United States
Environmental Protection
 Health Effects Research
 Research Triangle Park NC 27711
 Research and Development
 EPA-600/S1-84-019 Jan  1985
Project Summary
Water  Distribution  System  as  a
Potential Source  of  Mutagens  in
Drinking Water
Dipak K. Basu, Jitendra Saxena, Frederick W. Stoss, Joseph Santodonato, and
Michael W. Neal
  The primary objectives of this study
were to examine the changes in concen-
tration of six polycyclic aromatic hydro-
carbons (PAHs) and the possibility of
changes in mutagenic potential of
treated waters as a result  of leaching
during their passage through commonly
used distribution pipes in  the  United
States. With the exception of the fin-
ished water from one utility in which the
total concentration of the six PAHs was
138.5  parts per trillion (ppt), the total
PAH concentration in all other treated
water ranged from 0. to 13.4 ppt. The
corresponding total PAH concentration
in water after their passage through the
distribution pipes varied from 0 to 61.6
ppt. This demonstrates that PAH con-
centration in waters can increase  as a
result  of their passage through the
asphalt-line distribution pipes.
  M utagenic activity was also detected
in many of the treated water samples,
however, the levels of this  activity did
not correlate with either the transit of
water through the distribution system
or the levels of PAH in the water. There
was some evidence to indicate that the
water treatment process itself may have
contributed to the  mutagenicity  ob-
served in the finished water and that
compounds  responsible for the  ob-
served activity were different from the
low molecular weight chlorinated com-
pounds produced during chlorination.
  This Project Summary was developed
by EPA's  Health  Effects  Research
Laboratory, Research Triangle Park, NC,
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

   Finished  waters from the treatment
 sites are transported to the consumers
 through a  variety of  pipelines  In  the
 United States, three kinds  of pipes are
 commonly used as transmission/distri-
' bution Imesfor transporting treated water
 to  consumers, cast  iron/ductile  iron;
 asbestos/cement, and steel pipes. Other
 materials occasionally used as  water
 distribution lines are reinforced concrete,
 a variety of plastic materials, iron, copper,
 and rarely,  lead. With the  exception of
 plastic pipes, coating materials are often
 used for the interior of the water pipes to
 prevent corrosion or to aid in curing the
 pipes. For example, cast iron/ductile iron
 pipes are lined sometimes with asphalt,
 cement, or asbestos/cement coatings;
 concrete pipes with asphalt coating; and
 steel pipes with cement or coal-tar coat-
 ings. Depending upon the characteristics,
 manufacturing conditions, thickness, and
 age of the pipe or its lining, some leaching
 of materials can be expected to occur as
 water passes through the pipes. The
 leaching of volatile organic compounds
 from different types of water pipes has
 also been studied  in Europe. Cast iron
 pipes coated with polyurethane have been
 shown to add volatile contaminants such
 as  chlorobenzene  and aromatic com-
 pounds in drinking water.
  Both asphalt and coal tar are known to
 contain high levels of polycyclic aromatic
 hydrocarbons (PAHs), some of which are

