Environmental
Technology
Council
734
15th
Street,
N.
W.
 
Suite
720
 
Washington,
DC
20005
 
(
202)
783­
0870
____________________________________________________________________________________

January
17,
2006
Mr.
James
Berlow
USEPA
Headquarters
Ariel
Rios
Building
1200
Pennsylvania
Avenue,
N.
W.
Mail
Code:
5302W
Washington,
DC
20460
Re:
Amendments
to
Comparable
Fuels
Exclusion:
Options
Under
Consideration
Dear
Jim:

Thank
you
for
holding
the
public
stakeholder
meetings
on
the
options
under
consideration
for
possible
revisions
to
the
comparable
fuels
exclusion
(
CFE)
in
40
CFR
261.38.
At
the
meetings,
you
suggested
that
interested
parties
could
submit
their
comments
in
writing.
The
ETC
has
a
number
of
concerns
about
the
lack
of
an
adequate
basis
for
proceeding
with
this
proposed
rulemaking,
which
are
summarized
in
these
comments
for
the
record.
While
we
have
not
had
time
to
thoroughly
evaluate
the
options,
or
to
conduct
separate
studies
related
to
the
analytical
issues,
we
wanted
to
submit
our
preliminary
views
early
in
the
process.

Option
A:
Nondetect­
Based
Specifications
for
Volatile
Hydrocarbons
and
Oxygentates
 
Correct
the
Specifications
to
Base
Them
on
Gasoline
Quantitation
Limits
Rather
Than
Fuel
Oil
Quantitation
Limits
Option
A
would
base
the
detection
limits
for
certain
hydrogenated
and
oxygenated
constituents
in
CFE
waste
on
gasoline.
The
current
specifications
are
based
on
fuel
oil
detection
limits,
and
EPA
is
considering
whether
CFE
wastes
are
more
like
gasoline
than
fuel
oil,
and
therefore
these
detection
limits
might
not
be
achievable.

EPA
first
needs
to
address
many
unanswered
questions
before
proposing
such
revisions
for
the
following
reasons:

 
The
ACC
survey
does
not
indicate
that
detection
limits
are
a
problem
for
most
candidate
waste
streams,
so
it
is
unclear
why
EPA
should
undertake
a
major
new
rulemaking
to
revise
the
current
CFEs.

 
The
CFE
Background
Documents
and
ECI
paper
submitted
by
ACC
do
not
adequately
support
the
need
to
revise
the
detection
limits
for
candidate
wastes.
The
ECI
paper
was
based
on
tests
in
only
one
laboratory,
looking
at
one
2
technique
for
analyzing
CFE
waste
that
was
biased
towards
high
detection
limits.

 
The
assumption
that
CFE
waste
is
more
like
gasoline
than
fuel
oil
has
not
been
demonstrated,
either
in
the
comp
fuel
background
documents
or
other
studies,
and
requires
substantial
additional
analysis.

Each
of
these
points
is
discussed
in
detail
below.

1.
The
ACC
survey
does
not
indicate
that
detection
limits
are
an
issue
for
a
significant
number
of
waste
streams.

Most
of
the
96
respondents
to
the
ACC
survey
did
not
indicate
that
their
waste
failed
the
current
CFE
due
to
detection
limits,
but
rather
for
the
following
reasons:

 
The
waste
contains
high
percentage
levels
of
the
failing
constituent.
Those
respondents
claiming
that
their
wastes
have
a
problem
with
a
hydrogenated
or
oxygenated
constituent
are
seeking
very
high
percentage
levels
for
the
waste,
well
above
the
gasoline
detection
limit.
For
example,
respondent
D­
04­
13
would
like
to
see
the
toluene
standard
revised
to
50%
and
respondent
D­
05­
11
would
like
to
see
the
MEK
standard
revised
to
2%
and
the
isobutyl
alcohol
standard
revised
to
5%.
These
are
well
above
the
detection
levels
for
gasoline.

 
The
waste
contains
levels
of
metals,
halogenated
organics,
reactive
cyanide
or
some
other
non­
hydrogenated
or
non­
oxygenated
constituent
that
fails
the
CFE.
This
is
the
case
with
26
of
the
96
waste
streams,
or
23%
of
the
quantity
of
waste
surveyed
by
ACC.

 
Finally,
15
of
the
96
respondents
in
the
ACC
survey
indicated
no
issues
with
the
standards
at
all.
Several
respondents
had
not
evaluated
their
waste
relative
to
the
standards
and/
or
did
not
indicate
what
the
problem
was
with
meeting
the
CFE.

The
ETC
found
survey
responses
for
only
4
waste
streams
that
might
have
detection
limit
issues.
These
are
summarized
below:

 
Respondent
F­
01­
13
failed
for
only
methyl
methacrylate
and
is
seeking
a
limit
of
890
mg/
kg.

 
Respondent
F­
01­
16
failed
for
only
isobutyl
alcohol
and
is
seeking
a
limit
of
222
mg/
kg.

 
Respondent
K­
01­
12
stated
that
achieving
the
non­
detect
level
for
acrolein
and
isobutyl
alcohol
was
an
issue,
but
did
not
state
what
level
could
be
achieved.
3
 
Respondent
K­
04­
12
stated
that
only
acrolein
was
an
issue
and
could
meet
a
standard
of
500
mg/
kg.

Thus,
only
4
waste
streams
and
3
constituents
(
methyl
methacrylate,
isobutyl
alcohol,
and
acrolein)
appear
to
pose
any
problems
with
detection
limits.
For
the
three
problematic
constituents,
the
respondents
stated
that
they
could
meet
detection
levels
of
222
to
890
mg/
kg.
These
levels
are
far
below
EPA's
revised
specification
under
consideration
of
3,400
mg/
kg
based
on
gasoline.
Therefore,
none
of
the
ACC
survey
data
supports
any
need
to
revise
the
CFE
standards
to
the
levels
under
consideration
relative
to
a
gasoline
matrix.

Furthermore,
even
for
these
levels
of
222
to
890
mg/
kg,
it
is
not
clear
from
the
survey
if
these
were
non­
detect
levels,
or
the
levels
actually
present
for
which
the
respondents
wanted
the
exemption
raised.
It
is
possible
for
these
cases
that
the
respondents
could
detect
the
current
exemption
standard
of
39
mg/
kg,
and
merely
want
the
limit
raised
to
the
level
of
222
to
890
mg/
kg
since
this
is
the
level
present
in
their
waste.
The
reality
is
that
only
one
respondent
clearly
raised
detection
limits
as
an
issue,
but
did
not
provide
any
data
on
what
detection
limit
was
achievable
for
the
waste
matrix.

