Journal of Water Process Engineering 64 (2024) 105684
Available online 24 June 2024
2214-7144/© 2024 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
Assessment of the effectiveness of a full-scale trickling filter for the treatment of municipal sewage in an arid environment: Multiple linear regression model prediction of fecal coliform removal
Ahmed Osmane
a,b, Khadija Zidan
c,d, Rabia Benaddi
d, Sofyan Sbahi
c,e, Naaila Ouazzani
c,d, Moustapha Belmouden
a, Laila Mandi
c,d,*aLaboratory of Organic Chemistry and Physical Chemistry (Molecular Modeling and Environment), Faculty of Sciences, University Ibn Zohr, Agadir, Morocco
bLaboratory of Biomolecular and Medicinal Chemistry, Faculty of Science Semlalia, University Cadi Ayyad, Marrakech, Morocco
cNational Center for Studies and Research on Water and Energy (CNEREE), Cadi Ayyad University, Marrakech, Morocco
dLaboratory of Water, Biodiversity and Climate Change, Faculty of Sciences Semlalia, Cadi Ayyad University, Marrakech, Morocco
eNational Institute of Scientific and Technological Research in Water, City of Innovation Souss Massa, Ibn Zohr University, Agadir, Morocco
A R T I C L E I N F O Editor: Soroush Abolfathi Keywords:
Trickling filter Wastewater treatment Fecal coliforms removal Linear regression model Landscape irrigation reuse
A B S T R A C T
In response to water scarcity, Morocco faces the challenge of using treated wastewater for irrigation purposes.
This study evaluates the efficiency of a full-scale trickling filter (TF) system in Imintanout, Morocco. The system consists of three septic tanks, two TFs, and secondary decanters. Over a 5-year period, the system showed sig- nificant reductions in pollutants: 98 % of TSS, 94 % of BOD5, 98 % of COD, 41 % of TP, and 88 % of NH4+
, with these reductions being statistically significant at the 95 % confidence level. A multiple linear regression (MLR) model successfully predicted the removal of fecal coliform (FC) by the TF, and a reduction of 2.88 log units was achieved. High cos2 values indicated the importance of hydraulic loading rate (HLR), BOD5, and FC, which were particularly affected by seasonal variations. Positive correlations between FC and TSS in certain periods highlight the seasonal variability in the composition of urban wastewater, which is effectively captured by the MLR model (R2 =0.77). Although the treated water complied with Moroccan discharge standards, its high nitrate (140 mg L−1) and FC (4.32 log units) levels made it unsuitable for reuse in agricultural and landscape irrigation, as they exceeded safety limits. This work highlights the importance of optimizing treatment systems to produce high quality reclaimed water, which is essential to meet the challenges of water scarcity.
1. Introduction
Water resources have been declining in recent decades due to increasingly challenging environmental conditions. Arid and semi-arid regions, characterized by erratic rainfall patterns, frequent droughts, and other environmental stressors, have experienced an increase in water demand. This increase is particularly noteworthy given the seemingly anomalous phenomenon of urban irrigation in an area known for its arid expanses. At the same time, water stress has been exacerbated by the expansion of agricultural and industrial efforts to meet the de- mands of a growing world population [1]. In order to further improve the quality of life in the city, more water resources need to be made available for the irrigation of green spaces. Municipal wastewater can be an attractive alternative source of irrigation for cities that is less costly
than pumping groundwater. The recovery and reuse of wastewater is usually described in terms of standards published or proposed by regional authorities or international organizations. To meet these stan- dards, wastewater must be treated before it can be used for irrigation.
Treating wastewater has also helped make cities in the developing world safer and more resilient. Through physical, chemical, or biological processes, wastewater treatment plants concentrate the pollutants in wastewater in the form of residues called sludge that can be used in agriculture. They also discharge treated water that meets specific stan- dards and can be reused for irrigation and industrial purposes. Treat- ment processes include trickling filters, natural/aerated lagoons, biological discs, activated sludge, and multi-soil-layering systems [2–6].
It is from this perspective that this study is conducted, suggesting the use of a classic trickling filter filled with plastic material. The trickling filter
* Corresponding author at: National Center for Studies and Research on Water and Energy (CNEREE), Cadi Ayyad University, Marrakech, Morocco.
E-mail address: [email protected] (L. Mandi).
Contents lists available at ScienceDirect
Journal of Water Process Engineering
journal homepage: www.elsevier.com/locate/jwpe
https://doi.org/10.1016/j.jwpe.2024.105684
Received 28 November 2023; Received in revised form 16 June 2024; Accepted 17 June 2024
(TF) is a process, in which the problems of biomass adhesion and aeration are rarely observed [2,7]. Climatic variations, wastewater composition and intrinsic physicochemical properties are some of the factors affecting the process. TFs have been in use for more than a century for the biological treatment of wastewater. According to Godoy- Olmos et al. [8], a TF is a non-submerged fixed film biological reactor that uses a rock or plastic medium over which wastewater is continu- ously distributed. They were widely used in Poland in the 1970s and 1980s to treat food processing and municipal wastewater [9]. Trickling filters are characterized by lower energy consumption [2,7,8]. They can be classified into coarse filtration, carbon oxidation, nitrification and tertiary nitrification [2,7]. The value of the indicator kg BOD5•m−•3d−1 is used to categorize TF as low, medium or high load. A high load is 0.65 to 3.2 kg BOD5•m−•3d−1, and a low load is 0.07 to 0.22 kg BOD5•m−•3d−1 [8]. The surface hydraulic loading parameter, expressed as m3•m-2•d−1 [10,11], is also used in the classification of TFs.
For wastewater managers, prediction can be a useful approach. It is used to understand the relationships between input and output variables and to predict a given output. Multiple linear regression (MLR) has many
advantages, including mathematical simplicity, the ability to estimate input variable coefficients, and the ability to determine how they affect output variable variation [12]. Therefore, by using the MLR model to predict the coliform content in the output of the TF system, we can not only monitor the performance of the TF system, but also address the complex relationships between multiple variables within the TF system.
This approach can improve the overall treatment efficiency of the sys- tem. It will also allow us to predict the coliform content of the TF effluent, which can potentially be reused for agricultural irrigation purposes. The city of Imintanout (Morocco) suffers from water scarcity.
This is due to its location in a desert environment with an arid climate.
This has led the authorities to focus on wastewater treatment, as it offers the double benefit of protecting groundwater resources and creating a new source of water, i.e. treated wastewater that can be reused in agriculture.
