July
22,
2004
MEMORANDUM
From:
William
H.
Maxwell
CG/
ESD
(
C439­
01)

To:
Utility
MACT
Project
Files
Subject:
Meeting
with
TXU
Power,
July
22,
2004
Attendees
TXU
Power:

Shawn
Glacken
David
Duncan
EPA:
Robert
J.
Wayland,
OAR/
OAQPS/
ESD/
CG
Bill
Maxwell,
OAR/
OAQPS/
ESD/
CG
Discussion
At
the
meeting,
the
attached
materials
were
discussed.

Attachment
Clean
Air
Mercury
Rule
A
Summary
of
TXU
Power's
Comments
on
the
Proposed
Rule
July
22,
2004
3
Wind
1%

Nuclear
19%
Coal
43%

Natural
Gas/
Oil
22%
Coal
50%

Natural
Gas
and
Oil
21%
Nuclear
20%
Hydro
and
Other
Renewables
9%

Fuel
Mix­
TXU
*
Source:
Energy
Information
Administration,
Annual
Energy
Review
2002
Fuel
Diversity
by
Electric
Generation
Fuel
Mix­
US
Purchased
Power
(
Natural
Gas)

15%
4
Lignite
Belts
Monticello
Mine
Big
Brown
Mine
Martin
Lake
Mine
°
Texas
is
the
largest
consumer
of
coal
in
the
U.
S.

°
Texas
is
the
5th
largest
coal­
producing
state
°
9th
largest
coal
reserve
in
the
U.
S.

°
Texas
has
18
coal­
fired
power
plants
°
provide
reliable
energy
°
support
a
strong
economy
°
Coal­
mining/
Coal­
Fired
electric
generation:

°
$
10.5
billion
annually
in
total
expenditures
°
33,000
permanent
jobs
°
$
300
million
annually
in
state
and
local
revenue
Mining
in
Texas
5
TXU
Power
Coal­
Fired
Generation
°
9
Units
/
4
Plants
°
5825
Net
Megawatt
Capacity
°
60%
of
capacity
is
scrubbed
°
All
fire
Gulf
Coast
Lignite
°
8
also
fire
smaller
amounts
of
subbituminous
coal
°
3
mines
produce
>
25
million
tons
(
more
than
the
state
of
Ohio)
6
Existing
Emission
Control
Systems
SO2
Control
NOx
Control
Particulate
Control
Facility
Unit
Wet
Scrubber
(
FGD)
None
Over­

Fire
Air
Low
NOx
Burners
Baghouse
Electrostatic
Precipitators
Ammonia
Injection
Big
Brown
1
X
X
X
X
X
X
Big
Brown
2
X
X
X
X
X
X
Martin
Lake
1
X
X
X
X
Martin
Lake
2
X
X
X
X
Martin
Lake
3
X
X
X
X
Monticello
1
X
X
X
X
X
X
Monticello
2
X
X
X
X
X
X
Monticello
3
X
X
X
X
X
Sandow
4
X
X
X
X
Big
Brown
1
&
2
ESPs
and
Baghouses
are
in
series,
Monticello
1
&
2
ESPs
and
Baghouses
are
in
parallel.
7
Position
on
Proposed
CAMR
°
Strongly
support
a
national
cap
and
trade
program
under
Section
112(
n)(
1)(
A)
of
the
FCAA
°
Support
baseline
adjustment
ratios
as
proposed
utilizing
fuel
consumption
in
1999
°
Begin
monitoring
in
2010,
24
ton
cap
in
2015
and
15
ton
cap
in
2018
°
Most
cost­
effective
approach
°
No
"
hot
spots"
will
occur
°
Recommend
changes
if
final
rule
implements
a
site
average
MACT
rate:

°
Establish
an
additional
subcategory
for
Gulf
Coast
Lignite
°
Set
a
MACT
"
floor"
of
at
least
28
lb./
TBtu
for
the
Gulf
Coast
Lignite
subcategory
°
Must
have
at
least
one
year
automatic
extension
for
compliance
8
Considerations
for
Final
Rule
°
The
contribution
from
U.
S.
coal­
fired
power
plants
to
the
global
mercury
pool
is
extremely
small.

