Simplify Flash Calculations

Eubank, Philip T; Mannan, M Sam; Hall, Kenneth R . Chemical Engineering ; New York  Vol. 107, Iss. 5, 
(May 2000): 125-126,128,130.

ProQuest document link

ABSTRACT
 
Flash calculations are necessary for a number of process applications, such as distillation, reservoir depletion, and
also for process simulations. Practical plant designs call for multiple iterative flash-calculations to the point where
the engineer with the assignment begs for mercy. Even though today's computers are fast, these repetitive
calculations require tremendous amounts of computer time. As a result, process control schemes also are
affected. Thus, simplifications and improvements in speed for this onerous task are desirable. This article presents
a a simplified procedure for making flash calculations for binary and ternary mixtures. It has the advantage of
being relatively simple, especially when compared to the common methods. The flash calculations for a binary
mixture are reduced to solving a first order equation, and a ternary mixture requires solving a quadratic equation.

FULL TEXT
 
Headnote
Solve simple first-order and quadratic equations to determine the fraction vaporized from binary and ternary
mixtures
(ProQuest: ... denotes formulae omitted.)
Flash calculations are necessary for a number of process applications, such as distillation, reservoir depletion, and
also for process simulations. Practical plant designs call for multiple iterative flash-calculations to the point where
the engineer with the assignment begs for mercy.
Even though today's computers are fast, these repetitive calculations require tremendous amounts of computer
time. As a result, process control schemes also are affected. Thus, simplifications and improvements in speed for
this onerous task are desirable.
This article presents a a simplified procedure for making flash calculations for binary and ternary mixtures. It has
the advantage of being relatively simple, especially when compared to the common methods. The flash
calculations for a binary mixture are reduced to solving a first order equation, and a ternary mixture requires
solving a quadratic equation.
...
Derive equations
The derivations start with a material balance and then introduce a series of objective functions. A few examples
will verify the equations.
Equations (1), (2) and (3) show the material balance constraints:
... (1)
... (2)
... (3)
where z^sub i^, y^sub i^ and x^sub i^ are respectively the mole fractions of the ith component in the feed, vapor and
liquid, Ki is the equilibrium ratio of the ith component, V is the fraction (mole basis) in the vapor phase, L is the
fraction in the liquid phase, and L denotes the total number of components. The overall composition, z^sub i^, is

PDF GENERATED BY SEARCH.PROQUEST.COM Page 1 of 5

always known in practice.
Smith et al. [5] have devised a generalized approach to a flash calculation in terms of dimensionless objective
functions. These are shown in the next equations.
... (4)
... (5)
Rachford and Rice [4] offer the criterion for equilibrium from the obvious equality of (F^sub x^ - F^sub y^ to zero:
... (6)
Smith's team took advantage of a property of the Rachford-Rice criterion. The derivative with respect to V is
always negative (Equation 7).
... (7)
This means that finding the root, F - 0 in Equation 6 is easy. If one uses Newton's method, for example, it
converges rapidly. Even better, only one root exists in the physical region of V/0, 1/.
The iteration by Newton's method uses Equation (8) for the steps, Just one caveat: it is important to start with an
initial value (l = 0) of Vbetween 0.0 and 1.0:
... (8)
Simplified approach - ternary
There is a way to avoid running iterations and derivatives for ternary and binary mixtures. It uses the simple
identity: Xi = (Kt-1).
For a ternary mixture, after making that substitution, Equation. (6) yields Equation (9):
... (9)
This can be algebraically rearranged into Equation (10):
...
... (10)
Equation. (10) has the general polynomial form:
... (11)
where the ßterms are the multipliers of the vapor fraction terms in Equation ( 10). Solve for V by the quadratic
formula:
... (12)
The betas can be either positive or negative, so evaluate Equation (12) with both signs. The correct one will yield a
V that lies between 0.0 and 1.0.
One can divide Equation (11) by ß2 and then redefine ßo and ß^ This reduces the uncertainty of having to choose
roots, but at the peril of having divided by zero should one of the K-V alues be unity. Experience shows that
selecting the correct sign in Equation. (12) is never a serious problem with hand calculation. In computer
programs, it is best to use first one sign, and then the other, and then select the value of V that lies between 0.0
and 1.0.
Simplified approach - binary
A binary mixture uses 23 = O in Equation (6) as a special version of a ternary mixture. Do some algebraic
manipulation of Equation (6), remembering that Equation (1) applies, and Equation (13) is the result:
... (13)
A K-value of unity is not a problem here because one K-value must be greater than unity and one less than unity to
be in the two-phase region. Equation. (13) has the advantage of being a direct solution for V, avoiding the need to
solve the quadratic equation that would result from using Equations (4) and (5).
Equation (13) makes it easy to determine if the system is in the two-phase region before applying the flash criteria.
By definition, V = O at any bubble point; therefore, (XjA2) = -(Z2Iz1). At the dew point, by definition, V=I, therefore
(X1A2) = - (Z2Iz1) - (X1 Iz1).
Let us do an example. Take an equimolar (50 mol%) mixture, and assume that ?; = 2.4; then (X1A2) is -1.0 at the
bubble point and -5.8 at the dew point. For this system, X2 will lie between -2.4 and -0.4138 to be in the two-phase

