w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Full-scale validation of an air scour control system for energy savings in membrane bioreactors n b, Hector Monclu´s a, Montserrat Dalmau a, Sara Gabarro Giuliana Ferrero c, Ignasi Rodrı´guez-Roda a,b, Joaquim Comas a,b,* a
LEQUIA, Institute of the Environment, University of Girona, E17071, Girona, Catalonia, Spain ICRA (Catalan Institute for Water Research), Scientific and Technological Park of the University of Girona, H2O Building, c/ Emili Grahit 101, E17003, Girona, Spain c UNESCO-IHE Institute for Water Education, Westvest 7, 2611 AX, Delft, The Netherlands b
article info
abstract
Article history:
Membrane aeration represents between 35 and 50% of the operational cost of membrane
Received 12 January 2015
bioreactors (MBR). New automatic control systems and/or module configurations have
Received in revised form
been developed for aeration optimization. In this paper, we briefly describe an innovative
24 March 2015
MBR air scour control system based on permeability evolution and present the results of a
Accepted 25 March 2015
full-scale validation that lasted over a 1-year period. An average reduction in the air scour
Available online 17 April 2015
flow rate of 13% was achieved, limiting the maximum reduction to 20%. This averaged reduction corresponded to a decrease in energy consumption for membrane aeration of
Keywords:
14% (0.025 kWh m3) with maximum saving rates of 22% (0.04 kWh m3). Permeability and
Air scour
fouling rate evolution were not affected by the air scour control system, as very similar
Control
behavior was observed for these variables for both filtration lines throughout the entire
Energy savings
experimental evaluation period of 1 year.
Full scale
© 2015 Elsevier Ltd. All rights reserved.
MBR Optimization
1.
Introduction
Since the late 1990s, several improvements in filtration operational conditions, coupled with a decrease in the total capital expenditure for membrane bioreactors (MBRs), have led to the widespread adoption of MBR technology in municipal wastewater treatment facilities (Ferrero et al., 2012). As a result of the stabilization of market expansion at the turn of the 21st
century, MBR is currently a consolidated technology (Judd, 2008, 2011; Lesjean et al., 2009; Santos et al., 2011). The energy requirements of MBR systems are still greater than those of conventional activated sludge systems coupled with tertiary treatment. Membrane air scouring is a process that contributes markedly to the total operating costs (Fenu et al., 2010; Santos et al., 2011; Verrecht et al., 2010). Membrane aeration represents between 35 and 50% of operational costs (Krzeminski et al., 2012a, 2012b), but this range can be
* Corresponding author. LEQUIA, Institute of the Environment, University of Girona, E17071, Girona, Catalonia, Spain. Tel.: þ34 972418355. n), E-mail addresses: [email protected] (H. Monclu´s), [email protected] (M. Dalmau), [email protected] (S. Gabarro [email protected] (G. Ferrero), [email protected], [email protected] (I. Rodrı´guez-Roda), [email protected], [email protected] (J. Comas). http://dx.doi.org/10.1016/j.watres.2015.03.032 0043-1354/© 2015 Elsevier Ltd. All rights reserved.
2
w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
extended up to 80% if aeration for biological purposes is also taken into account (Barillon et al., 2013). The membrane air scour flow rates applied in practice are generally very conservative relative to the MBR manufacturers' recommendations. The most recent developments on the module configuration and aeration operational strategies have resulted in considerable aeration improvements (i.e., reusing the air scour between different modules, using aeration intermittently or using cyclic aeration for hollow fiber membranes (Barillon et al., 2013; Drews, 2010; Judd, 2011; Verrecht et al., 2008)). However, there are no robust control systems that can reduce membrane aeration requirements while maintaining optimum filtration performance. In most cases, the air scour flow rate remains fixed regardless of the sludge quality or the membrane permeability. Thus, automatic control systems are needed to optimize aeration in terms of the membrane performance and the sludge characteristics. Ferrero et al. (2012) compared different control systems, which had the common objective of reducing operational costs by modifying operational conditions (the flux, the cycle length, etc.). These authors noted that the number of patents and scientific publications on control systems for membrane bioreactors is still very limited. In addition, these patents are usually very general and include many assumptions to cover a wide spectrum of intellectual property. However, the statements in patents are not always scientifically proven or validated, at least over the long term (Ferrero et al., 2012). There are very few research publications on control systems that have been developed at the pilot scale, and none of these systems has been validated in closed-loop at full scale (Huyskens et al., 2011; Lorain et al., 2010). Ferrero et al. (2011b) developed an automatic control algorithm for reducing air scour based on permeability evolution, which was successfully tested at the pilot scale (Ferrero et al., 2011a). The aim of this paper is to monitor and discuss the effects of the implementation of this air scour control system in a full scale municipal MBR. Membrane filtration, sludge quality and biological nutrient removal were assessed by means of fouling behavior, fouling rates, energy consumptions, sludge characteristics and nutrient removal efficiencies.
