Evaluation of microbiological risks associated with direct potable reuse

Evaluation of microbiological risks associated with direct potable reuse

ARTICLE IN PRESS JID: MRAN [m5G;September 8, 2016;21:39] Microbial Risk Analysis 0 0 0 (2016) 1–12 Contents lists available at ScienceDirect Micr...

1MB Sizes 0 Downloads 49 Views

ARTICLE IN PRESS

JID: MRAN

[m5G;September 8, 2016;21:39]

Microbial Risk Analysis 0 0 0 (2016) 1–12

Contents lists available at ScienceDirect

Microbial Risk Analysis journal homepage: www.elsevier.com/locate/mran

Evaluation of microbiological risks associated with direct potable reuse Jeffrey A. Soller a,∗, Sorina E. Eftim b, Isaac Warren b, Sharon P. Nappier c a

Soller Environmental, LLC, 3022 King St., Berkeley, CA 94703, USA ICF International, 9300 Lee Highway, Fairfax, VA 22031, USA c U.S. Environmental Protection Agency, Office of Water, 1200 Pennsylvania Avenue, NW, Washington, DC 20460, USA b

a r t i c l e

i n f o

a b s t r a c t

Article history: Available online xxx

This work evaluates the potential microbial risks associated with various direct potable reuse (DPR) treatment train combinations for recycled water. The assessment methodology leverages readily available peer-reviewed pathogen density and treatment process removal data and extends a previously published statistical approach. The results illustrate clear quantitative human health-based advantages for DPR projects in which product water is introduced into the raw water supply upstream of a conventional drinking water treatment facility, compared to those in which product water is introduced directly into a potable water supply distribution system. The results also indicate that a single day can drive annual risks, highlighting the need for robust and reliable on-line monitoring of unit treatment processes within DPR facilities. The methodology is adaptable to other DPR treatment trains and could be iteratively refined as additional data become available. This work will be useful to federal and state regulators considering DPR as source water, state and local decision makers as they consider whether to permit a particular DPR project, and design engineers as they consider which unit treatment processes should be employed for particular projects. © 2016 Elsevier B.V. All rights reserved.

1. Introduction

have been asked to permit recycled water projects and municipalities are moving forward with IPR and DPR projects that use a variety of multiple barrier treatment strategies for contaminant removal (Dahl, 2014). For example, California has developed microbial regulations for groundwater replenishment IPR projects. California’s groundwater IPR regulations include a requirement of 12-log removal of enteric viruses, and 10-log removal of Cryptosporidium spp. and Giardia (also known as the “12-10-10 Rule” ), using at least three treatment processes (from wastewater treatment through advanced treatment) with no single process allowed a credit greater than 6-log (CDPH, 2011; NWRI, 2013). California may also use this 12-10-10 Rule for IPR surface water augmentation regulations and is considering statewide DPR regulations. The 12-10-10 pathogen log-reduction values are based on a risk goal of 1 infection per 10,0 0 0 people per year and are derived from the maximum reported densities of culturable enteric viruses, Giardia lamblia, and Cryptosporidium spp. found in raw sewage (Macler and Regli, 1993; U.S. EPA, 1998; Tchobanoglous et al., 2003; Sinclair et al., 2015). Recent literature, however, suggests that different reference viruses and new dose-response models also should be considered in the aforementioned log-reduction goal calculations (Pouillot et al., 2015; Messner and Berger, 2016; Messner et al., 2014; Teunis et al., 2008). For DPR projects in the State of Texas, the minimum log removal and/or inactivation targets are 8-log for enteric virus, 5.5-log for Cryptosporidium spp., and 6-log for Giardia for the ad-

As population growth, urbanization, droughts, and climate change continue to impact natural resources, water reuse is an increasingly important water supply option worldwide. For example, in 2009 the State of California approved a Recycled Water Policy, with the goal to increase the use of recycled water, over 2002 usage levels, by 2 million acre-foot/year by 2030 (Tchoganoglous et al., 2011). Moreover, there is a growing interest and willingness to use recycled wastewater as a drinking water source (WRRF, 2015; Richmond, 2016). The introduction of recycled wastewater into a potable water supply has, to date, most commonly included an environmental buffer, such as an aquifer or surface water reservoir. This reuse strategy is referred to as indirect potable reuse (IPR). Interest is also growing in direct potable reuse (DPR) in which recycled water is introduced into a potable water supply distribution system or into the raw water supply immediately upstream of a conventional drinking water treatment facility, without inclusion of an environmental buffer (WRRF, 2015). Currently, there are no national regulatory recommendations in the United States for the potable reuse of water. Nevertheless, States such as Texas, New Mexico, Arizona, California, and Virginia



Corresponding author. E-mail address: [email protected] (J.A. Soller).

http://dx.doi.org/10.1016/j.mran.2016.08.003 2352-3522/© 2016 Elsevier B.V. All rights reserved.

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

JID: MRAN 2

ARTICLE IN PRESS

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

vanced treatment system (not including the wastewater treatment plant). These targets are considered a starting point for the approval process and may be revised on a case-by-case basis taking into consideration data collected from the wastewater effluent in question. Texas also uses a health risk goal of 1 infection per 10,0 0 0 people per year as a benchmark (TWDB, 2014). Due to nature of the source water for DPR, there is a need to quantitatively evaluate the health risks associated with exposure to microbial contaminants resulting from implementation of the various multi-barrier treatment strategies under consideration. Given the low levels of target pathogens in product water that would be needed to comply with the health goals (i.e., ∼10−7 L−1 ), it is not feasible to determine whether adequate treatment is occurring via monitoring product water for pathogens from advanced water treatment facilities (AWTF) (Regli et al., 1991). To our knowledge, a quantitative evaluation of the microbial risks associated with the various multi-barrier IPR and DPR treatment strategies has not been conducted. To address this need, we applied and extended the probabilistic modeling approach suggested by Haas and Trussell (1998) and demonstrated by Olivieri et al. (1999). The objectives of this work are to: 1) document reference pathogen occurrence in raw wastewater and removal from various wastewater, AWTF, and drinking water unit treatment processes; 2) provide a methodology that can be used and updated for evaluating the microbial risks of infection associated with reference pathogens from DPR treatment trains (TT); and 3) provide context regarding the adequacy of the State log reduction goals under consideration for DPR. 2. Methods 2.1. DPR treatment trains This study encompasses four representative AWTF TT configurations that could be considered for DPR projects. These configurations are consistent with those recommended by the National Water Research Institute (NWRI) DPR expert panel (NWRI, 2013; NWRI, 2015). For three of the configurations (TT 1–3, Fig. 1), the DPR product water was assumed to directly enter the drinking water distribution system after AWTF treatment. In the fourth configuration (TT4, Fig. 1), DPR product water undergoes further treatment via a conventional drinking water treatment plant (DWTP) prior to entering the community drinking water distribution system. Additionally, municipalities and States are interested in identifying and evaluating treatment schemes that do not include reverse osmosis (RO) due to the difficulty and costs associated with disposal of RO brine concentrate (TWDB, 2014). Therefore, configurations without RO are also included in the analysis (TT3 and TT4, Fig. 1). TT1 is defined as a conventional activated sludge wastewater treatment plant (WWTP) that produces non-disinfected secondary effluent used as the feed water to an AWTF. The AWTF consists of microfiltration (MF), RO, advanced oxidation (UVAOP) (UV and hydrogen peroxide quenching), and an engineered storage buffer (ESB) with free chlorine disinfection (ESBCl). TT2 uses the same conventional secondary WWTP as the feed water to an AWTF that consists of ozonation, biologically active filtration (BAF), MF, RO, and UVAOP. TT3 uses the conventional secondary WWTP as the feed water to an AWTF that consists of ozonation, BAF, ultrafiltration (UF), UVAOP, and an ESBCl. TT4 is TT3 followed by a conventional DWTP comprised of flocculation, sedimentation, filtration, and disinfection via chlorination. This DWTP does not specifically address taste and odor, iron and manganese reduction, and/or the need to reduce disinfection by-product formation, all of which require enhanced water treatment that would be specific to the source water.

