Abstract
According to World Health Organization (WHO) statements, the boron concentration for drinking water should be less than 0.5 mg L−1. This study is aimed to tackle the challenge to obtain a low boron concentration at this level. In recent years, membrane capacitive deionization (MCDI) has started to attract attention as it allows the ions to be removed quickly and easily from the solutions. In this study, initially boron removal rates of ultrafiltration (UF) membrane were tested at different pressures. Then, MCDI system was operated to determine the optimum conditions for boron elimination. The optimum values were 40 ml min−1 flow rate, 1.1 V applied voltage, 10 min operating time, and pH 13. On the next step, an integrated UF membrane and MCDI system was developed to improve the total boron removal efficiency. The new integrated system exhibited excellent boron rejection performance in the range of 96%–100%. The technical aspects of design concept to reach low boron limits were discussed.
Export citation and abstract BibTeX RIS
1. Introduction
The need for fresh water increases due to population and economic growth, but fresh water resources are limited. The severity of this restriction varies between regions due to the physical availability of water resources and differences in local socio-economic conditions. Currently, more than two billion people live in places with water scarcity worldwide [1]. Water shortage has started to show its effects in the 19th century and has become one of the main problems of many societies in the world. Water demand increase is more than twice the population growth rate in the last century [2]. According to the United Nations report, if the necessary measures are not taken, 1.8 billion people in the world may suffer from water shortage in 2025 [3]. Fresh water is unevenly distributed across the world. Global climate change, rapid population growth, and the increasing demand for water in sectors such as agriculture, livestock, and energy are serious problems concerning this issue [4, 5].
Only 0.5% of the water resources in the world is fresh, and 97% is classified as sea water, and 2.5% as salty groundwater. It is clear that, by converting salt water into drinking water, an unlimited potential for water supply can be obtained. Desalination of seawater is much more important in regions with water shortage, like Middle East, and in countries with insufficient fresh water resources [6, 7]. In this context, desalination of sea water represents a feasible and effective method of obtaining clean water in some parts of the world. For effective use of desalination technologies, it is necessary to monitor and control the concentration levels of various natural compounds [8, 9].
One of the natural components in the sea water is boron, which is toxic at high concentrations. Boron occurs naturally and can be found in sea water at an average concentration of 4–5 mg L−1. Nonetheless, World Health Organization (WHO) Guidelines for Drinking Water Quality suggests a maximum boron concentration of 0.5 mg L−1 [10]. Boron is an important micro food for plants, animals and humans, despite the gap among insufficiency and redundancy is limited. It was shown in both animal and human studies that boron intake <1.0 mg per day could not reveal the beneficial effects of boron on health. However, excessive amount of boron is harmful to both human and animal health [11–13].
So far, boron elimination has been increased to complete elimination by thermal desalination processes such as multi-stage flash distillation (MSF) and multi-effect distillation (MED). Nevertheless, thermal technologies are increasingly applied less owing to their high energy footprints compared to membrane technology [14–16]. Alternative methods for boron removal can be given as adsorption, coagulation, ion-exchange, electrocoagulation and membrane methods [17–20].
Reverse osmosis (RO) membrane systems were studied thoroughly for boron removal [12, 21, 22]. RO membrane system is the mostly preferred technology for producing drinkable water from salty water resources. However, the main of problem of rapid membrane fouling reduces the economic efficiency of this system [23, 24]. Therefore, prior to RO membranes, it is necessary to use pre-treatment systems, in order to remove (completely or partially) present waste materials such as particles, colloids and organic matter [25, 26]. In recent years, ultrafiltration (UF) is preferred more than RO due to quick fouling of RO membranes. UF membranes, with pore size between 0.1 and 0.01 μm, can remove particles, colloids, microorganisms and some dissolved organic substances (usually with the aid of a coagulant dosage) and also produce high quality filtrate [27, 28].
