- Email: [email protected]
Removal of strontium ions from simulated radioactive wastewater by vacuum membrane distillation
Annals of Nuclear Energy 103 (2017) 363–368
Contents lists available at ScienceDirect
Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene
Removal of strontium ions from simulated radioactive wastewater by vacuum membrane distillation Fei Jia a, Junfeng Li a,b, Jianlong Wang a,b,⇑, Yuliang Sun a a b
Collaborative Innovation Center for Advanced Nuclear Energy Technology, INET, Tsinghua University, Beijing 100084, PR China Beijing Key Laboratory of Radioactive Waste Treatment, Tsinghua University, Beijing 100084, PR China
a r t i c l e
i n f o
Article history: Received 24 November 2016 Received in revised form 3 February 2017 Accepted 4 February 2017
Keywords: Vacuum membrane distillation Hollow fiber membrane Strontium Radioactive wastewater Dusty gas model
a b s t r a c t The removal of Sr2+ ions from simulated radioactive wastewater by vacuum membrane distillation (VMD) with hollow fiber membrane module was investigated. The removal efficiency of Sr2+ could maintain over 99.60%, and the membrane flux could maintain at 6.71 Lm2h1 when Sr2+ in the feed solution was about 10 mg/L. The effect of operating parameters on the membrane flux was examined, including feed temperature (30–70 °C), permeate side vacuum (0.10–0.98 atm) and feed flow velocity (10.5–41.8 L/h). Pressure buildup effect was observed during the process, which could interfere with the membrane flux. Pressure buildup effect was more serious when permeate side vacuum degree was over 0.9 atm. Considering energy consumption efficiency, the permeate side vacuum degree should be maintained at the turning point (0.9 atm) of the membrane flux. Dusty gas model could simulate the mass transfer of VMD process well with ARE (average relative error) of 5.31%. VMD is potential for the removal of Sr2+ ions from aqueous solution. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction After the tragedy of Fukushima Daiichi, the treatment of radioactive wastewater has received increasing attention. As one of main fission products of nuclear reactor, strontium has been regarded as a hazardous radionuclide with half-life of about 30 years (El-Kamash, 2008). A series of methods have been investigated to remove Sr2+ from aqueous solution, including adsorption (Chen and Wang, 2010, 2012; Park et al., 2010; Chen et al., 2014; Chen et al., 2016), ion exchange (El-Kamash, 2008) and membrane technology (Liu and Wang, 2013; Ding et al., 2015, 2016), etc. Compared with other methods, membrane technology could achieve both high salt removal efficiency and high concentration factor (Lawson and Lloyd, 1997). Among all membrane technologies, membrane distillation (MD) has some features in terms of its unique purification mechanism. Unlike series of pressure-driven membrane methods, MD employs the temperature difference across the membrane as the mass transfer driving force (Lawson and Lloyd, 1997). The membrane of MD system is made from hydrophobic materials which only allow volatile matters to pass through the membrane during the separation process (Lawson and Lloyd, 1997). Non-volatile matters like Sr2+ will be remained ⇑ Corresponding author at: Energy Science Building, Tsinghua University, Beijing 100084, PR China. E-mail address: [email protected] (J. Wang). http://dx.doi.org/10.1016/j.anucene.2017.02.003 0306-4549/Ó 2017 Elsevier Ltd. All rights reserved.
in the original wastewater. In previous research, we confirmed that MD could achieve high nuclide removal efficiency (Liu and Wang, 2013). Compared with pressure-driving membrane method, MD has several advantages including milder and safer operation condition, capability in processing high salinity wastewater, potential to use low quality heat sources like waste heat in nuclear power plant as mass transfer driving force (Lawson and Lloyd, 1997; Zuo et al., 2016). Classified by the membrane configuration, MD process could be categorized into four types: direct contact membrane distillation (DCMD), air gap membrane distillation (AGMD), sweeping gas membrane distillation (SGMD) and vacuum membrane distillation (VMD). DCMD is the original MD type with the simplest structure. However, as the permeate solution contacts with the membrane directly, the permeate will be polluted directly if there is a damage on the membrane or membrane wetting. The rest of three MD types decrease the risk of permeate pollution by separate membrane and permeate. For VMD method, it creates a vacuum atmosphere in the shell side of the membrane column to strip the permeate vapor from the shell. Besides decreasing the influence of membrane wetting, the application of vacuum also increase the mass transfer driving force of the system, which could increase the membrane flux of VMD process. There are three types of membrane column for VMD method: flat sheet membrane, spiral wound membrane and hollow fiber membrane (El-Bourawi et al., 2006).
364
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
Nomenclature A AGMD DCMD DGM Dm
DP Dt d
e J Jexp JDGM MD P
total effective membrane area (m2) air gap membrane distillation direct contact membrane distillation dusty gas model mass increase of the permeate collected in the condense (g) pressure difference across the membrane surface (Pa) sampling time (h) membrane thickness (m) membrane porosity permeate flux (Lm2h1) actual permeate flux (Lm2h1) simulated permeate flux (Lm2h1) membrane distillation mean pressure in membrane pores
Among these three types, hollow fiber type could achieve the highest surface density area. This results in the more compact system of the membrane module, which increase the real application potential of hollow fiber VMD method. There are a series of researches on the treatment of radioactive wastewater with MD method (Zakrzewska-Trznadel et al., 1999; Khayet, 2013; Liu and Wang, 2013, 2016; Wen et al., 2016). However, there were few researches focusing on Sr removal by VMD method. As discussed above, VMD has a potential to be an effective method to remove Sr2+ ions from solution. The objective of this study was to investigate the treatment of Sr2+-containing radioactive wastewater by hollow fiber VMD method. The effect of operating parameters on the performance of VMD process, including feed temperature, shell side vacuum degree and feed flow rate was studied. The mathematical simulation of VMD mass transfer process was discussed to predict the membrane flux.
