Filter materials for phosphorus removal from wastewater in treatment wetlands—A review

This paper aims to collect and analyse existing information on different filter media used for phosphorus (P) removal from wastewater in constructed wetlands. The most commonly used materials are categorized as natural materials (considered in 39 papers), industrial byproducts (25 papers) and man-made products (10 papers). A majority of studies on sorbents have been carried out in lab-scale systems as batch experiments, and only very few studies have highlighted results on full-scale systems. Among the great variety of filter media studied, most of materials had a pH level >7 and high Ca (CaO) content. The highest P-removal capacities were reported for various industrial byproducts (up to 420 g P kg⁻¹ for some furnace slags), followed by natural materials (maximum 40 g P kg⁻¹ for heated opoka) and man-made filter media (maximum 12 g P kg⁻¹ for Filtralite). We found a significant positive Spearman Rank Order Correlation between the P retention and CaO and Ca content of filter materials (R² = 0.51 and 0.43, respectively), whereas the relation of P retention to pH level was weak (R² = 0.22) but significant. There is probably an optimal level of hydraulic loading rate at which the P removal is the highest. Additional important factors determining the applicability of filter materials in treatment wetlands such as saturation time, availability at a local level, content of heavy metals, and the recyclability of saturated filter media as fertilizer should be taken into consideration.     

Effect of reactive substrates used for the removal of phosphorus from wastewater on the fertility of acid soils

Reactive substrates used in filter systems can reduce phosphorus (P) pollution and, once saturated with P, may be recycled in agriculture. These substrates are usually calcium carbonate derivates with high pH values, which may be particularly beneficial for acid soils. Three reactive substrates (Filtra P, Polonite and wollastonite) saturated with P were used as amendments to an acid soil in a pot experiment. Substrate amendments tended to improve ryegrass yield and P uptake compared with control and potassium phosphate treatments. Polonite produced the highest yield/amendment ratio, while Polonite and Filtra P significantly increased the concentrations of P and Ca in the ryegrass. Addition of all three substrates increased the pH, AL-extractable P and cation exchange capacity of soils during the experiment. These substrates can therefore be applied to acid soils in order to recycle P and improve soil properties.     

Biomass accumulation and carbon utilization in layered sand filter biofilm systems receiving milk fat and detergent mixtures

This research evaluated utilization of organic compounds and biomass accumulation in three different layered sand filter biofilm systems renovating a mixture of detergent and milk fat. Organic compounds were measured as chemical oxygen demand (COD) and biomass as ATP. A two-layer sand filter with coarse sand on the top and fine sand at the bottom; a three-layer sand filter with pea gravel on the top, coarse sand in the middle, and fine sand at the bottom; and a two-filters in-series sand filter biofilm system each received mixtures of butterfat and detergent. While the three-layer sand filter system had greater COD utilization capacity due to a longer filter run, the two-filters in-series sand filter system had greater COD utilization efficiency. The COD utilization was highly associated with the period of filter run to clogging. The decrease of COD utilization rate could lead to the clogging of sand filters. To maintain long filter runs without clogging, a COD loading rate of 2100 mg O(2)/m(2)/h was recommended for three-layer sand filters and 1800 mg O(2)/m(2)/h for two-layer sand filters for high fat content wastewater. In the two-layer sand filters and the three-layer sand filters, biomass did not seem to accumulate in any particular layer on a unit volume of sand. However, based on unit surface area, the coarse sand layer accumulated more biomass than the fine sand layer and the pea gravel layer.     

Treatment of turkey processing wastewater with sand filtration

This research investigated the feasibility of coarse/fine sand filtration for removing organic materials from turkey processing wastewater. Sand filtration was tested with three organic and hydraulic loadings. Six two-layer sand bioreactors were in three groups, each with 5 cm layer of pea gravel at the bottom to support layers of fine sand (46 cm) and coarse sand (15 cm) to a height of 66 cm. The bioreactors were inoculated with a mixture of 20% (vol/vol) of wastewater lagoon sludge, 40% (vol/vol) of turkey processing wastewater, and 40% (vol/vol) of BOD(5) dilution water before starting the column operation with turkey processing wastewater. The wastewater contained 1270+/-730 mg COD/L and was applied to each sand bioreactor at hydraulic loading rates of 94% during 80 days of column operation at low and medium hydraulic loading rates (132 L/m(2)/day). The removal at the highest hydraulic loading rate (264 L/m(2)/day) declined after the appearance of a black zone in the top layer of fine sand on day 30 for one reactor and day 50 for the other. The sand filtration in this study represents a feasible treatment for turkey processing wastewater and its efficiency and the life span of the process are associated with the extent of hydraulic loading of the sand bioreactors.     

