Ion-exchange tap water demineralization for process water preparation results in a saline regeneration wastewater (20–100 mS cm−1) that is increasingly problematic in view of discharge. A coupled nanofiltration–membrane distillation (NF-MD) process is evaluated for the recovery of water and sodium chloride from this wastewater. NF-MD treatment of mixed regeneration wastewater is compared to NF-MD treatment of separate anion- and cation-regenerate fractions. NF on mixed regeneration wastewater results in a higher flux (30 L m−2 h−1 at 7 bar) compared to NF on the separate fractions (6–9 L m−2 h−1 at 30 bar). NF permeate recovery is strongly limited by scaling (50% for separate and 60% for mixed, respectively). Physical signs of scaling were found during MD treatment of the NF permeates but did not result in flux decline for mixed regeneration wastewater. Final salt composition is expected to qualify as a road de-icing salt. NF-MD is an economically viable alternative compared to external disposal of wastewater for larger-scale installations (1.4 versus 2.5 euro m−3 produced demineralized water for a 10 m3 regenerate per day plant). The cost benefits of water re-use and salt recuperation are small when compared to total treatment costs for mixed regenerate wastewater.

INTRODUCTION

Demineralization by ion-exchange (IEX) is frequently used in industry to prepare process water from tap water (Helfferich 1962; Matsuzaki et al. 1987; Oosterom et al. 2000). The wastewater resulting from the regeneration of IEX resins consists of an aqueous mixture of salts, organic matter and metals (Clifford 1999) and has considerable salinity, typically ranging from 20 to 100 mS cm−1 (Gryta et al. 2005). Owing to the cost of external treatment and zero-liquid discharge technologies, IEX regeneration wastewater is typically neutralized and discharged (Mickley 2008). The discharge of brines, however, is continuously becoming more problematic due to increasingly stringent discharge limits. Production facilities discharging to small inland rivers are more affected than sites located near the sea or a brackish water body where discharge of brine may be environmentally acceptable. In specific cases the inability to discharge can become a real bottleneck for a plant.

Next to tightening discharge legislation, an increasing scarcity of freshwater and other resources is witnessed (Angelakis & Durham 2008) resulting in an increasing interest in new technologies and concepts for the treatment of brines (Kim 2011). Depending on brine flow and composition either recovery of inorganic salts and water or minimization of volume for final disposal is desired (Tang & Chen 2002). The objective of this paper is to evaluate a coupled nanofiltration–membrane distillation (NF-MD) process for the recovery of salts and water from demineralization wastewater. Sodium chloride (NaCl) is the main salt component in IEX regeneration wastewater. It has a wide range of industrial uses, for example in sodium carbonate and chlorine production, and can be applied as a road deicing salt. Owing to salt purity requirements, recycled NaCl from demineralization brines is expected to qualify mainly for medium- to low-grade applications in the chemical industry or as a road deicing salt (Fonnesbech 2001). Reuse of sodium chloride from demineralization regenerate requires first the removal of the other impurities (e.g. metals, organic matter (total organic carbon)) that are present in the brine. NF is used to separate the wastewater into a concentrate containing the undesired impurities and a purified, NaCl- permeate stream. MD (El-Bourawi et al. 2006; Alkhudhiri et al. 2012) is subsequently applied to further concentrate the NF permeate while reclaiming distilled water. The final product resulting from NF-MD is highly concentrated brine that can be further processed using conventional crystallization technologies. Application of NF before MD is expected to have additional benefits as NF will also soften the MD feed (Hassan et al. 1998; Karakulski & Gryta 2005; Gryta et al. 2006). The NF concentrate remains an unwanted by-product and is expected to have a composition similar to the feed, but with higher concentrations of the rejected components and a much lower volume (Van der Bruggen et al. 2008).

In this paper we present both the outcome of NF-MD experiments on mixed and neutralized regeneration wastewater (Figure 1(a)) and on separate anion- and cation-regenerate fractions (Figure 1(b)). Both the technical and economic feasibility are evaluated.

Figure 1

(a) Treatment of the mixed and neutralized wastewater. (b) Separate NF treatment of cation- and anion-regenerate fractions.

Figure 1

(a) Treatment of the mixed and neutralized wastewater. (b) Separate NF treatment of cation- and anion-regenerate fractions.