Capacitive deionization (CDI) is also a promising water treatment method regarding its simple design and easy operating conditions without chemical demand. CDI systems operate in adsorption and desorption cycles to obtain pure water and concentrated water, respectively. When electrical potential is given to the CDI cells, ions in the contaminated water are adsorbed on the surface of the porous carbon electrodes, resulting in purified water. For regeneration of ion saturated electrodes, the polarity is reversed to release back the caught ions as concentrate [29]. Membrane capacitive deionisation (MCDI), a modified version of CDI cell, is formed by coating the electrodes with two different selective ion exchange membranes (IEM), namely cation exchange membrane (CEM) and anion exchange membrane (AEM). In order to remove the ions from the feed water, MCDI system also operates with the electrical potential difference applied between two IEM coated electrodes. Nonetheless, MCDI systems are sensitive to fouling and scaling [30, 31].
The present study aimed to experimentally and theoretically investigate and describe the performance of a novel hybrid system. Initially, boron removal rates of UF membrane were tested at varying pressures. At a later stage, the removal efficiency of boron from synthetic solution was investigated at different operating conditions of MCDI system. On the next step, the elimination of boron was studied using an integrated laboratory scale UF membrane and MCDI system. Last but not least, a concept system that can produce its own energy during treatment operation was also developed. The results of this study are expected to expand the available knowledge to assist in the development of safe drinking water technologies and to remove environmentally relevant boron contaminants during water treatment and desalination process.
2. Materials and methods
2.1. Membrane system
In this study, a dead-end filtration cell loaded with flat sheet membrane (Sterlitech, USA) was used. The detailed structure of the membrane cell system was thoroughly described in a previous study [32]. Experimental tests were conducted to analyse the water permeability and boron mass transfer rate of UF membrane. The permeation volume of fluid deposited in the graduated cylinder within a definite time period (every 20 s during 10 min operating time). The permeation flux (L/m2.hr) was determined by using the permeation volume flowing across the active membrane. After 10 min, the UF membrane was replaced to begin the new experimental set and each set was repeated 3 times. The test was continued by increasing the operating pressure 0.2, 0.7, 1.2, 1.7, 2.1 and 2.6 bar. All measurements were carried out at room temperature.
2.2. MCDI system
The system consists of a feed vessel, a peristaltic pump (Watson Marlow 350, USA), a MCDI module and a potentiostat. Defined potential was applied to the MCDI cell via using a potentiostat (Gamry Interface 1000, USA). The MCDI module consists of cathode and anode carbon electrodes, current collectors (carbon cloth), IEMs for both cathode and anode; and the plexiglas case. For cathode, Ultrex CMI-7000 (Membranes International Inc., USA) was preferred as CEM, and dissimilarly Neosepta AMX (Astom Co., Japan) was used as AEM for the anode. The carbon electrodes, current collectors and IEMs were cut into dimensions of 12 cm × 12 cm. The operating flow rate of the MCDI system was 40 ml min−1. All experiments were repeated 3 times. Figure 1 shows the schematic view of the MCDI system.
Figure 1. Schematic view of the MCDI system.
Download figure:
Standard image High-resolution image2.3. Boron water characterization
Synthetic solutions were prepared by mixing ultrapure water and boric acid (Carlo Erba, Milan, Italy). Boron concentration in the solution was measured by atomic absorption spectroscopy (Perkin-Elmer Analyst 400). Boron removal (%) was calculated from inlet and outlet boron concentration (equation (1)).
where, C0 and Ci represent the initial and final boron concentrations in mg L−1, respectively.
3. Results and discussion
3.1. Effects of pressure on UF yield
In this experiment, the permeability of the UF membrane was tested using distilled water and synthetic boron solution as feed. The operating pressure was incrementally increased (0.2, 0.7, 1.2, 1.7, 2.1 and 2.6 bar) after each experimental set, to observe the change in permeate flux. The boron concentration in the synthetic solution was kept stable as 10 mg L−1. As seen in figure 2, the permeate flux graphs of distilled water and boron solution showed similar trends against varying pressures.
Figure 2. Permeate flux yielded by the UF membranes.
Download figure:
Standard image High-resolution imageMembrane fouling during the filtration process is an important issue that must be addressed for the practical application of membrane technology for boron removal [21, 33]. Flux is an important parameter in membrane systems, since high flux provides short operation time reducing both investment and operating costs. The basic requirement for the commercial application of membrane processes is to obtain high retention together with high flux values. As shown in figure 2, the permeate flux escalated with the increase of operating pressure. Various experiments to date define the same consistency that permeate would increase with rising applied pressure [22, 34].
