In this research, four types of low cost and high performance ceramic microfiltration (MF) membranes have been employed in an in-line adsorption–MF process for oily wastewater treatment. Mullite, mullite-alumina, mullite-alumina-zeolite and mullite-zeolite membranes were fabricated as ceramic MF membranes by low cost kaolin clay, natural zeolite and α-alumina powder. Powdered activated carbon (PAC) and natural zeolite powder in concentrations of 100–800 mg L−1 were used as adsorbent agent in the in-line adsorption–MF process. Performance of the hybrid adsorption–MF process for each concentration of PAC and natural zeolite powder was investigated by comparing quantity of permeation flux (PF) and total organic carbon (TOC) rejection during oily wastewater treatment. Results showed that by application of 400 mg L−1 PAC in the adsorption–MF process with mullite and mullite-alumina membranes, TOC rejection was enhanced up to 99.5% in comparison to the MF only process. An increasing trend was observed in PF by application of 100–800 mg L−1 PAC. Also, results demonstrated that the adsorption–MF process with natural zeolite powder has higher performance in comparison to the MF process for all membranes except mullite-alumina membranes in terms of PF. In fact, significant enhancement of PF and TOC rejection up to 99.9% were achieved by employing natural zeolite powder in the in-line adsorption–MF hybrid process.

INTRODUCTION

Oil in water emulsions (oily wastewaters or oily waters) is one of the most important environmental problems in oil and gas industries such as refineries, and petrochemical and transportation industries. These wastewaters with high emission to the environment contain oil and grease that are hazardous for humans, animals and plants (Emani et al. 2014; Venault et al. 2016).

These oily wastewaters also affect drinking water and groundwater resources, endangering aquatic resources and human health, cause atmospheric pollution, affect crop production, destroy the natural landscape and even, probably because of coalescence of the oil burner, cause safety issues (Yu et al. 2013). By treatment of these wastewaters, the treated water can be used in washing applications or flash tank water, etc., to alleviate the water crisis.

Obviously, these oily wastewaters should be treated before discharging them to the environment. The permitted oil and grease limits for discharging oily wastewater to surface and coastal waters are 10 and 20 mg L−1 respectively, set by the Central Pollution Control Board (CPCB) of India (www.cpcb.nic.in) (Zolfaghari et al. 2016). There are several methods for treatment of oily wastewater such as flotation, sedimentation, adsorption, coalescence, electric methods and membrane separation processes. Conventional methods are not useful for treatment of the wastewater due to high energy consumption, high cost and low efficiency (Arzani et al. 2016; Das et al. 2016). In the recent years, there has been high interest in employing ceramic microfiltration (MF) membranes for separation of oil in water emulsions due to high thermal, mechanical and chemical strength of theses membranes plus long life (Abbasi & Taheri 2013; Zhong et al. 2013; Medjemem et al. 2016). In this area, the combination of this MF process with other processes such as coagulation, adsorption and ion exchange (hybrid processes) has many benefits and improvements such as enhancing the quality of the produced water, reduction of fouling on membranes, energy savings, environmental friendliness, and reductions in the capital and operating costs of the plants (Abdulgader et al. 2013; Ang et al. 2015; Chen et al. 2015; Yang et al. 2015). In-line adsorption–MF is a process whereby adsorption agents are added to the feed immediately before the membrane module and omitting the stage of sedimentation. By using this process, filtration time reduces and oil rejection efficiencies are increased (Lee et al. 2009). In the literature, some studies were focused on integrating MF membranes with pre-treatments such as precipitation (Amaral et al. 2015), oxidation (Abdel-Shafy et al. 2016), adsorption (Basu et al. 2016), coagulation/flocculation (Guo & Hu 2011), photocatalysis (Manjumol et al. 2016), ozonation (Guo et al. 2016), electrocoagulation (Akarsu et al. 2016) and ion exchange (Shanmuganathan et al. 2014). There are a few papers in the literature that studied the in-line adsorption–MF hybrid process with powdered activated carbon (PAC) but there is no research about employing very cheap natural zeolite powders in this hybrid process. Natural zeolite is very cheap and has high performance in comparison to other adsorption agents such as PAC. Juang et al. (2004) studied removal of sodium dodecyl benzene sulfonate and phenol from water by a ceramic MF-PAC hybrid process. They investigated effect of trans-membrane pressure (TMP) and cross-flow velocity (CFV) on membrane performance. Results illustrated that increasing PAC concentration causes reduction of permeation flux (PF) and increase in CFV and TMP, leading to enhanced PF. Yang et al. (2011) investigated integration of ceramic MF membrane with PAC for advanced treatment of oil-in-water emulsion. They studied the role of PAC addition in the PF enhancement mechanism. Results showed that addition of PAC enhances PF in comparison with the ceramic MF process without PAC addition. They concluded that PAC reduces the gel layer that formed on the membrane surface, due to the mechanical scouring effect of particles, according to the estimation of hydrodynamic forces. Abbasi et al. (2011) studied oily wastewater treatment by MF-PAC hybrid process using mullite and mullite-alumina ceramic membranes plus PAC. They reported that at low concentration of PAC, PF increased while at higher concentration of PAC, PF reduced due to high membrane fouling. Also, results of their work show that employing PAC has a positive effect on improving total organic carbon (TOC) rejection of MF membranes for oily wastewater treatment. In another study, Wang et al. (2013) used PAC and accumulative countercurrent two-stage adsorption–MF hybrid process, for the removal of organics from the reverse osmosis concentrate produced in a refinery wastewater treatment plant. The results showed that approximately 70% of dissolved organic carbon was removed from the reverse osmosis concentrate. Also their analysis demonstrated that fine PAC particles adhering to the membrane surface and blocking the membrane pores dominated the irreversible membrane fouling process.

