- Email: [email protected]
Removal of boron from water by using reverse osmosis
Separation and Purification Technology 64 (2008) 141–146
Contents lists available at ScienceDirect
Separation and Purification Technology journal homepage: www.elsevier.com/locate/seppur
Removal of boron from water by using reverse osmosis Yunus Cengeloglu a,∗ , Gulsin Arslan a , Ali Tor b , Izzet Kocak a , Nesim Dursun c a
Selcuk University, Department of Chemistry, 42031 Campus, Konya, Turkey Selcuk University, Department of Environmental Engineering, 42031 Campus, Konya, Turkey c Selcuk University, Department of Soil, 42031 Campus, Konya, Turkey b
a r t i c l e
i n f o
Article history: Received 30 July 2008 Received in revised form 12 September 2008 Accepted 16 September 2008 Keywords: Boron removal Reverse osmosis Membrane
a b s t r a c t The removal of boron from water was investigated by using reverse osmosis (RO) technique with SWHR, BW-30 (FILMTEC) and AG (GE Osmonics) membranes, not considered previously. The effect of pH and concentration of the feed water and operating pressure on the boron rejection was investigated. The experimental results indicated that boron rejection mostly depends on membrane type, pH of the feed water and operating pressure. The results also showed that boron can be effectively removed only at feed water pH of 11. The lowest permeate boron concentration (the highest rejection, close to 99%) was obtained when SWHR membrane was used. The rejection efficiency of the membranes was found to be in the order of SWHR > BW-30 > AG. For all membranes, boron rejection increased with increasing operating pressure. Finally, two different natural (ground) water samples containing 24.8 and 9.4 mg/L of boron were treated by using RO with SWHR membrane and obtained results showed that RO could be efficiently used (with >95% rejection) for removal of boron from groundwaters. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The potential sources of boron contamination in water resources are either anthropogenic (pollution from sewage effluents, boron enriched fertilizers, landfill leachates and discharge from the soap and detergent manufacturing) or natural (e.g. water–rock interaction, sea water encroachment, mixing with fossil brines or hydrothermal fluids) [1]. Boron in drinking water is suspected as teratogenic. The 1993 WHO guidelines for drinking-water quality proposed a 0.3-mg/L standard for boron in drinking water. Later, this standard was revised to 0.5 mg/L because it would be difficult to achieve in areas with high natural boron levels with the treatment technology available [2]. Recently boron has been classified by the European Union (EU) as a pollutant of drinking water in national and international drinking water directives. The EU adopted a standard of 1 mg/L boron for drinking water [3]. Therefore, it is necessary to remove the excess amount of boron from water by a suitable method. The treatment methods used for the removal of boron from water can be divided into several categories. The first is coagulation and electrocoagulation processes [4–6]. The next two categories are adsorption [7,8] and ion-exchange processes [9–13]. The last cate-
∗ Corresponding author. Tel.: +90 332 223 2772; fax: +90 332 241 0106. E-mail address: [email protected] (Y. Cengeloglu). 1383-5866/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2008.09.006
gory consists of membranes processes such as Donnan dialysis [14], electrodialysis [15] and reverse osmosis (RO) [16–19]. In recent years, membrane manufactures have developed RO membranes with boron rejections of 91–96% [20,21]. However, most of the current desalination plants have to implement the additional treatment steps such as pH adjustment of feed water, post-treatment of RO permeate with ion exchange or several pass stage of permeate in order to improve boron rejection. In addition, several process configurations have been proposed to obtain the low boron concentration of the permeate from RO plant [17,21,22]. The objective of this work was to investigate and compare the boron removal efficiencies of three different RO membranes (SWHR, BW-30 and AG) using model solutions containing boron as single solute. The effect of pH and concentration of feed water and operating pressure on the boron rejection was determined. Finally, under optimal conditions, RO with SWHR membrane was applied to the natural (ground) water samples containing 24.8 and 9.4 mg/L of boron. The obtained permeate boron concentrations were compared with WHO and EU standards and permeate fluxes were evaluated. 2. Materials and methods 2.1. Reverse osmosis pilot plant The reverse osmosis pilot plant (Prozesstechnick GmbH) used in this study consists of a diaphragm pump controlled with a frequency converter (flow range: 1.8–12 L/min, pressure range: max
142
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
Fig. 1. Flow diagram of the reverse osmosis plot plant. (M1 and M2: Membrane housing, B1: Feed tank with heating/cooling jacket, V1 and V2: Emptying valve, V3 and V4: Pressure regulation valve, V5: Spring loaded valve, V6: Three way valve to select which membrane housing, P1: Pump, PI01 and PI02: Pressure gauge, DP1: Differential pressure indicator, LI01: Level indicator on the feed tank, TI01: Temperature indicator.)
40 bar), feed tank with heating/cooling jacket (5 L capacity), membrane housing for both spiral wound and flat-sheet membranes, different emptying and pressure valves. Fig. 1 shows the flow diagram of the described pilot plant. 2.2. Membranes Three membranes having 44 cm2 exposed area with a flat-sheet configuration were studied. The most relevant characteristics of these membranes are summarized in Table 1. 2.3. Experiments The boric acid solutions were prepared in distilled water by diluting the prepared stock solutions (1000 mg/L) to desired concentrations. H3 BO3 was obtained from R.P. Normapur® (Paris, France). NaOH and HCl were obtained from Merck Co. (Darmstad, Germany). All chemicals were the analytic grade reagents. In specific experiments, composition of the feed water and operating pressure were chosen as below:
Configuration Max temperature (◦ C) Max pressure (psig) Salt rejection (%) Chlorine tolerance (ppm)
At the beginning of each experiment, pH of the feed water (2 L) was adjusted to the desired pH level by addition of 0.1 M NaOH or 0.1 M HCl and it was placed in the feed water tank. The system was operated in the permeate recycle mode. A new membrane was used for each experiment after conditioning the membrane at least 2 h under the experimental conditions. Then the measuring sequence was started. Every hour, samples of permeate were taken and their boron concentrations were determined. The experiments were performed at 34 ± 1 ◦ C. The boron rejection was calculated according to the following equation: Boron rejection (%) =
C
permeate
Cfeed
× 100
(1)
where Cpermeate and Cfeed are the boron concentrations of the permeate and feed water, respectively.
