452-D-96-002
Regulatory Impact Analysis
of Implementation Requirements
for the Reduction of SO2 Emissions
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Air Quality Strategies and Standards Division
MD-15; Research Triangle Park, N.C. 27711
Draft Report
June 1996
-------
fr
I
*l
N
\
"*j
M
"j
(Disclaimer)
This report is issued by the Air Quality Strategies and Standards
Division of the Office of Air Quality Planning and Standards of
-the Environmental Protection Agency. This document is available
_in the public docket for the Proposed Implementation Requirements
-for the Reduction of Sulfur Oxide Emissions, which is available
at EPA's Air and Radiation Docket and Information Center,
Waterside Mall, Room M1500, Central Mall, 401 M. Street SW,
Washington, D.C. 20460. The EPA may charge a reasonable fee for
the copying of materials. Copies are also available through the
National Technical Information Services, 5285 Port Royal Road,
Springfield, Va. 22161. Federal employees, current contractors
or grantees, and non-profit organizations may obtain copies from
the Library Services Office (MD-35), U.S. Environmental
Protection Agency; Research Triangle Park, N.C. 27711; phone
(919)541-2777.
-------
TABLE OF CONTENTS
Page
TABLES
^FIGURES
--EXECUTIVE SUMMARY
SECTION 1. INTRODUCTION 1-1
1. 0 Background 1-1
1.1 Legislative History 1-2
1.2 The Short-Term S02 Externality 1-6
1.3 Proposed Resolution to the Externality 1-9
REFERENCES 1-11
SECTION 2 . STATEMENT OF NEED FOR ACTION 2-1
2.0 Characteristics of Emissions 2-1
2.1 Health Effects 2-5
2 . 2 Market Failure 2-11
REFERENCES 2-14
SECTION 3 . PROGRAM DESCRIPTION 3-1
3 . 0 The Intervention Level Program _ 3-1
3.1 Implementation Guidance 3-4
SECTION 4. COST ANALYSIS 4-1
4.0 Potential Actions and Costs Associated with
the IL Program 4-1
4 .1 Number of Exceedances 4-2
4 .2 Number of Predicted Actions 4-5
4 .3 Estimate of Costs Per Action 4-8
4 .4 Case Studies 4-12
4 .5 Summary 4-25
REFERENCES 4-27
-------
SECTION 5 . ECONOMIC IMPACTS 5-1
5 . 0 Introduction 5-1
5 .1 Industry Background 5-2
5.2 Facility Impacts 5-9
5.3 Summary of Economic Impacts 5-15
5.4 Impacts on Small Entities 5-16
REFERENCES 5-17
SECTION 6 . BENEFIT ANALYSIS 6-1
6 . 0 Introduction 6-1
6.1 Benefit Calculation Procedures 6-2
6.1.1 Step 1: Identify the Relevant
Concentration-Response Functions 6-3
6.1.2 Step 2: Identify the Improvement in
Ambient Air Quality 6-12
6.1.3 Step 3: Determine the Population
Affected by the Change in Air
Quality 6-15
6.1.4 Step 4: Valuation of the Improvement
in Human Health 6-28
6.1.5 Step 5: Estimate Benefits 6-34
6.2 Quantification of Estimates 6-36
6 . 3 Limitation of Analysis 6-40
6.4 Environmental Justice Considerations 6-46
REFERENCES 6-52
SECTION 7. BENEFIT-COST ANALYSIS 7-1
7 . 0 Net Benefit Analysis 7-1
APPENDIX A. Case Studies of Alternative Control
Strategies for the Intervention Level
Program A-1
in
-------
LIST OF TABLES
Table 1-1. Sulfur Dioxide NAAQS Designated Non-Attainment Areas 1-4
.Table 2-1. Monitored Ambient 5-Minute SO2 Peaks for Selected Sites,
1989-1993 2-4
Table 4-1. Summary of Case Studies 4-13
Table 5-1. Characteristics of Industries Selected for Analysis 5-4
Table 6-1. Prevalence of SRAW > 100% Associated With Short-Term
SO2 Levels for Exercising Asthmatics 6-7
Table 6-2. Symptom Prevalence Associated with Short-Term SO2 Levels
For Non-Exercising Asthmatics 6-8
Table 6-3. Symptom Probability Associated with Short-Term SO2 Levels
For Exercising Asthmatics 6-9
Table 6-4. Predicted Annual Exceedances of Alternative 5-Minute
SO2 Concentrations 6-15
Table 6-5. Population Characteristics 6-16
Table 6-6. Population Density Table 6-20
Table 6-7. Population Density Characteristics Used in Impact Analysis 6-21
Table 6-8. SO2 Plume Characteristics 6-21
Table 6-9. SO2 Plume Areas 6-24
Table 6-10. Exposed Population 6-26
Table 6-11. Asthmatic Population at Risk 6-26
Table 6-12. Comparative Indices of Severity of Respiratory Effects Symptoms.
Spirometry. and Resistance 6-31
Table 6-13. Asthma Symptom Severity Related to 5-Minute SO2 Exposure.... 6-31
Table 6-14. Benefits for Case Stud} 1 6-38
Table 6-15. Benefits for Case Study 2 6-38
-------
LIST OF TABLES (continued)
Table 6-16. Benefits for Case Study 3 6-38
Table 6-17. Benefits for Case Study 4 6-39
Table 6-18. Benefits for Case Study 5 6-39
Table 6-19. Sensitivity Analysis of Benefits for Case Study 4 6-39
Table 7-1. Quantified Benefits and Costs of Selected Case Studies 7-3
Table A-1.1. Predicted Annual Exceedances: Copper Smelter Model Plant A-7
Table A-l .2. Model Copper Smelter-3 Years of Exceedance Data A-10
Table A-l .3. Cost of Control (1993 dollars) A-l 1
Table A-l .4(a) Cost of New Intermittent Main Stack A-l 3
Table A-l .4(b) Cost of New Intermittent Main Stack (cont.) A-l 4
Table A-l .5(a) Cost of New Intermittent Slag Stack A-l 5
Table A-l .5(b) Cost of New Intermittent Slag Stack (cont.) A-l 6
Table A-2.1. Estimated SO2 Emissions from Model Primary Lead Smelter A-21
Table A-2.2. Cost of Control (1993 dollars) A-22
Table A-2.3(a) A-23
Table A-2.3(b) A-24
Table A-2.3© A-25
Table A-3.1. Case 3: Cost of Control (1993 dollars) A-30
Table A-3.2. Cost Effectiveness of Control A-30
Table A-3.3(a) Scrubber Costs A-31
Table A-3.3(b) Scrubber Costs A-32
Table A-3.4(a) Project Costs A-33
-------
Table A-3.4(b) Project Costs A-34
Table A-4.1. Emissions for Refinery in Case Study 4 A-36
-Table A-4.1 Annual Number of 5-Minute Exceedances Prior to SIP Controls A-39
Table A-4.2 Case 4: Cos of Control (1993 dollars) A-41
-Table A-4.3(a) Costs for Addition of Dry Scrubber Device A-42
Table A-4.3(b) Costs for Addition of Dry Scrubber Device A-43
Table A-5.1. Case 5: Cost of Control (1993 dollars) A-52
Table A-5.2. Annual Burden Cost for Report Review and Compliance Assurance. . . . A-52
Table A-6.1. Summary of Monitored 5-Minute Data: 1993-1994 A-55
-------
LIST OF FIGURES
Figure 4-1. Distribution of 5-minute Exceedances 4-4
Figure 5-1. Firm's Position in the Market Before Control Costs 5-12
Figure 5-2. Firm's Position in the Market After Control Costs 5-13
Figure A-1.1. Copper Smelter Model Plant Process Diagram A-4
Figure A-1.2. Annual Exceedances for Copper Smelter Model Plant A-6
Figure A-2.1. Annual Exceedances for Lead Smelter Model Plant A-18
Figure A-2.2. Lead Smelter Model Plant Process Diagram A-19
Figure A-3.1. Process Flow Diagram for Model Paper Mill A-27
Figure A-3.2. Annual Exceedances for Paper Mill Example A-29
Figure A-5.1. Annual Number of 5-Minute Exceedances Prior to SIP
Controls A-48
-------
Exective Summary
Sulfur dioxide (S02)is created during the combustion of
sulfur-containing fossil fuels and during the processing of
^natural ores. Data show that on occasion, bursts of SO2 are
.released from sources due to malfunctions, process upsets, and
^during start-up/shut-down procedures. Short-term emissions
(i.e., over a 5 to 10 minute period) can also occur at sources
Jthat use boilers to generate power that is used during facility
operations or for the sale to end-users. As such, short-term
emissions can occur at refineries, pulp and paper mills, copper
smelters, primary lead smelters, coke ovens, electric utilities,
and other facilities with similar operations. Short-term bursts
of S02 are generally disseminated within the local vicinity (less
than 20 kilometers) of the emitting source. When SO2 oxidizes in
water, it forms both sulfurous and sulfuric acids. If SO2
dissolves in the water of the respiratory tract of humans, the
resulting acidity is irritating to the pulmonary tissues.
Studies have demonstrated that acute exposures over a period of 5
to 10 minutes to elevated concentrations of SO2 can cause
respiratory responses in individuals with lung diseases, such as
asthma.
In April 1971, the U.S. Environmental Protection Agency
(EPA) established a primary National Ambient Air Quality Standard
(NAAQS) for S02that is set to protect public health, requiring
ambient air concentrations not to exceed 0.14 parts per million
(ppm) over a 24 hour period no more than once a year with a 0.03
ppm annual arithmetic mean. The EPA also promulgated a secondary
standard to protect the public welfare (i.e., buildings,
vegetation, ecosystems, and human discomfort) of 0.50 ppm not to
be exceeded in a 3 hour period more than once per year. In
addition, the EPA has also established a 24 hour significant harm
level (SHL) program that warns and protects against dangerously
high levels of S02.
During the review of the current NAAQS, the EPA proposed
three regulatory options in March 1995 to address the problems
associated with 5-minute peak S02 concentrations. The regulatory
options considered include: (1) augmenting implementation of the
existing standards by focusing monitoring on those sources or
source types likely to produce high 5-minute peak S02
concentrations, (2) establishing a regulatory program under
section 303 of the Act to supplement the protection provided by
-------
the existing NAAQS, and (3) revising the existing NAAQS by adding
a new 5-minute standard of 0.60 ppm. These regulatory options
were evaluated in a Regulatory Impact Analysis prepared in 1995.
