Environmental Protection Technology Series
  DETECTION AND  MOTIVATION  OF
ENTERIC VIRUSES  IN  WASTEWATER


              Environmental Monitoring and Support Laboratory
                      Office of Research and Development
                     U.S. Environmental Protection Agency
                              Cincinnati, Ohio  45268


-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has  been assigned  to the  ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate  instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                                    XX
                                              EPA-600/2-77-095
                                              May 1977
 DETECTION AND INACTIVATION OF ENTERIC VIRUSES

                 IN WASTEWATER
                      by

               Hill el I. Shuval
              Eliyahu Katzenelson
        Environmental Health Laboratory
   Hebrew University-Hadassah Medical School
               Jerusalem, Israel
              Grant No.  S-800990
             (formerly 17060 EAM)
                Project Officer

                  Gerald Berg
           Biological Methods Branch
Environmental Monitoring and Support Laboratory
            Cincinnati, Ohio  45268
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
            CINCINNATI, OHIO  45268

-------
                               DISCLAIMER
This report has been reviewed by the Environmental Monitoring
and Support Laboratory, U.S. Environmental Protection Agency,
and approved for publication.  Approval does not signify that
the contents necessarily reflect the views and policies of the
U.S. Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recom-
mendation for use.
                             11

-------
                                 FOREWORD
     Man and his environment must be protected from the adverse effects of
pesticides, radiation, pathogens, noise, and other forms of pollution,  and
the unwise management of solid waste.  Efforts to protect the environment
require a focus that recognizes the interplay between the components of our
physical environment—air, water, and land.   The Environmental Monitoring
and Support Laboratory-Cincinnati contributes to this multidisciplinary
focus through programs engaged in

       studies on the effects of environmental contaminants on the
       biosphere, and

       a search for ways to prevent contamination and to recycle
       valuable resources.

     The viruses that are discharged into sewage with the fecal wastes of
man, and thereby into the environment, constitute a hazard to all of those
who contact the waters that are insulted by the presence of these viruses.
Methodology that is necessary to detect and quantify small numbers of these
highly infective disease-producing agents in large volumes of water, and
methodology that is necessary to destroy them with ozone are described in
this report.
                                      Dwight G. Ballinger
                                      Director
                                      EMSL-Cincinnati
                                     111

-------
                                  ABSTRACT


This report covers studies on the development and evaluation of methods  for
concentrating and detecting low levels of viruses in large volumes of water.
In addition, it covers studies on the use of ozone in inactivating viruses
in water and wastewater.

Eight recognized virus concentration methods were evaluated under standardized
conditions and their virus recovery efficiency was determined.   Of particular
interest are the methods which proved efficient and also capable of concentra-
ting viruses from water samples of 100 gallons or more.   The most promising
methods in this category are filtration methods using cellulose nitrate  membranes,
aluminum hydroxide and PE-60.  A new method developed and evaluated in the  course
of the study used hollow fiber membranes.  Viruses from  large volumes of water
can be concentrated rapidly by this method without requiring pH adjustment.

A promising method for the rapid detection of viruses in water using the
fluorescent antibody technique was developed.  This method can provide
qualitative results in 6 to 9 hours and a quantitative estimate of virus
concentration in 18 hours.

Although much work remains to be done in developing and  evaluating virus
concentration and detection methods in water, there is good evidence that
practical methods for the virus assay of water and wastewater are an
achievable goal.

In studies on the use of  ozone as a virucidal agent in water and wastewater,
special  techniques for investigating this question were  developed, among
them an accurate spectrophotometric method for detecting very low concentra-
tions of ozone in small  samples (10 ml) of water.   A 0.3 ppm residual  of ozone
was found to inactivate over 99% of seeded poliovirus in clean  water in  less
than 10 seconds  as compared to 100 seconds required to achieve  the same
degree of inactivation by chlorine under equal conditions.

The kinetic curves of virus inactivation indicates a rapid first stage kill
in a matter of seconds followed by a slower kill  lasting minutes, until
complete inactivation is  achieved.   The role of virus clumps in explaining
this phenomenon  is considered.

Although no detectable dose response relationship could  be demonstrated
for ozone contact times greater than 10 seconds,  preliminary studies indicate
that such a relation may  exist for shorter contact times.   Ozone was also

-------
shown to kill  viruses rapidly and effectively  in wastewater effluent.
The virucidal  effect of various  possible ozone species was investigated
and has led to some preliminary  hypotheses  on  this  subject.  Ozone has
been shown to  be a rapid and effective virucidal agent of potentially
great value in field applications in the treatment  of water and wastewater.

This report was submitted in May 1975 in fulfillment of  Research Grant
No. S-800990 (formerly 17060 EAM) under the sponsorship  of the U.S.
Environmental  Protection Agency.   This report  covers a period from October
1969 to January 1975 and work was complete  as  of January 1975.

-------
                                  CONTENTS
                                                                    Fj

Abstract                                                             1V
List of Figures                                                     v1ii
List of Tables                                                       X1
Acknowledgments                                                      X1V
I     Introduction                                                   1
II    Conclusions                                                    27
III   Recommendations                                                36
PART A  Virus Detection Methods
IV    Review of Methods for the Detection of Enteric
      Viruses in the Water Environment                               42
V     Comparison of Eight Concentration Methods for
      Isolation of Viruses in Water                                  69
VI    Virus Concentration using Hollow Fiber Membranes               99
VII   Evaluation of Gauze Pad Method to Recover Viruses
      from Water                                                    117
VIII  A Rapid Fluorescent Antibody Method for Quantitative
      Isolation of Viruses from Water                               144
IX    Virus Types in Israel Sewage                                  166
PART B  Inactivation of Viruses in Water
X     The Chemistry of Ozone as a Disinfectant                      177
XI    Spectrophotometric Method for the Determination
      of Ozone in Aqueous Solutions
XII   Inactivation Kinetics of Viruses and Bacteria in
      Water by Ozone                                                "5
XIII  Disinfection of Viruses in Sewage by Ozone                    254
XIV   Coliphage  Inactivation in Seawater                            269
XV    List of Publications                                          286
                                   vii

-------
                               LIST  OF  FIGURES
 Number
 1    Detail  of Fiber Unit and System  Flow  Sheet                       101
 2   Result  of a Typical  Virus Concentration Experiment               102
 3   Scanning Electron Micrographs  of the  Cross-Section
     of the  Hollow Fiber  Membrane                                    106
 4   Concentration of Virus  in Feed Versus Volume  of
     Feed Gauze Sampler                                              109
 5   Gauze Sampler                                                   119
 6   Sampler in Operation                                            120
 7   Device  for Extracting Liquid for Gauze Pad                       123
 8   BGM Cells Infected with Poliovirus  Type I,
     Stained with Fluorescent Antibodies;  9 Hours
     after Infection                                                 150
 9   BGM Cells Infected with Poliovirus  Type I,
     Stained with Fluorescent Antibodies;  18 Hours
     after Infection                                                 150
10   The % of Poliovirus  in  Daily Grab Sampling  vs.
     Daily Gauze Pad in Sewage by Neutralization Test                 171
11    Inactivation of Poliovirus  I by  Ozone after 10
     Sec of  Various pH values (T=5°C)                                 190
12   Calibration Curve for Determination of Ozone  in Water            204

13   Stability of Color Intensity as  Function of
     Potassium Iodide Concentration                                  208

14   Ozone Aqueous Solution  Stability Versus Time,
     Temperature and Concentration                                    217
15   Influence of Stirring Rate on  Ozone Stability (at 5°C)           218
16   Kinetics of Ozone Demand of Different Virus Suspensions
     at 5°C  and stirring  rate 80 rpm                                  220

17   Inactivation Kinetics of Poliovirus I by 0.3  ppm
     Ozone at 5°C                                                    232
18   Inactivation Kinetics of Poliovirus I by 0.8  ppm
     Ozone at 5°C                                                    233
                                   vm

-------
Number                                                               Page


19   Inactivation Kinetics of Poliovirus  I  by  1.5  ppm
     Ozone at 5°C                                                    234

20   Virus Survival  Rate with Varying  Ozone Concentrations
     at 0.5 sec and  2.5 sec                                          235

21   Poliovirus I Survival after 40"  of Exposure
     to Various Concentrations of Ozone                              237

22   Inactivation Kinetics of Poliovirus  I  (118-b)
     Before and After Ultrasonication  by  o.l  ppm Ozone               239

23   Inactivation Kinetics of Coliphate "^2  by  Various
     Concentrations  of Ozone at 1°C                                  242

24   Inactivation Kinetics of E.  Coli  by  Various
     Concentrations  of Ozone at 1°C                                  243

25   Inactivation Kinetics of Poliovirus  I, added  in two
     Portions, at 0  Time and after 56", by  1.5 ppm Ozone              245

26   Correlation between Ozone Concentration and Redox
     Potential Values                                                246

27   Inactivation Kinetics of Poliovirus  I  by  o.4  mg/1  initial  0.,
     Concentration in the Presence of  5%  and 10% Filtered Sewage      258

28   Inactivation Kinetics of Poliovirus  I  by  0.8  mg/1  initial  03
     Concentration in the Presence of  5%  and 10% Filtered Sewage      259

29   Inactivation Kinetics of Poliovirus  I  by  1.3  mg/1  initial  0-
     Concentration in the Presence of  5%  and 10% Filtered Sewage      260

30   Inactivation Kinetics of Poliovirus  I  by  1.8  mg/1  initial  03
     Concentration in the Presence of  5%  and 10% Filtered Sewage      261

31   Kinetics of Increase in 03 Residual  Concentrations and  of
     Poliovirus I Inactivation During  Continuous 03 Bubbling
     Through Buffer  and Filtered Sewage                              263

32   Percent Survival of Poliovirus I  after Reaction with
     Different Initial 03 Concentrations  in 5% and 10%
     Filtered Sewage (Batch Experiments)                              266

33   Inactivation of Coliphage T£ in  Normal and Autoclaved
     Sea Water                                                       272

34   Inactivation of Coliphage T£ in  Sea  Water Containing
     Different Concentrations of Nutrient Broth                      273
35   The Effect of Nutrient Broth on  the  Anticol iphage
     Activity of a Mixture of Marine  Bacteria                         275
                                   IX

-------
Number


36   Anticoliphage Activity  in  Different  Isolates  of
     Marine Bacteria                                                 277
37   Different Rates of  Coliphage  Inactivation  by  an
     Isolate of Marine Bacteria                                       278
38   The Effect of Nutrient  Borth  on  the  Activation Capacity
     of an Isolated Marine Bacteria                                   279

39   The Inactivation of Coliphage in Preincubated Sea Water          280

-------
                                 TABLES


 Number                                                                page

 1     The  Recovery  Efficiency of  Four Methods for the
      Concentration of Seeded Attenuated Poliovirus I
      at Low  Input  Levels  in Tap  Water                                 78

 2     The  Recovery  Efficiency of  Five Methods for the
      Concentration of Seeded Attenuated Poliovirus I
      at High  Input Levels  in Tap Water                                79

 3     The  Recovery  Efficiency of  Four Methods for the
      Concentration of Seeded Attenuated Poliovirus I
      at Various  Input Levels in  Tap Water                             80

 4     The  Recovery  Efficiency of  Five Methods for the
      Concentration of Seeded Echovirus 7 at Various
      Input Levels  in Tap  Water                                        81

 5     Recovery Efficiency  of Four Methods for the
      Concentration of Enteroviruses from Sewage-
      Contaminated  Water                                               84

 6     Recovery Efficiency  of Four Methods for the
      Concentration of Enteroviruses from Sewage-
      Contaminated  Water                                               85

 7     Recovery Efficiency  of the  Gelman Cartridge Filter
      for  the Concentration of Seeded Attenuated Poliovirus  I
      at Various  Input Levels in  Tap Water                             86

 8     The  Average Recovery of Seeded Enterovirus and the
      Concentration Factor of Five  Methods with Tap Water              89

 9     The  Average Recovery of Seeded Enterovirus and the
      Concentration Factor of Four  Methods with Tap Water              92

10     Results of  Backwash  Experiments                                 103

11     Parametric  Results from Curve Fitting                           110
12     Results of  Virus Concentration Experiments                      113

13     A Comparison  of  Enterovirus Concentration by Two
      Sampling Methods:  Grab  (2  Liters) and  Pad  (24 Hours)           125
                                     xi

-------
Number
Page
14    The Average Percentage of  Enterovirus Detected on
      Gauze Pads Held  in  Sewage  for  1  Day Compared with
      the Percentage on Gauze  Pads Held for 3, 4, and 7 Days          126
15    Comparison Between  the Number  of Viruses Detected from
      3 Gauze Pads Immersed  in Parallel In Raw Sewage                 128
16    The Effect of Successive Gauze Pad Elutions on
      Poliovirus I Elution (Tap  Water-4 Elutions)                     130

17    The Effect of Successive Gauze Pad Elutions on
      Poliovirus I Elution (Tap  Water-5 Elutions)                     131
18    The Effect of Successive Gauze Pad Elutions on
      Enterovirus Recovery (Raw  Sewage)                               132

19    The Effect of Calf  Serum on Elution of  Poliovirus I
      From Gauze Pads  (Tap Water)                                    134

20    The Recovery of  the Gauze  Pad  Sample for Concentration
      of Poliovirus I  (Tap Water)                                    135

21    The Effect of Sample Volume on Poliovirus  I  Recovery
      (Tap Water)                                       J            137

22    Titers of Poliovirus I Obtained with the Fluorescent
      Antibody Method  after 9  and 20 Hour  Incubation,  and
      with the Plaque Count Method on Plates                          152

23    Titers of Poliovirus Type I Obtained with  the Fluorescent
      Antibody Technique and with the Plaque  Count Method             153

24    Comparison of the Fluorescent Antibody  Technique with
      the  Plaque Count Method for Quantitative Evaluation
      of Viruses  in 5  Liters of Water                                155
25    Comparison of the Fluorescent Antibody Technique with
      the  Plaque Count Method for the Quantitative Evaluation
      of Viruses  in 40 Liters of Water                               155

26    Virus  Types  Picked from Sewage  at Different Places
      and  Communities  in  Israel                                       169
27     Polio  and  Non-Poliovirus  in Grab and Gauze Pad Samples
       in Kiriat  Shemoneh Sewage by  Neutralization Test                170

28     Summary  of  the  Kinetics of Ozone Decomposition in Water         183

-------
Number

29    Influence of Potassium Iodide  Concentration on
      Intensity of Color                                              207
30    Reproducibility of Results  in  the  Spectrophotometric
      Method                                                          210
31    Comparison of Titremetric and  Spectrophotometric
      Methods for Determination of Ozone  in Water                      213
                                xiii

-------
                              ACKNOWLEDGEMENTS

This report represents a concerted effort by a research team over a five
year period.  Many of the staff members served during the whole period
while others participated in specific projects over shorter time periods.
It is not possible to list here the names of all those who contributed
each in his own way to the success of this project.

The scientists who played major roles in these studies are listed as
authors of  the sections as follows:

     Section I:      Mil lei I. Shuval
     Section IV:     Eliyahu Katzenelson & H.I. Shuval
     Section V:      Badri Fattal, E. Katzenelson, T. Hostovsky & H.I. Shuval
     Section VI:     Georges Bel fort, Y. Rotem-Borensztajn & E. Katzenelson
     Section VII:    Badri Fattal & E. Katzenelson
     Section VIII:   Eliyahu Katzenelson
     Section IX:     Badri Fattal
     Section X:      Mordechai Peleg
     Section XI:     Hannah Schechter
     Section XII:    Eliyahu Katzenelson, B. Kletter & H.I. Shuval
     Section XIII:   Eliyahu Katzenelson & N. Biedermann
     Section XIV:    Binyamin Kletter, E. Katzenelson & H.I. Shuval

 In  addition mention  should be made of  the contributions by M.  Green,  R.  Kalbo,
 J.  Sabag,  M. Hod,  T. Goldblum, M. Nevo and Dr.  M.  Nishmi.  The  active coop-
 eration  of the Staff of  the Ministry of  Health  and the Mekorot Water  Company
 in  carrying out  the  field  sampling programs  is  sincerely  appreciated.
                                      xiv

-------
                               SECTION I
                              INTRODUCTION

ENTEROVIRUSES  AND  WATER  QUALITY
The ultimate amount of water  annually available to a country is more
or less fixed  in volume.  Its rate of development is dependent on the
availability of financial resources and engineering technology.  Into
this fairly constant source of supply is spewed an ever increasina
amount of organic and inorganic chemical waste together with patho-
genic bacteria and viruses dispensed into the environment by steadily
growing populations massing in burgeoning urban centers.  This
expanding populace requires more and more water for urban and indus-
trial  use, but its very own wastes continually contaminate the
limited available supply, despoiling its quality.   This process
reaches, at times, a point where the water's utility for human
consumption is at risk.   Thus a built-in paradox of modern society
revolves around a situation where more and more water is needed by
the growing population but less and less becomes available at the
required quality as a result of the self-destructive process  of
pollution.   In this section we shall attempt to evaluate the impact
of the increasing burden of enteroviruses on water quality and
see what should be done  about it.

-------
THE GROWING DEMAND FOR WATER
Man's health and well-being are dependent in many ways both on the
quantity and the quality of water available to meet his varying needs.
For many years it has been accepted that the quality of the water
he drinks or uses for cooking or other domestic purposes may have a
direct effect on his health.  Today we are becoming no less concerned
with the burden of pathogenic bacteria and viruses in water used for
recreation, agricultural irrigation, or growing shellfish.   We are
also asking whether pathogens sprayed into the air during wastewater
treatment processes or land disposal systems may be carried as
aerosols to infect people in the vicinity.

As the population of the world grows at an average rate of  2 percent
per year, doubling every 30 years, more and more water must be
extracted from rivers, lakes and from the ground to supply  the
additional  people.   The extent of this increasing demand for water
can be understood in historical  terms in light of the statement
that the number of people alive today in the world is equal to the
total number of persons who ever lived on the face of the earth
since the beginning of mankind.   Each person uses directly  or in-.
directly from 50 to 500 liters of water per day and returns most of
that water to the drainage systems of the earth together with a
residual load of inorganic and organic pollutants as well as with a
burden of pathogenic microorganisms that are shed together  with the
human body wastes.  These once-used waters, together with their

-------
animate and inanimate residuals, return to the water sources  supplying
the essential  fluid of life to those living downstream.

THE LIMITS OF NATURAL SELF-PURIFICATION
In theory there is enough water in the world to dilute all  of the
domestic wastewater produced by man in a single year by a factor
greater than one million.  Even the most advanced efforts of  science
and technology could hardly expect to achieve such a high degree of
residual reduction in any controlled waste treatment process.

However, such a hypothetical calculation is highly misleading
since it assumes a continuous and uniform distribution of all
human wastes into the total volume of water of the world.  In fact,
97 percent of the world's water is in the oceans and is generally
not available as a direct repository or diluent of the wastes of
human population centers.  Of the remaining amount of sweet water,
three-fourths is locked up in the earth's polar ice caps and
glaciers.  It has been estimated that only about 0.01 percent of
the total water of the globe is actually in our rivers.

The self-purifying powers of the aqueous ecosystem have been
depended upon for many generations to rid our streams and rivers
of the pollutional load through such processes as dilution,
sedimentation, predation and biological decomposition.  However,
this capacity has its limits and can become severely overstressed
as the insult of pollution ever increases.

-------
The problem is even more complex than it might appear due to the
social forces which are at work and which are changing the
demographic distribution of people on the face of the nlobe.  One
hundred years ago 80 percent and more of the population was thinly
settled in rural areas while today more and more countries are
approaching a situation where 80 percent of the population is
crowded together in cities, most of them strung one after the
other along the lengths of the rivers of the world.

Today with some 317 billion people in the world, about 30 percent are
crowded together in urban areas with central sewage systems dis-
posing of their wastes mainly to the rivers.  By the year 2000 we
anticipate the total population of the world to double and during
the same period the urban population is expected to reach about 60
percent of the total.  That means an absolute increase of 400 percent
in the urban  population with a parallel increase in the pollution
burden of domestic wastes, including pathogenic bacteria and viruses
in our rivers and other bodies of water serving as repositories of
urban wastes.

We are already  faced with a situation where some of our rivers are
now so loaded with  such vast amounts of wastes of all kinds,
including human body wastes laden with the  full spectrum of entero-
pathogens, that by  the time they reach the  sea almost all, or all
of the flow has been pumped out for municipal or industrial use  at

-------
least once and returned to the river.    This means that many of our
sources of so-called fresh water are in effect partially diluted
wastewater which has undergone varying degrees of treatment.

THE PERSISTANCE OF VIRUSES IN THE WATER ENVIRONMENT
The scientific and technical  literature is replete with discussions
of refractory chemicals not removed by treatment processes and non-
biodegradable substances unaffected by the processes of natural
self-purification of rivers.   To this  list must be added a group of
enteroviruses which are amazingly resistant to environmental factors
and resemble in many ways the behavior of the so-called refractory
chemical wastes which persist so lonq.

Many laboratory and field studies have indicated that most enteroviruses
can persist for days and many even for months in the natural water
environment with viabilities longest in heavily contaminated water
                    2
during cold weather.   Under most river conditions, where downstream
consumers are rarely removed from upstream sources of fecal contami-
nation by more than a day or so of flow time, it must be accepted that
the enterovirus burden that reaches our water arteries is carried
along essentially undisturbed.  We have detected enteroviruses 25 km
downstream of a single isolated source of sewage flow into the Jordan
River,  and others have provided similar evidence of the resistant
nature of the virus burden in the water environment and their
persistence in rivers, lakes and in the sea even at points relatively
far removed from the point of their introduction.

-------
                                                 4
In 1962 an eminent panel of public health experts  concisely summarized
their views of the enterovirus burden on water quality as follows:
     More than 70 viruses have been detected in human feces.  All
     may be present in sewage.  Viruses pass through sewage
     treatment plants, persist in contaminated waters, and may
     penetrate the water treatment plants.  Numerous outbreaks of
     infectious hepatitis have been traced to contaminated drinking
     water.  The occurrence of such incidents appears to be
     increasing.

Three years later another group of scientists concluded that "the
capabilities of present water pollution control techonology are clearly
inadequate as far as viruses are concerned."

TEN YEARS OF  RESEARCH  FINDINGS
In 1965 when Gerald Berg and his colleagues initiated the first
international gathering of scientists to be devoted exclusively to
the problem of viruses in water under the heading,"Symposium on the
Transmission of Viruses by the Water Route," they asked: What is
the relationship among viral diseases, water supplies and water
pollution; what answers are needed to protect the public health from
the transmission of viral diseases; what answers do we have; and
what research is needed to obtain additional answers?

Ten years have passed since that first symposium and the handful of
active pioneer  researchers in the field of  viruses in water has grown

-------
considerably and much progress has been made in helping to answer
some of the questions that were then unanswered.

Virus monitoring techniques, although still  not perfected, have been
developed during this period to assay as much as  100 gallon samples
of water, which was then but a far-off goal.  Utilizing various
monitoring methods, surveys of virus concentrations in the natural
water environment have been conducted by researchers in many countries
and all continents.  We have a clearer picture today of the rate of
enteroviruses shed by communities into the water environment.   Raw
wastewater often carries enterovirus loads of 100-1000  pfu  per
       3
100 ml.   Rivers which serve as receiving waters  for raw sewage or
treated effluent often carry between 1-10 pfu    of enterovirus per
100 ml.  Other studies have detected the presences of varying
concentrations of  viruses in lakes and sea water used for recreational
purposes; in wells used to supply drinking water; in irrigation water
drawn from polluted rivers; in sea water in areas of shellfish
culture; and in the air in the vicinity of a wastewater spray
irrigation project.

In fact, it appears that when well-trained scientists use any one of
a number of new sensitive virus monitoring methods which can sample
large enough volumes of water, they are able to detect enteroviruses
in most phases of the natural water environment which is today
almost universally exposed to sources of fecal contamination.

-------
Here  it must be added that so far, enteroviruses have been detected
only  rarely in drinking water supplies which have undergone conventional
treatment.  Of course, the chances of detecting viruses in finished water
supplies have been slight, since for all intents and purposes there
has been to date no program for routine virus monitoring of such water
supplies except for the city of Paris, which has used the gauze pad
method shown to have a relatively low efficiency.

During these past ten years many excellent studies have advanced our
understanding of the virus removal  efficiency of conventional as well
as non-conventional water and wastewater treatment processes.

Here  again, without going into details, it appears that although many
advanced wastewater treatment processes such as lime precipitation,
ozonization and others can remove a very high percentage of the viruses
in wastewater, most normally designed and operated conventional bio-
logical wastewater treatment plants are capable of removing no more
than  80-90 percent of the viruses and even when chlorination of effluent
is practiced, it would be difficult to expect more than a 98.5
percent removal  under most conditions.

The virus removal  efficiency of most conventional  water treatment
plants is also not consistent although  here too laboratory and pilot
plant studies have demonstrated that much better removal  efficiences
can be obtained,  particularly if break  point chlorination or ozoni-
zation is practiced.
                                   8

-------
Studies indicate that where conventional  high rate sand filtration
plants followed by chlorination are treating heavily contaminated river
water containing high concentrations of ammonia, it is  unrealistic  to
rely on more than a 99.9 percent removal  of enteroviruses  from the
stream.

THE LAG IN WATER TREATMENT PRACTICES
There is no doubt today that the statement made 10 years ago
that "the capabilities of water pollutional control technology is
clearly inadequate as far as viruses are  concerned," is no longer
correct in light of new research findings and demonstration units.
However, it remains essentially correct as far as actual practice in
the field is concerned since most new plants designed and  constructed
during these past ten years have not taken into account upgraded
standards of treatment that would be required to meet the  virus
problems which were so clearly defined in 1965.

Some have said that the risk of transmission of viruses by water,
although being theoretically possible, is of no practical  importance
as a public health problem, but let us look briefly at  the record.

TRANSMISSION OF VIRUS DISEASE BY WATER

Craun and McCabe  reported that in the United States there was a
gradual increase from 1940 to 1970 in the number of cases  per year
of infectious hepatitis occurring in waterborne outbreaks  while a

-------
reverse trend is noted for  typhoid fever cases.   The Center for
Disease Control  in Atlanta has reported that during the two year period
of 1971-1972 there were 11 infectious hepatitis waterborne outbreaks
involving 266 persons.  These infectious hepatitis outbreaks
represent the most prevalent type of waterborne outbreaks among
those where the etiological  agent was specified.   It is true that
these reported waterborne infectious hepatitis cases represent only
about 0.3 percent of the total reported number of cases of infectious
hepatitis.  But one wonders  how many secondary infections or even
contact spread epidemics of the disease might have been initiated by
a single clinical or subclinical  case of infectious hepatitis intro-
duced into a community as a  result of ingesting a minimal infectious
dose of the agent carried by the water supply.  The massive outbreaks
of infectious hepatitis transmitted by shellfish grown in sewage con-
taminated coastal waters underscore the potential public health risks
associated with the growing  burden of viruses in the water environment.
Are we prepared to ignore this evidence even if it is not as
definitive as one might prefer?

The conclusions of Plotkin presented at the 1965 Symposium  that
"... one infective dose of tissue culture is sufficient to infect
                                                         g
man.  . ." still  remains to this day the basis for Berg's  conclusion
that "any amount of virus in drinking or recreational water that is
detectable in appropriate cell cultures constitutes a hazard to
those drinking the water."  Nothing that has been reported in the
literature since that time seems to contest that clear-cut conclusion.
                                  10

-------
THE CAUSE OF THE LAG
Why then has so little been done to translate the findings and
conclusions of ten years ago, which have been further strengthened
and reinforced by subsequent research, into public health policy
and engineering reality?  The knowhow is at hand and the technology
has been tried and tested.

It seems that the first cause of hesitancy in translating the recommend-
ations of the 1965 symposium into water pollution control policy was
the lack of a clear-cut epidemiological basis for establishing a
definitive virological standard for drinking and recreational water.
The second and possibly more central reason was the absence of an
acceptable standard procedure for assaying relatively large volumes
of water for viruses.  The availability of such a tried and proven
method is the sine qua non for establishing a virus standard for
water.  Let us see if these two reasons still hold today.

As to the first, the  lack of a clear-cut epidemiological basis, let
us look back and examine how the original microbial standard for
drinking water was established in its day.

The fact that water can serve as a highly effective vector of
enteric disease agents was reported by John Snow as early as 1854, even
before there was a full understanding of the nature of the causative
                                  11

-------
pathogenic microorganisms themselves.  Methods of treating water
by sand filtration to remove the causative agents were also practiced
prior to the time that the efficiency of such processes could be
controlled by evaluating the degree of removal of enteric micro-
organisms.  However, as time went on, it became the practice to
check the effectiveness of treating polluted water by bacterial
testing.

In 1900 the slow sand filter at Lawrence, Mass., treating the highly
polluted water of the Merrimac River, had been in operation for
several years and its installation had been followed by a drop in
the typhoid fever death rate in the city.  Based on their slow sand
filter performance and epidemiological insight they were able to set
one of the first microbial standards for drinking water which in today's
                                                            Q
terms would be equivalent to 69 colifcim bactera per 100 ml.   With
the introduction of water disinfection technology by chlorine, it was
possible in 1910 for the U.S; Public Health Service to establish its
first statutory standard for drinking water at 2 coliform bacteria
per 100 ml. of water.  In both these cases the standard set was
essentially a function of engineering feasibility based on the
assumption that the less enteric organisms ingested in drinking water*
the safer it would be.

Since that time, health authorities in many countries throughout the
world have set similar numerical standards for the microbial quality
                                   12

-------
of drinking water as one of the first administrative aids in the
monitoring and control of the safety of water supplies.   There can
be little debate today as to the wisdom and foresightedness of those
who established that early environmental quality standard which has
made a major contribution in the prevention and control  of waterborne
enteric disease and protecting the health of hundreds of millions of
people.

The bacterial standard for drinking water was established and put into
practice long before there was a full scientific evaulation of such
factors as:  the minimal infectious dose; the ratio of pathogens to
coliforms; the dose-response relationship; the reliability of coliforms
as an  indicator under various conditions in the natural  environment;
and the relative resistance of coliforms and pathogens in different
water  treatment processes.

In presenting this brief historical background of early microbial
water  standards, it is not to be implied that since the almost
universally accepted coliform drinking water standard was established
in its day, mainly on engineering feasibility criteria,  rather than
epidemiological and scientific data, that we should automatically
do likewise today.

There  is nevertheless something to be said for the basic logic and
direct approach of the early public health pioneers.  They had good
                                  13

-------
reason to believe that by ingesting pathogenic bacteria, carried by
water, people could and did become infected and they therefore strived
within the limits of the technology available to reduce the potential
risk to a reasonable minimum.   They then set standards based on
engineering feasibility so that others would have goals to meet
which had been proved to be obtainable.  These standards also enabled
them to monitor and control water treatment plants once they had been
built and put into operation so as to assure their continued hiqh level
of performance.  In the case of coliform bacteria in drinking water
where treatment costs were relatively small and benefits demon-
stratively large, the decision to set a microbial standard based on
the lowest level obtainable under reasonable engineering conditions
may in light of today's knowledge and experience by considered
obvious and simple enough.  However, 70 years ago there were many
who opposed spending the money needed to build the water treatment
plants required to meet those early standards.  What may be obvious
today was not so obvious to those who held the purse strings at the
time.

Enough has been said and written elsewhere as to the inadequacies of
the coliform standard in water as far as viruses are concerned.  Most
are in agreement today  that the absence of coliforms in water—particularly
water originating from  a highly polluted source—is not under all
circumstances  an assurance that viruses are not  present in low
concentrations.
                                    14

-------
ANALYTICAL APPROACH
Lacking a clear-cut epidemiological  basis for establishing  a  virus
standard for water, let us attempt to arrive at one analytically.   Let
us assume that one tissue culture infectious dose may indeed  lead  to
                                             i%
infection in a susceptible man and that one per hundred such  persons
exposed to such a dose will, in fact, become sick.

Let us also assume that the raw sewage stream contains 1000 pfu per
100 ml as had been demonstrated by a number of researchers  in various
parts of the world.  Such a raw sewage stream might be treated by
conventional methods leading to a 90 percent reduction of viruses  and
then diluted in a river by a factor of 100.  We could expect  a further
reduction in virus concentration by about 90 percent due to sedi-
mentation and natural die away in the stream so that at a downstream
water treatment plant intake the virus concentration might  now be
1  pfu    per 1000 ml.  If the water treatment plant is conventionally
designed and operated, it should be capable of providing a  further
reduction in virus concentration of 99.9 percent.  That would mean
there still might be something like 1  pfu   per 1000 liters  in the
community water supply.  Or in other words, about one person  per
thousand might ingest a tissue culture infectious dose per  day.  If
our assumption is correct that only 1 percent of those ingesting such
a  virus dose would actually become sick, this might mean one  person
per hundred thousand.  Such a low level of infection might  never be
detectable in small communities and even in very large cities might
                                 15

-------
not lead to the conclusion that the disease was waterborne, since most
cases would appear as sporadic clinical cases of various virus diseases.

When one or a number of the virus inactivating barriers of such a
system is inoperative and the virus concentration in the cormiunity
water supply is increased by several orders of magnitude, there is the
possibility that more easily detectable epidemic situations might
develop, particularly if a virulent virus is being shed into the
water from an upstream source.  All of this is based on many assumptions
but it is indicative of the type of situation that may exist today with
current enterovirus burdens in our water environment and current
water treatment technology.
THE NEED TO ESTABLISH VIRUS STANDARDS FOR WATER
It would not be unduly restrictive to draw the same conclusion as
the early public health pioneers reached when they established the
first coliform standard for drinking water based on engineering
feasibility.  It certainly should be technically feasible  today
to produce drinking, water containing no more than 1 pfu per
1000 liters.  Actually we know we can do better than that  and we
should insist on it wherever possible.

However, here we come to the second constraint--the limits of the
virus monitoring methods.  Despite  the fact  that we have not yet
selected a single standard virus monitoring  technique,  a number of
                                   16

-------
candidate methods  capable of assaying at least 100 gallons  of water
have been developed and are reviewed in this report.   Without too much
additional effort it should be possible today to reach a tentative
agreement on the use of one or more of the most reliable of  the methods
for routine water monitoring of 100 gallon samples.

The next step would be to set tentative virus standards for  water.
A number of proposals have been made.   The World Health Organization
has reconmended   that drinking water should not have any viruses in
a 10 liter sample.  Melnick   has suggested that no more than 1 pfu
                                      12
per 100 gallons be allowed, while Berg   has stated the premise that
the permissible level of viruses in water should be none.

Since it is obvious that any practical virus standard set today has
to relate to both engineering feasibility and the capability of monitor-
ing techniques to detect viruses in water, it would not be possible to
set a standard calling for no viruses in water.

It is, however, feasible today to call for the establishment of a
standard that we  know can be met by well operated conventional plants
and that we can verify by current monitoring techniques, even if we
know that we should be able to do better and may well want to require
a more rigorous standard later on as the technology develops, or for
special cases such as wastewater reuse.
                                   17

-------
A standard which requires that no viruses be present in a 100 gallon
sample of drinking water is a practical  and attainable goal  which we
are capable of establishing and putting  into effect right now.
There are many reasons why such a standard might not be good enough,
but we would make a lot of progress in upgrading water quality throughout
the world if this modest standard was achieved universally.   Far too
many urban water supplies do not meet it today!   To postpone setting
a standard today in order to wait until  we can establish a more
refined one later will only result in a  further continuation of the
current lax attitude towards controlling viruses in water.  A wise
man once stated that the worst enemy of  the "good"  is the "best!"
VIRUS STANDARD FOR RECREATIONAL WATER
It might also be desirable to consider a virus standard for recreational
water, although the epidemiological evidence is  much less convincing.
Studies have shown that bathers do actually ingest  from 10-50 ml per
bathing period and could swallow viruses present in the water.
Melnick   has proposed that 1  pfu   per 10 gallons be set as a
standard for recreational water, while Cookson   has proposed a similar
standard for effluent to be disposed into rivers used for recreational
purposes.  There is a particularly strong logic for a virus  standard
for seawater used for bathing since studies have shown that coliforms
disappear exceedingly rapidly in the sea as compared to enteroviruses.
We have shown that in the Mediterranean  Coast of Israel, coliforms
have a Tgo of one hour or less while poliovirus  has a TQQ of one to
two days.  We have been able to detect enteroviruses at a bathing beach
exposed to a known source of sewage although the coliform count at that
                                  18

-------
                    3
beach was quite low.    A direct virus standard for recreational  waters
of no viruses in 10 gallons of water is feasible and should likewise
be established today.

VIRUS STANDARD FOR RENOVATED WATER
The planned use of renovated wastewater for domestic consumption is
a special case which would justify an even more rigorous virus  standard
due to the special risks that might be associated with such use.
Here virus monitoring techniques are not yet sensitive enough to serve
as a basis for monitoring the'virus removal efficiency of the whole
treatment train since it would be reasonable here to insist on the
lowest level of viruses that can be obtained by today's most advanced
                                      12
wastewater treatment technology.  Berg   has suggested that renovated
water be treated in such a manner as to destroy 12 log units of
reference virus.  A treatment train which includes chlorinntion
with 10 mg/1 of HOC1 for one hour or an ozone residual of 0.5 mg/1 for
15 minutes as reported on in  the  report, may well be capable of
achieving this end.  Here the virus removal capability of each step
in the treatment train should be checked independently and the degree
of safety of the processes be established in advance.  In such an
effluent, no enteroviruses should be detectable in a 100 gallon sample.
When it becomes practical to monitor even larger samples, the standard
could be upgraded accordingly.  In order to provide the proper degree
of safety to the public in cases where renovated water is being used
directly for domestic supply, WHO has recommended that each batch be
                                  19

-------
monitored for viruses before release to the distribution system.  For
this, a rapid method of testina for viruses in water is essential since
conventional methods would require three days and more to obtain the
results.  The fluorescent antibody technique developed in our laboratories
by Katzenelson,   and reported upon here holds promise of nrovidino an
effective solution with Quantitative virus assay results in 24 hours.

PRACTICAL PROBLEMS
In advocatina that virus standards for water supolies and recreational
water be established now a number of oractical problems have to be
overcome.  Virus drinking water standard might be applied in the first
instance only to cities having populations of 100,000 or greater which
draw their raw water supplies from surface sources exposed to known
fecal contamination.  This would reduce the number of communities in
the world which would have to carry out mandatory virus assays while
concentrating on those that need it the most.

Although a few very large cities might find it feasible to establish
their own virus laboratory to carry out all stages of the virus
monitoring program it would not be practical or desirable for all
communities  to do this.  Specialized national, regional or state
laboratories could carry out the actual virus assay work for a number
of communities.  Under such a system the virus concentration procedure
would be carried out by local personnel who would send the concentrated
sample  or filter cartridge to the regional laboratory for tissue
                                    20

-------
culture assay.   WHO has already recommended that each country establish
national or regional  laboratories capable of carrying out virus  assays
of water.    Many countries have such facilities or are in the process
of organizing them.
There would be a need to establish one or several  tentative standard
procedures for concentrating and detecting viruses in water.   There
would also be a need to train additional laboratory personnel  in
virus concentration and detection techniques.   The E.P.A.  and the
World Health Organization in cooperation with  national health agencies
should take the lead in establishing the tentative standards  and
training personnel in their application.

In a matter of a few years' time it might well be possible to put the
proposed virus standards for water into practice, at least in a number
of the more highly developed countries where such a requirement could
best be justified.  Although trained personnel and laboratory facilities
may be a limiting factor at first, it should not take too long to
overcome.

CONCLUSIONS
A built-in paradox of modern society revolves  around a situation where
more and more water is needed by the growing  population but less and
less becomes available at the required quality as a result of the self-
destructive process of pollution.  The increasing burden of highly
                                   21

-------
persistent enteroviruses in the water environment on water quality is
of concern not only in relation to drinking water, but in water used
for  recreation, agricultural  irrigation and growinp shellfish.

Enteroviruses have been detected in all  phases of the water environment
and conventional water and wastewater technology is not capable of
removing such viruses under all circumstances in an entirely reliable
manner.  Although only a very small percentage of the total number of
reported infectious hepatitis cases per year is associated with water-
borned outbreaks, the number of infectious hepatitis cases per year
associated with such outbreaks is on the rise and is today the single
most prevalent type of waterborne disease of known etiology.   Massive
epidemics of infectious hepatitis transmitted by shellfish grown in
sewage contaminated water underscore the potential risks  associated
with the growing burden of viruses in the water environment.

It is justifiable today to establish a virus standard for water on the
basis of technological feasibility.  It should be possible even with
conventional  technology to produce a drinking water with  less  than
1   pfu    per 1000 liters.   It is also possible today to monitor 100
gallons of water for the presence of viruses.  A tentative standard of
no viruses in 100 gallons  of water would be a good first  step  in up-
grading water treatment facilities throughout the world.   Virus standards
should be set for recreational water and renovated water  as well.   We
                                   22

-------
have waited too long for research to provide the "perfect"  virus
monitoring technique and the most efficient method for removing viruses
from water.  In the interim little has been done in practice to apply
what we already know about the need to control viruses in water.   Let
us not wait for the "best" but try to apply today what can  be considered
"good enough."  The establishment of a virus standard in water will  be
a real step forward in achieving much needed improvements in water
quality.

RESEARCH PROGRAM
The sections that follow in this report review studies carried out at
the Hebrew University of Jerusalem over a five year period  in the
development and evolution of virus monitoring and detection techniques
and in the inactivation of enteroviruses in water with ozone.  These
studies have furthered the basic understanding of the problems of viruses
in water and their control and should provide a basis for both
practical applications and further investigations.
REFERENCES
 1.    World Health  Organization.   Reuse  of  Effluents:  Methods  of
      Wastewater Treatment and  Health  Safeguards.   Geneva,
      Technical  Report  Series No.  517,  1973.  63  p.
                                   23

-------
2.    Akin, E.W., W.H.  Benton,  and W.F.  Hill,  Jr.   Enteric  Viruses  in
     Ground and Surface Waters:   A Review  of  Their Occurrence  and
     Survival.   In:   Viruses  and Water  Quality:  Occurrence and
     Control, Snoeyink, V.,  (ed.).   Urbana, 111.,  University of
     Illinois,  1971.

3.    Shuval, H.I.   Detection  and Control of Enteroviruses  in the
     Water Environment.  In:   Developments in Water Quality Research,
     Shuval, H.I.,  (ed.).  Ann Arbor, Mich.,  Ann Arbor-Humphrey
     Science, 1970.

4.    U.S.  Department  of Health,  Education  and Welfare.  Report of  the
     Committee  on Environmental  Health  Problems to  the Surgeon General,
     U.S.  Government  Printing  Office, Washington,  D.C.  USPHS Publ.
     No.  908. 1962.

5.    President's Science Advisory Committee.  Restoring the Quality
     of Our Environment.   Report of the Environmental Pollution Panel,
     Washington, D.C.  1965. 317  p.

6.  Craun, G.F., and L.J. McCabe.  Review  of  Causes of Waterborne
    Disease  Outbreaks.  J. Amer. Water Works  Ass.   65:74-84,  1973.
                                  24

-------
 7.   Plotkin,  S.A.,  and  M.  Katz.  Minimal  Infective Doses of Viruses
     for Man by  the  Oral  Route.   In:  Transmission of Viruses by
     the Water Route,  Berg,  G.  (ed.).  New York, John Wiley and Sons,
     1967.

