vyEPA
United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-81-035 Mar 1981
Project Summary
Interrelationship of Bacterial
Counts With Other Finished
Water Quality Parameters
Within Distribution Systems
J. Kevin Reilly and Joyce S. Kippin
This study's objective was to obtain
realistic information concerning the
interrelationships among temperature,
chlorine, turbidity, coliforms, and
Standard Plate Count (SPC) densities
present in finished water after treat-
ment and distribution. Bacterial iden-
tifications were performed to deter-
mine types and densities of isolates
from the SPC and coliform tests.
The frequency of coliform isolation
was independent of free chlorine,
turbidity, and temperature. SPC's
were not contingent on low level
turbidity and varied with respect to
free chlorine residual and temperature.
SPC's exhibited no interrelationship
with coliform counts when the SPC
was less than 50 organisms/mL. A
slight inverse relationship was noted
between free chlorine residual and
turbidity. Of the physical and chemical
parameters measured, free chlorine
residual had the greatest influence on
the microbial population.
Encapsulated Klebsiella pneumo-
niae, Enterobacter agglomerans,
Enterobacter aerogenes and Entero-
bacter cloacae, which gave typical
coliform results, exhibited the ability
to survive a free chlorine residual of
0.2 mg/L or more. The diversity of
organisms identified by the SPC meth-
od strongly suggests the phenomenon
of an established microbial ecosystem
within the distribution networks.
This Project Summary was devel-
oped by EPA's Municipal Environ-
mental Research Laboratory. Cincin-
nati, Ohio to announce key findings of
the research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
Recently, increased attention has
been placed on water quality and the
monitoring procedures that ultimately
determine the quality of water delivered
to the consumer. Questions have been
raised about bacterial, chemical and
physical standards used in water quality
monitoring and their interrelationships
in the final product reaching the tap.
These standards and some of their
relationships have been studied and
documented m the laboratory and in
distribution systems. Previous studies,
however, have tended to focus on those
systems that do not provide full treat-
ment to their water supply and that have
not generally monitored seasonal fluc-
tuations of water quality parameters.
Water supplies providing full treatment
(chlorination, rapid mixing, flocculation,
sedimentation, filtration) have tended to
be ignored in previous studies, presum-
ably since they were assumed to be
supplying good and safe water to their
consumers. This study involved Salem
and Beverly (Massachusetts)—each
-------
Table 1. Effect of Free Chlorine on Co/iform Densities in Two Distribution Systems
Free chlorine (mg/L)
System
Salem
Beverly
>0.0
Coliform % samples
100 mL
<1 78
(900)*
>1 22
(257)
>5 9
(107)
<1 82
(863)
>1 18
(193)
>5 8
(60)
>0.2
% samples
80
(336)
20
(84)
6
(27)
78
(291)
22
(82)
10
(37}
>0.5
% samples
80
(176)
20
(44)
7
(16)
78
(106)
22
(30)
11
(15)
% samples
86
(80)
14
(13)
7
(5)
81
(13)
19
(3)
14
(3)
*The numbers in parentheses are the number of samples meeting the imposed test
statement criteria.
with its own distribution system but
sharing a common water source and
treatment plant.
Good quality water leaving the treat-
ment facility has long been known to
undergo deterioration within the distri-
bution system, but the extent of chemi-
cal, physical, and biological degradation
before the water reaches the consumer
has not yet been fully studied. The
objective of this research was to deter-
mine if monitoring the fundamental
parameters of temperature, chlorine
residual, turbidity, pH, coliforms, and
Standard Plate Count (SPC) adequately
characterized the microbial quality of
water as it traveled through a distribu-
tion system.
Results
During the study of the Salem and
Beverly distribution systems, the fre-
quency of coliform isolation was found
to be independent of the amount of free
chlorine present in the sample at the
time of collection. Despite the fewer
number of samples taken at each in-
creasing free chlorine level (read across,
Table 1), the frequency of coliform
isolation did not significantly decrease.
This occurred not only at the >1 coliform
level, but the maximum contaminant
level (MCL, >5) as well.