Because evidence suggests that high short-term SO2 concentrations
are a localized problem rather than a widespread national
concern, the current NAAQS was reaffirmed in May 1996 under CFR
Part 50.
Even with the existing programs to protect the public from
exposures to S02, a number of new studies have become available
that examine the potential health effects associated with short-
term exposures to SO2. Conclusions from the supplement to the
staff paper addendum indicate that effects of SO2 over a 5 to 10
minute period in a range of 0.60 to 1.0 ppm is of concern because
a substantial number of asthmatic individuals during elevated
breathing levels experience pronounced changes in lung function
that may be viewed as a mild asthma attack, cause discomfort,
prompt self-medication, and cause some individuals to alter their
activity.
Although 5-minute episodes are infrequent and affect only a
subset of the national population, it is clear that 5-minute SO2
concentrations above 0.60 ppm pose a health threat to sensitive
individuals, and the severity of the threat is a function of the
concentration and frequency of the peaks and population subject
to the episodes. To address the localized problem, the EPA is
proposing to implement a supplemental program under CFR Part 51
that effectively addresses valid concerns regarding short-term
S02 concentrations, while empowering States, local governments,
and communities with the ability and flexibility to address a
given situation appropriately. For these reasons, the EPA has
decided that in lieu of the three implementation options proposed
in 1995, it will propose a new "Intervention Level" (IL) program
under the authority of section 303 of the Act to supplement
protection provided by the existing SO2 NAAQS. Because the IL
program raises novel legal or policy issues, the following
Regulatory Impact Analysis (RIA) has been develope dto respond to
Executive Order 12866.
With the IL program, a range of concentrations is
established to bound the concentrations of concern for short-term
peaks of SO2. The two levels used to bound the concentrations
are: (1) a concern level of 0.60 ppm, and (2) and endangerment
level of 2.0 ppm. If the concern level is exceeded, the States
shall take action as appropriate giving consideration to risk
-------
criteria such as: concentration, frequency of episodes,
population exposed, and other site specific factors. As the
concentration level and frequency of the episode approaches the
endangerment level and the health effects are more pronounced, a
higher risk to the exposed population is anticipated, so State
-action will be increasingly more stringent.
Because the IL program is designed to address a localized
_problem, by providing more flexibility to the implementing
authority to protect the affected population from adverse health
impacts, there is tremendous uncertainty in determining the exact
response to the program by regulatory authorities, the
communities, and affected sources. Due to the numerous
uncertainties surrounding the implementation of such a program,
this document is unable to predict and quantify national impacts
of the IL program, but rather provides examples of a variety of
responses to the program through detailed case study analyses of
a sample of sources.
The cost analysis presents information on the number of
exceedances observed in the country based on best available data,
and the EPA's best judgement of the number of actions that will
occur. The control strategies that can be used in actions taken
.. can vary widely from a low cost alternative such as fuel
switching to a very costly alternative such as the installation
-of add-on control equipment. The cost of control is evaluated
through a series of case studies that present information on a
sample of control strategies that are viable under the IL
program. The types of actions and control strategies analyzed
are not exhaustive, however. Time and resource constraints
prevent an analysis of all possible control alternatives. In
addition, States and local communities while evaluating a 5-
minute SO2 problem may develop new and innovative ways of
addressing S02 concentrations.
Based on public comments received and the detailed
evaluation of existing monitor data submitted by States, the EPA
estimates that a total of ten areas throughout the country have a
potential to be evaluated for the level of public health risk
associated with short-term SO2 episodes. Several of these areas
show indications that the risk to public health would not warrant
action under the IL program due to the frequency and/or
concentration of the peaks, the location of sources vis-a-vis
population, or the time of day of S02 peaks. Overall, of the ten
areas indicated as having a potential short-term SO2 problem by
-------
ambient monitoring data, the EPA reasonably estimates that action
under the IL program could be warranted for approximately five
areas. In making this judgement about the likelihood of action
under the IL program, EPA is using several types of information,
including: 1) historical knowledge about the situation based on
interactions between the EPA Regions, States and local sources;
2) comments provided in response to the original proposals by
sources, States, and local agencies; 3) air quality and census
data; and 4) information about the industrial processes at
facilities in the locations of concern.
The case studies indicate the range of annualized cost for
solutions to different 5-minute SO2 problems to be from
approximately $300,000 to $2.2 million. In addition, some case
studies have no cost associated with the program since action is
not warranted under the IL program. Yet, some studies completed
for other analyses indicate the potential for a cost savings of
approximately $250,000 or a total annualized cost of $30 million.
The case studies demonstrate that the IL program provides a
significant amount of flexibility to regulatory authorities,
communities, and sources to achieve a reasonable solution to
short-term S02 problems at a substantially lower cost than other
potential regulatory vehicles. For example, the previously
proposed regulatory option of establishing a new short-term S02
NAAQS to eliminate exceedances of 0.60 ppm at any one time in a
given year was estimated to cost $1.75 billion. Several of the
sources assumed to incur costs under a NAAQS option would have
the potential to not have any regulatory action taken upon them
under the IL program and thus incur no compliance costs. Even if
all five of the actions predicted to occur under the IL program
have the highest end of costs estimated in the case studies of
this analysis ($2.2 million), the total cost of the IL program
would be $11 million, or $1.74 billion less than the NAAQS
option. Therefore, the IL program is a very cost-effective
solution to the public health risk associated with short-term
peaks of S02.
Given that implementation of the IL program will only occur
in those areas where a regulatory authority has determined that
there is a substantial risk to human health, it is unlikely that
a vast number of sources in any one industry discussed above will
be impacted. Typically, with the uniform implementation of the
cost of a regulation on several producers, an industry's marginal
producer is more likely to be affected causing the market supply
curve to shift, which allows producers to share the burden of a
-------
regulation with consumers through an increase in product prices.
With the IL program, there is a potential of only one or two
sources of an industry to incur additional control costs to
resolve a 5-minute S02 problem. If the sources affected by the
program are not the marginal producers of an industry, the market
supply curve is not likely to shift and the source would not
share the burden with consumers. Rather, the IL program is
likely to cause the source to absorb all of the compliance costs
and incorporate them into the cost of production to determine
thier optimal level of operation.
Given the uncertainties as to the number of actions taken
under the IL program and the types of sources impacted, it is not
feasible to interpret the potential impacts on small entities.
Small entities exist in nearly all of the industries potentially
impacted by the IL program. The cost analysis indicates that the
IL program may impact a total of 5 areas of the country*, which
lessens the likelihood of seeing a significant or
disproportionate impact on a small entities. If an action under
the IL program is taken on a small entity, the costs associated
with the action can be quite low if the state allows flexibility
in compliance methods for the program.
The quantified benefits of the case studies ranged in value
from $2,700 to $44,100. As such the costs exceed benefits by a
significant amount. The small magnitude of benefits results from
mainly two factors. First, the short-term peaks in SO2 under
consideration impact a fairly small geographic area within the
local vicinity of the model plants. The small geographic area
leads to a relatively small number of people being exposed to
these short term peaks. Second, the benefit estimates are
limited to the health benefits accruing to asthmatics who are
participating in activities that cause elevated ventilation
rates. Also, the controls that may result from an IL action
could reduce SO2 emissions year-round which creates benefits in
many other categories. The analysis is unable to consider
welfare benefits associated with any ecosystem, visibility, odor,
materials damage, or particulate matter improvements that may
result from control of short-term peaks in SO2. Although the
costs that are determined for the case studies exceed the
quantifiable benefits, the IL program achieves a reasonable
a Note that any one area affected by the IL program could impact only one or several
sources.
-------
solution to the short-term SO2 problem at substantially lower
cost than other potential regulatory vehicles, such as the
previously proposed new short-term SO2 NAAQS.
In addition to the lower cost of resolving short-term S02
problems, the IL program allows a regulatory authority to
consider environmental justice as a criteria to warrant action
under the IL program. Executive Order 12898 requires that each
Federal agency shall make achieving environmental justice part of
its mission by identifying and addressing, as appropriate,
disproportionately high and adverse human health or environmental
effects of its programs, policies, and activities on minority and
low-income populations.
A number of factors indicate that asthma may pose more of a
health problem among non-white, children, and urban populations.
Considering these factors, a general screening analysis is
conducted to examine the sociodemographic characteristics of the
case study areas potentially impacted by short-term SO2 peaks.
Overall, the populations in the case study areas do not show
any indications that a disproportionate number of non-white
individuals would be impacted by short-term SO2 ambient
concentrations greater than 0.60 ppm. This analysis, however,
does not cover all possible areas of the country with short-term
S02 peak concentrations greater than 0.60 ppm. Other areas of
the country may have a higher percentage of non-white citizens.
The analysis also indicates that there are twice as many children
residing in the case study areas as compared to the national
average, and potentially 595 could have asthma and thus
experience health impacts during peak S02 concentrations. In
addition to the large number of children potentially exposed to
peak SO2 concentrations, 27 percent of the households in the case
study areas are below the poverty level, which is twice the
national average. It should be noted, however, that it is not
known how many of the households below the poverty level contain
asthmatic individuals. Given the available data, there is an
indication that a disproportionate number of children and
households below the povercy level are exposed to short-term SO2
peaks. In general, children do not have the resources to
relocate or take action against sources of S02 emissions.
Similarly, households below the poverty level may be dependent on
local industrial sources for employment. In addition to having
limited resources to relocate or take action against sources of
SO2 emissions, they may be reluctant to do so if action would be
-------
a detriment to employment opportunities.
-------
SECTION 1. INTRODUCTION
1.0 Background
Sulfur dioxide (S02) , a strongly odorous gas, oxidizes in
water to form both sulfurous and sulfuric acids. When S02
dissolves in the water of the respiratory tract of humans, the
resulting acidity is irritating to the pulmonary tissues.
Similarly, when S02 dissolves in rain drops, the "acid rain" can
cause damage to both aq\;atic and terrestrial ecosystems as well
as corrode various materials. Therefore, the primary health
concern for short-term S02 emissions is response in the
respiratory tract of humans, which places individuals with asthma
at higher risk of responding to short-term S02 peaks.
