EPA-450/3-80-009b
Guidelines for Determining
Best Available Retrofit
Technology for Coal-Fired
Power Plants and Other
Existing Stationary Facilities
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
November 1980
-------
PREFACE •
Part I provides guidance on identifying those sources to be analyzed
for BART, assessing the anticipated improvement in visibility, conducting
an engineering analysis, and establishing emission limitations for BART.
Part II contains an explicit discussion of the engineering analysis
required by Part I. Part II is primarily for the analysis of fossil
fuel-fired power plants with a generating capacity in excess of 750 MW.
The procedures outlined in Part I, however, may be used for other existing
stationary facilities as well.
n
-------
PART I
GUIDELINE FOR. DETERMINING
BEST AVAILABLE RETROFIT TECHNOLOGY
I-i
-------
PART I. GUIDELINE FOR DETERMINING BEST AVAILABLE RETROFIT TECHNOLOGY
1.0 INTRODUCTION
1.1 Background
1.1.1 Pollutants of Concern
1.1.2 Phased Program
1.2 Identification of a Source Impairing Visibility
1
2
3
,3
2.0 VISIBILITY IMPACT ANALYSIS
2.1 Procedures
2.1.1 Source Information
2.1.2 Emission Rate Estimates
2.2.1 Primary Particulates
2.2.2 Oxides of Nitrogen
2.2.3 Sulfur Dioxide
2.2.4 Other Factors to be Considered
2.3 Engineering Analysis
2.4 Energy Impact
2.4.1 Energy Consumption
2.4.2 Impact on Scarce Fuels
2.4.3 Impact on Locally Available Coal
2.5 Environmental Impact
2.5.1 Air Pollution Impact
2.5.2 Water Impact
2.5.3 Solid Waste Disposal Impact
2.5.4 Irreversible or Irretrievable Committment of
Resources
2.6 Economic Analysis
2.6.1 Direct Costs
2.6.2 Capital Availability
2.6.3 Local Economic Impacts
2.7 Considering Alternative Control Systems
3.0 BART SELECTION
9
9
9
10
2.2 Preliminary Assessment of Improvement in Visibility 11
12
13
14
14
15
15
16
16
16
16
17
17
17
18
18
18
19
20
20
21
I-ii
-------
PART I. GUIDELINE FOR DETERMINING BEST AVAILABLE RETROFIT TECHNOLOGY
1.0 INTRODUCTION
Section 169A of the Clean Air Act, as amended in 1977, calls for
the protection of visibility in mandatory Class I Federal areas
*
where visibility is an important value. Section 169A specifically
requires affected States to remedy existing visibility impairment, in
part, through installation of Best Available Retrofit Technology (BART)
for certain existing stationary facilities.
EPA has promulgated regulations to be codified at 40 CFR 51.300
et seq that implement S169A. BART determinations must be performed on
a case-by-case basis considering such factors as the energy, environmental,
and economic impacts of alternative control systems. This document
provides guidance on identifying those sources to be analyzed for BART,
assessing the anticipated improvement in visibility, conducting an
engineering analysis of available control systems, and establishing
emission limitations for BART. The States must determine emission
limitations for fossil fuel-fired power plants with a total generating
capacity in excess of 750 megawatts pursuant to this guideline, which
'reflects EPA1s conclusion that the controls needed to meet the new source
performance standard (NSPS) for power plants (40 CFR Part 60, Subpart
Da) are generally available to these sources. The procedures outlined
herein are also appropriate for any other existing major stationary
source.
*
These areas are listed in 40 CFR Part 81, Subpart D. From this point
forward, they will be referred to as mandatory Class I Federal areas.
1
-------
1.1 BACKGROUND
Congress was concerned with the impairment of visibility in the
nation's parks and wilderness areas, but it realized remedying existing
impairment in these areas could not be reasonably accomplished overnight.
In order to assure that BART requirements will not be unduly burdensome
or costly, several provisions were included in Section 169A. These are:
(1) BART may not be required by the Administrator for existing
stationary facilities which have been in operation for more than fifteen
years as of August 7, 1977 unless the source was reconstructed after
August 7, 1962.
(2) BART for fossil-fuel fired power plants with a generating
capacity in excess of 750 megawatts must be determined pursuant to EPA
guidelines.
(3) The Administrator may exempt from BART requirements those
sources he determines do not cause or contribute to significant visibility
impairment in a Class I area. This exemption may not apply to fossil-
fuel fired power plants 750 megawatts or greater unless it is demonstrated
to the Administrator that the facility is located at such a distance
>%
from a Class I area as not to cause or contribute to significant visibility
impairment in any such area. Any exemption from BART will be effective
only upon concurrence by the appropriate Federal Land Manager.
(4) In determining BART for any existing stationary facility, the costs
of compliance, the energy and nonair quality environmental impacts of
compliance, any existing pollution control technology in use at the
source, the remaining useful life of the source, and the degree of
improvement in visibility anticipated to result from application of
controls shall be considered.
2
-------
1.1.1. Pollutants of Concern
Visibility impairment is caused by the scattering and absorption of
light by suspended particles and gases. NOp is a light-absorbing gas
and generally causes reddish or yellow-brown atmospheric discoloration
because it absorbs light at the blue end of the spectrum. Primary
particulates and secondary aerosols, formed from emissions of S02 and
NO , scatter light away from and into an observer's line of sight causing
/\
a reduction in visual range and atmospheric discoloration. These three
pollutants (primary particulates, NOX, and S02) have been identified
as the primary contributors to visibility impairment. Detailed background
information can be found in "Protecting Visibility: An EPA Report to
Congress."*
1.1.2 Phased Program
EPA has established a phased approach to visibility impairment. Phase
I focuses on controlling those sources which can currently be identified
as causing visibility impairment. Phase I visibility impairment primarily
includes visible plumes emitted from stacks, and single source haze.
Smoke, dust, or colored gas plumes obscure the sky or horizon. Single
source haze causes a general whitening of the atmosphere and reduction
of clarity of terrain features. Both forms of impairment when "reasonably
attributed" to a source must be regulated under Phase I. As our scientific
and technical understanding of source/impairment relationships improves,
future regulations will address more complex forms of visibility impairment
such as regional haze and urban plumes.
*This report is available through the National Technical Information
Service, 5258 Port Royal, Springfield, Virginia 22161.
-------
This guideline is directed toward Phase I analyses. Although the
number and kind of sources and the type of pollutants included in future
BART analyses may expand, the procedures outlined herein are unlikely to
change substantially. In performing BART analyses the State should be
cognizant of possible future requirements which could be imposed on
sources as a result of later phases of the program. For example, a
major power plant may have a coherent plume caused by primary particulate
emissions which must be analyzed under Phase I, and also contribute to
regional haze through emissions of sulfur dioxide which will be addressed
in later phases. Under Phase I, the source would be analyzed for BART
with respect to TSP because it causes visibility impairment in the form
of a distinct plume. However, since the source may also contribute to a
regional haze, the State would be well advised to also analyze control systems
for SOg to determine if a single system could more efficiently control
both pollutants than two separate systems and to evaluate whether alternative
TSP control systems would be compatible with future application of
control systems designed to control a different pollutant (e.g. S02).
The State is not required to impose S02 controls in this situation.
However, EPA intends at present that physical constraints, incompatible
particulate control, etc. resulting from limitations on Phase I requirements
will not serve as justification for not imposing S02 controls under
Phase II.
-------
PROCEDURES FOR IDENTIFYING SOURCES FOR BART ANALYSIS
UJ
Di
O
NO
NO
•NO-
NO
EXEMPTION
GRANTED EXEMPTION
NO -*« - | APPLICATION J~
YES
YES
YES
YES
REQUIREMENT FILED W/EPA
T EXEMPTION
DENIED
Federal Land Manager identifies
visibility impairment in class I area
State identifies source to which
impairment is "reasonably attributable"
Source in 28 source category with
"potential to emit" 250 tons/yr.
Source not in operation over
15 years as of August 7, 1977
List of sources to be analyzed for
BART provided to Federal Land Manager,
source, and EPA
Source believes it does not cause or
contribute to significant visibility
impairment
BART Analysis
Figure 1
-------
1.2 IDENTIFICATION OF A SOURCE IMPAIRING VISIBILITY
See Figure 1.
If a Federal Land Manager identifies visibility impairment in a
Class I area, the State must first determine, if possible, by visual
observation or any other monitoring technique it deems appropriate, the
existing stationary facility to which the impairment is reasonably
attributable. In other words, for the purposes of Phase I of the visibility
program, States need only identify impairment that can be physically
traced to a source.
States can use visual observation (either ground-based or with an
aircraft) or any other technique it deems appropriate to determine which
source causes the visibility impairment. An "Interim Guidance for
Visibility Monitoring", is available and describes current monitoring
methods. It is available through the National Technical Information
Service. Once the impact of the existing stationary facility on visibility
is identified as being reasonably attributable to that source, the State
must conduct an analysis to determine BART for that particular existing
stationary facility.
The Act limits the requirement for the installation of BART to
those existing stationary facilities which started operation after
August 6, 1962, and were existence as of August 7, 1977. An existing
stationary facility is any source which meets these requirements, is listed
in Table 1, and has a potential to emit 250 tons per year, or more, of
any air pollutant causing or contributing to visibility impairment.
A source which believes it does not cause or contribute to significant
visibility impairment in a Class I area may apply for an exemption from
BART. The exemption application must be submitted to the Administrator
*"Interim Guidance for Visibility Monitoring," U.S. Environmental Protection
Agency EPA-450/2-80-082
6
-------
TABLE 1
"EXISTING STATIONARY FACILITY"
fossil-fuel fired steam electric plants of more than 250 million British
thermal units per hour heat input,
coal cleaning plants (thermal dryers),
kraft pulp mills,.
Portland cement plants,
primary zinc smelters,
iron and steel mill plants,
primary aluminum ore reduction plants,
primary copper smelters,
municipal incinerators capable of charging more.than 250 tens of refuse
per day,
hydrofluoric, sulfuric, and nitric acid plants,
petroleum refineries,
lime plants,
phosphate rock processing plants,
coke oven batteries,
sulfur recovery plants,
carbon black plants (furnace process),
primary lead smelters,
fuel conversion plants,
sintering plants,
secondary metal production facilities,
chemical process plants,
fossil-fuel boilers of more than 250 million British thermal units per
hour heat input,
petroleum storage and transfer facilities with a capacity exceeding
300,000 barrels,
taconite ore processing facilities,
glass fiber processing plants,
charcoal production facilities
-------
PROCEDURES FOR BART ANALYSIS
NO REQUIREMENTS
BART emission limitation
established equivalent to NSPS
No further analysis
ENERGY IMPACTS
NO —
SOURCE IDENTIFIED
(See figure 1)
SOURCE INFORMATION
PRELIMINARY ASSESSMENT OF
IMPROVEMENT IN VISIBILITY
Is visibility improved by
meeting NSPS emissions levels?
yes
ENGINEERING ANALYSIS
Analysis of the impacts of
retrofitting
OTHER ENVIRONMENTAL IMPACTS
ECOMOMIC IMPACTS
ALTERNATIVE CONTROL SYSTEMS
if retrofitting to NSPS is foun|
unreasonable, other control systems
should be analyzed.
BART SELECTION
Emission limitation established!
SIP REVISION
Figure 2
8
-------
according to procedures outlined in 40 CFR 51.303. The Administrator,
after appropriate public review, will grant or deny the exemption. Any
exemption is only effective upon concurrence by the Federal Land Manager.
2.0 VISIBILITY IMPACT ANALYSIS
See Figure 2.
Upon identifying the existing stationary facility to which the visibility
impairment is reasonably attributable, a BART analysis for the pollutant(s)
causing the impairment must be performed. A visibility impact analysis
is the first step, necessary to determine if visibility is anticipated to
improve from the imposition of retrofit controls. The following sections
discuss how this is accomplished.
2.1 PROCEDURES •
2.1.1 Source Information
In order to conduct a visibility analysis the following data are
needed.
1. Plant size, capacity, mode of operation
2. Emission rates (actual and potential) for nitrogen oxides
(NOV), particulates, and sulfur dioxide (SOV), (grams per second)
X A
3. Remaining useful life of any existing pollution control systems
4. Remaining useful life of any specific units within the plant
5. Remaining plant life
6. Stack diameters (meters)
7. Stack heights (meters)
8. Actual gas velocity (meters per second)
9. Stack temperature (degrees Kelvin)
-------
The above data should be obtained from the plant and should be confirmed
by other data available to the State from in-house, Federal, and local
agency records. Data for full load conditions should be used for pre-
liminary visibility impact analysis. For visibility impact analyses in
conjunction with evaluation of BART alternatives, variations in emission
rates with changes in production may be considered if reliable data
are available. Other parameters which may also be useful are opacity
measurements and particle size distribution of emissions.
2.1.2 Emission Rate Estimates
A representation of current, actual emission rates, i.e., emission
rates with any existing control systems, is necessary so that the
expected improvement in visibility can be estimated. These emission
rates can be obtained from various places such as the source itself,
other control agencies, in-house data, or new emission test data. They
should represent actual emissions and not estimates based upon theoretical
control efficiencies.
This data should be thoroughly analyzed for its accuracy based on
present plant conditions. If the emission rates do not seem appropriate^
in light of the observed visibility impacts, the State should require
additional emission tests, and/or calculate a current emission rate
considering present plant processes, air pollution control systems
currently in use, and current fuel input. The differing emission rates
should then be compared and, using good engineering judgment, the one
which most accurately represents the current emission rate of the
source should be used.
10
-------
2.2 PRELIMINARY ASSESSMENT OF IMPROVEMENT IN VISIBILITY
After all appropriate data are collected and emission rates
established, the amount of improvement in visibility expected from
retrofitting must be assessed. This is accomplished by comparing the
existing visibility (based on existing emissions) with the visibility
anticipated from imposition of the maximum achievable control. Maximum
achievable control is generally represented by New Source Performance
Standards as published in 40 CFR Part 60, applicable to the source
under analysis. If the visibility impact analysis shows visibility
improves a perceptible amount under this level of control,* the BART
analysis then begins to consider alternative retrofit control schemes
for the source. If, after comparison of the visibility at existing
and maximum achievable control levels, n<3 perceptible improvement is
expected, the analysis need not continue. Additionally, if the State
chooses to impose a BART emission limitation equivalent to the NSPS the
analysis need not continue.
Both analytical techniques and empirical methods may be used to
estimate the degree in improvement in visibility anticipated from
control of certain pollutants. Analytical techniques which assess
visibility at various emission levels are now being refined by the Agency.
Two guideline documents, "Workbook for Estimating Visibility Impairment"
and "User's Manual for the Plume Visibility Model (PLUVUE)," discuss
a useful analytical technique to aid in assessing improvements in
visibility at various control levels. These documents have undergone
public review and are available through NTIS. Although this technique
has yet to be fully validated, preliminary results using data from EPA's
VISTTA program are promising, and the Agency believes this to be a valuable
*
Preliminary studies indicate a change in contrast in the range of 0.01
to 0.04 is capable of being perceived by a human observer. [See Protecting
Visibility: An EPA Report to Congress.]
11
-------
part of the decision-making process. Use of these two guideline documents
is not, however, required. States which in their discretion use the
guidelines should not consider any results obtained exclusively, but
should consider this information together with all other available information
in making a regulatory decision.
Empirical methods, i.e. comparison photographic techniques, can
also provide valuable input into the sum of information on which to
base a BART decision. A discussion of this technique follows.
2.2.1. Primary Particulates
Primary particulates are one of the major causes of visibility
impairment generally observed in the form of a distinct plume. The visibility
impairment caused by a primary particulate plume is usually localized and
can generally be traced back, by visual observation or monitoring, to
its source. The improvement anticipated from controlling primary particulate
emissions is (1) the plume disappears, (2) the effect becomes even more
localized, (3) the effect is reduced perceptibly or (4) the frequency of
the impairment decreases so as to improve visibility. A common sense
approach using comparison photographic techniques could adequately
demonstrate the impact of controlling emissions for the purposes of
Phase I BART determinations. These photographic techniques would involve
comparing the effects caused by a well controlled source versus those
caused by the source under consideration. This comparison would be of
similar sources of equivalent size under similar meteorological and
geographical conditions. For example, if a similar source has applied a
certain primary particulate control and its plume disappeared, or the
impairment was reduced, the source could be used as an example of the
12
-------
amount of improvement expected by application of that control technology.
For a more specific discussion of the proper use of photographs, see
Section 3.3.3 of the "Interim Guidance for Visibility Monitoring."
A more precise analysis of the effects of particulate matter on
visibility is accomplished through the use of mathematical and other
analytical techniques. The State may use these techniques at their
discretion. However, the Agency is not requiring their use. The workbook
and User's Manual referenced in the previous section describe these
techniques.
2.2.2 Oxides of Nitrogen
" Another major component of visibility impairment is N02. Gaseous
N02 absorbs blue light creating a reddish or yellowish-brown plume. NOX
can also act as a precursor of light scattering aerosols. As with primary
particulate plumes, the NOV plume is usually localized and can generally
A.
be traced back, by visual observations or monitoring, to its source.
Current techniques for reducing NO emissions may show some improvement
/\
in visibility, but evidence shows such techniques generally do not
reduce emissions sufficiently to render the plume unobservable or provide
substantial improvement in visibility. New, more effective control
techniques, at present available only under limited circumstances,
should become available within the next few years. Section 51.302 of the
regulations requires States to reanalyze any pollutant (such as NOX)
that has not previously been controlled by BART when the Administrator
determines that new, more effective control technology is available.
As with particulate matter, a precise analysis of the effects of
NO is accomplished with the analytical techniques mentioned previously.
X
13
-------
2.2.3 Sulfur Dioxide
Sulfur dioxide does not directly affect visibility, but is a
precursor of light scattering aerosols. These fine particles, (sulfates)
by scattering light in the observer - target path, reduce the contrast
and, therefore the clarity and detail, between the target and its
background. This general reduction in contrast caused by sulfate aerosols
is most often associated with regional haze, which will be dealt with
under Phase II, but sulfates can and do contribute to visible plumes and
single source haze. If the visibility impairment is "reasonably attributable"
to the source, as may be the case in isolated, rural environments, the
source should be required to implement BART to reduce SC^ emissions
where improvement in visibility is anticipated. However, since SOg is
most often a contributor to regional haze, an existing major facility
that emits SCL will generally not be subject to BART for that pollutant
for the first phase of the visibility program.
Analytical techniques are needed, but not required, for a precise
analysis of S02 and its effects on visibility. The Workbook and User's
Manual referenced previously provides information on this.
2.2.4 Other Factors To Be Considered
Frequency, duration, and time of occurrence refer to how often an
impairment impacts a class I area, how long this impairment lasts, and
when the impairment occurs. Relative improvement such as the model
predicts will not always present all the benefits that can be obtained.
For example, the model may show an overall improvement in sky-plume
contrast of 10 percent .from worst case impairment, but this may be
sufficient to reduce the frequency of the impairment so that its impact
14
-------
is substantially reduced during period of maximum visitor use. Oftentimes,
a reduction in frequency and duration will provide a maximum benefit for
a minimum control effort. Thus, the temporal extent of the impairment
is of great importance and should be considered when assessing anticipated
improvements in visibility.
2.3 ENGINEERING ANALYSIS
If visibility is expected to improve as a result of the imposition
of controls, available retrofit control systems should be analyzed so
that an emission limitation representing BART can be established. BART
determinations must be based on the cost of compliance, the time necessary
for compliance, the energy and nonair quality environmental impacts of
compliance, any existing air pollution control technology in use at the
source, the remaining useful life of the source, and the degree of
improvement reasonably anticipated to result from the use of such technology.
A general discussion of the economic, energy, and nonair environmental
impacts which should be considered is found in the following sections.
For the engineering analysis required by this part, information specific
to coal fired power plants is found in Part II. Part II provides information
on selecting alternative retrofit systems, and assessing the economic,
energy, and environmental impacts of retrofit alternatives.
2.4 ENERGY IMPACT
Energy impacts should address energy use associated with the control
system under investigation and the direct effects of such energy use on
the facility and the community. Some specific considerations for energy
impacts are presented below.
15
-------
2.4.1 Energy Consumption
The amount, type (e.g., electric, coal, natural gas), and source of
energy required by the control system under consideration should be
identified and compared. In analyzing for energy consumption, comparisons
can be made in terms of energy consumption per unit of pollution removed
(for example, Btu/ton particulate removed).
2.4.2 Impact on Scarce Fuels
The type and amount of scarce fuels (e.g., natural gas, distillate
oil) which are required to comply with the control requirement should be
identified and compared. The designation of a scarce fuel may vary from
area to area, but in general a scarce fuel is one which is in short
supply locally and can better be used for alternative purposes, or one
which may not be reasonably available to the source either at present or
in the future.
2.4.3 Impact on Locally Available Coal
A control system which requires the use of a fuel other than locally
or regionally available coal should be discouraged if such a requirement
causes significant local economic disruption or unemployment.
2.5 ENVIRONMENTAL IMPACT
The net environmental impact associated with the emission control
system should be determined. Both beneficial impacts (e.g., reduced
emissions attributed to a control system) and adverse impacts (e.g.
exacerbation of another pollution problem through use of a control
system) should be discussed and quantified. Indirect environmental
16
-------
impacts (such as pollution impacts at an off-site plant which manufactures
chemicals for use in pollution control equipment) normally need not be
considered. Some specific considerations are presented below.
2.5.1 Air Pollution Impact
The impact of air pollutants emitted from a gas stream or a fugitive
emission source can be assessed in terms of either quantity of emissions,
modeled effects on air quality, or both. If application of a control
system directly removes or releases other air pollutants (or precursors
to other air pollutants), then the pollutants affected and the impact of :
these emission changes should be identified. The analysis can consider
any pojlutant affecting air quality including pollutants which are not
currently regulated under the Act, but .which may be of special concern
regionally or locally.
2.5.2 Water Impact
Relative quantities of water used and water pollutants produced and
discharged as a result of use of the emission control system should be
identified. Where possible, the analysis should assess their effect on
such local surface water quality parameters as pH, turbidity, dissolved
oxygen, salinity, toxic chemical levels and any other important considerations,
such as water supply, as well as on groundwater. The analysis should
consider whether applicable water quality standards are met and the
availability and effectiveness of various techniques to reduce potential
adverse effects.
2.5.3 Solid Haste Disposal Impact
The quality and quantity of solid waste (e.g., sludges, solids) that
must be stored and disposed of or recycled as the result of the application
of an alternative emission control system, if considered, should be
17
-------
compared with the quality and quantity of wastes created if the emission
control system proposed as BART is used. The composition and various
other characteristics of the solid waste (such as permeability, water
retention, rewatering of dried material, compression strength, Teachability
of dissolved ions, bulk density, ability to support vegetation growth
and hazardous characteristics) which are significant with regard to
potential surface water pollution or transport into and contamination of
sub-surface waters or aquifers should be considered. The relative
effectiveness, hazard and opportunity for solid waste management options,
such as sanitary landfill, incineration, and recycling, should be identified
and discussed.
2.5.4 Irreversible or Irretrievable Commitment of Resources
The BART decision may consider the extent to which the emission
control system may involve a trade-off between short-term environmental
gains at the expense of long-term environmental losses and the extent to
which the system may result in irreversible or irretrievable commitment
of resources (for example, use of the scarce water resources).
2.6 ECONOMIC ANALYSIS
This analysis should address the economic impacts associated with
installing and operating control systems under consideration for BART.
Costs associated with New Source Performance Standards can be found in
the NSPS Background Information Documents. Other economic impacts which
should be considered follow.
2.6.1 Direct Costs
The direct cost for a control method should be presented. Investment
costs, operations and maintenance costs and annualized costs should be
presented separately. Costs should be itemized and explained. Credit
18
-------
for tax incentives should be included along with credits for product
recovery costs and by-product sales generated from the use of control
systems. The lifetime of the investment should be so stated. The costs
of air treatment, water treatment, and solid waste disposal should be
presented separately. When considering the addition of control equipment
to that already in place, the cost of incremental control should be
analyzed. Additionally, the expected useful life of any existing control
equipment should be evaluated on the basis of its expected retirement/
replacement schedule.
As a guide in determining when control costs become excessive,
comparisons can be made in terms of certain cost effectiveness ratios.
Such ratios may include the following: .-• . .
. ratio of total control costs to total investment costs
. cost per unit of pollution removed (for example, dollars/ton)
unit production costs (for example, mill/kw-hr, dollars/ton).
In some cases, the unit of production output may be difficult to determine,
as in the case of a plant producing many different products. In such
cases, unit production costs can be expressed as cost per dollar of
total sales.
The remaining useful life of the source will have an effect on the
amortized cost of the anticipated control equipment and, as such, should
be given strong consideration in determining BART.
2.6.2 Capital Availability ,
Capital availability addresses the difficulty that some sources may
face in financing alternative control systems. Proof of such claims
should be fully documented.
19
-------
2.6.3 Local Economic Impacts
Local economic impacts address the economic feasibility of BART
requirements and the impact on production decisions of the firm in
response to the level of control. For example, BART could alter the
economics of the plant to the point where the decision would be made to
cancel expansion of a facility, to reduce the scale of operation, or to
change the production mix. The local employment effects, including
number of jobs, dollars paid in salaries, and changes in employee skill
levels required should be evaluated. The guideline does not imply that
the BART decision should force a plant to the brink of shutdown. The
BART decision must be based on sound judgment, balancing environmental
benefits with energy, economic, and other impacts.
2.7 CONSIDERING ALTERNATIVE CONTROL SYSTEMS
As previously stated, for fossil fuel fired power plants with a
generating capacity in excess of 750 megawatts, the Agency believes that
the NSPS level of control can be met with technology that is generally
available to these sources, and that this level of control generally
represents the best these sources can install as BART.
In determining BART, and for inclusion in its SIP, the State must
explain in detail how it weighed the various BART factors required by
the Act (§169A(g)(2)), the regulations (§51.301(c)), and this guideline.
This explanation irust demonstrate that the emission limitation chosen
(if one other than the NSPS) reflects a reasonable balance of the various
BART factors. This explanation must set forth the visibility, energy,
economic, and other impacts associated with application of an NSPS level
of control, and compare those impacts to alternative levels of control
20
-------
including the level of control selected by the State as BART. Because
EPA believes that NSPS control generally represents the best these sources
can install as BART, if the State sets for a pollutant emitted by a
fossil fuel fired power plant with a generating capacity in excess of
750 megawatts a BART emission limitation equivalent to the NSPS level of
control, this detailed demonstration will not be required for the purposes
of EPA review.
3.0 BART SELECTION
An emission limitation that is BART must be established for each source.
This along with all evidence as to why this emission limit was chosen is
incorporated into the SIP submitted to EPA for approval. It is suggested
that if a range of alternative control systems were examined, the State
arrange these alternatives into an array. This array would include a
description of each alternative considered, the cost of the alternative,
the improvement, in visibility obtained, and any economic, energy, on nonair
environmental factors which affect the selection. This array would provide
a logical sequence by which the BART emission limitation was set. The State
must also present the logic network used in its final decision making
process.
21
-------
-------
PART II
RETROFIT GUIDELINES FOR
COAL-FIRED POWER PLANTS
Il-i
-------
This report has been reviewed by the Office of Air Quality Planning and .
Standards, Office of Air, Noise, and Radiation, Environmental Protection
Agency, and approved for publication. Msntion of company or product
names does not constitute endorsement by EPA. Copies are available free
of charge to Federal employees, current contractors and grantees, and
non-profit organizations - as supplies permit - from the Library Services
Office, MD-35, Environmental Protection Agency, Research Triangle Park,
NC 27711; or may be obtained, for a fee, from the National Technical .
Information Service, 5285 Port Royal Road, Springfeild, VA 22161.
Publication No. EPA-450/3-80-009b
Il-ii
-------
CONTENTS
Section
NOTE
.CONTENTS
FIGURES
TABLES
PART II.
RETROFIT GUIDELINES FOR COAL-FIRED POWER PLANTS
1.0 BACKGROUND INFORMATION
1.1 Introduction
1.2 Relation to Part I.
1.3 Utilization of Part II.
1.3.1 Purpose
1.3.2 Data Assumption and Technical
Approach
1.3.3 Content and Limitations
1.3.4 Method of Use
1.4 References
1-1:
1-1
1-1
1-2
1-2
1-3
1-4
1-4
1-5
2.0 RETROFIT EMISSION CONTROL TECHNIQUES
2.1 General
2.2 NO Emission Reduction Techniques
X
2.2.1 Low Excess Air
2.2.2 Staged Combustion
2.2.3 Low NO Burners
2.2.4 Flue Gets Recirculation
2.2.5 Burners Out-of-Service
2.2.6 Flue Gas Treatment
2.2.7 Derating
2.2.8 Reduced Air Preheat
2-1
2-1
2-3
2-5
2-7
2-8
2-8
2-9
2-9
2-9
2-10
11-111
-------
CONTENTS
Section
2.3 Particulate Emission Control
2.3.1 Electrostatic Precipitators (ESP)
2.3.2 Baghouses
2.3.3 Flyash Scrubbers
2.3.4 Effect of Acid Mist"on Particulate
Emissions
2.4 Emission Control of Sulfur Oxides
2.4.1 General
2.4.2 Description of Representative Wet and
Semi-dry Scrubbing Systems
2.5 Emission Monitoring
2.6 References
3.0 RETROFIT DESIGN AND COSTS
3.1 General
3.1.1 Emissions
3.1.2 Basis of Costs
'3.2 Retrofitting to Reduce N0y Emissions
3.2.1 Retrofit Techniques for NO Control
-A.
3.2.2 Retrofit Costs for N0v Control
3.3 Retrofitting Ductwork and Stacks
Page
2-9
2-9
2-10
2-11
2-12
2-13
2-13
2-14
2-28
2-30
3-1
3-1
•%
3-1
3-2
3-6
3-7
3-13
3-17
Il-iv
-------
CONTENTS
Section
Page
3.4 Retrofitting To Control Particulate Emissions 3-19
3.4.1 General
3.4.2 Electrostatic Precipitator Design
3.4.3 Baghouse Design
3.4.4 Retrofit Costs for Particulate Control
3.5 Retrofitting to Control S02 Emissions
3-19
3-22
3-26
3-2,8
3-32
3.5.1 Retrofit Costs for Wet S02 Control 3-32
3.5.2 Retrofit Costs for Lime Dry S02 Control 3-35
3.6 . Land Area Requirements
3.7 Emission Monitoring Costs
3.7.1 Retrofit Capital Costs
3.7.2 Operating, Costs
3.7.3 Annual Costs
3.8 Time Requirements For Retrofitting
3.9 Refernces
3-38
3-39
3-39
3-43
3-43
3-44
3-46
II-V
-------
CONTENTS
Section
Page
4.0 TECHNIQUES FOR ESTIMATING RETROFIT COSTS FOR
EMISSION CONTROL
4-1
4.1 General
4.2 Working Capital
4.3 Auxiliary Boiler Costs
4.4 Electrical Energy Penalty
4.5 Other Costs Not Estimated
4.6 Escalation
4.7 References
4-1
4-1
4-1
4-3
4-3
4-6
4-7
APPENDIX A
APPENDIX B
APPENDIX C
APPENDIX D
APPENDIX E -
RETROFITTING THE FOUR CORNERS POWER STATION
RETROFITTING THE MOHAVE POWER STATION
RETROFITTING THE NAVAJO POWER STATION
ANALYSIS OF FGD EFFICIENCY BASED ON EXISTING
UTILITY BOILER DATA, PREPARED FOR EPA BY
VECTOR RESEARCH, INCORPORATED
EPA RESPONSE TO PETITIONS FOR RECONSIDERATION
A-l
B-l
C-l
Il-Vi
-------
FIGURES
Figure
2-1 Most highly developed flue gas desulfurzation
processes
2-2 Typical process flow diagram for lime/limestone
scrubbing
2-3 General process flow diagram for semi-dry SC>2
scrubbing with lime
2-4 Semi-dry scrubbing system - Wheelabrator-Frye/
" Rockwell International
2-5 Semi-dry scrubbing system - Joy-Niro
2-6 Semi-dry scrubbing system - Babcock § Wilcox
3-1 Location of overfire air ports for C-E boilers
3-2 Arrangement of curtain air ports for F-W and R-S
boilers
3-3 Typical schedule for retrofitting large power
plants
Paj
2-16
2-21
2-23
2-27
2-28
2-30
3-12
3-14
3-45
Il-vii
-------
TABLES
Table
2-1 Characteristics of Commercial Throwaway FGD
Processes
2-2 Semi-Dry Scrubbing Systems; Characteristics of
Some of the Systems Presently Under Construction
3-1 B§W Low NO Burner Costs
A.
3-2 Overfire Air Port Costs
3-3 Values of A and b for Estimating the Cost of
Utility Boiler Stacks
3-4 ESP Specific Collection Area for Various Coals
3-5 Values of A and b for Estimating Capital and
Annual Costs of Wet Flue Gas Desulfurization
Systems
3-6 Electrical Energy Requirements for Wet Flue Gas
Desulfurization Systems
3-7 Minimum Land Area Requirements for Lime and
Limestone Scrubbing Systems
3-8 Sludge Generation for Lime and Limestone Scrubbing
Systems
4-1 Capital and Annual Costs for Auxiliary Boilers
Paj
2-19
2-25
3-15
3-16
3-18
3-24
3-33
3-36
3-40
3-41
4-2
-------
PART II
SECTION 1
BACKGROUND INFORMATION
1.1 INTRODUCTION
This part of the proposed BART guideline is for use in assessing
the effectiveness of retrofit control techniques and for estimating
cost. They are flexible with respect to specifying control systems for
implementation of BART.
1.2 RELATION TO PART I
Part I provides guidance on identifying those sources to be
analyzed for BART, assessing the anticipated improvement in visibility,
conducting an engineering analysis, and establishing emission limitations
for BART. Part II, as discussed below, contains an explicit discussion
of the engineering analysis required by Part I. Part I is general guidance
and is appropriate for the analysis of all existing major stationary source
categories.
This Part II provides specific engineering information on coal-fired
power plants having an operating capacity in excess of 750 megawatts. It
provides information for selecting alternative retrofit systems, and
assessing the economic, energy, and environmental impacts of retrofit
alternatives. Although this part is specifically for coal-fired power
plants, much of the engineering information and procedures may be helpful
when analyzing sources in other source categories.
1-1
-------
1.3 UTILIZATION OF PART II
1.3.1 Purpose
The guidelines in this document specify the emission
levels, emission reduction potential, and costs corresponding
to each of the retrofit systems discussed. By judicious
application of these data to any plant situation, an estimate
of cost and effectiveness of a control may be made for that
plant. The guidelines are not intended to provide comprehensive
cost estimates for retrofitting coal-fired steam generators.
Comprehensive cost estimates require extensive engineering
studies such as the preparation of specifications, bid criteria,
equipment layouts, and detailed drawings. Because the funds
needed for these types of studies are usually beyond the budgets
of most air pollution control agencies, the broad cost estimating
techniques of this document are recommended. The cost
estimating data and procedures of this document will generally
yield reasonable cost. Should one suspect that the cost
estimates of this document would lead to a false conclusion
on the cost feasibility of retrofitting certain control
systems, the more comprehensive cost (and more costly)
estimating techniques previously described should be used.
Although the precision of the cost estimates can be improved
by more costly studies, the accuracy of conclusions on the
effectiveness of the various systems for reducing emissions
would generally not be significantly improved by further
s tudy.
This document was prepared recognizing that there are techniques
other than those used as the basis for this document that are as
effective as those used for the cost estimates. Consequently,
the owner of a coal-fired steam generator should be allowed
to select other techniques as long as such alternate systems
perform at a level of effectiveness required by the BART.
determination.
1-2
-------
1.3.2 Data, Assumptions, and Technical Approach
This study resulted from the need to understand the basis
and methods of retrofit cost analysis that would cause
emission reduction of nitrogen oxides, particulate, and
sulfur oxides. The cost modules developed have been based
on the emission levels found in EPA background
documentation (1, 2, 3, and 4,). These levels are 210 and
260 nanograrns per joule heat input (0.5 lbs/10 Btu
and 0.6x lb/106 Btu) for NO from subbituminous
A
and bituminous coal respectively; 13 ng/J heat input (0.03
lbs/106 Btu)' for particulate emissions; and 90% :
removal of the sulfur oxides from the power plant flue
gas. These three pollutants are of prime visibility
concern although emissions from large, coal-fired, steam
generators also include carbon monoxide, halogens, trace
metals, and hydrocarbons (including polycyclic organic
matter). As stated in Part 1, it is doubtful that either NO
A
or SO control will be required in Phase 1 of the visibility
A
program. However, when this report was begun it was felt that control
systems for all of the major visibility impairing pollutants
should be investigated. All of that information is presented here
for reference and future use. The process and cost data were
obtained primarily from background information for new source perfor-
mance standards and from Pullman Kellogg in-house work (1,2,3,4, and
5). The data needed for establishing process requirements to
retrofit the example power plants were obtained from information
furnished by the power plants, from visual inspection of the
plant sites during plant visits, and from yearly reports prepared
by the utilities (FPC Form 67).
The methods considered for control of emissions are:
boiler modifications for reduction of nitrogen oxide
emissions; particulate control using baghouses and/or
electrostatic precipitators (hot or cold side); and flue
gas desulfurization by either wet or semi-dry scrubbing.
1-3
-------
The scope of work was directed to designs for retrofitting
power plants with 750 'MW, or larger, total plant capacity.
However, some of the designs can be applied to much
smaller plants. The costs developed here incorporate the
variations involved in attaining the plant capacity;
therefore, the study accomodates retrofitting most power
plants with emission controls.
1.3.3 Content and Limitations
The general content and the costs in this report describe
the method and choice of individual retrofit for emission
controls. The document also develops a method for
determining total retrofit investment and annual operating
costs. The content has been developed for engineering
personnel use such that the States and Federal government
can make best available retrofit technology decisions. It
is also intended for use by those interested industry
personnel involved in environmental control. The
appendices provide examples of retrofit costs and plant
layout requirements for three power plants. Reduction in
nitrogen oxide formation is achieved by boiler
modification only; no other control alternatives have been
selected. Particulate emissions control is limited to
baghouses and electrostatic precipitators (hot and cold
side). The flue gas desulfurization systems are designed
for wet or dry scrubbing.
1.3.4 Method of Use
Methods for developing cost data are described in Section
3. The technique for using these cost modules to
determine the total retrofit costs for a power plant is
described in Section 4. Examples are presented in the
Appendices.
1-4
-------
1.4 REFERENCES
1. EPA, "Electric Utility Steam Generating Units,
Background Information for Proposed NOX Emission
Standards." EPA-450/2-78-005a, July 1978.
2. EPA, "Electric UtiTity Steam Generating Units,
Background Information for Proposed Particulate Matter
Emission Standards." EPA-450/2-78-006a, July 1978.
3. EPA, "Electric Utility Steam Generating Units,
Background Information for Proposed S02 Emission
Standards." EPA-450/2-78-007a, August 1978
4. EPA, "Electric Utility Steam Generating Units,
Background Information for Proposed S02 Emission
Standards Supplement." EPA-4'50/2-78-007a-l, *ug 1979
5. Final report, "Retrofit Guidelines for Coal-Fired
Power Plants," Pullman Kellogg Division of Pullman
Incorporated, EPA Contract No. 68-02-2619, Work
Assignment 13, September 1979
1-5
-------
-------
SECTION 2
RETROFIT EMISSION CONTROL TECHNIQUES
2.1 GENERAL
The retrofitting technques for NOX, S02, and
particulate emissions considered in this document are based
only on commercially available methods for reducing these
pollutants. For NOX, the emission reduction techniques
considered include staged combustion (overfire air and/or
curtain air) and low NOX burners. The particulate
collection studies examined ESP's (cold or hot side) and
baghouses (fabric filters).
The maximum control effectiveness of the systems discussed
in this document is as follows:
Subbitumlnous coal
Bituminous coal
210 nanograms per joule
(0.5 lb/106 Btu)
260 nanograms per joule
(0.6 lb/106 Btu)
2-1
-------
Particulates
Fabric Filters
and Electrostatic
Precipitators
Scrubbers
sg2
Wet scrubbers
Dry scrubbers
13 nanograms per joule
(0.03 lb/106 Btu)
21 nanograms per joule
(0.05 lb/106 Btu)
90 percent removal o£ the SC>2
70 percent removal of the SC>2
As discussed in Section 2.2 and Section 3, it may not always
be possible to attain these NOX levels for all retrofit
situations. The EPA position on the operating effectiveness
of particulate and S02 retrofit control systems is
discussed in Appendices D and E of these guidelines.
Control of S02 emissions included studies of both wet
arid semi-dry scrubbing. The costs developed for the wet
scrubbing system include cases that use lime or limestone,
Wellman Lord, Mag-ox, or double alkali scrubbing. The
semi-dry scrubbing (lime) uses the Joy-Niro process. This
process uses a s-pray dryer followed by a baghouse for
particulate collection.
The control systems outlined above are discussed in detail
in the following sections.
2-2
-------
2.2 NOV EMISSION CONTROL TECHNIQUES
X
Section 2.2 discusses the technical aspects of retrofitting to
reduce NO emissions from coal-fired steam generators. There are
A
several expert and excellent documents in the public domain which
discuss the various aspects of control systems for reducing NO
A
emissions. Some of these discussions were submitted as comments
to the proposed visibility regulations. This guideline was not
designed to eliminate the consideration of any other information
available to the State. The State should use such information at
its discretion, with sound engineering judgment.
There are two distinct mechanisms for forming NO : one is fixation
X
of elemental nitrogen from the air, and the other results from
chemically combined nitrogen taken from fuel, in this case, coal.
Fixation of the nitrogen in air can be limited by reducing the
level of thermal excitation caused by flame temperature. The
principal methods for reducing thermal excitation are: (a) flue
gas recirculation, (b) staged combustion, (c) water or steam
injection, (d) reduced air.preheat, and (e) reduced heat-release
rate.
-------
Formation of NOX by oxidizing the nitrogen in coal
constitutes as much as 80 to 90 percent of the total NOX
emissions from pulverized-coal-fired boilers. By limiting
the combustion air supplied at or near the burners and by
directing air to limit high temperature mixing of
volatilized coal, nitrogen, and air, nitrogen oxide
formation from coal nitrogen can be retarded. As the fuel-
rich mixture cools by radiating heat to the surrounding
colder surfaces, the mixture flows into the air rich zone
and completes combustion at lower temperatures -
temperatures which are less favorable for developing
NOX. This method, one of several used to reduce NOX
emissions, is called staged combustion. There are several
NOx control techniques discussed in this document common
to all four boiler manufacturers including:
o Low excess air,
o Staged combustion,
o Low NOX burners,
o Flue gas recirculation,
o Burners out-of-service,
o Flue gas treatment,
o Derating,
o Reduced air preheat.
-------
The sections that follow discuss these techniques and
their potential adverse side effects. For new steam
generators, NOX emissions from burning Western
subbituminous coals can be reduced to a level of 210
nanograms per joule heat input (0.5 lb/10" Btu) and
emissions from burning bitumious coals can be reduced to
260 nanograms per joule (0.6 lb/10^ Btu) without
significant adverse side effects (1). NOX control for
new boilers has to be accomplished by custom design for
the steam generator in order to minimize adverse side
•effects while limiting NOX emissions.
For existing steam generators, it is not possible to
change the shape or size of the combustion chamber
substantially without replacing the steam generator. The
state of the art for reduction of NOX by combustion
modification is not developed to the extent that the
effectiveness of applying known control techniques to
existing steam generators can be predicted accurately. In
addition, it is not possible to predict the combustion
conditions where adverse side effects will become
intolerable for existing units. Consequently, reducing
NOX emissions from existing steam generators involves
trial and error as well as application of sound
engineering principles. Western subbituminous coals have
less tendency to cause tube wastage and slagging than some
high sulfur Eastern bituminous coals. Therefore, the
probability of success in reducing NOX emission by
applying combustion modification techniques is greater for
Western coal. Both Combustion Engineering (C-E) and
Babcock and Wilcox (B&W) steam generators have been
retrofitted to reduce NOX emissions to levels less
than 210 nanograms per joule (0.5 Ib/lO^ Btu) without
significant .-adverse side effects (1) .
2-4
-------
-------
2.2.1 Low Excess Air
About 10 to 20% excess air is needed in addition to the
theoretically required air to assure efficient, stable
operation of the boiler. This amount of excess air is
needed to cover the normal +3 percent fluctuations in
required combustion air, to aid soot burnout, to
.increase convective heat transfer, to harden the slag,
and to minimize tube wastage (corrosion). After
accommodating the air requirements established by these
operating conditions, if excess air can be reduced, then
NOX is reduced either because less oxygen is
available during volatilization, or thermal NOX is
retarded by low, oxygen radical concentrations.
However, under ideal conditions, well mixed, adiabatic
combustion systems respond adversely to lower excess
air, because higher NOX emissions result from higher
adiabatic flame temperatures. For pulverized-coal-fired
plants, the reduction in NOX emissions may be as
much as 20$ when applying the. low-excess-air control
method. When this method is applied, tight control of
individual burners must be made with respect to fuel/air
ratio. Although utility boiler systems usually show
NOX reductions with low excess air, the
effectiveness of reduced excess air varies for
individual boilers and there are some associated
problems which are discussed in the following
sub-sections.
2.2.1.1 Tube Wastage.- Operation in an environment with too low
excess air produces a fuel rich reducing atmosphere
which may accelerate the corrosion of the furnace tubes.
Therefore, there is a lower limit for reduction of
excess air below which potential adverse side effects
begin to accelerate (!_).
2-5-
-------
2.2.1.2 Slagging.- Potentially, low excess air can accelerate
slagging (I) . The molten component of the ash becomes
slag in the reducing atmosphere created by low excess
air.
2.2.1.3 Increased ash combustibles and CO content.- Low excess
air operation can have an adverse effect on particulate
loading if there is an increase in unburned carbon in
the ash. Any increase in particulate loading is
accompanied by changes in particulate characteristics
such as size distribution, and ash resistivity which can
affect the collection efficiency of ESP's. The increase
in unburned carbon may also result in energy losses in
the boiler. Opacity may be increased because of the
increase in particulate loading which causes overloading
of ESP's. Low excess air can also increase the CO
emissions (jL).
2.2.1.4 Reduced steam superheat.- When the amount of excess air
is reduced, the flue gas mass flowrate decreases. This
causes a decrease in the heat transfer rate and may lead
to a decrease in the superheated steam temperature. A
difference of up to 28°C (50°F) in the superheat and
reheat steam temperature may occur. Also, the reduced
superheat temperature -may require existing old plants to
reduce capacity as much as 30% of their rated power
output (I) ,
2.2.1.5 Reduced boiler efficiency.- Low excess air may not
reduce the boiler efficiency because the energy loss due
to the increase in unburned carbon and CO may be offset
by the decrease in energy loss to the stack resulting
from a lower flue gas temperature. Efficiency may be
improved in cases where CO and ash combustibles are not
increased significantly because combustion control
V
settings are tuned more finely (I) .
2.2.1.6 Flame Stability.- Flame stability can be affected by
reducing excess air. Flameouts and pulsations not only
disrupt electric power generation but also cause
potential safety hazards. Consequently, for all steam
generators, there are low excess air conditions that are
intolerable.
2-6
-------
2.2.2 Staged Combustion
Staged combustion can be accomplished in three ways. One
method is by maldistributing air (over-fire air), another
method is by maldistributing fuel (burner-out-of-service),
and the third method involves special burner design. A
typical utility 'boiler operates with an array of burners,
each of which operates at the percentage of excess air
specified for the boiler. The .flame, characteristics
differ with the individual boiler manufacturer's design.
Staged combustion is accomplished by redistributing the
air flow such that a cooler secondary combustion zone is
encountered by the fuel-rich combustion gases after they
leave the flame basket. Staged combustion has two effects
on NOX production;
o Fuel NOX is reduced because less oxygen is
available during volatilization.
o Thermal NOX is reduced because the flame
temperature does not reach as high a peak as when
all the heat release occurs in one stage.
The extent of staged air can be conveinently indexed by
the fraction of stoichiometrically-required air remaining
at the burner flame baskets. For example, suppose a six-
level burner boiler operating with 15$ excess air has five
operating burner levels with air supplied to six levels.
Then one-sixth of the air supply is staged, leaving the
burners operating at 96$- (115 x 5/6) of the required
stoichiometric air. This type of stag-ed combustion has
shown reductions of NOX production by as much as 20 to
40$ for pulverized-coal-fired utility boilers (_!) . Staged
combustion by special burners produces a lazy, fuel-rich
flame surrounded by an air envelope. Limited tests
indicate that this type of burner can reduce NOX
2-7
-------
emissions by as much as 50$. Potential adverse side
effects of staged combustion are much the same as those
discussed in Section 2.2.1.
2.2.3 Low N0y Burners
Low NOX burners are designed to reduce NOX
emission levels either alone or in combination with the
use of overfire air ports. With the low NOX burner
designs, it is possible to eliminate or alleviate the
potential problem that creates reducing atmosphere
pockets, and the tube wastage associated with reducing
atmospheres should be far less serious. Low NOx
burners are designed to reduce highly turbulent mixing
between the secondary and primary air streams. Because of
low turbulence intensity, the flame length in low NOX
burners will be longer 'than the flames of normal high
turbulence burners, and these increased flame lengths must
be evaluated by the manufacturer when retrofitting a
boiler. Potential adverse side effects are much the same
as .those discussed in Section 2.2.1.
2.2.4 Flue Gas Recirculation
The flue-gas recirculation method of NOX control
operates by recirculating flue gas to the windbox which
reduces the formation of thermal NOX by lowering flame
temperature and oxygen concentration at the burners. This
technique has been tried on an experimental basis and has
been found to be relatively ineffective in coal-fired
units. Also, the flyash problems in the recirculation
systems have not been solved sufficiently to warrant a
conclusion that the technique has been demonstrated
effectively U).
-------
2.2.5 Burners Out-Of-Service
This method .of NOX control is accomplished by shutting
off the pulverizers supplying the upper level burners.
The technique can be accomplished only if the remaining
pulverizers have enough spare capacity to supply the total
amount of coal required. Often, application of the
technique increases the minimum excess air requirements of
the boiler, and this may result in efficiency loss in the
boiler. Boiler derating may be required when fans are
already operating near their maximum capacity. Also
intolerable superheat conditions may occur because of a
resultant shift of the high temperature zone towards the
superheater. The technique could be useful, however, for
both owners and regulatory agencies, in helping to
determine what adverse side effects might ensue from
equipping a unit for overfire air while avoiding the need
for costly modifications.
2.2.6 Flue Gas Treatment ' .
The flue gas treatment method of NOX control has been
applied to oil and gas-fired units in Japan. EPA is
investigating the Japanese technology for potential
application to the U.S., coal-fired situation. At
present, this technology is not sufficiently demonstrated
on coal-fired units to be considered here (!_)..
2.2.7 Derating
Derating operates by causing a reduction in the volume of
flue gas which produces a reduction of NOX, both in
concentration and mass flow rate. Also, reducing the heat
release rate reduces combustion gas temperatures more
rapidly, thereby reducing the rate of formation of NOX
from the reactions between the nitrogen in the combustion
air and oxygen.
2-9.
-------
NOX control by derating older units will be more
difficult because of smaller combustion chamber design
practiced by boiler manufactures prior to 1971. The older
units, with a smaller firebox, have higher heat release
rates when compared to the units designed after 1971,
which have a 15 to 50% increase in combustion chamber
size.
Usually derating is an undesirable NOX control
technique. However, at times, it might develop that
derating would be a preferable alternative for utility
owners who otherwise would have to make costly steam
generator modifications.
2.2.8 Reduced Air Preheat
Reduced air preheat to control NOX emission is a
nonviable method because of the need for a hot air supply
to the pulverizer to heat the coal. The method causes
losses in boiler efficiency because of heat losses
attributable to increased stack gas temperature.
2.3 PARTICULATE EMISSION CONTROL
2.3.1 Electrostatic Precipitators (ESP)
Electrostatic precipitators function by charging and
collecting particles on. collection electrodes, and by
disposing of the collected ash. Primary electric power
supply is usually 240 or 480 volts of alternating current.
Transformer-rectifier sets are used to convert the current
from alternating to direct and to step up the secondary
voltage. High efficiency ESP systems are equipped with
power controls which regulate power at the optimum levels
for particulate collection. Secondary voltages range from
2-10
-------
10,000 to 40,000 volts, depending upon the particulate and
gas characteristics. Rapping systems dislodge the
collected ash. These systems are equipped with controls
which permit adjustment of both the frequency and
intensity of rapping. High efficiency ESP systems are
equipped with multiple hoppers and baffles to minimize the
tendency for gases- to bypass the electrical field
(sneakage). ESP systems are capable of reducing
particulate emissions to levels as low as 13 nanograms per
Joule heat input (0.03 lb/106 Btu) (2).
Performance of an ESP is affected by the following factors
(2):
Ov Coal ash characteristics - size distribution of
particulates, compositions such as sulfur content,
etc.
o ESP size - collection area and flow, cross-sectional
area for flue gas
o Grounded collection surface spacing
o Power control design
o Gas flow distribution
o Rapping control design
o Plyash handling system design
o Thermal expansion design
o Discharge electrode failure
o Maintenance practices
o Gas conditioning
2.3.2 Baghouses
Baghouses used for particulate emissions control are
effective to the extent of 13 ng/J heat input (0.03
lbs/10^ Btu) (2_) for best-controlled sources.
Pressure drop data show a range of less than 1.25
kilopascals (5" H20) to 2.5 kilopascals (10" H20),
2-11
-------
all at full load. Air-to-cloth ratios corresponding to
these pressure drops range from 0.58 to 0.91 actual cubic
meters per minute per square meter (1.9 to 3.0 ACFM/Ft2).
Data show that an air-to- cloth ratio of 0.61 actual
cubic meters per minute per square meter (2 ACPM/Pt2)
is a conservative criterion for sizing a baghouse for a
coal-fired steam generator with pressure -drops of less
than 1.25 kolopascals (5"H20) at full load gas volumes.
Although it is not demonstrated on large sized electric
utility steam generators, it is possible that precharging
the particulates before entry into the bags would permit the
use of much smaller baghouses with air to cloth ratios of
as much as 1.2 actual cubic meters per minute per square
meter (4 ACFM/Ft2)A. The life of a bag is estimated to be
at least 2 years if pressure drops are less than 1.25
kilopascals (5"H20).
2-1 la
-------
The key factors which affect baghouse performance are
(2);
o Bag material,
o Bag construction,
o Bag treatment,
o Baghouse size,
o Configuration of baghouse and 'its "control,
o Techniques of cleaning,
o Construction of tube sheet,
o Process characteristics,
o Maintenance practices.
2.3.3 Flyash Scrubbers
In common practice, using scrubbers to control
particulates from coal-fired power plants is done only in
conjunction with PGD systems. The FGD system can be
designed either for simultaneous removal of SC>2 and
particulates (Kellogg-Weir scrubbers, TCA scrubbers, etc.)
or for separate particulate and sulfur dioxide removal
(venturi-spray tower combination). Scrubbers are capable
of reducing particulate emissions to levels as low as
21 nanograms per joule heat input (0.05 lb/10°
Btu) (2).
2-llb
-------
A great variety of scrubber types and configurations
exist. Some of the most widely used scrubbers are
venturi, spray tower, orifice impingement, and
self-induced spray. The basic principle of scrubber
operation involves confronting particulates with impact
targets which can be either wetted surfaces or, and most
frequently, individual droplets.
The efficiency of a wet scrubber is a function of a number
of variables including particle characteristics (diameter,
density, viscosity) between the particle and the scrubbing
slurry droplet. To obtain efficient particulate removal,
the system is designed for an optimum combination of these
parameters. This combination is achieved by using a high
liquid to gas ratio (L/G) of scrubbing slurry to gas
stream, and by providing a high degree of atomization for
the scrubbing liquid. The choice of particulate collector
must take into account the particle size distribution, the
required collection efficiency, and the overall energy
consumption (as measured by the pressure drop across the
system).
2.3.4 Effect of Acid Mist on Particulate Emissions
Particulate emissions from coal can be affected by acid
mist concentrations. Reference 2 shows PGD units on well-
controlled coal-fired power plants do not increase
particulate emissions through sulfuric acid formation and
interaction.
2-12
-------
2.4
EMISSION CONTROL OP SULFUR OXIDES
2.4.1 General
The combustion products of the coal-fired, power-
generation station contain sulfur dioxide. There are
several sulfur removal alternatives. These alternates can
be grouped into these three principal categories:
o Desulfurization of coal prior to combustion - either by
ohemical or physical cleaning.
o Desulfurization of coal during combustion - fluidized
bed combustion.
.0 Flue gas desulfurization.
For this project only flue gas desulfurization will be
considered as a means for controlling emissions of sulfur
oxides to the atmosphere. .
There are several types of FGD processes that are
commercially available now, and they are grouped into two
major categories that are based on the product resulting
from the process:
o Throwaway processes where all waste streams are
discarded.
o Regenerable processes where the waste stream is treated
for regeneration of the sorbent and recovery of sulfur
compounds such as sulfur, sulfuric acid, and
concentrated H2S. .
2-13
-------
Both processes treat the flue gases using either wet,
semi-dry, or dry desulfurization modes. Since dry
processes, such as dry scrubbers using boiler injection
with limestone, are not within the scope of this study,
they will not be considered further. The wet and semi-
dry processes are accomplished using either slurry or
clear liquor, and these are further divided by the type
of alkali used for the scrubbing solution.
Figure 2-1 illustrates the general classification of
these PGD processes. This figure also indicates the
t
nine most highly developed processes at the extreme
right.
2.4.2 Description of Selected Wet and Semi-Dry._.Sc_rubbin.g_
Systems
2.4.2.1 S0g control by wet scrubbing.- Wet scrubbing
processes have gone through extensive research,
development, and full scale demonstration work. The
available data indicate that wet scrubbing will remove
' both S02 and particulates. Usually, the flyash is
collected upstream of the PGD unit to minimize the
volume of sludge and to prevent erosion of the
-------
FGD
PROCESSES
/
SEMI^DRYN.
'(SPRAY DRYER))—
REGENERABLE Y
SEMI-DR
SPRAY DRYER)
Figure 2-1.- Most highly developed flue gas desulfurization processes,
2-15
-------
moisture content increased by the scrubbing liquor, it
is often reheated before it enters the stack. Reheat
minimizes condensation and corrosion of the equipment
downstream of the scrubber (ductwork, fans), and it also
helps to avoid plume formation at the stack exit. The
product sludge is either discarded (throwaway) or
processed further (regenerable).
The types of process equipment and operating parameters
vary widely for these systems. The principal
differences are in the following equipment and
chemicals:
o Scrubbing equipment - The several successful systems
already used include packed tower, horizontal spray
towers, vertical spray towers, tray columns, and
venturi scrubbers.
o Types of alkali - For throwaway systems, the types of
scrubbing media that have been used are either a
slurry (lime, limestone, or alkaline fly ash) or a
clear liquid (sodium carbonate, double alkali, lime
chloride and dilute acids). Regenerable systems use
clear li'quids mostly (sodium sulfate, ammonia,
citrated potassium trisulfate). The alkali/
magnesium oxide, has been used as a scrubbing slurry
also.
o Process D e s i g_n P a .ram e?ib g r_ s^ - Parameters such as
liquid- to-gas ratio (L/G), gas velocity, scrubbing
medium, flow configuration (counter, cross or
cocurrent), and pressure drop for the entire
scrubbing system vary considerably for the FGD
processes. These variables have a large effect on
the PGD system operating costs.
2-16
-------
A. Wet scrubbing tfarowaway processes
In the throwaway processes, the product sulfur removed
is waste. Usually the waste stream is a sludge
containing sulfur compounds, unreacted alkali, fly ash,
and water. Table 2-1 shows some of the commercially
available FGD processes. Further discussion is limited
to wet lime/limestone processes.
