In this study, we determined the removal of a prototype human norovirus (Norwalk virus, NV) by bench-scale alum coagulation, flocculation and sedimentation processes using reverse transcriptase polymerase chain reaction (RT-PCR) for norovirus assays. After determining optimum conditions for the coagulation, flocculation and sedimentation processes in terms of turbidity reduction, jar tests were performed using the same waters seeded with test viruses. For comparison, two other important health-related viruses, poliovirus 1 (PV1) and coliphage MS2, were included in this study. The removal of NV by coagulation, flocculation and sedimentation processes based on RT-PCR assay in this study was 1.5 log10, which was similar to that of PV1 and a little lower than that of coliphage MS2 (2 log10) based on the same RT-PCR assay. The removal of NV in this study (1.5 log10) is considerably higher than the one in a recent study using recombinant norovirus virus-like particles (∼0.7 log10). Overall, the results of this study suggest that human noroviruses can be appreciably reduced by a properly-operated coagulation, flocculation and sedimentation processes and the contamination of drinking water by noroviruses should be controlled by conventional water treatment processes with conventional physico-chemical processes and disinfection.

INTRODUCTION

The basic physico-chemical processes in conventional water treatment consist of chemical coagulation, flocculation, sedimentation, and filtration. These processes were originally designed to remove particles from water (Amirtharajah & O'Melia 1990). However, these processes are also important barriers to waterborne pathogens, because all waterborne pathogens are themselves particles and some – especially viruses and bacteria – are usually associated with other solid particles (Amirtharajah & O'Melia 1990). In fact, previous studies have reported appreciable removals of some human enteric viruses and bacteriophages by these physico-chemical water treatment processes (Robeck et al. 1962; Sproul 1980; Rao et al. 1988; Sobsey et al. 1995; Nasser et al. 1995). However, the ability of these processes to remove noroviruses, an important group of viral agents of waterborne gastroenteritis, has not been reported.

Noroviruses are major agents of both endemic and epidemic gastroenteritis in the world and water is one of the most important routes of their transmission (Kapikian et al. 1996). In fact, there are several reports that noroviruses were associated with drinking water outbreaks (Kukkula et al. 1999; Nygard et al. 2003; Maunula et al. 2005). Despite their importance in both endemic and epidemic gastroenteritis and their association with drinking water outbreaks, little is known about the effectiveness of physico-chemical water treatment processes against these important waterborne viruses. This is probably because: (1) the availability of human noroviruses is quite limited due to the lack of animal or mammalian cell culture propagation systems for these viruses; and (2) the large amount of microorganism is necessary to conduct even bench-scale experiments to determine the effectiveness of physico-chemical water treatment processes against waterborne pathogens.

To circumvent those obstacles mentioned above, there was a recent study using recombinant norovirus virus-like particles (VLPs) that estimated the effectiveness of physico-chemical water treatment processes against noroviruses (Shirasaki et al. 2010). However, it is difficult to extrapolate the results of that study to true human noroviruses because there are significant differences between recombinant norovirus VLPs and actual human noroviruses in terms of their size, density, surface charges and other physical–chemical properties. For example, due to the lack of RNA in recombinant norovirus VLPs, there is a significant difference between the recombinant norovirus VLPs and true noroviruses in terms of their density – 1.28 g/cm3 vs. 1.38 g/cm3 for recombinant norovirus VLPs and Norwalk virus (NV, a prototype human norovirus), respectively. Therefore, it is possible that the denser and intact human NV could be more easily removed by coagulation, flocculation and sedimentation processes than the lighter and structurally incomplete recombinant norovirus VLPs. This and other differences between recombinant norovirus VLPs and true human noroviruses could make the behavior of the two quite different in physico-chemical water treatment processes. Therefore, in this study, we determined the removal of a prototype norovirus (NV) by bench-scale alum coagulation, flocculation and sedimentation processes using quantitative reverse transcriptase polymerase chain reaction (RT-PCR) for virus assays. For comparison, two other important and widely used health-related model viruses, poliovirus 1 (PV1) and coliphage MS2, were included in this study.

