ABSTRACT
This study examines the impact of incorporating a mobile bed into a membrane bioreactor (MBR) system on the treatment efficiency of dairy industry effluents. Initially, a conventional MBR system was operated for 60 days, followed by a modification that included a support material and ran for another 60 days under identical conditions. Performance was evaluated based on the removal efficiencies for soluble chemical oxygen demand (CODs), phenolic compounds, and oils and greases (OG), alongside measurements of solid content, dissolved oxygen, temperature, mixed liquor pH, and transmembrane pressure (TMP). The introduction of the mobile bed led to an increase in removal efficiencies for COD and phenolic compounds from 94.4 and 92.7% to 98 and 94.4%, respectively, marking statistically significant improvements (p < 0.05), while OG removal remained the same in both strategies (87.7%) (p > 0.05). Moreover, the modified system showed a more stable TMP profile, reducing the need for cleaning interventions compared to the conventional system, which experienced a notable TMP increase requiring cleaning at a 0.6 bar threshold. The findings suggest that integrating a mobile bed into MBR systems significantly enhances the treatment of dairy effluents, presenting an interesting solution for the upgrade of this type of system.
HIGHLIGHTS
The integration of a mobile bed into MBR systems significantly increases the removal efficiency of phenolic compounds.
Modified MBR systems with a mobile bed significantly reduce membrane fouling, making MBR operation more stable.
Integrating a mobile bed offers a promising upgrade for dairy effluent treatment.
INTRODUCTION
In 2022, Brazil's milk production soared to 34.6 billion L, positioning it as the world's fourth-largest milk producer. This achievement not only underscores Brazil's pivotal role in the global dairy sector but also highlights its significant contributions to the nation's social and economic fabric (IBGE 2022). However, the dairy industry is also associated with extensive water usage for both production and cleaning purposes, leading to the production of considerable amounts of wastewater. The improper disposal of this wastewater into aquatic ecosystems without adequate treatment raises grave environmental concerns, as highlighted by Carvalho et al. (2013).
Dairy effluents are characterized by their high organic load, a direct consequence of milk's nutrient-rich nature and the extensive sanitation processes involved in the dairy industry, including the cleaning of floors, tanks, and pipelines. These effluents are also laden with oils, greases, suspended solids, dissolved solids, and nutrients like nitrogen and phosphorus, which can trigger eutrophication in water bodies. The presence of hazardous compounds such as phenol, primarily from cleaning operations, adds to the complexity. Phenol's potential for bioaccumulation poses a significant threat to the local ecosystem and human health, necessitating efficient effluent treatment solutions to mitigate these environmental and health risks (Adulkar & Rathod 2014).
The generation of effluents in the dairy industry is significantly influenced by washing and cleaning activities, which can account for 50–95% of a dairy plant's total effluent volume. This substantial share emphasizes the urgent need for effective effluent treatment strategies to address the environmental impacts associated with these activities (Andrade et al. 2014).
Biological treatment methods, including aerobic systems, biological filters, stabilization ponds, and activated sludge processes, are among the primary strategies for dairy effluent management (Gupta 2011). The application of chemical processes in coagulation, flocculation, and flotation also plays a critical role in effluent treatment. However, challenges arise with conventional treatment systems, such as activated sludge processes and aerated lagoons, particularly in removing color and recalcitrant compounds like phenol, alongside their limited capacity to manage fluctuating organic loads (Izadi et al. 2018). These limitations have paved the way for the adoption of advanced technologies, notably membrane bioreactors (MBRs), which have demonstrated superior efficiency in treating both municipal and industrial effluents globally (Krzeminski et al. 2017).
MBRs merge physical membrane filtration with biological treatment, offering advantages such as reduced footprint, enhanced treatment efficiency, and the ability to retain biomass over extended periods, thereby eliminating the need for sedimentation tanks (Acarer 2023). The research strategies were to study the operation of the MBR in conventional mode and with the inclusion of support material, under the same operational conditions.