2.
Materials and methods
2.1.
WWTP La Bisbal d'Emporda
The membrane air scour control system was validated over a WWTP (LBE, North two-year period at La Bisbal d'Emporda East of Spain). This full-scale MBR plant was designed to treat a maximum daily flow rate of 3200 m3 of municipal wastewater. The treatment steps consist of a coarse screen (8 cm), a grit chamber, a buffer tank (1110 m3), a fine screen (1 mm), an oxidation ditch bioreactor (3636 m3), two parallel membrane lines (30 m3) with submerged ultrafiltration membranes (ZeeWeed 500C, Zenon, GE), and an additional secondary settler that is used during wet weather and peak flows. Each filtration line is equipped with an independent positive displacement air blower (GM 25S, Aerzen, Germany) that provides a design air scour flow rate of 17.8 m3 min1 operating at 3510 rpm (50 Hz), corresponding to a power consumption of 12.06 kW and an energy consumption of 289.5 kWh per day. When a variable frequency drive (VFD) was installed the power consumption improved from 12.15 kW to 12.06 kW, but to avoid overestimation and to evaluate objectively the fouling all the energy savings were calculated based on the difference between the power consumption at different air-scour flow rates minus the power consumption at 3510 rpm (12.06 kW) of the same blower A. A detailed description of the MBR is presented in Table 1. The aeration strategy adopted in this facility is 6 s on and 6 s off (Adams et al., 2011). The membranes were installed in 2003, and only 5.7% of the total area has been replaced over 10 years of operation. A visual inspection carried out in 2011 revealed that the membranes were in good condition (i.e., less than 1% of the membranes were damaged). Chemically-enhanced backflush (CEB) is performed as a maintenance routine procedure when the TMP exceeds a threshold of 0.4 bars (usually every week). First, a solution of 140 mg L1 sodium hypochlorite and 200 mg L1 EDTA is applied for 5 backflush pulses of 15 s, with a 5-
WWTP. Table 1 e Membrane characteristics at La Bisbal d'Emporda Parameter
Units
Membranes model Membrane filtration lines MBR capacity Total membrane area Pore size Design air scour flow rate Power consumption of Air-scour blowers
m3 day1 m2 mm m3 min1 kW
Aeration strategy Design SADma Design SADpb Filtration cycle Average permeate flux Average permeate net flux
seconds Nm3 m2 h1 Nm3 h1 m3 h1 minutes/seconds L m2 h1 L m2 h1
a b
SADm: Specific Aeration Demand with respect to membrane area, according to membrane manufacturer. SADp: Specific Aeration Demand with respect to permeate flow, according to membrane manufacturer.
ZeeWeed 500c (Zenon, GE) 2 3200 (1600 2) 5808 (2904 2) 0.04 17.8 Blower A (12.06 ± 0.3) Blower B (12.13 ± 0.5) 6 on/6 off 0.37 16.02 10 filtration/40 backpulse 23 ± 2 21.6 ± 2
3
w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
min relaxation between pulses, at a flux of approximately 23 L m2 h1. Every 6 months, recovery chemical cleanings are carried out by soaking the membranes in an alkaline solution (a 1500 mg L1 sodium hypochlorite solution) or an acid solution (a 1500 mg L1citric acid solution) for 6e12 h. Manual n et al., 2013). cleanings are performed every two years (Gabarro
Case 4: The long term permeability slope is positive or equal to zero (LT 0), and the short term permeability is negative (ST < 0), corresponding to a moderate or low energy saving opportunity.
2.3. 2.2.
The experimental study was performed in the LBE WWTP in two phases (two years).
