1
CONTROL
OF
MERCURY
EMISSIONS
FROM
COAL­
FIRED
ELECTRIC
UTILITY
BOILERS
Air
Pollution
Prevention
and
Control
Division
National
Risk
Management
Research
Laboratory
Office
of
Research
and
Development
U.
S.
Environmental
Protection
Agency
Research
Triangle
Park,
NC
Introduction
During
combustion,
the
mercury
(
Hg)
in
coal
is
volatilized
and
converted
to
elemental
mercury
(
Hg0)
vapor
in
the
high
temperature
regions
of
coal­
fired
boilers.
As
the
flue
gas
is
cooled,
a
series
of
complex
reactions
begin
to
convert
Hg0
to
ionic
mercury
(
Hg2+)
compounds
and/
or
Hg
compounds
(
Hgp)
that
are
in
a
solid­
phase
at
flue
gas
cleaning
temperatures
or
Hg
that
is
adsorbed
onto
the
surface
of
other
particles.
The
presence
of
chlorine
gas­
phase
equilibrium
favors
the
formation
of
mercuric
chloride
(
HgCl2)
at
flue
gas
cleaning
temperatures.
However,
Hg0
oxidation
reactions
are
kinetically
limited
and,
as
a
result,
Hg
enters
the
flue
gas
cleaning
device(
s)
as
a
mixture
of
Hg0,
Hg
2+,
and
Hgp.
This
partitioning
of
Hg
into
Hg0,
Hg
2+,
and
Hgp
is
known
as
mercury
speciation,
which
can
have
considerable
influence
on
selection
of
mercury
control
approaches.
In
general,
the
majority
of
gaseous
mercury
in
bituminous
coal­
fired
boilers
is
Hg2+.
On
the
other
hand,
the
majority
of
gaseous
mercury
in
subbituminous­
and
lignite­
fired
boilers
is
Hg0.

Control
of
mercury
emissions
from
coal­
fired
boilers
is
currently
achieved
via
existing
controls
used
to
remove
particulate
matter
(
PM),
sulfur
dioxide
(
SO2),
and
nitrogen
oxides
(
NOx).
This
includes
capture
of
Hgp
in
PM
control
equipment
and
soluble
Hg
2+
compounds
in
wet
flue
gas
desulfurization
(
FGD)
systems.
Available
data
also
reflect
that
use
of
selective
catalytic
reduction
(
SCR)
NOx
control
enhances
oxidation
of
Hg0
in
flue
gas
and
results
in
increased
mercury
removal
in
wet
FGD.

Table
1
shows
the
average
reduction
in
total
mercury
(
HgT)
emissions
developed
from
EPA's
Information
Collection
Request
(
ICR)
data
on
U.
S.
coal­
fired
boilers.
Plants
that
employ
only
PM
controls
experienced
average
HgT
emission
reductions
ranging
from
0
to
90
percent.
Units
with
fabric
filters
(
FFs)
obtained
the
highest
average
levels
of
control.
Decreasing
average
levels
of
control
were
generally
observed
for
units
equipped
with
a
cold­
side
electrostatic
precipitator
(
CSESP
hot­
side
ESP
(
HS­
ESP),
and
particle
scrubber
(
PS).
For
units
equipped
with
dry
scrubbers,
the
average
HgT
emission
reductions
ranged
from
0
to
98
percent.
The
estimated
average
reductions
for
wet
flue
gas
desulfurization
(
FGD)
scrubbers
were
similar
and
ranged
from
0
to
98
percent.

As
seen
in
Table
1,
in
general,
the
amount
of
Hg
captured
by
a
given
control
technology
is
greater
for
bituminous
coal
than
for
either
subbituminous
coal
or
lignite.
For
example,
the
average
capture
of
Hg
in
plants
equipped
with
a
CS­
ESP
is
36
percent
for
bituminous
coal,
3
percent
for
subbituminous
coal,
and
0
percent
for
lignite.
Based
on
ICR
data,
it
is
estimated
that
existing
controls
remove
about
36%
of
the
75
tons
of
mercury
input
with
coal
in
U.
S.
coal­
fired
boilers.
This
results
in
current
emissions
of
48
tons
of
mercury.
2
There
are
two
broad
approaches
to
mercury
control:
(
1)
activated
carbon
injection
(
ACI),
and
(
2)
multipollutant
control,
in
which
Hg
capture
is
enhanced
in
existing/
new
SO2,
NOx,
and
PM
control
devices.
Relative
to
these
two
approaches,
this
paper
describes
currently
available
data,
limitations,
estimated
potential,
and
Research
Development
and
Demonstration
(
RD&
D)
needs.
Depending
on
levels
appropriated
by
congress,
EPA
may
not
be
able
to
continue
it's
review
of
mercury
removal
technologies
in
fiscal
year
2004.

