x>EPA
           United States
           Environmental Protection
           Agency
           Office of Air Quality
           Planning and Standards
           Research Triangle Park NC 27711
EPA-45u/2-/8-007a-l
August 1978
           Air
Electric  Utility
Steam Generating
Units

Background
Information for
Proposed S02
Emission Standards

Supplement

-------
EPA-450/2-78-007a- I
Electric Utility Steam Generating Units
Background Information for Proposed SO 2
Emission Standards
Supplement
Emission Standards and Engineering Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Caro’ina 27711
August 1978

-------
This report is issued by the Environmental Protection Agency to report technical data
of interest to a limited number of readers. Copies are available - in limited quantities -
from the Library Services Office (MD-35), U S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or, for a fee, from the National Technical
Information Service, 5285 Port Royal Road, Springfield, Virginia 22161.
Publication No. EPA-450/2-78-007a-1
II

-------
                      Background Information Supplement
                   for Proposed S02 Emission Standards for
                   Electric Utility Steam Generating Units

                       Type of Action:  Administrative

                                 Prepared by:
Don R. Goodwin   /                                                (Date)
Director, Emission Standards and Engineering Division
Environmental Protection Agency
Research Triangle Park, North Carolina  27711
                                 Approved by:
Walter C. Barber                                                  (Date)
Director, Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, North Carolina  27711
Draft Statement Submitted to EPA's                           September,
Office of Federal Activities for Review on                        (Datel
This document may be reviewed at:

Central Docket Section
Room 2903B, Waterside Mall
401 M Street
Washington, 0. C.  20460

Additional copies may be obtained at:

U. S. Environmental Protection Agency Library (MD-35)
Research Triangle Park, North Carolina  27711

National Technical Information Service
5285 Port Royal Road
Springfield, Virginia  22161
                                    m

-------
TABLE OF CONTENTS
Pane
LIST OF ILLUSTRATIONS vii
LIST OF TABLES vii
1. INTRODUCTION 1-1
2. IMPACT ANALYSIS OF APRIL 1978 2—1
2.1 INTRODUCTION 2-1
2.2 INPUT ASSUMPTIONS USED BY COMPUTER MODEL 2-2
2.2.1 FGD Efficiency 2-2
2.2.2 Coal Sulfur Content Variability 2-2
2.2.3 Growth in Electric Demand 2—3
2.2.4 Cost Escalations 2—3
2.3 COST AND FINANCIAL IMPACTS 2-4
2.3.1. Cumulative Capital Expenditures 2—4
2.3.2 Annual Revenue Requirements 2—7
2.3.3 Average Consumer Rate 2—7
2.3.4 Average Residential Monthly Bill 2-8
2.3.5 Present Value of NSPS 2-8
2.3.6 Annualized Costs 2—10
2.3.7 Incremental Costs of Emission Reduction 2—11
2.3.8 Regional Annualized Costs 2—11
2.3.9 Annual Emissions of SO 2 2-12
iv

-------
Table of Contents (continued)
Page
2.4 IMPACT OF NSPS ALTERNATIVES ON FUEL PRODUCTION 2—12
2.4.1 Coal Production 2-12
2.4.2 Oil and Gas Consumption 2-12
2.5 EQUIPMENT CONSTRUCTION IMPACTS IN 1985 AND 1990 2-16
2.5.1 Generating Capacity 2-16
2.5.2 Coal Capacity With Scrubbers 2-16
2.6 MISCELLANEOUS IMPACTS 2-16
2.6.1 Percent of Flue Gas Scrubbed 2-16
2.6.2 Coal Plant Costs 2—18
3. IMPACT ANALYSIS OF AUGUST 1978 AND RELATED ANALYSES 3-1
3.1 INTRODUCTION 3-1
3.2 RESULTS OF THE AUGUST 1978 ANALYSES 3-4
3.2.1 National and Regional Utility SO 2 Emissions 3-4
3.2.2 Utility Oil and Gas Consumption 3-8
3.2.3 National Coal Production and Western Coal Shipments 3—12
3.2.4 Residential Bills, Capital Expenditures,
Present Values, and Annualized Costs 3—14
3.2.5 Sensitivity of Results to Assumed Oil Prices
and Rail Rate Escalation 3—16
3.3 OTHER ANALYSES 3—21
3.3.1 Joint DOI/DOE Study 3-21
3.3.2 DOE Proposal 3—24
3.3.3 UARG/NERA Analysis 3—26
4. DISCUSSION OF REGULATORY TOPICS 4-1
4.1 INTRODUCTION 4-1
4.2 FGD PERFORMANCE 4-1
V

-------
Table of Contents (continued)
Page
4.2.1 Overview 4-1
4.2.2 Prototype FGD Unit Performance 4-2
4.2.3 Full Scale FGD Systems 4-5
4.3.2.1 Vendor Guarantees and Other Statements 4-5
4.2.3.2 FGD Performance Test Results 4-6
4.2.4 Analysis of FGD Performance Range 4-il
4.2.5 Projected FGD Performance 4-13
4.2.5.1 Mean FGD Performance 4-13
4.2.5.2 Minimum FGD Performance 4-16
4.2.6 FGD Performance Control Improvements 4-19
4.2.7 FGD Control at Small Plants 4-21
4.3 DRY SO 2 CONTROL SYSTEMS 4-23
4.4 NONCONTINENTAL AREAS 4-25
4.5 PERFORMANCE TESTING 4-28
4.5.1 Particulate Matter 4-28
4.5.2 Sulfur Dioxide and Nitrogen Oxides Standards 4-29
4.5.2.1 Compliance Tests 4—29
4.5.2.2 Fuel Pretreatment Credits 4—31
4.5.3 Opacity 4-34
4.6 FGD COMPLIANCE 4-35
4.7 COAL IMPACTS 4-42
4.7.1 Production and Reserves 4-42
4.7.2 Anthracite Coal 4-46
vi

-------
LIST OF ILLUSTRATIONS
Figure Number Page
4-1 FGD 24-Hour Average Efficiency Distribution 4-14
4-2 Projected RSD for Small Plants 4-22
LIST OF TABLES
Table Number
2-1 National Costs in 1990 of Alternative New Source
Performance Standards. 2-5
2-2 Regional Costs and Residential Monthly Bills
Under Alternative New Source Performance Standards. 2-9
2-3a Regional Sulfur Dioxide Emissions in 1990 Under
Alternative New Source Performance Standards. 2-13
2—3b Sulfur Dioxide Emissions in 1990 Under Alternative
New Source Performance Standards. 2-14
2-4 Coal and Oil Consumption in 1990 Under Alternative
New Source Performance Standards. 2-15
2—5 Physical Description of Electric Utility Industry
in 1990 Under Alternative New Source Performance
Standards. 2-17
2-6 Physical Description of Electric Utility Industry
in 1985 Under Alternative New Source Performance
Standards. 2-19
3—1 Comparison of Assumptions, April 1978 and August
1978. 3-2
3—2 National and Regional Utility SO 2 Emissions. 3—5
3-2a Regional Utility SO 9 Emissions in 1990 and 1995 for
Plants Subject to the Revised NSPS. 3-9
vii

-------
List of Tables (continued)
Table Number Page
3—2b Regional Utility Coal Consumption in 1990 and 1995
for Plants Subject to the Revised NSPS. 3-10
3—3 Utility Oil/Gas Consumption. 3-li
3—4 National Coal Production and Western Coal Shipped
East. 3-13
3—5 Residential Bills, Capital Expenditures, Present
Values, and Annualized Costs. 3-15
3—6 A Comparison of Alternative Oil Price and Rail Rate 3-18
Assumptions Upon Results for the 0.2 (With Exemptions) and
Option in 1990. 3-19
3-7 1990 Regional Coal Production Under Different
Sensitivity Assumptions. 3-23
3-8 Comparison of EPA and DOE Proposals. 3-25
3-9 Description of Cases. 3-27
3- 10 Summary of Impacts as Analyzed by NERA for UARG. 3-28
4—1 List of U. S. Flue Gas Desulfurization (FGD)
Scrubbers Installed, Under Construction, or Awarded -
Designed for 85 Percent or Greater SO 2 Removal 4-7
U. S. Plants Reporting 90 Percent or Greater SO 2
4-2 Removal. 4-8
4-3 American Designed FGD Systems Operating in Japan. 4-9
4-4 SO 2 Removal Statistics. 4-18
4-5 Relationship Between Fuel Pretreatment and Post
Combustion Control for Removal of 85 Percent Sulfur
Dioxide. 4-32
4-6 Effects of Anthracite Use on Cost of Electricity. 4-47
viii

-------
1. INTRODUCTION
In February of 1978 EPA completed a preliminary analysis of options
for revising new source performance standards (NSPS) for new steam
electric generating units in the utility industry. The February analysis
is contained in the document Electric Utility Steam Generating Units:
Background Information for Proposed SO 2 Emission Standards (report
number EPA—450/2-78—007a). This supplement to the initial February
document provides results of additional impact analyses (sections 2 and 3)
and contains the results of new areas of investigation (section 4) used
in developing the proposed standards.
Sections 2 and 3 contain analyses of projected impacts of alternative
NSPS that include consumer and utility costs, national and regional
emissions, national coal and oil consumption, and new electric generation
construction. Impacts discussed in section 2 were developed in April
1978. After review of these data, it was apparent that many of the
impacts projected could be altered appreciably by the basic assumptions
used. A review of these assumptions indicated the need for changes based
upon additional or more up-to-date information. For example, future
electric demand was revised downward to reflect energy conservation
measures that have slowed recent growth in national electric demand.
These revised assumptions were then applied in August 1978 to develop

-------
the impacts given in section 3. A description of the assumptions used
in each analysis is given in introductions to the respective sections.
The impact analyses in sections 2 and 3 were projected from ICF
Incorporated’s coal and electric utilities model. These analyses also
contain impacts developed by Temple, Barker, and Sloane, Inc. (TBS)
using ICF data. The ICF model simulates the decision making process of
utility companies based upon the most economical choice of alternatives.
The number, size, and types (coal-fired boilers, oil-fired turbines,
etc.) of new power plant generating equipment to be constructed is
projected from electric demand growth, fuel availability and costs,
emission control costs, and electric generating equipment costs. Once
a physical description of the electric utility industry is projected,
the model determines the incremental operating costs of each electric
generating unit. Based upon the relative costs of operation, a load
dispatching order and unit load factors are projected. The model then
computes national and regional impacts on utility capital and annualized
costs, emissions, coal and oil consumption, etc. Because the model
projects utility decisions solely on the most favorable economic choice,
it does not take into account other factors that can influence decisions,
especially when the projected costs of two alternatives are very close.
Section 4 contains an assessment of EGO performance capabilities
and projected EGO performance in new affected units. Performance of dry
control systems is also included. Performance test requirements and
limitations on bypassing during FGD malfunctions are discussed. Other
topics addressed are application of the proposed standards to noncontinental
areas and effects of alternative standards on the utilization of high-
sulfur and anthracite coals.
1-2

-------
2. IMPACT ANALYSIS OF APRIL 1978
2.1 INTRODUCTION
In April 1978 an analysis of alternatives for preparing revised NSPS
for power plants was completed by EPA consultants Temple, Barker, and
Sloane Inc. Data supplied on April 17, 1978, from the CF, Inc. coal and
electric utilities model were used in the analysis. Some adjustments to
the SO 2 emission data were made at a later date after a computational
error was discovered. The corrected data are presented in this section.
The data presented are analyses of actual 1975 conditions, the impact in
1990 of the current NSPS, and future impacts of five alternatives for
revised NSPS.
The “floors 1 ’ analyzed are maximum control levels (24-hour average),
which, when attained, require no additonal percentage of SO 2 removel by
an FGD control system. A 0.2 lb/million Btu floor would require virtually
all coal to be subjected to full FGD control. For the purpose of this
analysis, full FGD control is 85% reduction(24-hour basis) except for 3
days/month during which 75 percent reduction is allowed. Higher floors
would permit some low-sulfur coals, principally Western coals, to be
burned without full FGD control (partial scrubbing). A portion of the
flue gas could be bypassed around the FGD system for reheating the stack
gas and improving plume buoyancy with the higher floors (0.5 and 0.8 lb/
million Btu). Bypassing of flue gas also reduces the energy penalty
associated with wet scrubbers since little or no reheat is required.

-------
The “ceilings” analyzed (also expressed in lb/million Btu) are maximum
emission limits (24-hour average) that cannot be exceeded except when
exemptions are considered. The alternatives with exemptions would allow
the ceiling to be exceeded for 3 days in a month, but would not allow
less FGD control than is required by the percent SO 2 removal standard.
The five alternatives considered are:
1. 0.2 floor, 1.2 ceiling, with exemptions.
2. 0.2 floor, 1.2 ceiling, without exemptions.
3. 0.5 floor, 1.2 ceiling, with exemptions.
4. 0.5 floor, 1.2 ceiling, without exemptions.
5. 0.8 floor, 1.2 ceiling, with exemptions.
2.2 INPUT ASSUMPTIONS USED BY COMPUTER MODEL
2.2.1 FGD Efficiency
In the analysis of alternatives, scrubbers were assumed to be 90 percent
efficient on a 30-day average basis when full FGD control was required to
attain the floor. In analysis of 0.2 floor alternatives, it was assumed
that full FGD control would be required. In analysis of 0.5 and 0.8
floors, it was assumed that less SO 2 percent reductions (partial scrubbing)
would be permitted when certain very low sulfur coals were burned and
that average SO 2 emissions would be equal to the floor each 24-hour period.
2.2.2 Coal Sulfur Content Variability
The variation in daily potential SO 2 emissions that must be con-
trolled by FGD was projected by a statistical distribution of 24-hour
averages based upon a relative standard deviation (RSD) of 15 percent.
Variation in the average sulfur content of coal consumed in a 24-hour
period by electric utility boilers is reviewed in section 4.2.7, where
RSD is discussed as a function of coal type and plant size.
2-2

-------
When no exemption in the ceiling was allowed, the assumption was that
the utilities would purchase coal that would be in compliance with the
ceiling when (1) the FGD control efficiency dropped to 75 percent and
(2) the projected 24-hour average coal sulfur content is highest (i.e.,
3 standard deviations above the long-run mean sulfur content). When an
exemption of the ceiling was allowed, it was assumed that the utilities
would purchase coal that would be in compliance with the ceiling at 85
percent efficiency (i.e., the ceiling constraint falls 2 standard deviations
above the long-run mean sulfur content because the higher SO 2 emissions
that would be projected by 3 standard deviations would peak through the
exemption in the ceiling).
2.2.3 Growth in Electric Demand
Growth in the nation’s electricity consumption was projected to be
5.8 percent per year through 1985 and 5.5 percent per year thereafter.
This is a relatively high growth assumption and therefore results in a
high estimate of the number of new plants affected by the revised standard.
Construction of nuclear units to satisfy this demand was projected to
increase national nuclear capacity to 108 gigawatts (GW) in 1985, 177 GW
in 1990, and 302 GW in 1995. Other increases in capacity are expected
from new coal or oil-fired electric-generating equipment and other
sources.
2.2.4 Cost Escalations
A general economic inflation rate of 5.5 percent per year was assumed;
however, costs in accompanying tables have been discounted to 1975 dollar
values. Present (1975) costs for crude oil ($13 per barrel), coal
transportation, and coal mining labor were not increased beyond the
general inflation rate. The oil price assumption has a major effect on
2—3

-------
results since it determines the mix of generation capacity which is used.
A low oil price results in more oil consumption and less new coal plant
construction as well as less coal consumption than would result if a high
oil price is assumed.
2.3 COST AND FINANCIAL IMPACTS
2.3.1 Cumulative Capital Expenditures
The cumulative capital expenditures given on Table 2-1 consist of
the sum of all capital expenditures anticipated from 1976 through 1990.
These expenditures include the amounts to be spent on the following
types of capital equipment:
• Power plants
• Transmission lines
• Distribution systems
Scrubbers
• Electrostatic precipitators
• Baghouses.
The capital expenditures include cash expenditures and allowances for
funds used during construction (AFDC). While AFOC is not a cash
expenditure, it has been included because it is capitalized and becomes
part of the consumer rate base. Excluded are the amounts spent on
pollution control equipment for pollutants other than SO 2 after 1976.
Some of the capital expenditures in the years 1985-1990 are for equipment
that will not be in service by 1990.
The differences in capital expenditures among alternative NSPS
reflect:
1. Number of scrubbers, precipitators, and baghouses constructed.
2-4

-------
Table 2-1. NATIONAL COSTS IN 1990 OF
ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
0.2 floor
0.5 floor
0.8 floor
1.2 ceiling
1.2 ceiling
1.2 ceiling
With Without
Exemptions Exemption
With Without
Exemptions Exemptions
Financial parameters
Cumulative capital
Expenditures:
1976-1990, $billions
Annual revenue
requirement, $billions
Average consumer rate,
mill s/kwh
Average monthly
residential bill*
Apartment
Rome without
electric heat
Home with
electric heat
Incremental cost of SO
reduction,*** $/ton 2
46.9
142.5
27.0
35.93
11.49
14.13
21.00
34.78
62.60
93.46
21.87
45.31
3.8
4.3
0.99
1.09
0.34
0.37
0.83
0.92
2.24
2.49
1.08
1.21
*Average monthly bills based on the following use in 1990: apartments 325 kwh/month; homes without electric heating,
800 kwh/month; homes with electric heating, 2,150 kwh/month; national average, 1,042 kwh/month.
Includes capital expenditures and operating costs through 1990 and operating costs from 1990 to 2020 for all SO 2
control equipment and capacity mix changes, discounted at a real rate of 5 percent.
***Incremental costs determined by dividing the incremental annualized cost by the reduction in SO 2 emissions,
using the current NSPS as a base.
(Source: Annualized costs and incremental costs per ton were prepared by ICF. All other figures prepared by lBS.
April 17, 1978.).
Alternative NSPS, 90 percent SO 2 removal (increase from current NSPS)
Actual Current
1975 NSPS
- 746.1
With
Exemptions
0 ,
National average
Economic parameters
Present value of NSPS,**
$billions
Annualized costs,
$bill ions
14.7
3.0
0.76
0.28
0.68
1 .85
0.89
15.6
1.3
l9
12.4
3.0
0.77
0.27
0.67
1 .83
0.88
18.1
1.5
750
2.9
0.5
0.15
0.05
0.12
0.34
0.16
0.7
0.3
300
- 35.5
— 95.7
0
26.1
2.0
833
31 .5
2.2
758