Air scour control system
The air scour control system, registered as Smart Air MBR, modifies the membrane air scour flow rate based on the permeability evolution, which is used as an indicator of the membrane performance and the sludge characteristics (Comas et al., 2010). The TMP and the permeate flow rate are monitored in real-time. These variables are first processed for data quality assurance and then cycle and daily averaged values are calculated. Long term (LT) and short term (ST) permeability trends are evaluated and compared to identify energy saving opportunities and alarms. The controlled variable is the slope ratio (SR), which is calculated as the ratio between the ST and the LT permeabilities. The ST was determined to be 4 days, and the LT was determined to be 14 days (Ferrero et al., 2011b). The SR is proportional to the control action, which in this case is the daily regulation of the air scour flow rate by increasing or decreasing the blower frequency (the manipulated variable) (Ferrero et al., 2011b). Normally, the permeability decreases in proportion to the cake layer resistance, which is roughly proportional to the volume of sludge filtered (Judd, 2011). However, corrective actions such as chemical cleaning or a decrease in the permeate flux can cause the permeability to increase. Operational conditions (the temperature, the suspended solids concentration, the sludge age, etc.) can also affect the sludge permeability. Thus, four different scenarios (i.e., relationships between ST and LT) have been defined for the air scour control system (see Fig. 1).
Phase 1: Data collection and calibration of the air scour control system (12 months from April 2011). During the first 8 months different implementation tasks were performed (monitoring energy consumption of blowers, frequency driver installation, monitoring filtration variables and adaptation of the control system). The air scour control system was implemented in filtration line A but operated in an open-loop regime for the last 4 months (Phase 1 Table 1) under the conditions recommended by the membrane suppliers, thus same operation conditions for filtration lines A and B. Data on the TMP, the permeate flux and the working hours per day of both process lines were collected in real time, processed and compared. The information acquired over this period was used to calibrate the control system parameters. Phase 2: Validation (12 months, from April 2012). The air scour control system was operated in a closed-loop regime for filtration line A, whereas line B was used as a reference with a constant air scour flow rate of 17.8 m3 min1 (as recommended by the membrane manufacturers). Fig. 2 is a schematic of the membrane tank that shows the two filtration lines and the equipment for the control system. In Phase 2, the maximum flow rate reduction allowed for the air scour control system in line A was increased progressively from 5% to 10%, 15% and 20% of the air scour flow rate.
Case 1: The long term permeability slope is negative (LT < 0), and the short term permeability slope is positive or equal to zero (ST 0), corresponding to a moderate energy saving opportunity. Case 2: The long term permeability slope is negative (LT < 0), and the short term permeability slope is negative (ST < 0, lower or higher than LT), corresponding to a low or a non-energy saving opportunity. Case 3: The long term permeability slope is positive or equal to zero (LT 0), and the short term permeability is positive or equal to zero (ST 0, lower or higher than LT) corresponding to a moderate or high energy saving opportunity.
Case 1
K
Experimental plan
Case 2
K
Chemical oxygen demand (COD), biochemical oxygen demand (BOD), nitrogen and phosphates species, mixed liquor suspended solids (MLSS) and sludge volumetric index (SVI) were analyzed in accordance with Standard Methods (APHA, 2005). The capillarity suction time (CST) (Triton electronics Ltd., type 304 B) (Scholz, 2005) was used to determine the sludge filterability. The permeability decay was used as an indicator of the general fouling evolution, and the fouling rate (which was measured as the slope of the permeability for each permeate
Case 3
K ST
LT
LT
ST time
Experimental monitoring
Case 4
K
ST
ST LT
2.4.
time
ST LT
time
Fig. 1 e Different relationships between ST and LT permeability.
ST
time
4
w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
TMP-A Flux-A
7
7
TMP-B Flux-B
9 10
Control Server Data acquisition and processing
1 3
2
1 Membrane filtration line A 2 Membrane filtration line B 3 Feed to membrane tank 4 Variable speed drive
Line-B 3
Line-A
5 Line A Blower (controlled by Smart Air MBR) 6 Line B blower 7 Sensors and probes
Set point of blower r.p.m.
