BEFORE
THE
UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
PETITION
OF
THE
CHEMICAL
MANUFACTURERS
ASSOCIATION
KETONES
PANEL
TO
REMOVE
METHYL
ISOBUTYL
KETONE
FROM
THE
LIST
OF
HAZARDOUS
AIR
POLLUTANTS
UNDER
SECTION
112(
b)
OF
THE
CLEAN
AIR
ACT
Carol
R.
Stack
David
F.
Zoll,
Esq.
Acting
Vice
President
Vice
President
and
CHEMSTAR
General
Counsel
Barbara
O.
Francis
Peter
G.
McHugh,
Esq.
Manager
Counsel,
CHEMSTAR
CMA
Ketones
Panel
Of
Counsel:

William
K.
Rawson
Jeffrey
R.
Holmstead
Gregory
S.
Slater
Latham
&
Watkins
1001
Pennsylvania
Avenue,
N.
W.
Suite
1300
Washington,
D.
C.
20004
April
22,
1997
CHEMICAL
MANUFACTURERS
ASSOCIATION
1300
Wilson
Blvd.
Arlington,
VA
22209
(
703)
741­
5000
i
EXECUTIVE
SUMMARY
Pursuant
to
Section
112(
b)(
3)
of
the
Clean
Air
Act
(
the
Act),
the
Ketones
Panel
of
the
Chemical
Manufacturers
Association
hereby
petitions
the
Administrator
of
the
Environmental
Protection
Agency
(
EPA)
to
remove
methyl
isobutyl
ketone
(
MIBK)
from
the
list
of
chemicals
that
are
regulated
as
hazardous
air
pollutants
(
HAPs)
under
the
Act.

Background
Information
MIBK
is
widely
used
as
a
solvent
and
chemical
intermediate.
It
is
a
highly
efficient
solvent
that
can
be
used
with
a
wide
variety
of
resins
and
is
therefore
particularly
valuable
in
the
formulation
of
high­
solids
coatings.
MIBK
is
manufactured
via
the
aldol
condensation
of
acetone
under
reductive
conditions
in
a
totally
enclosed,
continuous
process.

Delisting
Criteria
EPA
is
required
to
delist
a
substance
from
the
HAP
list
if
"
there
is
adequate
data
on
the
health
and
environmental
effects
of
the
substance
to
determine
that
emissions,
ambient
concentrations,
bioaccumulation
or
deposition
of
the
substance
may
not
reasonably
be
anticipated
to
cause
any
adverse
effects
to
human
health
or
adverse
environmental
effects."
Several
key
aspects
of
this
standard,
as
well
as
related
statutory
requirements,
must
be
taken
into
account.

First,
as
EPA
has
recognized,
the
Act
does
not
require
absolute
proof
that
a
substance
will
not
cause
any
adverse
effects.
Rather,
the
Agency
should
use
a
weight­
of­
theevidence
approach
to
determine
whether
it
is
"
reasonable"
to
anticipate
that
emissions
of
MIBK
will
cause
adverse
health
or
environmental
effects.
Second,
in
making
HAP
delisting
decisions,
the
Agency
is
not
to
consider
potential
accidental
releases,
which
are
regulated
under
Section
112(
r).
Rather,
it
must
consider
whether,
under
normal
operating
conditions,
emissions
can
reasonably
be
anticipated
to
cause
adverse
health
or
environmental
effects.
Finally,
Section
112(
b)(
2)
of
the
Act
makes
it
clear
that
MIBK
should
not
be
listed
as
a
HAP
solely
because
it
is
a
volatile
organic
compound
(
VOC).

Data
on
Health
and
Environmental
Effects
Acute
Health
Effects.
MIBK
is
not
acutely
toxic
at
the
low
concentration
levels
that
are
reasonably
likely
to
exist
beyond
facility
boundaries.
Studies
in
laboratory
animals
by
the
oral
and
inhalation
routes
of
administration
show
that
the
acute
toxicity
of
MIBK
is
low.
Exposures
of
humans
to
above
100
ppm
MIBK
may
cause
mild
eye
and
nose
irritation,
but
are
without
permanent
effects.
No
significant
acute
effects
are
expected
at
exposures
below
these
levels.
Results
from
air
dispersion
modeling
using
EPA­
approved
techniques
show
that
maximum
ambient
concentrations
beyond
industrial
site
boundaries
are
well
below
levels
of
concern.

Chronic
Health
Effects.
MIBK
also
may
not
reasonably
be
anticipated
to
cause
significant
chronic
health
effects
in
humans.
Inhalation
studies
conducted
with
rats,
mice,
dogs
and
monkeys
indicate
low
subchronic
toxicity.
The
results
from
many
different
mutagenicity
screening
assays
indicate
that
MIBK
exhibits
very
little,
if
any,
mutagenic
activity.
Existing
studies
also
indicate
that
MIBK
is
not
teratogenic
and
exhibits
low
developmental
toxicity.
ii
MIBK
may
cause
transient
anesthetic
effects
at
high
vapor
concentrations,
but
this
effect
is
readily
reversible
and
is
not
associated
with
any
evidence
of
neurotoxicity.
MIBK
has
not
been
tested
specifically
for
carcinogenicity,
because
data
on
its
structure
and
metabolism,
subchronic
health
effects
and
the
genotoxicity
studies
indicate
that
MIBK
is
not
likely
to
have
oncogenic
properties.
Thus,
the
weight
of
the
evidence
shows
that
MIBK
cannot
reasonably
be
anticipated
to
cause
cancer
or
other
chronic
health
effects
in
humans.

The
IRIS
database
does
not
contain
an
inhalation
reference
concentration
(
RfC)
for
MIBK.
However,
the
Ketones
Panel
has
calculated
an
RfC
for
MIBK
based
on
the
NOAEL
from
the
subchronic
inhalation
study
used
by
EPA
to
calculate
the
composite
score
for
MIBK
under
Section
112(
g)
of
the
Clean
Air
Act.
Based
on
this
study,
and
using
EPA's
1994
guidance
on
deriving
RfCs,
the
RfC
for
MIBK
is
2.4
mg/
m3.
This
value
represents
a
conservative
estimate
of
the
concentration
of
MIBK
in
air
to
which
an
individual
could
be
exposed
for
a
lifetime
without
adverse
effects,
and
far
exceeds
likely
human
exposure
levels
from
industrial
releases
of
MIBK.

Environmental
Effects.
MIBK
also
does
not
cause
significant
adverse
environmental
effects.
MIBK
occurs
naturally
in
many
plants
and
animals.
It
undergoes
rapid
degradation
in
the
atmosphere
and
is
readily
biodegradable.
Based
on
modeling
of
industrial
emissions
and
available
ambient
concentration
data,
the
levels
of
MIBK
likely
to
be
found
in
the
environment
are
well
below
the
lowest
toxicity
thresholds
for
micro­
organisms
and
aquatic
organisms.

Data
on
Emissions
and
Exposure
Emissions
Data.
The
Toxics
Release
Inventory
(
TRI)
shows
that
over
1,000
facilities
reported
emissions
of
MIBK
in
1994.
Most
of
these
sources
were
very
small,
however,
with
over
90
percent
of
them
reporting
emissions
of
less
than
25
tons.

Ambient
Monitoring
Data.
MIBK
has
been
reported
in
ambient
air
at
very
low
concentrations
at
a
limited
number
of
sites
in
rural
and
urban
locations.
Monitored
levels
of
MIBK
­­
even
in
industrial
areas
­­
typically
are
several
orders
of
magnitude
below
the
calculated
RfC.

Air
Dispersion
Modeling
Data
for
Industrial
Facilities.
The
Panel
funded
a
study
by
ENSR
Corporation
to
model
the
maximum
off­
site
concentrations
of
MIBK
at
a
wide
variety
of
facilities,
including
the
largest
sources
of
MIBK
emissions
in
the
country.
As
part
of
this
study,
the
Panel
identified
all
facilities
that
reported
MIBK
emissions
of
more
than
150
tons
in
1994.
It
contacted
each
of
these
facilities
(
12
based
on
1994
TRI
data)
to
gather
information
that
could
be
used
to
model
maximum
off­
site
concentrations.
Based
on
data
provided
by
these
companies,
as
well
as
information
available
from
public
sources
such
as
permit
applications,
ENSR
was
able
to
conduct
site­
specific
dispersion
modeling
for
10
of
these
12
facilities,
including
the
top
two
emitters.
ENSR
also
developed
a
generalized
modeling
approach,
based
on
the
methodology
employed
by
EPA
in
its
rulemaking
under
Section
112(
g)
of
the
Clean
Air
Act,
to
evaluate
smaller
sources
of
MIBK
emissions.
Based
on
its
study
of
both
large
and
small
sources,
ENSR
concluded
that
maximum
airborne
concentrations
beyond
facility
boundaries
are
well
iii
below
levels
of
concern
and
cannot
reasonably
be
anticipated
to
cause
adverse
health
or
environmental
effects.
The
Panel
also
analyzed
the
potential
that
groups
of
sources
might
collectively
result
in
levels
of
concern,
and
found
that
there
is
no
such
grouping.

Effect
of
Delisting
on
Emissions
and
Ambient
Concentrations.
If
MIBK
is
removed
from
the
list
of
HAPs,
use
of
MIBK
is
likely
to
increase.
For
several
reasons,
however,
MIBK
emissions
are
unlikely
to
increase
substantially.
MIBK
will
continue
to
be
regulated
as
a
VOC
and
is
often
used
in
blends
with
other
compounds
that
will
continue
to
be
regulated
as
HAPs.
In
addition,
MIBK
is
most
widely
used
in
paint
and
coating
applications,
where
performance
requirements
impose
inherent
limits
on
the
amount
of
MIBK
that
can
be
used.
Moreover,
based
on
the
available
monitoring
data
and
the
dispersion
modeling
analysis
conducted
by
ENSR,
any
reasonably
likely
increase
in
emissions
would
not
be
expected
to
result
in
ambient
levels
of
concern.
Perhaps
most
importantly,
removing
MIBK
from
the
HAP
list
is
likely
to
decrease
total
VOC
emissions
by
encouraging
the
use
of
MIBK
in
place
of
other
less
effective
solvents.

Other
Considerations
that
Weigh
in
Favor
of
Delisting
Delisting
MIBK
Would
Help
to
Reduce
VOC
Emissions
from
Many
Coating
Operations.
Over
the
last
several
years,
EPA
and
state
regulators
have
encouraged
or
required
the
use
of
high­
solid
coatings
as
an
effective
way
to
reduce
VOC
emissions
from
coating
operations.
It
is
well
known
that
MIBK
is
especially
valuable
in
the
formulation
of
high­
solids
coatings.
MIBK
effectively
dissolves
a
wide
variety
of
resins
and
is
a
more
efficient
solvent
than
the
available
non­
ketone
alternatives.
Thus,
the
use
of
MIBK
allows
the
formulation
of
coatings
with
higher
solids
content
and
lower
VOC
emissions.
In
EPA's
recent
rule
on
shipbuilding
coatings,
the
Agency
explicitly
recognized
that
the
use
of
highly
efficient
solvents
such
as
MIBK
is
the
most
effective
approach
for
reducing
VOC
emissions
in
some
coating
applications.
See
59
Fed.
Reg.
62681,
62688
(
Dec.
6,
1994).

EPA
Has
Recognized
in
Other
Contexts
that
MIBK
Has
Relatively
Low
Toxicity.
In
two
recent
rulemakings,
EPA
has
evaluated
the
health
effects
data
on
MIBK
and
concluded
that
MIBK
has
relatively
low
toxicity.
In
the
Agency's
proposed
rule
under
Section
112(
g)
of
the
Clean
Air
Act,
EPA
developed
a
methodology
for
ranking
the
relative
hazards
of
the
chemicals
listed
as
hazardous
air
pollutants
(
HAPs)
and
found
that
MIBK
was
among
the
least
toxic
of
the
listed
chemicals
(
approximately
187
out
of
189).
At
the
same
time,
EPA
also
proposed
"
de
minimis
values"
for
listed
HAPs.
These
de
minimis
values
were
intended
to
represent
the
amount
of
a
chemical
that
a
typical
facility
could
emit
without
posing
more
than
a
"
trivial"
health
risk.
Although
the
de
minimis
values
in
the
proposed
rule
were
"
capped"
at
10
tons
per
year
for
policy
reasons,
the
true
"
uncapped"
de
minimis
value
for
MIBK
based
on
EPA's
methodology
would
have
been
5,000
tons
per
year.
This
amount
is
approximately
an
order
of
magnitude
higher
than
the
emissions
of
the
facility
reporting
the
highest
MIBK
emissions
in
the
country
in
1994.
EPA
also
evaluated
the
toxicity
of
MIBK
under
its
Significant
New
Alternatives
Policy
(
SNAP)
program,
and
determined
that
it
has
"
comparatively
low
toxicity."

MIBK's
Inclusion
on
the
HAP
List
Was
Not
Based
on
a
Finding
of
Toxicity.
The
initial
HAP
list
was
developed
from
the
list
of
chemicals
that
must
be
reported
under
Section
313
iv
of
the
Emergency
Planning
and
Community
Right­
to­
Know
Act
of
1986
(
EPCRA).
MIBK
was
included
on
the
Section
313
list
solely
because
it
had
been
included
in
a
"
Survey
List"
of
chemicals
prepared
by
the
State
of
Maryland.
Inclusion
of
MIBK
in
the
Maryland
Survey
List
was
not
based
on
a
finding
of
toxicity
or
adverse
environmental
effects.
There
is
no
evidence
that
the
inclusion
of
MIBK
on
the
original
HAP
list
was
based
on
a
determination
by
Congress,
EPA
or
anyone
else
that
emissions
of
MIBK
can
reasonably
be
anticipated
to
cause
adverse
health
or
environmental
effects.

For
the
foregoing
reasons,
and
as
set
forth
in
greater
detail
in
this
Petition,
the
Ketones
Panel
respectfully
urges
the
Administrator
to
remove
MIBK
from
the
list
of
chemicals
that
are
regulated
as
HAPs
under
the
Clean
Air
Act.
v
TABLE
OF
CONTENTS
EXECUTIVE
SUMMARY..........................................................................................................
i
INTRODUCTION......................................................................................................................
1
I.
BACKGROUND
INFORMATION.................................................................................
3
A.
Chemical
and
Physical
Properties
.........................................................................
3
B.
Production
and
Use..............................................................................................
3
C.
Natural
Sources
of
MIBK
....................................................................................
4
II.
STATUTORY
CRITERIA
FOR
DELISTING.................................................................
4
A.
Standard
of
Proof
for
Delisting.............................................................................
5
B.
A
Substance
May
Not
Be
Listed
as
a
HAP
Unless
it
Reasonably
Can
Be
Expected
to
Cause
Adverse
Effects
Under
Normal
Conditions............................................................................................................
5
C.
MIBK's
Status
as
a
VOC
Is
Not
Relevant
to
the
Decision
of
Whether
it
Should
Be
Listed
as
a
HAP
.................................................................
6
III.
DATA
ON
HEALTH
AND
ENVIRONMENTAL
EFFECTS..........................................
7
A.
Inhalation
Is
the
Only
Significant
Route
of
Human
Exposure
to
MIBK
Emissions..................................................................................................
8
B.
MIBK
Cannot
Reasonably
Be
Anticipated
to
Cause
Adverse
Acute
Health
Effects
in
Humans
.....................................................................................
8
C.
MIBK
Cannot
Reasonably
be
Anticipated
to
Cause
Adverse
Chronic
Health
Effects
In
Humans......................................................................
15
1.
Subchronic
Studies
....................................................................................
16
2.
Mutagenicity
.............................................................................................
19
3.
Carcinogenicity..........................................................................................
20
4.
Reproductive
and
Developmental
Toxicity.................................................
21
5.
Neurotoxicity
............................................................................................
22
6.
Other
Effects
.............................................................................................
29
7.
Derivation
of
an
Inhalation
Reference
Concentration
(
RfC)
for
MIBK.................................................................................................
34
8.
Conclusions
Regarding
Potential
Chronic
Effects.......................................
34
vi
D.
MIBK
Does
Not
Cause
Significant
Adverse
Environmental
Effects
....................
35
1.
Biodegradation..........................................................................................
35
2.
Potential
For
Bioaccumulation...................................................................
36
3.
Effects
On
Microorganisms
.......................................................................
36
4.
Effects
on
Aquatic
Organisms....................................................................
37
5.
Effects
on
Plants........................................................................................
40
IV.
DATA
ON
EMISSIONS
AND
EXPOSURE.................................................................
40
A.
Emissions
Data...................................................................................................
40
B.
Ambient
Monitoring
Data...................................................................................
41
C.
Air
Dispersion
Modeling
Data
for
Industrial
Facilities.........................................
42
1.
Air
Dispersion
Modeling
of
the
Highest
Emitters
.......................................
43
2.
Air
Dispersion
Modeling
of
Smaller
Sources..............................................
48
3.
Potential
Impacts
from
Groups
of
Sources.................................................
51
D.
Effect
of
Delisting
on
Emissions
and
Ambient
Concentration..............................
52
V.
OTHER
REASONS
FOR
DELISTING
MIBK..............................................................
54
A.
Delisting
MIBK
Will
Help
to
Reduce
VOC
Emissions
from
Many
Coating
Operations
............................................................................................
54
B.
EPA
Has
Recognized
in
Other
Contexts
that
MIBK
Has
Relatively
Low
Toxicity
.....................................................................................................
58
1.
Proposed
Rule
Under
Section
112(
g)
of
the
Clean
Air
Act.........................
58
2.
Final
SNAP
Rule
.......................................................................................
60
C.
The
Inclusion
of
MIBK
on
the
Initial
HAP
List
Was
Not
Based
on
a
Finding
of
Adverse
Health
or
Environmental
Effects........................................
60
CONCLUSION.........................................................................................................................
62
vii
APPENDICES
Volume
I
A.
Table
of
References
B.
RfC
Calculation
for
MIBK
C.
Table
of
Ambient
Air
Concentration
Levels
of
MIBK
(
taken
from
a
study
conducted
by
the
State
of
New
Jersey
Department
of
Environmental
Protection
in
1979)

D.
Houston
Regional
Monitoring
Report
(
excerpts)

E.
Report
on
the
ENSR
Modeling
Study
F.
MIBK
Emissions
by
Zip
Code
Volumes
2,
3
and
4
G.
References
INTRODUCTION
Pursuant
to
Section
112(
b)(
3)
of
the
Clean
Air
Act
(
the
Act),
the
Ketones
Panel
of
the
Chemical
Manufacturers
Association
(
CMA)
hereby
petitions
the
Administrator
of
the
Environmental
Protection
Agency
(
EPA)
to
remove
methyl
isobutyl
ketone
(
MIBK)
from
the
list
of
chemicals
that
are
regulated
as
"
hazardous
air
pollutants"
(
HAPs)
under
Section
112
of
the
Act.
The
Ketones
Panel
includes
all
domestic
manufacturers
of
MIBK
as
well
as
manufacturers
of
several
other
ketone
solvents.
1
MIBK
is
a
highly
efficient
solvent
that
is
widely
used
in
a
variety
of
applications,
and
is
particularly
valuable
in
the
formulation
of
high­
solids
paints
and
coatings.

Under
Section
112(
b)(
3)(
C)
of
the
Act,
EPA
is
required
to
remove
a
substance
from
the
list
of
HAPs
upon
a
showing
that
"
there
is
adequate
data
on
the
health
and
environmental
effects
of
the
substance
to
determine
that
emissions,
ambient
concentrations,

bioaccumulation
or
deposition
of
the
substance
may
not
reasonably
be
anticipated
to
cause
any
adverse
effects
to
human
health
or
adverse
environmental
effects."
This
petition
reviews
the
considerable
body
of
literature
on
the
health
and
environmental
effects
of
MIBK,
along
with
extensive
data
on
releases
and
ambient
concentrations,
to
show
that
MIBK
meets
this
standard.