Alternatives to the above base TT configurations were also considered to evaluate the potential public health implications of DPR design choices. In one alternative, the ultraviolet (UV) disinfection unit processes are assumed to be operated in a manner consistent with conventional wastewater treatment disinfection (dose of 12 mJ/cm2 or less), rather than the use of UVAOP (dose of 800 mJ/cm2 ) typically employed for the purposes of disinfection by-product destruction. Each of the base TT configurations is considered for this alternative. In a second alternative, the impact of pathogen dose-response selection is evaluated by using recently published dose-response relationships for two of the reference pathogens, norovirus (NoV) and Cryptosporidium spp. (Messner and Berger, 2016; Messner et al., 2014). TT1a and TT1b (Fig. 1) base configurations are evaluated with the alternative dose-response models. 2.2. Pathogens The reference pathogens in this study include NoV, adenovirus (AdV), Cryptosporidium spp., Giardia lamblia, Campylobacter jejuni, and Salmonella enterica. Together these pathogens make up a large portion of all non-foodborne illnesses from known pathogens in the United States (calculated based on data from Mead et al. (1999) and Scallan et al. (2011)), are representative of other pathogens potentially of concern from the waterborne exposure route (Soller et al., 2010a,b; U.S. EPA, 2012, 2014), and have published corresponding dose-response relationships (Teunis et al., 2008; Crabtree et al., 1997; Teunis et al.; 2005; Haas et al., 1999; U.S. EPA, 2006a). The use of reference pathogens is an accepted practice in the field of Quantitative Microbial Risk Assessment (QMRA) (Regli et al., 1991; U.S. EPA, 2014; Roser and Ashbolt, 2007; Schoen et al., 2011; Soller and Eisenberg, 2008; Soller et al., 2003) to represent the potential adverse health effects of members of each broader microbial group, as well as the infectivity of known and unknown members of each microbial group (WHO, 2004). The corresponding standard dose-response models and parameter values are summarized in Table 1, along with recently published newer dose-response models for Cryptosporidium spp. and NoV (Messner and Berger, 2016; Messner et al., 2014). The dose-response relationship for Campylobacter is based on an outbreak study for which doses were inferred rather than measured (Teunis et al., 2005). It suggests much higher infectivity than a previous clinical trial study’s interpretation for which doses were measured (Medema et al., 1996). This dose response relationship was selected here as a precautionary approach to ensure that the risk of infection from Campylobacter was not underestimated. As indicated above, the relative impact of the dose-response relationship selections are evaluated through sensitivity analyses using methods published previously (Soller et al., 2015a). 2.3. Treatment efficacy of DPR unit processes for pathogens Peer-reviewed literature were collected and summarized to characterize the density of each of the reference pathogens in raw wastewater and the reduction of each of the reference pathogens across each of the individual unit treatment processes under consideration (Fig. 1). For each unit process that is “dose dependent” (such as disinfection via UV light, ozone, and free chlorine), a fixed dose was selected and applied for all pathogens across that particular unit process (discussed below in the results section). The reduction of each pathogen across each of the unit processes was generally assumed to follow a uniform distribution of the log reductions with minimum and maximum values consistent with the literature review findings (Eisenberg et al., 20 02, 20 06). While the literature review captures representative reductions of each reference pathogen across each unit process, it is not intended

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

ARTICLE IN PRESS

JID: MRAN

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

3

Treatment train 1.

Wastewater Treatment

MF

UV

RO

ESB + Cl2

Treatment train 2.

Wastewater Treatment

O3

BAF

MF

UV

RO

Treatment train 3.

O3

Wastewater Treatment

BAF

UF

ESB + Cl2

UV

Treatment train 4.

Wastewater Treatment

BAF

O3

Cl2

UF

Filter

UV

ESB + Cl2

Flocculaon/ Sedimentaon

Fig. 1. DPR treatment trains evaluated, legend: MF – Microfiltration, RO – Reverse Osmosis, UV – Ultraviolet disinfection, ESB + Cl2 – Engineered Storage Buffer with Free Chlorine (ESBCl), O3 – Ozonation, BAF – Biologically Active Filtration, UF – Ultrafiltration, Cl2 – Disinfection via chlorine. Table 1 Dose-response models and parameter values. Reference pathogen

Dose-response model

Equation2

Parameter values

Adenovirus Campylobacter jejuni Cryptosporidium spp.

Exponential (Crabtree et al., 1997) Hypergeometric (Teunis et al., 2005) Exponential (U.S. EPA 2006a) Fractional Poisson (Messer and Berger, 2016)1 Exponential (Haas et al., 1999) Hypergeometric (Teunis et al., 2008) Fractional Poisson (Messer et al., 2014)1 Beta-Poisson (Haas et al., 1999)

1 − e−rd 1-1 F1 (α , α + β , −d ) 1 − e−rd d P ∗ ( 1 − e− α ) 1 − e−rd 1-1 F1 (α , α + β , −d ) d P ∗ ( 1 − e− α ) 1 − (1 + βd )−α

0.4172 0.024, 0.011 0.09 0.737, 1 0.0199 0.04, 0.055 0.72, 1106 0.3126, 2884

Giardia lamblia Norovirus Salmonella enterica 1 2

Used in sensitivity analysis. In equations d represents dose.

to be exhaustive. These data and associated distribution can be updated in the future as additional data become available. 2.4. Numerical simulations A stochastic, static QMRA methodology was used to estimate infection from pathogenic microorganisms through ingestion of DPR product water for the DPR TTs described above (Soller and Eisenberg, 2008). Individuals are assumed to ingest a fixed amount of product water daily. This overall approach, referred to as quantitative relative risk assessment, has been used previously for the purpose of evaluating the relative risks associated with recycled water projects (Soller et al., 2015b, 20 0 0; Soller and Nellor, 2011a,b). Two-step Monte Carlo numerical simulations (Fishman, 1995) were conducted using Mathcad v13 (PTC, Inc.). For each TT, the first step was to compute an array of daily risk estimates (n = 365) for each reference pathogen. Pathogen specific daily risks (RPrisk ) are computed through standard QMRA calculations using the estimated daily density of each pathogen in the DPR product water, the volume of water ingested, and the corresponding doseresponse relationships (Haas et al., 1999). The volume of water

ingested was estimated to be 2.5 L/d which represents the 90th percentile of per capita water ingestion for adults (U.S. EPA, 2011). This ingestion value is intentionally precautionary. The estimated daily concentration of each pathogen in the DPR product water is computed as the product of randomly selected raw wastewater pathogen density and randomly selected attenuation values across each unit process following the approach originally suggested by Haas and Trussell (1998).

RPProducti = RPIn f luenti x 10−W W _RPi x

n 

10−DPRUnit Pr ocess_RPi

(1)

1

Where RPProducti = Reference Pathogen (i) concentration in DPR case study product water RPInfluenti = Reference Pathogen (i) concentration in raw wastewater WW_RPi = Reduction (log10 units) of Reference Pathogen i across conventional wastewater (WW) (starting with raw wastewater and including treatment through secondary treatment) DPRUnitProcess_RPi = Reduction (log10 units) of Reference Pathogen i across DPR unit process

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

ARTICLE IN PRESS

JID: MRAN 4

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

n = number of DPR unit processes (case study specific – Fig. 1) Cumulative daily risks from all of the evaluated pathogens were then computed as follows (Regli et al., 1991).

CumDailyRisk = 1 −

[m5G;September 8, 2016;21:39]

i 1

(1 − R Pr iski )

(2)

Where RPriski = Daily risk of infection from Reference Pathogen i Those daily risks are then combined in a similar manner to generate a single cumulative annual risk estimate:

CumAnnualRisk = 1 −

365  j=1

1 − CumDailyRisk j



(3)

The second step iterates the first step 10 0 0 times to generate a distribution of estimated annual risks. This assessment process is repeated for each of the TT configurations and alternatives. The cumulative annual risk distributions are then compared to the benchmark risk level of 1 infection per 10,0 0 0 people per year for finished drinking water (Macler and Regli, 1993; U.S. EPA, 1989). We also determine whether there are statistically significant differences between the results from various simulations using a MannWhitney-Wilcoxon non-parametric test (Mann and Whitney, 1947). 3. Results 3.1. Literature review for pathogen densities and reductions across unit treatment processes Peer reviewed scientific literature were obtained to quantify the density of reference pathogens in raw wastewater, and the attenuation of each reference pathogen across each individual unit treatment process. When disinfecting water with chlorine or ozone, disinfection residual concentration (mg/l) multiplied by contact time (min) (CT) is determined to indicate the achievement of a desired treatment efficacy. When disinfecting water with ultraviolet light, a UV dose in mJ/cm2 is determined to indicate the achievement of a desired treatment efficacy. Several factors can affect disinfectant efficacy such as pH, temperature, and turbidity. The studies reviewed did not typically detail all of these factors, with exception of controlled bench-scale experiments. Based on the range of doses and results found for each disinfection method, a threshold dose that is representative of AWTF use was selected: For chlorine, Ct of 12 mg-min/L or below is used; for UVAOP in an advanced water treatment facility, a UV dose of 800 mJ/cm2 or below is used; for UV in a typical wastewater disinfection application, a UV dose of 12 mJ/cm2 or below is used; and for ozone, a threshold Ct of 1.0 mg-min/L or below is used. Raw wastewater microbial densities. Cryptosporidium spp. has been detected in US wastewaters in numerous studies. Cryptosporidium spp. has been recognized to occur at levels as low as 0.3 or high as 5.0 × 104 oocysts/L (Yang et al., 2015; Crockett, 2007; Harwood et al., 2005; Nasser, 2016). Giardia spp. densities in raw wastewater range between 3.2–1.0 × 104 cysts/L (Harwood et al., 2005; Sykora et al., 1991; Wallis et al., 1996). Campylobacter spp. densities in raw wastewater range between 9.0 × 102 –4.0 × 104 Most Probable Number (MPN)/L (Stampi et al., 1993) and Salmonella spp. levels vary between 3.0–1.1 × 103 Colony Forming Units (CFU)/L (Lemarchand and Lebaron, 2003). The range of reported viral concentrations is wider than protozoa or bacteria. To account for the wide variability in the literature of NoV (Pouillot et al., 2015), a normal distribution for NoV based on occurrence in raw wastewater was created - NoV concentrations are estimated to be present at levels of 103.76 ± 100.93 copies/L (Eftim et al., 2016). Too few studies were available to develop an occurrence distribution for AdV. Thus, a range of 56–6.9 × 103 Infectious Units (IU)/L for AdV was used (Eftim et al., 2016; Hewitt et al., 2011; Hurst et al., 1988).