Transport and survival of bacterial and viral tracers through submerged-flow constructed wetland and sand-filter system

Untreated or improperly treated wastewater has often been cited as the primary contamination source of groundwater. The use of decentralized wastewater treatment systems has applicability around the world since it obviates the need for extensive infrastructure development and expenditures. The use of a submerged flow constructed wetland (CW) and a sand filter to remove bacterial and viral pathogens from wastewater streams was evaluated in this study Salmonella sp. and a bacteriophages tracer were used in conjunction with the conservative bromide tracer to understand the fate and transport of these organisms in these treatment systems. Viral breakthrough numbers in the sand filter and CW were similar with a Spearman Rank correlation of 0.8 (P<0.05). In the CW, the virus exhibited almost a 3-log reduction, while in the sand filter, the viruses exhibited a 2-log reduction. The bacterial tracers, however, did not exhibit similar reductions. Low numbers of bacteria and viruses were still detectable in the effluent streams suggesting that disinfection of the effluent is critical. The survival of the tracer bacteria and viruses was as expected dependent on the biotic and abiotic conditions existing within the wastewater. The results suggest that the microbial removal characteristics of decentralized wastewater treatment systems can vary and depend on factors such as adsorption, desorption and inactivation which in turn depend on the design specifics such as filter media characteristics and local climatic conditions.     

Population analysis in a denitrifying sand filter: conventional and in situ identification of Paracoccus spp. in methanol-fed biofilms.

The microbial community of a denitrifying sand filter in a municipal wastewater treatment plant was examined by conventional and molecular techniques to identify the bacteria actively involved in the removal of nitrate. In this system, denitrification is carried out as the last step of water treatment by biofilms growing on quartz grains with methanol as a supplemented carbon source. The biofilms are quite irregular, having a median thickness of 13 to 20 microns. Fatty acid analysis of 56 denitrifying isolates indicated the occurrence of Paracoccus spp. in the sand filter. 16S rRNA-targeted probes were designed for this genus and the species cluster Paracoccus denitrificans-Paracoccus versutus and tested for specificity by whole-cell hybridization. Stringency requirements for the probes were adjusted by use of a formamide concentration gradient to achieve complete discrimination of even highly similar target sequences. Whole-cell hybridization confirmed that members of the genus Paracoccus were abundant among the isolates. Twenty-seven of the 56 isolates hybridized with the genus-specific probes. In situ hybridization identified dense aggregates of paracocci in detached biofilms. Probes complementary to the type strains of P. denitrificans and P. versutus did not hybridize to cells in the biofilms, suggesting the presence of a new Paracoccus species in the sand filter. Analysis using confocal laser scanning microscopy detected spherical aggregates of morphologically identical cells exhibiting a uniform fluorescence. Cell quantification was performed after thorough disruption of the biofilms and filtration onto polycarbonate filters. An average of 3.5% of total cell counts corresponded to a Paracoccus sp., whereas in a parallel sand filter with no supplemented methanol, and no measurable denitrification, only very few paracocci (0.07% of cells stained with 4',6-diamidino-2-phenylindole) could be detected. Hyphomicrobium spp. constituted approximately 2% of all cells in the denitrifying unit and could not be detected in the regular sand filter. This clear link between in situ abundance and denitrification suggests an active participation of paracocci and hyphomicrobia in the process. Possible selective advantages favoring the paracocci in this habitat are discussed.     

Reuse potential of phosphorus-rich filter materials from subsurface flow wastewater treatment filters for forest soil amendment

The aim of this study was to test P-enriched filter materials from a wastewater treatment experiment regarding their fertilizing efficiency in pot experiments with silver birch (Betula pendula Roth.) seedlings. Tested materials included hydrated calcium-rich oil shale ash and well-mineralised peat. A mixture of peat and hydrated ash demonstrated the best results: seedlings grown on that mixture had the highest P concentration and an optimal N:P:K ratio in leaves. Hydrated ash has a low concentration of heavy metals and almost the same composition as common lime fertilizers. The hydrated ash sediment and its combination with well-mineralised peat enriched with phosphorus in wastewater treatment filters are promising materials for forest soil fertilization.     