known or  suspected carcinogens,  al-
though the level of PAH in coal tar may be
1000 to 100,000 times higher than in
asphalt  It is possible that the coating of
coal tar-based or asphalt-based paint or
sealant on iron, steel, or concrete pipes
could cause leach ing of small amounts of
PAH into drinking water. The concentra-
tion of any given PAH  leached into  the
water  is  dependent on its solubility,
concentration in the coating, contact time,
etc. There are reports in the literature that
undesirably high levels of PAH can result
In a German study, a 10-fold increase in
PAH concentration was reported at  the
end of water supply pipes resulting from
the leaching  of compounds from  the
coating used inside the  water pipes. In a
study in Pascagoula, Mississippi, more
than nine  PAHs, including fluoranthene
at a concentration  of  9.7  /ug/L, were
reported in water stored in a tank freshly
coated with  coal-tar  pitch. Similarly,
potable w-.'er in Portland, Oregon,  ob-
tained from a terminal  point of a water
distribution  pipe  coated with  coal  tar
showed the presence of PAH.
  Alt hough the leaching of coating mate-
rials from  the pipes can account for  the
presence of many compounds including
PAH,  it may not be the  only mechanism
which could contribute  to the deteriora-
tion of the water quality resulting from
passage of drinking waters through sup-
ply  networks. The reaction of residential
chlorine  with  organics  leached from
pipes, oxidation of leached materials, or
microbiological synthesis of compounds
on  pipe  walls and their  subsequent
release into the water may contaminate
the water  with compounds not originally
present in the finished water.
  The examination of changes in water
quality resulting from distribution pipes
can  be accomplished  by determining
changes in the levels of specific chemi-
cals Because of their reported  presence
m some coating materials used inside the
distribution  lines,  and  because  of  the
known carcinogenic potential of many of
the compounds in this class, PAHs were
chosen as a  class  of  chemicals to  be
monitored in order to determine possible
adverse effects of the distribution system
on water quality. To monitor the effects of
other harmful compounds that may be
leached from the pipes or produced in the
distribution system, an assay  was also
employed that was not specific for  a
particular class of chemicals. This general
assay was an in vitro microbiological
mutagenicity assay which has the capa-
bility of  detecting a  large variety of
compounds  with  genotoxic potential.
Therefore, the present investigation was
primarily oriented to provide information
on two aspects  of water  quality:  PAH
content and the presence of mutagens in
source waters, and the effect of treatment
and passage of those waters through the
distribution pipes on these measures of
water quality.
  The sampling locations for this  study
represent the largest water  supplies
within 24 selected counties in the United
States. These counties were selected by
the U.S. Environmental Protection Agency
for an epidemiological study of associa-
tion between cancer mortality rates and
drinking  water quality. Twelve pairs  of
counties  were selected with contrasting
(high  and low) total  mortality  rates for
colon, rectum, bladder, and liver cancer in
females during the years 1 968 to  1974.
Each pairof counties was matched on the
basis of five demographic characteristics
(geographic region, total population, per-
cent urbanization, median school  years
completed, and percent of work force in
manufacturing industry). Only counties
that were at least  50%  urban  were
included. The studies reported here, were
conducted in these counties to determine
if the occurrence of  PAH or mutagenic
activity correlated with the occurrence of
any of the types of cancer.
  The  monitoring of the expected low
levels of PAH in drinking water required a
sample collection method that  was cap-
able of concentrating  PAH from  large
volumes of water. A method developed in
a previous study was found to be suitable
and was  used; the organic contaminants
are adsorbed onto  polyurethane  foam
plugs  and subsequently  desorbed  by
elution with  an  organic  solvent. This
method also includes a final  chemical
separation procedure (thin layer chroma-
tography) capable of eliminating interfer-
ing impurities  and  a quantification
method (fluorescence spectrometry) sen-
sitive enough to detect small amounts of
PAH. The PAH detection limit for PAHs by
this method was determined to be  better
than with capillary-gas chromatography
with flame ionization detection.
  The overall mutagenic potential was
assessed by utilizing the Ames Salmonella
assay system.  This  assay  employs a
mammalian metabolic activation system,
thus allowing the detection of compounds
that are  biologically  active only  after
having been metabolized.  The system
was chosen  primarily for  two reasons.
First,  it provides a quick screening method
for testing mutagenic potential of com-
pounds.  Second,  a high correlation be-
tween positive results in the Ames assay
and  positive  results in  in  vivo animal
carcinogenicity assays have been estab-
lished for a large number of compounds.
  The scoring of samples for positive or
negative mutagenic activity was  inde-
pendently conducted by three individuals.
Objective evaluation criteria (i.e., doubling
of spontaneous  reversion rate,  and
demonstration of dose-related increase
in revertants) were employed in  the
interpretation of  all Ames assay results.
However,  in  certain assays where  the
results were not definitive (e.g.,  more
than a doubling but lack of dose-related
increase, or positive dose-response curve,
but number of revertants slightly less
than twice the spontaneous) a subjective
evaluation was also applied to the data. In
these cases,  a consensus of the  three
reviewers was obtained before declaring
the result to the positive or negative.