Psat PP
q r R SGMD T Tf Tf,m Tp Tp,m
s VMD
pure water vapor saturated pressure polypropylene density of permeate solution in the present study (kgL1) average pore radius (m) universal gas constant (8.314 Jmol1K1) sweeping gas membrane distillation mean temperature in membrane pores (K) temperature of feed side (K) temperature of membrane surface in feed side (K) temperature of permeate side (K) temperature of membrane surface in permeate side (K) pore tortuosity of membrane vacuum membrane distillation
During VMD process, the solution would be heated to the preset temperature at first and then be pumped into the membrane module. At the shell side of the module, the vacuum pump would create a vacuum atmosphere and the permeate vapor would be stripped away from the shell and be condensed into liquid in the condenser which was filled with chilled water. Then the rest of solution in the lumen side would be recycled into the container to supplement the heat loss. 2.3. Analytical methods Sr ions concentration was analyzed by flame AAS (Hitachi, ZA3000, Japan). Membrane flux of MD was obtained by detecting water mass collected in the condenser with precision balance. The membrane flux was deduced as follows:
J¼
Dm
qADt
ð1Þ
where J is the membrane flux (Lm2h1); Dm is the water mass collected in the condense (g); q is water in the present study (gm3); A is the membrane area (m2); Dt is the sampling interval (h).
2. Materials and methods 2.1. Chemicals and membrane modules SrCl2 (Sinopharm Chemical, China; analytical pure) solution of about 10 mg/L was employed as feed in this experiment. Hollow fiber membrane made from polypropylene (PP) (Wochi, WHPP96-21, China) was applied in the VMD process. Property of the membrane is listed in detail in Table 1. 2.2. Experimental set-up The VMD system was made up of feed container, feed heating system, MD module, condenser system, vacuum pump, peristaltic pumps, precision balance, flowmeter and thermometers (Fig. 1).
Table 1 Performance of the membrane module. Parameters
Units
Value
Hollow fiber number Effective membrane length Total effective inner area Packing density Porosity Mean pore diameter Inner radius of the fiber Outer radius of the fiber
– mm m2 – – lm mm mm
140 140 0.062 0.35 60% 0.18 0.50 1.36
3. Results and discussion 3.1. Effect of feed temperature In this section, feed temperature was varied from 30 to 70 °C, with 0.98 atm of permeate vacuum degree and 41.8 L/h of feed flow velocity. Fig. 2 shows the variation of permeate at different feed temperatures. It could be seen that at all temperatures, the permeate mass increased linearly with time, indicating that the VMD system reached the steady state for producing water. Therefore the sampling time was determined to be 10 min. Fig. 3 shows the variation of the membrane flux with feed temperature. It could be seen that the membrane flux increased exponentially (R2 = 0.9762) with the feed temperature, which could be elucidated by the variation of saturated vapor pressure in the lumen side. During the VMD process, the pressure difference across the membrane played as the mass transfer driving force. The saturated vapor pressure of lumen side Psat (Pa) at temperature of T(K) could be calculated by Antoine Equation:
3816:44 log Psat ðTÞ ¼ exp 23:1964 þ 46:13 T
ð2Þ
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
365
Fig. 1. Schematic diagram of VMD system used in this study.
Fig. 2. Variation of permeate mass at different feed temperatures.
Fig. 4. Variation of membrane flux with the mass transfer driving force.
membrane flux and the mass transfer driving force could be observed. Previous researchers also obtained the similar results (Liu and Wang, 2013; Zuo et al., 2016). 3.2. Effect of permeate side vacuum degree
Fig. 3. Variation of the membrane flux with feed temperature.
As shown in the equation, the pressure of the lumen side would increase exponentially with the temperature. Therefore, the mass transfer driving force would behave the same with the temperature. Fig. 4 shows the relationship of membrane flux with the mass transfer driving force. A linear relationship (R2 = 0.9953) between
In this section, the permeate side vacuum degree varied from 0.1 to 0.98 atm. The experiment was performed at 70 °C and 41.8 L/h of feed flow velocity. Fig. 5 shows the variation of membrane flux with permeate side vacuum degree. It could be observed that the membrane flux varied in S shape with permeate vacuum. The flux maintained about zero before 0.6 atm of vacuum degree and started to increase linearly from 0.6 to 0.9 atm while leveled off at about 6.5 Lm2h1 after 0.9 atm. For zero flux before vacuum degree of 0.6 atm, it could be ascribed to the feed solution saturation vapor pressure. According to Antoine Equation, at 70 °C, Psat of the lumen side could be about 31,000 Pa, 0.3 atm. Hence only when the permeate side vacuum degree was over 0.7 atm, there could be a positive driving force across the membrane. However, this turning point shifted to 0.6 atm in the study. This could be due to the determination method of feed temperature. In the study, the feed temperature was determined by the average temperature of the inlet and outlet temperature of the membrane module. When feed temperature was identified as 70 °C, the inlet
366
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
Fig. 5. Variation of membrane flux with permeate side vacuum degree.