Table 1 Characteristics of the membranes used. Characteristics
i. Feed water is a boron solution with different concentration (5, 10, 20 and 40 mg/L), at pH 5.5 and operating pressure 20 bar. ii. Feed water is a 40-mg/L of boron solution at pH ranging from 5.6 to 11 and operating pressure 20 bar. iii. Feed water is a 40-mg/L of boron solution at pH 5.5 under different operating pressure ranging from 16 to 35 bar.
Membranes
2.4. Natural (ground) water application
SWHR
BW-30
AG
Flat-sheet 45 1200 99.6 <0.1
Flat-sheet 45 600 99.5 <0.1
Flat-sheet 50 600 99 1000
The application of RO on the natural water samples taken from Kızıldere (Kütahya, Turkey) and Balc¸ova (I˙ zmir, Turkey) geothermal areas was performed under optimal conditions. The chemical composition of the natural water samples was determined by three times analyses [n = 3] and the results are given in Table 2.
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
143
Table 2 Chemical composition of the natural (ground) water samples from Kızıldere (pH 9.2) and Balc¸ova (pH 8.1). Ionic species
F− Cl− SO4 2− HCO3 − CO3 2− Li B Ca Mg Mn Fe Al
Concentration, mg/L [n = 3] Kızıldere (Kütahya)
Balc¸ova (I˙ zmir)
20.6 ± 0.8 76.7 ± 16.5 850 ± 18 1590 ± 24 213 ± 4 3.8 ± 0.1 24.8 ± 0.3 1.2 ± 0.1 Not detected Not detected Not detected Not detected
5.9 ± 0.6 125.9 ± 12.1 119.8 ± 5.9 965 ± 32 113 ± 6 1.1 ± 0.1 9.4 ± 0.1 9.4 ± 0.1 3.7 ± 0.2 Not detected Not detected Not detected
2.5. Instruments The concentration of boron and cations in the samples was determined by ContrAA 300 Atomic Absorption Spectroscopy (ContrAA 300, Analytikjena). The wavelength utilized for the determination of boron was 249.7 nm. Linearity for boron was observed in the concentration range of 1–20 mg/L. In addition, coefficient of regression (R2 ) and limit of detection (LOD) for boron were 0.9998 and 0.264 mg/L, respectively. Anions in natural water samples were determined by using the ion chromatography, a Dionex Model DX-100 system equipped with an AS16 separation column and an ASRS(R) Ultra II (4 mm) suppressor column (all supplied by Dionex). A solution of sodium carbonate (9 mM) was used as the eluent. pH of the samples was determined by an Orion ion meter with combined pH electrode. 3. Results and discussion 3.1. Effect of feed water concentration The effect of feed water concentration on the boron rejection is presented in Fig. 2. It is seen that boron concentration of feed water has no significant effect on the rejection. This result can be
Fig. 3. Dependancy of permeate concentration on the boron concentration of feed water. pH of feed water: 5.5, operating pressure: 20 bar, temperature: 34 ◦ C.
attributed to that permeate water concentration increases with increasing the feed water concentration (Fig. 3). In other words, boron rejection does not depend on the feed concentration. This finding is in agreement with previously reported result in Ref. [23]. Fig. 2 also shows that boron rejection is mainly affected by membrane type. The highest rejection was obtained by using SWHR membrane, whereas the lowest rejection was observed for AG membrane. The mean rejections for SWHR, BW-30 and AG membranes were 82, 73 and 55%, respectively. Dydo et al. [24] investigated the removal of boron from landfill leachate by using nanofiltration and RO with BW-30, TW-30, NF-90 and NF-45 (FILMTEC) membranes under maximum permissible operating pressure. Their results showed that boron rejection evidently depends upon membrane type and pH of the feed water. 3.2. Effect of pH of feed water The dependance of boron rejection upon pH of the feed water is presented in Fig. 4. For all studied membranes, an increase in the boron rejection was observed at pH > 9 of the feed water and the highest boron rejection was obtained at pH 11. This can be explained by considering dissociation of the boric acid. The value of the dissociation constant of boric acid is equal to pKa = 9.24, which means that boron predominantly exists as molecular species (boric acid) in aqueous solution at pH below 9 [25]. This neutral species easily permeates through the RO membrane, which results in lower boron rejection in comparison to the ionic species. Moreover, it was reported that boric acid is able to form hydrogen bridges with the active groups of the membranes and it can diffuse in a similar way to that of water [17]. The effect of pH can be better illustrated by the dissociation equilibrium given by the following equation: H3 (BO3 ) + 2H2 O ⇔ B(OH)4 − + H3 O+
Fig. 2. Dependancy of boron rejection on the boron concentration of feed water. pH of feed water: 5.5, operating pressure: 20 bar, temperature: 34 ◦ C.
(2)
When pH increases, H3 BO3 , which is a Lewis acid reacts with water resulting in the production of B(OH)4 − and H3 O+ . Especially, B(OH)4 − becomes the dominant species at pH between 9 and 10 and at pH 11 almost 100% of the boric acid exists as B(OH)4 − species [16,26]. In contrast to the neutral boric acid, which permeates easily the pores of the RO membrane, the anionic borate (B(OH)4 − ) is held in feed water and hence boron rejection is increased. Generally charged species are rejected to a greater extend by many RO mem-
144
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
Fig. 4. Dependancy of boron rejection on pH of feed water. Boron concentration of feed water: 40 mg/L, operating pressure: 20 bar, temperature: 34 ◦ C.
branes due to repulsive forces between membranes and anionic species [23,27–29]. Dydo et al. [24] pointed out that boron can be effectively removed only at a feed water pH, close to 11. Similar pH effects on the removal of boron from water by RO were also reported by other authors [17,19,23]. It should be also noted that operation at pH 11 with FILMTEC membranes is safe because in many situations the use of anti-scalants is not required. Therefore, it is obvious that operation at pH 11 is advantageous for boron rejection [27].
