United States     i       - Office of
Environmental Protection   Drinking Water
Agency
EPA-570/9-78-OOC
Water

-------

-------
United States
Environmental Protection
Agency
Official Business
Penalty for Private Use
$300
Fourth-Class Mai!
Postage and Fees Paid
EPA
Permit No. G-35
Washington DC 20460

-------

-------
               EVALUATION OF
      THE MICROBIOLOGY ISTANDARDS
            FOR DRINKING WATER
                     Edited by
                Charles W. Hendricks
            Criteria and Standards Division
               Office of Drinking!Water
               Washington, D.CJ20460
       U.S. ENVIRONMENTAL PROTECTION AGENCY
                Washington, D.C. 20460         :
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22151

-------
                         DISCLAIMER
This manual has been reviewed by the Office of Water and Hazardous Materials,
U.S. Environmental Protection Agency, and approved for publication. The opinions
expressed in the papers are those of the individual authors and do not necessarily
represent the views of the U.S. Environmental Protectioh Agency. Mention of trade
names, commercial products or techniques does not constitute endorsement or
recommendation for use.

-------
                            PREFACE
   Although the quality of America's drinking water is often taken for granted, safe
drinking water is a result of careful utilization of the latest technology and proper
water treatment. In the late 1800's it was discovered that microorganisms responsible
for diseases such as typhoid and cholera could pass through the intestinal tract and
find their way to the water systems. By 1900,30 people per 100,000 population died of
typhoid in the United States. The death rate from  typhoid fever  was substantially
reduced with the addition of filtration to water supply facilities and many cities had
this equipment on line by 1907. Chlorination was being practiced in most U.S. cities
by 1914, and during the period 1946-1970 waterborne outbreaks of typhoid fever had
dropped to a total of 53.
   It is apparent that the incidence of typhoid fever and cholera has been drastically
reduced  in the United States, but there is also information  to  indicate that
waterborne disease warrants further investigation. While there  were 53 cases of
typhoid during the years of 1946-1977, there were 297 outbreaks of gastroenteritis of
undetermined etiology. It is also significant to note that 71 percent of these outbreaks
occurred from  private  water supplies but at least 83 percent of the  illnesses were
associated  with community water systems.
   The above statistics indicate a risk of infectious disease associated  with drinking
water, but they do not address whether the water supplied met the 1962  Public Health
Service Drinking Water Standards. It is significant to note that one of the reasons for
the passage of the Safe Drinking Water Act of 1974, was the  lack of uniform
application of  the  1962 Standards to  all public water systems. Another largely
unknown factor is the extent  to which the  present microbiology standards are
adequate to protect against waterborne disease. Certainly the epidemiological data
supports the adequacy  of the standards for protection against cholera and typhoid
but the  prevalence of the less severe  diseases such as hepatitis, giardiasis and
gastroenteritis suggests that  some adjustment of the regulations may be required.
This volume focuses on the present standards for coliform bacteria,  coliform bacteria
detection and control, compliance with the coliform standard, and alternative means
of  determining compliance, and  lays the groundwork for future regulations
development.
   Sincere appreciation is acknowledged for the effort of Dr. Riley D.  Housewright,
Mr. Edwin E. Geldreich and Dr. S.M. Morrison for aiding in the planning of the
symposium. Special acknowledgments  are also  given to Cindy Moneymaker and
Lynette Simmons for their  secretarial  support  and to Barbara  Menking for her
editorial  assistance in preparing this Proceedings.
           Charles W. Hendricks
           August  1978
                                                                         ill

-------

-------
                            FOREWORD

   The Safe Drinking Water Act of 1974 provides a unique opportunity for the
 federal government to constructively add to the traditional and varied roles of State
 and local governments  in providing safe drinking water for this nation's public.
 Included in the Safe Drinking Water Act are requirements for the promulgation of
 Primary Drinking Water Regulations which will apply to all public water supplies.
 These regulations are to contain:
     1. a list of contaminants that have adverse  effects on human health.
     2. maximum contaminant levels (MCL's).
     3. criteria and  procedures  to "assure a supply  of drinking water which
       dependably complies with such maximum contaminant levels.
   On December 24, 1975, the U.S. Environmental Protection Agency promulgated
 National Interim Primary Drinking Water Regulations. These Regulations replaced
 the Public Health Service Drinking Water Standards of 1962 and for the first time
 applied maximum contaminant levels for microorganisms and selected chemicals to
 all public water supplies serving 25 or more persons. It is estimated that there are at
 least 50,000 community and more than 200,000 non-community supplies covered by
 the Interim Primary Regulations.
   The papers  in this Proceedings focus  on current microbiological  issues and
problems associated with the coliform MCL  and the means of determining
compliance.  It is anticipated that these discussions,  in  addition to the National
Academy of Sciences review, will provide the: Environmental Protection Agency with
new insights into the technical problems associated with the MCL's for coliform
bacteria and help to establish a firm basis for their revision.

                                      Joseph A. Cotruvo
                                      Director
                                      Criteria and Standards Division
                                      Office of  Drinking Water (WH-550)
                                      U.S. Environmental Protection Agency

-------

-------
                  TABLE OF  CONTENTS

 PREFACE

 FOREWORD

 WELCOME   Dr. Joseph A. Cotruvo, Director, Criteria and Stand-
                     ards Division, U.S. Environmental Protection Agency,
                     Washington,  D.C.      \

                            SESSION!
 EVALUATION OF THE COLIFORM STANDARD
 Chairman: Dr. Riley D. Housewright, National Academy of Sciences, Wash-
 ington, D.C.                                '
 Evaluating the microbial quality of potable waters 	   3
 M.J. Allen and E.E. Geldreich
 Interferences to coliform detection	   13
 E.E. Geldreich, M.J.  Allen and R.H. Taylor                '    ",
 Impact of the coliform standard on the transmission of disease	   21
 G.F. Craun                                                 '•
 Alternative indicators  of water contamination  and  some physiological
 characteristics of heterotrophic bacteria in water	.,	  37
 G.A. McFeters, J.E. Shillinger and D.G. Stuart!
 Some statistical considerations in water quality control.,..'....	   49
 L.R. Meunz
 Chlorine residual substitution—rationale  		....   57
 L.J. McCabe
 Public health considerations of the microbiology of "potable" water	  65
 R.R. Colwell, B. Austin and L. Wan

                            SESSION!      '••'-       ;


 COLIFORM DETECTION AND CONTROL
 Chairman: Mr. Edwin E. Geldreich, U.S. Environmental Protection Agency,
 Cincinnati, Ohio                                      ,       .

 Standard sample collection and preservation  ..;	   79
 H.D. Nash                                !

The EPA methods manual and the microbiological standards  	   85
R.H. Bordner and J.A. Winter

Evaluation of standard procedures—Standard Methods	   91
A.E. Greenberg                        i

                                             	   95
Virus detection  	
D.O. Cliver

Disinfectant resistant organisms
                                                                     101
P.T.B. Shaffer

                                                                     vii

-------
The relationship of turbidity to disinfection of potable water ..... ........  103
J.C. Hoff
Wastewater reuse and the problems of waterborne disease
H.W. Wolf
Evaluation  of  microbiological standards for drinking water— engineering
control practices
P.O. Haney
                                                                   119

                           SESSION 3

COMPLIANCE WITH THE COLIFORM STANDARD
Chairman: Dr. Charles W. Hendricks, U.S. Environmental Protection Agency,
Washington, D.Q
Compliance with drinking water regulations	•	  145
C.W. Hendricks
Experiences with the coliform standard under the interstate program  	  153
F. Taylor
Experiences with the coliform standard	  159
J. Harrison
Implementation of the coliform standard by a community water utility  	  169
R.J. Becker
Compliance with PL 93-523 coliform standards	  181
H.J. Ongerth
Canadian drinking water standards  	•	
T.P. Subrahmanyan, D.A. Schiemann, A.J. Rhodes, J. Henderson, E. Bowmer,
fc.R. Rozee and P. Payment
                                                1 !

                   PANEL DISCUSSION
ALTERNATIVE MEANS OF DETERMINING COMPLIANCE
Chairman: Dr. Sumner M. Morrison, Department of Microbiology, Colorado
State University, Fort Collins, Colorado
Alternative means of determining compliance  ......... ................. 197
S.M. Morrison
The coliform MCL— Can we defend it?  ............................ • • • 199
W. Litsky
Alternative means of determining compliance  ......... ....... • .......... 205
B.E. Gay
Alternative means of determining compliance  ......... ............ ...... 211
I. Markwood
Alternative means of determining compliance  .......................... 215
W.O. Pipes
Alternative means of measuring compliance with bacteriologic standards of
public water supplies  ............................... ................. 219
G.W. Fuhs
 VUl

-------
LIST OF ATTENDEES                     229




INDEX                            i      233
                                       IX

-------

-------
                      .   .    ..WELCOME  .             ,

                       Dr. Joseph A. Cotruvo, Director
                        Criteria and Standards Division         '
                           Office of Drinking Water            :
                    U.S. Environmental Protection Agency      :
                           Washington, D.C. 20460             :.

   I want  to  extend  a sincere welcome to all of you  in attendance at today's
 symposium entitled "Evaluation  of the Microbiology  Standards for Drinking
 Water". We are fortunate to have many of the nation's leading experts in water
 supply who will address both the  technical and regulatory aspects:of the current
 regulations. I would invite all of you to participate in the discussions to the maximum
 extent possible and everyone here should iind something of interest.
   The symppsium is sponsored  by the Criteria and Standards Division, Office of
 Water Supply and is intended to gather information in support of future regulations
 development. The Safe Drinking Water Act (PL-93-523) requires EPA to revise the
 Interim  Primary Regulations within  90 days after the receipt of the National
 Academy of Science review.  The presentations and discussions here will add to our
 understanding of the problems and issues and, together with the National Academy
 of Sciences review, will enable us to make the necessary revisions to the regulations.
   Despite  the advanced  technology and management practices applied to water
 treatment, waterborne disease outbreaks still occur and a significant portion of the
 time we cannot determine what etiological agent is involved. Most of the outbreaks
 and cases of illness resulted from contamination of groundwater and deficiencies in
 treatment but many of the outbreaks (31 %) were due to inadequacies or interruptions
 in the chlorination  process.  This indicates  that we  must constantly improve the
 operation of existing treatment and monitoring processes as well as consider new
 analytical methods and techniques.
   The preamble of the Interim Primary Regulations addressed three specific areas
 where changes in the primary regulations may be desirable. The month'ly average by
 which compliance is determined, is affected by the upper limit of coliform bacteria
 when the sample contains coliform bacteria too numerous to count. For example,
 some laboratories assume an  upper limit of 50 while other laboratories seek to count
 all the colonies on the plate. There is also a problem in dealing with spurious positive
 samples. All samples taken to determine compliance must be counted because a
 negative check sample taken days after a positive sample does not mean that the
 original positive result was in error. The third aspect to be resolved concerns relating
 monthly averages of coliform bacteria to monthly percentages of positive samples.
 Each routine sample taken during the month may be in compliance with the limit and
 a system could still fail the monthly average maximum contaminant level. I would
 anticipate,  in your discussions over the next two days, each of these questions could
 be discussed in detail as to provide the Criteria and Standards Division with a means
 of dealing with them without compromising the integrity of the present maximum
 contaminant levels.
  To help facilitate answering these and other compelling questions, the symposium
is divided into three sessions. Session I-Evaluation of the Coliform  Standard, will
have  Dr. Riley D.  Housewright from the National Academy of  Sciences, Safe
Drinking Water Committee, to act as chairman. His session will focus on microbial
water quality,  alternative indicators, and the problems encountered with testing
procedures and treatment. Session 2-Coliform Detection  and Control, chaired by
Mr.  Edwin E. Geldreich, Municipal and Environmental Research Laboratory,
Cincinnati, Ohio; will deal with the problems of the growth and detection of coliform
bacteria and other indicator  organisms. In addition, viruses, turbidity and water
reuse will be discussed. Dr. Charles W. Hendricks from my office will chair Session 3-

                                                                        xi

-------
Compliance with Coliform Standards. This discussion includes the compliance
aspects of the National Interim Primary Regulations and specific changes that should
be made in the primary regulations.  A panel discussion to be conducted by Dr.
Sumner M. Morrisdn from Colorado State University will consider alternative
means of determining compliance with bacteriological standards.
  I thank all of you for attending and hope this symposium will be beneficial for all
concerned as we work toward providing safe drinking water.
Xll

-------
Fly-Away Virus Concentrator (US Environmental Protection Agency)
SESSION  1
                 Evaluation of the Coliform Standard

Chairman: Dr. Riley D. Housewright, Project Director, Safe Drinking Water
Committee, National Academy of Sciences, Washington, D.C.

-------
     ,J   '         II'
„    j;   .      {;

-------
                                   MICROBIAL QUALITY OF WATER/M. J. ALLEN
            Evaluating the Microblal Quality of Potable Waters

                  Martin J. Allen and Edwin E. Geldreich, Jr.
                  Water Supply Research Laboratory/ MERL
                    U.S. Environmental Protection Agency
                           Cincinnati, Ohio 45268

 Bacterial Indicator Concepts

   Coliform  bacteria have  been the bacteriological tool used  to  measure the
 occurrence and intensity of fecal contamination in drinking waters for nearly 70 years
 (11,13). During this period, a mass of microbiological data has been developed to
 permit the conclusion that the absence of total coliforms in a potable water supply is
 evidence  of a safe supply. Although there  are some differences between various
 coliform  strains  with  regard  to  natural  survival  and  perhaps resistance to
 disinfectants, these are minor biological variations that are more clearly demonstrat-
 ed in the laboratory than in a water treatment system. The presence of any coliform
 bacteria, fecal or nonfecal, in treated water should not be tolerated, and when found,
 suggests either  inadequate  treatment or postchlorination contamination in the
 distribution system (5,15). For these reasons it is not surprising that the total coliform
 indicator system is part of this nation's drinking water standards.
   The coliform group of bacteria was first adopted in 1914 by the U.S. Public Health
 Service pursuant to the Interstate Quarantine Regulations for common carriers (21).
 Since the inclusion of this bacterial quality parameter for  potable waters, the
 coliform concept has remained  essentially unchanged in successive Public Health
 Service standards of water quality, i.e. 1925, 1942, 1946. During this time there have
 been a few minor refinements/clarifications in the multiple tube procedure.  The
 development and acceptance of the membrane filter analytical method in the 1962
 Public Health Service Standards added an alternate procedure, but since then no
 further advances  have  been made  in  methodology.  Most  recently,  with the
 promulgation of the National Interim Primary Drinking Water Regulations in 1975,
 the total coliform  group of bacteria was again selected as the indicator of sanitary
 significance for evaluating the microbial quality  of drinking water (20). In view of
 advances in methodologies for detection of enteric viruses and bacterial pathogens,
 the infallibility  of the coliform standard has been occasionally 'questioned  (18)
 although  it continues to be the basic biological test of water potability.
   To enumerate all waterborne pathogenic organisms, the microbiologist would
 have to perform a variety of complex, time-consuming, costly, and often tentative
 procedures for each sample analyzed. A more realistic approach is to use»a bacterial
 indicator  system which will detect and quantify fecal pollution from all warm-
 blooded animals. Insofar as bacterial pathogens  are concerned, the total coliform
 group of  organisms, which  comprise only a fraction  of the number and types of
 bacteria present in feces (12), is considered a reliable indicator of the adequacy of
 water treatment. From a practical standpoint, the standard  total coliform methods
 (MPN, MF) offer the finished water laboratory a microbial test parameter which is
 easy to perform, requires little time, modest equipment expenditures, is relatively
 inexpensive, and provides results in 24 hours for the MF or within 96 hours when
 using the  MPN procedure. Both the MPN and MF protocols can be easily.learned
 and performed by most water treatment plant personnel.         :
  While the coliform standard and its concomitant methodology are the preferred
microbial surveillance tool, there have been instances where such monitoring has not
signaled the presence of waterborne Giardia cysts. Newly reported data on non-
filtered  surface  water systems note that: the removal  of  coliforms during  the
disinfection process may not be an effective measure of cyst inactivation (9). Data

-------
                                                 I  I
                                                   II

MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS*


      10
Tl	1

  COXSACKIEViRUS A2
                                               M. FORTUITUM
             COXSACKIE-
                VIRUS    \
                    A9
                                                C. PARAPSILOSIS
               ADENOVIRUS
                   TYPES
    0.01
                            10                100
                          CONTACT TIME (MIN.)
                                    1000
  Figure 1.  FREE CHLORINE RESIDUALS AND CONTACT TIMES NECESSARY FOR
          99.9% INACTIVATION FOR SELECTED VIRUSES, BACTERIA, AND YEAST (2, 10).
collected to correlate virus occurrences with total coliforms in finished water support
the continued use of the coliform group as the sanitary standard.

The  Relative Resistance of Bacteria and Viruses
  There is a considerable amount
 of information available on the relative resistance
of bacteria and viruses to chemical disinfectants. Although a number of viruses may
occur in polluted surface waters. I.e. infectious hepatitis virus, poliovirus, reovirus,
adenovirus, echovirus, coxsackievirus and  bacteriophages, disinfection data are
limited principally to the enterip viral groups of polio and coxsackievirus (16).
Because chlorine is widely used as'the disinfectant of choice in most water treatment
plants, determination of the relatij/e resistance of these enteric viruses to free chlorine
and chloramines is of obvious practical importance. Ideally, it is desirable that the
general resistance of all enteric viniises to chlorination would be of the same order and
be similar to inactivation rates ofjthe indicator coliform bacteria. The literature on
studies using a variety of microorganisms and species of chlorine/chloramines show
that there are measurable differences in susceptibility to disinfectants (Figure 1).
While some enteric viruses, bactefia(Mycobacteriumfortuitum, M. phlei), and yeast
(Candida parapsilosis)  are  mo{re  resistant that  coliforms  to  inactivation by
conventional disinfectants (10), current recommended water treatment practices use

-------
                                   MICROBIAL QUALITY OF WATER/M. J. ALLEN
                                              i
                                              i
 chlorine concentrations greater than 1  mg/1 w;ith protracted contact periods to
 increase the likelihood of total inactivation of viral and bacterial pathogens.


 The Significance of Virus-rColiform Ratios ;

   The primary objective in the bacteriological examination of water is the detection
 of treatment deficiencies which can result in the eintry of pathogens into the finished
 water supply. However, the occurrence and density of viral and bacterial pathogens
 in polluted waters is highly variable. This variability reflects the intestinal diseases
 that are prevalent at a given time in the human or <|>ther animal subpopulations which
 contribute their wastes to a particular effluent or watershed, as well as the numerous
 physical and chemical conditions which aifect microbial virulence and survival.
   The relative number of coliforms and viruses in'feces, sewage, and polluted surface
 waters have been considered by some to be important in evaluating the waterborne
 disease  potential and validity of the coliform standard (6). While  coliforms are
 discharged  in the feces of nearly everyone:, viral excretion is  generally limited to
 children under the age of 15 years with approximately 10% of these children shedding
 viruses at any given time. About 30% of the population falls into this age group, and
 calculations based on a per capita daily discharge of virus and coliforms result in
 estimated  virus-coliform ratios of 1:65,000  foij feces, 1:92,000 for sewage, and
 1:50,000 in polluted surface water (1); studies: made  on Missouri River waters
 indicate a ratio of 1:500,000 (7). First year resu
ts of an in-house study (Brashear,
 unpublished) of polluted surface water have shown that even with improved viral
 methodology, no apparent relationship exists be'
tween the number of total or fecal
 coliforms in water and the number of viruses (Table 1). Viral isolates which included
 poliovirus 1,2,3, coxsackievirus B2, B4, and echovirus 1, 4, 25, were obtained from
 sample volumes of approximately 380 liters (1QO gallons); virus to, fecal coliform
 ratios ranged between 1:23 to 1:11,000. Although these ratios are considerably lower
 than previously reported, presumably due to improved recovery methodology, the
 coliform bacteria obviously greatly outnumber enteric viruses in sewage and polluted
 surface waters and thus still  appear to be  a more sensitive indicator of possible
 pathogen occurrences.                        i
   An important point which should be noted with respect to virus-coliform ratios is
 relative sample volumes.  Traditionally, coliform densities in waters  have been
 expressed as organisms per 100 ml or the approximate volume of a glass of water
 (Butterfield, personal communication). The standard drinking water sample volume
 for membrane filtration analysis was recently raispd from 50 ml to 100 ml. Byway of
 contrast, the methodology now being developed for isolating viruses from water
 (potable, surface, ground) processes from 1 to 1,9pO liters, and typical finished water
 sample volumes range from 380-1,900 liters. For equivalent density comparisons
 between viruses and coliforms, it is necessary to analyze comparable water volumes
 of a liter or more for each microorganism.


Efficacy of Water Treatment Processes

  A recent field study reported by Akin et al. (1) examined finished  water quality of
six municipal treatment plants in three states.  The treatment plants use surface
waters as  their  principal raw source water and  rely on conventional treatment
practices,  i.e. flocculation,  sedimentation, filtration, and chlorination (Table 2).
Bacteriological quality of the source waters was  variable, and at times exceeded the
recommended maximum fecal coliform density
(2000/100 ml) for surface water
sources (19). Although the primary objective of this study was to determine whether
human enteric viruses could be detected in finished waters using the equipment
/procedures  currently  under  development,  parallel  bacteriological  analyses

-------
                                                                                                           o
                                                                                                           §
                                                                                                           w
                                                                                                           HM
                                                                                                           O
TABLE 1. MICROBIOLOGICAL QUALITY OF OHIO RIVER WATER FROM CINCINNATI AREA*.                      £
Location1
Schmidt (467)3
Southside (475)3
Schmidt
Schmidt
Public Landing (470)3
Schmidt
Public Landing
Schmidt
Public Landing
Southside
Public Landing
Total Coliforms
(per 100 ml);
11,000
49,000
85,000
38,000
3,400
105,000
35,000
99,000
48,000
135,000
7,300
Fecal Coliforms
(per 100 ml)
6,500
17,000
23,000
11,000
370
43,000
3,500
42,000
7,200
33,000
1,200
.Standard Plate
Count per ml
37,000
62,000
79,000
48,000
20,000
162,000
112,000
193,000
46,000
79,000
8,300
Virus2
• Isolates
0
3
9
1
16
17
1
31
12
59
3
* From Brashear
1 Sampling period 9/8/75 through 12/1/75
2 From 100 gallon (380 liters) water samples
3 Ohio River Mile

                                                                                                            i
                                                                                                            i
                                                                                                            B

-------
  TABLE 2. PHYSICAL/BACTERIOLOGICAL SOURCE WATER CHARACTERISTICS AT
          WATER TREATMENT PLANTS*.
Location
Source Water(s)    Capacity/mean Output Fecal Coliforms F.C.
                 (10,000 mV day)    (MPN/lOOml) Geometric Mean
Columbus, OH
Sidney, UH
Muncie, IN
Seymour, IN
Kansas City, MO
St. Joseph, MO
Scioto River :
Tawana Creek, wells :
White River j
E. Fork White River ;
Missouri River .j
Missouri River ! '•
26.5 / 13.6
1.4 / I.I
6.0 / 4.9
0.56/ 0.45
79.5 / 41.5
11 / 5.3
38-3,000
40-2,800
50-2,000
1 1-14,000
1,700-4,300
280-7,000
280
263
210
536
3,010
1,890
                                                                                             s
                                                                                             b
                                                                                             69
                                                                                             I
*,Frpm Akin el al. (1)
                                                                           o
                                                                           •n
                                                                           $
                                                                                             1

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
   TABLES. TOTAL  COLIFORM/FECAL   COLIFORMS  RESULTS  OF
             FINISHED  WATfR  SAMPLES  FROM  SIX  TREATMENT
             .PLANTS*.
Treatment Plant
Location
Columbus, OH
Sidney, OH
Muncie, IN
Seymour, IN
Kansas City, MO
St. Joseph, MO
Membrane Fill
Total
Coliforms
0/4*
1/3
2/4
3/11
0/2
1/1
5r° Modified MPN* Large Volume Sampler'
Fecal Total
Colifcjrms Coliforms
o/
I/
o/
o/
o/
o/
I 2/4
3 2/3
4 0/4
11 5/11
2 2/2
1 0/1
Fecal Total Fecal
Coliforms Coliforms Coliforms
0/4 3/4 0/4
1/3 0/3 0/0
0/4 3/4 2/4
1/11 10/11 6/11
0/0 0/0 0/0
o/i . - -
* From Akin et al. (1)
" Up to 5 liters tested; negative results indicate a value <0.02/100 ml; the largest value obtained
 was 0.12 organisms/100 ml.       |
* 5.5 liters tested; negative results indicz te a value <0.02/100 ml; the largest value obtained was
 0.23 organisms/100 ml.
c 380 liters tested; negative results indicate a value <0.00059 organisms/100 ml; the largest
 value  obtained was >0.0043/100 ml.
J Occurrences/Number of Samples
involving large volume sampling protocols (380 liters) were performed to establish
coliform densities in finished waters.
  A summary  of the bacteriological findings is presented in Table 3. Using the
membrane filter procedure with up to  5  liters of finished water, coliforms were
detected in seven samples; one sample also contained fecal coliforms. The modified
MPN procedure (5.5 liter samples1) found  coliforms in 11 samples, 2 of which also
contained fecal coliforms. Finally], the large volume sampling method (380 liters)
found coliforms in 17 samples, 8 of which also contained fecal coliforms. In all cases,
coliform densities were  below the  maximum contaminant level (MCL) set  by
NIPDWR. More importantly, inJ56 finished water samples in which the average
volume of water analyzed was 1,100 liters, no viruses were isolated. Previous
laboratory evaluations of the sensitivity of the virus recovery procedure indicated a
detection level of 3-5 units (poliovirusjper 380 liters when 1,900 liters were sampled
(14). In contrast, the large volume sampling protocol for coliforms was able to detect
organisms using only 380 liter sample volumes.
  Thus,  the results of this study support the adequacy of conventional treatment
processes for removal of bacterial dnd viral pathogens from even the poorest quality
source waters (2,3,4)^ These data show that at least 99.9999% coliform reduction is
readily attainable in water treatment facilities with complete treatment; this percent
reduction is based on data in Table 3 where a negative large volume sampling result
from a highly polluted source water indicates a coliform density of < 0.00059/100
ml. Filtration, particularly when preceded by chemical floccuiation, is very effective
in removing microorganisms (17).
utilize rapid sand filtration prior
While the majority of water treatment facilities
to chlorination and distribution, a number of
utilities lack filtration capabilities  and rely solely on  chemical  disinfection to
eliminate microbial pathogens. Such treatment deficiencies have been associated
with increased outbreaks of giardiasis (8).

-------
                                  MICROBIAli QUALITY OF WATER/M.J. ALLEN
 Epidemiologica! Findings
   Twelve outbreaks of giardiasis including 4,860 cases in Rome, New York, were
 documented during the period  1971-1974 (8,9). all of these outbreaks, with the
 exception of Aspen, Colorado, involved municipal systems which used surface water
 supplies  that were only chlorinated  prior to 'distribution. Giardia, a flagellated
 protozoan, forms cysts which are not inactivated by chlorination at the dosages and
 contact times normally employed in water treatment processes; whether cdliforms
 were present or absent from systems associated with the outbreaks has not been
 established. Although high chlorine concentrations and protracted contact periods
 will inactivate the cysts,  it is recommended that nitration be implemented as part of
 the complete treatment of all surface waters. It is possible that with unfiltered surface
 waters, the coliform indicator system may not b e adequate to signal the presence of
 Giardia or other pathogenic cyst forming protozoans. Since the analysis for this
 organism is time-intensive, complex, and imprecise, the logical approach is  the
 elimination of the problem through improved water treatment practices, especially
 for finished water systems which have had a previous (chronic or intermittent) history
 of waterborne disease outbreaks attributed to trpatment deficiencies. Regarding the
 etiology of waterborne outbreaks of gastroenteritis, hepatitis-A, shigellosis, typhoid,
 and salmonellosis, Craun et al.  (9) concluded that treatment deficiencies, such as
 inadequate disinfection, interruption of disinfection or lack of treatment of ground
 waters were the primary causes.

Conclusions

  With the implementation of  the National  Interim Primary  Drinking. Water
Regulations, the coliform group of microorganisms  remains  the  most relevant
sanitary index for finished waters. Recent findings confirm earlier  work that the
coliform indicator concept is adequate in assuring bacterial and viral safety of water.
The inability of present  virological methodology to isolate viruses, from finished
waters results in  continuation of our reliance on and confidence in the coliform group
for monitoring finished water  quality.
  Although surface water quality may vary widely
through protection of receiving waterways, pres
and must be continually improved
ent water treatment practices are
capable of consistently producing pathogen-free water to the consumer. The .critical
biological safeguards in water treatment continue to be flocculation/filtration with
post-chlorination. Excluding protozoan cysts, all bacterial and viral pathogens are
readily inactivated by chlorine dosage and contact times routinely used in properly
operating water treatment facilities.  The key to eliminating waterborne disease
outbreaks, however, is constant surveillance of finished waters and their distribution
network and the maximum use of good treatment processes. Treatment plants with
any  deficiencies must be  upgraded  to remedy these inadequacies and assure
production of the best quality potable water.
                             REFERENCES

 1. Akin, E.W., D.A. Brashear, E.G. Lippy, and N.A. Clarke. 1975. A virus-in-water study of
   finished water from six communities. Health Ejects Research Laboratory, Office of
   Research and Development, U.S. Environmental Protection Agency, Cincinnati, Ohio.
 2. Berg, G.  1966. Virus transmission by the water vet icle: III. Removal of viruses by water
   treatment processes. Health Lab. Sci. 5:170-181.
 3. Chang, S.L., R.E. Stevenson, A.R. Bryant, R.L. Woodward, and  P.W". Kabler. 1958.
   Removal of Coxsackievirus and bacterial virus in  water by flocculation: II. Removal of
   Coxsackievirus and bacterial viruses and native bacteria in raw Ohio River water by
   flocculation with aluminum sulfate and ferric chlorate. Am. J. Public Health 45:159-169.

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


 4.  Chaudhuri, M., and R.S. Engelbrecht. 1970. Removal of viruses from water by chemical
    coagulation and flocculation. JAWWA 62:563-567.
 5.  Clark, H.F., and P.W. Kabler. 1964. Reevaluation of the significance of coliform bacteria.
    JAWWA 55:931-936.
 6.  Clarke, N.A., G. Berg, P.W. Kabler, and S.L. Chang. 1964. Human enteric viruses m water:
    source, survival, and removability. International Conference on Water Pollution Research,
    Pergamon Press, Oxford.
 7.  Clarke, N. A., H.D. Nash. 1972. Equating bacterial standards with viral numbers in potable
    waters. Proceedings Annual Meeting of the  American  Society for  Microbiology,
    Philadelphia, Pennsylvania.
 8.  Craun, G.F. 1976. Microbiology-waterborne outbreaks. J. Water Pollut. Control Fed.
    45:1378-1397.                                                       .  •
 9.  Craun, G.F., L.J. McCabe, and J.M. Huges. 1976. Waterborne disease outbreaks m the
    U.S. (1971-1974). JAWWA 55:420-424.                    ,
10.  Engelbrecht, R.S., D.H. Foster,  E.O. Greening, and  S.H.  Lee. 1974. New microbial
    indicators of wastewater chlorination  efficiency. Environmental Protection Technology
    Series EPA-670/2-73-082. U.S. Environmental Protection Agency, Washington, D.C.
11.  Geldreich, E.E. 1966. Sanitary significance of fecal coliforms in the environment. Water
    Pollution Research Service Publication No. WP-20-3. U.S. Department of Interior, Wash-
    ington, D.C.
12.  Geldreich, E.E. 1977. Bacterial  populations and indicator concepts  in feces, sewage,
    stormwater, and solid waste.  In: Berg, G. (Ed.), Indicators of viruses in water and food.
    Publication, Ann  Arbor,  Michigan.
13.  Geldreich, E.E., and N.A. Clarke. 1972. The coliform test: a criterion for viral safety of
    water. Proceedings 13th Annual Sanitary Engineering Conference, Urbana, Illinois.
14.  Hill, W.F., W. Jakubowski, E.W. Akin, and N.A.  Clarke.  1975. Detection of viruses in
    drinking water: sensitivity of the tentative standard method. Annual Meeting American
    Society Microbiology Proceedings, New York.
15.  Kabler, P.W. and H.F. Clark. 1960.  Cqliform  group and fecal coliform organisms as
    indicator of pollution in drinking water? JAWWA 52:1577-1579.
16.  Kabler, P.W., N.A. Clarke,  G.  Berg, and S.L.  Chang.  1961.  Viricidal efficiency of
    disinfectants in water. Public Health Reports 75:565-570.
17.  Robeck,  G.G., N.A. Clarke, and K.A. Dostal. 1962. Effectiveness of water treatment
    processes in virus removal. JAWWA 54:1275-1292.
18.  Shuval, H.I. 1976. Water need and usage, pp. 12-25.  In: Berg, G. et al (Eds.), Viruses in
    water. American Public Health Association, Washington, D.C.
19.  The Committee on Water Criteria. 1972. Water  quality criteria-1972.  Environmental
    Studies Board, National Academy of Sciences, National Academy of Engineering, Wash-
    ington, D.C.
20»  U.S. Environmental Protection Agency. 1975. National Interim Primary Drinking Water
    Regulations. Federal Register 40:59566-59574.
21.  U.S. Public Health Service.  1914. Bacteriological standard  for drinking water. Public
    Health Reports 29:2959-2966.
             QUESTION  AND  ANSWER SESSION.  '   .
                                                         i '
G.W. Fuhs, Division of Laboratories & Research, New York Department of Health,
Albany, New York
  Is a standard plate count on  groundwater more sensitive to contamination which
may have occurred one week previous  to sampling?
M.J. Allen, Water Supply Research Laboratory, U.S. Environmental Protection
Agency, Cincinnati, Ohio
Waters from undisturbed or protected aquifers, unlike those from surface sources,
have low indigenous  bacterial  populations. However, in shallow aquifers where
percolating water from surface sources can provide sufficient nutrients to support
microbial growth, an unusually high bacterialpopulation could indicate infiltration
of contaminated water  into the aquifer. More  importantly, abnormally  high
10

-------
                                  MICROBIAL QUALITY OF WATER/M J. ALLEN
microbial levels in shallow aquifers can adversely affect detection of the more specific
indicators. Thus, coliforms could be undetectable in ground water where excessive
bacterial densities occur. The failure to detect coliform bacteria results, unfortunate-
ly, in the assumption that these contaminated supplies are safe to drink. Therefore, it
could be necessary to employ additional micirobial indicator groups for ground water.
These indicator  organisms  may supplement or replace the coliform group when
ground water conditions  are such that erroneous results could  occur.
                                                                         11

-------
,,|'  	Ji,	HI,,!I   illli,,!,


-------
                  INTERFERENCES TO COLIFORM DETECIjION/E.E. GELDREICH


                  Interferences to Coliform Detection i n
                          Potable Water Supplies

         Edwin E. Geldreich, Martin J. Allen and Raymond! H. Taylor
                 Water Supply Research Laboratory/ MElf L    !
                    U.S. Environmental Protection Agency)
                           Cincinnati, Ohio 45268


  The monitoring of potable water supplies must always jitilize. bacteriological
procedures that are as sensitive for coliform detection as the stjite-of-the-art permits.
This goal is of particular importance in a. continuing surveiljance for low levels' of
coliform bacteria that signal water treatment or transmission  deficiencies which
could allow the breakthrough of waterbotne pathogens. White there has been much
effort  directed  towards  improving  methodology,  standardization of testing
procedures, and achieving greater levels of data reliability through quality control
and  laboratory certification programs,  little attention  has I been  given to those
peripheral  water characteristics that  may  impact on the detection of indicator
organisms. This concern has become more critical with the change of the official MF
test volume from 50 ml to 100 ml in the Drinking Water Standards, in order to obtain
better statistical reliability at the baseline limit of one coliform  per 100 ml. These
interferences relate to the nature and density of potable water  turbidity or to the
composition and magnitude of the finished water microbial flora.'

Turbidity in  Potable Water                            j     ] .
  Turbidity in potable water may be derived.from source Water, water treatment
practices, problems in the distribution system, or a combination of these factors.
Thus, turbidity may range from algal growth, natural siltation
industrial waste discharges to  surface v/aters, chemicals hi
filtration practices, residual coagulants in  water treatment,
                                                        , decaying vegetation,
                                                         ground water, poor
                                                         corrosion, microbial
growth or soil  intrusions in wlter disfiribution systems. As  a result.of these
conditions, a variety of materials may be represented in turbidjity, some of which can
impart a  chlorine demand. Other turbidities may be sources  of interference to
coliform detection, and only a few kinds of material will bje of little concern as
protective habitats for microbial survival.
                                                         of treatment barriers.
                                                         ral mineral  turbidity,
  \J i,\j\^l,L V W 4AC4-^/Jll.C*.l.O ± Vy »  ii«.*v* •*-• v Jtwi wv--. ••*—••-•
  Turbidity can be a transport vehicle for bacterial penetratior
While sand particles may be of little concern, other'natu
hydrated oxides from flocculation, and organic debris have bleen shown to provide
absorption sites for bacteria (16,17,18,22),, These conditions are of major concern in
those supplies that employ disinfection as the sole treatment rirocess for raw surface
waters (22). Without use of more extensive treatment processes, the barrier to
microbial penetration becomes inadequate because of turbidity fluctuations caused
by stormwater runoff, algal blooms, seasonal turnover of impounded waters and
water uses upstream. Where water treatment processes are poojrly controlled, natural
turbidity, alum floe particles or. cationic  polymers used in coagulation may break
through  filter beds into the distribution system.  These conditions will produce
turbidity levels higher than 1 NTU and may bring associated bacteria into the potable
water.      .                                            '

Turbidity Interferences                                i
  Suspended material (turbidity) in a potable water sample niay preclude use of the
membrane filter coliform procedure. Reasons for rejecting thi MF procedure relate
to the volume of samples that can be filtered, character of suspended material, and
the thickness of the  suspended material that deposits on the membrane filter surface.
                                                                           13

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                              	  ]
                                                       |          -      -,  '  '
 Relatively thin layers of gelatinous, finely divided, or hygroscopic materials, such as
 suspended ferrous, manganese, and alum floes or algal cell populations may clog
 filter pores or cause a confluent film of growth to develop during incubation. Thick
 surface layers of crystalline or siliceous materials may cause little or no difficulty.
   There is no practical method of removing turbidity without also removing some
 unpredictable portion of the bacterial population including coliform bacteria in the
 sample. Our experience has indicated that pre-filtering turbid water through a coarse
 filter or decanting the supernatant from settled turbidity in a water sample results in a
 20 to 80 percent loss of organisms. Turbidity limits for membrane filtration have
 never been determined because the limits vary with the type of material (sand VS.
 clay), chemical composition (phosphates, coagulants) and presence of various algal
 masses. While the multiple tube method may be the method of choice when testing
 potable water samples that contain turbidity interferences to membrane filtration,
 this approach may have some undesirable consequences. Selection of the multiple
 tube method automatically limits routine analyses to a 50 ml.test portion, rather than
 the 100 ml test portion used with the membrane filter procedure. Thus, the theoretical
 detection limit for total cqliforms in potable water becomes 2 organisms per 100 ml
 and assumes that  these organisms are uniformly  distributed in two  30 ml test
 portions. However, uniformity of bacterial suspension may not be achieved in turbid
 samples. Adsorption pf bagteria to turbidity particles and settling of heavier particles
 are  factoFs that must  be  rectified since the  random occurrenc.es  of only three
 coliforms in the 50 ml test portion is sufficient to produce an unsatisfactory result,
 Coliforms trapped in turbidity particles may  not produce gas in the presumptive
 medium unless released by vigorous sample agitation, This condition may be the
 cause of occasional differences in results from duplicate samples analyzed at the local
 water plant laboratory and at the State laboratory after indeterminant intervals  of
 agitation during transit,
                                                     :  !
                                              :         ! !
 Microbial Interferences
                                                     :  i
  While  water produced at the treatment  plant may be of high bagterjqlogiqa!
quality, finished water  entering a  distribution network is subject to continuous
bacteriological deterioration (1,4,7,8,10,12,19,20). Some factors which contribute to
the deterioration  include: 1) inadequate dosage or loss of disinfectant residual; 2)
incomplete removal of organics and/or turbidity due to treatment deficiencies; 3)
seasonal  temperature fluctuations; 4) extended retention periods  in reservoirs,
standpipes, and dead-end water mains; 5) sediment accumulations; and 6) line breaks
and  cross-connection  occurrences  in the  distribution network.  Although  the
microbiological flora  in finished waters is highly variable, bacterial groups most
frequently  associated with  finished water deterioration include: Pseudomonas,
Flavobacterium, Achromobacter, Proteus, Klebsiella, Bacillus,  Serratia, Coryne-
bacterium, Spirillum,  Clostridium, Gallionella.-Arthrobacter, and Leptothrix (6),
These organisms, in general, are able to survive and many may even multiply in
finished waters.  Such bacterial regrowth may result in  slime deposits, tubercle
formation, development of taste and odor problems, accelerated pipe deterioration
(6,12,20,21) increased health risk (14) and may also interfere with bacteriological
monitoring procedures.
  Although finished waters which have been shown to be CQljform free are generally
considered safe for human  consumption, there are occasions when the absence of
detectable coliforms cannot be equated with potability. Excessive densities of non-
coliform 'organisms in finished water may desensitize assay procedures for total
coliforms (5,9,11,23). These non-coliform organisms can coexist with coliforms in
water, but when introduced into lactose broth, many of the non-coliform organisms
may multiply at a rapid rate, thus intensifying the factor of coliform inhibition (15) or
simply outgrow a smaller coliform population.

 14

-------
                 INTERFERENCES TO COLIFORM DETECTION/E.E. GELDREICH
   In  laboratory tests conducted by Hutchison et al (17), suspensions of various
 antagonistic organisms (Pseudomonas, Actinomyces, Sarcina, and 'Micrococcus) in
 a density range of 1,000 to 2,000 per ml, when added to lactose tubes simultaneously
 to a suspension containing 10 E. coli per ml, resulted in reduced coliform detection.
 The loss of sensitivity ranged from 28 to 97 percent, depending upon the combination
 of mixed strains in these experiments.  Considering that the baseline for coliform
 detection is one organism per 100 ml, any loss of test sensitivity becomes critical.
  Analysis of bacteriological data from survey of distribution water samples from
 969 public water plants in this nation suggested that interference with coliform
 detection may be a problem (13). These data (Table 1) showed that the frequency of
 detecting total and fecal coliforms by the MF procedure, increased ,as the standard
 plate count increased to levels up to 500 per ml, but decreased in frequency when the
 non-coliform bacterial population exceeded  1,000 organisms per ml. These findings
 again suggest that high densities  of non-coliform bacterial populations have an
 adverse affect on detection of the coliform indicator.
  While provisions for a Standard  Plate Count measurement are  not currently
 included in the National Interim Primary Drinking Water Regulations, field studies
 in several different geographical areas  are now in progress to obtain additional
 information on the impact of the total microbial populations on MPN and MF
 methodology. One of these studies involves the Cincinnati water distribution system.
 This public water supply network, like many other systems, has numerous dead-end
 water mains. The water in these low flow mains is subject to abnormal deterioration,
 as demonstrated by the number of customer complaints in the areas, of a number of
 these  low  flow sections. For  this reason,  the  Water Distribution Division-
 Maintenance Section of the Cincinnati Water W.orks has maintained a routine
 flushing program  for at least 30 years.
  Weekly, a number of troublesome dead-end  water mains are flushed to clear out
 accumulated  sediments and bring fresher  waters  with free chlorine residuals into
 these  mains.  During a two year period, with the cooperation of the Maintenance
Section personnel, samples from these, flushes were ^obtained for'bacteriological
 analysis. All  samples  were  iced immediately and processed within  five hours of
 collection. Bacteriological analyses of 613 water samples from 32 sampling sites
 included the  total coliform multiple tube and membrane filter  procedures and a
   TABLE 1. BACTERIAL PLATE COUNT VS. COLIFORM DETECTION IN
             DISTRIBUTION  WATER  NETWORKS  FOR  969  PUBLIC
             WATER SUPPLIES.

General Bacterial Population*          Total Coliform          Fecal Coliform
Density Range     Number of
   per 1 ml           Samples  Occurrences  Percent  Occurrences  Percent
<1
11
31
101
301
501
>1
- 10
-30
- 100
-300
-500
- 1,000
,000
1013
371
396
272
120
110
164
47
28
72
48
30
21
31
4.6
7.5
18.2
17.6
25.0
19.1
18.9
22
12 :
28
20
11
9
5
2.2
3.2
7.1
7.4
9.2
8.2
3.0
  TOTAL            2446        277        —        107


'* Standard Plate Count (48 hrs. incubation, 35° C)
                                                                         15

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
standard plate count  (SPC agar pour plates, incubated at 35° C for 48 hours).
Physical/chemical  parameters measured  were free chlorine residual (DPD test),
turbidity, water temperature and pH.
   Bacteriological analysis of water samples obtained from dead-end water mains
(Table 2) showed that 76 of 613 samples contained coliforms when examined by the
MPN procedure. In contrast, the membrane nitration technique detected coliforms
in only 19 of these samples. Thus, the MF method detected coliforms in only 25% of
the  samples  that  were  positive  for  coliforms  by the  MPN  method. Closer
examination of the  data shows a correlation between excessive bacterial densities and
desensitization of the  membrane filter procedure.
   While detection of coliforms by both the M PN and M F techniques increased as the
standard plate count increased,  the MF method detected disproportionately fewer
coliforms as  noted in the range of MF coliform values compared to the MPN
coliform values per 100 ml. Although it is  not apparent from the data in Table 2 that
the MPN method is also subject to interferences by SPC organisms, a limited number
of water samples examined by the method displayed growth inconsistencies. Several
MPN tubes that showed no gas, but heavy growth, were inoculated into brilliant
green tubes and incubated at 35° C for  48 hours. These  tubes did produce gas,
indicating that the competing non-coliform  bacteria had suppressed the normal
metabolic pathway to gas production by coliforms in the presumptive medium. Such
false-negatives were also noted  with the  MF procedure. Membrane filter cultures
with excessive densities  of colonies but no  differentiated sheen colonies, were
aseptically  transferred into lauryl tryptose broth for evidence of gas production
within 48 hours at  35° C then confirmed in brilliant green bile broth. This extension
of the  MF  procedure  also revealed the presence of undetected coliforms that were
marked by high densities of the Standard Plate Count.

Controlling  Interferences Through Treatment
  Improved water  treatment practice is the first line of  defense against  problems
related to turbidity in finished water. With reduction of turbidity to below 1  NTU,
      TABLE 2. COLIFORM OCCURRENCES VS. STANDARD PLATE
                COUNT.
                                             '   '  • I
                                   Standard Plate Count Range (per ml)
 	__	< 500	501-1000	> 1000
 All Samples*                   502             •  35              51
 Samples with Coliforms**
   (MPN Method)               50               8                15
 Percent of Samples             10%            22.9%            29.4%
 MPN Range (per 100 ml)       2-49             5-79     ,        2-920
 MPN Mean (per 100 ml)         16               18              110
 Samples with Coliforms* * *
   (MF Method)                11               2                5
 Percent of Samples             2.2%            '5.7%            9.8%
 MF Range (per 100 ml)        < 1-18           < 1-50          < 1-90
 MF Mean (per  100 ml)          2                7                13
 MPN % Occurrence less
   MF % Occurrence           7.8              17.2              19.6

 *  588 of 613 samples had countable SPC
 ** 3 MPN positive samples did not have countable SPC
 *** 1 MF positive samples did not have countable SPC
                                                  •  !
16

-------
                  INTERFERENCES TO COLIFORM DETECTION/E.E. GELDREICH
  TABLE'S. MPN  COLIFORM  OCCURENCES  VS.  FREE  CHLORINE
             RESIDUALS (DPD)*
Free Chlorine               Number of Coliform         SPC Average
Residual (mg/1)	  Occurrences	(bacteria/ml)

<0 1                  .    .          58                     3040
0.1-0.5                        ,13                      860
0.6-1.0                                4                       50
>1.0                                  1     ,                 79   .
* Typical  values  for finished  waters were  1-2 mg I  free chlorine  and  Standard Plate
 Count < 1' ml.              ...       .

disinfection  efficiency  increases and enhances  control  of bacterial regrowth
throughout the distribution network. Analysis of the effectiveness of free chlorine
residuals in controlling bacterial deterioration, as demonstrated in this study (Table
3) and in the Community Water Supply Study (Table 4) prove the feasibility of this
approach for  minimizing the potential non-coliform interferences to coliform
methodology (6,8,22,23). Pipe sediment accumulation in low-flow water mains and
dead-end lines is a source of both intermittent turbidity and nutritive materials that
support  bacterial persistance and  possible regrowth problems that  may cause
coliform interferences. The total elimination of dead-end water mains is, of course,
impossible,' but implementation of  practices  such as  periodic flushings,  main
cleaning, maintenance of free chlorine residuals and modification  or addition of
various treatment processes can substantially improve distribution w&ter quality in
dead-end mains and low flow sections.                           '   .   ',   .
  Chemical characterization of 5 sediment samples from the Cincinnati distribution
system confirms'our argument that this material serves as a nutrient supply for
bacterial regrowth. Dried sediment samples were  pulverized with O.I0 gm of each

  TABLE 4 THE EFFECT OF VARYING  LEVELS OF RESIDUAL CHLO-
           ' RINE ON THE TOTAL PLATE COUNT IN POTABLE WATER
            DISTRIBUTION SYSTEMS*.
Standard Plate           ,'              Residual Chlorine (mg/1)
  Count**     ./               ...            '

                     0.0    0.01    0.1    0.2    0.3    0.4     0.5     0.6
< 1 .. .
1-10
11 - 100
101 - 500
501 - 1000
> 1000
Number of
Samples
8.1%,
20.4
37.3
18.6
5.6
10.0

520
14.6%
29.2
33.7 •
11.2
6.7
4.5

89
• 19.7%
38.2
28.9
7.9
1.3
3.9

76
12.8%
48.9
26.6
9.6
2.1
0

94
16.4%
45.5
23.6
12.7-
1.8
0

55
17.9%
51.3
23.1
5.1
0
2.6

39
'4,5%
59.1
31.8 .
4.5
,0
0

: 22
17.9%
42.9
28.6
10.7
0
0

28
 * Data from a survey of community water supply systems in 9 metropolitan areas (54)
** Standard Plate Count (48 hrs. incubation, 35° C)
                                                                         17

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                     I

 sample added to 200 ml of distilled-deionized water and allowed to solubilize for 3
 hours at ambient temperature. An aliquot of the supernatant was removed, acidified
 with HC1 and analyzed with Dohrmann DC-52 carbon analyzer. Additional aliquots
 of the supernatant were removed for Kjeldahl-nitrogen and nitrate-nitrogen analysis.
 Kjeldahl-N ranged from 250 to 575 mg/1, nitrate-nitrogen from 4400-5200 mg/1 and
 total organic carbon from 200 to'550 mg/1. Bacterial regrowth becomes a reality in
 slow flow sections or dead end lines where such high levels of bacterial nutrients occur
 in pipe sediments especially with water temperatures above 13° C, pH below 10, and
 no free  chlorine residual or ineffective chloramines present. Laboratory  growth
 studies using distilled  water extracts  of pipe sediments showed that nutrient
 concentrations were capable of supporting bacterial growth from initial densities of
 less than 100 SPC organisms per ml to 300,000 organisms per ml within 10 days at
 20°C.                         ;
                                  ' '	 'I                    '  !

Controlling Interferences Through Methodology
  Until all  interferences created  by inadequate water treatment and distribution
practices  are eliminated in  problem water supplies, the laboratory must remain
vigilant to the problem of coliform detection in potable waters of varying qualities.
Interferences to coliform detection often go undetected, partly because of a Standard
Methods concept that implies that no coliforms are present in a multiple tube test
when there is no evidence of gas production within 48 hours, or with the membrane
filter procedure when sheen colonies are absent.  In the multiple tube procedure,
consideration should be given to confirmation of test samples that produce 5
presumptive tubes with heavy growth but lack visible gas production. For membrane
filter cultures that lack diiferentiated sheen colonies and that have high background
densities of non-differentiated colonies, the entire membrane filter should be placed
in lauryl tryptose  broth to  test for gas production followed by confirmation  in
brilliant  green bile broth. This procedure would yield qualitative information on
suppressed coliform  occurrences that  might otherwise go undetected. Since some
types of turbidities between 1 and 5 NTU physically block membrane filter pores,
impede flow through and prevent development of discrete differential colonies, the
laboratory will be obligated to use the multiple tube procedure until the interference
problem is corrected through  improved treatment or distribution system mainte-
nance.
Conclusion
  In an effort to detect coliform occurrences at a baseline of one organism per 100 ml,
potable water must be free of interferences from turbidities that exceed the 1 NTU
limit specified iiUheJtetedm_£rjbiary Drinking Water Regulations and bacterial
densitiejjjy^u^blorganismspermhshould not be tolerated. To some extent, these
interferences  may  be circumvented  in  the  laboratory through alterations in
methodology. However, these  measures should only  be considered temporary
expedients to the monitoring program. The ultimate solution to these issues resides in
improved treatment practices anjd attention to minimizing deterioration of water
quality during distribution.     >
                            REFERENCES
                              j               ,      i  i •..                     .1
                              i              "	      i  j| • •    •                 !
 1. Bayliss, J.R. 1938. Bacterial aftergrowths in water distribution systems. Water Works and
   Sewage 55:720-722.           i
 2. Buelow, R.W., R.H. Taylor, E.E.,Geldreich, A. Goodenkauf, L. Wilwerding, F. Holdren,
   M. Hutchinson, and H. Nelson. 1976. Disinfection of new water mains. JAWWA 66:283-
   288.

 18                            .              "'     '   !|    ':

-------
                    INTERFERENCES TO COLIFORM DETECTION/E.E. GELDREICH
  3. Buelow, R. W., and G. Walton. 1971. Bacteriological quality vs. residual chlorine. JAWWA
    65:28-35.                                      :
  4. Committee on Water Supply. 1930. Bacterial aftergrowths in water distribution systems.
    Amer. J. Public Health 20:485-491.
  5. Fischer, G,  1950. The  antagonistic effect of aerobic sporulating bacteria on the coli-
    aerogenes group. Zeit. Immun. u. Exp. Then 707:1,6-20.
  6. Geldreich, E.E.  1973. Is the total count necessary? Sections VII-l-VII-2. Proceedings
    AWWA Water Qualtiy  Technology Conference, Cincinnati, Ohio.
  7. Geldreich, E.E., R.H. Taylor, and M.J. Allen. 1974. Bacteriological consideration in the
    installation and repair of water mains. Sections! VIII-j-VHI-5. Proceedings AWWA Water
    Quality Technology Conference, Dallas, Texas.    ;
  8. Geldreich, E.E.,  H.D. Nash, D.J.  Reasoner, and R.H. Taylor.  1972. The necessity of
    controlling bacterial populations in potable waters : Community Water Supply. JAWWA
    6^:596-602.
  9. Hutchinson, D., R.H. Weaver, and M. Scherago.  1943. The incidence and:significance pf
    microorganisms antagonistic to Escherichia coli in water. J. Bacteriol. 45:2S),
 10. Jewell, A.B. 1942. Bacterial aftergrowths in the  distribution system. Southwest Water
    Works J. 23:13-14.                             ;
 11. Kligler, L.J. 1919. Non-lactose fermenting bacterial from polluted wells and sub-soil. J.
    Bacteriol. 4:35-42.                              I
 12. Lee, S.H., and J.T. O'Connor. 1975. Biologically mediated deterioration of water quality in
    distribution  systems. Proceedings AWWA Water  Quality Technology Conference,
    Atlanta, Georgia.                              j
 13. McCabe, L.J., J.M. Symons, R.D. Lee, and G.G. Robeck. 1970. Surveylof community
 .   water supply systems. JAWWA 62:670-687.     •  I
 14. Peterson,  N., and  M. Favero. 1975.  Significance of gram-negative bacteria in wate;
    supplies. Proceedings AWWA Water Quality Technology Conference, Atlanta, Georgia.
 15. Reitter, R., and R. Seligmann. 1947. Pseudomonas aeruginosa in drinking water. J. Appl.
    Bact. 20:145-150.
 16. Sanderson, W.W.,and S. Kelly. 1962. Discussion of human enteric viruses in water: source,
    survival and removability;  pp. 536-541.  International  Conference:Water Pollution
    Research,  Vol. 2, London.                      !                   :
 17. Sen, R., and B. Jacobs. 1969. Pathogenic intestinal organisms in the unfiltered water supply
    of Calcutta and  the effect  of chioririatio'n. Indian  J. MecirRes. 57:1220-1227.
 18. Tracy, H.W., V.M.  Camerena, and  F. Wing.  .19,66. Coliform  persistence in  highly
    chlorinated water. JAWWA 58:1151-1159.        ,
 19. Victoreen, H.T. 1974. Control of water quality in transmission and distribution  mains.
    JAWWA 66:369-370.                           ;
 20. Victoreen, H.T.  1973. What's the significance of nuisance organisms? Sections IX-l-IX-6.
    Proceedings AWWA  Quality Technology Conference, Cincinnati, Ohio. :
 21. Victoreen, H.T.  1969. Soil bacteria and color problems in distribution systems. JAWWA
    61:429-431.                                    ;
 22. Walton, G. 1961. Effectiveness of water treatment processes as  measured by coliform
    reduction. Public Health Service Publication 898. United  States  Public Ijealth Service,
    Washington, D.C.
 23. Weaver, R.H., and T. Boiter. 1951. Antibiotic-producing species of Bacillus from well
    water. Trans. Kentucky Acad. Sci. 73:183-188.
             QUESTION AND  ANSWER SESSION
C.D. Cameron, Fairfax County Water Authority, Occoquan, Virgina
What effect would public announcements urging residents to use less water during
water shortages have upon the distribution system water quality? Would not a long
holding time (2 weeks versus 2 days) in system mains and tanks encourage loss of C12
and an increase in total population?              i                    ,

                                                                                19

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS

                                                  :  ! .                     '
 E.E. Geldreich, Water Supply Research Laboratory/MERL, U.S. Environmental
 Protection Agency, Cincinnati, Ohio
  There are several factors not denned that will impact on the status of the standard
 plate count population in a distribution network during slow flow periods. During
 periods of warm water temperature (above 13° C) and increased chlorine demand in
 the distribution lines, regrowth of the bacterial population will occur. Some of this
 regrowth potential might be suppressed where the finished water pH ranges from 9.0
 to  10.0.  If chloramines are carried as a residual, longer contact  time  in  the
 distribution network would result in some limited control over bacterial regrowth.
 As  a temporary counterforce,  less abrupt changes in  water pressure during this
 period would reduce the mixing of sediments into the water flow. However, our data
 would  suggest  the overall effect would be a gradual degradation of  the water
 regardless of limited benefits achieved by longer contact time and water quiescence.

                                                    I"
 R.R. Colwell, Department of Microbiology, University of Maryland, College Park,
 Maryland
 Since  90% of the total count  is missed when pour plates of water samples are
 incubated at 35° C, why not use lower incubation temperatures?

 E.E. Geldreich
  We  recognize that not all viable bacteria from a mixed  population in a water
 sample can be cultivated on a single medium, at any one incubation temperature or
 for one specified incubation time. The critical concern with the general population of
 bacteria in potable water lies primarily in the potential interferences they may create
 in the test for coliforms. The other determiningfactor is to attempt to include some
 indirect measurement  for  Pseudomonas,  Flavobacterium, and  other ^secondary
 pathogenic invaders which could pose a health risk^ in the ^aspttaTenvjracmient.
 Since the coliform test is incubated at 35° Cand the~potential secondary pathogenic
 invaders optimally grow at body temperature, the choice of a non-specific bacterial
 count in water at 35° C was logical. Furthermore, any attempt to specify a limit for
 this portion of the general bacterial population must follow strict adherence to a
 specific protocol for medium, incubation temperature and time. Thus the expression
 "Standard Plate Count" becomes appropriate and the bacterial population limited to
 growth on Standard Plate Count agar, with incubation of pour plate cultures at 35° C
 for 48  hours.
20

-------
                                        DISEASE TRANSMISSION/G.F. CRAUN
                 Impact of the Coliform Standard on the
                         Transmission of Disease

                             Gunther F., Craun
                         Chief, Epidemiology Branch
                           Field Studies Division
                     Health Effects Research Laboratory
. •                          Cincinnati, Ohio 45268


  Dr Hendricks, the Symposium Chairman, has asked that I review and summarize
the data on waterborne disease outbreaks in the United States, concentrating on the
water system deficiencies that allowed the outbreaks to occur. I will also discuss the
bacteriological quality of water provided by the water supplies involved in outbreaks
and whether  bacteriological surveillance as proposed in the National Interim
Primary Drinking Water  Regulations will prevent the  occurrence of waterborne
outbreaks.                                               '     ;   .    .
  The Center for Disease Control (CDC) and the Environmental Protection Agency
(EPA) jointly conduct surveillance of waterborne disease. Outbreaks are reported to
the CDC and the EPA by state and local health departments and by water  supply
agencies. This report summarizes data reported during  1971-1975. ;
  Only outbreaks associated with water used for drinking or domestic purposes are
included in this analysis. To be considered an outbreak at least two cases of infectious
disease  must be  reported before a common source can be noted and investigated.
Except  in unique circumstances, such as a case of chemical poisoning in which the
chemical was identified  in the water, a single case cannot be recognized as  having
been caused by drinking water. The waterborne outbreaks reported here are those in
which drinking  water has been  implicated epidemiologically as the vehicle ol
transmission of  the illness. In most of the outbreaks, the water was found to be
bacteriologically or chemically contaminated.                    '
   For analysis,  the water systems were classified as municipal, semi-public, or
individual. Municipal water systems are denned as public or investor-owned water
supplies that serve communities. Individual water systems are those used exclusively
by single residences in areas without municipal  systems or by persons travelling
outside of populated areas (e.g., backpackers). Semi-public water systems, located in
areas not served by municipal systems, are developed and maintained for a group ot
residences (e.g., subdivisions and trailer  parks) or at locations where the general
public  has  access  to  drinking water  (e.g.,  industries, camps,  parks,  resorts,
institutions, and hotels). The definition of municipal water system and semi-public
water system correspond to the Safe Drinking Water Act's definitions of community
water system and non-community water system, The  major exception is that small
subdivisions and trailer parks are included as semi-public systems in this analysis,
whereas the Safe Drinking Water Act defines these as  community water systems.
   During the period 1971-1975, 123 waterborne outbreaks were documented in the
 U.S., resulting in almost 28,000 cases of illness (Table 1). The three largest outbreaks
occurred in municipal water systems: Sewickley, Pennsylvania (5,000 cases ol
 gastroenteritis)  in 1975; Rome, New York (4,800 cases of giardiasis) m 1974; and
 Pico Rivera, California (3,500 cases of gastroenteritis) in 1971. Two deaths were
 associated with  outbreaks of waterborne disease during this period.
   The  mean  annual number of reported waterborne outbreaks during 1971-1975
 was 25. This is twice as many outbreaks as the mean annual number reported during
 the period 1951-1970 and equals the mean annual number reported during the period
 1920-1936 (Figure 1). The reason for this apparent increase in number of outbreaks is
 difficult to ascertain but is felt to be primarily the result of increased reporting and
 surveillance activities.         :
                                                                          21

-------
KJ
                                                 Source: G.F. CRAUN
                                                        U.S. Environmental Protection Agency
                                                                          Cincinnati, Ohio
     1920-251926-301931-36 1938-401941-451946-50  1951-551956-60 1961-65  1966-70
                                               YEAR
1971-75
           n
           §
           W
                                                                                            i
                                                                                            n
                                                                                            £
                                                                                            SB
                                                                                           I
              Figure 1. AVERAGE ANNUAL NUMBER WATERBORNE OUTBREAKS, 1920-1975.

