APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1977, p. 609-619Copyright C 1977 American Society for Microbiology

Vol. 33, No. 3Printed in U.S.A.

Virus and Bacteria Removal from Wastewater by RapidInfiltration Through Soil

STEPHEN A. SCHAUB* AND CHARLES A. SORBER

U.S. Army Medical Bioengineering Research and Development Laboratory, Ft. Detrick, Frederick, Maryland21701,* and The University of Texas at San Antonio, San Antonio, Texas 78285

Received for publication 29 September 1976

A rapid infiltration land wastewater application site, composed of unconsoli-dated silty sand and gravel, which has been in continuous operation for over 30years was examined for the accumulation and/or migration of a tracer virus(coliphage f2), indigenous enteroviruses, and enteric indicator bacteria in thesoils and underlying groundwater. Tracer f2 penetrated into groundwater to-gether with the front of percolating primary effluent and was not observed toconcentrate on the upper soil layers. The tracer virus concentration in a 60-foot(about 18.3-m)-deep observation well directly beneath the wastewater appli-cation area began to increase within 48 h after application to the soil. The tracerlevel in this well stabilized after 72 h at a level of approximately 47% of theaverage applied concentration. Indigenous enteroviruses and tracer f2 were

sporadically detected in the groundwater at horizontal distances of 600 feet(about 183 m) from the application zone. Laboratory soil adsorption studiesconfirmed the poor virus adsorption observed at the site. This was especiallytrue on surface soils when contained in wastewater. Enteric indicator bacteriawere readily concentrated on the soil surface by filtration on the soil surfacemat. However, during tracer f2 virus tests, comparison studies with fecalStreptococcus revealed that bacteria capable of penetrating the surface were

able to migrate into the groundwater. They were detected at the same locationsas tracer and enteric viruses.

Land application has been used as a waste-water disposal method for many years. Waste-water can be applied by spray irrigation, over-land flow, or rapid infiltration. Treatment ef-fectiveness is a function of the physical, chemi-cal, and biological properties of the soil as wellas wastewater application practices.There are several major advantages to waste-

water disposal by land application, includingrelatively low system capital and operatingcosts, unsophisticated facility requirements,the value of the nutrients imparted to the land,and recharge of groundwater by infiltration-percolation. Unfortunately, there are severalpotential problem areas which need more defi-nition. For example, the actual effectiveness ofland application of wastewater in removing en-teric microorganisms is not fully understood.A number of studies have attempted to evalu-

ate the enteric microbial removal potential ofsoils. McGauhey and Krone (9) summarizedexisting information and concluded that coli-forms and other bacteria move only a few feetin an unsaturated zone and several hundredfeet in a saturated zone. These investigatorsindicated that bacteria behave like other par-

ticulates in soil and are removed by straining,sedimentation, entrapment, and adsorption,and are influenced adversely by antagonisticenvironmental factors. They concluded that soilsystems are quite efficient in removing bacteriaand hypothesized that viruses can be removedwith equal effectiveness, principally by adsorp-tion.

Driver and co-workers (4) reviewed previousstudies and supported McGauhey and Krone'sconclusions. They suggested that the rapid in-filtration and spray irrigation modes of waste-water treatment should provide the most effi-cient removal of bacteria and viruses of anyland wastewater treatment method. Further,they concluded that, depending on the nature ofthe soil, soil depths ranging from 1 cm to 2.1 mare sufficient to remove bacteria and virusesfrom wastewater, and that the mechanisms ofremoval are similar in most soil systems, ex-cept that the relative effectiveness of the var-ious mechanisms differs according to the mode ofland application. They suggested that, for rapidinfiltration and spray irrigation systems, themost important microbial removal mechanismis filtration, followed by fixation or adsorption.

609

610 SCHAUB AND SORBER

A recent review of the fate of wastewatermicroorganisms in soil by Gerba et al. (6) sum-marizes the known mechanisms of their re-moval in soils. They indicate that whereas fil-tration on the soil is the principal mechanism ofbacterial removal, virus removal is predomi-nantly by adsorption to soils, especially clays.They cite recent research which indicates thatchanges in the water or soil characteristics mayallow virus de-adsorption and subsequentlypenetration to deeper soil layers.The requirement for research into potential

health and hygiene problems associated withland application of wastewater was recognizedand discussed by Sorber and co-workers (13).They pointed out that, during land applicationof wastewater, particularly when rapid infiltra-tion (RI) was the mode of application, entericviruses and bacteria, as well as potentiallytoxic chemicals (e.g., heavy metals and pesti-cides), may reach the groundwater.The purpose of this study was to determine

the effectiveness of removal of enteric bacteriaand viruses by the rapid infiltration mode ofwastewater treatment. This study was dividedinto two parts. The first part was designed toprovide base-line data on selected microbial pa-rameters which involved laboratory investiga-tions with soil and wastewater samples col-lected at the site. Soil and well sampling tech-niques were developed during two 1-day sam-pling periods.The second part of the investigation was de-

signed to determine the capability of the rapidinfiltration system to treat or remove entericbacteria and viruses over a total treatment cy-cle. A virus tracer study was performed. Indica-tor organisms were assayed in soil and ground-water samples. In addition, enteric viruseswere concentrated and assayed from groundwa-ter samples.