The objectives of this work are to: 1) evaluate the treatment effi- ciency and behavior of a full-scale TF plant in removing organic matter, nutrients, and fecal bacteria; 2) investigate the effect of seasonality on TF performance in an arid climate using Principal Component Analysis
Fig. 1.Presentation of the full-scale trickling filter system at Imintanout city (Morocco).
(PCA); 3) predict the fecal coliform concentration in TF effluent using the Multiple Linear Regression (MLR) technique and identify the rele- vant factors that can more accurately predict fecal coliform output in the TF system.
2. Material and methods
2.1. Description of the hybrid trickling filter plant
The current wastewater treatment plant is in operation since 2018. It serves the municipality of Imintanout (Morocco) with a population of 31,000. It is composed of a preliminary treatment consisting of a coarse screen followed by a fine screen and then two corridor grit chambers.
The primary treatment consists of three primary circular decanters with radial flow and sludge removal via a scraper bridge. The biological treatment consists of two circular trickling filters (TF) equipped with a rotary distributor. The full-scale TFs are identical in size (13 m ×2 m) and are filled with a random type of plastic media with a specific surface area of 90 m2 and a void ratio of 95 %. Two secondary circular and radial flow decanters have also been designed. The sludge is stabilized in three upper cylindrical digesters. These are coupled to a truncated conical lower section. The sludge is dewatered in twelve drying beds with a total surface area of 1100 m2. The WWTP was continuously fed with 1720 m3/day of HLR. Fig. 1 shows the layout of the WWTP.
2.2. Water sampling and analyses
Raw and treated effluents were collected from the experimental site in sterilized glass bottles every three months for five years. All samples were stored at 4 ◦C for bacterial and physicochemical analysis. The WTW multi 340i/multiparameter set probe (WTW Büro-Weilheim, Germany) was used for in situ analysis. The chemical oxygen demand (COD) was measured by the dichromate open reflux method [13], while the biochemical oxygen demand (BOD5) was analyzed by the Warburg method. The filtration method was used to quantify the total suspended solids (TSS) content, the indophenol technique was used to measure the NH4+concentration, and the cadmium‑copper column [14] was used to determine the NO3− concentration. Potassium peroxodisulfate digestion was used to determine total phosphorus (TP) [15]. For bacterial anal- ysis, lactose-2, 3, 5-triphenyl tetrazolium chloride TTC (Panreac, Spain) with Tergitol agar (HiMedia, India) was used to determine total co- liforms (TC) at 37 ◦C and fecal coliforms (FC) at 44 ◦C [16].
2.3. Ratio COD/BOD5
The COD/BOD5 ratio of the raw wastewater is 2.08, as shown in Table 1. The typical urban domestic wastewater is similar to this, with COD/BOD5 generally <3 [17]. Therefore, it can be concluded that the studied urban wastewater, although carrying a significant organic load, may contain only a fraction of organic matter that is readily biode- gradable [23,57]. This suggests that although biological treatment methods such as trickling filters can be effective, their efficiency may be limited by the proportion of biodegradable organic matter present.
2.4. Statistical analysis
The t-test is used to investigate significant differences. Then, the experimental data were analyzed and the influence of seasons on the
removal of pollutants was compared. The Principal Component Analysis (PCA) method was used to explore the experimental data set [58]. PCA was performed using the R programming language with the prcomp function and the factoextra package for visualization. Multiple linear regression (MLR) was used to establish relationships between fecal coliform (FC) concentration and the input variables, as well as to predict its level in the WWTP effluent. The MLR method was implemented using the lm() function. In addition, statistical metrics, including R-squared with a confidence level of 95 %, were calculated using the stat_cor () function from the ggpubr package, while the stat_regline_equation () function was used to add the regression equation.
3. Results and discussion 3.1. In situ parameters
Fig. 2 summarizes the five years of monitoring of the physicochem- ical characteristics under study, including DO, pH, EC and temperature.
Dissolved Oxygen (DO) is a key parameter used to control aerobic bio- logical processes. Its variation in the trickling filter (TF) system has a significant effect on the removal of various pollutants. The results showed that the DO concentrations were between 0.01 ±0.01 and 0.07
±0.02 mg L−1 at the inlet of the WWTP, while the DO scatter at the outlet was between 3.03 ±1.01 and 8.16 ±1.02 mg L−1 (Fig. 2). This increase in DO concentration at outlet of TFs could have been caused by aeration system installed at bottom of filter. Similar results were observed by La Motta et al. [18]. pH is also one of the most important control factors in aerobic treatment systems.
In order to improve the removal of pollutants, the pH values can possibly be used for the control of the TF aeration rate. The pH values between the inlet (7.65 ±0.13) and the outlet (7.1 ±0.28) of the TF system showed a significant change, as shown in Fig. 2. The pH remained in a suitable range for microorganism growth and nutrient adsorption (6.84–8.73) throughout the experiment. Nitrification pro- cesses within the TF system may be responsible for the decreasing pH at the system outlet [19].
The average values of electrical conductivity (EC) in the influent and effluent of the TF plant were between (1870 ±201 and 2730 ±222 μS/
cm) and (1410 ±220 and 2165 ±227 μS/cm), respectively (Fig. 2). In the influent, the degradation of organic matter by bacteria contributes to the production of salts and the increase of EC. According to Aziz and Ali [20], the decrease in EC was found to increase with increasing filter depth for the TF system. The obtained results showed satisfactory reduction efficiencies of EC, which were in compliance with the Moroccan limits for the quality of irrigation water (12,000 μS/cm).
3.2. Organic matter removal
Fig. 3 shows the evolution of total suspended solids (TSS) during the experimental period. The TSS concentrations at the inlet of the system were variable and ranged between 750 ±21.9 and 835 ±19.05 mg L−1, while the effluent TSS ranged between 9 ±1.05 and 45 ±1.01 mg L−1. The investigated system supports the removal of TSS up to 98 % and manages to produce a clear effluent with a high level of suspended solids' removal. Trickling filter systems have been shown to meet quality standards in systems using raw wastewater as a source of microorgan- isms in the filter medium [21]. The reduction in TSS is due to the retention of solid particles by the biofilm that develops in the pores of the TF media, thereby reducing the amount of solids in the effluent.
Biofilm can withstand the speed of the wastewater, creating an inter- action between the TSS and the microorganisms contained in the bio- film, where filtration begins, trapping and binding the suspended materials, thereby reducing the TSS [26]. This fact is consistent with that reported by Agustina et al. [22] in their research on wastewater treatment using biofilters to reduce BOD5, COD, TSS and pH. In addition, statistical analysis was performed and showed a significant difference (p Table 1
Basic characteristics of the raw domestic wastewater.