The
ACC
survey
does
not
indicate
a
widespread
problem
with
detection
limits
or
a
need
to
change
the
exemption
levels.
Before
embarking
on
a
new
rulemaking
that
will
consume
scarce
Agency
and
stakeholder
resources,
we
request
that
a
convincing
case
be
made
that
the
detection
limits
for
volatile
hydrocarbons
and
oxygenates
are,
in
fact,
preventing
a
substantial
use
of
wastes
as
comparable
fuels.
EPA
should
request
that
ACC
conduct
a
more
detailed
and
careful
survey
to
determine
the
actual
amount
of
hazardous
waste
that
will
qualify
for
the
CFE
under
Option
A
before
going
forward
with
a
time­
consuming
and
expensive
rulemaking.

2.
There
is
insufficient
evidence
of
analytical
issues
with
the
current
standards
based
on
fuel
oil,
and
the
ECI
paper
fails
to
demonstrate
a
major
problem.

Before
EPA
proposes
new
limits
based
on
gasoline,
the
agency
should
conduct
or
obtain
studies
clearly
showing
that
the
detection
limits
for
CFE
waste
are
not
achievable
at
a
reasonable
cost.
ACC
has
submitted
a
paper
prepared
by
Environmental
Chemistry,
Inc.
(
ECI),
a
commercial
laboratory
in
Houston,
as
support
for
raising
the
current
exemption
levels
based
on
detection
limit
issues.
However,
the
ECI
paper
fails
to
make
an
adequate
case
that
the
current
detection
limits
are
not
achievable.

The
ECI
paper
indicates
that
purge­
and­
trap
analysis
was
performed
on
the
samples
after
extensive
dilution
of
the
samples.
The
samples
were
first
diluted
or
extracted
in
methanol,
and
then
a
"
sub­
sample"
of
this
was
added
to
water
in
a
purge
vessel.
ECI
does
not
provide
any
information
on
the
amount
of
these
dilutions
relative
to
the
original
weight
of
sample,
making
it
difficult
to
determine
if
less
dilution
would
have
been
possible.
Less
dilution
would
have
resulted
in
better
detection
limits.
4
Having
pre­
diluted
the
samples,
ECI
could
have
used
direct
injection
techniques
to
analyze
the
samples.
This
would
have
provided
better
detection
limits.
Instead,
ECI
performed
purge­
and­
trap
after
diluting
the
samples
further
in
water
for
the
purge
vessel.
This
biased
the
results
towards
poorer
detection
limits.
ECI
does
not
adequately
justify
using
purge­
and­
trap,
merely
stating
that
direct
injection
was
not
considered
because
of
"
the
potential
problems
associated
with
it."
No
description
of
these
"
potential"
problems
is
provided.
ECI
does
admit
that
direct
injection
may
achieve
lower
(
better)
detection
limits.
Purge­
and­
trap
techniques
are
typically
used
for
water
and
wastewater
matrices
for
which
this
method
was
designed.
It
is
not
common
laboratory
practice
to
dilute
a
methanol
extract
of
a
hazardous
waste
matrix,
and
then
take
that
methanol
extract
and
introduce
it
to
water
to
do
purge­
and­
trap.
ECI's
choice
of
sample
preparation
and
instrument
introduction
of
the
sample
is
questionable,
and
appears
to
be
clearly
biased
towards
poorer
detection
limits.

It
is
also
noteworthy
that
in
developing
the
1998
CFE
exemption
levels,
EPA's
contractor
used
direct
injection
and
not
purge
and
trap
techniques
for
the
range
of
fuels
studied,
even
gasoline
(
see
Technical
Support
Document
for
HWC
MACT
Standards,
Volume
VI,
Development
of
Comparable
Fuel
Specifications,
pp.
2­
15
and
2­
16).
The
use
of
purge
and
trap
method
for
introducing
the
sample
to
the
GC/
MS
is
never
even
considered
and
discussed,
likely
because
it
is
not
the
right
technique
for
waste
samples,
and
it
would
bias
the
detection
limits
high.

Further,
ECI
provides
no
QA/
QC
data
to
support
its
findings.
The
paper
by
ECI
is
not
a
valid
MDL
study
because
no
QA/
QC
data
is
provided,
nor
any
kind
of
comprehensive
data
package
normally
required
for
the
most
basic
of
laboratory
reports
performed
for
regulatory
purposes.
In
the
early
1990s
the
ETC
performed
a
comprehensive
PQL
Study
on
Incinerator
Ash
matrices
in
response
to
certain
proposed
changes
to
the
land
disposal
restriction
regulations.
The
EPA
required
the
ETC
to
submit
extensive
data
quality
packages
and
extensive
QA/
QC
data
to
support
the
findings.
Before
initiating
the
study,
a
detailed
work
plan
had
to
be
submitted
to
EPA
for
the
PQL
Study
for
review,
comment
and
approval.
This
work
plan
as
well
as
the
final
data
package
underwent
extensive
review
by
David
Friedman
of
EPA's
Waste
Characterization
Branch
as
well
as
by
other
analytical
chemistry
experts
within
EPA.
EPA
issued
comments
on
the
study
and
additional
work
had
to
be
completed
and
submitted
to
further
substantiate
the
data.
EPA
appears
to
have
done
no
such
critical
review
of
the
ECI
work,
and
is
accepting
the
results
on
faith
without
any
QA/
QC
or
other
data
quality
support
documentation.

Even
if
ECI
had
used
non­
biased
techniques
and
had
provided
QA/
QC
and
other
data
quality
documentation,
EPA
should
not
be
basing
any
conclusions
on
the
work
of
only
one
laboratory.
The
ETC's
PQL
Study
had
to
be
designed
as
a
multi­
laboratory
study
to
validate
the
detection
limits
achievable
on
incinerator
ash
matrices
over
a
range
of
different
laboratories.
A
total
of
7
laboratories
had
to
be
used
to
justify
the
detection
limits
on
an
inter­
laboratory
basis.
The
detection
limits
achieved
in
one
lab
do
not
justify
changes
to
national
regulations,
particularly
one
that
is
exempting
waste.
5
Before
EPA
can
pursue
Option
A
further,
ACC
needs
to
develop
a
work
plan
with
clear
data
quality
objectives
and
QA/
QC
requirements.
It
must
be
performed
on
a
multiple
laboratory
basis
and
must
also
include
validation
work
within
EPA's
own
laboratory.
The
work
plan
should
include
evaluation
of
a
variety
of
sample
preparation
techniques,
including
direct
injection.
The
cost
factors
associated
with
completing
the
analysis
should
also
be
part
of
the
work
plan.

In
addition,
EPA
should
seriously
consider
basing
Option
A
detection
limits
on
waste
streams
that
have
qualified
for
the
current
CFE
exemption
and
not
gasoline.
EPA
should
use
samples
of
these
wastes
as
the
basis
for
detection
limit
studies.
The
question
is
what
detection
limits
are
achievable
for
hazardous
waste
matrices,
not
gasoline.