S02 is created during the combustion of sulfur-containing
fossil fuels and during the processing of natural ores. In the
atmosphere, S02 exists with a variety of particles and other
gases, and undergoes chemical and physical interactions with
them, forming sulfates and other transformation products. The
conversion of S02 into sulfates and other products is known to
contribute to problems with acid rain or particulate matter.
Data show that on occasion, bursts of S02 are released from
sources due to malfunctions, process upsets, and during start-
l-l
-------
up/shut-down procedures. Short-term emissions can also occur at
sources that use boilers to generate power that is used during
facility operations or for sale to end-users. As such, short-
term emissions can occur at refineries, pulp and paper mills,
copper smelters, primary lead smelters, coke ovens, electric
utilities, and other facilities with similar operations. Short-
term bursts of S02 are generally disseminated within the local
vicinity (less than 20 kilometers) of the emitting source.
Studies have demonstrated that acute exposures over a period of 5
to 10 minutes to elevated concentrations of S02 can cause
respiratory responses in individuals with lung diseases.
1.1 Legislative History:
In April 1971, the U.S. Environmental Protection Agency
(EPA) established a National Ambient Air Quality Standard (NAAQS)
for S02 under the authority of Sections 108 and 109 of the Clean
Air Act (CAA), which requires the regulation of criteria air
pollutants that may endanger public health or welfare. The
primary S02 NAAQS that is set to protect public health, requires
ambient air concentrations not to exceed 0.14 parts per million
(ppm) over a 24-hour period no more than once a year and a 0.03
ppm annual arithmetic mean. The EPA also promulgated a secondary
standard to protect the public welfare (i.e., buildings,
vegetation, ecosystems, and human discomfort) of 0.50 ppm not to
be exceeded in a 3-hour period more than once per year (38FR
25881, September 14, 1973). As Table 1-1 shows, there are
1-2
-------
currently 44 areas designated as not attaining the current NAAQS.
Periodically, EPA reviews the NAAQS to evaluate whether
rcrevision is necessary to adequately protect the public health and
^welfare. In 1988, EPA reviewed the NAAQS and concluded that the
^current 24-hour and annual standards were both necessary and
adequate to protect human health against S02 concentrations
associated with those averaging periods. These conclusions were
based on the scientific data assessed in criteria documents1'2'3
and staff papers4'5'6 and with the advice and recommendations of
the Clean Air Scientific Advisory Committee of EPA's Science
Advisory Board.
Additional protection is also provided under Title IV of the
1990 CAA Amendments, which requires electric utilities to
"reduce annual S02 emissions by 9 million metric tons (10 million
short tons) per year from a 1980 baseline of 23.3 million metric
tons. This reduction is implemented in two pfeases with the first
phase being completed in 1995 and a larger reduction is expected
in the second phase which will be completed by the year 2000.
While the primary objective of Title IV is to reduce the total
sulfate loadings resulting from regional sulfate transport, some
improvements in local SO ambient air quality will be realized as
a result of the reductions.
1-3
-------
Table 1-1. Sulfur Dioxide NAAQS
Designated Non Attainment Areas
AREA
Penobscot, ME
Warren , NJ
Allegheny, PA
Warren, PAb
Warren, PAb
Hancock, WVb
Hancock, WVb
Boyd, KY
Muhlenberg, KY
Bent on, TN
Humphreys , TN
Polk, TN
Tazewell, IL
Lake , IN
Laporte, IN
Marion, IN
VIGO , IN
Wayne , IN
AQCR 131, MN
Olmested, MN
Coshocton, OH
White Pine, NV
DESIGNATION3
P
P/S
P
P
P/S
P
P/S
P
s
P/S
P/S
P/S
P
P
P
P
P
P
P
P
P
P
AREA
Cuyahoga , OH
Gallia, OH
Jefferson, OH
Lake, OH
Lorain, OH
Lucas , OH
Marathon , WI
Oneida, WI
Grant , MM
Muscatine, IA
Lewis & Clark, MT
Yellowstone, MT
Cochise, AZ
Gila, AZ
Greenlee, AZ
Pima, AZ
Pinal, AZ
Pinal, AZ
Piti-Cabra, GM
Tanguisson, GM
Salt Lake, UT
Tooele, UT
DESIGNATION
P
P
P
P
P
P
P/S
P/S
P
P
P/S
P
P
P
P
P
P
P
P
P
P/S
P/S
The areas are indicated as being nonattainment for the
primary, secondary, or both NAAQS by P, S, P/S.
Because areas in Warren County, PA and Hancock, WV were
designated at different times, these counties each have two
separate nonattainment areas.
1-4
-------
Finally, EPA also has established a 24-hour significant harm
level (SHL) program that warns and protects against dangerously
high levels of S02. This program was designed to address
jremergency episodes that would occur where pollution levels build
i.up over a period of time to unhealthy levels. The program
^establishes four levels of concern, that if exceeded within a
,24-hour period, States must undertake various actions to remedy
the situation. The four levels established in the SHL are:
* Alert Level - 0.30 ppm,
* Warning Level - 0.60 ppm,
* Emergency Level - 0.80 ppm, and
* Significant Harm Level - 1.0 ppm.
The SHL program is a proactive program designed to prevent an
area from ever reaching the SHL. Between the Alter and Emergency
levels that are below the SHL, emission sources in the area are
required to take increasingly restrictive action to reduce
emissions as specified in the contingency plans with the approved
State implementation plan (SIP). Exceedance of the 1.0 ppm
concentration of the SHL requires urgent measures contained in
the SIP on the part of the State and emission source to correct
and prevent the episode from occurring again. From the Alert to
the Emergency levels that are below the SHL, emission sources in
the area are required to take increasingly restrictive action to
reduce emissions as specified in the contingency plan within the
1-5
-------
approved State implementation plan.
1.2 The Short-Term S02 Externality
Even with the existing programs to protect the public from
exposures to S02, a number of new studies have become available
that examine the potential health effects associated with short-
term (less than or equal to 1-hour) exposures to S02 (see the
staff paper supplement for a review of recent studies). In view
of these new studies and other relevant new information, EPA
prepared a supplement to the criteria document addendum8 and a
supplement to the staff paper addendum9. Conclusions from the
supplement to the staff paper addendum indicate that effects of
S02 over a 5 to 10 minute period in a range of 0.60 to 1.0 ppm
are of concern because a substantial number of asthmatic
individuals (approximately 25 percent) during oronasal (i.e.,
mouth and nose) breathing experience pronounced changes in lung
function that may be viewed as a mild asthma attack, cause
discomfort, prompt self-medication, and cause some individuals to
alter their activity. The response, however, generally is
resolved within an hour, and some individuals can still function
effectively despite whatever effects they perceive from the S02
exposure10.
The EPA currently has limited source oriented monitoring
information on 5-minute monitoring data for S02. However, EPA
evaluated data submitted from 16 States for SO-> ambient air
1-6
-------
monitors. The data from these monitors indicate that 43 percent
of the monitors registered 5-minute averages in excess of 0.60
ppm S0:. In addition, several of the monitors recorded multiple
exceedances of 0.60 ppm. Fifty percent of the monitors that
.indicated high .peaks of S02 recorded from 11 to 139 exceedances.
This evidence is likely to underestimate the national problem
because data were available from only a tenth of the SO2 monitors
nationally, and because the current monitoring network is set-up
in urban areas to measure ambient air quality for attainment of
the current 3-hour, 24-hour and annual NAAQS. If States decide
to relocate monitors to better evaluate 5-minute ambient SO.
concentrations around sources of concern, the number of measured
exceedances of S02 peaks could increase significantly.
During a review of the current NAAQS, EPA reviewed the
evidence and proposed three regulatory options in March 1995 to
address the problems associated with 5-minute peak S02
concentrations. The regulatory options considered include:
(1) augmenting implementation of the existing standards by
focusing monitoring on those sources or source types likely to
produce high 5-minute peak S02 concentrations, (2) establishing a
regulatory program under section 303 of the Act to supplement the
protection provided by the existing NAAQS, and (3) revising the
existing NAAQS by adding a new 5-minute standard of 0.60 ppm.
1-7
-------
These regulatory options were evaluated in a Regulatory Impact
Analysis prepared in 1995".
Compelling comments received from the March 1995 proposal
indicate that (1) there were a limited number of communities
showing evidence of a problem and (2) the emissions do not travel
far from the source when episodes occur. This suggests that high
short-term S02 concentrations are a localized problem rather than
a widespread national concern. Commenters argued that States
should be given the authority and the flexibility to impose
appropriate control requirements, especially in cases when the
short-term peaks are rare, and the potential for exposure is low.
Although 5-minute episodes are infrequent and affect only a
subset of the national population, it is clear that 5-minute SO2
concentrations above 0.60 ppm pose a health threat to sensitive
individuals, and the severity of the threat is a function of the
concentration and frequency of the peaks and population subject
to the episodes. Because every area that is subject to
significant short-term peaks has its own unique characteristics,
EPA agrees it is prudent to assess each individual situation, and
when necessary, act appropriately and efficiently to reduce the
risk to the public and that the States, being closest to each
individual situation, are in the best position to do so.
In general, the areas that are known to have high 5-minute
peak concentrations of S0: have market systems that have failed
1-8
-------
to deal effectively with air pollution. This occurs because the
ambient air has been treated as public goods and because most air
polluters do not internalize the full damage caused by their
emissions.
1.3 Proposed Resolution to the Externality
As a result of comments and additional information received,
EPA reaffirmed the current NAAQS program for S02 in May of 1996
under CFR Part 50. However, EPA is proposing to implement a
supplemental program under CFR Part 51 that effectively addresses
valid concerns regarding short-term S02 concentrations, while
empowering States, local governments, and communities with the
ability and flexibility to address a given situation
appropriately. For these reasons, EPA has decided that in lieu
of the three implementation options proposed in 1995, it will
propose a new "Intervention Level" (IL) program under the
authority of section 303 of the Act to supplement protection
provided by the existing S02 NAAQS.
With the IL program, a range of concentrations is
established to bound the concentrations of concern for short-term
peaks of S02. The two levels used to bound the concentrations
are: (1) a concern level of 0.60 ppm, and (2) and endangerment
level of 2.0 ppm. If the concern level is exceeded, the States
shall take action as appropriate giving consideration to risk
criteria such as: concentration, frequency of episodes,
1-9
-------
population exposed, and other site specific factors. As the
concentration level and frequency of the episode approaches the
endangerment level and the health effects are more pronounced, a
higher risk to the exposed population is anticipated, so State
action will be increasingly more stringent.
