Lignocellulosic biomass is the potential resource for production of raw materials and can be easily supplemented through reforestation. Conventionally, biochar is produced by slow pyrolysis and directly used in energy industries. However, current processes of pyrolysis are inefficient. Biochar, produced from slight pyrolysis “torrefaction”, has less volume and is cost effective in its transportation to power plant for pulverizer. The purpose of this study was to develop a coupled heat and mass transfer model ,including chemical kinetics. After that, the measurements of mass change during torrefaction of Cryptomeria japonica were compared with the simulation results. Simulation results shows that hemicellulose is the main reactant of the torrefaction. After the sample was heated for 3600s at 250℃, concentration of hemicellulose was reduced to nearly zero. After heating sample for 2000s at 270℃,and 1000s at 290℃, hemicellulose was completely reacted. However, concentration of cellulose and lignin changed slightly at 250℃, the change was significant when temperature was over 270℃. These findings indicate that hemicellulose is the main reactant of the torrefaction. After heating for 3600s, experimental results showed that yields of biochar at 250℃,270℃ and 290℃ were 77%, 69% and 65%. The model was successfully constructed with relative error 5% between the experiment and simulation. Additionally, the percentage curves of hemicellulose, cellulose, and lignin showed a correct approach with that of chemical analysis of the heated sample.