Wet lime/limestone processes
Wet lime and limestone systems make up a large
portion of the operating FGD scrubbers. Since
commercial installations have been operating for
several years, the overall operability and
reliability of both systems has been proven. The
basic process is fairly simple and very few process
steps are involved. In the first step, hot gases are
quenched to saturation temperature and passed along
to the scrubbers where a lime or limestone slurry
contacts the gases for SC>2 removal. The cleaned
gases are reheated and discharged to the atmosphere.
The lime/limestone scrubbing slurry is prepared using
slakers (lime) or ball mills (limestone). The
scrubbing slurry is collected in a tank and
recirculated to the scrubber. A purge stream
containing the net make of sulfite-sulfate can be
oxidized further by air or oxygen prior to entering
the thickener, for separation of the suspended solids.
Thickened slurry containing 20 to 40 percent solids
can be sent directly to a disposal pond or it can be
filtered. Filters are used for further reduction of
sludge volume by increasing the solids content to
about 70 wt$. Clear liquid is returned to the
system.
2-17
-------
CO
W
CO
co
Cd
O
C3
fa
P-.
w
0
S
0
e
CO
fa
a>
.
• O
0 \
i-H O
2-18
-------
The process has the following characteristics:
o
o
One of the lowest capital costs, as specific economic
studies have shown (3.).
Lime/limestone processes are not adversely affected
by fly ash in the system. They can remove both
S02 and particulates.
SC>2 removal efficiencies are generally high -
greater than 90% (.4,5.).
Reserves of lime/limestone are abundant in the U.S.
Some of these processes, if not designed properly,
have problems with plugging, scaling and corrosion .
Large liquid-to-gas ratio or substantial gas pressure
drop occurs in some processes.
The large quantity of waste requires a large, lined,
disposal pond.
Figure 2-2 illustrates a typical process flow diagram
for lime/limestone scrubbing.
2.4.2.2 SC>2 control by semi-dry scrubbing.- The concept of
semi-dry scrubbing for removal of sulfur dioxide from
flue gases is relatively new to the electric power
industry. As of late 1979 no commercial scale, semi-dry
scrubbing sytems were in operation. However, several
semi-dry scrubbing units are currently on order for
installation at'coal-fired power plants in the United
States.
2-19
-------
Cn
o
w
-------
Semi-dry FGD systems achieve SC>2 and participate
removal by using the combination of a spray dryer
followed by a conventional particulate control device.
Figure 2-3 illustrates the general process flow scheme
for semi-dry scrubbing with lime. Sulfur dioxide
removal takes place in the spray dryer, where the flue
gas that enters flows through a finely atomized spray of
alkali scrubbing solution. This solution uses either a
sodium based (soda ash) or calcium based (lime) type of
alkali. The SC>2 alkali reaction takes place between
the gas and liquid droplets and/or between the gas and
entrained particulates (6_,V • At the same tlme> the
water in the scrubbing slurry droplets evaporates due to
the-thermal energy provided from the incoming flue gas.
The treated flue gas leaves the spray dryer at a
temperature of about 28°C (50°F) above its dew-point or
about 77°C (170°F). This temperature can be controlled
by properly selecting the liquid-to-gas ratio at about
0.4 X 10-4 M3/ SM3, (0.3 Gal/MSCF). The
reaction product is a relatively dry, powder mixture of
sulfite, sulfates, and unreacted alkali. Collection of
the reaction products containing sulfur and particulates
is accomplished by using either an ESP or a fabric
filter (baghouse). The fabric filter also serves for^
additional S02 removal by reaction between the gas
and unused sorbent. Spent sorbent is usually
discarded.
Operating experience with spray dryers on utility
boilers has been limited to pilot-scale test programs
(8,9). Test results in these programs have indicated
S02 removals between 50 and 90% with lime alkali at
stoichometric ratios (SR) of 0.8 to 2.0. (SR = moles
lime/mole SC>2 entering). Process evaluations also
indicate that semi-dry processing is economically
attractive only in applications with moderate S02
2-21
-------
O 0)
en
w xJ
OJ -P
O -H
01
-------
removal requirements (approximately 70% max.) and with
coals of low to medium sulfur content. Some of the
characteristics of semi-dry scrubbing include:
o Dry powdered waste - no sludge
o High turndown capabilities
o No wet-dry interfaces
o No plugging or fouling problems
o Low pressure drop
o Power consumption about 50% of most wet-scrubber
consumption
o High reliability
o No corrosion problems ^
o No reheat requirements
The projected advantages of the semi-dry scrubbing
system and the positive results of pilot scale testing
have prompted utilities to give serious consideration to
spray drying as a means of SQ2 emission control.
There are several commercially available semi-dry
processes (6.) such as those offered by Wheelabrator-
Frye/Rockwell International; Joy Manufacturing Company's
Western Precipitation Division in conjunction with Niro
atomizer; Babcock & Wilcox; and Carborundum. Of these
four processes only Wheelabrator-Prye/Rockwell has both
regenerable and throwaway processes. The other three
are offered only as throwaway processes.
Characteristics of some of the semi-dry scrubbing units
currently on order are listed in Table 2-2.
2-23
-------
co
w
6-1
CO
X
tn
H
33
c o
boB
•H a>
co ce
Q)
a CM
ol
co|
I iH i-l
s- in a
o >J-5
J3 t. C
^ X
O O
O O
O rH
£1 -H
M (X)
O
L
•r)
O
•-3
O
in
O
CO
fc.
OS
O
CO M
O E-i
M O
E-i S
CO K
K CO
Cd 2
E-" O
O O
H CO
CO 63
CC
o a,
2
M
CQ .
aq
o
cc
o
CO
cc
Q
SL-
ID •
t-l bO
•H C
O -H
CD
CO
CC
ra j
o v«
•H 3
JJ *3
m co
••-<
t.
0}
*» £
o ra
a
ca
o
n)
o
o co
o
co
o
o
o
in
o
in
•
VO
to
e
o
in
in
o
o
a>
e
in
in
o
o
o
•
CO
s
Cd
co
CM
t
CM
Cd
J
CQ
>> C
4J O
r-1 -P
•H n)
o
o
CD
O
CU
•I-l
n)
a)
jj
aJ
O
n)
O
<-.
o
a> .-i
Cd
C O
•H >
to -H
cd (K
a>
o.
o
o
•I-l
-p
O i-l
rH
Cd t>>
CD
O iH
•H i-l
CO cd
cd S-
m
2-24
-------
A. Semi-Dry SC>2 Scrubbing Throwaway Processes
1. Wheelabrator-Frye/Rockwell International
The Wheelabrator system (spray dryer/baghouse) when
operating in an open loop is offered as a throwaway
process. The spray dryer is equipped with a multiple
atomizer and a roof dispenser. The alkaline scrubbing
slurry is a sodium carbonate solution. The Coyote
Station located near Beulah, North Dakota (10) has this
semi-dry PGD system scheduled for commercial operation in
1981. This system will serve a 410 MW lignite-fired
unit. The process flow diagram is shown in Figure 2-4.
2. Joy/Niro .
The alkaline scrubbing slurry uses either soda ash or
lime as alkali. The spray-dryer design uses a single
atomizer and compound gas dispenser to achieve adequate
gas mixing. Joy/Niro system (spray dryer/baghouse) has
developed a recirculation system in which lime slurry is
mixed with recirculated fly ash and spent reagent, for
reinjection into the spray-dryer (patent pending) the
process flow diagram is shown in Figure 2-5. The semi-
dry FGD system using a lime scrubbing slurry is being
installed at the Antelope Station Unit 1, a (440 MW gross
capacity). It is scheduled for commercial operation in
1982.
3. Babcock & Wilcox
This system (spray dryer/ESP) consists of a horizontal
reactor using a "Y" jet dual fluid atomizer followed
directly by an electrostatic precipitator. The system
uses a lime scrubbing slurry. The Laramie River Station
2-25
-------
!S
o
a
•8
H
(1)
-------
•H
S3
s-
-p
10
>1
0)
tn
C
•H
o
(0
•H
g
0)
CO
I
*
in
CM
Q)
I
•H
2-27
-------
Unit 3, whose gross capacity is 575 MW, is installing this
system (6.). Commercial operation should start in April
1982. The process flow diagram is shown in Figure 2-6.
4. Carborundum
A De Laval spray dryer in combination with a baghouse is
used for this system (6_). A test program using the process
is underway presently at the pilot plant located at Leland
Olds Station, Unit 1. The test program includes uses of
either dual, fluid-spray nozzles or a rotary atomizer. No
commercial operation plant is scheduled.
2.5 EMISSION MONITORING
In order to assure that the lowest NOX, particulate, and
S02 emission levels are achieved in a boiler system,
emission monitoring systems must be used. Currently no
continuous monitoring systems have been developed for
measuring mass particulate emissions. However, a variety of
instruments are available for measuring, indicating, and
recording opacity. These instruments as well as methods and
critera for evaluating performance are given in Reference 11.
Monitoring S02 and NOX emissions in units of mass
per unit heat input involves integrating the output of
S02 and NOX sensors with the output of C02 or
02 sensors. Various S02, NOX, C02, and
02 sensing instruments, that are suitable for this
purpose, are described in Reference 11. Reference 11 also
discusses how the various outputs are integrated to furnish
continuous indications and records of S02 and NOX
emission levels. In addition, the reference describes the
methods and criteria for evaluating the performance of
S02 and NOX emission monitoring systems.
2-28
-------
o
o
O
O
0
•8
ffl
CO
>1
(0
•H
i
M
O
Cfl
•d
I
•H
g
(1)
CQ
vo
I
CN
0)
Cn
-H
2-29
-------
2.6
REFERENCES
10
11
EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed NOX Emission
Standards." EPA-450/2-78-005a, July 1978
EPA, "Electric Utility Steam Generating Units -
Background Information for. Proposed Particulate Matter
Emission Standards." EPA-450/2-78-006a, July 1978
EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed S02 Emission
Standards." EPA-450/2-78-007a, August 1978.
EPA, "Alkali Scrubbing Test Facility; Summary of
Testing Through October 1974." EPA-650/2-75-041, June
1975
Lime/limestone wet-scrubbing test results at the EPA
alkali scrubbing test facility capsule report. NTIS
PB-258804, May 1975
Dickerman, J.C., et al, "Evaluation of Dry Alkali FGD
Systems" Radian Corporation. DCN 78-200-226-03. 31
March 1978.
"Spray-Dryer System Scrubs S02", Power Vol. 123.
No. 1, January 1979
Estcount, V.F. .et al, "Test of a Two-Stage Combined
Dry Scrubber/S02 Absorber Using Sodium or
Calcium", Proceedings of 40th annual meeting American
Power Conference, Chicago, Illinois, April 26, 1978
Janssen, K.E., Eriksen, R.L, "Basin - Electric's
Involvements with Dry Flue Gas Desulfurization." EPA
symposium on FGD - Las Vegas March 5-8, 1979.
"Coyote Station. First Commercial Dry FGD System",
Presented at the 41st Annual meeting APE, Chicago,
April 23-25, 1979
Handbook - "Continuous Air Pollution Source Monitoring
Systems", EPA 625/6-79-005, U.S. Environmental
Protection Agency, Cincinnati, Ohio, June 1979
2-30
-------
A. Letter N. Kotraba, Apitron Division, American Precision
Industries to D.R. Goodwin, U,S. Environmental Protection
Agency, Research Triangle Park, North Carolina,
February 29, 1980.
-------
-------
SECTION 3
RETROFIT DESIGN AND COSTS
3.1 GENERAL
The key pieces of equipment used to retrofit pulverized-
coal-fired, steam generators, for NOX reduction and for
control of S02 and particulates are fixed in
operational design. The functional design for sizing to
meet emission level requirements reduce relatively easily
to physical layout considerations and mathematical
analysis. Using the retrofit technology outlined in
Section 2, this section presents the guidelines for
determining retrofit costs. It also presents the
equations for prorating to other design conditions. This
is the basis for estimating costs for any desired- retrofit
situation. A typical schedule for retrofitting these
plants concludes the discussion.
3.1.1 Emissions
The cost modules for NO,, reduction are based on the
best available technology associated with boiler
modifications to reduce NOX formation. This document
presents costs based on these modifications. Emission
levels of 210 ng/J heat input (0.5 lb/106 Btu) for
subbituminous coal and 260 nanograms per joule
3-1
-------
heat input (0.6 Ibs/lO^ Btu) for bituminous coal are
the basis of the modification costs.1 Actual
implementation of the modifications discussed previously
may not permit this emission level to be reached, but it
presents the best potential for NOX emission
reductions.
The costs for S02 control are based on achieving
S02 reductions in the flue gas of up to 90$.2 For
particulate control, the cost modules are based on
achieving emission levels of 13 ng/J heat input (0.03
lb/106 Btu).3
3.1.2 Basis of Costs
The costs of an emission control systems are estimated as
capital costs and annualized cost. The capital cost
represents the initial investment necessary to install and
commission the system. All costs are based on 3rd-quarter
1979 dollars. Annualized costs represent the cost of
operating and maintaining the system and the charges
needed to recover the capital investment, which are
referred to as fixed costs. The cost of land for sludge
disposal is not included in this study. Land used for
sludge disposal is considered to have•zero value once
sludge disposal at that site has ceased.
Capital costs consist of direct and indirect costs
incurred up to the tie-in and startup of the retrofit.
Direct costs include the costs of various items of
equipment and the labor and material (construction costs
including field overhead) required for installing these
items and interconnecting the systems. Indirect costs
3-2
-------
include such items as freight, procurement, and
allocated costs associated with the purchase and
installation of the control equipment.
3.1.2.1 Direct costs.^ - The purchased cost of the
equipment and the cost of installing it are considered
direct costs. The cost of an equipment item is the
purchase price paid to the equipment supplier on a
free-on-board (f.o.b.) basis; this does not include the
freight charges. Installation costs cover the
interconnection of the system, which involves pipjng,
electrical, and the other work needed to commission it
such as the cost of securing permits arid the cost of
insurance for the equipment and personnel on site. The
costs of foundations, supporting structures, enclosures,
ducting, control panels, instrumentation, insulation,
painting, and similar items are attributed to
installation. Costs including site development,
relocation or alteration of existing facilities,
administrative facilities, construction of access roads
and walkways, and establishing rail, barge, or truck
facilities have not been included in developing the
retrofit costs except as noted; they must be determined
on an individual basis for a specific plant.
3.1.2.2 Indirect costs.^- The indirect costs include
freight from point of origin and indirect capital costs.
The indirect capital costs consist of several cost items
which are calculated as percentages of the total
installed cost (TIC), the direct costs as noted above.
The indirect capital costs include the following items:
3-3
-------
A. Interest - Interest covers costs accrued on borrowed
capital during construction. (About 10$ of the TIC.)
B. Engineering costs - These costs include administra-
tive, process, project, and general costs; design and
related functions for specifications; bid analysis;
special studies; cost analysis; accounting; reports;
procurement; travel expenses; living expenses;
expediting; inspection; safety; communications;
modeling, pilot plant studies; royalty payments
during construction; training of plant personnel;
field engineering; safety engineering; and consultant
services. (About 10% of the TIC.) ..
C. Taxes - Include sales, franchise, property, and
excise taxes. (About 1.45? of the TIC.)
Allowance for shakedown - Includes costs associated
D,
E,
F.
with system startup. (About 5% of the TIC.)
Spare parts - Represent costs of items stocked in an
effort to achieve 100 percent process availability;
such items include pumps, valves, controls, special
piping and fittings, instruments, spray nozzles, and
similar equipment not included in base cost modules.
(About 0.5% of the TIC.)'
Contingency costs - Includes costs resulting from
malfunctions,- equipment design alternations, and
similar unforeseen sources. (About 20% of the TIC).
Contractors fee and expenses - Includes costs for
field labor payroll, supervision field office,
administrative personnel, construction offices,
temporary roadways, railroad trackage, maintenance
and welding shops, parking lot, communications,
temporary piping, electrical, sanitary facilities,
rental equipment, unloading and storage of materials,
3-4
-------
travel expenses, permits, licenses, taxes, insurance,
overhead, legal liabilities, field testing of
equipment, and labor relations. Contractor fees and
expenses are about 5% of the TIC. The indirect cost
for a given estimate is about 58.6$ of the TIC.
Indirect costs have been added to all costs
presented in this document.
3-5
-------
3.2 RETROFITTING TO REDUCE NOX EMISSIONS
The effectiveness of applying currently available
retrofit control for N0v emissions to new coal-fired
6
power plants is 210 ng/J heat input (Q. 5 lbs/10
Btu) for subbituminous coal and 260 ng/J heat input
(0.6 lbs/106 Btu) for bituminous coal.1 However, these
levels may not always be achievable for existing units as a
result of intolerable adverse side effects. For new units
adverse side effects can be avoided by proper original
design, but with existing units it is more difficult to
apply the techniques while avoiding effects are discussed in
Section 2.2.
Expert advice from steam generator manufacturers and/or
combustion engineers is recommended in conjunction with decision
making on best available retrofit technology for N0x
control.
3-6
-------
3.2.1
Retrofit Techniques for NOY_Control
3.2.1.1 Plant data requirements.- When considering retrofitting
a particular boiler for NOX control in.a plant, the
following information related to existing boiler design
and operation should be gathered:
o Type of boiler (single-wall, opposed-wall,
tangential, or arch-fired)
o Manufacturer of the boiler
o Type of existing burners (arrangment, burner type,
and burner capacity)
o Existing NOX control and monitoring equipment
o Drawings of burner arrangement,
o Existing NOx emissions level, and State NOx
emissions limit
3-7
-------
3.2.1.2 Design parameters.- There are several key elements
involved in retrofitting steam generators for NOX
control. Each element is addressed separately in the
following paragraphs to clarify its importance.
A- Type of boiler - Steam generator designs vary
between manufacturers. Also individual
manufacturers may offer more than one type of
design, and they change their designs over the years
to accomodate changed fuels, design improvements,
and government regulations. • Consequently, retrofit
designs for coal-fired steam generators have to be
customized for the needs of each unit.
B. Overfire air - Overfire air may range from 15 to 30%
of the total air. The suggested value depends on
the manufacturer and type of boiler. Overfire air
is effective for Poster Wheeler, Riley Stoker, and
C-E boilers and is one of the key elements for
NOY control in these units. B&W (5) does not
A. ~™~
recommend overfire air. When applying overfire air,
ports must be cut in the firebox, and the wall tubes
have to be modified to make space for the air port
and tilt mechanism. Additional windbox
modifications for connecting the duct to the
overfire air ports and the individual air control
dampers are required to complete the overfire air
port retrofit. The overfire air jet velocity used
is 61 M/sec (200 feet per second), to provide
sufficient jet penetration depth.
3-8
-------
D,
Curtain air ports - Curtain air ports are applied
only to the Poster Wheeler and Riley Stoker boilers.
The main objective in using curtain air ports is to
minimize slagging and tube wastage problems. The
same air jet velocity, 61 M/sec (200 feet per
second) used for overfire air is used in sizing the
curtain air ports. The addition of ducts and
individual air-flow control dampers is needed in
addition to the wall tube modifications and firebox
wall cutting. Curtain air ports are designed and
' installed in a manner to surround the bottom, the
right and the left sides of the entire burners.
Overfire air ports occupy the top row of burners.
The amount of curtain air used is 10% of the total
air and this is distributed equally among all
curtain air ports. Again, the application and
effectiveness of using curtain air ports are
dependent on the type and year of manufacture of the
boilers.
Low NOX burners - B&W low NOx burners have
been tested and can be applied to existing units.
Poster Wheeler data on low NOX burners indicate,
it may not always be effective or possible to apply
low NOX burners to existing old units. When
retrofitting existing plants with low NOX
burners, rearrangement of the burners in the firebox
wall might be necessary. This modification involves
several operations including:
modification of the membrane t u b e s to
accomodate the low NOx burners
modification of the membrane tubes to close
the holes where the original burners existed
re-piping the coal feed lines from the
pulverizer ;
3-9
-------
o modification of the windbox to add a
compartment for better control of air/fuel
ratios
o tuneup or retrofitting the NOx emission
monitoring system.
E. Combustion air control
o Babcock & Wilcox uses a compartmentalized windbox
to provide secondary air flow metering and flow
control for each pulverizer. The result is a
rigid coal/air ratio control to each burner group
and flexibility to operate with lower total
excess air while maintaining an oxidizing
atmosphere around each burner. The Dual Register
burner and the compartmentalized windbox are
coupled together for this system. Excess air in
the system ranges from 15 to 20%.
•o C-E boilers use twenty percent excess air. The
air/fuel ratio is kept slightly above
stoichiometric at the burners. Twenty percent of
the total air is used as overfire air.
o Twenty percent excess air is used for Poster
Wheeler boilers. The percentage of overfire air
is 30$ and curtain air is 10$ of the total air.
o Twenty percent excess air is used for R-S
boilers. Overfire air is 30$ and curtain air is
10$ of the total air.
3-10
-------
3.2.1.3 Retrofit calculation procedures. -
A. Dual Register Burners for Babcock & Wilcox Steam
Generators - As mentioned in the previous sections,
for the B&W boilers, the Dual Register burners
coupled with modified windbox compartments is the
technique recommended for NOX control. The
existing cell burner, and each coal nozzle must be
replaced by an individual dual register burner. The
dual register burners have a rating equal
approximately that of a coal nozzle or one-third of
that of a cell burner, and therefore, the total
number of the Dual Register burners installed will
be three times that of the cell burn^^s.
B. Sizing of Overfire Air (OPA) Ports & Curtain Air
(CA) Ports - The burner arrangement dictates the
number and location of overfire air ports and
curtain air ports. For this study C-E, F-W, and
Riley Stoker boilers are retrofitted with
overfire-air ports, but only P-W and Riley Stoker
boilers will be retrofitted with curtain air ports.
The location of the overfire air ports is, in
principle, always at the top of each topmost burner
of -each vrtical burner column, for all three
boilers. (see Figure 3-D In the figure, (A) is an
example of the normal location of the OPA port in a
C-E boiler. where the firbox structure does not
allow the OPA ports to be located directly on top of
burners, the OPA port is offset but near the burner
top. C-E has retrofitted OPA ports in a manner
similar to (b) for NOX control tests at Alabama
Power Company's, Barry Unit 2 (6).
3-11
-------
o
(A)
OFA PORT
(TYP)
(OFFSET
CONFIGURATION)
OR
BURNER
(TYP)
O
(B).
Figure 3-;l- Location of overfire air ports for C-E boilers,
3-12
-------
3.2.2
Figure 3-2 shows the arrangement of the curtain air
ports for the Poster Wheeler and Riley Stoker
boilers.
Prom Pigure 3-2, you can see that the number of OPA
ports is equal to the number of burner columns, N,
and the number of curtain air ports is (N + 2M) where
M = number of burners per column. The size of each
OPA port is based on the overfire air required at an
inlet velocity of 61 M/sec (200 Pt/sec).
In order to minimize the number of wall tubes which
will be affected and to minimize impairment of the
firebox wall in retrofitting OPA ports, a 2:1 ratio
of height and wi-dth is used.
Retrofit Costs for NOY Control
The following sections provide guidelines for
estimating the costs of retorfitting for NOX
control. These cost estimates should be supplemented
whenever possible by consultation with boiler
manufacturers and/or other experts on boiler
modifications.
3.2.2.1 Burner module cost.-
A. Retrofit Capital Cost -
The low NOV burner module costs for B&W retrofit
A.
are given in Table 3-1. A boiler manufacturer's fee
of $60,000 for an engineering analysis of the firebox
should be added to the total retrofit cost for the
boiler once the burner module costs are determined.
-------
N=No. OF BURNER COLUMN-
X
TV.
X X
X X X X
X
X
OFA PORT
(TYP)
x
X X XX
M= No. OF
BURNER IN
ONE COLUMN
X : REPRESENTS BURNERS
: REPRESENTS CURTAIN AIR PORTS
Figure 3- 2.- Arrangement of curtain air ports, for F-W and R-S boilers,
3-14
-------
TABLE 3-1.-'B&W LOW NOX BURNER
Burnera(MW=3.5)
Piping
Instruments
Insulation, paint
Freight
Construction Cost
Total Installed Cost
Indirect Capital Costb
Total Module Cost Per Burner
COSTS
$ 61,800
1,000
8,000
100
4,100
59,000
134,000
79,000
$213,000
B. Annual Costs -
Annual costs for B&W low NOX -burner retrofit
systems should be estimated at 17.2 percent of the
total capital cost. No allowance should be made for
maintenance because maintenance costs for-the new
burners should be much the same as maintenance costs
for the old style burners. No allowance is made for
additional operation labor or energy costs.
3.2.2.2 Overfire air ports.-
A. Retrofit Capital Cost -
The cost of an overfire air port for C-E, FW, and
R-S retrofits includes port and membrane wall tubes
modifications, port tilt mechanism, air flow
controls, windbox and ducts additions, and the boiler
manufacturer's engineering analysis of the firebox. A
boiler manufacturer's fee of $60,000 for an
engineering analysis of the firebox should be added
to the total retrofit cost for the boiler once the
burner module costs are determined. The module cost
of an overfire air port is given in Table 3-2.
aBurner cost based on Dual Register burner, three burners
required for each cell burner replaced
bSee Section 3.1.2 for definition of indirect capital costs
3-15
-------
TABLE 3-2.- OVERPIRE-AIR PORT COSTS
Port fabrication $ 12,500
Duct work and windbox modification 4,500
Instruments and controls 6,500
Freight 1,500
Subcontract 4,500
Construction Cost 56,000
Total Installed Cost 85,500
Indirect Capital Costa 50,100
Total Module Cost Per Port $135,600
B. Annual Costs -
Annual costs for overfire and curtain air ports
should be estimated at 22.2 percent of total capital
costs. This allows 17.2 percent for amoritization
and 5.0 percent for maintenance and supplies.
3.2.2.3 Curtain Air Ports.-
A. Retrofit Capital Cost -
Cost of a.curtain air port is taken as 63$ of an OPA
port or $ 85,000/port. The boiler manufacturer's
engineering analysis should be included in the OPA
ports costs as given in Section 3.2.2.2
B.
Annual Costs -
See Section 3.2.2.2 B
3.2.2.4 Combustion air fan.- No costs for additional combustion
air fans are estimated for special burner, overfire air,
and curtain air modules.
aSee Section 3.1.2 for definition of indirect capital costs
3-16
-------
3.3
RETROFITTING DUCTWORK AND STACKS
For most existing power plants, there will not be enough
space to locate retrofit systems between the steam
generator and the existing stack. Although it might be
possible to remove existing air pollution control
systems to make room for the retrofit system between the
steam generator and the stack, this would usually make
it necessary to shut down the steam generator for more
than one year while the retrofit system is being
•installed. Consequently, retrofitting will usually
involve installing the retrofit system beyond the
existing stack. This will increase the length of duct
work required to retrofit an air pollution control
system to an existing steam generator and stack.
The additional length of ductwork required for
retrofitting is the length of ductwork needed to connect
the retrofit system to the existing ductwork near the
stack and the length of ductwork needed to return the
cleaned gases from the retrofit system to the existing
stack. Otherwise the lengths of ductwork needed for a
new or a retrofit air pollution control system are much
the same.
Analysis of the costs of ductwork for retrofit
situations indicates that 15 percent of the total
capital costs of a new air pollution control system is a
liberal allowance for the additional cost of ductwork
for a retrofit system.iS,!1* In special cases where
space limitations make it necessary to locate the
retrofit air pollution control system remote from the
existing stack,, it may be less costly to erect a new
stack in conjunction with the retrofit system thereby
decreasing the length of ductwork needed to cojmect the
retrofit system to the stack. Table 3-3 shows estimated
capital costs of new stacks.15
3-17
-------
The diameter of stack needed should be estimated using a
gas volume of 3760 actual cubic feet per minute per
gross megawatt of steam generator capacity and a linear
gas velocity at the outlet of 65 feet per second.
In cases where S09 control systems are to be retrofitted
£* —
it may be necessary to reline or to replace the existing
stack to provide for corrosion resistance. These costs are
not included in the cost estimates of this document.
TABLE 3-3. Values of A and B for Estimating the Cost
of Utility Boiler Stacks
Inside
Stack Diameter
at Outlet
Feet
15
20
30
40
15
A
0.185
0.393
2.184
4.377
B
2.625
2.535
2.330
2.262
Where Y = AHD
Y = Capital cost (direct and indirect) third quarter
1979 dollars
H - Stack height - feet (range between 250 and 1200
feet)
Costs include - concrete shell, foundations, and steel
liner
Designed for -
Windload - 40 pounds per square foot
soil bearing - 4500 pounds per square foot
seismic conditions - Zone 1 (minor risk)
3-ia
-------
3.4 RETROFITTING TO CONTROL PARTICULATE EMISSIONS
3.4.1 General
Electrostatic precipitators and baghouses are most
commonly used for high efficiency removal of
particulates from the combustion gases of coal-fired
steam generators. When flue gas desulfurization systems
are required, particulate scrubbing is often
incorporated into the air pollution control system,
although, of course, scrubbing can be used for
particulates in systems when sulfur is not a problem.
When an ESP is installed before the air preheax,er, it is
called a hot-side ESP. When it is installed between the
air preheaters and stack, it is called a cold-side ESP.
Baghouses for high efficiency particulate control have
become more widespread recently, especially when the
coal ash is difficult to collect with an ESP. Baghouses
are located downstream of the air preheater.
There are other particulate collectors such as cyclones
and settling chambers, but they are not efficient enough
to reduce particulate emissions to the levels required
by current new source performance standards.
The three alternatives to be explored for this
particulate emissions control retrofit study are:
0 Upgrading either cold-side or hot-side EPS systems
0 Installing baghouses
0 Installing scrubbers in conjunction with flue gas
desulfurization while retaining the existing
particulate controls.
3-19
-------
Various levels of particulate emission control can be
achieved by retrofitting ESP's based on various specific
collection areas. The efficiency achieved by baghouses has
been demonstrated to control particulate emissions to
levels less than 13 nanograms/joule (0.03 lb/106 Btu)
with an air to cloth ratio of 0.6l actual cubic meters per
minute per square meter (2 ACPM/ft2) and with pressure
drops of less than 1.25 kilopascals (5 in. H20) in
the filter at the full-load gas volume.3
The EPA position on aspects involved with the operating
effectiveness of the electrostatic precipitator and
baghouse systems of this document are discussed in
Appendix E of these guidelines.
Unlike techniques for NOX emission reduction, the
principal systems for particulate emissions control -
ESP's and baghouses, can be applied to all four
manufacturer's boilers. Key elements in retrofitting an
existing plant for particulate emissions control include
the following items:
o The total -collection area and arrangement of the
existing particulates collection equipment (ESP's or
baghouses) must be known.
o The available plant space for retrofitting ESP's, or
baghouses, or scrubbers must be established and must
be determined to be adequate. Unlike NOX emission
reduction retrofits, the retrofitting for particulate
emissions requires significant additional plant space.
Relocation of existing equipment, and addition of flue
gas ducts and fans may be involved. The existing plot
plan affects the retrofit cost because the choice of
arrangement for the retrofitted equipment is very much
3-20
-------
dictated by it. Detailed information about the
avilability of additional space is a vital element for
an economic retrofit.
Plue gas flow rate and its temperatures in and out of the
air preheater and out of the economizer must be known.
These data govern the size of the retrofitted system.
3-21
-------
3.4.2 Electrostatic Precipitator Design
The design of a new ESP system should be based on the
performance data of the existing ESP system'firing the
same coal whenever possible. In case there are no such
data available, coal ash analysis provides valuable data
for selecting the size and type of ESP required to achieve
the desired efficiency. Ash resistivity is a major factor
affecting the size of an ESP system. The resistivity of
ash entering the air preheater is much less when it is at
temperatures of 316 to 482°C (600 to 900°P) compared to
its resistivity at the air preheater outlet where the
temperature is about 149°C (300°F). Therefore, with high
resistivity ash coal, it is easier to collect on the hot
side than on the cold side. The following factors must be
taken into account in deciding whether a hot-side or a
cold-side ESP system is to be used (3) :
o Space available.for the retrofit system
o Temperature-resistivity characteristics of the fly ash
o Specific collection area requirements
o Differences in gas volume caused by temperature
differences
o Differences in gas volume caused by air leakage into
the air preheater
o Differences in construction requirements caused by
temperature differences
Analysis of the data on ESP's in Reference 3, shows that
ESP systems can limit particulate emissions from steam
generators to levels less than 13 nanograms per joule
(0.03 lbs/106 Btu). The size of the system required
to meet a given emission level depends upon ash
characteristics and the level of control required.
3-22
-------
The factors that have been discussed dictate that the
features in the following list need to be included in ESP
retrofit systems for control of particulate emissions from
large steam generators.
o Either a cold-side system or a hot-side system is
chosen and the choice depends on circumstances such as
the additional space required and available, and the
possibility for installing the ductwork on either the
entrance side or the downstream side of the air
preheater.
o Sufficient electrical .^ectionalization should be
included in the ESP system to ensure that adequate
collecting surface area will be available should a
breakdown occur in one of the sections.
o Automatic power controls should be provided as well as
instrumentation showing; 1) primary voltage, 2) primary
current, 3) secondary voltage, 4) secondary current, 5)
spark rate for each individual section.
o Insulation should be sufficient to minimize.temperature
drops which would cause acid attack.
o Enough flue-gas-flow, cross-sectional area should be
included to ensure maintaining the system in the high
collection efficiency range.
o Provisions should be made for good flue gas velocity
distribution in the gas passages, even at partial
loading.
o Sectionalization of the rapping system for difficult
dust should be about 139 square meter-s (1,500 ft2)
of collecting plate area per rapping section.
o Separate electrical Sectionalization should be about
one section for each 5 MW of gross generating capacity.
Retrofitting can achieve the following limits indicated in
Table 3-4 that follows (3_) :
3-23
-------
TABLE 3-4.-ESP SPECIFIC COLLECTION AREA (SCA) FOR VARIOUS GOALS3-
Coal Sulfur
Content %
.8
.8
.8
.8
.8
.8
3.5
3.5
3.5
Emission Limit
(lb/10 Btu)
.03
.03
.05
.05
.10
.10
.03
.05
.10
ESP Type
Hot
Cold
Hot
Cold
Hot
Cold
Colda
Colda
Colda
SCA 3
Ft2/1000 ACFM
650
1000
550
800
400
650
360
300
240
aHot side ESP's not normally used with 3.5$ S coal
The following data can'be used for estimating the flue
gas volume once the gross MW capacity of the unit is
known:
Coal Heating Value
(as burned):
Heat Rate:
"F" Factor (11):
Excess Air at Economizer:
Flue Gas Temperature out
of Economizer (Hot Side)
Variable, depending on
the coal
2.924 X 103 J/KWSec
(10,000 Btu/KW Hr)
2.64 x 10-7 Dry
Standard Cubic
metres per joule
(9,820 SCF/106 Btu)
at 0% excess air and
20°C (68°F)
25$
343°C (650°F)
-------
Excess Air Out of Air
Preheater:
Flue Gas Temperature out of
Air Preheater (Cold Side):
Water Content of Flue Gas:
C (300°P)
10$ of dry gas volume
3.4.2.1 Hot-side precipitator.- The major problem encountered
in retrofitting hot-side ESP systems to an existing
plant is insufficient space for fitting the hot-side ESP
between the economizer and air preheater. If there is a
space problem, either the air preheater has to be moved
and relocated to allow for duct work, or considerations
must be given to retrofitting with cold-side ESP
systems.
Duke Power Company has retrofitted both hot-and cold-
side ESP systems at their utility plants (16). .Duke
Power's hot-side ESP retrofits to their C-E boilers was
achieved without relocating the air preh-eaters because
space was available for ducting tie-ins.
For Duke Power hot-side retrofits, the flue gas duct was
partitioned into two parts at the horizontal section in
front of the air preheater. Hot flue gas was guided
into a new duct, stemmed from one side of the partition,
and was routed to the hot-side ESP system. The return
gas from the ESP system was directed into a new duct
that lead back to the other partition and flowed into
the air preheater.
When the existing gas passage ducting does not allow for
any modification, the only alternative to retrofitting
with hot-side ESP systems is to remove the air preheater
and relocate it in order to provide enough space for the
duct work. This installation involves shutdown time.
The retrofit work must be based on individual designs
for each situation.
3-25
-------
3.4.2.2 Gold-side precipitator.- Retrofitting cold-side ESP
systems to an existing plant involves connecting and
adding ducting to the existing flue-gas outlet of the
air preheater or the existing ESP. The exit flue-gas
duct from the ESP systems is connected to the ID fan,
stack, or scrubber system depending upon plant design.
3.4.2.3 Flue Gas Conditioning
The specific collecting surface area values of Table 3-4
do not reflect the possible beneficial effect of flue
gas conditioning agents such as sulfur trioxide,
sulfuric acid, or ammonia. Although it is possible that
use of these agents would reduce the size of the ESP
required to meet a given level of control, the
conservative estimates of Table 3-4 should be used for
cost estimating. However, if an owner elects to use
flue gas conditioning to implement a best available
retrofit technology decision, this should be allowed.
with a provision that the emission limitations of the
BART decision must be met.
3.4.3 Baghouse Design
The type of baghouse considered in this retrofit -study
is the inside-out, multicompartment, reverse-air design.
Of course other baghouse designs, such as outside-in
filtration, pulse-jet cleaning, and shake cle'aning can
be used as alternatives.
The design criteria and design features of the baghouse
systems to be used in this retrofit study are summarized
in the following list:
3-26
-------
o Air to cloth ratio is 0.61 actual cubic meters per
minute per square metre (2 ACFM/Pt2) to handle
full load gas volume at pressure drops less than
2.5 kilopascals (10" H20)
o Reverse-air cleaning
o Multicompartment system
o Provision for isolating each compartment for cleaning
or maintenance while the other compartments are on
line
o Provision for automatic cleaning on a compartment by
compartment basis with controls for adjusting the
quantity of reverse air, the frequency of cleaning,
and the duration of cleaning
o Provision for instruments indicatinp- pressure drop
and temperature in and out of the baghouse
o Adequate insulation to minimize temperature drops
which would cause acid attack on the baghouse
-------
3.4.4
Retrofit Costs For Particulate Control
3.4.4.1 Electrostatic Precipitators.- The following equation
should be used to estimate the capital cost of retrofit
electrostatic precipitator (ESP) systems.
Cold Side ESP Systems
y = 3635 (X)'6404
Hot Side ESP Systems
y = 414.5 (x)'8129
Where:
y = total capital cost in third quarter 1979
dollars
x = square feet of collecting surface area
These costs include the direct and indirect costs of
the ESP including ash removal and the additional cost
of retrofit ductwork (15 percent). The costs are derived
from the costs used to support the new source performance
•7
standards for new electric generating units.
The total collecting surface area required can be
determined using 1) the data of Table 3-4, 2) data on
the size of each steam generator to be studied, 3) data
on the ESP collecting surface area installed, and
4) data on the coal characteristics. With low sulfur
coal more collecting surface area is usually needed
than for high sulfur coal. In most cases it will not be
possible to retrofit a hot side ESP. However, since
some hot side ESP's have been retrofitted, cost data is
included. When the necessary data has been obtained
from the power plant, cost estimates can be made.
3-2g
-------
Once the levels of control to be studied are selected
(usually by those responsible for visibility impact
analysis) the specific collecting surface areas (SCA)
should be selected from Table 3-4. For cases not
covered by the table or where there is doubt that the
values of the table are applicable, ESP vendors should
be consulted for advice or engineering judgments should
be made. Total gas volume should be computed using the
following estimates.
-------
•! . !!
-------
Hot Side ESP Systems
Gas volume '= 4730 ACFM per megawatt at 650°F
Cold Side ESP Systems
Gas volume = 3760 ACFM per megawatt at 300°F
Although actual gas volume may be less than the foregoing
values, these estimates provide a safety factor for
cost estimating. These gas volume values should not be
used for enforcement purposes when test data show
different values.
Once the total collecting surface area requirements are
estimated using the data on gas volume and collecting
surface area, any installed collecting surface area
should be deducted to determine retrofit requirements.
The example on the Navajo plant of Appendix C shows' how
the foregoing technique is applied. Once the additional
ESP area has been estimated the capital cost of the
retrofit system should be computed usin.g^ the applicable
capital cost equation. *
Annual costs should be estimated using the following
equations:
Cold Side ESP Systems - y = 965.03 (x)0'6381
Hot Side ESP Systems - y = 111.54 (x)'8099
3-29
-------
y = annual costs - third quarter 1979 dollars
per year
x = retrofit collecting surface area required -
square feet
Annual costs can be estimated in units of mills per
kilowatt hour by dividing the dollar values by annual
power generation. In the absence of data from the plant,
annual power generation should be estimated assuming the
steam-electric generating system operates at 65 percent
of net (not gross) generating capacity during the year.
In determining net generating capacity the electric
power requirements of the retrofit systems should be
deducted from the net generating capacity of the plant
prior to retrofitting.
ESP retrofit costs should be estimated separately for
each steam generator. As previously stated, the costs
include both direct and indirect costs and provide an
allowance for the additional cost of retrofit ductwork.
Other costs that are not included but that might be
involved are discussed in,Sections 4.2, 4.4, 4.5, and 4.6,
The electrical energy requirements of ESP systems are
estimated to be 2.810 kilowatts per 1000 square feet of
collecting surface area added. This value or a value
obtained by consultation with ESP vendors should be
used to estimate the requirements for replacement of
electric generating capacity as discussed in Section 4.4.
3.4.4.2 Baghouses. - The following equations should be used to
estimate the capital cost of baghouses for any retrofit
situations:
y = 173420 (x)0'8384 without booster fan
or
y = 174987 (x)°-8563 with booster fan
3-30
-------
Where:
y = capital costs - third quarter 1979 dollars
x = size of baghouse - megawatts
These equations are applicable for full or partial gas
treatment. For example where a 500 megawatt unit is to
be retrofitted for filtration of 70 percent of the flue
gas, baghouse size would be 350 megawatts. A booster
fan should be provided in cases where only a baghouse .
is to be retrofitted. If the baghouse is to be
retrofitted in conjunction with S02 control, the cost
of the booster fan is included in the S02 control cost
estimate.
The baghouse capital cost equations include the 'direct
and indirect costs of the baghouse additional retrofit
ductwork, and the ash removal systems for a 2-to-l air
to cloth ratio, reverse air baghouse. The costs do not
include those discussed in.Section 4.2, 4.4, 4.5, and
4.6. The data needed for baghouse cost estimates are
data on the size of each steam generator (gross megawatts)
and the foregoing equations. The capital and annual
costs are derived from the cost data used to support the
new source performance standards for electric utility
steam generators and include a 15 percent allowance for
additional retrofit ductwork. Costs should be estimated
separately for each baghouse system.
Annual costs for baghouses should be estimated using
3
the following equations:
y = 31090 (X)'8494
without booster fan
y = 21934 (x)
.9358
without booster fan
3-31
-------
Where:
and
y = annual costs - third quarter 1979 dollars per year
x = baghouse capacity - megawatts
These costs can be converted to units of mills per
kilowatt hour using the techniques described in
Section 3.4.4.1.
The electrical energy requirements of baghouses are
estimated to be 6.615 kilowatts per megawatt of baghouse
13
capacity. This value should be used to estimate the
capital cost of replacing electric generating capacity
as discussed in Section 4.4 and for estimating unit
annual costs as discussed in Section 3.4.4.1.
Other costs that may be involved with retrofitting but
not included in the foregoing cost estimates are discussed
in Sections 4.2, 4.4, 4.5, and 4.6.
3.5 Retrofitting To Control SO- Emissions
3.5.1 Retrofit Costs For Wet S02 Control
Table 3-5 shows the values of A and b to be used in
estimating retrofit costs for various flue gas
2
desulfurization systems where:
and
y = Ax1
y = capital cost - third quarter 1979 dollars
x = size of system - megawatts
3-32
-------
TABLE 3-5.- VALUES OF A AND b FOR ESTIMATING CAPITAL AND
ANNUAL COSTS OF WET FLUE GAS DESULFURIZATION SYSTEMS
SYSTEM
CAPITAL COSTS
THIRD QUARTER
1979 DOLLARS
ANNUAL COSTS
THIRD QUARTER
1979 DOLLARS
PER YEAR
Lime FGD
Eastern 3-5%S
Eastern 7.0%S
Western 0.8IS
Anthracite
Lignite
1.715x10"
1.874x10°
1.450x10°
166.0x10:?
166.2X10-3
0.6612
0.6663
0.6546
1
1
494x10:
528.2x10:
405x10:
61.62x10:
61.92x10'
0.7107
0.7264
0.7052
1
1
Limestone FGD
Eastern 3.5%S
Eastern 7.0%S
Western 0.SIS
2.321x10
2.373x10
1.756x10
0.6375
0.6563
0.6455
656.4x10:
672.0x10:
508.3x10"
0.6803
0.7020
0.6828
Mag-Ox FGD
Eastern 3.5%S
Eastern 7.0%S
2.708x10
2.790x10
0.6279
0.6464
810.9x10:
811.6x10-
0.6623
0.6869
Double Alkali FGD
Eastern 3.5%S
Western 7.0%S
2.624x10
2.791x10
0.6194
0.6274
712.0x10:
735.6x10'
0.6745
0.6955
Wellman Lord FGD
Eastern 3.5%S
Western 7.0%S
2.573x10"
2.542x10°
0.6156
0.6307
784.1x10:
775.8x10'
0.6415
0.6539
3-33
-------
The costs include the direct and indirect costs of:
Feed storage and handling
Scrubbing
Plume reheat to 175°F
Liquor treatment
Sludge disposal
and Booster fans
The costs include one spare scrubbing module and booster
fan for each steam generator and allow 15 percent for
the additional cost of retrofit ductwork. The design
effectiveness of the systems shown in Table 3-5 is
90 percent removal of S07. The EPA position on aspects
Li
involved with the operating effectiveness of the wet SO-
scrubbing systems of this document is discussed in
Appendices D and E of these guidelines. The costs are
derived from the costs used to support the new source
performance standards for new steam electric generating
units.
3-34
-------
The cost equations can be used for partial scrubbing cost
estimates. The size of scrubbing system for partial scrubbing
should be calculated by the following equation:
SCR
x P
0.9
Where C
SCR
JSG
and
Size of scrubbing system for partial
scrubbing-megawatts
Size of steam generator-gross megawatts
Percent SCU removal desired
Once the size of the scrubbing system is computed, the capital
cost should be estimated using the applicable values of A and b
from Table 3-5. For most.cases the least costly system (usually
lime scrubbing) should be selected. Data on coal characteristics
and steam generator size (gross megawatts) are all that are
needed to estimate retrofit costs. Costs .should be estimated
separately for each steam generator.
For partial scrubbing of less than 75 percent of the flue
gases plume reheat is usually not necessary since the bypassed
flue gas will usually heat the combined gas from the scrubber
to at least the combined gas from the scrubber to at least
175°F. The capital costs estimated using Table 3-5 can be
adjusted by deducting $6000 for each megawatt of scrubber capacity
for cases where no plume reheat system is necessary. Case
2A of Appendices A, B, and C show examples of capital and
annual cost estimates for partial: scrubbing.
Annual costs should be estimated using the data of
2
Table 3-5 and the following equation:
3-3S
-------
Y = AXU
Where Y = annual scrubbing costs -
Third quarter 1979 dollars per year
X = scrubber capacity - megawatts
For partial scrubbing cases that do not include a reheater
$4200 per year per megawatt of scrubber capacity should be
13
deducted from annual costs. Unit annual costs can be
estimated using the techniques described in Section 3.4.4.1.
Electrical energy requirements for wet scrubbing systems
7
should be estimated as shown in Table 3-6 . Auxiliary boiler
capacity for reheating from 125° to 175°F should be estimated
at 210,000 Btu per hour heat input per megawatt of scrubber
capacity.
Other costs that may be involved with retrofitting but
are not included in the foregoing estimates are discussed in
Sections 4.2, 4.3, 4.4, 4.5, and 4.6.
3.5.2
Retrofit Costs for Lime Dry SO., Control
The total indirect and direct capital cost of lime dry
scrubbing systems should be estimated at 125 third quarter
17
1979 dollars per kilowatt of scrubber capaticy. These costs
include:
Feed Storage and Handling
Dry Scrubbing
Sludge Disposal
Booster Fans
and Retrofit Ductwork
The design effectiveness of the system is 70 percent removal
of SO
2'
The costs include 15 percent allowance for the
additional cost of retrofit ductwork.
3-36
-------
TABLE 3-6. ELECTRICAL ENERGY REQUIREMENTS
FOR WET FLUE GAS DESULFURIZATION2
System
Electrical Energy Required
Percent of Scrubber
Capacity (Gross Megawatts)
Lime Scrubbing
0 - 3.5 percent sulfur
* 7.0 percent sulfur
3.5
3.5
Limestone Scrubbing .
0-3.5 percent sulfur
7.0 percent sulfur
3.5
4.3
Double Alkali Scrubbing
0 - 3.5 percent sulfur
7.0 percent sulfur
3.1
3.1
Magnesium Oxide Scrubbing
0 - 3.5 percent sulfur
7.0 percent sulfur
5.9
9.4
Wellman Lord Scrubbing
0 - 3.5 percent sulfur
7.0 percent sulfur
13.3
25.9
3-37
-------
The annual costs of dry scrubbing are estimated at
41,500 third quarter 1979 dollars per year for each megawatt
of scrubbing system capacity.
Electrical energy requirements for dry .scrubbing should
be estimated at 16.9 kilowatts per megawatt of scrubber-
capacity.
*
Annual costs can be converted to unit annual costs usijig.
• * * ytf.'t i
the techniques described in Section 3.4.4.1. Other costs ' '
involved with retrofitting but are not.included in the fore-
going estimates are discussed in Sections 4.2, 4.4, 4.5, and
4.6.
3.5.3 Makeup Water Requirements
For some cases, especially for power plants
located in arid Western areas,'makeup water supply
can be a problem. Water requirements, for wet scrubbing
systems are estimated to be 1 gallon per minute per
megawatt.13 Water requirements for dry scrubbing are^
estimated to be 0.8 gallons per minute per megawatt.
In cases where water supply is limited, use of other
water sources-such as cooling tower blowdown should be
considered.
-------
LAND AREA REQUIREMENTS
Additional space is needed for retrofit air pollution
control systems except for N0x steam generator modifications
where the necessary hardware can be. fitted" within existing
structures. In most cases it will be necessary to locate
retrofit particulate and S02 emission control-systems on the
other side of the existing stack from the steam generator.
Although it might be possible to locate part of the retrofit
systems between the stack and the steam generator by removing
existing air pollution control systems, usually this is not
a viable option since removing existing equipment would, ijnv^jtvff
extended, costly shutdown of the steam generator. ; <•.
For electrostatic precipitators a minimum of 12.5 square
feet of space is needed for each 1000 square feet of collecting
surface area added. Minimum baghouse space requirements
should be estimated at 37 square feet per megawatt of baghouse
capacity. These space requirements are for a 2 to 1 air
to cloth ratio baghouse and do not include•space for ash
-------
-------
storage or disposal.
Table 5-7 shows space requirements for wet lime, wet
limestone, and dry lime scrubbing systems. These estimates
include space for feed storage and handling, scrubbers, and
liquor treatment but do not include space for sludge disposal.
Table 5-8 shows values for estimating S02 sludge generation.
A 50 percent solids content should be used for all cases
except for low sulfur Western coals or for Eastern coal cases
where a special oxidation system is to be installed. For
those cases 70 percent solids content should be used. Sludge
volumes can be converted to area by assuming a pond depth and
life. A pond depth of 50 feet and a life of 20 years shoulc
be used except for site specific situations where other values
are a more logical choice.
The foregoing values should be used to identify potential
space problems involved with retrofitting. If space problems
are identified, more comprehensive studies than those of this
document will be needed to indicate potential solutions.
5.7 EMISSION MONITORING COST
The retrofit cost for emissions monitoring systems
(18, 19, 20) is divided into two categories:
5.7.1 Retrofit Capital Costs
Monitoring Equipment
SOo/NO System (I/Boiler)
Analyzer, Remote Readout § Control =
Installation Cost
Total Installed Cost
Indirect Capital Costa
Field Certification Fee
Total S02/N0 System:
aRefer to Section 5.2.1 for definition
3-39
$50,000
22,000
52,000
50,000
10,000
92,000
-------
TABLE 3-7. MINIMUM LAND AREA REQUIREMENTS FOR
LIME AND LIMESTONE SCRUBBING SYSTEMS13
System
Minimum Land Area Required
(Square Feet Per
Megawatt Scrubber Capacity)
Wet Lime Scrubbing
0.5 percent sulfur
3.5 percent sulfur
80
180
Wet Limestone Scrubbing
0.5 percent sulfur
3.5 percent sulfur
80
140
Dry Lime Scrubbing
(Includes Baghouse)
80
3-40
-------
TABLE 3-8. SLUDGE GENERATION FOR LIME AND
LIMESTONE SCRUBBING SYSTEMS13
System
Sludge Generation
(Cubic Feet Per Year
Per Megawatt For
1 Percent Sulfur Coal)
Wet Lime Scrubbing
(50 percent solids)
6,000
Wet Lime Scrubbing
(70 percent solids)
4,000
Wet Limestone Scrubbing
(50 percent solids)
Wet Limestone Scrubbing
(70 percent solids)
Dry Lime Scrubbing
7,000
5,000
6,000
3-41
-------
Opacity System (I/Boiler)
Transmissiometer, Remote Readout,
Converter Unit, Air Flush Equip.,
and Miscellaneous
Installation Cost
Total Installed Cost
Indirect Capital Costa
Field Certification Fee
Total Opacity System:
Oxygen Analyzers (2/Boiler)
Probe, Shield, Umbilical §
Controller
Installation Cost
Total Installed Cost
Indirect Capital Cost a
Field Certification Fee
Total for Two Oxygen Analyzers:
Strip Chart Recorders 3/Boiler)
Installation Cost
Total Installed Cost
Indirect Capital Costa
Total for Three Strip Chart Recorders
Data Processor (I/Plant)
Processor, Software Pkg., Cabinet
§ Reason Code Panel
Installation Cost
Total Installed Cost
Indirect Capital Costa
Field Certification Fee
Total Data Proces:
$15,000
26,000
14,000
24,000
5,000
$68,000
$40,000
57,000
77,000
45,000
'5,000
$127,000
$ 6,000
6,000
12,000
7,000
$19,000'
$50,000
25.000
75,000
44,000
' 2,000
$121,000
In summary, capital costs are estimated at $506,000 per
steans generator plus $12'1,000 for a plant data processor.
a Refer to Section 5,2.1 for definition
3-42
-------
Operating Costs
Operating Costs Plant
Labor 1 man per shift with 3
shifts @ $11/MH
Annual Labor Cost
Material and Miscellaneous
Total
$96,100
5,900
$100,000
3.7.3 Annual Costs
Annual costs should be estimated at 17.2 percent of
capital costs plus.$100,000 per year per plant as given above
'Annual operating costs based on 365 days per year and
24 hours per day.
3-43
-------
3.8 TIME REQUIREMENTS FOR RETROFITTING
Figure 3-3 presents a typical engineering, procurement and
construction schedule for retrofitting a large power plant
for NOX, particulate, and SC>2 control. The elapsed
time from contract award to plant operation is 60 months.
The engineering time span is 26 months, and procurement is
36 months. Purchasing is completed when the final purchase
order is released; however, inspection, expediting, and
traffic are involved until the last of the materials are
delivered. The construction span is 33 months. This span
allows four months for tie-in to the existing equipment, and
it is based on staggering the shutdown of the units. No
special unit shutdowns would be required; normal plant
shutdowns would be utilized to tie into the existing
equipment. The time periods shown in Figure 3-3 can vary
considerably if other factors such as economic, political,
or international situations become involved.
3-44
-------
w
-p
5-1
a)
>
o
Q)
tP
5-4
(0
Cn
rt
•H
4J
O
M
-P
Q)
J-l
5-t
O
«4-4
-------
3.9 REFERENCES
1. EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed N0x Emission
Standards." EPA-450/2-78-005a, July 1978.
2. EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed S02 Emission
Standards." EPA-450/2-78-007a, August 1978.
3. EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed Particulate Matter
Emission Standards." EPA-450/2-78-006a, July 1988.
4. Wright, J., "Cost Analysis of Lime Based Flue Gas
Desulfurization Systems for New 500 MW Utility Boilers"
PEDCo Contract No. 68-02-2842, Assignment 25,
January 1979.
5. Meeting Notes, Dr. K. Hsiao, Pullman Kellogg - Meeting
with E.J. Campobenedetto, Babcock and Wilcox Co.,
Barberton, Ohio, March 19, 1979.
6. Meeting Notes, Dr. K. Hsiao, Pullman Kellogg - meeting
with D.J. Frey, Combustion Engineering, Windsor, CT,
March 21, 1979, and letter from D.J. Frey of March 26,
. 1979.
7. Meeting Notes, Dr. K. Hsiao, Pullman Kellogg - meeting
with A.H. Rawddn, et al., Riley Stoker Corporation,
Worcester, MA, March 20, 1979.
8. J. Vatsky, "Attaining Low NOV Emissions by Combining
.A.
low Emission Burners and Off-Stoichiometric Firing",
Foster-Wheeler Energy Corp. Paper presented at 70th
annual meeting of AIChE November 1977.
9. J. Vatsky, "Experience In Reducing N0x Emission on
Operating Steam Generators" Foster-Wheeler Energy
Corp., Livingston, NJ internal document.
10. Meeting Notes, Dr. K. Hsiao, Pullman Kellogg - meeting
with H.J. Melosh, III et al., Foster Wheeler Energy
Corporation, Livingston, NJ, March 2, 1979.
3-46
-------
11. C.E. Brackett and J.A. Barsin, "The Dual Register
Pulverized Coal Burner" paper presented to EPRI N0x
Control Technology Seminar, San Francisco, CA, Feb.
1976.
12. A.M. Rawdon and S.A. Johnson, "Control of N0x Emissions
from Power Boilers" Paper presented at the annual meeting
of the Institute of Fuel, Adelside, Australia,
November 1974.
13. Final Report, Retrofit Guidelines for Coal-Fired Power
Plants, Pullman Kellogg Division, EPA Contract No.
68-02-2619, Work Assignment No. 13, September 1979.
14. Letter N. Master, Pullman Kellogg to J. Copeland, EPA,
September 19, 1979.
15. Capital Costs of Free Standing Stacks, EPA Contract No.
68-02-099, Vulcan Corporation, Cincinnati, Ohio,
August 1973.
16. Telephone conversation Dr. K. Hsiao, Pullman Kellogg with
Mr. Franklin Jackson, Duke Power Co., May 1, 1979.
17. Letter John C. Buschmann, Niro Atomizer Incorporated to
Don R. Goodwin, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, February 26, 1980.
18. Letter from K.A. Kedtke, Leon Siegler, Inc., to
N. Master, Pullman Kellogg, Budgetary Quote, June 29,
1979.
19. Letter from R.F. Crowthen, Dynatron, Inc., to N. Master,
Pullman Kellogg, Budgetary Quote, June 26, 1979.
20. Letter from L.N. Roten, Thermo Electron to Dr. K. Hsiao,
Pullman Kellogg, Budgetary Quote, June 25, 1979.
3-47
-------
-------
SECTION 4
TECHNIQUES FOR ESTIMATING TOTAL RETROFIT
COSTS FOR EMISSION CONTROL
4.1 GENERAL
The cost of a power plant retrofit is estimated in terms
of capital cost and annualized cost (!_) . Capital cost
represents the initial investment necessary to install and
commission the retrofit, and the capital costs consist : of
the direct and indirect costs that are defined in
Section 3.2.1. Annualized costs are composed of direct
and fixed charges. Working capital,, that is the money
required to operate the plant after completion of the
retrofit, should also be included in the retrofit cost.
