ABSTRACT
The study analyses the performance of a pilot plant using a rotating hollow fibre (HF) membrane bioreactor system. The experiments evaluated the effect of operational parameters such as rotational speed, aeration strategies, and maintenance cleaning (MC) procedures on the efficiency of the system, in particular transmembrane pressure (TMP) and filtrate quality. The results indicate that the rotating membrane module reduces TMP increase and can operate for 48 days with satisfactory performance, even without aeration. This has the potential to significantly improve efficiency, resulting in significant energy savings. In addition, two MC methods, clean in air and clean in place, were tested and found to be efficient for weekly MC. It was observed that operating without aeration during colder seasons may not be effective. Therefore, adaptive strategies are needed to address seasonal temperature variations.
HIGHLIGHTS
Different membrane rotational speeds, aeration strategies, and maintenance cleaning methods were evaluated.
Membrane rotation played a significant role in maintaining the optimal transmembrane pressure, outperforming aeration.
The pilot plant operated continuously for 48 days without pre-screening and aeration.
There is significant potential for energy savings by reducing or eliminating aeration.
INTRODUCTION
The membrane bioreactor (MBR) system, integrating biological processes with membrane filtration, provides a robust approach to wastewater treatment (Judd 2016). Its compact design, achieved by eliminating the need for a secondary settling tank, confers a distinct advantage over conventional activated sludge (CAS) systems. The MBR process offers additional benefits, including the capability to operate at higher mixed liquor suspended solid (MLSS) concentrations, an extended sludge age, and reduced sludge production compared to CAS methods (Visvanathan et al. 2000; Pollice et al. 2008; Barreto et al. 2017). These benefits have encouraged the adoption of MBR technology across 200 countries by 2016, with the global market for MBR systems experiencing an annual growth rate of 15% (Judd 2016).
Nevertheless, fouling is a significant challenge associated with MBR systems, which is characterized by the accumulation of disruptive deposits on the membrane surface or within the pores. These deposits originate from retained salts, macromolecules, colloids, and particles (Ladewig & Al-Shaeli 2017). To deal with fouling, operational strategies such as chemical and mechanical cleaning strategies are necessary to prevent the membrane from fouling. However, these countermeasures add complexity to the system, resulting in increased capital and operational costs. The main cost drivers are the periodic need to replace membrane modules and the implementation of anti-fouling strategies, which collectively raise the financial burden of utilizing MBR technology (Judd 2011; Rahman et al. 2023).
The phenomenon of fouling represents the primary challenge encountered in an MBR system. As a result, solving or reducing the problem has the greatest impact on the efficiency of an MBR plant (Al-Asheh et al. 2021). Thereby, fouling is a process by which the membrane experiences a loss of performance due to the deposition of dissolved and/or suspended matter on the membrane surface, openings, or within the pores (Iorhemen et al. 2016). This leads to a decrease in flux, which is the quantity of material passing through a unit area of membrane per unit time, measured in litres per m2 per hour (or LMH) (Judd 2011). The fouling depends on a large number of factors such as the cleaning strategy, the operating conditions, the specific properties of the wastewater, and the membrane used (Al-Asheh et al. 2021). A common strategy to reduce membrane fouling is by using coarse bubble (CB) aeration under the membrane module. Yet the increased oxygenation leads to increased foam formation and increased ongoing energy requirements, which can be almost double that of the CAS process (Iorhemen et al. 2016; Judd 2016; Al-Asheh et al. 2021). The use of CB aeration compromises oxygen transfer efficiency, which could be significantly improved by using fine bubble (FB) aeration (Henkel et al. 2009; Mahdariza et al. 2023; Zuo et al. 2024).
Another possibility is to increase the shear force to prevent the attachment of biofilm to the membrane surface. A potential solution is to introduce a rotational membrane, which can minimize the formation of reversible fouling (Rector et al. 2006; Wu et al. 2008; Zuo et al. 2010; Jiang et al. 2012). Furthermore, a novel pilot-plant scale prototype of rotating hollow fibre (HF) MBR modules was built and studied in the batch process (Mahdariza et al. 2022, 2023). In contrast to conventional HF MBR modules, the new concept applies a continuing sheer force by rotation to the new arrangement of HF membrane modules. The results showed that the additional energy required for rotation can be overcompensated by the improved oxygen transfer efficiency driven by rotation.
In this study, a series of experiments was conducted with the rotating HF membrane module in operation. The objective was to evaluate the performance of the pilot plant under various operational parameters, with a focus on transmembrane pressure (TMP) and filtrate quality.