Fig. 6. Dependancy of permeate flux on the operating pressure. Boron concentration of feed water: 40 mg/L, pH of feed water: 5.5, temperature: 34 ◦ C.
In addition, increasing operating pressure also increased permeate flux (Fig. 6). Permeate flux for membranes was found to be in the order SWHR > BW-30 > AG. This result is expected because pressure is the driving force in RO system. Thus, higher operating pressure results in higher volume of permeate water. The same observation was indicated by Koseoglu et al. [32]. Permeate flux is important because higher flux gives the short operation time, which reduces the cost of RO system.
3.3. Effect of operating pressure 3.4. Natural (ground) water application Fig. 5 shows the effect of operating pressure on the boron rejection. Boron rejections for studied membranes increased with increasing operating pressure. Similarly, Koseoglu et al. [19], Sutzkover et al. [30] and Prats et al. [31] reported that higher boron rejection was observed when operating pressure was increased.
Fig. 5. Dependancy of boron rejection on the operating pressure. Boron concentration of feed water: 40 mg/L, pH of feed water: 5.5, temperature: 34 ◦ C.
In this study, the highest rejection and permeate flux were obtained by using SWHR membrane. Therefore, SWHR membrane was used for the removal of boron from natural waters by RO technique. Water samples were taken from Kızıldere (Kutahya, Turkey) and Balc¸ova (I˙ zmir, Turkey). Their boron concentrations were 24.8 and 9.4 mg/L, respectively. Prior to RO application, pH of the water samples was adjusted to 11 at which the highest boron rejection was obtained. Fig. 7 shows the time dependance of boron rejection for water samples. The mean boron rejections for samples from Kızıldere and Balc¸ova were 97 and 96%, respectively. Namely, permeate boron concentrations for samples from Kızıldere and Balc¸ova were 0.74 and 0.37 mg/L, respectively. As seen in Fig. 8, permeate fluxes increased during initial 2.5 h, which may indicate that dynamic membrane conditions were not achieved at initial 2.5 h. Then, fluxes for both samples reached a steady state values. Permeate fluxes for Kızıldere sample (20.9–30.5 L/m2 h) were found to be higher than those of Balc¸ova (17.2–24.1 L/m2 h). This result can be attributed to Ca and Mg contents of the samples. It is seen in Table 2 that Ca and Mg content of Balc¸ova sample is higher than that of Kızıldere. In literature, it was reported that lower permeate fluxes can be observed for the samples containing high amount of Ca and Mg because of the potential CaCO3 and Mg(OH)2 scaling [27,32]. WHO requires that boron concentration in drinking water is below 0.5 mg/L. In addition, EU defines the limit concentration of boron in drinking water as 1 mg/L. These requirements have affected the RO process design because of difficulty in achieving
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
145
4. Conclusion The results from this study can be concluded as follows: i. Removal of boron by RO depends greatly on the pH of the feed water. For all studied membranes it was found that boron can be effectively removed at pH 11. ii. Removal of boron increases when increasing the operating pressure. iii. The rejection of boron does not depend upon the feed water concentration. iv. Two different natural (ground) water samples containing 24.8 and 9.4 mg/L of boron were treated by using RO with SWHR membrane and obtained results showed that RO could be efficiently used (with >95% rejection) for removal of boron from groundwaters. Acknowledgments
Fig. 7. Boron rejections for natural samples Boron concentrations of samples from Kızıldere and Balc¸ova were 24.8 and 9.4 mg/L, respectively, pH of feed water: 11, operating pressure: 35 bar, temperature: 34 ◦ C.
such low boron concentrations. In order to overcome this problem, additional steps such as dilution of RO permeate with other sources, ion exchange post-treatment of RO permeate, and/or double-pass have been employed by most of the desalination plants [24,31,33]. For example, Dydo et al. [24] reported that high boron rejection (close to 99%), and low permeate concentration (<1 mg/L) were obtained at pH 11 by single stage RO with BW-30 membrane. Therefore, they proposed a two-stage RO system (at pH 11) to efficiently remove the boron from leachate samples. For the present study, permeate concentration of Balc¸ova sample (0.37 mg/L) was lower than WHO and EU recommendation limits. Permeate concentration of Kızıldere sample (0.74 mg/L) was also lower than EU limit. However, it was higher than WHO limit. Therefore, as suggested by Dydo et al. [24], a two stage RO system can be proposed for Kızıldere sample to reduce the permeate concentration below 0.5 mg/L.
Fig. 8. Permeate fluxes for natural samples boron concentrations of samples from Kızıldere and Balc¸ova were 24.8 and 9.4 mg/L, respectively, pH of feed water: 11, operating pressure: 35 bar, temperature: 34 ◦ C.