Specific cost estimating examples are given in
Appendices A, B, and C.
4.2 Working Capital
Working capital is the money set aside to operate the
plant after completion of the retrofit. The working
capital should be estimated as-25% of the total annual
operating costs (direct and fixed).
4.3 Auxiliary Boiler Costs
When plume reheat is required the capital cost of an
auxiliary boiler should be included in the total capital
cost estimate. Section 3-5.1 describes the techniques
that should be used to estimate the size of auxiliary
boiler needed. The annual costs of plume reheat steam are
included in^ the annual cost estimates of Table 3.4.
Table 4-1 should be used to estimate the capital cost of
auxiliary boilers.
4-1
-------
TABLE 4-1
CAPITAL AND ANNUAL COSTS (2) FOR AUXILIARY2
BOILERS GREATER THAN 250 x 106Btu/Hr HEAT INPUT
Capital Cost (a)
$/l()6 Btu/Hr Heat input capacity
Boiler 33,150
Pollution Control (b) 4.755
total 37,885
Annual Boiler Costs (a)
$/lo6 Btu/heat input (a)
Boiler Fixed Costs 1.00
Pollution Control Fixed Costs (b) 0.14
Boiler 0 & M 1.43
Pollution Control 0 & M(b) 0.43
total 3.00(a)
Steam Costs (a)
Boiler Less Fuel Cost
Fuel Cost (d)
$/lo6 Btu of steam (c)
3.75
0.63
total
4.38 (a)
(a) Third quarter 1979 dollars
(b) Includes systems for 90 percent S02 removal and
particulate emission reduction to 0.03 lb/10" Btu
(c) Assumes 80 percent boiler efficiency
(d) Assumes $0.50/10^ Btu fuel cost for Western power
plants. This value should be ad-justed for fuel costs
for plant studied.
4-2
-------
4.4 Electrical Energy Penalty
The total capital cost of a retrofit system includes the
capital cost of replacing the generating capacity lost
because of the electric power requirements of the retrofit
systems. This capital cost is $1,046 for each kilowat of
capacity required by the retrofit systems.Clj3,4)
Sections 3.4 and 3.5 describe the techniques to be used
.for estimating retrofit electric power requirements for
particulate and S02 control.
4.5
Other Costs Not Estimated
There are other capital and annualized costs involved in
conjunction with retrofits that are not estimated in this
document. This section identifies these cost elements and
provides guidance on factoring these costs into decision
making on best available retrofit technology (BART)
determinations.
Other potential costs that are not included in the
estimates of this document are identified as.follows:
1. Cost of land
2. Cost of relocating facilities to make room for the
retrofit systems
3. Cost of altering existing facilities to accommodate
the retrofit systems
4. Cost of providing additional facilities for additional
employees such as offices, locker rooms, etc.
4-3
-------
5. Cost of downtime for installing retrofits
6. Cost of stacks
The cost estimates of this document provide ample
allowances for grading, excavating, piling, and for
temporary construction facilities, etc.
In some cases, additional land may need to be purchased to
make up for the space needed for retrofit systems. Since
the cost of this land can usually be recovered when the
land is no longer needed, it is not included as a capital
cost. It is recognized, however, that necessary funds
would have to be made available for such land purchases
and that annual costs would result. In the case of land
for sludge disposal, it is assumed that once the land is
used, it would not be possible to reclaim the land for any
useful purpose. More study is needed to show that land
used for sludge disposal can be reclaimed for future use.
Since most power plants have not been designed for future
large retrofit systems, it is likely that most retrofit
cases will involve relocation of some facilities such as
shops, offices, or coal storage and handling systems.
These capital costs will also cause an increase in annual
costs.
Types of alterations that might be required to accommodate
retrofit systems are the cost of relining stacks to
compensate for more corrosive gas conditions or for
reinforcing existing ductwork to compensate for changed
flue gas pressure conditions, or costs for major changes
to structures to accommodate NOX combustion modifi-
cations. The costs of nominal alterations in conjunction
4-4
-------
with NOX combustion modifications is included in the
cost estimates of this document. Based on boiler
manufacturer's advice, it may also be necessary to modify
boiler pressure parts to control steam conditions to
specifications. These costs are site-specific and are not
estimated in this document.
The cost of downtime is also significant. The costs of
this document assume that no additional downtime is
required for retrofitting. The way downtime is avoided is
by making all necessary changes to the existing system and
by tieing in the retrofit systems during normal outages or
during unscheduled outages attributable to factors other
than retrofitting. As shown by Figure 3-8, these types of
changes can be made during a 5-year period. If downtime
is necessary, the following factors should be taken into
account in assessing costs.
1. The cost of purchased power. Usually purchased power
costs more than the cost of generating power within the
system. However, at times the added cost of purchased
power is reduced if the purchasing power system sells a
like amount at the same price in conjunction with an
exchange agreement.
i
2. The cost of power generation and distribution. Even
if it is not necessary to purchase power from another
system, downtime can involve significant additional costs.
Downtime may make it necessary for a power system to
generate power at a' less efficient plant or at a plant
firing more costly fuel. Power transmission losses also
need to be considered. For the plants of Appendices A, B,
and C, it is most likely that any downtime that would make
it necessary to generate power elsewhere would involve
significant additional fuel costs.
4-5
-------
3. Loss of Productivity. When a steam generator is down,
some labor, supplies, and services costs continue.
Although these costs are small in comparison to other
downtime costs, they should be considered in sufficient
depth to classify them in their proper perspective.
Another cost that is not estimated in this document is the
cost of transporting sludge from the liquor treatment
system to the disposal site. This cost is estimated at $2
per ton per mile. Such costs are not estimated because it
is not certain how far the sludge would have to be
transported.
4.6 Cost of Derating
As discussed in Section 2-2.7, derating is usually an
undesirable technique for reducing emissions. This is
because there are usually more cost effective methods
available. The costs of derating are very variable and
are site specific. Consequently, no generalizations can
be made on costs except to identify the potential cost
elements.
For cases where electrical energy demand is increasing
(this is almost always the situation)' the generating
capacity lost because of derating must be replaced. The
cost of replacing generating capacity is', given in
Section 4.4 For cases where generating capacity is limited
as compared with power system electrical demand, derating
costs can be the same as the costs for downtime discussed
in Section 4.5. For the most costly case it might be
necessary to purchase power from another power system for
several years until the power generating capacity lost by
derating is replaced. However in most cases derating
costs would not be this severe. For actual cases,
additional studie-s are necessary to estiamte the costs of
derating.
-------
4.7 Escalation
The costs of this document are based on September 1979
dollars. Section 3.8 presents data on schedules for
retrofitting which can be used in conjunction with
economic data not given in this document to estimate the
effect of escalation on capital costs.
-------
4.7 REFERENCES
Wright, J., "Cost Analysis of Lime Based Flue Gas
Desulfurization Systems for New 500 MW Utility Boilers",
PEDCo Contract No. 68-02-2842, Assignment 25, January
1979.
Impact Analysis of Selected Control Levels for New
Industrial Boilers, Preliminary Draft, Office of Air
Quality Planning and Standards, U. S. Environmental
Protection Agency, Research Triangle Park, North Carolina
June 1980.
EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed S02 Emission
Standards". EPA-450/2-78-007a, August 1978.
EPA, "Electric Utility Steam Generating Units -
Background Information for Proposed Particulate Matter,
Emission Standards". EPA-450/2-78-006a, July 1978.
4-7
-------
APPENDIX A
EXAMPLES FOR RETROFITTING THE FOUR CORNERS POWER STATION
A-i
-------
CONTENTS
SECTION
PAGE
CONTENTS
TABLES
FIGURES
A-ii
A-lv
A-vi
A-1.0 GENERAL
A-l
A-l.l Retrofit Alternatives
A-l.2 Plant Characteristics
A-l.3 Existing Facility Relocations
A-l.1! Flue Gas Ducting Requirements
A-2.0 BACKGROUND DATA
A-2.1 Plant Description
A-2.2 Steam Generator Description
A-2.3 Existing NOX Control
A-2.4 Existing Particulate Control
A-2.5 Existing S02 Control
A-3.0 PLANT SURVEY FORM
A-3.1 Company and Plant Information
A-3.2 Plant Data
A-3.3 Boiler Data
A-3-4 Fuel Data
A-3.5 Atmospheric Emissions
A-3.6 Particulate Removal
A-3.7 Scrubber Train Specifications
A-3-8 Calcining and/or Slaking Facilities
A-3.9 Disposal of Spent Liquor
A-l
A-ll
A-13
A-14
A-16
A-16
A-17
A-19
A-20
A-20
**
A-21
A-21
•A-21
A-22
A-28
A-29
A-30
A-31
A-33
A-33
A-ii
-------
CONTENTS
SECTION
PAGE
A-3.10 Cost Data
A-3.11 Major Problem Areas
A-3.12 Methods of Measuring Emissions
A-4.0 RETROFIT DESCRIPTION
A-4.1 NOX Emission Control
A-4.2 Particulate Emission Control
A-4.3 S02 Emission Control
A-5.0 RETROFIT COSTS
A- 6.0 REFERENCES
A-36
A-39
A-40
A-40
A-46'
A-46
A- 50
A-53
A-iii
-------
TABLES
Table
A-l
A-2
A-3
A-4
A-5
A-6
Retrofit for NOX Reduction
Retrofit Data for Electrostatic Precipitators and
Baghouses - Units 1 and 2
Retrofit Data for Electrostatic Precipitators and
Baghouses - Unit-3
Retrofit Data for Electrostatic Precipitators and
Baghouses - Units 4 and 5
Capital Investment Costs for Retrofitting
the Four Corners Power Plant
Annual Costs for Retrofitting the Four Corners
Power Plant
Page
A-41
A-47
A-48
A-49
A-51
A-52
A-iv
-------
FIGURES
FIGURE
A-l
A-2
A-3
A-4
A-5
A-6
• A-7
A-8
A-9
A-10
A-11
A-12
A-13
Addition of wet S02 scrubbing modules.
Arrangement of wet S02 scrubbing modules for
Units 1, 2, and 3.
Addition of wet S02 scrubbing and particulate
emission control modules.
Arrangement of cold side ESP's and wet S02
scrubbing modules for Units 1, 2, and 3-
Addition of dry S02 scrubbing modules with
baghouses.
Arrangement of dry S02 scrubbing and' baghouse
modules for Units 1, 2, and 3.
Arrangement of wet S02 scrubbing modules for
Units 4 and 5.
Arrangement of baghouse and wet S02 scrubbing
modules for Units 4 and 5.
Arrangement of semi-dry S02 scrubbing (spray-
dryer) and baghouse modules for Units 4 and 5.
General plot plan of the Four-Corners power station.
Retrofit arrangement of OFA and CA ports for Units
1 and 2 - Riley-Stoker boilers.
Retrofit arrangement of OFA and CA ports for
Unit 3 - Foster Wheeler boiler.
Retrofit for Units 4 and 5 with 54 sets of Dual
Register (B&W) burners and compartmentized windbox.
PAGE
A-2
A-3
A-4
A-5
A-7
A-8
A-9
A-10
A-12
A-18
A-43
A-44
A-45
A-v
-------
-------
SECTION A-l
GENERAL
A-l.l RETROFIT ALTERNATIVES
Pour alternatives for retrofitting each boiler at the Pour
Corners Power Plant were considered in this appendix to
demonstrate the use of the techniques described in Section 3
of the report. All alternatives include maximum NOX
control and the installation of opacity, S02, NOX
emission monitoring systems. Since some of the aspects of
these alternatives differ for the boilers ,to. which they
apply, they have been grouped by unit numbers for the
discussion. Alternatives for the Unit 1,2, and 3 boilers
are as follows: •
Alternative 1 - Add wet S02 scrubbing to each unit to
achieve 90% S02 removal. The existing venturi
scrubbers are retained to control particulate emissions to
a level of 21 ng/J heat input (0.05 lbs/106 Btu).
Figures A-l and A-2 show the general arrangement plot
plans with the addition of the S02 scrubbing modules.
Alternative 2 - Remove the existing venturi particulate
scrubbers and•add wet S02 scrubbing to achieve 90%
S02 removal for each unit. -Also, add high-efficiency,
cold-side electrostatic precipitators (ESP's) for control
of particulate emissions to a level of 13 ng/J heat input
(0.03 Ibs/lo6 Btu). Figures A-3 and A-4 show the
general arrangement plot plan with the added S02
scrubbing modules and cold-side ESP's.
Alternative 2a - The retrofit for particulate removal in
Alternative 2a is to the same emission level as for
Alternative 2, but baghouses have been used for cost
comparisons with Alternative 3. The wet S02 scrubbing
A-l
-------
CO
a)
•d
i
tn
a
•H
jQ
•9
M
O
CO
O
en
4J
0)
a
o
•H
4J
-H
<
0)
I
A-2
-------
!p
-------
cu
-P
(0
O
•H
-P
•g
fi
•H
d to
M ®
O rH
W 3
•O
CSO
o e
CO
^
-P O
(1) M
m o
o u
c c
o o
•H -H
-p
•H
•d
0)
M
A-4
-------
ro
& a
(0
-d
c -
nJ CN
ffl •»
- rH
Ol
co ca
W 4J
•H
•H
0)
rH
o in
o (U
rH
m r>
o -d
m (r>
s e
03 -H
O
cn
I
•
•^
<
0)
A-5
-------
in this case is based "on 70% SC>2 removal for cost
comparison with semi-dry scrubbing of Alternative 3.
Alternative 3 - The existing venturi particulate scrubbers
are removed in this case and semi-dry (spray drying)
S02 scrubbing is added for 70% S02 removal.
Baghouses are used as dry collectors, for control of
particulate emissions to a level of 13 ng/J heat input
(0.03 Ibs/lo6 Btu) . Figures A-5 and A-6 show the
general arrangement using. the dry scrubbing/baghouse
modules .
The three alternatives considered as examples for
retrofitting the Units 4 and 5 boilers are:
Alternative 1 - Add wet SC>2 scrubbing to achieve 90%
SC>2 removal and retain the existing ESP and
particulate removal for control to an emission level of 21
ng/J heat input (0.05 lb/106 Btu). Figures A-l and
A-7 show the general arrangement of the plant with the
added S02 scrubbing modules.
Alternative 2 - This option adds baghouses plus wet
S02 scrubbing for 90% SO^ removal, and it retains
the existing ESP's and particulate removal for control to
an emission level of 13 ng/J heat input (0.03 Ibs/lO^
Btu). Figures A-3 and A-8 show the plant's general
arrangement with the added SOg scrubbing modules and
baghouses.
Alternative 2a - Alternative 2a includes the same retrofit
for particulate removal emission levels as Alternative 2,
but baghouses have been used for cost comparison with
Alternative 3. The S02 scrubbing is based on 70%
SC>2 removal for cost comparison with semi-dry
scrubbing in Alternative 3.
A-6
-------
w
CD
CO
(0
-P
•H
CO
0)
i
tn
S
•H
O
CO
CM
O
CO
M
O
O
-H
-P
•H
t5
m
A-7
-------
Q)
W
3
O
,a
tr>
c
(ti
tn
PI
•H •
XI "
3 T)
o
to
(tf
CMO1
O
CO -
>4 10
15 -P
•H
m rt
o D
-------
*
^*
fll
J-l
o
M-l
CO
i
tn
C
•rH
U
tfl
CM
O
4J
I
-p -d
c! c
CD (d
tn
C to
(0 -P
M-H
S-i C
r-
k
OJ
A-9
-------
O
(0
CN
O
W
in
a) T
w
0) m
g
0) to
tn cu
C rH
(0 3
00
A-10
-------
Alternative 3 - Dry S02 scrubbing (spray drying) is
added for 70$ SC>2 removal. Baghouses are used for dry
collectors and, combined with the existing ESP's, provide
for particulate removal to a level of 13 ng/J heat input
...CO. 03 lbs/106 Btu). Figures A-5 and A-9 show the
general arrangement with the added dry scrubbing modules.
Section A-6 describes the techniques used for estimating
costs.
A.-1.2 PLANT CHARACTERISTICS
The characteristics of the plant site, existing equipment,
and the space requirements for the retrofit example are
shown in the following list (_!) :
A. The station is located on a 1000 acre tract of land.
B. Particulate removal for Units 1,2, and 3 is
accomplished using two venturi scrubbers per .unit.
C. Units 4 and 5 have cold side precipitators for
particulate removal.
D. Each boiler for Units 1,2 and 3 Is provided with two,
forced draft (PD); three, primary-air (PA); and two,
induced-draft (ID) fans.
E. Each boiler for Units 4 and 5 is provided with four,
forced-draft (FD) and two primary-air (PA) fans.
F. The number of S02 scrubbing modules is based on
the total calculated flue gas rate from each boiler.
G. One S02 scrubbing module is provided as a common
spare for each group of modules per boiler.
H. One flue gas reheater is required per wet S02
scrubbing module for 90% S02 removal.
I. One flue gas booster fan is required per scrubbing
module. •
J. The individual scrubbing modules are provided with
dampers. This provision allows the individual modules
to be isolated for maintenance.
•A-ii
-------
Q)
ti
10
in
tn
C T)
-H fl
o to
O..-JD
co
01
w
a)
4J CO
C 3
-------
K; Retrofit equipment tie-in to the power plant is based
on completion during a normal power plant maintenance
turn-around of 3 to 6 weeks.
L. An emergency bypass is provided around each S02
scrubbing system to allow emergency operation of the
boiler in the event of a major FGD malfunction.
Bypassing of the particulate control equipment is not
provided.
A-1.3 EXISTING FACILITY RELOCATIONS
Major revamp work that would be required to install the
equipment needed for S02 and particulate control
includes relocation of some existing buildings and/or
systems. The requirements for the alternatives being
considered are:
Alternative 1 - The following equipment relocation would
be necessary to^ allow for the space requirements of the
wet S02 scrubbing systems.
A. For Units 1,2, and 3, relocation of all buildings
situated east and north of the unit is necessary.
B. For Units 1,2, and 3, the scrubbers are located on
structures built over the cooling water discharge
canal.
C. For Units 4 and 5, no major equipment relocation is
required.
Alternative 2 - All equipment relocation considerations of
alternative 1 also apply to this alternative. In addition
the following items must be relocated.
A. For Units 1,2, and 3, relocation is needed for the
ready to use coal pile and some of the coal conveying
equipment.
A-13 .
-------
B. For Units 4 and 5, relocation of the blend coal pile
is required, to provide space for installing the
baghouses.
Alternative 3 - Equipment relocation requirements of
Alternative 1 also apply to this alternative. In
addition, the following items must be relocated.
A. For Units 1,2, and 3, relocation of the ready to use
coal pile and some of the coal conveying equipment is
required.
B. For Units 4 and 5, relocation of the blend coal pile
is required, to provide space for installing the
baghouses and booster fans.
A-1.4 FLUE GAS DUCTING REQUIREMENTS
Bypass duct and dampers are provided to enable the flue
gas to bypass the S02 scrubbing system completely
under emergency conditions. For all five units, the
bypass duct is upstream of the SC>2 scrubbing system.
Emergency bypassing of particulate control equipment is
not considered. Bypass duct locations for the three
alternatives are indicated below:
Alternative 1
Units 1,2 and 3.- The bypass duct is taken from the
plenum located between the ID fan and the booster fan.
Figure A-2 indicates this relationship.
Units 4 and 5.- The bypass duct is taken from the inlet
duct just between the existing ESP and the booster fan.
See Figure A-7 for a graphic indication of this
ducting.
A-14
-------
Alternative 2
Units 1,2 and 3-- Bypass duct is taken from the plenum
located between the scrubber and the booster fan. Figure
A-4 shows the elevation view for the cold side ESP and
S02 scrubbing modules.
Units 4 and 5-- Bypass duct is taken after the baghouse,
just before the scrubber. Figure A-8 shows the elevation
view for the baghouse and S02 scrubbing modules.
Alternative 2a
The bypass duct is the same as for Alternative 2. This
duct can also be used during normal operation to divert
about 22% of the total flue gas that does not require
treatment, since the module removes 90% S02, and' this
alternate requires only 70% SC>2 removal. The bypass
duct also provides the flue gas requirements 'for reheat.
Alternative 3
The bypass duct is taken just ahead of the spray dryer,
and it ties into the duct to the baghouses. This
location of the emergency bypass permits operation of the
baghouse for particulate control when a spray dryer is
out of service. Figure A-6 and A-9 show the elevation
views for the dry scrubbing and baghouse modules, and
they indicate this relationship.
A-15
-------
SECTION A-2
BACKGROUND DATA
A-2.1 PLANT DESCRIPTION (1,2:)
The Four Corners Power Plant consists of five thermo-
electric generation units .that develop a total plant
capacity of 2085 MW(net) and 2181 MW(gross). The capacity
for each boiler is:
Unit No.
Gross MW
Net MW
1
190
175
2
190
175
3
245
225
4
778
755
5
778
755
Units 1, 2, and 3 are owned and operated by Arizona Public
Service (APS). Units 4 and 5 are jointly owned by six
electric utilities and operated by APS. The participating
utilities are Southern California Edison Co., 48$; APS
15$; Public Service of New Mexico 13$; Salt River Project
10$; Tucson Gas and Electric Co., 7$; and El Paso Electric
Co., 7$.
All five units are pulverized-coal-fired, dry-bottom
boilers, where about 20$ of the ash in the coal is
retained as bottom ash. The remaining ash is entrained in
the flue gas and is collected by using either a Venturi
Wet Scrubber System or an Electrostatic Precipitator
(ESP). There are four flue gas stacks. Units 1 and 2 are
served by a common stack that is 250 feet high and has an
18.5-foot I.D. The stack for Unit 3 is also 250 ft high,
but it has a 15-foot I.D. The two stacks for Units 4 and
5 are 300 feet high and have 28.5-foot I.D's.
A-16
-------
Units 1 and 2 each have two, horizonta 1-shaft
regenerative, Ljungstrom air preheaters. Unit 3 has the
same preheaters as Units 1 and 2 but with vertical shafts.
U n i t s ' 4 and 5 each have two, horizontal-shaft,
regenerative, Ljungstrom air preheaters, and they each
have a tubular air preheater.
The bottom ash is conveyed hydraulically to dewatering
bins and then trucked either to the ash disposal pond
(bottom ash from Units 1, 2, and 3) or to the coal mine
(bottom ash from Units 4 and 5). The water aftsr
separation (decanted water) is returned to Morgan Lake.
Two, 100-foot I.D. thickeners serve Units 1, 2, and 3 for
clarifying spent ash liquor. The solids concentration of
the thickener underflow is about 40 wt%. The plan't has a
brine concentrator to treat the water from the ash pond.
Circulating, cooling-water discharge canal is located
between Units 3 and 4. The Instrument and Plant Air
Compressor Building is located between Units 2 and 3.
Units 4 and 5 have nine coal mills per unit, plus a spare
one. Units 1, 2, and 3 have three coal mills per unit,
working at full capacity, and they have no spares. Figure
A-10 shows a general plot plan arrangement for the Pour
Corners power atation.
A-2.2 STEAM GENERATOR DESCRIPTION
Units 1 and 2 have a total generating capacity of 350 MW
(net) and 380 MW (gross). Each unit consists of a
balanced-draft, Riley-Stoker (R-S) , 175 MW (net), 190 MW
(gross) boiler. Both units went into operation in 1963.
A-17
-------
-p
(d
•P
M
1
W
H
a)
c
n
o
o
S-l
3
o
fo
(U
M-l
O
-p
o
H
a.
0
fl
(U
o
i
4
O
.rH
<
0)
A-18
-------
Unit 3 uses a balanced-draft, Foster-Wheeler (F-W), 225 MW
(net), 245 MW (gross) boiler. Unit 3 started operating in
1964.
Units 4 and 5 have a total generating capacity of 1510 MW
(net) and 1556 MW (gross). Each unit consists of a
pressurized, Babcock & Wilcox (B&W), 755 MW (net), 778 MW
(gross) boiler. Unit 4 was started in 1969 and Unit 5 in
1970. Fireboxes for the five boilers were designed prior
to 1970.
For Units 1,2, and 3, there are two forced-draft (FD),
three primary-air (PA) and two induced-draft (ID) fans per
unit. Units 4 and 5 have four FD and two PA fans per
unit. Paddle-type fans are used for the units.
A-2.3 EXISTING NOX CONTROL
There are five boilers at the Four Corners Station. Units
1 and 2 use Riley Stoker boilers that are of the
horizontal, single-wall-fired type. The burners are
arranged in six vertical columns with three burners on
each column. Retrofit work' has been performed for NOX
reduction through burner modifications. The retrofit work
was developed, and installed by APS using a KVB spoiler
design (3_) . There are no data avaialble to show the
effectiveness and results of the retrofit. Unit 3 has a
Foster Wheeler, horizontal, single-wall-fired boiler. The
burners are arranged in four vertical columns with five
burners in each column. Plant data show that no retrofit
work for NOX reduction purposes has been done on
Unit 3. Units 4 and 5 are B&W boilers of the
opposed-wall-fired type. The burners on each unit are
arranged in six vertical columns on each opposite wall.
A-19
-------
The burners are the B & W's, high-intensity-cell,
turbulence burners. All burners are contained in a single
windbox. Units 4 and 5 were installed originally with
flue-gas (FG) recirculation systems; however, the FG
• recirculation system has been taken out because of severe
mechanical difficulties associated with the fans.
A-2.4 EXISTING PARTICULATE CONTROL (2_)
Boilers 1, 2, and 3 each use two venturi, wet-scrubber
systems for collecting flyash from the flue gas. The
venturi scrubber systems remove about 93% of the flyash
from the flue gas. Boilers 4 and 5 each are equipped with
two cold-side ESP's manufactured by Research Cottrell.
These ESP's collect about 97% of the flyash from the flue
gas. The total existing ESP collection area per boiler is
373,000 ft2.
A-2.5 EXISTING S02 CONTROL
S02, emission control is accomplished by using low
sulfur subbituminous coal as fuel.
A-20
-------
SECTION A-3
PLANT SURVEY FORM
A-3.1 COMPANY AND PLANT INFORMATION (1,1)
1. Company Name: Arizona Public Service Company
2. Main Office: P.O. Box 21666, Phoenix Arizona 85036
3. Plant Manager: Mr. B.E. Haelbig
4. Plant Name: Four Corners Power Plant
5. Plant Location: Fruitland, San Juan County, New Mexico
6. Person to Contact For Further Information: Mr. J. Weiss
7. Position: Senior Environmental Eng. Snecial Projects
8. Telephone Number: (602) 271-2292
9. Date Information Gathered: May 8 - May 10, 1979
10. Participants In Meeting Affiliation
D. Campbell
J. Weiss
N. Gonzalez
R. Redman
R. Roberts
APS/Acting Manager of Eng,
Four Corners Plant
APS/Senior Environmental
Eng. Special Projects
Pullman Kellogg
Pullman Kellogg
Pullman Kellogg
A-3.-2 PLANT DATA (APPLIES TO ALL BOILERS AT THE PLANT)
BOILER NO.
Capacity, MW, Net,
Service (Base,Peak)
FGD System Used?
175 175 225 755 755
BASE BASE BASE BASE BASE
NO NO NO NO NO
A-21
-------
A-3.3 BOILER DATA
1. Maximum Continuous Heat Input (MM Btu/Hr):
Unit 1 or 2 Unit 3 Unit 4 or 5
1760
2230
7040 (assuming 755 MW
net peak)
2. Maximum Continuous Generating Capacity:
Unit 1 • 2 3 4 5
Gross(MW) 190 190 245 778 778
Net (MW) 175 175 225 755, 755
2a. Maximum Heat Input (MM Btu/Hr): - .
Unit 1 or 2 Unit 3
1006 Btu/K WHR 2230
Unit 4 or 5
7460 (assuming 800 MW
net peak)
3. Flue Gas Temperature to Stack: 120°P(Units 1,2,3);
220°F(Units 4,5)
4. Maximum Continuous Flue Gas Rate to Stack:
Unit 1 or 2
Unit 3
Unit 4 or 5
640,000 ACFM at 120°F
800,000 ACFM at 120°P
3,000,000 ACFM at 220°F
These values are typical continuous flue gas flows at
full load. Flows can vary + 10%, depending on excess
air, coal composition, and water injection rate.
5. Flue Gas Analysis:
Flue Gas Component
(and inerts)
Units 1-2-3
4.3 - 7.0
12 - 14.8
80.6 - 81.5
Units 4-5
3.8 - 5.7
13-6 - 15.3
80.6 - 81.5
A-22
-------
6. Flue Gas Recirculation For NOX Control3-:
7. Boiler Manufacturer: R-S (Units 1,2); FW (3); B&¥ (4,5)
8. Years Boilers Placed In Service: 1963 (1,2); 1964 (3)
1969 (4); 1970 (5)
9. Boiler Service (Base Load, Peak, etc.): Base
9a. Wet Bottom __- Dry Bottom X
Units: 1/23 4 5
10. Stack Height Above Grade (Ft): 250 250 300 300
lOa. Stack Diameterb (Ft): 17.7 14 28.5 28.5
lOb. Velocity Of Gas (Exit):
Stack 1-2 Stack 3 Stack 4 Stack 5
Flue Gas Velocity (fps)
Leaving Stack 43-86 87 78 78
At Full Load (Net) (175-350• MW)(225 MW) (755 MW) (755 MW)
lOc. Exit Gas Temperature (°F):
115-122 122-130 205-245 205-245
lOd. Number Of Liners/Boiler: 1 for each stack
aFG recirculation originally installed for units 4 & 5. Later
was taken out due to severe mechanical difficulties.
bStack diameter at top. For units 1,2, and 3 includes liner.
A-23
-------
11. Boiler Operations (Hours/Year-1978):
Unit 1
6994
Unit 2
7933
Unit 3
7296
Unit 4
5300
Unit 5
4874
Ha. Boiler Operations (Hours/Year § full load): N/A
12. Boiler Capacity Factor3-: (1978):
Units
12345
67.60 83.87 72.55 49.62 48.35
13,
Boiler Operating Pressure (psig):
Units
1 & 2 3
1925
2125
4 & 5
3590
14. Boiler Superheat Heat Temperature (°F):
1005
1005
1000
I4a. Boiler Reheat Temperature (°F):
1005
1005
1000
I4b. Economizer B.F.W. Outlet Temperature (°F):
510
560
650
15. Ratio Of Fly Ash/Bottom Ash = 80/20
a Defined as:
KWH GENERATED IN YEAR
(Net) Max Cont. Generating Capacity in KW x 8760 HR/YR
A-24
-------
16. Burners:
Type
Manufacturer
No.Per Unit
Coal (#/Hr/
Burner)
Primary Air
(% of Total)
Secondary Air
(% of Total)
Total Excess
Air
Units
1&2
Flair
APS/
Design
18
3
Flair
Foster/
Wheeler
18
4&5
Cell
Bab co
Wile
18 ce
54 burner
nozzles
10,280
15%
85%
2Q%
14,450
10%
90%
20% :
50,000
N/A
N/A
18%
17. Fans (FD & ID):
Forced Draft
Type
Manufacturer
Induced Draft
Type
Manufacturer
Primary Air
Type
Manufacturer
(FD)
Air Foil
Green Fuel
Economizer Co.
(ID)
Paddle Wheel
American Std.
(PA)
Radial
Buffalo Forge
Company
Air Foil
Westinghouse/
Sturtevant
Paddle Wheel
American Std.
Radial
Westinghouse/
Sturtevant
Air Foil
Westinghouse/
Sturtevant
N/A
N/A
Air Foil
Clarage Fan
No. Per Unit
F.D.
I.D.
P.A.
2
2
3
2
2
3
4
N/A
2
A-25
-------
Units
1&2
4&5
Rating (CFM each)
F.D. 258,000
I.D. 387,000
P.A. 68,250
292,000 535.,000
435,000 N/A
75,920 (calc.) 218,000
Differential Pressure (inches of water)
F.D.
I.D.
P.A.
11.6-
16.0
24.5
10.25 23.7
12.04 ' N/A
24.5 (approx) 37
18. Windbox:
Compartment Single
Controlled Register
Mode of control Manual
Single
Register
Manual
Single
Register
Manual
19. Steam Temperature Control (Superheated & Reheated):
Attemporator
Cap. Ib/hr
73,000
40,000
225,000
20. Pressure Profile Throughout The Unit:
Units 1-2-3
Balanced Draft
Unit 4-5
Pressurized
A-26
-------
21. Air Preheater:
1&2
Units
4&5
Type
No. Per Unit
Flue gas ATe
Air ATe
Design flow:
Gas -Mlb/Hr
Air Mlb/Hr
DesignA P
(-in. of H20)
Flue gas inlet
Temp °F
Air preheater
outlet temp.
Flue gas pre-
outlet temp.
Lj ungstrom
Horizontal
2
434°F
562°F
Lj ungstrom
Vertical
2
4l8°F
516°F
Ljungstrom/
Tubular/Horiz,
2/1
4l4°F
485°F
1,790 ,
1,462
3.40
720
642°F
286°F
2,285
1,980
2.70
720
596°F
302°F
7,100
6,290
3.2
,655
565°
241°
F
F
22. Number Of Sections Of Air Ducts:
22a. Size Of Air Ducts:
Each of the air ducts on Units 1 and 2 has a cross sectional
area equal to approximately 120 square feet. Each of the
air ducts on Unit 3 has a cross sectional area equal to
approximately 191.8 square feet. Unit 4&5 not available.
e These figures correspond to normal operating
conditions at rated full load.
A-27
-------
23. Number Of Flue Gas Ducts:
1&2
Units
4&5
23a. Size Of Flue Gas Ducts:
Each of the flue gas ducts on Units 1 and 2 has a
cross sectional area equal to approximately 132 square
feet. Each of the flue gas ducts on Unit 3 has a cross
sectional area equal to approximately 186.4 square
feet. Unit 4&5 not available.
A-3.4 FUEL DATA
1. Coal Analysis (as received) (56): Max. Min. Avg,
s — 0.7
Ash 22
2. ' Total Ultimate Analysis (Wt5O
Component
Ash
S
Moisture
Oxygen
Hydrogen
Nitrogen
Carbon
Btu/lb
Average
22
.7
10.8
10,4
3.76
0.84
51.4
8800
A-28
-------
A-3.5 ATMOSPHERIC EMISSIONS
1. Applicable Emission Regulations: Particulates _SOp
a) Current requirements
Maximum allowable.emissions
Lb/MM Btu input to boiler
(1,2,3) 0.05
(4,5) 0.50
NO
- 0.7
FED
b) Future requirements
For Dec.1982 0.53 0.5
(4,5) 0.05
2. Plant Program For Particulates:
Units 1,2 and 3 have a total of six Chemico venturi wet
scrubbers, each scrubber module handling 400,000 -
500,000 ACFM @ 340°F.
Units 4 and 5 have Research Cottrell electrostatic
precipitators. The flue gas flow from Unit .4 or 5 to
the corresponding precipitator is 3.1 x 10^ ACFM §
250°F. . . • .
3. Plant Program For S02 Reduction: Use of low sulfur
coal
4. Plant Program For NOX Reduction: Modified burners
in Units 1 & 2. (KVB NOX spoilers installed).
A-29
-------
A-3.6 PARTICULATE REMOVAL
1. Type
Type
Manufacturer
Units 1-2-3
Venturi Wet Scrubber
Chemico
Efficiency %
design/actual 99.6/99.6
Specific collection
area N/A
Total collectiona
area N/A
Units 4-5
E.S.P.
Research
Cottrell
97/97
138 Ft2/1000
373,000 Ft2/BLR
Design basis, sulfur of fuel: 0.7%
Emission rate Lb/MM Btu (under normal operation
conditions):
0.9-1.5
0.58-0.7
0.9-1.5
0.8-1.4
1.4-1.9
0.7-1.1
Unit 1 or 2
S02
NOX
Total
particulates 0.02-0.05 0.02-0.05
2. Solids Collection System:
Present operating condition:15
At maximum capacity: YesX: No
4 or 5
0.25-0.5
a For each boiler there are two cold side ESP's with each
having 16 electrical sections for a total of 32 electrical
sections per boiler.
b Bottom ash is hydraulic conveyed to dewatering bins & then
trucked to coal mine (for units 4&5). The dewatered bottom ash
is trucked to ash disposal pond and used to sand blast. Fly ash
is pneumatically conveyed and sent to storage silo and then
trucked to the coal mines.
A-30
-------
The inlet fly ash loading range is 12-20 lb/106 Btu
The inlet bottom ash loading range is 3-5 lb/106 Btu
The inlet*flue gas temperature is 280-350°F for Units
1-2-3, and 230-280°F for Units 4-5.
A-3.7 SCRUBBER TRAIN SPECIFICATIONS (for venturi scrubbers)
1. Scrubber:
Type
Chemco wet venturi
Liquid/gas ratio 16.9 gpm/1000 ACFM (Units 1,2 and 3)
Gas velocity Information not available (Varies to
maintain pressure differential.
Materials of
construction
Top gas inlet - unlined. carbon steel
wetted center; 316 stainless steel
high gas velocity throat; 316
stainless steel intermediate plumb
carbon steel lined with polyester
glass reinforced resin. All other
scrubber shell sections and supports
are coated with polyester glass resin.
Internal mist eliminators are of
polypropylene glass reinforced.
A-31
-------
Internals:
Type
No. of stages
Internal mist eliminator
6 pass arrangement
Type and size of
packing material: Slats and spacers in a 30' OD x 13' ID
and 34' OD x 15' ID
Packing thickness
per stage: 2 9/16"
Material of construction:
Packing - polypropylene glass reinforced
Supports - 316 stainless steel/fiberglass
2. Clean Water Tray (at top of scrubber): N/A
3. Mist Eliminator (M.E.)
Type.
No. of passes
Space between vanes
Angle of vanes
Size of M.E.
Distance between top of liquid
inlet and bottom of M.E.
Position
Materials
Method of cleaning
Baffle
6
3"
55°
2'3 3/8"
78' (Units 1&2).
85' (Unit 3)
Vertical gas flow
Horizontal blade
eliminators
Polypropylene
Fiberglass reinforced
Annual overhaul and
periodic sprays
A-32
-------
A-3.8 CALCINING AND/OR SLAKING FACILITIES
One lime slaker, having an average on-line capacity of
0.42 tons per hour, serves the scrubber systems on Units
1-2-3.
1. Source Of Water For Slurry Make-up Or Slaking Tank:
Morgan Lake is the source of make-up water to the
slaking tank.
A-3.9 DISPOSAL OF SPENT LIQUOR
1. Transporting:
Fly ash from Units 1-2-3 is slurried from the
thickener to the ash ponds at the rate of 18 tons/hr
for Unit 1 and 2 and 20 tons/hr for Unit 3.
Fly ash from Units 4-5 is trucked to the Navaj o Mine
pit at the rate of 71 tpns/hr for each unit.
There is no scrubber sludge to dispose of.
2. Oxidizer:
This section is not applicable to the systems at the
Four Corners Power Plant.
A-33
-------
Source of water and pressure: Morgan Lake § 60-80 PSI
Flow rate during cleaning: Not available
Frequency and duration of cleaning: Not available
U. Reheater:
Type: Steam coils SS-316-L Units 1-2-3 (No longer in
operation)
5. Scrubber Pressure Drop Data (inches of water):
Units 1-2-3
25" W.G.
2" W.G.
Removed
Not available
Not applicable
28" W.G.
Particulate scrubber
Mist eliminator
Reheater
Ductwork
Total FGD system
Total part system
6. Fresh Water Make-up Flow Rates:
The Four Corners Power Plant has no S02 scrubbers, but
it does have particulate scrubbers on Units 1-2-3. Units 4
and 5 have no scrubbers. Steam and cooling water blowdown
are not available as make-up to the scrubber systems on
Units 1-2-3. Morgan Lake provides the make-up water for
the particulate scrubbing systems on Units 1-2-3.
7. Bypass System: None
A-34
-------
3. Clarifiers (thickeners):
Number: 2
Dimensions^ 100 Ft. I.D.
Concentration of solids in underflow:
Amount of flocculant: None
4. Rotary Vacuum Filter: N/A
5. Sludge Fixation: N/A
40 wt$ solids
A-3.10 COST DATA
The original estimated cost for the installation of
particulate scrubbers on Units 1-2-3 was $6,900,000. The
final installed cost, including the costs of replacement
of the recycle pump and installation of the lime feed
system was $27,780,000. The operating and maintenance
costs during the year of 1978 for the scrubbers on Units
l_2-3 were:
Operation of the facilities
Flyash & sulfur sludge removal
Replacement power costs
Operations Total
Maintenance
Total
$ 250,000
954,000
2,649.000
$3,853,000
2.798.000
$6,651,000
A-35
-------
A-3.11 MAJOR PROBLEM AREAS: (Corrosion, Plugging, etc.)
1. S02 Scrubber, Circulation Tank, Pumps, and Nozzles -
Problem/Solution:
Corrosion problems were experienced with the stainless
steel linings, and erosion problems occurred with the
coating used in the scrubber. Also, corrosion and
erosion problems were encountered with the pumps related
to the scrubbers. Initial operation of the scrubbers
resulted in severe pluggage which was reduced by the
addition of lime to control the pH. Pluggage of nozzles
within the system is still a problem with small chunks
flaking off within the scurbber system and being
recycled to the nozzles. The system does not have
S02 removal capabilities at this time, and no
circulation tank is being used.
2. Mist Eliminator - Problem/Solution:
The original mist eliminators were constructed from a
polyethylene material and were very flammable. As a
result of repairs in the scrubber areas, there were two
serious fires that resulted in extended down time to
repair damage. Mist modules have since been changed to
a fiberglass material that is less flammable. Another
problem related to the mist eliminators was-fouling,
that required more water to keep the modules clean than
the design called for. This problem has been alleviated
by the addition of lime for pH control.
3. Reheater - Problem/Solution:
The reheaters for these units were a 316L stainless
steel which was found to be very sensitive to the acid
mist in the flue 'gases. After a few months of use, the
A-36
-------
reheaters were no longer servicable and have since been
removed from service. Revoval of the reheaters caused
serious stack deterioration problems resulting in the
necessity to replace the stack liners.
4. Venturi Scrubber, Circulation Tanks and Pumps - Problem/
Solution:
Shortly after startup, the scrubber venturi and recycle
pumps had a serious erosion-corrosion problem and
required redesigning. The venturi of the scrubber w? 3
modified to include acid brick to prevent
erosion-corrosion in the throat area. The recycle pumps
were replaced with a rubber lined pump to improve the
reliability of the pump system.
5. I.D. Booster Fan and Ductwork - Problem/Solution:
The original ID fans were a 316 stainless steel. Due to
vibration problems these failed and had to be replaced
with inconel fans. Since that time, the .only problem
has been the erosion on the blades due to ash carryover
from the venturi scrubber. The duct from the ID fans to
the.stack was originally lines with 316L stainless steel
and was plug welded to a carbon steel sheel. This
arrangement resulted in deterioration and failure of the
stainless steel in the area of the welds. This lining
has since been replaced with a fiberglass lining and is
repaired on an annual basis.x
6. Limestone Milling System or Lime Slaking -
Problem/Solution:
'' *
The present system is a quick lime system with a batch
slaking process and problems in this area relate to the
size of the slaking unit and the relatively short
A-3 7
-------
slaking period required to fill the batch tanks. This
short time does not permit continuous operation of the
slaking unit at designed temperatures which has caused
some carryover of lime into the grits system. Scaling
of the system continues to be a problem in the lime
slaker area due to the hard water that is being used for
this process.
7. Sludge Treatment and Disposal - Problem/Solution:
This item is not applicable to the scrubbing system at
the Four Corners Power Plant. However, the following
comments apply to fly ash disposal from the scrubber
systems.
Flyash from the scrubber operation is settled in
thickeners and pumped to settling ponds for dewatering.
This has been a relatively reliable system. The
principal problem has been chips from the scrubber
operation causing plugging or fouling of the underflow
removal piping. On occasion, this has resulted in the
failure of the thickener rakes and a subsequent outage
of the thickener for clean out and repair.
8. Description of scrubber control methods under
fluctuating load.
Scrubber control consists of adding lime and
recirculating slurries until the specified pH limits and
percents of solids are obtained. Frequent scaling of
the pH probes which are located in the scrubber recycle
slurries have occured. Good pH control is essential for
scale-free operation, especially when the scrubber
liquids are recycled.
A-38
-------
A-3-12 METHODS OF MEASURING EMISSIONS
EPA methods 6,7 and 5 respectively, are used to measure
S02, N0x"and particulate emissions.
A-39
-------
SECTION A-4
RETROFIT DESCRIPTION
A-4.1 NOX EMISSION CONTROL
Table A-1 develops details of the retrofit examples
investigated for NOX reduction to boilers 1,2,3,4 and 5.
Figure A-11 shows the arrangement of the burners, the
overfire air (OFA) ports, and the curtain air (C.A.) ports
for the Number 1 and 2 boilers. Figure A-12 shows the
arrangement of the burners, the OFA ports, and the curtain
air ports for the Number 3 boiler.
Figure A-13 indicates the burner arrangement and windbox
compartment modifications for Boilers 4 and 5 after being
retrofitted. Eighteen sets of cell burners are in use, and
each cell burner has three cone nozzles. The number of sets
of cell burners is different on each wall. One wall has 8
sets'of burners, and the opposite wall has 10 sets.
Assuming each cone nozzle is equivalent to an individual,
dual-register burner, 54 sets of dual register burners would
s
be required for a retrofit. It may be necessary to provide
a total rearrangement of burners to accomodate the
different flame characteristics of the B&W, dual-register
burner. The windbox is compartmented horizontally • as part
of the retrofit. A vertical partition at the middle of each
compartmented windbox facilitates air distribution.
secondary air control dampers are provided at both ends of
each compartmented windbox.
A-40
-------
=a=
s
o
H
&^
o
CD
P
X!
O
EH
H
P4
O
cd
EH
oS
^fc
co <=a
t—
g 1
0
1 — 1
0
0
on
=8=
UN
^T i,
OJ 0
-P
CO
. O
OJ
=tfc
c&
rH
=s=
SH
0
O
P
O CO
CT\
t-H >a
0
rH
•rH
PC;
crj
1^
I
0
CQ •
O
ft-
ft!
O
bO
G
•H
SH
•rH
PH
rH
rH
crj
£S
\
0
rH
bO
CO
bO
G
•H
SH
•H
PH
rH
rH
crj
£^
1
0
1
1 — 1
bO
•H
CO
CQ
CD
CQ
-=r ;
UN
o
co
ON
CO *O UN
0 O OJ MD
rH OJ UN
on
£-,
CD
P
CO
•H
bO
0
CD
CQ
co
0
X
crj
LP\
O
co
«O MD
OnOJ ON
OJ LPv
O
'^f
rH
CQ O "UN
0 on UN
>H O
on
rH O
on
o
X •>
CO
0
UN
t- X!
rH V
O =
rH ON
rH
OO
CQ
0
X
cr-1
OJ
^o
on
->o oj
co CM -=r
on vo
co
oo
oo
ra o OJMD
0 oo •»
>H -=r
LO
OJ
O
03
>3 SH
EH <
. .
CO
£_,
0
G
;3
CQ
Column
rH CO
crj G
o a
•H 3
-P rH
S-. 0
0 O
*••->
X S-i
03 0
£_, r_J
0 SH
t, CQ
CQ OH
O
'v • •
GgsQi-
s. s
Boiler
r Outlet (°P
X 0
S -P
PH crj
O 0
CO ,G
0
S-. SH
•H CM
fe^. P
rH <
crj SH
O vH •
•H < ft
p a
0 CO 0
!-, co EH
O 0
0 O £H
£1 X iH
EH pq <
EH
H
PH
O
ecj
EH
DH
PH
o
w
PH
>H
EH
SH
•H
•
rH
CO P
-P SH
S SH 0
PH O PH
O PH
— 1
•
on
X!
o
£3
T3
G
•H
3£
•o
0
-P
G-
0
a
p
SH
crj
ft
a
0
O
•
•=r
A-41
-------
in
CO CQ
m co 0 «a!
53 53
•a
a>
3
G
•H
•P
ES
O
O
on
m ra
CD CD
53 53
o ra ra ra
CM o CD CD CD
rt
ra
o
-P
c
o
O
H
Q
W
K
O*
5s
CVJ
ra m
CD CD
o ra m m
CJ O CD CD CD
co
o
•H
ti-l
•H
•O
O
a
TJ
CD
CD
C •
O
Be.
W
fc.
O
CG
K
I
»
T—
I
O
PL*
ra
-p
ce
o
•a!
o
M
En
M
0
w
E-<
O O
O W
CO W
W 53
OH
« O rH
4i
cO C
O O
CD O
> cO
O O CD
bO C
bO hO C cfl
4^ 4J O
OO K
•H O t.
O C
S C -P
O rH
-H Q O O
bO
••H
B
n-i ra
O CD
^j
c ^
O E-i
•HCO
4-> a ra 53
CD
•H O
EnrH
CO
A-42
-------
1
^S^x>^
^N;^
1
f
^
•*»
j
H to
K EH
>, r
J V.
'
ft I
01
4J
•H
fi
CO
-p
3
£ to
-------
W
O
(3
D
Q) -H
S O
M
Q)
O 0)
M -P
4-> CO
0) O
(U
A-44
-------
FRONT VIEW OF BURNERS ON "A"-WALL
—©- -©—
—©— — ©—
._ .j _ _ _l_
DUAL REGISTER
BURNER (TYP)
(NEW)
— g>— -©—
:l
.-I
--H-
-®-
. 1
I
BURNER SECONDARY AIR
CONTROL DAMPER (NEW)
(TYP)
_t
COMPARTMENTED
WINDOW (NEW)
?LJRNER JL"(TYP)
k
T
1
i
i
i
EH
U
Q
M
X
-
X
*-" !
'
*^\
\
i
"A "-WALL
WINDBOX
'IREBOX #4,
UNITS
-
WINDBOX
"B"-WALL
5
'
'
^ <-*
\S*
\s*
s
s
EH
U
Q
(V
M
•
J
<=i
|
1
PARTITION
ID
Q!
-_-t--H--
-t- H —
—©_ —@_
r::r;
COMPARTMENTED
WINDOW (NEW)
4 1 4__
i ~ _ z. )_r_ i _
• (j>— —&—
H- h-
•-]•
:r:i~
—+—i--
—®— —@—
-T--+-
4-
«^
FRONT VIEW OF BURNERS ON "B"-WALL
Figure A-13.- Retrofit of Units 4 and 5 with 54 sets of Dual Register
(B&W) burners and compartment!zed windbox.
A-45
-------
A-4.2 PARTICIPATE EMISSION CONTROL
The related data and the calculated results for
particulate emission control retrofit work for Units 1 & 2
are shown in Table A-2. Unit 3 data are shown in Table
A-3, and Table A-4 has the information for Units 4 and 5.
For Units 1,2 & 3, hot side ESP retrofitting was not
considered due to space limitations and relocation
requirements for the air preheaters.
A-4.3 S02 EMISSION CONTROL
System requirements are based on 90% S02 removal for
wet scrubbing and 70% S02 removal for dry scrubbing
with 0.7 percent sulfur coal.
Sizes have been selected based on meeting S02 removal
requirements.
To evaluate the dry scrubbing systems, the corresponding
number of wet scrubber modules that would produce a 70%,
SC>2 removal has been used. Also, the flue gas rate to
be bypassed given in the table is the appropriate rate to
produce the effectively 70%, S02 removal when using a
90%, wet S02 removal system.
A-46
-------
TABLE A-2.- RETROFIT DATA FOR ELECTROSTATIC PRECIPITATORS
AND BAGHOUSES - UNITS 1 AND 2
Power Plant:
Boiler No.
Gross. MW/Boiler:
Flue Gas (ACFM/Boiler
300°F
68°F (SCFM)
170°pa
125°Fb
Four Corners
#1 & #2
190
2610 SCFM/MW § 68°F):
713,800
495,900
623,400
576,100
Per Boiler
Total Req'd Particulate
Collection Area '(FT2)
Existing Collection Area
(FT2)
Collection Area to be
Added (FT2)
No. of ESP Trains
No. of ESP Modules/Train
Total No. of Modules to be
Added
Total Area Added (F+2)
ESP Electrical Sectionalizing
Total No. of Electrical
Sections Required Per 5 MW
Existing Electrical Sections
No. of Elec. Sec/New Module
Total No. of Elec. Sections
Added
Total
Baghouse
Cold Side
ESP
713,800
0
713,800
2
11
22
712,800
Wet
Scrubbing
356,900
0
356,900
20
369,460
Dry
311,700
0
311,700
17
314,000
MW/Section
38
0
2
44
44
4.32
alncludes water from dry scrubbing
blncludes 13.3 percent water from wet scrubbing
A-47
-------
TABLE A-3.- RETROFIT DATA FOR ELECTROSTATIC PRECIPITATORS
AND BAGHOUSES - UNIT 3
Power Plant:
Boiler No.
Gross MW/Boiler:
Flue Gas (ACFM/Boiler)
300°F
68°F (SCFM)
170°Fa
125°Pb
Four Corners
#3
245
920,400
639,450
803,800
742,900
Per Boiler
Total Req'd Particulate
Collection Area (FT2)
Existing Collection Area
(FT2)
Collection Area to be
Added (FT2)
No. of ESP Trains
No. of ESP Modules/Train
Total No. of Modules to be
Added
Total Area Added (F + 2)
ESP Electrical Sectionalizing
Total No. of Electrical
Sections Required Per 5 MW
Existing Electrical Sections
No. of Elec. Sec/New Module
Total No. of Elec. Sections
Added
Total
Cold Side
ESP
920,400
0
920,400
3
10
30
972,000
50
0
2
60
60
Baghouse
Wet Dry
Scrubbing
460,200
0
460,200
402,000
0
402,000
25
461,800
22
406,400
MW/Section
4.13
alncludes water from dry scrubbing
blncludes 13.3 percent water from wet scrubbing
A-48
-------
TABLE A-4.
RETROFIT DATA FOR ELECTROSTATIC PRECIPITATORS
AND BAGHOUSES - UNITS 4 AND 5
Power Plant:
Boiler No.
Gross MW/Boiler:
Flue Gas (ACFM/Boiler)
300°F
68°F (SCFM)
170°Fa
125°Fb
Four Corners
#4 & #5
778
2,923,000
2,031,000
2,553,000
2,359,000
Baghouse
Per Boiler
Total Req'd Particulate
Collection Area (FT2)
Existing Collection Area
(FT2)
Collection Area to be
Added (FT2)
No. of ESP Trains
No. of ESP Modules/Train
Total No. of Modules to be
Added
Total Area Added (F+2)
ESP Electrical Sectionalizing
Total No. of Electrical
Sections Required Per 5 MW
Existing Electrical Sections
No. of Elec. Sec/New Module
Total No. of Elec. Sections
Added
Total
MW/Section
Cold Side
ESP
2,923,000
373,000
2,550,000
8
10
80
2,592,000
Wet
Scrubbing
1,461,500 1,
0
'1,461,500 1,
80
1,477,800 1,
Dry
276,500
0
276,500 '
70
293,000
155
32
2
160
192
4.03
alncludes water from dry scrubbing
^Includes 13.3 percent water from wet scrubbing
A-49
-------
SECTION A-5
RETROFIT COSTS
Retrofit capital and annualized costs for the alternatives
discussed in Section A-l are included in this section. The total
required plant investment costs given for each alternative do not
include the costs of removing and/or relocating any existing
equipment that may be associated with the particular retrofit
alternative. The cost for land required for sludge disposal
(from S02 removal) and the associated sludge transportion to
the disposal site are not included in the retrofit plant
investment or annualized operating costs. Also, the following
cost items are realized but not included in the total retrofit
capital and annualized costs:
(1) Site preparation
(2) Cost of down-time
(3) Additional stack lining if flue gas desulfurization is
installed
(4) Costs for removing ventiru scrubbers on Units 1, 2, and 3
Working capital, the money required to operate the new equipment
associated with the retrofit, has been calculated for each
retrofit alternative.
A-50
-------
TABLE A-5. CAPITAL INVESTMENT COSTS FOR RETROFITTING
THE FOUR CORNERS POWER PLANT - MILLIONS OF THIRD QUARTER
1979 DOLLARS (a)
Cost
Item
Alternative
2a
1. NOx Control
2. Particulate control
3. S02 control
4. Emission Monitoring
5' Auxiliary Boiler
6. Replacement of Power
Generating Capacity
7. Working Capital .
total
28.93
0.00
369.47
1.65
17.35
79.85
36.6-4
28.93
156.80
369.47
1.65
17.35
97.52
45.25
28.93
137.71
303.25
1.65
0
77.19
35.75
28.93
(b)
272.63(b)
1.65
0
38.54
24.04
533.89U) 7l6.97(c) 584.48(c) 365.79(c)
Dollars per Kilowatt
Gross Generating
Capacity
245
329
268
(a) Includes direct and indirect costs
(b) Costs of particulate and S02 control are combined
(c) See Section 4.5 for other costs not estimated
-------
TABLE A-6 ANNUAL COSTS FOR RETROFITTING THE FOUR CORNERS
POWER PLANT - MILLIONS OF THIRD QUARTER 1979 DOLLARS PER YEAR (a)
Cost
Item
Alternative
2a
1. NOx Control
2. Particulate Control
3. S02 Control
4. Emission Monitoring
total
Mills per kilowatt hour
of net power generation
(current net less retrofit
power requirements at
65 percent of maximum
net load)
5.265
0
140.924
.384 .
5.265
34.418
140.924
.384
5.265
26.439
110.912
.384
5.265
(b-)
90.510(b)
.384
I46.573(c) l80.99Kc) l43.000(c) 95.159(c)
12.2
15.2
• 11.9
7,9
(a) Includes fixed capital charges
(b) Costs of particulate and S02 control are combined
(c) See Section 4.5 for other costs not estimated
-------
4,
5.
SECTION A-6
REFERENCES
Meeting notes - N. Gonzalez/N. Master, Pullman Kellogg -
meeting with D.J. Campbell/J. Weiss, Arizona Pulbic
Service., Fruitland, NM, 8 May 1979
Letters from C.D. Jarman, Arizona Public Service to N.
Master, Pullman Kellogg, 5 July 1979
Meeting notes - N. Gonzalez/N. Master, Pullman Kellogg
- meeting with J. Weiss, et'al, Arizona Public Service
and S. Cuffe/J. Copeland, EPA,•Phoenix, Arizona, 17
July 1979
Letters from J.C. Evans, Snell & Wilmer to N. Master,
Pullman Kellogg, 1 June and 9 August 1979
Drawings received from APS:
Bechtel DWG 73005-2, Plot Plan, Rev 2, 2-20-70
Plot Plan, Rev 3-A, 9-17-63
General Arrangement Section A-A,
Ebasco DWG G-162385,
Ebasco DWG G-162390
Rev 5, 11-13-62
Ebasco DWT G-171673
General Arrangement Section A-A,
Rev 4, 3-6-64
Bechtel DWG 73496-1, General Arrangement Section, Rev 1,
2-14-68
A-53
-------
-------
APPENDIX B
EXAMPLES OF RETROFITTING THE MOHAVE POWER STATION
-------
CONTENTS
SECTION
PAGE
CONTENTS
FIGURES
TABLES
B-ii
B-iii
B-iv
B-1.0 GENERAL
B-l.l Retrofit Alternatives
B-1.2 Plant Characteristics
B-1.3 Flue Gas Ducting Requirements
B-l
B-l
B-4
B-6
B-2.0 BACKGROUND DATA B-l3
B-2.1 Plant Description B-13
B-2.2 Steam Generator Description B-14
B-2.3 Existing NOX Control B-16
B-2.4 Existing1Particulate Control B-16
B-2.5 Existing S02 Control B-17
B-3.0 PLANT SURVEY FORM B-l8
B-3.1 Company and Plant Information ' B-18
B-3.2 Plant Data B-18
B-3.3 Boiler Data B-19
B-3.4 Fuel Data B-21
B-3-5 Atmospheric Emissions . B-22
B-3.6 Particulate Removal B-22
B-3.7 Fresh Water Make-Up Flow Rates and Points .of
Addition B-23
B-4.0 RETROFIT DESCRIPTION
B-4.1 NOX Emission Control
B-4.2 Particulate Emission Control
B-24
B-24
B-24
B-5.0 RETROFIT COSTS
B-6.0 REFERENCES
B-29
B-32
B-11
-------
FIGURES
FIGURE
B-l Addition of wet S02 scrubbing modules.