METHODS

Sampling

Samples of cation-regenerate (acid), anion-regenerate (basic) and mixed wastewater (neutral) were collected from a large demineralization plant in the port of Antwerp, Belgium. Cation and anion regenerates were collected during the regeneration process at peak conductivity, while the mixed wastewater was collected from the final neutralization basin prior to discharge. The first batch of samples for screening tests contained 25 L per stream, and the second batch for module tests contained 150 L per stream.

NF

NF tests were carried out using 2.4 inch spiral-wound modules, following preliminary screening on meander cross-flow Amafilter TZA944 test cell (membrane selection: DOW-Filmtec NF270, KOCH TFC SR100, Desal 5 DK, Nadir NP030, KOCH MPF34 SelRO, and GE Duracid). DOW-Filmtec NF270-2540 (2.6 m2) was selected for NF module testing on mixed wastewater, KOCH SelRO MPS34-2540 A2X (1.6 m2) for the anion stream and Duracid NF2540F-30D (1.4 m2) for the cation stream. Conductivity (WTW 340i + Tetracon 325), turbidity (HACH 2100P ISO turbidimeter), pH (Knick Portamess 913 + SE101N), flux, pressure drop, circulation velocity and applied transmembrane pressure (TMP) were recorded during the module tests. Samples were collected for chemical analysis. Membrane modules were operated according to the manufacturers' guidelines. Two types of NF tests were performed: (1) batch module tests in which 25 L of feed was concentrated by increasing removal of permeate from the circulating feed flow; and (2) production tests, in which approximately 150 L of feed was filtered by continuously adding new feed to the feed tank, thereby maintaining the level in the feed tank as permeate was removed. Clean-water flux was evaluated before and after each test. Ion retention (R) was calculated as R = 1 − cp/cf where cf is feed concentration and cp is permeate concentration (Mulder 1996). Data analysis was performed using GraphPad Prism6 and Microsoft Excel. Genesys Membrane Master software was used to predict scaling behavior.

MD

A laboratory-scale direct contact (DCMD) unit was used for concentration of mixed NF permeates resulting from NF production tests. A feed temperature of 80 °C and delta T of 10 °C was used to generate a maximal and sustainable flux within the boundaries of the experimental setup. A standard flat-sheet Donaldson polytetrafluoroethylene (PTFE) membrane with an area of 10 × 30 cm was used. Ten litres of MD feed was concentrated in two steps. Flux, circulation velocity, mass balance, temperature, and TMP were recorded online. Samples were taken for chemical analysis.

Chemical analysis

Chemical analysis was performed by a commercial laboratory, using the following certified methods: discrete analyzer: HCO3, CO32− (WAC/III/A/006), NO3 (Wsaqno2), NH4-N, Cl, SO42−, ortho-PO43− (NEN 6604), total organic carbon (TOC) (WAC/III/D/50), inductively coupled plasma optical emission spectrometry: Na, K, Ca, Mg, Sr, B, Cd, Ni, Zn, Fe, Al (WAC/III/B/010), inductively coupled plasma mass spectrometry: B (WAC/III/B/001), Sb, Ba (CMA/2/I/B.1), ion selective F (CMA/2/I/C1.1 and 1.2), spectrophotometrically silica (SM4500-SiO2).

Cost estimation

The cost of NF-MD was estimated for a plant sized for the production of 20 and 200 m3 day−1 of demineralized water. The following assumptions were made:

  • The IEX resin bed operates at 95% water recovery.

  • Tap water costs 1.5 euro m−3 and electricity 0.1 euro kWh−1 and conforms to local rates.

  • Waste heat 0.003 euro kWh−1.

  • Technology and membrane capital expenses were obtained from technology suppliers.

  • A 10-year depreciation at a rate of 4% conforms to best available technique methodology and 350 working days/year.

  • Brine disposal costs 50 euro ton−1 (<40 g L−1 total dissolved solids (TDS)) and 200 euro ton−1 (>40 g L−1 TDS), including transport and according to local waste-treatment rates.

  • Produced road salt value of 20 euro ton−1.

  • No labor costs were incorporated.

The following input data were obtained from the experimental work described in this paper:

  • NF on mixed regenerate: 60% permeate recovery, 30 L m−2 h−1 flux and 7 bar TMP.

  • MD on mixed regenerate NF permeate at 95% distillate recovery.

  • NF on cation regenerate: 50% permeate recovery, 9 L m−2 h−1 flux and 30 bar TMP.