and outlet temperature were 75 °C and 65 °C respectively. Thus in the membrane column there was a part where the temperature was over 70 °C, indicating that in this part the mass transfer driving force should be positive. Therefore, the zero flux turning point occurred before the theoretical value. Another notable phenomenon was the membrane flux level-off when vacuum degree was over 0.9 atm. This should be ascribed to the pressure buildup effect in the shell side. During VMD process, large amount of vapor would permeate through the membrane and be wiped out by the vacuum pipe. For the vapor close the vacuum ventilation, it could be separated from the membrane soon. For the vapor far from the vacuum ventilation, it would accumulate in the shell which would result in the pressure recovery of the shell side. This effect would be more obvious with the increase of the vacuum degree. Similar effect was also obtained in previous study (Sun et al., 2014). Therefore, considering energy consumption efficiency, the permeate side vacuum degree should be maintained at the upper turning point of the S-shape curve of membrane flux with permeate side vacuum degree. In the present study, it was identified as 0.9 atm.
be ascribed to the variation of the boundary layer in the lumen side. Due to the hydrophobicity of the membrane, non-volatile maters like Sr2+ would be excluded from passing the membrane during the process. Thus there would be a boundary layer between bulk solution and the membrane in which Sr2+concentration was higher. This would attenuate the partial pressure of the water vapor which related to the mass transfer driving force. Thus the boundary layer had a negative effect on membrane flux. The feed flow velocity could interfere with the flow turbulence of the lumen solution which could interfere with the thickness of the boundary layer. During the present study, the feed solution was in laminar flow state, indicating that the thickness of the boundary layer could be inverse proportion to the feed flow velocity in laminar flow state. Previous studies also observed the similar results (Liu and Wang, 2013; Zuo et al., 2016). Sun et al. (2014) also found that there was an interference limitation of feed flow velocity on the membrane flux. When the feed flow turned into turbulent flow, the membrane flux would tend to be steady, suggesting that the feed flow velocity had a positive effect on membrane flux by interfere the boundary layer thickness. 3.4. Purification efficiency of Sr2+ In this section, about 10 mg/L Sr2+ solution was prepared and used as the feed solution, and the experiment was performed at 70 °C, 41.8 L/h of feed flow velocity and 0.90 atm of permeate vacuum degree. Fig. 7 shows the variation of the Sr2+ concentration in feed and permeate solution. It could be seen that Sr2+ removal efficiency kept over 99.60% at all time and the concentration in Sr2+ maintained below 0.05 mg/L, indicating that VMD process should be an efficient method to process Sr2+ in the aqueous solution. The variation of membrane flux during the continuous operation was also presented in Fig. 7. During the operation, the membrane flux maintained over 6.71 Lm2h1. The stable membrane flux and high Sr2+ removal efficiency indicated the membrane fouling and wetting could be ignored. Table 2 lists several wastewater treatment studies with different kinds of MD methods. It could be found that for different types of wastewater, VMD methods could achieve high water purification efficiency.
3.3. Effect of feed flow velocity 3.5. Mathematical simulation of mass transfer during VMD process Dusty gas model (DGM) was used to simulate the mass transfer process. When characteristics of the membrane and operation
14 Sr (mg/L)
In this section, the feed flow velocity varied from 10.5 to 41.8 L/h, the corresponding Reynolds Number was from 100 to 400. Fig. 6 shows the variation of membrane flux with Reynolds Number. It could be seen that the membrane flux increased linearly with the feed flow velocity. The linear flux increase should
12 10 Sr of feed Sr of permeate
0.2 0.1
Flux (L · m-2 · h-1)
10 0.0 8 6 4 2 0 20
40
60
80
100
120
140
Time (min) Fig. 6. Variation of membrane flux with Reynolds Number.
Fig. 7. Variation of Sr
2+
concentration and membrane flux.
160
367
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368 Table 2 Various MD researches in wastewater treatment. MD type/membrane material
Pollutants
Rejection rate
Reference
VMD PP DCMD/PTFE VMD/PP DCMD/PVDF DCMD/PES & PS with SMMs DCMD/(PTFE, PVDF, PP)
Sr (II) Low-level radioactive wastewater Dye solution Simulated radioactive wastewater Radioactive wastewater Municipal wastewater
Present study Zakrzewska-Trznadel et al. (1999) Criscuoli et al. (2008) Liu and Wang (2013) Khayet (2013) Kim et al. (2015)
DCMD/PTFE DCMD/PTFE DCMD/PE VMD/PP
Saline dairy wastewater Biologically treated coking wastewater Seawater Organic wastewater
99.74–99.60% 97.72–99.98% 99.99% 99.99% 99.69–99.93% COD 98% TP 99.99% 99% >99.10% 99.90% 98%
Kezia et al. (2015) Li et al. (2016) Zuo et al. (2016) Wang et al. (2016)
Table 3 Comparison of the experimental and simulated results. Tfeed (°C)
Tpermeate (°C)
Ppermeate (Pa)
Reynolds number
Jexp (Lm2h1)
JDGM (Lm2h1)
Relative error (%)
ARE (%)
67.6 68.1 68.1 69.3 70.0
31.6 31.3 34.4 31.6 33.2
2133 1973 5053 1973 2406
300 400 400 400 400
1.49 2.41 3.69 6.15 7.15
1.37 2.32 3.93 6.56 7.10
1.50% 5.52% 7.99% 3.36% 8.18%
5.31%
parameters, including feed and permeate side temperature, permeate side vacuum degree and permeate temperature, were determined, the membrane flux of VMD process could be estimated as follows:
J¼
8 re 3 ds
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 DP 2pRMT
ð3Þ
where d is the membrane thickness (m); r is the average pore radius (m); e is the membrane porosity; s is the pore tortuosity; R is the universal gas constant (8.314 Jmol1K1); T is the average gas temperature inside membrane pores (K); DP is the pressure difference across the membrane surface (Pa). During the simulation, s was used as a modulating parameter to control the deviation. Table 3 shows the experimental (Jexp) and simulated results (JDGM) under different operating conditions, relative errors and average relative errors (ARE). It could be seen that ARE was 5.31%, indicating that DGM could simulate the mass transfer of VMD process quite well. 4. Conclusions (1) VMD process was an efficient process to remove Sr2+ from aqueous solution. In the continuous operation, Sr2+ removal efficiency could reach 99.60% with about 10 mg/L Sr2+ in the feed solution and the membrane flux could maintain at 6.71 Lm2h1. (2) A pressure build-up effect was observed during the VMD process, which could interfere with the membrane flux. Pressure build-up effect was more serious when permeate side vacuum degree was over 0.9 atm, leading to the membrane flux level-off phenomenon. Considering energy consumption efficiency, the permeate side vacuum degree should be maintained at the turning point of the membrane flux. In the present study, it was identified as 0.9 atm. (3) DGM could simulate the mass transfer of VMD process well with ARE of 5.31%. (4) Though VMD process has exhibited the potential in the treatment of Sr2+-containing wastewater, it should be noticed that further study on membrane fouling is needed to be performed in the future.