Of
the
list
of
hydrocarbon
and
oxygenated
constituents
in
40
CFR
261.38,
the
vast
majority
of
constituents
in
the
ECI
paper
clearly
did
not
present
detection
limit
issues
relative
to
the
current
CFE.
There
were
only
8
constituents
that
the
ECI
study
showed
presented
detection
limit
problems,
even
with
the
flawed
analytical
methods
they
applied.
It
is
not
clear
from
EPA's
options
paper
if
the
changes
to
the
detection
limits
would
be
limited
to
this
subset
of
8
constituents.
EPA
needs
to
clarify
this,
and
if
so,
further
studies
on
these
8
constituents
are
needed
as
discussed
above.
Note
that
of
these
8
constituents,
ECI
did
not
have
completed
detection
limit
evaluations
for
2
constituents
(
2­
ethoxyethanol
and
propargyl
alcohol)
because
of
unspecified
"
analytical
problems
during
analysis."
For
the
remaining
6,
ECI
was
able
to
achieve
detection
limits
ranging
from
136
mg/
kg
to
816
mg/
kg
and
averaging
476
mg/
kg.
Therefore,
raising
the
current
CFE
limits
to
the
3,400
mg/
kg
level
of
gasoline
is
truly
not
justified,
even
by
the
flawed
ECI
study.

As
a
final
point,
it
is
worth
emphasizing
that
even
with
its
flaws,
the
ECI
study
did
not
justify
raising
the
current
exemption
levels
to
those
of
gasoline.
For
the
three
constituents
identified
in
the
ACC
survey
as
potentially
having
detection
limit
issues,
the
detection
levels
achieved
by
ECI
were
lower
than
the
level
EPA
is
suggesting
for
changes
of
3400
mg/
kg
based
on
gasoline.
For
acrolein,
isobutyl
alcohol
and
MEK,
the
ECI
study
showed
a
detection
level
of
544
mg/
kg,
far
lower
than
EPA's
suggested
revised
standard
of
3400
mg/
kg.
The
EPA
should
not
be
considering
detection
limits
that
are
substantially
higher
than
is
supported
by
the
data
in
both
the
ACC
survey
and
the
ECI
paper.

3.
The
assumption
that
CFE
waste
is
more
like
gasoline
than
fuel
oil
has
not
been
substantiated.

EPA
has
not
adequately
justified
why
gasoline
is
an
appropriate
benchmark
for
CFE
waste.
The
CFE
waste
that
is
covered
in
the
ACC
survey
is
not
gasoline­
like
and
no
industry
uses
gasoline
to
fire
industrial
furnaces
or
boilers.
EPA's
work
on
the
1998
CFE
included
a
survey
of
CMA
(
ACC)
members
(
see
Section
2.0
of
the
1998
Background
Document).
One
of
the
survey
questions
was
regarding
what
fossil
fuels
are
used
to
fire
their
industrial
boiler
and/
or
furnace
systems.
None
of
the
ACC
members
responded
that
gasoline
was
used
as
a
fuel
for
these
burner
units.
The
fuels
consisted
of
Number
2,
4
or
6
fuel
oil
or
natural
gas.
In
contrast,
there
are
actual
hazardous
wastes
that
have
qualified
6
for
the
CFE,
as
well
as
the
96
waste
streams
covered
in
the
survey,
that
would
represent
the
actual
matrix
and
be
better
candidates
for
analytical
detection
limit
studies.

If
a
given
waste
presents
detection
limit
issues,
this
is
usually
due
to
high
percentage
level
of
some
other
organic
constituent(
s)
that
present
a
response
background
against
which
the
target
analyte
is
difficult
to
distinguish.
In
such
cases
it
is
very
possible
that
the
major
constituent
causing
the
detection
limit
issue
is
a
hazardous
organic
constituent
that
would
preclude
the
waste
from
being
exempted.
For
each
waste
in
the
ACC
survey,
ACC
should
identify
what
these
major
constituents
are
and
if
these
are
RCRA
Appendix
VIII
constituents
the
waste
should
not
be
eligible
for
the
CFE.

Certain
constituents
in
the
CFE
are
so
highly
toxic
that
the
current
exemption
levels
should
not
be
raised
absent
a
compelling
scientific
basis.
Given
the
highly
toxic
nature
of
certain
carcinogens
like
acrolein,
benzene,
naphthalene,
and
other
PNA
constituents,
EPA
should
not
be
raising
these
exemption
levels
based
on
the
very
limited
record
before
the
agency.

EPA
should
consider
retaining
the
current
CFE
and
allowing
generators
to
justify
the
nature
of
any
interference
in
a
site­
specific
petition.
Given
that
the
testing
requirements
are
infrequent,
the
burden
for
doing
additional
analytical
work
is
a
small
price
given
the
nature
of
the
exemption.
Analytical
expertise
exists
within
EPA
and
most
States
to
critically
review
the
data
package
for
a
given
waste
stream
and
determine
the
validity
of
the
interference
or
other
analytical
aspects
that
are
causing
problems
with
the
current
levels.

Option
B:
Comparable
Emissions:
Conditional
Exclusion
of
Fuels
That
Are
Off­
Spec
for
Hydrocarbons
and
Oxygentates
Where
the
Conditions
Ensure
Emissions
Are
Comparable
to
Benchmark
Fuels
EPA
has
called
Option
B
a
"
Comparable
Emissions"
option
for
hydrocarbon
and
oxygenated
constituents
which
is
fundamentally
different
from
the
current
CFE
in
40
CFR
261.38.
Option
B
presumes
that
emissions
will
be
comparable
to
fossil
fuels
at
any
level
of
hydrocarbons
and
oxygenates
burned,
with
proposed
emission
restrictions
only
for
carbon
monoxide,
hydrocarbons,
and
the
heat
content
of
the
hazardous
waste.
For
example,
Option
B
would
allow
a
hazardous
waste
with
90%
benzene
to
be
fed
as
comparable
fuel,
with
the
expectation
that
the
emissions
would
be
no
different
from
feeding
a
waste
containing
the
3500
mg/
kg
(
0.35%)
benzene
normally
found
in
gasoline
or
the
21
mg/
kg
(
0.0021%)
normally
found
in
a
fuel
oil.

EPA
should
request
emissions
data
from
ACC
to
validate
this
assumption.
The
public
cannot
accept
on
faith
that
emissions
from
burning
90%
benzene
in
an
unregulated
unit
will
be
the
same
as
emissions
from
burning
less
than
0.35%
benzene
in
gasoline
or
fuel
oil.
EPA
argues
that
carbon
monoxide
or
total
hydrocarbon
monitoring
will
provide
assurances.
Such
continuous
monitoring
is
acceptable
for
a
RCRA
permitted
incinerator
or
cement
kiln
because
numerous
other
controls
are
in
place
such
as
maximum
permitted
feed
rates,
minimum
combustion
temperature,
minimum
combustion
air
flows,
and
a
7
multitude
of
parameters
governing
the
air
pollution
control
systems.
In
addition,
a
cement
kiln
or
incinerator
has
to
verify
on
a
site­
specific
basis
that
CO
or
THC
correlate
with
safe
overall
emissions
by
testing
individual
HAPs
during
Trial
Burn,
Risk
Burn
and
MACT
CPT
tests
performed
on
a
regular
basis.
Under
EPA's
proposal,
no
such
testing
would
be
required
of
units
burning
comparable
fuels.
EPA
has
not
done
any
validation
that
emissions
can
be
expected
to
be
"
comparable"
on
any
unit.

