UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
OFFICE
OF
PREVENTION,
PESTICIDES
AND
TOXIC
SUBSTANCES
MEMORANDUM
DATE:
April
10,
2006
SUBJECT:
Chlorine
Dioxide,
Sodium
chlorite
and
Sodium
Chlorate
(
Antimicrobial
Uses):
EPA
Response
to
Error
Correction
Comments
on
AD's
Risk
Assessment.

FROM:
Melba
Morrow,
DVM
(
Science
Team
Coordinator)
Antimicrobials
Division
(
7510C)

TO:
ShaRon
Carlisle,
Chemical
Review
Manager
Regulatory
Management
Branch
II
Antimicrobials
Division
(
7510C)

All
typographical
errors
and
other
agreed
upon
true
error
corrections
have
been
addressed
in
the
individual
scientific
chapters
and
the
risk
assessment
document.
Many
of
the
comments
provided
by
the
registrants
and
their
representatives
are
beyond
the
scope
of
error
correction
and
will
be
addressed
at
a
later
date.
These
comments
include
differences
of
opinions
on
Agency
procedures,
information
on
risk
mitigation,
comments
on
methods
used
in
conducting
dietary
assessments,
ecotoxicology
comments
and
comments
on
appropriate
/
inappropriate
values
used
in
occupational
and
residential
exposures.
Where
specific
comments
have
been
addressed
by
the
disciplinary
scientists,
specific
responses
are
provided.
The
following
comments
were
submitted
by
the
Chlorine
Dioxide
Panel
American
Chemistry
Council:
Comments
from
the
Chlorine
Dioxide
Panel
of
the
American
Chemistry
Council
Chlorine
Dioxide
Draft
Risk
Assessment
EPA
Docket
ID
Number
2006­
0328
ECOTOXICOLOGY
ASSESSMENT
Environmental
Modeling/
Exposure
Comment:
As
AD
is
aware,
all
once­
through
uses
of
sodium
chlorite/
chlorine
dioxide
require
a
National
Pollutant
Discharge
Elimination
System
(
NPDES)
permit.
The
Panel
believes
that
the
site­
specific
NPDES
permit
writer
is
better
able
to
evaluate
the
non­
target
organism
risks
that
may
result
from
the
once­
through
use
of
sodium
chlorite/
chlorine
dioxide
than
does
a
"
macro"
risk
assessment
by
AD.

This
is
beyond
the
scope
of
error
correction.

Comment:
To
address
the
issue
of
discharges
from
once­
through
systems,
the
Panel
supported
a
study
by
the
University
of
Maryland
to
assess
the
effects
of
chlorite
on
a
wide
variety
of
aquatic
species
[
Fisher
et
al.
2003].
This
data
was
reviewed
by
the
EPA
Office
of
Water.
[
EPA
Letter].
None
of
these
results
have
been
considered
in
this
section.

I
have
read
this
paper,
however
it
uses
a
different
method
of
risk
assessment
than
AD
uses
and
the
results
are
therefore
not
applicable
to
our
risk
assessment.

Comment:
The
Use
of
the
PDM4
Model.
If
the
aqueous
concentrations
calculated
by
the
PDM4
model
are
used
to
form
a
regulatory
position,
then
it
is
the
inappropriate
model,
for
two
reasons.
First,
OPPT
states
in
its
model
guidance
that
this
tier
I
model
should
not
be
used
for
regulation.
Second,
degradation
of
chlorine
dioxide
and
chlorite
can
be
significant.
The
PDM4
model
does
not
allow
for
degradation.
We
acknowledge
that
mathematical
modeling
would
be
appropriate
for
this
chemical/
use
pattern
combination,
but
PDM4
is
not
the
proper
choice.

This
is
beyond
the
scope
of
error
correction.
The
use
of
PDM4
is
consistent
with
the
approach
used
by
AD
to
model
once­
through
uses
in
the
past.