In
1988,
EPA
reviewed
the
health
and
environmental
effects
data
on
MIBK
in
connection
with
two
petitions
submitted
by
the
Panel
asking
that
MIBK
and
methyl
ethyl
ketone
(
MEK)
be
removed
from
the
list
of
chemicals
that
are
reportable
under
Section
313
of
the
Emergency
Planning
and
Community
Right­
to­
Know
Act
of
1986
(
EPCRA).
At
that
time,
EPA
1
The
members
of
the
Ketones
Panel
are:
Eastman
Chemical
Company,
Exxon
Chemical
Company,
Hoechst
Celanese
Chemical
Group,
Inc.,
Shell
Chemical
Company
and
Union
Carbide
Corporation.
2
identified
three
areas
of
concern
with
the
EPCRA
Section
313
petitions:
developmental
toxicity,

neurotoxicity
and
potential
liver
and
kidney
effects.
As
a
result,
the
petitions
were
withdrawn.

EPA's
concerns
with
the
1988
MEK
petition
have
now
been
resolved
in
the
IRIS
database.

Although
the
IRIS
database
for
MIBK
has
not
yet
been
updated,
the
Panel
believes
that
EPA's
treatment
of
the
MEK
data
in
the
IRIS
database
also
addresses
the
concerns
raised
about
MIBK
in
1988.
Moreover,
those
concerns
are
not
supported
by
the
extensive
body
of
literature
regarding
the
health
and
environmental
effects
of
MIBK.
Based
on
EPA's
current
methodology
for
deriving
inhalation
reference
concentrations
(
RfCs),
the
RfC
for
MIBK
is
2.4
mg/
m3.
This
value
is
indicative
of
MIBK's
low
toxicity,
and
is
consistent
with
the
Agency's
recognition
in
other
regulatory
contexts
that
MIBK
has
relatively
low
toxicity.

Part
I
of
this
Petition
provides
general
background
information
on
MIBK.
Part
II
discusses
the
statutory
criteria
for
delisting
substances
from
the
list
of
HAPs.
Part
III
reviews
the
data
on
the
potential
health
and
environmental
effects
of
exposure
to
MIBK,
and
also
explains
how
the
Panel
calculated
the
RfC
for
MIBK.
Part
IV
reviews
the
data
that
the
Ketones
Panel
has
developed
on
emissions
and
ambient
concentrations
of
MIBK,
and
shows
that
ambient
MIBK
levels
­­
even
maximum
off­
site
levels
around
the
largest
industrial
sources
of
MIBK
emissions
­­

are
well
below
the
RfC.
Finally,
Part
V
discusses
several
other
considerations
that
weigh
in
favor
of
removing
MIBK
from
the
HAP
list.
Based
on
the
information
presented
in
the
Petition,
the
Panel
respectfully
requests
that
the
Administrator
remove
MIBK
from
the
list
of
HAPs
under
Section
112
of
the
Clean
Air
Act.
2
2
The
Panel
has
also
submitted
a
separate
petition
asking
that
MIBK
be
delisted
under
Section
313
of
EPCRA.
3
I.
BACKGROUND
INFORMATION
A.
Chemical
and
Physical
Properties
MIBK
is
a
clear
colorless
liquid
with
a
sharp,
sweet
odor
and
a
molecular
weight
of
100.
Zakhari
et
al.
(
1977).
3
It
has
moderate
water
solubility
and
a
vapor
pressure
of
15
mm
Hg
at
20
°
C.
The
melting
point
and
boiling
point
of
MIBK
are
approximately
­
85
°
C
and
116
°
C,

respectively.
Id.
The
log
of
the
octanol/
water
partition
coefficient
is
1.19
(
HSDB
1996).

B.
Production
and
Use
MIBK
is
manufactured
via
the
aldol
condensation
of
acetone
under
reductive
conditions
wherein
dehydration
and
hydrogenation
immediately
follow
the
initial
condensation
to
form
MIBK.
The
process
and
equipment
are
continuous
and
enclosed,
and
product
is
transferred
to
bulk
storage
tanks
through
closed
lines.
The
equipment
and
tanks
customarily
are
vented
to
water
scrubbers
or
through
conservation
vents
to
prevent
loss
to
the
atmosphere
through
evaporation.
These
practices
help
to
minimize
releases
of
MIBK
to
the
environment.

MIBK
currently
is
produced
in
the
United
States
by
three
companies:
Eastman
Chemical,
Shell
Chemical
and
Union
Carbide.
Estimated
total
domestic
capacity
in
1995
was
approximately
220
million
pounds
(
Chemical
Marketing
Reporter,
August
5,
1996).

MIBK
is
used
both
as
a
solvent
and
as
a
chemical
intermediate.
It
is
a
highly
efficient
solvent
that
dissolves
a
wide
variety
of
resins
and,
therefore,
is
widely
used
in
surface
coatings,
adhesives,
inks,
and
traffic
marking
paint.
As
discussed
in
Section
V.
A
below,
it
is
especially
valuable
in
the
formulation
of
high­
solids
coatings,
which
increasingly
are
being
used
to
reduce
emissions
of
volatile
organic
compounds
(
VOCs)
from
many
types
of
coating
operations.

3
The
Table
of
References
is
found
in
Appendix
A.
Duplicate
copies
of
all
references
denoted
with
an
asterix
in
Appendix
A
have
been
submitted
with
this
Petition.
These
references
are
contained
in
Appendix
G.
4
MIBK
is
also
used
as
a
solvent
in
cleaning
fluids
and
dewaxing
agents,
and
as
an
extraction
medium
for
fats,
oils,
waxes
and
resins.

C.
Natural
Sources
of
MIBK
Methyl
ketones,
such
as
MIBK,
occur
naturally
in
plants
and
animals.
They
can
be
found
in
the
odorous
secretions
of
insects
and
in
dairy
products
including
milk,
butter,
and
cheese
(
Dumont
and
Adda
1978;
Gordon
and
Morgan
1979).
In
mammals,
MIBK
follows
well
known
intermediary
metabolism
pathways
including
reduction
and
oxidation
reactions
(
DiVincenzo
et
al.

1976).

II.
STATUTORY
CRITERIA
FOR
DELISTING
When
Congress
adopted
the
1990
Amendments
to
the
Clean
Air
Act,
it
placed
189
chemicals
and
chemical
categories
on
the
"
initial
list"
of
substances
to
be
regulated
as
HAPs.
See
Section
112(
b)(
1).
Congress
recognized,
however,
that
this
initial
list
was
not
necessarily
definitive,
but
should
be
reviewed
and,
if
appropriate,
revised
based
on
the
best
available
science.

Significantly,
Congress
authorized
the
Agency
not
only
to
add
to
the
list,
but
also
to
remove
substances
from
the
original
list.
It
thus
acknowledged
the
possibility
that
some
substances
on
the
initial
list
should
not
be
regulated
as
HAPs.

Under
Section
112(
b)(
3)
of
the
Act,
Congress
established
the
criteria
that
EPA
must
use
in
making
decisions
about
adding
or
removing
chemicals
from
the
list.
Under
Section
112(
b)(
3)(
C),
EPA
is
required
to
remove
a
substance
from
the
HAP
list
"
upon
a
showing"
that
there
is
adequate
data
on
the
health
and
environmental
effects
of
the
substance
to
determine
that
emissions,
ambient
concentrations,
bioaccumulation
or
deposition
of
the
substance
may
not
reasonably
be
anticipated
to
cause
any
adverse
effects
to
human
health
or
adverse
environmental
effects.
5
This
is
the
basic
standard
under
which
EPA
must
decide
whether
to
remove
MIBK
from
the
HAP
list.
As
discussed
below,
there
are
several
key
aspects
of
this
standard,
as
well
as
related
statutory
requirements,
that
must
be
taken
into
account.

A.
Standard
of
Proof
for
Delisting
The
delisting
standard
requires
that
there
be
"
adequate"
data
to
show
that
adverse
effects
"
may
not
reasonably
be
anticipated."
The
Agency
itself
has
recognized
that
Section
112(
b)
does
not
require
absolute
proof
that
a
substance
will
not
cause
adverse
effects:

The
EPA
does
not
interpret
section
112(
b)(
3)(
C)
to
require
absolute
certainty
that
a
pollutant
will
not
cause
adverse
effects
on
human
health
or
the
environment
before
it
may
be
deleted
from
the
list.
The
use
of
the
terms
"
adequate"
and
"
reasonably"
indicate
that
the
Agency
must
weigh
the
potential
uncertainties
and
their
likely
significance.

60
Fed.
Reg.
48081,
48082
(
Sep.
18,
1995)
(
proposal
to
remove
caprolactam
from
the
HAP
list).

Thus,
in
evaluating
both
the
exposure
data
and
the
data
on
health
and
environmental
effects,
the
Agency
should
use
a
weight­
of­
the­
evidence
approach
to
determine
whether
it
is
"
reasonable"
to
anticipate
that
emissions
of
MIBK
will
cause
adverse
health
or
environmental
effects.
The
Panel
believes
that
the
data
presented
below
clearly
show
that
"
emissions,
ambient
concentrations,

bioaccumulation
or
deposition
of
[
MIBK]
may
not
reasonably
be
anticipated
to
cause
any
adverse
effects
to
human
health
or
adverse
environmental
effects."

B.
A
Substance
May
Not
Be
Listed
as
a
HAP
Unless
it
Reasonably
Can
Be
Expected
to
Cause
Adverse
Effects
Under
Normal
Conditions
At
high
exposure
levels,
virtually
all
chemicals
can
cause
adverse
health
or
environmental
effects.
Under
Section
112(
b)(
3),
however,
a
substance
is
to
be
listed
as
a
HAP
only
if
"
emissions,
ambient
concentrations,
bioaccumulation
or
deposition"
of
the
substance
can
"
reasonably
be
anticipated"
to
result
in
levels
that
are
high
enough
to
cause
such
effects.
Thus,
if
6
emissions
of
a
listed
substance
are
not
reasonably
expected
to
result
in
ambient
levels,
deposition,

or
bioaccumulation
that
reasonably
can
be
anticipated
to
cause
adverse
health
or
environmental
effects,
then
that
substance
meets
the
Section
112(
b)
standard
for
delisting.
As
discussed
below,

there
is
no
appreciable
deposition
or
bioaccumulation
of
MIBK,
and
ambient
concentrations
are
far
below
levels
that
reasonably
may
be
expected
to
cause
adverse
effects.

In
this
regard,
it
is
significant
that
accidental
chemical
releases
are
addressed
in
another
part
of
the
Act,
Section
112(
r).
Section
112(
b)(
2)
specifically
states
that
accidental
releases
that
are
subject
to
regulation
under
Section
112(
r)
are
not
to
be
considered
in
HAP
listing
decisions.
Thus,
it
is
clear
that
listing
and
delisting
decisions
must
be
made
based
on
exposure
levels
that
result
from
normal
or
routine
emissions,
not
from
accidental
releases.

C.
MIBK's
Status
as
a
VOC
Is
Not
Relevant
to
the
Decision
of
Whether
it
Should
Be
Listed
as
a
HAP
Like
most
solvents,
MIBK
is
a
volatile
organic
compound
(
VOC).
VOCs
are
regulated
as
ozone
precursors
under
Title
I
of
the
Act
because
they
can
react
photochemically
with
other
pollutants
to
form
ground­
level
ozone.
Congress
made
it
clear,
however,
that
a
substance
is
not
to
be
listed
as
a
HAP
solely
because
it
is
a
VOC.
Section
112(
b)(
2)
of
the
Act
provides
that
a
substance
which
is
a
precursor
to
a
pollutant
(
such
as
ozone)
that
is
listed
under
Section
108(
a)
of
the
Act
may
not
be
included
on
the
HAP
list
unless
it
"
independently
meets"
the
HAP
listing
criteria.
The
listing
criteria
under
Section
112
are
focused
on
direct
toxic
effects,
not
on
secondary
effects
that
may
result
from
the
formation
of
ozone.
Substances
that
meet
the
HAP
listing
criteria
include
those
which
"
are
known
to
be,
or
may
reasonably
be
anticipated
to
be,

carcinogenic,
mutagenic,
teratogenic,
neurotoxic,
which
cause
reproductive
dysfunction,
or
which
are
acutely
or
chronically
toxic."
42
U.
S.
C.
§
7412(
b)(
2).
Thus,
the
fact
that
a
substance
may
be
7
an
ozone
precursor
is
not
relevant
to
the
decision
of
whether
it
should
be
listed
as
a
HAP.
The
Agency
implicitly
recognized
this
fact
by
removing
caprolactam,
which
is
a
VOC,
from
the
list
of
HAPs.
See
61
Fed.
Reg.
30,816
(
June
18,
1996).

As
a
practical
matter,
it
is
also
unnecessary
to
use
Section
112
of
the
Clean
Air
Act
to
regulate
VOC
emissions.
There
are
many
other
programs
under
the
Clean
Air
Act
that
are
specifically
designed
to
control
emissions
of
VOCs
and
other
ozone
precursors.
Under
Section
110
and
Part
D
of
the
Act,
any
state
that
does
not
meet
the
national
ambient
air
quality
standard
(
NAAQS)
for
ozone
must
adopt
a
state
implementation
plan
to
regulate
VOC
emissions
from
both
new
and
existing
sources.
In
addition,
VOC
emissions
are
regulated
under
Section
111
(
new
source
performance
standards)
and
Part
C
(
prevention
of
significant
deterioration).
In
light
of
the
other
programs
designed
specifically
to
control
VOC
emissions,
it
is
not
surprising
that
Congress
decided
that
VOCs
should
not
be
regulated
as
HAPs
unless
they
"
independently
meet"
the
listing
criteria
under
Section
112.

III.
DATA
ON
HEALTH
AND
ENVIRONMENTAL
EFFECTS
As
noted
above,
there
is
a
substantial
body
of
toxicological
literature
on
MIBK.

This
Section
of
the
Petition
first
reviews
the
potential
exposure
pathways
and
explains
why
inhalation
is
the
only
significant
route
of
human
exposure
resulting
from
MIBK
emissions.
The
Petition
then
reviews
the
literature
on
the
health
and
environmental
effects
of
MIBK,
and
demonstrates
that
MIBK
cannot
reasonably
be
anticipated
to
cause
adverse
health
effects
or
adverse
environmental
effects.
Finally,
the
Petition
explains
how
the
Panel
derived
an
RfC
of
2.4
mg/
m3
for
MIBK
based
on
EPA's
1994
guidance
for
setting
RfCs.
8
A.
Inhalation
Is
the
Only
Significant
Route
of
Human
Exposure
to
MIBK
Emissions
Section
112(
b)(
2)
indicates
that,
in
making
listing
decisions,
the
Agency
should
consider
whether
a
substance
may
reasonably
be
anticipated
to
cause
adverse
effects
"
through
inhalation
or
other
routes
of
exposure."
In
light
of
the
reasonably
anticipated
ambient
concentrations
(
described
in
Section
IV
of
the
Petition),
humans
would
not
be
expected
to
ingest
any
appreciable
amounts
of
MIBK
resulting
from
air
emissions.
Further,
because
of
MIBK's
relatively
rapid
biodegradation
and
volatilization
in
water
(
see
Section
III.
D),
it
is
highly
unlikely
that
humans
will
be
exposed
to
significant
amounts
of
MIBK
in
drinking
water.
In
addition,
given
its
lack
of
persistence
and
low
bioaccumulation
potential
(
also
described
in
Section
III.
D),
MIBK
emitted
to
the
air
would
be
unlikely
to
concentrate
in
food
sources.
Finally,
dermal
absorption
is
likely
to
be
insignificant
compared
to
inhalation,
both
because
dermal
absorption
is
a
less
efficient
exposure
route
to
humans
and
because
ambient
concentrations
of
MIBK
are
not
high
enough
to
make
this
route
toxicologically
relevant.
Thus,
it
is
clear
that
inhalation
is
the
only
route
of
human
exposure
with
potential
significance.

B.
MIBK
Cannot
Reasonably
Be
Anticipated
to
Cause
Adverse
Acute
Health
Effects
in
Humans
The
available
data
show
that
MIBK's
acute
toxicity
is
low.
The
concentration
levels
of
MIBK
that
are
likely
to
exist
beyond
facility
boundaries
are
several
orders
of
magnitude
below
levels
that
have
been
shown
to
cause
significant
adverse
acute
health
effects
in
animals
or
humans.
Thus,
MIBK
cannot
reasonably
be
anticipate
to
cause
acute
health
effects
in
humans.

MIBK's
acute
toxicity
in
animals
has
been
shown
to
be
very
low.
Smyth
et
al.

(
1951)
reported
that
the
lowest
lethal
concentration
for
rats
after
four
hours
of
exposure
was
4,000
ppm
MIBK;
the
rats
survived
4­
hr
exposures
to
2,000
ppm.
In
other
studies
with
rats,
9
exposures
of
21,000
ppm
were
lethal
within
53
minutes,
while
exposures
of
4,000
ppm
for
6
hours
caused
loss
of
coordination
and
prostration
but
not
death
(
Topping
et
al.
1994).

In
mice,
exposures
of
19,500
ppm
MIBK
resulted
in
anesthesia
within
30
minutes
and
rapid
recovery
upon
removal
to
fresh
air.
Concentrations
of
20,000
ppm
also
resulted
in
anesthesia,
but
deaths
occurred
after
30
minutes
of
exposure
(
Topping
et
al.
1994).
McOmie
and
Anderson
(
1949)
reported
that
no
mortality
was
seen
in
mice
exposed
to
vapors
of
MIBK
at
concentrations
of
24,000
ppm
for
15
minutes,
15,000
ppm
for
6
hours,
or
10,500
ppm
for
5
hours;
mortality
was
observed
after
exposures
of
20,000
ppm
for
30
minutes.
Zakhari
et
al.

(
1977)
reported
an
LC50
of
74.2
g/
m3
(
18,105
ppm)
for
mice
exposed
to
vapor
for
45
minutes.

Specht
(
1938)
and
Specht
et
al.
(
1940)
exposed
guinea
pigs
to
concentrations
of
1,000,
16,800,
and
28,000
ppm
MIBK.
Exposure
to
1,000
ppm
MIBK
for
24
hours
caused
little
or
no
irritation
of
the
nose
or
eyes.
The
guinea
pigs
showed
a
decreased
respiratory
rate
during
the
first
6
hours
of
exposure,
which
the
authors
attributed
to
low
grade
narcosis.
Exposure
to
16,800
ppm
caused
immediate
ocular
and
nasal
irritation
followed
by
ataxia
and
death
of
9
of
10
animals
within
6
hours
of
exposure.
Exposure
to
28,000
ppm
was
lethal
to
50%
of
the
guinea
pigs
within
45
minutes.

de
Ceaurriz
et
al.
(
1981,
1984)
examined
the
effects
of
acute
inhalation
exposure
of
MIBK
vapor
on
indicators
of
sensory
irritation
and
neurobehavioral
properties
in
mice.
In
acute
inhalation
studies
in
male
Swiss
OF1
mice
using
various
industrial
airborne
chemicals,
de
Ceaurriz
et
al.
(
1981)
measured
decreases
in
respiratory
rate
as
an
index
of
sensory
irritation.
For
MIBK
vapor,
the
concentration
that
induced
a
50%
decrease
in
respiratory
rate
in
the
mice
after
5
minutes
of
exposure
was
3195
ppm.
The
interpretation
of
this
finding,
however,
is
complicated
since
MIBK
may
also
produce
a
decrease
in
respiratory
rate
secondary
to
narcosis
(
Specht
et
al.
10
1940).
Therefore,
it
is
unclear
whether
the
reduction
observed
in
this
study
was
actually
the
result
of
narcosis
or
sensory
irritation.

