Sodium Hydroxide Production From Seawater Desalination Brine: Process Design and Energy Efficiency

Subscriber access provided by UNIV OF DURHAM
Sustainability Engineering and Green Chemistry

Sodium hydroxide production from seawater
desalination brine: process design and energy efficiency
Fengmin Du, David Warsinger, Tamanna Urmi, Gregory P Thiel, amit Kumar, and John H Lienhard
Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b01195 • Publication Date (Web): 18 Apr 2018
Downloaded from http://pubs.acs.org on April 18, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,

Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 37 Environmental Science & Technology
1 Sodium hydroxide production from seawater

2 desalination brine: process design and energy efficiency
3 Fengmin Dua, David M. Warsingera, Tamanna I. Urmia, Gregory P. Thiela, Amit Kumara, and
4 John H Lienhard Va*
a
5 Rohsenow Kendall Heat Transfer Laboratory, Department of Mechanical Engineering Massachusetts Institute of
6 Technology, 77 Massachusetts Avenue, Cambridge MA 02139-4307 USA
7 ABSTRACT
8 The ability to increase pH is a crucial need for desalination pretreatment (especially in reverse osmosis) and for
9 other industries, but processes used to raise pH often incur significant emissions and non-renewable resource
10 use. Alternatively, waste brine from desalination can be used to create sodium hydroxide, via appropriate
11 concentration and purification pretreatment steps, for input into the chlor-alkali process. In this work, an
12 efficient process train (with variations) is developed and modeled for sodium hydroxide production from
13 seawater desalination brine using membrane chlor-alkali electrolysis. The integrated system includes
14 nanofiltration, concentration via evaporation or mechanical vapor compression, chemical softening, further ion-
15 exchange softening, dechlorination, and membrane electrolysis. System productivity, component performance,
16 and energy consumption of the NaOH production process are highlighted, and their dependencies on
17 electrolyzer outlet conditions and brine recirculation are investigated. The analysis of the process also includes
18 assessment of the energy efficiency of major components, estimation of system operating expense and
19 comparison with similar processes. The brine-to-caustic process is shown to be technically feasible while offering
20 several advantages, i.e. the reduced environmental impact of desalination through lessened brine discharge,
1
ACS Paragon Plus Environment

