CHAPTER SEVEN

FLASHES AND DECANTER

Aspen Pluss model library contains two different rigorous flash blocks that solve the
appropriate material, energy balance, and equilibrium equations. The Flash2 block is
designed to produce a single vapor phase and a single liquid phase that are in equi-
librium when the flash conditions are specified such that a multicomponent mixtures
state is in the two-phase region. Similarly, the Flash3 block is designed to produce one
vapor phase and two liquid phases in equilibrium for suitably specified process condi-
tions. The Flash3 block is also capable of solving a liquidliquid equilibrium problem
under conditions where no vapor is produced. A similar block is Decanter, which is
designed to produce two liquid phases in equilibrium in the absence of a vapor phase.
The Flash2, Flash3, and Decanter blocks can be found in the model library under the
Separators tab.

7.1 Flash2 BLOCK

Figure 7.1 is a graphical depiction of a Flash2 block. The nomenclature used is as

follows: F represents the total feed flow, in moles/time; fi the flow of component
i in the feed, in moles/time; V the total vapor flow, in moles/time; vi the flow of
component i in the vapor, in moles/time; L the total liquid flow, in moles/time; li the
flow of component i in the liquid, in moles/time; and n the number of components.
Including the flash temperature Tf and pressure Pf results in 2n + 2 independent
variables given the feed state. These are equilibrium temperature; flash pressure; two
total flows, L and V ; and 2(n 1) component flows; or alternatively, 2n component
flows, excluding the total flows. Mole fractions are calculated from the independent

Teach Yourself the Basics of Aspen Plus By Ralph Schefflan

Copyright 2011 John Wiley & Sons, Inc.

93
94 FLASHES AND DECANTER

Vapor, V, v

Flash2
Feed, F, f P

Heat, Q
Liquid, L, l

Figure 7.1 Flash2 model.

variables by an equation such as

vi
yi = n (7.1)
j =1 vj

where yi is the mole fraction of component i in the vapor.

The applicable material balances are n componential equations such as

fi vi li = 0 (7.2)

or alternatively, n 1 equations such as equation (7.2) and one overall material balance
given by

F V L=0 (7.3)

Additionally, n equilibrium equations which describe the equality of the fugacities

of components in each phase are required. When the liquid fugacity is represented by
an equation of state, where L
i and i are the fugacity coefficients of component i in
V

the liquid and vapor phases, respectively, the result is

yi Vi xi L
i =0 (7.4a)

When the liquid fugacity is represented by an activity coefficient equation, where

i is the activity coefficient of component i and the vapor phase is represented by an
equation of state

yi Vi P i xi piv = 0 (7.4b)

where piv is the vapor pressure of component i , results. For the sake of simplicity, the
Poynting correction (see Prausnitz et al., 1999), which has a contribution only for very
light components, has been omitted.
 Flash2 BLOCK 95

An overall energy balance, where h with a suitable subscript represents the enthalpy
of the feed, liquid, and vapor, respectively, and Q represents the heat added, given by

hF F + Q hL L hV V = 0 (7.5)

completes the set of 2n + 1 equations.

The Flash2 block permits the specification of two of the four possibilities: flash tem-
perature, flash pressure, heat required, and the fraction of the feed vaporized, except the
combination Q and V/F . Various permutations of the specifications require other mod-
ifications of the equations and the list of unknowns. For example, if V/F is specified,
equation (7.5) is replaced by
  n
V j =1 vj
n =0 (7.6)
F specified j =1 fj

If Q, the heat added or removed, and the flash pressure are specified, the 2n com-
ponent flows and the flash temperature can be calculated. If the flash temperature and
pressure are specified, the sequence of the solution of the equations changes. Equations
(7.1) through (7.4) are solved simultaneously, after which equation (7.5) is solved
for Q.
An example of the use of a Flash2 block is given at Chapter Seven Examples/
Flash2Example. An important aspect of the setup, which is applicable to all simula-
tions, is the selection of the data bank from which the activity coefficient or equation
of state parameters were derived. Figure 7.2 shows that the binary interaction Wilson

Figure 7.2 Flash2 example Wilson parameters.