EPA
should
require
emissions
data
from
the
ACC
to
substantiate
the
claim
of
comparable
emissions.
In
addition,
EPA
should
collect
comprehensive
emissions
data
to
substantiate
that
the
burning
of
a
90%
benzene
waste,
as
well
as
any
waste
containing
high
percentage
levels
of
the
other
36
hydrocarbon
and
oxygenated
constituents,
are
truly
"
comparable"
in
emissions
to
fossil
fuels.
Certain
of
these
compounds,
such
as
benzene,
naphthalene,
several
PNAs,
and
acrolein
are
carcinogenic
or
highly
toxic
HAPs.
Therefore,
EPA
cannot
just
assume
that
emissions
resulting
from
unregulated
combustion
of
90%
levels
of
these
constituents
will
be
comparable
to
fuels
containing
levels
well
below
0.5%
of
these
constituents.

EPA
believes
that
carbon
monoxide
or
hydrocarbon
levels
in
the
combustion
gas
are
well
correlated
with
emissions
of
other
organic
pollutants.
The
basis
for
this
belief
is
the
Draft
Technical
Support
Document
for
HWC
MACT
Standards
(
NODA),
Volume
II:
Evaluation
of
CO/
HC
and
DRE
Database"
to
support
that
CO
or
HC
monitoring
is
sufficient.
A
careful
review
of
this
document
raises
several
questions
and
issues.
The
document
starts
off
with
a
review
of
an
October
1984
report
titled
"
Theoretical
Evaluation
of
Exhaust
Emissions
of
CO
and
HC
as
Indicators
of
Incinerator
Performance."
The
conclusion
of
this
report
is
that
"
in
most
cases"
CO
and
HC
"
should"
be
a
direct
indicator
of
DRE
failure.
This
report
then
presents
4
examples
of
DRE
Failure
Modes
in
which
CO
and
HC
may
be
helpful.
This
theoretical
report,
however,
does
not
conclude
that
CO
and
HC
would
always
serve
as
complete
indicators
of
DRE.
In
addition,
many
studies
after
1984
indicate
cases
in
which
CO
or
HC
are
not
complete
indicators
in
isolation,
as
discussed
further
below.
That
is
why
the
report
uses
statements
like
"
in
most
cases"
and
"
should"
to
qualify
the
tentative
conclusions.

The
support
document
then
reviews
the
1997
MACT
database
for
which
DRE
data
is
available
to
compare
with
HC
or
CO
data.
The
document
notes
the
following:

 
DRE
Failure
occurs
2%
of
the
time
when
CO
is
less
than
100
ppm,
and
14%
of
the
time
when
CO
is
greater
than
100
ppm.

 
DRE
Failure
occurs
3%
of
the
time
when
HC
is
less
than
12
ppm,
and
18%
of
the
time
when
HC
is
greater
than
12
ppm.

There
are
two
ways
to
look
at
these
data.
The
optimistic
way
is
to
say
that
there
are
more
cases
where
CO
or
HC
works
as
an
indicator
of
DRE.
This
is
what
EPA
is
doing
when
considering
the
data
on
a
percentage
basis.
Note,
however,
that
the
statement
in
the
support
document
is
equivalent
to
saying
that
84%
of
the
CO
cases
and
89%
of
the
HC
cases
are
indeterminable.
8
The
more
realistic
way
to
view
this
data
is
that
CO
and
HC
are
not
conclusive
indicators
of
DRE
performance
apart
from
other
control
parameters
that
apply
to
RCRA
permitted
combustion.
In
fact,
the
data
clearly
show
that
DRE
is
not
always
correlated
with
CO
or
HC.

Here
are
other
conclusions
that
can
be
drawn
from
the
same
data
set
evaluated
in
EPA's
support
document:

 
17
combustion
units
had
DRE
failures
even
though
HC
or
CO
were
below
12
ppm
or
100
ppm
respectively.

 
18
combustion
units
had
DRE
failures
when
HC
or
CO
were
above
12
ppm
or
100
ppm
respectively.
These
data
show
that
just
as
many
units
failed
DRE
above
or
below
the
HC
and
CO
thresholds
 
83
combustion
units
had
good
DRE
even
though
CO
was
above
100
ppm;
18
facilities
had
good
DRE
even
though
THC
was
above
12
ppm.
These
data
show
that
CO
or
HC
alone
do
not
support
good
DRE,
and
that
other
factors
are
influencing
DRE
as
well.

Therefore,
the
1997
Technical
Support
Document
should
not
be
viewed
as
conclusive
evidence
that
CO
and
HC
alone
assure
good
DRE.
Indeed,
17
combustion
units
is
not
a
small
number
of
units,
and
clearly
for
these
units
poor
DREs
were
experienced
despite
low
CO
or
HC
data.

The
support
document
does
try
to
explain
some
of
the
cases
in
which
CO
and
HC
did
not
work
as
indicators
of
DRE
performance.
One
of
the
parameters
noted
is
combustion
temperature.
The
document
states
on
page
8
that
DRE
failures
occur
more
often
at
low
combustion
temperatures.
Yet
temperature
is
not
considered
in
Option
B
as
important
to
control
to
ensure
good
DRE.

The
support
document
then
analyzes
only
7
of
the
17
data
points
where
DRE
was
poor
with
low
CO
or
HC.
Without
citing
specific
engineering
arguments,
for
5
of
these
7
cases
EPA
concludes
that
the
bad
DRE
performance
"
was
not
representative
of
the
performance
of
the
facility"
simply
because
other
passing
DRE
data
was
available
from
other
tests
at
these
same
facilities.
In
other
words,
the
support
document
arbitrarily
dismisses
the
data
as
outliers,
simply
because
the
facility
met
DRE
at
another
time.
But
clearly
the
DRE
failure
was
real,
and
the
CO
and/
or
HC
data
did
not
predict
the
performance.
For
the
other
2
of
the
7
cases
examined,
the
document
concludes:
"
This
is
a
classic
example
of
a
rare
situation
where
CO
and/
or
HC
will
not
catch
a
DRE
failure."
We
do
not
believe
that
2
out
of
7
cases
means
DRE
failure
is
"
rare,"
and
in
reality
this
sentence
is
an
admission
that
CO
and
HC
will
not
always
protect
against
a
DRE
failure.

Figure
3
of
the
1997
Support
Document
shows
Penetration
(
DRE)
plotted
against
CO
levels.
EPA
uses
this
graph
for
one
purpose,
to
identify
the
7
data
sets
discussed
9
above
for
further
discussion.
But
this
Figure
shows
one
overwhelming
picture
that
EPA
never
discusses;
namely,
that
a
plot
of
Penetration
(
DRE)
vs
CO
is
in
essence
a
scatter
diagram,
with
no
correlation
or
trend.
Figure
3
is
clear
evidence
that
CO
in
isolation
cannot
be
used
to
ensure
good
DRE
of
CFE
waste
burned
in
un­
regulated
units.

