Professional Documents
Culture Documents
PII: S0098-1354(18)30071-1
DOI: 10.1016/j.compchemeng.2018.02.005
Reference: CACE 6019
Please cite this article as: Raul Alejandro Vazquez Rojas , Francisco Javier Garfias Vásquez ,
Enrique Rodolfo Bazua Rueda , Simulation of a triple effect evaporator of a solution of caustic soda,
sodium chloride, and sodium sulfate using Aspen Plus, Computers and Chemical Engineering (2018),
doi: 10.1016/j.compchemeng.2018.02.005
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service
to our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and
all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Highlights
T
formation.
IP
Triple salt can be avoided by removing sodium sulfate from the caustic solution to a mass
fraction less than 0.002 at operating temperatures higher than 73 °C.
CR
US
AN
M
ED
PT
CE
AC
ACCEPTED MANUSCRIPT
Authors:
T
Francisco Javier Garfias Vásquez,1 PhD. garfiasv@unam.mx
IP
Enrique Rodolfo Bazua Rueda,1 PhD. erbr@unam.mx
1
Department of Chemical Engineering, Faculty of Chemistry, National University Autonomous of
CR
Mexico, Avenida Universidad 3000, Coyoacan 04510, Mexico D.F.
Keywords: Triple effect evaporator, Sodium hydroxide, Sodium chloride, Sodium sulfate
Simulation, Triple salt and precipitation
1. Abstract US
AN
Worldwide, the Chlor-Alkali process is the most well-known method for the production of chlorine
( ) and sodium hydroxide ( ). NaOH, also known as caustic soda, is a very important alkali
with many applications in the processing and production of paper, detergents, aluminum,
M
petrochemicals, inorganics, and in the food industry. The aqueous solution of caustic soda, known
as “cell liquor," produced in this process must be concentrated from 11 to 50 percent weight,
which is achieved through a multiple effect evaporator system. In some cases, the brine used as a
ED
raw material carries a few other components that cannot be separated before the feeding of the
brine into the process. The presence of sulfates and chloride ions in addition to high NaOH
concentrations (0.44 – 0.5 mass fraction) and high temperatures (above 60°C, at 86 mmHg) causes
the precipitation of a triple salt ( ). This work focuses on using and
PT
validating the model of a triple effect evaporator in Aspen Plus using plant data. According to our
results, lower temperatures and the extraction of sulfates could reduce the proportion of triple
salt that precipitates in the last stages of the evaporator.
CE
2. Introduction
AC
Multiple effect evaporation is the most common method for the industrial concentration of
inorganic solutions, which usually involves high boiling point elevation. Several case studies of
sodium hydroxide concentration can be found in evaporation technology texts (Holland, 1975;
Kern, 1965; Zain & Kumar, 1996), where the consideration of a single solvent-single solute mixture
has been adopted because of the availability of data concerning the thermodynamic and transport
properties of sodium hydroxide solutions. These documented data include Dühring’s rule
diagrams, heat of mixing diagrams, and overall heat transfer coefficient correlations. One of the
ACCEPTED MANUSCRIPT
main applications of these evaporation schemes is the evaporation of the caustic soda solution or
cell liquor from the Chlor-Alkali process (O'Brien et al., 2008). There are three main types of Chlor-
Alkali process: mercury cell, diaphragm cell, and membrane cell. Thus, depending on which is
installed in the processing plant, the cell liquor will have a different concentration of NaOH and
other components. For example, diaphragm cell technology yields a liquor with approximately 11
and 15 percent weight of sodium hydroxide and sodium chloride, respectively. In this situation,
the common examples of multiple effect evaporation cannot be applied because a two-solute
system and salt precipitation must be considered. Indeed, Kern (1965) includes the sodium
chloride precipitation and correlations for the calculation of a system with two solutes.
