Assessment of Sewer Source Contamination of ... - ACS Publications

Sep 7, 2010 - of human enteric viruses in municipal well water, focusing ... were sampled once for wastewater tracers and viruses. Wastewater tracers ...
4 downloads 0 Views 2MB Size
Environ. Sci. Technol. 2010, 44, 7956–7963

Assessment of Sewer Source Contamination of Drinking Water Wells Using Tracers and Human Enteric Viruses R A N D A L L J . H U N T , * ,† M A R K A . B O R C H A R D T , ‡,§ KEVIN D. RICHARDS,† AND S U S A N K . S P E N C E R ‡,§ U.S. Geological Survey Water Resources Discipline, 8505 Research Way, Middleton, Wisconsin, 53562, Marshfield Clinic Research Foundation, Marshfield, Wisconsin 54449

Received March 3, 2010. Revised manuscript received August 16, 2010. Accepted August 18, 2010.

This study investigated the source, transport, and occurrence of human enteric viruses in municipal well water, focusing on sanitary sewer sources. A total of 33 wells from 14 communities were sampled once for wastewater tracers and viruses. Wastewater tracers were detected in four of these wells, and five wells were virus- positive by qRT-PCR. These results, along with exclusion of wells with surface water sources, were used to select three wells for additional investigation. Viruses and wastewater tracers were found in the groundwater at all sites. Some wastewater tracers, such as ionic detergents, flame retardants, and cholesterol, were considered unambiguous evidence of wastewater. Sampling at any given time may not show concurrent virus and tracer presence; however, given sufficient sampling over time, a relation between wastewater tracers and virus occurrence was identified. Presence of infectious viruses at the wellhead demonstrates that high-capacity pumping induced sufficiently short travel times for the transport of infectious viruses. Therefore, drinkingwater wells are vulnerable to contaminants that travel along fast groundwater flowpaths even if they contribute a small amount of virus-laden water to the well. These results suggest that vulnerability assessments require characterization of “low yieldfast transport” in addition to traditional “high yield-slow transport”, pathways.

Introduction Human enteric pathogens are now recognized as potential contaminants of drinking-water wells, but sources of pathogens such as viruses are poorly understood. Recent studies have demonstrated occurrence of human enteric viruses in domestic and municipal wells in the United States (1-4), where more than half of waterborne-disease outbreaks attributable to groundwater consumption are believed to have a viral etiology (5, 6). Outbreaks related to viruscontaminated groundwater have also been documented in * Corresponding author phone: (608) 828-9901; fax: (608) 8213817; e-mail: [email protected]. † U.S. Geological Survey Water Resources Discipline. ‡ Marshfield Clinic Research Foundation. § Current Address: USDA-Agricultural Research Service, Marshfield, Wisconsin 54449. 7956

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 20, 2010

other parts of the world (7, 8), suggesting widespread hydrologic conditions suitable for virus survival and transport. Viruses are much smaller (27-75 nm) than bacterial and protozoan pathogens and thus are more easily transported through pores that physically filter larger pathogens. Virus adsorption onto sediment grains is the primary removal mechanism, although the strength of adsorptive forces depends on sediment and water chemistries (9). These factors notwithstanding, viruses may still be transported some distance, even into confined aquifers at travel rates relevant for human- health concern (10). The U.S. Environmental Protection Agency has listed several viruses on the third drinking water Contaminant Candidate List, emphasizing that waterborne viruses are a research priority (http:// www.epa.gov/ogwdw000/ccl/ccl3.html). Most studies on pathogens in wells have focused on sampling at the wellhead and through the distribution system. Yet investigation of virus sources and transport to the well is necessary to develop strategies that will reduce contamination at the wellhead. Travel time is also an important consideration because viruses are thought to lose their infectivity after one to two years in the subsurface (11-14). Such time information provides context for evaluating whether infectious viruses are expected in a given well. Although groundwater travel times commonly are longer than one year in unstressed systems, they can be much shorter near high-capacity pumping wells, such as those used for water supply (15, 16) because pumping increases the hydraulic gradient. Thus, in order to assess drinking-water well vulnerability to virus contamination, knowledge of virus source and source-to-well travel time is needed. Human enteric viruses are only shed from human gastrointestinal systems, limiting the number of potential sources to those containing human fecal wastes. Surface water downstream of a sewage treatment plant was identified as a source of viruses to a municipal well using stable isotopes of water and numerical groundwater modeling (16). These studies also identified viruses in drinking-water wells that did not have surface water contribution; leaking sanitary sewer lines in the vicinity of the wells were suggested as an additional source (17). Reported estimates of sanitary sewer leakage, or “exfiltration”, range from 1 to 56% of the dry weather flow (18). In the United States, exfiltration has been estimated as 30% of system flow as a result of infrastructure deterioration, and in local areas, sanitary sewer leakage has been reported to be as high as 50% of the system flow (19). The exfiltration rate for a sanitary sewer has been reported on the order of 1 L/m/day (20), and exfiltrated volumes for large municipalities are thought to reach tens of thousands of cubic meters per day (millions of gallons per day), exceeding the capacity of the sediments to filter, absorb, and immobilize contaminants carried therein (21). Even though more research is needed to make general system predictions (18, 22), local sanitary sewers have been related to drinking-water associated outbreaks of gastroenteritis (e.g., see refs 21, 23). Older, nonmaintained systems are thought to be more susceptible to exfiltration, as well as systems including pressurized by sewage lift stations (24, 25). Indeed, of the wells sampled by Borchardt et al. (17), the highest number of positive virus samples was obtained from a well near a pressurized lift station. When the water table is below the utility infrastructure, exfiltrated sewage is often concentrated and transported in the trenches surrounding sanitary sewers, especially during conditions of rainfall-induced infiltration, such that they can threaten drinking-water supplies (22). Sanitary sewer infra10.1021/es100698m