 8.   Berg,  G.  (ed.).   Transmission  of Viruses by the Water Route.
     New York, John  Wiley and  Sons,  1967. 484 pp.

 9.   Phelps, E.B.  Public Health  Engineering.  New York, John
     Wiley  and Sons,  1948.

10.   World  Health  Organization.   International Standards for
     Drinking  Water.   3rd ed.   Geneva, World Health Organization,
     1972.  70  p.

11.   Mel nick,  J.L.   Detection  of  Virus Spread by the Water Route.
     In:  Viruses  and  Water Quality:  Occurrence and Control,
     Snoeyink, V.,  (ed.).  Urbana,  111., University of Illinois, 1971.

12.   Berg G.   Integrated Apporach to Problem of Viruses in Water.
     J.  San. Eng.  Div.,  Proc.  Amer.  Soc. Civil Enqrs. 97: SA6:867-882,
     1971.

 13.   Cookson, J.T.  Report  on  the  Use of Temporary Wastewater
      Treatment Plants:   Standards  and Procedures for the Elimination
      of Health Hazards.  Maryland, Montgomery County Council, 1972.

-------
14.   Katzenelson,  E.,   Virologic  and Engineering Problems in
     Monitoring  Viruses in Water.   In:  Viruses and Water, Berg, G.,
     H.L.  Bodily,  E.H.  Lennete, J.L. Melnick and T.G. Metcalf,  (eds.)
     Washington, D.C.,  American Public Health Association, 1975.
                                 26

-------
                              SECTION II

                             CONCLUSIONS

1.   A methodology was developed to test the effectiveness of different
    systems  for concentrating water samples for the isolation and
    detection of viruses using standardized conditions  so as to obtain
    comparable results.   The following eight different  methods for
    virus concentration  and detection were studied for  sample concen-
    tration  ability and  virus recovery efficiency, using mainly 5  liter
    samples:
      1.    Aluminum alginate ultrafiltration (according to Gartner).1
                                                                           i
      2.    Cellulose nitrate membrane filtration (according to Berg et  al).'-
                                                                              3
      3.    Aluminum hydroxide precipitation (according  to Wallis and Melnick).
      4.    Phase separation (according to Shuval et al).
      5.    Flow-through  gauze sampler.
      6.    Cellulose nitrate membrane filter (according to Wallis  et al).
      7.    Insoluble polyelectrolytes (PE 60) (according to Wallis and  Melnick).
      8.    Cellulose nitrate membrane filter-eluting method (according  to
           Rao and Labzoffsky).
2.   The methods which showed consistently very high recovery efficiencies
    with poliovirus of about 100% were aluminum alginate ultrafiltration,
    aluminum hydroxide precipitation and phase separation.
                                     27

-------
3.  ilitn seeded Fchovirus 7 the Alginate and Phase Separation methods

    showed about 100% virus recovery efficiency while the other methods

    studied showed lower recoveries.


4.  Uhen natural wild types of enteroviruses from sewage were seeded

    in water to be tested by various concentration methods, the

    efficiency of virus recovery was generally lower except for tha

    membrane filtration method (accorciing to Rao et al).


5.  The tests for recovery efficiency with natural wild strains of entero-

    virus indicate that a number of the methods may be selective in

    concentrating enteroviruses which would mitigate against their use

    for field monitoring purposes.   Three methods, membrane filtration

    (according to Rao et al),  membrane filtration (according to Wall is
          7                                           C
    et «l)"and PE 60 (according to Hall is and Melm'ck), showed reasonably

    yood levels of virus recovery of 65% and more with both poliovirus

    alone and with natural  wild strains of enteric viruses.


6.  An in-depth study of the gauze pad method made it possible to provide a

    quantitative evaluation of this much-used, simple technique.  The

    efficiency of virus recovery for large volumes of water passed through

    the gauze pad was under 1%.


7.  Recovery efficiency of over 10016 was found in many experiements and

    fiidy be explained by the pressnce of virus clumps in the samples
                                   28

-------
     which  may become  disaggregated  during  the  concentration  procedure.
     Experimental  evidence  cited  from  other studies  supports  this
     possibility.

 8.   Of the above  eight methods studied  the alginate and  phase  separation
     methods are  limited for practical purposes to  concentrating samples
     of 10  liters  or less and are therefore not applicable  to monitoring
     drinking water samples which should have  a volume  of at  least  100
     gallons.  The cellulose nitrate membrane  filtration, aluminum  hydroxide
     and PE 60 methods are  suitable  for  such large  volumes  of water and
     demonstrate  reasonable recovery efficiencies.

 9.   The phase separation,  PE 60  and gauze  pad  method can be  used directly
     for samples  of sewage  and highly  contaminated  water  while  the  other
     methods are  effective  only with water  samples  relatively free  from
     suspended solids  unless special arrangements for prefiltration are
     provided.

10.   A new, promising  virus concentration technique using cellulose
     acetate hollow fiber ultrafiltration membranes was developed and
     evaluated.  Large volumes of water  can be concentrated rapidly by
     this method  with  reasonable  virus recovery efficiency.  The method
     has the advantage of not requiring  pH  adjustment of the feed water
     as required  with  the cellulose nitrate membrane filtration method.
                                     29

-------
     Thi:> method is not adversely affected by the presence of organic
     contaminants in the water.   It may prove to be an effective first
     stage in a sample concentration train to be followed by other sample
     concentration techniques.

11.   Attempts at a quantitative  evaluation of the efficiency of the gauze
     pad method under laboratory conditions,  indicate an  inverse relation-
     ship between efficiency of  recovery and  sample volume.   At low
     volumes  of 700 ml,  the efficiency of recovery was 7% while with
     50 liters only 0.5% virus recovery was achieved.

12.   In field studies using the  gauze pad, it was demonstrated to be more
     effective than 2 liter grab samples taken from the same sewage stream.
     The concentration of viruses in the liquid expressed from xjauze pads
     immersed in a sewage stream was about 100 times greater than that
     found in unconcentrated grab samples taken from the  sewage in which
     it was immersed.   No advantage was found in keeping  the pad
     immersed longer than 24 hours in the water stream being tested.
     Despite  its low-efficiency  of recovery and non-quantitative nature,
     the gauze pad can be of some value in field survey when other more
     refined  sample concentrating techniques  for large volumes are not
     available.

13.   A rapid  fluorescent antibody method for  quantitative isolation of
     viruses  from water has been developed using poliovirus type I as
     model.  This method has been shown to be capable of giving a
     qualitative answer as to the presence of viruses in a concentrated
                                   30

-------
     water sample in as little as 6-9 hours and a quantitative estimate
     of the virus concentration in 18 hours.  The FA quantitative assay
     has been shown to be comparable to that obtained by the conventional
     plaque forming method which requires about 3-7 days to complete.

14.   A study of virus types in sewage in Israel indicates that of 489
     strains isolated and identified, 74% proved to be poliovirus.
     Ten percent were coxsackie type B or Echo 9 while 16% were other
     Echovirus types or other viruses.  Thirteen percent of the polio-
     virus strains showed strong CPE when incubated at 40°C which is
     generally considered to indicate wild pathogenic strains as
     compared to the remaining 87% which grew only at 37°C.  These
     latter isolates are considered to be attenuated poliovirus vaccine
     strains.  Further field studies, using both grab samples and gauze
     pads checked for poliovirus by neutralization tests, also indicate
     that poliovirus is usually present.  Poliovirus is ubiquitous in
     Israel sewage and can be detected in essentially every sample tested.
     This is to be expected in a country practicing routine immunization
     of all infants against polio with live vaccine.  These findings may
     provide some support for the possible use of poliovirus as a virus
     indicator organism using the rapid FA method.

15.   The chemistry of ozone as a disinfectant is reviewed with possible
     explanations for the chemical pathways leading to the formation of its
     active radicals.  The possibility that a dissociation product,
     hydroxylradical (OH), rather than ozone itself is the active disinfec-
     tant 1s also presented.   Evidence is presented to show that the
                                    31

-------
     disinfectant power of ozone is pH dependent,  increasing  as  pH  rises.
     This may provide some support for the hypothesis  that the OH  radical
     is the active germicidal  species.
16.  A spectrophotometric method for the determination of ozone  in
     small samples of water was developed and evaluated.   This method
     has been shown to detect  ozone concentrations as  low as  0.01 mg/1
     in a 10 ml  sample.
17.  Studies on  the stability  of ozone in aqueous  solutions indicate
     at 5 C ozone concentrations show only minor decreases in 30
     minutes, while at 22°C the concentration was  reduced by  about
     50% in 30 minutes.  Ozone solutions stirred at 100 rpm resulted
     in a rapid  loss of ozone  while at 80 rpm the  ozone concentration
     was more or less stable over 30 minutes.  Studies also showed  that
     unless very carefully purified, virus stocks  caused  rapid  loss of
     ozone.  The results of these studies assisted in  establishing
     conditions  for virus inactivation kinetic studies.

18.  Ozone has been found to have a very rapid virucidal  effect  in  a low
     ozone demand system with  constant ozone levels.   For example,  99%
     of inoculated poliovirus  is inactivated in less  than 10  seconds
     with an ozone.concentration of 0.3 ppm as compared to 100  seconds
     required for the same degree of kill with 0.3 ppm of chlorine  or
     100 minutes for the same  concentration of iodine.

19.  The inactivation kinetics curve shows two stages.  The first  rapid
     kill stage  lasting less than 10 seconds with inactivations  usually
                                  32

-------
     of 99% and more.   The second stage shows a lower rate of kill
     with essentially  total  inactivation of all virus present being
     completed in several  minutes.

20.   Residual  ozone concentrations  under 0.1  ppm produce irregular
     rates of virus inactivation with no inactivation found in many
     tests, while increasing ozone  concentration above 0.2 ppm to
     1  ppm had very little effect on the virus inactivation rate.
     This finding might be interpreted to support the "all or
     nothing" effect attributed to  ozone.

21.   Some of the irregular and unpredictable inactivation patterns
     of ozone on viruses resulted from changes in the degree of
     resistance to ozone of certain virus stock cultures.  It has
     been shown that virus stock culture highly resistant to ozone
     result from transferring virus stock from -70 C to -18 C.
     Clumping may be the responsible mode in these resistant virus
     cultures.  It is  not clear whether wild virus strains in nature
     are of the resistant form or not.

22.   Ozone inactivation of Coliphage T~ and E. coli followed similar
     patterns as that found with viruses but some indication of dose
     response patterns could be detected.

23.   It is hypothesized that because of its extremely rapid virucidal
     action it was not possible to detect normal dose-response expected
     in any normal chemical disinfection reaction.  Preliminary  tests
     of virus kill in very short reaction periods were made and  provide
     the first tentative evidence of a true dose-response relationship.
                                   33

-------
24.  The virus inactivation ability of ozone was tested in clarified
     diluted sewage having B.O.D,  levels of 50 and 25 ppm.  In batch
     tests initial ozone doses of up to 1.8 mg/1 were rapidly
     dissipated.   However, with initial ozone doses of 0.8 mg/J ,
     a 90% inactivation was achieved in 10 seconds with the more
     concentrated sewage and a 99.5% inactivation was achieved with
     the more dilute sewage, while in neither case was an ozone
     residual detectible.   Organic matter concentrations  in the
     substrate were shown  to seriously interfere  with the inactivation
     process.

25.  In experiments with clarified raw sewage (B.O D. = 500 ppm)
     and continuous bubbling of ozone, virus inactivation occurred
     before ozone residuals could be detected.  A 99,9% inactivation
     of virus was achieved with an ozone residual of 0.6  mg/1.  Virus
     inactivation levels greater than 99,9% were achieved with ozone
     residuals of 1 ppm in a matter of minutes.   These data indicate
     that ozone is a very  rapid and effective virucidal agent even in
     wastewater high in organic content.

26.  Studies on natural inactivation of viruses  in the marine
     environment  show that specific marine bacteria possess antiviral
     activity.  The presence of organic matter in the seawater
     inhibits the antiviral effect.  The mechanism of this antiviral
     effect has not been determined,
                                 34

-------
REFERENCES
1,   Gartner, K.   Retention  and Recovery  of Polioviruses on a Soluble
    Utlrafilter,   In:   Transmission of Viruses by the Water Route.
    Berg, G  (ed.),   New York, Wiley  and Sons, 1967   p,  121-127,

2,   Berg, G., D.R.  Dahling, and 0=  Berman.   Recovery of Small
    Quantities of Viruses from Clean  Waters  on Cellulose  Nitrate
    Me.nbrane Filters.   Appl.  Microbiol.  22:608-614,  1967.

3.   Wallis, G,, and J.U Melnick.   Concentration  of  Viruses on
    Aluminum Phosphate and Aluminum Hydroxide Precipitates.   In:
    Transmission of Viruses by the Water Route,  Berg,  G,  (ed.).
    New York, Wiley and Sons, 1967,   p,  129-138,

4.   Shuval, H,I., B, Fattal, S. Cymbalista and N. Goldblum.   The
    Phase Separation Method for the Concentration and  Detection of
    Viruses in Water.  Water Res.  3:225-240, 1969.

5.  Wallis, G., M. Henderson, and J.L, Melnick.   Enterovirus  Concentra-
    tion  on Cellulose Membranes.  Appl.  Microbiol.  23:476-480,  1972.

6.  Wallis, G,, and J.L, Melnick.   Detection of Viruses in Large
    Volumes of Natural Waters by Concentration on Insoluble Poly-
    electrolytes. Water  Res,  4:787-796,  1970.
7.  Rao,  N.V., and N.A.  Labzoffsky.  A  Simple Method for the Detection
    of  Low  Concentrations  of  Viruses  in  Large Volumes of Water by the
    Membrane  Filter Technique.  Canadian Jour, of Microbiol. 15:399-403,
    1969.
                                  35

-------
                                SECTION  III
                              RECOMMENDATIONS

1.   While the comparative virus  concentration  and detection  studies
    reported upon here  were carried out  with 5 liter samples,  the  goal  for
    virus monitoring of water and wastewater should be  the assay of  larce
    volumes of water--400 liters and more.  A  number of the  methods
    evaluation are suitable for  sampling such  volumes,  particularly
    filtration methods  using cellulose nitrate membranes, aluminum hydroxide,
    PE 60 and the hollow fiber membrane  technioue.   Further  work in
    developing and evaluating these methods under controlled conditions
    with large volumes  of water  is called for.

2.   Further study is required to determine  the effect on  virus  recovery
    efficiency of organic matter, turbidity and variation in dissolved
    mineral content of  the water samples since preliminary data indicates
    that these factors  can be of major importance.

3.   Further studies should be carried out to determine  which of the  methods
    shows the least selectivity  in concentrating viruses  from  water
    samples.  In the studies reported upon  here, membrane filtration
    methods and PE 60 showed the least selectivity with natural wild
    strains of enteric  viruses.
                                   36

-------
4.  Further studies are required to determine the factors involved in
    virus recovery efficiencies which are apparently greater than 100%
    of the virus input.  The preliminary findings that this  phenomenon
    may be associated with virus clumping and disaggreaation should be
    fully investigated.

5.  Concentration of viruses from large volumes of water by  filtration
    and elution will invariably require a second step to reconcentrate
    the elute to a volume small enough to be assayed conveniently in
    tissue culture cell systems.  Since the efficiency of the second
    step is just as critical as the efficiency of recovery of the first
    step, it is essential that ways be found to maintain high recovery
    efficiency in it as well.  This question requires further investigation,

6.  The guaze pad method, while shown to have a low virus recovery
    efficiency of under 1% can be a useful  nonquantitative field survey
    technique when suspended in a flowing stream for a 24-hour period,
    particularly when more sophisticated methods are not available.
    Such a sampling technique is in most cases more effective in detecting
    viruses in the stream than small grab samples.

7.  The filtration methods which proved to  be most efficient and
    which are applicable to testing large volumes of water require some
                                   37

-------
    form of pretreatment when water high  in  suspended  solids  must  be
    assayed.   Appropriate prefiItration methods  must be  developed  and
    tested to determine  the extent that viruses  attached to  suspended
    particulates  are removed by  the pretreatment step.

8.   The hollow fiber membrane concentration  technique  developed and
    tested during this study holds particular promise  since  it has a high
    virus  recovery efficiency and can be  designed to handle  large  volumes
    of water.  In addition it does not require pH adjustment of the stream
    for virus adsorption nor is  it adversely affected  by the presence  of
    organic matter and suspended solids in the water.   The hollow  fiber
    membrane system may  prove to be an effective first stage in a  sample
    concentration system followed by other concentration techniques.
    Further intensive study should be devoted to upscalinq the system
    large volumes of 400 liters  and more  and testing its efficiency  under
    field conditions.

9.   The flourescent antibody (FA) method  for the rapid quantitative  assay
    of poliovirus in water should be of particular value where rapid
    results are required.  A study as to  the feasibility of using  this
    method based on the  concept of using  poliovirus as an "indicator
    organism" of virus contamination in water should be field tested  in
    various communities  where routine polio vaccination is practiced  and
                                      38

-------
     it can be assumed that poliovirus  is  continuously  present  in
     the sewage system.

     The possibility of expanding the technique to other  enteric viruses
     should be studied.

10.   Various tissue culture cell  lines  should be evaluated  and  compared
     to primary tissue culture cells to determine the most  effective
     system or combination of systems for  assaying natural  wild virus
     strains from field samples.

11.   Further study of the chemistry of ozone and ozone  species  in  aqueous
     solutions and their disinfection ability is essential  to gain a  better
     understanding of the use of ozone against viruses  under various
     conditions.  The present state of lack of predictability is  a major
     drawback in the use of ozone with natural raw waters or wastewater
     rich in organic compounds or other factors that may  react  with ozone.

12.   The effect of pH should be further studied to determine whether  the
     increased effectiveness of ozone with high pH is a result  of more
     active ozone species or some change in the virus which might increase
     its sensitivity to ozone.
                                    39

-------
     The role of clumping and disaggregation of viruses at various
     pH levels must be investigated in relation to this Question.

13.   Techniques must be developed to study the kinetics of ozone
     inactivation at very short time periods,  ranging from one-half
     to 10 seconds.   Preliminary evidence presented in this study
     indicates that while there is  little or no detectable dose
     response relationship for contact times greater than  8 seconds
     there may be such a relationship at shorter times. Such  studies
     are essential  to an understanding of this question.

14.   Findings relating to the  shifting of ozone sensitive  virus cultures
     to ozone resistant ones  as a result of transferring stock cultures
     from -80 C to  -18 C points to  the need to assure rigorous standar-
     dization of virus inactivation evaluation procedures.   Further
     investigation  should be  made as to the mechanism associated with
     this phenomenon particularly to determine the role of clumping.

     Studies  should  also be made to determine  whether viruses  in nature,
     i.e., in sewage, are of  the sensitive or  resistant form.
                                  40

-------
15.  Ozone has been shown to be a very rapid and effective virucidal

     agent in clean water as well as in wastewater effluent and holds

     promise of being of great practical  value in field application

     particularly where it is not considered desirable to use ever-

     increasing chlorine doses.  Further studies on the use of ozone

     as a virucidal agent under field conditions with various types

     of water are required.



16.  There is a need to investigate the mechanism of virus kill by

     ozone so as to allow for more rational  approach to its use.


                                                          *
17.  In cases where a residual disinfectant  is considered desirable

     in the water supply system, combinations of ozone and chlorine  can

     be considered.  The reactions between ozone and chlorine both in

     clean and raw water must therefore be studied.



18.  If ozone is to be used  instead of chlorine both because of its

     greater virucidal power and due to concern over the formation of

     toxic organohalides resulting from chlorination, there is a  need

     to determine whether the ozonization of organics in water results

     in the production of any deleterious compounds.  Such studies are

     an essential phase in the full evaluation of ozone disinfection

     for mass field application.
                                 41

-------
                                PART A
                       VIRUS  DETECTION METHODS

                              SECTION IV
     A  REVIEW  OF  METHODS  FOR  THE DETECTION OF ENTERIC VIRUSES
                    IN  THE  WATER ENVIRONMENT
INTRODUCTION
The possibility that water might serve  as the  vehicle for the trans-
mission of certain virus  diseases, particularly those whose infectious
agent is excreted through the enteric  tract,  has been considered
feasible for some time.  Mosely  has  pointed  out that over 50 documented
water-borne epidemics of  infectious hepatitis  have been recorded over
the years.   Apart from infectious hepatitis,  poliomyelitis and viral
gastroenteritis were the  only other viral infections that caused epi-
demics suspected of being transmitted  by water.  In most cases, the
epidemiological evidence  was  inconclusive.  An  exception is possibly
                                          o
a small polio epidemic in Nebraska in  1952  where strong evidence
suggested that it was caused  at least  partially by a water-borne virus.
The possibility that viral  gastroenteritis  may  be water-borne on
occasion cannot be ruled  out, however.

The massive waterborne epidemic of infectious hepatitis in Delhi, India,
in 19553 in which some 30,000 persons  became  infected by the contaminated
municipal drinking water which had undergone  what is generally considered
complete and adequate treatment, including  chlorination, emphasized
the need to develop new methods of monitoring water supplies for viruses.
                                    42

-------
Over the years, evidence has pointed to the fact that the usual
bacterial parameters of water purity, particularly the coliform group,
may not provide an adequate index as to the safety of water from a
virological point of view.  The need to monitor water specifically
for viruses of enteric origin presents many problems, and the develop-
ments in this field will be reported here.

THE PRESENCE OF ENTERIC VIRUSES IN WATER
Over 100 virus types are known to be excreted from humans through the
enteric tract and may find their way together with sewage into sources
of drinking water.  Many of these viruses are known to cause disease
in man.  However, the critical question is  whether these viruses can
survive long enough and in high enough concentration to cause disease
in people consuming such contaminated water.  The concentration of
enteric viruses in sewage and polluted water is an important factor
to consider.  Clarke and Kabler  calculated a theoretical average
number of enteric virus in infectious units in sewage and found it to
be about 500/100 ml.  In our own studies we have found the enteric
virus concentration in the sewage of communities in Israel  range from
          c
10-100/ml.   Based on these figures it can  be assumed that the virus
concentration in polluted river water would range from 1-10 viral
infectious units/100 ml as a result of physical dilution only.  The
number is lower during the cold months and somewhat higher in the late
summer and early fall due to seasonal variation of enteric virus
diseases.  It can be assumed that this concentration will be further
reduced both by processes of natural die-away and by water treatment,
imperfect as they may be in the removal and inactivation of viruses.
                                   43

-------
Under normal circumstances only a relatively small number of infective
units will, at worst, penetrate a water supply system which derives its
raw water from a heavily contaminated river.  Simultaneous infection of
a large number of people is therefore rather improbable under normal
conditions with modern water treatment methods.   Sporadic infections,
however,are possible, at least theoretically.   The latter becomes true
particularly in the light of Plotkin and Katz's   claim that "one in-
fective dose of tissue culture is sufficient to  infect men if it is
placed in contact with susceptible cells."  They were able to reach
this conclusion as a result of studies on attenuated polio viruses,
respiratory viruses, agents of ocular diseases,  viruses of animals,
and other agents.   This might mean that even when virus concentrations
as low as one virus infectious unit per 1000 ml  are present in water,
a certain number of individuals might well become infected by consuming
the normal daily intake of 1-2 liters per capita.

With this in mind, one may form a picture of water as playing a small
but important role in the spread of viral diseases in man in areas
provided with modern treatment facilities.  The  effect of such slightly
contaminated water may lead to sporadic cases  of disease dispersed
over a large area.  However, these occasional  cases may, in turn, act
as foci and through food or personal contact cause epidemics which
may involve much larger numbers of people.  In other cases when heavily
contaminated water reaches large population groups without adequate
treatment, explosive mass epidemics have occurred and may well occur
                                  44

-------
in the future.  It is therefore obvious that the development of methods
for the detection of viruses in water are required to allow for an
adequate evaluation of the virological  safety of water supplies and
treatment processes.  Bacterial evaluation of water as an indicator  of
contamination cannot replace such methods since it became apparent  that
viruses are not as sensitive as bacteria to hostile environmental  factors
or to standard purification procedures, and they may be present in  water
                                                       7 8
even when bacterial counts are at acceptable standards. '

TYPES OF VIRUSES IN WATER
Water being used for drinking and bathing can act as a vehicle for  the
transmission of most viruses.  The picorna group of viruses is the  one
most commonly found in sewage; it includes the polio, coxsackie, and
echo  viruses.  Adeno viruses which cause respiratory and eye infections,
and sometimes diarrhea, are commonly found in feces.

Infectious hepatitis is actually the only disease for which a water-borne
infection has been proven beyond any doubt.  However, its viral charac-
teristics are not yet clear.  There are some claims that the responsible
virus has been isolated from suspected cases of hepatitis, but most
viroloaists feel that these claims are as yet insufficiently established.
It must be remembered, however, that in spite of the latter there is
strong evidence supporting the view that this disease is actually caused
           Q
by a virus.
                                  45

-------
ISOLATION AND IDENTIFICATION OF ENTERIC VIRUSES
Viruses can only multiply inside living cells and therefore live organisms,
such as animals, chick embryos, or tissue culture must be used for their
isolation in the laboratory.  For the enteric viruses, tissue cultures
which may be of two types are generally used:  primary tissue cultures
and continuous cell cultures.  Primary tissue cultures are usually pre-
pared according to the method of Enders et al.    This method is based
on the fact that 0.25% trypsin acts on small cuts prepared from a tissue
(usually a kidney) by separating the cells from each other.  When put
inside a suitable glass or plastic flask, tube, or plate together with
a tissue culture nutrient medium, these cells attach to the wall of the
vessel and multiply.   As a result, a monolayer of cells is formed on
the wall.  Continuous cultures are very similar, but instead of a tissue
from an organ, a tissue culture is used as a source for the cells.
Initial isolation of enteric viruses is usually done on primary tissue
cultures prepared from monkey kidneys which typically have a higher
sensitivity than most cell lines.  However, any isolated virus can be
adapted to cultures of the continuous type.

After inoculation of a virus into a tissue culture some of the cells be-
come infected.   The virus multiplies within these cells and spreads to
the neighboring cells.  At the same time, the infected cell usually
undergoes morphological  and biochemical changes and dies.   The result is
a slow process of destruction of cells in the culture, a phenomenon
known as the cytopathic effect (CPE).  The process of viral spread from
cell  to cell  can be slowed down by adding a layer of agar together with
                                    46

-------
tissue culture medium over the cells.   As  a result,  instead  of  being
rapid and confluent, the CPE will  be limited to a smaller area  which  looks
macroscopically like a hole in the monolayer of cells.   These holes
are also known as "plaques".  A single plaque usually originates  from a
single infected cell which may be caused by a single virus infectious
unit.  This method is used for quantisation of enteric viruses  in tissue
cultures in the same manner as agar plates are used  in bacteriology
for the determination of bacterial counts.  The term "plaque forming
unit"  (pfu) was given to the lowest concentration of viruses that form
one plaque on a monolayer of cells.

Different viruses cause cytopathic effects which differ morphologically.
Also, plaques may be of different sizes and shapes.   However, this phe-
nomenon cannot be used for the final identification  of the isolates
since some viruses, belonging to different groups, cause identical CPE.

Final identification can only be achieved with specific antisera.  Here
the identification is based on the fact that specific antiserum will
neutralize the effect of the virus against which it  was prepared.

QUANTITATION OF ENTERIC VIRUSES
Two methods are available for the quantitative determination of enteric
viruses in a given sample of material  being assayed, both of which give
accurate results.  Selection of the method to be used is usually  based
on the experience and resources of the laboratory.  In the first  one,
                                  47

-------
the tube assay method, serial dilutions of the virus suspension to be
tested are prepared.  Groups of tissue culture tubes are inoculated with
each dilution.  After proper incubation at 37°C, the inoculated tubes
are examined for CPE.  Quantisation is obtained by finding the lowest
dilution of the virus suspension that caused CPE in 50% of the tubes.
The figure obtained is known as the TCID™ (tissue culture infectious
dose-50%) value of the virus suspension.   Using this same method, it
also is possible to calculate the virus concentration as a most pro-
bable number (MPN).

In the second, the plaque assay method, quantities of 0.3-1.0 ml  of
virus dilutions are inoculated into plates or bottles, the cells  of
which are later covered with an agar overlay.  After proper incubation,
usually at 37°C in a humid atmosphere containing 5% CO^. the inocu-
lated tissue cultures are examined for the presence of plaques.  Wien
plaques are present, they are counted and their number for each of
the dilutions is determined.  The number of pfu in the original virus
suspension is then calculated and the virus concentration is reported
as pfu/ml or other unit of volume.

A large number of methods for the isolation of virus from water have
been developed over the years.   A common  step in almost all these methods
is the concentration of the virus from the water sample.  This is an
integral  part of the procedure since the quantity of viruses in water
is often very small.  Raw domestic sewage, the main source of viral
                                  48

-------
contamination in water, contains about 10,000 plaque  forming  units  (pfu)
per liter.  This number, however, may sometimes  increase.   In our lab-
oratory we recently found up to 10  pfu/liter raw domestic  sewage.   When
the virus number reaches such high proportions their  isolation becomes
                                               2
relatively simple; but already at a level  of 10   pfu/liter  concentration
of the virus is advisable if an exact quantitative evaluation is  to be
achieved.  At lower virus levels concentration becomes  a necessity.
The basic difficulty in the isolation procedure is the  fact that  usually
the inoculum of a tissue culture has a volume of 0.1-1.0 ml.   In  other
words, when virus concentration in water is less than one per milliliter,
it will require many inoculations.  It is  much simpler  to concentrate
the viruses from large volumes of fluid and to use this concentrate as
the inoculum.  Most systems for the isolation of viruses from water are
therefore primarily concentration methods.  In fact,  the different
concentration methods can be divided into  seven main  groups.

1.  Sample incorporation
In this method the conventional culture medium is so  concentrated as to
allow for the incorporation in it of 10-60 ml of the  sample to be
assayed.    According to another approach, large volumes of a maintenance
medium, prepared from the water that is being tested, are inoculated
                   12
onto cell cultures.    No concentration of virus from the water prior
to inoculation is attempted in these cases.  Nevertheless,  by this
method virus detection is enhanced since with the same  tissue culture
tube or bottle a volume of sample 10-20 times larger than normally
inoculated can be assayed.
                                  49

-------
 2.   Ultrafiltration
 The  sample  is  passed  through  a  filter with  a  pore  size  small enough  to
 hold back the  virus,  which  remains  either on  or  in  the  filter, as  is
                                            13
 the  case with  the  aluminum  alginate filter.    After filtration, the
 filter  is dissolved in  a small  volume of 3.8% sodium citrate solution
 providing a  virus  concentrate.
 A  more  sophisticated method of ultrafiltration is to let the water flow
 over  the  surface of the filter thus preventing the virus and suspended
 solids  from entering the filter and clogging it.  The movement of the
 water is  caused by constant stirring or by tangential direction of
 the flow.    The end volume, which has not passed through the filter
 is small  and contains the concentrated viruses.  Cellulose acetate
 membranes are used in this system mainly.

 Hydroextraction,  ' 16 or its more advanced form, osmotic filtration,17
 also  belong in the ultrafiltration group.  Here, water is extracted
 from  the  sample through a dialysis membrane, either by hygroscopic
material  or by a concentrated salt solution.
3.  Freezing
The sample is slowly cooled to -15°C under constant stirring.  The
ice crystals formed during the freezing process are pushed toward
the periphery.  Ultimately, a small  volume of super cold water re-
mains in  the center containing all the substances that were present
in the original  sample:   salts, colloids, suspended solids including
        18
viruses.
                                  50

-------
4.  Two-phase separation
                                                   19
This method is based on the discovery of Albertsson   that the result of
certain mixtures made from two polymers such as dextran sulfate and
polyethylene glycol in an aqueous solution leads to the formation of
a two-phase system.  Introduction of particles and macromolecules into
this system will result in the partition of the particles in the two
phases, depending on their size and surface properties.  Viruses show
a nearly one-sided distribution and the method may therefore be used
                        20 21
for their concentration.  '    The concentration is accomplished by
adding polymer solutions to a virus suspension in such proportions
that almost all the virus particles are collected in a small volume
bottom phase and may be drained off separately.

5.  intracentrifugation
The centrifuge is mostly used for the concentration of small suspended
particles from fluids.  Being extremely small particles in the size
range 20-200mp, viruses are no exception, but because of their small
size, relatively high forces of the order of 60,000 x g for one hour
are required.  These are obtained by using an ultracentrifuge.  In
the usual procedure, the water sample is first centrifuged at a
relatively low speed to reduce the number of larger particles including
bacteria.  The supernatant is then centrifuged at high speed.  The
sediment obtained contains the virus and is  resuspended in a small
volume of tissue culture medium.  A high concentration factor may
                                22
thus be reached.  Anderson et al   described the development of a
complex centrifugation  system for  the  isolation and  separation  of
small numbers of virus  particles from  large  volumes  of fluid.
                                   51

-------
6.  Electrophoresis


                                                           23
Viruses are usually negatively charged at neutral  pH values   and will



therefore move toward the cathode when a virus suspension is placed in



an electric field.  This principle was used by Bier et al for concentra-


                             24
ting bacteriophages in water.    A simple procedure was developed by which



electrophoretic transport was used to bring about  adsorption of bac-



teriophages on dialyzing membranes.  With this method a sample of water



could be processed in a relatively short time and  a concentration of

                                              pc                    pc

viruses achieved.   Forced flow electrophoresis   and electro-osmosis



are two more advanced methods based on the same principle.





7.  Adsorption and Elution


                                                                    27
With this method viruses are adsorbed from water onto either cotton,

                   pn                   pg                           20

aluminum hydroxide,   protamine sulfate,   insoluble polyelectrolytes,



or cellulose nitrate,   which are amongst the adsorbants most often used.



The adsorbant is sometimes added to the water sample in powder form and



separated by centrifugation or simple filtration.   A second approach is



to pass the sample through a layer of adsorbant placed directly on a



filter, or the adsorbant itself fulfils the role of filter.  The viruses



are usually eluted from the adsorbant by a small volume of either


        32                  31
alkaline   or a protein-rich   solution.





The large variety of  methods is impressive and is still increasing,



which underscores the importance of viruses in water.  But, at the same



time, it also indicates  that the methods are still lacking in proficiency.
                                    52

-------
What do we expect from the ultimate virus isolation method?  It should
be able to quantitatively isolate in a very short time all  the different
viruses found in large volumes of various qualities of water.   Each of
these requirements present their own specific problems, and we shall
therefore discuss them one by one.
VIRUS TYPES AND RECOVERY EFFICIENCY
More than a hundred different virus types are found in domestic sewage.
Most of them belong to the picorna group, which include poliomyelitis,
coxsackie and echoviruses. But reo- and adeno viruses are also often
isolated from sewage.  This great divergency in virus types poses the
first difficulty—the lack of one single system for their isolation.
For example, many of the coxsackie viruses multiply in suckling mice,
but not in tissue cultures.  Echo and polioviruses, on the other hand,
multiply in tissue cultures only.  This is an extreme case.  Generally,
different virus types favor specific tissue cultures.

Two virus types found simultaneously in the sample, may prefer the same
kind of tissue culture, but may differ in their multiplication rates as
is seen with polio- and reoviruses.  The poliovirus proliferates at a
greater rate than the reovirus and thus destroys the cell culture while
the reovirus is still in its initial stages of development.  This phe-
nomenon is especially noticeable when using the plaque-count method.
On the other hand, the tissue culture infective dose (TCID5Q) method
                                                               33
allows the more slowly multiplying virus to appear.  Thus Nupen    found
                                 53

-------
the TCIDcn method for detecting and quatitation of viruses in sewage
        bu
superior to the pfu method.
In addition to the isolation system, the concentration method,  too,  may
                                 34
be selective. Grindrod and Oliver   found the two-phase technique a
significant aid in the detection of all  poliovirus types, coxsackie  virus
type B-3, and coxsackie virus type A-9.   The method was found to be   in-
effective for samples containing coxsackie virus type B-2 and echo virus
type 6.  The lower phase, into which the viruses concentrate, is
strongly inhibitory to these two agents  and to influenza A(PR-8) virus,
suggesting that the effect might extend  to other viruses as well. They
therefore warn users of this procedure in environmental virology studies
to be aware of the fact that their results may be biased significantly
by the selective action of the dextran sulfate employed in this techniuqe.

This selectivity was also demonstrated by a comparative study carried
out in our laboratory detailed elsewhere in this report.  The viruses
employed were laboratory strains of poliovirus I, echovirus 7 and
"natural" viruses.  The natural viruses  were, in fact, small amounts of
domestic sewage that were introduced into the sample.  The pronounced
difference in the recovery efficiency of laboratory strains and "nat-
ural" virus was striking.  Recovery of laboratory strains was
satisfactory to very good while "natural" viruses provided a relatively
low recovery.  It should be remembered that these "natural" viruses
are present in the water sources and it is their isolation that we
                                   54

-------
that we want to achieve.  It is therefore important that virus concentra-
tion methods be examined as to  their ability to concentrate not only
laboratory strains but also, and specifically, the "natural" virus,  as
was done by us.

THE ELUENT AND RECOVERY EFFICIENCY
Another outstanding fact is the discrepancy between input and recovered
virus.  The quantity of the recovered virus at times is larger than  the
inoculum.  The phenomenon cannot be ascribed to multiplication since
viruses multiply in cells only, contrary to bacteria.  In trying to
elucidate this finding, the question of the different eluates in the
methods was raised.  And indeed, suspension of the control virus in
the various eluates sufficed to increase the titer 2 times, and some-
times even more.  Hamblet et al   also describe similar findings.

The relationship between inoculum and recovered virus is probably
connected with the existence of virus clumps, whose dissociation is
caused by the eluent.  Such virus clumps were described by Galasso
          36
and Sharp.    Virus clumps are important in the monitoring of water
when exact quantitative results are desired.  Faulty quantisation
caused by such clumps may have serious consequences.  It should be
stressed therefore that such clumps have to be taken into account in
the concentration of viruses from water.

WATER QUALITY AND THE EFFICIENCY OF THE METHODS
In nature viruses are found in waters of different qualities:  raw
sewage, waste water effluents at various stages of treatment, rivers,
                                   55

-------
lakes, and seawater.  These waters may contain suspended participates,



organic matter or salts.   Their pH may vary from alkaline to acid.   All



these factors affect the virus concentration methods, specifically  the



large adsorption and elution group.  In these latter methods, adsorption



takes place as a result of electrostatic attraction between the adsor-



bant and the viruses.  Particulates, organic matter, salts and pH


                                                         37
differences affect the adsorption efficiency.  Berg et al   described



the strong inhibitory effect of a very small amount of organic matter



on the recovery efficiency of the cellulose nitrate membrane adsorption


                      3?
method.  Wall is et al,   on the other hand, showed that the addition



of a small quantity of aluminum chloride increases the efficiency of


                                              38
this particular method, and Rao and co-workers   were able to fully



recover inoculated entero-virus from raw sewage that had been steri-



lized by simply lowering its pH to 3.0.  On the basis of these results,


             39
Wall is et al   incorporated a special procedure into their portable



virus concentrator, which automatically lowers the pH of the sample.



Efficient use of the equipment is hampered, however, by the need for



adjustment with each change of water quality.
Another aspect of water quality is the presence of suspended solids.


These solids play an important role in methods involving filtration since


they clog the filters.   In our laboratory we did not succeed in filter-



ing more than 10 liters of prefiltered water through an alginate filter


with a 142 mm diameter.  Results were even worse with ordinary tapwater:



not more than 5 liters  could be passed through one filter.
                                   56

-------
The quantity of water necessary for good virological  evaluation is a
                 40
moot point.  Berg   suggested a minimal  volume of 100 gallons.   In the
light of current developments of concentration methods which enable the
processing of such large volumes of water, this is not an exaggerated
demand.
In trying to find a solution for the clogging of the filters, a pre-
filtration step was added to some of the methods.  In this context
            OQ
Wall is et al   suggested to serially pass the water through clarifying
filters with porosities of 1-5 urn to remove particulate matter and then
through a 1 umeter cotton textile filter to electrostatically remove
submicron ferric and other heavy metal complexes.  In experiments
carried out under controlled laboratory conditions, the authors
succeeded in recovering 80% of the virus that had been added to large
quantities of water.  It is reasonable to assume that under field
conditions, however, recovery efficiency will fall far below the above
                                        41
number.  This assumption rests on Berg's   finding that in natural
environment more than 50% of the viruses are sometimes adsorbed onto
solids, or manifest themselves in the form of large clumps.  Pre-
filtration may extract these viruses, and a false picture of the quantity
of virus in the water is thus obtained.
SAMPLE VOLUME AND RECOVERY EFFICIENCY
Some workers maintain that recovery of the entire virus population from
the water is superfluous and, anyhow, almost impossible to achieve.
                                   57

-------
Sampling of large volumes, even when recovery will  be low, should comp-
ensate for this flaw.

The gauze pad method,  advocated by many in this  connection, is  easy  to
use and facilitates large water volumes to be sampled.   Some re-
searchers       pointed to the feasibility of using the gauze pad under
extreme conditions.  In its original conception  the method was  meant
to be qualitative only.  Coin,   however,  built  a flow-through  gauze
sampler which allows a quantitative virus  estimation as well.   Liu et
   44
al   tested virus recovery with a similar  apparatus and came to the
conclusion that in spite of its low recovery, a  further development
of this device appears warranted since  (a)  the method is simple,  (b)
it is capable of sampling large volumes of water, (c) the cost  of
collecting samples is  low, and (d) it enables a  rough quantitative
assessment of viral pollutants in water.

This last proposition  was examined in our  laboratory with the aid of
a flow-through gauze sampler similar to the one  described by Liu  et
   44
al.    We tested the effect of sample volume on  the recovery efficiency
of the method.  In Section V,   we show that there  is an inverse  re-
lationship between sample volume and virus recovery:  the larger  the
former, the lower the  latter.  It follows  that the  method is unsuitable
for quantitative virus estimation.
                                  58

-------
THE TIME FACTOR IN VIRUS ISOLATION
The most prominent reason for monitoring viruses in water is to protect
the population from health hazards.  The results of the tests should
therefore be available before the water reaches the consumers.   Since
impoundment of vast quantities of water for long periods of time is un-
desirable, rapid results of water testing is therefore imperative.

Standard bacteriological tests take 24 hours, and even this relatively
short time is too long.  What should be said, then, about virological
tests which take days, and sometimes even weeks?  The reason for this
time lag is not inherent in the concentration methods.  Although some
methods are slow to perform, as for example, the phase-separation
                               45
method which needs 24-48 hours,   most concentration methods require
1-6 hours only.  The time taken up in virus isolation, on the other
hand, is much longer.  The accepted techniques involve inoculation  of
tissue cultures followed by incubation.  Virus is then demonstrated
by the appearance of cytopathic effect or plaques in the tissue
cultures.  The entire  process lasts from 4 to 7 days, or longer, de-
pending on the method.  In our laboratory, the plaque assay is used
and 5 days after sampling quantitative evaluation of virus  Is possible.
The time needed for plaque assay can be shortened, but then the
danger exists of incomplete plaque development, which is expressed by
a  too small number of  plaques or by false negatives.

In an effort to find a solution for the difficulties  related to the
time problems, a project for the rapid isolation and  identification  of
                                   59

-------
viruses from water by a fluorescent antibody  (FA) technique was initiated
in our laboratory, and is reported in Section  VIII.  Fluorescent anti-
bodies stain cells that contain viral antigens and thus enable
identification of eventual viruses in the cells considerably before the
development of a visible cytopathic effect.  We have shown that with
this method it is possible to achieve qualitative results in 9 hours
and a quantitative virus assay in 18-24 hours.