To boost the efficacy of MBR systems further, the integration of supporting materials like moving beds or fillers with large surface areas has been explored. These materials facilitate the growth of attached bacteria, thereby enhancing the concentration of biomass and improving the overall treatment efficiency. This evolution has led to the development of moving bed biofilm reactors (MBBRs), which are adept at degrading organic matter and nutrients in wastewater (Fujii et al. 2013).
Despite these advancements, there remains a significant gap in research, particularly concerning the treatment of dairy effluents using moving bed MBRs with mixed support materials and the removal of persistent pollutants like phenol.
METHODS
Experimental setup and characteristics of the membrane module
The methodology section of the article describes the design and operation of the MBR system used in the experimental study. The MBR system, central to this research, was housed in an acrylic rectangular tank with dimensions tailored to hold a total volume of 60 L, out of which 57 L were actively utilized for the experiments. This design choice ensured ample space for the treatment processes to occur within the reactor. To maintain an aerobic environment essential for the biological processes, the tank and the membrane within were aerated from the bottom using an air compressor. This setup guaranteed the continuous oxygenation of the system throughout the duration of the study.
The operational dynamics of the MBR were facilitated by gravity, which allowed for the seamless feed of wastewater into the system, while the removal of treated water (permeate) was managed by a digital peristaltic pump. This pump's operation was regulated by a float switch, ensuring a consistent flow rate for both the inlet and outlet, thereby maintaining a continuous system operation. The integration of a pressure sensor, connected to the permeate outlet pipe and linked to a digital vacuum gauge, provided real-time monitoring of the transmembrane pressure (TMP), a critical parameter in assessing membrane performance.
Automation of the system was achieved through the use of an Arduino board, which controlled the peristaltic pump responsible for the permeate extraction. This automation was pivotal in maintaining operational consistency and efficiency, with the extracted permeate being collected in a designated tank for further analysis or disposal.
The microfiltration process within the MBR was conducted using a submerged membrane module, characterized by a nominal pore size of 0.03 μm and an outer fiber diameter of 2.6 mm. The membrane, constructed from polyester, was of the hollow fiber type, offering a filtration area of 0.5 m2. This setup was designed to achieve a permeate flux rate ranging from 5 to 15 L/h, optimizing the treatment capacity of the system.
To enhance the biofilm development and increase the system's efficiency, the MBR incorporated mixed moving bed media, consisting of two types of support materials. These materials were selected based on their diameters (30 and 15 mm) and surface areas (700 m2/m3 and 550 m2/m3, respectively), providing a varied substrate for microbial attachment and growth.
Diagrammatic illustration of the membrane bioreactor employed in the study. Source: Gavlak et al. 2024. (1) Influent tank; (2) permeate tank; (3) air blower; (4) filling pump; (5) peristaltic pump; (6) power supply; (7) membrane; (8) aerobic reactor; (9) digital gauge.
Diagrammatic illustration of the membrane bioreactor employed in the study. Source: Gavlak et al. 2024. (1) Influent tank; (2) permeate tank; (3) air blower; (4) filling pump; (5) peristaltic pump; (6) power supply; (7) membrane; (8) aerobic reactor; (9) digital gauge.
Characteristics of the wastewater and start-up conditions of the MBR
The wastewater utilized in this study was sourced from a dairy processing facility, which specializes in cheese production alongside other dairy derivatives on a lesser scale. This effluent was a byproduct of both the dairy manufacturing processes and the cleaning routines for production equipment and facility floors. Prior to its use in the research, the effluent underwent preliminary treatment steps aimed at removing solid and fatty substances. This treatment sequence included passage through a grease trap, a flotation unit, and a settling pond, effectively reducing the content of solid and greasy materials.