This document analyzes the impacts of such a program on
affected sources. It will describe the IL program in detail and
evaluate the costs, benefits and economic impacts of the program.
Because the IL program is designed to address a localized
problem, by providing more flexibility to the implementing
authority (i.e., the States) to protect the affected population
from adverse health impacts, there is tremendous uncertainty in
determining the exact response to the program by regulatory
authorities, the communities, and affected sources. This
document, therefore, provides examples of a variety of responses
to the program through detailed case study analyses of selected
sources. Due to the numerous uncertainties surrounding the
implementation of such a program, this document is unable to
predict and quantify national impacts of the IL program, but
rather evaluates the potential national number of actions taken
for the IL program based on known exceedances of 0.60 ppm S02 and
provides a qualitative discussion of national impacts.
1-10
-------
REFERENCES
1. Air Quality Criteria for Particulate Matter and Sulfur
Oxides. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards; RTP, N.C. 27711. Document
no. EPA-600/8-82-029a-c; December 1982.
2. Second Addendum to Air Quality Criteria for Particulate
Matter and Sulfur Oxides (1982) : Assessment of Newly
Available Health Effects Information. U.S. Environmental
Protection Agency, Office of Air Quality Planning and
Standards; RTP, N.C. 27711. Document no. EPA/450/5-86-012,-
1986.
3. Supplement to the Second Addendum (1986) to Air Quality
Criteria for Particulate Matter and Sulfur Oxides(1982):
Assessment of New Finding on Sulfur Dioxide Acute Exposure
Health Effects in Asthmatic Individuals. U.S. Environmental
Protection Agency, Office of Air Quality Planning and
Standards; RTP, N.C. 27711. Document no. EPA/600/AP-93/002;
March 1994.
4. Review of National Ambient Air Quality Standards for Sulfur
Oxides: Assessment of Scientific and Technical Information -
OAQPS Staff Paper. U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards; RTP, N.C.
27711. Document no. EPA-450/5-82-007; November 1982.
5. Review of National Ambient Air Quality Standards for Sulfur
Oxides: Updated Assessment of Scientific and Technical
Information. U.S. Environmental Protection Agency, Office
of Air Quality Planning and Standards; RTP, N.C. 27711.
Document no. EPA-450/05-86-013; December 1986.
6. Review of National Ambient Air Quality Standards for Sulfur
Oxides: Updated Assessment of Scientific and Technical
Information - Supplement to the 1986 OAQPS Staff Paper
Addendum. U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards; Document no. EPA-452/r-
94-013; September 1994.
7. Ozone, Carbon Monoxide, Particulate Matter, Sulfur Dioxide,
Lead: Areas Designated Nonattainment. U.S. Environmental
Protection Agency, Office of Air Quality Planning and
Standards; RTP, N.C. 27711. July 1995.
8. Reference 3.
9 . Reference 6.
1-11
-------
10. Reference 6.
11. Regulatory Impact Analysis for the Proposed Regulatory
Options to Address Short-Term Peak Sulfur Dioxide Exposures.
U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards; RTP, N.C. 27711. March 1994.
1-12
-------
SECTION 2. STATEMENT OF NEED FOR ACTION
2.0 Characteristics of Emissions
When there are short-term episodes of S02, the emissions of
concern are disseminated in local vicinities to the source. All
emissions that travel beyond the local vicinity of the source are
generally diluted with ambient air to a concentration that will
not significantly impact public health.
In addition, there are also instances when short bursts of
S02 are emitted during thermal inversions, which traps the
emission in an area for prolonged periods of time. Thermal
inversions occur in unique cases of geography and meteorology.
If a source is located in a valley of hilly terrain, weather
conditions could exist in which colder air at-the elevated levels
of the hills or mountains traps warmer air in the valley closer
to the ground. Instead of allowing the warmer air to rise and
disseminate, it remains stagnant for prolonged periods of time.
There is a potential for numerous short-term S0_ episodes
around various sources that have sulfur as a component of
combustion or process operations. The Staff Paper Supplement1
examined available monitoring data from 1989 to 1993, which
2-1
-------
indicates the presence of peaks at or above both 0.50 ppm and
0.75 ppm S0:. Table 2-1 presents the number of hours during
which one or more 5-minute peaks at or above 0.50 and 0.75 ppm
were observed for a sample of source types based on information
contained in the Staff Paper Supplement2. In a subsequent study
completed in September 1995 by ICF Kaiser, Inc.3, monitoring data
from 16 States were submitted and analyzed for the existence of
5-minute peak S02 concentrations. Results of the analysis
demonstrate a prevalence of S02 concentrations in excess of 0.60
ppm for 43 percent of the monitors evaluated3.
Currently available information on 5-minute peaks of S02 is
limited for several reasons. The primary reason is that the
placement of monitors within the existing network is designed to
measure ambient air quality relative to the existing 3-hour, 24-
hour, and annual NAAQS. Therefore, use of hourly S02 data for
this analysis may underestimate the true potential for 5-minute
peaks. Additionally, because there are no requirements to
collect and submit 5-minute data, resources have been allocated
to other areas of monitoring. Overall, there is sufficient
evidence to determine that 5-minute peaks of S02 above 0.60 exist
and have the potential to affect the population surrounding the
a The concluding result of the analysis that 43% of the
monitors indicated a 5-minute problem is merely provided to
demonstrate that the problem exists. This result cannot be used
to determine the severity of a national problem because this
estimate is based on data that was voluntarily submitted by
States for S0: emissions around sources known to have 5-minute
problems.
2-2
-------
source of emissions.
2-3
-------
w
u
DJ
O
Id <*
EH CT>
D 'H
5 .
I
CTl
•7
ED AMBIEl
!D SITES,
0 g
1 — ( t1-!
2 "-1
0 w
5"* t/3
o
_] tj
rH *^
!
CM
W
•"-3
PH
ri]
.H
u
3 e
2&
o
CO "
0
•£ 2
a 2
^
S -^
a: w
o £
EH •<*•
2 ?5
2
VALUE
ppm
^in
O
3°'
[£] rfj
dl "
EH
g«
$ w
o %
H <
M "
V Oi
0 15
rS
01 —
C CM
-H *—
M
o -a
4-> O
•H -rH
C M
O rH
3 W
2 .0
O
C CM
-H —
0 T3
-tn -°
c ^
O CJ
ber of
rvances
(1)
g 01
3 in
2 X!
o
i
M
CU
C
•H
MH
(U
Q^ ^—
(V)
g —
3
0) X
-t CU
O rH
i-l p.
4J g
0) O
Oi U
It)
c
o
o
rH
5
00
rH
rH
•H
^
^1
(U
a
tfl
Oi
CU
jj
H
MH
rH
3
w
10
c
o
CM
cn
o
m
oo
(V)
— -
ff
O-i
^
^
4-)
C
3
O
U
>1
c
cu
_c
^
CU
1 — i
rH
cf,
m
c
o
in
CM
00
^-^
*3<
—
^1
0)
4-1
rH
CO
g
w
M
0)
a
a
o
o
in
rH
«
rH
in
CM
rH
m
rH
rH
CM
^1
0)
4J
rH
01
g
TI
10
CU
^
M
fT3
g
-H
i-l
04
0
"
tH
rH
in
o
rH
rH
iH
01
4J
rH
0)
g
U~l
i-l
.1)
a
a,
o
o
in
rH
CM
r-
in
CM
CM
PO
rH
r^
^
, 1
CU
CU
4-1
en
vo
rH
.
m
00
00
rH
m
m
rH
X
CU
rH
a,
e
o
u
rH
*
rH
M
4J
W
3
TJ
C
M
^^
t>-\
4J
H
-H
-H
4_J
D
rH
CO
.
O
CM
rH
01
o
rH
i-l
CO
a,
m
4J
m
to
M
-\
l_l
CU
rH
•H
O
00
rH
r--
VJ
4_)
ij]
3 r-;
*D • — !
C H
1 | *r>
O
"
rH
O
0
rH
O
£>.
M
0)
C
•H
>4H
0)
D-
e
3
CU
, !
o
M
4J
CU
Oi
o
"
«,
o
rH
o
^
M
CU
C
-H
m
(D
a
g
3
CU
,H
O
iH
4-1
(D
DJ
c
o •
e i-
£
a
U! C
c ^
o a
L "n 4-t
-i o co
-C -^ C
J u _.
4'
caw
01
r? -p
C 0 T>
jr 4- c
C-J
-------
2.1 Health Effects
To better understand the impact of short-term S0: emission
on human health, the following briefly characterizes asthma and
^discusses how people with such respiratory conditions would
respond to S02. .
Asthma is a disease that creates breathing difficulties for
individuals in response to a variety of environmental, chemical,
and physical conditions (i.e., cold or dry air, pollutants,
allergies, exercise). The disease can be classified as mild,
moderate, or severe and affects approximately 5 percent of the
national population5'4. The prevalence of asthma is higher among
African-Americans, older (8 to 11 year old) children, and urban
residents. Because there is a wide degree of variability of the
symptoms of asthma, some individuals may be unaware that they
have the disease, while others treat the disease through
medication and with doctor supervision. Asthma attacks can
result in a need to disrupt activities and rest, require self-
treatment with inhalers or medicine, or necessitate hospitaiization
and emergency room treatment5.
The most striking response to S0: for asthmatics and others
with hyperactive airways is bronchoccnstriction (airway
narrowing), usually evidenced as increased airway resistance, and
Many cases of mild asthma may be unreported, therefore, the true
prevalence of asthma may be as high as 7 to 10 percent of the
national population.
2-5
-------
the occurrence of symptoms such as wheezing, chest tightness, and
shortness of breath. The symptoms and response occurs quickly
(within 5 to 10 minutes of exposure). The response is also
generally brief in duration and if the stimuli are removed, lung
function usually returns to normal within 1-hour.
Healthy nonasthmatic individuals are essentially unaffected
by acute exposures to S02 at concentrations below 2.0 ppm.