-------
                                        DISEASE TRANSMISSION/G.F. CRAUN


          TABLE 1. WATERBORNE DISEASE OUTBREAKS.
                              1971-1975     i

                 1971       1972        1973        1974        1975   TOTAL

Outbreaks         19         29         26   .      25         24     123
Cases of Illness    5182       1638       1774  j     8356       10,879   27,829


   The 123 waterborne outbreaks were classified b,y type of water system (Table 2).
 More outbreaks occurred in semi-public water systems (57%) than municipal systems
 (30%) and individual systems  (13%);  however, outbreaks in municipal systems
 affected an average of 504 persons compared to 129 persons per outbreak in semi-
 public and 9 persons per outbreak in individual  systems.  Although most of the
 outbreaks (57%)  occurred in semi-public water systems, most of the illness (67%)
 resulted from outbreaks in municipal systems.   '•
     TABLE 2  WATERBORNE OUTBREAKS BY TYPE OF SYSTEM.
                                1971-1975    !
                                            i
                                        OUTBREAKS   CASES OF ILLNESS
 Municipal Systems                             37             18,633
 Semi-Public Systems                           70              9,058
 Individual Systems                             16               138
 Total                             ~         i  123             27,829


   An etiologic agent was determined in only 49% pf the outbreaks during the period
  1971-1975 (Table 3). The remainder were characterized as acute gastrointestinal
  illness of unknown etiology. Outbreaks in which ill persons had symptoms including
  abdominal cramps, nausea, vomiting, and diarrhe;a 24 to 48 hours after consumption
  of water and in which no specific etiologic agent could be determined were grouped in
  this entity. This includes a clinical entity known as "sewage poisoning" which is
  presumably caused by either coliform organisms or certain viruses that have yet to be
  fully characterized.                          ;                 '•
          TABLE 3. ETIOLOGY OFJVA.TERBORNE OUTBREAKS.
                                 -1971-1975 •!-
                                         OUTBREAKS   CASES OF ILLNESS
                                          .  i            -

  Acute Gastrointestinal Illness                 j   63             17,752
  Hepatitis-A                                :   14             '368
  Shigellosis                                 I   I4             2'803
  Giardiasis                                  !   13             5,136
  Chemical Poisoning                         1   12               511
  Typhoid                                   j  '  J               222
  Salmonellosis                              !    2                J/
  Enterotoxigenic E. coli                      I    1
  Total
123            27,829
                                                                       23

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


   The most commonly identified pathogen during this period was Shigella; there
 were 14 outbreaks and 2,803 cases of shigellosis:  Most of these occurred in non-
 municipal systems.
   Giardia lamblia is a flagellated protozoan responsible  for giardiasis. Clinical
 manifestations of Giardia infection can range from asymptomatic cyst passage to
 severe malabsorption. The mean duraction of illness is often 2 to 3 months. Only
 recently has G. lamblia been recognized as a cause of acute illness and as the etiologic
 agent in common-source outbreaks (3,2).  The 'disease has  received considerable
 attention since 1970, as a result of studies implicating tap water as the probable mode
 of transmission of giardiasis in travelers returning  from Leningrad. This increased
 awareness of physicians and public health workers has no doubt  been partially
 responsible for the increased reporting of this disease in .the U.S. Fifteen waterborne
 outbreaks  of giardiasis have been documented in  the U.S.  since the outbreak at
 Aspen, Colorado in  1965-1966. Thirteen occurred during the period 1971-1975,
 affecting 5,136 individuals.             .                                .
   Until the 1974 outbreak at Rome, New York, where 4,800 persons were affected,
 the waterborne outbreaks of giardiasis have generally  involved small municipal
 systems  or semi-public systems in recreational areas. Almost all of the outbreaks
 have occurred as the result of drinking untreated  surface water or surface water
 whose  only treatment  was disinfection.  Giardia cysts  are not  destroyed  by
 chlorination at dosages and contact times normally employed in water treatment, but
 it is felt  that they can be removed by well operated conventional treatment plants
 employing coagulation/flocculation, settling, and filtration.
   Although adequate disinfection data are not currently available, it is felt that
 Giardia cysts areas resistant to chlorination as cysts of E. histolytica. If this is correct,
 high concentrations of chlorine  and long contact  times are required in situations
 where settling and filtration cannot be employed; this may necessitate dechlorination
 prior to distribution (I). It has been reported that  G. lamblia cysts can survive in
 water and  remain infective for sixteen days (5). Rendtorff (6) conducted a series of
 studies with prisoner volunteers and found that £ of 22 men who received 10-25 G.
 lamblia cysts became infected and all 13 men who received from 100-1,000,000cysts
 became infected. Because an obvious source of gross human contamination could
 not be identified in most of the waters systems having giardiasis outbreaks, some wild
 and domestic animals are suspected as being important in the transmission of the
 disease to man. It is possible in water systems using surface water-whose only
 treatment is chlorination at conventional dosages and contact times that sufficient
 disinfection will be employed to produce a coliform-free water but not agiardia-free
 water.
   The were 14 outbreaks  of waterborne viral hepatitis affecting 368 people during
 1971-1975. There has been  .considerable controversy regarding the existence of
 viruses in treated water supplies and the possible health consequences. Hepatitis-A
 has been epidemiologically implicated in 66 waterborne outbreaks since 1946 and the
 data can be examined to determine how these viral outbreaks occurred.  Of the 22
 outbreaks occurring in municipal systems, three resulted from either inadequate or
 interrupted disinfection and five were related to the use of contaminated, untreated
 surface or  ground water. Half (eleven) of the outbreaks in  municipal systems
 occurred as the result of contamination of the distribution system, primarily through
 cross-connections and backsiphonage.  There  is  no evidence from waterborne
 outbreak data that hepatitis-A virus has been transmitted through water systems
 with properly operated  conventional treatment plants except where distribution
 system deficiencies have been found as the source of contamination.
   There  were  12 chemical  poisonings involving  fluoride,  chromate, selenium,
 phenol, furadan, arsenic, ethyl acrylate, a mixture  of lubricating oil and kerosene,
 cutting oil, fuel oil, and a herbicide.
   The four typhoid outbreaks affected 222 people and involved semi-public and
 individual water systems. The largest outbreak, responsible for 210 cases of typhoid,
                                             .':!.'
24

-------
                                        DISEASE TRANSMISSION/G.F. CRAUN
      TABLE 4. WATERBORNE DISEASE OUTBREAKS BY TYPE OF
                DEFICIENCY
                1971-1975

                                          OUTBREAKS CASES OF ILLNESS

.Untreated Surface  Water*                        19              5,729
Untreated Ground Water                         38              3,958
Treatment Deficiencies                      '     39             , 10,139
Distribution System Deficiencies                  15             • 7,468
Miscellaneous                                   12                535
Total                                          123             27,829
* Includes seven outbreaks of giardiasis in which surface water was chlorinated but
  not filtered                                                  ;


occurred in February and March 1973,  at the South Dade Migrant .Farm Labor
Camp, Dade County, Florida. This was the largest reported outbreak of typhoid
fever in the U.S. since 1939 (4). No deaths occurred, and the outbreak subsided with
few secondary cases and no transmission 1:o persons not connected with the camp.
  Two small outbreaks of salmonellosis occurred during this period: 34 cases of 5.
typhimurium in  a semi-public water system  and  3  cases of salmonellosis  in an
individual water system.
  At Crater  Lake National Park in 1975,.an estimated 1,000 cases of diarrhea
occurred following contamination of a spring, which served as the drinking water
source for the park, by overflow from an obstructed sewerage line. Enterotoxigemc
E. coli serotype 06:H16 were isolated from ill persons and from water samples.
  Untreated  or  inadequately treated water was responsible for the majority of
waterborne outbreaks during 1971-1975 (Table 4). Treatment deficiencies, primarily
inadequate or interruption of chlorination, and the use of untreated; contaminated
ground water accounted for 63% of all outbreaks and 51% of all illness.
  The outbreaks and cases of illness were also classified by cause and type oi water
system (Tables 5-6). As in the previous 25-year period, the major cause of outbreaks
in municipal water systems during 1971-1975 was contamination of the distribution
system- 41% of the outbreaks in municipal  water  systems occurred because of
deficiencies in the distribution of water. Generally, contamination of the distribution
system has  occurred  through  cross-connections and backsiphonage,  and  the
resulting  outbreaks have been  quite contained, affecting relatively few people.
However during 1971-1975, outbreaks caused by contamination of the distribution
system were responsible for 40% of the total cases of illness occurring in municipal
water systems.  Two large outbreaks accounted  for  most of the illness in  this
particular category: an estimated 5,000  cases  of acute gastroenteritis in Sewickley,
 Pennsylvania, which was felt to be related to contamination of an uncovered storage
 reservoir for treated water and 1,400 cases of a  similar illness in Sellersburg, Indiana,
which was traced to sewage contamination of a water main during construction.
   In the previous 25-year period, use of untreated ground water was responsible for
 most illness in municipal water systems. During 1971-1975, four outbreaks occurred
 in municipal systems because of the use of untreated ground water; however, only 74
 (0.4%) cases of illness resulted, because the outbreaks occurred in small water
 systems.  Nine '(24%)  outbreaks and 5,322 (29%)  cases of illness  occurring in
 municipal water systems were related to treatment deficiencies; six (16%) outbreaks
                                                                         25

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
      TABLE 5. MUNICIPAL WATER SYSTEMS DEFICIENCIES
                ASSOCIATED WITH WATERBORN^ OUTBREAKS.
                1971-1975
                                         OUTBREAKS  CASES OF ILLNESS
 Untreated Surface Water*                       6              5,404
 Untreated Ground Water                        4                74
 Treatment Deficiencies                          9              5,322
 Distribution Deficiencies                       15              7,468
 Miscellaneous                                  3               365
 Total  •                                       37             18,633
 * Includes four outbreaks of giardiasis in which surface water was chlorinated but
  not filtered
 and 5,404 (29%) cases of illness to use of untreated surface water. Four of the five
 giardiasis outbreaks occurring in municipal water systems were related to the use of
 surface water where disinfection was the only treatment. These were included in the
 "untreated surface water" category rather than the "treatment deficiency" category
 because specific treatment was not provided to remove or inactivate this pathogen.
 Conventional  disinfection  as the only treatment of surface water would not be
 expected to destroy Giardia cysts.
   Use of untreated ground water was responsible for most of the outbreaks and cases
 of illness that occurred in semi-public water systems in the past. Although this is still
 an important problem, the 1971-1975 data show that deficiencies in treatment were
 also responsible for many outbreaks and illness occurring in these water systems. Use
 of untreated ground water and treatment deficiencies accounted for 81%  of the

      TABLE 6. SEMI-PUBLIC WATER SYSTEMS DEFICIENCIES
                ASSOCIATED WITH WATERBORNE OUTBREAKS.
                1971-1975                  :'_...:
                                        OUTBREAKS  CASES OF ILLNESS
Untreated Surface Water*                       7               237
Untreated Ground Water                      27             3,859
Treatment Deficiencies                         30             4,817
Distribution Deficiencies                        0                 0
Miscellaneous                                  6               145
Total                                        70             9,058
* Includes three outbreaks of giardiasis in which surface water was chlorinated but
 not filtered

26

-------
                                        DISEASE TRANSMISSION/G.F. CRAUN


TABLE 7. SEASONAL DISTRIBUTION OF WATERBORNE OUTBREAKS.
                                 1971-1975


January
February
March
April
May
June
July
August
September
October
November
December
Unknown
*T^_4.~l
MUNICIPAL

1
2
3
2
1
5
4
5
4
3
4
2
1
«*•*
SEMI-PUBLIC
TOTAL AFFECTING VISITORS
2
1
2
4
9
14
12
12
4
3
3
4
^T/X
1
0
1
4
6
12
10
11
4
1
1
3
£/1
outbreaks and 96% of the cases of illness in semi-public systems. This is most likely
related to the size and nature of these systems, since most semi-public systems would
use ground water. Ground water is generally thought to be free of contamination and
little may be- done to,prevent intermittent contamination of these sources. Where
disinfection is provided, for either contaminated sources or as a preventative measure
for intermittent contamination, little  or nothing may  be done  to  insure that
continuous and effective disinfection is maintained. Many of these systems may be
operated for only a part of the year, and this may explain the lack of proper operation
and maintenance.
  A large number of waterborne outbreaks each year involves the travelling public
(Table 7).  In 1971-1975, 54 of the 70 outbreaks (77%) in semi-public water systems
affected travelers, campers, visitors to recreational areas  or restaurant  patrons.
Thirty-nine of the 54  outbreaks (72%)  involving a transient population occurred
during the months of May-August, the  period when outdoor activities such as
picnicking, camping and vacationing are most common. This implies that there is
either increased contamination of these water supplies during this period or, if it is
assumed that  the supplies  are always contaminated, use by greater numbers of
susceptible individuals during this period..Outbreaks in semi-public water systems
peaked during the summer months in 1971-1975,  which is consistent  with the
previous 25 years of record.  There appears to be little  seasonal variation for
outbreaks in municipal water systems, either during this period or the previous 25-
year period.
  Waterborne disease surveillance such as this is important for several reasons. First
and most important, it helps stimulate the recognition and investigation of ongoing
outbreaks and terminate them by identification of the water supply deficiencies that
allowed the outbreaks to occur. Second, it identifies deficiencies in water treatment
practices and leads to the formation of control measures. Regulatory agencies can
use the data for establishing water supply surveillance priorities based on outbreak-
related problems. For example, these data would indicate that additional emphasis
should be placed on surveillance  of smaller water supplies, particularly the semi-
                                                                        27

-------
MICROBIOL STAND Al IDS EVALUATION/C.W. HENDRICKS
public water supplies
that serve the travelling public. For these systems, increased
emphasis should be placed on protection of the source water from contamination,
disinfection of these water supplies, and proper operation of the systems to insure
that continuous, effecl ive disinfection is maintained. Outbreaks in municipal systems
would indicate that additional emphasis is needed in the area of cross-connection
control and prevention of distribution system contamination. Third, waterborne
disease surveillance can assist in the recognition of important or new etiologic agents
so that appropriate witer treatment can be provided. For example, the data indicate
that conventional  disinfection as the  only treatment for surface water sources is
ineffective in preventi ;ig waterborne transmission of giardiasis and  that to  protect
against transmission  of  this  disease all surface  water should  be treated by
sedimentation and filtration in addition to chlorination.
/I!, , - , ;: ^______ 	
TABLE 8. SUMM
SYSTE
WERE
OUTER
CAUSE OF OUTER
I. USE OF UN"
1. 1/100 ml,.2
*2. (-), 3/100 r
**3. (-),(-), 200
4. (-), 21 100 r
5. (-),(-),(-),
II. TREATMEN
!.(-),(-), 8/1
t2. 15/ 100 ml
tt3-<2,2MPN
ttK 12 samples
5. (-), 7200/1
f ft|6. 4 samples (
III. DISTRIBUT
1. (-)
2. (-)'s and lo
3. 44 samples
>39/100,<
IV. MISCELLAI
1. 121 100, 22
2. (-)
3.1/100,1/1'
Raw water data (coli
* 246/100
** 100/ 100
t 291 100 n
tt >16 MP
TTT >50/100
tfTt >50/100
28
\RY OF BACTERIOLOGICAL DATA FROM WATER
AS WHERE LOW NUMBERS OF TOTAL COLIFORMS
FOUND IN TAP WATER SAMPLES DURING THE
EAK INVESTIGATION.
! ' • i " :
EAK:
7REATED WATER
.1/100 ml, TNTC
il, 3/100 ml, 8/100 ml,' 19/ TOO ml, 32/100 ml
HOOml
il, 3/ 100 ml, 61 100 ml, 10/ 100 ml, 13/ 100 ml
162/100 ml
T DEFICIENCY
)0 ml, 20/1 00 ml
all(-)
)0 ml
-), 3/100, 10/100 ml
ION DEFICIENCY
w levels within DWS limits
(-), 5/ 100, 8/ 100, 91 100, 10/ 100, 10/100, 26/ 100, >39/ 100,
8/ 100, 521 100
IEOUS OR UNDETERMINED
'100
)0, 3/100, 4/100
,,,''' |j , | ,'], , , i n
brms) for above; not necessarily collected at the same time
nl
nl
.1
M . ' .
ml ,.
ml
i ,„ ::;, , , •„, ,.?; 	 j.,.r • |, ,: 	 ,,„ i ,„!.!,,„ ,,.;„ , ,(1,

-------
                                          DISEASE TRANSMISSION/G.F. CRAUN
    Historical water quality data from water systems having a
 can be used to determine the effectiveness of bacteriological su
 during 1971-1975 with sufficient bacteriological data were examined
 the  level of  contamination at. the  time of  the outbreak
 bacteriological surveillance could indicate if an outbreak woi|ld
    In  the majority of  instances  water samples collected
 investigation showed  high levels of either total or fecal colifi
 were outbreaks in which no or low numbers of coliforms we
 classfied according to the cause of the outbreak (Table 8). It
 these sample results are representative of the extent of conta
 occurred in the presence of low coliform contamination, or i
 actually larger and the samples were just not collected at the
   There were five outbreaks in which the drinking water
 implicated, but samples collected during the outbreak investiga
 for coliforms. One was caused by a cross-connection, one by
 distribution system, and another by contaminated ice. It is fe
 were most likely collected after the contamination had passed th
 are probably not representative of the water or ice which caused
 were, however,  representative of samples that might be cojllected
 bacteriological surveillance, and as such, the results would certa
 the remaining two  outbreaks with distribution samples neg
 contaminated water was distributed because of an interruption
 known only because the treatment deficiency was documented
 water was shown to be contaminated during the investigation
 samples were'collected at the time nf .the treatment failure,
          vaterborne outbreak
          veillance. Outbreaks
               todetermine(l)
           and (2)  if routine*1
             occur.
         during the  outbreak
         'c rms;. however, there
          e found. These were
          s difficult to judge if
          mination and illness
       if tlie contamination was
          right place or time.
             epidemiologically
          tion were all negative
          ontamination of the
          It that these samples
          •ough the system and
              the illness. They
                 for rountine
          nly be misleading. In
          ative for coliforms,
          of treatment. This is
            and; the untreated
           Unless distribution
      was
 OtHli L/JIWO YYWll< «^\_/il»^V> l,W»_l Clt-- VIIV- i.J.»ijy ii*""—ililflV ' 1 »-•"*.***»*** fc J.M.J.* v*«.yi^ •" ^-

 surveillance would not have warned of a pending outbreak
        ro atine bacteriological
 outBreaks where negative cotffbrm samples were collected at
 samples  wereHFOsMvF'To^^
       serve
           There were several
          th"e~sarne time other
             to ; illustrate  the
 importance of collectirfgTimely samples and collecting sample
 the systenTiranrouHneibacteriologlcal sampling program is
 timely  and^representative^sarripirng are a function oi the n
 samples required, and certainly a minimum of one or two samples
 quarter as permitted inJhe_NjaytjojiajJjalgjir|[LPrirnary Drinkin
 (7) cannot proggrly_assess_lhe quality of water being consumed.
"it required"two or more samples to find the contamination
 samples collected were  negative for coliforms. Unless samp
 continuously at a large number of points throughout the system, it is unlikely that
          s at proper points in
            be effective^ Both
          mimum number  of
       to
             per month or per
           Water Regulations—.
          n several outbreaks,
         a|nd in one, 44 of 54
          escanbeccjlseted
 routine bacteriological sU£V6iltaiiC.e will bj? userul in preventing the occurrence of
"outbreaks. Since it is technically and financially mfeasible to
 bacteriological monitoring, alternatives must be considered if
 prevented.  Monitoring for chlorine residual. is._a reasonabl
 especially important for small systems where relatively few
 are required.
   Outbreaks  having routine bacteriological surveillance dat
 determine if the occurrence of the outbreak could be
 These were categorized by the  cause  of the outbreak to illu
 (Tables 9-12).
   Previous bacteriological results were not available for those
 distribution deficiencies,  but several  observations can  be
 bacteriological surveillance for prevention of these outbreaks.
 examples of chemical cross-connections that have not resulted
 because the water had a distinct color or odor not because of any
 Bacteriologically contaminated water would not necessarily exh
 istics, and it is highly unlikely that samples could be collected f
 detect cross-connections and prevent  outbreaks due to this caus<
 occurred  during  1971-1975 where bacteriological
          provide continuous
          outbreaks are to be
           alternative and  is
       bacteriological samples
           were; examined to
predicted for that water supply.
          trate  several points

          mtbreaks caused by
        n)iade  about routine
          There are numerous
          n illness, but this is
          outine surveillance.
          ibit these character-
          :equently enough to
          :. Several outbreaks
                entered  the
contamnation
                                                                          29

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRlCKS
                                           '.. '      i "  '                     i ", '

   TABLE 9 TWO OUTBREAKS CAUSED BY USE OF
             UNTREATED, CONTAMINATED WATER.

                  Outbreak in Up'state New York Restaurant
                                    1971
                                                 : '  I                    '  : .'
 Acute Gastrointestinal Illness
 Overall attack rate 500-1000/581-1163 (86%)

   Tap water samples collected in restaurant before and after outbreak were negative
 for coliforms except for one sample collected on 1 M9; 71  which showed 200 total
 coliforms per 100 ml. Spring on several occasions prior to the outbreak yielded 100
 coliforms per 100 ml.

                          Big Sky Montana Outbreak
                                    1975

 Acute Gastrointestinal Illness (Yersinia enterocolitica was  isolated from the water.
   but stools were not examined for this  organism)
 Overall attack rate 139/ 314 (44%)

   It was reported that between 3/72 and 11 /18/74 samples were collected every 1-2
 months and that all had been  negative for coliforms. Water samples obtained on
  1/14-16/75  yielded  from  1 coliform  to  >  16 coliforms/200 ml.  During the
 investigation of the outbreak samples from the two wells  showed:
                                                    I          '            •:
    1. 4 samples-Negative for coliforms    	
    2. 8 samples - Positive for coliforms

 4 - 1/250 ml
 2-2/300 ml
  1 - 2/300 ml
  1 - 7/250 ml
                                                  , . il

distribution system and routine bacteriological sampling showed positive results.
This did not prevent an outbreak from  occurring because of the length of time
between sample collection and analysis and subsequent recognition that the positive
result represented a real problem. It is often assumed that a positive result is due to
accidental contamination in the system. The Interim Standards (7) require check
samples to confirm positive bacteriologic results and require State notification within
48 hours. It is unclear whether the 48 hours applies to the results of the check sample
or the initial positive-sample. Public notification can range from 7 days by radio and
TV to the time of the next water bill. If the coliform standard is exceeded,  by the time
check samples are collected for confirmation, and the maximum time is  utilized for
notification, it will be too late to take action to prevent an outbreak from occurring. If
an outbreak does occur, the only consolation will be that sample results are available
to assist the epidemiologist in his investigation.  However, with chlorine residual
monitoring, some action could be taken immediately by the operator after discovery
of a low or zero residual. It is unlikely the operator would assume that  he had just
made a sampling error when he measures chlorine residual.
   Surface  water should never be  used without treatment, but  there are smaller
communities and semi-public systems where  this is still done.  Bacteriological
surveillance of these distribution systems can only provide a false sense  of security.
There can  never be assurance that contamination will not  occur at any given time
even though the bacteriologic results have been negative. Systems using surface water
                                                 . •  i     ..               .• I ;
 30

-------
                                         DISEASE TRANSMISSION/G.F. CRAUN


                TABLE 10. STEELE ALABAMA OUTBREAK.
                                     1973


 Overall Attack Rate 50/100 (50%)
 Infectious Hepatitis

                        Bacteriological Qualtiy of Distribution Systems

           Date                   Samples                  Results
           2/72                       1                      2/JOO
         2/72-3/73           ,        30                       —

 Raw water samples 2/72-2/73 15 samples
         negative for coliforms; 7200 coliforms/100 ml found in late 3/73 during
         outbreak investigation
 and providing conventional chlorination as the only treatment are not uncommon in
 some parts of the country. Bacteriological surveillance  in these systems  is also
 meaningless, for conventional chlorination can provide a water free of coliforms and
 yet not be effective in destroying Giardia cysts if they are present.  ',
   In  many instances,  ground  water free of contamination  is  used without
 disinfection. Proper construction and frequent sanitary  surveys are required to
 prevent contamination of these sources. Two examples of outbreaks caused by use of
 undisinfected ground water illustrate this point (Table 9). The outbreak in the New
 York State restaurant occurred on November 13-14, 1971, but tap water samples
 revealed no contamination untirNovember 19. Although source water samples prior
 to the outbreak revealed coliform contamination, the only treatment provided was
 softening.  A  sanitary  survey  prior  to the  outbreak would have  noted the
 contaminated source and lack of proper treatment. The Big Sky, Montana outbreak
 (Table 9)  illustrates intermittent contamination of ground water felt to be of good
 quality and relatively low numbers of coliforms can be associated with illness. The
 water supply  was epidemiologically  implicated and chlorination  of  the  source
 stopped the outbreak. The previous bacteriological  history had not predicted the
 outbreak which occurred in December, 1974-January 1975, for all samples prior to
 that time  were  negative for coliforms.  Had  a complete sanitary survey been
 performed perhaps the sewage lines passing near the wells would have been noted as a
 potential hazard as they were; during the outbreak investigation.
  The remaining outbreaks to be discussed were all caused by treatment deficiencies,
 and show that routine  bacteriological surveillance  is of little use in  preventing
 outbreaks due to this cause (Tables 10-12). Frequent sanitary surveys and increased
 emphasis  on maintaining continuous and adequate  disinfection are important in
 preventing outbreaks such as these. Routine bacteriological surveillance data from
 the Steele, Alabama infectious hepatitis outbreak of February 21-March, 1973
 (Table 10) showed relatively little contamination either in the distribution system or
 the wells. The one positive distribution sample of 2 coliforms per 100 ml would most
 likely not  have caused any concern, as many of us would  dismiss this as a sample
 collector error. Contamination of the limestone aquifer and inadequate disinfection
 caused the outbreak. Routine bacteriological surveillance was of little help in
 preventing the outbreak. Monitoring  of chlorine residuals, however, might have
alerted  operators to  adjust the  level  of  chlorination  during  the  period of
contamination.

                                                                        31

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


                 TABLE 11. OUTBREAK IN A BOY'S CAMP.
                                     1975           '
                               t
      Acute Gastrointestinal Illness—3 separate outbreaks
      Overall attack rates of 12/73 (16%); 85/160 (53%); 67/100 (67%)

                             Bacteriological Quality of Distribution System
                              F   •   • .     '         i         ••
            '    Date   .       |         Samples                  Results

               5/27/75        [.           2         "             ~
               6/10/75        !            3                       —
               6/24/75        '            3                       —
              .7/15/75                    3                       -
               7/29/75                    2
               8/26/75                    I           ,          ./Too
                                                    i
        Raw Water quality 5/75:8/75: 2 samples negative for coliforms
   The outbreak at the Boys Camp  (Table  11) is an  excellent example  of how
 meaningless the results of a small number of bacteriological samples are in terms of
 Sutine surveillance. The bacteriological record for 1975 would certainly not indicate
 a problem with this supply, yet 3 outbreaks were *^™^*"™*^%£°*£
 time  in which the water supply ;was  epidemiologically implicated  Samples were
 collected every month at approximately 2 week intervals, and the outbreaks occurred
 on June 6, July 18, August 24, all between periods of sampling. The 3 wells used as a
 water source were chlorinated, but it was reported that the operator did not monitor
 for chlorine residuals. He relied  instead  on bacteriological results.  Additional
 information on this system was obtained in 1976 when another outbreak occurred. A
 thorough engineering evaluation showed one of the wells to be contaminated with 6
 to TNTC coliforms per  100 ml  and inadequate  control  of chlonnation.  me
 contamination that-a routine bacferiological surveillance program failed to uncover
 was finally found. The outbreaks most likely would, have been prevented with a
 thorough sanitary survey noting geological conditions  of the well, etc.  and
 monitoring of chlorine residuals to assist 5n proper maintenance of disinfection.
   the last example, an outbreak: at Richmond Heights, Florida, (Table 12) is one
 with which I am personally familiar. Again routine bacteriological sampling oi the
 distribution system showed no coliform contamination prior to the outbreak. On
 January 14-15,  1974, chlorination was interrupted and untreated well water was
 distributed  Routine surveillance noted 3  coliforms per 100 ml in a sample from the
 system on February  12. The samples  collected on February 18 appear to be check
 samples  but this is not certain. One sample on this date showed 10 coliforms per ,100
 ml The 19 negative samples on February 21 appear to be check samples also. During
 the period''of time that check samples were being collected, the community was
 experiencing an outbreak which affected 1,200 people. The outbreak investigation in
 March revealed that one of two'wells was continuously contaminated by effluent
 from a septic tank near the well.' A sanitary survey and sampling of the raw water
 could have detected these deficiencies  prior to the outbreak and a greater emphasis
 could have  been placed on continuous chlorination and monitoring of chlorine
 residuals.
 32

-------
                                         DISEASE TRANSMISSION/G.F. CRAUN
                   Table 12. Richmond Heights Florida outbreak.
                                      1974
   Acute Gastrointestinal Illness Compatible With Shigellosis
     (10 Culture Proven Cases)
   Overall Attack Rate 1200/.6500 (18%)       ^
   Date

   I/  9/74
   1/10/74
   2/12/74
   2/12/74
   2/18/74
   2/18/74
   2/21/74
   3/11/74
   3/26/74
   3/28/74
   3/29/74
  •3/31/74
Bacteriological Quality of Distribution System
                         i
                  Samples                  Results
                     9
                     2
                     1
                     1
                     1
                     2
                     19
                     9
                     2
                     6
                     3
                     1
3/ 100 ml coliform
10/100 ml coliform
  SPG - 3/26-31/-74 -'Average ISO/ml
  Raw Water - 3/27^4/74~~Total coliform
                             Fecal coliform

  * House with Water Softner
                                                       >50/100 ml coliform*
                                                          3/100 ml fecal*
                     >50/100 ml
                         to>50/100ml
 Conclusions                               '
                                            i
   Waterborne outbreaks continue to occur and vye should make use of these data to
 plan and develop surveillance programs. Water supply deficiencies that allowed the
 outbreaks to occur can be identified and the formulation of control measures should
 be based on outbreak related problems. Information on etiologic agents can be used
 to determine treatment  requirements.  Historical water quality data from systems
 having an outbreak can be used to determine the effectiveness of bacteriological
 surveillance.                               . i
   The data indicate that increased emphasis bd placed on surveillance of smaller
 water systems,  especially those serving the travelling public. The formulation of
 control  programs should include increased emphasis on (a) providing adequate
 treatment of surface water sources to prevent outbreaks of giardiasis, (b) assessing
 the potential for contamination of ground water/sources and providing appropriate
 treatment, (c) maintaining continuous disinfectipn, and (d) preventing  contamina-
tion from entering the distribution system.     ]
  Thorough clinical,  epidemiologic  and  laboratory   investigations must be
conducted to identify possible new etiologic agents responsible for outbreaks of acute
                                                                        33

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS

gastrointestinal illnesses of unknown etiology. Recognition of outbreaks and timely
investigation are necessary for collection and analysis of appropriate water samples
and clinical specimens in this regard.
  Routine bactgrj^^L*"'-vei11ance of water distnbution^stems for cohiorms,
                                                                   purposes
                                                               aks. Results ot
     roTEttle valrift in preventin^tlieoccurrenceofwaterborne
                                                         ou
                                                         rbetaEinTTFe length
 such sampling cannotje?ecervecLintimefor.corrective action.tQ_be_
-oTTmTe~T£—d^b^^
 i^irs-irrruicTn^o-ioii^^
               iou IUUK iu cuamt. my. v^...*^." - — -- ------ -• - - ;   •—*-.„  -
               s. The idea^fcheck sarnpleslboadds to the time delay, since the iirst
                            assumed t.o be a sampling error  rather than rea

contamination. It i
systems. The ^
momtSnri that provides-anlnaication ot
^g_-^r~amiT5Tr^^^
bacteriological results are misleading and meaningless.
   I propose, instead, an alternative monitoring program that would allow our
limited microbiological resources  to concentrate on research activities, such  as
identification of etiologic agents responsible for outbreaks of gastrointestinal illness
of unknown etiology. The emphasis  of this alternative program would be on
engineering evaluations  or sanitary surveys. A  thorough engineering evaluation
would include bacteriological monitoring of source water rather than water in the
distribution system. It is necessary to know the quality of source water and Potential
sources of contamination so that appropriate treatment can be recommended. The
emphasis then must be on providing and  maintaining this treatment, especially
continuous disinfection at adequate concentrations. For product quality control, a
simple test such as chlorine residual can be used. Of course, chlorine residual must be
defined for each system so that all factors of effective disinfection are included, such
as time, pH, temperature, species of chlorine, turbidity, and representative sampling
points and frequency of collection  must be determined. Bacteriological surveillance
of the distribution system should be used to check the validity of chlorine residual
monitoring, but this can be done at a lower frequency. This would also provide a
historical record which many seem to feel is important. If the primary disinfection oi
choice ceases to be chlorine,' residuals or markers of alternative disinfectants would
be employed, and the development of these disinfectants should include technology
to measure residuals. If supplies are not disinfected, chlorine residual measurements
would, of course, not work; however, it is  anticipated that a minimum number of
such supplies would exist.
   The real  importance  of measuring chlorine residual  is  that  the  primary
responsibility for taking corrective action is placed where it should be— with the
operator. Time is of the essence in preventing an outbreak from occurring and the
operator must.be the first to know if something is amiss, not the last to know. He can
provide the quickest and simplest response simply by adjusting the chlormator.
                             REFERENCES
                                    •'       '',,'   ''T " !       ii   '       ,          :
 1  Chang, S.L. and G.B. Fair. 1941. Viability and  destruction-of cysts of E. hystolytica.
  ' JAWWA 33:1705-1715.                                   .              .
 2. Craun, G.F., L.J. McCabe and J.M. Hughes. 1976. Waterborne disease outbreaks in the
   U S.—'1971-1974. JAWWA 55:420-424.
 3. Craun, G.F., and L.J. McCabe. 1973. Review of the causes of waterborne-disease outbreaks.
   JAWWA 65:74-84.
 4. Craun, G.F. 1974. Microbiology—Waterborne outbreaks. J. Water Pollut. Control t-ed.
   4(5:1384-1395.
 34

-------
                                            DISEASE TRANSMISSION/G.F. CRAUN
5. Rendtorff, R.C. and D.J. Holt. 1954. The experimental transmission of human intestinal
   protozoan parasites. IV. Attempts to transmit Endamoeba coli and Giardia lamblia cysts by
   water. Am. J. Hyg. 60:327-33%.                                   '
6. Rendtorff, R.C. 1954. The  experimental transmission  of human intestinal protozoan
   parasites. II. Giardia lamblia cysts given in capsules. Am. J. Hyg. 59:209-220.
7. United States Environmental Protection Agency. 1975. National interim primary drinking
   water regulations. Fed. Register 40:59566 -59588.
                                                                              35

-------
-  =y  ?   f-- i===

-------
                                 ALTERNATIVE INDICATORS/G.A. MCFETERS


  Alternative Indicators of Water Contamination and Some Physiological
            Characteristics of HeteroltropMc Bacteria in Water

       Gordon A. McFeiers, John E. Schillinger, and David G. Stuart
                       Department of Microbiology
                         Montana State University
                        Bozeman, Montana  59715

Introduction

   The purpose  of this presentation is to list some of the most promising
microorganisms and procedures  that have been suggested 4o  indicate water
contamination of public health significance. In addition, studies dealing with
the comparative die-off kinetics of selected indicator bacteria  and pathogens
will  be discussed, followed by a brief consideration  of  reported bacterial
growth in  certain aquatic environments.  These subjects will not be treated
exhaustively with- a complete listing of literature citations  at the end  of this
paper.
   In  discussing  the question of alternative indicator systems, it is important
to list certain criteria that  are appropriate  in judging their efficacy. Some may
seem  idealistic,  but they  are presented within the same altruistic frame  of
reference that is basic to  the indicator concept;  that is  safe drinking water.
The criteria follow:                          .
   A.  An indicator should be applicable  to .all types of water that may  be
      investigated.
   B.  If a  microorganism, it should be present in greater  numbers than the
      pathogen  in all cases when the  Jlatter. is: found.- Alternatively, positive
      indication should be present if -there is-a significant health hazard,
  C-  Any  indicator microorganism should not increase significantly  in the
      absence of a health hazard.
  D.  Indicator microorganisms should be more resistant to the physiological
      stress within aquatic environments and to  the action of disinfectants
      that  are commonly used therein. As  a result,  the  indicators should
      exhibit greater survival than pathogens under the "real world" condition
      of the  water that is to be  tested.
  E.  The  indicator reaction or test data should be unique and characteristic
      of that microorganism or determination.
  F.  The indicator methodology  should! be of minimal complexity,  rapid and
      inexpensive.                 " "'.....'".'""'....""_";„::_".
  G.  Indicator  mircoorganisms  should!  be harmless  to man  under  usual
      conditions.
  H.  The  indicator or test should be proportional to the health hazard that
      is present-


Indicator Microorganisms That Have Been Suggested

   In  this discussion we have arranged the more commonly used indicator
microorganisms first, with others  to follows.                   ;
   A.  Coliforms, fecal and E. coli.
      Since these organisms have already been discussed at this symposium,
      we have chosen not to emphasize them in this presentation except  to
      mention that some of these microorganisms satisfy most of our criteria

                                                                       37

-------
   MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                  	,	I	

       under a wide range of circumstances. They have also served us rather
       well for many years (10). However, on occasion they have not been
       detected when pathogens have been present, as in the case of the River-
       side, California, outbreak of 1,965, and some authors (4) have attached
       their efficacy, in part, on the'basis of their occasional  non-fecal origin
       in nature.
   B.  Fecal streptococci,           *                     '
       This group bacteria is frequently used as indicators  of sanitary sig-
       nificance-  However,  it must ;be acknowledged that some  species are
       capable of reproduction  exclusive  of warm-blooded animals. This was
       documented in two papers by;Mundt in 1963 (15, 16). However, others
       (7,  18) suggest that fec.al streptococci  provide  valuable supplemental
       data in determining fecal water contamination  if conducted  with, the
       fecal coliform test. Such a rriultibacterial detection scheme could em-
       body the fecal coliform/fecal stretptococcus ratio (8).
   C.  Standard plate count.
       Although  it is virtually  impossible to enumerate the  total population
       of heterotrophic bacteria within a water sample by a single test, the
       standard plate  count  may  provide useful water quality  data,  par-
       ticularly if it accompanies othbr complimentary bacterial measurements.
   V   The rationale for this  lies jin the  observation  that  a  great many
~""~-\ potential or secondary pathogenic invaders that are common to water
  C/^ may be detected in this way. As a result, the overall water  quality may
       be more accurately assessed in addition to the detection of pathogenic
       bacteria that are known  to reproduce in water.  Geldreich  (9) has sup-
       ported this contention by  adding that a standard plate  count limit
       may also reduce  interference in  counting low  numbers  of coliform
       bacteria.                      ...........
   D.  Clostridium perfringens
       The use of this bacterium as an indicator organism has been advocated
       by some European workers; most notably, Gunnar Bonde (2). Its spore-
       forming charactertistic has made it desirable as an  indicator in the
       presence of disinfectants or when water sample transportation time is
       protracted.  In addition,  its 'presence has been correlated with  high
       numbers of classical indicators (2). However, since many authors regard
       this organism ubiquitous in  nature, and it is a strict anaerobe, its wide-
       spread usefulness  as a sanitary indicator can be questioned.
   E.  Klebsietta pneumoniae.       \
       This bacterium  has been employed  as an indicator  of sanitary sig-
       nificance in Britain and water isolates were recently  demonstrated  as
       indistinguishable from clinickl isolates of this opportunistic  pathogen
       (20). It is not considered by! most to  be ubiquitous in nature since it
       is usually associated with E. \coli,  although it is found in carbohydrate
       rich wastes and on vegetables and seeds. In addition, it was found in
       the chlorinated  drinking water of Chicago  as  a fecal  coliform  (19).
       It has been pointed out that this organism is capable of proliferating
       in waters  under some environmental conditions  (14). This widely held
       view casts some doubt that  Klebsiella  species may independently serve
       as definitive indicators although it may be frequently enumerated as a
       total or fecal coliform.
   F.  Aeromonas species.
       Several species of Aeromonas are pathogenic to a variety of aquatic
       fauna and man. These microorganisms are widely distributed in nature,
    38

-------
                                  ALTERNATIVE INDICATORS/G.A. MCFETERS


        they are considered ubiquitous  by some authors (6) and are capable
        of prolonged survival in many aquatic environments. Because of these
        qualities, we regard Aeromonas species of limited  value as  indicator
        organisms in potable water.
    G.  Pseudomonas species.
        These opportunistic pathogens represent particular problems in stored
        water and some medical procedures!using water. They are ubiquitous
        in nature  and may  actively reproduce in potable and distilled water to
        populations  as high as 10 Vml (3). | Therefore, these microorganisms
        are a significant health hazard,  especially to the hospitalized patient.
        They may also persist in  such water for extended periods  of time.
        Because of this capability, thesie bacteria are frequently found in  the
        absence of coliforms;  one study reported that  one-half of the water
        samples containing  Pseudomonas aeruginosa lacked coliform bacteria
        (17).                               |                 :
    H.  Pathogenic microorganisms.         j
        Some pathogenic  microorganisms  like  Salmonella species occur in
        contaminated water in a statistically reproducible manner if the water
        is  grossly contaminated. However, under conditions where the con-
        tamination is  present in  a lower  level,  such as potable  water,  the
        occurrence of pathogens is less  predictable. No single'pathogen suffi-
        ciently satisfies the  criteria of an effective indicator organism-  There-
        fore, until one such pathogen is identified and appropriate detection
        methods are developed, we feel that  the  pathogenic microorganisms
        have many  shortcomings  as  indicators.  At some time in the future,
        developments may  allow a pathogen like one of  the enteric viruses to
        serve as such an organism.          i
. ...  1.  .Others,	.,.	_...,,..,.,.. .,_,.i.,.,„.,....,. ...,_...._.  . . 	
       Five additional'microorganisms will be mentioned. However, the present
       level  of  information on each regarding  distribution  and population
       dynamics in  water is inadequate  to  make a judgment as to their suit-
       ability as indicators. They arejCaryophanon latum,  the   Bifidobac-
       terium,  bacteroides, coliphage and certain  acid-fast bacterial      ~   ~
                                           i
 Other Indicator Procedures:  Rapid Tests ;

   The promise of rapid alternatives or supplementary indicator  systems is a
 highly desirous yet elusive  goal of water microbiology. This realization awaits
 the resolution of many theoretical ami practical problems  through research
 and testing  of  procedures  such as the ATP  biometer, liberation of  14CO2
 from radio-labeled lactose  and the Limulus 'lysate test.
   The Limulus lysate test  for bacterial endbtoxin warrants further mention.
 There is a growing body of  evidence that this test could be the most promising
 of the rapid indicator procedures. Studies that were done by Tom Evans  in
 our laboratories (5) illustrate  the relation o|f this test to some conventional
 bacterial  water  quality criteria. The results seen in Figure 1 are typical  of
 our findings  and suggest that the Limulus lysate assay for endotoxin parallels
 results of some conventional indicator bacterial tests in natural waters ranging
 in quality from  pristine streams to the discharge of a sewage treatment plant.
 Answers to questions such  as the  longevity of  endotoxin under conditions  of
 water  renovation and  disinfection along witli further refinement of the assay
 will bring this procedure within the scope of practical usefulness  in the near
 future.                                   i

                                                                        39

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
        10^
 s
 cc
 Ul
 o
 §

        10'
                                              --» CONCENTRATION Of ENDOTOXIN
                                                 NUMBER.OF HETEROTROPHIC
                                                 BACTERIA/ml
                                                 NUMBER OF TOTAL COLIFOHM
                                                 BACTERIA/100ml
                                                 NUMBER OF GRAM . NEGATIVE
                                                 BACTERIA/ml
                                                 NUMBER OF ENTERIC
                                                 'BACTERIA/mi
                                   JL
-L
                                              _L
                                                                        10°
                       EFI    HB   EG4   OF2   EG5   EG5a
                                      SITE
             Figure 1 BACTERIAL AND ENDOTOXIN PROFILE OF THE EAST GALLATIN RIVER
              3     DRAINAGE FOR SAMPLING DATES 7/11/75 AND 7/15 7Ei EACHI POINT
                   REPRESENTS THE GEOMETRIC MEAN OF FIVE REPLICATE SAMPLES
                   ENUMERATED BY THE SPREAD PLATE TECHNIQUE.
 Factors Influencing the Recovery and Enumeration oi Indicator Bacteria

    An understanding of the population  dynamics and physiological injury of
 heterotrophic bacteria in water is basic to the realistic  application of the
 bacterial  indicator concept. This assertion  is based  on the observation that
 the aqueous environment, from which the indicator bacteria are recovered is,
 at most, an extremely dilute solution  of bacterial nutrients. As a resuh,  when
 bacteria are released into the water from their nutrient-rich natural habitat,
 thev eo through a spectrum of/debilitation during which enumeration becomes
 progressively more difficult until they finally die. On the other hand, there is
 evidence that some heterotrophic indicator bacteria metabolize and even grow
 in some  aquatic environments that are oligotrophic by all other standards.
                                                                              cc
                                                                              la
                                      cc
                                      ui
                                      ca
                                      13
  40

-------
                                 ALTERNATIVE INDICATORS/G.A. MCFETERS


 An overview of these considerations will be given within the context of studies
 that have been carried out in our laboratories-
    In studying the physiological  behavior and the population! dynamics  of
 heterotrophic microorganisms in water, methodology can be very important.
 For example, coliform bacteria.are capable of cryptic growth,; using the or-
 ganic chemicals liberated  from dead cells that have undergone  autolysis, if
 the population is on the order of 106/iml in an enclosed container. For that
 reason we developed the membrane  diffusion  chamber  (Figure  2) to allow
 the retention of a population of microorganisms within an enclosure  that is
 immersed in a body of [ water (Figure 3) (12). Water is free to diffuse through
 the enclosure, carrying remnants of dead cells away and  allowing a. more
 realistic experimental approximation of what is likely to occur  in the natural
 aquatic environment. In addition, we found that the population of cells must
 not  be excessive  (Le.  less than lOVml)  in the  performance  of  these
 experiments.                     .                         ...
   Studies  were carried out using these chambers to investigate the  compara-
 tive  survival of classical indicator bacteria  and some • waterborne pathogens
 in water (13). It was demonstrated that coliforms  and enterococci  persisted
 E
 o
         2
                       2       4
10
                                   cim
Figure 2. SCHEMATIC DRAWING OF THE MEMBRANE DIFFUSION CHAMBER.

                                                                     41

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
Figure 3. PHOTOGRAPH OF MEMBRANE DIFFUSION
        CHAMBERS IN OPERATION SUBMERGED IN
        WATER.
                                    i
42

-------
                                  ALTEELNATWE INDICATORS/G.A. MCFETERS


 as well as or better than the pathogens that were studied (Table \ 1). Also, the
 fecal coliform to fecal streptococcus  ratio of human  fecal wastes decreased
 while  that from a bovine  source increased with time in water to  a point
 where the ratios were not useful  in distinguishing the  origin of the pollution
 within 24 to 48  hours. In another study that was carried  out in our labora-
 tories, largely by Gary Bissonnette, the physiological debilitation of indicator
 bacteria in water was examined (1). This  work demonstrated that indicator
 bacteria progressively lose the ability to grow and form colonies^ on selective
 media as they age  in natural water (Figure 4)  and that  an  appropriate
 nutrient environment and moderate temperatures  allow the repair  of  such
 injury within three hours (Figure 5).  These studies indicated that this occur-
 rence is  widespread among indicator bacteria in nature and that  injured cells
 may comprise as much as 90 percent of the population.
   An  ongoing investigation describing  the  interaction  between indicator
 bacteria  and algae in the high alpine zone also relates to the broad question
 of indicator bacterial behavior in water  to  the collection  and interpreta-
 tion  of. meaningful bacterial  water  quality data.  Relatively high numbers of
 total coliform bacteria in a mountain stream have been  shown to be associated
 with an algal mat community. In laboratory studies, indicator bacteria includ-
 ing total coliforms,  fecal coliforms and  members  of the genus Klebsietta
TABLE  1.  COMPARATIVE DIE-OFF  RATES  (HALF-TIME)*  OF
    FECAL INDICATOR BACTERIA AND ENTERIC PATHOGENS.
    .__'_'-	-••-"•" •••.•'•— "'.    .-'" - '"_"_"___'_	.   Bacteria         No. of
  "^Bacteria       ••--•--                         Half-time      I  strains
                                                   (h)           analyzed

Indicator bacteria
   Coliform bacteria (avg) 	       17.0            29
   Enterococci (avg)  	      22.0            20
   Coliform from raw sewage	       17.5
   Streptococci from raw sewage 	       19.5
,   Streptococcus equinus 	       10.0         :    1
   S, bovis	-       4.3             1

Pathogenic  bacteria
   Shigella dysenteriae '."!."."~.'.'.V.".'".'l	]'.'	.....       22A*            1
   S, sonnei 	       24.56            1
   S. flexneri	       26.86        .    1
   Salmonella  enteritidis ser. paratyphi A ........       16.06            1
   S. enteritidis ser. paratyphi D	       19.26            1
   S, enteritidis ser. typhimuriuni	       16.06        ',    1
   S,  typhi	       6.0             2
   Vibrio choleras	       7.2             3
   S. enteritidis ser. paratyphi B	       2.4             1

   "The  half-time  was determined grapbicsilly  from Fig. 2 and 4 as the time
required for a 50%  reduction in the initial population.
   6 The half-time was determined graphically from Fig. 4  as the time required for
a 50% reduction in  the population at 24 h.

                                                                        43

-------
  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                     DAYS
Figure 4. COMPARATIVE RECOVERY FOR E.COLI C320MP25 IN MEMBRANE
         FILTER CHAMBERS LOCATED AT SITES BR2 AND EG5 DURING A
         4- DAY EXPOSURE PERIOD. SAMPLES WERE SURFACE : OVERLAY
         PLATED USING TSY AND KF AGARl
  increased by two to three orders of magnitude at low temperatures using the
  products of algae from the mat community. The likelihood that these bacteria
  were reproducing this way in nature was further reinforced by the finding that
  they metabolized radioactive algal organic extracellular products in controlled
  laboratory experiments.  These findings  indicate  that  under some  circum-
  stances heterotrophic bacteria  are capable of growth in a high quality water.
  This was previously  suggested by  Hendricks  and Morrison  (11) who pro-
  posed that stream sediments adsorb organic substances  that are subsequently
  used by bacteria. The implications of these studies go far beyond the alpine
  streams in which they were done. For instance, surfaces within water storage,
  treatment  and distribution  facilities may adsorb and  concentrate potential
  44
                                               	,;	|	„

-------
                                ALTERNATIVE INDICATORS/G.A. MCFETERS
                          2      3~4
                       HOURS (in TSY broth)
Figure 5.  REPAIR OF INJURY IN TSY BROTH FOR E. COLI C320MP25 CELLS
         HAVING BEEN EXPOSED TO THE STREAM ENVIRONMENT OF SITE EG6
         FOR 2 DAYS. CONTROL OR O - TliME CELLS AND 2 - DAY EXPOSED
         CELLS WERE ENUMERATED OVER A 6 - HOUR GROWTH PERIOD IN TSY
         BROTH USING TSY AND DLA AGAR SURFACE - OVERLAY PLATES.
bacterial nutrients and provide favorable conditions for bacterial growth both
before and after disinfection. Activated carbon filters are an example of an
increasingly popular water treatment process that  is particularly susceptible
to this occurrence. With that in mind, the proliferation of bacterial indicators
or potential pathogens in potable  waters must  be considered  in examining
alternatives for determining water  quality. The growth of indicator bacteria
might  give a false signal of water contamination,  but multiplication of or-
ganisms such as Flavobacterium, Pseudomonas, and Klebsiella species, creates
a dangerous condition for  some consumers  that must  be recognized if the
test procedures are to serve our needs. With the  presently used bacterial in-
dicator surveillance protocol, this situation probably does exist in many finished
waters that are defined as potable.

                                                                     45

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                   I
                                                   I
Philosophical Basis of Indicator Bacteria and Recommendations

   A discussion  of "the  philosophical basis of the  indicator bacterial  concept
is essential in discussing alternative indicator protocols for determining water
quality. The philosophical basis that we consider fundamental to the indicator
concept -is that an indicator system that is more encompassing is better than
one that is more restrictive, in terms of the source of the contamination that
will be detected. Additionally, false negatives are to be  avoided more than
false positives if one has to select the lesser of the  two evils. As a justification
for this position, we cite the distinction between  sewage contamination and
feces  contamination and make .the generalization that  an effective indicator
system will detect not only feces but also sewage  contamination. The impor-
tance  of  this notion  is that water contamination from  sewage rnay  vary
widely in amounts  of fecal pollution and, more importantly, that  the  broad
spectrum of  health threatening chemicals and infectious agents that  may  be
transmitted through water are not exclusively associated with fecal material.
   Translated into choices that have been presented, E. coli is most restrictive
since it is most  frequently associated with feces followed  by fecal coliforms
that are usually  linked with fecal pollution, and then total coliforms that may
or may not be  associated with  fecal material. However, total coliforms are
a good indicator of sewage or the intrusion of surface or contaminated ground
water  if found  in finished drinking water. As another advantage, the total
coliforms are more  easily and economically detected with less chance  of
false  negatives than  fecal coliforms and E.  coli.  Also, the maximum con-
taminant level (MCL) for total coliform bacteria might be increased somewhat
because of the greater sensitivity in detecting a wide spectrum of contaminant
sources plus laboratory contamination with: this  group  of organisms. This
would help to satisfy some of the objections to the National Interim Drinking
Water Standards that have been voiced in many  quarters recently.
   Using such an indicator system or even E. coli, as in Britain, there is still
legitimate cause for concern regarding large  numbers of potential pathogens
and other microorganisms in drinking water that  are capable of reproducing
there.  This concern was recently expressed by the British (21) in an article
that described the proliferation and detection  of these organisms within water
distribution systems.  We propose  therefore that a test similar  to the present
standard  plate count be performed  as  a routine part of the microbiological
surveillance of all potable waters and that appropriate MCL's be set. Such a
test should be incubated for a minimum of  two  days  at  22 to  30°C to ac-
commodate slower growing bacteria that are  adapted to the reduced  ambient
temperature of their  aquatic environment. However valuable the information
gamed, the time lag required for the performance of this test presents a real
logistic problem since each re-test following  an excessive result will take at
least two  days. The implications of this are obvious in terms of the volume of
water  used in every two-day period.  For that reason the utilization of the
Limulus lysate test would be useful to check for excessive bacterial popula-
tions when the standard plate count  or coliform counts are high  and the
operator  needs  rapid assurance that his  remedial action has been effective.
This idea may be somewhat futuristic  since  the  Limulus  lysate test  has not
yet been perfected or widely  accepted.  However, we feel that this wiH become
a reality  within the next two to five years.
   Based  on  the arguments presented  in the preceding sections of this dis-
cussion, we propose the following recommendations for a bacterial indicator
system in finished water.
46

-------
                               ALTERNATIVE INDICA'
B.
C.
   A.  Total coliforms should be retained as the primary
       organisms with a slight relaxation of that MCL  (
       ml).
       A test similar to the standard plate  count should
       regular basis as an index of othex microorganisms
       in water. The MCL level might be  near the 207nil
       in Germany or as high as 500/ml.
       In the event either the  total coliform or  standarld
       exceed the MCL, a Limulus lysate  test should be
       becomes available.
       These recommendations  represent a realistic yet
       to use the enumeration  of  heterotrophic bacteria
       to measure the  safety of finished drinking  waters.
       there are imperfections in this plan, as in others
       posed, largely due to  the statistical  nature  of the
       inherent in searching for heterotrophic  bacteria within
       alien environment of treated water. In spite of these
       good reason to believe that this protocol will improve
       able health safety record of finished  drinking
       States.
   e determined on a
   that are  important
    limit that is used

    plate count data
  performed when it

  comprehensive plan
  and their products
   We 'recognize that
 lhat have  been pro-
  task and difficulties
      the hostile and
   difficulties, there is
    the already laud-
Water in  the  United
                         REFERENCES
                                                   Es:h
 1. Bissonnette, O.K., J.J. Jezeski, G.A. McFeters, and D.G. S
    of environmental stress  on enumeration  of indicator b;
  ""waters. Appl. Microbiol. 29:186-194.
 2. Bonde, G.J. 1962.  Bacterial indicators of  water
    quantitative estimation. Teknisk Forlag,, Copenhage.
 3. Carson, L.A., M.S.  Favero, W.W. Bondl, and N.J. Peterse;
    cal, biochemical, and growth characteristics of p. cepacia
    Appl.  Microbiol. 25:476-483.
 4. Dutka, B.J.  1973.  Coliforms are an  inadequate index
    Environ. Health 36:39-46.
 5. Evans, T.M.  1975. The Limulus  lysate assay  a rapid
    bacterial  water  quality. MS Thesis.  Montana State
    Montana.
 6. Fliermans, C.B., R.W. Gorden, T.C. Hazen,  and G.W.
    distributions  and survival in  a thermally altered lake. A;
    biol. 53:114-122.
 7. Galvand, M.M.  1974. Fecal  contamination—the water:
    sponsibility. Sewage Works. 12:66-69.
 8. Geldreich; E.E. and B.A;  Kenner. 1969. Concepts  of  l
    stream pollution. J. Water Pollut. Control Fed. 41:
 9. Geldreich,  E.E.  1973. Is the total count necessary?
    Proceedings  AWWA  Water  Quality   Technology
    Ohio.
10. Geldreich,  E.iE.  1967. Fecal coliform concepts  in stream
    Sewage Works. 114-.R98-R110.
11. Hendricks, C.W. and  S.M.  Morrison.  1967.  Multiplication
    selected enteric bacteria in clear mountain stream water. Water
12.  McFeters, G.A. and D.G. Stuart. 1972. Survival  of coliform
    waters; field and laboratory studies with membrane fill
    Microbiol.  24:805-811.
                                                         'ORS/G.A.MCFETERS
                                                       group of indicator
                                                       e. less  than 3/100
                                                      uart. 1975. Influence
                                                      .cteria from natural
                                                pollution. In:  A  study  of
                                                      .  1973. Morphologi-
                                                      firom distilled water.

                                                      of  water  quality. J.

                                                     j.nd sensitive test  of
                                                     University,  Bozeman,
   . 1977. Aeromonas
  ipl.  Environ. Micro-
                                                      part I, analysts' re-

                                                      ecal streptococci in
                                               :R336-R.352.
                                                    S< ction
                                                  Conference
        vii-i—yn-2.
           Cincinnati,
                                                       pollution. Water &

                                                          and  growth of
                                                          Res. 1:567-576.
                                                       bacteria in natural
                                                      er  chambers. Appl.
                                                                      47

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                 Coi nparatr
13. McFeters, G.A.,
    Stuart.  1974.
    gens in well wate
14. Menon, A.S. and
    microbiology of
    Environ. Technol.
15. Hundt, J.O. 1962.
    Appl. Microbiol.
16. Mundt, J.O. 1962
    Appl. Microbiol.
17. Nemedi, L. and
    aeruginosa in wa
18. Phirke, P.M. and
    of stream pollution,
19. Ptak, D.J., W.
    dence of Klebsiel a
20. Seidler, R.J.,
    environment:
    pneuriioniae
    29:819-825.
21. Water Research
    water  during  di
    Centre, London.
                 cu
                fron
L. J. McCabe, Jr., £
Protection Agency,
  What  are some cf
of the Limulus lysat
range of ambient w;
                     .K. Bissonnette, J.J. Jezeski, Carole A. Thomson, and D.G.
                         :ive'survival of indicator bacteria and enteric .patho-
                    r."Appl.  Microbiol. 27:823-829.
                   .  W.K. Bedford. 1973.  Study of the seasonal effects on the
                   :iorthern  pulp and paper  mill aeration lagoon. Can.  Dept.
                    Develop. Rpt. 4-AR-73-1. Ottawa, Canada.
                    Occurrence of  enterococci in animals in a wild environment.
                    11:136-140.
                     Occurrence of enterococci on plants m a wild environment.
                    77:141-144.
                    B. Layni.  1971.  Incidence  and hygienic importance  of  P.
                    er. Acta. Microbiol. Acad. Sci. Hung. 75:319-325.
                    S.R. Verma. 1972. Significance of enterococci as indicators
                    in. Indian J. Environ. Health 74:328-334.
                    jinsburg, and  B.F.  Willey.  1973.  Identification  and  inci-
                    i in chlorinated water samples. JAWWA 65:604-608.
                     . Knittel,  and  C. Brown.  1975. Potential pathogens in the
                    tural  reactions  and  nucleic  acid  studies  on  Klebsiella
                      clinical   and  environmental  sources.   Appl.   Microbiol.
                    Centre.  1976. Deterioration of  bacteriological quality  of
                    tribution. In:  Notes on water  research.  Water  Research
             QUESTION AND ANSWER  SESSION
                   health Effects Research Laboratory,  U.S. Environmental
                     iincinnati, Ohio
                     the more significant problems associated with the  use
                     assay to determine natural water quality over the broad
                    rters that might be tested?
G.  A. McFeters, Department  of Microbiology,  Montana State University,
Bozeman, Montana
  As you have suggested, there are still some potential -problems in the ap-
plication of the  Likiulus lysate test for determining water quality.  For in-
stance, we feel that more  needs to be known regarding .the persistence^ _of
endotoxin in  waters  of v3P"°LfI!fL chemical types as well  as in
water treatment and
to dinerent
  Beyond this, we
the assay procedure
                    renovation processes. JPossible references of the test
                   aiemistries or "additives Should  also be investigated.
                   also acknowledge  the need for
                                                                       due
are commercially a
allow the correlaticn
                     itself  (at present there are four different test  kits that
                    ailable)." Also, more studies need  to be performed  to
                      of this  new test_with_ other^slablisjig^^
\yaieE-qiiality.  Such
ditipns.
                    testing  should be done
                                                        world"water_can-
48

-------
                                  STATISTICAL CONSIDERATIONS/L. MEUNZ


       Some  Statistical ConsideratioBS in Water Quality  Control

                              Larry Muenz
                National Cancer Institute, Biometry Branch
           National Institutes of Health, Landow Building, C-509
                        Bethesda, Mainland 20014.

   As an outsider I  am particularly grateful for the opportunity  to  address
 you today regarding water quality control.  I  am a statistician, not a bac-
 teriologist  or  water quality expert,  and have  been asked to look at some
 statistical aspects  of water sampling. My  analyses are based on the use of
 the membrane filter method, although  they could  certainly be applied to
 other techniques. Indeed, what I have done is an example of the general class
 of quality control sampling analyses. Each  sample must be classified as being
 either good or bad depending  upon whether or  not one can detect  some rela-
 tively rare phenomena. My analyses today are  also based upon the idea that
 the use of  coliform indicators  is desirable, that  they really indicate something
 we want to know. I do not take into  account two facets of the problem that
 no doubt need ultimately to  "be considered. They are (a) in  measuring a
 certain number of E.coli> using them as indicator bacteria, we have indicated
 the presence  of  an uncertain number  of pathogens of various  sorts  and
 (b) the doses of  pathogens needed to induce illness in say 1% of the people
 drinking water with that level of  contamination changes radically from  one
 sort of bacterium to another. Regarding (a),  and this is an issue that the
 experimentalists need to consider, it would  be important to have a probabilis-
 tic relationship between measured number of E.  coli and probable numbers
 of various  sorts of pathogens  believed to be present.
 -"Specifically, I was asked to investigate  three topics. Theser are (1)  how
 many samples a  month  should be taken  for  various size towns, (2)  how
 much water should be taken hi each sample and (3)  are the present sampling
 rules well defined? My knowledge of the rules is based upon the EPA interim
 regulations of December  25,  1976, i.e.  specifically,  the  table appearing in
 paragraph 141.21 and the instruction of 141.14.              \
  The proposed  regulations  specify  a.  number  of samples  that increases
 steadily with population.  It is  not  evident that  this is necessary or desirable.
 For example, a nationwide voting poll doesn't need substantially more people
 hi its sample than one restricted to a single city. A critical aspect is homo-
 geneity: are we taking a sample from a homogeneous source?  If so, many
 big samples are no better than a smaller number of smaller samples. Figure 1
sHows'the "number of samples per 1,000 population according to the proposed
EPA regulations.  There are several observations. First, it evidently declines
with increasing town size.  This seems desirable.  There is a small insert in
 the beginning of the graph; this shows the number of samples per 1,000 popu-
lation for small towns up to a population  of approximately 90,000. The extent
 of this insert  on  the larger  graph is indicated  by a tick mark on the hori-
zontal axis of the larger graph.  The insert shows an extremely ragged  line,
definitely not a consequence of an analytic model. It  is not  necessarily bad
to use seat-of-pants procedures to determine the number of samples per head
but in this  case it appears not to be rational  for small communities.
  The next question refers to the amount  of water  per sample. A standard
sample is now 100 milliliters. There is nothing  unique  about this value other
than that  it corresponds to a convenient  size dish or test  tube. Standard
statistical theory can tell how large a sample is  needed to makej sure that we

                                                                      49

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 V)
 z
 8
 en
 tu
 a.
 Q

 I
 O
 o
 Ul
 _J
 a.
 <
 C/J
 cc
 Ul

 I
                                     1.2
                        1.1
\
   V
           090
                          2500


                  POPULATION (THOUSANDS)
                                     I
 Figure  1. NUMBER OF SAMPLES TAKEN PER 1, 000 PERSONS.
                                                                    5000
will  find something if something is there. Again, the critical issue is  the
same as previously raised. Do the tubes properly reflect the heterogeneity of
the source?                                                 .
  The regulations say that we should examine a  certain number or. tubes
containing 100 milliliters each, that number depending upon community size.
For  each batch of tubes, say 20 of them, the sampling rule is composed of
two parts. First, if the mean number of bacteria per tube is too high, declare
the sample to be contaminated. The threshold value  is now set at one bac-
terium per tube. Second, if more than 5% of the tubes contain more than 4
bacteria—even though the overall mean may be less  than one—also declare
the sample contaminated.
  To study 'this  two part sampling procedure we need an analytic model to
describe the number of caliform bacteria in a sample. With such a model
statistical theory can  find the best way of testing  the bacterial level of the
sample and then the proposed procedure can be compared with the optimal
one. It may be that the optimal procedure is excessively complex  and the
proposed procedure is an effective, simple approximation.
  The sampling problem is described by a number  of parameters. These are:
1. The level  of  contamination  that we consider just at  the border line of
acceptable and  not acceptable,  that is the threshold.  2. A measure  of vari-
 50

-------
                                  STATISTICAL CONSIDERATIONS/L. MEUNZ


 ability of contamination from one water sample to another. 3. The probabil-
 ity that with a given decision procedure and a certain level of contamination
 that we will "pull the alarm" at exactly the border line level.  4. The proba-
 bility that we will fail to detect some high level of contamination, one  sub-
 stantially  greater than the border  line value.  5. The historical  information
 on the frequency of waterborne disease in communities of various sizes.
   It is important to distinguish between the case  in which the water is uni-
 formly mixed and homogeneous and one in which the water has an important
 degree of heterogenity. With  uniform mixing a relatively small number of
 samples will suffice to indicate the average number of bacteria per volume of
 water. In the heterogeneous case we can consider an extreme, indeed un-
 realistic, example with  an  extraordinarily high degree of variability. This
 might be reflected hi a situation in which all the contamination was con-
 tracted into the form of a bolus. I call this the "bowling ball" model in which
 unless you  hit the particular focal point  of contamination with your sam-
 pling procedure you will detect no contamination at all. To have a reasonable
 probability  of detection  we would  need an extraordinarily large number of
 samples. In the more realistic situation in which contaminants are somewhat
 clumped it is evident that we will still meed a larger number of .samples than
 in the situation of homogeneous mixing.                      :
   When the water is  uniformly mixed the relevant distribution for the num-
 ber of particles per  standard volume of  water is the Poisson  distribution
 which has a single parameter and for which mean and variance are the same.
 If we assume that this mean varies from sample to sample and can  itself be
 considered a random  quantity we arrive at the negative binomial distribution.
 Professor Pipes of Drexel University and some of his graduate students have
 considered the practical problem of distinguishing between  these two possible
 distributions, Poisson and negative  binomial. Based  on a  large  number of
water samples they have almost invariably come to the conclusion  that the
negative  binomial fit the data better than does the  Poisson.  While this is
reassuring it isn't  at all surprising since the negative binomial distribution has
two parameters and the  Poisson only one. Thus we have more flexibility in
fitting with  the negative  binomial. Since the negative  binomial arises for the
case of the  Poisson where the mean is  allowed to  vary, in the extreme case
where the mean doesn't vary we get back  to the Poisson, that is, one  is a
special case  of the other.  The question then becomes: is the negative binomial
distribution  enough of an improvement relative to  the Poisson to  warrant the
addition of  another parameter?  Using the data collected by Dr. Pipes'  stu-
dents I conclude the answer is 'yes'. Making special provisions for a model
with a high  degree-of heterogeneity, namely the negative binomial, is justified
in real data.
   As I have indicated the sampling rule is in two  parts, one referring to the
mean and one to  the number of tubes containing an  excessive  number of
bacteria.  As was indicated in a handoult received prior  to this  meeting there
are problems with this rule for very small numbers of samples. For example,
one bad tube can guarantee that the mean will be too high  regardless of zero
counts in the others.
   Although this rule may not be ideal I have analyzed it using the negative
binomial model referred to  above.  The model is  parameterized  by  a mean
level of contamination, a coefficient  of variation for the prior  distribution of
the mean  and a sample size, -that is, number of tubes. I have calculated the
probability that a given water sample will  be accepted for various combina-
tions of the  parameters. Figure 2 shows the results of a computer simulation.

                                                                      51

-------
                                                           zs
                      PROBABILITY OF ACCEPTANCE OF SAMPLE
    I
    i
SSSSS
mnj_|3>m
3mX<2
m z m > 2
£232g

s|S8B
glsci
isi^q


aia

mil
= ro3oS
                                                           00

-------
                                  STATISTICAL CONSIDpRATIONS/L. MEUNZ
Because I have used a simulation rather than exact calculations the answers
have  a great deal of variability themselves and sometimes behave counter-
intuitively.
   Some general observations are that: 1. There is still a; high probability of
accepting a relatively bad sample, one with a mean of three coliform per 100
mis.,  if the number of tubes is small enough. 2. There is a general tendency
for the acceptance probability to increase as the variability of the mean in-
creases. This is entirely reasonable if we think back to my bowling ball model.
3. Things do  behave in a  rational fashion in that  the  probability of ac-
ceptance goes  down with  increased mean  and with increased numbers of
tubes.
   If  my model for bacterial  distribution is a reasonable  representation of
reality then this rule is not the optimal one.  It is suboptimal in the sense that
for a given probability of considering uncontaminated samples to be con-
taminated we could achieve a  higher probability  of cqrrectly detecting  a
contaminated sample with  some  other rule. Unfortunately the optimal  rule
is a good deal more complicated. It says "reject when the  mean is too high,
but the threshold that defines too high must be based on the variance of the
number of bacteria found per tube."
   The last topic I want to consider is historical data regarding the level of
water-borne disease by community  size,,   The  goal of quality control is to
      .0011
     .00055  •
m
<
m
O
cc
Q.
                           COMMUNITY SIZE (Log 10)
      Figure 3. PROBABILITY ANY ONE PERSON WAS INVOLVED IN AN OUTBREAK
              OF WATERBORNE DISEASE DURING 1946 - 197lO.

                                                                      53

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 keep to a low level the probability that some person will contract water-borne
 disease. In a community with a poor historical record for the incidence  of
 such disease we would have to be more stringent with the sampling rules  in
 order to keep the probability  down to the same level for rail citizens in all
 comknunities.     .       ,       ,	i   .,.,„•	, ,	
   The historical record is a confusing one. My last figure shows  the proba-
 bility that a single individual will be involved in an outbreak of water-borne
 disease at some time  in the time interval 1946-1970  as a function of com-
 munity size (on a Iog10 scale). Small  towns (3-5 thousand) and rather large
 towns  (50-100 thousand) have higher rates than towns of  other sizes.  A
 possible explanation for this is that the very smallest towns don't need elab-
 orate  quality control  mechanisms so  they  have a low disease rate. Larger
 towns (the first peak in the curve) do need such surveillance but don't intro-
 duce it until town size is near ten-thousand. Towns of 50-100 thousand popu-
 lation need quality control of  a higher order of sophistication but this  isn't
 to be found (the second peak) until  town size approaches one-half million.
 Of course,  this reasoning is entirely a product of having seen the data first.
   If we accept the odd behavior of this curve as representing reality we see
 that there could be objections  to a sampling plan which weighted the inten-
 sity of sampling using historical information. From a political viewpoint, it
 might be difficult to justify less sampling for towns of 10-thousand  population
 than for those which are either smaller or larger.
   After all this, please note that the maximum value  of the line indicates a
 probability of less t'han one part in a thousand over a 25 year span.
   I urge you to remember that sampling and quality control are problems to
 which the statistician can make a major  contribution  and his  advice should
 be solicited regarding any proposed regulations.
                                                   i


                                                   iESSl
             QUESTION AND  ANSWER  SESSION

S. M. Morrison, Department of Microbiology, Colorado State  University,
Fort Collins, Colorado

  How does your model take into account heterogenity with respect to small
and large towns that have their own treatment and  distribution systems?

L. Muenz, National Cancer Institute, National Institutes of Health, Bethesda,
Maryland

  It doesn't do so _in any explicit way. The coefficient-of -variation parameter
can be adjusted to agree with historical information from a town of a given
size but this sort of adjustment is not built into the model; it's at the discre-
tion of the user.  A more complicated model would already  have the com-
munity  size-variability  relation explicitly  included  although  I  wonder how
useful such a mixture of mathematics and survey information would be.
 P. D. Haney, Black and Veatch, Kansas City, Missouri
',     ,   '       "     ',-.'•     :." ..... " ...': ..... ;,:, .....    . ; , , i '  .       ,          " ..... :;
   In the 1940's the U.S. Public Health Service  published a graph giving the
 relation between number of water samples/month and population. J. K. Hos-
 kins, who was Assistant Surgeon Genera.! at the time  said that the mathemat-
 ical derivation wais so complicated that they only published the curve arid
 not the mathematics, Perhaps you could "track"' this down.
54
                                               jjiii^^  	:,.,. i.:,,,!!,:,!,.,'.,:,it

-------
                                 STATISTICAL CONSIDERATIONS/L. MEUNZ
L. Muenz
  My plot shows the number of samples/month/1000 persons as a function
of community size  using -information from  the E.P.A. interim regulations.
Assuming that this  is  the same number of samples/month as referred to in
the 1940's there is no way to ascribe a mathematical origin to this curve. The
demonstration of this  is given by looking at the  "ink blot" of points corre-
sponding to very low populations. This region is blown up in the insert and
a tick mark (') on the horizontal axis of the larger graph shows how far
the insert extends.  The behavior of  the insert  is completely  non-analytic
even if most of the larger graph seems rather smooth. No conceivable mathe-
matical  function could give rise to  the jagged line appearing in the insert.
Thus the whole curve is likely to be empirical; this, of course, doesn't mean
that it's bad,  just that it would be difficult to defend.
                                                                     55

-------
:l	
  Mi!*!/  I
                                                                    I, i

-------
                           CHLORINE RESIDUAL SUBSTITUTION/L.J. MCCABE


             CMorine Residual Substitution — Rationale

                          Leland 3.  McCabe
                  Health Effects Research Laboratory
                 U.S. Environmental  Protection Agency
                        Cincinnati, Ohio 45268
  Consideration of the use of chlorine residual as  an indication of the safety
of a  drinking water  supply was much easier before the discovery of  the
chloroform  problem.  In the past a chlorine residual was  an unquestioned
good. Now  the  value of the residual must be determined so  that trade-off
considerations can be given to the possible increase in  chloroform and  the
other trihalomethanes. Both toxicologies!  and epidemiological studies indi-
cate an adverse  effect of chloroform ingestion or the consumption of water
that has been chlorinated.
  The value of carrying a disinfection residual to  overcome  contamination
throughout  the  distribution system of  a community water  supply has  not
been  determined. Johns Hopkins University now has a research project to
determine the effectiveness of the residual to contain introduced  contamination
to a distribution system (3). The value of the residual in a  special situation
has been reviewed. It is believed that residual chlorine is a substantial safety
factor on shipboard distribution  systems but  credible evidence is not avail-
able to insist that such a residual be maintained (2).
  Besides the safety  factor that a  disinfection residual may  provide to a
distribution  system there 'is another benefit that  must be considered.  If  the
water supply operations is  set up to maintain a residual, the absence  of  the
residual can quickly signal  the inadequacy of treatment or the introduction
of foreign material into the distribution system. Thus, the measurement of
a disinfection residual can provide  a continuous-real-time monitoring  of  the
safety of the distributed drinking water.
  To control the spread of  waterborne  disease, information on water quality
must be available in  time  to influence decisions  concerning the use of  the
water. Microbiological methods are not adequate as they only provide his-
torical data  on the quality of water severail days -ago. Microbiological methods
can be very helpful in the investigation of the causes of waterborne disease
outbreaks but they do not lend themselves to monitoring.
  The need to rethink what we have  been calling water quality monitoring
or surveillance was demonstrated in the Community Water Supply Survey
(4). The  survey design called for sampling all water supplies  in one metro-
politan area of  each  of the Public Health Services regions. A  late change
was made in Region I to include the State of Vermont instead of New Haven,
Connecticut. Some virus studies  were being conducted  on the New Haven
supply  and  it was  thought best  not to include  this supply as  part of  the
national survey. Nationally, it was  found that at 85 percent of the water
systems, bacteriological samples were not collected  at the  rate prescribed in
the drinking water standards. Table  1 is a  copy of the table from the report
of the survey.  Samples were collected at less  than half  of the prescribed
frequency at 69% of  the  supplies surveyed. The prescribed frequency is
admittedly an arbitrary number but when  efforts were made to examine  the
required frequency the established frequency was universally  supported by
the state  sanitary engineers in 1968. The results of  the bacteriological sam-
pling are shown in Table  2  and it can be seen that  the problem of poor

                                                                       57

-------
in
09
                                                                                                                                 n
                  TABLE 1. BACTERIOLOGICAL SURVEILLANCE EVALUATION BY COMMUNITY SIZE.
                                                   Population Served in Thousands
                                                (All data are per cent of System Totals)
                                                     Overall System
                                                        Totals
                                            =.5
       .5-5
5-10  :  10-25   25-50  50-100  >100  Number Per Cent
                Met bacteriological
                   surveillance criteria
                Did not meet bacterio-
                   logical surveillance
                     criteria
                   Did not collect
                     samples at the
                     rate in the bac-
                     teriological  sur-
                     veillance  criteria
                Systems totals—number
        18     12      10
                      12      36    100       10
96      82     88      90     97      88     64     869      90

   94      74     83      85     92      88      64    827   55



  446    315     75     59     36      16     22    969     —
                   Note:   60 per cent of the survey population were served drinking water by
                 systems over which a bacteriological  surveillance program met the criteria.
                                                                                                                                 w
                                                                                                                                 o
                                                                                                                                 f
                                                                                                                                 S3
I
p
                                                                      I.