MATERIALS AND METHODS

Viruses and preparation

The prototype human NV strain 8FIIa (genogroup I) was obtained as stool samples from infected human volunteers and was stored at −80 °C until use. The stool samples were suspended in phosphate buffered saline (PBS) (10% weight/volume [w/v]) and NV was extracted by homogenizing in an equal volume of chloroform. The supernatant was recovered following low speed (5,000 g) centrifugation for 15 minutes at 4 °C. PV1 and F-specific coliphage MS2 were propagated as previously described (Shin & Sobsey 2008). Briefly, PV1 strain LSc and coliphage MS2 (ATCC# 15597-B1) were acquired from available stocks of the Environmental Microbiology Laboratory at the University of North Carolina at Chapel Hill. PV1 was propagated in the Buffalo Green Monkey Kidney (BGMK) cells and coliphage MS2 was grown in Escherichia coli C3000 (ATCC# 15597) by the double agar layer technique (Adams 1959). Stock suspensions of PV1 and coliphage MS2 were prepared by homogenizing cell lysates in an equal volume of chloroform followed by low speed (5,000 g) centrifugation at 4 °C (for 15 or 30 minutes for PV1 and coliphage MS2, respectively). All virus extracts (NV, PV1, and coliphage MS2) were further purified by centrifugal ultrafiltration (Centricon 100 ultraconcentrators, Amicon, Beverly, MA) to remove salts and small organic molecules in the suspension, and resuspended to its original volume in PBS.

Test waters

Raw water samples were obtained from the Orange Water and Sewer Authority drinking water treatment plant of Carrboro, NC. These waters had pH 6.6–7.0, turbidity 1.5–2.6 NTU, alkalinity 21–24 mg/L as CaCO3, and hardness 20 mg/L as CaCO3.

Jar tests

Preliminary jar tests on raw waters were conducted prior to and for each experiment with a jar test apparatus (Phipps and Bird, Inc., Richmond, VA) to optimize turbidity reduction with respect to coagulant dose, pH, flocculation time and flocculation speed. In both experiments, the coagulant dose that achieved the maximum reduction of turbidity was 40 mg/L, which resulted in the pH of the treated water between 5.5–6.0. The pH of water is one of the major factors affecting the removal of viruses in coagulation process and pH values between 5 and 6.5 are generally recommended for coagulation with metallic coagulants (USEPA 1980; Harrington et al. 2001; Bell et al. 2002). In this study, the pH of the water was changed to the recommended ranges (5–6.5) simply by adding an alum stock solution (without adding either NaOH or HCl stock solution), which was possible due to the alkalinity level of the raw water (see supplementary material for an example calculation (pH change of water after addition of alum stock solutions), available online at http://www.iwaponline.com/ws/015/100.pdf). Turbidity was measured with a nephelometer (Monitek, Model 21). The instrument was calibrated using the turbidity standards prepared according to Standard Methods (APHA 1985). In addition to turbidity, particles were measured with a particle counter (Met One, Water Grab Sampler). After determining optimum conditions for coagulation–flocculation–sedimentation, jar tests were performed using the same waters seeded with viruses. Briefly, 500 mL of virus-seeded water was placed in a 600 mL beaker, and a 35 mL sample was removed for viral and other physical analyses before adding coagulant. The experiment was started by injecting a 5 mL alum stock solution quickly with a syringe near the stirrer while stirring the water at 100 rpm (coagulation). After about 30 s, stirring was slowed to 25–30 rpm and continued for 20 min to allow coagulated particles to form flocs (flocculation). Finally, the floc was allowed to settle without stirring for 30 min (sedimentation) and a 35 mL sample was taken from 1 inch (2.54 cm) below the air/water interface for viral and other physical analyses. As mentioned above in the Introduction section, due to the limited availability of NV and the large amount of viruses needed for these experiments, only two experiments were performed.

PCR primers

The oligonucleotide primers used in this study for RT-PCR amplification for NV, PV1, and coliphage MS2 were previously described (De Leon et al. 1992; Shin & Sobsey 2008) and are listed in Table 1 along with their amplicon sizes.