°
An
even
smaller
amount
is
deposited
locally.

°
Only
a
fraction
is
converted
to
methylmercury.

°
The
reference
dose
(
RfD)
for
health
effects
for
mercury
is
likely
to
over­
estimate
rather
than
under­
estimate
risks.

°
No
commercially
available
technology
exists
that
can
specifically
remove
mercury
from
all
coal
types
and
unit
configurations
at
the
levels
required.
9
Contribution
of
the
Global
Background
to
Mercury
Deposition
(%)

0
to
20
20
to
40
40
to
60
60
to
80
80
to
100
Source:
Electric
Power
Research
Institute
Each
square
denotes
the
%
mercury
deposited,

which
comes
from
the
global
pool
10
Health
Risks
May
Be
Over­
Estimated
"
The
levels
reported
in
this
NHANES
1999
­
2000
subsample
for
maternal­
aged
females
were
below
levels
associated
with
in
utero
effects
on
the
fetus,
or
with
effects
in
children
and
adults."

°
Children
actually
measured
were
all
below
EPA's
Reference
Dose
(
RfD)

­
Second
National
Report
on
Human
Exposure
to
Environmental
Chemicals,
CDC,
2003.
11
No
Demonstrated
Mercury
Deposition
Near
TXU
Plants
°
Brookhaven
National
Laboratory
assessment
study
included
Monticello
Plant
°
Study
found
that
only
4%
­
7%
of
mercury
emitted
is
deposited
locally.

°
Study
used
very
conservative
elemental
mercury
emission
rate,
which
lead
to
a
greater
local
deposition
rate
than
we
believe
is
occurring.

°
Found
that
mercury
plant
by
plant
emission
goals
would
do
little
to
improve
human
health.

°
One
caveat
­
if
any
fish
mercury
level
increases
are
substantiated
through
data
collection,
there
may
be
a
justification
for
plant
specific
controls.

°
1996
Texas
Natural
Resource
Conservation
Commission
(
TNRCC)
Survey
of
Mercury
Concentrations
in
the
Cypress
Creek
Basin
°
Monticello
Plant
and
Reservoir
are
in
this
basin
°
Sediment,
soil,
water
and
fish
samples
were
collected
from
1993
to
1995
°
Concluded
­
"
None
of
the
data
collected
suggest
any
local
atmospheric
source
of
mercury."
12
No
Demonstrated
Mercury
Deposition
Near
TXU
Plants
°
2000
Texas
Department
of
Public
Health
(
TDH)
and
TNRCC
East
Texas
Lake
Study
°
Twenty
East
Texas
lakes
were
evaluated,
including
Monticello
Reservoir
°
Fish
tissue
from
Monticello
Reservoir
were
well
below
the
conservative
screening
level
of
0.7
mg/
kg.

°
Red
cedar
core
analysis
showed
no
increase
in
mercury
concentrations
over
time.

°
Mercury
concentrations
in
sediment
from
Monticello
Reservoir
were
less
than
the
screening
criteria.

°
No
mercury
advisory
has
ever
been
issued
for
a
power
plant
reservoir
in
Texas
°
In
2002,
TDH
sampled
Martin
Lake
Reservoir
and
found
that
"
eating
fish
from
this
water
body
poses
no
apparent
public
health
hazard."
13
No
Commercial
Mercury
Reduction
Technologies
°
TXU
Power
has
been
unable
to
secure
a
warranty
or
guarantee
on
the
performance
of
mercury
control
technologies
for
Big
Brown
Plant.

°
Requests
for
information
were
sent
to
3
control
equipment
vendors
asking:

°
What
type
of
mercury
control
they
would
propose
to
install
on
Big
Brown
Unit
1?

°
What
percentage
of
total
mercury
removal
they
would
predict?

°
What
removal
percentage
would
they
guarantee?

°
What
conditions
would
they
place
on
the
guarantee?

