CFD simulations are widely used to quantify the mixing performance of stirred tanks for various applications in chemical engineering and biotechnology. Due to advances in GPU computing, these simulations increasingly employ Large Eddy Simulation (LES), which explicitly resolves the dynamics of large-scale turbulence. Although such simulations are fully deterministic and therefore theoretically reproducible, small numerical variations induced by round-off errors, floating-point arithmetic, and differences in the distribution and ordering of operations in parallel computing lead to separation of trajectories i.e., different flow-field evolutions and consequently to significant run-to-run variability in predicted mixing times, even on the same hardware architecture. This work investigates the impact of repeated simulations, in the form of a case study, on the mixing-time distribution observed in a (Formula presented) stirred tank reactor using two commercial CFD packages operating with representative, production-level solver configurations. The analysis does not aim to assess the general performance of numerical method classes, but rather to quantify run-to-run variability under fixed solver settings and to compare the resulting numerical distributions to experimental variability. The results demonstrate that numerical variability is of comparable magnitude to the experimental spread, highlighting the necessity to treat LES-derived metrics as statistical ensembles rather than deterministic values. It is concluded that the reporting of confidence intervals is essential for methodological rigour in LES-based mixing studies.

Anticipating the occurrence and effects of mass transport limitations during fermentation scale-up is essential for commercialization, as heterogeneities might affect microorganisms. Tools like Computational Fluid Dynamics (CFD) aid this analysis but are computationally intensive, limiting design space exploration and consequently, fermentation optimization. Compartment models (CMs) based on CFD simulations offer an affordable alternative but require CFD recalibration with changing geometries or operating conditions, restricting their usage in optimization.
In this work, we introduce a hybrid machine-learning-aided compartment model (ML-CM) that accounts for flow pattern dynamics upon changes in both volume and stirring speed in a stirred tank bioreactor. The ML-aided dynamic compartment model (dyn-CM) enabled the spatiotemporal study of a process in 1/500th of the fermentation simulation time, maintaining reasonable accuracy. This method facilitates fed-batch fermentation modeling, process optimization, and scale-up effect analysis with modest computational resources, supporting reactor design and operational improvements within a defined operating space.

The oil palm industry has been under public scrutiny during the last decades due to environmental and social issues related to its practices. Oil palm (Elaeis guineensis Jacq.) trunks (OPTs) are of special interest as they are left idle in the field after the replanting process which is performed every 25 years. This common practice results in harvesting challenges, phytosanitary risks, and a loss of bioenergy potential. Due to their high moisture content and fibrous nature, OPTs present a problem for traditional conversion processes that require a dry and homogeneous material. This study evaluates the feasibility of converting OPTs into a bio-crude oil and biochar to increase the sustainability of the oil palm sector. To date, research efforts have primarily focused on hydrothermal liquefaction (HTL) of OPT without catalysts, resulting in a limited understanding of the potential of OPTs. Thus, the main novelty of this work is the evaluation of the effects of catalyst dosage (0–5 wt%) on the bio-oil yield, reaction temperature (260–300^∘C), and residence time (15–60 min) using a half-fraction experimental design methodology. For this, OPTs extracted from two plantations in Guatemala were used. The maximum bio-oil yield (26.77 ± 3.60 wt%) was found at 260^∘C for 15 min and 5 wt% catalyst with a high heating value (HHV) of 19.29 ± 1.33 MJ kg⁻¹. Nonetheless, the bio-oils produced without a catalyst at 300^∘C and 15 min have higher HHV (27.63 ± 1.35 MJ kg⁻¹) and are similar to Diesel fuel based on their H/C and O/C ratio. These results indicate that there is a potential trade-off between the bio-crude oil mass yield and HHV when using the catalyst.