-------
2. Type and capacity of new power plant constructed. As NSPS
become more stringent, a greater percentage of new plant additions will
utilize oil-fired turbines, which cost less to install than coal-fired
plants.
3. Type of coal burned in new coal—fired power plants (coal plant
design and construction costs vary with the type of coal).
In the 0.2 floor, 1.2 ceiling cases, cumulative capital expenditures
are similar. Although the 0.2 floor, 1.2 ceiling without exemptions
case has a large amount of less expensive, oil—fired turbine capacity;
turbine capacity is only slightly greater than that in the 0.2 floor,
1.2 ceiling with—exemptions case. The relative savings is offset by
lower construction costs for coal-burning plants in the case with
exemptions.
In the 0.5 floor, 1.2 ceiling cases, cumulative capital expenditures
differ. The 0.5 floor, 1.2 ceiling-without-exemptions case will have 4.9 GW
more turbine capacity in 1990 than the case with exemptions. This saving,
along with the saving attributable to a significantly lower percentage of
flue gas scrubbed, creates a relatively lower cumulative capital expendi—
ture in the without—exemptions case.
The 0.5 floor case results in higher total capital costs than the
0.2 floor case. This is due to the change in capacity mix under the two
alternatives. The 0.2 floor requires more expensive pollution control
equipment which results in utility decisions to delay construction of
some new coal-fired capacity and increase their oil consumption, thus
lowering their total capital costs. The 0.5 floor does not require as
much pollution control capital which reduces the cost of coal generation
2-6

-------
relative to oil-fired generation and results in more new coal capacity
being constructed. This increase in new plant construction offsets the
savings in capital associated with lower pollution control capital
expenses. Therefore, capital expenditures are higher under the 0.5
floor case. However, if total expenses including fuel and operating and
maintenance costs are examined, the 0.5 floor case is less expensive
than the 0.2 floor case.
2.3.2 Annual Revenue Requirements
The annual revenue requirements in Table 2-1 represent funds necessary
to cover all operating costs (fuel, operating expenses, and maintenance
expenses, capital costs AFDC, return on equity, and depreciation), and
taxes. The 1990 revenue requirements of the alternatives considered are
different because of changes in capital-related charges and total fuel
charges. These factors vary with the relative amount of energy generated
by coal-fired plants and oil-fired turbines. The generation mix as well
as total fuel charges are affected by fuel prices.
The annual revenue requirements can be expected to be different for
other years. This is due to the increasing total electric generating
construction requirement and the timing of the capital expenditures
which produces a shift between capital and 0&M expenses.
2.3.3 Average Consumer Rate
The average consumer electric rate is determined by dividing the
annual revenue requirements by the number of kilowatt-hours sold.
Under all alternative NSPS, the amounts of electricity sold are equal,
so that the percentage change in average consumer rate between alternative
NSPS is the same as the percentage change in required revenues.
2—7

-------
2.3.4 Average Residential Monthly Bill
The average residential monthly bill has been determined by estimating
the difference between the average electric rate and the historical
average residential electric rate and applying this rate to the estimated
electrical consumption. In 1976, residential consumers paid about 50
percent more for electricity than the average cost of each kilowatt-hour
produced. The differential in 1990 was estimated by assuming that
increased fuel and operating costs would be apportioned equally on a
per-kilowatt-hour basis to all customer classes and increased capital
costs would be apportioned at the same rate as they are currently, with
residential customers paying more than industrial customers. The figures
for apartments and homes with and without electric heat reflect the
different levels of usage by these types of residence. The figures reflect
an increase in electrical consumption of about 50 percent between 1975 and
1990.
The 1990 regional average monthly residential bills in Table 2-2 were
estimated by TBS using the national average monthly residential bills
reported in Table 2-1 and the regional annualized costs provided by ICF.
The regional annualized costs and electric generation figures were used
to calculate regional costs per kilowatt-hour. These costs were then
adjusted to represent regional residential costs. The ratio of each of
the regional residential kilowatt-hour costs to the national average
residential costs was used to adjust the national average monthly
residential bill.
2.3.5 Present Value of NSPS
The present values presented in Table 2-1 represent the real resource
cost of the sum of the capital expenditures for F1SPS pollution control
2-8

-------
Table 2—2. REGIONAL COSTS AND RESIDENTIAL MONTHLY BILLS
UNDER ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
Alternative NSPS, 90% SO 2 removal
Actual
1975
Current
NSPS
(chançie from current NSPS)
0.2 floor
0.5 floor
0.8 floor
1.2 ceiling
1.2 ceiling
1.2 ceiling
With
Exemptions
Exemptions
With
Exemptions
r; .)
Without
Exemptions
With
Exemptions
Annualized costs,
$bill ions
East
Midwest
West South Central
West
Total
Average residential
monthly bill, $
East
Midwest
West-South Central
West
National average
43.8
24.9
13.9
13.0
95.7
47.07
44.1 3
53.92
36.20
45.31
1.0
0.6
0.4
0.0
2.0
1 .22
1 .24
1 .75
0.12
1 .08
1.3
0.6
0.4
0.0
2.3
1 .54
1.23
1 .72
0.09
1 .21
0.7
0.4
0.3
-0.1
1.3
1 .02
1 .01
1 . 52
-0.07
0.89
21 .87
0.9
0.4
0.3
-0.1
1.5
1.15
0.91
1 .37
-0.12
0.88
0.1
0.0
0.2
0.0
0.3
0.12
0.03
0.83
0.04
0.16
(Source: Annualized Costs, ICE; Residential Bill,
IBS, April 17, 1978.).

-------
equipment and costs (or savings) resulting from generating-capacity
changes from 1976 to 1990. The present values include all capital and
operating expenses incurred over the 1976-1990 period plus fuel and
operation and maintenance expenditures for all NSPS equipment and
capacity mix changes. The costs of pollution control equipment consist
of capital expenditures for and operation and maintenance of scrubbers,
baghouses, and precipitators; and the capital expenditures for and
operating costs of additional generating capacity needed to make up for
the energy used by the pollution control equipment. Changes in generating
capacity mix can result in costs or savings in capital expenditures for
generating plants, as well as variations in coal price, and changes in
operating costs. The costs of capacity mix changes were computed as
changes from a base case: the current NSPS (l.2-lb/million-Btu SO 2
ceiling, 0.1 lb/million Btu particulate matter ceiling). In order to
calculate the “real reasource costs,” financial flows and transfer
payments were excluded since they do not measure resources allocated but
are really transfer payments. Among the accounting charges and cash
flows specifically omitted were: Allowance for funds used during
construction (AFDC), depreciation, interest, return on equity, and
taxes.
2.3.6 Annualized Costs
The 1990 annualized costs were calculated by ICF’s model, which
included fuel costs and operation and maintenance expenses for all capacity
in use, and a capital cost component for new capacity brought on—line
after 1975 and in use by 1990. The capital cost component does not include
2-10

-------
any charges for capacity installed earlier. Capital expenditures were
computed using a real fixed charge rate of 0.1, which includes depreciation,
dividends, interest, and taxes. Capital expenditures for electricity
distribution systems are not included; however, these should not change
between the alternative cases examined.
2.3.7 Incremental Cost of Emission Reduction
The incremental cost of emission reduction was calculated by
dividing the incremental annualized cost of each alternative NSPS by the
emission reduction in SO 2 emissions. The incremental annualized costs
and emissions reductions are the difference between the annualized cost
and emissions of each alternative NSPS and the current NSPS.
2.3.8 Regional Annualized Costs
The annualized costs of alternative NSPS in 1990 are given on a
regional basis in Table 2—2. They were calculated in the same manner as
described for the national annualized costs in Table 2-1.
The four regions selected were groupings of Bureau of Census regions:
1. East-—New England, Middle Atlantic, South Atlantic, and East
South Central
2. Midwest--West North Central and East North Central
3. West South Central--West South Central
4. West--Pacific and Mountain
Although most regional costs increase under the alternative NSPS,
annualized costs in the western regional decrease with a 0.5 floor. A
change to the 0.5 floor alternative would decrease demand for western
coal, and thereby drive down coal prices in the region. This coal price
saving is not offset by increased pollution control costs since State
regulations in the region are already more stringent than the current
NSPS.
2—11

-------
2.3.9 Annual Emissions of SO 2
The amounts of SO 2 emitted in 1990 by coal-fired generating plants,
are shown in Table 2-3a on a regional basis for each alternative NSPS.
Table 2-3a shows a reduction in national SO 2 emissions for all alternatives
considered relative to retaining the current NSPS.
Table 2-3b further details 1990 emissions by type of emitting plant.
The model predicts that as the costs of building and operating new plants
increases due to air pollution control expenses, construction of planned
new facilities will be delayed. Hence, the predicted increase in new plant
emissions (Table 2-3b) under the partial scrubbing options is due not only
to the higher emission rates, but also to the increased new capacity pre-
dicted under the less stringent standards. Similarly, the model predicts
that emissions from SIP/NSPS plants will increase under the revised stan-
dards as existing plants are utilized more in order to compensate for the
delays in bringing new capacity on line.
2.4 IMPACT OF NSPS ALTERNATIVES ON FUEL PRODUCTION
2.4.1 Coal Production
The effects of alternative NSPS on total and regional coal production
in 1990 are shown in Table 2-4. The amount of Western coal shipped
east, a large component of total Western coal production, is reported
separately. This figure is affected significantly by the assumptions
regarding rail rate increases. This analysis assumes no real price
increase and therefore results in a high estimate of the amount of Western
coal shipped East.
2.4.2 Oil and Gas Consumption
Oil and gas consumed in 1990 by electric utilities, in quads (1 x io15
Btu) is reported as the increment to oil and gas consumed in 1990 by
2—12

-------
Table 2-3a. REGIONAL SULFUR DIOXIDE EMISSIONS IN 1990
UNDER ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
Alternative NSPS, 90% SO 2 removal (thanqe from current NSPS)
0.2 floor
0.5 floor
0.8 floor
1.2 ceiling
1.2 ceiling
1.2 ceiling
Actual
1975
Present
NSPS
With
Exemptions
Without
Exemptions
With
Exemptions
Without
Exemptions
With
Exemptions
Annual emissions of
SO 2 (million tons)
East
9.1
10.8
—1.1
-1.4
—1.2
—1.1
—0.6
Midwest
8.8
8.7
-O.
-0.4
-0.3
-0.3
-0.1
West South Central
0.2
2.6
-0.8
-0.8
—0.6
-0.6
—0.3
West
0.5
1.3
—0.2
-0.2
0.1
0.1
0.0
Total
18.6
23.3
-2.2
-2.7
-2.0
-1.9
-1.0

-------
Table 2—3b. SULFUR DIOXIDE EMISSIONS IN 1990
UNDER ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
Alternative NSPS, 90% SO 2 removal (total emissions)
0.2 floor
0.5 floor
0.8 floor
1.2 ceiling
1,2 ceiling
1.2 ceiling
Actual Present With Without With Without With
1975 NSPS Exemptions Exemptions Exemptions Exemptions Exemptions
Annual emissions of
SO 2 (million tons)
SIP/NSPS Plantsa 16.8 17.2 16.9 16.7
New Plantsb 4.2 1.5 2.1 3.3
Oil/Gas Plants 2.3 2.5 2.3 2.3
TOTAL 18.6 23.3 21.1 21.3 22.3
Total Coal Capacity (GW) 205 465 444 460 460
aplants subject to existing state regulations or the current NSPS of 1.2 #S0 2 /MMBtu.
bPlants subject to the revised NSPS.

-------
Table 2-4. COAL AND OIL CONSUMPTION IN 1990
UNDER ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
Alternative NSPS, 90% SO 2 Removal (change from current NSPS)
0.2 floor
0.5 floor
0.8 floor
1.2 ceiling
1.2 ceiling
1.2 ceiling
Actual Current With Without With Without With
1975 NSPS Exemptions Exemptions Exemptions Exemptions Exemptions
Total Coal Production,
million tons 647.4 1766.8 -55.7 -49.2 -12.3 -1.0 +12.7
Regional coal
production
East 396.3 441.3 +25.3 +47.5 +22.6 ÷31.9 -23.5
Midwest 151.1 297.9 +76.9 + 8.0 +54.6 - 0.7 + 8.9
West 100.0 1027.5 —157.8 -104.6 —89.4 -32.1 +27.3
Western coal shipped
East 20.8 455.4 —156.2 -126.6 -109.7 -58.7 -26.3
Oil and gas con—
suwtption (quads) 6.5 6.3 +0.8 +0.9 +0.2 +0.2 +0.2

-------
electric utilities under the present NSPS. As noted earlier, these figures
are very sensitive to the oil price assumption which is used.
2.5 EQUIPMENT CONSTRUCTION IMPACTS IN 1985 AND 1990
2.5.1 Generating Capacity
Generating capacity in 1985 and 1990 is reported by type for each
alternative NSPS (Table 2-5.and 2-6). The largest differences between
cases occur in coal and turbine capacities since under more stringent
regulations companies defer coal plant additions, choosing instead to
add less expensive turbine capacity. Note that differences in new coal
capacity across the various cases. Nuclear capacity in a given year was
assumed to be about the same in all cases.
Coal capacity under alternative NSPS is reported in three categories,
each affected differently by NSPS: (1) Existing capacity in 1975 (not
subject to any NSPS), (2) new plants added from 1976 to 1982 (subject
to the current NSPS in all cases), and (3) new plants added from 1983
to 1985 or 1990 (subject to the current or alternative NSPS examined).
2.5.2 Coal Capacity With Scrubbers
Coal capacity with scrubbers under alternative NSPS is the sum of
existing capacity in 1975, new plants added from 1976 to 1982, and
new plants added from 1983 to 1985 or 1990. The amount of capacity
equipped with scrubbers through 1982 is similar in all cases. Differences
between cases arise because of coverage differences in new plants caused
by changes in the on—line dates for new plants and plants under construction
due to changes in the NSPS case examined.
2.6 MISCELLANEOUS IMPACTS
2.6.1 PercerTt of Flue Gas Scrubbed
The percent of flue gas scrubbed (Tables 2-5 and 2-6) are computed as
a weighted average of the percent of flue gas scrubbed by plants equipped
2-16

-------
Table 2—5. PHYSICAL DESCRIPTION OF ELECTRIC UTILITY INDUSTRY
IN 1990 UNDER ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
Alternative NSPS, 90% SO 2 Removal
0.2 Floor
0.5 Floor
0.8 Floor
1.2 Ceiling
1.2 Ceiling
1.2 Ceiling
Actual Current With Without With Without With
1975 NSPS Exemptions Exemptions Exemptions Exemptions Exemptions
Coal capacity,* GW
Existing 1975 204.6 204.6 204.6 204.6 204.6 204.6 204.6
New plants, 1976-1982 -- 88.4 88.8 88.8 87.7 88.8 87.3
New plants, 1983—1990 —- 171.7 151.0 148.3 166.2 164.3 167.3
Total coal, 1990 204.6 464.7 444.4 441.7 459.5 457.6 459.2
Oil/gas capacity, GW
Steam 129.0 143.5 143.5 145.6 145.6 143.2 143.5
Combined cycle 2.7 15.9 15.3 15.3 9.9 11.7 15.9
Turbine 42.1 142.5 165.2 166.8 153.4 156.5 148.5
Total oil/gas, 1990 173.8 301.9 324.0 327.7 308.9 311.4 307.9
Nuclear capacity, 1990, GW 38.3 176.7 176.7 176.7 176.7 176.7 176.7
Hydroelectric and other capacity, 1990, GW 68.3 87.7 86.2 85.2 86.8 85.5 86.8
Coal capacity with scrubbers, 1990, GW 2.5 97.4 204.9 204.9 220.8 219.4 110.5
Flue gas scrubber,** percent 81.6 95.4 96.1 84.6 78.7 87.0
Coal plant cost,*** $ 570.5 557.6 568.0 566.2 572.2 581.0
_________ __ __ ___ ___ I ’ ___ ___ _
* Existing plants are not subject to NSPS. New plants, 1976—1982, are subject to the current NSPS in all cases.
New plants, 1983-1990, are subject to the specific NSPS selected.
** Percentage of flue gas scrubbed is a weighted average of flue gas scrubbed in each coal capacity category. The percentage
of flue gas scrubbed in plants built prior to 1983 is similar in all cases; differences are due to dissimilarities in new
plants, 1983—1990.
Coal construction costs vary with the type of coal plants are designed to burn. Coal piant cost in 1990 is a weighted
average based on the projected mix of bituminous, subbitunionous, and lignite burning coal capacity.

