To: "'firstname.lastname@example.org'" <email@example.com>
Subject: Thermosyphon Reboilers
From: E. B. P (U. S. Consultant)
I have had discussions with several manufacturers regarding the amount of incoming feed a thermosyphon reboiler can vaporize. Based on my discussions, it appears that the upper limit is somewhere around 30-40%. What I have not been able to obtain is an adequate explanation as to why these reboilers can't vaporize a larger percent of the incoming feed. Do you have any information that would explain to me why they can't adequately vaporize higher percentages of the incoming feed?
Subject: Thermosyphon Reboilers: Vaporization Possible
The explanation takes some description but the vendors are correct. 30-40% is the typical limit. For any given exchanger the limit depends upon the construction details and the system involved. For some installations, I have observed limits as low as 20-25%. Others have achieve levels as high as 45-50% in special circumstances.
The mechanism behind this limit is similar to that in critical heat flux. Critical heat flux has been much more studied than the vaporization blanketing problem. Critical heat flux is the point that when the temperature driving force (LMTD) for heat-transfer increases that the overall heat transfer flux decreases. This seems backward from our common conception of heat-transfer. Normally, we define the duty in the familiar equation for heat-transfer as:
In standard practice we think of U, the overall heat-transfer coefficient, as a gradually and slowly changing variable with system properties. Therefore, when we increase the temperature difference and hold the area constant (A fixed), Q must increase. Practical work with heat-exchangers show that this is not always true. In some exchangers when the LMTD has been increased the duty transferred decreases. This is a common problem in units where a large increase in heat-transfer capability has been predicted based on a change in heating medium to one much hotter than previously used. Several units I have investigated owed the failure to meet their objectives to this problem.
Either the U or the effective A or both must have changed at the same time the LMTD changed. Most work on this problem expresses the difficulties in terms of a critical heat flux where:
Units for heat flux, q, are typically btu/hr-ft2 or kW/m2. Effective area is assumed as fixed.
The exact system behavior that creates critical heat flux limits differs between natural convection and forced convection surfaces. However, the three major predictive methods for critical heat flux are based upon (1) boundary layer separation, (2) bubble crowding and, (3) macrolayer evaporation. All three of these approaches have been used to predict critical heat flux. In fact, the physical action behind each of these theories probably occur all mixed together in different proportions: depending upon the system and equipment design. The linkage with vaporization limits for forced convection arises when we examine bubble crowding and boundary layer separation.
In both bubble crowding and the macrolayer evaporation approach to understanding the vaporization limits, the generation of bubbles at the heating surface prevents liquid from reaching the surface. The liquid cannot reach the heated surface and the gas (with a low thermal conductivity) does not effectively conduct heat to the liquid far away from the heated surface. In fact, the expanding vapor bubble pushes the liquid away from the surface. The main engineering difference between the models is the form of the correlation used to predict the critical heat flux.
The same procedure of bubble formation at the heated surface preventing liquid from reaching the exchanger tubes creates the vaporization limit in thermosyphon (and other forced-convection) exchangers. Once a certain volume of the fluid is made up of vapor, the vapor insulates the tube surface and prevents liquid from reaching it. Two general types of limits can be reached.
First option - operation with a wide boiling range mixture:
The residual liquid at the tube surface (that has not yet evaporated) becomes heavier when the bulk liquid cannot replenish it with light components. It's boiling point continues to rise as the lighter material volatizes. As vaporization from the film drops, the heat-transfer coefficient drops (the film is entering sensible heat-transfer). Net vaporization of the bulk stream stops because heat can no longer be effectively moved into the process.
Second option - operation with a narrow boiling range mixture:
Vaporization continues without liquid percolating, in sufficient quantity, to the heating surface. The surface eventually dries off (burnout). This shifts a significant part of the surface area in the exchanger from a vaporization service to a vapor superheating service. The vapor superheating service has a low heat-transfer coefficient. Net heat input drops. Overall vaporization of the bulk stream stops because heat can no longer be effectively moved into the process.
In practice, the situation is often much murkier. Equipment design, heating medium temperature profile, flow rates, system properties, and other factors all influence the operation. This creates the wide variation in exchanger performance in even apparently very similar systems.
An excellent detailed description of the factors involved can be found in Hewitt.
 Hewitt, G. F. Boiling: in Handbook of heat transfer, third edition.: editors: Rohsenow, W. M.: Hartnett, J. P.: and Cho, Y. I. McGraw-Hill Companies, Inc., New York, 1998. 15-123 to 15-137.