Intermittent backwashing and relaxation are mandatory in the membrane bioreactor (MBR) for its effective operation. The objective of the current study was to evaluate the effects of run-relaxation and run-backwash cycle time on fouling rates. Furthermore, comparison of the effects of backwashing and relaxation on the fouling behavior of membrane in high rate submerged MBR. The study was carried out on a laboratory scale MBR at high flux (30 L/m2·h), treating sewage. The MBR was operated at three relaxation operational scenarios by keeping the run time to relaxation time ratio constant. Similarly, the MBR was operated at three backwashing operational scenarios by keeping the run time to backwashing time ratio constant. The results revealed that the provision of relaxation or backwashing at small intervals prolonged the MBR operation by reducing fouling rates. The cake and pores fouling rates in backwashing scenarios were far less as compared to the relaxation scenarios, which proved backwashing a better option as compared to relaxation. The operation time of backwashing scenario (lowest cycle time) was 64.6% and 21.1% more as compared to continuous scenario and relaxation scenario (lowest cycle time), respectively. Increase in cycle time increased removal efficiencies insignificantly, in both scenarios of relaxation and backwashing.

INTRODUCTION

Rapid industrialization and urbanization is a great threat to fresh water availability. Approximately 700 million people live in water scarce conditions. Water resources should be properly managed in order to overcome water scarcity. Reuse of wastewater could be a better option to reduce the stress on fresh water resources (Maton et al. 2010; Tabraiz et al. 2015b). Many techniques are studied and being practiced for wastewater treatment, among them the biological treatment methods are economic and suitable (Tabraiz et al. 2015a). With respect to reuse context, membrane bioreactor (MBR) is an emerging wastewater treatment technology, and has been grown exponentially. MBR produce the effluent of reusable quality. It can remove >99% chemical oxygen demand (COD), 89% nitrogen (N), 88% phosphorus (Monclús et al. 2010). MBR has been used in treatment of sewage and industrial wastewater, i.e., pharmaceutical, tanneries, dairies, etc. Furthermore, it has also been used to remove the emerging pollutant, i.e., trace organics from wastewater (Sipma et al. 2010; Tadkaew et al. 2011). MBR combines the function of an activated sludge system with filtration in one unit tank. It is the combination of a bioreactor and membrane which act like a filter. In the process, organic matter degraded, and membrane filters out treated water from bioreactor while retaining biomass in it. Hence, increased solid retention time (SRT) achieved. Consequently, the removal of nutrients is increased (Ersu et al. 2010). It requires less hydraulic retention time (HRT) and less area as compared to activated sludge process. The main problem, which arises in the operation of MBR, is the frequent fouling of membrane. Energy requirement is increased when the membrane gets biofouled, frequent washing is required, and ultimately the membrane requires replacement. MBR could be a very sustainable wastewater treatment option, if the biofouling could be reduced, thereby, reducing the membrane replacement frequency can be controlled.

Environmental parameters, biological factors within bioreactor, operational conditions and membrane properties are the major factors affecting the fouling. It is challenging to establish a relation between these numerous factors and fouling (Maqbool et al. 2014). Operational conditions include: HRT, SRT, mixed liquor suspended solids (MLSS), aeration rate, aerators position, dissolved oxygen (DO), and flux rate directly affect the performance of MBR and membrane fouling (Cho & Fane 2002; Ji & Zhou 2006; Meng et al. 2007). Aeration rate and time have greater effect on the effluent quality and membrane fouling when compared to the aerators position. Previous studies revealed that the concentration of microorganisms increased in the bioreactor as the aeration-on time was increased, while the extracellular polymeric substances (EPS) concentration increased when aeration-off time was increased (Lim et al. 2007). Optimum aeration rate was determined, beyond which no reduction of fouling was observed by increasing aeration rate further (Ueda et al. 1997). DO level also influence the metabolic functions of microbial communities such as generation, composition as well as release of EPS (Gao et al. 2011; Wu et al. 2011). Low DO level resulted in more EPS production, in mix liquor, and its accumulation on membrane surface triggers higher membrane fouling (Gao et al. 2011).

Pollutants removal effectiveness and membrane fouling behavior have been reported in several recent studies on MBR. It was found that higher MLSS content causes the lowering of permeate flux and if MLSS increases then sludge can deposit on membrane surface more easily (Sven et al. 2007). Increasing SRT resulted in enhanced MLSS concentrations. Low SRT, less than 2 days, increased the fouling rate 10 times when compared to high HRT of 10 days. The main reason of fouling increment at very low SRT is considered to be EPS production in excess (Trussell et al. 2006; Zhang et al. 2006). Long HRT correspond to low permeate flux for a specific volume, while short HRT corresponds to a higher flux. It is expected that high concentration of dissolved organic content in bioreactor and higher permeate flux through membrane surface causes acceleration of membrane fouling and eventually declining permeate flux (Jeong et al. 2007).

Similarly, some reports observed that operational conditions; inorganic pollutant load, organic loading rate, cross flow velocity, trans-membrane pressure (TMP), etc. affect the membrane fouling (Trussell et al. 2006a; Kennedy et al. 2008). In addition, the feed water characteristics, pH and ionic strength of the particles, hydrophobicity of natural organic matter, particles size and shape of particles influence the fouling rates (Carroll et al. 2000; Abdelrasoul et al. 2013). The membrane characteristics, i.e., pore size, pores distribution, and hydrophobicity and hydrophilicity of membrane material, also affect the fouling rate (Gui et al. 2003; Sun et al. 2013).

Accordingly, many researchers applied different techniques to maintain relative stable membrane permeability by mitigating membrane fouling through physical, biological and chemical methods. Physical techniques investigated for fouling control are; vibration, air sparging, ultra-sonication, sponge ball cleaning, moving media, magnetic enzymatic carriers, relaxation in operation mode, and backwashing in operational mode (Zhao et al. 2000; Chen et al. 2006; Lee et al. 2006; Nywening et al. 2007; Ngo et al. 2008; Wu et al. 2008; Yeon et al. 2009).

