United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S1-85/017 Sept. 1985
Project Summary
Virus Removal During
Conventional Drinking
Water Treatment
Charles P. Gerba, Joan B. Rose, Gary A. Toranzos, Shri N. Singh, Lee M. Kelley,
Bruce Keswick, and Herbert L. DuPont
The reduction of enteroviruses and
rotaviruses was studied at a full scale
206 mgd water treatment plant involv-
ing chemical flocculation, sand filtra-
tion, and chlorination. Reduction of
enteroviruses and rotaviruses averaged
81% and 93%, respectively, for the
complete treatment process. The great-
est reduction of enteroviruses occurred
during pre-chlorination/flocculation
and filtration, and reduction of rota-
viruses occurred during pre-chlorina-
tion/clarif ication and final eMorinatton.
Enteroviruses or rotaviruses occurred in
24% of the finished water samples
containing levels of free chlorine (>0.2
mg/l), and meeting coliform bacteria
(1/100 ml) and turbidity (1 NTU)
standards. The results of this study
indicate that finished water having
measurable levels of free residual chlo-
rine and meeting standards for coliform
bacteria and turbidity cannot be assum-
ed to be virus free.
This Project Summary was developed
by EPA's Health Effects Research Labo-
ratory. Research Triangle Park, NC, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
Since the turn of the century, it has
been known that there is a significant risk
of contracting infectious disease from the
ingestion of sewage-contaminated water
and food. The spread of enteric bacterial
disease by this route has been well
controlled by the widespread application
of bacterial standards and modern treat-
ment processes for drinking water and
sewage.
The isolation of viruses from treated
drinking water has raised concerns that
water treatment methods may not always
adequately insure the removal of viruses
from water designated for human con-
sumption. Because viruses have been
isolated from drinking water that met
recommended levels of coliform bacteria,
chlorine and turbidity,1'*1' serious ques-
tions have been raised about the adequacy
of bacterial standards in judging the
sanitary quality of water relative to its
potential for transmitting viral disease.
The majority of processes used to treat
sources of potable water are capable of
reducing virus numbers. However, enteric
viruses have been shown to be less
effectively removed or inactivated than
indicator bacteria by many treatment
processes, and much of the available
information on the removal of enteric
viruses by water treatment processes is
based on laboratory studies with mem-
bers of the enterovirus group (particular-
ly, poliovirus type 1) and a few selected
coliphages. Not all enteric viruses are
likely to be removed with equal and high
efficiency by a given water process.
Laboratory studies have shown that polio-
virus is more effectively removed from
water by alum flocculation and from
sewage by adsorption to sludge floes than
rotavirus. Likewise, under field conditions
at an activated sludge sewage treatment
plant, enteroviruses appeared to be more
effectively removed than were rotavi-
-------
ruses. It is also recognized that naturally
occurring viruses may be more resistant
to chlorination than prototype laboratory
strains.
Few studies have been done on the
removal of naturally-occurring enterovi-
ruses by actual drinking water treatment
plants under field conditions. A previous
report3 describing research in which
enteroviruses and rotaviruses were iso-
lated from treated drinking water in a
distribution system and at a water treat-
ment plant initiated a more extensive
investigation at the same site. This study
reports on the occurrence of enteric
viruses within a full-scale water treat-
ment plant after treatment processes
including clarification, filtration, and
chlorination.
Procedure
The site in this study was a 205 mgd
water treatment plant supplied with raw
water from a river and conveyed approx-
imately 17 km by canal via two pumping
stations. The river is fed by a shallow lake
located in a watershed with a human
population of 4 to 8 million, where
untreated wastewater is discharged.
Each plant has a treatment sequence
consisting of chemical addition (liquid
alum and catfloc or superfloc) followed by
hydraulic mixing, flbcculation, clarifica-
tion with pre- and post-chlorination,
filtration through rapid sand filters or
automatic valveless sand filters and final
chlorination. Samples collected included
raw or intake water, water post-clarifica-
tion, water post-filtration, and finished
water from three of the five plants. Four
trips to the plant were made over a two-
year period and 20 to 40 samples were
collected during each trip.
