Prepared
by
Pat
Fair,
OW/
OGWDW
8/
10/
05
Chlorate
Ion
in
Drinking
Water
Sources
and
Control
of
Chlorate
Ion
Chlorate
ion
(
ClO
3
­)
is
primarily
present
in
drinking
water
as
a
result
of
the
use
of
chlorine
dioxide
or
hypochlorite
solutions
for
oxidation/
disinfection
in
the
treatment
process.
It
may
also
be
present
in
the
untreated
source
water,
but
the
ClO
3
­
concentrations
contributed
to
drinking
water
by
ambient
water
are
generally
much
lower
than
those
resulting
from
the
treatment
process.

The
American
Water
Works
Association
(
AWWA)
Disinfection
Systems
Committee
tracks
disinfection
practices
in
US
community
water
systems.
Their
most
recent
comprehensive
survey
(
completed
in
1998)
estimated
that
approximately
20%
of
the
systems
serving
populations
greater
than
10,000
use
sodium
hypochlorite
(
2%
generated
it
on­
site),
8%
use
chlorine
dioxide,
and
<
1%
use
calcium
hypochlorite.
(
AWWA,
2000a)
For
systems
serving
populations
less
than
10,000,
the
survey
estimated
that
approximately
34%
use
sodium
hypochlorite,
none
use
chlorine
dioxide,
and
at
least
4.5%
use
calcium
hypochlorite.
(
AWWA,
2000b)

Chlorine
Dioxide:
The
use
of
chlorine
dioxide
can
introduce
ClO
3
­
into
the
finished
water
by
several
routes.
Drinking
water
plants
generally
use
sodium
chlorite
as
a
starting
material
in
the
production
of
chlorine
dioxide.
Chlorate
ion
may
be
present
as
a
contaminant
in
the
feedstock
material
(
usually
less
than
four
percent
of
the
active
chlorite
is
chlorate).
A
typical
range
of
ClO
3
­

carryover
to
the
finished
water
from
chlorite
feedstock
contamination
is
about
50
µ
g/
L
for
a
1­
mg/
L
dose
of
chlorine
dioxide.
(
Gates,
1998)
Technology
to
generate
chlorine
dioxide
using
sodium
chlorate
is
now
available
to
the
drinking
water
industry,
which
introduces
the
possibility
of
ClO
3
­
carryover
to
the
finished
water
from
the
chlorate
feedstock.

Chlorate
ion
may
also
be
produced
due
to
inefficient
generation
of
chlorine
dioxide.
Excess
chlorine
will
favor
the
production
of
ClO
3
­
over
chlorine
dioxide,
as
will
keeping
the
generator
mixtures
at
highly
alkaline
(
pH
>
11)
or
acidic
(
pH
<
3)
conditions.
If
the
concentrations
of
feedstock
reactants
are
too
low
or
too
much
dilution
water
is
added
during
the
reaction,
ClO
3
­

formation
is
also
favored.

Chlorite
ion
(
ClO
2
­)
is
a
major
degradation
product
resulting
from
the
reaction
of
chlorine
dioxide
with
inorganic
and
organic
constituents
in
the
water.
When
free
chlorine
is
used
after
the
application
of
chlorine
dioxide
in
the
treatment
process,
ClO
2
­
is
oxidized
to
ClO
3
­.
This
conversion
will
continue
over
time
as
the
water
travels
through
the
distribution
system.
Chlorate
ion
is
also
formed
by
photodecomposition
of
chlorine
dioxide
when
treated
water
is
exposed
to
bright
sunlight
in
open
basins.

