United States
                    Environmental Protection
                    Agency
 Municipal Environmental
 Research Laboratory
 Cincinnati OH 45268
                                                                                 V f" . ./t" X
                    Research and Development
 EPA-600/S2-84-116 Sept. 1984
&ER&         Project Summary
                    Evaluation  of  Urban Runoff  and
                    Combined Sewer Overflow
                    Mutagenicity
                    Stuart J. Spiegel, Edwin C. Tifft, Jr., Cornelius B. Murphy, Jr., and Randy
                    R. Ott
                      A study was conducted to evaluate
                    combined sewer overflows and urban
                    runoff for  the  presence of chemical
                    mutagens.  The Ames Salmonella/m\-
                    crosome mutagenicity test was used as
                    a general biological effects test for the
                    qualitative detection of mutagens in the
                    sanitary  environment,  including rain,
                    urban runoff, sanitary wastewater,
                    combined  sewer overflows, sewage
                    treatment plant effluent, and receiving
                    waters. The Ames test is a relatively
                    sensitive and simple bacterial test for
                    detecting chemical mutagens. Its ad-
                    vantages over long-term animal tests
                    are speed, ease, and relatively low cost.
                    The test employs previously mutated
                    Salmonella typhimurium LT2 bacterial
                    strains that tend to mutate back to their
                    natural state when exposed to muta-
                    genic compounds.
                      Nineteen samples produced a detect-
                    able response to one or more of the five
                    S. typhimurium  test strains, with or
                    without metabolic activation. Nine of
                    these samples (47%) were of urban
                    runoff in the project area of metropoli-
                    tan Syracuse (Onondaga County), New
                    York,  and  they  produced 17  of 30
                    detectable responses (57%). Five of the
                    samples  (26%)  were  from combined
                    sewer overflows, and they produced 7
                    of 30 detectable responses (23%).
                      The results indicated that urban
                    runoff components that produce a
                    detectable response in the Ames test
                    may be diluted or inactivated in combi-
                    nation with sanitary sewage to form
                    combined sewage, since fewer responses
                    were detected in the latter than in urban
                    runoff.
  This Project Summary was developed
 by EPA's Municipal Environmental
 Research Laboratory, Cincinnati, OH.
 to announce key findings of the research
 project that is fully documented in a
 separate report of the same title (see
 Project Report ordering information at
 back).

 Introduction
  Only since the early 1960's have urban
 runoff and combined sewer overflows
 (CSO) been identified as pollutant contri-
 butors to receiving streams. And not until
 nearly 10 years later were urban runoff
 and CSO implicated as major sources of
 particulates,  nutrients, and microbial
 loading to urban waterways. In recent
 years, almost any potential pollutant to be
 found in the urban environment can be
 traced to  receiving  streams via this
 pathway.
  To decrease surface water pollution
 from these sources,  the U.S. Environ-
 mental  Protection Agency  (EPA) has
 engaged studies and provided funds for
 the design and construction of prototype
 CSO treatment facilities. These facilities
 are primarily designed to use some form
 of solids removal followed by disinfection.
  Though past studies have characterized
 urban runoff and  CSO, and identified
 constituent pollutants, they have not con-
 clusively  dealt with the  most recent
concerns about the fates and effects of
organic chemicals in these systems.
 Potentially toxic, mutagenic, carcinogenic,
or teratogenic chemicals are readily
identifiable in the urban environment, but
little attempt has been made to determine
their concentrations in urban  runoff and