In
a
second
short­
term
inhalation
study,
de
Ceaurriz
et
al.
(
1984)
exposed
mice
to
aliphatic
ketones
and
evaluated
neurobehavioral
effects
using
a
"
behavioral
despair"
swimming
test.
This
study
determined
that
exposure
of
mice
to
803
ppm
MIBK
vapor
for
4
hours
produced
a
decrease
by
50%
for
the
duration
of
immobility
in
the
3­
minute
forced
swimming
test.
A
4
hour
exposure
to
662
ppm
MIBK
decreased
the
duration
of
immobility
by
25%.
Interpretation
of
data
from
this
unconventional
test
designed
to
detect
the
efficacy
of
antidepressant
drugs
is
uncertain.

The
acute
toxicity
of
MIBK
has
been
studied
by
routes
other
than
inhalation
with
similar
results.
The
administration
of
MIBK
to
rodents
by
the
oral
and
dermal
routes
was
appreciably
lethal
at
only
very
high
dose
levels.
Smyth
et
al.
(
1951)
found
that
the
oral
LD50
of
MIBK
in
rats
was
2.08
g/
kg
when
administered
as
a
20%
emulsion
in
an
anionic
surfactant.

Batyrova
(
1973)
showed
that
the
oral
LD50
in
rats
and
mice
was
4.6
and
2.85
g/
kg,
respectively.

(
Topping
et
al.
1994)
reported
that
the
oral
LD50
of
MIBK
was
between
1.6
and
3.2
g/
kg
for
both
rats
and
guinea
pigs.
Union
Carbide
Corporation
(
1976)
reported
an
oral
LD50
in
the
rat
of
3.73
ml/
kg
(
2.98
g/
kg)
and
a
dermal
LD50
in
the
rabbit
of
greater
than
20
ml/
kg
(
16
g/
kg)
for
MIBK.

Human
data
also
indicate
low
acute
toxicity
of
MIBK.
Silverman
et
al.
(
1946)

exposed
12
people
to
MIBK
vapors
for
15
minute
periods
and
found
that
200
ppm
had
an
objectionable
odor
and
caused
eye
irritation;
100
ppm
was
the
highest
concentration
considered
acceptable
for
an
8
hour
exposure.

Elkins
(
1959)
reported
that
a
group
of
workers
exposed
to
approximately
100
ppm
while
water
proofing
boots
developed
headaches
and
nausea.
A
second
group
exposed
to
approximately
the
same
levels
at
a
different
facility
reported
only
respiratory
tract
irritation.
A
11
tolerance
was
said
to
be
acquired
during
the
work
week,
but
was
lost
over
the
weekend.
Most
of
the
effects
reported
at
100
ppm
were
not
noted
when
ventilation
reduced
the
MIBK
exposure
to
20
ppm
MIBK.
These
reports
are
described
in
a
single
paragraph
in
the
Elkins
reference.
No
details
or
documentation
are
provided
to
describe
the
calculation
or
measurement
of
exposure
levels,
duration
of
exposures,
range
of
exposure
levels,
number
of
people
exposed,
location
of
workers,
or
whether
the
reports
are
based
on
clinical
observations
or
merely
reflecting
subjective
responses.
These
reports
appear
to
be
personal
observations
by
the
author,
rather
than
a
careful
investigation
of
the
effects
of
exposure
to
MIBK.

Wayne
and
Orcutt
(
1960)
studied
the
eye
irritancy
potential
of
an
irradiated
mixture
of
MIBK
and
nitrogen
dioxide
using
human
volunteers.
Subjects
received
a
direct
ocular
exposure
to
the
reaction
products
formed
when
a
mixture
of
either
5
or
20
ppm
MIBK
and
1
ppm
nitrogen
dioxide
was
irradiated
for
1
hour
at
110
°
C
using
a
400
watt
mercury
vapor
lamp.
Under
these
conditions,
subjective
reports
of
eye
irritation
occurred
on
average
within
23
to
28
seconds
of
initiating
the
exposure
with
the
photochemical
reaction
products
in
the
mixture.
When
nonirradiated
control
mixtures
or
sunlight
irradiated
mixtures
were
used
to
conduct
the
exposures
there
were
few
reports
of
the
irritation
within
the
90
second
exposure
period.
The
latter
sets
of
conditions
were
judged
to
cause
negligible
or
only
a
very
slight
degree
of
eye
irritation
in
the
subjects.
An
increase
in
MIBK
concentration
in
the
reaction
mixture
did
not
cause
a
statistically
significant
reduction
in
the
response
time
because
of
the
large
amount
of
inter­
individual
variability
in
the
response.
Attempts
by
the
authors
to
relate
their
findings
to
smog­
induced
eye
irritation
in
urban
atmospheres
was
criticized
in
an
editorial
comment
published
at
the
end
of
the
article.
The
basis
for
the
criticism
was
the
difference
in
temperature
and
reaction
conditions
12
between
the
experimental
conditions
where
positive
effects
were
observed
and
the
ambient
conditions
that
exist
in
most
urban
environments.

Wigaeus­
Hjelm
et
al.
(
1990)
exposed
eight
male
volunteers
for
2
hours
to
10,
100,

and
200
mg/
m3
(
2.4,
24
and
48
ppm)
of
MIBK
or
to
a
combination
of
100
mg/
m3
of
MIBK
and
150
mg/
m3
of
toluene
on
four
separate
occasions.
All
participants
in
the
study
exercised
lightly
(
50
Watts)
during
the
exposures.
Two
neurobehavioral
tests
(
simple
reaction
time
and
an
arithmetic
test)
were
conducted
and
both
were
unaffected
by
the
MIBK
exposures.
Using
a
selfadministered
subjective
symptom
questionnaire
up
to
a
third
of
the
participants
in
the
study
reported
symptoms
of
irritation
to
the
nose
or
throat
or
headache
and
vertigo.
The
simple
nominal
scale
(
symptom
present/
not
present)
used
by
the
authors
to
record
the
prevalence
of
each
symptom
is
generally
regarded
as
the
least
valid
approach
for
collecting
and
recording
subjective
data
of
this
type
(
Berglund
and
Lindvall
1992).
Because
recent
studies
have
shown
that
odorous
volatile
chemicals,
such
as
MIBK,
are
capable
of
causing
false
symptom
reporting,
it
is
essential
for
studies
that
rely
on
subjective
methods,
such
as
symptom
questionnaires,
to
incorporate
a
nonirritating
odorant
control
to
account
for
the
non­
specific
responses
arising
from
an
unfamiliar
smell.
The
standardized
human
olfactory
threshold
for
MIBK
is
0.53
ppm
(
Devos
et
al.
1990).

In
a
follow­
up
study,
Iregen
et
al.
(
1993)
exposed
six
male
and
six
female
volunteers
to
10
and
200
mg/
m3
(
2.4
and
48
ppm)
of
MIBK
for
2
hours.
As
in
the
previous
study,
simple
reaction
time
and
arithmetic
ability
were
not
affected
by
the
MIBK
exposures.

Likewise,
there
were
no
effects
on
the
participants
heart
rate
or
mood.
A
17­
item
subjective
symptom
questionnaire
was
used
to
assess
irritancy
and
CNS
symptoms.
The
authors
did
not
describe
the
individual
questions
that
were
included
in
the
survey
or
how
the
answers
to
the
individual
questions
were
grouped
for
analysis.
Based
on
a
very
cursory
examination
with
poorly
13
described
methods,
the
authors
concluded
that
MIBK
produced
symptoms
of
irritation
and
discomfort.
The
authors
did
not
investigate
what
role
MIBK's
odor
had
in
evoking
a
positive
response
on
the
questionnaire.
Studies
have
shown
that
the
odor
from
a
chemical
can
be
a
highly
influential
factor
that
can
control
the
outcome
of
studies
that
rely
on
the
use
of
symptom
questionnaires
(
Cavalini
et
al.
1991;
Knasko
et
al.
1990).

Dick
et
al.
(
1992)
tested
subjects
for
neurobehavioral
performance
during
fourhour
exposures
to
100
ppm
MIBK
in
an
environmental
chamber.
Subjects
were
recruited
from
local
universities.
Neurobehavioral
effects
were
measured
using
a
series
of
five
psychomotor
tests,
one
sensorimotor
test,
and
a
test
of
mood.
Additionally,
subjects
were
asked
to
complete
a
questionnaire
in
which
"
yes/
no"
responses
were
required
to
a
series
of
questions
addressing
perceived
irritation.
A
total
of
seventeen
subjects
were
exposed
to
MIBK
as
part
of
a
larger
study
in
which
additional
subjects
were
exposed
to
50
ppm
MIBK
together
with
100
ppm
MEK,
200
ppm
MEK,
or
a
placebo
atmosphere.
Although
sensory
and
irritant
effects
were
noted
during
exposure,
no
interpretable
statistically
significant
performance
effects
were
identified
which
could
be
attributed
to
exposure
to
MIBK
(
or
a
combination
of
MIBK
and
MEK
or
MEK
alone).
While
70%
of
the
MIBK­
exposed
subjects
reported
a
strong
odor
during
exposure,
frank
irritation
(
throat
irritation,
headache,
tearing,
or
nausea)
was
reported
by
only
approximately
20­
30%
of
the
subjects.
Subsequent
statistical
analysis
of
the
subjective
effects
revealed
a
significant
difference
between
MIBK­
exposed
subjects
and
controls
only
with
respect
to
the
identification
of
a
strong
odor.
These
results
are
consistent
with
the
study
of
Wigaeus­
Hjelm
et
al.
(
1990)
and
Iregen
et
al.
(
1993),
which
also
failed
to
detect
neurobehavioral
effects
in
subjects
exposed
to
MIBK.
14
Gagnon
et
al.
(
1994)
exposed
four
healthy
subjects
for
7
hours
to
20
or
40
ppm
of
MIBK
and
examined
perceived
odor
intensity
and
the
frequency
of
subjective
symptom
reporting
during
and
after
the
exposure.
The
subjects
were
shown
to
rapidly
accommodate
to
the
odor
of
MIBK
after
entering
the
chamber
with
the
perceived
odor
intensity
(
subjective)
decreasing
and
odor
perception
threshold
(
objective)
increasing
as
adaptation
occurred.
The
odor
perception
thresholds
declined
towards
normal
levels
when
measured
55­
min
and
95­
min
following
the
exposure.
The
authors
failed
to
discuss
the
results
or
findings
from
the
subjective
symptom
questionnaire
in
any
detail.
It
is
important
to
note
that
the
adaptation
observed
in
this
study
is
a
well
known
sensory
phenomenon
that
is
relatively
specific
to
each
chemical
and
seldom
permanent
(
Cain
1970).

Linari
et
al.
(
1964)
presented
the
results
from
an
occupational
health
survey
involving
19
individuals
who
were
exposed
to
MIBK
for
20
to
30
minutes
a
day
for
up
to
1
year
at
concentrations
ranging
from
80
to
500
ppm.
In
a
companion
study
performed
5
years
after
the
original
investigation,
14
out
of
19
employees
were
reexamined
for
signs
and
symptoms
of
solvent
overexposure
(
Armeli
et
al.
1968).
Both
the
original
study
and
the
follow­
up
contain
many
serious
procedural
and
technical
flaws
that
severely
limit
their
usefulness
and
reliability.
Although
the
authors
attribute
many
of
their
findings
to
MIBK,
they
acknowledge
that
other
chemicals
were
used
in
what
appeared
to
be
a
chemical
isolation
process,
since
it
involved
a
final
centrifugation
step.
Few
details
were
given,
however,
on
the
nature
of
the
process,
the
other
chemical
exposures
involved,
the
monitoring
methods
for
MIBK,
or
the
extent
of
alcohol
consumption
and
cigarette
smoking
by
the
employees
involved
in
the
study.
4
4
A
wide
range
of
clinical
tests
was
performed
with
each
of
the
employees
in
the
study,
including
an
electrocardiogram,
a
chest
x­
ray,
liver
function
tests,
urinalysis,
serum
protein
15
The
American
Conference
of
Governmental
Industrial
Hygienists
(
ACGIH)
has
recommended
a
Threshold
Limit
Value
(
TLV)
for
MIBK
of
50
ppm
(
8­
hour
TWA).
The
current
OSHA
Permissible
Exposure
Limit
(
PEL)
for
MIBK
is
100
ppm.

In
summary,
human
exposure
to
high
concentrations
of
MIBK
in
the
range
of
several
thousand
ppm
may
result
in
narcosis,
which
is
readily
reversible
on
exposure
to
fresh
air;

exposures
in
the
range
of
100
to
several
hundred
ppm
may
be
irritating
to
the
eyes
and
nose;

exposures
below
100
ppm
appeared
to
be
tolerated
well,
but
are
easily
identified
by
the
characteristic
odor
of
MIBK.
Levels
below
0.1
ppm
are
below
the
odor
detection
level
of
the
human
nose
(
Verschueren
1977).
As
discussed
below,
ambient
air
levels
beyond
industrial
site
boundaries
as
a
result
of
continuous
or
frequently
recurring
releases
are
well
below
levels
that
would
raise
any
concerns
about
acute
human
health
effects.

C.
MIBK
Cannot
Reasonably
be
Anticipated
to
Cause
Adverse
Chronic
Health
Effects
In
Humans
There
is
a
considerable
body
of
data
which
shows
that
MIBK
does
not
cause
chronic
health
effects
in
humans
at
reasonably
anticipated
ambient
levels.

electrophoresis,
clinical
chemistry,
and
hematology
tests.
Except
for
an
occasional
high
or
low
reading,
the
authors
did
not
find
any
consistent
changes
in
their
battery
of
objective
tests
that
suggested
any
systemic
toxicity.
A
variety
of
subjective
health
complaints
were
noted,
however,
when
the
employees
were
given
a
health
questionnaire.
The
type
of
questionnaire
used
in
study
and
the
method
of
administration
were
not
given.
Subjective
health
complaints
included
gastrointestinal
functional
disorders,
dizziness,
headaches,
sensory
irritation,
and
muscle
weakness.
These
findings
are
highly
suspect,
however,
since
the
authors
failed
to
include
any
control
group
in
their
study.
In
addition,
the
authors
noted
that
respiratory
protection
was
mandatory
in
the
work
area,
so
it
is
uncertain
to
what
extent
the
employees
were
actually
exposed
to
MIBK.
It
is
essential
that
any
study
relying
on
the
use
of
a
subjective
symptom
questionnaire
also
include
a
reference
group
for
comparative
purposes,
because
many
subjective
complaints
of
the
type
observed
in
this
study
also
frequently
appear
in
unexposed
control
populations.
The
absence
of
a
control
group,
the
failure
to
consider
lifestyle
risk
factors,
and
the
acknowledged
co­
exposure
to
other
unmentioned
industrial
chemicals
are
all
serious
problems.
Therefore,
the
results
of
this
study
should
not
be
used
in
any
hazard,
exposure,
or
risk
determination
for
MIBK.
16
1.
Subchronic
Studies
MacEwen
et
al.
(
1971)
continuously
exposed
rats,
mice,
dogs,
and
monkeys
to
100
ppm
MIBK
for
two
weeks.
Animals
were
monitored
for
changes
in
body
weight,
organ
to
body
weight
ratios,
hematology,
clinical
chemistry,
pathology,
blood
gases,
blood
pH,

spontaneous
motor
activity,
and
EEG
(
one
monkey).
No
significant
signs
of
toxicity
were
observed
during
the
exposures
and
no
changes
in
EEG
activity
were
observed
absolute
and
relative
to
body
weight.
Kidney
weights
were
statistically
higher
for
the
rats
but
not
for
any
other
species,
and
rats
showed
a
slight
indication
of
reduced
growth.
The
rat
kidneys
showed
histological
changes
in
the
proximal
tubule
epithelium.
The
study
was
repeated
at
200
ppm
MIBK
and
the
changes
observed
at
100
ppm
were
also
observed
at
200
ppm
and,
in
addition,

absolute
and
relative
liver
weights
were
slightly
elevated.

Based
on
these
data,
a
90­
day
study
was
conducted
exposing
male
rats,
dogs,
and
monkeys
to
100
ppm
MIBK
continuously
under
reduced
atmospheric
pressure
which
simulated
space
cabin
conditions
(
MacEwen
et
al.
1971).
A
group
of
rats
was
also
used
to
study
recovery
from
MIBK
effects.
The
only
effects
observed
were
in
male
rats
which
showed
increased
liver
and
kidney
weights
and
the
presence
of
hyalin
droplets
in
proximal
renal
tubule
epithelial
cells.

Rats
removed
from
MIBK
exposure
after
two
weeks
showed
a
gradual
return
to
normal
structure
by
60
days
post
exposure,
while
those
exposed
for
90
days
also
exhibited
recovery
but
at
a
slower
rate.
Hyalin
droplets
were
not
associated
with
depressed
growth
or
any
indication
of
illness,
and
dogs
and
monkeys
showed
no
adverse
effects
related
to
MIBK
exposures.
Based
on
the
data,
the
authors
concluded
that
(
i)
100
ppm
of
MIBK
should
be
tolerable
by
man
for
90
days
and
(
ii)
the
60
minute
emergency
limit
of
100
ppm
and
the
90­
and
17
1,000­
day
provisional
limit
of
20
ppm
established
by
the
Space
Science
Board,
NAS/
NRC,

contained
a
wide
margin
of
safety.

As
part
of
a
testing
program
under
Section
4
of
the
Toxic
Substances
Control
Act,

the
Ketones
Panel
conducted
two­
and
14­
week
inhalation
studies
using
Fisher­
344
rats
and
B6C3F1
mice
(
Phillips
et
al.
1987).
In
the
two­
week
study,
rats
and
mice
(
males
and
females)

were
exposed
to
0,
100,
500
or
2,000
ppm
MIBK
for
6
hours/
day.
Male
rat
livers
were
slightly
heavier
in
the
2,000
ppm
group
and
regenerative
tubular
epithelium
and
hyalin
droplets
were
observed
in
the
kidneys
of
male
rats
from
the
500
and
2,000
ppm
groups.
A
low
incidence
of
lethargy
and
lacrimation
was
noted
in
both
rats
and
mice
at
the
2,000
ppm
exposure
concentrations.

The
14­
week
study
was
conducted
at
0,
50,
250
and
1,000
ppm
MIBK
vapors
6
hours/
day,
5
days
per
week.
At
1,000
ppm,
livers
from
male
rats
and
mice
were
slightly
heavier
than
control
livers.
Liver
weights
were
also
slightly
heavier
than
controls
for
the
250
ppm
male
mice.
Morphological
changes
were
not
associated
with
the
increase
in
liver
weights.
An
increased
incidence
and
extent
of
hyalin
droplets
were
seen
microscopically
in
the
proximal
renal
tubule
epithelium
of
male
rats
exposed
to
250
or
1,000
ppm
MIBK.
Control
kidneys
had
similar
droplets
but
to
a
lesser
degree.
The
increase
in
hyalin
droplets
was
not
associated
with
changes
in
kidney
function.
The
significance
of
renal
hyalin
droplet
formation
in
male
rats
is
uncertain
since
it
is
not
observed
in
female
rats,
castrated
male
rats
or
males
of
other
species.
Droplet
formation
occurs
in
normal
male
rats
and
appears
to
be
related
to
excretion
of
a
rat­
specific
protein
2­
microglobulin
by
the
kidneys
(
Alden
et
al.
1984).
There
is
no
evidence
that
humans
are
susceptible
to
a
chemically­
induced
protein
droplet
nephropathy
and
subsequent
renal
disease
similar
to
that
which
occurs
in
male
rats
(
Hard
et
al.
1993).
18
In
1985,
EPA
reviewed
the
Phillips
study
and
concluded
that
there
was
"
an
effect
(
increase
in
liver
weights)
but
no
conclusive
toxicity
under
the
current
condition"
(
Turner
1985
at
p.
4).
According
to
Turner,
"
no
gross
or
microscopic
hepatic
lesions
related
to
MIBK
exposure
were
seen.
There
were
no
microscopic
findings
in
any
of
the
tissues
examined
which
could
be
related
to
MIBK
exposures."
Id.
at
p.
1.
Dr.
I.
Baumel,
Director
of
EPA's
Health
and
Environmental
Review
Division
(
HERD),
concurred
with
Turner's
analysis:
"
We
agreed
that
there
was
no
significant
toxicity
seen
at
the
concentrations
tested"
(
Baumel
1985)
5.