-------
with scrubbers. As with coal capacity with scrubbers, the percent of
flue gas scrubbed through 1982 is similar in all cases. Differences
between cases again arise from differences in new plants installed after
1982. With the 0.2 floor, flue gas is 100 percent scrubbed at new plants
installed after 1982, whereas in other cases the percentage scrubbed
ranges from 89.9 to 76.3 percent.
2.6.2 Coal Plant Costs
Coal plant construction costs vary with the type of coal the plants
are designed to burn--bituminous, subbituminous, or lignite. Coal plant
costs in 1990 were computed as a weighted average based on the mix of
coal types under alternative NSPS. The costs, including AFDC, are
reported in 1975 dollars.
2-18

-------
Table 2-6. PHYSICAL DESCRIPTION OF ELECTRIC UTILITY INDUSTRY
IN 1985 UNDER ALTERNATIVE NEW SOURCE PERFORMANCE STANDARDS
Alternative NSPS, 90% SO 2 removal
0.2 floor
0.5 floor
0.8 floor
1.2 ceiling
1.2 ceiling
1.2 ceiling
Actual Current With Without With Without With
1975 NSPS exemption exemption exemption exemption exemption
Coal capacity*, GW
Existing 1975 204.6 204.6 204.6 204.6 204.6 204.6 204.6
New plants, 1976-1982 - 84.0 84.8 84.9 85.0 85.0 83.8
New plants, 1983—1985 — 31.6 24.1 19.2 25.5 22.8 29.9
Total coal, 1985 204.6 320.2 313.5 308.7 315.1 312.4 318.3
Oil/Gas Capacity, GW
r Steam 129.0 145.6 145.6 145.6 145.6 145.6 145.6
Combined cycle 2.7 12.2 11.6 11.6 11.6 11.6 12.2
Turbine 42.1 111.2 123.1 127.9 122.5 125.2 118.8
Total oil/gas, 1985 173.8 269.0 280.3 286.1 279.7 202.4 276.6
Nuclear capacity, 1985, GW 38.3 108.3 109.2 109.2 108.3 108.3 108.3
Hydro electric and other capacity, 1985, GW 68.3 88.2 88.2 88.2 88.2 88.2 88.2
Coal capacity with scrubbers, 1985, GW 2.5 67.5 76.6 74.7 77.6 76.7 69.8
Flue gas scrubbed,** percent 84.8 89.2 88.4 86.4 84.6 87.6
Coal plant cost 560.0 562.8 566.2 562.8 565.4 564.5
*Existing plants are not subject to NSPS. New Plants, 1976 - 1982, are subject to the current NSPS in all cases. New Plants, 1983-
1985, are subject to the specific NSPS selected.
**percentage of flue gas scrubbed is a weighted average of flue gas scrubbed in each coal capacity category. The percentage of
flue gas scrubbed in plants built prior to 1983 is similar in all cases; differences are due ‘to dissimilarities in new plants,
1983—1 985.
***Coal consturction costs vary with the type of coal plants are designed to burn. Coal plant cost in 1985 is a weighted
average based on the projected mix of bituminous, sub-bituminous and lignite burning coal capacity.

-------
3. IMPACT ANALYSIS OF AUGUST 1978 AND RELATED ANALYSES
3.1 INTRODUCTION
The August 1978 analysis described in this section differs from the
April 1978 analysis in section 2.0 in that a number of assumptions were
changed to provide a more accurate assessment of the impact of alternative
control options. These assumptions were changed as a result of internal
EPA discussions as well as discussions between EPA and other government
agencies such as the Department of Energy. The major changes are described
below and summarized in Table 3—1.
1. Growth rates - Growth in demand for electricity was estimated
to be 5.8% per year (1975—1985) and 5.5% per year (1985-1995) for the
April analysis and 4.8% per year (1975-1985) and 4.0% per year (1985-1995)
for the August analysis.
2. Nuclear capacity - Installed nuclear capacity in 1985, 1990, and
1995 was assumed to be 108, 177, and 302 GW, respectively, in the April
analysis and 97, 167, and 230 GW, respectively, in the August analysis.
3. 011 prices - Prices for residual oil were :assumed to be constant
at $13/barrel (1975 dollars) for the April analysis. The August analysis
used oil prices developed by the Department of Energy. These prices, in
1975 dollars, are $15/barrel in 1985, $20/barrel in 1990, and $28/barrel
in 1995. Further evaluation of these values has been conducted and downward
revisions resulted. A sensitivity analysis that uses oil prices lower

-------
Table 3-1. COMPARISON OF ASSUMPTIONS
April 1978 and August 1978
Assumption April August
Growth rates 1975-1985: 5.8%/yr 1975-1985: 4.8%/yr
1985-1995: 5.5% 1985-1995: 4.0%
Nuclear capacity 1985: 108 GW 1985: 97 GW
1990: 177 1990: 167
1995: 302 1995: 230
Oil prices (51975) 1985: $13/bbl 1985: $15/bbl
1990: $13 1990: $20
1995: $13 1995: $28
General inflation rate 5.5%/yr 5.5%/yr
Annual emissions @ 0.5 floor 1 0.5 lb S0 2 /nhillion Btu 0.32 lb S0 2 /millior, Btu
Coal transportation Increases at general Increases at general inflation
inflation rate rate plus 1%
Coal mining labor costs Increases at general Increases at general inflation
inflation rate rate plus 1%
Miscellaneous A number of miscellaneous changes were made between the A ri1 1978
study and the August 1978 study. These changes were either correc-
tions or refinements of values used in the April study. Examples
of these changes included revisions to the level of SIP control
assumed in the model, revisions to the scrubbing costs, changes in the
assumptions regarding industrial coal consumption, and changes to the
coal supply curves used in the April study.
1 See text for clarifying discussion

-------
than those quoted above was performed and these results are also presented
in this section.
4. Annual emissions for partial scrubbing - In the April analysis,
plants which could comply with the revised NSPS by partial scrubbing
were assumed to meet the emission floor (expressed in lbs S0 2 /million
Btu) on an annual average basis. The proposed NSPS, however, would not
allow that the floor be exceeded for any 24-hour period. To meet this
requirement, annual emissions would have to be less than the level of
the floor because of the variability of the sulfur content of coal.
Hence, in the August runs, the annual emission factors for partial
scrubbing were adjusted downward (e.g., annual emissions when meeting
a 0.5 floor were assumed to be 0.32 lb S0 2 /million Btu). This assumption
is also being questioned and more analysis of this issue will be made
before promulgation of the revised NSPS.
5. Coal transportation costs — In the April analysis it was assumed
that coal transportation costs would not increase at a rate different than
the assumed general inflation rate of 5.5% per year. In the August analysis,
however, it was assumed that coal transportation costs would increase
at a rate of 6.5% per year or 1% per year over the assumed general infation
rate.
6. Coal mining labor costs — In the April analysis it was assumed
that mining labor costs would not increase at a rate different than the
assumed general inflation rate of 5.5% per year. In the August analysis,
however, the new UMW contract was included and it was assumed that mining
3—3

-------
labor costs would increase at 6.5% per year, or 1% per year more than
the assumed general inflation rate.
7. Miscellaneous — A number of miscellaneous changes were made
between the April study and the August study. These changes were
either corrections or refinements of the values used in the April study
and include revisions to the level of SIP control assumed in the model,
revisions to the scrubbing costs, changes in the assumptions regarding
industrial coal consumption, and changes to the coal supply curves used
in the April study.
In addition to presenting the results of the August 1978 analysis
this section will also present the results of other studies that analyze
impacts relevant to the decision to revise the NSPS for utility boilers.
These analyses include studies by DOE and NERA and a joint study per-
formed for the Department of Energy and Interior.
3.2 RESULTS OF THE AUGUST 1978 ANALYSIS
The results of the August analysis are presented in terms of impacts
on national and regional SO 2 emissions, utility consumption of oil and
natural gas, national coal production impacts, shipment 0 f Western coal
to the East, utility capital expenditures, increases in residential
electric bills, annualized costs of control, and present value costs of
control.
3.2.1 National and Regional Utility SO 2 Emissions
Table 3—2 presents estimates of utility 502 emissions in 1990 and
in 1995 for seven different control scenarios. These scenarios are:
1. No change in the current NSPS,
3-4

-------
Table 3—2. NATIONAL AND REGIONAL UTILITY SO EMISSIONS
(millions of tons °2 per y ar)
Control Options
______________________ 0.5 0.67 0.8
With — Without With Without Wlth — ‘ With
1990 SO 2 emissions Current NSPS Exemptions Exemptions Exemptions Exemptions Exemptions Exemptions
Oil/gas plants a 1.05 1.40 1.43 1.24 1.27 1.24 1.23
Coal plants—exis ing 13.84 13.98 14.18 13.91 14.11 13.92 13.90
—liSPS 2.13 2.26 2.37 2.26 2.38 2.22 2.22
_ANSPSC 4.39 1.22 0.96 1.33 1.07 1.52 1.76
Total Nation 21.44 18.86 18.94 18.75 18.83 18.89 19.11
Eastern region 10.19 8.95 8.93 8.95 8.91 8.94 8.99
Midwest region 7.77 7.61 7.71 7.55 7.67 7.57 7.60
West South Central region 2.25 1.49 1.50 1.38 1.38 1.45 1.56
West region 1.25 0.81 0.81 0.87 0.87 0.93 0.96
1995 SO 2 emissions
Oil/gas plants 0.59 0.69 0.72 0.69 0.72 0.69 0.69
Coal plants’ exis inga 12.54 13.06 13.22 12.95 13.07 12.90 12.94
-liSPS 2.23 2.36 2.36 2.40 2.42 2.37 2.26
_ANSPSC 7.89 2.36 1.97 2.47 2.09 2.37 3.15
Total Nation 23.26 18.47 18.27 18.51 18.30 18.74 19.04
Eastern region 11.04 8.79 8.65 8.70 8.64 8.76 8.79
Midwest region 7.91 7.45 7.38 7.43 7.29 7.39 7.45
West South Central region 2.26 1.45 1.45 1.46 1.46 1.57 1.70
West region 1.68 0.78 0.78 0.92 0.91 1.02 1.10
aExisting plants subject to SIPs.
bNew Plants required to meet the current NSPS of 1.2 lb S0 2 /milllon•Btu. NOTE: Totals may not add due to rounding.
CPlants required to meet the revised (alternative) liSPS.

-------
2. A minimum emission limit of 0.2 lb 50,/million Btu with three
exemptions per month.
3. A minimum emission limit of 0.2 lb S0 2 /million Btu without
exemptions,
4. A minimum emission limit of 0.5 lb S0 2 /million Btu with three
exemptions per month,
5. A minimum emission limit of 0.5 lb S0 2 /million Btu without
exemptions,
6. A minimum emission limit of 0.67 lb S0 2 /million Btu without
exemptions, and
7. A minimum emission limit of 0.8 lb S0 2 /million Btu without
exemptions.
Emission estimates are presented for plants that are burning either
oil or natural gas, for existing coal-burning plants subject to SIP regu-
lations, for new plants that would come under the current new source per-
formance standard of 1.2 lb S0 2 /million Btu, and for those new plants
that would be subject to a revised new source performance standard. In
addition, SO 2 emissions are presented for the Eastern region of the
United States, for the Midwestern region, for the West South Central
region, and the Western region. Table 3-2 indicates that in 1990 total
national utility 502 emissions would be reduced from approximately 21.4
million tons under the current NSPS scenario to 18.8 to 19.1 million tons,
a reduction of approximately 11% to 12% depending upon which of the six
alternative control options is selected. As would be expected, emissions
from plants directly affected by the revised NSPS are reduced most
3—6

-------
dramatically. However, some of this reduction is offset by increases
in emissions from existing coal—fired plants and oil and gas plants.
Under the current NSPS, SO 2 emissions from plants subject to the re-
vised NSPS would amount to approximately 4.4 million tons in 1990.
This amount would decrease to 1.0 to 1.8 million tons in 1990, a re-
duction of 59 to 77%. For all other plants, 502 emissions amount to
approximately 17.0 million tons in 1990 under the current NSPS scenario.
These emissions are projected to increase to 17.4 to 18.0 million tons
under the various control options, an increase of 2 to 6%. Emissions
at the regional level in 1990 measured relative to the current F4SPS
decline the most in absolute terms in the East (1.2 to 1.3 million tons)
and in the West South Central region (0.7 to 0.9 million tons). In
relative terms, the greatest reductions over the current NSPS case are
seen in the West (23 to 35%) and in the West South Central region (31
to 39%).
Similar results are seen with regard to utility SO 2 emission in
1995. As compared to the current NSPS, a revised standard would result
in a decrease in national utility SO 2 emissions from approximately 23.3
million tons to 18.3 to 19.0 million tons, a decrease of approximately
18 to 21%. Again, the decrease is most noticeable in those plants that
would be directly subject to the revised new source performance standard
and emissions from plants not directly affected by the standard increase.
Emissions in 1995 for the ANSPS plants would decrease from approximately
7.9 million tons to 2.0 to 3.2 million tons, a decrease of 59 to 75%.
3—7

-------
Emissions at the regional level in 1995 measured relative to the
current NSPS decline the most in absolute terms in the East (2.3 to
2.4 million tons). In relative terms, however, the greatest decreases
are seen in the West South Central region (25-36%) and in the West
(35—54%).
Information on regional utility SO 2 emissions from those plants
subject to the revised NSPS is also relevant to the decision regarding
the appropriate level for the revised NSPS. Table 3—2A presents this
information. In Table 3-2B information is presented on the regional
amount of coal consumed by utilities subject to the revised NSPS.
3.2.2 Utility Oil and Gas Consumption
Total consuniption of oil and gas at utilities in 1990 and 1995 is
presented in Table 3-3. As was the case with the SO 2 emission informa-
tion presented in Table 3—2, information on utility oil and gas con-
sumption is presented for seven control options. In 1990, utility con-
sumption of oil and gas would increase from 2.6 quads to 2.9 to 3.4
quads, an increase of 12 to 31% depending upon which of the six alter-
native control options is selected. This translates into an increase
in oil cornsumption from 1.2 million to 1.4 to 1.6 million bbl/day. In
1995, utility oil and gas consumption increases from 1.7 to 1.9 quads
no matter which of the six alternative control options is selected.
This represents an increase of approximately 12%.
3-8

-------
Table 3—2A. REGIONAL UTILITY S02 EMISSIONS IN 1990 AND
1995 F0F PLANTS SUBJECT TO THE REVISED NSPS
Current Control Option (With Exemptio j _
__ 0 7 —
I. 1990 ANSPS
502 Emissions (thousand tons)
• East 2081 708 715 748 789
• Midwest 555 168 171 185 214
• West South Central 1161 239 280 354 461
• West 589 107 161 231 293
• TOTAL 4386 1222 1327 1517 1756
II. 1995 ANSPS
SO 2 Emissions (thousand tons)
East 4039 1316 1334 1382 1477
• Midwest 1180 415 409 457 521
• West South Central 1575 424 420 525 659
• West 1095 206 309 420 496
• TOTAL 7889 2360 2472 2784 T54

-------
Table 3—2B. REGIONAL UTILITY COAL CONSUMPTION IN 1990 AND
1995 FOR PLANTS SUBJECT TO THE REVISED NSPS
Current Control Option (With Exemption)
NSPS 0.2 0.5 0.67 0.8
I. 1990 ANSPS
Coal Consumption (QUADS)
East 3.468 3.405 3.427 3.479 3.467
Midwest 1.165 0.790 0.801 0.80/ 0. i08
West South Central 1.934 1.671 1.971 1.964 1.954
West 1.246 1.191 1.177 1.191 1.242
TOTAL 7.813 7.065 7.376 7.440 7.470
II. 1995 ANSPS
Coal Consumption (QUADS)
East 6.731 6.392 6.468 6.491 6.665
Midwest 2.211 1.935 1.924 1.990 1.996
West South Central 2.626 2.773 2.731 2.696 2.678
West 2.283 2.317 2.292 1.268 2.266
TOTAL 13.852 13.417 13.415 13.445 13.605

-------
Table 3-3. UTILITY OIL/GAS CONSUMPTION
Control Options
0.2 0.5 0.67 0.8
With Without WTth ldithout With WitTi
Current WSPS Exemptions Exemptions Exemptions Exemptions Exemptions Exemptions
I. 1990 utility oil/gas
consumption
- quads 2.6 3.3 3.4 3.0 3.0 3.0 2.9
— million bbl/day 1.2 1.5 1.6 1.4 1.4 1.4 1.4
II. 1995 utility oil/gas
consumption
— quads 1.7 1.9 1.9 1.9 1.9 1.9 1.9
— million bbl/day 0.8 0.9 0.9 0.9 0.9 0.9 0.9

-------
3.2.3 N tiona1 Coal Production and Western Coal Shipments
Table 3-4 compares the effects of alternative control options in
1990 and 1995 on coal production and the consumption of Western coal
in the Eastern portion of the United States. As can be seen from the
table, impacts on total national coal production in 1990 are relatively
minor no matter which control option is chosen. Total national coal
production drops from 1525 million tons for the current NSPS option
to between 1499 and 1523 million tons, a decrease of 0.1 to l.7% In
1995, total national coal production changes from 1865 million tons
under the current NSPS scenario to 1859 to 1872 million tons per year,
a change of -0.3% to +0.4%.
Depending upon the option chosen, Appalachian coal production in
1990 varies from an increase of 4 million tons (0.9%) to a decrease of
16 million tons (3.4%) when compared to the current NSPS option. Mid-
western production in 1990 varies between no increase and an increase
of 43 million tons (16%). Northern Great Plains production in 1990 de-
creases for all options when compared to the current NSPS with declines
of 3to 45 billion tons (1 to 8%). Finally, Western production is seen
to vary little in 1990 with the various options. Western production
varies from a decrease of 4 million tons (2.0%) to an increase of 3 bil-
lion tons (1.5%) depending on the option.
Analysis of total coal production in 1995 shows similar impacts.
The variation in total coal production from the current NSPS case is
slight (less than 0.5%) no matter which option is selected. Regional
impacts in 1995 also show little change from those described for the
1990 cases.
3—12