 2010 American Chemical Society

Published on Web 09/07/2010

FIGURE 1. Schematic diagram of additional instrumentation used in the intensive site-scale sampling. structure is often located near municipal wellheads, and carries a high viral load during periods of infections in a community (e.g. refs 26, 27). The objective of the present study was to evaluate sanitary sewers as a potential source of human enteric viruses in drinking-water wells in several communities in Wisconsin. The study (1) evaluated the relation of virus occurrence to wastewater tracer occurrence in a synoptic sampling of municipal wells in 14 communities, (2) instrumented and monitored the groundwater system near a municipal well in 3 of the 14 municipalities, and (3) performed more intensive sampling for virus and wastewater tracers at one of the instrumented sites. Instrumentation and sampling focused on shallow groundwater where sanitary sewer contributions were most likely. With a better understanding of the potential source and transport of sanitary-sewer derived viruses, better assessments of drinking-water well vulnerability can be made.

Experimental Section Sampling Schema. During the initial synoptic sampling in 2005 and 2006, 33 municipal wellheads in 14 communities were sampled once for the determination of viruses and wastewater compounds; 37 wells were sampled twice for stable isotopes of water. Based on these results, one well site in three different communities was chosen for more intensive study, including piezometer installation. Viruses were sampled monthly during four three-month periods in the spring and fall of 2006 and 2007. Wastewater compounds were sampled in the fall of 2006 (monthly, September-November) at all three sites, but in the fall of 2007 monthly sampling (September through November), was limited to one site, Community No. 3. Hydrogeology and Site Instrumentation. Detailed description of the site hydrology and instrumentation is given

in the Supporting Information (SI) Section 1. The municipal wells drew water from unconfined glacial sand and gravel aquifers, or unconfined transmissive bedrock aquifers overlain by sand and gravel aquifers. Detailed site instrumentation characterized subsurface water quality and transport environment at two locations at each well site (Figure 1): an area near the wellhead (“near”), and a second, more distant area along a flowline from a suspected source (“far”). The distance between the near and far wells ranged from approximately 8-25 m at the three sites, and the distance between the municipal supply well and the near wells ranged from 3 to 8 m. Instrumentation consisted of a water table well, data loggers installed in each water table well recording water level and temperature at 15 min intervals, with specific conductance data also included at Community No. 3. Estimates of travel time were obtained by evaluating inflection points in the specific conductance time series. At each of the three sites, piezometer nests were installed near each water table well for depth-discrete water-quality sampling. The total nest sampling consisted of six stainless-steel 15 cm long sampling ports with 15 cm of aquifer backfill separation. The nest was constructed to intersect the water table, leaving at least two sampling ports above the water table to accommodate increases in water table elevation. Sampling and Analyses. Groundwater samples were collected for stable water isotope, nitrate, wastewater tracer, and virus analyses. Detailed description of field collection, field and laboratory analysis, and related quality- assurance methods are described in SI Section 2. A brief description of that material is presented here. Unfiltered samples were collected from high-capacity wells for stable water isotopes δ2H (deuterium) and δ18O during the initial synoptic sampling to identify surface water contributions to the well. Water isotope results are reported VOL. 44, NO. 20, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7957