Following these initial treatment stages, sludge rich in biomass was collected from the facility's aerobic treatment pond. This sludge was then subjected to gravity thickening, a process aimed at concentrating the biomass to achieve a total suspended solids (TSS) content of approximately 8 g/L. Subsequently, the sludge was introduced into the research reactor, where it underwent an acclimatization phase. This phase was critical for adjusting the concentration of suspended solids within the reactor to a stable level of 5 g/L, a process that typically spanned approximately one month for each treatment strategy implemented.
To ensure optimal conditions for the acclimatization and operation of the biomass, the ambient temperature around the reactor was meticulously controlled using an air conditioning system. This temperature regulation was crucial for minimizing environmental fluctuations and creating an ideal setting for the biomass to adapt and thrive within the reactor.
Operating conditions applied to the BRM system
The experimental investigation of the MBR system was structured around two distinct operational strategies. The first strategy, referred to as Strategy 1 (S1), adhered to a conventional approach to MBR operation. In contrast, the second strategy, Strategy 2 (S2), introduced an innovative element by incorporating a mixed moving bed into the reactor setup. Despite these differences, both strategies were executed under identical operational conditions to facilitate a direct and meaningful comparison of their performances.
The adopted sludge age was 20 days, the hydraulic retention time was 20 h, permeate flow rate was 47.5 ml/min, and an operational time for each strategy was 60 days, both operated continuously. Throughout the systems operation period, the reactor operated under continuous flow but with intermittent filtration mode, adopting 8 min of filtration and 1 min for membrane relaxation (Wu et al. 2008). This allowed the solids to be carried away by air bubbles from the aerators, thereby slowing down membrane fouling. During the entire experimental period, the reactor temperature remained between 25 and 35 °C, and sludge disposal was carried out daily with a volume of 2.85 L, calculated based on the reactor's useful volume and the adopted sludge age. After 60 days of operation, the second operating strategy of the system was initiated, with the addition of a mixed moving bed comprising 30% of the reactor volume, as reported in the literature by Rusten et al. (2006) and Yang et al. (2009). This was divided into 15% of type one material, characterized by a diameter of 30 mm and a surface area of 700 m2/m3, and 15% of type two material, with a diameter of 15 mm and a surface area of 550 m2/m3, forming what was called a mixed moving bed.
Analytical methods and monitoring
In order to obtain a comprehensive analysis of the system's operation, samples were collected from three different points within the system: raw effluent, mixed liquor and permeate. Analytical procedures followed the methodologies outlined in the Standard Methods for the Examination of Water and Wastewater, 23rd Edition (APHA 2017). For the raw effluent and permeate, soluble chemical oxygen demand (CODs), total phenolic compounds, and oil and grease were monitored. For the mixed liquor, CODs, total phenolic compounds, TSS, volatile suspended solids (VSS), fixed suspended solids (FSS), dissolved oxygen, pH, and temperature were measured.
Prior to analysis, samples extracted from the mixed liquor of the aerobic reactor underwent a pre-filtration process using membranes with a pore size of 0.45 μm. This step was essential for the subsequent determination of CODs and the concentration of total phenolic compounds within the samples. Furthermore, to ensure vigilant monitoring of membrane fouling – a critical aspect of reactor operation – daily recordings of the TMP were obtained using a digital vacuum gauge. A TMP threshold of 0.6 bar was established as the indicator for the necessity of cleaning the membrane module. Additionally, the analysis of VSS present in the biofilm adhered to the moving bed was conducted, adhering to the methodologies outlined by Zhang et al. (2014) and Costa et al. (2018).
Additionally, respirometric assays were conducted using the methodology described by Ochoa et al. (2002) and Wolff et al. (2003) to assess and describe the metabolic activity of the biomass present in the reactor, as well as to determine the specific oxygen uptake rate (SOUR) by bacteria, through analysis of the decrease in dissolved oxygen concentration in the medium following substrate addition, as detailed in the experimental protocol. Respirometric analysis was performed three times during the operation of each adopted strategy, at the beginning, middle, and end of the operating period.
