See discussions, stats, and author profiles for this publication at: [Link]

net/publication/280080727

Is MVR an efficient alternative to optimize the energy use in an evaporation

unity of electrolytic caustic soda production? An exergy approach to the
diaphragm arrangement

Conference Paper · July 2015

CITATIONS READS

0 2,073

3 authors, including:

Carlos Eduardo Keutenedjian Mady Silvio De Oliveira Junior

University Center of FEI University of São Paulo
60 PUBLICATIONS   374 CITATIONS    222 PUBLICATIONS   2,168 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Techno-economic and environmental assessment of biorefinery technologies View project

Exergy and environmental ranking of bioethanol production routes View project

All content following this page was uploaded by Carlos Eduardo Keutenedjian Mady on 15 July 2015.

The user has requested enhancement of the downloaded file.

 PROCEEDINGS OF ECOS 2015 - THE 28TH INTERNATIONAL CONFERENCE ON
EFFICIENCY, COST, OPTIMIZATION, SIMULATION AND ENVIRONMENTAL IMPACT OF ENERGY SYSTEMS
JUNE 30-JULY 3, 2015, PAU, FRANCE

Is MVR an efficient alternative to optimize the

energy use in an evaporation unity of electrolytic
caustic soda production? An exergy approach to
the diaphragm arrangement

Carlos Eduardo Keutenedjian Madyab, Roberto Nicolas de Jardin Juniorc, Silvio de

Oliveira Juniora
a
Polytechnic School of the University of São Paulo, Av. Prof. Mello Moraes, 2231, 05508-900, São
Paulo, Brazil, cadumady@[Link], soj@[Link]
b
University of Campinas, Faculty of Mechanical Engineering, Mendeleyev St., 200, 13083-970 Campinas
- SP, Brazil, cadumady@[Link]
c
Unipar Carbocloro, nicolasjardin@[Link]

Abstract:
Energy and Exergy analysis were applied in an evaporation unity of electrolytic caustic soda production using
the diaphragm arrangement. The objective of the vaporization is increase the concentration of caustic soda
from entrance concentration which is at the order of magnitude of 10% (mass basis), until the ternary mixture
of caustic soda, water and sodium chloride reaches the commercial level, which is 50% in mass of caustic
soda. The evaporation process has three stages and one flash tank, which makes the steam consumption to
be high. To reduce this consumption it was proposed to use mechanical vapor recompression (MVR) in two
different stages of the process, making it possible to the decrease the steam consumption in 80% with the
necessity of additional 14.3MW electrical power. In energy basis the reduction in the plant consumption would
be about 50%. Moreover, there would be a reduction in the cost of the operation of the plant with the new
configuration. Nevertheless, from the exergy consumption point of view the original configuration consumes
75% of the plant configuration with the vapor recompression. This can be explained by the difference in the
quality of the electrical energy when compared with the steam used in the plant. When it is taken into account
the energy and exergy necessary to produce the electricity and the steam (using a boiler with natural gas
inside the factory), the mechanical vapor recompression uses less energy sources. Moreover, from the
perspective of exergy analysis the costs were lower.

Keywords:
Exergy analysis, Caustic soda production, Mechanical vapor recompression

1. Introduction
The Brazilian production of caustic soda in 2010 were estimated in 1,536 thousands of tons, being
63.4% originated from diaphragm process, 22,2% from the membrane process and 14.4% using
mercury process. All processes use the electrolysis of aqueous solution of sodium chloride to obtain
the caustic soda [1]. In the diaphragm process, the membrane has a poor permeability, and typically,
the resultant product of the electrolysis has a mass concentration of caustic soda of 12% and a high
concentration of sodium chloride. Therefore, there is a necessity of increase the concentration of
caustic soda and decrease the concentration of sodium chloride to the commercial level, which is 50%
and 1% in mass basis, respectively. To this aim, the method commonly used is a multi-stage
vaporization process, and since it is a ternary mixture, it is used hydrocyclones and centrifugal
systems for the removal of NaCl crystals. Nevertheless, the vaporization process has a high steam
consumption as driving energy.