°
Vendors
would
not
offer
a
guarantee.
65%
25%
20%
14
No
Commercial
Mercury
Reduction
Technologies
°
ALSTOM
Environmental
Control
Systems
Response
°
The
flue
gas
temperatures
of
the
baghouse
and
ESP
are
well
above
levels
where
mercury
sorbents
have
been
tested
at
full­
scale.
Therefore,
it
is
unknown
how
the
sorbents
will
function
at
Big
Brown.

°
New
high­
temperature
sorbents
have
been
tested
in
the
lab
and
are
showing
promise.
These
sorbents
would
have
to
be
screened
at
Big
Brown
to
determine
their
viability
for
this
application.

°
The
COHPAC
baghouse
at
Big
Brown
was
designed
at
an
air
to
cloth
ratio
of
12
ft/
min.
Because
of
the
high
pressure
drop
in
this
unit,
it
would
not
be
possible
to
add
additional
particulate
in
the
form
of
mercury
sorbents.

°
To
make
the
baghouse
compatible
for
sorbent
injection,
additional
baghouse
compartments
would
have
to
be
added
to
decrease
the
air
to
cloth
ration
to
6
ft/
min.
At
this
condition,
it
would
be
possible
to
add
sufficient
quantities
of
sorbent
that
90%
mercury
removal
could
be
possible.

°
The
alternative
would
be
to
inject
the
sorbents
upstream
of
the
ESP.
The
ESP
operates
at
a
higher
temperature
than
the
baghouse
so
temperature
issues
will
be
a
greater
concern.
Tests
have
shown
that
sorbent
injection
upstream
of
an
ESP
can
achieve
mercury
removal
on
the
order
of
70%.
15
No
Commercial
Mercury
Reduction
Technologies
°
Babcock
and
Wilcox
Company
Response:

°
Carbon
sorbent
injection
appears
to
be
a
reasonable
technology
source.

°
In
terms
of
mercury
removal
prediction
and
guarantee,
it
is
much
less
risky
to
predict
than
guarantee
specific
mercury
removal
levels
(
due
to
the
limited
number
of
and
a
high
degree
of
variability
in
pilot
and
field­
scale
test
results).

°
Mercury
control
technologies
are
highly
coal
and
boiler/
AQCS
configuration
dependent.

°
It
will
only
be
after
multiple
demonstrations
have
been
completed
before
all
the
anomalies
are
sorted
out
in
order
for
suppliers
to
take
on
the
risk
of
mercury
removal
guarantees.
16
No
Commercial
Mercury
Reduction
Technologies
°
Wheelabrator
Air
Pollution
Control
Response:

°
Would
most
likely
propose
activated
carbon
injection
upstream
of
the
fabric
filter
and
downstream
of
the
ESPs.

°
Reluctant
to
provide
a
guaranteed
mercury
removal
due
primarily
to
the
unique
nature
of
individual
facilities.

°
The
baghouse
air
to
cloth
ratio
is
higher
than
normally
designed
for
a
mercury
removal
system.

°
Other
issues
that
are
taken
into
consideration
when
providing
guarantees
­


Flue
gas
mixing
characteristics

Operating
temperatures

Fabric
filter
pressure
drop

Fuel
properties

Fan
capacity

Sorbent
injection
rates

Duct
residence
times

Filter
cake
contact
time
17
ICR
Data
Quality
°
Lack
of
quality
control
on
Part
II
testing
methods
°
Deadlines
did
not
allow
time
to
qualify
laboratories
°
No
NIST
mercury
standard
was
available
for
high
ash/
moisture
content
coals
°
Insufficient
and
biased
sampling
of
steam
generating
units
°
Part
III
sample
was
skewed
toward
wet
and
dry
scrubbed
units
°
Sized
of
the
data
set
(
80+
units
representing
1,140
units)
was
too
small
°
Failure
to
use
an
established
reference
method
with
known
variability
°
Ontario
Hydro
method
was
only
a
draft
reference
method
for
Part
III
°
Current
ASTM
D
6784­
02
has
a
relative
standard
deviation
of
34%

°
Failure
to
account
for
sampling
errors
due
to
poor
sampling
port
location
°
Inadequate
experience
of
testing
companies
with
required
tests
°
Large
number
(
31
0f
81)
of
Part
III
tests
exhibited
negative
removal
results
18
A
New
Subcategory
and
MACT
Floor
for
Gulf
Coast
Lignite
°
Mercury
content
of
Gulf
Coast
lignite
is
much
higher
than
ICR
Part
II
data
indicate.