Relaxation refers to stopping the production mode of the membrane for a short span of time, to give a break in continuous operation. Consequently, fouling is naturally reduced when reversibly attached foulant diffuse away from the membrane surface due to concentration gradient. Maqbool et al. (2014) evaluated the effect of filtration modes on the fouling while treating synthetic wastewater at 15 L/m2-h flux. The results showed the low fouling rate at lower ratio of filtration time to relaxation time (Maqbool et al. 2014). On the other hand, backwash is the back flushing of permeates in the membrane which removes the foulant from membrane due to shear. Yigit et al. (2010) studied the effect of backwash on membrane fouling at different backwash scenarios at 20 L/m2-h flux. The study revealed that the lower ratio of filtration time to backwashing time reduced the fouling rates (Yigit et al. 2010). A study compared backwashing after a fixed interval, with backwashing initiated on 1.5 KPa TMP increment. Results revealed that fixed interval backwashing is more effective in removing fouling (Smith et al. 2006). Chemical enhanced backwashing was also applied to reduce fouling in MBR, i.e., sodium hydroxide, sodium hypochlorite. The addition of chemical in backwashing water reduced fouling more effectively (Lee et al. 2012). A study evaluated the effect of backwash and relaxation on fouling pattern. Synthetic wastewater was used, and the flux rate of 20 L/m2-h was maintained. The backwashing and relaxation intervals were not constant; hence, the comparison is not justified. Furthermore, study did not compare the effect of backwash and relaxation on pore and cake resistance. Additionally, the effect of cycle time on pore and cake resistance was not evaluated substantially (Wu et al. 2008).

This study aims to fulfill this gap by comparatively evaluating the effect of relaxation and backwashing cycle time on membrane fouling. The effect of backwashing and relaxation is evaluated on cake resistance, pore resistance and average fouling rates in short and long cycle times, keeping the filtration time/backwashing time and filtration time/relaxation time ratio constant, at flux of 30 L/m2-h. Additionally, a real wastewater (sewage) was used in the study as this type of study has never been experimented on sewage. The research would help to decide the best option (optimal option) to reduce membrane fouling viz. (relaxation or backwashing) and the associated cycle time in the operation of high flux MBRs for sewage treatment.

MATERIALS AND METHODS

Bench scale MBR

A submerged membrane system was assembled (Figure 1). A feed tank with capacity of 70 L, containing domestic sewage was used to feed the MBR. A solenoid valve was attached between the feed tank and bioreactor to regulate the flow. The operation of solenoid valve was controlled through level sensors, placed at different levels in the bioreactor. These sensors were linked to solenoid valve through a relay unit. Bioreactor tank volume was 8.4 L (working volume). A hollow fiber membrane (polyvinylidene difluoride) of 0.04 micron pore size and 0.07 m2 area (Hinada Water Treatment Tech., Co., Ltd, China) was submerged in the bioreactor. The permeate flux of 30 L/m2/hr was maintained with HRT of 4 h.
Figure 1

Schematics of bench scale MBR.

Figure 1

Schematics of bench scale MBR.

A peristaltic pump (BT100M, Boading Chuangrui Precision Pump Co., Ltd, China) was used to pump out effluent from the bioreactor through membrane. The pump flow rate range was 0.0015–380 mL/min, to set the required flux. Digital timer's circuit was connected with pump to control the timings of forward, backwash and relaxation time of pump. A digital manometer (82152, AZ, China) was installed on pipe in between membrane and pump to measure TMP. To prevent manometer from water, a water trap box was attached ahead of it.

Activated sludge was obtained from a wastewater treatment plant situated in Islamabad, and mixed in a bioreactor. Before starting the study, the bioreactor was run in the laboratory for 1 month. The designed sewage flow and aeration was continuously provided to bioreactor. Content of the wastewater were kept well mixed using the mixer. The biomass concentration was kept between 5.5 and 6 g/L by withdrawing the excess biomass while maintaining SRT of 20 days. The DO concentration in the bioreactors was kept between 3 and 4 mg/L.

Operational plan

The study conducted three trials to compare the effect of relaxation and backwashing on biofouling. The trial details are listed below in Table 1.

Table 1

Operational plan

Trial Scenario Description Run time Relaxation time Backwashing time 
Trail 1 Scenario-1 MBR Forward Continuous 
Trail 2 (Relaxation) Scenario-1 MBR5 + 0.5R 5 min 0.5 min 
Scenario-2 MBR10 + 1R 10 min 1 min 
Scenario-3 MBR15 + 1.5R 15 min 1.5 min 
Trail 3 (Backwashing) Scenario-1 MBR5 + 0.5B 5 min 0.5 min 
Scenario-2 MBR10 + 1B 10 min 1 min 
Scenario-3 MBR15 + 1.5B 15 min 1.5 min 
Trial Scenario Description Run time Relaxation time Backwashing time 
Trail 1 Scenario-1 MBR Forward Continuous 
Trail 2 (Relaxation) Scenario-1 MBR5 + 0.5R 5 min 0.5 min 
Scenario-2 MBR10 + 1R 10 min 1 min 
Scenario-3 MBR15 + 1.5R 15 min 1.5 min 
Trail 3 (Backwashing) Scenario-1 MBR5 + 0.5B 5 min 0.5 min 
Scenario-2 MBR10 + 1B 10 min 1 min 
Scenario-3 MBR15 + 1.5B 15 min 1.5 min 

Wastewater composition

The sewage was collected from the community sewage collection tank four times a day and mixed to make a composite sample to feed the MBR. The composite was characterized on a daily basis. The characteristics of the sewage are given in Table 2. The effluent from the MBR was collected in a tank at the end of each scenario, and the required volume of sample was taken from the effluent tank for the analysis of each parameter.

Table 2

Sewage characteristics

Sr. no. Parameters Mean valuen Standard deviation 
pH 6.8–7.9 – 
COD 375.8 ±50.5 
TSS 340.5 ±120.5 
TP 18.36 ±3.2 
TKN 45.5 ±5.7 
Sr. no. Parameters Mean valuen Standard deviation 
pH 6.8–7.9 – 
COD 375.8 ±50.5 
TSS 340.5 ±120.5 
TP 18.36 ±3.2 
TKN 45.5 ±5.7 

n is number of sample that were 30.