According to Lara et al. [2] the vapor compression has been used in desalination problems because
of its capacity to treat large volumes of water with a wide range of salt concentrations, nevertheless
it has high operating and capital costs. Moreover, separation problems such as distillation [3] and
evaporation are widely used in chemical industry and the separation problems can represent about
one third of the total energy consumption of the overall process [4]. Therefore, reduce the energy
consumption of the vaporization process can lead to significant savings.

Authors such as Hamed et al. [5] claim that the unique characteristic of vapor compression is the
energy reuse of the vapor generated in the last effect to act as an energy source in the first effect.
There are two different ways to recompress the vapor, thermally in a steam ejector and mechanically
in a compressor. Nhien et al. [3] applied mechanical vapor recompression (MVR) in a mixture of
acetone, methanol and water, from which it was proposed the use of two recompressions integrated
with the distillation system. The saving obtained by the authors achieved values as high as 60% in
energy basis when compared with regular extractive distillation. It is important to highlight that the
pressure ratio was 2.5 for each compressor used.

Several authors applied mechanical vapor recompression for different processes, from desalination
processes [5-9] to chemical industry using distillation process [3]. Some of these articles applied the
energy, exergy and thermoeconomic analyses. In the desalination process [9] obtained that multi
effect evaporation with mechanical vapor compression resulted in lower unit product costs in
comparison of thermal vapor compression, whereas [6] obtained that the later results in lower
destroyed exergy and higher exergy efficiency.

2. Original arrangement and mechanical vapor recompression

Figure 1 indicates a schematic representation of the vaporization process, from which, the liquor
enters as a mixture of sodium hydroxide, sodium chloride in water, represented by stream 0. This
schematic representation is classified by [10] as a typical triple effect diaphragm cell evaporator.
Therefore, this schematic representation will be named in the text as conventional evaporation system.

The diaphragm electrolysis produces a solution with low concentrations of caustic soda and with
impurities. According to [10] a typical concentration of the solution stream in the output of the
diaphragm cell, in mass basis, varies from 10-13% and 10-15% of NaOH and NaCl, respectively. The
concentrations adopted in the present analysis are 10.2% of NaOH and 16.5% of NaCl. For the sake
of simplicity, the components for the crystals removal, such as hydrocyclones and centrifugal systems
were removed from the schematic representation of Figure 2, as well as the heat exchangers and
preheaters for the solution vaporization process. Nevertheless, they were taken into account for mass,
energy and exergy balances.

The stream 1 is the end of the vaporization process, with a concentration of 49.4% of sodium
hydroxide and 2.8% of sodium chloride. Other processes are involved until the final product of 50.4%
of NaOH and 0.9% of NaCl. The pressure and temperature at stage 3, 2, 1 and flash tank are,
respectively: (9.9kPa, 58oC), (30.6kPa, 87.1oC), (98.1kPa, 140oC) and (10.7kPa, 91.62oC).

Ejectors and barometric columns are used to guarantee the vacuum, where steam and liquid water are
used to operate these components. The stream 7 that leaves the evaporation process undergoes
condensation process into liquid water.

2
 Figure 1. Conventional evaporation plant schematic representation

The liquor enters in the process by the evaporator 3, as indicated in Figure 1 (in red), from which the
vapor that comes from the evaporator 2 remove a fraction of the water. In this process, the heat
exchangers and pre-heaters are used to increase the temperature of the solution before it enters the
evaporator. These processes are repeated until the solution reaches the evaporator 1 where the input
energy is provided by the process steam (which is indicated by the stream 5), bought or produced
elsewhere. Moreover, the last effect is the flash tank that does not use heat exchangers, because of
the low temperature and pressure that lead to spontaneous vaporization of the water content in the
solution. Therefore, the energy inputs in the process of vaporization are the enthalpy of steam 5, the
steam to operate the ejectors (stream 6), and the liquid water (stream 4) for the barometric column.