In this study, based on our previous experience, cheap natural zeolite powder has been employed for enhancing oily wastewater treatment performance in the adsorption–MF hybrid process. In fact, natural zeolite powder has been used for two aims: first, used in the structure of mullite and mullite-alumina membranes for enhancing PF as new fabricated membranes and, second, employed as adsorbent. The objective of this investigation is to reduce production cost of ceramic membranes by using cheap natural zeolite and comparing natural zeolite and PAC performance in the in-line MF-adsorption hybrid process for enhancement of PF and TOC rejection during oily wastewater treatment. The importance of this work can be considered (i) from an environmental view by reducing hazardous content of oily wastewaters, (ii) from an economical view by decreasing production cost of ceramic membranes and (iii) in terms of saving water resources by reuse of treated wastewater in some applications. The novelty of this paper is fabrication, application and characterization of mullite-alumina-zeolite and mullite-zeolite ceramic MF membranes and use of natural zeolite as adsorption agent in the adsorption–MF hybrid process for oily wastewater treatment. Therefore, four types of low cost and high performance ceramic membranes, mullite, mullite-alumina, mullite-alumina-zeolite and mullite-zeolite, were fabricated by extrusion method. These membranes were used in the combined process of MF and in-line adsorption by PAC and natural zeolite powder for treatment of oil-in-water emulsions. Also, effect of PAC and natural zeolite powder concentration on membrane PF and TOC rejection was investigated.

THEORY

In the membrane separation process, parameters such as PF, TOC rejection (R) and porosity analysis are very important. PF is volume (V) of accumulated permeation flow in unit of time (t) and surface (A) of membrane, which is given by Equation (1): 
formula
1
R is the difference between oil concentration in feed (Cf) and permeation flow (Cp) divided by oil concentrations in feed (Cf) as follows: 
formula
2
Then the membrane porosity is calculated as follows (Mohammadi et al. 2005; Bakhtiari et al. 2011): 
formula
3
W2, W1 and Vm are dry and wet masses and volume of the membrane, respectively, and ρw is the water density at the experiment temperature.

MATERIAL AND METHODS

Membrane fabrication

Four types of tubular and symmetric ceramic MF membranes were fabricated by extrusion method using local powders. Mullite membranes were fabricated by mixing 69 wt.% kaolin clay and 31 wt.% distilled water. Kaolin was supplied by Zenooz mine, Tabriz, Iran, with purchased price 0.16 USD per kilogram. Chemical analysis of kaolin is listed in Table 1. The major component and phase in kaolin clay are SiO2 and kaolinite respectively, was also reported by Nandi et al. (2008) and Bouzerara et al. (2009) for preparation and application of ceramic membranes with kaolin clay. The mixture of kaolin clay and distilled water was extruded by an electrical extruder manufactured in Tehran, Iran, with 32 belts, power of 2 hp and dimension (length × width × height) 2 m × 1 m × 1.75 m. A tubular membrane with inner diameter of 10 mm, outer diameter of 14 mm and 10 cm length was made. Home-made membranes were then dried for 24 hour in room temperature followed by sintering in a programed furnace. Table 2 shows sintering procedure for preparation of membranes. In sintering temperature 550 °C, metakaolinis formed from the kaolinite phase under an endothermic reaction (Equation (4)). 
formula
4
Table 1

Chemical analysis of kaolin clay used in preparation of all ceramic membranes

Component Percent Phases Percent 
SiO2 61.62 Kaolinite 64 
TiO2 0.4   
Al2O3 24–25 Illite 2.4 
Fe2O3 0.45–0.65   
K20.4 Quartz 27 
Na20.5   
Loss on ignition 9.5–10 Feldspar 6.6 
Total 100  100 
Component Percent Phases Percent 
SiO2 61.62 Kaolinite 64 
TiO2 0.4   
Al2O3 24–25 Illite 2.4 
Fe2O3 0.45–0.65   
K20.4 Quartz 27 
Na20.5   
Loss on ignition 9.5–10 Feldspar 6.6 
Total 100  100 
Table 2

