Abstract
Membrane filtration technology is widely applied in conventional surface water treatment plants because of its low power load and high filtration accuracy. This study follows the transition from a power-driven to a gravity-driven siphon-submerged ultrafiltration membrane system in a drinking water treatment plant in Shandong Province, China. The proportion of space membrane areas in the tank was further developed, by increasing the membrane areas of a single tank to 1.6 times that before the transformation, and the design production capacity of a single tank is increased from 10,000 m3/d before the transformation to 16,700 m3/d. The recovery rate of a single tank was also increased from 96.28 to 97.77%. The energy consumption per ton of water decreased from 0.06 to 0.017 MJ. The annual carbon emissions reduction reached 197.13 tons per tank. In terms of the water quality, the algae removal rate from surface water by the ultrafiltration membrane reached 100%. Further, the removal rates of 2 μm particles, turbidity, and chlorophyll were above 99%. Our results suggest that gravity-driven siphon-submerged ultrafiltration promotes environmental protection, energy conservation, and emissions reduction. This technology introduces a new development direction for membrane filtering in drinking water treatment plants.
HIGHLIGHTS
A power-driven water treatment system was upgraded to be gravity-driven.
The gravity siphon membrane modules required excellent hydrophilicity properties.
The flow and transmembrane pressure were linearly related to the water level.
Energy consumption and CO2 emissions decreased with the gravity-driven system.
INTRODUCTION
The current situation regarding drinking water safety is concerning. Problems such as water shortages, pollution risks, and quality changes are becoming increasingly serious. People are growing more concerned than before about the safety and quality of drinking water, which pushes drinking water quality indicators to become increasingly stricter. Because of its efficient filtration accuracy and stable effluent quality, membrane technology has attracted extensive attention in the field of drinking water purification (Cevallos-Mendoza et al. 2022).
The microfiltration or ultrafiltration membrane used in municipal drinking water in China is a mature technology that has been applied on a large scale (Li & Chen 2018; Yuan et al. 2020; Shen et al. 2021). For most of these applications, power-driven external pressure column and submerged ultrafiltration processes were adopted. Both membrane treatment processes are designed to separate and purify contaminants in the water using pump pressurization or suction. Figure S1 shows the 300,000 m3/d pressure ultrafiltration process in the Ling Zhuang zi water plant in Tianjin, China. Figure S2 shows the 30,000 m3/d submerged ultrafiltration process in the Dingwu water plant in Laoling, Shandong Province, China.
In recent years, some researchers have studied the application of gravity-driven siphons to the membrane filtration process (Patterson & Pease 2005; Janson et al. 2006; Crossley et al. 2007; Song et al. 2020). Gravity-driven siphon water production is achieved through a height difference between the water levels of the membrane tank and the clean water tank, following the Bernoulli equation. Compared with the traditional power-driven water production process, the gravity-driven siphon has a lower energy consumption and simpler operation and maintenance (Capitaine 1984; Shorney-Darby et al. 2007; Tang et al. 2018). Once the technology is developed further, it will be well worth promoting. In Singapore, Water Treatment and Process Group of General Electric Company carried out the transformation of a gravity-driven siphon-submerged membrane system for a 182,000 m3/d-Phase I V-shaped filter in the Choa Chu Kang Water Plant in 2008. The membrane system occupied only four of its nine filters, and the three filters are transformed into supporting equipment rooms, reagent rooms, backwashing, cleaning water tanks, etc. The other two filters are reserved spaces. The system carried out the maximum transmembrane pressure (TMP) difference of the membrane through a 4.65 m water level difference between the membrane and clean water tanks. The energy consumption of the entire membrane system after the transformation was only 0.0178 MJ/m3 (Luo 2008). Gravity-driven siphons have also appeared in China in recent years. In 2020, Tianjin Motimo Membrane Technology Co., Ltd adopted this technology for the reconstruction of a 125,000 m3/d water plant in Tangshan, Hebei Province. In this project, the Phase I sand filters of the water plant were transformed, and the energy consumption per ton of water was reduced from 0.045 to 0.0185 MJ/m3 (Hu et al. 2021). Also in 2020, the Ningbo Taoyuan Water Plant, the largest submerged waterworks in China with a capacity of 500,000 m3/d, adopted the gravity-driven submerged membrane process located in a mountainous area. However, the above-mentioned application cases and published articles rarely mention the design details of the gravity siphon process. Taking the gravity siphon membrane process transformation project of a water supply plant in Shandong Province as an example, the study analyzes the progressiveness of the gravity siphon-driven immersion membrane technology from three aspects: (1) component selection to process design; (2) changes in water volume, energy, consumption and recovery rate before and after system transformation; (3) stability of membrane system operation and water quality of membrane inlet and outlet. It provides a new reference for the development of the membrane water treatment process in surface water plants.
