Recovery of Water Using Other Membrane Technologies

As shown in Table 9.4, there are several other types of membrane technologies based on other driving forces instead of pressure. This section of Chapter 9 discusses and gives typical examples of the use of two very important membranes in the treatment of AMD - electrodialysis and membrane distillation (Abhanga et al., 2013).

9.3.2.1 Concepts of Electrodialysis and Membrane Distillation Technologies

Amongst the different technologies, electrodialysis has been tested and proven to be an effective technology for water recovery from acidic solutions (Wisniewski and Wisniewska, 1999; Cifuentes et al., 2006; Agrawal and Sahu, 2009; Cifuentes et al., 2009). Electrodialysis is a membrane process where the driving force is electrochemical and ions are transported across a water swollen ion-exchange membrane under the influence of electrical potential (Mortazavi, 2008). Rodrigues et al. (2008) define electrodialysis as a membrane-separation process based on the selective migration of aqueous ions through ion-exchange membrane as a result of an electrical driving force. In this technology, the direction of transport and rate of each ion depends on its charge, mobility, solution conductivity, relative concentrations, applied voltage, etc. (Rodrigues et al., 2008; Benvenut et al., 2013). Electrodialysis provides a means of selective separation of anions and cations and is considered as a clean technology which does not require the addition of chemicals; it can be operated in continuous mode and allows obtaining profitable by-products (Mortazavi, 2008; Marti-Calatayud et al., 2013).

Membrane distillation process is an over five-decade old technique which is emerging commercially and is currently being explored for various applications including concentrating acid and recover fresh water from acidic waste solutions, desalination, water and wastewater treatment, removal of volatile organic compounds, and food processing (Tomaszewska, 2000; Tomaszewska et al., 2001; Alkhudhiri et al., 2012; Kesieme et al., 2012; Camacho, 2013; Shirazi et al., 2013; Simate and Ndlovu, 2014; Kesieme, 2015). Furthermore, Kesieme et al. (2012) confirm that membrane distillation has been used to recover water and concentrate acid and metal values from mining wastewater and process solutions. Ideally, the process combines both the conventional distillation and membrane-separation processes (Shirazi et al., 2014a).

Membrane distillation is a thermally driven separation process that utilises hydrophobic, microporous membranes as a contactor (Shirazi et al., 2014b). The driving force in the membrane distillation is the vapour pressure difference induced by the temperature difference across the hydrophobic membrane (Alkhudhiri et al., 2012). In other words, the temperature difference existing across the membrane results in a vapour pressure difference, thus vapour molecules are transported from the high-vapour pressure side to the low- vapour pressure side through the pores of the membrane (Tomaszewska, 2000). The hydrophobicity of the membrane prevents the transport of liquid across the pores of the partition while water vapour can be transported from the warm side, and condensing at the cold surface (Kesieme et al., 2012). Since the separation mechanism is based on the vapour/liquid equilibrium, it means that the component with the highest partial pressure will exhibit the highest permeation rate (Tomaszewska, 2000). Moreover, mass transfer in membrane distillation is controlled by three basic mechanisms - Knudsen diffusion, Poiseuille flow (viscous flow), and molecular diffusion (Alkhudhiri et al., 2012). Hull and Zodrow (2017) explain the operation of membrane distillation in simple terms as follows for the case of AMD. In membrane distillation, heated AMD is separated from a cooled distillate by a hydrophobic, water-excluding membrane. Due to the fact that water only passes through the membrane in the vapour phase, non-volatile sulphates and heavy metals are retained in the concentrate stream.

Membrane distillation processes have several configurations which are as follows (Kesieme et al., 2012): (1) direct contact membrane distillation (DCMD), (2) air gap membrane distillation (AGMD), (3) sweeping gas membrane distillation (SGMD), and (4) vacuum membrane distillation (VMD). The different membrane distillation configurations are represented in Figure 9.4 (Foureaux et al., 2020).