The effect of free chlorine residuals
on SPC was also analyzed. The results
demonstrate a very definite reduction of
the SPC with increased free chlorine
residuals (read across, Table 2), in
contrast to the coliform results presented
in Table 1. Increased chlorine residual
levels from >0.0 mg/L to >0.1 mg/L
effectively dropped the SPC percentage
more dramatically in Salem than in
Beverly. It may be speculated that the
older, more encrusted and slower flow-
ing (higher retention time) Salem distri-
bution system had a greater proportion
of its pipe network harboring SPC
organisms than the cleaner Beverly
system. Beverly's results deviated from
Salem's at the >1.0 mg/L percent level,
again on the lower side, probably because
of Beverly's newer distribution system
not allowing a suitable environment for
the establishment of the microorganisms.
If this is correct, then why were the
>0.1 mg/L and >0.5 mg/L results from
Beverly similar to those of Salem?
Possibly the threshold of "effective"
disinfection in the "clean" Beverly
distribution system was lower than it
would have been in the older, more
encrusted Salem system. The "effec-
tive" level for Beverly was somewhere
between 0.5 mg/L and 1.0 mg/L, where
a sharp drop in the percentage was
noted. Salem's "effective" level must
have been higher than 1.0 mg/L be-
cause no such percentage reduction
occurred between the 0.5 mg/L and 1.0
mg/L levels.
Another analysis (Table 3) revealed
the coliform frequency was approxi-
mately the same throughout the SPC
ranges of 0,3,10, and 50. The 500 level
may be statistically invalid since only 24
samples in Salem and 8 in Beverly
exceeded the 500 level. The table reveals
three conclusions. First, the SPC levels
do not affect the frequency of coliform
recovery (read across). Second, the fact
that coliforms appear with the same
frequency, regardless of SPC levels,
strongly suggests that the coliforms are
a part of the distribution system's
microbiological flora. The third conclu-
sion is that high or low SPC densities do
not indicate either the presence or
absence of coliform organisms.
The diversity of organisms isolated
from the distribution systems (Table 4)
strongly suggests these organisms have
established an ecosystem within those
pipe networks. The ability of organisms,
including coliforms, to establish and
maintain this ecosystem is not surprising
when it is realized howthe environment
and the microorganisms genetic poten-
tial combine to form an ideal habitat for
bacteria. If the distribution systems
have a diversified microbial flora con-
sisting of the organisms in Table 4 with
their varied characteristics, then that
flora cannot and should not be expected
to react in an absolute manner to one
parameter(e g., free chlorine residual or
turbidity) with any sort of consistency.
These organisms, including coliforms,
are able to survive a variety of physical,
chemical, and biological phenomena by
being encapsulated. This dense poly-
saccharide capsular coat, that is gener-
ally absent from the pure laboratory
Table 2. Effect of Free Chlorine on SPC Densities in Two Distribution Systems
Free chlorine (mg/L)
System
Salem
Beverly
SPC/mL
<3
>3
>10
>50
>500
<3
>3
>10
>50
>500
>0.0
% samples
46
54
27
8
1
55
45
18
5
0
% samples
57
43
17
4
0
59
41
15
4
0
>0.5
% samples
60
40
11
2
0
62
38
11
2
0
% samples
65
35
9
1
0
73
27
0
0 i
0 '
-------
cultures but not from environmental
strains, enables the various environ-
mental organisms to protect themselves
from the "hostile" conditions of a dis-
tribution system
Encapsulation is essential to the
success of bacteria in natural environ-
ments because the capsule coat collects
useful materials and also binds harmful
ions and molecules in the environment.
The implication is clear, the dense
polysaccharide coat has not only the
physical barrier capacity to protect itself
from free chlorine molecules and ions
but the chemical capability as well. With
the "neutralization" of free chlorine, the
water utility's primary defense mecha-
nism, the distribution system, becomes
an ideal environment for the survival
and growth of microorganisms.
Conform organisms are no different
from other groups of microorganisms
listed in Table 4 with regard to their
sustaining and replicating capability
within the distribution system. They are
able to replicate in water with trace
organics present as evidenced by the
fact that 22.4% of the microorganisms
randomly selected from the SPC during
this study were coliforms. The coliform
group makes up a rather remarkably
large percentage of the SPC population
identified in Table 4. Although taken in
the context of the large number of
coliforms isolated throughout the study
period by the membrane filtration meth-
od and the hypothesis that these organ-
isms comprise part of the ecosystem in
the distribution system, it is not a
remarkably large percentage and it is in
fact a normal phenomenon that might
be predicted. With a microbiological
flora established throughout the distri-
bution system, the generally indepen-
dent nature of the results concluded for
coliform and SPC populations, when
compared with temperature, turbidity,
and chlorine, may be understood.