For enteric virus collection and concen-
tration, samples of 9.8 to 756 liters were
collected using the adsorption-elution
method and 1 MDS Virosorb filters [AMF
CUNO, Meriden, CT]. Primary filter elu-
ates were reconcentrated from a volume
of 1L to approximately 36 ml by an organic
flocculation method. All samples for
enterovirus analysis were inoculated
onto monolayers of BGM cells, 2 ml/75
cm2 plastic flask, overlaid with mainte-
nance medium and observed for cyto-
pathic effects (CPE) for a period of 21
days. Those samples positive for CPE
were passed a second time in the BGM
cell line and then plaqued as a final
confirmatory test using an agar overlay
method. Subsequently, virus isolates
were identified by serum neutralization
tests using the Lim-Melnick enterovirus
Antibody Serum Pools. Rotaviruses were
detected by using the immuno-fluores-
cent assay described by Smith and
Gerba.4
One-L grab samples and virus samples
were taken simultaneously. Bacterial
densities, standard plate count bacteria,
total and fecal coliform bacteria, and fecal
streptococci bacteria, were determined
using membrane filtration. Bacteriophage
concentrations were determined in the
grab samples as well as in each 1 MDS
filter eluate, before freezing, using Es-
cherichia coli Hfr (ATCC 15597) as the
host bacterium. Physical-chemical meas-
urements of the water included total and
free chlorine levels, pH and turbidity.
Results and Discussion
The biological and physico-chemical
characteristics of the water after the
treatment steps were examined over a
two-year period. Four sampling periods
were chosen, three during the dry season
(3/82, 1/83 and 4/83) when the plant
was operating more optimally and one
during the rainy season (7/82) when the
plant could not produce water meeting
minimum coliform and turbidity stand-
ards.
The mean values for turbidities, bac-
teria and viruses were calculated for each
sampling trip for the dry season only. In
general, the total plate count bacteria
were found to be the most numerous,
ranging from 9/ml in the finished water
to 465/ml in the raw water, followed by
the fecal streptococci, (20 to 4453/100
ml), enterococci (6 to 1165/100 ml), total
coliforms (15 to 216/100 ml) and fecal
coliforms (1 to 66/100 ml). Coliphage
concentrations ranged from 1/L in the
finished water to 1769/L in the raw
water.
Table 1 summarizes the data on enteric
virus isolation from all the samples during
the dry season. Viruses were recovered in
68% of the raw water samples and after
clarification and filtration; 50% and 37%
of the samples were positive for viruses,
respectively. Viruses were detected in
20% of the finished water samples. It was
found that eight (24%) of the finished
water samples that contained free chlo-
rine levels of greater than 0.2 mg/L and
met turbidity and bacterial standards
were also positive for enteric viruses.
Average percent reductions for all
water quality parameters studied were
determined after each step in the water
treatment processes (Table 2). The major
reduction in bacteria and coliphage, (90%
to 99%) came after pre-chlorination/
clarification with slight decreases durinj
the other treatment steps. The initia
reduction of microorganisms during th<
treatment process may be due to the
addition of chlorine, although adsorptior
to the floe during clarification undoubtedly
is capable of removing large numbers o
organisms. The greatest reduction ir
turbidity occurred during the clarification
step and averaged 87% after complete
treatment.
The greatest removal of enteroviruses
appeared to occur during pre-chlorina-
tion/clarif ication and filtration steps. The
major reduction of rotaviruses occurred
after pre-chlorination/clarification and
post-chlorination of the finished water.
Of all the groups of microorganisms
studied, enteric viruses were the least
effectively removed.
From this study and others1'2'3, it
appears that enteric viruses can occur at
detectable levels in finished drinking
water meeting current coliform, turbidity,
standards and containing levels of free
chlorine (0.2 mg/L). The isolation of
human enteric viruses in these waters is
not a condemnation of conventional
drinking water treatment but an indication
that water quality parameters currently
accepted to ensure the production of
microbially safe water do not necessarily
ensure the absence of enteric viruses.
Although the plant was able to produce
water with a turbidity of 1 NTU and less
than one total coliform per ml, major
operational and design deficiencies were
apparent. The most serious problem was
in the filters where cracking, mudballs,
and sand boils were observed, indicating
shortcircuiting. The plant may have been
operating near or above its design capac-
ity (hydraulically overloaded)' but the
extent of the problem could not be deter-
mined due to the lack of metering. The
chlorine contact time could not be deter-
mined; however, operators reported it as
less than 30 minutes at peak flows
dropping as low as 15 minutes. This could
explain the occurrence of enteric viruses
in the finished water. Still it is significant
that these deficiencies were not always
reflected by the quality of the finished
water during the dry season. It may be
that, in marginally operated water treat-
ment plants or plants with a heavily
contaminated raw water source, meas-
ureable levels of enteric viruses are able
to penetrate the treatment process, even
when current water quality standards are
achieved.
-------
Table 1. Summary of Enteric Virus Isolation from All Samples" After Successive Drinking Water Treatment Processes
Site Raw Clarified Filtered
Finished
Samples Meeting''
Standards
Number of
Samples Collected 19 14 8 54
Number Number Number Number
Virus Positive % Positive % Positive % Positive %
Enterovirus 9 47 4 29 0 059
Rotavirus 9 47 5 36 3 37 7 13
Both Entero &
Ftotaviruses0 5 26 2 10 0 0 1 2
Total Enteric
Viruses 13 68 7 50 3 37 11 20
33
Number
Positive %
4 12
5 15
1 3
8 24
" = All samples collected during the dry season only.