We thank the Scientific and Technical Research Council of Turkey (TUBITAK-Grant number 105M048) and Selcuk University Research Foundation (SUAF) for the financial support of this work. References [1] Agency for Toxic Substances and Disease Registary, Division of Toxicology/Toxicology Information Branch (ATSDR), Toxicological Profile for Boron, ATSDR, Atlanta, USA, 2007, pp. 1–249. [2] World Health Organization (WHO), Boron, Environmental Health Criteria Monograph 204, WHO, IPCS, Geneva, 1998. [3] E. Weinthal, Y. Parag, A. Vengosh, A. Muti, W. Kloppmann, The EU drinking water directive: the boron standard and scientific uncertainty, Eur. Environ. 15 (2005) 1–12. [4] A.E. Yilmaz, R. Boncukcuo˘glu, M.M. Kocakerim, B. Keskinler, The investigation of parameters affecting boron removal by electrocoagulation method, J. Hazard. Mater. 125 (2005) 160–165. [5] A.E. Yilmaz, R. Boncukcuo˘glu, M.M. Kocakerim, A quantitative comparison between electrocoagulation and chemical coagulation for boron removal from boron-containing solution, J. Hazard. Mater. 149 (2007) 475–481. [6] A.E. Yilmaz, R. Boncukcuo˘glu, M.M. Kocakerim, An empirical model for parameters affecting energy consumption in boron removal from boron-containing wastewaters by electrocoagulation, J. Hazard. Mater. 144 (2007) 101–107. [7] Y. Cengeloglu, A. Tor, G. Arslan, M. Ersoz, S. Gezgin, Removal of boron from aqueous solution by using neutralized red mud, J. Hazard. Mater. 142 (2007) 412–417. [8] N. Öztürk, D. Kavak, Boron removal from aqueous solutions by adsorption using full factorial design, Fresen. Environ. Bull. 12 (2003) 1450–1456. [9] A.E. Yilmaz, R. Boncukcuoglu, M.T. Yilmaz, M.M. Kocakerim, Adsorption of boron from boron-containing wastewaters by ion exchange in a continuous reactor, J. Hazard. Mater. B 117 (2005) 221–226. [10] R. Boncukcuoglu, A.E. Yilmaz, M.M. Kocakerim, M. Copur, An empirical model for kinetics of boron removal from boron-containing wastewaters by ion exchange in a batch reactor, Desalination 160 (2004) 159–166. [11] M. Badruk, N. Kabay, M. Demircioglu, H. Mordogan, U. Ipekoglu, Removal of boron from wastewater of geothermal power plant by selective ion-exchange resins. I. Batch sorption–elution studies, Sep. Sci. Technol. 34 (1999) 2553–2569. [12] M. Badruk, N. Kabay, M. Demircioglu, H. Mordogan, U. Ipekoglu, Removal of boron from wastewater of geothermal power plant by selective ionexchange resins. II. Column sorption–elution studies, Sep. Sci. Technol. 34 (1999) 2981–2995. [13] U. Beker, A. Cergel, O. Recepoglu, Removal of boron from geothermal power plant wastewater in Kizildere, Turkey, Energy Source 18 (1996) 645–654. [14] H.F. Ayyildiz, H. Kara, Boron removal by ion exchange membranes, Desalination 180 (2005) 99–108. [15] Z. Yazicigil, Y. Öztekin, Boron removal by electrodialysis with anion-exchange membranes, Desalination 190 (2006) 71–78. [16] G. Georghiou, I. Pashalidis, Boron in groundwaters of Nicosia (Cyprus) and its treatment by reverse osmosis, Desalination 215 (2007) 104–110. [17] M. Rodriguez-Pastor, A.F. Ruiz, M.F. Chillon, D.P. Rico, Influence of pH in the elimination of boron by means of reverse osmosis, Desalination 140 (2001) 145–152. [18] N. Özturk, D. Kavak, T.E. Köse, Boron removal from aqueous solution by reverse osmosis, Desalination 223 (2008) 1–9. [19] H. Koseoglu, N. Kabay, M. Yuksel, M. Kitis, The removal of boron from model solutions and seawater using reverse osmosis membranes, Desalination 223 (2008) 126–133.
146
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
[20] M. Taniguchi, Y. Fusaoka, T. Nishikawa, M. Kurihara, Boron removal in RO seawater desalination, Desalination 167 (2004) 419–426. [21] P. Glueckstern, M. Priel, Optimization of boron removal in old and new SWRO systems, Desalination 156 (2003) 219–228. [22] M. Busch, W.E. Mickols, Reducing energy consumption in seawater desalination, Desalination 165 (2004) 299–312. [23] Y. Magara, A. Tabata, M. Kohki, M. Kawasaki, M. Hirose, Development of boron reduction system for sea water desalination, Desalination 118 (1998) 25–33. [24] P. Dydo, M. Türek, J. Ciba, J. Trojanowska, J. Kluazka, Boron removal from landfill leachate by means of nanofiltration and reverse osmosis, Desalination 185 (2005) 131–137. [25] W. Stumm, J.J. Morgan, Aquatic Chemistry, 3rd edn., Wiley, NY, 1996. [26] J. Lou, Modelling of boron sorption equilibrium and kinetic studies of ion exchange with boron solution, Ph.D. Thesis, Oklahoma State University, USA, 1997. [27] J. Redondo, M. Busch, J.P. De Witte, Boron removal from seawater using FILMTECTM high rejection SWRO membranes, Desalination 156 (2003) 229– 238.
[28] N. Nadav, Boron removal from seawater reverse osmosis permeate utilizing selective ion exchange resin, Desalination 124 (1999) 131–135. [29] C. Bartels, R. Franks, S. Rybar, M. Schierach, M. Wilf, The effect of feed ionic strength on salt passage through reverse osmosis membranes, Desalination 184 (2005) 185–195. [30] I. Sutzkover, D. Hasson, R. Semiat, Simple technique for measuring the concentration polarization level in a reverse osmosis system, Desalination 131 (2000) 117–127. [31] D. Prats, M.F. Chillon-Arias, R.M. Pastor, Analysis of the influence of pH and pressure on the elimination of boron in reverse osmosis, Desalination 128 (2000) 269–273. [32] H. Koseoglu, N. Kabay, M. Yüksel, S. Sarp, Ö. Arar, M. Kitis, Boron removal from seawater using high rejection SWRO membranes—impact of pH, feed concentration, pressure, and cross-flow velocity, Desalination 227 (2008) 253– 263. [33] M. Wilf, C. Bartels, Optimization of seawater RO systems design, Desalination 173 (2005) 1–12.