Table
1.
Average
mercury
capture
by
existing
post­
combustion
control
configurations
used
for
PC­
fired
boilers
Average
Mercury
Capture
by
Control
Configuration
Coal
Burned
in
Pulverized­
coal­
fired
Boiler
Unit
Post­
combustion
Control
Strategy
Post­
combustion
Emission
Control
Device
Configuration
Bituminous
Coal
Subbituminous
Coal
Lignite
CS­
ESP
36
%
3%
0
%
HS­
ESP
9
%
6
%
not
tested
FF
90
%
72
%
not
tested
PM
Control
Only
PS
not
tested
9
%
not
tested
SDA+
CS­
ESP
not
tested
35
%
not
tested
SDA+
FF
98
%
24
%
0
%
PM
Control
and
Spray
Dryer
Adsorber
SDA+
FF+
SCR
98
%
not
tested
not
tested
PS+
FGD
12
%
0
%
33%
CS­
ESP+
FGD
75
%
29
%
44
%
HS­
ESP+
FGD
49
%
29
%
not
tested
PM
Control
and
Wet
FGD
System(
a)
FF+
FGD
98
%
not
tested
not
tested
CS­
ESP
=
cold­
side
electrostatic
precipitator
(
a)
Estimated
capture
across
both
control
devices
HS­
ESP
=
hot­
side
electrostatic
precipitator
FF
=
fabric
filter
PS
=
particle
scrubber
SDA
=
spray
dryer
absorber
system
State­
of­
the­
art
of
Controlling
Mercury
Emissions
by
Activated
Carbon
Injection
ACI
has
the
potential
to
achieve
moderate
to
high
levels
of
Hg
control.
The
performance
of
an
activated
carbon
is
related
to
its
physical
and
chemical
characteristics.
Generally,
the
physical
properties
of
interest
are
surface
area,
pore
size
distribution,
and
particle
size
distribution.
The
capacity
for
Hg
capture
generally
increases
with
increasing
surface
area
and
pore
volume.
The
ability
of
Hg
and
other
sorbates
to
penetrate
into
the
interior
of
a
particle
is
related
to
pore
size
distribution.
The
pores
of
the
carbon
sorbent
must
be
large
enough
to
provide
free
access
to
internal
surface
area
by
Hg0
and
Hg2+
while
avoiding
excessive
blockage
by
previously
adsorbed
reactants.
As
particle
sizes
decrease,
access
to
the
internal
surface
area
of
particle
increases
along
with
potential
adsorption
rates.
3
Carbon
sorbent
capacity
is
dependent
on
temperature,
the
concentration
of
Hg
in
the
flue
gas,
the
flue
gas
composition,
and
other
factors.
In
general,
the
capacity
for
adsorbing
Hg2+
will
be
different
than
that
for
Hg0.
The
selection
of
a
carbon
for
a
given
application
would
take
into
consideration
the
total
concentration
of
Hg,
the
relative
amounts
of
Hg
°
and
Hg2+,
the
flue
gas
composition,
and
the
method
of
capture
[
electrostatic
precipitator
(
ESP),
FF,
or
dry
FGD
scrubber].

ACI
may
be
used
either
in
conjunction
with
existing
control
technologies
and/
or
with
additional
control
such
as
the
addition
of
an
FF.
To
date
ACI
has
only
been
evaluated
during
short­
term
tests
on
commercially
operating
electrical
generating
plants.
Longer­
term
tests
of
ACI
have
been
limited
to
continuous
operation,
24
hr/
day­
7days/
week,
for
a
period
of
less
than
two
weeks
at
four
field
test
sites.
Also,
combustion
modification,
such
as
coal
reburning
technology,
may
increase
the
carbon
in
fly
ash
and
yield
enhanced
Hg
capture
in
PM
control
devices
The
Department
of
Energy/
National
Energy
Technology
Laboratory
(
DOE/
NETL),
the
Electric
Power
Research
Institute
(
EPRI)
and
a
group
of
utility
companies
have
funded
projects
to
evaluate
the
use
of
ACI
as
summarized
in
Table
2.
The
Hg
removal
via
ACI
is
measured
between
the
inlet
and
outlet
of
the
particulate
matter
control
device.
Note
that
these
projects
represent
ACI
applications
that
can
be
used
to
control
Hg
emissions
from
units
that
(
1)
are
currently
equipped
with
an
ESP,
and
(
2)
burning
bituminous
or
subbituminous
coals.
The
tests
at
Alabama's
Gaston
Plant
show
the
potential
Hg
control
levels
that
can
be
achieved
by
installing
a
compact
hybrid
particulate
collector
(
COHPAC)
or
small
pulse­
jet
FF
downstream
of
an
existing
ESP
and
injecting
activated
carbon
upstream
of
the
COHPAC
unit.

Table
2.
ACI
field
test
projects.

Test
Site
Information
Mercury
Capture,
%
Test
Site
Coal
Particulate
Control
Baseline
ACI
Test
Results
Long­
term
Test
Duration
PG&
E
NEG
Brayton
Point,
Unit
1
Low­
sulfur
Bituminous
Two
CS­
ESPs
in
Series
90.8
94.5
ACI
for
two
5­
day
periods
PG&
E
NEG
Salem
Harbor,
Unit
1
Low­
sulfur
Bituminous
CS­
ESP
90
94
ACI
for
one
4­
day
period
Wisconsin
Electric
Pleasant
Prairie,
Unit
2
Subbituminous
CS­
ESP
5
65
ACI
for
one
5­
day
period
Alabama
Power
Gaston,
Unit
3
Low­
sulfur
Bituminous
HS­
ESP
+
COHPAC
0
25­
90
ACI
for
one
9­
day
period
University
of
Illinois
Abbott
Station
High­
sulfur
Bituminous
CS­
ESP
0
73
4
A
mobile
sorbent
injection
system
and
a
mobile
test
laboratory
were
constructed
for
use
at
all
test
sites
except
Abbott.
Norit
lignite­
based
carbon,
Darco­
FGD,
was
used
as
the
benchmark
sorbent
at
all
test
sites.
Tests
at
the
sites
generally
included:

 
the
use
of
Apogee
Scientific
semi­
continuous
emission
monitors
(
S­
CEMs)
for
measurement
of
Hg0
and
total
vapor­
phase
Hg
(
Hgv);
 
periodic
measurements
of
Hgp,
Hg2+
and
Hg0
with
the
Ontario­
hydro
(
OH)
method;
 
laboratory
and
slipstream
sorbent
screening
tests;
 
baseline
tests
without
the
use
of
sorbents;
 
parametric
tests
to
evaluate
the
effects
of
process
conditions
and
sorbent
variables;
and
 
4­
to
9­
day
tests
with
Darco­
FGD.

The
purpose
of
tests
at
each
site
was
to
determine
the
performance
and
costs
of
activated
carbon
sorbents
for
controlling
Hg
emission
from
coal­
fired
electrical
generating
plants
equipped
only
with
an
ESP.
The
field
tests
are
summarized
below.

Brayton
Point
ACI
testing
was
conducted
on
the
245­
MW
Unit
1,
which
fired
a
low­
sulfur
bituminous
coal
with
0.03
ppm
Hg
and
2000­
4000
ppm
chlorine.
The
unit
is
equipped
with
low­
NOx
burners
and
typically
has
high
levels
of
unburned
carbon
(
UBC)
in
the
fly
ash
as
indicated
by
loss
on
ignition
(
LOI)
measurements.
The
PM
control
system
at
the
unit
is
unusual
in
that
it
consists
of
two
CSESPs
in
series
and
long
duct
runs.
Carbon
was
injected
between
the
ESPs.