Conventional secondary wastewater treatment. There are numerous approaches that can be used to produce secondary treated wastewater. Most consist of physical removal, primary settling, anaerobic and/or aerobic activated sludge biological treatment, and secondary clarification or settling. For this evaluation, secondary treatment that does not include disinfection or filtration was assumed. Activated sludge in an oxidation ditch or in extended aeration tanks achieved 0.7–1.5-log reduction for Cryptosporidium spp. and 0.5–1.5-log reduction for Giardia spp. (Cheng et al., 2009). However, removals as high as 3.3-log are possible for Giardia spp., e.g., following activated sludge treatment (TaranBenshoshan et al., 2015). Conventional secondary treatment of viruses can achieve 0.9–3.2-log removal of AdV and 0.8–3.7-log removal of NoV (Pouillot et al., 2015; Francy et al.; 2012). Although these published log removal values are used, it is noteworthy that secondary treatment also has been reported to be ineffective at removing some viruses and protozoa (WHO, 2006). Secondary treatment of bacteria can achieve between 1.3–1.7-log removal for Salmonella typhimurium (Zhou et al., 2015) and treatment with trickling filters, activated sludge plants, and oxidation ponds can achieve between a 0.6–2.0-log removal of Campylobacter spp. (Whiley et al., 2013). Other bacteria can be more susceptible to treatment and secondary wastewater treatment approaches vary considerably. For example, secondary treatment can achieve log removals as high as 3.0-log for enteric bacteria (EPHC NHMRC and NRMMC, 2008) or 7.0-log for other bacteria (Rose et al., 2001). Ozone disinfection (Ct less than 1.0 mg-min/L). U.S. EPA (1991) guidelines for ozonation during drinking water treatment define temperature-dependent virus and Giardia disinfection credits for ozone. Giardia spp. inactivation is greater than 3-log at 1.0 mgmin/L ozone at 20 °C, and virus inactivation is in excess of 4-logs at the same dose. Results from laboratory studies appear to support these guidelines. For example, ozone concentrations of 0.07–0.60 mg-min/L at 5 °C and pH 7 achieved a 4-log reduction of AdV in spiked, demand free water experiments (ThurstonEnriquez et al., 2003). At least 2-log Giardia spp. inactivation was observed at ozone doses of 0.17 mg-min/L at 25 °C and 0.55 mgmin/L at 5 °C in an experiment using clinical specimens of Giardia spp. (Wickramanayake et al., 1985). Exposure to 5 mg-min/L ozone can achieve 1-log reduction of Cryptosporidium spp. (Korich et al., 1990), which indicates Cryptosporidium spp. is more resistant than Giardia spp. to ozone (U.S. EPA, 1991). Higher doses of ozone can achieve log reductions of protozoa as high as 4-log (EPHC NHMRC and NRMMC, 2008). Reviews that do not specify the ozone dose suggest higher log removals are possible for viruses (WHO, 2006; EPHC NHMRC and NRMMC, 2008; NRWI, 2013; Ishida et al., 2008). In the absence of pathogen-specific data, NoV removal by ozone was assumed to be equivalent to that of male-specific coliphage, or 5.4 logs at 0.22 mg-min/L ozone (Tanner et al., 2004). Similarly, log removal data for bacterial species are scarce, yet reviews that do not specify the dose of ozone suggest that log removals between 2–6-log are possible (WHO, 2006; EPHC NHMRC and NRMMC, 2008). Given these uncertainties, and the availability of log removal data for E. coli, Salmonella spp. and Campylobacter spp. reductions are assumed to be equal those reported for E. coli, which are 4-log reductions at CT values less than 1.0 mg-min/L (Sigmon et al., 2015). Biologically active filtration. Biologically active filtration treatment typically includes sand or other filter media interspersed with microbes, activated carbon, or diatoms. Of all treatment types included in this study, the available data to characterize microbial attenuation across biologically active filtration was the scarcest. Thus, conventional filtration data of wastewater was used as a surrogate. Due to their large size, protozoa are more susceptible to filtration than viruses. Dual-media filtration can reduce proto-

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

JID: MRAN

ARTICLE IN PRESS

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

zoa between 1–3-log (WHO, 2006; EPHC NHMRC and NRMMC, 2008). Reported reductions of Giardia spp. via sand filtration are 0.9–3.9-log removal (Taran-Benshoshan et al., 2015; Fu et al., 2010). Reported reductions of Cryptosporidium spp. through sand filtration are approximately 0.9-log reduction (Taran-Benshoshan et al., 2015; Fu et al., 2010). These recently published values are used as maxima for the simulations, along with minimum conservative lower estimate reductions of 0-log, consistent with the data reported by Rose et al. (2004). Adenovirus was reduced 0–0.6-log by full-scale granular media filtration during drinking water treatment (Linden et al., 2012) and enterovirus was reduced by 1-log by sand filtration (Goddard and Butler, 1980). In the absence of pathogen-specific data, removal of NoV was assumed to be similar to enterovirus by sand filtration, with a minimum of 0-log as a conservative lower estimate. Pilot-scale rapid sand contact filtration and biological-chemical contact filtration can achieve 0.5 and 2-log reduction of Salmonella spp., respectively (Koivunen et al., 2001). Campylobacter spp. were assumed to have similar removal. Microfiltration. MF, UF, and RO treatments are broadly defined by their filtration media and pore size (i.e., with MF > UF > RO based on the pore sizes). Reduction across these treatment processes is typically roughly proportional to the size of the pathogen. Bacteria and protozoa are removed more readily than viruses, which are smaller but can adhere to suspended particles. Pathogen-specific data are scarce for MF, UF, and RO, thus general removal for each pathogen type are reported. MF can achieve 3–9-log removal of bacteria and 4–7-log removal of protozoa (Reardon et al., 2005). MF in a membrane bioreactor can achieve 2.4–4.9-log removal of AdV and 1.5–3.3-log removal of NoV (Francy et al., 2012). Reviews that do not specify the virus species suggest that filtration can generally achieve between 0.5–6-log removal of viruses (WHO, 2006; EPHC NHMRC and NRMMC, 2008). Reverse Osmosis. RO achieves up to a 4-log removal of Salmonella spp. bacteria (National Research Council, 2012), which is also used for Campylobacter spp.. RO with pore sizes of 0.0 0 01–0.0 01 μm achieved between 2.7–6.5-log removal for viruses (Smeets et al., 2006; Kruithof et al., 2001; Jacangelo et al., 2005). Protozoa were conservatively assumed to have removals equivalent to those of viruses. Ultrafiltration. UF membranes of achieved removals for 5.6–9log for E. coli, 4.4–6.0-log for Cryptosporidium spp., and 4.7–7.4-log of Giardia spp. (EPHC NHMRC and NRMMC, 2008; Jacangelo et al., 1997; U.S. EPA, 2001; Kachelesky and Masterson, 1995). Salmonella spp. and Campylobacter spp. removal were assumed to be equal to that of E. coli. UF of secondary effluent has been shown to achieve 4.5-log removal for NoV and 4.9 log removal of AdV (Qiu et al., 2015). Ultraviolet disinfection (UVAOP dose at800 mJ/cm2 and UV at 12 mJ/cm2 ). UV disinfection efficacy can vary due to several factors such as temperature, pH, turbidity, UV transmittance, and UV lamp pressure (U.S. EPA, 2012). Some microorganisms can associate or bind to particles that shield them from UV light. After UV disinfection, once microorganisms are exposed to visible light, their cells can begin to repair DNA damaged by the UV light in a process known as photo reactivation (U.S. EPA, 2012). UVAOP disinfection is used as an AWT process for potable reuse. In these advanced treatment applications, the most common purpose of this unit process is the destruction of disinfection byproducts such as 1,4-dioxane and N-nitrosodimethylamine (NDMA). For this purpose, UV doses in the range of approximately 800 mJ/cm2 are common. At 800 mJ/cm2 , greater than 6-log of inactivation of each of the reference pathogens can be achieved (Gerba et al., 2002; U.S. EPA, 2006b). For disinfection of wastewater via UV light, pathogen inactivation is less than for UVAOP because significantly lower doses of