Liver
and
kidney
effects
were
also
observed
following
repeated
oral
gavage
exposure
of
rats
to
MIBK
(
Microbiological
Associates,
1986
or
Levine
et
al.
1987).
Sprague­

Dawley
rats
given
13
weeks
of
oral
gavage
administration
of
MIBK
at
dosages
of
0,
50,
250
or
1,000
mg/
kg
daily
were
evaluated
for
effects
on
body
weight,
feed
consumption,
organ
weight,

morbidity,
clinical
chemistry,
hematology
and
histopathology.
Exposure
to
1,000
mg/
kg
day
produced
a
reduction
in
body
weight
gain
and
evidence
of
lethargy
following
dosing.
Kidney
weights
were
increased
as
were
BUN
(
blood
urea
nitrogen
concentration),
urinary
protein
and
ketones,
serum
potassium
levels,
and
the
incidence
of
nephrosis
in
male
and
female
rats.
Liver
weights
and
enzymes
also
were
increased
in
both
sexes,
but
in
the
absence
of
corresponding
histopathological
lesions.
Exposure
to
250
mg/
kg
day
produced
similar
but
less
pronounced
effects
to
the
liver
and
kidney,
and
exposure
to
50
mg/
kg
day
produced
no
effects
to
the
rats.

5
In
reviewing
the
same
study
in
1988,
EPA
found
a
NOAEL
of
50
ppm
based
on
increased
liver
weights
in
rats
and
mice.
EPA's
reviewer
stated,
"
No
alterations
in
liver
weights
were
noted
from
female
rats
or
mice.
Gross
and
histopathological
examination
revealed
no
treatment­
related
lesions
in
either
species.
Based
on
the
elevations
in
urinary
glucose
and
protein,
the
NOAEL
for
this
study
appears
to
be
50
ppm..."
Griffin
(
1988)
at
p.
7.
However,
evidence
of
liver
and
kidney
enlargement
in
an
animal
study,
without
any
evidence
of
pathological
damage,
does
not
demonstrate
an
adverse
effect
in
the
animals,
and
is
of
no
use
in
a
human
health
hazard
evaluation.
19
Differences
between
the
study
results
for
the
90
day
inhalation
and
the
oral
gavage
study
could
be
related
to
differences
in
kinetics
of
oral
bolus
dose
versus
continuous
inhalation
exposure
or
differences
in
the
strain
of
rat
employed.
The
continuous
inhalation
exposure
scenario
is
considered
more
relevant
to
potential
human
chronic
exposure.

Batyrova
(
1973)
reported
that
20
to
30
ppm
of
MIBK
4
hours/
day
for
4.5
months
interfered
with
detoxification
by
the
liver
and
increased
the
eosinophil
counts
in
rats.
The
reliability
of
this
study,
however,
is
questionable.
Subsequent
repeated
exposure
studies
(
Phillips
et
al.
1987;
Microbiological
Associates,
1986)
have
not
reproduced
any
MIBK
associated
effects
on
hematological
parameters.

In
summary,
the
principal
health
effects
observed
in
the
above
subchronic
studies
were
an
increase
in
liver
and
kidney
weights
and
in
hyaline
droplets
in
the
kidneys
of
male
rats.

HERD
has
determined
that
these
effects
do
not
indicate
significant
toxicity.
No
gross
or
microscopic
hepatic
lesions
related
to
MIBK
exposure
were
seen
at
any
concentrations
or
in
any
species
tested.

2.
Mutagenicity
In
general,
MIBK
exposure
has
not
been
associated
with
genotoxicity
in
vitro
or
in
vivo.
The
testing
program
under
TSCA
Section
4
also
involved
a
battery
of
genotoxicity
tests,

including
the
Salmonella/
mammalian
microsome
(
Ames)
assay,
the
L5178Y
TK+/­
mouse
lymphoma
assay,
the
unscheduled
DNA
synthesis
in
rat
primary
hepatocyte
assay,
the
micronucleus
cytogenetic
assay
in
mice,
and
the
BALB
3T3
mouse
embryo
cell
transformation
assay.
The
results
of
these
assays
indicate
that
MIBK
has
very
little,
if
any,
mutagenic
activity
(
O'Donoghue
et
al.
1988).
20
Brooks
et
al.
1988
also
tested
MIBK
for
genotoxic
activity
in
bacterial
mutagenicity
assays
using
Salmonella
typhimuriumi
and
Escherichia
coli
WP2
and
WP2uvr
A,
a
yeast
cell
(
Saccharomyces
cerevisiae
JD1)
mitotic
gene
conversion
assay,
and
in
in
vitro
assays
for
induction
of
chromosome
damage
in
cultured
rat
liver
cells
or
in
Chinese
hamster
ovary
cells.

The
bacterial
and
yeast
assays
were
tested
both
in
the
presence
and
in
the
absence
of
rat
liver
metabolic
activation
system
(
S9
fraction).
MIBK
gave
a
negative
response
in
all
of
these
genotoxicity
tests.

3.
Carcinogenicity
MIBK
has
not
been
tested
specifically
for
carcinogenicity.
Data
on
its
structure
and
metabolism,
subchronic
health
effects,
and
the
genotoxicity
studies
reviewed
above
indicate
that
MIBK
is
not
likely
to
have
carcinogenic
properties.

MIBK
neither
belongs
to
a
class
of
chemicals
known
to
react
with
DNA
nor
is
metabolized
to
materials
which
are
likely
to
react
with
DNA.
Materials
which
are
oncogenic
for
mammals
appear
to
cause
cancer
either
by
interacting
with
the
genetic
material
(
DNA)
(
that
is,

they
are
genotoxic
and,
therefore,
are
probably
initiators
of
the
carcinogenic
process),
or
they
produce
chronic
toxic
effects
which
result
in
increased
cell
turnover
and,
therefore,
produce
effects
by
epigenetic
mechanisms
and
are
probably
promoters
of
the
carcinogenic
process.
The
data
available
for
MIBK
indicate
that
this
chemical
substance
is
not
genotoxic
and
does
not
produce
significant
chronic
toxicity.
Accordingly,
MIBK
is
unlikely
to
be
carcinogenic
by
either
genetic
or
epigenetic
mechanisms,
and
is
thus
unlikely
to
be
either
an
inducer
or
promoter
of
carcinogenicity.
21
4.
Reproductive
and
Developmental
Toxicity
Phillips
et
al.
(
1987)
studied
the
effects
of
exposures
of
0,
50,
250,
or
1,000
ppm
MIBK
vapor
given
6
hours/
day,
5
days/
work
for
14
weeks
on
the
morphology
of
male
and
female
reproductive
organs
in
mice
and
rats.
No
effects
were
found
in
testes
weights
or
the
histology
of
the
testes,
epididymides,
prostate
gland,
seminal
vesicles,
ovaries,
uterus,
oviducts,
vagina,
cervix,

and
mammary
glands.

The
testing
program
sponsored
by
the
Panel
under
TSCA
Section
4
included
developmental
toxicity
studies
in
rats
and
mice
(
Tyl
et
al.
1987).
Timed­
pregnant
CD­
1
mice
and
Fischer
344
rats
were
exposed
to
MIBK
vapors
by
inhalation
on
gestational
days
6
through
15
at
concentrations
of
0,
300,
1,000
or
3,000
ppm.
The
animals
were
sacrificed
on
gestational
day
21
(
rats)
or
18
(
mice),
and
live
fetuses
were
examined
for
external,
visceral
and
skeletal
alterations.

In
mice,
exposure
to
3,000
ppm
resulted
in
maternal
toxicity
(
apparent
exposure­
related
increases
in
deaths
(
12.0%)
and
clinical
signs),
increased
absolute
and
relative
liver
weight,
and
fetotoxicity
(
increased
incidence
of
dead
fetuses,
reduced
fetal
body
weight
per
litter
and
reductions
in
skeletal
ossification).
No
treatment­
related
embryotoxicity
was
seen.
No
treatment­
related
increases
in
fetal
malformations
were
seen
at
any
exposure
concentration
tested.
There
was
no
evidence
of
treatment­
related
maternal,
embryo,
or
fetal
toxicity
(
including
malformations)
at
300
or
1,000
ppm.

In
rats,
exposure
to
3,000
ppm
resulted
in
maternal
toxicity
(
clinical
signs,

decreased
food
consumption,
and
decreased
body
weight
and
body
weight
gain),
increased
relative
kidney
weight
and
fetotoxicity
(
reduced
fetal
body
weight
per
litter
and
reductions
in
skeletal
ossification).
No
increase
in
fetal
malformations
was
observed
in
any
exposure
group
relative
to
controls.
At
300
and
1,000
ppm
there
was
no
maternal,
embryo,
or
fetal
toxicity
22
(
including
malformations).
Reduced
fetal
body
weight
was
observed
in
rats
at
300
ppm,
but,
as
explained
below,
this
apparent
finding
was
confounded
by
litter
size
and
should
not
be
considered
treatment­
related.

The
reduction
in
fetal
body
weights
seen
in
the
rat
at
300
ppm
is
an
artifact
resulting
from
the
fact
that
the
litters
in
the
300
ppm
group
contained
more
fetuses
than
the
controls.
In
1985,
the
Agency
evaluated
these
data
and
reached
the
same
conclusion.

Specifically,
the
Agency
stated:
"
The
data
show
that,
in
the
rat
and
the
mouse,
MIBK
causes
significant
developmental
effects
(
fetal
death,
reduced
fetal
weight,
delayed
ossification)
in
the
conceptus
at
the
high
dose
tested
only
(
3,000
ppm).
No
effects
were
noted
at
lower
doses
(
1,000,
300,
0
ppm)....
The
NOEL
derived
from
the
data
is
1,000
ppm."
(
Letter
of
Ottley
to
Kariya,
1/
18/
85).
6
In
summary,
the
weight
of
the
evidence
indicates
that
MIBK
does
not
cause
any
significant
reproductive
or
developmental
toxicity
in
mice
or
rats
exposed
to
1,000
ppm
MIBK
vapors
by
inhalation.

5.
Neurotoxicity
Numerous
studies
have
been
conducted
to
assess
the
neurotoxic
potential
of
MIBK.
These
studies
show
that
MIBK,
like
many
other
solvents,
causes
transient
pharmacologic
6
In
1988,
the
Agency
concluded
that
"
it
seems
most
prudent
to
...
use
300
ppm
as
the
probable
LOAEL
for
developmental
toxicity."
EPA
Memorandum
from
M.
Campbell
to
E.
Dage,
Chemical
Review
and
Evaluation
Branch
(
November
1,
1988)
at
p.
2.
The
Agency,
however,
did
not
state
why
it
disagreed
with
its
previous
determination
to
disregard
the
effects
seen
in
rats
only
at
300
ppm.
The
Panel
believes
the
Agency's
interpretation
in
1985
was
correct.
The
questionable
finding
in
the
rat
at
300
ppm,
which
did
not
arise
in
the
rat
at
1,000
ppm
or
in
the
mouse
at
either
dose,
clearly
is
not
sufficient
to
establish
that
MIBK
can
reasonably
be
anticipated
to
cause
serious
or
irreversible
developmental
toxicity
in
humans.
23
effects
at
high
exposures.
The
studies
do
not
demonstrate
that
MIBK
produces
nervous
system
damage,
even
following
repeated
exposure
at
relatively
high
concentrations.

Three
neurobehavioral
studies
in
humans
following
acute
exposure
at
levels
ranging
from
2
to
100
ppm
have
already
been
described
(
see
Section
III.
B,
pp.
12­
13).
The
investigators
did
not
detect
neurobehavioral
effects
in
the
test
subjects
in
any
of
the
studies.

The
study
by
Phillips
et
al.
(
1987),
described
in
Section
III.
C.
1
(
subchronic
studies)
found
no
evidence
of
neurotoxicity
in
mice
or
rats
exposed
to
0,
50,
250
or
l,
000
ppm
MIBK,
6
hours/
day,
5
days/
week
for
14
weeks.
No
signs
of
neurotoxicity
were
observed
clinically
and
histologic
examinations
of
the
brain,
spinal
cord,
and
peripheral
nerves
were
normal.

The
Ketones
Panel
recently
sponsored
a
schedule­
controlled
operant
behavior
(
SCOB)
study
in
rats
under
an
enforceable
consent
agreement
under
TSCA
(
Bernard
and
David
1996).
The
study
consisted
of
two
sets
of
animals,
male
Sprague­
Dawley
(
SD)
rats
restricted
to
13­
18
g
of
feed
per
day
and
used
for
SCOB
testing
and
male
ad
libitum­
fed
SD
rats
used
to
evaluate
systemic
toxicity.
Both
sets
of
animals
were
exposed
to
vapor
concentrations
of
0,
250,

750,
or
1500
ppm
of
MIBK
over
a
13­
week
period.

Animals
were
observed
for
signs
of
toxicity
prior
to
exposure,
once
per
hour
during
exposure,
and
30
minutes
to
one
hour
after
exposure.
Animals
exposed
to
1500
ppm
exhibited
transient
reduced
activity
of
minimal
to
minor
severity
during
exposure
for
Weeks
1
to
10
only.
The
severity
of
reduced
activity
during
exposure
decreased
over
the
course
of
the
study.

No
signs
of
reduced
activity
were
observed
immediately
after
exposure.
Animals
exposed
to
750
ppm
exhibited
transient
reduced
activity
of
minimal
severity
during
the
exposure
period
for
Weeks
1
to
8
only.
The
severity
of
reduced
activity
during
exposure
was
unchanged
over
the
course
of
8
24
weeks.
No
signs
of
reduced
activity
were
observed
immediately
after
exposure.
No
reduction
in
activity
was
observed
during
exposure
for
the
control
or
250
ppm
animals.

The
mean
absolute
liver
and
kidney
weights
of
all
ad
libitum­
fed,
test
substanceexposed
groups,
and
the
relative
(
to
body
weight)
liver
and
kidney
weights
for
the
750
and
1500
ppm
ad
libitum­
fed
groups
were
statistically
higher
(
p
<
0.05)
when
compared
to
the
ad
libitumfed
control
group.
Mean
terminal
body
weights
for
the
1500
ppm
SCOB
group
were
statistically
higher
(
p
<
0.05)
than
for
the
SCOB
control
group.
The
mean
absolute
liver
weights
for
the
750
and
1500
ppm
SCOB
groups,
and
the
mean
relative
(
to
body
weight)
liver
weights
for
the
250
and
750
ppm
groups
were
statistically
higher
(
p
<
0.05)
when
compared
to
the
SCOB
control
group.
No
other
terminal
body
weight
or
organ
weight
differences
were
observed
for
either
the
ad
libitum­
fed
or
the
SCOB
animals.

SCOB
testing
occurred
daily
in
feed­
restricted
male
rats
during
Weeks
1­
13
of
exposure
and
Weeks
14
and
15
following
the
cessation
of
exposure.
Fixed­
ratio
(
FR)
running
rate,
FR
pause
duration,
fixed­
interval
(
FI)
response
rate,
and
index
of
curvature
values
for
each
animal
were
calculated
as
percent
of
baseline
activity
and
the
percents
from
Weeks
4,
8,
13,
and
15
were
compared
across
groups.
No
significant
differences
were
seen
in
SCOB
values
at
any
test
vapor
concentration,
and
there
was
no
apparent
change
in
activity
after
cessation
of
exposure.

No
exposure­
related
changes
were
detected
during
gross
necropsy
examinations
of
ad
libitumfed
or
SCOB
male
rats
exposed
to
the
test
substance.
No
tissues
were
processed
for
microscopic
examination.

Spencer
et
al.
(
1975)
exposed
rats
to
1,300
ppm
of
methyl
n­
butyl
ketone
(
MnBK)

for
4
months
or
1,500
ppm
MIBK
for
5
months.
While
MnBK
produced
central­
peripheral
distal
axonopathy,
only
minimal
distal
axonal
changes
were
seen
in
rats
exposed
to
MIBK.
The
axonal
25
pathology
seen
after
MIBK
exposure
may
have
been
due
to
the
3
percent
MnBK
which
was
contained
in
the
MIBK
or
more
likely
was
due
to
a
neuropathy
induced
by
the
wire
mesh
caging
the
rats
were
housed
in.

Spencer
and
Schaumburg
(
1976)
gave
cats
subcutaneous
injections
of
150
mg/
kg
MIBK
or
MIBK/
methyl
ethyl
ketone
(
9/
l)
twice
daily,
five
days/
week
for
up
to
8.5
months.
No
nervous
system
damage
was
found
following
these
exposures.
Dogs
were
similarly
given
subcutaneous
injections
of
150
mg/
kg
MIBK,
twice
a
day
for
11
months
without
producing
electromyographic
changes
or
other
evidence
of
neurotoxicity.
(
Topping
et
al.
1994).

Rats
were
given
intraperitoneal
injections
of
MIBK
or
a
mixture
of
MIBK/
methyl
ethyl
ketone
(
9/
1)
five
times
a
week
for
35
weeks.
Dose
levels
of
10,
30
or
100
mg/
kg
were
increased
to
20,
60
or
200
mg/
kg
after
two
weeks
of
exposure.
Except
for
body
weight
suppression
after
3­
4
weeks
of
exposure,
the
only
other
effect
noted
was
transient
anesthesia
during
the
first
month
of
treatment
in
the
200
mg/
kg
animals
(
Topping
et
al.
1994).

Garcia
et
al.
(
1978)
reported
that
2
of
7
rats
exposed
to
25
ppm
MIBK
vapor
exhibited
increases
in
lever­
pressing
response
rates.
The
significance
of
this
finding
is
unclear
since
statistical
analyses
of
the
data
were
not
conducted
and
the
single
rat
exposed
to
50
ppm
did
not
exhibit
an
increase
in
response
rate.

Geller
et
al.
(
1979­
a)
describes
operant
behavior
studies
conducted
on
an
undetermined
number
of
rats
using
the
same
procedures
as
in
Garcia
et
al.
(
1978).
Data
are
presented
for
one
rat
only,
which
was
exposed
to
25
ppm
MIBK.
In
this
animal,
the
response
rate
increased
to
58%
over
the
control
level.
Seven
days
post
exposure,
the
response
rate
remained
16%
above
the
control
level.
The
presentation
of
data
for
a
single
rat
makes
this
report
very
difficult
to
interpret.
26
Geller
et
al.
(
1979­
a)
also
describes
operant
behavior
tests
on
four
baboons
exposed
to
MIBK
at
25,
35,
50
and
75
ppm
for
1
week.
In
the
presence
of
50
ppm
MIBK,
one
baboon
had
an
increased
response
rate
during
the
delay
interval
on
all
five
days
the
tests
were
conducted.
The
increased
response
rate,
however,
was
not
observed
uniformly
throughout
the
2­

hour
test
session,
and
an
increase
in
the
MIBK
concentration
to
75
ppm
for
an
additional
48
hours
did
not
alter
the
response
rate.
The
authors
speculated
that
differences
in
response
rates
were
due
to
anxiety
levels
in
the
test
animals.
There
is
no
indication
that
the
data
were
reproducible
and
in
fact,
the
results
from
the
one
baboon
with
an
increased
response
rate
at
50
ppm
actually
contradict
the
results
(
decreased
response
rate)
seen
at
50
ppm
in
a
subsequent
study
by
Geller
et
al.
(
1979­
b)
described
below.

