Quantcast
Last updated on April 23, 2014 at 1:22 EDT

Pathogen and Indicator Organism Reduction Through Secondary Effluent Filtration: Implications for Reclaimed Water Production

September 21, 2008

By Levine, Audrey D Harwood, Valerie J; Farrah, Samuel R; Scott, Troy M; Rose, Joan B

ABSTRACT: The reduction of pathogens and indicator organisms through secondary effluent filtration was investigated at six full- scale treatment facilities, ranging in capacity from 0.04 to 1 m^sup 3^/s (1 to 25 mgd). Grab samples were assayed for pathogens (cultivable enteric viruses, Giardia, and Cryptosporidium) and indicator organisms (coliforms, enterococci, Clostridium perfringens, and coliphages) quarterly under peak flow conditions from each facility over the course of 1 calendar year. Log^sub 10^ removals resulting from filtration averaged 0.3 to 0.8 log^sub 10^ for cultivable enteric viruses, 0.4 to 1.5 log^sub 10^ for protozoan parasites, 0.01 to 3.7 log^sub 10^ for indicator bacteria, and 0.3 to 1.1 log^sub 10^ for coliphages. In addition to filter design (cloth, monomedium shallow- or deep-bed, or dual-media filters), differences in reduction of pathogens and indicators could be attributed to the combined effects of hydraulic loading rates, chemical addition practices, backwashing and postbackwashing operating strategies, and the effectiveness of upstream biological treatment and sedimentation.

Water Environ. Res., 80, 596 (2008).

KEYWORDS: reclaimed water, pathogens, wastewater filtration, prechlorination.

doi: 10.2175/106143008X266742

Introduction

With increasing use of reclaimed water for public access applications, such as landscape irrigation, urban water uses, wetlands and surface water augmentation, and recirculating cooling water for buildings and other industrial applications, the effectiveness of treatment practices for control of pathogens has become more important. In many facilities, nitration is used between secondary biological treatment and disinfection (either chemical or photochemical) to reduce the concentration of suspended solids and particle-associated microorganisms that may interfere with disinfection effectiveness. Because it is now recognized that some pathogens, such as Cryptosporidium, are resistant to disinfection, secondary effluent filtration represents a critical step to control pathogens that are not removed by biological treatment. The effectiveness of pathogen reduction through filtration is influenced by filter characteristics and operating practices, microbial properties (size, surface properties, and degree of association with other microorganisms or particles), and water quality variables. A wide variety of filters has been adopted for water reclamation applications, including conventional deep-bed monomedia or dual- media filters, continuously backwashed upflow filters, disk (cloth) filters, fuzzy filters, and membranes (Bourgeous et al., 2003; England et al., 1994). To meet requirements for process reliability and maintenance of flow, secondary effluent filtration typically consists of a minimum of two filters operated in parallel, in either a declining rate or constant rate mode. Typically, effluents from individual filters are combined before disinfection. Direct monitoring of the microbial characteristics of filter effluents is not widely practiced, as the point of compliance for most reclaimed water systems is after disinfection.

The purpose of this paper is to evaluate the role of filtration for removal of pathogens and indicator organisms during production of reclaimed water and the degree to which traditional models of surface water filtration can be applied to reclaimed water filtration scenarios. The data presented in this paper are derived from testing conducted at six full-scale water reclamation facilities that use either granular media filtration (anthracite and/ or sand) or cloth filters (Rose et al., 2004). Relationships between filter characteristics (depth, hydraulic loading, and backwashing practices); water quality; and removal or persistence of viruses, protozoan pathogens, conventional indicator bacteria, and alternative microbial indicators through full-scale secondary effluent filtration are evaluated.

Background

Individually and collectively, treatment processes provide a combination of physical, chemical, and biological barriers to reduce pathogen levels in reclaimed water. To verify treatment effectiveness, disinfected effluents are routinely assayed for bacterial indicators and sampled intermittently for viruses and/or protozoan pathogens (U.S. EPA, 2004; York and Walker-Coleman, 2000). Coliform bacteria (either the total coliform group or thermotolerant [fecal] coliforms) are used almost universally to benchmark the microbiological quality of reclaimed water. In the United States, monitoring requirements (microbial analyte[s] and frequency) and specific numerical limits are set by state and local regulatory agencies and, typically, are based on daily monitoring of disinfected effluents.

Filtration Removal Mechanisms. Removal of microorganisms through granular media filtration is governed by particle-media interactions that result from diffusion, interception, and sedimentation (Camp, 1964; Darby et al., 1991; Elimelech, 1992; Ghosh et al., 1975; Glasgow and Wheatley, 1999; Habibian and O’Melia, 1975; Hall and Fitzpatrick, 2000; Iwasaki, 1937; Jegatheesan and Vigneswaren, 2000; Kavanaugh et al., 1978; O’Melia and Ali, 1978; Stevenson, 1997; Tchobanoglous and Eliassen, 1970; Trussell and Chang, 1999; Yao et al., 1971). The deposition of microorganisms and other particles in filters is dependent on effective transport to the filter media coupled with retention within the filter pores or attachment to the media surface.

The single collector efficiency concept has been widely used to evaluate the overall efficiency of filtration for removal of particles by defining a nondimensional single collector efficiency, eta, which reflects the ratio of particles striking the filter media to the total particle flux by incorporating the combined effects of diffusion (eta^sub D^), interception (eta^sub I^), and gravity sedimentation (eta^sub G^) on particle-media interactions, as defined in eq 1 (Camp, 1964; Elimelech, 1992; Habibian and O’Melia, 1975; Harvey and Garabedian, 1991; Iwasaki, 1937; Logan et al., 2001; Redman et al., 2001; Yao et al., 1971).

Where

eta = overall collision frequency coefficient;

eta^sub D^ = collision frequency coefficient resulting from diffusion;

eta^sub I ^= collision frequency coefficient resulting from interception, eta^sub D^;

eta^sub G^ = collision frequency coefficient resulting from gravitational sedimentation;

k = Boltzman constant (1.38 x 10^sup -23^ J/K);

T = absolute temperature (K);

mu = dynamic viscosity of water (N-s/m^sup 2^);

d^sub p^ = effective particle (microorganism) diameter (m);

d = effective diameter of media grain (m);

v = water velocity = flowrate/surface area (m/s);

rho = density of water (kg/m^sup 3^); and

rho^sub P^ = particle (microorganism) density (kg/m^sup 3^).