Acknowledgments The research was supported by the National S&T Major Project (2013ZX06002001), the National Key Research and Development Program (2016YFC1402507), and the Program for Changjiang Scholars and Innovative Research Team in University (IRT-13026). References Chen, C., Wang, J.L., 2010. Removal of heavy metal ions by waste biomass of Saccharomyces Cerevisiae. J. Environ. Eng. 136, 95–102. Chen, H., Shao, D.D., Li, J.X., Wang, X.K., 2014. The uptake of radionuclides from aqueous solution by poly (amidoxime) modified reduced graphene oxide. Chem. Eng. J. 254, 623–634. Chen, Y.W., Wang, J.L., 2012. Removal of radionuclide Sr2+ ions from aqueous solution using synthesized magnetic chitosan beads. Nucl. Eng. Des. 242, 445– 451. Chen, Z., Wu, Y., Wei, Y., Mimura, H., 2016. Preparation of silica-based titanate adsorbents and application for strontium removal from radioactive contaminated wastewater. J. Radioanal. Nucl. Chem. 307, 931–940. Criscuoli, A., Zhong, J., Figoli, A., Carnevale, M.C., Huang, R., Drioli, E., 2008. Treatment of dye solutions by vacuum membrane distillation. Water Res. 42, 5031–5037. Ding, S.Y., Yang, Y., Li, C., Huang, H.O., Hou, L.A., 2016. The effects of organic fouling on the removal of radionuclides by reverse osmosis membranes. Water Res. 95, 174–184. Ding, S.Y., Yang, Y., Huang, H.O., Liu, H.C., Hou, L.A., 2015. Effects of feed solution chemistry on low pressure reverse osmosis filtration of cesium and strontium. J. Hazard. Mater. 294, 27–34. El-Bourawi, M.S., Ding, Z., Ma, R., Khayet, M., 2006. A framework for better understanding membrane distillation separation process. J. Membr. Sci. 285, 4– 29. El-Kamash, A.M., 2008. Evaluation of zeolite A for the sorptive removal of Cs+ and Sr2+ ions from aqueous solutions using batch and fixed bed column operations. J. Hazard. Mater. 151, 432–445. Kezia, K., Lee, J., Weeks, M., Kentish, S., 2015. Direct contact membrane distillation for the concentration of saline dairy effluent. Water Res. 81, 167–177. Khayet, M., 2013. Treatment of radioactive wastewater solutions by direct contact membrane distillation using surface modified membranes. Desalination 321, 60–66. Kim, H.C., Shin, J.W., Won, S., Lee, J.Y., Maeng, S.K., Song, K.G., 2015. Membrane distillation combined with an anaerobic moving bed biofilm reactor for treating municipal wastewater. Water Res. 71, 97–106. Lawson, K.W., Lloyd, D.R., 1997. Membrane distillation. J. Membr. Sci. 124, 1–25. Li, J.F., Wu, J., Sun, H.F., Cheng, F.Q., Liu, Y., 2016. Advanced treatment of biologically treated coking wastewater by membrane distillation coupled with precoagulation. Desalination 380, 43–51. Liu, H.Y., Wang, J.L., 2013. Treatment of radioactive wastewater using direct contact membrane distillation. J. Hazard. Mater. 261, 307–315. Liu, H.Y., Wang, J.L., 2016. Separation of ammonia from radioactive wastewater by hydrophobic membrane contactor. Prog. Nucl. Energy 86, 97–102.
368
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
Park, Y.J., Lee, Y.C., Shin, W.S., Choi, S.J., 2010. Removal of cobalt, strontium and cesium from radioactive laundry wastewater by ammonium molybdophosphate–polyacrylonitrile (AMP–PAN). Chem. Eng. J. 162, 685–695. Sun, A.C.G., Kosar, W., Zhang, Y.F., Feng, X.S., 2014. Vacuum membrane distillation for desalination of water using hollow fiber membranes. J. Membr. Sci. 455, 131–142. Wang, J., Sun, X.G., Yuan, Y., Chen, H., Wang, H.Y., Hou, D.Y., 2016. A novel microwave assisted photo-catalytic membrane distillation process for treating the organic wastewater containing inorganic ions. J. Water Process Eng. 9, 1–8.
Wen, X., Li, F.Z., Zhao, X., 2016. Removal of nuclides and boron from highly saline radioactive wastewater by direct contact membrane distillation. Desalination 394, 101–107. Zakrzewska-Trznadel, G., Harasimowicz, M., Chmielewski, A.G., 1999. Concentration of radioactive components in liquid low-level radioactive waste by membrane distillation. J. Membr. Sci. 163, 257–264. Zuo, J., Bonyadi, S., Chung, T.S., 2016. Exploring the potential of commercial polyethylene membranes for desalination by membrane distillation. J. Membr. Sci. 497, 239–247.