The
average
baseline
removal
efficiency
across
both
ESPs
averaged
90.8
percent,
as
measured
during
three
tests
with
the
OH
method.
During
parametric
tests,
a
variety
of
activated
carbons,
including
Darco­
FGD,
were
injected
just
downstream
of
the
first
ESP.
Incremental
Hg
removal
efficiencies
across
the
second
ESP
ranged
from
3
to
93
percent
depending
on
the
carbon
injection
concentration.
Total
average
Hg
removal
efficiencies
across
both
ESPs
as
determined
by
the
S­
CEMs
averaged
94.5
percent
during
injection
of
Darco­
FGD
at
10
lb/
MMacf.

Longer­
term
performance
tests
involved
the
continuous
injection
of
Darco­
FGD
24
hours/
day
for
10
days
at
two
different
injection
concentrations.
Five
days
of
injection
at
10
µ
g/
dncm
was
followed
by
five
days
of
injection
at
20
µ
g/
dncm.
The
average
removal
efficiency
across
both
ESPs
during
ACI
concentrations
of
10
lb/
MMacf
was
94.5
percent
as
measured
during
3
OH
method
tests.
These
high
Hg
capture
efficiencies
are
considered
to
be
atypical
of
other
CS­
ESP
units
because
of
the
high
UBC
concentrations,
the
two
ESPs,
and
the
long
duct
runs.

Salem
Harbor
Tests
were
conducted
on
Unit
1,
an
88
MW
single
wall­
fired
unit
which
is
equipped
with
low­
NOx
burners,
a
selective
noncatalytic
reduction
(
SNCR)
system
for
NOx
control
and
a
CS­
ESP.
Salem
Harbor
fires
a
South
American
low­
sulfur
bituminous
coal
with
0.03­
0.08
ppm
Hg
and
206
ppm
chlorine.
The
resulting
fly
ash
had
an
LOI
of
20
to
30
percent.

Parametric
tests
at
reduced
loads
that
lowered
fly
ash
LOI
to
15
to
20
percent
did
not
significantly
reduce
Hg
capture.
Increasing
the
ESP
inlet
temperature
from
300
°
F
to
350
°
F
reduced
Hg
removal
from
approximately
90
percent
to
the
10­
20
percent
range.
The
effects
of
5
changes
in
LOI
over
test
range
of
15
to
30
percent
were
not
as
strong
as
the
effects
of
temperature
changes.

During
November
2002,
four
days
of
long­
term
sorbent
injection
tests
were
conducted
with
Darco­
FGD
at
an
injection
concentration
of
10
lb/
MMacf.
The
average
Hg
capture
efficiency
during
3
OH
tests
was
94.0
percent.
The
Hg0
concentrations
for
all
inlet
and
outlet
samples
were
below
the
method
detection
limit.
More
than
95
percent
of
the
total
inlet
Hg
was
measured
as
Hgp,
indicating
nearly
complete
in­
flight
capture
of
Hg
upstream
of
the
ESP.
The
very
high
inflight
Hg
capture
by
the
UBC
in
fly
ash
and
injected
activated
carbon
are
not
believed
to
be
representative
of
plants
equipped
with
a
CS­
ESP.

Pleasant
Prairie
ACI
testing
was
conducted
on
the
600­
MW
Unit
2,
which
fired
a
PRB
coal
with
0.11
ppm
Hg
and
8
ppm
chlorine.
The
unit
is
equipped
with
an
ESP.
Testing
was
conducted
on
one
ESP
chamber
(
1/
4
of
the
unit).
The
plant
sells
its
fly
ash
for
use
in
concrete.

Baseline
tests
using
the
OH
method
exhibited
Hg
capture
in
the
ESP
of
about
5
percent
with
more
than
70
percent
of
the
Hg
at
the
ESP
inlet
being
Hg0.
Major
parametric
test
variables
included
sorbent
properties
and
sorbent
injection
concentration.
At
low
ACI
concentrations,
Hg
reductions
across
the
ESP
were
higher
than
expected,
reaching
60
to
65
percent
at
injection
concentrations
near
10
lb/
MMacf.
Increasing
sorbent
injection
concentrations
to
20
to
30
lb/
MMacf
increased
Hg
reduction
efficiencies
to
only
about
70
percent.
Subsequently,
in
longterm
tests
carbon
was
injected
continuously
at
24
h/
day
for
5
days.
OH
measurements
confirmed
that
about
60­
70%
mercury
removal
could
be
achieved
at
a
carbon
injection
concentration
of
10
lb/
MMacf.

Gaston
ACI
testing
was
conducted
on
the
270­
MW
Unit
3,
which
fired
low­
sulfur
eastern
bituminous
coals
with
0.14
ppm
Hg
and
160
ppm
chlorine.
The
unit
is
equipped
with
low­
NOx
burners,
a
HS­
ESP
and
a
COHPAC,
which
was
retrofit
earlier
to
capture
residual
fly
ash
escaping
the
ESP.
Testing
was
conducted
on
one­
half
of
the
flue
gas
stream.