Statistical methods
For the processing and analysis of data collected from the two operational strategies implemented within the system, the trial version of STATISTICA 7 software (2005) was utilized. To assess the data's homoscedasticity, normality, and independence, the Bartlett test, Shapiro-Wilk test, and Durbin-Watson test were employed, respectively. Following this preliminary analysis, an Analysis of Variance (ANOVA) was conducted with a 95% confidence level to facilitate a comparative analysis of the data. This was further supplemented by the Tukey HSD test, utilizing an alpha level (α) of 0.05, to discern any significant differences between the operational strategies based on the parameters evaluated.
RESULTS AND DISCUSSION
Throughout the operational phase, the effluent being treated displayed average concentrations of CODs (1,010 ± 150 mg/L); total phenolic compounds (12 ± 2 mg/L); oil and grease 1,358 ± 96 and pH (7.3 ± 0.3).
Dairy industry effluents are known for their high organic content, which includes lipids, carbohydrates, and proteins, alongside significant amounts of other organic materials. These effluents, when released into aquatic environments without adequate treatment, drastically decrease the water's dissolved oxygen levels. This reduction is primarily due to the oxygen consumption by aerobic microorganisms that decompose the organic matter, thereby endangering local and adjacent aquatic ecosystems (Villa et al. 2007).
A significant environmental concern associated with dairy effluents is the high concentration of phenolic compounds, which are byproducts of cleaning processes involving chemical agents. Phenols are characterized by their structure, consisting of one or more hydroxyl groups attached to an aromatic ring. The environmental and health risks posed by phenolic compounds are linked to their hydrophobic nature, which leads to low water solubility and a tendency to accumulate in fatty tissues. Additionally, these compounds are capable of forming free radicals, exhibit acidic properties, are resistant to degradation (recalcitrant), bioaccumulate, and present significant environmental and public health risks (Prasad et al. 2019). Given these factors, Brazil enforces strict regulations on the permissible levels of phenolic compounds in discharged effluent, setting the limit at 0.5 mg/L. Thus, the treatment and effective removal of these compounds are critical for protecting environmental health and biodiversity.
The effluent's high levels of oils and grease, primarily derived from the substantial amounts of animal fat in raw effluent (given that cow's milk is approximately 3.8% fat, according to Blowey et al. 1992), represent another challenge. Oils and grease not only degrade slowly but can also hinder biological degradation processes and decrease oxygen transfer to biological flocs. This interaction can adversely affect degradation rates, leading to diminished efficiency in the treatment process (Chipasa & Mechzyeka 2006).
Biomass monitoring analysis
Throughout the operational period, the levels of TSS, FSS, and VSS remained relatively stable across both implemented operational strategies. During the initial strategy, TSS levels averaged at 4.99 ± 0.3 g/L, with FSS and VSS at 1.4 ± 0.1 g/L and 3.61 ± 0.22 g/L, respectively. For the second strategy, TSS, FSS, and VSS levels were recorded at 5.5 ± 0.3 g/L, 1.4 ± 0.18 g/L, and 3.81 ± 0.3 g/L, respectively. Notably, the substantial VSS concentrations in both strategies highlight the abundance of microbial presence (Jordão & Pessôa 2011). Von Sperling (2005) notes that in extended aeration activated sludge systems, VSS concentrations in aeration tanks can range between 4,500 to 5,000 mg/L. Consequently, higher VSS concentrations in the aerobic reactor imply increased biomass availability for biological assimilation, potentially reducing the required system footprint. However, this heightened VSS level might also exacerbate membrane fouling issues (Damayanti et al. 2011).
When analyzing the VSS/TSS ratio of the reactor, which indicates the degree of mineralization of the mixed liquor, it was observed that in operation condition S1, the average was 0.8 ± 0.03, while in condition S2, this average was 0.79 ± 0.02. However, in both conditions, the values obtained approached 0.85, a value recommended by Metcalf & Eddy (2003), as well as the value of 0.9 reported by Belli (2011) when using MBRs in sequential batch mode for the treatment of sanitary effluent, and similar to the value found by Kellner (2014) in their study of sanitary effluent treatment by MBR with mobile bed in sequential batch mode, where this value was 0.78.