Environmental Science & Technology Page 2 of 37
21 and the increase in the overall water recovery ratio of the reverse osmosis facility. Additionally, best-use
22 conditions are given for producing caustic not only for use within the plant, but also in excess amounts for
23 potential revenue.
24 INTRODUCTION
25 As the global population grows and economies develop, demands on the world’s fixed fresh water supply are
26 increasing. Both the growing demand and regional water stress—often punctuated by a changing climate—are
27 driving the rising use of seawater desalination to access the 97% of Earth’s water found in the oceans.1 With
28 considerable reductions in energy consumption and cost over the past several decades, the dominant choice for
29 new seawater desalination facilities is reverse osmosis (RO).2
30 Due to the implementation of more energy-efficient pumps and improved membranes, RO processes are
31 coming closer to the thermodynamic minimum energy consumption.3,4 Despite these technological advances,
32 however, seawater RO (SWRO) – and all desalination systems currently in use today – produce a large quantity
33 of concentrated brine that is discharged back to the sea.5 This discharge has been reported to threaten marine
34 ecosystems6 in several ways, including upsetting the physiochemical balance, causing thermal plumes, and
35 amplifying contaminant concentration.7 One class of solutions to this problem involves more extensive brine
36 post-treatment, possibilities for which have been comprehensively reviewed elsewhere.8-10 For instance, zero
37 liquid discharge (ZLD) has been considered as a way to eliminate brine discharge. ZLD, however, presents high
38 energy and capital costs,11 largely as a result of the high concentrations which require energy intensive thermal
39 desalination technologies.12-15 The reuse of the brine to produce useful and valuable chemicals can be a more
40 sustainable solution, although this approach is rarely applied due to the variety of impurities in the brine,
41 leading to complex separation and purification needs.16 Nevertheless, such a solution could limit harmful
42 environmental impacts by reducing (or even eliminating) brine discharge; and this approach could cut plant
43 costs or generate revenue, making fresh water more affordable.17
2
44 In this study, we focus on one example of brine chemical recovery: converting NaCl in brine to NaOH,
45 commonly known as caustic or caustic soda, which can be re-used within the RO facility. NaOH is widely used to
46 increase pH during pretreatment of seawater feed. At higher pH, aqueous boron compounds, toxic to human18
47 and plant19,20 health, exist primarily as charged borate species, which are better rejected by RO membranes.21,22
48 Undesirable heavy metals and hardness, which can cause membrane scaling, can also be precipitated at high
49 pH.23,24 The use of caustic soda can hinder biofouling as well.25 Further, caustic soda is used in the makeup of
50 cleaning solutions for removing organic foulants and scales.26 Converting RO brine to NaOH therefore can
51 benefit both the environment and the financial bottom line by reducing brine discharge while simultaneously
52 supplying in-plant chemical demand. Excess caustic soda produced might even be a profitable side-product of
53 high-capacity seawater desalination plants: world annual consumption of NaOH is constantly growing, from 53
54 million tons in 2002 to over 65 million tons in 2015 and 82 million tons in 2020 (expected).27,28
55 The vast majority of NaOH used in RO plants today is manufactured by the chlor-alkali process, which
56 accounts for 99.5% of caustic production worldwide.28 The process electrolyzes near-saturated NaCl brine,
57 producing caustic soda as well as chlorine and hydrogen gas. Three main variants of the process exist, but the
58 membrane variant, which uses a cation exchange membrane as the separator between catholyte and anolyte, is
59 the most widely used and is considered the best available technology.29,30 Membrane electrolyzers produce
60 caustic soda of 32-35 wt% concentration, but the process requires a very pure feed brine of about 290-310 g/L
61 NaCl concentration.31 The major purity requirements are listed in the following:28,31-33
62 • Hardness ions (Ca2+ + Mg2+) < 0.02 ppm, as Ca2+ reduces current efficiency and both ions increase
63 electrolyzer operating voltage;
64 • Free chlorine (chlorine and its active hydrolyzed forms, i.e. hypochlorite33 < 0.1 ppm, as it damages ion-
65 exchange resin (used for pretreatment) and reduces life of equipment and piping;
66 • Sulfate as Na2SO4 < 4-8 g/L, as they reduce current efficiency of the electrolyzer.
3
67 These stringent purity requirements, particularly on hardness (< 0.02 ppm for Ca2+ and Mg2+), indicate that any
68 RO-brine reused will require considerable treatment before being suitable for the chlor-alkali process.
69 Previous studies have suggested alternative approaches to treat seawater and/or its concentrated brine to
70 obtain NaCl as intermediate product and ultimately to produce NaOH.34,35 Others have focused more on the
71 process of converting brine directly to a chlor-alkali feed without going through the intermediate stage of salt
72 production.36,37 Thiel et al.38 provide a more detailed review of these and other technologies, along with
73 thermodynamic benchmarks for each technology.
74 Beyond these past studies, no comprehensive model has been found in the literature that describes a full
75 system to convert RO brine to caustic soda, including pretreatment and production. By combining the individual
76 components, interrelations between components are to be considered and model parameters are to be chosen
77 and optimized in order to fit the whole system for caustic production directly from seawater. In this work, an
78 Aspen Plus® model for the brine-to-caustic system, including purification and concentration components and
79 membrane electrolysis cell, is constructed. Using the model, system productivity and the energy consumption of
80 each component are determined.
81 In the section “Modeling methodology,” we introduce the overall modeling approach, system physics and the
82 models employed for each component. In the section “System Parameterization, boundary Conditions,” we
83 present and justify the values used to parameterize the components. In the section “Results and discussion,” we
84 discuss the results from the model and the advantages of the process. Finally, the environmental implications of
85 the present process are considered in the last section.
86 MODELING METHODOLOGY
87 In this section, we first introduce the modeling approaches and assumptions made. Thereafter we describe
88 the process chain and its individual components to convert SWRO brine to caustic soda. To achieve the desired
4
89 purification and concentration for the membrane electrolysis, essential pretreatment components are chosen
90 based on their reliability and maturity.
91 MODELING APPROACH AND ASSUMPTIONS
92 Steady-state simulations in Aspen Plus are conducted with the focus on system-level performance of the
93 conversion process. Details at the component-level (i.e., heat and mass transport phenomena, reaction kinetics)
94 are not taken into consideration.
95 The ENRTL-RK model, which is implemented in Aspen, is applied to simulate the non-ideal thermodynamic
96 behavior of brine and other relevant electrolyte streams. This model combines ENRTL (Electrolyte non-random
97 two-liquid) model for the non-ideal electrolyte liquid phase and the Redlich-Kwong (RK) equation of state for the
98 gas phase. Originally developed by Chen39, the ENRTL model is a widely applied property model for process
99 simulation of electrolyte systems with mixed solvent. Song and Chen40 concluded that the ENRTL is preferred for
100 process modeling applications, compared to other models (Pitzer, OLI MSE, UNIQUAC). In Aspen Plus, the
101 improved ENRTL model from Song and Chen40 is implemented.41
102 The following assumptions are made during the modeling:
103 • Seawater is considered as NaCl solution, contaminated with Ca2+, Mg2+ (hardness ions), bromide (Br–)
104 and sulfate ions (SO42–);
105 • Each stream is ideally mixed without temperature, pressure and concentration gradients;
106 • Chemical equilibria of aqueous phase reactions, including electrolyte dissociations and salt
107 precipitations, are reached for every stream. These equilibria are predicted automatically by Aspen.
108 OVERALL PROCESS CHAIN
5
109 Figure 1 shows a block flow diagram with relevant components and streams: the SWRO brine is first purified
110 by nanofiltration (NF) where most sulfate ions and a fraction of hardness ions (Ca2+ and Mg2+) are removed. The
111 NF permeate is concentrated by electrodialysis (ED), and further concentrated by evaporation or mechanical
112 vapor compression up to NaCl saturation, as required by the electrolyzer. The remaining hardness ions in the
113 brine are removed by chemical softening and ion exchange (IX). Finally, the sufficiently pure and concentrated
114 brine is acidified and sent to the membrane electrolyzer to produce the products: NaOH, Cl2, and H2.
115 Nanofiltration (green box in Figure 1) is used to remove sulfate ions. High concentration of sulfates may cause
116 CaSO4 fouling42 in concentration units (ED, Evaporation). This explains why NF is put at the first step in the
117 process chain. In addition, NF lowers Ca and Mg content of the brine,43,44 reducing other fouling concerns.45
118 Electrodialysis (orange box in Figure 1) is chosen as the next step for a primary concentration. The employed
119 NF-ED chain is also suggested by Garriga,36 and ED is also recommended by Casas et. al.17 The feed is split into
120 two streams which are fed into a diluate and a concentrate channel. ED stack transports NaCl (and water) from
121 the former to the latter. For industrial ED systems, the highest concentration achieved is around 200 g/kg (20
122 wt%).46 The diluate outlet is set as 3.5 wt%. This outlet concentration corresponds to normal seawater and
123 allows the ED diluate to be recycled as a feed into the RO plant.
124 Since the membrane cell requires nearly saturated brine, evaporation is used after ED as a final concentration
125 step (pink box in Figure 1). The combination of ED and evaporation might be considerably more economic than
126 evaporation alone, as suggested by Leitz.47
127 Mechanical vapor compression (MVC) is an alternative for evaporation. While the evaporator is simply heat-
128 driven, MVC compresses the vapor produced by evaporation and powers the evaporator using this compressed
129 vapor. With a preheating heat exchanger integrated, the only energy input is the mechanical work associated
130 with the compressor.
6
131 After being concentrated to saturation, the brine goes through the chemical softening stage. Na2CO3 and
132 NaOH are added to precipitate Ca2+ and Mg2+ ions as calcium and magnesium salts. These precipitates are then
133 removed by filtration.
134 Chemical softening achieves a ppm-level hardness level which still does not meet the required purity for the
135 electrolyzer cells.28 Thus, a cation ion-exchange step is used as a final purification step to remove hardness down
136 to ppb-level. The process chain of chemical softening, ion-exchange and electrolysis is a standard process in a
137 typical membrane chlor-alkali plant.28 Before entering the electrolyzer, the purified brine is acidified with HCl.
138 In the membrane electrolyzer, production of Cl2, H2, and caustic soda takes place, resulting in a depleted brine
139 stream that typically contains up to 20 wt% NaCl. The depleted brine from the electrolysis cell must be
140 dechlorinated (light blue component in Figure 1) before being discharged to the environment29 or recycled.28
141 After dechlorination, the depleted brine is fed into a splitter and split into a purge stream (see “Purge stream”
142 in Figure 1, mid left) and a recycled stream. The purge stream leaves the system. We define the ratio of the
143 purged stream and the total stream fed to the splitter as the “purge ratio”. The role of the purge is to prevent
144 impurities (primarily sulfate) from accumulating in the system (see section “Purge ratio”). This is also mentioned
145 by the literature28 as one of the methods to control sulfate. Other possible sulfate removal/controlling methods
146 would be an additional NF stage or the usage of barium salts in chemical softening. However, the former leads
147 to NaCl depletion48 and the latter involves the usage of expensive and toxic barium salts.29 Thus, both options
148 are not favored here.
149 The recycle stream has a brine concentration of around 20 wt%28 and does not require the primary
150 concentration of ED, so it is fed back to the evaporation/MVC component.
151
152 MODELING OF MEMBRANE ELECTROLYZER
7
153 The ion-exchange membrane in the electrolyzer separates the electrolysis cell into anode and cathode
154 chambers. The overall reaction is (R1). On the anode, chloride is oxidized to chlorine via (R2). On the cathode,
155 water is reduced to hydrogen and hydroxide ions according to (R3).

Overall 2NaCl + 2H O 2NaOH + H + Cl (R1)

Anode: 2Cl Cl + 2e (R2)

Cathode: 2H O + 2e H + 2OH (R3)
156
157 A typical membrane electrolysis cell is illustrated in Figure 2. On the anode (blue), chloride is oxidized to
158 chlorine gas, part of which is dissolved in the anolyte.49 The side reaction is water oxidation, producing oxygen.
159 Extents of both reactions depend on the process chlorine efficiency and the anode current efficiency .50 The
160 definition of these efficiencies is elaborated in the Supplementary Information (SI) section “Anode chamber.” On
161 the cathode (red), water is reduced to hydrogen, producing hydroxide ions. Part of them migrates back to the
162 anolyte, reducing the sodium transport by a factor of , the cathode current efficiency (see SI section “Cathode
163 chamber and membrane transport”). Typically, the anode side brine has depletion of NaCl from 26 wt% to 20
164 wt% and the feed caustic concentrates NaCl from 30 wt% to 32 wt%. Some of the 32 wt% caustic is taken as
165 product; the rest is slightly diluted to 30 wt% using deionized water and recirculated.
166 The detailed Aspen modeling for the electrolyzer is elaborated in the SI section “Modeling of membrane
167 electrolyzer in Aspen Plus”. Additionally, model validation is conducted against a set of reference plant data in
168 the literature28, which is described in SI section “Validation of membrane electrolyzer model”. The agreement
169 with literature values shows the accuracy of the electrolyzer model and the applied thermodynamic property
170 model.
171
8
172
173 ENERGY CONSUMPTION
174 The energy consumption of the electrolyzer can be calculated as follows:
Δ% &'()
= ⋅ " = ⋅#⋅ ( 1)