Baseline
test
results
showed
that
neither
the
HS­
ESP
nor
COHPAC
captured
a
significant
amount
of
Hg.
During
ACI
parametric
tests,
Hg
capture
efficiencies
ranged
from
25
to
more
than
90
percent,
depending
on
the
carbon
injection
rate.
ACI
concentrations
of
3
lb/
MMacf
resulted
in
gas­
phase
Hg
reductions
greater
than
90
percent
across
the
COHPAC.
However,
it
was
determined
that
ACI
resulted
in
a
significant
increase
in
COHPAC
cleaning
frequency.
The
different
activated
carbons
used
in
the
parametric
tests
produced
Hg
capture
efficiencies
similar
to
Darco­
FGD,
the
benchmark
sorbent.
Differences
in
sorbent
particle
size
or
base
material
(
bituminous
coal
or
lignite)
did
not
result
in
appreciable
performance
differences.
Subsequently,
in
long­
term
tests,
carbon
was
injected
continuously
at
24
h/
day
for
9
days.
The
COHPAC
cleaning
frequency
and
ACI
rate
was
kept
at
a
reduced
level
to
avoid
adverse
impacts
on
COHPAC
bag
life.
Relatively
short
duration
OH
measurements
reflected
about
90%
removal
of
mercury,
but
measurements
taken
with
S­
CEMS
reflected
about
78%
removal
over
the
period
of
the
long­
term
testing.
6
Abbott
In
the
summer
of
2001,
EPRI
sponsored
ACI
tests
at
the
Abbott
Power
Plant
located
in
Champaign,
Illinois.
Unit
5,
the
test
unit,
is
a
stoker­
fired
unit
followed
by
air
heater
and
a
CSESP
During
the
tests,
Unit
5
burned
an
Illinois
Basin
coal
with
nominal
sulfur
and
chlorine
contents
of
3.8
and
0.25%,
respectively.
Activated
carbons
used
during
the
parametric
tests
included
Darco
FGD,
fine
FGD
(
size
segregated
Darco
FGD),
and
an
experimental
Corn
Char
sorbent.

During
the
parametric
tests
ACI
concentrations
were
varied
from
5.1
to
20.5
lb/
MMacf.
The
ESP
inlet
temperatures
ranged
from
340
to
390
°
F.
The
performance
of
Darco
FGD
and
the
corn
char
sorbents
were
similar,
showing
increases
in
Hg
capture
proportional
to
the
ACI
concentration.
The
fine
FGD
sorbent
exhibited
improved
performance
relative
to
the
standard
FGD.
The
best
performance,
73%
Hg
capture,
was
achieved
by
injection
of
fine
FGD
at
13.8
lb/
MMacf
at
an
ESP
inlet
temperature
of
341
°
F.
The
high
sulfur
flue
gas
appeared
to
impair
the
performance
of
the
activated
carbon.
This
is
consistent
with
bench­
scale
research
that
shows
that
high
SO2
concentrations
diminished
the
adsorption
capacity
of
activated
carbons.

Recently,
EPA
has
estimated
cost
for
ACI­
based
controls.
1
These
estimates
range
from
0.03­
3.096
mills/
kWh.
However,
the
higher
costs
are
usually
associated
with
the
plant
configuration
utilizing
SDA+
CS­
ESP
or
HS­
ESPs.
Excluding
the
costs
associated
with
the
plant
configurations
involving
SDA
+
ESP
or
HS­
ESP,
cost
estimates
are
from
0.03
to
1.903
mills/
kWh.
At
the
low
end
of
this
cost
range,
0.03
mills/
kWh,
it
is
assumed
that
no
additional
control
technologies
are
needed,
but
mercury
monitoring
will
be
necessary.

RD&
D
Needs
for
Sorbent
Injection
Systems
In
order
to
enhance
the
cost
effective
capture
of
Hg
by
ACI,
and
other
sorbent
injection
systems,
for
the
important
coal
type/
retrofit
control
combinations,
the
following
RD&
D
efforts
are
needed.

°
Research
efforts
on
Hg
speciation
and
capture
should
be
continued.
These
efforts
will
include
bench­
and
pilot­
scale
investigations
on
the
effects
of
flue
gas
composition,
fly
ash
properties
(
UBC
content
and
catalytic
metal
content),
flue
gas
quench
rates,
and
other
important
parameters.
Speciation
and
capture
computer
models
must
be
developed
to
evaluate
field
test
results
and
for
application
to
other
utility
sites.

°
Development
and
demonstration
of
low
cost
sorbents,
impregnated
sorbents
and
innovative
sorbents
that
are
effective
in
controlling
Hg
emissions
from
subbituminous
coal
and
lignite
should
continue.
High
temperature
sorbents
for
use
with
HS­
ESPs
also
should
be
investigated.

°
Development
and
demonstration
of
techniques
to
improve
Hg
capture
in
units
equipped
with
an
ESP,
SDA/
ESP
or
SDA/
FF
and
burn
subbituminous
coal
and
lignite
is
needed.

1
Performance
and
Cost
of
Mercury
and
Multipollutant
Emission
Control
Technology
Applications
on
Electric
Utility
Boilers,
EPA/
600/
R­
03/
110,
October
2003,
United
States
Environmental
Protection
Agency,
Office
of
Research
and
Development,
National
Risk
Management
Research
Laboratory,
Research
Triangle
Park,
NC.
7
This
will
include
evaluation
of
coal
blending,
combustion
modifications,
use
of
oxidizing
reagents,
and
use
of
impregnated
sorbents.

°
Evaluation
and
demonstration
of
cost­
effective
ESP
retrofit
approaches
including
installation
of
ducting
to
increase
residence
times
and
use
of
circulating
fluidized
bed
absorbers
for
optimal
utilization
of
sorbents
should
be
conducted.
The
use
of
multipollutant
sorbents
that
capture
SO2
and
Hg
should
also
be
investigated.

°
Determination
and
demonstration
of
optimum
design
and
operating
conditions
for
COHPAC
applications
on
a
range
of
boiler
operating
conditions
is
needed.
This
will
include
evaluation
of
the
effects
of
air­
to­
cloth
ratios,
fabric
filter
material,
cleaning
frequencies,
and
baghouse
arrangements
on
Hg
capture.
COHPAC­
based
tests
should
be
conducted
with
both
mercury
and
multipollutant
sorbents.

°
Continued
evaluation
of
potential
leaching
or
re­
emission
of
mercury
from
sorbent/
ash
residues
that
are
disposed
of
or
utilized
is
needed.

Mercury
Control
by
Enhancing
the
Capability
of
Existing/
New
SO2/
NOx
Controls
Implementation
of
fine
PM
standards,
EPA's
Interstate
Air
Quality
Rule,
Utility
MACT
rulemaking
to
control
mercury
emissions
from
utility
boilers,
the
Clear
Skies
legislation
and
other
multi­
pollutant
reduction
bills
in
the
Congress
are
focusing
on
future
reductions
of
NOx,
SO2,
and
mercury
emissions
from
power
plants.
Also,
a
significant
fraction
of
existing
boiler
capacity
already
has
wet
or
dry
scrubbers
for
SO2
control
and
/
or
SCR
for
NOx
control.
As
such,
multipollutant
control
approaches
capable
of
providing
SO2/
NOx/
Hg
reductions
are
of
great
interest.
These
approaches
and
their
potential
impact
on
mercury
reductions
are
discussed
below.