Comment:
Overly
Conservative
Scenarios.
Typical
OPP
regulatory
risk
assessments
attempt
to
calculate
the
upper
90th,
95th,
and
99th
percentiles
for
exposure,
depending
on
the
target
population
and
the
toxicological
endpoint.
The
7Q10
approach
used
in
the
draft
RED
is
more
conservative
than
even
the
upper
99th
percentile,
because
it
is
based
on
the
lowest
seven­
day
flow
in
a
10
year
period,
i.
e.,
the
lowest
seven
days
of
3,644
seven­
day
periods
(
not
Sunday
to
Sunday
weeks).
Comments
from
the
Chlorine
Dioxide
Panel
of
the
American
Chemistry
Council
Chlorine
Dioxide
Draft
Risk
Assessment
EPA
Docket
ID
Number
2006­
0328
This
is
beyond
the
scope
of
error
correction.
Again,
the
7Q10
approach
is
consistent
with
many
previous
assessments
for
this
use
conducted
by
AD.

Comment:
Dosing
for
Model
Input.
The
use
of
800
ppm
as
the
maximum
dose
for
both
dosing
scenarios
(
intermittent
and
continuous)
is
inappropriate.
This
dose
rate,
appears
only
on
a
single
label
(
EPA
Re.
No.
10707­
32,
issued
to
Baker
Petrolite).
This
product,
a
combination
of
chlorite
and
chlorate,
is
reportedly
obsolete,
and
will
be
cancelled
or
amended
by
the
registrant
to
delete
this
dose.
The
next
higher
dose
rate
(
25
ppm)
appears
on
a
number
of
registered
labels,
however
this
rate
is
still
significantly
higher
than
any
actual
use
rate
for
these
products.
Dose
rates
of
ClO2
at
this
level
would
make
this
product
economically
unacceptable.

As
previously
reported
to
the
Agency,
the
typical
dose
rate
for
oncethrough
systems
is
0.25­
3.0
ppm.
The
Panel
believes
that
a
maximum
dose
rate
of
5
ppm
is
more
realistic,
and
is
sufficient
for
achieving
efficacy
in
a
once­
through
system.
The
Panel
therefore
urges
the
Agency
to
not
permit
dose
rates
exceeding
5
ppm.

The
assessment
will
be
changed
to
include
a
statement
saying
that
it
is
believed
that
the
single
label
with
the
800
ppm
dose
rate
will
be
either
cancelled
or
amended
by
the
registrant
to
delete
this
dose.
As
long
as
the
25
ppm
dose
rate
remains
on
the
labels,
it
will
need
to
be
assessed
as
a
worst
case
scenario.
If
the
maximum
rate
is
changed
on
the
labels
to
5
ppm,
the
risk
assessment
can
be
updated.

ECOLOGICAL
HAZARD
AND
RISKS
Comment:
The
panel
has
reviewed
the
data
used
by
the
EPA
in
this
chapter
and
notes
the
following
discrepancies
(
See
attached
data
matrix):
 
EPA
endpoints
do
not
agree
with
study
endpoints
in
9
out
of
27
cases.
Some
of
the
differences
are
minor
such
as
incorrect
rounding.
Other
discrepancies
involve
numbers
very
different
from
the
study.
It
is
not
clear
whether
these
are
typos
or
study
reevaluations.
If
the
agency
has
reevaluated
the
studies,
no
explanation
for
the
methods
used
was
provided.

ACC
252854
 
The
endpoint
from
the
study
was
reevaluated
in
the
Data
Evaluation
Record
(
DER)
using
different
statistical
methods
determined
to
be
more
appropriate
for
the
data.
Comments
from
the
Chlorine
Dioxide
Panel
of
the
American
Chemistry
Council
Chlorine
Dioxide
Draft
Risk
Assessment
EPA
Docket
ID
Number
2006­
0328
MRID
94068008
 
The
NOAEC
value
in
the
assessment
was
determined
based
on
the
fact
that
there
was
no
mortality
and
no
sign
of
toxicity
or
abnormal
behavior
in
any
of
the
groups
of
organisms
and
the
highest
concentration
tested
was
5000
ppm.

