Environmental Protection Agency
Office of Air Programs
Washington, D.C.
Contract No. PH 22-68-60
SYSTEMS GROUP
j
-------
11130-W003-RO-00
SENSITIVITY ANALYSIS OF SELECTED AIR QUALITY
IMPLEMENTATION PLANNING PROGRAM INPUT PARAMETERS
W. D. Dickerson
July 1971
Prepared for
Environmental Protection Agency
Office of Air Programs
Contract No. PH 22-68-60
TRW SYSTEMS GROUP
7600 Colshire Drive,
-------
The work upon which the publication is based
was performed pursuant to Contract No. PH-22-
68-60 with the U. S. Public Health Service,
Department of Health, Education and Welfare.
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
2.0 PARAMETER SENSITIVITY WITH RESPECT TO MAXIMUM
POLLUTANT CONCENTRATION 3
3.0 PARAMETER SENSITIVITY WITH.RESPECT TO DEVICE
ANNUAL CONTROL COST 19
4.0 PARAMETER SENSITIVITY SUMMARY 41
5.0 REFERENCES 43
APPENDICES
A DIFFUSION MODEL DESCRIPTION 45
B CONTROL COST FORMULATION 55
C EMISSION INVENTORY DATA 69
-------
ILLUSTRATIONS
Page
2-1 Maximum Concentration Variation with
Normalized Plume Rise 7
2-2 Maximum Concentration Variation with
Effective Stack Height (Average) 8
2-3 Maximum Concentration Variation with
Effective Stack Height (Regional). 9
2-4 Maximum Concentration Ratio Variation with
Effective Stack Height Ratio 11
A-l Source Coordinate System for Diffusion Model 45
A-2 Interpolation of Wind Directions 46
A-3 Virtual Point Source Concept 53
A-4 Area Utilization Concepts 54
C-l Stack Height Range for Powerplants 70
C-2 Exit Velocity Range for Powerplants 71
C-3 Exit Diameter Range for Powerplants. 72
C-4 Exit Temperature Range for Powerplants 73
C-5 Effective Stack Height Range for Powerplants 76
C-6 ACFM Range for Powerplants 77
C-7 Effective Stack Height Range for Fuel Combustion
Sources (Excluding Powerplants) 78
C-8 Exit Velocity Range for Fuel Combustion Sources
(Excluding Powerplants) 79
C-9 Exit Diameter Range for Fuel Combustion Sources
(Excluding Powerplants) 80
C-10 Exit Temperature Range for Fuel Combustion Sources
(Excluding Powerplants) 81
C-ll ACFM Range for Fuel Combustion Source (Not Including'
Powerplants) 82
-------
ILLUSTRATIONS (Cont'd)
Page
C-12 Effective Stack Height Range for Industrial
Process Sources 83
C-13 Exit Velocity Range for Industrial Process
Sources 84
C-14 Exit Diameter Range for Industrial Process
Sources 85
C-15 Exit Temperature Range for Industrial Process
-------
TABLES
Page
2-1 Diffusion Model Input Parameters 4
2-2 Effective Stack Height Parameters For Example Case 12
2-3 Summary of Parameter Sensitivity for Example Case 13
2-4 Nominal Effective Stack Height Parameter Values 15
2-5 Percent Parameter Change Required to Produce a
5 Percent x Change 16
2-6 Suggested Maximum Value Range for Effective
Stack Height Parameters 17
3-1 Control Cost Input Parameters 20
3-2 Control Device Description 24
3-3 Preset Device Data 27
3-4 Control Costs Equations 29
3-5 Regional Cost Sensitivity
- Small Plant (20-103 ACFM) 30
- Medium Plant (150-103 ACFM) 31
- Large Plant (600-103 ACFM) 32
- Extra Large Plant (2000-103 ACFM) 33
3-6 Example Control Cost Values 34
3-7 Typical Control Cost Breakdown (% of Total) 37
3-8 Percent Annual Control Cost Deviation Due to
50% Deviation in Average Parameter Values 38
3-9 Suggested Maximum Control Cost Parameter Uncertainty 39
4-1 Suggested Maximum Parameter Range 42
A-l Coefficients for a Calculation 50
z
B-l Control Measure Purchase Cost Equations 56
B-2 Existing Device Correction Factors. . i 65
C-l Average Parameter Range for Powerplants 75
C-2, Effective Stack Height Diameter Range 87
-------
1.0 INTRODUCTION
A major problem in the analysis of regional air pollution abatement
(control) strategies is the collection and verification of an adequate data
base. If computer programs (such as the Air Quality Implementation Planning
Program [IPP 1970]) are used, the scope of the data base must be expanded to
include: (1) effective stack parameters* (stack height, exit velocity, exit
diameter, and exit temperature) and (2) regional cost data (interest rate,
utility costs, labor rates, and fuel costs). In practice these data are
often difficult to define; the effective stack parameters because of the
physical measurements involved and the cost parameters because regionwide
average values, suitable for all sources, must be found.
As an aid in the definition and use of these parameters, their sensi-
tivity, with respect to pollutant concentration values and/or annual device
cost are determined. The study utilizes the atmospheric diffusion model
and the control cost model of the IPP (described in detail in Appendices A
and B). Since these models use generally accepted techniques, and have
been verified through extensive use, the study conclusions and recommenda-
tions are generally applicable. The study results are presented such that:
(1) the accuracy requirement of each parameter can be determined as a func-
tion of the remaining parameters and a maximum allowable pollutant concen-
tration and/or annual control cost error, and (2) given the complete set of
parameters and parameter accuracy estimates, the resulting maximum possible
error in pollutant concentration and/or annual control cost can be estimated.
Chapter 2.0 presents the sensitivity analysis of the effective stack
height parameters for individual sources. The source parameter sensitivity
is determined with respect to the maximum annual average pollutant concen-
tration value (x ) produced by a given source. A general formulation for the
the variation of x as a function of the uncertainty in any one or all of
the effective stack parameters is obtained. The formulation is then used to
determine the parameter sensitivity for several source category-size com-
binations. The parameter values used in the analysis are derived from a
survey of 5 detailed emission inventories (Appendix C).
*
The term parameter is used here to denote inventoried (i.e., input) data
-------
Chapter 3.0 presents the sensitivity analysis of the regional cost
parameters with respect to the annual control cost of each physical device
available in the IPP control cost model. This general analysis does not
include fuel substitution measures since these are highly dependent on
existing source configuration and user input sulfur limitations. Annual
control cost equations are summarized for each device (subject to the pre-
set data assumptions of the model). The sensitivity of the annual cost for
each device with respect to interest rate, labor rate, electricity cost,
water cost, and natural gas cost, is then determined for a range of plant
sizes.
Chapter 4.0 presents a general summary of the parameter sensitivity
results together with recommendations for the maximum variability of the
parameters investigated.
This study has shown that relating parameter uncertainty to variable
error on a per-source basis is a valid approach as an aid in the determina-
tion and use of the effective stack parameters and regional cost data. In
view of this, it is recommended that; (1) in studies in which average param-
eters are determined (existing data summaries, literature searches, etc.),
the parameter uncertainty be related to variable error as a measure of the
average parameter validity, and (2) an extension of the cost sensitivity
study be made to include the preset data within the cost equations (purchase
cost parameters, labor quantity required per device, etc.) and the fuel sub-
stitution measures. The last item is particularly important since these
-------
2.0 PARAMETER SENSITIVITY WITH RESPECT
TO MAXIMUM POLLUTANT CONCENTRATION
In this chapter those point source emission inventory stack parameters
used by the IPP diffusion model are analyzed to determine their sensitivity
with respect to.maximum long term pollutant concentration values. Since
this study is for individual source parameters, the maximum pollutant con-
centration values are for a single source and not the region wide maximum
concentration value. This measure of the parameter sensitivity was chosen
because it is the variable most sensitive to input parameter changes. In
addition, the maximum concentration value provides the best single indica-
tion of the impact of a given source on the regional air quality.
The complete set of input data required by the model, summarized in
Table 2-1, consists of the emission inventory data and the meteorological
data. Since the meteorological data are, in general, region dependent, a
nominal set was selected for this study. The determination of this nominal
set was done as follows:
• Stability Wind Rose - As described in Appendix A, the
stability wind rose gives the relative frequency of
occurrence of each combination of the 16 wind speed
directions, 6 wind speed categories and 5 stability
categories. For the study of point source parameters,
however, only a single wind direction is required. To
determine a nominal set of data, average stability-speed
data was obtained from the stability wind roses used in
the Chicago, Cincinnati, Denver, St. Louis, and Washington
regions. This was accomplished by averaging the wind speed
distribution for each wind speed direction-stability combination.
• Mixing Height - For the range of parameters considered in this
study, downwind distances to the maximum concentration point
do not exceed 10 kilometers. For such travel distances, the
mixing heights above 500-600 meters do not effect the maximum
concentration value. Since the normal range of average after-
noon mixing heights is from 700 to 2,000 meters, a value of
1,000 meters was artibrarily picked for this study (check runs
-------
Table 2-1. Diffusion Model Input Parameters
Parameter
Symbol
Meteorological Parameters
Units
Stability Wind Rose—
Relative frequency of
occurrence for each
combination of the
standardized 16 wind
directions, 6 wind
speed classes and
5 stability classes.
Mixing Height—
Height above surface
through which the
pollutant is mixed.
Ambient Pressure
Ambient Temperature
Pollutant Decay Factor—
Time, after release , at
which pollutant concen-
tration has decayed to
one-half its original
value.
Meters
Millibars
Degrees Kelvin
Hours
Stack height
Exit Velocity
Exit Diameter
Exit Temperature
-Emission Inventory Parameters_
h
V
D
T
Meters
Meters/second
Meters
-------
heights indicate less than 2% error occurs in the maximum
concentration). The reader should note, however, that the
mixing height value can effect air quality on a regionwide
scale. This is caused by pollutant travel distances large
enough to allow vertical dispersion interference at the
mixing height.
• Ambient Temperature and Pressure - The normal ranges of the
ambient temperature and pressure are: 270 to 300 degrees
Kelvin and 950 to 1,050 millibars, respectively. Since
these parameters are used explicitly in the effective stack
height calculation, their sensitivity with respect to maximum
concentration will also be determined. The nominal values
used for these parameters are 288° Kelvin and 1,000 millibars.
• Pollutant Decay Factor - Although this parameter can have a
pronounced effect on regionwide air quality, the commonly used
pollutant decay factors of three hours or greater have very
little effect on maximum concentration values (less than 1%
for the parameters used in this study). This is due to the
short travel distance (and therefore travel time) to the
maximum concentration point. For this study, then, an infinite
pollutant half-life value was used.
The effective stack height (H), defined by Equation (1) is composed
of the physical stack height and the plume rise (which accounts for the
buoyancy and vertical momentum of the effluent).
This formulation, due to Holland, is used in the IPP Diffusion Model.
H=h+Ah=h+^. (Im5 + 2.68 E-3PD(1 - T /T) (1)
U a
where
h = stack height
AH = plume rise (meters)
V = exit velocity (meters/second)
D = exit diameter (meters)
T = exit temperature (degrees Kelvin)
P = ambient pressure (millibars)
T = ambient temperature (degrees Kelvin)
Si
-------
The analysis procedure was to first establish a relationship between
the maximum concentration (Y ) and the effective height of release (H).
m
Then the variations in H (and therefore x ) with respect to the parameters
m
defined in Equation (1) were determined. To establish the y versus H
m
relationship, a series of runs were made varying h from 20 to 150 meters
and UAh (normalized plume rise) from 30 to 750 meters. Figure 2-1 shows
the results of these runs in terms of the normalized maximum concentration
(maximum concentration divided by the emission rate in tons/day).
