Library Subscription: Guest
Home Archives Officers Future meetings American Society of Thermal and Fluids Engineering
4th Thermal and Fluids Engineering Conference

ISSN: 2379-1748


Kevin J. Farrell
Heat Transfer Research, Inc., PO Box 1390, Navasota, Texas, USA 77868-1390

DOI: 10.1615/TFEC2019.cmd.027542
pages 287-296

KEY WORDS: kettle, reboiler, entrainment, droplet distribution, Eulerian-Lagrangian, design rules


Accurate prediction of kettle entrainment is crucial to sizing kettles effectively, reducing equipment cost, and increasing column efficiency. Many geometry and process parameters affect the prediction of liquid carryover:
· vapor velocity distribution
· droplet size and distribution
· volume and shape of disengagement space
· number and location of nozzles
· physical properties
Simple design sizing rules have been developed based on operating experience. However, with an Eulerian-Lagrangian computational approach, we can compute the amount of entrainment directly from first principles. First, we calculate the turbulent vapor flow field above the bundle using the vapor generation profile from the design. Then, an assumed distribution of droplets is released from the liquid/vapor interface near the top of the bundle. The paths of the droplets are tracked with an equation of motion which includes all relevant forces. Our initial application assumed no slip between the liquid and the vapor at the interface boundary and liquid mass flow rates calculated via an entrainment correlation for a submerged vapor jet. Our initial application of the method to the entrainment study of Kister and Chaves showed a surprisingly good comparison to the measured entrainment in the dropout drum, using a Rosin-Rammler droplet distribution with a 50-µm reference diameter and an exponent of 2. The computational approach seems well suited to parametric entrainment studies of kettles and nozzles of various numbers, sizes, and locations for deriving design guidance. Additional measurements of entrainment and droplet sizes and distributions will help fine tune the approach for absolute predictions.

Purchase $20.00 Check subscription Publication Ethics and Malpractice Recommend to my Librarian Bookmark this Page