In this model, transport by diffusion is inversely proportional to the media grain size (d^sub p^) and filtration velocity (v). Transport by interception is controlled by the ratio of the microorganism size (d^sub p^) to the grain size (d) and is independent of filtration velocity, while transport by sedimentation is proportional to the ratio between the particle settling velocity and the filtration velocity (v) or hydraulic loading rate. The model does not directly incorporate water quality variables, such as pH, ionic strength, or organic content.

A comparison of the theoretical single collector removal efficiency for control of protozoa, bacteria, and viruses through sand and anthracite media typically used for production of reclaimed water suggests that individual bacterial cells are likely to have the lowest removal efficiency compared with freely suspended viruses, protozoa, or aggregates (flocs) of microorganisms and other particulates (Figure 1). Essentially, removal of freely suspended viruses and other nanoscale particles is controlled by diffusion, while removal of protozoa is influenced through the combined effects of differential sedimentation and interception. Removal of freely suspended bacterial cells is influenced by the cumulative effects of diffusion, differential sedimentation, and interception. The effective grain size of the media affects the collector efficiency for removal of viruses and bacteria, whereas removal of protozoa and/ or microbial aggregates is more affected by hydraulic loading rates than by media characteristics. It should also be noted that removal of bacteria through filtration is likely to be less sensitive to the influence of variable hydraulic loading rates than removal of viruses or protozoa (see eq 1).

These calculations suggest that the use of bacterial indicators may provide a conservative assessment of filter performance for removal of microorganisms, if bacteria are present as discrete particles (not associated with other particulate matter), and disinfectant chemicals are not added upstream. From a process design perspective, this model suggests that smaller grain size media would affect removal of freely suspended viruses and other nanoparticles, whereas lower hydraulic loading rates would be more effective at improving removal efficiency for protozoan pathogens. In addition, the use of upstream coagulation to aggregate freely suspended viral and bacterial particles should be designed in conjunction with optimization of hydraulic loading rates. This model does not address the development of headloss through filter runs, the interplay of media size and hydraulic loading rates on headloss development, or potential effects of coagulant chemicals and water quality on headloss and filter run length. While the collector efficiency concept describes the likelihood of particles colliding with filter media, the net removal depends on retention within the filter. Factors that are likely to influence attachment include the net surface charge on the filter media and microbial surfaces; media properties (type, size, and depth); hydraulic loading rates; upstream chemical use (oxidants and/or coagulants); water quality variables; flow control; and backwashing and post-backwashing practices. The characteristics of secondary effluent that might affect filter and microbial surface characteristics include pH, ionic strength, temperature, particle characteristics, and characteristics of dissolved organics (concentration, molecular size, and charge density). High ionic strength and residual organics act to decrease the electric double layer surrounding microorganisms, particles, and filter media, thereby increasing the potential for particle attachment (Brown and Abramson, 2006; Elimelech and O’Melia, 1990; Hsu and Huang, 2002; Hsu et al., 2001). Ionic strength varies with total dissolved solids levels in reclaimed water, which can range from 200 to 2500 mg/L, depending on the characteristics of the service area source water quality, water usage patterns, stormwater effects (infiltration and inflow), and sources of dissolved solids from industrial discharges and point-of- use water treatment discharges. The presence of residual organics may affect adsorption to filter media or may serve to improve transport through filtration (Foppen et al., 2006).

Typically, secondary effluent contains active biomass that consists of microorganisms that are either freely dispersed or entrapped in microbial agglomerates (floes) that can colonize the filter media. During a filter run, the buildup of accumulated material within the filter coupled with the development of biofilms can serve to augment filter efficiency; however, a side effect of particle accumulation and biofilm formation is an increase in specific velocities within pores, resulting in pore blockage, channeling effects, higher effective hydraulic loading rates, and particle remobilization (Ahmad and Amirtharajah, 1998; Hall and Fitzpatrick, 2000; Hozalski and Bouwer, 1998). Some particles/ pathogens may be released during filtration, depending on the duration of the filter run, hydraulic inconsistencies, and effectiveness of upstream treatment (Ahmad and Amirtharajah, 1998; Darby et al., 1991; Glasgow and Wheatley, 1999; Jegatheesan and Vigneswaran, 2000; Jiminex et al., 2000).

Backwashing and postbackwashing practices vary with filter designs and plant operating practices. Backwashing operations are designed to dislodge and scour accumulated particles from filter media, by fluidizing the media and applying a combination of highvelocity water with or without air scour (Ahmad and Amirtharajah, 1998; Hall and Fitzpatrick, 2000; Hozalski and Bouwer, 1998). Operating variables, such as backwashing frequency and duration, degree of fluidization, backwash source water (chlorinated or nonchlorinated), degree to which air scour is used and monitored, and approach for returning a filter to service following backwashing, can affect the efficiency of subsequent filtration cycles. In general, backwashing is not designed to remove biofilms (Wang et al., 1995). Particle/pathogen release may occur following backwashing, resulting from either remnant particles (Amburgey et al., 2004) or hydrodynamic inconsistencies in the filter media (Glasgow and Wheatley, 1999).

Modeling of filtration effectiveness for reclaimed water applications is complicated by the heterogeneity of the particles in secondary effluent coupled with the lack of tools to quantify the effects of particle accumulation and filter porosity on interstitial fluid velocities (Redman et al., 2001). One approach that has been widely used is to assume that pathogen deposition through filtration can be modeled as pseudo-first-order removal, as shown in eq 2, where C is the number concentration of particles or microorganisms entering a unit volume of the filter, e is a filter coefficient, and L is depth (Brown and Abramson, 2006; Camp, 1964; Harvey and Garabedian, 1991; Iwasaki, 1937; Reddi, 1997; Trussell and Chang, 1999). The filter coefficient, e, reflects the composite effects of filter media characteristics, hydrodynamic factors, and water quality variables.

Where

C = number concentration of microorganisms (#/100 mL or #/100 L),

L = depth from surface of filter (m), and

lambda = filter coefficient (m^sup -1^).

Under steady-state conditions, the concentration of microorganisms in the filter effluent can be estimated as follows:

C = C^sub o^e^sup -lambdaz^ (3)

Where

C^sub o^ = concentration of particles or microorganisms applied to the surface of the filter (#/100 mL or #/100 L), and

z = filter depth (m).