Parameter Unit Min Max Mean
COD mg L−1 1290.00 1554.00 1352.00
BOD5 mg L−1 620.00 900.00 650
COD/BOD5 – – 2.08
Fig. 2. Evolution of DO, pH, EC, and Temp ◦C during the treatment period by the trickling filter plant. Symbol **** p <0.0001 indicates statistical significance.
Fig. 3. Evolution of TSS, BOD5 and COD concentrations at input and output of the trickling filter plant. Symbol **** p <0.0001 indicates statistical significance.
<0.001) in the removal of TSS at the outlet of the TF system.
The COD concentration decreased from 1352 ± 104.86 to 75 ± 27.72 mg L−1 and the mean removal was 94 %. Furthermore, the BOD5
content was variable, ranging from 620 ±90 to 1532 ±102 mg L−1. The concentration of BOD5 was removed by approximately 98 % during the stable phase of the experience. The better removal efficiency was due to the gradual increase of wastewater load feeding the system, unique design, and optimal operating conditions under which the study was conducted. For example, the high reduction in COD is due to the high concentration of microorganisms and the similarity of the microor- ganism sources to the treated wastewater, so that biofilm formation is better. In the biofilm layer, the organic compounds are degraded by the aerobic microorganisms, and then the COD value is reduced. Kornaros &
Lyberatos [23] and Bouchelkia & Belarbi [24] reported that the high removal efficiency is directly related to the presence of a mixture of microorganisms that effectively oxidize the organic compounds under good aeration conditions provided by air stripping at the inlet. La Motta et al [18] reported that 30-60% of the total COD removal was due to air stripping caused by the air supply at the bottom of the filter. The
remaining COD was clearly removed by biological action. COD measures the total amount of oxidizable organic matter in water, while BOD5 measures the amount of biodegradable organic matter over a 5-day period. COD therefore includes organic compounds that are not neces- sarily biodegradable under the specific conditions of the BOD5 test.
Some organic compounds present in wastewater may be oxidized but not necessarily biodegradable within five days, which explains why COD removal may be higher than that of BOD5 [59,60]. In addition, trickling filters allow wastewater to be exposed to bacteria for a longer period of time, which may result in more complete degradation of organic matter [61]. Furthermore, the quality of the treated effluent in terms of COD and BOD5 concentrations met the Moroccan wastewater discharge standards (COD: 250 mg L−1; BOD5: 120 mg L−1) [43].
3.3. Nitrogen and total phosphorus removal
The evolution of the influent and effluent NO3− concentrations is shown in Fig. 4. The influent NO3− concentration varied from 0.01 ± 0.001 to 0.26 ±0.01 mg L−1. On the other hand, the average nitrate
Fig. 4.Evolution of NH4+, NO3− and TP concentration at input and output of the trickling filter plant. Symbol **** p <0.0001 indicates statistical significance.
concentration in the effluent varied between 46.02 ±0.5 and 253.01 ± 0.10 mg L−1. It was higher and did not meet the Moroccan standards of water quality for irrigation (30 mg L−1). The presence of aerobic zones has a significant effect on the removal mechanisms of NO3− within the TF system [27–29]. In other studies it was reported that a low organic load in the raw wastewater hinders the denitrification process during the treatment. The unavailability of a carbon source for denitrifying bacteria resulted in an increase of nitrate concentration in the treatment water [28,30]. Similarly, in the vertical flow constructed wetland combined with trickling filter (CW-TF), the lack of anoxic zones hinders the denitrification process. This explains the poor elimination of nitrate within the CW-TF [25]. Schipper et al. [31] and Rambags et al. [32]
reported that the majority of denitrifying bacteria are heterotrophic and use organic matter as a carbon source for their growth. It can be concluded that high nitrification was achieved by the TF system. How- ever, the design of the system and the operating conditions were not conducive to achieving good nitrate removal.
The phosphorus concentrations measured in the influent and effluent of the trickling filter are shown in Fig. 4. The influent concentrations ranged from 12.2 ±0.69 mg L−1 to 19.36 ±0.45 mg L−1, while the effluent concentrations ranged from 5 ±0.4 mg L−1 to 12.2 ±0.3 mg L−1. The performance of the system in terms of phosphorus removal reached 41 %. This was due to a combination of adsorption and pre- cipitation phenomena with calcium (Ca), aluminum (Al), iron (Fe) and other clay minerals present in the system. Kim et al. [25] and Mis- bahuddin & El-Rehaili [33] reported that the removal of phosphorus is explained by the retention of phosphorus-bearing particles by filtration and by the potential sorption or precipitation of P species in solution due to the oxygenation of the wastewater when it is injected at the bottom of the TF system. The removal of phosphorus by these phenomena has also been reported in the literature review for systems such as multi-soil- layering, sand filter and constructed wetlands [6,34–36]. Due to the biofilm development in the trickling filter, it was likely that some phosphorus was partially retained in the trickling filter in the form of microbial biomass. In the long term, excess biomass was probably being removed from the trickling filter, but this was irregular and could not be fully accounted for by the monitoring protocol [25,37].
3.4. Microbial parameters removal
Fig. 5 shows the average levels of bacterial indicators of fecal contamination obtained at the inlet and outlet of the WWTP. In the influent, the average concentration of FC was between 6.36 ±0.02 and 6.69 ±0.07 log units, while in the effluent, it has reached values be- tween 4.26 ±0.03 and 4.70 ±0.03 log units, showing a reduction of 2.88 log units. The results obtained showed that the trickling filter system achieved a moderate removal of FC. Accordingly, the average concentration of FC in the final effluent quality was higher than the acceptable concentration limits for irrigation recommended by Moroc- can standards (3 log units). The central pathways for coliform removal
in the TF system are physical filtration, adsorption, and other mecha- nisms, including predation and microbial cell death (die-off). The di- versity of the biofilm bacterial community is directly related to the accumulation of nutrients in the system. In this regard, increasing wastewater residence time plays a critical role in reducing bacteria [38,39]. Many research studies suggest that filtration, adsorption, decomposition, and the natural death of bacteria are the main paths for FC elimination [38]. Protozoa, rotifers, nematodes, and phage-feeding bacteria could play an important role in pathogen removal in con- structed wetlands and multi-soil-layering systems [6,41,42]. The coli- form population in biofilms may be regulated by Bdellovibrio and Ensifer adhaerens. The phenomenon of natural mortality can also have a sig- nificant impact on the ability of the system to reduce coliform levels [40,42,52]. On the other hand, many additional parameters, including effluent quality and temperature, affect the ability to eliminate bacteria [36,40].