T
Furthermore, if there are sulfate salts within the brine fed to the process, a triple salt
IP
(Na2SO4·NaCl·NaOH) precipitates because of the presence of sodium sulfate, sodium chloride, and
sodium hydroxide (O'Brien et al., 2008). This particular salt causes several problems in the
equipment because of its abrasiveness, necessitating regular shut-downs for cleaning tasks and
CR
the need for nickel-plated reinforced surfaces.
As noted before, previous works concerning the multiple effect evaporation of caustic
solutions and its modeling and optimization have focused on solving design problems rather than
US
operating issues. For example, the additional equipment needed to deal with energy recovery and
salt precipitation has been reported in filed patents (Svance, 1958; Zabotto et al., 1978).
AN
3. Simulation model
A triple effect evaporator was designed in Aspen Plus, which is a modular process simulation
software with an extensive database for many chemical systems including electrolyte solutions.
M
In particular, the electrolyte non-random two liquids (ENRTL) model, based on Chen's work
(Chen and Evans, 1986; Chen and Song, 2004; Song and Chen, 2009), was used to fit experimental
data for numerous electrolyte systems.
ED
Taking advantage of the database and thermodynamic methods available for caustic solutions,
a model of a multiple effect evaporator for cell liquor evaporation was developed in the Aspen
PT
4. Process description
CE
A common type of evaporator for this process has three evaporation stages, called effects, and
a flash tank in countercurrent configuration, which means the cell liquor or diluted stream is fed to
the third effect while the live steam is fed to the first effect. This configuration is normally
AC
employed when the viscosity of the solution increases with increasing concentration (Reinhard,
1989).
The liquor extracted from the first effect is finally concentrated in a flash tank operated at
vacuum conditions and is then cooled and filtered for final storage.
The third effect and the flash tank operate under vacuum conditions 64 mmHg ( 8 kPa) using
a barometric condenser. The first effect is set to operate at pressures slightly above atmospheric
pressure ( 120 kPa). Meanwhile, in the second effect, the absolute pressure is around 26.6 kPa.
Each effect is assembled with a body, a one-pass shell and tube heat exchanger, a circulation
ACCEPTED MANUSCRIPT
pump system, a feed pump system and a hydro-cyclone. In both the third and second effects,
there is an elutriation column attached to the bottom of the body of the effect to separate a
crystal-rich slurry or magma. This magma is centrifuged and filtered for liquor recovery and crystal
separation. The condensate of the second and first effects is also used to pre-heat the liquor fed to
each of those effects, as shown in Figure 1.
The model was set up using four phase separators for the three effects and the flash tank (1EF,
2EF, 3EF, and FLASH) according to a common design of this type of evaporator. In addition, three
heaters were coupled with the flash modules to represent the heat exchange between steam and
T
liquor and four extra heaters for preheating the liquor fed to the second and third effects. A
component separator was used for the solid extraction in each effect. Finally, pumps, mixers, and
IP
splitters were introduced to match the layout of a real evaporator unit. The salt decomposer inlet
and outlet streams allow the bypass of a small fraction of the feed to the second effect that is
CR
treated by a subprocess named "salt decomposer” in which NaCl and Na2SO4 crystals are removed
by crystallizers and centrifuges.
The following process flow diagram shows the proposed simulation of a triple effect
evaporator in Aspen Plus.
US
AN
M
ED
Fig. 1. Triple effect evaporator model in Aspen Plus. 1EF, 2EF, 3EF correspond to the first, second and third effect,
respectively.
The mass and heat stream names displayed in the simulation diagram in Figure 1 are defined
CE
in Table 1. The tear streams selected and specified with chemical plant mass balance information
were M2, Q1, and the de-superheating condensate stream.
AC
ACCEPTED MANUSCRIPT
ID Description
Feed Liquor fed to the third effect
Steam Superheated steam fed to the third effect
F Feed stream
V Vapor stream
C Condensate
L Liquor
T
M Magma
Q
IP
Heat transferred from steam
QPH Heat transferred from condensate
The defined input streams in the simulation model were the following: steam, feed, and
CR
salt decomposer outlet.