Figure 2 indicates a proposal to reduce the vapor consumption using the mechanical vapor
recompression to increase the temperature and the pressure from the thermodynamic state of the
steam leaving the evaporator (stream 7) to the process steam. This pressure ratio would be extremely
high (P5/P7=37.4) to perform MVR in a single stage. The present solution breaks the process in two
stages, first, from the stream 7 to the state that the steam leaves evaporator 1 (stream 8), which results
in a compression ratio of 9.9. The second compression is from the stream 8 to the process steam state
(stream 5), which results in a pressure ratio of 6.7. The energy input would be electrical current
(compression power) – Wmvr. Since the compressions are in different stages (Figures 3 and 4), the
term ∆Hcool is the intermediate heat transfer to the environment between the stages of compression.
Additionally, the exergy content in ∆Hcool could be used as a source in some process in the factory.

Another option of vapor compression could be the use of thermal vapor compression, but since the
objective is to reduce steam consumption and the flow rates of 7, 8, 9 and 10 are the same order of
magnitude of the process steam flow rate, this technique would not lead to a significant reduction in
the vapor consumption.

3
 Figure 2. Modification using mechanical vapor recompression - MVR

The pressure ratios of 6.7 and 9.9 are still high for regular compressors; therefore, it was adopted that
each compression stage must not exceed a ratio of 2.0. This solution leads to the necessity of the use
of different stages with intermediate cooling. The schematic representation of the compressors that
takes the vapor from the state of the stream 7 to the state of the stream 8 is indicated in Figure 3. The
state after the compression is called cpi and after the cooling rI, where i indicates the number of
compressions and cooling already performed, i=1, 2 and 3 (cp4 is the stream 8).

Figure 3. Mechanical vapor recompression in four stages with intermediate cooling

Using the same process, Figure 4 represents a schematic representation of the recompressions that
take the vapor at state of the stream 7 to the state of the stream 5, which is the process vapor (usually
bought from other plants). At this point depending of the actual disposal of the evaporators it is
possible to connect directly the stream 8 to the compressors, or reuse it in evaporator 2 and compress
the steam of evaporator 1 in compressors 2. The difference m8-m9 should be added to the stream that
enters the compressors 2. This last solution is the one that is indicated in Figure 2.

Figure 4 Mechanical vapor recompression in three stages with intermediate cooling

4
To calculate the thermodynamic states between the stages it was used an isentropic efficiency of 85%
from which it was possible to determine these several thermodynamic states after each compression
and cooling. The cooling had the objective to take the steam to the same temperature before the
compression, or when this temperature is lower than saturation temperature, take to the to the
saturation temperature at the same pressure. This energy can be used as an input to other processes in
the factory, nevertheless, here this exergy is considered as lost since it has not be applied to a process
yet. Table 1 represents these intermediate states.

Table 1. Thermodynamic states of the recompression, cooling process

Comp. from 8 to 5 Comp. from 7 to 8
o
P (kPa) T ( C) P (kPa) T (oC)
cp1 185 219 cp1 17.6 116
r1 185 140 r1 17.6 58
cp2 349 218 cp2 31.6 116
r2 349 140 r2 31.1 70
cp3 659 214 cp3 55.3 130
r3=5 659 163 r3 55.3 84
- - - cp4=8 98.1 145

2.1. Energy and Exergy analysis

The mass, energy and exergy equations can be applied for Figures 1 and 2, considering:
 Steady state operation.
 Heat transfer to environment negligible.
 No pressure drops in pipes; therefore, the pumps power has been neglected.

For Figure 1 it is possible to calculate the destroyed exergy as: Bd = Binput - Boutput. Hence, the exergy
analysis is indicated according to Equation 1.

Bd  B0  B5  B6  B4  B3  B2  B1 , (1)

The exergy efficiency allows a comparison between the performance of different system and
technologies in one single thermodynamic basis. Nevertheless, there is still not a consensus about its
addressing in processes such as a separation, evaporation and others. In the present analysis, the
exergy efficiency is indicated by Equation 2, and can be defined as the ratio of the exergy output to
the exergy input.

Bout B1  B2  B3
η  , (2)
Bin B0  B5  B4  B6

For Figure 2 the destroyed exergy and exergy efficiency are calculated by Equations 3 and 4. For
these equations, the exergy associated with the intermediate cooling, Bcooling, is considered to be
destroyed in the environment or used in some other process in the factory. Moreover, Wmvr is the total
energy used in the vapor compression, from stream 7 to 10.