CONCLUSIONS
In this review only part of the problems related to virus monitoring
in water were dealt with.   Items such as cost and simplicity of the
methods were barely mentioned, although these are important factors.
The aspects discussed here are characterized by their direct effect
on the ability of the monitoring methods to quantitatively and rapidly
isolate thp great variety of viruses found in large volumes of water
of different qualities.

Some of the problems are relatively easy to solve.   For example, the
inverse effect of the sample volume on the recovery efficiency is
characteristic for the gauze method only.   It was mentioned here to
serve as a warning to be taken into account when developing new
methods.   On the other hand, clumps and viruses adsorbed onto solids
are intricate matters.  Comprehensive research is required to deter-
mine how far these factors are implicated in the quantitative
estimation of viruses in water.
                                   60

-------
Although many concentration methods are available, none of them can be
said to be superior in the full sense of the word.  Many workers have
directed their research activities toward adsorption and elution methods,
since these are easy to perform.  But these methods are sensitive to
changes in water quality.  An answer to this problem may be the latest
ultrafiltration methods such as reverse osmosis.  In our laboratory we
are testing the ability of cellulose acetate hollow fiber units to
concentrate viruses from water (See Section VII ).  These units provide
large filtration areas in relatively small space.  One of the units
tested is approximately 5 cm in diameter and 50 cm in height.  In
spite of its small size, it dehydrates at the rate of about 500 ml per
minute.  This means that 100 liters are reduced into a reasonably
small volume in less than 4 hours.  Since these units are not affected
by salinity, pH differences or organic matter, the method is promising
and continued research into this direction is recommended.

The need for quick results of virological tests cannot be underestimated.
Somehow, this subject has not been given the attention it deserves and
as a result it lags behind.  This is apparently one of the reasons why
virological examinations are not accepted as required tests for potable
water.  Interest should be focused on the time factor and priority be
given to research in this area.  Our findings with the fluorescent
antibody technique point toward a possible solution.

The problem of isolation of the numerous virus types and  their  recovery
still remains insurmountable.   It is unlikely that a single  system,
                                   61

-------
sensitive enough to isolate the great variety of viruses in water will
be discovered in the near future.   In this context it should be noted
that coliform bacteria are used as indicators for fecal  pollution of
water.  These bacteria are very well  suited for this purpose since they
are always present in feces and, consequently, in raw domestic sewaqe.
Of the great number of enteroviruses, only the 3 polio types are always
present in domestic sewage of urban areas in developed countries be-
cause of the widespread routine administration of live polio virus
vaccine to infants.  It would be advisable, therefore, to weigh the
possibility of having these viruses serve as indicators  of viral
pollution of water.  It may be added  that such choice will  certainly
solve many of the problems related to the monitoring of  viruses in
water.

REFERENCES
1.  Mosley, J.W.  Transmission of Viral  Diseases by Drinking Water.  In:
    Transmission of Viruses by the Water Route, Berg, G. (ed.). London,
    Interscience, 1962.

2.  Bancroft, P.M., W.E. Engelhard and C.A. Evans.  Poliomyelitis in
    Huskerville (Lincoln), Nebraska.   J. Am. Ass. 164:836, 1957.

3.  Dennis, J.M.  Infectious Hepatitis Epidemic in New Delhi India.  J.
    Am. Wat. Wks. Ass. 51:1288-1296,  1959.

4.  Clarke, N.A. and P.W. Kabler.   Human Enteric Viruses in Sewage.
    Hlth. Lab. Sci. 1:44, 1964.
                                  62

-------
5.   Shuval,  H.I.   Detection and Control  of Enteroviruses  in  the  Water
    Environment.   In:   Development in Water Quality Research,  Shuval,
    H.I.  (ed.).   Ann Arbor, Ann Arbor-Humphreys  Publ.  1970.

6.   Plotkin, S.A. and M.  Katz.   Minimal  Infective Doses of Virus for
    Man by the Oral  Route.   In:  Transmission of Viruses  by  the  Water
    Route, Berg,  G.  (ed.),  London, Interscience, 1967.
7.   Kjellander,  J. and E.  Lund.  Sensitivity of  E.coli  and Polio virus
    to Different Forms of Combined Chlorine.  J. Am.  Wat.  Wks. Ass.
    57:893,  1965.

8.   Shuval,  H.,  S. Cymbalista,  A. Wach,  Y. Zohar and N. Goldblum.  The
    Inactivation of Enteroviruses in Sewage by Chiorination.  Proc. 3rd
    Int.  Conf. Water Pollution  Res., January 1967.
9.    Kissling, R.E.  Laboratory Status of the Infectious  Hepatitis Agent.
      In:  Transmission of Viruses by the Water Route, Berg,  G. (ed.).
      London,  Interscience,  1967.

10.   Enders, J.F., T.H. Weller, and  F.C. Robbin.  Cultivation of the
      Lansing Strain of Poliomyelitis Virus  in Cultures of Various Human
      Embryonic Tissues.  Science. 109:85, 1949.

11.   Rawal, B.D.  and S.H. Godbole.   A Simple Tissue Culture  Method for
      Detection of Viruses in Drinking Water;  Sample Incorporation
      Method.  Environmental Health.  6:234,  1964
                                  63

-------
12.   Berg, G., D.  Berman, S.L.  Chang,  and N.A.  Clarke.   A Sensitive
     Quantitative  Method for Detecting Small  Quantities  of Virus  in
     Large Volumes of Water.   Am.  J.  Epiol.  83:196,  1966.

13.   Gartner,  H.   Retention and Recovery of  Polioviruses on a
     Soluble Ultrafilter.  In:   Transmission  of Viruses  by the  Water
     Route, Berg,  G.  (ed.).   London,  Interscience, 1967.

14.   Shohmair, K.   Virus Concentration by Ultrafiltration.   In:
     Methods in Virology, Maramorosk,  K.and  H.  Kopeowski, (eds.).
     New York, Academic Press, 1967.   Vol. 11:245-274.

15.   Shuval, H.I., S.  Cymbalista,  B.  Fattal  and N. Goldblum.  Concen-
     tration of Enteric Viruses in Water by  Hydro-Extraction and Two-
     Phase Separation.   In:   Transmission of  Viruses by  the Water  Route,
     Berg.  G.  (ed.).  London,  Interscience, 1967.
 16.  Cliver,  D.O.  Detection of Enteric  Viruses by Concentration with
     Polyethylene Glycol.  In:  Transmission of Viruses by the Water
     Route, Berg. G. (ed.). London, Interscience, 1967.

 17.  Sweet, B.H., J.S. McHale, K.S. Hardy, F. Morton,  J.K. Smith,  and
     E. Klein.  Concentration of Virus from Water by Osmotic Ultra-
     filtration.  Water Res. 5:823-829,  1971.

 18.   Rubenstein, S.H., H.G. Orbach, N. Shuber,  E. King and J.  Zachler.
     Freeze Concentration of Viral Agents from Large Volumes of Water.
     J.A.W.W.A.  63:301-302, 1971.
                                 64

-------
19.   Albertson, P.A.   Partical  Fractionation  in  Liquid  Two-Phase
     Systems.   Biochim.  et Biophys.  Acta.  27:378,  1958.

20.   Albertson, P.A.   Partition of Proteins  in Liquid Polymer Two-
     Phase Systems.   Nature.  182:709,  1958.

21.   Frick ,  G. and  P.A.  Albertson.  Bacteriophage Enrichment in  a
     Liquid Two-Phase System  with  a  Subsequent Treatment with
     "Freon"  113.  Nature.  183:1070, 1959.

22.   Anderson, N.6.,  G.B.  dive, W.W.  Harris  and J.G. Green.  Iso-
     lation of Viral  Particles  From  Large  Fluid  Volumes.   In:
     Transmission  of Viruses  by the  Water  Route, Berg.  G.  (ed.).
     London,  Interscience, 1967.

23.   Brinton,  C.C. and M.A.  Lauffer.  In:   Electrophoresis, Bier, M.
     (ed.).  New York, Academic Press, 1959.
24.  Bier, M., G.C.  Bruckner, F.C. Cooper and H.E. Roy.  Concentration
     of Bacteriophage by Electrophoresis.   In:   Transmission  of Viruses
     by the Water Route, Berg, G.  (ed.).  London,  Interscience, 1967.

25.  Ellender, R.D., F. Morton, J. Whelan and B.H. Sweet.   Concentra-
     tion  of Virus from Water by  Electro-Osmosis and Forced-Flow
     Electrophoresis.  Prep. Biochem.  2:215-228, 1972.

26.  Sweet, B.H. and R.D. Ellender.   Electro-Osmosis:  A New Technique
     for  Concentrating Viruses from Water.  Water Res.  6:775-779, 1972.
                                  65

-------
27.  Melnick, J.L., J. Emmons, E.M. Opton and J.H. Coffey.   Coxsackie
     Viruses from Sewage.  Methodology Includinq an Evaluation of
     Grab Sample and Gauze Pad Collection Procedures.   Am.  J.  Hyg.
     59:185-195, 1954.

28.  Wallis, C. and J.L. Melnick.  Concentration of Viruses on
     Aluminum and Calcium Salts.  Am. J.  Epidemic!. 85:459, 1966.

29.  England, B.  Concentration of Reovirus and Adenovirus  from
     Sewage and Effluents by Protamine Sulfate Treatment.   Appl.
     Microbiol. 24:510-512, 1972.

30.  Wallis, C., S. Grinstein, J.L. Melnick and J.E.  Field.
     Concentration of Viruses from Sewage and Excreta  on Insoluble
     Polyelectrolytes.  Appl. Microbiol.  18:1007, 1969.

31.  Cliver, D.O.  Factors in the Membrane Filtration  of
     Electroviruses.  Appl. Microbiol. 13:412, 1965.
32.  Wallis, C., M. Henderson and J.L. Melnick.  Enterovirus
     Concentration on Cellulose Membranes.   Appl. Microbiol. 23:476-480,
     1972.

33.  Nupen, E.M.  Virus Studies on the Windhoek Wastewater  Reclamation
     Plant.  Water. Res. 4:661-672, 1970.

34.  Grindrod,  J.  and D.O.  Cliver.  Limitations of the Polymer Two-Phase
     System for Detection of Viruses.   Arch.  Ges. Virusforsch.
     28:337-347, 1969.
                                  66

-------
35.   Hamblet,  F.E.,  W.F.  Hill,  Jr.,  and  E.W.  Akin.   Effect  of  Plaque
     Assay Diluent upon Enumeration  of Poliovirus  Type  I.   Appl.
     Mi:: rob id.  15:208, 1967.

36.   Galasso,  G.L. and  D.G.  Sharp.   Virus  Aggregation and the  Plaque-
     Forminq Unit.  J.  Immunol.  88:339,  1962.

37.   Bero, G., D.  R.  Dahling and D.  Berman.   Recovery of Small
     Quantities  of Viruses from Clean  Waters  on  Cellulose Nitrate
     Membrane  Filters.   Appl.  Microbiol.   22:608-614, 1971.

38.   Rao, V.C.,  U. Chandarkar,  N.U.  Rao,  P.  Kumaran  and S.B. Lakhe.
     A Simple  Method for Concentrating and Detecting Viruses in
     Wastewater.   Water Res.  6:1565-1576, 1972.

39.   Wall is, C. ,  A.  Hanna and J.L. Melnick.   A Portable Virus  Con-
     centrator for Testing Water in  the  Field.   Water Res.  6:1249-1256,
     1972.
 40.   Berg, G.  An Integrated Approach to  the Problem of Viruses in Water.
      In:  Proc.  of the National  Specialty Conference on Disinfection.
      New York,  A.S.C.E., 1971.  p339-364.

 41.   Berg, G.  Reassessment of the  Virus  Problem  in Sewage and  in
      Renovated Waters.  In:  Progress in Water  Technology, Jenkins,
      S.H. (ed.).   New York, Pergamon  Press,  1973.  Vol. 3, p. 87-94.
                                   67

-------
42.   Lund E., and C.E.  Heckstrotn.   A Study on Sampling  and Isolation
     Method for the Detection of Viruses in Sewage.   Water Res.  3:823-832,
     1969.

43.   Coin,  L.  In:  Transmission of Viruses by the Water Route,  Berg,
     G. (ed.).  New York, Wiley Interscience, 1967.

44.   Liu, O.C., D.A.  Brashear, H.R.  Seraichekes,  J.A. Barwick  and  T.6.
     Metcalf.  Viruses  in Water.  Appl.  Microbiol.   21:405-410,  1971.

45.   Shuval, H., B. Fattal ,  S. Cymbalista and N.  Goldblum.   The  Phase-
     Separation  Method for  the Concentration and Detection  of Viruses
     in Water.   Water Res.   3:225-340,  1969.
                                   68

-------
                              SECTION V



            COMPARISON OF EIGHT CONCENTRATION  METHODS  FOR



                    ISOLATION OF VIRUSES IN WATER





INTRODUCTION



Various methods for the detection of small  quantities of viruses  in  water


                                         1                             9
have recently been developed.  Hill  et al   and Shuval and  Katzenelson^



have reviewed and compared different individual studies of methods



for concentration and detection of viruses  in water.   However,  for



proper evaluation of different methods, it  is essential to use  the  same



water sample in controlled tests of recovery efficiency.  Different



research workers have reported studies in which two or more concentration



procedures were compared on split samples.  The results of these studies

                          3

are summarized by England.







This study presents a controlled comparison of eight different  methods



for the concentration and detection of seeded enteroviruses in  a  known



water sample and is aimed to determine which of these method(s) would



be most appropriate for routine monitoring  of polluted water or



renovated wastewater.





MATERIALS AND METHODS



Viruses



Attenuated poliovirus I was obtained and treated as described by Shuval

      A

et al.   Echovirus 7 (Wallace, AGKP 8A), was obtained from G. Berg,
                                   69

-------
Environmental Protection Agency, Cincinnati, Ohio.  Both viruses were
stored in 2 ml volumes at -70°C, thawed shortly prior to each experiment,
and diluted in M-199 medium to the concentration required.
Cell Cultures and Virus Assays
Vero Cells -
A continuous line of African green monkey kidney cells (Flow Laboratory
Ltd., Scotland) was used for studies with poliovirus.  The cells were
grown in M-199 medium, supplemented with 10% calf serum.
BGM cells -
A continuous line of monkey kidney cells (kindly supplied by Dr. G.
Berg) was grown in M-199 medium supplemented with 10% fetal bovine
serum.  These cells have been characterized by Barron et al  and
Danling et al.    The cells were used in experiments with  echovirus 7
and with viruses from diluted raw sewage.  Virus was titrated by
the plaque method, using disposable plastic petri dishes, and was
expressed as plaque forming units (pfu).  Virus assays were carried
out as described by Shuval et al;  the incubation period  was 3 days.

Antiobiotic Solution
One ml of antibiotic stock solution contained 200 mg streptomycin,
2.0 x 10  units of penicillin, 4 mg neomycin and 5 mg kanamycin.  The
solution was sterilized by filtration.  All  media or solutions used
in this study contained 1 ml of antibiotic solution per liter unless
otherwise stated.
                                   70

-------
Aluminum Hydroxide
The suspension was essentially prepared as described by Wall is and
Melnick.   However, the stock suspension was twice as concentrated
as the above described, according to B. England (personal  communication)

Membrane and Filtration Equipment
These consisted of pressure vessels and filter holders of stainless
steel.  Filters used were either type MF (Millipore Corp), 47 mm
diameter, 0.45y pore size, or Sartorius (Cat. #SM 11306), 142 mm
diameter, 0.45u pore size, or aluminum alginate Sartorius ultra
filters (Cat. # 12710), 142 mm diameter, pore size 0.05y.   For some
experiments Gelman Cartridge filters were used:  (Cat. No. #12104,
3p  pore size and Cat. No. 12106, 0.22y  pore size, both 16 aim
length).

Ultrasonic Treatment
This was carried out with the aid of a Measlring and Scientific
Equipment Ltd., London, MSC 100 watt ultrasonic disintegrator, probe
#25925, using a 20 ml tube at maximum power and peak activity for
10 min.

Gauze Sampler
A special gauze sampler was constructed for this study as described
by Fattal et al8 (See also  Section VII).

Concentration Methods and Procedures
This  study was carried out in two  stages:  the first was to compare
                                   71

-------
 five different methods for the concentration and detection of seeded virus
 in  5 equal aliquots of the same tapwater sample.  The procedures of these
 methods are described in test procedure A.  The second stage was to
 compare 4 methods one of which had been employed in stage one.  This
 stage  is described in test procedure B.  In both stages, we used the
 aluminum hydroxide precipitation method (Wallis and Melnick ) thus enabling
 us  to  compare methods from the two stages.
 Test Procedure A-
 A typical experiment was carried out as follows:  25 liters of tap water
 were dechlorinated with 15 ml of 10% sodium thiosulfate, the pH adjusted
 to  7.0, and then filtered through a membrane filter (Mi Hi pore Corp) with
 porosity of 0.45pi.  The virus stock was diluted in M-199 medium and
 added  to the water sample until  the desired concentration was attained.
 The sample was divided into five equal  aliquots, each of which was
 treated by one of the methods described below.   The virus stock diluted
 in the M-199 medium and assayed for viruses served as control.

 1.  Aluminum Alginate Ultrafiltration - (according to Gartner9).   Water
 samples of 5 liters were filtered through an aluminum alginate ultra
 filter at a positive pressure of 2 atmospheres.  After filtration,
 the membrane was separated from the supporting  filter and dissolved in
 5 ml sterilized 3.8% citrate solution and 0.5 ml antibiotic solution.
The final  volume was 6 ml  of concentrate.
                                   72

-------
2.  Cellulose Nitrate Membrane Filtration - (according  to  Berg  et al   ).
A 5 liter sample was passed through a cellulose nitrate membrane  filter
(Mi Hi pore Corp) at a negative pressure of about 700 mm Hg.   Each
filter was then homogenized in a Sorval Omni-Mixer (# 17106)  at high
speed for 5 min with 10 ml  of a 3% Difco beef extract solution  and
sonicated.  The suspension  was centrifuged at 250 x g for  5 min and
decanted.  One ml antibiotic solution was added to the  supernatant.
The final volume was 12 ml  of concentrate.

3.  Aluminum Hydroxide Precipitation - (according to Wall is and Mel nick  )
Aluminum hydroxide suspension (12 ml) was added to a 5  liter  sample.
The mixture was stirred gently with a magnetic stirrer  for 60 min at
room temperature and then filtered by positive pressure through a
Sartorius filter.  The precipitate was trapped on the filter, recovered
with a spatula and suspended in 5 ml saline containing  0.5 ml anti-
biotic solution.  The final volume was 5.5 ml of concentrate.

4.  Phase Separation - (according to Shuval et al  ).  To  a  5 liter
water sample, the following compounds were added:  10 gr sodium dextran
sulfate 500 (Pharmacia, Sweden), 322 gr polyethelene glycol  (Carbowax
6,000) and 105 gr sodium chloride.  The mixture was stirred with  a
magnetic stirrer, transferred to a separatory funnel and kept overnight
at 5 C.  The lower- and interphase were removed; their combined volume
averaged 35 ml.  To each milliliter, 0.3 ml of 3 M KC1  was added  and
the solution centrifuged at 15,000 x g for 30 min.  One ml antibiotic
solution was added to every 10 ml of supernatant.  The final  volume
was 35 ml of concentrate.
                                   73

-------
5.  Flow-through Gauze Sampler - A 5 liter water sample was filtered
through the gauze sampler at a positive pressure of about 0.1 atmosphere.
After filtration, the gauze pad was removed from the sampler with the
aid of tweezers and transferred into a plastic bag.  The pH was adjusted
to 8.0 with a 1% NaOH solution.  The water in the pad was measured
and calf serum was added to a final concentration of 5%.  The pad was
then hand-pressed 5-7 times.  The pad, placed in a specially designed
                                                   Q
polypropylene insert, as described by Fattal et al,  with a perforated
bottom, was placed in a standard 250 ml centrifuge tube (see Section VII
in this report) and centrifuged at 4,000 x g for 10 min.  Nearly all
of the water content in the pad could thus be recovered.  The dry pad
was replaced in the plastic bag and 20 ml saline, pH 8.0, containing
5% calf serum were added.  The processes of absorption, expression
and centrifugation were repeated to obtain a second eluate; the
liquid from this elution was mixed with that of the first one.  The
average volume of the combined eluates was 82 ml.  One ml antibiotic
solution was added to every 10 ml of eluate.  The final volume was
90 ml of concentrate.

The five different concentrates and the diluted control virus were
kept at 4°C and assayed the following day, or kept at -70°C until
assayed.  Nearly the entire volume of each concentrate  using the
first three methods was assayed.  However, only about one third of
the volume of the concentrates of the fourth method and about one
quarter of the fifth method were assayed.
                                   74

-------
Test Procedure B -
Twenty liters of tapwater were dechlorinated, pH adjusted, prefiltered
and virus added, as described in test procedure A.   The sample was
divided into four equal aliquots, each of which was treated by one of
the following methods:

1.  Aluminum Hydroxide Precipitation - as described in test procedure A.
                                                                  1 o
2.  Cellulose Nitrate Membrane Filter - (according to Wall is et al  ).
A 5 liter sample was adjusted to pH 3.5 with HC1 and then A1C13 was
added to give a final concentration of 0.0005 M AlCI-j.  The sample was
filtered through a 47 mm cellulose nitrate membrane filter (0.45p pore
size).  The membrane was then washed with 5 ml of saline at pH 7.0
and then eluted with 5 ml of 0.05M glycine buffer at pH 11.5.  Five
ml of glycine buffer, 0.05M pH 2 neutralized the previous buffer.  One
ml antiobiotic solution was added.  The final volume was 11 ml of
concentrate.

3.  Insoluble Polyelectrolytes (PE 60) -  (according to Wall is and Melnick  )
A 5 liter sample was adjusted to pH 5.0 with HC1 and then filtered through
polyelectrolyte layers of 47 mm diameter.  The PE 60 pad  layers were
prepared by  adding 2 ml of a 10% suspension  of PE 60 (prepared as de-
scribed by Wallis  et al7) to 300 ml of H20 stirred with the aid of a
magnetic stirrer for 5 min and filtered by negative pressure  through  an
AP 20 pad.   The layer  contained  100 mg of PE 60.  After filtration,  the
virus which  was adsorbed  to the  layer, was eluted with 10 ml  of 0.05  M
borate buffer pH   9.0 containing  10% bovine calf  serum.
                                     75

-------
One ml of antiobiotic solution was added.   The final volume was 11  ml
of the concentrate.
4.  Cellulose Nitrate Membrane Filter - (eluting method according to Rao
and Labzoffsky14).   A 5 liter sample was adjusted to pH 3 with HC1  and
then filtered through a 47 mm cellulose membrane 0.45y pore size.  Ten
ml of a sterile 3% solution beef extract pH 8 was added to the membrane
to elute the virus.  One ml of antiobiotic solution was added to
eluate.  The final  volume of the concentrate was 11 ml.

Concentration of Attenuated Poliovirus I By Gelman Membrane Filters
In addition to a comparison of different concentration methods, some
experiments were carried out with the aid of Membrane Cartridge filters
supplied by Gelman,  using water samples of between 160-360 liters in
40 liter aliquots.  Each 40 liter sample was dechlorinated and the pH
adjusted to 3.5 with concentrated HC1.  The virus stock was diluted
in M-199 medium and added to the water sample until the desired
concentration was  attained.  The sample was filtered through the
Cartridge filter by a positive pressure of 0.5-1.5 PSI.  The flow
rate was 1.2 liter per min.  The virus was eluted from the cartridge
by adding 80 ml of 0.05 M glycine buffer at pH 11.5. The pH was then
immediately adjusted to 7.0 with an HC1 solution.  Eighty ml of 0.05 M
glycine   buffer at pH 11.5 were again added to the filter for a
second and third elution.  The pH was adjusted to 7.0 with HC1.
One ml of antiobiotic solution was added to each 9 ml of eluate.  The
final volume of the three eluates were 300 ml.
                                  76

-------
RESULTS
Recovery of Attenuated Poliovirus I
The results of the recovery efficiency of attenuated poliovirus  I  in
5 liter water samples are summarized in Tables 1 ,  2 and 3.   Table 1
shows the recovery for four methods described in test procedure  A  of
virus concentrations at low input levels (8-64 pfu/5 liters).  Table  2
describes the recovery for five methods from test procedure A  with an
input of 390-790 pfu/5 liters.   Table  3 shows the recovery for methods
described in test procedure B,  with virus inputs of 40-827 pfu/5 liters.

There are only slight differences among the recovery efficiencies  of
the first four methods, while that of the gauze pad is extremely low
(about 1%).  It should be pointed out that the results for the pauze
pad method were not presented in Table 1  because of its very low
concentration factor and recovery.  By employing this method,  viruses
would not be detectable when found in such low concentrations  as were
used in experiments in Table 1.

Table 1 clearly shows the high recovery of the alginate filter and of
the aluminum hydroxide method.   A recovery of over 100% in each
experiment is particularly noticeable, with an average of 247% for the
alginate and 213% for the aluminum hydroxide method.  The average recovery
with the cellulose nitrate method,according to Berg et al   ,was 63%.
The phase separation method shows an average recovery of more than
100% while that of 5 out of 7 experiments was more than 80%.  With all
four methods, 8 pfu/5 liters could be detected.
                                    77

-------
Table  1.  THE RECOVERY EFFICIENCY OF FOUR METHODS FOR THE CONCENTRATION



           OF SEEDED ATTENUATED POLIOVIRUS I AT LOW INPUT LEVELS IN



           TAP WATER
Exp.
no.
1
2
3
4
5
6
7
Ave.
S.D.
Calculated
total
poliovirus
input
pfu/5 L
8
10
15
16
27
35
64
8-64

% Virus recovered
Alginate
filter
160
243
226
268
464
266
101
247
+113
Cellulose
nitrate MF
(Berg et al10)
18
100
63
71
72
70
45
63
+26
A1(OH)3
111
263
105
236
284
366
127
213
+101
Phase
separation
169
43
41
242
123
85
81
112
+ 73
                                  78

-------
Table  2.  THE RECOVERY EFFICIENCY OF FIVE METHODS FOR THE CONCENTRATION
           OF SEEDED ATTENUATED POLIOVIRUS I AT HIGH INPUT LEVELS IN
           TAP WATER
      Calculated
      total
      poliovirus
Exp.  input
no.   pfu/5 L	
                               % Virus  recovered
                        Cellulose
             Alginate   nitrate MF         .   .      Phase     Gauze
              filter    (Bergetal'u)    M|1UH'3  separation    pad
1      391
2      567
3      792
                 44
                 86
                165
             36
             58
            107
               93
               89
              176
           48
           83
           89
           0.4
           1.0
           1.5
Ave.
S.D.
390-790
 98
+61
 67
+36
119
+49
 73
+22
 0.97
+0.55
                                   79

-------
 Table  3.  THE RECOVERY EFFICIENCY OF FOUR METHODS FOR THE CONCENTRATION



            OF SEEDED ATTENUATED POLIOVIRUS I  AT VARIOUS INPUT LEVELS  IN



            TAP WATER

Exp.
No.
1
2
3
4
5
6
7
8
9
10
Ave.
S.D.
Calculated
total
poliovirus
Input
pfu/5 L
40
54
101
101
153
180
183
320
322
827

40-887


A1(OH)3
65
30
70
99
53
30
-
23
23
60
50
+25

% Virus
MFa
(Rao et al)
93
57
101
89
167
62
36
29
65
67
77
+40

recovered
MFb
(Wall is et al)
53
48
119
57
29
107
56
51
44
83
65
+29


PE 60
83
65
69
55
144
74
40
40
53
76
70
+30
a
  MF = Cellulose nitrate-elution according to the Rao et al method
                                                                  14
b MF = Cellulose nitrate according to Wall is et al
                                    80

-------
Table 2 indicates an average recovery of about 100% with the alginate
filter and aluminum hydroxide as compared to the average recovery of
about 70% with cellulose nitrate and phase separation.  The gauze pad
method shows a very low recovery of about 1%.

Table 3 shows that the recovery efficiency of MF according to Rao et
   14
al,   is higher than the others.  The average recovery with A1(OH)3
was 50%.        The MF Rao method shows an average recovery of 77%
while that of the MF Wall is et al method12 was 65% and PE 60-70%.

The Effect of Sample Volume on  Recovery Efficiency of the Gauze
Pad Sampler
To obtain a better understanding of the mechanism by which the gauze
sampler concentrates viruses, the effect of  sample volume on attenuated
poliovirus I was tested.  As can be seen in  Table 21, Section VII, virus
recovery  is dependent on sample  volume.  In  quantities of 50 liters, recovery
averaged 0.5%, and in volumes of 700 ml it was 7%.  The larger the
volume, the lower the efficiency of the pad.

Table21 shows the sharp decrease in recovery efficiency relative to the
sample volume in each experiment.  On the other hand, the concentration
factor of the gauze is obtained  by dividing  the pfu/ml of gauze  fluid
after filtration  by the pfu/ml  of the  control fluid  before filtration.
This factor increases in direct  proportion to the sample volume:   the
factor is 1.2 for 700 ml versus  6.6 for 51,500 ml.  The experiments
                                     81

-------
were performed simultaneously.  Each gauze pad was eluted 4 times and
centrifuged; the recovery efficiency is the total of these four elutions,

Recovery of Echovirus 7
Since poliovirus is only one of the various viruses found in sewage, an
additional enterovirus present in sewage, echovirus 7,was also tested.
Table 4 shows the results of virus recovery by the five methods de-
scribed in test procedure A for concentration of seeded echovirus 7 at
various input levels (5-2,000 pfu/5 liters).

The alginate filter and phase separation methods show higher recovery
than the other methods (average 101% and 125%, respectively).
Aluminum hydroxide shows slightly higher results than the cellulose
nitrate filter method.  It is worthwhile to mention that in Experiments
1 and 2, three methods were successful:  alginate, A1(OH)3 and phase
separation, and we were able to detect 5 and 7 pfu/5 liters of water
at good recovery efficiency.  The gauze pad shows again an extremely
low recovery, even with high virus concentration.

Recovery of Enteroviruses from Sewage Contaminated Water
A comparison of virus detection methods using laboratory virus strains
is not complete unless supported by experiments with natural wild
viruses usually found in wastewater.  Eleven experiments, 4 performed
in test procedure A, and 7 in test procedure B, were conducted in which
raw sewage was added to tapwater samples.  Two to 20 ml of domestic
raw sewage were added to 20 liters of dechlorinated filtered tapwater,
                                   82

-------
Table 4    THE RECOVERY EFFICIENCY OF FIVE METHODS FOR THE CONCENTRATION



           OF SEEDED ECHOVIRUS 7 AT VARIOUS INPUT LEVELS IN TAP WATER
Exp.
no.
1
2
3
4
5
6
7
8
9
Ave.
S.D.
Calculated
total
virus input
pfu/5 L
5
7
17
21
207
323
478
783
2,000
5-2,000
% Virus recovered
Alqinate
f 1 1 ter
202
59
128
97
215
69
45
45
50
101
i 67
Cellulose
nitrate MF ,n
(Berg et allu)
< 8
< 7
72
48
35
8
16
11
7
22
+25
A1(OH)3
24
32
23
129
75
73
52
60
20
54
+35
Phase
separation
177
112
127
239
176
74
80
68
79
125
+ 60
Gauze
pad
N.D.a
N.D.
N.D.
N.D.
N.D.
1.5
< 0.6
< 0.3
0.1
^ 0.6
  not done
                                   83

-------
  adjusted to pH 7.0.   Four virus test procedures were then carried out.
  The results are presented in Tables 5 and 6.

  Table 5 shows an average recovery in all  four methods of less than 50%,
  with a slight advantage, in most experiments, for alginate and aluminum
  hydroxide methods.   The lowest recovery is found with the Phase Separation
  method in all experiments.  Table 6 shows a high recovery efficiency of
  MF according to Rao et al (average of 111%).  On the other hand, the
  average recovery efficiency of PE 60 is 80% compared to 67% according
  to the MF Wall is method.  An average of 40% was obtained with A1(OH)3.

  Table 5.   RECOVERY EFFICIENCY OF FOUR METHODS FOR THE CONCENTRATION OF
             ENTEROVIRUSES FROM SEWAGE CONTAMINATED WATER
Dilution Sewage
of sample control

Exp.
no.
1
2
3
4
Ave.
S.D.
with
sewage
%
0.01
0.02
0.02
0.03
-
before
dilution
pfu/ml
45
27
35
38
-
Calculated
total input

enterovi ruses
pfu/5 L
23
27
35
57
23-57
Alg.F.
26
74
15
72
47
+31



% Virus recovered
a MFb
14
5
14
26
15
+ 9
A1(OH)3
11
32
31
60
34
+20
phase
separation
8
N.D.C
10
5
5
+5
a Alg.F = Alginate filter
b MF = Cellulose nitrate, according to Berg et al
c N.D. = not done
                                      84

-------
Table   6.  RECOVERY EFFICIENCY OF FOUR METHODS FOR CONCENTRATION OF
           ENTEROVIRUSES FROM SEWAGE-CONTAMINATED WATER



No.
1
2
3
4
5
6
7
Ave.
S.D.
Dilution
of sample
with
sewage
%
0.010
0.015
0.025
0.025
0.1
0.1
0.1


Sewage
control
before
dilution
pfu/ml
6
9
11
6
4
16
24


Calculated
total input
entero-
vi ruses
pfu/5 L
3
7
12
7
19
82
118
3-118





A1(OH)3
67
43
--
29
32
21
48
40
+17
I Virus
a
MFa
(Rao
et al
method)
167
114
242
57
21
101
75
111
+74
recovered
h
MFD
(Wall is
et al
method)
167
71
108
29
11
45
36
67
+54




PE 60
--
71
142
71
68
71
54
80
+31
   MF = cellulose nitrate-elution according to Rao et al
                                                    12
   MF = celluslose nitrate according to Wallis et al
 Recovery of Seeded Attenuated Poliovirus I by Cartridge Filter
 In order to concentrate and detect seeded virus from large volumes of water,
 we decided to test also Membrane Cartridge filters from Gelman Company.  By
 this method we succeeded  in filtering up to 360 liters of tapwater.
 Seven  preliminary experiments were carried out with the use of 2 kinds of
 filters:  six experiments with  a 3y  pore size filter and one experiment
 with a 0.22y pore size filter.  The  results are summarized in Table  7.
 From this table  it can be seen  that  the recovery  efficiency of the 3u
                                       85

-------
filter is between 27-168% compared to 36% for the 0.22y filter.   The
                                               2
lowest number of seeded poliovirus was 3.8 x 10  pfu/360 liters,  i.e.
approximately 1  pfu/liter with a recovery efficiency of 58%.

Table 7.   RECOVERY EFFICIENCY OF THE GELMAN CARTRIDGE FILTER FOR THE
           CONCENTRATION OF SEEDED ATTENUATED POLIOVIRUS I  AT VARIOUS
           INPUT LEVELS IN TAPWATER.
Exp.
no.
1
2
3
4
5
6.
7
Filter
no.
a 12104
12104
12104
12104
12104
12104
b 12106
Sample volume
filtered
(L)
160
200
200
200
360
360
200
Total poliovirus
input
pfu
1.1 x 107
9.2 x 106
1.3 x 106
1.1 x 106
9.5 x 102
3.8 x 102
1.0 x 106
% Virus
recovered
27
45
168
75
47
58
36
a 12104 = pore size 3y
b 12106 = pore size 0.2y
DISCUSSION
The number of enteroviruses detectable in sewage is considerable;  their
concentration changes from month to month and from one community to
another.  With direct inoculation tests carried out in our laboratory
on Jerusalem domestic sewage, up to 95,000 pfu viruses/liter were
found.  This means that the average total number of viruses in the
                                    86

-------
sewage flow from Jerusalem, with a population of 250,000,  might be
                    12
approximately 3 x 10   pfu/day.   In spite of the biological,  chemical
and physical treatment methods and the various disinfection processes,
viruses may at times be present in water supplies, albeit  in  small
numbers.  Contaminated rivers will most likely contain viruses in
various concentrations during most of the year.  The quantity of
water actually ingested by the public in cities with populations of
one million and more is in the range of thousands of cubic meters per
day.  A small number of viruses that pass through the treatment barriers
may therefore constitute a potential health hazard, particularly if
we bear in mind that even one pfu of virus is capable of producing
infection in susceptible hosts.

From the epidemiological point of view, it is important that sensitive
methods for the detection of small quantities of virus in large
volumes of water be developed.  This would allow the use of a viral
indicator as a test for pollution in routine monitoring of water.  Such
a system is particularly important in view of the fact that under
certain conditions the conventional bacterial indicators of pollution
may be  absent in treated water which nevertheless can still carry
the more resistant enteroviruses.  These virus detection methods
should  be inexpensive and simple  enough to allow their use in most
routine water monitoring laboratories, and should be able to detect
one virus in a relatively large volume of water, preferably more
than 100 liters.
                                  87

-------
Many methods have been developed during the last few years.  Several
                123
research workers      have reviewed and compared different methods for
detecting viruses in water.  They compared various parameters, such
as the concentration factor, recovery efficiency and time required
for each method.  However, few controlled comparison studies of methods
have as yet been described.  The present study tries to fill this gap.
Eight methods, used today, for the concentration and detection of
small quantities of viruses in water were compared. Two virus strains
(attenuated poliovirus I and echovirus 7) and enteroviruses from
sewage were employed.

Table  8 summarizes Tables  1,  2,  4  and  5.   It  can be  seen  that  the
first four methods provide relatively good recovery of seeded poliovirus
and echovirus in water samples as compared to the poorer recovery
obtained from samples of diluted sewage.

This finding can possibly be explained by the selectivity of the various
concentration methods which may not concentrate all virus types with
equal efficiency.  The virus concentration of the sewage samples used
in calculating recovery efficiency was made by the direct inoculation
method, which leads to the least possible loss of viruses.  Grindrod
et al   have already shown that the phase separation method concentrates
viruses selectively.  As for the MF method according to Berg et al   and
the aluminum hydroxide method, it appears possible that the organic
matter in the sewage may compete for adsorption sites, thus reducing the
efficiency of  these methods when used with contaminated water.

-------
Table  8.  THE AVERAGE RECOVERY OF SEEDED ENTEROVIRUS AND THE

           CONCENTRATION FACTOR OF FIVE METHODS WITH TAP WATER

Virus
Polio-
Virus I
Echo-
virus 7
Entero-
virus
from
sewage
Average
Average
factor

pfu/5 L

8-790

5-2000



23-57
No.
of
exp.

10

9



4
final volume
concentrati

on


Alq.F.9

202

101



47
6

850
% Average
virus
MFb A1(OH)3

64

22



15
12

425

185

54



34
6

850
recovered
Phase
separation

100

125



5a
35

140

Gauze
pad

1.0°
H
o.r



-
90

55
 3  Alg.F.  =  alginate filter

    MF = cellulose nitrate according to the Berg  et al  method
 c
    = 3 experiments

    = 4 experiments
                                    89

-------
The average recovery of the alginate filter, the aluminum hydroxide
method and the phase separation method was high with poliovirus (100%
and over), as compared to the MF cellulose nitrate method of Berg et
al (64%), and the gauze pad (only 1%).  The average recovery of the
alginate filter and phase separation methods were high with echovirus 7
(more than 100%), while the MF cellulose nitrate method (Berg et al)
and aluminum hydroxide methods gave relatively low recoveries (22%
and 54%, respectively).  The gauze pad was again extremely low (^0.5%).
The average recovery of enteroviruses from sewage was low with 4 methods
and very low with the phase separation (5%) and MF cellulose nitrate
(15%) methods (Berg et al).

Comparing our results with those of other workers, we find that Gartner
succeeded in recovering 25% of 106 pfu/10 liter and 100% of 100 pfu/10 liter
with an alginate filter, which is similar to our findings for low
concentrations of poliovirus.

Although there were differences in the technique used for MF cellulose
»
nitrate (Berg et al), our results with poliovirus and echovirus 7
correlate well with those obtained by Berg et al.    Comparing the
results of the aluminum hydroxide method with those of Wallis and
Melnick,  the latter recovered about 80% of 249 pfu / liter, using
poliovirus, while we found an average of 50% (Table  9) to 185%
(Table  8).  The recovery with the phase separation method in this
study correlates well with the results obtained by Shuval et al.
                                   90

-------
The recovery of the gauze pad was extremely low compared  to  the  other
four methods, which is in keeping with the results  reported  by Hoff  et
  17               18
al   and Liu et al.    Furthermore, recovery efficiency becomes  lower as
sample volume increases.   It follows that the method is unsuitable for
quantitative virus estimation.
Table 9  summarizes Tables 3  and 6 .   It can be seen  that  the MF  method
                      14
according to Rao et al    and PE 60 method provide good recovery  of seeded
poliovirus as well  as diluted sewage:   70% and over on the  average.   The
                                   12
MF method according to Wall is et al   provides an average recovery of 65-
67%, in both seeded poliovirus and diluted sewage, while with PE 60 the
average recovery is 70-80%.   A comparison of the data  from  the 8 methods
presented in Tables 8  and  9,  clearly shows the advantage of the three
methods:  MF-Rao,  MF-Wallis and PE 60—for both seeded and diluted
sewage.  The elution of the  virus from MF according to Rao  method  seems
to be superior to the others.  Our results correlated  with  those
obtained by Rao et  al,14 MF  Wallis et al12 and with Wallis  et al.13
In all experiments, the concentration factor of the eight  methods
ranged from 55 to 850.  However, the MF,  PE 60 and one  variation of
the aluminum hydroxide method can achieve concentration factors  higher
than 1,000.

For technical  reasons, the alginate and phase separation methods are
suitable for concentrations of not more than 10 liters  per sample.
                                    91

-------
The first method required 6 hours (at 3 atmospheres pressure)  and the
second 18-24 hours.  On the other hand, the MF cellulose nitrate, the
aluminum hydroxide and gauze pad methods can be used for filtering large
volumes of water, 100 liters and more.


Table  9.  THE AVERAGE RECOVERY OF SEEDED ENTEROVIRUS AND THE
           CONCENTRATION FACTOR OF FOUR METHODS WITH TAP WATER.

Virus
Polio
virus I
Enterovirus
from sewage
No.
of
pfu/5 L exp.

40-827 10

3-118 7
Average of final volume
Average of
concentration

factor
%

A1(OH)3

50C

w
6

850
Average

MFa
Rao et

77

111
11

450
virus recovered

MFb
al Wall is et al

65

67
11

450


PE 60

70

80
11

450
 MF = cellulose nitrate-elution according to Rao et al
 MF = cellulose nitrate according to Wallis et al
c = 9 experiments
  = 6 experiments

All the methods described here are simple and not expensive to employ.
For the alginate, MF and Al(OH)., methods, pure water, free of suspended
particles, is required, while with the phase separation, the gauze pad
method and PE 60 contaminated water can be used as well.
                                   92

-------
Recovery of more than 100% was obtained in many experiments.   This
phenomenon may be explained by the presence of virus clumps in the
sample, which may have been disaqgregated during the concentration
         19 20
procedure  '   and is very similar to the effect of diluent upon
                                                        21
enumeration of poliovirus as described by Hamblet et al.     They
found that nutrient broth when compared with Hank's balanced  salt
solution and an assaying diluent gave four to six times more  plaques
under similar conditions.  They postulated that nutrient broth
promotes disaggregation of virus clumps rather than enhances  host
cell susceptibility.