3.5. Statistical analysis of seasonal effects on trickling filter plant efficiency
By identifying individuals based on the removal dataset, Principal Component Analysis (PCA) was used to investigate the effect of season on pollutant removal efficiency (Fig. 6). The two significant dimensions [Dim1 (68.5 %) and Dim2 (13.2 %)] accounted for 81.7 % of the vari- ation in pollutant removal. Thus, Fig. 6a shows the squared cosines for dimensions 1 and 2, which represent the quality of the representation of the variables on the principal components. A higher squared cos2 (close to 1) indicates that the principal component represents the variable well.
For example, most of the variables studied have a high squared cos2 for Dim1. Furthermore, we can see that BOD5 has a relatively high loading on Dim1 and NH4+has a high loading on Dim2, indicating that these variables contribute significantly to the variation captured by these respective principal components. Similarly, we can see that some ob- servations (Fig. 6b) have higher values for Dim.1 and Dim.2. This means that Dim.1 and Dim.2 capture a significant proportion of the variance in the experimental data.
Seasons such as summer and spring had a positive effect on Dim1 and were closely associated with each specific parameter: HLR, Temp, pH, BOD5, COD, TP, and FC (Fig. 6c). However, a negligible relationship was found with Dim2. The PCA cos2 plot also shows the average difference between the spring and summer seasons in terms of the variability of the distribution of HLR, T, pH, BOD, COD, TP, and FC.
It was shown that HLR, BOD5, and FC all had high cos2 values ranging from 0.86 to 0.96, indicating that they are significant seasonal variables. The significant effects of the two seasons on HLR and the removal of BOD5 and FC can be used to explain these data, indicating that HLR and the methods used to remove organic particles and bacterial indicators in TF systems are likely seasonal. The concentration of COD and FC showed a significant relationship. In the presence of organic matter, the fecal bacterial indicator grows significantly faster [40].
Fig. 5. Evolution of FC concentration at input and output of the trickling filter plant. Symbol **** p <0.0001 indicates statistical significance.
Furthermore, a significant effect of organic matter on the abundance of E.coli was observed by Bouteleux et al. [45]. However, it may be difficult for FC bacteria to survive due to the high removal rate of organic matter in the TF system. In addition, Arora & Kazmi [46] showed how the warm season enhances the decomposition of organic matter by increasing the metabolic activity of bacterial cells. The TF system removed organic matter very well and did not clog during operation.
The structure and design of TF materials maintain high water permeability, which reduces the potential for clogging [47]. CW, sand filters and lagoons are other unconventional technologies with known seasonal effects on pollutant removal. Ouellet-Plamondon et al. [48]
found that COD reduction in CW was slightly better in summer than in winter. Seasonality is known to play a significant role in the removal of coliforms and pathogens in wastewater treatment processes [34,40].
Furthermore, FC bacteria was the variable most affected by spring and summer variations (cos2 >0.86).
This result indicates that microbial degradation was the primary FC removal mechanism in the TF system. El Hamouri et al. [49] found that high rate algal pond method resulted in higher coliform removal during summer in arid region. Higher temperatures in the CW system improve the removal of coliform indicator bacteria [50]. TSS shows negative contributions to Dim1 (cos2 >0.85), but only a negligible relationship with Dim2. There is a good correlation with the six individuals during fall and winter and summer (individual PCA). The reason for this finding was the highest TSS concentration ever recorded in municipal waste- water during this period.
3.6. Multivariate linear relationship
A comprehensive analysis of the linear relationship was performed to identify the primary variables that significantly impact FC output, with a 95 % confidence level. The results indicate that the parameters DO, NO3−
and TSS_in are particularly influential in the effective reduction of FC levels in the TF system. As shown in Fig. 7, these variables can be used to
explain the FC removal rate.
The relationship between DO concentration and FC removal is highlighted in Fig. 7. This suggests that an aerobic environment would be more suitable for FC removal in a TF system. This conclusion is supported by other studies, such as that of Speitz et al. [51], which showed significant relationships between bacterial mortality and DO levels in the aerobic zone. Better conditions for such predator commu- nities were provided by the improved aeration of the gravel layers in the vertical flow multi-soil-layering system [34,35]. Predator communities would benefit from improved aeration. Dissolved oxygen can affect the survival of microorganisms within the lagoon system [4]. Similarly, a significant relationship between DO and bacterial mortality in the aquatic environment has been reported [52]. Based on the MLR model, FC and NO3− are interrelated in the TF system. Denitrifying microor- ganisms are responsible for denitrification by converting NO3− to Fig. 6. Effect of Seasonal Variations on TF Plant Removal Effectiveness: A Principal Component Analysis (PCA) Perspective. a) Variables-PCA shows the relationships between the variables and the principal components. (b) Individuals-PCA shows the projection of individual data points onto the principal components to indicate similarities or differences among the observations. (c) PCA Biplot visualizes the relationships between variables and observations in the data set across seasons.
Fig. 7.Linear relationship plot depicting the relationship between the relevant input variables and the level of FC in TF effluent. The x-axis represents the input variables concentration, and the y-axis represents the FC_out level. Data points are represented by colored dots, with the color gradient indicating the FC_out level (from orange to purple). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
dinitrogen gas N2 in the TF system [27–29].
According to Rambags et al. [32], denitrifying bioreactors can effectively minimize microbial pollutants in wastewater by reducing N loading. In other situations, the prevalence of microeukaryotes, such as protozoa, in wastewater treatment results in clearer effluent [53]. Pro- tozoa accelerate nutrient recycling and contribute to organic floccula- tion. According to Laurin et al. [54], protozoa favorably affect the ability of the denitrification system to produce cleaner effluent. The reduction of total coliforms, fecal coliforms, and fecal streptococci (FS) in the TF system depended on the physicochemical characteristics of the porous environment, such as filtration and bacterial adsorption processes, fol- lowed by microbial degradation and natural death (die-off). The pro- tozoan species found in the substrate may be crucial in the bacterial predation of municipal wastewater [55]. The significant relationship between coliform concentrations and TSS for the TSS predictors, as shown by Morgan et al. [56], suggests that mineralization and sedi- mentation are important factors influencing coliform levels in natural methods of treating domestic wastewater. TSS, organic debris, and nu- trients have been shown to inhibit the adsorption of bacterial cells in a porous environment by competing for adsorption sites [40].