as ‰ (per mil) differences from the concentrations in Vienna standard mean ocean water (VSMOW). During the initial 2006 synoptic sampling, wastewater tracers were analyzed using filtered samples and USGS National Water-Quality Laboratory (NWQL) Schedule 1433. During the subsequent intensive instrumentation and sampling in 2006-2007, unfiltered water samples were analyzed for wastewater compounds using NWQL Schedule 4433, a new analysis approved by the USGS during the study. Schedule 4433 contains seven more compounds than 1433, and is appropriate for whole (unfiltered) water samples. Both Schedule 1433 and 4433 use polystyrene-divinylbenzene solid-phase extraction and capillary-column gas chromatography/mass spectrometry (28, 29), and include analyses for the following: nonionic surfactants and their degradates, food additives, fragrances, antioxidants, flame retardants, plasticizers, industrial solvents, disinfectants, fecal sterols, polycyclic aromatic hydrocarbons, and commonly used domestic pesticides (SI Section 3, Table 3-1). Municipal well water was collected at the wellhead sampling port; piezometer nests were sampled using a peristaltic pump, and were pumped for at least three well volumes before sampling to ensure a sample representative of the aquifer of the depth sampled. A subset of the groundwater samples were also analyzed for nitrate using in-field spectrophotometric methods. Water samples for virus analyses from municipal wells were sampled by connecting a filter to the wellhead tap while the pump was running (mean sample volume ) 1151 L, n ) 68, range 530-2067 L). Piezometers were sampled by peristaltic pump with the tubing sterilized between samples (piezometer mean sample volume ) 715 L, n ) 5, range 636-757 L). Viruses were concentrated from pumped groundwater using glass wool filtration (30); groundwater pH was reduced to 7.0 during sampling when the ambient pH was g7.5. Filters were shipped on ice to the laboratory within 24 h of sample collection and viruses were eluted with beef extract followed by further concentration to 2 mL by flocculation with polyethylene glycol. Details on the postfiltration elution and concentration methods are described in Lambertini et al. (30). Samples were analyzed by two-step quantitative reverse transcription polymerase chain reaction (RT-qPCR) for enterovirus, norovirus genogroups I and II, hepatitis A virus, rotavirus and by qPCR for adenovirus. Standard curves and reference controls for quantifying each virus type were established as described previously (30). Crossing thresholds were calculated by the second derivative maximum method; all virus concentrations are reported as genomic copies/ liter. Samples positive by qPCR for enterovirus or adenovirus were further evaluated for virus infectivity by observation of cytopathic effect in cell culture. An additional test for infectivity was conducted by integrated cell culture-PCR (ICCPCR) in which, after each passage and at the end of incubation, an aliquot of the lysed cell sheets was analyzed for enterovirus RNA or adenovirus DNA following the same quantitative PCR procedures used for the water samples. The viral nucleic acid concentration in the inoculum increasing 10-fold or more during cell culture was considered indicative of infectious virions even though CPE may not have been observed. Enterovirus and adenovirus serotypes were determined by direct sequencing. Consensus sequences were analyzed with Lasergene software (DNAStar, Madison, WI) and submitted to the National Center for Biotechnology Information BLAST Web site for homology searches (31).

Results I. Synoptic Sampling of Municipal Wells. Eight of 37 wells sampled for water isotopes showed measurable surface water contributions (yielded samples that resided off the meteoric water line, SI Section 2, Figure 2-1). Only wells not exhibiting 7958