The discrepancies found between the recovery efficiency using the
Al(OH)., in the first and second stage tests may be explained by the
fact that the experiments in the second stage were carried out two
years after the first stage experiments.  During this period the
stock  polio  virus which  had  been  kept  at  -20°C  probably  under-
went aggregation thus in turn bringing about poorer recovery
efficiency at the second stage.  On the other hand, there were no
significant differences in recovery efficiency between the 2  stages
for the Al(OH)- concentration of the diluted sewage.

The experiments carried out with the Gelman membrane cartridge filter
indicate that the filter affords a good recovery efficiency,  and is
suitable for filtering large volumes (up to 360 liters) of water.
                                 93

-------
Hill et al22 who also used cartridge filters, though larger,  succeeded
in filterina 400 liters of water and recoverd 1-2 pfu poliovirus/20 liters
water.  The above group used a combined technique, i.e.,  membrane
cartridge filter adsorption followed by an aqueous polymer two steps
phase separation method according to Shuval  et al.    In  conclusion,
the cellulose nitrate membrane filter and the aluminum hydroxide
methods are preferable for volumes of about 100 liters of water, while
cartridge filters are preferable for 400 liters.   Clean and organic-
matter free water is a prerequisite when using these methods.  Large
volumes of unclean water may be used with the PE  60 method which is
suitable for filtration of large volume of water  containing organic
matter.  This certainly applies to drinking  and  renovated water.
However, for samples of contaminated water, the alginate and aluminum
hydroxide methods can be used effectively provided smaller water
samples are used to avoid rapid clogging of the filters.

Although many improvements in virus detection methods can be
anticipated, we feel that these results imply that even at the present
stage of development, routine monitoring of drinking water is feasible
and can provide a considerable degree of public health protection.
                                   94

-------
REFERENCES
1.  Hill, W.F., W.E. Akin, and W.H.  Benton.   Detection of Viruses in
    Water:  A Review of Methods and  Application.   Water Res.  5:967-995,
    1971.

2.  Shuval, H.I., and E. Katzenelson.   The Detection of Enteric Viruses
    in the Water Environment.  In:  Water Pollution Microbiology.,
    Mitchell, R. (ed.), New York, Wiley and Sons, 1972.  p.347-361.

3.  England B.  Recovery of Viruses  from Waste and Other Waters by
    Chemical Methods.  In:  Symposium on Detection of Viruses in Waste
    and Other Waters, Berg., G. (ed.).  Developments in Industrial
    Microbiology, 1975, Vol. 5.  p.  174-183.

4.  Shuval, H.I., A. Thompson, B. Fattal, S. Cymbalista, and  Y. Wiener.
    Natural Virus Inactivation Processes in Seawater.  J. San.  Eng.
    Div. 97:587-600, 1971.

5   Barron, A.L., C. Olshevsky, and M.M. Cohen.  Characteristics of
    the BGM Line of Cells from African Green Monkey Kidney.  Archiv.
    fur die gesamte Virusforschung.  32:389-392, 1970.

6.  Dahling, D.R., G. Berg, and D. Berman.  BGM, a Continuous Cell
    Line More Sensitive than Primary Rhesus and African Green Kidney
    Cells for the Recovery of Viruses from Water.  Health Laboratory
    Sciences. 11:275-282, 1974.
                                 95

-------
 7.   Wallis,  G.,  and  J.L.  Melnick.   Concentration of Viruses on Aluminum
     Phosphate  and  Aluminum  Hydroxide  Precipitates.  In:  Transmission
     of Viruses by  the  Water Route,  Berg, G.  (ed.). New York, Wiley
     and Sons,  1967.  p.  129-138.

 8.   Fattal,  B.,  E. Katzenelson,  and H.I. Shuval.  Comparison of Methods
     for Isolation  of Viruses  in  Water.   In:   Virus Survival in Water
     and Wastewater Systems, Malina, J.F., Jr.  and B.P. Sagik, (eds.).
     University of  Texas  at  Austin,  Center for Research in Water
     Resources, 1974.   p.  19-30.

 9.   Gartner, H.  Retention  and Recovery  of Polioviruses on a Soluble
     Ultrafilter.   In:   Transmission of Viruses  by the Water Route.
     Berg, G. (ed.).  New York, Wiley  and Sons,  1967.  p. 121-127.

10.   Berg, G.,  D.R. Dahling, and  D.  Berman.   Recovery of Small
     Quantities of  Viruses from Clean  Waters  on  Cellulose Nitrate
     Membrane Filters.   Appl.  Microbiol.  22:608-614, 1967.

11.   Shuval,  H.I.,  B. Fattal,  S.  Cymbalista and N. Goldblum.  The
     Phase Separation Method for  the Concentration and Detection of
     Viruses  in Water.   Water Res.  3:225-240,  1969.

12.   Wallis,  G.,  M. Henderson, and  J.L. Melnick.  Enterovirus
     Concentration  on Cellulose Membranes.  Appl. Microbiol.
     23:476-480,  1972.
                                  96

-------
13.   Wall is, G., and J.L.  Melnick.   Detection of Viruses  in  Large  Volumes
     of Natural  Waters by  Concentration on Insoluble Polyelectrolytes.
     Water Res.  4:787-796, 1970.

14.   Rao,  N.V.,  N.A. Labzoffsky.   A Simple Method for the Detection  of
     Low Concentrations of Viruses  in Large Volumes  of Water by the
     Membrane Filter Technique.   Canadian Jour,  of Microbiol.
     15:399-403, 1969.

15.   Plotkin, S.A., and M. Katz.   Minimal Infective  Doses of  Viruses
     for Man by Oral Route.   In:  Transmission of  Viruses  by Water Route,
     Berg, G. (ed.).  New York, Wiley and Sons,  1967. p.  151-166.
16.   Grindrod, J. and D.O. Cliver.   Limitation of Polymer Two Phase
     System for Detection  of Viruses.  Arch Ges. Virusforsch.
     28:227-247, 1969.

17.   Hoff, J.C., R.D. Lee, and R.C. Becker.  Evaluation of Methods for
     Concentration of Microorganisms in Water.  Water Hyg. Div. p.  37,
     1971.

18   Liu,  O.C.,  D.A. Brahsear, H.R. Seraichekas, J.A. Barnick, and T.G.
     Metcalf.  Virus in Water:  A Preliminary Study  on Flow-through  Gauze
     Sampler for Recovering Virus from Water.  Appl. Microbiol. 21:405-410,
     1971.

19.   Galasso, G.L. and D.G. Sharp.   Virus Aggregation and the Plaque-
     Forming Unit.  J. Immunol.  88:339-347, 1962.
                                    97

-------
20.  Gerba, C.P. and G.E.  Schaiberger.   Aggregation as  a  Factor in
     Loss of Viral  liter in Seawater.   Water Res.  9:567-571,  1975.

21.  Hamblet, F.E., W.F. Hill  Jr., and  E.W.  Akin.   Effect of  Plaque
     Assay Diluent  Upon Enumeration of  Poliovirus  Type  I.  Appl.
     Microbiol.  15:208, 1967.
22.  Hill, W.F., E.W.  Akin, W.H.  Benton,  and T.G.  Metcalf.  Virus
     in Water:   Evaluation of  Membrane  Cartridge  Filters  for
     Recovering  Low Multiplicities of Poliovirus  from Water.  Appl.
     Microbiol.  23:880-888, 1972.
                               98

-------
                              SECTION VI
           VIRUS CONCENTRATION USING HOLLOW FIBER MEMBRANES

INTRODUCTION
The purpose of this study was to determine the ability of cellulose
acetate hollow fibers, commonly used as an ultrafiltration membrane
for biological applications  to concentrate viruses.   The well  known
advantages associated with the fiber membrane permeators include
compactness, i.e. very high active membrane surface area per unit
volume of permeator, light weight, low water hold-up and low
     2
cost.   Virus concentration has been successful using flat ultra-
                               3 4
and hyper-filtration membranes. '  The advantages of the method as
compared to others are no need to adjust pH and lack of interference
by organic matter or salinity.  It is the compactness, light weight
and low water hold-up of the fiber unit that renders it attractive
for the first (concentration) step in the detection of virus from
large volumes of water.  Once the viruses are concentrated into
small volumes, the second step (virus enumeration) can be used to
quantitate virus levels.

In this section, we present the results of an experimental study to
concentrate viruses using a hollow fiber membrane permeator.  The
effect on permeator performance of the following independent variables
direction of flow through the fiber, feed flow rate, initial virus
                                  99

-------
concentration in the feed, cell  type, and backwashing was studied.

MATERIALS AND METHODS
Apparatus
A commercially available hollow fiber membrane permeator (Bio-Fiber 80
Mini-plant from Bio-Rad Laboratories, Richmond, California, 94804,
U.S.A.) was used.  The dimensions of the unit and fibers with the flow
sheet are presented in Fig. la.   The hollow fiber membranes were made
of cellulose acetate with a nominal  surface area of 10,000 sq. cm.
and a nominal molecular weight cut-off of 30,000 (Dow product designation
C/HFU-10).  The feed, seeded with polio-type 1 virus was pumped (Cole-
Palmer Model 7015) through the core of the fibers at a linear flow rate
of 9.1 and 3.16 ml s  , and continuously recycled.  The permeate,
which passed radially from the inside to the outside of the fibers
was drawn into a buchner funnel by a vacuum pump (Millipore XX60 220
50) operating at 25  in Hg  (0.836 atm).  Thus, the driving force for
permeation was about  1 atm, since a  small positive  pressure was also
exerted  on the feed  side.
Cell  Cultures and Medium
Two continuous lines  of African green monkey  kidney cells, Vero, (Flow
Laboratory,  Ltd., Scotland) and BGM  (kindly supplied  by  Dr. G. Berg)
were  used.   The  cells were grown on  tissue culture  plastic plates
(Nunc, Denmark).  The medium, Eagle's minimum essential  medium, con-
tained 10 per cent  foetal  bovine serum  and was  incubated at 37°C in
90-95 per cent relative humidity and 5% C02 atmosphere.
                                   100

-------
                             HOLLOW FIBERS
                   HOLLOW  FINE  FIBERS
                                                                 CONCENTRATED
                                                                     OUT
                        MINIPLANT  ULTRAFILTER
                                   PERMEATE RESERVOIRS
VIRUS ' FEED
RESERVOIR
              (b)  FLOW SHEET
             Fig. 1.  Detail of fiber unit and system flow sheet
                                  101

-------

T 25.0
O
X
i
E
20.0
z>
u.
Q.

O
LU
S 15-0

z

•^
o
< lo-°
o:
z
UJ
o
z
O 50
0 °'U
D
(£
-» i n
o




^_ \ FLUX 5
X X*NN ./x x'x«— / r»
'^•.....'•x XV !;
i *
o
1
1
1
1
f
1
1
o
1
FEED/
/
o'

ff
s A
X ^
'*°*' ^s'
- -"A-^-^^'PERME ATE
a o _Qfcf?" **
f 1 i 1 i 1
       432

       VOLUME   OF  FEED    V


Fig. 2.  Result of a typical  virus concentration experiment
                                                  100  x


                                                       Lu



                                                  80  g-

                                                       fe  'c
                                                       LJ  c

                                                       K

                                                  60  g  i
                                                        2  E
                                                  200^  D
                                                        cc  t
                                                  100  g£
                                                        o <
                                                        a) a:
                                                        D LU
                                                  0    £ ^
                     102

-------
                                   Table 10. RESULTS OF BACKWASH EXPERIMENTS
o
co
Exp.
no.
14

15

16

17

19

20

21

22

Initial
virus count3
pfu
7.415 x 106

2.335 x 106

2.265 x 106

7850

5550

5.165 x 106

9.165 x 106

235

Percent virus received in aliquots of permeate(%/ml )
12345
34.5%
130 ml
53.8%
100 ml
26.9%
100 ml
19.1%
50 ml
22.3%
90 ml
28.2%
50 ml
6.55%
40 ml
54.5%
60 ml
1.7%
140 ml
12.1% -
100 ml
1.9%
100 ml
7.6% -
95 ml
5.6% -
70 ml
10.7% 0%
100 ml 100 ml
4.0% 1.8% 0.8% 0%
40 ml 75 ml 65 ml 20 ml
0% -
110 ml
Total percentage
recovered (%)
36.2

65.9

28.8

26.7

27.9

38.7

13.2

54.5

         aln 5000 ml feed

-------
Virus Assays
Virus assays were made by the plaque-forming unit (pfu) method, using
an overlay medium consisting of Eagle's minimum essential medium, 2%
foetal bovine serum and 1 per cent special agar-noble (Difco Labs.,
Detroit, Michigan, U.S.A.).  Staining was done with Neutral Red.

Experimental Procedure
The required virus concentration (pfu ml  ) was prepared in 5 L  of feed
water and placed in the virus reservoir (see flow diagram in Fig. Ib )•
At time t = 0, both the peristaltic pump and the vacuum pump were
switched on at a preset flow rate and vacuum setting, respectively.
Ten ml samples for virus assays were taken at decreasing time intervals
from both the feed and permeate reservoirs.  The volume of permeate
collected during each of these time intervals was measured and the
average permeation flux (ml min  )  was noted.  The results were then
plotted, as shown in Fig.  2,  after each experiment.
RESULTS
The results for eleven experiments  are presented in Table  10.   The
effect of each independent variable—flow direction (radially) through
the hollow fiber, feed flow rate, initial virus concentration in the
feed, cell type, and virus recovery using washing techniques—on the
performance of the hollow fiber permeator is discussed below.

Radial Flow Direction
One of the major drawbacks 1n using hollow fiber membrane permeators for
                                   104

-------
turbid waters in wastewater desalination is the high rate of flux-
decline(Belfort, 1974).   This has been attributed to the poor
hydrodynamic condition when the feed or turbid water is passed at
high pressures (^0 atm) on the outside of the filaments.  The best
way these inherently weak fibers can withstand such high pressures
is in compression with the feed solution on the outside.

Thus, it would seem logical from the outset, for the experiments
discussed here, to operate the system with the feed solution on the
inside rather than outside of the filaments.  Firstly, because high
pressures are not used here and secondly, well-controlled hydrodynamics
as in tubular flow is essential so as to minimize virus loss.  Another
factor that should be considered is the structural characteristics  of
the membrane fiber wall.  Depending on the method of preparation,
fiber membranes can be made with a dense skin (or active layer) and
an open spongy understructure, according to the original fabrication
method of Loeb and Sourirajan  .  Not all membranes have skins; while
some may have a skin on either the bore surface or outer fiber surface.
Skins are usually less  than 1  per cent of  the total membrane thickness.
From Fig.  3, we see that the  fibers used  in these experiments are
homogeneous and semi-dense throughout without a dense  skin or open
spongy understructure.  The close-up SEM (Cambridge Scientific
Instruments, Model S4-10) of the wall cross-section (Fig. 3b) displays
a  structure consisting  of small  pinholes and nodules  very similar  to
                                  105

-------


Fig. 3.   Scanning electron  micrographs  of the  cross-section of  the hollow fiber membrane.
(a)   Total  fiber cross-section  magnification:  x  418, o.d.-170um; wall thickness-18um.
(b)   Close-up of wall  cross-section  magnification:  x 550, wall  thickness~13ym.
                                          106

-------
that described by Kesting .   The significance of this  semi-dense



structure is discussed later with regard to the backwashinq experiments,
The average permeation rate for about 1  atm vacuum for radial  flow from



inside to outside, and outside to inside of the fiber was 94.8 and 86.2



ml min  , respectively.  Since we wanted to maximize the permeation



flux and minimize virus loss due to poor hydrodynamics, all  experi-



ments were run with the feed solution in the fiber capillaries and



the vacuum on the outside of the fibers  (see Fig.  1).





Feed Flow Rate



From the initial experiments (see Table 10, experiments 7, 11 and 12) we



observed poor virus recovery efficiencies  (< 40/0  and suspected the



well known phenomenon, concentration polarization, to be responsible



for virus build-up and subsequent adsorption onto the membrane



surface.  Concentration polarization can be reduced by increasing the



bulk flow rate  (or Reynolds Number)  inside the fiber thereby increasing



the shear at the solution-membrane interface .  Experiment 14 was



conducted at one third(3.16 ml s~ )  the usual feed operating flow rate

           _i

of 9.1 ml s  .  These  flow  rates correspond to Reynold's Numbers



(NR  = dv/n, where d is the internal fiber diameter, cm, v is the



average internal flow  velocity,  cm s  , and n the kinematic  iscosity


                 2  -1
of the liquid,  cm  s   ) of  about 2.58 and  7.27, respectively.  For



tubular flow both are  in the laminar regime.
                                  107

-------
From Fig. 4 and Table 11, we note that the concentration experiment
can be adequately (certainly within the repeatability of the virus assay
                                                   Q
technique) described by the following relationship.
                               C = AV"R
where C = virus concentration in feed reservoir,
                               PFU ml"1
                               V = volume of the feed solution, ml
                                 - (V0 - Qt)
                               A = preexponent
                                 • CoVoR
                               CQ= initial C, pfu ml
                               VQ= initial V, ml
                               Q = permeation rate,  ml min -1
                               t = time, min
                               R = rejection parameter
                               C = virus concentration in permeate, PFU
                                    ml'1
The rejection parameter R is obtained from the slope of the linear plots
in Fig.4 and describes the ability of a membrane to reject solute (virus).
R  varies from zero (poor rejecting membrane) to one (excellent rejecting
membrane).  From Table 11, of all the experiments listed only Experiment
12 had a significantly different slop R.  The average coefficient of
permeability L  as presented in Table 10, however, is not significantly
                                   108

-------
CNJ 5; in to 50jOOO
a. o. Q. a
X XXX
UJU UIUJ
   u.
   Q.
   O
   UJ
   UJ
   o
   h-
   <

   h-

   LU
   O
   z
   o
   u
          100
            IO.OOO 5,000          1,000   500

                        VOLUME  OF FEED  (ml"1 )
                 Fig. 4.  Concentration of virus in feed versus volume of feed
                                   109

-------
           Table 11.  PARAMETRIC RESULTS FROM CURVE FITTING3



                                  7T
Exp.                                 _-,                      Rejection
no.	pfu ml 	R

12                             303,000                        0.54(1)

14                           1,151,000                        0.76(7)

15                             515,200                        0.78(3)

16                             327,800                        0.76(0)

17                                 775                        0.75(0)


19                                 485                        0.73(6)

a  See Fig. 4
                                  110

-------
different for Experiment 12 than for the other experiments  listed in
Table 11.  For Experiment 14 (slow flow rate), the value of L  is on  the
low end of the range of values reported.
We thus conclude that for the feed flow rates 3.16 and 9.1  ml  s~ ,  no
significant difference with respect to virus rejection was  detectable.
However, decreasing the average feed flow rate slightly reduced the
permeation flux, indicating an increased resistance due possibly to
surface adsorption.  Moreover, the recovery of virus from washing
experiments is of the same order of magnitude for the slow  and fast
feed rate experiments.  (See the difference between the last two
columns in Table 10 for, say Experiments 14 and 20).  Thus,  the large
percentage of virus held-up in the fiber system must be due to some
phenomenon other than only adsorption onto the fiber surface.   A
suggested explanation for this effect will be provided later in
the discussion on backwashing.

Initial Virus Concentration and Cell Type
From Tables 10 and 11 and Fig.  4  , we note that for a range  of three
orders of magnitude of initial feed concentration (Experiments 14,
15, 16, 17 and 19) the R parameter is virtually identical.   Experiment
12, however,  has a feed concentration of 3000 pfuml"  and  shows a  20
per cent drop in the rejection parameter R.  Based on this  result,  all
we can make is a preliminary statement to the effect that from 3000
pfu ml   and  above the membrane rejection is probably adversely
                                   111

-------
affected by the initial virus concentration.  This concentration level
is, however, far above that for sewage samples (45 pfu ml~ ) or
secondary treated effluents (57 pfu ml"  ).
Although  it is known that the same virus concentration will plate about
three times more pfu ml"  for the BGM than the Vero cells, no signifi-
cant difference is observed between the cells used in the assay
technique for the concentration rate experiments presented here (see
Experiments 15 and 16 in Fig.  4).

Backwash ing experiments
From the first few experiments (see Experiments 7, 11 and 12 in Table  10)
it became obvious that over 60 per cent of the virus was disappearing,
either due to inactivation or adsorption on or within the membrane.
The first possibility was evaluated by conducting the concentration
experiment for 75 min without vacuum to see if continual recycle of
the feed solution caused virus inactivation.   No detectable inactivation
or virus loss was observed.  This experiment also indicated that the
virus did not have a propensity to adsorb onto the inner fiber surface.
This second possibility, virus adsorption onto the inner fiber surface,
was also dismissed as a major cause of virus loss by the feed flow
rate experiments.   The third possibility, virus adsorption within the
fiber wall, was tested by drawing a vacuum on the inside of the
fibers and placing virus free water on the outside.  Aliquots of
water that permeated the fibers were collected and assayed for virus.
The results are presented in Table  12.
                                   112

-------
                  Table 12.  RESULTS OF VIRUS CONCENTRATION EXPERIMENTS
Exp.
no.
7
11
12
14
15
16
17
19
20
21
22
Speed
of .
feed3 Cells0
F
F
F
S
F
F
F
F
F
F
F
?
Vero
BGM
BGM
BGM
Vero
BGM
BGM
BGM
BGM
BGM
Feed
concentration
(pfu ml"1)
667
767
3000
1438
467
453
1.57
1.11
1053
1833
0.04(7)
Final
volume ,
of feed Duration
(ml) (min)
130
100
90
115
90
95
90
92
115
130
47
61.5
71.5
58.5
63.5
63.0
59.0
50.5
56.0
54.5
56.5
58.1
Average permeability6 ,V,1rus recovery
[ Without With
o P i -i washing washing
(ml crrf^ h~' atm"1) (%) (%}
5.70 x 10"1
4.93 x 10"1
6.05 x 10"1
5.54 x 10"1
5.62 x 10"1
5.99 x 10"1
7.00 x 10"1
6.31 x 10"1
6.44 x 10"1
6.20 x 10"1
6.13 x 10"1
37.6
32.6
34.2
56.3
44.2
49.3
35.8
54.0
51.1
64.6
24.5
--
--
--
92.5
110.1
78.1
62.5
81.9
89.8
77.8
78.87
.   The feed stream was pumped either fast(F=9.1 ml s~ )  or slow (5=3.16 ml s~ )
   Viruses were assayed on either Vero or BGM cells
   Initial feed volume was 51
"  Final volume before washing
                        _p                   _2-l
e  To convert to (gpd ft  ), multiply (ml cm    h~ ) by 5.8897.  An important parameter is L ,
   water permeability, which is obtained by dividing the average flux by the applied        P
   pressure difference (about 0.8333 atms here).

-------
From Table 12, we see that large percentage  of the  missing  virus  can  be
recovered with relatively small  volumes  of wash water.   Indeed, most  of
the wash-recovered virus is collected in the first  and  second  aliquots
(in about 150 ml).  From the last column in  Table 10, we note, however
that from 10 to 20% of the original  virus is still  missing.   It is
possible that virus degradation  and  subsequent inactivation  due to  the
forced movement into and out of  the  semi-dense membrane (see SEM  in Fig.
3b) is responsible for this small  virus  loss.

The reason such a large amount of the original virus  is held-up in  the
fiber wall could be explained by the inherent structure of  the fiber
itself.  The lack of a dense skin which  would prevent virus  intrusion
into the membrane and the existence  of a homogeneous  semi-dense wall,
could explain this hold-up.  Ideally, these  experiments have indicated
that a fiber with a dense skin located on the bore  surface  or inside  of
the fiber wall would serve to minimize virus intrusion  and  maximize
the permeation or dehydration rate.   Experiments using  this  system  are
at present underway in our laboratories.
CONCLUSIONS
1.  The core-feed cellulose acetate  hollow fiber unit was successfully
tested as a potential virus concentrator for the first step of a  two
step sequential process to quantify  virus counts in large volumes of
water.
                                  114

-------
2.   Dehydration ratios of greater than 50 in about 1  h  and  an  average
polio type 1  virus recovery of 84 per cent (average of  all  the
experiments)  were obtained.
3.   Feed polio type 1  virus concentrations below 1500 pfu ml~  ,  feed flow
rate and cell type had no significant effect on the virus  rejection.
At 3000 pfu ml"  however, the rejection was decreased by 20 per  cent.
4.   Linear plots were obtained for log (virus concentration) versus
log (volume of feed).   An average polio-1 virus rejection  of 76  per
cent was obtained with the fiber unit.
5.   From 13 to 54 per cent of initial virus was recovered  by backwashing.
An average of 16 per cent of the virus was inactivated or damaged
during the experiments due to intrusion of the virus into  the semi-
dense fiber membrane wall.
6.   The method does not require adjustments of the pH of the water and
there is no interference by organic matter or salinity.
REFERENCES
1.  Bio-Rad  Laboratories  (Richmond,  California 94808, U.S.A.)  The New
    Hollow Fiber Technique.   Bulletin  1004,  1972.
2.   Bel fort  G.   Pressure-driven  membrane processes  and  wastewater
     renovation.  In:   Wastewater Renovation and Reuse,  Shuval, H.I.
     (ed.). New York, Academic Press, 1975.
 3.   Sorber,  C.A., J.F. Malina, and B.P.  Sagik.  Virus  Rejection by
     Reverse Osmosis—Ultrafiltration Processes.  Water Res. 6:1377-1388,
     1972.
                                   115

-------
4.  Sweet B.H., J.S.  McHale, K.J.  Hardy, F.  Morten,  J.K.  Smith and E.
    Klein.   Concentration of Virus from Water by Osmotic  Ultrafiltration-
    I.  Water Res. 5:823-829, 1971.
5.  Sourirajan.  Reverse Osmosis.  London, Logos Press, 1970.

6.  Kesting, R.E.  Concerning the  Microstructure of  Dry-RO Membranes.
    J. Appl. Poly. Sci.  17:1771-1784, 1973.

7.  Brian P.T.L.   Desalination by Reverse Osmosis.  In:  Mass Transport
     in  Reverse Osmosis, Merton, U. (ed.).  Cambridge, Mass., MIT Press,
     1966.

8.  Deinzer, M., R. Melton and D.  Mitchell.   Trace Organic Contaminants
    in Drinking Water:  Their Concentration  by Reverse Osmosis, submitted
    to Water Res., 1974.
9.  Buras, N.  Recovery of Viruses from Waste-water and Effluent by the
    Direct Inoculation Method. Water Res. 8:19-22, 1974.
                                   116

-------
                             SECTION  VII

     EVALUATION OF GAUZE PAD METHOD TO  RECOVER  VIRUSES  FROM  WATER
INTRODUCTION
In 1948 Moore  described for the first  time the use  of  gauze pads
for the isolation of microbial  pollutants  from  sewage.   MacCallum
     2
et al   adapted this method for the detection of poliomyelitis virus
                         3
in sewers.   Melnick et al  were the  first  to point out  that  a gauze
pad soaked in flowing sewage, yields  a  significantly higher
percentage of positive tests for viruses  than the grab  sample method.
However, the findings obtained, using the  gauze pad, are qualitative
and not quantitative, therefore some  groups tried to evaluate gauze
pad efficiency in the recovery of viruses  from water.   Coin  et al
designed a gauze pad sampler for the  quantitative measurement of
viruses in water.  Hoff et al  designed a  modified sampler which
was evaluated by Liu et al  for its  efficiency in concentration of
poliovirus from fresh and seawater and for some parameters affecting
the efficiency, e.g. the pH, and calf serum.
The simplicity of the gauze pad method, as well as its hiqh percentage
of positive tests, prompted us in 1970 to undertake a study to
quantitate this method.  The aim of this study was to evaluate the
gauze pad method for virus concentration from water and also to
compare it to the grab method in field studies.
                                  117

-------
MATERIALS AND METHODS
Virus and Virus Assays
Attenuated poliovirus type 1  served as a model for laboratory studies.
The virus was kindly supplied by the Ministry of Health's Virus
Laboratory in Tel-Aviv.   The stock virus was seeded on Vero cells,
harvested and subdivided for storage at -70°C into volumes of 2 ml.
Virus was assayed according to the plaque forming unit (pfu) method
as described by Shuval et al  in 1970.

Cell Culture
1.  A Vero cell line obtained from Flow Laboratories Ltd. Scotland,
    was grown and maintained in M-199  medium with the addition of
    10% calf serum.
2.  Primary monkey kidney cells, Rhesus obtained from the
    Ministry of Health's Virus Laboratory, Jaffe^Tel-Aviv, were
    trypsinized, grown and maintained  in M-199 medium with the
    addition of 10% calf serum.

Antibiotics
The antibiotic stock solution contained 200,000 penicillin units,
200 mg of streptomycin,  5 mg kanamycin, and 4 mg neomycin per 1 ml H^O.

Gauze Pad
Two kinds of sterile surgical gauze pads were used.  For field studies,
a pad weighing about 5 grams was folded into 16 layers and tied with
                                  118

-------
a nylon cord to which a weight was attached.   For laboratory experiments,
a pad weighing about 6 grams was folded into 24 layers with a diameter
of 10 cm.

Gauze Pad Sampler Apparatus
The gauze sampler is shown in Fig. 5.    All parts of the sampler are
made of stainless steel except for the rubber disc.  The sampler
contains:  a. 5 liter stainless steel container, 23 cm diameter, 22 cm
length; b. and c.  2 parts of the gauze holder, each of 0.5 liter
capacity, diameter 10 cm; d.  rubber disc; e.  gauze pad; f.  stainless
steel screen, each of its pores having a 2 cm diameter.
            Water
                                                      Effluent
                          Fig.  5.   Gauze  sampler
                                   119

-------
The sampler in operation is  shown  in  Fig.  6.    The water from  the
tank (c) enters the gauze sampler  (d)  through  an  opening in  the  bottom.
The water level rises slowly passes  through  the gauze  pad (3), overflows
and is discarded.
                      Fig. 6.  Sampler  in operation
                                  120

-------
PROCEDURES FOR SAMPLING AND ELUTIQN

Field Study -

Two sampling methods were used and compared for concentration and

detection of viruses from sewage:   (1)  Grab Sample Method and

(2) Gauze Pad Method.



(1)  Grab sample method - Twenty-nine daily samples of raw urban sewage

were taken during the period of June 6  to July 12, 1970, and weekly

samples were taken in the period from July 13 to September 27, 1970.

Near the outlet of the sewage pipes, a  composite grab sample of

3 liters sewage was collected for 60 minutes between 7:00 and

8:00 a.m.  Every 10 minutes, 500 ml of sewage was collected and

transferred to a sterile plastic vessel.  The vessels were kept

in an ice bucket, flown to the laboratory on the same day, and

refrigerated at 4°C until the following day.  Two liters of each
                                                             Q
sample were then concentrated by the phase separation method.

After concentration, the viruses were stored at -20°C until

seeded on primary monkey kidney cells.



(2)  Gauze pad method - During the above period from June to

July, 29 gauze pads were each immersed in the sewage outlet

mentioned previously, for 24 hours.  In addition, 10 gauze pads

were immersed at the same outlet during the second  period from

July to September for various periods of time:  3 pads  for 3

days, 2' pads for 4 days and  5 pads for 7 days.
                                  121

-------
After the pad was removed from the sewage, it was transferred to a
nylon bag and the latter was placed in a plastic bottle, cooled and
conveyed to the laboratory as described.  Several drops of a 1%
solution of NaOH were added to the nylon bag containing the gauze
to adjust the pH to 8.0.  The liquid in the gauze was hand
expressed.  The expressed liquid was sterilely transferred to a
glass cylinder.  The gauze was expressed twice more, once after the
addition of 10 ml of sterile distilled water, pH   8.0, and once
after the addition of saline, pH   8.0.  The 3 eluants were pooled
and the gauze was transferred to a specially designed polypropylene
insert with a perforated bottom, placed in a standard 250 ml
centrifuge tube (Fig. 7 ) and centrifuged at 4000 x g for 10
minutes in a cooled centrifuge.   Nearly all of the liquid content
in the pad could thus be recovered.  The volumes collected ranged
from 50-80 ml.  In order to sediment the bacteria, the liquid was
recentrifuged for 30 minutes at 12,000 x g.  After centrifugation,
1 ml of the stock antibiotic solution was added per 10 ml super-
natant.  The samples were kept at -20°C until seeded on primary
monkey kidney cells.
                                122

-------
      1   -  250 ml plastic
            centrifuge  bottle
      2   -  Polypropylene  insert
            perforated  bottom
      3   -  Screw cap
      A   -  The expressed  liquid
      5   —  Gauze  pad
Fig.  7.  Device for extracting
         liquid for gauze pad
               123

-------
 Laboratory Experiments  -
 For laboratory experiments,  various  volumes  of  dechlorinated  tap  water,
 pH  6.0,  containing  different virus  concerrcrations,  v/ere  filtered  through
 the gauze sampler at a  positive  pressure  of  about 0.1  atmospheres.
 The rate of sample  flow was  2.5  liters  per minute.   After  filtration,
 the gauze pad  was removed  from the  sampler with the aid  of tweezers
 and transferred into a  plastic bag.   The  pH  was adjusted to 8.0
 with a  U NaOH solution.   In some experiments,  sufficient  calf serum
 was added to attain 5%  of  the total  volume of the fluids.   The pad
 was then hand-pressed 5-7  times  and  centrifuged at  4000  x  g for 10
 minutes.

 The dry  pad was replaced  in  the  plastic bag  together with  20  ml saline,
 pH  8.0,  to which, in some  experiments,  5% calf  serum was added.   The
 processes of absorption, expression  and centrifugation were repeated
 to  obtain second, third, fourth  and  fifth eluatss.   One  ml  stock
 antiobiotic solution was added to every 10 ml of eluate.   Each
 eluate was  stored at -70°C until assayed.

 RESULTS
 Field Study
 A Comparison of Virus Concentration  by  the Two  Sampling  Methods -
 Grab and  pad -  Table 13 presents the results of enterovirus concen-
 tration  in  raw  sewage by daily grab  samples  together with  the results
.of  enterovirus  concentration  by  24 hour pad  samples.   Since the
                                  124

-------
Table 13.  A COMPARISON OF ENTEROVIRUS CONCENTRATION BY TWO SAMPLING



           METHODS:  GRAB (2 LITERS) AND PAD (24 HOURS)
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Average
S.D.
Grab
pfu/2 L
2340
3120
2666
4680
1680
6360
6280
1480
5560
1180
2200
1760
1800
2040
2600
440
1140
780
2400
3600
2160
1660
1280
1600
1220
1780
2920
4800
2620

24 Hours Pad
ofu/pad
5414
468
7016
15400
5670
11365
16500
12333
10241
5445
8039
7429
6967
4517
6453
244
9821
2053
294
948
2195
2020
6641
7084
2955
4437
10240
6622
8213

Pad
Grab
2.31
0.15
2.63
3.29
3.38
1.79
2.63
8.33
1.84
4.61
3.65
4.22
3.87
2.21
2.48
0.55
8.61
2.63
0.12
0.26
1.02
1.22
5.19
4.43
2.42
2.49
3.51
1.38
3.13
2.92
±2.08
 Average of "rab samples on day of immersing the pad and day of removal
                                125

-------
 volume  of  the grab samples collected was always 2 liters, the concen-
 tration of viruses in 2 liters sewage sample was compared to the
 concentration in the liquid expressed from pads immersed in the
 same  sewage outlet for 24 hours.

 From  Table 13it can be seen that in 4 cases only (No. 2, 16, 19 & 20)
 the recovery of viruses using grab sampling method was higher than
 that  in the pad samples and in 87% of the gauze samples, the virus
 recovery is higher than in the 2 liter grab samples.  On the
 average, the number of viruses detected with the aid of the 24
 hour  gauze pad is about 3 times larger than the average number
 found in 2 liters of the grab sample at the moments of immersion
 and removal of the pad.

 A Comparison Between the Concentration of Viruses in Gauze Pads Kept
 in Sewage  for 24 Hours as Opposed to Those Kept 3. 4. and 7 Days at
 the Same Time -
 Several pads were immersed in raw sewage for different periods of time:
 3 pads for 3 days, 2 pads for 4 days and 5 for 7 days.  For purposes
 of comparison, the average percentage of enteroviruses detected in
 gauze pads  immersed in sewage for 24 hours was assumed to be 100%.
 The results of this comparison are summarized in Table 14.

 From Table 14 it can be seen that in 9 out of 10 pads held for more than
 24 hours,  the percentage concentration of virus was equal  to or less
 than that  of pads immersed for 24 hours.  In only one sample (7 days),
was the percentage concentration twice that of the 24 hour sample.
                                 126

-------
The average percentage of virus concentration in pads kept for 3,  4 or 7
days ranged between 50-80%, i.e., less than the 24 hour average for the
same period.  It is worthwhile to mention that the amount of viruses in
pads kept for 24 hours ranged on the average between 2511-6599 pfu/pad.

Table  14.  THE AVERAGE PERCENTAGE OF ENTEROVIRUS DETECTED ON GAUZE PADS
             HELD IN SEWAGE FOR 1 DAY COMPARED WITH THE PERCENTAGE ON
                   GAUZE PADS HELD FOR 3, 4, AND 7 DAYS
                        (1 day assumed as 100%)
No.
1
2
3
4
5
Average
A Comparison
Average of
1 day
%
100.0
100.0
100.0
100.0
100.0
100.0
Between Three
3 days
%
5.1
105.0
41.2
--
--
50.4
Gauze Pads Run
4 days
%
107.0
18.4
--
--
--
62.7
in Parallel
7 days
I
213.9
107.5
9.0
33.3
36.2
80.0
in Raw Sewage-
Two  sets of triplicate gauze pads were immersed for 24 hours in canals
of flowing raw  sewage and then similarly treated in the laboratory for
virus  detection.  The results are presented in Table  15.  The results
show that there are  no significant differences among the triplicates
of each set.
                                  127

-------
   Table  15.  COMPARISON BETWEEN THE NUMBER OF VIRUSES DETECTED FROM 3
              GAUZE PADS IMMERSED IN PARALLEL IN RAW SEWAGE
No.
1
2
Laboratory

Pad
I
4575
8772
Experiments
pfu/PAD
Pad
II
6605
8934


Pad
III
7138
9504

Ave.
pfu
6106
9070

S.D.
11352
t 384

The field study has shown that the pad method is superior to the grab
method.  However, the gauze pad method is not a quantitative due to
the fact that:  the flow rate of the water passed through the gauze,
the amount of virus adsorbed to the pad, the number passed right through
it and the amount of unrecovered virus from the total  adsorbed—are all
unknown.

Because of the above and since it has not been possible to quantitate
field results, we decided to test the pad method under controlled
laboratory conditions in order to evaluate field results quantitatively.

The Effect of Repeated Elutions -
The effect of successive elutions on virus recovery from gauze pads
immersed in tap water and flowing sewage was determined.
                                  128

-------
Tap water - Various volumes (700-51,500 ml)  of dechlorinated tap water
were filtered through the gauze in the sampler shown in Fig. 5.   In
some experiments each gauze was eluted 4 times (table 16);  and in others,
each gauze was eluted 5 times (Table 17).

From Table 16, it can be seen that the first eluate contained, on the
average, 47% of the total virus obtained from gauze in 4 elutions.
The second eluate contained, on the average, 25%; i.e., 72% is eluted
after the first 2 elutions.  Table 17 shows also on the average, 72%
after the first 2 elutions.  However, the average in the first eluate
is 42%.  The fourth eluate in Table 16 and the fifth eluate in Table 17
contained the least amount of virus.

Sewage  - The effect of repeated elutions with sewage was tested.  Pads
were immersed in flowing sewage for 24 hours.  Each gauze was eluted
four times.  The results are presented in Table 18.

As shown in Table  18,  the  first eluate contained an average of 50% of
the total virus obtained from the  gauze in 4  elutions, whereas the
second  eluate contained  an average of  26%, i.e. 75% is eluted after
the first two elutions.  It  should be  pointed out that Table  18
presents data for  wild enteroviruses  from sewage, whereas Tables  16
and 17  present  those  for seeded attenuated poliovirus  I.
                                  129

-------
CO
o
         Table 16.  THE EFFECT OF SUCCESSIVE GAUZE PAD ELUTIONS ON POLIOVIRUS I ELUTION


                    (TAP WATER-4 ELUTIONS)3
Eluate
1st
2nd
3rd
4th
Total
Exp
pfu
73
13
18
4
108
. 1
.%
67.6
12.1
16.2
4.1
100.0
Exp
pfu
122
37
42
9
210
. 2
%
57.9
17.7
20.0
4.4
100.0
Exp.
pfu
2709
2736
1430
1017
7872
3
%
34.3
34.7
18.1
12.9
100.0
Exp.
pfu
2250
3103
1511
1573
8437
4
%
26.7
36.8
17.9
18.0
100.0
Average
46.6
25.4
18.1
9.9
100.0
S.
+ 19
+ 12
+ 1
+ 6

D.
.3
.3
.6
.8

         aln this table and in the subsequent ones, each elution is expressed as a percentage of


          the sum total virus (100%) obtained from the gauze in the sum total elutions.

-------
Table 17.  THE EFFECT OF SUCCESSIVE GAUZE PAD ELUTIONS ON POLIOVIRUS I  ELUTION  (TAP  WATER  -  5  ELUTIONS)
	 ExpTT 	
Eluate
1st
2nd
3rd
4th
5th
Total
pfu
1466
2076
812
949
612
5915
%
24.8
35.1
13.7
16.1
10.3
100.0
Exp. 2
pfu
4965
2508
976
632
340
9421
%
52.7
26.6
10.4
6.7
3.6
100.0
Exp.
pfu
46499
44145
22648
12600
14196
1.4xl05
3
%
33.2
31.5
16.2
9.0
10.0
100.0
Exp. 4
ofu
71402
49776
30100
11332
10000
1.7xl05

%
41.4
28.8
17.8
6.6
5.8
100.0
Exp.
pfu
96000
44800
21200
9068
5506
1.8xl05
5
%
54.4
25.4
12.0
5.1
3.1
100.0
Exp.
pfu
83335
65400
21332
17868
6019
1.9xl05
6
%
43.2
33.4
11.0
9.3
3.1
100.0
Average
Of
to
41.6
30.1
13.5
8.8
6.0
100.0
S.D.
+ 11.3
+ 3.8
+ 3.0
+ 3.9
+ 3.4


-------
Table 18.  THE EFFECT OF SUCCESSIVE GAUZE PAD ELUTIONS
           ON ENTEROVIRUS RECOVERY
                     (Raw sewage)
Eluate
1st
2nd
3rd
4th
Total
pfu
5250
1611
1244
667
8772
%
59.8
18.4
14.2
7.6
100.0
pfu
4500
2417
1371
646
8934
%
50.
27.
15.
7.
100.
i
4
1
3
2
0
pfu
3825
3125
1915
639
9504
Average
% % S
40.2
32.9
20.2
6.7
100.0
50
26
16
7
100
.1
.1
.6
.2
.0
+9
+7
+3
1°

.D.
.8
.3
.2
.5

                          132

-------
The Effect of Calf Serum on Elution Recovery Efficiency of Viruses
From Gauze Pads -
Since it has been found that the addition of serum to elutinq fluid
significantly affects recovery efficiency of virus from pads (Liu et
al ), experiments were performed in order to compare the effect of
calf serum (5%) added to a saline eluant.  In each experiment, equal
volumes of tap water, pH = 6.0, containing equal concentrations of
attenuated poliovirus I were passed through separate gauzes.  The
two gauzes were treated as follows:  one gauze was eluteu 4 times
with saline containing 5% calf serum, pH = 8.0, the othe'  was similarly
treated but without the addition of calf serum.  The result? are
summarized in Table 19.