UV light are applied. The U.S. EPA reviewed research studies to determine Cryptosporidium spp., Giardia spp., and AdV inactivation due to UV treatment (U.S. EPA, 2003). That review found that at 12 mJ/cm2 , Cryptosporidium spp. was reduced from 2.0–3.5-log, Giardia spp. removal ranges were between 2.0–3.5-log, and AdV levels were reduced by 0–0.5-log. Reviews that do not specify the virus species or UV dose suggest that UV disinfection can achieve between 1–3-log removal of viruses (WHO 2006, EPHC NHMRC and NRMMC, 2008). Data on NoV reduction across UV treatment in the range of 12 mJ/cm2 are sparse. Incremental mean reduction due to UV treatment is reported to be on the order of 0.8-log (Pouillot et al., 2015; Lee and Shin, 2011). This estimate is similar to other enteric viruses reductions at UV doses in the range of 12 mJ/cm2 , which are in the range of approximately 0.5–1.5-log (U.S. EPA, 2003). Based on these data, NoV reductions were estimated to be 0.5–1.5-log. UV disinfection at less than 10 mJ/cm2 can achieve removals of 4.0-log for E. coli (U.S. EPA, 2003). In the absence of pathogen-specific data, Campylobacter spp. or Salmonella spp. removals are assumed to equal that of E. coli. Conventional drinking water treatment. Full-scale conventional drinking water treatment can achieve between 1.5–4.0-log removal of Cryptosporidium spp. and 0.3–4.0-log reduction of Giardia spp. (Linden et al., 2012; Betancourt and Rose, 2004). The lower bound for Giardia spp. likely differs from Cryptosporidium spp. due to differing test conditions between the cited studies, although actual reductions for these two reference pathogens are likely similar (U.S. EPA, 1991). Slow sand filtration conventional drinking water treatment can reduce E. coli, Campylobacter, and MS2bacteriophages by 2–3, 3–4, and 1.5–2 logs, respectively (Hijnen et al., 2005; Albinana-Gimenez et al., 2006). In the absence of pathogen-specific data, AdV and NoV removals are assumed equal that of MS2-bacteriophage, and Salmonella spp. removal to equal that of E. coli. These reported values are used to characterize pathogen reduction across TT4. Disinfection with free chlorine (Ct less than 12 mg-min/L). Bacteria and viruses are more sensitive to free chlorine disinfection than protozoa. US EPA guidelines suggest a 4-log reduction of viruses can be achieved at free chlorine doses as low as 2 mg-min/L, although more recent data suggest that there are some conditions (i.e., pH > 8) under which some viruses require higher CTs to achieve a 4-log reduction (Keegan et al., 2012). Giardia spp. are much more resistant, with a maximum of 0.5-log reduction at 12 mg-min/L (U.S. EPA, 1991). Cryptosporidium spp. is highly resistant to free chlorine disinfection, with doses estimated as high of 80 0–90 0 mg-min/L chlorine providing only 1-log reduction (Nasser, 2016; Hirata and Hashimoto, 1998). No reduction data were available for Cryptosporidium spp. at or below a dose of 12 mg-min/L; therefore, it was assumed that chlorine at a dose of 12 mg-min/L does not reduce Cryptosporidium spp. Chlorine doses of 0.35 mg-min/L can achieve between 4–5-log reduction of AdV (Baxter et al., 2007). Specific data were not available to determine the efficacy of free chlorine disinfection relative to NoV. Reviews that do not specify the virus species or free chlorine CT suggest that free chlorine can achieve between 1–3-log inactivation of viruses and 2–6-log inactivation of bacteria (WHO, 2006; EPHC NHMRC and NRMMC, 2008). NoV reductions were assumed to be similar to other viruses, with reductions in the range of 1-4-log. In our analyses, we assume that all removal and/or inactivation in the ESB was solely due to the Cl disinfection. Literature review summary. A summary of the parameter values selected based on the results of the literature review is provided in Table 2. Parameters were generally characterized with uniform distributions, with lower and upper bounds corresponding to the minimum and maximum values reported in the literature (Soller et al., 2010b,a; Soller and Eisenberg, 2008; Eisenberg et al., 2004; Soller et al., 2006; Soller et al., 2014). In some cases, a single

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

5

ARTICLE IN PRESS

JID: MRAN 6

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

Table 2 Summary of reference pathogen raw wastewater influent densities and reductions across wastewater, drinking water, and DPR unit treatment processes pathogen densities in raw wastewater and log10 reductions across unit treatment processes. Adenovirus

Campylobacter

Cryptosporidium

Giardia

Norovirus

Salmonella

Min

Max

Min

Max

Min

Max

Min

Max

Min

Min

Max

Raw wastewatera Conventional secondary wastewater treatment

56 0.9

6.9E+03 3.2

900 0.6

4.0E + 04 2.0

0.3 0.7

5.0E + 04 1.5

3.2 0.5

1.0E + 04 3.3

3.76 0.8

3 1.3

1.1E + 03 1.7

Ozonation Biologically active filtration Microfiltration Reverse osmosis Ultrafiltration Ultraviolet disinfection with advanced oxidation (800 mJ/cm2) Ultraviolet disinfection with peroxide (12 mJ/cm2) Conventional drinking water treatment Disinfection with free chlorine

4.0 0 2.4 2.7 4.9 6.0 0.0 1.5 4.0

a b

4.0 0.5 3.0 4.0 5.6 6.0 4.0 3.0 4.0

0.6 4.9 6.5

0.5 2.0 5.0

1.0 0 4.0 2.7 4.4 6.0 2.0 1.4 0.0

2 9.0 9.0

4.0

3.0 0 4.0 2.7 4.7 6.0 2.0 0.3 0.0

0.85 7.0 6.5 6.0 3.5 4.0

Max b

5.4 0 1.5 2.7 4.5 6.0 0.5 1.5 1.0

3.9 7.0 6.5 7.4 3.5 4.0 0.5

0.93 3.7 1 3.3 6.5

1.5 2.0 4.0

b

4.0 0.5 3.0 4.0 5.6 6.0 4.0 2.0 4.0

2 9.0 9.0

3.0

Adenovirus IU/L, Campylobacter MPN/L, Cryptosporidium oocysts/L, Giardia cysts/L, Norovirus log 10 copies/L, Salmonella CFU/L. Values shown for raw wastewater are mean and standard deviation of normal distribution in log10 copies. Annual risk estimate Benchmark Risk

Benchmark Risk 10-4

10-4 Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

Daily Risk of Infection

10-6 10-7 10-8

10-6

Annual risk estimate

10-9 10-10 10-11

10-14

10-7 10-8 10-9 10-10 10-11 10-12

10-12 10-13

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-5

Daily Risk of Infection

10-5

10-13

a

10-15 0.001 0.01

10-14

b

10-15 0.1

1

10

30

50

70

90

99

99.9

Percent of Values Less than Corresponding Value

0.001 0.01

0.1

1

10

30

50

70

90

99

99.9

Percent of Values Less than Corresponding Value

Fig. 2. Daily pathogen risks for TT1 (WWTP – MF – RO – UV – ESBCl), (a) Simulation using UVAOP dose of 800 mJ/cm2 ; (b) Simulation using UV dose of 12 mJ/cm2 .

point estimate value was used where appropriate data were not available to specify a uniform distribution. 3.2. Simulation results Fig. 2 presents the results for one simulation for TT1 (estimated distributions of cumulative annual risks of infection based on 10 0 0 simulations for each of the TT configurations are summarized in Fig. 6). Fig. 2 shows the estimated daily risks of infection for each of the reference pathogens, the estimated cumulative daily risk from all reference pathogens, and the estimated annual risk of infection (2.5 × 10−9 , red line, Fig. 2a). In this simulation, the annual risk estimate is driven by the highest individual daily risks estimates for NoV. Together, the seven highest daily risk estimates account for approximately 80% of the annual risk estimate. When the UV unit treatment process in this TT is operated at the lower dose (12 mJ/cm2 ), consistent with traditional wastewater disinfection rather than for destruction of disinfection byproducts, the estimated pathogen risks are significantly higher (p = < 0.001) (Fig. 2b). In this simulation, the estimated annual risk of infection is 5.0 × 10−4 (red line, Fig. 2b), which is above the benchmark risk level. The annual risk estimate is driven by the five highest daily risks estimates for NoV, which together account for approximately 80% of the annual risk. In the TT2 base scenario simulation, the annual risk (1.0 × 10−11 ) is dominated by one NoV and five Cryptosporidium spp. daily risk values, together accounting for approximately 90% of the total

annual risk estimate (Fig. 3a). The estimated annual risk of infection associated with the lower UV dose simulation is 1.8 × 10−7 (Fig. 3b), with the annual risk estimate driven by the two highest daily risks estimates for NoV, three highest daily risks estimates for AdV, and the single highest daily risk estimate for Cryptosporidium spp. Together, these six daily risk estimates account for approximately 70% of the annual risk. The estimated annual risk of infection associated with the TT3 base simulation is 1.2 × 10−8 (Fig. 4a). The annual risk estimate is dominated by the Cryptosporidium spp. daily risk estimates, with the twenty days with the highest Cryptosporidium spp. daily risks accounting for approximately 70% of the estimated annual risk estimate. The estimated annual risk of infection associated with the lower UV dose simulation is 3.9 × 10−5 (Fig. 4b). The annual risk estimate is driven by the Cryptosporidium spp. risks estimates, which accounts for the vast majority of the annual risk. The estimated daily risks of infection for each of the reference pathogens for the TT4 base simulation (TT3 followed by a conventional DWTP) along with the estimated cumulative daily risk from all reference pathogens are presented (Fig. 5). The estimated annual risk of infection is 8.7 × 10−11 . Similar to the results presented for TT3, the annual risk is dominated by the Cryptosporidium spp. daily risk estimates. The estimated annual risk of infection associated with the lower UV dose simulation is 2.5 × 10−7 (Fig. 5b). These results illustrate that circulation of the DPR water into a conventional DWTP, prior to consumer exposure, provides risk reduction levels of approximately 2-log for the scenarios evaluated.