MATERIALS AND METHODSStudy site. The site selected for study was located

at Fort Devens, Mass., a U.S. Army installationsituated northwest of Boston. During this study theinstallation had a permanent resident population ofapproximately 10,000 in addition to a daily workforce of 5,000.The sewage treatment facility has been in contin-

uous operation since 1942. Collected sewage is com-minuted and pumped from a central pump station toImhoff tanks which have a retention time of approx-imately 6 h. Generally, the wastewater flow is be-tween 1.0 and 1.5 million gallons/day (3,785 to5,678 m3/day).The Imhoff tank effluent is distributed through a

dosing tank to the RI cells without chlorination.There are 22 RI cells, each with an average area ofabout 0.8 acre (0.324 hectare), constructed on a steep-

APPL. ENVIRON. MICROBIOL.

sided kame. The soil beneath the RI cells is princi-pally unconsolidated silty sands and gravel of gla-cial origin, underlaid with bedrock.Under normal operating conditions, three RI cells

receive wastewater for a 2- to 3-day period. A subse-quent period of 2 days is required for complete ef-fluent infiltration. At that time, the cells are al-lowed a resting period of 10 days. During the yearprior to this study, a 7-day application period fol-lowed by a 14-day resting period was used. The RIcells are cycled on a regular basis and receive ef-fluent for approximately 50 days each year. Winterconditions, although reducing infiltration some-what, do not hamper normal operations.

During the past 30 years, the major operationalproblem has been clogging of the soil on the rapidinfiltration beds. To alleviate this problem, the bedsare renovated by removal of 1 to 4 feet (0.305 to 1.22m) of the surface soils and replacing them with bankrun sand and gravel after loosening the exposedsubsurface layers. The most recent renovation oc-curred in 1968. During this renovation, a black as-phaltic-appearing layer was observed at a soil depthof between 18 and 24 inches (44 to 59 cm).Numerous observation wells were installed at the

study site. They were wash-drilled with diamondcore drills and they were cased during the drillingoperation. Tap water was used for substrate removalfrom the casing. Fine sediments were removed bypumping. Observation wells no. 3, 7, and 8 havemetal riser pipe and no. 10 metal screens. All otherwells were installed with PVC riser pipes and no. 20PVC screens. Protective well casings to a depth of atleast 4 feet (1.22 m) below the ground surface wereprovided to all of the wells except no. 3 and 8. In allcases, riser pipes were extended 36 inches (92 cm)above the surface and metal caps were installed toprevent contamination. Specifications of the obser-vation wells used during this study can be found inTable 1.

Virus stock preparation and assay. Encephalo-myocarditis virus (EMC) and poliovirus I (Chatstrain) were utilized for laboratory studies of entero-virus adsorption to soils. The EMC virus was grownand assayed on mouse L continuous culture cell linewhereas poliovirus I was grown and assayed onBGM and KB continuous cell lines. Procedures usedin the growth and maintenance of cell lines as wellas cell culture techniques, virus filtration, and virusplaque assay have been detailed elsewhere (11).The f2 bacteriophage was grown as described by

Loeb and Sinder (8), as modified by Kruse et al. (7).Eleven- to 23-liter stocks were grown in a laboratoryfermentor where a concentration of over 1012 plaque-forming units (PFU)/ml was attained. Chloroformwas added as a preservative and the stocks weretemporarily stored under refrigeration. The f2 wereassayed by the soft agar overlay method (tryptoneyeast extract) utilizing a lawn of male Escherichiacoli, strain K-13 (American Type Culture Collectionno. 15766). Incubation was at 35°C for 18 h. Fieldsamples were heated to 370C for 5 min prior to assayto eliminate the chloroform preservative added inthe field.

Virus adsorption studies. The two enteric viruses

MICROBIAL TREATMENT BY RAPID INFILTRATION 611

TABLE 1. Observation well specificationsaObser- Ground Well Wellvation surface de,h screen Well diamwell elevation u(feet) length (inches)no. (feet) (feet)

3 209.30 9.5 3.0 1.257 223.18 28.4 10.0 1.258 219.91 19.0 3.0 1.25

10 218.91 22.8 10.0 1.2511 212.39 14.9 10.0 1.2515 269.31 94.5 40.0 2.5017 213.41 18.0 10.0 1.2519 211.56 22.5 10.0 1.2520 210.89 16.6 10.0 1.25

a One foot = 0.348 m; 1 inch = 2.54 cm.

and f2 bacteriophage were used to evaluate the ad-sorptive capacity of the various soils. Autoclaved,weighed samples (10 g) of soil from unrenovated orrenovated (1968 renovation) cells were mixed with40 ml of autoclaved primary treated sewage ordeionized water containing 10-3 M CaCl2. The mix-ture was stirred vigorously for 5 min with a mag-netic stirrer. At that time, virus to a concentrationof 104 PFU/ml was added and the samples werestirred slowly for a 30-min adsorption period. Ali-quants (10 ml) were centrifuged at 12,000 x g for 10min to separate the solids and associated virionsfrom the liquid phase. The supernatant portion wasassayed for infectious virus on the appropriate cellculture. Virus titers were expressed as PFU pergram of sample.