Analytical methods

COD and phosphate were measured by calorimetric method using spectrophotometer (DR 600, HACH, USA). Total organic carbon (TOC) was measured using the TOC analyzer (TOC-L, Shimadzu, Japan). Total Kjeldahl nitrogen (TKN) was analyzed using Kjeltec™ 2100 (FOSS analytical AB, Sweden). DO in the bioreactor was measure using DO meter (Model 4320, Control Company, USA). Nitrate and MLSS were measured using Standard Methods (APHA, AWWA, WEF 2005). Each scenario was repeated three times. The results reported are the mean of three values with standard deviation.

Cake, pore and intrinsic resistance

Total resistance, cake resistance, pore resistance and intrinsic resistance were measured by dead end filtration. Before the start of a new scenario or trial, washed membrane was placed in a tab water tank and a pump was operated at the same flux which was maintained in the study. The pressure developed across the membrane (ΔPi) was observed. After reaching the critical pressure, the membrane with cake was taken out from the bioreactor carefully and placed in the tab water tank. The pump was operated at the same flux and the pressure developed across the membrane was noted (ΔPt). After that, the cake was removed from the membrane by shaking the membrane well in the bioreactor tank. Membrane was then again placed in a tab water tank and a pump was operated at the same flux. Similarly, the pressure developed across the membrane was noted (ΔPp).

The cake, pore and intrinsic resistances were calculated by using the following models (Khan et al. 2012; Maqbool et al. 2014). 
formula
1
 
formula
2
 
formula
3
 
formula
4
 
formula
5
 
formula
6
where Rt, Ri, Rp, and Rc is total resistance (m−1), intrinsic resistance (m−1), pores resistance (m−1), and cake resistance (m−1), respectively. Whereas J is permeate flux (m3/m2 s), μ is the dynamic viscosity of permeate (N sec/m2), ƒt is temperature correction factor, and T is bioreactor temperature (°C). Rc is cake resistance, that is caused by the formation of cake on the membrane surface and causes membrane fouling. While Rp is pores resistance that results due to adsorption of EPS and small particles, which penetrate in membrane pores, and Ri is the membrane intrinsic resistance (Jarusutthirak & Amy 2006).

Membrane cleaning

After each scenario, fouled membrane was cleaned to restore the intrinsic TMP, to reuse it in the same scenario, as each scenario was repeated three times. It was drenched in aqueous solution of 2% sodium hydroxide and 4 g/L sodium hypochlorite (Sigma-Aldrich), for 30 min (Maqbool et al. 2014). At last, distilled water was filtered through the membrane for 30 min for its complete cleansing. The cleaning of the fouled membrane by above mentioned method is common and it restores intrinsic TMP by removing organic fouling (Judd et al. 2004). For new trial/scenario, new membrane was used.

RESULTS AND DISSCUSION

Relaxation run

Membrane fouling trends

Trends of membrane fouling at different scenarios of relaxation were shown in Figure 2. Profile reveals the fouling behavior of different scenarios: MBR(Forward), MBR(5+0.5R), MBR(10+1R), and MBR(15+1.5R). The membrane fouled in 19.5, 27, 25, and 21.5 h in MBR(Forward), MBR(5+0.5R), MBR(10+1R) and MBR(15+1.5R) scenario, respectively. The MBR(5+0.5R), MBR(10+1R) and MBR(15+1.5R) operation time was 1.38, 1.28, and 1.1 times greater than MBR(Forward), respectively. The results showed that membrane in scenario MBR(15+1.5R) fouled early as compared to MBR(5+0.5R), MBR(10+1R). The MBR(5+0.5R) fouled after longer time as compared to the other scenarios. The steady phase in the MBR(5+0.5R) was elongated as compared to MBR(Forward), MBR(10+1R) and MBR(15+1.5R), which caused the longer run.
Figure 2

Membrane fouling patterns of different scenarios of relaxation.

Figure 2

Membrane fouling patterns of different scenarios of relaxation.

Pressure range, phase time, and fouling rates in different phases of TMP profile: maturation phase, steady phase and jump phase, are given in Table 3. The highest fouling rates were observed in the jump phase, on contrast, the lowest fouling rates were observed in the steady phase. Maximum steady phase time observed was 19.3 h for MBR(5+0.5R), while the steady phase times of MBR(10+1R), MBR(15+1.5R), and MBR(Forward) were 15.6, 15.7 and 11.6 h, respectively. Similarly, the fouling rates observed in the maturation phase of MBR(5+0.5R), MBR(10+1R) and MBR(15+1.5R) and MBR(Forward) were 1.3, 2.15, 2.1, and 2.98 KPa/h, respectively. The fouling rates, in the steady phase of MBR(5+0.5R), MBR(10+1R) and MBR(15+1.5R) and MBR(Forward), were observed 0.5, 0.56, 1.07, and 1.68 KPa/h, respectively. Jump phase has higher fouling rates as compared to the maturation and steady phases. The fouling rates in the jump phases of MBR(5+0.5R), MBR(10+1R) and MBR(15+1.5R) and MBR(Forward), observed were: 1.221.34, 1.45 and 1.64 KPa/h, respectively. It was concluded that increase in the cycle time increased the fouling rates in all the phases, even if the runtime and relaxation time ratio was kept constant. Lowest fouling rates and highest phase times were observed in MBR(5+0.5R). A previous study reported the effect of filtration modes on the fouling while treating synthetic wastewater at flux of 15 L/m2-h. The 8 min filtration followed by 2 min relaxation showed the average fouling rates of 2.03 KPa/day (Maqbool et al. 2014). The fouling rates in the reported study were less as compared to the present study. The low fouling rates may be caused due to low flux rates, different wastewater type, and different ratio of runtime to relaxation time. Similarly, in another study, the fouling rates in relaxation modes were 8–10 kPa/h. The higher fouling rates were due to high concentration of MLSS, 9–10 mg/L (Wu et al. 2008).