Bd  Wmvr  B0  (m6  m10 ) B5  B3  B2  B1  Bcooling , (3)

5
 Bout B1  B2  B3  Bcooling
η  , (4)
Bin B0  m5  m10 B5  Wmvr

To apply Equations 1 to 4 it is necessary to calculate the total exergy of each stream. For the vapor,
it is possible to calculate accordingly to Equation 5, taking into account physical (bph) and chemical
(bch) exergy. The water was considered as liquid state in the reference environment.

Bvap  mvapbch,water  mvapb ph,water  mvap bch,water  h  h0  T0 s  s0  , (5)

Considering the solution of sodium hydroxide, sodium chloride in water, Equation 6 can be used to
calculate de physical exergy of the solution taking into account the presence of crystals, when
necessary.

  T 
B ph,sol  msolcsol  mcryst ccryst T  T0  T0 ln   , (6)
  T0 

From the enthalpy of the solution it was possible to calculate the specific heat csol, considering that in
the reference state (0oC) the enthalpy of the solution and crystals are considered zero. Thus, Equation
(7) demonstrates the determination of specific heat of the solution (NaOH, NaCl and H2O).

H sol  H sol,ref H sol

c sol   , (7)
Tsol  Tsol,ref Tsol

The crystals are analyzed in a similar manner, in which the specific heat can be obtained by the
expression ccryst  H cryst /Tcryst .
The chemical exergy of the solution is represented by the two terms in Equation 8, where the first is
the standard chemical exergy [11]. In this equation, xi is the mole fraction; γi is activity coefficient of
the solution and the index i represents: water, sodium chloride and sodium hydroxide. These
coefficients of the different components were obtained in [11, 12, 13]. For sodium chloride and
caustic soda it was adopted as the activity coefficient the value of 0.6, and as a recommended by [11],
for water it was adopted 1.0.

Bch,i  i i xi bch,i  RT0 xi ln γi xi  ,

m
(8)
Mi
The total exergy of the solution is the summation of physical and chemical exergy indicated
respectively by Eq. (6) and (7). Therefore: Bi  Bch,i  B ph,i .
Regarding the energy and exergy indicator, it was proposed three mathematical relations to compare
the MVR with project condition. The first is a traditional index used in the field of sodium chloride,
which is the ratio of mass of vapor consumed to produce certain mass of pure NaOH in the product
(cmass). Based on this index, two other indicators were proposed accordingly to Eq. (9) and (10), but
in energy and exergy basis. Equation 9 is also traditionally used in the caustic soda production sector.

Econsumed
Cenergy  , (9)
mNaOH , product
Bconsumed
Cexergy  , (10)
mNaOH , product
6
3. Results and discussion
3.1. Evaporation unity
Table 2 indicates the exergy inlet and outlet of the project conditions and the modification using
mechanical vapor recompression – MVR. The value of the exergy inlet in the project condition, which
is 42% higher than the mechanical vapor recompression, is related to the high consumption of water
in the barometric column. The destroyed exergy in the project condition is higher than using MVR;
nevertheless, the percentage of the exergy input that is destroyed is lower in the conventional
condition.

Table 2. Exergy input, exergy output and destroyed exergy for the conventional condition and MVR
Project MVR
MW % MW %
Exergy input 81.1 100 34.1 100
Exergy output 69.0 85.0 22.5 66
Destroyed exergy 12.2 15.0 11.6 34

In table 3 it is considered the values of mass and energy consumption of the project condition and
using MVR. It is important to highlight that despite the power of compression be higher than the
exergy variation of the process steam (Table 4), there is a reduction in the energy consumption of the
vaporization process. The steam consumption would decrease from 38.5MW to approximately
14.4MW of electricity. The exergy content of the steam in the ejector is 263.8kW. From Table 3 it is
possible to conclude that there is a significantly reduction of steam consumption and if the price of
the electricity is lower than the price of the steam, there would be a considerable reduction in
evaporation costs. Nevertheless, regarding the evaporation unity as control volume, the exergy used
in mechanical vapor recompression is higher than in the project condition.