Virus field studies. The tracer study at the sitewas performed by continuous addition of stock f2bacteriophage into the wastewater at the dosingtank with a peristaltic pump over the 7-day applica-tion period. The pump flow was adjusted daily toprovide a virus concentration of 105 PFU/ml ofwastewater applied to the RI cells. Cells no. 10, 15,16, and 17 were utilized for the tracer tests. Effluent,soil, and groundwater samples were analyzed for f2bacteriophage prior to, during, and after dosing ofthe wastewater.The migration of indigenous enteric viruses from

the wastewater applied to the RI cells was evaluatedduring the f2 tracer virus field study. One-liter com-posite samples of wastewater standing on the bedsand 4-liter samples of groundwater obtained fromobservation wells were concentrated for enteric vi-ruses by the bentonite adsorption procedure. A 70-mg amount of powdered bentonite clay per literfollowed by CaCl2 to provide a concentration of 10-2M was added to the samples. Samples were inter-mittently shaken for 20 min to allow the virions toadsorb to the clay. The samples were then filteredthrough a 142-mm-diameter AP25 fiber glass (Milli-pore Corp.) prefilter to collect the bentonite andadsorbed virions. The adsorbed virions were elutedfrom the clay on the prefilters by the addition of 10ml of tryptose phosphate broth, and eluates weredrawn through the filter by suction and collected ina small vacuum flask. The samples were placed in 10ml of polypropylene snap-lock tubes and frozen untilassay. Upon thawing, ethylenediaminetetraacetic

acid (0.1 M) was added to dissolve a precipitatewhich formed during freezing, and 0.25 ml of samplewas placed upon replicate sets of BGM and KB cellmonolayer tissue culture plates. After a 60-min vi-rus attachment period, the residual sample was as-pirated from the plates to lessen bacterial over-growth and the solid overlay was added. Four daysof incubation were allowed to enhance detection ofenterovirus plaques.Grab samples of primary effluent, ponded waste-

water in RI cells and the Nashua River, were takenwith a dip spoon. The RI wastewater samples were acomposite of equal portions from each test cell. Sam-ples for microbiological analysis were placed in 1-gallon (3.785-liter) plastic containers or 32-ounce(0.95-liter) sterile prescription bottles and iced untilassay. Samples collected during the f2 bacterio-phage tracer study were placed in 5-ml polypropyl-ene snap-lock tubes. Chloroform and 0.5 ml of 5xnutrient broth were added, and the tubes were fro-zen until analysis.

Sampling of observation wells was conducted byplacing a weighted plastic tube into the wells to adepth of 6 inches (15.2 cm) beneath the groundwatersurface. Suction was applied by a hand pump whichdelivered the samples to a 1-liter fleaker bottle. Thefirst liter of sample was discarded, and subsequentportions were collected for analysis. In observationwells no. 2 and 15, the depth to groundwater was toogreat to obtain the samples by suction; thus, thesewells had to be sampled with a 400-ml-capacity,Kemmerer-type water sampler (7). Again, the firstsample volume collected was discarded. Sampleswere placed in containers as previously described.

Water table measurements were made routinelyat the observation wells during the bacteriophagetracer study. A measured line connected to a con-ductivity probe and meter was utilized for thesedeterminations. Measurements were taken from thesurface of the groundwater table to the top of thewell stand pipes.

RESULTSVirus adsorption to soils. Poliovirus I, EMC

virus, and f2 bacteriophage were utilized tocompare the capacity of renovated and unreno-vated RI cell soil layers to adsorb the virions.

Enteric indicator bacteria studies. Indige-nous indicator bacteria concentrations of thewastewater, soil, and groundwater were deter-mined. Total coliform, fecal coliform, and fecalStreptococcus analyses were performed utiliz-ing membrane filter techniques as describedpreviously (1). Results were expressed as col-ony-forming units (CFU) per volume of liquidor dry weight of soil.

Microbiological and hydrological samplingand preparation. Soil sampling for microbio-logical analysis was performed on five visiblydistinct soil layers extending to approximately30 inches (73.5 cm) beneath the surface. Shoveland pick were used to dig the proper depth, andthen one side of the hole was carefully cut to

VOL. 33, 1977

612 SCHAUB AND SORBER

provide a vertical face for sampling. Samplesfrom each soil layer were taken from the mid-point of the layer by scooping into the verticalcut face with a Teflon spoon. The first severalscoops of each layer were discarded to minimizecontamination from other soil layers. All soilsamples were sealed in doubled plastic bags,and placed on ice until analysis. Samples formicrobiological analyses were processed within8 h of collection.Total coliform, fecal coliform, fecal Strepto-

coccus, and f2 bacteriophage were eluted fromthe soils by the addition of nutrient broth at pH7.2 (5 g of soil and 10 ml of eluent) followed byvigorous mixing with a magnetic stirrer for 5min. Heavy particles were allowed to settle for1 min, and the supernatant portions were as-sayed as described previously.

Separate aliquants of the soil samples wereweighed into petri dishes, dried overnight at1050C, and reweighed. The soil dry weight wasused in the calculation of virus and indicatorbacteria.

Figure 1A illustrates EMC virus adsorptionin deionized water containing 10-3 CaCl2 forsoils from renovated and unrenovated cells.Neither surface soil was capable of extensiveadsorption. Below the surface layer, 60 to 80%of the virus was adsorbed by the soil from therenovated cell. The soil from the unrenovatedcell, on the other hand, did not retain signifi-cant virus in the upper subsurface layers but

A.EMC Virus in Distilled B.EMC Virus in

Deionized H20 + Primary Effluen10-3 MCoC12

0 Lin

0

401!

began to approach the adsorptive capacity ofrenovated cell soil at the 18-inch (45.7-cm)depth. At the 30-inch (76.2-cm) layer, the soiladsorptive capacity of renovated and unreno-vated cells was equal.Replacement of the suspending fluid with do-

mestic, primary effluent resulted in a signifi-cantly different relationship between both soilsand EMC virus. Figure 1B shows that, withsewage, the adsorptive capabilities of soils ob-served in the distilled water tests were notattained. A maximum of 7.5% adsorption wasobserved. Comparative studies utilizing polio-virus I also reveal poor virus adsorption capa-bilities for the soils saturated with primaryeffluent (Fig. 1C). Virus adsorption reached24% at the mid-depths, but overall adsorptionwas about 10% for soils from both cells. Studiesutilizing f2 bacteriophage with primary ef-fluent indicated that there was essentially noadsorption of this virus to any of the soil layers(Fig. 1D).