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

ARTICLE IN PRESS

JID: MRAN

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

7

Benchmark risk

Benchmark risk

10-4

10-4

10-5

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-7 10-8

10-5

10-9 10-10 Annual risk estimate

10-11 10-12

10-7 10-8

Annual risk estimate

10-9 10-10 10-11 10-12

10-13 10-14

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-6

Daily Risk of Infection

Daily Risk of Infection

10-6

10-13

a

10-14

10-15 0.001 0.01

0.1

1

10

30

50

70

90

99

b

10-15 0.001 0.01

99.9

Percent of Values Less than Corresponding Value

0.1

1

10

30

50

70

90

99

99.9

Percent of Values Less than Corresponding Value

Fig. 3. Daily pathogen risks for TT 2 (WWTP – O3 – BAF – MF – RO - UV), Simulation using UVAOP dose of 800 mJ/cm2 ; (b) Simulation using UV dose of 12 mJ/cm2 .

Benchmark risk

10-5

Daily Risk of Infection

10-6 10-7 10-8

Benchmark risk 10-4

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-6 Annual risk estimate

10-9 10-10 10-11 10-12 10-13 10-14

10-7 10-8 10-9 10-10 10-11 10-12 10-13

a

10-15 0.001 0.01

Annual risk estimate

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-5

Daily Risk of Infection

10-4

10-14

0.1

1

10

30

50

70

90

99

99.9

b

10-15 0.001 0.01

Percent of Values Less than Corresponding Value

0.1

1

10

30

50

70

90

99

99.9

Percent of Values Less than Corresponding Value

Fig. 4. Daily pathogen risks for TT 3 (WWTP – O3 – BAF – UF – UV – ESBCl), (a) Simulation using UVAOP dose of 800 mJ/cm2 ; (b) Simulation using UV dose of 12 mJ/cm2 .

Benchmark risk

10-4

Benchmark risk

10-4

10-5

10-7 10-8 10-9

Annual risk estimate 10-10 10-11 10-12 10-13 10-14

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-6

Daily Risk of Infection

Daily Risk of Infection

10-5

Norovirus Adenovirus Cryptosporidium Giardia Campylobacter Salmonella Combined Daily Risk

10-6

10-7 10-8

Annual risk estimate

10-9 10-10 10-11 10-12 10-13

a

10-14

10-15 0.001 0.01

0.1

1

10

30

50

70

90

Percent of Values Less than Corresponding Value

99

99.9

b

10-15 0.001 0.01

0.1

1

10

30

50

70

90

99

99.9

Percent of Values Less than Corresponding Value

Fig. 5. Daily pathogen risks for TT4 (WWTP – O3 – BAF – UF – UV – ESBCl – DWTP), (a) Simulation using UVAOP dose of 800 mJ/cm2 ; (b) Simulation using UV dose of 12 mJ/cm2 .

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

JID: MRAN 8

ARTICLE IN PRESS

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

Fig. 6. Cumulative annual pathogen risks of infection for DPR TTs.

The estimated distributions of cumulative annual risks of infection (based on 10 0 0 simulations) for each of the TT configurations are summarized in Fig. 6. Overall, the results illustrate that under normal operating conditions, the TTs can be configured to be extremely effective in removing reference pathogens to levels that present low public health risk. The results also reveal that when the UV unit treatment process is operated at the higher UVAOP dose, the associated relative pathogen risk is significantly less than the risk associated with the lower UV dose (p < 0.001, comparison of TT1a to TT1b). Finally, a statistical comparison of TT3 with TT4 reveals that circulating DPR product water through a DWTP provides incremental pathogen risk reductions of approximately two logs for the TT evaluated (p < 0.001).

3.3. Sensitivity analysis results The sensitivity analysis evaluated the relative effect of using alternative dose-response relationships (fractional-Poisson) for NoV and Cryptosporidium spp. on the risk estimates for TT1 (Messner and Berger, 2016; Messner et al., 2014). Because aggregation of NoV particles is a significant source of uncertainty in the doseresponse relationship (Teunis et al., 2008), an aggregation size of 1106 viral particles was assumed in the sensitivity analysis, as compared to an aggregation size of 1 (i.e., disaggregated) in the base simulations NoV (Teunis et al., 2008). Fig. 7 presents the estimated daily risks of infection for NoV and Cryptosporidium spp. for TT1a. The results indicate that the estimated daily risks of infection differ significantly for both NoV and Cryptosporidium spp. when then alternative dose-response relationships are employed (Cryptosporidium spp., p < 0.001; NoV, p < 0.001). However, the use of the alternative Cryptosporidium spp. dose-response relationship significantly elevates the daily risk

estimates, while the alternative NoV dose-response relationship significantly lowers the daily risk estimates. The effect of the alternative dose-response models on the cumulative annual risks of infection for TT1a and TT1b was also explored (Fig. 8). The overall annual risks are reduced by approximately 1.5 log10 and 2.5 log10 units when the Fractional Poisson dose-response relationship is used for only NoV for TT1a and TT1b, respectively. Overall annual risks are relatively unchanged when the Fractional Poisson dose-response relationship is used for only Cryptosporidium spp. in both TT1a and TT1b. When the Fractional Poisson model is used for both NoV and Cryptosporidium spp., overall annual risks are reduced (compare TT1a to Both FP, Fig. 8). This sensitivity analysis indicates that use of the alternative dose-response models influences the predicted overall annual risk estimates and illustrates the overall impact that uncertainty in the dose-response relationships can have on the simulation results.

4. Discussion Recycled water is an increasingly important and valued resource worldwide and potable reuse is an important component of those plans. DPR is now being considered as a viable option in many communities. Communities may lack an alternative water supply or suitable hydrology for IPR groundwater recharge, DPR may be less costly than the use of tertiary treated water for irrigation once the capitol costs of distribution are considered, and/or DPR may require less energy than other water supply options (Tchoganoglous et al., 2011). DPR projects can take two distinct forms – those in which the product water is introduced directly into a potable water supply distribution system, and those in which the product water is introduced into the raw water supply immediately upstream of a conventional DWTP. The relative health

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

ARTICLE IN PRESS

JID: MRAN

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

9

Benchmark Risk

10-4 10-5

Daily Risk of Infection

10-6

NoV (Hypergeometric DR) NoV (Fractional Poisson DR) Cryptosporidium (Exponential DR) Cryptosporidium (Fractional Poisson DR)

10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 0.001 0.01

0.1

1

10

30

50

70

90

99

99.9

Percent of Values Less than Corresponding Value Fig. 7. TT1a (WWTP – MF – RO – UVAOP – ESBCl) Daily Norovirus and Cryptosporidium spp. Risks for alternative Dose-Response (DR) Relationships.

Fig. 8. Cumulative annual risks of infection based on alternative dose-response relationships for Norovirus and Cryptosporidium spp., TT1a and TT1b.

protection afforded by both types of DPR strategies was evaluated in this study. There are several notable findings from this work and some uncertainties that should be considered in context. First, our results illustrate statistically significant predicted human health-based advantages for DPR projects in which AWTF product water is introduced into the raw water supply immediately upstream of a conventional drinking water treatment facility compared to those in which AWTF product water is introduced directly into a potable water supply distribution system (compare TT3 to TT4, Fig. 6). The TT evaluated here demonstrated ∼2 log reduction in risk. This incremental benefit of the conventional drinking water treatment facility could vary across treatment systems because pathogen reductions (Table 2) can vary for the pathogens of primary concern, which can change from TT to TT. Additional benefits to this

approach (inclusion of conventional drinking water treatment) may include public perception benefits, source water diversification benefits, and a time dimension that would be unavailable (or much more limited) in a system in which AWTF product water is introduced directly into the drinking water distribution system. Second, annual risk estimates for any particular TT are driven by the highest daily risks for any of the individual reference pathogens. This finding indicates that even a single day can drive annual risks and highlights the need for robust and reliable online monitoring of unit treatment processes within DPR facilities. Higher relative risks on any particular day could stem from high concentrations in the raw wastewater, treatment efficacy at the lower end of the expected range, or due to an excursion from design operations. In this context, this study highlights the potential public health importance of even short term excursions