The
primary
ways
in
which
water
systems
can
control
the
levels
of
ClO
3
­
in
the
finished
water
is
through
high
efficiency
operation
of
their
chlorine
dioxide
generators
and
by
reducing
ClO
2
­

concentrations
prior
to
the
addition
of
free
chlorine.
Careful
control
of
the
generation
process
minimizes
ClO
3
­
formation.
Ferrous
ion,
which
is
a
coagulant
aid,
can
be
used
to
convert
ClO
2
­
to
chloride
ion
and
thus
prevent
it
from
reacting
with
free
chlorine
to
form
ClO
3
­.
Prepared
by
Pat
Fair,
OW/
OGWDW
8/
10/
05
Hypochlorite:
Some
water
systems
use
sodium
hypochlorite
or
calcium
hypochlorite
as
their
source
of
free
chlorine.
Chlorate
ion
can
be
formed
in
these
products
during
the
manufacturing
process,
but
the
decomposition
of
hypochlorite
solutions
during
storage
is
the
more
significant
source
of
ClO
3
­
in
systems
using
hypochlorite.
Sodium
hypochlorite
is
usually
purchased
as
a
solution,
and
ClO
3
­
concentrations
increase
between
the
time
of
manufacture
and
delivery
to
the
water
plant.
Calcium
hypochlorite
is
a
solid,
and
thus
ClO
3
­
concentrations
don't
increase
until
calcium
hypochlorite
solutions
are
prepared
for
use
at
the
water
treatment
plant.

The
rate
at
which
hypochlorite
ion
(
OCl­)
disproportionates
to
ClO
3
­
is
influenced
by
concentration
of
OCl­,
pH,
and
temperature.
The
rate
of
decomposition
increases
as
the
concentration
of
OCl­
increases,
so
water
systems
can
use
dilution
as
one
control
strategy.
The
pH
should
be
in
the
12
to
13
range
to
minimize
decomposition;
a
pH
below
11
greatly
increases
the
rate
of
decomposition.
Hypochlorite
solutions
should
be
protected
from
high
temperatures
and
sunlight.
Storage
time
should
be
minimized;
both
from
the
time
of
manufacture
to
delivery
and
from
the
time
of
delivery
to
use.

Chlorate
Ion
Occurrence
Data
Data
on
the
occurrence
of
ClO
3
­
in
drinking
water
are
available
from
two
primary
sources:
the
Information
Collection
Rule
(
ICR)
Auxiliary
1
Database,
Version
5.0
(
USEPA,
2000)
and
the
AwwaRF
research
study
on
the
control
of
ClO
3
­
in
hypochlorite
solutions
(
Gordon
et
al,
1995).

Information
Collection
Rule:
The
most
extensive
data
on
the
occurrence
of
ClO
3
­
in
drinking
water
is
from
the
ICR
(
USEPA,
1996).
Source
water
and
drinking
water
were
monitored
for
ClO
3
­
between
July
1997
and
December
1998.
Water
systems
serving
a
population
of
at
least
100,000
were
required
to
monitor
for
ClO
3
­
at
treatment
plants
using
chlorine
dioxide
or
hypochlorite
solutions
in
the
treatment
process.
Plants
using
chlorine
dioxide
collected
monthly
samples
of
the
source
water
entering
the
plant,
the
finished
water
leaving
the
plant,
and
at
three
sample
points
in
the
distribution
system
(
near
the
first
customer,
an
average
residence
time
and
a
maximum
residence
time).
Plants
using
hypochlorite
solutions
were
only
required
to
collect
quarterly
samples
of
the
water
entering
and
leaving
the
plant.
If
chlorine
dioxide
or
hypochlorite
solutions
were
used
intermittently
at
a
plant,
ClO
3
­
samples
were
only
required
in
sample
periods
in
which
they
were
in
use.