A comparative analysis of microbial growth on two distinct support materials utilized in the second operational strategy revealed significant differences in microbial proliferation rates, as confirmed by statistical significance (p < 0.05). The first type of support material exhibited a microbial growth rate of 0.027 gVSS/L·day, with an average VSS concentration of 0.88 ± 0.475 g/L during the second strategy phase. In contrast, the second type of support material demonstrated a halved growth rate of 0.013 gVSS/L·day and an average VSS concentration of 0.48 ± 0.35 g/L, suggesting that support materials with larger diameters and surface areas are more conducive to biofilm development. The observed variability in concentrations is attributed to the weekly analytical frequency and the continual biomass expansion.
The dissolved oxygen levels in the mixed liquor of the MBR exhibited an average of 5.6 ± 0.48 mg/L during the first strategy, statistically similar to the 5.5 ± 0.41 mg/L observed in the second strategy (p > 0.05). These values, which consistently ranged between 4 and 6 mg/L, align with the typical operational standards for MBRs (Cicek et al. 2001) and underscore the vital role of dissolved oxygen in supporting the metabolic processes of microorganisms for organic matter oxidation (Albornoz 2017).
The pH levels in the mixed liquor remained stable at an average of 7.7 (± 0.2 for the first strategy and ±0.1 for the second), indicating reactor stability across both configurations without necessitating adjustment. This pH range is ideal for aerobic effluent treatment processes, ideally positioned between 6 and 8, as recommended by Von Sperling (2014), thus affirming that the operational conditions were within optimal parameters for efficient treatment.
Temperature measurements of the mixed liquor averaged 26.5 ± 1.8 °C during the first strategy and 27 ± 1.2 °C during the second, showing no significant variation between the two. This consistency can be attributed to the controlled environment facilitated by air conditioning, with the recorded temperatures falling within the ideal biological treatment range of 25–35 °C, as suggested by Jordão & Pessoa (2011).
The average values of total endogenous SOUR obtained from the three analyses performed in each strategy were 33.5 ± 2 mgO2/gVSS·h for condition S1 and 31.1 ± 3 mgO2/gSSV·h for condition S2, considered statistically equal (pvalue = 0.0996). This suggests that there were no significant changes in oxygen demand to maintain the cellular functions of microorganisms after the addition of support material, despite a possible higher biomass concentration due to attached growth. Similarly, heterotrophic SOUR, related to oxygen consumption from carbonaceous substrates, showed no significant variation after the addition of attached growth (pvalue = 0.9226). The mean SOUR values were 48.1 ± 4 (S1) and 46.1 ± 3.1 (S2) mgO2/gVSS.h.
When evaluating the SOUR of total autotrophic oxygen consumption (related to oxygen consumption from the degradation of nitrogenous compounds), it was observed that during condition S1, the average SOUR for autotrophic total was 2.1 ± 0.9 mgO2/gVSS·h, while in condition S2, the average value obtained was 9.2 ± 1.1 mgO2/gVSSV·h. Therefore, there was a significant increase (pvalue = 0.0014) in the total autotrophic SOUR following the addition of support material in the second strategy compared to conventional operation. This increase was attributed to the creation of anaerobic, anoxic, and facultative zones in the support material, promoting the growth of nitrifying bacteria in the medium, which utilize part of the oxygen and are responsible for consuming nitrogenous compounds.
