(76a) Efficient Modeling of Feedstock Variability in Biomass Fractionation

Feedstock variability remains a major barrier to consistent performance and profitability in lignin-first biorefineries. Real-world scenarios–such as the surge in paper product demand during the COVID-19 pandemic–highlight how supply chain uncertainties exacerbate challenges in processing structurally diverse biomass species [1-3]. Although numerous experimental efforts have advanced lignin extraction and conversion, they are often insufficient for systematically addressing the structural and compositional heterogeneity introduced by variable feedstocks.

To address this, appropriate modeling approaches become essential. While kinetic Monte Carlo (kMC) methods offer detailed multiscale insights into lignin structure evolution, their computational intensity limits practical deployment [4, 5]. In contrast, we introduce a population balance model (PBM) designed for both computational efficiency and speed, enabling fast adaptation to diverse feedstocks. The proposed PBM captures critical lignin features–molecular weight distribution, monomeric compositions, and bond configurations–using a set of differential equations. It also handles multimodal behaviors commonly arising from mixed biomass inputs. This efficiency makes PBM highly suitable for tasks such as parameter estimation and real-time integration, which are otherwise infeasible with heavier models.

This capability equips biorefineries with a flexible and predictive tool to mitigate the risks associated with feedstock variability, opening the door to the model-based control strategies that enable to control lignin structural properties dynamically during the fractionation process. By linking fast simulation with process feedback, the approach ultimately contributes to increased robustness and profitability in next-generation lignin-first biorefineries.

Literature cited:
[1] Luo Z., Qian Q., Sun H., Wei Q., Zhou J., & Wang K., (2023) Lignin-first biorefinery for converting lignocellulosic biomass into fuels and chemicals. Energies, 16, 125.

[2] Jung J., Choi H.-K., Son S.H., Kwon J.S.-I., & Lee J.H., (2022) Multiscale modeling of fiber deformation: Application to a batch pulp digester for model predictive control of fiber strength. Comput. Chem. Eng., 158, 107640.

[3] Kumari S., Jeble S., Venkatesh V.G., Nagarajan C., & Shi Y., (2023). Antecedents of agriculture supply chain performance during COVID-19: An emerging economy perspective. Oper. Manag. Res., 16, 489-510.

[4] Lee C.H., Kim J., Ryu J., Won W., Yoo C.G., & Kwon J.S.-I., (2024). Lignin structure dynamics: Advanced real-time molecular sensing strategies. Chem. Eng. J., 487, 150680.

[5] Kim J., Ryu J., Yang Q., Yoo C.G., & Kwon J.S.-I., (2024). Real-time model predictive control of lignin properties using an accelerated kMC framework with artificial neural networks. Ind. Eng. Chem. Res., 63, 20978-20988.