  • NF on anion regenerate: 50% permeate recovery, 6 L m−2 h−1 flux and 30 bar TMP.

RESULTS AND DISCUSSION

Demineralization process

Wastewater from a large demineralization plant located in the port of Antwerp was used for this study. The plant produces ultra-pure water from tap water by using cation- and anion-exchange resins. The cation-exchange resin is regenerated with 4% HCl, while the anion-exchange resin is regenerated using 3% NaOH. During the first phase of the regeneration process (i.e. 20–25% of total regeneration time), a regenerate with a low salinity is released. A first optimization of the process should be to separately collect this first wash water, reducing the final wastewater volume. Following the first rinsing phase, highly concentrated brine is released from both the cation- and anion-resin columns. The pH of the cation regenerate decreases to pH 0.6 (minimum), and that of the anion regenerate increases to 13.1 (maximum). Both fractions are mixed together in a large collection basin, and pH of the wastewater is adjusted to pH 7 before discharge. Chemical analysis of the regenerate samples (Tables 1 and 2) shows that the cation regenerate contains a relatively high concentration of metals such as aluminum, strontium, barium, and zinc, while the anion regenerate contains high concentrations of TOC. These components are to be removed by NF in order to obtain a reusable NaCl-rich salt.

Table 1

Regenerate composition for NF-MD on separate anion and cation fractions

 Anion regenerate Cation regenerate Mixed NF permeates    
 NF NF MD  Crystallizationa 
 Feed Permeate Feed Permeate Distillate Concentrate  Salt composition 
Qb 2.5 2.5 4.75 0.25 m3 day−1 84 kg day−1 
pH 12.5 12.6 0.63 0.44 7.32 6.85 – Sor.  
Na+ 13,200 3,580 2,680 2,790 48.8 1,17,000 mg L−1 347 g kg−1 
K+ 114 68.8 551 538 4.6 5,280 mg L−1 15.7 g kg−1 
Cl 9,880 2,240 33,400 20,200 2.03 1,83,000 mg L−1 543 g kg−1 
F 3.41 0.7 n/a 0.66 <0.20 19 mg L−1 0.056 g kg−1 
NO3 4,600 2,410 n/a <0.89 2.64 29,700 mg L−1 88.1 g/kg 
Ca2+ 751 <10.0 9,570 130 <10.0 750 mg L−1 2.22 g kg−1 
Mg2+ 38 <0.50 673 3.2 <0.50 27 mg L−1 0.080 g kg−1 
SO42− 4,720 29.9 n/a <3.0 <3.0 1,180 mg L−1 3.50 g kg−1 
CO32− 58.9 4.55 <0.010 <0.010 <0.010 <0.010 meq L−1 – g kg−1 
HCO3 0.19 0.012 <1.00 <0.010 <0.010 <0.010 meq L−1 – g kg−1 
PO43− 0.33 <0.050 n/a <0.050 <0.050 <0.050 mg P L−1 – g kg−1 
NH4+ 1.21 0.87 0.45 n/a 1.14 0.68 mg N L−1 2.02 mg kg−1 
SiO2 <0.20 0.52 22.7 <0.20 0.34 54.7 mg L−1 162 mg kg−1 
Al 0.53 0.023 6.63 0.026 0.063 0.086 mg L−1 0.255 mg kg−1 
Fe 0.33 <0.15 n/a <0.15 <0.15 <0.15 mg L−1 – mg kg−1 
Cd 0.0018 <0.001 0.0509 0.026 <0.001 0.196 mg L−1 0.581 mg kg−1 
Ni 0.029 <0.010 0.228 <0.010 <0.010 0.012 mg L−1 0.036 mg kg−1 
Zn 0.175 <0.025 2.88 2.2 <0.025 5.04 mg L−1 14.9 mg kg−1 
Ba 0.236 <0.050 2.59 0.037 <0.050 0.28 mg L−1 0.83 mg kg−1 
Sr 1.53 <0.050 25.3 0.176 <0.050 2.1 mg L−1 6.23 mg kg−1 
5.33 <0.500 n/a n/a 0.114 9.63 mg L−1 28.6 mg kg−1 
TOC 630 5.7 n/a <2.0 97 mg C L−1 288 mg kg−1 
 Anion regenerate Cation regenerate Mixed NF permeates    
 NF NF MD  Crystallizationa 
 Feed Permeate Feed Permeate Distillate Concentrate  Salt composition 
Qb 2.5 2.5 4.75 0.25 m3 day−1 84 kg day−1 
pH 12.5 12.6 0.63 0.44 7.32 6.85 – Sor.  
Na+ 13,200 3,580 2,680 2,790 48.8 1,17,000 mg L−1 347 g kg−1 
K+ 114 68.8 551 538 4.6 5,280 mg L−1 15.7 g kg−1 
Cl 9,880 2,240 33,400 20,200 2.03 1,83,000 mg L−1 543 g kg−1 
F 3.41 0.7 n/a 0.66 <0.20 19 mg L−1 0.056 g kg−1 
NO3 4,600 2,410 n/a <0.89 2.64 29,700 mg L−1 88.1 g/kg 
Ca2+ 751 <10.0 9,570 130 <10.0 750 mg L−1 2.22 g kg−1 
Mg2+ 38 <0.50 673 3.2 <0.50 27 mg L−1 0.080 g kg−1 
SO42− 4,720 29.9 n/a <3.0 <3.0 1,180 mg L−1 3.50 g kg−1 
CO32− 58.9 4.55 <0.010 <0.010 <0.010 <0.010 meq L−1 – g kg−1 
HCO3 0.19 0.012 <1.00 <0.010 <0.010 <0.010 meq L−1 – g kg−1 
PO43− 0.33 <0.050 n/a <0.050 <0.050 <0.050 mg P L−1 – g kg−1 
NH4+ 1.21 0.87 0.45 n/a 1.14 0.68 mg N L−1 2.02 mg kg−1 
SiO2 <0.20 0.52 22.7 <0.20 0.34 54.7 mg L−1 162 mg kg−1 
Al 0.53 0.023 6.63 0.026 0.063 0.086 mg L−1 0.255 mg kg−1 
Fe 0.33 <0.15 n/a <0.15 <0.15 <0.15 mg L−1 – mg kg−1 
Cd 0.0018 <0.001 0.0509 0.026 <0.001 0.196 mg L−1 0.581 mg kg−1 
Ni 0.029 <0.010 0.228 <0.010 <0.010 0.012 mg L−1 0.036 mg kg−1 
Zn 0.175 <0.025 2.88 2.2 <0.025 5.04 mg L−1 14.9 mg kg−1 
Ba 0.236 <0.050 2.59 0.037 <0.050 0.28 mg L−1 0.83 mg kg−1 
Sr 1.53 <0.050 25.3 0.176 <0.050 2.1 mg L−1 6.23 mg kg−1 
5.33 <0.500 n/a n/a 0.114 9.63 mg L−1 28.6 mg kg−1 
TOC 630 5.7 n/a <2.0 97 mg C L−1 288 mg kg−1 