This model suggests that, for a given filter coefficient, lambda, filter efficiency should increase with depth. However, in practice, filter depth and filtration run length are limited by headloss and other practical constraints. Another way of increasing filter performance is to increase the effective filter coefficient, lambda, through upstream chemical addition of coagulants, flocculants, and/ or oxidants.

Review of Pathogen Removal Through Surface Water Filtration in the Context of Water Reclamation. Pathogen removal through filtration has been studied extensively in the production of drinking water from surface water sources (Amburgey et al., 2004, 2005; Bustamante et al., 2001; Edzwald et al., 2000; Emelko et al., 2003; Hamngton et al., 2003; Hsu and Huang, 2002; Logan et al., 2001; Nasser et al., 1995; Nieminski and Ongerth, 1995; Patania et al., 1995; Swertfeger et al., 1999; Xagoraraki et al., 2004). In potable water systems, it has been observed that removal of Giardia cysts and Cryptosporidium oocysts is influenced by the degree of filter maturation; use of coagulant chemicals (Koivunen et al., 2003; Mosher and Hendricks, 1986; Patania et al., 1995); ionic strength; pH; zeta potential (Bustamante et al., 2001; Hsu and Huang, 2002); and filter grain size (Logan et al., 2001 ; Stevenson, 1997). In general, the surface charge of cysts and oocysts is negative under neutral and basic pH conditions, with the zeta potential for Cryptosporidium approximately double that of Giardia at neutral pH (Hsu and Huang, 2002; Ongerth and Pecoraro, 1996), suggesting that removal of Cryptosporidium through filtration may not parallel Giardia removal, particularly under low-ionic-strength and/or alkaline pH conditions. While reclaimed water filtration is similar to surface water filtration, in principle, limited data are available on the relative roles of pH, ionic strength, and organic matter in reclaimed water applications. Because indicator organisms (coliform bacteria and coliphages) are used in conjunction with indirect measures of microbial concentrations (suspended solids and turbidity) to assess the performance of reclaimed water filtration, it is important to identify the appropriate parameters that should be monitored to facilitate assessment and prediction of pathogen removal through filtration.

Comparison of Regulatory Approaches for Surface Water Filtration and Reclaimed Water Production. Because of concerns about the protozoan pathogens Giardia and Cryptosporidium in water supplies, the U.S. Environmental Protection Agency (Washington, D.C.) (U.S. EPA) has implemented rules that mandate that surface-water- treatment facilities include robust filtration systems that meet stringent turbidity requirements and achieve at least a 2-log reduction of protozoan pathogens (40 CFR 141.721 [U.S. EPA, 2006]). For example, under the Long-Term 2 Enhanced Surface Water Treatment Rule enacted in 2006, at least 95% of the combined filter effluent turbidity measurements must be less than or equal to 0.15 NTU (40 CFR 141.727) (U.S. EPA, 2003). While turbidity levels do not correspond to pathogen concentrations, the rule is based on conclusive evidence from a wide array of surface water filtration studies, which demonstrated that turbidity levels below 0.2 NTU are likely to correspond to low or nondetectable levels of Giardia and Cryptosporidium (Edzwald et al., 2000; Emelko et al., 2005; Xagoraraki et al., 2004). In contrast, turbidity (or suspended solids) requirements for water reclamation facilities are based on monitoring of the disinfected effluent, with typical turbidity limits ranging from 2 to 5 NTU or suspended solids limits of 5 mg/ L, depending on permit requirements (U.S. EPA, 2004; York and Walker- Coleman, 2000). Thus, surface-water-filtration practices cannot be directly extrapolated to reclaimed water production, as a result of the order of magnitude differences in turbidity limits and other confounding variables.

Role of Filtration in Reclaimed Water Production. In reclaimed water production, filtration is intended to provide for removal of microorganisms and associated turbidity, as a pretreatment to disinfection. Disinfection systems are designed to inactivate pathogens that are not removed through upstream processes. Historically, wastewater treatment facilities have relied on the use of chlorine as a disinfectant. However, because of increasing concerns about the presence of disinfection byproducts in treated wastewater and the need to dechlorinate effluents that are discharged to receiving waters, there has been a shift towards the use of UV and other alternative disinfectants. In addition, some pathogens are resistant to chlorine. As disinfection systems become more streamlined, the role of filtration becomes more important for pathogen control. In addition, pathogen removal through filtration is even more critical for control of pathogens that are resistant to chemical or photochemical disinfection. Methodology

The reduction of pathogens through filtration was assessed through sampling of six wastewater treatment facilities under peakflow conditions. A summary of the characteristics of the treatment facilities is given in Table 1. The filters tested in this project included granular media (deep-bed and shallow-bed) and cloth filters. Prechlorination was applied in one of the facilities (S11); cationic polyelectrolyte was applied upstream of filtration in another (A48); and a third facility used alum in the secondary clarifier (A24S48), while the remaining facilities did not use any upstream chemicals during the testing program. Backwashing practices varied among the facilities, but four of the six facilities used chlorinated effluent on a routine basis, as detailed in Table 1. A minimum of four sampling events were conducted at each facility, over a 1-year period, to assess pathogen and indicator concentrations associated with untreated wastewater and effluents from biological treatment (postsedimentation), filtration, and disinfection. For each sample event, filter operating data were compiled, including hydraulic loading rates, backwashing history, and operational anomalies. Routine water quality monitoring data, such as the 5-day test for carbonaceous biochemical oxygen demand (C- BOD^sub 5^), total suspended solids (TSS), turbidity, ammonia, and phosphorus were obtained from each facility; however, all facilities did not monitor all parameters on the same schedule as the microbial sampling program.

A side-by-side comparison of the effects of prechlorination on removal of indicator organisms was conducted over a 10-hour filter run at the facility that prechlorinates (S11). In addition, intensive testing of a single filter run, to evaluate the consistency of removal of bacterial indicators, was conducted at three of the facilities (S11, A48, and A24S48). The effect of backwashing on protozoan pathogen concentrations (Giardia and Cryptosporidium) in filter effluents immediately before and after putting a filter back in service was evaluated at two facilities (A48 and A24S48).