A carbon free filter for collection of large volume samples of cellular biomass from oligotrophic waters

Isotopic analysis of cellular biomass has greatly improved our understanding of carbon cycling in the environment. Compound specific radiocarbon analysis (CSRA) of cellular biomass is being increasingly applied in a number of fields. However, it is often difficult to collect sufficient cellular biomass for analysis from oligotrophic waters because easy-to-use filtering methods that are free of carbon contaminants do not exist. The goal of this work was to develop a new column based filter to autonomously collect high volume samples of biomass from oligotrophic waters for CSRA using material that can be baked at 450°C to remove potential organic contaminants. A series of filter materials were tested, including uncoated sand, ferrihydrite-coated sand, goethite-coated sand, aluminum-coated sand, uncoated glass wool, ferrihydrite-coated glass wool, and aluminum-coated glass wool, in the lab with 0.1 and 1.0 μm microspheres and Escherichia coli. Results indicated that aluminum-coated glass wool was the most efficient filter and that the retention capacity of the filter far exceeded the biomass requirements for CSRA. Results from laboratory tests indicate that for oligotrophic waters with 1×10(5) cells ml(-1), 117l of water would need to be filtered to collect 100 μg of PLFA for bulk PLFA analysis and 2000 l for analysis of individual PLFAs. For field sampling, filtration tests on South African mine water indicated that after filtering 5955l, 450 μg of total PLFAs were present, ample biomass for radiocarbon analysis. In summary, we have developed a filter that is easy to use and deploy for collection of biomass for CSRA including total and individual PLFAs.     

Biomass allocation and nutrient use in fast-growing woody and herbaceous perennials used for phytoremediation

This study assessed the suitability of two deciduous woody perennials (Salix spp. and Populus spp.) and two summer green herbaceous perennials (Phragmites australis and Urtica dioica) for purification of nutrient enriched wastewater. The main hypothesis tested was that species with a particular trait combination of high relative growth rate (RGR), low nutrient productivity (A) and high mean residence time (MRT) of nutrients would be most effective in accumulating nutrients. The nitrogen and phosphorus use efficiency at the whole plant level was analysed. Four treatments comprising two possible phytoremediation substrates (municipal wastewater and landfill leachate) and two control plant nutrition situations (balanced nutrient solution and pure water) were applied in four replications to the four plant species. Generally, all four species studied showed a high RGR and a low P productivity in the balanced nutrient solution treatment, while the opposite (low RGR and high P productivity) was seen in the phytoremediation substrate and pure water treatments. The general conclusion is that if P is present in marginal proportions in the wastewater, a vegetation filter with Phragmites would have an advantage since biomass and nutrient accumulation in Phragmites does not decrease as much during phytoremediation as that in deciduous woody perennials.     

Integrated constructed wetland systems employing alum sludge and oyster shells as filter media for P removal

This research aimed to investigate the technical feasibility of integrated constructed wetland system consisting of a pre-filter unit and a constructed wetland (CW), in series; packed with alum sludge (AS) and oyster shells (OS) as the filter media, respectively, for nitrogen and phosphorus removal from domestic wastewater. Based on the 240 days of operation from January to August 2007, this integrated system was highly effective in removing BOD, N, P and TSS compounds which were found to be 89.5%, 68.8%, 99.4% and 89.9%, respectively. After this period, the integrated system was modified as the CW and post-filter unit, in series. The post-filter of this modified integrated system was operated during 60 days with cover for light shield and during another 60 days with no cover from September to December 2007. The treatment performance of modified integrated system was effective in removing BOD₅, N, P and TSS compounds which were found to be 91.4%, 86.8%, 99.7% and 73%, respectively, during which the post-filter had operated with no cover. To simulate high rainfall conditions, the integrated system was tested under hydraulic shock loading at the overall hydraulic retention time of 0.7 day during one day. This hydraulic shock loading conditions made BOD₅, TN, TSS concentration increase, but made no effect on P concentration. Integrated system combined a pre-filter and a CW unit or a CW unit followed by a post-filter is recommended for use in domestic wastewater which should result in high treatment performance, especially on P removal.     