METHODS
Project overview
implement energy-saving transformations on 6 of the 12 existing membrane tanks;
install Polyvinylidene fluoride (PVDF) ultrafiltration membranes with high filling density, high flux, and low transmembrane pressure difference;
transform and replace part of the process equipment, pipes, and electrical and automatic control; and
achieve the water production and energy conservation goals for the gravity-driven siphon water system.
Efficiency optimization of the process design
The power-driven submerged ultrafiltration system was transformed into a gravity-driven siphon ultrafiltration system. In addition to the approved upgrade of existing equipment parameters and the dismantling of the pipeline, the four key elements were as follows:
- (1)
selecting ultrafiltration components with better membrane performance;
- (2)
selecting an appropriate membrane flux;
- (3)
ensuring a water level difference to act as a driving force;
- (4)
designing a stable, reliable, and flexible process operation sequence.
Membrane performance requirements
The material, filtration accuracy, pure water flux, and hydrophilic contact angle must be improved when selecting membrane modules. Further details are provided in Table 1.
Project . | Before replacement . | After replacement . | Note . |
---|---|---|---|
Membrane component material | PVDF | PVDF | High chemical stability and pollution-resistant materials |
Average pore size (μm) | 0.03 | ≤0.03 | High filtration precision ensures the safety of water quality |
Pure water flux (L/m2·h/bar) | ≥800 | ≥1,200 | High pure water flux ensures a stable production capacity |
Hydrophilic contact angle (°) | ≤65 | ≤35 | The hydrophilicity is related to the membrane flux properties |
Project . | Before replacement . | After replacement . | Note . |
---|---|---|---|
Membrane component material | PVDF | PVDF | High chemical stability and pollution-resistant materials |
Average pore size (μm) | 0.03 | ≤0.03 | High filtration precision ensures the safety of water quality |
Pure water flux (L/m2·h/bar) | ≥800 | ≥1,200 | High pure water flux ensures a stable production capacity |
Hydrophilic contact angle (°) | ≤65 | ≤35 | The hydrophilicity is related to the membrane flux properties |
Selection of the membrane flux
To guarantee membrane performance, the selection of the membrane flux is key. If the membrane flux selected is too high, the membrane operation load will be too heavy, contamination will increase rapidly, the cleaning cycle will be frequent, and operation costs will rise. Selecting a small membrane flux will lead to increased investment costs and space occupation. We investigated the membrane flux of several submerged membrane process water plants that treat similar water sources. See Table S1 for details, which is used as the basis for the selection of membrane flux for this transformation project. Considering that the membrane area space can be expanded according to the existing membrane tank, the final instantaneous flux was selected to be 24.33 LMH (L/m2·h) (specific data are shown in Table S1).
Design of the membrane tank and water level
Because there was no extra space in this project, the renovation was based on an existing membrane tank. Table 2 compares the main parameters before and after the renovation.
Project . | Original membrane tank design . | Film tank transformation design . |
---|---|---|
Membrane tank size L * W (m) | 10.6 × 5.5 | 10.6 × 5.5 |
Number of membrane frames (table) | 16 | 22 |
Number of membrane components (branch) | 512 | 572 |
Single-branch assembly membrane area (m2) | 35 | 50 |
Total membrane area (m2) | 17,920 | 28,600 |
Design flux (L/m2·h) | 23.25 | 24.33 |
Total capacity (m3/d) | 10,000 | 16,700 |
Water production supervisor pipe diameter (mm) | DN400 | DN500 |
Aeration pipe diameter (mm) | DN200 | DN250 |
Project . | Original membrane tank design . | Film tank transformation design . |
---|---|---|
Membrane tank size L * W (m) | 10.6 × 5.5 | 10.6 × 5.5 |
Number of membrane frames (table) | 16 | 22 |
Number of membrane components (branch) | 512 | 572 |
Single-branch assembly membrane area (m2) | 35 | 50 |
Total membrane area (m2) | 17,920 | 28,600 |
Design flux (L/m2·h) | 23.25 | 24.33 |
Total capacity (m3/d) | 10,000 | 16,700 |
Water production supervisor pipe diameter (mm) | DN400 | DN500 |
Aeration pipe diameter (mm) | DN200 | DN250 |
Notes:L, length; W, Width.