Table 19 shows that in 2 experiments U ^nd 4), the addition of serum
was slightly advantageous, whereas in 2 other experiments (] and 3),
saline was the better eluent.  The fifth experiment shows r,o effect
of the serum on elution.  At any rate, after 2 elutions, with or
without serum, about 75% of the total virus obtained from 4 elutions
is released.

The Efficiency of Recovery of the Gauze Sampler for Viruses in Water-
Twenty experiments were carried out in order to determine the efficiency
of the gauze pad sampler.  This was done by comparing the concentration
of viruses before passage through the gauze to that found on the gauze
after filtration.  The results of these experiments are summarized in
Table  2n.
                                  133

-------
           Table 19.  THE  EFFECT OF  CALF SERUM  ON  ELUTION OF POLIOVIRUS I FROM GAUZE PADS (TAP WATER)


                     Exp.  1          Exp. 2Exp. 3          Exp. 4         Exp.~5
                  with  with-     with     with-   with    with-   with    with-  with     with-  with    with-
                  C.S.   out     C.S.     out     C.S.    out     C.S.    out    C.S.     out    C.S.    out
         El uate    %      %	2	%	%	%__	%	%	%	%	%	%

         1st      57.9  67.6     54.7     27.6    26.7    34.3    56.1    36.9    44.6    43.9    48.0   42.1
                                                                                               + 13.0  +15.4

         2nd      17.7  12.1     27.6     39.2    36.8    34.7    26.2    35.1    34.5    30.6    28.6   30.3
                                                                                                + 7.5  +10.6
co
*"        3rd      20.0  16.2     10.7     15.3    17.9    18.1    12.4    18.0    11.4    18.5    14.6   17.2
                                                                                                +4.2   +1.4


         4th       4.4    4.1      7.0     17.9    18.0    12.9     5.3    10.0     9.5     7.0     8.8   10.4
                                                                                                + 5.5   +5.0

         Total    ToO  ToOToO    ToO   ToOTooTo   ToOToO   ToOTooTo   ToO  100.0
         Total"""'-
         virus       1.6  x  10      7.2 x 10J      8.2 x 10J       1.5 x 10°       1.7 x 10b
         on  pad	

-------
Table  20.  THE RECOVERY OF THE GAUZE PAD SAMPLER FOR CONCENTRATION  OF



           POLIOVIRUS I (TAP WATER)
Exp.
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Vol ume
ml
700
700
700
7500
7800
7800
9000
10000
10000
10000
19000
30000
30000
30000
30000
30000
30000
51500
51500
51500
Before concentration
Total
pfu jfu/ml
9-OxlO3
4.9xl04
5.4xl04
9.7xl04
5.5xl05
7.0xl05
9.3xl03
2.8xl03
2.8xl05
4.0x.05
2.5xl04
l.lxlO6
1.9xl06
l.lxlO7
1.3X107
2.4xl07
2.9xl07
6.6xl05
3.6xl06
4.0xl06
12.9
70.0
77.5
12.9
70.0
77.5
1.0
1.0
28.3
41.7
1.3
38.1
59.4
350.0
446.7
840.0
966.7
12.9
70.0
77.5
After concentration
pfu/ml
13.3
80.0
105.0
22.9
162.5
185.0
2.7
5.3
197.3
210.9
3.5
147.9
235.5
4E09.0
3bOO.O
4750.0
4250.0
45.0
450.0
725.0
Total
pfg/pad
531
3.2xl03
4.2xl03
917
6.5xl03
7.4xl03
108
211
7.9xl03
8.4xl03
141
5.9xl03
9.4xl03
l.SxlO5
1.^:1 05
1.9xl05
1.7xl05
l.SxlO3
l.SxlO4
2.9xl04
Recovery
efficiency
5.9
6.6
7.7
1.0
1.2
1.1
1.2
2.2
2.8
2.1
0.6
0.5
0.5
1.6
1.1
0.8
0.6
0.3
0.5
0.7
                                    135

-------
From Table  20, it can be seen that:
(1)  The recovery efficiency of viruses by the gauze pad is low:
     0.3 - 7.7%
(2)  The recovery efficiency of the viruses is a reciprocal function
     of the sample volume and not of the total number of viruses
     passed through the gauze:  the larger the volume, the lower
     the efficiency (an efficiency of 7.7% for 700 ml versus 0.3%
     for 51,500ml).
(3)  On the other hand, a comparison between the amount of virus
     per ml (pfu/ml) found in the expressed gauze fluid and the
     amount of virus per ml  (pfu/ml) in the control  sample (before
     concentration)  shows that in each experiment the amount of
     viruses after concentration in the expressed gauze fluid is
     higher than that of the control.   (Compare columns 4 and 5 in
     Table 20).

The Effect of Sample Volume on Virus Recovery Efficiency-
In order to verify the recovery efficiency of viruses as a reci-
procal function of the sample volume,  controlled experiments were
carried out as follows:

Attenuated poliovirus I ac a concentration of 12.9 pfu/ml was added
to 60 liters of dechlorinated tap water.   The 60 liters were then
divided into 3 aliquots: 700, 7,800 and 51,500 ml, each of which
was passed through a separate gauze.  Each gauze was subsequently
uniformly eluted 4 times.
                                 136

-------
                                      Table 21.   THE EFFECT OF SAMPLE VOLUME  ON POLIQVIRUS  1 RECOVERY (TAP WATER)
Experiment no. 1





CO
-4
Virus
concentration
pfu/ml

i n
Volume in gauze
ml control eluate
700 12.9 13.3



Virus
in
gauze
virus
in %
control recovery
1.0 5.9

Experiment no. 2 Experiment no. 3
Virus
concentration
pfu/ml

in
in gauze
control eluate
70.0 80.0



Virus
in
gauze
virus
in
control
1.1

Virus
concentration
jjfu/ml Virus
in
gauze
in virus
% in gauze in %
recovery control eluate control recover"
6.6 77.5 105.0 1.4 7.7

Ave. of
Virus
concentration
pfu/ml

in
in gauze
control eluate
53.5 66.1
.-35. 3 i47.4
3 exp.


Virus
in
gauze
virus
in
control
1.2
io.2





recovery
6.7
io.9
 7800    12.9     22.9     1.8       1.0      70.0    162.5     2.3        1.2      77.5    185.0     2.4       1.1       53.5   123.5     2.2



                                                                                                                       135.3   187.8    10.3







51500    12.9     45.5     3.5       0.3      70.0    450.0     6.4        Q.5      77.5    725.0     9.4       0.7       53.5   406.8     6.4        0.5



                                                                                                                       ^35.3  ±341.8    -3.0       io.2

-------
 In  a  second  and  third  set  of  experiments,  the  virus  concentration  under
 similar  conditions was  70.0 and  77.5  pfu/ml  respectively.  The  results
 of  the 3 sets of experiments  are summarized  in Table  21.

 Table  21 shows the sharp decrease  in  recovery  efficiency relative  to the
 sample volume in each experiment:  average recovery  efficiency  of  6.7%
 for 700  ml and 0.5% for  51.5  L.     The concentration factor of the
 gauze  is obtained by dividina the  pfu/ml of gauze fluid after filtration
 by  the pfu/ml of the control  fluid before filtration.  This factor
 increases in direct proportion to  the sample volume:  the factor is 1.2
 for 700  ml versus 6.6 for  51,500 ml.
 DISCUSSION
 This study is divided into two parts:  (a) field study for comparison of
 two sampling methods, pad and grab samples; and (b) laboratory  experiments
 for evaluation of the gauze pad method for concentration of viruses
 from water.

 For the  second part,  we designed a flow-through gauze sampler (Figs. 5 and
 6) in order to quantitate the gauze pad method.  A new method involving
centrifugation for the expression of gauze liquid was used (Fig. 7).
The centrifugation method enabled the expression of 95% or more of the
liquid absorbed by the gauze.

The findings  of this  study show that the recovery efficiency for
concentration of viruses by gauze pads from tap water is very low and
                                    138

-------
dependent on sample volumes (0.5% for 51  liters  versus  7%  for  700 ml).
These findings are in agreement with those of  Liu  et  al.    On  the other
hand, we found that the concentration factor of  the pad is dependent
on the virus concentration in volumes up  to 30 liters:   volumes
greater than 30 liters show no effect (1.2 for 700 ml  and  6.4  for 30
or 51 liters).
The addition of calf serum to the wash fluid did not  significantly affect
elution of viruses from pads immersed in tap water, see Table 19.
Liu et al  found that the addition of calf serum also enhances
elution from tap water pads, dependent on pH.   On  the average, about
50% of the viruses are eluted in the first wash as compared to about
70% in the first two washes, see Tables 16  and  17.

Accordinn to our results (Tablel3), we find that  the concentration  of
virus in 2 liters of sample by the grab method is  3 times lower, on
the average, than that of liquid expressed from the 24 hour oad.
Obviously, different results would have been obtained if we had used
gauze of different size or used different volumes  of grab samples.
If we compare equal volumes of gauze and grab, i.e., the concentration
of virus in 1 ml of liquid expressed from a gauze pad kept in sewaae
for 24 hours compared with the average concentration in 1 ml of grab
sample on the day of immersion and the day of pad removal, we find
that the gauze  liquid contains 90 times more virus than the number
found in the sewage in which it was  immersed.
                                    139

-------
In our field work, two sample methods were employed:   grab sample and
gauze pad.  Composite samoles were taken by the grab  method and
                                                         o
concentrated by the phase separation method (Shuval  et al  ).   This
method, thouqh it is quantitative, has been shown to  be somewhat
selective in its ability to concentrate all enteroviruses  (Grindrod
        g
& Cliver ).  On the other hand, the gauze pad method  thouqh advantageous
in sampling for a continuous and long period of time  (24 hours and
more)tis limited by beinq qualitative, since we don't know the flow
rate, how much virus is adsorbed to the pad, and how  much  recovered.
In addition, wes as well as other groups (Hoff et al,   Liu  et al )
find that only a very low percentaae (about 1%) of the entire entero-
virus population can be concentrated by this method.   In spite of
these limitations, various investigators have found the qauze pad
method to be superior to the grab method in the detection  of viruses.
It may be that the limitations are overcome by the large volume of
sewage which passes through and by the pad's ability  to pick up
clumps of viruses, excreted by individual carriers, which  enter the
sewage intermittently.  Grab samples may completely miss such peaks
in virus concentration.
In addition, the concentration of viruses in sewage during 24 hours
may differ from that found at the moment of grab sampling.  However,
it was found that the concentration of virus in the gauze eluent is
always higher than the concentration of virus in the sewage surround-
ing the gauze, i.e., 3 to 410 times higher, and this means that the
                                  140

-------
viral  recovery of the pad in sewage is quite effective despite the low
efficiency.

It was found that there is no advantage in immersina the gauze pads for
more than 24 hours, see Table 14.  Lund and Hedstrom   found that
pads immersed for 24 contained more viruses than those maintained for
48 hours.

A comparison between the number of viruses detected in aauze pads
immersed in parallel in raw sewage reveals only slight differences
(Table  15).  Lund and Hedstrom   found it advantageous to maintain two
gauzes  simultaneously as they discerned small differences in the
concentration as well as different strains in the two gauzes.

In spite of the  limitation of the gauze pad method  (low recovery  efficiency
and qualitative  results) we  think that this method  is useful and may
be of practical  value in epidemiological  studies for early discovery,
or epidemiol ogic intelligence work  on viral  outbreaks even before  clinical
signs can  be  diagnosed,  since virus excretion into  sewage precedes
the clinical  syndromes.
                                    141

-------
REFERENCES
1.   Moore, B.   The detection of Paratyphois  in  towns  by means  of  sewage
    examination.   Mon.  Bull. Minist.  Hlth.  Publ.   Hlth. Lab. Serv.
    (directed by Med.  Res.  Council),  7:241-248,  1948.

2.   MacCallum, P.O., W.C.  Cockburn,  E.H.R.  Smithard,  and  S.L.  Wright.
    The Use of Gauze Swabs  for the Detection of  Poliomyelitis  Virus
    in Sewers.  Second Int.  Poliomyelitis  Conf.,  Lippincott, Phila.
    p. 484-487, 1952.

3.   Melnick, J.L., J.  Emmons, M.  Opton,  and  J.H.  Coffey.   Coxsackie
    Viruses from Sewage.   Methodology Including  an Evaluation  of  the
    Grab Sample and Gauze  Pad Collection Procedures.   Amer. Jour.
    Hygiene 59:185-195, 1954.

4.   Coin, L., M.L. Menetrier, J.  Labonde,  and M.C. Hannoun.  Modern
    Microbiological and Viroloqical  Aspects  of  Water  Pollution.
    Direction de THygiene Sociale,  Paris,  1961.

5.   Hoff, J.C., R.D. Lee,  and R.D. Becker.   Evaluation of Method  for
    Concentration of Microorganisms  in Water.  Water  Hygiene Division,
    1971,  p.37.

6.   Liu, O.C., D.A. Brashear, H.R. Seraichekas,  J.A.  Barnick and
    T.G. Metcalf.  Virus  in Water:  A Preliminary Study on a Flow-
    Through Gauze Sampler  for Recovering Virus  from Waters.  Appl.
    Microbiol. 21:405-410,  1971.
                                 142

-------
 7.   Shuval,  H.I., A. Thompson, B. Fattal, S. Cymbalista and Y. Wiener.
     Natural  Virus Inactivation Processes in Sea Water.  In:  Proceedings
     of the National Specialty Conf. on  Disinfection, New York,
     American Society of  Civil Engineers, pp 429-452, 1970.
 8.   Shuval,  H.I., B. Fattal, S.  Cymbalista, and N. Goldblum.  The
     Phase-Separation Method for  the Concentration  and Detection of
     Viruses  in  Water.  Water Research 3:225-240, 1969.
 9.   Grindrod, J.  and D.O.  Cliver.  Limitation of the Polymer Two Phase
     System for  Detection of Viruses.  Archive fur  die gesamte Virusfor-
     schung 28:337-347, 1969.
10.   Lund,  E. and  C.E.  Hedstron.   A Study on Sampling and  Isolation Method
     for the  Detection  of Viruses in Sewage.  Water Res.,  3:823-832,
     1967.
                                    143

-------
                             SECTION VIII
     A RAPID FLUORESCENT ANTIBODY METHOD FOR QUANTITATIVE ISOLATION
                         OF VIRUSES FROM WATER

INTRODUCTION
The accepted techniques for virus isolation involve inoculation of
tissue cultures followed by incubation.  Virus is then demonstrated by
the appearance of cytopathic effect or plaques in the tissue cultures.
The entire process continues for 3-7 days or sometimes even longer,
depending on the method.

It is obvious that a method for testing potable water sources requiring,
at least, 3 days does not provide an adequate degree of protection,
especially when contaminated river water is distributed to large
populations within hours after passing through a treatment plant.
Such tests should be as rapid as possible, and the assay should be
completed before the water is released to the distributing system.

Rapid qualitative methods for demonstration of enteroviruses based on
the fluorescent antibody (FA) technique have already been
described.   ''   In this study efforts were made to develop it one
step further, thus affording a rapid quantitative determination of
enteric viruses in water.  As a model, poliovirus type 1 was chosen.
                                  144

-------
MATERIALS AND METHODS
Virus Stock
Poliovirus type I (Brunhilda) was used throughout this study.   The
virus was grown in Vero cells (Flow Laboratories, Scotland), accumulated
and concentrated by the phase separation method.      Vials containing
0.5 ml of the concentrated viruses were stored at -80°C.   Before each
test a sample was defrosted and diluted according to the experimental
requirements.  The virus was assayed on BGM cells,  as described
elsewhere.

Antiserum
Rabbits were injected in the footpad with 1.0 ml  emulsion containing
equal volumes of concentrated poliovirus I (2.3 x 10   pfu/ml) and
complete Freund's adjuvant (Difco Laboratories).   Second, third and
fourth identical injections were administered after 2, 6 and 10 weeks,
respectively.  Fourteen days after the last injection, blood was
withdrawn from the ear veins and the serum separated.

Purification and Concentration of Gamrnaglobulin
                                            Q
This was carried out as described elsewhere.

Labeling of  Gammaqlobulin
Labeling was performed as  described previously   with  the exception  of
the  temperatures  (about 20°C) and the  stirring which  was  continued for
4 hrs. Isomer I of fluoresceinisothiocyanate (FITC: BDH, England) was
used  throughout  this study.
                                   145

-------
Vero Cell Powder
Powder was prepared by acetone-drying from saline-suspended  cells  in  the
same manner as with liver powder.

Removal of Non-specific Staining
The labeled gamma globulins were adsorbed by the Vero cell  powder to
remove non-specific staining in the following manner:  100  mg  powder
were suspended in saline and centrifuged at 1,000 x g for 15 min;
5 ml labeled gamma globulin were added to the sediment and  mixed.
This suspension was incubated at 37°C for one hour, followed by
centrifugation at 40,000 x g for 30 min.  The supernatant was considered
adsorbed serum.  Additional adsorptions were sometimes necessary until
complete removal of non-specific staining was achieved.   A  solution
of rhodamine-labeled bovine albumin (Microbiological Associates)
was added to the adsorbed serum (final dilution 1:40) to obtain a
clear differentiation between positive and negative reactions.
Micro Tissue Culture for FA Staining
A suspension was prepared containing 2-3 x 106/ml BGM cells in M-199
medium with Hank's salts (Flow Laboratories, Scotland),  20% fetal
bovine serum and antibiotic solution (penicillin 200 units, strep-
tomycin 200 yg, kanamycin 5 yg, neomycin 4 yg  per milliliter final
concentration). This mixture was inoculated with virus according to
the requirements of each experiment.  Five drops (0.02 ml/drop)
were put on a standard microscope slide, several slides  per experi-
ment.  The slides were  then  placed  into  specially designed vessels
                                 146

-------
containing small  volumes of water to prevent dehydration,  and  incubated
in a C02 incubator at 37°C.  Unless otherwise stated,  the  length  of  in-
cubation was 18-20 hrs.  After incubation,  the slides  were rinsed with
0.15 M phosphate buffer, pH 7.2, followed by three washings in acetone.

Fluorescent Antibody Staining
The direct method was used,   with the staining continuing for one
hour.  The stained preparations were examined under a  Zeiss WL Research
microscope with fluorescent attachment.

Concentration of Viruses from Water and  Their Inoculation  into
Tissue Cultures
Viruses were concentrated from the water by filtration through a
cellulose nitrate membrane filter (pore  size o.45 y; Sartorius Co.,
Germany).  For volumes of 5 liters, the  technique of Rao and
          12
Labzoffsky   was utilized.  The membrane diameter was  47 mm.  The
adsorbed virus was eluted from the membrane with 7 ml  3% beef extract.
For the testing of 40 L volumes, concentration of the  viruses was
carried out in two stages, according to  Sobsy et al,   with the follow-
ing modifications:  the pH of the water  sample was adjusted to 3.0
and filtered through a cellulose nitrate membrane filter (diameter
142 mm, pore size 0.45 y).  To elute the adsorbed virus, 100 ml
glycine buffer (0.2 M, pH  11.5) was filtered through the membrane.
The pH of the eluate was adjusted to 3.0 with 1 M HC1, and the
suspension was again filtered (filter diameter 47 mm).  The adsorbed
                                            13
virus was eluted  with 7 ml 3% beef extract.
                                   147

-------
One ml sterile M-199 medium (concentrated 10 fold), 2 ml fetal  bovine
serum and 0.1 ml antibiotic mixture (200,000 U penicillin, 200,000 yg
streptomycin, 5,000 yg kanamycin, 4,000 yg neomycin per ml) were
added to each of the 7 ml concentrate obtained in both procedures.
BGM cell (0.1 ml, 4-6 x 106) were added to 2 ml  of the concentrate.

Micro tissue cultures were prepared on microscope slides with this
virus/cell  suspension as described above.  The remainder of the virus
                                                          p
was used for inoculation of the tissue cultures  on plates.
RESULTS
Growth of Poliovirus in Micro Tissue Cultures
To determine the optimal time required for FA staining of the cultures,
infected micro tissue cultures were prepared and incubated.  The first
slide was removed after incubation of one hour,  washed, fixed and
stained.  The second slide was treated in the same manner after in-
cubation of two hours.  This procedure was followed for 24 consecutive
hours.  All slides were examined microscopically.  Incubation of one
hour  showed  the cells to be attached to  the  slide; nearly  all were
still spherical and separated one from the other.  All cells stained
reddish-brown, the color of the rhodamlne bovine albumlne used as the
counter stain.

Four hours  after Incubation, the cells appeared  as in mature tissue
cultures; they were flattened and attached to one another, giving the
                                  148

-------
impression of monolayers.   The cells were still  stained with  the  counter
stain.

Individual cells stained with the green fluorescence of the fluorescent
antibodies started to appear after 6-7 hours of incubation, and after
9 hours a maximum of individual stained cells was reached.   They
appeared as isolated green spheres, surrounded by the reddish brown
stained cells (Fig. 8).

After an incubation period of 16 hours the positive cells started to
appear in foci of 5-30 cells, looking like clusters of green fluorescent
spheres (Fig. 9 ).  The number of positive cells increased in each
infected foci until, after 22-24 hours, the clusters had grown into
small plaques, each comprising scores of cells.

Quantitative  Estimation of Poliovirus with FA Staining
The  foregoing experiments indicated that the growth process of polio-
virus 1n micro tissue cultures,  as  revealed by the FA staining, can be
divided into  two  stages:  stage  1,  reaching Its peak at 9 hours,
during which  positive cells  remain  single entities, and stage 2,
with the  peak at  about  18-24 hours, when the positive cells form
clusters.   It is  reasonable  to assume  that both the single positive
cells appearing at 9 hours and the  clusters at 18-24 hours, each
represent  one plaque forming unit  of  the original  suspension.  To
test this  contention, a number of  experiments were carried out.
                                   149

-------
Fig. 8.  BGM Cells infected with  poliovirus  type  I,
         stained with fluorescent antibodies;  9 hours
         after infection.   Two  positive  (white) single
         cells are clearly  seen (x 100)
Fig.  9.   BGM  cells  infected  with  poliovirus  type  I,
         stained with fluorescent  antibodies;  18 hours
         after infection.   The  cluster of  positive
         (white) cells  is  clearly  defined   (x 100)
                        150

-------
Micro tissue cultures, infected with serial  10-fold dilutions of
poliovirus, were prepared as described in methods and divided into two
batches.  One batch was incubated for 9 hours and the other for 20
hours.   The FA stained preparations were examined thoroughly under the
microscope.  The single positive cells (9 hrs) and the clusters of
positive cells (20 hrs) were counted.  Simultaneously, the same poliovirus
stock was titrated on plates as a control (Table 22).  There is agree-
ment between the FA preparations incubated for 20 hours and  the controls.
The 9-hour incubate, however, shows less satisfactory results, with a
virus titer far below that of the controls.   Another important factor
with the 20-hour incubate was the readily discernible clusters, which
should enable any technician with minimal experience to read results
with ease.  The single stained cells  (9 hours), on the other hand,
are often difficult to recognize or to distinguish from non-specific
fluorescence.  In view of these obstacles, 9-hour incubates were dis-
continued and the results described below concern only samples taken
at 18-24 hours.

TableZS compares the poliovirus titers of the FA stained cells and
the plaque assay method.  The results  of the plaques were read two
days after those of the  FA.  The virus titers  in both methods are
nearly  identical.

Quantitative  Detection of Poliovirus  in Water
After  it was  established that the  above  procedure enables  a  rapid  and
quantitative  assay of poliovirus in  controlled experiments,  it was
                                  151

-------
en
ro
        Table  22. TITERS OF POLIOVIRUS  I  OBTAINED WITH  THE  FLUORESCENT ANTIBODY  METHOD  AFTER 9 AND 20


                   HOUR  INCUBATION, AND WITH  THE PLAQUE COUNT METHOD ON PLATES
Exp.
no.
1


2


3


Incubation
period
(hours)
20
9

20
9

20
9

"DiTutTorf
of viruses
sample
lO'4
io-5
io-2
io-3
io-4
ID'5
io-1
io-2
io-4
io-5
io-2
TO'3
No. of positive
cells or cell
groups per drop
3, 4, 7, 3
2, 1, 1, 2
169, 190, 164
23, 21, 18
2, 3, 1, 2
1, 1, 0, 1
78, 61, 67
8, 5, 2
4, 3, 4, 2
2, 2, 1, 0
58, 44, 53, 26
6, 4, 3, 4
Calculated
virus titer
2.1 x IO7
7.5 x IO7
8.7 x IO6
1.0 x IO7
1.0 x IO7
3.8 x 107
3.4 x IO5
2.6 x IO5
1.6 x IO7
6.3 x IO7
2.1 x IO6
2.1 x IO6
Virus titc-r
on plates

3.4 x IO7


3.0 x IO7


2.6 x IO7


-------
Table  23.  TITERS OF POLIOVIRUS  TYPE  I  OBTAINED WITH THE FLUORESCENT

           ANTIBODY TECHNIQUE  AND WITH THE  PLAQUE COUNT METHOD


Exp.                    	    '           V1rus"t'lter""
no*	Plaque count           "FA technique

1                                3.0 x  107              2.6 x 107

2                               2.7 x  107              6.2 x TO6

3                               1.5 x  TO7              1.1 x 107

4                               2.0 x  107              1.2 x 107

5                               1.1 x  107              1.2 x 107

6                               2.2 x  107              1.3 x 107

7                               1.8 x  107              1.5 x 107

8                               1.4 x  107              1.4 x 107
                                 153

-------
decided to investigate its practical  application—that is,  the  quanti-
tative isolation of the virus from large volumes  of water.   Two accepted
methods for the concentration of viruses from water were utilized  in
this study:  one designed for volumes of 5 liters,  and the  other for
larger bodies of water, in our case,  a volume of  40 liters.   Tap water
was contaminated with 10-1,000 pfu poliovirus I,  according  to the  re-
quirements of the experiment, concentrated and titrated simultaneously
on micro cultures and plates.  Tables 24 and 25 summarize the results
of experiments with the 5- and 40-liter samples,  giving good correlation
between the two methods.  Because of contamination, no results  could
be obtained in three experiments with the plaque  assay, a factor which
did not come into play with the FA method (Table   25).
DISCUSSION
The main reason for virus monitoring of water is  to protect the popu-
lation from health hazards.  It would be desirable if the quality  of
the water be known before it reaches the consumer.   Since impoundment
of vast quantities of water for long periods is undesirable, rapid
results of water tests are imperative.

Standard bacteriological tests require 24 hours,  but even this  rela-
tively brief period should be shortened for most  water distribution
systems.  Virological tests which take days, and  sometimes  weeks,  are,
of course, even less efficient 1n providing early warning of
contamination.  The reason for this time lag is not inherent in the
                                   154

-------
Table  24.  COMPARISON OF THE FLUORESCENT ANTIBODY TECHNIQUE WITH THE



           PLAQUE COUNT METHOD FOR QUANTITATIVE EVALUATION OF VIRUSES



           IN 5 LITERS OF WATER
Exp.
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14

FA counts
800
800
110
90
100
100
100
28
55
55
24
20
26
32
pfu recovered
Plaque counts
920
624
942
140
150
167
133
37
35
66
66
61
44
61
                                  155

-------
Table  25.  COMPARISON OF THE FLUORESCENT ANTIBODY TECHNIQUE WITH THE
            PLAQUE COUNT METHOD FOR THE QUANTITATIVE EVALUATION OF
            VIRUSES IN 40 LITERS OF WATER
Exp.
no.
1
2
3
4
5
6
7
8
9

FA counts
22
36
3
7
21
14
42
8
8
ofu recovered
Plaque counts
33
8
7
contaminated
contaminated
contaminated
6
13
19
                                 156

-------
concentration methods.   Although some of these methods  are  slow,  as  for
                                                             5 6
example, the phase separation method which needs 24-48  hours, '   most
concentration methods require 1-6 hours only.   On the other hand, the
time needed for the isolation of virus is far more extended.   The usual
techniques involve inoculation of tissue cultures followed  by incubation.
Viruses are then demonstrated by the appearance of the  cytopathic
effect (CPE) or plaques in the tissue cultures.  The entire process
lasts from 3 to 7 days or longer, depending on the particular method.
In our laboratory, the plaque assay is used, which enables  quantitative
evaluation of the virus 3-5 days after sampling.  It is possible to
shorten the time needed for the plaque assay, but the danger of incom-
plete plaque development then exists, which is expressed by too low a
number of plaques or by false negatives.

The main objective of  the present study was the development of a
quantitative method  for the  isolation  of  viruses  from water which
would be at  least as rapid as the current bacteriological methods.

Viruses enter drinking water sources  by way of  domestic sewage.
                                                     15
In  the  latter,  various types  of  viruses  are present     most
of which belong to  the enterovirus  groups,  including the following
sub-groups:  polio,  coxsackie and echo viruses.   They  are
mostly  found in sewage and their concentration  ranees  from 400-2,000
pfu/1,     sometimes  reaching over 10,000  pfu/1.   Of  the entire
enterovirus  group,  only the  three polio types should always  be
                                   157

-------
present in domestic sewage of urban areas in developed countries.   The
reason is the wide-spread routine administration of live poliovirus
vaccine to infants.  The viruses multiply in the intestines and are
excreted with the feces for several weeks after the initial administration.
Their number in the feces may reach 10  pfu/gm.    The presence of other
types of enteroviruses in sewage depends on the degree of their distri-
bution amongst the population at a given time, and this of course
fluctuates considerably.  This fact should be borne in mind when water
is tested for the presence of viruses.  Moreover, a single system  that
could be used to isolate all  the different types of enteroviruses  does
not exist.  Most of the coxsackie A virus types do not multiply in tissue
cultures but require suckling mice.  Polio- and echovirus, on the  other
hand, grow in tissue cultures.  Also, the length of time  required for
the development of the cytopathic effect varies for each virus type.
For example, CPE of poliovirus  is displayed in 3-5 days, while that  of
reoviruses  appears considerably later.   The isolation of the different
virus types found in water requires a wide range of techniques and
systems, a fact which makes routine practical  use too complicated.  In
bacterial examinations, only one type of bacteriurn—coliforms--is
taken as being representative for other  enteric bacteria present in the
feces, thus becoming the indicator for bacterial fecal pollution.   It
would therefore be logical to select a viral indicator, the poliovirus
being the most suitable candidate.  Such an indicator for viral pollu-
tion of water could simplify the technique since it is based on a
single system.  Furthermore,  it would also significantly shorten the
identification period.
                                    158

-------
However, there are certain limitations reciardinq the use of poliovirus
as viral indicator.  Unlike E.  coli, thf.y do not necessarily comprise
tne majority of viral population in sewage, and at times are not
present at all (see Section IX  of this report).  Therefore, negative
results obtained with tests based on polioviruses only, would not be
sufficient proof for the absence of other dangerous viruses in water.
On the other hand, the lack of  a rapid virus test slows down the imple-
mentation of routine viral examination of potable water.  The use of the
three polioviruses as indicators is suggested for a rapid examination
of water as part of a complete  and comprehensive virological test that
will includp all possible viruses in water.  Thus, a rapid and preliminary
answer i.iav he nl tained as to the presence of viruses in water designated
for hu»ian consumption.  With this objective in mind, poliovi ruses were
selected for the present study, with poliovirus I as model.

The process of the development of the CPE is slow and includes the
following staaes:  infection of susceptible cells, multiplication and
liberation of the progeny, upon which the cycle starts again.  In the
case of polioviruses, each cycle lasts for 6-9 hrs ' at 37°C.  Several
such cycles are needed before the CPE can be detected visually; in
other words, the results are detectable only after a few days.  A reagent
that would enable identification of the virus  in the tissue culture,
before the CPE becomes visible, would therefore allow for  a much
shortened test time.  Fluorescent antibodies could here fulfill the
role of such a reagent.
                                   159

-------
Final identification of viruses >s currently  earned out either by
the use of specific antibodies which neutralize the appearance of CPE
in infected tissue cultures or by  fluorescent antibodies,   The latter
stain the cells containing vi>a!  antigens, thus enabling identifcation
of viruses in the cell  considerably before the visible CPE,   It is
noteworthy that poliovirus anf.gens were already demonstrated approxi-
mately 6 hours post infection and  reached their maximal  concentration
3 hours thereafter.  Thus, by using fluorescent antibodies it is
theoretically possible to determine the presence or absence of polio-
virus in tissue cultures 6-9 hours after infection.  The FA technique
was suggested as a method for typing polioviruses in 1959  and it was
                                          124
later established as a reliable technique- ''  '   The value of this
technique for the identification of poliovirus is beyond dispute^

The FA method requires a special  microscope,  not suitable for tissue
cultures grown m bottles or on plates.   A method of micro tissue
cuUures on microscope slides was  therefore developed,   The cells
were inoculated with the virus while still in suspension and measured
drops (0 02 ml) were placed on a slide thus making quantitative deter-
mination of the virus possible,.
BGM cells,  used routinely in our laboratory for virus Isolation from
water, were employed throughout this study   It is reasonable that
other types of cells may be suitable here, although the number of
cells necessary to obtain a monolayer within a short period of time
should then first be determined   The number of 2-3 x 10  cells/ml  was
chosen with BGM cell* after trial  and error.   This number may not apply,

                                   160

-------
however, to other cell types.  Using BGM cells in the micro tissue
culture method resulted in a monolayer after 4 hours.  Twenty-four
hours after the start of the experiment (the time needed for com-
pletion of the test) the cells still formed a monolayer.

Positive cells, stained by the FA technique, were already observed after
incubation of 9 hours, but the results were not yet sufficiently quanti-
tative.  Furthermore, the finding and identifying of positive cells at
a time when they are still scarce, enhanced the danqer of 'false positive1
or 'false negative'  identification.   On the other hand, after incubation
of 18-24 hrs, the cell culture was at the end of the second virus growth
cycle e.g., the progeny of each primarily infected cell had, in turn,
infected the neighboring cells.  Viewing these FA stained preparations,
under the microscope demonstrated groups of positively stained cells.
It is extremely simple to identify such cell groups, and even a relatively
inexperienced person could easily recognize them.  The results after
18-24 hrs incubation are almost identical to those obtained by the plaque
count technique, with the difference that the latter method requires
two additional days.

The rapid method could be utilized for various virological procedures and
is not necessarily limited to water testing.  Since our motivation was
the virological examination of water, the proposed method was tried in
combination with various techniques for concentration of viruses from
                                                             12 14
water.  Two accepted techniques, based on membrane filtration  '   were
chosen for this purpose.  They are simple to perform and have a high
concentration factor  of 1,000-10,000.
                                  161

-------
A 40-liter volume was concentrated to 7 ml  of fluid.   But even this  small
volume required scores of slides with the micro tissue culture system.
Ten slides resulted in the demonstration of approximately 10 pfu/40  liters,
which was the upper limit of sensitivity.  Techniques with a higher  con-
centration factor would enhance the sensitivity of the rapid method  and
simplify the procedure.  For practical  routine application, 0.5 ml
final concentrate would be most suitable.

The rapid method and the accepted plaque technique, using concentrated
water samples, yielded nearly identical numbers of viruses.  Furthermore,
apart from the fact that the rapid method showed results 2-3 days before
the plaque technique, plates with the latter were often contaminated with
bacteria, thus prohibiting reading of results.   The incubation period in
the rapid method was too short to allow contamination.

The objective of the study was to develop a rapid method for the
isolation and identifcation of viruses from water.  A time interval
of 18-24 hours should satisfy most requirements for water distribution.
It should be pointed out that, in fact, this rapid method allows for
the demonstration of viruses in the water sample after only 6-9 hours of
incubation.  However, the possibility of 'false negatives' or 'false
positives' has then to be taken into account, probably due to the
impurities in the reagents used in the FA technique.  Improving the
quality of the reagents should overcome these obstacles and may enable
virus identification in water within 9 hours or less.
                                    162

-------
REFERENCES
1.    Brown,  G.C.   Fluorescent Antibody  Techniques for the Diagnosis of
     Enteric Infections.   Arch.  Ges.  Virusforsch. 13:30-34, 1963.

2.    Hatch,  M.H., S.S.  Kalter and  G.W.  Ajello.   Identification of
     Poliovirus Isolates  with Fluorescent  Antibody.  Proc. Soc. Exp.
     Biol. Med. 107:1-4,  1961.

3.    Kalter, S.S., M.H. Hatch, and G.W.  Ajello.  The Laboratory
     Diagnosis of Poliomyelitis  with  Fluorescent Antibodies.  Bacteriol
     Proc. p. 89-90, 1959.

4.    Riggs,  J.L., and G.C. Brown.   Application  of Direct  and  Indirect
     Immunofluorescence for Identification of Enteroviruses and
     Titrating Their Antibodies.  Proc. Soc.  Exp. Biol. Med.
     110:833-837, 1962.
5.    Shuval, H.I., S. Cymbalista,  B.  Fattal  and N.  Goldblum.
     Concentration of Enteric Viruses in Water  by Hydro Extraction
     and Two Phase Separation.  In: Transmission of Viruses by the
     Water Route, Berg, G., (ed.). New York,  Interscience Publ.,
     1967. p. 45-55.
6.    Shuval, H.I., B. Fattal, S. Cymbalista,  and N. Goldblum. The
     Phase Separation Method for the Concentration  and  Detection  of
     Viruses in Water.  Water Res. 3:225-240, 1969.
                                 163

-------
 7.   Barren, A.L., C.  Olshevsky,  and M.M.  Cohen.   Characteristic  of
      the B.G.M.  Line of Cells from African Green  Monkey  Kidney.
      Arch. Ges.  Virusforsch.  32:389-392,  1970.

 8.   Shuval, H.I., A.  Thompson,  B.  Fattal, S.  Cymbalista,  and Y.  Wiener.
      Natural Virus Inactivation  Processes  in Seawater.   J. San. Eng.
      Div.  ASCE 97:587-600, 1971.

 9.   Katzenelson, E.,  and H.  Bernkopf.   Studies of the P.L.T. Agents
      with  the Aid of the Agar Immunodiffusion  Technique.   Amer. J.
      Ophthal. 63:1483-1487,  1967.
10.   Katzenelson, E.,  and H.  Bernkopf.   Serologic Differentiation of
      Trachoma Strains  and other  Agents  of  the  P.L.T. Group with the
      Aid of the  Direct Fluorescent Antibody Method.  J.  Immunol.
      94:467-474, 1965.
11.   Nairn, R.C.  Fluorescent Protein Tracing.  3rd Ed.  London,
      E. and S. Livingstone, Ltd., 1969.

12.   Rao,  N.U.,  and N.A.  Labzoffsky.  A Simple Method for  the Detection
      of Low Concentrations of Viruses in  Large Volumes of  Water by
      the Membrane Filter Technique.   Can.  J. Microbiol.  15:399-403,
      1969.

13.   Berg, G., D.R. Dahling,  and  D.  Berman. Recovery of Small
      Quantities  of Viruses from  Clean Waters on Cellulose  Nitrate
      Membrane Filters.  Appl. Microbiol.,  22:608-614, 1971.
                                   164

-------
14.    Sobsey, M.D.,  C.  Wallis,  M.  Henderson,  and  J.L.  Melnick.
      Concentration  of  Enteroviruses  from  Large Volumes of Water.  Appl.
      Microbiol.  26:529-534,  1973.

15.    Shuval, H.I.   Detection and  Control  of  Enteroviruses in  the
      Water Enteroviruses in  the Water Environment  Development.  In:
      Water Quality  Research.   Ann Arbor-London,  Ann Arbor Humphrey
      Science Publ.  , 1970.  p. 47-71.

16.    Sabin,  A.B.  Recent Studies  and Field Tests with a  Live  Attenuated
      Polio Virus  Vaccine.  First  International Conf.  on  Live  Polio
      Viruses Vaccines.  Pan  American Sanitary Bureau, 1959. p.  14-33.
17.    Lwoff,  A.,  R.  Dulbecco, M. Vogt, and M. Lwoff.   Kinetics of  the
      Release of  Poliomyelitis  Virus  from  Single  Cells.   Viroloay.
      1:128-139,  1955.
                                  165

-------
                              SECTION IX
                     VIRUS TYPES IN ISRAEL SEWAGE

INTRODUCTION
Poliovirus and other enteroviruses are present in sewage during epi-
demic and nonepidemic periods (Ramos Alvarez et al).   In Israel where
oral Sabin vaccination is administered routinely it is to be expected
that poliovirus can be isolated regularly from sewage.  In the period
1968-1970, we isolated and typed enterovirus from sewage of several
communities in Israel in order to gain information of the general  fre-
quency of enterovirus types which were detected by the methods we
were studying.
MATERIALS AND METHODS
Virus Assay
The sewage samples were collected by grab and gauze pad methods.
Grab samples were concentrated by the phase separation method  (Shuval
       2
et al),  and assayed for enteroviruses on primary monkey  kidney
monolayers by the plaque forming method  (pfu).  In  the case of the gauze
pad method, the liquid was sterilely expressed from the pad and assayed.
Cell Culture
Primary monkey kidney cells, Rhesus or Vervet, obtained from the
Ministry of Health Virus Laboratory, Oaffe-Tel-Aviv, were trypsinized
and maintained in M-199 medium with the  addition of 10% calf serum.
                                  166

-------
Virus Typing
Two methods were used:
1.  Picked Virus - A number of plaques from positive samples  were trans-
ferred to individual tubes which were frozen and kept at -20°C for
identification.  Virus typing procedures were similar to those used in
                                                    o
routine clinical virus typing work (Lennette et al).    For the purpose
of this study, viruses were classified in gross categories only.  They
were either identified as one of the three types of poliovirus (P,, P2,
P.,) or, when found negative in neutralization tests against poliovirus
antisera but producing paralysis when injected in suckling mice, they
were classified as Coxsackie virus type B or Echovirus type 9.  Strains
found negative with polio antisera as well as in suckling mice, but
exhibiting a cytopathogenic effect (CPE), were considered as possible
strains of Echovirus (ECHO) or other unidentified enteroviruses.

Preparation of Antipolio Serum - Antiserum against poliovirus types 1,
2 and 3 was prepared by intramuscularly injecting a rabbit with 1 ml
                                                     g
of a solution composed of 0.5 ml poliovirus (titer 10  pfu/ml) and 0.5 ml
 complete    Freund's adjuvant (Difco Laboratories).  The two components
of the suspension were mixed just prior to injection.