The results obtained are supported by the studies of Kellner (2014). The author evaluated the treatment of sanitary effluents using a MBR with moving bed biofilm reactor in sequential batch mode, varying the percentage of media fill in the reactor according to the useful area. From the data, different values of SOUR for endogenous, autotrophic, and heterotrophic processes were obtained. When 30% of the useful reactor volume was utilized, the values found were 9.20, 8.71, and 7.88 mgO2/gVSS·h, respectively. Conversely, when 40% of the volume was employed, the values for endogenous, autotrophic, and heterotrophic SOUR were 2.97, 11.24, and 21.73 mgO2/gVSS·h, respectively. These results underscore the importance of adding the moving bed biofilm reactor in increasing autotrophic SOUR, thereby promoting the proliferation of nitrifying bacteria and contributing to the removal of nitrogenous compounds, for example, although not assessed in the present study, it is important to emphasize.
Evaluation of organic matter reduction
Variations in COD concentrations in raw effluent, mixed liquor, and permeate across evaluated treatment strategies.
Variations in COD concentrations in raw effluent, mixed liquor, and permeate across evaluated treatment strategies.
During the implementation of the first operational strategy, a notable organic matter removal was achieved by the reactor's biomass, resulting in an average removal efficiency of 92 ± 1.1% and a mean concentration of CODs at 84 ± 6.5 mg/L. The introduction of a mixed mobile bed and an enhanced bacterial concentration within the system (encompassing both dispersed and attached growth) led to a significant improvement in aerobic degradation efficiency, which escalated to an average of 94 ± 0.6% and a reduced mean soluble COD concentration of 66 ± 6.2 mg/L, indicating a statistically significant enhancement (pvalue = 0.0045). In terms of the overall system efficiency during the first strategy (S1), an exemplary average removal rate of 94.3 ± 0.61% was recorded, culminating in a final treated effluent concentration of 55 ± 3 mg/L. This result was primarily attributed to the aerobic bacterial action, with a substantial portion, precisely 32.5 ± 2.8% of the soluble CODs being intercepted by the microfiltration membrane.
In the execution of the second strategy, the system's total efficiency soared to an average of 98 ± 0.6%, presenting a significantly lower concentration of 20 ± 4.2 mg/L, surpassing the results achieved under conventional operational conditions (pvalue 0.0013). This outcome is supported by the findings of Erkan et al. (2018) and Fraga et al. (2017), who reported COD removal efficiencies of 98, 98.2, and 94.1%, respectively, in their studies focusing on the treatment of dairy industry effluents using conventional MBRs. The data derived from both operational strategies underscore the substantial reductions in COD levels, affirming the potential of these methodologies as viable solutions for the treatment of such effluents.
Analysis of phenol reduction
Dynamics of phenolic compound concentrations in raw effluent, mixed liquor, and permeate in the membrane bioreactor for the operational strategies studied.
Dynamics of phenolic compound concentrations in raw effluent, mixed liquor, and permeate in the membrane bioreactor for the operational strategies studied.
In the initial phase of the system's first operational strategy, following aerobic degradation and subsequent filtration, the permeate exhibited an average phenol concentration of 0.75 ± 0.05 mg/L, achieving a total removal efficiency of 92.7 ± 0.71%. Viero et al. (2007) reported a 100% removal of phenol using a conventional MBR with a submerged module for refinery effluent treatment, albeit starting from a lower initial concentration of 1.8 mg/L. Advancing to the second strategy, which incorporated a mixed mobile bed, the overall removal efficiency improved to 94.4 ± 1.68%, with the permeate showing an average phenol concentration of 0.64 ± 0.13 mg/L. The enhancement in phenol removal attributable to the introduction of the mixed mobile bed is statistically significant, demonstrating a notable difference in the efficiency of aerobic degradation and the overall performance of the implemented strategies (pvalue = 0.0016).
A commendable phenol removal was observed during the first strategy (S1) through aerobic digestion, where the inlet phenol concentration of 9.9 ± 1.62 mg/L was reduced to 2.16 ± 0.4 mg/L following degradation and biological assimilation processes, registering an average removal efficiency of 78.13 ± 2.32%. The introduction of the mixed mobile bed in the second strategy (S2) further enhanced bacterial degradation efficiency, achieving an average of 88.55 ± 4.81% with an inlet concentration of 10.37 ± 1.95 mg/L.