Nine independent measurements of microbiological water quality were used to assess filtration performance. A summary of the characteristics of the indicator organisms (coliform bacteria, enterococci, Clostridium perfringens, and coliphages) and pathogens (cultivable enteric viruses, Giardia, and Cryptosporidium) tested in the influent and effluent from each filter is shown in Table 2. The frequency with which each of the tested microorganisms was detected in each filter effluent is also reported in Table 2.

All microbial assays were carried out as previously reported (Harwood et al., 2005; Rose et al., 2004). One-liter grab samples were collected for analysis of the bacterial and viral indicators, while the parasite and virus samples were collected and concentrated by pumping up to 100 L of water through appropriate filters. All chlorinated samples were dechlorinated by adding Na^sub 2^S^sub 2^O^sub 3^ to a final concentration of 0.001% immediately after collection. Cartridge filters (IMDS, Cuno Inc., Meriden, Connecticut) were used for collection and concentration of viruses, while Pall-Gelman Envirochek HV capsule filters (Pall Gelman Laboratory, Ann Arbor, Michigan) were used for collection and concentration of parasites.

Bacteria were quantified using membrane filtration (47 mm; 0.45- [mu]m pore size). When necessary, samples were diluted with phosphate-buffered saline, and all were filtered in triplicate. Total coliforms and fecal coliforms (FC) were enumerated according to Standard Methods (APHA et al., 1998)-total coliforms on mEndoLES agar at 37[degrees]C and fecal coliforms on mFC agar at 44.5[degrees]C. Entemcoccus spp. were enumerated using method 1600 (U.S. EPA, 1996) on mEI agar at 41[degrees]C. Clostridium perfringens was enumerated on m-CP plates (Acumedia Manufacturers Inc., Baltimore, Maryland) using anaerobic gas packs (BBL GasPak, Beckton Dickinson, Cockeysville, Maryland) at 45[degrees]C for 24 hours, followed by exposure to ammonium hydroxide fumes. Colonies that turned red or dark pink were enumerated as Clostridium perfringens (C. perfringens) (Bisson and Cabelli, 1979).

Coliphages were analyzed by the agar overlay method of Adams (1959) using two E. coli host strains in separate assays-E. coli HS (pFamp) R (ATCC #700891) and E. coli C3000 (ATCC #15597)-and were enumerated as plaque-forming units (PFU)/100 mL. Viruses were concentrated by filtration followed by organic flocculation (U.S. EPA, 1996). Cultivable enteric viruses were detected by the observation of cytopathic effects on recently passed (

Data were sorted by filter type and microorganism concentration. Microbial data were log^sub 10^-transformed before data analysis. In cases where values were below detection limits, the detection limit was used for statistical analyses. The performance of the individual filters was evaluated based on calculating microbial persistence through filtration (C/C^sub o^) and log^sub 10^ reduction (log^sub 10^ C^sub o^ – log^sub 10^ Q. The validity of using parametric statistics to evaluate log^sub 10^-transformed microbial data was tested using the Kolmogorov-Smirnov test for normality, with an alpha value of 0.05, using Graphpad Prism statistical software (Graphpad 2000, GraphPad Software Inc., San Diego, California). Statistically significant differences among filters were identified using either analysis of variance (ANOVA) or the Kruskal Wallis test (nonparametric). Post-tests included Tukey, Newman-Kuels, and Bonferroni for normally distributed data or the Dunn’s test for nonparametric analyses. Pairwise comparisons of filter performance were conducted using either t-tests (normally distributed data) or the MannWhitney test (nonparametric).

Comparison of Filtration Performance

To facilitate comparison of the filters, they were grouped into two categories. Group I includes filters that either use chemicals directly upstream of filtration (S11 [chlorine] and A48 [polymer]) or are influenced by surface filtration mechanisms (CLOTH), while group II includes deep-bed filters that contain sand (SU40, A30S10, and A24S48 [alum used in secondary clarifier]). Differences among the filters were evaluated based on influent concentrations and removal of cultivable enteric viruses, protozoan parasites (Giardia and Cryptosporidium), and bacterial indicators. Detailed analyses were conducted on group II filters to assess the effects of hydraulic loading rate and sand depth on reduction of pathogens.

A summary of the water quality associated with sampling of each facility is given in Table 3. As shown, C-BOD5 values were typically below 5 mg/L for most of the facilities, except A48. The TSS levels were typically below 4 mg/L. Nitrogen (ammonia, nitrite, and nitrate) and phosphorus levels varied with the extent of nutrient removal.

The overall performance of the individual treatment facilities for reduction of pathogens and indicators was assessed by comparing the average log^sub 10^ reduction for removal of each microbial parameter through each filter, as shown in Table 4. A different pattern was observed for removal of pathogens compared with the indicator organisms. The highest log^sub 10^ removal was associated with the filter that has the deepest layer of sand, hydraulic loading rates ranging from 30 to 36 m^sup 3^/m^sup 2^ * min (1.3 to 1.5 gpm/ft^sup 2^), and uses coagulation during secondary sedimentation (A24S48). Filtration without prechlorination achieved average cultivable enteric virus removals of 0.3 to 0.8 log^sub 10^ and parasite removals of 0.4 to 1.5 log^sub 10^. In contrast to surface-water-filtration systems designed to provide at least a 2- log reduction of protozoan pathogens (40 CFR 141.721), the log reductions associated with the filters tested in this project were typically below 2-log, and, in some cases, higher concentrations were observed in filter effluents than influents, resulting in negative log reductions.