B-2 Addition of baghouses and wet S,02 scrubbing
modules.
B-3 Addition of dry S02 scrubbing modules with
baghouses.
B-4 Plan arrangement of wet S02 scrubbing modules.
B-5 Elevation arrangement of wet S02 scrubbing
modules.
B-6 Arrangement of ducting for baghouses and wet S02
scrubbing systems.
B-7 Arrangement of wet S02 scrubbing and baghouse
modules.
B-8 Arrangement of dry S02 scrubbing and baghouse
modules.
B-9 General plot plan of the Mohave power station.
B-10 Location of OFA ports and ducts for a twin furnace
Mohave boiler.
PAGE
B-2
B-3
B-5
B-7
B-8
B-9
B-10
B-12
B-15
B-27
-------
TABLES
Table
B-l
B-2
B-3
B-4
Retrofit for NOX Reduction
Retrofit Data for Electrostatic Precipitators and
Baghouses - Units 1 and 2
Capital Investment Costs for Retrofitting the
Mohave Power Plant
Annual costs for Retrofitting the Mohave
Power Plant
Page
B-25
B-28
B-30
B-31
B-iv
-------
SECTION B-l
GENERAL
B-l.l RETROFIT ALTERNATIVES
The four alternative examples considered in this appendix
follow. All alternatives include maximum NOX control
and the installation of emission monitoring systems for
opacity, S02, and NOX.
Alternative 1 - Add wet scrubbing to achieve 90% S02
removal and retain the existing ESP's.for particulate
emissions control to a level of 21 ng/J heat input (0.05
lb/10^ Btu),. Figure B-l shows the general arrangement
of the plant with the addition of the wet S02
scrubbing modules.
Alternative 2 - This option adds baghouses plus wet
S02 scrubbing for 90% S02 removal, and it keeps
the existing ESP's for control of particulate emissions to
a level of 13 ng/J heat input (0.03 lbs/106 Btu).
Figure B-2 shows the plant's general arrangement with the
added S02 scrubbing modules and baghouse modules.
Alternative 2a - The retrofit for particulate control in
Alternative 2a is the same as for Alternative 2, using
baghouses for cost comparison with Alternative 3. The wet
S02 scrubbing in this case is based on 70% S02
removal for cost comparison with semi-dry scrubbing in
Alternative 3.
B-l
-------
(fl
-------
tfl
Q)
§•
01
c
•H
O
01
CM
O
CO
4J
-------
Alternative 3 - Semi-dry, S02 scrubbing (spray drying)
for 70% S02 remoyal is used in this case. Also
baghouses are used as dry collectors with the existing
ESP's for control of particulate emissions to a level of
13 ng/J heat input (0.03 lbs/106 Btu). Figure B-3
shows this general arrangement with the addition of the
S02 scrubbing modules and baghouses.
B-1.2 PLANT CHARACTERISTICS
Major revamp work required to install the S02 and
particulate control equipment does not involve relocation
of existing equipment. Characteristics of the plant site,
existing equipment, and space requirements for each
retrofit example are shown in the following list (!_).
A. The existing electrostatic precipitators are located
downstream of the air preheater (cold side).
B. Each of the two boilers has two primary-air (PA) fans
and two forced-draft (FD) fans.
C. The number of S02 scrubbing modules used is based
on the total calculated flue-gas rate from each
boiler.
One S02 scrubbing module per boiler is provided as
a spare.
One flue-gas reheater is required for each wet S02
scrubbing module for 90$? S02 removal.
One flue-gas, booster fan is required for each
scrubbing module.
The individual scrubbing modules are provided with
dampers. This provision allows the individual modules
to be isolated for maintenance.
Tie-in of retrofit equipment to the power plant is
based on completion during normal power plant
maintenance turn-arounds of 3 to 6 weeks.
D
E
P
G
H
B-4
-------
oo
DO
[Q
Q)
W
(0
X!
,
-P
•H
W
(U
i
tn
£
-H
5-1
U
03
CM
O
CO
o
s
o
•rH
I
PQ
0)
B-5
-------
I. Tie-ins to the existing stack are the basis for the
retrofit examples and are realized to be extremely
difficult. Addition of a new stack is a possible
alternative but was not considered for estimating the
costs for retrofitting the plant.
J. An emergency bypass is provided around each SC>2
scrubbing system to allow operation of the boiler in
the event of a major FGD malfunction. Bypassing of the
particulate control equipment is not provided.,
B-1.3 FLUE GAS DUCTING REQUIREMENTS FOR RETROFITTING
Bypass duct and dampers are provided to enable the flue
gas to completely bypass the S02 scrubbing system.
Bypass duct locations for the three alternatives are
indicated below:
Alternative 1 - The emergency bypass duct is located
adjacent to the stack upstream of the wet SC>2
scrubbing module booster fans. Figure B-4 shows the
additional ducting requirements for the retrofit. Figure
B-5 shows the elevational view for the SC>2 scrubbing
module.
»v
Alternative 2 - The emergency bypass duct is taken after
the baghouse, just before the scrubber modules as shown in
Figure B-6. Figure B-7 shows the elevation vie'.w for the
S02 scrubbing and baghouse modules.
Alternative 2a - The bypass duct is the same as used in
Alternative 2. This duct can also be used during normal
operation to divert about 22.2$ of the total flue gas that
does not need to be treated, since the module remove 90%
S02 and this alternate requires only 70% overall
S02 removal. The bypass duct also provides the flue
gas requirements for reheat.
B-6
-------
K
S^
-------
UJ
in
Hi
^J
UJ
!i
-1"
»^
I
18-
IX!
CO
I
•H
43
O
CO
CN
o
-p
I
o
-p
cu
&1
O
•H
4J
(0
>
(U
ffl
5-1
&
-H
B-8
-------
CN
o
CO
4J
CU"
en
0)
en
f3
o
&
a
tn
d
-H
4J
O •
f3 to
T3 g
i
-P W
(0
O
to
I
pq
a;
s-i
B-9
-------
CO
(U
01
tn
(0
O
CO
0)
OJ
g
-------
Alternative 3 - The bypass duct is taken ahead of the
spray dryer and tied into the duct going to the baghouses.
This location of the emergency bypass permits operation of
the baghouse for particulate control when a spray dryer is
out of service. Figure B-8 shows the elevation view -for
the dry-scrubbing and baghouse modules.
B-ll
-------
IU
1/sA
8
1
V.
Q
§i^
3
Uj
Cu
CO
0)
0)
(0
2
I
(0
c
(0
tn
c!
-H
I
M
O
CO
CN
O
CO
(D
(0
po
PQ
-------
SECTION B-2
BACKGROUND DATA
B-2.1 PLANT DESCRIPTION (1,1)
The Mohave Generating Station is owned jointly by Southern
California Edison, (56$), The Los Angeles Department of
Water and Power, (20$) ,. Nevada Power Company, (14$), and
the Salt River Project, (.10$). It is operated by Southern
California Edison-(SCE). .
Total plant capacity is 1580 MW(net) and 1640 MW(gross).
-' It consists of two, 790 MW(net), 820 MW(gross),
forced-draft, pressurized-firebox, pulverized-coal-fired,
steam generators. The two, Combustion Engineering (CE)
boilers are equipped with tilting tangential 'burners.
Each boiler has 8 burner sets, and each burner set has 10
burner nozzles. There are, therefore, 80 burners per
boiler. The coal feed rate per boiler is approximately
856,000 Ibs/hr.
Pulverized subbituminous coal slurry is dewatered in
centrifuges before the coal is fed= into the burners.
The flue gas duct runs vertically down from the rear side
of each boiler, and it turns 90° to the horizontal. The
duct is then branched into two ducts, each connected to an
electrostatic precipitator. Ducts from each precipitator
combine into one duct and are connected to one side of the
single 500 ft. stack.
B-13
-------
Bottom ash from the boiler bottoms is mixed with water and
is transferred, as slurry, to three settling tanks. The
ratio of fly ash to bottom ash is 70/30. The water
overflow from each tank is stored in a fourth tank for
further settling of the ash, and then it is recycled to
complete the "closed-loop" slurry loop. The flyash from
the precipitators is pneumatically transported to a dry
ash storage bin.
The ash disposal site is sized for 30 years of disposal
capacity.
The flyash and the closed-loop, wet-ash, collecting
systems are currently operating at full capacity.
Therefore, the capacity of each collecting system has to
be increased for any increase in the amount of collected
particulates that occurs as a result of retrofitting.
Figure B-9 shows the general plot plan arrangement for the
Mohave power -station.
B-2.2 STEAM GENERATOR DESCRIPTION
Each boiler uses a forced draft system, and the firebox is
operated at 21" H20, positive pressure. Any further
increase in the firebox pressure to compensate for
pressure drop with the additional emission control
equipment may result in problems of flue gas leakage.
Therefore, the addition of induced draft (ID) fans have
been considered for the retrofit examples. The existing
boiler design does not allow for the use of new FD fans as
an alternative to the addition of the ID fans. Each
boiler has two primary air fans (PA) and two forced-draft
fans (FD) operating in parallel.
B-14
-------
oo
DO
fl
o
•H
4J
(0
-P
M
U
1
Q)
>
fd
X!
s
0)
X!
-P
M-l
O
C
(d
rH
Pu
-P
O
•H
o,
Q)
(U
O
CQ
0)
M
•H
PM
B-15
-------
The two, CE 790 MW (net) boilers are tangentially fired
and twin furnace designs. Neither boiler is equipped with
over fire air ports. The boilers are the dry bottom
type.
Vertical-shaft, regenerative, Ljungstrom air preheaters
are used. There are two air preheaters per boiler.
B-2.3 DESCRIPTION OF EXISTING NOX CONTROLS (3.)
Presently there are no specific controls for NOX
emission in the plant. Originally, an NOX monitoring
device was designed and installed at the plant by SCE, but
it is no longer in operation because of maintenance
problems. No retrofit work for NOX reduction has ever
been done.
B-2.4 DESCRIPTION OF EXISTING PARTICULATE CONTROLS (4_)
The existing particulate controls consist of two sets per
boiler of cold side, Research-Cot trell, Inc.,
electrostatic precipitators. The manufacturer's guarantee
was 97.9? removal efficiency, at 2,300,000 ACFM per
boiler, with flue gas temperature at 268°F, and with coal
sulfur contents over 0.3 wt$ . Expected performance was
98.6? under similar conditions.
Installation of the cold side ESP's was completed in 1970
and they have a life expectancy of 35 years. Two sets of
ESP's are operated in parallel for each boiler. Each
precipitator unit is enclosed in a 3/16" thick steel shell
and is segmented into four mechanical units. Each
precipitator is equipped with sixteen ash hoppers. Each
mechanically segmented unit has 37 flue-gas passages (38
collecting plates) that are 9-inches wide. The units are
B-16
-------
180'
i
OO
DO
O
V?
X
td
o
•rH
. -p
to
O
&t
0)
^
J3
i
0)
O
d
nS
iH
D4-
-P
O
•H
CU
0)
fl
-------
The two, CE 790 MW (net) boilers are tangentially fired
and twin furnace designs. Neither boiler is equipped with
over fire air ports. The boilers are the dry bottom
type.
Vertical-shaft, regenerative, Ljungstrom air preheaters
are used. There are two air preheaters per boiler.
B-2.3 DESCRIPTION OF EXISTING NOX CONTROLS (3.)
Presently there are no specific controls for NOX
emission in the plant. Originally, an NOX monitoring
device was designed and installed at the plant by SCE, but
it is no longer in operation because of maintenance
problems. No retrofit work for NOX reduction has ever
been done.
B-2.4 DESCRIPTION OF EXISTING PARTICIPATE CONTROLS (4_)
The existing particulate controls consist of two sets per
boiler of cold side, Research-Cottrell, Inc.,
electrostatic precipitators. The manufacturer's guarantee
was 97.9? removal efficiency, at 2,300,000 ACFM per
boiler, with flue gas temperature at 268°F, and with coal
sulfur contents over 0.3 wt$. Expected performance was
98.6$ under similar conditions.
Installation of the cold side ESP's was completed in 1970
and they have a life expectancy of 35 years. Two sets of
ESP's are operated in parallel for each boiler. Each
precipitator unit is enclosed in a 3/16" thick steel shell
and is segmented into four mechanical units. Each
precipitator is equipped with sixteen ash hoppers. Each
mechanically segmented unit has 37 flue-gas passages (38
collecting plates) that are 9-inches wide. The units are
B-16
-------
30-feet high and 21-feet long, and they are segmented Into
three sections. The inlet section is 9-feet long in the
flue gas flow direction and the second and third sections
are each 6-feet- long. The total effective collecting
plate surface area is 220,000 square feet for each
precipitator.
Each precipitator has 1,,036 discharge electrodes.
Electrically, each unit consists of four sections in the
gas flow direction, while there are only three mechanical.
The high voltage, uni-directional, power supply for the
discharge electrodes is supplied by silicon transformer-
rectifier (T-R) sets. One electrical control unit is
provided for each T-R set which senses spark rate,
current, and voltage. The control unit maintains optimum
current and voltage conditions automatically and limits
the voltage and current to the T-R set rating. The
collecting plates in each precipitator are cleaned of fly
ash by means of 64 magnetic impulse, gravity impact
rappers. The discharge electrodes of each precipitator
are cleaned of fly ash by means of 32 magnetic impulse,
gravity impact rappers. The precipitator control room is
situated on the roof of each precipitator. Control panels
for the existing precipitators occupy the entire control
room, and no space is available for additional controls
for additional precipitators. Fly ash from the
precipitator hoppers are pneumatically transported to a
dry ash storage bin. Flue-gas exit ducts from two
precipitators and merge into a single duct that connects
to the stack. The corresponding duct from the other
boiler is connected to the other side of the stack.
B-2.5 EXISTING S02 CONTROL DESCRIPTION
The S02 emission control is the use of low sulfur
subbituminous coal as fuel for the steam generators
&-J7
-------
SECTION B-3
PLANT SURVEY FORM
B-3.1 COMPANY AND PLANT INFORMATION (1)
1,
2,
3-
4,
5,
6,
7.
8,
9
10
Company Name:
Main Office:
Southern California Edison Company
2244 Walnut .Grove Avenue, P.O. Box 800,
Rosemead, California 91720
Plant Manager: Mr. R.S. Currie, Mgr. of Steam Generation
Plant Name: Mohave Generating Station
Plant Location: Clark County, Laughlin, Nevada 89046
Person to Contact For Further Information: Mr. Lee Brothers
Position: Senior Engineer
Telephone Number: (213) 572-1630
Date Information Gathered: April 23 - April 25, 1979
Participants in Meeting Affiliation
L. Brothers •
N. Gonzalez
K. Hsiao
N. Master
R. Redman
SCE
Pullman Kellogg
Pullman Kellogg
Pullman Kellogg
Pullman Kellogg
B-3.2 PLANT DATA (APPLIES TO ALL BOILERS AT THE PLANT)
Boiler No.
Capacity, MW (net)
Service (Base,Peak)
FGD System Used?
790
Base
No
790
Base
No
B-18
-------
B-3.3 BOILER DATA
1. Maximum Continuous Heat Input;
II,
16,700 MM BTU/HR
8.35 x 109 BTU/HR
8.35 x 109 BTU/HR
1a. Maximum Heat Input: 16,700 MM BTU/HR
2. Maximum Continuous Generating Capacity (Gross) 820 MW
(Net) 790 MW ' .
3. Flue Gas Temperature: 270-300 (,@ STACK) °F
4. Maximum Continuous Flue Gas Rate 4.2 x 106_ ACFM @ 60°^.
I. 2.1 x 106 SCFM
II. 2-.1 x 106 SCFM
5.
6.
7.
8.
9.
9a.
9b.
10.
lOa.
lOb.
lOc.
10d.
11.
11a,
12.
13.
Not Available
YES
NO
X
Flue Gas Analysis:
Flue Gas Recirculation
For NOX Control:
Boiler Manufacturer: C.E. (I,II)
Years Boilers Placed In Service: 1971 (I), 1971 (II)
Boiler Service (Base Load, Peak, etc.): Base
Wet Bottom Dry Bottom X
Firing Type: PCTA
Stack Height Above Grade: 500 FT.
Stack Diameter (ft.): 33' ID at Outlet
Velocity Of Gas (Exit): 120 Ft/Sec at Full Load
Exit Gas Temperature: 300°F
Number of .Liners/Boiler: One (One stack for two boilers)
Boiler Operations: Hours/Year (1977): (I) 6194, (II) 6872,
Boiler Operations: Hours/Year @ Full Load: Not Available
Boiler Capacity Factora:(I) 56.1$,(II) 66.1%
Boiler Operating Pressure: 3500 PSIG @ Turbine Inlet
aDefined as:
KWH GENERATED IN YEAR
(Net) Max Cont. Generating Capacity in KW x 8760 HR/YR
B-19
-------
14. Boiler Superheat Heat Temperature; 1000
I4a. Boiler Reheat Temperature: 1000
I4b. Economizer .B.F.W. Outlet Temperature: N/A °F
I4c. Superheater AP = 200 psi
15. Ratio of Fly Ash/Bottom Ash = 70/30
16. Burners:
Type: Tilting.Tangential
Manufacturer:
No. Per Unit:
Rating:
Coal:
C.E.
80
130.5
10,700
Primary Air: 24.4
Secondary Air;
Tertiary Air: NONE
Total Excess Air:
15
_MM BTU/HR
_#/HR/Burner
_% of Total
_% of Total
_% of Total
% of Total
17. Fans [F.D. & ID]:
Type: FD, PA
Manufacturer: American Standard (FD), Westinghouse (PA)
No. Per Unit:- 2-PA, 2-FD
Rating; 305,000 ACFM Each (PA), 125,000 ACFM (FD)
p: 32.5 (FD), 31.2 (PA)
HP: FD 7000, PA 1750
"H20
B-20
-------
18. Steam Temperature Control (Superheated & Reheated): By
Burner Tilt, Attemporator Controls Superheated & Reheat
Steam
Attemporator (Capacity): Not Available
19. Air Preheater:
Type: Lj ungstrom Vertical Shaft, Regenerative
No.: 2/Boiler
Flue Gas AT: 425°F ;
Air AT 553°F (Secondary Air)/595°F (Primary Air)
Calculated Flow Rate: 4890M Ibs/Hr per A.P. (Gas)
4360M Ibs/Hr Air A.P. (Air)
CalculatedAP 3.2" H20 (Air Side)
Flue Gas Inlet Temp.: 730 °F
B-3.4 FUEL DATA
1. Coal Analysis (as received) (%): MAX. I MIN. I AVG.
I I
—- I —— I 0-5
s
Ash
2. Total Ultimate Analysis (wtl)
As Received
I
Dry Basis
I
l.Q
As Mined
S:
Ash:
N:
Moisture :
C:
02:
H2:
HHV (BTU/LB):
0.19
4.85
0.53
57.59
28.44
6.29
2.11
0.44
11.43
1.24
12.70
67.07
14.85
4.97
12,200 12,200
B-21
-------
B-3.5 ATMOSPHERIC EMISSIONS
1. Applicable Emission Regulations Partioulates SO NO
a) State of Nevada
Current Requirements
Max. Allowable Emissions
Lb/MM Btu Input To Boiler
0.0675
0-. 6 NONE
2. Plant Program For Particulates: Cold Side ESP's
3. Plant Program For S02 Reduction: Use of Low Sulfur Coal
4. Plant Program For NOX Reduction: No Overfire Air. Use
of Tangential Burners
B-3.6 PARTICULATE REMOVAL
1. Type
Mech.
E.S.P.
FGD
Manufacturer...
Efficiency: Design/Actual...
Emission Rate Lb/Hr (Total)a
Gr/SCFM...
Lb/MM Btu...
Specific Collection Area (ft2/1000 ACFM)
(Design/Operating) :
Total Collection Area: 440,000 Ft2/Boiler
Design Basis, Sulfur Content
Of Fuel...
Research
Cottrel
97.9/98.6
850
0.24
0.05
191/143
0.5
aAt maximum continuous load
B-22
-------
2. Solids Collection System
Present Operating Condition:
At Maximum Capacity:
Bottom Ash Goes To Dewatering
Bins and Then Trucked Awaya
Closed Loop Bottom Ash13
YES X NO
Minimum Particulate Capturing Size:
Particulate Loading Into Collector:
Unburned Carbon %: N/A
1 Micron
GR/Cu-Ft flue gas
B-3.7 FRESH WATER MAKE-UP FLOW RATES AND POINTS OF ADDITION
Steam Slowdown Rate0:
None
Availability of Blowdown:
C.W. Blowdown Rate0: N/A
_GPM
None GPM
GPM
aAsh disposal area based on 35 years
bFlyash from ESP is pneumatically conveyed to storage silos,
wetted down in the unloader & trucked away
°Available water source for future S02 wet scrubber
B-23
-------
SECTION B-4
RETROFIT DESCRIPTION
B-4.1 NOX EMISSION CONTROL
The retrofit work investigated for NOX reduction
involves installation of overfire air (OPA) ports and the
associated tilt drive mechanisms, duct work, and control
air dampers for each branch ai-r duct to each OPA port.
Also, a new NOX control and monitoring instrument
system would be required for each boiler unit. Based on
discussions with Combustion Engineering, 20% excess air is
recommended for boiler operation, and 20% of this total
air is the design rate for the OPA, The air temperature
at the inlet to the firebox is 635°F, and the pressure is
estimated at 34.2" H20. The calculated required OPA
per boiler is 672,540 ACPM. Using,200 feet per second for
OPA Jet velocity, 16 square OPA ports, each 22 1/2" x 22
1/2", are required. Two OPA ports are installed together
at the top of each tangential burner set. The associated
work required includes redesign of pressure ports, cutting
the firebox wall, ductwork, and added air dampers. These
items are shown in Table B-l. The locations of OPA ports
are shown in Figure B-10.
B-4.2 PARTICULATE EMISSION CONTROL
The calculated number of modules for ESP and baghouse use,
for particulate control, and the data related to these
modules are shown in Table B-2.
B-24
-------
TABLE B-1.- RETROFIT FOR NOY REDUCTION
Units
Boiler MW (MW/Boiler)
Boiler Manufacturer
Burners: Type•
Arrangement
No. Burner-s/Vertical Column
No. of Burner Columns
Theoretical Air SCFM/Boiler
Excess- Air %
Air Temp. At'Preheater Outlet (°F)
&
820
C-E
Tangential
10
8
1,464,700
20
635
TYPE OF RETROFIT
1. Overfire Air (OFA)
% of OFA to Total Air
Total OFA ACFM
No. of OFA Ports
Size of OFA Port
Air ACFM per OFA Port
2. Curtain Air (C.A.)
% of C.A. to Total Air
Total C.A. ACFM
No. of C.A. Port
Size of C.A. Port
Air ACFM/C.A. Port
3. Low NOX Burners
4. Compartmented Windbox
Yes
20
672,540
16
H W
22 1/2" X 22 1/2"
42,034
N/A
H
H
it
n
N/A
N/A
B-25
-------
TABLE B-1.- RETROFIT FOR NOX REDUCTION (Continued)
Units
#1
&
ASSOCIATED MODIFICATION WORK
REQUIRED:
Cutting of OFA Port
Cutting of C.A. Port
Removing & Modification of
Membrane Wall Tubes/Boiler
(Approximate No. of Tubes)
Windbox Modification
Duct Connection (Addition)
OFA Tilt Drive Mechanism
Control To Each OFA & C.A. Port
Yes
No
120
No
Yes
Yes
Yes
B-26
-------
(U
o
•H
>
4J
w
^
0
CO
CO
o
o
K
l-l
s
w
) —
O CO
t< «
H
• 3
CC £3
H
H
EC
Kg
•^ pi
_
pj
^aj
H
S3
M tf
< O,
H
ffi pq
to
4J
O
01
-p
o n
O-H
o
axi
o
•H 0)
-p >
(0 (0
O.C
o o
J S
o
rH
I
B-27
-------
TABLE B-2.- RETROFIT DATA FOR ELECTROSTATIC PRECIPITATORS
AND BAGHOUSES
Power Plant:
Boiler No.
Gross MW/Boiler:
Flue Gas ACFM/Boiler:
300°F
170°F
125°F
68°F
Mohave Power Generating Station
#1 & #2
820
3,080,600
2,690,300a
2,486,300^
2,140,200°
Per Boiler
Total Req'd Particulate
Collection Area (FT2)
Existing Collection Area
(PT2)
Collection Area to be Added
(FT2)
No. of ESP Trains
No. of ESP Modules/Train
Cold Side
ESP
.3,080,600
440,000
2,640,600
, 8
10
Total No. of Modules to be Added 80
Total Area Added (F+2) • 2,592,000
Baghouse
Wet Dry
Scrubbing
1,540,300 1,345,150
1,540,300 1,345,150
84 74
1,551,732 1,367,002
ESP Electrical Sectionalizing
Total No. of Electrical Sections
Required Per 5 MW 164
Existing Electrical Sections 32
No. of Elec. Sec/New Module 2
Total No. of Elec. Sections
Added 160
Total 192
MW/Section 4.27
a Includes water from dry scrubbing
b Includes water from wet scrubbing
c 2610 SCFM/MW § 820 MW
B-28
-------
SECTION B-5
RETROFIT COSTS
Retrofit capital and annualized costs are included in this
section for the alternatives discussed in Section B-l. The total
required plant investment costs given for each alternative do not
include the costs of removing and/or relocating any existing
equipment that may be associated with the particular retrofit
alternative. The cost for land required for sludge disposal
(from S02 removal) and the associated sludge transportion to .
the disposal site are not included in the retrofit plant
investment or annualized operating costs. Also, the following
cost-items are realized but not included in the total retrofit
capital and annualized costs:
(1) Site preparation
(2) Cost of down-time
(3) Additional stack lining, if flue gas desulfurization is
installed
Working capital, the money required to operate the new equipment
associated with the retrofit, .has been calculated for each
retrofit alternative.
B-29
-------
TABLE B-3. CAPITAL INVESTMENT COSTS FOR
RETROFITTING THE MOHAVE POWER PLANT , ,
MILLIONS OF THIRD QUARTER 1979 DOLLARS (-a)
Alternative
1.
Cost Item
Control
2. Particulate
Control
3. S0_ Control
4. Emission
Monitoring
4.54
0
234.30
0.73
5. Auxiliary Boiler 13.05
6. Replacement of 60.04
Power Generating
Capacity
7. Working Capital 23.2-8
TOTAL
335.-94
2
4.54
96.18
234.30
0.73'
13.05
71.39
27.92
2a
4.54
96.18
191.1-1
0.73
0
58.05
22.85
3
4.54
--(b)
205.00^
0.73
0
28.98
17.32
373.46
(c)
256.57
(c)
Millions of Dollars 205
Per Kilowatt of
Gross Generating
Capacity
273
228
156
(a) Includes direct and indirect costs.
(b) Costs for particulate and S07 control are combined.
h
(c) See Section 4.5 for other costs not estimated.
B-30
-------
Cost Item
TABLE B-4. ANNUAL COSTS FOR RETROFITTING
THE MOHAVE POWER PLANT - MILLIONS
THIRD QUARTER 1979 DOLLARS PER YEAR
Alternative
1 .
2.
3.
4.
NOV Control
X
Particulate Control
S02 Control
Emission Monitoring
1
0
91
0
1
.0
08
.000
.8
.2
98
26
1
18
91
0
2
.008
.563
V
.898
.226
1
18
71
0
2a
.008
.563
.616
.226
1
68
0
3
.008
. _ _
.060
.226
(b)
(b)
TOTAL
:c). 91.413^ 69.,
Mills Per Kilowatt Hour 10.3
of Net Power Generation
(Current Net Less Retrofit
Power Requirements at
65 Percent of Maximum
Net Load)
12.5
10.1
7.5
(a) Includes fixed capital charges.
(t>) Costs of Particulate and S02 control are combined.
(c) See Section 4.5 for other costs not estimated.
B-31
-------
SECTION B-6
REFERENCES
Meeting notes - N. Gonzalez/N. Master, Pullman Kellogg -
meeting with L.E. Brothers, Southern California Edison,
Mohave Plant Site, 24 April 1979
Meeting notes - N. Gonzalez/N. Master, Pullman Kellogg -
meeting with L.E. Brothers, et al, Southern California
Edison and S. Cuffe/J. Copeland, EPA, 19 July 1979
Technical data for Mohave units 1 and 2 Electrostatic
Precipitators, SCE, internal document
Drawing received from SCE Bechtel DWG 74212-17, Plot
Plan, Rev 17, 3-19-79
B-32
-------
APPENDIX C -
EXAMPLES FOR RETROFITTING THE NAVAJO POWER STATION
-------
CONTENTS
SECTION
PAGE
CONTENTS
FIGURES
TABLES
C-ii
C-iii
C-iv
C-1.0 GENERAL
C-l.l Retrofit Alternatives
C-1.2 Plant Characteristics
C-1.3 Equipment Location Changes for
Retrofitting
C-l. 4 Flue Gas Ducting Requirements
C-2.0 BACKGROUND DATA
C-2.1 Plant Description
C-2.2 Steam Generator Description
C-2.3 Existing NOX Control
C-2.4 Existing Particulate Control
C-2.5 Existing S02 Control
C-3.0 PLANT SURVEY FORM
C-3.1 Company and Plant Information
C-3.2 Plant Data
C-3.3 Boiler Data
C-3.4 Fuel Data
C-3.5 Atmospheric Emissions
C-3.6 Particulate Removal
C-3.7 Fresh Water Make-Up Flow Rates and
Points of Addition
C-l
C-l
C-4
C-6
C-7
C-12
C-12
C-14
C-14
C-16
C-17
C-18
C-18
C-18
Q-19
C-21
C-21
C-22
C-22
C-4.0 RETROFIT DESCRIPTION
. C-4.1 NOX Emission Control
C-4.2 Particulate Emission Control
C-23
C-23
C-23
C-SO RETROFIT COSTS
C-6.0 REFERENCES
C-25
C-28
C-ii
-------
FIGURES
FIGURE
C-l Addition of wet S02 scrubbing modules.
C-2 Addition of ESP's and S02 scrubbing modules.
C-3 Addition of dry S02 scrubbing modules with
baghouses.
C-4 Arrangement of wet S02 scrubbing modules.
C-5 Arrangement of wet S02 scrubbing and ESP modules,
C-6 Arrangement of dry S02 and baghouse modules.
C-7 General plot plan of the Navajo power station.
C-8 Schematic of air, flue gas, and coal conveying
for the twin furnace Navajo boiler.
PAGE
C-2
C-3
C-5
C-8
C-9
C-11
C-13
C-15
-------
TABLES
Table Page
C-l Retrofit Data for Electrostatic Precipitators and
Baghouses - Units 1, 2, and 3 C-24
C-2 S02 Scrubber Modules for Navajo Power Plant C-25
C-3 Navajo Plant, Raw Materials and Utilities
Requirements - Alternative 1 and 2, Wet Scrubbing C-27
C-4 Navajo Plant, Raw Materials and Utilities
Requirements - Alternative 2a, Wet Scrubbing at 70%
S02 Removal C-28
C-5 Navajo Plant, Semi-dry Scrubbing Raw Materials
and Utilities Requirements - Alternative 3, Semi-dry
Scrubbing at 70% SC>2 Removal C-29
C-6 Navajo Plant - Estimated Flue Gas Ductwork
Requirements C-30
C-7 Navajo Plant - Capital and Investment Costs for
Alternative 1 C-32
C-8 Navajo Plant - Annual Operating Cost for
Alternative 1 C-33
C-9 Navajo Plant - Capital and Investment Costs for
Alternative 2 C~34
C-10 Navajo Plant - Annual Operating Cost for
Alternative 2 C-35
C-ll Navajo Plant - Capital and Investment Costs for
Alternative 2a C-37
C-12 Navajo Plant - Annual Operating Cost for
Alternative 2a C-38
C-13 Navajo Plant - Capital and Investment Costs for
Alternative 3 c-4°
C-14 Navajo Plant - Annual Operating Cost for
Alternative 3 c~41.
C-15 Summary of Retrofit Capital Costs - Navajo
Power Plant C-43
C-l6 Summary of Retrofit Annual Costs - Navajo
Power Plant C-44
C-iv
-------
SECTION C-l
GENERAL
C-l.l RETROFIT ALTERNATIVES
• Four alternative examples for retrofitting the Navajo
Power Plant were considered in this appendix to
demonstrate the use of the methods developed in Section 3
of the report. Alternatives for the Units 1,2, and 3
boilers follows. Since the =Navajo steam generators' are
designed for maximum NOX control, no NOX
retrofitting is necessary. All alternatives include the
installation of emission monitoring systems for opacity,
S02, and NOX.
Alternative 1 - Wet S02 scrubbing for 90% S02
removal is added to existing ESP's to provide 21 ng/J heat
input (0.05 lb/106 Btu) particulate levels. Figure
C-l shows the plant general arrangement with the addition
of the S02 scrubbing modules.
Alternative 2 - Alternative 2 upgrades the existing ESP
collection area by adding of high-efficiency cold-side
ESP's. Wet S02 scrubbing is also added. The resultant
system provides 90% S02 removal and particulate
emissions limitation to a leel of 13 ng/J heat input (0.03
Ib/lO^ Btu). Figure C-2 shows the general arrangement
when the S02 scrubbing and cold side ESP modules are
included.
Alternative 2a - The retrofit for particulate control is
the same as for Alternative 2. The wet S02 scrubbing
is based on 70% S02 removal for cost comparison with
semi-dry scrubbing of Alternative 3.
-------
tfi
i
c
•H
O
0)
O
CQ
c
o
•H
-P
•H
u
C-2
-------
01
O
to
O
CO
t!
C
(0
0)
en
w
G
O
•H
-P
-H
I
U
0)
C-3
-------
Alternative 3 - Semi-dry S02 scrubbing (spray drying)
to 70% S02 removal is provided for this option. By
using baghouses as dry collectors with the existing ESP's.
particulate emissions are controlled to a level of 0.03
Ib/MM Btu. Figure C^3 shows the general arrangement plot
plan with the addition of the S02 scrubbing modules
and baghouses.
C-1.2 PLANT CHARACTERISTICS (I,2_,3_)
Characteristics of the plant site, existing equipment, and
space requirements for retrofit are presented in the
following list:
A. The Navajo Station is located on a 1021 acre tract of
land.
B. Major revamp work to install the equipment for S02
and particulate control requires relocation of existing
equipment.
C. The existing electrostatic precipitator is on the hot
side of the air preheater system.
D. The existing ash disposal site is located on 765 acres
about two miles east of the plant.
E. Each boiler has two primary-air (PA) fans, four
forced-draft (FD) fans, and four induced-draft (I.D.)
fans.
F. The number of S02 scrubbing modules used is based
on the total calculated flue gas rate from each
boiler.
G. One S02 scrubbing module per boiler is provided as
a spare.
H. One flue gas reheater is required for each wet S02
scrubbing module for Alternatives 1 and 2.
I. One flue gas booster fan is provided per scrubbing
module.
C-4
-------
(9i-l)2-9EOt d
W
0)
-p
•H
CO
-------
J. The individual scrubbing modules are provided with
dampers. This provision allows the individual modules
to be isolated for maintenance.
K. Tie-in of retrofit equipment to the power plant is
based on completion during normal power plant
maintenance turn-arounds of 3 to 6 weeks.
L. An emergency bypass is provided around each S02
scrubbing system to allow operation of the boiler in
the event of. a major FGD malfunction. Bypassing of the
particulate control equipment is not provided.
C-1.3 EQUIPMENT LOCATION CHANGES FOR RETROFITTING
Major revamp work to install the equipment for S02 and
particulate control requires relocation of some existing
buildings and/or systems. The requirements for the
alternatives being considered are:
Alternative 1
1. Requires the relocation of the warehouse to a point
some 300 yards south of the present location.
Alternatives 2 and 2a
1. Requires relocation of warehouse.
2. Requires relocation of ash settling and dewatering
tanks, fly ash bins and associated auxiliary equipment now
located in the ash unloading area. This would require
extensive grade work to facilitate such a move. The area
suggested would require an estimated 2,000,000 cubic yards
of earth removal which is mostly rock". This would be in
-------
addition to extensive modifications to the ash handling
systems. It may be possible to totally redesign the ash
system and avoid the excavation. The cost for work of
this nature is not included in this study.
Alternative 3 -
1. Requires relocation of warehouse.
2. Requires relocation of ash unloading area, settling
tanks, dewatering bins, fly ash bins and associated
auxiliary equipment. This relocation has the uame
excavaton requirements as Alternative 2 and 2a.
No costs have been included for a sludge disposal area
acquisition. This would probably require negotiations
with the Navojo Tribe and could be very complicated and
costly.
C-1.4 FLUE GAS DUCTING REQUIREMENTS
Bypass ducting and dampers are provided to enable the flue
gas to go around the S02 scrubbing system. Bypass
ducting is located for the three alternatives as indicated
in the following paragraphs:
Alternative 1 - The bypass duct is taken from the plenum
located between the ID fan and the booster fan. After
bypassing the S02 scrubbing system, the bypass duct
joins the duct to the stack. Figure C-4 shows the general
arrangement of the scrubbing module and associated ducting
with tie-ins to the existing breeching.
Alternative 2 - The bypass duct is taken after the cold
side ESP's. Figure C-5 shows the general arrangement of
the precipitators and SC>2 scrubbing modules including
location of the emergency, S02 scrubbing system
bypass.
-------
CO
0)
H
1
tn
rt
•H
ft
•§
M
O
M
CN
O
CQ
•P
0)
M-l
O
-P
n
(U
PI
m
i
o
0)
C-8
-------
w
Q)
i
CO
(0
fl
•H
S-l
O
CO
O
CO
-p
I
-P
fi
1
Q)
cd
s-i
m
I
O
CD
S-l
C-9
-------
Alternative 2a - The bypass duct is arranged in the same
manner as for Alternative 2. This duct can also be used
during normal operation to divert about 22.2$ of the total
flue gas that does not require treatment, since the modules
remove 90% S02 and this alternate requires only 70%
S02 removal. The bypass duct also provides the flue gas
requirements for reheat.
Alternative 3 - The bypass duct is taken ahead of the spray
dryer and then ties into ducting to the baghouses. This
location of the emergency bypass permits operation of the
baghouses for particulate control when a spray dryer is out
of service. Figure C-6 shows the general arrangement of the
dry scrubbing module.
C-10
-------
CO
•d
i
0)
(d
CM
O
1
§
m
M
M
I
O
0)
C-ll
-------
SECTION C-2
BACKGROUND DATA
C-2.1 PLANT DESCRIPTION (J.,.2)
The Navajo station is jointly owned by several
governmental and private utilities. The station is
operated by Salt.River Project Agricultural Improvement
and Power District (SRP).
Total plant capacity is 2250 MW (net) and 2415 MW (gross).
The station consists of three 750 MW (net), 805 MW
(gross), pulverized-coal-fired, supercritical-steam
generating units manufacturered by Combustion Engineering
(C.E.). The units use balanced draft systems.
A typical subbituminous coal burned at Navajo has a sulfur
and ash weight percent of 0.5 and 10 to 12 respectively.
The coal is mined, by Peabody Coal Company, in the Black
Mesa area located in Northern Arizona. The coal feed rate
to each boiler is in the range of 300-335 TPH (wet basis).
The coal ash content varies with coal deliveries. At the
present time ash content rejection point is 16 wt$.
Each boiler has a 775-foot high, 25-foot I.D. stack. The
flue gas from each steam generator passes through a high
efficiency (+99.5$) Joy-Western hot side electrostatic
precipitator and is discharged to atmosphere from the
stack at a velocity of about 100 ft/sec (at full load),
and at a temperature of 300°F. Figure C-7 shows the
general plot plan for the Navajo power station.
C-12
-------
o
-H
-P
(6
-P
W
0
5
O
cu
O
X!
-P
nJ-
•H
cu
S-i
(U
C!
0)
O
I
O
-------
C-2.2 STEAM GENERATOR DESCRIPTION (2,3)
The three Combustion Engineering boilers are equipped
withr tilting tangential burners rated at 157.5 x 106
MM BTU/Hr/burner, and they are twin furnace designs.
There are 56 burners per boiler. Each burner fires coal
at a rate of 12,000 (maximum) and 11,428 (normal) Ibs/hr.
The total continuous heat input to the plant is 21,665.55
MM Btu/Hr with boilers I, II, III having 7,455.44,
6,789.21, and 7,420.9 MM BTU/Hr respectively. Each boiler
is equipped for overfiring. •
Each of the boiLers has two, primary-air (PA) fans, two
forced-draft (FD) fans, and four induced-draft (ID) fans.
The PA fans (Howden) are used for coal conveying, and they
utilize about 15% of the total air. The FD and ID fans
are manufacturered by Westinghouse. Each boiler has two
vertical shaft, regenerative Ljungstrom air preheaters.
The burner rating is 7.5 tons/hr. Each mill feeds coal to
eight burners, and there are seven mills per boiler. The
number of burners per boiler is 56. The coal feed rate
per boiler is 300 to 335 tons per hour. The three C-E
boilers are of dry bottom type with a Fly Ash/Bottom Ash
ratio of 80/20. The bottom ash slurry is dewatered and is
sent by truck to a disposal pond with 35 years of disposal
capacity. Figure C-8 shows a schematic of the air, flue
gas, and coal conveying provision for each boiler.
C-2.3 DESCRIPTION OF EXISTING NOX CONTROLS
Presently, there are 16 overfire air ports on each boiler.
The dimensions of the OFA ports is unknown. The OFA is
presently set at minimum for air-flow cooling of the
registers. Automatic control is available for opening the
dampers. The current operation involves hand-loaded,
c-14
-------
COAL PIPING
TO BURNER
TYP. EACH
'PULVERIZER
CORNER
WINDBOX
u
BURNER 4
s
NOTE:
7 BURNERS PER
CORNER FOR A
TOTAL OF 56
BURNERS PER
BOILER.
Figure C-8.-
Schematic of air, flue gas, and coal conveying for
the twin furnace Najavo boiler.
c-15
-------
Bailey positioners for damper control of the OFA. Manual
control is used for adjusting the tilt. One handwheel is
provided for each OFA register. Tests conducted by SRP
indicated no change in NOX emission level when going
from the minimum to the maximum openings of the OFA
ports.
C-2.4 EXISTING PARTICULATE CONTROLS (.2,3.)
Each boiler is equipped with hot-side precipitators
manufactured by Western Precipitators, a division of Joy
Manufacturers. There are 16 mechanical flow sections,
each containing 48 collection panels. There are 6
electrical sections for each mechanical section or a total
of 96 electrical sections for each boiler. This is
equivalent to 8.39 MW per electical section. The design
efficiency of these precipitators is 99.5/6. The specific
collection area is 307 (design)/270 (operating) square
feet per 1000 ACFM flue gas. The design basis for these
ESP's is based on coal with a 0.5% (wt) , sulfur content.
The particulate emission rate, at maximum continuous load
from these ESP's, is 468 Ibs per hour per boiler, or 0.02
Ibs per 10^ Btu .heat input (the designed emission
rate). Total collection area per boiler is 1,209,600
ft2. Controls for each precipitator includes the
following items:
o A-C supply volt meter
o A-C current meter
o Spark rate meter
o D-C volt meter
o D-C current meter
C-16
-------
There are 6 field monitors for each two chambers or a
total of 48 monitors per boiler. Each monitor can ground
out one chamber to take readings of one chamber and one
field. This allows a read out of any one of the 96 fields
per precipitator.
C-2.5 EXISTING S02 CONTROL DESCRIPTION
There are no current S02 emission controls except that
low sulfur coal is used as fuel for the steam generators.
C-17
-------
SECTION C-3
PLANT SURVEY FORM
C-3.1 COMPANY AND PLANT INFORMATION (1,2!,3.)
1. Company Name: Salt River Project
2. Main Office: P.O. Box 1980, Phoenix, Arizona 85001
3. Plant Manager: Mr. Harold Voeple
4. Plant Name: Navajo Generating Station
5. Plant Location: Page, Arizona 86040, Coconino County
6. Person To Contact For Further Information:
Mr. John McNamara
7. Position: Associate General Manager-Power
8. Telephone Number: (602) 273-2851
9. Date Information Gathered: April 17 - April 20, 1979
10. Participants in Meeting Affiliation
Richard F.- Durning
John R. McNamara
•Donald W. Moon
Richard H. Silverman
Gregory T. Whalen
Norman Master
Nora Gonzalez
Ronnie Redman
SRP
SRP
SRP
SRP
SRP
Pullman Kellogg
Pullman Kellogg
Pullman Kellogg
C-3.2 PLANT DATA (APPLIES TO ALL BOILERS AT THE PLANT)
BOILER NO.
CAPACITY, MW (NET)
SERVICE (BASE,PEAK)
FGD SYSTEM USED?
750
BASE
NO
750
BASE
NO
750
BASE
NO
C-18
-------
C-3.3 BOILER DATA
1. Maximum Continuous Heat Input: 21,665-55 MM Btu/Hr
I. 7,455.44 x 10°
II. 6,789.21 x 10°
III. 7,420.9 x 10°
la. Maximum Heat Input: 21,665.55
2,
Btu/Hr
3
4
Maximum Continuo.us Generating Capacity (Gross) 805 MW
(Net) 750 MW
Flue Gas Temperature: 300 (AT STACK) °F
5.
6,
7
8
Maximum Continuous Flue Gas Rate 9.34 x 10 AGFM § 300°F
BOILER I. 3.23 MMACFM % 307°F
BOILER II. 2.94 MMACFM @ 282°F
BOILER III. 3.17 MMACFM % 298°F
Flue Gas Analysis: Before Precipitator 0? (.>"- 3.5%)
Flue Gas Recirculation: Yes No ' X
For NOX Control
Boiler Manufacturer: C.E. (I,II,III)'
Years Boilers Placed in Service: Initial Firing 1974 (I),
1974 (II), 1975 (HI)
9. Boiler Service (Base Load, Peak, Etc.): Base Load
9a. Wet Bottom Dry Bottom_ X
9b. Firing Type
PCTA
10. Stack Height above Grade: 775 ft. Per Boiler
lOa. Stack Diameter (Ft.): 25' ID at Outlet
lOb. Velocity Of Gas (Exit): 100-110 Ft/Sec at Full Load
lOc. Exit Gas Temperature: 300°F '
lOd. Number Of Liners/Boiler: One (One stack per boiler)
11. Boiler Operations: Hours/Year (1978): (I) 7640, (II) 8077,
(III) 7295 ' -.
lla. Boiler Operations: Hours/Year § Full Load: (I) 7640,
(II) 8077, (III) 7295
12. Boiler Capacity Factored) 76.4%, (II) 82.1*, (III) 76.0%
13. Boiler Operating Pressure: 3800 PSIG @ Water Wall
Outlets3-
aBase Load - Operates @ 100% Capacity
bDefined as:
KWH GENERATED IN YEAR
(Net) Max Cont.Generating Capacity in KW x 8760 HR/YR
C-19
-------
14. Boiler Superheat Heat Temperature: 1005
I4a. Boiler Reheat Temperature: 1002
I4b. Economizer B.F.W. Outlet Temperature: 631
I4c. Superheater delta P = 200 psi
15
16
F 6 Full Load
Ratio of Fly Ash/Bottom Ash = 80/20
Burners c:
Type: Tilting Tangential
Manufacturer: Combustion Engineering (C.E.)
No. Per Unit: 56
Rating: 157.5 x 10° MM Btu/Hr
Coal: 15,000/11,428
Primary Air: 15
Secondary Air:
Tertiary Air: None
Total Excess Air: 20
17. Fans [F.D. & I.D.
#/Hr/Burner (Max./Normal)
7.5 T/HR/BURNER
% of Total
% of Total
% of Total
% of Total
Type: Square Cage (FD,PA,ID)
Manufacturer: Westinghouse (FD,ID), Howden (PA)
No. Per Unit: 2-PA, 4-FD, 4-ID
Ratine *
Regenerative Air Preheater Inlet = 10.80" H20
Regenerative Air Preheater Outlet = 5.50" H20
Windbox = 3.50" H20
Furnace = 0.0" H20
Draft At Regenerative Or Air Preheater Outlet = -14.90
H20
18. Windbox
Main: Equipped With Overfire Air
Branch:
Control:—
19. Steam Temperature Control
(Superheated & Reheated)
Attemporator (Capacity): 216M Ibs/hr for superheat
temperature control; 4$ MCR (Capacity) Feed Water To
Firing Rate Ratio
G 7 Mills/Boiler 56 Burners = 300-335 TPH (Operating)
8 Burners/Mill . . .
Each Mill 60 Tons/Hr (Design) aat Design Loads
d At design loads
C-20
-------
20. Air Preheater
Type: Vertical Shaft, Regenerative Lungstrom
No.: 2/Unit
Flue Gas T = 398°F
Air T 507°F (Secondary)/534°F (Primary)
Design ffi ow Rate: 6.875M Ibs/hr GAS 5,950M Ibs/hr
Design P 5.3" H?0 (Airside)
Flue Gas Inlet Temp. : 650 °F
AIR
C-3.4 FUEL DATA
1. Coal Analysis (as received)
S %
'Ash %
2. Total Ultimate Analysis (wt$)
MAX . | MIN.
AVG
0.5
22
8
o. . .
Ash...
N. ..
Moisture...
o...
02 • • •
ilp • . •
Chlorine...
HHV (Btu/Lb)...
0.50
10.43
1.00
10.27
61.29
12.13
4.37
0.01
10,725
C-3.5 ATMOSPHERIC EMISSIONS .
1. Applicable Emission Regulations P ARTICULATES SO
a) Current Requirements 0.1 1.0
Max. Allowable Emissions
Lb/MM Btu Input To Boiler
NOX
0.7
2. Plant Program for Particulates : Hot Side ESP's
3. Plant Program For S02 Reduction: Use of Low Sulfur* Coal
4. Plant Program For NOX Reduction: C.E. Boiler with
Tangential Firing and Overfire Capacity
-------
C-3.6 PARTICIPATE REMOVAL
1. Type
Manufacturer...
Efficiency: Design/Actual...
Emission Rate
MECH.
E.S.P.
Joy
Western8J1—
99.5/98.5
PGD
Average Lb/MM Btu... 0.06
Specific Collection Area (ft2/1000 ACFM)
(Design/Operating): 307/270
Total Collection Area: 1,209,600 SF
Design Basis, Sulfur Content
Of Fuel... 0.5 WT$ I
2. Solids Collection System:
Present Operating Condition:
SBottom Ash Slurry to Dewatering Bins and Then
Trucked Away To Disposal Area, Ply Ash Pneumatic
Conveyed To Hopper (Truck 75 tons)
At Maximum Capacity... YES X NO
'Minimum Particulate Capturing Size... 1 Micron
Unburned Carbon %• <0.03
C-3.7 FRESH WATER MAKE-UP PLOW RATES AND POINTS OP ADDITION:
None
GPMh
Steam Slowdown Rate
Availability of Slowdown None GPM
C.W. Slowdown Rate N/A GPMh
eAt Maximum Continuous Load
fESP-l6 Sections (Chambers)/Unit (48 Panels/Unit), 96 Total
Electrical Sections
SAsh Disposal Pond Based On 35 Years Life
^Available Water Source For Future S02 Wet Scrubber
-------
SECTION C-4
RETROFIT DESCRIPTION
C-4.1 NOY EMISSION CONTROL
A.
The existing OFA ports require added automatic, tilt-drive
mechanisms, tied to the burner tilt mechanism, for NOX
control. Twenty percent excess air is recommended with
the OPA set at a flow rate of 20% of the total air. New
NO,
monitoring systems will be installed for NO
x
emission monitoring for each boiler.
C-4.-2 PARTICULATE EMISSION CONTROL
The retrofit work to be done for particulate emission
control is shown in Table C-l.
C-23
-------
TABLE C-l.- RETROFIT DATA FOR ELECTROSTATIC PRECIPITATORS
AND BAGHOUSES - UNITS 1, 2, AND 3
Power Plant:
Boiler No.
Gross MW/Boiler:
Flue Gas ACFM/Boiler:
300°F
170°F
125°F
68°F
Navajo Generating Station
#1, 2 & 3
805
3,024,250
2,641,015
2,440,890
2,101,051
Baghouse
Per Boiler
Total Req'd Particulate
Collection Area (FT2)
Existing Collection Area
(FT2)
Collection Area to be Added
(FT2) - Hot Side
No. of ESP Trains
No. of ESP Modules/Train
Total No. of Modules to be
Added
Total Area Added - (F+2)
ESP Electrical Sectionalizing
Total No. of Electrical Sections
Required Per 5 MW
Existing Electrical Sections
No. of Elec. Sec/New Module
Total No. of Elec. Sections
Added
Total
MW/Section
Cold Side
ESP
3,024,250
1,209,600
1,814,650
8
7
56
1,814,400
Wet
Scrubbing;
1,512,125 1,
0
1,512,125 1,
82
1,514,786 1,
Dry
320,508
0
320,508
72
330,056
161
96
2
112
208
3.87
C-24
-------
SECTION C-5
RETROFIT COSTS
Retrofit capital and annualized costs are included in this
section for the alternatives discussed in Section C-l. The total
required plant investment costs given for each alternative do not
include the costs of removing and/or relocating any existing
equipment that may be associated with the particular retrofit
alternative. The cost for land required for sludge disposal
(from S02 removal) and the associated sludge transportion to
the disposal site are not included in the retrofit plant
investment or annualized operating costs. Also, the following
cost items are realized but not included in the total retrofit
capital and annualized costs:
(1) Site preparation .
(2) Cost of down-time
(3) Additional stack lining, if flue gas desulfurization is
installed
Working capital, the money required to operate the new equipment
associated with the retrofit, has been calculated for each
retrofit alternative.
C-25
-------
TABLE C-2. - CAPITAL INVESTMENT COSTS FOR RETROFITTING
THE NAVAJO POWER PLANT - MILLIONS OF THIRD QUARTER 1979
DOLLARSLaj
Cost Item
1
1. NOY Control 0.00
Jn,
2. Particulate Control 0.00
3. S02 Control 347.25
4. Emission Monitoring 1.04'
5. Auxiliary Boiler 19.21-
6. Replacement of Power
Generating Capacity 88.41
Working Capital
TOTAL
34.09
489.98
(c)
Alternative
2a
0.00 0.00
111.21 111.21
347.23 283.29
1.04
19.21
104.43
41.23
1.04
84.79
33.73
(c)
0.00
301.88
1.04
42.68
25..13
370.73
(c)
Dollars per Kilowatt of
Gross Generating
Capacity
203
259
213
154
(a) Includes direct and indirect costs
(b) Costs for particulate and SO- control are combined
(c) See Section 4.5 for other costs not estimated
C-26
-------
TABLE C-3. ANNUAL COSTS FOR RETROFITTING THE NAVAJO
POWER PLANT - MILLIONS OF THIRD QUARTER 1979
DOLLARS PER YEAR
Cost Item
Alternative
1..
2.
3.
4.
NOX Control
Particulate Control
S02 Control
Emission Monitoring
0
0
136
0
1
.000
.000
.064
.278
0
28
136
0
2
.000
.562
.064
.278
0
28
106
0
2a
.O'OO
.562
.077
.278
0.
-
100.
0.
3
000
(b)
222(b)
278
Total
136.342
164.904^134.917^
100.500
(c)
Mills per Kilowatt hour
of net generation
(current net less
retrofit power require-
ments at 65 percent of
maximum net load)
10.3
12.5
10.2
7.43
(a) Includes fixed capital charges
(b) Costs of particulate and S09 control are combined
£
(c) See Section 4.5 for other costs not estimated
C-27
-------
SECTION C-6
REFERENCES
Meeting notes - N. Gonzalez/N. Master, Pullman Kellogg -
meeting with R.H. Silverman, et al, Salt River Project,
Phoenix, Arizona, 16 April 1979
Meeting notes - N. Gonzalez/N. Master, Pullman Kellogg -
meeting with G. Whalen, et .al, Salt River Project, Page,
Arizona, 17/18/19 April 1979
Letters from J.R. McNamara, Salt River Project to N.
Master, Pullman Kellogg, 27 April, 17 May and 27 August,
1979
Meeting notes - N. Master/W. Talbert, Pullman Kellogg -
meeting with J.O..Rich, et al, Salt River Project and J.
Copeland, EPA, Phoenix, Arizona, 20 July 1979
Drawings received from SRP:
Bechtel DWG A-665-C125 Site General Arrangement Plan,
Rev 5, 7-8-74.
Bechtel DWG A-665-M421 General Arrangement Section, Rev
6, 3-29-76
C-2B
-------
APPENDIX D
ANALYSIS OF FGD SYSTEM EFFICIENCY BASED ON EXISTING UTILITY BOILER DATA,
PREPARED FOR EPA BY VECTOR RESEARCH, INCORPORATED
-------
VRI-EPA7.3-FR79-1
OAQPS-78-LVI-B-13
ANALYSIS OF FGD
SYSTEM EFFICIENCY BASED ON
EXISTING UTILITY BOILER DATA
R. FARRELL
T. DOYLE
N. ST.CLAIRE
NOVEMBER 1979
TECHNICAL REPORT
Prepared for
Offica of Air Quality Planning.and Standards'
Emission Standards and Engineering Division
Environmental Protection Agency
VECTOR RESEARCH, INCORPORATED
Ann Arbor, Michigan
-------
CONTENTS
Page
1.0 INTRODUCTION AND SUMMARY ...... 1-1
2.0 PREDICTED BEHAVIOR OF THIRTY-DAY AVERAGES OF EFFICIENCY ... 2-1
2.1 Scope Of Analysis ' 2-5
2.2 Analysis Results • 2-7
2.3 Methodology - 2-37
3'.0 DESCRIPTIVE STATISTICS ON FGD 'SYSTEM EFFICIENCY DATA .... 3-1
3.1 Data Set .' 3-1
3.2 Lognormal Transformation 3-2
3.2.1 The Untransformed Variable . 3-2
3.2.2 The Transformed Variable 3-5
3.3 Estimated Parameters and Comparability Among Units . . 3-7
3.3.1 Means and Standard Deviations 3-7
3.3.2 Autocorrelation 3-11
3.3.3 Autoregressive Model . 3-13
3.4 Possible Confounding Factors 3-13
4.0 COMPARISON WITH ENTROPY RESULTS . . 4-1
4.1 Predicted Exceedences 4-1
4.2 Process Structure . 4-2
' 4.3 Differences Among Sites 4-4
-------
-------
I) 1-1
1.0 INTRODUCTION AND SUMMARY
The Environmental Protection Agency (.EPA) promulgated new standards
of performance for electric utility steam generating units, on
June 11, 1979. In addition to restricting the levels of pollutants that
these units emit into the atmosphere, the standards require a 90 percent
reduction in potential 502 emissions if they exceed 0.60 Ib/million
BTUs of heat input. On August 10, 1979, a petition for reconsideration
of these standards was submitted to EPA by the Utility Air Regulatory
Group (UARG).1 Part of this petition requested that EPA reconsider the
90 percent removal requirement. This request was based on analyses per-
formed by- Entropy Environmentalists, Incorporated,.which were documented
•in Appendix. B of the UARG Petition entitled "A Statistical Evaluation of
the EPA FGD System Data Base Included in the Subpart DA MSPS Docket".