-------
                          CHLORINE 1RESIDUAL SUBSTITUTION/L.J. MCCABE


TABLE 2.  WATER  SUPPLY SYSTEMS IN VARIOUS  POPULATION
     SIZES WHERE DELIVERED WATER EXCEEDED THE COLIFORM
     DENSITY LIMIT.
Population Range
^500
501-5,000
5,000-10,000
10,001-25,000
25,001-50,000
50,001-100,000
> 100,000
Totals
Number of
Systems
446
315
75
59
36
16
22
969
Number of
Systems
Exceeding
79
29
5
5
1
1
0
120
Per Cent of
Systems
Exceeding
18
10
7
9
3
6
0
12
Per Cent of
Survey Population
Exceeding
15
8
: 7
7
3
6
0
2
quality  water is  greater with the smaller supplies. Something needs to be
developed that can address that problem.
  The usefulness of the chlorine residual to maintain the water within the
bacteriological limits of the standards was also demonstrated in the  CWSS.
Although the practice  of chlorination  caused a dramatic decline in the fre-
quency of the samples containing colifoims, these organisms are not eliminated
merely by claiming that chlorination is practiced.  Unless a chlorine residual
was maintained in the distribution system, a significant portion of the samples
from  the distribution system contained coliforms.  The coliforms are elimi-
nated nearly completely in systems providing at least a  trace  of chlorine
residual  (1).
  Buelow and Walton (1) also analyzed Cincinnati's water surveillance data,
during a period when a free chlorine residual was being maintained in the
plant  effluent. This  is illustrated in Figure 1 when the free residual was car-
ried in 1969-1970.  All monthly average coliform counts were below 0.1 per
100 ml.  and in most cases the  samples  with coliforms were collected  from
locations without chlorine residuals.
  If the reality of these data is faced,  it is obvious that something new must
be done in  the way of improving  bacteriological surveillance.  The idea of
substituting  the chlorine residual  for coliform testing emerged from con-
sidering the  results of the Community Water Supply Survey. The concept was
outlined  by  Robeck (5) but  he cautioned that five specific problems should
be addressed before the substitution is  allowed for a water supply.
  1.  The number and location of  samples  for which chlorine residuals are
      to  be  substituted.
  2.  The form and concentration of chlorine residuals to be maintained.
  3.  The frequency of chlorine residual determination.
  4.  The analytical method  to be used.
  5.  The past bacteriological history of sampling location.
  The problems were considered and paragraph(h) of Section 141.21 of the
National Interim Primary  Drinking Water Regulation (6)  allows for _the
substitution  of chlorine residual for 75%  of the microbiological sampling.
At least  4 chlorine  residuals  are substituted for each  microbiological  sample
dropped  but the residual must be measured daily. The benefit of the use of
this provision to  the  safety  of  the  smaller water  supplies  is obvious. Daily
monitoring of a water supply would not be achieved by the coliform test,

                                                                      59

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS

 I
 o
 u
 a.
 e
 8
 u.
 Id

1.6
K
A
.3
.2
.1
0
JA







N.



iru
—










f


_


_j_r




1

12 MONTH AVG. MF COUNT/100 ml
1966 0.11
1969-1970 0.02




_n _T~~^ Ln-J-T-r-1

MAR. MAY JUL. SEP. NOV. JAN. MAR. MAY JUU SEP. NOV. *MAY JUL. SEP. NOV. JAN. MAR.
1966 1967 1969 1970
          Figure 1. MONTHLY AVERAGE COLIFORM COUNTS PER 100m1
unless the supply served 25,000 people. For a population served of 25,000 to
28,000, 30 microbiological samples are collected per month.
  If the rules are followed the chlorine residual procedure will assure much
quicker action in correcting a problem of unsafe water.  "When a particular
sampling point has been shown to have a free chlorine residual less than 0.2
mg/1, the water at that location shall be retested as soon as practicable and
in any event within one hour." Thus, the operator has one hour to increase
or restore the chlorine application and make sure  the residual reaches  the
sampling point.  With coliform testing, considering sample transit  times, time
to perform  the  test, and time to make the notification, the reaction  to bad
water might be half a week's delay. A good example of the response follow-
ing the loss of chlorine residual  was the investigation  of a cross connection
in Tampa, Florida (Communication  from Hillsborough County Health De-
partment to the State Division of Health, November 27, 1973).  A pressure
sewer had been mistakenly connected to the water main and the location of
the problem was accomplished with the chlorine residual testing.
  Chlorine  residual testing could have been  required in addition to the re-
quired microbiological monitoring but to promote the  use of the more effec-
tive  monitoring it was decided to eliminate the more  costly microbiological
testing. Savings effected can be used for other  water  quality testing. Responsi-
bility for first line  quality  control must rest with the  water works operator.
The present system of collecting  limited samples at  small water supplies and
sending them  to the state laboratory  for testing introduces  a mystical aspect
tp quality control. The operator is not involved personally and reports come
back a week or more later that may indicate deficiencies in the past. When
the operator determines a  chlorine residual and does not find  a pink color
     '     '  '  ,  , Ml ,      .                , ,   	,,    I |   I ||
 60

-------
                          CHLORINE RESIDUAL SUBSTITUTIQN/L.J. MCCABE
 (DPD test), he is the only one to know there is a problem and he must make
 a  decision to protect the public's health. The responsibility  for  an acute
problem is where .it should be and at the only place where necessary correc-
tive  action can be taken. The publication  of the causes of waterborne  out-
breaks must be intensified to reinforce the responsibility of the water-works
operator in the  prevention  of disease.  This  will require the reporting  and
investigation  of  outbreaks and  this activity  can be supported with  EPA's
water supply program grants.  The state  laboratory  microbiolbgist  should no
longer be  overburdened with routine bacteriology samples and should be  able
to participate  in outbreak investigations to determine etiological agents  and
their source. Table 3  shows a tabulation of waterborne disease outbreaks by
agent.  The large  percent  attributed  to the  nonspecific  category  gastro-
enteritis points out the need for more extensive  microbiological workups
during outbreak  investigation.
  Section  141.21 paragraph (h) requires the collection of coliform samples
when the  chlorine residual cannot be restored within one hour. This sample
has  a  great likelihood to contain collforms  but it is to be averaged in to
indicate the monthly  means for compliance  with  section 141.14. This pro-
vides a much stricter  level of quality for a  supply using the chlorine residual
procedure. For example a water supply serving  5,000 population is required
to collect  6 microbiological  samples  and may substitute the chlorine test for
4 of them. Thus 2 coliform samples would be  collected per month  but  sur-
veillance would be provided daily by the chlorine residual test. If the residual
were low  on one day of the month and the follow  up coliform  count was 4
per  100 ml the  system would not meet the  drinking water standards even
though  the supply likely had good quality  water all but  one day of the
month. To promote the use of the chlorine residual substitution, which pro-
vides really a better  monitoring system, the days with adequate  chlorine
residual should be considered zero coliform days .in computing an average.
          TABLE 3.  WATERBORNE DISEASE OUTBREAKS
                   DISTRIBUTION BY ETIOLOGY.
                               1946-1974


Gastroenteritis
Infectious Hepatitis
Shigellosis
Chemical Poisoning
Giardiasis
Typhoid
Salmonellosis
Amebiasis
Poliomyelitis
Enteropathogenic E. coli
Tularemia
Leptospirosis
Number of Outbreaks
Community
• - 'Water Systems

52.6
16.3
9.6
5.9
5.2 - .
4.4
4.4
0.7
0.7
0.0
0.0
0.0
135
Other
\Vater Svfitems
TT **iw|. kJj^utiJ.ji3
47.5
13.7
10.2
4.0
2.5
15.8
2.8
1.2
0.0
1,2
0.6
0.3
322
                                                                      61

-------
f! P *
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


                                 Recommendation

                                   I would  think that complete chlorination substitution would be usable for
                                 the smaller supplies. If this were allowed for supplies up to 9400 P°P^ion
                                 served it would also take care of the averaging problem with a small number
                                 of microbiological samples. A bacteriological sample is required when the
                                 chlorine cannot be restored within an hour. The operator of the system wou d
                                 of course continue to work to get the residual restored and it will faave^o
                                 be assumed that .this will be done as diligently as possible When  he getsthe
                                 residual back up  to  the prescribed level the public is again protected. What
                                 WoSd be the best microbiological test to perform on this  sample that was
                                 Taken before the problem was corrected? A sample for Cl^dmm perfr^es
                                 would seem to provide data on what was needed to be known and the sample
                                 could be just mailed to the laboratory without  a concern **««"«***;*
                                 these organisms  are more resistant to chlorine and should be ubiquitous
                                 enough to  indicate when polluted water enters the distribution system.
                                    For large water systems the full benefits of the chlorine -residua  should.be
                                 promoted  Continuous monitoring is possible with feedback  controls o adjust
                                 The treatment and give alarms when departures  occur. In these  systems the
                                 microbiology applied to  samples collected when  the residual chlorine cannot
                                 ™e restoredI within one hour  should be rather complete. The testing should
                                 assist in determining  the  source  of contamination  and ought  to be  more
                                 elaborate than just a coliform test.
                                                           REFERENCES

                                 1  Buelow,  RiW., and  G. Walton.  1971.  Bacteriological  quality  vs.  residual
                                 2  CoesCorol.  1977.  Final report  of  the  drinking  water
                                    Section  ad HcS Advisory Committee. Department of Health,  Education
                                    and Welfare, Atlanta, Georgia.                                         ,
                                 3  Kruse  CW  1976.  Grant Application  R804307— Biological evaluation  of
                                  ' the  benefits  of maintaining a  chlorine residual in public water systems.
                                    US Environmental Protection Agency, Washington, D.C.
                                 4. McCaS, L.J., J.M. Symons, R.D. Lee, and G.G. Robeck.  1970.  Survey  of
                                    community water supply  systems. JAWWA 62:670-687.
                                 5  Robeck  G G.  1974.  Substitution  of  residual  chlorine measurement tor
                                    SSbution  bacteriological  sampling.  Proceedings  AWWA  Water  Quality
                                    Technology  Conference,  Dallas,  Texas.                          ....
                                 6. United  States  Environmental  Protection Agency.  1975 .National  interim
                                    primary drinking water regulations. Fed. Register 40:59566-59574.
                                              QUESTION  AND ANSWER SESSION

                                    R. R. Colwell, Department of Microbiology, University of Maryland, Col-
                                  lege Park, Maryland
                                    How can chlorine  residuals supply information concerning actual bacteria
                                  populations in drinking water? If chlorine is that effective in the treatment of
                                  pathogens in sewage,  why do  we need to chlorinate the drinking water if we
                                  do it to sewage?
   "U,« I",	!•(!	•
                      ::•	.hi,, Sim,:, , 'Iff"
                                  62
                                 ''Jr'.'iWJI	H!*ifc'.; tW;ii.t!(>'
n' H1'™ 
-------
                           CHLORINE RESIDUAL SUBSTITUTION/L.J. MCCABE


   L. J.  McCabe, Health Effects  Research Laboratory, U.S. Environmental
Protection Agency, Cincinnati, Ohio

   The chlorine residual test is not supposed to give a quantitative evaluation
of the bacterial population. When  the chlorine residual exceeds the specified
amount  it is an immediate  qualitative evaluation that  pathogens  are  not
present.
   Pathogens are  also derived from non-point sources  that are not treated. It
will be a great day when all  sewage is disinfected adequately and with reli-
ability, until  this  Utopia, it  would seem desirable to disinfect drinking water.
A residual disinfectant provides  some measure of protection in  distribution
systems.

   W. Ginsberg, Water Purification Laboratories, Chicago Bureau of Water,
Chicago, Illinois

   Do you believe water  treatment plant  operators are likely to take  four
times the number of chlorine residual samples when they often do not take a
sufficient number of coliform samples now?

   L. J. McCabe

   We want to provide to the operator the  most effective test for him to use.
Their use of  any  test is a matter of personal modification and; enforcement. I
think these would be enhanced with a. chlorine residual measurement. In the
case where neither test were used, the chlorine residual requirement would
waste less money  because it  would not be necessary to have an unused micro-
biology capability standing  by. With the  use  of the  chlorine residual,  the
microbiological capability can  be used for more in .depth analyses of the most
meaningful  samples and be  ready  for   waterborne  disease  outbreak  in-
vestigations.
                                                                      63

-------
1.    <-!'       '          i	ii

-------
                         MICROBIOLOGY OF POTABLE WATER/R.R. COLWELL
          Public Health Considerations of the Microbiology  of
                             "Potable"  Water            •   •      •    '•

                   R.R. Colwell, B. Austin and L. Wan x
                       Department  of Microbiology
                          University of  Maryland
                      College Park, Maryland 207.42
                    1 Department  of Civil Engineering
                     Howard University
                     Washington, D.C. 20001

   Water is essential for every biological function and the convenience of ob-
 taining a ready supply of drinking water  from municipal supplies is accepted
 without question in civilized societies. In fact, the occasional odor of chlorine
 when a tap is turned on  has  been interpreted by  the layman as a guarantee
 that the water has been purified and is,  therefore,  safe to consume.  Unfor-
 tunately,  there are problems, other  than microorganisms, as,  for example,
 taints,  odors, floccules and disturbing colorations^  which cause  temporary
 distress and,  perhaps, a telephone call of complaint to the supplying agency;
 otherwise,  in North  America water is  available  and used without  further
 thought.
   Natural watejrs_which flow in rivers  and streams, and settle in wells and
 lakes, Contain a Jffetincjnini^^
 un^lssolYed^jjiOT^^                                                 micror
 flo7ararF"potential pathogens  and,  therefore, hazardous lo the health of the
 ctSnTBTImi^whereas ioAer__jmcrop^uiis^s_areJjeneficial  in mineralizing
 dfganic matter, concentrating heavy -metals  in tficMlng~~filte"rs,  and 'acting"
               ~~'
function of "water supply agencies to provide  "clean"  water, in quantities
necessary to cope with the demand of the community. This water is collected
from rivers and  wells and sometimes from estuaries or the  sea. The water
that is collected is treated with flocculants. e.g. alum, to precipitate suspended
material, filtered  through sand beds to remove precipitates and floes of living
organisms, and chlorinated Jo_gliminate potentially harmful organisms. Addi-
tional  compounds often  are  added to "the. water  (such as~3iuprides;  and the
finished water is distributed to storage  reservoirs  or through1 pipes to con-
sumers. The water  is not •microbiolo.gicallv  sterile since spores and many
Gram-negativerods  survive chlorination, and it is  not pure, as fluorides_and
other compoundlTmav be  added. However, thT quality ot tn¥ finished water
'should be safe for_drinkin~a_and-usa^. in.hospiMs.^phannacies  and  industry.
 Periodically, samples of water are collected from ffiecfistribution  system and
chemical analyses done, but, unfortunately, rather primitive techniques are
used to assay microbiological quality.  The microbiological analyses are recog-
nized to be time-consuming and  may provide questionable  information, par-
ticularly since the few bacteriological tests that are most commonly done are
performed  by inexperienced  or inadequately  trained personnel. Very  rarely
is a registered microbiologist employed to  carry out the bacteriological tests.
  Increasingly, problems  with drinkiing water  are  being brought  up  for
public discussion. Setting aside the problems of toxic chemicals  and  speaking-
only to microbiological criteria for waiter quality, there are several problems
that need to  be addressed, as, for example, the  difficulties in .establishment
of  standard  laboratory  procedures,  standardization of , laboratories, and
development of criteria  for qualification of professional environmental and

                                                                      65

-------
               ff
:* M. "  . !
Ji H1,,  :• :
if-!!	
f ft
                           4:'
                           "Si1:"
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


sanitary  microbiologists. The major problems  are, clearly, (1) standardiza-
tion of personnel and training; (2)  standardization of  laboratories carrying
out the analyses; (3) standardization of reporting results and banking of the
data for coordination and intercalibration around the country; and (4) stand-
ardization of media and reagents^ as well as procedures for sample gathering,
plating, culturing, reading tests, interpretation of test results, and recording
of data.  Clearly, the need for professionalization of water  testing and  water
quality determination is most urgent. Papers in this series deal with this prob-
lem and it  can be seen that there are solutions,  but that effort, time and
money wiilbe required.                                             .   .
  The microbiological  standard for drinking water  has not changed  in  its
essentials since  1914 (10), and is determined  by  reference to the presence
of coliform bacteria. The definition of coliform is  vague and includes  many
asporogenous, Gram-negative rods  (21). Their isolation from water is de-
pendent  upon the antiquated Most-Probable-Numbers Technique (1) or  by
the newer  mebrane-filtration  method (1). The presence of "coliforms"  in
numbers greater than 1/100 ml (27) is taken to mean that the water is con-
taminated.  The solution is often re-chlorination of  reservoirs and pipe water,
and the problem, thereafter, quickly forgotten. It takes at  least  48 h for
coliforms to be recognized by the  standard procedures  employed (23),  .by
which time  the contaminated water will have been  distributed and consumed
by  many individuals in the community. Thus, at best, the  bacteriological
methods presently employed provide only an indication that trouble occurred
in the system and the event may have  been 2  or  3  days ago, hardly  to  be
considered satisfactory  water quality control. Thankfully, water-borne infec-
tions are rare in the U.S.A., even though the potential  exists for dissemina-
tion of disease-producing.microorganisms. Thus, modern microbiology tech-
niques should be incorporated in routine examination  of  water. Gas-liquid
chromatography  appears to offer  immediate detection of  bacteria and bac-
terial products by direct analysis.  Methods such as pyrolysis and laser scan-
ning  also offer  potential value in direct  assessment  of water quality. The
application  of molecular genetic techniques to water bacteriology appears to
be  almost totally lacking. Reasoner (23) correctly stated that  "the develop-
ment: of quantitative procedures for pathogenic bacteria has  lagged behind the
development of quantitative procedures for non-pathogenic bacteria." Un-
fortunately, it can also be stated that water bacteriology has not advanced to
a degree commensurate with that  of  other aspects of  microbiology, most
notably of  microbial genetics and physiology.  If  the aim of  water quality
determination is to  locate the presence of pathogens or indicator bacteria,
then it is essential not only to be aware of other types  of organisms occurring
m water, but also to understand their physiology, ecology,  and  role in public
health.

Bacterial Microflora in Treated  Water

  Although it is relatively easy to  remove some microorganisms from water
by  filtration and chlorination, it is difficult to reduce  the_contgn^ of organic
material to a level below which it w2LJMJl_supparJ_mic]X>^^
                                  	^	  	„_-.- -, means of disinfectants are generally
                                  ineffective and invalidate the description of pure ^water(13JL-The  viable,
                                  aerobic, heterotrophic bacteria which  can be routinely isolated from  treated
                                  water include Acfiromobacter, Arthrobacter, Bacillus, Clostridium,  coryne-
                                  forms, Flavobacterium,  Gallionella, Klebsiella, Leptothrix, Micrococcus, My-
                                 rcobact.erium, Proteus. Pseudomonas and Serratia  (4, 26, 29,  30, 32). These
                                  66

-------
                          MICROBIOLOGY OF POTABLE V
 rATER/R.R. COLWELL
  survive inadequate treatment or enter via contaminatior
  within the distribu-
  tion system. Although none of these organisms are primary pathogens, some
  may  be responsible for taints and odors,  spoilage of commercial products
  for which water is used, and also may act as secondary invaders in human
  disease processes.
    Pseudomonas and Flavobacterium are problematic genera, some species of
  which are opportunistic pathogens  (14). The latter genus, a: heterogeneous
  taxon of convenience (3, 20), comprises a  variety of sp;cies, some of which
  are tolerant to high levels of chlorine, with certain strains  resisting as much
  as 100 mg/L. Pseudomonas and  Flavobacterium spp. have  been noted to
  cause heavy growth in  the sprinkler heads of scrub  siiiks  and water baths
  in hospital operating theatres and preparatory areas (15|). Yellow-pigmented
  rods, described as flavobacteria, have been isolated frota patients with sys-
  temic infections  following heart-lung  surgery.  One  species,  F. meningo-
  „ -	I*	  *_ _A1_4-».«.!*•* MAnlnt-n w£ +rt rt *"»*l! !•». « J-»+-I rtf  /* O O ^ rt +*A 1  t 4-CT »-»»*£lCie» »*»^-»O  P M/%1 t I ft
 septicum, is relatively resistant to antibiotics  (22) and
 not be tolerated in treated water. The dangers resulting f;x>m bacterial metab-
 olites, such as pyrogens,  should not be overlooked,  anc  in one study (24)
 it was demonstrated that water used for pharmaceutical
 viable bacteria  producing pyrogens. The problems caused  by bacteria in
 water extends to scientific laboratories where, deionized '
 have  been shown to contain pigmented  Gram-negative
 bacterial growth occurs in  the resin columns, with  the
 comprising as many as 10s  cells/ml (2:5). Surely,  meth
  its presence should
  purposes contained
 and soft water units
  rods. In particular,
  bacteria population
 ids mtist be devised
"to ensure the sterility of all water that is to be used as drinking water.

 Purification ol Water

   The effectiveness  of chlorination justifiably can, be  criticized because of
 the resistance of many viruses and  bacteria to chlorine  treatment. Treatment
 with ozone is  considered  to be an alternative  method of water treatment
 since  ozone is  bacteriocidal  and removes taints and oc.ors without leaving
 unpleasant  residues  (11).  Methods not presently employed in water treat-
 ment, but of potential'Value, include heat sterilization a; id membrane filtra-
 tion, both of which  can  be adapted for purification  of  large  volumes of
 water.  Reverse osmosis  and  ultrafiltration are  perhaps
 and effective in reducing bacterial  numbers  (6). In all
 claving and ultrafiltration, viable organisms remain in ft
 ment and this fact raises several questions (13):

    (a)  are  the  microorganisms remaining in drinking
        tional treatment merely harmless  indicators of an
    (b)  are  they beneficial by  assisting in the breakdown
     .  pounds  transported by  the water?
    (c)  are* they a hazard to the system, the user or as
        end product of industrial processes intended for  i
        by  humans?
   the  most  feasible
  cases,; except auto-
  e water  after treat-
 yater after  conven-
  unsterile condition?
    of organic corn-

 contaminants in  an
ingestion or injection
 Microbiological Problems  with Water Treatment  and Distribution

   McCabe and co-workers (19)  found that, of 969 public:water supply
 systems serving 18.2 million individuals, 613 utilized grpumd-vyater systems,
 67% of which distributed water to 2.8 million consumers without any disin-
 fection. Of 621 wells,  9%  produced water that contained cbliforms,  2%
 contained fecal coliforms,  and 11% of the wells dispensed water with >

                                                             :          67

-------
III	I'.j I ,1 II
in ::R I- -  iii
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


500  heterotrophs/ml as  determined by plate counts. How many pathogens,
particularly  organisjms  causing  gastro-enteritis, were distributed  in those
waters can  only be speculated.                                      .     .
   It is particularly disquieting  to  realize that even  water in hospitals  is
far from pure. In one interesting study (18) it was recorded that as much
as 30-40%  of water  contained  in so-called  sterile tanks and  pipes  serving
operating theatres derecbnt~ammate
-------
                         MICROBIOLOGY OF POTABLE WATER/R.R. COLWELL


 before a sanitary problem is detected, when it may already be too late for
 action to avoid health hazard.

 Tap  Water

   Bacteriological  examinations of  tap  water frequently reveal the presence
 of large numbers of bacteria. Goliforms and fecal  streptococci occur infre-
 quently in  chlorinated  water,  and from their  presence  is  inferred 'recent'
 contamination.
   In  a bacteriological  study of tap  water undertaken  in  our  laboratory,
 (Table 1 ) , it was shown that samples collected in the Washington, B.C. area
 were  negative for coliforms,  as determined by incubation for up to 7 days at
 35 °C in lactose broth.  However,  after this extended incubation time  the
 plate counts on glucose tryptone yeast-extract agar (GTYEA) were alarming-
 ly high,  although lower than for bottledjwate£-gP^|3III5fler incubation
 for only 48 h these plate counts were very low, yet the prolonged incubation
 period at 25 °C was necessary for  development of the majority of the bac-
 teriai~c"olomes. ihis trend was  repeated, in parr, on iiosin-MetnyIene~-Blue
 agar (JtJMBjTwhich allows the selective growth of coliforms. Several samples
 (Table 1)  contained /3-haemolytic colonies on blood agar  incubated  aerobic-
 ally, and anaerobically  under hydrogen;. Thus,  it is recommended that plate
                            , the incubation periods  be  increased  to' 7-14
                         of bacteriological  media  be employed in routine
examination of water..
Drinking Fountains

  The convenience"of drinking ^fountains  for  providing refreshing,  cool
water is  accepted as part of everyday  life, yet the bacteriological standard
of the water at public drinking fountains is low. The drinking fountain head,
unless regularly cleaned, is often dirty.  In 1965, 78% of drinking fountains
tested (12)  were  contaminated, yet in 1975,  76% were  in  a similar poor
state. It is interesting to note that a wide range of microorganisms were re-
covered from  the  fountains. Although it is assumed that drinking water al-
ways contains a residual chlorine level, the  water from the  drinking foun-
tains tested by Herman (12) was found to be devoid of the disinfectant.  This
may be explained by the separate water delivery systems for drinking foun-
tains, in which the water is cooled and constantly circulated,  with the result
that the chlorine level  usually falls below the effective level  of  0.6  mg/L,
thus permitting a diverse range of microbial growth at the air-water interface
on  the fountain head. Clearly, the bacteriological  quality  of water at com-
munal drinking fountains needs  to be  constantly monitored and,  where
necessary, the water delivery systems modified.

Bottled Water

  Bottled water has gained popularity  in the  U.S.A.  with annual sales ex-
ceeding $150  million  (31).  In 1972,  there were an estimated  700 water
bottling plants  (8), with  California  the largest  customer.  The  quality
standards of bottled  waters varies  between states, with some complying  with
the Public Health  Service Water Standards (22), whereas others have lesser
or no standards. However, from the bottle labels it is  often implied that the
quality is superior to the  municipal water supply, and that this quality will

                                                                       69

-------
TABLE 1.  BACTERIOLOGICAL EXAMINATION ;OF TAP WATER OBTAINED FROM THE DISTRIBUTION SYSTEM
                          WITHIN THE WASHINGTON, D. C. METROPOLITAN AREA"
. ; No. of colonies/ml
Presumptive • '
colifonns/
100 i
Source after
U.S. Capitol . .
U.S. Supreme Court
Washington-Suburban Sanitary Station
Alexandria City Hall
Fairfax County Offices :
Montgomery County Offices .-
Prince Georges County Court House
nl
7 days
0
0
0
0
0
0
0
GfYEA"

2 days
5
20
0
0
5
10
0

7 days
7.8X102
1.4X104
1.1 XlO2
7 XlO3
4 XlO3
50
0
EMB°

2 days
0
40
0
0
0
0
0

7

4

9
1



days
60
XlO3
5
XlO2
.2X10S
0
0
Hektoen
Agar

7 days
0
0
0
0
0
0
0
Sheep Blood Agar
(Aerobic) (Anaerobic)

2 days
5
30
0
0
0
20
0

7 days
10
30
5
0
0
30
0

3 days
40
70
0
0
0
60
0
 a Water was allowed to run for.2-3 min before collecting samples in sterile botfles. Inoculated media, except for the presumptive coliform
    test, were incubated at 25°C for the time periods indicated.
 "Glucose tryptone yeast-extract agar (Difco).
 c Eosin methylene-blue agar (Difco).                                                                     .
o
I
8
                                                                                                                    a

                                                                                                                    1

-------
TABLE  2. BACTERIOLOGICAL EXAMINATION OF BOTTLED WATER. INCUBATION  OF MEDIA,  EXCEPT  THE
  PRESUMPTIVE COLIFORM TESTS, WAS AT 25°C.
No. of colonies/ml
Presumptive
coliforms/ GTYEA °
100 ml
Source (Brand name) after
Aqua Pure0
Deer Park
Great Bear
Mountain Laurel
opting vvaicrui9
Mountain Valley
Mineral Water
Poland Water obtained from :
Grocery Store
Manufacturer
U.S. Supreme Court
7
0
0
0

\j

0

0
0
0
days 2 days
10'
4.4X10°
5 XlO2

u

0

4 XlO6
6.6X105
1.8X10°
7 days
102
con"
1.5X108

JU

0

con
con
con
EMB"
2 daysj |
0
2 XlO5
6.3 XlO5

6 A1U

0

5 XlO4
3.5X104
1.8X104
7 days
0
con
6.6X105

z xiir

0

7 XlO*
3.5X10*
1.8X10*
Hektoen
"Pnfpru*
Agar
2 days •
0
5.1 XlO3
0

u

0

0
0
0
7 days
0
6.4X10°
0

u

0

0
0
6 XlO2
Sheep Blood Agar




(Aerobic) (Anaerobic)
2 days
50
2xlOa
3 XlO3 .

i.axiir

0

1.7X10*
1.5X10*
3.7X10*
7 days
50
con
5.4X104

J..JXHT

0

1.8X104
1.5X10*
6.4X10*
3 days
0
1.1X10"
3 XlO3

6 XlO-

0

0
0
1.8X10°

§
o

R
o


S

w
w
*fc>*
a Glucose tryptone yeast-extract agar (Difco)
" Eosin methylerie-blue (Difco)
0 Pasteurized
d con = confluent growth
"From hot springs, >90°C
* Refrigerated
8

-------
           'I!,,11" 'I   H,1 lift!
'ffl i
it	
                   MICROBIOL STANDARDS EVALUATION/CW. HENDRICKS


                   be  maintained throughout the shelf life  of  the product. Unfortunately, this
                   is usually far fronji true,  and the bacteriological quality of  freshly bottled
                   water varies  from [brand  to brand  (9). Furthermore,  the  bacterial count
                   tends to increase With storage. The usefulness of  the bottled water standards
                   adopted by the Fo'od and Drug Administration,  which became effective on
                   May 22,  1974 (28 j, is questionable. They do not include a plate count limit
                   for bacteria,  but  tjbie coliform  limit of  less  than .2.2 organisms/102 ml  is
                   similar to the dririking water standard. It  is a  help  th.at bottled  water  is
                   categorized as a fiod, but it is  a pity that bottlers often recommend it for
                   preparing baby food, pharmaceutical  prescriptions  and hot  and cold  bev-
                   erages! Thus, certain members of the community, namely the young, aged
                   and infirm, may bz more susceptible to infection by opportunist organisms
                   which could occur in the water.
                     In a study undertaken  in this  laboratory, the bacteriological  quality  of
                   bottled water obtained in Washington,  D.C.  was examined  (Table  2). Al-
                   though the coliform numbers, as determined  by  the production of  acid and
                   gas (in lactose brbth)  within  48 h, were  negligible;  in  some  cases  there
                   were large numbers of aerobic heterotrophs  as determined from plate counts
                   	j_ ~_ c-v-r'Vrj A  .^^^iirv,  TV.Q einnififant  nninf tn hp marlfi IS that. CXCeOt
                  medium. The significant point to be made is that, except
                   coliform tests, the test media were  incubated  at  25°C.
                   37°C, little or no growth was obtained. Thus,  it is mis-
                    made on GTYEA
                    hi  the case of  the
                    When incubated at  -*/  •»-', ***.».*^ v* *+~ e,~—..~ ..—	
                    leading  to enumerate  bacteria in water  by incubating the inoculated test
                    media at 37°C.
                      It can be seen (table 2)  that pasteurized water, i.e.,  Aqua  Pure, and re-
                    frigerated mineral [water, was  of  good bacteriological quality;  however, the
                    numbers of bacteria present hi other brands was disturbingly high.  These
                    data were obtained several years ago  (Washington Star, Washington, D.C.,
                    July 11, 1971). Monitoring of the microbiological quality of  bottled water
                    should be done on! a regular basis to ensure the public health  safety of this
                    product. Of particular concern was the large number of /Hiaemolytic colonies
                    which developed oln blood agar. These colonies may reflect  the presence of
                    potentially harmful  organisms  in the water, and the numbers greatly exceed
                    those found in  tap! water (Table  1). Furthermore, it is  readily obvious that
                    the bacterial  counlts for stored, bottled water, viz.,  Poland water obtained
                    from a  grocery st|ore  and from  the U.S. Supreme  Court at  the tune the
                    study was carried put,  were higher than in water obtained directly from the
                    bottler. Clearly, long-term storage of bottled water can have an  adverse effect
                    in  providing for gjrowth of bacteria. It would seem appropriate for bottlers
                    to  either pasteurize  the water, or  require refrigeration during storage.
Recommendations

  Too little is known about the microbial ecology of water at  source, treat-
ment, distribution iuid storage. It is essential that the normal microbial flora
of drinking water lie regularly monitored, and  the taxoriomic composition of
the microorganisms present in source  water" be determined on a regular
basis. Thus, seasonal variations in the flora can  be  measured so  that altera-
tions  in the flora lan be recognized to be abnormal increases in the -total
number of microorganisms or selective increases in a few taxa due to season-
ality.
  It is desirable not to have any  viable organisms  in  drinking  water, and
methods to improve the  treatment of  water, including removal  of viruses,
     , i    ",,,   *   I 	;;	;	M	 	| „„,,,•	  	:	
                 i   	'   	'

-------
                          MICROBIOLOGY OF POTABLE WATER/R.R. COLWELL


 and  maintain its purity until  consumption, should  be employed  in water
 treatment systems.
   Methods presently employed for the examination of water are inappropri-
 ate  and give a poor idea  of water quality. It is important to  examine the
 effectiveness of a wide range of media, such as blood agar and low  nutrient
 media, pre-treatment at low temperatures, and of using a wide  range of in-
 cubation temperatures,  especially low temperature  incubation  of 20-25 °C
 for  enumeration of total viable  bacterial populations. Thus,  psychrophiles,
 mesophiles and  thermophiles  should be  considered  because of -the  variety
 of conditions  and uses for drinking water. Non-selective  enrichment tech-
 niques have been  successful in the recovery of Salmonella spp.  (17)  and
 Vibrio cholerae  (5)  from  the  Chesapeake  Bay,  and these  methods  may be
 effective  in  determining the sanitary  quality of drinking  water.  It is  im-
 portant to remember that no single technique is  suitable for the isolation of
 all pathogens; therefore, a combination of methods is  recommended, with
 the minimum being inclusion of total viable counts at 25 °C.
   The concept of a residual chlorine  level in  drinking  water needs  to be
 carefully scrutinized because of chlorine resistance amongst viruses and many
 bacteria. Although chlorine-resistant coliforms  may not  occur, the presence
 of Flavobacterium and Pseudomonas is undesirable and represents a potential
 health hazard. Examination of drinking water for the presence of Pseudo-
 monas should  be incorporated  in the routine analyses of the  water since  a
 suitable method exists for the enumeration of P. aeruginosa (16). Also, other
 Pseudomonas spp. are reasonably easily identified and enumerated.
   The  taxonomy of Flavobacterium must be radically improved, and isola-
 tion procedures for pathogenic  species of this genus should be developed..
   The  problems appertaining to antibiotic  resistance and  plasmid transfer
 between taxa should be  studied and the significance of these phenomena to
 public health^safety determined.
   In summary, the present methods for bacteriological analysis  of drinking
 water, although unsatisfactory  in  many respects, should not  be discarded
 in favor of the chlorine residual or other  such  chemical test. Rather,  the
 bacteriological  testing should  be improved,  expanded and modernized. It is
never wise to throw the baby out with the bath water,  and substitution of
 the  chlorine residual  for  sound  microbiological analyses  would  do just
that!
                           REFERENCES

 1. American Public Health Association. 1971. Standard methods  for  the  ex-
    amination of water  and wastewater, 13th ed. American  Public Health As-
    sociation Inc., New York.
    Bayliss, J.Ri 1930. Bacterial aftergrowth in water distribution systems. Water-
    works and Sewerage  77:335-338.                                  '  .'
 3. Bean,  P.G.,  and  3JR..  Everton.  1969. Observations on  the  taxonomy of
    chromogenic bacteria isolated  from cannery environments.  J. Appl. Bacteriol.
    52:51-59.
" •. Clark, P.M.,  R.M. Scott, and  E. Bone.  1967.  Heterotrophic iron-precipitating
    bacteria. J. Am. Water  Works Assn. 59:1036-1042.
 5. Colwell, R.R., and J. Kaper.  1977.  Vibrio species as bacterial indicators of
    potential health hazards associated with, water. In Bacterial indicators/health
    hazards associated with  water. Am.  Soc. Testing Mat. pp. 115-125.


                                                                       73
&•

-------
                                      MICROBIOL STANDARDS EVALUATION/C.W. HENDRlCKS


                                      6. De Roos, R.L., D.R. Vesley, A.G. Du Chene, and L.J..  Hart.  1976.  High-
                                        purity water supplies for biomedical research laboratories. Health Lab.  Sci.

                                      7. Geldreich,  E.E.,  H.D.  Nash, D.J. Reasoner,  and R.H.  Taylor.  1972. The
                                       ' necessity of controlling bacterial populations  in potable waters:  Community
                                        water supply. J. Am. Water Works Assn. 64:596-602.
                                      8 Geldreich,  E.E.,  H.D. Nash,  D.J.  Reasoner,  and R.H. Taylor.  1975a  The
                                        bacterial quality  of  bottled  water.  News  Environ.  Res.  Cincinnati,  U.S.,

                                      9 GeldreichPE.E.,  H.D. Nash,  D.J.  Reasoner,  and R.H. Taylor.  1975b. The
                                        necessity of controlling bacterial populations in potable waters—bottled  water
                                        and emergency water supplies. J. Am. Water Works  Assn.  67:117-124.
                                     10 Haney, PJD. 1977. Evaluation of microbiological standards for drinking water.
                                      •'••/„ Engineering Control  Practices Technical Symposium, Office of Water Sup-.
                                        ply. U.S. E.P.A. pp. 13-14.
                                     11. Hann, VA. 1956. Disinfection of drinking  water  with ozone. J. Am. Water
                                       ' Works Assn. 48: 1316-1320.                                            .
                                     12. Herman, L.G.  1976.  Sources  of the  slow-growing pigmented water bacteria.
                                     /•;' Health Lab. Sci. 15:5-10.                                .
                                    /(J3) Herman, L.G.  The  slow growing  pigmented water bacteria: Problems  and
                                    VKsources. Adv.  Microbiol. In press.
                                    /(14 /Herman  LG., and  H.J. Fournelle. 1964. Flavobacteria:  A water-borne po-
                                    "V7tential pathogen. In  Illrd International  Congress of  Chemotherapy;  Inter-
                                        national  Society of Chemotherapy, pp.  1352-1555. G.T. Verlag.
                                     15. Herman, L.G. and C.K. Himmelsbach. 1965. Detection and control of hospital
                                        sources of  flavobacteria. Hospitals 39:72-76.
                                     16.*Highsmith, A.K., and R.L. Abshire. 1975. Evaluation of  a Most  Probable
                                        Numbers technique  for  the  enumeration of Pseudomonas  aerugfnosa.  Appl.
                                        Microbiol.  50:596-601.                                       ,„*,.•*
                                     17. Kaper, J.B., G.S. Sayler, M.M.  Baldini,  and  R.R.  Colwell. 1977. Ambient-
                                        temperature primary nonselective  enrichment  for  isolation of  Salmonella
                                        spp  from  an estuarine environment. Appl. Environ^  Microbiol. 55:829-835.
                                     ! g Kelsey, J.C., and M.M. Beeby. 1964.  Sterile water "Ibr  operating theatres.
                                       "The Lancet, July 11, No. 7350:82-84.
                                     19. McCabe, L.J.,  J.M.  Symons,  R.D. Lee  and  G.G. Robeck  1970 Survey of
                                        coramunity  water supply systems.  J. Am.  Water Works  Assn.. 62:670-687.
                                     20 McMeekin,  TA., DJB.  Stewart, and J.G.  Murray. 1972. The Adansonian
                                        taxonomy and the  DNA base  composition of some  Gram negative, yellow
                                        pigmented  rods. J. Appl. Bacteriol. 55:129-137.
                                     21. Orndorff, S.A., B. Austin,  LA. McNicol, and R.R. Colwell. 1978. Isolation
                                        of  lactose-fermenting Pseudomonas  spp.  from  the  aquatic  environment.
                                        Can.  J. Microbiol. In press.
                                     22. Pederson,  M.M.,  MA.  Marso, and M.J.  Pickett.  1970.  Nonfermentative
                                        bacilli associated with man. Am. J. Clin. Pathol.  54:178-192.
                                     23. Reasoner,  D.J.  1974.  Microbiology-detection  of bacterial  pathogens  and
                                       ' their  occurrence. J.  Water Pollution Control  46:1395-1408.
                                     24. Saunders, L.,  and E.  Shotton.  1956.  Water for pharmaceutical purposes.
                                        J. Pharm. Pharmacol. £: 83 2-847.
                                     25. Stamm,  J.M.,  WJE. Engelhard,  and  J.E.  Parson.  1969.  Microbiological
                                        study of water softener  resins. Appl. Microbiol. 13:376-386.
                                    H|k Tennessee Dept.  Public Health. 1966. Epidemiologic studies of atypical  and
                                     ^ acid fast bacilli in Tennessee. P.H.S. Grant No. CC00078; 133 pp.
                                     27 United States  Public Health  Service.  1962.  Drinking Water  Standards.  Dept.
                                        Health, Ed. and  Welfare, P.H.S.  Pub.  No. 956, Washington, D.C.
                                     28. United  States Food  and Drug Administration. 1973.  Bottled water quality
                                       'standards:  addition  of  fluoride and current good manufacturing  practice
                                        regulations. Dept.  Health,  Ed. and Welfare, Federal  Register  38,   226:
                                    r.,  32558-32565.
                                    \29\ Van der Kooij,  D.  1977.  The occurrence  of Pseudomonas spp. in surface
                                     V
|||;,;  -''I  ,      (;v   _   (   ,,,-11  ;|'|^   ,  45.187:197.	    ' "	      ;    ";    ;;	
                ; ,           ' ' ii!	  f<,ii   '        !'  '          ...];•• ; •  ,    •     , ;  , :' ' ,;  "i ,:     :
                1            " '	:"    74

-------
                          MICROBIOLOGY OF POTABLE WATER/R.R. COLWELL
5^9. Victoreen, H,T.  1969. Soil bacteria arad  color problem in distribution systems.
    J. Am. Water Works Assn.  62:429-431.
 31. Waters, E.P. 1974. What about bottled  water? F.D.A. Consumer., U.S. Dept.
     Health, Education and Welfare, Washington, D.C.
    Willis,  A.T. 1957. Anaerobic  bacilli in  a  treated  water supply.  J. Appl.
    BacterioL 20:61-64.
                                                                         75

-------
                                                       llliillL	  |i|J
                               I ...... ,  '  :"!.
                                                    " Wf
It ,PK	'     ' : "1
                               i!'!	It:'.
f    |
                                                    •;< .f	"
I'lit,  :  t ft
                                                                                                                                                                                                                              1 II'NIII	!„	
                                                                                                                                                                                                                              ,Lii-"«   *  i

-------
Water Analysis by the Membrane Filter Procedure
SESSION 2
                  Coliform Detection and
Control
Chairman: Mr. Edwin E. Geldreich,  Water Supply Research Laboratory/
MERL, U.S. Environmental Protection Agencj
,  Cincinnati, Ohio
                                                                 77

-------
friiLL .......... :kliii;id£ ..... lili ........ li^     ........ k ..... ii
si, ..... i!i ..... kiih ..... Mi:* .......
                                                                 J H' <;
                                                                          :r'lj

-------
                           SAMPLING COLLECTING AND HANDLING/H.D. NASH
                     Sampling Collection and Handling

                              Harry D, Nash
                 Water Supply Research Laboratory/MERL
                   U.S. Environmental Protection Agency
                           Cincinnati, Ohio 45268

    The ideal transported water sample is  one which reaches  the  laboratory
  with  its  bacterial  flora unchanged,  both  qualitatively and quantitatively, as
  it was at  the  time  of collection. The type  and degree  of change  is  best
  determined by analyzing the  samples at the time and location of collection
  for comparison of results obtained  by the same  procedural analysis  after
  transit. Unfortunately  this  action is  impractical for  a routine drinking water
  monitoring program. Changes which may occur are dependent upon many
  variables  and combinations of  the  individual variables. Factors  which in-
 fluence the bacterial flora during transit  are  quality of the sample  bottle,
  bacterial population, density of each type comprising the population, concen-
 tration of  nutrients both organic and inorganic, fluctuations in temperature
 during transit,  turbidity, clumping  and antagonistic  action; all of these ex-
 hibit their effect on the bacterial flora during the transit/storage period. These
 factors can and do vary for every water sample, regardless of water classi-
 fication  as to source,  surface, ground or potable water. Sampling  locations
 must  be  carefully  selected  to  insure representative sampling, collection pro-
 cedures must be conscientiously followed  to prevent  accidential contamina-
 tion, and every attempt must  be made to transport samples promptly to the
 laboratory for processing. Failure to comply with these facts regarding sample
 collection  and handling will produce data which are confusing, misleading
 and detrimental to  any monitoring program.

 Collection of Potable Water Samples

   Proper collection procedures and techniques  are fundamental to  obtaining
 reliable results for  any water sample.  Errors relating to sampling are usually
 greater than those  associated with  laboratory analysis.
   The laboratory is usually responsible for supplying sterile sample bottles.
 Each bottle must be  properly  prepared with respect  to the  type water being
 sampled  with  thiosulfate  added to  bottles for sampling  chlorinated water
 supplies,.or a chelating agent (EDTA)  for  waters containing metals, copper
 and/or zinc.  Quality control  tests  must  be conducted on each  batch of
 bottles sterilized to assure sterility of all bottles. Sterile bottles should not be
 stored more than 30 days to reduce possible contamination due to accidental
 damage during  storage. After  sample collectors receive  sterile bottles, it is
 their responsibility  to prevent accidental contamination during the collection
 procedure. Bottles must not be carried in vehicles unprotected  and certainly
 not for several days to even weeks. Samples should be collected from a  distri-
 bution system tap which does  not have any attachments. Taps which have
 anti-splash devices  and/or home purification attachments should be avoided
 or removed  before sample  collection.  It should  be  known  that the tap is
 supplying  water from a service pipe directly connected with the main and is
not served from any type of storage tank (1,2,5). The tap should be opened
and water allowed to run to waste for 2 to 3 minutes, or for a time sufficient
to clear the  service line. The recommended time for flushing varies from 2 to
5 minutes. Flushing time is  relative; the intent is to flush the service line to

                                                                       79

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                    i
                                                    I

assure that the sample collected is representative of water in the main. Treat-
ment of water taps before collecting potable water samples is unnecessary if
water is allowed to  run to waste and reasonable  consideration given to the
choice of sampling taps (5). Thomas et al  (13)  reported that flammg taps
resulted in no significant differences for both total  and fecal conforms nor for
the standard plate count.
   The sample bottle must remain closed until it is to be  filled. The stopper
must be protected from  contamination  when removed and during the actual
sample collection. The  bottle should be  filled without  rinsing, leaving an
ample air space so that the sample can be thoroughly mixed before analysis,
and the stopper replaced immediately  (1,5).
   CompHance with  recognized sample collection protocol can prevent ac-
cidental contamination and reduce the occurrence of spurious samples. Sample
collection is an important  part of the testing procedure; the test is no better,
nor  more reliable  than the sample  which arrives at  the  laboratory for
analysis. Laboratories must be certified, evaluated,  approved—choose  your
word—the intent is the same; however,  there are few programs which provide
adequate  and repeated  training for  sample  collectors. Such  training is
recognized as worthwhile but economically impractical;  whereas  failure to
have such a program may really be the  economical impractabihty. Serious
consideration must be given to the implementation of a program designed to
train and  periodically recertify  sample collectors,  similar to  the  milk pro-
ran. m1, ;  , i  '.,«
                                  grin'application, transit time and storage are so  interrelated, they should be
                                  considered together. Standard  Methods  (1) provides recommendations for
                                  transit/storage times for stream, source and potable water samples. Stream and
                                  source water samples should be held  below 10°C during a  maximum transit
                                  time of 6 hours. Such samples  should be refrigerated when received m the
                                  laboratory and processed within two hours.  When local conditions dictate
                                  delays in 'sample delivery  longer  than  6  hours, consideration  should be
                                  given to field examination or by use of the  tentative delayed-incubationtotal
                                  cbliform procedure. In a study on sample  storage time-temperature effects,
                                  Geldreich et al (6)  observed that MPN coliform counts conducted on samples
                                  held  at  5°C for periods  up  to  48 hours were in acceptable  agreement
                                  with initial MPN results than counts  observed on samples stored at 13-32°C
                                  or ^t 35°C. Also, they observed better stability for samples held at 5°C and
                                  examined by the MF technique. The more precise results obtained by the
till'',   '<• I   .     Ili  •      ! i;i  !t"'j|  • direct MlF count as compared to the statistical MPN estimate revealed a more
                                  subtle change in bacterial  densities during  storage. These  recommendations
                                  and information basically apply to stream  and source  water  samples.  Since
                                  these requirements  may not be realistic for individual potable water samples
                                  transported to the laboratory by mail service,.the time interval between.col-
                                  lection and analysis should  in no case exceed 48 hours, preferably within 30
                                  hours as specified in Standard Methods. Even the recommendation for refriger-
                                  ation for potable  water samples can  be  interpreted  as optional and not
                                  a requirement.
                                     Presently, most State programs do not enforce the 30-hour transit/storage
                                  time  limit.  Because of a multitude  of reasons,  e.g. inefficient mail service
                                  and inability to utilize other means of  transportation  due to geographical
                                  locations  of remote water supplies, State  programs  allow examination  of
                                  samples  which  arrive within  48  hours, or  more  realistically, within two
                                  calendar  days. Few, if any State programs require refrigeration of potable
                                  water samples during the transit period.
                                   80

-------
                           SAMPLING COLLECTING AND HANDLING/ELD, NASH
    Conscientious investigations were conducted  by many workers  in  an at-
  tempt to determine a transit/storage time of water samples which would not
  adversely affect the bacterial flora either qualitatively  or quantitatively. Cox
  and Clairborne  (4) reported that refrigerated water samples showed relatively
  little change after 48 hours, while paired  samples stored  at room tempera-
  ture showed a noticeable change.  However, the lowest coliform density ob-
  tained at zero hours  storage was 390/100  ml  which would not qualify their
  samples as  potable water. Shipe and Fields (12)  indicated that a chelating
  agent  (EDTA)  added to stream samples known to contain copper and zinc,
  0.16 mg/1 and  6.5 mg/1, respectively,  reduced the rate of decrease  in E.
  coJi density. They concluded that this  chelating  agent  would.be of value
  in maintaining  the coliform index near the  level existing at  the time of
  sample collection for a  storage period of 24  hours. Data  indicated  that
  addition of  EDTA at the time of sampling resulted in approximately  69%
  of the cells  remaining viable after 2 hours and 40% after 24 hours.  How-
  ever, there is no indication of survival of coliform organisms in water when
  the initial density was between 1 and 10/100 ml. Calwell and Parr (3)  and
 Hoather (8) reported a gradual decrease in coliforms with increased storage
  time whether samples were iced or not. Lucking (10) concluded that the time
 element between collection and examination is  important  and testing must
 begin within 48  hours of collection. Lansome et al (9) concluded that sam-
 ples having a coliform density between 1-20/100 ml need not be refrigerated
 up to a 48-hour  storage period with no adverse effects on coliform detection.
   Previous investigations relating  to transit/storage time  of'samples  have
 omitted information  regarding the  influence of increased nonrcoliform  den-
 sities on coliform survival and detection, especially with respect to potable
 water supplies. However, Lucking (10)  and Lansome et  al  (9)  certainly
 recognized this as a factor directly relating to the over-all  problems of transit
 and storage. They also recognized that this and other factors such as changes
 in temperature, availability of nutrients, oxygen tension and  chemical concen-
 tration are extremely varied in widely separated geographical areas, therefore
 making the transit/storage problem unique  for each water supply. Excessive
 non-coliform densities may interfere with coliform detection  when examining
 potable  waters expected to have  low coliform  densities  (7).  Analyses of
 bacteriological data  from  a  survey of distribution water transmitted by
 969 public water plants did suggest that high non-coliform densities can in-
 terfere  with  detection  of low levels of coliform. Data in Table 1 shows that
 the frequency of detecting total and fecal coliforms increases as the standard
 plate count increased  to levels up to 500/mI; but decreases  in frequency
 when the non-coliform densities exceed 1000/ml. These findings indicate that
 high densities  of non-coliform bacterial populations  may  adversely  affect
 coliform detection. The factor of high density non-coliform  organisms versus
 low density coliform organisms is just being investigated and  may be cor-
 related  to transit/storage  time for potable water as  well  as  stream  and
 source water supplies. Data obtained on the  distribution system water quality
 in Northern Kentucky reflects the potential interference of coliform detection
 exhibited by high densities of non-coliforms after a 24-30 hour transit/storage
 period.  The  Northern Kentucky treatment plant laboratory data, based on
 samples examined less than six hours after collection, indicated  satisfactory
 water quality throughout the distribution system. Data obtained on similar
samples collected  during the same sampling period and from the same distribu-
tion system but examined after a 24-30 hours transit time, were reported by the
 State Laboratory  as being potentially unsatisfactory. Results  of these samples
showed excessive or confluent bacterial growth.  In some instances sheen was

                                                                       81

-------
'; 1	i1!""1 II!1 !,». XI i,
                              MCROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                       PLATE COUNT VS. COLIFORM DETECTION
                                                         WATER  NETWORKS  FOR   969  PUBLIC
                                  WATER SUPPLIES.
                           -.! .
                                   General Bacterial
                                                               Total Coliform
                                                       Fecal Coliform
Density Range
per 1 ml
-10
11-30
31-100
101-300
301 - 500
501 - 1,000
> 1,000
Number of
Samples
1013
371
396
272
120
110
164
Occurrences
47
28
72
48
30
21
31
Percent
4.6
7.5
18.2
17.6
25.0
19.1
18.9
Occurrences
22
12
28
20
11
9
5
•« /\T
Percent
2.2
3.2
7.1
7.4
9.2
8.2
3.0
                               TOTAL
                  2446
                                                             277
                                        * Standard Plate Count (48 hrs. incubation, 35°C)
observed associated with the confluent growth, but distinct cohform colonies
did not develop which prevented quantitation  of  the coliform density  Un-
published data obtained during a study of this Northern  Kentucky distribu-
tion system  indicate a statistical significant  increase  of the non-coMorm
density after a 24-hour storage period.  The density increase appears to be
temperature  dependent and is greater during  the  summer months  when the
distribution sysLn water temperature  reaches 26°-27°C. This  increase in
non-coliform density occurs even though the free chlorine residual is main-
tained at a concentration usually above  1 nag/1 throughout the  distribution
system.                                                           ,   ,
  Limited  data is available relating  to  transit/ storage  time specifically for
potable waters/However, the data available does indicate that samples ex-
amined immediately  after  collection will provide  the most  valid  results
Since this is  impractical for potable water, data indicate that such samples
should be examined as soon as possible  after collection. The transit/ storage
period ranges from 6 to as many as 72 hours. However, 24 hours for potable
waters appears to be acceptable by the majority of  investigators while  a
few indicate  that 48 hours is also acceptable. All of the  studies  ^ indicate
that the transit/storage problem is  dependent upon a multitude of  factors
and the influence of these  factors, individually or  combined, will vary for
every water examined, be it stream, source or  potable.
  The bacteriological flora of a water sample will either  increase  or de-
crease upon  storage. Which change and  the degree of change are difficult to
predict and will vary with each  sample,  water supply and geographical  loca-
tion.  If these  changes, which are related to  transit/storage  result ra un-
acceptable data, .there are alternate steps  which should be considered. Samples
could be sent to laboratories nearer the water supply. Additional laboratories
or regional laboratories could be established. Alternate means of transporting
samples other than the mail service, such as bus  lines, private  courier or
United Parcel Service,  should be seriously investigated.  The delayed mem-
                               82

-------
                            SAMPLING COLLECTING AND HANDLING/H.D. NASH
  brane filter total coliform test could also be considered, however, on a case
  by  case  basis and not adopted as  a cure-all  procedure.
    Perhaps the transit/storage dilemma for  potable water cannot be resolved
  until research is conducted dealing only with potable water samples. Such
  a research project must address many types of water supplies, in many geo-
  graphical areas  to  obtain  results which  can resolve  the problem.  Until
  definitive data  are compiled,  it is  necessary to  make  recommendations
  based upon the data presently available, taking into consideration the con-
  flicting interpretations of that data  as applied to potable  water samples. A
  strong  recommendation is  made  that potable water  samples, be processed
  within  thirty hours after collection; .this is  a maximum transit/storage time
  period. If the laboratory is  required by State regulations to examine samples
  after 30  hours  and up  to  a 48-hour transit/storage period, the laboratory
  should indicate that the data may not be valid  because of excessive delay be-
  fore sample processing. Samples arriving for processing after a 48-hour period
  should  be refused without exception and a new sample requested.
                             REFERENCES

  1. American Public Health  Association.  1975. Standard method^ for the ex-
    amination of water and wastewater. 14th ed.  American Public Health Asso-
    ciation Inc., Washington,  D.C.                  '-        -
  2. Bordner, R.H., P.V. Scarpino, and J.A. Winter.  (Eds.).  1977.  (In  prepara-
    tion).  Microbiological methods for monitoring  the  environment,  part  I—
    water and waste. Environmental Monitoring and  Support  Laboratory, United
    States Environmental  Protection Agency, Cincinnati,  Ohio.
  3. Caldwell,  EX.,  and  L.W.  Parr.  1933.  Present  status  of  handling  water
    samples—comparison  of  bacteriological  analyses  under  varying  tempera-
    tures  and holding conditions  with si>ecial  reference  to  the direct  method.
    Amer. J. Public  Health 25:467-472.
  4. Cox, K.E., and  F.B. Claiborne.  1949. Effect of age and storage  tempera-
    ture on bacteriological water samples.  JAWWA  47:948-952.   ;
  5. Geldreich, E.E.  1975. Handbook for evaluating water bacteriological labora-
    tories. 2nd ed. United States Environmental Protection Agency; Washington,
    D.C.
  6. Geldreich, E.E., P.W, Kabler, H.L.  Jeter, and H.F. Clark.  1953. A delayed in-
    cubation  membrane  filter  test for  coliform  bacteria in  water, Amer.  J.
    Public  Health 45:1462-1474.
  7. Geldreich, E.E.,  H.D. Nash,  D.J.  Reasoner,  and R.H.  Taylor. 1972. The
    necessity  of  controlling bacterial populations" in  potable waters: Community
    water supply. JAWWA 64:596-602.
 8. Hoather,  R.C.  1952.  The  bacteriological  examination of  water.  J.  Inst.
    Water Eng.  61:426-442.                                     :
 9. Lonsame, B.K., N.M. Parhad, and N.U. Rao. 1967. Effect of storage  tempera-
    ture and time on the coliform  in water samples. Water Res. /:309-316.
10. Lucking, H.E.  1967. Death rate of coliform bacteria in stored Montana wa-
    ter samples. J. Environ. Health 29:576-580.
11. Public Health  Laboratory Service  Water  Sub-Committee.  1953.  The  effect
    of  sodium thiosulfate on  the  coliform  and bacterium coli counts  of non-
    chlorinated water samples.  J. Hyg. 51:572-577.
12. Snipe,  E.L. and  A. Fields.  1956.  Chelation as  a method for  maintaining
    the coliform index in water samples. Public Health Reports. 71:974-978.
13. Thomas, S.B., CA.  Scarlett, and W.A. Cuthbert.  1954. The  flaming of tap
    before sampling  of  the bacteriological  examination  of farm  water  supplies.
    J. Appl. Bacteriol.  77:175-181.                              :

                                                                         83

-------
ifcii'iiiffr,1'!11,  'SS!^                  '""
                                                                                                                                                                                                                                                                                                                     ii   ' ,"l<,i!!: 111:!

-------
                                      EPA METHODS MANUAL/R.H. BORDNER
      The  EPA Methods Manual  and the  Microbiological Standards

                   Robert H. Bordner and John A. Winter
             Environmental  Monitoring and  Support Laboratory
                   U.S. Environmental Protection Agency
                          Cincinnati, Ohio 45211

    Uniform application of the best  available methods is essential for generat-
  ing the valid  data  needed to  enforce water quality standards.  The  EPA
  methods manual, Microbiological Methods for Monitoring the Environment,
  Part I Water and Wastes (2) provides these methods in support of  the special
  needs of the Agency.
    The purpose of  this report  is  to  preview the microbiology manual in
  preparation, show how it supports the Agency's responsibility to monitor the
  quality of our Nation's waters, and demonstrate particularly how it performs
  this function for the National Interim Primary Drinking Water Regulations
  (NIPDWR).

  Responsibility for  Standardization

    The Federal Water Pollution Control Act Amendments of 1972, the Marine
  Protection, Research and Sanctuaries Act of  1972 (commonly referred to as
  the  Ocean Disposal  Act) and the Safe Drinking Water  Act of 1974 direct
  the Environmental Protection Agency to protect and improve the  quality  of
  our water resources and to control pollution. To achieve these goals  the Agen-
  cy must set, enforce and monitor standards for water supplies, ambient wa-
  ters  and wastewater  discharges. In EPA, the  Environmental Monitoring and
  Support Laboratory  (EMSL)—Cincinnati is responsible for selecting, vali-
  dating and standardizing analytical methods  for  water and for establishing
  quality control procedures.
    With the publication of the microbiological methods manual, EMSL—Cin-
  cinnati  has completed its mission to provide standardized methods  for chem-
  istry, biology and microbiology.  In  addition,  EMSL has prepared  a quality
  control handbook for water  and  wastewater laboratories.
    These manuals are not replacements for nor in competition with Standard
  Methods  for the Examination of  Water and Wastewater, (1) but rather
  provide the basic methods for EPA to carry out its enforcement responsi-
  bilities. The EPA microbiology  manual cites as valuable  sources of further
 detail and background information Standard Methods and EPA's Handbook
-for Evaluating  Water  Bacteriological Laboratories (3).

 Development of the Manual

    The standardization process  for  microbiological procedures began  at  a
 seminar held in San  Francisco January, 1973 where  the Agency's  needs for
 standard microbiological methods were agreed on and a  steering committee
 was formed to develop the EPA manual. It represented the Office of Research
 and Development, the Regional Offices, the Office of Water  Programs and
 the National Environmental  Investigations Center (Office  of  Enforcement).
 Committee members were assigned  reisponsibility for preparing sections on
 selected microbiological parameters.  First drafts were  presented and reviewed
 at  the second meeting of the  committee in January, 1974 at Cincinnati where
 the basic design, format and content were also established for the manual.

                                                           :           85

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 The editors, Robert Bordner and John Winter, EMSI^Cincinnati, developed,
 formatted and edited these materials and added technical  detail and infor-
 mation reflecting Agency policy. This draft was reviewed by EPA programs
 and Regional offices. The many constructive review comments, changes and
 corrections are now being incorporated into the final version of the  manual
 which is  scheduled for distribution in September, 1977.
 Legal Authority
    The manual will be the basic* reference lor  monitoring water and wastes
 in compliance with applicable water arid effluent standards established by the
 Agency.  Publication in the Federal Register will establish the legal authority
 for the methods used in the NPDWR program as was done for  the  pro-
 cedures  used  in  the National  Pollution  Discharge  Elimination  System
 (NPDES)  for  the analysis of effluents.
    The Code of Federal  Regulations  (CFR) Title 40, Part 136 (6)  lists the
 procedures required for Section 304 (g) of the  Federal Water Pollution Con-
 trol Act (NPDES). Table 1, an excerpt  from that Federal Register, shows
 the approved test procedures for bacterial parameters. References are made to
 Standard Methods  for the Examination of  Water and  Wastewater (1) and
 the methods of the U.S. Geological  Survey (5). When released, the EPA
 microbiology  manual will be  referenced in the column provided  in  the
 table. The amendments also  provide blanket approval  of alternate  test
 procedures by  applications to EMSI^-Cincinnati and for individual approval
 by application through the EPA Regions.
    It  is  anticipated  that  the  alternate  test  procedures  for the  Naional
 Primary Drinking Water Regulaions  will  be referenced similarly by amend-
 ment to the Federal Register.
 ,
I",
;,:,,,  ;
.  '•  '',','jff''
''    "'
          l.;:i
 Objectives  of  the Manual

   To carry out the  legal requirements  set forth in the three water quality
 acts  the EPA manual must provide:
 1. "fhe  best  available  'standardized  methods for  water quality  determina-
 tions, enforcement actions,  compliance  monitoring, resource planning and
 research  activities. ....... The procedures will be the reference  methods  for  the
 ' evaluation of modified ..... or'new meffiods.
 ^'^e'accompIls!imen^of the objectives of ..... trie manual will  assure the uni-
 form application of the most reliable procedures in all laboratories. Uniform
 procedures  facilitate  comparison, interchange   and  common use  of data
 in cooperative field surveys, interlaboratory studies and information storage
 and  retrieval systems.
 2, Methods that have  the general consensus of microbiologists within EPA.
 ]^[any methods are in use by researchers or by water analysts that have  not
 been proven by application in the various waters and geographical areas of
 this  country. These,  methods are not necessarily acceptable  for  uise in  the
                      '
 3. Methods that are tailor-made  to meet the  specific needs of the Agency.
 Procedures acceptable  for some agencies  may not meet  the needs of EPA.
 For example, if a decision is made for more stringent holding time require-
 ments  for potable water samples than has  previously been practiced, such
 specification will be put in the manual.
        I                               I I
 86

-------
                     TABLE 1.   LIST OF APPROVED TEST PROCEDURES FOR WASTEWATER *
Parameter and units Method EPA
4.
5.
6.
7.
8.
BACTERIA
Coliform (fecal)5, number MPN; membrane filter
per 100 ml.
Coliform (fecal)5, in pres- .. _ do 6- «
ence of chlorine, num- ;
berperlOOml. ;
Coliform (total)5, number 	 do6.
per 100 ml.
Coliform (total)5, in pres- MPN; "« membrane filter
ence of chlorine, num- with enrichment
ber per 100 ml. - 	 • I • •
Fecal streptococci5, num- MPN; « membrane filter;
ber per 100 ml. plate count.
r
References
(page nos.)
USGS
methods 14th ed. PT. 31 methods7
standard 1975
methods ASTM
	 922 	
937 45
	 922
	 928,937 	
	 916
	 928 	 35
	 916 	
933 	
	 943 	 	
. 	 944 	 5Q._ .
	 947 	
                                                                                                                              I
00
     5 The method used must be specified.
     6 The 5 tube MPN is used.
     7 Slack,  K.V.  and others,  "Methods for  Collection and  Analysis  of Aquatic
      Biological and Microbiological Samples: IIS. Geological Survey  Techniques  of
      Water-Resources Inv. book 5, ch. A4 (1973)".      :              icwiumues  m
     8 Since the membrane filter technique usually yields low and variable  recovery
      from chlorinated wastewaters,  the MPN method will be required  to resolve any
      controversies.

1 FEDERAL REGISTER, VOL. 40, NO. 232-40 CFR 136, DECEMBER 1,  1976
                                                                                                                              ffi
                                                                                                                              w

-------
                                  MICROB*OL STANDARDS EVALUAridN/c.w.
i	f:,;,
.111 ...... j ..... v.
                                  4 Procedures with limited options:  For enforcement purposes the analyst is
                                  required to use  one preferred' procedure whenever possible. At the First
                                  Microbiology Seminar on Standardization of Methods in San Francisco (4)
                                  in 1973 Jeter reported four variations of membrane filter tests and sixteen
                                  ways to conduct multiple tube tests  for total coliforms in Standard Methods.
                                  The EPA  manual must provide a  single, direct method wherever possible.,
                                  For example the use  of lactose broth for the MPN analysis of total coli-
                                  forms will not be included in (the EPA manual.              ,  ~ •*  •   fnr
                                  5 Validation of methods previous to publication in the  manual. Criteria  tor
                                  validation  are specified for nevf or additional methods to be included in  the
                                  manual. No methods will be added  to the manual without sufficient support-

                                  6D|'Quick response to changing needs of the Agency is possible by the single
                                  agency action of EPA. For example the Water Supply Laboratory Certifica-
                                  tion  Program requires criteria; and  procedures to  evaluate all  water supply
                                  laboratories. This very timely and  nationally-important information was in-
                                  serted inthe microbiology mamial without delay. If the Maximum Contami-
                                  nant Levels or methods of  establishing levels for total  coliforms  in drinking
                                  water should change,  the manual could easily reflect these changes.


                                  Special Features of the Manual
                                  •"'.•: '	, ,  '.-:' •  	-'. ,:::   ••  	": ;.••-,  ,-j	 „!	      	
                                   The EPA  manual features several special characteristics:
                                   1  A looseleaf style is planned'for the manual so that method improvements
                                   or additions can be  made easily.  For  example,  a recent development in
                                   the manufacture of "membrane filters produced a filter of larger pore size for
                                   microbiological analyses. A statement was added to the manual allowing the
                                   use of this filter, which is not permitted by Standard Methods. If the National
                                   Academy  of Science recommends  additional parameters for drinking water,
                                   for example the standard plate count, the methods would be added to the
                                   manual.                      I                                      f    '
                                   2. New editions are  planned as  demanded  to  keep  the contents ot the
                                   manual current These revisions can be published in as little as 18 months.
                                   As  improved test procedures are  developed and validated and other parame-
                                   ters are proved useful they will, be considered for the manual.
                                   3. The manual guides the 'analyst in selection  and use of the best method for
                                   his  needsi In response to the recent controversy concerning use of the mem-
                                   brane filter procedures for chlorinated effluents, the EPA manual limits the
                                   methodology.                                                   ...
                                   4. Important supporting information  for good  laboratory practice is in-
                                   cluded in  the manual. Examples  are  a section on quality control of micro-
                                   biological procedures,  an area jthat  has been largely ignored. Another section
                                   is devoted to  laboratory and field  safety. A  third applies basic statistics to
                                   microbiological analyses, and a fourth discusses legal considerations for the
                                    micfobiologist.                                                      .
                                   5. The manual will be widely distributed to Federal, State, local and private
                                   laboratories performing water"analyses'.
                                   6. it is written in a simple, direct, stepwise style that is  easily understandable
                                   and prepared hi a "cook book]' form so that it can be used at the bench by
                                   the analyst. The looseleaf form will also be conducive to use at the bench.
                                   7. The format assigns  a decimal number  to  each  paragraph  for   easy
                                   reference  and  location,  similar to the American  Society  for  Testing  and
                                   Materials  Books of Standards.
                                   88
ll|M
                 I!;,;;;!!	,,„!,	,JilK .; iiiilllil	Illiillll

-------
                                      EPA METHODS MANUAL/R.H. BORDNER
 Summary

    In summary, the EPA microbiological  methods  manual  fills the specific
 needs and legal requirements of the Agency! to standardize  methods in sup-
 port of its enforcement, compliance monitoring, ambient monitoring and re-
 search programs.  The methods chosen are the best available,  allow limited
 options and represent  the  consensus of EPA microbiologists. The manual
 format allows  for rapid  response  to changing Agency  needs,; offers  oppor-
 tunity for frequent revisions, guides the analy'st in the selection  and use of the
 best methods for his/her  purposes  and features a stepwise "cook book" form
 for application at the bench level.          j
                            REFERENCES

  1. American Public Health Association. 1975. Sjtandard methods for the examina-
    tion of water and wastewater. 14th ed. The American Public Health Associa-
    tion.
  2. Bordner, R.H., P.V. Scarpino and  J,A. Winter  (Eds.).  1977  (In prepara-
    tion). Microbiological methods for monitoring the  environment, part I—
    water and wastes.  Environmental  Monitoring  and Support Laboratory, U.S.
    Environmental Protection Agency, Cincinnati* Ohio.
  3. Geldreich, E.E. 1975. Handbook for evaluating water bacteriological labora-
    tories, EPA  No. 670/9-75-006.  2nd ed. Municipal Environmental Research
    Laboratory,  Office  of Research  and Development,  United States Environ-
    mental Protection Agency,  Cincinnati, Ohio..
  4. Jeter, H.L.  1973. Total coliforms, p. 20-24.f Proceedings of the First Micro-
    biology Seminar  on Standardization of Methods, EPA No. R4-73-022. Office
    of  Research  and  Development,  United States   Environmental  Protection
    Agency,  Washington, D.C.                 ;
  5. Slack, K.V.  et al.  1973.  Methods  for  collection and  analysis  of  aquatic
    biological and microbiological  samples, Book 5, chapter A-4.  In:  Tech-
    niques  of water-resources  investigations of  the   United  States  Geological
    Survey. United States Printing Office, Washington, D.C.      i
  6. United States  Environmental Protection  Agency.  1976.  Guidelines  estab-
    lishing test procedures for the analysis of /pollutants—amendments. Federal
    Register  40:52780-52786.                  {


             QUESTION  AND ANSWER  SESSION'
                                            i
C.W; Hendricks, Office of Drinking Water,  U.S. Environmental Protection
Agency, Washington,  D.C.                  :

When will the manual be available?          '

R.H.  Bordner, Environmental Support Laboratory,  U.S. Environmental  Pro-
tection Agency, Cincinnati, Ohio             j,
                                            I          "       '      "
The target date for printing is September, 1977. The manual should be avail-
able by the end of the calendar year.
                                            i'
S.M.  Morrison,  Department  of Microbiology,  Colorado  State  University,
Fort Collins, Colorado                      1                 :

In response  to the Drinking Water Law many[ laboratories are buying equip-
ment  and instrumentation and  preparing  to jperform tests  they  have never
done  before  with  technicians  who are completely unfamiliar with the tests

-------
1	ilff,	f'f	]>T
ill! ........... i ......... ....... • ...... am ...... y;:im .......... ~
       •w ....... » ....... ; ................... w ...... wm-im ..... HIT ...... rw,1*:™ ..... r: .......... • ...... ..... rsis ..... «! ....... wmx ...... IWSHKI? ........... iis?. ........... '
               i!	!!;
	'!	i	I	HBW	•	Hill
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKSi
and  equipment.  This is particularly  true of the smaller laboratories. Have
any  provisions been made for this problem?
 " ..... '  ..    .         .    .     1  ................. il ...........
R.H. Bordner

EPA recognizes the need for training as an important part of the program to
upgrade the laboratories. Training by the States is certainly encouraged; to
pur knowledge funding has not been provided by EPA.
 "' ........... ' .......... .............                  •                    !
S.M. Morrison

I would like to see the use of regional of district laboratories which do have
this  expertise encouraged.