Sintering procedure of all fabricated ceramic membranes

Temperature (°C) Temp. gradient (°C min−1
25–550 
550 0 (for 60 min) 
550–975 
975 0 (for 60 min) 
975–1,100 
1,100 0 (for 60 min) 
Temperature (°C) Temp. gradient (°C min−1
25–550 
550 0 (for 60 min) 
550–975 
975 0 (for 60 min) 
975–1,100 
1,100 0 (for 60 min) 
In sintering temperature 1,050 °C, kaolin transforms to mullite phase and free silica under Equation (5). 
formula
5

Bouzerara et al. (2009) sintered ceramic membranes which were fabricated by kaolin clay under the temperature of 1,250 °C for an hour. Nandi et al. (2008) sintered ceramic membranes which were made of kaolin clay at four different temperatures, 850, 900, 950 and 1,000 °C to verify the effect of sintering temperature on the membrane properties. Free silica can be washed by strong alkali solution (20 wt.% NaOH) for 5 hours at 80 °C in an oven to increase porosity of fabricated membranes. For fabrication of mullite-alumina membranes, α-alumina powder with 99.6% purity, 75 μm average particle size and purchased for 2.08 USD per kilogram was provided by Semnan mines, Iran. After that, equal amounts of α-alumina powder and kaolin clay (50 wt.% α-alumina powder and 50 wt.% kaolin clay) were added to distilled water, then extruded and sintered for preparation of mullite-alumina membranes. Natural zeolite supplied by Semnan mine, Iran, with purchased price 0.38 USD per kilogram, was used for fabricating mullite-alumina-zeolite and mullite-zeolite membranes and use as adsorbent in the hybrid MF-zeolite process. For preparation of mullite-alumina-zeolite membranes, 30 wt.% α-alumina powder and 20 wt.% natural zeolite powder were added into a mixture of 50 wt.% kaolin and distilled water. This was followed by the extrusion and sintering processes used for mullite and mullite-alumina membranes. Chemical analysis of natural zeolite is presented in Table 3. From this table it can be observed that SiO2 and Al2O3 are the most common materials in the structure of natural zeolite. Analysis of natural zeolite reported by Shavandi et al. (2012) also showed high amount of SiO2 and Al2O3 in its structure. For preparation of mullite-zeolite membranes, 40 wt.% natural zeolite powder was added into a mixture of 60 wt.% kaolin clay and distilled water, followed by the extruding and sintering method. Composition and porosity of all fabricated membranes are shown in Table 4. As this table shows, porosity of mullite-alumina membranes is enhanced in comparison to mullite by using α-alumina powder in its structure. Mullite-zeolite membranes with porosity of 46.6% are more porous than others. PAC and NaOH with high purity were supplied by Merck Company, Germany. Properties of PAC and natural zeolite are listed in Table 5. According to this table, BET (Brunauer–Emmett–Teller) surface area, mean pore volume and mean pore diameter of PAC and natural zeolite were 554 m2 g−1 and 79 m2 g−1, 0.138 cm3g−1 and 0.119 cm3g−1, and 4.6 μm and 26.8 μm respectively. Natural zeolites generally have lower surface area than PAC but they can compete with PAC for separation of oil from water due to their low cost, convenient accessibility and hydrophilic properties (Bandura et al. 2015). Final production cost of mullite, mullite-alumina, mullite-alumina-zeolite and mullite-zeolite ceramic membranes taking into account the sintering procedure was about 63 to 83 USD m−2, so these four types of ceramic membranes are cheaper in comparison to commercial ceramic membranes. Economic estimations are presented in the Appendix (available with the online version of this paper).

Table 3

Chemical analysis of natural zeolite for preparation of mullite-alumina-zeolite and mullite-zeolite membranes

Component Percent 
SiO2 68.5 
Al2O3 11 
Na23.8 
K24.4 
CaO 0.6 
Fe2O3 0.2–0.9 
Loss on ignition 10–12 
Component Percent 
SiO2 68.5 
Al2O3 11 
Na23.8 
K24.4 
CaO 0.6 
Fe2O3 0.2–0.9 
Loss on ignition 10–12 
Table 4

Composition and porosity of all fabricated membranes

Membrane Kaolin% α-alumina% Zeolite% Porosity(%) 
Mullite 100 32.6 
Mullite-alumina 50 50 36.43 
Mullite-alumina-zeolite 50 30 20 34.02 
Mullite-zeolite 60 40 46.62 
Membrane Kaolin% α-alumina% Zeolite% Porosity(%) 
Mullite 100 32.6 
Mullite-alumina 50 50 36.43 
Mullite-alumina-zeolite 50 30 20 34.02 
Mullite-zeolite 60 40 46.62 
Table 5