Two booster doses of the same composition were injected at intervals of
14 days.  Blood samples were withdrawn from the ear 7-10 days after
each booster in order to determine the titer.  When the titer was high,
larger amounts of serum were withdrawn directly from the heart with a
                                   167

-------
50 ml syringe.  The serum was separated from the samples, inactivated at
56°C for 30 minutes and subsequently stored at -20°C until further use.
2.  Neutralization Test for Determination of Poliovirus in Sewage
A 0.15 ml volume of sewage sample, diluted to contain not more than 20
pfu of virus, was mixed with 0.15 ml rabbit serum containing antibodies
against poliovirus types 1, 2 and 3, and diluted to 4 times the con-
centration necessary for neutralization of the amount of virus present.
The control contained the same diluted sewage as well as an equal volume
of M-199.  After incubation at 36°C for 60 minutes, 0.3 ml of the mixture
was seeded on a plated monolayer of monkey kidney.   The results are ex-
pressed as pfu/ml.  The amount of poliovirus in a 0.15 ml sample is the
difference between the amount of virus neutralized in the sewage sample
and the amount of virus in the control sample.   The neutralization test
                                                                 p
was used for sewage grab samples concentrated by phase separation  as
well as for sewage expressed from gauze pads Immersed in sewage for 24
hours.
RESULTS
Virus Types Picked from Sewage
The results of virus strains picked from sewage at different places in
Israel are shown 1n Table 26.   During 1968-1970, 489 plaques were
isolated.  Seventy-four percent of all strains  proved to be polio-
viruses:  25% polio 1, 12% polio 2 and 37% polio 3.  Thirteen per-
cent of the poliovirus strains showed a strong  CPE  when incubated at
40°C.  The remaining 87% grew at 37°C only.  The latter are considered
                                  168

-------
Table 26.  VIRUS TYPES PICKED FROM SEWAGE AT DIFFERENT PLACES AND COMMUNITIES IN ISRAEL
KIRIAT SHEMONEH
Virus Raw Sewage
type No. %
Polio I
Polio 2
Polio 3
Cox. B
or Echo 9
Echo or
Other
TOTAL
20
7
36
13
29
105
19.0
6.7
34.3
12.4
27.6
100.0
Raw Sewaqe
No. %
15
19
34
11
11
90
16.7
21.1
37.8
12.2
12.2
100.0
TIBERIAS
Imhof
No. %
23
7
29
7
8
74
31.0
9.5
39.2
9.5
10.8
100.0
Effluent
No. %
53
18
68
10
15
164
32.3
11.0
41.5
6.1
9.1
100.0
JERUSALEM
Raw Sewage
No. %
11
2
11
2
13
39
28.2
5.1
28.2
5.1
33.4
100.0
BET SHEMESH
& RAMLEH
No. I
1
1
0
1
4
7
14.3
14.3
' 0
14.3
57.1
100.0
HAIFA
No. %
1
3
3
3
0
10
10.0
30.0
30.0
30.0
0
100.0
Total
No.
124
57
181 ..
47
80
489
c;
25.3
11.7
37.0
9.6
16.4
100.0

-------
attenuated poliovirus vaccine strains.   Ten percent of the strains  were
Coxsackie Type B or Echo 9 and 16$ were Echovirus or other virus  strains.

Polio and Non-Polio Viruses Tested by Neutralization
During the period 11  June 1970 through  July 1970, 52 samples  were taken
from the main sewage pipeline of Kiriat Shemoneh, a settlement in northern
Israel.  Twenty-six grab samples and 26 gauze pads, which  were held for
24 hours, were taken daily.  These samples  were tested for polio  and non-
poliovirus by the neutralization test.   The average results are in  Table 27.

Table 27.  POLIO AND NON-POLIOVIRUS IN  GRAB AND GAUZE PAD  SAMPLES IN
           KIRIAT SHEMONEH SEWAGE BY NEUTRALIZATION TEST

No.
26
S.D.
26
S.D.
Sam pi i ng
method
Grab

Gauze
Pad
Ave. no. of
enterovirus
7364 pfu/L
+9559
17543 pfu/pad
+17197
Ave. no. of
non-polio
5111 pfu/L
+8440
12097 pfu/pad
+12247
%
Poliovirus
36.9
+19.9
37.0
+28.1
From Table 27 it can be seen that the amount of the total  non-poliovirus
in grab as well as in gauze pad samples was higher than that of the total
amount of poliovirus.  An average of 37% of the daily grab and gauze pad
samples was identified as poliovirus.  These findings are  lower than
those found in Table 26.    In Table  26, 60% of the picked plaques of
Kiriat Shemoneh sewage was identified as poliovirus.
                                   170

-------
ID
cr

>

o


o
Q.
100



 90



 80



 70



 60



 50



 40



 30



 20



  10
                          % POLIO IN GRAB


                         % POLIO IN GAUZE
        12
              I  Q   I
                 20    24   28
10
              JUNE    1970          JULY 1970


              DATE   OF  GAUZE   REMOVAL


            Fig. 10. The % of poliovirus in daily grab sampling vs.

                  daily gauze pad in sewage by neutralization test
                          171

-------
In Figure 10  we compare poliovirus in daily grab samples with gauze pad
samples kept for 24 hours on the same day and from the same sewage.  From
this figure it can be seen that the percentage of the amount of polio-
virus from daily gauze pads is not clearly correlated to that of the
daily grab sampling.  However, the average percentage of poliovirus found
in 26 gauze pads and 26 grab samples was identical (37%).

DISCUSSION
Isolation and identification of enteroviruses in sewage of a particular
community is very important from an epidemological point of view because
                                                                     4-9
it depicts the general situation of enteroviruses in that population.
They also can serve as a forewarning of an approach!na enterovirus
epidemic or of a carrier of the same.

In this study we isolated and  identified enteroviruses in sewage from
various communities in Israel.  Our findings show that the percentage of
poliovirus found by neutralization tests as well as by picked plaques is
                                                   12
higher than that found by other researchers.  Palfi   found 19% polio-
                                                         13
virus from 317 isolated enteroviruses and Horbowska et al   found 22%
polioviruses from 139 enteroviruses.

In Figure  10, it can be seen that in using the gauze pad sampling or
grab sampling, there are daily fluctuations in poliovirus levels, and
that in almost every sample checked, poliovirus was detected, in the
picked plaques as well as in neutralization tests.
                                   172

-------
 These  findings  seem  to  be  reasonable  in  Israel  sewaqe where Sabin polio
 vaccinations  are  administered  routinely  to  infants,  four  times within
 their  first year.

 It  is  difficult to reach a  significant conclusion  from  these results,
 because  of the  variance of  enterovirus concentration in sewage due to
                      14
 the change of season,   polio  vaccinations  administered within a short
 period of time,  or to enteroviral outbreak epidemic.    It is worthwhile
 to  emphasize  that our field and  laboratory  methods were selective for
 poliovirus.   The  concentration method,   '   the tissue  culture cells,
 as  well  as the  short, incubation  time  are not able  to detect certain
 enteroviruses,  such  as  Reoviruses,Adenoviruses  or  Rhinoviruses,  and  this
 can explain the high percentage  of  polioviruses that we found in the
 sewage.  Regular virus monitoring of sewage  can  give  a reliable picture
 of  enteroviruses  in  sewage  as  well  as of the prevalence of enteroviral
 diseases in the population.
 REFERENCES
1.    Ramos-Alvarez,  M.  and A.B.  Sabin.   Intestinal Viral  Flora  of
      Healthy Children Demonstrated by Monkey Kidney Tissue Culture.
      Amer.  J. Public Health.  46:295, 1956.

2.    Shuval,  H.I.,  B. Fattal,  S. Cymbalista and N. Golciblum.   The
      Phase-Separation Method for the Concentration and Detection of
      Viruses  in Water.   Water  Res.  3:225-240,   1969.
                                  173

-------
3.     Lennette,  E.H.  and N.  Schmidt.   Diagnostic  Procedures  for  Viral
      and Rickettsial  Infections.   American  Public Health  Assoc.  Inc.,
      New York,  4th Edition, 1969.
4.     Horstmann, D.M., J.  Emmons,  L.  Gimpel,  T. Subrahmanyan and J.T.
      Riordan.   Enterovirus  Surveillance Following a  Community-wide
      Oral Poliovirus Vaccination  Program:   A Seven-Year Study.   Amer.
      J. Epidemic!. 97(3):173-186,  1973.
5.     Pittler,  H.,  W.  Hopken, and  K.W.  Knocke.  Enteroviruses and Adeno-
      viruses in Sewage:  Demonstration and  Epidemiological  Study 1963-
      1965.  Z  BAKT Orig.(Germany).  204:33-48, 1967.
6.     Wiley, J.S.,  T.D.Y.  Chin,  C.R.  Gravelle and S.  Robinson.   Entero-
      viruses in Sewage During a Poliomyelitis  Epidemic. In: Transmission
      of viruses by the Water Route,  Berg,  G. (ed.)   New York, Wiley and
      Sonss 1967.   p. 168.
 7.    Wilterdink,  J.B., H.T. Weiland,  and J.D. Verlinde.   A Longitudinal
      Study on  the Significance of Examining Sewage for the  Presence of
      Polioviruses in the Population.  Arch. Ges. Virusforsch. 32:82-90,
      1970.
 8.    Tarabcak, M.,  I.  Kratochvil, and A. Milosovicova.  Effect  of Vacci-
      nation with  Live Poliovirus  Vaccine on the Circulation of  Entero-
      viruses in the  Population.   J. Hyg. Epidemiol. Microbiol.   Immun.
      15(7):258-266,  1971.
                                   174

-------
 9.     Wilterdink,  T.B.,  H.T. Weiland  and T.D. Verlinde.  A Longitudinal
       Study on the Significance  of  Examining Sewage for the Presence of
       Polioviruses in  the  Population.  Arch. Ges. Virusforsch. 32:82-90,
       1970.

10.     Kuwert,  E.  P.G.  Hohler,  and C.A. Primaveri.  Intratypic Differ-
       entiation of Poliovirus  Strains  Isolated  from Sewage:  Consequences
       for the  Practice of  Oral Vaccination.  ZBL BAKT Orig. (Germany).
       215:16-25,  1970.

11.     Nelson,  D.B., R.  Circo,  and A.S. Evans.   Strategic Viral Surveil-
       lance of Sewage  during and Following  an Oral Poliovirus Vaccine
       Campaign.  Am. Jour, of  Epidemiology.  86:641-652, 1967.

12.     Palfi, A.B.   Virus Content of Sewage  in Different Seasons in
       Hungary.  Acta Microbiol.  Acad.  Sci.  (Hungary). 18(4):231-237,
       1971.
  13.   Horbowska, H., H.  Wielopolska, H. Krolak.   Virologic Studies on
       Sewage in Warsaw  in  the  Years 1966-1971.  Presegl Epidemiol.
       (Poland).  27(3):103-108,  1973.
  14.   Melnick, J.L., J.  Emmons,  J.H.  Coffey, and H. Schoof.  Seasonal
       Distribution of  Coxsackie  Viruses in  Urban Sewage and Flies.  Am.
       J.  Hyg.  59:164-184,  1954.
                                    175

-------
15.   Zdrazilek,  J.,  K.  Zacek,  J.  Chvapil,  V. Mikesova, L. Pokorna, V.
     Tomanova, J.  Tranc,  and J.  Vrabkova.   Virological Surveys for
     Enteroviruses in Prague at  the  End  of 1960  and  in 1969.  CS Epidemiol
     Microbiol.  Immun.  (Czechoslovakia). 20(3):67-72, 1971.

16.   Grindrod J. and D.O.  Oliver.  Limitation  of the Polymer Two Phase
     System for Detection of Viruses.  Archiv. Fur Die Gesamte
     Virusforschung, 28:337-347,  1969.

17.   Duff, M.F.   Isolation of  Ether  Resistant  Enteroviruses from
     Sewage:  Methodology. Appl.  Microbiol.  19(1):120-127, 1970.
                                  176

-------
                                PART B
                  IMACTIVATION  OF VIRUSES  If!  WATER
                               SECTION X
               THE CHEMISTRY  OF OZONE AS A DISINFECTANT
INTRODUCTION
Ozone has been used for the disinfection of water supplies since the
beginning of the century when it was applied to the treatment of water
                     1                          2
for the City of Paris  and also at Nice, France.   At present, there
are nearly 1000 installations in operation, mainly in Europe, but also
including 20 in Canada where the largest is operating on drinking
water supplied to the City of  Cuebec, treating flow rates up to 60 mgd.
Thus, while ozone is known to be an even more powerful disinfectant
             4
than chlorine  very little is known of the chemistry of ozone as a dis-
infectant.  In contrast, thorough studies  have been carried out on the
chemistry of the halogens in water and the disinfection efficiencies of
the different species present in the solution.   It is the purpose of
this section to review the chemistry of ozone in water as it is known
and to consider the potential disinfection efficiencies of ozone and
its probable dissociation species in water.

DISSOCIATION OF OZONE IN WATER
Ozone is known to be an unstable gas that decomposes slowly in the
gaseous phase to ordinary oxygen; the decomposition is slow at ordinary
temperatures and low ozone concentrations but is greatly accelerated
by heat.  The homogeneous thermal gas phase decomposition of ozone
can be described by a very simple mechanism:
                                    177

-------
        M + 0, <—>  00 + 0 + M                  -24.6 Kcal
             3  k2    2
        0 + 03 —>   202                         +93 Kcal
where M represents all the substances (including ozone) present in the
gas phase.  Such substances could be 02, N2, C02, Ne or other "foreign"
gases present in the ozone.6  Each of the above gases has a different
kinetic effect on the activation and deactivation of ozone.  For dilute
solution of ozone in oxygen, the rate of dissociation of ozone can be
written:
                 -d(03)/dt = 2k1k3(03)2/k2(02)
thus, the more dilute the ozone, the slower the dissociation.   Axworthy
and  Benson   found that a sample containing 5% ozone in an atmosphere
of oxygen could be stored at room temperature for nearly two months be-
fore the ozone concentration would fall below 4%.

However, both the mechanism and kinetics of the dissociation of ozone in
water are uncertain.  Weiss  observes that in increasingly alkaline
solutions of ozone in water, absorption in  the ultraviolet decreases,
finally disappearing  in strongly alkaline medium, at -40°C.  On the
basis of this observation,  and the fact that decomposition increases
rapidly with increasing alkalinity, Weiss proposed as  a first step the
reaction:
        03 + OH" —>  02 + H02                                          3
                                 178

-------
being followed by the chain reactions:
          03 + H02 —> 202 + OH                                        4
          03 + OH  —> 02  + H02                                       5
          H02 + H02—> 03  + H20                                       6
          H02 + OH —> 02 + H20                                        7
Weiss therefore calculated that the kinetics of ozone decomposition  should
be a 3/2 rate constant with respect to the ozone concentration.   However,
              Q
Alder and Hill  on the basis of their kinetic studies suggested  a first
order reaction with respect to ozone concentration and proposed  the  follow-
ing mechanism as consistent with their results:
          03 + H20  —>H03+ + OH"                                      8
          H03+ + OH"<~> 2H02                                          9
          03 + H02 —> HO + 202                                       10
          H02 + HO —> H20 + 02                                       11
Recently Gorbenko-Germanov and Kozlova   have investigated the decomposition
of ozone in basic aqueous media (-50°C and 8M KOH) using the electron spin
resonance technique and also absorption spectroscopy.  By comparison with
spectra obtained from potassium ozonide (KO.J and potassium superoxide
(K02), they were able to confirm the presence of the ozonide and superoxide
radical -ion in 8M caustic potash solutions ozonized at 50°C.  '     On
the basis of their studies, they suggested the following mechanism for
ozone decomposition:
                                     179

-------
Step I    (-50°C)     303  +  20H" —>   203"  +  H20  + 202
                                                                     12
Step II   (-50°C)     303" +  H20  —>   02" + 303  + 20H"           13
Step III  (27.5°C)    202" +  H20  —>   H02" + OH"  +  02           14
Step IV   (27.5°C)            H02" —>   OH" + %02                   15
the overall reaction being 20-
                                     30
Gorbenko-Garmanov and Kozlova   also suggested that the Step I was
probably initiated by the fol lowing reaction:
                         03 + OH"  — •>  03"  + OH                    16
the hydroxyl radical then reacting further.  This latter stage would
explain the increased dissociation of ozone with increasing alkalinity.
Other studies have been carried out on the kinetics of ozone decomposition
in water and Table  28 shows the range of variables covered and the varying
conclusions concerning the reaction order relative to ozone.

Except for the work of Gorbenko-Germanov et al,  '   no direct studies
(in comparison to kinetic study inferences) have been made of the decom-
position of ozone in water.  However, the reaction between water and
                                                                21 22
ozone in the vapor phase has been investigated by Norrish et al,  '
who observed strong OH absorption bands in the flash photolysis of ozone-
water mixtures and proposed that OH radicals are formed by reaction with
the 0 radical:
                                    180

-------
          03 + hv —> 02 + 0                                           17
          0 + H20 —> 20H                                              18
Based on reactions 1  and 18 it appears reasonable therefore to assume that
an alternative mechanism for ozone decomposition in water might be:
          03 + H20  —>02 + 20H                                        19
                      22                23
Both Norrish and Wayne   and also Demore   have suggested that the hydroxy
radical (OH) would then react as follows:
          OH + 03 —> H02 + 02                                          5
to produce the hydroperoxyl radical (H02) which could react:
          H02 + 03 —> OH + 202                                         4
                                                            O
similar to the reaction scheme previously proposed by Weiss.
The reactions of the hydroxyl and hydroperoxyl radicals have been studied
by investigators interested in the radiation chemistry of water.   The
hydroxyl radicals are reported to dimerize to form hydrogen peroxide:
          OH + OH —> H202                                             20
                                            24
and to further react with hydrogen peroxide:
          OH + H202 —>H02 + H20                                       21
The hydroxyl and hydroperoxyl radicals are also reported to react with
                                   25
each other in the following manner:
          OH + H02 —> H20 + 02                                        22
                                   181

-------
In alkaline solutions the following reaction becomes increasingly


          26
important:



          OH + OH" — > 0" + H20                                        23



The oxide radical, in contrast to the hydroxyl  radical, can react with


                               27
oxygen to form the ozonide ion:



          0" + 02 — > 03~                                              24




As stated previously, the ozonide ion has been  identified as an intermediate



decomposition product of ozone in alkaline aqueous media (8M KOH) at below



zero temperatures by investigation using the electron spin resonance tech-



nique.    These authors   also noted the presence of the superoxide radical



ion (02~) which had been previously suggested as a possible decomposition

                           ?ft                         -              o
produce of the ozonide ion.    The half-life of the 03  radical at 25 C


                                                            29
amounts to several milliseconds and is longer with higher pH   and concen-



tration of oxygen in solution.    The mechanism of dissociation of the



ozonide ion is uncertain, although it has been  suggested that it dissociates



to give oxygen and a peroxide ion   or by thermal dissociation back to



the oxide ion (rate constant 3.3 x 103sec"1),   similar to that found by



Garbenko-Germanov and Kozlova.    Like the hydroxyl radical, the hydro-


                                                                         25
peroxyl radicals have also been reported to dimerize in aqueous solutions
as follows:
                H02 — >H202 + 02                                       25
                                                  3
disagreeing with reaction 6 as suggested by Weiss.   Taking into consideration



the above mentioned reactions, the following stages for ozone decomposition
                                      182

-------
Table 28.  SUMMARY OF THE KINETICS OF OZONE DECOMPOSITION IN WATER
Reference
Rothmund et al12
1 3
Sennewald
Weiss8
Adler & Hill9
14
Stumm
Kil patrick et al15
Kilpatrick et al15
Rankas et al
Hewes & Davison
Hewes & Davison
Hewes & Davison
Czapski18
19
Rogozhkin
20
Merkulova et al
pH range
O V.
5.3 — >
O «si
1 — >
7.6 — >
0 — >
13
5.4 — >
2 —>
6
8
10 — >
9.6 — >
0.22 — >

4
8
8
2.8
10.4
6.8

8.5
4


13
11.9
1.9
Temp, range
(°C)
0
0
0
0 — > 27
1.2 -> 19.8
25
25
5 — > 25
30 — > 60
10 •— > 50
10 — •> 20
25
25
5 — > 40
Reaction order
with respect to
°3
2
2
3/2
1
1
3/2
2
3/2
2
3/2 — > 2
1
1
1
1 or 2
                                  183

-------
in aqueous solution can therefore be suggested:
          03 + H20  —>02 + 20H                                        19
          03 + OH2  —>Q2 + H02                                         5
          03 + H02  —>02 + OH                                          4
          OH + OH  —> H202                                            20
          OH + H02 —> H20  +  02                                      22
          H02 + H02 —> H202 + 02                                      23
          OH + OH" —>  0"  +  H20                                     24
          0" + 02  _^  Q3"                                            25
the further dissociation of the ozonide ion probably following reaction
13 to 15.  At high pH values the initiation state could well  be:
       303  + 20H" ~> 203" + H20 + 202                                12
as demonstrated by Garbenko-Germanov and Kazlova,   followed  by reaction
13 to 15.
As can be observed, the decomposition behavior of ozone in water is compli-
cated depending on the alkalinity of the solution and possibly also on the
oxygen content.  All of the intermediate species formed are very reactive
                                               32
and possess very short half-lives. Ivanov et al   have recently also pro-
posed reaction 19 as a chain initiation step.  However, since only an
abstract was available to the author, it is difficult to know on what
bases the proposed initiation step was suggested.
                                    184

-------
DISINFECTION POTENTIALS OF OZONE AND ITS DISSOCIATION SPECIES  IN WATER
From the above survey, the possible species to be found in aqueous ozone
solution are 0-,, OH, HO,,, 0~, 03~ and possibly the free oxygen atom if
the ozone decomposes as in reaction 1 before reacting with the water.
                        •50
While Morris has stated,   and correctly, that there should be no re-
lationship between the oxidation potential of a substance and its
germicidal activity, however, it can be stated that substances that do
not possess a high oxidation potential will not be germidically active.
Conversely, chemical species that have high oxidation potential may
possess germicidal potential.

Since almost no direct studies  (except for H,,02), as yet, have been
carried out on  the germicidal activity of the above species,  it is of
interest  to consider  from a  theoretical point of  view the possible
disinfection potential of the various species present in ozonated
water.  Hydrogen  peroxide can be discarded at once as being the respon-
sible species for the  strong germicidal activity  of ozonated  solutions
                                           34
due to  its  slow disinfection effectiveness.

Ozone has one of  the  highest oxidation  potentials known  (2.07 volts in
acidic  solutions  and  1.24 volts in  basic  solutions).   However,
Baxendale35 gives a  value for the  oxidation  potential  of the  hydroxyl
radical  as  2.8  volts  at  H+  = 1.0 M and  1.7V for  the hydroperoxyl
radical,  which  suggests  that the OH radical  in  water might be the species
responsible for the strong  germicidal  activity  of ozonated solution and
not the free  ozone  itself.
                                    185

-------
Comparison of the reactions of the hydroxyl radical and ozone indicate
a strong similarity.  Ozone is well known to reduce the organic carbon
content of wastewater effluents   and a similar effect has been noted
using the OH radical.    In the latter study, the hydroxyl radical was
produced by using Fenton's reagent which is the reaction of ferrous
ions with excess of hydrogen peroxide:
          Fe2+ + H202 ->Fe3+ + OH" + OH                              21
Likewise, the ozonation of phenol in aqueous solution has been shown to
                                                            38
produce as intermediate products both catechol  and o-quinine   and
similar oxidation products were observed in the reaction between the
                            39
hydroxyl radical and phenol.    In the oxidation of various amino acids
both the ozone   and OH radical   show a preference for sulfur containing
amino acids.  An intermediate product in the oxidation of cysteine by
                                                                   40 41
both ozone and the OH radical has been demonstrated to be  cysteine.  *
One of the few simple carbon-hydrogen bond ozonation reactions which
has been studied in water solution is that of malonic acid
(HOOC)CH2(COOH).42  The ozone attacked the methylene group (CH2) converting
it to an alcohol (hydroxymaIonic acid) and a ketone(ketomalonic acid)
function.  The hydroxyl likewise is known  to react with fully saturated
organic compounds by simple hydrogen abstraction to form water and a
               41
carbon radical.    Thus, the reaction of the hydroxyl radical with malo-
nic acid in aqueous solution could well proceed in a similar manner, i.e.,
          (HOOC)  CH2(COOH) + OH —>  (HOOC) CH  (COOH) +  H20
          (HOOC) CH  (COOH) + OH —> (HOOC) CH  (OH) (COOH)
                                            hydroxy-malonic acid
                                    186

-------
Malonic acid is known to react readily with the OH radical,   however,
no study was made of the resulting by-products.
Recently Hofgne et al,  '   have shown experimentally that as the pH
increases the kinetics of ozonization of organic materials changes and
that the relative reaction rates for the competing oxidation reaction
in alkaline solutions is similar to that obtained in radiation studies,
thus implying that the same species present in irradiated water is
present in ozonated alkaline aqueous solution—namely the hydroxide
ion, probably in the manner suggested by Garbenko-Germanov et al
(reaction 16).

The chemical reactivity of the basic form of the hydroxyl radical, the
oxide radical ion 0", differs markedly from that of the OH in many
reactions.  While the hydroxyl radical adds readily to aromatic molecules,
the reactivity of 0" toward aromatic compounds is lower by at least
three orders of magnitude in the specific rate constant.    The OH
radical is much more effective than is the oxide radical ion in oxidizing
a number of inorganic anions.    Hydrogen abstraction reactions of 0~, on
the other hand, exhibit a specific reactivity only slightly lower than
that of OH.30

The limited studies on the ozonide radical   have shown that it is almost
totally unreactive to aromatic molecules (such as the benzoate  ion) as
well as to methanol and ethanol.  Likewise, the hydroperoxyl radical  (H02)
                                    187

-------
has been shown to be almost inert in aqueous solution towards organic
                                46             47
substances such as ascorbic acid   or cysteine.
There are a number of facts that seem to indicate that the dissociation
product(s) are more potent oxidents than the ozone itself.  Hewes and
Davison   have shown that the speed of ozonation of organic compounds
in wastewater is pH dependent and increases with increasing temperature.
Further, the oxidation process can be catalyzed by adding certain in-
organic cations.  The above factors all  affect the ozone decomposition
rate and thus it appears that it is the decomposition products that
                                                               48
affect the oxidation rate.  Likewise, Reicherter and Sontheimer   have
shown that both pH and catalysts affect the rate of ozone purification
of wastewater systems and that a radical mechanism is probably respon-
sible for ozone oxidation in aqueous solutions.
A limited number of preliminary experiments have been carried out in our
laboratories to investigate the effect of pH on ozone inactivation of
poliovirus I in aqueous solutions.  For those experiments in the acidic
pH range the method used was identical to those described 1n this report (see
Section XIII).    the virus being added to an ozonated solution already
of the required pH.  However, for those measurements carried out at pH
greater than 8 a variation in the above method was necessary due to
the rapid decomposition of the ozone.  A 400 ml sample of unbuffered
water was prepared containing the required amount of ozone.  Then
one millilitre of the virus suspension was added simultaneously with
                                     188

-------
a sufficient quantity of concentrated sodium hydroxide solution.   The
alkaline solution produced the required pH.

The results obtained at the various pH studied (2, 4,   8.5,  9.5,  10)
                                                                   A
showed the same two stage structure as previously reported at pH  7.
The first stage was shorts 10 sec, with a rapid virus  kill,  followed
by a slow second stage.  Increasing the ozone concentration  between
0.08 mg/1 to 0.2 mg/1 had little effect on the inactivation  rate.  A
graph showing inactivation of the virus versus pH after 10 sec and
3 min is shown in Fig. 11.  The results show a definite dependence on
pH with the slowest activation rate occuring at a pH around  4.  A
                                                        48
similar result was obtained by Reicherter and Sontheimer   for the
ozonization of wastewater effluents at various pH values.

The above results would appear to agree with the conclusions of Reicherter
              48
and Sontheimer   and suggest that the pH change affects the  ozone dis-
sociation and that the species thus produced inactivate the  virus.
However, an alternative explanation may be possible.  It is  known that
under various conditions clumping of viruses occurs, and it  is
possible that clumps are broken up in the higher pH solution, and thus
making them more susceptible to ozone inactivation.

In order to evaluate the above point, an inorganic catalyst was added
at pH 4 to see if this affected the inactivation rate.  It was observed
                                    189

-------
VO
o
                  100
                   10
              to
                   1.0
I   O.I
>
or
D


# 0.01
                0.001L
                          I   2   3  4  5  6   7  8   9  10  II   12  13 14

                                           PH   VALUE,
          Fig.  11.  Inactivation of Poliovirus 1  by ozone after 10 sec of various pH values (T=5°C)

-------
that the addition of 10 mg Fe^O^/lOO ml solution greatly increased the
ozone decomposition rate and also the virus inactivation rate.   The
survival rate for the virus was decreased by a factor of about ten.
Control experiments showed that Fe?0., alone did not affect the virus
stability.

In conclusion it therefore appears that it is the decomposition products
of ozone that affect primarily the inactivation rate of viruses.  How-
ever, while the ozone concentration decreases as quickly on addition
of Fe203 as at pH=10, a smaller inactivation rate was observed.  Thus
it may be that at different pH values different species are formed and
these species have different inactivation abilities.
DISCUSSION AND CONCLUSIONS
In conclusion, it appears as if the dissociation products of ozone in
water may be more powerful oxidization agents than ozone itself.  Due
to the similarity between the reactions of organic compounds with ozone
in aqueous solution and with the  hydroxyl radical, it appears  that  it may
be the hydroxyl radical that is mainly responsible for  the high oxida-
tive potential of ozone in water.  Likewise the hydroxyl radical may
be thought of as giving rise to the high germicidal action of  ozone.
                                                                  49
However,  this last statement must be treated with caution.  Hoigne
has shown that the hydroxyl radical will react  primarily with  an
organic solute when this  is in solution together with a microorganism.
Thus while in clean water systems the  hydroxyl  radical  may well be
                                  191

-------
available to inactivate viruses, once organic substances are present
they will preferentially react with any high reactive species (such as
OH) produced by the ozone on decomposition in water.

In summary, more experimental work, such as is presently being carried
out in our laboratories is obviously needed before more definitive
statements can be made about the germicidal properties of ozone and
its dissociation species in water.  An attempt must be made to further
examine the effect of pH changes and increased catalytic decomposition
of the ozone on the disinfection ability of ozone.  Further studies
need to be initiated, if possible, on the direct effect of the dis-
sociation species of ozone on disinfection, such as OH, H02, 03", 02~
and 0".
REFERENCES
 1.   Guinvarch, P.  Three Years of Ozone Sterilization of Water  in
     Paris.   In:  Ozone Chemistry and Technology, Advances  in Chemistry
     Series No. 21, 1959.   p. 416-429.

 2.   Lebout, H.   Fifty Years of Ozonation  at Nice.   In:  Ozone
     Chemistry and Technology, Advances  in Chemistry Series No.  21,
     1959.   p. 450-452.
 3.   Draper,  E.W.J.   Practical Aspects  of  Water and Waste  Water
     Treatment by Ozone.   In:  Ozone Water and  Wastewater  Treatment,
     Evans  III. F.L.  (ed.). Ann  Arbor  Science, 1972.   p.  145-179.
                                  192

-------
4.   Katzenelson, E., B. Kletter, H. Schecter and H.I. Shuval.
     Inactivation of Viruses and Bacteria by Ozone.  In:  Chemistry
     of Water Supply, Treatment and Distribution, Rubin, A.J. (ed.).
     Ann Arbor Science,  1974. p. 409-421.

5.   Kinman, R.N., A.P.  Black, and W.G. Thomas, Jr.  Studies of Water
     Disinfection with  Iodine.  National Speciality Conference on
     Disinfection, Amer. Soc. of Civil Engr., p. 11-35, 1970.

6.   Benson, S.W. and A.E. Axworthy, Jr.  Implications of Data on
     the Gas Phase Decomposition of Ozone.   Advances in Chemistry
     Series No.  21,  1959.  p. 398-404.
7.   Axworthy, A.E.,  Jr.  and S.W. Benson.  Mechanism of Gas Phase
     Decomposition of Ozone.  In:  Thermal  and  Photochemical Reactions
     Advances in Chemistry Series No.  21, 1959.  p. 388-397.

8.   Weiss, J.   Investigations on the  Radical H02  in Solution.
     Trans. Far. Soc. 31:668-681. 1935.

9.   Adler, M.6. and G.R. Hill.  The Kinetics and  Mechanism  of
     Hydroxide  Ion Catalyzed  Ozone  Decomposition in Aqueous  Solution.
     J.  Amer. Chem.  Soc. 72:1884-1886, 1950.

10.   Gorbenko-Germanov, D.S.  and  I.V.  Kozlova.   Mechanism of Ozone
      Decomposition  in Alkaline Aqueous Media.   Dokl.  Akad.  Nauk.
     USSR, 210:851-854, 1973.
                                 193

-------
11.    Gorbenko-Germanov,  D.S.  and I.V.  Kozlova.   Intermediate
      Decomposition Products of Ozone in  Alkaline  Aqueous  Media
      Investigated by Electron Spin Resonance.   Zhur.  Fiz.  Khim.
      48:166-168,  1974.

12.    Rothmund,  V.  and A.  Burgstaller.  Decomposition  Velocity of
      Ozone in Aqueous Solutions.   Monatsch  Chem.  34:665-692, 1913.

13.    Sennewald, K.  Decomposition  of Ozone  in Aqueous Solutions.
      Z.  Phys. Chem.  (Leipzig) A164:305-317, 1933.

14.    Stumm, W.   The Solubility of Ozone  in  Aqueous  Solution.
      Helvetica Chimica Acta,  37:773-778, 1954.
15.   Kilpatrick, M.L.C, C.C. Herrick, and M. Kilpatrick.   The
      Decomposition of Ozone in Aqueous Solutions.   J. Amer. Chem.
      Soc., 78:1784-1789, 1956.
16.    Rankas, M.M., E.K.  Siirode, and S.R. Kyulm.   Rate  of Ozone
      Decomposition in Different Waters.   Trudy Tallinskoga Politek-
      nich  Inst. Ser.  A, 198:219-232.  Chemical Abstracts 60:2633e


17.    Hewes,  C.G.  and R.R. Davison.  Kinetics of Ozone Decomposition
      and Reaction with Organics in Water.  AIChE Journal  17:141-147,
      1971.

18.    Czapski,  6., A. Samuni, and R. Yell in.  The Disappearance of
      Ozone in  Alkaline Solution.   Isr. J.  Chem. 6:969-971, 1968.
                                 194

-------
19.    Rogozhkin,  G.I.   Kinetics  of Ozone Absorption  by  Alkaline
      Solutions.   Chem.  Abs.  73:91108y.

20.    Merkulova,  V.P., V.S.  Lovchikov, and M.D.  Ivanovskii.   Kinetics
      of the Breakdown of Ozone  in Sulfate Solutions.   Izv.  Vyssh.
      Ucheb. Zaced. Khim. Tekhnol. 14:(6) 818-823, 1971.

21.    McGrath, W.D. and R.G.W.  Norrish.   Influence of Water  on the
      Photolytic Decomposition of Ozone.  Nature 182:235-237, 1958.
22.    Norrish, R.G.W.  and R.P.  Wayne.  The Photolysis  of Ozone by
      Ultraviolet Radiation II.   The Photolysis of Ozone Mixed with
      Certain Hydrogen-Containing Substances.  Proc. Roy. Soc. Ser.
      A. 288:361-370, 1965.
23.    Demore, W.B.   Rate Constants for the Reactions of Hydroxyl  and
      Hydroperoxyl  Radical with Ozone.   Science,  180:735-737,  1973.

24.    Thomas, J.K.   Rates of  Reaction of the Hydroxyl Radical.  Trans.
      Far.  Soc.,  61:702-707,  1965.

25.    CzapsKi, G. and B.H.J.  Biehki.  The Formation  and Decay of  H203
      and HOp in Electro-Irradiated Aqueous Solutions.   J. Phys.  Chem.
      67:2180-2184, 1963.

26.    Hochanandal,  C.J.   Photolysis of Dilute Hydrogen  Peroxide
      Solutions in the Presence  of Dissolved Hydrogen  and Oxygen.
      Radiat. Res.  17:286-301,  1962.
                                  195

-------
27.    Adams, G.E., J.W.  Boag and B.D.  Michael.   Rate Constant for
      Reaction of 0" with Oxygen.   Nature,  205:898-899,  1965.
28.    Felix, W.D., B.L.  Gall and L.M.  Dorfman.   Pulse Radiolysis
      Studies.  IX.Reactions of the Ozonide Ion in Aqueous  Solutions.
      J. Phys. Chem. 71: 384-392,  1967.
29.    Czapski, G.  and L.M.  Dorfman.  Pulse  Radiolysis Studies.  V.
      Transient Spectra  and Rate Constants  in Oxygenated Aqueous
      Solutions.   J. Phys.  Chem. 68:1169-1177,  1964.

30.    Czapski, G.   Pulse Radiolysis Studies in  Oxygenated Alkaline
      Solutions.   J. Phys.  Chem. 71:1683-1688,  1967.
 31.    Gall,  B.L.  and L.M. Dorfman.  Pulse  Radiolysis Studies. XV.
       Reactivity  of the Oxide  Radical Ion  and  of  the Ozonide Ion in
       Aqueous  Solution.  J. Amer.  Chem. Soc. 91:2199-2204, 1969.
 32.    Ivanov,  Yu.  E., G.P.  Nikitina, M.F.  Pushlenkov and V.G.
       Shunkov.  Ozone in Aqueous Solutions.  I. Kinetics and Mechanism
       of  the Decomposition  of  Ozone in Water and  in  Nitric Acid
       Solutions.  Zhur.Fiz.  Khim. 46:2163,  1972.
 3?.    Morris,  J.C.   Disinfectant Chemistry and Blocidial Activities.
       In:   National  Specialty  Conference on  Disinfection.  Amer.  Soc.
       of  Civil Eng., 1970.  p.  609-633.

 34.    Yoshpe-Purer, Y.  and  E.  Eylan.  Disinfection  of Water  by Hydrogen
       Peroxide.   Health Laboratory Science 5:233-238, 1968.
                                  196

-------
35.    Baxendale, J.H.   Effects  of Oxygen and pH in the  Radiation
      Chemistry of Aqueous  Solutions.   Radiat.  Res.  Suppl.  4:114-135,
      1964.

36.    Hewes, C.G.  and  R.R.  Davison.   Renovation of Wastewater by
      Ozonation.  AIChE Symposium Series 69:71-80, 1972.

37.    Bishop, B.F., 6.  Stern, M.  Fleischman, and L.S.  Marshall.
      Hydrogen Peroxide Catalytic Oxidation of Refractory Organics  in
      Municipal Waste  Waters.  I&EC Proc. Des.  & Develop. 7:110-113,
      1968.
38.    Eisenhauer, H.R.   The Ozonization of Phenolic Wastes.  J. WPCF.
      40:1887-1899, 1968.

39.    Stein, G. and J.  Weiss.  Some Free Radical Reactions of Phenol.
      The Action of Hydrogen Peroxide-Ferrous Salt Reagent and of
      X-Rays on Aqueous Solutions of Phenol.  J. Chem.  Soc.  3265-3274,
      1951.

40.    Mudd,  J.B., R. Leavitt, A.  Orgun, and T.T. McManus.  Reaction of
      Ozone with Amino Acids and Proteins.  Atmos. Envir. 3:669-681,
      1969.

41.    Dorfman, L.M. and G.E. Adams.  Reactivity of the Hydroxyl
      kadical in Aqueous Solutions.  U.S. Department of
      Commerce, National Bureau of Standards, NSRDS-NBS46.  1973

42.    Dobinson, F.  Ozonation of Malonic Acid in Aqueous Solution.
      Chemistry and Industry, 1959. p. 853-854.
                                  197

-------
46.   Barr, N.F. and C.G. King.  The X-Ray Induced Oxidation of Ascorbic
      Acid and Ferrous Ion.  J. Amer. Chem.  Soc.  78:303-305, 1956.

47.   Swallow, A.J.  The Action of X-Radiation on Aqueous Solutions
      of Cysteine. J. Soc. 76:1334-1339, 1952.

48.   Reicherter, A. and H. Sontheimer.   Investigation of the Use of
      Ozone for Water and Wastewater Purification.  Vom Wasser.  41:369-
      380,  1973.

49.   Hoiqne, J.  Aqueous Radiation Chemistry in  Relation to Waste
      Treatment.  IAEA Symposium, Munich.  SM-194-401.   March, 1975.
43.   Scholes, G. and R.L. Wilson.   Radiolysis of Aqueous Thymine
      Solutions.  Determination of Relative  Reaction Rates of OH
      Radicals.   Trans. Far.  Soc. 63:2983-2993, 1967.

44.   Hoigne, J. and H. Bader.   The Rate of  the Hydroxyl  Radical  in
      the Ozonization of Aqueous Solutions.   Chimica.  29:20, 1975.

45.   Hoigne, J.-  Ozone vs Irradiation  Treatment of Organic
      Impurities in Water.  IAEA Symposium,  Munich.   SM-194-410.
      March, 1975.
                                  198

-------
                             SECTION  XI
                    SPECTROPHOTOMETRIC METHOD FOR THE
               DETERMINATION OF OZONE IN AQUEOUS SOLUTIONS

INTRODUCTION
IhG "idin reasons for developing methods of ozone determination were to
study the natural occurence of ozone in the atmosphere and its presence
as b result of air pollution.  The methods for the detennination of
aunospheric ozone involve a variety of analytical techniques.  Chemical,
electrochemical, ''  and optical methods     have been developed, each of
theiii with its advantages ind disadvantages.

Although the use of ozone as a disinfectant for water was reported on
as early as 1895,   its application in water purification has been
                               1 o
more or less limited to Europe.    At present, ozone is actively being
investigated for use in water and waste water treatment.  Although
this technique is very effective, so far no rational and scientific
basis has been given for its practical application.

The  impetus to develop a suitable procedure for the determination of
dissolved ozone  in water arose from the requirement of the kinetic
study carried out in this laboratory for the inactivation of  viruses
and  bacteria by  ozone.  For  this purpose,  the method must be  able to
detect  ozone concentrations  as low as 0.01 ppm  and  should also  use
small samples in order to overcome the  problem  of  overly  large  volumes
                                  199

-------
in the inactivation reactor vessel.   The method should be sufficiently
rapid so as to allow for frequent sampling in the course of kinetic
experiments.

Since many methods for ozone determination in air have been previously
employed, a survey of the pertinent literature has been made and the
iodide chemical method selected.   The iodometric method is accepted as
the method of reference, since it agrees with the absolute physical
                            13                                 14
methods based on gas density   or pressure change measurements.

The classical procedure described in Standard Methods   for the
determination of residual ozone in water, requires large volumes of
sample under normal conditions and would require samples as large as
1 liter for the detection of ozone concentration as low as 0.03  ppm.
Moreover, thiosulfate titration of iodine liberated by ozone has also
some limitations.

To avoid these difficulties, the spectrophotometric method initially
reported upon by Salzman   for the determination of ozone in small
volumes of water was evaluated and refined.   This involves oxidation
of a buffered iodide solution and spectrophotometric measurement of
the triiodide ion liberated by ozone.  The concentration of iodide and
the pH of the solution have a marked effect on the results.
                                 200

-------
The recommended method corresponds to the requirements mentioned earlier
for our experimental conditions.   In addition, it has the advantage of
allowing the preparation of a calibration graph with a stable iodine
solution, instead of using the unstable and reactive solution of
ozone in water, as a standard.

MATERIALS
Apparatus
Ozonizer -
The ozone is generated from air in a Fischer-Laboratory Ozonator
(OZ III) based on the electrical discharge method.  An appropriate
suction pump is used for drawing the air through the ozonator.  The
air stream  is dried initially by means of a "Koy Senior" air puri-
fier and flow equalizer, and by means of two  columns containing
granulated  potassium hydroxide and silica gel, respectively.  The
ozone generated by  this equipment is passed through  a  sintered glass
dispersion  tube (Corning grade) into water, contained  in a two
liter Pyrex glass reactor.   The tubing used is of Teflon or glass,
since these materials have  the  least effect on ozone.