" = Number of finished water samples positive for virus which had free chlorine residual and met U. S. turbidity, and bacterial standards.
c = Number of samples positive for both Entero- and Rotaviruses.
Table 2. Average Percent Reductions of Turbidity, Bacteria and Viruses After Successive Drinking Water Treatments
From flaw
to Site Date
Prechlorinated/
Clarified 3/82
1/83
Grand Mean
Filtered 1/83
Finished 3/82
1/83
4/83
Grand Mean
Collec-
tion
1
3
3
1
3
4
Turb.
NTU
11
0*
4
73
81
76
90
87
Total
Plate
Count
47.5
89.1
78.9
98.9
91.7
98.6
98.6
98.1
Total
Col/-
forms
100.0
94.3
98.9
98.5
100
98.6
98.9
99.3
Bad
Fecal
Con-
forms
99.5
100
97.1
100
100
WO
99.3
98.5
'ena
Fecal
Strep
99.5
92.6
97.2
99.7
99.7
99.5
99.7
99.5
Entero-
cocci
NO*
97.3
97.3
99.8
NO
99.5
NO
99.5
Coli-
phage
(Direct)
28.5
97.0
90.4
99.8
100
99.9
99.9
99.9
Coli-
phage
(Concen-
trate)
99.0
99.91
99.7
99.9
99.8
99.9
99.8
99.9
Entero-
virus
CPE
25
78
55
100
45
92
100
81
Rota-
virus
95
0
61
0
99
0
93
93
* = No removal.
" = Not determined.
c - Mean reductions determined using mean counts weighted by number of samples per trip.
Conclusions
1. Both enteroviruses and rotaviruses
could be isolated from finished
drinking water containing chlorine
levels of X>.2 mg/L and meeting
coliform bacteria (1/100 ml) and
turbidity (1 NTU) standards.
2. Enterovirus and rotavirus removal
averaged 81 % and 93%, respective-
ly, for the complete treatment pro-
cess involving prechlorination/
flocculation, sand filtration, and final
chlorination.
3. The complete water treatment pro-
cess was more effective in the
removal of turbidity, coliform, fecal
coliform, fecal streptococci, stand-
ard plate count bacteria, and coli-
phage than enteric viruses.
Pilot plant studies as well as field
studies should be conducted in
order to determine the occurrence
and significance of enteric viruses
after water treatment processes
and to evaluate coliphages as an
indicator system for animal viruses
as well as other water quality
parameters.
Recommendations
1. Studies should be conducted to
determine the efficiency of human
rotavirus removal by drinking water
treatment processes.
2. Better methods should be developed
for the detection and concentration
of coliphage from water.
3. Longer term studies should be
conducted on the occurrence of
enteroviruses and rotaviruses in
treated drinking water taking ad-
vantage of the continual develop-
ment of detection methods. This
-------
would allow the development of a
larger data base which could be
used to determine if stronger corre-
lations exist between the presence
of coliphage, total plate count bac-
teria and enteric viruses.
4. Research should be conducted on
water quality and operational treat-
ment conditions that will assure
essentially complete removal of
viruses.
References
1. Melnick, J. L. and Gerba, C. P.
(1982). Viruses in surface and
drinking waters. Environ. Intl. 7:3-
7.
2. Payment, P. and Trudel, M. (1984).
Detection and health risk associated
with low level virus concentration
in drinking water. Water Sci. and
Techn. (In press).
3. Deetz.T. R. era/. (1984). Occurrence
of rota- and enteroviruses in drink-
ing and environmental water in a
developing nation. Water Res.
18:567-571.
4. Smith, E. M. and Gerba, C. P. (1982).
Development of a method for the
detection of human rotavirus in
water. Appl. Environ. Microbiol.
43:1440-1450.
Charles P. Gerba, Joan B. Rose. Gary A. Toranzos, Shri N. Singh, and Lee M. Kelle
are with the University of Arizona. Tuscon, AZ 85721; Bruce Keswick and
Herbert L. DuPont are with the University of Texas Sciences Center, Houstoi
TX 77025.
Elmer W. Akin and John C. Hoff are the EPA Project Officers (see below).
The complete report, entitled "Virus Removal During Conventional Drinking
Water Treatment," {Order No. PB 85-227 510/AS; Cost: $10.00, subject to
change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officers can be contacted at:
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S1-85/017
0000329 PS
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO IL 60604
------- |