Chlorine
dioxide
was
used
by
22
water
systems
(
29
treatment
plants)
during
at
least
one
of
the
18
monthly
ICR
sampling
periods.
Data
from
413
samples
collected
at
the
entry
point
to
the
distribution
system
showed
ClO
3
­
concentrations
ranging
from
<
20
µ
g/
L
to
1,600
µ
g/
L.
The
ClO
3
­
concentrations
ranged
from
<
20
µ
g/
L
to
2,200
µ
g/
L
in
the
1084
samples
collected
in
the
distribution
system.
The
distribution
of
average
ClO
3
­
concentrations
calculated
for
each
treatment
plant
and
sample
point
are
summarized
in
Table
1.
The
distribution
system
average
concentrations
determined
for
each
water
plant
by
averaging
the
data
from
the
three
distribution
system
sample
points
are
summarized
in
the
last
column
of
Table
1.
The
median
distribution
system
average
concentration
is
129
µ
g/
L
with
a
range
from
<
20
µ
g/
L
to
691
µ
g/
L.
Prepared
by
Pat
Fair,
OW/
OGWDW
8/
10/
05
Sodium
hypochlorite
solutions
were
in
use
in
44
water
systems
(
61
treatment
plants)
during
the
six
quarterly
ICR
sampling
periods.
(
None
of
the
systems
reported
using
calcium
hypochlorite
as
the
source
of
their
chlorine
solutions.)
Data
from
312
samples
were
reported
with
concentrations
ranging
from
<
20
µ
g/
L
to
a
maximum
of
1,400
µ
g/
L.
The
average
ClO
3
­
concentration
in
the
finished
drinking
water
for
each
treatment
plant
ranged
from
<
20
µ
g/
L
to
502
µ
g/
L
with
a
median
concentration
of
99
µ
g/
L.
Table
1
summarizes
the
distribution
of
average
ClO
3
­

concentrations
calculated
for
each
plant.

Table
1.
Chlorate
Concentrations1
(
µ
g/
L)
­
ICR
Data
Hypochlorite
Plants
Chlorine
Dioxide
Treatment
Plants
Finished
Finished
Near
First
Customer
Average
Retention
Time
Maximum
Retention
Time
Distribution
System
Average2
10th
Percentile
23
56
52
55
35
52
20th
Percentile
37
77
95
84
71
79
50th
Percentile
99
119
126
132
138
129
80th
Percentile
155
195
203
232
230
217
90th
Percentile
239
226
239
282
301
264
Maximum
502
687
632
735
707
691
#
WTPs
61
29
27
27
27
27
#
PWSs
44
22
21
21
21
21
1The
average
chlorate
concentration
was
calculated
for
each
sample
point
at
each
WTP
over
the
entire
ICR
monitoring
program.
The
distribution
of
these
averages
is
presented
in
this
table.
2The
distribution
system
average
chlorate
concentration
was
calculated
for
each
WTP
using
the
three
distribution
system
sample
points.
The
distribution
of
these
averages
is
presented
in
this
column.

AwwaRF
Hypochlorite
Project:
The
American
Water
Works
Association
Research
Foundation
sponsored
a
project
to
study
how
water
systems
could
minimize
ClO
3
­
formation
in
the
hypochlorite
solutions
they
use
for
disinfection.
As
part
of
the
data
gathering
effort,
they
obtained
information
from
185
water
systems
concerning
their
use
of
hypochlorite
solutions.
Samples
of
source
water,
hypochlorite
solution,
and
finished
drinking
water
from
111
of
the
water
systems
were
analyzed
for
ClO
3
­.
Only
one
set
of
samples
was
collected
for
each
system.

Background
information
on
the
subset
of
111
water
systems
that
provided
samples
was
not
reported
separately
from
the
185
systems
who
answered
the
questionnaire.
Therefore,
the
ClO
3
­
Prepared
by
Pat
Fair,
OW/
OGWDW
8/
10/
05
concentrations
cannot
be
directly
related
to
the
size
of
the
water
system
or
type
of
hypochlorite
solution
in
use.
However,
73.5
%
of
the
systems
who
responded
to
the
questionnaire
served
populations
less
than
100,000
with
a
subset
of
66%
serving
populations
less
than
10,000.
There
is
a
possibility
that
a
few
systems
using
calcium
hypochlorite
were
sampled
in
the
AwwaRF
project,
since
13%
of
the
185
systems
reported
using
calcium
hypochlorite
and
85%
reported
using
sodium
hypochlorite.