Table 3

Pressure range, phase time and fouling rates of maturation, steady and jump phases in relaxation filtration scenarios

Filtration scenario (relaxation) Maturation phase Steady phase Jump phase Average fouling rates (KPa/hr) 
MBR(Forward) 
 Pressure range (KPa) 3.34–9.0 9.0–19.5 19.5–35.3 1.64 ± 0.22 
 Phase time (hours) 1.9 ± 0.15 11.6 ± 1.1 5.7 ± 0.33 
 Fouling rate (KPa/h) 2.98 ± 0.13 1.68 ± 0.11 3.65 ± 0.14 
MBR(5+0.5R) 
 Pressure range (KPa) 2.4–4.75 4.75–14.27 14.27–35.3 1.22 ± 0.19 
 Phase time (h) 1.3 ± 0.19 19.3 ± 1.3 6.4 ± 1.1 
 Fouling rate (KPa/h) 1.3 ± 0.15 0.5 ± 0.085 3.28 ± .21 
MBR(10+1R) 
 Pressure range (KPa) 2.6–6.0 6.0–14.7 14.7–35.2 1.34 ± 0.13 
 Phase time (h) 1.58 ± 0.2 15.6 ± 1.1 7.22 ± 1.2 
 Fouling rate (KPa/h) 2.15 ± 0.17 0.56 ± 0.05 2.84 ± 0.17 
MBR(15+1.5R) 
 Pressure range (KPa) 3.8–8 8–16.9 16.9–35.1 1.45 ± 0.15 
 Phase time (h) 2.0 ± 0.33 15.7 ± 1.57 3.95 ± 0.67 
 Fouling rate (KPa/h) 2.1 ± 0.14 1.07 ± 0.07 4.6 ± 0.37 
Filtration scenario (relaxation) Maturation phase Steady phase Jump phase Average fouling rates (KPa/hr) 
MBR(Forward) 
 Pressure range (KPa) 3.34–9.0 9.0–19.5 19.5–35.3 1.64 ± 0.22 
 Phase time (hours) 1.9 ± 0.15 11.6 ± 1.1 5.7 ± 0.33 
 Fouling rate (KPa/h) 2.98 ± 0.13 1.68 ± 0.11 3.65 ± 0.14 
MBR(5+0.5R) 
 Pressure range (KPa) 2.4–4.75 4.75–14.27 14.27–35.3 1.22 ± 0.19 
 Phase time (h) 1.3 ± 0.19 19.3 ± 1.3 6.4 ± 1.1 
 Fouling rate (KPa/h) 1.3 ± 0.15 0.5 ± 0.085 3.28 ± .21 
MBR(10+1R) 
 Pressure range (KPa) 2.6–6.0 6.0–14.7 14.7–35.2 1.34 ± 0.13 
 Phase time (h) 1.58 ± 0.2 15.6 ± 1.1 7.22 ± 1.2 
 Fouling rate (KPa/h) 2.15 ± 0.17 0.56 ± 0.05 2.84 ± 0.17 
MBR(15+1.5R) 
 Pressure range (KPa) 3.8–8 8–16.9 16.9–35.1 1.45 ± 0.15 
 Phase time (h) 2.0 ± 0.33 15.7 ± 1.57 3.95 ± 0.67 
 Fouling rate (KPa/h) 2.1 ± 0.14 1.07 ± 0.07 4.6 ± 0.37 

Numbers of samples were 3.

Fouling resistances

Total resistance is the main factor that was affected by increasing the cycle time of run-relaxation trial, as the cycle time increased, the total resistance increased. The total resistance of different scenarios: MBR(5+0.5R), MBR(10+1R) and MBR(15+1.5R) and MBR(Forward) was 6.03 × 1012, 6.63 × 1012, 7.25 × 1012 and 7.81 × 1012 m−1, respectively. In the same fashion, the pore resistance and cake resistance increased with increase in the cycle time of run-relaxation trial (Table 4).

Table 4

Cake, pore and intrinsic resistances of membrane in relaxation filtration scenarios

Resistance ( × 1012MBR(5+0.5R) MBR(10+1R) MBR(15+1.5R) MBR(Forward) 
Cake layer resistant 3.01 ± 0.13 3.32 ± 0.11 3.61 ± 0.15 3.9 ± 0.2 
Pores resistance 2.28 ± 0.09 2.57 ± 0.075 3.01 ± 0.11 3.17 ± 0.17 
Intrinsic resistance 0.74 ± 0.05 0.74 ± 0.044 0.63 ± 0.08 0.74 ± 0.034 
Total resistance 6.03 ± 0.09 6.63 ± 0.076 7.25 ± 0.11 7.81 ± 0.13 
Rc/Rt (%) 50 50 50 50 
Rp/Rt (%) 37.8 38.8 41.5 40.6 
Resistance ( × 1012MBR(5+0.5R) MBR(10+1R) MBR(15+1.5R) MBR(Forward) 
Cake layer resistant 3.01 ± 0.13 3.32 ± 0.11 3.61 ± 0.15 3.9 ± 0.2 
Pores resistance 2.28 ± 0.09 2.57 ± 0.075 3.01 ± 0.11 3.17 ± 0.17 
Intrinsic resistance 0.74 ± 0.05 0.74 ± 0.044 0.63 ± 0.08 0.74 ± 0.034 
Total resistance 6.03 ± 0.09 6.63 ± 0.076 7.25 ± 0.11 7.81 ± 0.13 
Rc/Rt (%) 50 50 50 50 
Rp/Rt (%) 37.8 38.8 41.5 40.6 

Numbers of samples were 3.

Cake resistance contributes ∼50% of the total resistance in all the scenarios of relaxation run. On the other hand, the pore resistance contribution to the total resistance increases as the cycle time increases. The percentage ratio of pore resistance to total resistance (Rp/Rt) of MBR(5+0.5R), MBR(10+1R), MBR(15+1.5R) and MBR(Forward), was 37.8, 38.8, 41.5 and 40.6%, respectively. A study was conducted to evaluate the effect of relaxation (filtration <2 s, relaxation <1 s) on fouling, revealed the fouling cake resistance up to 50 × 1011 m−1 for flux of 50 L/m2·h, less as compared to continuous run (Defrance & Jaffrin 1999). Similarly, another study conducted to evaluate different filtration-relaxation modes. The total resistance of the best scenario (8 min run, 2 min relaxation) was 2.7 × 1012 (m−1), which was low as compared to the present study results. It was due to the lower flux (15 L/m2·h) of the reported study (Maqbool et al. 2014).