Properties of PAC and natural zeolite used as adsorbent agents

  BET surface (m2 g−1Mean pore volume (cm3 g−1Mean pore diameter (μm) 
PAC 554.70 0.138 4.6 
Natural zeolite 79.26 0.119 26.8 
  BET surface (m2 g−1Mean pore volume (cm3 g−1Mean pore diameter (μm) 
PAC 554.70 0.138 4.6 
Natural zeolite 79.26 0.119 26.8 

Membrane characterization

Scanning electron microscopy (SEM) micrographs (Tescan Vega 3, Czech Republic), with acceleration voltage 20 kV, of fabricated membranes are shown in Figure 1. As the figure shows, surface morphology for mullite and mullite-zeolite membranes is similar. This is due to the fact that kaolin clay and natural zeolite powder are the main constituents of mullite and mullite-zeolite membranes and the composition of these two mineral powders is similar. Unlike the other two membranes, surface morphology of the mullite-alumina-zeolite and mullite-alumina membranes are less similar to each other because of difference between α-alumina and natural zeolite powder composition. All membranes show a porous and rough surface morphology. Mullite-zeolite membranes are more porous than the others. SEM observation showed that membranes' surfaces did not have any defects in their structures. X-ray diffraction (XRD) analysis was carried out using a Philips PW1800 diffractometer with Cu Kα radiation, manufactured in the Netherlands, for identification of formed phases in the structure of fabricated membranes (mullite, mullite-alumina, mullite-alumina-zeolite and mullite-zeolite) after sintering procedures. XRD patterns for membranes are shown in Figure 2. In sintering temperature 1,050 °C, kaolin under temperature is transformed to mullite, quartz and cristobalite phases. Quartz and cristobalite phases can be washed by strong alkali solution for increase in porosity of membranes. As the XRD pattern shows, in mullite and mullite-zeolite membranes, major phases are quartz and mullite and the minor phase is cristobalite, while in mullite-alumina and mullite-alumina-zeolite membranes, major phases are corundum and quartz and minor phases are mullite and cristobalite. Corundum is the most stable Al2O3 phase at all temperatures under ambient pressure. Other metastable Al2O3 polymorphs are often formed before reaching the most stable α-Al2O3 phase, in circumstances such as decomposition of aluminum hydroxide and aluminum oxy-hydroxide, high temperature oxidation of alumina-forming alloys, and recrystallization of amorphous alumina (Adachi et al. 2015). The membrane average pore diameter was measured using Image J software (version 1.44p) by SEM analysis (Nandi et al. 2009; Suresh & Pugazhenthi 2016). The membrane average pore diameter from SEM analysis was calculated using the following equation: 
formula
6
Figure 1

SEM micrograph of prepared ceramic membranes: (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite.

Figure 1

SEM micrograph of prepared ceramic membranes: (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite.

Figure 2

XRD pattern of fabricated ceramic membranes: (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite.

Figure 2

XRD pattern of fabricated ceramic membranes: (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite.

In this equation, n is the number of pores, daverage is the membrane average pore diameter in μm and di is the diameter which corresponds to each pore of the membrane. Mean pore diameter of mullite, mullite-alumina, mullite-alumina-zeolite and mullite-zeolite is 0.538, 0.74, 1.01 and 0.809 μm respectively.

Oily wastewater preparation

Synthetic oily wastewater was prepared by mixing crude oil provided by the Gachsaran oil field, Iran, distilled water and Triton X-100 (0.01 wt.%) as emulsifier agent. Triton X-100 was supplied by Merck Company, Germany. Table 6 presents physical and chemical analysis of the crude oil. In all experiments, 1,000 mg L−1 oily wastewater was prepared by utilizing a homogenizer for mixing the emulsion with 19,000 rpm speed for 30 min. The size of oil emulsion droplets in the feed was measured by a dynamic light scattering analyzer device (Microtrac, MN402-NS-0000-0000-000-4M, USA). The emulsion sample (1 mL) was analyzed immediately after preparation. Droplet size distribution of synthetic oily wastewater is shown in Figure 3. As shown, average droplet size of emulsion is small enough (96.5 nm) to remain stable for at least 12 hours. It is expected that the emulsion would become more stable as oil droplets in the emulsion become smaller.
Table 6

Physical and chemical analysis of crude oil for preparation of synthetic oily wastewater