Contents lists available at ScienceDirect
Separation and Purification Technology journal homepage: www.elsevier.com/locate/seppur
Removal of boron from water by using reverse osmosis Yunus Cengeloglu a,∗ , Gulsin Arslan a , Ali Tor b , Izzet Kocak a , Nesim Dursun c a
Selcuk University, Department of Chemistry, 42031 Campus, Konya, Turkey Selcuk University, Department of Environmental Engineering, 42031 Campus, Konya, Turkey c Selcuk University, Department of Soil, 42031 Campus, Konya, Turkey b
a r t i c l e
i n f o
Article history: Received 30 July 2008 Received in revised form 12 September 2008 Accepted 16 September 2008 Keywords: Boron removal Reverse osmosis Membrane
a b s t r a c t The removal of boron from water was investigated by using reverse osmosis (RO) technique with SWHR, BW-30 (FILMTEC) and AG (GE Osmonics) membranes, not considered previously. The effect of pH and concentration of the feed water and operating pressure on the boron rejection was investigated. The experimental results indicated that boron rejection mostly depends on membrane type, pH of the feed water and operating pressure. The results also showed that boron can be effectively removed only at feed water pH of 11. The lowest permeate boron concentration (the highest rejection, close to 99%) was obtained when SWHR membrane was used. The rejection efficiency of the membranes was found to be in the order of SWHR > BW-30 > AG. For all membranes, boron rejection increased with increasing operating pressure. Finally, two different natural (ground) water samples containing 24.8 and 9.4 mg/L of boron were treated by using RO with SWHR membrane and obtained results showed that RO could be efficiently used (with >95% rejection) for removal of boron from groundwaters. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The potential sources of boron contamination in water resources are either anthropogenic (pollution from sewage effluents, boron enriched fertilizers, landfill leachates and discharge from the soap and detergent manufacturing) or natural (e.g. water–rock interaction, sea water encroachment, mixing with fossil brines or hydrothermal fluids) [1]. Boron in drinking water is suspected as teratogenic. The 1993 WHO guidelines for drinking-water quality proposed a 0.3-mg/L standard for boron in drinking water. Later, this standard was revised to 0.5 mg/L because it would be difficult to achieve in areas with high natural boron levels with the treatment technology available [2]. Recently boron has been classified by the European Union (EU) as a pollutant of drinking water in national and international drinking water directives. The EU adopted a standard of 1 mg/L boron for drinking water [3]. Therefore, it is necessary to remove the excess amount of boron from water by a suitable method. The treatment methods used for the removal of boron from water can be divided into several categories. The first is coagulation and electrocoagulation processes [4–6]. The next two categories are adsorption [7,8] and ion-exchange processes [9–13]. The last cate-
∗ Corresponding author. Tel.: +90 332 223 2772; fax: +90 332 241 0106. E-mail address: [email protected] (Y. Cengeloglu). 1383-5866/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2008.09.006
gory consists of membranes processes such as Donnan dialysis [14], electrodialysis [15] and reverse osmosis (RO) [16–19]. In recent years, membrane manufactures have developed RO membranes with boron rejections of 91–96% [20,21]. However, most of the current desalination plants have to implement the additional treatment steps such as pH adjustment of feed water, post-treatment of RO permeate with ion exchange or several pass stage of permeate in order to improve boron rejection. In addition, several process configurations have been proposed to obtain the low boron concentration of the permeate from RO plant [17,21,22]. The objective of this work was to investigate and compare the boron removal efficiencies of three different RO membranes (SWHR, BW-30 and AG) using model solutions containing boron as single solute. The effect of pH and concentration of feed water and operating pressure on the boron rejection was determined. Finally, under optimal conditions, RO with SWHR membrane was applied to the natural (ground) water samples containing 24.8 and 9.4 mg/L of boron. The obtained permeate boron concentrations were compared with WHO and EU standards and permeate fluxes were evaluated. 2. Materials and methods 2.1. Reverse osmosis pilot plant The reverse osmosis pilot plant (Prozesstechnick GmbH) used in this study consists of a diaphragm pump controlled with a frequency converter (flow range: 1.8–12 L/min, pressure range: max
142
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
Fig. 1. Flow diagram of the reverse osmosis plot plant. (M1 and M2: Membrane housing, B1: Feed tank with heating/cooling jacket, V1 and V2: Emptying valve, V3 and V4: Pressure regulation valve, V5: Spring loaded valve, V6: Three way valve to select which membrane housing, P1: Pump, PI01 and PI02: Pressure gauge, DP1: Differential pressure indicator, LI01: Level indicator on the feed tank, TI01: Temperature indicator.)
40 bar), feed tank with heating/cooling jacket (5 L capacity), membrane housing for both spiral wound and flat-sheet membranes, different emptying and pressure valves. Fig. 1 shows the flow diagram of the described pilot plant. 2.2. Membranes Three membranes having 44 cm2 exposed area with a flat-sheet configuration were studied. The most relevant characteristics of these membranes are summarized in Table 1. 2.3. Experiments The boric acid solutions were prepared in distilled water by diluting the prepared stock solutions (1000 mg/L) to desired concentrations. H3 BO3 was obtained from R.P. Normapur® (Paris, France). NaOH and HCl were obtained from Merck Co. (Darmstad, Germany). All chemicals were the analytic grade reagents. In specific experiments, composition of the feed water and operating pressure were chosen as below:
Configuration Max temperature (◦ C) Max pressure (psig) Salt rejection (%) Chlorine tolerance (ppm)
At the beginning of each experiment, pH of the feed water (2 L) was adjusted to the desired pH level by addition of 0.1 M NaOH or 0.1 M HCl and it was placed in the feed water tank. The system was operated in the permeate recycle mode. A new membrane was used for each experiment after conditioning the membrane at least 2 h under the experimental conditions. Then the measuring sequence was started. Every hour, samples of permeate were taken and their boron concentrations were determined. The experiments were performed at 34 ± 1 ◦ C. The boron rejection was calculated according to the following equation: Boron rejection (%) =
C
permeate
Cfeed
× 100
(1)
where Cpermeate and Cfeed are the boron concentrations of the permeate and feed water, respectively.
