1
Cost
Analysis
of
Mercury
Monitoring
Techniques.
Draft
report
prepared
by
Arcadis
for
Dr.
Ruben
Deza,
U.
S.
Environmental
Protection
Agency,
Washington,
DC.,
November
3,
2003.
COST
ESTIMATES
FOR
MERCURY
EMISSIONS
MONITORING
The
cost
information
gathered
in
this
document
is
largely
based
on
documents
generated
by
Arcadis
Inc.,
on
November
2003,
for
the
Clean
Air
Markets
Division,
specifically
the
"
Cost
Analysis
of
Mercury
Monitoring
Techniques"
1,
which
is
also
in
the
CAMR
docket;
a
cost
analysis
performed
by
MACTEC
Inc.;
and
additional
information
obtained
by
the
Agency
in
the
development
of
the
rule.

This
paper
provides
an
overview
of
the
cost
information
and
provides
an
estimate
of
the
monitoring
costs,
focusing
on
the
utility
boilers
that
would
be
required
to
monitor
mercury
emissions
under
the
CAMR.

In
developing
these
calculations,
EPA
considered
comments
from
utility
sources
and
monitoring
system
vendors.
The
paper
assumes
that
utility
sources
will
take
advantage
of
those
monitoring
components
already
installed
as
part
of
complying
with
other
rules
under
the
Clean
Air
Act,
such
as
CEMS
shelters,
platforms,
flow
meters,
etc.
Therefore,
such
equipment
is
not
not
included
in
this
paper.
In
addition,
it
is
estimated
that
no
major
structural
modifications
are
necessary
for
access
or
support
equipment,
and
that
sufficient
on­
site
power
is
available
to
operate
additional
mercury
monitoring
equipment.

The
paper
addresses
monitoring
costs
on
the
basis
of
the
main
options
reflected
by
the
proposed
regulation
and
that
are
likely
to
be
relied
on
by
sources
for
purposes
of
complying
with
the
cap
and
trade
rule.

Methodology
Several
monitoring
options
and
cut­
off
levels
were
chosen
in
an
attempt
to
structure
the
cost
analysis.
The
analysis
relies
on
the
data
available
from
the
1999
Mercury
Information
collection
Request
(
ICR).
This
ICR
provided
mercury
emission
information
on
a
total
of
1120
utility
sources,
and
based
on
this
information
a
cumulative
mercury
distribution
curve
was
developed.

Estimates
were
then
developed
for
units
using
the
Low
Mass
Emissions
(
LME)
approach
representing
one
percent
of
the
cumulative
mercury
emissions
in
the
distribution
curve.
The
LME
provision
represents
a
more
flexible
monitoring
option
were
qualifying
sources
would
not
be
required
to
use
CEMS
or
sorbent
traps
but
instead
would
be
required
to
once­
a­
year
stack
testing
and
continuous
flow
measurements.
From
the
emissions
distribution
curve,
228
low
emitting
sources
would
have
the
LME
option.

A
second
group
of
sources,
corresponding
to
those
emitting
between
1%
and
5%
of
the
cumulative
mercury
emissions,
would
be
allowed
to
use
a
variation
of
the
LME
option,
which
differs
from
the
first
one
in
that
it
requires
semi­
annual
stack
testing
instead
of
annually.
A
total
of
207
units
would
be
able
to
use
this
option.

Based
on
available
cost
information,
including
that
in
the
ARCADIS
study,
the
Agency
estimates
that
the
stack
testing
provision
costs
will
average
about
$
5,500
each.
An
additional
$
1,500
annual
are
estimated
for
technical
calculations,
labor
and
other
associated
expenses.
This
would
bring
the
annual
expenditure
per
unit
to
about
$
7,000
for
LME
units
in
the
first
group,
and
about
$
12,500
for
the
ones
in
the
second
group.

For
CEMS,
there
are
two
major
components
to
be
considered,
capital
costs
and
operational
costs.
Based
on
a
survey
of
11
manufacturers,
Arcadis
estimated
an
average
capital
cost
of
$
60,000.
In
addition,
the
analysis
assumes
that
each
source
would
purchase
a
Hovocal
or
Mercal
unit
to
generate
ionic
mercury,
which
is
necessary
for
CEMS
calibration.
The
approximate
cost
of
either
instrument
is
$
10,000.
Based
on
this,
the
capital
cost
is
estimated
at
about
$
70,000.

The
operational
costs
can
also
be
broken
into
two
components,
i.
e.,
routine
maintenance
and
operation
and
quality
assurance/
control
procedures.
Based
on
discussions
with
industry,
these
costs
are
estimated
at
$
87,000.

For
units
using
a
sorbent
trap,
in
compliance
with
the
final
draft
rule,
the
system
would
include
a
carbon
trap
with
a
three­
section
cartridge:
a
primary
capture
cartridge,
backup
for
breakthough,
and
a
portion
for
spiking
for
QA/
QC,
and
two
carbon
traps
would
be
run
in
parallel
for
one
week
runs.
The
costs
for
a
trap
system
is
estimated
at
$
8,500.
Capital
costs
including
the
installation
of
the
complete
system
is
estimated
at
about
$
20,000.
The
operational
costs
can
be
broken
down
in
three
components:
routine
operational
costs,
laboratory
analysis
costs,
and
QA/
QC
procedures,
and
are
estimated
at
about
$
113,000.

The
Agency
estimates
that
about
50
percent
of
the
sources
not
using
LME
procedures
will
use
each
of
the
other
two
available
monitoring
technologies.
Based
on
the
ICR
data,
the
number
of
units
able
to
use
one
monitoring
option
or
the
other
are
:

°
1%
of
cumulative
mercury
distribution
curve:
228
units
use
LME
1
approach
°
5%
of
cumulative
mercury
distribution
curve:
207
units
use
LME
2
approach
°
The
rest
of
the
units
use
evenly
CEMS
and
sorbent
traps
methodologies:
°
342
use
CEMS
and
343
use
sorbent
traps
Based
on
the
number
of
units
with
options
to
use
the
various
monitoring
approaches,
a
number
of
scenarios
is
presented
in
the
following
table.
Table
1
Option
Capital
Costs
($
1000)
Annual
Operating
Costs
($
1000)
Annualized
capital
costs
($
1000)
Annualized
costs
($
1000)
Number
of
Units
Total
Annualized
Costs
($
1000)

1%
0
7.0
7.0
7.0
228
1,596
5%
0
12.5
12.5
12.5
207
2,588
CEMS
70
87.0
8.4
95.4
342
32,627
Sorbent
Traps
20
113,0
2.4
115.4
343
39,582
Totals
76,393
Then,
the
estimated
annual
impact
of
the
measure
is
$
76,392,500.