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 20, 2010

surface water influence were considered for further investigation, to eliminate potential surface water-derived virus sources (16, 17). Of the 33 municipal wells sampled for wastewater tracers, 12 wells had at least one detection of a wastewater tracer (Table 1), all at concentrations below levels of regulatory concern. However, not all of the detections could be considered unambiguous indicators of impacts from a sanitary sewer source. For example, some of these compounds, such as DEET and caffeine, are widely used and thus have high contamination potential. Other constituents common to wastewaters are not unique to sanitary sewer systems; phenol is a common disinfectant and is used in manufacturing resins, fertilizers, paints, and dyes; phenathrene, naphthalene, and methylnaphthalene may reflect subsurface contamination from historic industrial use, such as coal gasification. For the purposes of this study, the detergent and flame retardant compounds were considered to be unambiguous tracers of sanitary sewer contributions (see SI Section 3). Although not included as a tracer for the purposes of site selection, caffeine is often considered a ubiquitous wastewater tracer and thus is reported when it was detected. Four of the 33 wells sampled in the synoptic sampling were positive for at least one of the unambiguous tracers of sanitary sewer contributions (Table 1). Wells were considered candidates for further intensive monitoring if the synoptic sampling showed unambiguous tracers, or if the virus concentrations in a well were relatively high (described below). Four of the municipal well sites sampled for wastewater tracers during the synoptic sampling were chosen for additional field instrumentation and sampling. Of these four well sites, sufficient quantities of water could not be obtained in the shallow subsurface at one site; therefore only three sites have reported results (italicized rows, Table 1). Among the 33 wells sampled for viruses during the 2006 synoptic wastewater sampling, five were qPCR-positive for viruses (Table 1): two wells were positive for enteroviruses (1.7 and 4.8 genomic copies/L), two were positive for adenoviruses 0.8 and 5.4 genomic copies/L), and one was positive for genogroup II norovirus (77 genomic copies/L). The two enteroviruses tested positive for infectivity by ICCPCR. Concordance between virus and wastewater tracer detection taken at any one time was poor; only two wells were simultaneously positive for both viruses and wastewater tracers when sampled concurrently (Table 1). II. Intensive Characterization, Sampling, and Results from Three Communities. Hydraulic gradients along the piezometer transects were toward the wellhead at all three sites; the municipal well’s effect on the water table varied from extensive in Community No. 1 where the pumping well was screened shallow, to minimal at Community No. 3 in wells pumping from deeper bedrock (SI Section 1, Table 1-1 and Figure 1-1a and 1-1b, respectively). Specific conductance measurements from the Community No. 3 site provide an estimate of travel time based on inflection points in the time series data (SI Section 1, Figure 1-2). The lags in specific conductance inflection from the “far” to “near” monitoring wells were consistent with a flow direction toward the municipal well, and ranged from about 42 days during the relatively higher pumping in the fall to about 66 days during the lower pumping period that followed (NovemberFebruary). Based on these lag times and the distance of 8 m between the far and near piezometers, the Darcy flux is estimated at approximately 0.2 m/d and 0.1 m/d for the August-October and November-February periods, respectively. Assuming similar average travel times for the entire site, viruses move at the same velocity, and 2-year survival of viruses in the subsurface, viruses transported from sources within 100-150 m could be expected to be infectious at the municipal wellhead. Indeed, of 14 PCR-positive samples for enteroviruses

TABLE 1. Results of Jan-Feb 2006 Synoptic Sampling of 14 Communities in Wisconsina virus detected (concentration)

wastewater detects in sample

community

well number

1 1

2 4

neg EV (4.8)

0 3

none phenathrene, tetrachloroethylene, DEET

2 2 2

1 4 5

neg AdV (5.4) neg

0 1 1

none diethoxyoctylphenol pyrene

3

1

neg

3

4-cumylphenol, DEET, Isophorone

3 4 4 5 5 6 7 7 8

2 2 3 3 4 5 2 3 2

neg neg neg neg NV GII (77) neg neg neg neg

0 0 0 1 0 0 0 0 2

none none none 4-cumylphenol none none none none tetraclorethylene, tributyl phosphate

8 8 8

5 6 7

neg neg neg

0 0 5

none none anthracene, fluoranthene, phenathrene, pyrene, carbazole

9 9 10 10 11 11

1 2 2 3 1 2

neg neg EV (1.7) neg neg neg

0 0 0 0 0 3

12

1

neg

1

none none none none none napthalene, 1-methylnapthalene, 2-methylnapthalene tetraclorethylene

12

3

neg

1

tetraclorethylene

12 12

4 5

neg neg

0 2

none tetraclorethylene, DEET

13

1

neg

1

tetraclorethylene

13 13 14 14

4 5 1 2

AdV (0.8) neg neg neg

0 0 0 0

none none none none

analytes detected

compound uses (1) coal tar, diesel, crude oil (2) solvent/degreaser, veterinary anthelmintic (3) nonionic detergent metabolite component of coal tar and asphalt (1) nonionic detergent metabolite, mosquito repellent, solvent for lacquer plastic, oil,

nonionic detergent metabolite

(1) solvent/degreaser, veterinary anthelmintic, (2) flame retardant (1,2,3,4) wood preservative, component of tar/deisel, crude oil, (5) insecticide/dyes/ explosives/lubricants