ACC
252854
 
The
endpoint
from
the
study
was
reevaluated
in
the
DER.
The
EPA
NOAEC
has
been
deleted
from
the
assessment
since
it
was
not
determined
in
the
study.

ACC
252854
 
The
endpoint
from
the
study
was
reevaluated
in
the
DER.
Although
no
NOAEC
was
determined
in
the
study,
the
highest
concentration
at
which
no
mortality
occurred
was
417
ppm
and
this
is
the
NOAEC
value
used
in
the
assessment.

MRID
131351­
The
DER
for
this
study
states
that
79%
active
ingredient
was
used.
The
confidence
limits
have
been
adjusted
slightly
in
the
assessment
to
reflect
the
exact
values
in
the
study.
The
EPA
NOAEC
has
been
deleted
from
the
assessment
since
it
was
not
determined
in
the
study.

ACC
253743­
The
report
date
has
been
changed
to
1985.
The
DER
for
the
study
indicates
that
there
was
no
mortality
in
any
of
the
dilutions
for
either
rainbow
trout
or
bluegill
so
the
LC50
for
both
is
greater
than
100
ppm.

ACC
252854
 
Although
no
NOAEC
was
determined
in
the
study,
the
highest
concentration
at
which
no
mortality
occurred
was
186
ppm
and
this
is
the
NOAEC
value
used
in
the
assessment.

ACC
69810
 
This
was
an
omission.
The
EPA
NOAEC
has
been
changed
to
reflect
the
study
value:
less
than
70
ppm
rather
than
70
ppm.

ACC
253379
 
The
DER
for
this
study
states
that
80%
active
ingredient
was
used
as
the
test
material.
The
NOAEC
has
been
added
into
the
appropriate
table
in
the
assessment.

MRID
94068006
 
The
DER
(
as
well
as
the
study
summary)
for
this
study
states
that
80%
active
ingredient
was
used
as
the
test
material.
The
NOAEC
has
been
added
into
the
appropriate
table
in
the
assessment.

MRID
94068009­
The
spelling
of
the
author's
name
has
been
changed.
The
endpoint
from
the
study
was
reevaluated
in
the
DER.
The
value
was
determined
to
be
0.146
ppm
using
probit
analysis
instead
of
the
least­
square
linear
regression
model
that
was
used
in
the
study.
Although
the
NOAEC
was
not
Comments
from
the
Chlorine
Dioxide
Panel
of
the
American
Chemistry
Council
Chlorine
Dioxide
Draft
Risk
Assessment
EPA
Docket
ID
Number
2006­
0328
calculated
in
the
study,
mortality
occurred
at
all
levels
tested
except
60
ppb,
therefore
the
NOAEC
was
determined
to
be
0.06
ppm
as
stated
in
the
assessment.

MRID
43259401­
The
96
hr
LC
50
value
from
the
study
was
used
in
the
assessment
and
the
confidence
limits
have
been
adjusted
slightly
in
the
assessment
to
reflect
the
exact
values
in
the
study.

MRID
43259402
­
The
96
hr
LC
50
value
from
the
study
was
used
in
the
assessment
and
the
confidence
limits
have
been
rounded
to
the
nearest
hundredth.
The
EPA
NOAEC
has
been
deleted
from
the
assessment
since
it
was
not
determined
in
the
study.

MRID
41843102
 
Change
to
number
of
organisms
made.