W. Litsky, Department of Environmental Sciences, University of Massachu-
setts, Amherst, Massachusetts

You described  briefly the relationship of  the EPA  methods  to  Standard
Methods.  How do  the methods in the EPA manual relate to those being
developed by other organizations such  as ASTM (the American Society for
Testing and Materials)?
                      ......  •  !  .................................. ]•
R.H. Bordner
EPA members  are  participating  in  the  activities of  ASTM  as  well  as
Standard Methods and .other methods — standardizing groups. The methods
must be reviewed by the EPA Microbiological  Methods Steering  Commit-
tee.  Criteria  for  inclusion of new methods  in  the  manual, in the future
state that the method should satisfy the needs of the Agency,  show some
advantages over  present methods, be validated by the  developer and other
laboratories, be relatively easy to use and practical, etc.
                     , ,    M        ..... ,,   *  ...... ',;,;,,," ,  ',,!,, J| ,   .....       ,    , „

Q,W. Hendricks!               j
You mentioned internal review j of the  manual.' Are "there provisions for the
manual to be  reviewed by  experts  in  aquatic microbiology  outside  the
Agency?                      !
  *,  ""-     ,   ,               .. ................. ,• ........   ; .......              ,
R.H.         '       '         .................... ~' ........................ " ............... ' ................ ......... ....................... ' ................ '' ...... ' ....... ................. ...... ........... ..................................... ..........................
                                 We dp plan pn an external review.
                                 90

-------
                                     STANDARD METHODS/A.E. GREENBERG


        Evaluation of Standard Procedures:  Standard  Methods

                          Arnold £. Greenfoerg
                  Bioenvironmental Laboratories Section
                    California  Department of  Health
                       Berkeley, Caltfornia 94704

  In  1905, after almost ten years  of deliberations and discussions, the
American Public Health Association published  the first edition of what  is
now called Standard Methods for the Examination of Water and Wastewater.
It was a book devoted largely to the enumeration and identification of bac-
teria in water but it did include methods for chemical analysis  as well. The
intervening 72 years have seen  tremenous expansion of the  manual, wide-
spread legal  or quasi-legal  acceptance of the  methods it contains, and  a
total of fourteen  editions.  The 15th edition is in preparation an;d expected to
be available by 1978.
  Although APHA has  continued as the publisher,  it was  joined  by the
American Water  Works Association in 1925 and in 1947 by the Water Pol-
lution Control Federation in preparing this widely used and widely accepted
manual of methods. The three  sponsoring societies have different internal
procedures for dealing with the administrative and  technical  details of pro-
ducing Standard  Methods but all of these are  coordinated by  a Joint Edi-
torial Board that  consists of the  chairman of each society's standard methods
committee. Within this framework, APHA has  major responsibility for the
sections on microbiology.
  The current edition includes 131 pages (out of a total of 1193) devoted to
the bacteriological examination of water and wastewater. Principal emphasis is
given to counting organisms indicative of fecal  pollution such as total coli-
form  bacteria,  fecal coliforms,  and fecal  streptococci by  either a multiple
tube dilution  or  a membrane  filter technique  and to enumeration  of the
total bacterial count. Brief  sections are devoted to detection of pathogens
such as Salmonella, Shigella, enteropathogenic E. coli, and leptospires. Tenta-
tive procedures are included for viruses  in water, Staphylococcus and Pseu-
domonas  aeruginosa in swimming  poods, fungi, yeasts, actinomycetes, and
nematodes. A section on the identification of nuisance iron and sulfur bac*
teria is included  also.
  The sections devoted  to  indicator organisms do not emphasize a  strict
taxonomic approach to identification since that has never been considered of
primary importance in public health. Tfcius, for example, methods for Escheri-
chia coli or Enterobacter aerogenes are much less important than procedures
for fecal  coliforms which are defined as coliform bacteria capable of pro-
ducing gas from lactose at 44.5° C. This difference stresses  the pragmatic ap-
proach to counting organisms originating in the  gut of warmblooded animals
rather than specific identifications.                           :
  On both a  national and  an international level in recent years there has
been  a proliferation of organized interest in standardization  of methods hi
aquatic microbiology. In the United States, as we have  heard, the Environ-
mental Protection Agency is in process of preparing a manual of bacteriologi-
cal  methods  that  will  be a companion  volume to their  already published
Methods for Chemical Analysis of Water  and Wastes (EPA-625-/6-74-003,
1974)  and  Biological Field and Laboratory Methods  for Measuring the
Quality of Surface Waters and Effluents (EPA-670/4-73-001, 1973). It has
been argued that EPA as a regulatory agency requires procedures that will

                                                                      91

-------
IE il1!1:*1 :' .1 • I
!l!i lilritt i   II	'
If" 111* •'.'  •  I "• ("I1
 I ill 1
               
-------
                                    STANDARD METHODS/A.E. GREENBERG


Joint Task Group responsible for preparing  the section on quality control,
the remedy is obvious  and immediately at hand.
  In  conclusion, the Water Supply Program  of  EPA  has  available to it
the best  efforts  of bacteriologists, including  those from EPA, in providing
analytical methods  for evaluating  microbiology  standards. These exist  in
Standard Methods for  the  Examination of  Water  and Waste-water  which
should be adopted as  the  sole methods source. Exclusion of unacceptable
alternative .methods can be achieved by appropriate page or section citations
in EPA  regulations.
                                                                    93

-------
 II! 1't.ir
 1  g1	lull,"    I    I,




 Bi&J I'  f  ii
111 11)1!':'     I    i
a, 'illlilllll I1!1"  i   !" »'i,  !!," !" ''           if >i

.mi1 mil'',.'.  '".i  t:'i           at:1'
"!• tih   •  5 : i', •';i;        •    HI
it IIP     "iinii  ;i,i	B"1:
                                                         :'l	/'lull'      /'Illli!
 ." s     i!|i:.ii!
 ,  ill, III"     •',, ijlll
''Kir     ",!<: is. ?
                                                                                                                                                                                                                                              1*",);1-, iihii'ii   'ii•"'<', •
                                                                                                                                                                                                                                                   'i .1 ;-,iiilr   W '	i'JH	I
                                                                                                                                                                                                                                                   i Hill",  i,	ill.  .'• "-!"M

                                                                                                                                                                                                                                                    • Hffi1^	:t	;>:'!	I

-------
                                              VIRUS DETECTION/D.O. COVER
                             Virus Detection
                              Dean O. Cliver

                          Food Research Institute
                          University of Wisconsin
                        Madison, Wisconsin  53706
Field Virus Survey Unit (US Environmental Agency)
   I have been assigned the topic of "Virus Detection", as it relates to Evalua-
tion of the Microbiology Standards for Drinking Water.  I  plan to interpret
this charge rather liberally.                                  ,
   I shall begin  by noting that the purpose of standards is  to .protect public
health, that viruses are a class of pathogens sometimes transmitted by water,
and that I have  been unable to find the; word "virus" in the  National Interim
Primary Drinking Water Regulations.  In the spirit of  the  Regulations,  I
had better define the term before I use it further. The viruses share three most
important  properties:
   1. They are transmitted from host to  host in the form of discrete particles
     too small to be seen- with a light microscope.
   2. They multiply only inside suitable  living host cells.
   3. They  cause mysterious,  incurable diseases  (in  the sense  that when
     your physician  tells you, you have a  "virus", he is saying he doesn't
     think he can do  anything for you).  Fortunately, most individuals, if
     they are otherwise healthy, recover from most virus infections without
     any  substantive  medical assistance.
  The only virus disease proved to be 'transmitted currently in  water in  the
United States is hepatitis A, formerly known  as  infectious hepatitis (11).

                                                                       95

-------
•i Jsif,  ::' I
                       i '1,,, !	if 1,,
                              • Hi
 ll Bill1 ,  li"i 111 I,,, I
I'lS   ri.si	•
    «i	life
I;,11 Ulln , , i I'"1 , II. M.I i',I. .,<

li:l'?l:tl&
is ill'''", i .. i!' '•*•:
                          •i:l! li  I,1	ill
                          i
               	ill
               IIP
                          	'Hill  i	I
  Illllll
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 Still some of the gastroenteritis which results  from drinking contaminated
 water is probably caused by virus (2) and many viruses known to cause other
 diseases  are  detected in sewage-polluted waters (10).  I don't want  to be-
 labor the point, but I think it is worth noting that all of the viruses thought
 to be transmissible in drinking water are shed in human feces, and that these
 virus-containing feces have been diluted by a factor of approximately 2000
 as  they  occur in domestic  wastewater. Therefore,  virtually all threat oi
 virus transmission through drinking water could be eliminated by perfecting
 and requiring the use  of  some ^of the alternatives to common water-flush
 toilet (also known as the "W.C.").
    We can assume that this is unlikely to happen in the near future and that
 viruses will continue to occur in raw water sources supplying some com-
 munity systems,  I'ougnt,  then, to point out that conventional water treat-
 ment techniques were  not specifically  developed to remove or inactivate
 viruses, and that viruses have a perverse tendency to adhere to each other (5)
 y wM as to suspended solids in water (12). These factors, and the  relative
 resistance of viruses to chemical disinfectants (9), would  appear to compli-
 cate the  task of producing virus-free drinking water from  contaminated raw
 water or from wastewater. Still, known incidents of virus  disease associated
 ..              "water have resulted  either from chronic under-treatment of
                was ".known to be contaminated, or. from cross-contamination
                g^^'s'^f^^^ .............. virus-containing  wastewater into  other-
 ••'jjjyjgg safe finished/ water during distribution  (4). Rather than viewing vkus
 " if ansmission ...... through ...... water as an exotic problem requiring new  and unprece-
 lente! countermeasuresT I suggest that  it  be seen as another potential con-
 |feguence ........ of making some of the same  old mistakes. Such lapses are infre-
 quent  and are not equally probable:  under-treatment  has most often been
 seen to  occur  in private or semi-public water systems; incidents  involving
 community water supplies have  most  frequently entailed  recontamination
 of the finished water during  distribution (4).
    One would like to have some btHer meahs than human disease to  detect
 the presence of  virus in  drinking  water. The task of recovering  virus  is
 complicated  by the  extremely  small quantity of virus material  that is likely
 to be present in a water sample and  by the tendency  of  the virus  particles
 to  adsorb to suspended solids (13).  However, quite elegant methods have
 been devised for ' recovering virus particles  from  samples  as large  as 1900
 liters (~ """"500  gallons) or greater.  The virus particles  in the  concentrated
 ||^pe g-g«| g— st-fl to be detected by the infections they may produce in
 tissue cultiires  (3).  Such infections become perceptible only slowly:  inocu-
 lated tissue cultures must usually be  Incubated  and observed for days or
 weeks. The potential for success varies: no tissue culture is known  to sup-
 port replication  of  the  hepatitis  A virus,  but  tissue cultures  are probably
 incp sensitive than the host's body to some of  the other intestinal  viruses
•''wnicn' ' imglit" occur "in water  (Oliver,  unpublished data).  Although, there
 will always  be room  for improvement in  such techniques, we have now
 reacned  the point  where a  "Tentative  Standard Method" for detecting
 viruses in water exists  (3). The method has been shown  to recover  experi-
 mentally-inoculated  viruses efficiently from large-volume  samples of water
  (6) . No virus  has been detected in approximately 175  large-volume samples
 of finished community drinking water by  this method, even though the raw
 water sources of some of these communities were known to be highly con-
 taminated (1; E.W. Akin, EPA Cincinnati, personal  communication).
    The most obvious explanation  for  these negative test results is that vkus
 was absent from the samples, although  other reasons can be envisioned.
              i          1    I       II        I
                                  96

-------
                                            VIRUS DETECTION/D.O. CUVER

   Because it is difficult to detect viruses in water, it is  reasonable to seek
"indicators"  (organisms or substances whose presence  in water is likely to
accompany that of virus). In fact, there are  many possible indicator systems:
Coliforms, fecal coliforms, and  fecal streptococci suggest fecal contamina-
tion of water; however,  they  differ from viruses in the attrition they are
likely to undergo in  the  environment or during water  treatment. Turbidity,
standard plate count, and free  residual chlorine  (or lack of  it) have, at
best, an indirect association with the  likelihood that virus has contaminated
water. Coliphages (bacteriophages specifically infectious  for Escherichia coli)
as a group are indicative  of fecal contamination of water  and  may lose their
infectivity at a rate comparable to animal viruses (8); however, there seems
to be some problem  in selecting a  "best" host strain  of E. coll in which to
detect the coliphages (14).
   Finally,, the vaccine strains of the human  polioviruses have  been proposed
as indicators because they are almost always present in  urban sewage, are
likely to lose infectivity at approximately the same rates  as many other in-
testinal  viruses,  and may  be more rapidly detectable  than viruses in  general,
using a fluorescent antibody test procedure  (7). Unfortunately, polioviruses
are not  always present in sewage, and the detection methods are a good deal
more complex and  exacting than  most Standard Methods microbiological
procedures.
   By now, you  are probably wondering how it is ever possible to determine
that a community's drinking water does  not contain virus. In fact, it  is not
possible to demonstrate the complete  absence of virus from each day's total
production of finished water. The most direct test of the safety of drinking
water is drinking the water. However, given the twd-liter-per-day consumption
figure upon which the Maximum Contaminant  Levels are  supposed  to  be
based, only about 1 % of the total water  used at home (and, therefore, even
less of  the daily per  capita total water consumption)  is. drunk.. The rest is
used in a variety of ways that  are unlikely to  allow manifestation of any
virus it  contains.
   The  water could  also  be tested  for  virus by  the procedures mentioned
previously. I believe  that the  surveys now  in progress represent the most
proper use of the Tentative Standard Method and its counterparts—these are
not procedures  appropriate for  routine  quality control  of finished waters
by individual suppliers; neither  can they reasonably  serve as the basis for
establishing or enforcing Maximum  Contaminant Levels  for  viruses.  What
the survey results seem to show is that virus-free water is attainable by means
of the  best current treatment methods,  if these  are  conscientiously  applied,
even when raw  water sources are far from ideal. Less rigorous  treatment is
likely to be necessary with high-quality, protected  raw water sources. The
evidence indicates that  the means to produce .virus-free  finished  water are
at hand, at least for raw waters that do not consist principally of sewage
effluent. The most important task at this time is to ensure that  these means
are employed, absoluely without fail,  by  water suppliers.  Whether virus
testing  involves  one  "large" sample or several small ones per  day,  only a
minute  portion of a community's daily water output  could be monitored in
this way, at a cost in resources that  could better be applied to other surveil-
lance tasks. The  findings  of the EPA community water survey indicate that
the results  of such testing are  likely to be negative,  but this obviously does
not guarantee that the. consumer is receiving virus-free drinking water.
   I believe that the most consumer  protection that  can be obtained for the
consumer's dollar will derive from means already at hand. Suppliers should
be helped in  selecting treatments appropriate to the  quality of their raw wa-

                                                                       97

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
ter  For any source other than extremely well-protected ground water, treat-
ment  should probably include  at least  the following:  1)  coagulation  and
sedimentation,  2)  filtration, and 3)  chemical disinfection  (I  am not  now
prepared to recommend treatments for direct reclamation of urban effluents).
The physical: treatments which  are  suggested should serve to remove proto-
zoan cysts, a very real cause for concern, as well as most solids  that might
protect virus from chemical disinfection. Beyond selection and adaptation of
treatments  appropriate to the quality of the raw water, adequate consumer
protection  requires that  the supplier  by  supervised to ensure that the treat-
ments are properly performed all of  the time. Microbiological testing of the
finished product alone won't do the job.  There must be monitoring  of  both
microorganisms and turbidity, as welt as  a continuous record of the proces-
sing as it Is done. Then the quality of the finished water must be sustained all
the way to the user's home. The presence of a free chlorine residual, or the
absence of coliform organisms,  at the tap is  reasonable evidence  that cross-
contamination  or  back-siphonage has not occurred to  degrade the  finisnecl
water during distribution. '	_	_.  _	  ' .'"
   I believe that this discussion supports a  few  conclusions:  First,  it is
reasonable to  test'for".viruses"in finished drinking water on occasion, but it
is not 'reasonable to expect drinking water suppliers to-do it. Second, Maxi-
;m;gm*'contaminant levels  (as  defined in the Safe  Drinking Water Act of
 1974), are not appropriate  for  viruses. Finally,  the basis for protecting the
public from water-borne virus  disease lies rather  in detecting lapses in the
practices of treatment and distribution than  in  detecting  the virus in the
^finished product.


                ;..   .-.  •  •  ; REFERENCES       '    '  ,
  1. Akin, E.W., D.A.  Brashear, and  N.A. Clarke. 1975.  A  virus-in-water study
    Of finished water from six  communities. Health Effects Research Laboratory,
    Office of Research and Development, United States  Environmental  Protection
    Agency, Cincinnati, Ohio.
  2. Albrey,  M.B.,  and A.M.  Murphy.  1976.  Rotaviruses  and  acute  gastroen-
    teritis of infants and  children.  Med. J. Aust.  7:82-85.
  3 American  Public  Health  Association.  1976.  Section   913:  Detection  of
    enteric viruses  in water and wastewater,  pp. 968-975. In: Standard methods
    for the examination of water and wastewater. 14th ed. American Public Health
    Association  Inc., Washington,  D.C.
  4 Craun, G.F., and  L.J. McCabe. 1973.  Review of the causes of waterborne-
   ' disease outbreaks.  JAWWA. 65:74-84.                              .
  5. Floyd™ R., and D.G. Sharp. 1977.  Aggregation of poliovirus and reovirus by
    dilution in water. Appl. Environ.  Microbiol. 55:159-167.
  6. Hill, WK W. lakubowski, E.W. Akin, and N.A. Clarke. 1976. Detection  of
    virus  in  water: sensitivity of  the tentative  standard  method  for drinking
    water. J. Appl. Environ. Microbiol. 31:254-261.
  7. Katzenelsoii, E. 1976. A rapid method for quantitative  assay .of  poliovirus
    frpm  water  with the  aid of the fluorescent antibody technique. Arch. Virol.
    50:197-206.   •	  t	 . •" 	
 •g. Kott,  Y., R. Netta, S. Sperber, and N. Betzer. 1974. Bactenophages as  viral
    pollution indicators. Water Res. 8:165-171.
  9. Liu, O.C., H.R. Seraichekas, E.W.  Akin, D.A.  Brashear, E.L. Katz, and  WJ.
    Hill,  Jr.  1971. Relative resistance .of  twenty  human  enteric viruses to free
    chlorine in  Potomac  water, pp. 171-195. In:  Snoeyink,  V., and  V.  Griffin
    (Eds.), Virus  and  water quality:  occurrence and  control. Proceedings of the
    Thirteenth  Water  Quality  Conference. University of Illinois Bulletin,  Vol.
    69, No. 1, Urbana, Illinois.

 98

-------
                                               VIRUS DETECTION/D.O. COVER
10. Metcalf,  T.G.,  C. Wallis, and  J.L.  Melnick. 1974.  Virus enumeration and
    public  health  assessments in polluted surface water contributing  to trans-
    mission of virus in nature, pp. 57-83,, In: Virus survival in water and waste-
    water systems.  Center for Research  in Water Resources,  The University of
    Texas,  Austin, Texas.
11. Mosley, J.W. 1967.  Transmission of Viral Diseases by Drinking Water, pp.
    5-23. In: Transmission of viruses by the  water route. Interscience Publishers,
    New York.
12. Schaub,  S.A.,  and  C.A.  Sorber.  1976.  Viruses  on solids  ,in  water, pp.
    128-138.  In: Viruses  in  water.  American  Public  Health Association Inc.,
    Washington, D.C.
13. Sobsey, M.D. 1976. Methods for  detecting enteric viruses in water and waste-
    water,  pp.  89-127.  In:  G. Berg et al.  (Eds.),  Viruses  in water. American
    Public  Health Association Inc.,  Washington, D.C.
14. Vaughn,  J.M.,  and T.G.  Metcalf.  1975.  Coliphages as indicators  of enteric
    viruses  in shellfish and shellfish raising estuarine waters. Water Res. 9:613-616.
                                                                           99

-------
                                                                                                                                                                                                      i.   ','!, I1!1!1!*'  "mi  .ill< i' !t    	        ii.  i.l*  ',    Hi''1"  I'i  "I	!Fi
                                                                                                                                                                                                                                                                                                                                                           '"'II ,'      "  111 III1    i'1 II"    II1  Wl;lll
'' :  I'll

-------
                     DISINFECTANT RESISTANT ORGANISMS/P.T.B. SHAFFER
                   Disinfectant  Resistant Organisms *

                           Peter T. B. Shaffer
                         Carbo.rundum Company
                     Niagara Falls, New York 14302

  Much of the dialogue at this symposium has been based on  laboratory
disinfection studies in which  the die off  rates of standard strains  of viruses
have  been measured as  a  function  of time  at  various  levels of  chlorine
residuals. While this specific  information is technically sound,  it overlooks
the variations  in  resistance to  chlorine  inactivation  among various strains
of viruses or among individual  members of a specific virus strain.  Thus,  it
becomes questionable to measure  a  chlorine residual and  assume that this
will assure a given level of disinfection of all viruses. Any  indirect standard
must be viewed with some misgivings. If a water is to be assured to be free of
a particular organism, that organism must be monitored even if infrequently.
  Virus  isolates recovered  from well-treated finished  water (2) have  been
questioned, to  no  small extent on the  basis that they were recovered from a
water of very  low turbidity in  the presence  of chlorine  residuals in excess
of one milligram  per liter  (1). We have examined isolates from one  such
study in considerable detail and  are convinced that they  are sufficiently dif-
ferent from any available laboratory strains to assure that  they are natural
isolates  and not artifacts resulting  from contamination (3).
  Most recently,  two of these isolates were  grown to high titers, carefully
purified by sedimentation in an ultracentrifuge, and subjected to chlorine in-
activation.  Their   inactivation  rates  were  compared with  those  of  two
standard poliovirus 1 strains  (LSc and Mahoney) grown  up and purified  in
the same way. They showed a  chlorine inactivation  rate orders  of magni-
tude  lower than  that of either of the  two standard controls. These data
shown,  in part below, will be published in detail  shortly.
VIRUS LOG SURVIVAL
STRAIN FRACTION (30 MINUTES)
Mahoney
LSc
Isolate 1
Isolate 2
-4.6
-7.79
-1.74
- -3.83
FREE CHLORINE RESIDUAL
(mgll)
0.39
1.34
0.92
1.31
  The point to be noted is that  all viruses,  in  fact all  strains, or even  all
members within a  single standard strain, do not respond  in ,the same way
when subjected to  disinfection conditions typical of those  applied hi a well
operated drinking  water plant. To  assume that all viruses will  exhibit the
same die off rate  as that shown by standard laboratory strains is not tech-
nically sound. Therefore, to replace  a direct standard, whether it be coliform
or virus, with an indirect standard of turbidity and  chlorine residual can not
be defended on the basis of experimental evidence.
   * This paper was submitted after the Conference.

                                                                      101

-------
                                    MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                               REFERENCES
      in. ..... i';
     ''   '
nil	ilH  i, II '!•!:,;•
 liiiii."iii/iiii1' i  ' I '•,'" I".''

 lili'MflllNIIUI!1 i i " ill,  n/f
fi JH"i> : i , 1
                 1!'
Hf'!!'
1	'"31
1. Akin, E.W.  and W.  Jakubowski. 1976.  Viruses in finished water—the Occo-
 ' quan experience. Proceedings AWWA Water  Quality technology Conference,
   San Diego, California.
   Hoehn",  R;C., C.W. Randall, F.A. Bell  and  P.T.B. Shaffer. Trihalornethanes
   and viruses in a water supply. J. Environ. Eng. (to be published).
   Shaffer^  P;Y.B., .RJE.  Meierer and C.D. McGee.  1976.  Isolation  of natural
   viruses from a  variety of waters. Proceedings AWWA Water Quality Tech-
   nology Conference, San Diego,  California.
4, %|ffer;  'P;T.B., "T.G.  Metcalf,  O.J. Sproul  and  D.G.  Sharp,  Chlorine  re-.
   sistant poliovirus  from finished drinking water (to be published).
                                              ",• 	!!
                            ' li'lll  N "Ml,,,',	,|
)  lifl
lit'ill	„
          i'(>"   ,  «;<;
                                            i „!;;,: ,.  '   j.. .,;.,

                                            !,"': '' *l  '   ' "i1
                                                                                                                 l|l "U;1!'!!' 'Vlr   "	ill
                                    102

-------
                                   TURBIDITY AND DISINFECTION/J.C. HOFF
     The Relationship of Turbidity to Disinfection of Potable Wafer

                              John C. HoS
                 Water  Supply Research Division/MERL
                  U.S. Environmental Protection Agency
                         Cincinnati, Ohio  45268
Introduction

  Although turbidity has been used  as a  measure of potable water quality
for many years, increased attention has been focused on this  parameter as a
result  of its inclusion. at a  Maximum Contaminant  Level  (MCL)  of 1
Nephelometric  Turbidity Unit  (NTU)  in  the National  Interim  Primary
Drinking Water Regulations. Turbidity in drinking water can  be caused by a
large variety of particles which can originate from many sources. Turbidity
may be present as a result of incomplete removal by treatment! processes of
particles present in the raw water source or may result from problems with
water treatment techniques such as coagulation. Turbidity may also originate
within the distribution system as a  result of corrosion, growth of micro-
organisms, line repairs, etc.                                  ;
  Because of the current emphasis on turbidity in potable water, the signif-
icance of turbidity, particularly as it  relates  to  disinfection  efficiency during
water treatment, will be discussed from  the  historical point of view as well
as in the light of recently developed information. Problems associated with the
use  of  turbidity as  a contaminant limit  and some possible  alternatives for
future consideration also will be discussed.

Turbidity as a Potable Water Standard

  Turbidity in water is caused by the presence of suspended matter such as
clay, silt,  finely divided  organic  and  inorganic matter  and;  microscopic
microorganisms. Turbidity is defined and expressed as an optical property
of water and is measured by determining the degree of light scattering by
the particulates present  in  the  sample (1). Because of the nature of the
measurement,  turbidity  determinations  cannot be  related  either  to the
quantity (mg/1)  or characeristics  (size, shape, chemical nature)  of the
particles present in  the  sample.  Another problem is that different types of
turbidimeters, after being standardized on the same turbidity standard, will
usually give readings differing by as  much as  500% when used to measure
the  turbidity of a sample containing another  turbidity causing substance
(12).
  Despite these limitations, turbidity has been found to  be a useful parameter
for measuring the quality of public water supplies. Turbidity first became a
part of  the  USPHS  Drinking Water Standards in 1942 when  a tolerance
limit of 10 units was established. In 1946 this was  changed to a tolerance
limit of  10 for filtered water only and a recommended limit of 10 for other
waters. In 1962, the recommended limit was reduced to 5 and the tolerance
limit was dropped  (2).  The most recent  change embodied hvthe National
Interim Primary Drinking Water  Regulations establishes a turbidity Maximum
Contaminant Level (MCL) of 1 with up  to 5 units allowed if the supplier
can  demonstrate that the  higher limit does not do any of the following:
   1. Interfere with  disinfection.

                                                                      103

-------
rvnri	(Fffn	'•	s	rvt
                                                               v-wsHcmf/n	IK1	nil	iisii
1' Mi :'••['
|H 'lilll
ll llll''1";: :,:
I"!	I	if I/:
               I'1::"';,!: ,•.;"•,  ;;';	SI;
...... ill
 lli'ii:ii
toiiEL
("I i,   (h I) i
iiT't: •!'
                           i ill n i  i	in i
I! IB	  I  Mi:	Ill
iiif
 i Mi1"
  II
lii ("IP1"!, ' 'II
                                                                               	•[ '::,! I • '/"	•>*•'"!
                                                                                ''si'.;]* ;;;i, , •-«)
                                                                                 ' ill' ill i|U". .1 Ji,! hi" ill"!
                           ^' ^ "  MICROBIOLSTANpiiLpSE^ALUATI6^^^
        2.  Prevent maintenance of effective disinfection throughout the distribution
           system.
        3.  Interfere with microbiological determinations.
        Of the three, the most criti«d'and'^Vpne'_which has generated most con-
     cern is the possible interference of turbidity with disinfection efficiency. The
     second criterion relates  to  complex problems  of turbidity as  a source of
     'nutrients"and available surface for growth of microorganisms in distribution
     lines, as "well as the more directrelationshipbetweenturbidity and disinfectant
     demand™ Significancei"of'the "tKird 'cnterion'''h"as'been"shown at high turbidities
     but an efiect in the turbidity range of 1 to 5 has not been clearly demonstrated.

     Theoretical Aspects of the Relationship Between Turbidity and
     Disinfection
        For disinfection to be effective, there must be contact between the agent and
     the organisms that  the disinfectant is to kill. Some  of the particles  which
     make up part of the turbidity may be particle-microorganism complexes in
     which a microorganism is adsorbed onto a larger particle.  In other cases one
     or more smaller particles may "be "adsorbed onto a microorganism or a par-
     ticle may completely encase a microorganism. In each of these cases varying
     degrees of protection of the  microorganism  from the  disinfectant  may  be
     expected.      	      	[
       The microorganisms of concern in this respect include pathogenic protozoa,
     bacteria and viruses. SeVeral characteristics of these groups  of microorganisms
     must be considered  relative to possible effects  of turbidity on  their inactiva-
     tion. One of these is size. The sizes of representatives of  these three groups
     differ by a factor of over 100X (poliovirus 0.03 /^m diameter,  E.  coli 0.3  X
     1-5 /im, E. histolytica  cysts 5-20 /*m). This would indicate that  in the case
     of virus protection  by  turbidity, contact between the disinfectant  and the
     microorganism eould be prevented  by much  smaller  complexes than would
     be needed for protection  of enterobacteria or protozoans. Size  is also im-
     portantin relationship to the significance of surface charge  phenomena in
     causing  aggregation  or adsorption of  particles. These  factors  become in-
     creasingly  important as size decreases  and are very  important with regard
     to particles in the size range of enteroviruses.  Another factor  of  importance
     is the relative innate  disinfection  resistance  of the  microorganisms  them-
     selves. In general, the  enterobacteria are  the  most sensitive to disinfectants',
     followed by the enteroviruses, while the  protozoans  exhibit the greatest re-
     sistance  (19).
       Based on these factors, several assumptions regarding  turbidity  effects can
     be made. In order to protect ah enterovirus from disinfectant  action, a par-
     ticle, whether  produced by adsorption of many extremely small particles onto
     one virus particle or from dissolution of fecal  or other material to form
     small particles some of which may  completely enclose a virus  particle, must
     be larger than 0.03 ju,m hi diameter. Particles which will protect  bacteria  or
     cysts musf be correspondingly larger. Thus,  the particles present in  water
     which  are  of  concern because of possible disinfection efficiency  effects are
     those ranging  upwards from 0.03 /xm.
     "... ^Tffie^extensive} body  of  literature  on  the physio-chemical mechanisms
     involved in virus  adsorption onto surfaces and the influence  of these inter-
     actions  on the fate of  viruses in the environment was recently reviewed by
     Britton   (5).   The  adsorption  phenomenon  is  used  for  virus  removal
     in various waste water and water treatment processes  (coagulation, floccula-
                                   104

-------
                                    TURBIDITY AND DISINFECTION/J.C. HOFF


  tion, filtration and also is applied in methods developed for concentration
  and detection of low levels of vkus from large volumes of water. Because of
  the way in which viruses, particularly enteroviruses, are produced,  excreted,
  and transported (intestinal tract —^  feces —> water)  and the number and
  variety of particles present in natural waters it is likely that many entero-
  viruses exist in the environment as adsorbed, rather than free, particles (4).

  Results of Research on Protective Effects of Turbidity

    Although the concept that certain  particles present in water may protect
  microorganisms from inactivation by disinfection certainly  is logical, little
  direct evidence of such protection has been demonstrated. Clarke and Chang
  (9)  stated  that turbidity was responsible for disinfection failure which re-
  sulted in the Delhi infectious  hepatitis outbreak. An AWWA  committee (3)
  stated that "Effective  disinfection can  be carried out only on water free from
  suspended material." Gulp (11) stated that for good viral disinfection turbidi-
  ties of 0.1 JTU were  preferable. Cookson. (10) stated that the ability of dis-
  infection to produce virus-free water  depends on removal of turbidity to no
 more than 1 JTU.
    A  few studies  have provided evidence that  the  presence of  particulate
 matter interferes with effective inactivation of viruses and bacteria in water.
 In an early rather crude study,  Neefe et al (15) showed that feces  from an
 infectious hepatitis patient when suspended hi water caused hepatitis in two
 of five  human volunteers after  exposure to  a chlorine residual of 1.1  mg/1
 for 30 minutes.  In a similar experiment, when the  water was first coagulated
 and then filtered before disinfection to the same chlorine  residual,  none of
 five volunteers contracted hepatitis. .Change et al (7) demonstrated that enteric
 bacteria and enteroviruses ingested by aquatic nematodes were  completely
 protected  against  high residual  chlorine  levels (80-90  mg/1) even when
 more than 90% of the nematodes were  immobilized. Tracy et al (18) re-
 ported  the  persistence of coliforms  in  drinking  water which was highly
 chlorinated. The problem  was most severe  when turbidities  were  between
 5 and 10 units. They concluded that the most likely  explanation was that
 the coliforms had been ingested by Crustacea present in the water and were
 thus protected from contact with the disinfectant.
   In a discussion included as  a  part of a paper by Clarke et al  (8), W. W.
 Sanderson and S. Kelly of the New York State Health Department, Division
 of Laboratories presented  evidence that turbidity interfered with disinfection
 in a  water supply. They studied an impounded water supply receiving no
 treatment other than chlorination. The  concentration of free chlorine residual
 in samples from  household taps after a minimum of 30 minutes contact time
 varied from 0.1 to 0.5 mg/1 and the total chlorine  residual was between 0.7
 and 1  mg/1. These samples consistently yielded confirmed coliform organisms.
 Turbidities in these samples varied from 4 to 84 TU's and microscopic exami-
 nation showed iron rust and plankton to be present. They concluded  ". . .
 coliform bacteria were imbedded in particles of turbidity and were probably
 never in contact with  the active agent. Viruses, being smaller than bacteria,
 are much more likely to escape the action of chlorine in a natural water. Thus,
 it would be  essential to .treat  water by coagulation and filtration  to nearly
 zero turbidity if chlorination is to be effective as a viricidal process."
   Results of our research on  this  problem as well as  the results of other
recent research provide more definitive  information regarding the relationship
of low levels of turbidity and disinfection. Our research in this area has
been  designed  to determine whether  significant  differences in disinfection
                                                                      105

-------
                             MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
-'!/' -1
                       1 ft,  ,	Ill'1"11
                                  -1
                                  -2
                             S

                             cc

                             <
                             >

                             OC
                      lif"1"!"!'. " 'liilll!  i,i.".!,'
,-4
                                  -5
                                  -6
                                  „„,  ,*y
                        D VIRUS ONLY - 0.28  NTU
                           HOCI (0.50 - 0.46 mg/l)

                        A VIRUS * BENTONITE -  7.1 NTU
                           HOCI (0.50 - 0.44 mg/l)
                                                         A
                                                         n
                             A
                             D
                                       A
                                       D
                                                                                             D
                                                                                                  n
468
     TIMI=(min.)
                                                                                       10
                                                                12
                                       Figure V.' iMACif IVATl"oN OF BENTONITE - ASSOCIATED POLlOVlRUS 1 BY
                                              CHLORINE (pH6.0, 5° C).
                              efficiency can be specifically related to turbidities of 5 NTU and 1 NTU. All
                              experiments were conducted at 5°C and pH 6.0. Under these conditions vir-
                              tually 100%  of the free chlorine residual is present as hypo-chlorous acid
                              (HOCI) which is the most  effective chlorine  species for disinfection. The
                              LScZab  strain of attenuated ppliovirus type 1, treated to remove chlorine
                              demand," 'was'used.	'	
                                Studies  of  effects of inorganic turbidity on  disinfection were done using
                             	;|Virus	"SorSe'd	l!;"on"":"oentonite	clay	and "virus  precipitated  with  aluminum
                              phosphate (A1PO4). Volumes of virus associated with bentonite or A1PO4
                              sufficient to produce a turbidity of ~ 5 NTU were added to stirred reactors
                              containing 0.5 mg/l  HOCI.  Centrifugation results  showed that 65-80% of
                              the bentonite-yirus mixture was bentonite associated and over 90% of the
                              AlPO4-virus  mixture  was"".'AlPO4  associated.  An  example of  results using
                              106


-------
                                  TUBSIDITY AND DISINFECTION/J.C. HOFF
     -1
    -2
 o

-------
 MICROBIOL STANDARDS EVALUATION/ClW. HENDRlCiKS
   O)
   E  1
   O
   O
   X
     -1
 3
 g
 cc
 cc
 '3
 05
    -6
                                                          s
                                                          O
© VIRUS-BENTONITE
      TURBIDITY-5-5 NTU
A VIRUS-AIPO4
      TURBIDITY-4.2 NTU
E VIRUS ONLY,
       TURBIDITY-0.2 NTU
                                                 G
                               TIME (min)
        Figures. INACTIVATION OF POLIVIRUS ASSOCIATED WITH AIPO4 AND
              BENTONITE BY 1.5 mg/1 HOCI.
conducted at higher HOCI levels (-1.5 to 3.0 mg/1). Results of an experi-
ment comparing inactivation of clay associated, A1PO4  associated,  and free
virus at 1.5 mg/1 HOCI are shown in Figure 3. Inactivation rates were very
similar, with bentonite associated virus showing a slightly lower rate of in-
activation than AlPOY associated or free virus. Figure  4  shows a  com-
108
                                          	I,

-------
                              TURBIDITY AND DISINFECTION/J.C. HOFF
O
g  1
                      HEP 2 CELL-ASSOCIATED VIRUS
                       TURBIDITY-1.4 NTU
                      ARTIFICIAL HEP 2 CELL-VIRUS
                       MIXTURE TURBIDITY-1.O NTU
                      VIRUS ONLY  O.15 NTU

                    El HEP 2 CELLS PRESENT, VIRUS ADDED
                       SEPARATELY TURBIDITY-1.4 NTU
                             3       4
                              TIME (min)
          Figure 4. EFFECT OF CELLULAR MATERIAL ON RATE OF POLIVIRUS
                INACTIVATION BY CHLORINE.

parison of disinfection rates of free virus, cell-associated  virus and an
artificial cell-associated virus mixture  prepared  by mixing virus with cells
which had been harvested,  washed, and stored at 5°C for  1-2 days. The
artificial mixture was stored ovemite -at 5°C before use. The  results show a
pronounced protective effect of virus  associated with cells. Also  shown in
                                                            109

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 O) 2
                                       T-T
-G
                               A VIRUS  ONLY
                                   TURBIDITY-0.15 NTU
                               © HEP 2 CELL
                                   ASSOCIATED VIRUS
                                       TURBIDITY 1.4 NTU
                               30       40
                             TIME (min)
          Figure 5. INACTIVATION OF CELL- ASSOCIATED POLIVIRUS BY
                 CHLORINE (60 min.).
Figure 4 is an inactivation curve in which washed cells were added to the
disinfection reactor followed immediately by virus. This simulated conditions
in which virus inactivation was occurring in the presence of chlorine demand
created by the cellular material. The results show that the presence of chlorine
demand had no effect on the inactivation of the virus. Figure 5 shows the
results of a longer term experiment on  inactivation of cell associated virus.
110

-------
                                  TURBIDITY AND DISINFECTION/J.C. HOFF


HOCI levels declined from an initial level of 3.0 mg/1 to 1.5 mg/1, most of
the decline  occurring in the first 5 minutes of the experiment. Even though
the turdidity was only  1.4 NTU, virus was detected even after  50 minutes
exposure  to 1.5 mg/1  HOCI. Free virus  exposed to 3.0 mg/1  HOCI  was
reduced more than 5 logs in less than two minutes.
   In two other recent  studies,  the  effect of inorganic particles  on virus
disinfection has been studied. Boardman (6) reported that inorganic particles
including alum, calcium  carbonate and kaolin did not interfere with bac-
teriophage T7 inactivation by chlorine but did interfere  with the rate of in-
activation of poliovirus. The poliovirus results are shown in Table 1.  The
experiments were conducted at  22.5°C  at a residual chlorine level of 0.21
mg/1. The chlorine residual determinations indicate that  most of the residual
was present as combined rather than free chlorine. Although turbidity was not
measured, the  amounts of alum (20  ing/1)  and kaolin (50  mg/1) were
similar  to the  levels of A1PO4  (50  mg/1)  and bentonite (35; mg/1)  used
in our  research. Boardman related the protection shown to  extensive virus
aggregation  enhanced by the^presence of particulates serving as foci for virus
adsorption.  Experiments designed to show  this were inconclusive. Boardman
also noted that the protective effects wens greatest in systems in which infor-
mation  of particulates occurred in the presence of the virus (CACO3 & alum).
  Stagg et  al (17)  compared  the inactivation by HOCI of bacteriophage
MS-2 associated with bentonite clay and in the freely suspended  state. Both
freely suspended and clay associated phages  were inactivated ; rapidly  but
the rate of inactivation  of clay associated phage was retarded. At equivalent
HOCI concentrations ranging from 0.02 to 0.6  mg/1,  approximately twice
the length of time was  required  for equivalent inactivation (99% ) of clay-
associated phages as for freely suspended phages. Reaction rates in both cases
were  essentially first-order.  Time required  for 99% inactivation of freely
suspended phage was  10 seconds and for clay associated phage 19 seconds at
0.5 mg/1 HOCI at 22°C. Turbidity levels  for the experiments were in  the
range of 2 to 4 Jackson Turbidity Units (JTU).
  Use of cell associated virus in  studies of disinfection of combined sewer
overflows  by chlorine and chlorine dioxide was also reported by Moffa  and
Smith (13). Cell associated virus, when mixed with simulated combined sewer
overflow water was  much more  resistant to inactivation by  ClOs than  was
TABLE  1.  EFFECT OF  VIRUS  ADSORPTION  ON  POLIOVIRUS
 INACTIVATION BY 0.21  mg/1 RESIDUAL CHLORINE^ AT 22.5°C.

Preparation                                     pH     Percent Inactivation
                                                         in 10 Minutes
Polio Only
Polio Only
Polio + Kaolin (50 mg/1)
Polio -1- Alum (20 mg/1)
Polio + Ca CO3 (600 mg/1)
7.0
10.1
7.0
7.0
10.1
99.7
97,3
95.8
78.4
84.7
  'free residual range—0.02-0.08 mg/1
   combined residual range—0.12-0.14 mg/1
   From: Boardman (6)
                                                                     111

-------
MlCROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


 freely suspended virus. The results  are shown in Figures 6 and 7. Turbidity
 and disinfection demand characteristics of the water were not given. However,
 since the cell  associated  virus was in simulated combined sewer overflow
 water, it is likely that a high chlorine demand was  present.
   More recently, we have initiated some  studies on the effects of turbidity
 in  the range of 1  to 5 NTU  on coliform inactivation  by chlorine.  These
 experiments also were  conducted at pH  6.0, 5.0°C in 0.05  M phosphate
 buffer.  The coliforms used in these experiments  were  naturally  occurring
 coliforms associated with solids in unchlorinated primary sewage effluent. The
 effluent was collected and centrifuged at 2,000 RPM for 15 minutes to sedi-
 ment the solids. The sedimented solids were washed to remove soluble chlorine
 demanding substances by repeated centrifugation and resuspension in 0.05 M
 phosphate buffer, pH 6.0. The concentrated  washed solids, when diluted to
 produce turbidities of 1 and 5 NTU contained coliform  concentrations of
 1 X 10s —  IX 10V100 ml.  Inactivation of the coliforms associated with
 the solids was  studied in  stirred batch  reactors containing 0.5  mg/1 HOC1.
 Coliform levels were determined by standard  coliform MPN tests.  Completed
 tests were conducted on all positive tubes. Inactivation  of a washed culture
 of  E. coli (ATCC  11229) by the same level of HOCl was  also determined.
                                                  ,       ..
                                       SYMBOL    DOSAGE mg/l
       o
       D.
             10
             10
              10 -
                0   30   60  90  120 150 180  210   240  270  300

                                TIME, seconds
 Figure 6. SINGLE-STAGE DISINFECTION OF POLIVIRUS 1 WITH
         CIO2 IN A NO-DEMAND SYSTEM. (ADAPTED FROM: MOFFA, 13)
112

-------
                                  TURBIDITY AND DISIlSfFECTION/J.C, HOFF
Typical results are shown in Figure 8. It is evident that the paturally occurring
coliforms associated with effluent solids were much more resistant to chlorine
than was the washed E. coli culture. At both 1 and 5 NTU's survival ratios
were  similar,  with coliforms still  detectable  after 60  minutes exposure to
0.5 mg/1 HOC1.                                      ;
   The results of the recent virus studies described above | indicate that virus
associated with inorganic particles such as  clays or  A1JPO4  is inactivated
by HOC1 at rates similar or perhaps  slightly lower than  freely  suspended
virus. The rates in both cases are essentially first order witri no evidence of an
extremely resistant fraction. In contrast, disinfection rates of cell associated
viruses were not first order, with inactivation rates beconiing slower as time
of  exposure increased.  This indicates  that  various degjrees of  protection
existed with a few complexes being extremely resistant to Disinfection.
    u.
    a.
   O
   Q_
                                      SYMBOL  DOSAGE mq/l
          10
          10
10
          10
             0   30   60   90  120 150   180 210  240J   270 300
                              TIME,  seconds
Figure 7.  EFFECT OF C!O2 ON CELL-ASSOCIATED POLIVIRUS 1.
           (ADAPTED FROM: MOFFA, 13)

                                                                   113

-------
                •Ill
                                     	•	";' •	" «' 1	'I' '' fTTirWffl	19' Iff? -	fIIITT 5	W\'"!!, :'T: ""! '"*; 1|l:!li;Ti	" I!	5 V'	"!	I: ] 'f!'!!	"~m
".. Si:
                                                                                          ir •;" i" y	;' i:* i	u! •" linn1 »F •,' ,in i'  nwi! •' n 11- IT; ,; , i|, i| nniri" • TJ/ ,rn •	«T asm
                                    MICROBIOL STANDARDS EVALUATION/C.W. liENDRICKS
                                         1.0
                                          -1
                                         -2
                                     o
                                     o
                                     at
                                     ii
                                     CO
                                         -4
                                         -6
                                                                 20         30
                                                                       TIME (MIN.)
                                 40
           50
                            O WASHED £. CPU

                               WASHED  PRIAAARY
                               EFFLUENT SOLIDS

                            A 5NTU

                            D 1NTU
                                                                 J_
10         20         30
                TIME  (MIN )
40
                                                                                                  50
60
                                            Figure 8. INACTIVATION OF COLIFORMS ASSOCIATED WITH PRIMARY EFFLUENT
                                                    SOLIDS BY CHLORINE (pH 6.0, 5.0°.C).
                                    114
                                                                               ll.lf ,!	'I '','	•, • i !'i i'll<» '4, .* Inl, l,i, l i111!1!1"] ,, I li1"1: < Illlli' ,, ,i, , l'i il'F / >,'! L! .1,,' '

-------
                                   TURBIDITY AND DISINFECTION/J.C. HOFF
Discussion
   In these studies, the measure of the presence  or absence of viable virions
after exposure  to  the disinfectant was  the ability  of the virus-particulate
complex to initiate infection of a cell, thereby resulting in a visible plaque.
Recent studies have  shown that  enteroviruses associated with j solids remain
infective for cell cultures (14) and for animals (16). The mechanism by which
virus-particulate  complexes initiate infection is not known.  It is possible that
the complex is broken at the cell surface and the virion enters the cell  as an
unadsorbed particle or that the entire complex  enters the  cell. It  is known
that various chemicals, notably serum protein can prevent adsorption of virus
to inorganic particles and  also can elute virus previously  adsorbed to such
particles.  Because of the  nature of the assay system  Giving cells)  it is in-
evitable that some soluble protein will be present at the cell surface. Because
of. this, it is difficult  to determine the details of cell  infection by adsorbed
virus. It is also possible that viable virions, completely encased in feces, cell
debris,  etc. and  thereby completely protected from disinfection, may exist
and remain undetected. In a few instances in the  studies we  conducted, efforts
were made to release protected cell associated virus by freezing and thawing
of samples collected  at various intervals during  disinfection. The results of
these studies were equivocal.  Because  of  the complex digestive  processes
which occur following ingestion,  such  virus could be freed  from encasement
and  being viable, initiate infection in the  host.  Thus,  virus which remains
viable but undetectable in cell culture systems could possibly: be detectable
in an animal assay system.     -        -
   While we do not know  how well cell associated  viruses simulate viruses
as they exist in natural waters, it is likely that this model is more representa-
tive  of natural conditions than the model employing freely suspended viruses.
The results of the disinfection studies of  naturally occurring solids associated
coliforms show inactivation curves simiilar to those shown  by cell associated
viruses. This would indicate that  similar  effects would be seen with naturally
occurring viruses present in such solids,
   In interpreting the  results of these studies, particularly the virus disinfec-
tion research, it should be kept  in mind that the protective effects  shown,
while they well may reflect what actually occurs during water treatment, repre-
sent a tremendous  magnification  of the  virus levels which  might be present
under "real world"  conditions. Whereas in  laboratory experiments, we can
begin studies with tremendous numbers of viruses (~  10s  PFlT/ml), actual
numbers present in some  of  the worst  raw waters used  as potable  water
sources may contain on the  order of 1 PFU/gallon.
   The information presented above has  been limited to the consideration of
turbidity interferences with disinfection of potable water. Several other very
important aspects of turbidity as related to health  hazards associated with pot-
able water have not been addressed. These include  the use of turbidity as an in-
dicator of the presence of particulates such as  asbestos which may  pose a
chemical hazard or as an  indicator of the presence of  microbiological forms
such as E. histolytica, bacterial spores,,  and perhaps Giqrdia  lamblia which
have or are believed to have a high innate  resistance  to disinfection. These
aspects must be taken into account  in considering requests! for allowing
turbidity limits higher than 1  NTU.
                                                                       115

-------
         MICROBIOL STANDARDS EVALUATION/C.W. iSNDRICK
         Conclusions
                                             I  , .  j  ' » :   I''1' ! j'  ] , !' , "  .   [ -:   ,  '    •"  '•
           As stated previously, turbidity measurement is well established as a useful
         operational guide for assessing" the performance of  various processes such as
         coagulation, sedimentation and filtration in water treatment plant  operation.
         Unfortunately because of its lack of specificity in determining the nature of the
         turbidity particles present, turbidity does not appear to be the ideal indicator of
         _•*• •_ J!	A _ t • 1! .*._.. _.C __ A4.nl>1A •••*•*.». •*•<•**. n-1-n.nri 4-t-» a.  C n+avfckranf+at  fff "f llT'Kirllt'XT' "Wit n  filCITI-
        disinfectability of potable water
since the interference of turbidity with disin-
        fection depends much more on the types of turbidity present than the number
        of turbidity units present.  The Bacteriological data derived from the primary
        effluent studies indicate  similar [disinfection curves at .turbitities  of  5 and 1.
        Based on this, one could assume tiSat'water Seatment processes which reduce
        this type of turbidity to 1 NTIJT rather than 5  NTU  would result in an ap-
        proximate 5 fold  increase in safety factor. In  other words the disinfectant
        resistant portion would still  be jpresent but at a 5 fold lower concentration.
        Although cell-associated virus studies were not  conducted at both turbidities,
        it seems reasonable that the saiiie  observation could be made in the case of
        turbidity caused by clays and AJ1PO4. Interference with disinfection efficiency
        is either negligible or relatively! minor. It is likely that  little in the way of
        increased safety factor would be  gained  from  reducing  turbidity caused by
        these types of particles from 5 NTU to 1  NTU.
           While  it is  evident  that turb
:dity is  at present the  established parameter,
         future  consideration should  b&  given to  other parameters as  a  possible
         measure  of participate interference with disinfection  for use as regulatory
         criteria. Parameters such as total organic carbon or nitrogen in the suspended
         solids present in a particular  water might be used to determine whether the
         solids are organic or inorganic.  Determination  of chlorine demand of the
         solids present in a  particular water might also  be used to distinguish in-
         organic from organic particles.
           Further consideration of the
relevance of coliforms as indicators  of viral
        safety of disinfected water may also be in order. The results of the studies
        on inactivation  of naturally occurring coliforms  associated with primary
        effluent  solids  indicate  that sjich  coliforms  are  far  more resistant  than
        freely suspended  coliforms.  Thus,  the  coliform  group  may  be a  more
        rigorous  disinfection  indicator  'than is indicated by  disinfection studies of
        pure  cultures.                   !
                                        CFERENCES

          1. American Public Health  Association.  19751 Standard-methods for the ex-
            amination of water and wast ewater. 14th ed. American  Public Health Asso-
            ciation, Inc., New York.
          2. American Water Works  Association.  1971. Water  quality  and  treatment.
            3rd ed. McGraw Hill, New York.
          3. American Water Works  Coriiniittee on  Viruses  in Water.  1969.  Viruses in
',	;    ,';   '    water.  JA'WWA 61:491-494. \   	^  "[	'_';	,",„	  .'	  	;
          4. Berg, G. 1973. Reassessment I of the virus problem  in sewage arid in surface
            and renovated waters. Progress in Water Technol. 5:87-94.
          5. Bkton,  G.  1975. Adsorption of viruses  onto  surfaces  in  soil  and  water.
            Water  Res.  9:473-484.
          6. Boardman, G.D. 1976.  Protection of  waterborne viruses by virtue  of their
            affiliation with particulate  matter. Ph.D. Thesis.  Univ. of Maine.
          7. Chang, S.L.,G. Berg, N.A. J Clarke, and P.W.  Kabler.  1960.  Survival,  and
            protection against chlorination,  of human  enteric pathogens  in  free-living
            nematodes isolates from watejr supplies. Am. J.  Trop. Med. Hyg.  9:136-142.

-------
                                     TURBIDITY AND DISINFECTION/J.C. HOFF
  8. Clarke, N.A., G.  Berg,  P.W. Kabler, and  S.L. Chang. :l 964. Human enteric
    viruses in water: source, survival, and reir ovability. International Conference
    on  Water Pollution  Research, Pergamon Press, Oxford.
  9. Clarke, N.A. and S.L.  Chang.  1959. Enteric  viruses in  water.  JAWWA
    57:1299-1317.
 10. Cookson, J.T., Jr. 1974. Virus and water s'upply. JAWWA. 66:707-711.
11. Gulp,  R.L.  1974.  Breakpoint chlorination
for virus inactivation. JAWWA
    65:699-703.
12. Hach,  C.C.   1972.  Understanding  turbidity  measurement.  Indust.  Water
    Eng. 5:18-22.
13. Moffa, P.E.  and  J.E. Smith,  1974.  Ben,ch-scale  high-rate  disinfection  of
    combined sewer overflows with chlorine and chlorine dioxide.  Final Report
    EPA  Project  #8-802400. United  SCates  Environmental  Protection  Agency,
    Washington, D.C.
14. Moore,  B.E.,  B.P. Sagik,  and J.F. Molind,  Jr. 1974. Viral association with
    suspended  solids. Water Res. 9:197-203.
15. Neefe, J.R.,  J.B. Baty, J.G. Reinhold,  and J. Stokes, Jr. 1947. Inactivation
    of the virus  of  infectious  hepatitis in drinking water. Am. J. Public Health.
    37:365-372.                                                 '.
16. Schaub, S.A., and  B.P. Sagik. 1975. Associition of  enteroviruses with natural
   and artificially introduced colloidal solids in water and infectivity of solids-asso-
    ciated virions. Appl. Microbiol. 30:212-222.                  '•
17. Stagg, C.H.,  C.  Wallis, and C.H. Ward. 1977. Inactivation of \clay-associated
    bacteriophage MS-2 by chlorine. AppL Environ. Microbiol. 33:385-391.
18. Tracy, H.W.,  V.M. Camarena,  and F.  Wi|ng.  1966.  Coliform persistence in
    highly chlorinated  waters.  JAWWA. 55:1151-1159.            '
19. White, G.C.  1972. Handbook ,oi .Chlorination. Van Nostrand Reinhold Co.,
    New  York.                               i

                                                                          117

-------
                                                                                                                                                                                           II	'INiW  "II	':,'  '  I	I' '  I  ''lif'l! "'
                                                                                                                                                                                                                                                                                                    iiiiiii!j"i»  "	j    mi nil
                                                                                                                                                                                                                                                                                                                                                                                                    'I'l'lIU'illl! Ililllllli   I  I1  i1   '•*>  '"
                                                                                                                                                                                                                                                                                                                                                  ifi'i
II                              II
                                                                                                                                                                                                                                                                                                                                                                                                                                                                     li	I	in
                                                                                                                                                                                                                                                                                                                                                                                                                                                                    f":   !'!'
                                                                                                                                                                                                                                                                                                                                                                                                                                                                     I'B   i'iv

-------
                                            WASTEWATER REUSE/H.W. WOLF
                            Wastewater Reuse
                                  ami the
                    Problems  of Water-Borne  Disease

                             Harold W. Wolf
                         Environmental Engineering
                          Texas A&M University
                        College Station, Texas 77843
 Introduction

   Much has been written about wastewater reuse,  a topic that seems to be-
 come more popular with time. Since there are many types of reuse applications
 made of wastewater, perhaps the first thing we should do is; define the type
 of reuse that concerns  this  presentation:  direct, potable reuse. But this still
 isn't as sharp a definition as I'd like. Webster  (10)  says that potable means
 "suitable for drinking." I personally believe that there may be other uses  of
. water that are more demanding than just drinking (11). For example, a water
 could conceivably not present any trouble to a person drinking it if it contained
 a substantial population of Staphylococcus aureus, but prepare a salad with it
 and allow it to incubate a few hours and  the results could be quite different.
 Hence, I usually generalize the type of reuse that I have in mind  as:  for
 intimate human usage.
   Another definition is essential  to  our discussion  today.  Webster (10)
 defines disinfection as "the freeing from infection esp. by destroying harmful
 microorganisms." The Glossary of Water and Wastewater Control Engineer-
 ing (3) defines it as "the art of killing the larger portion of microorganisms
 in or on a substance with the probability that all pathogenic bacteria are killed
 by the agent used." The Glossary definition thus:
   1)  introduces the concept of disinfection being a rate function—which it is
  ' 2)  introduces the concept of probability,  which is also proper, and
   3)  excludes viruses.
   When applied to drinking water, the Glossary definition must be updated
 to include pathogenic viruses. Until it is,  we must be careful  that  we define
 the interpretation of disinfection we are talking about.

 Organisms of Concern

   Since the wastewater  reuse under consideration concerns the  most intimate
 human usages,  and since wastewater can conceivably include anything,  we
 must address virtually all pathogenic  agents. It would be inordinately naive
 to be concerned only with enteric organisms even though these are expected
 to be the most abundant.
   From a disinfection point of view, the organisms of concern can be grouped
 as follows:       .
                            Bacteria
i                              Vegetative
                                Encapsulated
                                Non-encapsulated
                              Spores                       :
                      .-•-.-. .Fungi & Yeasts
                            Rickettsiai

                                                                      119

-------
  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                              Viruses
                                Free
                                Within tissue
                              Protozoans
                                Trophozoites  (vegetative state)
                                Cysts (spore state)
                              Helminths & Flukes
                                Adults or intermediate forftis
                                      "      '  ' " ..... "' "       "
                           .          . . .     ,    .
  I believe such a grouping of microorganisms is more useful for our purposes
  than a long listing of the many genera and species of microorganisms that
  can cause  disease. For an excellent discussion of  waterborne pathogens, see
  Geidreich's chapter in Mitchell's book  (5). I would be reluctant  to make up
  a list of pathogens because those I overlooked would surely be the next to be
  involved in a waterborne outbreak.
     One can look at the grouping presented and ask, "How well does the coli-
  form group represent all^of these?" The answer is  obviously "Npt nearly well
> enough."
     One can also pick out the most resistant forms rather quickly;  e.g., the en-
  capsulated vegetative bacteria, spores, viruses  protected by tissue, cysts, and
  ova.  A  classically  durable  ovum  is that  of  the  roundworm,  Ascaris
  lumbricoides.                                               -
     Lastly,  one can look at the list and realize what a  formidable task, we're
  expecting of any one disinfectant.  Viewed from this perspective, hypochlorous
  acid (HOC1)  looks quite spectacular.

'''Problems  . .      '                 '      ..........       '

     1)  Drug resistance
          Any review of water-borne disease problems that might be associated
        with potable reuse of wastewater effluents should, in my opinion,  start
       with the  following statement from  Grabow, Prozesky, and  Smith of
       South Africa:
          "The therapeutic value of antimicrobial drugs is  diminishing due to
        the  rapid increase of resistant  bacteria. A current  prominent type of
       resistance  is mediated by R factors  (extra-chromasomal  nucleic  acid
       elements) which may cause  high  level resistance  to many drugs. These
       factors  may also  provide resistance  to other antibacterial  agents such
       as u.v.  light, heavy metals, bacteriocins  and  phages,  and may enhance
       the  virulence and infectivity of pathogens.  Intestinal Gram-negative
       bacteria like coliforms may  act as reservoirs of R factors and transfer
       them to pathogens. There is evidence that sewage polluted water  may
       play  an important role in the  spread of coliform  and other bacteria
       carrying R factors. Since coliforms have joined forces with bacteria
       increasingly  involved  in disease, they can no longer  be  regarded as
       harmless indicators of fecal  pollution. This calls  for  a re-evaluation of
       water quality standards and for  more  advanced purification of sewage
       prior to discharge into the en viroment" (6).
         Drug-resistant coliforms  are  not  the exclusive  domain  of  South
       Africa. A  few weeks  ago, Dr. Sagik (Dean  of Science, University of
       Texas Health Science  Center, San Antonio) mentioned that he  was
       looking for a coliform to use in tracing studies and searched the Austin
       Wastewater Treatment Plant for coliforms resistant to streptomycin +
       penicillin + nalidixic acid  and  found from  1:10 to l:iOO  coliforms
„'!'  ,    '    .......        ,,  ......... , ..... „  ,   ;'.;.' .; .......    ; ,  j     ;          '
 , 120
iiiiiiiiit jsst\.:',tja ..... i:;,.: '.i!1;!; ....... ;„;„ jj ^ ........ ;i; .,f ^..atii ..... iiiiiii.!: j j.. *. ,
-------
                                         WASTEWATER REUSE/H.W. WOLF
   resistant to all three (Sagik,  B. Telephone communication, Feb.  8,
   1977).
      Some of us in Texas look at our long border with Mexico with some
   trepidation because we know how readily available antibiotics  are across
   that border, we know that is a large movement of people across that
   border, and we read periodically in Morbidity and Mortality—and  in
   Science  (4)—increasingly frequent  accounts of illness; caused by anti-
   biotic resistant strains of  bacteria—many of which occur across that
   border.
2) Viruses                                             ;
      I think that public health engineers, in general, are fairly comfortable
   with  the fact that viruses can be  inactivated by HOC1.  Now and then
   we see an article that questions  our faith  (7). The most recent such
   occurrences have been discussed by Akin and Jakubowski (1).
      The question of virus-presence in  finished drinking waters is going  to
   be with us a long time. It will be with us until we can cultivate the virus
   of infectious hepatitis and it will be with us until we learn a  great deal
   more than we know now  about the slow-growing viroids.  I am not
   aware, however,  that these organisms diifer structurally from  the larger
   viruses, i.e., they all appear to consist of a nucleic acid in a protein coat.
   Consequently, we should not expect the viroids to differ very much from
   the larger viruses in their response to HOC1. However, other factors
   could contribute to a finding of significant differences, e.g., a small virus
   could be protected from the action of HOG1 by being  imbedded in a
   particulate which would not be; large enough to provide the same pro-
   tection to a larger virus.
      Unfortunately,  HOC1  is  a factor  in  the production  of trihalo-
   methanes—substances that we would  like  to minimise the presence
   of in our  drinking  water. Fortunately, we have three other powerful
   agents that might be considered. Morris (8) reports that ozone is even
   more effective  than HOC1, and  that viruses—although  more resistant
   than vegatative bacteria—can also be readily handled by, ozone. Chlorine
   dioxide appears capable of handling viruses, and so does one other tech-
   nique which is  not discussed very often—high pH  (12)  With all these
   techniques available, there appears  to be ample methodology for the
   handling of viruses in a reuse situation. The concern really boils down
   to other qualities of the water and whether these might interfere with
   the disinfection process.
3) Pseudomonads
      Most of us .are aware of the ubiquity_of the pseudomonads. I think
   they can be found almost anywhere that you look. Alexander  (2) states
   that only a small number of bacterial species can bring,about denitrifica-
   tion,  and  of these;  pseudomonads—including the  pathogenic Pseudo-
   monas aeruginosa—are the more important. Possibly because of then-
   relationship to denitrification, pseudomonads are important  organisms
   in soils and in wastewater treatment.
      Even though the pathogenicity of Ps. aeruginosa is determined more
   by the patient's state  of  resistance than  by any  inherent  virulence
   (9),  the  thought of creating conditions  for these  microorganisms  to
   proliferate in our distribution and plumbing systems is cause for some
   soul-searching in any discussions of reuse for potable purposes.
                                                                   121

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 Recommendations

   Drug-resistant bacteria
      1. EPA should  support  periodic  surveys of  the  drug-resistance of
         coliforms isolated from public water supplies.  By doing  this,  a
         picture should evolve which will better define the potential of the
         problem  to our water supplies.
      2. EPA should  not adopt  new  coliform standards, but  should take
         a hard-nosed  stance about  any infraction  of existing  standards.
         The  monitoring program for example,  might call for the follow-
         ing:
           a)  If a coliform  determination is >1 or  >2.2 per  100 ml  (de-
              pending upon methodology), immediate resampling  should be
              done. (If  it is  above  4/100 ml,  public  notification  should be
              made)
           b)  |f the resampling  is  >1  or  2.2  per  100  ml,  both public
              notification and immediate study should be done.
  Viruses
      1. Any  studies of virus presence in finished waters must pay more at-
         tention to those confounding factors  that are known  to interfere
         with  the disinfection processes used. Turbidity, pH, and tempera-
         ture  are  not the only factors; NH3 organic-N, chlorine  species,  a
         measure  of  organics,  and particle-size are  also  important  (in-
         tuitively).
      2. Routine coliphage determinations  at select water treatment plants
         should be instituted. If  the  disinfection process is working as it
         should, there should be  no viable coliphage in the finished water.
         If coliphage are found, why look for pathogenic viruses? The con-
         ditions for pathogenic virus survival will have existed.
      3. EPA should support basic studies  of  viruses and basic  studies of
        • virus   disinfection  by  a variety  of  disinfecting  or  inactivating
         methods.
  Pseudomonads
    The .problem which I have referred to as a Pseudomonas problem should
  really be addressed as a problem of maintaining water quality in distribu-
  tion systems.  Recommendations for  EPA  activity here would include:
      1.  Basic studies of water quality and organism growth.
      2.  Strict application of the rules of  sanitary surveys with  regard to
         covered reservoirs, dead ends, adequate pressures, main disinfection,
         etc.   i_       	   '
-------
                                              WASTEWATER REUSE/H.W. WOLF


 2. Alexander, M. 1961. Introduction  to soil microbiology, p. 299. John Wiley  &
    Sons, Inc., New York.
 3. American  Public Health  Association.  1969.  Water  and  wastewater control
    engineering.  American  Public Health Association,  Inc., New  York.
 4. Calliton, B.J. 1976. Drug  resistance growing worse. Science 194:1396.
 5. Geldreich, E.E. 1972. Water-borne pathogens, p. 207. In:  Mitchell R., (Ed.),
    Water pollution  microbiology.  Wiley-Interscience, New  York.
 6. Grabow,  W.O.K.,  O.W.  Prozesky,  and  C.S.  Smith. 1974., Drug resistant
    coliforms call for review of water quality standards.  Water Res. 5:1-9.
 7. Liu,  D.C., H.R.  Seraichekas,  E.W.  Akin, D.A. Brashear,  E.L.  Katz, and
    W.J. Hill, Jr.  1971. Relative resistance of twenty human enteric viruses  to
    free  chlorine  in  Potomac  water, pp.  171-195. In:  Snoeyink, V.,  and V.
    Griffin  (Eds.), Virus and  water quality: occurrence  and control. Proceedings
    of  the  Thirteenth Water  Quality Conference. University of  Illinois Bulletin,
    Vol. 69, No. 1, Urbana, Illinois.
 8. Morris, J.C. 1976. The role  of ozone in water  treatment. Paper presented at
    the  96th Annual Conference, AWWA, New Orleans, Louisiana.
 9. Smith, D.T.  and N.F. Conant. 1960.  Zinsser microbiology, p. 425.  Appleton-
    Century-Crofts, Inc., New York.
10. Webster's New Collegiate Dictionary. 1976. G & C Merriam Company, Spring-
    field, Massachusetts.
11. Wolf,  H.W. 1972.  The coliform  count as a  measure  of water quality,  p.
    344. In:  Mitchell,  R.  (Ed.),  Water  pollution microbiology.  Wiley-Inter-
    science, New York.
12. Wolf,  H.W., R.S.  Safferman, A.R. Mixson, and C.E. Stringer.  1974.  Virus
    inactivation during  tertiary  treatment. In:  Malina, J.F.,  Jr.  and B.P. Sagik,
    (Eds.), Virus survival in water and wastewater  systems. Center  for Research
    in water resources, The University of Texas, Austin, Texas.
                                                                           123

-------

-------
                            ENGINEERING CONTROL PRACTICES/P.D. HANEY


    EVALUATION OF MICROBIOLOGICAL STANDARDS FOR
                         DRINKING WATER             :

                      Engineering Control Practices
                                  .by
                             Paul D.  Haney
                  Black and Veatch, Consulting Engineers
                       Kansas City, Missouri 64114

Introduction                                                 .
   It seems  appropriate to hold this symposium at a time when  one of the
foremost citizen concerns of at least half the  Nation is water, or perhaps,
more accurately, the lack of it.  Regional water crises are not new. In my
lifetime, I have acquired some knowledge of several,  and have had personal
experience with serious  water shortage and  water quality problems in the
West and  Midwest.  Assuming  my remarks  are  of  some  interest and
significance  to symposium participants,  please, in assessing them,  keep  in
mind my western and middlewestern perspective (or provincialism, if you
will).
   My experience has been in the fields  of public health — state sanitary engi-
neering  and  the. U.S.  Public  Health  Service — teaching  certain  sanitary
engineering  subjects .and, in  more recent  years, the practice of  engineering
in a professional engineering firm specializing in energy and water projects.
.1 emphasize energy  because  water and  energy  are closely  intertwined. Each
impacts the  other.
   With this brief background, I shall now proceed to the  business at hand,
engineering  control  practices, with the  full knowledge that I am  tackling a
complex matter that cannot be brought into focus in 20  minutes,  or  even
20 hours or 20 days.  Therefore, please bear with me while I report briefly
on what is  expected of professional engineering and  water utility managers
and operators  in the vital fields of public water supply and water quality
control.  I  shall make  occasional reference  to Dr.  Murphy's law  and  dts
corollaries.