In order to calculate an effective stack height from this data, an
appropriate value of U was required. Although somewhat arbitrary, this
value should reflect average conditions in order to produce realistic
effective stack height values. An average value of 4.5 meters/sec, was
obtained from the stability-speed profile used, by averaging over all
stability-speed combinations. Then, combining the stack height and normal-
ized plume rise values into effective stack height values lead to the de-
sired maximum concentration versus effective stack height curve shown in
Figure 2-2. Applying a curve to fit to the data in Figure 2-2 yields the
relationship:
X /Q = BH~A; A = 2.562, B = 47-10 , Q = emission rate (tons/day) (2)
This figure also presents the downwind distance at which the maximum
concentration occurs. The "fan" at the high H end of the distance curve
indicates that, for large effective stack heights, there exists an area
of maximum concentration as opposed to a single point. Although the
distance curve is not required in the analysis of this chapter , it does
provide useful data for locating the Y point for a given source. Using
this curve to get distances and the regional stability wind-rose to deter-
mine the most active wind direction radial should provide the Y location
for each source.
As a check on the validity of the results in Figure 2-2, a set of
X /Q versus H data was obtained by making a series of runs using the most
active wind direction stability-speed values from the Washington, D. C.,
Cincinnati, Philadelphia and St. Louis regions. A plot of these data is
shown in Figure 2-3. No regional calibration values were used in these
runs. Note that the slopes of the "actual" and average curves are the
same. This is the key factor in this study since changes and not absolute
-------
10
9
10
10
9
8
7
H .:
•v» ^
m
0
60
3.
h=50
h=20
10
9
8
7
6
6
10
100
Normalized Plume Rise (meters /second)
3 4 5676910
3 4 5*70910
3 4 6 6 7 • 9 10
Figure 2-1. Maximum Concentration Variation
-------
10
9
8
7
6
5
10
9
8
7
6
5
no
-- Cf
clOr
V
.10
0) •
t*
0) :
4J :
=58:
o
10
9
8
7
6
6
.1
Distance
ttrt
Height Q
iiiiniiimini
3 4 & 6789 10
4 6 6 7 8 9 10
4 5 6 7 8 9 10
Figure 2-2. Maximum Concentration Variation
-------
10
8
7
6
6
4
3
2
10
9
8
7
6
6
4
3
2
8
6
6
4
3
2
lllll[ll[i I:(T
10-:
: C ::: ::::::::: : =
O ::: ;:::: ::;; ::
cr> — --
a --- --
- (
*
•^
oo :::::: .:
•y l['[[jjj|[|[|l[[-!
>'
1 --
• !••• '
- i n
J.U
• = ::;;!j
I
1
:;::; j:::::::
J
C
\
\
, .1
.
. . , ,
^
Effective Stack Height
--\--
*
i . . j
.\.-,
. . C
\
i
-lot
III
1
k
LI
_
CMeters ]
-
1
3 4 6 6789 10
346679910
3 4 6678
Figure 2-3. Maximum Concentration Variation With
-------
The sensitivity of the effective stack height with respect to changes
in x is obtained from Equation (2) as
Figure 2-4 shows this relationship for a range of H/H from .2 (-80%)
to 4 (+400% change in H). The use of ratios is used throughout this
section and has the advantage of producing a direct variation in x
m
with respect to a given variation in H. For example, if the effective
stack height is assumed to be 200 meters + 10 meters, (possible error of
10%), Figure 2-4 [or Equation (3)] shows that x will, at most, be in
m
error by 25% (corresponding to H/H = 190/210 =..905).
To relate the effective stack height parameters to the height ratio,
Equation (1) is used. As an example, the variation of H/H with respect
to a new exit temperature (T..) is given by
h + 7^ [Ii5 + 2.68E-3PD(1-T /T,)]
"•
_
h + -^j [1.5 + 2.68E-3PD(l-Ta/T)]
Obviously, the height ratio is dependent on both the variation of the
parameter under consideration and the nominal values chosen for all of
the parameters. To illustrate, the variation of a typical set of stack
parameters (Table 2-2) is calculated. Using this data and Equations (3)
and (4), the sensitivity of each parameter is obtained. Table 2-3
summarizes the parameter sensitivities for several parameter accuracy
values.
The maximum possible error in Y for this example can also be
calculated if the accuracy limits of each of the parameters is known.
This is accomplished by using the maximum and minimum parameter values
to determine the minimum value of H/H (and thus the maximum value of
o
Y /X )• As an example, assume that all of the parameters in Table 2-2
have an accuracy limit of +_ 10%. Then,
72 + 18*3^6 [1.5 + 2.68*.99*3.6(1-291/455)]
minimum (H/H ) =
88 + ^ [1.5 + 2.68*1.01*4.4(1-285/465)]
= .65
Now, from Figure 2-4, /X =2.9. Thus a 20% accuracy limit for all
-------
10
9
8
7
6
6
•-lOOh
10
9
8
7
1QU
10
9
8
7
6
6
.1
H/H
3 456789 10
8 48678910
3 466781 10
Figure 2-4. Maximum Concentration Ratio Variation
With Effective Stack Height Ratio
-------
Table 2-2. Effective Stack Height Parameters
For Example Case.
Parameter
Stack Height (h)—Meters
Exit Velocity (V)—Meters/second
Exit Diameter (D)—Meters
Exit Temperature (T)—°K
Ambient Pressure (P)—mb
Ambient Temperature(T )—°K
Si
Value
80.
20.
4.
460.
1000.
288.
-------
Table 2-3. Summary of Parameter Sensitivity
For Example Case.
Maximum Possible Error in
Parameter Parameter Error=lQ% =20% =50%
h 12. 25. 90.'
V 10. 21. 120.
D 20. 75. 300.
T 25. 65. *
P 10. 27. *
T 20. 40. *
*Exceeds parameter range.
-------
of the stack parameters results in a maximum possible error in x of almost
200%. It should be noted that the "maximum possible error" value is quite
conservative since it assumes that all of the parameter values are under-
estimated (or overestimated) by the maximum variation.
To develop the sensitivity of the effective stack height parameters
in general, nominal parameter data sets applicable to various size power-
plants, fuel combustion sources (except powerplants) and industrial process
sources were utilized. These data, shown in Table 2-4, were obtained from
a survey of 6 detailed emission inventories (see Appendix C).
Table 2-5 shows the individual parameter deviation (from the nominal)
required to produce a 5% deviation in x • In general, these results show
that the exit diameter and temperature are the most sensitive. The maximum
possible error in x for all cases is less than 50%. This suggests that
the set of parameter errors in Table 2-5 could be used as a guide for
determining inventoried parameter requirements. Averaging the lowest
deviation values in each category results in the suggested parameter limits
shown in Table 2-6.
-------
Table 2-4. Nominal Effective Stack Height Parameter Values.
Parameter Values
Parameter
h(meters)
V(meters/sec)
D (meters)
T(deg. Kelvin)
H (meters)
o
Plant Size =
Rated Capacity (BTU/hr) =
ACFM =
,PP2
| FC
^ IP
X PP
< FC
^ IP
. PP
| FC
' IP
/ PP
] FC
' IP
/ pp
"
1 FC
Small
50 -106
20 -103
-
41.
18.
_
5.
6.
_
2.
2.
_
525.
500.
_
50.
28.
Medium
400 -106
150 -103
64.
47.
35.
7.
9.5
6.
3.5
2.5
4.5
465.
485.
425.
92.
69.
67.
Large
1500 -106
600 -103
83.
51.
65.
12.
13.5
6.
4.
3.
6.
440.
460.
375.
139.
92.
107.
Extra-Large
6000 '106
2000-103
122.
55.
91.
22.
17.5
6.
5.5
3.5
7.5
425.
435.
300.
290.
119.
114.
Defines plant size.
PP = Powerplants, FC = Non-Powerplant Fuel Combustion, IP = Industrial Process
-------
Table 2-5.
Percent Parameter Change Required to Produce a 5 Percent x Change.
Parameter Change (%)
Paremeter
H(meters) <
V (meters/sec) \
D (meters) \
T(deg. Kelvin)
P (millibars)
T (deg. Kelvin)
Si
I
PP*
FC
IP
PP
FC
IP
PP
FC
IP
PP
FC
IP
PP
FC
IP
PP
FC
IP
Small Plant
-
2.1
2.9
-
11.4
5.0
-
6.5
3.5
-
•12.0
6.4
-
16.5
19.2
-
13.5
6.8
Medium Plant
3.7
3.8
3.6
6.0
6.2
3.9
3.7
3.5
2,2
3.2
5.7
2.0
8.2
9.2
5.3
3.1
6.1
2.5
Large Plant
3.0
3.3
2.9
4.6
4.1
4.5
2.5
2.7
2.5
3.2
1.0
1.8
. 6.0
6.7
6.4
3.1
3.5
2.1
Extra Large Plant
4.3
4.1
2.2
3.4
4.7
9.2
2.3
2.0
8.5
2.1
2.6
1.0
4.3
5.2
26.0
2.1
2.8
1.0
-------
Table 2-6. Suggested Maximum Value Range for Effective
Stack Height Parameters.
Parameter Value Range
h(meters) + 3%
V(meter/sec) + 5%
D(meters) + 3%
T(deg. Kelvin) + 3%
P (millibars) + 5%
a —
T (deg. Kelvin) + 3%
Si
-------
3.0 PARAMETER SENSITIVITY WITH RESPECT
TO DEVICE ANNUAL CONTROL COST
In this study, an investigation is made of the sensitivity of the
regional cost parameters with respect to the annual cost of the various
control devices specified in the Implementation Planning Program. Table
3-1 gives a description of the regional cost parameters and other data
required in the control cost equations. The devices available within the
IPP are described in Table 3-2. The IPP control cost formulation, as.
described in Appendix B, together with the preset data used in the model
(Table 3-3) are used as the basis for the parameter sensitivity determina-
tion.
The annual cost for each control device is composed of the total
installed cost and the operating and maintenance cost (O&M). The total
installed cost includes the purchase cost, installation cost, and interest
charges. In general, the O&M costs consist of: (1) the electrical power
cost necessary to maintain the effluent gas flow through the control de-
vice, (2) the cost of water, chemical, and natural gas used by the device,
(3) the cost (or credit) resulting from disposal of the collected pollu-
tant, and (4) the labor cost associated with device maintenance. The O&M
costs for each device is composed of various combinations of the general
O&M formulation. The specific equations used to determine the annual
control cost for each device are summarized in Table 3-4. Subject to one
approximation, these equations were taken directly from Appendix B. The
exception is the way in which the annualized capital charge is calculated.
This calculation utilizes a Capital Recovery Factor (C.R.F.) calculated
as
C.R.F.
where i is interest rate and N is the rated life of the device. Since
the rated life (N) of the devices considered is between 15 and 20 years,
the C.R.F. can be approximated by a straight line, for an interest range
of 5 to 10%, to within 1/2% of the exact value. This approximation
allows easy manipulation of the capital charge terms.
-------
Table 3-1.
Control Cost Input Parameters
: Regional Cost Input Parameters
Labor Cost Rate (C , $/hr)
JG
This is the average hourly wage rate for unskilled labor within the
region. Although the quantity of labor required depends on the character-
istics of the emission source and the control device applied, the labor
cost rate is determined only by the regional labor market.
Interest Rate (C., %)
The prevailing interest rate. This item is used in computing an
annual cost based on the purchase and installation costs of control
devices.
Utility Costs
The cost of electricity (C , $/kwhr), natural gas (C , $/cu.ft.),
and water (C , $/gallon). Separate electricty and natural gas cost values
w
may be input for power plants (SIC 4911), industry (SIC 3XXX and 2XXX),
and commercial (all other SIC numbers).
-Control Cost Preset Parameters-
Price Coefficients
The cost of each control measure to a particular source is
determined by multiplying some measure of the source size by input price
coefficients. Usually the measure of size is the number of cubic feet of
gas per minute to be treated (ACFM). For some devices, however, the
power-generating capacity, fuel usage, or some other measure is used.