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

JID: MRAN 10

ARTICLE IN PRESS

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

from design operations and the potential public health implications of high pathogen loading that can occur, particularly during outbreak conditions (e.g., when sewage loading from an upstream community may be highest for one or more pathogens). Moreover, these particular results, based on a probabilistic approach raise questions about the interpretation of the more simplistic “log removal credits” that States often use to determine the adequacy of proposed TTs. To effectively understand the likelihood of achieving the benchmark level of public health protection, a robust and rigorous method for TT evaluation within a public health context is needed. The methodology employed here offers a transparent approach that can be used for this purpose. Third, this analysis indicates that NoV is an important reference pathogen for several of the TTs evaluated, and thus, should be considered carefully. There has been resistance among recycled water professionals to specifically consider NoV risks related to the use of recycled water. Part of the resistance has been due to the fact that NoV cannot be readily cultured and is thus monitored by molecular methods (i.e. qPCR) that identify viral RNA, rather than infectious viral particles. Herein our sensitivity analyses capture a representative range of NoV infectivity reported in the literature (Van Abel et al., 2016). While the differences between alternative and commonly used dose-response models may be statistically significant, they are not enough to change the relative risk comparisons and overall conclusions of this work. They are sufficient, however, to raise questions about which pathogens drive the estimated risks for specific treatment train configurations. Additionally, the literature indicates that NoV has been reported in densities as high as 109 genome copies/L in raw wastewater (da Silva et al., 2007). If NoV were to be present in source water used for DPR at densities of this magnitude, the previously suggested 12-log of viral reduction may not adequately ensure the benchmark level of public health protection (data not shown). Therefore, understanding NoV presence in raw wastewater, NoV attenuation across individual unit treatment processes, the relationship between viral RNA and infectious viral particles, and NoV infectivity after DPR treatment should be areas of active research and considered in DPR project approval. Fourth, our results illustrate that proposed DPR project designs need to carefully consider reduction of both Cryptosporidium spp. and human enteric viruses, such as NoV. For example, DPR TTs that employ non-reverse osmosis processes for advanced treatment are preferred by some inland communities due to the brine disposal issues associated with RO treatment. However, this work indicates that treatment trains lacking RO should be configured upstream of a conventional drinking water treatment facility and/or with a UVAOP unit process operating at high UV doses (800 mJ/cm2 ) to reliably achieve the benchmark level of infection due to risks from Cryptosporidium spp. On the other hand, an AWTF treatment system comprised of MF-RO-UV-ESBCl may require the use of high UV doses to reliably achieve the benchmark level of infection due to risks from NoV. Alternatively, configuring an MF-RO-(low dose) UV-ESBCl DPR facility upstream of a conventional drinking water treatment facility may also provide confidence that the desired level of protection is achieved, although that configuration was not specifically evaluated here. Finally, our results indicate that expected performance of individual unit treatment processes within an integrated TT can be critical for consistently achieving the benchmark level of public health protection. This observation reinforces the importance of robust and reliable monitoring systems for critical unit treatment processes, and highlights the need to understand the potential vulnerability of integrated systems with respect to unit treatment process failures or upsets. A series of assumptions were made regarding the efficacy of unit treatment processes that are “dose dependent”, such as UV light treatment, chlorination, and ozonation. Selected doses rep-

resent common usage. As illustrated above, changing the selected dose values can have a profound influence on the associated level of public health protection. This point was highlighted by simulating the UV unit treatment process under two very different, yet feasible operational conditions. Similar effects could also be seen by varying the type or dose of chlorine and the dose of ozone. It was also assumed that pathogen reduction across each of the unit treatment processes could reasonably be estimated based on the minimum and maximum vales reported in the literature. Targeted future work may refine these estimates, potentially influencing the interpretation of the results. For example, because the upper tails of the statistical distributions influence the annual risks, updated pathogen-specific log removal data or statistical distributions may be useful refinements to (or confirmation of) this work. The most commonly employed peer reviewed dose-response relationships were used for the reference pathogens. However there is uncertainty embedded in each of those relationships, particularly as the probability of adverse health effects from a relatively homogenous feeding trial subpopulation is extrapolated to the widely heterogeneous general population. Nevertheless, when the results of this work are viewed within the context of a relative health assessment rather than an estimate of absolute risk, the uncertainty surrounding these issues is clearly overshadowed by importance of the overall findings of this work.

5. Conclusions This study leveraged readily available peer-reviewed data and extended a previously published statistical approach to estimate daily and annual infection risks associated with the consumption of product water from a series of DPR treatment trains. Overall, the methodology resulted in several important findings for DPR implementation, is adaptable to other DPR treatment trains, and could be iteratively refined as additional data become available. This work should be useful for federal and state regulators considering DPR as source water, state and local decision makers as they consider whether to permit a particular DPR project, risk managers identifying the impact of treatment failures, and design engineers as they consider which unit treatment processes should be employed for particular projects.

Acknowledgments The research described in this article was funded by the U.S. EPA Office of Water, Office of Science and Technology under contract # EP-C-11-005 to ICF International, LLC. This work has been subject to formal Agency review but does not necessarily reflect the views of the Agency, and no official endorsement should be inferred. The authors gratefully acknowledge the valuable contributions of Michael Messner, Margaret Nellor, Stig Regli, and Jamie Strong for their critical review of the manuscript and insightful comments. References Albinana-Gimenez, N., Clemente-Casares, P., Bofill-Mas, S., Hundesa, A., Ribas, F., Girones, R., 2006. Distribution of human polyoma-viruses, adenoviruses, and hepatitis E virus in the environment and in a drinking-water treatment plant. Environ. Sci. Technol. 40 (23), 7416–7422. Baxter, C.S., Hofmann, R., Templeton, M.R., Brown, M., Andrews, R.C., 2007. Inactivation of adenovirus types 2, 5, and 41 in drinking water by UV light, free chlorine, and monochloramine. J. Environ. Eng. 133 (1), 95–103. Betancourt, W.Q. and Rose, J.B. (2004) Drinking water treatment processes for removal of Cryptosporidium and Giardia veterinary parasitology 126(1-2), 219. CDPH, 2011. Groundwater Replenishment Using Recycled Water. Department of Public Health, California http://www.cdph.ca.gov/certlic/drinkingwater/Documents/ Recharge/DraftRechargeReg- 2011- 11- 21.pdf.

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

JID: MRAN

ARTICLE IN PRESS J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

Cheng, H.-W.A., Lucy, F.E., Graczyk, T.K., Broaders, M.A., Tamang, L., Connolly, M., 2009. Fate of Cryptosporidium parvum and Cryptosporidium hominis oocysts and Giardia duodenalis cysts during secondary wastewater treatments. Parasitol. Res. 105, 689–696. Crabtree, K.D., Gerba, C.P., Rose, J.B., Haas, C.N., 1997. Waterborne adenovirus: A risk assessment. Water Sci. Technol. 35 (11-12), 1–6. Crockett, C.S., 2007. The role of wastewater treatment in protecting water supplies against emerging pathogens. Water Environ. Res. 79 (3), 221–232. da Silva, A.K., Le Saux, J.C., Parnaudeau, S., Pommepuy, M., Elimelech, M., Le Guyader, F.S., 2007. Evaluation of removal of noroviruses during wastewater treatment, using real-time reverse transcription-PCR: different behaviors of genogroups I and II. Appl. Environ. Microbiol. 73 (24), 7891–7897. Dahl, R., 2014. Advanced thinking: potable reuse strategies gain traction. Environ. Health Perspect. 122 (12), A332–A335. Eftim, S., Hong, T., Ichida, A., Warren, I., Soller, J., Nappier, S.P., 2016. Systematic literature reviews and development of distribution curves for viral densities in raw wastewater. Univeristy of North Carolina. Eisenberg, J.N., Brookhart, M.A., Rice, G., Brown, M., Colford Jr., J.M., 2002. Disease transmission models for public health decision making: analysis of epidemic and endemic conditions caused by waterborne pathogens. Environ. Health Perspect. 110 (8), 783. Eisenberg, J.N.S., Soller, J.A., Scott, J., Eisenberg, D.M., Colford, J.M., 2004. A dynamic model to assess microbial health risks associated with beneficial uses of biosolids. Risk Anal. 24 (1), 221–236. Eisenberg, J.N.S., Moore, K., Colford, J., Soller, J.A., Eisenberg, D., 2006. Application of a dynamic model to assess microbial health risks associated with beneficial uses of biosolids. Water Environment Research Foundation Report 98-REM-1A. IWA Publishing. EPHC NHMRC, NRMMC, 2008. Australian Guidelines for Water Recycling Augmentation of Drinking Water Supplies, Environment Protection and Heritage Council, National Health and Medical Research Council, Natural Resource Management Ministerial Council. Canberra, Australia. Fishman, G.S., 1995. Monte Carlo: Concepts, Algorithms, and Applications. Springer, New York, NY, USA ISBN 0-387-94527-X. Francy, D.S., Stelzer, E.A., Bushon, R.N., Brady, A.M., Williston, A.G., Riddell, K.R., Borchardt, M.A., Spencer, S.K., Gellner, T.M., 2012. Comparative effectiveness of membrane bioreactors, conventional secondary treatment, and chlorine and UV disinfection to remove microorganisms from municipal wastewaters. Water Res. 46 (13), 4164–4178. Fu, C.Y., Xie, X., Huang, J.J., Zhang, T., Wu, Q.Y., Chen, J.N., Hu, H.Y., 2010. Monitoring and evaluation of removal of pathogens at municipal wastewater treatment plants. Water Sci. Technol. 61 (6), 1589–1599. Gerba, C.P., Gramos, D.M., Nwachuku, N., 2002. Comparative inactivation of enteroviruses and adenovirus 2 by UV light. Appl. Environ. Microbiol. 68 (10), 5167–5169. Goddard, M., Butler, M., 1980. Viruses and Wastewater Treatment: Proceedings of the International Symposium on Viruses and Wastewater Treatment, Held at the University of Surrey, Guildford. Pergamon Press, Oxford, pp. 15–17 September 1980. Haas, C.N., Trussell, R.R., 1998. Frameworks for assesing reliability of multiple, independant barriers in potable water reuse. Water Sci. Technol. 38 (6), 1–8. Haas, C.N., Rose, J.B., Gerba, C.P., 1999. Quantitative microbial risk assessment. J.W. Wiley, Inc. 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. Hewitt, J., Leonard, M., Greening, G.E., Lewis, G.D., 2011. Influence of wastewater treatment process and the population size on human virus profiles in wastewater. Water Res. 45 (18), 6267–6276. Hijnen, W.A., Brouwer-Hanzens, A.J., Charles, K.J., Medema, G.J., 2005. Transport of MS2 phage, Escherichia coli, Clostridium perfringens, Cryptosporidium parvum, and Giardia intestinalis in a gravel and a sandy soil. Environ. Sci. Technol. 39 (20), 7860. Hirata, T., Hashimoto, A., 1998. Experimental assessment of the efficacy of microfiltration and ultrafiltration for Cryptosporidium removal. Water Sci. Technol. 38 (12), 103–107. Hurst, C.J., McClellan, K.A., Benton, W.H., 1988. Comparison of cytopathogenicity, immunofluorescence and in situ DNA hybridization as methods for the detection of adenoviruses. Water Res. 22 (12), 1547–1552. Ishida, C., Salveson, A., Robinson, K., Bowman, R., Snyder, S., 2008. Ozone disinfection with the HiPOX reactor: streamlining an old technology for wastewater reuse. Water Sci. Technol. 58 (9), 1765–1773. Jacangelo, J., Adham, S., Laine, J.M., 1997. Membrane Filtration for Microbial Removal. AWWA Research Foundation, Denver, CO. Jacangelo, J., Madec, A., Schwab, K.J., Huffman, D.E., Mysore, C.S., 2005. Advances in the use of low-pressure, hollow fiber membranes for the disinfection of water. Water Sci. Technol 5, 109–115. Kachelesky, L.A., Masterson, T., 1995. Membrane Filtration of Sewage Treatment Plant Effluent. American Water Works Association, Reno, NV. Keegan, A., Wati, S., Robinson, B., 2012. Chlor(am)ine disinfection of human pathogenic viruses in recycled waters. Australian Water Quality Centre SWF62M-2114.