Treatment efficiency evaluation

The removal efficiencies of COD, TOC, TP, and TKN, in the relaxation run trial, are presented in Figure 3. The COD removal efficiencies were highest in the MBR(Forward). The COD removal efficiencies in MBR(5+0.5R), MBR(10+1R), MBR(15+1.5R) and MBR(Forward) were 89.32 ± 2.1%, 91.43 ± 2.6%, 92.34 ± 3.3%, and 94.55 ± 2.57%, respectively. The removal efficiency increased as the cycle time of run-relaxation was increased. The COD was effectively reduced (90–95%) in each scenario of run-relaxation. A study reported 90–93% COD removal efficiencies at different run relaxation modes of MBR at flux of 15 L/m2-h and HRT of 4 h (Maqbool et al. 2014). Similarly, the removal efficiencies for the TOC, TP, and TKN increased as the cycle time of run-relaxation was increased. The highest removal efficiencies were in MBR(Forward); TOC (85.1 ± 3.3%), TP (48.76 ± 4.13%), and TKN (50.5 ± 3.72%). The removal efficiencies in MBR(5+0.5R) were: TOC (80.4 ± 2.3%), TP (43.96 ± 3.35%), and TKN (40.5 ± 2.9%). The MBR(10+1R) removal efficiencies observed were: TOC (82.5 ± 3.1%), TP (44.21 ± 2.75%), and TKN (43.3 ± 2.7%). Similarly, the MBR(15+1.5R) removal efficiencies were: TOC (84.1 ± 2.7%), TP (46.45 ± 3.77%), and TKN (45.2 ± 2.8%). The removal efficiencies were slightly higher in the MBR(Forward) as the cake and pore resistance is higher. Consequently, the wastewater had to pass through more thick cake (more biological activity) and small pores (improved retaining of particles), which causes increase in removal efficiencies. Similarly, in other scenario more cake and small pores increased the removal efficiency a little.
Figure 3

Treatment efficiencies of different scenarios of relaxation.

Figure 3

Treatment efficiencies of different scenarios of relaxation.

Backwashing Run

Membrane fouling trends

Trends of membrane fouling at different scenarios of backwashing are given in Figure 4. The trends explore the fouling behavior of different scenarios; MBR(Forward), MBR(5+0.5B), MBR(10+1B), and MBR(15+1.5B). The membrane fouled in 19.5, 32.7, 31.3, and 27.74 h in MBR(Forward), MBR(5+0.5B), MBR(10+1B) and MBR(15+1.5B) scenario, respectively. The MBR(5+0.5B), MBR(10+1B) and MBR(15+1.5B) operation time was 1.68, 1.60, and 1.42 times greater than MBR(Forward), respectively. The results showed that membrane in scenario MBR(15+1.5B) fouled early as compared to MBR(5+0.5B) and MBR(10+1B). The MBR(5+0.5B) fouled after longer time as compared to the other scenarios (Table 5). The maturation phase time of the MBR(5+0.5B), MBR(10+1B), MBR(15+1.5B), and MBR(Forward) was 5.3, 2.6, 2.3, and 1.9 h. The steady phase time of the MBR(5+0.5B), MBR(10+1B), MBR(15+1.5B), and MBR(Forward) was 22.1, 23.3, 19.54 and 11.6 h, respectively. The jump phase time of the MBR(5+0.5B), MBR(10+1B), MBR(15+1.5B), and MBR(Forward) was 5.3, 5.4, 5.9 and 5.7 h. The maximum fouling rates observed in the jump phase. Overall, the fouling rates increased as the cycle time of the run-backwashing trial was increased in each phase. The average fouling rates in the MBR(5+0.5B), MBR(10+1B), MBR(15+1.5B), and MBR(Forward) were 0.94, 0.96, 1.06 and 1.64 KPa/h, respectively. A study evaluated the backwashing effect reported the 8–10 KPa/h for different run time to backwashing time ratios. The flux of 20 L/m2·h was maintained, while the MLSS concentration (9–10 g/L) was very high as compared to present study (Wu et al. 2008).
Table 5

Pressure range, phase time and fouling rates of maturation, steady and jump phases in backwash filtration scenarios

Filtration scenario (relaxation) Maturation phase Steady phase Jump phase Average fouling rates (KPa/h) 
MBR(Forward) 
 Pressure range (KPa) 3.34–9.0 9.0–19.5 19.5–35.3 1.64 ± 0.22 
 Phase time (h) 1.9 ± 0.15 11.6 ± 1.1 5.7 ± 0.33 
 Fouling rate (KPa/h) 2.98 ± 0.17 1.68 ± 0.11 3.65 ± 0.15 
MBR(5+0.5B) 
 Pressure range (KPa) 4.2–9.5 9.5–16.2 16.2–35.0 0.94 ± 0.13 
 Phase time (h) 5.3 ± 0.24 22.1 ± 2.1 5.3 ± 0.37 
 Fouling rate (KPa/h) 1.0 ± 0.013 0.3 ± 0.01 1.75 ± 0.13 
MBR(10+1B) 
 Pressure range (KPa) 4.87–8.77 8.77–16.45 16.45–35.03 0.96 ± 0.09 
 Phase time (h) 2.6 ± 0.11 23.3 ± 2.3 5.4 ± 0.73 
 Fouling rate (KPa/h) 1.6 ± 0.043 0.33 ± 0.02 3.44 ± 0.22 
MBR(15+1.5B) 
 Pressure range (KPa) 5.65–9.0 9.0–19.0 19.0–35.0 1.06 ± 0.14 
 Phase time (h) 2.3 ± 0.087 19.54 ± 1.77 5.9 ± 0.16 
 Fouling rate (KPa/h) 1.45 ± 0.135 0.51 ± 0.12 2.7 ± 0.27 
Filtration scenario (relaxation) Maturation phase Steady phase Jump phase Average fouling rates (KPa/h) 
MBR(Forward) 
 Pressure range (KPa) 3.34–9.0 9.0–19.5 19.5–35.3 1.64 ± 0.22 
 Phase time (h) 1.9 ± 0.15 11.6 ± 1.1 5.7 ± 0.33 
 Fouling rate (KPa/h) 2.98 ± 0.17 1.68 ± 0.11 3.65 ± 0.15 
MBR(5+0.5B) 
 Pressure range (KPa) 4.2–9.5 9.5–16.2 16.2–35.0 0.94 ± 0.13 
 Phase time (h) 5.3 ± 0.24 22.1 ± 2.1 5.3 ± 0.37 
 Fouling rate (KPa/h) 1.0 ± 0.013 0.3 ± 0.01 1.75 ± 0.13 
MBR(10+1B) 
 Pressure range (KPa) 4.87–8.77 8.77–16.45 16.45–35.03 0.96 ± 0.09 
 Phase time (h) 2.6 ± 0.11 23.3 ± 2.3 5.4 ± 0.73 
 Fouling rate (KPa/h) 1.6 ± 0.043 0.33 ± 0.02 3.44 ± 0.22 
MBR(15+1.5B) 
 Pressure range (KPa) 5.65–9.0 9.0–19.0 19.0–35.0 1.06 ± 0.14 
 Phase time (h) 2.3 ± 0.087 19.54 ± 1.77 5.9 ± 0.16 
 Fouling rate (KPa/h) 1.45 ± 0.135 0.51 ± 0.12 2.7 ± 0.27 