It is crucial to note that, although the average concentrations did not meet the regulatory standards, from the 43rd day of the second strategy, phenol concentrations complied with the thresholds set by Brazilian legislation (CONAMA Resolution 430/2011), maintaining below the maximum limit of 0.5 mg/L until the end of the experiment, with a final concentration of 0.39 mg/L (BRASIL 2011).
The observed increase in phenol removal in MBRs can be attributed to the enhanced acclimatization and assimilation capacity of the bacteria adhered to the mixed mobile bed and the increased biomass concentration, thanks to the added support material. This improvement is particularly significant, considering that phenolic compounds are especially difficult to degrade through biological treatment processes. Moreover, MBRs with biomedia enhance the degradation of complex organic compounds by offering a larger surface for biofilms, creating varied microenvironments for specialized microorganisms, and increasing biomass retention. However, it is remarkable the scarcity of research focused on the removal of phenolic compounds from dairy industry effluents, using MBR configurations, whether conventional, mobile bed, or mixed mobile bed, highlighting the need for more studies in this area.
Analysis of oil and grease reduction
Variability in oil and grease concentrations in raw effluent, mixed liquor, and permeate across diverse treatment strategies.
Variability in oil and grease concentrations in raw effluent, mixed liquor, and permeate across diverse treatment strategies.
During the operation of Strategy 1 (S1), the initial average concentration of oil and grease in the untreated effluent was notably high at 1,372.4 ± 115.70 mg/L. Utilizing a conventional MBR for treatment resulted in an average removal efficiency of 87.74 ± 1.20%, leading to an average permeate concentration of 170.3 ± 7.55 mg/L. The introduction of a mixed mobile bed maintained the removal efficiency relatively stable, with an average of 87.79 ± 0.36%. The concentrations of oil and grease in the untreated effluent and the treated permeate for this modified strategy were 1,338 ± 73.78 mg/L and 165.74 ± 5.29 mg/L, respectively, marking a statistically significant reduction (pvalue = 0.342).
In both operational strategies, the removal efficiencies were deemed high. Nonetheless, the integration of the mixed mobile bed did not markedly enhance the reduction of oil and grease concentrations in the treated effluent (p > 0.05). This suggests that the presence of oil and grease compounds primarily influences biological degradation processes. Consequently, the principal mechanism driving the removal of these substances was identified as the membrane separation process. This observation aligns with the findings of Galvão & Gomes (2018), who reported an average removal efficiency of 98.39% through the use of a microfiltration membrane separation process in the treatment of effluents from the dairy industry. These results underscore the potential of physical separation methods in the effective removal of oil and grease from wastewater.
Assessment of membrane fouling
Variations in transmembrane pressure throughout different operational strategies.
Variations in transmembrane pressure throughout different operational strategies.
The contaminants present in wastewater can lead to the obstruction of membrane surfaces, resulting in the formation of a cake layer on these surfaces and an increase in TMP, as observed in the present study. These contaminants tend to accumulate in the membrane pores, leading to the clogging of these structures, such as through the accumulation of iron and manganese oxides, as well as other particulate aggregates according to Demirkol et al. (2021).
An elevated growth rate of TMP was noted during strategy S1, where the TMP reached the critical threshold of 0.6 bar on two occasions. Zsirai et al. (2012) emphasize that membrane filtration processes should ideally maintain TMP below 0.6 bar, as recommended by manufacturer guidelines, necessitating two cleaning interventions during this phase. Upon integrating a mixed mobile bed in strategy S2, the operation did not hit the 0.6 bar TMP threshold, thereby obviating the need for membrane cleaning sessions required in S1. Nonetheless, the peak TMP observed was 0.55 bar, nearing the critical limit. The integration of support material played a pivotal role in mitigating membrane fouling.