US
the pressure in the three effects (1EF, 2EF and 3EF),
pressure and temperature in the flash tank (FLASH),
temperature change in the feed preheater heat exchangers,
AN
pressure and vapor fraction in the heat exchangers of each effect,
outlet pressure from the feed bombs,
the fraction of separation per component in each of elutriation columns.
M
The input variables for the model are presented in Tables 2 and 3. Additional input as tear
streams, cooling water, and recycle streams were added to match the mass balance with the
design data. The input data was extracted from a plant operation sheet provided by Mexichem
ED
(Vazquez, 2015).
Table 2. Input values for the inlet streams of the evaporator model.
PT
T
The SYSOP15M database was selected from the Aspen Plus package files, which contains the
ENRTL-RK model parameters adjusted from published data (Silcock, 1979). According to the
IP
software information, this database includes the parameters for a system based on the reactions
shown in Table 4 of the next section for the following conditions.
CR
Temperature less than 200 °C.
Pressure under 1013.25 kPa (10 atm).
NaOH concentration up to 60 wt%.
NaCl concentration up to saturation.
US
AN
Na2SO4 concentration up to saturation.
5. Thermodynamic method
The selected thermodynamic method for the activity coefficient calculation was the ENRTL
M
method with parameters for caustic systems, which include NaOH, NaCl, and Na2SO4. The
database for these systems is labeled SYSOP15M, which according to the software documentation
has the parameters fit for a multicomponent solution and salt precipitation following the
ED
Reaction Description
Sodium hydroxide
dissociation
Sodium chloride dissociation
CE
Sodium chloride
precipitation
Sodium sulfate precipitation
Decahydrated sodium
sulfate (Glauber´s Salt)
precipitation
Double salt precipitation
The dissociation and precipitation constants among with the ENRTL parameters used in this
study were those from the SYSOP15M database.
These component parameters are present in the default ENRTL method database; however,
we found that by using those parameters, the calculations of NaOH solutions above 40 wt% were
highly inaccurate. So, the SYSOP15M database was selected because it better approximated the
T
boiling point increase data found in the Dühring as shown in Figure 2 a) and b).
IP
250
Boiling point of aqueous solution of NaOH, °C
CR
200
150
100 US
AN
50
0
M
Fig. 2 a). Plot of boiling point for caustic aqueous solutions calculated with default parameters (continuous lines)
comparison with adjusted data (markers)(Holland, 1975).
CE
AC
ACCEPTED MANUSCRIPT
200
150
T
100
IP
50
CR
0
0 20 40 60 80 100 120 140 160
10%
10%
20%
20%
US
Boiling point of pure water, °C
30%
30%
40%
40%
50%
50%
60%
60%
AN
Fig. 2 b). Plot of boiling point for caustic aqueous solutions calculated with SYSOP15M parameters (continuous lines)
comparison with adjusted data (markers)(Holland, 1975).
M
6. Results
As stated before, the objective of this work was to analyze the behavior of the caustic liquor in
the evaporator. Particularly, the precipitation of a triple salt is a very sensitive problem in multiple
ED
effect evaporators that contain sulfates in the feed stream. Because very limited data is available
concerning the formation of the triple salt, first, a phase diagram of the phases present in the
solution in a certain range of temperatures was calculated. In addition, the concentration of each
PT
component was normalized so that the behavior of this system could be represented in a 2D-plot.
The simulation model presented in Figure 1 was run by setting the temperature and pressure
in the flash tank modules that represent each of the effects of the evaporator. The values of the
concentration of each component in the outlet liquor were compared with the data provided by
the plant (Vazquez, 2015), and the percent error was calculated according to Eq. (1).
AC
| |
(1)
ACCEPTED MANUSCRIPT
T
Na2SO4 0.27% 2.78% 9.60%
IP
An actual triple effect evaporator works with inputs and outputs other than the liquor feed
CR
stream and the live steam, unlike the more idealized models found in the literature. Most of the
additional streams and equipment are used for energy recovery purposes and to handle salt
precipitation.