Table 1

Oligonucleotide primers for RT-PCR amplication of virusesa

Viruses Primersb Amplicon size Position 
NV (5′ primer) 5′-CAAATTATGACAGAATCCTTC-3′ 260 4,601–4,621 
 (3′ primer) 5′-GAGAAATATGACATGGATTGC-3′  4,840–4,860 
PV1 (5′ primer) 5′-CCTCCGGCCCCTGAATG-3′ 197 449–465 
 (3′ primer) 5′-ACCGGATGGCCAATCCAA-3′  627–645 
MS2 (5′ primer) 5′-GCAACCTCCTCTCTGGCTAC-3′ 233 2,118–2,137 
 (3′ primer) 5′-CCCTACAACGAGCCTAAATTC-3′  2,310–2,330 
Viruses Primersb Amplicon size Position 
NV (5′ primer) 5′-CAAATTATGACAGAATCCTTC-3′ 260 4,601–4,621 
 (3′ primer) 5′-GAGAAATATGACATGGATTGC-3′  4,840–4,860 
PV1 (5′ primer) 5′-CCTCCGGCCCCTGAATG-3′ 197 449–465 
 (3′ primer) 5′-ACCGGATGGCCAATCCAA-3′  627–645 
MS2 (5′ primer) 5′-GCAACCTCCTCTCTGGCTAC-3′ 233 2,118–2,137 
 (3′ primer) 5′-CCCTACAACGAGCCTAAATTC-3′  2,310–2,330 

aThe primers used in RT step: 3′ primers for NV and PV1 and random hexamers for coliphage MS2.

bReferences for viruses: NV (De Leon et al. 1992), PV1 (De Leon et al. 1992), MS2 (Shin & Sobsey 2008).

Virus assays

Infectivity assays

PV1 was assayed by the plaque technique on confluent layers of BGMK cell cultures grown in 60 mm diameter Petri dishes as previously described (Sobsey et al. 1988) and coliphage MS2 was assayed by the double agar layer plaque technique on host E. coli C3000 as previously described (Adams 1959).

RT-PCR assays

All the viruses (NV, PV1, and coliphage MS2) were assayed by RT-PCR using the Perkin-Elmer-Cetus RNA Core kit (Perkin Elmer-Roche, Alameda, CA) as previously described (De Leon et al. 1992). Briefly, viral RNAs were released by direct heat release at 95 ⁰C and RT was performed at 42 °C for 60 minutes by Moloney leukaemia virus (MoLV) reverse transcriptase with specific 3′ primers for NV and PV1 (Table 1) or random hexamers for coliphage MS2. PCR was performed by Taq polymerase with additional specific 5′ primers for NV and PV1 and both 5′ and 3′ primers for coliphage MS2 (Table 1). A total of 40 cycles of PCR was carried out using the following thermal profile: denaturation for 1.5 min at 95 °C, annealing for 1.5 min at 55 °C, and extension for 1.5 min at 72 °C. A 15 μl portion of RT-PCR product was analyzed by 2% agarose gel electrophoresis, the electrophoresed gel was stained with ethidium bromide, and resolved PCR products were visualized by UV light using a transilluminator. The virus titer based on RT-PCR assays (PCR unit (PCRU)/mL) was obtained by an end-point dilution method (dilution to extinction of RT-PCR-amplifiable RNA) and titer was determined by the last dilution that gives a positive signal (or its corresponding dilution factor) and the sample volume used.

RESULTS AND DISCUSSION

Table 2 shows the removals of the test viruses by alum coagulation, flocculation and sedimentation processes under optimized treatment conditions. The removals of test viruses based on infectivity assays and RT-PCR assays varied among test viruses. The removal of PV1 based on RT-PCR assay (1.5 log10) corresponded relatively well to that based on infectivity assay (1.9 log10). Conversely, the removal of coliphage MS2 based on RT-PCR assays (2 log10) was somewhat lower than that based on infectivity assay (3 log10). Meanwhile, the removal of NV by the alum coagulation, flocculation and sedimentation processes based on RT-PCR assay was 1.5 log10, which was the same as that of PV1 and a little lower than that of MS2 (2 log10) based on the same kind of RT-PCR assay.

Table 2

Removals of NV, PV1, and MS2 by alum coagulation, flocculation and sedimentation processes (40 mg/L alum, pH 5.5–6.0, and room temperature)

Virus reduction (−log10 value) 
Virus Infectivity RT-PCR 
 Exp. 1 Exp. 2 Mean Exp. 1 Exp. 2 Mean 
NVa – – – 1.5 
Poliovirus 1 1.9 1.9 1.9 1.5 
MS2 NDb 3.0 3.0 ND 
Virus reduction (−log10 value) 
Virus Infectivity RT-PCR 
 Exp. 1 Exp. 2 Mean Exp. 1 Exp. 2 Mean 
NVa – – – 1.5 
Poliovirus 1 1.9 1.9 1.9 1.5 
MS2 NDb 3.0 3.0 ND 

aInitial virus concentrations.