Results and Conclusions
  The results indicate that in 67 percent
of the total  water supplies studied,  a
measurable increase in concentration of
PAH in  water does occur as a conse-
quence of treated water travelling through
petroleum asphalt-lined distribution lines.
The  increases  in PAH  concentration,
however, were small and amounted to a
maximum of 61 ppt in  water. It was also
determined that  the maximum leaching
effect occurred with distribution  pipes
that had recently been coated with asphalt
linings.  The  recently  installed  asphalt-
coated pipes predominantly added  the
more water-soluble PAH, whereas leach-
ates from the older pipes showed higher
ratios of  less  water-soluble  to  more
water-soluble PAHs. In older pipes this is
probably due to depletion of the more
water-soluble PAH from  the surface of
the coating. Waters from a few distribu-
tion  lines showed reductions  in PAH
concentration instead of the  expected
enhancements of PAH  concentration.
This reversing trend can be explained by
assuming sorption of PAH on the surface
of the  distribution pipes or  chemical
interaction with oxidants in water. Distri-
bution pipes that are depleted of  the
leachable PAH may act  as  sorbents for
these compounds. Thus, depending  on
the age of the pipes, it is possible  to get
either desorption or sorption effects. In
addition  to  the  pre-selected six  PAHs
identified and quantified in this study,
many other compounds were detected in
the water samples. No attempt was made
to identify these  compounds.
  The results of bacterial mutagenicity
assays  indicated the presence of  low

levels of mutagenic activity in many of the
waters  sampled. In general,  the  raw,
finished, and distributed drinking waters
tested in this investigation demonstrated
no  significant or consistent mutagenic
activity in the Ames test system. Thus, it
is not clear whether the leachate  from
certain  distribution pipes actually  con-
tributes to the presence of mutagens or,
on the other hand, whether some pipes
adsorb mutagens which had not previous-
ly been removed during treatment. Of the
24 water supplies tested, two showed no
detectable mutagenic  activity  in either
the raw water or water taken from any of
the collection points. Five other water
supplies had a number of sampling sites
in  which  an  individual  assay point
demonstrated a mutation ratio (number
of revertants in experimental plate divided
by number of revertants in control plates)
of between  1.5 and  2.0  The random
appearance of these  weak mutagenic
responses and the concomitant lack of a
dose-related increase  in mutation rate
indicate that these values fall within the
normal variation of the Ames assay
  When samples produced no mutagenic
response the reason was probably that
there were  insufficient quantities  of
mutagens present. There was also the
possibility, though,  that some of these
samples contained mutagenic compounds
but the  biological activity of these com-
pounds was masked by the presence of
toxic substances in the samples. Since
toxicity tests were not performed on any
of the water samples  or their concen-
trates, the only detectable evidence of
severe toxicity in our investigation would
be the absence of a bacterial lawn (the
limited  growth  of  histidine-dependent
bacteria  allowed by the trace amount of
histidine supplied  by  the  media in the
minimal plate). The absence of a bacterial
lawn did  not result from the testing of any
samples  in this study. However, slight
bacterial toxicity was occasionally encoun-
tered  as  indicated by a dose-dependent
decrease in the number of revertants
with some samples. An example of this
toxic effect was observed with the raw
water samplefrom Fremont, Ohio, where
average number of revertants for strain
TA1535 declined from 79.5 for control to
62.0,  50.0,  56.5, and  8.4  as  the dose
increased. It is apparent from the variable
effect of metabolic activation on bacterial
toxicity (in some samples S-9 increased
activity  while  in others  it decreased
toxicity)  that  the chemical  composition
and/or amounts of the toxic agents in
individual samples varied  extensively
among the different collection locations.
In most samples, the lowest dose tested
appeared to be nontoxic to the bacteria.
  The types of pipe associated with the
two water samples that demonstrated no
mutagenic activity were cast  iron and
asbestos/cement.  Although  many  vari-
ables are unknown about the pipes  such
as age, condition, and the physical prop-
erties (pH, types of solutes) of the water, it
is apparent that  this study shows no
consistent relationship between the type
of pipe and  the  absence of  observed
mutagenic activity in the water which has
been transported by these pipes.
  Fifteen of the water supplies provided
at least one sampling point which either
produced a mutagenic response or a weak
mutagenic response in at least one of the
five  strains of  bacteria tested. In this
study a response was considered to be
weakly  mutagenic if the reversion  ratio
fell between 1.5 and 2.0 and was accom-
panied by a positive dose-response trend.
A clear mutagenic response in this study
was  considered to occur  when the rever-
sion  ratio equaled or exceeded  2.0. It
should be noted that it is not unexpected
to observe mutagenic responses  with
environmental samples  where the  high
dose point has a  lower ratio  than the
intermediate  and low dose point. Such
data  are generally indicative of the toxicity
to the tester strain of bacteria used in the
Ames assay. Even in samples demonstra-
ting  clear  mutagenic potential in  this
study, the response was not  great,  with
most ratios being  only  slightly greater
than 2.0.
  On the basis of the  results of the
present project, the investigators recom-
mend  the  following areas for further