f2 Bacteriophage adsorption at the fieldsite. For comparison ofthe laboratory results toactual site conditions, the soils in renovatedcell no. 16 and unrenovated cell no. 17 wereanalyzed prior to and during the f2 bacterio-phage tracer study. Samples from soil layerssimilar to those analyzed in laboratory virusadsorption studies were treated with tryptosephosphate broth for elution of f2 bacteriophage.Table 2 shows the soil concentrations of f2 bac-

O 3 10 18 30 0 3 10 I8 301 0 3 10 18 30' 0Soil Loyer Mid-depth (Inches)

FIG. 1. Virus adsorption to soils from various depths. The selected liquids and soils were mixed together ata 5:1 ratio. Virus was added and allowed to adsorb to the soils for 30 min. The mixture was centrifuged topellet the soils and associated virions. Unadsorbed virions in the supernatant portion were assayed. Thepercent virus adsorption was determined from the supernatant virus as a fraction of virus assayed in themixture prior to centrifugation.

APPL. ENVIRON. MICROBIOL.

MICROBIAL TREATMENT BY RAPID INFILTRATION 613

teriophage in terms of soil dry weight. A sur-

prisingly high background concentration of in-digenous bacteriophage was observed at thesoil surface in both cells on day zero. Once pastthe surface, the concentration of virus was

much lower. On day 13 of the test, the concen-

tration of virus on cell no. 16 was uniform over

the entire soil profile. At this time, all theeffluent had infiltrated past the soil surface andthe soil had a high water content. The unreno-

vated cell no. 17 still retained 6 to 8 inches (15.2to 20.3 cm) of water on this sampling day. Onday 21, a uniform distribution of virus was

observed in cell no. 16, albeit somewhat lowerthan on day 13. Cell no. 17 had a dry surface atthis time, and comparable but somewhat lessuniform virus concentrations with depth were

observed. It is important to note that, afterinfiltration was complete in each cell, the ob-served bacteriophage concentration was ap-

proximately that which could have been pre-

dicted from laboratory test results based on theaverage concentration of virus dosed into thesystem (2.5 x 105 PFU/ml of sewage). Sinceonly minor amounts of virus were retained atthe surface, it would appear that the viruspassed through these upper soil layers withease.

Indicator bacteria association with soils atthe field site. The soils obtained for enterovirusassociation experiments also were examined for

enteric indicator bacteria. Bacteria were elutedfrom separate soil samples as previously de-scribed. The eluates were assayed for total coli-form, fecal coliform, and fecal Streptococcus.Table 3 shows that the concentration of micro-organisms at the soil surface was significantlygreater than it was in the subsurface soils.Once past the soil surface, there was a more

gradual decline in the numbers of organismswith RI cell depth. In all samples there werehigher concentrations of total coliform than fe-cal coliform and there were more fecal coliformthan fecal Streptococcus. The ratio of both fecalcoliform and fecal Streptococcus to total coli-form increased once past the soil surface.During the f2 tracer study, indigenous fecal

Streptococcus analyses were conducted to com-

pare the movements of indicator bacteria tothose of the bacteriophage. The average fecalStreptococcus concentration found in the ef-fluent applied to the cells was 105 CFU/100 ml.Table 4 shows that the zero time samples con-

tained relatively few fecal Streptococcus ineither cell no. 16 or 17, and those detected wereprincipally in the surface layer. However, nei-ther cell had received effluent for at least 4weeks. After flooding and infiltration/percola-tion, there were appreciable bacterial concen-trations on the soil surface, but the concentra-tions diminished once past the surface. Com-parison of the day 13 and day 21 sampling pe-

TABLE 2. f2 Bacteriophage tracer at various soil depthsPFU/g of dry soil

Samplea Day 0 Day 13, Day 21

Cell 16 Cell 17 cell 16 Cell 16 Cell 17

Surface 8.7 x 103 1.8 x 103 8.6 x 102 4.1 x 102 9.9 X 1023 inches 7.4 x 102 5.6 x 101 7.8 x 102 2.1 x 102 2.6 x 10210 inches 1.1 x 101 1.4 x 101 9.7 x 102 1.3 x 102 6.8 x 10118 inches 8.9 x 101 4.5 x 101 1.1 X 103 3.7 x 102 2.9 x 10130 inches 4.7 x 101 2.1 x 102 2.7 x 102 9.0 x 101

a One inch = 2.54 cm.

TABLE 3. Preliminary analysis for indigenous indicator organismsCFU/g (dry wt) of soil or CFU/ml of effluent

Sample" Total coliform Fecal coliform Fecal Streptococcus

Mar. 1974 June 1974 Mar. 1974 June 1974 Mar. 1974 June 1974

Effluent on 4.2 x 105 4.2 x 104 (0.1)a 8.1 x 102 (0.002)RI cells

Soil surface 9.6 x 106 2.5 x 107 3.5 x 104 (0.004) 2.5 x 102 (0.1) 1.8 x 104 (0.002) 6.2 x 104 (0.003)6-inch depth 2.5 x 105 7.8 x 104 1.8 x 104 (0.07) 2.5 x 104 (0.3) 1.4 x 103 (0.006) 6.6 x 102 (0.009)12-inch depth 2.3 x 104 7.5 x 103 (0.3) 1.5 x 102 (0.007)

a Numbers in parentheses indicate the ratio of the indicator organism concentration to total coliform concentration forthe same test period and sample.

b One inch = 2.54 cm.