Spectrophotometer -
Zeiss P. MQII,  equipped with stoppered 20  and 40 mm quartz cuvettes,
suitable  for  use  in the U.V.  region.

Glassware  -
 It is very important  to keep the  glassware scrupulously clean,  since
 traces  of  impurities  may  cause  very serious  errors.  In the  preparation
                                   201

-------
of ozone-demand free glassware, all the glassware used in the tests is
cleaned with dichromate-concentrate sulfuric acid mixture, rinsed in tap
water, then by distilled water, soaked in a strong ozone solution and
finally dried at 180°C to ensure the elimination of ozone.  Only exper-
iments conducted in a clean, ozone demand free system, would expose
viruses to a relatively constant concentration of ozone during the
course of the experiment.

Reagents
Chemicals -
Highest  quality analytical  grade chemicals are used.
Distilled Water -
Distilled water of high purity must be used; demineralized water from a
mixed-bed ion exchanger is not suitable for this work.  The distillation
must be carried out in an all glass still, in the presence of alkaline
potassium permanganate and redistillation is necessary.  The dilution
water used has to be ozone-demand free.  For this, it is convenient to
ozonize bidistilled water and then to ensure its dissipation by boiling.

Standard iodine solution 0.01 N -
Dissolve successively 6.4 g potassium iodide and 1.2692 g iodine, to a
volume of 1000 ml.  It is advisable to mix the solid  iodine and pota-
ssium iodide in a relatively small volume of water until the Iodine
is completely dissolved, then to dilute the solution to 1000 ml.
Age one day before use.  The solution may be standardized by titration
                                   202

-------
with arsenious oxide.  One ml of 0.01  N iodine (triiodide) is equivalent
to 24 yg ozone.  A standard iodine solution 0.01  N may be alternatively
prepared from an ampoule.

Neutral potassium iodide reagent -
Dissolve 13.61 g potassium dihydrogenphosphate, 14.20 g anhydrous di-
sodium hydrogen phosphate, and 20.0 g  potassium iodide.  Bring mixture
to 1000 ml.  This solution must be stored in a dark bottle in the
refrigerator and must avoid exposure to sunlight.  This reagent is used
in the procedure "B" for high ozone concentrations.  In the procedure
"A" for low ozone concentrations, the  same reagent is used, but with a
higher concentration of potassium iodide:  5%, instead of 2%.

METHOD
Ozone reacts with the neutral potassium iodide solution and liberates
iodine and in an excess of potassium iodide, iodine is in the complexeH
triiodide form.  The concentration of  triiodide liberated is determined
spectrophotometrically at a wave length of 352 my.

Procedure
Two similar procedures are recommended:  procedure "A" for low ozone
concent*ations (0.01 to 0.30 ppm) and  procedure "B" for high concen-
trations (0.30 to 2.0 ppm).

In procedure "A", 10 ml of the ozone containing samples are introduced
into a test tube which contains 2 ml of 5% neutral potassium iodide
                                  203

-------
1.0

iu 0.8
o
£ 0.6
QC
O
GO
< 0.4
0.2
o
: V x
o X
/ /
/ y
f\* ^r
S
/ /
f ^r
/o •'
/ /
/\/
" .^^
rf* 1 1 1 1











A 0 O.I 0.2 0.3 0.4 0.5
* cm
B 0 0.5 1.0 1.5 2.0 2.5
2 cm . . .
        OZONE   (ppm)
Fig. 12.  Calibration curve for determination of
       Ozone in water
            204

-------
buffered reagent.  The test tube is left for about 30 minutes in a cool
and dark place.  Triiodide is quantitatively liberated and the intensity
of absorbance is then read, using cells of 40 mm light path, against a
distilled water blank.

In procedure "B" for higher ozone levels, 5 ml of the sample are intro-
duced in a test tube containing 5 ml of 2% neutral potassium iodide.
After about 30 minutes, the intensity of the absorption is read using
cells of 20 mm light path.

Calculation
The stoichiometry of the ozone iodide reaction, upon which all calcu-
lated ozone concentrations are based, is 1:1, one mole of ozone
liberates one mole of iodine at neutral pH.    '
Calibration graph
A standard curve of absorbance versus iodine  (or ozone) concentration
is plotted from readings of a series of freshly prepared standards.
The stock iodine solution 0.01 N ( 1 ml = 240 yg of ozone) is diluted
with neutral potassium iodide reagent to obtain a solution of iodine
0.00004 N ( 1 ml = 0.96 ug ozone).  Different portions of this solu-
tion are transferred to test tubes and diluted with neutral potassium
iodide reagent to 12 ml final volume (curve A) or to 10 ml final volume
(curve B).  Mix and immediately read the absorbance at 352 my, using
distilled water as the reference.  Calibration graphs for both
procedures are shown in Fig. 12.   Beer's Law is obeyed over  the whole
range investigated in both cases.
                                  205

-------
RESULTS AND DISCUSSION
Procedural Variables
An evaluation of several nrocedural variables was undertaken in order
to ascertain the effects they would have on ozone determination.

Concentration of Potassium Iodide Reagent
The effect of different concentrations of potassium iodide on the intensity
of the colour development was investigated.  Table 29 presents the
absorbance values at  different ozone levels utilizing solutions of 2%,
5% and 10% potassium  iodide concentration.  The obtained values are
comparable, but there are increasina absorbance values with higher
potassium  iodide concentrations, especially at the low ozone concentration
levels.

Changing  the concentration of the  phosphate buffer affects the
                                                 19
stoichiometry of the  ozone-iodide  reaction, Scott   found lower ozone
values by  using a weaker buffered  solution, possibly by allowing a
rise  of the pH during the reaction  of  ozone and  potassium  iodide.

Stability  of the Colour Intensity
The  rate  of colour  development was  also  checked  at five minute  intervals
between time zero and one hour.  The  results  are presented in  Fig. 13.
By  using  a 2% potassium iodide solution, the  intensity of  the  colour
increases  slightly  during the first 25-30 minutes and  then begins  to
fade  slowly.  More  than 90%  of the iodine is  liberated at  time  zero
                                    206

-------
        Table 29.  INFLUENCE OF POTASSIUM IODIDE CONCENTRATION ON



                          INTENSITY OF COLOUR
Potassium Iodide
concentration
2%
5%
10%
Absorbance
a*
0.060
0.070
0.075
(Immediate lecture")
b*
0.205
0.210
0.220
c*
0.505
0.515
0.525
*  Different ozone concentrations levels
                                  207

-------
               430
rs>
o
o>

            Sc 390
            o
m
< 370
350


   0
                       i    i

                                                   X
                                           k*x.
                                                    \
                                                 '"A 2%  Kl
                                               >—o5% Kl
                             I
i    i   i
                  15        30       45

                       TIME (min)
                                                             60
                          Fig. 13. Stability of colour intensity as function


                                 of potassium iodide concentration

-------
and the remainder is liberated during the first half hour.   By using a
more concentrated potassium iodide solution (5%), almost all the iodine
is liberated during the first five minutes, then the colour begins to
fade gradually.

Since our investigations require multiple determinations to be performed
during the first fifteen minutes of ozone action, optimal results are
obtained by using the 2% potassium iodide solution and reading the
absorbance approximately 30 minutes after the triiodide ion is
liberated by ozone.

METHOD EVALUATION
Reproducibility
In order to establish the precision of the proposed method for
determination of ozone in water, six parallel determinations were
carried out, in a consecutive order, at different ozone concentrations.
The standard deviation of the results and the 95% confidence limits
are calculated and  shown in Table  30. The ozone concentration levels
examined are 0.05 to 0.33 ppm in the procedure used for low ozone
levels and 0.23 to  1.92 ppm in  that used for higher ozone levels.
The fluctuations of the results are minimal (less than 4%) when  the
procedure used corresponds to the  ozone level for which it  is
intended.  For instance, at 0.20 ppm level, the  relative standard
deviation is 0.9% with the suitable procedure "A" and 3.4%  if
procedure "B"  is used.
                                   209

-------
Table 30.  REPRODUCIBILITY OF RESULTS IN THE SPECTROPHOTOMETRIC METHOD
Procedure
A



B



Mean Value
O^ ppm
0.058
0.125
0.217
0.329
0.236
0.418
1.077
1.920
Range of Results
min. and max.
0.055-0.060
0.120-0.130
0.216-0.220
0.327-0.331
0.23-0.25
0.41-0.43
1.06-1.09
1.90-1.95
Standard
ppm
0.002
0.003
0.002
0.002
0.008
0.008
0.008
0.024
Deviation
%
3.4
2.4
0.9
0.6
3.4
1.7
0.8
1.2
95% Confidence
limits
0.0020
0.0027
0.0012
0.0015
0.0063
0.0057
0.0075
0.0180
a  Six replicates at each level
                                     210

-------
Sensitivity
The sensitivity of the spectrophotometirc method applicable to low
ozone levels is between the limits 0.01  to 0.30 ppm.   For the method
of higher ozone levels, it is between 0.05 to 2.00 ppm.

Correlation with Other Methods for Ozone Determination
The results obtained from spectrophotometric ozone determination usinc,
the neutral potassium iodide reagent were compared with those obtained
using the Mast reagent and also with the standard volumetric method.

The Mast reagent* recommended for the Mast Ozone Meter** was used by
            20
Jones, et al   in a toxicolog.ical experimental study.  In this method,
ozone reacts initially with the potassium bromide solution and the
liberated bromine then reacts with potassium iodide; the free iodine
is then measured spectrophotometrically.

Repeated determinations were carried out comparinq the neutral
potassium iodide reagent with the Mast reagent, at different ozone
levels.   No significant differences were observed.
                                                  15
The ozone determination  in the standard volumetric   as well as in the
spectrophotometric method   is based on the measurement of the triiodide
 *   Mast Reagent:   10 gr potassium  iodide;  25 gr  potassium bromide;
    1.25 g  sodium  phosphate monobasic and 5.0 g sodium phosphate dibasic;
    made up to 500  ml with distilled water.
 **  Mast Development Company,  Davenport,  Iowa.
                                   211

-------
ion liberated by ozone from potassium iodide solution.  In the
spectrophotometric method the amount of triiodide ion is measured
directly.  The precision and sensitivity of this instrumental analysis
permits it to be chosen as an independent variable.   On the other hand,
in the volumetric method, after the liberation of triiodide ion by
ozone, two additional variables are introduced.  These are (1)
acidification, in order to obtain the well defined stoichiometry of
the triiodide-thiosulfate reaction, and (2) end point detection, by
formation of starch iodine complex.  To compare the results obtained
by these two methods parallel samples of different ozone concentrations
were analysed carefully by both procedures.  The results are presented
in Table 31.

The comparative results obtained by both methods show the volumetric
readings to be greater than those of spectrophotometry.  A reasonable
explanation of this difference can be derived with reference to the
stoichiometric reaction:
               03 + 2KI + H20<=^  12 + 02 + 2KOH
               I2 * I" -> I3-
The liberated oxygen 02, which appears here as a by product, does not
modify the absorbance readings in the spectrophotometric method, at
neutral pH.  Since in the volumetric method acidification is used,
the 02 supplied by the mechanism of the reaction may liberate
supplementary I3~ ions:
                                  212

-------
       Table 31.  COMPARISON OF TITR1METR1C AND SPECTROPHOTOMETRIC

                 METHODS FOR DETERMINATION OF OZONE IN WATER
Ozone Level
  Range
   ppm
                                           Mean of results,  ppm
Number of
 samples
Standard
 method
Spectrophotometric
     method
  0-0,49

  0,5-0.99

  1.0-1.99

  2,0-4.99


    Total
    5

    8

   12

   15


   40
  0,33

  0.80

  K44

  3.97
      0.28

      0-74

      1,37

      3,75
                                    213

-------
               61"  +  02  +  4H+ ^	>  2I3"  + 2H20

which are subsequently measured and serve  as a source of error.
It is known,  '    that even the dissolved atmospheric oxygen  may be
a potential source of error in the volumetric titration of iodine, but
fortunately this reaction has a low velocity.  However.it may  be
catalyzed by light, metal ions or other factors.   In the case  of ozone
determination, the newly formed oxygen may be in  an activated  form or
some other intermediate species may also be produced, which in acid
solution could liberate additional iodine.

If the reaction between ozone and iodide takes place via some  intermediate
                                                           23
forms, i.e. ozonide ion 0 3> as some authors have reported,   the
stoichiometry of the reaction in neutral medium remains one to one, in
agreement with the classical equation:
               °3  +  r  ~ * °3~  +  ]/2 12

               °3~  +  l~  +  H2°  - * 20H +  °2  +
However, if the same reaction takes place in acidic-medium,  the ozonide
ion may be transformed to H03~, which may be able to oxidize additional
iodide to iodine, i.e. ,
                           H*  +  03~  - - >  H03
               H03  +  I"  - > H03"  +  1/2 I2
               H03" +  I"  + 2H+ - >H02 + H20 + 1/2 I2
                                   214

-------
Thus, in the above presented two mechanisms, the classical  equation and
the ozonide pathway, additional  iodine is liberated in the  standard
volumetric determination   This  fact correlates with our observations.

The second possible source of error in the titrimetric standard method
is the end point detection-  The starch end point is not sufficiently
sensitive for titration with very dilute thiosulfate solution.  Further-
more, the sensitivity of this end point detection depends also on the
amount of iodide ion present in solution.

APPLICATIONS FOR INACTIVATION STUDIES
The  inactivation process is an interaction between qermicidal agent and
microorganism, analogous to a chemical reaction, that is assumed to
follow the course of a  first order reaction.   The classic Chick's  law
of germicidal action is valid when only  the microorganism is  the
critical  reactant.  The practical approach  to  this condition  would be
to ensure that the  concentration of the  disinfectant and the  other
mi leu factors remain constant during  the period of exposuref

Different experiments were carried out in order to optimize  these
working   parameters for inactivation  experiments using  ozone as  a
germicidal  agent.

One  of  the  difficulties in the  work with ozone is  its  low  stability
                                  24           25
in water solution,  Alder  and Hill    and Stumm  showed that ozone in
                                  215

-------
solution is more stable at temperatures near the freezing point of water.
To better understand the conditions required to obtain stable ozone
concentrations the following variables were examined:   the influence
of  temperature and the influence of stirring rate as  a function of
time and of ozone concentration.

The Influence of Temperature
Fig.14 presents the influence of temperature on stability of ozone
concentration versus time.  The two levels were those  generally used
in the study of ozone's disinfecting activity.   It appears that at
5°C the ozone solution is evidently more stable than at 22°C.  The
concentration of ozone solution at 5°C remains  almost  constant
during the first five minutes and decreases slightly during the
following 15-20 minutes.  At room temperature (22°C) ozone decay
begins immediately and becomes significant after the first five minutes.

The Influence of Stirring Rate
Fig. 15 presents the influence of stirring rate on ozone stability in
kinetic experiments, at three ozone levels: 0.1, 0.6,  and 1.0 ppm.
Stirring is at 60, 80 and 100 rpm, while maintaining a constant tempera-
ture of 5°C in a buffered solution of pH 6.8.

The kinetics of ozone decomposition under these conditions was found to be
correlated with the rate of stirring and the concentration of ozone.
The percentage of loss in ozone rises with the Increase 1n ozone
                                   216

-------
                 1.0
IN5
             E 0.8r
             Q.     A
             Q.
                0.6
            LU
O
N 0.4
O
                0.2
        _   X
                               —A
           *•  °""
                                    	o	o	o---
                                        A	A	2	
                                                               5°C
                          10     20     30    40
                                  TIME  (min)
                                          50    60
                            Fig. 14.  Ozone aqueous solution stability versus

                                    time, temperature and concentration

-------
                  o-
                  A-
                             6O rpm
                         — ° SOrpm
                         -A iQOrpm
  I.OJT
  0.5
  0.2 -
                                           ••
                                           •o
 £
 Q.

    0.2 h
LU
o  O.I
N
O
   0.10

   0.05

  0.02
              .-i	-•—
                 u — — o -
                                      —-o
0.01
    0
                   10    15    20   25
                    TIME  (min)

              fig. 15.  Influence of stirring rate
                     on Ozone stability (at 5°C)
30
                  218

-------
concentration.   Although  the stirring  at  60  and  80  rpm  slightly  reduces
the stability of the ozone solution,  the  effect  of  100  rpm  is  far more
accentuated.  During 15 minutes, at 60 and 80 rpm,  the loss of ozone
is loss of ozone is less than 15%.  At 100 rpm, the loss of ozone is
30-40% after only 2-3 minutes and after  15 minutes this has reached
75-80%.

It may be concluded that for an experiment of 15 minutes duration, 80
rpm is the optimal  stirring rate for  these conditions.

Ozone Demand of Virus Culture Suspensions
The stability of ozone concentration  may also be influenced by  the
presence of organic matter  in virus culture  suspensions.   A partially
purified poliovirus culture was added, in different dilutions,  to
ozone solutions at  5°C and  at a stirring rate of 80 rpm.   The residual
ozone was determined during  a period  of  15 minutes.  The results are
presented in  Fig. 16.  After  a ten minute contact of the partially
purified virus  with the  ozone solution,  the  loss of ozone  is  about
                                o                                -4
90% for virus dilution of 5,10   , 75% for virus dilution of 5.10
and 10% for  a dilution of 5.10"5.
 Highly purified stock cultures,  containi/ig about 10 pfu  per nil,
 showed very low, ozone demand even at high virus concentration, with
 the stock diluted as  little as  10"3.
                                   219

-------
                      UJ
ro
ro
o
^
o
ISI
0

^
ID
O
C/)
tu
cr

Q
0»^

80


60

40

20

0
ra-n™ Q-— a 	 a 	 a 	 Q___ 	 /»
i w~-*r- 	 a >-*
r "• c
|
ii

J4.4 	 A_4^^^
V 4^--A
VV ^^~~**^~**-A

| | 1 1 1 1 1 1 1 1 1 1 1 1 1
1 3 5 7 9 II 13 15
                                                           TIME   (min)
                                           Fig. 16.   Kinetics of ozone demand of  different virus
                                                     suspensions at 5 C  and  stiring rate  80 rpm.

                                           a - Stock  culture dilution 5 x 10~3;
                                                                           o
                                           b - Stock  culture dilution 5 x 10" ;
                                           c - Stock  culture dilution 5 x. 10"3;
                                           A - Purified stock culture dilution 5 x 10~3

-------
 As  a  result  of  these  studies  on the factors affecting the stability of
 the ozone solutions we were able  to establish optimal conditions for
 our inactivation  experiments.
 REFERENCES
1.   Byers, D.H.  and B. Saltzman.   Determination  of Ozone  in Air  by
    Neutral and  Alkaline  Iodide Procedures.  Advan.  Chem.  Ser.
    21:93-101, 1959.

2.   Hofmann,  P.  and P. Stern.   Photometrische  Bestimmung  Niedriger
    Ozonkonzentrationen in Wassern.  Anal.  Chim.  Acta 47:113-120, 1969.

3.   Layton, R.F. and R.N.  Kinman.  A New Analytical  Method for Ozone,
    Nationality  Speciality Conference  on Disinfection, Am. Soc.  Civil
    Eng., p.  285,  1970.

4.   Ehmert, A.  Chemical  Ozone Measurement.  Advan.  Chem.  Ser.
    21:128-132,  1959.

5.   Hersh, P. and  R. Deuringer.  Galvanic Monitoring  of Ozone  in
    Air.  Anal.  Chem.  35:897-899,  1963.

6.   Alway, C.D.  and G. Slomp.   Ozone Concentration Meters.  Advan.
    Chem. Ser.  21:108-118, 1959.

7.   Hanst, P.L., E. Stephens,  W.  Scott,  and R.  Doerr. Adsorption
    for the Infrared Determination of  Trace Amounts of Ozone.
    Anal. Chem.  33:1113-1115,  1961.
                                  221

-------
8.    Bersis, D.  and E.  Vassiliou.   A chemiluminescence  method  for
     determining ozone.   Analyst 91:499-505,  1966.

9.    Amos, D.  Specific Spectrofluorimetric Determination  of Atmospheric
     Ozone Using 2-diphenylacetyl  1-3 Indandione  1  Hydrazone.  Anal.
     Chem. 42:842-844,  1970.

10.  Sachdev, S.L., J.P.  Lodge, and P.W.  West.  Specific Method  for the
     Determination of Ozone in the Atmosphere.  Anal. Chim. Acta 58:
     141-147, 1972.

11.  Van Ermengen, E.   Sterilization of Water by  Ozone.  Ann.  Inst.
     Pasteur 9:673-680,  1895.

12.  Hann, V.A.   Disinfection  of Drinking Water with Ozone.  J.A.W.W.A.
     46:1316-1320, 1956.

13.  Boelter, E.D., C.L.  Putnam and E.  Lash.   lodometric Determination
     of Ozone of High  Concentration.   Anal. Chem. 22:1533-1535,  1950.

14.  Birdsall, C.M.,  A.C.  Jenkins, and  E. Spadinger.  lodometric
      etermination of  zone.   Anal.  Chem. 24:662-664, 1952.
15.  Standard Methods  for the  Examination of  Water  and  Waste Water,
     p.  271. 13th edn.   APHA,  AWWA,  WPCT, New York, 1971.

16.  Saltzman, B.E. and  N.  Gilbert.   lodometric Microdetermination of
     Organic Oxidants  and Ozone.   Anal. Chem  31:1914-1920, 1959.
                                 222

-------
17.   Kopczynski,  S,L.  and J,J<  Bufalini.   Some  Observations  on
     Stoichiometry of  lodometric Analyses  of Ozone at pH 7.0.   Anal.
     Chem.  43:1126-1127,  1971

18.   Hodgeson, J  A ,  R E.  Baumgardner,  B.E.  Martin and R,A.  Retime,
     Stoichiometry in  the Meutral lodometric Procedure for Ozone by
     Gas-Phase Turation  with Nitric Oxide   Anal, Chem, 43:1123-1126,
     1971.

19.   Scott, D-W-   Measurements of fotal Oxidants in the Atmosphere.
     Report, Warren Spring Laboratory,  England, 1970.

20.   Jones, R.A., L.J. Jenkins, R.A. Coon, and J- Siegel.  Effects of
     Long Term Continuous Inhalation of Ozone on  Experimental  nimals
     Toxicol. Appl. Pharmacol. 17:189-202,  1970.

21,   Skoog, D.A.  and D.M, West, Fundamentals of Analytical Chemistry,
     New York.  Holt, Rinehart and Winston.  1963,   p.  458,

22.   Fisher, B.R. and D.G. Peters.  Basic Theory  and  Practice of
     Quantitative Chemical Analysis.   London.  W.B.  Saunders Company,
     3rd edn, 1968. p, 568.

23.   Boyd A.W., C. Willis and R. Cyr.  New  Determination of Stoichiometry
     of the lodometric Method for Ozone Analysis  at  pH  7.0. Anal.  Chem.
     42:670-672.
                                   223

-------
24.  Alder, M.G. and G.R. Hill.  The Kinetics and Mechanism of Hydroxide
     Ion Catalyzed Ozone Decomposition in Aqueous Solution*  J. Amer.
     Chem.  Soc. 72:1884-1886. 1950.

25.  Sturnm, W.  Der Zerfall  von Ozon in Wassriger Losung.  Helv.  Chim.
     Acta.  37:773-778, 1954
                                 224

-------
                             SECTION XII
    INACTIVATION KINETICS OF VIRUSES AND BACTERIA IN  WATER  BY  OZONE
INTRODUCTION
As long ago as the end of the last century ozone (0-J was  suggested
as a disinfectant for drinking water   Today,  this chemical is being
used to disinfect water supplies in many European countries,   A
review of the literature  suggests that in many ways  ozone is
superior to other chemical disinfectants: it acts quicker  and  in
lower concentrations under certain conditions, and has few known
side effects such as taste, odor and toxic by-products which are
                                                       p
characteristic for some other chemicals e,g ,  chlorine-    These
positive qualities have led to a renewed interest in  ozone as  a
disinfectant of water for the inactivation of  viruses which,
under certain circumstances, appear resistant  to chlorine.

It is striking that relatively little basic research  has been  done
on the inactivation kinetics of microorganisms by 0,  at constant
concentrations.  In particular, the research with viruses  has  been
neglected.   Such research is important in order to understand  the
time-concentration relations associated with the killing power of
the disinfectant, and would make it possible to calculate  the
quantities  of ozone needed for different kinds of water.

In this study efforts were made to develop methods by which the
disinfection abilities of ozone at constant levels of concentration
                                 225

-------
can be tested.  Various microorganisms found in polluted water were chosen,
These included E. coli  coliphage T^. and poliovirus type I.   The main
thrust of the research was with viruses, since these are known to be more
resistant to disinfectants than bacteria.

MATERIALS AND METHODS
Microorganisms
Poliovirus stock -
The same batch of poliovirus type I (Brunhilda) was used throughout
this study.  The virus was grown in Vero cells (Flow Laboratories),
accumulated, and concentrated by the phase separation method,  and by
two subsequent purification cycles of the concentrated virus.  A
cycle consisted of centrifugation of the suspension for 30 min at
12,000 x g.  The supernatant was then centrifuged for one hour at
100,000 x g.  The pellet was resuspended in 27 ml 0.005 M phosphate
buffer, pH 7.0, and ultrasonicated (20 Me/sec) for 2 min.  Finally,
the purified virus suspension was centrifuged at 4,000 x g for 30 min
and stored in glass vials (0.25 ml) at -70°C.  Titration of the virus
                                                           Q
was done on Vero cells, which indicated a count of 5.0 x 10  pfu/ml .

Coliphage T-
Coliphage T« was grown on E. coli.  The bacteria were inoculated in an
Ehrlenmeyer flask containing nutrient broth, incubated in a shaker bath
                                    226

-------
at 37°C for 5 hours.  Subsequently, 10  pfu bacteriophage were added
and the suspension was incubated for an additional  3 hours.   At the end
of the entire incubation period (8 hrs) the suspension became almost
clear.  This suspension was centrifuged for 45 min  at 12,000 x g;  the
supernatant was centrifuged for an additional  3-4 hours at 40,000  x g.
The sediment  was resuspended in phosphate buffer,  pH 7.5, and again
centrifuged.  The cycle was repeated 4 times.   After the last washing,
                                                                   o
the sediment was resuspended in phosphate buffer, diluted to 5 x 10
pfu/ml and kept under refrigeration.

E. coli (ATCC 11229)-
Before each experiment the bacteria were inoculated into 30 ml
nutrient broth and incubated overnight at room temperature, followed
by 4-5 hrs in a shaker bath at 37°C.  The bacteria  were harvested
and washed 4 times by centrifugation in 50 ml  0.5 M phosphate buffer
(pH 7.0).   The final sediment was resuspended in 50 ml of the same
buffer solution.  The suspension was then filtered  through a Milli-
pore filter (3.0y) to remove possible clumps.   The  resulting solution
                        q
usually contained 2 x 10  bacteria/ml.

Ozone
The ozone was generated from the oxygen in the air by means of a
Fisher-Lab Ozonizer (OZ III).  A suction pump (Charles Austen Pumps,
Ltd.) was used to pass the air through the ozonizer.  The air stream
was dried by means of a Koty "Senior" air purifier and flow equalizer.
The ozone-rich air was bubbled through a 0.05 M phosphate buffer,
pH 7.0, until a high concentration of 03 suspension was reached.  The
                                 227

-------
(L excess remaining in the air stream was neutralized by passing the air
stream through two traps, the first of which contained a concentrated
solution of KI and the second, thiosulfate.  Before each experiment,
the concentrated solution was diluted until the appropriate ozone
concentration was obtained.

Ozone concentration determination -
This was done by utilizing a spectrophotometric method developed in our
                                   4
laboratory described in Section XI.   Two similar procedures were
developed:  procedure "A" for low ozone concentrations (0.01-0.30 ppm)
and procedure "B" for high concentrations (0.30-2.0 ppm).

In procedure "A", 10 ml of the ozone containing samples were intro-
duced into a test tube containing 2 ml of 5% neutral potassium iodide
buffered reagent.  The test tube was left for about 30 min in a cool
and dark place.   Triiodide was quantitatively liberated and the in-
tensity of absorbance was then read with a spectrophotometer (Zeiss
P. MQ II, equipped with stoppered 20 and 40 mm quartz cuvettes,
suitable for use in the U.V. region) using cells of 40 mm light path,
against a distilled water blank.

In procedure "B" for higher ozone levels, 5 ml of the sample were
introduced in a test tube containing 5 ml of 2% neutral potassium
iodide.  After about 30 min the Intensity of the absorption was read
using cells of 20 mm light path.
                                  228

-------
Ozone-deniand-1 ree water -
Water used as, dilutant in the ozone experiments was made ozone-demand-
free by distillation in an all-glass still in the presence of alkaline
potassium-permanganate   This was followed by ozonizing the bidistilled
water and dissipation of the ozone by boiling.
Ozone-demand-tree g 1 assware_ -
In the course of this study it was found that c/one-demand-free glass-
ware is a prerequisite for maintaining a constant level of ozone
concentrations during the experiments.  All glassware was therefore
cleaned with dichromate-concentrate sulphuric acid mixture, after
wmch the glassware was rinsed in tap water and distilled water,
soaked in a strong ozone solution and finally dried at 180 C.

Inactivation of poliovirus
One ml of virus stock suspension diluted  1:10 in ozone-demand-free
water was put into a beaker containing 400 ml phsophate buffer (0.05 M,
pH 7.2) and dissolved ozone in the desired concentration.  Before and
during the experiment the beakers were kept at 5°C + 0.1°C or 1°C +
0.1°C in a thermostatically controlled water bath.  The contents of the
beakers were constantly  stirred  at 80 rpm by glass paddles connected
to a mechanical stirrer  (Phipps  and Bird).  The beakers were covered
to prevent the escape of ozone.  Samples  for virus determination were
withdrawn by pipettes through orifices in the covers.  The first
sample (5 ml) was  taken  8 seconds after the  introduction  of  the virus.
Further samples (5 ml each) were taken at 8-second  intervals for  the
first two minutes,  followed by sampling a 3,  5,  10  and 15 minutes.
The samples were introduced into a 5 ml neutralizer  (solution  of  0.002 M
                                  229

-------
tetra sodium pyrophosphate and 0.0004 M sodium sulphite)  immediately
upon withdrawal.  "Sampling time" was recorded as the very moment the
sample was blown from the pipette into the neutralizer.   Pi pettinq
never took more than 2 seconds at the most.  An additional beaker con-
taining 200 ml phosphate buffer and 0.5 ml of the diluted virus stock
(without ozone) was stirred simultaneously in the same water bath to
act as control.  Samples for ozone determination were drawn immediately
before  and 1, 2, 5, 10 and 15 min after virus addition.   Neutralized
samples were assayed virus in tissue cultures on the day  of the experi-
ment.  The preparation of dilution and inoculation  into tissue cultures
                              5
have been described elsewhere.
In order to decrease the sampling time to less than 8 seconds, a
simple fast-mixing apparatus was set up.  This apparatus  consisted of two 10 ml
glass syringes Cone containing ozone solution and the other the virus sus-
pension)  which could be injected simultaneously into a mixing device.
The two injection parts of the mixing device were of 4mm i.d. and
led into a 2 mm diameter mixing chamber, the resulting mixed
solution being forced out at right angles through a 2 mm tube.  The
entire mixing assembly was made of Teflon.  With the present simple
set-up, which was operated manually, it was possible to sample
after reaction times of about half a second.  The reaction being
quenched by flowing into the neutralizing sulfite solution.
Longer reaction times could be obtained by increasing the length
travelled by the mixed solution before being quenched.  While the
above apparatus and the manual operation are obviously not capable
                                      230

-------
of accurate reproducibility it represents  a simple,  quick  and  inexpensive
way to examine the profitability of virus inactivation studies  in  the
one-half to three second range.

Inactivation of coliphage T?
             ------ "---r        j^
The procedures followed here were as with poliovirus, with  the  exception
of the sampling frequency, which was done every 10 seconds.  For assays
of bacteriophages, samples were diluted in nutrient broth.   One ml of
each dilution was mixed with 1 ml 1.5% liquified agar (52°C) containing
E.col-  bacteria.  After stirring, the mixture was poured onto  petri
dishes containing agar   Plaques were counted after 18 hr incubation
at 37°C.

Inactivation of E. coli
The experiments were essentially the same as with coliphage Tg.
Bacterial assays were done by the pour plate method in nutrient agar.
Redox Potentials
These were measured with a potentiometer (Radiometer Type PHM 26C)
with calomel- and platinum electrodes.  The potentials are given as
the uncorrected values in millivolts, using the calomel electrode
as reference cell.
RESULTS
Inactivation Kinetics of Poliovirus I by Ozone
Concentrations of 0.07, 0.1, 0.15, 0.2, 0.3, 0.5, 0.8, 1.5  and 2.5
ppm ozone were used.  Figures  17,  18 and  19 depict the inactivation
                                    231

-------
              100
rv>
CO
r«o
        ID


        >

        u_
        o
        cr

        w    0.01
            0.001
                     1   1   1    1   1   1
                           1   1    1   1    1   1   1
0     16    32    48     64   80    96

                 TIME   (seconds)
                                                               112
                   Fig. 17.  Inactivation. kinetics of poliovirus I by 0.3 ppm Ozone at 5°C

-------
ro
CO
CO
              100
         g:
         >
<

>    O.I
             0.01
           0.001
                 0     16
                     32    48    64    80     96

                         TIME   (seconds)
                                                        I	i
                   Fig. 18.  Inactivation. kinetics of poliovirus I by 0.8 ppm Ozone at 5°C

-------
            100
ro
          0,001
                0     16    32    48    64    80    96    112
                                 TIME   ( seconds )
                  Fig. 19.  Inactivation kinetics of poliovirus  I by 1.5 ppm Ozone at 5°C

-------
                CONTACT TIME,  0.5 sec
                CONTACT TIME,  2.5 sec
 O.I   0.2   0.3  0.4   0.5   0.6
03  CONCENTRATION, (mg/L)
  Fig. 20. Virus survival rate with varying Ozone
         concentrations at 0.5 sec and 2.5 sec
                  235

-------
kinetics of the viruses with ozone concentrations  ranging from 0.3-1.5
ppm at 5°C.  Each graph sums up 3 experiments at least.   A significant
feature is the two-stage action of ozone.   The first stage is  short,
less than 8 seconds, with a virus kill  of  99-99.5%.   Stage 2 lasts  from
1 to 5 min, and in this period the remaining viruses are inactivated.
Increasing the ozone concentration from 0.2 ppm to 1.0 ppm had very
little effect on the inactivation rate, while at ozone concentration
of 1.5 ppm and above there was a somewhat  higher rate of inactivation
during the second stage.

What appears to be a true dose response is, however, seen in Fig. 20.
In this preliminary experiment, the effect of several concentrations
of 03 on poliovirus I, during very short periods of time (0.5 and 2.5
seconds) was measured.  The results clearly show a higher percent
kill with the increased 0^ concentrations.  The effect is already
noticeable after 0.5 second  of contact and is very strong after
2.5 seconds.

During the initial phase of this study, it was found that ozone
concentrations lower than 0.15 ppm did  not appear to cause virus
inactivation, as shown in Figure 21  which  depicts  the percent
survival of poliovirus type I after 40  seconds of exposure to
various  concentrations of ozone.  In later experiments, inacti-
vation was obtained with 0.1 and 0.05 ppm  ozone with the same
virus stock.  There was, however, a lack of consistency in these
results and in i^^\e of the experiments  no  inactivation was seen.
                                 236

-------
ro
CO
                    ID
                    a:
U.
o

_J
<

>
o:
Z)
                         IOO
                           10
                           I.Or
0.1 r
                        O.OIr
                       0.001
                             0 .1   .2  .3  .4  .5  .6  .7  .8  .9  1.0 I.I  1.2  1.3 1.4 1.5


                                                OZONE     (ppm)


                                     ria  21.  Poliovi'-us I survival dfter 40' ;.f oxcos'.r,-

                                             t.n various ( cn^entrn-iO'V, of Ozcru

-------
Effect of Ultrasonic Treatment
Berg et al  have suggested formation of virus clumps as a possible
explanation for the differences in sensitivity of the virus stocks
to chemical disinfectants.  Such virus clumps can be expected to be
more resistant to low concentrations of disinfectants.  Ultrasoni-
cation causes the break-down of clumps of microorganisms.  A
preparation of poliovirus was therefore ultrasonicated for 2 min
at 20 me/sec (M.S.E. 100 watt) and the ultrasonicated stock was then
tested in inactivation experiments with 0.1 ppm 03 (Fig. 22).  The
effect of the ultrasonic treatment may be said to have been dramatic.
While before the treatment the initial kill (after 8 seconds contact)
was less than 90% (and even after contact of 15 minutes 0.2-0.5% of
the virus could still be found), after ultrasonication the virus
became extremely sensitive to the same 03 concentration.  This found
its expression in an initial kill (after 8 seconds) of 99.5% and in
a complete disappearance of the virus after 3 minutes i.e., a
decrease of more than 99.999%.  The effect of ultrasonication alone
on the virus titer of the stock was negligible.
Effect of Storage Temperature
Ultrasonication experiments provided a possible explanation for the
fluctuation in virus sensitivity in 03, but 1t did not explain how
and when these shifts occur.  Certain circumstantial evidence, which
became apparent on careful reconstruction of the records concerning
the storage of the virus stock, helped throw some light on this
question.
                                238

-------
    100
     10
CO
o:
>
U.
O
     1.0
     0.
*   0.01
CO
<£
   0.001
0.0001
        ^ V
                             BEFORE  ULTRASONICATION
                 AFTER  ULTRASONICATION
                   J	L
                                              J	L
       01   2  3 4  5  6  7  8  9  10  II   12 13 14  15
                        TIME    Imin)
            Fiq. 22.  Inactivation kinetics of  poliovirus  I (118-b),
                   before and after ultrasonication, by 0.1 ppm Ozone
                             239

-------
For technical reasons the virus stock had to be transferred from its
original -70°C storage to -15°C for a period of several  weeks.   The
records indicated that only after this change did this particular
batch show an increased resistance to 0.1 ppm 03>  This  observation
led to the question whether storage temperature affects  poliovirus
sensitivity to 03.  In trying to find the answer, the following
experiment was carried out:  10 ml of virus stock solution were
ultrasonicated for 2 min and divided into ampoules of 0.25 ml  each
(a quantity sufficient for one inactivation experiment).  Half of
the ampoules were then stored at -70°C and half at -15°C.   Samples
were removed at different time internals from both freezers and
inactivation kinetics experiments with 0.1 Og were carried out in
the usual manner.

The virus stored at -70°C revealed a high sensitivity to 03 treatment,
identical to the reaction of the freshly prepared and ultrasonicated
virus batch (Fig. 22).  Its sensitivity did not undergo  any change even
after storage of 2 months.   In contradistinction, the virus batch
stored at -15°C showed a drastic change in sensitivity after only 5
days of storage.  This change expressed itself in the virus returning
to a resistant inactivation pattern similar to that seen before ultra-
sonication.  In further experiments virus stock from the -15°C freezer
was ultrasonicated for a second time and became sensitive again.
Transferring the sensitive ultrasonicated virus from -70°C to -15°C
                                 240

-------
resulted in a shift to the resistant form,  a  reaction  similar  to  that
shown by ijntreatec! \nruf

Inactivation Kinetics of Coliphaoe L,
Oj concentrations ranging from 0,01-1.3 opm were used.   The  inactivation
kinetics of coliphage T? with 0.01, 0.09 and  0,26 ppm  Og at  1°C are
depicted in Figure 23    Similar results were obtained at 5°C.   As with
poliovirus, both the two-stage inactivation curve and  the lack of a
clear dose-response above concentrations of 0.2 ppm 03 are seen.   Lower
concentrations caused a limited "dose response" occurring in the first
stage only, and was expressed in an initial kill of (A) 99-99.9% at
03 concentrations ranging between 0.01-0.06 ppm; (B) kill of 99.9-
99 99% at 0 07-0.19 ppm 03; (C) kill of 99.9-99.999% at 0.2  ppm 03.

Inactivation Kinetics of E.co'li
The 03 concentrations ranged from 0.02-1.55 ppm.  Figure 24 shows the
inactivation kinetics at 1°C.  Similar results were obtained at 5°C.
Again, the two-stage inactivation curve is seen.  Above 0.1  ppm 03 no
dose response was obtained   The lowest concentration causma inacti-
vation was 0.04 ppm 03-  Some sort of "dose response"  could be seen
when the concentrations ranged between 0.44-0.07 ppm 03<

Introduction of Virus at Two Different Times
A possible explanation for the two-stage inactivation curve could be
provided by a chemical change of ozone e.g., the appearance of some
intermediate decomposition species, triggered  by the microorganisms
                                  241

-------
     100
      10
CO

9i
>
LJ_
o
>
(T
D
CO
  0.000!
0.01 r
   0.001 =
                                   0  0.01 ppm 03
                                         0.09 ppm 03
                                           0.26 ppm 0-
        0    20  40   60   80    100   120

                 TIME  (seconds)


          Fig 23.  Inactivation kinetics of coliphage Tp by var

                 concentrations of Ozone at 1°C
                          242

-------
   100
                                           0.04 ppm  03
                                        a  n
                                        0.07 ppm 03
0.0001
       0    20   40   60   80   100  120

               TIME   (seconds)

        Fig.  24. Inactivation kinetics of E. coli by various
               concentrations of Ozone at 1°C
                          243

-------
or organic contaminants introduced with the virus stock.  A second
portion of poliovirus was therefore added 56 seconds after the first
administration.  The inactivation rate of this additional virus
inoculum was the same as of the first portion, which implies that the
inactivating power of the ozone, whose measured concentration did not
change, was not affected by the experimental conditions (Fig. 25).
Redox Potentials of Ozone Solutions
To verify a possible connection between the redox potentials and the
disinfection ability of ozone,the redox potentials of different 03
concentrations were measured.   Figure26 shows the results of many
such measurements.  Redox potentials increase rapidly with increasing
03 concentrations from 0.03-0.2 ppm, after which there is a levelling
off.

DISCUSSION AND CONCLUSIONS
During the search for disinfectants of bacterial contamination in
water, ozone was one of the first chemical agents found to possess
germidical properties.   However, its mode of action, particularly
against viruses, has been investigated less than that of other dis-
infection agents.  Since there is a dearth of data on its mode of
action, it was decided to study the behavior of ozone in clean water.
Ozone is extremely labile; even a very small amount of organic matter
in the water causes its rapid  dissipation.  Extreme care was required
to build an experimental system in which ozone loss could be kept at
a minimum. Toward this purpose, ozone-demand-free glassware, microorganism
                                   244

-------
0.001
             16
32    48     64    80
    TIME   (seconds)
96
112
            Fig. 25.  Inactivation kinetics of poliovirus I,
                    added in two portions, at 0 time and
                    after 56", by 1.5 ppm Ozone
                              245

-------
l\3
1000
> 800
E
~ 600
QL
o400
200
• • • « J| 	 —
/•
•
T
	 1 	 1 	 1 	 1 	 1 	 L 1 'I i i i 1 1 i I , i II |
                        0.5           l.O           1.5          2.0
                          OZONE  CONCENTRATION    Ippm)
                 Fig. 26. Correlation between Ozone concentration and redox potential values
2.8

-------
stock and water had to be prepared.   Further measures were covering  ihe
beakers, maintaining the temperature at 5°C or lower during the experi-
ments, and stirring not exceeding 80 rpm.   Stirring above 80 ypm leadj
                         A
to a rapid loss of ozone.   In spite of these meticulous preparations,
it was not always possible to keep the ozone at a constant level.  In
this study results are reported only for those experiments where ozone
dissipation did not exceed 20% during the 15-minute period of the
experiment.
The main object of this work was to elucidate the inactivation kinetics
of microorganisms in water.  Since pollovlrus type I 1s ubiquitous 1n
sewage it was chosen as a prototype for viral contamination in water.
Another virus indicator used 1n this work was the bacterial virus
coliphage Tg.  Finally, E.  coli  a bacterial  indicator of fecal contam-
ination of water was also Included.