The analysis included a numerical simulation of 1,000 years of flue gas
desulfurization (FGD) efficiency to examine the impact of the 90 percent
efficiency standard promulgated by EPA. ,
Vector Research, Incorporated, (VRI) is under contract to EPA to
provide statistical and analytical support to the Agency on an as needed
basis. On November 1, 1979, VRI was tasked to simulate or otherwise
analytically describe FGD system efficiency to permit examination of the
questions raised by the Entropy findings. The primary purpose of the
task was to determine the levels of system efficiency and variability in
^-Petition for Reconsideration, Docket Number OAQPS-78-1.
-------
1-2
this efficiency that would be necessary to maintain at most one exceed-
ence per year for a thirty-day rolling average on a 90 percent efficiency
standard. The VRI simulation was to be based on analysis of data pro-
vided by EPA describing the efficiency of 11 flue gas desulfurization
units and to additionally describe results over a wide range of facility
parameters. The data analysis and simulation results were to be supplied
to EPA within two weeks of initiation of the task. The authors were
supported in this effort by Dr. Richard Cornell, a VRI associate, and
other VRI staff.
This report presents the results of VRI's analysis activities and is
organized into four chapters. This introductory chapter provides a
description of the task and a summary of major results. The second chap-
ter describes the results obtained concerning the behavior of various
thirty-day averages for parametrically described FGD systems. The range
V
of parameters used in generating these results was based in part on the
statistical analysis of the data. This analysis is discussed in chapter
three. The final chapter then discusses comparisons between VRI's
results and those reported by Entropy Environmentalists, Incorporated.
The major conclusions of this analysis were as follows:
(1) The use of thirty-day moving averages of efficiency results in
low-variability efficiency measurements at a facility, even
when the daily data shows much larger variability. This
results in averages which cluster much more closely around the
central value of. the efficiency measurements .than do the, daily
efficiencies.
(2) Existing facilities show 'significant correlations in the
efficiencies of sulfur removal on successive days. These
autocorrelations, as well as the median levels of efficiency
-------
J>l-3
and the fundamental variability of the process, influence the
closeness with which thirty-day averages will remain clustered
about their mean.
(3) The minimum long run average efficiency levels (described here
in terms of the geometric mean) at which a facility must be
operated in order that the ratio at which thirty-day rolling
averages occur below 90, 89„ 88, 87, 86, or 85 percent be held
to one per year are shown in exhibit 1-1 for facilities with
autocorrelations of 0.7 and various fundamental variability
levels, some of which clearly represent good engineering and
operating practice and some of which may not. Exhibit 1-2
shows similar data but for a failure rate of one failure per
ten years. As the exhibits show, the rate of occurence of
30-day rolling averages below 90 percent would be above one per
jf-
" " ' f
year for facilities wiht a 92 percent geometric mean efficiency
and daily variaility anywhere from 0.20 to 0.60. These
facilities would, however, have rates below one per year if the
*
threshold were 89 percent and the daily variability were no
greater than 0.26, or if the threshold were 88 percent and the
daily variability was no greater than 0.32, or if the threshold
were 87 percent and the daily variability was no greater than
0.38, or if the threshold were 86 percent and the daily
variability was no greater than 0.43, or if the threshold were
85 percent and the. daily variability was no greater than 0.48.
-------
EXHIBIT 1-1:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED
TO MAINTAIN NO MORE THAN ONE FAILURE PER YEAR
Daily
Std. Dev
(in
.
*
.
.
*
*
m
*
*
*
.
9
*
*
.
9
*
*
.
.
t
[
*
*
*
]
log)
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
'.58
.59
*
60
Std. Dev.
of 30-Day
Average1
(.0068)
(.0071)
(.0075)
• (.0079)
(.0082)
(.0086)
(.0090)
(.0093)
(.0097)
(.0101)
(.0105)
(.0109)
(.0112)
(.0116)
(.0120)
(.0124)
(.0128)
(.0133)
(.0137)
(.0141)
(.0145)
(.0150
(.0154)
(.0158)
(.0163)
(.0167)
(.0172)
(.0177)
(.0182)
(.0186)
(.0191) .
" (.0196)
(.0201)
(.0206)
(.0212)
(.0217)
(.0222)
' (.0228)
(.0233)
(.0239)
(.0245)
Minimum Efficiency
For Threshold Shown
<90%
-------
EXHIBIT 1-2:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER TEN YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
'.24
.25 .
126
.27-
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40 •
•- .41
.42
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
. 55
.56
. .57
.58
.59
.60
Std. Dev.
of 30-Day
Average
(.0068)
(.0071)
(.0075)
(.0079)
(.0082)
(.0086)
(.0090)
(.0093)
(.0097)
(.0101)
(.0105)
(.0109)
(.0112)
(.0116)
(.0120)
(.0124)
(.0128)
(-.0133)
(.0137)
(.0141)
(.0145)
(.0150)
(.0154)
(.0158)
(.0163)
(.0167)
(.0172)
(.0177)
(.0182)
(.0186)
(.0191)
(.0196)
(.0201)
(.0206)
(.0212)
(.0217)
(.0222)
(.0228)
(.0233)
(.0239)
(.0245)
Minimum Efficiency
For Threshold Shown
<90% <89%
<87% <86% <85%
92.6
92.7
92.8
92.9
93 0
«r O • \J
93.1
93.2
93.3
93.4
93.5
93.6
93. 7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.8
94.9
95.0
95.0
95.1
95.2
95.3
95.4
95.5
95.5
95.6
95.7
95.8
95.8
95.9
96.0
96.1
96.1
96.2
91.8
91.9
92.1
92.2
92 ^
92.4
92.5
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
•93.5 :
93.6
93. 8
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94. 7
94.8
94.9
95,0
95.1
95.2
95.3
95.3
95.4
95.5
95.6
95.7
95.8
95.8
91.1
91.2
91.3
.91.5
01 5
91.7
91.9
92.0
92.1
^2.2
92.4
92. 5
92,6
92.7
92.8
93.0
93.1
93.2
93.3
93.4
93.5
.93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94. 6
94.7
94.8
94.9
95.0
95.1
95.2
95.3
95.4
95.5
90,3
90.5
90.6
90.8
90.9
91.0
91.2
91.3
91.5
91,6
91.7
,91.9
92.0
92.1
•92.2
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
.94.7
94.8
94.9
95.0
95.1
89.6
89.7
89.9
90.1
90.2
90.4
90.5
90.7
90.8
90.9
91.1
91.2
91.4
.91.5
91.6'
91.8
91.9
92.0
92.. 2
92.3
92.4
.92.6
92.7
92.8
92.9
93 .-1
93. 2
93.3
93.4
93.5
93.6
93.7
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
88.8
89.0
89.2
89.3
89.5
89.7
89.8 .
90.0
90. 1
90.3
90.5
90.6
90.8
90.9
91.1
91.2
91,3
91.5
,-91. 6'
91.8
91.9
92.0
92.2
92.3
92.4
92.6
92.7
92.8
9?. 9
93.1
93.2
93.3
93.4
93.5
93.7
93.8
93.9
94.0
• 94.1.
94.2
• 94.3
Facility autocorrelation =0.7
*In computing the 30-day average variability, a geometric mean
emission level of 92% was assumed.
-------
1-6
The rates would be below one occurence per ten years for
combinations of thresholds and daily variabilities as follows;
Threshold Daily Variability
89% •
88%
87%-
86%
85%
no greater than 0.21
no greater than 0.27
no greater than 0.32
no greater than 0.37
no greater than 0.41
Data for auto correlations other than 0.7 can be found in the
body of the report.
(4) There is very little change in these estimates of minimum effi-
ciencies when the assumptions concerning the type of statisti-
cal distribution used to represent the efficiency data are
varied. Both normal and lognormal distributions provide rea-
sonable fits to the existing daily efficiency data, with the
lognormal probably slightly better than the normal. (Because
j
the lognormal distribution appears to fit the data better than
the normal, it has been used in generating exhibits 1-1 and
1-2, and in general throughout the analyses.). Both distribu-
*
tional assumptions produce very similar results in terms of the
predicted behavior of thirty-day averages taken on a rolling
basis.
These conclusions, as well as many other observations, are discussed in
more detail in the body of-this report.
-------
£2-1
2.0 PREDICTED BEHAVIOR OF THIRTY DAY AVERAGES OF EFFICIENCY
This chapter describes the main results of this analysis. The
principal question of interest is the behavior of thirty-day moving
averages of efficiency, and specifically the rate at which such averages
would dip below selected thresholds. The behavior of the moving or
rolling average was examined,for various true (average) efficiencies,
variabilities, and time dependencies.
In a setting where penalties could be imposed when such averages
fell below a regulatory threshold, EPA would expect to sft the threshold
level so that facilities designed, constructed, and operated in
accordance with good engineering practice would produce very infrequent
threshold crossings, while facilities not in accord with good engineering
practice woul'd show averages below the threshold on a more frequent
basis. That is, the threshold should correspond to some value
approximately at the minimum expected to be seen regularly from
well-engineered and operated facilities. This analysis is not designed^
to analyze what levels of performance correspond to good engineering
practice, but to show the relation between the operating characteristics
of a facility and the rates at which various threshold values of
thirty-day averages would be crossed. This information can then be
combined by EPA with expert knowledge of the achievable levels of
engineering and operating performance in designing regulatory policies.
Although the precise method of computing the thirty-day average
might vary somewhat, this analysis has assumed that a daily average
efficiency is generated each day from more frequent measurements of
emissions, and that these daily averages are then averaged for a period
-------
£2-2
of thirty days. Such thirty-day averages might be computed each day,
each week, each month, or at any other frequency, based on the thirty-day
period ending with the computation day.. The behavior of averages at
various computation frequencies will be discussed. We believe that this
general scheme contains most policies of interest. In the case of
possible changes in the precise methods of computing averages from hourly
or more frequent data, the analysis encompasses policies with essentially
the same effects as those which might be adopted. All the analyses have
assumed that data would be available for each day of operations.
In order to predict the behavior of the averages involved, assump-
tions must be made about several basic properties of the measurements of
scrubbing efficiency at a facility. These assumptions concern the long-
run level of scrubbing efficiency achieved, the type and amount of daily
variability which will be observed, and any temporal'patterns or Correla-
tions which might be expected in the observed efficiency.
Before presenting any numerical analyses of the issues, it is neces-
sary to define the various types of measurements which were used in
describing and analyzing the process. The level of scrubbing efficiency
achieved will be discussed in terms of several different related quanti-
ties. For some ourposes, it is necessary to consider the measured daily
efficiency: this quantity is produced by reducing more frequent measure-
ments of inlet and outlet sulfur concentrations to a daily efficiency
figure. These measurements may also be considreed in terms of the equiv-
alent measurements of emissivity, which is 1-efficiency, so that an
efficiency of 90 percent corresponds to an emissivity of 10 percent.
Daily efficiency or emissivity measurements (which were the basic
data used in the detailed data analyses of actual facilities, as
-------
D 2-3
described in chapter 3.0, and which also form a basis in terms of which
all these analyses were conducted) are observed to vary when measured
repeatedly at a single facility. This variation is stochastic or
probabilistic, rather than deterministic, in nature. .That is, the exact
measurement which will be obtained at some future time is not completely
determined from our knowledge of the process, but includes, elements of
randomness.
Describing the randomness in the daily measurements involves
describing the distribution of the daily measurements (that is, the
frequencies with which the measurement takes on various values) and the
interrelations among the daily measurements for different days. The
distribution of the daily measurements is tyoically described in-terms of
a measure of the center of the measurements observed (such as the
mean, the geometric mean, or the median) a measure of the variability of
the measurements about this center (such as the standard deviation or
geometric standard deviation), and the particular shape or type of .
distribution which descirbes the variability (such as the normal or
. ^ ^ -ILJ JIUU-T ^
logjTormal_ distribution). The interrelationships between measurements on
various days are typically measured in terms of the correlation between
measurements'on successive days.
The mean (sometimes called the arithmetic mean) of the measurements
is simpl-y the long-run average of the measurements. The geometric mean
is the value which would be obtained by taking the antilogarithm of the
mean of the logarithms of the measurements. The geometric mean of
measurements is always less than the arithmetic mean, no matter how the
measurements are distributed. The median of measurements is the value-
such that 50 percent of the measurements are above it and 50 percent
-------
P2-4
below. The standard deviation of measurements is the root-mean-square
average of the deviations of the measurements about their own mean. The
geometric standard deviation is the root-mean-square average of the
deviations of the logarithms of the measurements about the mean of the
logarithms. The correlation (or autocorrelation), of a sequence of
measurements varies between -1 and +1. . a correlation of +1 indicates
perfect correlation — that is, in our case, successive measurements at a
single facility would be identical. A correlation of 0 indicates no
dependence between successive measurements. Correlations below 0
indicate that high measurements are followed by low and low by high.
All of these terms may be applied to any sequence of measurements.
In the specific problem at hand, they may be applied_to daily efficiency
measurements, daily emissivity measurements, or thirty-day averages of
either. Generally, daily efficiencies are discussed in this analysis in
terms of the geometric mean emissivity (or the equivalent efficiency) and
the geometric standard deviation of emi'ssivity. This geometric standard
deviation may be thought of as a percentage variability in the measure-
*
ments so that a geometric standard deviation of 0.20 would indicate a
daily variation of about 20 percent of the daily mean. These scales of
measurement1 were chosen because they were those which had been used in
past studies of the same general topics. The thirty-day averages are
typically discussed in terms of the frequencies with which particular
levels of emissivity would be exceeded by the thirty-day averages or in
terms of their mean and standard deviation (arithmetic, not geometric).
-------
/) 2-5
2.1 SCOPE OF ANALYSES
In the specific problem at hand, the evidence supports the use of a
model in which observed dependencies in sequences of efficiency measure-
ments are viewed as produced by correlations between immediately succes-
sive days. The evidence on this point is discussed in the next chapter.
In such a model (an autoregressive model of lag one) the only correlation
parameter required to describe the pattern is the basic correlation
between the observations on successive days. All other dependencies are
then computable from this correlation coefficient. In terms of these
parameters, the region of the parameter space examined in this analysis
was:
(1) Long-run geometric mean emissivities of six percent to nine
percent, with particular attention to the value of eight
percent, corresponding to a 92 percent efficiency.^
(2) Daily geometric standard deviations of 0.20 to 0.50 and
distributions of measurements described by a probability
distribution of emissivities similar to the 1ognormal or normal
distribution, probably having more similarity to the 1ognormal
(see chapter 3.0). It must be remembered that these daily
\ ' m ii imiimi—i
variabilities in emissivity lead to much smaller variabilities
in the thirty-day-efficiency. For example, a typical facility
with daily emissivities of the order of nine percent with a
IAIthough the 92 percent figure is not the geometric mean efficiency
but the efficiency corresponding to the geometric mean of eraissivity,
we will,, when approoriate, refer to such values as geometric means
.without intending to nislead.
-------
2-6
50 percent variability would have daily efficiencies of 91
percent, with a daily error of 4.5 percent, and thirty-day
average efficiencies of about 91 percent with a variability of
only about one percent.
(3) Oay-to-day correlations between successive observations of 0.0
to 0.7.
The results of this analysis address three topics:
(1) The average number of times per year that thirty-day average
efficiencies, computed daily (360 times per "year"), would be
below various thresholds as a function of the facility operat-
. ing parameters assumed.
(2) The minimum long-run level of efficiency which' a facility would
have to maintain to limit its average threshold crossings on
the same rolling average to one per year, one per two years,
•r
one per five years, or one per ten years as a function of the
level of variability and correlation of daily observations at
the facility. These efficiencies are presented in terms of
4>
geometric means, keeping the method of description for all
daily data consistent. At these levels, the long-run rate of
excessive emissivity measured in terms of thirty-day rolling
averages, would be held to the one per year or other rata as
given. The actual number of excesses in a specific year would,
of course, vary, so that at a rate of one per year, some years
would have two, for example, and others zero.
(3) The potential effects of changing the frequency of computation
of the averages on the rate at which threshold crossing would
occur.
-------
£2-7
Following the presentation of these results, a very brief section
discusses the methods of computation used to generate the' estimates.
2.2 ANALYSIS RESULTS
The most basic and fundamental results of this analysis simply
describe the mean, standard deviation, and distribution of the thirty-day
averages as functions of the elementary process parameters describing the
level of efficiency, the variability of the daily observations, and the .
autocorrelation. Exhibit 2-1 shows the means and standard deviations of
the thirty-day rolling averages for a sampling of parameter values in the
region examined. Several observations can be made from that data. The
most basic is simply that the mean efficiency is different than the
efficiency level described by the geometric mean emissivitv. This
difference simply reflects the differences in meaning between the^mean
and the geometric mean. The difference would remain even if the data had
beem normally distributed: the geometric mean of a normally-distributed
datum is not identical to its mean, and the relation between the two
r
values in the parameter region of interest is almost precisely the
relation between the same parameters in the lognormal distribution..
A second observation is that the variabilities of the thirty-day
averages are much lower than the variabilities of the daily data. This
reduction in variability is the basic reason why taking averages of
sequences of observations is useful in obtaining consistent estimates of
actual performance levels. The third observation which can be made from
the exhibit is that both the mean and the standard deviation of the
thirty-day averages are clearly influenced by the variability and
autocorrelation in the efficiency process, as well as by the level of
efficiency.
-------
D 2-8
EXHIBIT 2-1: MEAN AND STANDARD DEVIATION OF 30-DAY AVERAGES
Process Parameters
Geom. Mean Geom. Std. Dev.
.9100
.9100
.910CF
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9100
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9200
.9300
.9300
.9300 '
.9300
.9300
.9300
.9300
.9300
.9300
.9300
.9300
.9300
.9300
.9300
.9300
.9300
. 9400
.9400
.9400
.9*00
.9400
.9400
.9400
.9400
.24CO
.9400
.9400
.9*00
.9*00
.9100
. 9400
.9100
.2000
.2000
.2000
.2000
.3000
.3000
.3000
.3000
.4000
.4000
.4000
.4000
.5000
.5000
.5000
.5000
.2000
.2000
.2000
.2000
.3000
.3000
.3000
. 3000
.4000
.4COO
. 4000
.4000
..5000
.5000
.5000
.5000
.2000
.2000
.2000
.2000
.3000
.3000
.3000
.3000
.4000
.4000
.4000
.4000
.5000
.5000
.5000
.5000
.2000
.2000
.2000
.2000
.3000
.3000
.3000
. 3000
.4000
.4000
.4000
.4000
.5000
.5000
.5000
.5000
Autocor.
0.0000
.3000
.5000
.7000
0.0000
.3000
.5000
.7000
0.0000
.3000 .
.5000
7000
0.0000
.3000
.5000
.7000
0.0000 •
.3000
.5000
.7000
0.0000
.3000
. 5000'
.7000
0.0000
,.3000
.5000
.7000
0.0000
.3000
.5000
.7000
0.0000
.3000
.5000
.7000
0.0000
.3000
-5000
.7000
0 0000
.3000
-5COO
.7000
0.0000
.3000
.5000
.7000
0.0000
.3000
.5000
.7000
0.0000
.3000
.5000
.7000
0.OCOO
.3000
.5000
.7000
o.ooco
.3CCO
.5000
.7000
Thirty-Day Average
Mean StcU De'v.
.9082
• 9032
.9082
.9082
.9059
.9059
.9059
.9059
.9025
.9025
.9025
.9025
.8980
'.8980
.8930
.8980
.9184
.9184
.9184
.9184
.9163
.9163
.9163
9163
.9133
.9133
.9133
.9133
.9093
.9093
.9093
.9093
.9286
.9286
.9286
.9286
-9263
.9268
.9258
.9268
.9242
.9242
.9242
.9242
9207.
.9207
.9207
.9207
.9338
.9388
.9338
.9388
.9372
.9372
.9372
.9372
9350
.9350
.9350
.9350
.9320
.9320 •
.9320
.9320
.0034
.0045
.0057
.0075
.0053
.0070
.0088 ,
.0118
.0074
.0098
.0122
.0163
.0099
.0130
.0162
.0215
.0030
.0040
.0051
.0063
.0047
.0063
.0073
.0105
.0066
.0087
.0109
.0145
.0088
.0116
.0144
.JU91
^0026
.0035
.0044
.0059
.0041
.0055
.0068
.0092.
.0058
.0076
.0095
.0127
.00,77
.0101
.0125
.0167
.0023
.0030
.0038
.0051
.0035
. 0047
.0059
.0073
.0049 •
.0065
.0082
.0109
.0066
.0087
.0108
.0143
-------
2>2-9
Additional analyses not easily presented in tabular form addressed
the shape of the distribution of the thirty-day rolling averages.
Questions had been raised about whether these averages would be distrib-
uted normally. The distribution was found to be very nearly, although
not exactly, normal. Although the averages were much more nearly normal
than the approximately lognormal daily measurements, all of the analyses
took account of the remaining non-normality; no results were based on
normal approximations.
The data in exhibit 2-1 was presented in terms of facility operating
parameters which were simply chosen to sample the region uf greatest
interest. The actual values of the basic process parameters are avail-
able for some experiments at specific facilities. Exhibit 2-2 shows the
parameters describing the processes at these facilities. The actual
statistical analysis of the data to produce these estimates of the
parameters is described in chapter 3.0. Exhibit 2-3 shows the means and
standard deviations of thirty-day average efficiency observations which
would be expected if a new facility with a 92 percent geometric mean
efficiency had the same operating conditions {process variability and
autocorrelation) as with each of the individual existing facilities.
As can "be seen in these exhibits, there is considerable variation .
among the results at the individual sites. There cannot be a strictly
statistical decision as the degree to which any. particular site repre-
sents good engineering and operating practices, state-of-the-art systems,
well-calibrated and maintained measuring equipment, and otherwise is
appropriate for use in extrapolations to future facilities. Any analyses,
of these issues must be made by engineers rather than statisticians.
Accordingly, the remaining analyses' of the behavior of the thirty-day
-------
D 2-10
EXHIBIT 2-2: PROCESS PARAMETERS OF ACTUAL FACILITIES
Unrt
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesville B
Lawrence
Geometric
Mean
• 84.4
83.3
80.8
85.4
97.0
89.2
88.5
96.0
86.0
92.5
95.4
Geometri c
Standard
Devi ati on
.295
.343
.234
.212
.359
.118
.182
.368
.447
.474
.835
Auto-
Correl ati on
.6955
.6949
.4683
-.1428
.2524
.6983*
' .5995
.8897
.7131
.6255
.6386
-------
z-n
EXHIBIT 2-3: THIRTY-DAY AVERAGE MEAM AND STANDARD
DEVIATION FOR 92%-EFFICIENT FACILITIES WITH
VARIABILITY AND AUTOCORRELATION OF ACTUAL
FACILITIES
Variability and
Autocorrelation
from:
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesville B
Lawrence
Mean
91.64%
91.52%
91.78%
91.82%
91.47%
91.94%
91.87%
91.44%
91.16%
91.05%
88.70%
Standard
Deviation
1.03%
1.22%
0.57%
0.32%
0.73%
0.39£4
0.52%
2,05%
1.66%
1.48%
3.70%
-------
2-12
average processes will continue to be presented, as was the initial
material in exhibit 2-1, in general parametric terms. The appropriate
cases from these parametric results may then be selected by engineers to
be used in any further .analyses.
In using the parametric results, it may be appropriate to examine
the expected behavior of processes with one or more parameters equal to
those, of specific existing facilities (as was done in generating exhibit
2-3), or to consider the fact that the measurements from existing
facilities are from finite, and generally fairly limited, data samples,
and to consider the possible errors in estimation which may be present.
When this second technique is used, it may be of interest to know that
the Shawnee TCA and Pittsburgh II (taken together, assuming that their
true long-run levels of variability are identical as the data suggests)
have a 95 percent confidence interval on the long-run geometric standard
deviation running from 0.16 to 0.23, and that Lousiville North and South
taken together have a 95 percent confidence interval from 0.29 to 0.36.
(The corresponding 99 percent intervals are from 0.15 to 0.25 for Shawnee
•> .
TCA and Pittsburgh II and 0.28 to 0.38 for the Louisville facilities.)
Exhibit 2-4 shows the rate (in occurrences per 360-day year) at
which 30-day- averages of efficiency computed daily would fail to meet a
threshold level of 90 percent efficiency for a facility with, an actual
efficiency level of 92 percent^ and variability parameters as shown.
Each estimated rate is shown with an associated standard error of
estimate in parentheses. These estimates are for a facility with a
lognormal distribution of emissivity. Facilities with high values of
^-Corresponding to a geometric mean emissivity of eight percent.
-------
-D 2-13
EXHIBIT 2-4: FREQUENCY OF OCCURENCE (OCCASIONS PER YEAR)
OF BELOW - 90% AVERAGES IN A 92% EFFICIENT
FACILITY WITH LOGNORMAL OBSERVATIONS
PROCESS AUTOCORRELATION
0 - ' 0.3 ' 0.5
0.7
0.0
0.002 (.002) 0.189 (.031) 2.514 -(.095)
G
E
0
M
E
I
R
r
c
D
A
I
L
y
V
A
R
I
A
B
I
L
I
T
Y
0.320 (.0215) 2.670 (.0865) 9.900 (.332) 25.045 (.77(&)
10.233 (.180) 26,3935 (.186) 41.2375 (.3975) 62.4455 (.7366)
52.241 (.2655) 72.1565 (.3950) 87.608 (.5515) 102.496 1.9325)
Lognormal distribution.
Figures in parentheses are standard errors.
-------
J> 2-14
either variability (40 percent or greater) or day-to-day correlation (0.7
or greater) would be expected to fail to meet the threshold more than one
time per year, with facilities with high values of both variability and
correlation failing to meet the threshold for major fractions of their
operating days.
Exhibit 2-5 shows a comparison of these results with those which
would be expected on similar facilities where the variability of the
enrissivity was normal! rather than lognormal. As can be seen in the
exhibit, the pattern of dependency between' the plant operating parameters
and the rate at which the threshold is not met remains essentially the
same. That is, the rate of threshold failures does not depend in any
major way on the shape of the statistical distribution of the
observations (within the general area of reasonability).
Exhibit 2-6 shows the expected rate at which thirty-day averages
j
below thresholds other than 90 percent would occur for various
variability and correlation parameters. Exhibits 2-7 through 2-9 show
this same information for geometric mean emissivities other than eight
percent (corresponding to more or less efficient facilities). All of
these exhibits were derived using the lognormal distribution of emis-
sivity observations; rates of threshold failure for the normal case
differ by only small amounts, just as in the 92 percent-efficient cases.
Exhibits 2-10 through 2-13 show the efficiency levels (1.00 -
geometric mean emissivities) at which facilities with various variability
and correlation parameters would maintain a rate of threshold failure no
higher than one per year (with rolling averages computed daily). These
1Truncated at 0 efficiency.
-------
EXHIBIT 2-5:
D 2-15
FREQUENCY OF OCCURENCE (OCCASIONS PER YEAR) OF
BELOW-90% AVERAGES IN A 92% EFFICIENT FACILITY
WITH NORMAL OR LOGNORMAL OBSERVATIONS
PROCESS AUTOCORRELATION
0.3 0.5
0.7
Lognormal :
G
E
0
M
E
T
R
I.
f*
0
A
I
L
Y
V
A
R
I
A
B
I
L
I
T
Y
0.0 0.002 (.002)
•2 Normal:
0.0 0.009
Lognormal :
0.320 (.0215) 2.670 (.0865)
.3
Normal :
0.090 1.639
Lognormal :
A 10.233 (.180) 26.3935 (.186)
Normal :
7.742 22.777
I r\ n v» f\ v»m 3 T «
o.i ay i.u^ij £.011- -uusor
0.051 1.206
9.900 (.332) 25.045 (.7705)
-
6.. 678 2.1.403
41.2375 (.3975) 62.4455 (.7365)
.
39.689 64.527
52.241 (.2655) 72.1565 (.3950) 87.608 (.5515) 102.496 1.9325)
Normal:
52.061
75.449
92.764
112.50
Lognormal distribution cases above
normal cases.
Figures in parentheses are standard
errors.
-------
D 2-16
EXHIBIT 2-6: FREQUENCY OF OCCURENCE (OCCASIONS PER YEAR) OF
BELOW-THRESHOLD AVERAGES IN A 92% EFFICIENT FACILITY
(with standard errors in parentheses)
PROCESS AUTOCORRELATION
G
E
6
M
E
I
R
I
C
Q
A
I
L
Y
V
A
R
I
A
8
I
L
I
T
Y
30-day u
30-day a
eff<90*
" <39*
•2 „ <88%
11 <87%
11 <86%
" <85%
30-day u
30-day a
eff<90%
, " <89%
" <88»
" <87%
" <86%
" <85%
30-day u'
30-day cr
eff<90%
- " <89%
• • " <88S»
" <37%
11 <36%
'" '<85%
30-day u
30-day a
eff<90%
e " <39%
•° " <38%
11 <87%
11
-------
D 2-17
FXHIBIT 2-7- FREQUENCY OF OCCURENCE (OCCASIONS PER YEAR) OF
EXHibli /. jjELjjw_THRESH(JLD AVERAGES IN A 94% EFFICIENT FACILITY
(with standard errors in parentheses)
G
E
0
M
£
I
R
I
C
Q
A
I
L
Y
V
A
R
I
A
3
I
L
I
PROCESS AUTOCORRELATION
MMH
30-day u
30-day a
eff<90%
? " <89%
II -07W
II Sfitf
" <85%
'•
30-day u
30-day cr
eff<90%
3 II OOW
" <87%
" <86%
30-day u
30-day a
4 " <89%
11 <37?
11 <86%
II
-------
2-18
EXHIBIT 2-8:
G
E
0
M
E
I
R
I
C
0
A
I
L
Y
V
A
R
I
A
B
I
L
I
T
Y
FREQUENCY OF OCCURENCE (OCCASIONS PER YEAR) OF
BELOW-THRESHOLD AVERAGES IN A 93% EFFICIENT FACILITY
(with standard errors in parentheses)
PROCESS AUTOCORRELATION
30-day u
30-day a
eff<90%
,, " <89%
•* " <88%
11 <87%
11 <86%
" <85%
30-day u
30-day cr
eff<90%
M ^Q QC/
** ^QZ7 W
•3 " <88%
" <87%
" <86%
11 <85%
30-day u
30-day a
eff<90%
4 " <83%
M ^.ja
i, ,,052
11 <85S
30-day u
30-day cr
eff<90%
Q " <39%
" <88»
11 <&7%
" <85S
0
0.9286
.0026
o.'o
0.0
0.0
0.0
0.0
0.0
0.9268
• .0041
0.0
0.0
0.0
0.0
0.0
0.0
0.9242
.0058
0.0675 (.0155)
0.0
0.0
0.0
0.0
0.0
0.9207
.0077
2.799 (.175)
0.107 (.022)
0.0045" (.0045)
0.0
0.0
0.0
0.3
0.9286
.0035
0.0
0.0
0.0
0.0
0.0
0.0
0.9268
.0055
0.006 (.006)
0.0
0.0
0.0
0.0
0.0
0.9242
.0076
0.887 (.203)
0.0255 (.0075)
0.0
0.0
0.0
0.0
0.9207
.0101
10.9885 (.0935)
1.5165 (.097)
0.186 (.0195)
0.036 (.0155)
0.0045 (.0035)
0.0
" 0.5
0.9286
.0044
0.0
0.0
0.0
0.0
0.0
0.0
0.9268
.0068
0.1475 (.030)
O.OQ65 (.0065)
0.0
0.0
0.0
0.0 . , ,*
0.9242
.0095
4.3635 (.1895)
.0.450 (.045)
0.0395 (.0195)
0.015 (.0085)
0.0
0.0
0.9207
.0126
22.9385 (.635)
6.2575 (.0925)
1.4525 (.073)
0.299 (.039)
0.0585 (.0165)
0.0
' 0.7
0.9286
.0059
0.022 " (.0195)
0.0.
0.0
,0.0
0.0
•o.o
0.9268
.0092
2.0385 (.0485)
0.1465 (.042)
0.0015 (.0015)
0.0
0.0 .
0.0
0.9242
.0127
15.3025 .418)
4".031 .197)
' 0.8*42 .0795)
0.1755 (.035)
0.385 (.011)
0.0045 (.004)
0.9207
.0167
41.890 '"*(. 932)
18.276 (.611)
7.522 (.3035)
2.764 (.150)
1.036 (.0745)
0,385 (.0425)
Conditions:
Facility with 7 % geometric
mean enrissivity (93S efficiency)!
Lognarma! distribution of
observations
-------
£>2-19
EXHIBIT 2-9:
b
E
0
M
£
I
R
Q
A
I
L
Y
V
A
R
I
A
8
I
L
I
T
Y
FREQUENCY OF OCCURENCE (OCCASIONS PER YEAR) OF
BELOW-THRESHOLD AVERAGES IN A 91% EFFICIENT FACILITY
(with standard errors in parentheses)
30-day u
30-day cr
.4
11 <89%
" J " <83%
" <87%
" <35%
0.982
0.004
0.0
0.0
0.0
0.9082
.0034
3.696 (0.2160)
0.0
0.0
0.0
0.0
0.0
0.9059
. .0053
48.869 (0.604)
(0.1015)
(0.004)
30-day u 0.9025
30-day cr ^0074
eff<902 128.6885 (1.47)
<89%
<86%
18.807
1.028
0.0275
0.0005
0.0
(0.5955)
(0.0335)
(0.0155)
(0.0005)
.5.
30-day u 0.8980
30-day a '-0099
of-?<90% 201.053 (1.0585)
~" <39fo 72.932 (0.6025)
••
-------
EXHIBIT 2-10:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED
TO MAINTAIN NO MORE THAN ONE FAILURE PER YEAR
Minimum Efficiency
For Threshold Shown
<90% <89%
-------
2-21
EXHIBIT 2-11:
Daily
Std. Dev.
(in log.)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
\41
.42
.43
.44
.45
.46
.47 .
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Dev.
of 30-Day
Average1
(.0058)
(.0061)
(.0064)
(,0067)
(.0070)
(.0073)
(.0076)
(.0080)
( . 0083 )
(.0086)
(.0089)
(.0093)
(.0096)
(.0099)
(.0103)
(-0106)
(.0110)
(.0113)
( 0117)
(.0121)
(.0124)
( 0128)
(.0132)
(.0135)
(.0139)
(.0143)
(.0147)
(.0151)
(.0155)
(.0159)
" (.0164)
(.0168)
(.0172)
(.0177)
(.0181)
(.0186)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED
TO MAINTAIN NO MORE THAN ONE FAILURE PER YEAR
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87%
91.9
92.0
92.1
91
91
91
,1
,2
.3
92.2 91.4
,3
,3
.4
.5
92.
92.
92.
92.
92.6
92.7
92.8
92.9
93.0
93.0
93.1
93.2
93.3
93-4
93.5
93.6
93.6
93.7
93.8
93.9
94.0
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94.7
94.8
94.8
94.9
95.0
95.1
95,1
95.2
,5
,6
.7
91
91
91
91.8
91.9
92.0
92.1
92
92
92
,2
,3
,4
,5
7
92.4
92.5
92.6
92.7
92-8
92.9
93.0
1
2
,3
,4
.4
.5
.6
7
93.
93-
93.
93,
93.
93.
93.
93,
93.8
93.9
94.0
94.1
94.1
94.2
94.3
94.4
94.5
94.6
94.6
94.7
90.3
90.4
90.5
90.6
90.7
90.8
90.9
'91.0
91.1
91.
91,
91.
91.
91.
91 8
91.9
92.0
92.1
92-2
92,3
92.4
92.5
92.6
92.7
92.8
92.9
92-9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.2
89.5
89.6
89.7
89.8
89.9
90.0
90.2
90.3
90.4
90.5
90.6
90.7
90.8
91.0
91.1
91-2
91 3
91.4
91-5
91-6
91.7
91-8
91.9
92.0
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
88.6
88.8
88.9
89.0
89.2
89.3
89.4
89.5
89.7
89.8
89.9
90.0
90.1
90-3
90-4
90-5
90-6
90-7
90-9
91.0
91-1
91-2
3
,4
,5
.7
1
,2
,3
93.
93.
93.
93.4
93.5
93.6
93.7
93.3
91-
91.
91.
91,
91.8
91.9
92.0
.92.1
92.2
92.3
92.4
92.5
92.7
92.8
92.9
93.0
93.1
93.2
93.3
<85%
87.8
88.0
88.1
88.2
88.4
88.5
88.6
88.8
38.9
89.0
89. £
89.3
89.4
89.6
89-7
89.8
89-9
90.1
90.2
90.3
90.5
90-6
90.7
90.8
90.
91,
91,
91
91
91
91
9
1
2
,3
,4
,5
,7
91.8
91.9
92.0
92.
92.
92.
92.
92.6
92.7
92.8
Facility autocorrelation = 0.60
computing the 30-day average variability, a geometric mean emission
level of 92% was assumed.
-------
D 2-22
EXHIBIT 2-12:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED
TO MAINTAIN NO MORE THAN ONE FAILURE PER YEAR
Daily Std. Dev.
Std. Dev. of 30-Day
(in log) Average'
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.33
.39
.40
.41
.42
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54 i
.55 {
.56 i
.5? (
.53 I
.59 1
.60 I
(.0062)
(.0066)
(.0069)
(.0072)
(.0076)
(.0079)
(.0082)
(.0086)
(.0089)
(.0093)
(.0096)
(.0100)
(.0104)
(.0107)
(.0111)
(.0115)
(.0118)
(.0122)
(.0126)
(.0130)
(.0134)
(.0138-)
(.0142)
(.0146)
'.0150)
.0154)
.0159)
[.0163)
[.0167)
[.0172)
.0176)
.0181)
.0185)
'.0190)
[.0195)
.0200)
.0205)
.0210)
.0215)
.0220)
.0226)
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87% <86% <85%
92.0
92.1
92.2
92.3
92. "4
92.5
92 ..6
92.7
92.8
92.3
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.6
93.7
93.8
93.9
94.0
94.1
94.1
94.2
94.3
94.4
94.5
94.6
94.5
94.7
94.8
94.9
94.9
95.0
95.1
95.2
95.2
95.3
95.4
91.2
91.3
91.4
91.5
91.6
91.7
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94.7
94.8
94.9
94.9
90.4
90.5
90.6
90.8
90.9
91.0
91.
91.
91.
91.
91.
91.6
91.7
91.8
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92k 9
93.0
,1
,2
,3
93,
93.
93.
93.4
93.5
93.6
93.7
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
89.6
89.7
89.9
90.0
90.1
90.2
90.3
90.5
90.6
90.7
90.8
90.9
91.
91.
91.
91.4
,5
.6
,7
91.
91.
91.
91.8
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
88.8
88.9
89.1
89.2
89.3
89.5
89.6
89.7
89.9
90.0
90.1
90.2
90.4
90.5
90.6
90.7
90.9
91.0
91.1
91.2
.3
,5
91
91
91.6
91.7
91.8
91.9
92.0
,1
,3
92.
92.
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.5
93.6
88.0
88.2
88.3
88.4
88.6
88.7
88.9
89.0
89.1
89.3
89.4
89.5
89.7
89.8
89.9
'90.1
90.2
90.3
90.5
90.
90.
90.8
91.0
91.
91.
91.
91.
91,
91.
91.8
91.9
92.
92.
92.
92.4
92.5
92.6
92.8
92.9
93.0
93.1
,6
7
,1
,2
,3
,5
,6
,7
.1
,2
.3
Facility autocorrelation =0.65
In computing the 30-day average variability, a geometric mean emission
level of 9255 was assumed.
-------
D 2-23
EXHIBIT 2-13:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED
TO MAINTAIN NO MORE THAN ONE FAILURE PER YEAR
Daily
Std. Dev.
(in log)
.20 '
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
,38
.39
.40
...41
.42
.43
.44
.45 .
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
. 38
.59
.60
Std. Dev.
of 30-Day
Average'
• (.0068)
(.0071)
(.0075)
(.0079)
(.0082)
(.0086)
(.0090)
(.0093)
(.0097)
(.0101)
(.0105)
(.0109)
(.0112)
(,0116)
(.0120)
(.0124)
(.0128)
(.0133)
(.0137)
(.0141)
(.0145)
(.0150)
(.0154)
(.0158)
(.0163)
(.0167)
(.0172)
(.0177)
(.0182)
(.0186)
- (.0191)
(.0196)
(.0201)
(.0206)
(.0212)
(.0217)
(.0222^
(.0228)
(.0233)
(.0239)
(.0245)
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87% <86%
92.
92.
92.
92.
92,
92.
92.
92.8
92.9
93.0
93.
93.
93.
93.4
93.5
93.6
93.7
93.8
93.8
93.9
94.0
94.1
94.2
94.
94.
94.4
,5
,6
,7
94.
94.
94.
94.8
94.9
94.9
95.0
95.1
95.2
95.2
95.3
95.4
95.5
95.5
95.6
91.4
91.5
91.6
91.7
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.2
94.3
94.4
94.5
94.6
94.7
94.8
94.9
94.9
95.0
95.1
95.2
90.6
90.7
90.8
90.9
91.1
91.2
91.3
91.4
91.5
91.6
91.7
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93..S
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.7
89.8
89.9
90.1
90.
90,
90.
90.6
90.7
90.8
90
91
.9.
.1
91.2
,3
,4
,5
91
91,
91
91.6
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.6
92-7
92.8
92 9
93.0
93.1
93.2
93.
93.
93,
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
,3
,4
,5
89.0
89.2
89.3
89.4
89.6
89.7
89.8
90.0
90.1
90.2
90.4
90.5
90.6
90.8
90.9
91.0
91.1
•91.3
91.4
91.5
91.6
91.7
91 9
92.0
92.1
92.2
92.3
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93
93
93.4
93.6
93.7
93.8
93.9
2
,3
88.2
88.4
88.5
88.7
88.8
89,0
89.1
89.3
89.4
89,5
89.7
89.8
90.0
90.1
90.2
90.4
90.5
90.6
90.8
90.9
91.0
,2
.3
91.
91.
91.4
91 5
91.7
91.8
91.9
92.0
92.
92.
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93.2
93.3
93.4
Facility autocorrelation =0.70
'in computing the 30-day average variability, a geometric mean emission
level of 92" was assumed.
-------
2-24
minimum efficiency critical values are accurate to within at least 0.2
percent (two tenths of one percent). Exhibits 2-14 througn 2-25 show
similar data for threshold failure rates of one per two years, one per
five years, and one per ten years. (Given the randomness of the process,
there is no set of operating conditions that can achieve a true zero rate
of failure; some failures will occur randomly under any conditions.)
Policies in which averages are computed less frequently than daily,
but are still thirty-day averages for the last thirty-days at the time of
computation (for example, averages computed weekly or monthly) would, of
course, result in fewer threshold failures per year for all facilities,
whether or not operated in accordance with good practice, simply because
there would be fewer occasions per year on which failures could occur.
The effect on the rate of failures per year is, in fact, exactly
proportional to the frequency of computation of the average.* Thus, if
•r
weekly averaging were used, in which a thirty-day average was computed
for the thirty-day period ending, for example, on each Friday, the rate
of threshold failures per year for any set'of operating parameters would
*
simply be one-seventh of that shown in the preceding exhibits. If
averages are computed once every thirty days, the rate of failures per
year would be one-thirtieth of that in the exhibits, etc. The exhibited
critical operating levels at which one failure per year would occur, of
course, no longer apply if the frequency of average computation is
changed.
2-This fact can be proven completely mathematically for all the pro-
cesses considered here, whether involving the normal, lognormal, or
other distribution. Somewhat in violation of intuition, the proposition
remains true no matter what the correlation structure of the daily
observations.
-------
EXHIBIT. 2-U:
P2-25
MIMIMUM GEOMETRIC MEAM EFFICIENCIES REQUIRED TO
MAINTAIN MO MORE THAN ONE FAILURE PER TWO YEARS
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87% <86%
(.0054)
(.0057)
(.0060)
(.0063)
(.0065)
(.0068) .
( 0071)
i • \J \J f «t j
(.0074)
(.0077)
(.0080)
(.0083)
(.0087)
(.0090)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0109)'
• (.0113)
(.0116)
(.0119)
(.0123)
(.0127)
^ 9 \S J. fc- - f
(.0130)
(.0134)
(.0138)
(.0141)
\ " w •*• . /
(.0145)
(.0149)
\ * v -fe * •* /
(.0153)
(.0157)
^ • « *• •** • /
(.0161)
\ * w •• •* /
(.0165)
(.0169)
(.0174)
\ * ** •** * . /
(.0178)
(.0182)
\ » v •*• ^* — /
(.0187)
(.0192)
(.0196)
91.9 91.1 90.3 89.5 88.7 87.9
92.0 91.2 90.4 89.6 88.8 88.0
92.1 91.3 90.5 89.7 88.9 38.1
92 2 g1^ 90.6 89.3 89.0 88.3
92.3 91.5 90.7 89.9 89.2 88.4
92.4 91.6 90.8 90.1 89.3 88.5
92.4 91.7 9C-9 90.2 89.4 88.7
92,5 91.8 91.0 90-3 89-5 88.8
92.6 91.9 91.1 90.4 89.7 88.9
92,7 92.0 91 2 90.5 89.8 89,1
92.8' 92.1 91.4 90.6. 89.9 89.2
92.9 92.2 91 5 90." 90.0 89.3
93.0 92.3 91.6 90.9 90.2 89.5
93.1 92.4 91.7 91.0 90.3 89.6
93.1 92.5 91.8 91.1 90.4 89.7
93.2 92.5 91.9 91.2 90.5 89.3
93.3 92.6 92.0-91.3 90.6 90.0
93.4 92.7 92.1 91.4 90.3 90.1
93.5 92,8 92.2 91.5 90.9 90.2
93.6 92.9 92.3 91.6 91.0 90.3
93.6 93.0 92.4 91.7 91.1- 90.5
93.7 93.1 92,5 91.8 91.2 90..6
93.8 93.2 92-6 92.0 91.3 90.7
93.9 93-3 92.7 92.1 91.4 90.8
94.0 93.4 92.8 92.2 91.6 91.0
94.1 93.5 92.9 92.3 91.7 91.1
94.1 93.5 93.0 92.4 91.3 91.2
94.2 93.6 93.1 92.5 91.9 91.3
•94.3 93.7 93.2 92.6 92.0 91.4
94.4 93.8 93.3 92.7 92.1 91.6
94.5 93.9 93.3 92.8 92.2 91.7
94.5 94.0 93.4 92.9 92.3 91.8
. 94.6 94.1 ' 93.5 93.0 92.5 91.9
94.7 94.2 93.6 93.1 92.5 92.0
94.8 94.2 93.7 93.2 92.7 92.. 1
94.8 94.3 93.8 93.3 92.8 92.3
Q4.9 94.4 93.9 93.4 92.9 92.4
95.0 94.5 94.0 93.5 93.0 92.5
Q5 1 94- 6 94.1 93.6 93.1 92.6
95.1 94.7 94.2 93,7 93..2 92.7
Q5.2 94.7 94.3 93.8 93.3 92.8
Facility autocorrelation = 0.5.5
computing the 30-day average -variability, a geometric mean
emission level of 92% was assumed.
-------
EXHIBIT 2-15:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED' TO
MAINTAIN NO MORE THAN ONE FAILURE PER TWO YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
• .39
.40
.41
.42
.43 '
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Dev.
of 30 -Day
Averaged
(.0058)
(.0061)
(.0064)
(.0067)
(.0070)
(.0073)
(.0076)
(.0080)
(.0083)
(.0086)
(.0089)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0110)
(.0113)
(.0117)
(.0121)
(.0124)
(.0123)
(.Q13Z)
(.0133)
(.0139)
(.0143)
(.0147)
(.0151)
(.0155)
(.0159)
(.0164)
(.0168)
(.0172)
(.0177)
(.0181)
(.0186)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
<90%
Minimum Efficiency
For Threshold Shown
<88% <87%
-------
EXHIBIT 2-16:
D 2-27
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN MO MORE THAN ONE FAILURE PER TWO YEARS
Daily
Std.- Dev.
(in loq)
*
•<
•
*
•
*
••
•
•
•
•
•
20
21
22
23
24
25
26
27
28
29
30
31
32
33
..34
9-
«
a
»
*
«
•
*
*
•
*
•
35
36
37
38
39
40
41
42
43
44
45
46
47
.48
•
•
49
50
.51
«
•
52
,53
,54
4
4
t
i
1
t
,55
.56
.57
,58
.59
.60
Std. Dev.
of 30-Day
Average1
(,
(,
.0062)
.0066)
(.0069)
(.0072)
(.0076)
(.0079)
(.0082)
(
(
(
(
(
(
f
(
(
I
(
(
(
(
(
(
(
(
f
f
(
{
(
(
(
(
(
(
(
(
(
.(
.0086)
.0089)
.0093)
.0096)
.0100)
.0104)
.0107)
.0111)
.0115)
.0118)
.0122)
.0126)
.0130)
.0134)
.0138)
.0142)
.0146)
.0150)
.0154)
.0159)
.0163)
.0167)
.0172)
.0176)
.0181)
.0185)
.0190)
.0195)
.0200)
.0205)
.0210)
.0215)
.0220)
.0226)
Minimum Efficiency
For Threshold Shown
<90% <89%
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
94.
94.
94.
94.
94.
94.
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
6
7
8
9
0
1
2
3
3
4
5
6
7
7
94.8
94.
95.
.9
.0
95.1
95.
95.
95,
95.
95.
95.
95.
,1
,2
.3
,4
,4'
. 5
.5
91.3
91.5
91.6
91.7
91,8
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92,8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.0
94,1
94.2
94.3
94.4
94.5
94.6
94.7
94.7
94.8
94.9
95.0
95.1
95.2
<88% '
90.
90.
90.
90.
91.
91.
91.
91.
91.
91.
91.
91.
91.
92.
92,
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93,.
93.
93.
93.
93.
93.
6
7
8
9
0
2
3
4
5
6
7
8
9
0
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
94.0
94.
94.
94.
94.
94.
1
,2
,3
,4
,4
94.5
94.6-
94,
,7
<87% <86%
89.8
89.9
90.0
90.2
90.3
90.4
90.5
90.7
90.8
90.9
91.0
91.1
91.3
91.4
91.5
91.6
"91.7
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
89.0
89.1
89.3
89.4
89.5
89.7
89.8
89.9
90.1
90.2
90.3
90.5
90.6
90.7
90.9
91.0
91. 1'
91.2
91.3
91.5
91.6
91.7
91.8
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
<35%
88.2
88.4
88.5
88.. 6
88.8
88.9
89.1
89.2
89.4
89.5
89.6
89.8
89.9
90.1
90 ..2
90.3
90.5
90.6
90.7
90.9
•'91.0
91.1
91.2
91.4
91.5
91.6
91.8
91.9
92.0
92.1
92.2
92.4
92.5
92.6
92.7
92.8
92.9
93.1
93.2
93.3
93.4
Facility autocorrelation = 0.65
ln computing the 30-day average variability, a geometric mean
emission level of 92^ was assumed.
-------
^2-28
EXHIBIT 2-17:
MINIMUM GEOMETRIC MEAM EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER TWO YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.23
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
' ..41
.42 •
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Dev.
.of 30-Day
Average^
(.0063)
(.0071)
(.0075)
(.0079)
(.0082)
(.0086)
(.0090)
(.0093)
(.0097)
(.0101)
(.0105)
(.0109)
(.0112)
(.0116)
(.0120)
(.0124)
( 0128)
( 0133)
(,0137)
(,0141)
(.0145)
(.0150)
(.0154)
(.0158)
(.0163)
(.0167)
(.0172)
(.0177)
(.0182
(.0186)
(.0191)
(.0196)
(.0201)
(.0206)
(.0212)
(.0217)
(.0222)
(.0228)
(.0233)
(.0239)
(.0245)
Minimum Efficiency
For Threshold Shown
<90%
92.3
92; 4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.6
94.7
94.8
94.9
95.0
95.1
95.1
95.2
95.3
95.4
95.4
95.5
95.6
95.7
95.7
95.8
<89%
91.5
91.6
91.7
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93,3
93,4
. 93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.6
94.7
94.8
94.9
95.0
95.1
95.2
95.2
95.3
95.4
<88%
90.7
90.9
91.0
91 1
91.2
91.4
91.5
91- .6
91.7
91.8
91.9
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.1
93.2
93.3
93.4
93.5
93.6
93.7.
93.8
93.9
94.0
94.1
94.2
94.3
94.3
94.4
94.5
94.6
94.7
94.8
94.9
95.0
<87%
90.0
90.1
90.2
90.4
90.5
90.6
90.8
90.9
91.0
91.1
91.3
91.4
91.5
91.6
91.8
91.9
"92 0
92.1.
92.2
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
<86%
89.2
89.3
89.5
89.6
89.8
89.9
90.1
90.2
90.3
90.5
90.6
90.7
90.9
91.0
91.1
91,3
91 4
91.5
91.6
91.8
91.9
92.0
92.1
92.3
92.4
92.5
92.6
92.7
92.8
93.0
93. 1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.2
<85%
88.4
88.6
88.7
88.9
89.0
89.2
89.3
89.5
89.6
89.8
89.9
90.1
90.2
90.4
90.5
90 6
• 90.8
90.9
91.0
91.2
-•'91.3
91.4
91.6
91.7
91.8
92.0
92*, I
92.2
92.3
92:5
92.6
92.7
92.8
92.9
93.1
93.2
93.3-
93.4
93.5
93.6
93.7
Facility autocorrelation = 0.70
In computing the 30-day average variability, a geometric mean
emission level of 92% was assumed.
-------
2-29
EXHIBIT 2-13:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN MO MORE THAN ONE FAILURE PER FIVE YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
.41 •
.42
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Oev.
of 30-Day
Average*
(.0054)
(.0057)
(.0060)
(.0063)
(.0065)
(.0068)
(.0071)
(.0074)
(.0077)
(.0080)
(.0083)
(.0087)
(.0090)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0109)
(.01x3)
(.0115)
(.0119)
(.0123)
(.0127)
(.0130)
(.0134)
(.0138)
(.0141)
(.0145)
(.0149)
(.0153)
(.0157)
(.0161)
(.0165)
(.0169)
(.0174)
(.0178)
(.0182)
(.0187)
(.0192)
(.0196)
Minimum Efficiency .
For Threshold Shown
<90% <89% <88% <87% <86%
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94.7
94.8
94.3
94.9
95.0
95.1
95.1
95.2
95.3
95.4
95.4
91.2
91.3
91.5
91.6
91.7
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94,0
94.1
94.1
94.2
94.3
94.4
94.5
94,6
94.7
9*1.7
94.8
94.9
95.0
90.4
90.6
90.7
90.8
90.9
91.0
91.1
91.2
91.4
91.5
91.6
91.7
91.8
91.9
92.0
92.1 •
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93. 7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
89.7
89.8
89.9
90.0
90.1
90.3
90.4
90.5
90.6
90.7
go.1"
91.0
91.1
91.2
91.3
91.4
91.6
91.7
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.5
93.7
93.8
93.9
94.0
94.1
88.9
89.0
89.1
89.3
89.4
89.5
89.6
89.8
89.9
90.0
90.2
90.3
90.4
90.5
90.7
90.8
90.9
91.0
91.2
91.3,
91.4
91.5
91.6
91.7
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.7
92.3
92.9
93.0
93.1
93,2
Q-5 3
•J W » W
93.4
93.5
93.5
88.1
88.2
83. 3
88.5
88.6
88.8
88.9
89.0
89.2
89.3
89.5
89.6
89.7
89.9
90.0
90.1
90.3
90.4
90.5
90.7
90.8
90.9
91.0
91.2
91.3
91.4
91.5
91.7
•;.8
9L.9
92.0
92.1
92.3
92.4
92.5
92.5
92.7
92.3
92.9
93.1
93.2
Facility autocorrelation = 0.55
ln. computing the 30-day average variability, a. geometric mean
emission level of 92^ was assumed.
-------
O 2-30
EXHIBIT 2-19:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER FIVE YEARS
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87% <86% <85%
(.0058)
(.0061)
(.0064)
(.0067)
(.0070)
(.0073)
( . 0076)
(.0080)
(.0083)
(.0086)
(.0089)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0110)
(.01N13)
(.0117) •
(.0121)
(.0124)
(.0128)
(.0132)
(.0135)
(.0139)
(-0143)
(.0147)
(.0151)
(.0155)
(.0159)
(.0164)
(.0168)
(.0172)
(.0177)
(.0181)
(.0186)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
.94.
94.
94.
94.
94.
94.
94.
94.
94.
95.
95.
95.
95.
95.
95.
95.
95.
95.
2
3
4
5
6
6
7
8
9
0
1
2
3
4
5
6
7
8
9
9
0
1
2
3
4
4
5
6
7
8
9
9
0
1
2
2
3
4
5
5
6
91.
91.
91.
91.
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
94.
94.
94.
94.
94.
94.
94.
94.
95.
95.
95.
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
3
4
5
6
7
8
9
g
0
1
2
90.
90.
90.
90.
91.
91.
91.
91.
91.
91.
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
94.
94.
94.
94.
94.
6
7
8
9
1
2
3
4
5
6
8
9
0
1.
2
3-
4
5
6
7
8
a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
.6
7
3
89.8
89.
90.
90.
90.
90.
90.
90.
90.
90.
91.
91.
91.
91.
91.
• 91,
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
g
I
2
3
4
6
7
8
9
1
2
3
4
5
7
8
9
0
1
f\
3
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
89.
89.
39.
89.
89.
89.
89.
90.
90.
90.
90.
90.
90.
90.
90.
91.
91.
91.
91.
91.
91.
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93,
93.
93.
93.
0
2
3
4
6
7
8
0
1
2
4
5
6
8
9
0
1
3
4
5-
6
8
9
0
1
2
3
5
6
7
8
9
0
1
2
3
5
6
7
8
9
88.2
88.4
88.5
88.7
88.8
89.0
89.1
.89.3
89.4
89.5
89.7
89.8
90.0
90.1
90.2
90.4
90.5
90-. 5
90.8
90.9
91.0
91.2
91.3
91.4
91.5
91.7
91.8
91.9
92.0
92.2
92.3
92.4
92.5
92.6
92.8
92.9
93.0
93.1
93.2
93.3
93.4
Facility autocorrelation^ 0.60
ln computing the 30-day average variability, a geometric mean
emission level of 92% was assumed.
-------
£> 2-29
EXHIBIT 2-13:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN MO MORE THAN ONE FAILURE PER FIVE YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
.41
.42
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Oev.
of 30-Day
Average^
(.0054)
(.0057)
(.0060)
(.0063)
(.0065)
(.0068)
(.0071)
(.0074)
(.0077)
(.0080)
(.0083)
(.0087)
(.0090)
(.0093)
(.0095)
(.0099)
(.0103)
(.0106)
(.0109)
(.0113)
(.0116)
(.0119)
(.0123)
(.0127)
(.0130)
(.0134)
(.0138)
(.0141)
(.0145)
(.0149)
(.0153)
(.0157)
(.0161)
(.0165)
(.0169)
(.0174)
(.0178)
(.0182)
(.0187)
(.0192)
(.0196)
Minimum Efficiency .
For Threshold Shown
<90% <89% <88% <87% <86% <85%
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94.7
94.8
94.3
94.9
95.0
95.1
95.1
95.2
95.3
95.4
95.4
91.2
91.3
91.5
91.6
91.7
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.0
93.1
93.2
93.3
93.4
93.5
93,6
93.7
93.8
93.9
94.0
94,1
94.1
94.2
94.3
94.4
94.5
94,6
94.7
94.7
94.8
94.9
95.0
Facil
90.4
90.6
90.7
90.8
90.9
91.0
91.1
91.2
91.4
91.5
91.6
91.7
91.8
91.9
92.0
92.1 -
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
•93. 7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
89.7
89.8
89.9
90.0
90.1
90.3
90.4
90.5
90.6
90.7
90."