°
Predominant
analytical
method
used
to
determine
mercury
in
coal
has
significant
problems
with
high
ash/
moisture
coals.

°
ICR
Part
III
coal
analysis
with
more
accurate
methods
shows
a
higher
mercury
content
that
is
approximated
by
the
Ontario
Hydro
"
inlet"
testing.

°
Recent
analysis
of
lignite
cores
and
plant
samples
also
demonstrate
higher
mercury
content.

°
Proposed
MACT
floor
for
lignite
was
established
with
units
firing
the
lower
mercury
content
Fort
Union
lignite.

°
Fort
Union
ICR
Part
II
coal
analysis
and
Part
III
coal
analysis
and
"
inlet"
gas
analysis
data
show
a
relative
consistency
of
mercury
levels.

°
The
"
inlet"
Ontario
Hydro
tests
on
the
5
Gulf
Coast
lignite
ICR
Part
III
units
measured
mercury
levels
5
times
higher
than
those
of
the
5
Fort
Union
units
used
to
establish
the
lignite
floor.
19
Gulf
Coast
Lignite
Mercury
Content
Average
Lignite
Hg
Analysis
Results
in
lb./
Trillion
Btu
(
Minimum
 
Maximum)

Plant/
Unit
ICR
Part
II
ICR
Part
III*
TXU
Core
TXU
Plant
Big
Brown
1
13.88
(
3.13
 
24.52)
32.92
(
31.89
 
33.59)
22.42
(
7.97
 
39.16)
23.18
(
11.26
 
41.10)

Big
Brown
2
13.76
(
3.13
 
24.52)
22.42
(
7.97
 
39.16)
24.19
(
9.82
 
46.49)

Martin
Lake
1
13.15
(
3.07
 
41.16)
20.30
(
8.60
 
32.34)
23.29
(
9.13
 
51.35)

Martin
Lake
2
13.15
(
3.07
 
41.16)
20.30
(
8.60
 
32.34)
22.67
(
13.83
 
37.60)

Martin
Lake
3
13.15
(
3.07
 
41.16)
20.30
(
8.60
 
32.34)
25.50
(
9.35
 
78.05)

Monticello
1
19.44
(
5.35
 
40.13)
46.29
(
38.55
 
61.21)
44.04
(
22.68
 
61.40)
29.29
(
10.11
 
78.86)

Monticello
2
19.44
(
5.35
 
40.13)
44.04
(
22.68
 
61.40)
32.37
(
10.99
 
61.37)

Monticello
3
19.44
(
5.35
 
40.13)
48.48
(
44.92
 
54.96)
44.04
(
22.68
 
61.40)
29.70
(
6.80
 
55.33)

Sandow
4
9.72
(
2.60
 
20.17)
30.87
(
23.40
 
41.92)

*
TXU
found
an
error
in
the
calculation
of
mercury
concentration
in
coal
(
lb./
TBtu)
in
the
stack
test
reports.

The
"
as­
received"
heat
content
was
used
instead
of
the
"
dry"
heat
content.
The
values
reported
in
these
comments
have
been
corrected.
20
Gulf
Coast
Lignite
Mercury
Content
(
cont.)