US
According to the assumptions made for the evaporator model to reduce convergence issues,
the process flow diagram was adjusted to include some of the inputs and outputs found in a real
operating evaporator. Therefore, the simulation should provide results similar to the overall
conditions of each effect. A component separator was used to model the elutriation columns.
AN
Because the quantity of each component in the solution and the salt flows are known from plant
data, the amount of each molecular component (H2O, NaOH, NaCl, and Na2SO4) can be calculated
from a mass balance equation and introduced into the simulation.
M
ED
Table 6. Percent error between the flash tank calculation and real data.
PT
The error between the real and the calculated data was shown to be less than 1% (Table 6),
except for the composition of sodium chloride and sodium sulfate in the solution, which had errors
ACCEPTED MANUSCRIPT
of 8.43 and 12.9%, respectively. For the purposes of this work, an error less than 10% is
acceptable. While the sodium sulfate concentration percent error is slightly higher than this value,
it can be tolerated because the concentration of this component in the solution never exceeds 1
wt%. Therefore, it is acceptable to assume that these deviations will not influence the calculation
of properties for the liquid streams significantly.
The variations in the concentrations within each of the effects are presented in Table 5. The
greater accuracy in this part was expected because of the statement presented earlier regarding
the influence of salt precipitation.
T
Nevertheless, our simulation presents a good approach to model the conditions of the real
IP
operating evaporator in terms of phase equilibrium calculations.
CR
The next plots were calculated to determinate the salt precipitation with the three electrolytes
studied before being fed to the flash tank. The mass flow of one electrolyte was varied while the
other two were held constant at temperatures from 0 to 140 °C. The initial input specifications for
US
Figures 3–5 are listed in Table 7. Then using the Aspen Plus vapor-liquid-solid-equilibrium routine
several points were calculated over the temperature range to determine the presence of solids.
Table 7. Flash inlet stream component values at a pressure of 1 atm (101.3 kPa).
AN
Flash inlet stream values
Mass
Component Mass Flow (kg/h)
Fraction
M
180
160
Na2SO4(s) + NaCl(s) + Solution +
Vapor
140
120
Temperature , °C
T
NaCl(s) +
80 Solution
IP
60
CR
NaCl(s) + Na2SO4(s) + NaCl(s) +
40 Solution Solution
20
0
0 0.1
US
0.2 0.3 0.4 0.5
AN
Mass fraction of NaOH in the liquid at flash tank inlet
Fig. 3. Phase diagram at the flash inlet liquor concentration on varying the NaOH mass fraction at 1 atm.
M
In Figure 3, the upper zone contains three phases as the solution boils at higher temperatures.
In contrast, in Figures 4 and 5, there is no vapor phase because of the large increase in boiling
point. As noted, sodium chloride precipitates over the entire temperature range. According to this
ED
phase diagram and in agreement with reference data found in the literature (Schroeder, Gabriel
and Partridge, 1935; Silcock, 1979), the triple salt is present at ranges between 0.33 and 0.55 mass
fraction of sodium hydroxide and 40 to 160°C.
PT
CE
AC
ACCEPTED MANUSCRIPT
120
Solution
T
80
IP
60
CR
40 Na2SO4(s) + NaCl(s)
+Solution
20
0
0 0.1 0.2
US 0.3 0.4 0.5 0.6
AN
Mass fraction of NaCl in the liquid at flash tank inlet
Fig. 4. Phase diagram at the flash inlet liquor concentration on varying the NaCl mass fraction at 1 atm.
M
140
120
ED
100
Solution
PT
Temperature , °C
80
CE
60
40
AC
0
0 0.005 0.01 0.015 0.02 0.025
Mass fraction of Na2SO4 in the liquid at flash tank inlet
Fig. 5. Phase diagram at the flash inlet liquor concentration on varying the Na2SO4 mass fraction at 1 atm.