VOL. 33, 1977

614 SCHAUB AND SORBER

riods on cell no. 16 suggest that the 7-daydrying period did not severely affect fecal Strep-tococcus levels, especially in the upper soil lay-ers.Groundwater migration. Groundwater lev-

els in the observation wells were measureddaily during the tracer study from day 0through day 7 and again on days 10, 13, and 20.Groundwater contours were developed fromthese data. Of particular significance are thevarious stages of groundwater mounding dur-ing the infiltration period. At day zero nomound existed below the beds being flooded andthe groundwater flow was generally in the di-rection of the Nashua River (Fig. 2). A ground-water mound began to build within 2 days andit became relatively stable after 5 days of infil-tration (Fig. 3). After flooding of the beds

TABLE 4. Indigenous fecal Streptococcus at varioussoil depthsCFU/g of dry soil

Soildepth Day 0 Day 13, Day 21(inches)"ael1

Cell 16 Cell 17 celi 16 Cell 16 Cell 170 1.6 1.0 x 102 2.9 x 103 1.5 x 103 1.3 x 1023 3.0 3.0 4.9 x 102 5.3 x 102 1.7 x 10110 <3.0 <3.0 1.9 x 102 1.3 x 102 1.3 x 10118 <3.0 <3.0 8.5 x 101 1.6 x 101 <3.030 <3.0 3.0 x 10' <3.0 <3.0

a One inch = 2.54 cm.

ceased, the mound began to decay and by day 20only a small portion of the mound remained.

Hydraulic conductivity, void ratio, and grainsize analyses were conducted on soil samplesfrom the RI site. Hydraulic conductivity wasapproximately 10-2 cm/s under unit head condi-tions. This would result in a field penetrationtime of 48 to 72 h for the wastewater to reachthe groundwater table. Since hydraulic conduc-tivity would change slightly with changes invoid ratio and grain size, the laboratory datacannot predict lateral movement within thesaturated aquifer.

Figure 4 illustrates the elevation of thegroundwater mound directly beneath theflooded RI cells as observed in observation wellno. 15. It can be seen that the mound decayed ata linear rate upon cessation of flooding of thebeds. Through extrapolation it would appearthat the water table would return to its initialelevation within 17 days after termination ofwastewater application.Tracer movement. The f2 bacteriophage was

introduced in the primary effluent to RI cellsno. 10, 15, 16, and 17 to determine migrationinto the groundwater. An effective concentra-tion of 105 PFU/ml was maintained over the 7-day wastewater application cycle.The standing wastewater from all four beds

was composited every 8 h during the bacterio-phage application period. Samples from obser-

FIG. 2. Groundwater contour prior to wastewater application to study cells. The groundwater level wasdetermined with a measured conductivity probe at the observation wells 1 h before wastewater application tothe cells commenced.

APPL. ENVIRON. MICROBIOL.

MICROBIAL TREATMENT BY RAPID INFILTRATION 615

FIG. 3. Groundwater contour on the fifth wastewater application day. The groundwater level was deter-mined with a measured conductivity probe at the observation wells after 120 h of wastewater application.

210

*210^

X 208.

206

2040 5 10 15 20 25

Time After Wostewoter Applicotion (doys)

FIG. 4. Groundwater mound at observation welldirectly beneath wastewater application zone. Obser-vation well 15 was observed for groundwater mound-ing with a measured conductivity probe for 21 daysafter testing was initiated.

vation wells no. 10, 11, 15, 19, and 20 were takenat the same times. Observation wells no. 3, 7, 8,17, and 21 were sampled every 24 h. Also, sam-ples from all wells were obtained on days 10, 13,and 21. Results can be found in Table 5. Figure

5 illustrates bacteriophage movement into thegroundwater beneath the RI beds as observedat well no. 15. The virus reached the well afterapproximately 48 h, the same time as thegroundwater mound began to form. The viruslevels peaked and stabilized between days 3and 6. The average bacteriophage concentra-tion in well no. 15 during this peak period was1.17 x 105 PFU/ml, which was approximately47% of the composite average on the RI cells.After application to the cells ceased, the tracerconcentration declined steadily. By day 20, thelevel was only 2% of the peak concentration.A comparison of tracer bacteriophage in the

outlying observation wells with well no. 15 andthe composite is provided in Table 5. It is inter-esting to note that frequent "spikes" of f2 wereobserved in the perimeter wells most directly inthe path of groundwater flow. High concentra-tions were noted in well no. 20 at 72 h (100% ofthe composite) and in well no. 10 at 156 h (7.2%of the composite). The tracer concentrations inobservation wells no. 3, 7, 8, 17, and 21 duringthe test period were relatively low.

Indigenous enteric virus movement. Duringthe f2 tracer study, samples from observationwells no. 10, 11, 15, 19, and 20 plus the compos-ited wastewater on the cells were concentratedfor indigenous enteric virus analyses. Table 6shows that the average indigenous enteric vi-rus concentration for the composite effluent

VOL. 33, 1977

TABLE 5. Concentrations of tracer f2 bacteriophage on RI cells and in observation wells