3.7. Multiple linear regression prediction
In this study, the Multiple Linear Regression (MLR) method was implemented using the lm() function in the R programming language.
This method serves as a reliable benchmark when the assumptions are true and there is a linear relationship between the variables that are being studied. However, if the assumptions are violated or there is a nonlinear relationship, more complex models may have better perfor- mance than MLR [62]. Once the significant input variables affecting FC removal in the TF system were identified, the Multiple Linear Regression (MLR) model was used to predict the variation in TF sanitation effi- ciency. The performance of the MLR model was evaluated by comparing the predicted and actual values using the R2 metric. The results indi- cated a strong linear relationship between the actual and predicted TF performance, with a correlation coefficient of >0.77, demonstrating this relationship at a 95 % confidence level. This high correlation is an indication that the MLR model developed in this study is suitable for the prediction of TF content.
Fig. 8 shows the regression analysis of the MLR model, which further supports its effectiveness in predicting FC removal. As a result, the proposed MLR model can be used as a valuable tool to evaluate the
behavior of the TF system in the face of FC contamination.
In addition, the MLR model has successfully predicted the FC con- centration at the outlet of the system. This provides valuable insights for future TF efficiency management. Therefore, MLR serves as a simple and reliable model to interpret the effect of input variables on the output variable through its coefficients.
3.8. Water quality and environmental implications
Results of this study showed that a hybrid TF plant effectively removes organics, nutrients, and fecal coliforms from municipal wastewater. In addition, during the five-year monitoring period, no obvious problems (such as clogging, odor, insects, etc.) were observed.
Table 2 shows that post-treatment pH, EC and TSS values are in the high range of limits allowed for direct discharge in Morocco [43].
However, the average nitrate concentration in the effluent was higher and was not in compliance with the Moroccan irrigation water quality standards (30 mg L−1) [44]. The lack of denitrification may be attributed to the absence of anaerobic conditions and the short residence time in the hybrid trickling filter system [63,64]. To meet stringent effluent limitations, additional treatment may be required to provide sufficient denitrification capacity using technologies such as subsurface horizontal flow MSLs or constructed wetlands (CWs). These systems are predominantly anaerobic due to limited oxygen transport to saturated media [6,34], and have demonstrated denitrification capacity [6,65,66].
On the other hand, the average concentration of fecal coliforms (FC) in the final effluent was higher than the acceptable concentration limits for irrigation recommended by the Moroccan standards (3 log units) [44]. According to the literature, trickling filters are ineffective in removing fecal bacteria from municipal wastewater due to limited contact time between wastewater and microbial biofilms, the prevalence of aerobic conditions that do not support fecal bacteria removal, incomplete treatment of solids that contain fecal bacteria, a lack of predation or competition on fecal bacteria, and microbial communities that prefer to break down organic matter rather than fecal bacteria removal under high organic loading conditions [67]. In order to achieve a water quality that can be safely used for irrigation without posing a sanitary risk, it is necessary to implement complementary treatment processes to improve the removal of fecal coliforms and pathogens.
4. Conclusion
This study focuses on investigating a large-scale hybrid trickling filter (TF) technology for treating urban wastewater in an arid climate.
The effectiveness of the hybrid TF system was confirmed, achieving
Fig. 8. Performance of MLR model in prediction of FC level based on com- parison of predicted and observed removal of FC in TF system (Period covered 5 years).
Table 2
Level recommended for treated wastewater reuse in irrigation.
Variable Unit Influent Effluent Admissible limits for direct discharge [43]
Admissible limits for wastewater Reuse [44]
pH unit 7.65 ±
0.33 8 ±0.37 – 6.5–8.4
EC mS/
cm 2219 ±
220 1939 ±
190 – 12,000
TSS mg
L−1 790 ±
30.97 12.02 ±
1.75 150 100
BOD5 mg
L−1 650 ±
13.05 10.50 ±
0.50 120 –
COD mg
L−1 1352 ±
56.77 75.04 ±
13.70 250 –
NO3-N mg
L−1 0.08 ±
0.01 140.03
±12.09 – 30
FC log
CFU/
100 mL
6.57 ±
0.70 4.32 ±
0.60 – 3
significant removal rates for TSS, BOD5, COD, TP, and NH4+at a 95 % confidence level. Seasonal variations had a significant impact on the performance of the TF, especially in the summer and spring, with vari- ables such as HLR, BOD5, and fecal coliforms showing high cos2 values.
The positive correlation between the TSS concentration and the sea- sonality implies that the composition of the wastewater varies throughout the year.
Key parameters were identified as significant contributors to the removal of fecal coliform (FC) in the TF system, including DO, NO3−, and TSS. The developed Multiple Linear Regression (MLR) model, with an R2 value of 0.77, shows promise in predicting FC removal in similar wastewater treatment plants, although further validation in different contexts is needed to ensure its reliability.
Despite compliance with Moroccan standards for direct discharge, treated municipal wastewater does not meet safe reuse criteria for irri- gation due to high nitrate levels above the allowable limit of 30 mg L−1 and fecal coliform levels above the allowable limit of 3 log units. This makes upgrading the wastewater treatment plant (WWTP) a challenge for the Imintanout community. Additional treatment measures are essential in order to achieve a water quality that meets Moroccan standards for the irrigation of agricultural crops and landscaping.
CRediT authorship contribution statement
Ahmed Osmane: Writing – original draft, Validation, Methodology, Investigation, Conceptualization. Khadija Zidan: Writing – original draft, Visualization, Investigation, Conceptualization. Rabia Benaddi:
Writing – original draft, Visualization, Investigation. Sofyan Sbahi:
Writing – original draft, Software, Methodology, Formal analysis.
Naaila Ouazzani: Writing – review & editing, Validation, Methodology.
Moustapha Belmouden: Writing – review & editing, Supervision, Project administration, Conceptualization. Laila Mandi: Writing – re- view & editing, Writing – original draft, Validation, Supervision, Re- sources, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
Data will be made available on request.
References
[1] K. Belhassan, Water scarcity Managementin the Maghreb Region’, drought-impacts and management, IntechOpen (2022), https://doi.org/10.5772/
intechopen.103788. Sep.07.
[2] G.T. Daigger, J.P. Boltz, Trickling filter and trickling filter-suspended growth process design and operation: a state-of-the-art review, Water Environ. Res. 83 (5) (2011) 388–404, https://doi.org/10.2175/106143010x12681059117210.