In
a
similar
study
also
using
four
baboons,
Geller
et
al.
(
1979­
b)
reported
delayed
behavioral
response
times
in
baboons
exposed
to
50
ppm
MIBK
alone,
but
no
alteration
of
response
was
seen
when
MIBK
was
combined
with
100
ppm
methyl
ethyl
ketone.
Variability
in
the
response
of
the
baboons
was
high
and
the
profiles
of
the
responses
over
time
and
among
the
test
subjects
were
dissimilar.
Furthermore,
the
pattern
of
the
response
time
measured
between
exposures
demonstrated
no
persistence
of
any
effects.
A
variety
of
mechanisms,
all
speculative,

were
proposed
by
Geller
to
account
for
these
findings.
It
should
be
noted
that
this
study
used
only
one
concentration
of
each
substance
or
combination
of
substances,
no
statistics
were
conducted,
and
the
effects
were
fully
reversible.

MacEwen
et
al.
(
1971)
reported
no
EEG
changes
in
monkeys
(
two)
exposed
continuously
to
100
or
200
ppm
MIBK
for
two
weeks.
Spontaneous
activity
measurements
were
also
unaffected
by
MIBK
exposure,
but
details
of
the
experimental
procedure
were
not
provided.
27
In
a
short­
term
inhalation
study,
de
Ceaurriz
et
al.
(
1984)
exposed
mice
to
aliphatic
ketones
and
evaluated
neurobehavioral
effects
using
a
"
behavioral
despair"
swimming
test.
This
study
determined
that
exposure
of
mice
to
803
ppm
MIBK
vapor
for
4
hours
produce
a
decrease
by
50%
for
the
duration
of
immobility
in
the
3­
minute
forced
swimming
test.
A
4
hour
exposure
to
662
ppm
MIBK
decreased
the
duration
of
immobility
by
25%.
Interpretation
of
data
from
this
unconventional
test,
which
is
designed
to
detect
the
efficacy
of
antidepressant
drugs,
is
uncertain.

Abou­
Donia
et
al.
(
1985­
a)
studied
the
effects
of
MIBK,
n­
hexane,
and
combinations
of
the
two
solvents
on
induction
of
neurotoxicity
in
hens.
Continuous
exposures
to
1,000
ppm
MIBK
for
90
days
followed
by
a
30­
day
observation
period
resulted
in
leg
weakness
with
recovery
but
no
evidence
of
axonal
damage.
Simultaneous
exposures
to
MIBK
and
nhexane
resulted
in
potentiation
of
n­
hexane
neurotoxicity
apparently
by
induction
of
hepatic
p­
450
enzymes.

Batyrova
(
1973)
reported
that
20
to
30
ppm
of
MIBK
administered
4
hours/
day
for
4.5
months
produced
disturbances
in
the
conditioned
reflexes
of
rats.
The
reliability
of
this
report,
however,
is
questionable.
In
a
limited­
design
drinking
water
study,
MIBK
administered
to
female
rats
at
a
dosage
of
1
g/
kg
day
for
120
days
did
not
produce
adverse
effects
on
the
nervous
system
(
CMIR
1977).

Two
in
vitro
studies
have
been
conducted
for
MIBK.
Selkoe
et
al.
(
1978)
studied
the
effects
of
methyl
n­
butyl
ketone,
n­
hexane,
25­
hexanedine,
methyl
ethyl
ketone,
and
MIBK
on
an
in
vitro
model
used
to
study
neurotoxins
which
produce
neurofilamentous
hyperplasia.
The
model
system
was
a
murine
neuroblastoma
cell
line
which
develops
cytoplasmic
filamentous
hyperplasia
when
exposed
to
aluminum
ions.
In
this
model,
methyl
n­
butyl
ketone
and
n­
hexane
28
induced
a
highly
reproducible
series
of
cytotoxic
effects
and
adversely
affected
extension
or
maintenance
of
neuritic
processes.
In
contrast
to
these
effects,
MIBK
did
not
produce
cytopathological
changes
analogous
to
methyl
n­
butyl
ketone
or
n­
hexane
and
did
not
produce
morphologic
changes
in
neuroblastoma
cells
at
concentrations
which
routinely
produced
widespread
cytoplasmic
and
nuclear
changes
when
methyl
n­
butyl
ketone
was
present
in
the
culture
medium.

Huang
et
al.
(
1993)
studied
MIBK
and
other
monoketones
(
carbon
chain
length
from
3
to
10)
for
their
in
vitro
effects
on
synaptosomal
membrane
proteins.
Specifically
evaluated
were
Na+­
K+­
adenosine
triphosphatase
(
Na+­
K+­
ATPase),
a
well
known
cell
membrane
integral
enzyme
often
used
as
a
membrane
toxicity
model,
and
3H­
DHA­
labeled
beta­
adrenergic
receptor
binding,
a
synaptic
plasma
membrane
function
indicator.
The
study
found
that
all
the
monoketones
tested
produced
dose­
dependent
inhibition
of
Na+­
K+­
ATPase
activity
and
3H­
DHA
binding
to
mouse
synaptosomes.
Thus
monoketones,
like
other
lipophilic
solvents,
exert
their
membrane
effects
by
perturbation
of
the
membrane
micro­
environment
surrounding
membrane
receptors.

The
literature
contains
two
case
reports
of
men
developing
peripheral
neuropathy
following
abuse
of
spray
paint
or
lacquer
thinner
purported
to
contain
MIBK
(
AuBuchon
et
al.

1979;
Oh
and
Kim
1976).
In
these
cases,
there
were
no
confirmations
of
the
presence
of
MIBK
in
the
abused
products
and,
therefore,
there
is
the
distinct
possibility
that
the
reports
are
erroneous
in
referring
to
MIBK
as
the
potential
causative
agent
of
peripheral
neuropathy.
7
7
Further,
the
possibility
of
adverse
effects
following
intentional
abuse
of
a
solvent
product
does
not
indicate
a
hazard
from
reasonably
anticipated
environmental
releases.
29
In
conclusion,
the
available
data
indicate
that
MIBK
may
cause
transient
pharmacologic
effects
at
high
vapor
concentrations,
but
these
effects
are
readily
reversible
and
are
not
associated
with
any
evidence
of
neurotoxicity.
The
available
studies
do
not
demonstrate
that
MIBK
causes
damage
to
the
nervous
system,
even
following
repeated
exposure
at
high
concentrations.
Accordingly,
MIBK
cannot
reasonably
be
anticipated
to
cause
neurotoxic
effects
in
humans
under
realistic
exposure
scenarios.

6.
Other
Effects
Several
studies
have
evaluated
the
neurotoxicity
and
chemical
interactions
of
MIBK
with
chemicals
associated
with
neurologic
dysfunction.
While
MIBK
administered
alone
did
not
produce
the
specific
neurotoxic
effect,
when
administered
simultaneously
with
known
neurotoxic
chemicals,
the
subsequent
neurotoxicity
was
enhanced.
This
enhancement
was
the
likely
result
of
MIBK
induction
of
cytochrome
P­
450
enzymes
and
subsequent
increased
metabolism
of
these
chemicals
to
their
neurotoxic
metabolites.

Abou­
donia
et
al.
(
1985­
a)
studied
the
effects
of
MIBK,
n­
hexane,
and
combinations
of
the
two
solvents
on
induction
of
neurotoxicity
in
hens.
Continuous
exposures
to
1,000
ppm
MIBK
for
90
days
followed
by
a
30­
day
observation
period
resulted
in
leg
weakness
with
recovery
but
no
evidence
of
axonal
damage.
Lapadula
et
al.
(
1991)
examined
the
role
of
liver
microsomal
cytochrome
P­
450
in
the
mechanism
of
the
synergism
of
n­
hexane
neurotoxicity
by
MIBK.
The
results
of
this
study
suggested
that
MIBK
selectively
induces
cytochrome
P­
450
isozymes
leading
to
the
metabolic
activation
of
the
weak
neurotoxicant
n­
hexane
to
the
potent
neurotoxicant
2,5­
hexanedione.

Hens
simultaneously
treated
5
days/
week
for
90
days
with
technical
methyl
butyl
ketone
vapor
(
70%
methyl
n­
butyl
ketone
and
30%
methyl
isobutyl
ketone)
and
dermally
applied
30
O­
ethyl
O­
4­
nitrophenol
phenylphosphonothioate
(
85%)
(
EPN)
exhibited
greatly
enhanced
neurotoxicity
compared
to
the
neurotoxicity
produced
by
either
chemical
when
applied
alone
(
Abou­
Donia
et
al.
1985­
b).
The
authors
attributed
this
enhancement
in
part
to
MIBK
induction
of
cytochrome
P­
450
and
the
resulting
increase
in
the
formation
of
neurotoxic
products
from
methyl
n­
butyl
ketone
and
EPN.

In
a
subsequent
study,
Abou­
donia
et
al.
(
1991)
examined
the
joint
neurotoxic
action
of
simultaneous
exposure
to
vapors
of
MIBK
and
n­
hexane
and
dermally
applied
EPN
in
groups
of
hens.
As
was
noted
in
their
previous
studies,
increased
neurotoxicity
was
found
after
the
concurrent
exposure.
This
finding
was
specifically
related
to
MIBK
induction
of
phenobarbital­
inducible
cytochrome
P­
450
isozymes
and
the
subsequent
metabolic
activation
of
EPN.

MIBK
was
examined
for
effects
on
the
duration
of
ethanol­
induced
loss
of
righting
reflex
and
on
ethanol
elimination
in
mice
(
Cunningham
et
al.
1989).
MIBK,
at
a
dose
of
5
mmol/
kg
dissolved
in
corn
oil
and
injected
intraperitoneally
30
minutes
before
injection
of
4
g/
kg
ethanol,
produced
a
significantly
prolonged
duration
of
ethanol­
induced
loss
of
righting
reflex.

The
concentration
of
ethanol
in
blood
and
brain
upon
return
of
the
righting
reflex
were
similar
in
MIBK­
treated
and
control
animals
suggesting
the
interaction
altered
ethanol
metabolism
rather
than
increased
CNS
sensitivity.
MIBK
was
found
to
reduce
the
activity
of
mouse
liver
alcohol
dehydrogenase
in
vitro;
however,
this
effect
was
not
confirmed
in
vivo
by
a
reduction
in
the
elimination
rate
of
ethanol.

MIBK
administered
alone
produces
an
increase
in
liver
weight
in
laboratory
animals
but
no
associated
microscopic
evidence
of
hepatotoxicity.
Alternatively,
when
MIBK
is
administered
before
various
hepatotoxicants,
the
resulting
hepatic
injury
induced
by
the
31
hepatotoxicant
is
enhanced.
This
potentiation
of
hepatic
injury
has
been
best
documented
for
MIBK
and
haloalkanes.
In
addition
to
the
potential
of
liver
effects,
MIBK
has
also
been
demonstrated
to
enhance
the
nephrotoxicity
produced
by
select
haloalkanes.
The
explanation
for
these
interactions
is
considered
in
part
related
to
ketone
induction
of
monooxygenase
and
subsequent
secondary
formation
of
haloalkane
reactive
metabolites.

In
male
Sprague­
Dawley
rats,
a
single
oral
dose
of
MIBK
enhanced
the
hepatotoxicity
of
a
single
intraperitoneal
dose
of
chloroform
given
24
hours
later.
The
noobserved
effect
and
minimal­
effect
levels
of
MIBK
were
375
and
560
mg/
kg
body
weight,

respectively
(
Vezina
et
al.
1985).

Pilon
et
al.
(
1988)
demonstrated
that
the
extent
of
potentiation
of
carbon
tetrachloride
liver
toxicity
(
as
shown
by
an
increase
in
plasma
alanine
transaminease
activity
and
bilirubin
concentration)
was
dependent
on
the
concentration
of
both
MIBK
and
carbon
tetrachloride
in
male
rats.
The
minimum
effective
MIBK
dose
decreased
10­
fold
when
the
carbon
tetrachloride
dose
was
increased
from
0.01
mL/
kg
to
0.1
mL/
kg.
These
findings
suggest
that
liver
injury
is
determined
by
the
product
of
MIBK
and
carbon
tetrachloride
doses.

In
addition
to
necrosis,
effects
to
bile
flow
are
potentiated
by
ketone
solvents.

MIBK
and
two
metabolites,
4­
methyl­
2­
pentanol
and
4­
hydroxy
methyl
isobutyl
ketone
potentiated
the
cholestasis
induced
in
rats
by
a
combination
of
manganese­
bilirubin
or
manganese
alone
(
Vezina
and
Plaa
1987,
1988).
Pretreatment
of
rats
with
doses
of
375
mg/
kg
MIBK
for
at
least
three
days
enhanced
the
reduction
of
bile
flow
caused
by
acutely
toxic
doses
of
chloroform,

taurolithocholate,
manganese
(
4.5
mg/
kg
I.
V.)
or
a
manganese­
bilirubin
combination.
By
increasing
the
manganese
dose
to
6
mg/
kg,
given
intravenously,
the
minimally
effective
dose
of
MIBK
that
further
reduced
bile
flow
could
be
lowered
to
94
mg/
kg.
A
single
dose
of
MIBK
as
32
high
as
1500
mg/
kg
or
multiple
doses
of
750
mg/
kg
for
3
days
or
188
mg/
kg
for
7
days
did
not
reduce
bile
flow
in
the
absence
of
one
of
these
other
agents.
When
the
taurolithocholate
dose
was
reduced
from
20
mg/
kg
(
which
was
a
cholestatic
dose)
to
10
mg/
kg
(
which
was
not
a
cholestatic
dose),
MIBK
pretreatment
did
not
alter
bile
flow.

Vezina
et
al.
(
1990)
assessed
the
ability
of
MIBK
and
its
two
major
metabolites
to
potentiate
the
liver
injury
induced
by
chloroform
in
rats.
MIBK
and
both
metabolites
significantly
increased
the
liver
damage
induced
by
chloroform.
The
minimally
effective
dosage
to
potentiate
the
chloroform­
induced
hepatotoxicity
was
approximately
5
mmol/
kg
for
the
three
compounds.

The
induction
of
hepatotoxicity
was
demonstrated
to
be
associated
with
MIBK's
capacity
to
induce
cytochrome
P­
450.

In
addition
to
the
enhancement
of
haloalkane
hepatotoxicity,
MIBK
has
also
been
demonstrated
to
enhance
the
hepatotoxicity
of
1,2­
dichlorobenzene.
Brondeau
et
al.
(
1989)

examined
the
ability
of
acetone,
methyl
ethyl
ketone,
MIBK
and
cyclohexanone
vapors
to
influence
the
hepatotoxicity
of
1,2­
dichlorobenzene
in
rats
and
mice.
Sprague­
Dawley
rats
were
exposed
to
MIBK
vapor
at
concentrations
of
0,
595,
1280,
and
3020
ppm
and
OF1
mice
were
exposed
to
levels
of
0,
664,
1477,
and
3260
ppm
MIBK
for
4
hours
followed
18
hours
later
by
a
4­
hour
exposure
to
1,2­
dichlorobenzene.
Exposure
to
MIBK
alone
increased
hepatic
glutathione­

S­
transferase
activity
(
40
to
65
percent)
and
cytochrome
P450
levels
but
did
not
alter
glutamate
dehydrogenase
activity
in
rats
or
liver
glucose­
6­
phosphatase
activity
in
mice.
Following
pretreatment
with
MIBK,
1,2­
dichlorobenzene­
induced
increases
in
glutamate
dehydrogenase
and
glucose­
6­
phosphatase
activities
were
potentiated.
MIBK
was
concluded
to
potentiate
1,2­

dichlorobenzene
liver
injury
upon
reaching
a
threshold
value
of
liver
cytochrome
P­
450.
33
MIBK
potentiation
of
haloalkane­
induced
nephrotoxicity
is
described
by
Raymond
and
Plaa
(
1995).
Sprague­
Dawley
rats
pretreated
with
MIBK,
methyl
ethyl
ketone
or
acetone
(
13.6
mmol/
kg)
exhibited
a
significant
increase
in
chloroform
induced
kidney
toxicity.
Of
the
three
ketones
tested,
MIBK
demonstrated
the
lowest
potency
ranking
for
chloroform
nephrotoxicity.

The
interaction
effects
between
MIBK
and
the
hepatic
porphyrinogen,

hexachlorobenzene
was
examined
by
Krishnan
et
al.
(
1992).
This
study
found
that
when
the
two
chemicals
were
administered
simultaneously,
MIBK
reduced
the
severity
of
hexachlorobenzeneinduced
porphyria,
but
when
given
sequentially
after
hexachlorobenzene
accumulation,
MIBK
enhanced
the
porphyrinogenic­
response.
These
results
suggest
that
the
effect
of
combined
exposure
to
hexachlorobenzene
and
MIBK
on
hepatic
porphyria
depends
on
the
sequence
of
the
administration
of
both
chemicals,
and
the
mechanism
involved
in
the
interaction
may
invoke
both
the
induction
and
inhibition
of
specific
hepatic
isoenzymes
by
MIBK.

Krishnan
et
al.
(
1989)
demonstrated
that
MIBK
potentiated
the
methemoglobinemia
induced
by
N,
N­
dimenthylaniline.
Groups
of
male
Sprague­
Dawley
rats
were
pretreated
(
orally)
with
7.5
mmol/
kg
of
MIBK
followed,
18
hours
later,
by
the
intraperitoneal
administration
of
0.8
or
2.4
mmol/
kg
of
N,
N­
dimethylaniline.
Pretreatment
with
MIBK
enhanced
significantly
the
methemoglobinemia
produced
by
N,
N­
dimethylaniline.
Similar
enhancements
of
N,
N­
dimethylaniline
methemoglobinemia
were
observed
with
pretreatments
of
microsomal
enzyme
inducers,
suggesting
that
ketones
act
by
the
same
mechanism
as
the
enzyme
inducers.