-------
Table 3-4. NATIONAL COAL PRODUCrION AND WESTERN COAL SHIPPED EAST
(million tons)
Control Options
0.2 0.6 O 67 0.8
With Without With Without With With
I. 1990 national coal production Current NSPS Exemptions Exemptions Exemptions Exemptions xemptions _ Exemptions
Appalachia 465 449 469 450 467 450 449
Midwest 275 318 277 316 275 294 290
Northern Great Plains 587 542 558 552 572 5B4 583
West 198 194 195 200 201 195 201
Total 1 1!17 1 2
Western coal consumed
in East 149 118 135 117 140 147 152
II. 1995 NatIonal coal production
Appalachia 523 533 557 534 544 512 516
Midwest 332 397 333 390 326 373 364
Northern Great Plains 815 760 799 751 818 795 795
West 195 176 180 184 184 189 192
Total
Western Coal consumed
in East 210 130 173 133 204 190 196

-------
Western coal consumed in the East in 1990 under the current NSPS
scenario is projected to be 149 million tons. For all control alter-
natives other than the 0.8 option, decreases in this value are seen
which range from 2 to 32 million tons (1 to 21%). For the 0.8 option
an increase of 3 million tons (2%) over the current NSPS scenario is
seen. In 1995 all control options result in a decrease in the amount
of Western coal consumed in the East when compared to the current NSPS.
These decreases range from 6 to 80 million tons (3 to 38%).
3.2.4 Residential Bills, Capital Expenditures, Present Values, and
Annualized Costs
Table 3-5 compares the effects of alternative control options in
1990 and 1995 upon monthly average residential bills, utility capital
expenditures, present values, and annualized costs. With regard to
monthly national average residential bills, little change is seen in
either 1990 or 1995 when the current NSPS case is compared to any of
the six alternatives. In 1990 the maximum increase in monthly resi-
dential bills is $0.70/month, an increase of 1.6 percent. In 1995
the maximum increase is $1.18/month, an increase of 2.6 percent.
Comparing total utility capital expenditures in 1990 for the cur-
rent NSPS versus the four control options shows that capital expendi-
tures increase at the most by $8 billion, or 1.7 percent. In 1995,
the corresponding figures are $32 billion and 4.4 percent.
Present values increase above the current NSPS case in 1990 by
a maximum of $15 billion. In 1995 the present value increases by a
maximum of $25 billion.
3—14

-------
Table 3—5. RESIDENTIAL BILLS, CAPITAL EXPENDITURES, PRESENT VALUES,
AND ANNUALIZED COSTS
Control Options
0.2 0.5 0.67 0.8
With Without With Without With With
Current NSPS Exemptions Exemptions Exemptions Exemptions Exemptions Exemptions
I. 1990 Monthly national avg
residential bills, S/mo 43.89 44.22 44.59 44.48 44.59 44.38 44.38
Utility capital expenditures,
$ billions 478 478 477 486 485 482 483
Present value — increase over
current NSPS — $ billions — — -10 15 9 15 10 10
Annualized cost — $ billIons 91.5 93.8 93.2 93.6 92.8 92.6
Il. 1995 Monthly national avg
residential bills, $/mo 45.34 46.22 46.52 46.13 46.35 46.12 46.10
Utility capital expenditures
$ billions 733 765 765 759 757 753 752
c . ., Present value — increase over
current NSPS, $ billions —— 16 25 16 24 17 17
Annualized cost, $ billions 125.6 128.2 129.2 127.9 128.7 127.6 127.5

-------
Finally, annualized costs increase in 1990 from $91.5 billion
to a maximum of $93 8 billion, an increase of 2.5 percent. In 1995
the corresponding figures are $125.6 billion, $129.2 billion, and 2.9
percent.
3.2.5 Sensitivity of Results to Assumed Oil Prices and Rail Rate
Escalation
One phase of the August 1978 analysis conducted by EPA included
consideration of the sensitivity of the various impacts to changes in
assumptions regarding future oil prices and the amount by which rail
rates for coal transportation would increase over time. This section
will briefly discuss the results of these sensitivity runs.
The discussion of the August analysis presented up to this point
was based upon oil prices of $15/barrel in 1985, $20/barrel in 1990,
and $28/barrel in 1995. These values were decided upon by EPA after
consultation with DOE and others. After these values were specified
they came under some criticism both from people within EPA and also
from the Council of Economic Advisors. Discussions ensued which led
to the specification of two alternative oil price scenarios. One
scenario fixed oil prices at $12.30/barrel in 1985 and $17.00/barrel
in 1990. Another scenario specified the prices in 1985 and 1990 to
be $12.30/barrel and l2.70/barre1, respectively.
In addition to performing sensitivity analysis on oil prices,
EPA also decided to test the sensitivity of the various results to
assumptions regarding the rate at which coal transportation rail
3-16

-------
costs would increase. As noted earlier, the basic set of August
analyses specified that rail rates would escalate at a rate of 6.5%
per year (general inflation plus 1% per year). An alternative scenar-
io was then specified that set rail rate escalation equal to the gen-
eral inflation rate of 5.5% per year.
The results of these sensitivity analyses are shown in Table 3—6
for the 0.2 (with exemptions) control option. In 1990, with regard to
national utility SO 2 emissions, a low oil price scenario has the result
of increasing projected emissions from oil/gas plants and coalplants
subject to SIPs. Emissions from coal plants subject to the existing
NSPS and the revised NSPS decrease relative to the base case, however,
but not enough to offset the increases noted above. As a result, total
national utility SO 2 emissions increase for the low oil price scenario
relative to the base. This increase in national emissions is accompanied
by increases in each of the four regions of the country being considered
in this analysis.
The results of a very low oil price scenario are similar to the
low oil price scenario. Again, emissions increase over the base case
for oil/gas plants, coal plants subject to SIP, at the national level,
and for each of the four regions. Emissions decrease for coal plants
subject to the revised NSPS for this scenario as they did for the low
oil price scenario. With regard to emissions from coal plants subject
to the revised NSPS, however, the very low oil price scenario shows a
slight increase in emissions whereas the low oil price scenario showed
a decrease in emissions.
3-17

-------
Table 3—6.
A COMPARISON OF ALTERNATIVE OIL PRICE AND RAIL RATE
ASSUMPTIONS UPON RESULTS FOR THE 0.2 (WITH EXEMPTIONS)
OPTION IN 1990
BASE
CASE 1
LOW
OIL
PRICE
VERY LOW
OIL
PRICE
NO RAIL
RATE
ESCALATION
I. National Utility SQ 2 Emissions
(millions of tons)
(A,
Oil/Gas Plants
Coal Plants — Exis ing 5
— NSPS
1.40
13.98
2.26
1.87
14.15
2.19
2.41
14.35
2.27
1.22
13.81
2.16
— ANSPS
1.22
0.99
0.65
1.26
TOTAL NATION
18.86
19.20
19.68
18.45
Eastern Region
8.95
9.08
9.35
8.84
Midwest Region
7.61
7.68
7.80
7.46
West South Central Region
1.49
1.58
1.63
1.33
West Region
0.81
0.86
0.90
0.82
II.
Utility Oil/Gas Consumption
Quads
3.3
4.4
6.1
3.0
.
Million bbl/day
1.5
2.1
2.9
1.4
III.
National Coal Production
(million tons)
Appalachia
449
440
422
436
Midwest
318
303
281
289
Northern Great Plains
542
518
498
618
West
194
192
182
194
TOTAL
1502
1454
1382
1537
IV. Western Coal Consumed in the East
(million tons)
118 116
118 174

-------
Table 3—6. A COMPARISON OF ALTERNATIVE OIL PRICE AND RAIL RATE
ASSUMPTIONS UPON RESULTS FOR THE 0.2 (WITH EXEMPTIONS)
OPTION IN 1990 (CONTINUED )
LOW VERY LOW NO RAIL
BASE 1 OIL 2 OIL 3 RATE
CASE PRICE PRICE ESCALATION
V. Selected Economic Information 8
• Monthly National Avg.
Residential Bill, $/mo 44.22 43.66 42.62 NA
• Utility Capital Expenditures,
$ Billions Cumulative 478 453 408 NA
• Present Value — Increase Over
Current NSPS, $ Billions 10 15 11 NA
• Annualized Cost, $ Billions 93.4 91.3 87.1 NA
CA)
‘ .0
1990 Oil Price of $20.00 /bbl; Rail Rate Escalation of 1% over inflation.
2 1990 Oil Price of $17.00 /bbl; Rail Rate Escalation of 1% over inflation.
1990 Oil Price of $12.70 /bb l; Rail Rate Escalation of 1% over inflation.
1990 Oil Price of $20.00 /bbl; Rail Rate Escalation of 0% over inflation.
Existing plants subject to SIPS.
6 New plants required to meet the current NSPS of 1.2 lb S0 2 /million BTU.
Plants required to meet the revised (alternative) NSPS.
8 All economic data is presented in 1975 dollars. All cost information includes costs for controlling
particulate emissions for new sources to a level of 0.03 lbs per million BTU.

-------
The alternative rail rate scenario displayed on Table 3—6 shows,
relative to the base case, decreases in SO 2 emissions from oil/gas
plants, coal plants subject to SIPs, coal plants subject to the cur-
rent NSPS, and at the national level. An increase in emissions is
seen for plants subject to the revised NSPS. At the regional level,
decreases in emissions relative to the base case are shown for all
regions with the exception of the West where a slight increase is seen 0
Table 3—6 further goes on to show, as expected, that both the low
oil price scenario and the very low oil price scenario result in in-
creased oil/gas consumption at utilities relative to the base case,
The results of the alternative rail rate scenario, however, is to de-
crease the amount of oil/gas consumption at utilities relative to the
base case.
Also presented on Table 3-6 is information on coal production.
As expected, as oil prices decrease demand for coal decreases at the
national level and for each of the four regions under consideration.
As rail rates decrease relative to the base case coal becomes rela—
tively less expensive and more coal is prodUced at the national level.
This increased national production is solely attributable to increased
Northern Great Plains production, however, since the other three re-
gions show either decreased production or no change in production rela-
tive to the base case.
Reducing the price of oil has relatively little effect on the con-
sumption of Western Coal in the East, however, as seen on Table 3-6.
3-20

-------
The effect of the alternative rail rate scenario is to increase the
base case consumption of Western coal in the East by 6 million tons,
or 47 percent.
Finally, Table 3-6 shows the effects of alternative oil price
scenarios on various economic factors. Information on the effect of
the alternative rail rate scenario is not available. It is seen that
lowering the oil price results in a decrease in all of the economic
indicators shown on Table 3-6 except for the present value. In the
case of this statistic a low oil price scenario results in a present
value increase of $15 billion over the current NSPS, a very 10w oil
price results in an increase of $11 billion, and the base case has an
increase of $10 billion.
3.3 OTHER ANALYSES
In addition to the analyses described above, three other major
studies have been conducted which analyze certain impacts of a revised
NSPS for power plants. One of these studies was performed by ICE, In-
corporated, under a joint contract with the Department of Interior and
the Department of Energy. A second study was performed by the Depart-
ment of Energy and the third study was performed for the Utility Air
Regulatory Group (UARG) by National Economic Research Associates, Inc.
tNERA).
3.3.1 Joint DOIJDOE Study
EPA has been using the report developed by ICF for the Departments
of Interior and Energy in order to study the sensitivity of certain re-
suits to changes in input assumptions. This section will reproduce one
3—21

-------
table from the June 1978 draft report entitled The Demand For Western
Coal And Its Sensitivity To j y Uncertainties and also describe the
major assumptions underlying the study.
Table 3-7 is taken directly from the
tioried above. The following is a list ot
the table:
1. High severance tax in the West: North Dakota, Montana,
Wyoming, Colorado, Utah, and New iJexico 30 percent severance
tax whereas base case has actual state severance taxes.
2. Low severance tax in West: same as above except severance
tax is 5 percent.
3. High electricity growth rate: one percentage point higher
national electricity growth rate than base case, which is
4.8 percent for 1975-1985 and 4.0 percent for 1985—1990.
4. Low electricity growth rate: one percentage point lower
national electricity growth rate than the base case.
5. Higher oil prices: same as DOE high dase: $20 per barrel in
1995, $30 per barrel in 1990 in 1975 dollars (relative to l5
and S20 per barrel for 1985 and 1990 respectively used for
base case).
6. Lower oil prices: same as DOE low case —— $13 per barrel in
1985 and 1990 in 1975 dollars.
7. Revised new source performance standards: require full scrub-
bing (i.e., 90 percentremoval), whereas partial scrubbing was
allowed (i.e., 0.5 pounds of sulfur dioxide per million BTU’s
June 1978 draft report men-
the major assumptions behind
3-22

-------
Table 3-7
*990 at010lini. COAL Plth)UC’I*ON
18*088 DIFFONEWr S NS1T1V1TY ASSUMPTIONS
(IA ton i,)
__________________________________________ S c enar to
H igh L ou High Lo High Lou High tow Nigh Cc bined
Severance Severance Electricity Ele ,,tricity hAL Oi.I 905 1.7 lb. Libo Labor MU Cycle
_ egion ge_ Tan Tox Loi . A iJ o 1 Sten ,1.srd Escalation FeceSetion MACI AlI e4
Northern
Appa3achla 225.6 233.0 206.5 240.1 204.6 240.2 206.8 221.5 207.6 211.5 232.1 242.3 214.3
Ce ’, A rat
Apjnlachia 207.1 216.2 107.1 219.0 195.4 215.0 202.5 203.4 251.5 194.0 214.7 227.5 205.8
Southern
ApNiach.ia *3.8 13.8 13.8 13.8 13.8 13.8 13.8 13.8 13.8 *3.8 13.0 13.8 13.0
Fatal 446.5 463.0 407.9 490.9 413.7 469.0 423.2 444.7 476.0 425.3 460.6 403.5 443.8
Mjduent 313.5 332.7 217.9 339.5 277.9 328.7 296.5 323.5 254.5 287.7 323.1 350.2 310.5
Total 313.5 332.1 277.9 339.5 277.9 228.7 296.5 323.5 254.5 287.1 323.1 350.2 310.5
Eu 0 A e rn
Northern
Great Plai’,s 22.7 20.2 21.9 26.1 21.9 22.5 21.9 21.9 21.8 23.0 22.) 23.0 22.1
West em
Oor them
Cre..t Plain. 528.0 510.1 652.5 656.3 455.4 540.3 492.4 517.9 590.2 548.2 502.6 424.2 519.5
Total ¶ .Sl.5 538.4 614.4 602.1 417.3 567.9 514.3 539.8 612.0 622.0 526.3 448.1 541.6
fentmal West 10.3 10.7 9.6 10.3 9.6 10.3 10.3 9.6 9.7 9.6 9.9 *0.9 9.9
Gulf *04.1 107.9 004.4 107.9 95.1 104.1 504.5 104.1 94.6 104.1 104.1 *07.9 *04.1
8ocky
Nountainp 53.0 36.6 45.5 53.3 53.6 53.0 52.5 52.6 56.7 52.1 53.0 52.4 53.0
SOuttoo st 47.6 37.4 43.2 64.0 41.6 47.6 30.1 48.2 53.7 40.5 48.3 38.4 48.1
horthuest 3.7 3.7 3.7 3.7 3.7 3.7 4.4 6.0 3.7 3.7 4.6 6.0 4.4
Total 210.6 156.1 206.3 239.2 203.6 2*8.6 201.5 220.6 217.9 217.9 219.8 215.6 219.4
NAT I oNAL
0’ )TA1. 1,510.5 1,530.3 1,565.9 1,741.9 1,372.6 1,584.1 1,435.5 l.520.S 1.560,4 1,552.9 1,529.9 1,497.3 2
We*tnrfl Coil
Cono ed in
East 1)3.0 *02.1 233.5 198.3 109.4 144,1 *28.5 159.1 195.3 191.3 *11.3 49.2 130.1
National
Ut Ill op
yU ,,l ron —
I u. Is I
— CQ O A 21.3 21.3 71.3 15.5 *9.1 22.5 19.3 21.3 21.9 21.3 21.5 21.7 21.0
— O Il 2.9 2.9 2.9 3.2 2.5 1.5 5.0 2.9 1.9 2.9 7.7 3.0 3.1

-------
floor) in base case.
8. Current new source performance standards: (i.e., 1.2 pouncis
per million (BTU’s) which can be met with low sulfur coal
without scrubbing.
9. High labor cost escalation: same as DOE high range — two
percent per year real escalation after 1980, relative to
one percent in base case.
10. Low labor cost escalation: same as DOE low range — no real
escalation after 1980.
11. High rail rates: 50 percent increase in rates by 1985,
relative to no increase in base case.
12. New combined cycle oil plants allowed by federal regulation,
relative to such plants being prohibited in the base case.
3.3.2 DOE Proposal
In a memo to EPA dated July 8, 1978, DOE proposed that the revised
NSPS for power plants should reflect 85% removal of SO 2 on a monthly
basis and an emission floor of 0.5 lb SO 2 per million BTU averaged on a
monthly basis. In a subsequent memo to EPA dated August 11, 1978, DOE
reproposed that that NSPS reflect an emission floor of 0.8 lbs SO 2 per
million BTU averaged on a daily basis. Table 3—8 is taken directly from
the DOE memo to EPA dated July 8, 1978 and compares the current NSPS to
the EPA full scrubbing (0.2 floor) control option and the alternative
partial scrubbing options of 0.5 and 0.8 lbs SO 2 per million BTU, aver-
aged on a monthly basis. Impacts for the 0.8 daily averaging time option
were not presented in either of the DOE memos referred to above .
3-24

-------
TABLE 3-8. COMPARISON OF EPA AND DOE PROPOSALS 1
Current EPA 0.5 2 0.8
NSPS Proposal Floor Floor
Coal Capacity in 1990
(l000s of megawatts) 465 430-444 460 460
Present Value Cost
through 1990 ($billions) 36 62-73 49—51 46
SO Emissions in 1990
mi11ions of tons) 21.3 19.0-19.7 19.3-20.0 20.3
Oil and Gas Consumption
in 1990 (millions of
bblfday 3.1 3.5-3.8 3.2 3.2
Coal Production in 1990
(millions of tons)
East 441 464-466 464 449
Midwest 298 373—375 353 313
West 1027 835—869 _ 938 998
TOTAL 1767 1672—1711 1755 1760
Western Coal Consumed
East of the Mississippi
River (millions of tons) 455 275—300 345 381
1 From DOE memo to EPA dated July 8, 1978.
2 Values from Tables 4 and 5 of the July 8 memo.

