木質素水熱液化產製生物油之動力學與製程模型開發

Abstract

木質素是生質能源領域中豐富但未充分利用的資源，其高效轉化對發展再生能源具有重要意義。水熱液化（Hydrothermal Liquefaction, HTL）技術在亞臨界水條件下將木質素轉化為高熱值生物油，被視為極具潛力的木質素高值化途徑。然而，目前對其水熱液化動力學機制與產物生成行為的瞭解仍不完整。本研究針對上述挑戰，開發了一套嚴謹的木質素水熱液化動力學模型並建立完整的製程模擬，以提升對反應行為的預測能力並最佳化生物油產製效率。 在研究方法上，首先對既有總集式動力學模型進行系統性改良，提出不同於過往文獻的反應路徑與動力學架構，使之更準確地描述木質素在水熱環境下分解為油相、水相、固相及氣相產物的速率。接著，本研究建立了首創的木質素HTL嚴格反應動力學模型，引入多種模型化合物，描述各相中具體產物間的平衡與轉化反應。在模型建立完成後，將其輸入Aspen Plus軟體中模擬連續式木質素HTL製程，設計其分離與熱整合單元，並最佳化其操作條件。 研究結果顯示，改良後的總集模型在木質素各相產物產率的預測上較文獻模型更加準確，平均絕對誤差為2.03 wt%。而嚴格動力學模型能精確模擬不同反應條件下的各相產率與產物分佈，平均絕對誤差為4.51 wt%。在製程模擬與最佳化方面，利用嚴格模型輸出的產物組成資料進行Aspen Plus製程模擬，使用多目標最佳化，結果顯示生物油的產率與熱值之間存在明顯拮抗關係，在最佳條件下（溫度270 – 340 ˚C、滯留時間120分鐘、水過量比5）隨著溫度升高有助於提升油品熱值但使產率下降。本研究選定溫度302 ˚C、120分鐘滯留時間及水過量比5作為製程與熱整合設計之示範操作條件，對應約41 wt%的生物油產率與約30.2 MJ/kg的高熱值。而熱整合設計顯著降低了外部加熱器的能耗，與未整合之製程相比減少高達95 %的能耗。

Lignin is a rich but underutilized resource in the field of biomass energy, and its conversion is crucial for the development of renewable energy. Hydrothermal liquefaction (HTL) process converts lignin into high-energy bio-oil under subcritical water conditions and is considered a highly promising method for lignin valorization. However, our understanding of the kinetics mechanism and product formation behavior remains insufficient. This study aims to address these challenges by developing a rigorous kinetic model for lignin HTL and establishing a complete process simulation to enhance the predictive capability of reaction behavior and optimize bio-oil production efficiency. In terms of research methodology, this research initially refined existing lumped kinetic models by proposing different reaction pathways and kinetic frameworks to more accurately describe the decomposition rates of lignin into oil, water, solid, and gas phase products in hydrothermal environment. Subsequently, this research established a rigorous kinetic model, introducing various model compounds to describe the equilibrium and conversion among specific products in each phase. After completing the model, it was integrated into Aspen Plus software to simulate the continuous lignin HTL process, design its separation and thermal integration units, and optimize its operating conditions. The results showed that the enhanced lumped model was more accurate in predicting the yields of lignin's various phase products compared to existing literature models, achieving a mean absolute error of 2.03 wt%. The rigorous kinetic model effectively predicted the yields and product distribution across each phase under different reaction conditions, with a mean absolute error of 4.51%. Regarding process simulation and optimization, the product distribution data output from the rigorous model was used for Aspen Plus process simulation, employing multi-objective optimization. The results indicated a significant antagonistic relationship between the yield and heating value of bio-oil. Under optimal conditions (temperature 270 – 340 ˚C, residence time 120 minutes, water-to-biomass ratio 5), an increase in temperature enhanced the heating value of the oil while concurrently reducing the yield. This study selected a temperature of 302 ˚C, a residence time of 120 minutes, and a water-to-biomass ratio of 5 as the operating conditions for process and thermal integration design, yielding approximately 41 wt% bio-oil with heating value around 30.2 MJ/kg. The thermal integration design significantly reduced the energy consumption of external heaters, achieving up to a 95% reduction in energy consumption compared to the non-integrated process.