aSalt composition is derived from MD brine composition.

bQ: flow rate (crystallization: mass flow rate).

NF tests

Cation-regenerate

The NF batch module test on cation regenerate started with an initial flux of 32 L m−2 h−1 at 20 bar TMP and 600 L h−1 circulation velocity (Duracid NF2540F-30D NF). Flux decreased strongly with increasing permeate recovery and occasionally reduced to zero (Figure 2(a)). TMP was stepwise increased up to 45 bar at 75% permeate recovery (PR) while the membrane has a typical operating pressure ranging from 27 bar to 55 bar. Flux remained stable (19 L m−2 h−1) during iso-concentration mode at 25% PR. Flux decline was observed during iso-concentration at higher recovery ratios, being indicative of fouling/scaling. Negative retention was observed for K+ and Na+, while Cl and Mg2+ retention was relatively constant with increasing permeate recovery. Ca2+ exhibited lower retention with increasing permeate recovery. Negative ion retention is commonly found in NF on complex mixtures, caused by the effect of the Donnan distribution of the salt between the solution and the membrane (Gilron et al. 2001). Owing to the decrease in calcium retention at high permeate recoveries, it was decided to run subsequent NF production tests at 50% permeate recovery. A clean-water test was performed before and after the batch module test. A clean-water flux decline of 14.5% was witnessed indicating loss of membrane permeability and possible fouling/scaling issues.

Figure 2

Flux decline and ion retention for NF module tests on (a) cation regenerate, (b) anion regenerate and (c) mixed regenerate.