175 The cell voltage, , is assumed constant at 3.2 V, a value taken from reference plant data given in the
176 literature.28 It shall be noted that in real cells, the cell voltage increases with increasing current due to the ohmic
177 loss in the cell and the over-voltage on both electrodes.28
178 MODELING OF OTHER COMPONENTS
179 The modeling of other components used in the process chain, including brine acidifier, ion-exchanger,
180 chemical softening, evaporator, MVC, electrodialysis, nanofiltration, dechlorination and brine purge, are
181 elaborated in the Supplementary Information (SI) section “Modeling of other components in Aspen Plus”.
182 SYSTEM PARAMETERIZATION, BOUNDARY CONDITIONS
183 In this section, the feed brine composition is given as a boundary condition of the model. Parameter values
184 used in each individual component of the Aspen modeling are listed in Table 1. More details including
185 parameters of individual components are given in Table S6 of the SI.
186
187 RESULTS AND DISCUSSION
188 This section introduces the results obtained by this study. First, we evaluate the energy consumption and
189 thermodynamic efficiency of each component and give some suggestions for system improvement. Next, we
9
190 show the dependency of system performance on two important process parameters based on a sensitivity study
191 and justify our chosen parameter values. Then, we use the obtained chemical dosage requirements and energy
192 consumption numbers to estimate the operating expense (OPEX) of the system and compare it with standard
193 chlor-alkali. Last, we discuss the advantages, best-use examples, and challenges of the proposed process.
194 OVERALL PRODUCTIVITY
195 From 17.5 t/h brine feed to the process, the membrane electrolyzer produces 208.4 kg/h of 32 wt% NaOH
196 (66.7 kg/h as equivalent dry product). While 1.9 kg/h of it (around 3 %) should be used internally for the
197 chemical softening and ion-exchange components, the real caustic output amounts to 64.8 kg/h (as dry). A
198 summary of mass flows at each stage can be found in Table S7 of the SI. The model is scalable with mass, thus
199 allowing its application to a different NaOH production rate.
200 ENERGY CONSUMPTION AND SECOND LAW EFFICIENCY OF COMPONENTS
201 In industry, energy costs dominate the chlor-alkali process and most high-concentration processes. For this
202 process chain, the energy consumption, least work, and second law efficiencies of various components are
203 illustrated in Figure 3.
204 Notably, two options are given for the brine saturation component: the evaporator or the MVC. The
205 evaporator consumes 190.8 kW heat (2.94 kWh/kg NaOH) whereas the MVC component consumes 15.75 kW
206 electricity (0.24 kWh/kg NaOH). Regardless of this concentration component, the rest of the system (NF, ED,
207 electrolyzer) consumes about 189.9 kW (2.93 kWh/kg NaOH) electricity.
208
209 The second law efficiency, given by Equation (2), is a quantitative measure for the thermodynamic efficiency
210 of a given process. Equation (2) can be applied for processes including both work and heat input + :56,57
10
'
,, = ( 2)
/
+ -1 − 0 2 ⋅ +
/1
211 where 34 is the ambient temperature, and 35 is the temperature of the heat source (both in K). In the
212 evaporator, we assume the use of 1.2 bar saturated steam (123.5 °C) as the heat source, which is reasonable
213 condition for industrial waste steam. ' is the thermodynamic least work. Its calculation is elaborated in the
214 SI (equation S-10).
215
216 The chlor-alkali electrolysis consumes the most energy in the process. However, it shows good
217 thermodynamic efficiency (68.2 %), limiting potential improvements. Irreversibilities are mainly caused by
218 limited current efficiency and voltage losses (due to overvoltages on the electrodes and the ohmic losses in the
219 solution, the membrane, and the metal hardware).28 Nevertheless, this efficiency assumes that Cl2 and H2 are
220 desired products beside NaOH. If, instead, NaOH is the only product of interest and HCl is produced as side
221 product (by combusting H2 and Cl2), the chlor-alkali electrolysis efficiency decreases to 39.4 % when no energy is
222 recovered from the formation of HCl.
223 The evaporator, as the second highest energy consumer in the process, shows a poor second law efficiency
224 (3.5 %). Better use of the steam produced (i.e., to preheat before the electrolyzer) would increase the efficiency
225 of this component significantly.
226 Overall, improvement of the energy efficiency of brine concentration components represents the most
227 feasible route to significant energy savings. Processes not requiring as much brine concentration as electrolysis
228 could yield very significant energy savings.38,58
229
11
230 SENSITIVITY STUDIES
231 Modeling of the brine-to-caustic process in Aspen allows the variation of certain parameters to reveal their
232 impact on the system performance. The key parameters from the process chain chosen here are the anolyte
233 outlet concentration of the electrolyzer (depleted brine in Figure 2) and the “purge ratio” of the recycling brine
234 (Figure 1, after the dechlorination block). During the sensitivity study, all system parameter values (see Table 1)
235 are kept constant except for the varied parameter.
236
237 ANOLYTE OUTLET CONCENTRATION
238 System performance is heavily influenced by the concentrations of streams related to the recirculation of
239 depleted brine from the electrolyzer back into the concentration stages. A lower anolyte outlet concentration
240 corresponds to a higher conversion of NaCl to NaOH in the electrolyzer cell, and therefore high caustic
241 productivity. However, this requires that the recycled brine must be concentrated more, resulting in a higher
242 absolute energy consumption in the evaporator/MVC component. This high absolute energy consumption is
243 offset by the higher caustic productivity in the electrolyzer. Overall, the electrolyzer productivity wins this
244 competition: the specific energy consumption, normalized by the caustic production is lower at low anolyte
245 outlet concentrations, as shown in Figure 4.
246 It is evident from Figure 4 that by lowering the anolyte concentration from 22 wt% to 18 wt%, the evaporator/
247 MVC component is consuming about 10% less specific energy. Technically, this means a high applied current and
248 a great depletion of the feed NaCl in the electrolysis cell are beneficial for the whole brine-to-caustic process in
249 terms of productivity and energy consumption. However, as shown in Figure 4, a lower limit exists at 18.2 wt%
250 (200 g/L), below which a stable operation of the membrane electrolyzer cell is no longer feasible.28
12
251 Based on the results of this sensitivity study, we chose 19 wt% as the anolyte outlet concentration of the
252 electrolyzer (Table 1) which allows high system productivity at low specific energy consumption while avoiding
253 the danger of unstable operation by providing a safe margin between the operation point and the lower limit.
254
255
256 PURGE RATIO
257 Purging refers to removing a fraction of the depleted brine from the system using a simple splitter, allowing
258 only a remaining fraction to reenter the process (see Figure 1). This prevents impurities, primarily sulfate, from
259 accumulating in the system. To limit the sulfate concentration in the membrane cell under the allowable
260 tolerance, a minimal purge fraction is needed, which is determined based on the sensitivity study here.
261 The purge ratio (removed stream by the splitter divided by total stream entering the splitter) is varied from
262 0.2 to 1. The electrolyzer feed mass flow as well as its sulfate content with respect to the purge ratio is shown in
263 Figure 5.
264 Figure 5 shows that the mass flow into the electrolyzer is decreasing with increasing purge ratio (horizontal
265 axis). As shown, less purge and more recycling leads to higher system productivity. The recycling, however, is
266 limited by the sulfate accumulation in the system. The impurities bromine and chlorate also accumulate, but at
267 lower concentrations and further below tolerances than sulfate. As seen in Figure 5, a purge ratio between 0.3
268 and around 0.45 risks adverse effects (yellow zone) by high sulfate concentration (red line), and a ratio below
269 0.3 risks extreme impact by sulfate (red zone).
270 Based on this sensitivity study, a 0.5 purge ratio is selected for the system-level process, corresponding to a
271 1:1 ratio of recycle and purge streams. According to Figure 5, this purge ensures an electrolyzer feed sulfate
13
272 concentration of about 3 g/L Na2SO4 (in the green zone) which maximizes cell performance while preventing
273 possible adverse effects of sulfate accumulation.
274 OPERATION COSTS
275 With the energy consumption and chemical dosage obtained from the process model, the operational cost
276 (OPEX) of the proposed system can be estimated. The calculations are summarized in Table 2.
277
278 Evidently, energy costs are dominant in the OPEX of the whole process. The membrane electrolyzer, as the