Table 1 Characteristics of the membranes used. Characteristics
i. Feed water is a boron solution with different concentration (5, 10, 20 and 40 mg/L), at pH 5.5 and operating pressure 20 bar. ii. Feed water is a 40-mg/L of boron solution at pH ranging from 5.6 to 11 and operating pressure 20 bar. iii. Feed water is a 40-mg/L of boron solution at pH 5.5 under different operating pressure ranging from 16 to 35 bar.
Membranes
2.4. Natural (ground) water application
SWHR
BW-30
AG
Flat-sheet 45 1200 99.6 <0.1
Flat-sheet 45 600 99.5 <0.1
Flat-sheet 50 600 99 1000
The application of RO on the natural water samples taken from Kızıldere (Kütahya, Turkey) and Balc¸ova (I˙ zmir, Turkey) geothermal areas was performed under optimal conditions. The chemical composition of the natural water samples was determined by three times analyses [n = 3] and the results are given in Table 2.
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
143
Table 2 Chemical composition of the natural (ground) water samples from Kızıldere (pH 9.2) and Balc¸ova (pH 8.1). Ionic species
F− Cl− SO4 2− HCO3 − CO3 2− Li B Ca Mg Mn Fe Al
Concentration, mg/L [n = 3] Kızıldere (Kütahya)
Balc¸ova (I˙ zmir)
20.6 ± 0.8 76.7 ± 16.5 850 ± 18 1590 ± 24 213 ± 4 3.8 ± 0.1 24.8 ± 0.3 1.2 ± 0.1 Not detected Not detected Not detected Not detected
5.9 ± 0.6 125.9 ± 12.1 119.8 ± 5.9 965 ± 32 113 ± 6 1.1 ± 0.1 9.4 ± 0.1 9.4 ± 0.1 3.7 ± 0.2 Not detected Not detected Not detected
2.5. Instruments The concentration of boron and cations in the samples was determined by ContrAA 300 Atomic Absorption Spectroscopy (ContrAA 300, Analytikjena). The wavelength utilized for the determination of boron was 249.7 nm. Linearity for boron was observed in the concentration range of 1–20 mg/L. In addition, coefficient of regression (R2 ) and limit of detection (LOD) for boron were 0.9998 and 0.264 mg/L, respectively. Anions in natural water samples were determined by using the ion chromatography, a Dionex Model DX-100 system equipped with an AS16 separation column and an ASRS(R) Ultra II (4 mm) suppressor column (all supplied by Dionex). A solution of sodium carbonate (9 mM) was used as the eluent. pH of the samples was determined by an Orion ion meter with combined pH electrode. 3. Results and discussion 3.1. Effect of feed water concentration The effect of feed water concentration on the boron rejection is presented in Fig. 2. It is seen that boron concentration of feed water has no significant effect on the rejection. This result can be
Fig. 3. Dependancy of permeate concentration on the boron concentration of feed water. pH of feed water: 5.5, operating pressure: 20 bar, temperature: 34 ◦ C.
attributed to that permeate water concentration increases with increasing the feed water concentration (Fig. 3). In other words, boron rejection does not depend on the feed concentration. This finding is in agreement with previously reported result in Ref. [23]. Fig. 2 also shows that boron rejection is mainly affected by membrane type. The highest rejection was obtained by using SWHR membrane, whereas the lowest rejection was observed for AG membrane. The mean rejections for SWHR, BW-30 and AG membranes were 82, 73 and 55%, respectively. Dydo et al. [24] investigated the removal of boron from landfill leachate by using nanofiltration and RO with BW-30, TW-30, NF-90 and NF-45 (FILMTEC) membranes under maximum permissible operating pressure. Their results showed that boron rejection evidently depends upon membrane type and pH of the feed water. 3.2. Effect of pH of feed water The dependance of boron rejection upon pH of the feed water is presented in Fig. 4. For all studied membranes, an increase in the boron rejection was observed at pH > 9 of the feed water and the highest boron rejection was obtained at pH 11. This can be explained by considering dissociation of the boric acid. The value of the dissociation constant of boric acid is equal to pKa = 9.24, which means that boron predominantly exists as molecular species (boric acid) in aqueous solution at pH below 9 [25]. This neutral species easily permeates through the RO membrane, which results in lower boron rejection in comparison to the ionic species. Moreover, it was reported that boric acid is able to form hydrogen bridges with the active groups of the membranes and it can diffuse in a similar way to that of water [17]. The effect of pH can be better illustrated by the dissociation equilibrium given by the following equation: H3 (BO3 ) + 2H2 O ⇔ B(OH)4 − + H3 O+
Fig. 2. Dependancy of boron rejection on the boron concentration of feed water. pH of feed water: 5.5, operating pressure: 20 bar, temperature: 34 ◦ C.
(2)
When pH increases, H3 BO3 , which is a Lewis acid reacts with water resulting in the production of B(OH)4 − and H3 O+ . Especially, B(OH)4 − becomes the dominant species at pH between 9 and 10 and at pH 11 almost 100% of the boric acid exists as B(OH)4 − species [16,26]. In contrast to the neutral boric acid, which permeates easily the pores of the RO membrane, the anionic borate (B(OH)4 − ) is held in feed water and hence boron rejection is increased. Generally charged species are rejected to a greater extend by many RO mem-
144
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
Fig. 4. Dependancy of boron rejection on pH of feed water. Boron concentration of feed water: 40 mg/L, operating pressure: 20 bar, temperature: 34 ◦ C.
branes due to repulsive forces between membranes and anionic species [23,27–29]. Dydo et al. [24] pointed out that boron can be effectively removed only at a feed water pH, close to 11. Similar pH effects on the removal of boron from water by RO were also reported by other authors [17,19,23]. It should be also noted that operation at pH 11 with FILMTEC membranes is safe because in many situations the use of anti-scalants is not required. Therefore, it is obvious that operation at pH 11 is advantageous for boron rejection [27].
