Abstract
Increasing stringency of environmental discharge standards has triggered an industry-wide inclination towards membrane bioreactors over conventional activated sludge processes to ensure fulfilment of environmental discharge criteria. Yet, despite its plentiful advantages, high aeration costs remain as a key deterrent to the widespread adoption of the MBR technology. This backdrop created an impetus for a wastewater treatment company to develop an efficient MBR air scouring protocol that can be realized in existing plants without retrofitting. Known as pulsed cyclic aeration, plant trial applications have demonstrated that fouling control and aeration savings can be improved by >30%, resulting in scouring energy consumptions that can be as low as 0.049 kWh/m3.
INTRODUCTION
As the environmental issues of mainland China have global repercussions, the Chinese authorities have undertaken great efforts to ensure that wastewater discharge standards are in pace with the water environment degradation brought about by decades of rapid urbanization and industrialization (Li et al. 2012). Specifically, these circumstances led to the rising preference of membrane bioreactors (MBRs) over conventional activated sludge processes (ASPs) to guarantee superior effluent qualities that meet environmental discharge criteria. In detail, MBRs refer to an advanced wastewater treatment process where a membrane filtration process is integrated with the ASP technology, bringing about the following advantages over ASPs (Stephenson et al. 2000; Metcalf & Eddy 2003; Judd 2010):
- (i)
Better effluent quality
- (ii)
Smaller footprint
Despite the aforementioned advantages, the principal drawback of the MBR technology is its elevated levels of energy consumption, where process aeration can make up to 60% of the total energy consumption, with membrane aeration contributing 60% of total aeration demands (Yufen et al. 2010). Fortunately, due to optimizations of the MBR technology over the last 50 years, overall energy demands have reduced from about 5.0 kWh/m3, which was needed for the first side-stream MBRs, to about 1.0 kWh/m3 by 2005, and more recently, to approximately 0.4–0.6 kWh/m3 beyond year 2010. Despite the improvements, conventional ASP plants still cost less energy to operate than MBRs, at 0.3–0.4 kWh/m3 (Yufen et al. 2010; Krzeminski et al. 2012). With membrane scouring contributing a huge proportion of the total energy demands, exceptional opportunities to enhance MBR feasibilities lie within approaches that seek to reduce air scouring demands.
These considerations as elaborated have shaped the context that influenced Yangli Sewage Treatment Plant (YSTP), a 200,000 m3/d large-scale sewage STP in Fuzhou (China), to reduce the aeration needs of its MBR system. Commissioned by CITIC Envirotech Limited (CEL) in 2015, the MBR system was designed using a proven aeration configuration that is employed in all the wastewater treatment plants owned by the conglomerate. Specifically, this paper aims to discuss the successes of a plant trial in YSTP elucidating the effectiveness of an innovative aeration protocol to control membrane fouling at several operational (membrane) fluxes.
METHODS
YSTP treatment train design
Figure 1 below illustrates the treatment train design of the YSTP where the use of A2O process have been employed to allow for biological nutrient removal (BNR) and, combined with membrane filtration to enhance the overall effluent quality at the outfall. Details of the MBR system is as detailed in Table 1.
Summary of MBR system parameters
Parameters . | Units . | Value . | Comments . |
---|---|---|---|
Membrane brand | – | Memstar | |
Membrane type | PVDF (3G-TIPS) | 3rd generation thermally-induced phase separated PVDF membrane | |
Pore size | μm | <0.1 | |
Tensile strength | MPa | 8 | Stronger than the predecessor NIPS tensile strength of 2 MPa |
Membrane area | m2 | 614,400 | |
Membrane flux | LMH | 15–30 | Temperature corrected to 25 °C |
MLSS | mg/L | 4,000–6,000 | |
Years of operation | years | 3 | Since 2015. No membrane replacements yet. |
Parameters . | Units . | Value . | Comments . |
---|---|---|---|
Membrane brand | – | Memstar | |
Membrane type | PVDF (3G-TIPS) | 3rd generation thermally-induced phase separated PVDF membrane | |
Pore size | μm | <0.1 | |
Tensile strength | MPa | 8 | Stronger than the predecessor NIPS tensile strength of 2 MPa |
Membrane area | m2 | 614,400 | |
Membrane flux | LMH | 15–30 | Temperature corrected to 25 °C |
MLSS | mg/L | 4,000–6,000 | |
Years of operation | years | 3 | Since 2015. No membrane replacements yet. |
MBR system parameters
Aeration protocol
The conventional aeration configuration used by CEL is the X/X strong-weak cyclic aeration, which involves switching the magnitude of the airflow between high and low levels at fixed time intervals of X seconds. As it has been well documented that air scouring controls membrane fouling by inducing flow fluctuations and local tangential shear turbulences (Judd 2010; Braak et al. 2011), it was hypothesized that if the original continuous scouring can be made intermittent by eliminating either the high or low airflow component, greater shear forces can be generated to enhance scouring intensities, thus improving fouling control. This alternative aeration protocol is named X/X cyclic pulsed aeration, where high airflow is introduced for X seconds and aeration stops for the next X seconds.