In general, logic reduction of bacteria (coliforms, enterococci, and Clostridium) was 2- to 9-fold greater than log^sub 10^ reduction of pathogens, suggesting that, in contrast to Figure 1, the use of bacterial indicators as a monitoring tool may overpredict pathogen reduction. The highest overall removal of the indicator organisms was associated with S11, the prechlorinated filter (cumulative log^sub 10^ reduction of indicators of 13.7); this filter ranked second (after A28S48) for cumulative pathogen reduction. The cloth filter, which performs as a depth filter, had the second highest log^sub 10^ reduction of the tested indicators (cumulative log^sub 10^ reduction of indicators of 10.6) and the lowest reduction of protozoan pathogens. Among the other filters, the two filters that used coagulants either during secondary sedimentation (A28S48) or upstream of filtration (A48) had greater log^sub 10^ reduction of indicators and pathogens than the filters with no chemical pretreatment (A30S10 and SU40). In general, the weakest removal of indicators and pathogens was associated with the filter with the shallowest bed of sand and higher hydraulic loading rates (1.5 to 3.9 L/s [2.2 to 5.8 gpm/ft^sup 2^]), which did not apply upstream chemicals (A30S10). Comparison of Pathogen Removal Through Deep-Bed Filters that Contain Sand. Further analysis of data collected from facilities that operated deep-bed filters that contain sand (SU40, A30S10, and A24S48) was conducted to evaluate the filter collection efficiency, eta, and filter coefficient, lambda, for cultivable enteric viruses, Giardia, and Cryptosporidium (Figure 2). Based on theoretical values of collection efficiency (Figure 1), one would expect eta values for protozoan pathogens to be higher than observed values for viruses, as a result of differences in particle size and dominant removal mechanisms. In addition, variations in eta values among the filter types may reflect differences in apparent particle size, as a result of association with larger particles. The collection efficiency, eta, was significantly higher for Giardia and Cryptosporidium associated with the dual-media filter A30S10 compared with the other filters (p

Relationships between filter coefficients, lambda, and hydraulic loading rates for cultivable enteric viruses, Giardia, and Cryptosporidium are shown in Figure 3 (for hydraulic loading rates below 1.4 L/s [2 gpm/ft^sup 2^]). Even though the range of hydraulic loading rates that were tested in this project was controlled by operating parameters at the individual facilities (as opposed to systematic variation through experimental design), filter coefficients were inversely correlated to the hydraulic loading rate (p

The data were also evaluated to determine if there were any relationships between the depth of filtration media and log^sub 10^ reduction of indicators and pathogens from deep-bed filters operated without upstream chemical addition. For Giardia and Cryptosporidium, weak but significant correlations were observed. Increased depth was associated with improved log^sub 10^ reduction, which ranged from approximately 0.005 log^sub 10^ removal for Giardia (total) to approximately 0.01 log^sub 10^ removal for Cryptosporidium (total) per incremental centimeter of sand. While these correlations were statistically significant, these results suggest that it would require an additional 100 cm (3.3 ft) of sand to provide filtration capacity for an additional 0.5 log^sub 10^ reduction of Giardia and 1 log^sub 10^ reduction of Cryptosporidium. The additional headless associated with this increased depth makes this concept impracticable for control of protozoan pathogens, except in the context of groundwater recharge and soil-aquifer treatment. Sand depth did not have a significant effect on removal of cultivable enteric viruses or indicators (p > 0.05). No other significant correlations were observed between depth and microbial removal (total depth or anthracite depth).

Comparisons Between Indicator Organisms and Pathogens in Filter Effluents. Coliphages have been widely used as indicators for viral pathogens because of their similarity in size. To evaluate the validity of this relationship for the filtration data collected in this project, ANOVA tests were conducted to compare log^sub 10^ reductions and C/C^sub o^ for coliphages and cultivable enteric viruses for all filters tested in this project. No significant differences were observed for the individual organisms among the filters; therefore, the data were pooled for further comparisons. A comparison of log^sub 10^ reductions of coliphages and cultivable enteric viruses is given in Figure 4 in a boxplot format. In general, higher log^sub 10^ reductions were associated with coliphages than cultivable enteric viruses. In addition, coliphages and cultivable enteric viruses did not display consistent patterns of detect versus nondetect (see Table 2). The highest frequency of nondetects for cultivable enteric viruses (80%) was associated with S11, while the highest frequency of nondetects for coliphages (86%) was associated with A30S10. These discrepancies may be the result of the characteristics of the wastewater associated with each facility coupled with removal mechanisms upstream of filtration, resulting from differences in the biological treatment system.

To test the efficacy of using coliphage concentrations to predict cultivable enteric virus concentrations, filter effluent concentrations of cultivable enteric viruses are shown as a function of coliphage concentrations in Figure 5. The data are weakly correlated (R^sup 2^ = 0.4); however, in approximately 44% of the cases where cultivable enteric viruses were detected, coliphages were not detected (blue squares to the left of coliphage detection limit line). Thus, for this study, the absence of coliphages did not correspond to the absence of viruses. Perhaps this relationship could be improved if larger volume assays for phage were used (1 L versus 10 mL). Improvements in coliphage detection limits might provide an opportunity for revisiting this issue. However, because the sources of enteric viruses in wastewater differ from the sources of coliphages, the use of coliphages as a viral indicator should be verified for individual facilities.

Effect of Prechlorination on Removal of Indicator Organisms. To evaluate the effect of prechlorination, two filters at the facility that prechlorinates (S11) were isolated and operated in parallel. One filter received secondary effluent amended with approximately 2 mg/L chlorine, and the second filter received nonchlorinated secondary effluent. Hydraulic loading rates were similar for the two filters. Grab samples were collected at 2-hour intervals and evaluated for selected indicators. A comparison of C/C^sub o^ for fecal coliforms, enterococci, coliphages, and C. perfringens with and without prechlorination is shown in Figure 6. Prechlorination resulted in significantly greater removal of fecal coliforms and enterococci (p

Effect of Backwashing Practices on Removal of Protozoan Pathogens. To evaluate the effects of backwashing practices on removal of protozoan pathogens, samples were collected before and after backwashing of two filters-A48 and A24S48 (Figure 7). In each case, chlorinated filter effluent was used as a source of backwash water, and the filters were backwashed and returned to service using normal protocols. For both filters (A48 and A24S48), the concentrations of Giardia and Cryptosporidium increased following backwashing. There was almost a 1 logio increase associated with the dual media deep-bed filter (A24S48). While the cost of pathogen testing precluded additional testing of this issue in the context of this project, similar findings have been reported for filters used to process drinking water, where the presence of “remnant” particles or the lack of filter ripening can result in pathogen breakthrough following backwashing (Ahmad and Amirtharajah, 1998; Amburgey et al., 2004, 2005; Darby et al., 1991; Emelko et al., 2003; Hall and Fitzpatrick, 2000; Hozalski and Bouwer, 1998; Jegatheesan and Vigneswaran, 2000; Logan et al., 2001). Optimization of backwashing procedures by using extended terminal subfluidization backwash and/ or more gradual postbackwashing startup procedures, such as delayed start or slow start, have been suggested for improving the consistency of microbial removal through filtration (Amburgey et al., 2004).