-------
3.3.3 UARG/NERA Analysis
The Utility Air Regulatory Group of the Edison Electric Institute
(UARG) requested that the National Economic Research Associates, Inc.
(NERA) conduct a study to evaluate the impacts of various alternative
proposals for the revised NSPS for power plants. NERA evaluated six
alternative proposals. These are described in Table 3 -9. The major
results of the NERA analysis are presented in Table 3-10.
3-26

-------
Table 3—9.
DESCRIPTION OF CASES
SOz Emissions Standards 1
Allowable
Percent Averaging Exemptions Particulate
Case Floor Ceiling 2 Removal Time per Month Standard
(lbs. S0 2 /MMBtu) (Percent) (lbs./MMBtu)
(1) (2) (3) (4) (5) (6)
Base
NSPS 1.20 30—day None 0.10
Proposed Revisions
to Current NSPS :
UARG1 1.20 20—85 30—day None 0.08
UARG2 1.20 20—85 30—day None 0.03
EPA1 0.20 1.20 85 24—hour None 0.03
EPA2 0.20 - 1.20 85 24—hour 3 0.03
DOE1 0.50 1.20 85 24—hour 3 0.03k
DOE2 0.80 1.20 85 24—hour 3 0.03’
1 For all cases scrubber systems are assumed to be 100 percent available and are
designed with two redundant modules unless only one operating module is re-
quired, in which case there is one redundant module.
2 Phe ceiling is 1.20 pounds of sulfur dioxide per million Btu or the SIPS,
whichever is more restrictive.
3 while DOE recommends a 30—day rather than 24—hour averaging time, the 24—hour
averaging time with 3 allowable exemptions was used for the purposes of com-
paring the DOE and EPA percent removal proposals.
‘While DOE recommends a 0.05 to 0.08 particulate standard, a 0.03 standard was
used for purposes of comparing the DOE and EPA SO standards.

-------
Table 3—10. SUMMARY OF IMPACTS AS ANALYZED BY NERA FOR UARG
1
1990 Utility SO 9 Emissions
(millions of tbns)
1990 Annual Cost
($ Billions 1977)
SO 2 Removed
(millions of tons)
$/ton $02 Removed
Present Value Cost Above
Current NSPS
($ Billions 1977)
1990 Oil/Gas Consumption
(miflion bbl/day)
1990 Total Coal Production
(million tons)
Current
NSPS
“Analysis of the Economic Impact of Alternative Proposed Revisions to
National Economic Research Associates, Inc., August 29, 1978
New Source Performance Standards”,
cJ
UARG 1
18.24
7.15
23.87
300
13.50
0.79
530.1
329.8
642.1
1501.8
20.02
5.96
22.09
270
0.67
531.0
312.5
665.8
1509.3
UARG 2
18.26
7.52
23.85
315
17.70
0.79
530.1
331.4
640.2
1501.4
EPA 1
16.84
9.28
25.27
367
37.66
0.94
502.9
385.0
598.4
1486.0
EPA 2
16.86
• 9.15
25.25
362
36.18
0.94
502.9
385.0
604.3
1491.9
DOE 1
16.88
8.09
25.23
321
24.16
0.93
507.4
359.8
624.7
1491.5
Appalachia
Midwest
West
1OTAL
DOE 2
17.36
7.54
24.75
305
17.92
0.79
531.9
326.4
643.4
1502.2

-------
4. DISCUSSION OF REGULATORY TOPICS
4.1 INTRODUCTION
This section summarizes many factors considered by EPA in evaluating
various regulatory topics. Discussion of these issues gives additional
insight into several regulatory requirements drafted and the impacts
they may have on electric utility companies and coal producers. The
performance of flue gas desulfurization (FGD) systems in the U. S. and
Japan and the projected performance of improved systems is discussed in
section 4.2. New dry control systems that may reduce SO 2 control costs
for low-sulfur coal to levels even less than those for partial scrubbing
are discussed in section 4.3. Other topics discussed are SO 2 control in
noncontinental areas (4.4), use of continuous monitoring systems for
compliance determinations (4.5), procedures for shifting electric load
and maintaining electric system reliability during FGD malfunctions
(4.6), and potential impacts upon coal production, Midwestern coal
reserves, and anthracite coal (4.7).
4.2 FGD PERFORMANCE
4.2.1 Overview
EPA has evaluated potential systems of continuous SO 2 emission
reduction (considering the cost, and non-air quality health and environ-
mental impact, and energy requirements of such reduction) and concluded
that FGD systems used to remove SO 2 from flue gases have been adequately

-------
demonstrated. This section summarizes the test results of these demonstrations
and discusses the performance of FGD systems currently available.
FGD systems have been designed to use several different types of
absorbers and absorbents. All FGD designs installed and tested have not
been equally successful. To evaluate the relative performance of FGD
system designs, EPA has tested several absorber designs and absorbents
at the Shawnee 10-MW test facility. The Shawnee test results give
valuable information regarding FGD system design for high percentage SO 2
removal.
EPA has also surveyed the performance test results of full-scale
FGD systems and has closely evaluated the continuous, long-term performance
of FGD systems by use of continuous monitoring systems at several full-
scale power plants. A sumary of these tests, surveys, vendor guarantees,
and EPA’s analysis of requirements for attainment of high percentage SO 2
removal are given in the following sections.
4.2.2 Prototype FGD Unit Performance
At the Shawnee test facility, several FGD system designs and
absorbents were evaluated for 6- to 8-hour or 1-week test periods under
controlled operating conditions. The short-term (6 to 8 hours) tests
were directed at determining SO 2 absorption efficiencies without considering
potential long-term operating problems; however, during other portions
of the Shawnee test program, these FGD systems were operated for extended
periods without scaling problems. A summary of these test runs and
conclusions was prepared for EPA by Bechtel Corporation in their report,
Flue Gas Desulfurizatiofl Systems: Design and Operating ConsiderationS ,
report number EPA-600/7-030b.
4-2

-------
By controlling the operating conditions at Shawnee te specific
levels, direct comparisons between performance of different types of
absorbers and absorbents were obtained, and the principal factors that
have the greatest effect upon absorption efficiency were identified.
These data show that a proper balance between (1) absorber design, (2)
absorber operating conditions, and (3) type of absorbent must be achieved
in order to attain a specific FGD control efficiency. The absorber
design determines the amount of contact at the gas—to-liquid interface.
The operating conditions determine the concentration and amount of
absorbent available to the absorber, and the type of absorbent determines
the reactivity or rate of reaction with the SO 2
In the Shawnee tests, venturi absorbers were found to be less
effective in comparison to turbulent contact absorbers (TCA) or spray
tower absorbers (STA). When tested with less reactive calcium—type
absorbents under the same operating conditions, the venturi absorber
efficiency was only 40 to 50 percent. However, venturi absorbers can be
effective when used in series or when a more reactive absorbent is used.
Bechtel has projected that two venturi absorbers used in series can
attain greater than 95 percent mean SO 2 removal with lime or limestone
absorbents. The Shawnee data also contain tests where more reactive
absorbents were used. The single venturi absorber tested removed 85
percent SO 2 with magnesium oxide enriched limestone and 95 percent SO 2
with two sodium-type absorbents during the tests.
The improved performance of venturi absorbers with a more reactive
sodium-type absorbents was also demonstrated by an 8-month test program
of a 20-MW double-alkali EGO module at the Scholz station of Gulf Power
Company. With a single venturi absorber, the FGD module was able to
attain over 90 percent average SO 2 removal at several pH levels. When a
tray tower absorber was added in series with a venturi absorber, up to
4-3

-------
99 percent average SO 2 removal was obtained. Tests of a few hours
duration were performed at each operating condition to determine FGD
average efficiencies with each configuration.
In the Shawnee tests, STA and TCA absorbers demonstrated 90 percent
average SO 2 removal efficiency with either lime or limestone absorbents
when the pH levels selected were the maximum that could be used without
scale formation. The liquid-to-gas (L/G) ratios needed to attain 90
percent efficiency were 30 to 40 percent lower with the TCA absorber
than were needed with the STA.
A second series of tests were run with more reactive magnesium oxide
enriched limestone. At the same hG ratio used in the limestone tests,
the TCA absorber efficiency was increased from 90 to 97 percent. The
STA absorber also attained 97 percent efficiency and unlike the previous
limestone tests, required no greater L/G ratio than the TCA.
An additional factor that can affect operating conditions is the gross
load of SO 2 removed by the FGD system. This load increases with higher
sulfur fuels. Each 1 percent sulfur in bituminous coal is roughly
equivalent to 800 ppm of SO 2 at the absorber inlet. The flue gas treated
at Shawnee typically contained 2000 to 3000 ppm SO 2 prior to treatment;
however, a few additional tests were run at SO 2 inlet concentrations of
less than 1000 ppm. The percentage removal efficiency increased for the
same absorber, absorbent, and operating conditions at reduced 502 inlet
concentrations. The data show that the absorber design and operating
conditions to attain high SO 2 absorption efficiency are more moderate
for low-sulfur coals. Tests at Mohave power station with full-scale FGD
modules (170 MW) also show moderated operating conditions. At 200 ppm SO 2
inlet concentrations, L/G ratios of only 20 gal/Mcf were necessary to
obtain 95 to 99 percent SO 2 removal.
4-4

-------
In addition to Shawnee, several other small FGD modules (10 to 20
MW) have demonstrated greater than 90 percent 502 removal efficiency.
In PEDCo’s report prepared for EPA, Flue Gas Desulfurization System
Capabilities for Coal-Fired Steam Generators , Table 4-7, report number
EPA-600/7-78-032b, two prototype units at Gulf Power Company and one
operational unit owned by the U. S. Air Force are reported to have
attained 95 to 99 percent average SO 2 control.
The prototype-size EGO units described in Bechtel’s and PEDCo’s
reports attained removal efficiencies up to 95 percent with high sulfur
coal and up to 99 percent efficiency with low sulfur coal during short-
term tests. These tests show that absorber design, operation, and
abosrbent type can be properly matched to attain high mean FGD efficiency.
4.2.3 Full-Scale EGO Systems
4.2.3.1 Vendor Guarantees and Other Statements
Through EPA’s contractor, PEDCo Environmental, Inc., the performance
guarantees offered by EGD module suppliers were solicited. The results
of this survey are summarized in Table 3-10 of the report Effects of
Alternative Sulfur Dioxide New Source Performance Standards on Flue Gas
Desulfurization System Supply and Demand , report number EPA—600/7-78- .
033. Ten suppliers of EGO modules reported that they would guarantee °2
removal performance greater than 90 percent. Five specifically mentioned
guarantees of 95 percent. The performance guarantees offered are
generally based upon short-term acceptance tests, which are typically 6
to 8 hours duration. Two-thirds of the respondents to the survey stated
that under contract, they would be willing to operate and maintain the
EGO system after installation.
In December of 1977, EPA held a meeting of the National Air Pollution
Techniques Advisory Committee. At that meeting a spokesman for TVA who
is also a member of the Utility Air Regulatory Group stated that an 85
4-5

-------
percent SO 2 removal requirement (24-hour basis) could be attained. A
second spokesman from a major vendor of lime FGD systems agreed.
4.2.3.2 EGO Performance Test Results
As of November 1977, 32 operational FGD systems had been installed
at utilities in the U. S. These 32 systems serviced 10,717 megawatts of
electric generating plant capacity. An additional 34 FGD systems
servicing 14,219 megawatts of plant capacity were under construction and
contracts were awarded for 20 FGD systems for power plants designed to
produce 9,758 megawatts. Table 4-1 lists some of the U. S. FGD units
designed for 85 percent, or greater, SO 2 removal which are installed,
under construction or for which contracts have been awarded. In PEOCo’s
report, 10 utility size (65 to 170 MW) FGD modules were listed as having
demonstrated greater than 90 percent SO 2 removal on coal-fired boilers.
Table 4-2 lists some of the U. S. facilities reporting 90 percent or
greater SO 2 removal.
At the beginning of 1978, there were over 500 power plants with FGD
systems in Japan. These systems control SO 2 emissions from about
31,000 megawatts of electric generating capacity. About half of this
capacity consisted of utility boilers. Seven of the boiler installations
burn coal with a reported sulfur content of 0.6 percent to 2.5 percent.
There are 25 utility and industrial applications of U. S. scrubber
technology in Japan. The FGD systems in Japan (both U. S. and Japanese
design) on coal-fired boilers reportedly operate with SO 2 removal efficiencies
ranging from 90 percent to 93 percent and, within a few months after
startup, achieve EGO system availabilities of 95 to 100 percent. Table
4-3 provides pertinent data concerning the operation of U. S. design FGO
systems installed on coal-fired boilers in Japan. The performance of
FGD systems on five utility boilers and two industrial boilers are
4-6

-------
Table 4-1. LIST OF U.S. FLUE GAS DESULFURIZATION (FGD) SCRUBBERS INSTALLED
UNDER CONSTRUCTION, OR AWARDED-DESIGNED FOR 85 PERCENT OR GREATER SO 2 REMOVAL
Under construction or contract awarded
No. Capacity Coal
FGD’s MW SUl ’fur,%
FGD
‘Company type
Existing
—4
Design SO
removal
Startup
date
Arizona Public Service
Limestone
1
115
0.4—1.0
92
10/73
Columbus & Southern Ohio Electric
Lime
1
400
4.5-4.9
90
2/77
Duquesrie Light
Lime
2
920
1-2.8
85
10/75
Louisville Gas & Electric
Lime
1
178
3.5-4.0
85
8/76
Northern Indiana Public Service
Weilman Lord
1
115
3.5
90
77
Pennsylvania Power Company
Lime
2
1670
4.7
92
4/76
Philadelphia Electric
MgO
1
120
2.5
90
9/75
Nevada Power
Sodium Carbonate
3
375
0.5-1.0
85—90
4/74
Allegheny Power System
Columbus & Southern Ohio Electric
Lime
Lime
2
1
1250
400
4.5
4.5-4.9
90
90
3/80
77
Public Service Co. of New Mexico
Weliman Lord
2
715
0.8
90
78
Southern Mississippi Electric
Deirnarva Power Company
Louisville Gas & Electric
Niagara Mohawk Power Coop.
Limestone
Weilman Lord
Double Alkali
Aqueous Carbonate
2
1
1
1
360
180
277
100
1.0
7.0-8.0
3.5—4.0
2.5-4.5
(coke)
85
85-90
90
90
78
4/SO
1l/?
6/78
Pennsylvania Power Co.
Lime
1
825
4.7
92
4/80

-------
Table 4-2.
U. S. PLANTS REPORTING 90 PERCENT OR GREATER SO 2 REMOVAL
Utility
Station
MW FGD Process Fuel SUlfura,% SO 2 removal,%
Cholla No. 1
Phillips
Mitchell
Ed dyston e
Mys t i C
Dickerson
Mohave 1 & 2
Green River
Conesvi lie
115
410
115
120
150
95
Each 170
Limestone
Lime
Wellman Lord
MgO
MgO
MgO
Lime
0.4-1.0
1.0-2.8
3.5
2.5
2.5 (oil)
2.0 (oil)
0.6
3.8
4.5-4.9
92
90+
go
95-98
90
90
95
90+
90+
Arizona Public Service
Duquesne Light
Northern Indiana Public Service
Philadelphia Electric
Boston Edison
Potomac Electric Power
Southern California Edison
Kentucky Utilities
Columbus & Southern Ohio
a Unless otherwise noted, fuel
c o
64 Lime
400 Lime
burned is coal.