Application of alkali-activated materials for water and wastewater treatment: a review

Alkali-activation (or geopolymer) technology has gained a great deal of interest for its potential applications in water and wastewater treatment during the last decade. Alkali-activated materials can be prepared via a relatively simple and low-energy process, most commonly by treating aluminosilicate precursors with concentrated alkali hydroxide and/or silicate solutions at (near) ambient conditions. The resulting materials are, in general, amorphous, have good physical and chemical stability, ion-exchange properties, and a porous structure. Several of the precursors are industrial by-products or other readily available low-cost materials, which further enhances the environmental and economic feasibility. The application areas of alkali-activated materials in water and wastewater treatment are adsorbents/ion-exchangers, photocatalysts, high-pressure membranes, filter media, anti-microbial materials, pH buffers, carrier media in bioreactors, and solidification/stabilization of water treatment residues. The purpose of this review is to present a comprehensive evaluation of the rapidly growing prospects of alkali-activation technology in water and wastewater treatment.

     

Use of blast furnace granulated slag as a substrate in vertical flow reed beds: field application

Research was conducted at Middle East Technical University (METU), Ankara, Turkey in 2000 to determine whether a reed bed filled with an economical Turkish fill media that has high phosphorus (P) sorption capacity, could be implemented and operated successfully under field conditions. In batch-scale P-sorption experiments, the P-sorption capacity of the blast furnace granulated slag (BFGS) of KARDEMIR Iron and Steel Ltd., Co., Turkey, was found to be higher compared to other candidate filter materials due to its higher Ca content and porous structure. In this regard, a vertical subsurface flow constructed wetland (CW) (30 m(2)), planted with Phragmites australis was implemented at METU to treat primarily treated domestic wastewater, at a hydraulic rate of 100 mm d(-1), intermittently. The layers of the filtration media constituted of sand, BFGS, and gravel. According to the first year monitoring study, average influent and effluent total phosphorus (TP) concentrations were 6.61+/-1.78 mg L(-1) and 3.18+/-1.82 mg L(-1); respectively. After 12 months, slag samples were taken from the reed bed and P-extraction experiments were performed to elucidate the dominant P-retention mechanisms. Main pools for P-retention were the loosely-bounded and Ca-bounded P due to the material's basic conditions (average pH>7.7) and higher Ca content. This study indicated the potential use of the slag reed bed with higher P-removal capacity for secondary and tertiary treatment under the field conditions. However, the P-sorption isotherms obtained under the laboratory conditions could not be used favorably to determine the longevity of the reed bed in terms of P-retention.     

Retention and removal of the fish pathogenic bacterium Yersinia ruckeri in biological sand filters

AIMS: To investigate the retention and removal of the fish pathogenic bacterium Yersinia ruckeri in biological sand filters and effects on the microbial community composition. METHODS AND RESULTS: Sand filter columns were loaded (70 mm day(-1)) with fish farm wastewater and a suspension (10(8) CFU ml(-1)) of Y. ruckeri. Bacterial numbers and protozoan numbers were determined by plate counts and epifluorescence microscopy, respectively, and microbial biomass and community composition were assessed by phospholipid fatty acids (PLFA) analysis. Concentrations of Y. ruckeri in the filter effluent decreased from 10(8) to 10(3)-10(5) CFU ml(-1) during the experiment. Numbers of Y. ruckeri in the sand decreased from 10(6) CFU g(-1) dry weight (DW) sand to 10(4) CFU g(-1) DW sand. In contrast, microbial biomass determined with plate counts and total PLFA increased during the whole experiment. Principal component analysis (PCA) revealed a change in microbial community composition with time, with the most pronounced change in surface layers and towards the end of the experiment. Protozoan numbers increased from ca 0-600 cells g(-1) DW sand, indicating the establishment of a moderate population of bacterial grazers. CONCLUSIONS: The removal of Y. ruckeri improved during the experiment. Introduction of Y. ruckeri to the sand filter columns stimulated growth of other micro-organisms, which in turn caused a shift in the microbial community composition in the sand. SIGNIFICANCE AND IMPACT OF THE STUDY: This study increases the understanding of the dynamics of sand filters subjected to a high loading of a pathogenic bacterium and can therefore be used in future work were the overall aim is to provide a more reliable and efficient removal of pathogenic bacteria in biological sand filter systems.     