-------
  Two alternative sites were also chosen  m
in case difficulties arose at any time in the
collection  of samples from the primary
sites:
 (a) Site 004 - a large basin drainage area
    characterized by mixed residential
    and commercial land use; and
 (b) Site  074  -  a  drainage area with
    predominantly  commercial  and in-
    dustrial land use.
  In addition, three CSO samples were
collected in Washington, D.C., one at each
of three  separate sites. Because  of
difficulties in obtaining  manual grab
samples  at these  sites, an automatic
sampler   was used. Also, two CSO
samples  were  taken  from  a site  in
Rochester, New  York (Site 007, Maple-
wood Avenue), during two storm events.
  The sample collection sites  were
chosen using three selection criteria:
    1. Uniqueness of drainage area char-
      acteristics. Drainage  areas with
      percent  imperviousness,  popula-
      tion density, size, and land use mix
      representative of several others in
      the city were evaluated for selec-
      tion. In addition,  drainage  areas
      that were  unique in the city (i.e.,
      predominantly open space, such
      as Site 046) were selected.         -
    2. Ease and safety of sample col lee-  fl
      tion. Where several sites of similar
      physical  characteristics could
      have been selected, ease of sample
      collection  was considered to pre-
      serve sample integrity. In addition,
      both receiving streams traverse
      busy urban  areas where sample
      collection would be  performed
      through manholes  in  streets.  To
      ensure the  safety  of personnel,
      accessibility  during high-traffic
      periods was considered.
    3. Availability of baseline data. Phy-
      sical and  chemical analyses  of
      CSO's  in the Syracuse area have
      been performed during previous
      characterization  studies. These
      data could serve as  a baseline for
      this study.  This criterion was
      considered least important of the
      three in site selection.
  CSO samples were  collected during
four wet-weather  events  in glass con-
tainers lowered  from manholes directly
above the overflow pipe. Sufficient
sample was transferred to fill an amber
glass bottle for  Ames  testing,  a  50-ml
plastic container with nitric acid preser-
vative for TOC and heavy metals analysis,
and a plastic half  gallon  container  for
other chemical/physical analyses. This
sample storage  and preservation proce-
dure was followed for all samples. Glass
CSO. Several  reasons exist for  this
omission. Such studies would require the
analysis of long  lists of compounds, as in
priority pollutant analysis, and these lists
are constantly expanding. Technological
difficulties exist in the analysis of such
compounds. The  detection limits  of
available instrumentation may not be
sufficiently low to detect minute concen-
trations of contaminants at  levels  that
may still pose a potential health hazard.
The  cost of  analyzing a single sample
under these conditions is high, and the
cost would  be  staggering to  examine
enough samples to provide  statistically
significant data for a long list of compounds.
Furthermore, such determinations would
still hot provide  any biohazard risk index;
that is, any  measure of the biological
effects these compounds might have in
their environmental medium,  particularly
effects on the human system or synergis-
tic or multiplicative effects.
  At this time,  most treatment systems
for CSO are still in the initial  planning or
evaluation phases. Though it  is generally
agreed that  treatment should consist of
solids removal and disinfection, there has
been little   evaluation of the  need for
additional levels of treatment. Because of
the highly variable nature and occurrence
of urban runoff and CSO, it  is doubtful
that the need for advanced treatment can
be determined solely by characterization
of a list of  specific chemicals. Instead,
general information  describing the po-
tential biohazard  risk of these systems
would be a  valuable criterion for deter-
mining the need for advanced treatment.
  This study has thus been based on the
premise that urban runoff may substan-
tially contribute substances  of potential
risk to the human environment and that a
preliminary  analytical survey using the
Ames Salmonella/microsome mutagen-
icity test might outline the presence and
scope  of that  potential. To this  end,
samples were taken from seven sources
of  environmental contributors  to the
Onondaga County (Syracuse), New York,
CSO  system. These sources  included:
rain, urban  runoff, dry-weather sanitary
flows,  and CSO, as well as influent and
effluent from the Metropolitan Syracuse
Treatment   Plant (Metro) and samples
from Onondaga Lake, the ultimate recep-
tor of the overflows. Samples  were tested
by the Ames  test  and subjected to
analyses for other chemical parameters,
including total  suspended solids (TSS),
total  organic carbon  (TOC),  5-day bio-
chemical oxygen demand (BOD5), oil and
grease (O&G), and  the heavy  metals
cadmium (Cd),  chromium (Cr), lead (Pb),
and copper  (Cu). The results of the Ames
test were compared with the baseline
chemical  analyses to determine any
correlations that might help identify the
source of mutagenic activity.
  Since inorganics (especially metals)
are removed  in the sample  extraction
procedure, the primary purpose of heavy
metal  analysis is to  characterize the
samples  and  evaluate whether  their
concentrations were typical or atypical of
the sample type.
  This study  was  completed in  two
phases. Phase  I involved the collection
and analysis of 74 samples during 1980.
Based  on the results of those analyses,
modifications were made to some of the
procedures. Phase II involved the collec-
tion  and analysis of  an  additional 11
samples  in 1981 from those drainage
areas that elicited detectable  responses
during Phase I.

Sample Collection
  As mentioned earlier, provisions were
made in the original study plan to collect
seven  types of samples in  Onondaga
County (Syracuse), New York (CSO, dry-
weather sanitary flow,  urban runoff,
influent and effluent to Metro, Onondaga
Lake, and rain).