Drinking Water Standards— Past, Present, and Future

   To navigate successfully, one must have good  charts, know where he is
going, where he is,  and where he has been. There must be a few questions
about where we are and where we are going or we wouldn't be holding this
symposium. Hopefully, at its conclusion, we shall have some better charts
than now. Let's examine briefly where we have been  and where we  seem to
be.
   The first  U.S. drinking water standards that enjoyed, official status were
those established by theJJ.S. Public Health Seryicejn 1 9 1.4.-Q&en called the
"Treasury Standards", they covered only the  bacteriological aspects of water
quality. A most interesting feature_ of these standards  was the inclusion of a
                                                     Wnyjjga&Jtlkopped?
What is its jnipOT-tanee? - .
   TEe~1914 standards bulletin, which sold for five cents, briefly discussed the
sanitary  survey  and its importance.  The sanitary survey,  well known  to all
older sanitary engineers, has been a unifying influence in the standards for
over half a century.
                                                                     125

-------

        t*-
  TABLE 1.  CHANGING STANDARDS FOR DRINKING WATER.
[Tabulated numerical values are in milligrams per  liter (mg/1) except  for
          Bacteria,  pH, Taste/Odor, Turbidity, Year,  Color]
               1914    1925           1942       1946           1962
1975-1977 *
Conform Bacteria/ 100 ml
Total Bacteria/ml
Copper
Iron
-Lead
Manganese
.Magnesium
Zinc
Arsenic
j Chromium
( Selenium
1 Barium
^-\ OJ_Y.2 iim
V^SClIIllUT II
; Mercury
Silver
;pH
Chloride
: Sulf ate
Total Dissolved Solids
Taste/Odor (Threshold Number)
Turbidity (Turbidity Units)
Color Units
<2 <1
100 —
— 0.2
— 0.3
— 0.1
— 0.05
— 100
— 5.0
— —
— —
_ _ _
— —

""•--"— ' T
— —
,„ ^ , -....
	 	
— 250
— 250
— 1000
— None
— • "Clear"
— "Colorless"

-------
Hydrogen Sulfide                     —       —
Fuoride                              —       	
Phenols                              __       	
Cyanide                              —       	

Nitrate (N)                          	       	
Foaming Agents                     —       	
Carbon Chloroform Extract .  ,        —       	

Radioactivity                         —       	
                          ;              -            i  i
Pesticides                            —       	
                          i
Sanitary  Survey           ; ;          Yes       Yes '  ;

Viruses                                               :
Asbestos                                       .        |
Chlorinated Organics  ;     :                          j  \
What is next?             i ;    '                     '!  :

  * Federal Register, Dec. 24, 1975; July 9, 1976; March 31, 1977.
  1.0
0.001
 Yes
  1.5
0.001
 Yes
0.8-1.67
 0.001
   .01

 10.
  0.5
  0.2

 Yes
 Yes
 0.05
1.4-2.4
                                           10.
                                            0.5
                                           Yes
                                        .00
                                                             o
                                                             z
                                                             PI
                                                             w
                                                             o
                                                             o
                                                             i
                                                                                                                           b
                                                                                                                           33

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


  Table 1  is an attempt to summarize past Public Health Service (1914-62)
and present Environmental  Protection Agency standards. This table reflects
some sixty years  of changing  culture.  It is  apparent that bacteriological
quality  is still of concern  but the bacteriological standards have not changed
greatly  since 1914. Probably Ian engineer who could design a plant meeting
the 1914 coliform standards Icould readily  design one that could meet cur-
rent coliform  standards although much has been learned about water supply
and treatment since 1914. The table also tells us  that, over the years,  there
has been greatly increased concern  over things of  a chemical,  or better, a
microchemical nature. This upsurge in interest in microchemistry is essentially
a post-World  War II happening. This era was marked by an upsurge in in-
dustrial chemistry, notably petrochemicals,  and wastes associated with their
manufacturing and use were bound to have an impact on the water resources
of the  world. Parenthetically,  it should be noted that these  new chemical
products were welcomed  by consumers  who found them effective and  con-
venient. Accompanying this  industrial revolution  was  almost  overwhelming
progress in analytical chemistry, which is truly the cornerstone of knowledge
about this  complex, universal, solvent we call "water". We have, in about 30
years, moved from the era of'[microbiology into microchemistry. Yet, with the
emphasis on chemical quality, we  must not forget,  not for a moment, the
fundamental and  continuing; importance  of  microbiological water quality.
Speaking very generally, microbiological  contamination produces acute health
effects.  In  a susceptible population, water-borne disease will often  manifest
itself in a matter of days  or  weeks.  On the other  hand, usual microchemical
pollution; i.e., excluding chemical accidents involving massive contamination,
is  chrome  m nature, sometimes requiring  years  to  make  its effects  felt. I
repeat,  these  effects, while  very serious, are usually long-term rather  than
short-term. This, of course,  makes  the  investigation of microchemical  con-
tamhianfs exceedingly difficult.
   Regarding the future of our drinking water standards, I find the crystal ball
quite cloudy. I am certain, however, that they will reflect the inexorable march
of science. They will be more complex,  not less. The difference between the
standards of the year 2000 and those of the present probably will be greater
than the difference,  noted in the table,  between 1914 and 1977. (The pro-
posed upper limit for  pH given in the new standards seems unreasonable.)
   I would  be remiss, if I did hot remind you that at least one very important
facet of the standards does not lend itself well to numerical expression. Con-
sequently,  it is not included  in the table I  have presented.  This is the  state-
ment, and  I quote from  the| 1962 standards:  "The water  supply should be
obtained from the most desirable source which is feasible,  and effort'should
be  made to prevent or control pollution of the source. If  the  source is not
adequately protected by natural  means,  the supply shall  be adequately pro-
tected by treatment".        ;
The Microbiology Standard

  This symposium concentrates on microbiological water quality, which is
quite appropriate. I am sure the organizers recognized the impossibility of dis-
cussing microbiology, per se.jln thinking about the microbiology of drinking
Water, it is essential to think! of many other matters, not the least of which
is disinfection. As is so often the case in this imperfect world, the solution
of one problem can create another. The notable example is  chlorination^ for
bacteriological quality improvement which can  be attended by  potentially

128

-------
                             ENGINEERING CONTROL PRACTICES/P.D. HANEY
 harmful, side reactions. Chlorine  has a good. bacterial toxicity but a  poor
 chemical specificity, and this fact is very impoiitant in many situations.
   Over the years the  coliform test has served us well and I have little doubt
 that it will continue to do so.  It  is  by no means perfect, but 'what is? The
 standard test which is the heart of  the  standard itself may be  criticized as
 covering too broad a spectrum, too  difficult and time consuming,  too sensi-
 tive  in  some respects,  not sensitive enough in others. True,  the standard
 test has on occasion caused a false alarm and'much embarrassment but this
 has been the result of trying to short-cut the  standard procedure and not that
 of any inherent defect. Yet, for all of this, the coliform test and the coliform
 standard .have certainly been a .major force in our  conquest of  epidemic
 disease.                                     ;                  i
   As a state sanitary engineer, I was  once responsible for checking the quality
 of some 400 public water supplies. My predecessors had done a good job and
 in general these supplies were quite "satisfactory from a sanitary standpoint.
 Most were from protected wells. Water from [surface sources,  With one ex-
 ception, was filtered. Water samples collected by health department personnel
 and  those shipped to the department laboratory by water utility  personnel
 were competently examined and the results  promptly  transmitted to  the
 utility.  "Bad" samples, i.e., those  showing  coli'forms, were reported  by  tele-
 phone or telegram and confirmed  by letter. Quarterly summaries of the  bac-
 teriological quality of  all supplies were prepared  and posted. It was dismaying
 to note that each quarter,  some 100  of the supplies failed to meet the drink-
 ing water standards. What was the  reason.for this  rather alarming  state of
 affairs? I hate to admit it, but I  couldn't answer  this question .then and I
 can't now. I can only  make a few clarifying  comments. As I recall, none of
 the surface water supplies, all subject to microbiological contamination,  and
 treating waters  containing from  about  10  to; far above  100,000  coliform
 organisms (MPN) per 100 ml, ever failed to meet the, standards. Neither did
 those ground water supplies where fairly extensive treatment systems had been
 installed; e.g., iron removal, softening. The principal things in common about
 these utilities were filtration and chlorination.  O'f these two, I did then and do
 now, believe chlorination was of major importance. Invariably, the supplies
 that failed to meet  the  standards  Were unchloirinated supplies  derived from
 wells. In general,  the wells were properly located and protected (there were,
 of course, exceptions)  and they supplied, water that met the bacteriological
 standards. However, we encouraged water utility personnel to obtain samples
 from various points throughout  the  distribution system  and they complied.
 The problem of "bad" samples could,  in general,  be identified with the  fol-
 lowing:
   1.  Poor sampling                           :
        a.  bad technique                      :
        b.  improper sampling point; e.g., a fire, hydrant.
   2.  Water main, service line, plumbing repairs; failure to flush  and disinfect.
   3.  Tank painting and repairs; failure to flush! and disinfect. Improper vents
      and access covers.                       '                  :
   4.  Well, well pump, etc. repairs; failure to  flush and disinfect.
   5.  Gross connections; backflow problems.
   6.  Low or negative pressures in system because of operating defects or in-
      adequate hydraulic capacity.             ;
   7.  Ground storage repairs; improper vents,  covers, access manholes.
   Other than bad sampling, which is a matter of training and judgment,  the
problems seemed to arise principally from mindr contamination in the distri-
bution system. Now, there were probably about 4s many defects in the distribu-

                                                                      129

-------
FI,
                                  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


                                  tion systems  of surface sources  as  there were in ground sources.  Yet the
                                  surface supplies had good records while many of the unchlorinated ground
                                  supplies had  poor records. The difference apparently  was chlorination.  The
                                  results of a campaign to  secure  chlorination  of  all supplies supported this.
                                  When chlorination was applied, a system that  formerly gave an unacceptably
                                  high proportion of bad samples suddenly turned around and consistently met
                                  the standards. Simple, continuous chlorination of all'wells made the difference,
                                  which was dramatic. In fact, the clean-up was so sudden and complete  that
                                  the state sanitary engineer was accused of falsifying bacteriological data simply
                                  to prove his point that chlorination would go  a long ways toward improving
                                  water safety. The conclusion: the bacteriological standards can be met, in fact
                                  exceeded, by the exercise of reasonable care in system operation coupled with
                                  simple continuous chlorination. To summarize, most of the supplies  involved
                                  were acceptable until they entered the distribution system. Here, minor con-
                                  tamination occurred from a multiplicity of sources. Chlorination was gen-
                                  erally adequate to insure compliance with the  standards then in  force, which
                                  were  essentially the same as those published  by EPA in 1975. As a brief
                                  epilogue, I can report that the state's program  of insistence on chlorination of
                                  all  supplies was continued and as of  today,  all  public water supplies, over
                                  700, are chlorinated.
                                    As  to surface waters,  the program included  "selling" free-residual  pre-
                                  chlorination  to  surface water  utilities.  Generally, this was well received.
                                  Chlorine should be recognized  as a water treatment chemical. It  is more than
                                  a disinfectant. Prechlorination to  a free residual provides values beyond those
                                  associated with excellence in disinfection.
                                    I do not expect to pursue the  matter of nematodes in drinking water but they
                                  have been found and they may be more common than we realize in supplies
                                  derived from  some surface waters. Their source probably is the raw water,
                                  but the possibility that they may find a "home" in the treatment  plant cannot
                                  be  excluded. Nematodes  are macroscopic,  easily detected, and resistant to
                                  chlorine. Their greatest health significance seems to be the fact that they may
                                  ingest pathogenic bacteria and  then serve as a carrier and protector for these
                                  pathogens. The water  industry  is very much  aware of nematodes. They are
                                  indeed a  "can of worms". Chang and others have conducted  extensive in-
                                  vestigations of nematodes in water.
                                    Coliform criteria have  been proposed by organizations other than EPA.
                                  The American Water  Works Association has  established "quality goals" for
                                  drinking water and the AWWA rationale states: "Modern disinfection control
                                  procedures are such that  a practical goal can be destruction of  all  coliform
                                  organisms". The quality goal is "no  coliform organisms". This is, I think, an
                                  admirable goal, but proof of compliance probably would be  virtually impos-
                                  sible. We cannot deal in absolutes in the water safety business.
                                    The City of Chicago, well known for excellence in the control of water treat-
                                  ment systems, adopted a  "Chicago Standard"  of  not more than 0.5% of 10
                                  ml standard portions positive in contrast to an allowable 10%.
                                    There is some  relationship  between raw water quality and  that of the
                                  finished water. Over the years, guidelines for  maximum allowable raw-water
                                  coliform concentrations have been recommended. These have posed problems
                                  and, in large measure,  have been ignored, especially on "flashy" midwestern
                                  streams where coliform levels are usually fairly low during low and moderate
                                  flow periods but go out of sight during high flow periods. The latest  coliform
                                  criterion for raw waters is that of the National Academy of Science—National
                                  Academy of Engineering published in the "blue book" in 1973. This  calls for
                                 . a coliform limit of 20,000/100 ml (2,000/100 ml fecal coliform). It should
                                  130

-------
                            ENGINEEMNG CONTROL PRACTICES/P.D. HANEY
 be understood that this limit is based upon a "defined treatment process", as
 specified in the NAS-NAE publication "Water  Quality Criteria 1972". With
 additional treatment, a much higher coliform level can be handled without
 difficulty.
   We need to understand that "coliforms" represent a fairly broad spectrum
 of organisms derived from many sources, not necessarily disease related.
 Studies (1970) by the Department of Health and Environment, State of
 Kansas, (Soldier Creek Basin), have shown that the "heaviest bacterial pollu-
 tion  (coliforms) existed during  high flow periods" when dilution was the
 greatest. Natural and agricultural sources  were obviously  significant contrib-
 utors. The watershed had an area of 290 square miles and consisted prin-
 cipally of some 400 farms plus a 120 square mile Indian Reservation,  largely
 undeveloped. There were no "point sources" of pollution in the watershed;
 i.e., no municipal or industrial outfalls arid no confined animal feeding opera-
 tions.
   I am especially familiar with water treatment plants located along the Mis-
 souri River between Omaha, Nebraska and the Missouri-Mississippi confluence
 at St. Louis. These plants at times "see" very high apparent coliform densities
 yet have no difficulty meeting the bacterial standards. They are, of course,
 operated by competent personnel and provide multistage treatment including
'pre  and post chlorination. Many of the  plants also include lime softening, a
 process long known for its effectiveness in  the reduction of bacterial densities.
   I believe the coliform test has been and still is a valuable water quality
 monitoring tool. It  has some shortcomings  which are well known to this audi-
 ence. Despite these, if I were faced with  a decision regarding water quality
 and the constraint of being allowed only one test, I believe the  coliform  test
 is the one I would  select. In applying the test to treated water, one short-cut
 to sure trouble is failure to carry the test through the "completed" phase when
 coliforms are. "confirmed" in a finished water in which they are  normally not
 found.  One of my  gripes about the standard test is the "confirmed" designa-
 tion  for the  second  phase. The word "confirmed" is  a  strong word  and
 easily misinterpreted. Coliforms are not "confirmed" until the "completed" test
 is conducted—a rather complex  and time-consuming procedure. I wish we
 could come up with a truly descriptive word for the "confirmed" test.
   The  relationship between coliform density and  pathogens  is  understood
 principally in a qualitative sense. Some thirty years  ago a paper was published
 in which the authors, Kehr  & Butterfield, considered the relationship between
 coliforms  and enteric pathogens. For a typhoid rate of .01  to 30/1000 popu-
 lation/year,  Kehr &  Butterfield suggested a Salmonella  typhi  (E.  typhosa)
 per  million coliforms ratio of 3 to 120. They  presented  evidence  that  this
 ratio might remain relatively constant through natural purification processes,
 approximating 99.9%. They concluded that their studies ". . . emphasize the
 basic value of the  coliform test as an indicator of the possible presence of
 pathogens and indicate that a very real danger may exist when  coliforms, in
 even moderate concentrations, are present. The factor of safety provided by
 the ratio of a million or so  coliforms present for each E. typhosa would, it is
 believed, take care  of the usual fluctuations in the ratio of  E. typhosa to coli-
 forms provided the density  of coliforms , . . be kept quite low or eliminated
 by methods which reduce the general bacterial population". I commend  this
 paper to you for its interest and the imagination displayed by the authors, both
 of whom I knew and liked as individuals  and respected as competent scien-
 tists.
   In regard to viruses,  Berger, et al, .reporting  for the American Society of
 Civil Engineers in 1970, summarized water  treatment practice as follows:

                                                                      131

-------
                                    MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                                                      i
                                          "Ideally,  the  water treatment plant should be capable of reducing
                                        turbidity to less  than 0.1 Jackson Unit and provide an HOC! residual of
                                        1 ppm after contact period of 30 minutes. In practice, it may be assumed
                                        that for waters having a pH less than 8, an acceptable level of protection
                                        will be provided by treatment producing a turbidity less than 5 Jackson
                                        Units and a free .chlorine residual of 0.3-0.4 ppm with a contact period of
                                        301  minutes, "'for waters with higher pH, either the  chlorine residual or
                                        the; contact period should."be increased By half. Because  the water treat-
                                        ment plant is the  major line of defense against  introduction of viruses
                                        into the drinking water system, it is necessary to emphasize the absolute
                                        necessity of constant vigilance in  careful plant operation."
                                      I believe these recommendations still provide  sound guidance to the engi-
                                   neering  profession.
                                      My conclusions:  Chlorine  is  a  proven disinfectant. Prechlorination to a
                                   free residual provides excellence in disinfection  and is a valuable treatment
                                   adjunct. Ground  water supplies  need  at least  marginal chlorination as a
                                   safety factor. Such supplies, if not chlorinated, will have  difficulty complying
                                  1 with existing bacterial standards, principally because of "problems" associated
                                   with distribution. As  of now and assuming we can continue to use chlorine as
                                   in the,past, I see no serious  problems  in complying with  the  intent  of the
       , •„,-     s ;       .   ,"   • ;;    standards.              •         ...        	,  ,
                                      Supplementing  these remarks, I do  foresee some problems  for the small
                                   utilities who collect only one sample per month. Chance contamination  re-
                                   sulting from faulty collection, poor judgment in selecting the sampling point,
                                   and laboratory errors can easily throw a small utility out  of compliance. This
                                   sets up a chain reaction of notification that could prove extremely burdensome
                                   and possibly result in loss of public confidence if repeated frequently.  Public
                                   notification has both  good and bad features. It is a delicate matter. Judgment
                                   is involved in this entire matter  of standards compliance and we all know the
                                   futility of trying to write judgment into any sort  of law, rule,  regulation, or
                                   guideline. Some better  publications  about the standards would be helpful. I
                                   commend Region VII,  EPA,  for their booklet titled  "Answers to Questions
                                   About the Safe Drinking Water Act". Hopefully, the standards themselves will
                                  .be  pubished  in better  format than that provided by the  Federal Register,
                                   which is  poor.                                                             .

                                   Water Distribution

                                     Here is a real problem. The  distribution system, the most costly portion of a
                                   public water supply, has in full measure been treated like  a "poor relative", in
                                   the past. .This attitude is  now changing. There can be many troubles  in the
                                   distribution system. I have already mentioned some of them. Dr. Murphy's
                                   law, which tells us that: "In any field of scientific endeavor,  that which can
                                   go wrong will go. wrong" certainly finds application in the water distribution
                                   system.
                                     The distribution system presents a large water-pipe interface. I once roughly
                                   .estimated this at 0-6 to 0.7 acre  per 1,000 population.  Strange things can and
                                   do happen at this very large interface. Corrosion is one, but hi keeping with the
                                   theme of this conference,  I shall emphasize biological effects. These effects,
                                   which produce dramatic water quality changes, are identified in the literature
                                   and reach their peak  in distribution zones of poor circulation, notably  "dead
                                   ends". Here chlorine disappears and remaining  bacteria make the most  of
                                   their opportunity to grow and produce changes, all of which are displeasing to
                                   the water customer. Perhaps there is no health  problem but the customers

	        ••        	    132

-------
                            ENGINEERING CONTROL PRACTICES/P.D. HANEY
 don't like turbid, smelly water and they won't suffer it in silence either. On the
 other hand, perhaps there are health problems. These  bacteria are seldom
 studied to any extent. Possibly they have greater direct or indirect health sig-
 nificance than we realize. An indirect effect that has been identified is  inter-
 ference with the standard coliform test. This reminds me that the present and
 past standards, the 1914 edition excepted, are silent on the  matter of the
 total numbers of bacteria allowable. I know there  are good reasons for this
 omission; but, nevertheless, consider it unfortunate. As an alternative, I hope
 EPA will strengthen its statements and recommendaions as to the value of the
 "standard plate count" and will also  recommend a guide number.  If I were
 running a water system, I would certainly include the plate count as a routine
 test (9).
   Another  problem in water distribution,  the significance of which has not
 been generally appreciated, is that of bacterial slime formation on the interior
 of water  transmission mains. Th|s is serious for at least two reasons; the hy-
 draulic capacity of the pipe is significantly reduced; energy requirements for
 pumping are significantly increased.  It is almost unbelievable what a thin, rip-
 pling deposit of bacterial slime will do to a pipe's flow characteristics. A  very
 thin, rippling, slime layer averaging no more than about 1/32  inch in thick-
 ness can  reduce the capacity of a fairly large pipe as much as 50% without
 appreciably changing its internal diameter. The  slime growth is microbial in
 nature and  has high  adsorptive capacity. It often contains considerable  iron
 and manganese. The slime forming  organisms  can live and thrive on trace
 amounts of organics and minerals. The film builds insidiously over a period of
 years by extracting nutrients from the water and excreting waste products into
 it.
   Consider a 12-inch water pipe, connected to  a pumping station. The pipe,
 when installed,,has. a Hazen-Williams  "C"  value of 135. If, in five years, this
 "C" value is reduced to, say, 80, the capacity loss for the same  pressure drop
 is 100(135-80)7135 — 41%. To maintain the  original flow in the pipe, the
 pressure at the pumping station must be increased to the extent of! [135/80]1-85
 = 2.6 times its initial value or an increase of 160%. Considering that energy
 and energy costs, associated  with  pumping, vary  directly as  the pressure
 against which the pumps are working, it is evident that bacterial slime growths
 in pipelines are an  extremely  serious matter.  These  slime,  or  so-called
 "nuisance",  organisms may or may not have health significance.  (I doubt  that
 we can answer this question entirely satisfactorily.)  On the other hand, they
 do have practical operating and  economic significance.  The alternative to
 higher pumping pressures is to clean the line and restore the "C" value. This
 is expensive and difficult, and,  obviously, the pipe must be out ^ of service
during cleaning. Furthermore, cleaning is,  at best, a  temporary remedy. Un-
less the organisms are controlled somehow, the whole process starts over the
 day the clean pipe is put back  into service.  Pipe slime brings the water-energy
 relationship  into sharp focus.
  Another objectionable "occurrence that  can-result from  pipe slime  is a
 sloughing off of quantities of the slime. This can produce an offensive  and
probably unwholesome condition at what the U.S. Public Health Service,  and
now the EPA,  call the "free flowing outlet  of the consumer".
  The distribution system is indeed  a labyrinth, replete  with traps for  the
unwary. It of ten illustrates a well kifown corollary of Dr. Murphy's law: "Left
to themselves, things always go from bad to worse".
                                                                      133

-------
                                  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                   The Anatomy of a Water-Borae Epidemic

                                     Let's look at what can happen in distribution. This ''happening", or one like
                                   it, might well form the topic of another symposium. My treatment of it will
                                   be necessarily brief. This city, located in the western Midwest, had a population
                                   of about 11,000 and was an important railroad center.  (The latter was im-
                                   portant in at least two respects.) The water supply was  obtained from eight
                                   wells having a depth  of somewhat more than 100  feet. They were located
                                   several miles from the city and tapped a major aquifer known to be of good
                                   mineral and biological quality. The well water was pumped to a concrete
                                   storage reservoir, covered, as I recall, from which it was repumped through
                                   about 8 miles of 12 inch and  14 inch transmission pipes to  the city's water
                                   distribution  system. Early in  September,  the state health department  was
                                   notified by a practicing physician that a severe outbreak of intestinal disorder
                                   was occurring in the  city. Samples of  water were collected  and  coliform
                                   tests performed. Samples  from the  wells  were "clear".  Samples  from the
                                   distribution system all contained fairly large numbers of  coliform  bacteria.
                                   The water system did not include  chlorination, and emergency chlorination
                                   equipment was installed. "Boil-the-water"  advice  was distributed.  The rail-
                                   road, a major line, was ordered to  cease taking on city water. World War II
                                   was in progress and war industries and members of the armed forces traveling
                                   on troop trains were involved, as well as the city population. Corrective ef-
                                   forts were top late to  avert an epidemic and there were more  than 3,000
,'U JM1:'1 ,  'HI. I
                                             HYDRANT
                                                       SEWAGE OVERFLOWED

                                                       i. *- VALVE HANDLE
                                                          GROUND LINE ELEV.,100 5
                                     VALVE
                                                Figure 1. CROSS- CONNECTION SCHEMATIC DRAWING.
                                                       (American Journal of Public Health, 1944)
                                  134

-------
HYPOCHLORITE
 SOLUTION
                          CALIBRATED PUMP
                         R HYPOCHLORINATOR
                                                                                                              CHLORINE RESIDUAL
                                                                                                                   TESTING
                                                                         FLOW MEASUREMENT
                               INJECTION TUBE AND
                                CORPORATION COOK                                                  VALVE CLOSED


                                        TYPICAL SET UP FOR WATER DISINFECTION WITH A HYPOCHLORINATOR.
                             Figure  2. DISINFECTION OF NEW OR REPAIRED WATER MAINS  (Office of
                                      Civilian Defense, 1943}
                                                                                                                                     W

                                                                                                                                     §
                                                                                                                                     P3
                                                                                                                                     sa
                                                                                                                                     O
                                                                                                                                     8
                                                                                                                               O
                                                                                                                               w
                                                                                                                               CO
                                                                                                                               r«
                                                                                                                               b
in

-------
               i ..... it'
               	I"
               .ii;:11*';
 III
I,:. illU!	 . i . , H
         MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                         RAW WATER FROM CLINTON
                                         REISEIRVOIR & PUMPING STATION
                                                     CARBON FACILITIES

                                                     FUTURE FLOW SPLITTER
       .	       T'
	. fcOAGULANT
   NPOLYMER
    /PERMANGANATE
	^ (__              I
                                                              LJ
                                                                   FUTURE
                                                                   RAPID MIX
                                                                   NOJ
                                                           /FUTURE
                                                          / PRESED-V
                                                                     1
      WASH
      WATER
      RECOVERY
                   I    ilMENTATION!
                   i    \ BASIN J
                   I      Xs.   ^**^
                   iJYPASSJ^j-^' V



f                          FUTURE*.
                          PRIMARY)
                         yBASINy

                         '   ""* FUTURE
                               RAPID MIX
II;. innr  i  .|l  '!'
Tilitll111 i H . i 'li!.l<
                                       RAPID MIX
                                       NO.
                                               PRESED-
                                              IMENTATION
                                                BASIN
                                               (FLOC.-SED^
                                                      BYPASS
                                               PRIMAR
                                               BASIN
                                                                  POLYMER
                                                                  CARBON
                                                                  SODA ASH
                                                                  COAGULANT
                                                                  PERMANGANATE
                            .RAPID MIX
                            NO. 2
                     ECONDARY
                      BASIN
                     (FLOC.-SED.1
                                                                                     SECONDARY
                                                                                     .  BASIN
     FLUORIDE
     POLYPHOSPHATE
     CHLORINE
                                                                                                  SLUDGE
                                                                                                  CONTROL
                                                                                                  BUILDING
                                    POLYPHOSPHATE
                           TRANSFER,
                           PUMP
                                   WEST HILLS
                                   SERVICE
                                                            CLEARWELL
         CENTRAL
         SERVICE
                                                                                                  •FUTURE
                                                                                                  RESERVOIR
t'HMi:
                                         Figure 3. LAWRENCE, KANSAS - CLINTON WATER TREATJyiENT PLANT FLOW
                                                DIAGRAM AND POINTS OF CHEMICAL APPLICATION.
                                  cases of dysentery. Public health laboratory tests showed that the predominant
                                  organism was tShigelia paradysenteriae. How did this  organism enter  the
                                  syslem? Figure  1 gives essential  details. A  14 inch supply main was out of
                                  service for repairs.  Pressure was "off"  for  12 hours or more.  During  this
                                  period there was a sewer stoppage in a small railroad village where  railroad
                                  maintenance employees were quartered. The sewage backed  up, overflowed
                                  frost-proof nydraritpits and flowed backward into the 14  inch transmission
                                  main. The head available to cause this flow  of sewage into the empty trans-
                                  mission main was at least six feet. How much sewage entered the water main
                                  is unknown, but it was enough. When the transmission "main  was restored to
                                  service, it was given only a minimum flush-out and was not  disinfected.  Re-
                                  calling  that this was a transmission main, when service was  restored, what-

                                  136
                                                                                                                  ;
                                                                                                                  ' ..... i tri
1 111	lililiiiij	J	Is	;;,	,^
	ill	ilill
                            ^	ii	JLAiiiL'i-&.	!,'..!•	I:

-------
                             ENGINEERING CONTROL PRACTICES/P.D. HANEY


 ever sewage was stored in the main, probably quite a bit,  was distributed
 throughout the city. Here is a thing that should not have happened and the
 probability of its occurrence was very low. Nevertheless, it did happen. Here
 is a distribution system disaster that chlorination of the supply  would not have
 averted. One lesson derived from this experience is  that pipes Out of service
 for repair must be flushed thoroughly and disinfected  before1 they can be
 assumed to furnish  safe water. How often is this procedure followed? Often?
 Seldom? Not at all? The answer is, I believe, "often" but not "always". Re-
 member, a water main  out  of service is a serious matter and there is great
 pressure to restore service for public  fire  safety as well as convenience. It is
 the duty of  responsible  professionals to guide repair crews as to the correct
 procedure. Disinfection  of a new or repaired main is no mean feat. It  often
 requires mechanical ingenuity and technical competence of  a  high  order
 (Figure 2).  Another lesson  in "preventive medicine" is that if "frost-proof"
 facilities are installed, they  should not fee drained to a sewer.
   Referring again to Murphy's Law, this  disaster seems to be an illustration
 of the important corollary: "If there is a possibility of things going wrong, the
 one that will go  wrong is the one that will do the most damage". The distri-
 bution system often seems to be a living, thinking thing, well aware of  all of
 Murphy's laws and  how to apply them.                       ;
   If you are not familiar with water works practice  and want tp know more
 about it, I recommend to you an old publication, produced by :the Office of
 Civilian Defense during World War II (1943)—OCD Publication 2022,
 titled "Waterworks  Engineering in Disaster";

 The Lawrence, Kansas Clinton Reservoir Water Treatment Plant

   Figure 3 illustrates this water treatment plant, recently  designed by the
 author's firm. It will have a  capacity of  10 million gallons daily and includes
 softening, as  well as other standard treatment systems.
   The raw water will be obtained from CJinton Reservoir, a 7,000 acre  lake,
 now being constructed by the Corps  of Engineers, four miles southwest  of
 Lawrence,  Kansas. This is  a multi-purpose  reservoir. Its functions include:
 flood control, water  quality  control, recreation, public water supply, and bio-
 support. The dam will  control  367  square miles  of the watershed of the
 Wakarusa River. Storage allocations are  as  follows:

          Flood  Control              267,800 acre-feet
          Multipurpose               129,400
         Sediment                    28,500            ~
         Water Supply               :i 10,400 (36 billion gallons)

   Land use in the watershed is principally  crop land and pasture. Two small,
 sewered  communities are located about  15 miles above the reservoir. Their
 combined population  is about 2500. Both have sewage treatment  plants.
There are also  a few rural mobile home  parks. There  are no significant  agrir
 cultural or industrial waste  point sources. Estimated  mineral  characteristics
 of the impounded water  are  as follows:
                                                                     137

-------
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


                                                          Long-Term Mean Values
                                                 Cations                             Anions
                                                   mg/1     meq/1                     rng/1     meq/1
                                      Ca«*          88-       4.39        HC03-        280       4.59
                                      Mg**          14        1.15        S042-           69       1.44
                                      Na+           14        0.61
                                      K+             3.4      0.09        Cl-             10       0.28
                                                              6.24                  |               6.31

                                    Total hardness is expected to range from 100 to 382 mg/1, carbonate hard-
                                 ness, 100 to 332 mg/1. Coliform and fecal coliform densities will vary widely
                                 with runoff. Coliform MPN values in the range of 200,000 to 300,000/100
                                 ml are likely during peak  inflow periods. The estimated  long-term  fecal-
                                 coliform, goemetric-mean density is expected to be about 50/100 ml. Profe-
                                 ably the most difficult water treatment conditions  will be  those  associated
                                 with algal blooms and related taste-and-odor,  coagulation and filtration prob-
                                 lems. Taste-and-odor  control will, as usual,  be the most  difficult and the
                                 most unpredictable.
                                    Note from the schematic that free residual chlorination will be employed,
                                 with provision for alternative chlorine feed points.  The water will be  am-
                                 moniated after filtration and the free chlorine converted to chloramine.
                                 "  Detention time in the various basins at 10  mgd will be as follows:
                                      Presedimentation

II I 111 ,   II 'i-   .  '	in,  •
Primary *
                                      Secondary

                                    * Based on 20% "split" treatment bypass
                                    This is a symposium on microbiology and is, therefore, concerned with all
                                  those things that contribute to rendering the water "safe". I have no feeling
                                  of uncertainty about the bacterial quality of the water this plant will deliver
                                  to public service. We should have no difficulty meeting the coliform "MCLs  .
                                  In fact, I would be surprised if we did not do much better than, the drinking
                                  water standards require. For viruses, we have  no standards, but  I believe
                                  that if we did have, the plant  effluent would readily meet them.  The plant
                                  will employ excess lime treatment and high lime treatment has been shown
                                  to be quite effective in virus reduction. This may be due to the  combination
                                  of high pH plus the excellent  coagulation  achieved by lime softening  of a
                                  fairly  hard water.  In addition, we  have free-residual chlorination  and  con-
                                  siderable contact time.  We also have,  as back-up, dual media filters. If there
                                  are any viruses present, we should remove  them, destroy  them, or at  least
                                  inactivate them to the point that they are  incapable of producing disease.
                                  While I am  no  authority on viruses, my judgment, based on experience with
                                  water plants and  literature reports, tells me this  plant will  produce "safe"
                                  water.  'The water should  also  be esthetically pleasing and  soft enough for
                                  general domestic use. The turbidity standard also can be met without  diffi-
                                  culty. Continuous turbidity monitoring equipment has been provided.
                                    •••.   :••'   •     •;     •••••"'  .  ,;•  . • '';•,'.,.     :  li1" •  M:I-| !•.      ••'• :,;!''    - •  •••"  :i i
                                  Impact o£ the Standards on Engineering Practice

                                    Turbidity, a light  scattering phenomenon,  is  closely  related  to micro-
                                  biological quality.  When turbidity is measured,  we are indirectly  measuring
                                  138

-------
                             ENGINEERING CONTROL PRACTICES/P.D. HANEY
 participates in the water and particulatess  are important. Turbidity has long
 been viewed as a sign of potential  danger. Certainly,  turbid water, whatever
 the cause, is  unappealing,  if not actually repulsive. Good turbidity removal
 at a treatment plant is reassuring insofar as bacterial and virus hazards are
-concerned. Turbidity is appropriately classed as a  health hazard  and while
 not all turbidity is related to microbiological quality,  often it is closely related.
 Turbidity in finished water may  interfere with chlorine disinfection, prevent
 maintenance of a  residual during distribution, interfere with  bacteriological
 testing, and be a sign of haphazard plant operation and consequent danger of
 water-borne disease.
   I do not believe there is anything in the current EPA microbiological drink-
 ing water standards that changed the design of the  Lawrence, Kansas  plant*
 If, in the future,  a virus standard  is set, monitoring  for viruses will be re-
 quired, and this will call for additional effort by  laboratory personnel. There
 is, however, looming on the horizon the matter  of  microchemical effects of
 chlorination.  Organic  chemists have long known  of chlorine's affinity for
 organics and its ability to enter organic molecules,  forming a variety of chloro-
 substitution products. Only recently have  health authorities  and  the  water
 industry begun to delve into these  rather subtle,  attenuated  chlorine-organic
 reactions in water. We had some experience in the past with chlorophenols
 but this  was considered principally a taste-odor problem without health sig-
 nificance. Regrettably,  we  didn't  look farther,  but the analytical equipment
 and the competence to  run  it have not been generally available for very many
 years.
   We are now confronted with the real possibility  of a standard for halo-
 genated  compounds  in water, such  as:  carbon  tetrachlofide, : chloroform,
 dichloromethane, bromoform, dibromomethane,  and  probably others.  This
 standard doubtlessly will set limits in terms of micrograms rather than  milli-
 grams per liter, or-possibly micromoles per liter. If this standard cannot be
 met by  conventional chlorination  practice,  then modifications will be re-
 quired; however, in making these modifications, we must be ever mindful of
 their effects on microbiological quality  of the finished water. The new micro-
 chemical  standards involving halogenated organics are closely  related to the
 microbiological standards because they  may influence, arid possibly drastically
 alter, chlorination practice.  Engineers ami the water industry have  had a lot
 of  experience with chlorine and  we are not going  to cast it  aside  without
 a great deal of careful  consideration. It is in this  area that engineering  prac-
 tice and  drinking water standards are likely to collide  head-on. Remember,  I
 am speculating about a probable standard  of  unknown magnitude, not an
 existing standard. If we do, in fact, have to curtail chlorination somehow, or
 use untried substitutes, there  will  be great professional concern about micro-
 biological quality. None of us want to solve a microchemical water problem
 and at the same time create a microbiological hazard. Surely,  this  would be
 the negation of sound professional practice.
   What can be done?  Are there chlorine  substitutes  that avoid the micro-
 chemical problem? There apparently are and they are being studied.  Examples
are ozone and chlorine dioxide. Both are powerful disinfectants that insofar
 as we know avoid the possibly harmful organic  reactions that result when
we use chlorine. Both of these substitutes have engineering and possibly other
 disadvantages. First, although, we have had  some experience  with chlorine
 dioxide, we have not had much, and  even less with  ozone. The latter has
been used for many years in Europe and elsewhere,  but there has been little
real engineering experience  with its  generation and application  in U.S. water
works practice. Other potential problems with ozone are lack of  a residual

                                                                      139

-------
         MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                       	        I ••        '       	

         and the inability to store reserve supplies. Ozone must be generated at the
         rate used. Furthermore, we know very little about the reactions of  either
         ozone or chlorine dioxide with trace substances,  notably organics, in water.
         They are both powerful oxidizing agents,  probably fully capable of  reacting
         with many trace organics. What are the reaction products? Are they innocuous
         or  harmful? I don't believe we  can answer this  question today,   and the
         answers may be several years away.
           What other means are available for dealing with the problems that arise
         from present chlorination practice? Remove the organics. This can  be done
         by granular carbon filtration but only at an almost staggering cost and only
         after application of considerable engineering effort. While we know a good
         deal about carbon filtration, there is probably as much to be learned  about
         practical day-to-day, reliable operation, as has been learned.
           We can reduce the reactivity of the chlorine added to water by converting
         it in the water, to chloramine. Here, again, we encounter a penalty. Reduced
         reactivity means  reduced disinfecting power. This means higher residuals and
         a requirement for much longer reaction time. The problem of higher residuals
         would be easyto handle, but providing  the necessary long reaction time can
         be very difficult. Also, chloramines are not spectacularly  effective for de-
         struction or inactivation of viruses.
                                                           j

         Summary
           The  current EPA microbiological standards are essentially those of  past
         years. They are proven standards that can readily be met by sound engineer-
         ing design and good operating practice. The standards  call for considerable
         paperwork and record keeping, to "which many  small  utilities are  unaccus-
         tomed. The  required  paperwork may be a nuisance but  it is  not  all bad.
         It is dismaying and alarming to learn by experience how many  small water
         utilities, and some not so small, keep poor or practically  no water use, quality,
         and equipment records and have no up-to-date maps of their  water  systems.
         The matter of public notification has good and bad features and, doubtlessly,
         there will be considerable controversy about this requirement.
           Present bacteriological standards have not had a major impact on engineer-
         ing design. With  the aid of coagulation, sedimentation, filtration, and chlorina-
         tion, we can meet or exceed these standards without great difficulty. We can
         probably also meet any reasonable virus  standards proposed. Achieving the
         turbidity standard should not be difficult, except for some utilities using un-
         filtered surface supplies. For those,  the technology, and engineering  expertise
         to apply it, are available.  The  problem  is,  therefore,  one of management,
         finance, and public support. The job can  be done, but consumers  must ex-
         pect to pay more for water than in the past. One of the problems  in the water
         industry is that  water,  a truly precious substance, has always,  hi  the past,
         been plentiful and cheap—probably too cheap. This is  changing.
           Any  alteration required  in  chlorination practice, as,  for example, curtail-
         ment or elimination of prechlorination, will have a major impact on engineer-
         ing design. We are ready, willing, and able to apply proven new systems,  as
         required, but we can't do the whole job  alone. I am pleased that the  scientific
         community is, at long last, showing much more  interest in water. Engineers
         must look to the scientists to  define basic water  quality needs  and we invite
         them to work with us in charting a true course. Scientists need the help  of
         engineers and engineers certainly need the help of scientists. Working alone,
         neither group can be completely effective, but working together, there is very
il  4  !  'little we can't do.
         140

-------
                             ENGINEERING CONTROL PRACTICES/P.D. HANEY


                            REFERENCES

 1. Berger,  Bernard B.  et  al. 1970, Engineering evaluation of virus  hazard in
   water. J. Sanit. Eng. Div., Proc. Am. Soc. Civ. Eng. 96:111-161.
 2. Board  of  Directors, AWWA.  Quality  goals  for potable  water.  JAWWA
   €0:1317-1322.
 3. Chang, Shih Lu, G.  Berg, N.A.  Clarke:, and P.W. Kabler.  1960. Survival  and
   protection  against  chlorination  of human enteric  pathogens in  free-living
   nematodes  isolated from water supplies.  Am. J. Trop. Med. Hyg. 9:136-142.
 4. Kehr, Robert W., and C.T. Butterfield. 1943. Notes on  the relation between
   coliforms and enteric pathogens.  Public Health Reports  55:589-607.
 5. Kinnaman,  C.H., and F.C. Beelman.  1944.  An  epidemic  of 3000 cases of
   bacillary dysentery   involving  a  war industry, and  members of  the armed
   forces.  Am. J.  Public Health 34:948-954.
 6. Stoltenberg, G.A.  1970.  Water  quality  in  an  agricultural  watershed.  In:
   Transactions of the 20th Annual Conference on  Sanitary Engineering, Bulle-
   tin No.  62. School of Engineering and Architecture, Lawrence, Kansas.
 7. The  Committee on  Water Quality Criteria. 1972. Water  quality criteria—
   1972, Environmental Studies Board, National Academy of Sciences, National
   Academy of  Engineering,  Washington,  D.C.
 8. United  States Pubjic Health  Service.  1962. Public health  service drinking
   water standards. Public Health Service Publication No. 956. U.S. Department
   of  Health,  Education and Welfare, Washington, D.C.
9. Water  Research Centre.  1976.  Deterioration  of bacteriological  quality  of
   water during  distribution. In:  Notes on water  research. Water  Research
   Centre,  London.
                                                                         141

-------
1          '                   ii

-------
Water Treatment Plant, Jackson, Mississippi
 SESSION 3
               Compliance with the Coliform Standard

   Chairman: Dr. Charles W. Hendricks, Criteria & Standards Division, Office
 of Drinking Water, U.S. Environmental I'rotection Agency, Washington, D.C.
                                                                143

-------
,
             'if

-------
                           DRINKING WATER REGULATIONS/C.W, HENDRICKS


             Compliance with Drinking Water Regulations

                          Charles W. Ilendricks
                         Office of Drinking Water
                  U.S. Environmental Protection Agency
                         Washington, D.C. 20460            '.

   In 1970, the Department of Health, Education and Welfare undertook a
study of public water supply systems (4). The results of this study indicated:
   1.  36% of the tap samples contained one or more bacteriological or chemi-
      cal constituents exceeding the limits in the 1962 Public Health  Service
      Drinking Water Standards.
   2.  85% of the systems did not analyze a sufficient number of bacteriologi-
      cal samples.
   3.  77%  of the plant operators were inadequately trained in fundamental
      water microbiology.
   4.  46% of the plant operators were deficient in knowledge of the chemistry
      relating to their plant operation.
   5.  79% of the systems were not inspected by State or county authorities
      in the year prior to the study.
   For these and other compelling reasons,  Congress  enacted the Safe Drink-
ing Water  Act (2) (PL 93-523) in 1974, "to assure that the water supplied
to the public is safe to drink." The basic structure of the Act, as it pertains
to public water systems, calls for .the establishment of: maximum contaminant
levels for most contaminants in drinking water;  and  criteria and  procedures,
including quality control measures, to assure compliance with such maximum
contaminant levels.

Provisions. of the Safe Drinking Water Act  and The Interim Primary
Drinking Water Regulations

   The Safe Drinking Water Act applies  to  each "Public  Water  System,"
defined in  Section 1401 (4) of the Act as "a system  for the provision to the
public of piped water for human consumption, if such  system has at least
fifteen service connections or regularly serves at least twenty-five individuals."
Both privately and publicly owned water systems are covered,  and the U.S.
Environmental Protection Agency has interpreted service "to the public" to
include factories, churches, schools, and private housing developments  (3).
  The definition of a "public water system" in the National  Interim Primary
Regulations (3) seeks to explain the  meaning of the statutory reference to
"regular" service. It was originally proposed to interpret  the term to include
service for three months out of the year, but this  definition would have ex-
cluded many public  accommodations  such  as campgrounds, lodges and sea-
sonal recreation areas which serve large  numbers  of tourists but  are  in
operation for perhaps less than three months each  year. The definition  in
the final  version covers systems  serving at least twenty-five individuals at least
60 days out of the year.
  Two types of "public water  systems" are  recognized  in the Interim Pri-
mary Regulations—those serving  community residents  and  those  serving
transients for intermittent users. This  differentiation is based on whether the
water served is an individual's major source of drinking water. Although
this concept has more applicability for  chemical than microbiological  con-
taminants,  it is, never the less,  based  upon the fact  that the possible health
effects  of a contaminant in drinking water  in many cases are quite different

                                                                    145

-------
Ill1 11111 1 '1 III 1' ii '•: " " Ml '1
1 111 1 1 ; i 1 i
in Hill i ' i ','.••:'• • 'i1 iii
1 i iiiih i 1 ' i
'1 I
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                             •
     for a person drinking the water for a long period of time than for a person
   •  drinking the water only briefly or intermittently. For microbiological pur-
     poses, however, the risk of infectious disease  is dependent on whether the
     organism is present in the water, and little distinction is made in the Interim
     Primary Regulations for the microbiological monitoring of very small  com-
     munity water systems serving 1000 or -less residents as compared to the non-
     community supplies which serve the more transient users.
       The  Maximum Contaminant Levels (MCL), for  coliform bacteria con-
^i'/'tamed in the Interim Primary Regulations  (Table  1) are  based upon the
     analytical  technique used. A  laboratory may  employ the  membrane  filter
     procedure  or either the ten or one-hundred milliliter fermentation tube proce-
     dure. Any of these procedures may  be  employed at the discretion of the
     State, but only one procedure can be used during any  compliance period
     since the limits prescribed for  the fermentation tube methods are based upon
     tubes containing gas and not values taken from the "most probable number"
     table.
       Limits for coliform bacteria, whether the sample has been analyzed by
     the membrane filter oir the fermentation tube technique, are  expressed in two
     parts. The first  part of the MCL establishes the arithmetic  mean concentra-
     tion of coliform bacteria that can be allowed in water from the system per
     month. In essence this number represents a limit for the general water quality
     of the system as a whole! Tfie• s'econH portion of the MCL seeks to place a
           i                   ;     i  II   II     ill!	^ i iII j || i        i i 	   ,, .  •!	 " • n|j|r.

        TABLE 1.   MAXIMUM MICROBIOLOGICAL CONTAMINANT
            "  .   	'"  LEVELS*
Jilllli 1|	ill'Mi, Illil, ' '
                        	UK „  i'illil
        A. A mean of 1 coliform per 100 ml of water per month.
        B. Four coliforms per 100 ml of water in more than one sample when less
           than, 20 samples are examined per month; or
           Four coUforms per 100 ml of water in more than five  percent of'the
           samples when 20 or more samples  are examined per month.

     FERMENTATION TUBE TECHNIQUE         |

        10 ML Procedure
        A. No coliform bacteria in more than 10 percent of the portions examined
           per month.
        B. No coliform bacteria in three  or  more portions in more than  one
           sample when less than 20 samples are examined per month; or
           No coliform bacteria in three  or more  portions in more than five per-
           cent of the samples when 20 or more samples are examined per month.

        100 ML Procedure

        A. No coliform bacteria in 60 percent of the portions examined per month.
        B. No coliform bacteria  in five  portions in  more  than  one sample
           :y?hen less  than five samples are examined per month; or
           No coliform In  five portions in more  than 20 percent  of the samples
           when five  or more samples are examined per month.
                                    * National Interim Primary Drinking Water Regulations  (40 CFR 141.14)

                                  146"   '             	'   "•  '	'  "	  '-  '	  	'    '

-------
                         DRINKING WATER REGULATIONS/C.W. HENDRICKS
limit  on the number of coliform bacteria that would be  acceptable in any
individual  sample and  represents a limit for water quality at a  specific
sampling point within the distribution system. Failure of any portion of the
limit is sufficient cause to fail the MCL.
   Table 2 is a summary of the major sampling and analytical requirements
for coliform bacteria that are contained in the Interim Primary Regulations.
These provisions instruct the utility in how many samples to take, the analyti-
cal procedure to use and under what  conditions should follow-up (check
samples) be taken. Also included is a provision for the substitution of chlorine
residual  determinations for some of the required coliform determinations
should the States desire to employ it.                        ',
   The sampling  and  analytical requirements  of the  Regulations determine
the extent to which compliance with the maximum contaminant levels will
be maintained. It has been and continues to be the Office of Water Supply
policy to keep the Primary Regulations flexible and to allow for State discre-
tion where possible. For  this reason the analytical procedures  other than
those cited in Standard Methods may be employed if it  can be  shown that
the alternative analytical technique is substantially equivalent to the prescribed
test in precision and accuracy as it relates to the determination of compliance
with the MCL. Likewise, some monitoring requirements may be  reduced by
the State  for small community and non-community systems if local condi-
tions  warrant the reduction.

Problem Areas in the Interim Primary Regulations

   The Maximum Contaminant Levels  for coliform bacteria and their sam-
pling and analytical provisions  as contained in the Interim Primary Regula-
TABLE 2.   MICROBIOLOGICAL CONTAMINANT SAMPLING AND
                  ANALYTICAL REQUIREMENTS.*

MAJOR PROVISIONS

  A.  Analytical  procedures  are  from  Standard Methods,  APHA,  13th
      Edition.
  B.  Sampling frequency is based on population and is specified.
  C.  Sampling frequency for small community systems serving persons 25-
      1000 may.be reduced to  1 per quarter at State discretion.
  D.  Sampling frequency for non-community systems of 1 per quarter may
      be reduced on the basis of a sanitary survey.
  E.  Two consecutive, daily check samples are to be taken if:
      —The  MF count exceeds four coliforms per 100  ml of water.
      —-Coliforms occur in three or more .10 ml portions.
      —Coliforms occur in five 100 ml  portions.
  F.  Chlorine residual analyses may be substituted for  up to 75 percent of
      the coliform determinations at a rate of four chlorine residual determi-
      nations taken for each coliform test substituted. At  least one chlorine
      residual test must be taken per day. _
  G.  Public  notification is required if a MCL or a monitoring frequency is
      violated.
  * National Interim Primary Drinking Water Regulations (40 CFR 141.21)

                                                                    147

-------
               I!1!'' I	I'!1'	"Bit!1!	.'• .aillllt; ' ' nUIIIi ""•!•	 "'"'	i .,- ''In'1!- •,'	';'!';  .1?	H't'iBR	PI	lib I'll! SI! I	rv'ffl	;	IP1;'	t''»!iif'5i'l!iri«	|i|t".ii!|i!''»l™!i|'!"!|" .,'Kfjyw;!' i ':'l
               i;,
               !.
 liiiiii " i •'! '•! "  "

 wr : •;.,! • ' :
!' jilt	
                                               i  \
    MICROBIOL STANDARDS EVALUATlON/C.. HENDRICKS
11 ,11'	,  , i,',  '..	 	!!!  .• "! '  ' " i   i" ;" .1. •'  •'     i ,„ , •   , • ij i' :• ,31	"I	, , ,.''' .,•• II :'!!
                                  tions are certainly not the last wdrd in developing drinking water regulations.
                                  Fqr example, the promulgated MCL's for coliform bacteria are basically the
                                  1962 Public Health Service Standards (5)  with, minor refinements and clari-
                                  fications. However, further changes may be desirable before Revised Regula-
                                  tions are developed. For example
                                  cal methods do not resolve the r
                                  assumed to be present in  a singl
                                  tories assume an upper limit of
                                  individual bacteria to a level of
                                  upper limit assumed will  affect
                                  determine compliance with the
                                   , the MCL's for the membrane filter analyti-
                                   uestion of how many coliform bacteria are
                                   ; highly contaminated sample. Some labora-
                                   50, while others seek to  continue  to count
                                   100 or even higher in a single sample. The
                                   he monthly average which is  calculated to
                                   MCL's.  Another question relating to  the
                                  As  the  regulations are written,
                                  coliform bacteria MCL's is the ^natter of possible spurious positive samples.
                                   all  routine samples  must be counted even
    though  subsequent check samples  may not  indicate  contamination.  The
    reason for this is that bacterial contamination is often intermittent or transient.
    Negative follow-up samples  taken a  day or  more afer a  positive routine
    sample has been analyzed cannot demonstrate  that the positive  result was in
    error. It may be possible, however,  to prescribe  a means  of  dealing  with
    spurious  positive results without compromising the integrity of MCL's.
      A third  question concerning the MCl's for coliform. bacteria is  the  rela-
    tionship  of monthly  averages J>f coliform bacteria  levels to monthly per-
    centages of positive samples. For example, the monthly average MCL for
    the membrane filter Method is violated if the monthly  average exceeds one
    coliform bacterium per sample. However, for purposes of determining whether
    the month-percentage-of-positivejsamples MCL  is violated, a sample is counted
    as positive only if it contains nkore than four coliform bacteria.  Thus, it is
    possible,  particularly when a relatively small number of samples is taken, for
    a system to  fail  the  monthly Average MCL  even when no  single sample
    taken during the month is out i»f compliance  with the limit
      In addition to the questions listed above, this Office has been asked to con-
    sider an amendment to the Interim Primary Regulations that would specifically
    address the case where one positive sample would cause the supply to violate
    the MCL. These requests cite th 2 following.

       1.  For small utilities  (perhaps those  less  than  10,000 persons served
          although the cited  examples are often  those serving 2,400  or  less),
          the collection of one sample requiring the collection of check samples
          may  cause a supply  to :;ail to mee.t the MCL for  that compliance
          period, thus triggering tb.4  public  notification requirements.
      2.  The collection of one such sample followed by  two negative check
          samples should  not  be counted a failure of the MCL  because it could
          have resulted  from improper  sampling technique, from  accidental
          contamination in the laboratory, or the lack of precision in the  culturing
          procedures.                  '	 •	
      3.  Public notification by way of the news media would serve no purpose
          because it would be too long after the fact for consumers to take  self-
          protective actions (even if such action had  been  justified at  the  time
          of  collection of the  sample  in question).  The notice may also unneces-
          sarily upset and alarm consumers, and would ultimately be self-defeating
          as  consumers might develop  a  "no-hum"  attitude when  it  became
                                        apparent no waterborne
                                        tamination notices.

                                    Certainly these comments are
                                  is not addressed in the Interim
                                  148
                                   disease transmission accompanied  the  con-
                                  significant because they point to an area that
                                  Primary Regulations. I would hope that dis-
                                                                              t • w	

-------
                          DRINKING WATER REGULATIONS/C.W. HENDRICKS
cussion at this symposium would focus  on
results especially for small systems taking 1
what to  do about false positive
!ss than ten samples per month.
 In addition  to this  question, the  following  also warrant discussion  and
 comment.

   1.  Validity of the Coliform Indicator Syjstem
      —Are we using the proper indicator system?
      —Do the present limits protect health?
      —Is it possible to calculate risk from these limits?
      -^-Can non-biological tests be sulbstitutjed?              ;
   2.  Sampling Frequency and the Arithmet c Mean         |
      —Is the  calculation of a mean appro] >riate for  small  systems?
      —How many samples are enough?
      —How firm is the technical basis; for t le present monitoring frequency?
   3.  Compliance With the  Coliform Bacteia MCL
      —Is the  two part limit practicable?
      —Should follow-up (check) samples be used in calculation of the mean?
      —What is the significance of two negative follow-up samples?
      —Is the MCL properly established on I a technical basis?
      —Can the limit be met for all systems?

   In  conclusion,  the  Interim Primary  Regulations  for microbiology  are
basically the same as those contained in  I he 1962  Public Health Service
Standards with certain minor  clarifications. These Regulations are adequate,
in my opinion, to protect the public health providing the water comes from a
protected, uncontaminated source. There ar; certain  desirable modifications
that can be made by amendment, others, hotoever, must await further study.
Hopefully this symposium will provide guidance to the Agency on  how these
modifications should be made.
                           REFERENCES
 1. House of Representatives. 1974. Safes Driri
    No. 93-1185. 93rd Congress, 2nd session,
 2. The safe  drinking  water act, title XIV-Sa:
ding Water Act, pp. 16-17.  Report
iVashington, D.C.
rety of public water  systems,  (PL
                                         States, Washington, D.C.
                                          Agency.  1975.  National
                        interim
   93-523). 1974. Congress  of the  United
3.  United  States Environmental  Protection
   primary drinking water regulations. Fed. Register 40:59566-59588.
4.  United States Public Health Service. 1970.1 Community water supply study-
   significance of national findings. United States Department of Health, Educa-
   tion, and -Welfare, Washington, D.C.
 5. United States Public Health Service. 1962.
   States Department of Health, Education, and Welfare, Washington, D.C.

            QUESTION AND ANSWER  SESSION

  W.B. Jeffcoat, Alabama Department of Public Health,  Montgomery, Ala-
bama.
  Your .statement...!!! session^ 3,  concerning
the regulation is quite true: We in Alabama have  relayed our concern to
Region IV on the same matters. You onlj
however. The numerical value of '"TNTC"
mentioned. You stated  that help had  been
 Drinking water standards. United
 the three problem areas with
  addressed one of these areas,
 and the dual MCL's  were not
sought to solve  these problems.
Help from where? EPA? States? Universities? Water Systems? Did the help

                                                                     149

-------
IMF!	BISEilB1!"
             HIM '• iipj'ii ..... p, j"s "" > ....... j ...... if ' .: ......... "P • ..... ; 'K" ; :T! ..... ..... t ....... ] ...... 'Wi ,s ..... 7i:i«K w ..... ' ? :"i ...... in ..... ' f >. ' ...... • m f'-
                                                             t ...... iiRiisf iiiisif am>" iiiifE ...... «' :VKV ..... f .• ...... •'• ..... i ....... :^
                                                                                           • • w: ?! -SK
               MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
               come from  people  who  are familiar with the operation  of  a water supply
               system or. did it come from EPA officials  and University personnel whose
               only contact with a water system  is an occasional  visit.
                 I'm sure you  realize that federal agencies especially EPA  have low levels
               of credibility. This is something we wish to  avoid on the State level. Forcing
               a water supply to comply with unrealistic requirements will  place the State
               in the same category of EPA. The goal of the  SDWA is "safe drinking water
               for all"  and to  accomplish this  goal  cooperation  is needed between  the
               State  and water system: There is  more to operating a water system  than
               merely complying with  the standards. I agree with  Mr. Taylor and Mr.
               Harrison that the mean  should not be used in determining  public notifica-
               tions. Check samples should also be counted as they do with other parame-
               ters hi the National Interim Primary Drinking Water Regulations. My main
               question  is  will  EPA consider the proposals  of the utilities and  States  or
               will EPA rely on information provided by those unfamiliar with the actual,
               daily operation of public water systems and with the actual,  daily  operation
               of a state public water supply  supervision program?

                 C.W. Hendricks, Office of Drinking Water,  ILS. Environmental Protection
               Agency, Washington, D.C.
                 This meeting  was designed  to provide the U.S. Environmental Protection
               Agency with the best information that experts in  Water Supply  from EPA,
               the States, Universities and water  utilities can provide. I do  not believe that
               any one  person, or profession  for that matter, has all of the information at
               hand needed to correct  the deficiencies that  we see in the  colifonn maxi-
               mum contaminant  levels, I would anticipate that any  amendments to the
               National Interim  Primary Regulations would  be thoroughly reviewed  by
               State and utility personnel before they are promulgated.
                                                                   i

                 G.W. Fuhs, New York Department of Health, Albany, New York.

                 How did you determine there was  no  danger  connected with  coliform
           '"'si  standard?
           !|i,,!l!lll!  ' ''  '  '.	'"''I' ill.'1.  I1 	       ".  II,         ,' ,, ,   i.11" J  . fill	 	'» .ill1  I I  'I    i1 I'P  '  I    ':
                 C.W. Hendricks.
                 I would refer  you to Mr. Floyd Taylor's paper elsewhere in this manual for
               a description  of how  our present colifonn standard originated.  I  suspect
               that the  standards were empirically derived without any study to determine
               what the risk to health might  be. I might add that experience has  shown us
               that the present  coliform standards adequately protect the public  health pro-
               viding a  high quality uncontaminated source water is used. Any substantive
               changes  m  the  MCL will require careful  evaluation before they  are pro-
               mulgated.
                 Comment by  E.E. Geldreich, Water Supply Research Laboratory/MERL,
               U.S. Environmental Protection Agency, Cincinnati,  Ohio.

                 If I may, I would like to respond to this  presentation by Dr. Hendricks in
               a series of short comments on the many aspects  of the subject.
                 a. Two Classes of Water Quality
                 The proposed amendments  in the bacteriological regulations are  drifting
               into  a direction  of establishing two classes of public water quality in the
               Nation with a disregard for identifying poor quality waters in smaller sys-
               tems. While this may ease  the burden of small  water supply surveillance on
               the State agencies,  the proposed changes will weaken the goal of achieving

            "!	  150   '           '"            "         ':'''	   '  '   !":" 	'':   "'  '     "'   ' "''"

-------
                          DRINKING WATER REGULATIONS/C.W. HENDRICKS
 uniform, high quality Water through minimal regulations for all public water
 supply systems. Furthermore, these relaxed  regulations  would be applied to
 the  very water systems  that our 1969  national  community water supply
 survey indicated were  most subject  to producing marginal  quality potable
 water.
   b.  Variance in MF-MPN Test Application
   As  the regulations now  stand  there is a  variance in the  application of
 rules related  to the  membrane filter versus  the  multiple tube test and an
 uneven interpretation of bacteriological results when using  either  of these
 two alternative  bacteriological procedures.
   Since either of these procedures (MF  or  MPN) may be used in the ex-
 amination of  potable water, all  results should be reporting MF  results in
 terms of coliforms per  100 ml. The: present practice of reporting  MF re-
 sults in terms  of coliforms per  100 ml while  restricting MPN reports to only
 the number of  positive tubes, results  in  some inequities of  interpretation.
 Every laboratory that is capable of performing the multiple tube test is also
 capable of computing the proper  MPN value. The use  of a system  that re-
 ports only the number of positive  tubes per test is archaic, and easily subject
 to manipulation to meet a monthly minimum  and should be discarded.
   c. Recognition of  Low-Level Contamination Occurrences
   The  definition of  an unsatisfactory sample should be "1  per 100 ml"
 for the membrane filter test and "2.2 per 100 ml" for the multiple tube test.
 These limits are equivalent  since the  MPN test value is  a statistical estimate
 and uses a table of  numbers elevated by incorporation of a positive bias.
 This revision  in the  definition  of an unsatisfactory sample would  direct at-
 tention to the occurrence  of  persistent  low level contamination  problems
 that have been ignored in the past. This change would not alter the intent of
 the proposed  amendment as  it relates to repeat  sampling since  the term
 "negative" is  interpreted to mean that no coliforms were  detected  in  the
 official  test portions  by  either the MPN or MF procedure. The  change
 would require mandatory  check  samples for all detected occurrences of
 coliforms in public water supplies.
   d. Sample Frequency for Small Supplies
   The proposed  reduction of  sampling  frequency  for small  community
 systems serving 25  to 1000 persons to one sample per  quarter- and further
 reducing for non-community systems of one  per  quarter on  the basis of a
 sanitary  survey  should  not be permitted. Since  the  number of  supplies
serving 25 to  100 persons is largely unknown and these supplies are mostly _
 comprised of small trailer parks and roadside facilities that serve a transient
population, sampling  should only .be reduced to  once each quarter during
non-recreational periods  of  the year,  however, increased to once per month
 during the normal recreation period.
   Since  the burden of increased  monitoring and  corrective  action  will be
on public water supplies serving populations  between 25 and 9,400, guide-
line dates should be  established to intensify  program direction first  to sup-.
plies serving 1,000 to 9,400 people,  .then to supplies in the  100 to 1,000
population category,  leaving the  smallest supplies  (25 to  100 population
served) as the last group to be phased into the program.
   e. Check Sample Incentives
  Check  samples  should be included  in  the monthly  average  of  routine
samples for all water supplies, regardless  of size. This would; offer incentive
to water plant management to increase their  sampling program beyond  the
minimum two consecutive  daily check samples required following  an un-
satisfactory sample  report. Because of the urgency to detect contamination

                                                                      151

-------
                  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
               'in'1'1'!  'I "   " „, '  !  ;   ..      ,,', .....   •.''''.  '' '":  '  'i';1  '• •  '?  'ij;/1' i ""I"!1,: '  ,:'*•) '''  li '
f"	:'
S'        >:   ,    ,     •  • ,    • ' ,  ,  ',',' i  • ,i .' i " '   ''. , , • i  , ,   • . , i<,;  ,.,11 .M,: •   ,  :•  ,•  >!„
 breakthroughs into the potable water, requests for a repeat sampling following
 unsatisfactory results should  be made i. promptly and the  sampling initiated
 within .no more than 48 hours. The answer to the perennial concern about
 the integrity of the sampling during collection  could  best be resolved by
 stating a requirement that sample  collectors  must be  certified  by the ap-
 propriate state agency following successful completion of a 1-2 day training
 course. Thereafter, recertification could be related to observed performance
 of sample collectors  during the recertification of the  laboratory.
   f. Public  Notification
   The public  notification requirements  should  be  restructured to  include
 two levels of  response.  The first level of  response occurs  when one  to  four
 cbliforms are  detected  by the MF  method or 2.2 to  5.1 coliforms  are re-
 ported in the  MPN procedure.  In this case, the  water system  must institute
 aft immediate review of treatment practices, including examination and clean-
 ing (if necessary) of the distribution system. Daily check samples are  manda-
 tory during both  the sanitary  survey and the period  of corrective action
 until  two consecutive satisfactory samples are obtained.  A  written  record
 of  the survey report,  corrective action(s) and bacteriological  results are
 maintained  for five years. The second level  of response  involves any oc-
 currences of 5 or more coliforms per 100 ml  on the MF or above  5.1 per
 100 ml as measured by the  MPN  test in one or more  of the daily check
 samples. In  this later situation, public notification would be mandatory, in-
 cluding a  "boil water order," if necessary, immediate increases in disinfec-
 tion to achieve a disinfection residual  in. the most remote sections  of the
 system, hourly measurements of disinfectant residuals and daily bacteriologi-
 cal, sampling at these remote sections. As soon  as the monitoring data on
 the distribution network demonstrates a disinfectant residual for 24 continu-
 ous hours, and two consecutive days of satisfactory bacteriological results, a
 public notice cancelling the "boil water order" should be  issued.
   g. Program Approach
   In essence,  while portions  of the regulations need to  be changed to re-
 solve  problems in monitoring and interpretation  of laboratory results, these
 modifications must not  drift toward establishing two classes of public water
 quality in the  Nation with a disregard for identifying and responding to
 poor quality waters in smaller systems. A systematic approach to attacking
 the increased program responsibility should be geared to solving current con-
 straints in state and  local water  supply staffing and monitoring  costs, not
 in minimizing objectives defined by intent of the law.
                 152

-------
                   COLIFORM STANEiARD: INTERSTATE CARRIERS/F. TAYLOR


             Experiences With the Coliform Standard Under the
                        Interstate Carrier Program

                               Floyd Taylor

                       Chief,  Water Supply Branch
                   U.S. Environmental Protection Agency
                       Boston, Massachusetts 02203

 Introduction

   The past often provides the  record  of experience which  is  useful  in
 guiding decisions on present and future courses of action. History is also
 littered with the wreckage of men and systems who ignored its lessons and
 this of course is what We seek  to  avoid by the intensive examinations we are
 giving this  "Evaluation of the  Microbiology Standards  for Drinking Water."
 Our particular part of this examination deals with experience with the coli-
 form  standard  under the interstate carrier certification program conducted
 for 56 years by the U.S. Public Health Service and for seven; by the Environ-
 mental Protection  Agency. I believe it will be helpful to  briefly review the
 development of the coliform standard as it appeared in successive issues of
 •the U.S.P.H.S. Drinking Water Standards.
 The Coliform Standard

   In  1914 the entire  document,  commonly  referred to as  the  Treasury
 Standards, was entitled "Bacteriological Standard for Drinking Water," under
 which Secretary McAdoo informed the Surgeon General that "In the future
 common  carriers will be required to furnish water for passengers in inter-
 state traffic which will conform to this standard."  (5) The sections of that
 standard  pertinent  to today's discussion read  as follows:  "Not more  than
 one out of five lOcc portions of any sample examined shall show the presence
 of  organisms of the Bacillus coli group when  tested as follows:" and "The
 total number of bacteria developing on  standard agar plates,  incubated 24
 hours  at 37°C  shall not  exceed 100 per cubic centimeter." (6) It should be
 noted  that no mention was made of the number of samples to be taken nor
 of the deviations from the above stated requirements which would appear in
 later reyisions._The  next Treasury Department document was entitled "Drink-
 ing Water Standards" and we see  the beginning of the present day practice
 for it  read, in part, as follows:
             "II. AS TO BACTERIOLOGICAL QUALITY
  "1.  Of all the standard (lOcc)-portions examined in accordance  with the
 procedure specified  below, not more than 10 percent shall show the presence
 of organisms of the> JB. coli group.   .
  "2.  Occasionally  three or more  of the five equal (lOcc)  portions consti-
 tuting  a single standard sample may show the presence of E. '• coli. This shall
 not be allowable if it occurs in more than
  (a)  Five percent of the standard samples when 20 or more samples have
     - been examined;
  (b)  One standard sample  when less than 20 samples have been examined.
Note—It is to be understood that in the examination of any water supply the
series of samples must conform to both the above requirements 1 and 2.  For
example,  where the total number of samples is less than six, the occurrence of

                                                                    153

-------
lIlliKf" f ! ! 1 I!11!"1!:,,
Pi!!' '' , ".
!('»:;• ' i .
1 	 IIPK 1 	 i 	 Sfil'i'i!' 	 'i1'"1! 	 •III!1!!1111"' lillill11'!'!!1!';'!:'- 	 If '• 	 Ti 	 !' ' 	 I1:';
" ' si;!, :;;,, .'-yv f 'H'f fia,,"!' 	
	 in;; 	 	 i 	 	 I'1!, J ;•,!" Ij
- !? "i !;' "'i 	 " ! i: &<'• 'ijj 	 !• "CM" ' " fjf'i .«•. . • i",,1";- 	 "«' •); 	 !i;i W'i ("iji •!? '!" ' ' Hjf i I'-faf!
: .; ! ."'• ..'..', 	 	 '' t:- .i*'''' ''!' f-'W
;l . . . .•:,'..' ' t I' J
llliriiliTlLiniljiiliil 	 I'lli'llT ; r .."!:» '.ii;. .IIIUII1 1. ::i :' .1 n 	 I.'.)1;.. II II inn II null
' !!!-.' "'- I;-1' ' Si1'!!:,'!"'' ' ' :i 1 ' 'I" I1 1 1 1
'" . . : '' "'::' '"; ' '" ' " ', "'"";,' ! ''''i"11' !':"' • 'l'"l':!|;!.li
")! I'1;;	, I	fl '
                               MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
t 111! ! .;! 1
                          til'
positive tests in three or more of the five portions of any single sample, al-
though it would be permitted under requirement 2, would constitute a failure
to meet requirement  1."
   In the 1942 and 1946 document by then called the Public Health Service
Drinking Water Standards — standard portions and standard samples  were
defined  and the 100 ml standard  sample was introduced as were the  first
requirements as  to the  number of samples which should be collected.  The
\yell known graph showing the relation between population density and num-
ber of samples to be collected  per month appeared in  the  1942 standards.
It will be helpful to note the slight changes  in the allowed limits as they
have bearing on today's  discussion. The requirement was as follows:
   "12.203 (b) (1) of  all  the standard ten milliliter (10 ml) portions examined
per month  in accordance with the specified procedure, not more than ten
(10) percent shall show the presence of organisms of the coliform group."
   "12.203 (b) (2) occasionally three (3) or  more  of the five (5) equal ten
milliliter (10 ml) portions constituting  a single standard sample may show
the presence of organisms of the coliform group, provided that this shall not
be allowable if it occurs in consecutive samples or in more than
   (i) Five (3) percent of the standard  samples when twenty (20) or more
      samples have been examined per month.
   (ii) One (1)  standard  sample when  less than  twenty (20) samples  have
      been examined per month.
   Provided further that when three or  more of  the five equal ten milliliter
(10 ml) portions constituting a single  standard sample show the presence
of organisms of  the  coliform group, daily samples from the same sampling
point shall be collected promptly  and  examined until  the  results obtained
from at least two consecutive samples show the  water to be of  satisfactory
quality." It is not  necessary to  emphasize the  requirements for  100  ml
sampling as this system has not  been widely adopted even though it has been
carried over into the EPA Interim Primary  Drinking Water Regulations. It
will be realized that  the 10' ml multiple tube  series for the  1925 and  1942
standards were specifying an average of  less than 2.2 coliform organisms per
100 ml with occasional allowances as high as over 16  organisms per  100
ml, but, beginning m 1942, only if check samples (two consecutive)  were
negative. In other words  a high sample was  the signal to take corrective
action.
   In the last (1962) PHS Standards  the general wording of the 1946  edi-
tion was continued but with the further  restriction that "the presence of the
coliform group in three or  more  10  ml portions  shall not be allowable if
this occurs in two consecutive  samples" (2).  Also in 1962 the membrane
filter technique was recognized with the requirements not to exceed an average
of one colony per 100 ml (monthly arithmetic)  and on occasion not more
than 4/100 ml.  And that is how we arrived  where we are today with the
bacteriological requirements of the Interim Primary Drinking Water Standards
wnick are  essentially those  of  the  1962 PHS Standards.  I  would however
point out that the 1914 document and its successors,  or rather the applica-
tion of  them together with the practice of  filtration  and disinfection  were
responsible for the  dramatic reduction of  enteric disease  in this  country
during the  early 1900's.

