United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-82-054 August 1982
Project Summary
Persistence and Detection of
Coliforms in Turbid Finished
Drinking Water
Ramon J. Seidler and Thomas M. Evans
The Safe Drinking Water Act requires
public water systems with surface water
sources to monitor turbidity, as weH as
conforms, on a routine basis. The results
of these two measurements provide an
indication of the water quality and treat-
ment efficiency of the system. In several
regions of the country, surface waters
are not filtered and precipitation carries
turbidity to the consumer's tap. No
experimental evidence exists to define
the impact of such turbidity in systems
using chtorination as the only treatment.
Models were developed to define the
quantitative interrelationships between
total organic carbon, disinfection effi-
ciency, chlorine demand, and turbidity in
surface waters entering a distribution
system. The results illustrate that tur-
bidity and its associated total organic
carbon exert specific and predictable
levels of chlorine demand. Turbidity is
also associated with a decrease in disin-
fection efficiency, that is, turbidities in
excess of 5 NTU inhibit the elimination
of conforms even when a free residual
chlorine is maintained for 1 hour. Finally,
at turbidities in excess of 2 NTU, a strik-
ing interference with coliform detection
by membrane filtration was documented.
The models will provide water treatment
operators with the necessary guidelines
to compensate for the undesirable ef-
fects of turbidity when they occur in the
source water. The relevance of the
models in different regions can be con-
firmed through measurement of water
turbidity, chlorine demands and coflform
persistence.
Other types of interference with coli-
form detection were found to occur in
the standard most probable number
(S-MPN) technique. Up to 50 percent of
the coflform-contaminated drinking water
samples can be missed by the S-MPN
technique. Interference was not linked
to turbidity but seems to be due to inade-
quacies in S-MPN media formulation. An
abbreviated MPN technique was field
tested and was found to be superior to
the S-MPN in conform detection.
This Prefect Summary was developed
by EPA'3 Municipal Environmental Re-
search Laboratory, Cincinnati, OH, to
announce key findings of the research
project that is fully documented in a sep-
arate report of the same title (see Project
Report ordering information at back).
Introduction
The quality of finished drinking water
in the United States leaves much to be
desired even though it is probably better
than that of most other industrialized
nations. In the period between 1972 to
1976, some 27,000 individuals be-
came ill from consumption of contami-
nated drinking water. In 1978 there
were an additional 11,435 cases of
waterborne disease. Some experts be-
lieve that as many as 90 percent of these
outbreaks go unreported and that many
other victims suffer, but fail to associate
their illnesses with contaminated drinking
water.
-------
Epidemiological studies have shown
that most of the waterborne disease out-
breaks occur in semipublic water systems.
These include systems serving camp-
grounds, parks, hotels, and restaurants
that have their own water system avail-
able for use by the traveling public. These
small systems also have more deficien-
cies in equipment design, maintenance,
and monitoring than the larger municipal
systems.
Many of the water supplies in the
Pacific Northwest rely on the abundant
surface waters as the raw water source.
Many of these smaller water systems
simply chlorinate the water as the only
form of treatment. Fall and winter rains
bring runoff into the surface streams
causing them to carry considerable
amounts of turbidity. Since the National
Primary Drinking Water Regulations re-
strict the average monthly turbidity to a
maximum contaminant level (MCL) of 1
nephelometric turbidity unit (1 NTU),
many of these systems are not in compli-
ance during precipitation periods. How-
ever, there is no presently available
method to assess the impact of turbidity
that enters distribution systems having
no flocculation or filtration treatment. A
study by the National Academy of
Sciences concluded that, "fundamental
information is needed on the interactions
between viable and nonviable compo-
nents of particles in drinking water and
particularly on their resistance to disin-
fection and to other water treatment
processes."
Under Section 141.13 of the Act,
public water suppliers may request a tur-
bidity MCL relaxation to 5 NTU monthly
average. To qualify for such a relaxation,
the supplier must show that the turbidity
does not interfere with disinfection, bac-
teriological measurements, or with the
maintenance of a satisfactory disinfec-
tant in the distribution system. There are
no guidelines available on which to base
such practical decisions, and there is no
scientific information to allow an assess-
ment of how turbidity will quantitatively
affect each of these parameters. The
goal of the present study was to develop
a quantitative assessment of the physical,
chemical and bacteriological parameters
associated with turbid surface waters as
they relate to the regulations in Section
141.13 of the Act. The results of this
cooperative agreement illustrate that
turbidity has a definable effect on chlorine
demand and also exerts an interference
with coliform detection using the MF
technique. The relationship can be ex-
pressed by mathematical models which,
in the opinion of the investigators, clearly
justify the MCL for turbidity in the Safe
Drinking Water Act. The results ex-
pressed in the models can be used as
guidelines for judging the impact of tur-
bidity on relevant drinking water quality
parameters.