91.0
91.1
91.2
91.3
91.4
91.6
91.7
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.5
93.7
93.8
93.9
94.0
94.1
88.9
89.0
89.1
89.3
89.4
89.5
89.6
89.8
89.9
90.0
90.2
90.3
90.4
90.5
90.7
90.8
90.9
91.0
91.2
91.3.,
91.4
91.5
91.6
91.7
91.9
92.0
92.1
92.2
92.3
92.4
92.5
92.7
92.3
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
ity autocorrelation
88.1
88.2
88.3
88.5
88.6 .
88.8
88.9
89.0
89.2
89.3
89.5
89.6
39.7
89.9
90.0
90. 1
90.3
30.4
90.5
90.7
90.8
90,9
91.0
91.2
91.3
91.4 .
91.5
91.7
:-8
S<<,9
92.0
92.1
92.3
92.4
92.5
92.5
92.7
92.3.
92.9
93.1
93,2
= 0.55
computing the 30-day averaqe variability, a. geometric mean
emission level of 92% was assumed.
-------
D 2-30
EXHIBIT 2-19:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER FIVE YEARS
Minimum Efficiency
For Threshold Shown
<90% <89% <88%
<86% <85%
(.0058)
(.0061)
\ »•—•-—•— *
(.0064)
(.0067)
(.00.70)
(.0073)
( . 0076)
(.0080)
(.0083)
(.0086)
(.0089)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0110)
(.0113)
(.0117)
(.0121)
(.0124)
(.0128)
(.0132)
(.0135)
(.0139)
(.0143)
(.0147)
(.0151)
(.0155)
(.0159)
(.0164)
(.0168)
(.0172)
(.0177)
(.0181)
(.0186)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
.94.
94.
94.
94.
94.
94.
94.
94.
94.
95.
95.
95.
95.
95.
95.
95.
95.
95.
2
3
4
5
6
6
7
8
9
0
1
2
3
4
5
6
7
8
9
9
0
1
2
3
4
4
5
6
7
8
9
9
0
1
2
2
3
.4
5
5
6
91.
91.
91.
91.
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
94.
94.
94.
94.
94.
94.
94.
'94.
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
s\
3
3
.4
5
.6
.7
,8
9
g
95.0.
95.
95.
1
2
90.
90.
90.
90.
91.
91.
91.
91.
91.
91.
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
94.
94.
94.
94.
94.
6
7
8
9
i
2
3
4
5
6
8
9
0
1
2
3-
4
5
6
7
8
a
6
1
2.
3
4
5
6
7
8
9
0
1
2
3
4
s
6
7
8
89.8
89.
90.
90.
90.
90.
90.
90.
90.
90.
91.
91.
91.
91.
91.
• 91,
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
93.
94.
94.
94.
94.
0
1
2
3
4'
6'
7
8
9
1
2
3
4
5
7
8
9
0
1
2
3
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
89.
89.
89.
89.
89.
89.
89.
90.
90.
90.
90.
90.
90.
90.
90.
91.
91.
91.
91.
91.
91.
91.
91.
92.
92.
92.
92.
92.
92.
92.
92.
92.
93.
93.
93.
93.
93.
93.
93.
93.
93.
0
2
3
4
6
7
8
0
1
2
4
5
6
8
9
0
i
3.
4
5-'
6
3
9
0
1
2
3
5
6
7
8
9
0
1
2
3
5
6
7-
8
9
88.2
88.4
88.5
88.7
88.8
89.0
89.1
89.3
89.4
89.5
89.7
89.8
90.0
90.1
90.2
90.4
90.5
90-. 6
90.8
90.9
91.0
91.2
91.3
91.4
91.5
91.7
91.8
91.9
92.0
92.2
92.3
92.4
92.5
92.6
92.8
92.9
93.0
93.1
.93.2
93.3
93.4
Facility autocorrelation =0.60
ln computing the 30-day average variability, a geometric mean
emission level of 92% was assumed.
-------
D 2-31
EXHIBIT 2-20:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER FIVE YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
.41
.42
.43
.44
.45
.46
' .47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Dev.
of 30-Day
Average^
(.0062)
(.0066)
(.0069)
(.0072)
(.0076)
(,0079)
(.0082)
(.0086)
(.0089)
(.0093)
(.0096)
(.0100)
(.0,104)
(.0107)
(.0111)
(.0115)
(.0118)
(.0122)
(00126)
(.0130)
(.0134)
(.0138)
(.0142)
(.0146)
(.0150)
(.0154)
(.0159)
(.0163)
(.0167)
(.0172)
(.0176)
(.0181)
(.0185)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
(.0215)
(.0220)
(.0226)
Minimum Efficiency
For Threshold Shown
<90% <89% <88%
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5"
93.6
93.7
93.8
93.9
94.0
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.7
94.8
94.9
95.0
95.1
95.1
95.2
95.3
95.4
95.5
95.5
95.6
95.7
95.8
95.3
91.5
91.6
91.8
91.9
92cQ
92,1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.7
94.8
94.9
95.0
95.1
95.2
95.3
95.3
95.4
90.8
90.9
91.0
91.1
91.2
91.4
91.5
91.6
91.7
91.8
92,0
92.1
92-. 2
92.-
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93,2
93.. 3
93.4
93.5
93.6
93.7
93.8
93.9
•94.0
94.1
94.2
94.3
'94.4
94.5
94.5
94.6
94.7
94.8
94.9
95.0
90.0
90.1
90.3
90.4
90.5
90.6
90.8
90.9
91.0
91.2
91.3
91.4
91.5
91.7
91.8
91.9
92.0
92.1
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
89.2
89.4
39.5
89.6
89.8
89.9
90.1
90.2
90.3
90.5
90.6
90.8
90/9
91.0
91.1
91.3-
91.4.
91.5
91.7
91.8
91.9
92.0
92.2
92.3
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.9
94.0
94.1
94.2
88.4
88.6
88.8
88.9
89.1
89.2
89.4
89,5
89.7
89.8
89.. 9
90. 1
90.2
90.4
90.5
90,7
90.8
90.9
91.1
91.2
"91.3
91.5
91.6
91.7
91.8
92.0
92.1
92.2
92.4
92.5
92.6
92.7
92.8
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93,7
Facility autocorrelation = 0.65
3-In computing the 30-day average variability, a geometric mean
emission level of 92% was assumedl
-------
£? 2-32
EXHIBIT 2-21:
MINIMUM GEOMETRIC MEAM EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER FIVE YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.33
.39
.40
-.41
.42
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Dev.
of 30-Day
Average^
(.0068)
(.0071)
(.0075)
(.0079)
(.0082)
(.0086)
(.0090)
(.0093)
(.0097)
(.0101)
(.0105) -
(.0109)
(.0112)
(.0116)
(.0120)
(.0124)
(.0128)
(.0133)
(.0137)
(.-0141)
(.0145)
(.0150)
(.0154)
(.0158)
(.0163)
(.0167)
(.0172)
(.0177)
(.0182)
(.0186)
(.0191)
(.0196)
(.0201)
(.0206)
(.0212)
(.0217)
(.0222)
(.0228)
(.0233)
(.0239
(.0245)
Minimum Efficiency
For Threshold Shown
<907, <89%
<87%
<85%
92.4
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94.7
94.8
94.9
95.0
95.1
95.1
95.2
95.3
95.4
95.5
95.5
95.6
95.7
95.8
95.8
95.9
96.0
96.1
91.7
91.8
91.9
92.1
92.2
92.3
92.4
92.5
92.6
92.7
92.8
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.6
94.7
94.8
94.9
95.0
95.1
95.2
95.3
95.3
95.4
95.5
95.6
95.7
90.9
"91.1
91.2
91.3
91.5
91.6
91.7
91.8
92.0
92.1
92.2
92.3
92.4
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.5
94.6
94.7
94.8
94.9
95.0
95.1
95.2
95.3,
90.2
90.3
90.5
90.6
90.7
90.9
91.0
91.2
91.3
91.4
91.5"
91.7
91.3
91.9
92.1
92.2
92.3
92.4
92.5
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5'
94.6
94.7
94.8
94.9
89.4
89.6
89.7
89.9
90.0
90.2
90.3
90.5
90.6
90.8
90.9
91.0
91.2
91.3
91.4
91.6
91.7
91.8
92.0
92.1
92.2
92.3
92.5
92; 6
92.7
92.8
93.0
93.1
93.2
93. a
93.4-
93.5
93.6
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
88.7
88.8
89.0
89.2
89.3
89.5
89.6
89.8
89.9
90.1
90.2
90.4
90.5
90.7
90.8
91.0
91.1
91.3
91.4
.91.5
91.7
91.8
91.9
92.1
92.2
92. -3
92.5
92.6
92.7
92.8
92.9
93.1
93.2
93.3
93.4
93.5
93.7
93.8
93.9
94.0
94.1
Facility autocorrelation =0.70
computing the 30-day average variability, a geometric mean
emission level of 92% was assumed.
-------
EXHIBIT 2-22:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER TEN YEARS
Dailv
*-* V* * » J
Std. Dev.
(in log)
.20
.21
.22
.23
.24
!26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
.41
.42
.43
.44
.45
.46
.47
.48
• ^w
49
• ^-7
.50
.51
.52
.53
• •*/.-+*>
.54
lie
« ^ v
.57
.58
.59
.60
Std. Dev.
of 30-Day
Average1
(.0054)
(.0057)
(.0060)
(.0063)
(.0065)
(.0068)
(.0071)
(.0074)
(.0077)
(.0080)
(.0083)
(.0087)
(.0090)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0109)
(.0113)
(.0116)
(.0119)
\ /
(.0123)
(.0127)
(.0130)
(.0134)
(.0138)
(.•0141).
\ » * ••"/
(.0145)
^ V V *. y
(.0149)
(.0153)
(.0157)
(.0161)
(.0165)
^ w w *. -— f
(.0169)
(.0174)
(.0178)
(.0182)
(.0187)
(.0192)
(.0196)
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87% <86% <85%
92.1 91.3 90.6' 89.8 89.0 88.2
Q2.2 91.5 90.7 89.9 89.1 88.3
92.3 91.6 90.8 90.0 89.3 88.5
92.4 91.7 90.9 90.2 89.4 88.6
92.5 91.8 91.0 90.3 89,5 88.8
92.6 91.9 91.1 90.4 89.7 88.9
92.7 92.0 91.3 90.5 89.8 89.1.
92.8 92.1 91.4 90.7 89.9 89.2
92.9 92.2 91.5 90.8 90.1 89.4
93.0 92.3 91.6 90.9 90.2 89.5
93.1 92.4 91.7 91. U 90.3 89.6
93.2 92.5 91.8 91.1 90.5 89,3
93.3 92.6 91.9 91.3 90.6 89.9
93.4 92.7 92.0 • 9U4 90.7 90.1
93.5 92.8 92.2 91.5 90.8 90.2
93.5 92.9 92.3 91.6 91.0 90.3
93.6 93.0 92.4 91.7 91.1 90.5
93.7 93.1 92.5 91.8 91. Zv 90.6
93.8 93.2 92.6 92.0 91.3 90.7
93.9 93.3 92.7 32.1 91.5 90.9
94.0 93.4 92.8 92.2 91.6-91.0
94.1 93.5 92.9 92.3 91.7 91.1
94.2 93.6 93.0 92.4 91.8 91.2
94.2 93.7 93.1 92.5 91.9 "1.4
94.3 93.8 93.2 92.6 92.1 91.5
94.4 93.9 93.3 92.7 92.2 91.6
94.5 93.9 93.4 92.8 92.3 91.7
94.6 94.0 93.5 9.3.0 92.4 91.9
94.7 94.1 93.5 93.1 92.5 92.0
94.7 94.2 93.7 93.2 92.6 92.1
. 9d.3 94.3- 93.8 93.3 92.7 '92.2
94.9 94.4 93.9 93.4 92.9 92.3
95.0 94.5 94.0 93.5 93.0 92.5
95.1 94.6 94.1 93.6 93.1 92.6
95.1 94.6 94.2 93.7 93.2 92.7
05. 2 9^.7 94.3 93.8 93.3 92.3
95.3 94.8 94.3 93.9 93.4 92.9
95.4 94.9 94.4 94; 0 93.5 93.0
Q5.4 95.0 94.5 94.1 93.6 93.2
95.5 95.1 94.6 94.2 93.7 93.3
95.o 95.1 94.7 94.3 93.8 93.4
Facility autocorrelation = 0.55
•In computina the 30-day average variability, a geometric
emission level of 92% was assumed.
-------
J) 2-34
EXHIBIT 2-23:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN NO MORE THAN ONE FAILURE PER TEN YEARS
Daily
Std. Dev.
(in log)
.20
.21
.22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
.34
.35
.36
.37
.38
.39
.40
.41
.42
.43
.44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Oev.
of 30-Day
Average^
' (.0058)
(.0061)
(.0064)
(.0067)
(.0070)
(.0073)
(.0076)
(.0080)
(.0033)
(.0086)
(.0089)
(.0093)
(.0096)
(.0099)
(.0103)
(.0106)
(.0110)
(.0113)
(.0117)
(.0121)
(.0124)
(.0123)
(.0132)
(.0135)
(.0139)
(.0143)
(.0147)
(.0151)
(.0155)
(.0159)
(.0164)
(.0168)
(.0172)
(.0177)
(.0181)
(.0186)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
Minimum Efficiency
For Threshold Shown
<90%
<87% <86% <85%
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.3
94.4
94.5
94.6
94.7
94.8
94.8
94.9
95.0
95.1
95.2
95.2
95.3
95.4
95.5
95.6
95.6
95.7
95.8
91.5
91.6
91.7
91.8
91.9
92.0
92.1
92.3
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93,4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.8
94.9
94.9
95.0
95.1
95.2
95.3
95.4
90.7
90.8
91.0
91.1
91.2
91.3
91.4
91.6
91.7
91.8
91.9
92 ..0
92.1
92.2
92.4-
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.8
94.8
94.9
89.9
90.1
90.2
90.3
90.5
90.6
90.7
90.8
91.0
91.1
91.2
91 ..3
91.5
91.6
91.7
91.8
92.0
92.1
92.2
92.3
92.4
92.5
92.6
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
89.2
89.3
89.4
89.6
89.7
89.9
90.0
90.1
90.3
90.4
90.6
90.7
90.8
90.9
91.1
91.2
91.3
91.5
91.6:
91.7
91.8
92.0
92.1
92.2
92.3
92.4
92.5
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.5
93.6
93.7
93.8
93.9
94.0
94.1
88.4
88.5
88.7
88.8
89.0
89.1
89.3
89.4
89.6
89.7
89.9
90.0
90.2
90.3
90.4
90.6
SO. 7
90.8
- 91.0
91.1
91.2
91.4
91.5
91.6
91.8
91.9
92.0
92.1
92.3
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93.2
93.3
93.4
93.6
93.7
Facility autocorrelation =0.60
ln computing the 30-day average variability, a geometric mean
emission level of 92% was assume'd.
-------
/> 2-35
EXHIBIT 2-24:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO .
MAINTAIN NO MORE THAN ONE FAILURE PER TEN YEARS
Daily
Std. Dev.
(in 1 og )
.20
.21
.22
.23
.24
.25
• .26.
''.27
.28
.29
.30
.31
.32
.33
.34 .
.35
.36
.37
.38
.39
.40
.41
.42
.43
. .44
.45
.46
.47
.48
.49
.50
.51
.52
.53
.54
.55
.56
.57
.58
,59
.60
Std. Dev.
of 30-Day
Average^
(.0062)
(.0066)
(.0069)
(.0072)
(.0076)
(.0079)
(.0082)
(.0086)
(.0089)
(.0093)
, (.0096)
(.0100)
(.0104)
(.0107)
(.0111)
( 0115)
(.0118)
(,0122)
(.0126)
(.0130)
(.0134)
(.0138)
(.0142)
(.0146)
(.0150)
(.0154)
(.0159)
(.0163)
(.0167)
(.0172)
(.0176)
(.0181)
(.0185)
(.0190)
(.0195)
(.0200)
(.0205)
(.0210)
(.0215)
(.0220)
(.0226)
Minimum Efficiency
For Threshold Shown
<90% <89% <88%
<86%
92.4
92.5
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93 9
94.0
94.1
94.2
94.3
94.4
94; 5
94.5
94.6
94.7
94.8
94.9
95.0
95.1
95.1
95.2
95.3
95.4
95.5
95.5
95.6
95.7
95.3
95.8
95.9
96.0
91.6
91.8
91.9
92.0
92.1
92.2
92.3
92.4
92.6
92.7
92.8
92.9
93.0
93.1
93 2
93,3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94,2
94,3
94.4
94.5
94.6
94-6
94.7
94.8
94.9
95.0
95.1
95.2
95.3
95.3
95.4
95.5
95.6
90.9
91.0
91.1
91.3
91.4
91.5
91.6
91.8
91.9
92.0
92.1
92.2
92.3
92.5
92.6
92 7-
92.8
92. -9
93.0
93.1
93.2
93.3
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94 2
94.3
94.4
. 94.5
94.5
94.6
94.7
94.8
94.9
95.0
95.1
95.2
90.1
90.3
90.4
90.5
90.7
90.8
90.9
91.1
91.2
91.3
91.5
91.6
91.7
91.8
92 0.
92-1
92.2
92.3
92.4
92.6
92.7
92.8
92.9
93.0
93.1
93.2
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.5
94.7
94.8
89.4
89.5
89.7
89.8
89.9
90.1
90.2
90.4
90.5
90.7
90.8
90.9
91.1
91.2
91 3
91.5
91.6
91.7
91.9
92.0-
92.1
92.2
92.4
92.5
92.6
92.7
92.8
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.9
94.0
94.1
94.2
94.3
94.4
88.6
88.8
88.9
89.1
89.2
89.4
39.5
89. 7 •-
89.8
90,0
90.1
90.3
90.4
90.6
90.7
90.9
91.0
91.1
91.3
91.4
91.5
91.7
91.8
91.9
92.1
92.2
92.3
92.5
92.6
92.7
92.8
92.9
93.1
93.2
93.3
93.4
93.5
93.6
93.8
93.9
94.0
Facility autocorrelation = 0.65
n computing the 30-day average variability, a geometric mean
emission level of 92% was assumed. . .
-------
2-36
EXHIBIT 2-25:
MINIMUM GEOMETRIC MEAN EFFICIENCIES REQUIRED TO
MAINTAIN HO MORE THAN ONE FAILURE PER TEN YEARS
Daily
Std. Dev.
(in log)
.20
.21
,22
.23
.24
.25
.26
.27
.28
.29
.30
.31
.32
.33
. .34
.35
.36
.37
.38
.39
.40
.'41
.42
.43
.44
.45
.46
.47
.48
.49
.50-
.51
.52
.53
.54
.55
.56
.57
.58
.59
.60
Std. Dev.
of 30-Day
Average^
(.0068)
(.0071)
(.0075)
(.0079)
(.0082)
(.0086)
(.0090)
(.0093)
(.0097)
(.0101)
(.0105)
(.0109)
(.0112)
(.0116)
(.0120)
(.0124)
(.0128)
(.0133)
(.0137)
(.0141)
(.0145)
(.0150)
(.0154)
(.0158)
(.0163)
(.0167)
(.0172)
(.0177)
(.0182)
(.0186)
(.0191)
(.0196)
(.0201)
(.0206)
(.0212)
(.0217)
(.0222)
(.0228)
(.0233)
(.0239)
(.0245)
Minimum Efficiency
For Threshold Shown
<90% <89% <88% <87%
<85%
92.6'
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.8
94.9
95.0
95.0
95.1
95.2
95.3
95.4
95.5
95.5
95.6
95.7
95.8
95.8
95.9
96.0
96.1
96.1
96.2
91.8
91.9
92.1
92.2
92.3
92.4
92.5
92.7
92.8
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.8
93.9
94.0
94.1
94.2
94.3
94.4
94.4
94.5
94.6
94.7
94.8
94.9
95.0
95.1
95.2
95.3
95.3
95.4
95.5
95.5
95.7
95.8
95.8
91.1
91.2
91.3
91.5
91.6
91.7
91.9
92.0
92.1
92.2
92.4
92.5
92.6
92.7
92.8
93.0
93.1
93.2
93.3
93.4
93.5
93.6
93.7
93.8
93.9
94.0
94.1
94.2
'94.3
94.4
94.5
94.6
94.7
94.8
94.9
95.0
95.1
95.2
95.3
95.4
95.5
90.3
90.5
90.6
90.8
90.9
91.0
91.2
91.3
91.5
91.6
91.7
91.9
92.0
92.1
92.2
92.4
92.5
92.6
92.7
92.9
93.0
93.1
93.2
93.3
93.4
93.5
93.7
93.8
93.9
94,0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
94.8
94.9
95.0
95.1
89.6
89.7
89.9
90.1
90.2
90.4
90.5
90.7
90.8
90.9
91.1
91.2
91.4
91.5
91.6
91.8
91.9
92.0
92.2
92.3
92.4
92 ."6
92.7
92.3
92.9
93.1
93.2
93.3
93.4
93. 5 '.
93.6
93.7
93.9
94.0
94.1
94.2
94.3
94.4
94.5
94.6
94.7
88.8
89.0
89.2
89.3
89.5
89.7
89.8
90. Q
90.1
90.3
90.5
90.6
90.8
90.9
91.1
91.2
91.3
91.5
91.6
91.8
91.9
92.0
92.2
92.3
92.4
92.6
92.7
92.8
92.9
93.1
93.2
93.3
93.4
93.5
93.7
93.8
93.9
94.0
94.1
94.2
94.3
Facility autocorrelation = 0.70
ln computing the 30-day average variability, a geometric mean
emission level of 92% was assumed.
-------
2-37
2.3 METHODOLOGY
Monte-Carlo simulation techniques were used to generate the data in
for the lognormal-distribution processes in exhibits 2-4 through 2-9.
The IBM Scientific Subroutine Package uniform random number generator
RANDU was used to generate the basic pseudo-random number stream for the
analyses. Box-and Muller's technique was used for generating
pseudo-random normal random deviates (with an accuracy in the resultant
distribution of at least six digits).1 Lognormal deviates were
generated by. the exponential function from these normal deviates. All
the estimates were generated using non-overlapping random-number streams
of 720,000 days (2,000 years). The standard errors of the estimates were
estimated by treating the 2,000 years as four replicated experiments of
500 years each. The computations were performed to 32 and 64 bit
accuracy on a Hewlett-Packard Series 1000 Model F computer, and the runs
consumed about 40 CPU hours of computation. The simulation was checked
by comparing statistics for which exact results were known from theory,
and all cases agreed to three or more digit accuracies (with sample
periods of 8,000,000 days in this testing).
The normal-distribution estimates were generated by exact solution
of the mathematical system, to accuracy of five, or more decimals.
•Completely exact solutions of the lognormal case were not available,
which led to the use of Monte-Carlo simulation. The critical values
given in exhibits 2-10 through 2-25 could not be found with the required
accuracy by simulation in the two-week term of this analysis, because
technique is significantly more accurate in its results than
those usually used in good statistical practice. It was used because of
the requirement to estimate very small probabilities.
-------
£> 2-38
such a determination by simulating all points necessary to search for the
critical values would have required approximately 2000 hours of computer
time. Accordingly, mathematical methods were used to compute these
values to within 0.2 percent. These methods, although derived from
standard techniques, were developed specifically for this analysis. The
techniques involve first using series approximations to the lognormal
distribution function and to its thirtieth convolution with itself, so as
to obtain accurate estimates of the third and fourth moments and
cumulants of the statistical distribution of the thirty-day averages.
(The first and second moments are known exactly in closed form.) These
estimates are then used in Edgeworth and Cornish-Fisher series expansions
of the distribution of the thirty-day averages, from.which expected rates
of threshold failures and critical values can be completed. It was found
that only one non-normal term of the Edgeworth expansion was required to
achieve the desired accuracy. These methods were compared with the
simulation techniques to verify their accuracy (and the accuracy of the
computer implementations'used.) All results were within 0.1 percent of^
the correct values as determined by simulation, indicating that the
expansions are somewhat more accurate in the region of interest than the
guaranteed bound of 0.2 percent we obtained analytically. The exact
expression used to compute the critical minimum-efficiency values
reported above is given in exhibit 2-25.
-------
I) 2-39
CJ
H—I
-r- uJ
eC UJ
>•
CJ I
^: CD
uj s:
t—11—i
o _i
u. o
11 r^*
UJ
U_
•ZL o
UJ UJ
-
ll 1 I I I f>"
CD uj a.
UJ *—*
CD 3: UJ
UJ
1 O
CD
sr-S uj
*~*' CQ
J— 00
«C CO IT)
21 UJ UJ
i— i cj a:
I— o rs
co c;_i
UJ CX I— i
> '
% "i
CO - S^
° 'i -
" T- r— i
^—^ 01
O -a ^1^
O ^ <_
r^i *•- *~
§ - * ^r
"~ CM
*-»» CM
1 CM
r—' O
— o
C3^ 4-
co
i- no
i ro
r— - — CO
. . , — O
>, •
rM ^— ^
D -— •
C . I CJ i —
in ,— > ~-~ o
i — S- i — O
• r— CM 1 CM CM
^-^ C- CM CU
4- i. O D
» — • CO t I
ro li
[~-. --^ II II II
CM V--
H- i- '5 X
_.
1 .t^ "i
'2 ^
1/1
U1
CD
£
CL
ro
E
s_
o
E
cn
o
CO.
3=.
O
sz
o
£ 1
o 1^
u
o v\
3 CO
re .0
-o o
R3
(/)
>1 »r-»
4-5
•r- 01
F*~* £3
S *>
'si ^
U ^
o3 o
_, .?-_
01 S
S_ , .
.5 »
S^
2 |g
s. s
s^ "*
^3 -1-3
a ^.
«3 *">
^ 01
•— T-*
T3 jvi
(C -Q
c
t) ra
S-
a
"S
*
IT)
1 —
,_!
-CJ
(^
CO
r-.
f*H^
•
CM
01
4^
C
ca
t/i
O
0
CO
JZ
s=
-r"
01
O)
0)
§
CJ
0)
,2;
o
>
CM
•p*
01
CO
•M
s-
I-™
S3
0
J^-
•I—
S-
u
f
s
-£^
o
1
-------
-------
£3-1
3.0 DESCRIPTIVE STATISTICS ON FGD SYSTEM EFFICIENCY DATA
Basic descriptive statistics were required in construction of the
model simulating the variable efficiency of steam generating .units. The
appropriate model structure and statistical distribution characteristics
were determined from an examination of observations reported from eleven
operating units. In addition, operating system parameters were varied
over ranges determined partly on the basis of parameter estimates made
from the data. This chapter consists of four sections describing the
observations and statistical analyses of them.
Section 3.1 defines the variable analyzed and describes the data
base used. A lognormal description of the analysis variable was used by
EPA and Entropy in previous analysis of this data. Section 3.2 discusses
the aopropriateness of such a description. As was shown in the analysis
reported in chapter 2.0, the issue of distributional form has little
influence on the principal results. In section 3.3 the means, standard
deviations, and autocorrelation factors are presented for each of the
eleven units. Differences in these parameters among the eleven units are
also noted. Additionally, the appropriateness of a first-order
autogressive model is discussed. Section 3.4 discusses possible
confounding of results caused by variation in the sulfur content of
untreated emissions.
3.1 DATA SET
Data on the efficiency factor from eleven electric utility steam
generating units were provided to VRI by the EPA. The data which was
received in printed tabular form was believed to. be that previously
-------
analyzed by EPA and Entropy. The eleven units, the number of observa-
tions from each and the time period in which the observations were made
are described in exhibit 3-1. Each observation represents a twenty-four
hour average of FGD system efficiency calculated from the unput and
output emission levels at each unit. (Efficiency was defined as the
percentage of S02 removed from the gas flow through the scrubbing
process.)
As shown in exhibit 3-1, the amount and time frame of the data
differed significantly from one unit to the next. The limited number of
observations from the Philadelphia and Pittsburgh II units make the data
from these two facilities of limited use. The twenty-four data points
from Conesville A and the twenty-one from Conesville B represent the only
measurements taken over a six-month period. Further, the data set for
any individual unit was generally characterized by intermittant data
voids. This scattering of data points limits the degree of certainty
with which any inferences concerning the correlation structure of the
process should be reviewed.
3.2 LOGNORMAL TRANSFORMATION
3.2.1 THE UNTRANSFORMED VARIABLE
An analysis of the distribution of the efficiency values for each of
the units indicated that at least four were clearly negatively skewed
(see exhibit 3-2). Skewness, the third moment about the mean, measures
the degree to which a distribution is unbalanced or "off-center". A
negative skewness factor indicates a distribution with a long left-hand
tail. A variable with a normal distribution is balanced and has a
skewness of zero. Two of the units with significant skewness were also
-------
/>3-3
EXHIBIT 3-1: ANALYSIS DATA BASE DESCRIPTION
Steam Generating
Unit
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesville B
Lawrence
Number of
Observations
Time Period During Which
Observations Were Made
66
89
20
11
8
52
42
31
24
21
30
July
Ouly
Sept
Nov.
Sept
Aug.
July
Dec.
Dec.
June
June
Jan.
21,
21,
. 14
10,
. 18
9,
30,
7,
7,
15,
15,
16,
1977
197-7
, 1977
1977
, 1977
1977
1978
1978
1978
1978 .
1978
1979
- Dec.
- Dec.
- Nov.
- Dec,
- Oct.
- Nov.
- Sept
- Jan.
- Jan.
- Dec,
- Dec.
- Feb.
23
23
9,
6;,
9»
23
. 8
25
29
13
13
21
, 1977
, 1977
1977
1977
1977
, 1977
,1978
, 1979
, 1979
, 1978
, 1978
,1979
(156
(156
(57
(27
(22
days)
days)
days)
days)
days)
(107 days)
(41 days)
(49
(51
days)
days)
(183 days)
(183 days)
(37
day?)
-------
o>
•r—
O
03 LO
O O
£= 03
cn
3-4
o
O
to
10
•r-
to
O
CO
10 r~i
o
(M
CM
CM
oo
LO
oo
en
c
o
CM
to
to
cu
LU
10
CO
03 LO
O O
-a
03
CO
LU
CO
LU
O>
•r-
to
CM
o
CO
CVl
LO
CM
CM
OO
CM .
CT>
r-1
CM
cn
LO
CM
CM
a>
cu o T-
> -r-
•r- -M C
CM
to
to
Q
CO
to
LU
LU
S«J
to
CM
CO
CQ
O
c
cu
o
LU
CD
O
LU
U_
to
cn
•r-"
to
in
*!-"
Wl
o
CU
o
O
•r- >-
-»-> O
s- o
•r- O
-a s~
0)
U1 +->
tO O
CU 01
•t- 03 -C
-a
W
CU
CU
to
CD
CO
CO
CM
r-
en
CM
LO
cr>
co
oo
CM
OO
00
o
•p
a> o
CD
CU
•a
CU
"3 .
cu cu
e
CJ
C
O
O)
N
03
o s_
S O
cu
cu
to 03
cu
i- 01
3 CU
03 ^3
CU CJ
o cu
OO
-^ — 03
to
O
(O tj
O S_
•r- O
c rct
cr. s_
= o
oo
-------
EXHIBIT 3-1: ANALYSIS DATA BASE DESCRIPTION
Steam Generating
Unit
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesville B
Lawrence
Number of
Observations
Time Period During Which
Observations Were Made
66
89
20
11
8
52
42
31
24
21
30
July
July
Sept
Nov.
Sept
Aug.
July
Dec.
Dec.
June
June
Jan.
21,
21,
. 14
10,
. 18
9,
30,
7,
7,
15,
15,
16,
1977
1977
, 1977
1977
, 1977
1977
1978
1978
1978
1978
1978
1979
- Dec.
- Dec.
- Nov.
- Dec.
- Oct.
- Nov.
- Sept
~ Jan.
- Jan.'
- Dec.
- Dec.
-. Feb.
23,
23,
9,
6,
9,
23,
. 8,
25,
29,
1977
1977
1977
1977
1977
1977
1978
1979
1979
13, 1978
13.,
1978
21, 1979
(156
(156
(57
(27
(22
(107
(41
(49
(51
days)
days)
days}
days)
days)
days)
days)
days)
days)
(183 days)
(183 days)
(37
days-)
-------
CO
tU
O
es
K-l
to
a
CO
a;
u.
CM
co
»-H
CO
a
CO
CO
LU
Ul
Cvl
CO
CQ
UJ
U
LU
CQ
LU
Lu
CO
fO LO
O O
O)
O
LU
LU
_J
CQ
es:
co
CVJ
oo
co
CVi
CM
o
oo
CM
LO
CM
CM
CM . CM
O
CM
Lf>
OO
LO
CM
oo
O)
O)
LO
CO
CO
CM
r~-
cn
cn
CsJ
LO
CO
oo
O)
O •!-
CJ
i— O
o
o co
CD CO
ra c
OJ
CO
LO
cn
cn
cn
cn
CO
cn
oo
LO
O
S. .O
•i- O
-a s_
QJ
a N
•i-' S-
o
CO 4->
to u
CO «
CO 01
CO
O
O
CO
QJ
CO •!-
CO
(U
"3 ...» CO
O
-------
3-5
found to have a significantly non-zero kurtosis. Kurtosis, a function of
the fourth moment about the mean, is often considered to measure the
degree of peakedness in the distribution. A positive value indicates a
higher peak (and longer tails).than in the normal distribution and a
negative value indicates a flatter peak. A variable with a normal
distribution has a kurtosis of zero.
Since the negative skewr.ess was a significant and consistent feature
of the efficiency variable, the loge transformation performed by both
EPA and Entropy in previous analyses of the data might be expected to
produce a variable with a more normal distribution.
3.2:2 THE TRANSFORMED VARIABLE
the transformation variable used is log (1-efficiency). For most of
the units, the transformation improved the normality of the distribution
significantly. This improvement can be seen in the skewness and kurtosis
values for the untransformed and transformed variable, displayed in
exhibit 3-2. The significance column of the. display indicates the
certainty with which the sample statistic implies an actual departure
from the normal distribution.
Exhibit 3-3 presents the arithmetic medians, means, and standard
deviations predicted for the observations under the lognormal assumption.
Comparison of these predicted values with the actual sample statistics
provides an intuitive feel for the goodness of fit of the lognormal
distribution. The lognormal assumption results in accurate predictions
. except in the estimates of standard deviations at the Conesville and
Lawrence units.
-------
D 3-6
EXHIBIT 3-3:
Unit
COMPARISON OF ARITHMETIC VALUES PREDICTED BY
THE LOGNORMAL DISTRIBUTION ASSUMPTION WITH
ESTIMATES FROM THE OBSERVATIONS
Arithmetic Values Predicted
By Lognormal Assumptions
Median
.Standard
Mean Deviation
Observed Estimates From
Untransformed Variable
Median
Standard
Mean Deviation
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesville B
Lawrence
84.4
83.3
80.8
85.4
97.0
89.2
88.5
96.0
86.0
92.5
95.4
83.8
82.2
80.2
85.0
96.8
89.1
88.3
95.8
84.5
91.6
93.4
4.9
6.2
4.5
3.2
1.2
1.3
2.2
1.6
7.3
4.2
6.6
84.6
83.3
81.2
86.1
96.7
88.9
88.5
95.7
84.1
91.9
95.3
83.8
82.3
80.3
85.1
,96.8
89.1
88.3
95.8
•*»•
84.7 '
91.7
93.6
4.7
5.9
4.6
•3.4
1.2
1.3
2.2
1.5
6.1
3.5
5.3
Ipor lognormal distributions: (the quantity'(1-efficiency) is lognormally
distributed).
Median = e^1 y= mean of logarithmic variable.
o-2/2 a= Standard deviation of log-
Mean
Standard Deviation = eyea /2(ea -1)1/2
arithmic variable.
-------
D 3-7
In spite of the apparent better agreement between the lognormal
distribution and that data, Kolmogorov-Smirnov tests comparing both
normal and lognormal distributions with the data indicated that either
assumption could be accepted.
Overall, then, the lognormal distribution presents a slightly better
characterization of the efficiency data than the normal. However, from
the available data, it is evident that the lognormal description is not
an ideal fit for all cases, and that the distribution is also very nearly
normal in many of the cases.'
3.3 ESTIMATED PARAMETERS AMD COMPARABILITY AMONG UNITS
3.3.1 MEANS AND STANDARD DEVIATIONS
Exhibit 3-4 presents the medians, means, and standard deviations of
the transformed variable, log (1-efficiency). The differences in the
means and standard deviations among the eleven units can readily be seen
from examination of the exhibit. Statistical tests1 were performed on
the differences in means and variances for each pair of units. (The
variance is the square of the standard deviation.) The results of these
tests are presented in exhibits 3-5 and 3-6. The level of significance
indicates the probability of the observed difference occurring by chance
if, in reality, there was no difference between the two means (or
variances). For example, the significance of the difference in variances
between the Louisville South and Pittsburgh I units is .0305. This means
that if there were really no difference in the variances at these units,
lT-tests were performed on the means and F-tests on the variances.
-------
3-8
EXHIBIT 3-4: ESTIMATED PARAMETERS OF TRANSFORMED VARIABLE
UNIT
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shav/nee .Venturi
Conesvilie A
Conesvilie B
Lawrence
MEDIAN
MEAN (y) STANDARD DEVIATION (a]
-1.8836
-1.7910
-1.6885
-1.9729
-3.5143
-2.2047
-2.1840 '
-3.1353
-1.8798
-2.5170
-3. -0791
-1.8608
-1.7868
-1.6492.
-1.9223
-3.4927
-2.2217
-2.1608
-3.2270
-1.9626
-2.5884
-3.0714
.295
.343
.234
.212
.359
.118
.182''
.368
.447
.474
.835
-------
co
co uj
uj i—<
C_> H-
2: t-<
UJ _J
a: i—i
LU o
u_
-------
CO
LU
CO I— «
LU J—
o >-<,
2: _J
UJ I— <
e: o
LU -«
CO
2: to
I— I LU
LU
CO
CJ
ot-
CO O
t-4 LU
J— CC
CO
UD
I
CO
X
UJ
-------
3-11
3.05 percent of random samples drawn from these units would produce a
difference in sample variance of the observed magnitude. A significance
level of .05 or lower is usually considered to be clear evidence"of a
difference. . .
The variances at the Chicago and Shawnee TCA units were signifi-
cantly lower than the variances at almost all of the other units. EPA
officials noted that both of these units are well run and a low
variability in efficiency was expected. The Pittsburgh II unit was :
described as being similar to the Shawnee TCA units, but because of the
limited number of observations the results are of less interest. The
significantly high variance at the Lawrence unit is believed by EPA
officials to be the result of an unusually low sulfur content of the
coal. .
Because of the highly significant differences in the variances among
the units examined and the inaccurate estimation of variance at the
Conesville and Lawrence units, it is not appropriate to combine these
variances for analysis. • ' -
3.3.2 AUTOCORRELATION
The lag-one autocorrelation estimates for each of the eleven units
are presented in exhibit 3-7, along with the number of observations from
which the estimates were drawn and the significance of the factor. (The
observations included were those for which there was also an observation
on the preceding or succeeding day.) The level of significance is
dependent on the number of observations, hence' the autocorrelation factor
of 0.6255 at the Conesville B unit is not significant because it is based
on only seven observations while the autocorrelation factor of 0.5995 at
-------
EXHIBIT 3-7: FIRST-ORDER AUTOCORRELATION FACTORS
ON THE VARIABLE LOG (1 - EFFICIENCY)
UNIT
Autocorrelation
Significant at
.05 level
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesvilie B
Lawrence
49
72
11
7
5
37
37
25
13
7
27
.6955
.6949
.4683'
-.1428
.2524
.6983
.5995
.8897
.7131
.6255
.6386
yes
yes
no
no
no
yes r*
yes
yes
yes
no
yes
1
The autocorrelation was determined by comparing day 't' with day 't-T;
the data was not collapsed and missing data was not filled-in, so that
only the observation days which were preceded or followed by another
observation day were included.
-------
£ 3-13
Shawnee TCA is significant. It seems almost certain that first-order
autocorrelation does, in fact, exist at most or all units. Entropy used
an estimate of 0.7 in their simulation model. This appears to be an
appropriate value if the model is dealing with a unit similar to one of
the Louisville units. However, for units more similar to the Shawnee TCA
unit, 0.6 would be a more reasonable estimate. Differences in opera-
tional procedures at the units are an unknown but probably relevant
factor.
-3.3.3 AUTOREGRESSIVE MODEL
The possibility of autocorrelation factors associated with lags of
two or more was also examined. A first-order autogressive model is one
in which the variable in time "t" is a function of the same variable in
time "t-1". A second-order autogressive model was compared with a
•-* -
first-order autogressive model. A comparison of the residual led to the
conclusion that the first-order autogressive model is appropriate. A
further examination of partial correlations up to a lag of ten led to the
»
conclusion that the first-order autogressive model is appropriate.
3.4 POSSIBLE CONFOUNDING FACTORS
It is recognized that many other factors may be related to the
efficiency variable. It was suspected that the efficiency factor at a
given unit might be related to the level of sulfur in the raw emissions.
Data was available for all but the Lawrence unit on the pounds per
million BTUs of sulfur in the gas before processing. The Pittsburgh I
and Conesville scrubbers processed gas with a significantly higher
average sulfur content than the other units (see exhibit 3-8). Mo
-------
3-14
EXHIBIT 3-8: COMPARISON OF MEAN SULFUR CONTENT OF
INPUT EMISSIONS AND MEAN EFFICIENCY
UNIT
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Philadelphia
Chicago
Shawnee TCA
Shawnee Venturi
Conesvilie A
Conesvilie B
Lawrence
MEAN SULFUR CONTENT
OF INPUT EMISSIONS
(Ib/MMBTU)
5.653
5.687
6.647
5.462
5.049
5.643
5.555
5.660
7.793
7.359
NA
MEAN OF EFFICIENCY
(Arithmetic Equivalent
of Transformed Variable)
83.8
82.2
80.2
85.0
96.r8
89.1
88.3
95.8
84.5
91.6
93.4
-------
3-15
relationship appeared to exist, however, between mean efficiency at a
unit and the mean level of sulfur before scrubbing.
Within individual units, statistically significant correlations
between efficiency and sulfur content were found at two units, the
Chicago unit and the Shawnee TCA unit, At the Shawnee TCA unit, the
relationship was the expected negative one (-.45) with increasing sulfur
content leading to decreasing efficiency. At the Chicago unit, however,
a positive correlation (.47) was found, with increasing sulfur content
leading to increasing efficiency.
On the basis of the evidence, then, one must concise that there is
no predictable relation between the actual levels of sulfur emissions
before scrubbing and the efficiency of the scrubbing operation, and that
the analyses reported here are not contaminated by any confounding effect
of this nature.
••>
Many additional factors are of probable relevance in determining the
efficiency levels of scrubbers. Operating procedures can be altered to
compensate for high or low sulfur content as well as high or low electri-
*
city demands. The location and type of measuring device used can affect
efficiency readings. The age, type, and condition of. the scrubber
equipment may also affect efficiency. The present data set does not
offer any evidence of the types or magnitudes of any effects from these
or other sources.
-------
-------
4.0 COMPARISON WITH ENTROPY RESULTS
This chapter summarizes the degree to which the findings in the
preceding chapters appear to agree with the results developed by Entropy.
Environmentalists, Incorporated. It is organized into two sections which
parallel the material presented in chapters 2.0 and 3.0. In the first
section the number of exceedences predicted by Entropy are compared to
those predicted by VRI, with a potential explanation of the observed
differences. The second section compares the VRI and Entropy descrip-
tions of the statistical structure characterizing the efficiency of
eleven flue gas desulfurization (FGD) units at eight electric utility
sites. The disparities between the Entropy and VRI estimates of process
parameter values are examined, and rationales for these differences are
discussed. -
4.1 PREDICTED EXCEEDEMCES
Although the details of Entropy's 1,000 year simulation were not .
available, VRI believes the material presented in chapter 2.0 nearly
replicates the Entropy approach. Some differences between the VRI and
Entropy simulated data are attributable to the inherent random nature of
the simulation process itself and the slight improvement in confidence
levels of VRI's figures produced as a consequence of the doubling of the
number of simulated years (2,000 instead of 1,000). Where VRI used
parameters comparable to those reported by Entropy, reasonably similar
numbers of exceedences were predicted.
Although these results show generally the same pattern of effects,
there are differences greater than can be explained by chanca effects.
-------
D 4-2
In view of the great care taken in this analysis, including special
rechecking of the disparate results, we suspect that the Entropy results
are probably less accurate where differences exist, possibly d-a to the
use of less accurate random number generation and transformation tech-
niques. In this connection, it is worth noting that VRI's estimates were
generated using methods considerably more precise than usually found in
good statistical practice. This extra precision was required in view of
the requirements to make accurate estimates O-F extremely small
probabilities.
Despite these minor differences, VRI's results substantiate
Entropy's conclusion that the number of exceedences per year is extremely
sensitive to the median (or mean) F6D system efficiency and the varia-
bility in this efficiency. YRI-simulated values nearly replicate
Entropy's findings that the degree of autocorrelation can affect,-the
number of- exceedences although with less impact than variation in the
mean and variance. VRI's analyses also provide information not provided
by Entropy such as the data in exhibits 2-10 through 2-15; in these-
areas, no comparisons are possible. '
4.2 PROCESS STRUCTURE
Analysis of the 24-hour FGD efficiency data indicate that the
measured values of efficiency are not symetrically distributed about
their mean, generally weakening any normal distribution hypothesis.
VRI's analysis agrees with the Entropy and EPA findings that the quantity
(1-efficiency) has a distribution which can be reasonably approximated by
a lognormal distribution. There are many other candidate distributions
-------
£> 4-3
which might equally well'be used to describe the observed distribution of
efficiency values. As shown in chapter 2.0, adoption of other distribu-
tions would not significantly influence the analysis results, but instead
might confuse major differences between the Entropy and VRI results with
insignificant discrepancies. Consequently, the above analysis used pri-
marily the lognormal distribution hypothesis proposed by EPA and con-
curred with by Entropy.
Entropy further found that the FGD efficiency data had significant
first-order autocorrelation. VRI's results upheld this finding even
though VRI's estimate of autocorrelation was based on consecutive calen-
dar days rather than the method suggested by Entropy's statistical con-
sultant which collapsed serial data into a string of days for which data
were available. In addition, VRI's negative finding on the presence of
higher order autocorrelation helped to validate the Entropy implicit
assumption that first-order (one day) lags were sufficient to describe
process time dependencies.
VRI used a data base which appeared to be approximately, but not
exactly, the same as that employed in the Entropy analysis. Specific
differences between the data provided are evidenced: (1) by disparities
in the numbers of observations at particular sites; and (2) by differ-
ences in numerical estimates. Disparities in the numbers of observations
occurred for two of the utilities reported, i.e.: .
Number of Observations
Site VRI Entropy
Chicago 52 35
Shawnee TCA 42 : 37
-------
4-4
Entropy does hot report the number of observations from the Lawrence
unit, so comparisons cannot be made. VRI-estimated parameter values for
a and y generally differ from Entropy's estimates by no more than two
percent except for the following sites.
Logarithmic Parameter Values
Site
Chicago
Shawnee TCA
Lawrence
UVRI
-2.222
-2.161
-3.. 071
a£
-2.206
-2.168
-3.437
UVRI
.118
.182
.835
v
.106
.186
.676
y VRI
.698
.600
.639
CTE
.86
.65
N/A
As noted above, VRI and Entropy were not using identical data bases for
the Chicago and Shawnee TCA sites*. It is expected chat the differences
at the Lawrence site may also be the result of a different data base.
Finally, the Entropy .data base combined observations from the Louisville
north and south units into a single site (Cane Run) while they were
treated separately in VRI's analysis. Entropy notes that averaging the
results of these two units reduces the overall variability of the com-
-------
which may not represent the future state of the art for boiler units.
VRI therefore did not combine all the data together to estimate future
site variability. Entropy, in its analysis of these differences, did
combine the data to generate forecasting intervals, discussed in terms of
levels of correctness. In this analysis, Entropy assumed that future -
sites would have levels of variability distributed as broadly as the
variabilities observed at existing sites. Thus, Entropy assumed that the
data from each of the existing sites constitutes a sample representatin^
appropriate state of the art design and operating practices which would
be used in future facilites. Without this assumption, t'.are is no
justification for using forecasting intervals based on the complete range
of variabilities. ' .
Rather than adopt this strong assumption, VRI has chosen to present
the bulk of its results in parametric form covering the range of
variabilities, leaving engineering analysis (combined with the data from
chapters 2.0 and 3.0) to identify the levels of variability which should
actually be expected at future sites. EPA personnel suggested that
Shawnee TCA and Pittsburgh II might be the best representatives of future
practices. Statistical analysis of these two sites suggests that they
had a common-variability. Accordingly, a confidence interval for the
variability at these sites was presented in chaoter 2.0. A confidence
interval is also presented there for the Louisville units.
-------
-------
APPENDIX E
EPA RESPONSE TO PETITIONS FOR RECONSIDERATION
-------
-------
Wednesday
February 6, 1980
'esi
Protection Agency
Standards of Performance for
Stationary Sources for Electric Utility
Steam Generating Units; Decision in
Response to Petitions for
Reconsideration
E-l
-------
0210 Federal Register / Vol. 45, No. 28 / Wednesday, February 6,1980 / Rules and Regulations
ENVIRONMENTAL PROTECTION
AGENCY
40 CFR Part 60
Standards of Performance for New •
Stationary Sources; Electric Utility
Steam Generating Units; Decision in
Response to Petitions for
Reconsideration :
AGENCY: Environmental Protection
Agency (EPA).
ACTION: Denial of Petitions for
Reconsideration of Final Regulations.
SUMMARY: The Environmental Defense
Fund. Kansas City-Power and Light
Company, Sierra Club, Sierra Pacific
Power Company and Idaho Power
Company. State of California Air
Resources Board, and Utility Air
Regulatory Group submitted petitions
for reconsideration of the revised new
source performance standards for
electric utility steam generating units ;
lhal were promulgated on June 11, 1979
(44 FR 33580). The petitions were
evaluated collectively since the
petitioners raised several overlapping
issues. When viewed collectively, the
petitioners sought reconsideration of the
standards of performance for sulfur
dioxide (SO»), purliculate matter, and
nitrogen oxides (NOX). In denying the
petitions, the Administrator found that
tha petitioners had failed to satisfy the
statutory requirements of section
307(dj(7)(B) of the Clean Air Act. That
Is, the petitioners failed to demonstrate
cither (1) lhat.it was impractical to raise
their objections during the period for
public comment or (2) that the basis of
their objection arose after the close of
the period for public comment and the
objection was of central relevance to the
ouicome of the rule. This notice, also
responds to certain procedural issues
raised by the Environmental Defense
Fund (EOF). It should be noted that the
N'atunil Resources Defense Council
(N'RDC) f.led a July 9. 1979. letter in
which they concurred with the
procedural issues raised by EOF.
DATES: Effective February 6. 1980.
Interested persons may advise the
Agency of any technical arrors by
March 7, I960.
ADDRESSES: EPA invites information
front interested persons. This
information should be sent to: Mr. Don
R. Goodwin, Director. Emission
Standards and Engineering Division
(MD-13), Environmetslal Protection
Agency. Research Triangle Park. North
Carolina 27711, telephone [919] 541-
5271.
Docket Number OAQPS-78-1
contains all supporting materials used
• by EPA in developing the standards, .
including public comments and .. ... •
materials pertaining to the petitions for
reconsideration. The docket is available
for public inspection and copying ' .
between 9:00 a.m. and 4:00 p.m., Monday
through Friday at EPA's Central Docket
Section, Room 2903B, Waterside Mall,
401 M Street, SW., Washington, D.C. ",
20460. ; ;
FOB FURTHER INFORMATION CONTACT:
Mr. Don R. Goodwin, Director, Ernissiop
Standards and Engineering Division-
(MD-13), Environmental Protection
Agency, Research Triangle Park. North •
Carolina 27711. telephone (919) 541- . •.
5271.
SUPPLEMENTARY INFORMATION:
Background •
On September 19,1978, pursuant to
Section 111 of the Clean Air Act
Amendments of 1977, EPA proposed ,
revised standards of performance to
limit emissions of sulfur dioxide (SOi),
particulate matter, and nitrogen oxides
(NO*) from new, modified, and '
reconstructed electric utility steam
generating units (43 FR 42154). A public
hearing was held on December 12 and
13,1978. In addition, on December 8,
1978, EPA published additional
information on the proposed rule (43 FR
57834). In this notice, the Administrator
set forth the preliminary results of the
Agency's analysis of the environmental,
economic, and energy impacts
associated with several alternative
standards. This analysis was also
presented at the public hearing on the
proposed standards. The public
comment period was extended until
January 15,1979, to allow for comments
on this information. -
After the Agency had carefully
evaluated the mora than 600 comment
letters and related documents, the
Administrator signed the final standards
on June 1,1979. In turn, they were
promulgated in the Federal Register on
June 11,1979.
On June 1,1979, the Sierra Club filed a
petition for judicial review of the
standards with the United Sta.tes Court
of Appeals for the District of Columbia.
Additional petitions were filed by
Appalachian Power Company, et al., the
Environmental Defense Fund, and the
State of California Air Resources Board
before the close of the filing period on
August 10. 1979.
In addition, pursuant to section
307(d)(7)(B) of the Clean Air Act, the
Environmental Defense Fund, Kansas
City Power and Light Company, Sierra
Club, Sierra Pacific Power Company and
Idaho Power Company, State of
California Air Resources Board, and
"Utility Air Regulatory Group petitioned
the Administrator for reconsideration of
the revised standards.
Section 307(d)(7)(B) of the Ant
provides that:
Only an objection to a rule or procedure
which was raised with reasonable specificity
during the period for public comment
(including any public hearing) may be raised
during judicial review. If the person raising
an objection can demonstrate to the
Administrator that it was impracticable to
raise such objection within such time or if the
grounds for such objection arose after the
period for public comment (but within the
tirasf specified for judicial review) and if such
objection is of central relevance to the
outcome of the rule, the Administrator shall
convene a proceeding for reconsideration of
the rule and provide the same procedural
rights as would have been afforded had the
information been available at the time Shu .
rule was proposed, tf the Administrator
refuses to convene such a proceeding, such
person may seek review of such refusal in the
United Stales Court o£ Appeals for the ,
appropriate circuit (as provided in subsection.
(b)). ': . . ' '.
The Administrator's findings and
responses, to the issues raised by the
petitioners are presented in this notice.
Summary of Standards .' '
Applicability
The standards apply to electric utility
steam generating units capable of firing
more than 73 MW (250 million Btu/hour)
.heat input of fossil fuel, for which
construction is commenced after
'September 18,1£78. Industrial
cogeneration facilities that sell less than
25 MW of electricity, or less than one- '
third of their potential electrical output
capacity, are not covered. For electric
utility combined cycle gas turbines.
, applicability of the standards is
determined on the basis of the fossil-fuel
fired to the steam generator.exclusive of
the heat input, and electrical power
contribution of the gas turbine.
SO2 Standards
The SO2 standards are as follows:
(1) Solid and solid-derived fuels
(except solid solvent refined coal): SO3
emissions to the atmosphere are limited
to 520 ng/J (1.20 Ib/inillion Btu) heat
input, and a 90 percent reduction in
potential SO2 emissions is required at all
times exc«pt when emissions to the
atmosphere are less than 260 ng/J (0.60
Ib/miilion Btu) heat input. When SO-
emissions are less than 260 ng/J (0.60 lb/
million Btu) heat input, a 70 percent
reduction in potential emissions is
required. Compliance with the emission
E-2
-------
Federal Register / Vol. 45, No. 26 / Wednesday,
limit and percent redaction requirements
is determined on a continuous basis by
using continuous monitors to obtain a
30-day rolling average. The percent
reduction is computed on the basis of.
overall SOa removed by all types of SO2
and sulfur removal technology, including
flue gas desulfurization (FGD) systems
and fuel pretreatment systems (such as
coal cleaning, coal gasification, and coal
liquefaction). Sulfur removed by a coal
pulverizer or in bottom ash and fly ash
may be included in the computation.-
(2) Gaseous and liquid fuels not
derived from solid fuels: SO2 emissions
into the atmosphere are limiteed to 340
ng/J (0.60 Ib/million Btu) heat input, and
a 90 percent reduction in potential SO2
emissions is required. The.percent
reduction requireme'nt does not apply if
SO2 emissions into the atmosphere are
less than 86 ng/J (0.20 lb/mil!ion Btu)
beat input. Compliance with the SOa
emission limitation and percent
reduction is determined on a continuous
basis by using continuous monitors to
obtain a 30-day rolling average.
(3) Anthracite coal: Electric utility
steam generating units firing anthracite
coal alone are exempt from the
percentage reduction requirement of the
' SOa standard but are subject to the 520
ng/| (1.20 Ib/million Btu) heat input
emission limit on a 30-day rolling
average, and all other provisions of the
regulations including the particulate
matter and NO* standards.
(4) Noncontinental areas: Electric
utility steam generating units located in
noncontinental areas (State of Hawaii,
the Virgin Islands, Guam, American
Samoa, the Commonwealth of Puerto
Rico, and the Northern Marina Islands)
are exempt from the percentage
reduction requirement of the SQ~
standard but are subject to the
applicable SO= emission limitation i>'.td
all other provisions of the regulatioas
including the particulate matter and NOX
standards.
(5) Resource recovery facilities:
Resource recovery facilities which
. incorporate electric utility steam
generating units that fire less than 25
percent fossil-fuel on a quarterly (90-
day) heat input basis are not subject to
the percentage reduction requirements
but are subject to the 520 ng/J (1.20 lb/.
million Btu) heat input emission limit.
Compliance with the emission limit is
determined on a continuous basis using
continuous monitoring to obtain a 30-
day rolling average. In addition, such
facilities must .monitor and report their
heat input by fuel type.
(8) Solid solvent refined coal: Electric
Utility steam generating units firing solid
Solvent refined coal (SRC I) are subject
to the 520 ng/I (1.20 Ib/million Btu) heat
input emission limit (30-day rolling
average) and all requirements under the
NOX and particulate matter standards.
Compliance with the emission limit is
determined on a continuous basis using
a continuous monitor to obtain a 30-day
rolling average. The percentage
reduction requirement, which is
obtained at the refining facility itself, is
85 percent reduction in potential SO2
emissions on a 24-hour (daily) averaging
basis. Compliance is to be determined
by Method 19. Initial full-scale
demonstration facilities may be granted
a commercial demonstration permit
establishing a requirement :of 80 percent
reduction in potential emissions on a 24-
• hour (daily) basis. .
Particulata Matter Standards
The particulats matter standard limits
emissions to 13 ng/J (0.03 Ib/million Btu)
heat input. The opacity standard limits
the opacity of emissions to 20 percent (6-
minute average). The standard's are
based on the performance of a well-
designed and operated baghouse or
electrostatic precipitator.
A'OX Standards
The NOX standards are based on
combustion modification and vary
according to the fuel type. The
standards are: .
(1) 86 ng/J (0.20 Ib-million Btu) heat
input from the combustion of any
gaseous fuel, except gaseous fuel
derived from coal;
(2) 130 ng/J (0.30 Ib/million Btu) heat
input from the combustion of any liquid
fuel, except shale oil and liquid fuel
derived from coal;
(3) 210 ng/J (0.50 Ib/million Btu) heat
input from the combustion of
subbituminous coal, shale oil, or any
solid, liquid, or gaseous fuel derived
from coal;
(4) 340 ng/J (0.80 Ib/million Btu) heat
input from the combustion in a slag tap
furnace of any fuel containing more than
2fi percent, by weight, lignite which has
been mined hi North Dakota, South
Dakota, or Montana;
(5) Combustion of a fuel containing
more than 25 percent, by weight, coal
refuse is exempt from the NO, standards
and monitoring requirements; and
(8) 260 ng/J (O.eOib/miliion Btu) heat
input from the combustion of anthracite
coal, bituminous coal, or any other solid
fuel not specified under (3), (4), or (5).