Figure 2

Flux decline and ion retention for NF module tests on (a) cation regenerate, (b) anion regenerate and (c) mixed regenerate.

Anion regenerate

The batch module test on anion regenerate started with an initial flux of 9 L m−2 h−1 at a TMP of 30 bar and 500 L h−1 circulation velocity (KOCH SelRO MPS34-2540 A2X). Flux gradually decreased with increasing permeate recovery. A strong flux decline was observed at 75% permeate recovery; a representative sample for 88% permeate recovery could not be obtained (Figure 2(b)). The carbonate ion displays negative retention. For nitrate, chloride and sodium a strong decrease in retention with increasing permeate recovery is observed, while sulfates are retained almost completely for all permeate recoveries. Despite the low flux (2–4 L m−2 h−1) at 50 and 75% permeate recovery it was decided to run the subsequent production test at 75% in order to obtain a sufficiently high permeate concentration of NaCl. A clean-water test was performed before and after the batch module test. A clean-water flux decline of 12.9% was witnessed. This indicates loss of membrane permeability and possible fouling/scaling issues.

Mixed regenerate

The batch module test on mixed regenerate wastewater started with an initial flux of 37 L m−2 h−1 at 650 L h−1 circulation velocity and 7 bar TMP (DOW-Filmtec NF270–2540). In comparison to the anion- and cation-regenerate wastewater fractions, TMP is low and flux is found not to decline as strongly with increasing permeate recovery. (Figure 2(c)). With increasing permeate recovery, the retention of both monovalent and bivalent ions decreases. At 50% permeate recovery, Ca2+ retention was found to drop significantly, though flux remained high at 30 L m−2 h−1. A clean-water test was performed before and after the filtration test. A low clean-water flux decline of 5.2% was witnessed. Scaling projection indicates that the mixed stream has high scaling potential (e.g. BaSO4). The maximum obtainable permeate recovery is therefore limited to 60% while dosing antiscalants. A higher permeate recovery can only be maintained when additional pretreatment is performed.

NF production tests

NF production tests were performed on each regenerate fraction to (1) produce sufficient feed for MD experiments and (2) confirm the results of the previous NF module tests on a larger volume. Production tests on all three streams revealed a similar permeate composition compared to previous module tests. For the cationic and anionic stream, the flux declined by 10% with each run. NF concentration of the anion-regenerate stream resulted in highly viscous slurry. Production tests confirm that cation- and anion-regenerate streams are more difficult to treat by NF compared to the mixed regenerate stream. TMP of 30 bar results in 6 L m−2 h−1 at 50% PR for anion regenerate, and 30 bar TMP is required to generate 9 L m−2 h−1 at 50% PR for the cation regenerate. The higher pressure required for the treatment of the separate streams is attributed to feed composition (higher TDS for separate streams, TOC foulants for anion regenerate stream) and to the characteristics of the membranes used (permeability and retention). According to the manufacturers' specifications the NF270 membrane (mixed stream) produces a flux of 53 L m−2 h−1 under standard test conditions (2,000 ppm MgSO4, 4.8 bar TMP) while the Duracid NF membrane (cation regenerate) produces a flux of 10 L m−2 h−1 only (2,000 ppm MgSO4, 7.6 bar TMP). Different standard test conditions are given for the MPF 34 SelRO membrane (glucose/sucrose solutions 3%/3%) result in 60 L m−2 h−1 at 30 bar TMP.

MD

A 1:1 mixture of cation- and anion-regeneration wastewater NF permeates was prepared and neutralized by adding 5.84 mg L−1 of HCl. MD was performed on this mixture until a concentration factor of 14.4 was reached. During the final batch, a flux of 16.5 L m−2 h−1 was observed. Flux remained constant up to a concentration factor of 10.1 (220 mS cm−1 concentrate electrical conductivity), whereupon it slowly declined to 15.5 L m−2 h−1, indicating scaling. Final concentrate salinity was 250 mS cm−1. MD on the NF permeate of the mixed regenerate resulted in a constant flux of 17 L m−2 h−1. No flux decline was observed until a 26-fold volumetric concentration was obtained. Excessive scaling of the tubing was observed in the last stages of the experiment and the membrane showed signs of abrasion. The electrical conductivity of the final concentrate was 130 mS cm−1. Chemical analysis of the two MD concentrates confirms that the mixture of separate anion and cation NF permeates has lower scaling potential (Tables 1 and 2). Additional pilot testing is required in the case of further up-scaling to confirm the maximal concentration factor and the long-term process stability of MD.