CSO  Samples
  The project plan required the selection
and collection of one CSO sample during
four wet-weather events from each of six
CSO discharges. The object was to select
six drainage areas to be examined, each
with divergent characteristics of size,
land use, population density, and/or
percent imperviousness.  Selection was
made based on previous studies performed
in the Syracuse  area to  characterized
CSO sites.
  The  physical characteristics of the six
CSO sites that were selected are sum-
marized as follows:
 (a) Site 005 -  a small drainage area with
    a  high  percentage  of impervious
    surface (86%)  and predominantly
    commercial land use (however, few
    historical chemical data were avail-
    able from previous studies to charac-
    terize this area);
 (b) Site 019 - a large  drainage area with
    mixed land use;
 (c) Site 027 -  the only drainage area that
    includes  a significant  amount  of
    industrial  land use;
 (d) Site 037-  a drainage area with a high
    percentage of  impervious surface
    and mixed land use;
 (e) Site 043  -  a  large, high-density
    residential area; and
  (f) Site 046 - the only drainage area that
    is primarily open  space.

-------
containers were used for the collection
and  storage of  Ames samples, since  it
was unknown what effect plastic would
have on the samples.

Dry-Weather Sanitary Flow
Samples
  Sanitary sewage samples were collected
from the combined sewer either before or
after a storm event, as appropriate to the
collection schedule, in the same manner
as the CSO samples.

Urban Runoff Samples
  Urban runoff samples were collected
from catchbasins in their corresponding
CSO sites. Where necessary, curbside
runoff was used.

Influent and Affluent to Metro
  Six samples were  collected from the
Metro  plant. These included one set of
influent and corresponding effluent
samples during  storm flow, a similar pair
of samples during dry-weather flow, and
two influent  samples collected during
separate wet-weather flow periods.

O no n dag a Lake Samples
  Samples were collected from the north
and  south basins of Onondaga Lake 9 m
from the Lake's surface, the lower level of
the epilimnon.

Rain Samples
  Rain  samples were collected at the
O'Brien & Gere  Syracuse office in 36-in.
diameter stainless steel  pans prerinsed
with dichloromethane. These pans were
placed away from buildings, cars, trees,
and pavement to minimize contamination
from extraneous sources. Rain from two
pans was combined until a sufficient
sample was collected for analytical pur-
poses.  The pH of this sample was taken
immediately.

Mutagenicity Testing

Phase I
  Criteria used in this study for determin-
ing  a  significant level of  mutagenicity
were recommended by EPA. The signifi-
cance for each strain is determined by the
formula
        MAR=E-C
where  MAR=mutagenic activity ratio
         E    =the number of induced re-
              vertants
         C    =the number of spontane-
              ous revertants on the day
              that the E revertants were
              induced
         c   =the historical rate of spon-
              taneous reversions in the
              testing laboratory.
MAR values  were calculated  for  all
samples with all five test strains, without
and with S-9 liver activation. A negative
MAR results when E is less than C for a
given test strain on a particular day. A
zero value is obtained when E equals C.
And a positive value results when E is
greater  than C,  although this positive
value may only be a fraction if E is only
slightly greater  than C. A sample is
considered  positive  or detectable if the
MAR is 2.5 or  greater.
  Table 1 summarizes only those samples
and the particular strain(s) with a MAR of
2.5 or greater. The MAR values  presented
are for  100 /ul of sample concentrated
200 times (the highest concentration of
any  sample available in this study and
therefore the most likely sample dosage
to induce mutagenic activity), with 500//I
of S-9 reaction mixture and 100//I of the
appropriate test culture. Nineteen sam-
ples induced a detectable response in one
or more of the  test strains by this
criterion.
  Nine of these 19 positive samples (or
47%) were of  urban runoff; seven were
collected from Syracuse and  two from
Rochester,  New York.  Five of the  19
samples (or 26%) were CSO samples —
three from Onondaga County and  two
from Washington, D.C.
  Nine of the  various  samples listed in
Table 1 as inducing a positive MAR were
mutagenic in TA1538 yet not in the more
sensitive  counterpart, TA98.  Some
indications are that this is not an unusual
test  result; TA1538 may show  greater
sensitivity than TA98 with some chemi-
cals.
  No MAR of 2.5 or greater was obtained
from analyses  of the following samples:
CSO  from  Sites 4 and 46; sanitary
wastewater from  Sites 37, 43, and 46;
urban  runoff  from Sites  27 and 46;
influent to  the Metropolitan  Treatment
Plant;  and  CSO from Rochester, New
York.
  The results of all applied dosages  for
samples that produced a detectable MAR
(equal to or greater  than 2.5)  appear in
Table 2. Three concentrations  (2X, 20X,
and 200X) were  applied at a dosage of
100//I/plate at each concentration. Nine
of the 19 samples that produced detect-
able  responses (a positive MAR) did so
across the range of all three concentra-
tions; six of these  responses resulted
from metabolic activation. Four of these
nine samples produced a MAR of 2.5 or
greater for all three concentrations. The
four curves produced from the responses
of these samples are generally horizontal
with little slope,  indicating that these
samples did not show an increase in the
mutagenic response with an increase in
the amount of sample applied. Generally,
a response that increases with dosage or
concentration would be expected; but a
response such as that shown by these
samples  is not unusual. Only  sample
number 95871  exhibited a logarithmic
response curve.
  Of the 30 detectable responses (those
from more than one test strain for six of
19 samples), 10 were  from Site  004 (in
three samples), and four were from Site
037 (also in three samples). A more
interesting observation is that 17 responses
of the total of 30 (57%) were from nine
runoff samples (47% of the 19 samples
from which detectable responses were
obtained). CSO samples produced seven
responses from five samples, and dry-
weather  sanitary  samples elicited four
responses from three samples. Thus,
urban stormwater runoff may contribute
potentially mutagenic substances to CSO
and receiving waters. But since mutagen-
ic activity is not as evident  in  the
respective CSO samples as it is in the
urban  runoff, it  is probable that the
concentrations of mutagenic substances
in the  runoff  are diluted below the
detectable limits  of this study  as the
runoff mixes with sanitary waste to form
CSO.