Contents lists available at ScienceDirect
Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene
Removal of strontium ions from simulated radioactive wastewater by vacuum membrane distillation Fei Jia a, Junfeng Li a,b, Jianlong Wang a,b,⇑, Yuliang Sun a a b
Collaborative Innovation Center for Advanced Nuclear Energy Technology, INET, Tsinghua University, Beijing 100084, PR China Beijing Key Laboratory of Radioactive Waste Treatment, Tsinghua University, Beijing 100084, PR China
a r t i c l e
i n f o
Article history: Received 24 November 2016 Received in revised form 3 February 2017 Accepted 4 February 2017
Keywords: Vacuum membrane distillation Hollow fiber membrane Strontium Radioactive wastewater Dusty gas model
a b s t r a c t The removal of Sr2+ ions from simulated radioactive wastewater by vacuum membrane distillation (VMD) with hollow fiber membrane module was investigated. The removal efficiency of Sr2+ could maintain over 99.60%, and the membrane flux could maintain at 6.71 Lm2h1 when Sr2+ in the feed solution was about 10 mg/L. The effect of operating parameters on the membrane flux was examined, including feed temperature (30–70 °C), permeate side vacuum (0.10–0.98 atm) and feed flow velocity (10.5–41.8 L/h). Pressure buildup effect was observed during the process, which could interfere with the membrane flux. Pressure buildup effect was more serious when permeate side vacuum degree was over 0.9 atm. Considering energy consumption efficiency, the permeate side vacuum degree should be maintained at the turning point (0.9 atm) of the membrane flux. Dusty gas model could simulate the mass transfer of VMD process well with ARE (average relative error) of 5.31%. VMD is potential for the removal of Sr2+ ions from aqueous solution. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction After the tragedy of Fukushima Daiichi, the treatment of radioactive wastewater has received increasing attention. As one of main fission products of nuclear reactor, strontium has been regarded as a hazardous radionuclide with half-life of about 30 years (El-Kamash, 2008). A series of methods have been investigated to remove Sr2+ from aqueous solution, including adsorption (Chen and Wang, 2010, 2012; Park et al., 2010; Chen et al., 2014; Chen et al., 2016), ion exchange (El-Kamash, 2008) and membrane technology (Liu and Wang, 2013; Ding et al., 2015, 2016), etc. Compared with other methods, membrane technology could achieve both high salt removal efficiency and high concentration factor (Lawson and Lloyd, 1997). Among all membrane technologies, membrane distillation (MD) has some features in terms of its unique purification mechanism. Unlike series of pressure-driven membrane methods, MD employs the temperature difference across the membrane as the mass transfer driving force (Lawson and Lloyd, 1997). The membrane of MD system is made from hydrophobic materials which only allow volatile matters to pass through the membrane during the separation process (Lawson and Lloyd, 1997). Non-volatile matters like Sr2+ will be remained ⇑ Corresponding author at: Energy Science Building, Tsinghua University, Beijing 100084, PR China. E-mail address: [email protected] (J. Wang). http://dx.doi.org/10.1016/j.anucene.2017.02.003 0306-4549/Ó 2017 Elsevier Ltd. All rights reserved.
in the original wastewater. In previous research, we confirmed that MD could achieve high nuclide removal efficiency (Liu and Wang, 2013). Compared with pressure-driving membrane method, MD has several advantages including milder and safer operation condition, capability in processing high salinity wastewater, potential to use low quality heat sources like waste heat in nuclear power plant as mass transfer driving force (Lawson and Lloyd, 1997; Zuo et al., 2016). Classified by the membrane configuration, MD process could be categorized into four types: direct contact membrane distillation (DCMD), air gap membrane distillation (AGMD), sweeping gas membrane distillation (SGMD) and vacuum membrane distillation (VMD). DCMD is the original MD type with the simplest structure. However, as the permeate solution contacts with the membrane directly, the permeate will be polluted directly if there is a damage on the membrane or membrane wetting. The rest of three MD types decrease the risk of permeate pollution by separate membrane and permeate. For VMD method, it creates a vacuum atmosphere in the shell side of the membrane column to strip the permeate vapor from the shell. Besides decreasing the influence of membrane wetting, the application of vacuum also increase the mass transfer driving force of the system, which could increase the membrane flux of VMD process. There are three types of membrane column for VMD method: flat sheet membrane, spiral wound membrane and hollow fiber membrane (El-Bourawi et al., 2006).
364
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
Nomenclature A AGMD DCMD DGM Dm
DP Dt d
e J Jexp JDGM MD P
total effective membrane area (m2) air gap membrane distillation direct contact membrane distillation dusty gas model mass increase of the permeate collected in the condense (g) pressure difference across the membrane surface (Pa) sampling time (h) membrane thickness (m) membrane porosity permeate flux (Lm2h1) actual permeate flux (Lm2h1) simulated permeate flux (Lm2h1) membrane distillation mean pressure in membrane pores
Among these three types, hollow fiber type could achieve the highest surface density area. This results in the more compact system of the membrane module, which increase the real application potential of hollow fiber VMD method. There are a series of researches on the treatment of radioactive wastewater with MD method (Zakrzewska-Trznadel et al., 1999; Khayet, 2013; Liu and Wang, 2013, 2016; Wen et al., 2016). However, there were few researches focusing on Sr removal by VMD method. As discussed above, VMD has a potential to be an effective method to remove Sr2+ ions from solution. The objective of this study was to investigate the treatment of Sr2+-containing radioactive wastewater by hollow fiber VMD method. The effect of operating parameters on the performance of VMD process, including feed temperature, shell side vacuum degree and feed flow rate was studied. The mathematical simulation of VMD mass transfer process was discussed to predict the membrane flux.