MATERIAL AND METHODS
The research explored a range of operational variables, such as diverse module rotational speeds and aeration systems, to assess their impact on system performance. After each experimental session, the system underwent maintenance cleaning (MC), which involved chemically enhanced backflush with 20 L of sodium hypochlorite (NaOCl). This procedure was essential for maintaining the efficiency of the system.
In the first phase of the experiment, two different ratios of filtering to backflushing processes and fluxes were examined. The outcomes of this phase guided the subsequent experimentation during the second phase, which focused on testing different aeration strategies and membrane module rotations. Each experimental set was scheduled to run for approximately 1 week (6–7 days), after which the MC of the membrane module was conducted. This phase also aimed to evaluate the effectiveness of MC by exploring different concentrations of NaOCl. Following the methodologies outlined by Judd (2011) and Wang et al. (2014), two distinct approaches were compared: cleaning in air (CIA), involving emptying the reactor, and cleaning in place (CIP), where the reactor remained filled with wastewater. The final phase extended over 7 weeks, during which the pilot plant operated to assess the feasibility of the MBR module in practical applications.
RESULTS AND DISCUSSIONS
The measurement of the quality of sludge and filtrate
A series of parameters were subjected to periodic laboratory sampling during all experiments conducted within this study. In addition to monitoring the quality of the input sludge, it was necessary to observe any changes in the filtrate quality that may have occurred when different operational setups were applied during the operational period of the pilot plant.
As illustrated in Table 1, the pilot plant demonstrated effective performance in terms of soluble chemical oxygen demand (sCOD) removal and filtrate turbidity even in the absence of pre-screening. Despite the MLSS concentration exceeding 8 g/L on several occasions during the 48-day operation period (third phase), the turbidity of the filtrate remained consistently below 1 NTU. Furthermore, the pilot plant demonstrated the ability to maintain an average sCOD removal rate of 54%, which aligns with reported values for microfiltration MBR systems from other studies, which range between 25 and 98% (Ahn & Song 1999; Baek & Pagilla 2006; You et al. 2007; Lin et al. 2012; Deowan et al. 2019; Kabuba et al. 2023).
The characteristic of sludge and the filtrate during the experiment
Parameters . | First phase . | Second phase . | Third phase . | |||
---|---|---|---|---|---|---|
Sludge . | Filtrate . | Sludge . | Filtrate . | Sludge . | Filtrate . | |
MLSS concentration (g/L) | 5.9 (±0.4) | – | 6.2 (±0.4) | – | 7.0 (±2.2) | – |
Turbidity (NTU) | – | 0.6 (±0.1) | – | 0.8 (±0.2) | – | 0.4 (±0.1) |
pH | 6.9 (±0.1) | 7.0 (±0.0) | 6.9 (±0.1) | 7.0 (±0.1) | 6.8 (±0.1) | 6.8 (±0.1) |
Conductivity (μS/cm) | 1,203 (±54) | 1,230 (±58) | 1,165 (±153) | 1,206 (±146) | 734 (±211) | 849 (±215) |
sCOD (mg/L) | 50.0 (±3.5) | 21.6 (±2.6) | 48.8 (±2.4) | 21.0 (±1.8) | 24.8 (±8.1) | 14.5 (±3.9) |
Parameters . | First phase . | Second phase . | Third phase . | |||
---|---|---|---|---|---|---|
Sludge . | Filtrate . | Sludge . | Filtrate . | Sludge . | Filtrate . | |
MLSS concentration (g/L) | 5.9 (±0.4) | – | 6.2 (±0.4) | – | 7.0 (±2.2) | – |
Turbidity (NTU) | – | 0.6 (±0.1) | – | 0.8 (±0.2) | – | 0.4 (±0.1) |
pH | 6.9 (±0.1) | 7.0 (±0.0) | 6.9 (±0.1) | 7.0 (±0.1) | 6.8 (±0.1) | 6.8 (±0.1) |
Conductivity (μS/cm) | 1,203 (±54) | 1,230 (±58) | 1,165 (±153) | 1,206 (±146) | 734 (±211) | 849 (±215) |
sCOD (mg/L) | 50.0 (±3.5) | 21.6 (±2.6) | 48.8 (±2.4) | 21.0 (±1.8) | 24.8 (±8.1) | 14.5 (±3.9) |
The impact of filtration-to-backflush time ratio and flux on TMP
In the initial phase of the experiment, the investigation focused on the impact of the filtration-to-backflush ratio and the flux on the increase of TMP. The objective was to achieve a TMP increase that would allow stable operation with one MC per week.
TMP increase across varied filtration-to-backflush time ratios and filtration fluxes.
TMP increase across varied filtration-to-backflush time ratios and filtration fluxes.