279 primary energy consumer, contributes almost two thirds of the total operation costs. Concentration
280 components ED, evaporator or MVC also has significant energy costs. Nevertheless, if the desalination plant is
281 co-located with a power-plant (this is not uncommon), low grade heat could potentially be diverted to help
282 lower the concentration cost.
283 Besides operating costs, the capital expenditure (CAPEX) also plays a dominant role in such a long-chain
284 process. However, its estimation is highly site-specific, depending on several economic and other factors, and
285 thus having very limited accuracy in a general case. We will not further focus on it but point out that it could be
286 a direction for further studies.
287
288 COMPARISON OF THE PROCESS CHAIN WITH SIMILAR PROCESSES
289 In this section, our model for the process chain is compared against other existing models and the
290 improvements and suggestions for future design are noted.
291 COMPARISON WITH STANDARD CHLOR-ALKALI
14
292 The major difference of the proposed process from the standard chlor-alkali industry lies in the feed stream.
293 The feedstock in chlor-alkali industry is usually highly pure rock salt which requires no concentration step and
294 fewer purification steps. As shown in Table 2, the pretreatment steps (NF, ED, MVC, chemical softening) in the
295 brine-to-caustic process would cost around 0.0588 $ per kilogram caustic soda produced. If caustic soda were
296 produced in excess of the brine-to-caustic needed internally at the RO plant (i.e., to generate extra revenue),
297 this pretreatment cost should be compared with local rock salts (including the availability of rock salts locally,
298 and the cost and quality of available salts) to decide whether it is economically feasible to produce excess
299 saturated NaCl solution from RO brine.
300
301 COMPARISON WITH MELIAN -MARTEL37
302 In Melian-Martel’s model of caustic production from RO brine, instead of controlling the sulfate concentration
303 in the system by purging, barium is used to precipitate sulfate salts. Melian-Martel et al. proposed the following
304 process: Softening-MEE-IX-Electrolysis, where MEE is multi-effect-evaporation. The depleted brine from the
305 electrolyzer is recycled after a dechlorination process. Additionally, they have calculated flows and
306 concentrations in each stage.
307 While the final objective of their study is similar to ours, our process has the following differences/
308 improvements:
309 • Removal of sulfate: In our process, nanofiltration is used to reduce sulfate along with a brine purge,
310 instead of using toxic and expensive barium salts. The sulfate concentration does not need to be
311 lowered to the level of chemical precipitation as membrane cells are somewhat tolerant (purity
312 requirements given in the section “Modeling details”).
15
313 • Method of concentration: Melian-Martel used multi-effect evaporation (MEE) as a single-step
314 concentration while we used ED as primary step and evaporation/MVC as final step. This saves energy.
315 • NaCl concentration: Melian-Martel’s system concentrates the NaCl to 36 wt% in the MEE, and then
316 dilutes it to 30.6 wt% for input to the membrane electrolyzer. In our system, maximal brine
317 concentration is between 26 and 27 wt%. We suggest that high NaCl over-saturation (> 27 wt%)
318 should be avoided as the precipitation of NaCl crystals may cause scaling in evaporators and pipes.
319 Overall, our process is technically more feasible due to the absence of oversaturated brines. Brine purging to
320 remove sulfate is also easy to operate and reduces chemical treatment costs.
321 COMPARISON WITH GARRIGA36
322 The primary difference between the process chain suggested by Garriga and the present work is the addition
323 of brine recycling. Garriga’s process chain is NF-ED-Saturation-Softening-IX-Electrolysis. Saturation is achieved by
324 adding solid NaCl, as in the chlor-alkali industry. However, as salt is not presumed to be available as feedstock,
325 the present process uses evaporation/MVC instead. Our process also has higher productivity and lower effluent
326 volume due to the recycling of depleted brine.
327 ADVANTAGES, BEST-USE CASE, AND CHALLENGE OF THE PROCESS
328 Overall, the brine-to-caustic process investigated has the following advantages:
329 • Reduction of effluent from the RO facility:
330 Overall, this process can reduce up to 29% of the RO facility effluent. The effluents of the process (see
331 Figure 1) are the NF retentate (rate of 12.20 t/h) and the brine purge of (0.26 t/h) which in Figure 1 are
332 the green box right next to RO bring inlet and the green arrow coming out of the teal box of
16
333 dechlorination. The effluents have a higher concentration than the RO brine and can possibly be further
334 concentrated in zero-liquid discharge processes, such as evaporation ponds.
335 • Reusing ED diluate:
336 The ED step not only concentrates the brine, but also can improve the production of pure water thus
337 reducing pure water costs as well. The ED diluate can be fed back to SWRO for increased recovery, since
338 the diluate’s concentration of 3.5 wt% matches regular seawater. As this stream amounts for 26.2% of
339 the RO brine, the actual water recovery of a regular RO plant (50%) would increase to 57.5%.
340 Additionally, since this recycle stream comes from the RO brine, it does not require the costly extra
341 pretreatment in the SWRO system which is necessary for fresh seawater.
342 • Avoiding of concentration-dilution cycle:
343 The proposed process avoids the steps of NaCl and NaOH concentration, transportation and dilution
344 that are standard in chlor-alkali industry and desalination plants.
345 In standard chlor-alkali systems, solid salt is used as the feedstock and mostly originates from
346 energy intensive evaporation and crystallization of an NaCl solution. Yet no energy is recovered when
347 the salt is transported to a chlor-alkali facility and dissolved, resulting a loss of about 2.58 kWh heat per
348 kg NaOH produced.
349 In RO plants, 50 wt% NaOH solution is typically obtained from chlor-alkali facilities and diluted
350 for internal use. However, the 50 wt% solution is produced by evaporating the 30-35 wt% caustic soda
351 from the electrolyzer.28 This fraction of energy input (about 0.70 kWh heat/kg NaOH) is wasted by
352 dilution.
353 • Potential revenue from side products:
354 The side products of the chlor-alkali process, H2 and Cl2, can generate substantial revenue. First, for
355 every kilogram of caustic soda produced, 27.6 grams of hydrogen with a heat value of 0.92 kWh is co-
356 produced. This heat could be directly applied in the evaporation step. Second, a relatively large amount
17
357 of chlorine is produced, namely 0.87 kg per kg NaOH. Assuming a price of 250 $/t61 for chlorine, the
358 potential revenue amounts to 0.22 $/kg NaOH, nearly covering the operation costs listed in Table 2. It is
359 crucial to mention, however, that the chlorine obtained through this process is a crude product
360 (contains O2, H2O, possibly N2 and CO2) that can either be sold for a low price or has to be purified and
361 liquefied yielding additional operation and capital costs.
362 Additionally, besides H2 and Cl2, it is also possible to produce more valuable side products on
363 site, such as sodium hypochlorite, a typical bleaching chemical that is produced from NaOH and Cl2.
364 • Reduction of transportation costs:
365 The onsite production eliminates the transportation costs of NaOH for internal usage in SWRO plants.
366 Additionally, as water is produced in RO facility as well as in the evaporator/MVC component of
367 the brine-to-caustic system, the transportation costs of deionized water for the membrane electrolyzer
368 (see Figure 1) can be saved.
369 The best-use case of the proposed brine-to-caustic process would be an SWRO desalination plant, satisfying
370 some of the following points:
371 • Locations with scarce or expensive rock salt resources, corresponding to high price for standard chlor-
372 alkali feedstock, and therefore making the brine-to-caustic process economically feasible not only for
373 internal usage, but also for producing extra NaOH and Cl2 for revenue.
374 • Remote desalination plants which are far from the nearest chlor-alkali facility, causing high
375 transportation costs.
376 • High restrictions of boron content in the freshwater leading to high internal caustic usage.
377 • Limitations for RO brine discharge or legal benefits for low-impact discharge methods.
378 • Possible co-existing power plants that can supply waste heat for concentration processes.
18
379 • Possible co-existing chemical plants that can utilize the produced chlorine and excess NaOH.
380
381 Nevertheless, several challenges related to the process remain, for example the required know-how for each
382 component in such an integrated process, as well as operation and control issues, especially during start-up and
383 shut-down.
384 ENVIRONMENTAL IMPLICATIONS