[3] G. Di Bella, et al., Occurrence of microplastics in waste sludge of wastewater treatment plants: comparison between membrane bioreactor (MBR) and conventional activated sludge (CAS) technologies, Membranes (Basel). 12 (4) (2022), https://doi.org/10.3390/membranes12040371.
[4] A. Osmane, K. Zidan, R. Benaddi, I. Sebah, Nadia Elmouraille, M. Ait Ali, M. Belmouden, Study the effectiveness of the natural lagoons for removal of organic matter, nutrients and fecal coliform from urban domestic wastewater under arid climate (Morocco), J. Ecol. Eng. 24 (2) (2023) 196–207, https://doi.
org/10.12911/22998993/156518.
[5] S. Waqas, N.Y. Harun, M.R. Bilad, T. Samsuri, N.A.H.M. Nordin, N. Shamsuddin, A.
B.D. Nandiyanto, N. Huda, J. Roslan, Response surface methodology for optimization of rotating biological contactor combined with external membrane filtration for wastewater treatment, Membranes 12 (2022) 271, https://doi.org/
10.3390/membranes12030271.
[6] K. Zidan, L. Mandi, A. Hejjaj, S. Sbahi, A. El, A. El, N. Ouazzani, A. Assabbane, Efficiency of a new hybrid multi-soil-layering eco-friendly technology for removing pollutants from domestic wastewater under an arid climate, J. Water Process Eng.
51 (2023) 103482, https://doi.org/10.1016/j.jwpe.2022.103482.
[7] C.D. Henrich, M. Marggraff, Energy-efficient wastewater reuse– the Renaissance of trickling filter technology, in: Proc. 9th Int. Conf. Water Reuse, 2013, pp. 27–31 [Online]. Available: http://www.expoval.de/sites/default/files/download/hen rich_marggraff_wastewater_reuse_by_trickling_filters_iwa_water_reuse_conference_
2013_in_namibia.pdf.
[8] S. Godoy-Olmos, S. Martínez-Llorens, A. Tom´as-Vidal, M. Jover-Cerd´a, Influence of filter medium type, temperature and ammonia production on nitrifying trickling filters performance, J. Environ. Chem. Eng. 4 (1) (2016) 328–340, https://doi.org/
10.1016/j.jece.2015.11.023.
[9] R.M. Zyłka, W. Dabrowski, E. Gogina, O. Yancen, Trickling filter for high efficiency treatment of dairy sewage, J. Ecol. Eng. 19 (4) (2018) 269–275, https://doi.org/
10.12911/22998993/89657.
[10] S. Ambika, Chapter 7 - integrated sustainability impact assessment of trickling filter, Editor(s): Thendiyath Roshni, Pijush Samui, Dieu Tien Bui, Dookie Kim, Rahman Khatibi, Risk, Reliability and Sustainable Remediation in the Field of Civil and Environmental Engineering, Elsevier, 2022, pp. 93–109, ISBN
9780323856980, doi:https://doi.org/10.1016/B978-0-323-85698-0.00003-4.
[11] Z. Dymaczewski, E.S. Kempa, M.M. Sozanski, Coagulation as a structure-forming separation process in water and wastewater treatment, Water Sci. Technol. 36 (4) (1997) 25–32, https://doi.org/10.1016/S0273-1223(97)00415-0.
[12] W. Luo, et al., Preliminary study on the application of near infrared spectroscopy and pattern recognition methods to classify different types of apple samples, Food Chem. 128 (2) (2011) 555–561, https://doi.org/10.1016/j.
foodchem.2011.03.065.
[13] APHA, Standard Methods for the Examination of Water and Wastewater, 21st ed., American Public Health Association/American Water Works Association/Water Environment Federation, Washington, DC, USA, 2005. Int. J. Adv. Agric. Res.
[14] J. Rodier, L’analyse de l’eau: eaux naturelles, eaux r´esiduaires, eau de mer, 7`eme
´edition, Dunod, Bordas, Paris, 2009, p. 1365.
[15] AFNOR, Recueil de normes françaises: eau, m´ethodes d’essai, 2`eme ´edition, 1997.
Paris.
[16] Moroccan Standards, Moroccan Standard Approved by Order of the Minister of Industry, Trade and Economy Last Level, Moroccan Industrial Standardization Service, Rabat, 2006.
[17] M. Allaoui, E.L.H. Mohati, M. Saadallah, A. El Harfi, Performance ´epuratoire de STEP de Dar El Gueddari (Lagunage naturel) [STEP Purifying Performance of Dar El Gueddari (Natural Lagoons)] 16, 2016, pp. 704–717, no. 4.
[18] E.J. La Motta, J.A. Jim´enez, J.C. Josse, A. Manrique, The effect of air-induced velocity gradient and dissolved oxygen on bioflocculation in the trickling filter/
solids contact process, Adv. Environ. Res. 7 (2) (2003) 441–451, https://doi.org/
10.1016/S1093-0191(02)00011-4.
[19] D. Nourmohammadi, M.B. Esmaeeli, H. Akbarian, M. Ghasemian, Nitrogen removal in a full-scale domestic wastewater treatment plant with activated sludge and trickling filter, J. Environ. Public Health 2013 (2013), https://doi.org/
10.1155/2013/504705.
[20] S.Q. Aziz, S.M. Ali, Treatment of synthetic dairy wastewater using disposed plastic materials as trickling filter media: optimization and statistical analysis by RSM, Adv. Environ. Biol. 13 (10) (2019) 1–16, https://doi.org/10.22587/
aeb.2019.13.10.1.
[21] C. Joel, K. Ezekiel, L.A. Mwamburi, Effects of seasonal variation on performance of conventional wastewater treatment system, J. Environ. Earth Sci. 5 (1) (2021) 1–7, https://doi.org/10.7176/jees/11-7-06.
[22] A. Agustina, I.E. Suprihatin, J. Sibarani, Pengaruh Biofilm Terhadap Efektivitas Penurunan Bod, Cod, Tss, Minyak Dan Lemak Dari Limbah Pengolahan Ikan Menggunakan Trickling Filter, Cakra Kim. 4 (2) (2017) 137–145.
[23] M. Kornaros, G. Lyberatos, Biological treatment of wastewaters from a dye manufacturing company using a trickling filter, J. Hazard. Mater. 136, no. 1 SPEC.
ISS (2006) 95–102, https://doi.org/10.1016/j.jhazmat.2005.11.018.
