CFD modeling of Biomass Hydrolysis Reactor

The intensification of climate change has stimulated the research for sustainable energy sources, among which second generation (2G) bioethanol has emerged as a promising alternative. This biofuel can be obtained through the fermentation of simple sugars derived from lignocellulosic biomass. A critical step in the biomass hydrolysis process is the liquefaction, where pretreated biomass is converted into glucose through the action of enzymes. To ensure economic feasibility, this step must be conducted at high solids loading to minimize water usage and reduce the energy demand for distillation of ethanol in the last step of the process. However, under such conditions the biomass slurry exhibits non-Newtonian behaviour, leading to significant challenges in the mixing process and in the description of system evolution during enzymatic hydrolysis. To overcome these challenges, this work developed the computational fluid dynamics (CFD) model of the reactors used to conduct biomass liquefaction in the experimental tests. The study began with a literature review aimed to have a better understand of the main difficulties associated with modelling of biomass hydrolysis reactors and the mixing of non-Newtonian fluids. A reference study was then selected to build a validated model: the geometry of the reactor described in the paper was reconstructed, and the torque values on the impeller at different rotational speeds were compared with the reported experimental data. This validation procedure confirmed the reliability of the model for both Newtonian and non-Newtonian fluids and enabled the estimation of some important hydrodynamic parameters such as the power input, the average shear rate, the elongation and the fraction of dead zones during the agitation. After this section, CFD models of the experimental reactors were developed, including the IKA systems equipped with Paravisc and Anchor impellers, as well as the TORNADO reactor. From these simulations, it was possible to construct the Power Number-Reynolds curve (Np-Re) in the laminar regime, different for each geometry. In the end, the several hydrodynamic parameters that may have an impact on the biomass conversion were extracted. The subsequent analysis revealed that all the hydrodynamic parameters investigated may have an influence on the glucose conversion, particularly during the first two hours of the liquefaction. Beyond this initial phase, their impact on the conversion rate diminished significantly, suggesting that other factors become more important. Nonetheless, it was not possible to determine the relative contribution of each hydrodynamic parameter using a single system and one set of experimental data. To achieve this, further simulations and experimental studies under different conditions are required, particularly with the Anchor IKA and TORNADO reactors, which will provide complementary information about the relation between glucose conversion and the reactors hydrodynamic. In this work the CFD model of three different reactors has been created and validated leading to the possibility to build the Power Number-Reynolds curve for different geometry and extract important hydrodynamic parameters for different operating conditions. This demonstrates the value of CFD modeling as a tool for analysing the liquefaction of lignocellulosic biomass at high solids loading, giving the possibility to calculate parameters that are impossible to measure in laboratory.