The
manner
in
which
the
data
were
filtered
to
remove
a
substantial
number
of
DRE
data
points
that
were
considered
to
be
difficult
POHCs
is
also
highly
questionable.
It
is
possible
that
many
more
DRE
failure
cases
were
filtered
out
by
EPA
that
would
have
shown
additional
poor
correlation
with
CO
and/
or
HC.

In
sum,
the
1997
Technical
Support
Document
does
not
support
that
CO
and/
or
HC
will
ensure
DRE
performance
or
low
organic
emissions
for
CFE
waste.
Before
proceeding,
EPA
should
collect
a
great
deal
more
actual
emissions
data
to
justify
that
CFE
waste
will
not
produce
increased
exposure
to
toxic
organic
contaminants
to
the
public
and
the
environment.
The
support
document
clearly
does
not
indicate
that
the
emissions
from
burning
a
waste
containing
90%
benzene
will
be
"
comparable"
to
emissions
from
a
fuel
containing
under
0.5%
benzene.
Comparable
emissions
have
not
been
demonstrated
and
will
not
be
guaranteed
with
controls
only
in
place
for
solely
CO
or
HC.

In
addition,
EPA's
Final
Technical
Support
Document
for
HWC
MACT
Standards,
Volume
IV,
dated
July
1999,
discusses
in
Chapter
10
the
operating
parameter
limits
needed
to
control
DRE.
This
section
clearly
notes
that
besides
CO
or
HC,
controls
are
also
needed
for
minimum
combustion
temperature,
maximum
combustion
gas
flow
rate,
maximum
waste
feed
rates,
and
other
control
parameters
associated
the
operation
of
each
hazardous
waste
firing
system.
These
must
be
monitored
on
a
continuous
basis.
This
Background
Document
therefore
provides
further
support
that
CO
or
HC
alone
are
not
adequate
to
provide
for
DRE.
EPA
must
therefore
require
similar
controls
under
Option
B.

In
addition,
the
control
of
CO
or
HC
does
not
protect
against
increase
emissions
of
Products
of
Incomplete
Combustion
(
PICs)
as
well
as
dioxin
and
furan
(
D/
F)
emissions.
Both
PIC
and
D/
F
emissions
can
be
substantially
higher
for
CFE
wastes
burned
in
non­
RCRA
regulated
units.
EPA
does
not
have
any
data
that
characterizes
emissions
of
D/
F
and
PICs
from
the
burning
of
CFE
wastes,
or
that
correlates
CO
and
HC
with
PIC
and
D/
F
emissions.
The
papers
cited
below
show
that
this
is
a
substantial
concern.

Hall,
Dellinger,
Graham
and
Rubey1
performed
studies
on
a
12
component
mixture
in
a
laboratory
scale
thermal
destruction
unit.
The
mixture
included
certain
of
the
constituents
in
the
CFE
including
toluene,
MEK,
1,4­
dioxane
and
acetonitrile.
The
1
Hall,
D.
L.,
Dellinger,
B.,
Graham,
J.
L.,
and
Rubey,
W.
A.
"
Thermal
Decomposition
Properties
of
a
Twelve
Component
Organic
Mixture."
Hazardous
Waste
&
Hazardous
Materials,
Volume
3,
November
4
1986,
pg
441­
449.
Lieber,
Inc.
Publishers.
10
data
showed
substantial
emissions
of
these
compounds
under
a
wide
range
of
CO
levels.
The
authors
concluded
as
follows:

"
Although
changes
in
CO
concentrations
as
measured
in
the
effluent
of
a
full
scale
thermal
destruction
device
may
be
an
indicator
of
severe
system
upset,
data
presented
here
indicates
that
CO
concentration
correlates
poorly
with
gas­
phase
destruction
efficiencies
for
this
mixture
of
12
organic
compounds.
The
apparent
lack
of
correlation
between
CO
levels
and
changes
in
the
decomposition
of
the
constituents
of
the
mixture
places
doubt
on
the
usefulness
of
CO
measurements
as
an
indicator
of
destruction
efficiency."

Dellinger
has
performed
numerous
studies
on
the
mechanisms
of
PIC
formation.
His
studies
show
that
the
concentrations
of
PICs
in
stack
emissions
of
combustors
can
be
as
high
or
higher
than
POHCs.
2
Dellinger
presents
data
on
PIC
levels
in
thermal
degradation
studies
showing
levels
comparable
to
POHCs,
and
notes
that
"
the
parameters
affecting
PIC
formation
are
numerous
and
complex."
In
1989
EPA's
Science
Advisory
Board
evaluated
the
data
available
at
the
time
regarding
PIC
emissions
and
EPA's
proposed
strategy
to
control
PICs
using
HC
and
CO
limitations.
3
The
SAB
concluded
that
CO
was
not
adequately
correlated
with
PIC
emissions,
and
that
a
dual
control
strategy
of
both
CO
and
HC
monitoring
may
be
more
protective.
However,
the
SAB
pointed
to
the
need
to
control
other
combustion
parameters
including
temperature
and
combustion
air.
Dellinger's
work
also
demonstrates
the
need
to
control
temperature,
residence
time,
combustion
gas
velocity
and
combustion
air
to
minimize
PIC
formation.
Dellinger's
work
clearly
shows
that
sole
reliance
on
CO
is
not
adequate
to
control
PIC
emissions,
4
citing
work
conducted
by
other
EPA
scientists
at
EPA's
Risk
Reduction
Engineering
Laboratory
in
Cincinnati.
The
cited
study
concluded
that
"
no
correlation
between
CO
and
organic
emissions
could
be
discerned."
Therefore,
it
is
clear
that
relying
on
CO
or
HC
alone
will
not
ensure
"
comparable"
emissions
or
protection
of
health
and
environment
when
burning
CFE
waste
in
non­
RCRA
regulated
units.