Exp. 1: 1.1 × 105 PFU/mL (PV1), and 106 PCRU/mL (NV).

Exp. 2: 9.7 × 105 PFU/mL (PV1), 1.9 × 106 PFU/mL (MS2), and 105 PCRU/mL (NV).

bND: not determined.

Table 3 shows the reductions of turbidity and particle counts by the alum coagulation, flocculation and sedimentation processes with the same water and under the same treatment conditions as the virus removal experiments. The reduction of turbidity (0.75 log10) was lower than the removals of the test viruses (1.5–3 log10) under the same treatment condition regardless of virus assay. Conversely, the reductions in particle counts (1.65 log10) corresponded relatively well to the virus removals based on RT-PCR assays (1.5–2 log10), but were somewhat lower than the virus removal based on infectivity assays (1.9–3 log10) under the same treatment condition.

Table 3

Reductions of turbidity and particle counts by alum coagulation, flocculation and sedimentation process (40 mg/L alum, pH 5.5–6.0, and room temperature)a

Reduction (−log10 value) 
Turbidity Particles (2–200 μm) 
Exp. 1 Exp. 2 Mean Exp. 1 Exp. 2 Mean 
0.5 1.0 0.75 1.7 1.6 1.65 
Reduction (−log10 value) 
Turbidity Particles (2–200 μm) 
Exp. 1 Exp. 2 Mean Exp. 1 Exp. 2 Mean 
0.5 1.0 0.75 1.7 1.6 1.65 

aInitial and final values of turbidity and particle counts.

Exp. 1: turbidity (from 2.5 NTU to 0.7 NTU: 72% reduction), particle counts (from 7,692/mL to 160/mL).

Exp. 2: turbidity (from 5.8 NTU to 0.6 NTU: 90% reduction), particle counts (from 10,634/mL to 256/mL).

The removals of PV1 and coliphage MS2 by the alum coagulation, flocculation and sedimentation processes based on infectivity assays were ∼2 log10 for PV1 and 3 log10 for coliphage MS2. The substantial removals of PV1 and coliphage MS2 by alum coagulation, flocculation and sedimentation processes were similar to those reported in previous studies (Thorup et al. 1970; Rao et al. 1988; Sobsey et al. 1995). Conversely, the removals of PV1 and MS2 based on RT-PCR assays (1.5 log10 and 2 log10, respectively) were somewhat lower than those based on infectivity assay (1.9 log10, and 3.0 log10, respectively) in this study. This discrepancy is possibly due to underestimation in virus titers by plaque assay. That is, coagulation process causes virus aggregation and association with particles, which could result in apparent reduced infectivity titers if those virus aggregates produced only single plaques. Such underestimation of apparent virus infectivity would result in the overestimation of virus removals by coagulation, flocculation and sedimentation processes based on results by infectivity assay. Conversely, the RT-PCR assay is less sensitive to this virus aggregation and particle association phenomena because it determines the total copy numbers of viral nucleic acids in a sample. Therefore, it seems that RT-PCR assays are possibly a more conservative measure of virus removals by coagulation–flocculation–sedimentation processes than plaque-forming infectivity assays.

The results of this study shows that human NV appears to be removed by coagulation, flocculation and sedimentation processes to a similar extent as the other viruses tested. The removal of human NV by coagulation, flocculation and sedimentation in this study (1.5 log10) based on RT-PCR assay was the same as that of PV1 (1.5 log10) and a little lower than that of MS2 (2 log10) based on the same RT-PCR assay. Meanwhile, the removal of NV by coagulation, flocculation and sedimentation in this study (1.5 log10) was similar to the reduction of particles (1.65 log10), but somewhat higher than the reduction of turbidity (0.75 log10). Traditionally, turbidity reduction has been recommended as a process control (treatment indicator) for virus removal by coagulation process (USEPA 1980). In fact, many previous studies have shown that maximum virus removal occur at or near the maximum turbidity reduction (Chaudhuri & Engelbrecht 1970; York & Drewry 1974; Rao et al. 1988). Although the reduction of turbidity by coagulation process in this study was relatively low (0.75 log10) possibly due to the low initial turbidity (2.5–5.8 NTU), there was a significant removal of the test viruses (1.5–3 log10) under the same treatment condition. Meanwhile, the reduction of particles (1.65 log10) by coagulation process in this study was similar to the removal of test viruses (1.5–3 log10) under the same treatment condition. As mentioned in the Introduction section, all waterborne pathogens are themselves particles and some – especially viruses and bacteria – are usually associated with other solid particles in water. Therefore, it is quite possible that there is a good correlation between the reduction of particles and the removal of viruses in coagulation process. Overall, it appears that particle count could be a suitable surrogate and turbidity could be a conservative surrogate for predicting the removal of human norovirus by coagulation, flocculation and sedimentation processes.