Among these configurations, DCMD is the most widely used because it is convenient to set up, consumes relatively low energy, gives high water flux (Kesieme et al., 2012), and has 100% (theoretical) macromolecules rejection

FIGURE 9.4

Different membrane distillation configurations (Foureaux et al., 2020). (a) direct contact membrane distillation (DCMD); (b) air gap membrane distillation (AGMD); (c) sweep gas membrane distillation (SGMD); and (d) vacuum membrane distillation (VMD). HE01: Heat exchanger. P01: Vacuum pump. Hot and cold side represented as red and blue streams, respectively.

ability, in addition to non-volatile compounds and inorganic ions (Foureaux et al., 2020). Membrane distillation processes, in general, have several advantages including the following: (i) the ability of 100% (theoretical) rejection of non-volatile compounds and inorganic ions; (ii) the possibility of operating at a relatively low temperature when compared to the conventional distillation process; (iii) lower operating pressures when compared to the classical membrane-separation processes that have the pressure gradient as the driving force and (iv) less influence of the fouling phenomenon due to less chemical interaction between process solution and membrane surface (Wang and Chung, 2015; Manna and Pal, 2016; Biniaz et al., 2019; Foureaux et al., 2020). Moreover, membrane-separation processes can operate without pre-treatment since it is able to treat a high solute concentration in the feed stream and it is not so sensitive to polarisation concentration effects like other classical membrane-separation processes (Alkhudhiri and Hilal, 2018; Biniaz et al., 2019; Foureaux et al., 2020).

9.3.2.2 Selected Typical Studies of Electrodialysis and Membrane Distillation Technologies

In a study of the use of electrodialysis, Buzzi et al. (2013) investigated the possibility of employing the technique to treat AMD generated by the mining of coal for water recovery purposes. The AMD samples that were used in the study were collected from a carboniferous area in Criciiima, Brazil at different locations throughout the carboniferous area to represent different situations where AMD is generated. After collection, any solids in each of the six AMD samples were allowed to settle. Thereafter, each sample was filtered through a membrane of 0.45-pm pores before the samples were chemically characterised for the following parameters: pH, conductivity, and levels of Na+, K+, Mg2-, Ca2+, Fe3+, Cu2+, Zn2+, Mn2+, Fe2+, F~, CT, N03-, and S042-. The electrodialysis experiments were conducted in a laboratory cell with five compartments. The results of the study by Buzzi et al. (2013) showed that electrodialysis is suitable for recovering water from AMD, with contaminant (e.g., Fe, Al, Mn, Pb, Zn, Cu, etc.) removal efficiencies that are greater than 97%. However, the precipitation of iron at the surface of the cation-exchange membrane constitutes a problem for the system because it causes a blockage of the membrane through the scaling phenomenon, which reduces the process efficiency. It is proposed that the elimination of iron from AMD prior to electrodialysis process would facilitate smooth recovery of water from the AMD.

In a 2011 study, Buzzi et al. conducted experiments to ascertain the possibility of using MF membrane, as pre-treatment, followed by electrodialysis technique for the treatment of AMD aimed at obtaining water for reuse. The AMD sample used in this experiment was collected from a carboniferous area in Criciuma in Brazil. The pluvial drainage from the coal was a more diluted drainage when compared to the drainage from the percolation of deposits of wastewater in the same carboniferous area. The AMD sample in the study was analysed for the following chemical compositions: pH, conductivity, total dissolved solids, Na+, K+, Mg2+, Ca2+, Fe3+, Cu2+, Zn2+, Mn2+, Fe2+, F‘, Cl', N03‘ and S042". One pre-treatment with MF was necessary to prevent fouling and scaling of the electrodialysis, so the entire AMD sample used in the experiment was filtered through a membrane of 0.45 pm before the experiments. Electrodialysis tests were conducted in a laboratory cell with five compartments. The results showed that pre-treatment of AMD with MF membrane combined with electrodialysis was efficient to extract over 97.5% of cations and anions after 55 h when 2.6 mAcnr2 current density was applied. The results indicated that water could be recovered by electrodialysis from AMD after a single pre-treatment with a MF membrane.