Fig. 6. Dependancy of permeate flux on the operating pressure. Boron concentration of feed water: 40 mg/L, pH of feed water: 5.5, temperature: 34 ◦ C.
In addition, increasing operating pressure also increased permeate flux (Fig. 6). Permeate flux for membranes was found to be in the order SWHR > BW-30 > AG. This result is expected because pressure is the driving force in RO system. Thus, higher operating pressure results in higher volume of permeate water. The same observation was indicated by Koseoglu et al. [32]. Permeate flux is important because higher flux gives the short operation time, which reduces the cost of RO system.
3.3. Effect of operating pressure 3.4. Natural (ground) water application Fig. 5 shows the effect of operating pressure on the boron rejection. Boron rejections for studied membranes increased with increasing operating pressure. Similarly, Koseoglu et al. [19], Sutzkover et al. [30] and Prats et al. [31] reported that higher boron rejection was observed when operating pressure was increased.
Fig. 5. Dependancy of boron rejection on the operating pressure. Boron concentration of feed water: 40 mg/L, pH of feed water: 5.5, temperature: 34 ◦ C.
In this study, the highest rejection and permeate flux were obtained by using SWHR membrane. Therefore, SWHR membrane was used for the removal of boron from natural waters by RO technique. Water samples were taken from Kızıldere (Kutahya, Turkey) and Balc¸ova (I˙ zmir, Turkey). Their boron concentrations were 24.8 and 9.4 mg/L, respectively. Prior to RO application, pH of the water samples was adjusted to 11 at which the highest boron rejection was obtained. Fig. 7 shows the time dependance of boron rejection for water samples. The mean boron rejections for samples from Kızıldere and Balc¸ova were 97 and 96%, respectively. Namely, permeate boron concentrations for samples from Kızıldere and Balc¸ova were 0.74 and 0.37 mg/L, respectively. As seen in Fig. 8, permeate fluxes increased during initial 2.5 h, which may indicate that dynamic membrane conditions were not achieved at initial 2.5 h. Then, fluxes for both samples reached a steady state values. Permeate fluxes for Kızıldere sample (20.9–30.5 L/m2 h) were found to be higher than those of Balc¸ova (17.2–24.1 L/m2 h). This result can be attributed to Ca and Mg contents of the samples. It is seen in Table 2 that Ca and Mg content of Balc¸ova sample is higher than that of Kızıldere. In literature, it was reported that lower permeate fluxes can be observed for the samples containing high amount of Ca and Mg because of the potential CaCO3 and Mg(OH)2 scaling [27,32]. WHO requires that boron concentration in drinking water is below 0.5 mg/L. In addition, EU defines the limit concentration of boron in drinking water as 1 mg/L. These requirements have affected the RO process design because of difficulty in achieving
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
145
4. Conclusion The results from this study can be concluded as follows: i. Removal of boron by RO depends greatly on the pH of the feed water. For all studied membranes it was found that boron can be effectively removed at pH 11. ii. Removal of boron increases when increasing the operating pressure. iii. The rejection of boron does not depend upon the feed water concentration. iv. Two different natural (ground) water samples containing 24.8 and 9.4 mg/L of boron were treated by using RO with SWHR membrane and obtained results showed that RO could be efficiently used (with >95% rejection) for removal of boron from groundwaters. Acknowledgments
Fig. 7. Boron rejections for natural samples Boron concentrations of samples from Kızıldere and Balc¸ova were 24.8 and 9.4 mg/L, respectively, pH of feed water: 11, operating pressure: 35 bar, temperature: 34 ◦ C.
such low boron concentrations. In order to overcome this problem, additional steps such as dilution of RO permeate with other sources, ion exchange post-treatment of RO permeate, and/or double-pass have been employed by most of the desalination plants [24,31,33]. For example, Dydo et al. [24] reported that high boron rejection (close to 99%), and low permeate concentration (<1 mg/L) were obtained at pH 11 by single stage RO with BW-30 membrane. Therefore, they proposed a two-stage RO system (at pH 11) to efficiently remove the boron from leachate samples. For the present study, permeate concentration of Balc¸ova sample (0.37 mg/L) was lower than WHO and EU recommendation limits. Permeate concentration of Kızıldere sample (0.74 mg/L) was also lower than EU limit. However, it was higher than WHO limit. Therefore, as suggested by Dydo et al. [24], a two stage RO system can be proposed for Kızıldere sample to reduce the permeate concentration below 0.5 mg/L.
Fig. 8. Permeate fluxes for natural samples boron concentrations of samples from Kızıldere and Balc¸ova were 24.8 and 9.4 mg/L, respectively, pH of feed water: 11, operating pressure: 35 bar, temperature: 34 ◦ C.