Results and Discussion
Statistical Models Explaining
the Impact of Turbidity
Models were developed through mul-
tiple linear regression computer analyses
in order to define the relationships be-
tween turbidity, chlorine demand, total
organic carbon, and disinfection effi-
ciency (Iog10 coliform decrease). The
models were based on data collected in
six watersheds over a two-year period
(Table 1).
In developing the models, the data ob-
tained from each of the watersheds were
not found to be significantly different.
Chlorine demand (CLDMD), which is
most important in predicting disinfection
efficiency, was found to be a function of
the turbidity and the associated total
organic carbon (TOO content of the
water (Model 2). The variables, turbidity
and TOC, explained 95 percent of the
variation in CLDMD. Turbidity was also
found to be an accurate predictor of TOC
levels in the raw water (Model 1). Disin-
fection efficiency Oog10 of the coliform
decrease) was found to be influenced by
the season, water turbidity, chlorine de-
mand, and the initial coliform density in
the raw source water (Model 3). The
coefficient of multiple determination,
expressed as a percentage, indicated
that 66 percent of the variation in disin-
fection efficiency was explained by the
variables in the model. The inclusion of a
numerical term that describes the impact
of seasonal TOC concentrations was
necessary because of the rainfall pat-
terns in the Pacific Northwest. This may
not be necessary in other geographic
regions. The seasonal correction factor
does not have a strong numerical impact
on the overall relationships. Unaccounted
for variation in Model 3 may be due to
such unmeasured parameters as sea-
sonal variations in coliform populations,
varying sensitivities of coliforms to
chlorine, and coliform masking in turbid
samples.
Impact of Turbidity on
Microbiological Determinations
by Membrane Filtration
Turbidity was also found to be asso-
ciated with failures of the standard mem-
brane filtration (MF) technique to detect
coliforms. The MF coliform detection
failures were assessed by placing filters
without typical colonies (often without
any visible colonies) into tubes contain-
ing lauryl tryptose broth (LTB) and pro-
cessing in a manner similar to the S-MPN
method. The incidence of false negative
MF results (failures) was a strong func-
tion of the turbidity level of the water.
Thus, at < 2 NTU, only 5/36 samples
were false negative. However, at 5 NTU,
15/36 samples were MF negative, but
actually contained completed coliforms
when the filter was placed into tubes of
LTB. At turbidities in excess of 10 NTU,
over 80 percent of the filters apparently
free of typical colonies were found to be
coliform positive by this method.
Inadequacies of the Standard
MPN (S-MPN) Technique
In addition to the influence of turbidity
on the MF technique, interference with
coliform detection was documented in
the S-MPN technique. This interference
was documented by using a modified
MPN (M-MPN) technique. In the M-MPN,
gas negative presumptive and confirma-
tory tubes were processed to m-endo
agar LES and examined for coliform
colonies. In addition, the completed step
was expanded to include two secondary
broth media. Interference (or false nega-
tive results) could occur at all stages in
the S-MPN technique. The M-MPN de-
tected completed coliforms in 41 drink-
ing water samples, while only 22 of
these samples were coliform positive by
the S-MPN technique. Coliform interfer-
ence in the S-MPN was found to affect
compliance with the Safe Drinking
Water Act, especially in marginal water
supply systems.
The Abbreviated MPN (A-MPN)
Technique
Analysis of data from field trials using
the S- and M-MPN provided possible al-
ternatives for improving coliform recov-
ery with less time and expense than with
the M-MPN. The procedures developed
for the A-MPN utilize LTB as the pre-
sumptive medium, m-endo agar LES as
the confirmatory medium, and LTB as *
the secondary completion broth. All \
highly turbid but gas negative presump-
tive tubes were also streaked onto
-------
Table 1. Models Derived to Predict Impact of Turbidity on Drinking Water Quality.