The modified plan and section of the membrane tank are shown in Figures S3 and S4, while the submerged ultrafiltration module for the gravity-driven siphon is shown in Figures S5 and S6. The capacity of the membrane tank was increased by 10,680 m2 with the same floor area owing to the vertical space in the tank being fully utilized. Simultaneously, a membrane assembly with better performance was used in the membrane tank, which improved the average membrane flux. The water production capacity of a single tank was expected to increase 1.67 times.
Water level difference design
We calculated the inlet channel water level, existing membrane tank operating level, highest clean water tank level, and possible head loss in the water treatment process. The difference between the available membrane and clean water tank levels was 2.9 m. After subtracting the various pipe and head losses, the water level difference obtained by the submerged membrane system was 2.617 m. The driving force of this difference could reach up to 26.17 kPa. To increase the driving force of the water level difference, the height of the adjustable weir plate of the inlet channel was raised.
Process operation step sequence
Serial No . | Equipment/valve . | Function . |
---|---|---|
1 | Inlet channel | Water storage to quickly replenish water for membrane tank |
2 | Inlet valve | Control the level of membrane tank |
3 | Membrane tank | Place for membrane production, backwashing, and chemical cleaning |
4 | Membrane modules | Used to filter impurities and pollutants in water |
5 | Vacuum generator | It is used to extract the air in the pipeline to avoid breaking the vacuum environment |
6 | Siphon production valve | Control the flow of gravity-driven siphon membrane production |
7 | Filtrate valve/suction pump | When siphon cannot meet the membrane water production, the water production can be pumped |
8 | Aeration valve/roots blower | Control the membrane air washing |
9 | Backwash valve/backwash system | Control the membrane backwashing |
10 | Chemical dosing valve | Dosing chemicals during EFM |
11 | Drain valve | Control the discharge of membrane tank |
12 | Circulating valve1/2/3 and circulating pump | Drive the chemicals to circulate inside and outside the membrane to enhance the cleaning effect |
13 | Vent valve | Control the discharge of cleaning waste liquid |
Serial No . | Equipment/valve . | Function . |
---|---|---|
1 | Inlet channel | Water storage to quickly replenish water for membrane tank |
2 | Inlet valve | Control the level of membrane tank |
3 | Membrane tank | Place for membrane production, backwashing, and chemical cleaning |
4 | Membrane modules | Used to filter impurities and pollutants in water |
5 | Vacuum generator | It is used to extract the air in the pipeline to avoid breaking the vacuum environment |
6 | Siphon production valve | Control the flow of gravity-driven siphon membrane production |
7 | Filtrate valve/suction pump | When siphon cannot meet the membrane water production, the water production can be pumped |
8 | Aeration valve/roots blower | Control the membrane air washing |
9 | Backwash valve/backwash system | Control the membrane backwashing |
10 | Chemical dosing valve | Dosing chemicals during EFM |
11 | Drain valve | Control the discharge of membrane tank |
12 | Circulating valve1/2/3 and circulating pump | Drive the chemicals to circulate inside and outside the membrane to enhance the cleaning effect |
13 | Vent valve | Control the discharge of cleaning waste liquid |
To avoid membrane fouling – the main problem in the application of ultrafiltration membranes – an effective cleaning method to reduce membrane contamination was key (Qiu & Zhang 2002; Liu et al. 2014; Huang et al. 2021). This project refers to the operating experience of the submerged membrane system before the transformation and realizes the settable and revisable functions on some key parameters. In this way, when the water quality changes or the physical/chemical cleaning effect is poor, the parameters can be adjusted to delay membrane pollution.
The gravity-driven siphon-submerged ultrafiltration process included the following stages:
Water replenishment
Raw water went through the inlet channel into the membrane tank after opening the inlet valve. To improve the water production efficiency, the time it took for water to reach the set operating level was constrained at 60 s.
Water production
The siphon production valve was opened after the set level was reached, there was feedback between the valve and the set water-flow rate. The system will automatically adjust the siphon production valve opening through the proportional integral differential according to the set flow and start to produce water. During this stage, the vacuum generator would also start and stop automatically according to the flow switch supervisor of the generator. This prevented the water production pipeline from affecting the gravity-driven siphon with air bubbles.