VSD
5
8
6
4
8 Recirculated stream to oxidation ditch 9 Permeate pipes 10 Server
Fig. 2 e Schematic diagram of the control system of the LBE WWTP.
cycle, as in Monclu´s et al., 2011), was used to estimate the fouling accumulated in each filtration cycle to ensure that there were no differences between the filtration line that was regulated by the air scour control system (line A) and the reference line B. Differences on FRs among both lines were determined by Student's t test for paired samples, where a pvalue
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Full-scale validation of an air scour control system for energy savings in membrane bioreactors n b, Hector Monclu´s a, Montserrat Dalmau a, Sara Gabarro Giuliana Ferrero c, Ignasi Rodrı´guez-Roda a,b, Joaquim Comas a,b,* a
LEQUIA, Institute of the Environment, University of Girona, E17071, Girona, Catalonia, Spain ICRA (Catalan Institute for Water Research), Scientific and Technological Park of the University of Girona, H2O Building, c/ Emili Grahit 101, E17003, Girona, Spain c UNESCO-IHE Institute for Water Education, Westvest 7, 2611 AX, Delft, The Netherlands b
article info
abstract
Article history:
Membrane aeration represents between 35 and 50% of the operational cost of membrane
Received 12 January 2015
bioreactors (MBR). New automatic control systems and/or module configurations have
Received in revised form
been developed for aeration optimization. In this paper, we briefly describe an innovative
24 March 2015
MBR air scour control system based on permeability evolution and present the results of a
Accepted 25 March 2015
full-scale validation that lasted over a 1-year period. An average reduction in the air scour
Available online 17 April 2015
flow rate of 13% was achieved, limiting the maximum reduction to 20%. This averaged reduction corresponded to a decrease in energy consumption for membrane aeration of
Keywords:
14% (0.025 kWh m3) with maximum saving rates of 22% (0.04 kWh m3). Permeability and
Air scour
fouling rate evolution were not affected by the air scour control system, as very similar
Control
behavior was observed for these variables for both filtration lines throughout the entire
Energy savings
experimental evaluation period of 1 year.
Full scale
© 2015 Elsevier Ltd. All rights reserved.
MBR Optimization
1.
Introduction
Since the late 1990s, several improvements in filtration operational conditions, coupled with a decrease in the total capital expenditure for membrane bioreactors (MBRs), have led to the widespread adoption of MBR technology in municipal wastewater treatment facilities (Ferrero et al., 2012). As a result of the stabilization of market expansion at the turn of the 21st
century, MBR is currently a consolidated technology (Judd, 2008, 2011; Lesjean et al., 2009; Santos et al., 2011). The energy requirements of MBR systems are still greater than those of conventional activated sludge systems coupled with tertiary treatment. Membrane air scouring is a process that contributes markedly to the total operating costs (Fenu et al., 2010; Santos et al., 2011; Verrecht et al., 2010). Membrane aeration represents between 35 and 50% of operational costs (Krzeminski et al., 2012a, 2012b), but this range can be
* Corresponding author. LEQUIA, Institute of the Environment, University of Girona, E17071, Girona, Catalonia, Spain. Tel.: þ34 972418355. n), E-mail addresses: [email protected] (H. Monclu´s), [email protected] (M. Dalmau), [email protected] (S. Gabarro [email protected] (G. Ferrero), [email protected], [email protected] (I. Rodrı´guez-Roda), [email protected], [email protected] (J. Comas). http://dx.doi.org/10.1016/j.watres.2015.03.032 0043-1354/© 2015 Elsevier Ltd. All rights reserved.
2
w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
extended up to 80% if aeration for biological purposes is also taken into account (Barillon et al., 2013). The membrane air scour flow rates applied in practice are generally very conservative relative to the MBR manufacturers' recommendations. The most recent developments on the module configuration and aeration operational strategies have resulted in considerable aeration improvements (i.e., reusing the air scour between different modules, using aeration intermittently or using cyclic aeration for hollow fiber membranes (Barillon et al., 2013; Drews, 2010; Judd, 2011; Verrecht et al., 2008)). However, there are no robust control systems that can reduce membrane aeration requirements while maintaining optimum filtration performance. In most cases, the air scour flow rate remains fixed regardless of the sludge quality or the membrane permeability. Thus, automatic control systems are needed to optimize aeration in terms of the membrane performance and the sludge characteristics. Ferrero et al. (2012) compared different control systems, which had the common objective of reducing operational costs by modifying operational conditions (the flux, the cycle length, etc.). These authors noted that the number of patents and scientific publications on control systems for membrane bioreactors is still very limited. In addition, these patents are usually very general and include many assumptions to cover a wide spectrum of intellectual property. However, the statements in patents are not always scientifically proven or validated, at least over the long term (Ferrero et al., 2012). There are very few research publications on control systems that have been developed at the pilot scale, and none of these systems has been validated in closed-loop at full scale (Huyskens et al., 2011; Lorain et al., 2010). Ferrero et al. (2011b) developed an automatic control algorithm for reducing air scour based on permeability evolution, which was successfully tested at the pilot scale (Ferrero et al., 2011a). The aim of this paper is to monitor and discuss the effects of the implementation of this air scour control system in a full scale municipal MBR. Membrane filtration, sludge quality and biological nutrient removal were assessed by means of fouling behavior, fouling rates, energy consumptions, sludge characteristics and nutrient removal efficiencies.
