Docket
No.
A­
79­
15
*
Singh,
Kamal
VAO1,11/
2/
99
6:
24
PM
­
0500,

.'
From:
"
Singh,
Kamal
VAO1"
<
KSingh@
icfconsulting.
com
To:
'
DavidAilor'
<
dailor@
accci.
org>
Cc:
"
Palmer,
Walter
VAO1"
<
WPalmer@
icfconsulting.
cow
Subject:
Date:
Tue.
2
Nov
1999
18:
24:
37
­
0500
David:
Please
find
attachled
the
list
that
Walt
Palmer
discussed
with
you
earlier
today.
Kamal
Singh.

inputAssumptions.
doc
Item
No.
XII­
C­
4
EPA
AIR
DOCKET
ic­
I
Printed
for
David
Ailor
<
dailor@
nopa.
org>
1
Preliminary
Draft
List
of
Model
Input
Values
ICF
Consulting
is
conducting
screening­
level
atmospheric
dispersion
modeling
to
project
atmospheric
concentrations
and
surface
deposition
rates
of
gaseous,
vapor
phase,
and
particulate
air
pollutants
emitted
from
coke
oven
facilities.

The
model
first
simulates
plume
rise
using
USEPA
guideline
Buoyant
Line
Source
Model
(
BLP)
and
second
simulates
dispersion
and
deposition
using
USEPA
endorsed
Industrial
Source
Complex
(
ISCST3)
model.
ICF
has
conducted
a
very
preliminary
screening
run
with
this
model,
using
hypothetical
or
default
data
from
a
combination
of
sources,
and
conservative
assumptions.
The
inputs
used
for
the
screening
run
and
the
corresponding
sources
of
information
are
listed
below.

Model
Parameter
Input
Units
Value
Surface
temp
of
offtake
422
K
Ambient
temp
293
K
Average
coal
throughput
0.90
ton
coal/(
hroven

PMlO
emission
factor
0.0004
(
Ib
PMlO/
ton
coal)
PMlO
per
unit
volume
per
3.00E­
07
lb/
fi3
opacity
%

Percent
opacity
60
%

Exit
temperature
626
K
PMlO
emission
factor
0.02
(
lb
PMlO/
ton
coal)
PMlO
per
unit
volume
per
3.00E­
07
Ib/
ft3
opacity
%
Percent
opacity
60
%

Exit
temperature
626
K
Source
Geneva
Steel.
Engineeringjudgement.
Section
114
Coke
Facilities
Survey.
Table
2:
Summary
of
Battery
Design
and
Operation.
1997
data
for
A
K
Steel
Middletown.

Draft
AP­
42
Table
12.2­
2.
Assuming
post­
NESHAP
controls.
Allegheny
County
Health
Department
(
ACHD).
1995.
Liberty
Borough
PM­
10
SIP
Buoyancy
Calculations.
Henceforth
referred
to
as
 
ACHD. 
ACHD.
Geneva
Steel.

Draft
AP­
42
Table
12.2­
2.
Assuming
post­
NESHAP
controls.
ACHD.

ACHD.
Geneva
Steel.
Parameter
Lid
and
Offtake
Leaks
PMlO
emission
factor
PMlO
per
unit
volume
per
opacity
%
Percent
opacity
Exit
templeratwe
Pushing
PMlO
emission
factor
PMlO
per
unit
volume
per
opacity
YO
Percent
opacity
Exit
temperature
Travel
Car
PMlO
emission
factor
PMlO
per
unit
volume
per
opacity
%
Percent
opacity
Exit
temperature
Decarbonizing:
PMlO
emiission
factor
PMlO
per
unit
volume
per
opacity
%
Percent
ouacitv
Exit
temperature
Model
Input
Units
Value
0.000376
(
Ib
PMlO/
ton
coal)
3.00E­
07
Ib/
ft3
60
%

626
K
0.09
(
lb
PMlO/
ton
coal)
3.00E­
07
Ib/
fu
5
%

1,033
K
0.458
(
lb
PMlOhrtravel
car)
3.00E­
07
Ib/
fU
Source
Draft
AP­
42
Table
12.2­
2.
Assuming
post­
NESHAP
controls.
ACHD.