Coverage of the Standards
                                                     I
   When the  first federal  drinking water standards were promulgated there
were in 1916, 4077 U.S. public water supplies  which came under their juris-
                               154

-------
                  COLIFORM STANDARD: INTERSTATE CARRIERS/F. TAYLOR


 diction. It is estimated that they then served about  35% of Americans  who
 utilized public water supplies. In 1975 the number of Interstate Carrier  Wa-
 ter Supplies was 689 serving 90,000,000 people or about 55% of the popula-
 tion on public water supplies. I would also cite the historic policy by which
 the standards were  applied namely through a  cooperative, federal-state  pro-
 gram. Even more importantly,  the  standards were adapted by nearly all
 the states as applicable to public water supplies generally within their jurisdic-
 tions.  However, judging from  experience in New  England, this application
 was administratively and not legally  enforced. Now a word about  the  gen-
 eral procedures for  applying the bacteriological standards.

 Application of Standards

   Records  of the  precise methods  of  applying  the  first  drinking water
 standards were difficult to find but Treasury Department Circular No. 48,
 October 5,  1912 directed  PHS  commissioned  officers traveling on common
 carriers to make observations regarding sanitary conditions  and submit ap-
 propriate reports  (1).  More importantly, the  states  were required to make
 reports to .the Service on the  quality of the  sources and presumably gave
 warning when the quality of a  supply failed  to meet the requirements. Let
 me here state for the benefit of those who  are not  aware of how the Inter-
 state Carrier Water  System worked,  the enforcement phase has always been
 an action between the federal government and the  common: carrier  not be-
 tween the federal government and the water  purveyor.  Furthermore, since
 the  early days of the system  the  states made recommendations as to the
 type of action which should be taken.
   Sometime between 1925 and 1938, when my personal experience with the
 system began, a three level method for approving carrier water supplies  was
 adopted.  First there -was - full approval, - second  provisional approval  and
 third, prohibited. The first  meant that all was well, the second that there  was
 some question about the capability of the supply to continuously provide safe
 water  and prohibited meant  what it  said. What the classification would be
 was somewhat arbitrarily determined  by  discussions  between the state health
 engineer and the PHS district engineer.
   Beginning  in 1942 the  concept  of numbers of samples per month  and
 allowable ratios of coliforms were adopted  and an  annual report form  was
 introduced which required the State to  show for each month what the bac-
 teriological  record had  been. However  these  reports were  submitted  only
 once a year to PHS  and_"bad" results (which  by that time could be a year
 or more old) were usually explained by a footnote. Invariably the supply was,
 at worst,  "provisionally  approved." With the  great  shift during this period
 of personnel and respurces to water pollution  control activities at both state
 and federal levels, interest in water supply  declined  almost proportionately.
 By 1959 many state and Federal water  supply programs had  fallen into  a
 state of low effectiveness. In that year the Public Health Service water supply
 program was placed  under new management.
  In  1962 Public Health Service Headquarters Water Supply  Program  ad-
 vised all the regional offices that the bacteriological requirements *  of  the
 standards meant what they said and would be enforced. In the ensuing years
 the numbers of provisional  and prohibited actions increased  markedly. It
was also during those years that the Service began to keep track of  the wa-
  * HEW General Counsel determined that only the bacteriological provisions of
the standards were enforceable.

                                                                     155

-------
i inijjjii,!!1 '  , r :h;>i'»,, i	ill"   ,i,  if, |
i!'1 iiiiiis i  i ',i:' iiS,  • \  i'1"1
 111,1''.	"i
Ik1 ill
 (ill
               II!!1,
 II Illlli
                                MICROBIOL STANDAR 3S EVALUATION/C.W. HENDRICKS
                                ter supply  improvementsthat resulted  from  all enforcement actions  and
                                although I cannot quote  you exact data I can  tell you that the rate of  new
                                water  supply construction resulting from the program was of the order of
                                several millions of d
                                   In 1971 when the
                                to  EPA the first sy
                       liars per year.
                       interstate Carrier Water Supply Program had gone over
                       tematic guide to the Interstate Carrier Water Supply
                                 Certification Program appeared entitled "A  Guide to the Interstate Carrier
                                 Water Supply Certification  Program (7). It provided valuable criteria  for
                                 certifying Interstate Carrier Water Supplies including both bacteriological and
                                chemical. Again the
   and the basic elemen s were  as follows:
     ':'"!, ' " "!'!•!' ''"l; iJit'",""" 'i" ' ": iii"
    For Approved:
            ;, ,|,|i,i ,;,, ,,,,; • ,,      ,   „,          ,,,, ,,„„ , „	, ,„	,   , ,,	 „,,,,, l , ,, ,    ,        ....
            r quality must meet all the limits of Section 3.2  (sections  as they
             "          2 standards)  and must meet the  sampling ratios  of
                                   appear in  the  19
                                   Section 3.1 at leas
                                  For Provisional A
                                     Water quality
                                   12 month  reportir
                                   for 2 or more months
                       iproval:

                     f iled  to meet Section 3.2 requirements for one of any
                         period and/or monitoring  requirements were failed
                                  Use Prohibited:
       Water quality :
     months  of any  1
     violated by  failure
     samples  for  3 mori
  To get back to a  "p
  ply must have a recc
  bacteriological resultr
     I can not speak fo:
  Regjon I but I can si
  grading of monitorin
  tions. In one state w
I Hill :*••	.'Hi.jvwjjs.,	a	i 'i	i
  the public water supp
                                               equipments
                                                treatment,
  '50%  did meet them.
    I would also like
  bacteriological' 'n
  hibited actions which
  field,  Massachusetts
  1969  Springfield, a ci
  tion as the only
  the PHS, standards
  adequate watershed j
  $125,000 automatic,
  teriological record ha
    In, the summer of
  of its distribution  53
  100,000  inhabitants.
  refused  to	change tl
 'I'"' Today'Fall" River "is
  million most of whid
                                156
               "i	
                              	I	
                       ormer are of particular interest to us in this discussion
                        11 out of each 12 month period.
                                                     iled  to  meet Section 3.2 requirements for 2 or  more
                                                      month period  and/or monitoring  requirements  were
                                                      to obtain  less than 50% of the required number of
                                                     hs of any  12 month period.
                                                     rovisional"  from a "prohibited" classification the sup-
                                                     rd of at least three consecutive months of satisfactory
                                                     and  sampling frequency.
                                                     the impact of the systematized requirements outside of
                                                     y that there it has resulted in better reporting, and up-
                                                      frequency and the improvement of plants and opera-
                                                      have seen an improvement  from 1970 when 70% of
                                                     ies did not  meet bacteriological standards to 1976 when
                                                     to  cite two  cases  in  point  where failure  to  meet the
                                                           of the Drinking Water Standards resulted in pro-
                                                     encouraged needed improvements. The first was Spring-
                                                    und the second Pall River,  Massachusetts. Until about
                                                      of some 160,000 provided water with slow sand filtra-
                                                         . Frequent instances of  coliform levels in excess of
                                                     d been noted  and a sanitary survey revealed less than
                                                     rotection.  The city was requested to, and did, install a
                                                     chlorination station since which installation the bac-
                                                      been excellent.
                                                      969 Fall River experienced widespread contamination
                                                     stem and the  state issued  boiled water  orders to its
                                                     The PHS issued a "Prohibited Use"  certification  and
                                                     e classification until  adequate  treatment  was assured.
                                                    using a modern water  filter plant which cost some $4.5
                                                      was paid by  a HUD grant.
                                                                             vilillt'-V
                                                                             ,,=4*1	-H

-------
                  COLIFORM STANDARD: INTERSTATE CARRIERS/F. TAYLOR
Conclusions
   I believe I have complied with my instructions to  tell you how the bac-
teriological limits of the interstate carrier standards were implemented from
Public Health Service Headquarters and regional offices and what were some
of the results. The third  question is more difficult. "Should we continue to
implement the limit for communities  in  the  same way  as  indicated in the
interim regulations: my recommendation  is "yes"  as far as the numbers of
allowable coliforms are concerned and; "no" as concerning the  monitoring
and public notification requirements.
   Let me next  say that  some of you would have recognized  a similarity
between the "provisional approval" of the old  ICWS system and the variance
and exemption  provisions of  the SDWA. These allowed room  or  time in
which to negotiate. I firmly believe this room to negotiate: plus the Use Pro-
hibited  action  clause, where needed, accounted for our
ICWS program.  These principles  should continue to govern our program.
                           REFERENCES

 1. Kent,  F.S.  1959. Administration  of  the  interstate quarantine regulations
   governing  drinking water.  Interstate  Carrier  Section,  Creneral  Engineering
   Program,  United States Public  Health  Service, Washington, D.C. (Available
   from Tloyd Taylor, United States Environmental Protection Agency, Boston,
   Massachusetts)
                                       success  under the
 2. United States Public Health Service.  1962. Drinking wat
                                      ;r standards. United
   States Department of Health, Education, and Welfare, Washington, D.C.
 3. United States Public Health Service. 1925. Drinking water
                                       standards. Treasury
   Department. Public Health Reports 40:696.
 4. United States- Public Health-Service.  1947. New  interstate quarantine regu-
   lations.  Federal Security Agency. Fed. Register 12:3189-3197.
 5. United  States  Public  Health Service.  1914.  Bacteriolc
   drinking water. Public Health Reports 29:2959.
 6. Ibid, p. 2960.
 7. Water  Supply Division.  1973. A guide to  the interstate
   certification, p.
   ington, D.C.
                                      igical  standard for
                                      carrier water supply
4,  5. United States Environmental Protection Agency,  Wash-
                                                                      157

-------
                                                                                                                                                                                                               ii!1:
                                                                                                    .11" t
                                                                                                                                                                                                      m
I;1  ..;  "I'l,	;	!"  .a;.; [ *MI'
   ':,-:.-:;  i:,	I;   IP luf ,,,!<:,,v,  ,
            '!'!*'   iiat i'1-;
                  "
                                                   -f      V
                                                                                                                                                                                                                                                              'I       !  "..|il.ir!!  lill
                                                                                                                                                                                                                                                        .•'hi!'  1  :    i  "li iiiifl

                                                                                                                                                                                                                                                        ":|{I;I!"1 !!"  !'  •  iilflli ;
                                                                                                                                                                                             \'  •     Ills:   'lif
                                                                                                                                                                                              ••' ,;   Hi:*-!     f:	
                                                                                                                                                                                                      ll|>!llll  !>    "Mlliv
                                                                                                                                                                                                                                                              i   •    ,  •  ji'ij,	;,;!
lull ill
1 *

-------
                               COLIFORM STANDARD: REGIONS/J. HARRISON


                  Experiences with the Coliform Standard

                              Joseph Harrison
                           Water Supply Branch
                  U.S. Environmental Protection Agency
                          Chicago, Illinois 60604
 Introduction
   When Dr. Charles Hendricks invited  me—almost two  months  ago—to
 participate in this symposium  regarding standards for bacteriological quality
 of drinking water, I in turn asked each of  the nine other EPA Regions and a
 number of States to help  me  gather the facts and figures that would docu-
 ment  experiences with the historic coliform standard.  I have  received re-
 sponses from most  Regions and about eleven different State agencies.
   I will describe some typical experiences that public water systems have in
 finding coliform bacteria,  and how various State agencies and  EPA regions
 have treated these findings in  the past. There has been more diversity than
 consistency in the  supervision of bacteriological  quality. But  I think you
 will agree that  no State or EPA Region  has ever been  as  stringent as the
 National Interim Primary  Drinking Water Regulations (NIPDWR)  are now
 asking us to be.
   It is not fair for me to take the entire credit for this paper as it is almost
 entirely pieced together from the inputs of others".  A  number of people have
 written me ease history descriptions  of  bacteriological problems including
 follow-up findings and actions. I will quote a number of them  directly. The
 cases illustrate some important points  and  I am sure  you will enjoy them as
 I did.                 ; ;    	   -  -
   I guess one reason-that  I am here toda.y is that I  have some very strong
 feeling about the  current  microbiological  standard. It is not logical; water
 systems cannot comply with it; and it cannot be enforced as currently written
 in 40 CFR 141.14 and 141.21.
   Coliform bacteria in drinking water represent an acute  rather than a long
 term health effect. A  single swallow of coliform is just  as  dangerous as a
 monthly average exposure. The bacteriological safety of drinking water should
 not be defined by an arithmetic mean of all samples  or percent of positive
 portions over a month's period. The individual sample limits define the safety
 of drinking water, whereas a  monthly or yearly  average coliform limit  per-
 tains to the reliability of drinking water systems.  By applying both limits, as
 we have in the December 24 NIPDWR, we end up with a double standard.
 Public water  system consumers  can,  according to our  current  standards,
 safely  drink water with four coliforms per 100 milliliters every day of the
 month until the  last  day when this same water suddenly becomes unsafe be-
 cause the computed monthly average exceeds one coliform per 100 ml.
  It is a good idea to use long term averages or percentages of bacteriological
 data to judge  the adequacy of water system facilities  and operations. But I
 am absolutely convinced we must use only the individual sample limits to
 define safe drinking water. Monthly summaries of bacteriological data should
 do no more than trigger investigative action to ascetrain that a serious health
hazard is not lurking on the sideline arid waiting to emerge. It is not proper
 to define water with four coliforms per 100 milliliters  as safe  to drink and at
the same time say water with more than one coliform is unsafe.
  The  second major  flaw  regards  public notification.The  December 24
regulations require that the public be notified on the basis of  unverified bac-

                                                                     159

-------
Il1' lil1 !,,,
iiui ni'M
               ;':*
(I	)
I,Bill	
:,: sit
mini
81
                                     MICROBIOL STANI >ARDS EVALUATION/CJW. HENDRICKS
                                     teriological data.
       ICJLlUJliJKIwCU. vicii-ci.  LI..L  juiAWfcj  «u  MV——.~-- — - -	^  f            j          ^
       unsatisfactory bacteriological sample result will force nearly  all public water
       systems in this country to give public notice.
                                       P.L. 93-523 pu
        ,, i* | f_ y j~*j£*j I-"-*- -*nw .i,LWfc.*i*wt*fcAv/AJ. **p*»jjf *"*»*»*Q —— —— —	
       such as bacteria and nitrates is .not meant as a warning for consumers to take
            	  __i:	:* lt.~:_',»«,'„.  V.S.a1f  Qtofpo will still TiaVfi to 1SSU6 dlSCre-
                                     emergency action
                                     pendent of SDWA
                        .n  fact,  as  Section 141.21 (f)  is currently written, a single
                        ilic notification respecting acute health effect contaminants
                    their own behalf. States  will still have to issue  discre-
       cmcr&cLiw y  O.UUX/.LI m niv-*.*  w T? **  ^ w ».*«.....•-• ~«~ -—-  •• — 	 -               _
       tionary boil orders and other  emergency  advisories for this  purpose, inde-
                         requirements.
       LJCJL1UWUL \Ji. \JL-f YY.I . JH^V*l.4.wi**«.*..«.
         The Safe Drinling Water Act (SDWA) notice, on the other hand, will be
       included in the'^ater bill and is rather a more casual news letter and
       torical record of
       and what is bein,
       that  a State  qua
                                                       a  past  water quality  problem, the problem's  significance,
                                                        done  about it.  The State implementation regulations say
                                     mat it  ocaLC  i^fying  for  primary enforcement responsibility need only
                                     require notice in the water bills or, if the bills are less frequent than quarterly,
                                                 "of direct mail. The water bill  announcements will be at
                                     icast a  wc«. -  e  rhaps months later than the occurrence of a bacteriological
                                     problem.  This no1 ice should  further include "a balanced  explanation  of the
                                     significance or  seriousness to the  public health, a fair explanation  of ^ steps
                                     taken to correct anv problem, and the results of any additional sampling." This
                                                       **. -   ...   •  .   .      	J •  TT	  T* M_.A*.* O1__1 1 Q*»  TI7T*l/-»Vl
                                     is all in keeping
                       with the  intent expressed in House Report 93-1185, which
                                        A guide to the
       i^ OIL 111. j\.w^j^/JLri£f »»it** *.i..i.w  **^fcv**-	^	—	           *.
       states: "The purpose  of this notice requirement is to educate the public as to
       the extent to which public water systems  serving them are performing . .  .,
       to develop public awareness 'of the problems facing public water systems, and
       to encourage a willingness  to support greater expenditure at all levels of gov-
       ernment to assist in solving these problems."
         It £ important therefore, that these information  notices be  factual and
       valid We,.must bi certain they are not triggered by spurious results that would
       unduly alarm the public. There is, in fact, good argument for public notice to
       be determined oaly after  the  results of a  comprehensive  investigation. But.
       with the inheren: error  we will always have  in  microbiological testing, the
       SDWA violationf and public notice  should NEVER be based on the  result
       of a single colifo rm test. If we ever start enforcing  on the basis of non-valid
       results and no actual health  hazard,  water system operators will soon  begin
       to  bias samples toward negative results rather than search for water quality
       problems.                                                        .
         I  will explain  my thoughts  and  recommendations  for correcting  these
       anomalies in a wiile.  But first I would like to relate some specific experiences
       with the coliform standard that a number of people have  helped me document
       for this symposiv m.                                                •

       EPA Regions' Experience With the Coliform Standard
                        Interstate Carrier Water Supply Certification Program was
written in  1971 by EPA's  water supply program.  It was, perhaps, the first
attempt  to  establish uniform criteria and procedures  for  interpreting the
drinking water standards (DWS) and for  classifying interstate carrier water
supplies  (ICWS). But even with the use of this uniform guide, the 20%  of
ICWS recorded ia the latest tally as "Provisionally Approved" does not reflect
the total number of bacteriological discrepancies  that were found nor the
number  that would have failed the present microbiological MCL (maximum
contaminant level). Professional judgment can never be completely eliminated
from evaluation [of the bacteriological safety  and reliability  of public water
systems.
                                     160

-------
                             COLIFORM STANDARD: REGIONS/J. HARRISON
   Region III  gave two good examples of where judgmenjt
to override a classification that the numbers alone would
                                                        was wisely .used
                                                       lave  dictated.
                                                       gical  samples per
                                                       7 analysis. In July
                                                       sample was found
                                                        sample made the
                                                        100 ml and  thus
                                                       :ollected and  they
                                                         was a sampling
                                                       to a "Provisional"
                                                         present and no

                                                        latest changes we

                                                            ny—900,000
                                                         in the following
  Chesapeake City, Virginia—50,000 population served
     This supply collects  between 70 and  100  bacteriolc
  month from its distribution system  and performs the M'
  of 1976 they collected 92 samples. That month, one bad
  with a coliform density of  520 per  100 ml.  This one
  mean density for  that month calculate out to 5.7 per
  exceed the standard.  Appropriate  check samples .were i
  came out negative. It is  a very good possibility that theire
  error. This should have resulted in an ICWS downgrade
  classification. In our judgment there was no health hazatd
  action  was taken to downgrade the system.
  The December 24 NIPDWR are written—even with the
heard about today—to take this latitude away on June 24
  Another case involves the Philadelphia Surbah Water Compa:
population served. Region III accepted the explanation givejn
letter without any downgrade  in classification.

  Mr. Walter Stanley     '       :
  Pennsylvania Dept. of Environmental Resources
  1875 New Hope Street
  Norristown, PA 19401

  Dear Mr. Stanley:
     This will explain, for the record, why we reported a mean  coliform
  density of 2.3 organisms/100 ml for the week  ending July 28, 1973.
     During  that week a number of  samples from our system  showed the
  presence of coliform organisms. However, we  began to doubt  the veracity
  of the  results  since they reflected  points widely scattered throughout our
  several divisions (300 square mile area).
     A thorough investigation made early  in August  showed that the "bad"
  samples were not  due to the source—but to one  particular sample man.
  We established this fact by having a second man follow the first  one and
  sample the same points for several days. Without exception, the first man's
  samples showed coliforms, whereas those collected by the follow-up man
  were negative.
     Even a  thorough medical examination failed to show how  the  first
  sampler's samples  became contaminated. Nevertheless, we took this man
 .,, _,rt* ' j.1. ._ —--	TI	 -_»..	=- — .^. -3 - -«.«•< r,*. 1+ lvvt -- -J-1-. rx-*«/^ii fVti 4"fn imnrt  if»  r^O^*'f O1*1J"\lrt 
-------
                      !? ;;; ?: f j if >wi i
 MICROBIOL STANDARDS EVALUATION/C.W. HMDRICKS
   	             ,   .       	  i ,  .               . ,  . ,

 determining the degree of health risk and appropriate follow-up when bac-
 teriological monitoring raises a question about the reliability of water system
 facilities or operation.
    Region  IV uses a case history concerning Florence, South  Carolina, to
 point out two key weaknesses in the microbiological standard:
      On February 24, 1972, the Region  was advised by the State  that the
    Florence System had had eleven highly contaminated samples on February
    14 and  15, such that they would  surely fail for the month. At that time,
    it also became known that the system  had failed in October 1971, based
    on one sample out of 36 unsatisfactory with a count of 47 for a mean coli-
    form density of 1.3. Subsequent sampling (not rigorous  check sampling)
    showed  no problems. After a joint USEPA-State  investigation, the system
    was classified USE PROHIBITED on February, with appropriate telegrams
    and a press release.
      This instance  illustrates two problems with the Microbiological MCL:
    (1) One single unsatisfactory sample in October caused the system to fail
    and would  have necessitated public  notice  under  the  Interim  Primary
    Drinking Water Regulations;  (2) Official action  on  the  flagrantly  bad
    quality detected in mid-February  was not warranted until  the end of the
    month, when problems had disappeared.
    It should be -noted in this instance that although the Florence system was
 classified  "Use  Prohibited",  the State gave public  assurances of no health
 danger. State agencies in the past have not rigorously determined compliance
 with bacteriological standards. Even though State standards are very similar
 or identical to the PHS standards, enforcement actions are only contemplated
. when contamanition is substantially  more serious than that  minimally neces-
 sary to fail the MCL. Many States feel the real value of the bacteriological
 test is to paint a monthly arid yearly picture of water quality which can be
 considered, along  with knowledge of the  system and  its operation, to reach
 a conclusion as to whether that system is adequate. Some States do require
 resamples whenever routine results indicate that coliform is present. But the
 most effective follow-up is that provided through direct one-on-one contact
 between the field  engineer and  the  water supply. This follow-up is by  and
 large at the discretion of the engineer. One such person in Region V who, I
 know, is doing an outstanding job gave me this description of his operation:
      It is my policy that when one sample  has a count of 10/100 ml or more,
    or  two or more samples show contamination, to call the results to the
    supply operator. At this  time we attempt to determine if anyhing has hap-
   pened which might have caused the contamination. Based  on our telephone
    conversations, we make specific recommendations  such  as: boil orders, in-
   creased  chlorine residuals, collect samples from proper taps, resampling,
    etc. Recommendations obviously vary  considerably  depending upon  infor-
   mation obtained.	
   I would  not want to see this work  interrupted  for the purpose of enforcing
 additional monitoring, reporting, and public notification where water quality
 is not impaired or a health hazard does not exist.
   One of the most striking statements I received was written by Leroy Stratton
 of the Illinois Department of Public Health. He noted that facilities which
 would be regulated  by his Department "are primarily non-community sup-
 plies with none exceeding 1,000 population." These  systems will be required
 to sample on a quarterly basis, therefore  "requiring public notice any time a
 contaminated sample is received, regardless of the level of contamination." I
 would add that two-thirds or more of the nation's community water systems
 are also less than 1,000 population,  likely to be sampling once every three

 162

-------
                              COLIFOBJM STANDARD: REGIONS/J. HARRISON
 months, and therefore require public notice every time a contaminated sample
 is received.
   Don Taylor, in EPA's Region IV office, examined  12 months  of records
 for 122 public water supplies and  reached these conclusions  based  on the
 December 24 regulations:
      "1.  The "average" public water supply will fail the monitoring require-
          ments almost 5 (4.77) months each year.
       2.  The "average" public water supply will fail the bacteriological qual-
          ity provisions approximately one month  in each 21  months  of
          record.
       3.  The "average" public water supply  would go to public  notice ap-
          proximately 5 or 6 times per year."
   One final example again from Region IV points out the fallacy of the
 monthly average limit for dictating water quality safety.
      An interstate carrier water system, Ingalls Shipbuilding-West in Pasca-
   goula,  Mississippi, showed chronic  friability to  consistently maintain bac-
   teriological quality in 1972 and 1973. Strict application of the PHS Drink-
   ing Water Standards and  the "Guide to the Interstate Water Supply Certifi-
   cation Program" resulted in the following sequence of actions:
      3/10/72—System  classified USE PROHIBITED  (with press release)
               based on failure to meet standards 11/71 and 2/72.
      5/11/72—System  reclassified  PROVISIONALLY  APPROVED  (with
               press release) based on 3 months good record.
  ,.2/15/73—-System  classified USE iPROHIBITED  (with press release)
               based on failure in  1/73.
      8/ 1/73—System  reclassified  PROVISIONALLY APPROVED  based
               on 4 months good record. (System also failed in 2/73).
      9/14/73—System  classified USE PROHIBITED  (with press release)
       - :-     based on failure in 8/73'.                      •
  """ 1/17/74—SystemT reciassified  PROVMONALLY APPROVED  based
               on 4 months good record.
      9/19/74—System  reclassified APPROVED based on 12  months good
               record.
     The  classification  actions outlined above  resulted  from five  individual
   months of failure to meet standards. In only one of these months was the
   failure  the result of a single unsatisfactory sample  (1 unsatisfactory out  of
   4). In  the other four instances, at least two samples were  unsatisfactory.
     This example points out the  problems of small  systems meeting the
   standards, even with onus of unfavorable publicity.
	? The  official record "includes' another noteworthy  month. In  August 1972,
   83  samples  were collected, one was  unsatisfactory with a  count  of 80.
   During the previous and subsequent months four samples had been col-
   lected monthly.  The unsatisfactory sample would have meant  another USE
   PROHIBITED  situation,  and evidently approximately 79  additional sam-
   ples were taken to bring  the system into compliance.
  The best way to eliminate the monthly average dilemma is to  eliminate the
 monthly average limit. A better monthly summary is simply the  number  or
 percentage of positive coliform results in «ny one month. But  an important
 distinction must be made between any monthly  summary coliform limit and
 an individual sample limit. The first pertains tothe reliability of water systems,
 and the second pertains to the safety of drinking water.
  The occurrence  or recurrence of any coliform in drinking water should be
 investigated to ascertain and correct potential health hazards. But only one
 limit  should actually define  safe drinking water.  We must remember that un-

                                                                    163

-------
                                                                                             	:wir	"";	r^i
                                   MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
I!'1 i;i!*iil >, ' I <:!l ni
                                   reliable water systems  can produce  safe  drinking water, and discrepancies
                                   respecting a low monthly summary limit may be only a potential rather than
                                   an actual or existing health hazard.
               	'•• •    '""  ';»''"	•	j   ''"Proposals
               d!	|	  , : ri,jij:  iiiiili!   i-1''(it	'"  . 'Li. 1'.. •	S	' ;.	,;•' r! -i;/vt •'•(  	 "iv:  ,""11-, ".•. * ;m-" folOi	"*! 1».: ;.pt)   li'th	" 'Wi,!' ; ::",,!n,i!i  ',:•:«;*»(
                                      As stated in the  preamble to the December 24 National  Interim Primary
                                   Drinking  Water Regulations, "further changes in the  microbiological  regula-
                                   tion may  be desirable." A number of people other than  myself have given
                                   this  considerable thought, and have had input into  the  following two pro-
                                   posals. These two alternatives I feel are logical. Either one, if chosen, is prac-
                                   ticalenough to  allow  implementation.
                                      The text of Section  141.14  and  parts (d), (e), (f), and (g) of Section
               'iiji':;    ••	^//yE;  =  'ii 14121	in the December 24 regulations should be deleted.  Part  (h) of Section
               *:'•.!'.'  ••'•<:*'-:)*!•  *':   ,": M S'l""should be	labeled (g).  The	following then inserted as replacement
               lliiiiii"       '   I'iiii'll't:,  "viis
 III    i I   P     3	'    	;  r-:	B  sit
                                   and change:
                                   	;:: Section 141.14   	.  " ' "   	
                                          The following are the maximum contaminant levels tor colirorm
                                       bacteria applicable to community water systems and non-community
                                   '••V.''w||er systems'; Compliance with maximum contaminant levels for
                                       cbliform bacteria is determined pursuant to Section 141.21.
                                     The last sentence  is consistent with the MCL. regulations for organic and
                                   inorganic chemicals. It allows compliance to  be defined by follow-up activi-
                                   ties," an.3it prevents causing violations and public notice on the basis  of a
                                   single sample result.

               ;;	;  "	;';;;;;,    '  ""    '          "       -.   PROPOSAL i       ,'      .         ,	'"|

                                       Section 141.14 (Cont'd)
                                        (A)  When  the  membrane  filter  technique  pursuant to  Section
                                             141.21 (a)  is used, the number of coliform bacteria shall be
                                             less than one  per 100 milliliters in any  sample collected and
                                             analyzed pursuant to 141.2l(b) or (c).
                                        (B)  When the fermentation tube method and either ten milliliter
                                             standard portions or 100  milliliter standard portions pursuant
                                       f     to Section  141.21 (a)  are  used,  coliform bacteria shall not be
                                             present in any portions in any sample collected and analyzed
                                             pursuant to Section 141.21 (b) or (c), ...'..'..
                                     Bacteria in drinking water represent an acute health effect and the presence
                                   of any coliform organism in finished drinking water suggests  either inade-
                                   quate treatment or  access  of undesirable materials to thejwater after treat-
                                   ment. Coliform also may be indicative  of dangerous contamination and  of
                                   Water that is not safe to drink.  The limit therefore  simply pertains  to the
                                   presence of any coliform  in drinking water". It eliminates the need  for  an
                                   arbitrary allowable number and it is easily understood.
 	      ;:."•!|	:',;•r*!i  "i  -,...;.,-Section 141.21
                                        (d)(ij  When  a sample exceeds;"a maximum contaminant level set
                                               forth in Section 141.41 (a)  or (b), one check sample shall
                                               be collected from the same sampling  point and examined.
                                               If a subsequent sample has.  already been  taken from the
                                               sampling point, it shall be  considered the check sample.
                                     If a supply monitors regularly as prescribed in (b) and  (c) of this section
                                   and there does not appear to be serious contamination, then one check sample
                                   should be enough. However, paragraph (3) will allow the  State to require

"I",  '"',',  :"     	            '164               '        "    	"	'

-------
                              COLIFORM STANDARD: REGIONS/J. HARRISON
additional check samples if a health hazard is suspected. Counting the next
subsequent sample as the check sample makes a great deal of sense but wasn't
spelled out in the December 24  regulations. Many  water systems will have
already resampled  from  the  same  tap  before  the  results of the  previous
sample are known.
     (2) When the  examination  of the check sample required in Sec-
         tion 141.21(d)(l) shows the presence  of  coliform organisms,
         the supplier of water shall:
         (I)    Report  to the State within 48 hours;
         (II)   Initiate  an investigation,  including the  collection  and
                examination of additional check  samples  from the same
                point and other sampling points in the area,  to  define
                the extent of the problem; and
         (III)   Notify the public in the area affected by the  indicated
                contamination as prescribed in  Section  141.32  or the
                State public notification requirements,  unless  the State
                determines that  no health hazard has  actually existed.
  Based on  the  results  of the check sample, the  supplier of water must
1) report to  the State, 2) initiate an investigation, and  3) notify the public.
An. important provision regards public notice requirements only  to the area
affected by the indicated contamination. For large regional water systems it
is folly to require notification to perhaps one-half  the State's population when
the problem may be limited to one section  of one town, one waiter main, one
building, or  one faucet.  Of course, the utilities  would have the option to
notify all consumers if they so choose.
     (3)  The State,  at its discretion, may require that additional check
         samples be  collected  at a  specified frequency from the same
         sampling point  and other sampling points in the area and ex-
        ., amined to  identify  and  eliminate suspected health hazards
         when a sample exceeds a maximum contaminant  level pursuant
         to Section 141.14(a) or (b), even if the check sample required
         in Section  141.21(d)(l)  does not indicate the presence  of
         coliform bacteria.
  Some sample results or bacteriological histories may raise a question about
the reliability of the system to continuously supply safe drinking water even
though no unsafe water has yet been found. In this case the State may  re-
quire additional samples to identify and eliminate  suspected health hazards.
     (4)  When the  cause of the indicated contamination has been deter-
         mined and  corrected, additional  check samples  shall be col-
 	— lected at a frequency directed by  the State.           :
  Just to confirm that the correction "has been effected,  additional samples
may be wanted for a specified  time  period.
     (5)  The location at which the check sample was taken pursuant to
         Paragraph  (d)(l) of  this Section shall not be eliminated from
         future sampling without approval of the State.
  Same requirement as in the December 24  regulation.
     (E) The  State may determine that unreliable  examination  results
          for a sample collected in a discreet monitoring period pursuant
          to  Sections 141.21(b)  or  (c) were caused by factors beyond
„_„„::__„ the control of the water" supplier.  Such factors  could be ex-
          cessive transit time between collection and examination of the
          sample, samples being  broken in transit,  or interference test
        ""results When the membrane filter technique is used. If this is
          the case,  another sample collected  immediately  upon learning

                                                                     165

-------
III"!) i1  I	! '  I'll!' 'I'liillllllllllllll'llli
                        MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                                                                                   ill
111!! I
llni '
»'ii1'!'!1:'!!!1 ' 'I

!|i:!!!:'<::! L
I'll*.' •'•!  in      tii;"
o

I
          ii	  "-nil.  Mil  :  t  •;{:''t1"'!1	;l""!
                                 w
 j"!!;.:,, ,l!"'.!l,'!.f :\if-if,	.Ji'i1.':"1'!,'"' {3
.;	::	:	I
             ::         ;:  •  -'i '.
              'Sill  ''''Hiiil  !'•. .   l;-»>! :::., 'if
              i ....... sill' ...... Illiallll   ,'    r", ' (  -i

                "         1
                      '•(• ::j .-!!,: ! t
             !'• Tail1'   i! !!!,!"'  ,  ' ''iJilim
                                      1
                                       A

                                             0.0


                                             fi S
                                             
                                             y <
                                                                    11
                                                             •B8J8
                                                             •a •* o o
                                             aE*

                                                                             33
                                                                         t-    OOVD    C
-------
                              COLIFORM STANDARD: REGIONS/J. HARRISON
           of these results may be used in determining compliance with
           sampling requirements in  Section  141.21(b)  or  (c).  How-
           ever, a single sample may not be attributed to more than one
           monitoring period.
   Overgrowths, "too long  in  transits"—over 48 hours not  30 hours, and
 broken bottles frequently account for 10% of the samples being invalid or
 the  results unreliable. These samples  should be  disregarded,  but unless the
 operator is then allowed  to collect and count-another sample, he will  be in
 violation of the monitoring requirement through no fault of his own.
     (F)  Check samples, samples with unreliable examination results,
           and special purpose  samples,, such as those taken to determine
           whether disinfection practices following water main place-
           ment, replacement,  or  repair have been sufficient,  shall not
           be used to determine compliance with Section 141.21 (b) or
           (c).
   Same as in the December 24 regulation, except samples with unreliable
 examination results are not counted  as well.
   Table 1 was compiled with data  from  10 different  Regional and  State.
 agencies. It shows  that ten percent of public water systems experience one
 to four coliform organisms per 100 milliliters at least once per year. Seldom
 have these occurrences been found to present a health risk. Another ten per-
 cent of public water  systems experience over four coliforms  per 100  milli-
 liters at least once per year and twenty percent of those occurrences  have
 been found on follow-up to be  of signiJicance.
   These data point out that a limit of one coliform may be more regulation
 and  more  expense than  is needed to define safe drinking water. Federal
 drinking water standards have  for over 50 years defined water  with  four
 coliforms per 100  milliliters as safe  to drink and there is no evidence now
 saying it is not. Some State agencies  from a practical ^standpoint may  want
 to or have to prioritize follow-up regarding bacteriological results. This lati-
 tude to four  has historically existed  in the Public Health Service standards
 and  may be needed in the future when SDWA  requirements  are applied to
hundreds of thousands of small public water systems. Proposal 2 divides the
 standard into two parts—141.14(a) and (b) addresses the safety of  drinking
 water and 141.14(c) addresses  the reliability of drinking water systems. The
 first  paragraph of  Section 141.14  and changes to Section 141.21  would re-
 main as I have discussed with Propsal  1.
                            PROPOSAL 2

     (Section 141.14 cont'd)
     (A)     When  the membrane filter technique is used, the number
             of coliform bacteria shall not exceed four per 100 milli-
             liters in  any sample  collected  and analyzed pursuant to
             Section 141.21 (b)  or (c).
     (B)(l)  When  the  fermentation tube method  and ten milliliter
             standard portions  are used, coliform bacteria  shall not
             be present in three or more portions  in any sample col-
    ~  . • .    lected and  analyzed pursuant to Section 141.21 (b)  or (c).
     (B)(2)  When  the  fermentation  tube method  and  100 milliliter
             standard portions are used, coliform bacteria shall not be
             present in  five portions in any sample  collected  and
             analyzed pursuant to Section 141.21 (b) or (c).

                                                                     167

-------
iMIfiii!!
               1P1
               II	
                         fliii!  Nli'K'i
!	s
 Ilin;,.!!!!,. 	 i 'ill
l;i i'illtt  ',! I	,.	»   .  Illli
iiK   M:n;i   '	I!!5
                                  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
     (C)      The State and public water system shall initiate definitive
        ••',-    action to identify the cause of the positive bacteriological
              sample results  and to eliminate potential health  hazards
              which migh  exist in "the  system when monitoring  pur-
              suant to Section 141.21 (b) or (c) shows  the presence of
              any coliform organisms in either  of the following:
     (C)(l)  more  than ten  percent of  the samples  when 20 or more
              samples are examined per month; or
     (C)(2)  more  than two samples when less  than 20 samples are
              examined per month.
    " ' ''"'i ,„;' in !!!!,i,	I	J	 : 'tli	'	i  * j	"	'B Ii1" iV'lii1!1  ''''liilv/ffliNH"!1!;11!!	,:: /I"." i H» 4 ii.dl	I'1*'1 "' 111, iiJI'iT •'  i,  ;:•';  '!',.' T "iiv

             .QUESTION  AND ANSWER SESSION
  - ,''. ; •'••.. ;;s •';,;."   •:• < ;;sf 4f
C.W. Hendricks, Office of Drinking Water,  U.S. Environmental Protection
Agency, Washington, D.C.
  Do you have any recommendations for a simpler standard of sample col-
lecting to eliminate the confusion caused by terminology of  "less"  or "more
than 20 samples per month"?
         II  III III!  II     ,' I'llV "iii'i'ii iiii!1!!1 i,' " i:jj; ' '? " i ill;1 »!|l ,» "'iii!!!,1!;!'1 ',!'!'lii!li?''''iii'ii	liiiii* «"''ii 1 • '!">i: ^nii' itrA Km 'i'lj!' r, N v ' , "!!ij|,iii,!"!'',    ,. M •',' 	 ," i £
J. Harrison, Water Supply  Branch, U.S. Environmental Protection Agency,
Chicago, Illinois.
  My proposal 1 and the first parts of proposal 2 simply refer  to  the result
of any sample,  and  eliminate confusion  caused by  different standards .for
those systems collecting less or more than 20 samples per  month. This  ade-
quately covers a definition of safe drinking water.  However, if addressing
the reliability of water systems, one must be concerned about repetitive sample
results. For those systems collecting 20 or more samples per month, it is best
to regulate  repetitive results  by a  percentage of  all samples collected; for
those collecting less  than 20 samples, by  a simple number.

MJ. Allen, U.S. Environmental Protection Agency, Cincinnati, Ohio.

  Are you  defining  a "spurious" water sample  as one sample which  results
in raising the monthly average greater than maximum MCL? If not, what is
your definition of a "spurious" water sample?

J. Harrison
  I used the word "spurious" in the context that Safe Drinking Water Act
notices to the public  should not be  "triggered by spurious results that would
unduly alarm the public." By  a spurious  result I mean a false or illegitimate
analytical result that does not  reflect quality of the water but  is rather caused
by an error in  analysis,  sample collection,  transit, or other anomalies not
directly related to the water supply system. Such anomalies and false results
do occur and they are the reason that nitrate  as well as all other maximum
contaminant levels, except coliform bacteria,  are never violated  on the basis
of a single  unverified sample  result. It is not proper that the December 24
Regulations  allow spurious results to raise the monthly average  greater than
the MCL, and require over  90  percent  of  all public water systems—those
required to  collect  only one sample each reporting period—to notify the
public every time a positive bacteriological sample result is obtained. My pro-
posals 1 and 2 would both define  a microbiological violation and  attendant
notification  on the basis of two—the check sample also being unsatisfactory—
analytical	results.'     '  '	

168

-------
                              IMPLEMENTATION: COMMUNITIES/RJ. BECKER


                 Implementation of the Coliform Standard
                       by a Community Water Utility

                             Robert J. Becker
                        Indianapolis Water Company
                        Indianapolis,  Indiana 46202

   As the sole representative of the public water supply industry,  I  feel it
 is appropriate to begin with a disclaimer statement. Namely, that the views
 and  opinions expressed by the  speaker do  not  necessarily represent those
 of all  community  water supplies  governed  by   the Safe  Drinking  Water
 Act.
   I must assume my presence here today reflects the desire of the Environ-
 mental Protection Agency to obtain and evaluate the bacteriological operating
 data from a representative and responsible water utility. Thanks to a number
 of illustrious predecessors, the Indianapolis Water Company  has a  continuous
 record of bacteriological data on raw, treated, and distribution system samples
 dating back over fifty  years.
   The Indianapolis  Water Company is an investor owned utility  serving ap-
 proximately 700,000 water consumers  in  Central Indiana.  Currently 95%
 of our annual supply  comes from two surface  streams  which during  low
 flow  periods  become rather scrawny.  Each  supply  is supplemented from
 upstream  storage  impoundments  during  these  low  flow  periods.  These
 streams  receive a significant proportion of municipal wastewater discharge
 as well as agricultural runoff. While they are undoubtedly not ideal supplies,
 I feel certain they are far  from  the worst.
   Our treatment facilities would  be considered fairly conventional. Coagu-
 lation is with alum, and the mixing and settling basins provide  a designed
 retention time of between  two  and onie-half to four hours. Disinfection is
 achieved  with chlorine application  prior to  mixing and  settling  at a rate
 sufficient to maintain a free chlorine  residual through the filters.  All  filters
 are of the rapid sand type with a six inch anthracite coal cap.
   As is  the common practice of almost all water utilities in this  country
 we rely  almost totally  on chlorine to  provide a bacteriologically  safe  water
 to the consumer's  tap.  During the winter months chlorine residuals leav-
 ing our  treatment facilities are maintained slightly in excess of  1.00  ppm.
 This residual is increased to about 1.50 ppm during the hot summer months.
 Normally,  we apply anhydrous ammonia to the filter effluent to  convert to
 a  chloramine residual  prior to  delivery to  the distribution system. We
 realize that a chloramine residual is" a less potent bacteriocide than a free
 chlorine  residual. Therefore, for about a four week period each spring and
 fall we  discontinue  ammonia application, and  treat the entire system with
 a free chlorine residual. It is important that this occur while water tempera-
 tures  are not  much  greater than  50 °F.
  As a system supplying  a population  of approximately 700,000,  we are
 required  to collect  250  samples monthly. Actually, we  will  run  about 375
 samples per month. Our distribution system sampling points  are selected for
 their  strategic location,  general sanitary  conditions, and availability  for our
 regular  entrance. At .the present time,  we have 53 sample sites  located
 throughout "our service areaT Due to distances involved to  the extremities of
 the system, we have four samples trips. Each trip involves  15 to 20 samples.
 Obviously, some  locations are sampled, more frequently than others.  It  Is
 our intention to have  the  preparation  of all samples  completed within
four hours of collection. All of our outlying  booster stations and elevated

                                                                     169

-------
!r Tf '""S*!!!!*
 Li "If!11,
                                                                             I1- ran1
                                                                                     •	If I!"* "I i!	in
.• • '. ••: ;:,;i .I . ''">;,: ' .,.• iiiiiF1 	 m 	 "!;. •
!' ''I.1' ' ' ' ; i ""If U,1' 111! 1 •!'"' '|l"
i " " ;! i, ;; . "'.ii,,!' : ' ijliii'1!,,,, I'ly,:1 "''
; 	 ' III!1)!:,] !« ' .Uj'i, 	
'. ' I „ ' • " , I si: I" kft'ni
: , •; , 'i 	 i i • P; 	 i1 ,„ 1 i , iV" " i ji!'S;," If:,"!, :, ' '
, 'I' 111''! II'1 , . „ : i, ^l < . . ', llllll'llilj III!" 	 '•':
j iiil
'|L VI
.'•••! 	 1 "
„: i ,:;)!]
1 ' •:•
i1' i1
i) •: i . ; : '-jiii iiii:':
'.jii i,i' i n 	 iiiii,
" :•'! ' I ' !' ' '.ill I!!1
•ifi1 ! '„(;,; I,,;! 	 "Eir|ii
i, i „ ; '"''Si1 I-111
i ii : ii i , iiii"1
n^llt i , '. 	 iiji'liiii'
               -AVD/NoixYmvAa SCTHVONVXS loiacraoiw
                                                                               	si!
                                                                                      ;"i'i '. il  i  	iPil'.f
                                                                                              li <
                                                               IIIII  III I 111   i!ii!l!,;Liill'i!'!ii:lllllll»ailr ,

-------
                               IMPLEMENTATION: COMMUNITIES/!?..!. BECKER
 tanks serve as sampling sites. Other sites include fire stations, filling stations,
 and  the  like. We  purposely  have a few  sajmpling sites on  or near  dead
 ends.  We are trying to 'maintain a knowledge of water quality throughout
 the system, and not just obtain a good recor'd to impress the public or the
 public health  authorities. We  do not now  have nor  ever  had  during our
 period of record any open  or  uncovered finished water reservoirs.
   There  are  many factors  which influence t|he  water quality  which is de-
 livered to the  consumer's tap. One of the grejat limitations in achieving uni-
 form quality throughout  an entire distribution system lies  in our inability to
 maintain  adequate levels of chlorine rissiduals indefinitely.  We have docu-
 mented this occurrence in  a series of laboratory  experiments. Sterile gallon
 containers of the White River  treatment plant effluent were stored at 68°F
 and  80°F. Identical  containers  were used jfor  both waters  with  a  free
 chlorine  residual and a  chloramine  or cornbined  residual.  Figure  1  il-
 lustrates  the persistence of  chlorine residuals in  relation  to both time and
 form. Chlorine residuals were determined by starch-iodide titration as speci-
 fied by Standard Methods.  As you can readily  see, the water  temperature
 and the form of residual make a substantial idifference in the persistence of
 the residual. This work was done  under fairly ideal laboratory conditions.
 I  doubt that  chlorine  residuals in  our distribution  piping would persist  as
 well.                                        |
   For comparative  purposes, we ran  a similar test on our  best well water
 supply. Figure 2 illustrates the  difference in the persistence of free chlorine
 residuals  between our  surface water treatment plant effluent and  this particu-
 lar well water supply. These findings  correlate perfectly with the work done
 in  Detroit by  Mike Taras on  the  chlorihation of nitrogenous  compounds.
   Another major obstacle for the maintenance] of high water quality through-
 out the distribution system is  the. design criteria.  Pipe sizing  is  designed
 usually to meet fire flows ~6n fop of  normal (system  demands. This  unques-
 tionably is  a  deterrent to desirable displacement, particularly  in a remote
 residential or  low water  use  area. The location of  sizeable  elevated or
 ground storage facilities on this  piping network  adds  still another obstacle
 as  positive circulation is  not normally achievjed.  In a  system such as ours,
 which includes 15 different pressure  districts, the  division  valving creates
 additional dead  ends  to  those normally  encountered  in  a  system that  is
 subject to substantial growth  in all  directions.  We have a total of  over
 1750  dead end  situations  in  our  system.  I mention  these  factors which
 adversely affect circulation within the system  only  to emphasize the com-
 plexity and magnitude of the problem we face.
   Despite our. utilization of a chloramine residual during the summer months,
 we still lose our residuals  in distribution system storage  reservoirs and in
remote portions of the system.  In recognition! of this problem,  we have in-
stalled supplemental chlorination  equipment in all booster pumping  stations
 and elevated storage tanks that have  been constructed  on the  system since.
 1956. These installations have become more sophisticated  as our experience
and confidence have  grown with  unmanneli automatic chlorine  feeding
systems. They are normally in operation from about May through November.
  We  do perform all  of  our own bacteriological  analysis generally  as  pre-
scribed in Standard Methods. The laboratory personnel,  equipment, and pro-
 cedures are all inspected and. approved annuaHy by the Indiana; State Board
of Health. Following an .extensive study conducted in bur laboratory in the
mid 1940's  we selected Lauryl  Ttryptose Broth for all  sample tubes on the
basis  that it resulted in less false presumptives than Lactose Broth  on  our
particular waters. We do make practice of carrying to completion the testing
of  gas formers from finished water samples.1 For over ten  years we have

                                                                      171

-------
sili  I!,!!!
ill!  ilsM
!
                                                                 ;j II  I ;|! i!i |! ii
                                                                                                                   15
                                                                           DAYS
                                            Figure i PERSISTENCE OF FREE CHLORINE RESIDUALS AFTER 5 HOUR CONTACT TIME.
                                                                                                                                                                     §

                                                                                                                                                                     I
                                                                                                                                                                      WJ

                                                                                                                                                                      I
                                                                                                                                                                      I
                                                                                                                                                                      kH
                                                                                                                                                                      i
                                                                                                                                                                      "n
                                                                                                                                                                      ^
                                                                                                                                                                      a
                                                                                                                                                                      o
                                                                                                                                                                      1
                                                                                                                                                                      GO-

-------
                              IMPLEMENTATION: COMMUNITIES/RJ. BECKER


 run  parallel  comparisons  between  the  membrane  filter  technique  and
 multiple tube for various selective locations. As a result of this study we
 have elected to stay with the .multiple tube procedure  for  all of our  bac-
 teriological analyses. We believe it is  the  more  sensitive for our particular
 situation.  This  should become  apparent  as  we examine  the  data  which
 will follow.
   Certainly one consideration in the bacterial quality of the finished water
 is .the bacterial quality of  the  sources. Since  the White River provides ap-
 proximately  two-thirds of  our  total  production, let's  examine  this supply
 from a bacteriological standpoint. Table 1 enumerates the monthly average
 MPN for  the  10-year period 1967-76. From the  table it is fairly obvious
 that the quality is subject to considerable variation. Generally  the months
 with  the greater  numbers  reflect a period  of  higher  surface runoff.  The
 last two months of 1976 correspond to an extremely dry period with neglible
 surface runoff. By some  standards this would not be considered a particu-
 larly desirable  source. However, it  happens to  be  the major waterway
 through the central portion of Indiana,
   In recent years we have  also been  testing  our raw waters for the more
 specific type fecal coliform  organisms. Table 2 illustrates the results of this
 particular series based on twice  a week sampling for the past year. Monthly
 average  numerical comparisons  are shown for the MPN as determined by
 the multiple tube  procedure, the MPN  as  determined by  the  membrane
 filter technique utilizing M-Endo  Broth,  the  fecal coliform count  utilizing
 M-FC Broth  and  incubation at 44.5-° C,  and fecal strep, counts  utilizing
 both  M-Enterococcus and  KF  media. As you  can readily see the fecal
 coliform numbers are significantly below those  obtained for the total coli-
 form.  For the membrane  filter,  fecal coliform,  and fecal  strep,  determina-
 tions between  .5  and 5  ml of raw  water are  used  depending  upon  raw
 water  turbidity and bacteriological quality.        --.-—?•.-

   Table 3  illustrates  the bacteriological results of the treatment  plant efflu-
 ent and  distribution system  samples over  a 50 year period at  5 year inter-
 vals.  WR  indicates  the  effluent samples  from the White River treatment
 plant,  and  FC  indicates the  effluent samples from the  Fall Creek treatment
 plant.  The Fall Creek rapid  sand  treatment plant was placed  in service as
 a  secondary supply  in 1941. At the: White  River plant the initial  rapid
 sand filters were placed in service in 1926.  Prior  to that all filtration  was
 through slow sand filters  since about  1905. Additional rapid'sand filter ex-
pansion at this plant occurred in 1953, 1959, and  1972. Back in 1925  and
 1930 the  chlorine residuals  leaving the  plant  were in the range of  .10
ppm. This would  be considered an extremely low range- according to  our
 current standards. I  would  sincerely  doubt that it was a breakpointed or
free chlorine residual. A good understanding of this chlorination phenomena
 did not occur until the mid to late 1930's. In reviewing this data I feel it
is  apparent the conventional treatment process  with responsible operators
can provide an exceptionally safe bacteriological product  regardless of varia-
 tions  in the quality  of  the  source. Likewise  the data indicates that  this
 quality does tend to deteriorate slightly in the distribution system. While we
 are able to show a significant and consistent improvement in the coliform
recovery from  the system  samples, we.have not  shown the same degree of
 success with regard to the total (standard) plate count. Undoubtedly the  im-
provement in coliform recovery is largely the result of leaving the treatment
plants with higher levels of chlorine residuals over the years.
                                                                     173

-------
                     inn 'iii	I in	mi	iiiiiiiiiiiiii'iiii iiiiiiiiii
                                                                         	I  '(I I'1!!!	 I  	I   I	i I1"!'!!  I	I  i Ml	l|i|i||l	I	id
                                             MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                                                                                                                                      1 lillil|li  inn
          O
          00

          §
                                              vo



                                              vo   vo
                                              ON   r-
     W
                                                    •*
                                                    r~
                                              J>   r-
                                              &
                    ,  ..... .1 1  •:,*  , .. ..... .;»', .  , af  • aiiis • •.
                     [[[
                             •.•f»:v
 iii  '  ;>
                                                 "'::
,
sill  "E   t-i
     "    ~
                                          "1   O
                                             i ,,«   O

                                             ' :'
IV • RliJi  ,  ! , !' n,i M""
.i!'!1!!!:1 i  i-i   {;.:,
                                    i!1 ..... "H  i. "luiiifl '  .
                     f ..... ;  -it • ;,>y:
                                               .._.
                                   	&
                                            ,   g
                                               i
                                               13
                                                    l
                                                        VO VO
                                                              '*"
                                                              CS
                                               JO vi  oo
                                               ro O\  CN
                                                        10
                                                           Cl
                                                                                               oo


                                                                         r,«oCvo-o
                                                        10 O 00 t-l  «0 CO O  t- O O, 00^ r-^VO^

                                                        «o" ON ON r-T T-? CO" rf vT rt" •* C<1 O JO
                                                        _J vo              ra     1-1        r~- CN
               X^ ^H" ^ 1/-T CO" U-T VO VO . V

               1-1 10    i-l              ^H
               ?2 18

               cQvJvOVOOi-iCNVDVO OO,1'"!

               eo" r-" i-n" vT oo" c-f i-T       t-H T-<
               C^i "^1" CO C^ ^^
                                                    s;

                                                                                    fe-    fe
                                                                                   ^?  ^ >5

 o
'cf
 CO



                                                                                                                             B
                                                                                                                             p
                                                                                                                                    O i-l  CO ^ JO
                                                                                                                                 oo ON ON  cN O "^r
                                                                                                                                 » oo vo  •* V)

                                                                                                                                   "i co r--  o
                                                                                                                                 VO CO  V> CO VO  CO
                                                                                                                                    oo  fN vo co
                                                                                                                                     CM c^l ON ^5  JO
                                                                                                                                   .  ON ON r>J O  '*
                                                                                                                                 co  o\ 1-1 '"fr f^
                                                                                           O t-- -t
                                                                                           O •—I O\
                                                                                           •^- OJ •"*

&

<
                                                                               S

                                                                                                                              fe
                                                                                                                              fe
                                                                                        t-. r--  O oo
                                                                                           o  c»i r~
                                                                                           Tj-  •* CO
                                                                                                                                     CO ^ 0

                                                                                                                                  00 CO 04 O
                                                                                                                                  co co
                                                                                                                                                  vi
                                                                                                         0
                                                                                                         V)
                                   co oo O »n  ON vo
                                   oo tN ON ON  1-3 '^r
                                   vo co fN TH  t^
                                                                                                                                     t«. in vi o 10
                                                                                                                                     r—I C-l t~ O 'T
                                                                                                                                     VO fN   . C*
                                                                                         5 oo m vo  ON oo
                                                                                         t-.. t-» ^j- co  vo

                                                                                         cTof I-H"
                                                                                         co^

                                                                                         vo
                                                                                         ON_ ON vj_ oo (N
                                                                                                                                      *j
                                                                                                                                      a, g
                                                                                                                                   o  o
                                                                                                                                   U  (U
                                                                                                                                  OH PH
       iiiiiuiii ( ..... iii iiiiiii i  ii in i iiiiiiiiiiL i, iii I iiiiiii .......... 'in 1 1 iiiii
                                               174
                                             i ..... i ....... mi , i ii iiiiti .......... iiii
                                                                         am im, ..... \IU:M liii ; s ...... IIIIMII v^  ..... [in.;. :iii:ii:ii£! ...... '

-------
                             IMPLEMENTATION: COMMUNITIES/R J. BECKER
                 TABLE 3.  ANNUAL AVERAGES.
35° Plate Count
Plant Effluent
WR FC
1975
1970
1965
1960
1955
1950
1945
1940
1935
1930
1925
AVG.
1
2
2
2
1
2
2
3
5
4
. 10
3
3
1
2
2
1
1
2




2 .
% Coliiform
Plant Effluent
WR FC
.06
.01
.02
.01
.01
.05
.18
.36
.00
.00
.00
.07
.28
.01
.17
.00
.01
.03
.03




.08
35° Plate Count % Coliform
Distribution System
System Distribution
75
300
18
100
300
110
60
85
11
210
325
140
.14
.29
.29
.32
.44
.67
.80
: 1.48
1.40
1.50
2.32
.88
  Table 4 examines the average monthly distribution of system bacteriologi-
cal  samples  for the 25-year period (1951-1976). We examine approximately
375 samples or 1875  tubes per month from the distribution system. This is
125 samples over that required  by the Drinking Water Standards on the
basis  of population  served.  The numbers  under  the  coliform heading
represent the  average number  of tubes  that completed  as coliforms for
that particular month. Naturally you recall  that  a total plate count  limi-
tation was proposed in the  Interim Primary  Drinking Water Standard.. We
objected rather strenuously on the  basts that our_historical  record^, indicated
we  regularly recovered" counts in excess of this number, and we were un-
aware of any  health related significance. The second row  of numbers  indi-
cates the monthly average total plate  counts.  With a few individual samples
reaching  10,000  colonies  it is not difficult to reach a high monthly aver-
age. The lower row of numbers  indicates the number of plates or samples
which would have exceeded for  the  month the proposed  500 plate  count.
Quite obviously  the  deterioration  in  bacteriological quality is a seasonal
one for us peaking during the hot summer months.  Water temperature plays
a part in a  number of influential parameters.
  Table 5 provides an analysis of all distribution system samples on a monthly
basis for the past ten years (1967-76). The  seasonal trend which^appeared.
so traditionally in the previous slide is not so  apparent in this one.  However,
we  can determine that  during this particular  period that gas formers  were
recovered in .29% of the  tubes examined. Likewise that the number of  coli-
forms which completed averaged .19% of the tubes examined or  almost
exactly two-thirds of the gas formers recovered. The irregular pattern  shown
in this slide perhaps introduces  still  another variable.  Certainly the  mis-
handling  of the sample in either collection or preparation  can  lead  to en-
tirely erroneous results which nevertheless  influence  the quantitative measure
of water quality or safety. While this does  have  a slight  impact on  our
monthly averages, it might be critical  for a small utility with a limited num-
ber of samples.
  Table 6 analyzes each sample  which resulted in gas formation in three or
more  10  ml portions  for the years 1970-76.  For each sample is shown the
chlorine residual, total place count, number of tubes* showing gas "formation,
and the number  of tubes  completing  as coliforms.  Almost  identical data is

                                                                     175

-------
TABLE 4  25-YEAR MONTHLY AVERAGE (1951-76) DISTRIBUTION SYSTEM SAMPLES.
Jan. Feb.
COLIFORMS

TOTAL PLATE
NO. PLATES
— EXCESS 500






Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Total
Avg.



COUNT




TABLE

No.
SAMPLES
3,782
3,596
3,841
3,690
3,779
3,769
3,765
3,882
3,623
3,961
3,440
3,698
44,826
3,736

3 4

8 8
1 1



Mar Apr.
.3 4

6 6
0 1



May June
6 9

41 78
5 10



5. DISTRIBUTION SYSTEM SAMPLE

No.
TUBES
18,910
17,980
19,205
18,450
; 18,895
18,845
18,825
19,410
18,115
19,805
17,200
i 18,490
224,130
18,680


GAS
FORM
26
44
22
53
64
27
98
77
110
53
40
46
660
55
.29%


0/5
3,764
3,575
3,831
3,667
3,756
3,760
3,733
3,853
3,596
3,950
3,423
3,667
44,575
3,715

July
10

318
25











ANALYSIS


1/5
13
11
7
13
12
7
21
17
17
8
13
28
167
14



Aug. Sept.
12 9

439 262
38 31



(1967-1976).

No. of Samples
Oct.
7

133
19






2/5 3/5 4/5
5
4
1
3
3
0
5
5
3
2
2
2
35
3

0
3
0
3
2
0
3
1
3
0
0
0
15
1.3

0
1
0
1
2
0
2
4
4
0
2
0
16
1.3

Nov.
3

52
7






5/5
0
2
2
3
4
2
1
1
0
1
0
1
17
1.4

Dec.
3

18
6





TOTAL
COLI
23
• 42
19
47
52
17
53
51
48
17
25
37
431
36
.19%
o
w
s
f
§
§
o
G ; :•
H
o • i "
55 ' "_'
p ; . ^:
s • • "
55 ' ";"
l? V
o . i1.
on ; 't\'
s


j ,, -„-







-------
     TABLE 6.  DISTRIBUTION SYSTEM COHFORM RECOVERY (3  OR MORE—10 ML)  ANALYSIS.

Date
5-20-76
8-18-76
9- 9-76
4-25-75
5- 1-75
9- 2-75
5- 9-74
5-31-75
9-23-74
11-18-74
12-11-74
5- 9-73 '
2-22-72
I J.-7A-T).
: 4-17-72
!j 7-17-72 , •
j 4-29-71
4-30-71:
6-15-71
'• 7-12-71 '
8-31-71 |
2- 9-70 •
— 	 : 	 5-14-79- 	
I 5-15-70
6-19-70 '.
7- 8-70
8-21-70
8-25-70
11-30-70
Sample '.
No. |
52 , ! '
140 J •
146 !
151
139
302
139 i :
140
302
302
' 42 : i i.
33 i '< ;
52 ! • !'
139 ; ;
139 P. ;
148 ;
145
43 .
302
58
;56 '.'
148 !
	 302 	
'61 i-. f
150 i ' !
150 i ; !
146 •' ;
62
146
C12
Res.
.80
.36
.12
1.09
.92
.76
.88
.40
..76
.72
1.00
1.04
1.04
=80
.64
1.08
.28
.40
.64
.08
.04
.76
	 n j

.20
.72
.84
0.00
.12
.40
SAMPLE
TPC1 GF2
35 5
11 3
5,400 ••'•' 4
1 : 3
110 5
1 i 3
240 ' 1 5
70 ' 4
1 i 5
1 4
115 j 5
2 ,
6
19
650 :
1
1 ;
10 ';
1 •
5,400
55 ':
i .';
	 1 ,.
55
55 C
4
•3
5
5
3
5
: 5
5
4
3
5
1 3
5(
... . '5
4 :i j 3
14,000 ' 4
6 ! 4
1 4
COLI
5
3
4
3
5
3
5
3
4
4
5
4
3
5
5
3
5
5
5
4
3
5
3
3
5
3
4
4
4
CHECK SAMPLE
TPC1
0
1,200
85
2
1
0
1
1
1
1
1
1
5
5
1
1
1
1
1
11,000
600
1
	

1
18
N.A.
•N.A.
45
1
GF2
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0~,"
1
0
2
b
0
0
0
COLI
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
	 ••—'"•" /\ 	 " 	 • 	
u
0
1
0
0
0
0











§
a
g
. &
§
H
a
o
5?
o
o
1
h3
M
tn
t/1
p
J-1
s
1. Total Plate Count
2. Gas Formation
                                                                                                           O

-------
   >   -   ^, :.     t   -  ,  s  -   i  ^*-*     •    ----.
TABLE 7. DEAD END SAMPLE COMPARISON WITH TIME.
         IWC LAB (2 HR.)
EPA LABORATORY
Date
6-24-75

7-16-75




No.
1
2
3
4
5
6
7
8
9
Gl, Res.
0.00
.54
.20
.12
.12
.12
0.00
0.00
.12






TPC*
4,900
1
0
2
0
1
44
300
300
GF
0
0
0
0
0
0
0
1
0
2 COLI







0
0
0
0
0
0
0
1 '
0
TPC
110,000
540
730
650
240
480
21,000
760
30,000
GF
5
0
0
3
0
0
2
0
0
COLI







3
0
0
2
0
0
2
0
0
1. Total Plate -Count
2. Gas Formation
TABLE 8.