385 The energy consumption and environmental impact of desalinated water production has become a major
386 concern, especially due to the relative energy demands and required disposal of concentrated brine. The brine is
387 usually discharged as wastewater, which can create a local imbalance in the ocean’s salinity with negative
388 impacts on marine ecosystems. Additionally, the treatment of water requires the use of caustic soda, of which
389 the production and transportation contributes indirectly to the energy consumption of water production.62
390 In this study, a recovery process for creating caustic soda from desalination brine has been developed focusing
391 on purification requirements and energy efficiency.
392 The present research not only develops a feasible process to produce caustic soda on-site at seawater
393 desalination plants, but also explores the productivity, the energy use of the overall process, and the
394 thermodynamic performance of the individual components. This process, as proposed at small-scale, can
395 produce hundreds of metric tons of NaOH (a dry equivalent) per year, an amount that would fulfill the caustic
396 soda needs of a typical, large-scale SWRO plant.38 Moreover, if our NaOH recovery process was used on all the
397 brine from a large-scale SWRO plant (about 10,000 kt/year), over 35,000 t caustic could be produced for
398 commercial sale. The excess caustic production could be economically attractive, especially for desalination
399 plants in remote locations, or with expensive local salt resources. Additionally, desalination plants using this
400 process for the brine could increase typical water recovery from 50% to around 58%, reduce brine disposal
19
401 volumes by 29%, save on pretreatment costs of feed water, and thus lower the normalized cost of freshwater.
402 Furthermore, our process has a potential to reduce transportation, disposal costs, and emissions involved in
403 brine disposal in other desalination applications that require zero liquid discharge.
404 ACKNOWLEDGEMENTS
405 The authors of this work gratefully acknowledge sponsorship by Cadagua, a Ferrovial subsidiary, through the
406 MIT Energy Initiative. Aspen Plus from AspenTech, Inc. is used in this research. The authors also wish to thank
407 Kishor G. Nayar for his suggestions on electrodialysis as well as Alicia Gómez González and Luis Alberto Letona
408 Cabrida at Cadagua for providing important input parameters and giving practical insights.
409 CONTENTS OF SUPPORTING INFORMATION
410 • Brief introduction of modeling in Aspen Plus
411 • Modeling details of the membrane electrolyzer, validation and parametric study of the model
412 • Modeling details of the pre- and post-treatment components in the brine-to-caustic process chain
413 • Summary of mass flow, temperature and concentration at each stage of the process
414 • Brief introduction of least work analysis
415 NOMENCLATURE
416 Roman Symbols
6 - Water transport number
# C/mol Faraday’s constant (96,485 C/mol)
" A Electric current
% kmol/h Molar flow
20
+ kW Heat flow
3 °C Temperature
V Voltage
kW Work flow
417
418 Greek Symbols
- Cathode current efficiency