Specification Unit Value Specification Unit Value 
Specific gravity @ 15.6 °C 0.8750 Reid vapor pressure psi 4.83 
API gravity  30.21 Asphaltene content Wt.% 3.60 
Sulfur content Wt.% 1.62 Wax Wt.% 6.06 
H2Wt. mg L−1 <1.0 Degree of melting point of wax °C 56 
Total nitrogen Wt.% 0.21 Conradson carbon residue Wt.% 5.19 
Basic sediment and water Vol.% 0.05 Ash content Wt.% 0.025 
Water content Vol.% 0.05 Acidity mgKOH g−1 0.05 
Salt content Pound per thousand barrel Nickel mg L−1 29 
Kinematic viscosity @10 °C cST 28.27 Vanadium mg L−1 105 
Kinematic viscosity @20 °C cST 16.82 Iron mg L−1 2.6 
Kinematic viscosity @40 °C cST 8.410 Lead mg L−1 <1 
Pour point °C −18 Sodium mg L−1 10 
Specification Unit Value Specification Unit Value 
Specific gravity @ 15.6 °C 0.8750 Reid vapor pressure psi 4.83 
API gravity  30.21 Asphaltene content Wt.% 3.60 
Sulfur content Wt.% 1.62 Wax Wt.% 6.06 
H2Wt. mg L−1 <1.0 Degree of melting point of wax °C 56 
Total nitrogen Wt.% 0.21 Conradson carbon residue Wt.% 5.19 
Basic sediment and water Vol.% 0.05 Ash content Wt.% 0.025 
Water content Vol.% 0.05 Acidity mgKOH g−1 0.05 
Salt content Pound per thousand barrel Nickel mg L−1 29 
Kinematic viscosity @10 °C cST 28.27 Vanadium mg L−1 105 
Kinematic viscosity @20 °C cST 16.82 Iron mg L−1 2.6 
Kinematic viscosity @40 °C cST 8.410 Lead mg L−1 <1 
Pour point °C −18 Sodium mg L−1 10 
Figure 3

Droplet size distribution of synthetic oily wastewater.

Figure 3

Droplet size distribution of synthetic oily wastewater.

Experimental setup and operation

Experimental setup is shown in Figure 4. A simple setup was designed and all operating conditions (temperature, pressure and CFV) were controlled easily. The effects of different operating parameters such as pressure (0.5–4 bar), CFV (0–2 m s−1), temperature (15–55 °C), on PF, fouling resistance, fouling and rejection of mullite membranes for treatment of synthetic wastewaters were investigated in a previous paper (Abbasi et al. 2010). The optimum operating conditions (TMP = 3 bar, CFV = 1.5 m s−1 and T = 35 °C) from this previous study were used in the present study. In order to control temperature of fluids in tanks, a heater and cooling water were placed into each tank. PAC was poured in-line into the oily wastewater and wastewater was filtered in the adsorption–MF hybrid process for 90 minutes. For all fabricated ceramic MF membranes, PF and R during oily wastewater treatment were obtained. It must be noted that most of the experiments were conducted carefully twice and average values of PF and oil rejection percent are reported in this paper. The results of these three experiments for all cases in the adsorption–MF hybrid process with PAC and natural zeolite showed very good repeatability with error lower than 1% between three runs. All permeate samples were taken in the final 5 minutes of filtration time. The emulsion has the maximum adsorption at 254 nm. Therefore, 254 nm UV (UV254) was used for the UV spectrometer to measure the oil concentration in permeation flow. All operating conditions during experimental procedures were fixed and did not change. About an hour after filtration ends, PAC or natural zeolite flocs aggregated on the bottom of the wastewater tank and settled out. These precipitated flocs after conventional filtration were regenerated. Adsorption agents (PAC or natural zeolite) were dried in 105 °C for 3 hours by using an electrical furnace. Then high temperature desorption and decomposition in 550 °C under an inert atmosphere for 3 hours was carried out. Finally residual organic matters by using steam at 800 °C were gasified and these regenerated adsorbents will be used in future investigations.
Figure 4

Schematic diagram of experimental setup.

Figure 4

Schematic diagram of experimental setup.

RESULTS AND DISCUSSION

Performance of membranes in MF-PAC hybrid process

Figures 5 and 6 present performance of all ceramic membranes in the MF-PAC hybrid process in terms of PF plus TOC rejection for oily wastewater treatment. As Figure 5(a) illustrates, by addition of PAC with concentration of 100–800 mg L−1 in the feed tank, PFs of mullite membrane are enhanced continually while they are lower than PF of the MF only process. These results are not consistent with Juang et al. (2004), who reported a decreasing PF by increasing PAC dosage for sodium dodecyl benzene sulfonate and phenol removal from water. One reason for this phenomenon is that very small PAC powders plus oil droplets fill membrane pores and decrease PF. Of course, by increasing PAC concentration to 800 mg L−1 membrane PF increases because PAC powders crush the membrane surface and decrease membrane fouling. Abbasi et al. (2011) reported an increasing trend in PF by increasing PAC concentration up to optimum value and then a descending trend of it by further enhancing PAC dosage. Yang et al. (2011) also reported enhancing PF by using PAC with an oil in water emulsion. In the case of TOC rejection, addition of different concentration of PAC has a positive effect due to adsorption of oil droplets. Similar results are also reported in the literature, which showed enhancing TOC rejection with increasing PAC concentration (Juang et al. 2004; Abbasi et al. 2011). In fact, as shown in Figure 6(a), for treatment of oily wastewater, TOC rejection in MF process with mullite membrane is 98.4% while in MF-adsorption with PAC at concentration of 400 mg L−1, it is 99.4%.
Figure 5