- Temp-F
_Turb.-Raw
MPN-Raw
Algae-Raw
Alum-$/MG
CMor-#/MG
Turb.-Sett
Algae-Sett
Turb.-PE
Algae-PE
t Coli-D.S.
:TPC-D.S.
:TPC>500
Jan.
38
46
49,231
1,252
224
50
1.5
212
.36
25
4
12
1
Feb.
38
90
38,383
934
260
43
1,6
139
.35
14
1
9
2
Mar.
43
55
17,504
1,485
133
44
1.9
62
.23
9
0
3
0
Apr.
51
85
23,488
5,779
193
46
1.5
450
.26
25
7
11
3
WHITE
May
66
90
6,415
13,167
215
60
1.4
1,232
.34
94
9
28
4
RIVER SUPPLY— 1975.
June
73
105
3,151
7,632
182
64
1.7
551
.25
54
0
3
0
July
77
90
5,709
10,953
249
81
1.9
1,497
.57
165
0
180
13-
Aug.
79
70
7,010
12,431
403
94
2.1
1,744
.85
158
1
450
39
Sept.
68
45
2,451
15,503
185
72
1.9
1,463
.37
99
4
90
16
Oct.
58
35
2,579
7,358
116
60
1.5
1,274
.30
90
0
31
7
Nov.
52
25
5,018
3,982
114
46
1.6
272
.26
23
o
65
4
MF §
	 03
0 §
0 CO
•H
0 &
u o
0 >
0 o
0 n
0 >

|
|
*; ,33-
Dec. §
40 rl
30-— &
57,085
3,263
172
59
2.2
328
.35
26,

4' :
0

-------
                             IMPLEMENTATION: COMMUNITIES/R.J. BECKER


also shown on each check sample. As you examine  this information, it be-
comes apparent that regardless  of your theory or beliefs there exists some
exceptions. My own interpretation is that  a majority of these unsatisfactory
results are  due to mishandling  either  during sample  collection or sample
preparation.  This impression primarily  centers  around the  relatively  high.
chlorine residuals occurring in some of the samples at the  time of collection,
and the conviction that this range of residual is not  conducive to significant
bacterial growths. Likewise the generally satisfactory  results recovered in the
check samples.
  During the summer of 1975 an EPA, Cincinnati Laboratory team was at
our utility making some virus studies. The bacteriological laboratory back in
Cincinnati requested that we collect several dead end water samples, and re-
turn them to Cincinnati with the virus team. Two sets of samples were col-
lected from  nine separate locations. Each  sample was collected from the
nozzle flow from a dry barrel fire hydrant as soon as it was anticipated that
the branch line had been displaced. This is obviously an undesirable sample
point from a sanitary standpoint. Sample  preparation in our lab was within
two hours. I have no definite knowledge of the time interval prior to prepa-'
ration in  the EPA lab, but it could have  been close to 24 hours after col-
lection. Table 7 illustrates the results from both labs with the primary variable
being time. This is still another obstacle which a small utility faces in getting
a water sample to  a central  laboratory for analysis,  and according to some
authorities presents a  major handicap  in obtaining representative quality
results.
                                     '
  Table 8 is a brief monthly summary  looking af a number of raw water
parameters which may have an effect on finished water and distribution sys-
tem water quality. The year of 1975 was not a particularly bad year relative
to water quality, :butit does rillustr ate one previously  unmentioned parameter
which reportedly has a significant impact  upon bacteriological quality.  The
parameter that I would like to focus on ,is  that of turbidity. Turbidity is one
of  the most nebulous for quantitative  comparison.  Turbidity comes in all
sizes, shapes,  and  charges.  Although cold water  temperatures do  interfere
with coagulation, silt particles  accompanying soil runoff, are normally relaT
lively easy to remove. However the smaller the particle size, the more  diffi-
cult it is  to remove. This is demonstrated  by comparing the algae  counts
for raw, settled, and filter effluents. For our supply  the most normally  pre-
dominant  organisms are diatoms such as Cyclotella,  Synedra,  and Navicula.
As  you can see from the data presented we fairly consistently remove 88%
of the algae by coagulation and  settling. The datar also indicates that 92% of
the remainder are removed:by filtration. Raw water  algae counts as high as
50,000 to 100,000 organisms per ml have been  recorded in  the past.  Our
record over several years on raw, settled,  and , filter effluent algae  samples,
indicates  the percent removal statistics  hold fairly uniform. Naturally,  the
greater the raw water concentration the higher the concentration will be in
the finished water resulting in higher turbidity readings. If you will look closely
at  the month of August, you will. see  it  w.as  the  month with the highest
effluent turbidity even with a twice average alum dosage for about an average
raw water turbidity. In the statement of  Basis and Purpose for the Proposed
National  Interim Primary Drinking Water Standards it states "A  properly
operated water treatment plant employing  coagulants and  granular  filtration
should have no difficulty in consistently producing a finished water conforming
to this limit." We are one utility  that can meet the  limitation of one (1) Tur-
bidity Unit on an annual basis, but from  our historical data we know we will

                                                                      179

-------
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
 exceed that limitation on  a  monthly basis under certain  raw water  quality
 conditions.
                                :o confuse the issue by the presentation of
                                  to  the  subject  of bacteriological quality.
                                2X one requiring a knowledgeable interpreta-
  It has not been my intention
limited operating data pertinent
Rather that the subject is -a compl
 tion of perhaps misleading results. The Indianapolic Water Company histori-
 cally has met and expects to contmue to meet the proposed coliform standard
 in the Primary Drinking Water Regulations. Please remember that PL 93-523
 is known  as the "Safe Drinking Water Act"  and not the "Pure Drinking
 Water Act."
   The most  distasteful portion of  the Act to me and  many other water
 utility operators  is that of public notification. No responsible operator would
 hesitate to forewarn their consumers when a known hazard exists, but to do
 so based upon an erroneous test does nothing but discredit the utility or the
 industry. It is my personal observ ition that public notification should be made
 only after  a careful review of dai:a indicating an MCL has been exceeded by
 a knowledgeable public health  authority, and that a hazard does or did exist.
  Another area of the Act that [ question is the responsibility of the utility
 to maintain water quality to the tap. This really is beyond the limits of any
 control by the utility, and actually falls on private property. The water utility
 does not establish plumbing codei,  approve building plans, or perform build-
 ing inspections during  or following  construction. Likewise it  has absolutely
 no  control over the  remodeling or  repair of  existing plumbing  within the
 consumers property. We know from first hand experience that these actions
 can result  in unsatisfactory and anrepresentative bacteriological quality.
  From the standpoint of elapse :d time and our operating  data the second
 daily check sample would appear superfluous. In the event  of plumbing repair
 contamination introduced from within the sample  structure,  samples  col-
 lected from adjacent sites may prove to be much more representative of actual
 water quality in the system. If a bad sample is recovered in a small system
 late in the  month, there is no reasonable way an operator  can collect enough
 samples  to improve his monthly average! If the bad sample occurs early in
 the month, he may have a change.  If there is  a  suspected cause  of sample
 contamination in the original sample, it makes more sense  to me to disregard
 it in the record and retain the  results of a satisfactory check sample. Rather
 than retain the results of an unrspresentative sample, and not even consider
 the results  of the check sample. I realize this all boils down to the numbers
 game in trying to meet  a hard and fast regulation. I think  it would be well to
 point out that most responsible w ater utilities operate according to  well docu-
 mented and proven practices ani principles evaluated to  fit their own par-
 ticular situation.  As long as all these systems perform properly, satisfactory
 bacteriological results will generslly follow.  However, if a  fault occurs in the
 system and unsatisfactory bacteriological quality results, no number of addi-
 tional samples will correct the fault.
  I sincerely appreciate the  opportunity to express my obviously  prejudiced
 views on this subject to this group today. I  am  equally grateful that it is not
 my responsibility to establish bacteriological quality standards to protect the
public's health under  all the known variable conditions that exist throughout
the nation.
180

-------
                                   IMPLEMENTATION: STATE/HJ. ONGERTH
             Compliance with PL 93-523 Coliform Standards

                            Henry J. Ongeith
                     California Department of Health
                        Berkeley, California 94704           :

   Many people have taken the coliform standards for granted. The advent
of PL 93-523 and its mandatory  compliance! requirements have now  begun
to raise questions in the minds of some about problems of attainability and
compliance  with these standards. This  presentation will briefly  review the
development of certain  aspects of  the  coliform  standards  and discuss the
program of compliance in California.        j
  In  1925 Treasury Department standards for drinking water were revised
and the coliform standards as we know them] were first presented. The  mem-
bership of  the  advisory  committee reads  like a Who's Who of Sanitary
Engineering  including: George C.  Whipple  C.   A. Emerson, George W.
Fuller, Charles  G. Hyde, Edwin O.  Jordan, Harry E.  Jordan,  Milton J.
Rosenau, Robert S. Weston, C.-E. A.  Winslqw, and Abel Wolman—a most
impressive group.
  The report of this committee  is  brief and  exceedinly interesting, and is
presented in full as follows:
       "The task referred to this committee by the Surgeon General of
    the  Public Health  Service is to  forrmlate  definite specifications
    which may be used by  the Public Health Service in the administra-
    tive action which it is required  to take upon  the supplies of drink-
     ing  water offered by common carriers
for the use of passengers
     carried in interstate traffic. The recommendations submitted apply,
     therefore, only to this special case,  and
     general application.              :
       "Since  the purpose of the supervisior
ire not"proposed for more

.which the Public Health
     Service exercises over these water supplies is to safeguard the health
     of  the  public, the examinations  and specific requirements  herein
     proposed have reference chiefly to forming a judgment of  safety,
     and are designed especially to afford  protection against  the most
     serious danger which is associated with water supplies, namely, that
     of infection with typhoid fever and other diseases of similar origin
     and transmission. Less emphasis has been placed upon physical arid
     chemical characteristics affecting  the  acceptability  of  water with
     respect of appearance, taste, and  odor,  because  these  are matters
     of less  fundamental importance and becpuse, in actual  experience,
     the water supplies which come under consideration, if  satisfactory
     from the standpoint of safety, will usually be  found  satisfactory
     with respect to physical and chemical characteristics.
      "The first step toward the establishment of standards which will
     insure the safety of water supplies conforming to themris--to agree
     upon some  criterion of safety. This is necessary because  "safety"
     •_	j.	—....——I!—...  ~* „  *!• AW f, Mn f\ A't-Tinl f i*  f*t*r**3ii /•»£»/-!  tr« t*cfc1»s-friTrja  
-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
     safety of a water supply, it is well established that the water supplies
     of many of our large cities are safe in the sense stated above, since
     the large populations using them continuously have,  in recent years,
     suffered only a minimal incidence of typhoid fever and other poten-
     tially water-borne infections. Whether or not these water  supplies
     have had any part whatsoever in the conveyance of such infections
     during the period referred to is a question that can not be answered
     will full certainty; but the total incidence of the diseases has  been
     so low that even though the water supplies be charged with responsi-
     bility for the  maximum share which may reasonably be suggested,
     the risk of infection through them is  still very small compared to
     the ordinary hazards  of  everyday life.1
        "The committee has, therefore, taken this better class of municipal
     water supplies as its standard of comparison with respect to safety
     and proposes,  as  a fair objective, that the water supplies furnished
     by common carriers to passengers in interstate  traffic be  of  com-
     parable safety. As regards protection of the traveling public, such a
     standard is fair, since it  implies that the  use of  the water  supplied
     to them in travel shall not add to the almost negligible risk which
     is ordinarily incurred at home by  those  who habitually use water
     supplies of somewhat better than average quality. From the stand-
     point of the carriers also, this standard is believed to be  fair and
     reasonable,' since it refers  to water  supplies which .are  actually
     obtainable in all sections of the country  and from a great variety
     of sources,
        'The  next  and principal task of the committee  has been to set
     up  objective   requirements  which will  conform  to this  general
     standard of safety; that  is,  requirements which will ordinarily be
     fulfilled by the municipal  supplies  of  epidemiologically demon-
     strated safety which  constitute the  standard of comparison, but
     will exclude supplies of less assured safety.  Since there is no single
     and measurable characteristic  of  water supplies which  bears any
     known and constant relation to actual safety, the standard recom-
     mended is composite, including certain requirements relative to the
     source and protection of  the water supplies in question as indicated
     by a careful sanitary survey, and certain other requirements relative
     to bacterial content as shown by standard tests.
        "It is anticipated that little objection  will be raised  to the re-
     quirements laid down as to source and protection,  at least to then-
     general  intent, because they are based upon well recognized  prin-
     ciples of  sanitary  engineering, and  because they  are  necessarily
     stated in general terms  which imply  a rather broad consideration
     of each supply from all angles and  the exercise of discretion  in
     forming  an ultimate  judgment of  its fitness. The bacteriological
     standard, on the other hand, is stated  in. definite quantitative terms.
     This is unavoidable if such a  standard be included  at all, since the
     methods of bacteriological examinations  are quantitative and yield

   1This evidence actually proves only that the  water supplies in  question have
 been generally "safe" in the past during the period of low prevalence of infection.
 The likelihbcxi that they will continue to be equally or more safe .in the future must,
 of course, be  reckoned from "other considerations,  such  as the probability  of
 future  change in the pollution of their watershed, the character and  consistency of
 their protection, etc.  ••

 182

-------
                                     IMPLEMENTATION: STATE/H.J. ONGERTH
    .  results in the definite terms used in the standard. However, in view
      of the well-recognized principle  that the significance of bacteriologi-
      cal examinations is variable, and must be  interpreted with  due
      regard to  all other facts known about the particular water supply
      in question, the objection may be raised that a rigid application of
      this standard  will  arbitrarily exclude  a  considerable  number of
      water supplies which conform to all other requirements and which
      competent  opinion  will consider to  be quite safe. The validity of
      this criticism  is recognized, but it  is  not considered  of sufficient
      force  to require or  justify the lowering of  the bacteriological stand-
      ard proposed.  This  viewpoint appears  proper when it is recognized
      that  the definite terms of  bacteriological quality in  which  this
      standard is  expressed represent only agreement as to  safety,  and
      not as to limiting values beyond which demonstrable or even  pre-
      sumptive danger lies. Between  the point on  which the committee
      is in agreement  as  to the assured safety of water  supplies  and the
      point  at which agreement  could be reached as to their dangerous
      quality is  a wide zone. Within  this zone  lie many water  supplies
      which, if  considered in the light of  available evidence from all
      angles, are believed to be as safe as other supplies which conform
      to all  the bacteriological requirements.
        "The committee,  therefore, considers it preferable to recommend
      that in actual practice  the  bacteriological standard be  applied, as
      are other  requirements,  with some  latitude;  in  other words,  that
      supplies which,  on rigid inspection are  found to be satisfactory
   "   in  other  respects  but fail  to meet  the  bacteriological standard,
      may be accepted in the discretion of the certifying authority. In view
      of the character of the personnel entrusted with the responsibility for
;-:-.investigation arid administr:ative  action; the committee feels assured
      this procedure is preferable to  the  alternative of rigid  and auto-
      matic  application."
   In one of the appendices  of the  report, the committee's  philosophy on
 attainability of  the  standards is  presented as follows:
        "As to the reasons for specifying these particular limiting values
      rather than some other values, either higher or lower, it is  obvious
      that the assignment of any definite limits of bacterial content as a
      criterion of the safety of water supplies of diverse origin and history
      must necessarily be an  arbitrary  procedure,  because the  relation
      which the ^terminable  bacterial content bears to the  actual safety
 ....'_ of._a._water_.supply_is^ variable  and to  some extent indeterminate.
      Therefore, all  that may be  claimed  for the  standards  proposed is
      that, hi the judgment of  this committee, they are reasonable; that is,
      are consistent with the other requirements specified as to source  and
      protection  of the water supplies  in question, afford an ample guar-
      antee of safety, and.can be met without too costly and burdensome
      effort. In connection with  this  hist consideration,  the committee
      has  analyzed the records of daily ^examinations  of a considerable
      number of  municipal water  supplies for  the  years 1919: to  1922
      and finds .that!.a substantial majority of,them conform to both re-
      quirements of the standard."
   Whether  or  not, these statements reflected  general  professional  opinion
 on the subject or whether they  molded! professional opinion cannot  be said,
 but most certainly  as I saw the standards applied  in the decade of  the 30's

                                                                        183

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


and on,  practices in application of the standards  were consistent with these
views.
  The drinking water standards, from the very beginning, were adopted
under tlie  authority of the Interstate  Quarantine  Regulations and were  the
standards for the  water to be used on common  carriers, railroads,  vessels,
etc.;  carrying passengers  iri  interstate  traffic.  In addition, these coliform
standards (of course, the  organisms were called B. coli in the 20's and early
30's) were generally recognized as the nation's drinking water standards and
applied by the states. I do not know the details of the history of the use of
these standards  state-by-state. I believe  that gradually more states formally
adopted  these standards,  as their own, by reference. As to  compliance and
enforcement, undoubtedly this was a  mixed bag  around the country. First,
speaking just about compliance for waters used on interstate carriers: Until
1942,  sampling  for compliance with the standards was  done partly at  the
source and partly  hi distribution  systems. Unquestionably, practices varied
from place to place.  It is likely that,  as  with  places  studied for the  1925
standards,  most  of  the sampling was done at the  point of production of  the
water, at treatment facilities or at ground water sources. It is significant that
as to compliance, water systems will more readily comply if sampling  is done
at sources  than if sampling is  done in distribution  systems.
  The 1925 standards  contain no directions or limitations with  relation to
sampling location. It is likely that the practices at the time  were to  sample
at the point of water production and to a lesser degree in distribution  sysems.
The 1942  standards,  in recognition of the Chicago amoebic dysentery out-
break attributed  to  cross connections, required that the bacteriological exami-
nation of water be of samples collected throughout the distribution  system.
In  other words, source samples  could not be utilized in considering  com-
pliance.  Thus, even though  the  coliform  numbers for the  standards  were
unchanged  from the  1925 standards, in  fact  these  standards  were more
difficult to attain.
  At this tune compliance was based on annual  results—a summary of all
samples collected during the reporting year.  Sometime in the late 1940's  the
Public Health Service administratively  changed the ground rules and there-
upon required compliance on  a month-by-month  basis. This also increased
the difficulty of compliance.
  The remainder of  this  discussion concerns  the conduct of the coliform
fnonitoring-compliance program in California through  the years. The  Bureau
of Sanitary Engineering was created in 1915. At that time little water was
receiving treatment. The  city of Sacramento obtained water without treat-
ment from the  polluted  Sacramento  River. According to  the  July  1915
monthly bulletin of the California State Board of Health,  Sacramento had
a death rate from  typhoid averaging about  53^  per  100,000 annually  for
each  five year period  since 1895 although some  of these were  "imported"
cases. The case  rate, as contrasted to  the death  rate, was  about ten times
as high.  The typhoid death rate for the entire  State was 13.6 per 100,000
population. The program emphasis for the Bureau of Sanitary Engineering
for the next 15 years was to get sewage out of water supply sources, to  get
adequate treatment for surface water  sources,  and to  eliminate hazardous
wells.  The first  statewide comprehensive  sanitary engineering   survey  of
community water systems was made in 1930. Beginning in the mid-30's with
the advent of WPA and  PWA projects, program  emphasis was transferred
to development  of sewage treatment  facilities  to replace raw  sewage  dis-
charges.  When I came to  work in the State Health Department in 1938,  the
water supply program was focused on (1) dealing with "problem" systems
184

-------
                                   MPLEMENTATION: STATE/H.J. ONGERTH
 to get correction of major defects, and (2) annual inspections of the interstate
 carrier systems.  I  estimate that at that time the Department  had probably
 four or five hundred systems under surveillance.
   Efforts were  made to bring systems into compliance  with the coliform
 standards when they were not met, but. the major emphasis was on improving
 physical works and operating procedures that were likely to  result in,  or had
 the  potential for,  production  of  poor quality  water.  Compliance with the
 bacteriological standards  was (and still is) a secondary matter. Public health
 is founded on the solid rock of prevention. Our philosophy is that the safety
 of public water  supply is assured through attention to the development  of
 satisfactory facilities and  existence of capable management and operator per-
 sonnel. Sampling and analysis is only one part of the surveillance process and
 provides after-the-fact information, analogous to watching the  scenery  from
 the rear observation platform of a train. In 1938 a few utilities—the larger
 muncipally owned ones, such as Sacramento, San Francisco,  East  Bay MUD,
 Los  Angeles, and San Diego—and 2 or 3 major investor-owned utilities  such
 as California Water  Service, carried out routine bacteriological  monitoring
 programs with samples analyzed  in their own  laboratories.  For  the  rest  of
 the systems,  the  State Health Department did infrequent  sampling. Samples
 were collected at least once a year at the interstate carrier systems, which
 numbered 100 or so. For the rest, water systems were visited  at least  on a
 decennial basis,  and more often where the problems were known to exist.  At
 the time of  the  decennial visit,  a thorough investigation was  made  of the
 physical works,  facilities operation,  and relevant  operating records  to the
 extent that any existed. Bacteriological and chemical sampling also was done.
 The  water at some of the places met the coliform standards,  and at other
 places, it did not.
 , .As the years went on, and = additional manpower became available—and
 this  was always a hard struggle—the program  was tightened up. Gradually
 a policy evolved of requiring all  water  utilities to  collect water  samples  in
 compliance with  the Drinking Water Standards  frequency  requirements, and
 to have  those samples analyzed  in "approved" laboratories. Beginning  in
 the early 60's, we  initiated an  aggressive program  of  requiring compliance
 with the frequency standards. As more; places complied, we  began collecting
 and  organizing statewide . statistics with relation  to  compliance with the
 bacteriological  standards. As  of  about  1968  approximately 30%   of the
 water systems under  our jurisdiction—1100 with 200  or  more service  con-
 nections—failed to meet the coliform standards one or more months  out  of
 the year: Perhaps ~a third of those-not in compliance failed multiple months,
 some failing as much as 6 or~more months of the year.  For the past 8 years,
 we have worked very aggressively  to get compliance, following up month-by-
 month with those places which failed to meet the coliform standards the
previous month. Now, after about 8 years  of quite energetic effort, though
with no efforts at legal action, the statistics for 1976 show about 8.9% (97)
 of approximately 1100 systems failing to meet the coliform standards  one  or
 more months. Last year 2 places failed to meet the  standards for 5 months;
2 failed for 3 months; 14 failed 2 months; the remainder (79)  for only 1
month. These statistics are  low since  compliance has been  calculated from
 running series of the last  20 samples.
  As I have indicated, the "state" program deals  with water  utilities  with
200  or more  service connections.  Local health departments deal with the
less than 200 service  connection systems, and the current inventory records
about 4400  community water systems under local health department juris-
diction. These systems are  only  partly under  routine bacteriological  sur-

                                                                     185

-------
    ']! "•''» III1  '"I1 In	VMrnii Illlllltl 1'i il'llll^^          llllllillllp1   "'  i!	'» IT  I	 'II1 'I'' ', 'j'l	, i .nil	.iiilllllpllliinllili	„	': ' ,,i'i|	 rl	Ti1,,,"  	Ill  , 'illil",!'v ,i ' iiil'l" i|i|i|ll!ll""i" IH.11111"1 IT'1; ,,'!' "IIII	I' J""i1 Tl'iiii!" • 	'',  ' '  '' :l '"'i'llll'j',  IK1,11» I1,'1!!'! 1'l,''ll,ll!lll!ili!ll'
                                                                                       I         . .
                              " ,          ,      ,          '.     .       ,      I   	' .    Ill'1
                                                                                       !!    ' i" ,
                           11  ;                      ,     '                 • '    IF ,     I .,  I      ,   :
                                  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


                                  veillance,  and statistics are  available  for calendar 1976  for 2136 of these
                                  systems. The results show that 32 systems  failed the coliform standards 6
                                  or more months; 209  systems failed 2-5 months; and 162 systems failed 1
                                  month. In summary, 403  systems failed 1  or more months, amounting to
                                  approximately 981  system/month failures. This  would have  produced  that
                                  many public notifications  of failure to meet coliform  standards. As noted
                                  above, these statistics  are on  the  low  side  because compliance  was  cal-
                                  culated from running  series  of the last 20 samples. In addition, it is likely
                                  that these  statistics in part represent  the better small water systems. If all
                                  systems were under bacteriological surveillance,  the  proportion of  failure
                                  to comply would be higher. Most certainly there is a big problem of bringing
                                  these systems into compliance with the  standards.
                                    A final  comment—and an important one—what are the causes for failure?
                                  Of course, they  are varied. Almost no failures occur at the point of produc-
                                  tion at water filtration facilities. Some failures are traced to malfunctioning
                                  of cUormation facilities at small  places  with surface water sources  and no
                                  filtration  facilities. A  few failures are  traced  to  coliform  "contamination"
                                  of wells.  (We will not accept  chlorination of coliform-producing well water
                                  as a method of compliance. Wells must produce  coliform-free  water or be
                                  cut  off the system.) Other  causes for "high"  coliform  are  storage  facility
                                  defects; for example, birds gaining  entry, work on  distribution lines  (main
                                  breakers,  etc.)—coupled with unsuccessful  main  disinfection—and,  finally,
                                  the  elusive "sampling error."   ..--..
                                    Follow-up samples are required at all places failing to meet the standards.
                                  In a substantial  fraction of places, the follow-up samples are free from coli-
                                  form, and the original positives are attributed.to "sampling error"—or to some
                                  transitory  condition  of water  quality.  If the cause of the positive  sample
                                  is a real thing, further positives may result—though sometimes  sporadically.
                                  As to cross connections producing coliform  contamination, few if any cross
                                  connections have been  discovered through  bacteriological sampling. This is
                                  not to say that cross connections have  not resulted in contamination of water
                                  systems—-the reverse is our experience.
                                    Since 1970 California statutes have  contained a public notification  section
                                  as follows:
                                         "When it is determined by  the department that it is in the public
                                       interest, -the department  shall  notify a person, who has not met the
                                       standards of the  department  for  water quality of such noncom-
                                      pliance, and shall require such person to notify in writing each  of
                                       his customers of  the  department's determination that  the quality
                                       pf tfte water fails to comply with the standards,  requirements,
                                '      or conditions established  by the department arid to include any
                                      comments of the department regarding the possible dangers because
                                      of such noncompiiance. The content  of such  statement  shall be
                                       approved by the department  prior to distribution. Notification by
                                       such  person shall be  repeated at intervals as required by  the de-
                                      partment or until the department determines that there is compliance
                                      with the standards or  requirements of the department."
                                    In  1973  the  Legislature  added  the  following  with relation  to  public
                                  notification:
                                        "Upon determination by any person who  furnishes  or supplies
                                      water to a user for domestic  purposes that a  significant rise in the
                                      bacterial count of water has occurred in water which is being sup-
                                      plied  to a user for domestic purposes within the State of California,
                                      the person  who furnishes or  supplies such water shall provide,  at
;„;  -:.,            ;;;,           .    ,;;:;                        ,             ,      •   •         j '
                               	  186

-------
                                   IMPLEMENTATION: STATE/H.J. ONGERTH
     such person's expense,  an analysis of such water  within 24 hours
     to the county health officer and the State Department of Health.
     As used in this subdivision 'significant  rise in the bacterial count
     of water' means such rise in the bacterial count of water as the de-
     partment determines, by regulation, represents an immediate danger
     to the health of the water users."
   Regulations  adopted  pursuant to  this section define significant  rise  as
 follows:
        "(a) Significant rise in bacterial count representing an immediate
     danger to the health of  water users is defined as an  increase in
     coliform bacteria,  when associated with  a suspected  waterborne
     illness or disruption of  physical works or operating procedures.
        "(b) The  following  criteria shall be  used as  an indication of a
     significant rise in bacterial count, and is applicable to routine bac-
     teriological samples . . .:
            "(1)  Whenever  all five portions of more than one sample
         collected during a week show the presence  of coliform or-
         ganisms; or
            "(2) Whenever  50 percent or more of  the portions  of  a
         group of two or more samples  collected during a day show the
         presence of coliform organism; or
            "(3) Whenever  the daily samples, collected following un-
     . •— satisfactory: samples .  . .,  show   the  presence of  coliform
         organisms  in three or more portions  in each sample if more
  ;       than  one sample is taken,  or  in all portions  if only a  single
         sample is taken."
  Upon a  finding of significant  rise, notification of both the Department
of'Heaith and  the water users is required, as follows:-
       "(a)  When the coliform "limits  specified  above  are reached or
     exceeded,  the water utility shall immediately notify the department
     by telephone  of the bacteriological  findings and, at the  same tune,
     shall furnish  information on the current status  of physical works
     and operating  procedures which may  have caused  the elevated
     bacteriological findings, or any information on  community illness
     suspected  of  being waterborne.
       "(b) Upon receipt of  informatiion specified in  (a), the depart-
     ment,  taking cognizance of all available  information shall deter-
     mine whether or not continued delivery of water would constitute
     anTimminent  danger "Ma the health of the  users.
 	--"--:----*'(1)-Upon- finding  an imminent  danger,  the department
         shall  immediately  notify the water  utility of  this finding and
         of the  necessity  to implement  the  emergency  notification
         plans ..."
_ These sections have recently been, modified where necessary to bring Cali-
 fornia statutes  into compliance with PL-93-523.
  In conclusion, an opinion  on the attainability of the current bacteriological
 standards:   These standards are stringent—only very  minor  defects can
 produce failures to meet.the., standards-—but they are attainable when  every-
 thing is in  good order.  Unless there is  universal chlorination with chlorine
 residuals carried to  the  end of all distribution systems, a  fraction of water
 systems will always fail to meet the standards. I consider that this eventuality
 does not represent a "public health problem of any significance. The problem
 with compliance is that with  current waterworks practices,  a  great deal of

                                                                      187

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
regulatory effort is necessary  in order to  get a reasonable degree of com-
pliance. Meanwhile, I am sure some states have a  long way to go.
                                                                     	;t
188

-------
              CANADIAN DRINKING WATER STANDARDS/T.P. S1JBRAHMANYAN
                   Canadian  Drinking Water  Standards

            T. P. Subrahmanyan, D. A. ScMemann, and A. S. Rhodes
                         Laboratory Services Branch
                         Ontario Ministry of Health
                         Toronto, Ontario, Canada

                               3. Henderson
                    Canadian  Public  Health Association
                          Ottawa, Ontario, Canada

                                E.  Bowmer
                          Division of Laboratories
                   British Columbia Department of Health
                    Vancouver, British Columbia, Canada

                               K. R. Rozee
                       Department of Microbiology
                           Dalhousie University
                       Halifax, Nova Scotia, Canada

                               P. Payment
                         Institut Armand Frappier
                     	Cite de Laval
              .                Quebec,  Canada

       ^desirability of having uniform  drinking water  standards throughout
 Canada has long  been recognized. In early attempts to achieve this, the
 considerable amount of  data  and expertise available  in the United States
 was  utilized.  This was done mainly through cooperation between  national
 agencies  and by consultation  with many  scientists active  in  the  relevant
 fields. Our present attempt to  coordinate U.S. and  Canadian efforts  follows
 this tradition.
   We do  not claim to represent  the opinions of every agency involved in
 the regulation  and enforcement  of drinking  water standards in Canada.
 However,  as members of a major committee (formed by the Canadian Public
 Health Association)  involved  in the revision of  Canadian Drinking Water
 Standards,-we-hope-to present  a-Canadian consensus arrived  at  by  this
 committee : in its presentation  to  the  Department of-National Health  and
 Welfare, Canada.
   The history .of drinking water standards  in  Canada  dates back to 1923,
 when the  Federal Cabinet  passed  an order-in-council  which provided  a
 standard of bacteriological quality applicable to water used for drinking  and
 culinary purposes  in vessels using  Canadian  inland  waters including  the
 Great Lakes.-In-1930rthese-regulations v?ere extended to all types of common
 carriers  crossing international  borders into Canada  and  inter-provincial
 borders within Canada.  Canadian coastal shipping was included in 1937.
,  .Drinking water  standards became more  generally  applicable  when  the
 Department of. National  Health  and  Welfare, with  the approval  of  the
 Dominion  Council  of Health, began  to  apply the 1943  United States  Public
 Health Seryicet.JStandards,  pending an examination of the.situation in Canada.
 In the, interim, Canadian  authorities referred to the  U.S.  standards of 1943
 and 1962 as well as those published by the World Health Organization in

                                                                     189

-------
,11 '!'	i;  ,
li! 	i, t'i!1
                               MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


                               1958  and  1963.  The Canadian Public Health Association (CPHA)  also
                               recognized the need for standards suitable for Canadian conditions which are
                               not always the same as those  in. the  United States  and other parts of the
                               developed world.  It formed a Drinking Water Committee in 1966 to develop
                               appropriate guidelines. A year later, the Department of National Health and
                               Welfare convened an Advisory Committee on  Public Health Engineering.
                               The task  of drawing up drinking water standards was  delegated to a sub-
                               committee which decided to combine forces with the earlier group formed
                               by CPHA With  the able  assistance of Mr. S. K. Rrishnaswami, who was
                               appointed principal  investigator, the  joint committee prepared preliminary
                               reports and drafts for desirable standards.
                                 Three levels of drinking water quality were designated on the basis ot tne
                               premise that water  for domestic use should  be free from pathogenic or-
                               ganisms and their  indicators  as well as from other deleterious  substances
                               besides  being palatable and  free  from  objectionable  odours,  tastes and
                               colours. The poorest of these quality levels  was  called  'maximum per-
                               missible"  the next "acceptable" and the most desirable level, which should
                               be the goal was called "objective". It was emphasized that the poorest quality
                               should not be permitted for any length of time.  On the basis of the detailed
                               recommendations of the committee and taking  into  account other  national
                               and international standards,  the 1968 Canadian Drinking Water Standards
                               and Objectives  was drawn up and published by  the Department of National
                               Health and Welfare.                                               .„
                                 The original committee emphasized the need for continuous surveillance
                               of changes in water quality and for studies  to develop basic information  on
                               Canadian conditions and requirements. In  1975, the  Department of National
                               Health and Welfare established a Federal-Provincial working group for  re-
                               vising the Canadian Drinking Water Standards,  a task which is expected to
                               be completed by 1978." As part of this process,  a number of submissions
                               have  been made  including the  report by the CPHA advisory committee of
                               which the authors are members. One of the  authors (J. H.) prepared a draft
                               for a Criterion  Document  entitled  "Microbiological Quality of Drinking
                               Water" based on current knowledge drawn from  national and international
                               sources.  The  criteria document with recommendations  was finalized  and
                               Submitted to the  Department of National Health and Welfare.
                                  Regulation  and  enforcement of  water  quality  standards in  Canada  are
                               traditionally vested  in provincial agencies with the territories in the north
                               falling under federal jurisdiction.  Differences  can  and  do exist  between
                               provinces. The situation in Ontario, for example, is complex. The Ontario
                               Ministry  of the Environment (formerly Ontario Water Resources Commis-
                               sion)  is responsible  for the overall control of regulation and enforcement of
                               water quality in  public supplies. They  work in cooperation with the  Min-
                               istry  of Health, especially when potential health problems  are encountered.
                               Both  Ministries maintain large  water microbiology  laboratories centrally in
                               Toronto  and  smaller ones in  various  regions  of the  province. To avoid
                               duplication of efforts, the Ministry of the Environment laboratories  limit
                               their  activities  to public piped  systems  which are under  their control. The
                               Ministry of Health  laboratories deal with specimens collected by local and
                               provincial health agencies. In addition, major municipalities such as  Toronto
                               have  their own testing facilities and in these cases,  the Ministries play con-
                               sultative roles.                                                          .
                                  Similar jurisdictional delineations with  some  unavoidable overlaps exist
                               in other  parts of the  country. The  regulatory agencies  at provincial and
                               municipal levels follow the current Canadian Drinking Water Standards with

                               1 I9°                                        """     .!  '',',              '
                                                                             iii,;i'Jikiiil	ili.'4iii'i!!,,f'.i Jl'illnld	n; •;.!	•./1,.., i,.',i. .viii'ililJuiK..\,„,. _,,,L.... ..ViL'.-_I'«\L_
|.ilji!B||ihhijlli! lLhiiillJii.1,1 .ilL,.,,,  IIIIIIIBIill iiliii iJiiii....,/,,, ' •, iliii;!, iiiillliLililllllllll" I. ,1 Jliniilillllllllili., ,!!• •, i III jil	4,1 Jlllllill	II .«• ill	liliillliililiiil n, n If1!. I,. „,	;l L,, ^ L _. .4. ^i1' k. jyjiiu :«. i ..* .J'l,'.!._..!_,.. iL ;..,' j, ,_,*_,_ ^l _ju jlJUi, iMiill, • Ji.iiilniii.i n I 'i .MI .' • 'i, Jllli'! &. •iiil>;! i'JmlJiil	Ml Mitilli idd	«: •;.!	' /1, n ^ ivHliI i.il»i i ,„... _,, iL.... ..VillL'.- < '<> i.i. J	••	 '.I. > iHO® •• 'ii i1' '"4 r. iil'.. .•.!; u!i il.1 jil

-------
              CANADIAN DRINKING WATER STANDARDS/T.P. SUBRAHMANYAN


  modifications designed to suit  perceived local needs  and to improve the
  service.  Some,  such as the  Ontario  Ministry of  the  Environment,  have
  developed their own objectives, which  conform to the national standards but
  specify details of improved sample  collection  or additional examinations.
  Thus the results of  the  attempts  to  attain a  measure  of uniformity are
  encouraging.
    The 1968  Canadian Drinking Water Standards specify that  methods given
  in the current edition of "Standard Methods for the Examination of Water
  and Wastewater (APHA)" should  be  employed in the routine examination
  of water.  It is also recommended that siampling frequencies should conform
  to the U.S. Public Health Services recommendations of 1962 with the proviso
  that frequency and location of sampling be established by the control agency
  or  Medical  Officer of  Health after adequate  investigation of  the source,
  method of treatment and protection of the water supply. Raw waterjs  con-
                                         of the samples STany consecuUve
30-day B^nod^have^je^l^oJiforja^ejasjities of less than  IQOjref  100 ml and
an^aTcolifo~                          ~
             ^^
        colifonn~density of IRSS than 1 nn(riper~TUu ml. although the organisms
                                               relationship.
  •"FmisHeddrinking water is considered "acceptable" if 95% or more of the
 samples tested in a 30-day period  are negative for  total coliform organisms
 when enumerated as a Most Probable Number (MPN) index by the multiple
 tube fermentation method.  None of the positive samples should have MPN
 indices greater than 4 per 100 ml.  Waters with MPN indices of 4 to 10 per
 100 ml in more than three consecutive  examinations should be immediately
 investigated in respect of source, treatment and other parameters. According
 to  the  1968 standards, the membrane  filter (MF)  coliform test  should be
 used to determine the potability of a given water supply only after adequate
 parallel testing  has demonstrated  that  the  MPN  and  MF methods yield
 equivalent information-relative to  the  sanitary quality of the  water supply
 system.  The MF method  should  be carried out  on  a minimum  sample
 volume of 200 ml. Drinking water is considered "acceptable" if at least 95%
 of  the  samples  hi  a  30-day period are negative and none  of the positive
 samples show1 MF counts greater than 4 per 200 ml or 10 per 500 ml.
  Any  water with MF counts  greater than 6 per 200 ml or 15 per 500 ml
 or  approaching these limits in  three consecutive tests should be immediately
 investigated. A special series of samples and the commencement of suitable
 and effective remedial measures must follow any deviation  from 'MPN or
 MF standards.  In such an event, all samples must be carried to the "com-
 pleted"  test" procedures.  In addition, the demonstration of any pathogenic
 organism in finished water must lead to the rejection of the supply until effec-
 tive remedial treatment and disinfection have been completed.  The Medical
 Officer of Health or the control agency  may issue an order requiring boiling
 of water or other corrective action in appropriate situations.
  There is no published information on whether water quality standards  are
 in fact  being met. However, we were able to draw  upon the experience of
 Mr. L.  Vlassoff,  Chief, Microbiology Section,  Ontario Ministry of the En-
 vironment. In his  experience of many years involving over 500 water treatment
 plants,  enforcement 0* standards  has mot presented serious  problems.  In
 these examinations, a  simpler  sampling  system is used. A  minimum of 8
 samples per month are collected at weekly intervals from small communi-
ties but  for large centres, the number of samples is at least the same as in the
 1968 standards. MPN, MF or Presence-Absence tests are carried out on 100
ml  samples. About 5%  of the specimens require  detailed  testing,  which
includes tests for  several individual bacteria. As a result  of the good rapport

                                                                     191

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


established between  microbiologists  and  engineers  managing  public  water
systems in Ontario, there is good cooperation in the enforcement of stand-
ards and in taking rapid remedial action as well  as in carrying out additional
examinations when indicated. The Drinking Water Standards specified for
public supplies may not be applicable to other water supplies which'are also
likely to have enforcement problems. In revising  standards, this aspect should

*^££±±-«f the recent CPHA committee on  "Microbiologi-
cal Quality of Drinking  water"  are  summarized  below  and  the more  im-
portant sections compared with the corresponding secttons in the 1975 V*.
National Interim Primary Drinking Water Regulations. The detailed recom-
mendations of the CPHA Committee are available  from the Canadian  De-
partment of National Health and Welfare.*
     1. Since modern  technology is capable of purifying water even from
         heavily polluted  sources,  it was  felt that the major concerns should
         be adequacy of treatment and safety of the finished water.
     2. The total conform group is preferred  as an indicator of treatment
         adequacy and  fecal coliform. measurement is preferred for  monitor-
        ing raw water quality and to  indicate  the  potential presence ol
         pathogens at source. The present total  coliform standards should be
         retained but a minimal sample volume of 200 ml should be stipulated.
         The more detailed U.S. regulations differ from this in  that the recom-
         mended sample volumes  are not the same and that maximum per-
         missible contaminant levels  (MCLs) are given in  detail. When the
         new Canadian Standards are drawn up, the final recommendations
         are likely to show, a greater similarity to the U.S. regulations.
     3   The general bacterial population  should be monitored regularly. The
         "acceptable" level should be limited to 500 organisms per  ml since
         at higher levels, coliform measurements may  not  be reliable. An
         additional recommendation  is that the definition of coliforms should
         include the cytochrome oxidase negative  characteristic in  order to
         exclude members of the genus Aerompnas.
  "  4  A free chlorine  residual of 0.3  mg per litre should  be maintained
       '  throughout  the system,  the turbidity  of water entering the distri-
         bution  system  should  be limited to one Jackson turbidity unit. The
         pH should not exceed  7.5. If higher pH levels have to be maintained,
         it would be necessary to specify higher chlorine residuals for more
         alkaline  conditions.   When  chlorine  residual  measurements are
         substituted  for  bacteriological  examinations,  they  should  not ex-
         ceed 75%  of  the required  examinations. In place of each colirorm
         test omitted,  three chlorine residual determinations  should be car-
         ried  out  It  was  considered  desirable that  the  chlorine  residual
         should be  determined daily but  it was left  to the  control  agency
         to determine the frequency  of tests. CPHA recommendations differ
         from the U.S. regulations  in specifying a higher level  of chlorine
         residual but this is lower than the  0.5  mg per litre required by the
         1968 Canadian  Standards.  Another difference  is in  the number  ol
         chlorine  residual  measurements  recommended for  each  substitu-
         tion  The mcrease in the use of ozone  as a  disinfectant  by water
         treatment plants,  especially in  Quebec,  has  to be considered  m

         •, :   f    ••  •      ".' ,     :  • ,     ,';':.-    i; j.: _  ; _  ;;"._   '  • :L • .- j^
   ••'* Environmental Standards Division, Environmental Health Centre, Health and
 Welfare Canada, Tunney's Pasture, Ottawa, Ontario K1A OL2.
 192

-------
              CANADIAN DRINKING WATER STANDARDS/T.P. SUBRAHMANYAN


          evaluating the merits of substituting bacteriological tests by chlorine
          residual monitoring.
      5.  While  sampling frequency  based on population served  might be
          realistic and adequate for large  centres, it was felt that the require-
          ments for smaller communities are inadequate and that the minimum
          sampling requirement should be raised to 10 samples per  month.
          Where this is not practical and  less than  10 samples are examined,
          the present standards should be  modified  to  stipulate that none of
          the samples should  be positive  for coliform organisms. When a
          sample is positive,  corrective action should  be .taken immediately
          and tests on daily samples carried  out until no coliform organisms
          are found in two consecutive samples. Greater emphasis should be
          placed on sanitary surveys to detect problems in the system. Sampling
          from trouble spots  should receive greater attention than the  col-
          lection of representative  samples.
      6.  In view of the unacceptable degree of variation in the efficiency of
          different membrane  filters in  recovering coliforms, specifications for
          membrane filters are required; these  should cover all the parameters
          that affect the recovery of these  organisms.
      7,  Another  aspect requiring attention  is quality control. The  use of
          simulated water specimens for comparative testing in different  labora-
          tories  and  the  development of  reference  procedures with which
          methods used in the different laboratories may be compared were
 ~  7 .    suggested.                                          :
      8.  Research  into techniques for simple methods  for the  detection of
          streptococci  of exclusively  fecal origin  has  been  recommended.
          Health hazards  due  to the presence  of Klebsiella organisms  in  wa-
          ter and the possible use of fecal  sterols as an absolute indicator of
    '"""   fecal pollution" merit further study.                  :
      9.  Virological  examination of water was discussed  at  length.  It was
          decided not to recommend  virological examination  as  a routine
          procedure although  the  present  bacteriological standards may  not
          reflect viral contamination. It  was felt that since recent  technological
          developments  are encouraging,  virological  studies  should be  car-
          ried out selectively  and that virological surveys  of  drinking water
          should receive priority when developing research programmes.
     10. It is necessary to  determine whether standards are being met,
         whether bacteriological surveillance  programmes  meet  established
         criteria  and whether the water supply systems  are adequate for pro-
'~	~  "viding  safe  drinking wafer.  Therefore,  a  drinking water  supply
         survey  which  encompasses" "different parts of Canada  has been
         recommended.
   Finally,  some of the  differences between the  Canadian  position and  the
 revised U.S. regulations reflect the  fact  that the  revision process in  the
 United States is at a more advanced stage. Further, the Canadian Standards
 provide only guidelines  for  regulatory agencies. When the  new Canadian
 StanHanls" are finalized on the basis of the CPHA recommendations and other
 submissions,  some of these  difference!?  may disappear.  Those  based  on
 differences between  our  situations will,  however,  remain.  We  wish to  ac-
 knowledge the generous  cooperation we  received from Dr. P.  Toft,  Chief,
 Environmental Standards Division, Health and Welfare Canada,  Ottawa and
 Mr. L. Vlassoff, Chief, Microbiology  Section, Ontario Ministry of the  En-
 vironment, Toronto.


                                                                      193

-------
                                MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                 ,.	:  QUESTION AND ANSWER  SESSION

                                 C.W. Hendricks, Office of Drinking Water, U.S. Environmental  Protection
                                 Agency, Washington, D.C.

                                 Do you think suppliers will sample at the rate of 8 to 10 per month?
                                 T.P.  Subrahmanyan,  Enteric  Virus  Laboratory,  Ontario  Department  of
                                 Health, Toronto, Canada.
                                 The water supply  utilities in Ontario, from which most of the Canadian
                                 experience is drawn, are run either by the Ontario  Ministry of the Environ-
                                 ment or by other public  agencies in cooperation with the Ministry. In the
                                 case of the major municipalities, the collection of 8 to 10 samples  per month
                                 has not presented significant problems. I am not in a position to comment
                                 on utilities run by  private  companies.
                                 194
«|l|> llllllu'll II K : Jll I ill) Hi	 ;, "' , liiiiiU , l-l'Lr . ,l . ., '	. ,	Iliiiii - . lalllll ! Jill /,!':,;	';

-------
Kansas City, Missouri Water Treatment Plant
PANEL DISCUSSION


           Alternative Means of Determining Compliance

Chairman: Dr. Sumner M. Morrison, Department of Microbiology, Colorado
State University, Fort Collins, Colorado
                                                             195

-------
K	?,••!  SB  iilIKi  itid!	
                                                                              S;,;.,-!.!,:i; L 'iiiiiti '& mi11  •;., s '.-i "ii,!>	   .  „:;:::	Hi it Hit; i;•«	1 •	 • i, Ji:     AJU,	.:tv
                                                                                                                                                                                            llllllli    iiiiiii    ii	111 ill	nil
                                                                                                                                                                                                                                            III  III
                                                                                                                                                                                                                                                             •nil

-------
                      ALTERNATIVE MEANS OF COMPLIANCE/S.M. MORRISON


              Alternative Means of Determining Compliance

                           Simmer M. Morrison
                       Department of Microbiology
                        Colorado State  University
                       Fort Collins, Colorado 80527

   Microbiologists, in the area of laboratory work for the surveillance of
 water supplies,  are sort of in the analogous position of the common expres-
 sion "painting yourself into a corner".  Well developed, simple,  inexpensive
 tests for coliforms exist  and are being productively  applied  in; the control
 of drinking water quality for  a large portion of  our population.  In con-
 junction with the work of the  engineers and health officials,  the laboratory
 work has contributed to  the decline in the  waterborne illness over  the past
 50-60 years. Now the mandate is to expand this  supply of safe water to all
 public systems, including many, many small ones. This philosophic or political
 change in the need for safe water for  all has created our problem. What was
 considered  a simple, inexpensive, expeditious  and effective  procedure for
 metropolitan population centers now  is considered a procedure, for many
 reasons  already  discussed, too  cumbersome, costly and ineffective  for the
 small, systems. Some inconvenience,  cost,  dislocation and personnel train-
 ing  is going to occur in  meeting the charge that Congress has  given EPA
 and us as scientists and professionls with a  stake in near  zero risk drinking
 water for all.
   I  would like to comment about a few points that may be alternate ways
 of establishing compliance. Some of these  ideas may be "far out"  and not
 worthy of consideration, but this panel is to explore all possibilities.
   1. While population served seems to be a very logical way to assign re-
 quired tests, is there  reason to consider volume treated or volume delivered
 for  domestic use as a means  of determining required numbers of tests?
 Systems that supply a high proportion of then- water for  industrial, cooling
 or irrigation uses are different  from systems  that supply domestic  water
 only.
   Similarly,  is there reason to  establish categories of requirements  based
 upon type of water and type of treatment system, or perhaps the  availability
 of engineering support?  Perhaps  the  information we now  have  would not
 allow such an approach, but if better statistical risk  data were to be accumu-
 lated we might  be better  able  to provide the same degree  of protection  to
 all people with a more equitable cost  to each.
   2. The use  of monthly enumeration  and averaging of samples is an
 arbitrary use of  time periods.  A system  using  a  30 day,  or even a 60 or
90 day running average would better indicate  deteriorating water systems.
 The arbitrary calendar month will cause system operators to  establish test-
 ing times to avoid getting caught with a high test in the last part of the month
 and  other similar ways of not using good statistical  sampling.
   3.  A major fear that seems to have been expressed is that compliance pro-
cedures do not really  take into account possible sampling or laboratory error.
A procedure allowing immediate retesting,  perhaps  for 3  consecutive days,
would allow the  discard of the single high coliform test without setting in
 motion reaction from the State  and possibly public disclosure.
   4. There seems to be an attempt to directly link the MCL of the bacteria
to day-to-day prevention of disease. From  an epidemiological standpoint it
should be placed into proper  context. Periodic sampling of small portions of

                                                                     197

-------
                                     MICROBIOL STANDARDS EYALUATION/C.W. HENDRICKS
                                     water for coliforms, when coupled with proper design and operations man-
                                     agement, has proven quite effective in preventing frequent  large scale water-
                                     borne disease  outbreaks. It has stood  the test of time. Other alternatives to
                                     bacterial testing, as frequent chlorine residual sampling or engineering surveys
                                     only, need to  be tried buiin an experimental mode. Until these alternatives
                                     have stood the test of time, as has the cpliform test,  there  should be both
                                     short-term and long-term surveillance research. To avoid  again not  having
                                     the answers to pertinent questions, the research must  be well-designed  and
                                     directed  by outside as well as  in-house expertise; as free as possible from
                                     self-serving bias.             !..'.'                        ....        f
                                      "5. We have heard  some  previous speakers  analyze and  criticize some ot
                                     our'long-established testing procedures on  the  basis  of modern statistical
                                     analysis.  Among the very important research areas that needs development in
                                     4 above is an  attempt to put such items as sampling frequency, sampling de-
                                     sign, minimal required types  of tests, correlations  with human waterborne
                                     illness patterns, etc. on a solid  defensible statistical basis.
                                       6  To  me, the concept of public notification when a water system appears
                                     to be in  non-compliance, is a two edged sword. While it might provide ap-
                                     piopriate pressure on the water supplier to  be in continuing compliance,  it
                                     could also serve to create complacency in the population  of a community.
                                     In some places  it could lead to severe disruption of the  local community,
                                     endless litigation,  suits for all  types of real and  imagined  damage, political
                                     ictribution,  scare  tactic  competition  and  advertising  and  wasteful and un-
                                     ethical methods to avoid notification.  In  other places,  with no penalties be-
                                     yond  notification, the local people will be  successfully sold on  the idea
                                     that the  notices are bureaucratic garbage and they will evoke no response.
                                     In either extreme case the end  result will not be safer drinking water but
                                     rising consumer costs for the legal and political maneuvers.
                                       Far more effective would be a State program of compliance by cooperative
                                     effort with  a review  and  emergency team  to  attain  the compliance. By
                                     charging the costs for these emergency reaction measures back to the water
                                     supplier, the pressure would exist for  staying in compliance and the chances
                                     for meeting the safe  drinking  water objective would be greater.
                                       With these  few remarks, I would like  to  introduce this  afternoon's panel
                                     speakers.                                               .
                                     The panel members for the remainder of this session are:
                                     Dr. Warren Litsky, University of Massachusetts; Mr. Berry E.  Gay, Jr., of
                                     the Division of Laboratories  in the Illinois  Department of Public Health;
                                     Mr;  Ira  Markwood,  Director of Public  Water  Supplies  for the State of
                                     Illinois; Dr. G.  Wolfgang  Fuhs, Director of Laboratories  and Research for
                                     the New York Department of Health;  and, Dr. Wesley  O. Pipes, Drexel Uni-
                                     versity.  These individuals  will  present  their ideas concerning alternative
                                     methods of determining  compliance to the microbiology standards.
UK 	lliMyiiliil
                    198

i.jkit,uiiU	Jill	iiJM^	i: .Livi!'iiii.	1.:.'.lia	Hiii	ii i. -^	• .,; l,,:;,!i...,,1, ilii. jl-: i.	't	i	it	:,	;!	liliiilf 	a	i,iilii,	 tllilllliiiiiii.iliii::

-------
                                             THE COLIFORM MCL/W. LITSKY


                  The Coliforin MCL: Can We Defend It?

                              Warren Utsky
                   Department of Environmental Sciences
                               Marshall  Hall:
                        University of Massachusetts
                      Amherst, Massachusetts 02003

   During the past day and a half we have heard some  excellent papers con-
 cerning evaluation of the coliform standard, detection and control of coliform
 bacteria and compliance with the coliform standard as they apply  or pertain
 to .the National Interim Primary Drinking Water Regulations. We have heard
 the pros and cons of  many ageless scientific debates and we have also been
 privileged to the  arguments  of a new breed  of ecologically oriented micro-
 biologists and. their rational for adopting alternative indicators of water con-
 tamination. We have listened to the  advantages and disadvantages,  the
 rational and the  public health considerations of substituting the chlorine
 residual for the traditional microbiological procedures. After listening in awe
 to the parade of expert  speakers I asked  myself, "What will the outcome be
 for  all this rhetoric?"
   This question was not prompted by sarcasm or smugness but rather by in-
 creased activity of  consumerism and consumer groups. The days are long
 gone when recommendations or regulations of  Government Agencies were
 universally accepted. Unfortunately, the trend,now is to distrust any govern-
 ment official, challenge  any document put forth and  above all ignore  the
 scientists since "they are the group that got us into this present day mess."
 Many legal actions  initiated by consumer groups against government in  the
 name of _.the ..people have been brought to my _ attention with the result that
 a few of my colleagues  are becoming extremely wealthy by acting as  scien-
 tific advisors to  law firms. Because  of this  trend and because of Section
 141.32 which mandates that persons served by a community water system
 be notified when  the  system fails to comply with the applicable MCL,  I
 predict a rash  of legal entanglements,  the likes of which  have never been
reached, and EPA better be ready for this deluge. As a mental exercise during
 the past few days I have asked myself, "What advice would I give to a lawyer
 who has been retained by a water supplier ajccused of exceeding the bac-
 teriological MCL?" Realizing that the object of any court  action is to prove
 that your opponent or the regulations, etc;, that he is enforcing are  incorrect,
 illogical, or just plain  stupid, the scenario which I present may be a  trifle
biased, to. say .the  least.
   With these introductory remarks concluded, let me present a few items on
 which I would base a case were I acting in the. capacity of a legal advisor.
   I would  most  definitely attack the substitution  of the chlorine residual for
the coliform standard. The data in Table I demonstrate that there is  most
definitely fecal coliform bacteria in water  showing a  high  chlorine residual.
While  these  data are based on sewage  effluent and. combined   chlorine
residual, as measured  by the OT test, they suggest that  chlorine residual
monitoring is not as infallible as its proponents wish us to believe.
   Leaving this subject  I would-then advise the court that while we fully sup-
port the present coliform standard, we  challenge  the  procedure  by which
this standard is measured! In as much as the present regulations give an un-
fair edge to those using the  Multiple Tube Technique over the Membrane
Filter Technique  for  enumerating .the coliform bacteria, I  contend that

                                                                     199

-------
                                               	I	
MICROBIOL STANDARDS EVALUAf ION/C.W. HENDRICKS
                    ,             ,, 	 • .  ..:     '.       .  ,  ,        . ,
TABLE 1   EFFECT OF CHLORINE RESIDUAL AND CONTACT TIME
    ON FECAL COLIFORM DENSITIES IN SEWAGE EFFLUENTS.
Plant
A
A
A

A

S
S
N
N
Effluent
primary
primary
primary

primary

secondary
secondary
primary
primary
CI Residual Contact F. C. MPN
rng/1 mm
1.5
1.0
1.25

0.75
• •: : :. .
1.5
0.5
1.2
1.1
10 7900
12 7700
10 9200
15 270
20 20
11 35,000
15 4,900
20. 2,400
OF* 330
OF 1,100
25 5,400
20 16,000
   * outfall
 the former method will be universally employed and, in the interest of time,
 will limit the remainder of my discussion to the Multiple Tube  Method.
 While the definition of-the coliform group has been accepted for many years I
 w.puld question whether the techniques are actually capable of isolating and
 enumerating  this particular  group. We  all have heard the outcries of Dr.
 Victor Cabelli of  EPA that members of the  Aeromonas  group, a  common
 soil inhabitant, can  not  be  distinguished  from the  coliform bacteria  using
 the  present standard methods.  Thus, are we  measuring  coliform  bacteria,
 Aeromonas species, a mixture of the two,  or God knows  what?
   The use of the MCLs intrigues me and may also amuse the court. Case in
 point, Section 141.14 defines the  MCL  in actual numbers per 100ml  when
 the  Membrane Filter Technique is used while when the Fermentation Tube
 Method is used  the MCL is defined as coliform  present in a 10ml  standard
 portion. In other words the  MCL for the  latter is defined as the number of
 tubes showing gas. By employing  this system, the calculation of the  Most
 Probable Number (MPN) and its very embarrassing  95 percentile confidence
 limits range are avoided. Whatever the reasons for this may be, the  fact still
 remains that both the MCL and the MPN are based on positive fermenta-
 tion tubes and,  call it what you wish, the MCL are going to be correlated
 with the MPN  which we have witnessed many times during the  past day
 and a half.                                                ,    ,   ,
   Now for the  MPN or MCL of coliform bacteria. As  one who  has been
 teaching sanitary bacteriology for the past twenty-five years, I have become
 disenchanted with the accuracy and  precision of the MPN technique for
 coliform bacteria  as  outlined in Standard Methods.  I have  seen too  much
 variation on replicate samples by technicians and students employing flawless
 technique. The precision of the MPN just is not there and the data  in  Table
 2 will illustrate this. Fecal coliform  determinations  were made on marine
 watersj chlorinated sewage effluents and laboratory chlorinated sewage efflu-
 ents hi triplicate by a technician with ten years laboratory experience. As can
 be seen in Table 2, the resulting triplicate MPNs show that a ten-fold  range
 200

-------
                                            THE COLIFORM MCL/W. LITSKY


      TABLE 2. TRIPLICATE FECAL COLIFORM MPN  RANGE.

         SAMPLE                         MPN/100 ml
Marine




Chlorinated
Effluent

Lab
Chlorinated
Effluent
49
79
94
8
700
2,200
49,000
490,000
35,000
240,000
49,000
310
350
220,
1,300
1,700
7,900
130,000
790,000
. 110,000
540,000
130,000
310
540
350
1;300
7,900
13,000
490,000
2,400,000
240iOOO
2,400,000+
490,000
 is not uncommon. These data should not surprise anyone working with the
 MPN technique  and yet the standard is based on this type of an analysis.
   Following this I would attempt to inform the court that other investigators
 have-published numerous papers warning-of the  fallacies  concerning  the
 MPN technique; but.first let us consider the statistics on which the MPN
; was originally based.  ......                ,
   McCrady  (4)  in  1915 proposed the MPN  based on Bayesian statistics, a
 branch of statistics  that modern books do  not even  mention.  I find _ it  im-
 possible to explain what  Mr.B ayes' theorem was and how it is involved with
 the MPN. However, my  friends  in the statistics department informed me that
 the  following statement  from McCrady's paper is the basic assumption of
 Bayes'  statistics as applied to the  MPN. I hope that you are able  to make
 more sense out of it than I could:
        "There  is a certain assumption involved in this principle which
     must be recognized. This  assumption  is that all numbers of B.
     coli are equally probable; that is, that in the long run  of  samples,
     one number of B.  coli will  appear  about  as often  as any, other
     number.
 	It.is to be noticed that this assumption is the same, in kind, that
     is  involved in the determination of the 'Most Probable Number.'
     There,  it  is assumed that  the 'Most  Probable  Number',  so  de-
     termined, is as  probable as any other number."
   In 1957, Woodward (7), then  Chief of Water  Supply at the Robert A.
 Taft Sanitarian Engineering Center, USPHS, discussed the precision of  the
 MPN and he concluded  that:
     .-  "The  lack of precision of MPN estimates of bacterial densities
     is  generally  recognized—at least by those who  perform the tests.
     Like many  other measurements used  and interpreted  by others
     than the laboratory  workers who fashion them, MPN estimates are
     often credited with  a precision they" do not deserve.
       Precision confidence limits have been  computed for  three-tube
     and five-tube tests using multiple decimal dilutions and for the five-
     tube smgle dilution test 'These limits are broader than the approxi-
     mate ones based  on a log  normal distribution  of MPN's. For a

                                                                    201

-------
i	rri ifw "  • i	in111  'i I	i	inni1 ' i	i  ••   "    •• '.'• T  v7^   • : 	•'!	l;l"!i	'.JT "Ji" ••!  	v1" • <. •. '    ;it*	v' •'vM^'m	'n'j	;"1"l; "!"	  *    «•"•	'	i't:>:	r iw;	ira1	HI!
                              MICROBIOL STANDARDS EVALUATION/C.W.HENDRICK&
                                   three-tube multiple dilution test, the 95 percent confidence limits
                                   cover approximately  a 33-fold range from approximately  14 to
                                   458  percent of the MPN  estimate.  For a five-tube multiple dilu-
                                   tion  test, the" 95  percent confidence  limits  cover approximately a
                                   thirteen-fold range from approximately 24  to  324 percent of the
                                   MPN. In short, MPN estimates are not precise and many of them
                                   are inherently improbable."                                      .
                                 Laubusch  (2)  in  1958 noted several limitations in the dilution technique
                              for the determination of MPN of coliform bacteria in water:
                                     "Not  the least of these is that the results depend on the number
                                   of tubes observed in  each dilution and  may be greater  than the
                                   "true" coliform density (as much as 23 percent high for a series of
                                   5-5-5 tube tests and 43 percent high for a series of 3-3-3 tests).
                                   Other limitations  include deficiencies in laboratory  technique  and
                                   personal errors  associated therewith,  differences  of  sensitivity
                                   of alternate procedure methods, influence of one  variety of organisms
                                   on another,  lack of true randomness of bacterial distribution,  etc.
                                   For these reasons, the MPN by the dilution method should not be
                                   treated with sacred reverence."
                                As to the standards Laubusch wrote:
                                     "The  currently  accepted  'Standards,' (6) which involve planting
                                   of a  minimum  of five  10-ml  portions, were not altogether acci-
                                   dental. There appears  to be no rational basis in  the establishment
                                   of a minimum of five  portions, rather than say ten or more, except
                                   that  it is necessary to  limit the amount of labor  and materials
                                   involved to such an extent  that the results are  commensurate with
                                   the requisite accuracy—an entirely reasonable  approach.
                                   "  The most accurate method of evaluating fermentation tube test re-
                                   sults  hi terms of a determinate number is  still open to dispute, al-
                                   though in more recent years this is less obvious due to the general
                                   acceptance and convenience of  tables of  most probable  numbers
                                   from the Hoskins  (1)  adaption of the Reed (5) formula based on
                                   the Poisson  random distribution."
                                 Laubusch then summarizes by quoting Hoskins whose MPN tables we all
                              have used throughout  the  years:
                                     Twenty years ago Hoskins asserted that ". . . it seems reasonable
                                   to conclude that  the  dilution tube method cannot result in an ac-
                                   curate enumeration of coli-aerogenes  organisms and that, therefore,
                                   the need for an acceptable and dependable method still remains."
                                   He added further that ". . . if correspondingly as much study (as
                                   the MPN) were  devoted to the  development of a suitable plating
                                   phase of coli-aerogenes enumeration, it is quite reasonable to believe
                                   that  this effort would yield a method superior  to the present one."
                                 Finally a short quotation from a 1975 paper by. Oblinger and Koburger
                              (5)  illustrating the mathematical bias in MPN values:
                                     McCarthy  et al. .(3)  demonstrated a  substantial  mathematical
                                   bias  in  MPN values relative to plate counts. With the agar plate
                                   count method as  the  control  estimator,  10 replicate MPN's indi-
                                   cated an average bias of +29%, +10%, +6%, and —4% when
                                   compared  to the plate counts'  arithmetic average, geometric mean,
                                   median, and harmonic mean, respectively.  The precision of  10.
                                   replicate plate  counts was  at  least  three  times that of replicate
                                   MPN values. Perhaps Woodward summed it  up the best: "The
                               202

-------
                                             THE COLIFORM MCL/W. LITSKY


     lack of precision of  MPN estimates of bacterial densities is  gen-
     erally recognized—at least by those who perform these tests."
   After disposing  of the  MPN technique I would focus on the subject of
 check samples.  Section 141.21  (D)  (2)  states that when the coliform bac-
 teria occur in three or more 10ml portions of a single sample,  two consecu-
 tive daily check samples  shall  be  collected  and examined  at  the same.
 sampling point.  It can be argued that the check samples are required because
 of the belief that most positive results are due to an error in. the collection
 process or in the laboratory analysis but seldom due  to the fact that coli-
 form bacteria were  actually present in the  water. I  have heard speaker
 after speaker refer to .spurious positive  analysis, laboratory  errors, unsatis-
 factory collection  procedures  .as  if they were extremely common occur-
 rences and thus the need for  check samples. I contend that the collection
 routine and the laboratory analysis  should  be such  that a positive  tube
 indicates just what it should indicate—trouble! I submit that there can be no
 room for  error in a clinical  laboratory especially  during the ;typing  and
 matching of blood; likewise, there should be very little room for error when
 analyzing finished  waters. Like it or hot, the check sample does not  and
 cannot indicate  the actual condition  of the original sample. I see no reason
 why these common errors in the collection of the sample or in the laboratory
 analysis should be condoned as routine happenings. They need not occur  and
 moreover they  should not occur to the  degree that  has  been  indicated at
 this symposium.                         ^             .
"I will be the first to admit that my  arguments presented here are  ex-
 tremely biased and somewhat facetious.  Nevertheless, these are  the types of
 arguments that  may  use in future legal  confrontations—and EPA  had bet-
 ter be ready for them.
                           REFERENCES

 I, Hoskins,  J.K.,  and C.T.  Butterfield.  1935. Determining the bacteriological
    quality of drinking water. JAWWA 27:1101-1109.
 2. Laubusch, E.J. 1958.  As  a  yardstick  for pollution,  how good  is the MPN
    coliform index? Water & Sewage Works J05:334-338.
 3. McCarthy, J.A. et al. 1958.  Evaluation of the reliability of coliform density
    tests. Am. J. Public Health 48:1628-1635.                  .
 4. McCrady,  M.H.  1915.  The  numerical interpretation of fermentation  tube
   TesultsT J. Infect.  Dis. 17:183-212.
 5. Oblinger, J.L.,  and J.A.  Koburger. 1975. Understanding and  teaching the
    most probable number technique. J. Milk Food Technol. 35:540-545.
 6. United States Public  Health Service.  1946. Drinking water standards. U.S.
    Government  Printing  Office, Washington,  D.C.
 7. Woodward, R.L.  1957. How  probable is the most probable number? JAWWA
    49:1060-1068.
     ---.-  -QUESTION AND ANSWER SESSION

Mr. Kimball
What is the problem of false  samples occurring  from improperly sterilized
bottles which,  are done by commercial labs?

                                                                     203

-------
Ill 11111111 1 1 III 1 ' !".''• :"~ 111 < 1 III |i Hill IIP II 111 " ; " •• ,'' ;:""; • 	 1* ' "'" ' 1 1 'in 1 	
Hill II
I'l ' 1 1 '" :''i:, 1 111 111 1 1
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                      	      •   ! •

W. Litsky, Department of Environmental Sciences, University of Massachu-
setts,  Amherst, Massachusetts

It is assumed that employing standard methods for th'e analyses of water  the
results obtained  reflect the condition of the  water. This  requires that all
chemicals, solutions, media and glassware be sterile. A further requirement
is that the analyses be carried out by a  qualified technician m  order to pre-
vent  contamination of the sample,  glassware, diluent  or  media from^the
environment, fingers, etc.  In other words, the bacteria which are counted on
the plate or by  the MPN technique must  originate  from the sample and
from nowhere-else. An improperly sterilized sample bottle can in itself, be a
source of bacteria thereby negating the value of the water analyses. Improperly
sterilized bottles have no place in the water sampling routine. Any legitimate
commercial laboratory should be  able to supply sterile sampling bottles If
they  can not accomplish this simple  feat they should not be tolerated. They
should, like the desert dwellers, fold  their tents and silently slip away.
                                                  i                  , •   '..
Comment by G.W. Fuhs, Division of Laboratories &  Research, New York
Department of Health, Albany, New York
In addition to  laboratory imprecision  of  MPN results,  environmental
variability in natural surface waters (patchiness) can cause a standard  devia-
tion  (in a lognormal distribution) corresponding to a factor of 4, and  some-
times corresponding to one order of magnitude and more. This was de-
termined  at optimal  laboratory  precision  (20 to  80  colonies per  plate,
membrane filter  technique).
                    ...•-.,..  •...•.... , 	 , :....  .. .„ ; .....v|	,	,  „               ..,!'!

Participant from  New Jersey
Collectors' error  had been found to occur  in  the  form of  a batch of
nonsterile  bottles which gave false-positive  results. This meant that isolated
positive results could very well  be  in error.