We thank the Scientific and Technical Research Council of Turkey (TUBITAK-Grant number 105M048) and Selcuk University Research Foundation (SUAF) for the financial support of this work. References [1] Agency for Toxic Substances and Disease Registary, Division of Toxicology/Toxicology Information Branch (ATSDR), Toxicological Profile for Boron, ATSDR, Atlanta, USA, 2007, pp. 1–249. [2] World Health Organization (WHO), Boron, Environmental Health Criteria Monograph 204, WHO, IPCS, Geneva, 1998. [3] E. Weinthal, Y. Parag, A. Vengosh, A. Muti, W. Kloppmann, The EU drinking water directive: the boron standard and scientific uncertainty, Eur. Environ. 15 (2005) 1–12. [4] A.E. Yilmaz, R. Boncukcuo˘glu, M.M. Kocakerim, B. Keskinler, The investigation of parameters affecting boron removal by electrocoagulation method, J. Hazard. Mater. 125 (2005) 160–165. [5] A.E. Yilmaz, R. Boncukcuo˘glu, M.M. Kocakerim, A quantitative comparison between electrocoagulation and chemical coagulation for boron removal from boron-containing solution, J. Hazard. Mater. 149 (2007) 475–481. [6] A.E. Yilmaz, R. Boncukcuo˘glu, M.M. Kocakerim, An empirical model for parameters affecting energy consumption in boron removal from boron-containing wastewaters by electrocoagulation, J. Hazard. Mater. 144 (2007) 101–107. [7] Y. Cengeloglu, A. Tor, G. Arslan, M. Ersoz, S. Gezgin, Removal of boron from aqueous solution by using neutralized red mud, J. Hazard. Mater. 142 (2007) 412–417. [8] N. Öztürk, D. Kavak, Boron removal from aqueous solutions by adsorption using full factorial design, Fresen. Environ. Bull. 12 (2003) 1450–1456. [9] A.E. Yilmaz, R. Boncukcuoglu, M.T. Yilmaz, M.M. Kocakerim, Adsorption of boron from boron-containing wastewaters by ion exchange in a continuous reactor, J. Hazard. Mater. B 117 (2005) 221–226. [10] R. Boncukcuoglu, A.E. Yilmaz, M.M. Kocakerim, M. Copur, An empirical model for kinetics of boron removal from boron-containing wastewaters by ion exchange in a batch reactor, Desalination 160 (2004) 159–166. [11] M. Badruk, N. Kabay, M. Demircioglu, H. Mordogan, U. Ipekoglu, Removal of boron from wastewater of geothermal power plant by selective ion-exchange resins. I. Batch sorption–elution studies, Sep. Sci. Technol. 34 (1999) 2553–2569. [12] M. Badruk, N. Kabay, M. Demircioglu, H. Mordogan, U. Ipekoglu, Removal of boron from wastewater of geothermal power plant by selective ionexchange resins. II. Column sorption–elution studies, Sep. Sci. Technol. 34 (1999) 2981–2995. [13] U. Beker, A. Cergel, O. Recepoglu, Removal of boron from geothermal power plant wastewater in Kizildere, Turkey, Energy Source 18 (1996) 645–654. [14] H.F. Ayyildiz, H. Kara, Boron removal by ion exchange membranes, Desalination 180 (2005) 99–108. [15] Z. Yazicigil, Y. Öztekin, Boron removal by electrodialysis with anion-exchange membranes, Desalination 190 (2006) 71–78. [16] G. Georghiou, I. Pashalidis, Boron in groundwaters of Nicosia (Cyprus) and its treatment by reverse osmosis, Desalination 215 (2007) 104–110. [17] M. Rodriguez-Pastor, A.F. Ruiz, M.F. Chillon, D.P. Rico, Influence of pH in the elimination of boron by means of reverse osmosis, Desalination 140 (2001) 145–152. [18] N. Özturk, D. Kavak, T.E. Köse, Boron removal from aqueous solution by reverse osmosis, Desalination 223 (2008) 1–9. [19] H. Koseoglu, N. Kabay, M. Yuksel, M. Kitis, The removal of boron from model solutions and seawater using reverse osmosis membranes, Desalination 223 (2008) 126–133.
146
Y. Cengeloglu et al. / Separation and Purification Technology 64 (2008) 141–146
[20] M. Taniguchi, Y. Fusaoka, T. Nishikawa, M. Kurihara, Boron removal in RO seawater desalination, Desalination 167 (2004) 419–426. [21] P. Glueckstern, M. Priel, Optimization of boron removal in old and new SWRO systems, Desalination 156 (2003) 219–228. [22] M. Busch, W.E. Mickols, Reducing energy consumption in seawater desalination, Desalination 165 (2004) 299–312. [23] Y. Magara, A. Tabata, M. Kohki, M. Kawasaki, M. Hirose, Development of boron reduction system for sea water desalination, Desalination 118 (1998) 25–33. [24] P. Dydo, M. Türek, J. Ciba, J. Trojanowska, J. Kluazka, Boron removal from landfill leachate by means of nanofiltration and reverse osmosis, Desalination 185 (2005) 131–137. [25] W. Stumm, J.J. Morgan, Aquatic Chemistry, 3rd edn., Wiley, NY, 1996. [26] J. Lou, Modelling of boron sorption equilibrium and kinetic studies of ion exchange with boron solution, Ph.D. Thesis, Oklahoma State University, USA, 1997. [27] J. Redondo, M. Busch, J.P. De Witte, Boron removal from seawater using FILMTECTM high rejection SWRO membranes, Desalination 156 (2003) 229– 238.
[28] N. Nadav, Boron removal from seawater reverse osmosis permeate utilizing selective ion exchange resin, Desalination 124 (1999) 131–135. [29] C. Bartels, R. Franks, S. Rybar, M. Schierach, M. Wilf, The effect of feed ionic strength on salt passage through reverse osmosis membranes, Desalination 184 (2005) 185–195. [30] I. Sutzkover, D. Hasson, R. Semiat, Simple technique for measuring the concentration polarization level in a reverse osmosis system, Desalination 131 (2000) 117–127. [31] D. Prats, M.F. Chillon-Arias, R.M. Pastor, Analysis of the influence of pH and pressure on the elimination of boron in reverse osmosis, Desalination 128 (2000) 269–273. [32] H. Koseoglu, N. Kabay, M. Yüksel, S. Sarp, Ö. Arar, M. Kitis, Boron removal from seawater using high rejection SWRO membranes—impact of pH, feed concentration, pressure, and cross-flow velocity, Desalination 227 (2008) 253– 263. [33] M. Wilf, C. Bartels, Optimization of seawater RO systems design, Desalination 173 (2005) 1–12.