In the case of this MBR-based STP in Fuzhou, X is 24 seconds and a major advantage of this cyclic pulsed protocol is that it can be implemented with great ease in all existing treatment plants of CEL through valve manipulation along aeration pipelines (that are of standardized design). Table 2 below summarizes the critical differences between strong-weak cyclic aeration and cyclic pulsed aeration.
Differences between strong-weak cyclic aeration and cyclic pulsed aeration
Aeration protocol . | Airflow per module (Nm3/h.module) . |
---|---|
24/24 strong-weak cyclic aeration | Strong flow: 6.25 |
Weak flow: 2.5 | |
Average @ 4.375 | |
24/24 cyclic pulsed aeration | Strong flow: 5.47 |
Weak flow: 0 | |
Average @ 2.73 |
Aeration protocol . | Airflow per module (Nm3/h.module) . |
---|---|
24/24 strong-weak cyclic aeration | Strong flow: 6.25 |
Weak flow: 2.5 | |
Average @ 4.375 | |
24/24 cyclic pulsed aeration | Strong flow: 5.47 |
Weak flow: 0 | |
Average @ 2.73 |
Experimental plan
The plant trial aims to elucidate the effectiveness of an alternative scouring protocol by first establishing a baseline response using the current 24/24 strong-weak cyclic aeration at various membrane fluxes and then compare results obtained with the new 24/24 cyclic pulsed aeration. Thus, the plant trial is a two-phased approach with details as summarized in Table 3.
Experimental conditions at different phases
Phase . | Aeration configuration . | Operational flux (LMH) . |
---|---|---|
1 | 24/24 strong-weak cyclic aeration | 15, 25, 30 |
2 | 24/24 cyclic pulsed aeration | 15, 25 |
Phase . | Aeration configuration . | Operational flux (LMH) . |
---|---|---|
1 | 24/24 strong-weak cyclic aeration | 15, 25, 30 |
2 | 24/24 cyclic pulsed aeration | 15, 25 |
RESULTS AND DISCUSSION
The results of Phase 1 are illustrated in Figure 2 and Table 4. Evidently, the higher the membrane flux, the higher the initial TMP and the sharper the TMP gradient (averaged daily TMP increments). When the flux was increased by 20% from 25LMH to 30LMH, fouling rates increased by >120% (from 0.0722 kPa/d to 0.1645 kPa/d), indicating that 30LMH is likely the maximum sustainable flux for the current set of operational conditions (under strong-weak cyclic aeration). Additionally, while SADp has improved by 25% (from 4.375 Nm3/m3 at 25LMH to 3.285 Nm3/m3 at 30LMH), the significant jump in TMP increment rates suggested an undesirable compromise of long-term sustainable operations.
Fouling rates of baseline scouring protocol at different fluxes
Flux . | Nm3/h.module . | SADm (Nm3/h.m2) . | SADp (Nm3/m3) . | TMP gradient (kPa/d) . |
---|---|---|---|---|
30 LMH | Strong flow: 6.25 | 0.219 | 3.285 | 0.1645 |
25 LMH | Weak flow: 2.5 | 4.375 | 0.0722 | |
15 LMH | Average @ 4.375 | 7 | 0.0628 |
Flux . | Nm3/h.module . | SADm (Nm3/h.m2) . | SADp (Nm3/m3) . | TMP gradient (kPa/d) . |
---|---|---|---|---|
30 LMH | Strong flow: 6.25 | 0.219 | 3.285 | 0.1645 |
25 LMH | Weak flow: 2.5 | 4.375 | 0.0722 | |
15 LMH | Average @ 4.375 | 7 | 0.0628 |
Plot of TMP data against time for various membrane fluxes using strong-weak cyclic aeration.
Plot of TMP data against time for various membrane fluxes using strong-weak cyclic aeration.