(1,2,3) Major component of gasoline (1) solvent/degreaser, veterinary anthelmintic, (1) solvent/degreaser, veterinary anthelmintic, (1) solvent/degreaser, veterinary anthelmintic, 2) mosquito repellent (1) solvent/degreaser, veterinary anthelmintic,

Well sites selected for intensive instrumentation are in italics; Virus designations: NV GII ) norovirus genogroup II; AdV ) adenovirus; EV ) enterovirus; Adenoviruses and enteroviruses were not serotyped or tested for infectivity during the synoptic sampling. Virus concentrations reported in parentheses as genomic copies/liter. Neg means the sample was tested and found negative for all viruses. a

or adenoviruses, seven were positive for infectious virus (Table 2). However, the potential source area for contamination could be larger because viruses travel in pathways faster than the apparent average groundwater velocity (32). Nitrate was commonly found in the groundwater sampled from piezometers at the three sites, ranging from 1.3 - 7.9 mg/L-N (SI Section 3, Table 3-5). No discernible relation was evident between nitrate concentration and virus or waste-

water tracer occurrence. During the fall 2006 wastewater sampling, all three sites showed unambiguous wastewater tracers in both the municipal wells and associated piezometer nests, but usually not during the same month (SI Section 3, Table 3-3). The “near” piezometer at Community No. 3 was notable for the largest number of wastewater tracer detections for a single sample collected during the 2006 sampling, and the largest number of detections (4) for nonionic detergent VOL. 44, NO. 20, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7959

TABLE 2. Virus Serotypes, Concentrations, And Infectivity at the Three Community Sitesa year

sample month

2006

2007

April May June September October November March April May September October November

community 1

community 2

well

well

neg NV GI (7.3) neg EV (0.7) neg neg neg neg neg neg neg

Echo 18 (3.9) neg neg EV (0.5) neg neg neg neg neg neg AdV 5 (0.31) neg

community 3 piezometer 1 (far)

AdV 2 (0.02) AdV 5 (0.06)

piezometer 2 (near)

well

neg Echo 30 (0.13) AdV 2 (0.01)

neg AdV (1.3) NV GI (13.6) neg neg neg Cox B2 (0.44) neg neg AdV 41 (0.02) AdV 6 (0.03) AdV 5 (0.02) AdV 2 (0.07)

a Piezometer nests in Communities 1 and 2 were not sampled for viruses. Virus designations: NV GI ) norovirus genogroup I; AdV ) adenovirus; EV ) enterovirus; Echo ) echovirus; Cox ) coxsackievirus. EV or AdV without a listed serotype indicates unable to sequence. Virus concentrations reported in parentheses as genomic copies/liter. Italics indicates the virus tested positive for infectivity by ICC-PCR. A missing table entry means the sample was not collected; neg means the sample was tested and found negative for all viruses.

metabolites (Piezometer N24, SI Section 3, Table 3-3). There are multiple nearby sanitary sources, including an upgradient sanitary sewer line, a suspected improperly abandoned septic system, and a pressurized lift station. There are no upgradient potential surface water sources of viruses at Community No. 3. Although this community was characterized by the high number of tracer detections, none of the wastewater tracer detections exceeded regulatory concentrations. Subsequent monthly sampling during September-November 2007 at the Community No. 3 municipal well and associated piezometers showed wastewater tracers in both the municipal well and piezometer nests, but not necessarily during the same month or the same tracers (SI Section 3, Table 3-4). A relatively high number of wastewater tracers (6) were again detected at piezometer N24 in the October sample, similar to the seven detections in the October 2006 sampling. During 2006-2007, samples for virus analyses were collected monthly from the three community municipal wells. Each well had two or more virus-positive samples (Table 2). Six of the twelve samples from the Community No. 3 well were positive for adenoviruses, enterovirus (Coxsackievirus B2), or genogroup I norovirus. In the 2007 sampling of piezometers at this site, both the near and far piezometer nests had two virus-positive samples; one piezometer contained infectious adenovirus 2 and the other piezometer was positive for infectious echovirus 30 (Table 2). Along the hydraulic gradient from the far to the near piezometer to the well, two of the three detected serotypes (adenoviruses 2 and 5) were common to all three sampling locations, suggesting a common source. Within a given sample location, well or piezometer, concordance in the occurrence of viruses and unambiguous wastewater tracers was excellent. That is, every virus-positive well and piezometer was also positive at some point during the study for an unambiguous wastewater tracer.