However, for individuals with asthma or hyperactive airways the
effects of S02 increases with both increased overall ventilation
rates and an increased proportion of oral ventilation in relation
to total ventilation. Oral ventilation is thought to accentuate
the response because the scrubbing of S02 by the nasal passages
is bypassed6. Ventilation rates that trigger oronasal breathing
can occur from activities such as climbing about three flights of
stairs, light cycling, shoveling snow, light jogging, playing
tennis, or walking up a moderate hill. Moderately higher
breathing can occur from activities such as moderate cycling,
chopping wood, or light uphill running. Even though such
exercise is not strenuous per se. it has been determined that
these activities are enough to cause some bypassing of nasal
passages in breathing which exposes SO—sensitive individuals to
a risk of bronchoconstriction. Risk is also present for
individuals who are obligate mouth breathers, or who may be
breathing through their mouth due to nasal congestion from
temporary conditions7. In contrast, individuals with more severe
2-6
-------
asthmatic conditions have poor exercise tolerance and, therefore,
are less likely to engage in sufficiently intense activity to
achieve the requisite breathing rates for notable S02-induced
-respiratory effects to occur8'9.
The health effects associated with exposures to the proposed
concern level, 0.6 ppm S02/ 5-minute block average, were the
focus of EPA's most recent review of the primary national ambient
air quality standards for sulfur oxides (measured as sulfur
dioxide). The health effects and the Administrator's conclusions
about the public health risks associated with exposure to 0.60
ppm S02 are thoroughly discussed in the EPA documents generated
during that review: the criteria document supplement10, the staff
paper supplement11, the November 15, 1994 proposal notice (59 FR
58958) and the [insert date of publication] final decision notice
[insert FR cite].
The EPA's concern about the potential public health
consequences of exposures to short-term peaks of S02 arose from
the extensive literature involving brief (2- to 10-minutes)
controlled exposures of persons with mild (and, in some cases
moderate) asthma across a range of concentrations of S02 while at
elevated ventilation rates. The major effect of S02 on sensitive
asthmatic individuals is bronchoconstriction, usually evidenced
in these studies by decreased lung function and the occurrence of
clinical symptoms such as wheezing, chest tightness, and
2-7
-------
shortness of breath. The proportion of asthmatic individuals who
respond, the magnitude of the response and the occurrence of
symptoms increase as S02 concentrations and ventilation rates
increase. The criteria document supplement contains a summary of
the literature on the health effects associated with brief
exposures to S02, some details of which are provided in the
benefits analysis of this document.
Taking into account the available health effects studies and
the body of comments on the health effects, the Administrator
concluded in the final decision notice [FR cite] that a
substantial percentage (20 percent or more) of mild-to-moderate
asthmatic individuals exposed to 0.6 to 1.0 ppm S0: for 5 to 10
minutes at elevated ventilation rates (such as would be expected
during moderate exercise) would be expected to have lung function
changes and severity of respiratory symptoms that clearly exceed
those experienced from typical daily variation in lung function
or in response to other stimuli (e.g., moderate exercise or
cold/dry air). For many of the responders, the effects are
likely to be both perceptible and thought to be of some health
concern; that is, likely to cause some disruption of ongoing
activities, use of bronchodilator medication, and/or possibly
seeking of medical attention.
During the regulatory review process of the current NAAQS,
there was some agreement by medical experts that at this
2-8
-------
concentration, 0.60 ppm S02, the frequency with which such
effects are experienced may affect the degree of public health
concern that is appropriate. After taking into account the broad
..range of opinions expressed by CASAC members, medical experts,
Pand the public,, in the final decision notice [FR cite] the
:Administrator concluded that repeated occurrences of such effects
should be regarded as significant from a public health
standpoint, and that the likely frequency of occurrence of such
effects should be a consideration in assessing the overall public
health risk in a given situation.
The severity of respiratory symptoms and lung function
changes are greater than normal when asthmatic individuals are
exposed to S02 concentrations of 0.6 to 1.0 ppm S02. At 0.60
ppm, some mild or moderate asthmatic individuals at elevated
ventilation are likely to respond with bronchoconstriction and
effects are likely to be thought of as an immediate health
concern. At 1.0 ppm, the effects are likely €o be more
pronounced. Individuals experience more substantial changes in
pulmonary function accompanied by symptoms and may also
experience mild bronchoconstriction while at rest, which may
cause disruption of ongoing activities, use of medication, and/or
possibly seeking medical attention. At concentration levels
above 1.0 ppm, concern is increased. At 1.5 ppm, there is an
increased fraction of mild and moderate asthmatics who are likely
to respond with more pronounced effects, and there is increased
2-9
-------
concern for more severe asthmatic individuals who have poor
exercise tolerance. At 2.0 ppm, approximately 80 percent of the
at risk population are likely to respond with effects ranging
from moderate to incapacitating. Asthmatic individuals at rest
are likely to experience moderate bronchoconstriction that would
necessitate medication or hospitalization. At 3.0 to 5.0 ppm,
nonasthmatic adults at mild exercise will experience
bronchoconstriction, and asthmatic individuals at rest will
likely experience pronounced bronchoconstriction.
Many asthmatics take medication to relieve symptoms and
functional responses associated with exacerbation of this
disease. One of the most commonly used asthma medications (beta-
agonists) also inhibits S02. This has led to suggestions that
asthmatic individuals may be protected from responses to S02
because they medicate prior to exercise. However, most mild
asthmatic individuals use medication only when symptoms arise.
(SP, p. 16). Therefore, pre-exercise bronchodilator use would
not be likely to occur for many potentially SO;- sensitive
individuals. In addition, many moderate asthmatics who come from
low socioeconomic status may not have adequate access to the
health care system, may have poor medication use based on lack of
finances to purchase medication and thus may be prone to frequent
deterioration of their lung function. Such individuals would be
at increased risk from S02 exposure because of their potentially
lower baseline level of lung function.
2-10
-------
2.2 Market Failure
The analysis of recent data also indicate that 5-minute
peaks of S02 occur in areas that also violate the current S02
.JJAAQS program. As culpable sources strive to attain the current
*iNAAQS, some 5-minute peaks will be resolved, however, there are
^several occurrences that will not be captured by the current
NAAQS. These incidences generally occur in local areas and can
be corrected by actions taken by States or local regulatory
authorities. Unfortunately, the constitutions of 16 States
declare that regulatory measures placed on citizens and
businesses of the state may not be any more stringent than
federal regulations. This precludes several States from taking
independent action for known problems with short-term S02 peaks.
In general, the areas that are known to have high 5-minute
peak concentrations of S02 have market systems that have failed
to deal effectively with air pollution. This occurs because the
ambient air has been treated as public goods and because most air
polluters do not internalize the full damage caused by their
emissions. Although States and the Federal government have
several programs in place to limit emissions of S0: to the
atmosphere (and thus help sources internalize the costs of any
damages to the environment), bursts of S0; continue to be emitted
in some areas. Once in the atmosphere, citizens around these
sources incur real costs associated with the pollution. In
economic theory, this is referred to as a negative externality.
2-11
-------
In theory, affected parties could participate in
negotiations with the polluting sources to receive compensation
for damages incurred, or resolve the pollution problem at the
source. However, such resolutions might not occur in the absence
of regulatory action because of two major impediments which block
the correction of pollution inefficiencies and inequities by the
private market. The first is the high transaction costs that
occur when a large number of individuals who are affected by the
pollution, act independently to negotiate and resolve the problem
with the source(s)c. In return, the source faces transaction
costs to compensate individuals adversely impacted by air
pollution by contacting the individuals affected, apportioning
injury to each from the various polluting sources, and executing
the appropriate damage suits of negotiations. If left to the
private market, each polluter and each affected individual would
have to litigate or negotiate on their own or else organize into
groups for these purposes. The transaction costs involved would
be high and could probably exceed the benefits of any reduction
in pollution.
The second factor discouraging private sector resolution of
any air pollution problems is that pollution abatement tends to
be a public good. That is, once emissions from a particular air
c It should be noted that the source(s) that citizens would
negotiate with for resolution are often the primary employer for
the local area and are vital components of the local economy.
Citizens may not feel at ease to cause difficulties for such a
source.
2-12
-------
pollutant have been reduced through abatement measures, the
benefits of the abatement can be enjoyed by additional people at
no additional cost. This constitutes the classic "free rider"
^problem. As such, any one individual that is adversely impacted
by bursts of short-term S02 may be reluctant to contribute their
time or money to reduce pollution knowing that the potential
- exists for him to enjoy the benefits of reduced pollution (at no
cost) if another person took abatement action. As a result,
without community support or group participation in areas
affected by short-term S02,action to resolve the problem is
unlikely to occur.
Based on comments received from the previous proposal,
mechanisms to establish a national regulation to correct the
externality has been argued to also be too burdensome and an
inefficient use of society's resources. For instance, one
regulatory option for national control that was evaluated in the
previous RIA was the establishment of a new 9O2 NAAQS. While
this option would eliminate the problem at known sources, it
would be inefficient for sources in areas with a low public
health risk from 5-minute peak concentrations to be required to
.-install control equipment. This document demonstrates that
providing a more flexible program that allows States to monitor
and remediate short-term peaks based on public health risk
appears to provide a more efficient solution to the problem than
the three other implementation options evaluated in the previous
2-13
-------
References
1. Review of National Ambient Air Quality Standards for Sulfur
Oxides: Updated Assessment of Scientific and Technical
Information - Supplement to the 1986 OAQPS Staff Paper
Addendum. U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards; Document no. EPA-452/r-
94-013; September 1994.
2. Regulatory Impact Analysis for the Proposed Regulatory
Options to Address Short-Term Peak Sulfur Dioxide Exposures.
U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards; RTP, N.C. 27711. March 1994.
3. Summary of 1988-1995 Ambient 5-Minute S02 Concentration
Data, Draft Final Report. ICF Kaiser, Systems Applications
International; RTP, N.C. Prepared under contract no. 68-D3-
0101, work assignment 7 for the U.S. Environmental
Protection Agency. September 1995.
4. 1994 National Prevalence Rates for Asthmatics. National
Center for Health Statistics; March 1996.