The
ClO
3
­
concentrations
reported
in
the
finished
water
are
summarized
in
Table
2.
The
median
concentration
in
the
finished
water
is
161
µ
g/
L.
The
ClO
3
­
concentrations
in
the
hypochlorite
solutions
ranged
from
0.03
to
113
g/
L.

Table
2.
Chlorate
Concentrations
­
AwwaRF
Project
(
PWSs
using
Hypochlorite
Solutions)

Finished
Water
Chlorate
Concentration
(
µ
g/
L)

Minimum
<
10
10th
Percentile
15
20th
Percentile
41
50th
Percentile
161
80th
Percentile
611
90th
Percentile
1,160
Max
9,180
#
PWSs
111
#
States
13
Chronic
Exposure
to
Chlorate
Ion
Even
though
the
ICR
ClO
3
­
sampling
was
targeted
to
systems
suspected
of
having
ClO
3
­

contamination
due
to
the
treatment
process
in
use,
it
is
reasonable
to
assume
that
there
were
not
significant
ClO
3
­
levels
in
the
systems
in
the
same
size
category
that
were
not
sampled.
This
is
based
on
earlier
drinking
water
studies
that
found
ClO
3
­
concentrations
in
source
water
were
too
low
to
impact
the
levels
in
drinking
water
on
the
same
scale
as
treating
the
water
with
either
chlorine
dioxide
or
hypochlorite
(
Bolyard
et
al,
1993).

The
ICR
data
confirm
the
presence
of
ClO
3
­
in
source
water
(
75
of
744
samples
of
water
entering
the
treatment
plants
contained
measurable
ClO
3
­),
but
also
demonstrate
that
the
concentrations
are
Prepared
by
Pat
Fair,
OW/
OGWDW
8/
10/
05
generally
very
low,
can
vary
considerably
over
time
at
the
same
sample
site,
and
are
minor
compared
to
those
observed
from
chlorine
dioxide
or
hypochlorite
use.
Data
were
reported
from
105
treatment
plant
influent
sample
points
in
the
ICR
and
samples
from
33
of
those
sites
contained
ClO
3
­
concentrations
of
20
µ
g/
L
or
greater.
Chlorate
concentrations
were
reported
in
influent
samples
from
both
surface
and
ground
water
sources.
Samples
from
fifteen
of
the
33
sites
contained
measurable
ClO
3
­
in
more
than
one
sampling
period,
but
with
one
exception,
the
concentrations
were
all
#
120
µ
g/
L;
70%
were
between
20
and
50
µ
g/
L.
One
influent
water
had
a
ClO
3
­
concentration
of
944
µ
g/
L
in
one
sample
period,
but
the
concentrations
were
#
100
µ
g/
L
in
the
other
sample
periods.
Three
influent
waters
contained
a
high
ClO
3
­
concentration
(
1,300
to
1,600
µ
g/
L)
in
one
sample
period
and
none
in
the
other
sample
periods.
The
ICR
data
indicate
that
the
influence
of
source
water
ClO
3
­
(
as
reflected
by
the
influent
samples)
on
the
concentrations
in
finished
drinking
water
is
minimal
compared
to
the
contribution
from
using
chlorine
dioxide
or
hypochlorite
solutions
in
the
treatment
process.

The
ICR
data
set
provides
the
best
available
estimate
of
long
term
exposure
to
ClO
3
­
from
drinking
water,
because
multiple
samples
were
collected
over
an
18
month
period.
Only
systems
serving
populations
of
at
least
100,000
were
sampled
during
the
ICR.
Even
though
this
size
category
includes
roughly
one
percent
of
the
total
number
of
drinking
water
systems
in
the
United
States,
it
serves
almost
60
percent
of
the
population.
During
the
ICR,
there
were
296
water
systems
in
this
size
category;
7%
used
chlorine
dioxide
and
15%
used
hypochlorite
solutions.