The price coefficients for most devices were taken from the Control Tech-
niques Documents for Particulates and Sulfur Oxides prepared by the
National Air Pollution Control Administration.
-------
Table 3-1. (Continued)
Installation Cost Factor
This factor is used in determining the total cost of installing a
control device on a particular pollution source. It accounts for those
cost elements which depend on the characteristics of the source and of
the device itself. The installation cost factor includes the costs of
transportation, erection, architectural and engineering, auxiliary equip-
ment, utility connection, startup, spare parts, and land and buildings".
For each control device preset in the program, a corresponding, installa-
tion cost factor, which represents a fraction of the total purchase price,
is also preset. The preset factors are those presented in the 1960 NAPCA
document, Control Techniques for Particulates.
Expected Life (years)
The length of service expected from each control device is included
in each device record. This item is used to apportion the total installed
cost of the control device into an annual charge value.
Labor Quantity (hrs/year)
This is the yearly amount of labor necessary to operate and maintain
each device. Labor quantities associated with four different plant sizes
are required. The appropriate value for each particular emission source
is selected and applied by the Control Cost Program. The total labor
expense, which is one of the major components of device operating and
maintenance costs, is calculated as the product of the quantity of labor
and the labor cost rate. Labor quantity is computed by converting skilled
and supervisory hours into equivalent unskilled hours.
Operating Cost Factors
A number of factors relating to each device are considered in the
computation of operating cost. The pressure drop (in. H-0) associated
with each device is a function of the resistance to gas flow through the
device and therefore provides a measure of the amount of power necessary
-------
Table 3-1. (Continued)
to operate it. Chemical costs ($/ton) give the relative costs of
additives used in wet scrubbing when gases in the effluent stream are
used, a fuel factor must be included. This factor corrects the computed
fuel usage to account for the various heat input requirements (see
Particulate Control Techniques Document).
Disposal Cost or Credit ($/ton)
For each pollutant collected, there must be a charge for its
disposal. This charge is associated with the control device, 'since the
form in which the pollutant is collected is quite important. If the
collected pollutant has value, either directly or by return to a process,
then the disposal charge will be negative relative to the other device
costs.
Rated Efficiency (%)
These data elements represent the basic removal efficiencies of the
control device under consideration for the various pollutants. These
efficiencies are modified within the Control Cost Program to reflect
variations in source characteristics. The modifications are made on the
basis of preprogrammed engineering criteria which take into consideration
the effect of existing control devices.
-Emission Inventory Parameters-
Effluent Gas Stream
The type and size of control devices applicable to a source are
strongly affected by the two gas stream parameters: maximum exhaust gas
volume (ACFM) and stack exit temperature (T, degrees Kelvin). For ex-
ample, considerably more expensive fabric-filter installations are re-
quired when high-temperature gas must be cleaned.
Existing Pollution Controls (n, %)
The efficiency of existing air pollution control devices. When
adding new control devices, they must be compatible with existing equipment,
-------
Table 3-1. (Continued)
and their efficiencies must be adjusted to account for pollutant removal
by the existing device.
Operating Schedule (H, hours/year)
The annual hours of plant operation.
Fuel Usage. For each combustion source, the types, quantities,
and sulfur and ash contents of all fuels presently used must be "
supplied.
Maximum Process Rate (R, BTU/hr). For industrial processes, the
maximum process rate is used as the measure of plant capacity.
Emission Rates (E, ton/day). The actual amount of each pollutant
emitted by each source.
-------
Table 3-2.
Control Device Description
Wet Scrubbers or Wet Collectors (Device Codes: 001, 002, 003; used for
particulates.
These are devices which use a liquid, commonly water, to remove
particulates or gases directly from a gas stream. The removal is accom-
plished by contact, or through an increase in the collection efficiency
of a second-stage collector by increasing the effective particle diameter.
The simplest wet scrubber is a spray chamber into which water is injected
by a spray nozzle. Wet scrubbers applied to a hot gas stream have the
additional property of gas cooling and humidification.
Mechanical (Centrifugal) Collectors (007-009; used for particulates)
These are devices which remove particulate matter from a gas stream
by using only the centrifugal force. Nothing is injected into the gas to
combine with or entrain the particles. These devices use the centrifugal
force created by spinning the exhaust gas stream to drive particulate
matter from the gas. The spinning gas stream is set up inside a cylindrical
container by tangential gas inlets, vanes, or fan action.
Electrostatic Precipitators (010, Oil, 012; used for particulates)
These are devices which remove particles from a gas stream by elec-
trically charging the suspended particles as the gas passes through a
corona discharge. The charged particles are then collected on a grounded
collection plate. High-voltage precipitators, which are the only electro-
static devices considered by the program, can be operated over a relatively
wide range of collection efficiencies. These devices produce a low pres-
sure drop and consequently have low operating costs for large gas volumes.
Low-voltage precipitators are not considered since they are generally used
for industrial hygiene purposes and not for air pollution control.
Gas Scrubber (013, used for sulfur oxides)
This is a device in which water (or another liquid) is brought into
intimate contact with the effluent gas stream. The pollutants are removed
by absorption or by chemical reaction. The cost elements of this device
-------
Table 3-2. (Continued)
were expressly tailored to gas removal. While this device can be applied
for particulate or mist removal, it is generally more costly than the
alternate devices listed here. Consequently, the program applies this
device for gas removal only.
Mist Elminators (014, 015; used for particulates)
These are devices designed to remove liquid droplets. The principal
collection mechanisms are interception and impaction. Operation is much
like that of fabric filters for dry particulates.
Fabric Filters (016, 017, 018; used for particulates)
These are devices which remove particles from a gas stream by passing
the gas through fabric tubes or envelopes. A cake of collected particles,
supported by the fabric, accomplishes the filtration. Very high efficien-
cies can be attained using fabric filters. A wide range of fabrics and
cleaning techniques are available.
Afterburners (019, 020, 021, 022; used for particulates)
These are devices which remove pollutants by combustion. The direct-
flame afterburner brings the gas stream into contact with a high-temperature
flame to achieve rapid oxidation. The catalytic afterburner operates at a
lower temperature by oxidizing the pollutants on.the surface of a catalyst.
Flue Gas Desulfurization (039, 041, 042; used for sulfur oxides)
These control measures rely upon injection of various chemicals into
the exit gas stream (alternatively, the reactive or catalytic material may
be employed in a fixed of fluidized bed). The injected chemicals either
react with sulfur oxide directly to produce a readily collectible solid or
gas, or absorb the sulfur oxide by producing a physically bonded sulfate.
The more promising methods for cleaning flue gases are:
Catalytic Oxidation (039)
This type of control for sulfur dioxide in the flue gas is a
recently developed desulfurization process. After high-temperature
oxidation of the SO- to S0_, an absorbing tower procedure using
-------
Table 3-2. (Continued)
sulfuric acid is employed. Sulfur is recovered as commercially
saleable sulfuric acid. This process is reported to be applicable
to larger existing installations and to new power generating
stations, [Stites, et al., 1969].
Limestone Injection - Dry Process (041)
This control measure employs finely ground limestone which is
injected into the boiler combustion zone. The limestone reacts
with the sulfur oxides to form calcium sulfate. The sulfate is then
removed from the gas stream by a precipitator or other particulate
control device. The process is applicable to both new and existing
coal fired electric generating facilities.
Limestone Injection - Wet Process (042)
This control measure employs the same basic injection of lime-
stone as the dry process. However, instead of collecting calcium
sulfate in the dry state, a liquid slurry of lime is used to remove
the sulfate and any unreacted sulfur oxide. The process is appli-
cable to basically the same sources as the dry process described
above. The more expensive collection equipment needed may indicate
that the wet process may be more often applied to new installations.
By-Product Manufacturing (043, 044, 045; used for sulfur oxides)
This capitalizes on the fact that either elemental sulfur or sulfuric
acid may be manufactured from a sulfur oxide-rich gas. Both products are
in relatively high demand in industrial markets, and often can bring a
price high enough to substantially offset the cost of controlling emissions.
Cases in which sulfuric acid and sulfur plants are used as control measures
are not to be confused with cases where similar plants are operated inde-
pendently in such a manner as to be emission sources themselves. Clearly,
a sulfur plant will not be considered a control measure for a sulfur plant
emission source.
-------
Table 3-3. Preset Device Data
ID
1
2
3
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
39
41
42
43
44
45
DEVICE
NAME
WET SCRUBBER HI EFFIC
WET SCRUBBER MED EFFIC
WET SCRUBBER LO EFFIC
CYCLONE HI EFFIC
CYCLONE MED EFFIC
CYCLONE LO EFFIC
ELECT PRECIP HI EFFIC
ELECT PRECIP MED EFFIC
ELECT PRECIP LO EFFIC
GAS SCRUBBER
MIST ELIMINATOR HI VEL
MIST ELIMINATOR LO VEL
FABRIC FILTER HI TEMP
FABRIC FILTER MED TEMP
FABRIC FILTER LO TEMP
CATALYTIC AFTERBURNER
CATALYTIC AB-WITH HE
DIRECT FLAME AFTERBURNER
DIRECT FLAME AB WITH HE
CATALYTIC OXYDATION
DRY LIMESTONE INJECTION
WET LIMESTONE INJECTION
H2504 PLANT-CONTACT
H2504 PLANT-2 CONTACT
SULFUR PLANT
RATED EFF
so2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
80.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
90.0
50.0
80.0
97.5
99.5
95.0
PART
98.0
90.0
80.0
85.0
75.0
60.0
99.0
95.0
90.0
80.0
99.0
85.0
99.0
99.0
99.0
95.0
95.0
95.0
95.0
0.0
0.0
98.0
0.0
0.0
0.0
MANUFACTURERS PRICE
1
0.289E 01
0.289E 01
0.126E 01
0.241E 01
0.151E 01
0.244E 00
0.424E 02
0.312E 02
0.197E 02
0.317E 01
0.266E 01
0.177E 01
0.145E 01
0.348E 01
0.266E 01
0.755E 01
0.755E 01
0.571E 01
0.571E 01
-0.124E 02
0.0
0.111E-05
0.0
0.0
0.0
2
0.228E 00
0.228E 00
0.145E 00
0.197E 00
0.157E 00
0.990E-01
0.623E 00
0.441E 00
0.318E 00
0.251E 00
0.325E 00
0.217E 00
0.838E 00
0.448E 00
0.325E 00
0.151E 01
0.15 IE 01
0.117E 01
0.117E 01
0.167E 01
0.995E-06
0.481E-16
0.800E 00
0.800E 00
0.810E 00
COEFF
3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-0.603E-16
0.0
0.0
0.0
0.0
INST
FACT
2.00
2.00
2.00
1.00
1.00
1.00
1.00
1.00
1.00
2.00
1.00
1.00
1.00
1.00
1.00
1.00
2.00
1.00
2.00
0.0
0.0
0.0
0.0
0.0
0.0
CHEM DISP COST LABOR QUANTITY*
COST S02
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
2.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-27.6
0.0
0.0
-27.6
-27.6
-15.0
PART SML
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.0
0.0
1.0
1.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.50
0.38
0.38
0.25
0.25
0.25
0.50
0.50
0.50
0.38
0.25
0.25
0.50
0.50
0.50
0.25
0.25
0.1.3
0.13
0.0
0.0
0.0
0.0
0.0
0.0
MED
1.00
0.75
0.75
0.50
0.50
0.50
1.50
1.50
1.50
0.75
0.50
0.50
1.25
1.25
1.25
0.50
0.50
0.25
0.25
0.0
0.0
0.0
0.0
0.0
0.0
LGE
2.00
1.50
1.50
1.00
1.00
1.00
3.00
3.00
3.00
1.50
1.00
1.00
2.50
2.50
2.50
1.00
1.00
0.50
0.50
0.0
0.0
0.0
0.0
0.0
0.0
X-LGE
4.00
3.00
3.00
2.00
2.00
2.00
6.00
6.00
6.00
3.00
2.00
2.00
5.00
5.00
5.00
2.00
2.00
1.00
1.00
0.0
0.0
0.0
0.0
0.0
0.0
DEV
LEE
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
20
20
20
PRS
DRP
20.0
4.0
3.0
4.0
3.0
2.0
.5
.5
.5
5.0
10.0
5.0
5.0
5.0
5.0
1.0
1.0
1.0
1.0
41.5
0.0
0.0
0.0
0.0
0.0
FUEL
FACT
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.49
0.24
1.00
0.40
0.0
0.0
0.0
0.0
0.0
0.0
-------
The cost equations were then used to determine the changes in annual
control cost for each device as a function of the regional cost input
parameters: interest rate, electricity cost, water cost, natural gas cost,
and labor cost for four plant size ranges. These results are presented
in Table 3-5 for small (ACFM<40E3), medium (40E3 • Operating time = 8760 hours
Labor cost J
General Preset data (Table 3-3)
This table is used to construct the Annual Control Cost variation
for a given variation in the input cost parameters. For example, if the
regional labor cost was known to the nearest $.2 and the electricity
cost to the nearest $.005, the maximum annual control cost variation for
the wet scrubbers would be $1,750 plus $8,289 or about $10,000. To ob-
tain a general relations ihp between the variation of control cost with
variations in the cost parameters, specific cost data must be assumed.