-------
Table 4.3. AMERICAN DESIGNED FGD SYSTEMS OPERATING IN JAPAN
Developer, -
Power type Number Capacity Date of Coal Co S Average so 2 Availability,
Company station FGD FGD MW startup source sulfur,% removal ,, %
EPOC Takasago Mitsul-Chemico
Limestone 2 500 1975 & Blended 2 93 98.6
1976 Domestic
EPDC Isogo Chemico-IHI
Limestone 2 530 1976 Domestic 0.4 93 99+
EPDC Takehara Babcock-Hitachi 1 256 1977 Blended 2 93 97
Limestone Domestic
Mitsui Miiki Mitsui-.Chemico 1 156 1972 Blended 2.4 90+ 100
Carbide Lime Domestic
Mitsui Miiki Mitsul-Chemico 1 175 1975 Blended 2.4 90+ 99
Limestone Domestic

-------
described. On oil-fired boilers, up to 98 percent SO 2 removal has been
reported.
The PEDCo report and a report by J. Ando, SO 2 Abatement for Stationary
Sources in Japan (report number EPA-600/7-77-103a), discussed several
oil—fired boiler FGD systems that have attained high efficiencies.
Of the nine FGD systems attaining greater than 90 percent control listed
(Table 3-9) in the PEDCo report for oil-fired power plants, four attained
greater than 95 percent SO 2 removal. Control of oil-fired-boiler 502
emissions is technically less troublesome than those from coal-fired
boilers because (1) the flue gases are not so heavily laden with particulate
matter and (2) the physical properties of refined and blended fuel oil
(e.g., percent sulfur) are more uniform than coal.
Tests of coal-fired-boiler EGO systems in the U. S. have demonstrated
95 percent or greater control. With the less difficult to scrub low
sulfur Western coals, a 170-MW horizontal crossflow spray scrubber
module using lime absorbent demonstrated 95 percent average SO 2 removal
at Arizona Public Service Company’s Four Corners station. This same
module and a 170-MW TCA absorber each demonstrated 95 percent average
removal at Southern California Edison’s Mohave station when applied to
control of low-sulfur-coal emissions.
Two FGD units owned by the Louisville Gas and Electric Company have
demonstrated greater than 95 percent removal of SO 2 from high-sulfur-
coal emissions during short-term tests when magnesium oxide was added to
the lime absorbent. At the Cane Run station MgO was added to lime until
the SO 2 removal efficiency reached 95 to 96 percent for 3 days. At the
Paddy’s Run station MgO additions to the lime produced 99.7 to 99.9
percent SO 2 removal during 8-hour tests on 2 separate days.
The U. S. system tests were relatively short—term, lasting a few hours
or up to 3 days, and may not adequately estimate the range of 24—hour
4-10

-------
average oerformance or the mean performance of these systems. The
Japanese data reported in Table 4-3 are representative of mean performance
because these data are averages of long-term continuous-monitoring
system results.
The data averages presented in this section do not define the range
of FGD performance expected during normal operation. The results of
more extensive test programs to define this range are given in section
4.2.4.
4.2.4 Analysis of FGD Performance Range
Although many FGD modules have been tested, few systems are reported
to have been continuously evaluated for more than 3 days. Short-term
performance tests under controlled conditions provide good design information
that can be used to predict absorber Derformance under fixed conditions,
but they do not reveal how an FGD system would respond to the uncontrolled
variations in SO 2 concentrations at the absorber inlet that are experienced
during boiler operation. These variations are typically due to the fuel
quality. Refined fuel oil is typically homogeneous, but coal is not.
Variability of sulfur in coal is an inherent property created
during the fossilization process, and all naturally occurring coals can
be expected to produce variations in the SO 2 concentrations of boiler
flue gas. Sulfur in coal is unevenly distributed within a coal seam.
Coal shipments, even from the same mine, will have a range of sulfur
content, which has been described by statistical distributions in an EPA
report ( Preliminary Evaluation of Sulfur Variability in Low—Sulfur Coals
from Selected Mines , report number EPA-450/3-77-044, Nov. 1977) and in a
paper by C. Nelson and J. Dragos “Coal Variability and Sulfur Compliance.”
An EPA review of the variation in coal sulfur content is given in section
4.2.7.
4-11

-------
Two methods are currently feasible for reducing the variation in
coal quality. One is coal washing, which can remove pockets of pyritic
sulfur, but is ineffective in removing sulfur chemically bound to the
organic constituants of the coal. The second method is coal blending,
which can be effective when the mean sulfur content of separate coal
shipments is monitored. A third method, solvent refining of coal, is
currently under development, but may not be economically feasible unless
it can completely replace the FGD system for SO 2 control. Coal-fired
units with EGO systems in Japan are fired with blended coal to reduce
the variation in potential emissions resulting from use of coal supplies
with different sulfur contents. At least one unit in the United States is
installing coal blending facilities specifically for control of coal
sulfur variability.
The effect that coal quality can have on FGD performance is evident
in tests at Shawnee. In three tests of up to 20 days each, sharp increases
in inlet SO 2 concentrations lead to corresponding decreases in SO 2
removal efficiency. During these tests, no attempt was made to counter
this effect by manual or automatic adjustments to the FGD system.
In order to more closely evaluate the SO 2 variability effect and
the impact of FGD process controls, continuous SO 2 emission monitors
were placed in three full—scale operational units; Louisville Gas and
Electric’s Cane Run 4, Pennsylvania Power’s Bruce Mansfield 1, and
Philadelphia Electric’s Eddystone 1. None of these FGD systems has an
automated control system, but adjustments to slurry pH were made manually
by the process operators. Continuous monitoring performance data were
also obtained from Northern Indiana Public Service Company’s full—scale
Weliman-Lord EGO unit at the Mitchell station.
The continuous monitoring data taken at each of the facilities
tested are summarized and evaluated in EPA’s report entitled First Interim
4-12

-------
Report: Continuous Sulfur Dioxide Monitoring at Steam Generators, ”
June 1973 (EMB Project Number 77SPP23A). The report evaluates the
variability i n potential emissions at the absorber inlet, in FGD system
502 removal efficiency and in emissions to the atmosphere. With minor
exceptions, all data were found to best fit a log—normal statistical
distribution. The report describes the geometric mean and geometric
standard deviations of the EGO inlet data, FGD SO 2 r noval efficiency,
and the FGD outlet data, which represent the emissions to the atmosphere.
Figure 4-1 suninarizes the percent SO 2 reduction data for each facility.
The data displayed are a probability distribution of EGO 24-hour average
502 removal efficiencies. These data, which do not include periods of
startup and shutdown, are representative of the range in EGO 24-hour
average performance normal for these systems.
The data in Figure 4-1 give the mean 502 removal efficiency for
each EGO system and the probability of occurrence for higher and lower
average EGO efficiencies during 24-hour compliance periods. These data
show about the same performance for the lime EGO systems at the Can Run
and Bruce Mansfield stations. The regenerable FGD systems at Mitchell
(Wellman-Lord) and at Eddystone (t4g0) stations show the superior
performance of more reactive absorbents.
4.2.5 Projected FGD Performance
4.2.5.1 Mean FGD Performance
A “line of improved performance” is projected on Figure 4-1 through
a mean SO 2 percent removal level of 92 percent. This projected increase
in mean EGO performance level is supported by available information.
Shawnee data show that attaining 92 percent efficiency with an figO
enriched-lime or liinestone-absorbant is technically easier (requires
4-13

-------
99
I I I 11111
I I I I III
5 10 20 30 40 50 60 70 80
PROBABILITY, percent
Figure 4 -1. FGD 24-hour average efficiency distribution.
I I I
11111’
90 95 98 99 99.9
II I I
98 —
97 —
96 —
4. . 95
=
a,
U
>.
C.,
2
I i . ’
;390
U-
U-
UJ
85
U-
80—
70 —
60 —
50
40
30
20
10
—I I I I
0.01 0.05 0.1 0.5 1
4-14

-------
lower L/G or pH levels), but that FGD systems can attain 92 percent
efficiency without enriching the absorbent when the proper absorber is
used. With the TCA absorber and no MgO enrichment, tests at Shawnee
showed 92 percent or greater removal when lime or limestone were used.
With TCA or similar absorbers using sufficient gas/liquid residence
time, reaction tank residence time, and other design factors, the less-
reactive calcium-type absorbents enriched with MgO have attained 95
percent mean SO 2 removal. With the most reactive absorbents, sodium or
magnesium types, all absorbers tested at Shawnee were capable of achieving
95 percent mean SO 2 removal.
In addition to the 10-MW pilot plant at Shawnee, several full-scale
EGO units have demonstrated 92 percent efficiency. Five American designed
limestone systems installed on Japanese coal-fired boilers (Table 4-3),
four units installed on Japanese oil-fired boilers, and several tests in
the U.S. have demonstrated 92 percent mean EGO efficiency. This level
of performance was demonstrated by the full-scale FGD systems at Paddy’s
Run and at Cane Run by using MgO additions to the lime absorbent. The
lime FGD system at the Mohave and Four Corners tests are also examples.
The feasibility of attaining 92 percent mean FGO efficiency is also
supported by the performance of regenerable and double-alkali FGD systems.
The full-scale units operated at the Mitchell station (Weliman—Lord
system) and at the Eddystone station (MgO system) were evaluated with
continuous monitoring systems. When the FGD performance data were
reduced to 24-hour averages, the mean of these average SO 2 removal
efficiencies was 90 percent at the Mitchell station and 97 percent at
the Eddystone station (Figure 4-1). The ability of the more reactive
4-15

-------
sodium or magnesium type absorbents to attain over 95 percent removal is
also documented by several Shawnee and Scholz station (double alkalI)
pilot-scale tests.
Double alkali or regenerable FGD systems consistently have higher
performance, and their use may be needed for certain applications
(midwestern plants using high sulfur coals, plants with limited sludge
disposal area, etc.). However, there may not be sufficient vendor
capacity to supply only double alkali or regenerable systems to meet the
total national demand anticipated for FGD systems resulting from
the proposed standards. A survey of vendor capacity is given in
the report Effects of Alternative New Source Performance Standards on
Flue Gas Desulfurization System Supply and Demand , report number EPA-600/7-
78-033. Because of limited regenerable system vendor capacity and
higher costs, lime and limestone systems, which have lower performance
capabilities, will have wide application and must be relied upon as a
basis for a nationally applicable NSPS.
FGD systems capable of attaining 92 percent mean SO 2 removal efficiency
are currently available from at least 5 to 10 suppliers of lime, limestone,
regenerable, and other FGD equipment. Numerous demonstrations of technology
have provided design information for such systems and have demonstrated
high SO 2 removal efficiency on full-scale as well as pilot-scale systems.
4.2.5.2 Minimum FGD Performance
FGD systems should be designed for the most severe operating conditions
anticipated, such as maximum boiler load and maximum SO 2 inlet concen-
trations. The SO 2 removal efficiency is largely fixed by the absorber
design, absorbent type, and by the level of operating parameter selected.
Once a mean SO 2 percent removal capability has been selected for design
4-16

-------
into the FGD system, consideration should be given to the minimum 24—
hour average performance because a daily (24-hour) basis has been selected
for averaging emissions or FGD control efficiency for determining
compliance. Thus, compliance during routine operation will depend upon
maintaining a minimum percent SO 2 reduction each day of power plant
operation and will depend on the control of variation in 24-hour average
EGO SO 2 removal efficiency. The EGO system should achieve a certain
minimum SO 2 removal (the proposed standard) at maximum boiler load and
SO 2 inlet concentration and be able to operate at higher absorption
efficiencies for all other operating conditions less severe than peak
boiler load or peak SO 2 inlet concentration.
Once the design capability has been incorporated, process controls
must be applied to exercise the capability in response to changing
inputs (coal sulfur, boiler load, etc.) and to maintain critical process
operating parameters (pH, L/G, etc.) within prescribed ranges. The FGD
system cannot be adjusted to control emissions under the worst operating
conditions (maximum inlet SO 2 concentration and boiler load) and left to
function by itself. For example, if the inlet SO 2 concentration fell
and the lime addition rate to the system was not reduced, pH would rise
above safe limits and the system would develop scale.
The variation in FGD efficiency was evaluated by EPA using continuous
monitoring data from full-scale boilers equipped with FGD. These data
were reduced to 24-hour averages, and probability distributions of 24-
hour average FGD efficiency were plotted for each facility (Figure 4—1).
The geometric standard deviation (variability) for each distribution
4-17

-------
(Table 4-4) was found to be roughly the same (1.014-1.060 percent) for
each facility; however, the Cane Run and Bruce Mansfield (test I) data
did show higher geometric standard deviations (1.057-1.060 percent).
This may be due to factors known to exist at these facilities during the
tests. At Cane Run, the boiler is used as a peaking unit and the data
base contains results measured during load changes. The data bases for
the other facilities also contain results measured during load changes,
but the load changes were not as frequent as at Cane Run. At Bruce
Mansfield during test I, pH instrumentation problems were experienced.
The pH instrumentation was improved prior to test U. Before the pH
instruments were repaired and relocated (test I), pH control of the FGD
process was hampered.
Table 4-4. SO REMOVAL STATISTICS
(24-hour A ERAGING PERIOD)
Geometric
No. of Mean FGD efficiency standard deviation
Site points ( percent SO 2 removal) ( percent SO 2 removal )
Cane Run 89 89.8 1.060
Bruce Mansfield —
Test I 20 81.4 1.057
Bruce Mansfield -
Test II 11 85.3 1.029
Eddystone 8 96.8 1.014
Mitchell 25 90.0 1.015
4-18

-------
A “line of improved performance” is projected on Figure 4-1 using
approximately the same geometric standard deviation recorded for the
four facilities tested. The projected performance of an improved FGD
system with 92 percent mean SO 2 removal is predicted to range from 75 to
97+ percent (24-hour average basis). The probability distribution
projects that less than 10 percent of the 24-hour averages (less than
3 days per month) would fall below 85 percent SO 2 removal and that the
probability of any 24-hour average falling below 75 percent is only 0.01
percent (about once very 4 years assuming 70 percent boiler operability).
Less than 75 percent control can easily be avoided by reducing boiler
load and shifting electric generation load to another unit or by using
improved process controls discussed in section 4.2.6.
4.2.6 EGO Performance Control Improvements
The monitoring data for four full-scale EGO systems tested by
continuous monitoring systems have shown that the EGO SO 2 removal efficiency
variability can be controlled, but not eliminated, by manual operation
of the system. Additional automatic control systems can be designed
into new FGD controlled steam generators to assist in the control of
effects of inlet SO 2 concentration variation, on FGD efficiency. The
controls currently used by EGO system designs in the United States use
pH instruments which do not detect these concentration variations before
the slurry reaction tank chemistry has been upset. Instrumentation at
the FGD absorber inlet is needed to keep the FGD system in balance when
inlet conditions change. A signal from the inlet SO 2 continuous monitoring
equipment can be used to produce more rapid action by the slurry tank
lime feeders for adding absorbent prior to a significant depletion of
the absorbent in the reaction tank. Currently used pH controls do not add
4-19

-------
absorbent until a depletion has occurred. Additional monitoring of
inlet gas velocity for automatic control of the mass rate of slurry
pumped to the absorber (L/G ratio control) would also contribute to FGD
efficiency stability. More automated controls of this type that anticipate
impacts of changes in SO 2 inlet concentration and changes in gas flow due
to boiler load changes have the potential to reduce the variation in
24-hour average 502 absorber efficiency and to prevent process upsets
that can cause scaling. In Japan, these automatic process controls are
currently in use (see Sulfur Oxides Control Technology in Japan , Maxwell,
Elder, and Morasky, June 30, 1978). Measurement of flue gas volume and
SO? concentration at the scrubber inlet are used to automatically
determine slurry make-up volume requirements. Fine tuning of the make-up
feed rate is maintained by pH control. The FGD process chemistry stays
in balance and the FGD removal efficiency is stabilized.
Steadier FGD operating conditions in Japan have also resulted from
using coal blending to reduce variations in the sulfur content of the coal
fuel. For example, at the Takehara station, blending of coal controls
inlet SO 2 between 1150 and 1650 ppm. The blending reduces the burden
upon the automatic control systems in adjusting the FGD operation to inlet
SO 2 concentration conditions. Blending can reduce variations in coal sulfur
content in all coal supplies, even shipments from a single mine. Blending
of coal at a cost of about $7 per ton (1980 dollars) can produce steadier
FGD performance and would allow fewer adjustments to operating controls
(pH, lime rates, liquor rates, etc.) to maintain scrubbing efficiency.
In addition to more constant SO 2 control efficiencies, greater FGD
availability due to less scaling that results from process chemistry upsets
would be experienced. Availabilities of Japanese FGD units of 97—100
percent are reported.
4-20

-------
4.2.7 FGD Control at Small Plants
The JaDanese units show the potential for improved FGD operating
control with blended coal and automated instrumentation. With small-
size power plants, these technologies may be needed to attain the proposed
percent SO 2 reduction standard. Various coal types have different ranges
of sulfur content and in smaller power plants the coal sulfur variability
can be magnified. The opportunity for coal sulfur variations to average
themselves away is less in small plants because they fire smaller quantities
of coal in a 24-hour compliance period.
A major U. S. coal company supplied data on Wyoming low-sulfur raw
coal to EPA. In their judgment, the data typify Wyoming coals. Assuming
5 percent sulfur retention in the ash, this coal would average 0.84 lb
S0 2 /million Btu for the total amount of coal sampled. The data received
were for 306 lots of 600 tons, which were combined into lots weighing
1200 tons, and then further combined into lot sizes of 5000 tons and
10,000 tons. For each lot size group, the mean, standard deviation, and
relative standard deviation (RSD) were determined. One important feature
of their data is the small lot sizes, which enabled extension of the RSD
curve without extrapolation (Figure 4-2).
A second major U. S. coal company supplied data on Illinois high-
sulfur washed coals. These data were for 140 lots of coal varying in
size from 1000 tons to 14,000 tons, and included lot size, sulfur content,
and heating value for each lot. With this information, the RSD versus
lot size for Midwestern washed coal was determined (Figure 4-2). The
smallest lot sizes used in developing this curve were those in the
interval 1200 to 1299 tons.
A 25 MW plant burns an average of only 264 tons of coal in a 24-
hour period, but the smallest lot size for which data were obtained was
4-21

-------
COAL LOT SIZE tons
Figure 4-2. Projected RSD for small plants.
4-22
2O
12
I-
C d ,
uJ
I-
_18
w
96 240
480
960 2400 4800

-------
600 tons (about 60 MW plant capacity). For illustrative purposes, both
RSD curves were extrapolated for small lot sizes (Figure 4-2). The
dotted portions of the curves in this figure are not based on data for
the tonnages they represent.
The proposed percent SO 2 reduction standard can be achieved on
many large power plants without coal blending and improved process
control instrumentation to reduce EGO efficiency variation. However,
in small plants or in plants which are firing certain coal types, closer
control of EGO operations may require application of technologies to
improve absorber efficiency control (e.g., coal blending and improved
process controls). Technologies to attain higher mean SO 2 percent
removal, (e.g., coal washing plus FGD or regenerable FGD systems) are
also available. An appropriate combination of these technologies will
enable any power plant regardless of size or coal type to attain the
percent SO 2 reduction standard.
4.3 DRY SO 2 CONTROL SYSTEMS
Dry SO 2 scrubbing systems have made significant advancements in the
past few years and many variations of the system have been developed.
One of the more effective systems incorporates the use of a spray dryer
and baghouse. In this system a spray dryer (similar to a wet SO 2
scrubber) is used with lime, soda ash or other reactants to scrub SO 2
from the flue gases. Because of the very minimal use of water in the
Spray dryer (by design) the flue gas is quickly reheated from a saturated
wet gas into a dry particulate-laden gas stream before leaving the spray
dryer. Following the spray dryer a baghouse is used to collect all
particulate matter (including so 2 reactants). A significant portion of
the overall 502 removal by the dry control system takes place on the
filter cake in the baghouse.
4-23