Psat PP
q r R SGMD T Tf Tf,m Tp Tp,m
s VMD
pure water vapor saturated pressure polypropylene density of permeate solution in the present study (kgL1) average pore radius (m) universal gas constant (8.314 Jmol1K1) sweeping gas membrane distillation mean temperature in membrane pores (K) temperature of feed side (K) temperature of membrane surface in feed side (K) temperature of permeate side (K) temperature of membrane surface in permeate side (K) pore tortuosity of membrane vacuum membrane distillation
During VMD process, the solution would be heated to the preset temperature at first and then be pumped into the membrane module. At the shell side of the module, the vacuum pump would create a vacuum atmosphere and the permeate vapor would be stripped away from the shell and be condensed into liquid in the condenser which was filled with chilled water. Then the rest of solution in the lumen side would be recycled into the container to supplement the heat loss. 2.3. Analytical methods Sr ions concentration was analyzed by flame AAS (Hitachi, ZA3000, Japan). Membrane flux of MD was obtained by detecting water mass collected in the condenser with precision balance. The membrane flux was deduced as follows:
J¼
Dm
qADt
ð1Þ
where J is the membrane flux (Lm2h1); Dm is the water mass collected in the condense (g); q is water in the present study (gm3); A is the membrane area (m2); Dt is the sampling interval (h).
2. Materials and methods 2.1. Chemicals and membrane modules SrCl2 (Sinopharm Chemical, China; analytical pure) solution of about 10 mg/L was employed as feed in this experiment. Hollow fiber membrane made from polypropylene (PP) (Wochi, WHPP96-21, China) was applied in the VMD process. Property of the membrane is listed in detail in Table 1. 2.2. Experimental set-up The VMD system was made up of feed container, feed heating system, MD module, condenser system, vacuum pump, peristaltic pumps, precision balance, flowmeter and thermometers (Fig. 1).
Table 1 Performance of the membrane module. Parameters
Units
Value
Hollow fiber number Effective membrane length Total effective inner area Packing density Porosity Mean pore diameter Inner radius of the fiber Outer radius of the fiber
– mm m2 – – lm mm mm
140 140 0.062 0.35 60% 0.18 0.50 1.36
3. Results and discussion 3.1. Effect of feed temperature In this section, feed temperature was varied from 30 to 70 °C, with 0.98 atm of permeate vacuum degree and 41.8 L/h of feed flow velocity. Fig. 2 shows the variation of permeate at different feed temperatures. It could be seen that at all temperatures, the permeate mass increased linearly with time, indicating that the VMD system reached the steady state for producing water. Therefore the sampling time was determined to be 10 min. Fig. 3 shows the variation of the membrane flux with feed temperature. It could be seen that the membrane flux increased exponentially (R2 = 0.9762) with the feed temperature, which could be elucidated by the variation of saturated vapor pressure in the lumen side. During the VMD process, the pressure difference across the membrane played as the mass transfer driving force. The saturated vapor pressure of lumen side Psat (Pa) at temperature of T(K) could be calculated by Antoine Equation:
3816:44 log Psat ðTÞ ¼ exp 23:1964 þ 46:13 T
ð2Þ
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
365
Fig. 1. Schematic diagram of VMD system used in this study.
Fig. 2. Variation of permeate mass at different feed temperatures.
Fig. 4. Variation of membrane flux with the mass transfer driving force.
membrane flux and the mass transfer driving force could be observed. Previous researchers also obtained the similar results (Liu and Wang, 2013; Zuo et al., 2016). 3.2. Effect of permeate side vacuum degree
Fig. 3. Variation of the membrane flux with feed temperature.
As shown in the equation, the pressure of the lumen side would increase exponentially with the temperature. Therefore, the mass transfer driving force would behave the same with the temperature. Fig. 4 shows the relationship of membrane flux with the mass transfer driving force. A linear relationship (R2 = 0.9953) between
In this section, the permeate side vacuum degree varied from 0.1 to 0.98 atm. The experiment was performed at 70 °C and 41.8 L/h of feed flow velocity. Fig. 5 shows the variation of membrane flux with permeate side vacuum degree. It could be observed that the membrane flux varied in S shape with permeate vacuum. The flux maintained about zero before 0.6 atm of vacuum degree and started to increase linearly from 0.6 to 0.9 atm while leveled off at about 6.5 Lm2h1 after 0.9 atm. For zero flux before vacuum degree of 0.6 atm, it could be ascribed to the feed solution saturation vapor pressure. According to Antoine Equation, at 70 °C, Psat of the lumen side could be about 31,000 Pa, 0.3 atm. Hence only when the permeate side vacuum degree was over 0.7 atm, there could be a positive driving force across the membrane. However, this turning point shifted to 0.6 atm in the study. This could be due to the determination method of feed temperature. In the study, the feed temperature was determined by the average temperature of the inlet and outlet temperature of the membrane module. When feed temperature was identified as 70 °C, the inlet
366
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
Fig. 5. Variation of membrane flux with permeate side vacuum degree.
and outlet temperature were 75 °C and 65 °C respectively. Thus in the membrane column there was a part where the temperature was over 70 °C, indicating that in this part the mass transfer driving force should be positive. Therefore, the zero flux turning point occurred before the theoretical value. Another notable phenomenon was the membrane flux level-off when vacuum degree was over 0.9 atm. This should be ascribed to the pressure buildup effect in the shell side. During VMD process, large amount of vapor would permeate through the membrane and be wiped out by the vacuum pipe. For the vapor close the vacuum ventilation, it could be separated from the membrane soon. For the vapor far from the vacuum ventilation, it would accumulate in the shell which would result in the pressure recovery of the shell side. This effect would be more obvious with the increase of the vacuum degree. Similar effect was also obtained in previous study (Sun et al., 2014). Therefore, considering energy consumption efficiency, the permeate side vacuum degree should be maintained at the upper turning point of the S-shape curve of membrane flux with permeate side vacuum degree. In the present study, it was identified as 0.9 atm.