Then, percent control cost deviation as a function of percent cost param-
eter deviation can be determined. Table 3-6 presents typical control
cost data for four plant size ranges defined for Table 3-5. This data
was obtained from surveying data from several regions. Table 3-6 also
shows the resulting annual control cost for each device plant size
-------
Figure 3-5. Control Costs Equations.
OPERATING AND MAINTENANCE COST
Dev Total Installed Cost Electricity Liquid Chemical Fuel
Disposal
Labor
ID Purchase Cost 1+N C.R.F. Quantity Cost Quantity Cost Quantity Cost Quantity Cost Quantity Cost Quantity Cost Othtr
001 (2886. +.288 ACFM) 3 (.00702 C.+. 06088) + 1.955 10~4
-4
002 (2886.+. 288 ACFM) 3 (.00702 C.+. 06088) + 1.955 10
-4
003 (1257.+ .145 ACFM) 3 (.00702 C^. 06088) + 1.955 10
007 (2413.+.197 ACFM) 2 (.00702 C.+.06088) + 1.955 10~4
-4
008 (1507.+.157 ACFM) 2 (.00702 C.+. 06088) + 1.955 10
-4
009 ( 244.+. 099 ACFM) 2 (.00702 C^. 06088) + 1.955 10
20 ACFM H C + 3. 10 ACFM H C + N/A + N/A + E n' H/(8 Shift) 1 + H
. e _2 w
4 ACFM H C + 3. 10 ACFM H Cw "l"
3 ACFM H C + 3. 10" ACFM H GU +
4 ACFM H C + N/A +
6
3 ACFM H C +
6
2 ACFM H C +
010 (42413.+.623 ACFM) 2 (.00702 C +.06088) + (1.055 10~4 .5 ACFM H
+.34 10" 3
ACFM H) C +
Oil (31243. + .441 ACFM) 2 (.00702 C +.06088) + (1.955 10~4 .5 ACFM H
+.26 10~3
ACFM H) C +
012 (19695.+. 318 ACFM) 2 (.00702 C<+. 06088) + (1.955 10~4 .5 ACFM H
+.19 10"3
013 (3175.+ .25 ACFM) 3 (.00702 C^. 06088) + 1.955 10"5
014 (2658. + .325 ACFM) 2 (.00702 C.+. 06088) + 1.955 Ifl"4
-4
015 (1772.+.217 ACFM) 2 (.00702 C +.06088) + 1.955 10
*• /.
016 (U48.+.838 ACFM) 2 (.00702 C^. 06088) + 1.955 10 H
017 (3478.+. 448 ACFM) 2 (.00702 C.+. 06088) + 1.955 10~4
-4
018 (2658.+. 325 ACFM) 2 (.00702 C^. 06088) + 1.955 10
019 (7550.+1.515 ACFM)1 2 (.00702 C^. 06088) + 1.955 10"4
020 (7550.+1.515 ACFM)2 3 (.00702 C.+. 06088) + 1.95510~4
_ 1 ,
021 (5713 .+1.174 ACFM) 2 (.00702 GJ+. 06088) + 1.95510
022 (5713.+1. 174 ACFM)4 3 (.00702 ^+.06088) + 1.955 10~4
039 ALog (-12.41 ,
+ 1.671og1Q R+3) 1 (.00702 Cjt. 06088) + 1.955 10
041 (9.95 10"7 R-6.03
+
+
+
+
+
+
+
.).
+
+
+
ACFM H) C + N/A + N/A +
e .
5 ACFM H C£ + 3 10 ACFM H GU + EsngH/(8 Shift) 10 +
10 ACFM H C + N/A + N/A +
e
5 ACFM H C +
5 ACFM H C +
5 ACFM H Ce +
5 ACFM H C +
1 ACFM1!! Cfi +
1 icFM2H Ce +
1 UPH Ce +
1 ACFM^H Ce +
41.5 ACFM H Cfi +
10"17 R2) 1 (.00702 Ct+. 06088) + N/A +
042 (1.108 10~6 R-4.81
10~17 R2+9.572 10"8
S R-4.16 10"13 S R2) 1 (.00702 C^. 06088) +
043 ALog Q[a+.81og -(E /2)
^U iU °
+61 1 (.00746 C.+.0424) +
1
044 ALog_n[a4-. Slog, n «»
+6] 1 (.00746 C±+.0424) +
045 ALog10[a+.81og1()(Es/2)
+
+
+
+
+
*
+
+
+
+
+
+
+
+
+
+
+
+
t
L C + "/*
P P e
+ E n' H/(8 Shift) 1 + H L C +
p p e
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ N/A + H
+ N/A + H
+ E n' H/(8 Shift) 1 + H
P P
+ E n' H/(8 Shift) 1 + H
P P
+ N/A + E n' H/(8 Shift) 1 + H
• — 1 P P
t .49 ACFM H C + N/A + H
+ .245 ACFM tl C +
+ 1. ACFM3 H C +
+ H
+ H
L ce *
L C +
6
L C +
L C +
"
L C +
e
L C +
e
L C +
e
L Ce +
L C +
e
L Ce +
L Ce +
L C +
L Ce +
L Ce +
L C +
L C +
+ .404 ACFM4 H C + N/A + N/A * . H/A
+ N/
+'
+
+
+
A -En' H/(8 Shift) 27.6 +
+ N/A +
+ N/A +
- Es n^ H/(8 Shift) 27.6 +
- EB r^ H/(8 Shift) 27.6 +
+ (AI*,10<-12.*m.671o,10*«l .05
i
+ 365 P (.17+. 07 S 2)+25 10
+ 365 P (.196+. 0382 S 2>+130 IO3
- ALog1(Jl«+.8 Log10(Eg/2)+6]/20
+ALog10[b+.75Log10Eg+3]
- ALog10la+.8 Log1Q(Eg/2)+6]/20
+ALog10[b+.75Log10Eg+3)
+ 6] 1 (.00746 Cj+,0424) + N/A + N/A + N/A + N/A ~ Es "s H/ (8 Shift) 15. + N/A - ALog10{a+.81Log10(Es/2)+6) /IO
+ALog,.[b+. 75Log. _E +6)
1U 10 s
NOTES- NOTES: (continued) NOTES: (continued
For devices 010,011,012,016,017,018; if T > 532°K, new ACFM ACFM = ACFM (.097 + 266. 4/T ) n1, n1 = applied efficiencies %
value must be calculated and cooling cost°must be added to ==1 ° V a "PP-1160 e«iclencles, *
O&M cost. ACFM - ACFM (1 t 288/TQ) E , tg - emission rates, ton/day
N - installation factor (pre-set) ACFM " ACFM U + i39/TQ) R = rated capacity, BTU/hr
H • operating time - hr/yr ' ACFM - ACFM (1 + 556/To) a,b • calculated cost coefficients
L - labor quantity (pre-set) ACFM " ACFM C1 + 222/TQ) PC = coal burned, ton/day
Ci'Ce'CwfCe " C08t of interestC!), electricity ($/kwhr) ,
S - coal sulfur content, %
water($/gal), and labor ($/hr)
- antl logarithm to the base 10.
-------
Table 3-5. Regional Cost Sensitivity - Small Plant (20-10 ACFM)
3
Change in Annual Control Cost - $ 10
u>
o
Device
ID
001
002
003
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
039
041
042
043
044
045
Per %
Interest
.161
.161
.088
.089
.065
.045
.770
.563
.366
.173
.129
.086
.256
.175
.129
.777
.975
.777
.834
.022
.348
.479
N/A
N/A
Per .01 $/kwhr Per .001 $/gal Per .0001 $/ft3 Per $/hr
Elec. Cost Water Cost Gas Cost Labor Cost
4
1
1
1
3
1
1
1
1
14
.310 .311 N/A 4.370
.863 .311
.647 .311
.370 N/A
.027
.685
.766
.626
.503 N/A
.078 .311
.424 N/A
.712
.712
.712
.712
.540
.438
.723
.494
.042
N/A
N
3.325
3.325
2. .187
2.187
2.187
4.376
4.376
4.376
3.325
2.187
2.187
4.376
4.376
N/A 4.376
8.585 2.187
4.291 2.187
17.522 2.187
7,079 2.187
N/A ' N/A
/A N/A N/A N,
-------
Table 3-5. (continued) - Medium Plant (150 10 ACFM)
Change in Annual Control Cost - $ 10'
u>
Device Per % Per .01 $/kwhr Per .001 $/gal Per .0001 $/ft3 Per $/hr
ID Interest Elec. Cost Water Cost Gas Cost Labor Cost
001
002
003
007
970 37
970 7
485 5
449 10
008 .352 7
009
010 1.
Oil 1.
012
013
014
015
016 1.
017
018
019 5.
020 6.
021 5,
022 5.
039
041 2.
042 3.