96 FLASHES AND DECANTER

Figure 7.3 Flash2 example specifications.

Figure 7.4 Flash2 example results.

parameters were obtained from the data source apv70vle-lit. The upper and lower tem-
peratures of applicability are also shown. If this set is deemed inappropriate, selection
of the tab Databanks permits the user to select an alternative source of data. The input
specifications for this example are shown in Figure 7.3, and the flash results are given
in Figure 7.4.
A useful feature of the Flash2 and some other equilibrium stage blocks is the ability
to specify entrainment; that is, a specified fraction of the liquid phase is carried by
the flashing vapor into the overhead. A comparison of the performance of a Flash2
block with and without entrainment is given at Examples/Flash2EntrainmentExample.
The flowsheet is shown in Figure 7.5. A comparison of the overheads produced by
the flash is given in Figure 7.6. The composition of stream 22 is lower in the more
volatile component, methanol, and its fraction vapor is less than 1.0.

7.2 Flash3 BLOCK

Figure 7.7 is a graphical depiction of a Flash3 block. The nomenclature used is as

follows: F represents the total feed flow, in moles/time; fi the flow of component
i in the feed, in moles/time; V the total vapor flow, in moles/time; vi the flow of
component i in the vapor, in moles/time; L1 the total flow of the first liquid phase, in
moles/time; Iil , the flow of component i in the liquid, in moles/time; L2 the total flow
 Flash3 BLOCK 97

F1

D1 3
1
DLPL 22

44

F2

33

Figure 7.5 Flash2 entrainment example.

Figure 7.6 Flash2 entrainment results.

98 FLASHES AND DECANTER

Vapor, V, vi

Flash3
Feed, F, fi

Second Liquid, L2, l2i

Heat, Q
First Liquid, L1, l1i

Figure 7.7 Flash3 model.

of the second liquid phase, in moles/time; and Ii2 the flow of component i in the liquid,
in moles/time. Including the flash temperature Tf and pressure Pf results in 3n + 2
independent variables, where n is the number of components given the feed state. The
variables consist of the equilibrium temperature, the flash pressure, and the three total
flows (i.e., and L1 , L2 , and V ), and 3(n 1) componential flows or, alternatively, 3n
componential flows, excluding the total flows. Mole fractions are calculated from the
independent variables by an equation such as equation (7.1).
The applicable material balances are n componential equations, such as

fi vi li1 li2 = 0 (7.7)

or alternatively, n 1 equations such as equation (7.7) and one overall material bal-
ance, given by

F V L1 L2 = 0 (7.8)

Additionally, 3n equilibrium equations which describe the equality of the fugacities

of components in each phase can be written, but only 2n are independent. The vapor-
phase fugacity is represented by an equation of state where the Vi are the fugacity
coefficients of component i in the vapor phase, and the liquid phases are represented
by an activity coefficient equation where i1 and i2 are the activity coefficients of
component i and piv is the vapor pressure of component i . Then any two of

i1 xi1 i2 xi2 = 0 (7.9a)

yi ViP i1 , xi1 piv = 0 (7.9b)
yi ViP i2 , xi2 piv =0 (7.9c)

apply. For the sake of simplicity the Poynting correction (see Poling et al., 2000), which
has a contribution only for very light components, has been omitted from equations
(7.9b) and (7.9c). The overall energy balance given by equation (7.5) completes a set
of 3n + 1 equations:
 DECANTER BLOCK 99

hF F + Q hL1 L1 hL2 L2 hV V = 0 (7.10)

The Flash3 block permits the specification of two of the four possibilities: flash tem-
perature, flash pressure, heat required, and the fraction of the feed vaporized. Various
permutations of the specifications require other modifications to the equations and the
list of unknowns. For example, if Q, the heat added/removed and the flash pressure
are specified, the 3n component flows and the flash temperature can be calculated.
When specifying the block it is important to identify the key component in the
bottom takeoff (liquid 2) such that the block conforms with its physical behavior (i.e.,
the heaviest of the solvents should be selected). An example of the use of a Flash3
block is given at Examples/Flash3Example.