7 - Current utilization factor in electrodialysis
,, - Second law efficiency
- Chlorine current (anode current) efficiency
- Process chlorine efficiency
419
420 Subscripts
∞ Ambient
cell Electrolyzer cell
least Least (work)
+ Heat source
421
422 Abbreviations
ED Electrodialysis
ENRTL Electrolyte- Non-random-two-liquid
IX Ion-exchange
MEE Multi-effect evaporation
MVC Mechanical vapor compression
NF Nanofiltration
21
ppb Parts per billion (µg/kg)
ppm Parts per million (mg/kg)
RK Redlich-Kwong (equation of state)
RO Reverse osmosis
SWRO Seawater reverse osmosis
ZDL Zero liquid discharge
423
424
425 REFERENCES
426 (1) Khawaji, D.; Kutubkhanah, I. K.; Wie, J.-M. Advances in seawater desalination technologies.
427 Desalination 2008, 221 (1-3), 47–69.
428
429 (2) Ghaffour, N.; Missimer, T. M.; Amy, G. L. Technical review and evaluation of the economics of water
430 desalination: current and future challenges for better water supply sustainability. Desalination 2013,
431 309, 197–207.
432
433 (3) Semiat, R. Energy issues in desalination processes. Environ. Sci. Technol. 2008, 42 (22), 8193–8201.
434
435 (4) Warsinger, D. M.; Tow, E. W.; Nayar, K. G.; Maswadeh, L. A.; Lienhard, J. H. Energy efficiency of
436 batch and semi-batch (CCRO) reverse osmosis desalination. Water Res. 2016, 106, 272–282.
437
438 (5) Adham, S.; Burbano, A.; Chiu, K.; Kumar, M. Development of a NF/RO knowledge base, California
439 Energy Commission. Public Interest Energy Research Program Report, 2005.
440
441 (6) Roberts, D. A.; Johnston E. L.; Knott, N. A. Impacts of desalination plant discharges on the marine
442 environment: A critical review of published studies. Water Res. 2010, 44, 5117–5228.
443
444 (7) Lattemann, S.; Höpner, T. Environmental impact and impact assessment of seawater desalination.
445 Desalination 2008, 220 (1-3), 1–15.
446
447 (8) Pérez-González, A.; Urtiaga, A. M.; Ibáñez, R.; Ortiz, I. State of the art and review on the treatment
448 technologies of water reverse osmosis concentrates. Water Res. 2012, 46, 267–283.
449
22
450 (9) Rodríguez-DeLaNuez, F.; Franquiz-Suárez, N.; Santiago, D. E.; Veza, J. M.; Sadhwani, J. J. Reuse and
451 minimization of desalination brines: a review of alternatives. Desalin. Water Treat. 2012, 39, 137–
452 148.
453
454 (10) Xu, P.; Cath, T. Y.; Robertson, A. P.; Reinhard, M.; Leckie, J. O.; Drewes, J. E. Critical review of
455 desalination concentrate management, treatment and beneficial use. Environ. Eng. Sci. 2013, 30 (8),
456 502–514.
457
458 (11) Tong, T.; Elimelech, M. The global rise of zero liquid discharge for wastewater management: drivers,
459 technologies, and future directions. Environ. Sci. Technol. 2016, 50 (13), 6846–6855.
460
461 (12) Warsinger, D. E.; Swaminathan, J.; Lienhard, J. H. Effect of module inclination angle on air gap
462 membrane distillation. Proceedings of the 15th International Heat Transfer Conference 2014.
463
464 (13) Rezaei, M.; Warsinger, D. M.; Lienhard, J. H.; Samhaber, W. M. Wetting prevention in membrane
465 distillation through superhydrophobicity and recharging an air layer on the membrane surface. J.
466 Membrane Sci. 2017, 530, 42–52.
467
468 (14) Warsinger, D. M.; Swaminathan, J.; Morales, L. L.; Lienhard, J. H. Comprehensive condensation flow
469 regimes in air gap membrane distillation: visualization and energy efficiency. J. Membrane Sci. 2018
470 (in press) https://doi.org/10.1016/j.memsci.2018.03.053
471
472 (15) Warsinger, D. E.; Swaminathan, J.; Maswadeh, L. A., Lienhard, J. H. Superhydrophobic condenser
473 surfaces for air gap membrane distillation. J. Membrane Sci. 2015, 492, 578–587.
474
475 (16) Van der Bruggen, B.; Lejon, L.; Vandecasteele, C. Reuse, treatment, and discharge of the concentrate
476 of pressure-driven membrane processes. Environ. Sci. Technol. 2003, 37 (17), 3733–3738.
477
478 (17) Casas, S.; Aladjem, C.; Cortina, J.; Larrotcha, E.; Cremades, L. Seawater reverse osmosis brines as a
479 new salt source for the chlor-alkali industry: integration of NaCl concentration by electrodialysis.
480 Solvent Extr. Ion Exc. 2012, 30 (4), 322–332.
481
482 (18) Murray, F. J. A human health risk assessment of boron (boric acid and borax) in drinking water.
483 Regul. Toxicol. Pharm. 1995, 22 (3), 221–230.
484
485 (19) Lenntech BV, Desalination Post-treatment: Boron Removal Process.
486 http://www.lenntech.com/processes/desalination/post-treatment/post-treatments/boron-
487 removal.htm (Revised on 2017.07.04).
488
489 (20) Koç, C.; Effects on environment and agriculture of geothermal wastewater and boron pollution in
490 Great Menderes Basin. Environ. Monit. Assess. 2007, 125 (1), 377–388.
491
23
492 (21) Magara, Y.; Kawasaki, M.; Sekino, M.; Yamamura, H. Development of reverse osmosis membrane
493 seawater desalination in Japan. Water Sci. Technol. 2000, 41 (10-11), 1–8.
494
495 (22) Chua, K.; Hawlader, M.; Malek, A. Pretreatment of seawater: results of pilot trials in Singapore.
496 Desalination 2003, 159 (3), 225–243.
497
498 (23) Rahardianto, A.; Gao, J.; Gabelich, C. J.; Williams, M. D.; Cohen, Y. High recovery membrane
499 desalting of low-salinity brackish water: integration of accelerated precipitation softening with
500 membrane RO. J. Membrane Sci. 2007, 289 (1), 123–137.
501
502 (24) Valavala, R.; Sohn, J.; Han, J.; Her, N.; Yoon, Y. Pretreatment in reverse osmosis seawater
503 desalination: a short review. Environmental Engineering Research 2011, 16 (4), 205–212.
504
505 (25) Lin, H.-W.; Rabaey, K.; Keller, J.; Yuan, Z.; Pikaar, I. Scaling-free electrochemical production of caustic
506 and oxygen for sulfide control in sewers. Environ. Sci. Technol. 2015, 49 (19), 11395–11402.
507
508 (26) Nitto Hydranautics, Foulants and cleaning procedures for composite polyamide RO membrane
509 elements. Technical Report Nitto Hydranautics, 2011.
510
511 (27) Global consumption of caustic soda to reach 82-MT by 2020.
512 https://www.oceanicpharmachem.com/global-consumption-of-caustic-soda-to-reach-82-mt-by-
513 2020 (Revised on 2017.07.04).
514
515 (28) O’Brien, T. F.; Bommaraju, T. V.; Hine, F. Handbook of Chlor-Alkali Technology. Springer, 2005.
516
517 (29) Brinkmann, T.; Santonja, G. G.; Schorcht, F.; Roudier, S.; Sancho, L. D. Best available techniques
518 (BAT) reference document for the production of chlor-alkali. Technical Report European
519 Commission, Joint Research Centre, 2014.
520
521 (30) EuroChlor, Chlorine industry review 2014-2015: Maintaining momentum in uncertain times.
522 Technical Report EuroChlor.
523
524 (31) Bommaraju, T. V.; Lüke, B.; O’Brien, T. F.; Blackburn, M. C. Chlorine. Kirk-Othmer encyclopedia of
525 chemical technology, 2004.
526
527 (32) Schmittinger, P.; Florkiewicz, T.; Curlin, L. C.; Lüke, B.; Scannell, R.; Navin, T.; Zelfel, E.; Bartsch, R.
528 Chlorine. Ullmann’s Encyclopedia of Industrial Chemistry, 1986
529
530 (33) O’Brien, T. Dechlorination of brines for membrane cell operation. In Modern Chlor-Alkali
531 Technology, pp. 251–270, Springer, 1990.
532
24
533 (34) Kobuchi, Y.; Terada, Y., Tani, Y. The first salt plant in the middle east using electrodialysis and ion
534 exchange membranes. In Sixth International Symposium on Salt ll 1983 (pp. 541–555).
535
536 (35) Al-Mutaz, I. S.; Wagialla, K. M. Techno-economic feasibility of extracting minerals from desalination
537 brines. Desalination 1988, 69 (3), 297–307.
538
539 (36) Garriga, S. C. Valorization of brines in the chlor-alkali industry. Integration of precipitation and
540 membrane processes. PhD thesis, Universitat Politècnica de Catalunya, 2011.
541
542 (37) Melián-Martel, N.; Sadhwani, J. J.; Báez, S. O. P. Saline waste disposal reuse for desalination plants
543 for the chlor-alkali industry: The particular case of pozo izquierdo SWRO desalination plant.
544 Desalination 2011, 281, 35–41.
545
546 (38) Thiel, G. P.; Kumar, A.; Gómez-González, A.; Lienhard, J. H. Utilization of desalination brine for
547 sodium hydroxide production: Technologies, engineering principles, recovery limits and future
548 directions. ACS Sustain. Chem. Eng. 2017, 5 (12), 11147–11162.
549
550 (39) Chen, C.-C.; Evans, L. B. A local composition model for the excess Gibbs energy of aqueous
551 electrolyte systems. AIChE J. 1986, 32 (3), 444–454.
552
553 (40) Song, Y.; Chen, C.-C. Symmetric electrolyte nonrandom two-liquid activity coefficient model. Ind.
554 Eng. Chem. Res. 2009, 48 (16), 7788–7797.
555
556 (41) Aspen Technology Inc., Benefits of multi-solvent NRTL models in Aspen Plus. Technical Report Aspen
557 Technology Inc., 2012.
558
559 (42) Warsinger, D. M.; Tow, E. W; Swaminathan, J.; Lienhard J. H. Theoretical framework for predicting
560 inorganic fouling in membrane distillation and experimental validation with calcium sulfate. J.
561 Membrane Sci. 2017, 528, 381–390.
562
563 (43) Roy, Y.; Warsinger, D. M.; Lienhard, J. H. Effect of temperature on ion transport in nanofiltration
564 membranes: Diffusion, convection and electromigration. Desalination 2017, 420, 241–257.
565
566 (44) Warsinger, D. M.; Servi, A.; Van Belleghem, S.; Gonzalez, J.; Swaminathan, J.; Kharraz, J.; Chung, H.
567 W.; Arafat, H. A.; Gleason, K. K.; Lienhard, J. H. Combining air recharging and membrane
568 superhydrophobicity for fouling prevention in membrane distillation. J. Membrane Sci. 2016, 505,
569 241–252.
570
571 (45) Warsinger, D. M.; Tow, E. W.; Maswadeh, L. A.; Connors, G.; Swaminathan, J.; Lienhard J. H.
572 Inorganic fouling mitigation by salinity cycling in batch reverse osmosis. Water Res. 2018 (in press)
573 https://doi.org/10.1016/j.watres.2018.01.060
574
25
575 (46) Asahi Glass and Chemicals Co., Personal Communication to Kishor G. Nayar, October 2016.
576
577 (47) Leitz, F. B. Electrodialysis for industrial water cleanup. Environ. Sci. Technol. 1976, 10 (2), 136–139.
578
579 (48) Mousavi, S. E.; Moosavian, S. M. A.; Jalali, M.; Zahedi, P.; Karimi, E. Sulfate removal from chlor-alkali
580 brine using nanofiltration: Parameters investigation and optimization via Taguchi design method.
581 Desalin. Water Treat. 2017, 100, 75-90.
582
583 (49) Mirzazadeh, T.; Mohammadi, F.; Soltanieh, M.; Joudaki, E. Optimization of caustic current efficiency
584 in a zero-gap advanced chlor-alkali cell with application of genetic algorithm assisted by artificial
585 neural networks. Chem. Eng. J. 2008, 140 (1-3), 157–164.
586
587 (50) Jalali, A.; Mohammadi, F.; Ashrafizadeh, S. Effects of process conditions on cell voltage, current
588 efficiency and voltage balance of a chlor-alkali membrane cell. Desalination 2009, 237 (1-3), 126–
589 139.
590
591 (51) Walha, K.; Amar, R. B.; Quemeneur, F.; Jaouen, P. Treatment by nanofiltration and reverse osmosis
592 of high salinity drilling water for seafood washing and processing abstract. Desalination 2008, 219
593 (1-3), 231–239.
594
595 (52) Liu, J.; Yuan, J.; Ji, Z.; Wang, B.; Hao, Y.; Guo, X. Concentrating brine from seawater desalination
596 process by nanofiltration–electrodialysis integrated membrane technology. Desalination 2016, 390,
597 53–61.
598
599 (53) Dow Water & Process Solutions, Design Software: ROSA (Reverse Osmosis System Analysis).
600 http://www.dow.com/en-us/water-and-process-solutions/resources/design-software (revised on
601 2017.08.29).
602
603 (54) Dow Water & Process Solutions, ROSA Help File.
604 https://dowac.custhelp.com/app/answers/detail/a_id/205 (revised on 2017.08.29).
605
606 (55) Dow Water & Process Solutions, AmberliteTM selective resins for brine purification in the chlor-alkali
607 industry, Technical Report Dow Water & Process Solutions.
608
609 (56) Lienhard, J. H.; Mistry, K. H.; Sharqawy, M. H.; Thiel, G. P. Thermodynamics, Exergy, and Energy
610 Efficiency in Desalination Systems. In Desalination Sustainability: A Technical, Socioeconomic, and
611 Environmental Approach, Chpt. 4, H. A. Arafat, ed. Elsevier Publishing Co., 2017.
612
613 (57) Swaminathan, J.; Chung, H. W.; Warsinger, D. M.; Lienhard, J. H. Energy efficiency of membrane
614 distillation up to high salinity: Evaluating critical system size and optimal membrane thickness. Appl.
615 Energ. 2018, 211, 715-734.
616
26
617 (58) Kumar, A.; Katherine, G.; Thiel, G.; Schroder, U.; Lienhard, J. H. Direct electro-synthesis of sodium
618 hydroxide and hydrochloric acid from brine streams. Nature Catalysis 2018, in revision following
619 review.
620
621 (59) Electric power monthly with data for October 2017.
622 https://www.eia.gov/electricity/monthly/archive/december2017.pdf (Table 5.6.A., Revised on
623 2018.02.04).
624
625 (60) Natural gas weekly update. https://www.eia.gov/naturalgas/weekly (Revised on 2018.02.04).
626
627 (61) Indicative Chemical Prices A-Z. https://www.icis.com/chemicals/channel-info-chemicals-a-z (Revised
628 on 2018.02.04).
629
630 (62) Santana, M. V.; Zhang, Q.; Mihelcic, J. R. Influence of water quality on the embodied energy of
631 drinking water treatment. Environ. Sci. Technol. 2014, 48 (5), 3084–3091.
632
633 TOC figure
27
Retentate Diluate Distillate Precipitate
Input: Evaporation or Chemical