212 5
908 5
368 4
947 3
860 9
722 25
482 12
760 12
993 12
722 12
459 4
513 3
788 5
767 3
728 105
.177 2.854 N/A 8.750
.435 2.854
.577 2.854
.271 N/A
.704
.136
.747
.696
.777 N/A
.295 2.854
.677 N/A
.840
.840
.840
840
087
234
623
709
231
727 N/A
850
043 N/A
044
045 N
6.563
6.563
4.376
4.376
4.376
13.126
13.126
13.126
8.751
4.376
4.376
10.939
10.939
N/A 10.939
64.388 4.375
32.180 4.375
131.412 2.188
53.092 2.188
N/A N/A
-------
Table 3-5. (continued) - Large Plant (600-10 ACFM)
Change in Annual Control Cost - $ 10"
Device
ID
001
002
003
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
039
041
042
043
044
045
Per %
Interest
3.700
3.700
1.859
1.694
1.344
.838
5.845
4.155
2.956
3.239
2,776
1.853
7.082
' 3.824
2.776
22.681
26.393
24.647
23.739
6.680
9.538
13,736
14,690
17,690
26,810
Per .01 $/kwhr Per .001 $/gal Per .0001 $/ft3 Per $/hr
Elec. Cost Water Cost Gas Cost Labor Cost
165.774 12.73 N/A 17.502
33.154 . 12.73
24.868 12.73
41.100 N/A
30.828
20.550
23.006
18.806
15.121 N/A
41.445 12.73
102.756 N/A
51.378
51.378
51.378
51.378
18.153
14.078
24.130
16.355
421.251
N/A
13.127
13.127
8.751
8.751
8.751
26.254
26.254
26.254
13.127
8.751
8.751
21.878
21.878
N/A 21.878
257.551 8.751
128.730 8.751
525.647 4.376
212.368 4.376
N/A N/A
-------
Table 3-5. (continued) - Extra Large Plant (2000-10 ACFM)
Change in Annual Control Cost - $ 10"
Device
ID
001
002
003
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
039
041
042
043
044
045
Per %
Interest
12.195
12.195
6.136
5.566
4.431
2.784
18.095
12.825
9.209
10.642
9.166
6.120
23.558
12.632
9.166
83.511
93.543
94,159
86.186
68.299
26.678
43.460
25,610
30,760
47,100
Per .01 $/kwhr Per .001 $/gal Per .0001 $/ft3 Per $/hr
Elec. Cost Water Cost Gas Cost Labor Cost
674.743 51.772 N/A 35.000
134.947 51.772
101.219 51.772
137.000 N/A
102.760
68.500
76.686
62.686
50.406 N/A
168.691 51.772
342.520 N/A
171.260
171.260
171.260
171.260
67.133
50.110
97.720
59.597
1257.586
N/A
N
26.254
26.254
17.502
17.502
17.502
52.507
52.507
52.507
26.254
17.502
17.502
43.756
43.756
N/A 43.756
858.502 17.502
429.099 17.502
1752.156 8.751
707.892 8.751
N/A N/A
•
-------
Table 3-6. Example Control Cost Values
Parameter
.
ACFM
(2)
H-hrs/yrU;
T-°k
C -%
i
C -$/kwhr
e
C -$/gal
w
C -$/cu.ft.
g
C -$/hr
a
R-BTU/hr
E -ton/day
P ^o ^
N -%
P
E -ton/day
M -%^
s
P -ton /day
S— 7
fo
Device ID
001
002
003
007
008
009
010
Oil
012
013
014
015
016
017
Small
Plant
3
20-10
8760
500
8.5
.018
.002
.44-10"3
3.
6
50-10
I.
0.
1.
0.
60.
3.
24,693
15,295
13,608
9,966
9,129
8,343
27,710
23,971
20,436
18,766
12,662
9,986
19,833
18,442
Medium
Plant
•3
150-10
8760
425
8.5
.017
.002
.44-10~3
3.
6
400-10
10.
0.
5.
0.
300.
3.
A
"Annual Device
115,452
58,008
46,143
41,417
35,021
27,703
85,558
74,358
65,377
80,079
69,203
43,251
88,955
75,340
Large
Plant
3
600 -10
8760
375
8.5
.016
.002
.44-10"3
3.
6
1500-10
100.
0.
50.
0.
3000.
3.
fo**1 f1
442,556
314,313
165,787
152,150
126,060
95,440
252,138
214,937
186,634
362,005
238,344
140,303
305,619
249,649
Extra Large
Plant
3
2000 • 10
8760
300
8.5
.016 .
.002
.44-10~3
3.
6
6000-10
300.
0.
100.
0.
6000.
3.
1,604,917
706,202
537,266
460,425
375,185
275,656
699,433
582,190
495,967
1,014,586
757,948
431,641
918,269
730,700
(contd.)
(1) Defines plant size.
(2) No typical values apply, therefore conservative value used.
-------
Table 3-6. (continued)
Annual Device Cost
Small Medium Large Extra Large
Device ID Plant Plant Plant Plant
018 17,651 70,694 231,700 671,192
019 59,003 400,551 1,612,460 5,474,634
020 43,318 275,511 1,102,932 3,730,472
021 98,639 697,113 2,822,121 9,611,703
022 53,274 348,919 1,415,832 4,819,625
039 16,751 151,315 .382,906 2,764,309
041 43,900 136,421 834,787 1,775,094
042 147,446 242,198 826,819 1,803,098
043 N/A N/A 568,667 815,534
044 | | 588,517 849,934
045 N/A N/A 1,396,614 2,393,477
-------
combination. Devices 042, 043, and 045 (by-product manufacturing) are
not applied to the small and medium plant sizes because they failed to
meet the minimum efficiency requirements. Table 3-7 shows the average
(over plant size) percent contribution to total cost for each of the
control cost equation components. It is obvious from this table that
the fuel, labor, and electricity costs are the dominant parameters.
Now, using the data from Table 3-6, the percent deviation in annual
control cost due to a specified deviation in each cost parameter can be
calculated for each device plant size combination. Table 3-8 presents
the maximum annual control cost deviation (from the four plant sizes) for
each device resulting from a ten percent cost parameter deviation. These
results show that a ten percent deviation for all cost parameters results
in less than a nineteen percent error in total annual control cost (ob-
tained by summing the individual deviations of the worst case, device 039),
Since the device costs utilize the cost parameters in different ways, the
ten percent restriction may be too severe for some of the cost parameters.
Assuming a total control cost deviation of twenty percent as an upper
limit, the maximum uncertainty of each cost parameter can be determined
from Table 3-8. These values are shown in Table 3-9.
-------
Table 3-7. Typical Control Cost Breakdown (% of Total)
DEVICE
ID
001
002
003
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
039
041
042
043
044
045
ANNUAL
CAPITAL
CHARGE
14
29
19
19
18
14
39
32
26
17
19
21
37
24
19
24
41
15
28
36
19
24
37
45
26
OPERATIONS AND MAINTENANCE
ELECT LIQ CHEM FUEL DISP
57 5 N/A N/A 6
53 11
22 13
42 N/A
38
33
13
12
11 N/A N/A
19 7 29
66 N/A N/A
54
26
31
33
2
2
1
2
N/A
11
12
14
15
16
9
11
12
6
N/A
N/A
8
10
N/A 10
71 1
51 2
82 1
67 2
N/A N/A
N/A
N/A
Credit
Credit
N/A N/A N/A N/A Credit
COST
LABOR OTHER
18 N/A
26
34
25
29
37
39
45
51
22
15
25
29
35
38
2
4
1
1 N/A
N/A 64*'.
81
76
63*
55*
N/A 174*
*Includes disposal credit.
-------
Table 3-8. Percent Annual Control Cost Deviation Due To 10% Cost Parameter Deviation
% Annual Cost Deviation (By Cost Parameter)
00
uevice
ID
001
002
003
007
008
009
010
Oil
012
013
014
015
016
017
018
019
020
021
022
039
041
042
Interest
1/2
1
1
1
1
1/2
1
1
1
1/2
1
1
1
1
1
1
1
1/2
1
1
1
1
Electricity
7
3
3
5
4
4
2
2
2
3
7
6
3
4
4
2
2
2
2
7
N/A
N/A
Water Chemical Natural Gas
1 N/A N/A
1
2
N/A
N/A N/A
1 3
N/A N/A
N/A
7
5
8
6
N/A
N/A N/A N/A
Disposal
1
2
2
2
2
2
2
2
2
1
N/A
N/A
1
1
2
2
3
1
2
11
N/A
N/A
Labor
5
7
7
7
8
5
5
6
5
5
7
7
7
7
1
2
1
1
N/A
-------
Table 3-9.
Suggested Maximum Control Cost Parameter Uncertainty
Parameter Maximum Parameter Uncertainty (%)
Interest Rate 20
Electricity Cost 10
Water Cost 20
Chemical Cost 20
Natural Gas Cost 10
Disposal Cost 10
Labor Cost 10
-------
4.0 PARAMETER SENSITIVITY SUMMARY
The analysis of Chapters 2.0 and 3.0 presented techniques for
determining the sensitivity of the effective stack height parameters
(with respect to maximum pollutant concentration, Y ) and control cost
parameters (with respect to annual control device cost). As a general
guide, the set of maximum parameter sensitivities which would result in
less than 20% error in Y or total annual control device cost were obtained
m
(Table 4-1). A more common situation arises when only a few parameters
are not known with sufficient accuracy. To determine the impact, in such
cases, an analysis using the specific parameter values in question must be
performed.
Effective Stack Parameters
The sensitivity of each of the effective stack height parameters:
h, V, D, T, P and T is highly dependent on the nominal values assumed for
cl
the complete set. For this reason, the sensitivity of a given parameter
should be determined by using the best estimate for each remaining param-
eter. The procedure is to first calculate H from Equation (1) using the
nominal parameter values.
H = 7— [1.5 + 2.68E - 3PD (1 - T /T)] (1)
T1 • -J Si
Then, H is calculated using the limit value of the parameter in question.
Finally, the corresponding maximum concentration sensitivity is determined
from Equation (2) (or from Figure 2-4, Chapter 2.0).
Xm / u \-2.502
Xm
o
(t)
To maintain less than a maximum possible error of 20%, the ratio H/H
should be l.OS^H/H >..95.
o
Control Cost Parameters
To determine the cost sensitivity for a particular source device
combination, the basic control cost equations presented in Table 3-4
should be used. If the general impact of a large cost parameter uncer-
tainty is desired, Table 3-8 should be reconstructed using the appropriate
parameter uncertainty values.
-------
Table 4-1. Suggested Maximum Parameter Range
Parameter
Stack height (meters)
Exit velocity (meters/second)
Exit diameter (meters)
Exit temperature (degrees Kelvin)
Ambient pressure (millibars)
Ambient temperature (degrees Kelvin)
Interest rate (percent)
Electricity cost ($/kwhr)
Water cost ($/gallon)
Chemical cost ($/ton)
3
Natural gas cost ($/ft )
Disposal cost ($/ton)
Parameter Range (%)
±3
+5
+3
+3
+5
+5
20
10
20
20
10
10
-------
5.0 REFERENCES
"Air Quality Implementation Planning Program, Volume I,
Operator's Manual," Environmental Protection Agency,
National Air Pollution Control Administration, Washington,
D. C., November 1970.
CLARKE, J. F. "Nocturnal Urban Boundary Layer Over Cincinnati,
Ohio," Monthly Weather Review (1969).
GIFFORD, F. A., JR., "Use Of Routine Meteorological Observations
For Estimating Atmospheric Dispersion," Nuclear Safety, 2, 47.,
1961.
HOLLAND, J. Z. "A Meteorological Survey Of The Oak Ridge Area,"
Atomic Energy Commission Report ORO-99, pp. 554-559, (1953).
HOLZWORTH, G. C. "Estimates Of Mean Maximum Mixing Depths In
The Contiguous United States," Monthly Weather Review, Vol. 92,
No. 5, pp. 235-242, (May 1964).
MARTIN, Delance 0., and Joseph A. TIKVART, "A General Atmospheric
Diffusion Model For Estimating The Effects On Air Quality Of One
Or More Sources," APCA Journal, pp. 68-148, (June 1968)
PASQUILL, F. "The Estimation Of The Dispersion Of Windborne
Material," Meteorol Magazine, 90, 1063, pp. 33-49, (1961).
PASQUILL, F. "Atmospheric Diffusion," London, D. Van Nostrand
Company (1962).
TURNER, D. B. "A Diffusion Model Of An Urban Area," Journal of
Applied Meteorology, Vol. 3, pp. 83-91, (February 1964).
-------
APPENDIX A
DIFFUSION MODEL DESCRIPTION
The atmospheric diffusion model is based upon a diffusion model
developed by Martin and Tikvart [1968]. The basic output of the model
is in the form of calculated long term average pollutant concentrations
at ground level.
The Martin-Tikvart model calculates concentrations downwind from a
set of point and area sources on the basis of the Pasquill [1962] point
source formulation. A description of the process follows:
The stack in Figure A-l represents a typical elevated point source.
The coordinate system for this source has its origin at ground level at the
base of the stack (i.e., directly below the effective point of emission).