  1.  Samples that showed mutagenicity
     should  be fractionated  to isolate
     and identify the component(s) re-
     sponsible  for the mutagenic effect.

  2.  Since no  dramatic  PAH leaching
     was observed with  the  petroleum
     asphalt-lined pipes, further  study
     should  be directed towards the
     leaching effects from distribution
     pipes lined  with coal  tar-based

  3.  In the present investigation, the age
     of the installed pipes was found to
     be one of the most critical factors in
     the leachability  of  PAH from the
     pipe linings. In future studies, the
     age of pipes, condition of coating,
    and contact time should be given
    primary considerations in the selec-
    tion  of  sampling  sites. Recently
    installed pipes are  expected to
    provide  more  dramatic enhance-
    ment  of PAH  concentration than
    the older pipes.

4.  Factors, such as pH, residual chlo-
    rine, aggressive index of water, and
    conditioning of the pipes after ap-
    plication of coat ings (all factors that
    may effect the leaching of PAHs),
    should be morethoroughly studied.

5.  The effect of chlorination of drinking
    water on  some of  the leached
    compounds should be investigated.
    A few recent studies suggest that
    the leached compounds are con-
    verted to chlorinated or oxidized
    products in the presence of chlorine.

6.  The apparent increase in mutagenic
    activity resulting from conventional
    water treatment processes  should
    be studied in  greater detail  This
    should include an investigation of
    the production of  high  molecular
    weight compounds having biologi-
    cal activity as  a result of chlorina"-

7.  The mutagenic activity of trace
    contaminants  in treatment  chemi-
    cals  used  for  the production  of
    finished water should be investi-
    gated as a possible source of muta-
    genic activity in drinking water.

    D. K. Basu, J. Saxena, F.  W. Stoss, J. Santodonato, and M. W. Neal are with
      Syracuse Research Corporation, Syracuse, NY 13210-4080.
    Frederick P. Kopfler is the EPA Project Officer (see below).
    The complete report, entitled "Water Distribution System as a Potential Source of
      Mutagens in Drinking Water,"(Order No. PB 85-125 474; Cost: $19.00, 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:
            Health Effects Research Laboratory
            U.S. Environmental Protection Agency
            Research Triangle Park, NC 27711
  * U S GOVERNMENT PRINTING OFFICE, 1985  559-016/7875
United States
Environmental Protection
Center for Environmental Research
Cincinnati OH 45268
   PERMIT No  G-3
Official Business
Penalty for Private Use $300