Table 2

Regenerate composition for NF-MD on mixed regeneration wastewater

 NF MD  Crystallizationa 
 Feed Permeate Distillate Concentrate  Salt composition 
10 5.7 0.3 m3 day−1 101 kg day−1 
pH 7.25 7.15 n/a n/a – Sor.  
Na+ 2,410 2,260 51.8 57,600 mg L−1 285 g kg−1 
K+ 81.8 88.1 8.8 2,160 mg L−1 10.7 g kg−1 
Cl 5,460 4,780 80 1,23,000 mg L−1 610 g kg−1 
F n/a n/a <0.20 5.6 mg L−1 0.028 g kg−1 
NO3 256 218 <0.89 5,590 mg L−1 27.7 g kg−1 
Ca2+ 1,270 530 <10.0 12,300 mg L−1 61 g kg−1 
Mg2+ 142 17 <0.50 422 mg L−1 2.09 g kg−1 
SO42− 920 9.2 <3.0 220 mg L−1 1.09 g kg−1 
CO32− <0.010 <0.010 <0.010 <0.010 meq L−1 – g kg−1 
HCO3 1.21 <1.00 <1.00 <1.00 meq L−1 – g kg−1 
PO43− n/a n/a <0.15 <0.15 mg P L−1 – g kg−1 
NH4+ 0.16 <0.15 0.19 1.59 mg N L−1 7.88 mg kg−1 
SiO2 88.9 87.65 0.29 72.8 mg L−1 361 mg kg−1 
Al 0.059 0.036 <0.010 0.13 mg L−1 0.645 mg kg−1 
Fe n/a n/a <0.15 <0.15 mg L−1 – mg kg−1 
Cd 0.004 0.0031 <0.0010 0.0113 mg L−1 0.056 mg kg−1 
Ni 0.04 <0.010 <0.010 0.019 mg L−1 0.094 mg kg−1 
Zn 0.246 0.0415 <0.025 0.104 mg L−1 0.516 mg kg−1 
Ba 0.327 0.119 <0.050 2.94 mg L−1 14.6 mg kg−1 
Sr 2.95 0.9995 <0.050 27 mg L−1 134 mg kg−1 
0.638 0.644 <0.017 20.4 mg L−1 101 mg kg−1 
TOC n/a n/a 20 210 mg C L−1 1,041 mg kg−1 
 NF MD  Crystallizationa 
 Feed Permeate Distillate Concentrate  Salt composition 
10 5.7 0.3 m3 day−1 101 kg day−1 
pH 7.25 7.15 n/a n/a – Sor.  
Na+ 2,410 2,260 51.8 57,600 mg L−1 285 g kg−1 
K+ 81.8 88.1 8.8 2,160 mg L−1 10.7 g kg−1 
Cl 5,460 4,780 80 1,23,000 mg L−1 610 g kg−1 
F n/a n/a <0.20 5.6 mg L−1 0.028 g kg−1 
NO3 256 218 <0.89 5,590 mg L−1 27.7 g kg−1 
Ca2+ 1,270 530 <10.0 12,300 mg L−1 61 g kg−1 
Mg2+ 142 17 <0.50 422 mg L−1 2.09 g kg−1 
SO42− 920 9.2 <3.0 220 mg L−1 1.09 g kg−1 
CO32− <0.010 <0.010 <0.010 <0.010 meq L−1 – g kg−1 
HCO3 1.21 <1.00 <1.00 <1.00 meq L−1 – g kg−1 
PO43− n/a n/a <0.15 <0.15 mg P L−1 – g kg−1 
NH4+ 0.16 <0.15 0.19 1.59 mg N L−1 7.88 mg kg−1 
SiO2 88.9 87.65 0.29 72.8 mg L−1 361 mg kg−1 
Al 0.059 0.036 <0.010 0.13 mg L−1 0.645 mg kg−1 
Fe n/a n/a <0.15 <0.15 mg L−1 – mg kg−1 
Cd 0.004 0.0031 <0.0010 0.0113 mg L−1 0.056 mg kg−1 
Ni 0.04 <0.010 <0.010 0.019 mg L−1 0.094 mg kg−1 
Zn 0.246 0.0415 <0.025 0.104 mg L−1 0.516 mg kg−1 
Ba 0.327 0.119 <0.050 2.94 mg L−1 14.6 mg kg−1 
Sr 2.95 0.9995 <0.050 27 mg L−1 134 mg kg−1 
0.638 0.644 <0.017 20.4 mg L−1 101 mg kg−1 
TOC n/a n/a 20 210 mg C L−1 1,041 mg kg−1 

aSalt composition is derived from MD brine composition.

bQ: flow rate (crystallization: mass flow rate).