The
common
conclusion
in
all
of
these
studies
is
that
enhanced
bioactivation
of
the
co­
administered
chemical
caused
by
MIBK
appears
to
be
mainly
responsible
for
these
interactions.
34
7.
Derivation
of
an
Inhalation
Reference
Concentration
(
RfC)
for
MIBK
The
IRIS
database
does
not
contain
an
inhalation
reference
concentration
(
RfC)

for
MIBK.
However,
the
Ketones
Panel
has
calculated
an
RfC
for
MIBK
based
on
EPA's
1994
guidance
for
deriving
RfCs.
See
EPA
Office
of
Research
and
Development,
"
Methods
for
Derivation
of
Inhalation
Reference
Concentrations
and
Application
of
Inhalation
Dosimetry,"

EPA
No.
600/
8­
90/
066F
(
October
1994)
(
hereinafter
the
"
RfC
Guidance").
The
RfC
is
based
on
the
1983
subchronic
inhalation
studies
in
rats
and
mice
sponsored
by
the
Ketones
Panel
as
part
of
the
testing
program
under
Section
4
of
TSCA.
This
study,
which
is
also
referenced
in
an
April
1987
document
prepared
by
EPA's
Environmental
Criteria
and
Assessment
Office,
was
subsequently
published
by
Phillips
et
al.
(
1987).
Significantly,
it
also
was
used
by
EPA
to
calculate
the
composite
score
for
MIBK
as
part
of
the
relative
hazard
ranking
for
MIBK
and
other
compounds
under
Section
112(
g)
of
the
Clean
Air
Act.
Based
on
the
Phillips
study,
the
RfC
for
MIBK
is
2.4
mg/
m3.
This
RfC
represents
a
conservative
estimate
of
the
concentration
of
MIBK
in
air
to
which
an
individual
could
be
exposed
for
a
lifetime
without
adverse
effect.
8
8.
Conclusions
Regarding
Potential
Chronic
Effects
In
summary,
the
weight
of
the
available
evidence
suggests
that
MIBK
can
not
reasonably
be
anticipated
to
cause
chronic
health
effects
in
humans.
Inhalation
studies
conducted
with
various
animals
indicate
low
subchronic
toxicity.
MIBK
is
not
teratogenic,
and
it
exhibits
low
developmental
toxicity
and
very
little,
if
any,
mutagenic
activity.
Based
on
its
structure
and
metabolism,
MIBK
is
not
likely
to
have
oncogenic
properties.
The
Ketones
Panel
has
calculated
8
A
further
explanation
of
how
the
Panel
calculated
the
RfC
for
MIBK
is
presented
in
Appendix
B.
35
an
RfC
for
MIBK
of
2.4mg/
m3,
which
far
exceeds
likely
human
exposure
levels
from
industrial
releases
of
MIBK.

D.
MIBK
Does
Not
Cause
Significant
Adverse
Environmental
Effects
Under
Section
112(
b)(
3)(
C)
of
the
Act,
EPA
must
also
consider
whether
emissions
of
a
substance
may
reasonably
be
anticipated
to
cause
"
adverse
environmental
effects."
The
term
"
adverse
environmental
effect"
is
defined
as:

any
significant
and
widespread
adverse
effect,
which
may
reasonably
be
anticipated,
to
wildlife,
aquatic
life,
or
other
natural
resources,
including
adverse
impacts
on
populations
of
endangered
or
threatened
species
or
significant
degradation
of
environmental
quality
over
broad
areas.

Section
112(
a)(
7).
Thus,
to
qualify
as
an
"
adverse
environmental
effect"
for
purposes
of
delisting
decisions,
the
effect
must
be
both
"
significant
and
widespread."
As
discussed
below,
MIBK
emissions
clearly
do
not
cause
significant
or
widespread
adverse
effects
on
the
environment.

1.
Biodegradation
MIBK
is
readily
biodegradable.
Bridié
et
al.
(
1979)
determined
the
biochemical
and
chemical
oxygen
demands
(
BOD
and
COD)
for
MIBK.
The
BOD
test
was
conducted
for
five
days
using
a
standard
dilution
method
and
seeded
with
coarse
filtered
effluent
from
a
biological
sanitary
waste
treatment
plant.
The
COD
tests
were
conducted
using
the
standard
potassium
dichromate
method.
The
BOD5
was
76%
of
the
theoretical
oxygen
demand
(
ThOD)

and
the
COD
was
79%
ThOD.
Price
et
al.
(
1974)
reported
a
BOD20
of
69%
ThOD
for
MIBK
using
non­
acclimated
seed
from
settled
domestic
wastewater
and
a
BOD20
of
53%
ThOD
in
synthetic
seawater.
An
earlier
study
by
Lamb
and
Jenkins
(
1953),
using
settled
sewage
seed,

reported
a
BOD20
of
56.6%
ThOD
for
MIBK.
Since
the
reported
BOD
values
are
approximately
36
equal
to
or
greater
than
60%
ThOD
within
28
days,
MIBK
is
considered
readily
biodegradable
according
to
EPA
environmental
fate
guidelines
(
see
40
C.
F.
R.
§
796.3200).

2.
Potential
For
Bioaccumulation
Because
MIBK
readily
biodegrades
(
Bridié
et
al.
1979;
Price
et
al.
1974;
Lamb
and
Jenkins
1952)
and
enters
the
intermediary
metabolism
of
mammals
(
DiVincenzo
et
al.
1976),

it
would
not
be
expected
to
bioaccumulate.
Thus,
it
is
not
surprising
that
the
calculated
ecological
magnification
for
fish
was
zero
in
Metcalf's
Model
Aquatic
Ecosystem
(
Lande
et
al.

1976).

The
bioconcentration
potential
of
MIBK
is
also
related
to
its
octanol­
water
partition
coefficient
(
Kow).
The
reported
values
of
log
Kow
for
MIBK
ranged
from
1.09
to
1.31
(
Ginnings
et
al.
1940;
Hansch
et
al.
1968;
Sangster
1989).
Bioconcentration
factors
(
BCF)
of
2­
5
and
a
log
BCF
of
0.38
were
estimated
for
MIBK
using
linear
regression
equations
from
Lyman
et
al.
(
1982)
based
on
a
log
Kow
of
1.19
and
a
water
solubility
of
20400
mg/
L
at
20
°
C
(
HSDB
1996).
Chemicals
with
BCF
<
100
have
a
low
potential
to
bioconcentrate
based
on
EPA's
criteria
(
Zeeman
1995).

3.
Effects
On
Microorganisms
Egyud
(
1967)
reported
that
a
concentration
of
1
x
10­
3
M
MIBK
had
only
mild
transient
inhibitory
effects
on
the
growth
of
E
coli.
Bringmann
and
Kühn
(
1980)
reported
a
16­

hour
toxicity
threshold
of
275
mg/
L
for
Pseudomonas
putida.
Using
the
Microtox
®
toxicity
analyzer
to
conduct
bacterial
bioluminescense
bioassays,
Curtis
et
al.
(
1982)
obtained
a
5
min
EC50
of
80
mg/
L
for
Photobacterium
phosphoreum.
37
Protozoans
are
less
susceptible
than
bacteria
to
the
effects
of
MIBK,
as
indicated
by
the
cell
multiplication
inhibition
test
toxicity
thresholds
shown
in
Table
1.
The
lowest
toxicity
threshold
for
growth
inhibition
of
protozoa
was
450
mg/
1
for
Entosiphon
sulcatum.

TABLE
1
Acute
Toxicity
of
MIBK
to
Protozoa:
Cell
Multiplication
Inhibition
Test
Toxicity
Threshold
Protozoa
mg/
l
Duration
of
Experiment
Saprozoic
flagellate
(
Chilomosas
paramaecium)
>
800
48
hours
Bacteriovorous
flagellate
(
Entosiphon
sulcatum)
450
72
hours
Bacteriovorous
ciliate
(
Uronema
parduczi)
950
20
hours
______________________

References:
Bringmann
and
Kühn
(
1980,
1981).

4.
Effects
on
Aquatic
Organisms
MIBK
has
a
low
acute
and
low
chronic
toxicity
to
aquatic
organisms
based
on
EPA's
criteria
(
Zeeman
1995)
of
>
100
mg/
L
and
>
10
mg/
L
for
low
acute
and
low
chronic
toxicity,
respectively.
The
acute
toxicity
values
ranged
from
460
to
890
mg/
L
for
several
species
of
freshwater
fish
and
from
240
to
4280
mg/
L
for
two
invertebrate
species
(
Table
2a).
The
chronic
toxicity
of
MIBK
was
studied
in
an
early
life
stage
test
with
fathead
minnows
(
Call
and
Geiger
1992)
and
in
a
life
cycle
test
with
Daphnia
magna
(
Kühn
et
al.
1989).
The
no­
observed
38
effect
concentrations
(
NOEC)
were
56
and
168
mg/
L
for
fish
growth
and
survival,
respectively,

and
78
mg/
L
for
daphnid
reproduction
(
Table
2b).

MIBK
also
has
a
low
toxicity
to
algae.
EC50
values
reported
for
two
genera
of
green
algae
ranged
from
400
to
2000
mg/
L,
and
no­
effect
or
threshold
concentrations
of
136
and
725
mg/
L
were
found
for
a
blue­
green
and
green
algae
species
(
Table
3).

TABLE
2a
Acute
Toxicity
of
MIBK
to
Aquatic
Organisms
Organism
Species
LC50
or
EC50
(
mg/
L)
Duration
of
Test
Reference
Freshwater
fish
Golden
orfe
(
Leuciscus
idus
melanotus)
675­
890
48
hrs
Juhnke
and
Lüedemann
1978
Goldfish
(
Carassius
auratus)
460
24
hrs
Bridié
et
al.
1979
Rainbow
Trout
(
Salmo
gairdneri)
600
96
hrs
Stephenson
1983
Fathead
minnow
(
Pimephales
promelas)
505­
540
780
96
hrs
24­
48
hrs
Veith
et
al.
1983
Brooke
et
al.
1984
Waggy
and
Payne
1974
Freshwater
invertebrate
Water
flea
(
Daphnia
magna)
240­
4280
24
hrs
Bringmann
&
Kühn
1977
and
1982
Kühn
et
al.
1989
Marine
invertebrate
Brine
shrimp
(
Artemia
salina)
1230*
24
hrs
Price
et
al.
1974
_____________________

Median
toxicity
threshold
(
TLm)
39
TABLE
2b
Chronic
Toxicity
of
MIBK
to
Aquatic
Organisms
Organism
Species
NOEC
(
mg/
L)
Duration
of
Test
Reference
Freshwater
fish
Fathead
minnow
(
Pimephales
promelas)
168
survival
33
days
Call
and
Geiger
1992
56.2
weight
33
days
Call
and
Geiger
1992
<
56.2
length
33
days
Call
and
Geiger
1992
Freshwater
invertebrate
Water
flea
(
Daphnia
magna)
78
reproduction
21
days
Kühn
et
al.
1989
TABLE
3
Acute
Toxicity
of
MIBK
to
Freshwater
Algae
Species
EC50
(
mg/
L)
Duration
of
Test
Reference
Bluegreen
algae
(
Microcystis
aeruginosa)
136*
8
days
Bringmann
and
Kühn
1978
Green
algae
(
Scenedesmus
quadricauda)
725**
7
days
Bringmann
and
Kühn
1980
Green
algae
(
Scenedesmus
subspicatus)
980
biomass
48
hours
Kühn
and
Pattard
1990
2000
growth
rate
48
hours
Kühn
and
Pattard
1990
Green
algae
(
Selenastrum
capricornutum)
400
96
hours
Stephenson
1983
*
ECO
**
Cell
multiplication
inhibition
test
toxicity
threshold
(
TT).
40
5.
Effects
on
Plants
As
noted
in
Section
I.
C
above,
MIBK
occurs
naturally
in
plants.
The
effects
of
MIBK
on
lower
order
plants
have
been
discussed
in
the
preceding
section.
Toxicity
to
higher
order
plants
due
to
MIBK
exposure
has
not
been
reported.
Higher
order
plants
are
unlikely
to
be
exposed
to
deleterious
levels
of
MIBK
because
MIBK
is
readily
degraded
in
the
atmosphere
and
by
biodegradation.

IV.
DATA
ON
EMISSIONS
AND
EXPOSURE
A.
Emissions
Data
Because
MIBK
emissions
currently
must
be
reported
under
Section
313
of
EPCRA,
the
Toxics
Release
Inventory
(
TRI)
is
a
good
source
of
information
about
MIBK
emissions
from
industrial
facilities.
Table
4
below
summarizes
reported
emissions
of
MIBK
based
on
1994
TRI
data,
and
indicates
the
number
of
TRI
reporting
facilities
with
MIBK
air
emissions
in
different
reporting
ranges.

TABLE
4
MIBK
Air
Emissions
No.
of
Facilities
Reporting
Range
(
lbs/
yr)
%
Distribution
778
0
­
20,000
75.5
162
20,001
­
50,000
15.7
46
50,001
­
100,000
4.5
21
100,001
­
200,000
2.1
11
200,001
­
300,000
1.1
5
300,001
­
400,000
0.5
4
400,001
­
700,000
0.4
4
above
700,000
0.4
1,031
Reference:
1994
TRI
Data.
41
As
discussed
above
in
Section
I.
B,
MIBK
is
widely
used
in
many
types
of
solventbased
systems
because
of
its
effectiveness.
Table
4
shows
that,
although
MIBK
is
used
at
a
large
number
of
facilities,
the
vast
majority
of
them
have
very
low
emissions
of
MIBK.
Over
75
percent
of
the
facilities
reporting
MIBK
emissions
emitted
less
than
10
tons
in
1994.
Over
90
percent
of
those
facilities
reported
1994
MIBK
emissions
of
less
than
25
tons.

B.
Ambient
Monitoring
Data
MIBK
has
been
reported
in
ambient
air
at
very
low
concentrations
at
a
limited
number
of
sites
in
rural
and
urban
locations.
Data
on
ambient
air
levels
of
MIBK
are
presented
in
Appendices
C
and
D.
The
following
is
a
brief
overview
of
the
information
presented
in
those
appendices,
as
well
as
additional
information
found
in
the
published
literature.

Appendix
C
includes
a
table
taken
from
a
study
conducted
by
the
State
of
New
Jersey
Department
of
Environmental
Protection
in
1979
(
Bozzelli
et
al.
1980).
MIBK
was
detected
in
only
4
out
of
168
samples
taken
in
1979
during
a
study
conducted
by
the
Air
Pollution
Research
Laboratory
of
the
New
Jersey
Institute
of
Technology.
The
highest
measured
value
of
MIBK
in
the
study
was
1.00
ppb.
The
six
sample
sites
included
densely
populated
areas
where
large
chemical
complexes
were
located,
including
the
Exxon
Bayway
facility
which,
at
the
time,

produced
MIBK
(
Bozzelli
and
Kebbekus
1983).
Thirty­
six
samples
were
taken
in
Elizabeth,
New
Jersey,
near
the
Exxon
Bayway
refinery,
which
at
the
time
was
the
largest
MIBK
producer
in
the
United
States.

An
ambient
air
monitoring
study
that
included
MIBK
has
been
conducted
in
the
industrial
(
ship
channel)
area
of
Houston,
Texas
since
January,
1987.
The
results
from
the
seven
monitoring
stations
in
the
ship
channel
during
the
period
of
January
1,
1987
to
December
31,
42
1995
show
24­
hour
average
airborne
concentrations
of
MIBK
from
below
the
level
of
detection
to
a
high
of
5.77
ppb.
The
mean
(
long­
term
average)
airborne
concentrations
of
MIBK
at
the
seven
monitoring
sites
during
the
same
eight­
year
period
ranged
from
0.13
ppb
to
0.18
ppb.
See
Houston
Regional
Monitoring
Report,
included
in
Appendix
D.

C.
Air
Dispersion
Modeling
Data
for
Industrial
Facilities
Over
the
last
two
years,
the
Ketones
Panel
has
undertaken
a
program
to
gather
data
on
the
maximum
airborne
concentrations
of
MIBK
to
which
the
public
may
be
exposed.
As
part
of
this
program,
the
Panel
funded
a
study
by
ENSR
Corporation
to
model
the
maximum
offsite
concentrations
of
MIBK
at
a
wide
variety
of
facilities
emitting
MIBK,
including
the
largest
sources
of
MIBK
emissions
in
the
country.
The
findings
of
the
ENSR
study,
along
with
a
detailed
description
of
the
methodology
employed
by
ENSR,
are
contained
in
the
report
attached
at
Appendix
E
(
hereinafter
referred
to
as
the
ENSR
Report).
This
study
shows
that,
even
at
the
largest
industrial
emitters
of
MIBK,
maximum
airborne
concentrations
beyond
facility
boundaries
are
well
below
levels
of
concern
and
do
not
pose
a
risk
to
human
health
or
the
environment.

The
ENSR
study
was
divided
into
three
parts.
First,
because
airborne
concentrations
are
likely
to
be
highest
around
facilities
with
the
highest
emission
rates,
ENSR
separately
evaluated
maximum
off­
site
concentrations
of
MIBK
around
each
of
the
facilities
reporting
MIBK
emissions
of
150
tons
or
more
in
1994.
Second,
ENSR
used
a
generic
model
to
make
conservative
estimates
of
maximum
off­
site
concentrations
around
smaller
facilities
that
emit
MIBK.
Third,
ENSR
analyzed
the
possibility
that
groups
of
facilities
located
in
the
same
area
might
collectively
cause
airborne
levels
of
concern.
The
three
parts
of
ENSR's
analysis
are
discussed
below.
43
1.
Air
Dispersion
Modeling
of
the
Highest
Emitters
As
the
starting
point
for
its
modeling
program,
the
Ketones
Panel
sought
to
model
the
maximum
off­
site
concentrations
for
all
facilities
reporting
MIBK
emissions
greater
than
150
tons
per
year.
The
Panel
selected
this
threshold
based
on
the
methodology
that
EPA
developed
to
set
de
minimis
values
for
hazardous
air
pollutants
under
Section
112(
g)
of
the
Clean
Air
Act.
See
Documentation
for
De
Minimis
Emission
Rates
for
Proposed
40
CFR
part
63,
subpart
B
(
EPA­

453/
R­
93­
035).
Under
Section
112(
g),
the
de
minimis
value
for
a
chemical
is
the
amount
that
an
EPA
model
facility
could
emit
without
posing
more
than
a
"
trivial"
risk
to
human
health
or
the
environment.
(
The
de
minimis
value
for
MIBK
is
discussed
further
in
Section
V.
B
of
this
Petition.)
In
the
Section
112(
g)
rulemaking,
EPA
proposed
to
"
cap"
de
minimis
levels
at
10
tons
per
year
(
tpy),
but
at
the
same
time
recognized
that,
for
low
toxicity
chemicals,
emissions
of
more
than
10
tpy
would
still
pose
only
a
trivial
risk.
59
Fed.
Reg.
15,504,
15,527
(
April
1,
1994).

Significantly,
EPA's
methodology
may
also
be
used
to
calculate
true
"
uncapped"
de
minimis
values
for
different
compounds.

EPA's
methodology
requires
the
use
of
an
RfC.
At
the
commencement
of
its
modeling
exercise,
the
Panel
believed
that
the
RfC
for
MIBK
was
0.75
mg/
m3
based
on
EPA's
pre­
1994
approach
for
setting
RfCs.
(
As
discussed
above
at
p.
35,
however,
the
correct
RfC
for
MIBK,
based
on
EPA's
1994
RfC
Guidance,
is
actually
2.4
mg/
m3.)
Using
the
lower
RfC
in
EPA's
model
for
calculating
de
minimis
values,
the
Panel
derived
a
de
minimis
value
for
MIBK
of
1,500
tons
per
year.
In
order
to
establish
a
meaningful
cutoff
point
for
its
modeling
exercise,

the
Panel
decided
that
it
would
seek
to
model
all
facilities
with
reported
emissions
that
were
more
than
10
percent
of
this
amount.
Thus,
it
sought
to
model
the
maximum
off­
site
concentrations
for
44
all
facilities
reporting
MIBK
emissions
greater
than
150
tons
(
or
300,000
pounds)
per
year.

Based
on
1993
TRI
data,
the
Panel
identified
13
such
facilities.

The
Panel
also
worked
with
ENSR
to
develop
a
detailed
questionnaire
to
gather
the
information
that
would
be
necessary
to
model
the
maximum
off­
site
concentrations
at
each
facility.
This
questionnaire,
along
with
a
cover
letter
explaining
the
Panel's
modeling
program,

was
sent
to
the
13
facilities,
and
representatives
from
the
Panel
also
contacted
each
of
the
facilities
to
encourage
their
participation.
By
the
time
the
Panel
received
the
necessary
data
and
ENSR
began
its
modeling
exercise,
the
TRI
data
for
1994
had
become
available.
Based
on
the
1994
data,
two
of
the
original
13
facilities
were
below
the
150
ton
threshold,
but
two
additional
facilities
now
exceeded
the
threshold.
During
the
course
of
its
discussions
with
the
individual
facilities,
the
Panel
also
discovered
that
one
facility
had
incorrectly
calculated
its
reported
MIBK
emissions
and,
in
fact,
was
well
below
the
150­
ton
threshold.
As
a
result,
this
facility
was
not
included
in
the
modeling
study,
and
the
ENSR
Report
covers
12
rather
than
13
individual
facilities.