Water Quality Variables. This study was based on evaluation of full-scale wastewater reclamation facilities that produce reclaimed water that is used for nonpotable public access reuse. In general, all of the facilities produce a high-quality effluent with BOD^sub 5^ and TSS values below 10 mg/L (see Table 3). Two of the facilities are designed for biological nutrient removal (SU40 and A24S48). Water quality data from each facility were evaluated to determine if there were any relationships between reduction of pathogens and indicators through filtration and secondary effluent characteristics. Removal of indicator bacteria, coliphages, cultivable enteric viruses, and Giardia by filtration were not significantly correlated to any of the monitored water quality parameters (p > 0.05). However, effluent concentrations of Cryptosporidium were weakly correlated to effluent BOD^sub 5^, TSS, ammonia, and phosphate (i.e., increased removal was associated with higher water quality) (linear correlation coefficients, R^sup 2^, were 0.2 [p = 0.02] for BOD^sub 5^, 0.6 [p = 0.0002] for TSS, and 0.3 [p = 0.03] for NH^sub 3^-N and PO^sub 4^-P). Summary and Recommendations

Granular media filtration of secondary effluent is widely used for production of reclaimed water. Key process variables include the filter media characteristics (i.e., depth and particle size), use of upstream chemicals (oxidants and/or coagulants), hydraulic loading rates, and backwash practices. In this study, data from six different full-scale wastewater reclamation facilities were compared to evaluate filter performance. Because of the nature of the grab sampling conducted in this project and differences in operating parameters of the individual filters, it was not possible to sample each filter under similar operating conditions, thus introducing confounding variables that may affect data interpretation. For example, the facility that uses prechlorination (S11) also had the lowest hydraulic loading rates and smallest media, making it difficult to separate the effects of prechlorination and hydraulic loading rate under routine plant operations.

Filtration of secondary effluent differs from current practices that have been adopted for filtration of surface water. Key differences include the degree to which upstream chemical amendments are used, turbidity limits, and filter backwashing and postbackwashing protocols. While the higher ionic strength and organic content of secondary effluents may serve to enhance particle removal through filtration, the log^sub 10^ reduction of pathogens observed through this study was significantly lower than values reported for surface water filtration. Based on the data from this study and others, some degree of pathogen reduction was observed through all of the filters that were tested; however, the results suggest that optimization of filtration operations (including pretreatment, loading rates, and backwashing practices) can improve pathogen reduction through filtration and thereby improve the overall microbial safety of reclaimed water. In addition, because of differences in source characteristics and removal mechanisms through biological treatment and filtration, removal patterns associated with traditional bacterial indicators (coliforms) and coliphages do not correspond to reduction of protozoan pathogens and cultivable enteric viruses. Adoption of alternative monitoring approaches and extrapolation of design and operation strategies from surface-water filtration may result in more robust filtration systems for the production of reclaimed water.

Conclusions

Based on the data derived from this project, filter loading rates, use of upstream chemical addition, backwashing practices, and water quality of the secondary effluent can influence the efficiency of granular media filtration for removal of pathogens from secondary effluent. The major conclusions are as follows:

(1) The use of traditional models of single collector efficiency and filter coefficients can be applied to provide a qualitative comparison of pathogen removal from granular media filtration of secondary effluent; however, supplemental water quality data and operations information are needed to optimize filter performance.

(2) Higher rates of removal through granular media filtration were observed for conventional bacterial indicators compared with pathogens, suggesting that the use of bacterial indicators is not an effective predictor of filtration performance for pathogen reduction.

(3) The suite of microbiological indicators tested in this project did not correlate to concentrations, removal (log reduction), or presence/absence of cultivable enteric viruses, Giardia, or Cryptosporidium in the filtered secondary effluents tested in this project. Coliphage removal or concentrations did not correspond to removal or concentrations of cultivable enteric viruses,

(4) For deep-bed filters that contained sand and were not prechlorinated, removal of pathogens was inversely proportional to hydraulic loading rates and directly proportional to the depth of sand.

(5) The use of prechlorination resulted in improved removal of microorganisms that are susceptible to chlorine inactivation (coliforms and enterococci) but had minimal effect on chlorineresistant bacteria, such as the endospore-forming Clostridium perfringens.

(6) The return of filters to service directly following backwashing may result in higher effluent concentrations of Giardia and Cryptosporidium until filter ripening occurs. The extent of the increase varies with filter characteristics (media, depth, and upstream chemical use). Continuous backwash filters may not experience these performance fluctuations.

(7) Water quality variables that were correlated to removal of Cryptosporidium through granular media or cloth filtration include effluent concentrations of BOD^sub 5^, TSS, ammonia nitrogen, and phosphate. No statistically significant water quality relationships were associated with removal of the other microorganisms tested in this project (cultivable enteric viruses, Giardia, and indicator organisms).

(8) For the filters tested in this project, removal of conventional indicator organisms through filtration tended to exhibit different patterns than those observed for viral and protozoan pathogens, suggesting that alternative monitoring practices may be needed to assess and manage microbial health risks associated with reclaimed water applications.

Credits

This study was funded by the Water Environment Research Foundation (Alexandria, Virginia) as project 00-PUM2T. Personnel from wastewater reclamation facilities in Phoenix, Arizona; Santa Barbara, California; and Eustis, St. Petersburg, and Hillsborough County, Florida assisted with sample collection and data compilation. Vasanta Chivulaka, Molly McLaughlin, Stefica Depovic, and Angela Gennaccaro conducted microbiological testing for the project.

Submitted for publication November 12, 2006; revised manuscript submitted September 9, 2007; accepted for publication December 11, 2007.

The deadline to submit Discussions of this paper is October 15, 2008.

References

Adams, M. H. (1959) Bacteriophages; Interscience Publishers: New York.

Ahmad, R.; Amirtharajah, A. (1998) Detachment of Particles During Filter Backwashing. J. Am. Water Works Assoc., 90 (12), 74-85.

Amburgey, J. E.; Amirtharajah, A.; Brouckaert, B. M.; Spivey, N. C. (2004) Effect of Washwater Chemistry and Delayed Start on Filter Ripening. J. Am. Water Works Assoc., 96 (1), 97-110.

Amburgey, J. E.; Amirtharajah, A.; York, M. T.; Brouckaert, B. M.; Spivey, N. C.; Arrowwood, M. J. (2005) Comparison of Conventional and Biological Filter Performance for Cryptosporidium and Microsphere Removal. J. Am. Water Works Assoc., 97 (12), 77-91.