RCRA
permitted
incinerators
and
industrial
furnaces
are
required
to
do
comprehensive
risk
burns
and
trial
burns
to
establish
operating
conditions
for
numerous
other
parameters
besides
CO.
These
include
minimum
combustion
temperature,
minimum
residence
time,
combustion
gas
velocity
indicators,
combustion
air
flow,
and
feed
rates
all
to
ensure
that
DRE
and
PIC
risk
burn
results
are
reported
on
an
ongoing
basis.
Site­
specific
PIC
Risk
Assessments
were
supported
by
the
1989
Science
Advisory
Board
paper
in
particular
as
an
effective
strategy
to
protect
against
PIC
emissions.
Yet
2
Dellinger,
B.,
et
al.
"
Pathways
of
Formation
of
Chlorinated
PICs
from
the
Thermal
Degradation
of
Simple
Chlorinated
Hydrocarbons."
Journal
of
Hazardous
Materials,
Elsevier
Publishers,
1989.
3
Report
of
the
Products
of
Incomplete
Combustion
Subcommittee,
Science
Advisory
Board,
USEPA,
October
10,
1989.
Docket
number
BBSP­
S0077A.
4
Dellinger,
B.,
et
al.
"
On
the
Relationship
Between
CO,
POHC
and
PIC
Emissions
from
a
Simulated
Hazardous
Waste
Incinerator."
JAPCA
39:
321­
327,
1989.
11
EPA's
CFE
waste
Option
B
would
place
blind
faith
in
CO
or
HC,
and
not
require
any
of
these
other
vital
controls
to
ensure
emissions
are
protective.
The
importance
of
residence
time,
temperature
and
oxygen
in
controlling
PIC
and
organic
emissions
is
supported
by
other
studies
by
Dellinger.
5
These
parameters
must
be
combined
with
limitations
on
CO
and
HC
to
control
organic
emissions.

ACC
therefore
needs
to
do
emissions
studies
and
provide
the
results
to
EPA
to
prove
their
underlying
premise
that
emissions
will
be
"
comparable."
On
the
surface,
this
cannot
be
concluded
merely
by
controlling
CO
or
HC.
Even
if
DRE
performance
is
good,
comparable
emissions
cannot
be
assumed.
For
example,
the
same
combustion
device
that
achieves
99.99%
DRE
for
both
a
fossil
fuel
with
0.5%
benzene
and
a
CFE
waste
with
90%
benzene
will
emit
180
times
higher
levels
of
benzene
when
fired
on
the
CFE
waste.
The
0.01%
fraction
emitted
in
the
fossil
fuel
case
is
0.00005
and
in
the
CFE
waste
case
is
0.0090.
Clearly
emissions
are
not
comparable.
Other
considerations
that
must
be
addressed
by
EPA
are
as
follows:

 
A
PM
limitation
is
also
needed
that
is
at
the
same
level
as
the
current
MACT
standards.
As
noted
in
a
memorandum
dated
February
22,
1996,
from
Larry
Gonzalez
of
EPA
in
the
RCRA
Docket
F­
96­
RCSP­
FFFFF,
organic
HAPs
are
present
on
particulate
matter.

 
EPA
and
the
states
should
have
authority
to
require
site­
specific
risk
burns
and
trial
burns
as
part
of
the
condition
for
units
burning
CFE
waste,
if
there
is
a
reasonable
basis
to
believe
there
may
be
significant
risks.
There
is
no
other
mechanism
to
assure
safe
emissions
or
even
comparable
emissions.
Public
notice
and
comment
must
also
be
allowed.

 
Regular
confirmatory
emissions
testing
should
be
performed
on
some
minimum
frequency
(
i.
e.,
every
5
years),
demonstrating
DRE,
PIC
emissions
and
D/
F
emissions.

 
The
HWC
MACT
D/
F
emission
limitation
should
be
in
place
for
units
burning
CFE
waste,
as
CO
and
HC
are
not
adequate
surrogates
for
D/
Fs.

 
Note
that
D/
F
formation
does
take
place
at
low
chlorine
levels
as
shown
in
numerous
studies.
The
current
CFE
halogen
limit
of
540
mg/
kg
is
sufficient
chlorine
to
allow
formation
of
dioxin
and
furan.

5
Dellinger,
B.,
et
al.,
"
PIC
Formation
Under
Pyrolytic
and
Starved
Air
Conditions."
EPA
/
600/
S2­
86/
006
July
1986,
Hazardous
Waste
Engineering
Research
Laboratory.
Dellinger,
B.,
"
Theoretical
DRE
Calculations
for
Incinerators
and
Thermal
Oxidizers,"
IT3'
02
Conference,
May
2002.
Dellinger,
B.,
et
al.,
"
Minimization
and
Control
of
Hazardous
Combustion
Byproducts,"
EPA/
600/
S2­
90/
039.
12
 
ETC
agrees
under
B4
that
firing
rate
must
be
restricted.
However
this
firing
rate
cannot
be
universally
defined
but
must
be
determined
in
a
comprehensive
risk
burn
and
trial
burn
that
establish
the
maximum
feed
rates
for
CFE
waste.

Minimum
Heating
Value.
The
8,000
Btu/
lb
level
is
too
low
and
does
not
conform
to
the
ACC
survey
data,
which
indicates
an
average
heat
content
of
13,750
Btu/
lb.
Only
11%
on
a
volume
basis,
or
7
out
of
96
waste
streams
surveyed
by
ACC,
had
heating
values
at
or
below
8,000.
Of
these
one
waste
had
no
problem
meeting
the
CFE,
and
the
rest
failed
for
metals
(
As,
Cr),
halogens,
acrolein,
reactive
cyanide
but
not
hydrocarbons
or
oxygenated
constituents.
Only
one
waste
had
an
issue
with
benzene
which
is
a
highly
toxic
carcinogen.
Therefore,
none
of
the
ACC
respondents
with
values
of
8,000
Btu/
lb
or
lower
should
have
been
used
to
establish
minimum
heating
value.
EPA
must
filter
the
survey
results
for
constituents
failing
current
exemption
that
were
not
oxygenates
or
hydrocarbons.

If
EPA's
intent
is
to
set
a
comparable
fuels
exemption,
then
heat
content
data
must
also
be
"
comparable"
to
fossil
fuels.
EPA
should
be
setting
a
minimum
heating
value
of
15,000
Btu/
lb
or
higher,
if
the
intent
is
really
to
promote
energy
recovery.
Note
also
that
if
the
ACC
survey
data
is
filtered
to
remove
the
7
streams
at
or
below
8,000
which
do
not
belong
there
(
because
these
failed
for
metals,
cyanide,
halogen
or
parameters
other
than
hydrocarbons
or
oxygenates),
then
the
average
heat
content
of
the
remaining
waste
is
closer
to
15,000
Btu/
lb.

In
justifying
any
claim
of
"
comparable
emissions,"
EPA
must
also
evaluate
other
pollutants
such
as
PM,
NOX,
SOX,
and
other
HAPs.
Given
that
EPA
is
considering
exempting
waste
containing
high
percentage
levels
of
hazardous
constituents,
the
impact
on
emissions
of
these
other
HAPs
must
also
be
evaluated
relative
to
fossil
fuels.
EPA
must
factor
these
emissions
from
other
HAPs
into
the
Risk/
Benefit,
Economic
and
other
impact
evaluations
or
studies
performed
in
support
of
the
proposal.
The
cost
of
increased
emissions
and
impacts
from
PM
and
other
HAPs
should
be
fairly
evaluated
as
part
of
this
rulemaking.