be ascribed to the variation of the boundary layer in the lumen side. Due to the hydrophobicity of the membrane, non-volatile maters like Sr2+ would be excluded from passing the membrane during the process. Thus there would be a boundary layer between bulk solution and the membrane in which Sr2+concentration was higher. This would attenuate the partial pressure of the water vapor which related to the mass transfer driving force. Thus the boundary layer had a negative effect on membrane flux. The feed flow velocity could interfere with the flow turbulence of the lumen solution which could interfere with the thickness of the boundary layer. During the present study, the feed solution was in laminar flow state, indicating that the thickness of the boundary layer could be inverse proportion to the feed flow velocity in laminar flow state. Previous studies also observed the similar results (Liu and Wang, 2013; Zuo et al., 2016). Sun et al. (2014) also found that there was an interference limitation of feed flow velocity on the membrane flux. When the feed flow turned into turbulent flow, the membrane flux would tend to be steady, suggesting that the feed flow velocity had a positive effect on membrane flux by interfere the boundary layer thickness. 3.4. Purification efficiency of Sr2+ In this section, about 10 mg/L Sr2+ solution was prepared and used as the feed solution, and the experiment was performed at 70 °C, 41.8 L/h of feed flow velocity and 0.90 atm of permeate vacuum degree. Fig. 7 shows the variation of the Sr2+ concentration in feed and permeate solution. It could be seen that Sr2+ removal efficiency kept over 99.60% at all time and the concentration in Sr2+ maintained below 0.05 mg/L, indicating that VMD process should be an efficient method to process Sr2+ in the aqueous solution. The variation of membrane flux during the continuous operation was also presented in Fig. 7. During the operation, the membrane flux maintained over 6.71 Lm2h1. The stable membrane flux and high Sr2+ removal efficiency indicated the membrane fouling and wetting could be ignored. Table 2 lists several wastewater treatment studies with different kinds of MD methods. It could be found that for different types of wastewater, VMD methods could achieve high water purification efficiency.
3.3. Effect of feed flow velocity 3.5. Mathematical simulation of mass transfer during VMD process Dusty gas model (DGM) was used to simulate the mass transfer process. When characteristics of the membrane and operation
14 Sr (mg/L)
In this section, the feed flow velocity varied from 10.5 to 41.8 L/h, the corresponding Reynolds Number was from 100 to 400. Fig. 6 shows the variation of membrane flux with Reynolds Number. It could be seen that the membrane flux increased linearly with the feed flow velocity. The linear flux increase should
12 10 Sr of feed Sr of permeate
0.2 0.1
Flux (L · m-2 · h-1)
10 0.0 8 6 4 2 0 20
40
60
80
100
120
140
Time (min) Fig. 6. Variation of membrane flux with Reynolds Number.
Fig. 7. Variation of Sr
2+
concentration and membrane flux.
160
367
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368 Table 2 Various MD researches in wastewater treatment. MD type/membrane material
Pollutants
Rejection rate
Reference
VMD PP DCMD/PTFE VMD/PP DCMD/PVDF DCMD/PES & PS with SMMs DCMD/(PTFE, PVDF, PP)
Sr (II) Low-level radioactive wastewater Dye solution Simulated radioactive wastewater Radioactive wastewater Municipal wastewater
Present study Zakrzewska-Trznadel et al. (1999) Criscuoli et al. (2008) Liu and Wang (2013) Khayet (2013) Kim et al. (2015)
DCMD/PTFE DCMD/PTFE DCMD/PE VMD/PP
Saline dairy wastewater Biologically treated coking wastewater Seawater Organic wastewater
99.74–99.60% 97.72–99.98% 99.99% 99.99% 99.69–99.93% COD 98% TP 99.99% 99% >99.10% 99.90% 98%
Kezia et al. (2015) Li et al. (2016) Zuo et al. (2016) Wang et al. (2016)
Table 3 Comparison of the experimental and simulated results. Tfeed (°C)
Tpermeate (°C)
Ppermeate (Pa)
Reynolds number
Jexp (Lm2h1)
JDGM (Lm2h1)
Relative error (%)
ARE (%)
67.6 68.1 68.1 69.3 70.0
31.6 31.3 34.4 31.6 33.2
2133 1973 5053 1973 2406
300 400 400 400 400
1.49 2.41 3.69 6.15 7.15
1.37 2.32 3.93 6.56 7.10
1.50% 5.52% 7.99% 3.36% 8.18%
5.31%
parameters, including feed and permeate side temperature, permeate side vacuum degree and permeate temperature, were determined, the membrane flux of VMD process could be estimated as follows:
J¼
8 re 3 ds
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 DP 2pRMT
ð3Þ
where d is the membrane thickness (m); r is the average pore radius (m); e is the membrane porosity; s is the pore tortuosity; R is the universal gas constant (8.314 Jmol1K1); T is the average gas temperature inside membrane pores (K); DP is the pressure difference across the membrane surface (Pa). During the simulation, s was used as a modulating parameter to control the deviation. Table 3 shows the experimental (Jexp) and simulated results (JDGM) under different operating conditions, relative errors and average relative errors (ARE). It could be seen that ARE was 5.31%, indicating that DGM could simulate the mass transfer of VMD process quite well. 4. Conclusions (1) VMD process was an efficient process to remove Sr2+ from aqueous solution. In the continuous operation, Sr2+ removal efficiency could reach 99.60% with about 10 mg/L Sr2+ in the feed solution and the membrane flux could maintain at 6.71 Lm2h1. (2) A pressure build-up effect was observed during the VMD process, which could interfere with the membrane flux. Pressure build-up effect was more serious when permeate side vacuum degree was over 0.9 atm, leading to the membrane flux level-off phenomenon. Considering energy consumption efficiency, the permeate side vacuum degree should be maintained at the turning point of the membrane flux. In the present study, it was identified as 0.9 atm. (3) DGM could simulate the mass transfer of VMD process well with ARE of 5.31%. (4) Though VMD process has exhibited the potential in the treatment of Sr2+-containing wastewater, it should be noticed that further study on membrane fouling is needed to be performed in the future.