7.3 DECANTER BLOCK

Figure 7.8 is a graphical depiction of a Decanter block. The nomenclature used is as

follows: F represents the total feed flow, in moles/time; fi the flow of component i
in the feed, in moles/time; L1 the total flow of the first liquid phase, in moles/time; li1
the flow of component i in the liquid, in moles/time; L2 the total flow of the second
liquid phase, in moles/time; and li2 the flow of component i in the liquid, in moles/time.
Including the decanter temperature Td and pressure Pd results in 2n + 2 independent
variables, where n is the number of components given the feed state. The variables
consist of the equilibrium temperature and the decanter pressure, the two total flows
L1 and L2 , and 2(n1) componential flows or, alternatively, 2n componential flows,
excluding the total flows. Mole fractions are calculated from the independent variables
by an equation such as (7.1).
The applicable material balances are n componential equations, such as

fi li1 li2 = 0 (7.11)

Feed, F, fi

Decanter First Liquid, L1, li1

Second Liquid, L2, li2

Figure 7.8 Decant model.

100 FLASHES AND DECANTER

or alternatively, n1 equations such as equation (7.11) and one overall material bal-
ance, given by

F L1 L2 = 0 (7.12)

Additionally, n equilibrium equations can be written which describe the equality of the
fugacities of components in each phase, where i1 and i2 are the activity coefficients
of component i and xi1 and xi2 are the mole fractions of component i in phases 1 and
2, respectively:

i1 xi1 i2 xi2 = 0 (7.13)

An overall energy balance,

hF F + Q hL1 L1 hL2 L2 = 0 (7.14)

completes a set of 2n + 1 equations.

Most decantation processes are carried out isothermally, and the heat associated
with solute moving between phases is usually negligible, thus eliminating the need
for equation (7.14) and reducing the number of unknowns to 2n when the pressure
is specified. This is reflected in the blocks input forms, where the only allowable
specifications are temperature and pressure. As in the Flash3 block, care must be
taken to specify the key component in the second liquid product, which should be
the heavier phase. The feed from Flash3Example was used to create an example of
the use of a Decanter block, which may be found in Examples/DecanterExample. It
would be useful to compare results of Flash3 and Decanter since both are carried out
at the same conditions: 1 psia and 62 F. The Flash3 specification is V /F = 0.1, which
produces a temperature of 61.9 F, which is close enough for a comparison. Figure 7.9
shows the data sources for the Uniquac parameters, which are LLE data compared
to VLE data for Flash3. This is an example of a conundrum. Which data produce
the correct answer? Both seem reasonable. Aspen Plus has solved the appropriate
equations, but it is up to the user to select the correct result. Perhaps some laboratory
work will be required.
The block also permits an alternative formulation of the basic equations as a Gibbs
free energy minimization problem. This formulation and some methods of solution are
described by Walas (1985). Additional details are provided in Chapter Ten.
An interesting aspect of the Decanter block is the possibility of its use to simulate a
batch extraction process. The describing equations are identical to those of a continuous
decanter if the flows per unit time are replaced with a batch charge as a feed and the
batch removal of the two products. A common practice in industry is to place a solvent
A containing a quantity of C, the product desired, into a vessel. Following this, the
solvent B is charged to the vessel. After agitation the products are allowed to settle. This
is the equivalent of feeding both streams as a mixture to the decanter, producing the
products L1 and L2 . Common industrial practice is to remove the equilibrium B stream,
now containing extracted C, and recharging the vessel with fresh B. This process is
easily simulated with several Decanter blocks in series, as shown in Figure 7.10.
The combined feeds are repeatedly settled, the extract and raffinate removed, and the
resulting product sent to the next stage. Each stage represents the same vessel one
 WORKSHOPS 101

Figure 7.9 Alternative data sources.

SOLV-C-1 SOLV-C-2

A-PLUS-C DECANT1 DECANT2 DECANT3

RAFF1 RAFF2 RAFF3

EXTRACT1 EXTRACT2 EXTRACT3

MIX COMBINED

Figure 7.10 Batch extraction process.

time step later in the process, when it contains the remaining raffinate and is charged
with fresh solvent C. The composition of the mixed extracts is the product of the
mixer.