On­
Site
or
Limited
Off­
Site
Burning.
There
should
be
no
off­
site
burning
of
CFE
waste.
This
is
important
to
ensure
that
all
of
the
above
issues
are
controlled
and
enforced.
EPA
is
considering
the
option
of
allowing
hazardous
wastes
potentially
containing
high
levels
of
carcinogenic
constituents
(
benzene,
acrolein,
etc.)
to
be
shipped
by
the
generator
to
"
affiliated
off­
site
facilities,"
and
possibly
to
any
off­
site
burners,
under
the
exclusion
for
comparable
emissions
in
Option
B.
Setting
aside
the
question
of
whether
emissions
are
in
fact
comparable,
there
can
be
no
dispute
that
the
hazardous
wastes
fuels
being
shipped
to
the
burner
are
exactly
like
other
hazardous
waste
fuels,
and
not
like
comparable
fuels.

As
we
have
emphasized
above
in
these
comments,
under
Option
B
the
so­
called
comparable
emissions
fuel
can
have
90%
benzene
or
other
hazardous
compounds
and
still
qualify,
as
long
as
the
emissions
are
comparable.
But
it
is
apparent
that
this
fuel
is
as
risky
during
storage,
handling,
and
transport
as
any
other
hazardous
waste
fuel
that
is
13
currently
subject
to
RCRA
Subtitle
C
standards
for
good
reasons.
We
cannot
envision
any
good
reason
why
EPA
would
allow
hazardous
wastes
destined
for
a
qualified
burner
under
Option
B
to
be
shipped
without
manifests
to
ensure
tracking,
emergency
response,
and
secure
delivery.
We
cannot
imagine
why
EPA
would
allow
storage
in
tanks
and
containers
that
do
not
meet
standards
for
hazardous
waste
fuels,
such
as
compatibility,
integrity,
containment,
inspections,
etc.
Ironically,
off­
spec
waste
oil
would
be
subject
to
safer
controls
under
40
CFR
Part
279
than
would
these
comparable
fuels.

We
do
not
see
a
significant
distinction
between
off­
site
affiliated
facilities
and
any
other
off­
site
burners.
In
both
cases,
the
hazardous
waste
fuel
must
be
stored,
handled,
and
shipped
to
the
qualified
burner,
posing
the
same
risks
of
leaks,
spills,
explosions,
fugitive
emissions,
and
harm
to
the
public
that
justify
careful
RCRA
controls.
We
understand
the
argument
that
affiliated
burners
have
the
same
corporate
parent,
and
therefore
potential
liability
may
be
broader,
but
we
are
also
aware
of
how
affiliated
companies
are
often
separately
incorporated
to
limit
liability,
and
how
difficult
it
can
be
to
pierce
the
corporate
veil
absent
a
high
threshold
showing
of
fraud.
We
note
that
the
storage,
handling,
and
shipping
requirements
for
hazardous
wastes
in
general
do
not
make
any
special
allowances,
much
less
exemptions,
for
shipments
between
affiliated
corporations.
None
is
warranted
with
respect
to
hazardous
waste
fuels
that
may
meet
the
CFE
under
Option
B.

In
conclusion,
the
current
Option
B
provisions
are
far
from
adequate
to
allow
exclusion
of
CFE
waste.
EPA
must
gather
real
emissions
data
and
not
assume
that
CO
or
HC
will
serve
as
default
surrogates.
EPA
must
also
look
hard
at
the
fact
that
PIC
and
D/
F
emissions
are
impossible
to
predict
and
require
real
testing
in
trial
burns
and
risk
burns
on
a
site­
specific
basis.
EPA
must
also
expand
the
criteria
to
include
combustion
temperature,
residence
time,
combustion
gas
velocity,
minimum
oxygen
and
other
operating
parameters.
Certain
highly
toxic
and
persistent
compounds
such
as
benzene,
PNA
compounds,
acrolein,
naphthalene,
phenolic
compounds
and
toluene
must
be
excluded
from
the
scope
of
the
CFE.
Limitations
on
PM
correlated
to
ash
content
should
also
be
set.
Finally,
heating
value
truly
comparable
to
fuel
and
consistent
with
the
ACC
survey
results
on
the
order
of
15,000
Btu/
lb
should
be
set
as
minimums
for
CFE
waste.

Option
C:
Blending
to
Meet
the
Specifications
for
Hydrocarbons
and
Oxygenates.

Currently,
the
blending
of
wastes
to
meet
the
CFE
specifications
is
prohibited,
except
for
viscosity.
Option
C
would
allow
blending
of
wastes
that
exceed
the
specs
for
hydrocarbons
and/
or
oxygenates
with
comparable
fuel
or
fuel
oil
to
meet
the
specifications.
Importantly,
blending
to
meet
the
specifications
for
halogenated,
sulfonated,
or
nitrogenated
organics
would
still
be
prohibited.

How
will
EPA
enforce
blending­
only
for
hydrocarbons
and
oxygenates?
We
agree
that
blending
to
meet
specs
for
halogenated,
sulfonated,
or
nitrogenated
organic
constituents
would
constitute
improper
dilution.
The
specification
for
these
organics
is
nondetect
at
a
minimum
detection
limit,
and
it
should
be
impermissible
to
meet
the
specs
by
blending
a
detected
quantity
of
these
other
organics
to
a
level
below
the
minimum
14
detection
limit.
Dilution
to
meet
the
metal
specs
must
also
be
prohibited.
But
without
specific
requirements
for
the
waste
analysis
plan,
recordkeeping,
and
reporting
with
respect
to
blending,
it
will
be
impossible
for
EPA
and
states
to
detect
and
prevent
abuses.

The
waste
analysis
plan
(
WAP)
requirements
for
comp
fuel
generators
in
40
CFR
261.38(
c)(
7)
are
very
general
and
drafted
primarily
for
the
CFE
determination.
The
generator
needs
only
a
written
WAP
that
specifies
"
the
parameters
for
which
each
hazardous
waste
will
be
analyzed,"
the
test
methods,
sampling
method,
and
frequency
of
analyses.
Generators
need
only
retain
the
WAP
at
the
facility;
there
is
no
review
or
approval
by
EPA
or
the
state.

Generators
who
blend
must
be
required
by
the
CFE
regulations
to
sample
and
analyze
their
hazardous
waste
streams
before
and
after
blending
to
show
that
impermissible
dilution
does
not
take
place.
Generators
who
blend
should
at
least
be
required
to
submit
their
WAPs
to
the
regulatory
authority
for
review
and
approval.
Most
importantly,
generators
should
be
required
to
obtain
a
standardized
permit
for
the
tanks
and
equipment
used
for
hazardous
waste
blending.
EPA's
option
does
not
expressly
address
the
status
of
such
tanks,
but
since
the
materials
being
blended
are
hazardous
wastes
prior
to
meeting
the
CFE,
the
tanks
should
meet
the
minimum
tank
standards
for
structural
integrity,
secondary
containment,
etc.,
set
forth
in
a
standardized
permit.