PFU/mlSample

O.la 1.0 1.3 1.7 2.0 2.3 2.7 3.0

RI cellsb 3.9 x 105 3.6 x 105 2.7 x 105 2.3 x 105 1.6 x 105 2.4 X 105 2.1 X 105 1.4 X 105Well 15 1.3 x 102 3.3 x 100 1.9 x 101 2.3 x 10' 7.3 x 102 2.6 x 103 5.8 x 104 1.3 x 105Well 10 4.2 x 10° 1.0 x 100 2.3 x 100 5.6 x 100 4.3 x 100 5.2 x 101 2.3 x 101 1.0 x 100Well 11 4.0 x 100 6.0 x 100 4.8 x 101 1.0 x 10' 1.7 x 103 4.6 x 10° 4.0 x 100 2.5 x 103Well 19 6.5 x 10° 4.0 x 100 4.6 x 100 1.3 x 100 1.8 x 102 0.3 x 100 2.3 x 101 9.8 x 10'Well 20 <1.0 x 100 4.0 x 100 4.0 x 100 1.5 x 102 0.3 x 100 <1.0 x 100 2.7 x 105

Sample 3.3a 3.7 4.0 4.3 4.7 5.0 5.3

RI cells 2.1 x 105 1.4 x 105 2.6 x 105 3.2 x 105 3.5 x 105 2.3 x 105 3.0 x 105Well 15 1.4 x 105 6.1 X 104 1.5 x 105 8.4 x 104 7.5 x 104 1.6 x 105 8.7 x 105Well 10 1.3 x 100 1.2 x 101 2.2 x 101 1.3 x 100 7.5 x 104 1.6 x 105 8.7 x 104Well 11 3.6 x 100 5.0 x 100 2.3 x 100 1.0 x 100 9.3 x 10° 3.3 x 100 8.0 x 100Well 19 1.6 x 100 <1.0 x 10° 9.0 x 100 2.0 x 100 3.0 x 101 1.0 X 100 0.3 X 100Well 20 1.1 x 10' 2.0 x 10' 1.7 x 100 3.6 x 100 1.0 X 103 2.5 x 103 1.0 x 100

Sample 5.7a 6 6.5 6.9 11 14 21

RI cells 1.6 x 105 3.7 x 105 1.7 x 105 2.0 x 105Well 15 1.3 x 105 1.3 x 105 1.2 x 105 3.8 x 104 1.9 X 104 2.3 x 104 2.4 x 103Well 10 8.5 x 104 1.7 x 100 1.8 x 104 >3.0 x 102 1.2 x 101 <1.0 x 100Well 11 0.6 x 100 5.0 x 10' 0.6 x 100 <1.0 x 100 5.6 x 101 <1.0 x 100 1.0 x 100Well 19 <1.0 x 100 2.6 x 100 2.0 x 100 1.6 x 102 3.0 x 100 <1.0 x 100Well 20 <1.0 x 10° 0.3 x 100 0.6 x 100 <1.0 x 10" 8.3 x 100 <1.0 x 100 <1.0 x 100

aTime after wastewater application (days).b Composite of wastewater after applied to RI cells.

TABLE 6. Indigenous enteric virus concentrations

PFU/literSampling Observation well

day Effluentcomposite 15 10 11 19 20

1 420 10 38 12.52 60 0 5 20 1.33 230 0 30 0 1.3 54 255 25 10 0 0 7.35 245 0 5 7.66 15 7.8 7 157 445 3.8

11 0 3.221 0 0

Avg 276 8.3 i10.6 i8.3 i7.5 i4.2

Time after Tracer Virus Application(days)

FIG. 5. Tracer virus (f2) in observation well 15.Tracer virus f2 was sampled in observation well 15directly beneath the wastewater application zone witha Kemmerer water sampler. Samples were taken ev-

ery 8 h for the first 7 test days and once each on days10,13, and 21.

samples was 276 PFU/liter. The virus isolationsfrom well samples were sporadic, although attimes concentrations were as high as 10% of thecomposite average. Values observed for well no.

15 were thought to be lower than anticipatedbecause of difficulties encountered in eliminat-ing a precipitate from the sample prior to assay.

Fecal Streptococcus movement. Compositewastewater samples and samples from observa-tion wells no. 10, 11, 15, 19, and 20 were evalu-ated. Figure 6 illustrates that the fecal Strepto-coccus concentrations declined as the wastewa-ter rested on the RI cells. This reduction couldbe due to die-off or to settling. It is interestingto note that Streptococcus reached observationwell no. 15 in 48 h despite their considerablylarger size. The peak concentration in well no.

15 was reached at day 5 and declined constantly

iL

5.;0

9

168I

0C

616 SCHAUB AND SORBER APPL. ENVIRON. MICROBIOL.

MICROBIAL TREATMENT BY RAPID INFILTRATION 617

-

.10l

E

0 5 10 15 20

T, 'e After Wastewoter AppIccohon to cellsldoys)

FIG. 6. Fecal Streptococcus concentrations duringtracer virus test. Fecal Streptococcus levels were de-termined in the ponded primary wastewater effluentand in observation wells by membrane filter tech-niques. The bars are read on the right margin. Thesymbols A and () are read on the left margin.

thereafter. Samples from the other observationwells in line with the groundwater movementindicated some sporatic "spikes."

DISCUSSIONDirect laboratory comparison of enterovirus

and f2 bacteriophage adsorption to renovatedand unrenovated RI cell soil layers indicatedthat viruses suspended in primary effluentwere poorly held by all soil layers. Bacterio-phage adsorption was minimal. When con-tained in deionized water with metal cation toenhance adsorption, the soils from the reno-

vated cells adsorbed significant numbers of vi-rus except in the high organic content surfacelayer. The soils from the unrenovated cell werenot able to adsorb significant numbers of virusin the upper subsurface layers, indicating somedepletion of adsorptive sites. Previous experi-mentation has indicated that at low cation andhigh soluble organic concentrations and moder-ate pH (such as would be found in wastewater),clays, soils, and other inorganic minerals do notadsorb virions extensively (2, 3, 12). Mutuallyexclusive electronegative surface charges com-

pete with short-range attractive forces on thevirions and solids. Additionally, small-molecu-lar-weight organics effectively compete with vi-rus for soil adsorption sites. It is apparent that,

even in the absence of sewage, the adsorptivesites in the upper soil layers of unrenovatedcells were depleted. Studies by Moore and co-workers (10) would indicate that the f2 bacte-riophage could not be expected to adsorb toeither soils or organic solids in the upper soillayers at the pH used in the present study. Intheir studies they observed that, in almost allinstances, poliovirus was more readily ad-sorbed to clays and sewage solids than was f2bacteriophage.