Discussion Characterizing Source and Transport. Wastewater tracers were generally in relatively restricted depths within the aquifer, usually near the water table (Table 3), indicating (1) the source is nearby, and (2) that characterization of a wastewater plume or virus source would require detailed vertical sampling of the shallow saturated zone. The notable exception to the limited vertical distribution was observed at piezometers near the wellhead at Community No. 1 (Table 3), likely a result of the dynamic water table at the site. When wastewater tracer and virus results are evaluated by month 7960

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 20, 2010

and by tracer, it is difficult to find correspondence between location, tracer, and virus occurrence at any one sampling time/event. Wastewater tracers may not correspond along the expected flowpath due to the study design that limited tracer sampling to the fall of 2006 and 2007, thus potentially missing the tracer movement along flowpaths during the other seasons. It is also possible that the near and far piezometer nests may not be along the same flowline or that transience in the groundwater flowpaths causes them to include different flowlines over the study period. Nitrate was ubiquitous throughout the profile (SI Section 3) and thus not a useful indicator of source and transport. The poor correspondence between colocated/concurrent wastewater tracers and virus results is likely a result of method sensitivity and source history. Both virus and wastewater tracer analyses can suffer from poor or variable recovery. However, virus analyses are more sensitive than wastewater tracer analyses, and preconcentrate many liters of raw water onto a filter during sampling. Thus, dilution of wastewaterderived water by nonwastewater-derived water could reduce wastewater tracer concentrations to levels below detection (microgram/L minimum detection limit), yet viral RNA or DNA could remain above detection because the detection limit is in the femtogram/L range. Viral sources, on the other hand, are more variable than wastewater tracer sources because of their clinical seasonality and the episodic nature of infections and fecal shedding. Wastewater tracers such as flame retardants and detergents have more consistent use and related presence in the waste stream, and would likely be detected more frequently if the analytical methods were more sensitive. Even though the waste stream concentrations may not vary widely, the large monthly variability in wastewater tracer occurrence in Community No. 3 piezometer N24 (SI Section 3, Tables 3-3 and 3-4) demonstrates that large temporal variability might still be expected due to other processes such as temporal changes in groundwater flowpaths, or differing removal properties of viruses and associated tracer. Much better correspondence between wastewater tracers and virus occurrence is observed when groundwater is sampled multiple times over a sufficiently long period (Table 3). Given more sampling, all sampling locations that had virus positive results also had unambiguous tracer occurrence during at least one of the sampling events. Sampling over longer time periods provides a more robust characterization of well vulnerability and transport because it can account for confounding processes such as transience in the ground-

TABLE 3. Summary of Results of 2006 and 2007 Wastewater and Virus Intensive Sampling (Brackets Show Composite Sample Used for Virus Analysis)

water flow field, variability in source distance and concentration, and episodic groundwater recharge. This approach may be preferred to use of standard microbiological indicators of viruses (e.g., Escherichia coli, coliform bacteria) that have been shown to give false negatives (e.g., ref 17), likely due to the physical straining in the aquifer of the larger bacterial indicators. Although a universal sampling design is likely not tenable, multiple virus samples over several seasons is necessary to adequately characterize well vulnerability to virus contamination. Given the recent advances in virus-

related analytical techniques, however, the capability to run large sample numbers has increased such that the need for more frequent sampling is less problematic than in the past. The high number of detects in a single well of tracers and viruses unique to wastewater demonstrates the potential importance of leaking sanitary sewers as a contamination source to drinking-water wells. Thus, assessments of drinkingwater vulnerability should include proximity and condition of sanitary sewer lines, in addition to the possibility for surface water sources such as those identified by Borchardt et al. VOL. 44, NO. 20, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7961

(17) and Hunt et al. (16). Such a finding is consistent with the relatively high rates of leakage expected in community systems, and the common close proximity of sanitary sewer lines to municipal wells as urbanization surrounded and moved beyond historic well fields in the three communities. Given the expected