Multipollutant
Removal
in
Wet
FGD
More
than
20
percent
of
coal­
fired
utility
boiler
capacity
in
the
United
States
uses
wet
FGD
systems
to
control
SO2
emissions.
In
such
systems,
a
PM
control
device
is
installed
upstream
of
the
wet
FGD
scrubber.
Wet
FGD
systems
remove
gaseous
SO2
from
flue
gas
by
absorption.
For
SO2
absorption,
gaseous
SO2
is
contacted
with
a
caustic
slurry,
typically
water
and
limestone
or
water
and
lime.

Gaseous
compounds
of
Hg2+
are
generally
water­
soluble
and
can
absorb
in
the
aqueous
slurry
of
a
wet
FGD
system.
However,
gaseous
Hg0
is
insoluble
in
water
and
therefore
does
not
absorb
in
such
slurries.
When
gaseous
compounds
of
Hg2+
are
absorbed
in
the
liquid
slurry
of
a
wet
FGD
system,
the
dissolved
species
are
believed
to
react
with
dissolved
sulfides
from
the
flue
gas,
such
as
H2S,
to
form
mercuric
sulfide
(
HgS);
the
HgS
precipitates
from
the
liquid
solution
as
sludge.

The
capture
of
Hg
in
units
equipped
with
wet
FGD
scrubbers
is
dependent
on
the
relative
amount
of
Hg2+
in
the
inlet
flue
gas
and
on
the
PM
control
technology
used.
ICR
data
reflected
that
average
Hg
captures
ranged
from
29
percent
for
one
PC­
fired
ESP
plus
FGD
unit
burning
subbituminous
coal
to
98
percent
in
a
PC­
fired
FF
plus
FGD
unit
burning
bituminous
coal.
The
high
Hg
capture
in
the
FF
plus
FGD
unit
was
attributed
to
increased
oxidization
and
capture
of
Hg
in
the
FF
followed
by
capture
of
any
remaining
Hg2+
in
the
wet
scrubber.
8
RD&
D
Needs
for
Wet
FGD
Systems
to
Enhance
Mercury
Capture
°
Achieving
high
Hg
removal
efficiencies
in
a
wet
scrubber
depends
on
mercury
in
the
flue
gas
being
present
in
the
soluble
Hg2+
form.
While
the
majority
of
mercury
in
bituminous
coalfired
boilers
exists
as
Hg2+,
the
fraction
available
as
Hg2+
varies.
Further,
as
discussed
above,
flue
gases
from
subbituminous
and
lignite
coal­
fired
boilers
predominantly
contain
Hg0,
which
is
insoluble.
Therefore,
to
ensure
high
levels
of
mercury
capture
in
wet
scrubbers
in
a
broad
range
of
applications,
process
means
for
oxidizing
Hg0
in
coal
combustion
flue
gas
are
needed.
RD&
D
efforts
should
be
conducted
with
the
objective
of
making
available
oxidizing
catalysts
and
reagents
by
2015.
Also,
RD&
D
efforts
should
be
undertaken
to
examine
coal
blending
as
a
means
to
increase
oxidized
mercury
content
in
flue
gas.

°
Scrubber
design
and
operating
conditions
may
require
modification
to
optimize
Hg
dissolution
in
the
scrubber
liquor.
Therefore,
optimization
research
should
be
undertaken
at
pilot­
scale
and
then
demonstrated
at
full­
scale.

°
It
has
been
noted
that
in
some
scrubbers
dissolved
Hg2+
is
reduced
to
Hg0,
which
can
be
stripped
from
the
scrubbing
liquor
and
entrained
in
the
stack
gas.
RD&
D
efforts
should
be
conducted
in
this
area
with
additives
developed
in
bench­
and
pilot­
scale
testing
and
demonstrated
at
full­
scale.

°
Since
a
significant
portion
of
the
absorbed
Hg
may
end
up
in
the
spent
scrubber
liquor
in
the
form
of
dissolved
aqueous­
phase
Hg2+,
RD&
D
should
be
conducted
to
develop
Hg
removal
techniques
from
wastewater.

°
RD&
D
efforts
be
should
be
conducted
to
make
available
multipollutant
scrubbers
capable
of
removing
SO2,
Hg,
and
NOx
from
flue
gases
of
coal­
fired
boilers.
Research
conducted
in
the
1970s
through
90s
has
investigated
removal
of
NOx
in
wet
scrubbers.
Since
use
of
wet
scrubbers
at
power
plants
is
expected
to
increase
in
the
near
future
in
response
to
regulatory
requirements,
it
is
very
desirable
to
develop
wet
scrubber­
based
technologies
capable
of
providing
simultaneous
SO2­
Hg­
NOx
control.
Such
technologies
would
not
only
make
wet
scrubbers
more
cost­
effective,
but
would
avoid
the
need
for
installing
additional
control
equipment,
especially
at
constrained
plant
layouts.

°
Full­
scale
demonstrations
should
be
conducted
to
achieve
high
levels
of
mercury
control
using
ACI
with
wet
FGD,
with
or
without
additional
oxidizing
agents.
This
is
especially
relevant
to
subbituminous­
and
lignite­
fired
boilers.

Multipollutant
Removal
in
Dry
Scrubbers
More
than
10
percent
of
the
U.
S.
coal­
fired
utility
boiler
capacity
uses
spray
dryer
absorber
(
SDA)
systems
to
control
SO2
emissions.
An
SDA
system
operates
by
the
same
principle
as
a
wet
FGD
system
using
a
lime
scrubbing
agent,
except
that
the
flue
gas
is
mixed
with
a
fine
mist
of
lime
slurry
instead
of
a
bulk
liquid
(
as
in
wet
scrubbing).
The
SO2
is
absorbed
in
the
slurry
and
reacts
with
the
hydrated
lime
reagent
to
form
solid
calcium
sulfite
and
calcium
sulfate.
Hg2+
may
also
be
absorbed.
Sorbent
particles
containing
SO2
and
Hg
are
captured
in
the
downstream
PM
control
device
(
either
an
ESP
or
FF).
If
the
PM
control
device
is
a
FF,
there
is
the
potential
9
for
additional
capture
of
gaseous
Hg0
as
the
flue
gas
passes
through
the
bag
filter
cake
composed
of
fly
ash
and
dried
slurry
particles.