Field studies utilizing tracer f2 bacterio-phage and indigenous enteroviruses demon-strated that viruses move past the upper soillayers as was predicted by the laboratory ad-sorption studies. The field studies demon-strated that, in addition to poor adsorption,other soil phenomena such as filtration orstraining were not a factor, mainly because ofthe grain size of the sandy, silty, gravelly soilsin relation to the extremely small virus parti-cles. Even the finer surface mat provided byfiltration and deposition of sewage-suspendedsolids had little capability to retain virions.Thus, virus penetration and groundwatermound formation at observation well no. 15occurred in the same time frame (approxi-mately 48 h after wastewater application wasinitiated). Importantly, the f2 bacteriophagestabilized in the groundwater beneath the cellsat almost 50% of the applied virus concentra-tion. Indigenous enteric viruses were also ob-served in the wastewater, but their relative re-covery in observation well no. 15 was sporadicand was approximately 10% ofthe virus concen-tration in the wastewater. Part of this lowerrecovery may be due to the greater soil adsorp-tion of enterovirus as observed with poliovirusin laboratory studies. If indeed this was theprincipal operating factor in the field, then theobserved results could have been an-ticipated. However, the procedural difficulty ineliminating flocs which had formed in the virusconcentrates after frozen storage may have re-sulted in a blocking of virus attachment andinfection of cell monolayers during assay.At times enteric viruses and f2 bacteriophage

were observed in relatively high numbers atobservation wells located in the general path ofthe groundwater movement. However, it wouldappear that at this site there is some heteroge-neity in the nature of the aquifer. The irregu-larities may be important influences on thegroundwater flow patterns resulting in differ-ential flow conditions. Thus, viruses dispersedin the wastewater would be subjected to theconditions and would occur in sporadic highconcentrations over considerable distances. Thefarther the water travels, the greater the ef-

VOL. 33, 1977

618 SCHAUB AND SORBER

fects of differential flow on virus migration and,therefore, the greater the divergence of ob-served virus levels. It is also possible that, overthe years, channelization has occurred in theaquifer. Such conditions would sporadicallyshort circuit the flow in the aquifer until a newequilibrium in the groundwater system isachieved. These probable phenomena need ver-ification through more detailed aquifer charac-terization by the selective installation of addi-tional groundwater monitoring points.

It is important to realize that, in the RI sys-tem studied, viruses are capable of horizontalmigration in the groundwater to distancesgreater than 600 feet (183 m) from their source.Also, the results emphasize the importance offrequent or continuous composite sampling toascertain the potential for virus contaminationof the groundwater. The fact that peak indige-nous enteric virus concentrations observed inthe perimeter wells did not have the same tem-poral relationships as did the f2 bacteriophagemay be due to the interaction of groundwatercontaminated by previous wastewater applica-tions to other RI cells.The bacteriological indicator organisms be-

haved somewhat differently than did the vi-ruses at the RI site. The soil surface retained avery large concentration of total coliform, fecalcoliform, and fecal Streptococcus organismsafter the wastewater had completely infiltratedthe soil surface. Once past the surface zone, thebacterial concentration dropped significantlybut tended to decline at a slower rate at succes-sive sampling depths. Unlike the viruses, thebacteria were filtered or strained at the soilsurface, most likely because of their muchlarger relative size. It is doubtful that adsorp-tion played an important role in retaining thebacteria in the soil. Fecal Streptococcus deter-minations revealed that the number of viableorganisms declined in the standing wastewateron the RI cells. Also, it was observed that therewas a reduction of fecal Streptococcus on thesoils with drying, although it was not deter-mined if this was due to die-off or the continuedmigration of the organisms through the soil.The fecal Streptococcus organisms were de-

tected in significant numbers in the ground-water mound under observation well no. 15 inthe same time frame as were bhe viruses. How-ever, unlike the tracer virus, previous applica-tions to the site could have resulted in themaintenance of residual organisms in thegroundwater. Organism regrowth is a furtherpossibility. Consequently, an accurate assess-ment of their movement is difficult. On occa-sion, fecal Streptococcus organisms were ob-

served in the observation wells in line withthe groundwater flow. Their occurrences didnot correspond with tracer bacteriophage occur-rence in the wells; however, this could be ex-plained by different migration times and alsothe fact the indigenous microorganims like theenteric viruses are always in the wastewaterand their presence may result from previouswastewater applications to other RI cells.The results of this study can be compared