Phase II
  For this verification phase of the study,
an additional 11 samples were collected in
Syracuse in  the late summer of 1981.
These samples included two runoff and
two CSO samples from both Sites 004
and 037, two samples  of Metro influent,
and a rain sample. One gallon of each
sample was extracted  and brought to a
volume of 20 ml with dimethyl sulfoxide
(DMSO),  resulting  in  a  concentrate of
189.25 times the original sample. Five
dosages (500, 400, 300, 200, and 100//I)
were applied.
  In two instances, the samples did elicit a
detectable response. One  sample was
CSO from Site 004, and the  other was
Metro influent. The CSO sample (OBG
No. 28524) had a  MAR of 5.2 for tester
strain TA 1538 at a dosage  of  200 /ul;
higher dosages resulted in  a  weaker
response. The Metro influent had a MAR
of 6.1 for TA1538  at the highest  applied
dosage of 500 /t/l; lower dosages  did not
result  in a MAR greater than  2.4.

-------
Table 1.    Phase I Mutagenicity Results
MAR*
TA98
Sample
Number
94492
95366
95368
95475
95476
95477
94491
95739
94493
94497
95745
95871
95873
95474
95369
94765
94766
94662
94663
City
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
4
4
Source
1
1
1
2
2
2
3
3
3
3
3
3
3
4
7
1
1
3
3
ID
Location
19
27
43
4
19
27
4
4
19
37
37
37
43
2
.
24
34
7
7
CODE
Descriptor
1
1
1
2
2
2
1
1
1
1
1
1
1
2
.
1
1
1
1
Without
S-9
-t
-
-
-
-
-
-
3.0
-
2.7
-
-
.
_
_
-
-
-
-
With
S-9
2.9
.
-
-
-
-
3.3
2.6

-
-
-
-

_
-
-
-
-
MAR MAR MAR MAR
TA100 TA1535 TA1537 TA1538
Without
S-9
2.8
-
-
-
-
.
-
-
-
_
-
6.8
.
_
_
-
76.2
-
-
With Without With Without With Without With
S-9 S-9 S-9 S-9 S-9 S-9 S-9
.
72
9.1
2.7
3.9
11.2 - - - - - 5.1
4.1 2.6 3.4 2.5
4.0 4.8
2.6 - - - -
_
29
2.9 ----- -
10.0
6.1
2.6
4.4
4.2 -
4.5
3.8
*MAR - mutagenic activity ratio.
t - No activity at 2.5 or above.
Table 2.    Detectable Responses from Phase I Analyses
Sample
Number
94492
94492
95366
95368
95475
95476
95477
95477
94491
94491
94491
94491
94491
95739
95739
95739
95739
94493
94497
95745
95871
95871
95873
95474
95369
94765
94766
94766
94662
94663
Organism
TA98
TA100
TA 1538
TA 1538
TA 1538
TA 1538
TA100
TA 1538
TA98
TA 1537
TA 1537
TA 1538
TA 1538
TA98
TA98
TA 1537
TA 1537
TA 1535
TA98
TA 1538
TAWO
TA100
TA 1538
TA 1538
TA 1538
TA 1538
TAWO
TA 1537
TA 1538
TA 1538
S-9 Reaction
Mixture fal)
500
-
500
500
500
500
500
500
500