Pending
Legal
Challenges
to
the
Comparable
Fuel
Exclusion
in
40
CFR
261.38
As
EPA
is
aware,
the
lawfulness
of
the
comparable
fuels
exclusion
is
the
subject
of
pending
litigation.
American
Chemistry
Council
v.
EPA,
No.
98­
1375
and
consolidated
cases
(
D.
C.
Circuit).
While
this
case
is
in
abeyance
at
this
time,
we
anticipate
that
legal
challenges
to
the
proposed
revisions
will
be
filed
and
consolidated
with
the
pending
action,
and
the
case
will
then
be
active
again,
rather
than
dismissed.
Some
of
the
issues
that
may
be
raised
with
respect
to
the
1998
CFE
are
as
follows:

 
The
flawed
nature
of
certain
of
the
metals
standards
that
are
incorrectly
based
on
an
assumed
composite
value.

 
The
small
number
of
samples
that
were
evaluated
in
establishing
the
exemption
levels
from
a
limited
number
of
geographic
locations.

 
The
allowance
for
5,000
Btu/
lb
as
a
minimum
heat
content.
A
5,000
Btu/
lb
waste
is
not
comparable
to
a
15,000
Btu/
lb
fuel.
The
recent
ACC
survey
would
indicate
strong
evidence
to
raise
this
threshold
to
15,000
Btu/
lb.

 
The
normalization
to
10,000
Btu/
lb
intended
to
adjust
for
fuel
substitution
raises
several
issues:

o
This
is
an
admission
that
5,000
Btu/
lb
threshold
is
not
truly
based
on
fuel.
How
can
5,000
Btu/
lb
be
considered
a
minimum
energy
recovery
threshold
when
normalization
is
being
based
on
10,000
as
an
energy
15
recovery
threshold?
EPA
has
arbitrarily
picked
a
number
when
it
comes
to
defining
thresholds
for
energy
recovery
with
no
real
science
or
consideration
of
real
fossil
fuel
values.

o
The
normalization
should
be
based
on
19,500
not
10,000
as
the
waste
is
being
substituted
for
fossil
fuel,
not
50%
fossil
fuel.
Therefore,
EPA
has
not
properly
set
the
standard
so
as
to
adjust
for
the
increased
environmental
loading
from
the
burning
of
CFE
waste.
The
goal
covered
on
page
1­
4
of
the
Comparable
Fuel
Background
Document
of
not
increasing
the
net
environmental
loading
of
hazardous
constituent
emissions
from
the
comparable
fuel
in
comparison
with
the
fossil
fuel
has
not
been
satisfied
with
the
10,000
Btu/
lb
normalization
basis.

 
Allowance
of
knowledge
as
opposed
to
requiring
sampling
and
analysis.
Generators
cannot
possibly
know
that
their
hazardous
waste
meets
the
comp
fuel
specifications
without
adequate
analysis,
and
any
use
of
"
knowledge"
must
be
grounded
on
such
testing
and
analysis.

 
The
unspecified
frequency
of
confirmatory
testing.

 
Verification
testing
needs
to
be
overseen
with
requirements
analogous
to
the
delisting
program
in
which
results
are
submitted
to
regulatory
agencies
for
review
and
verification.

 
The
lack
of
a
specification
for
total
ash
content
to
control
PM.
This
is
important
as
many
HAPs
are
absorbed
onto
or
integral
to
particulate
matter.

 
The
failure
to
consider
that
certain
highly
toxic,
carcinogenic
or
persistent
hazardous
constituents
must
not
be
present
at
any
concentration
in
any
exempted
waste.
The
current
rule
allows
even
dioxin,
PCBs,
mercury
and
other
highly
toxic
constituent
to
be
present
in
CFE
waste,
even
though
such
highly
toxic
constituents
are
not
found
in
fossil
fuels.

 
The
presumption
that
hazardous
constituents
that
are
not
associated
with
fossil
fuels
can
be
presumed
to
be
present
up
to
their
statistically
adjusted
detection
limits
in
fuel
oil
matrices.
For
example,
dioxin
and
PCBs
are
not
associated
with
fossil
fuels
yet
EPA
allows
a
comparable
fuel
waste
to
be
burned
in
place
of
fossil
fuel
based
solely
on
a
flawed
adjusted
detection
limit
construct.
40
CFR
261.38
must
totally
ban
certain
toxic
constituents
from
waste
at
any
level,
if
these
toxic
constituents
are
not
associated
with
fossil
fuels.
Otherwise,
EPA
is
allowing
toxics
along
for
the
ride
and
is
increasing
environmental
impact
substantially
over
the
burning
of
fossil
fuels.

 
The
1998
rule
lacked
any
rigorous
environmental
impact
or
risk
assessment
evaluations
regarding
the
increased
emissions
that
result
from
burning
exempted
waste
over
fossil
fuels.
The
cost/
benefit
economic
evaluation
failed
16
to
consider
negative
environmental
impacts
resulting
from
the
burning
of
exempted
waste
in
place
of
fossil
fuels.
This
analysis
must
also
include
emission
studies
for
CFE
waste
that
has
been
excluded
since
1998
for
all
HAPs
to
evaluate
what
increased
environmental
burden
has
resulted.

 
The
1998
rule
allows
for
treatment
to
meet
the
comparable
fuel
specifications
without
adequate
controls
for
such
treatment,
particularly
controls
to
guard
against
dilution.

 
The
1998
rule
fails
to
address
at
all
the
issue
of
residues
from
the
burning
of
CFE
waste
in
unregulated
units.
This
is
a
concern
particularly
because
there
is
no
ash
limitation,
and
the
amount
of
residues
in
exempted
waste
can
be
substantial.
The
rule
fails
to
specify
that
residues
from
burning
CFE
waste
remain
hazardous
waste
subject
to
proper
RCRA
disposal.

Conclusion
Once
again,
the
ETC
appreciates
the
opportunity
to
provide
stakeholder
input
on
the
various
options
that
EPA
is
considering
for
changes
to
the
CFE
regulations.
However,
we
must
insist
that
EPA
first
obtain
additional
information
from
the
proponents
of
these
changes
to
address
the
many
questions
and
uncertainties
that
arise
from
these
options,
even
at
this
early
stage
of
consideration.
In
particular,
we
believe
EPA
should
ascertain
with
some
certainty
how
much
hazardous
waste
will
in
fact
become
eligible
for
the
CFE
under
these
options,
and
whether
any
substantial
energy
recovery
and/
or
legitimate
cost
reductions
will
be
realized.
We
fear
that
hazardous
wastes
will
simply
be
diverted
from
regulated
units
to
unregulated
burning,
with
little
or
no
additional
energy
recovery.
Most
disturbing,
we
fear
that
any
cost
reductions
will
result
only
from
eliminating
safeguards
and
protections
for
the
storage,
transport,
and
burning
of
hazardous
wastes
that
are
vitally
necessary
to
protect
public
health
and
the
environment.

Respectfully
submitted
David
R.
Case
Executive
Director
cc:
Mr.
Bob
Holloway
Steven
Silverman,
Esq.