Numbers of samples were three.

Figure 4

Membrane fouling pattern of different scenarios of backwashing.

Figure 4

Membrane fouling pattern of different scenarios of backwashing.

Fouling resistances

Cake and pore resistances are the main reasons for the shorter run of the membrane in MBR. Cake resistance in the MBR(5+0.5B), MBR(10+1B), MBR(15+1.5B), and MBR(Forward) were 3.69 × 1012, 3.6 × 1012, 3.39 × 1012, and 3.9 × 1012 m−1, respectively (Table 6). It showed some aberrant trend. The increase in the cycle time reduced the cake resistance. In the longer cycles of run-backwash trial, the backwash lasted for longer time which effectively reduced the cake layer, while in the shorter cycles, the backwash lasted for shorter times, comparatively, which could not remove the cake as well as the longer time backwash. Pore resistance in different scenarios: MBR(5+0.5B), MBR(10+1B), MBR(15+1.5B), and MBR(Forward) were 2.18 × 1012, 2.51 × 1012, 2.70 × 1012, and 3.9 × 1012 m−1, respectively. Pore resistance increased gradually as the backwash-run cycle time increased. A study reported total resistance of 5 × 1012 (m−1) for the continuous run, and 2.6 × 1012 (m−1) for the backwashing-run scenario (240 s continuous run, 20 s backwashing with 50 L/m2·h flux). The study operated at 20 L/m2·h flux, and MLSS concentration was 9–10 g/L (Wu et al. 2008). Another study evaluated the effect of backwashing (run time 300 s, backwashing 5 s) at different fluxes using an ultrafiltration membrane. The total resistance was 2.6 × 1011 (m−1) at 18 L/m2·h flux. The MLSS concentration of study was low (3 g/L) and air sparging was also applied for fouling removal (Jiang et al. 2003). A similar study evaluated the scenario of backwashing at different fluxes. The total resistance at flux 20 L/m2·h flux was 2.2 × 1012 (m−1) for a scenario (5 min run followed by 15 s backwashing) (Bouhabila et al. 2001).

Table 6

Cake, pore and intrinsic resistances of membrane in backwash filtration scenarios

Resistance (X 1012MBR(5+0.5B) MBR(10+1B) MBR(15+1.5B) MBR(Forward) 
Cake layer resistant 3.69 ± 0.25 3.6 ± 0.19 3.39 ± 0.18 3.9 ± 0.2 
Pores resistance 2.18 ± 0.17 2.51 ± 0.13 2.70 ± 0.15 3.17 ± 0.17 
Intrinsic resistance 0.63 ± 0.08 0.76 ± 0.04 0.77 ± 0.077 0.74 ± 0.034 
Total resistance 6.5 ± 0.17 6.87 ± 0.12 6.86 ± 0.135 7.81 ± 0.13 
Rc/Rt (%) 57 52 49 50 
Rp/Rt (%) 34 37 39 40.6 
Resistance (X 1012MBR(5+0.5B) MBR(10+1B) MBR(15+1.5B) MBR(Forward) 
Cake layer resistant 3.69 ± 0.25 3.6 ± 0.19 3.39 ± 0.18 3.9 ± 0.2 
Pores resistance 2.18 ± 0.17 2.51 ± 0.13 2.70 ± 0.15 3.17 ± 0.17 
Intrinsic resistance 0.63 ± 0.08 0.76 ± 0.04 0.77 ± 0.077 0.74 ± 0.034 
Total resistance 6.5 ± 0.17 6.87 ± 0.12 6.86 ± 0.135 7.81 ± 0.13 
Rc/Rt (%) 57 52 49 50 
Rp/Rt (%) 34 37 39 40.6 

Numbers of samples were three.

Treatment efficiency evaluation

Removal efficiencies of backwash run trial are presented in Figure 5. Removal efficiencies of COD, TOC, TP, and TKN in the MBR(5+0.5B), were 85.22 ± 2.7%, 73.4 ± 2.7%, 30.89 ± 2.9%, and 38.6 ± 3.7%, respectively. Similarly, the removal efficiencies of COD, TOC, TP, and TKN, in the MBR(10+1B), were 87.32 ± 3.1%, 76.8 ± 3.5%, 34.5 ± 3.3%, and 41.7 ± 4.1%. The MBR(15+1.5B) resulted the following removal efficiencies; COD (88.21 ± 3.2%), TOC (77.4 ± 2.8%), TP (35.22 ± 3.1%), and TKN (42.7 ± 3.1%). The increase in the removal efficiency is due to the decrease in the pore size caused by the pore higher blockage rates in longer cycle time. These trends revealed the gradual reduction in the treatment efficiencies as the cycle time of the backwash run increased. Nitrogen can be removed by nitrification and denitrification processes or assimilation. In the longer aeration times the removal efficiency of nitrogen reduced due to the inhabitation of denitrification process (Han et al. 2005; Guadie et al. 2014). The longer aeration time in the present study reduced the nitrogen removal efficiency. The phosphorus removal is dependent on assimilation or the luxury uptake of phosphate without any need (Rosenberger et al. 2002). The removal efficiencies of the phosphate were also low due to the lower HRT (4 hr).
Figure 5

Treatment efficiencies of different scenarios of backwash trial.