[24] H.S. Hamid Bouchelkia, Fadila Belarbi, Etude des performances d’un lit bact´erien classique `a garnissage en pouzzolane de Beni Saf, no. October 2014, 2013, p. 185.
[25] B. Kim, M. Gautier, S. Prost-Boucle, P. Molle, P. Michel, R. Gourdon, Performance evaluation of partially saturated vertical-flow constructed wetland with trickling filter and chemical precipitation for domestic and winery wastewaters treatment, Ecol. Eng. 71 (2014) 41–47, https://doi.org/10.1016/j.ecoleng.2014.07.045.
[26] E.H. Eding, A. Kamstra, J.A.J. Verreth, E.A. Huisman, A. Klapwijk, Design and operation of nitrifying trickling filters in recirculating aquaculture: a review, Aquac. Eng. 34 (3) (2006) 234–260, https://doi.org/10.1016/j.
aquaeng.2005.09.007.
[27] S. Biesterfeld, G. Farmer, L. Figueroa, D. Parker, P. Russell, Quantification of denitrification potential in carbonaceous trickling filters, Water Res. 37 (16) (2003) 4011–4017, https://doi.org/10.1016/S0043-1354(03)00302-6.
[28] R. Kanda, N. Kishimoto, J. Hinobayashi, T. Hashimoto, Effects of recirculation rate of nitrified liquor and temperature on biological nitrification–denitrification process using a trickling filter, Water Environ. J. 30 (3–4) (2016) 190–196, https://
doi.org/10.1111/wej.12196.
[29] Y. Zhang, Y. Cheng, C. Yang, W. Luo, G. Zeng, L. Lu, Performance of system consisting of vertical flow trickling filter and horizontal flow multi-soil-layering reactor for treatment of rural wastewater, Bioresour. Technol. 193 (2015) 424–432, https://doi.org/10.1016/j.biortech.2015.06.140.
[30] W. Luo, C. Yang, H. He, G. Zeng, S. Yan, Novel two-stage vertical flow biofilter system for efficient treatment of decentralized domestic wastewater, Ecol. Eng. 64 (2014) 415–423, https://doi.org/10.1016/j.ecoleng.2014.01.011.
[31] L.A. Schipper, W.D. Robertson, A.J. Gold, D.B. Jaynes, S.C. Cameron, Denitrifying bioreactors-an approach for reducing nitrate loads to receiving waters, Ecol. Eng.
36 (11) (2010) 1532–1543, https://doi.org/10.1016/j.ecoleng.2010.04.008.
[32] F. Rambags, C.C. Tanner, R. Stott, L.A. Schipper, Bacteria and virus removal in denitrifying bioreactors: effects of media type and age, Ecol. Eng. 138 (June) (2019) 46–53, https://doi.org/10.1016/j.ecoleng.2019.06.021.
[33] M. Misbahuddin, A. El-Rehaili, Phosphorus removal from trickling filter effluents by fly ash, J. King Saud Univ. - Eng. Sci. 7 (2) (1995) 185–197, https://doi.org/
10.1016/S1018-3639(18)30626-3.
[34] K. Zidan, S. Sbahi, A. Hejjaj, N. Ouazzani, A. Assabbane, L. Mandi, Removal of bacterial indicators in on-site two-stage multi-soil-layering plant under arid climate (Morocco): prediction of total coliform content using K-nearest neighbor algorithm, Environ. Sci. Pollut. Res. 0123456789 (2022), https://doi.org/
10.1007/s11356-022-21194-x.
[35] C.J. Wei, W.Z. Wu, Performance of single-pass and by-pass multi-step multi-soil- layering systems for low-(C/N)-ratio polluted river water treatment, Chemosphere 206 (2018) 579–586, https://doi.org/10.1016/j.chemosphere.2018.05.035.
[36] S. Cao, W. Chen, Z. Jing, Phosphorus removal from wastewater by fly ash ceramsite in constructed wetland, Afr. J. Biotechnol. 11 (16) (2012) 3825–3831, https://doi.
org/10.5897/ajb11.2719.
[37] H. Kodera, M. Hatamoto, K. Abe, T. Kindaichi, N. Ozaki, A. Ohashi, Phosphate recovery as concentrated solution from treated wastewater by a PAO-enriched biofilm reactor, Water Res. 47 (6) (2013) 2025–2032, https://doi.org/10.1016/j.
watres.2013.01.027.
[38] I. Naz, D.P. Saroj, S. Mumtaz, N. Ali, S. Ahmed, Assessment of biological trickling filter snystems with various packing materials for improved wastewater treatment, Environ. Technol. (United Kingdom) 36 (4) (2015) 424–434, https://doi.org/
10.1080/09593330.2014.951400.
[39] A. Schwager, M. Boller, Transport phenomena in intermiitent filters, Waf. Sc. TIt 35 (97) (1997) 13–20, https://doi.org/10.1016/S0273-1223(97)00090-5.
[40] T.K. Stevik, K. Aa, G. Ausland, J.F. Hanssen, Retention and removal of pathogenic bacteria in wastewater percolating through porous media: a review, Water Res. 38 (6) (2004) 1355–1367, https://doi.org/10.1016/j.watres.2003.12.024.
[41] J.T. Jasper, D.L. Sedlak, Phototransformation of wastewater-derived trace organic contaminants in open-water unit process treatment wetlands, Environ. Sci.
Technol. 47 (19) (2013) 10781–10790, https://doi.org/10.1021/es304334w.
[42] H. Wand, G. Vacca, P. Kuschk, M. Krü Ger, M. K¨a Stner, Removal of bacteria by Filtration in Planted and Non-planted Sand Columns, 2006, https://doi.org/
10.1016/j.watres.2006.08.024.
[43] Valeurs Limites Sp´ecifiques de Rejet Domestique, Arrˆet´e n◦1607–06 du Joumada II 1427, Secr´etariat d’Etat Aupr`es du Minist`ere de l’Energie, des Mines, de l’Eau et de l’Environnement, Charg´e de l’Eau et de l’Environnement, D´epartement de l’Eau, Royaume du Maroc (2006) 50.
[44] Normes de qualit´e des eaux destin´ees `a l’’irrigation, Secretariat d’´etat aupr`es du minist`ere de l’´energie, des mines, de l’eau et de l’environnement, charg´e de l’eau et de l’environnement, Maroc, 2007.
[45] C. Bouteleux, S. Saby, D. Tozza, J. Cavard, V. Lahoussine, P. Hartemann, L. Mathieu, Escherichia coli behavior in the presence of organic matter released by algae exposed to water treatment chemicals, Appl. Environ. Microbiol. 71 (2) (2005) 734–740, https://doi.org/10.1128/AEM.71.2.734-740.2005.