When
chlorine
dioxide
is
the
source
of
ClO
3
­
in
drinking
water,
it
is
appropriate
to
use
the
average
concentration
in
the
distribution
system
to
estimate
exposure.
This
is
because
the
concentration
is
expected
to
change
within
the
system
due
to
the
conversion
of
ClO
2
­
to
ClO
3
­
in
the
presence
of
chlorine.
Fifty
percent
of
the
chlorine
dioxide
plants
had
average
distribution
system
ClO
3
­

concentrations
of
#
129
µ
g/
L.
Ninety
percent
had
concentrations
#
264
µ
g/
L.

The
average
ClO
3
­
concentration
at
the
entry
point
to
the
distribution
system
can
be
used
to
estimate
exposure
when
hypochlorite
solutions
are
the
source
of
the
ClO
3
­
contamination.
No
additional
ClO
3
­
is
expected
to
be
formed
in
the
distribution
system.
Fifty
percent
of
the
plants
using
hypochlorite
solutions
had
finished
water
ClO
3
­
concentrations
of
#
99
µ
g/
L.
Ninety
percent
had
concentrations
#
239
µ
g/
L.

The
AwwaRF
data
set
is
much
smaller
than
the
ICR
data
set,
because
the
111
systems
from
13
states
were
only
sampled
once.
Low
levels
of
ClO
3
­
were
measured
in
almost
20%
of
the
source
waters
with
90
percent
of
the
samples
having
concentrations
less
than
35
µ
/
L.
(
Over
30%
of
the
source
waters
sampled
during
the
ICR
contained
measurable
concentrations
of
ClO
3
­
with
90
percent
having
concentrations
less
than
23
µ
/
L.)
The
finished
water
ClO
3
­
concentrations
measured
in
the
AwwaRF
study
are
generally
higher
than
those
observed
in
the
ICR.
This
difference
could
be
the
result
of
a
number
of
factors
such
as:
1)
The
AwwaRF
data
represents
a
single
point
in
time
while
the
ICR
data
reflects
an
average
over
18
months;
2)
Most
of
the
AwwaRF
samples
were
collected
from
utilities
that
served
population
of
less
than
100,000,
while
all
of
the
ICR
samples
were
from
utilities
serving
at
least
100,000;
and
3)
Hypochlorite
treatment
Prepared
by
Pat
Fair,
OW/
OGWDW
8/
10/
05
practices
may
have
changed
between
when
the
AwwaRF
samples
were
collected
(
1993)
and
the
ICR
samples
were
collected
(
1997­
98).

References
AWWA
Water
Quality
Division
Disinfection
Systems
Committee,
Committee
Report:
Disinfection
at
Large
and
Medium­
Size
Systems,
May
2000a,
p
32­
43.

AWWA
Water
Quality
Division
Disinfection
Systems
Committee,
Committee
Report:
Disinfection
at
Small
Systems,
May
2000b,
p
24­
31.

Bolyard,
M.,
Fair,
P.
S.,
and
Hautman,
D.
P.
"
Sources
of
Chlorate
Ion
in
US
Drinking
Water,"
Journal
AWWA
Vol
85(
9)
81­
88,
1993.

Gates,
D.
J.
The
Chlorine
Dioxide
Handbook.
American
Water
Works
Association,
Denver,
CO,
1998.

Gordon,
G.
G.,
Adam,
L.,
and
Bubnis,
B.
Minimizing
Chlorate
Ion
Formation
in
Drinking
Water
When
Hypochlorite
Ion
is
the
Chlorinating
Agent.
American
Water
Works
Association
Research
Foundation,
Denver,
CO,
1995.

USEPA,
1996.
National
Primary
Drinking
Water
Regulation:
Monitoring
Requirements
for
Public
Drinking
Water
Supplies:
Cryptosporidium,
Giardia,
Viruses,
Disinfection
Byproducts,
Water
Treatment
Plant
Data
and
Other
Information
Requirements.
Final
Rule.
FR
61:
94:
24354­
24388
(
May
14,
1996).

USEPA,
2000.
ICR
Auxiliary
1
Database.
EPA
815­
C­
00­
002.
Office
of
Water,
Cincinnati,
OH,
April
2000.