Figure 5

Treatment efficiencies of different scenarios of backwash trial.

Comparison of relaxation and backwashing run

Comparison of membrane fouling trends

It was observed that the fouling rates in the relaxation trial are much higher as compared to the backwashing trial. The fouling rate of MBR(5+0.5B) was 23% less as compared to the MBR(5+0.5R). Similarly, the fouling rates of the MBR(10+1B) and MBR(15+1.5B) were 28.3% and 27% less as compared to the MBR(10+1R) and MBR(15+1.5R), respectively. Different scenarios of relaxation and backwashing operation run time was more as compared to MBR(Forward); MBR(5+0.5R) (7.5 h), MBR(10+1R) (5.5 h), MBR(15+1.5R) (2 h), MBR(5+0.5B) (13.2 h), MBR(10+1B) (11.8 h) and MBR(15+1.5B) (8.24 h).

It is revealed from the results that the smaller run cycle time prolongs the membrane operation, whether it is relaxation-run trial or backwashing-run trial. The MBR(5+0.5B) and MBR(5+0.5R) fouling rates were 42.7% and 25.6%, respectively, less as compared to MBR(Forward). The MBR(5+0.5B) observed the longest run of 33 h, while the MBR(5+0.5R) run was 27 h, the longest one in the relaxation-run trial. Linear regression analysis showed a strong linear relation between the cycle time and fouling rates, i.e., R2 = 0.999 for relaxation trial and R2 = 0.87 for backwashing trial (Figure 6). The pores and cake fouling rates of relaxation and backwashing at different cycle times can be observed in Figure 6 as well. The fouling rates in the backwashing scenarios are less as compared to the relaxation scenarios. The backwashing of permeate removes the cake layer and pores foulant effectively as compared to the relaxation.
Figure 6

Regression analysis – cycle time versus fouling rates.

Figure 6

Regression analysis – cycle time versus fouling rates.

Comparison of fouling resistances

Cake layer resistances in the relaxation run scenarios were less as compared to the backwash run scenarios, which was due to longer run of backwashing trails. On the contrary, the pore resistances in the relaxation run scenarios were more as compared to the backwash run scenarios, which revealed that backwashing effectively removed pores blockage, even after a long run, compared to relaxation trials (Figures 7 and 8). The Rc/Rt decreased gradually as the cycle time increased, in the backwash scenarios: MBR(5+0.5B) (57%), MBR(10+1B) (52%), and MBR(15+1.5B) (49%). On the other hand, the Rc/Rt values in the relaxation run scenarios remain constant, i.e., 50%. In the case of Rp/Rt, the percentage decreased in both cases. A strong linear relation of cake resistance (R2 = 0.999) and pore resistance (R2 = 0.986) was established with cycle time in relaxation trial (Figure 7). Similarly, for the backwashing trial, the linear relation was established between cake resistance (R2 = 0.946) and pore resistance (R2 = 0.976) versus cycle time (Figure 8).
Figure 7

Regression analysis: filtration-relaxation cycle time versus pore and cake resistance.

Figure 7

Regression analysis: filtration-relaxation cycle time versus pore and cake resistance.

Figure 8

Regression analysis: filtration-backwashing cycle time versus pore and cake resistance.

Figure 8

Regression analysis: filtration-backwashing cycle time versus pore and cake resistance.

Comparison of treatment efficiency evaluation

It was observed that removal efficiencies in the relaxation run scenarios were slightly higher as compared to the backwashing run trial scenarios. The reason was higher pores fouling and less cake removal in the relaxation run scenarios, which reduce the pore size, as compared to the backwashing run trial scenarios. The cake microorganisms are very effective in the removal of nutrients, so the nutrients removal in the relaxation scenario were higher.

For the optimal design of the MBR operation to reduce the biofouling, and to increase the operation time without washing, a short run cycle of backwashing and relaxation are feasible and effective, as short as possible. Nevertheless, the short cycles would require more frequent on and off of pumps. So the pumps with shorter cycle time would be required to pump out the wastewater through the membranes. Furthermore, the backwashing is the better option as compared to the relaxation, even though both help to optimize the operation.

CONCLUSION

The study revealed that the backwashing was more appropriate technique to control the biofouling in MBR as compared to the relaxation. Furthermore, both backwashing and relaxation were affective, if applied after a small interval of time, i.e., small cycle time. The lowest fouling rates were observed in MBR(5+0.5B), due to effective removal of cake and pore foulant. The fouling rates of the backwashing scenarios were less as compared to the relaxation scenarios. On the contrary, the removal efficiencies of organic pollutants and nutrients were insignificantly higher in relaxation scenario.

ACKNOWLEDGEMENTS

The author thankfully acknowledges the research grant from the University of Engineering and Technology, Taxila, Pakistan.