Meanwhile, the removal of NV by coagulation, flocculation and sedimentation processes in this study (1.5 log10) is greater than that of recombinant norovirus VLPs (∼0.7 log10) by the same process, as reported in a recent study (Shirasaki et al. 2010). There are several possible explanations for this difference in human NV and norovirus VLP reductions by coagulation, flocculation and sedimentation processes. First, the sizes of the recombinant norovirus VLPs and human NV are somewhat different: 35.7 nm vs. 27 nm for the recombinant norovirus VLPs and human NV, respectively. Second, due to the lack of RNA in recombinant norovirus VLPs, there is a significant difference between recombinant norovirus VLPs and human NV in terms of their buoyant density: 1.28 g/cm3 vs. 1.38 g/cm3 for the recombinant norovirus VLPs and human NV, respectively. Third, there might be some difference in terms of surface charges between human norovirus and the recombinant norovirus VLPs. That is, there are three types of proteins in human norovirus virion (VP1, VP2, and VPg), but the recombinant norovirus VLPs is made of only one type of protein (VP1) (Green 2007). Although the proportion of VP2 and VPg in human norovirus virion is small, it is still possible that the lack of VP2 and VPg makes the surface charge of the recombinant norovirus VLPs different from that of human norovirus. Overall, it is possible the denser and smaller human NV could be more easily removed by coagulation, flocculation and sedimentation processes than the lighter and larger recombinant norovirus VLPs. It seems that the physical–chemical difference between the recombinant norovirus VLPs and human NV make the behavior of the two quite different in coagulation, flocculation and sedimentation processes.

Overall, the results of this study suggest that human noroviruses could be appreciably reduced by properly-operated coagulation, flocculation and sedimentation processes although additional studies on the reduction of different genogroups of human noroviruses are recommended. Nonetheless, considering the high infectivity (low infectious doses) of human noroviruses (Ward & Kin 1984), even very low levels of such noroviruses in drinking water can still pose considerable risks to the public who consume such drinking water. Hence, it is necessary to provide additional treatment in the form of filtration and disinfection to further remove and/or inactivate those viruses escaping the conventional physico-chemical treatment processes of alum coagulation, flocculation and sedimentation in order to assure the virological quality of drinking water.

CONCLUSION

In this study, the ability of the widely used coagulant, alum (aluminum sulfate, Al2(SO4)18H2O), to remove a prototype norovirus (NV) was investigated in bench-scale experiments using a standard jar test apparatus. The results of our study indicated the following:

  • NV, the prototype human genogroup 1 norovirus can be appreciably reduced by properly-operated alum coagulation, flocculation and sedimentation processes.

  • It appears that particle count could be a suitable surrogate and turbidity could be a conservative surrogate for predicting the removal of human norovirus by coagulation, flocculation and sedimentation processes.

ACKNOWLEDGEMENT

This research was funded by the National Water Research Institute (HRA 699-512-92 and WQI 699-527-95).