Gulf
Coast
Lignite­
Fired
Units
Plant/
Unit
ICR
Part
II
Avg.
Hg
(
lb/
TBtu)
Test
Method
ICR
Part
III
Avg.
Hg
(
lb/
TBtu)
Test
Method
ICR
Part
III
Avg.
"
Inlet"

Hg
(
lb/
TBtu)

Big
Brown
1
13.88
D3684
32.92
D6414
27.96
Big
Brown
2
13.76
D3684
Dolet
Hills
7.62
Unknown
Limestone
1
15.39
D3684
15.98
D3684
28.03
Limestone
2
15.39
D3684
Martin
Lake
1
13.15
D3684
Martin
Lake
2
13.15
D3684
Martin
Lake
3
13.15
D3684
Monticello
1
19.44
D3684
46.29
D6414
44.08
Monticello
2
19.44
D3684
Monticello
3
19.44
D3684
48.48
D6414
35.41
Pirkey
1
30.81
Microwave
/
CVAA
San
Miguel
1
6.67
D3684
Sandow
4
9.72
D3684
TNP­
One
1
3.61
D3684
TNP­
One
2
3.61
D3684
26.99
D6414
25.88
21
Fort
Union
Lignite
Mercury
Content
Fort
Union
Lignite­
Fired
Units
Plant/
Unit
ICR
Part
II
Avg.
Hg
(
lb/
TBtu)
Test
Method
ICR
Part
III
Avg.
Hg
(
lb/
TBtu)
Test
Method
ICR
Part
III
Avg.
"
Inlet"

Hg
(
lb/
TBtu)

Antelope
Valley
1
6.56
Unknown
6.03
EPA
7471
5.80
Antelope
Valley
2
6.56
Unknown
Coal
Creek
1
9.87
Unknown
10.47
Unknown
9.58
Coal
Creek
2
9.87
Unknown
Coyote
1
12.43
Unknown
10.21
EPA
3050/
7471
13.72
Heskett
2
7.97
Unknown
7.99
EPA
7471A
7.48
Leland
Olds
1
6.83
Unknown
Leland
Olds
2
6.83
Unknown
4.86
EPA
7471
5.28
Lewis
&
Clark
1
9.03
Unknown
11.46
EPA
7471A
17.00
Milton
R
Young
1
10.67
Unknown
Milton
R
Young
2
10.67
Unknown
7.59
Unknown
8.91
Stanton
1
8.29
Unknown
7.66
EPA
3051/
7471
6.68
Stanton
10
8.29
Unknown
6.89
EPA
3051/
7471
6.30
Unit
names
listed
in
bold
font
are
the
units
upon
which
the
lignite
MACT
floor
was
set.
22
New
MACT
Floor
for
Gulf
Coast
Lignite
Plant/
Unit
Id.
Control
Configuration
Emission
Rate
(
lb./
TBtu)

Big
Brown
Unit
1
ESP­
CS
+
Baghouse
30.12
Limestone
Unit
1
ESP­
CS
+
Wetscrubber
15.89
Monticello
Unit
1
ESP­
CS
&
Baghouse
44.08*

Monticello
Unit
3
ESP­
CS
+
Wetscrubber
22.36
Average
28.11
*
Used
inlet
analysis
since
outlet
was
about
27%
higher
than
the
inlet
measurement
and
the
coal
content.

Average
with
Twin
Oaks
Power
Unit
2
(
TNP­
One
2)
included
is
25.76
lb./
TBtu
with
a
18.36
standard
deviation.
This
unit
uses
fluidized­
bed
technology.
23
Review
of
Position
on
Proposed
CAMR
°
Strongly
support
a
national
cap
and
trade
program
under
Section
112(
n)(
1)(
A)
of
the
FCAA
°
Support
baseline
adjustment
ratios
as
proposed
utilizing
fuel
consumption
in
1999
°
Begin
monitoring
in
2010,
24
ton
cap
in
2015
and
15
ton
cap
in
2018
°
Most
cost­
effective
approach
°
No
"
hot
spots"
will
occur
°
Recommend
changes
if
final
rule
implements
a
site
average
MACT
rate:

°
Establish
an
additional
subcategory
for
Gulf
Coast
Lignite
°
Set
a
MACT
"
floor"
of
at
least
28
lb./
TBtu
for
the
Gulf
Coast
Lignite
subcategory
°
Must
have
at
least
one
year
automatic
extension
for
compliance