1) TOC =
2) CLDMDa =
3) LFDC* =
Model
1.070
+0.153(NTUf
-0.075
+0.029f NT U)
+0.405 (TOC J
1.6951
+0.0549 (Season/
-0.1676(NTUJ
+0.7763 (CLDMD)
+O.OOO3 (TCP
Coefficient of
Multiple
Determination
(R*J
b
0.936
0.663
Number
of
Observations
IN)
.23
17
32
Mean Squared
Error
(MSE)
_
0.0186
0. 1965
Total
Squared
Error
(C)
—
3.0
4.5
Standard Error
of
Regression
Coefficients
—
0.0989
0.0413
0.0780
0.2448
0.0257
0.0325
0.3829
0.0011
t
Values
—
-0.77
2.02
5.09
6.92
2.13
-5.15
2.03
2.76
"Total organic carbon in mg/l.
"Multiple regression terms do not apply to linear regression involving one independent variable.
cNephelometric turbidity units.
^Chlorine demand in mg/l.
* Log-fold decrease in coliforms.
*A numerical term (1-12, where [1] and December [12]) used to explain seasonal effects on LFDC.
g Verified total coliforms measured by MF technique.
m-endo agar LES. Typical or atypical pre-
sumptive coliform colonies were inocu-
lated into the secondary LIB. Based on
an analysis of 155 drinking water
samples, the geometric mean number of
coliforms/100 ml was 1.6 for the S-MPN
and 5.7 for the A-MPN. Statistical anal-
yses confirmed that the A-MPN was
superior to both the S-MPN and MF tech-
niques for coliform recovery from drinking
water.
Other Reports Based
on This Research
Additional published material based
on research conducted under this coop-
erative agreement includes:
Evans, T.M., M.W. LeChevallier, C.E.
Waarvick, and R.J. Seidler. 1981.
Coliform species recovered from un-
treated surface drinking water and
drinking water by the membrane filter,
standard, and modified most probable-
number techniques. Appl. Environ.
Microbiol. 41:657-663.
Evans, T.M., R.J. Seidler, and M.W.
LeChevallier. 1981. Impact of verifi-
cation media and resuscitation on ac-
curacy of the memberane filter total
coliform enumeration technique.
Appl. Environ. Microbiol. 41:1144-
1151.
Evans, T.M., C. Waarvick, and R.J.
Seidler. 1980. Occurrence of false
negative results in the most-probable-
number technique used for total coli-
form detection in surface and drinking
water supplies. Bact. Proc. p. 206.
Evans, T.M., C.E. Waarvick, R.J. Seidler,
and MW. LeChevallier. 1981. Failure
of the most-probable-number tech-
nique to detect coliforms in drinking
water and raw water supplies. Appl.
Environ. Microbiol. 41:130-138.
LeChevallier, M.W., T.M. Evans, and
R.J. Seidler. 1980. Effect of turbidity
on disinfection efficiency and bacterial
resistance in finished drinking water.
Bact. Proc. p. 200.
LeChevallier, M.W., T.M. Evans, and
R.J. Seidler. 1981. Effect of turbidity
on chlorination efficiency and bacterial
persistence in drinking water. Appl.
Environ. Microbiol. 42:1 59-1 67.
LeChevallier, M.W., R.J. Seidler, and
T.M. Evans. 1980. Enumeration, and
characterization of standard plate
count bacteria in chlorinated and raw
water supplies. Appl. Environ. Micro-
biol. 40:922-930.
Seidler, R.J., T.M. Evans, J.R. Kaufman,
C.E. Waarvick, and M.W. LeChevallier.
1980. New directions in coliform
methodology. AWWA 8th Annual
Technology Conference Proceedings
161-172.
Seidler, R.J., T.M. Evans, J.R. Kaufman,
C.E. Waarvick, and M.W. LeChevallier.
1981. Limitations of standard coli-
form enumeration techniques. J. Am.
Water Works Assoc. In Press.
The full report was submitted in fulfill-
ment of Cooperative Agreement CR-
806287 by Oregon State University,
Corvallis, OR, under the sponsorship of
the U.S. Environmental Protection
Agency.
US.aOVERNMEKTPraNTINSOFFICE.U«S-559-On/075i
-------
R. J. Seidler and T. M. Evans are with Oregon State University. Corvallis, OR
97331.
Harry D. Nash is the EPA Project Officer (see below).
The complete report, entitled "Persistence and Detection of Conforms in Turbid
Finished Drinking Water." (Order No. PB 82-227 752; Cost: $9.00. subject to
change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project 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
PS 000032JTtClIoM M>
------- |