The impact of aeration strategies and module rotation on TMP
In the subsequent stage of the investigation, the focus shifted to assessing the impact of three different aeration strategies: without aeration, FB aeration, and CB aeration. This assessment was further enhanced by adjusting the rotational speeds of the membrane module to 0, 20, and 30 rpm. The limitation to 30 rpm was based on insights from a previous study on the same membrane system conducted by Mahdariza et al. (2022). The study identified that standard aeration efficiency peaked at this speed, with efficiency declining at higher rotation speeds.
TMP increase across varied aeration strategies and membrane rotational speeds (Mahdariza et al. 2024).
TMP increase across varied aeration strategies and membrane rotational speeds (Mahdariza et al. 2024).
A number of studies have highlighted the enhanced physical cleaning benefits of CB aeration. These include studies by Judd (2005), Phattaranawik et al. (2007), Braak et al. (2017), and Zhao et al. (2021). Furthermore, a study conducted by Jones (2017) demonstrated that the rotational mechanisms in a rotating MBR system contributed to a mere 12% of fouling prevention by removing the cake, with the majority of the removal achieved through air scouring. However, the findings from this study revealed no substantial differences in the increase of TMP among the various aeration strategies when membrane rotation was implemented. This indicates that the efficacy of membrane cleaning and fouling prevention may not be significantly influenced by the type of aeration employed as assumed. This thereby emphasizes the pivotal role of membrane rotation in maintaining optimal membrane performance. In addition to facilitating oxygen transfer, membrane rotation has been demonstrated to increase shear force, thereby limiting the build-up of a cake layer on the surface of the membrane. This finding is consistent with previous research on this specific rotating HF membrane module (Mahdariza et al. 2022, 2023).
MC strategy
Furthermore, the influence of MC on TMP reduction was inspected through experiments employing three different concentrations of NaOCl solution while maintaining the solution temperature between 30 and 38 °C. In each cleaning cycle, a 5 L volume of NaOCl solution was introduced to the membrane module four times, each followed by a 5-min soaking period.
Table 2 demonstrates that increasing the concentration of NaOCl did not result in a proportional decrease in TMP within the CIP approach. In contrast, the CIA approach demonstrated a direct correlation between increased NaOCl concentration and TMP reduction, highlighting its effectiveness. It is important to note the variation in initial TMP values prior to MC in different experimental setups. Despite the observed variations, the data obtained suggest that the CIA method is more effective in reducing TMP when compared to the CIP method.
TMP before and after MC (Mahdariza et al. 2024)
NaOCl solution concentration (ppm) . | CIA . | CIP . | ||||
---|---|---|---|---|---|---|
TMP before cleaning (bar) . | TMP after cleaning (bar) . | TMP reduction (bar) . | TMP before cleaning (bar) . | TMP after cleaning (bar) . | TMP reduction (bar) . | |
250 | 0.58 | 0.32 | 0.26 | 0.55 | 0.36 | 0.19 |
500 | 0.56 | 0.30 | 0.26 | 0.54 | 0.41 | 0.13 |
1,000 | 0.69 | 0.33 | 0.36 | 0.37 | 0.30 | 0.07 |
NaOCl solution concentration (ppm) . | CIA . | CIP . | ||||
---|---|---|---|---|---|---|
TMP before cleaning (bar) . | TMP after cleaning (bar) . | TMP reduction (bar) . | TMP before cleaning (bar) . | TMP after cleaning (bar) . | TMP reduction (bar) . | |
250 | 0.58 | 0.32 | 0.26 | 0.55 | 0.36 | 0.19 |
500 | 0.56 | 0.30 | 0.26 | 0.54 | 0.41 | 0.13 |
1,000 | 0.69 | 0.33 | 0.36 | 0.37 | 0.30 | 0.07 |
This finding is consistent with the results from Brepols et al. (2008), which demonstrated that CIA achieved twice the permeability recovery compared to CIP with the same NaOCl dosage. However, it is noteworthy that the overall CIA process in this pilot plant required an additional hour compared to CIP, which was attributed to the time necessary for emptying and refilling the tank. Therefore, the CIP method is considered to be sufficiently effective for routine weekly MC.
Further experiment without aeration
TMP increase for 48 days of operation with membrane rotation and without aeration.
TMP increase for 48 days of operation with membrane rotation and without aeration.