2.
Materials and methods
2.1.
WWTP La Bisbal d'Emporda
The membrane air scour control system was validated over a WWTP (LBE, North two-year period at La Bisbal d'Emporda East of Spain). This full-scale MBR plant was designed to treat a maximum daily flow rate of 3200 m3 of municipal wastewater. The treatment steps consist of a coarse screen (8 cm), a grit chamber, a buffer tank (1110 m3), a fine screen (1 mm), an oxidation ditch bioreactor (3636 m3), two parallel membrane lines (30 m3) with submerged ultrafiltration membranes (ZeeWeed 500C, Zenon, GE), and an additional secondary settler that is used during wet weather and peak flows. Each filtration line is equipped with an independent positive displacement air blower (GM 25S, Aerzen, Germany) that provides a design air scour flow rate of 17.8 m3 min1 operating at 3510 rpm (50 Hz), corresponding to a power consumption of 12.06 kW and an energy consumption of 289.5 kWh per day. When a variable frequency drive (VFD) was installed the power consumption improved from 12.15 kW to 12.06 kW, but to avoid overestimation and to evaluate objectively the fouling all the energy savings were calculated based on the difference between the power consumption at different air-scour flow rates minus the power consumption at 3510 rpm (12.06 kW) of the same blower A. A detailed description of the MBR is presented in Table 1. The aeration strategy adopted in this facility is 6 s on and 6 s off (Adams et al., 2011). The membranes were installed in 2003, and only 5.7% of the total area has been replaced over 10 years of operation. A visual inspection carried out in 2011 revealed that the membranes were in good condition (i.e., less than 1% of the membranes were damaged). Chemically-enhanced backflush (CEB) is performed as a maintenance routine procedure when the TMP exceeds a threshold of 0.4 bars (usually every week). First, a solution of 140 mg L1 sodium hypochlorite and 200 mg L1 EDTA is applied for 5 backflush pulses of 15 s, with a 5-
WWTP. Table 1 e Membrane characteristics at La Bisbal d'Emporda Parameter
Units
Membranes model Membrane filtration lines MBR capacity Total membrane area Pore size Design air scour flow rate Power consumption of Air-scour blowers
m3 day1 m2 mm m3 min1 kW
Aeration strategy Design SADma Design SADpb Filtration cycle Average permeate flux Average permeate net flux
seconds Nm3 m2 h1 Nm3 h1 m3 h1 minutes/seconds L m2 h1 L m2 h1
a b
SADm: Specific Aeration Demand with respect to membrane area, according to membrane manufacturer. SADp: Specific Aeration Demand with respect to permeate flow, according to membrane manufacturer.
ZeeWeed 500c (Zenon, GE) 2 3200 (1600 2) 5808 (2904 2) 0.04 17.8 Blower A (12.06 ± 0.3) Blower B (12.13 ± 0.5) 6 on/6 off 0.37 16.02 10 filtration/40 backpulse 23 ± 2 21.6 ± 2
3
w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
min relaxation between pulses, at a flux of approximately 23 L m2 h1. Every 6 months, recovery chemical cleanings are carried out by soaking the membranes in an alkaline solution (a 1500 mg L1 sodium hypochlorite solution) or an acid solution (a 1500 mg L1citric acid solution) for 6e12 h. Manual n et al., 2013). cleanings are performed every two years (Gabarro
Case 4: The long term permeability slope is positive or equal to zero (LT 0), and the short term permeability is negative (ST < 0), corresponding to a moderate or low energy saving opportunity.
2.3. 2.2.
The experimental study was performed in the LBE WWTP in two phases (two years).