ICR
data
reflected
that
units
equipped
with
SDA
scrubbers
(
SDA/
ESP
or
SDA/
FF
systems)
exhibited
average
Hg
captures
ranging
from
98
percent
for
units
burning
bituminous
coals
to
24
percent
for
units
burning
subbituminous
coal.

RD&
D
Needs
for
Dry
Systems
to
Enhance
Mercury
Capture
°
SDA
is
considered
to
be
quite
effective
in
removing
Hg2+
from
flue
gases.
Full­
scale
demonstrations
of
SDA
and
ACI
should
be
conducted
to
achieve
high
levels
of
SO2
and
mercury
controls
on
subbituminous
and
lignite­
fired
boilers.
These
demonstrations
should
include
both
ESP
and
FF
PM
controls.

°
Circulating
fluidized
bed
absorber
technology
appears
promising
to
provide
high
levels
of
SO2
and
Hg
control.
Recent
applications
of
this
technology
reflect
SO2
control
in
excess
of
90%.
As
for
mercury
control,
limited
pilot­
scale
experience
has
shown
high
mercury
removal
rates.
This
technology,
with
or
without
ACI,
should
be
demonstrated
for
mercury
control
in
several
full­
scale
tests
using
a
range
of
coals.

Multipollutant
Removal
Via
SCR
and
Wet
FGD
As
mentioned
above,
the
speciation
of
mercury
is
known
to
have
a
significant
impact
on
the
ability
of
air
pollution
control
equipment
to
capture
it.
In
particular,
the
oxidized
form
of
mercury,
mercuric
chloride
(
HgCl2),
is
highly
water­
soluble
and
is,
therefore,
easier
to
capture
in
wet
FGD
systems
than
Hg0
which
is
not
water­
soluble.
SCR
catalysts
can
act
to
oxidize
a
significant
portion
of
the
Hg0,
thereby
enhancing
the
capture
of
mercury
in
downstream
wet
FGD.

Several
studies
have
suggested
that
oxidation
of
elemental
mercury
by
SCR
catalyst
may
be
affected
by
the
following:

 
The
space
velocity
of
the
catalyst;
 
The
temperature
of
the
reaction;
 
The
concentration
of
ammonia;
 
The
age
of
the
catalyst;
and
 
The
concentration
of
chlorine
in
the
gas
stream.

DOE,
EPRI,
and
EPA
have
co­
sponsored
a
field
test
program
that
evaluated
mercury
oxidation
across
full­
scale
utility
boiler
SCR
systems.
Testing
was
performed
at
four
coal­
fired
electric
utility
plants
having
catalyst
age
ranging
from
around
2500
hours
to
about
8000
hours.
One
plant
fired
subbituminous
coal
and
three
other
plants
fired
Eastern
bituminous
coal.
The
test
results
showed
high
levels
of
mercury
oxidation
in
two
of
the
three
plants
firing
eastern
bituminous
coal
and
insignificant
oxidation
at
the
other
two
plants
(
one
firing
bituminous
coal
and
the
other,
subbituminous).
For
the
bituminous
coal­
fired
plant
with
low
mercury
oxidation,
over
50
percent
of
the
mercury
at
the
SCR
inlet
was
already
in
the
oxidized
form.
It
is
also
noted
that
the
SCR
system
at
this
plant
was
operated
with
significantly
higher
space
velocity
(
3930
hr­
1)
that
those
of
10
the
other
plants
(
1800­
2275
hr­
1).
Finally,
ammonia
appeared
to
have
little
or
no
effect
on
mercury
oxidation.

The
two
bituminous
coal­
fired
plants
at
which
high
levels
of
mercury
oxidation
across
SCRs
was
observed
were
retested
in
the
following
year
(
2002).
Again,
similar
high
levels
of
oxidation
were
observed.
Two
additional
plants
firing
bituminous
coals
were
also
tested
in
2002.
Results
of
the
tests
showed
high
levels
of
mercury
oxidation,
similar
to
the
two
plants
tested
previously.
Currently,
a
DOE­
sponsored
field
test
program
is
further
evaluating
the
potential
effect
of
SCRs
and
FGDs
on
mercury
removal.

RD&
D
Needs
for
SCR
and
Wet
FGD
Systems
to
Enhance
Mercury
Capture
 
Aging
of
SCR
catalyst
with
regard
to
mercury
oxidation
should
be
examined
in
bench­,
pilot­,
and
field
tests.

 
SCR
impact
on
mercury
oxidation
should
be
examined
for
subbituminous
and
lignitecoal
fired
boilers
and
boilers
firing
coal
blends.
These
impacts
should
be
evaluated
on
pilot­
and
field­
scales.

 
Bench­
and
pilot­
scale
research
on
understanding
the
science
behind
SCR­
Hg
interactions
should
be
continued.
This
research
has
the
potential
to
provide
valuable
information
for
optimizing
SCR
catalysts
for
combined
NOx
and
mercury
control.

Potential
Impact
of
Coal
Use
and
Availability
of
NOX/
SO2
Controls
on
Mercury
Control
In
general,
the
extent
to
which
mercury
control
approaches
discussed
above
may
be
utilized
in
the
future
would
depend
on
the
extent
to
which
coal
would
be
used
in
U.
S.
power
plants
and
the
availability
of
existing/
new
NOX/
SO2
emission
controls
in
response
to
potential
emission
reduction
requirements.