Process evaluation

The composition of the regenerate following different treatment steps is evaluated based on the outcome of the NF and MD experiments (Tables 1 and 2). Separate treatment of cation- and anion-regenerate fractions results in a lower overall water and salt recovery. The resulting salt composition is not sufficiently pure for reuse in the chemical industry (confirmed by local chlorine production plant) but is expected to qualify as a road de-icing salt. Salt resulting from NF-MD on mixed regeneration wastewater followed by crystallization contains 61% Cl, 28.6% Na+, 6.1% Ca2+, 2.8% NO3 and 0.1% TOC (Table 2) as main constituents. The presence of metals in the final product is low when compared to German standards for deicing salt, for example, with 90 ppm Ni present versus 5,000 ppm allowed, and 60 ppm Cd versus 2,000 ppm allowed. Of concern is the relative high level of nitrates, 28 g kg−1 for mixed wastewater and 80 g kg−1 for separate NF treatment of anion and cation regenerate. Further validation with regulatory agencies is required.

Cost estimation

Cost estimation was performed for two different sized IEX demineralization processes (20 and 200 m3 day−1 of demineralized water produced) and for both proposed process trains, respectively (Figure 1). These plant sizes represent a regeneration wastewater production of 1 and 10 m3 day−1. The lowest flow is representative for small and medium enterprises, whereas the largest flow is typically encountered in larger chemical plants. Cost estimation (Table 3) indicates that separate treatment of the cation and anion regenerate is more expensive when compared to treatment of the mixed stream. NF-MD is not economically viable for small-scale installations, while being more attractive for larger-scale installations compared to external disposal rates (50 euro ton−1). NF-MD on mixed regeneration wastewater results in an additional cost of 1.4 euro m−3 of process water produced for larger-scale installations compared to 2.5 euro m−3 for external disposal (local rate). The cost benefits of water reuse and salt reuse resulting from NF-MD on mixed regeneration wastewater are small when compared to the overall cost (Table 3). The main economical benefit is the reduction in the costs for external treatment.

Table 3

NF-MD cost estimate for different sized demineralization plants

 Mixed regenerate Anion/cation regenerate  
Q demin water 20 200 20 200 m3 day−1 
Q regenerate 10 10 m3 day−1 
Capex NF stage 13.17 1.30 31.56 3.16 euro m−3 regenerate 
Opex NF stage 1.24 0.27 51.68 5.39 euro m−3 regenerate 
Disposal NF concentrate 20.00 20.00 25.00 25.00 euro m−3 regenerate 
Capex MD stage 19.81 2.95 19.81 2.20 euro m−3 regenerate 
Opex MD stage 4.97 3.54 4.94 3.03 euro m−3 regenerate 
Water reuse −0.86 −0.86 −0.71 −0.71 euro m−3 regenerate 
MD brine treatment 1.4 1.4 1.2 1.2 euro m−3 regenerate 
Salt reuse −0.12 −0.12 −0.1 −0.1 euro m−3 regenerate 
Total treatment cost 59.61 28.47 133 39 euro m−3 regenerate 
 Mixed regenerate Anion/cation regenerate  
Q demin water 20 200 20 200 m3 day−1 
Q regenerate 10 10 m3 day−1 
Capex NF stage 13.17 1.30 31.56 3.16 euro m−3 regenerate 
Opex NF stage 1.24 0.27 51.68 5.39 euro m−3 regenerate 
Disposal NF concentrate 20.00 20.00 25.00 25.00 euro m−3 regenerate 
Capex MD stage 19.81 2.95 19.81 2.20 euro m−3 regenerate 
Opex MD stage 4.97 3.54 4.94 3.03 euro m−3 regenerate 
Water reuse −0.86 −0.86 −0.71 −0.71 euro m−3 regenerate 
MD brine treatment 1.4 1.4 1.2 1.2 euro m−3 regenerate 
Salt reuse −0.12 −0.12 −0.1 −0.1 euro m−3 regenerate 
Total treatment cost 59.61 28.47 133 39 euro m−3 regenerate 

Capex: capital expenditure; Opex: operating expenditure.