REFERENCES

REFERENCES
Adams
M.
1959
Bacteriophages
.
Interscience Publishers Inc
.,
New York
.
American Public Health Association
1985
Standard Methods for the Examination of Water and Wastewater
.
17th edn
,
American Public Health Association
,
Washington, DC
.
Amirtharajah
A.
O'Melia
C. R.
1990
Coagulation Process: Destabilization, Mixing, and Flocculation
. In: Water Quality and Treatment.
4th edn
,
McGraw-Hill, Inc
.,
USA
.
Bell
K.
LeChavallier
M. W.
Abbazadegan
M.
Amy
G. L.
Sinha
S.
Benjamin
M.
Ibrahim
E. A.
2002
Enhanced and Optimized Coagulation for Particulates and Microbial Removal
.
AWWA Research Foundation
,
Denver, CO
.
Chaudhuri
M.
Engelbrecht
R. S.
1970
Removal of viruses from water by chemical coagulation and flocculation
.
Journal of the American Water Works Association
62
(
9
),
563
567
.
De Leon
R.
Matsui
S. M.
Baric
R. S.
Herrmann
J. E.
Blackow
N. R.
Greenberg
H. B.
Sobsey
M. D.
1992
Detection of Norwalk virus in stool specimens by reverse transcriptase-polymerase chain reaction and nonradioactive oligoprobes
.
Journal of Clinical Microbiology
30
(
12
),
3151
3157
.
Green
K.
2007
Caliciviridae: the norovirus
. In:
Fields Virology
(
Knipe
D. M.
Howley
P. M.
, eds).
5th edn
,
Lippincott-Raven
,
New York
.
Harrington
G. W.
Chen
H. W.
Harris
A. J.
Xagoraraki
I.
Battigelli
D.
Standridge
J. H.
2001
Removal of Emerging Waterborne Pathogens
.
AWWA Research Foundation
,
Denver, CO
.
Kapikian
A. Z.
Estes
M. K.
Chanock
R. M.
1996
Norwalk group of viruses
. In:
Fields Virology
(
Fields
B. N.
Knipe
D. M.
Howley
P. M.
, eds).
3rd edn
,
Lippincott-Raven
,
New York
.
Kukkula
M.
Maunula
L.
Silvennoimen
E.
von Bonsdorff
C. H.
1999
Outbreak of viral gastroenteritis due to drinking water contaminated by Norwalk-like viruses
.
The Journal of Infectious Disease
180
(December
),
1771
1776
.
Maunula
L.
Miettinen
I. T.
Bonsdorff
C. H.
2005
Norovirus outbreaks from drinking water
.
Emerging Infectious Disease
11
(
11
),
1716
1721
.
Nasser
A.
Weinberg
D.
Dinoor
N.
Fattal
B.
Adin
A.
1995
Removal of hepatitis A virus (HAV), poliovirus and MS2 coliphage by coagulation and high rate filtration
.
Water science and Technology
31
(
5–6
),
63
68
.
Nygard
K.
Torven
M.
Ancker
C.
Knauth
S. B.
Hedlund
K. O.
Giessecke
J.
Andersson
Y.
Svensson
L.
2003
Emerging genotype (GGIIb) of norovirus in drinking water, Sweden
.
Emerging Infectious Disease
9
(
12
),
1548
1552
.
Rao
V. C.
Symons
J. M.
Ling
A.
Wang
P.
Metcalf
T. G.
Hoff
J. C.
Melnick
J. L.
1988
Removal of hepatitis A virus by drinking ware treatment
.
Journal of the American Water Works Association
80
(
2
),
59
67
.
Robeck
G. G.
Clarke
N. A.
Dostal
K. A.
1962
Effectiveness of water treatment processes in virus removal
.
Journal of American Water Works Association
54
(
10
),
1275
1292
.
Shin
G.
Sobsey
M. D.
2008
Inactivation of norovirus by chlorine disinfection of water
.
Water Research
42
(
17
),
4562
4568
.
Sobsey
M. D.
Fuji
T.
Shields
P. A.
1988
Inactivation of hepatitis A virus and model viruses in water by free chlorine and monochloramine
. In:
Proceedings of International Conference for Water and Wastewater Microbiology
,
IAWPRC, Pergamon Press
,
New York
.
Sobsey
M. D.
Battigelli
D.
Handzel
T. R.
Schwab
K. J.
1995
Male-Specific Coliphages as Indicators of Viral Contamination of Drinking Water
.
American Water Works Association
,
Denver
.
Sproul
O. J.
1980
Critical Review of Virus Removal by Coagulation Processes and pH Modification. EPA- 600/2-80-004
.
Thorup
R. T.
Nixon
F. P.
Wentworth
D. F.
Sproul
O. J.
1970
Virus removal by coagulation with polyelectrolytes
.
Journal of American Water Works Association
62
(
2
),
97
101
.
USEPA
1980
Critical Review of virus removal by coagulation processes and pH modifications. EPA-600/2-80-004
.
Ward
R. L.
Kin
E. W.
1984
Minimum infectious dose of animal viruses
.
CRC Critical Review of Environmental Control
14
,
297
310
.
York
D. W.
Drewry
W. A.
1974
Virus removal by chemical coagulation
.
Journal of the American Water Works Association
66
(
12
),
711
716
.