Boron removal rates at different pressures are given in the table 1. As can be seen in table 1, boron removal efficiency of UF membranes declined with increase in pressure. Similar observation was also reported by Chon et al 2013 [25].
Table 1. Boron removal rates of UF at different pressures.
| Pressure (bar) | Boron rejection (%) |
|---|---|
| 0.2 | 95.4 |
| 0.7 | 84.3 |
| 1.2 | 52 |
| 1.7 | 24.2 |
| 2.1 | 12.5 |
| 2.6 | 9.7 |
3.2. Effects of different operating conditions on MCDI system
3.2.1. Influence of boron concentration on MCDI system
In this experiment, the effect of feed concentration on single MCDI system was analysed. To anticipate the effect of boron concentration on MCDI performance, solutions containing different boron concentrations (1, 4, 7, 10 and 13 mg L−1) were fed to the system. As shown in figure 3, the boron removal efficiency changed from 74% to 93% for diverse boron concentrations.
Figure 3. Effect of boron concentration in feed water (0.9 V operating voltage, pH 7, 6 min operating time).
Download figure:
Standard image High-resolution imageThe results showed that, in single MCDI cell, boron removal efficiency rises with the increase in boron concentration of feed water. Kim et al (2015) reported 100% salt removal with 5 min min−1 adsorption/desorption time and also 98% salt removal with 3 min min−1 cycle in MCDI cell. The results presented here are consistent with the literature [35].
3.2.2. Influence of pH on MCDI system
Previous studies shows that, more than 99% of the boric acid in the solution dissociates in borate ion when pH reaches 11.5 [22, 35]. In the literature, boron rejection rises from 50%–75% at pH 7–8 to 95% at pH 10.5, while the concentration of borate ions escalates with rise in pH [18, 22]. The effect of solution pH to enhance boron removal in membrane systems was noticed extensively in the literature [36–38]. However, the influence of pH for boron elimination has not studied in MCDI systems.
Results of boron removal with single MCDI cell at different pH values are shown in figure 4. With the rise of pH, boron removal rate showed an increasing trend. At pH 13, boron rejection rates were more than 83% for all different boron concentration. These results are well in line with the reported data from prior studies [37, 39].
Figure 4. The impact of pH on boron removal (0.9 V applied voltage and 6 min operating time).
Download figure:
Standard image High-resolution image3.2.3. Influence of applied voltage on MCDI system
The applied voltage is an important parameter for the removal performance of the MCDI cell. Applied voltage was adjusted in range of 0.3–1.1 V to determine the optimum operating potential. Figure 5 shows the effect of varying applied voltage (0.3, 0.6, 0.9 and 1.1 V) on boron removal performance.
Figure 5. The influence of applied voltage on boron removal (6 min operating time and pH 7).
Download figure:
Standard image High-resolution imageAs seen in figure 5, boron removal efficiency increased with rise of the applied voltage. For the solution with 13 mg L−1 boron concentration, the removal efficiency was 95.5% even at low applied voltage of 0.3 V and the elimination rate reached to 99.8% for 1.1 V. MCDI process was driven by the potential variation among a pair of activated porous carbon electrodes. In this regards higher voltages leads to better boron removal performance, so the optimum operating potential was 1.1 V. The results are also consistent with other studies [40, 41]. Li et al (2011) achieved high removal efficiency of 97% in MCDI system with 110 μs cm−1 initial conductivity and 1.2 V potential [42].
3.2.4. Influence of operating time on MCDI system
Other significant parameter for optimization of MCDI is the operating time. In this experimental set, the effect of longer operating times on boron removal was investigated. As presented in figure 6, boron removal in the MCDI system has increased gradually with escalating operating times.
Figure 6. The influence of operating time on boron removal (0.9 V applied voltage and pH 7).
Download figure:
Standard image High-resolution imageAs shown in figure 6, the boron removal efficiency ranged from 68.3% to 98.7% for the operating time between 2 min and 10 min. In approximately 10 min of operation, boron removal was higher than 98.4% for 13 mg L−1 concentration. This achieved removal rate is sufficient for water quality, and further increase of operating time is likely to increase the cost of water treatment with MCDI. Same results were documented in a former study for optimizing the salt adsorption ratio in MCDI indicating similar aspects with varying operating times [43].