Table 3. Comparison in mass and energy basis for the conventional configuration and MVR
proposal
Conventional MVR
∆H (MW) ∆H (MW)
Process steam 38.5 3.12
Steam to ejector 0.96 -
Liquid water to ejectors - -
Intermediate cooling - 11.5
Compression power - 14.3

In Table 4 it is demonstrated the difference between the steam consumption (process and ejectors) in
the original configuration of the evaporator and using the mechanical vapor recompression (MVR);
and it is also demonstrated the electricity consumption in the new configuration. This last decreases
the energy consumption of evaporation plant in 53% (modification to use electricity instead of vapor).
From the exergy analysis point of view, the steam exergy variation is 12MW, whereas the
compression power is 14MW. In other words, if it is considered only the frontier of the evaporation
unity as control volume, from the exergy analysis it is possible to conclude that the MVR has higher
exergy costs that the original condition.

7
Table 4. Comparison of the energy and exergy consumption for the conventional configuration and
MVR proposal
Project MVR
MW MW
Steam enthalpy variation 39.6 3.1
Steam exergy variation 12.0 0.9
Destroyed exergy 12.2 11.6
Compression power - 14.0
Electricity consumption - 15.6
Energy consumption 39.6 18.7
** Mechanical and electrical compression efficiency of 90%

Other possible solution not studied with details in this article is to use a desuperheater instead of the
last cooling process of the Figure 4. The idea of this process is to increase the enthalpy of the steam
adding some amount of water to the last stream, to guarantee that the specific enthalpy is in state 5,
or the same absolute enthalpy, H5. With this procedure the energy available for the process of
vaporization would be 37.6MW, being necessary only 2% of the original vapor consumption of the
vaporization process. Nevertheless, it would demand some modifications in the heat exchanger in the
first effect; therefore, these analyses were not conducted with higher details, bearing in mind that is
a possible solution to decrease more the energy consumption of the vaporization process.

In Table 10 it is indicated the ratio of the indicators in mass (tvapor/tsoda), energy (kJenergy/kgsoda) and
exergy (kJexergy/kgsoda) basis in the original configuration and considering MVR. The indexes are
defined in Eqs. (9) and (10). The first two columns indicate that the mass and energy indexes would
improve if the MVR were used. Nevertheless, the same indicator, but in exergy basis shows that the
original configuration consume less exergy to produce the same amount of NaOH.

Table 5. Consumption indicator for the factory

Ratio cmass cenergy cexergy
MVR/Original 0.08 0.49 1.39

3.1. Effect of the energy generation

As indicated in [14] the Brazilian energy matrix has about 82% of its basis of hydro-electric power,
the remain of the energy source is biomass and natural gas (10%), nuclear power (3%) and others.
The same authors considered the conversion efficiency in a hydroelectric plant as 82%, whereas a
cogeneration system using natural gas has a conversion efficiency of 45%. With these values, it is
possible to make a comparison between the energy and exergy necessary to produce electricity
(considering the energy source as hydroelectric or as natural gas), and the energy and exergy
consumption of a boiler operated with natural gas as fuel (energy efficiency of 90% and exergy
efficiency of 30%).

Table 6 indicates the results of energy and exergy consumption to produce the steam at the condition
of the evaporator. For electrical energy (hydro and natural gas), it also considered the exergy
necessary to produce the surplus of steam to satisfy the evaporator demand since MVR do not attend
all the requirements as indicated by Table 3. From Table 6 it is possible to infer that the exergy costs
considering the production of the steam is lower using mechanical vapor recompression. If
transmissions lost were 16% for hydroelectric generation [14], the exergy required to evaporation

8
process would be 24 MW. Accordingly, considering the production chain of the electricity and steam,
the MVR would be the best solution from the energy and exergy point of view. Nevertheless, this
technique must be recommended in countries where a considerable part of the electricity matrix are
obtained from hydro plants, nevertheless cogeneration system with natural gas would also have lower
exergy cost if compared with steam production in the factory.