Zheng et al. (2015) used a laboratory scale electrodialysis system with an effective area of 88 cm2 to remove copper and cyanide in simulated and real gold mine effluents. The electrodialysis stack consisted of two electrodes made of a titanium plate coated with ruthenium, five anion-exchange membranes, and six cation-exchange membranes. The gold mine effluent with a concentration of 47 mg/L and a cyanide concentration of 242 mg/L was provided by Zhaoyuan gold smelter plant (Shandong, China), whereas the simulated solutions were made from CuCN and NaCN. The concentrations of total copper and cyanide in simulated solutions, in mg/L, were as follows: sample 1 (Cu = 23.5, CN = 121), sample 2 (Cu = 47, CN = 242), sample 3 (Cu = 70.5, CN = 363), and sample 4 (Cu = 94, CN = 484). The effects of applied voltage, initial concentration, and flux rate on the removal rate of copper and cyanide were investigated. The results showed that the highest copper (99.41%) and cyanide (99.83%) removal rates were achieved under the following conditions: applied voltage of 25 V, initial concentration for copper and cyanide of 47 mg L 1 and 242 mg L 1 (sample 2 and real AMD), and a flux rate of 4.17 mL s'1. The results of the study also showed that the lowest concentrations of copper (0.44 mg L '), cyanide (0.48 mg L'1), and zinc (0.34 mg L ') found in the treated effluent water were all below regulatory limits (copper, cyanide < 0.5 mg L *, zinc < 2.0 mg L4) which makes it suitable for reuse. In addition, ion-exchange membrane fouling was studied and the results showed the presence of CuCN, [Cu(CN)3]2~, Cu(OH)2, and Zn(OH)2 in the precipitate, and the fouling of anion-exchange membranes could be decreased significantly via pH adjustment.

The recovery of water has also been studied in other acidic systems containing various metals (Wisniewski and Wisniewska, 1999; Agrawal and Sahu, 2009; Cifuentes et al., 2009; Benvenut et al., 2013). In all these studies, electrodialysis has been found to be an effective method for water recovery. In fact, membrane modules on average have a water recovery rate ranging from 50% to 80% (West et al., 2011). However, the only major disadvantage is that electrodialysis membrane technologies can be very expensive in applications where the wastewater contains elevated hardness and sulphate near gypsum saturation (West et al., 2011). The high cost of operation is due to the large reagent requirement for upstream pre-treatments, high power demand, and expensive disposal options for the concentrated brine solution produced by the membranes. At a time when responsible energy and environmental practices are under the spotlight, technologies with high energy use and that produce a large volume of wastewater are becoming less appealing.

In a study published almost a decade ago, Kesieme et al. (2014) claimed to have carried out experiments for the first time using DCMD for acid and water recovery from a real leach solution generated by a hydrometallurgical plant. In other words, the aim of the study conducted by Kesieme et al. (2014) was to assess the opportunity of using DCMD to recover fresh water and acids from real acid leach solutions generated from hydrometallurgical plants. Two different real leach solutions containing HC1 or H2S04 were obtained from a Jervois Mining process plant in Melbourne, Australia. The membranes used were flat sheet polytetrafluoroethylene (PTFE) supported on polypropylene scrim backing. The membranes had an active area of 0.0169 m2, pore size of 0.45 pm. A cartridge filter with filtration size of 0.5 pm was used on the hot loop to collect precipitated matter prior to entering the MD module. The flow rate into the hot and cold sides of the module was 900 mL/min. The feed temperature was 60°C and the cold temperature was maintained at 20°C. For the solution containing non-volatile substances only water vapour was transferred across the membrane and the non-volatile compounds such as H2S04 were retained by the membrane. Solutions containing volatiles compounds such as HC1 and water vapour could pass through the membrane as permeate. The vapour (or permeate) was condensed directly into the solution (distillate) in which HC1 was dissolved. The HC1 flux was calculated from the material balance of HC1 in the distillate collected every hour taking into account the changes in volume and the acid concentration in the distillate. The water flux was calculated based on Equation 9.16, and recovery was calculated as shown in Equation 9.17.

The results of the test work by Kesieme et al. (2014) showed that fluxes were within the range of 18-33 kg/m2/h and 15-35 kg/m2/h for the H2S04 and HC1 systems, respectively. In the H2S04 leach system, the final concentration of free acid in the sample solution increased on the concentrate side of the DCMD system from 1.04 M up to 4.60 M. The sulphate separation efficiency was over 99.9% and overall water recovery exceeded 80%. In the HC1 leach system, HC1 vapour passed through the membrane from the feed side to the permeate side. The concentration of HC1 captured in the permeate side was about 1.10 M leaving behind only 0.41 M in the feed from the initial concentration of 2.13 M. In all the experiments, salt rejection was > 99.9%. The results of this study clearly showed that DCMD was viable for high recovery of high water quality. The concentrated H2S04 and metals remaining in the feed may be selectively recovered using solvent extraction. The HC1 can be recovered for reuse using only DCMD.

