Towards a multizone vortex dryer for dairy sprays
Umair Jamil Ur Rahman is a PhD student in the research group Energy Technology. Supervisor is prof.dr.ir. G. Brem and co-supervisor is dr.ir. A.K. Pozarlik from the Faculty of Engineering Technology.
Conventional co-current spray dryers are widely used in the dairy industry to produce milk powder. These dryers are known for their high capital costs and low thermal efficiencies and are responsible for almost 27-55% of the total energy consumption of the dairy industries. In comparison to co-current dryers, counter-current dryers have higher thermal efficiencies but they are not employed in the food industry due to the risks of product degradation.
In order to develop an alternative commercially viable and process-intensified spray-drying technology, high drying rates in a small volume must be achieved and the residence time of droplets/particles must be reduced to maintain product quality. This can be done by operating the dryer as a multizone vortex chamber unit wherein high air temperatures and high-G acceleration are employed. In this configuration, hot air enters axially into the central zone (counter-current to the droplets flow) while a cold vortex flow is created via relatively cold airflow entering the dryer through the tangential channels. The process intensification is established as an outcome of both: multizone drying operations with high and low temperature air feeding zones and high-G acceleration. Hence establishing two drying zones: a fast-drying zone in the radial center and a final slow drying zone in the periphery of the chamber. The former leads to enhanced drying rates, and the latter intensifies air–droplet contact and air–solid separation. The key features of this technology can be utilized for the integration of different unit operations, bigger feed capacities, in situ separation of particles, and improved product qualities.
The objective of this thesis is to investigate the phenomena interacting in the counter-current multizone vortex spray dryer and identify the most relevant parameters for an optimal spray drying operation. This is done by performing experimental and numerical research.
Experimental research
To get insight into the counter-current spray drying process and the related phenomena, droplet size measurements were carried out using the Particle Droplet Image Analysis technique. Then, based on the data obtained for the droplet size, velocity, and distribution, a lab-scale counterflow dryer was designed and studied for water and milk feeds.
The results showed a strong dependency and interaction between the atomization process and drying gas flow. The water experiments confirmed that to obtain process intensified spray drying with maximum evaporation efficiency, a high air inlet temperature together with fine atomization is required. For instance, applying an air inlet temperature of 360 °C combined with a spray SMD of 30-50 µm resulted in a stable spray drying operation with a complete evaporation of a 20 kg/h of feed.
The drying tests with milk feed showed an analogy to the water tests. The results showed that smaller particles, due to their low glass transition temperature, were more prone to deposition and sticking to the walls. Moreover, two deposition zones were identified: a top zone where dried particles were found sticking on the walls due to the high air temperatures and a bottom zone where semi-dried deposits were observed due to higher particle moisture content and low air temperatures. The recovered particles exhibited typical wrinkled morphology attributed to the low solids content and small droplet sizes.
CFD investigations
Multiphase CFD simulations were performed to gain insight into the counter-current spray drying process. The developed CFD model was initially applied to simulate the spray drying behavior wherein hot air and liquid feed are injected into the counterflow dryer, hence, investigating only the hot zone of the setup. The model predictions were in close agreement with the experimental temperature measurements with a maximum average error of 5%. Then, comparative simulations were conducted to evaluate the dryer performance by simulating the influence of the nozzle position in the dryer, the initial spray SMD, the inlet air temperature, and the feed rate. It was found that placing the atomizer too close to the gas outlet increases the product deposition on the bottom wall, while its placement too close to the hot air inlet results in droplets/particles entering the hot air inlet section. Moreover, increasing the spray momentum by increasing the feed rate resulted in a minimum product deposition and an increased mean particle diameter and residence time. The evaporation profiles in the dryer, i.e., temperature and moisture fields, were found to be highly dependent on the SMD of the spray parameters. Increasing the air inlet temperature intensified the evaporation rates only in the initial 10–20 cm of the atomizer, where the droplet concentration was highest. Further downstream of the atomizer (> 20 cm), the droplets/particles moved radially outwards in the mixing zone (recirculation zones), where evaporation rates were similar for all investigated cases.
The model was further extended to simulate the spray drying process in a multizone vortex chamber. The parameters controlling the process were identified by conducting comparative simulations. The results showed that the position of the solids outlet has a strong effect on the asymmetries generated in the flow as well as on the particle losses via the gas outlet. Furthermore, the simulations showed a clear distinction between the high and low-temperature zones in the dryer, with significant evaporation occurring only in the central fast-drying zone of the dryer. Finally, it was concluded that in order to obtain a product separation efficiency of 85 wt.% or higher, an average centrifugal acceleration of 250g is required in combination with a narrow initial droplet size distribution of 45-85 µm. The in situ particle separation and segregation of small and large-sized particles leading to distinct drying histories were also revealed. The majority of the bigger particles (>32 µm) left via the solids outlet with a mean residence time of 0.67 seconds, whereas the bulk flow of the smaller particles (<32 µm) left the reactor via the gas outlet with a mean residence time of 0.04 seconds.
The CFD modeling approach and knowledge gained from the lab-scale vortex chamber study were applied to simulate spray drying in the pilot-scale dryer. The results of the large-scale reactor simulation showed that most of the evaporation occurs close to the atomizer in the hot central region, while the air temperatures and moisture values are uniform in the outer radial region of the dryer and other parts of the chamber. The data also revealed a strong momentum transfer between the air and particles with the average G-acceleration in the main drying chamber equal to 70g, while this was reduced to 30g in the presence of particles. Furthermore, it was shown that most of the small-sized particles (< 25 µm) left the reactor with a mean residence time of 0.5 seconds, whereas the majority of the bigger-sized particles (> 34 µm) exited via the bottom solids outlet. These particles had a mean residence time of 2 seconds.
The study confirmed that by employing vortex chamber technology, process intensification of the spray drying process is indeed possible. The high-G multizone drying operation with high and low-temperature zones can result in enhanced drying rates and small particle residence times. Furthermore, the in situ separation and segregation of different sized particles led to distinct drying histories and narrow residence time distributions, therefore, ensuring optimum product quality.
The data and knowledge obtained in this research provide a framework for further development of counter-current dryers for dairy sprays; the findings reveal that the two-step spray drying process in a vortex chamber configuration can definitely be done efficiently in a small reactor volume and without product degradation, but requires special attention to the initial droplet size and atomization conditions as well on the separation efficiency of the product. Further research is recommended to investigate the quality of the product obtained, to validate the proposed models in more detail, to improve the CFD model prediction on the wall deposition effects, and to improve the understanding of the two-step spray drying process on the pilot scale.