7.4 WORKSHOPS

Workshop 7.1 Use the Flash2 block to flash the feeds in Table 7.1 adiabatically at
constant pressure for the purpose of reducing the acetone composition in the bottoms
product. The feed temperature is 70 C and its pressure is 760 mmHg.
Be sure to set up appropriate units. Use Aspen Pluss stored Uniquac, NRTL, and
Wilson parameters as the basis for the calculations. What do you observe?
102 FLASHES AND DECANTER

TABLE 7.1 Feeds for Workshop 7.1

Component Feed 1 (lbmol/hr) Feed 2 (lbmol/hr)
Ethanol 40
Water 50
Acetone 10

Workshop 7.2 Based on the solution to Workshop 7.1, using 760 mmHg as a base
case, create a sensitivity study for pressure varying between 10 and 760 mmHg. Show
the effect of varying flash pressure on fraction vaporized, mole fraction acetone in the
vapor and liquid, and flash temperature using Aspen Pluss stored Uniquac parameters.

Workshop 7.3 The feeds in Table 7.2 are to be fed to a Flash3 block at a pressure
of 760 mmHg and a temperature of 80 C. Using Aspens stored Uniquac parameters,
use a sensitivity study to find the temperature at which the vapor phase disappears
when operating at 760 mmHg. The temperature and pressure of both feeds are 25 C
and 760 mmHg.

TABLE 7.2 Feeds for Workshop 7.3

Component Feed 1 (lbmol/hr) Feed 2 (lbmol/hr)
Ethanol 10
Water 40
Toluene 50

Workshop 7.4a Using the feeds from Workshop 7.3, employ the Decant block
to calculate the composition and quantity of the resulting phases at 25 C using the
parameters provided in Table 7.3. Be sure to set up appropriate units.

TABLE 7.3 Uniquac Parameters for a TolueneWaterEthanol System

Parameter/ Parameter Estimate
Element No. Component Pair (SI units)
UNIQ/2 TolueneH2 O 926.6623
UNIQ/2 H2 OToluene 258.8106
UNIQ/2 EtOHH2 O 301.8589
UNIQ/2 H2 OEtOH 119.5998
UNIQ/2 EtOHToluene 49.37053
UNIQ/2 TolueneEtOH 100.2406

Workshop 7.4b Develop a sensitivity study to determine the effect of temperature

on this operation. Vary the temperature from 30 to 100 C, displaying the amount of
ethanol extracted into the water-rich phase and the ethanol composition of the toluene-
rich phase. Be sure that the base case is executed last. Note all observations concerning
this sensitivity study. Are the results reasonable?
 REFERENCES 103

Workshop Notes
Workshop 7.1 A duplicator block will permit all three simulations to be done at once,
and the results will facilitate easy comparison. Two of the results are comparable. The
third displays a similar trend, but which are correct?

Do the activity coefficient models work for this system?

Were the parameters in the database fit properly?
Should new data be located and custom-fit?
Would an in the field design be acceptable for controlling the product compo-
sition (e.g., a secondary sensor loop feeding the set point of the primary control
loop, to make up for model deficiencies)?

Workshop 7.3 A sensitivity study with a plot should be a good guide for a design
specification. See the solutions to Workshops 7.4a and 7.4b.

Workshop 7.4 All Uniquac parameters except Uniq/2 equal must be set to zero.

Where did the list of Uniquac parameters come from? How can their quality be
assessed?
Results show that extraction improves as the temperature gets colder.

REFERENCES

Poling, B. E., Prausnitz, J. M., and OConnell, J. P., The Properties of Gases and Liquids, 5th ed., McGraw-
Hill, New York, 2000.
Prausnitz, J. M., Lichtenthaler, R. N., and de Avezedo, E. G., Molecular Thermodynamics of Fluid-Phase
Equilibria, 3rd ed., Prentice Hall, Upper Saddle River, NJ, 1999, p. 41.
Walas, S. M., Phase Equilibria in Chemical Engineering, Butterworth, Woburn, MA, 1985, p. 388.