The x-axis extends horizontally in the mean wind direction. The y-axis is
horizontal in the cross-wind (and cross-plume) direction. The z-axis is
vertical.
EFFECTIVE
STACK HEIGHT
Figure A-i. Source Coordinate System for Diffusion Model.
-------
Although the change in wind direction is a continuous function over
the long-term period, for computation purposes discrete wind directions
are specified with respect to a 16-point compass, corresponding to 22.5-
degree sectors. For seasonal or longer periods it is often assumed that
all wind directions within a given 22.5-degree sector occur with equal
frequency. Thus, the effluent could be assumed to be uniformly distributed
in the horizontal within the sector. However, this assumption would result
in discontinuities in calculated concentrations at sector boundaries. A
more reasonable distribution is obtained by using a linear interpolation
between sector centerlines. Thus, the concentration at a given receptor
location is composed of proportional contributions from both the sector
containing the receptor and from the nearest adjacent sector. The linear
interpolation term is given by (c-y)/c, where y is the crosswind distance
between the receptor and the sector centerline, and c is the sector width
at the receptor location. This concept is illustrated in Figure A-2. Note
that a SSW wind affects the receptor to the NNE of the source.
SSW WIND VECTOR (d)
/ RECEPTOR
NORTH , ,*d .x sw WIND VECTQR (
-------
The "effective stack height" (or effective height of emission
release) is the height at which the plume center line becomes horizontal.
The effective stack height is the sum of the physical stack height and an
incremental factor related to the buyancy and vertical momentum of the
effluent.
The concentration, x» at a position (x,y,z) for pollutants emitted
at (0,0,h) is given by
106Q f !/. \2l
X(x,y,z; h) - . exP rilT"/ expl~I
£ I- U U U I *• \ V i • 4, » w • •
yz L\y/J L V * / J
r / \21 (1)
*-[-i(-e)J
where:
X(x,y,z; h) =• pollutant concentration, micrograms/meter , at point
x,y,z for an effective stack height h.
Q - emission rate, grams/sec.
u - mean wind speed, meters/sec.
a »a « standard deviation of the plume concentration distri-
bution in the cross-plume and vertical directions,
meters. (
-------
Use of the linear crosswind distribution requires that the form of
Equation (2) be changed to reflect a univariate Gaussian distribution:
azu I/ITT (2)rx/16)
When an elevated stable layer occurs locally, the estimated pol-
lutant concentrations are calculated with the assumption that all the
effluent remains within the mixing layer height L, where L is defined as the
vertical distance from the ground to the bass of the stable layer. For the
model calculations, o is considered to increase in the downwind direction
until it reaches a distance x. at which a * 0.47L. Up to this distance,
L Z
the Gaussian vertical distribution is assumed, and Equation (3) is appro-
priate. At distance x, the trapping effect of the elevated stable layer
begins to be effective, and uniform mixing below the base of the stable
layer is assumed to occur at downwind distance 2x,. For distances x > 2x- ,
'I* *"— w
the average concentration is calculated with the assumption that the plume
is uniformly mixed in the vertical:
106Q(c-y)/c ' f..
X Lu(2»x/16) ^ '
For distances between x. and 2x-, x is determined by a linear inter-
polation between Equation (3), evaluated at x., and Equation (4), evaluated
at 2xL.
For a specific receptor (r) and source (s) configuration, an esti-
mate of x for each pollutant is obtained by choosing a representative
ITS
wind speed for each wind speed class and solving the appropriate equation
for every wind speed and stability class appropriate for the time period
in the geographical area of interest. The average concentration, x » *-s
obtained by summing all concentrations and weighting each one according
to its frequency for the particular wind direction, wind speed class, and
stability class. The expression for average concentration for a given
pollutant is:
-jfctt
(5)
d-1 n»l m-1
-------
where:
F, « normalized frequency during the period of interest for
wind direction interval d, wind speed class n, and
stability class m
X , = average ground-level concentration calculated from Equation
(3) or (4) as appropriate.
For each of the 16 wind direction intervals, wind speed is defined
in six categories and stability class in five categories. Thus a three-
dimensional array of 480 categories is established. However, only a few
of these wind directions result in non-zero contributions for specific
source-receptor pairs. Thus the computation time is reduced significantly.
Vertical variations in wind speed and wind direction are not accounted
for in the present model.
The representative speeds associated with the six climatological wind
speed categories (0-3, 4-6, 7-10, 11-16, 17-21 and >21 knots), are given
by the five mid-interval values of 0.67, 2.46, 4.47, 6.93 and 9.61 meters
per second, and by 12.52 meters per second (25.5 knots) for the >21 knots
category.
The five stability categories (S = 1, 2, 3, 4 and 5, in order of
increasing atmospheric stability) are defined on the basis of the criteria
stated by Turner [1964]. Stability in the lowest part of the atmosphere
is determined primarily by the net radiation and local wind speed.
Turner's classification is based upon ground-level meteorological
observations only (surface wind speed, cloud cover, ceiling), supplemented
by solar elevation data (latitude, time of day, and time of year); thus
the stability estimates can be obtained for any Weather Bureau station at
which continuous observations have been made.
The total concentration at a specific receptor r, x~ > for a given
pollutant is
where
X = the average estimated concentration at receptor r from
source s, as given by Equation (5).
-------
Using the calibration constants and background concentration values, the
total calibrated mean concentration value at each receptor, x'. (for each
pollutant) is now given by,
The values of a (x,S) used in the program are those of Pasquill
[1961] and Gifford [1961]. For computation these are represented in the
form;
az - axb + c (8)
where
a, b and c values are constants for each stability class, as shown
in Table 4-2
No a, b and c values are stated for S » 5 in Table A-l'.
Stability class S = 5 is associated with nighttime, surface inversion
conditions, and the a values for this case are the smallest normally used.
However, because of the thermal and mechanical influences of urban areas,
the lowest part of the typical urban atmosphere is less stable than its
rural counterpart. Since the Atmospheric Pollutant Concentration Program
is intended for use in urban regions, the a values for S » 4 are always
Z
used when the meteorological criteria indicate S = 5.
The minimum value of x used in calculating a is 100 meters. If x
Z
<100 meters, x is set equal to 100 meters prior to the a Calculation.
TABLE A-l
COEFFICIENTS FOR a CALCULATION
z
Stability Class (S) a b . c
1 ~ Very Unstable .001 1.890 9.6
2 - Moderately Unstable .048 1.110 2.0
3 ~ Slightly Unstable .119 .915 0.0
x > 1000 meters 2.610 . .450 -25.5
x <_ 1000 meters .187 .755 -1.4
50
-------
The mixing layer height L, has a marked diurnal, daily and seasonal
variation. However, since it is impractical to account for all these
variations, a procedure reflecting only major changes is used in the model.
The procedure determines mixing height by modifying the average afternoon
mixing height values, as tabulated by Holzworth [1964], according to the
stability class being considered. Stability classes S = 1, 2 and 3 are
afternoon conditions, with S = 1 corresponding to very unstable conditions.
When S = 1, the value of L is assumed to be 50% greater than the climato-
logical value tabulated by Holzworth; when S = 2 or 3, the climatological
value is adopted. According to Turner's criteria, S = 5 can occur only when
night-time ground-based inversion conditions exist. Since a shallow layer of
neutral or weak lapse conditions has been found to occur over urban areas
(even with strong nocturnal surface inversions in the surrounding rural
area), a mixing height of L = 100 meters is adopted for stability class
S » 5, when this class is indicated by the objective criteria. The 100-meter
value is based upon observations of Clarke [1969]. Stability class S = 4
is a.neutral stability condition which occurs either with high wind speeds
or with cloudy conditions. To find the mixing height for the transition
period between day and night, the afternoon mixing height values are averaged
with the 100 meter night time mixing height for 40 percent of the class S =
4-occurrences. The remaining 60 percent of the class utilizes the climato-
logical value.
The effective stack height, h, appearing in Equation (3) , is de-
fined as the height of the plume centerline when it becomes horizontal.
Thus h ** h* + Ah, where h* is the physical stack height and Ah is the
plume rise. The effective stack height does not appear in Equation (4) be-
cause the height of emission is immaterial after the pollutant is uniformally
mixed in the interval.
The plume rise equation used in the program (when actual point
sources are being considered) is due to Holland [1953], and is given by
V d
Ah
s
1.5 + 2.68 • 10"3P
T - T
_s a
T
s
(9)
-------
where:
V * stack gas-exit velocity (meters/sec)
s
d - stack exit diameter (meters)
u « mean wind speed (meters/sec)
P * atmospheric pressure (mb)
T • stack gas-exit temperature (°K)
s
T » ambient air temperature (°K)
d
Since this equation is appropriate for the neutral stability con-
dition, it must be modified when applied over a range of stability condi-
tions. The following modification is used to allow for a range of from
1.3 Ah, for very unstable conditions, to Ah for neutral stability.
h - h* + Ah(1.4 - 0.1S) (10)
For some point sources (e.g., power plants with tall stacks) when
the mixing height is low the effective emission height will be above the
mixing height. Based on the assumption that the plume will not disperse
downward through the stable layer, these cases are identified and elimi-
nated from consideration by the program. For area sources an average
effective height of emission must be estimated.
The program computes area source contributions by converting the
area sources to equivalent, or "virtual," point sources. In the conversion
process, both the downwind distance and source strength are dependent on
the particular source-receptor configuration.
If the total emission is assumed to be concentrated at the center of
an area source, concentrations downwind tend to be overcalculated, especially
for nearby receptors. Since uniform spread of the plume across the sector
Is assumed, it is logical to proceed a step further and assume a virtual
point source at such a distance upwind that the 22.5-degree sector used
-------
subtends the area width. This concept is illustrated in Figure A-3'. Here
the program would use x instead of x as the source-receptor downwind dis-
tance in Equation (3) or (4). The vertical spread, as measured by
still calculated using x , the actual downwind distance.
is
WIND
DIRECTION
VIRTUAL POINT
SOURCE LOCATION
AREA SOURCE
RECEPTOR
Figure A-3'. Virtual Point Source Concept.
In a similar manner, nearby receptors are affected by emissions from
only a portion of the source area and, therefore, would show excessive con-
centration values if the total area emission value were used. To correct
this, the source emission rate is multiplied by an "area utilization"
factor, Q*, which is the ratio of that portion of the source area lying
within a 22.5-degree sector upwind of the receptor (A') to the total area
(A). For example, in Figure A-4, Receptor 1 would use the total area-
emission value, Q, while Receptor 2 would use the proportional amount,
QQ* - A(A'/A). Note that in this figure the virtual point source would be
defined by the reduced area width, W.
-------
A'=AREA "SEEN" BY
RECEPTOR 2
22.5 SECTOR
WIND
DIRECTION
RECEPTOR 1
VIRTUAL SOURCE
LOCATION FOR A1
SOURCE AREA A
Figure A-4. Area Utilization Concepts.
To account for decay of the pollutant from the atmosphere, a pollutant
decay factor is applied to the concentration value X, as determined from
Equation (3) or (4). The time-based decay factor is given by exp[-.693(x/u)/
(3600 T)], where T is the pollutant half-life in hours.
-------
APPENDIX B
CONTROL COST FORMULATION
The control cost formulation presented here is taken from the
Implementation Planning Program (IPP) , Volume I, Chapter 5. Within the
IPP framework, the control cost program utilizes data from the Source
File (emission inventory data), card input to the control cost program
and pre-set data in the control cost program (cost and device data).
The total annual cost resulting from the assignment of a control
device to a source involves the component costs: purchase cost, installa-
tion cost, interest charges, and operating and maintenance cost. The
calculations involving each of these component costs are described below.
