Liquid nitrogen enabled direct freeze concentration
Zhuo Zhang is a PhD student in the Department of Energy, Materials and Systems. Promotors are prof.dr.ir. S. Vanapalli and prof.dr.ir. H.J.M. ter Brake from the Faculty of Science & Technology.
This thesis investigates the feasibility of using liquid nitrogen as a medium for performing direct-contact freeze concentration across various scales: micro, meso, and macro scales to understand the underlying mechanisms and practical applications of liquid nitrogen in the freeze concentration processes.
From a micro-scale perspective, the research primarily focused on the evaporation dynamics and heat transfer mechanisms associated with a liquid nitrogen droplet within an immiscible liquid pool. The aim was to thoroughly understand the evaporation processes to assess the performance of using liquid nitrogen in freeze concentration applications. In Chapter 2, we introduced a 2D quasi-steady-state theoretical model and examined the behavior of liquid nitrogen droplet heat transfer dynamics in detail. The model introduces scaling laws that describe the rate at which the liquid nitrogen droplet shrinks and the surrounding bubble grows. These scaling laws are crucial for understanding how the droplet evolves and how this evolution affects the efficiency of the cooling process. Accurate prediction of these physical quantities allows for better control of the cooling process, which is essential for optimizing the freezing of the surrounding liquid. Furthermore, the model also predicts the heat transfer rate between the droplet and the surrounding liquid, a key factor in determining the effectiveness of freeze concentration. Initially, the model aligns well with the experimental data from open literature, particularly during the early stages of evaporation where heat transfer is dominated by conduction. This agreement validates the model’s accuracy in representing the initial phases of the droplet’s evaporation process.
The results from Chapter 2 identify that as the evaporation process progresses, convection might begin to play a more significant role, particularly in the later stages. This shift leads to deviations from the model’s predictions, suggesting that the heat transfer mechanism is not solely conduction-dominated at the later evaporation stage. The findings indicate the necessity for further investigation into this transition. To address the limitations observed in the 2D model at later stages of evaporation, Chapter 3 further investigates the heat transfer mechanism between an eccentric sphere and its spherical enclosure, as an analogy to the liquid nitrogen droplet and the bubble surrounding it.
The research in Chapter 3 experimentally examines the impact of Rayleigh numbers and eccentricities on the heat transfer mechanism. By comparing cases with varying Rayleigh numbers, the finding shows that at low Rayleigh numbers, heat transfer remains conduction-dominated, confirming the validity of the assumptions made in the initial droplet evaporation model. The experimental results are also in line with a quasi-steady-state conduction-based model. However, even at a moderate Rayleigh number, a stronger influence of buoyancy-driven convection is observed, and the need to account for convective heat transfer becomes evident. The introduction of a refined Nusselt number law that covers a wider range of Rayleigh numbers and eccentricities could provide more accurate predictions across the entire evaporation process. This adjustment is crucial for developing more reliable models that can be used to optimize the freeze concentration process using liquid nitrogen, ensuring that the liquid nitrogen’s cooling potential is fully utilized.
During continuous cooling from the liquid nitrogen droplet, the liquid pool would eventually solidify which significantly alters the situation compared with which the pool remains liquid for the entire time. From a meso-scale perspective, an improved 1D modeling approach was developed to predict the dynamics of ice growth and solute redistribution at the propagating ice front. This research focuses on the characterization of the intrinsic partition coefficient (K0), which varies as a function of the characteristic velocity (v∗) of the freezing front. A significant achievement of this work is the identification of a Sigmoidal relationship that accurately describes the variation in the partition coefficient as the freezing front progresses, providing a more refined understanding of the solutes inclusion mechanism during the freeze concentration process. The study primarily utilized a sucrose-water system to develop and validate the model, demonstrating that the Sigmoidal function provides a robust approximation of the intrinsic partition coefficient across a range of characteristic velocities. However, the framework established in this research is not limited to this particular solute-solvent system; it can be extended to other systems by determining the appropriate parameters through experimental data. These parameters would depend on the specific mass transfer properties of the solute molecules and the cooling duty required during the freezing process.
In Chapter 5 and Chapter 6, we shift our focus to macro-scale level. In Chapter 5, a proof-of-concept laboratory direct-contact freezer utilizing the cold energy of liquid nitrogen by direct injection was developed, and experimental trials for various concentrations of sugar solution were investigated. It successfully demonstrated the formation of ice particles where the average size of the ice particles decreased with an increase in the concentration of the sugar solution. An ice growth rate of around 4 g/s with an initial 1 kg sucrose solution was observed in the trials, along with an average partition coefficient of approximately 0.6. This study demonstrated the potential of using liquid nitrogen as a ‘direct-contact’ coolant in freeze concentration. In addition, the challenges of using liquid nitrogen as a direct-contact coolant are identified and valuable insights are provided for improving the design and operation of the system.
Once the feasibility of utilizing liquid nitrogen as the direct-contact coolant is proved, a comprehensive and systematic study to quantify the operational parameters and efficiencies is necessary for practical application. Therefore, in Chapter 6, we proposed a conceptual design and developed a detailed material and energy model of a direct contact freeze concentration system using liquid nitrogen as the coolant. In the proposed operation system, the main components include the ice generator, liquid nitrogen storage unit, ice filter, ice melter, and heat exchanger. The influence of the mass flow rate of the liquid nitrogen, the inlet temperature, and the mass flow rate of the liquid solution into the ice generator are considered as the parametric input parameters in this study. Sample calculations with the basis of 2.6 kg/s liquid nitrogen mass flow rate can produce 4.72 kg/s of the concentrated liquid with the total mass flow rate of the liquid solution at 7 kg/s. The annual production of the ice and concentrated solution are 65.66 kton and 135.94 kton, respectively, while assuming the annual operation time is 8000 hours.