American Public Health Association; American Water Works Association; Water Environment Federation (1998) Standard Methods for the Examination of Water and Wastewater, 20th ed.; American Public Health Association: Washington, D.C.

Bisson, J. W.; Cabelli, V. J. (1979) Membrane Filter Enumeration for Clostridium Perfringens. Appl. Environ. Microbiol, 37 (1), 55- 66.

Bourgeous, K. N.; Riess, J.; Tchobanoglous, G.; Darby, J. L. (2003) Performance Evaluation of a Cloth Media Filter for Wastewater Reclamation. Water Environ. Res., 75 (6), 532-538.

Brown, D. G.; Abramson, A. (2006) Collision Efficiency Distribution of a Bacterial Suspension Flowing Through Porous Media and Implications for Field Scale Transport. Water Res., 40, 1591- 1598.

Bustamante, H. A.; Shanker, S. R.; Pashley, R. M.; Karaman, M. E. (2001) Interaction Between Cryptosporidium Oocysts and Water Treatment Coagulants. Water Res., 35 (13), 3179-3189.

Camp, T. R. (1964) Theory of Water Filtration. ASCE J. Sanit. Eng. Div., 90 (SA4), 1-30.

Darby, J. L.; Lawler, D. F.; Wilshusen, T. P. (1991) Depth Filtration of Wastewater: Particle Size and Ripening. J. Water Pollut. Control Fed., 63 (2), 228-238.

Edzwald, J. K.; Tobiason, J. E.; Parento, L. M.; Kelley, M. B.; Kaminski, G. S.; Dunn, H. J.; Galant, P. B. (2000) Giardia and Cryptosporidium Removals by Clarification and Filtration Under Challenge Conditions. J. Am. Water Works Assoc., 92 (12), 70-84.

Elimelech, M. (1992) Predicting Collision Efficiencies of Colloidal Particles in Porous Media. Water Res., 26 (1), 1-8.

Elimelech, M.; O’Melia, C. R. (1990) Kinetics of Deposition of Colloidal Particles in Porous Media. Environ. Sci. Technol., 24, 1528-1536.

Emelko, M. B.; Huck, P. M.; Coffey, B. M. (2005) A Review of Cryptosporidium Removal by Granular Media Filtration. J. Am. Water Works Assoc., 97 (12), 101-115.

Emelko, M. B.; Huck, P. M.; Douglas, I. P. (2003) Cryptosporidium and Microsphere Removal During Late-in-Cycle Filtration. J. Am. Water Works Assoc., 95 (5), 173-182.

England, S. K.; Darby, J. L.; Tchobanoglous, G. (1994) Continuous- Backwash Upflow Filtration for Primary Effluent. Water Environ. Res., 60 (2), 145-152.

Foppen, J. W.; Okletey, S.; Schijven J. F. (2006) Effect of Goethite Coating and Humic Acid on the Transport of Bacteriophage PRD1 in Columns of Saturated Sand. J. Contam. Hydrol., 85 (3-4), 287- 301.

Ghosh, M.; Jordan, T. A.; Porter, R. L. (1975) Physicochemical Approach to Water and Wastewater Filtration. ASCE J. Sanit. Eng. Div., 101 (EE1), 71-86.

Glasgow, G. D. E.; Wheatley, A. D. (1999) Observations of Rapid Flow Rate Disturbances in Drinking Water Filters and Their Effect on Solids Removal. J. Water SRT-Aqua, 48, 257-262.

Habibian, M. T.; O’Melia, C. R. (1975) Particles, Polymers, and Performance in Filtration. ASCE J. Environ. Eng. Div., 101 (EE4), 567-583.

Hall, D.; Fitzpatrick, C. S. B. (2000) Suspension Concentration Profiles During Rapid Gravity Filter Backwashing. Chem. Eng. J., 80, 197-201.

Harrington, G. W.; Xagoraraki, I.; Assavasilavasukul, P.; Standridge, J. H. (2003) Effect of Filtration Conditions on Removal of Emerging Pathogens. J. Am. Water Works Assoc., 95 (12), 95-104.

Harvey, R. W.; Garabedian, S. P. (1991) Use of Colloid Filtration Theory in Modeling Movement of Bacteria Through a Contaminated Sandy Aquifer. Environ. Sci. Technol., 25 (1), 178-185. Harwood, V. J.; Levine, A. D.; Scott, T. M.; Chivukula, V.; Lukasik, J.; Farrah, S. R.; Rose, J. B. (2005) Validity of the Indicator Organism Paradigm for Pathogen Reduction in Reclaimed Water and Public Health Protection. Appl. Environ. Microbiol., 71 (6), 3163-3170.

Hozalski, R. M.; Bouwer, E. J. (1998) Deposition and Retention of Bacteria in Backwashed Filters. J Am. Water Works Assoc., 90 (1), 71- 85.

Hsu, B.-M.; Huang, C. (2002) Influence of Ionic Strength and pH on Hydrophobicity and Zeta Potential of Giardia and Cryptosporidium. Colloids Surf. A, 201, 210-206.

Hsu, B.-M.; Huang, C.; Pan, J. R. (2001) Filtration Behaviors of Giardia and Cryptosporidium-Ionic Strength and pH Effects. Water Res., 35 (16), 3777-3782.

Iwasaki, T. (1937) Some Notes on Sand Filtration. J. Am. Water Works Assoc., 29 (10), 1591-1602.

Jegatheesan, V.; Vigneswaran, S. (2000) Transient Stage Deposition of Submicron Particles in Deep Bed Filtration Under Unfavorable Conditions. Water Res., 34 (7), 2119-2131.

Jiminex, B.; Chavex, A.; Leyva, A.; Tchobanoglous, G. (2000) Sand and Synthetic Medium Filtration of Advanced Primary Treatment Effluent from Mexico City. Water Res., 14 (2), 473-480.

Kavanaugh, M. C.; Toregas, G.; Chung, M.; Pearson, E. A. (1978) Particulates and Trace Pollutant Removal by Depth Filtration. Prog. Water Technol., 10 (5/6), 197-213.

Koivunen, J.; Siitnonen, A.; Heinonen-Tanski, H. (2003) Elimination of Enteric Bacteria in Biological-Chemical Wastewater and Tertiary Filtration Units. Water Res., 37, 690-698.