ACCEPTED MANUSCRIPT
The three diagrams show that precipitation of the triple salt is more temperature dependent
than dependent on the concentrations of sodium sulfate and sodium chloride. Because
concentrations of sodium hydroxide above 44 wt% are required by the product specification, the
reduction of sodium sulfate to a concentration of less than 0.002 seems to be the first alternative.
Operation at temperatures less than 73 °C will also avoid the triple salt precipitation zone, but this
option is not possible because the steam recovered in the third effect must be at a higher pressure
than the other two effects to maintain the countercurrent scheme.
In the next analysis, the process variables were varied to observe their influence on the
T
formation of the triple salt within the effects. A similar analysis can be carried out using phase
diagrams, but it is more difficult to observe the variations because of the four-component system.
IP
8. Sensitivity analysis
CR
The presented model was sufficiently accurate to calculate the concentrations of liquor and
salt flows between each effect and the final product according to the design data.
The sensitivity analysis was run by varying the mass flow of the live steam feed to the heat
US
exchanger in the first effect and the operating pressure in the flash tank. In each scenario, the
calculated variables were the mass flow of precipitated salt, the stage temperature, and the outlet
vapor. According to the phase diagrams shown in Figures 3, 4, and 5, high temperatures promote
AN
the precipitation of the triple salt and represent an increase in the steam required to be fed to the
system, thereby reducing the economy of the evaporator.
The first approach was to observe the behavior of the evaporator while reducing the amount
of live steam fed to the first effect and keeping the pressure profile in the system constant.
ED
160
155
First effect temperature, °C
PT
150
CE
145
140
AC
135
130
68 70 72 74 76 78 80 82 84
Live steam flow feed to the first effect, 103 kg/h
Fig. 6. The temperature increase in the first effect with increasing steam feed.
ACCEPTED MANUSCRIPT
2.50
Triple salt flow from the first effect, 103 kg/h
2.00
1.50
T
IP
1.00
CR
0.50
0.00
68 70 72
US
74 76
Live steam flow feed to the first effect,
78
103 kg/h
80 82
AN
Fig. 7. Salt precipitation with increasing steam feed.
Figure 7 shows that by reducing the temperature in the first effect, it is possible to reduce the
M
quantity of triple salt formed in this stage. Nevertheless, the caustic soda concentration in the
effluent leaving this process was reduced, based on the results shown in Figure 8.
ED
0.50
0.48
Mass fraction of NaOH in the flash tank
0.46
PT
0.44
0.42
CE
0.40
0.38
0.36
AC
0.34
0.32
0.30
68.00 70.00 72.00 74.00 76.00 78.00 80.00 82.00
Live steam flow feed to the first effect, 103 kg/h
Fig. 8. The concentration of NaOH in the flash tank at a constant temperature and pressure.
ACCEPTED MANUSCRIPT
Because a product with less than 50% weight (or 0.5 mass fraction) is not allowed at the end
of the evaporator, another process condition must be adjusted.
T
in the high temperatures of this stage, the triple salt is formed. In these calculations, the vapor
IP
fraction was held constant, so the ratio between the liquid and vapor phases remained constant.
CR
2300
Mass flow of triple salt precipitated in the flash
2250
2200
US
AN
tank, kg/h
2150
M
2100
ED
2050
10 110 210 310 410 510 610 710 810
Absolute pressure in the flash tank, mmHg
PT
21500
Total mass flow of salt precipitates, kg/h
21000
20500
20000
19500
19000
T
18500
IP
18000
CR
17500
0 100 200 300 400 500 600 700 800
Pressure in flash tank, mmHg
0.504
US
Fig. 10. Total precipitated solid stream in the flash tank.