Figures
1
and
2
depict
projected
Unites
States
coal
consumption
and
production
trends
for
the
United
States,
respectively.
It
is
evident
from
Figure
1
that
the
majority
of
coal
consumed
in
the
U.
S.
is
by
the
electric
power
generation
sector
and
that
this
consumption
rate
is
expected
to
increase
in
the
future.
Figure
2
reflects
that
the
amount
of
low­
sulfur
coals
(
e.
g.,
subbituminous
coals)
produced
has
been
significant
and
this
production
is
expected
to
increase
in
the
future.
Based
on
these
data,
it
can
be
deduced
that
consumption
of
low­
sulfur
coals
in
the
power
generation
sector
is
expected
to
increase
in
the
future.
As
discussed
above,
control
of
mercury
emissions
from
boilers
firing
low­
rank
(
subbituminous
and
lignite)
coals
is
more
difficult
that
from
boilers
firing
bituminous
coals.
Considering
the
projected
increase
in
use
of
low­
sulfur
(
i.
e.,
low­
rank)
coals,
it
is
important
that
cost­
effective
approaches
for
controlling
mercury
emissions
from
boilers
firing
such
coals
be
developed
via
focused
RD&
D
efforts.
11
Figure
1.
Electricity
and
other
coal
consumption
(
million
short
tons).
2
Figure
2.
Projected
coal
production
(
million
short
tons)
by
sulfur
content.
2
The
trends
in
coal­
fired
capacity
equipped
with
SCR
and
scrubbers
based
on
EPA's
analysis
of
Clear
Skies
Act
are
shown
in
Figures
3
and
4,
respectively.
It
is
clear
from
these
figures
that
current
and
future
NOX
and
SO2
emission
reduction
requirements
are
expected
to
result
in
large
capacities
(
about
100
GW
each)
of
SCR
and
scrubber
systems
for
coal­
fired
utility
boilers,
as
early
as
2005.
Further,
these
capacities
are
expected
to
increase
at
steady
and
significant
rates.
These
projections
underscore
the
need
to
engage
in
focused
RD&
D
efforts
to
determine
costeffective
means
for
optimizing/
tweaking
these
NOX/
SO2
controls
to
achieve
mercury
control
as
a
co­
benefit
with
small
incremental
costs.

2
Source:
Annual
Energy
Outlook
2003
with
Projections
to
2025,
DOE/
EIA­
0383(
2003),
Energy
Information
Administration,
Office
of
Integrated
Analysis
and
Forecasting,
U.
S.
Department
of
Energy,
Washington,
DC
20585,
January
2003.
12
Figure
3.
Projected
capacity
of
SCR
applications
on
U.
S.
coal­
fired
utility
boilers.
3
Figure
4.
Projected
capacity
of
scrubber
applications
on
U.
S.
coal­
fired
utility
boilers.
3
3
Source:
"
2003
Technical
Support
Package
for
Clear
Skies:
Section
D:
2003
projected
impacts
on
generation
and
fuel
use,"
available
at
http://
www.
epa.
gov/
air/
clearskies/
technical.
html.
13
Summary
and
Conclusions
Although
the
potential
Hg
emissions
are
calculated
to
be
75
tons
per
year
based
on
the
Hg
content
in
coal,
the
actual
current
emissions
are
estimated
to
be
48
tons
per
year
due
to
Hg
capture
with
pollution
controls
for
PM
and
SO2.
The
reduction
at
any
individual
plant
ranges
from
0
to
98%
dependent
on
coal
type,
control
technology
type,
and
other
unquantified
factors.

A
very
limited
set
of
short
term
full­
scale
trials
of
activated
carbon
injection
have
been
carried
out
as
described
earlier
in
this
white
paper.
These
trials
do
not
cover
a
representative
range
of
control
technology/
fuel
combination
that
would
be
required
to
demonstrate
the
widely
achievable
levels
of
Hg
control
that
might
be
achieved
in
a
cost
effective
manner.
Furthermore,
they
represent
short­
term
(
4­
9
day)
continuous
operation
and
do
not
address
all
of
the
operational
issues
and
residue
impacts
that
may
be
associated
with
commercial
operation.
Therefore,
these
technologies
are
not
currently
commercially
proven
to
consistently
achieve
high
levels
of
Hg
control
on
a
longterm
basis.

These
data
provide
a
basis
for
hypothesizing
the
levels
of
Hg
reductions
that
might
be
achievable
using
technology
specifically
for
Hg
control
alone
or
enhanced
capture
in
existing
or
new
systems
for
control
of
SO2
and
NOx.
These
estimates
contained
in
Table
3
are
based
on
best
engineering
judgment
and
the
assumption
that
a
focused
RD&
D
program
is
carried
out
in
an
effective
and
expeditious
manner.

Key
observations
are
as
follows:

1.
The
database
clearly
indicates
that
Hg
emission
controls
for
low­
rank
(
subbituminous
and
lignite)
coal­
fired
boilers
are
more
difficult
than
for
bituminous­
fired
boilers.
Further,
a
significant
amount
of
low­
rank
coal
is
currently
being
used
by
the
electric
utility
industry,
and
this
use
is
expected
to
increase
in
the
future.
Accordingly,
it
is
important
to
engage
in
focused
RD&
D
efforts
aimed
at
developing
emission
controls
for
low­
rank
coal­
fired
boilers.

2.
Assuming
sufficient
development
and
demonstrations
are
carried
out,
by
2010,
ACI
with
an
ESP
has
the
potential
to
achieve
70%
Hg
control.
ACI
with
an
ESP
and
a
retrofit
fabric
filter,
or
a
fabric
filter
alone,
has
the
potential
to
achieve
90%
Hg
reduction.
Proper
design
and
consideration
of
operational
and
residue
impacts
need
to
be
incorporated
into
the
effort.