In terms of MC, the pilot plant was effectively operated with four times CIP and one time CIA over the course of this period. An operational pause on the 26th day, necessitated by pump repairs at the WWTP, briefly halted inflow to the MBR pilot plant. As seasonal temperatures began to fall, leading to cooler wastewater, there was a noticeable acceleration in the increase of the TMP after the 28th day of operation. The experiment showed that a higher concentration of NaOCl for CIP was necessary due to this condition. By the end of the experiment, TMP levels had risen to over 0.6 bar. This result emphasizes the need for increased NaOCl concentrations and more frequent MCs during colder months, which is particularly important for operations without aeration.
The impact of sludge temperature on membrane fouling
Several studies have found a correlation between decreasing temperatures and increased membrane fouling. This is attributed to the enhanced release of soluble microbial products and extracellular polymeric substances by filamentous bacteria (van den Brink et al. 2011; Ma et al. 2013; Iorhemen et al. 2016). Therefore, temperature differentials are considered a significant factor that could either exacerbate or alleviate fouling during filtration and backwash operations. To investigate this hypothesis, the study repeated two experiments under varying sludge temperatures caused by seasonal changes. Figure 6 illustrates the comparison of these experiments, demonstrating the effect of temperature on TMP increase.
This outcome demonstrates that experiments conducted with sludge at a temperature 6 °C lower resulted in a faster increase in TMP. Hence, for this specific rotating membrane module configuration, it is not recommended to use a configuration without aeration during colder seasons. To overcome the challenges posed by lower temperatures, it is suggested to adopt one or more alternative strategies. Itokawa et al. (2008) recommended doubling the frequency of MC in colder months compared to summer. Other strategies, such as the use of aeration systems, reducing operational flux, and applying higher concentrations of NaOCl for MC, can also significantly mitigate fouling rates and ensure optimal membrane system performance during colder months.
CONCLUSIONS
This study evaluated the performance of a pilot plant MBR system utilizing a rotating HF membrane module by measuring the increase in TMP and the quality of the filtrate during the filtration operation. The operational variables of filtration flux, filtration-to-backflush time ratio, aeration strategies, and membrane rotational speeds were found to exert an influence on the dynamics of TMP during membrane filtration. Furthermore, two distinct methodologies for conducting MC were employed, and their efficacy in reducing TMP was evaluated. The most significant findings of this study are as follows:
(a) Based on the initial phase of the experimental series, the optimal configuration for the pilot plant to operate for 7 days of filtration was a flux of 24 LMH and a filtration-to-backflush time ratio of 9:1 min. The pilot plant also demonstrated the capacity to operate at elevated filtration flux; however, it is not recommended for extended periods of operation.
(b) The pilot plant exhibited the ability to operate and achieve the desired level of TMP increase through membrane module rotation even in the absence of aeration. This was corroborated by a subsequent extended period of operation utilizing this configuration. The results indicated that membrane rotation had a more pronounced effect on the control of fouling than the type of aeration employed.
(c) Both FB and CB aerations performed in the same manner for the TMP increase behaviour during the experiments. This suggests that FB aeration can be used to enhance oxygen transfer without compromising fouling mitigation and should the membrane module be integrated as a submerged MBR system into an existing aeration tank of a CAS system.
(d) Temperature effects on fouling dynamics are correlated with an accelerated TMP increase at lower temperatures. Consequently, it is recommended that MC protocols and operating strategies be adjusted during colder seasons.
Further research could be directed towards optimizing the operating parameters, such as varying the membrane rotational speeds and applying relaxation, to improve the performance of this rotating HF MBR module for its application in wastewater treatment.
ACKNOWLEDGEMENTS
The authors would like to express their sincere appreciation to the Ministry of Finance of the Republic of Indonesia for granting the LPDP scholarship to Fathul Mahdariza (ref. no. S-1289/LPDP.4/2018). Special recognition is also given to Dipl.-Ing. Klaus Strätz (Enwat GmbH) and Andreas Scharf (frapp GmbH) for their role in the initial construction and automation of the pilot plant. The authors are deeply indebted to Dr.-Ing. Jochen Henkel for his invaluable contributions and manifested in insightful discussions that greatly enriched the research. Appreciations are also due to Dipl.-Ing. Frank Koch, Dietmar Landgrebe, and the dedicated team at WWTP Kassel for their unwavering technical support. Further acknowledgement is extended to Dr.-Ing. Ursula Telgmann, Andrea Brandl, and Monika Degenhardt from the Chair of Urban Water Engineering, the University of Kassel, for their meticulous laboratory work. The authors are equally appreciative of the contributions made by Dipl.-Ing. Ralf Feldner and Lukas Marmucki for their technical assistance. Lastly, the authors express profound gratitude to PD Dr.-Ing. Stephan Fuchs and the esteemed team at the Department of Water Quality Management, Karlsruhe Institute of Technology, whose assistance during the initial stage of this study was invaluable.
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.