In another study by Kesieme and Aral (2015) which acted as a follow-up to the 2014 study, an assessment of the potential and opportunities for DCMD to concentrate H2S04 and recover fresh water from acidic process solutions was conducted. The study was also aimed at identifying how membrane distillation can work in combination with solvent extraction in the mineral processing industry for acid recovery. Table 9.7 shows the acid and metal compositions to the membrane distillation and solvent extraction systems used in the study. Experiments were conducted in DCMD mode to confirm the viability of membrane distillation to concentrate a 4-L synthetic acidic waste solution and to recover fresh water. The membrane had an active area of 0.0169 m2 with pore size of 0.45 pm, and a cartridge filter with filtration size of 0.5 pm was used on the hot loop to collect precipitated matter prior to entering the membrane distillation module. The flow rate into the hot and cold sides of the module was 900 mL/min. The temperature on the hot side of the membrane distillation module was 60°C and the cold side temperature was 20°C. Permeate build-up was measured by the accumulated mass of water in the permeate tank.

The organic system consisting of 50% tris-2-ethylhexylamine (TEHA) and 10% ShellSol A150 (a 100% aromatic diluent) in octanol was used in the

TABLE 9.7

The Acid and Metal Compositions to the Membrane Distillation and Solvent Extraction Systems

Species

Feed to the Membrane Distillation System

Feed to the Solvent Extraction System Mimicking Concentrate from Membrane Distillation

Acid (M)

0.850

2.450

Aluminium (g/L)

0.056

0.250

Cobalt (g/L)

0.071

0.289

Copper (g/L)

0.269

1.040

Calcium (g/L)

0.218

0.307

Iron (g/L)

2.780

11.350

Magnesium (g/L)

0.050

0.220

Manganese (g/L)

0.002

0.008

Nickel (g/L)

0.065

0.259

Source: Kesieme and Aral, 2015.

solvent extraction system. The feed composition mimicking an acidic process solution after concentration using membrane distillation (see Table 9.7) was made by dissolving AR grade 245 g/L H2S04 and sulphates of metals including Fe, Ni, Zn, Mg, Co, and Cu in distilled water. All batch solvent extraction tests were carried out in 100-mL hexagonal glass vessels immersed in a temperature-controlled water bath. The solution temperature was maintained at the desired temperature (±1°C) during testing. For the acid extraction test, the organic system was mixed with concentrated solution at an A/О ratio of 1:2 and a temperature of 22°C. The loaded organic solution was stripped twice at О/A ratios of 2:1 and 1:5 at 60°C. The raffinate and the loaded strip liquors were titrated to determine acid concentrations for extraction and mass balance calculations.

The results of the experiments by Kesieme and Aral (2015) confirmed that the membrane distillation was capable of concentrating H2S04 and recover fresh water from process acidic solutions. The DCMD experiment showed that H2S04 was concentrated from 0.85 to 4.44 M whereas the water recovery exceeded 80%. The sulphate and metal separation efficiency was >99.99%. After recovery of water with DCMD, over 80% of H2S04 was extracted in the solvent extraction system in a single contact from the waste solution (i.e., the concentrated solution from the membrane distillation) containing 245 g/L H2S04 and metals with various concentrations. After three stages of successive extraction, nearly 99% of acid was extracted, leaving only 2.4 g/L H2S04 in the raffinate. The extracted acid was easily stripped from the loaded organic solution using water at 60°C. After scrubbing the loaded organic solution at an О/A ratio of 10 and 22°C, 98-100% of entrained metals were removed in a single contact with only 4.5% acid lost in the loaded scrub liquor. It was also found that the phase disengagement time was in the range of 2-4 min for both extraction and stripping which indicates a reasonable fast phase separation. In summary, the results from the study showed that membrane distillation and solvent extraction can be applied to recover acid and fresh water for reuse and metal values from mining and acidic process solutions.

Kang et al. (2019) investigated the techno-economic feasibility of using membrane distillation to recover clean water from AMD employing both renewable and non-renewable energy sources. The bench-scale set-up was used in batch tests in the study to establish whether membrane distillation is a viable technology to treat AMD. The bench-scale set-up was later modified to an open-loop continuous mode operation. Five membranes (two polypropylene, two polytetrafluoroethylene, and one polyvinylidene fluoride) that showed very high water flux of >35 kg/hm2 and high liquid entry pressures of 25-40 psig were used in the bench-scale set-up. The bench-scale set-up employed simulated AMD solution made up from soluble metal sulphates (Cd, Cr, Fe, and Zn), sodium arsenate, sodium selenate, and adjusted to pFl 2 with dilute sulphuric acid. The best of the five membrane distillation identified in the bench-scale set-up was investigated in an open-loop continuous process. In an open-loop continuous process, fresh simulated AMD was introduced into a tank and concentrated AMD was withdrawn at a predetermined rate so as to maintain a constant volume of the AMD feed in the tank, while purified water (i.e., the distillate) was also taken out at a predetermined rate in order to maintain it at a constant volume in the distillate tank.