[m5G;September 8, 2016;21:39] 11

Koivunen, J., Lanki, E., Rajala, R.L., Siitonen, A., Heinonen-Tanski, H., 2001. Determination of Salmonellae from municipal wastewaters. Water Sci. Technol. 43 (12), 221–224. Korich, D.G., Mead, J.R., Madore, M.S., Sinclair, N.A., Sterling, C.R., 1990. Effects of ozone, chlorine dioxide, chlorine, and monochloramine on Cryptosporidium parvum oocyst viability. Appl. Environ. Microbiol. 56 (5), 1423–1428. Kruithof, J.C., Kamp, P.C., Folmer, H.C., Nederlof, M.M., van Hoof, S.C.J.M., 2001. Development of a membrane integrity monitoring strategy for the UF/RO Heemskerk drinking water treatment plant. Water Sci. Technol. 1 (5-6), 261–271. Lee, J., Shin, G., 2011. Inactivation of human adenovirus by sequential disinfection with an alternative UV technology and free chlorine. J. Water Health 9 (1), 53. Lemarchand, K., Lebaron, P., 2003. Occurrence of Salmonella spp. and Cryptosporidium spp. in a French coastal watershed: relationship with fecal indicators. FEMS Microbiol. Lett. 218 (1), 203–209. Linden, K., Salveson, A.T., Thurston, J.A., 2012. Study of Innovative Treatments for Reclaimed Water. WateReuse Research Foundation, Washington, D.C.. Macler, B.A., Regli, S., 1993. Use of microbial risk assessment in setting United States drinking-water standards. Int. J. Food Microbiol. 18 (4), 245–256. Mann, H.G., Whitney, D.R., 1947. On a test of whether one of two random variables is stochastically larger than the other. Ann. Math. Statist. 18 (1), 50–60. Mead, P.S., Slutsker, L., Dietz, V., McCaig, L.F., Bresee, J.S., Shapiro, C., Griffin, P.M., Tauxe, R.V., 1999. Food related illness and death in the United States. Emerging Infect. Dis. 5 (5), 607–625. Medema, G.J., Teunis, P.F., Havelaar, A.H., Haas, C.N., 1996. Assessment of the dose-response relationship of Campylobacter jejuni. Int. J. Food Microbiol. 30 (1-2), 101–111. Messner, M.J., Berger, P., Nappier, S.P., 2014. Fractional Poisson–a simple dose-response model for human norovirus. Risk Anal. 34 (10), 1820–1829. Messner, M.J., Berger, P., 2016. Cryptosporidium Infection Risk: Results of New DoseResponse Modeling. Risk Anal. doi:10.1111/risa.12541. Nasser, A.M., 2016. Removal of Cryptosporidium by wastewater treatment processes: a review. J. Water Health 14 (1), 1–13. National Research Council, 2012. Water Reuse: Expanding the Nation’s Water Supply Through Reuse of Municipal Wastewater. National Academy Press, Washington, D.C.. NWRI, 2013. Examining the Criteria for Direct Potable Reuse. Recommendations of an NWRI Independent Advisory Panel, Porject 11-02. National Water Research Institute, Fountain Valley, CA. NWRI, 2015. Framework for Direct Potable Reuse. National Water Research Institute, Fountain Valley, CA. Olivieri, A.W., Eisenberg, D.M., Soller, J., Eisenberg, J.N.E., Cooper, R.C., Tchobanoglous, G., Trussell, R.R., Gagliardo, P., 1999. Estimation of pathogen removal in an advanced water treatment facility using Monte Carlo simulation. Water Sci. Technol. 40 (4-5), 223–233. Pouillot, R., Van Doren, J.M., Woods, J., Plante, D., Smith, M., Goblick, G., Roberts, C., Locas, A., Hajen, W., Stobo, J., White, J., Holtzman, J., Buenaventura, E., Burkhardt, W., 3rd Catford, A., Edwards, R., DePaola, A., Calci, K.R., 2015. Meta– analysis of the reduction of norovirus and male-specific coliphage concentrations in wastewater treatment plants. Appl. Environ. Microbiol. 81 (14), 4669–4681. Qiu, Y., Lee, B.E., Neumann, N., Ashbolt, N., Craik, S., Maal-Bared, R., Pang, X.L., 2015. Assessment of human virus removal during municipal wastewater treatment in Edmonton, Canada. J. Appl. Microbiol. 119 (6), 1729–1739. Reardon, R., DiGiano, F., Aitken, M., Paranjape, S., Kim, J.H., Chang, S.-Y., 2005. Membrane Treatment of Secondary Effluents for Subsequent Use. Water Environment Research Foundation/International Water Association. Alexandria, VA. Regli, S., Rose, J.B., Haas, C.N., Gerba, C.P., 1991. Modeling the risk from Giardia and viruses in drinking-water. J. Am. Water Works Assoc. 83 (11), 76–84. Richmond, G., 2016. Global science engagement. Science 351 (6272), 427. Rose, J.B., Huffman, D.E., Riley, K., Farrah, S.R., KLukasik, J.O., Hamann, C.L., 2001. Reduction of enteric microorganisms at the upper Occoquan Sewage Authority water reclamation plant. Water Environ. Res. 73 (6), 711–720. Rose, J.B., Nowlin, H., Farrah, S.R., Harwood, V., Levine, A., Lukasik, J., Menendez, P., Scott, T.M., 2004. Reduction of pathogens, indicator bacteria, and alternative indicators by wastewater treatment and reclamation processes. Water Environment Research Foundation Report 00-PUM-2T. IWA Publishing. Roser, D., Ashbolt, N., 2007. Source Water Quality Assessment and the Management of Pathogens in Surface Catchments and Aquifers. Research Report 29, Cooperative Research Centre for Water Quality and Treatment. Australia. Scallan, E., Hoekstra, R.M., Angulo, F.J., Tauxe, R.V., Widdowson, M.A., Roy, S.L., Jones, J.L., Griffin, P.M., 2011. Foodborne illness acquired in the United States— major pathogens. Emerging Infect. Dis. 17 (1), 7–15. Schoen, M.E., Soller, J.A., Ashbolt, N.J., 2011. Evaluating the importance of faecal sources in human-impacted waters. Water Res. 45 (8), 2670–2680. Sigmon, C., Shin, G.-A., Mieog, J., Linden, K.G., 2015. Establishing surrogate–virus relationships for ozone disinfection of wastewater. Environ. Eng. Sci. 32 (6), 451–460. Sinclair, M., O’Toole, J., Gibney, K., Leder, K., 2015. Evolution of regulatory targets for drinking water quality. J. Water Health 13 (2), 413–426. Smeets, P., Rietveld, L., Hijnen, W., Medema, G., Stenstrom, T.A., 2006. Efficacy of water treatment processes. Microrisk: a scientific basis for managing drinking water safety from source to tap.