with a tritium tracer study conducted at theFort Devens rapid infiltration site by Stewart(14). The tritium tracer test utilized 75 Ci oftritium applied to the same RI cells as wereused in the study reported herein. Tritium wasapplied over a 4-day period. The investigatorexperienced various problems which resulted inunexpectedly high tritium dilution on the beds.The tritium was detected mainly in observationwells no. 10, 11, 19, and 20 and, to a lesserextent, in wells no. 5, 6, 7, 8. The tritiumshowed up in the same wells as did the bacterio-phage used as a tracer in this study. It wasinteresting to note that in the perimeter wellsthe tritium tracer was detected in sporadic"spikes" much like the bacteriophage, enterovi-rus, and fecal Streptococcus organisms were inthis study. Stewart calculated that a higherthan anticipated dilution existed in the obser-vation wells. The greatest tritium recovery re-vealed a dilution of at least 75-fold which indi-cated that either the wells are not completelyin the main stream of groundwater flow or thatthe main pulse had not arrived within 10 daysof application. The latter possibility wouldseem unlikely in light of the observed infiltra-tion rates and the relatively low affinity ofthese soils for tritium. The present study sup-ports his observation that the hydraulic conduc-tivities of the soil layers in the unsaturatedzone beneath the beds were highly variable. Itis possible that in some wells at the site thestratified layers within the depth of the wellscreens would conduct water at various rates.Also, the water table may have exceeded theelevation of the well screen and therefore stag-nant water in the pipe above the well screenmay have been sampled.

Wellings et al. (15) demonstrated circum-stances similar to those in the present study.They investigated enterovirus movement in acypress dome after the spray irrigation of apackaged treatment plant effluent. They ob-served sporadic "spikes" of natural enteric vi-ruses in groundwater beneath the applicationarea which may be similar in nature to the"spikes" noted in the present study. They sug-gested that the virus migration observed in

APPL. ENVIRON. MICROBIOL.

MICROBIAL TREATMENT BY RAPID INFILTRATION 619

their study may have been the result ofchangesin the soil/water ratio (such as by rainfall),causing selective virus elution from the soils.Significant movement of poliovirus containedin distilled water and treated wastewater wasobserved in sandy soil cores by Duboise et al.(5). Simulated rainfall enhanced elution of vi-rus from soil particles although the movementwas transient with each water application andreassociation was observed at greater soildepths. High soil pH (9.0) also enhanced virusmigration through the soil columns.

Presently, studies are being conducted on op-erational spray irrigation and overland runoffland wastewater application sites to determinethe capabilities of various soil classes andwastewater applications to remove enteric andseeded tracer viruses during treatment.

LITERATURE CITED

1. American Public Health Association. 1971. Standardmethods for the examination of water and wastewa-ter, 13th ed. American Public Health Association,Inc., New York.

2. Carlson, G. F., F. E. Woodward, D. F. Wentworth, and0. J. Sproul. 1968. Virus inactivation on clay parti-cles in natural water. J. Water Pollut. Control Fed.40:R89-R106.

3. Drewry, W. A., and R. L. Eliassen. 1968. Virus move-ment in soil. J. Water Pollut. Control Fed. 40:R257-R271.

4. Driver, C. H., B. F. Hrutfiord, D. E. Spyridakis, E. B.Welch, and D. D. Wooldridge. 1972. Assessment ofthe effectiveness and effects of land disposal metho-dologies of wastewater management. Wastewatermanagement report 72-1, U.S. Army Corps of Engi-neers, Washington, D.C.

5. Duboise, S. M., B. E. Moore, and B. P. Sagik. 1976.Poliovirus survival and movement in a sandy forest

soil. Appl. Environ. Microbiol. 31:536-543.6. Gerba, C. P., C. Wallis, and J. L. Melnick. 1975. Fate of

wastewater bacteria and viruses in soil. J. Irrig.Drain. Div. Am. Soc. Civ. Eng. Proc. 101:IR3:157-174.

7. Kruse, C. W., K Kawata, V. P. Olivieri, and K. E.Longley. 1973. Improvement in terminal disinfectionof sewage effluents. Water Sewage Works 120:57-64.

8. Loeb, T., and N. D. Sinder. 1963. A bacteriophage con-taining RNA. Proc. Natl. Acad. Sci. U.S.A. 47:284-289.

9. McGauhey, P. H., and R. B. Krone. 1967. Soil mantleas a wastewater treatment system. Sanitary Engi-neering Research Laboratory report no. 67-11, Uni-versity of California, Berkeley.

10. Moore, B. E., B. P. Sagik, and J. F. Malina, Jr. 1975.Viral association with suspended solids. Water Res.9:197-203.

11. Schaub, S. A., E. P. Meier, J. R. Kolmer, and C. A.Sorber. 1975. Land application of wastewater thefate of viruses, bacteria and heavy metals at a rapidinfiltration site. Report no. TR 7504, AD A011263,U.S. Army Medical Bioengineering Research and De-velopment Laboratory, Ft. Detrick, Frederick, Md.

12. Schaub, S. A., C. A. Sorber, and G. W. Taylor. 1974.The association of enteric viruses with natural tur-bidity in the aquatic enviromment, p. 7143 In J. F.Malina and B. P. Sagik (ed.), Virus survival in waterand wastewater systems. Center for Research in Wa-ter Resources, University of Texas, Austin.

13. Sorber, C. A., S. A. Schaub, and K. J. Guter. 1972.Problem definition study: evaluation of health andhygiene aspects of land disposal of wastewater atmilitary installations. Report no. 73-02, AD 752122,U.S. Army Medical Environmental Engineering Re-search Unit, Edgewood Arsenal, Md.

14. Stewart, G. L. 1974. Infiltration and tritium tracer testsconducted at the Fort Devens wastewater treatmentplant. Report to U.S. Army Corps of Engineers, NewEngland Division, Waltham, Mass.

15. Wellings, F. M., A. L. Lewis, C. W. Mountain, and L.V. Pierce. 1975. Demonstration of virus in groundwa-ter after effluent discharge onto soil. Appl. Microbiol.29:751-757.

VOL. 33, 1977