Variation of PF with time during oily wastewater treatment by MF-PAC hybrid process using (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 5

Variation of PF with time during oily wastewater treatment by MF-PAC hybrid process using (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 6

TOC rejection of ceramic membranes in MF-PAC hybrid process for oily wastewater treatment with (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 6

TOC rejection of ceramic membranes in MF-PAC hybrid process for oily wastewater treatment with (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Performance of mullite-alumina membranes in MF-PAC hybrid process is illustrated in Figures 5(b) and 6(b). Performance of mullite-alumina membranes during filtration of oily wastewater is similar to that of mullite membranes. As the results show, addition of different PAC concentration has a negative effect on PF while TOC rejection increased from 97.1% in MF to 99.5% in MF-PAC with 400 mg L−1 PAC concentration. In addition, membrane PF decline versus filtration time for mullite-alumina-zeolite membranes in MF-PAC process is presented in Figure 5(c). In this case, by increasing the concentration of PAC from 100 mg L−1 to 400 mg L−1, PF of mullite-alumina-zeolite membrane is enhanced continually but is lower than PF of the MF only process. By further increasing PAC concentration from 400 mg L−1 to 800 mg L−1, it is observed that PF of mullite-alumina-zeolite membrane is higher than that of the MF only process. Due to the high hydrophilic nature of zeolite in the membrane structure, fouling of oil on the membrane surface is not serious and PAC powders at high concentration (800 mg L−1 of PAC) crush the surface, detach and remove fouling on the membrane surface. Therefore, the membrane fouling is reduced and PF is enhanced. Similar to mullite and mullite-alumina membranes, TOC rejection increases from 96.2% in MF to 99.2% in MF-PAC with 400 mg L−1 PAC as shown in Figure 6(c). Finally, for mullite-zeolite membrane in adsorption–MF, with addition of all PAC concentrations (100–800 mg L−1) PF is enhanced and is more than that of the MF only process (as shown in Figure 5(d)). This is because of the hydrophilic nature of zeolite in the membrane structure, weak fouling on the membrane surface and crushing of the membrane surface with PAC. Also, as shown in Figure 6(d), TOC rejection was increased from 96.6% in MF process to 98.1% in MF-PAC hybrid process with 800 mg L−1 PAC concentration.

Performance of membranes in MF-zeolite hybrid process

Figures 7 and 8 show comparisons between performance of MF and adsorption–MF process with natural zeolite powder using all ceramic membranes in terms of PF and TOC rejection for oily wastewater treatment.
Figure 7

Variation of PF with time during oily wastewater treatment by MF-zeolite hybrid process using (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 7

Variation of PF with time during oily wastewater treatment by MF-zeolite hybrid process using (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 8

TOC rejection of ceramic membranes in MF-zeolite hybrid process for oily wastewater treatment with (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 8

TOC rejection of ceramic membranes in MF-zeolite hybrid process for oily wastewater treatment with (a) mullite, (b) mullite-alumina, (c) mullite-alumina-zeolite and (d) mullite-zeolite membranes.

Figure 7(a) demonstrates that by integration of natural zeolite powder with the MF process, mullite membrane PFs was enhanced in comparison to the MF only process. At the best conditions, final PF of mullite membrane increased from 176.1 L m−2h−1 in MF process to 305.4 L m−2h−1 in adsorption–MF hybrid process by addition of 200 mg L−1 natural zeolite powder. In fact, by addition of 200 mg L−1 natural zeolite in oily wastewater, PF of mullite ceramic membrane was enhanced by up to 73% in comparison to MF process without addition of any natural zeolite powder. It must be noted that by increasing zeolite concentration from 200 to 800 mg L−1 in adsorption–MF hybrid process, PF reduced slightly because natural zeolite particles deposit on the membrane surface and increase membrane fouling (Abbasi et al. 2011). Also, TOC rejection in this case increased from 98.4% in MF to 99.9% in MF-zeolite with 800 mg L−1 natural zeolite powder as shown in Figure 8(a). Performance of mullite-alumina membranes in MF-zeolite hybrid process is illustrated in Figures 7(b) and 8(b). As shown, unlike mullite membranes, for mullite-alumina membranes, addition of zeolite to the oily wastewater tank causes reduction of PF in comparison to the MF process only. On the other hand, at low concentrations of zeolite (100 and 200 mg L−1) PF reduction is not significant but at high concentrations of natural zeolite powder (400 and 800 mg L−1) it is meaningful due to deposit of natural zeolite particles and oil droplets on the membrane surface and increased fouling. SEM analysis proved these observations. Figure 9(a) and 9(b) show SEM micrographs of the surface and cross-section of the mullite-alumina membrane after filtration of oily wastewater with 800 mg L−1 natural zeolite in the hybrid adsorption–MF process. As the micrograph shows, high fouling on the membrane surface and pores affect the membrane performance. In addition, zeolite adsorbs the oil on the membrane surface; therefore, it can improve TOC rejection (Abbasi et al. 2011). As shown in Figure 8(b), by employing MF-zeolite hybrid process with 400 mg L−1 zeolite, TOC rejection increased from 98.3% in MF process to 99.7%. Performance of mullite-alumina-zeolite membranes in MF-zeolite hybrid process is shown in Figures 7(c) and 8(c). Results illustrate that addition of zeolite powder up to 200 mg L−1 in adsorption–MF hybrid process has a positive effect for enhancement of PF. By addition of 100 mg L−1 natural zeolite powder, PF of mullite-alumina-zeolite membrane increases from 354 L m−2 h−1 to 418.5 L m−2 h−1 (18.2% increase in PF) in MF process without zeolite addition. Also, as Figure 8(c) indicates, by increasing zeolite concentration from 0 mg L−1 to 800 mg L−1, TOC rejection increased from 96.2 to 99.2%.
Figure 9

