Container design optimization for efficient industrial freezing process of coffee extracts

Freezing is essential not only for ensuring food safety and extending shelf life but also for preserving the intrinsic aroma and flavor profiles. In this study, we developed and validated a phenomenological model to simulate the industrial-scale freezing process of coffee extract. The model was implemented using the finite element method in SfePy, incorporating temperature-dependent thermophysical properties and phase change behavior. Experimental validation was conducted using temperature data collected from a 25 kg coffee extract in a container frozen at –25 °C. The simulated temperature profiles closely matched the experimental results, with a mean squared error (MSE) of 1.32 °C${}^2$. Subsequently, the validated model was integrated with an optimization routine to determine the optimal container dimensions that minimize freezing time. The results indicate that increasing the container diameter and reducing its height significantly shorten the freezing time. For instance, a container with a diameter of 0.8 m and a height of 0.4 m demonstrated a substantial reduction in freezing time compared to the current design. Together with the optimization program, this phenomenological model serves as a cost-effective virtual tool to support the design of more efficient freezing processes in the coffee industry, enabling engineers to test container modifications and process conditions without incurring the costs of physical trials.