MRID
41880403
 
The
endpoint
from
the
study
was
reevaluated
in
the
DER
using
EPA's
Toxanal
program
to
calculate
percent
inhibition
of
growth
using
cell
count.
The
96
hour
EC50
value
using
this
method
is
1.32
ppm
as
shown
in
the
assessment.
The
value
in
the
study
was
based
on
binomial/
nonlinear
interpolation
which
may
be
unreliable
for
use
with
tests
that
have
these
kinds
of
results.
EPA
used
Toxstat
version
3.3
(
square
root
transformation,
Bonferroni's
t­
test
and
Dunnett's
ANOVA)
to
determine
the
NOAEC
for
this
study
which
was
determined
to
be
less
than
0.62
ppm.
The
value
in
the
assessment
has
been
amended
to
show
that
the
NOAEC
is
less
than
0.62
ppm.

 
The
EPA
lists
NOAEC
levels
when
the
study
does
not
in
8
out
of
27
cases.
In
two
other
cases,
the
study
determined
a
NOAEC
value
and
the
agency
failed
to
list
it.

See
above
explanations.

 
Many
studies
list
EC50
values
for
24,
48,
72
and
96
hours,
yet
the
EPA
has
primarily
chosen
to
rely
on
96
hours,
but
not
consistently.

As
a
rule,
the
endpoints
are
based
on
96
hour
EC50
values
for
all
organisms
but
the
freshwater
invertebrates.
For
freshwater
invertebrates,
the
endpoint
is
based
on
a
48
hour
EC50
because
of
the
organism's
short
life
cycle.
If
no
96
hour
LC50
is
provided
in
the
study
(
if
the
study
does
not
last
for
96
hours),
we
list
the
endpoint
that
was
gathered
at
the
end
of
the
study.

 
Some
studies
did
not
go
to
96
hours.
In
these
cases,
the
EPA
just
lists
the
values
without
any
explanation.
Therefore,
the
table
comparison
is
similar
to
comparing
apples
to
oranges.
Comments
from
the
Chlorine
Dioxide
Panel
of
the
American
Chemistry
Council
Chlorine
Dioxide
Draft
Risk
Assessment
EPA
Docket
ID
Number
2006­
0328
See
above
explanation.

 
In
four
studies,
the
test
material
was
assumed
100%
active
by
the
lab
conducting
the
study.
It
appears
that
an
adjustment
should
be
made.
The
maximum
concentration
of
sodium
chlorite
AI
in
a
registered
product
is
80%,
therefore,
the
actual
test
results
should
be
adjusted.

See
above
explanations.
Since
the
actual
percent
active
that
was
used
in
the
studies
is
noted
in
the
tables,
comparisons
among
endpoints
can
be
made.

Not
all
of
the
cited
studies
were
immediately
accessible
to
our
reviewers;
however
given
the
number
of
discrepancies
noted
in
the
available
studies,
we
have
some
concerns.

TOXICOLOGY
Re:
Prenatal
developmental
toxicity
study
in
rabbits.

The
registrant's
edit
was
not
agreed
to.
The
author
chose
to
edit
the
summary
as
follows:
"
In
a
developmental
drinking
water
study
(
MRID
417157701),
time
 
mated
female
New
Zealand
White
rabbits
(
16­
17/
dose)
were
administered
sodium
chlorite
(
purity
80.58%)
from
gestation
days
7
through
19
at
0,
200,
600,
or
1200
ppm
(
calculated
intake
0,
12.2­
13.8,
30.1
­
38.1,
or
32.0­
60.1
mg/
kg/
day)".

DIETARY
RISK
ASSESSMENT
Many
of
the
comments
contained
in
the
dietary
chapter,
as
suggested
by
industry,
have
not
been
accepted
and
will
require
further
discussion,
as
they
involve
policy
changes
and
coordination
of
our
policies/
methodologies
with
those
of
other
regulatory
agencies.
They
are
beyond
the
scope
of
error
correction.

OCCUPATIONAL/
RESIDENTIAL
EXPOSURE
The
comment
regarding
the
inappropriate
use
of
the
RfC
for
chlorine
dioxide
has
not
been
addressed
under
error
correction.
In
addition,
no
changes
will
be
made
in
reference
to
the
comment
on
products
that
should
be
removed
from
the
residential
market.