3.
Projections
reflect
that
current
and
future
NOX
and
SO2
emission
reduction
requirements
are
expected
to
result
in
large
capacities
(
over
100
GW
each)
of
SCR
and
scrubber
systems
for
coal­
fired
utility
boilers,
as
early
as
2005.
Further,
these
capacities
are
expected
to
increase
at
steady
and
significant
rates.
Ongoing
R&
D
has
the
potential
to
provide
the
basis
for
enhanced
Hg
removal
in
retrofitted
system
by
2010.
Assuming
sufficient
research
development
and
demonstration
of
representative
technologies,
by
2015
new
and
existing
systems
installed
to
control
NOx
and
SO2
(
e.
g.,
SCR+
FGD+
FF)
have
the
potential
to
achieve
90
to
95%
control
of
Hg.
Subbituminous
and
lignite
systems
may
require
Hg
oxidation
technology
and/
or
additional
advanced
sorbents
to
achieve
these
levels.
The
longer
timeframe
for
these
systems
is
driven
by
the
fact
that
more
R&
D
is
required
to
optimize
Hg
control
approaches
before
demonstrations
are
conducted.
14
4.
Cost
estimates
fall
in
a
wide
range.
It
is
projected
that
the
Hg
removal
capabilities
projected
in
Table
3
would
add
no
more
than
about
3
mills/
kWh
to
the
annualized
cost
of
power
production.
Control
by
an
enhancing/
optimizing
FGD
and
SCR
has
the
potential
to
reduce
such
costs
substantially,
since
optimized
systems
may
require
little
additional
investment
and/
or
operational
costs,
especially
for
bituminous
coals.

5.
The
projected
performance
in
Table
3
represents
the
date
by
which
the
demonstration
of
the
most
difficult
case
(
e.
g.,
lignite)
for
the
particular
technology
would
be
completed.
The
demonstrations
of
the
technology
for
easier
situations
(
e.
g.,
high­
chlorine
bituminous
coal)
could
be
completed
somewhat
earlier.
It
is
important
to
note
that
completion
of
such
demonstrations
would
represent
only
the
potential
initiation
of
the
retrofit
program
which
would
take
a
number
of
years
to
fully
implement,
assuming
of
course,
both
successful
demonstrations
and
a
regulatory
driving
force.
The
time
it
would
take
to
fully
deploy
such
technologies
would
depend
on
a
number
of
factors,
including
the
specifics
of
the
regulatory
mandates,
available
vendor
capability
to
meet
the
hardware
demand,
and
the
time
for
design
and
construction
of
the
specific
retrofit
technologies
selected.

Based
on
our
experience
with
coal­
fired
utility
boiler
retrofit
technologies,
we
estimate
that
once
a
utility
has
signed
a
contract
with
a
vendor,
installation
on
a
single
boiler
could
be
accomplished
in
the
following
timeframe:

­
ACI
on
an
existing
ESP
or
FF
could
be
installed
in
approximately
1
year;
­
ACI
and
a
retrofit
fabric
filter
(
e.
g.,
COHPAC)
could
be
retrofitted
to
an
existing
ESP
in
approximately
2
years;
and
­
a
new
SCR/
FGD/
PM/
Hg
control
system
could
be
retrofitted
in
3­
4
years
dependent
on
the
retrofit
difficulty.
­
existing
SCR
or
FGD
to
enhance
Hg
control
could
be
retrofitted
in
about
one
year
6.
Table
3
also
reflects
the
existing
capacities
associated
with
key
coal
type/
control
technology
combinations.
These
capacities,
with
the
exception
of
CS­
ESP
+
retrofit
FF
and
PM
+
dry
FGD,
are
significant,
thereby
underscoring
the
fact
that
development
of
mercury
control
approaches
would
need
to
take
into
consideration
these
key
coal
type/
control
technology
combinations.
The
relatively
low
capacity
associated
with
the
CS­
ESP
+
retrofit
FF
combination
is
not
surprising
because
in
the
absence
of
mercury
reduction
requirements,
relatively
few
plants
have
used
this
combination
to
control
residual
amounts
of
fly
ash
escaping
their
ESPs.
Again
the
relatively
low
capacity
associated
with
PM
+
dry
FGD
is
a
result
of
the
present
economics
associated
with
sulfur
reduction
via
wet
or
dry
FGD
or
firing
low­
sulfur
coal.
However,
as
discussed
above,
in
the
presence
of
mercury
reduction
requirements,
these
latter
combinations
will
offer
attractive
mercury
control
approaches.
15
Table
3.
RD&
D
goals
for
projected
cost­
effective
mercury
removal
capability
(%)
for
key
coal
type/
control
technology
combinations.
4
Projected
Hg
Removal
Capability
in
2010
by
the
Use
of
ACI4
Projected
Hg
Removal
Capability
in
2010
by
Enhanced
Multipollutant
Controls4
Projected
Hg
Removal
Capability
in
2015
by
Optimizing
Multipollutant
Controls4
Control
Technology
Existing
Capacity
(
MW)
in
20035
Bitum­
inous
(
Bit)
Low­
rank
coals
Bit.
Coals
Low­
rank
coals
Bit.
Coals
Low­
rank
coals
PM
Control
Only­
CS­
ESP
153133
706
706
NA7
NA
NA
NA
PM
Control
Only­
CS­
ESP
+
retrofit
FF
2591
90
90
NA
NA
NA
NA
PM
Control
Only­
FF
11018
90
90
NA
NA
NA
NA
PM
+
Dry
FGD
8919
NA
NA
908
60­
708
90­
958
90­
958
PM
+
Wet
FGD
48318
NA
NA
909
70­
809
90­
959
90­
959
PM
+
Wet
or
Dry
FGD
+
SCR
22586
NA
NA
90
70­
8010
90­
9510
90­
9510
4
Based
on
the
assumption
of
aggressive
RD&
D
implementation
as
outlined
elsewhere
in
this
white
paper.
5
Capacity
values
have
been
obtained
from
EMF
controls
available
in
"
EPA's
2003
Clear
Skies
Act
parsed
file
for
2010"
available
at
http://
www.
epa.
gov/
airmarkets/
epa­
ipm/
results2003.
html.
The
capacity
values
have
been
rounded
to
the
nearest
whole
number.
6
This
control
level
is
based
on
data
from
the
Pleasant
Prairie
field
tests.
7
NA
=
not
applicable.
8
Assumes
that
additional
means
to
ensure
oxidation
of
Hg0
or
innovative
sorbents
will
be
used
as
needed.
9
Assumes
that
means
to
oxidize
Hg0
will
be
used
as
needed.
Note
that
in
some
cases
this
may,
in
part,
be
accomplished
by
FF.
10
Assumes
that
additional
means
to
ensure
oxidation
of
Hg0
or
innovative
sorbents
will
be
used
as
needed.