5. Reference 1.
6. Reference 1.
7. Reference 1.
8. Supplement to the Second Addendum (1986) to Air Quality
Criteria for Particulate Matter and Sulfur Oxides(1982):
Assessment of New Finding on Sulfur Dioxide Acute Exposure
Health Effects in Asthmatic Individuals. U.S. Environmental
Protection Agency, Office of Air Quality Planning and
Standards; RTP, N.C. 27711. Document no. EPA/600/AP-93/002;
March 1994.
9. Reference 1.
10. Reference 8.
11. Reference 1.
2-14
-------
SECTION 3. PROGRAM DESCRIPTION
a.-3 .0 The Intervention Level Program
Given that 5-minute peak S02 emission events pose a health
threat, but tend to be localized problems in areas scattered
throughout the United States, the intervention level program
allows for placement of resources and efforts precisely where the
problem occurs, instead of requiring a blanket nationwide
. approach that might call for unnecessary administrative effort.
The Intervention Level (IL) program is derived in part from
the SHL program, which has served in the past as a means for
- implementing the authority granted under section 303 of the CAA.
Whereas the SHL program is proactive, establishing measures in
advance to prevent .pollution levels form exceeding the SHL, the
IL program is a reactive approach to prevent future occurrences
of unhealthy pollution events once these levels have been
reached. The intervention level program establishes a range of
concentrations in the Code of Federal Regulations with the lower
boundary being the concern level, set at 0.60 ppm SO.., and the
upper boundary being the endangerment level, set at 2.0 ppm SO.'.
1 The measurement of the concentrations are based on a 5-
minute block average, which is a 5-minute hourly
maximum value for SO, obtained by the highest of the 5-
3-1
-------
These boundary levels are based on health criteria discussed in
chapter 2, and their objective is to protect the population at
risk from "...imminent and substantial endangerment to public
health or welfare, or the environment...", as stated in section
303 of the CAA..
In the event that the concern level concentration is
exceeded in a given area, the State should assess the situation
to determine whether intervention is appropriate. In making this
determination, the State should consider the concentration of the
5-minute peaks, the frequency of the episodes (based on monitor
data and an estimate of the number of 5-minute peaks not recorded
by the monitoring network), the history and nature of any citizen
complaints, available information on potential population
exposure (inferred in part by the population in the vicinity of
the source), the type of process being used, a history of past
upsets or malfunctions, the type of fuel used, knowledge of how
well the source is controlled, and any other considerations
deemed necessary by the State.
Because the health effects become more severe as the
5-minute S0: concentration approaches the endangerment level, it
is expected that the State will respond with more intensive
corrective measures as the endangerment level concentration is
minute averages from the 12 possible nonoverlapping
periods during a clock hour.
3-2
-------
approached. If the State determines that the circumstances
surrounding a source of high 5-minute peaks pose an unacceptable
risk of harm, it should take remedial action as appropriate. For
example, if the endangerment level is exceeded, the States could
sconsider taking action to shut down the facility until the cause
sof the high 5-minute peaks can be remedied. If necessary, EPA is
-prepared to take action under the authority of section 303 if the
endangerment level is exceeded in a given area, and the State
fails to address the problem.
Like the previously proposed implementation alternatives, a
key element of this new implementation strategy is the relocation
of existing S02 monitors to areas near point sources where peak
S02 concentrations may exist. The existing S02 monitoring
•network was designed to characterize urban ambient air quality
^associated with 3-hour, 24-hour, and annual S02 concentrations,
and cannot adequately measure peak S02 concentrations from point
sources. To allow for the measurement of short-term peaks, EPA
proposed revisions to the ambient air quality surveillance
requirements (40 CFR, Part 58) and proposed certain technical
changes to the requirements for Ambient Air Monitoring Reference
•^and Equivalent Methods (40 CFR, part 53) in November 1994 and
March 1995 notices.
The EPA believes that these changes to the monitoring
requirements will give the States the flexibility to locate
3-3
-------
monitors in areas where they are concerned about 5-minute peaks,
and to respan the monitors to measure these peaks. Under the
intervention level program, the States would be able to identify
areas to be monitored, based on State priorities, source
emissions, citizen complaints, or other variables. The EPA will
assist the States' efforts to identify and prioritize areas for
monitoring 5-minute peak concentrations by providing information
compiled from various databases. The EPA leaves the discretion
on how best to utilize this information to the States.
Unlike the program originally proposed by EPA under section
303, the intervention level program does not require States to
submit revised contingency plans to EPA requiring specific
actions for the State and source to undertake when an exceedance
occurs. The EPA presumes that the SIPs currently in force
provide the States with adequate general authorities to implement
the intervention level program without submittal of revised
contingency plans because section 110(a)(2)(g) of the CAA
requires that the SIPs contain adequate authorities to implement
section 303 programs. Elimination of the requirement to submit
revise contingency plans is expected to minimize the potential
administrative burden on the States.
3.1 Implementation Guidance
The EPA believes the concern level of 0.60 ppm averaged over
a 5-minute period is the concentration at which States should be
3-4
-------
concerned about the health impacts of a peak emission episode.
Although a detailed guidance document will be developed and
provided to regulatory authorities and the public, the following
-provides a general description of implementation procedures for
*the IL program.
f
Once the concern level has been exceeded in a given area,
the State should investigate the episode, and consider the number
of episodes (both observed and predicted), the concentration
levels, the nature and location of the source (or sources), the
proximity of the source to population, and other pertinent
factors to characterize risk to the public health. Based on the
concentration and frequency of the 5-minute peak concentration
events, the State may wish to carry out a compliance inspection
-of the culpable source(s). If the source is out of compliance
-with its exiting emission limits (based on the NAAQS or other air
pollution requirements), then the State would take necessary
steps to bring the source into compliance. I-f, however, the
State determines that bringing the source into compliance with
its existing emission limits would not be likely to prevent
further exceedances of the concern level, or the State determines
- the source to be in compliance with all applicable emission
limits, then further action may be needed. In such
circumstances, the next step would be for the State and source to
examine the cause of the emissions, the nature of the peaks, the
potential for exposure, and the risk to public health. Once
3-5
-------
these are determined, the State, source, and community would
determine what a course of corrective action, if any, would need
to be developed to address the cause of the 5-minute peaks.
Under the intervention level program, EPA is not specifying
a time limit in which States and sources must take corrective
action, although EPA expects that development of control
strategies and implementation of any course of corrective action
will occur in an expeditious and efficient manner. The State
should determine what is considered to be expeditious for each
individual situation, based on the risk to public health,
specific processes or operations at the source that cause the
peak episodes, the available options for control, the reasonable
lead time necessary for planning design, procurement and
installation of control devices or process modifications, and
other pertinent considerations. Control measures needed to
prevent recurrences of 5-minute S02 peaks may include better
operation and maintenance of control equipment, better capture of
fugitive emissions, raising the stack height for intermittent
control, restriction of operations during times of peak exposure
(i.e., conducting activities during hours when fewer people are
outside, or when weather conditions are unfavorable), or other
innovative control measures.
In determining the course of corrective action, States may
also consider the appropriateness of control alternatives. When
3-6
-------
the health risk does not warrant the application of specific
control measures, States and sources may wish to consider
addressing the health risk through alternative approaches. The
tState must ensure that any corrective action, including non-
icontrol approaches, are Federally enforceable against the source.
In the event that a State does not take action once the
intervention levels have been exceeded, the EPA would consult
with the State to discuss the basis for the State's decision.
After consulting with the State, if EPA determines that
corrective action is warranted to protect public health, EPA will
take action.
The intervention level program also provides a mechanism for
-involvement by members of the local community to a source of
potential emissions. When States evaluate the potential for a
short-term S02 problem, they should also take into account the
number and nature of citizen complaints received, and apply
suitable resources to receiving, reviewing, and responding to the
concerns of citizens and community groups. Citizens who express
concerns about the health and welfare effects due to high ambient
concentration peaks should be given every opportunity to present
-and clarify their concerns to the State. Citizens, in turn,
should be made aware of what types and levels of information will
be most helpful in determining links between peaks and health
effects, and given every opportunity to gather and provide that
3-7
-------
information. In assessing citizen complaints, and the
information provided by citizens, States should be mindful that
individual citizens and community groups may not have the
resources available to regulatory agencies, or industry. The EPA
will serve as an information resource for States and citizens,
and provide technical consultation regarding health effects, risk
analysis, ambient air concentrations, monitoring, modeling, and
other issues, if requested.
After the State completes its assessment of the potential
health risk of an emission peak, it; may determine one of three
things, based on the frequency, magnitude, and nature of 5-minute
peak concentrations in an area: (1) corrective action is needed;
(2) corrective action is not needed; (3) more information is
needed to reasonable determine if corrective actions is needed.
The EPA expects that local citizens and community groups will be
kept informed during the decision-making process, be informed of
the factors and information used to support the decision, and be
given and opportunity to comment if they disagree with the
decision.
If the State decides that corrective action is necessary,
the recommended corrective action should be developed through a
collaborative process involving the State, industry, and the
local community.
3-8
-------
SECTION 4. COST ANALYSIS
4.0 Potential Actions and Costs associated with the IL Program
The IL program is designed to give wide discretion to States
and local areas for the implementation and enforcement of the
program. There are three steps to estimate the total cost to
society of the IL program: (1) develop an estimate of the
frequency of exceedances of 0.60 ppm or higher that occur across
the nation, (2) predict the number of actions taken by States
that would result from these exceedances, and (3) apply the
appropriate cost of control for the action to arrive at an
estimate of the total cost to society of the program. Because
the IL program provides a large amount of flexibility for its
implementation, significant uncertainty exists in a cost analysis
of the program and it is not possible to complete all three steps
above.
The following analysis presents information on the number of
exceedances observed in the country based on best available data,
and the EPA's best judgment of the number of actions that will
occur. The control strategies that can be used in actions taken
can vary widely from a low cost alternative such as fuel
switching to a very costly alternative such as the installation
4-1
-------
of add-on control equipment. Because of the huge uncertainty
surrounding the control strategies to be chosen for an action, a
national cost estimate is not provided. Alternatively, this
analysis evaluates the cost of control through a series of case
studies that present information on a sample of control
strategies that are viable under the IL program. The types of
actions and control strategies analyzed are not exhaustive,
however. Time and resource constraints prevent an analysis of
all possible control alternatives. In addition, States and local
communities, while evaluating a 5-minute S02 problem, may develop
new and innovative ways of addressing S02 concentrations.