Conclusions
The similarity of the ecosystems'
bacterial isolates from the two separate
and distinct distribution systems of
Salem and Beverly was probably related
to sharing the same source of water and
treatment—the Salem and Beverly
Water Supply Board's filtration plant.
Coliform bacteria were found to be a
part of the ecosystem established within
the distribution systems, and the occur-
rence of coliforms in the distribution
networks was independent of free
Table 3. Effect of SPC Densities on Coliform Densities in Two Distribution Systems
SPC/mL
System
Salem
Beverly
Coliform
WOmL
<1
>1
>5
<7
>;
>5
>0
% samples
78
(900)*
22
(257)
9
(107)
82
(863)
18
(193)
6
(60)
>3
% samples
74
(521)
26
(183)
11
(73)
81
(431)
19
(101)
10
(39)
>10
% samples
72
(242)
28
(94)
14
(42)
79
(147)
21
139)
10
(17)
>50
% samples
71
(64)
29
(26)
10
(7)
75
(45)
25
(15)
12
(4)
>500
% samples
75
(18)
25
(6)
0
(0)
75
(6)
25
(2)
25
(D
The number of samples meeting the imposed test statement criteria.
Table 4.
Organisms Identified from Salem and Beverly Distribution Systems
m-Endo agar LES
Klebsiella pneumoniae
Klebsiella rhinoscleromatis
Klebsiella ozaenae
Enterobacter cloacae
Enterobacter aerogenes
Enterobacter agglomerans
Escherichia coli
Citrobacter freundii
Serratia liquifasciens
Acinetobacter calcoaceticus
CDC Group 11K
Aeromonas hydrophila
Plate Count Agar
Klebsiella pneumoniae
Enterobacter agglomerans
Enterobacter cloacae
Enterobacter hafnia
Serratia marcescens
Proteus
Pseudomonas cepacia
Pseudomonas fluorescens
Pseudomonas maltophilia
Pseudomonas put/da
Pseudomonas vesicularis
Bacillus
Bacillus subtilis
Streptomyces
Streptococcus
Lactobacillus
Arthrobacter
A chromobacter
Achromobacter xylosoxidans
Corynebacterium
Flavobacterium
Moraxel/a
Rhizobium
Nitrococcus
Micrococcus
Acinetobacter antratum
Actinomyces
Clostridium
Vibrio alginolyticus
Aeromonas hydrophila
CDC Group 11F
CDC Group UE1
Alcaligenes
Yeast
„ US GOVERNMENT PRINTING OFFICE 1981-757-012/7052
-------
chlorine residuals and turbidity fluctua-
tions of less than 2.0 Turbidity Units.
Also, SPC of 50 colonies or less did not
reflect the presence or absence of
coliforms.
The best quality water may be pro-
duced at a treatment facility and a high
chlorine residual may be employed
throughout the delivery system, but if
the distribution system has a microbial
ecosystem throughout its network, then
regardless of that high quality chlori-
nated water, microorganisms from that
flora, including coliforms, may be iso-
lated from that system.
Analysis of all the parameters in the
above conclusions proved difficult be-
cause of the variation of physical and
chemical processes at the filtration
plant and the inherent complex nature
of the dynamic ecosystems within the
distribution networks. It is recommended
other distribution systems should be
studied with respect to the ecosystems
established within them. Also studies
should be instituted to evaluate the
ability of environmental organisms,
specifically coliforms, to withstand the
effects of free chlorine residuals. Critical
considerations of these studies would
be confirmation of encapsulated bac-
teria. The authors consider a study of pH
and its effect on the environmental
organisms, in conjunction with chlorine,
to be very critical. Finally, additional
research should be undertaken to study
the SPC enumeration procedures along
with improved media and new recovery
methods and to define the health signif-
icance and impact on coliform popula-
tions by the SPC population
The full report was submitted in ful-
fillment of Grant No. R 804724 by the
Salem and Beverly Water Supply Board
under the sponsorship of the U.S
Environmental Protection Agency.
J. Kevin Reilly and Joyce S. Kippm are with the Salem and Beverly Water Supply
Board, Beverly, MA 01915
Raymond H. Taylor is the EPA Project Officer (see below).
The complete report, entitled "Interrelationship of Bacterial Counts with Other
Finished Water Quality Parameters Within Distribution Systems," (Order No.
PB 81-168 726; Cost: $800, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at.
Municipal Environmental Research Laboratory
U S Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
Third-Class
Bulk Rate
C'i
5
------- |