Outskirts Press, INC.

Vapor Liquid Separation

Author Contact :



Ronald J. Robichaux



1522 Savannah Drive, Garland, Texas 75041



E-mail: rjrcajun339@aol.com



My experience with separators led me to compile design information for the precise design of this type of equipment. These devices seem simple in their operations. However I have seen a number of them fail to provide the performance required in the process. Working with consultants, their many suggestions and rules of thumb to resolve the issues, has left me with concern as to how to best approach a design for a proper operating separator.



I studied the works of many Engineers and designers and collected information from all the manufacturers I worked with in designing separators for my projects. I have read and collected many articles on the subject of separator design.



In 1984 I had the opportunity to accept a position with Perry Equipment Corporation in Mineral Wells, Texas. The range of designs, research efforts and field trouble shooting left me with the understanding that "rules of thumb" are useful in a limited set of circumstances. Fortunately for the industry these set of circumstances are sufficient at about 75 to 90 percent of the time.



One of the tasks I was given, while in their employ, was to attempt to place a scientific approach as to why these devices proved so successful in most situations, but failed in other situations. In order to develop a better understanding of the design parameters I wanted to know the physical properties of the fluids and their effect associated with separator failures and efficiencies.

Chapter 1

Introduction



Over the years the author has sized many separators, for new installations, and performed trouble-shooting on failed situations. Most of the sizing was based on standard industry “K” or “C factor based formulas. Most were successful, 95% of the time, and a few were unsuccessful or hard to analyze. The source of the constants in the settling of particle sizing formula was hard to determine in some applications.



After the four-year stint with Perry Equipment Corporation, one of the leading filter separator manufacturers, the author has a better insight as to how the constants were determined. This interval provided experience with laser particle counters, which provided valuable information to the author.



As stated by industry-recognized specialist, such as Norman P. Lieberman, the principle formula for equipment sizing is



Va = C x ((rl - rv) /rv)^1/2. Eq. (26)

Or Va = K x ((rl - rv) /rl)^1/2



Va is the allowable settling velocity for a given size particle.

rl being the density of the liquid forming the droplet at operating conditions

rv is the density of the vapor providing the buoyancy and resistance to settling at operating conditions.

C is the settling constant for a given sized particle being considered and is given in ft/hr.

K is the settling constant for given sized particle being considered and is given in ft/sec



This is a relatively simple equation and is very successful in 95% of the sizing cases, as long as the proper “C or “K” factor is available. Herein lies the problem for the Engineer, which “C” or ”K” factor to use and how was it determined. Another problem is determining the particle size to be separated, and it’s relative “C” or “K” factor. Most published “C” and ”K” factors are based on 150-micron droplet sizes. In later chapters we will see that much smaller droplet sizes may dominate the droplet distribution.



In his book Norman P. Lieberman provides the following “K” values with characterization statements;





Without Demister With Demister



Minimum Entrainment 0.15 0.30

Moderate-pressure service

Minimum entrainment 0.10 0.20

High (100 + psi) H2 service

Moderate entrainment 0.25 0.45

Severe Entrainment 0.50 ___

De-entrainment of steam ___ 0.20



In another of his books Norman P. Lieberman states;



“Thousands of experiments have been run in the plant and laboratory to determine K values for different levels of entrainment. Here are the results:”



· K = < 0.15: very slight entrainment

· K = 0.23: normal to low entrainment

· K = 0.35: high entrainment

· K = > 0.50: very severe entrainment



Which he follows with the profound statement “I feel sure, dear reader, that you are objecting to my use of such terms as “slight” and “severe,” as these are qualitative Terms. Well, we do not know enough about entrainment to quantify it.”



He goes on further to say “All I am saying, is don’t take Eq. (26) or the tabulated K values as the ultimate truth.”



The Lieberman’s have produced very informative and useful books, but the statement above leaves one to wonder as to when a value of “K” or “C “, depending on the author’s interpretation of nomenclature, should be applied. They state precisely that the engineer needs to take more into consideration in designing a separator.



The intent of this book is to divulge more of the data and calculation methods to be used in defining droplet sizes, entrainment and separation quantities and to expose the reasons for the failure of a minority of separator designs.



We will attempt to do justice to the Lieberman’s statement. See Chapter 26 for a comparison of calculated versus proposed “K” and “C “ Values.