(a) SEM micrograph of surface of mullite-alumina membrane after oily wastewater treatment in the hybrid adsorption–MF process using 800 mg L−1 natural zeolite. (b) SEM micrograph of cross-section of mullite-alumina membrane after oily wastewater treatment in the hybrid adsorption–MF process using 800 mg L−1 natural zeolite.

Figure 9

(a) SEM micrograph of surface of mullite-alumina membrane after oily wastewater treatment in the hybrid adsorption–MF process using 800 mg L−1 natural zeolite. (b) SEM micrograph of cross-section of mullite-alumina membrane after oily wastewater treatment in the hybrid adsorption–MF process using 800 mg L−1 natural zeolite.

Finally, addition of different natural zeolite concentrations has positive effect on PF enhancement in MF-zeolite hybrid process with mullite-zeolite ceramic membranes (see Figure 7(d)). It can be attributed to the synergistic effect of natural zeolites which are used in the structure of mullite-zeolite membranes and in the hybrid process as adsorption agent. Results show that PF of mullite-zeolite membranes increases from 60 L m−2 h−1 in MF process to 502 L m−2 h−1 in MF-zeolite hybrid process by addition of 100 mg L−1 natural zeolite powder. In fact, at low concentration of natural zeolite powder, by crushing of the membrane surface with zeolite powder, fouling on the membrane surface was reduced, which led to enhancement of PF of membranes. This result can be attributed to the synergistic effect of natural zeolite powder employed in the mullite-zeolite ceramic membrane structure and added into the oily wastewater tank in the hybrid process. Also, as seen in Figure 8(d), TOC rejection increased from 96.6% in MF process to 98.95% in MF-zeolite with 800 mg L−1 zeolite concentration.

CONCLUSIONS

In this paper, four types of cheap and novel ceramic membranes were used in in-line MF-PAC and MF-zeolite hybrid processes for synthetic oily wastewater treatment. Tubular and symmetric mullite, mullite-alumina, mullite-alumina-zeolite and mullite-zeolite membranes were fabricated using very cheap raw materials such as kaolin clay, natural zeolite and α-alumina powder and employed in experiments.

Results showed that by addition of PAC in the adsorption–MF process for treatment of oily wastewater, TOC rejection increased in all concentrations of PAC for all membranes in comparison to the MF process only. Also, addition of PAC at 800 mg L−1 in the adsorption–MF process has a positive effect on enhancement of PF of membranes except mullite and mullite-alumina ceramic membrane. In addition, by combination of in-line adsorption using natural zeolite powder with the MF process during oily wastewater treatment, PF of all membranes except mullite-alumina membranes was enhanced significantly. Also, in all experiments, TOC rejection with the MF-zeolite hybrid process was higher than for the MF process. Finally, it can be concluded that employing PAC and cheap natural zeolite powder as adsorbent agent in the in-line adsorption–MF hybrid process improves performance of oily wastewater treatment in terms of PF and TOC rejection.