4.1 Number of Exceedances
As is discussed in the staff paper supplement1 and the 1994
reproposal of the health standard", the occurrence of short-term
peaks of S02 are relatively infrequent and highly localized
around point sources of S02. In 1993 and again in 1994, EPA
requested that States collect and submit 5-minute S02 ambient
monitoring data from source-based monitors. Available data have
been compiled and statistical parameters calculated in a report
for the EPA by ICF Kaiser, Inc3.
The monitored measurements submitted for the analysis were
evaluated for the maximum concentration occurring in any 5-minute
block of an hour. The data indicate that concentrations of S02
occur in a range from 0.0 ppm to greater than 2.5 ppm. The
4-2
-------
number of observations recorded at any monitor ranged from 308 to
48,795 hours, with the mean number of observations equaling 7,641
hours (Note that a complete year of hourly maximum 5-minute
-.averages would contain 8,760 observations). There were 63
^monitors, located in 16 States, with continuous data sets of
- either the maximum 5-minute block average per hour or all of the
5-minute block averages per hour. For data sets containing all
of the 5-minute block averages per hour, the maximum 5-minute
block average for each hour was extracted and that parameter was
used throughout the analysis. Of the 63 monitors, 27 (or 43
percent) registered one or more concentrations greater than the
proposed concern level of 0.60 ppm S02 during the time periods
represented for the monitors involved. Of the 27 monitors that
recorded exceedances of 0.60 ppm, the number of such exceedances
ranged from 1 to 139, which corresponds to 0 to 3 percent of the
hours represented in the data. Of the 27 monitors measuring at
least one exceedance, 12 monitors recorded from one to five
exceedances, while eight monitors recorded from 25 to 139
exceedances. Figure 4-1 displays the distribution of hourly
maximum 5-minute SO2 peaks that exceed 0.50 ppm.
While these data came from source-based monitors, the
existing S02 monitoring network is designed to characterize
ambient air quality associated with 3-hour, 24-hour, and annual
S02 concentrations rather than to detect short-term peak S02
levels. This could have resulted in underestimates of the
4-3
-------
Hourly Maximum 5-Minute Concentrations (ppm)
Figure 4-1. Distribution of 5-Minute Exceedances
maximum 5-minute block averages recorded. Therefore, changes in
monitor siting and density near SO2 sources most likely to
produce high 5-minute peaks could increase both the number of
exceedances and the concentrations of the maximum 5-minute block
averages recorded.
4-4
-------
4.2 Number of Predicted Actions
The EPA received varied comments from industry groups,
States and communities for the 1994 proposal of the
implementation of a new NAAQS under CFR Part 51. Some commenters
stated that several sources are already well controlled and a new
JSAAQS would require redundant controls at sources that do not
affect a substantial population because they are located in rural
areas. Other commenters applauded the proposal of a new NAAQS,
as it would resolve problems with frequent exposures to short-
term S02 episodes. Still other commenters acknowledged the
presence of 5-minute peak S02 concentrations, but indicated that
the episodes occurred during hours of the day in which the
at-risk population would not likely be participating in
activities that induce oronasal breathing and, therefore, there
was little risk to public health. Additionally, many States were
concerned that the administrative burden imposed by a traditional
regulatory program where risks to public health were minimal,
might adversely impact their ability to effectively implement
programs for other pollutants.
Based on public comments received and the detailed
.evaluation of existing monitor data submitted by States, EPA
estimates that a total of ten areas throughout the country have a
potential to be evaluated for the level of public health risk
4-5
-------
associated with short-term S02 episodes3. Several of these areas
show indications that the risk to public health would not warrant
action under the IL program due to the frequency and/or
concentration of the peaks, the location of sources vis-a-vis
population, and the time of day of S02 peaks. For instance, a
source may have a superior record of controlling S02 emissions
and complying with the current NAAQS, but has an unusual process
malfunction that is recorded by a nearby monitor as a 5-minute
peak concentration greater than 0.60 ppm. After conferring with
the source, the regulatory authority in this instance may decide
that due to the infrequency of such malfunctions action is not
warranted under the IL program. Alternatively, a source located
in a rural area that attains the current NAAQS but has regular or
repeated 5-minute peaks may not have action taken because of a
low potential for exposures of concern. Similarly, if 5-minute
S02 peaks occur at night, it could be determined that the
potential for exposure to the at-risk population is low and no
action needs to be taken.
Overall, of the ten areas indicated as having a potential
short-term S02 problem by ambient monitoring data, EPA reasonably
estimates that action under the IL program could be warranted for
approximately five areas. In making this judgment about the
Because the IL program is designed to be implemented at
the States' discretion, this document will not present specific
information that implicates an area as violating the concern
level of the IL program and thus prescribe to the State when and
what type of action should be taken., if any.
4-6
-------
likelihood of action under the IL program, EPA is using several
types of information, including: 1) historical knowledge about
the situation based on interactions between the EPA Regions,
--States and local sources; 2) comments provided in response to the
^original proposals by sources, States, and local agencies, which
tnot only provide information about the situation, but also the
regulatory agency's likely response (because in this assessment,
EPA is not only making a provisional judgment about the potential
public health risk engendered in these situations, but also is
trying to gauge the responsiveness of the regulatory agency in
charge); 3) air quality and census data; and 4) information about
"the industrial processes at facilities in the locations of
concern.
It should be noted, however, that the uncertainties
-surrounding the estimate of actions to be taken for the IL
program are tremendous. One major restriction in the ability to
provide a clearer estimate of actions is the fack of data. As is
stated above, EPA has evaluated data for 63 S02 monitors in the
existing network. This represents only a tenth of all SO2
monitors in operation. The collection of 5-minute data has only
recently been undertaken by some States due to interest in a
potential short-term S0: regulation. Previous data collection
efforts were to demonstrate compliance with the 3-hour, 24-hour,
and annual NAAQS and as such does not provide sufficient
information on 5-minute peaks. Additional uncertainty exists on
4-7
-------
how the states will prioritize areas for monitoring the public
health risk associated with a specific area, and how the
negotiations between the State, source, and citizens will result
in remedial action.
In addition, as monitors are relocated to better measure
5-minute S02 concentrations, additional actions for the IL
program could result. This outcome would indicate that the
current estimate of the number of actions taken for the IL
program is underestimated. At another extreme, due to budgetary
constraints, a State could set priorities for environmental
actions based on the severity of the problem and decide that
other issues such as particulate matter and ozone will utilize
all available resources. This decision would result in little
effort applied to the IL program and consequently zero actions
would be taken. If this happens, then the estimate of five
actions taken nationwide could be an overestimate.
4.3 Estimate of Costs per Action Taken
In the previous regulatory impact analysis of proposed
implementation plans for a new NAAQS (regulatory option 1) or a
program under Section 303 (regulatory option 2), the cost
analysis assumed that if an area indicated exceedances of 0.60
ppm at any one time during a year, then controls would have to be
installed at sources contributing to the problem. In that
analysis, the costs estimation was based on a worst case
4-8
-------
assumption that add-on control technology, such as SO2 scrubbers
would be applied to resolve short-term S0: problems. The upper
bound of total cost to society in the analysis was estimated to
toe $1.75 billion (1993 dollars) based on the cost of implementing
.a new short-term S02 NAAQS with 1 allowable exceedance4. The
fcost of the Section 303 option was demonstrated to be more
-.flexible as far as control alternatives, and therefore, was
assumed to cost significantly less than a new NAAQS. In a
supplemental analysis, the cost of implementing the Section 303
option at two model utility sources was evaluated for several
control alternatives, which demonstrated the wide variance in
potential cost of a more flexible regulatory alternative.
The IL program is also proposed as an option under the
authority of Section 303, but provides substantially more
flexibility for its implementation as compared to the previously
proposed regulatory option 2. While the regulatory options
described in the previous proposal would be implemented to all
areas of the country that show one exceedance of 0.60 ppm, the IL
program is to be implemented locally by States or local
regulatory agencies based their assessment of public health risk.
Control alternatives which may be considered to resolve a short-
term SO; problem include, but are not limited to' :
b The list of control alternatives is not exhaustive and
the EPA anticipates that given the flexibility of the IL program,
the States and sources will develop new and innovative ways to
control for short-term S0_.
4-9
-------
• additional add-on control equipment,
• intermittent control technology to reduce emissions
during 5-minute peak episodes,
• improved operating and maintenance procedures,
• various.dispersion techniques, and
• switching to combustion fuels with low sulfur content.
The EPA has observed that each scenario of potential action
under the IL program is unique based on the types of sources
involved, the concentration of emissions, the frequency of
emissions, the geographical surroundings and meteorological
conditions of the area, and concentration of population.
Although in EPA's best judgment, five actions under the IL
program will occur, the choice of control strategies chosen in
each action is dependent on the negotiation of resolution between
the regulatory authority, the source, and the community. When
taking action under the IL program, the regulatory authority
could (for reasons specific to the situation) insist on the use
of add-on control equipment to remediate the 5-minute problem, or
they could provide flexibility to the source to propose an
innovative solution to the problem. Because of the huge
uncertainty surrounding the control strategies to be chosen for
an action, it is not feasible to estimate the total cost of the
IL program. Alternatively, this analysis evaluates the cost of
control through a series of case studies that present information
4-1C
-------
on a sample of control strategies that are viable under the IL
program.
Appendix A presents detailed analyses of seven case studies,
Awhile a summary of each case study is provided in the immediate
-pages that follow. In five of the case studies, the State
decides that the risk to public health warrants action under the
IL program, while the remaining two studies demonstrate
situations in which short-term S02 emissions were evaluated by
the State but no action is taken.
The selection of the sources and actions investigated in the
case studies is primarily based on data availability0. The
studies utilize information from the report of monitoring data
along with prior studies conducted by EPA and public comments
received with regard to prior S0: proposals. The case studies
attempt to evaluate a variety of industries that are known to
emit SO2, but the selection of these industries does not indicate
EPA's intent to target any particular industry for control. In
addition, the method of evaluation and control strategies that
are discussed should not be viewed as guidance on how the IL
program should be implemented. Supplemental guidance documents
c While this report presents five case studies of action
under the IL program, it is not intended to correlate with the
total estimate of five actions presumed to be taken nationally.
The selection of case studies was independent of the
determination of the total number of actions to be taken for the
IL program.