Purchase Cost
The purchase cost of control devices depends, in general, upon the
characteristics and complexity of the control device and the size of the
pollution source to be controlled. In the Control Cost Program, these
parameters are used to determine a basic purchase cost equation for each
control device. The general form of this equation is:
2
y = a + bx + ex ,
where y = purchase cost in thousands of dollars, and a, b, and c are
pre-set* coefficients in the device specification. Experience has shown
that the coefficient c in the above equation can be set equal to zero in
most instances. Table B-l displays the pre-set coefficients associated
with each control device including in the Control Cost program. The para-
meter x represents the measure of the source's size, as described below.
*The term pre-set indicates values which exist in the program but may be
changed by user input.
-------
Table B-l. Control Measure Purchase Cost Equations
Control Measure
Wet Collector
001 High Efficiency )
002 Medium Efficiency /
003 Low Efficiency
Mechanical (Centrifugal)
007 High Efficiency
008 Medium Efficiency
009 Low Efficiency
Electrostatic Precipitators
010 High Efficiency
. Oil Medium Efficiency
012 Low Efficiency
Gas Scrubber
013
Mist Eliminator
014 High Velocity
015 Low Velocity
Fabric Filters
016 High Temperature Type
017 Medium Temperature Type
018 Low Temperature Type
Purchase Cost Equation
y = 2.886 + .288x
y - 1.257 + .I45x
y = 2.413 + .197x
y '- 1.507 + .157x
y = .244 + .099x
y - 42.413 + ,623x
y = 31.243 + .441x
y - 19.695 + .318x
y = 3.175 + .251x
y - 2.658 + .325x
y = 1.772 + .217x
y - 1.448 -I- .838x
y - 3.478 + .448x
y = 2.658 + .325x
See Table 3-1 for device descriptions.
"Cost Functions; y
3 3
10 dollars, x = 10 ACFM unless noted otherwise.
-------
Table B-l. Control Measure Purchase Cost Equations
(Contined)
Control Measure Purchase Cost Equation
Increased Combustion Efficiency (Afterburner)
019 Catalytic Combustion „ y = 7.550 + 1.515x
020 Catalytic with Heat Exchanger y = 7.550 + 1.515x
021 Direct Flame Combustion y = 5.713 + 1.174x
022 Direct Flame with Heat Exchanger y = 5.713 + 1.174x
3
Flue Gas Desulfurization
039 Catalytic Oxidation logio y = ~12-41 + 1.67 I°g10 x
041 Limestone Injection (dry) y = 9.95-10~7 x -6.03-10~17x2
U — 8 —182
042 Limestone Injection (wet) + 9.572-10 sx - 4.16-10 sx
By-Product Manufacture
043 Sulfuric Acid Plant (Contact Process)
044 Sulfuric Acid Plant (Double Contact) Io8iny = a + bx
045 Sulfur Plant
1 33
Cost Functions; y = 10 Dollars, x = 10 ACFM unless noted otherwise.
2
Heat exchangers are considered accessory equipment and their costs are
included under installation costs.
2
The "x" terms in these equations are the rated capacity of the source
(BTU/hr); y = 103 dollars.
4
In practice, this process only applies to coal. The parameter "s"
appearing in the cost equation is the sulfur content in percent by
weight (i.e., 0 <_ s <_ 100).
See page B-5 item (f) for parameter definitions.
-------
(a) Devices 001, 002, 003, 013
For these devices, x is exhaust gas volume in thousands
of ACFM. Since a temperature change occurs in these
wet collection devices, a new exhaust volume based on
the original gas volume is computed (for use in the
O&M calculations) as follows:
T = 294 + .097 (T -294)
n o
where T = new exit temperature (°K)
n
T = exit temperature as obtained from the
° Source File (°K)
then the new exhaust gas volume, V, is given by
V = V (T /T )
ov n o'
where V = exhaust gas volume as obtained from the
° Source File (ACFM)
(b) Devices 004-012 and 014-018
The parameter x is the exhaust gas volume in thousands of
ACFM as obtained from the Source File with no corrections
applied. If, however, the exit temperature is greater than
532°K (500°F), gas cooling must be utilized prior to the
application of devices 010-012 and 016-018. As a result,
the purchase (and O&M) costs for these devices use an adjusted
ACFM value, calculated as follows:
V = V (526.2 + .023 T )/T
o o o
T = 500°K
n
The cost associated with the cooling requirement is
determined by the following equations and is added to
the O&M costs.
If exhaust gas volume is greater than or equal to 2000
ACFM but less than 6800 ACFM,
Cost ($) = 1131.7 [e[-436 ln
-------
If exhaust gas volume is greater than or equal to 6800 ACFM but
less than 100,000 ACFM
Cost ($) = 720.17 [ e ['668 ln
-------
where T = existing exit temperature (°K)
t = 288 for device 019
139 for device 020
556 for device 021
222 for device 022
then the new (output) ACFM is determined from
V = V (T /T )
o n o'
(d) Device 039
For this device, x is the logarithm (base 10) of the
rated capacity of the plant (BTU/hr.), and y is the
logarithm (base 10) of the purchase cost (103 $).
(e) Devices 041-042
For these devices, x is the rated capacity of the plant
(BTU/hr.) as obtained from the Source File, and y = 103
dollars.
(f) Devices 043-045
For these devices, x is the logarithm (base 10) of the
amount of sulfur in the exhaust gas in tons/day, calcu-
lated as follows:
x = log1Q (Es/2)
where E = annual average S02 emission rate as obtained from
the Source File (tons/day).
The price coefficient "a" is determined by the program on the
basis of the device and the percent by volume of sulfur dioxide
in the exhaust gas. This percent is calculated using the
following formula:
P = 18.48E /(V shifts)
s
where E = annual sulfur dioxide emission rate as obtained
from the Source File (tons/day)
T = exit temperature as obtained from the Source
File (°K)
V = exhaust gas volume as obtained from the Source
File (ACFM)
Shifts = number of 8-hour shifts per day from the Source
File
P = percent sulfur dioxide by volume in the exhaust
gas
-------
a «
a •
a -
a »
-0.8218
-1.0647
-1.2937
-1.3449
a - -0.7426
a - -0.9855
a - -1.2145
a - -1.2657
Device 045 . -
a -
a -
a »
-0.5779
-0.7069
-0.7629
The quantity "a" (to be used in the cost equation) is then
determined as follows:
Device 043 Device 044
0_< P _< 3.0
3.0< P <_ 6.0
6.0< P _<10.0
P >10.0
°1 p 1 5-°
5.0< P <_ 7.0
P > 7.0
In these cases the "y" value, produced by the cost equation,
is the logarithm (base 10) of the purchase cost in millions
of dollars.
In summary, the cost equation for devices 043, 044, and 045 is:
Iog10 y 10~6 = a + b log10(Es/2)
where y = purchase cost
a = number determined by device type and
percent S02 in exhaust gas.
b = user input Manufacturer's Price Coefficient
No. 2.
A final test is made to eliminate very low efficiency devices
(based on purchase cost). If the purchase cost is less than
1.0-106 for device 043, 1.2-106 for device 044; or 2.5-106 for
device 045, the device is not applied.
When the corrosive gas, streams from acid manufacturing facilities
(SIC 2819, Process Codes 01 or 02) are controlled, more expensive control
device construction materials must be used. For these cases the Control
Cost Program utilizes the following factors to increase the device purchase
price:
If exit temperature < 330°F, multiply purchase cost by 1.7.
If exit temperature >_ 330°F, multiply purchase cost by 5.0.
-------
Installation Cost
The cost of installing a control device on a particular source is
calculated on the basis of an installation cost factor. This factor is
expressed as a percentage (decimal equivalent) of the purchase price and
is a pre-set item for each control device. The program sums purchase cost
and installation charge to determine the total installed cost for each
source-device combination.
For some control device installations, however, extensive modifica-
tions of the source are required and it is not possible to identify a
purchase cost figure independent of the installation charge. For these
devices, the total installed cost is estimated by the purchase price equa-
tions described under Purchase Cost. No additional factor is required in
these cases to account for installation at a source site. Devices in this
category are the flue gas desulfurization units (039, 041, 042) and
by-product manufacturing facilities (043, 044, 045).
Annual Capital Charge
After the total installed cost (purchase plus installation) of a
device has been determined, an annualized capital charge is calculated,
based on the pre-set rated life of the control device and the user input
prevailing interest rate. Therefore, the annual cost comprises both
depreciation of the initial total investment (installed cost from
Installation Cost) and the interest costs.
The particular accounting technique employed here is called the
Capital Recovery Factor (C.R.F.). The use of the C.R.F. allows the write-
off of the initial investment to be divided into a uniform series of
end-o f-year payment s.
The C.R.F. multipled by the initial investment cost (debt) determines
the uniform end-of-year payments necessary to repay the debt in "N" years
(the rated life of the device) with an interest rate "i" (decimal equivalent)
The C.R.F. is calculated by the following relation:
(1 + i)N - 1
-------
The annual capital charge is then computed to be the product of the total
installed cost and the C.R.F. value.
In practice, the capital recovery concept operates as follows:
(1) Payments are made at the end of each year in the amount
determined by multiplying the capital recovery factor by
the initial debt.
(2) Interest payments are made to the bondholders at the end
of each year in the amount determined by multiplying the
interest rate by the initial debt.
(3) The difference between the equal annual payments and the
interest payments to the bondholders is placed into a
depreciation account where it is assumed to draw interest
at the same rate paid to the bondholders.
The total yearly cost to the source for controlling pollution by
means of a specified device is then the sum of the annual capital charge
and the annualized operating and maintenance expense discussed in the next
subsection.
Operating and Maintenance Expense
Operating and maintenance (O&M) expenses are usually a major portion
of the annualized control device cost. The factors which must be con-
sidered in determining the annual O&M cost of a particular source-device
combination consist of: the amount of power (calculated as electrical
power) necessary to maintain the effluent gas flow through the control de-
vice, the quantity of labor required, the cost of liquid or additional fuel
used by the device, and the cost or credit resulting from disposal of the
collected pollutant. These factors are used in the following equations
to determine the O&M cost for each type of control device pre-set in the
program.
For control devices 001 through 022, a single equation is used to
compute the annual O&M cost:
O&M = AB + CD + EF + GI + JK + LM,
where A - electricity quantity (kwhr/yr) (a)*
B = electricity cost ($/kwhr) (input as regional
data)
The letters refer to the following items which describe the computation
of these data elements.
-------
C = water quantity (gal/yr) (b)
D = water cost ($/gal) (input as regional data)
E » chemical quantity (ton/yr) (c)
F = chemical cost ($/ton) (pre-set as device data)
G = fuel quantity (cu.ft./yr) (d)
I = fuel cost (S/cu.ft.) (natural gas value, input
as regional data)
J = disposal quantity (ton/yr) (e)
K = disposal cost ($/ton) (pre-set as device data)
L = labor quantity (hr/yr) (f)
M = labor cost ($/hr) . (input as regional data)
The computation of these data elements are described below. In each case
for which the exhaust gas volume is required,the new computed value should be
be used fas determined under Purchase Cost, items (a), (b), and (c)].
(a) Electricity Quantity (kwhr/yr) = (1.955-10~4)pVH
where p = pressure drop (in lUO)
V = new exhaust gas volume (ACFM)
H = operating hours (hr/yr)
If the device is one of the electrostatic precipitators
(010, Oil, 012), the following quantities are added to
the electricity quantity:
Device 010, add (0.34 10~3) VH
Device Oil, add (0.26 10~3) VH
Device 012, add (0.19 10~3) VH
(b) Liquid Quantity (gal/yr) = (3.0 10~2) VH
The liquid quantity is only computed for wet devices
(001, 002, 003, 013). For these cases the exhaust gas
volume (V) will always be the new computed volume.
(c) Chemical Quantity (ton/yr) = P H/*8-shifts
s
where P = amount of S00 removed by the device P is
s 2 s
calculated as S0_ emission rate times applied efficiency
(see Table B-2) H/(8-shifts) gives the operations time per year.