An outstanding characteristic of ozone action is its rapid Inactivation
of the microorganisms  tested and the low concentrations of 03 needed.
Comparing the results  obtained  in this study with those of Scarplno et
al,  who used chlorine, and with those of Berg et al,  who used Iodine,
the  superiority of ozone 1s striking.  With  0.3 ppm  of disinfectant,
for  example, a 99% kill of pollovlrus was reached in less  than 10 seconds
when using ozone as  compared to approximately 100 seconds  needed with
chlorine (at pH 10)  and about 100 minutes with Iodine.   In fact, ozone
acts so quickly that it 1s practically Impossible to measure  the  time
                                    247

-------
required for 99% kill.  It was not possible to obtain curves showing the
                                                                      o
inactivation rate as a first order reaction according to Chick's Laws.
A noteworthy phenomenon was the two-stage shape of the inactivation
curve.  Stage 1 lasted less than 10 seconds with an inactivation of
more than 99%.  Stage 2 continued for several minutes during which
period final inactivation occurred.  This feature cannot be ascribed
to changes in the composition of ozone, since poliovirus added 56
seconds after the first inoculum was inactivated at the same rate.
A possible explanation for the two-stage inactivation kinetics might
be that 0.5-1.0% of the poliovirus consists of clumps.  Galasso and
Sharp  were the first to mention such clumps, and Berg et al
described these virus clumps as impeding the inactivation rate.  It
is interesting to note that this two-stage action also appeared in
experiments with coliphage T2 and E.coli.   It may be assumed that
here, too, clumps were the cause of this phenomenon.  An attempt was
made to obliterate the clumps in the Tp phage and E.coli cultures and,
indeed, an accelerated rate of kill with these two microorganisms
was seen during the first stage.  03 concentrations as low as 0.01
ppm (the lowest concentration that could be measured) resulted in a
rapid inactivation of the coliphage, while E.coli was inactivated by
0.04 ppm 03.

Increasing the 03 concentration to above 0.2 ppm had very little
effect on the inactivation rate (as far as could be discerned by us).
                                 248

-------
Since, for technical  reasons, the earliest sample could be taken only
8-10 seconds after the introduction of the microorganisms, it was
impossible in these experiments to determine the processes that take
place during 0-10 seconds-   Dose response during this period may
occur.  To determine this,  s;x>cial apparatus for obtaining contact times
of one to ten seconds would be required   A preliminary experiment tends
to confirm this hypothesis, as is seen in Fiq  20.  These results,
however, were obtained with a crude method   Better equioment is now
under development for future study

With concentrations lower than 0,2 ppm 03, the  inactivation curves of
poliovirus were not sufficiently  reproducible to allow an accurate inter-
pretation of the results,  In some experiments  these ozone concentrations
had no effect, while  in others, concentrations  even  lower than 0.05 ppm
caused rapid inactivation.

Berg et al  maintain  that clumping affects virus  resistance.  Ultra-
sonication  (said to disintegrate  the  clumps) of  a relatively resistant
batch of  poliovirus I indeed  resulted in  an  increased  sensitivity to
0.1 ppm Oo.  It  is interesting  that,  while the  sensitivity of the
virus to  ozone increased, the  titer did  not  change although  it would
be  reasonable  to expect  an  increase  in  the  titer after disintegration
of  the clumps.  This  may be  explained by  a  simple  calculation    Let
us  assume that 5% of  the virus  was  in the form  of  clumps, and  that  an
average clump  consists of  10 virions.   Even  the complete disintegration
                                   249

-------
of all  clumps would thus result in an increase of about 50% in virus
number, an increase not considered to be significant by the usual  methods
of virus quantisation.   A pronounced increase in virus sensitivity,
however, would be shown by the techniques used in the present study.

The phenomenon of sensitive virus becoming resistant after having  been
stored in a freezer still remained unexplained.   There was reason  to
suspect that the storage temperature may have been the causative
factor.  To test this assumption, a series of experiments were per-
formed in the course of which it became apparent that storage at
-15 C is inadequate for poliovirus I in inactivation experiments with
CL.  This inadequacy expressed itself in transformation of the virus
from sensitive to resistant after only 3 days at -15°C.  On the other
hand, storing for 2 months at -70°C did not cause the slightest change
in sensitivity of the virus.   It should be remarked that samples
transferred from -70°C  to -15°C--without thawing—also shifted to  the
resistant form, which might well explain how our poliovirus changed
its level of sensitivity.

The most acceptable explanation for the virus stock shifting from
sensitive to resistant  appears to be that a significant percentage of
the viruses form clumps.  However, other possible physical changes
resulting from storage  of virus at various subfreezing temperatures
cannot be ruled out at  this stage.
                                 250

-------
An association may exist between the mode of action  of 0.,  and its  redox
potential.'   Redox potentials increased with increasing concentrations
of ozone from 0.03 to 0.2 ppm, but about 0,2 ppm Oo  the potentials
levelled off.  The fact that only a minor increase in redo;: potentials
was seen above 0.2 ppm 0^ correlates well with the lack of dose
response in the inactivation curves.  However, redox potential values
do not explain our results of ozone action on the microorganisms with
concentrations lower than 0.2 ppm Og.

Lowering the temperature from 5°C to 1°C had nearly no effect on the
                                                        •
inactivation kinetics of the microorganisms, contrary to chlorine.
It should be noted that 03  is chemically very active, even at temper-
atures as  low as -70°C,

The work described here should  be regarded  as preliminary.  Special
techniques are being developed  which will enable  us to measure the in-
activation rate of microorganisms by ozone  during the first 10 seconds
(first stage).  For  research  on water disinfection, however,  these
findings are  important  since  they clearly show  it to be imperative to
store the  virus, used  in the  experiments, under uniform conditions at
-70°C if changes in  sensitivity during  the  course of  the  experiments
are to be  avoided.

The major  unanswered question is, however,  whether  viruses  in the
natural aqueous environment are of  the  sensitive  type or  whether  they
                                   251

-------
include a siqnifleant percentage of clumps, thus possibly following the
resistant pattern of inactivation under practical  field conditions of
water and wastewater disinfection.  Our plans now call  for a further
investigation of this critical point as well  as studies usinq ozone in
contaminated water and effluent.  But, whatever remains to be done,
the rapidity with which ozone acts at low concentrations, as was
demonstrated here, promises positive utilization of ozone in the
disinfection of water and wastewater.
REFERENCES

1.   Venosa, A.D.  Ozone as a Water and Wastewater Disinfectant.  In:
     Ozone in Water and Wastewater Treatment.  Evans, F.L. (ed.).
     Ann Arbor, Science Publications, 1972.  p. 83-100.

2.   Berg, G.  An Integrated Approach to the Problem of Viruses in
     Water.  In:  Proc. Natl. Specialty Conf. on Disinfection, Amer.
     Soc. Civil Engineers Publication, 1970.    p.  339-364.

3.   Shuval, H.I., B. Fattal, S. Cymbalista,  and N. Goldblum.  The
     Phase-Separation Method for the Concentration and  Detection of
     Viruses in Water.  Water Res., 3:225-240, 1973.

4.   Shechter, H.  Spectrophotometric Method  for Determination of
     Ozone in Aqeous Solutions.   Water Res.,  7:729-739, 1973.

5.   Shuval, H.I., A. Thompson,  B. Fattal, S. Cymbalista, and Y.
     Weiner.   Natural Virus Inactivation Processes in Seawater.  0.
     Sanitary Eng.  97:587-600,  1971.
                                  252

-------
 6.    Berg,  G.,  S.L.  Chang,  and  E.K.  Harris.  Devitalization of Micro-
      Organisms  by Iodine.   Virology  22:469-481,  1964.

 7.    Scarpino,  P.V., G.  Berg, S.L. Chancj,  D. Dahling, and M. Lucas.
      A Comparative Study of Inactivation of Viruses  in Water by
      Chlorine.   Water Res., 6:959-965,  1972.
 8.    Chick, H.   An Investigation  of  the Laws of  Disinfection.  J.
      Hyg.   8:92-158, 1908.

 9.    Galasso,  G.L. and D.G. Sharp.   Virus  Particle Aggregation and the
      Plaque Forming Unit.   J. Inwunol.  88:339-347,  1962.
10.    Chang. S.L.  Modern Concept  of  Disinfection.  In:   Proc. Natl.
      Specialty Conf. on Disinfection, Amer. Soc. Civil Engineers
      Publication, 1970.   p. 635-680.
                                   253

-------
                             SECTION XIII
               DISINFECTION OF VIRUSES IN  SEWAGE  BY  OZONE
INTRODUCTION
In a previous study describing ozone (03)  action on  microorganisms  in
ozone-demand-free water,   it was found that low concentrations  of  0-
rapidly inactivate these entities.   Compared to chlorine,  which  is
widely used for disinfection of water, 03  acts  ten times  as  fast.   The
above study was performed under optimal conditions,  using  water  and
glassware that were ozone-demand-free. Such  situations certainly do
not exist in the "true-world".  The different kinds  of water that
have to be disinfected (drinking water, surface water, renovated
water, or treated sewage effluents) at all times contain  a certain
amount of organic matter which reacts with the  03 thus causing a
decrease in 03 concentration thereby creating ozone  demand.   A
                   234
number of articles, '  '  ' indeed point out that the  03 concentrations
required to inactivate viruses in water possessing 03 demand, are
considerably higher than concentrations needed  under optimal conditions,

In the present study,  03 was tested for its  ability  to inactivate
water borne viruses in the "true world".  As a  model, poliovirus
type I was chosen, which was introduced into filtered sewage.

MATERIALS AND METHODS
Poliovirus Stock
The same batch ~f poliovirus Type I (Brunhilda) was  used throughout

                                  254

-------
the study.   Preparation and titration of the virus  stock  has  been
described elsewhere.

Ozone
Ozone was generated from the oxygen in the air,  as  described  pre-
viously,

Analysis for Ozone in the Presence of Sewage
The method used for determining 0., in effluents  was the same  iodometric-
spectroscopic technique as described for the clean  water system,   but
with the following modifications due to the changing color of the
effluent solution under investigation.  The original color of the
effluent was a pale yellow, which on ozonization slowly disappeared.
It was therefore not possible to use the original solution as a blank
for spectroscopic measurement.  Instead, together with each sample
removed for analysis, a duplicate sample was taken  and diluted with
an identical buffered solution without potassium iodide, and  this
sample was then used as the reagent blank.  With this modification it
was possible to measure the 03 concentration of  the solution  under
investigation.

Ozonization Vessel
The ozonization experiments were performed in a 500 ml water jacketed
vessel.  A Teflon stop-cock near the bottom of the vessel allowed for
the removal of samples for analysis.  The virus and 03 were introduced
into the flask via a glass stoppered opening at the top of the vessel.
                                  255

-------
Continuous ozonization of the eft'lumi was 
-------
withdrawn at 10 sec, one min and 5 min after introduction  of the
virus-effluent mixture.   The samples for virus  determination were added
to a reducina solution to stop the ozone activity,
                                                                  Q
(b) Continuous ozone bubbling, in which 1  ml virus  stock (1.0 x 10  pfu/ml)
was added to 400 ml of undiluted filtered sewage in the ozonization
vessel   Magnetic stirring was then started and continued throughout the
experiment.  Ozone-rich air at a rate of 0,5 L/min was bubbled through
the effluent-virus admixture.  Samples for virus titration and 0-, deter-
mination were withdrawn as  in method  (a), but at different time
intervals:  the first sample was taken prior to 03 bubbling, the
following samples  (after 0., bubbling  had been started) were withdrawn
at  15-second  intervals up to two minutes; the tenth sample was taken
 at  three minutes.  The samples for virus determination were added to a

                                                                       5
reducing solution.    In both methods a constant temperature of 5  C  was
 kept.   Virus  assay was done on the same  day, according to Shuval et al.'

 RESULTS
 Batch  Experiments
 The  interaction  of 0.4,  0,8,  1.3  and  1.8 mg/L 03  (initial concentration)
 and  the virus in the  diluted  and  filtered  sewage  was examined.  Each
 of these 03  concentrations was tested with 5% and 10% effluent  in
 buffer.  The most striking observation was the  complete  disappearance
 of 03  immediately upon  the addition of the virus-effluent admixture.
 At the same  time, a sharp decrease  (90-99,9%)  in  virus titer by most
 0  concentrations was seen  (Figs. 27, 28,  29,  30).  This virus
                                    257

-------
               lOOLo-
                                         IO%  EFFLUENT
r>o
tn
oo
ir

>

u_
o

-J
           QC
                 10
                 1.0
      O.I
                                          5%  EFFLUENT
               0.01
                                             1
                                1234


                                        TIME     ( min  )



                        Fig. 27. Inactivation kinetics  of poliovirus  I by 0.4 mg/L


                                initial  0., concentration in the presence of 5% and


                                10% filtered sewage

-------
                IOO
            cr
                 10
                 L0
ro
en
vo
            GC
                 O.I
        /0% EFFLUENT
                                            5% EFFLUENT
                                            2           3

                                           TIME  (min)
                          Fig. 28.  Inactivation kinetics of poliovirus I  by 0.8 mg/L


                                   initial  03 concentration in the presence of 5% and
initial v.,


10% filtered sewage

-------
                100
            cr

            >

            u.
            O
                        >—o
ro
o>
o
o:
z>
            #
                  .01
                                           /O  %   EFFLUENT
                                            5%  EFFLUENT
                                 1234


                                           TIME  (min  )



                          Fig.  29.  Inactivation kinetics of poliovirus  I by 1.3 mg/L


                                   initial 03 concentration in the presence of 5% and


                                   10% filtered sewage

-------
ro
           (/>
           I    10
           U_
           O
                1.0
tr
co
              0.01
                                          10%  EFFLUENT
                                           5% EFFLUENT
                                                        I
                               1234
                                         TIME  (mln  )

                       Fig. 30.  Inactivation kinetics  of poliovirus I by 1.8 mg/L
                               initial 03 concentration in the presence of 5% and
                               10% filtered sewage

-------
inactivation took place only during the first ten seconds of the experi-
ment, after which period no significant change in virus titer occurred.

The degree of inactivation correlates with the strength of the 03
concentration; at concentrations above 1.3 mg/LO^, however, this
correlation disappears, and increasing the concentration causes only
slight, if any, increase in the inactivation capacity of 03.  Virus
inactivation was also considerably less with the 10% than with the
5% effluent, employing the same 0., concentration.

Continuous Ozone Bubbling
The virus inactivation obtained with batch experiments was only
partially due to the rapid depletion of the residual ozone.  Since our
objective was to verify whether 0_ would be able to completely
inactivate viruses under conditions similar to those in treatment plants,
it was decided to introduce ozone in a continuous manner into the
filtered sewage.  This sewage was either undiluted or diluted 1:2.
Virus in ozone-demand-free buffer was used as control.  Typical results
are depicted in Fig. 31.  Ozone residual concentrations in the controls
began to rise immediately after bubblino had been started and reached
its maximum (3mg/L) in less than three minutes.  Simultaneously, the
virus titer became sharply reduced, as expressed by the inactivation
of 90% in 10 seconds and 99.999% in less than one minute, while the
0  residual concentration at that time was about 1 mg/L.
                                   262

-------
    I00r»*«	
I    'o
     1.0
QC
     o.i
    0.01
                        X
\
                                 2           3
                               TIME    (min)
          Fig. 31.   Kinetics of increase in 0,  residual concentrations  and
                   of  poliovirus I Inactivation during continuous 0,
                   bubbling through buffer and filtered sewage
                                          263

-------
Ozone residual in the effluent was first detected at 90 seconds after
bubbling had been started, after which its concentration increased
rapidly.  A concentration of 3 mg/L 03 was reached at the same point
in time as noticed with the buffer e.g., about three minutes.   Virus
inactivation also started after a lag period (30 seconds) of bubbling.
Inactivation of 99% of the virus took about one minute from the start
of the experiment, at a time when no 0., residual was detected  in the
filtered sewage.  At two minutes, inactivation of 99.999% of the
virus was obtained, while the 03 residual  concentration measured was
0.6 mg/L.

DISCUSSION
Early in the course of ozone experimentation, one encounters the
phenomenon of its extreme instability, making it difficult to  maintain
a constant 0., level.  In a previous study in ozone demand free
distilled water  special efforts were made to keep a constant  0-, level
during experiments designed for inactivation of microorganisms, which
allowed for reproducible results.  In this earlier study, it was
found that very low 03 concentrations effectively inactivated  various
microorganisms, among them polioviruses.  As to the latter, a  high
degree of inactivation was obtained with the use of 0.1  mg/LOg.
          2
Coin et al  found a threshold value of 0.7 ml/L residual 03 in surface
water samples, below which inactivation was not satisfactory.   Their
initial 03 concentration, however, was 5 mg/L.  Majumdar et al '
mention a threshold concentration of 1 mg/L 03 for distilled water.
                                 264

-------
The main difference between these three latter studies  and our previous
work is that these authors did not maintain a constant  03 level.   The
present study utilized our previous techniques, while employing water
possessing ozone demand, which does not allow for constant 03 residuals,
One might therefore expect our results to be more similar to those in
the works quoted above.

A further goal of the present study was to test the ability of 03 to
disinfect water in a "true world" situation in which it always shows
ozone  demand.

With the objective of reproducible  results in mind, the same batch
of  filtered  effluent was  used throughout the  present study.  This was
achieved by  filtering a large quantity of  sewage and storing  it at
-80°C.

A  very noticeable  occurrence with  the batch  experiments was  the nearly
 immediate  disappearance of the  03  residual,  with partial  viral inacti-
 vation during this extremely  short time  interval.   The degree of
 inactivation depended on  both  the  initial  03 concentration and the
 quantity of organic matter in  the water.   Figure 32 depicts  the
 degree of  inactivation of poliovirus by  different  initial 03 concen-
 trations in 5% and 10% filtered sewage.   The amount of organic matter
 in the effluent played an important role in the ability of 03 to
                                  265

-------
100
            0.5        I        1.5       2
      INITIAL  03  CONCENTRATION  (mg/l)
  Fig. 32.   Percent survival  of poliovirus I after reaction with
          different initial 03 concentrations in 5% and 10%
          filtered sewage  (Batch experiments)
                        266

-------
inactivate microorganisms.   A large quantity of organic matter (BOD of
50 ppm, COD of 78 ppm, total  Nitrogen of 3 8 ppm and Ammonia of  4 ppm)
in the liquid impeded the inactivation ability of 0^   This  is very
clear with low initial 03 concentrations;  increase in 03 concentration
led to a diminished effect of the organic matter   This dose response
of 03 action on viruses was not seen in our previous study with ozone-
demand-free water.     It is reasonable to assume that 03 largely reacts
with the organic matter, leaving only a small amount to attack the
virus.  By increasing the initial 03 concentration, therefore, more 03
will be available for virus inactivation

This interference by organic matter with virus inactivation  was also
observed in the continuous ozone bubbling experiments.  In organic-
matter free water, detectable 03 started to appear  immediately after
the experiment had begun, accompanied by a very rapid virus  inactivation.
In the presence of organic matter, on the other hand, both the
appearance of 0, residual and the virus  inactivation were delayed
Interestingly enough, in the continuous  bubbling experiments virus
inactwation occurred before the 03  residual could  be detected.  A
meaningful  inactivation of more  than 99.9% was achieved at a residual
of 0.6 mg 03/1 .  The  range of 03 residual concentrations (0.6-1.0 mg/Lj
which  led to a nearly complete  inactivation  is, in  fact, identical to
                                                                     234
those  described by others, also working  on systems  with ozone  demand. ' '

The  overall  results  attained with ozone  show that  it  has  a  high  potential
as  a  viral  disinfectant, even where  sewage  effluent high  in  organic
content  is  concerned.
                                   267

-------
REFERENCES
1,  Katzenelson, E , B-Kletter and H-1  Shuval    Inactivation  Kinetics
    of Viruses and Bacteria in Water by Use of  Ozone,  J.A,W.W.A.
    66:725-729, 1974.

2.  Coin, L , C  Hannoun and C-  Gomerrla   The  Inactivation  of the
    Poliomyelitis Virus Present in Water,  By Use  of Ozone.   La Presse
    Mediale, 37:72-74, 1964

3.  Ma.iumdar, S.B., W.H  Ceckler and 0 J  Sproul,   Inactivation of
    Poliovirus in Water by Ozonization,  J-W.P  C.F., 45:2433-2443,
    1973,

4.  Majumdar, S.B,, W.H.  Ceckler and 0-J  Sproul.   Inactivation of
    Poliovirus in Water by Ozonization  J W-P-C.F., 45:2048-2055,
    1974.

5.  Shuval, H.I., A.  Thompson, B.  Fattal., S- Cymbalista,  and  Y.
    Wiener.  Natural  Virus Inactivation Processes  in Sea Water.  J.
    San  Engin, Div.  97:587-600, 1971.

6 •  Schechter, H,  Spectrophotometnc Method for Deternination  of
    Ozone in Aqueous  Solution.  Water Res.  7:729-739,  1973.
                                 268

-------
                            SECTION  XIV
                  COLIPHAGE INACTIVATION  IN  SEAWATER

INTRODUCTION
Pathogenic bacteria and viruses can be carried by sewage into the sea.
It is therefore Important to study the inactivation  of non-marine
microorganisms which enter the marine environment.

loday we have considerable information on the physical, chemical  and
                  123
Biological factors  '*' involved in the  die-away of enteric bacteria
in marine environment.  The fate of enteric viruses  is much less  known.

                                       A
Previous studies done 1n our laboratory  have shown  that there is a
definite biological marine anti-viral activity (MAVA) which causes
1n seven days a tnree to six loci reduction of polio-virus type I
seeded Into seawater. while in heat-treated seawater, used as control,
only a one log reduction was detected.

The MAVA activity was removed from seawater by filtration through a
membrane filter with a 0.45 pore size.  This activity 1s also ether
sensitive.

The present report 1s a further study on  the factors involved in
virus Inactivation 1n seawater.  Collphage T2 was chosen as a model
                                  269

-------
because working wjth bacteriophages instead of enteroviruses has the
advantage of easy performance of the titration and the short time needed
to gain results.
METHODS AND MATERIALS
Assay of Bacteriophage
Samples were diluted in nutrient broth.  One ml of each dilution was
mixed with 1 ml of 1% liquified agar at 52°C to which 0.1  ml E. coli
culture in the logarithmic growth phase was added.  After stirring,
the mixture was poured on petri dishes containing agar.  Plaques were
counted after 18 hours incubation at 37°C.

Isolation and Counts of Bacteria from Seawater
Seawater samples were taken from the Palmachim Beach 30-40 meters from
shore, 80 cms. depth.  No known source of pollution exists near the
sampling point.  This location can be considered as characteristic for
an unpolluted area of the Eastern Mediterranean.  The samples were
inoculated on Zo Bell medium 2216 and incubated at 20°C.   The formula
of the medium was as follows: 0.5 gr Bacto peptone, 1.0 gr Bacto yeast
extract, 0.2 gr sodium thiosulfate, 1,000 ml seawater (pH  of the
mixture after autoclavinq 7.4).  Descrete colonies which  grew on the
media, were picked and subcultured on agar slopes of Zo Bell media.  All
pure cultures were stored at 4°C.  Counts of marine bacteria were done
by diluting the sample in sterile seawater.  0.1 ml of the dilution of
the original  sample were spread with a bent glass rod on  Zo Bell agar.
The plates were incubated at 20°C.
                                  270

-------
Assays for MAVA in Seawater
Ehrlenmeyer flasks of 250 ml, containing 50 ml  seawater, were inoculated
             Q
with 2.5 x 10  coliphage particles suspended in 1  ml  phosphate buffer
pH 7.5 (final bacteriophage concentration 5 x 10  particles/ml).   The
                                       o
inoculated seawater was incubated at 20 C.  Counts of phage and of
marine population were made immediately after inoculation (0 time) and
afterwards in intervals of 24-48 hours.
Sterilization of Seawater
Seawater was sterilized by filtration through a Sartorius membrane
filter of 0.45/pore size.

Washing of Marine Bacteria
The bacteria were washed by centrifugation.  A bacterial suspension in
sterile seawater was centrifugated 10,000 rpm for 10'.  The sediment
was resuspended in sterile seawater and centrifugated again.  This
cycle was repeated five times.

RESULTS
Assays for Anticoliphage Activity in Seawater
Studies of the anticoliphage activity of seawater were carried out under
laboratory conditions.  An inactivation curve of coliphage in fresh
seawater samples is shown graphically in Figure 33.  The inactivation
process involves two stages.  A lag phase lasting 4 to 8 days after
which a decrease in the titer occurs.
                                  271

-------
  100
   '0
u.  -
o
   O.I
                                         	STERILE

                                            * SEA WATER
                                             ,NORMAL

                                              SEA WATER
         —J—I—I—I—I—I—I—I	I	I   I  I   I  I  I

         48  96   144  192 240 288 336 384
   10
<
o:
UJ

fe  I0

m


fe  10
            BACTERIAL  COUNTS
   10
       -I—i—i—i—i—i—i	i	i	i  i   i  i   i  i  i
     0   48   96
                   144  192  240 288  336  384

                      HOURS
 Fig. 33. Inactivation of coliphage T9 in normal and autoclaved sea water
                          272

-------
          100
fV>
^j
co
o  i.o -
    n NB  1/10


    * STERILE  SEA WATER



           NB  I/ 100



-^0 NORMAL  SEA WATER

     NB I/1000
                  24  48  72  96
                                120  144  168   192  216  240  264

                                HOURS
                   Fig. 34.  Inactivation of coliphage T2 in sea water containing


                           different concentrations of nutrient broth

-------
In experiments in which marine bacteria were  suspended  in  sterile
seawater and in concentrations of 10  bacteria/ml  the decrease  in  phage
titer took place without any lag phase (Figures  35, 36, 37 and 38).
                                                                4    5
The initial  number of marine bacteria in seawater  ranged  from  10 -10
bacteria per/ml  (Fig. 34   bottom).   This number decreased  during  the
assay period in  1-2 logs.
The Effect of Organic Matter on the  Anticoliphage  Activity  in  Seawater
To study the effect of added organic matter on the self-purification
processes in seawater, assay for anticoliphage activity were performed
in seawater containing different amounts of nutrient broth.  Figure  34
shows that addition of organic matter to seawater  inhibits  its anti-
coliphage activity of normal seawater.  Figure 35  shows graphically  the
effect of nutrient broth on the anticoliphage activity of a mixture  of
serially washed marine microorganisms added to sterile seawater in the
concentration of 10  bacteria/ml.   Addition of 0.3% nutrient broth
powder inhibited the anticoliphage activity of the microorganisms.

Anticoliphage Activity of Isolated Marine Microorganisms
Five different pure cultures of marine bacteria in the stationary  growth
phase were washed and suspended in sterile seawater 1n the concentration
of 10  bacteria/ml.  The anticoliphage activity of these  isolates  is
depicted in Figure 36.   Three of these Isolates showed anticoliphage
activity.  The isolates having anticoliphage activity differed widely
in the morphological characteristics of their colonies.  All of them
                                   274

-------
              IOO
               10  -
en
           cr
           ID
U.
O

£
               O.I  -
              0.0
                                         9   STERILE
                                         -• SEA WATER

                                              /ml  BACTERIA
                                               0.3% NB
                                         IN  STERILE SEA WATER
                                                       \0f/m\  BACTERIA ADDED
                                                       TO  STERILE SEA WATER
       0   24   48  72  96   120 144
                       HOURS
                                                    168
              Fig. 35.  The effect of nutrient broth on the anticoliphage
                      activity of a mixture of marine bacteria

-------
were Gram negative.  One isolate was a vibrio and the two others  were
rods of different sizes.

The positive cultures isolated in this study showed reduced activity when
assayed again after several transfers on Zo Bell  medium and stored at
5°C for three months.  Figure 37 shows the inactivation activity  of the
same bacteria (isolate IV) suspended, in the first case, in sterile
seawater immediately after isolation and in the second case, after several
transfers on Zo Bell medium and storage at 4°C for 3 months.

In a separate experiment another marine bacterium was isolated.   This
organism showed identical activity to MAVA but unlike the previous
isolate, did not lose its activity when transferred on Zo Bell  medium.
The bacterium is apparently a Flavobacterium sp.   It grows well on Zo
Bell medium at 20°C; at 37°C however, it does not multiply.

The Effect of Organic Matter on the Anticoliphaqe Activity of Isolates
Cjf Marine Bacteria
To examine some factors in seawater that might reduce the rate  of coli-
phage inactivation by marine bacteria, 0.3% nutrient broth powder was
added to a suspension of marine bacteria (isolave IV) in concentration
of 7 x 10  bacteria/ml in sterile seawater.  This suspension was  assayed
for anticoliphage activity.  The same isolate, suspended in sterile
seawater without nutrient broth, was used as a control.  Figure 38 shows
that addition of nutrient broth inhibited the antiphage activity  of
the marine bacteria, although it stimulated growth of the bacterium.
                                   276

-------
ro
           100
            10
         _1
         I

         I  i.o
         u_
         O
            O.I
          O.OI
O
                                        	ft--..  A	• STERILE SEA WATER
                                             A—*--
                     *<~"\ ISOLATE  II
                       "•a
                      ISOLATE  X
                                   X ISOLATE I


                                   A ISOLATE IV
I	 I
                                             I     I
                    • ISOLATE III

                    I     I     I     I
              0   24   48   72  96
    120  144  168  192  216 240  264
     HOURS
              Fig.  36. Anticoliphage activity in different isolates of marine bacteria

-------
         100
ro
~-j
CO
                                                 I  STERILE
                                               A I  NORMAL
                    48    96     144     192     240   288   336     384
                                        HOURS
                                                                                 STERILE
                                                                          O M NORMAL
          Fig. 37.  Different rates of coliphage inactivation  by an isolate of marine bacteria

-------
  100
    10
*  1.0
O)
        -  *ox
LL
O
   O.I
  0.(
                                         0.3 NB  ADDED
                                   O NORMAL SEA WATER
       0  24 48  72 96  120 144 168 192
    10
  10
cr
LJ
H
O
<
m

UJ

E
u.
O
10 9
10 8
    10
    10 6
              BACTERIAL  COUNTS
  0.3 NB  ADDED
                                NORMAL SEA WATER
                          O
                       i   i
 O

_l	I
        0  24 48 72 96 120 140 168 192

                   HOURS
 Fig.  38. The effect of nutrient broth on the activation

         capaHtv of an isolated marine bacteria
                           279

-------

O)
   100
     10
U.
0   1.0
    O.I
     PHAGE
     INOCULATION
       0   48 96  144 192 240288336 384 432 482 528
    I05
t
                BACTERIAL  COUNTS
III
tArf
h-
U

CD
U.
O


z
in4
IU




I03


JO2
*• \

^k \
~ \ ^
X N
\ \ 3
A-^A ^V_A--A— °
A=^JJ 	 ^Ag6
Ohrs

• ' ' 1 1 1 1 1 1 1 1 1 1 ' 1 ' 1 ' ' I i i I
                                                   360 hrs
       0   48 96  144 192 240 288 336 384 432 482 528
                            HOURS
     Fig. 39.  The inactivation of coliphage in preincubated

             sea water and the indicated hours of preincubation

             ir ?0°C before inoculation of the phage
      i

                             280

-------
The Effect of the Physiological  State of the Marine  Bacteria  on  Their
Anticoliphage Act ivity
Bacterial counts in seawater to which coliphaqe were added  showed  that
the coliphage inactivation was associated with a decrease in  the marine
bacterial population (see Figure 33)  and it usually  occurred  after a lao
period.  In order to study the effect of this phenomena on  coliphage
inactivation in seawater, the following experiments  were performed.
Several Erlenmyer flasks containing seawater were incubated at 20  C
in  the dark.  Each flask was inoculated with coliphaqe at a different
time interval after incubation had begun.  The rates of colinhaae
inactivation in the pre-incubated seawater are summarized in  Figure 39
The inactivation curve of the coliphaqe, inoculated in seawater pre-
incubated for 360 hours, lacks the lag period which characterized  the
inactivation in fresh seawater.  In  that suspension the bacterial
population  was  already in the decline phase  at the  time the phase
was introduced.

DISCUSSION
In  this  report  some  factors  involved in  the  self-purifving capacity of
vitro  seawater  were  studied.   The  coliphage inactivation model  was
chosen because  T  bacteriophages,  like enteroviruses,  are found  in  domestic
sewaqe and  are  discharged into seawater.

Mitchell5noted  that the  antiviral  activity of natural  seawater  was  stronger
 than  that of seawater polluted with  organic matter.  Lycke et al1
                                     281

-------
described a heat sensitive antiviral factor in seawater.  Proteins and
amino acids were found to inhibit this factor.  They suggested that the
inhibition was caused by reaction of the proteinious material with the
antiviral factor.

Our observations correlate quite well with those mentioned above.
Addition of nutrient broth inhibited the antiviral activity of seawater.
The inhibitory effect of nutrient broth could be caused by its action
on either the marine microorganisms or on their products.

Phages can be used as a source of food for bacteria when they are
digested by proteolytic enzymes produced by the bacteria.  Such a
possibility was described by Oliver and Herrmann7 who showed that  pro-
teolytic enzymes of Pseudomonas aeruginosa can be involved in nutrition
or autolysis.  Nutrient broth may protect the bacterio-phages by
competitive inhibition.

A loss of the antiviral activity by marine isolates after several
transfers and storage was also observed by other authors.  Magnusson
et al  8isolated a marine microorganism with a virus inactivating
capacity.  This capacity was lost after several  transfers on artificial
media and storage at 25°C.   We suspect that nutrients,  which had been
stored in the bacteria during the passages on rich media, caused the loss
of phage inactivating activity.  This may also be the reason for the
reduced antiphage activity  in some of our experiments.
                                  282

-------
Studies on the kinetics of coliphage inactivation in pre-incubated
seawater showed that the leg phase was lost after a long pre-incubation
of natural seawater at 20°C.  During that time, the number of marine
bacteria decreased at least in one lag.  It is possible that during
this period bacterial exocellular polymers were produced and these
caused phage flocculation or absorption, thus decreasing the number
of phages in the suspension.  Such exocellular polymers were shown by
             q
Pavoni et al.

SUMMARY
Seawater has been shown to possess a self-purifying capacity which
enables it to  inactivate foreign microorganisms.   The present report
deals with biological and chemical factors involved in the inactivation
of coliphage T~ in seawater.  The aims of this study were:
1.  Study of kinetics of T2 phage inactivation in seawater.
2,  Isolation  of marine microorganisms involved in phage inactivation.
3.  Investigation of factors that inhibit the antiviral activity in
    seawater.

The main findings were:
1.  A typical  inactivation curve of T« phage in fresh seawater involved
    two stages:  a lag phase after which a decrease in the titer occurs.
2.  When marine bacteria were added to sterile seawater the decreases in
    phage titer occurred without a lag phase.
3.  Some marine microorganisms were isolated and their anticoliphage
    activities in sterile seawater was examined.
                                  283

-------
4.  Nutrient broth in low concentration was found to inhibit the anti-
    viral activity of seawater.
5.  Pre-incubated seawater exhibits the anticoliphaqe activity
    without any lag period.
6.  Two alternative explanations for the anticoliphage activity in
    seawater were suggested:  a) phage are digested by proteolytic
    enzymes of marine bacteria, and b) bacterial  excellular polymers
    cause biological flocculation.
REFERENCES

1.  Magnusson, S., C.E.  Hedstrom, and E.  Lycke.   The Virus  Inactivation
    Capacity of Seawater.  Acta Path. Mlcrobiol.  Scand.   66:551-559,  1966.
2.  Matossian, A.M. and  G.A. Garabedtan.   V1ruc1dal  Action  of Seawater.
    Amer. J. Epidemiol.  85: 1-8, 1967.

3.  Mitchell, R.   Factors Affecting the Decline of Nonmarlne Micro-
    organisms in  Seawater.  Water Res. 2:535-543, 1968.
4.  Shuval, H.I., A. Thompson, B. Fattal, S.  Cymballsta  and Y.  Wiener.
    Natural Virus Inactivation Processes  in Seawater.  J. Sanitary Eng.
    Div. 97:587-600, 1971.
5.  Mitchell, R.  and H.W. Jannasch.  Processes Controlling  Virus
    Inactivation  in Seawater.  Environ.  Sci. Technol. 3:771-773, 1969.

6.  Lycke, E., S. Magnusson and E. Lund.   Studies on the Nature of
    Virus Inactivating Capacity of Seawater.  Arch.  Ges. Virus.
    Forsch. 17:409-413,  1966.
                                  284

-------
7.  Oliver, O.D.  and J.E.  Herrmann.   Proteolytic and Microbial
    Inactivation  of Enteroviruses.   Water Res.  6:797-805,  1972.

8.  Magnusson, S., K.  Gunderson, A.  Brandberg,  and E.  Lycke.  Marine
    Bacteria and  their Possible Relation to the Virus  Inactivation
    Capacity of Seawater.   Acta Path. Microbiol. Scand.  71:274-280,  1967,

9.  Pavoni, J.L., M.W. Tenney and W. F.  Echelberger.  Bacterial
    Exocellular Polymers and Biological  Flocculation.   J.W.P.C.F.
    44:414-431, 1972.
                                 285

-------
                          LIST OF PUBLICATIONS
 l.   Shuval, H.I. and E. Katzenelson, "The Detection of Enteric Viruses
     in  the Water Environment" in:  Water Pollution Microbiology,  (ed.)
     R.  Mitchell, John Wiley, New York, pp. 347-361, 1972.

 2.   Shechter, H., "Spectrophotometric Method for Determination of Ozone
     in  Aqueous Solutions", Water Res. Pergamon Press, Vol.  7,  pp. 729-739,
     1973.

 3.   Katzenelson E., B. Kletter, Hanna Schecter and H.I.  Shuval,
     "Inactivation of Viruses and Bacteria by Ozone", in:  Chemistry  of
     Water Supply, Treatment and Distribution, (ed.) A.J. Rubin, Ann
     Arbor Science Publishers,  pp., 409-421, 1974.

 4.   Katzenelson, E., B. Kletter and H.I. Shuval, "Inactivation Kinetics
     of  Viruses and Bacteria in Water by Use of Ozone", Jour. American
     Water Works Association, Vol. 66, pp. 725-729, 1974.

 5.   Fattal, B., E, Katzenelson and H.I. Shuval, "Comparison of Methods
     for Isolation of Viruses in Water", in:  Virus Survival in Water and
     Wastewater Systems, (eds.) Malina and Sagik, Center for Research in
     Water Resources, The University of Texas at Austin, pp 19-30, 1974.

 6.   Belfort,  G., Y, Rotem and E. Katzenelson, "Virus Concentration Using
     Hollow  Fiber Membranes,"  Water Res., Vol. 9, pp. 79-85, 1974.

 7.   Katzenelson, E«, "A Rapid Method for Quantitative Assay of Poliovirus
     from Uater with the Aid of the Fluorescent Antibody Technique",
     Archives  of Virology. Vol. 50, pp. 197-206, 1976.

 8.   Belfort,  G., Y. Rotem and E. Katzenelson, "Virus Concentration
     Using Hollow Fiber Membranes,II", Water Res. Vol. 10, pp.  279-284,  1976,

 9.   Kletter,  B. and E. Katzenelson, "Coliphage Inactivation in Seawater",
     Acta Adriatica, 1976, accepted for publication (15 pages).

10.   Katzenelson E. and N. Biedermann, "Disinfection of Viruses in Sewage
     by  Ozone", Water Res., 1976, accepted for publication (10 pages).
                                     286

-------
                                  TECHNICAL REPORT DATA
                           /Please read laaructions on the reverse before completing)
  REPORT NO.
                             2.
                                                          3. RECIPIENT'S ACCESSION-NO.
  EPA-600/2-77-095	
4. TITLE AND SUBTITLE

  DETECTION AND INACTIVATION OF  ENTERIC VIRUSES IN

  WASTEWATER
               5. REPORT DATE
               6. PERFORMING ORGANIZATION CODE
                Mav  1977 issuing date
7 AUTHOR(S)
                                                          8. PERFORMING ORGANIZATION REPORT NO.
  Hi 11 el  I.  Shuval and Eliyahu  Katzenelson
9. PERFORMING ORGANIZATION NAME AND ADDRESS

  Environmental  Health Laboratory
  Hebrew University-Hadassah Medical  School
  Jerusalem,  Israel
                                                           10. PROGRAM ELEMENT NO.
                  1BD713
               11. CONTRACT/GRANT NO.
                S-800990
                (formerly 17060EAM)
12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental  Monitoring and Support Lab.
 Office of  Research and Development
 U.S. Environmental Protection Agency
 Cincinnati,  Ohio   45268
 - Cin., OH
13. TYPE OF REPORT AND PERIOD COVERED
 Final (Oct. 1969-Jan. 1975)
               14. SPONSORING AGENCY CODE
                      EPA/600/06
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report covers studies on the  development and evaluation of methods  for concentra-
ting and assaying low levels of  viruses  in large volumes of water  as  well  as studies
on the use of ozone in inactivating  viruses in water and wastewater.

Of the eight virus concentration methods evaluated, filtration with cellulose nitrate
membranes, aluminum hydroxide and  PE-60  proved most promising.  The feasibility of
using hollow fiber membranes was demonstrated.  A rapid method capable of  detecting
viruses in water in less than 24 hours using fluorescent antibodies was  developed. A
spectrophotometric method of detecting low concentrations of ozone in small  (10 ml)
samples of water was developed.  Kinetic studies show that ozone  inactivates entero-
viruses more rapidly than chlorine under comparable conditions. With  a 0.3 ppm residua
ozone inactivates 99% of seeded  poliovirus in clean water in less  than 10  seconds as
compared to 100 seconds for chlorine.   Although no detectable dose-response relation-
ship could be demonstrated for ozone contact times greater than 10 seconds, preliminary
studies indicate that such a realtionship may exist for shorter contact  times.  The
kinetic curve of virus inactivation  having a rapid first stage lasting seconds followe
by a slower stage lasting minutes  may be associated with virus clumping  phenomenon.
Ozone has been shown to kill viruses rapidly and effectively in wastewater effluent.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
i.                 uea^mriuna
 Viruses,* water, wastewater,  public health,
 filtration, ozone,  disinfection, water
 quality
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                                                                           COSATl Field/Group
  Concentration detection,
  inactivation monitoring,
  methods,  Israel
                06M
18. DISTRIBUTION STATEMENT

   Release  to  public
  19. SECURITY CLASS (This Report)
    Unclassified
                                                                         21. NO. OF PAGES
                301
  20. SECURITY CLASS (This page)
    Unclassified
                             22. PRICE
EPA Form 2220-1 (9-73)
287
                                                                     mm* OTOfc 1177-757-056/6402

-------