ACHD.
Geneva
Steel.

Draft
AP­
42
Table
12.2­
8.
Assuming
emissions
captured
by
travelling
hood
and
a
baghouse.
ACHD.

ACHD.
Geneva
Steel.

1/
20"
ofACHD
emission
factor
to
account
for
capture
efficiency
of
Coke
side
sheds.
ACHD.

I
10
I
%
I
ACHD.
1,033
I
K
I
Geneva
Steel.

I
0.05
I
(
lbPMlO/
ton
I
ACHD.

3.00E­
07
lb/
ft3
ACHD.

60
%
ACHD.

I
1,255
I
K
I
Geneva
Steel.

Figure
1
presents
the
battery
configuration
assumed
for
modeling
purposes.

General
Assulm~
tionsand
InDut
Values
Used
for
the
ISCST3
Modeling
For
purposes
of
ISCST3
modeling
the
coke
oven
emissions
were
assumed
to
be
released
at
the
BLPsimulated
final
plume
rise
height
plus
battery
height,
directly
above
the
coke
oven
batteries.

Modeling
options
used
in
the
ISCST3
screening
mode
include:
No
stack­
tip
downwash
(
plume
rise
determined
in
BLP),

0
Use
of
buoyancy­
induced
dispersion,

0
Use
of
calms
processing,
No
building
downwash
(
already
incorporated
as
part
of
BLP),
Default
wind
speed
profile
exponents,

0
Default
vertical
potential
temperature,

0
Flat
terrain
Dry
deposition
(
gas
and
particles)

Input
values/
assumptions
used
in
the
ISCST3
screening
include:

0
No
additiolnal
buoyancy
or
momentum
plume
rise,
Actual
dimensions
and
orientation
of
the
battery
line,
All
possible
wind
speedhtability
combinations
every
10
degree
wind
direction
from
zero
to
360,

0
Mixing
heights
set
to
a
nominal
value
of
5,000
meters,
Ambient
temperature
of
293
K,
Anemometer
height
of
lO­
m,
Surface
pressure
of
1000
mb,
Rural
dispersion
curves,
A
nominal
value
of
0.10
meters
(
suburban)
for
surface
roughness
(
zo),
Friction
velocity,
u,=
ku
/
In
(
dz&
where
k
is
von
Karman
constant
of
0.4,
No
precipitation,
Monin­
Obukhov
length
as
a
function
of
surface
roughness
and
stability
class
as
defined
by
Golder
(
19721,
State
of
Vegetation
­
Unstressed
and
Active,
Default
values
for
gases
in
vegetatiodland:
cuticle
resistance,
ground
resistance
and
reference
resistance
of
pollutant
to
reactivity
through
leaf
­
expressed
in
terms
of
SOZ:
cuticle
resistance
30
s/
cm,
ground
resistance
=
10
dcm,
pollutant
reference
reactivity
=
8,
Incoming
solar
radiation:
based
on
scheme
implemented
in
MPRM
(
Irwin
et
al.,
1988)
­
per
Table
2,
Leaf
Area
Index
­
ratio
of
leaf
surface
area
divided
by
ground
surface
area;
0.2
based
on
a
urban
or
built­
up
area,
Deposition
over
land
only,
Pollutant­
specific
parameters
for
the
two
primary
gas­
phase
pollutants
from
coke
oven
emissions,
benzene
and
naphthalene,
per
Table
3
(
molecular
difhsivity
from
Fletcher
et
al.,
1997;
mesophyll
resistance
from
EPA,
1993a),
Benzene
Soluble
Organic
(
BSO)
emission
rates
for
hgitive
emission
points
per
Table
4.

Table
2:
Assumed
Solar
Radiation
Values.