Air scour control system
The air scour control system, registered as Smart Air MBR, modifies the membrane air scour flow rate based on the permeability evolution, which is used as an indicator of the membrane performance and the sludge characteristics (Comas et al., 2010). The TMP and the permeate flow rate are monitored in real-time. These variables are first processed for data quality assurance and then cycle and daily averaged values are calculated. Long term (LT) and short term (ST) permeability trends are evaluated and compared to identify energy saving opportunities and alarms. The controlled variable is the slope ratio (SR), which is calculated as the ratio between the ST and the LT permeabilities. The ST was determined to be 4 days, and the LT was determined to be 14 days (Ferrero et al., 2011b). The SR is proportional to the control action, which in this case is the daily regulation of the air scour flow rate by increasing or decreasing the blower frequency (the manipulated variable) (Ferrero et al., 2011b). Normally, the permeability decreases in proportion to the cake layer resistance, which is roughly proportional to the volume of sludge filtered (Judd, 2011). However, corrective actions such as chemical cleaning or a decrease in the permeate flux can cause the permeability to increase. Operational conditions (the temperature, the suspended solids concentration, the sludge age, etc.) can also affect the sludge permeability. Thus, four different scenarios (i.e., relationships between ST and LT) have been defined for the air scour control system (see Fig. 1).
Phase 1: Data collection and calibration of the air scour control system (12 months from April 2011). During the first 8 months different implementation tasks were performed (monitoring energy consumption of blowers, frequency driver installation, monitoring filtration variables and adaptation of the control system). The air scour control system was implemented in filtration line A but operated in an open-loop regime for the last 4 months (Phase 1 Table 1) under the conditions recommended by the membrane suppliers, thus same operation conditions for filtration lines A and B. Data on the TMP, the permeate flux and the working hours per day of both process lines were collected in real time, processed and compared. The information acquired over this period was used to calibrate the control system parameters. Phase 2: Validation (12 months, from April 2012). The air scour control system was operated in a closed-loop regime for filtration line A, whereas line B was used as a reference with a constant air scour flow rate of 17.8 m3 min1 (as recommended by the membrane manufacturers). Fig. 2 is a schematic of the membrane tank that shows the two filtration lines and the equipment for the control system. In Phase 2, the maximum flow rate reduction allowed for the air scour control system in line A was increased progressively from 5% to 10%, 15% and 20% of the air scour flow rate.
Case 1: The long term permeability slope is negative (LT < 0), and the short term permeability slope is positive or equal to zero (ST 0), corresponding to a moderate energy saving opportunity. Case 2: The long term permeability slope is negative (LT < 0), and the short term permeability slope is negative (ST < 0, lower or higher than LT), corresponding to a low or a non-energy saving opportunity. Case 3: The long term permeability slope is positive or equal to zero (LT 0), and the short term permeability is positive or equal to zero (ST 0, lower or higher than LT) corresponding to a moderate or high energy saving opportunity.
Case 1
K
Experimental plan
Case 2
K
Chemical oxygen demand (COD), biochemical oxygen demand (BOD), nitrogen and phosphates species, mixed liquor suspended solids (MLSS) and sludge volumetric index (SVI) were analyzed in accordance with Standard Methods (APHA, 2005). The capillarity suction time (CST) (Triton electronics Ltd., type 304 B) (Scholz, 2005) was used to determine the sludge filterability. The permeability decay was used as an indicator of the general fouling evolution, and the fouling rate (which was measured as the slope of the permeability for each permeate
Case 3
K ST
LT
LT
ST time
Experimental monitoring
Case 4
K
ST
ST LT
2.4.
time
ST LT
time
Fig. 1 e Different relationships between ST and LT permeability.
ST
time
4
w a t e r r e s e a r c h 7 9 ( 2 0 1 5 ) 1 e9
TMP-A Flux-A
7
7
TMP-B Flux-B
9 10
Control Server Data acquisition and processing
1 3
2
1 Membrane filtration line A 2 Membrane filtration line B 3 Feed to membrane tank 4 Variable speed drive
Line-B 3
Line-A
5 Line A Blower (controlled by Smart Air MBR) 6 Line B blower 7 Sensors and probes
Set point of blower r.p.m.
VSD
5
8
6
4
8 Recirculated stream to oxidation ditch 9 Permeate pipes 10 Server
Fig. 2 e Schematic diagram of the control system of the LBE WWTP.
cycle, as in Monclu´s et al., 2011), was used to estimate the fouling accumulated in each filtration cycle to ensure that there were no differences between the filtration line that was regulated by the air scour control system (line A) and the reference line B. Differences on FRs among both lines were determined by Student's t test for paired samples, where a pvalue