REFERENCES

REFERENCES
Abdelrasoul
A.
Doan
H.
Lohi
A.
2013
Fouling in membrane filtration and remediation methods
. In:
Nakajima
H.
(ed.),
Mass Transfer-Advances in Sustainable Energy and Environment Oriented Numerical Modeling
,
InTech
, pp.
195
218
.
American Public Health Association, American Water Works Association, Water Environment Federation
2005
Standard Methods for the Examination of Water and Wastewater
.
American Public Health Association (APHA)
,
Washington, DC
,
USA
.
Bouhabila
E. H.
Aїm
R. B.
Buisson
H.
2001
Fouling characterisation in membrane bioreactors
.
Separation and Purification Technology
22
,
123
132
.
Carroll
T.
King
S.
Gray
S.
Bolto
B. A.
Booker
N.
2000
The fouling of microfiltration membranes by NOM after coagulation treatment
.
Water Research
34
(
11
),
2861
2868
.
Defrance
L.
Jaffrin
M.
1999
Reversibility of fouling formed in activated sludge filtration
.
Journal of Membrane Science
157
(
1
),
73
84
.
Gao
D.-W.
Fu
Y.
Tao
Y.
Li
X.-X.
Xing
M.
Gao
X.-H.
Ren
N.-Q.
2011
Linking microbial community structure to membrane biofouling associated with varying dissolved oxygen concentrations
.
Bioresource Technology.
102
,
5626
5633
.
Han
S.-S.
Bae
T.-H.
Jang
G.-G.
Tak
T.-M.
2005
Influence of sludge retention time on membrane fouling and bioactivities in membrane bioreactor system
.
Process Biochemistry
40
(
7
),
2393
2400
.
Jarusutthirak
C.
Amy
G.
2006
Role of soluble microbial products (SMP) in membrane fouling and flux decline
.
Environmental Science & Technology
40
(
3
),
969
974
.
Jeong
T. Y.
Cha
G. C.
Yoo
I. K.
Kim
D. J.
2007
Characteristics of bio-fouling in a submerged MBR
.
Desalination
207
,
107
113
.
Jiang
T.
Kennedy
M. D.
van der Meer
W. G.
Vanrolleghem
P. A.
Schippers
J. C.
2003
The role of blocking and cake filtration in MBR fouling
.
Desalination
157
(
1
),
335
343
.
Judd
S.
Robinson
T.
Holdner
J.
Alvarez-Vazquez
H.
Jefferson
B.
2004
Impact of membrane material on membrane bioreactor permeability
. In:
Proceedings of the Water Environment-Membrane Technology Conference
,
Seoul, South Korea
.
Kennedy
M.
Kamanyi
J.
Rodríguez
S.
Lee
N.
Schippers
J.
Amy
G.
2008
Water treatment by microfiltration and ultrafiltration
. In:
Li
N. N.
Fane
A. G.
Ho
W. S. W.
Matsuura
T.
(eds),
Advanced Membrane Technology and Applications
.
John Wiley & Sons, Inc.
,
Hoboken, NJ, USA
, pp.
131
170
.
Lee
E.-J.
Kim
K.-Y.
Lee
Y.-S.
Nam
J.-W.
Lee
Y.-S.
Kim
H.-S.
Jang
A.
2012
A study on the high-flux MBR system using PTFE flat sheet membranes with chemical backwashing
.
Desalination
306
,
35
40
.
Maton
L.
Psarras
G.
Kasapakis
G.
Lorenzen
J. R.
Andersen
M.
Boesen
M.
Bak
S. N.
Chartzoulakis
K.
Pedersen
S. M.
Kloppmann
W.
2010
Assessing the net benefits of using wastewater treated with a membrane bioreactor for irrigating vegetables in Crete
.
Agricultural Water Management
98
(
3
),
458
464
.
Monclús
H.
Sipma
J.
Ferrero
G.
Rodriguez-Roda
I.
Comas
J.
2010
Biological nutrient removal in an MBR treating municipal wastewater with special focus on biological phosphorus removal
.
Bioresource Technology
101
(
11
),
3984
3991
.
Nywening
J.-P.
MacPhail
M.
Zhou
H.
Behmann
H.
2007
Influence of permeation relaxation on particle dispersion in wastewater membrane bioreactor processes
.
Proceedings of the Water Environment Federation
2007
(
11
),
6496
6519
.
Rosenberger
S.
Krüger
U.
Witzig
R.
Manz
W.
Szewzyk
U.
Kraume
M.
2002
Performance of a bioreactor with submerged membranes for aerobic treatment of municipal waste water
.
Water Research
36
(
2
),
413
420
.
Sipma
J.
Osuna
B.
Collado
N.
Monclús
H.
Ferrero
G.
Comas
J.
Rodriguez-Roda
I.
2010
Comparison of removal of pharmaceuticals in MBR and activated sludge systems
.
Desalination
250
(
2
),
653
659
.
Smith
P. J.
Vigneswaran
S.
Ngo
H. H.
Ben-Aim
R.
Nguyen
H.
2006
A new approach to backwash initiation in membrane systems
.
Journal of Membrane Science
278
(
1
),
381
389
.
Sven
L.
Al-Halbouni
D. &
Thomas
W.
Lyko
S.
Al-Halbouni
D.
Wintgens
T.
Janot
A.
Hollender
J.
Dott
W.
Melin
T.
2007
Polymeric compounds in activated sludge supernatant-- characterisation and retention mechanisms at a full-scale municipal membrane bioreactor
.
Water Research
41
,
3894
3902
.
Tabraiz
S.
Nasreen
S.
Qureshi
L. A.
Zeeshan
M.
Ahmad
I.
Ahmed
S.
Hassan
Z.
2015b
Sewage land disposal and unpaved drains: threat to groundwater quality
.
Desalination and Water Treatment
57
(
43
),
20464
20469
.
Tadkaew
N.
Hai
F. I.
McDonald
J. A.
Khan
S. J.
Nghiem
L. D.
2011
Removal of trace organics by MBR treatment: the role of molecular properties
.
Water Research
45
(
8
),
2439
2451
.
Ueda
T.
Hata
K.
Kikuoka
Y.
Seino
O.
1997
Effects of aeration on suction pressure in a submerged membrane bioreactor
.
Water Research.
31
,
489
494
.
Wu
J.
Le-Clech
P.
Stuetz
R. M.
Fane
A. G.
Chen
V.
2008
Effects of relaxation and backwashing conditions on fouling in membrane bioreactor
.
Journal of Membrane Science
324
(
1
),
26
32
.
Yigit
N.
Civelekoglu
G.
Harman
I.
Koseoglu
H.
Kitis
M.
2010
Effects of various backwash scenarios on membrane fouling in a membrane bioreactor
. In:
Survival and Sustainability
.
Elsevier
,
Oxford, UK
, pp.
917
929
.
Zhang
J. S.
Chuan
C. H.
Zhou
J. T.
Fane
A. G.
2006
Effect of sludge retention time on membrane bio-fouling intensity in a submerged membrane bioreactor
.
Separation Science and Technology
41
,
1313
1329
.
Zhao
Y.-J.
Wu
K.-F.
Wang
Z.-J.
Zhao
L.
Li
S.-S.
2000
Fouling and cleaning of membrane-a literature review
.
Journal of Environmental Sciences-Beijing
12
(
2
),
241
251
.