Product/Service

Energy Optimization of Spray Drying Plant

Aerotak

About Author

Fluid Mechanics and Computational Fluid Dynamics (CFD) simulation specialists

See partner

This case study shows how Aerotak™ is able to simulate and map the flow field and particle trajectories within a commercial spray drying pilot and from this optimize the energy consumption.

Objective

Accurate prediction of particle paths, drying behavior, and estimation of product deposition on chamber walls. The overall purpose is to apply the CFD simulation as a tool for optimizing operating conditions and design to improve performance and product characteristics of a spray drying plant.

Simulation setup for a spray drying plant

The fluid domain is restricted to the spray drying chamber itself, consisting of a cylindrical part and a conical part transitioning to a bent outlet pipe. Downstream cyclone for product separation is not included in the fluid domain as well as the gas disperser located on top of the ceiling. The inlet annulus (marked in red) is provided with a set of velocity components for producing the drying air mass flow rate of 350 kg/h and the swirling conditions provided by the gas disperser. The inlet face is imposed with an absolute humidity and temperature corresponding to that of the operating condition. The outlet face (marked in blue) is imposed with a zero gauge pressure. All walls are imposed with a no-slip condition and a wall-to-ambient heat transfer coefficient to simulate the loss of energy due to convective heat transfer to the surroundings.

A k-omega SST turbulence model is adopted for turbulent flow structures and mixing between hot and cold regions of the chamber. Using a steady-state flow assumption on a volume mesh of 2 million cells, a fully developed flow solution is obtained prior to injection of computational parcels, representing droplets being atomized by the centrifugal atomizer. The atomization process is simplified by injecting particles composed of a liquid (evaporative) and a solid (non-evaporative) component. Parcels are imposed with an empirically estimated radial velocity component and a tangential velocity component equal to the peripherical speed of the rotary atomizer. Atomizer rotation is accounted for by applying a moving wall condition to the surface of the blunt atomizer geometry.

Spray drying simulation results

Converged flow and particle field results are obtained after 24 hours of simulation time. Flow and temperature fields are presented in the figures below. A vertical plane through the geometric center of the chamber and six horizontal planes are used for vizualization. Variation in the temperature field influences the particle drying behavior, while the flow field is crucial for prediction of particle trajectories.

https://images.squarespace-cdn.com/content/v1/59eefd1a29f1875b3b70a35b/1550568723952-6WGG4BM57XEGEPXTGVRP/ke17ZwdGBToddI8pDm48kM21R5Mq7WEWb3dW4XIe25x7gQa3H78H3Y0txjaiv_0fDoOvxcdMmMKkDsyUqMSsMWxHk725yiiHCCLfrh8O1z5QPOohDIaIeljMHgDF5CVlOqpeNLcJ80NK65_fV7S1UWOmW_SxXhMah5Cd5-paehDkrQW_cPMuBfgvKiI8j0TGpC969RuPXvt2ZwyzUXQf7Q/spray+dryer+lagrangian+particle+simulation+geometry+field+transient+response+computational+fluid+dynamics+CFD+velocity+field?format=750w

https://images.squarespace-cdn.com/content/v1/59eefd1a29f1875b3b70a35b/1550568697229-LX22K2YNTVLPMUT6L3I2/ke17ZwdGBToddI8pDm48kM21R5Mq7WEWb3dW4XIe25x7gQa3H78H3Y0txjaiv_0fDoOvxcdMmMKkDsyUqMSsMWxHk725yiiHCCLfrh8O1z5QPOohDIaIeljMHgDF5CVlOqpeNLcJ80NK65_fV7S1UWOmW_SxXhMah5Cd5-paehDkrQW_cPMuBfgvKiI8j0TGpC969RuPXvt2ZwyzUXQf7Q/spray+dryer+lagrangian+particle+simulation+geometry+field+transient+response+computational+fluid+dynamics+CFD+temperature+field?format=750w

Instantaneous particle trajectories of individual parcels with residence time up to 0.1 seconds are plotted on the left hand side of the figure below.

Figure 4

Averaged particle statistics can be gathered to provide general information on trajectories, drying curves, residence times, mass flow rates etc. The right hand side the figure above indicates averaged location and concentration of particles escaping the domain, with percentages expressing the particle mass weighted average residual moisture, which are directly related to wall deposition and its extent. For the current chamber configuration and operating condition, large particle impact rates on the cylindrical part of the cylinder are observed at high residual moistures, implying that wall deposition is likely to be most extensive in the upper part of the chamber. Adjusting the gas disperser to compensate for this could decrease the wall deposits and hence a potential plant capacity increase could be gained.

Did you like this solution?

Aerotak
  • 1 Solutions
Primary contact
Kasper Berthu Damkjær
+4527121013

Location

Looking for our logo?

Find it in the State of Green Toolbox together with loads of materials to help you promote Denmark’s green solutions abroad.
Visit toolbox