Table 6. Energy and exergy consumption of the factory considering the electricity and steam
production
Energy (MW) Exergy (MW)
Boiler with natural gas 42.7 40.0
Electricity production natural gas 34.3 34.0
Electricity production Hydroelectric 20.9 20.6

4. Conclusions
In this article, mechanical vapor recompression technique was analyzed as an attempt to reduce steam
consumption in an evaporator of sodium hydroxide and sodium chloride solution in water. It was
performed the energy and exergy analysis of the evaporation system considering as a control volume
the factory and the production chain of electricity, which is distinguished feature of this article. From
the range analyzed, it is possible to conclude that:
 Mechanical vapor recompression saves 50% of the total energy consumption of the factory.
The steam consumption of the solution reduces to 8% if compared with the consumption of
the original configuration.
 From the exergy analysis point of view, the compression power is higher than the exergy
variation of the steam, indicating that this modification has higher exergy costs to produce the
same amount of caustic soda. Nevertheless, if the performance of production and transmission
of electricity and steam were included in the analysis, the exergy costs of MVR would be
lower than the conventional condition. This results highlights that, from the exergy point of
view, MVR is a better solution.

Acknowledgments

The authors would like to acknowledge UNIPAR-Carbocloro S.A. for the financial support and to
allow the publication of the data presented in the article.

Nomenclature
b specific exergy, kJ/kg
B exergy flow rate, kW
c specific heat, kJ/(kg K)
C mass, energy and exergy index, kg/kJ, kJ/kJ
h specific enthalpy, kJ/kg
H enthalpy flow rate, kW
m mass flow rate, kg/s
P pressure (kPa)
T temperature, K
x mole fraction, -

9
 Greek symbols
η efficiency, -
γ activity coefficient, -
Subscripts and superscripts
cool cooling
ch chemical
MVR mechanical vapor recompression
ph physical
sol solution

References
[1] Brum da Silva I.M.C., Hidróxido de Sódio, Revista Virtual de Química, Vol 4, Num 1, 2012.
[2] Lara J.R., Osunsan O., Holtzapple M. T., Advanced Mechanical Vapor-Compression Desalination
System, Desalination, Trends and Technologies, Michael Schorr (Ed.), InTech, 2011 DOI:
10.5772/14711.
[3] Le Cao Nhien G.K., Andika R., Husnil Y.A., Lee M., Application of Mechanical Vapor
Recompression to Acetone–Methanol Separation. International journal of chemical engineering and
application 2014;5(3):216-218.
[4] Linnhoff B., Dunford H., Smith R., Heat integration of distillation columns into overall processes.
Chemical Engineering Science 1983;38(8):1175-1188.
[5] Hamed O.A., Zamamiri A.M., Aly, S., Lior, N. Thermal performance and exergy analysis of a
thermal vapor compression desalination system. Energy conversion and management 1996;37(4):
379-387.
[6] Nafey A.S., Fath H.E.S., Mabrouk A.A., Thermoeconomic design of a multi-effect evaporation
mechanical vapor compression (MEE–MVC) desalination process. Desalination 2008;230(1):1-15.
[7] Mabrouk A.A., Nafey, A.S., Fath, H.E.S. Analysis of a new design of a multi-stage flash–
mechanical vapor compression desalination process. Desalination 2007;204(1):482-500.
[8] Ettouney H., Design of single-effect mechanical vapor compression. Desalination 2006;190(1):
1-15.
[9] Al-Juwayhel F., El-Dessouky H., Ettouney H., Analysis of single-effect evaporator desalination
systems combined with vapor compression heat pumps. Desalination 1997;114(3):253-275.
[10] Rutherford J., Ver Hoeve, R.W., U.S. Patent No. 4,459,188. Washington, DC: U.S. Patent and
Trademark Office 1984.
[11] Szargut J., Morris D.R., Steward F.R., Exergy analysis of thermal, chemical, and metallurgical
processes. Nova York: Hemisphere Publishing Corporation, 1988.
[12] Campbell A.N., Bhatnagar O.N., Osmotic and Activity Coefficients of Sodium Hydroxide in
Water from 150 to 250oC. J. Chem. Eng. Data 1984;29:166-168
[13] Pitzer K.S., Pelper J.C., Busey R.H., Thermodynamic properties of aqueous sodium chloride
solutions. J. Phys. Chem. Ref. Data, 1984;13(1), 1-102.
[14] Flórez-Orrego D., Silva J.A., Oliveira Jr S.D. Renewable and non-renewable exergy cost and
specific CO2 emission of electricity generation: The Brazilian case. Energy Conversion and
Management 2014;85:619-629

10

View publication stats