Pollutant
Molecular
Pollutant
Reactivity
Mesophyll
Resistance
Term
(
r,)
Diffusivity
@,)
Parameter
(
A)
Benzene
0.0912
cm2/
s
10
(
moderate
value)
10
(
high
degree
of
solubility)
Naphthalene
0.0590
cm2/
s
10
(
moderate
value)
100
(
insoluble
in
water)

Table
4:
Emission
Rate
Inputs
Used
for
the
ISCST3
Modeling.
I
Emission
Point
I
Best
Estimate
of
I
Source
BSO
Emission
Rate
(
Ib/
hr)
Doors
1.0943
EPA
method
303
observations
0.0017
0.0069
0.0393
0.01
for
A
K
Steel,
Middletown.
EPA
method
303
observations
for
A
K
Steel,
Middletown.
EPA
method
303
observations
for
A
K
Steel,
Middletown.
EPA
method
303
observations
for
A
K
Steel,
Middletown.
Engineering
judgment.

Within
ISCST3
a
set
of
virtual
stacks
was
input
to
represent
emission
releases.
Each
virtual
stack
is
a
group
of
19
ovens
covering
an
area
of
246.5
1
m2,
with
18.46
m
spacing
between
each
virtual
stack.
The
modeled
battery
consists
of
4
virtual
stacks.
The
effective
stack
diameter
for
each
stack
is
17.72
m.

Co­
Located
Facilities
Emissions
from
the
byproduct
recovery
plant,
combustion
stack
and
quench
tower
are
also
included
in
the
analysis,
but
arle
modeled
using
only
the
ISCST3
(
as
the
buoyant
line
plume
rise
model
algorithm
is
not
needed).
These
co­
located
facilities
were
modeled
in
separate
model
runs.

The
byproduct
recovery
plant
was
modeled
as
an
area
source
(
90m
x
90m,
­
2
acres
­
see
Figure
2­
2)
with
an
average
of
height
of
12
feet.
This
represented
an
average
source
emission
height
of
10
feet
from
the
various
pumps,
valves
and
flanges
and
2
additional
feet
for
a
small
plume
rise.

The
combustion
stack
and
quench
tower
were
modeled
as
point
sources
with
stack
parameters
as
in
Table
5.

Combustion
Stack
250
13.8
137,400
500
Combustion
Stack
Quench
Quench
Tower
Information
Source
Tower
Information
Source
A
K
Steel,
Middletown.
100
A
K
Steel,
Middletown.
A
K
Steel,
Middletown.
14
A
K
Steel,
Middletown.
A
K
Steel,
Middletown.
183,678
Engineering
judgment.
A
K
Steel,
Middletown.
200
Engineering
judgment.

Figure
2
indicates
the
relative
location
of
these
co­
located
facilities
to
the
coke
oven
battery,
as
assumed
for
purposes
of
the
base
case
model
run.

Benzene
and
Benzene­
Toluene­
Xylene
(
BTX)
post­
NESHAP
emission
rates
for
by­
product
plant
emission
points
are
presented
in
Table
6.

Table
6:
By­
product
Plant
Emissions
(
Source:
Draft
AP­
42,
Table
12.2­
19,
Emission
Factors
for
Coke
By­
product
Recovery
Plants
­
Benzene
and
BTX,
and
Coke
Production
Rate
for
A
K
Steel,
Middletown
batterv
3
­
48.89
tonshr).
Figure
1:
Coke
Battery
Configuration
for
Coke
Oven
Modeling
(
each
rectangle
represents
one
oven).

2.2'
1
,
O'

+
i
41
­

0
0
0
1.......­........

0
t­­
242.2'
//
/

Figure
2:
Layout
of
Coke
Plant
with
Coke
Ovens,
Combustion
Stack,
Quench
Tower
and
Byproduct
Recovery
Plant.

combustion
Stack
200m
I
­
Battery
Ovens
I
Quench
Towet