With the addition of the mobile bed, its continuous circulation within the mixture was facilitated by its buoyancy and the system's air currents, as described by Mannina & Viviani (2009), enhancing the exposure and interaction of the support material with the mixed liquor (Costa et al. 2018). This likely improved the homogenization within the reactor and facilitated the shearing action of these elements against the membrane fibers, aiding in the removal of solids and adhered fats. As a result, there was a notable reduction in TMP growth, thereby diminishing fouling incidents.
This observation aligns with findings from Duan et al. (2015), who noted that the friction generated by the collision between the support material and membrane fibers lessens biofilm formation on the membrane's outer surface, thereby enhancing permeability. Similarly, Liu et al. (2012) and Alves (2016) found that support material presence significantly curtails membrane fouling. Hence, the data suggest that employing a mixed mobile bed can effectively address one of the primary challenges in MBRs: the decline in hydraulic permeability due to material deposition and biofilm accumulation on the membrane surface (Erkan & Engin 2017; Bokhary et al. 2018).
This assertion gains further credence from a statistical correlation analysis between TMP values and the effluent concentrations of COD, phenolic compounds, and oils and greases for each evaluated strategy, with the analytical results presented in Table 1.
Correlation matrix analysis: TMP relationships with COD, total phenols, color, turbidity, and oils and greases variables
. | TMP (S1) . | TMP (S2) . | ||
---|---|---|---|---|
Parameters . | rKendal . | p-value . | rKendal . | p-value . |
CODs | 0.27 | 0.35 | 0.27 | 0.35 |
Phenolic compounds | 0.38 | 0.18 | 0.22 | 0.47 |
Oils and greases | 0.44 | 0.11 | 0.38 | 0.18 |
. | TMP (S1) . | TMP (S2) . | ||
---|---|---|---|---|
Parameters . | rKendal . | p-value . | rKendal . | p-value . |
CODs | 0.27 | 0.35 | 0.27 | 0.35 |
Phenolic compounds | 0.38 | 0.18 | 0.22 | 0.47 |
Oils and greases | 0.44 | 0.11 | 0.38 | 0.18 |
The data's statistical analysis demonstrates an absence of significant correlation between the concentrations of the parameters evaluated and the membrane fouling process across all strategies examined (p > 0.05). Thus, the observed decrease in the pronounced rise of TMP during the application of the second strategy cannot be attributed to alterations in the characteristics of the raw effluent concerning these specific parameters. Rather, it is likely that the effect was produced by the shearing action provided by the mixed bed, which, when combined with improved homogenization and circulation of the mixed liquor, could explain the reduction in TMP acceleration. This suggests that mechanical actions and fluid dynamics within the mixed bed play a crucial role in influencing membrane fouling processes, independent of the chemical composition changes in the effluent.
CONCLUSIONS
The study demonstrates that conventional MBR technology, when applied to dairy industry effluent treatment, achieves high efficiency in removing CODs phenolic compounds, and oils and greases, with initial removal rates of 94.4, 92.7, and 87.8%, respectively. The integration of a mixed bed into the MBR process significantly enhances these removal efficiencies to 98.0, 94.4, and 87.8% for soluble COD, phenolic compounds, and oils and greases, respectively. Statistical analyses confirm that the mixed bed's incorporation markedly boosts the treatment's effectiveness for soluble COD and phenolic compounds, highlighting its role in improving overall treatment capacity.
Regarding membrane fouling, the study found no significant correlation between the studied parameters and the increase in TMP in both the conventional and mixed bed MBR setups. However, the addition of a mixed bed was important to mitigate TMP rise, reducing the frequency of membrane cleaning. In contrast, the conventional MBR configuration showed a significant increase in TMP over time, necessitating more frequent cleaning. This evidence suggests that MBR systems, particularly those incorporating mixed beds, are effective in treating dairy effluent while addressing membrane fouling challenges, offering a promising solution for enhancing dairy effluent treatment efficiency.
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.