Continuous compliance with the NO*
standards is required, based on a 30-day
rolling average. Also, percent reductions
in uncontrolled NOX emission levels are
required. The percent reductions are not
controlling, however, and compliance
with the NO. emission limits will assure
compliance with the percent reduction
requirements.
Emerging Technologies
The standards include provisions
which allow the Administrator to grant
commercial demonstration permits to
allow less stringent requirements for the
initial full-scale demonstration plants of
certain technologies. The standards
include the following provisions: ,
(1) Facilities using SRC I are subj.ect to
an emission limitation of 520 ng/J (1.20
Ib/million Btu) heat input, based on a • •
30-day rolling average, and an emission
reduction requirement of 85 percent,.
based on a 24-hour average. However,
the percentage reduction allowed under
a commercial demonstration permit for
the initial full-scale demonstration plant
using SRC I would be 80 percent (based
on a 24-hour average). The plant
producing the SRC I would monitor to
ensure that the required percentage
reduction (24-hour average) is achieved
and the power plant using the SRC I
would monitor to ensure that the 520 ng/
J heat input limit (30-day rolling
average) is achieved.
(2) Facilities using fluidized bed
combustion (FBC) or coal liquefaction
xvould be subject to the emission
limitation and percentage reduction
requirement of the SO* standard and to
the particulate matter and NO*
standards. However, the reduction in
potential SO2 emissions allowed under a
commercial demonstration pentlit for
the initial full-scale demonstration
plants using FBC would be 85 percent
(based on a 30-day railing average). The
NOX emission limitation allojwed under a
commercial demonstration permit for
the initial full-scale demonstration
, plants using coal liquefaction would be
3GO n»/J (0.70 Ib/miliion Btu) heat input,
based on a 30-day rolling average.
(3) No more than 15.000 MW
equivalent electrical capacity would be
allotted for the purpose of commercial
demonstration permits. The capacity
will be allocated as follows:
Equivalent electrical
Pollutant capacity MW
So-id 5oivent-r»fined coal „
FlLfJc-ad beci combustion
(atmospheric)
RL'itiized !>3d combustion
(prassurceil) . .
Ccal liquefaction
S07
SO*
SO,
NO
6.0CC-10.000
4CO-3 000
'400-1,200
750-10 CCO
Compliance Provisions
Continuous compliance with the SOj
and NO* standards is required and is to
be determined with continuous emission
monitors. Reference methods or other
approved procedures must be used to
supplement the emission data when the
E-3
-------
8212 Federal Register / Vol. 45, No. 26 / Wednesday, February 6, 1980 / Rules and Regulations
C'j:stlnuous emission monitors
malfunction in order to provide emission
data for at least 10 hours of each day for
at least 22 days out of any 30
consecutive days of boiler operation.
A malfunctioning FGD system may be
bypassed under emergency conditions.
Compliance with the particulate
standard is determined through
performance tests. Continuous monitors
arc required to measure and record the
opacity of emissions. The continuous
opacity data will be used to identify
CNCCSS emissions to ensure that the '
particulate matter control system is
being properly operated and maintained.
Issues Raised in the Petitions for
Reconsideration
/. SOs Maximum Emission Limitation of
52Qn§//(1.2 Ib/Million Btu) Heal Input
The Environmental Defense Fund
(EOF), Sierra Club, and State of
CnHfornia Air Resources Board (GARB)
requested that a proceeding be
convened to reconsider the maximum
SOi emission limitation of 520 ng/J {1.2
Ib/million Btu) heat input. In their
petition, EOF set forth several
procedural questions as the busts for
thefr request. First, they maintained that
they did not have the opportunity to
comment on curtain information tvhich
was submitted to EPA by the National
Coal Association at an April 5; 1979, ,
meeting and In subsequent
correspondence. The information
pertained to the impacts that different
emission limitations will have on coal
production in the Midwest and Northern
Apptilachia. They argued that this . ';;•
information materially influenced the,
Administrator's final decision. Further,
they maintained that the
Administrator's decision in setting the••-
(.•mission limitation was based on ex
parte communications and improper
OJMressional pressure.
The Sierra Club also raised objections
lo information developed during the
post-comment period. They cited the
Information supplied by the National •
Co?il Association, and the EPA staff
analysis of the impact that different
emission limitations would have on
burnable coal raserves. In addition, they
challenged the assumption that
conservatism in utility perceptions of
• scrubber performance could create a
significant disincentive against the
burning of high-sulfur coal reserves. The
Sierra Club maintained that this
information is of "central relevance"
sines it formed the basis of the
tstablishment of the final emission
limitation and that the Sierra Club was
denifid She opportunity to comment on
this information. Finally, the Sierra Club
and GARB subscribed fully to arguments
presented by SDF concerning ex parte
communications. •
Background • .
The potential impact that the emission
' limitation may have on high-sulfur coal
reserves did not arise for the first time in
the post-comment period. It was an
issue throughout the rulemaking. In the
proposal, the Agency stated that two
factors had to be taken into
consideration when selecting the ,
emission limitation—FGD efficiency and
the impact of the emission limitation on
high-sulfur coal reserves (43 FR 42160,
middle column]. The proposal also
indicated that, in effect, scrubber
performance determines the maximum
sulfur content of coals that can be fired
in compliance with emission limitation
even xvhen coal preparation is
employed. From the discussion it is clear
that the Administrator recognized that
midwestern high-sulfur coal reserves
could be severely impacted if the
emission limitation was npt selected
writh care (43 FR 42160, middle column).
In addition, the Administrator also
specifically sought comment on the
related question of new coal production
as it pertained to consideration of coal
impacts in the final decisipn (43 FR
42155, right column).
At the December 1978 public hearing
oh the proposed standards, the Agency
specifically sought to solicit information
on the impact that lower SO2 emission
limits (below 520 ng/J (1.2 Ib/millipn
Btu) heat input) would have on high-
sulfur coal reserves. In response to
questions from an EPA panel member
and the audience, Mr. Hoff Stauffer of
ICF, Inc. (an EPA consultant) testified
that the potential impact of lower
emission limitations on high-sulfur coal •
reserves would be greater in certain
states than was indicated by the results
of the macroeconomic analysis
conducted by his firm. He added further
that if the degree of reduction
achievable through coal preparation cr
scrubbers changed from the values
assumed in the analysis (35 percent for
coal preparation on high-sulfur coal and
90 percent for scrubbers) the coal
impacts would vary accordingly. That is.
if greater reduction could be achieved
by either coal preparation or by
scrubbers the impacts would be
reduced. Conversely, if the degree of
reduction achievable by either coal
preparation or scrubbers was less than
the values assumed, the impacts would
be more severe (public hearing
transcript, December 12,1978, pages 46-
47).
The subject was broached again when
Mr. Richard Ayres, representing the
Natural Resources Defense Council and
- serving as introductory spokesperson for
other public health and environmental
organizations, was asked by the panel
what effect lowering the emission
limitation would have on local high-
sulfur coal reserves. Mr. Ayres
responded that a lower emission
limitation may have the effect of
. requiring certain coals to be scrubbed •
more than required by the standard. He
added that the utilities would have an
economic choice of either buying local
high-sulfur coal and scrubbing more or
buying lower-sulfur coal which may not
be local and scrubbing less. He further
indicated that it was not clear that a
lower limitation would have the effect of
precluding any coal. In doing so, he
noted that the "conclusion depended
entirely on assumptions about the
possible emission efficiencies of
scrubbers." Finally, Mr. Ayres was
asked whether as long as production in
a given region increased that the
requirement of the Act to maximize the
use of local coal was satisfied. He
responded that it was a "matter of
degree" and that he would not say as
long as production in a given region did
not decline the statute was served
(public hearing transcript, December 12, •
1978, pages 77-^80).
Mr. Robert Rauch, representir'j the
Environmental Defense Fund, also
recognized in his testimony that
lowering the emission limitation to the
level recommended by EOF (340 ng/J
(0.8 Ib/million Btu) heat input) would
adversely impact high-sulfur coal
reserves. In his testimony he stated
"Adoption of the proposed lower ceiling"
would result in the exclusion of certain
high-sulfur coal reserves from use in
power plants subject to the revised
standard." He added that the use of
adipic acid gud clhjsr £lv.rry addilivsa
would enhance scrubber performance,
thereby alleviating the impacts on high-
sulfur coal (public hearing transcript,
December 13,1978, pages 189-1&1).
Mr. Joseph Mullan of the National
Coal Association testified in response to
a question from the hearing panel that
lowering the emission limitation from
520 ng/J (1.2 Ib/million Btu) heat input
would preclude the use of certain high-
sulfur coals. He added that the National
Coal Association would furnish data on
•-• such impacts (public hearing transcript, ,
December 13,1978, page 246).
Turning now to the written comments .
on the proposed standard submitted
jointly by the Natural Resources
Defense Council and the Environmental
Defense Fund, we see that they carefully
assessed the potential impacts on high- •
sulfur coal reserves that could result
E-4
-------
Federal Register / Vol. 45, No. 26 / Wednesday, February 6, 1980 / Rules and Regulations 8213
from •various emission limitations. They
concluded, "Generally, the higher the
percent removal requirement, the
smaller the percentage of coal reserves
which are effectively eliminated for use
by utility generating units/1 They went
on to argue that if their recommended
standard of 95 percent reduction in
potential SO2 emission was accepted a
lower emission limitation could be
adopted without adverse impacts on
coal reserves (OAQPS-78-1, IV-D-631,
page V-128). , -• :•, "!.";.
•Rationale for the Maximum Emission
Limit
The testimony presented at the public
hearing and the written comments
served to confirm the Agency's initial
position that scrubber performance.and
potential impacts on high-sulfur coal
reserves had to be carefully considered
when establishing the emission
limitation. Meanwhile, it became
apparent that the analysis performed by
EPA's consultant on emission limits
below 520 ng/J (1.2 Ib/million Btu) heat,
input might not fully reflect the impacts
on major high-sulfur coal production
areas. This finding was evident by study
of the consultant's report (OAQPS-73-1,
IV-A-5, Appendix D) which showed
that the model used to estimate coal
production in Appalachia and the
Midwest was relatively insensitive to
broad variations in the emission ceiling.
The Agency then concluded that the
macroeconomic model was adequate for
assessing national impacts on coal use,
but lacked the specificity to assess
potential dislocations in specific coal
production regions. In effect the analysis
tended to mask the impacts in specific
coal producing regions through
aggregation. Concern was also -raised as
to the validity of the modeling -
assumption that a 35 percent reduction
. in potential SO2 emissions can be
achieved by-coal washing on all high-
sulfur coal reserves. "
In view of these concerns, EPA -
• concluded shortly after the close of the
comment period that additional analysis
was needed to support the final
emission limitation. In February, EPA
began analyzing the impacts of
alternative emission limits on .local high-
sulfur coal reserves. To account for
actual and perceived efficiencies of
scrubbers, the staff assumed three levels
of scrubber control—-85 percent, 90
percent, and 95 percent. In addition, two
levels of physical coal cleaning were
reflected. The first level was crushing to
1.5 inch top-size and'the second was
crushing to % inch top-size, both
followed by wet beneficiation. In
addition, by using seam-by-seam data
on coal reserves and their sulfur
reduction potential (developed for EPA's
Office of Research and Development} it
was possible to estimate the sulfur
content of the final product: coal based
on reported chemical properties of coals
in the reserve base (OAQPS-78-1, IV-E-
12). Since this approach did not require
the staff to assume a single level of
sulfur reduction for all coal preparation
plants, it introduced' a major refinement
to the analysis previously per formed by
EPA's consultant. The analysis was
substantially completed in March 1979
(OAQPS-78-1, IV-B-57 and IV-B-72).
The April 5,1979, meeting was called
to discuss coal reserve data and the
degree of sulfur removal achievable
with physical coal cleaning (OAQPS-
78-1, IV-E-10). The meeting gave EPA
the opportunity to present the results of
its analysis and to verify the data and
assumptions used with those persons
who are most knowledgeable on coal.
production and coal preparation. EPA
sought broad representation at the
meeting. Invitees including not only the
National Coal Association but
representatives from the Environmental
Defense Fund, Natural Resources •
Defense Council, Sierra Club, Utility Air
Regulatory Group, United Mine Workers
cf America, and other interested parties.
The invitees were furnished copies of
the materials presented at the meeting,
subsequent correspondence from the
National Coal Association, and minutes
of the meeting. ...
The meeting served to confirm that .
the coal reserve and preparation data
developed independently by the EPA
staff were in close agreement with those
prepared by the National Coal
Association [NCA]. In addition, the
discussion led EPA to conclude that coal
preparation technology which required
crushing to %-inch top-size would be
unduly expensive, lead to unacceptable •
energy losses, and pose, coat handling .
problems (OAQPS-78-1, IV-E-11}. As a
result, the Administrator revised his
assessment of state-of-art coal cleaning
technology (44 FR 33596, left column).
In an April 19,1979, letter to the
Administrator (OAQPS-70-1. IV-D-763),
attorneys for the Environmental Defense
Fund and the Natural Resources
Defense Council submitted comments on
the information presented by the
National Coal Association at the April 5,
1979, meeting and in a subsequent NCA
letter to the Administrator dated April 6,
1979. In their comments, they were .
critical of the National Coal
Association's assumptions concerning
scrubber performance and the removal
efficiencies of coal preparation plants.
They also noted that the Associaton's
data was based on a small survey of the
total coal reserves in the Midwest and
Northern Appalachia. They argued
further that coal blending could serve to
reduce the adverse impact on high-sulfur
coal caused by a lower emission limit. In
doing so, they recognized that the
application of coal blending would have
to be undertaken on a case-by-case .
basis. Finally, they maintained that
there is no evidence that the coal
industry would be unable to meet
increases in coal demand even if the
. National Coal Association's reserve •:
data on coal preclusions were accepted.
In conclusion, they noted that the
Association's data was of questionable
, relevance since it was predicated on a
maximum removal efficiency of 90
percent
Subsequent correspondence from the
National Coal Association serv 1 to
reaffirm a point that had been made
earlier in the rulemaking. That is,
utilities would have a choice of either •
buying lower-sulfur coal and scrubbing
to meet " a percent removal requirement
or buying higher-sulfur coal and
scrubbing more than required by the
standard in order to meet the emission
limitation. In addition, they cited-the .
conservative nature of utilities and
stressed that this would be reflected in
tlisir coal buying practices. As was
discussed at the public hearing and in
the written comments such behavior by
utilities would result in adverse impacts
on the use of certain local high-sulfur '
coals.
In reaching final conclusions about
the impact of the SOi standard on coal
production, the Administrator judged
that utilities would be inclined to select
coals that would meet the emission limit
with no more than 90 percent reduction
hi potential SO* emissions l (44 FR
33596, left column). With this
• assumption, the analysis revealed that
an emission limit of less than 520 ng/j
create a disincentive to burn a
significant portion of the coal reserves
in the Midwest and Northern
Appalachia (OAQPS-78-1, IV-B-72). If
the emission limit had been set at 430 .'
ng/J (1.0 Ib/million BtuJ heat input, 15
percent of die total reserve base in the
Eastern Midwest .(Illinois, Indiana, and
Western Kentucky) would have been
impacted. The impact in Northern
Appalachia would be 6 percent and this
impact would have been concentrated in
the areas of Ohio and the northern part "
of West Virginia. If only currently
1 The pre vious version of the EPA analysis had •
assumed either 85 or 90 percent control levels in
addition to coal washing. That approach
disregarded the fact that the net reduction in
potential SOj emissions may have been greater than
90 percent in some cases.
E-5
-------
B214 Federal Register / Vol. 45, No. 26 / Wednesday. February 6, 1980 /Rules and Regulations
owned coal reserves are considered, up
to 19 percent of the high-sulfur coals in
some regions would be impacted
(OAQPS-78-1.1V-B-72). The
Administrator judged that such impacts
are unacceptable.
The final point made by NCA was
that utility coal buying practice typically
incorporates a margin of safety to
ensure compliance with SO2 emission
limitations. Rather than purchasing a
high-sulfur coal that would barely ..
comply with the emission limit, the
prudent utility would adopt a more
conservative approach and purchase
coal that would meet the emission limit
with a margin of safety in order to
account for uncertainly in coal sulfur
variability. This approach, which
reflects sound engineering principles,
could result in the dislocation of some
high-sulfur coal reserves.
The Administrator determined that
consideration of a margin of safely in
coal buying practice was reasonable.
Using NCA's recommendation of an 8.5
percent margin (reported as "about 10
percent'1 in the preamble to
promulgation], coal impacts were
reanalyzed. This study showed
additional coal market dislocations
(OAQPS-78-1, IV-B-72). For example, in
Illinois, Indiana, and Western Kentucky,
the impact on coal reserves by a 430 ng/
J (1.0 Ib/million Btu) heat input emission
limit increased from 15 percent without
the margin to 22 percent when the
margin was assumed. Considering only
currently owned reserves, the impact
increased from 19 percent to 30 percent.
Even with the margin, the analysis
predicted no significant impact for a 520
ng/J (1.2 Ib/milHon Btu) heat input
standard.
Having determined the extent of the
potential coal impacts associated with a
lower emission limit, the Agency then
assessed the potential environmental
bnnefits. The assessment revealed that
by 1995 an emission limit of 430 ng/J (1.0
lli/million Btu) heat input would reduce
national emissions by only 50 thousand
tons per year relative to the fUO ng/f (1.2
lb/million Btu) heat input limit. That is,
Itv projected emissions from new plants
would be reduced from 3.10 million tons
So a 3.05 million tons as a result of the
more stringent emission limit (OAQPS-
70-1,1V-B-75).
The petitions providing no information
to either refute the assumptions or the
findings of the final coal impact
analysis. The Sierra Club argued that
EPA had misinterpreted its own analysis
of coal impacts (Sierra Club petition,
page 9). They maintained that the EPA
figures presented at the April 5 meeting
(OAQPS-78-1, IV-E-11. attachment 3)
supported establishment of a 340 ng/J
(0.3 Ib/miliion Btu) heat input standard.-
In doing so the Sierra Club ignored the
analysis performed by the Agency after
the April 5 meeting, particularly with
respect to the Administrator finding that
utilities would purchase coal which
would meet the emission limit (with
margin) with no more than 90 percent
reduction in potential SOZ emissions. •
In conclusion, the decision as-to the
appropriate level of emission limitation
rested squarely on two factors. First, the
Administrator's finding that a 90 percent
reduction in potential SO2 emissions,
measured as a 30-day rolling average,
represented the emission reduction
achievable through the use of the best
demonstrated system of emission
reduction, and second, that the marginal
environmental benefit of a 430 ng/J (1.0 —•
Ib/ir.illion Btu) heat input standard
coupled with a 90 percent reductiort-In •
potential SO- emissions could not be .
justified in light of the potential impacts
on high-sulfur coal reserves. If he had
determined, as some petitioners
suggested, that higher removal
efficiencies were achievable on high-
sulfur coals, the emission limitation
could have been established at a lower
level without significantimpacts on
local high-sulfur coal reserves.
Environmental Defense Fund Procedural
Issues
EDF's petition objected to the fact that
after the close of the public comment
period, representatives of the National
Coal Association and a number of
members of Congress talked to EPA
officials and submitted documents to
EPA arguing that the ceiling should be
set at 520 ng/J (1.2. Ibs/million Btu) heat
input. EDF objected to these
communications on a number of
grounds. First, they argued that it was
improper, under Section 307(d) of the
Act, for the Agency to consider
information submitted more than 30
days.after the public hearing. Second,
they objected that the Agency failed to
make transcripts of the oral
communications, and that, in any event,
the summaries of those communications
that the Agency placed in the docket
were inadequate. Third, they implied
that Agency officials received additional
oral communications which wtvu not
documented in the rulemaking docket.
Fourth, they objected that these written
and oral communications were ex part?,
and therefore improper, citing, for
example. United States Lines, Inc. v.
FMC, 584 F. 2d 519 (B.C. Cir., 1978).
Fifth, they argued that the
Administrator's decision on the ceiling
was based in part on information
obtained in ex parte discussions and
thus not placed in the docket as of the
date of promulgation, in violation of
Section 307(d). Finally, they argued that
the communications from members of
Congress constituted improper pressure
on the Administrator's decision, citing.
for example, D.C. Federation of Civic
Associations v. Volpe, 459 F. 2d 1231
(D'.C. Cir. 1972). EDF argued that these
alleged procedural errors were of
central relevance to the outcome of the
rule, and that the Agency should
' therefore convene sc proceeding to
reconsider.
The Administrator does not believe
that the procedures cited by EDF were
improper. Moreover, as discussed
below, any arguable errors were not of
central relevance to the outcome of the
rule, and therefore do not constitute
grounds for granting EDF's petition to
reconsider.
First, it was not improper for the
Administrator to consider information
submitted more than 30 days after tlie
public hearing. Section 307(d)(5) requires
that the Administrator consider
documents submitted up to 30 days after
the hearing. It does not forbid the
Administrator to consider additional
comments submitted after that 30-day
period.
Second, the Agency's summaries of
oral communications were adequate.
Section 307(d)f5) does not require, as
EDF argues, that Agency .officials keep
transcripts of their oral discussions with
persons outside the Agency. It simply
requires the Agency to make a transcript
of the public hearing on a proposed
rulemaking. Third, Agency officials
wrote memoranda of all significant oral
communcations between Agency
officials and persons outside the
executive branch, such as the two
meetings with Senator Byrd, and the
memoranda were promptly placed in the
rulemaking docket. These memoranda
accurately reflect the information and'
arguments communicated to the Agency.
Fourth, the oral and written-
communications cited by EDF were not
ex parte. The Agency promptly placed
'the written comments in the rulemaking
docket where they were available to,the
public. Also, the NCA sent copies of its
written comments directly to the
principal parties to the rulemaking,
including EDF and NRDC. Similarly, the
Agency placed the memoranda of oral
communcations in the docket where
they were available to the public. Any
member of the public has had the
opportunity to submit a petition for
reconsideration if that information was
used erroneously by EPA in setting the
standard, and several persons ha ve
done so.
Fifth, contrary to EDF's assertion, the
Administrator's decision on the
E-6
-------
Federal Register / Vol. 45,. No.-26-/ Wednesday,. February 6. 1980-/ Rules and Regulations 8215
emission ceiling WAS not based on any
information not in the docket.
Finally, it was not-improper for the
Administrator to listen. to« and. consider
the views of Senators, and Congressmen,.
including Senator Byrd. It is. not unusual -
for members of Congress to express
' their views on the merits o£ Agency
rulemaking, and it is. entirely proper for
the Administrator to consider those
views.
EOF objects particularly, to a meeting
. the Administrator attended with Senator
Byrd on April 26,1979; arguing that the
contact was ex parte and! improperly
influenced the Administrator's decision.
Neither contention is correct. A
:memorandum summarizing the
discussion at the meeting was placc-tl in
the docket, and members of the public
have had the opportunity to comment on
it, as EOF has done. No new information
was presented to the Administrator at
the meeting. -
Senator Byrd's comments at this
meeting also tlid not improperly
influence the Administrator. Although
the Senator strongly urged the
Administrator to set the emission ceiling
at a level that would not preclude the-
use of any significant coal reserves, the
Administrator had already-concluded
from the 1977 Amendments to the Clean
Air Act that trie revised standards
should not preclude significant reserves.
This view was based an the
Administrator's interpretation of the
legislative intent of the 1977
Amendments and was reflected in the
proposed emission ceiling of 520 ng/J
{1.2 tbs/rnillion Btu) heat Input, as-
, discussed in the preamble to the
proposed standards (43 FR 42160).
This vi-jw was reaffirmed in. the final
rulemaking, based on the intent of the
1977 Amendments (44 FR 33595-33596}.
Although the Administrator was aware
(as he would have been even in-the
absence of a meeting) of. Senator Byrd's
concern t'.:at a ceiling lower then 520-'ng/
] f'j .'2. Ibs/milUoa Btu}heat input would
inappropriately preclude significant coal
reserves, ihs Administrator's decision
was not based on Senator Byrd's
expression of concern. The-
Administrator had already concluded
that anything more than a mininirfi
preclusion of the- use of particular ooaS
reserves would, in the absence of
significant resulting emission reductions,
be inconsistent with the intent of the
1977 Amendments. Because the
Agency's analysis showed that even an
'emission limit of 430 ng/J (1.0 Ibs/
• million Btu) heat input couk! preclude
the use of up to 22 percent of certain
coal reserves without significantly
reducing overall emissions, the
Administrator's judgment was that a
ceiling lower than 520 ng/J (1.2.Ibs/
million Btu} heat input was not justified.,
Thus, the views of Senator. By'rdand.
other members of Congress, at most,
served io reinforce, the Administratoc's
own judgment that the proper level fos
the standard was 520 ng/J. (1.2 Ibs/
million Btu). heat input Even assuming,
therefore, that it was improper for the
Administrator to consider the views of
members of Congress,..this procedural
"error" was not of central relevance, to
the outcome of the rule- •' :
II. SQX Minimum Controt. Level (70
Percent.Reduction of Potential
Emissions}
The Kansas City Power and Light
•Company (KCPL), Sierra- Club, and
Utility Air Regulatory Group (UARG)
requested that a proceeding be
convened to'reconsider the 70 percent
minimum control level which1 is-
applicable when burning tow-sulfur
coals. Both the Sierra Club and UARG
maintained that they did not have- an
opportunity to fully comment on. either
the final regulatory analysis ordry SO3
scrubbing technology since the-phase 3
macroeconomic analysis of the standard
(44 FR 33603, left column) and
supporting data were entered into the
record after the close of the public
comment period. Both claimed that their
evaluation of this additional information
provided insights which are of-central
relevance to the Administrator's final
decision and that reconsideration of the
standard is warranted. The KCPL
petition did not allege improper
administrative procedures, but asked for.
reconsideration based on their-
evaluation of the merits of the standard.
In. seeking a more stringent minimum
reduction requirement, the Sierra Club •
contended that dry SO2 scrubbing is not
a demonstrated technology and,.
therefore, no basis exists for a-variable
control standard. Alternatively, the
Sierra Club maintained that if dry
technology is considered demonstrated
the rscord supports a more stringent
minimum control level. With respect to
-the regulatory analysis, the petition- .
charged that faulty analytical
methodology and assumptions led Ihe
Agency- to erroneous conclusions about
the impacts of the promulgated standard
relative to the more stringent uniform or
full control alternative. They suggested
that analysis performed using proper
assumptions would support adoption of
a uniform standard.
In support of a less stringent minimum.
reduction requirement, the UARG
petition presented a regulatory analysis
which was prepared by their consultant,
National Economic Research Associates
(NERA). Based on this study, UARG-
argued that a 50 percent minimum
requirement would be superior in terms
of emissions, costs, and energy impacts.
Finally, they argued that a lower percent
reduction would provide greater . .
opportunity to develop dry SOa
scrubbing technology.
In their petition-KCPL sought either an
elimination of the percent seduction
requirement when emissions are 520
ng/J. (1.2 Hi/million Btu), heat input or
less, or, as an. alternative, a reduction in
the.70 percent requirement.. liLsapportof
their request, KCPL set forth several'
arguments. First, they cited; tha '
economic and energy impacts
associated with the application of ' • _
scrubbing technology on low sulfur:
coals.. Second, they noted that a.
significant portion of sulfur-in tha coal
they plan to burn xvill be removed in the.
fly ash. Finally, they asserted that health
and welfare considerations tlo not
warrant scrubbing of low sulfur coals.
s:rtce their uncontrolled SO* emissions
are less than the emissions allowed by
the standard for high-sulfur coals with
90 percent scrubbing.
The primary basis for the- UARG"and
Sierra Club requests for reconsideration
of the minimum control level was the
Agency's phase 3 economic modeling
analysis (4-3>FR 33602). Because ihd
phase 3 analysis was completed after
the close of the public comment penod,
it is important that tha results of that
study are viewed ia proper perspective
to their role in the Administrator's;
decision. The petitioners implied that
the adoption of the 70 percent variable
control standard was-based solely on
the phase 3- analysis and that the phase
3 analysis was a naw venture by the
Agency, and therefore, the public was
excluded from active participation in the
decision process. This notion is false.
Contrary to! views of the UARG and
the Sierra Club, the phase 3 study did. ,
not mark a significant departure from.
ths Agency's earlier analysis of the
issue of uniform, versus-variable! control.
No new economic modeling conce^s
were introduced nor were any modeling
input assumptions changed from those.
presented ia the- phase 2 analysis.
Instead, the phase 3. study served merely
•(a) to refine the analysis by
incorporating consideration of dry SO»
scrubbing in response to public. -
comments and (b) to facilitate
spi?cification of the final standard. In
effect, phase 3 brought together the
results of an analysis that had-
proceeded, under close public scrutiny
for more than a year. In order to
consider the full ranga of applicability of
dry SOa scrubbing systems, it was
necessary to introduce a new alternative
standard—the variable control standard
E-7
-------
0216 Federal Register / Vol. 45, No. 28 / Wednesday, February 6, 1980 / Rules and Regulations
^* ******"*'*'*g*"*^*B>***'''^**|?!^y^^rTO**Mn' """"" """ JOTMmt-Mm^Mp^piiffltagMlBMiMnMiMCTffl^MmMytMW mmniiin- MiiiMMawMamwBttaMgyawMjwWwiwmMawt wmvoinmmm®is»4ttmtrini*&tiy*eni&ais
-------
Federal Register /Vol. 45, No.26/Wednesday.Februaiy 6. 1980 / Rules and Regulation^
of California Air Resources Board, also
recognized that it was not sufficiently
advanced at this time to be considered.
Instead, they merely recommended that
the standard require plants to set aside
space so that catalytic.ammonia
injection could be added at some future
date (OAQPS-78-1, rV-D-268). In
comparison, dry scrubbing has
undergone extensive testing at pilot
' plants, and there are three full-scale
facilities under construction that will .
begin operation in the 1981-82 period.
With respect to commercial ,
demonstration permits for solvent
.refined coal and fluidized bed
combuMion, the standard merely allows
initial, full-scale demonstration units
some flexibility. Subsequent commercial
facilities will be required to meet the
final standards. In adopting this
provision, the Administrator recognized
that initial full-scale demonstration units
often do not perform to design
specification, and therefore some
flexibility was required if these capital
intensive, front-end technologies were to
' be pursued. On the other hand, the
Agency concluded that more
conventional devices such as dry
scrubbers could be scaled up to
commercial-sized facilities with
. reasonable assurance that the initial
facilities would comply with the-
applicable requirements. In view of this,
the inclusion of dry scrubbing under the
commercial demonstration permit
provision was not appropriate.
Finally, in a letter dated September 17,
1979, to the Administrator, the Sierra
Club submitted additional information
to buttress its argument that dry
scrubbing is not demonstrated
technology. This letter cited EPA's "FGD
quarterly Report" of Spring 1979. The
report indicates that the direct injection
of dry absorbents (such as nahcolite)
into the gas stream may be a
breakthrough, yet it calls for further pilot
plant studies. The*inference the Sierra
Club drew from the article waathat the
EPA technical staff does not believe dry
scrubbing is sufficiently developed to be
considered in the rulemaking. The Sierra
Club failed to recognize that there are
several different dry scrubbing
approaches in different stages of
development. The "FGD Quarterly
Report" does not pertain to the
approach employing a spray dryer and
baghouse with lime absorbent which
- serves as the basis for the
Administrator's finding (EPA-450/3-7Q-
021 at 3-61). -
The Sierra Club also cited an article in
the Summer 1979 "FGD Quarterly
Report" on vendors' perspectives
toward dry scrubbing. In doing so, the
Sierra Club noted^that the article
indicates that vendors expressed an
attitude of caution toward dry scrubbing
which led the Sierra Club to conclude >
that the technology is not available. It ;
should be noted from the article that
only one of the vendors present was
actively engaged in dry scrubbing and
that firm was quite positive in their
remarks. Babcock and Wilcox, who had
conducted spray dryer pilot plant
studies and is pursuing contracts for
full-scale installations, commented that
"while the dry scrubbing approach is
new, the technology is proven."
Economic Modeling,
The Agency's regulatory analysis
concluded that the variable control
standard with a 70 percent minimum
control level would result in equal or
lower national sulfur dioxide emissions
than the uniform 90 percent standard
while having less impact on costs, waste
disposal, and oil consumption (44 FR
33807, middle column and 33608}.. The
Sierra Club petition charged that the
Agency used.an unrealistic model and
faulty assumptions in reaching these
conclusions. The petition alleged that /
utility behavior as predicted by the EPA
model is "incredible" and that this
incredible behavior leads to "the
outlandish notion that stricter emission
controls will lead to more emissions."
The Administrator finds this allegation
to be without merit.
The principle modeling concept being
challenged is whether or not increased
costs of constructing and operating a
new plant (due to increased pollution
control costs) will affect the utility
operate r's decisions on boiler retirement
schedules, the dispatching of plants to
meet electrical demand, and the rate of
construction of new plants. The model
used for the analysis assumed that
utility companies over the long term will
make decisions that minimize the cost of
electricity generation. That is, (1) under -
any demand situation utilities will first
operate their equipment with the lowest
operating costs, and (2) existing
generating capacity will be replaced
only if its operating costs exceed the
capital and operating costs of new
equipment. While political, financial, or
institutional constraints may bar cost-
minimizing behavior in individual cases,
the Administrator continues to believe
that the assumption of such behavior is
the most sound method of analyzing the
impacts of alternative standards.
Under this approach, the model
simultaneously adjusts both the
utilization of existing plants and the
construction schedule of new plants
(subject to Subpart Da) based on the
relative economics of generating
electricity under alternative standards.
Hence, average capacity factors for the
population of n:r,v plants may vary
among standards due to variations in
the mix of base and intermediate loaded
plants which are brought on line in any
one-year. But this does not mean, as
concluded in the Sierra Club petition at
page 8, that the model predicted that
utilities would permit new base-loaded
units to remain idle while they continue
to build s till .more new units.
The petition also alleged that this
modeling concept was introduced in the
phase 3 analysis, which Was completed
after the close of the public comment
period, and hence the modeling
rationale was not subject to public
review. The petition went on to criticize
some of die assumptions in th aiodel
charging that they were not even
mentioned in the record.
The Administrator finds no basis for
the Sip»^a Club's assertion that the
modeling methodology and input ' . •
assumptions were not exposed for
public review. First, the same model
was used for the phase 1, 2, and 3
analyses. The basic model logic was
explained in the preamble to the
September proposal and comments were
solicited specifically on the use of a cost
optimization model for simulating utility
decisions (43 FR 42162, left column).
Secondly, the model's input
assumptions were subjected to broad
review. Assumptions were presented in
the September preamble and in even
greater detail in the consultant's reports
which are part of the record (OAQPS-
78-1, II-A-^2, D-A-90, and H-A-91).
Followirg proposal, the Agency
convened an interagency working group
to review the macroeconbmic model and
the Agency's.input assumptions (44 FR
33604, left column). Members of the
group represented -a spectrum of
expertise and interests (energy,
employment, environment, inflation,
commerce). The group met numerous
times over a period of two months.
including meetings with UARG, NRDC,
and Sierra Club. As a result of the
group's recommendations, the phase 2
analysis was conducted. A full
description of the analysis including
changes to the modeling assumptions
\vas presented at the public hearing and f
a detailed report was put into the record *
(OAQPS-78-1, IV-A-5). For the phase 3
analysis accompanying promulgation,
the only change in modeling
assumptions from phase 2 was the
. introduction of dry scrubbing
technology. Based on the detailed record
4 established, the Administrator
concludes that the Sierra Club had
ample opportunity to analyze and
E-9
-------
K218
i*e°'sler J . V()l- 45' No- 26 / Wednesday. February 6, 1980 / Rules and Regulations
comment on the Agency's analytical
approach and did so by incorporating
the EOF ur.d N'RDC comments into their
January 1979 comments (OAQPS-78-1.
tV-D-828).
The Sierra Club also criticized the
conclusions of the Agency's regulatory
analysis because the assumed oil prices
were too low and the nuclear plant
growth rate was too high. To assist in
evaluating the petitions, two sensitivity
tests were performed on the Agency's
regulatory analysis. Using the phase 3
assumptions as a base, the analysis was
rerun first assuming higher oil prices
and then assuming both higher oil prices
and a lower nuclear growth rate
IOAQPS-78-1, VI-B-16). The studies
addressed the promulgated standard,
the full control option (uniform SO
percent control), and a variable control
standard with a 50 percent minimum
control requirement as recommended by
t'ARG. The predictions for 1995 are
summarized in Tables 2 and 3,
respectively. For comparison, the phase
3 results are repealed in Table 1.'
With respect to energy input
assumptions, the oil prices used by the
Agency for the phase 3 analysis were
tmed on the Department of Energy's
estimate of future crude oil prices. These
estimates are now probably low
becausn of tho 1979 OPEC price increase
which occurred after promulgation of
the standard. For the sensitivity -.
analysis, the following oil prices in 1979
dollars were assumed:
Assumed Oil Prices
(Dollys per Barrel)
Sensitivity Ptiasa 3
analysis
1935..
'1990 _
1995" -' .
25 '
30
38
IB
20
26
These prices were obtained from
conversations with DOE's policy
analysis staff. The prices may appear
low in1 comparison to the example of
$41.00 per biursl spot market oil given in
the Sierra Club petition, but the Sierra
Club figure is misleading because .
utilities seldom purchase spot market
oil. The meaningful parameter is the
average refiners' acquisition cost, which
was $21/barrel at the time of this
analysis. The original nuclear capacity
assumptions were based on the
industry's announced plans for new
capacity. For sensitivity testing, these
estimates were modified by excluding
nuclear power plants in the early
planning stages while retaining those
now under construction or for which,
based on permit status, plans appear
firm. The following assumptions of total
nuclear capacity resulted:
Table I.—Sum/nary of 1995 Impacts With Phase 3 Assumptions'
Level of control with 520 ng/J maximum'emission limit
Current Variable con- Variable con- Full
standards trol. 50 pet trol. 70 pet control
minimum minimum
felfconil SOt E«-»l»ionj {.•rtTion tons) .„„...
vita
Hiestmnat Anou»te«-ert««n'a) Cow at SO, BMoctNjn (1978 S/ton)..
let* Casl Caooei1^ (G AI ..,,.,.,. ,..,,,,....,_i
238
11^
8.3
2.8
1.7
. 1.4
1,767
554
20.6
9.7
8.0
1.8
1.1
2.9
9M
1.6
1.745
537
20.5
9.7
S.O
1.7
1.1
3.3
1.036
1.6
1.752
537
20.7
10.1
7.9
1.7
0.9
4.4
1,428
1.8
. 1.761
520
C" P~CC* NycXstr
|3 !3?S| Capacity
(GW)
two..
S12.10 07
1640 I6S
21.CO 228
•S<>* M ta J36C3 lor jteignaSon ol census regions
E-10
-------
Federal Register / Vol. 45, No. 26 / Wednesday, February 8, 1980 / Rules and Regulations
MfflgiaESjcoKaQiaa^aiai'i^^
8219
Table 2.—Summary of 1935 Impacts With Higher Oil Prices'
Levei of control with 520 ng/J maximum emission limit
Current
standards
Variable con-
troi, SO pet
minimum
Variable con-
trol, 70 pet
minimum
Fun
control
National SOi Emissions.(million tons).,
East'.
Midwsst...
West South Central..
West
Incremental Annualized Cost (billions 1978 S)
Incremental Cost of SO, Reduction (1978 S/ton)...__
Oil Consumption (million bb)/day)_ ;_—„,,
Coal Production (million tons) ,,. .-. !
Total Coat Capacity (GW) ™_
23.2
10.9
8.2
2.6
1.6
0.9
1,800
588
19.8
9.1
7.9
1.7
1.1
3.3
967
0.9
1,797
687
19.S
9.1
7.8
1.8
1.0
3.6
977
0.9
1,802
SS7
19.7
9.5
7.8
1.5
0.9
S.O
1,049
0,9
1,832
537_
•With wet ant dry scrubbing and the tallowing energy assumptions:
rear:
1985..
1990..
1995..
Oil prices
(5 1975)
S20.20
24.20
30.70
Nuclear
capacity
-------
322Q Federal Register / Vol. 45, No. 26 / Wednesday, February 6, 1980 / Rules and Regulations
million tons.per.year in contrast to _.
about 3.5 million tons per year under
b'»th She phase 3 and the high oil price
sensitivity projections.
While emission levels were roughly
the same as under the phase 3 energy
assumptions, the relative impacts of the
alternative standards changed
somewhat. National emissions were ;
predicted to be 100,000 tons less under
f«!l control than, under the standard.
Relative to full control, the standard .
was still predicted to reduce emissions
by about 400,000 tons in the East, but on
a national basis this was offset by
emission increases in the other regions.
"WJlh higher oil prices and less nuclear
capacity, the environmental benefit of
full control in the West and West South
Central was greater by about 100,000
tons, but this impact is masked in Table
3 due lo rounding. The variable standard
with a 50 percent minimum control level
resulted in about 400,000 tons per year
more emissions than full control and
about 300,000 tons per year more than
•he standard.
The total cost of all the alternatives
increased due to the increased coal,
capacity. Relative to the standard, the
cost of the 50 percent variable control
standard remained about the same. The
full control standard, however, was
s'gnlficantJy more expensive. The _
marginal cost of full control (relative to
(he standard] increased from Sl.l billion
under the phase 3 energy assumptions to
Sl.8 billion.
Energy impacts were about the same
as those predicted in the high oil price
Sensitivity runs. Oil consumption was
;,UI1 predicted at about 900.000 barrels
per day under all alternative standards.
Coal production under all alternatives
increased by about 100 million tons per
\ ear.
Even considering the uncertainty of
fu'ura oil prices and nuclear capacity,
the Administrator found no basis for
convening a proceeding on the modeling
issue. The sensitivity runs did not show
s'^nificanl changes in the relative
impacts of the alternatives. Under the
sensitivity test with both high oil prices
i;id slowed nuclear growth, full control
for the first time showed lower
emissions nationally than the standard.
But the cost of this additional 100,000
tons of control was estimated at $1.8
b!i'ion, which represents more than a 40
percent increase in the incremental cost
of th« standad (Table 3). The principal
environmental benefit of full control
would be felt in the \Vest and West
§uuth Central. Through case-by-case
new source review ample authority
exists to require more stringent controls
as necessary to protect our pristine
ateas and national parks (44 FR 33584,
left column). As a result, the
Administrator continues to believe that
the flexibility offered by the standard
will lead to the best balance of energy,
environmental, and economic impacts
than either a uniform 90 percent
standard or a 50 percent variable
. standard and hence better satisfies the
purposes of the Act.
On the other side of the modeling
issue, UARG charged that the Agency's
regulatory analysis does not support a
70 percent minimum requirement. The
petition called the Agency's control cost
estimates unrealistic and presented a
macroeconomic analysis which
concluded that a 50 percent minimum
requirement would result in a more
favorable balance of cost, energy, and
environmental impacts.
Response to the UARG petition was
difficult because the UARG position was
presented in two separate reports
submitted at different times, and the two
reports reached different conclusions. In
the formal petition, UARG
recommended 50 percent minimum
control and promised a detailed report
by NERA supporting their position.
When the NERA report arrived six
weeks later, if recommended 30 percent
control. In light of this confusion, it was
decided to revieiv each report
separately based on its own merits, but
•devote primary attention to the 50
percent recommendation. After
reviewing UARG's macroeconomic
analysis, the Administrator finds no
convincing arguments for altering the
conclusion that the 70 percent minimum
removal requirement provides the best
balance of impacts. In the formal
petition, UARG's conclusion that a 50
percent standard is superior was based
on a NERA economic analysis which
adswned that only wet scrubbing
technology was available to utilities. A
detailed analysis of the NERA results •
was not possible because only summary
outputs were supplied in their
comments. But the results of this
analysis seem to coincide with the
Agency's conclusions that there are
energy, environmental, and economic
benefits, associated with standards that
provide a lower cost control alternative
for lower sulfur coals. The problem with
the UARG initial analysis is that it
overlooked the economic benefits of dry
scrubbing.
In recognition of this shortcoming,
UARG presented their estimate of the
costs of dry scrubbing made by Battelle
Columbus Laboratories (UARG petition.
page 25) and then hypothesized without
supporting analysis that "with realistic
cost assumptions the advantages of a
lower percent removal are likely to
increase even further" (UARG petition.
page 27}. Table 4 compares BnUelle's
costs to those used in the EPA
regulatory analysis. The two estimates
are almost the same. More importantly.
the two estimates agree that the cost of
a 70 percent efficient dry system is not
significantly greater than the cost of a 50
percent efficient system, and this
conclusion had important implications
in the specification of the standard.
Based on these comparisons, the
Administrator finds that the UARG
petition supports the Agency's dry
scrubbing cost assumptions and the
finding that no significant cost benefit.
will result from a,standard with a 50
percent minimum control level.
Table 4.—Comparison of UARG and £P4 dry SO,
Scrubbing Costs' (Mills/ktrh)
Percent removal
50 _
70
Inlet sulfur (Ibs
SC^/milliort
Blu)
0.80
2.00
08Q
2,00
UARG
'1.68
«2.t3
1 9?
2.54
EPA '
2.06
2.44
268
2,66
' Wet scrubbing costs range up to S mifls/hwh
= UARG costs bas;xl on 55 percent removal
In their second report, UARG
presented additional economic analyses
by NERA. In those analyes, the impacts
of 30, 50, and 70 percent minimum
control standards were tested assuming
that both wet and dry scrubbing
technology were available. The analyses
were performed with three different sets.
of control cost assumptions—EPA's
costs, Battelle's costs, and an additional
set of costs specified by NERA. The „
report concluded that the 70 percent
standard is superior using EPA's costs
but that under the other cost estimates
the 30 percent standard is better. The
cost effectiveness of alternative
standards (dollars per ton of pollutant
removed) was their principal basis of
evaluation. UARG then alleged that EPA
overestimated the differences in cost
between wet and dry scrubbing and that
this error led to the wrong conclusion
about the impacts of the 70 percent
minimum removal requirement. The EPA
cost assumptions were criticized
primarily because different methods
were used to estimate dry and wat
scrubbing costs. To justify their position,
UARG presented estimates of wet and
dry scrubbing costs developed by
Battelle. UARG believes that Battelle
understand scrubber costs, but that .
Battelle's relationship between wet and
dry scrubbing costs is more accurate
than EPA's (UARG petition, page 7). As
noted above, Battelle agreed with, the
Agency's dry scrubbing costs, but for •
E-12
-------
Federal Register. / Vol. 45, No. 26 / Wednesday. February 6,
wet scrubbing the Battelle costs were
substantially lower than the Agency's.
Typically, when comparing results of
studies, the Agency has detailed
documentation with which to compare
the methods of costing and analysis. In
this case, the Administrator had
documentation for neither the NERA •
costs nor the Battelle costs. The NERA
costs were unreferenced and supported
by neither engineering analysis nor
vendor bids. They assumed that the
capital cost of a dry scrubber is 10
percent less than that for a comparable
wet scrubber and that the operating
costs and energy requirements are the
same for the two systems. The UARG
petition promised a detailed report from
Battelle, but the report was not
delivered. Without a basis for
evaluation, the Battelle and NERA costs
can only be considered as hypothetical
data sets for the purpose of sensitivity .
testing of the economic analysis. They
cannot be considered as new
information on SO2 control costs.
The EPA cost estimates, on the other
hand, have withstood several critical
tests. The PEDCo cost model for wet
scrubbers which was used by EPA was
thoroughly reviewed by Department of
Energy (DOE) consultants, and DOE
concurred with the EPA estimates
through the interagency working group. .
Later, the Agency's costs were again
reviewed in detail against wet scrubber
costs predicted by the Tennessee Valley
Authority's scrubber design mode!.
While the two models initially seemed
to produce divergent results, careful
analysis of the respective costing
methodology showed that for similar
design specifications the two models
produced costs that were very close, the
major difference stemming from
different assumptions about the
construction contingency fee (OAQPS-
7.8-VIV-&-5Q). The Administrator
concluded from these cost comparisons
that the Agency's flue gas .-.--
desulfurlzation cost assumptions are
-reasonable.
The FJ-A dry scrubbing costs were
based primarily on engineering studies
submitted by electric utility companies
and equipment vendors for the full-scale
utility systems now on order or under
• construction. Using these studies, the
EPA cost estimates were made in lull
cognizance of the basic assumptions
. used in the FSDCo wet scrubbing model.
The result was that for economic
modeling purposes (OAQPS-78-1, TV-
A-2S, page B-17) the dry scrubbing cost
estimates in the background document
(EPA 450/5-7G-021, page 3-67) were
increased to reflect similar fuel
parameters, local conditions, and degree
of design conservatism as reflected in
the wet scrubbing tosts. Since care was
taken in aligning these costs, the
Administrator does not accept UARG's
allegation that EPA's costs for wet and
dry scrubbing are invalid because they
were developed on an inconsistent
basis.
Even if EPA accepted UARG's
unsubstantiated cost assumptions, the
NERA sensitivity analyses provided no
new insights nor did they materially
contradict the Agency's basic
conclusions about the standard. Using,
the Battelle costs and NERA's
"alternative scrubber costs" as a range,
NERA predicted that relative to 50
percent minimum control, a 70 percent
standard would reduce national SOi
emissions by an additional 250 to 450 .
thousand tons per year compared to
about 100 thousand tons estimated by
EPA (Table 1). NERA predicted the
additional costs of a 70 percent
minimum standard relative to a 50
percent requirement would be between
$300 million and $400 million per year
compared to $300 million predicted by
EPA. It was only hi moving to 30 percent
control that the NERA results showed a
distinct cost savings ,($600 to $900
million) over tha 70 percent level, but .
the 30 percent standard produced an
additional -700 thousand tons per year of
SO- under both of their control cost
scenarios. Tha Administrator rejects the
30 percent standard advocated by
UARG because the potential cost
savings do not justify the potential
emission increases. In conclusion, the
trade-offs between costs and emissions
shown fay UARG are generally similar to
those predicted by EPA in promulgating
the standard and therefore do not
support a different standard from the 70
percent variable standard adopted.
Other Issues
Kansas City Power and Light
Company sought either an elimination: of
the percent reduction requirement when
emissions are 520 ng/J (1.2 Ibs/milh'on
Btu) heat input or-less or as an
ahornative a reduction in the 70 percent
minimum control requirement. In their
arguments, KCPL cited annualized
control costs for wet scrubbing of $11.4
million and an energy penalty of 70
thousand tons of coal per year to
operate a scrubber. Second, they noted
that 14 percent of the potential SOa
emissions from the coal they plan to
burn will be removed by the fly ash.
Taking these two factors in account,
KCPL computed a cost effectiveness
ratio for a hypothetical 650 MW unit to
be $3,600 per ton of sulfur removed and
concluded that such control was too
expensive. Finally, they .concluded that
scrubbing low-sulfur coals is not
warranted since uncontrolled SO* ».
emissions from their new plants will be
less than the emissions allowed by the
•standard for high-sulfur coals with 90
percent scrubbing.
After careful review, the
Administrator finds that the KCPL
petition provided no legal or technical
basis for reconsidering the final rule.
First, the question of whether a plant
burning low-sulfur coal should be
required to meet the same percentage
reduction requirement as those burning
high-sulfur coal has been a central issue
throughout this decision-making. Since
this issue was raised in the proposal (43
FR 42155, left column), KCPL had ample
opportunity to make their points during
the public comment period. In fact, it
was the recognition of this trade ,ff in
emissions between high-sulfur and low-
sulfur coal that led the Administrator to
first consider the concept of variable
control standards (43 FR 42155, right
column). \A.'hile sulfur removal by fly ash
does not represent best demonstrated
technology for SQj control, sulfur
removal by fuel pretreatment, fly ash,
and bottom ash may be credited toward
meeting the 70 percent requirement.
Second, the KCPL petition does not
allege the requisite procedural 'error that
the standard was based on information
on which they had no opportunity to
comment. Their objections center
primarily on the economic and energy
impacts of wet SO* scrubbing on low-
sulfur coal. These issues were clearly
identified by the Agency in the
background document supporting
proposal (OAQPS-78-1, III-B-3,
Chapters 5 and ?}. Furthermore, the
preamble to proposal specifically
requested comments on the Agency's
assumptions for the regulatory analysis
(43 FR 42162, left column).
Finally, and more importantly» the
major points made by KCPL are not of .
central relevance to the outcome of the
rule because the information presented
does not refute the Agency's data base
on wet scrubbing. Consider the . •
following comparisons to the
assumptions of the EPA regulatory
analysis.
(a) The control costs quoted by KCPL
for a 650 MW unit were $31 million in
capital and $6.2 million in operating
expenses. The EPA assumptions applied
to a comparably sized unit result in $55
million in capital costs and $7 million in
operating expense.
(b) KCPL quoted an energy impact of 8
tons of coal per hour to operate the
scrubbetf Considering their operating
requirement of 460 tons of coal per hour,
the energy penalty of SO3 control is 1.7
E-13
-------
8222 Federal Register / Vol. 45. No. 26 / Wednesday. February 8, 1980 / Rules and Regulations
percent. The Agency's economic model
assigned 2.2 percent.
(n)'iCCPL computed cost effectiveness
of the standard at S3f>00 per ton of sulfur
removed. This figure is baaed on a ,
misunderstanding of the application of
the fly ash removal credit toward the 70
percent removal requirement According
to the standard, the scrubbing
requirement when assuming a, 14 percent
SO2 removal in flyash is 65 percent
rather than 56 percent as calculated by -
KCPL. At 65 percent scrubbing, the cost
per ton of sulfur removed is $3100. This
converts to a cost of $1550 per ton of
sulfur dioxide removed which is similar
to the costs estimated by EPA for low-
suJfurcoal applications (OAQPS-78-1,
M-B-3 and IV-B-14). ,
Thus, the Administrator has already
concluded that energy and economic
costs greater than those cited by KCPL
are justified to achieve the emission
reductions required by the standard.
Conclusions on Minimum Control Level
After carefully xvsighing the
arguments by the three petitioners, the
Administratcrcan find no new
information or insights which are of
central relevance to his conclusions
about the benefits of a variable control
standard with a 70 percent minimum
removal requirement. The Sierra Club
and UARG correctly point out that the
Ajjeucy'jr phase 3 analysis was
completed after the close of the public
comment period and that they were
therefore unable to comment on the final
step of the regulatory analysis. But in
assessing these comments it is important
to put the phase 3 analysis in proper
context with its role in the final
decision. The Administrator's
conclusions about the responses of the
utility industry to alternative standards
ivera not based on phase 3 alone, but a
scries of economic studies spanning
more than a year's effort. These
analyses were performed undeta range
of assumptions of economic conditions,
regulatory options, and flue gas
dcsulfurizalion parameters/The phase 3
analysis was merely a fine tuning of the
regulatory analysis to reflect dry
scrubbing technology. .
No new modeling concepts or
assumptions were introduced in phase 3.
The fundamental modeling concept as
introduced in the September proposal
{•53 FR 42181, right column) has not
changed. The model input assumptions
were the same as those of the phase 2
analysis presented on December 8.1978
{•M FR 5-J834, middle column), and at the
December 12 and 13,19"8, public
hrarlng. Detailed consultants' reports on
the modeling analyses were available
?vr comment before the close of the
public comment period. This public
record provided adequate opportunity
for the public to comment both on the
principal concepts and detailed
implementation of the regulatory
analysis before the close of the public
comment period.
Even though new information was
added to the record after the close of the
comment period, none of the petitions
raised valid objections to this
information or cast any uncertainty that
is germane to the final decision. The
Administrator has very carefully
weighed the petitioners comments on
dry scrubbing and the UARG sensitivity
analysis on pollution control costs. Not
only did the UARG analysis generally
confirm the conclusions of the EPA
regulatory analysis, but it established
that even if dry scrubbing costs vary
substantially, the relative impacts of a '
50 versus 70 percent minimum removal
requirement change very little. The 70
percent s.tandard was estimated to
produce lower emissions for only
slightly higher costs. Differences in cost
effectiveness, which UARG seem to
weigh most heavily, varied by only S2 to
a maximum of $50 per ton of SO»
removed across alternative cost
estimates. In the final analysis none of .
the petitions repudiated the Agency's
findings on the state of development,
range of applicability, or costs of dry
SO2 scrubbing. In light of these findings,
the Administrator finds the information
in the petitions not of central relevance
to the final ruls and therefore denies the
requests to convene a proceeding to ,
reconsider the 70 percent minimum
removal requirement.
///. SOa Maximum Control Level (90
percent reduction of potential SO3
emissions)
Petitions for reconsideration
submitted by the Utility Air Regulatory
Group (UARG) and the Sierra Club
questioned the basis for the maximum
control level of 90 percent reduction in
potential SO- emissions, 30-day rolling
average. The other petitions did not
address this issue. However, in a July 18,
1979, letter, the Environmental Defense
Fund (EDF) requested EPA to review
utilization of adipic acid scrubbing
add! lives'as a basis for a more stringent
maximum control level. An additional
analysis by UARG xvas forwarded to
EPA on January 28,1980. Although it
was reviewed by EPA, a detailed
response could not be prepared in the
three days afforded EPA for comment
prior to the court's deadline of January
31, 1980, for EPA to respond to the
petitions. However, the only issue not -
previously raised by UARG (boiler load
variation) has been addressed by this
response.
With their petition, UARG submitted a
statistical analysis of flue gas
desulfurization (FGD) system test data
which purportedly revealed certain
flaws in the Agency's conclusions. Hie
UARG petition maintained that a
scrubber with a geometric mean
(median) efficiency of 92 percent could
not achieve the standard because of
variations in its performance. UARG
also maintained dial the highest removal
efficiency standard that can be justified
by the Agency's data is 85 percent, 30-
day rolling average. In the alternative.
they suggested that the 90 percent, 30-
day rolling average standard could be
retained if an adequate number of
exemptions were permitted during any
given 30-dgy averaging period. On the
other hand, the Sierra Club questioned
why the standard had been established
;it 90 percent when the Agency had
documented that well-designed,
operated, arid maintained scrubbers
could achieve a median efficiency of 92
percent. la doing so, they argued that a
90 percent, 30-day rolling average
standard was not sufficiently stringent.
After reviewing their petitions, the
Administrator finds that the Sierra Club
and UARG overlooked several
significant factors which were of critical
importance to the decision to
promulgate a SO percent, 30-day rolling
average standard. The Sierra Club
position was based on a
misunderstanding of the statistical basis.
for the standard. The UARG analysis
was flawed because it did not consider
the sulfur removed by coal washing,
coal pulverizers, bottom ash, and fly ash
(hereafter, collectively referred to as"
sulfur reduction credits). Instead the
UARG petition based its conclusions on
the performance of the FGD system
alone. In short, UARG did not analyze
the promulgated standard (44 FR 33582,
center column). Furthermore. UARG
underestimated the minimum
performance capability of scrubbers by
assuming that future scrubbers would
not even achieve the level of process
control demonstrated by the best
existing systems tested by EPA.
EPA has prepared two reports which
re-analyze the sanaa FGD test data
considered in UARG's analysis. One
report identified the important design
and operating differences in the FGD
systems tested lOAQPS-78-1. VI-B-14)
by EPA and the second report provided
additional statistical analyses, of these
data (OAQPS-73-1. VI-B-13). The
results of the EPA analyses showed that:
1. Flue gas desulfurization systems
can achieve a 30-day rolling average
efficiancy between 88 percent and 89
E-14
-------
Federal Register / Vol. 45, No. 26 / Wednesday, February 8, 1980 / Rules and Regulations 8223
percent (base loaded boilers) or
between 86 and 87 percent (peak loaded
boilers) with no improvements in
currently demonstrated process control.
2. Even if a new FGD system attained
only 85 percent efficiency {30-day rolling
•average), a 90 percent reduction in
potential SOa emissions.can be met
when sulfur reduction credits are
'considered.