As
described
in
the
ENSR
Report,
either
facility­
specific
data
or
pre­
existing
modeling
results
were
obtained
from
9
of
the
12
highest­
emitting
MIBK
sources
in
the
country.
9
One
of
the
remaining
facilities
declined
to
participate
in
the
ENSR
study
because
it
had
already
committed
to
phasing
out
the
use
of
MIBK.
In
order
to
conduct
modeling
for
the
other
two
facilities
identified
as
top
emitters,
the
Panel
and
ENSR
attempted
to
obtain
site­
specific
9
Four
facilities
provided
the
full
set
of
facility­
specific
data
needed
for
modeling
and
two
facilities
provided
results
of
dispersion
modeling
of
MIBK
emissions
that
had
already
been
conducted
independent
of
the
ENSR
study.
In
addition,
three
facilities
provided
limited
information
that
was
supplemented
with
publicly
available
data
from
the
AIRS
database
and
Title
V
permit
applications.
45
information
from
public
sources,
including
EPA's
Aerometric
Information
Retrieval
System
(
AIRS),
permit
applications,
and
site
plans
on
file
with
local
zoning
boards.
This
effort
generated
sufficient
data
to
allow
ENSR
to
conduct
site­
specific
modeling
for
one
of
the
two
remaining
facilities
that
did
not
provide
information
directly
to
ENSR.
Thus,
the
data
necessary
for
sitespecific
modeling
was
obtained
­­
either
from
the
facilities
or
from
public
sources
­­
for
10
of
the
12
highest
emitters
of
MIBK.
10
Using
this
data,
ENSR
performed
air
quality
modeling
analyses
for
each
facility
using
EPA's
"
Tiered
Modeling
Approach
for
Assessing
Risks
Due
to
Sources
of
Hazardous
Air
Pollutants"
(
1992).
This
approach
uses
three
successively
more
rigorous
modeling
techniques.

Tier
1
requires
only
limited
source
information
and
an
EPA
look­
up
table,
and
provides
the
most
conservative
predictions
of
maximum
concentrations.
Tier
2
requires
additional
source
information
and
an
EPA
screening
level
computer
program,
and
generates
predictions
that
are
somewhat
more
realistic
than
Tier
1
predictions.
Tier
3,
which
requires
extensive
data
from
the
source
and
uses
EPA's
most
advanced
dispersion
modeling
techniques,
provides
the
most
realistic
predicted
concentrations.
Tier
3
modeling
was
performed
for
a
facility
if
Tier
2
modeling
predicted
maximum
annual
concentrations
above
1.0
mg/
m3
or
maximum
24­
hour
concentrations
above
5.0
mg/
m3.
The
results
of
the
ENSR
modeling
study
of
the
highest
emitters
are
shown
on
Table
5
and
6.
Table
5
shows
maximum
annual
off­
site
concentrations;
Table
6
shows
maximum
10
For
the
two
remaining
facilities,
ENSR
estimated
maximum
annual
off­
site
concentrations
using
the
generic
approach
developed
to
evaluate
ambient
concentrations
around
smaller
sources.
This
approach
is
described
fully
in
the
ENSR
Report
and
is
discussed
in
Section
IV.
C.
2
below.
Even
using
conservative
assumptions,
the
maximum
predicted
annual
offsite
concentrations
at
these
2
facilities
were
less
than
0.75
mg/
m3.
ENSR
Report
at
pp.
4­
3
(
Table
4­
2)
and
5­
1.
46
24­
hour
off­
site
concentrations.
In
both
tables,
facilities
are
listed
according
to
Tier
2
modeling
results
in
descending
order.

TABLE
5
Air
Dispersion
Modeling
Results
for
Highest­
Emitting
MIBK
Sources
Maximum
Annual
MIBK
Concentrations
(
mg/
m3)
(
RfC
=
2.4
mg/
m3)

Site*
Tier
1
Tier
2
Tier
3
B10
14.75
6.58
0.35
B8
9.88
2.81
0.13
B15
8.04
2.07
0.22*
B2
5.85
1.18
0.41*
B6
4.1
0.73
­­
B13
1.21
0.60
­­
B17
2.54
0.57
­­
B11
3.39
0.38
­­
B9
­­
0.008**
­­
B4
­­
­­
0.004**

*
Simplified
conservative
Tier
3
modeling
conducted
assuming
all
emissions
from
a
single
point
source
with
the
minimum
distance
to
property
boundary
in
all
directions.

**
Based
on
modeling
results
provided
by
the
individual
company.

*
Companies
submitted
the
information
necessary
to
conduct
the
modeling
under
conditions
of
confidentiality.
For
this
reason,
facilities
are
not
identified
by
name
and
modeling
results
cannot
be
presented
side­
by­
side
with
emissions
data.

As
shown
on
Table
5,
airborne
concentration
levels
of
MIBK
are
clearly
below
levels
of
concern.
Based
on
the
Tier
3
data,
even
the
highest
predicted
concentration
is
almost
an
order
of
magnitude
below
the
RfC.
The
data
as
a
whole
demonstrate
that
emissions
of
MIBK
do
not
result
in
ambient
concentrations
that
may
reasonably
be
anticipated
to
cause
chronic
adverse
health
effects.
47
TABLE
6
Air
Dispersion
Modeling
Results
for
Highest­
Emitting
MIBK
Sources
Maximum
24­
hour
MIBK
Concentrations
(
mg/
m3)
(
Benchmark
=
24
mg/
m3)
*

Site**
Tier
1
Tier
2
Tier
3
B10
59.2
32.9
2.73
B8
51.8
19.70
1.71
B15
44.9
14.46
2.56*
B2
58.3
8.28
3.23*
B13
6.8
4.23
­­
B6
17.6
4.04
­­
B17
14.2
3.97
­­
B11
18.6
2.67
­­
B9
­­
0.04**
­­
B4
­­
­­
0.02**

*
Simplified
conservative
Tier
3
modeling
conducted
assuming
all
emissions
from
a
single
point
source
with
the
minimum
distance
to
property
boundary
in
all
directions.

**
Based
on
modeling
results
provided
by
the
individual
company.

*
This
health
benchmark
is
based
on
the
RfC
of
2.4
mg/
m3,
modified
only
to
eliminate
the
uncertainty
factor
of
10
for
extrapolating
from
subchronic
to
chronic
exposures.

**
Companies
submitted
the
information
necessary
to
conduct
the
modeling
under
conditions
of
confidentiality.
For
this
reason,
facilities
are
not
identified
by
name
and
modeling
results
cannot
be
presented
side­
by­
side
with
emissions
data.

Maximum
24­
hour
off­
site
MIBK
concentrations
were
compared
to
a
health
benchmark
of
24
mg/
m3
­­
the
RfC
modified
to
eliminate
the
uncertainty
factor
of
10
that
is
used
to
extrapolate
from
subchronic
to
chronic
exposure.
As
shown
on
Table
6,
the
maximum
24­
hour
concentrations
were
all
well
below
this
benchmark.
Based
on
tier
3
values,
the
highest
predicted
24­
hour
value
was
3.23
mg/
m3
­­
again
almost
an
order
magnitude
below
the
health
benchmark.

It
should
be
noted
that
the
maximum
predicted
24­
hour
concentrations
are
based
on
the
worstcase
conditions
occurring
in
any
24­
hour
period
over
the
last
5
years,
and
are
therefore
very
conservative.
48
It
is
also
important
to
recognize
that
the
methodology
used
by
ENSR
was
not
intended
to
represent
actual
population
exposure.
Rather,
it
was
designed
to
identify
the
highest
annual
and
24­
hour
off­
site
concentrations
that
might
occur
around
each
facility.
In
all
cases,
the
modeled
maximum
concentrations
are
near
facility
boundaries,
and
it
is
unlikely
that
there
is
continuous
exposure
at
any
of
these
locations.
Further,
because
the
methodology
is
designed
to
predict
maximum
off­
site
concentrations,
it
incorporates
a
number
of
conservative
assumptions.

Actual
average
concentrations
are
likely
to
be
lower,
and
could
be
lower
by
an
order
of
magnitude
or
more.
Thus,
the
results
of
ENSR's
air
dispersion
modeling
analyses
likely
overpredict
actual
exposures.

2.
Air
Dispersion
Modeling
of
Smaller
Sources
The
Panel
also
recognized
that
there
could
be
relatively
high
off­
site
exposures
around
smaller
MIBK
sources
due
to
such
things
as
unusual
dispersion
climatology,
lower
emission
release
heights,
and
proximity
of
nearby
residents.
Therefore,
ENSR
developed
an
approach
for
analyzing
maximum
off­
site
concentrations
around
smaller
facilities
that
emit
MIBK.

All
facilities
that
reported
MIBK
emissions
of
10
tons
or
more
on
the
1994
TRI
were
divided
into
source
categories
based
on
their
two­
digit
SIC
codes.
As
discussed
below,
ENSR
developed
model
facilities
for
each
source
category
in
which
no
facility
was
individually
modeled
under
the
first
part
of
its
analysis.
It
then
used
a
generic
EPA
model
to
predict
maximum
off­
site
concentrations
for
facilities
in
each
source
category.

As
described
more
fully
in
the
ENSR
Report,
the
air
dispersion
model
used
to
evaluate
potential
exposures
around
smaller
sources
was
based
on
the
model
developed
by
EPA
as
part
of
the
Agency's
rulemaking
under
Section
112(
g)
of
the
Clean
Air
Act.
This
is
a
conservative
model
that
allows
the
prediction
of
maximum
annual
exposures
based
on
two
49
parameters:
emissions
release
height
and
distance
to
the
nearest
receptor.
The
EPA
model
incorporates
the
following
conservative
assumptions:

 
emissions
all
emanate
from
a
single
point;

 
emissions
have
negligible
exit
velocity
(
10
cm/
second);

 
emissions
are
released
at
ambient
temperature;
and
 
emissions
are
subject
to
worst­
case
aerodynamic
building
downwash.

For
purposes
of
the
Section
112(
g)
rulemaking,
the
Agency
applied
the
model
based
on
median
dispersion
climatology
developed
from
314
weather
stations
located
throughout
the
United
States.
To
predict
maximum
annual
concentrations
around
smaller
sources
of
MIBK,
ENSR
adjusted
the
model
to
incorporate
worst­
case
dispersion
climatology.

ENSR
then
assigned
each
of
the
source
categories
identified
with
MIBK
sources
to
one
of
the
following
"
dispersion
categories"
based
on
the
assumptions
set
forth
below:

 
Heavy:
Major
facilities
located
in
industrial
areas
on
relatively
large
sites.
(
Stack
height
=
15
meters;
distance
to
receptor
=
200
meters)

 
Medium:
Moderate
size
facilities
located
in
light
industrial
or
commercial
areas
on
smaller
sites.
(
Stack
height
=
10
meters;
distance
to
receptor
=
150
meters)

 
Light:
Smaller
facilities
located
on
relatively
small
sites
in
mixed­
use
areas,
where
emissions
are
likely
to
be
released
from
roof
vents
in
one­
story
buildings.
(
Stack
height
=
5
meters;
distance
to
receptor
=
100
meters)

The
model
was
then
used
to
predict
maximum
annual
concentrations
for
facilities
representing
each
source
category
in
which
at
least
one
facility
reported
more
than
10
tons
of
MIBK
emissions
in
1994.
The
predicted
maximum
concentrations
for
each
source
category
are
based
on
the
emission
rates
reported
by
the
facilities
reporting
the
highest
and
second
highest
emissions
in
each
category.
The
results
of
this
analysis
are
presented
in
Table
7.
50
TABLE
7
Air
Dispersion
Modeling
Results
for
Smaller
MIBK
Sources
Maximum
Annual
MIBK
Concentrations
(
mg/
m3)
(
RfC=
2.4
mg/
m3)

SIC
Code
Source
Category
Description
Dispersion
Category
Highest
Emission
Rate
in
SIC
Code
(
lbs)
Predicted
Concentration
from
Highest
Rate
(
mg/
m3)
2nd
Highest
Emission
Rate
in
SIC
Code
(
lbs)
Predicted
Concentration
from
2nd
Highest
Rate
(
mg/
m3)

23
Apparel
Light
127,930
0.525
28,514
0.117
24
Lumber
Medium
180,000
0.164
58,160
0.053
25
Furniture
Light
88,750
0.364
78,687
0.323
26
Paper
Heavy
114,492
0.045
NA
NA
27
Printing
Medium
24,885
0.023
NA
NA
29
Refining
Heavy
236,000
0.092
112,470
0.044
31
Leather
Medium
27,962
0.025
24,200
0.022
32
Concrete
Medium
31,840
0.029
NA
NA
34
Fab.
Metal
Medium
207,502
0.189
83,336
0.076
35
Ind.
Machine
Heavy
93,523
0.036
46,700
0.018
36
Electronic
Medium
56,000
0.051
48,700
0.044
38
Measuring
Inst.
Medium
46,200
0.042
21,701
0.020
39
Misc.
Mfg.
Medium
43,450
0.040
32,900
0.030
87
Engineering
Light
31,050
0.127
NA
NA
97
Nat.
Security
Heavy
20,610
0.008
NA
NA
51
Based
on
this
analysis,
ENSR
concluded
that
"[
g]
eneralized
dispersion
modeling
indicates
that
maximum
annual
off­
site
concentrations
are
below
1
mg/
m3
in
the
vicinity
of
these
lesser
emitting
facilities
[
and]
are
well
below
1
mg/
m3
in
most
cases."
ENSR
Report,
Executive
Summary.
Thus,
based
on
an
assessment
of
a
wide
range
of
sources
and
source
categories,

including
both
large
and
small
sources,
it
appears
that
maximum
off­
site
concentrations
of
MIBK
are
well
below
levels
of
concern.

3.
Potential
Impacts
from
Groups
of
Sources
The
ENSR
study
also
evaluated
the
possibility
that
there
might
be
groups
of
facilities
that
collectively
have
significant
health
or
environmental
impacts,
even
though
no
single
facility
reported
emissions
that
would
raise
concerns.
ENSR
began
this
part
of
its
analysis
by
noting
that,
based
on
the
results
of
the
modeling
study
for
the
highest
emitters,
"
for
sources
to
contribute
to
off­
site
receptor
concentrations
above
the
RfC,
at
least
3
major
MIBK
emitters
would
need
to
be
side­
by­
side
with
a
receptor
located
between
them."
ENSR
Report
at
p.
6­
1.

ENSR
then
pointed
out
that
no
two
of
the
top
emitting
facilities
were
even
located
in
the
same
geographic
area.
As
a
result,
it
concluded
that
it
was
"
highly
improbable"
that
combined
emissions
from
multiple
facilities
would
result
in
concentrations
above
the
health
benchmarks.
Id.

ENSR
nevertheless
conducted
an
evaluation
to
identify
possible
clusters
of
MIBK
sources.
As
a
first
step,
it
used
the
TRI
database
to
identify
every
facility
in
the
country
that
reported
MIBK
emissions
of
more
than
10
tons
in
1994,
including
both
point
and
fugitive
emissions.
For
the
year
1994,
there
were
253
facilities
above
this
threshold.
Of
this
number,
242
facilities
are
located
in
postal
ZIP
codes
in
which
there
was
no
other
facility
reporting
MIBK
emissions
greater
than
10
tons
in
1994.
Each
of
the
remaining
11
facilities
are
listed
in
Appendix
52
F,
which
is
organized
by
ZIP
code.
Appendix
F
shows
the
reported
emissions
for
each
facility,

along
with
the
total
emissions
for
each
ZIP
code.

This
analysis
shows
that
there
is
no
significant
grouping
of
sources
that
emit
MIBK.
The
11
facilities
are
distributed
fairly
evenly
over
5
different
ZIP
codes.
Only
one
of
these
ZIP
codes
contains
more
than
2
facilities,
but
total
MIBK
emissions
from
facilities
in
that
ZIP
code
were
just
over
40
tons
in
1994.
ENSR
noted
that,
even
if
all
the
facilities
in
any
single
zip
code
"
were
combined
into
one
co­
located
facility,
both
site
specific
and
generalized
modeling
suggests
that
such
a
facility
would
not
cause
ambient
levels
that
would
approach
the
health
benchmarks."
Id.
Thus,
ENSR
concluded
that
"
there
is
no
group
of
MIBK
emitting
facilities
that
collectively
would
result
in
maximum
off­
site
concentrations
approaching
the
[
health]

benchmarks."
Id.
at
p.
7­
1.

D.
Effect
of
Delisting
on
Emissions
and
Ambient
Concentration
Because
of
MIBK's
status
as
a
HAP,
companies
currently
are
discouraged
from
using
it.
In
some
cases,
product
formulators
and
users
must
comply
with
regulatory
limits
on
the
HAP
content
of
their
products.
In
other
cases,
companies
are
likely
to
reduce
their
use
of
HAPs
in
order
to
avoid
the
need
to
install
maximum
available
control
technology
(
MACT)
under
Section
112(
d)
of
the
Act.
Even
in
the
absence
of
regulatory
requirements,
companies
often
try
to
avoid
the
use
of
chemicals
that
are
labeled
as
HAPs.
Therefore,
if
MIBK
is
removed
from
the
HAP
list,
usage
of
MIBK
is
likely
to
increase.

For
several
reasons,
however,
MIBK
emissions
are
unlikely
to
increase
substantially.
First,
MIBK
is
used
primarily
in
paints
and
coatings
and,
to
a
lesser
extent,
in
inks
and
adhesives.
In
these
and
other
solvent
applications,
MIBK
is
rarely
used
by
itself.
Typically,

MIBK
is
part
of
a
solvent
blend
that
must
be
carefully
formulated
to
achieve
the
proper
53
performance
characteristics,
including
such
things
as
evaporation
rate,
surface
tension,
solvent
balance,
and
flash
point.
Although
there
is
flexibility
to
increase
the
use
of
MIBK
in
many
solvent
blends,
there
are
inherent
limits
on
the
amount
of
MIBK
(
or
any
other
single
solvent)
that
can
be
used
in
any
formulation.

Second,
because
MIBK
is
used
primarily
in
solvent
blends,
there
will
often
be
other
HAPs
that
are
used
in
the
same
application.
In
many
cases,
the
facilities
where
such
applications
are
used
will
be
required
to
meet
"
maximum
available
control
technology"
(
MACT)

standards
and
will
need
to
install
control
technology
to
reduce
their
HAP
emissions.
In
most
cases,
such
technology
will
reduce
all
solvent
emissions,
including
emissions
of
MIBK.

Therefore,
although
MIBK
would
no
longer
be
listed
as
a
HAP,
the
implementation
of
MACT
standards
will
control
MIBK
emissions
along
with
emissions
of
other
chemicals.

Third,
MIBK
will
continue
to
be
regulated
as
a
VOC.
Many
areas
of
the
country
have
not
yet
reached
attainment
with
the
national
air
quality
standard
(
NAAQS)
for
ozone
and
must
reduce
VOC
emissions
in
order
to
meet
the
NAAQS
standard.
Emissions
of
solvents,

including
MIBK,
are
subject
to
increasingly
stringent
standards
under
both
federal
and
state
programs
designed
to
control
ozone
formation.
Therefore,
emissions
of
MIBK
will
continue
to
be
regulated
even
after
it
is
removed
from
the
HAP
list.