PDF GENERATED BY SEARCH.PROQUEST.COM Page 2 of 5

region. However, the definition of X imposes an additional restriction: X2 is bounded by (-1, °). That chops off -2.4
to -1. As a result, X2 must fall between -1 and -0.4138 to be outside the superheated vapor region.
Application to vapor liquid data
Using the methodology developed in this work, we have performed flash calculations for two binaries and a
ternary. The results appear in Tables 1, 2, and 3. Tables 1 and 2 show the application of Equation (13) when
calculating the vapor fraction for methane + n-heptane with vapor-liquid-equilibrium data from [3] and for nitrogen
+ n-decane data from [1] respectively. Table 3 applies Equation (12) to calculate the vapor fraction from nitrogen +
methane + n-decane vapor-liquid equilibrium data [2].
Edited by Peter M. Silverberg
References
References
1. Azarnoosh, A., and McKetta, J. J., Nitrogenn-Decane System in the Two-Phase Region, J. Chem. Eng. Data, pp.
494-^96, 8, 1963.
2. Azarnoosh, A., and McKetta, J. J., Vapor Liquid Equilibrium in the Methane-n-Decane-Nitrogen System, J. Chem.
Eng. Data, 8, pp. 513-519, 1963b.
3. Chang, H. L., others, Vapor-Liquid Equilibria of Light Hydrocarbons at Low Temperatures and High Pressures:
The Methane-nHeptane System, AIChE J., 12, pp. 1212-1216, 1966.
4. Rachford, H. H., Jr., and Rice, J. D., Procedure for Use of Electronic Digital Computers in Calculating Flash
Vaporization Hydrocarbon Equilibrium, J. Petroleum Technology, 1, 19, 1952.
5. Smith, J. M., others, "Introduction to Chemical Engineering Thermodynamics," 5th ed., p. 457, McGraw-Hill, New
York, 1996.
AuthorAffiliation
Philip T. Eubank*, M. Sam Mannan, and Kenneth R. Hall
Texas A&M University
AuthorAffiliation
* Send all correspondance to this author.
AuthorAffiliation
Authors
Philip T. Eubank is professor of chemical engineering at Texas A&M University (Chemical Engineering Dept.,
College Station, Tex. 778433122; Phone: 979-845-3339; Fax: 979-845-6446; E-mail: [email protected]). He has a
B. S. from Rose-Hulman Institute and a Ph.D. from Northwestern University. His research interests are in the
thermophysical properties of fluids and fluid mixtures plus electrical discharge machining.
M. Sam Mannan is associate professor of chemical engineering at Texas A&M University (Phone: 979-862-3985;
Fax: 979-458-1493; E-mail: [email protected]) and director of the Mary Kay O'Connor Process Safety Center. He
received the B. S. from the Engineering University in Dhaka, Bangladesh and the M. S. and Ph. D. from the
University of Oklahoma. His research interests include process safety and risk management, quantitative risk
assessment, reactive chemistry and transport modeling of chemical releases.
Kenneth R. Hall is the GPSA professor of chemical engineering and director of the Thermodynamics Research
Center at Texas A&M University (Phone: 979-845-3357; Fax: 979-845-6446; E-maü: [email protected]). During his
career, he has worked with Amoco and ChemShare and taught at the University of Virginia. He was a visiting
professor at Katholeike Universitiet te Leuven and Director of the Chemical and Transport Systems Division of the
National Science Foundation.

DETAILS

Subject: Chemical plants; Simulation; Process controls; Problem solving

PDF GENERATED BY SEARCH.PROQUEST.COM Page 3 of 5

Location: United States--US

Classification: 9190: United States; 8640: Chemical industry; 5310: Production planning &control

Publication title: Chemical Engineering; New York

Volume: 107

Issue: 5

Pages: 125-126,128,130

Number of pages: 4

Publication year: 2000

Publication date: May 2000

Section: Design

Publisher: Access Intelligence LLC

Place of publication: New York

Country of publication: United States , New York

Publication subject: Chemistry, Engineering--Chemical Engineering

ISSN: 00092460

Source type: Trade Journals

Language of publication: English

Document type: Feature

Document feature: Diagrams Tables Equations References

ProQuest document ID: 1019056110

Document URL: http://proxy.library.nyu.edu/login?url=https://search.proquest.com/docview/101905

6110?accountid=12768

Copyright: Copyright Access Intelligence LLC May 2000

Last updated: 2016-11-19

Database: ProQuest Central

LINKS

PDF GENERATED BY SEARCH.PROQUEST.COM Page 4 of 5

Linking Service

Database copyright  2020 ProQuest LLC. All rights reserved.

Terms and Conditions Contact ProQuest

PDF GENERATED BY SEARCH.PROQUEST.COM Page 5 of 5