To clarify the basis for the Agency's
conclusions, the following discussion
reviews the development of information
used to establish the final percent
•reduction standard. Initially, EPA
studied the application of FGD systems x
for the control of SOz emissions from
power plants. As part of that effort,
information which described the status
and performance of FGD systems in the
U.S.. and Japan was inventoried and
evaluated. These evaluations included
the development of design information
on- how to improve the median
efficiency of FGD systems based upon
an extensive testing program at the
Shawnee facility (OAQPS-78-1, II-A-
75). The Shawnee test data and other
data (OAQPS-78-1, II-A-71) on existing
FGD systems were generated by short-
term performance tests. These data did
not define the expected performance
range (the minimum and maximum SOa
percent removal) of state-of-the-art FGD
' systems.
Because a continuous compliance
standard was under consideration,
information about the process variation
of FGD systems was needed to project,
the performance range of scrubber
efficiency around the median percent.
removal level. For the purpose of
measuring process variation, several
existing FGD systems were monitored
with continuous measurement
instrumentation. The selection of FGD
systems to be tested was limited
principally to the few FGD systems
available which were attaining 80 to 90
percent median reduction of high-sulfur
coal emissions. When examining the
results of these tests, it should be
recognized that they do not reflect the
performance of a new FGD system
specifically designed to attain a
continuous compliance standard.
When the percent reduction standard
was proposed. EPA projected the
.performance of newly designed FGD
systems. The projection, referred to as
the "line of improved performance" in
the analysis, was principally based on
the information on how to improve
median system performance (OAQPS-
78-1, IH-B-4). The line showed that
compliance with the proposed standard
(85 percent reduction in potential SO2
emissions, 24-hour average) could be
attained with an FGD system if the only
improvement made relative to an
existing FGD system was to increase the
median efficiency to 92 percent. The
"line of improved performance" did not
reflect the sulfur reduction credits that
could be applied towards compliance
with the proposed standard or the
improvements in process control (less
than 0.289 geometric standard deviation)
that could be designed into a new
facility. While these alternatives were
discussed in detail and included within
the basis for the proposed standard
(OAQPS-78-1, III-B-4), the purpose of
the "line of improved performance" was
to show that even without credits or
process control improvements, the
proposed standard could be met. Upon
proposal, the source owner was
provided a choice of complying with the
percent reduction standard by.(l) an
FGD system alone {85 percent reduction,
24-Jiour standard), or by (2) use of sulfur '
reduction credits together with an FGD
system attaining less than 85 percent
reduction,
After proposal, EPA continued to
analyze regulatory options for
establishing the final percent removal
requirement. On December 8,1978,
economic analyses of these additional
options were published in the Federal
Register (43 FR 57834) for public
comment. In this notice EPA stated that-
Reassessment of the assumptions made in
the August analysis also revealed that the
impact of the coal washing credit had not
been considered in the modeling analysis.
Other credits allowed by the September
proposal, such as sulfur removed by the
pulverizers' or in bottom ash and flyash, were
determined not to be yigmficant when viewed
at the national and regional levels. The coal
washing credit, on the other hand, was found
to have a significant effect on predicated
emissions levels and, therefore, was taken
into consideration in the results presented
here.
This statement gave notice that the
effect of the coal washing credit on
emission levels for the proposed control
options had not been properly assessed
in previous modeling anayses. In the
economic analyses completed before
proposal, the environmental benefits of
the proposed standard were optimistic
because it was assumed that all high-
sulfur coal would be washed, but a
corresponding reduction in the level of
scrubbing needed for compliance was
not taken .into account. This error
resulted in the analyses understimating
the amount of national and regional SO*
emissions that would have been allowed
by the proposed standard. This problem
was discussed at length at the public
hearing on December 12,1978 (OAQPS-
78-1, IV-F-l, p. 11, 22, 28, and 29).
UARG addressed this question of coal
washing in comments submitted in
response to recommendations presented
at the-public hearing by the Natural
Resources Defense Council (OAQPS-78-
1, IV-F-l, p. 65,12-12-78) that the final
standards be based upon the removal of
sulfur from fuel together with the
removal of SOa from flue gases with a
FGD system. In their comments
(OAQPS-78-1, IV-D-725, Appendix A,
p. 23), UARG had three main objections:
(1) All coals are not washable to the •
same degree.
(2) Coal cleaning may not be • • ..• • .
economically feasible. .,
(3) The Clean Air Act and the
Resource Conservation and Recovery
Act may preclude the construction of .
coal washing facilities at every mine. •
EPA has reviewed these comments
again and does not find-that they change
the Administrator's conclusion that
washed coal caa be used in conjunction
with FGD systems to attain a 90 percent
reduction in potential SOs emissions.
First, EPA realizes that all coal is not
equally washable. In the regulatory
analaysis, the degree of coal washing .
was a function of the rank and sulfur
content of the coal. Moreover, because
of tha variable control seals inherent in
. the standard, 75 percent of U.S coal
reserves would require less than 90
percent reduction ih potential SO*
emissions. The remaining 25 percent are
high sulfur coals on which the highest
degree of sulfur removadjby coal
washing are acheived. Second, the
washing assumptions used by the
Agency reflected the percentage of
sulfur removal currently being attained
by conventional coal washing plants in
the U.S. (OAQPS-78-1, IV-D-756).
These washing percentages were
therefore cost-feasible assumptions
because they are typical of current
. washing practices. Finally, the Agency
does not believe that environmental
regulations will prohibit the cleaning of
coal. The Clean Air Act and the
Resource Conservation and Recovery
Act may impose certain environmental
controls, but would not prevent the
routine construction of coal washing
plants. Thus, the Agency concluded that
the basis for the promulgated standard
could be a combination of FGD control
and fuel credits.
Based on these findings, EPA stated '
(44 FR 33582) that the 90 percent
reduction standard "can be achieved at
the individual plant level even under the
most demanding conditions through the
application of flue gas desulfurizalion
(FGD) systems together with sulfur
reductions achieved by currently
practiced coal preparation techniques.
Reductions achieved in the fly ash and
E-15
-------
S224
Federal Register' / Vol. 45, No. 26 / Wednesday, February 6, 1980 / Rules and Regulations
•tumuli »iiiiiiiiiiliiiiiii*i« Mii*MM>^m»iiiiiiiiiii ^iimiiitraMffliiiiMi»iiiiiiM»'iMMaMMmm^^
boltom ash are also applicable". Thus,
FGD systems together with removal of
.sulfur from the fuel was the basis for the
final standard. The standard prohibits
the emission of more than "10 percent of
the potential combustion concentration
(90 percent reduction)." That is, the final
standard requires 90 percent reduction
of the potential emissions (the
theoretical emissions that would result
from comjbustion of fuel in an uncleaned
state), not 90 percent removal by a
scrubber. • :
Since UARG failed to take into
consideration sulfur reduction credits, ,' •
UARG analyzed a more stringent.
standard than was promulgated.
Furthermore,-EPA's review revealed that
while the statistical methodology in the
UARG analysis was basically correct, it
was flawed by UARG's assumption
about the process variation of a new
FGD system. As a result, the statistical
anaysis was improperly used by UARG
to project the number of violations
expected by a new FGD system.
* To elaborate on the variability issue,
page 14 of the UARG petition states:
The range of efficiency variability values
resulting from this analysis represents the
range of efficiency variabilities that can be
expected to be encountered at future FGD
sites.
This assumption artificially inflated
the amount of variability that would
reasonably be expected in a new FGD
system because it presumed that there
xvere no major design and operational
differences in the existing FGD systems -
tested and that the performance of new
systems would not improve beyond that
of systems tested by EPA. To estimate
process variability of new FGD systems,
UARG simply averaged together all data
from all systems tested including
malfunctioning systems (Conesville).
EP.Vs review of these data showed that
there were major design and operating
(Inferences in the existing FGD systems
tasted and that the process control could
hit improved in new FGD systems
(OAQPS-78-1, VI-B-14). Therefore, not
ull of the FGD systems tested by EPA-
were representative of best
c, Ttnonstrated technology for SOS
control.
These major differences in the FGD
systems tested are apparent when the
test reports are examined (OAQPS-78-1,
VI-B-14),One of the tests was
conducted when the FGD systems were
nol operating properly (Conesville]. Two
tests were conducted on regenerative
FGD systems (Philadelphia and
Chicago) which are not representative of
a limo or limestone FGD system.
Another test was on an adipic acid/lime
FGD system (Shawnee-venturi). None of
these tests were representative of the
process variation of well-operated, lime
or limestone FGD systems on a high-
sulfur coal application (OAQPS-78-1,
VI-B-14). . .
. Only three systems were tested when
(1) the unit was operating normally, and
(2) pH control instrumentation was in
place and operational (Pittsburgh,
Shawnee-TCA, and Louisville). Only at
Shawnee did EPA purposely have -
skilled engineering and technician -.-•• •
personnel closely monitor the operation
during the test (OAQPS-78-1, VI-B-14).
Data from these systems best describe
the process control performance of
existing lime/limestone FGD systems.
During the Pittsburgh test, there were
some problems with pH meters. The..
data was separated into Test I (pH
meter inoperative) and Test II (pH meter
operative). During Test I, operators ' •
measured pH hourly with a portable _
instrument (OAQPS-78-1, VI-B-14).
Analysis of these data show low
process variation during each test period
(OAQPS-78-1, VI-B-13). Although the
process variation during the second test
was 10 percent lower, the difference
was not found to be statistically
significant. Data taken during each test
(I and II) are representative of control
attainable with pH controls only. Boiler
load was relatively stable during the
test. Average process variation as
described by the geometric standard
deviation was 0.21 and 0.23,
respectively.
At Shawnee, only pH controls were in
use, but additional attention was given
to controlling the process by technical
personnel. Boiler load was purposely
varied. Geometric standard deviation
was 0.18,-which was similar to that
recorded at Pittsburgh. UARG
acknowledged that careful attention to
control of the FGD operation by skilled
personnel was an important factor in
control of the Shawnee-TCA scrubber
process (OAQPS-78-1, H-D-440, page •
3). It was at the Shawnee test that the
bast control of FGD process variability
by an existing FGD system was
demonstrated (OAQPS-78-1, II-B-13).
The Louisville test appears to
represent a special case. The average
process variation was significantly
higher (0.30 and 0.34 for the two units
tested) than was recorded at the two
other tests (Pittsburgh and Shawnee).
An EPA contractor identified two
factors which potentially could
adversely affect process control at
Louisville (OAQPS-78-1, VI-Brl4). First,
they noted that Louisville was originally
designed in the 1960's and has had
significant retrofit design changes which
could affect process control. Second, the
degree of operator attention given to
process control is unknown. In addition.
UARG showed that an additional factor
which may affect the FGD process
control is boiler load changes. Unlike a
new boiler, the Louisville unit is an
older boiler which has been placed into
peaking service and therefore
experiences significant load changes
during the course of a day. As was the
case with Pittsburgh and Shawnee,
Lpuisville only uses pH controls to
regulate the process. The process ' "
variation was analyzed and the
maximum process variation of the
Louisville system, at a 95 percent
confidence level, was determined to be
0.36 geometric standard deviation
(OAQPS-78-1, VI-B-13). This estimate
of process variation represents a "worst
case" situation since it reflects the •
degree of FGD variability at a peaking
unit rather than on the more easily ;- •
controlled immediate- or base-loaded
applications. , . "
In addition to basing their projections
on nonrepresentative systems, UARG
has also ignored information in a
background information document
(OAQPS-78-1, II-B-4, section 4.2.6) on "
feasible process control improvements
which were currently used in Japan
(OAQPS-78-1, II-I-359). An appraisal of
the degree of process instrumentation
and control in use at the existing FGD
systems tested and a review of the
feasible process control improvements
which can be designed into new FGD
systems was also reviewed (OAQPS-
78-1, VI-B-14). As described in this
review, none of systems tested had
automatic process instrumentation ,
control in operation. All adjustments to
scrubber operation were made by
intermittent, manual adjustments by an
operator. Automatic process controls,
which provide immediate and
continuous adjustments, can reduce the
process control resp'onse time and the
magnitude of FGD efficiency variation.
Even the best controlled FGD systems
tested (the Shawnee FGD system, which
was designed in the 1960's) employed •
only feedback pH process control
systems (OAQPS-78-1, IV-J-20). None
of these existing FGD systems were
designed with the feedforward process
control features now used in Japan
(OAQPS-78-1, H-I-359) for the
automatic adjustment of scrubber make-
up in response to changing operating
conditions. These systems respond to
boiler load changes or the amount of
SO> in the flue gases to be cleaned.
before they impact the scrubbing
system. The use of such systems would
improve the control of short-term FGD
efficiency variation. At the FGD systems
tested, the actual flue gas SO2
E-16
-------
Federal Register / Vol. 45, No. 28 / Wednesday, February 6, 1980 / Rules and Regulations S225
concentration (affected by coal sulfur
content) and gas volume (affected by
boiler load] was not routinely monitored
by the FGD system operators for the
purpose of controlling the FGD
operation as is currently practiced in
Japan (OAQPS-78-1, n-I-359). Thus,
even the best controlled existing •. •
; systems tested were not representative
of the control of process variation that
would be expected in the performance
of new FGD systems to be operated in
the 1980's (OAQPS-78-1, VI-B-14). For
the purpose of describing the range'in
performance of an FGD system using
only feedback pH control and which are
known to have received-close attention
by operating personnel, the data -••" -
recorded at these two existing FGD
systems (Pittsburgh, test II and
Shawnee-TCA) have been used by.EPA
to project the maximum process
variation that would result (0.23
geometric standard deviation) at a 95
percent confidence interval fora base
loaded boiler. The data from Louisville
was used to represent performance of a
peak loaded boiler (0.36 geometric
standard deviation at the 95 percent
confidence level). These values are
conservative because the data collected
at the existing FGD systems tested are
not representative of the lower process
variation that would be expected in
future FGD systems designed with
improved process control systems
(OAQPS^78-1, VI-B-14).
F.PA's statistical analysis of scrubber
efficency is in close agreement with the
UARG analysis when the same process
variation and amount of autocorrelation
was assumed. EPA's analysis showed
about the same autocorrelation effect '
(the tendency for scrubber efficiency, to
follow the previous day's performance)
as UARG's analysis. A "worst-case" 0.7
autocorrelation factor was used in both
analyses even though a more favorable
0.5 factor could have been used based
upon the measured autocorrelation of
the data at the Shawnee-TCA and
Pittsburgh tests. A comparison of the
minimum 30-day average performance
of a FGD system based upon EPA and
UARG process variation assumptions is
givsn-ia Table 5a.
The EPA analysis (OAQPS-78-1. VI-
B-13) summarized in Tables 5a and 5b
shows the median scrubbing efficieny
required to achieve various minimum 30-
day rolling average removal levels
(probability of 1 violation in 10 years).
The three sets of estimates shown are
based on (1) the same process control
demonstrated at Pittsburgh, test II and
loaded, well-operated existing plant
(o-R=0.20 on average and o-g=0.23 at the
95 percent confidence level), (2) the
same process control demonstrated at
Louisville which represents a peak
loaded, existing plant (o-e=0.32 on
average and crg=0.38 at the 95 percent
confidence level), and (3) the poor
process control projected by UARG
(o-g=0.29 on average and Estimates a/a based on probability of only 1 violation in 10 years. Process variation Estimates are based on probability of onty 1 violation in
10 years. Process variation (
-------
8228 Federal Register / Vol. 45, No. 26 / Wednesday, February 6, 1980 / Rules and Regulations
be substantial, are summarized as
follows:
1. Coal washing. On high-sulfur
midwestetn coals that would be subject-
to the 90 percent reduction requirement,
an average of 27 percent sulfur removal
was achieved by conventional coal
washing plants in 1978 (OAQPS-78-1),.,
IV-D-761). Even La Ohio where th 85 percent reduction are sufficient to
attain compliance with the final SOX
percent reduction standard as is shown
in Table 8:
TaMa 6.—Impact of Sulfur Reduction Crsdils
on Rstiuirad FGD Control Efiiciancies to
Attain 90 Percent Overall SO, Reduction
Cotnpfcanco
Option
C
Co*J wisJtiog rtmcvai. p^Hcwt _». 27
fif-ttfcir. By Mh. and botton ash
_ 10
_ 85
SO, tedue'icfi In polor.tial
™___ 90
20.
4
87
90
0
B9
Table 6 illustrates that even if the ,
FGD system attained only 85 percent-'
reduction as UARG has claimed, the 90
percent removal standard would be ' •
achieved (Option A) even if a coal
washing plant attained only 27 percent
reduction in sulfur (the average
reduction reported by the National Coal.
Association for conventional coal
washing plants, OAQPS-78-1, IV-D- '
761). In addition, Table 6 illustrates that
less fuel credit is needed when the FGD
system attains more than 85 percent
reduction (Options B and C). For
example, even if the minimum amount of
coal washing curenfly being achieved
(20 percent in Ohio) is attained, only 87
percent FGD reduction would be
needed. Thus, less than average or only
average sulfur reduction credits (i.e.,
only 8-27% coal washing and 0-10% -
pulverizer, bottom ash and fly ash
credits) would be needed to comply with
the 90 percent reduction standard even
if the FGD system alone only attained 85
to 83 percent control. Moreover, for 75
percent of the nation's coal reserves
which have, potential emissions less
than 260 ng/J (8.0 Ibs/million Btu) heat
input (OAQPS-78-1, IV-E-12, page 13),
less than SO percent reduction in
potential SO* emissions would be .
needed to meet the standard.
The statistical analysis submitted by
UARG does not address the basis (FGD
and sulfur reduction credits) o'f the
standard and therefore does not alter
•the conclusions regarding the
achievability of the promulgated percent
reduction standard. The prescribed level
can be achieved at the individual plant
level even under the most demanding
conditions through the application of
scrubbers together with sulfur reduction
credits.
Finally, UARG's petition (p. 15) states
that the final standard was biased by an
error in the preamble (see table, 44 FR.
33592) which incorrectly-referred to
certain FGD removal efficiencies as
"averages" rather than as geometric
"means" (medians). These removal
efficiencies were properly referred to as
"means" in the EPA test reports. This
discrepancy had no bearing on EPA's
decision to promulgate a 90 percent SO3
standard. Even though UARG claims a
bias waa introduced, their consultant's
report states (see Appendix B,Page 46):
Therefore, even though EPA mistakenly
used the term "average SO» removal" in She
promulgation, it is obvious that when the
phrase "mean FGD efficiency" ia used. EPA Is
correctly referred to the mean (or median) of
the long-normal distribution of (1-eff).
Thus, even though Entropy (UARG's
consultant which prepared their
statistical analysis in Appendix B)
"discovered a discrepancy" as UARG
alleges, they did not reach a conclusion
as UARG has done, that a simple
transcription error in preparation of the
preamble undermined the credibility of
EPA's analysis of the test data. In fact,
the analysis of test data performed by
EPA (OAQPS-78-1, H-B-4) used correct
statistical terminology.
The Sierra Club also submitted a
petition that questioned the promulgated
90 percent, 30-day rolling average
standard. The petition asks "why the
final percentage of removal for 'full
scrubbing* was set at only 90 percent for
a 30-day average" in view of the
preamble to the proposal which
mentions a 92 percent reduction (43 FR
42159). The petition states that "EPA
indicated that 85 percent scrubbing on a
24-hour average was equivalent to 92
percent on a 30-day average." This
statement is a misquotation. The
preamble actually stated that "an FGD
system that could achieve a 92 percent
long-term (30 days or more) mean SO2
removal would comply with the
proposed 85 percent (24-hour average)
requirement" The long-term mean
referred to'is the median value
(geometric mean) of FGD system
performance, not an equivalent
standard. Reference in the preamble
was made to the background
information suppIement-(OAQPS-78-l,
III-B-4) which provided "a snore
detailed discussion of these findings."
The 92 percent removal is described
therein as the median (geometric mean)
of the statistical distribution defined by
the "line of improved performance" hi
Figure 4-1. A median is the middle
number in a given sequence of numbers.
Thus for a sequence of 24-hour or 30-day
rolling average efficiencies, the median
SOj removal (92 percent) is a level at
which one-half of the. 30-day rolling .
average FGO*j'?i,tp.in.P-ffir.'.aw.rjfis.,wo,uljd.
be higher and one Jialf would be lower.
Since one-half of the expected removal
efficiencies would ba lower than the 92
percent median, a standard could not be
set at that level. The standard must
recognize the range of 30-day rolling -
average FGD efficiencies that would be
expected. The petition is based upon a
misconception as to the meaning of the
92 percent value (a median) and is
therefore not new information of central
relevance to this issue.
The Environmental Defense Fund'
requested that EPA consider the
relevance of the lime/limestone-adipic
acid tests at Shawnee to this
rulemaking. Adipic acid has been found
to increase FGD system performance by
limiting the drop hi pH that normally
occurs at the gas/liquid interface during
SO» absorption. Test runs at Shawnee
E-18
-------
Federal Reglstec / Vol. 45, No.
showed increased FGD performance (in
one test series She efficiency increased
from 71 percent to 93 percent) with no.
apparent adverse impact upon FGD
system operation.
EPA agrees that use of adipic acid
additive in lime/limestone scrubbing --
solutions appears very promising and is
currently planning a full-scale FGD
system demonstration. Several
important areas are to be evaluated in
the EPA test program. The handling and
disposal characteristics of waste sludges
from the scrubber must be evaluated to
see that adipic acid does not affect
control of leachates into groundwater. In
addition, the consumption rate of-adipic
acid by the FGD system and its ultimate
disposition must be evaluated.
Furthermore, tests must.be conducted to
. show whether or not the concentration
of adipic acid in FGD system sludge
po=es significant environmental
• problems. In the absence of such data,
EPA does not believe it prudent to
include adipic acid ais a basis for the
current revised standard.
IV. Particulata Matter Standards
Only one of the four petitions for
reconsideration raised issues concerning
the particulate matter standard. In their
petition, the Utility Air Regulatory
Group (UARG) argued principally that
baghouse technology was not -•;:'
demonstrated on large coal-fired utility
boilers and that the 13 ng/J (0.03 lb/
million Btu) heat input standard could
not be achieved at reasonable costs •
with electrostatic precipitators on low-
sulfur coal applications. They also noted
that emission test data on a 350 MW
baghouse application was placed in the
record after the close of the comment
period. In response to these data, UARG
.presented operating information on
baghouse systems obtained from, two
coal-fired installations. In addition they
restated arguments that had been raised
in their January 1979 comments
concerning EPA's data base and the
potential effects of NO, and SO2
emission control on particulate
emissions.
In reaching his decision that baghouse
technology is adequately demonstrated,
the Administrator took into account a
number of factors. In addition to the
emission test data and other technical
information contained in the record, he
placed significant weight on the fact that
at least 26 baghouse-equipped coal-fired
electric utility steam generators were
operating prior to promulgation of the
standard and that 28 additional units
were planned to start operation by the
end of 1982. He also noted that some of
the utility companies operating
baghouses on coal-fired steam •'
generators were.ordering more
baghouses and that none of them had -
. announced plans to decommission or
retrofit a baghouse controlled plant
because of operating or cost problems.
The Administrator believed that this • :
was a strong indication that some
segments of the utility industry believe
that baghouses are practical, •
economical, and adequately ••
demonstrated systems for control of
particulate emissions. These electric
utility baghouses are being applied to a
wide range of sizes of steam generators
and to coals of varying rank and sulfur
content. The Industrial Gas Cleaning
Institute, speaking for the manufacturers -
of.baghouses submitted comments
(OAQPS-78-1, IV-D-247) confirming
that baghouses are adequately
demonstrated systems for control of
particulate emissions from coal-fired
steam-electric generators of ail sizes and
types. • •
In the proposal, EPA acknowledged
that large baghouses of the size that
would typically be used to meet the
standard had only been recently
activated. Further, the Agency
announced that it planned to test a 350
MW unit (43 FR 42169, center column).
The validated test data from this unit,
located at the Harrington Station,
demonstrated that the standard could be
achieved at large facilities (OAQPS-78-
1, V-B-1, page 4-1). The Agency also
became aware .that the operators of the
facility were encountering start-up
problems. After carefully evaluating the
situation, the Agency concluded that the
problems were temporary in nature (44
FR 33585, left column). ,
Furthermore, Appendix E of the
UARG petition supports the Agency
finding. According to .Appendix E, the ,.
start-up problems experienced at
Harrington Station (Unit #2) have not
affected unit availability nor have they
altered the utility's plans for equipping
another large coal-fired steam generator
at the site (Unit #3) with a baghouse.
Appendix E noted, "The company feels
that the baghouse achieved an
availability equal to that of the
electrostatic precipitator installed in
unit 1" (UARG petition, Appendix E,
page 2). The Appendix also examined
two retrofit baghouse installations on
boilers firing Texas lignite at the
Monticello Station (Unit #1 and Unit
#2). While the first unit that came on
line experienced problems, Appendix E
notes, "Since the start-up of Unit 2. bag
filter, the baghouse has been operational •
at all times the boiler was on Una (due
to the solution of the majority of the
problems associated with Unit 1
baghouse)" (UARG petition, Appendix
E, page 5). These findings served to
reinforce the Agency's conclusion that
problems encountered at these initial
installations are .correctable.
Based on the Harrington and
Monticello experience, UARG-
maintained that EPA did not properly
consider the cost of activating and
maintaining a baghouse. Contrary to
UARG's position, the cost estimates -
developed by EPA provide liberal
allowances for start-up and continued
maintenance. For example, the Agency's
cost estimates for a baghouse for a 350
MW power plant provided over $1.4
million for start-up and first year -
maintenance of which $440,000 was
included for bag replacement (OAQPS-
78-1, II-A-84 and VI-B-12). For
subsequent years, $710,000 per year was
allowed for routine maintenance of
which $440,000 was included.for bag
replacement. In comparison, the UARG
petition indicated that bag replacement
costs during the first year of operation of
the baghouse at the Harrington Station
.{350 MW capacity) would be $250,000
and the bag replacement costs at the
two Monticello baghouse units (610 MW
capacity total) would total about - . -
$542,000. From the information provided
by UARG, it appears that the Agency
has fully accounted for any potential
costs that may be incurred during start-
up or annual maintenance.
UARG further maintained'that higher
. pressure drops encountered at these
initial installations would increase the
cost of power to operate a baghouse
beyond those estimated by the Agency.
The Administrator agrees that if higher
.prsssurs &cpc 4iJC'.3isasaai2£sd 5®.. : • .
increase in cost will be incurred.
However, even assuming that the.
pressure drops initially experienced at
the Harrington and Monticello Stations
occur generally, the annual cost will not
increase sufficiently to.affect the .
Administrator's decision that the
standard can be achieved af a
reasonable cost. For example, the
increase in pressure drop reported by
UARG (UARG petition, page 43) at the
Harrington station would result in a cost
penalty of about $191,000 per year,
which represents only a 4.5 percent ' • •
increase in the total annualized
baghouse costs projected by EPA
(OAQPS-78-1, II-A-64; page 3-18) and
E-19
-------
{1228 Federal Register / Vol. 45, No. 26 / Wednesday, February 8, 1980'/ Rules and Regulations
Itiiti than one percent increase in
relation to utility operating costs. It
^should be reported, however, that as a
result of corrective measures taken at
Harrington station since start-up, the
operating pressure drop reported by
UARG has been reduced. If the pressure
drop stabilizes at this improved level 2
kilcpascals (8 inches HaO) rather than
the 2.75 kilopascals (11 inches H2O]
suggested by UARG the $191,000 cost
penalty would be reduced by some
S80.000 per year (OAQPS-78-1, VI-B-11
and UARG petition, page 43).
UARG also maintained that a period
longer than 180 days after start-up is
required to shake down new baghouse
installations, and that EPA should
r.mfsnd 40 CFR 60.8, which requires
compliance to be demonstrated within
180 days of start-up. UARG based these
comments on the experience at the
Harrington and Monlicello Stations. It is
important to understand that 40 CFR
6Q.B only requires compliance with the
emission standards within 180 days of
start-up and does not require, or even
suggest, that the operation of the facility
be optimized within that time period.
Optimization of a system is a continual
process based on experience gained
with time. On the other hand, a system
may be fully capable of compliance with
the- standard before it is fully optimized.
In the case of the Harrington station
the initial performance test was
conducted by the utility during October
1978 (which was within four months of ,
start-up). The initial test and a
subsequent one were found however, to
b_% invalid due to testing errors and
Ihcrefore it was February 1979 before
valid test reauits were obtained for the
Harrington Unit (OAQPS-78-1, IV-B-1,
page 42). This lest clearly demonstrated
achievement of the 13 ng/J (0.03 lb/
million Btu) heat input emission level.
These results were obtained even
'.hough the unit was still undergoing
operation and maintenance refinements.
With respect to the Monticello station,
UARG reported that no actual
performance test data are available
(UARG petition. Appendix E, page 6).
UARG also maintained that
Iwghouses are not suitable for peaking
u«!ts because of the high energy penalty
associated with keeping the baghouse
above the dew point. EPA recognizes
thiit baghouses may not be the best
control device for all applications. In
t!»03C instances where high energy
penalties may be incurred in heating the
bnuhouse above the dew point, the
utility would have the option of
employing an electrostatic precipitator.
However, some utilities will be using
b.ighouites for peaking units. For
example, the baghouse control system
on four subbituminous, pulverized coal-
fired boilers at the Kramer Station have
been equipped with baghouse preheat
systems and that station will be placed
in peaking service in the near future
(OAQPS-78-1, VI-B-10).
UARG also argued that it may be
necessary to install a by-pass system in
conjunction with a baghouse to protect
the baghouse from damage during
certain operation modes. The use of
such a system during periods of start-up,
shutdown, or malfunction is allowed by
the standard when in keeping with good
operating practice.
The UARG petition implied that the
test data base for electrostatic
precipitator systems (ESP) is inadequate
for determining thai such systems can
meet the standard. Contrary to UARG's
position, the EPA data base for the
standard included test data obtained
under worst-case conditions, such as (ij
when high resistivity ash was being
collected, (2) during sootbiowing, and (3)
when no additives to enhance ESP
performance were used (OAQPS-78-1,
III-B-1, page 4-11 and 4-12). Even when
all of the foregoing wqrst-case
conditions were incurred
simultaneously, particulate matter
emission levels were still less than the
standard. It should also be understood
that none of the ESP systems tested
were larger than the model sizes used
for estimating the cost of control under
worst-case conditions.
The UARG petition also questioned
the Administrator's reasoning in failing
to evaluate the economic impact of
applying a 197 square meter per actual
cubic meter per second (1000 ft -flOOO
ACFM) cold-side ESP to achieve the
standard under adverse conditions such
as when firing low-sulfur coal. The
Administrator did not evaluate the
economic impact of applying a large,
cold-side ESP because a smaller, less
costly 128 square meter per actual cubic
meter per second (650 ft -/MOO ACFM) •
hot-side ESP xvould typically be used.
The Administrator believed that it
xvould have been non-productive to
investigate the economics of a cold-side
ESP when a hot-side ESP would achieve
the same level of emission control at a
lower cost.
The UARG petition also suggested
that hot-side ESP's are not always the
best choice for low-sulfur coal
applications. The Administrator agrees
with this position. In some case, low-
sulfur coals produce an ash which is
relatively easy to collect since flyash
resistivity is not a problem. Under such
conditions it would be less costly to
apply a cold-side ESP and therefore it
would be the preferred approach.
However, when developing cost impacts
of the standard, the Agency focused on
typical iow-sulfur coal applications
which represents worst case conditions,
and therefore assessed only hot-side
precipitators.
The UARG petition suggests that in
some cases the addition of chemical
additives to the flue gas may be required
to achieve the standard with ESPs, and
the Agency should have fully assessed
the environmental impact of using such
additives. The Administrator, after
assessing all available data, concluded
that the use of additives to improve ESP
performance would not be necessary
(OAQPS-78-1, III-B-1, page 4-11).
Therefore, it was not incumbent upon
EPA to account for the environmental
impact of the use of additives other than
to note that such additives could
increase SO3 or'acid mist emissions. In
instances where a utility elects to
employ additives as a cost saving
measure, their potential effect oa the
environment can be assessed on a case-
by-case basis during the new source
review process.
UARG also maintained that there are
special problems xvith so.me low-sulfur
coals that would preclude the use of hot-
side ESPs and attached Appendix F in
support of their position. Review of
Appendix F reveals that while the
author discussed certain problems
related to the application of hot-side
ESPs on some western iow-sulfur coal,
he also set forth effective techniques for
resolving these problems. The author
concluded, "The evidence of more than
11 years of experience indicates that hot
precipitators are here to stay and very
likely their use on all types of coal will
increase." .,
UARG also argued that the data base
in support of the final particulate
standard for oil-fired steam generatir.g
units was inadequate. The standard is
based on a number of studies of
particulate matter control for oil-fired
• boilers. These studies were summarized
and referenced in the BID for the
proposed standard (OAQPS-78-1, HI-B-
1, page 4-39). These earlier studies
(Control of Particulate Matter from Oil
Burners and Boilers. April 1976, EPA-
4."0/3-7G-C05; and Particulate Emission
Contra! Systems for Oil-fired Boilers,
December 1974, EPA-450/3-74-063)
support the conclusion that ESP control
systems are applicable to oil-fired steam
generators and that such emission
control systems crm achieve the
standard. The achievability of the
sta,ndnrd was also confirmed by the
Hawaiian Electric Company, a firm that
xvould be significantly affected by the
standard .since virtually ali their new
E-20
-------
Federal Register / Vol. 45, No, 28 / Wednesday, February 6. 1980 / Rules and Regulations 8229
jenerating capacity will be oil-fired due
o their location. In their comments the
company indicated, "Hawaiian Electric
Company supports the standards as '
proposed in so far as they impact upon
he electric utilities in Hawaii" ••
[OAQPS-7B-1, IV-D-159).
UARG'also argued that the
Administrator had little or no data upon
which to base a conclusion that the
particulate standard is achievable for
ignite-fired units. In. making this
assertion, UARG failed to recognize that
he Agency had extensively analyzed •
ignite-fired units in 1976 and concluded
that they could employ the same types '
of control systems as those used for
other coal types (EPA-450/2-76-030a.
jage II-29). Additionally, review of the
.iteratura and other sources revealed no
new data that would alter this finding
Some of the data considered includes
OAQPS-78-1, II-I-59, II-I-312, and II-I-
322) and the Agency continues to
aelieve ihat the emission standards are
achievable when firing all types of coal
including lignite coal. UARG has not
provided any information during the
comment period or in their petition
which would suggest any unique
problems associated with the control-of
particulate matter from lignite-fired
units. , .•
The UARG petition alleged that the •
Administrator did not take into account
•the effect of NOX control in conjunction
with promulgation of the particulate-
standard. In developing the N'OS
standard, the Administrator assessed'
the possibility that NOX controls may •
increase ash combustibles and thereby
affect the mass and characteristics of
particulate emissions. The •
Administrator concluded, however, that
the NO* standard can be achieved
without any increase in ash
combustibles or any significant change
in ash characteristics and therefore •
there would be. no impact on the
particulate standard (OAQPS-78-1, III-
B-2. page &-14).
UARG also raised the issue of sulfate
carryover from the scrubber slurry and
its potential effect on particulate .
emissions. EPA initially addressed this
issue at proposal and concluded that
with proper mist eliminator design and
maintenance, liquid entrainment can be
controlled to an acceptable level [43 FR
42170, left column). Since that time, no
new information has been presented
that would lead the Administrator to
reconsider that finding.
In summary, UARG failed to present
any new. information on particulate •
matter control that is centrally relevant
to the outcome of the rule.
V, NOt Standards v,
The Utility Air Regulatory Group
(UARG) sought reconsideration of the
NO, standards. They maintained that
the record did not support EPA's
findings that the final standards could
be achieved by all boiler types, on a
variety of coals,-and on a continuous
basis without an unreasonable risk of
adverse side effects. In support'of this
position, they argued that while EPA's
short-term emissions data provided
insight into NO* levels attainable by
utility boilers under specified conditions
during short-term periods, they did not
sufficiently support EPA's standards
based on continuous compliance.
Further, they maintained that the
continuous monitoring data relied on by
the Agency does not support the general
conclusions that all boiler types can
meet the standards on a variety of coals
under all operating conditions. They
also argued that the Agency failed to
collect or adequately analyze data on
the adverse side effects of Iow-NOx
operations. Finally,-they contended that
vendor guarantees have been shown not
to support the revised standards. The
arguments presented in the petition
were discussed in detail in an
accompanying report prepared by
UARG's consultant.
. In general, the UARG petition merely
reiterated comments submitted in
January 1979. Their arguments
concerning short-term test data, the
potential adverse side effects of low-
NOX operation, and manufacturer's
guarantees did not reflect hew
information nor were they substantially
different from those presented earlier.
For example, hi their petition, UARG
asserted that new information received
at the close of comment period revealed
that certain data EPA relied upon to
conclude that Iow-NOx operations do
not increase the emissions of polycyciic
organic matter (POM) are of
questionable validity (UARG petition,
page 56). This comment repeats the
position stated in UARG's January 15,
1979, submittal (QAQPS 78-1, IV-D-611,
attachment—KVB report, January 1979,
page 86). More importantly, UARG
failed to recognize that EPA did not rely
on the tests in question and that the
Agency noted in the BID for the
proposed standards (OAQPS-78-1, III-
B-2, page 6-12) that the data were
insufficient to draw any conclusion on
the effects of modern, low-NOx Babcock
and Wilcox burners on POM emissions.
Instead, EPA based its conclusions in
regard to POM on its finding that
combustion efficiency would not
decrease during low-NO* operation and
therefore, there would not be an
increase in POM emissions (43 FR 42171.
left column arid OAQPS-78-1, Iii-B-2,
page 9-6).
Similarly, UARG did not present any
new data in regard to boiler tube
corrosion. They merely restated the
arguments they had raised in their
January 1979 comments which
questioned EPA's reliance on corrosion
test samples (coupons). EPA-believes
that proper consideration has been
given to the corrosion issues and
substantial data exist to support the
Administrator's finding that the final
requirements are achievable without
any significant adverse side effect (44
FR 33602, left column), In addition.
UARG also maintained that the Agency
should explain why it dismissed the 190
ng/J (0.45 Ib/million BtuJ heat inpi•' NO*
emission limit, (44 FR 33602, right
column) applicable to power plants in
New Mexico. In dismissing the
recommendation that the Agency adopt
a 190 ng/J . ussion limit, the
Administrator noted that the only
support for such an emission limitation
was in the Form of vendor guarantees.
In relation to vendor guarantees,
UARG maintained in their January
comments and reiterate in their petition
that EPA should not rely on vendor
guarantees as support for the revised
standards. EPA cannot subscribe to
UARG's narrow position. While vendor
guarantees alone would not provide a
sufficient basis for a new source
performance standard, EPA believes
that consideration of vendor guarantees
when supported by other findings is
appropriate. In this instance, the vendor
guarantees served to confirm,EPA
findings that the boiler manufacturers
possess the requisite technology to
.achieve the final emission limitations.
This approach was described by Foster
Wheeler in their January 3,1970, letter
i y Y ft r*f-» r^-»> •» *S»-^c« T--* " **r •*•> <•.-.-
lO w«-nvv*,' \*JXitt^nrcr no—I, *•»—u—Oj l,
attachment—KVB report, January 1979,
page 119) that states, "When a
government regulation, which has a
major effect on steam generator design.
is changed it is unreasonable to judge •
the capability of a manufacturer to meet
the new regulation by evaluating
equipment designed for the older less
stringent regulation."
This observation is also germane to
the arguments raised by UARG with
respect to EPA data on short-term
emission tests and continuous
monitoring. In essence, UARG
maintained that the EPA data base was
inadequate because boilers designed
and operated to meet the old 300 ng/j
(0.7 Ib/million BJu) heat input limitation
under Subpart D have not been shown
to be in continuous compliance with the
. E-21
-------
8230 Federal Register / Vol. 45, No. 26 / Wednesday, February Q, 1980 / Rules and Regulations
new standard under Subpart Da. While.
this statement is true, these units, which
were designed and operated to meet the
old standard, incurred only five
exceedances of the new standards on a •
monthly basis. Moreover, a review of
the available 34 months of continuous
monitoring data from six utility boilers
revealed that they all operated well
below the applicable standard (OAQPS-
70-1, y-B-Z).
In addition, UARG argued that the
available continuous monitoring data
demonstrated that the Agency should .
not have relied on short-term test data.
Citing Colstrip Units 1 and 2, they noted
that less than one-third of the 30-day
averags emissions fell below the units'
performance test levels of 125 ng/J (0.29
Ib/million Btu) heat input and 165 ng/J
(0.38 Ib/million Btu] heat input,'
respectively. They further maintained
that this had not been considered by the
Agency. In fact, the Administrator
recognized at the time of promulgation
that emission values obtained on short-
term tests could not be achieved
continuously because of potential
adverse side effects and therefore
established emission limits well above
the values measured by such tests (44
PR 42171, left column]. In addition, EPA
took into account the emission
variability reflected by the available
continuous monitoring data when it
established a 30-day rolling average as
the basis of determining compliance in
the standards (44 FR 33538, left column).
UARG also maintained in their
petition that EPA should not rely on the
Colstrip continuous monitoring data
because it was obtained with uncertified
monitors. The Administrator recognized
that the Colstrip data should not be
relied on in absolute terms since
monitors were probably biased high by
approximately 10 percent (OAQPS 78-1,
HI-B-2, page 5-7]. EPA's analysis of
data revealed, however, that it would be
appropriate to use the data to draw
conclusions about variability in
emissions since the shortcoming of the
Colstrip monitors did not bias such
findings. This data together with data
obtained using certified continuous
monitors at five other facilities (OAQPS
78-1, V-B-1, pays 5-3) and the results
from 30-day test programs (manual tests
performed about twice per day) at three
additional plants (OAQPS 78-1, II-B-62
and II-B-70) enabled the Administrator
to conclude that emission variability
under low-NOx operating conditions
was small and therefore the prescribed
emission levels 'are achievable on a
continuous basis.
UARG argued that since the only
continuous monitoring data available
was obtained from boilers manufactured
by Combustion Engineering and on a
limited number of coal types, the
Agency did not have a sufficient basis
for finding that the standards can.be
achieved by other manufacturers or
when other types of coals are burned.
The Administrator concluded after
reviewing all available information that
the other three major boiler . .:.
manufacturers.can achieve the same .
level of emission reduction as
Combustion Engineering with a similar
degree of emission variability (43 FR.
42171, left column and 44 FR 33586,
middle column). This finding was
confirmed by statements submitted to
UARG and EPA by the other vendors
that their designs could achieve the final
standards, although they expressed
some concern about tube wastage
potential (OAQPS-78-1, IH-D-611,
attachment-KVB report, pages 116-121
and IV-D-30). EPA has considered tube
wastage (corrosion) throughout the
rulemaking and has determined that it
will not be a problem at the NOX
emission levels required by the
standards (44 FR 33602, left column).
With respect to different coal types, the
Agency concluded from its analysis of
available data that NOX emissions are
relatively insensitive to differing coal
characteristics and therefore other coal
types will not pose a compliance
problem (43 FR 42171. left column and
OAQPS-78-1, IV-B-24). UARG did not
submit any data to refute this finding.
UARG also.argued that the continuous
monitoring data should have been
accompanied by data on boiler
operating conditions. EPA noted that the
data were collected during extended
periods representative of normal
operations and therefore it reflected all
operational transients that occurred. In
piirticular, at Colstrip units 1 and 2 more
than one full year of continuous
monitoring data was analyzed for each
unit. In view of this, EPA believes that
the data base accurately reflects the-
dcgree of emission variability likely to
be encountered under normal operating
conditions. UARG recognized this in
principle in their January 15 comments .
(Part 4, page 15) when they stated that
"continuous monitors would measure all
variations in NOX emissions due to
operational transients, coal variability,
pollution control equipment degradation,
etc."
In their petition, UARG restated their
January 1979 comments that EPA's
short-term test data \vere not
representative and therefore should not
serve as a basis for the standard. As
noted earlier, EPA did not rely
exclusively on short-term test data in
setting the final regulations, in addition,
contrary to the UARG claim, EPA
believes that the boiler test
configurations used to achieve low-NOx
operations reflect sound engineering
judgement and that the techniques
employed are applicable to modern
boilers. This is not to say that the boiler
manufacturers may not choose other
approaches such as low-NOx burners to
achieve the standards. While
recognizing that EPA's test program was
concentrated on boilers from one
manufacturer, sufficient data was
obtained on the other major
manufacturers' boilers to confirm the
Agency's finding that they would exhibit
similar emission characteristics (44 FR
33586, left column). Therefore, in the
absence of new information, the
Administrator has no basis to
reconsider his finding that the
prescribed emission limitations are
achievable on modern boilers produced
by all four major manufacturers.
VI. Emission Measurement and
Compliance Determination
The Utility Air Regulatory Group ,
(UARG) raised several issues pertaining
to the accuracy and reliability of
continuous monitors used to determine
compliance with the SO2 and NOX
standards. UARG particularly
commented on the data from the
Conesvilie Station. In addition, they also
maintained that the sampling method for
particulates was flawed. With respect to
compliance determinations, UARG
maintained that the method for
calculating the 30-day rolling averages
should be changed so that emissions,
before boiler outages are not included
since they might bias the results. In
addition, UARG argued that the
standards were flawed since EPA had
not included a statement as to how the
Agency would consider monitoring
accuracy in relation to compliance
determination. With the exception of the
method of calculating the 30-day rolling
average and the comments on the
Conesvilie station, the petition merely
reiterated comments submitted prior to
the close of the public comment period.
As to the reliability and durability of
continuous monitors, information in the
docket (OAQPS-7a-l, II-A-88, IV-A-20,
IV-A-21, and IV-A-22) demonstrates
that on-site continuous monitoring
systems (CMS) are capable and have
operated on a long-term basis producing
data which meet or exceed trre minimum
data requirements of the standards.
In reference to the Conesyille project,
UARG questioned why EPA dismissed •
the continuous monitoring results since
it was the only long-term monitoring
effort by EPA to gain instrument
E-22
-------
Federal Register /Vol. 45, No. 26 / Wednesday. February 6, 1980 / Rules and
operation experience. UARG maintained
that this study showed monitor
degradation over time and that it was
representative of state-of-the-art
monitoring system performance. This
conclusion is erroneous. EPA does not
consider the Conesville project adequate
for drawing conclusions about monitor
reliability because of problems which
occurred during the project. •
To begin with, UARG is incorrect in
suggesting that the goal of the project
was to obtain instrument operating
experience. The primary purpose of the
project was to obtain 90 days of . •
continuous monitoring data on FGD
system performance. Because of
intermittent operation of the steam
generator and the FGD system, this
objective could not be achieved. As the
. end of the 90-day period approached, a
decision was made to extend the test
duration from three to six months. The
intermittent system operation continued.
As a result, when the FGD outages were
deleted from the total project time of six
months, the actual test duration was
similar to those at the Louisville,
Pittsburgh, and Chicago tests and did
not, therefore, represent an extended
test program.
EPA does not consider the Conesville
results to be representative of state-of-
the-art monitoring system performance.
Because of the intermittent operation
throughout the test period (OAQPS-78-
1, IV-A-19, page 2), it became obvious
that the goals of the program could not
be met. As a result, monitoring system
maintenance lapsed somewhat. For
example, an ineffective sample -
conditioning system caused differences
in mcnitor and reference method results
(OAQPS-78-1, IV-A-20, page 3-2). If the
. EPA contractor had performed more
rigorous quality assurance procedures,
such as a repetition of the relative
accuracy tests after monitor.
maintenance more useful results of the
monitor's performance would have been
obtained. Thus, the Conesville study re-
emphasized the need for periodic
comparisons of monitor and reference
method data and the inherent value of
sound quality assurance procedures.
The UARG petition suggested that the
standards incorporate a statement as to
how SPA will consider monitoring
system accuracy during compliance
determination. More specifically, UARG
recommended that EPA define an error
band for continuous monitoring data
and explicitly state that the Agency will
take no enforcement action if the data
fall within the range of the error band.
The Agency believes that such a-
.provision is inappropriate. Throughout
this rulemsking, EPA recognized the
need for continuous monitoring systems
to provide accurate and reproducible
data. EPA also recognized that the
accuracy of a CMS is affected by basic
design principals of the CMS and by
operating and maintenance procedures.
For these reasons, the standards require
that the monitors meet (1) published
performance specifications (40 CFR Part
60 Appendix B) and (2) a rigorous
quality assurance program after they are
installed at a source. The performance
specifications contain a relative
accuracy criterion which establishes an
acceptable combined limit for accuracy
and reproducibility for the monitoring
system. Following the performance test
of the CMS, the standards specify
quality assurance requirements with
respectrto daily calibrations of the
instruments. As was noted in the
rulemaking (44 FR 33611, right column),
EPA has initiated laboratory and field
studies to further refine the performance
requirements for continuous monitors to
include periodic demonstration of
accuracy and reproducibility. In view of
the existing performance requirements
and EPA's program to further develop
• quality assurance procedures, the
' Administrator believes that the issue of
continuous monitoring system accuracy
' was appropriately addressed. In doing
so, he recognized that any questions of .
accuracy which may persist will have to
be assessed on a case-by-case basis.
The UARG petition also raised as an
issue the calculation of the 30-day
rolling average emission rate. UARG
maintained that the use of emission data
collected before a boiler outage may not
be representative of the .control system
performance after the boiler resumes
operation. UARG indicated that boiler
outage could last from a few days to
several weeks and suggested that if an
outage extends for more than 15 days, a
new compliance period should be
initiated. UARG also suggested that if a
boiler outag'e is less than 15 days
duration and the performance of the
emission control system is significantly
improved following boiler start-up, a
new compliance period should be
initiated. UARG argued that the data
following start-up would be more
descriptive of the current system
performance and hence would provide a
better basis for enforcement.
A basic premise of this rulemaking
was that the standard should encourage
not only installation of best control
systems but also effective operating and
maintenance procedures (44 FR 33595
center column, 33601 right column, and
3359,7 right column). The 30-day rolling
average facilitates this objective. In
selecting this approach, the Agency
recognized that a 30-day average better
reflects the engineering realities of SOj
and NO* control systems since it affords
operators time to identify and respond
to problems that affect cor.trol system
efficiency. Daily enforcement (rolling
average) was specified in order to
encourage effective operating and
maintenance procedures. Under this
i approach, any improvement in emission
control system performance following
start-up will be reflected in the
compliance calculation along with
efficiency degradations occurring before
the outage. Therefore, the 30-day rolling
average provides an accurate picture of
overall control system performance.
On the other hand, the UARG
suggestion.would provide a distorted
description of system performance since
it would discount certain episodes of
poor control system performance. That
is, the system operator could allow the
control system to degrade and then shut-
down the boiler before a violation of the
standard occurred. After start-up and
any required maintenance, anew
compliance period would commence,
thereby excusing any excursions prior to
a shut-down. In addition, since a new
averaging period would be initiated the
Agency would be unable to enforce the
standard for the first 29 boiler operating
days after the boiler had resumed
operation. In the face of this potential
for circumvention of the standards, the
Administrator rejects the UARG
approach.
UARG also reiterated their previous
comments that EPA did not properly
consider the accuracy and precision of
Reference Method 5 for measuring
participate concentrations af or below
13 ng/J (0.03 Ib/million Btu) heat input.
EPA has recognized throughout this
rulemaking that obtaining accurate and
precise measurements of very low
conrj?n.trat}nn
-------
8232 Federal Register / Vol. 45, No. 26 / Wednesday, February 6, 1980 / Rules and Regulations
and this has not been evaluated. EPA
hds addressed the question of
determining representative locations ; .
and the number of sampling points in
some detail in the reference methods - •-
and appropriate subparts. These •
procedures were designed to assure .
accurate-measurements. EPA has also
evaluated the effects of less then ideal, <
sampling locations and concluded that ,.; ,
generally the results would be biased : - .
bfclow actual emissions. Assessment of ...
Hie extent.of possible biases in * .. : -&j
measurement data, however, must be ... •'
made on a case-by-case basis.
UARG raised again the issue of acid .
mist generated by the FGD system being
collected In the Reference Method 5
sample, therefore rendering the emission
limit unachievable. EPA has recognized
thl3 problem throughout the rulemaking.
In response to the Agency's own
findings and the public comments, the
standards permit determination of
psrtlculate emissions upstream of the
scrubber. In addition, E?A announced
that it is studying the effect of acid.mist
on participate collection and is . . t.
developing procedures to correct the
collected mass for the acid mist portion.
Vll. Applicability of Standards
Sierra Pacific Power Company and ,..
Idaho Power Company (collectively, „ '..-. ..
"Sierra Pacific") petitioned the
Administrator to reconsider the
definition of "affected facility," asking -
that the applicability date of the
standards be established as the date of
promulgation rather than the date of „
proposal. 40 CFR 60.40a provides:
(a) The affected facility to which this "
iiubpart applies is each electric utility steam
(2) For whkh construction or modification
is commenced after September 18, 1978. • ':
September 19, 1978, is the date on
which the proposed standard was
jniblished in the Federal Register. EPA
based this definition on sections
1 1 i{.i}(2) and lllfbKG) of the Act
Section lll{a)(2) provides:
Tlw term "new source" means any
SMttonary source, the construction or
£;CK!if!cation of which Is commenced after the
,.v,V.'i;,ltion ui , . "^lations fur, if earlier,
proposatl refutations) prescribing a standard
of performance under this section which will
be applicable to such source.
Section lll(b](6) includes a similar
provision specifically drafted to govern
the applicability date of revised
stundards for fossil-fuel burning sources
(of which this standard is the chief
example.) It provides:
Any nsw or modified fossil fuel-fired
S!>tltonary source which commences
construction prior to the date.of publication
of the proposed revised standards shall not
be required to comply with such revised
standards. . ,
Sierra Pacific does not dispute that
the Agency's definition of affected , .
facility complies with the literal .terms of
sections lll(a)(2} and lll(b)(6)..Sierra
Pacific maintains, however, that the - .
definition is unlawful, because the .'
standard was promulgated more than 8
months after the proposal, in violation of
sections lllfb)(a)(B) and.307(d)(10).
Section lll(b)(l)(B) provides that a
standard is to be promulgated '/vithin 90
days of its proposal. Section 307(d)(10).
allows the Administrator to extend
promulgation deadlines, such as the 90-
day deadline in section Ul(b)(l)(B), to
up to 6 months after proposal. Sierra
Pacific argues that section lll(a)(2) does
not apply unless the deadlines in
sections lll(b)(l)(B) and 307(d)(10) are
met. In this case the final standard was
promulgated on June 11,1979, somewhat
less than 9 months after proposal. (It
was announced by the Administrator at
a press conference on May 25,1979, and
signed by him on June 1,1979.)
• In the Administrator's view, the
applicability date is properly the date of
proposal. First, the plain language of
section lll[a)(2) provides that the
applicability date is the date of
proposal. Second, the legislative history
of section 111 shows that Congress did
not intend that the applicability date-
should be the date of proposal only
where a standard was promulgated
within 90 days of proposal. Section
lll(a)(2) took its present form in the
conference committee bill that became
the 1970 Clean Air Act Amendments, '
whereas the 90-day requirement came '
from the Senate bill, and'there is no
indication that Congress intended to link
these two provisions;2 , .
Moreover, this interpretation
represents longstanding Agency
practice. Even where responding to
public comments delays promulgation
more than 90 days, or more than 6
months, after proposal, the applicability
dates of new source performance
. standards are established as the date of
proposal. See 40 CFR Part 60, Subparts
D et seq.
Sierra Pacific argues that its position
has been adopted by EPA in
"analogous" circumstances under the
Clean Water Act. This is inaccurate.
Section 306 of the Clean Water Act
specifically provides that the date of
*ln any event, in the Administrator's view the 90-
day requirement in section lll(b)(l](B) no longer
governs the promulgation or revision of new source
standards. It has been replaced by procedures set
forth in section 111(0 enacted by the 1977
uraendmsnts.
• proposal of a new source standard is the
applicability date only if the standard is
promulgated xvithin 120 days of proposal
(section 306(a)(2), (b)(l)(B)}.
Sierra Pacific suggests that utilities
are "unfairly prejudiced" by the
applicability date, but does.not submit
any information to support this claim. In
any event, there does not seem to be
any substantial unfair prejudice. At the
time of proposal, the Administrator had
not decided whether a full or partial
control alternative should be adopted in
the final SO2 standard. As a result, the
Administrator proposed the full control
alternative stating (43 FR 42154, center
column): ,,,
* *' * the- Clean Air Act provides that new
source performance standards apply from the
date they are proposed and it would be easier
for power plants that start construction
during the proposal period to scale down to
partial control than to scale up to full control
should the final standard differ from the
proposal.
In fact; the final SO? standard was less
stringent than the proposed rule.
In this case, utilities were on notice on
September 19,1978, of tb.8 proposed
form of the standard, and that the
standard would apply to facilities-
constructed after that date. In March
1979, it became clear to the Agency that,
it would not be possible to respond to
all the public comments and promulgate .
the final standards by March 19, as
required by the consent decree in Sierra
Club v. Coslle, a suit brought to compel
promulgation of the standard. (The
comment period had only closed on
January 15; EPA had received over 625
comment letters, totalling about 6,000 • •'.
pages, and the record amounted to over
21,000 pages.) The Agency promptly
contacted the other parties to Sierra
Club v. Cdstle, and all the parties jointly
filed a stipulation that the standand
should be signed by Jans 1 caiihaJ Jhs
Administrator should not seek "any
further extensions of time." This
' stipulation was well-publicized (see, for
example, 9 Environment Reporter
Current Developments 2246, March 30,
1979). Thus utilities such as Sierra
Pacific had reasonable assurance that
the standard would be signed by June 1,
as it was.
Even assuming, as Sierra Pacific does,
that section 111 required the standard to
be promulgated by March 19, utilities
had to wait only an additional period of
84 days to know the precise form of the
promulgated standard. This delay is not
substantial in light of the long lead 'times
required to build a utility boiler, and in
light of the fact that the pollution control
techniques required to comply with the,
promulgated standard are substantially
E-24
-------
Federal Register /Vol. 45. No. 26 / Wednesday, February 6^ I960 /Rjeaand Regulations ^ ^233
the same as those required by the
proposed standard. . •
Sierra Pacific's proposal that the
applicability date be shifted to the date
of promulgation is also inconsistent with
Congress' clear desire that the revised
standard take effect promptly. See
section lll(b)(6). ' • .
In conclusion, Sierra Pacific has
submitted no new information, has not
shown that it has been prejudiced in any
way, and has simply presented an
argument that is incorrect as a matter of
Jaw. Its objection is therefore not ot
central relevance and its petition is
denied.
Dated: January 30,1980. -
Douglas M. Costle.
Administrator.
JFR Doc. SO-3,'32 FiiBll 2-5-60: 8:45 am|
B!LUNG CODE 6580-01-M
E-25
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1, REPORT NO.
EPA-45C/3-SC-009b
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Proposed Guidelines for Determining Best Available
Retrofit Technology for Coal-Fired Power Plants and
Other Major Stationary Sources
5. REPORT DATE
M
lovember 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
t, PERFORMING 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. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
DAA for Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
14. SPONSORING AGENCY CODE
EPA/200/04
16. SUPPLEMENTARY NOTES
16. ABSTRACT
Guidelines for the effectiveness and costs of retrofitting coal-fired power
plants and other major stationary sources for control of particulates, NO , and SO,
per the provisions of Section 169A of the Clean Air Amendments of 1977.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI field/Group
Air Pollution
Particulate Matter
Nitrogen Oxides
Sulfur Dioxide
Air Pollution Control and Costs
Steam Generating Units
Air Pollution Control
13-B
8, DISTRIBUTION STATEMENT
Unlimited
19. SECURiTY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITV CLASS /This page)
I Unclassified
22. PRICE
EPA Faint 2220—1 (Rev. 4—77) PREVIOUS EDITION is OBSOLETE
------- |