3.3. Boron removal with integrated UF & MCDI system
After optimum working conditions of the single MCDI cell were detected, boron removal rate of integrated UF & MCDI system was also tested. In this case, boron contaminated water was passed through UF membrane as pretreatment before feeding into the MCDI cell. The experimental set-up is schematized in figure 7. As described above, the optimum conditions for maximum boron extraction in single MCDI cell were detected as 40 ml min−1 flow rate, 1.1 V operating potential, 10 min operating time and pH 13. The same operating conditions for MCDI were also kept stable for the integrated UF & MCDI system, and only operating pressures were alternated to determine the effect on the whole system. Table 2 shows boron removal performance of the integrated UF & MCDI under different pressures.
Figure 7. Schematic view of integrated UF & MCDI system.
Download figure:
Standard image High-resolution imageTable 2. Boron removal rate of integrated UF & MCDI system.
| UF membrane | MCDI cell | |
|---|---|---|
| Pressure (bar) | Boron rejection (%) | Boron rejection (%) |
| 0.2 | 95.4 | 100 |
| 0.7 | 84.3 | 100 |
| 1.2 | 52 | 100 |
| 1.7 | 24.2 | 99 |
| 2.1 | 12.5 | 98.3 |
| 2.6 | 9.7 | 96.7 |
As described above, the optimum conditions for maximum boron extraction in single MCDI cell were detected as 40 ml min−1 flow rate, 1.1 V operating potential, 10 min operating time and pH 13. The same operating conditions for MCDI were also kept stable for the integrated UF & MCDI system, and only operating pressures were alternated to determine the effect on the whole system. Table 2 shows boron removal performance of the integrated UF & MCDI under different pressures.
As seen in table 2, high boron removal rates were obtained in the range of 96.7%–100% under varying operation pressures. The integrated system had better removal efficiency values under low pressure values. High flux with increase in pressure causes the membrane to loosen and resulting in higher transition. Low pressure membranes are more preferable as having better treatment results and lesser membrane investment cost. Low pressure also reduces operating costs, while less energy is required for driving force of water transport across the membrane.
3.4. Power production from low temperature application
Organic Rankine cycle (ORC) system is generally able to produce power at lower pressure and temperature, while refrigerant fluid is used instead of water like in traditional Rankine cycles [44]. The overall objective of ORC is to produce power (electricity) at lower temperatures. The system consists of 4 main elements. (1) Heat exchanger (HEX): the waste heat of the external system is transferred to the refrigerant fluid. After the HEX, the refrigerant fluid enters into the turbine as pressurized hot steam. (2) Turbine: by taking the energy of superheated steam, the system produces power while reducing the pressure. (3) Condenser: the fluid & vapour mixture is condensed as saturated liquid. (4) Pump: the saturated liquid is compressed at high pressure and transferred to the HEX. These systems can be used for different purposes (heating, cooling, power) such as cogeneration, trigeneration, multigeneration, and especially for the reuse of waste heat [45].
In scope of this study, a system concept was developed to mount ORC power system to the integrated UF & MCDI system. By using HEX, the waste heat of the hot MCDI discharge water is transferred to the refrigerant fluid and ORC system produces power as explained above. In this regards, the integrated system is able produce its own energy that is mostly needed for creation of potential and pressure differences during treatment. Figure 8 shows the schematic view of integrated UF & MCDI system mounted with concept ORC power system.
Figure 8. Integrated UF & MCDI system mounted with concept ORC power system.
Download figure:
Standard image High-resolution image
4. Conclusion
In this paper, different operating conditions were tested to increase the boron extraction rate from water by using a single UF membrane cell, a single MCDI cell and also integrated UF & MCDI system. Boron was eliminated from solution with high efficiency at low pressures. MCDI performances at varying cell voltages from 0.3 V to 1.1 V showed that, this system has the potential to be used in a wider range of voltages. Providing adequate process optimization with optimum concentration, operating time and applied voltage, MCDI can be extremely effective for recovery of water with relatively high boron concentrations. It can be concluded that, integrated UF & MCDI system has the potential to be a promising technology alternative for effective boron extraction from water sources.