-------
Spray drying has been tested at pilot plants, and it appears
capable of achieving 85 percent SO 2 removal (24- hour average basis)
with lime, soda ash, and other reactants. The system is principally
limited by economics to coals with less than 1.5 percent sulfur if lime
is used. For low sulfur coal applications, spray dryers appear to have
several advantages in comparison to wet scrubbers. Experience with
spray dryers for SO 2 control has been limited to several pilot plants,
but these demonstrations have been sufficiently convincing for some
utility companies. Full size spray drying units for commercial power
plants have been ordered and will initiate operation in the early 1980’s.
Some of the potential advantages currently claimed for spray dryer,
dry SO 2 control systems are:
1. A dry SO 2 scrubber does not require certain equipment required for
disposal of wet sludge (thickeners, centrifuges and mixers).
2. The dry system is expected to use less costly low carbon steel for
most of the equipment because the stack gases are not cooled as much as
with a wet scrubber. Less fan corrosion and imbalance are expected.
3. The dry system may have more flexibility of operation. Dry scrubber
slurry feed rates can be adjusted with less concern for pH control and
dry scrubbers are expected to have more turn-down capability.
4. The capital cost of the dry system is projected to be $20 to $40
per kilowatt less than a comparable wet scrubbing system with full control.
5. A dry SO 2 control vendor has estimated that about one-half as many
operators and one-third as many maintenance personnel will be required
for the dry system as compared to wet scrubbers.
4-24

-------
6. The dry system is expected to have only 50 percent of the energy
requirements of the wet system. No flue gas reheat is necessary.
7. Annualized costs of particulate matter and SO 2 control may be reduced
about 15 percent with dry controls in comparison with wet scrubbers
(full control).
4.4 NONCONTINENTAL AREAS
Several island areas will be affected by the proposed regulations:
Hawaii, Guam, American Samoa, the Northern Marianas, Puerto Rico, and
the Virgin Islands. A small amount of electric generating capacity will
be constructed to meet future needs in these noncontinental areas. In
Hawaii, several small units will be constructed on separate islands to
provide an additional 1141 MW capacity during the period 1980 to 2000.
Electric generating units in these islands now fire low sulfur oil
or waste materials such as bagasse from their sugar industry. In
Hawaii, the Virgin Islands, and municipal areas in Puerto Rico, fuel oil
containing less than 0.5 percent sulfur is used to comply with State
Implementation Plans. New units are also projected to fire low sulfur
oil notwithstanding the coal conversion legislation currently being
considered by Congress. The coal conversion bill has been drafted with
provisions that would exempt Hawaii and other noncontinental areas from
requirements to use coal in new electric generating units.
The application of the same proposed standard for all facilities in
the United States regardless of geographic location would require FGD
systems to be installed on oil—fired units in these islands. Although
use of FGD on oil-fired units in Japan has been well-demonstrated,
several unique features of U.S. noncontinental areas distinguish them
from larger islands such as Japan as well as from the U.S. mainland.
4-25

-------
Limited land area is available for FGD sludge disposal, and the ground—
water level on many islands is near the surface. Sludge disposal ponds
must be lined and must be shallow. Shallow ponds require up to five
times more acreage than comparable ponds on the U.S. mainland. Because
of the large size pond required, high cost of land and the need for a pond
liner, a sludge pond on the island of Oahu, State of Hawaii, is estimated
to cost up to $12 million for a 141-MW unit. In contrast, a sludge pond
for a 1414114 unit located on the U.S. mainland is projected to cost only
$132,000. The increased risk of contaminating ground water levels close
to the surface and the much greater cost of sludge ponds detract from
the technical and economic feasibility of using throw-away absorbents in
FGD systems in these islands.
Alternative disposal methods were also considered. An EPA study
of ocean disposal of FGD sludge identified several promising options,
but disposal of sulfite-rich FGD wastes was not considered acceptable
until more definitive data are available. A second alternative considered
was to produce a usable by-product rather than a throw-away sludge, as
is typically done in Japan. This would be done either by forced oxidation
conversion of sludge to gypsum or construction of a regenerable FGD
system that produces either sulfuric acid or a solid sulfur product.
There is no significant industrial intrastructure on these islands that
could make use of FGD by-products. Although these FGD by-product
materials are potentially usable on the U.S. mainland, shipment to the
mainland would be costly.
Applying FGD to units -in these islands would be more expensive than
similar applications on the mainland, and unique environmental problems
would be encountered. In addition to potential groundwater contamination,
reliance upon FGD controls in these islands could create short-term
4-26

-------
excursions of high SO 2 emissions be cause there would be little opportu-
nity on a small island to shift electric generating load away from a
unit with a malfunctioning FGD system to a second generating source.
Use of low-sulfur oil instead of FGD would produce low and constant
emissions, would be much less costly, and would avoid potential ground
water contamination. The Hawaiian Electric Company has projected that
application of FGD on the island of Oahu alone would increase their
annual revenue requirements by $56 million.
4-27

-------
4.5 PERFORMANCE TESTING
4.5.1 Particulate Matter
Compliance with the particulate matter standard would be determined
by using EPA Method 5 operated at a filter temperature up to 160°C. EPA
Method 3 would be used to determine oxygen or carbon dioxide concentrations.
These concentration measurements would then be used to compute particulate
emission in units of the standard as specified in proposed EPA Method
19.
EPA relies primarily upon Method 5 for gathering a consistent data
base for particulate matter standards. Method 5 meets the above criteria
by providing detailed sampling methodology and includes an out-of—stack
filter to facilitate temperature control. The latter is needed to
define particulate matter on a common basis since it is a function of
temperature and is not an absolute quantity. If temperature is not
controlled, and/or if the effect of temperature upon particulate formation
is unknown, the effect on an emission control limitation for particulate
matter may be variable and unpredictable.
As applied to the steam generator new source standard, EPA Method 5
originally employed a filter system located out-of-stack and operated at
a temperature of 120°C. In October 1975, EPA revised the performance
test requirements for steam generators to allow operation of the filter
system at temperatures up to 160°C. The purpose of this revision was to
prevent collection of condensable gaseous compounds, which would not be
coAtrollable by dry control devices operating at stack temperatures
found at modern boilers.
In February 1978, EPA promulgated Method 17 for the determination
of particulate from sources when specified in applicable subparts.
4-28

-------
Method 17 uses an in-stack filter and, therefore, collects particulate
matter at the temperature of the stack gas. Since Method 17 measures
particulate matter at the stack gas temperature, it is considered to
yield comparable results in comparison to EPA Method 5 at stack temperatures
of less than 160cC.. Method 17 is, therefore, acceptable to demonstrate
compliance with the steam generator standard. The method allows a flexible
connection between the probe and the sample box and thereby has the ad-
vantage of eliminating traversing with the sample box. The method also
eliminates possible imprecision in recovering samples from long stainless
steel probes used in very large-diameter stacks. Since Method 17 is not
applicable for stack gases containing saturated water vapor, it
is not applicable to stacks following wet scrubber systems unless demisting
and reheat treatment are sufficient to raise the stack gas above its
dew-point.
4.5.2 Sulfur Dioxide and Nitrogen Oxides Standards
4.5.2.1 Compliance Tests
Compliance with the proposed SO 2 and NO standards would be based
upon the data obtained from a continuous monitoring system. If FGD were
used for 502 control) continuous SO 2 emission monitors would be required
both upstream and downstream of the FGD system to determine compliance
with the 85 percent SO 2 reduction requirement As an option, compliance
with the SO 2 standards could be determined using both an “as fired’ fuel
sampler to determine the sulfur content and heating value of the fuel
fired to the boiler and a continuous SO 2 emission monitor after the FGD
system to measure SO 2 emissions discharged into the atmosphere. In
addition to crediting the SO 2 removed by the FGD system, this option
would provide credit for sulfur removed by coal pulverizers and in the
4-29

-------
bottom ash and fly ash. The SO 2 percent reduction requirement and
emission limitation would both be averaged over a 24-hour (daily)
period. If fuel is treated prior to combustion to reduce SO 2 emissions,
a sulfur removal credit would also be allowed.
Performance testing to determine compliance with the percent reduc-
tion requirements for NO would not be required. An affected facility
would be assumed to be in compliance provided the facility is in compliance
with the applicable NO emission limitation.
For SO 2 and NO continuous monitoring, EPA requires the pollutant
and diluent gas analyzer to measure the gas concentrations at least once
every 15 minutes (Subpart A, General Provision, 40 CFR 60.13(n)). For the
purposes of this standard, the 15-minute measurements would be averaged
each hour. The hourly average concentrations are recorded (printed out)
and then averaged each 24-hour period to determine compliance in accordance
with the procedures in EPA Method 19. A minimum of 23, one-hour averages
are required to calculate the daily average emission rates.
Emission determinations are not required 1-hour per day to allow
for daily zero, span, calibration checks, and adjustments to the continuous
monitoring system and for an additional 8 hours per month to allow for
routine service. Fewer than 23, 1-hour emission determinations will be
averaged on days containing periods of plant startup, shutdown, inoperation,
and emergency conditions. One-hour average emissions would be determined
during each of these periods except plant inoperation, but would not be
included in the daily arithmetic emission average for determining
compliance with the proposed SO 2 and NO standards. Thus, only the data
averaging method specified within EPA Method 19 may be used to determine
compliance.
When the compliance monitoring system fails to operate properly,
4-30

-------
the same source owner or operator would obtain emissions data by
(1) operation of a second monitoring system or (2) conducting manual tests
using EPA reference methods. if a second monitoring system is used, the
source owner would have to keep the second system in operation at all
times. When conducting the manual tests, the source owner would have
to keep trained manpower available to collect the samples.
EPA requires continuous monitors to meet performance specifications
promulgated in 40 CFR 60, Appendix B. Since compliance with the SO 2 and
NO standards would be determined by continuous monitors, EPA is currently
developing additional quality assurance procedures. These procedures would
not change the present performance specifications, but would provide
additional periodic field tests to assure the accuracy of the monitoring
data. Sections within the proposed regulations are being reserved for
the future addition of these requirements. This matter should not pose a
problem since new sources affected by the revised standards are not
expected to begin operation until 1984.
4.5.2.2 Fuel Pretreatment Credits
Pretreatment of a fuel to remove sulfur or increase heat content
would be credited toward an 85 iercent SO 2 reduction requirement. For
example, by pretreatment of a fuel with 1000 ng/J potential SO 2 emissions
(e.g., about 2.3% S coal) to remove 25 percent of the sulfur, the FGD
system SO 2 removal requirement would be reduced to 80 percent (750 ng/J
reduced to 150 ng/J). An 85 percent emission reduction (1000 ngIJ
reduced to 150 ng/J) would be necessary if the fuel were burned untreated.
Table 4-5 shows the amount of SO 2 removal that would be needed in conjunction
with fuel pretreatment to achieve an overall 85 percent SO 2 reduction.
4-31

-------
Table 4-5. RELATIONSHIPS BENEEN FUEL PRETREATMENT AND POST
COMBUSTION CONTROL FOR REMOVAL OF 85% SULFUR DIOXIDE
Fuel pretreatment credits would be
fuel and increase in fuel heat content.
or processes for which credits would be
1. Physical coal cleaning.
2. Solvent refining of coal.
3. Liquefaction of coal.
4. Claus processing for removal of sulfur from gasified coal.
5. Hydrotreating of oils by refineries.
Rotary breakers used to separate rock and other material from raw
coal prior to processing or shipment by a coal preparation plant are
considered an integral part of the coal mining process and this use would
not be considered as fuel pretreatment. Sampling of raw input coal to
determine fuel credits would be performed after the coal has passed
through a rotary breaker or a course screen rather than from the tipple.
Pretreatment
SO 2 removal %
0 (no fuel pretreatment)
10
20
30
40
50
60
70
80
85
Post combustion SO 2
removal efficiency, %
85
83
82
79
75
70
63
50
25
-0-
(no post combustion
SO 2 control
removal of sulfur from
of the type of equipment
given for
Examples
given are:
4-32

-------
The proposed standard would not require that fuel be pretreated
before firing but, as indicated, would allow credit for pretreatment.
The amount of sulfur removed by a fuel pretreatment process would be
determined by applying Method 19 (Appendix A). The owner or operator of
the electric utility who would use the credit would be responsible for
insuring that the Method 19 procedures are followed in determining SO 2
removal credit for pretreatment.
The fuel monitoring procedures proposed under Method 19 could be
made part of the fuel supply contract. Because of the small impact of
coal pretreatment on FGD performance, the uniformity of highly processed
fuels (such as solvent-refined coal), and the desirability to develop a
viable fuel pretreatment crediting procedure, a 90-day quarterly average
is being proposed for determining fuel cleaning credits. Through use of
Method 19, the fuel supplier could provide a certificate of credits to
the coal purchaser with each lot of fuel delivered or for all lots delivered
for the calendar quarter. The certificates would have to show (1) sulfur
analysis of coal input and output from the preparation plant (ng/J lbs/million
Btu), (2) quantity delivered, (3) heat content, and (4) calculation of the
pretreatment credit. For example, if the analysis of coal to and from
the preparation plant were 5 percent sulfur (3,800 ng/J) and 4 percent
sulfur (3,000 ng/J), respectively, on a dry basis,a 20 percent pretreatment
credit would reduce FOG removal requirements from 85 percent to 83 percent.
After the utility company receives the fuel and the pretreatment
certificate, a summary of credits for all fuel received in a calendar
quarter would be prepared for pretreatment credit. The credits for all
fuel deliveries would be averaged on a weighted-Btu basis to determine
an average credit for the quarter. For example, if half of the heating
4-33

-------
value of all fuels received were cleaned coal with a 30 percent pre-
treatment credit and half were for coal with a 20 percent pretreatment
credit, the average credit for the quarter would be 25 percent.
After the average pretreatment credit is determined for a quarter
(e.g., the first quarter of a year), it would be applied toward deter-
mining compliance with the 85 percent removal standard for all 24-hour
(daily) periods during the next quarter (e.g., the second quarter of a
year). Thus, the average fuel cleaning credit from the first quarter
would be considered together with each 24-hour (daily) average percent
removal achieved by the FGD system during the second quarter and used to
determine compliance with the 85 percent overall SO 2 removal requirement.
For example, a 25 percent fuel cleaning credit for a first quarter would
reduce the SO 2 reduction requirement to 80 percent for all 24-hour daily
periods during the following quarters, except for three days per month
where 67 percent rather than 75 percent would be allowed.
The approach of determining the credit on the basis of the previous
quarter’s fuel receipts would allow the utility to know before the fact
the amount of SO 2 removal that must be attained by the FGD system.
Without this procedure, the utility owner could not determine until the
end of a quarter whether he had used sufficient FGD control to comply
with the emission standard.
4.5.3 Qpacity
Compliance with opacity standards could be determined at any time
by visual observations using EPA Method 9. Except during startups,
shutdowns, and emergency energy conditions, all data from visual observations
would be used for determining compliance with the opacity standard.
4-34

-------
A continuous monitoring system for opacity would be required in the
stack except when only gaseous fuels are fired. The opacity data from
the continuous monitor would not be used to determine compliance with
the opacity standard. It would be used to assist in assuring that the
particulate matter control system is properly operated and maintained at
levels observed at the time the particulate matter performance test is
conducted.
If interference with opacity monitoring is expected because of
condensed water vapor in the stack, the monitor would be located upstream
of the FGD system. If interference with the opacity measurements is
expected upstream of the FGD system, the opacity monitor would not be
required and operating parameters of the particulate matter control
device would be monitored.
4.6 FGD COMPLIANCE
FGD systems are composed of FGD modules, each of which is separate
scrubbing system. Because FGD modules are not generally manufactured in
sizes over 125-MW capacity, large power plants use multiple EGO modules
in parallel.
EPA has analyzed the availability of EGO modules and systems. In
PEDCo’s report Flue Gas Desulfurization System Capabilities for Coal—Fired
Steam Generators , EPA-600/7-78-032b, March 1978, it was stated that with
best operation and maintenance FGD modules would be available for service
up to 90 percent of the time. A review of the availability of other mature
coal-fired electric generating components was also performed for comparative
purposes. In Radian’s report The Effect of Flue Gas Desulfurization
Availability on Electric Utilities , EPA-600/7—78—O3lb, March 1978, components
such as boilers, and turbines were reported to have average availabilities
between 80 and 97 percent.
4-35

-------
When FGD modules, even those with 90 percent availability, are
integrated into an FGD system, the probability that all modules in the
system will be available simultaneously diminishes in proportion to the
number of modules. For example, an FGD system consisting of four modules
without spares could have an availability of only 66 percent =
0.656) if all modules were operated simultaneously.
To address this problem, spare FGD modules may be needed in systems
where all FGD modules are often called upon to operate. Even when high
FGD module availabilities (90%) are attained, the FGD module will be
out of service a significant amount of time (10%) for regularly scheduled
maintenance, forced outages, and repairs. When the power plant is
operated at maximum capacity, the maintenance problem is compounded
because all modules are needed to treat the flue gas. To maintain high
FGD system availability, one or more spare FGD modules may be necessary.
The Radian report concludes that a base-loaded power plant cannot meet
consumer demand without spare FGD module capability.
EPA has concluded that compliance with the proposed standards using
FGD control systems could be attained by (1) using 90 percent available
FGD modules, (2) using spare FGD modules in base-loaded units, and (3)
providing a limited allowance for emergency conditions when the electric
generating plant output must be continued to maintain continuity of
electric service.
Although a high-quality routine maintenance program will keep FGD
modules 90 percent available, the amount of time for such maintenance
can be considerable, even continuous. With spares, a module can be
rotated out of operation for maintenance even when the power plant is
operating at full electrical load. Switching FGD modules need not have
4-36