Significantly,
the
monitoring
and
modeling
data
discussed
above
show
that,
even
if
MIBK
emissions
were
to
increase
significantly,
ambient
concentrations
would
be
expected
to
remain
well
below
levels
of
concern.
Actual
monitoring
data
in
a
number
of
areas,
including
industrial
areas,
indicate
that
ambient
concentrations
of
MIBK
are
very
low.
The
air
dispersion
modeling
study
conducted
by
ENSR
showed
that,
in
most
cases,
maximum
annual
off­
site
concentrations
around
facilities
emitting
MIBK
should
be
below
the
RfC
by
an
order
of
54
magnitude
or
more.
Even
at
the
worst­
case
off­
site
locations
around
the
largest
sources
of
MIBK,
ambient
levels
are
below
the
RfC
by
a
factor
of
3
or
more.
Therefore,
even
significant
increases
in
MIBK
emissions
would
not
pose
an
appreciable
risk
to
human
health
or
the
environment.

Finally
­­
and
perhaps
most
importantly
­­
any
increase
in
MIBK
emissions
is
likely
to
be
more
than
offset
by
decreases
in
emissions
of
other
solvents.
As
discussed
in
the
next
section
of
this
Petition,
MIBK
is
a
highly
effective
solvent.
In
many
applications,
relatively
small
amounts
of
MIBK
may
be
used
to
perform
the
same
function
served
by
other,
less
efficient
solvents.
For
this
reason,
it
is
widely
used
in
high­
solids
coatings,
which
offer
the
only
feasible
approach
for
reducing
VOC
emissions
from
certain
types
of
coating
operations.
Because
of
MIBK's
efficiency
as
a
solvent,
increases
in
MIBK
emissions
will
in
many
cases
represent
an
overall
decrease
in
VOC
emissions.

V.
OTHER
REASONS
FOR
DELISTING
MIBK
As
discussed
above,
it
is
clear
that
MIBK
meets
the
statutory
delisting
criteria
and
should
therefore
be
removed
from
the
HAP
list.
EPA
should
be
aware,
however,
that
there
are
additional
considerations
that
weigh
in
favor
of
removing
MIBK
from
the
HAP
list.
These
additional
considerations
are
discussed
below.

A.
Delisting
MIBK
Will
Help
to
Reduce
VOC
Emissions
from
Many
Coating
Operations
As
noted
above,
MIBK
is
especially
valuable
in
the
formulation
of
high­
solids
coatings,
which
are
increasingly
used
to
reduce
VOC
emissions
from
industrial
and
commercial
coating
operations.
MIBK
not
only
dissolves
a
wide
variety
of
resins,
but
is
a
more
efficient
solvent
than
the
available
non­
ketone
alternatives.
Thus,
compared
to
the
alternatives,
a
smaller
55
amount
of
MIBK
may
be
used
to
perform
the
same
function.
The
use
of
MIBK
therefore
allows
the
formulation
of
coatings
with
higher
solids
content
and
lower
VOC
emissions.

In
addition
to
its
solvent
properties,
MIBK
has
unique
chemical
properties
which
are
used
in
the
manufacture
of
compliant
high­
solids
coatings.
MIBK
is
a
good
polymerization
solvent
for
low
molecular
weight
resins,
which
form
the
basic
building
blocks
of
high­
solids
(
low
VOC)
coatings.
Key
characteristics
of
MIBK
­­
including
its
hydrophobicity,
good
stability,
low
hydrogen­
bonding
attributes,
low
surface
tension,
low
viscosity,
and
low
density
­­
make
it
a
leading
choice
for
high­
solids
polymer
manufacturing.
The
resultant
low
viscosity,
high­
solids
polymers
can
then
be
used
to
produce
low
VOC
coatings.

Over
the
last
decade,
EPA
and
many
state
agencies
have
sought
to
reduce
VOC
emissions
from
coating
operations
and
other
commercial
applications
that
involve
the
use
of
organic
solvents.
In
some
cases
­­
particularly
those
involving
large­
scale
coating
operations
­­

the
most
effective
approach
for
reducing
VOC
emissions
is
to
install
a
solvent
recovery
system
or
other
type
of
control
device.
In
other
cases,
companies
have
reduced
their
VOC
emissions
by
switching
from
solvent­
based
technologies
to
alternative,
non­
solvent
technologies.
A
number
of
coating
operations,
for
example,
have
switched
from
conventional
solvent­
based
coatings
to
waterborne
or
powder
coatings.

In
many
cases,
however,
these
options
are
simply
not
feasible.
For
example,
in
many
wood
coating
applications,
water­
based
finishes
cannot
be
used
because
they
are
absorbed
into
the
substrate
and
raise
the
grain
of
the
wood.
Although
a
control
device
may
be
technically
feasible
for
some
wood
finishing
operations,
EPA
has
acknowledged
that
many
such
operations
are
simply
too
small
to
justify
the
installation
of
a
control
device.
Where
it
is
not
practical
to
use
a
control
device
or
a
non­
solvent
technology,
EPA
has
recognized
that
the
best
alternative
is
to
56
use
products
that
can
accomplish
a
given
task
with
the
least
possible
amount
of
solvent.
For
coating
applications,
this
generally
means
a
switch
from
conventional
coatings
to
"
high­
solids"

coatings.
In
several
recent
rulemakings,
EPA
has
adopted
standards
that
will
effectively
require
the
use
of
such
coatings
in
certain
industries.
See,
e.
g.,
61
Fed.
Reg.
19005
(
April
30,
1996)

(
proposed
rule;
automobile
refinishing
coatings);
60
Fed.
Reg.
62930
(
Dec.
7,
1995)
(
final
rule;

wood
furniture
coating
operations);
60
Fed.
Reg.
64330
(
Dec.
15,
1995)
(
final
rule;
shipbuilding
coating
operations).

The
amount
of
solids
in
a
coating
is
limited
by
the
ability
of
the
solvent
to
dissolve
the
resins
and
retain
them
in
solution
until
the
coating
is
applied.
After
the
coating
is
applied,
the
solvent
evaporates
into
the
air,
leaving
behind
a
hard,
uniform
finish.
Thus,
the
more
effective
the
solvent,
the
higher
the
proportion
of
solids
and
the
lower
the
emissions
into
the
air.

EPA
recognized
this
fact
in
its
recent
rule
to
reduce
emissions
from
shipbuilding
operations.
See
60
Fed.
Reg.
at
64330.
In
this
rulemaking,
EPA
acknowledged
that
the
use
of
highly
efficient
solvents
such
as
MIBK
is
the
preferred
environmental
alternative
in
many
coating
applications,
even
though
such
solvents
may
be
listed
as
HAPs.
Although
the
primary
purpose
of
the
rule
is
to
control
HAP
emissions,
EPA
designed
the
rule
to
minimize
VOC
emissions
as
well.

Thus,
the
Agency
adopted
regulatory
standards
that
effectively
require
the
use
of
higher­
solids
coatings
in
the
shipbuilding
industry.

MIBK
and
methyl
ethyl
ketone
(
MEK)
­­
another
ketone
solvent
with
similar
solvency
­­
are
two
of
the
most
efficient
solvents
available
to
coating
formulators.
Because
they
are
currently
listed
as
HAPs,
a
formulator
may
need
to
increase
the
HAP
content
of
a
coating
in
order
to
reduce
its
VOC
content.
The
Agency
explicitly
recognized
this
tradeoff
in
the
preamble
to
the
proposed
rule.
EPA
noted
that
a
coating
reformulated
to
reduce
its
HAP
content
may
have
57
"
higher
VOC
content
than
the
one
it
replaces,"
and
went
on
to
say
that
"
the
HAP
to
VOC
ratio
may
even
increase
when
a
company
develops
a
new
reformulation
with
lower
VOC."
59
Fed.

Reg.
at
62688.
The
Agency
also
noted,
however,
that
even
where
the
HAP
to
VOC
ratio
in
the
coating
increases,
"
the
absolute
HAP
emissions
are
likely
to
go
down,"
presumably
because
higher
solids
coatings
allow
more
coverage
per
gallon
of
coating.
Id.
(
emphasis
in
original).

The
Agency
addressed
this
issue
by
setting
identical
limits
for
the
VOC
and
HAP
content
of
the
coatings
covered
by
the
shipbuilding
rule.
The
rule
sets
a
limit
on
the
amount
of
"
volatile
organic
hazardous
air
pollutants"
(
VOHAPs)
that
can
be
used
in
specified
types
of
coatings.
Because
VOHAPs
are
defined
to
include
both
HAPs
and
VOCs,
a
formulator
may
use
any
solvent
up
to
the
VOHAP
content
limit
for
each
coating.
This
approach
encourages
the
use
of
higher
solids
coatings
and
eliminates
the
incentive
for
formulators
to
use
less
efficient
solvents
that
must
be
used
in
greater
volumes.
Thus,
the
rule
recognizes
that
highly
efficient
solvents
such
as
MIBK
are
the
preferred
environmental
approach
for
reducing
overall
emissions
from
many
coating
applications.

Because
MIBK
is
currently
listed
as
a
HAP,
however,
companies
are
discouraged
or
even
prevented
from
using
it
­­
even
where
it
would
allow
them
to
reduce
their
VOC
emissions
by
switching
to
higher­
solids
coatings.
In
some
industries,
such
as
the
wood
furniture
industry,

facilities
must
comply
with
regulatory
limits
on
the
HAP
content
of
their
coatings.
In
other
industries,
as
already
stated,
companies
are
likely
to
reduce
their
use
of
HAPs
in
order
to
avoid
the
need
to
install
maximum
available
control
technology
(
MACT)
under
Section
112(
d)
of
the
Act.
Even
in
the
absence
of
regulatory
requirements,
companies
often
try
to
avoid
the
use
of
chemicals
that
are
labeled
as
HAPs.
For
example,
the
Ketones
Panel
is
aware
of
a
major
manufacturing
company
that,
as
a
matter
of
corporate
environmental
policy,
is
seeking
to
58
eliminate
the
use
of
MIBK
and
other
listed
chemicals
from
its
manufacturing
processes,
even
though
the
company's
toxicologists
have
recognized
that
MIBK
does
not
pose
significant
toxicity
concerns.
The
simple
fact
that
MIBK
is
listed
as
a
HAP
discourages
companies
from
using
it
­­

even
in
applications
where
it
would
provide
a
clear
environmental
benefit.
Removing
MIBK
from
the
HAP
list
would
eliminate
this
disincentive
and
benefit
the
environment
by
facilitating
the
use
of
higher­
solids
coatings.

B.
EPA
Has
Recognized
in
Other
Contexts
that
MIBK
Has
Relatively
Low
Toxicity
In
two
recent
rulemakings,
EPA
has
reviewed
the
health
effects
of
exposure
to
MIBK.
In
both
cases,
EPA
has
recognized
that
MIBK
has
relatively
low
toxicity.

1.
Proposed
Rule
Under
Section
112(
g)
of
the
Clean
Air
Act
On
April
1,
1994,
the
Agency
published
a
proposed
rule
under
Section
112(
g)
of
the
Act
that
included
a
detailed
system
for
ranking
and
setting
"
de
minimis
values"
for
the
various
chemicals
listed
under
Section
112(
b),
including
MIBK.
59
Fed.
Reg.
15504.
Under
the
proposed
hazard
ranking
system,
the
Agency
developed
a
list
of
"
threshold
pollutants"
that
were
not
considered
"
high
concern"
pollutants,
and
were
believed
to
pose
the
least
risk
of
any
of
the
listed
HAPs.
Not
surprisingly,
MIBK
was
listed
as
a
threshold
pollutant.
For
ranking
the
relative
risk
of
the
compounds
on
the
threshold
list,
the
Agency
assigned
a
"
composite
score"
for
each
chemical
based
on
the
severity
of
any
health
effect
caused
by
the
chemical
in
test
animals
and
the
dose
at
which
the
effect
is
likely
to
occur.
Under
this
system,
a
chemical
could
receive
a
composite
score
from
1
­
100,
although
the
pollutants
on
the
threshold
list
all
had
scores
between
2
and
46.

Based
on
this
proposed
hazard
ranking
system,
the
Agency
assigned
a
composite
score
of
4
to
MIBK,
indicating
that
it
is
among
the
least
hazardous
of
the
chemicals
on
the
list
59
(
approximately
187
out
of
189).
Only
two
compounds
had
a
lower
composite
risk
score
than
MIBK,
and
both
of
them
were
scored
at
3.
Thus,
under
the
hazard
ranking
proposed
by
EPA,

MIBK
was
one
of
the
least
hazardous
chemicals
on
the
HAPs
list.

EPA
also
proposed
a
system
for
setting
de
minimis
values
for
the
various
chemicals
listed
as
HAPs.
The
de
minimis
value
was
the
amount
of
a
chemical
that,
based
on
an
EPA
model,
a
typical
facility
could
emit
without
posing
more
than
a
"
trivial"
risk
to
human
health
or
the
environment.
For
compounds
such
as
MIBK
that
are
non­
carcinogens,
the
values
were
designed
to
ensure
that
public
health
was
protected
with
an
"
ample
margin
of
safety."
Id.
at
15,

525.
The
proposed
de
minimis
values
ranged
from
0.0000006
tons
per
year
to
10
tons
per
year.

For
policy
reasons
unrelated
to
risk,
EPA
"
capped"
de
minimis
levels
at
10
tons
per
year,
but
at
the
same
time
recognized
that,
for
several
low
toxicity
chemicals,
emissions
of
more
than
10
tons
a
year
would
still
pose
only
a
trivial
risk.
Id.
at
15,527.
Not
surprisingly,
the
proposed
de
minimis
level
for
MIBK
was
set
at
the
10
ton
cap.

Significantly,
however,
EPA's
methodology
may
be
used
to
calculate
the
true
"
uncapped"
de
minimis
value
for
MIBK.
This
approach
is
still
conservative
for
at
least
two
reasons.
First,
as
noted
above,
EPA's
approach
for
setting
de
minimis
values
was
specifically
designed
to
allow
an
"
ample
margin
of
safety."
Second,
although
the
EPA
model
used
to
calculate
the
de
minimis
values
was
not
a
"
worst­
case"
model,
the
Agency
recognized
that
it
incorporated
a
number
of
conservative
assumptions.
Id.
at
15,526.
Therefore,
based
on
this
methodology,
the
uncapped
de
minimis
level
for
MIBK
derived
from
an
RFC
of
2.4
mg/
m3
would
have
been
5,000
tons
per
year.
Based
on
the
most
recent
TRI
data,
the
facility
with
the
highest
MIBK
emissions
in
the
country
emits
only
about
10
percent
of
this
amount.
60
2.
Final
SNAP
Rule
Under
Section
612
of
the
Clean
Air
Act,
EPA
has
developed
a
program
­­
called
the
Significant
New
Alternatives
Policy
(
SNAP)
program
­­
to
identify
acceptable
substitutes
for
chemicals
that
are
being
phased
out
of
production
because
they
deplete
the
stratospheric
ozone
layer.
59
Fed.
Reg.
13,044
(
March
18,
1994).
Under
the
SNAP
program,
the
Agency
specifically
evaluated
the
toxicity
of
MIBK
and
listed
it
as
an
acceptable
substitute
in
a
number
of
applications.
In
the
final
SNAP
rule,
EPA
discussed
concerns
about
possible
risks
posed
by
petroleum
hydrocarbons
and
concluded
that
these
risks
were
relatively
small
and
were
adequately
addressed
by
existing
regulations
and
work
practices.
The
Agency
then
discussed
the
use
of
oxygenated
hydrocarbons
and
stated
that
"
two
of
the
typical
oxygenated
hydrocarbons
examined
in
the
Agency's
risk
screen,
methyl
ethyl
ketone
and
methyl
isobutyl
ketone,
also
have
comparatively
low
toxicity."
59
Fed.
Reg.
at
13,120.
Thus,
EPA
has
recognized
that
MIBK
has
relatively
low
toxicity
and
that,
under
some
circumstances,
the
use
of
MIBK
as
a
substitute
actually
helps
to
protect
the
environment.

C.
The
Inclusion
of
MIBK
on
the
Initial
HAP
List
Was
Not
Based
on
a
Finding
of
Adverse
Health
or
Environmental
Effects
The
inclusion
of
MIBK
on
the
initial
HAP
list
was
not
based
on
a
finding
of
adverse
health
or
environmental
effects.
The
initial
HAP
list
was
developed
from
the
list
of
chemicals
that
are
reportable
under
Section
313
of
EPCRA.
The
Section
313
list
was
a
compendium
of
the
New
Jersey
"
Environmental
Hazardous
Substance
List"
and
a
Maryland
"
Survey
List."
These
two
lists
were
combined
in
Committee
Print
No.
99­
169
of
the
Senate
Committee
on
Environment
and
Public
Works,
entitled
"
Toxic
Chemicals
Subject
to
Section
313
of
the
Emergency
Planning
and
Community
Right­
to­
Know
Act
of
1986."
The
combined
list
61
constitutes
the
initial
list
of
chemicals
subject
to
the
reporting
requirements
of
Section
313.
See
EPCRA
Section
313(
c).

At
the
time
EPCRA
was
enacted,
MIBK
was
not
on
the
New
Jersey
list.
When
New
Jersey
enacted
the
Worker
and
Community
Right­
to­
Know
Act
(
codified
at
N.
J.
A.
C.

§
7:
1G­
1,
et
seq.),
it
compiled
a
list
of
250
chemicals
for
careful
review
based
on
three
criteria:

whether
the
chemical
(
1)
presented
a
public
health
hazard;
(
2)
was
an
environmental
contaminant;

or
(
3)
was
present
in
the
State
in
quantities
of
10,000
pounds
or
more.
New
Jersey
then
gathered
information
about
the
production,
use
and
effects
of
these
250
chemicals
from
a
number
of
sources
and
evaluated
each
chemical
for
inclusion
on
its
list
based
on
two
criteria:
(
1)
evidence
of
significant
production
in
New
Jersey
and
(
2)
evidence
of
health
or
environment
effects
such
as
carcinogenicity,
teratogenicity,
mutagenicity,
acute
toxicity,
persistence
and
ability
to
bioaccumulate.
See
New
Jersey
Department
of
Environmental
Protection,
"
Worker
and
Community
Right­
to­
Know
Basis
and
Background
Document."
Notably,
New
Jersey
did
not
include
MIBK
in
its
final
Hazardous
Substance
list.

MIBK
was
included
in
the
Maryland
Survey
List,
but
its
inclusion
did
not
reflect
a
finding
of
adverse
health
or
environmental
effects.
The
Maryland
Survey
List
was
informally
developed,
based
on
a
variety
of
federal
and
state
lists,
for
purposes
of
information
gathering
by
the
State
regarding
chemical
usage
in
Maryland.
The
Survey
List
eventually
was
used
to
survey
Maryland
businesses
to
determine
the
production
and
use
levels
in
the
State
of
Maryland
for
each
chemical.

Therefore,
the
inclusion
of
MIBK
on
the
initial
HAP
list
does
not
reflect
a
determination
by
Congress,
EPA,
the
states
of
New
Jersey
or
Maryland,
or
anyone
else
that
62
MIBK
meets
the
listing
criteria.
Moreover,
the
circumstances
surrounding
the
creation
of
the
initial
list
do
not
create
any
presumption
against
delisting.

CONCLUSION
The
Ketones
Panel
respectfully
submits
that
MIBK
meets
the
delisting
criteria
set
forth
in
Section
112(
b)(
3)(
C)
of
the
Clean
Air
Act.
Accordingly,
the
Administrator
should
grant
this
Petition
and
remove
MIBK
from
the
list
of
chemicals
that
are
regulated
as
hazardous
air
pollutants
under
the
Clean
Air
Act.

DC_
DOCS\
18399.1
[
W97]