As Phase 1 data for 30LMH conditions revealed incompatibility with stable long-term operations, Phase 2 focused on quantifying the effectiveness of cyclic pulsed aeration on fouling control for 15 and 25LMH. With reference to Figure 3 and Table 5, it is evident that when aeration demands have fallen by >37% when cyclic pulsed aeration is used (from 4.375 to 2.73 Nm3/h.module), fouling rates remained relatively unchanged for 15LMH (0.0628 kPa/d to 0.0698 kPa/d) and decreased by >30% for 25LMH, a possible indication that the effect of fouling control by cyclic pulsed aeration is more pronounced when fouling propensities are higher (at higher membrane fluxes).
Fouling rates of cyclic pulsed protocol at 15 & 25LMH
Aeration protocol . | Flux . | Nm3/h.module . | SADm (Nm3/h.m2) . | SADp (Nm3/m3) . | TMP gradient (kPa/d) . |
---|---|---|---|---|---|
Strong-weak cyclic aeration | 15 | Average @ 4.375 | 0.219 | 7 | 0.0628 |
25 | 4.38 | 0.1278 | |||
Cyclic pulsed aeration | 15 | Average @ 2.73 | 0.1365 | 4.37 | 0.0698 |
25 | 2.73 | 0.0872 |
Aeration protocol . | Flux . | Nm3/h.module . | SADm (Nm3/h.m2) . | SADp (Nm3/m3) . | TMP gradient (kPa/d) . |
---|---|---|---|---|---|
Strong-weak cyclic aeration | 15 | Average @ 4.375 | 0.219 | 7 | 0.0628 |
25 | 4.38 | 0.1278 | |||
Cyclic pulsed aeration | 15 | Average @ 2.73 | 0.1365 | 4.37 | 0.0698 |
25 | 2.73 | 0.0872 |
Plot of TMP data against time for various membrane fluxes using cyclic pulsed aeration.
Plot of TMP data against time for various membrane fluxes using cyclic pulsed aeration.
The successful application of the cyclic pulsed aeration allowed for enhanced fouling control that made higher fluxes of 25LMH sustainable and consequently, brought about a desirable reduction in the electric energy consumption per ton of permeate produced (Table 6). And when compared with aeration consumption of several other STPs owned by CEL, cyclic pulsed aeration has proven to be a competitive protocol, especially at higher fluxes.
Unit electric energy consumption (for membrane aeration) between different MBR plants
. | CITIC Envirotech Assets . | |||||||
---|---|---|---|---|---|---|---|---|
Project . | Yangli STP . | Chengdu STP . | Huaifang STP . | Jingxi STP . | Jiaxing STP . | |||
Location . | Fujian . | Sichuan . | Beijing . | Guangzhou . | Zhejiang . | |||
Aeration protocol | Strong-weak cyclic aeration | Cyclic pulsed aeration | Strong-weak cyclic aeration | Cyclic pulsed aeration | ||||
Flux (LMH)a | 15 | 25 | 15 | 25 | 20 | 15 | 18 | 19 |
Membrane aeration's electric energy consumption (kWh/m3) | 0.131 | 0.081 | 0.079 | 0.049 | 0.069 | 0.082 | 0.098 | 0.054 |
. | CITIC Envirotech Assets . | |||||||
---|---|---|---|---|---|---|---|---|
Project . | Yangli STP . | Chengdu STP . | Huaifang STP . | Jingxi STP . | Jiaxing STP . | |||
Location . | Fujian . | Sichuan . | Beijing . | Guangzhou . | Zhejiang . | |||
Aeration protocol | Strong-weak cyclic aeration | Cyclic pulsed aeration | Strong-weak cyclic aeration | Cyclic pulsed aeration | ||||
Flux (LMH)a | 15 | 25 | 15 | 25 | 20 | 15 | 18 | 19 |
Membrane aeration's electric energy consumption (kWh/m3) | 0.131 | 0.081 | 0.079 | 0.049 | 0.069 | 0.082 | 0.098 | 0.054 |
aTemperature corrected to 25 °C.
CONCLUSIONS
An unique air scouring protocol called pulsed cyclic aeration has been successfully tested in YSTP, a large-scale MBR system in China. Through analysis, it was discovered that aeration demands and fouling control can be both improved by >30% due to the likely creation of stronger shear forces at membrane surfaces than the conventional strong-weak cyclic aeration used throughout CEL's water assets. Interestingly, the effectiveness of the cyclic pulsed aeration protocol seemed to be more pronounced at higher membrane fluxes, where fouling rates remained relatively unchanged for 15LMH (0.0628 kPa/d to 0.0698 kPa/d) and decreased by >30% for 25LMH. More importantly, pulsed cyclic aeration can be easily realized in existing plants via manipulation of valves along the aeration lines which are of standardized design without any retrofitting.