CONCLUSION

Comparison of NF-MD on mixed regeneration wastewater and NF-MD on separate anion- and cation-regenerate fractions shows that NF-MD on mixed regeneration wastewater is technically and economically preferable. Both wastewater mixtures, however, exhibit severe scaling/fouling potential resulting in limited NF permeate recovery (50 and 60%), which results in lower recovery of salts and water and higher residual NF concentrate volume. Higher permeate recoveries are desired but require additional pretreatment (not evaluated in this work). NF-MD is not economically viable for small-scale installations but can provide an economically viable alternative to external disposal of wastewater for larger-scale installations where a zero liquid-discharge policy needs to be applied. The cost benefits resulting from water and salt recuperation are small when compared to the total treatment cost.

ACKNOWLEDGEMENTS

This study forms part of the Flemish innovation and cooperation project ‘The Blue Circle’ (IWT VIS-traject De Blauwe Cirkel), which focuses on concepts for recovery of water and resources from inorganic concentrates. T. Jir˘íc˘ek completed this work under the support of the ‘National Programme for Sustainability I’ LO 1201 and the OPR&DI project ‘Centre for Nanomaterials, Advanced Technologies and Innovation’, CZ.1.05/2.1.00/01.0005.

REFERENCES

REFERENCES
Alkhudhiri
A.
Darwish
N.
Hilal
N.
2012
Membrane distillation: a comprehensive review
.
Desalination
287
,
2
18
.
Clifford
D. A.
1999
Ion-exchange and inorganic adsorption
.
Water Quality and Treatment
4
,
561
564
.
El-Bourawi
M. S.
Ding
Z.
Ma
R.
Khayet
M.
2006
A framework for better understanding membrane distillation separation process
.
Journal of Membrane Science
285
(
1–2
),
4
29
.
Fonnesbech
J. K.
2001
Ice control technology with 20 percent brine on highways
.
Transportation Research Record
1741
(
1
),
54
59
.
Gilron
J.
Gara
N.
Kedem
O.
2001
Experimental analysis of negative salt rejection in nanofiltration membranes
.
Journal of Membrane Science
185
(
2
),
223
236
.
Gryta
M.
Karakulski
K.
Tomaszewska
M.
Morawski
A.
2005
Treatment of effluents from the regeneration of ion-exchangers using the MD process
.
Desalination
180
(
1
),
173
180
.
Gryta
M.
Karakulski
K.
Tomaszewska
M.
Morawski
W.
2006
Demineralization of water using a combination of MD and NF (RO)
.
Desalination
200
(
1
),
451
452
.
Hassan
A. M.
Al-Sofi
M. A. K.
Al-Amoudi
A. S.
Jamaluddin
A. T. M.
Farooque
A. M.
Rowaili
A.
Dalvi
A.
Kither
N. M.
Mustafa
G. M.
Al-Tisan
I. A. R.
1998
A new approach to membrane and thermal seawater desalination processes using nanofiltration membranes (part 1)
.
Desalination
118
(
1
),
35
51
.
Helfferich
F. G.
1962
Ion-Exchange
.
McGraw-Hill
,
London
.
Karakulski
K.
Gryta
M.
2005
Water demineralisation by NF/MD integrated processes
.
Desalination
177
(
1
),
109
119
.
Matsuzaki
H.
Kubota
M.
Ebara
K.
Takahashi
S.
Kuroiwa
M.
1987
Apparatus for producing ultra-pure water by ion-exchange resin. US Patent 4,698,153
.
Mickley
M.
2008
Survey of High-Recovery and Zero Liquid Discharge Technologies for Water Utilities
.
Water Reuse Foundation
,
Alexandria, VA, USA
.
Mulder
M.
1996
Basic Principles of Membrane Technology, 2nd edn
,
Kluwer Academic
,
Dordrecht, The Netherlands
.
Van der Bruggen
B.
Mänttäri
M.
Nyström
M.
2008
Drawbacks of applying nanofiltration and how to avoid them: a review
.
Separation and Purification Technology
63
(
2
),
251
263
.