G.W. Fuhs
This  observation  reflects  a typical quality  control problem and is not in
conflict with Dr.  Litsky's  and my own views. In the case mentioned, posi-
tive results were discarded upon  investigation  and for a good reason.  Dr.
Litsky and I argue that isolated positive results should  never be discarded
without cause.
 204

-------
                                     OTHER MEANS OF COMPLIANCE/B.E. GAY
                 Alternative Means of Determining Compliance

                             Berry E. Gay, Jr.
                           Division  of Laboratories            !
                     Illinois Department of Public Health
                           Springfield, Illinois 62701

     I understand there is a lawyer here in Washington, D.C. who is attempting
   to aid bureaucrats  so they will write  their agencies'  regulations in plain
   English. With  that in  mind I'll make no attempt.to write a regulation.
     I realize that the federal government has not stipulated by regulation that
   all community water  supplies shall be chlorinated. Illinois  Environmental
   Protection Agency regulations do require that public water supplies shall be
   chlorinated.  Based upon  the number of public  water supply samples  sub-
   mitted to two  of the three Illinois Environmental Protection Agency labora-
   tories from July 1, 1975 to June  30, 1976,  I have documented from the sum-
   maries provided by these two  laboratories the following—together the two
   laboratories analyzed 60,665 potable water samples and  4,131  of them were
   positive for total coliform, in terms of percentage  that's about  7%, but after
   verification procedures the actual percentage was about 2%.  The two per
   cent verified  positive for coliforms may have been spurious results due to a
   fluke in sample collection, an error by laboratory or a miscue by water treat-
   ment plant personnel.	    -                  ,  ...   .,...:
	In the Statement of Basis and  Purpose for the Proposed National Interim
   Primary  Drinking Water  Standards  (3), on page 9, there is  the  following
   statement,  "The presence  of any coliform bacteria,  fecal or  non-fecal, in
  treated water should not be tolerated". In. the Illinois Pollution Control Board
  —Rules and  Regulations'—Chapter  6 Public Water Supplies (2) there is  a
  definition of  the word  safe—"Safe"  means that the  water contains no sub-
  stances or organisms which are or may be injurious to a person in normal
  health who ingests the water". Currently the application of check samples (4),
  whenever a positive coliform result is  obtained whether >4 coliform colonies/
  100 ml by  membrane filter (MF) or three or more tubes with gas by most
  probable number (MPN)  an immediate resampling from the same  sampling
  point is required until two consecutive daily check samples show no coliform
  colonies by MF or positive tubes  by  MPN. Thus we can assume  the goal is
  zero coliform.
    Each Environmental Protection Agency  laboratory sends out procedures
  for collecting  samples, but  at times it appears to be a wasted effort as samples
  arrive without complete sampling data,  bottles  are completely  full, and no
  one has signed the collection sheet. Before  going any further, it ;appears we
  have a weak link to discuss. The weak link is the sample collector. No where
  in the National Interim Primary Drinking Water Regulations (4) is  this per-
  son's responsibility described. We know in a few months the final criteria for
  Water Laboratory Certification will be published  and a reference to approved
  laboratories is provided  in Section  141.28 of NIPDWR (4). Those of us that
  have  reviewed the proposed Criteria and  Procedures  for Certification  of
  Water Laboratories Involved in Public Water Supply  Analyses noticed there
  is a definite procedure given for collecting water samples when the laboratory
 is assigned the responsibility—unfortunately  not all laboratory personnel col-
 lect samples. Perhaps upon rewriting of NIPDWR someone might add a
 statement such as this—"At the discretion of the State Regulatory Agency—
 water sample collectors, after a successful training  period, may  be issued a
                                                                     205

-------
MICROBIOL STANDARDS EVALUATION/CAvl HENDRICKS





        of mistakes, cutting coUs or "improving "JJ^^^J™

                  when U.S. Public Health Service wrote Drmhng Water

                                             <£=
               es

                                                              A
to tnt eaenand their
many commercial laboratories
  One omy co-workers lives
and pays in the neighborhood
           water supplier sh
cndar
                 since one S£.
the water supply. This bring
         Samplmg and Analy
              «The  supplier
             density samples
                             main function is  to be of service. We have
                             approved for water analysis and we feel rtn
 tt-st
 indfcatine presence of coliforik whether within or exceeding the MCLs.
   One olmy co-workers lives in a small community of about 650 popdation
                            of $20.00 per month for water He feels his
                            uld provide safe water  (no conform bacteria
          y waer suppe
           and further believds one  sample per month is  not  adequate  I
                                         th          **»
                           .mple per month
                             me to  Section 141.21
                            ical Requirements  I have a definite dislike the
                                                                  Con-
     ina  Samplmg an   nay ca   e
  way the words monitoring and frequency are used m this section.
    Sd£Lm b-«The supplier of water for a community water
                            of water  for a community water  system shall
                             at regular time intervals and in number pro-
                  sy sa
                  population Served. In no event shall the frequency be less
             forth below". Then there is  the chart showing Population and
       um number of sampl
                       .
                           3  that are required  to  be taken  each month.
                           jr means to watch or check on, monitoring is
JYLUilxlujL aa u.v.ui-1-" ^j  <•—	'L "«"•"                  A~fi*»*A \\\r WphstFT
the action word meaning checking on. Frequently also ^defined by^Webster
means happening at short intervals, and frequency the fact orCondition of
"   " ."• ff    °.^  _T___.  .4	^iam \X7~ VIQVP cnmft laboratories capable
  Monitor as defined by Webst
  occurring frequently. Here's t
  of examining all the required
  206
                              problem. We have some laboratories capable
                            minimum number of samples in one day. For

-------
                                   OTHER
 example, in one of our metropolitan laboratories there are fourteen analysts
 approved for total coliform procedures and
 required for them to analyze is 480 per rnonth.  If their  collection; unit de-
 cided to collect all samples in one day and
 these  fourteen analyists could  easily anal1
 community of 10,600. population the Registered Sanitarian could easily col-
 lect all required twelve samples in one day
 yst in the laboratory who has the capability to  process them after delivery.
 In both cases cited, both laboratories; have
 the minimum number of samples required
 MEANS OF COMPLIANCE/B.E. GAY
the minimum number of samples
wrought them in to the laboratory,
rze them.  Likewise,  in  a smaller
 and deliver them, to a lone anal-
 the capabilities of examining all
 per month in one  day. Thus, as
 it is now written, this would be satisfactory, but it is hardly monitoring at
 short intervals.
   I suggest the minimum number of samples per month be increased to four
 (one per week until all four quadrants of the community are sampled.) If only
 one sample is taken, the community could be in trouble the rest of month
 and no one would detect anything until the  next month. Besides, how  could
 those that go for bacterial result averaging do it with one sample? Not only
 that,  it is hardly what  one would call monitoring when  frequency means
 monthly!
   As of this  date, our small and large community water suppliers ,are in a
 state  of chaos—they all have to secure tie  services  of an approved  water
 laboratory or spend the monies  to esitablis! i their, own. A "most unfortunate
 situation. Our state has approximately 1,650 public water systems—there are
 1,348 below the 5,000 population level, 119 hi the 5,000 to 10,000 level and
 183 greater  than 10,000  level.  Of  Chose 1,348 communities  below  5,000
 level, we have approximately 700 small cdrimunities in the 25-1,000 popula-
 tion level. Our non-community supplies nuriber approximately 11,000-12,000.
 A majority of these non-community supp ies are  currently  being sampled
 monthly.                            1                  -
   Since  March  14,  1977,  my office  has hid approximately 90 requests for
 laboratory certification information. Many  requests  are  from  individuals
 wanting to start a commerical laboratory str ctly for the total coliform param-
 eters. Some requests have come from county health departments,  county re-
 habilitation  centers,  factories, schools,  smell  and large communities. Three
 small community representatives  expressed their desire to establish a central
 laboratory and do samples  for other comriunities within  a 25 mile radius.
 They also felt sure  the  various  mayors involved would agree to having a
 central laboratory with each community sharing expenses. As it appears  now,.
 there  are numerous groups  with  the same  hought,  and this, in my,opinion,
 is a feasible arrangement. As an individual community, the cost would be too
 great, but together it is affordable. There is also another route to go. We can
 establish reciprocity  with our neighboring  states. We have  two of our  com-
 mercial laboratories  in Northern Illinois aialyzing water samples for  com-
 munities in Wisconsin and Iowa, so  for tie  southern area of our state we
 will enter into a reciprocity agreement  witi water laboratories approved by
 states of Missouri, Indiana and Kentucky.                   ,
   We have the technology to send men to  the moon, and we have the  tech-
 nology to have a safe water supply.
   While I decline to write a regulation, I am sure the Regulatory Engineering
 Staffs can. I do suggest small communities shouldn't be singled out any  more
than large metropolitan areas. The presence  of coliform bacteria in public
drinking  water La either location is very undesirable. A Federal Chlorination
Regulation would  be of  great value.  We'vfe had  laboratory certification for
years, but no one state has really  insisted water plant laboratories obtain cer-

                                                                     207

-------
                                                      •I	'
 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


  tification. The certification program to date as carried out by Geldreich and
  Nash is of the highest caliber. Those states  that have approval programs
  only surveyed laboratories upon request—as a voluntary program  Some,  ol
  us take pride in our programs and feel we have accomplished standardization
  of procedures so as to produce reliable data for the enforcement agencies and
  weV°done so with  the  aid of Geldreich and Nash. I take umbrage that
  spurious positive results are solely a laboratory error  and I assume ^the engi-
  neering sfaff in the water treatment plants would take  offense i^boratory
  personnel responded  with a statement "someone goofed in the plant . While
  the sample collector  may be a weak link, I am sure they too feel .resentment
   n being blamed. It is a matter of educating all concerned as to their respectrve
  responsibilities-to, provide a safe water supply  by  plant personnel,  proper
  procedures that must be used when collecting water samples, and the proper
  methodolosv that must be adhered  to by laboratory personnel.
    Ke comments from the various  regions of U.S. Environmental Protec-
  tion Agency I was surprised to note a desire to give  a n equal  ranking  of
  MF wkh  MPN  results-I suggest a  reading of HoW Probable is  the Most
  frobable Number? by Richard L.  Woodward (5)  Mr Edwin  Geldreich.*
  Handbook for Evaluating Water Bacteriological Laboratories  (1)  page^izu
    For those rewriting  141.21-why is it necessary to notify  the public based
  upon the averaging for the month? It seems to me when the first check sample
  ffpositive for Ae presence of  coliform the State Regulatory Agency should
  be notified by the laboratory performing the analyses.
    As for substituting  residual  chlorine determinations  for  a bacteno ogical
  sample, I  suggest  deletion.  Bacteriological  examinations  detect  pollution,
  ch£;  residuals do  not. State Regulatory Agencies  should be given the
  option or allowed to use their own  discretion regarding chlorine residuals,
  but  not as substitutes  for bacteriological • sampling. Perhaps  sample collector
  will determine chlorine residual at point of sampling and include information
  on collection  sheet.  Sewage treatment plants have been given the go ahead
  s'm to  use MF on  chlorinated effluents and  on proving results should use
  MPN for fecal  coliform determinations. Thus coliform bacteria can be de-
  tected in the presence of chlorine.
    We have no laboratories in our state using five 100 millihter standard por-
   tions for MPN. Perhaps the use of  the five 100 ml portions  could be deleted.
    Those of us with  active laboratory approval programs have a  shortage or
   survey forms. Mr. Geldreich's  supply is very limited. Will each region allow
   the states to come up with the supplemental procedures  survey form for SPC,
   Fecaf Coli and Fecal Strep and permit attaching it to the federal forms that
   will be distributed for use? At  times,  it is necessary to supplement total coil-
   form counts with results by Standard Plate Counts  (SPC).  SPC's are also
   necessary for quality control within the laboratory. Fecal coliform procedures
   are  also used as a supplement to  determine  quality of source waters The
   new survey forms are directed only to total coliforms. I guess Id better go
   now and see the lawyer that is teaching bureaucrats to write in plain English.
'"	!	•'••   -;"	  '       '"!	•	'••"'•••"   :i:	":iiir  '•   '	;    ;
.*;!                         :   REFERENCES

    !   Geldreich, E.E. 1975. Handbook for evaluating water bacteriological labora-
     '  tones. 2nd ed. Environmental Research Laboratory, Cincinnati,  Ohio.
    2.  Illinois Pollution Control Board, Rules  and Regulations,  Chapter 6, Public
     '  Water Supplies—November  22,   1974.
    3  Office of Water Supply.  1975. Statement of basis  and purpose  for the pro-
     '  pOSed national interim primary drinking  water  regulations. U.S. Environ-
       mental Protection Agency,  Washington,  D.C.
   208

-------
                                   OTHER MEANS OF COMPLIANCE/B.E. GAY


  4. U.S. Environmental Protection Agency.  1975. National interim primary drink-
     ing water regulations. Fed. Register 40:59566-59574.
  5. Woodward, R.L. 1957. How probable is the  most probable number? JAWWA
     49:1060-1068.
             QUESTION  AND ANSWER SESSION

   VJ. Cabelli, U.S. Environmental Protection Agency, Narragansett, Rhode
 Island.

   Is it advisable to have a certification program for  sample collectors due
 to the turnover in personnel?

   B.E.  Gay, Division of Laboratories:, Illinois Department of Public Health,
 Springfield,  Illinois.

   Yes,  it is most advisable to have a certification program  for sample col-
 lectors.
   As each public water supply treatment plant is evaluated for requirements
 by State EPA Engineers and the laboratories are  surveyed  for compliance
 with Standard Methods by Sjtate Water Microbiology  Survey Officers, some
 responsible person such as the plant engineer, chief operator or laboratory
 supervisor should be evaluated by the State authority for proper sample col-
lection procedures. Upon certification toy the State authority this person would
then become responsible for training, surveying and certifying his  own water
supply sample collectors. A report of these  surveys should  be sent to the
State EPA office,  the Laboratory Survey Officer, the  sample collector cer-
 tification authority, and the Regional USEPA office.
  This will tie in the sample collector to  the program," "as he may make or
break the monitoring quality control program in the field.
                                                                   209

-------
          •	•  t   '	liar  :t
;	I
'.if

-------
                             OTHER MEANS OF COMPLIANCE/I. MARKWOOD
                                                   i
             Alternative Means of Determining Compliance
                            Ira Mssrkwood
                     Director, Public Water Supplies !
                        Springfield, Illinois 62706   j
                                                   i
  I suppose you expect  me to talk profoundly on thje questions  that were
just proposed earlier. I don't intend to do so; I am not sjire I really understand
the questions, let alone the answers. I was supposed to come here and give
some alternate  methods to improve the bacteriological . standards or replace
them. Joe Harrison got here first. I therefore will say I agree with  Joe's pro-
posal No. 1, but there is one other thing I would like to see done;  that is
that  the sampling period  be a running period rather tian  a calendar month,
for example.
  If you go by a  calendar month,  a supply can take samples the last week
of the month and  the first week of the following mon:h and then  go  almost
two months before they have to take samples again. Oq if they should happen
to miss the end of the month by one day because their bottles came late, then
they are listed as not having  sampled that month  everi though their samples
came to the laboratory the following  day. That is one| change that I haven't
heard specifically mentioned,  which I would like to se
-------

  MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


  Montitpring is the last line of defense, it is really only there to see that what
  should have  been done before has  been done, and  that nothing has been
  overlooked. So, while it is important, I wouldn't say it is the most important
  thing that has to be done. I don't think any of the parts are more important
  than another. They all have  to be done  and done right if we are to assure
  the safety of a water  supply.
    I suppose  we are no different than other State agencies. We  don t have
  enough money and enough people to do the job the way we think it should
  be done,  but  we  do the best we can and,  because we can't make all the engi-
  neering evaluations we would like to, and send our people out to  all the in-
  vestigations we would like to, we have to rely quite  a bit on the  monitoring
  more than I  think we really  should.  But that is one of the facts  of  life and
  we have  to live with it.                                         '
    Now, getting to the monitoring itself. The ideal for a test would be some-
  thing that could be run accurately by an operator, would give a definite—
  let's say binary—result, either yes or no, nothing in between. He would then
  know just what to do. He wouldn't have  to use his judgment. We don't have,
  that. Instead  we have the coliform test. As  we all know, this is not  perfect;
  it has its problems. We spent most of two  days here trying to discuss how
  to interpret the results, so that it is obvious that there must be judgment with
  the knowledge of the system, with  the  knowledge  of  the  people who are
  ppe      that system, and of any other conditions  which may affect it.  It
  can't be done by the ordinary operator.
    One of the things we can't do, as you  all know, is shut off a supply. That
  is a last resort. We must try to keep the supply in operation and yet make
  it safe, which obviously is the reason for boil orders. However, this  in itself
  creates problems because if we know  that a system can be contaminated, then
  how can we  keep it operating? If we issue a boil order, we have problems
  with compliance. We know people don't boil water. We  had  a case where we
  notified a supply that  their water was contaminated and a boil order should
  be issued. The water superintendent said, yes, he would  see that it was done.
  We checked back a couple of hours later and it hadn't been done. We  told
  hhn to issue it immediately and he said, "Yeah, okay." We checked back an
  hour later and found that he  still hadn't issued it; he was going to wait until
  the next day.  We gave him a few choice threats, and insisted that it had to be
  done immediately. He went on television and said,  "The Illinois EPA says
  you should boil your water, but I don't think you have to and I am not going
  to boil mine." That is  almost an exact quote.
    When we take a sample we hope  that it is representative. But  really it is
  only representative of the water that  is passing by  the sampling point  at
  that instant. We  will never be able to get a representative sample of that par-
  ticular water again because  as it  goes through the  distribution  system,  it
  changes and  it mixes with other water in  the system, is dispersed,  so the
  sample we  get is a sample of one discrete portion of water. However much
  Ve might try, it  is not representative of the  whole system and never  will be.
  AVe have  to tal^e  that in consideration when we set a standard.
    We have to be very careful when we  have public notification.  I can give
  you an example  in a slightly  different area.  We have mandatory fluoridation
  in Illinois. I had a woman call me up  to ask me if  there was some way  in
  which a petition  would get her municipality released from mandatory fluori-
  dation because she didn't want it. I talked with her for a little  while  and
  found out  that she was drinking bottled water? They  were buying bottled
  water for drinking purposes—she, her husband  and the children—and they
  were only using  the public water supply  for bathing and so forth. We got a

  212
	^^^	i	

-------
                               OTHER MEANS OF COMPLIANCE/I. MARKWOOD
 sample of the brand  of bottled water that she was using. The municipality
 was holding pretty well between  .9 and 1.2 parts  of fluoride. The  bottled
 water had about 5 parts.
   When people start distrusting their water, t iey'11 go elsewhere, and almost
 invariably the other source will  be less safe tian the public water  supply. If
 we start giving public notice with  little or  no real backing for the  notice, in
 other words, if we give  it just because a  number has been exceeded (a num-
 ber which has been selected somewhat arbitrarily), then we are going to have
 two results. We will probably have first  one, land  then the second. The first
 result will be panic and will have  people going off in  all directions trying to
 avoid the use of the water at all costs,  including using the nearby pond, and
 that stream that looks so nice and clear as it runs past the sewage plant outfall.
 Then, they are going to  decide that this is just| a lot of crying  of "wolf", and
after a while, when we have a major emergency,, we will send out a notice and
nobody will pay  attention. And  that is one of the things that I am worried
about. I would hate to see that attempts to implement the Safe Drinking Water
 A -.*. •.*._..1J	t.	it__   _  1 , •            i      1         _    _         "
Act would eventually result in poorer or less s;
this country.
  Thank you.
e water for the communities in
                                                                     213

-------
                                                                                                                                      !	I.1 :,„!!,; ,,"ll	 nlffl""1!1'!1, : '  "!!'i'''i||lp;''
                                                                                                                                                                                                                                                                       t1'!"•f|'?:;'i  '•'.13i'li'"'iiiiiiii!IFilii
If  fii
                                        1	•,
                                        II	
               IK; i. liiiiiiibiii,   •!,  I"!),,,  iiiiiiiiiiiiiiii „ it ' !t i IMIK hi..  tiiifllB^^^^^^^ illllliii i, ii'	:. in
                                                                                                                    ., lip,;,'  ; ....... i'  ;.:\  , j.' ..... "tfi   ..... i ^ ;
                                                                                                                                                                                                                     ..... jliii    il'liiiiiliiiidiii' ''
                                                                                                                                                                                                                                                                                                       ik.  i   i'l ..... mll

-------
                                    OTHER MEANS OF COMPLIANCE/W.O. PIPES
               Alternative Means of Determining Compliance
                              Wesley O. Pipe's
                      Department of Biological  Science
                             Drexel  University
                      Philadelphia, Pennsylvania 19104

    Over a period of a day and a half it has become quite clear that there is a
  real problem associated with establishing microbiology standards for drinking
  water. What is  not clear is the  exact nature of the problem. It may be a
  public health problem.  It  may  be a microbiology problem. It  may  be a
  statistics problem. It may be the engineering problem of providing adequate
  treatment  facilities and distribution  systems
 whose integrity is  difficult  to
  broach. It may be the personnel problem of fiading capable people and train-
  ing them properly as operators, water samplers, and laboratory technicians. It
  may be a regulatory problem. It is probably k mixture of all of these.
    Previous, speakers  have concentrated mostlj? on the technical parts of  the
  overall problem. Thus, it is appropriate for ire to make a few comments on
  the nature of the regulatory problem.
    First, I would like to say  that I have never seen a technical problem solved
  by a regulation. Technical  problems  cannot ie legislated  (or regulated) .out
  of existence. They have to have technical solutions if they are to be solved at
  all. However, regulations can create  technical  problems. There needs to be
  some very serious  thought about separating the regulatory part of the overall
  problem  from the technical parts and about how to word the  regulation so
  that it motivates efforts to solve the technical parts of the problem.
    There  is a basic conflict between regulatory  approaches and technical ap-
 proaches.  Lawyers write  regulations and we can  be sure that whatever sub-
 jects and numbers technical people put into the regulation to begin with, the
  lawyers will have  the last chance to determine  the  wording.  Lawyers talk
  about "good draftmanship". What they mean is that in wording a  law or a
 regulation it should be made clear that, if  a person does A, B, and C, he is
 operating perfectly legally and within the law,
 he has violated the law and is subject to a fin
 other penalty they can dream up.  They want
 from C to D so high that few people will take
   In technical matters, usually A is excellent,
 but if he does A, B, C, and D
; or a prison term or whatever
 to make the penalty of going
 the risk.
B is good, C is poor, D is bad,
 E is worse, and F is awful. Somewhere within the range of C, D and E, a line
 between the acceptable and the unacceptable is passed but the exact location of
-this line is difficult to  determine precisely. As an example, if the monthly
 average concept is used  in the  microbiology  standards for drinking water,
 a specific number needs to be selected. That number may be 1 coliform per
 100 ml or 2.2 per 100, or 4 per 100. AH of these numbers are in the gray
 area  between acceptable and unacceptable  and probably none of them is
 technically defensible in an absolute, quantitative sense. If any one of those
 numbers is selected, it is quite likely that someone will come up with a num-
 ber of examples of water supplies which exceeijl the monthly average standard
 in communities  where there is no evidence ofl  waterborne infectious disease.
 On the other hand, an  average  of 16 coliforms per 100 ml is probably  far
 over the black and would be running too great a risk of outbreaks of water-
 borne disease in too many communities.
  The Primary  Drinking Water Regulations are written in terms of maxi-
 mum contaminant levels (MCL's). It appears that  there is some  attempt to
 apply the MCL concept to different types of parameters on a somewhat uni-
                                                                     215

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS


form basis, f his  may be  a source of part of the problem  Water treatment
Processes are not in particular designed to remove arsemc barium, cadmium
cSomum, lead, mercury, .nitrate, selenium, or silver although some Amount
of °ome of these substances ma^ get removed during treatment. Water treat-
ment processes are in particular designed to  remove ^^.
bacteria from water.  There is  k statement in Standard Methods (1

         . no coliform bacteria of any kind should be tolerated in a
                snene  to differentiate between  the use  of  the coliform
                 coliforms as indicators rather than to establish a drmkmg
 water standard  and it is an  exireme viewpoint. However  it does emphasize
     'wlr the concept that treied water should have no  coliforms m it.
      properly operated treatment system usually turns  out finished water
         coliforms present.  When there are no coliforms m the finished water
     cforl found in the distribution system indicate that water torn some
 other source  has been introduced. The discovery of coliforms  m either the
 SsnedTater or in the distribution system indicates that something is wrong
 and corrective action should bl taken. If coliforms  are found in .to _ finished
 water at the  treatment plant, tne treatment system  should be investigated to
 Sd  oS wSy ^£e letting through and adjustments  should be made »
 *at they don't. Ifcoliform organisms are found in the .dutnbutom system .the
 sbiirce M the water containing1 coliforms should be located and etaunated.
 On the other hand, if the positive coliform tests  are the result  o : poor  sam-
 pling or poor laboratory technique, this needs to be  discovered and corrected
 Regulations written in terms  bf monthly averages and  percent of  samples
 Slowing positive results just don't get the message ac ^     .            »
 lowing positive
OK to have some coliforms
                                                          .
                             present. The regulations should be
       o  ave so                    .
 that it is clear that positive resflts are not O.K.; they indicate a problem that
               _   -    .»    Jt   _  1 __  _ ^lv*n .3 •w.-rt-rv-V'rt'f tIT
              _   -    .»     Jt
needs to be solved and needs

  Testing finished water at a
bacteria  is not a very rewardi
                                                                       so
                             :o be solved promptly.
                             well run water'''tr'eatm'eht plant for coliform
                              Lg activity because practically  all the results
 bactera s no  a very rng, a^umj www«,~ — r~ ------- j  — %
 are 0 coliforms per  100 ml. That is why my research to determine the ire-
 cmency distributions  for colifortn organisms in water has been done using raw
 water samples. The frequent result  of 0 coliforms per 100 ml is  a reason to
 reconsider the use of monthly kverages in the regulations. That result doesn t
 mean that there are no coliforkns present; it just means that the  sample was
 too small. If a liter, or  10 litlrs,  or 100 liters, etc., of water were filtered
 sooner or later a coliform  org^ism would turn up. The 0 coliforms per 100
 ml iust indicates less than 1 coliform per 100 ml. Averaging a string of num-
 bers which  include  several zeijos which  really aren't zeros creates problems
 which the statisticians haven't solved yet.
    There are statistical problem^ in  estimating the number (or range of num-
 bers) wnicn" best represent the results of a series of coliform determinations
 I say this in full knowledge of the  work of Clarence  Velz during the  1940s
 (7)  and Harold Thomas during the 1950's  (6). Their work really was ex-
 cellent  considering  that they were sanitary  engineers doing statistics. The
 questions we are asking now are somewhat different, we are demanding more
 precise answers, and  we  have more serious and more  economically far-
 reaching decisions to base on those answers. Also the field of statistics has
 seen several new development^ in the last 15 years and some of the newer
 statistical techniques offer better answers to our questions if we had  the right
 kind of data.

  216

-------
                                    OTHER MEANS OF COMPLIANCE/W.O. PIPES


     Dr. Litsky presented a strong indictment of the proposed coliform stand-
  ards based on the lack of precision of the most probable number  (MPN).
  What he presented is  correct;  however, he didn't  give a fair and balanced
  presentation. It is true that McCrady (5) was the first to use the term, "most
  probable number", and it is true that he used Bayes' concept of inverse proba-
  bility in his derivation of the MPN equation. It is also true that the Bayesian
  approach has been discounted  by mathematicians  and  no statistician would
  use that approach now. McCrady was a bacteriologist not a statistician and
  his entire mathematical approach was crude. However, the same MPN equa-
  tion was derived more elegantly by Greenwood and Yule (4) two years  later.
  It  was also derived by R.  A. Fisher  (3)  in  his classic paper  on parameter
  estimation in 1921 and there are several other derivations of  the MPN equa-
  tion which have  been  published  since. I  think that we have to accept the
  mathematical soundness of the  MPN equation  if we accept  the  assumption
  on which it is based.
    The term, "most probable  number," is an unfortunate one because statis-
  ticians have rejected the Bayesian concept of inverse probability and because
  the term implies a precision the  number doesn't have. It should be called the
  "maximum likelihood estimator  of the mean coliform density" but, of course,
  McCrady didn't know  this when he wrote his  paper because  Fisher didn't
  invent the term "maximum likelihood estimator" until six years later. It is re-
  markable  that a  bacteriologist  anticipated Fisher's  general  result for  this
  specific case by  six years.
   The Bother  references which Dr. Litsky cited were trying to point out  that
  the precision of the MPN is not  as high as most people who use the multiple
  tube fermentation  test think it is. They \vere not saying that  the MPN value
  had no precision at all. It would,  of course, be  a good thing for the people
 who use MPN values  to  understand what its precision actually is and for
 this to.be taken into consideration in .the regulations.
   Standard Methods (1)  does give the 95 % confidence interval range for
 MPN values. This  range in most cases is a factor of 5; that is,  using the table
 in Standard Methods we can  distinguish between water with  1 coliform per
 100 ml and water  with 10 coliforms per 100 ml but we Can't  distinguish be-
 tween water with 1 coliform per  100 ml and water with 5 coliforms per  100
 ml.  The table in Standard Methods is a bit optimistic. The assumption behind
 the  MPN equation is that the numbers  of bacteria in small volume aliquots
 from a water sample will fit a Poisson distribution. The characteristic of the
 Poisson distribution is that the variance is  equal to  the mean. Anyone who
 has  had much experience doing coliform counts on water samples knows that
 the  variance is much greater  than the  mean. It even says so  in Standard
 Methods (1). That was  the main point Dr. Muenz was trying to make yes-
 terday when he was discussing the data I had sent to him.
   The  statistical  model  which Dr. Muenz developed is really rather simple
 compared with models other statisticians have suggested to me and it is very
 powerful. If we had the right data on  coliform counts in treated water and
 water in distribution systems his  model could be  used to put the sample fre-
 quency table of the regulations on  a rational basis. I suspect that his "bowl-
 ing  ball" model;  that is, most of the water with no coliforms and only an
 occasional  slug of water with some coliforms is the correct one. From that,
 I suspect that if. the sampling frequency table were  put on a rational basis,
 it would indicate a need for a much greater sampling frequency than anyone
has  suspected  so  far.
  I have talked mostly about the  regulatory  problem and the statistical prob-
lem because the other technical aspects of the overall problem have been well

                                                                      217

-------
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                 ,     .       ,      „      ,         .1    .       .         ,     ,          . ,,

                                 covered by other speakers. I would like to end by agreeing with a couple of
                                      grew       McCabe about the need for real time response to bacterial
                                 conation of a public water supply. Such  a contamination «j^ to be
                                 detected as soon as possible and needs to be corrected as soon as P"»"«.The
                                 bacteriological tests we have are not suited for a monitoring program because
                                 SteWSmnd  time. Some parameter such  as residual chlorine needs to
                                 be used ior monitoring purposes. It would be more rational to  use the micro-
                                 b o^gtaHesTng for short-term, intensive studies of the level .oi ^contamination
                                        raw water, of the effectiveness of the  treatment  system,  of the m-
                                        3 ^distribution system, and of  the validity of some alternate mom-
                                               n  I want to say is that I agree with what Archie Greenberg
                                  said thi  morn?ng about flexibility in the regulations. Colif orms are surrogates
                                  for  patiiSens or for public health effects. Standards  and regulations are
                                  urroPgates for judgments. The regulations attempt to prejudge every sihiafaon
                                  which might arise. We just aren't that smart. A regulation might coyer 90%
                                  S the^fes adequately  but there always will be exceptional cases m which
                                  following  a regulation literally  will lead to the wrong response to the situ-
                                  ation  The regulation must be written  with enough flexibility so that the
                                  SicaT peopfe who have the direct responsibility  for water supply can find
                                  the  right response and solve the problem.

                                                            REFERENCES
                                                                              1!,1- ,'.?; u ••••
                                   3  ££;TA.^
                                      Philos. Trans. R. Soc. London (Series A) 222:309-368.
                                   4. Greenwood, M., Jr., and G.U. Yule. 1917. On  the statistical interpretation of
                                      some ^ bacteriological methods employed in water analysis!. Hyg./6. 36O4.
                                   5> McCrady, M.H., 1915. The numerical interpretation of fermentation tube  re-
                                    ' suits. J.  Infect. Dis.  17:183-212.
                                   6. Thomas, H.A.,  1955. Statistical  analysis of coliform  data. Sewage and Ind.
                                   7.  Velz, C.J., 1951. Graphical approach to statistics. IV. Evaluation of bacterial
                                      density. Water and Sewage Works 95:66-73.
l|l|| I	Ill

-------
                                   OTHER MEANS OF COMPLIANCE/G.W. FUHS


        Alternate Means of Measuring Compliance with Bacteriologic
                     Standards of Public Water Supplies

                             G. Wolfgang Fufas
                        Environmental  Health Center
                    Division of Laboratories and Research
                   New York State Department of Health
                         Albany, New York 12201

   • The question of what constitutes an adequate measurement of compliance
  with bacteriological  standards in  water supplies  should be  extended to a
  consideration of the standard itself and of its  utility for the protection of
  public health. For the  purpose of this conference, therefore, we should at-
  tempt to develop  a logical  connection  between the absolute , measurement of
  performance by bacteriological testing and the meaning of that performance in
  terms of public health.                        .         .
    In an earlier  paper (1)  I attempted to define mathematically the relation-
  ship  between bacteriological standards and  public  health  risk for bathing
  beaches. A distinction was  made  between two situations, governed by what
  might be called the laws of, large and small numbers. These lines of reason-
  ing  are easily applied to considerations  of water  system safety.
    The terms large numbers and small numbers refer to the size of the human
  population in a watershed. For sources  of  potable water polluted  by large
:  populations  (e.g., the Great Lakes  and the Mississippi and  lower Hudson
  rivers), the probability of ingesting an infectious dose can be calculated from
  the probability of ingesting a certain number of viable infectious particles and
«the probability,  of infection .from. :suc.h_ a particle.. This calculation must in-
  clude proportonality  factors, such  as . (-a) -the abundance  of excretersjof Jj£__
  fectious particles in the human population  a"nd j5T~tBe~ratib of infectious
                                            bacteriaTA large polluting popu-
                               ^
 lation  virtually"" assures the presence of the infectious agent, but in numbers
 that are. reasonably stable and predictable and generally quite low.
   When water polluted by a large population serves as a source of potable
 water, the public health risk  associated with its use depends to a great extent
 upon the abundance  of excreters in the watershed population, which deter-
 mines  the ratio of infectious .particles to indicator bacteria. The density of
 the infectious  agents can be reduced, often  by several or many orders of
 magnitude,  by precipitation, filtration andean initial disinfection step.  For
 infectious agents which can be  effectively removed by  this treatment, the
 product will often be  quite safe for drinking even without final disinfection —
 not because the pathogens  are  absent, but  because  their numbers  are so
 small as to virtually preclude ingestion of an infectious dose by any one con-
 sumer.  Final disinfection can then be considered mainly protection against
 sources of contamination affecting the treated water.
   The law  of  small numbers applies  to sources of drinking water affected
 by small  populations  and to drinking water  distribution systems. Whenever
 the polluting population is small, the ratio of  pathogens to indicator bacteria
 is subject to wide fluctuations  depending  on the presence  or absence of
 excreters in  the watershed. The  safety of the  supply with regard  to a speci-
 fied disease  is  assured  when there are no excreters in the population, but
 the water may be highly unsafe if and when only a few or even one excreter
 is present. In the extreme case the probability . of one excreter being present
 at any time  is virtually identical with the probablity of an outbreak. Assum-

                                                                       219

-------
Figure 1. AREA COVERED IN SURVEY OF WATER SUPPLIES IN NEW YORK
       STATE. COUNTY LINES ARE SHOWN.
                                                                         SCAIE IN MIIES

                                                                         10 0  10 20 30 40

-------
                                    OTHER MEANS OF COMPLIANCE/G.W. FUHS

  ing that  excreters are randomly  distributed  and as mobile as the  general
  population, this  probability will obey  a Poisson distribution with  time, with
  the ratio  of  excreters to  general population as the  mean.  Therefore the
  public health risk in this case also is  predicated upon the prevalence of in-
  fection in the general population.
    This principle explains three phenomena: (a) short episodic outbreaks of
  waterborne disease related to occasional pathogenic contamination of a water
  supply  whose treatment  system fails intermittently  or contiuously,  (b) the
  absence of waterborne disease in certain small, isolated communities,  despite
  almost  constant  contamination of their water  systems with fecal bacteria,
  and (c) the occasional water system which produces illness whenever a vital
  disinfection step  fails, indicating access to the system by a resident excreter.
    In dealing with situations governed  by this law of small numbers^ and for
  tbejpurpose of protecting (rather than predicting) public safety, the^nuniber
  o'F^^ters_ for any given ^d7s^a?en^rH^>o^lat|on^houId be abitrarily set at
  the  theoreticaj^gxpectea" number plus one. This additional excreter repre-
  sents thF1B^ividlIal~wE6~may  haveTccess to the system at some time in the
  immediate future and thereby create  an  acute  hazard. This approach,  in
  effect, leads to a  policy of zero permissible contamination for all potable and
  recreational waters affected by small populations and, in particular,  for all
  public water  supplies, since  their distribution  systems  are susceptible  to
  localized  access  or failure.
    The strategy for a zero-contamination policy consists in setting up one or
  several  protective barriers:  watershed  protection with  fences  and patrols;
  impermeability of sources, reservoir structures, and distribution systems; and
  disinfection. Each barrier is intended ,to be absolute but  is in fact  not so,
  due to the imperfection of all human endeavor. The effect of any small num-
  ber of failures of any barrier, can be reduced, often dramatically, by a second
 protective barrier, as  long as the failures of .^either barrier-are randomly dis—
 tributed with time and their causes are independent of one another.  In this
 case, the over-all  probability of failure of the barrier system is the product of
 the probabilities of failure of the individual barriers (together with the proba-
 bility that some person having access to the system is excreting pathogens, if
 fecal bacteria are the indicators of system failure).
   Since these probabilities combine in a multiplicative fashion, we should not
 be surprised to find a lognormal distribution of the  number of incidents of
 contamination with fecal bacteria per unit of time. This indeed is the case.
   Bacteriologic records from 726  public water supplies in 42 mostly rural
 counties of New York State_(Fig. 1).were..examined for reports of conform
 bacteria.  (This is the geographical area served by the sanitary bacteriology
 laboratories of the New York State Department of  Health.)  Detection of
 any coliform bacteria in one or more samples in any calendar month  during
 which samples were received was counted as one  occurrence. This was done
 in  an attempt to  assign equal weight to failures of the source, of the treat-
 ment, and  of. the distribution system, .although the use of the calendar  month
 is rather arbitrary. Most records covered 3 years  (July  1972-August 1975),
 with a minimum  of 12 months. Sampling frequency  was based on popula-
 tion according  to the 1962 Public Health Service table, but with a  minimum
 frequency  of 12 rather than 24 per year.
  The cumulative frequency distribution of the average number of  occur-
rences per year, plotted on log-probability paper,  shows a straight line over
most of the range (Fig. 2). This reflects the multiplicative behavior of com-
ponent probabilities, as explained above. TLie different slope ofjhe_gurve in
the range of supplies  which fail most  fri3quentiy~may  indicate a  different

                                                                      221

-------
                                 MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
                                              re^singie  jLuu-im aam.k'i^j LWI.WV..         ..
                                              wTthatT7% of the supplies wefe free of coliforms for 3 years;
                                        an value was 1 event every 13 months, or 0.9 -ents per year How-
                                 ever  10% of the systems .showed cohforms. every 3  months or  more ire
                                 quStly"and 1% were more or less continuously contaminated Typical ex-
                                 ampks of poorly performing systems are <«)  a supply with a defectBranch
                                 inlne distribution system which continuously delivers contaminated water to
                                 an outlyng section of the community  and  (b)  defective small supplies m
                                 economically very depressed regions where improvements cannot reasonably
                                 be expected without financial assistance from the outside.
                                 .  This method  of determining- the performance  of a  water ^ »«"£
                                 more  stringent  than the  existing  regulatory procedure, which allows the
                                    99.9
                                 •0 99.8
                                 H 99.5
                                 E1   99
                                 %   98

                                 m-  95
                                 U_
                                 O   ,90
                                  loJ
                                  O
                                  tr
                                  UJ
                                  O
80
70
60
50
40
30
20

 10

 5
                                        O.i,
                                               MONTHS  BETWEEN/OCCURRENCES
                                        100  '   60
                  30   20
                                                        T
 15  12
~1—T~
 6
T
 4
~T
                                                                                         2  1.5   I
                                                                                         _L__J—L
               0.3   '0.5       I        2    3  4  5 6  8  1012
              - OCCURRENCES PER  YEAR
                                           ..  ,    ,   :	,  	.	     ,..  „,
                                      Figure 2. CUMULATIVE AVERAGE FREQUENCY blStRIBUTION OF COLIFORM
                                              OCCURRENCES IN NEW YORK STATE WATER SUPPLIES, FOR
                                              DEFINITION OF OCCURRENCE AND PERIOD COVERED, SEE TEXT.
                                              FOR AREA doVEREb, SEE Figure 1.
II1II 111 111 || tiliiililPliiLH!	llki; ,,mki. „,, IKJjL.ili.i ,b	iL 
-------
                                 OraER MEANS OF COMPLIANCE/G.W. FUHS
        0
I            2          3          4
OCCURRENCES  PER  YEAR
    Figure 3. CUMULATIVE FREQUENCY DISTRIBUTION OF COLIFORM OCCURRENCES
            IN FIVE LOCAL JURISDICTIONS IN NEW YORK STATE. THE SUMMARY
            CURVE (HEAVY LINE) APPROACHES THE CURVE OF THE LOGNORMAL
            DISTRIBUTION IN Figure 2.
averaging of coliform counts over all samples taken in any given month. It
also is suitable for characterizing the status of an entire region or jurisdiction.
Differences among five local jurisdictions in northeastern New York State are
illustrated in Figure 3. While the summary  curve approaches the lognormal
distribution, the individual curves show markedly different  degrees of per-
formance,  which are suggestive of  both  regional socioeconomic differences
and the varying intensities with which  corrective actions are sought on the
local level.
                                                                  223

-------
                                                                                                            I  i  ii 11 (nil;    I     «          Ij||        !l
                                                                                                            ;  i
to
                                       I'll	1	1	1	1
                                                                                                                                            o
                                                                                                                                           •g
                                                                                                                                            w
                                                                                                                                            s
                                                                                                                                            f
                                                                                                                                            co
                                                                                                                                            H
                                                                                                                                             I
                                                                                                                                             O
                                   j      FMAMJ      JA      SOND
                                          Figure 4 SEASONAL DISTRIBUT10N.OF COLIFORM OCCURRENCES IN THE FIVE
                                                 JURISDICTIONS. HEAVY LINE: ACTUAL OCCURRENCES. LIGHT LINES:
                                                 CONSTANT FUNDAMENTAL, AND FIRST FIVE HARMONICS OBTAINED
                                                 BY FOURIER ANALYSIS. DOTTED LINE: LONG-TERM MEAN MONTHLY
                                                 TEMPERATURES FOR ALBANY, NEW YORK. THE DISCREPANCY
                                                 BETWEEN ACTUAL EVENTS AND PREDICTIONS IN DECEMBER AND
                                                 JANUARY WAS AN ARTIFACT OF THE SAMPLING SCHEDULE.

-------
                                   GIBER MEANS OF COMPLIANCE/G.W. FUHS
    The 726 supplies in our sample produced an average of 650  occurrences
  per year, or almost 2,000 occurrences over the 3-year period covered by this
  study.  Yet there  were only  two outbreaks of  disease.  One  occurred in a
  larger supply which had  about 4 occurrences a year but was mostly in com-
  pliance by regulatory standards because these are based on average counts,
  which in this supply were generally low. The outbreak was ascribed to water-
  borne Giardia lamblia. The other outbreaw  of disease was of a particular
  freaky  nature and can  hardly  be called waterborne in  a strict  sense.
  It  was ascribed to  a rare pathogen,  Yersinia enterocolitica,  which was
  traced to a supply with an almost flawless bacteriologic record.  The supply
  itself did not carry an infectious dose  for any consumer,  but the  bacteria
  multiplied in a commercial dairy product, causing  the  outbreak. This low
  ratio  of outbreaks  to system failures shows that the  pattern of outbreaks of
  waterborne disease is governed by the: laws of probability.
    The seasonal distribution  of coliform  occurrences in  the five  local juris-
  dictions covered in the previous sample is shown in Figure 4. Fourier analysis
  (the first six harmonics are shown) indicates a seasonal pattern which closely
  follows the long-term mean monthly temperature curve for the  area. The
  abnormality  in December-January is an  artifact caused  by  the sample col-
  lection schedule.
    Another important consideration  in  formulating  regulations on bacterio-
  logic  sampling and in determining compliance is the statistical power of the
  bacteriologic test. The most powerful statement that can be derived from the
  standard tests is the proof  that coliforms (or any  other selected group  of
  bacteria or agents)  were present in the sample and therefore were present  in
 the system at the time of sampling. Whenever we deal  with small numbers
 of bacteria in a small number of samples, the quantitation of  a positive result
 is  characterized by  poor precision. But the power of the negative  test is also
 quite  poor. A negative result for a single 100-ml sample, if interpreted at the
 95%  confidence level  (which is  standard in scientific work), should read
 "fewer than three  bacteria per 100  ml"  and  not "fewer than one,"  as has
.been customary in  water  supply monitoring. The probability that the state-
 ment  "fewer than one" is correct  is only about 70%.  (The remaining per-
 centage points are  attributable to the possibilities that there  are one  or two
 bacteria per 100 ml.) Even this statement, which is based on  the Poisson dis-
 tribution, .holds only for small parcels of  water, which are likely  to show a
 random and  homogeneous distribution  of bacteria,  and  does  not hold for
 entire  distribution systems.
   Whenever the statistical power of  a. single test is  small,  we are told by
 statisticians to continue testing until a satisfactory conclusion  can be reached.
 This is exactly the purpose of the long-term evaluation shown  in Figure 2. As
 long as  a  water system does  not  undergo  a  significant change the record
 reflects  "present Conditions" except  for seasonal variations  and  random
 events. If  the sampling period spans at least  1 year  and  if at least 24
 samples  were taken, a rating of  the system by  this method is possible.  I
 tend to  prefer a series of monthly samples taken during 3  years over the
 same number of samples  taken more frequently  during  1 year because the
 3-year series may include  events such as  floods, droughts, or other unusual
 stresses on  the water supply.                     	           —
   This type of long-term  evaluation  has  the advantage that  a single occur-
 rence of bacterial contamination is  placed into proper perspective.  Individual
 occurrences of bacterial contamination may be ascribed  to unusual circum-
 stanes  and may be excusable, but a high rate of recurrence of such incidents
 is an indication of  an unreliable source, ailing equipment, sloppy  operation,

                                                                      225

-------
         MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS!


         or any combination of these anjd other factors. The single incident also rarely
         produces an outbreak of disease, but the aggregate of all incidents is a direct
         measurement of the tightness of the system and therefore of the public health
         risk associated with it,        !     -                                     .,
           In this  context I feel Strongly that no positive bacteriologic results should
         be discarded without cause. In  our experience,  isolated positive  results tend
         to become confirmed in the long run, although  not necessarily in the check
         samples immediately following*  which may not represent the same parcel of
         water. Also, in our experienced false-positive results by laboratories are less
         frequent than  false-negative  results, but  this may in part be  due, not to
         laboratory error, but to the low statistical power of the negative  result, as
         explained above.             j                                      .""  . .
           Long-term evaluation of the type indicated can be used to set priorities
         for a detailed  investigation of'systems that  are performing poorly.  Such an
         investigation would include sanitary surveys with intensive sampling, using a
         more complete spectrum  of  bacteriologic and chemical  techniques.  These
         should include nonstandard but nevertheless  firmly established microbiologic
         procedures for the identification  of bacterial  types  indicative of  specific
         sources of contamination.  (The summary  methods of coliform determination
         and standard plate counts are|  appropriate for  determining the  presence or
         absence of any number of prjoblems, but they do relatively  little to .reveal
         the exact nature of the problem.)
          1 The detailed investigation of poor or marginal supplies should be directed
         at finding  the most effective ways  of upgrading  them. The investigation of a
         supply with average performance  should  aim at determining  whether a sig-
         nificant problem exists and  whether performance  could  be improved at a
         reasonable cost.              j
           Long-term  evaluation and  investigation are directed at finding ways  for
         long-term  improvement of water supply operations and long-term improve-
         ment of the prevalence of infections in  the general population, which, as we
         have seen, is essential in assuring that the fewest possible system failures re-
         sult in disease. Nothing should i discourage  regulatory peronnel from following
         u£  positive bacteriologic results immediately and taking all action  necessary
         tp protect the public from acute  health hazards. But an  equally important
         long-terni goal is a conclusivej  rating of each water supply. This cannot be
         obtained  by a highly variable,  month-to-month rating, which only  breeds
         confusion, indecisiveness, and, [if combined with a public notification scheme,
         public apathy.               j
           As the problems of many water supply systems are chronic or of a long-
         term nature, the solutions are often not  of the quick-and-easy  type. Disinfec-
         tioncreates only-one barrier  where several  are  needed. Chlorination cannot
         compensatecprnpletely for the!shortcomings of deteriorating distribution sys-
         tems, and theformation of chlorinated organic materials has become a cause
         for concern. There is need forj improved watershed protection, for  treatment
         by  advanced techniques, and particularly for the rehabilitation  of old dis-
         tribution  systems. The need  for costly, long-term solutions involving con-
         structionprojects is particularly'obvious in areas in the northeast which are
         already economically depressed.
           Public notification of a poor long-term and overall performance of a  supply
         —nas  opposed to proclamation {of an acute public' health hazard—can be an
         important  instrument to obtain  support from an informed public for long-
         term improvement of a water Supply. The prohibited-use classification of the
         U.S. Interstate Carrier  Sanitation Program was  a valuable precedent in this
         regard. Jn the absence of a construction  grant  program for water  supplies,
         226
i mi nil in  11 n ill      n in     i  in     n  i  i i n           i     n     • n 11  i  n i H  i i in

-------
                                OTOER MEANS OF COMPLIANCE/G.W. FUHS
 correct handling of the public notification issue assumes particular importance
 if activities under  the Safe Drinking Water Act are to become more than a
 rather fruitless, frustrating exercise.


                         REFERENCES
 1. Fuhs, G.W.  1975.  A  probabilistic  model of  bathing  beach  safety.  Sci.
   Total Environ. 4:165-175.
            QUESTION AND ANSWER SESSION
  V.J. Cabelli, U.S. Environmental Protection Agency, Narragansett, Rhode
Island.
  What is the appropriate  action for a one-barrier system to provide ab-
solute safety from an outbreak "of disease?
  G.W. Fuhs, Division of Laboratories & Research, New  York Department
of Health, Albany, New York.
  Make sure that the barrier works at all times. If the barrier is a chlorina-
tor, it must be watched 24 hours a day, 7 days a week.
                                                                   227

-------
                                                                                                                                                                       JIB  it
                                                                                                                                                                      • iiK ijii
                                                                                                                                        :i:	
                                                                                                                                                                       'JIB!',*
                                                                                                     Hull III"     Hi":'!11'1
'"(;"Hi!;"'n,,	lib'illlllS^^^'Ii'iisiiiiiiit'i'i'liii1,' t'   :'i'    "''Jill'.!'!!;::'!' .:.>'• 'iillili;i!"'i'ii
.v 'tSS**,.&.i. ..... iS ....... 6, "Li •:.; .; ..... ., 1 ^'M, 'Slat ...... li ........... Uv I ! .:*:  , I li ..... si A£, ...... nMO . » at ..... : "i '
                                                                                                                                                   ........... !!lliiili:ii ......

-------
                       JLIST OF ATTENDEES

        EVALUATION OF THE MICROBIOLOGY  STANDARDS
                        FOR DRINKING WATER

                 OFFICE OF  DRINKING WATER
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
                        WASHINGTON, D.C.
                           APRIL 13-14, 1977
 Martin J. Allen
 Water Supply Research Laboratory
 U.S. Environmental Protection Agency
 26 W. St. Clair Street
 Cincinnati, Ohio  45268

 Ray E. Anderson, Jr.
 Head, Public Water Supply. Section
 201 W. Preston Street
 Baltimore, Maryland 21203

 Robert Becker
 Indianapolis Water Company
 1220 Water Way Blvd.
 Indianapolis, Indiana 46202

 James W. Berry
 Fairfax County Water Authority
 P.O. Box 151
 Occoquan, Virginia  22125

 Robert H. Bordner
 Environmental Support Laboratory
 U.S. Environmental Protection Agency
 26 W. St. Clair Street
 Cincinnati, Ohio 45268

 Fran Brezenski
 Region II
 U.S. Environmental Protection Agency
 Edison, New Jersey  08817

 Victor J. Cabelli
 U.S. Environmental Protection Agency
West Kingston, Rhode Island  02892
Craig D. Cameron
Fairfax County Water Authority
P.O. Box 151
Occoquan, Virginia 22125
  Dean Cliver
  Food Research Institute
  University of Wisconsin
  1925 Willow Drive
  Madison, Wisconsin 53706

  Rita R. Colwell
  Department of Microbiology
  University of Maryland  '
  College Park, Maryland 20742
  Steve Cordle
  Water Supply Research Staff
  Office of Research and Development
  U.S. Environmental Protection Agency
  Washington, D.C. 20460
 Joseph A. Cotruvo
 Criteria & Standards Division
 Office of Drinking Water (WH-550)
 U.S. Environmental Protection Agency
 Washington, D.C. 20460
 Gunther F. Craun
 Health Effects Research Laboratory
 U,S. Environmental Protection Agency
 26 W. St. Clair Street
 Cincinnati, Ohio  45268

 Patricia Creamer
 U.S. Environmental Protection Agency
 Region — II
 26 Federal Plaza
 New York, New York  10007
 Rodney S. DeHar
 2562 Executive Center East
Tallahassee, Florida 32303
                                                                 229

-------
                                             in	ill ",:.,: wi" i n vf, \wn\vr
i
    MICROBIOL STANDARDS EVALUATION/C.W. HENDRlCKS
    John E. Delaney
    Lawrence Expert nent Station
    37 Shattuck Street
    Lawrence, Massachusetts  01843

    Robert J. Drye, Jr.
    Environmental Sciences Branch
    North Carolina Division of Health
      ,il Services
    P.O. Box 28047 ,
    Raleigh, North Carolina 27611
G. Wolfgang
Division of
New York
       Funs
      Laboratories
     Department
Scotland Avenue
                           & Research
                          of Health
Albany, New Ydrk 12201

Barry E. Gay, Jr.
Division of Labc ratories .  .
niinpis Departmmlof Public Health
134 N: 9th Street
Springfield, lilinbis 62701
     Edwin E. Geldreich, Jr.
     Water Supply Research Laboratory.
     U.S. Environmental Protection Agency
     26 W. St. Clair
     Cincinnati, Oio
              Street
               45268
     Ralph E. Gentrf
     Environmental Protection Agency
     College Station Road
     Athens, Georgia
                30605
     Arnold E. Greenberg
     Chief, Bibenvironmental Laboratories
     California Department of Health
     2151 Berkeley Way
     Berkeley, Calif drnia 94704

     David R.Goff
     Waterwork
     180? Park Road N.W.
     Washington, D.b.  20010
                      ;" ;;.';.; •• :,;: •• • "•:.-•	•; ••••
     Joseph Harrison
     IIS. Environmental Protection Agency
     230 Sputn Dearborn Street
     Chicago, Illinois 60604

     Paul Haney
     Black and Veatch
     P.O. Box 8405
     Kansas City,
      230
             Missouri 64114
 Department of Natural Resources
 270 Washington Street
 Atlanta,' Georgia 30334

 Charles W. Hendricks
 Criteria & Standards Division
 Office of Drinking Water (WH-550)
 U.S. Environmental Protection Agency
• Washington, D.C. 20460

 Morris L. Hennessey
 Division of General Sanitation
 201 W. Preston Street
 Baltimore, Maryland 21203

 John C. Hoff
 Water Supply. Research Laboratory/
 „	'MERL
 U.S. Environmental Protection Agency
 26 W. St. Clair Street
 Cincinnati, Ohio 45268

  Riley D. Housewright
  Safe Drinking Water Committee
  National Academy of Sciences
    (JH-308)
  2101 Constitution Avenue, N.W.
  Washington, D.C. 20418

  William J. Jamieson
  North Carolina Department of
    Human Resources
  Raleighj North Carolina 27611

  W. Blake Jeffcoat
  Division  of Public Water Supplies
  State Office Building
  Montgomery, Alabama  36130

  Bill Kesler
  Department of Natural  Resources
  P.O. Box 1368
   Jefferson City, Missouri

  Arnold M. Kuzmack
   Office of Drinking Water (WH-550)
   U.S. Environmental Protection Agency
   Washington, D.C. 20460

   Frank Lambert
   Division Consolidated Lab Services
   Box  18"77	1":'	!""	;	:	:	
   Ricmond, Virginia  23215

   Roger Lee
   (ME) PNW Regional Office
   901 Fourth & Pike Building
   Seattle,' Washington 98101

-------
   Judith Lewis
   Maryland State Health Department
   201 West Preston Street
   Baltimore,  Maryland 21203

   Warren Litsky
   Department of Environmental
     Sciences
   Marshall Hall
   University of Massachusetts
  Amherst, Massachusetts 02003

  Ira Markwood
  Director Public Water Supplies
  2000 Churchill Road
  Springfield,  Illinois  62706

  Leland J. McCabe, Jr.
  Health Effects Research Laboratory
  U.S. Environmental Protection Agency
  26 W. St. Clair Street
  Cincinnati, Ohio 45268

  Gordon A. McFeters
  Department of Microbiology
  Montana State University
  Bozemari, Montana  59715      "

  A. Thomas Merski
  EPA —Region III
  6th & Walnut Street
  Philadelphia, Pennsylvania 19106

  Cameron Moffat
 Washington Aqueduct
 5900 MacArthur Blvd., N.W..
 Washington,  D.C.  20016

 S. M. Morrison
 Department of Microbiology
 Colorado State University
 Fort Collins,  Colorado 80527

 Larry Muenz
 Field Studies & Statistics Program
 National Cancer Institute
 7910 Woodmont Avenue
 NIH, Landow Bldg. C-403      ... ,..._._
 Bethesda, Maryland 20014

Harry D. Nash, Jr.
Water Supply Research Laboratory/
   MERL
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
   R. Warren Norris, Jr.
   Water Supply .Branch, Region VI
   U.S. Environmental Protection Agency
   1201  Elm Street
   Dallas, Texas 752J70

   Otmar O. Olson
   U.S. Environments 1 Protection Agency
   Region VII
   1735  Baltimore
   Kansas City, Missouri 64108
  Henry J. Ongerth,
  Bureau of Sanitary
 3uef
Engineering
  California Department of Public
    Health
  2151 Berkeley Waj
  Berkeley, Califormfi 94704

  Wesley O. Pipes
  Department of Bio
  Drexel University
  Philadelphia, Penn yl
 Louis A. Resi
 Technical Support
 U.S. Environrrienta
 5555 Ridge Avenue
 Cincinnati, Ohio 4 268
Division, (OWS)
Protection Agency
 Pamela Rhodes
 Fairfax County Wa
 Lorton, Virginia
 Charles E. Rundgrea
 Water Supply Brancjh
 Route 4, Box 160
 Hillsborough, Nortl
 Augustus Ruser III
 Florida Division of Health &
   Rehabilitative Sejjvice
 P.O. Box 210
 Jacksonville, Florida 32201

 Michael E. Sawson
 ERGO
 185 Alewipe Brook Parkway
 Cambridge, Massachusetts 02138

 Stephen Schmidt
 Pennsylvania Department of
  Environmental Resources
 16tfi Floor, Fulton Building
Harrisburg, Pennsylvania  17120
ogical Science

 Ivania  19104
er Authority
Carolina 27278
                                                                      231

-------
MICROBIOL STANDARDS EVALUATION/C.W. HENDRICKS
      Peter T. B. Shaffer
      Carborundum Company
      P.O. Box 1054
      Niagara Falls, New York 14302

      M. C. Snead
      Johns Hopkins School of Hygiene
      615 N.Wolfe Street
      Baltimore, Maryland 21205

      Eloise W. Stewart
      Water Supply Laboratory
      47 Trinity Avenue, S.W.
      Atlanta, Georgia  30334

      Stanley B. Stolz
      Division of Natural Resources
      P.O. Box 776
      St. Croix, U.S. Virgin Islands 00820

      T. P. Subrahmanyan
      Enteric Virus Laboratory
      Laboratory Services Branch
      P.O. Box 9000
      Ontario Department of Health
      Toronto, Ontario, CANADA M5WIR5

      Clifford Summers
      Nebraska Department of Health
      301 Centenial Mall South
      Lincoln, Nebraska 68509

      Floyd Taylor
      Water  Supply Branch
      U.S. Enyironmenial Protection Agency
      Boston, Massachusetts 02203

	 ;I   Peter" Toft
      Environmental Standards Division
      Department ofNational Health &
	  >•••'   Welfare
      Ottawa, Ontario,  CANADA KlA OL2
L	i  i „ ,   *i" si  «i '   .,•„,'          ,"„ •     ., 'I,,,:;	,  ' „„., , :,' HI
      Richard S. Tobin
      Health & Welfare Canada
      Environmental Health Center
      Ottawa, Ontario CANADA KIA OL2

      Joseph E. Uridil
      Nebraska Department of Health
      301 Centenial Mall South
      Lincoln Nebraska 68509

      Kenneth W. Whaley
      Tennessee Department of
      .  Public Health
      Cordell Hall Building
      Nashville, Tennessee  37219
                                           Roger D. Wible
                                           Carborundum Company
                                           Research & Development Division
                                           Niagara Falls, New York  14302

                                           John' Wilford
                                           Division Water Resources
                                           P.O. Box 2809
                                           Trenton, New Jersey 08625

                                           John Winter
                                           EMSL — Cincinnati
                                           Environmental Protection Agency
                                           Cincinnati, Ohio  45211

                                           Michael Wojton
                                           Maryland State Health Department
                                           201 West Preston Street
                                           Baltimore, Maryland  21203

                                           Harold M. Wolf
                                           Civil Engineering Building, R.m. 202
                                           Texas A&M University
                                           College Station, Texas 77843

                                           Richard Woodhull
                                           Conn-State Health Department
                                           79 Elm1 Street
                                           Hartford, Connecticut 06115
 232

-------
                                     INDEX
 Achromobacter 14, 66
 Acid-fast bacteria  39
 Actinomyces  15
 Adsorption:
   Aeromonas  92
   algae 43, 79
   microbial interference  14
   nutrients 44
 Aeromonas 38, 39, 192,  200
 American Water Works Association 105
 Analytical considerations:
   false positive results 149, 204
   low-level contamination 151
   negative results 151
   MF vs. MPN  spurious results 80,
     168,  173,  197
   survival 105
   temperature  68
 Analytical methods:
   American Public  Health Association
     91
   EPA Manual 85	_•- ....
   Standard Methods 85, 91       -"-':,-..
   U. S. Geological  Survey 86
   virus 96
Analytical models:
   "bowling ball" 51, 53
   negative binomial 51
   Poisson 51,  217,  221, 225
   statistical significance 198, 201, 215
Arthrobacter  14, 66
Ascaris lumbricoides 120

                  B
Bacillus 14, 66
Bacteria, specific
   Achromobacter 14, 66
   Actinomyces 15
   Aeromonas 38, 39, 192, 200
   Arthrobacter 14,  66
   Bacillus 14,  66         '
   Bacter-aides 39
  Bifidobacterium 39
   Caryophanon 39
  Clostridium  14, 38, 62,  66
  Corynebacterium  14
  Enterobacter  aerogenes  91
  Escherichia coli 15, 25, 37, 42, 46,
    49, 91, 94, 97, 104,112-114,  153. ...
  Flavobacterium 14, 20, 45, 66, 67, 73
    Gallionelfa 14, 66
    Klebsiella 14, 3S, 43, 45, 66,  193
    Leptothrix  14, 66
    Micrococcus  15, 66
    Mycobacterium 4. 66
    Proteus 14, 66
    Pseudomonas  14, 15, 20, 39, 45, 66,
      67/68, 73, 91, 121, 122
    Salmonella 39, 43, 73, 91
    Salmonella typhi 43, 131
    Sarcina 15
    Serratia 14, 66
    Shigella 43, 91
    Shigella paradysenteriae  136
    Spirillum 14
    Staphylococcus 91, 119
    Streptococcus 38, 43,  193
    Vibrio  cholerae 43, 73
    Yersinia enterocolitica 30, 225
  Bacterial  distribution 53
  Bacterial  nutrients 40
  Bacterial  regrowth  18, 20, 45
  Bacteroides 39
  Bifidobacterium  39
  Bottled water  69, 71
 Candida parasilosis 4
 Caryophanon 39
 Check samples 34,  180, 206
 Chelating agents  81
 Chlorine demand 16
 Chlorine residual:
   DPD test 61
   continuous monitoring 62
 Chlorine-residual  substitution  101, 147,
      199, 208
 Chlorine-resistant organisms 4, 96,  101
 Cholera iii
 Clostridium perfringens 14, 38, 62
 Coliform bacteria 3, 43, 112
 Coliform detection  77
 Coliform-pathogen ratio 130, 219
 Coliform standard:
   implementation 159, 164,  169
   limit 191
.Coliphage 39, 97
 Community  Water  Supply Survey  17,
     57, 59  '
 Community water systems 21
 Compliance with regulations 195
                                                                           233

-------
Contaminated samples 129
Corynehacterium 14
Crustacea 105

       .";   ":      D

Data calculation 164,  207, 215
Dead-end mains 15, 16, 132,  178, 179
Diatoms 179
Disease, specific:
  amoebic dysentery 184
  giardiasis 8, 23
  infectious hepatitis 23, 31, 95, 105
  salmpnellpsis  23, 25
  shigellosis 23, 24, 33
  typhoid Hi; 23, 24, 131,  181
Disinfectants:
  chlorine, demand 110
  chlorine dioxide 111, 121, 139
  hypochlorous  acid  106-114,  120, 121
  .bzbrie, 121, 139
Disinfection:
  chlorinatibn 32, 128, 132
  interference 103
  effectiveness  104,  119
  residuals  17 1
Distribution systems 15, 67, 70, 81, 132,
     178, 179    .
Drinking fountains 69
Drinking Water Standards:
  see Legislation
Drug-resistance  factor 120
 Engineering  control practices 125
 Entamoeba histolytica 24, 115
 Enteric bacteria
 Enterobacter aerogenes 91
 Enterotoxic Escherichia  coli 23, 25
 Esdierichia coli  15, 23, 25, 37, 42, 46,
     49, 91,  94, 97, 104, 112-114, 153
 Fecal coliform bacteria 8,,37, 43
 Fecal coliform-fecal streptococci  ratio
     43
 Fecal sterols  193
 Fecal streptococci  38, 69, 193
 Filters:   '	'   "     '           .'  ""_	
   carbon 45
   dual media 138
 Flavobacterium meningosepticum 67
 Gallionella 14, 66
 Gastrointestinal illness iii, 23,' 32, 34, 96
 Giardia 9, 24, 31, 39, 115
 Giardiasis 8, 21, 23,  26
 Ground water 31
                   H
 Hepatitis A 24
 234
  Heferotrophic bacteria:
    growth 40, 41
    injury 40
 ..... , lh ........    ,, „  , I , ,
  Indicator organisms
    characteristics 37
    concept 3
    specific organisms  37
    rapid  methods 66
  Interstate Carrier Program  153, 160

 ................. '. .    K

  Klebsiella 14, 38, 43, 45,  66, 193
  Laboratory certification  88,  207,  208,
      209
    Legislation:
    California  185-187
    Canada 189-193
    EPA Interim Primary Drinking Water
      Regulations (1975)  v, xi, 3, 9, 18,
      29, 30, 59, 103, 140, 145, 146-149,
      154, 159, 164, 166, 175, 179, 183,
      205
    Public Health Service Drinking Wa-
      ter Standards (1942)  3,  154, .184
    Public Health Service Drinking Wa-
      ter Standards (1946)  3, 154
    Public Health Service Drinking Wa-
      ter Standards (1962) iii, 145, 154,
      206
    Safe prinking Water Act (1974)  iii,
      v, ix,2i, 145, 160, 167, 168, 180,
      213                         •  .
    Treasury  Standards  (1914)  3, 125,
      153
    Treasury  Standards  (1925)  3, .181,
      184
  Leptpthrix  14,  66
"'• Limuliis  lysate 39, 46,  47, 48


              '      M        ;
  Maximum  Contaminant Level  (MCL)
      46, 47, 97, 103, 146, 160-162, 164,
      168, 180,  197, 199, 206, 215
  Membrane diffusion chamber 41, 42
  Membrane filter  procedure 3,  15, 66,
      80, 146, 167, 191,  199, 200
  Microbial contamination 50
  Microbial ecology 72
  Micrococcus 15, 66
  Monitoring:
    chlorine residual 29,  57, 192, 218
    frequency  55 •           ' '    •
    quality control  54, 193
    statistical considerations 49
  Monthly averages 148, 163,  168,  173,
       176, 197,  208

-------
 Most probable number procedure 3, 15,
     66, 146, 167, 173, 191, 199, 200-
     204, 208, 217
 Mycobacterium 4, 66

                  O

 Organic halogens 139

                  P

 Pipe "C" values' 133
 Poliovirus 97
 Potable water 116
 Proteus 14, 66
 Protozoa, specific:
   Entamoeba histolytica 24, 104, 115
   Giardia  lamblia 3, 9, 21, 24, 26, 31,
     39, 115, 225
 Pseudomonas 14, 15, 20, 39, 45, 66-68,
     73, 91, 121,  122
 Public  notification 148, 159,  160, 180,
     186, 198, 206, 212, 213, 226
 Public  water supplies 226
                  R
Rapid indicator systems 39
Ratios:
   coliform-pathogen  130, 219
   fecal coliform-fecal streptococci 43
   virus-coliform 5
Safe Drinking Water Act iii, 21, 45
Salmonella 39, 43, 73, 91
Salmonella typhi 43, 131
Salmonellosis 23, 25
Sample collection  79, 205
Sampling:
  frequency 79, 149, 193, 207
  techniques 79
Sanitary survey  32
Sarcina 15
Semi-public systems 27
Serratia 14, 66
Sewage contamination 46
Sewage poisoning  23
Shigella 43, 91
Shigellosis 24, 33
Shigella paradysenteriae 136
Spirillum 14
Standard  Methods  85-87,  91-93,  97,
    -147, 171, 191, 200, 216, 217
 Standard plate count 15, 20, 42, 175,
     208
 Standards evaluation 58, 145
 Staphylococcus 91, 119
 Streptococcus 38, 43* 193
 Tap water 69
 Temperature dependence 82
 Transit time 80
 Treatment deficiencies:
   cross connections 25, 96
   slime formation 133
 Total  coliform  bacteria 3, 42, 43, 131,
     193
 Turbidity:
   control practices 139
   interferences  with  measurement  13,
     18
   research 105, 111
 Typhoid iii, 23, 24, 131

                  V

 Vibrio cholerae 43, 73
 Virus 4, 91, 101, 104, 113, 121, 138
 Virus, specific
   adenovirus 4
   bacteriophage 4
   coliphage 39, 97
   coxsackievirus 4
   echovirus 4
   hepatitus A 4, 24,  31, 95
   poliovirus 4,  97, 104, 111
   reovirus 4
 Virus-coliform  ratios  5
 Virus concentrator 1
 Virus detection  95

                  W

 Water distribution system 11, 32
 Water quality:
   characteristics 65
   data 6,  7
   goals 130
 Water softening 33
 Water treatment 44, 98, 104,  137
Waterborne  disease:
   epidemics 21, 22,  134
   etiology 21, 61, 95
   indicator organisms  21
   surveillance 27
Wells 32
•j}{- U.S. GOVERNMENT PRINTING OFFICE: 1979 O 275-621
                                                                           235

-------
f
aiiiBSSt"!!11!!!!!!!1111 :i II|;!!!I!!I!!I!!;;;!!|I!I|I,IB! 'UKI'1!!!!!;;,:"!1!;!1!:;,1  !	!,/

 i  >''"% ''''•  '  i ",  "   ',' '.''
                                       =:~i~^^^^^^^^                	RI5!!1R
                                                                                                               .     •           	1	Wmi	K	1	WF	if! "I'llMB!	rifflSBIfimffl	MttlllllMnSff	''
                                                                                                             i;i"  '   ,'""'"!";I:':   '"   "'irl,':1",'  ;  =;,;  ":•	is  :::'i:    ,'  ''if;'	:i-;    iiiii"1   ill""1: °""1,    '"'
                                                                                                                                                                                             111"']    'I! I!1:,!!
                                                                                                                                                                                      -	ItHi!'!    -ililirii',  ,    !l
                                                                                                                                                                                             iliii'ji ....... '   .'I,:1!1 ...... ......... IF:i' :«' :      IT
                                                                                                                                                                                                                                                , T!l ,?  1, i,T   I   :!j|liil!!!i! W'
                                                                                                                                                                                                                                                          !""  i   "111* 111,
                                                                                                                                                                                                                                XT
                                                                                                                                                                                                                                  !'i	(I
                                                                                                                                                                                                                                  	1

-------