-------
a major impact on plant operation. At reduced electrical loads, all FGD
modules may not be needed for SO 2 control and some modules could undergo
maintenance at such times or at times when the entire plant is taken out
of service for repair of non-FGD related components.
The Radian report states that mature, coal—fired electric generating
plants are reported to have average availabilities of only 70 to 77 percent,
even when FGD systems are not used. Thus, a considerable amount of time
would also be available for FGD module maintenance when the power plant
is not being operated.
If a power plant is base loaded and must therefore operate with a
minimum of interruptions, spare FGD modules can keep the availability of
the total FGD system of modules above 90 percent. EPA has projected
that one spare may be needed for plants up to 500 MW and two spares may
be needed for larger plants up to 1000 MW.
Even with a good maintenance program and use of spare FGD modules,
complete EGO control may not be maintained for a portion of the plant
operating hours. At these times, the proposed regulations would require
that the electric generating load be shifted when possible to another
electric generating plant. This procedure is necessary to minimize by-
passing of uncontrolled SO 2 emissions to the atmosphere. Because frequent
load shifting is not economical, it also provides an incentive for
maintaining the FGD control system. The uncontrolled, incremental SO 2
emissions that would result from shifting the load is expected to be no
greater than the SO 2 emissions that would result from bypassing.
Discharging untreated flue gas to the atmosphere (bypassing of
emissions) because of SO 2 control system malfunctions would be allowed
4-37

-------
only during emergency conditions, and even then SO 2 emissions from the
facility would have to be minimized by operating all available FGD
modules to the maximum extent feasible. Emergency conditions are
considered to be periods when a power plant and its associated utility
system are being operated at full operable capacity except for an amount
of capacity equal to the largest single steam generator in the system.
Because it may be necessary to meet consumer electric demand, would
reduce SO 2 emissions, and would be economically reasonable in large
power plants, the emergency condition provisions would apply only
to (1) plants that have installed at least one spare FGD module or to
(2) small plants (< 125 MW capacity). Small plants would need only one
FGD module (no spare) to attain 90 percent availability, and use of a
spare module on a small plant would be economically unattractive. The
added cost of a spare module has been evaluated in PEDCo’s report,
Particulate and Sulfur Oxides Emission Control Costs for Large Coal-Fired
Boilers , EPA-450/3-78-007, February 1978.
The emergency condition provisions are necessary to maintain the
capability to meet electric demand when adequate generating reserves
are not available. A minimal amount of spinning reserves (not greater
than the largest unit in the system) must be kept separate from the
load shifting procedures to prevent “blackouts” or serious impairment
of service continuity.
During periods of emergency, the derating and load shift procedures
cannot be implemented without seriously jeopardizing the electric
reliability of the utility system. Interconnection capacity (for
purchasing power) and spinning reserve capacity must be kept available
to handle sudden losses of generating capacity within the utility system.
This amount of spinning reserve can typically vary from 75 to 150 percent
4-38

-------
of the capacity of the largest single unit within the utility company
depending upon the characteristics of individual company systems.
The interconnection capacity represents the amount of power that
can be brought into the utility from neighboring companies. This amount
can be limited by the load carrying capacity of switch gear, transmission
lines, etc., used to interconnect the utility company’s grid with that
of neighboring companies. The interconnected system is automatically
the first source of replacement power in the event of a sudden electric
generation loss. Power is immediately pulled into the utility system
through the interconnections, which spreads much of the lost load over
many electric generators in several other utility systems. A small,
noncritical frequency drop is typically experienced over a wide area
rather than a severe service problem or a blackout in a localized area.
Because the interconnection capacity is so important for immediate
reaction to an electric generation loss, a minimum amount of interconnect
capacity must be kept available. To maintain or restore this minimum,
energency power brought into the utility through the interconnection
should be reduced within 10 minutes and entirely replaced as soon as
possible by loading spinning reserves or bringing nonspinning reserves
into operation. An individual unit in spinning reserve, already synchronized
and partially loaded to the electric system, can immediately start
assuming additional load; however, a unit cannot be instantly loaded.
The rate at which a unit can be loaded (ramp rate) is not the same for
all units. Because of the time needed to bring spinning reserves into
operation, interconnected capacity must be kept available and the spinning
reserves are not usually concentrated in any one unit to minimize response
time to satisfy load demands.
4-39

-------
In addition to the simple replacement of power generation described
above, there are other factors that can at times be critical to the
reliable and efficient operation of a utility system. Because localized
power factors, transmission line capacity, and other problems contribute
to making the overall operation very complex, an amount of capacity
defined as “system emergency reserves” should not be involved in the
load shifting procedures that are to be implemented to avoid bypassing
emissions around an FGD module. These system emergency reserves which
equal the rated capacity of the largest single electric generating unit
within the electric utility system may be distributed among several
electric generating units within the utility company. This procedure is
usually necessary (1) to keep reserves available in each local service
area and (2) to minimize the amount of time needed to get the emergency
reserves into full operation because power increase is typically limited
to 2- to 15-MW/mm per unit. By bringinq several units into operation at
once, more capacity can be brought into ooeration per minute during an
emergency.
Emergency condition provisions would not prohibit the installation
of bypasses around an FGD system; however, the use of these bypasses can
place the owner or operator in violation of the proposed standards
unless their use is strictly limited to periods of emergency conditions
when only system emergency reserves are available to prevent blackouts.
At any time the utility has available reserves (either cold units or
spinning units) in excess of emergency reserves, the utility is required
to bring them into operation to replace a malfunctioning unit and to
simultaneously purchase power in the interim period of time it takes to
bring available reserves on-line. During emergency conditions or the
4-40

-------
time needed to bring a spare module on line, SO 2 continuous monitoring
emission data would not be included in any 24-hour performance test
average. Although these periods would not be included in performance
tests, they would be subject to requirements under 40 CFR 60.11(d),
which require the:maximuni control of SO 2 emissions feasible even under
adverse conditions. Load shifts to other generating units will usually
be feasible with purchase power being used in the interim period when
units are being brought into service.
Normal procedure would not include the routine use of emission control
system bypass during (1) the time a spare FGD module is being dampered into
the system to replace a malfunctioning module that is attaining substandard
SO 2 control, (2) the time a module is being removed for scheduled maintenance
service, or (3) the time a module is being brought on line because of an
increase in boiler load. A spare FGD module should be brought on line
before other modules are removed from service or before load changes are
made. Bypass of emissions would only be necessary during emergency
conditions. During startup and shutdown all necessary modules should be
brought on line to minimize emissions prior to system changes. These
procedures may cause utilities to change some operating procedures including
those for dispatching load, but will not significantly affect the operation
of the electric system or the ability of the utility company to maintain
service continuity.
4-41

-------
4.7 COAL IMPACTS
4.7.1 Production and Reserves
The effect of the proposed standards on coal production has been
projected by EPA for the period 1985 to 1995 (Section 3). In comparison
to coal production under the current NSPS, an 85 percent reduction
requirement, a 0.2 lb/million Btu floor, and a 1.2 lb/million Btu ceiling
are expected to have almost no impact on total national coal production
(<2 percent decrease) in comparison to the current NSPS (Table 3-4).
Regional coal production patterns may shift, however, because of the
increased utilization of higher sulfur coal. Midwestern coal production
is projected to increase up to 15 percent more than it would have had
the current L JSPS remained in effect for all new units. About 6 percent
decrease in coal production in the Northern Great Plains area and
virtually no change in other Western or Appalachian coal production are
expected in 1990. Shipments of Western coal to the Midwest and areas
further east are expected to decline by about 20 percent. Even with
FGD controls, shipment of Western coal to the midwest and other areas
will not be totally stopped, but will be reduced. Under the proposed
standard, the amount of Western coal shipped east is expected to be less
than 10 percent of total national coal production.
A 85 percent reduction requirement would not affect national coal
production, but an SO 2 emission limitation (ceiling) of 520 ng/J (1.2
lb/million 6th) would restrict the sulfur content of coal that can be
utilized even when FGD controls are applied. Although Midwestern coal
production does not decline in comparison with the current NSPS, mining
of many coal desposits would be restricted by a 1.2 lb/million Btu
emission limitation.
4-42

-------
Continuous monitoring of SO 2 emissions would require strict, daily
compliance. The composition of coal varies appreciably even when all
shipments are from the same mine (see sections 4.2.4 and 4.2.7). To
ensure compliance with the ceiling, SO 2 emissions should be based on minimum
FGD performance (24-hr basis) and maximum daily average coal sulfur
content since these conditions may coincide.
EPA has analyzed the impact on national and regional coal reserves
of 24-hour minimum FGD efficiency conditions coinciding with peaking
coal sulfur content. Minimum FGD efficiencies were projected form the
line of improved performance in Figure 4-1, and coal peak sulfur contents
were projected using 15 percent RSD. Two options for selection of the
SO 2 emission limitation were analyzed: 1.2 lb/million Btu with 3 exemptions
per month (option 1) and with no exemptions (option 2). With each of
these options, the mean sulfur content of the coal that could be used to
comply with the ceiling was projected and used to compute (1) future
coal production and (2) the amount of coal reserves that would be restricted
by the ceiling. The sulfur content of the coal that would comply with
Option 2 is the same as would be required by a 0.8 lb/million Btu ceiling
with exemptions and the coal production impacts would. also be the same.
The 3-day exemption used in option 1 is based upon variation in
FGD performance that would be expected from a unit not using blended
coal or automated controls that monitor inlet EGO conditions. An EGO
system that achieves 85 percent reduction on a daily basis is expected
to have up to 3 days per month of 75 percent efficiency as well as
several days of very high (95 to 99 percent) performance. An exemption
4-43

-------
in the ceiling would allow relief during periods of 24-hour average
FGD performance between 75 and 85 percent which can coincide with peaks
in coal sulfur content.
An analysis (Section 3) of national and regional coal production in
1985, 1990, and 1995 was performed for each option. There were no
significant differences in total national production with either option
(Table 3-4). The analysis included use of washed, Midwestern coal when
coal washing was necessary to attain compliance with the ceiling.
Sufficient national reserves were available to satisfy national demand
with either option.
On a regional basis, a ceiling without exemptions (option 2) had the
effect of dislocating coal production in the Midwest. Midwestern coal
production in 1995 would be reduced by about 14 percent in comparison
to option 1. The lost Midwestern production was replaced by increased
production in the Northern Great Plains and Aopalachian regions and by
a small production increase in the West. Thus, option 2 would produce
no shortage of coal to meet power plant demand in the Midwest, but would
transfer some mining operations to other areas.
Analysis of Midwestern coal reserves restricted by a ceiling without
exemptions (option 2) verifies the production dislocations projected.
The analysis was performed using coal reserve statistics classified by sulfur
content. In the States of Ohio, and Illinois, and in western Kentucky
60 to 90 percent of reserves would be restricted even if coal cleaning
were used. Without coal cleaning, over 85 percent of reserves in these
States would be restricted by option 2.
This impact upon midwestern coal reserves and production would be
avoided by option 1 (1.2 lb/million Btu with a 3 day-per-month exemption).
4-44

-------
Even with no coal washing use, 40-65 percent of reserves would not be
restricted and option 1 would maximize local coal utilization in
Midwestern coal-fired power plants. The power plant would not be
exempted from controlling SO 2 emissions during the 3-day exemption
because the FGD percent reduction requirement would apply.
This analysis of Midwestern coal reserve impacts assumes 15 percent
RSD coal sulfur content variability. Eastern and Midwestern bituminous
coals often exhibit less coal sulfur variation. In Figure 4—2, washed
Midwestern coal, > 2000 ton lot sizes, has a RSD of 5 percent or less.
High sulfur, Midwestern coal that has a potential problem complying with
the 1.2 lb/million Btu ceiling will probably be washed and the 5 percent
RSD would be more typical than 15 percent. In addition, coal washing is
projected to remove up to 35 percent of the sulfur from Midwestern coal
using conventional technology. When consideration is given to these
factors, well over half of Midwestern coal reserves are expected to be
useable in new power plants subject to the proposed standards.
Use of coal blending would make even more of the Midwestern reserves
useable. Very high sulfur coals can be washed and then blended with
lower sulfur coals, including lower sulfur Midwestern coals, to produce
a fuel that could comply with the ceiling (option 1) when 85% FGD control
is applied.
Another alternative available for utilizing high sulfur coals is to
install an FGD system capable of greater than 85% efficiency. A few
EGO systems (e.g., Weliman-Lord, Magnesium Oxide, Double Alkali, etc.)
have shown this capability during performance tests (section 4.2.5).
With 90% minimum FGD efficiency, a Midwestern coal containing about 6
to 7 percent average sulfur after coal washing would be expected to
comply with option 1. Thus, with (1) coal washing and coal blending
4-45

-------
or (2) coal washing and a 90% efficient FGD system, virtually all
Midwestern coal reserve could be utilized in new power plants.
In addition, solvent refined coal (SRC) plants are being constructed
which would utilize high sulfur (>5% average) Midwestern coal. These
SRC plants are expected to produce a product that will comply with
the option 1 ceiling when fired in new power olants.
4.7.2 Anthracite Coal
The proposed standard would cover anthracite coal in the same
manner as all other coals under a uniform SO 2 control requirement.
Anthracite coal has not been in significant demand since the 1940’s
and may not become an economically feasible utility fuel in the near
future unless its use is encouraged. Although anthracite coal has SO 2
emission characteristics similar to those of low-sulfur Western coal,
it has not become a common utility fuel to date because (1) it costs
approximately 75 percent more to mine in comparison to local bituminous
coal, and (2) anthracite coal-fired power plants are 10 to 15 percent
more expensive to construct than plants firing bituminous coal.
The suggestion has been made that anthracite coal should be subject
to SO 2 control requirements less stringent than those for other coals.
One recommendation included a “flexible interpretation” of the emission
standards for anthracite coal in which no greater FGD efficiency would
be required than that amount (approximately 55 percent control) which
would give an SO 2 emission rate equivalent to that which would result
from using 85 percent FGD control on the local bituminous coal. The recommended
“flexible interpretation” for anthracite coals was based on its “inherently
low-sulfur content and local socioeconomic conditions.” The increased use of
anthracite coal would ostensibly increase strip mining jobs in economically
4-46

-------
depressed areas of Pennsylvania and would assist in correcting acid mine
drainage problems associated with many abandoned mines in the area.
When considering reduced SO 2 control for anthracite coal, it is
appropriate to consider anthracite coal in relation to low—sulfur Western
coals. Anthracite coal does not have SO 2 emission rates significantly
different than those of high-quality low-sulfur Western coal required to
meet the proposed 85 percent reduction requirement. With full or partial
FGD, low—cost Western low-sulfur coal will have a market, but high—cost
anthracite coal may not. The use of low-sulfur anthracite coal in its
local market. Since Congress intended that SO 2 emissions from low—sulfur
Western coals be fully controlled, anthracite coal emissions should be
fully controlled as well. Anthracite coal use could be required by
State regulation if its use is of local importance.
The use of local anthracite coal under the 85 percent SO 2 removal
requirement would result in an aoproximately 4 percent higher cost of
electricity to the consumer than if a local bituminous coal were used
with 85 percent SO 2 removal.
Table 4-6 gives electrical costs associated with partial and full
FGD treatment for anthracite coal.
Table 4-6. EFFECTS OF ANTHRACITE USE
ON COST OF ELECTRICITY
Type of Electricity
local coal FGD, % rate c/KWh Variation, %
Bituminous 85 3.46 baseline
Anthracite 55 3.40 —1.7
Anthracite 85 3.59 +3.8
4-47

-------
TECHNICAL REPORT DATA
(Please read Inrtj ucriong on the reverse before completing)
1. REPORT NO. 2.
EPA-45 0/2-78-O 07a-l
3. RECIPIENTSACCESSIOP .NO.
4. TITLE AND SUBTITLE
Electric Utility Steam Generating Units: Background
Information for Proposed 502 Emission Standards -
Supplement.
5. REPORT DATE
1978
6.PERFORMINGORGANIzATI ONc0 OE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORM$NG ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
.
11.cON -rRAcT/GRANTN O.
.
12. SPONSORING AGENCY NAME AND ADDRESS -—
{)AA for Air Quality Planning and Standards
Office of Air, Noise, and Rad.iation
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND P Rl0p COVERED
l4.SPONSoRIN ENcyCoDE
.
EPA 200/04
1.5. SUPPLEMENTARY NOTES
Revised Standards of Performance for the control of emissions of particulate matter and
nitrogen oxides from electric utility steam generating units are also being proposed.
These standards are supported in separate Background Information documents, numbered
EP -4FiO/2-7 -c1r!5a for nitrogen oxides and EPA-450/2-78-006a for particulates.
lb. J\bstract
Revised Standards of Performance for the control of sulfurdioxide emissions from..
electric utility power plants are being proposed under the authority of section 111
of the Clean Air Act. These standards would apply only to electric util.ity . team
generating units capable of combusting more than 73 MW heat input (250 million Btu)
of fossil fuel and for which construction or modification began on or after the-date
of proposal of the regulations. This document contains background information,
environmental and economic impact assessments, and the rationale for the standards,
as poroosed under 40 CFR Part 60, Subpart Da.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
C. COSATI Field/Group
Air pollution
Air Pollution Control
Pollution control
-
Standards of performance
.
Electric utility power plants
.
Steam generating units
Sulfur dioxide
.
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
21. NO. OF PAGES
Unlimited
Unclassified
—
20. SECURITY CLASS (This page)
22. PRICE -
Unclassified
EPA Form 2220.1 ( .73)

-------