[46] S. Arora, A.A. Kazmi, The effect of seasonal temperature on pathogen removal efficacy of vermifilter for wastewater treatment, Water Res. 1 (74) (2015) 88–99, https://doi.org/10.1016/j.watres.2015.02.001.
[47] W.T. Mook, M.H. Chakrabarti, M.K. Aroua, G.M.A. Khan, B.S. Ali, M.S. Islam, M.
A. Abu Hassan, Removal of total ammonia nitrogen (TAN), nitrate and total organic carbon (TOC) from aquaculture wastewater using electrochemical technology: a review, Desalination 285 (2012) 1–13, https://doi.org/10.1016/j.
desal.2011.09.029.
[48] C. Ouellet-Plamondon, F. Chazarenc, Y. Comeau, J. Brisson, Artificial aeration to increase pollutant removal efficiency of constructed wetlands in cold climate, Ecol.
Eng. 27 (3) (2006) 258–264, https://doi.org/10.1016/j.ecoleng.2006.03.006.
[49] B. El Hamouri, K. Khallayoune, K. Bouzoubaa, N. Rhallabi, M. Chalabi, High rate algal pond performances in faecal coliforms and helminthe egg removals, Water Res. 28 (1994) 171–174, https://doi.org/10.1016/0043-1354(94)90131-7.
[50] J. Faulwetter, V. Gagnon, C. Sundberg, F. Chazarenc, M. Burr, J. Brisson, A. Camper, O. Stein, Microbial processes influencing performance of treatment wetlands: a review, Ecol. Eng. 35 (2009) 987–1004, https://doi.org/10.1016/j.
ecoleng.2008.12.030.
[51] R.L. Spietz, C.M. Williams, G. Rocap, M.C. Horner-Devine, A dissolved oxygen threshold for shifts in bacterial community structure in a seasonally hypoxic estuary, PLoS One 10 (8) (2015) e0135731, https://doi.org/10.1371/journal.
pone.0135731.
[52] A. Fern´andez, C. Tejedor, A. Chordi, Effect of different factors on the die-off of fecal bacteria in a stabilization pond purification plant, Water Res. 26 (8) (1992) 1093–1098, https://doi.org/10.1016/0043-1354(92)90145-T.
[53] C.R. Curds, The ecology and role of protozoa in aerobic sewage treatment processes, Ann. Rev. Microbiol. 36 (1982) 27–46, https://doi.org/10.1146/
annurev.mi.36.100182.000331.
[54] V. Laurin, N. Labb´e, S. Parent, et al., Microeukaryote diversity in a marine methanol-fed fluidized denitrification system, Microb. Ecol. 56 (2008) 637–648, https://doi.org/10.1007/s00248-008-9383-x.
[55] A.M. Kadam, G.H. Oza, P.D. Nemade, H.S. Shankar, Pathogen removal from municipal wastewater in constructed soil filter, Ecol. Eng. 33 (1) (2008) 37–44, https://doi.org/10.1016/j.ecoleng.2007.12.001.
[56] J.A. Morgan, A.E. Hoet, T.E. Wittum, C.M. Monahan, J.F. Martin, Reduction of pathogen indicator organisms in dairy wastewater using an ecological treatment system, J. Environ. Qual. 37 (2008) 272–279, https://doi.org/10.2134/
jeq2007.0120.
[57] X.L. Jin, M. Jing, X. Chen, Z.X. Zhuang, X.R. Wang, X.R. Wang, F.S. Lee, A study on the relationship between BOD5 and COD in coastal seawater environment with a rapid BOD measurement system, Water Sci. Technol. 60 (12) (2009) 3219–3223, https://doi.org/10.2166/wst.2009.810.
[58] A. Tharwat, Principal component analysis-a tutorial, Int. J. Appl. Pattern Recogn. 3 (3) (2016) 197–240.
[59] M. Kornaros, G. Lyberatos, Biological treatment of wastewaters from a dye manufacturing company using a trickling filter, J. Hazard. Mater. 136 (1) (2006) 95–102, https://doi.org/10.1016/j.jhazmat.2005.11.018.
[60] J.J. García-Mesa, J.M. Poyatos, F. Delgado, et al., The influence of biofilm treatment systems on particle size distribution in three wastewater treatment plants, Water Air Soil Pollut. 212 (2010) 37–49, https://doi.org/10.1007/s11270- 009-0320-5.
[61] D. Tonon, A.L. Tonetti, B. Coraucci Filho, D.A. Camargo Bueno, Wastewater treatment by anaerobic filter and sand filter: hydraulic loading rates for removing organic matter, phosphorus, pathogens and nitrogen in tropical countries, Ecol.
Eng. 82 (2015) 583–589, https://doi.org/10.1016/j.ecoleng.2015.05.018.
[62] S. Sbahi, N. Ouazzani, L. Latrach, A. Hejjaj, L. Mandi, Nitrogen modeling and performance of multi-soil-layering (MSL) bioreactor treating domestic wastewater in rural community, J. Water Proc. Eng. 13 (2021) 6–11, https://doi.org/10.1016/
j.jwpe.2021.102389.
[63] R.X. Hao, S.M. Li, J.B. Li, C.C. Meng, Denitrification of simulated municipal wastewater treatment plant effluent using a three-dimensional biofilm-electrode reactor: operating performance and bacterial community, Bioresour. Technol. 143 (2013) 178–186, https://doi.org/10.1016/j.biortech.2013.06.001.
[64] G. Tchobanoglous, H.D. Stensel, R. Tsuchihashi, F. Burton, Wastewater Engineering: Treatment and Resource Recovery, McGraw-Hill education, New York, 2003.
[65] Y. Zhang, Y. Cheng, C. Yang, W. Luo, G. Zeng, L. Lu, Performance of system consisting of vertical flow trickling filter and horizontal flow multi-soil-layering reactor for treatment of rural wastewater, Bioresour. Technol. 193 (2015) 424–432, https://doi.org/10.1016/j.biortech.2015.06.140.
[66] A.N. Bezbaruah, T.C. Zhang, Quantification of oxygen release by bulrush (Scirpus validus) roots in a constructed treatment wetland, Biotechnol. Bioeng. 89 (2005) 308–318.
[67] CIWEM, Handbooks of UK Wastewater Practice: Biological Filtration and Other Fixed-Film Processes, CIWM, London, 2000.