500
-
500

500
-
500
-
-
500
-
500
500
500
500
-
-
-
-
-

200-Xt
2.9
2.8
7.2
9.1
2.7
3.9
11.2
5.1
3.2
4.1
2.6
3.4
2.5
3.0
2.6
4.0
4.8
2.6
2.7
2.9
68.
29.
10.
6.1
2.6
4.4
16.2
4.2
4.5
3.8
MAR*
20X
0.5
0.3
3.2
6.0
2.8
4.0
0.1
4.6
0.2
2.5
0.9
2.8
0.7
1.1
0.6
3.2
1.9
-0.1
0.3
-0.7
12.
1.7
2.1
5.6
1.3
0.8
-0.5
-0.5
0.6
0.6

2X

-
1.2
2.3
3.4
4.0
-
5.4
-
-0.7

0.2
-
-
-
0.8
-
-
-
-
1.7

-
6.5
-
-
-
-
-
-
* MAR = mutagenic activity ratio.
t Concentrated 200 times.

-------
  Two samples produced toxicity; runoff
from  Site 004 (OBG  No.  28885)  with
TA100 (no activation) and Metro influent
(OBG No. 28888) with TA100 (with and
without activation).
  The four aforementioned samples that
produced  positive or toxic  responses
were re-analyzed for verification. None of
the analyses resulted in a MAR of 2.5 or
greater at any applied dosage.

Conclusions
  The following conclusions resulted
from the project:
 1. Nineteen samples out of 85  total
    samples (22%) may be interpreted to
    have induced a detectable mutagenic
    response from one or more of the test
    strains,  with  or without metabolic
    activation.  Nine of these samples
    were of urban runoff. Some samples
    required a 200-fold concentration for
    detection  of  this activity.  Of  30
    detectable responses, 17 (57%) were
    found in urban runoff, and seven
    (23%) were found in CSO.
 2. The samples with a detectable
    response  did not appear to  be
    correlated  to  levels  of any of the
    chemical parameters measured or to
    any particular CSO site. The experi-
    mental  design involved too few
    samples and chemical parameters to
    obtain correlations between  detect-
    able mutagenic activity and chemical
    characteristics of samples or sample
    types (CSO, urban runoff, etc.).
 3. Fewer responses were detected in
    CSO  than in urban  runoff. Thus
    substances present in  urban runoff
    that produce a detectable response
    in the Ames test may be diluted or
    inactivated when the runoff is added
    to sanitary sewage to form combined
    sewage (CSO).
 4. This study indicates the need for a
    more comprehensive survey of the
    subject.  Definitive conclusions are
    difficult  because of problems  en-
    counterd in analytical methodology,
    sample toxicity to test bacteria,  and
    the inherent chemical  variability of
    CSO and urban runoff.
 5. The spot test (preliminary screening)
    may not  be usable as a preliminary
    step in  analyzing  polluted  aquatic
    samples  with the Ames Salmonella/
    microsomemutagenicity test because
    of sample toxicity at ambient concen-
    trations.
 6. Dichloromethane (methylene chlo-
    ride) may serve as both a disinfec-
    tant and an extractant. But if extrac-
    tion or concentration is not required
    in other aquatic environmental
    samples,  a more suitable disinfec-
    tion procedure may be necessary.
  The  full report was submitted in
fulfillment of Cooperative Agreement No.
CR-806640 by O'Brien & Gere Engineers,
Inc., and by the Department of Drainage
and Sanitation, Onondaga County, under
the sponsorship of the U.S. Environmental
Protection Agency.
  Stuart J. Spiegel. Edwin C. Tifft, Jr., and Cornelius B. Murphy, Jr.  are with
    O'Brien and Gere Engineers, Inc., Syracuse, NY 13221; and Randy R. Ottis with
    the County of Onondaga, Department of Drainage and Sanitation, North
    Syracuse, NY 13212.
  Richard Field (see below) and Robert Turkeltaub are the EPA Project Officers.
  The complete report, entitled "Evaluation of Urban Runoff and Combined Sewer
    Overflow Mutagenicity." (Order No. PB 84-211 168; Cost: $14.50, 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, Richard Field, can be contacted at:
            Storm and Combined Sewer Program
            Municipal Environmental Research Laboratory—Cincinnati
            U.S. Environmental Protection Agency
            Edison, NJ 08837
                                                                                            *USGPO:  1984-759-102-10671

-------
United States
Environmental Protection
Agency
                         Center for Environmental Research
                         Information
                         Cincinnati OH 45268
Official Business
Penalty for Private Use $300
Pb   OuO
U  6  E^y/l*
KFbluu   >
                             HkUTtCTlON
                          IU  fr 060^4

-------