Physical backwash
The flux of the ultrafiltration membrane gradually decays as the membrane pores get plugged with organic matter, inorganic chemicals, and microbial pollutants (Zheng et al. 2011). Physical backwashing slowed down the formation of fouling and filter cake layers and eliminated pollutants accumulated on the membrane surface in the short-term through aeration, air washing, and drainage.
- (1)
Aeration. After the water level of the membrane tank was lowered to the backwash level, the water production control valve was closed, triggering the air washing procedure. A blower started working and the aeration valve of the membrane tank was opened. Air rose gradually from the bottom of the membrane frame along the membrane assembly. The large air bubbles removed the suspended particles accumulated on the surface of the filaments.
- (2)
Air washing. The backwash pump and backwash valve were opened at the same time. Backwashing water is generally produced by a membrane, and the backwashing flow is 1.5 times that of the produced water flow. In this step, the pollutants on the membrane surface can be further peeled off through the synergetic effect of air and water.
- (3)
Drainage. The drain valve of the membrane tank was opened, and the backwash sewage was discharged by gravity. The purpose was to remove the contaminants that had fallen into the water from the membrane tank. Simultaneously, the rapid drop in the water level during the discharge process could also scour the membrane surface.
The conventional water production and backwashing process steps are shown in Figure S7.
Chemical cleaning
Conventional physical backwashing cannot remove all the contaminants from the membrane surface (e.g., microbial pollution, organic pollution, and inorganic salt scales). Therefore, to guarantee production capacity, the membrane operating pressure difference exhibits a slow growth trend. Because of the accumulation of membrane contamination, this trend rises sharply in later stages, which has a substantial impact on the continuous production capacity. To ensure the recovery of the membrane performance and achieve long-term stable operation, chemical agents aimed at eliminating pollution must be added. Two units of chemical cleaning were used in this project: enhanced flux maintenance (EFM) and cleaning-in-place (CIP).
- (1)
EFM
This type of maintenance cleaning involved the use of chemical cleaning equipment to add low concentrations of chemicals. It could effectively reduce the contamination by organic and inorganic colloids on the membrane surface after a short cycle and submergence. After the flux was restored, the stable operation of the system was maintained for 3–7 days.
- (2)
CIP
This restorative cleaning is the use of a chemical cleaning system to add high concentrations of chemical agents. Through multiple cycles and prolonged submerging, chemical reactions to separate and dissolve the compound contaminants accumulated on the membrane surface.
The EFM and CIP process steps of maintenance cleaning and chemical online cleaning are shown in Fig S8 and Fig S9. Cleaning agents are commonly used for different pollutants. The selection of cleaning agents for different pollutants is shown in Table S2.
Theoretical basis of the gravity-driven siphon operation
Considering that Z2 > 0, u2 > 0, and hf > 0, we can obtain P2 < 0 for section 2–2 to generate negative pressure and a siphon effect.
RESULTS AND DISCUSSION
Flow and pressure changes with different water levels
Stability of membrane system operation
SF0 is the membrane specific flux before temperature correction, L/(m2 · h · kPa); J is the membrane flux, L/(m2 · h); J = Q/A, Q: membrane system flow m3/h; A: Membrane area, m2; △P is the measured transmembrane differential pressure, kPa.
SF is the membrane specific flux corrected to 15 °C, L/(m2 · h · kPa); T is the water temperature, °C.
It can be seen from Figure 6 that the membrane SF is relatively stable at the initial stage of operation, showing a slow decline, and each physical backwash can play a certain recovery effect. When operating for more than 200 days, the SF drops rapidly. It can be seen from the reference flux and TMP curve that the membrane flux maintains a low operating state during this period, which may be due to the relatively small water demand during this period. At the same time, the TMP has obviously increased in 200–290 days, indicating that the membrane has been deeply polluted, and it is difficult to recover with conventional physical cleaning. So the water plant carried out the first chemical cleaning after the system was put into operation, using 3,000 ppm sodium hypochlorite and 5,000 ppm citric acid for chemical cleaning. After cleaning, the recovery effect of TMP and SF is ideal, and the subsequent operation parameters are relatively stable.
Removal effect of the main membrane filtration pollutants
In addition, the inlet and outlet water quality of the membrane system is also an important indicator for stable operation of the membrane system. The quality of influent water directly affects the speed of membrane pollution, while the membrane itself has a high filtering accuracy, and the quality of produced water is generally relatively stable. If the quality of produced water from the membrane fluctuates greatly or the quality index of produced water is generally poor, the high probability indicates that the integrity of the membrane module is damaged and the operation of the membrane system is defective.