REFERENCES

REFERENCES
Abbasi
M.
Mirfendereski
M.
Nikbakht
M.
Golshenas
M.
Mohammadi
T.
2010
Performance study of mullite and mullite–alumina ceramic MF membranes for oily wastewaters treatment
.
Desalination
259
(
1
),
169
178
.
Abbasi
M.
Reza Sebzari
M.
Mohammadi
T.
2011
Enhancement of oily wastewater treatment by ceramic microfiltration membranes using powder activated carbon
.
Chemical Engineering & Technology
34
(
8
),
1252
1258
.
Abdel-Shafy
H. I.
El-Khateeb
M. A.
Mansour
M. S. M.
2016
Treatment of leather industrial wastewater via combined advanced oxidation and membrane filtration
.
Water Science and Technology
74
(
3
),
586
594
.
Abdulgader
H. A.
Kochkodan
V.
Hilal
N.
2013
Hybrid ion exchange – Pressure driven membrane processes in water treatment: a review
.
Separation and Purification Technology
116
,
253
264
.
Adachi
S.
Ishimaru
M.
Sina
Y.
McHargue
C. J.
Sickafus
K. E.
Alves
E.
2015
Corundum-to-spinel structural phase transformation in alumina
.
Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms
358
,
136
141
.
Akarsu
C.
Ozay
Y.
Dizge
N.
Elif Gulsen
H.
Ates
H.
Gozmen
B.
Turabik
M.
2016
Electrocoagulation and nanofiltration integrated process application in purification of bilge water using response surface methodology
.
Water Science and Technology
74
(
3
),
564
579
.
Amaral
M. C. S.
Pereira
H. V.
Nani
E.
Lange
L. C.
2015
Treatment of landfill leachate by hybrid precipitation/microfiltration/nanofiltration process
.
Water Science and Technology
72
(
2
),
269
276
.
Bakhtiari
O.
Samei
M.
Taghikarimi
H.
Mohammadi
T.
2011
Preparation and characterization of mullite tubular membranes
.
Desalination and Water Treatment
36
(
1–3
),
210
218
.
Basu
S.
Kaushik
A.
Saranya
P.
Batra
V. S.
Balakrishnan
M.
2016
High strength distillery wastewater treatment by a PAC-MBR with low PAC dosage
.
Water Science and Technology
73
(
5
),
1104
1111
.
Bouzerara
F.
Harabi
A.
Condom
S.
2009
Porous ceramic membranes prepared from kaolin
.
Desalination and Water Treatment
12
(
1–3
),
415
419
.
Chen
B.
Xiong
X.
Yao
Z.
Yin
N.
Low
Z.-X.
Zhong
Z.
2015
Integrated membrane process for wastewater treatment from production of instant tea powders
.
Desalination
355
,
147
154
.
Das
B.
Chakrabarty
B.
Barkakati
P.
2016
Preparation and characterization of novel ceramic membranes for micro-filtration applications
.
Ceramics International
42
(
13
),
14326
14333
.
Guo
J.
Wang
L.
Shi
J.
Guan
Y.
Wang
K.
Liu
H.
Liu
J.
Wang
B.
2016
Impact of AC/O3 pretreatment on ultrafiltration processing sewage effluent for fouling alleviation
.
Desalination and Water Treatment
57
(
56
),
27113
27124
.
Medjemem
N.
Harabi
A.
Bouzerara
F.
Foughali
L.
Boudaira
B.
Guechi
A.
Brihi
N.
2016
Elaboration and characterization of low cost ceramics microfiltration membranes applied to the sterilization of plant tissue culture media
.
Journal of the Taiwan Institute of Chemical Engineers
59
,
79
85
.
Mohammadi
T.
Pak
A.
Nourian
Z.
Taherkhani
M.
2005
Experimental design in mullite microfilter preparation
.
Desalination
184
(
1
),
57
64
.
Nandi
B. K.
Uppaluri
R.
Purkait
M. K.
2008
Preparation and characterization of low cost ceramic membranes for micro-filtration applications
.
Applied Clay Science
42
(
1–2
),
102
110
.
Nandi
B.
Uppaluri
R.
Purkait
M.
2009
Treatment of oily waste water using low-cost ceramic membrane: flux decline mechanism and economic feasibility
.
Separation Science and Technology
44
(
12
),
2840
2869
.
Shanmuganathan
S.
Nguyen
T. V.
Shim
W.
Kandasamy
J.
Vigneswaran
S.
2014
Performance of submerged membrane–ion exchange hybrid system with Purolite A502PS in treating reverse osmosis feed
.
Separation and Purification Technology
122
,
24
31
.
Shavandi
M.
Haddadian
Z.
Ismail
M.
Abdullah
N.
Abidin
Z.
2012
Removal of residual oils from palm oil mill effluent by adsorption on natural zeolite
.
Water, Air, & Soil Pollution
223
(
7
),
4017
4027
.
Suresh
K.
Pugazhenthi
G.
2016
Development of ceramic membranes from low-cost clays for the separation of oil–water emulsion
.
Desalination and Water Treatment
57
(
5
),
1927
1939
.
Venault
A.
Jumao-as-Leyba
A. J.
Chou
F.-C.
Bouyer
D.
Lin
I. J.
Wei
T.-C.
Chang
Y.
2016
Design of near-superhydrophobic/superoleophilic PVDF and PP membranes for the gravity-driven breaking of water-in-oil emulsions
.
Journal of the Taiwan Institute of Chemical Engineers
65
,
459
471
.
Yu
L.
Han
M.
He
F.
2013
A review of treating oily wastewater
.
Arabian Journal of Chemistry
.
DOI: 10.1016/j.arabjc.2013.07.020
.
Zolfaghari
R.
Fakhru'l-Razi
A.
Abdullah
L. C.
Elnashaie
S. S. E. H.
Pendashteh
A.
2016
Demulsification techniques of water-in-oil and oil-in-water emulsions in petroleum industry
.
Separation and Purification Technology
170
,
377
407
.