猪场污水的A2/O-深度处理和中水回用

猪场污水采用A²/O-生物接触氧化-混凝沉淀-砂滤消毒组合方法处理,可以实行中水回用。原污水经过固液分离、调解池、竖流式初沉池、厌氧消化池、缺氧池、传统活性污泥曝气池、竖流式二沉池、接触氧化池、药剂混合池、斜板沉淀池、砂滤池、pH调解和消毒池处理后进入排放池,系统正常运行后处理废水可以达到GB8978-1996中规定的二级以上标准,处理水没有检测到病原微生物。中水引入中水塔,通过独立设计的中水道,可以应用到猪舍冲栏使用。而污泥通过污泥干化场得到干化。     

The Effect of Hydraulic Loading Rate and Influent Source on the Binding Capacity of Phosphorus Filters

Sorption by active filter media can be a convenient option for phosphorus (P) removal and recovery from wastewater for on-site treatment systems. There is a need for a robust laboratory method for the investigation of filter materials to enable a reliable estimation of their longevity. The objectives of this study were to (1) investigate and (2) quantify the effect of hydraulic loading rate and influent source (secondary wastewater and synthetic phosphate solution) on P binding capacity determined in laboratory column tests and (3) to study how much time is needed for the P to react with the filter material (reaction time). To study the effects of these factors, a 2² factorial experiment with 11 filter columns was performed. The reaction time was studied in a batch experiment. Both factors significantly (α = 0.05) affected the P binding capacity negatively, but the interaction of the two factors was not significant. Increasing the loading rate from 100 to 1200 L m⁻² d⁻¹ decreased P binding capacity from 1.152 to 0.070 g kg⁻¹ for wastewater filters and from 1.382 to 0.300 g kg⁻¹ for phosphate solution filters. At a loading rate of 100 L m⁻² d⁻¹, the average P binding capacity of wastewater filters was 1.152 g kg⁻¹ as opposed to 1.382 g kg⁻¹ for phosphate solution filters. Therefore, influent source or hydraulic loading rate should be carefully controlled in the laboratory. When phosphate solution and wastewater were used, the reaction times for the filters to remove P were determined to be 5 and 15 minutes, respectively, suggesting that a short residence time is required. However, breakthrough in this study occurred unexpectedly quickly, implying that more time is needed for the P that has reacted to be physically retained in the filter.     

Short Communication: Antibiotic resistance among faecal coliform bacteria isolated from wastewater before and after treatment by an experimental sand filter

During the functioning period of a sand filter, 254 and 234 faecal-coliform strains were isolated from raw wastewater and reclaimed effluent, respectively. When tested for their susceptibility to ampicillin, kanamycin, chloramphenicol, streptomycin and tetracycline, 53% of the strains from the raw sewage and 84% of those isolated from the effluent were resistant to at least one antibiotic. The high incidence of drug resistance in the effluent was connected with a high proportion of Klebsiella pneumoniae, Enterobacter cloacae and Citrobacter freundii.     

Bioremediation of Vegetable Oil and Grease from Polluted Wastewater Using a Sand Biofilm System

Pseudomonas sp. (L1), P. diminuta(L2) were among eight bacterial strains isolated from vegetable grease and oil-contaminated industrial wastewater, four of which only were found to have the ability to degrade oil and grease. They were identified and investigated for oil and grease degradation either individually or in combinations in previous unpublished work by the authors. Since the combination M1 (Pseudomonas sp. andP. diminuta) produced the highest degradative activity, it was used in the present study in a biofilm sand filter system for vegetable oil and grease removal. This system was tested either as one unit or two units in sequence where different flow rates (30, 50, 100 ml/h) were applied compared to a control unit(s). Results showed that both biofilm systems reduced oily wastewater, even in cases of high degree of pollution (fat, oil & grease (FOG), 7535 ppm; biochemical oxygen demand (BOD₅), 525 ppm; chemical oxygen demand (COD), 1660 ppm). Results also showed a removal of FOG with efficiency at 100%; BOD₅ at 95.9% and COD at 96%, at 50 ml/h flow rate using one unit of biofilm system. On using two units in sequence, a complete removal of FOG, BOD₅ and COD with efficiency 100%, at flow rate 100 ml/h was achieved. In conclusion, the previous biofilm results indicated the efficiency of such a system in treating oily polluted wastewater (vegetable oil origin) on the basis of bacterial isolates being used, the optimum flow rate, and the number of biofilm units used in sequence to obtain the highest removal capacity of such a system. This revised version was published online in July 2006 with corrections to the Cover Date.     