NF ED
Brine MVC softening
2+ 2+
Green: Purification steps Ca , Mg
IX
Orange: Concentration steps
Depleted brine Cl2

(recycled)
Membrane
Electrolyzer
Deionized water
H2
Product: NaOH
634
635
28
Retentate Diluate 3.5% Distillate Na2CO3, NaOH Precipitate
Permeate Concentrate Saturated

20% Evaporation or 27% Chemical
7 wt% brine NF ~5% ED
MVC softening
2+ 2+
Ca , Mg
Recycled IX
brine
Purge stream HCl
Dechlorination
NaHSO3, HCl Depleted brine Cl2 (O2)
20%
Membrane
Deionized water Caustic soda Electrolyzer H2
30%
Caustic soda
32%
Product NaOH
636
637 Figure 1. Brine to NaOH system block flow diagram for this study. Successive components purify (green) or
638 concentrate (orange) the feed to reach suitable input conditions for a membrane electrolyzer. Typical brine and
639 caustic concentrations are shown (percentage given as wt% solute). Color key of streams: Dark blue = brine
640 streams; light blue = water; red = product; yellow = chemical dosage; green = effluent. Width of arrows indicates
641 the mass flows. Starting from the inlet (left to right), the successive process includes nanofiltration (NF),
642 electrodialysis (ED), evaporation or mechanical vapor compression (MVC), chemical softening, and ion-exchange
643 (IX) as pretreatment components; the last step is the membrane electrolyzer as the NaOH production unit, with
644 dechlorination as a post-treatment step. Components are chosen for purification capabilities (by concentration)
645 and energy efficiency for concentration steps. Membrane electrolysis produce the primary product (NaOH) as
646 well as the gases Cl2 and H2 at the electrodes.
647
29
B: anode current efficiency

@ BC : process chlorine efficiency
@BC B
O2 Cl2
H2
A =: cathode current efficiency
D A
Ion-exchange ?: water transport number
Anode membrane Cathode
– Diluted
Feed brine <1 – BC > H
+ Cl2 (aq) OH
26 wt% NaCl 30 wt% NaOH
e-↑ e-↓
= Na
+
Depleted brine Concentrated

–
20 wt% NaCl <1 – => OH 32 wt% NaOH
Anolyte ?= H2O Catholyte

648
649
650 Figure 2. Schematic diagram of a typical cell for the chlor-alkali process, containing an ion-exchange membrane
651 as well as anode and cathode for the reaction. The ion-exchange membrane separates the anode (left, blue) and
652 the cathode chamber (right, red). Black arrows in the figure show species transported through the membrane
653 and generated at the electrodes. Molar amount of species (blue formulas) are based on 1 mole of electrons (1
654 Faraday) consumed by the electrolyzer.
655
656
30
160
140
Least Work Real system
120
Exergy input [kW] 68.2%
100
80
60
40
20
11.1% 3.5% 14.2%
4.3%
0
Nanofiltration Electrodialysis Evaporator MVC Electrolyzer
657
658 Figure 3. Least work compared to total consumed exergy (denominator in equation (2)) in the nanofiltration,
659 electrodialysis, evaporation, MVC, and electrolyzer components. Percentages given are second law efficiencies
660 (least work divided by actual exergy consumption).
661
662
31
663
664 Figure 4. Evaporator (right, red) and MVC (left, blue) energy consumption normalized by the caustic production
665 amount (as 32 wt% solution) with respect to the anolyte outlet concentration of the electrolyzer. A stable
666 operation of the membrane electrolyzer is no longer feasible with anolyte concentration lower than 18.2 wt%
667 (red zone).
668
32
669
670 Figure 5. Impact of depleted brine purge on brine flow into electrolyzer and sulfate accumulation in the cell.
671 Red and yellow zones indicate that the sulfate concentration exceeds tolerance in most membrane cells (> 8 g/L
672 Na2SO4), or some membrane cells (> 4 g/L Na2SO4), respectively.28,31,32 Green zone is free of adverse effects of
673 sulfate.
674
675 Table 1. Selected modeling parameters and their values of the final system-level process.
Component Parameter Value Unit Source/Rationale
Temperature 25 °C Ambient
Mass fraction H2O 91.87 % Boundary condition.
Feed brine Mass fraction NaCl 6.39 % Corresponds to 0.69 g/L Ca2+, 2.2 g/L Mg2+, 5.3
g/L SO42–, 85 mg/L Br– (based on density of
Mass fraction CaCl2 0.18 %
1059 kg/m3 calculated by Aspen).
Mass fraction MgCl2 0.81 %
33
Mass fraction Na2SO4 0.74 % 2.75 wt% as Na+ is the stoichiometric amount
of typical 7.0 wt% SWRO brine
Mass fraction NaBr 0.01 %
Isothermal, temperature after pumping is