Amongst all commercial membranes tested by Kang et al. (2019), 0.45-pm pore polypropylene membrane exhibited the highest water flux ('62 kg/hm2) and achieved 90% water recovery under optimal open-loop continuous membrane distillation conditions employing a realistic simulated AMD feed (total dissolved solids = 900 mg/L, pH = 2.4). However, the results indicated that the polypropylene membrane would need to be replaced once every 150 days under real process conditions. The cost analysis of the membrane distillation plant was performed for two major components of cost (i.e., capital cost and annual operating cost), and for four different energy sources, including local utility, photovoltaic, solar thermal, wind, and natural gas. The capital cost was defined as the cost associated with plant construction, process equipment purchases, and installation charges. The operating costs included amortisation, fixed charges, operating and maintenance costs, and membrane replacement costs. The economic analysis indicated pipelined natural gas and local electricity to be the most economical energy sources for heating AMD water and resulted in the total treatment costs of $0.476/m3 and $0.607/m3 of AMD, respectively.

In a study by Ryu et al. (2019), the performance of the following systems was evaluated: (1) natural and modified (heat treated) zeolite for heavy metal removal from AMD, (2) submerged DCMD for producing water for reuse from AMD, and (3) integrated submerged DCMD/sorption system for simultaneously removing heavy metals and producing water for reuse from AMD. The synthetic AMD solution used in the study was prepared by dissolving analytical grade CaS04/ MgS04-(3H,0), NaOH, FeO(OH), Fe(S04)-7H20, ZnS04-7H20, CuS04-5H20, Al2(S04)3-18H20, and Ni(N03)2-6H20 in Milli-Q water. The pH of the solution was adjusted using concentrated H2S04 (10 M). The natural zeolite in powder form (particle size < 75 mm) used in the experiments had a bulk density of 2.7g/cm3 and was mainly composed of clinoptilolite (—85 wt%) with minor quantities of quartz and mordenite (~15 wt%). Heat treatment method carried out at four different temperatures of 300, 400, 500 and 600°C for 24 h was used to potentially enhance the performance of natural zeolite. Heat treatment was chosen as it requires no additional chemicals and complex modification processes. The set-up of the DCMD consisted of a double-walled feed tank containing AMD solution with a submerged hollow fibre membrane made of polyvinylidene fluoride. The membrane pore size, inner and outer diameters, wall thickness, and contact angle were 0.1 pm, 0.7 mm, 1.2 mm, 250 pm, and 106 + 2°, respectively. The membrane module was made of 18 fibres of 0.2-m length (active membrane area of 0.0136 m2). The outer wall of the double-walled feed tank was circulated with heated water connected to a heating system, thus enabling an AMD feed solution to be maintained at a temperature of 55.0 ± 0.5°C. The permeate solution was maintained at 22.0 ± 0.5°C using a cooling system.

The results of a study by Ryu et al. (2019) showed that modified (heat treated) zeolite achieved 26-30% higher removal rate of heavy metals compared to natural untreated zeolite. Heavy metal sorption by heat treated zeolite followed the order of Fe > Al > Zn > Cu > Ni and the data fitted well to Langmuir and pseudo second-order kinetics model. A slight pH adjustment from 2 to 4 significantly increased Fe and Al removal rate (close to 100%) due to a combination of sorption and partial precipitation. An integrated system of submerged DCMD with zeolite for AMD treatment enabled to achieve 50% water recovery in 30 h. The integrated system provided a favourable condition for zeolite to be used in powder form with full contact time. Likewise, heavy metal removal from AMD by zeolite, specifically Fe and Al, mitigated membrane fouling on the surface of the hollow fibre submerged membrane. The integrated system produced fresh water of high quality while concentrating sulphuric acid and valuable heavy metals (Cu, Zn, and Ni).

 
Source
< Prev   CONTENTS   Source   Next >