AN
Mass fraction of NaOH in the liquid phase within
0.502
0.5
M
0.498
the flash tank
0.496
ED
0.494
0.492
PT
0.49
CE
0.488
0.486
0 100 200 300 400 500 600 700 800
AC
In the plots shown in Figures 9–11, an absolute pressure closer to atmospheric pressure would
imply an increase in the boiling point of the solution. As shown in Figure 9, operation from 64
mmHg to almost atmospheric pressure only reduced the total mass flow of triple salt by about
8.2%. The difference between the mass flow values in Figures 9 and 10 correspond to the values
for solid or precipitated sodium chloride because no other salt seems to be present under these
ACCEPTED MANUSCRIPT
conditions. Finally, Figure 11 shows that an increase in the absolute pressure in the flash tank will
reduce the concentration of caustic soda in the final product. Although a decrease of
concentration until 49.5% is affordable, the operating pressure only results in a triple salt
precipitation reduction of 1.1%.
11.Conclusions
A triple effect evaporation model proposed in Aspen Plus was able to predict the
concentration of the liquor in the effect exits with relative errors of less than 10%. Evaporated
water and precipitated salt were also compared to the plant data, and the results were used to
T
establish the base case scenario. Although solid salt flow was accurately calculated with the
IP
simulator (±10% error), the evaporated water from the first and second effect was not. Therefore,
the heat exchanger used for preheating the feed to each effect was adjusted to match the results
to the available data. These deviations could be attributed to the assumption of the preheater as a
CR
simple heat transfer module, but this was necessary to achieve convergence.
In addition, the calculated boiling temperatures of the liquor presented deviations of up to
±3°C, which is highly accurate considering the assumptions that were made. Sodium chloride
US
precipitation is not avoidable, so the evaporator is designed to deal with its presence in all parts of
the process. Nevertheless, the specific conditions for the formation of the triple salt, as shown in
Figures 3, 4 and 5, can be found; that is, at high concentrations of sodium chloride and sodium
AN
hydroxide, low concentrations of sodium sulfate (around 0.005 mass fraction), and temperatures
above 74°C, the presence of the triple salt in the caustic liquor is unavoidable.
According to the data displayed in the previous section, the precipitation of a triple salt can be
M
avoided only if the sodium sulfate is removed from the caustic solution to a mass fraction less than
0.002 at operating temperatures higher than 73 °C. Reducing the operating temperature to less
than 73 °C is a feasible solution, if the system vacuum capacity allows it, to prevent compromising
ED
12.Acknowledgment
We are grateful to Mexichem Derivados S.A. de C.V. and the Research & Development Center
CE
of Mexichem for the data provided to compare the results of the simulation presented in this
work. In addition, we thank the National Council of Science and Technology (CONACYT) for the
support to realize this investigation.
AC
ACCEPTED MANUSCRIPT
References
Chen, C., & Evans, L. B. A local composition model for the excess gibbs energy of aqueous
electrolyte systems. AIChE Journal, 32(3) (1986), 444-454.
Chen, C., & Song, Y. Generalized Electrolyte-NRTL Model for Mixed-Solvent Electrolyte Systems.
AIChE Journal, 50(8) (2004), 1928-1941.
T
evaporation and extraction. New Jersey: Prentice-Hall Inc (1975).
IP
Kern, D. Process Heat transfer. New York: McGraw-Hill (1965).
O'Brien, T., Bommaraju, T., & Hine, F. Handbook of Chlor-Alkali Technology. USA: SpringerScience,
CR
Business Media Inc, (2008).
US
Reinhard, B. Evaporation technology: principles, applications & economics. US: Weinheim, (1989).
Schroeder, W. C., Gabriel, A., & Partridge, E. P. Solubility Equilibria of Sodium Sulfate at
Temperatures of 150 to 350°C. Effect of Sodium Hydroxide and Sodium Chloride. J. Am.
AN
Chem. Soc., 9(57) (1935), 1539-1546.
Press.
Song , Y., & Chen, C. Symmetric Electrolyte Nonrandom Two-Liquid Activity coefficient Model. Ind.
ED
Zabotto, G., Guichard, J., & Fourier, D. US Patent No. 4087253 A, (1978).
CE
Zain, O. S., & Kumar, S. Simulation of a multiple effect evaporator for concentrating caustic soda
solution-computational aspects. Journal of chem eng of Japan, 29(5) (1996), 889-893.
AC