Temperature 26 °C
25.6 °C
Pressure feed 20 bar Typical value for high-salinity feeds36,51,52
Water recovery 30.3 %
Rejection of Cl– 16.1 %

ROSA Simulation53,54: Two-stage system with
Nanofiltration
Rejection of SO42– 95.7 % six NF270-400 elements per stage, see SI
“Nanofiltration”
Rejection of Ca2+ 38.3 %
Rejection of Mg2+ 40.1 %
Rejection of Br– 16.1 % Assumed same as Cl–
Pump efficiency 80 % Estimation
Eletrodialysis Temperature 26 °C Assumed isothermal
(C: Outlet concentration C 20 wt% Industrial upper limit46

concentrate, D:
Outlet concentration D 3.5 wt% Can be fed back to RO plant
diluate)
Brine temperature at 1 bar with saturated

Temperature 108 °C
NaCl
Evaporator
Brine concentration
27 wt% Ensure saturated brine for the electrolyzer
outlet
High enough to speed up the precipitation

Chemical process (source IPPC, cited by Garriga36
Temperature 60 °C
Softening
same as ion exchange
34
Ion exchange Temperature 60 °C Recommended temperature55
Temperature 88 °C Reference plant28
Pressure anolyte 1.09 bar Reference plant28
Pressure catholyte 1.05 bar Reference plant28
See SI “Anode and cathode current

Feed brine pH value 3 -
efficiencies in electrolyzer”
Product NaOH
32 wt% Reference plant28
concentration
Recycle NaOH
30.3 wt% Reference plant28
Electrolyzer concentration
Cathode current 94 % Reference plant28

efficiency
Anode current efficiency 96 % Estimation
Water transport number 4.25 - Reference plant28
Anolyte outlet
19 wt% See section “Anolyte outlet concentration”
concentration
Voltage of the 3.2 V Reference plant28

electrolyzer
Purge splitter Purge ratio 0.5 - See section “Purge ratio”
Dechlorination Temperature 88 °C Adiabatic mixing
676
677
678 Table 2. Estimated operation expenditures of the purposed system. Cost normalized by the caustic production of
679 64.8 kg/h.
35
Component Consumption Cost per unit Cost per produced NaOH
NF 10.9 kWh electricity 0.0695 $/kWh59 0.0117 $/kg
ED 26.8 kWh electricity 0.0695 $/kWh59 0.0288 $/kg
Electrolyzer 152.1 kWh electricity 0.0695 $/kWh59 0.1631 $/kg
0.0102 $/kWh (based on

Evaporator/ 190.8 kWh heat / 0.0301 $/kg
natural gas price60 /
MVC 15.8 kW electricity 0.0169 $/kg
0.0695 $/kWh59
Chemical 16.2 kg/h Na2CO3

0.165 $/kg dry61 0.0014 $/kg
softening (3.6 wt%)
0.154 kg/h HCl

Ion-exchange 0.243 $/kg (dry base)61 <0.0001 $/kg
(6 wt%)
1.356 kg/h HCl

Acidifier 0.243 $/kg (dry base)61 0.0018 $/kg
(37 wt%)
4.321 kg/h NaHSO3

Dechlorination 0.529 $/kg dry61 0.0130 $/kg
(38 wt%)
0.2501 $/kg (Evaporator)

TOTAL
0.2369 $/kg (MVC)
680
36
TOC figure
Retentate Diluate Distillate
Permeate Concentrate Saturated

Evaporation Chemical
RO brine NF ED
/MVC softening
IX
Post-treatment
Depleted brine Cl2 (O2)
Membrane +
Deionized water Electrolyzer

–
H2
Product (NaOH) NaOH

Sodium Hydroxide Production From Seawater Desalination Brine: Process Design and Energy Efficiency

Uploaded by

Copyright:

Available Formats

Sodium Hydroxide Production From Seawater Desalination Brine: Process Design and Energy Efficiency

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sodium Hydroxide Production From Seawater Desalination Brine: Process Design and Energy Efficiency

Uploaded by

Copyright:

Available Formats

Subscriber access provided by UNIV OF DURHAM

Sustainability Engineering and Green Chemistry

is published by the American Chemical Society. 1155 Sixteenth Street N.W.,

1 Sodium hydroxide production from seawater

4 John H Lienhard Va*

ACS Paragon Plus Environment

29 new seawater desalination facilities is reverse osmosis (RO).2

43 costs or generate revenue, making fresh water more affordable.17

63 electrolyzer operating voltage;

73 thermodynamic benchmarks for each technology.

80 each component are determined.

85 the present process are considered in the last section.

90 based on their reliability and maturity.

91 MODELING APPROACH AND ASSUMPTIONS

94 are not taken into consideration.

101 improved ENRTL model from Song and Chen40 is implemented.41

102 The following assumptions are made during the modeling:

104 and sulfate ions (SO42–);

108 OVERALL PROCESS CHAIN

123 allows the ED diluate to be recycled as a feed into the RO plant.

126 evaporation alone, as suggested by Leitz.47

130 with the compressor.

133 removed by filtration.

148 are not favored here.

150 concentration of ED, so it is fed back to the evaporation/MVC component.

152 MODELING OF MEMBRANE ELECTROLYZER

155 water is reduced to hydrogen and hydroxide ions according to (R3).

173 ENERGY CONSUMPTION

174 The energy consumption of the electrolyzer can be calculated as follows:

177 loss in the cell and the over-voltage on both electrodes.28

178 MODELING OF OTHER COMPONENTS

182 SYSTEM PARAMETERIZATION, BOUNDARY CONDITIONS

185 parameters of individual components are given in Table S6 of the SI.

187 RESULTS AND DISCUSSION

194 OVERALL PRODUCTIVITY

199 allowing its application to a different NaOH production rate.

200 ENERGY CONSUMPTION AND SECOND LAW EFFICIENCY OF COMPONENTS

203 illustrated in Figure 3.

207 electrolyzer) consumes about 189.9 kW (2.93 kWh/kg NaOH) electricity.

214 SI (equation S-10).

222 recovered from the formation of HCl.

225 of this component significantly.

228 could yield very significant energy savings.38,58

230 SENSITIVITY STUDIES

235 are kept constant except for the varied parameter.

237 ANOLYTE OUTLET CONCENTRATION

245 outlet concentrations, as shown in Figure 4.

256 PURGE RATIO

269 0.3 risks extreme impact by sulfate (red zone).

273 possible adverse effects of sulfate accumulation.

274 OPERATION COSTS

282 lower the concentration cost.

286 a direction for further studies.

288 COMPARISON OF THE PROCESS CHAIN WITH SIMILAR PROCESSES

290 improvements and suggestions for future design are noted.

291 COMPARISON WITH STANDARD CHLOR-ALKALI

299 saturated NaCl solution from RO brine.

301 COMPARISON WITH MELIAN -MARTEL37

306 concentrations in each stage.

% kmol/h Molar flow

- Cathode current efficiency

,, - Second law efficiency

- Chlorine current (anode current) efficiency

- Process chlorine efficiency