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003

JID: MRAN 12

ARTICLE IN PRESS

[m5G;September 8, 2016;21:39]

J.A. Soller et al. / Microbial Risk Analysis 000 (2016) 1–12

Soller, J.A., Eisenberg, D., Olivieri, A. and Galitsky, C. (20 0 0) Groundwater replenishment system water quality evaluation. Prepared for EOA, Inc., Orange County Water District and Orange County Sanitation District. Soller, J.A., Olivieri, A., Crook, J., Parkin, R., Spear, R., Tchobanoglous, G., Eisenberg, J.N.S., 2003. Risk-based approach to evaluate the public health benefit of additional wastewater treatment. Environ. Sci. Technol. 37 (9), 1882–1891. Soller, J.A., Eisenberg, J.N.S., DeGeorge, J.F., Cooper, R.C., Tchobanoglous, G., Olivieri, A.W., 2006. A public health evaluation of recreational water impairment. J. Water Health 4 (1), 1–19. Soller, J.A., Eisenberg, J.N.S., 2008. An evaluation of parsimony for microbial risk assessment models. Environmetrics 19 (1), 61–78. Soller, J.A., Bartrand, T., Ashbolt, N.J., Ravenscroft, J., Wade, T.J., 2010a. Estimating the primary etiologic agents in recreational freshwaters impacted by human sources of faecal contamination. Water Res. 44 (16), 4736–4747. Soller, J.A., Schoen, M.E., Bartrand, T., Ravenscroft, J., Ashbolt, N.J., 2010b. Estimated human health risks from exposure to recreational waters impacted by human and non-human sources of faecal contamination. Water Res. 44 (16), 4674–4691. Soller, J.A. and Nellor, M.H. (2011a) Development and Application of Tools to Assess and Understand the Relative Risks of Regulated Chemicals in Indirect Potable Reuse Projects Task 1 - the Chino Basin Groundwater Recharge Project, WaterReuse Foundation, Alexandria, VA. Soller, J.A. and Nellor, M.H. (2011b) Development and Application of Tools to Assess and Understand the Relative Risks of Regulated Chemicals in Indirect Potable Reuse Projects Task 1 - the Montebello Forebay Groundwater Recharge Project, WateReuse Research Foundation, Alexandria, VA. Soller, J.A., Schoen, M.E., Varghese, A., Ichida, A.M., Boehm, A.B., Eftim, S., Ashbolt, N.J., Ravenscroft, J.E., 2014. Human health risk implications of multiple sources of faecal indicator bacteria in a recreational waterbody. Water Res. 66C, 254–264. Soller, J.A., Bartrand, T., Molina, M., Whelan, G., Schoen, M., Ashbolt, A., 2015a. Estimated human health risks from recreational exposures to stormwater containing animal faecal material. Envrion. Modell. Softw. 72, 21–32. Soller, J.A., Nellor, M.H., McDonald, E.T., Cruz, C.J., 2015b. Human health risk associated with direct potable reuse - evaluation through quantitative relative risk assessment case studies. Environ. Sci. Water Res. Technol. 1 (5), 679–688. Stampi, S., Varoli, O., Zanetti, F., De Luca, G., 1993. Arcobacter cryaerophilus and thermophilic campylobacters in a sewage treatment plant in Italy: two secondary treatments compared. Epidemiol. Infect. 110 (3), 633–639. Sykora, J.L., Sorber, C.A., Jakubowski, W., Casson, L.W., Gavaghan, P.D., Shapiro, M.A., Schott, M.J., 1991. Distribution of Giardia cysts in wastewater. Water Sci. Technol. 24 (2), 187–192. Tanner, B.D., Kuwahara, S., Gerba, C.P., Reynolds, K.A., 2004. Evaluation of electrochemically generated ozone for the disinfection of water and wastewater. Water Sci. Technol. 50 (1), 19. Taran-Benshoshan, M., Ofer, N., Dalit, V.O., Aharoni, A., Revhun, M., Nitzan, Y., Nasser, A.M., 2015. Cryptosporidium and Giardia removal by secondary and tertiary wastewater treatment. J. Environ. Sci. Health. Part A, Toxic/Hazard. Subst. Environ. Eng. 50 (12), 1265–1273. Tchobanoglous, G., Burton, F.L., Stensel, H.H., 2003. Wastewater Engineering, Treatment and Reuse, 4th ed. McGraw Hill, New York, NY. Tchoganoglous, G., Leverenz, H., Nellor, M., Crook, J., 2011. Direct Potable Reuse - A Path Forward. WateReuse Research Foundation, Alexandria, VA. Teunis, P., Van den Brandhof, W., Nauta, M., Wagenaar, J., Van den Kerkhof, H., Van Pelt, W., 2005. A reconsideration of the Campylobacter dose-response relation. Epidemiol. Infect. 133, 583–592. Teunis, P.F., Moe, C.L., Liu, P., S, E.M., Lindesmith, L., Baric, R.S., Le Pendu, J., Calderon, R.L., 2008. Norwalk virus: how infectious is it? J. Med. Virol. 80 (8), 1468–1476.

Thurston-Enriquez, J.A., Haas, C.N., Jacangelo, J., Gerba, C.P., 2003. Chlorine inactivation of adenovirus type 40 and feline calicivirus. Appl. Environ. Microbiol. 69 (7), 3979. TWDB, 2014. Direct Potable Reuse Resource Document Texas Water Development Board. Austin, TX, USA. U.S. EPA, 1989. Drinking water: National primary drinking water regulations; filtration; disinfection; turbidity, Giardia lamblia, viruses, Legionella, and heterotrophic bacteria. Final rule. U.S. Environmental Protection Agency, Washington, D.C., p. 27486. U.S. EPA, 1991. Guidance Manual for Compliance with the Filtration and Disinfection Requirements for Public Water Systems Using Surface Water Sources. EPA570391001. U.S. Environmental Protection Agency, Washington, D.C.. U.S. EPA, 1998. Interim enhanced surface water treatment rule. Proposed rule. U.S. Environmental Protection Agency, Washington, D.C., p. 69477. U.S. EPA, 2001. Low Pressure Membrane Filtration for Pathogen Removal: Application, Implementation, and Regulatory Issues. EPA 815-C-01-001. U.S. Environmental Protection Agency, Office of Water, Washington, D.C.. U.S. EPA, 2003. Ultraviolet Disinfection Guidance Manual (Draft). U.S.EPA Office of Water, Washington, D.C.. U.S. EPA, 2006a. National primary drinking water regulations: Long term 2 enhanced surface water treatment rule (LT2ESWTR); Final Rule. 40CFR Parts 9, 141 and 142, 71 (654). U.S. Environmental Protection Agency, Washington, D.C.. U.S. EPA, 2006b. Ultraviolet Disinfection Guidance Manual for the Final Long Term 2 Enhanced Surface Water Treatment Rule. Washington, D.C. EPA 815-R-06-007. U.S. EPA, 2011. Exposure Factors Handbook. U.S. Environmental Protection Agency, Washington, D.C. EPA/600/R-090/052F. U.S. EPA, 2012a. In: Technology, O.o.S.a. (Ed.), Microbiological risk assessment tools, methods, and approaches for water media. U.S. Environmental Protection Agency, Washington, D.C. In preparation. U.S. EPA, 2012b. Guidelines for Water Reuse, EPA/600/R-12/618. Office of Water (ed). Washington, D.C.. Van Abel, N., Schoen, M., Meschke, J.S., 2016. Comparison of risk predicted by multiple norovirus dose-response models and implications for quantitative microbial risk assessment. Risk Anal. In Press. Wallis, P.M., Erlandsen, S.L., Isaac-Renton, J.L., Olson, M.E., Robertson, W.J., van, K.H., 1996. Prevalence of Giardia cysts and Cryptosporidium oocysts and characterization of Giardia spp. isolated from drinking water in Canada. Appl. Environ. Microbiol. 62 (8), 2789. Whiley, H., van den Akker, B., Giglio, S., Bentham, R., 2013. The role of environmental reservoirs in human campylobacteriosis. Int. J. Environ. Res. Public Health 10 (11), 5886–5907. WHO, 2004. Guidelines for drinking-water quality, Third edition Volume 1 World Health Organization, Geneva. WHO, 2006. Guidelines for the safe use of wastewater, excreta, and greywater – Volume 2: Water use in agriculture. World Health Organization, Geneva. Wickramanayake, G.B., Rubin, A.J., Sproul, O.J., 1985. Effect of ozone and storage temperature on Giardia cysts. J. Am. Water Works Assoc. 77, 74–77. WRRF, 2015. Framework for Direct Potable Reuse, Project 14-20. WateReuse Research Foundation, Alexandria, VA. Yang, J., Schneider, O.D., Jjemba, P.K., Lechevallier, M., 2015. Microbial Risk Modeling for Main Breaks. J. Am. Water Works Assoc. 107 (2), E97–E108. Zhou, J., Wang, X.C., Ji, Z., Xu, L., Yu, Z., 2015. Source identification of bacterial and viral pathogens and their survival/fading in the process of wastewater treatment, reclamation, and environmental reuse. World J. Microbiol. Biotechnol. 31 (1), 109–120.

Please cite this article as: J.A. Soller et al., Evaluation of microbiological risks associated with direct potable reuse, Microbial Risk Analysis (2016), http://dx.doi.org/10.1016/j.mran.2016.08.003