Acknowledgments The research was supported by the National S&T Major Project (2013ZX06002001), the National Key Research and Development Program (2016YFC1402507), and the Program for Changjiang Scholars and Innovative Research Team in University (IRT-13026). References Chen, C., Wang, J.L., 2010. Removal of heavy metal ions by waste biomass of Saccharomyces Cerevisiae. J. Environ. Eng. 136, 95–102. Chen, H., Shao, D.D., Li, J.X., Wang, X.K., 2014. The uptake of radionuclides from aqueous solution by poly (amidoxime) modified reduced graphene oxide. Chem. Eng. J. 254, 623–634. Chen, Y.W., Wang, J.L., 2012. Removal of radionuclide Sr2+ ions from aqueous solution using synthesized magnetic chitosan beads. Nucl. Eng. Des. 242, 445– 451. Chen, Z., Wu, Y., Wei, Y., Mimura, H., 2016. Preparation of silica-based titanate adsorbents and application for strontium removal from radioactive contaminated wastewater. J. Radioanal. Nucl. Chem. 307, 931–940. Criscuoli, A., Zhong, J., Figoli, A., Carnevale, M.C., Huang, R., Drioli, E., 2008. Treatment of dye solutions by vacuum membrane distillation. Water Res. 42, 5031–5037. Ding, S.Y., Yang, Y., Li, C., Huang, H.O., Hou, L.A., 2016. The effects of organic fouling on the removal of radionuclides by reverse osmosis membranes. Water Res. 95, 174–184. Ding, S.Y., Yang, Y., Huang, H.O., Liu, H.C., Hou, L.A., 2015. Effects of feed solution chemistry on low pressure reverse osmosis filtration of cesium and strontium. J. Hazard. Mater. 294, 27–34. El-Bourawi, M.S., Ding, Z., Ma, R., Khayet, M., 2006. A framework for better understanding membrane distillation separation process. J. Membr. Sci. 285, 4– 29. El-Kamash, A.M., 2008. Evaluation of zeolite A for the sorptive removal of Cs+ and Sr2+ ions from aqueous solutions using batch and fixed bed column operations. J. Hazard. Mater. 151, 432–445. Kezia, K., Lee, J., Weeks, M., Kentish, S., 2015. Direct contact membrane distillation for the concentration of saline dairy effluent. Water Res. 81, 167–177. Khayet, M., 2013. Treatment of radioactive wastewater solutions by direct contact membrane distillation using surface modified membranes. Desalination 321, 60–66. Kim, H.C., Shin, J.W., Won, S., Lee, J.Y., Maeng, S.K., Song, K.G., 2015. Membrane distillation combined with an anaerobic moving bed biofilm reactor for treating municipal wastewater. Water Res. 71, 97–106. Lawson, K.W., Lloyd, D.R., 1997. Membrane distillation. J. Membr. Sci. 124, 1–25. Li, J.F., Wu, J., Sun, H.F., Cheng, F.Q., Liu, Y., 2016. Advanced treatment of biologically treated coking wastewater by membrane distillation coupled with precoagulation. Desalination 380, 43–51. Liu, H.Y., Wang, J.L., 2013. Treatment of radioactive wastewater using direct contact membrane distillation. J. Hazard. Mater. 261, 307–315. Liu, H.Y., Wang, J.L., 2016. Separation of ammonia from radioactive wastewater by hydrophobic membrane contactor. Prog. Nucl. Energy 86, 97–102.
368
F. Jia et al. / Annals of Nuclear Energy 103 (2017) 363–368
Park, Y.J., Lee, Y.C., Shin, W.S., Choi, S.J., 2010. Removal of cobalt, strontium and cesium from radioactive laundry wastewater by ammonium molybdophosphate–polyacrylonitrile (AMP–PAN). Chem. Eng. J. 162, 685–695. Sun, A.C.G., Kosar, W., Zhang, Y.F., Feng, X.S., 2014. Vacuum membrane distillation for desalination of water using hollow fiber membranes. J. Membr. Sci. 455, 131–142. Wang, J., Sun, X.G., Yuan, Y., Chen, H., Wang, H.Y., Hou, D.Y., 2016. A novel microwave assisted photo-catalytic membrane distillation process for treating the organic wastewater containing inorganic ions. J. Water Process Eng. 9, 1–8.
Wen, X., Li, F.Z., Zhao, X., 2016. Removal of nuclides and boron from highly saline radioactive wastewater by direct contact membrane distillation. Desalination 394, 101–107. Zakrzewska-Trznadel, G., Harasimowicz, M., Chmielewski, A.G., 1999. Concentration of radioactive components in liquid low-level radioactive waste by membrane distillation. J. Membr. Sci. 163, 257–264. Zuo, J., Bonyadi, S., Chung, T.S., 2016. Exploring the potential of commercial polyethylene membranes for desalination by membrane distillation. J. Membr. Sci. 497, 239–247.