Economic analysis
The original pressure-driven submerged membrane system was transformed into a gravity-driven siphon one. There were clear, observable effects of energy conservation and emissions reduction (Young & Perri 2015). The energy consumption per ton of water, chemical consumption, and recovery rate were improved to varying degrees. Table 4 shows the advantages of the gravity-driven submerged ultrafiltration system.
Project . | Before the transformation . | After the transformation . | Remarks . |
---|---|---|---|
Daily output (t/d) | 10,000 | 16,700 | – |
Energy consumption (MJa/d) | 2,160 | 1,022 | – |
CO2 emission due to power consumptionb (tons/tank·year) | CO2 emission reduction can reach 120.09–56.82 = 63.27 | ||
Energy consumption per ton of water (MJ/t) | 0.216 | 0.061 | Energy consumption can reach 0.216–0.061 = 0.155 |
Chemical consumptionc (Yuan/t) | NaClO: 0.005 Citric acid: 0.033 Total: 0.038 | NaClO: 0.003 Citric acid: 0.02 Total: 0.023 | NaClO price: 800 RMB/Td Citric acid price: 15,000 RMB/T |
Recovery rate (%) | 96.28 | 97.77 | e |
Project . | Before the transformation . | After the transformation . | Remarks . |
---|---|---|---|
Daily output (t/d) | 10,000 | 16,700 | – |
Energy consumption (MJa/d) | 2,160 | 1,022 | – |
CO2 emission due to power consumptionb (tons/tank·year) | CO2 emission reduction can reach 120.09–56.82 = 63.27 | ||
Energy consumption per ton of water (MJ/t) | 0.216 | 0.061 | Energy consumption can reach 0.216–0.061 = 0.155 |
Chemical consumptionc (Yuan/t) | NaClO: 0.005 Citric acid: 0.033 Total: 0.038 | NaClO: 0.003 Citric acid: 0.02 Total: 0.023 | NaClO price: 800 RMB/Td Citric acid price: 15,000 RMB/T |
Recovery rate (%) | 96.28 | 97.77 | e |
aMJ, Million joules, 1 MJ = 0.28 kWh.
bAccording to the Statistical Bulletin of the People's Republic of China on National Economic and Social Development in 2020 issued by the National Bureau of Statistics, thermal power generation in China will account for 68.51% in 2020; According to the Statistical Data of National Electric Power Industry in 2020 issued by the National Energy Administration, the standard coal consumption per kilowatt hour of electricity under the premise of dynamic power generation is 305.5 g; According to the Construction Scheme of National Carbon Emission Trading Market (Power Generation Enterprises), the carbon dioxide emission per ton of standard coal shall be 2.6 tons; The comprehensive carbon dioxide emission per kilowatt hour is 0.6851 × 0.3055 × 2.6 = 0.544 kg.
cThe difference in the consumption of chemicals per ton of water in membrane tanks before and after transformation is mainly caused by the difference in the production flow.
dRMB/T means Renminbi (RMB) per ton.
eF0, Inlet flow; F1, Production flow; F2, Backwash flow.
CONCLUSIONS
- (1)
The membrane modules used in the gravity-driven siphon-submerged ultrafiltration system required excellent hydrophilicity and antipollution properties.
- (2)
The upper filtration pressure limit of this system depended on the absolute height difference between the membrane and clean water tanks. The maximum membrane filtration pressure difference of this project was 26.17 kPa.
- (3)
The flow velocity and TMP of the gravity-driven siphon system had a linear relationship with the water level.
- (4)
The algae removal rate of the siphon membrane system was 100%. The removal rates of 2 μm particles, turbidity, and chlorophyll were all >99%. The average TOC removal rate was 18.12%. The average permanganate index removal rate was 20.43%.
- (5)
Compared with the traditional power-driven submerged membrane system, the energy consumption per ton of water of the gravity-driven system was reduced by 0.155 MJ. The annual CO2 emissions reduction was of 63.27 tons per tank. The chemical cost was reduced by 0.015 yuan per ton of water and the recovery rate increased from 96.28 to 97.77%.
ACKNOWLEDGEMENTS
This study was supported by the National Key R&D Program of China (grant number: 2021YFC3201304).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.