Using 2,6-dichlorobenzamide (BAM) degrading Aminobacter sp. MSH1 in flow through biofilters—initial adhesion and BAM degradation potentials

Micropollutants in groundwater are given significant attention by water companies and authorities due to an increasing awareness that they might be present even above the legal threshold values. As part of our investigations of the possibility to remove the common groundwater pollutant 2,6-dichlorobenzamide (BAM) by introducing the efficient BAM degrader Aminobacter sp. MSH1 into biologically active sand filters, we investigated if the strain adheres to filters containing various filter materials and if the initial adherence and subsequent degradation of BAM could be optimized. We found that most of the inoculated MSH1 cells adhered fast and that parameters like pH and ionic strength had only a minor influence on the adhesion despite huge influence on cell surface hydrophobicity. At the given growth protocol, the MSH1 strain apparently developed a subpopulation that had lost its ability to adhere to the filter materials, which was supported by attempted reinoculation of non-adhered cells. Analysis by quantitative PCR showed that most cells adhered in the top of the filters and that some of these were lost from the filters during initial operation, while insignificant losses occurred after 1 day of operation. The inoculated filters were found to degrade 2.7 μg/L BAM to below 0.1 μg/L at a 1.1-h residence time with insignificant formation of known degradation products. In conclusion, most filter materials and water types should be feasible for inoculation with the MSH1 strain, while more research into degradation at low concentrations and temperatures is needed before this technology is ready for use at actual waterworks.     

Mechanical mode floating medium filters for recirculating systems in aquaculture for higher solids retention and lower freshwater usage

The aim of this work was to develop a better understanding of a floating medium in a mechanical filtration mode. The experiments were carried out using a commonly available polystyrene floating medium filter with the grain size of 1mm. A sand medium filter with the similar grain size was also tested for the comparison. A short-term trial of 2h and a long-term of 20 days filtration times were conducted with three custom manufactured pressurized filters of 16l. The filters were operated under three different configurations: (i) upflow with floating media (UFMF), (ii) downflow with floating media (DFMF) and (iii) downflow with a sand medium (DSF). The results of the long-term trial indicated that at a flow rate of 22 m/h, the UFMF and DSF had similar solid removal capacity with an average total suspended solids (TSS) removal efficiency of 60%. The DFMF could only remove 33% of TSS. However, during the short-term trial, TSS removal efficiency of the UFMF was better compared to the DSF (e.g., 71%, 56% and 57% of TSS removal in UFMF compared to 66%, 49% and 41% in the DSFF at the flow rates of 20, 25 and 31m/h, respectively). The energy requirements of each filter were compared by measuring the pressure differential across each filter. The long-term trial indicated that the UFMF had a significantly less pressure differential (44 kPa) compared to the DSF (80 kPa) (p<0.001). This was further confirmed that at different flow rates whereby the DSF displayed higher pressure differentials for filtration rates at 350, 450, 550 and 800 l/h. The study indicated that floating medium filter was better and more applicable to recirculating aquaculture systems than conventional pressurized sand filter.     

Integrated Metagenomic and Physiochemical Analyses to Evaluate the Potential Role of Microbes in the Sand Filter of a Drinking Water Treatment System

While sand filters are widely used to treat drinking water, the role of sand filter associated microorganisms in water purification has not been extensively studied. In the current investigation, we integrated molecular (based on metagenomic) and physicochemical analyses to elucidate microbial community composition and function in a common sand filter used to treat groundwater for potable consumption. The results revealed that the biofilm developed rapidly within 2 days (reaching ∼10¹¹ prokaryotes per gram) in the sand filter along with abiotic and biotic particulates accumulated in the interstitial spaces. Bacteria (up to 90%) dominated the biofilm microbial community, with Alphaproteobacteria being the most common class. Thaumarchaeota was the sole phylum of Archaea, which might be involved in ammonia oxidation. Function annotation of metagenomic datasets revealed a number of aromatic degradation pathway genes, such as aromatic oxygenase and dehydrogenase genes, in the biofilm, suggesting a significant role for microbes in the breakdown of aromatic compounds in groundwater. Simultaneous nitrification and denitrification pathways were confirmed as the primary routes of nitrogen removal. Dissolved heavy metals in groundwater, e.g. Mn²⁺ and As³⁺, might be biologically oxidized to insoluble or easily adsorbed compounds and deposited in the sand filter. Our study demonstrated that the role of the microbial community in the sand filter treatment system are critical to effective water purification in drinking water.