Logan, A. J.; Stevik, T. K.; Siegrist, R. L.; Ronn, R. M. (2001) Transport and Fate of Cryptosporidium parvum Oocysts in Intermittent Sand Filters. Water Res., 35 (18), 4359-4369.

Mosher, R. R.; Hendricks, D. W. (1986) Rapid Rate Filtration of Low Turbidity Water Using Field-Scale Pilot Filters. J. Am. Water Works Assoc., 78 (12), 42-51.

Nasser, A.; Weinberg, D.; Dinoor, N.; Fattal, B.; Adin, A. (1995) Removal of Hepatitis A Virus (HAV), Poliovirus, and MS2 Coliphage by Coagulation and High Rate Filtration. Water Sci. Technol., 31 (5- 6), 63-08.

Nieminski, E. C.; Ongerth, J. E. (1995) Removing Giardia and Cryptosporidium by Conventional Treatment and Direct Filtration. J. Am. Water Works Assoc., 87 (9), 96-106.

O’Melia, C. R.; Ali, W. (1978) The Role of Retained Particles in Deep Bed Filtration. Prog. Water Technol., 10 (5/6), 167-182.

Ongerth, J. E.; Pecoraro, J. P. (1996) Electrophoretic Mobility of Cryptosporidium Oocysts and Giardia Cysts. J. Environ. Eng., 122 (3), 228-231.

Patania, N. L.; Jacangelo, J. G.; Cummings, L.; Wilczak, A.; Riley, K.; Oppenheimer, J. (1995) Optimization of Filtration for Cyst Removal; AWWA Research Foundation: Denver, Colorado.

Reddi, L. N. (1997) Particle Transport in Soils: Review of Significant Processes in Infrastructure Systems. J. Infrastruct. Syst., 3 (2), 78-86.

Redman, J. A.; Grant, S. B.; Olson, T. M.; Estes, M. K. (2001) Pathogen Filtration, Heterogeneity and the Potable Reuse of Wastewater. Environ. Sci. Technol., 35 (9), 1798-1805.

Rose, J. B.; Farrah, S. R.; Harwood, V. J.; Levine, A. D.; Lukasik, J.; Menendez, P.; Scott, T. M. (2004) Reduction of Pathogens, Indicator Bacteria, and Alternative Indicators by Wastewater Treatment and Reclamation Processes, 00-PUM-2T; Water Environment Federation: Alexandria, Virginia.

Slifko, T. R.; Huffman, D. E.; Rose, J. B. (1999) A Most- Probable-Number Assay for Enumeration of Infectious Cryptosporidium Parvum Oocysts. Appl. Environ. Microbiol, 65 (9), 3936-3941.

Sobsey, M. D. (1989) Inactivation of Health-Related Microorganisms in Water by Disinfection Processes. Water Sci. Technol., 21, 179-195.

Stevenson, D. G. (1997) Flow and Filtration Through Granular Media-The Effect of Grain and Particle Size Dispersion. Water Res., 31 (2), 310-322.

Swertfeger, J.; Metz, D. H.; DeMarco, J.; Braghetta, A.; Jacangelo, J. A. (1999) Effect of Filter Media on Cyst and Oocyst Removal. J. Am. Water Works Assoc., 91 (9), 90-100.

Tchobanoglous, G.; Eliassen, R. (1970) Filtration of Treated Sewage Effluent. ASCE J. Sanit. Eng. Div., 96 (SA2), 243-265.

Trussell, R. R.; Chang, M. (1999) Review of Flow Through Porous Media as Applied to Head Loss Through Filters. J. Environ. Eng., 125 (11), 998-1006.

U.S. Environmental Protection Agency (2004) Guidelines for Water Reuse, EPA-625/R-04-108; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (1995) ICR Protozoan Method for Detecting Giardia Cysts and Cryptosporidium Oocysts in Water by a Fluorescent Antibody Procedure, EPA-814/B-95-003; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (2003) Long Term 2 Enhanced Surface Water Treatment Rule Toolbox Guidance Manual, EPA-815/ D-03- 009; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (2005) Method 1623: Cryptosporidium and Giardia in Water by Filtration/IMS/FA, EPA-815/ R-05-002; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (1996) Microbial Laboratory Manual. ICR: Information Collection Rule. U.S. Environmental Protection Agency: Cincinnati, Ohio.

U.S. Environmental Protection Agency (2006) Ultraviolet Disinfection Guidance Manual for the Final Long-Term 2 Enhanced Surface Water Treatment Rule, EPA-815/R-06-007; U.S. Environmental Protection Agency: Washington, D.C.

Wang, J. Z.; Summers, R. S.; Miltner, R. J. (1995) Biofiltration Performance: Part 1, Relationship to Biomass. J. Am. Water Works Assoc., 87 (12), 55-63.

Xagoraraki, I.; Harrington, G. W.; Assavasilavasukul, P.; Standridge, J. H. (2004) Removal of Emerging Waterborne Pathogens and Pathogen Indicators by Pilot Scale Conventional Treatment. J. Am. Water Works Assoc., 96 (5), 102-113.

Yao, K.; Habibian, M. T.; O’Melia, C. R. (1971) Water and Waste Water Filtration: Concepts and Applications. Environ. Sci. Technol., 5 (11), 1105-1112.

York, D. W.; Walker-Coleman, L. (2000) Pathogen Standards for Reclaimed Water. Water Environ. Technol., 12 (1), 58-61.

Audrey D. Levine1*, Valerie J. Harwood2, Samuel R. Farrah3, Troy M. Scott4, Joan B. Rose5

1 Office of Research and Development, U.S. Environmental Protection Agency, Washington D.C.

2 Department of Biology, University of South Florida, Tampa, Florida.

3 Department of Microbiology & Cell Science, University of Florida, Gainesville, Florida.

4 BCS Laboratories Inc., Miami, Florida.

5 Department of Crop and Soil Sciences, Michigan State University, East Lansing, Michigan.

* Office of Research and Development, U.S. Environmental Protection Agency, 1200 Pennsylvania Avenue NW 8101R, Washington D.C. 20460; e-mail: Levine.Audrey@epa.gov.

Copyright Water Environment Federation Jul 2008

(c) 2008 Water Environment Research. Provided by ProQuest LLC. All rights Reserved.