(d) Fuel Quantity = fVH
where f = 0.490, if device 019
0.245, if device 020
1.000, if device 021
-------
Table B-2. Existing Device Correction Factors
Correction Factor
Existing
Efficiency
> 0 and 140
>40 and
>60 and
>65 and
>70 and
>75 and
>80 and
>83 and
>86 and
>89 and
>91 and
>93 and
>95 and
>97 and
>99 and
160
I65
<70
17 5
18 0
183
£86
<89
l91
193
£95
£97
£99
11 00
Wet Scrubber Dry Cyclone
(001,002,003,013) (007.008,009)
1.0,
1.1
1.2
1.3
1.5
1.8
2.2
2.6
3.1
3.5
4.1
4-7
5.5
7.3
10.0
1.0
1.3
1.8
2.3
2.7
3.2
3.8
4.4
5.1
5.8
6.6
7.6
9.0
11.0
12.0
Electrostatic
Precipitator
(010,011,012)
1.0
1.0
1.1
1.2
1.3
1.6
2.0
2.5
2.9
3.4
3.9
4.4
5.4
7.5
11.0
Applied Eff. - 1. - (1. - Rated Eff. ) (Correction Factor)
-------
0.404, if device 022
0.000, all other devices
V is the output (i.e., new) ACFM
It should be noted that additional fuel is only used for
the catalytic and direct flame afterburners. The program
assumes that the additional fuel is natural gas.
(e) Disposal Quantity - Emission rate for pollutant in
question (tons/day) times new device efficiency (after
adjustment) for same pollutant times days of operation
per year.
This computation is performed for each of the pollutants.
(f) Labor Quantity = LH
Where H = operating time (hr/yr)
L = labor (labor hours/operation hours)
There are four, pre-set, labor values which relate
to the source size being controlled (small, medium,
large, and extra large). Source size is measured by V,
actual cubic feet of exhaust gas per minute (as corrected).
The program selects the appropriate value as follows:
Use the "small" value for V <_ 40,000
Use the "medium" value for 40,000 < V <_ 250,000
Use the "large" value for 250,000 < V <_ 1,000,000
Use the "extra large" value for V > 1,000,000
For any particular control device, only some of the terms in the above
equation will be non-zero. An electrostatic precipitator, for instance,
will not use liquid, chemicals, or additional fuel. The quantity JK how-
ever, will always be calculated.
Operating and maintenance costs for the remaining control devices
cannot be estimated by the above calculations. These devices are generally
more complex (involving sulfur oxide removal), and separate equations have
been developed to compute their O&M expenses. These equations are based
on manufacturers' specifications and operating experience reported in the
literature.
-------
For the three flue gas desulfurization measures considered, the fol-
lowing calculations of operating cost are used.
Device 039 - Catalytic Oxidation
O&M = 0.05'(total installed cost)
+ AB + JK
Where A, B, J, and K are as defined
above.
Device 041 - Dry Limestone
O&M = 365 PC (0.17 + 7.0 SC ) + 25,000
Device 042 - Wet Limestone
O&M = 365 P (0.196 + 3.82 SC ) +
130,000 ° °
where P = coal burned in ton/day (Source File)
S = coal sulfur content, decimal (Source File)
C = chemical cost in $/ton (input as device data)
The O&M costs associated with the use of sulfuric acid or sulfur
by-product manufacturing as a pollutant control measure (devices 043, 044,
and 045) are computed from the following equation:
O&M = JK - (Purchase Cost)/(Rated Life) + y
where, for each device:
J = disposal quantity, as calculated in item (e) above.
K = input disposal cost
Purchase Cost = value computed on page 60, item (f)
Rated Life = input as device data
y = value as calculated below. The variable P used in these
computations is the percentage concentration of sulfur
dioxide, by volume, in the exhaust gas as given on Page 60 ,
item (f).
-------
Devices 043,044-Sulfuric Acid Plant-Single Contact and
Double Contact Process
Iog10y =0.75 log1Qx + a
where y = annual cost (in 10^ $)
x = sulfur emission in tons/day
if 0 <_ P < 3.0, then a = 1.6903
3.0<.P<6.0, then a =» 1.4445
6.0 <_ P < 10.0, then a = 1.2076
P > 10.0, then a = 1.721
Device 045 - Sulfur Plant
Iog1()y - 0.81 Iog1()x + a
where y = annual cost (in 10 $)
x « sulfur emission (above)
if 0 £ P < 5.0, then a » -1.210
5.0 <_ P < 7.0, then a » -1.375
P >_ 7.0, then a = -1.435
The total annual device cost, output by the program, is then the
annual capital cost (defined under Annual Capital Charge) plus the O&M cost
and, if applicable, the gas cooling cost.
-------
APPENDIX C
EFFECTIVE STACK HEIGHT PARAMETER RANGES
Since the sensitivity of each of the effective stack height parameters
is dependent on the nominal set of parameters chosen, a realistic nominal
set is required to properly rate the parameter sensitivity. To determine
a set of nominal values, the detailed emission inventories for six pre-
viously inventoried regions were utilized (Chicago, Cincinnati, Denver,
Philadelphia, St. Louis, and Washington, D.C.). Although several of these
inventories were in the final checkout stage (and, therefore, subject to
a limited number of changes), no attempt was made to investigate apparently
spurious data. It is anticipated that the sample size was large enough
that a small number of data errors would not affect the parameter trends
established.
The inventory data were divided into the three categories: Power Plants,
Fuel Combustion (except Power Plants), and Industrial Process. For each
category the stack height (h), exit velocity (V), exit diameter (D), and
exit temperature (T) were plotted versus plant size (rated capacity for
fuel burning source and ACFM for industrial process sources). Although
no precise relationship could be established between the various stack
parameters and source size (as to be expected for such broad source
groups), the nominal values and ranges over which these parameters occurred
were, in general, defined. The user should note, however, that the nominal
values presented here should not be used in place of missing inventory data.
These results provide only the general parameter range and a nominal set
for comparison purposes. To illustrate the pitfall of using such nominal
values in an inventory,the techniques of the parameter sensitivity analysis
(Chapter 2.0) were used to relate the Power Plant parameter ranges to a
corresponding deviation in maximum annual average pollutant concentration
values.
Figures C-l through C-4 show the plots of Power Plant stack height,
exit velocity, exit diameter, and exit temperature versus rated capacity,
respectively. These data were taken from 102 powerplants located through-
out the six regions. In each figure, loci of parameter extremes are shown
-------
-I 0> CO o
RATED-CAPACITY (BTU/HR)
-------
-------
RATED CAPACITY CBTU/hr)
-------
oo to o
f 1000 • 10
RATED CAPACITY (BTU/hr)
Figure C-4. Exit Temperature Range for Power Plants
-------
which include 90 percent of the data points. Table C-l summarizes the
average parameter values and their ranges for small to medium (100*10 to
600*106BTU/hr), medium to large (600*10 to 2000*106 BTU/hr), and large
to extra large (2000*10 to 10,000*10 BTU/hr) power plants. Using these
data to determine x /X [Equations (3) and (4),Chapter 2.0], for each
parameter in each plant°size group, shows that significant pollution
concentration errors could be introduced if an average V or D value
were used in place of measured (or carefully estimated) data. The maxi-
mum pollutant concentration errors that could result from use of average
values for each of the stack parameters are shown in Table C-l as "Maximum
Concentration Error."
Figures C-5 and C-6 are plots of the power plant effective stack
height [calculated using Equation (1)] and ACFM values versus rated
capacity. The apparently narrow parameter ranges for these plots are
deceptive. Using the same analysis as above, the use of an average
effective stack height value from Figure C-5 could produce a maximum
pollutant concentration error of 500 percent. From the control cost
analysis of Chapter 3.0, the use of an average ACFM value from Figure C-6
could produce an average annual control cost error of around $100,000 for
each device.
Figures C-7 through C-ll show the Fuel Combustion plots of h, V,
D, T, and ACFM versus rated capacity, respectively. Figures C-12 through
C-15 show the industrial process plots of h, V, D, and T versus ACFM,
respectively. In each figure, loci of parameter extremes are shown.
Table C-2 contains a summary of the parameter ranges and nominal
values taken from the parameter plots. The nominal values presented
are, in general, median values.
-------
Table C-l. Average Parameter Range for Powerplants
Parameter Value
Small to
Parameter Medium Plant
Stack Height (meters)
Exit Velocity (m/sec)
Exit Diameter (meters)
Exit Temp ("Kelvin)
60
7
3-
480
+ 20
+ 5
+ 2
+ 60
Maximum
Concentration Medium
Error Large Plant
40%
>400%
>400%
30%
70
15
4
450
+
+
+
+
30
14
2
40
Maximum •
Concentration Large
Error X-Large
80%
>400%
>400%
20%
110
20
5
440
+
+
+
+
to
Plant
50
16
3
30
Maximum
Concentration
Error
100%
>400%
>400%
-------
RATED CAPACITY (BTU/hr)
6 ( 7 « t «•
Figure C-5. Effective Stack Height Range for Power Plants
-------
RATED CAPACITY (BTU/hr)
1000 • 10
I I 111 ill I I I I
8 t 10
Figure C-6. ACFM Range for Power Plants
-------
at **i oo to
RATED CAPACITY (BTU/hr) E
f 10,000-10
-------
in o» *»J oo co o
rRATED CAPACITY (BTU/HR)
-------
oo
RATED CAPACITY CBTU/hr)
-------
00 CO O
-•RATED CAPACITY (BTU/hr)
10,000 • 10
-------
'010,000-10
«ATED CAPACITY (BTU/hr)
3 4 5 6799 10
S 4 ft 6789 10
100-10
1000-10
10,000-10
Figure C-ll. ACFM Range for Fuel Combustion Source (Not Including Power Plants)
-------
O» ^J 00 CO o
CO
U>
t Lz-iir: MAXIMUM EXHAUST GAS VOLUME (ACFM) I
-------
MAXIMUM EXHAUST GAS VOLUME (ACFM) • i
en en ** GO to
ffii
"100'IQ-
:1000-10-
-------
00 CO O
r."::::.i;:-L;£ldi MAXIMUM EXHAUST GAS VOLUME (ACFM)
1000-10-
-------
MAXIMUM EXHAUST GAS VOLUME (ACFM)^ «•
^250
-------
Table C-2. Effective Stack Height Diameter Range
Plant Size
CO
—i
Rated Capacity (BTU/hr)
ACFM1
h (meters)
V (meters/sec.)
D (meters)
T (deg kelvin)
PP2
FC
IP
PP
FC
IP
PP
FC
IP
PP
FC
IP
Small
50E6
20E3
9-94(41)
5-70(18)
.1-15(5)
.5-21(6)
.5-3.5(2)
.5-4.5(2)
460-590(525)
280-925(500)
Medium
400E6
150E3
35-96(64)3
17-84(47)
8-95(35)
1-22(7)
.1-19(9.5)
.5-25(6)
.5-5.5(3.5)
.5-5 (2.5)
1-8 (4.5)
400-525(465)
425-545(485)
280 (425)
Large
1500E6
600E3
46-118(83)
26- 77(51)
10-130(65)
2- 34(12)
6- 21(13.5)
.5- 26(6)
2. -6. 5(4)
.5-6 (3)
1.5-10.5(6)
390-480(440)
400-520(460)
280 (375)
Extra-Large
6000E6
2000E3
76-162(122)
37- 70(55)
12-130(91)
4-46(22)
13-23(175)
.5-27(6)
3-8(5.5)
1-7(3.5)
2-12(7.5)
385-465(425)
380-490(435)
280-525(300)
Defines plant size.
2
PP = Power Plant, FC = non-Power Plant Fuel Combustion, IP = Industrial Process
------- |