4-11
-------
for the program will be issued by the EPA in the future.
4.4 Case Studies
Table 4-2 presents a summary of the case studies prepared
for this analysis. The table displays the type of source
evaluated, whether action is taken and why, the control strategy
imposed and the total annualized cost in 1993 dollars. Specific
summaries of each case study is provided below.
Case 1
The first case study evaluates one source whose 5-minute S0:
emissions exceed 0.60 ppm, which are impacting the local
community around the source. The study evaluates a typical
copper smelter facility that is located in a valley which creates
frequent thermal inversions, thus trapping emissions in the
valley for prolonged periods of time. Based on a statistical
distribution of available monitoring data at copper smelters,
there are 74 exceedances of the concern level (0.60 ppm),26
exceedances of 1.0 ppm, and 34 exceedances at the endangerment
level (2.0 ppm).
During the evaluation of the problem, it was discovered that
the exceedances were seasonal in nature, occurring primarily
between the months of September and February (which contributes
to the conclusion that the exceedances are associated with
4-12
-------
Table 4-2. Summary of Case Studies
Case
Study/Source
Type
1 . Copper
Smelter
2. Paper Mill
3. Lead
Smelter
4. Petroleum
Refinery
5. Multiple
Sources
6. Several
Coke Oven
Facilities
7. Utility
Action Taken and Why
Yes - one source impacting small
community with frequent violations of
the concern level.
Yes - one source impacting moderate
size community and school aged-
children frequently.
Yes - one source impacting small
community at the endangerment level.
Yes - one source impacting several
populated communities across State
borders.
Yes - several sources deteriorating
community ambient air at levels greater
than 0.60 ppm.
No - exceedances occur at night when
people not exercising.
No - rural location of facility does not
present risk to population.
Control
Strategy
Two taller
stacks
Double
Contact Wet
Scrubber
Packed Bed
Scrubber
Continuous
monitoring
and Dry
Scrubber
Trading
program
N/A
N/A
Annual Costs
(1993 dollars)
$1,870,000
$1,150,000
$344,000
$2,224,000
$243,029 to
$280,964
Minimal for
monitoring
$0
thermal inversions). The source was assumed to be adequately
controlled to attain the current NAAQS. As the process was
already controlling for emissions, most S02 is already removed
from the emission stream. The addition of add-on controls to
ensure against future exceedances of the concern level would be
redundant and result in costs comparable to the original control
equipment yet remove relatively less SO , yielding a
4-13
-------
prohibitively high measure of cost-effectiveness. The source
recognizes that compliance with the NAAQS precludes consideration
of stack heights greater than Good Engineering Practice (GEP),
which is 213 feet or 65 meters. However, the IL program as
proposed under Section 303 of the CAAA permits the use of
intermittent controls such as greater stack heights as long as
the source continues to comply with ambient air requirements at
the permitted stack heights. The State and source agree that
taller stacks would be used during the period of the year likely
to produce thermal inversions. During warmer months the taller
stacks would be used during stagnant weather conditions only. As
a result, costs of constructing two new stacks at the facility
are evaluated to total $14 million in capital costs, which
equates to $1.87 million annually.
Case 2
The second case study evaluates short-term S02 emissions
from a paper mill that impacts a populated community that had
submitted complaints of a shortness of breath to the State. The
source of short-term S02bursts is from the ending of the sulfite
pulping digestion cycle which is a batch operation at the
facility. The cycle usually runs for 6 hours and then emissions
are vented over a 5 to 10 minute period. In addition to the
local community affected by the emissions, the facility is
located adjacent to an elementary school, so the school yard
receives a large portion of the short-term emissions. Monitor
4-14
-------
data demonstrates frequent exceedances of both the concern and
endangerment levels. Because of the numerous exceedances of the
endangerment level and the impact on school-aged children, EPA
sdecided to work with the State to evoke prompt remediation of the
public health risk.
The addition of a double contact wet scrubber (along with
retrofitting of the digester) was the only alternative available
to resolve the air quality problems. The cost of rebuilding the
digestor to accept the scrubber, and installing the scrubber, is
calculated to be $9.45 million in capital costs, which is
annualized to be $1.15 million per year.
Case 3
The next case study evaluates a single source impacting a
less populated area than that of case 2, but the existence of
frequent exceedances of the concern and endangerment level
coupled with violations of the NAAQS raises concern with the
State as to the areas public health risk. The State first
investigates improvement in public health risk that can be
achieved by attainment of the NAAQS, and discovers that tighter
adherence to the current SIP requirements will not provide
adequate protection of the short-term ambient conditions. In
addition, the source is located in a hilly terrain, so the
concern that thermal inversions could cause ambient S02 to stay
at elevated levels for prolonged periods of time also existed. A
4-15
-------
primary lead smelter was modeled for this case study and the
source and State negotiated the installation of add-on control
equipment to the blast furnace. Specifically the analysis looks
at the addition of a packed bed scrubber to the blast furnace
exhaust. The cost of installing the scrubber and modeling its
effectiveness is calculated to be $0.28 million in capital costs
or $0.344 million annually.
Case 4
The fourth case study evaluates a petroleum refinery located
in a populated area in which the State has received numerous
complaints of asthma and respiratory difficulties, plus burning
eyes and throats. In addition, emissions from the source are
known to transport across State boundaries to another community
close to the facility. Coordination of both States with the
facility is required to remedy the problem. During the
investigation the States found that the facility has an old piece
of equipment that has been grandfathered from control
requirements. Even with this uncontrolled equipment, the
facility usually complies with the NAAQS, however, to resolve the
instances when the NAAQS is violated the source is required to
practice additional monitoring plus operating and maintenance
(O&M) practices to eliminate excess emissions. However,
exceedances of the concern level are projected to be
approximately 150 per year, with several additional exceedances
at 1.0 ppm and 2.0 ppm. To resolve the problems that remain and
4-16
-------
trigger the IL program, the source is asked to use the currently
installed Continuous Emission Monitors (CEMs) to provide
continuous data (which is incremental to hourly data provided to
-show attainment of the NAAQS) for exceedances of the IL program,
*and report why exceedances occurred plus show O&M practices in
...place to avoid future exceedances. The cost for increase O&M
^practices and reporting and record keeping of the CEMs totals
$0.034 million annually. The source is also required to add a
dry scrubber device to the uncontrolled unit at a capital cost of
$20.4 million (or $2.19 million annually). Combining the control
strategies results in a total cost of this solution of $20.5
. million in capital costs, or $2.224 million annually.
Case 5
The fifth case study evaluates an area that has several
industrial sources that contribute to frequent exceedances of
both the NAAQS and the IL program. As a result, the regulatory
authority implements .stricter enforcement of 43IP requirements to
meet the NAAQS, including a requirement for the installation of a
CEMs, and implements additional requirements under the IL
program. The sources impacted by the action include two oil
^ refineries, two sulfur recovery plants that support the
refineries, and a coal burning power plant.
Prior to enforcing stricter SIP requirements, monitor data
indicate an average of 32 instances in an hour of 5-minutes S02
4-17
-------
concentrations exceeding the concern level. After installation
of the CEMs, data show an average of twelve violations of the
concern level in an hour. Thus, it is concluded that the current
SIP strategy does not eliminate 5-minute episodes. Under the SIP
requirements, the sources are required to use the CEMs to record
and report hourly monitor data to show compliance with the NAAQS.
For the IL program, sources are to provide continuous monitor
data to record periods of time when 5-minute violations are
frequent. After 2 years of collection by the regulatory
authority and the sources, the area will implement a trading
program among sources to provide intermittent control during
periods when 5-minute exceedances are likely. Control strategies
considered by the sources to reduce the combined affect on
ambient S02 concentrations include the temporary scaling back of
production and the use of cleaner combustion fuels. The sources
that can control at the least-cost would do so in exchange for
compensation by other sources that do not control.
Costs associated with this strategy include increased burden
for reporting and record keeping of 5-minute continuous
monitoring01 and the cost of a 10-20 percent reduction in
emissions beyond the NAAQS emission limits. A unit cost of $270
per ton of emissions reduced is assumed from the market rate for
the S0: Allowance Trading Program, because a source will either
d However, equipment costs for the CEMs are attributed to
meeting the NAAQS.
4-18
-------
opt to pay an allowance price of $270 per ton or control at an
amount less than $270 per ton. The annualized cost of
monitoring, reporting, and record keeping and a 10 percent
-rollback of emissions is calculated to be $210,855, while a 20
-^percent rollback in emissions totals $248,890 per year. In
-^addition to costs imposed on sources, the case study estimates
-that the regulatory authority incurs $32,173 annually for report
review and compliance assurance. Therefore, the total cost
associated with the case study is from $243,029 to $280,964 per
year.
Cases 6 and 7
The case studies presented to this point have demonstrated
* some situations in which there was a need for implementation of
the IL program and have discussed the costs associated with
implementation. In the two case studies that follow, the
regulatory authorities (State or local agencies) investigate
situations where exceedances of the concern level are known, but
as a result of a simplified risk assessment, they have determined
that the risks associated with the violations were not
significant enough to warrant action under the IL program.
In the sixth case study, monitor data are evaluated for a
Metropolitan Statistical Area (MSA) with three coke oven
facilities (small, medium, and large) in close proximity to each
other. Although there is only one recorded violation of the
4-19
-------
NAAQS in the past 4 years, exceedances of the concern level of
the IL program can occur during shut-downs or malfunctions of the
desulfurization plants at these facilities. A review of the
monitor data indicated exceedances in the area ranging from 0.60
and 1.0 ppm with a majority of exceedances occurring around 0.80
ppm. A total of 68 exceedances were recorded during 29 hours.
Fourteen of the hours that recorded exceedances did not have
desulfurization plants in operation at some of the facilities.
With concern for the number of exceedances that could affect the
highly populated area, the regulatory authority (a local agency)
closely examined the data to determine if action under the IL
program was necessary. During the investigation, they discover
that 55 percent of the hours with exceedances were between the
times of 11:00 p.m. and 5:00 a.m., while nearly 80 percent of the
exceedance hours occurred between 9:00 p.m. and 6:00 a.m. In
addition, the local agency or State had not received any citizen
complaints pertaining to a short-term exposure to S02. As a
result, the local agency concluded that the public health risk
from exposures of concern is very low due to time of day when
these peaks occur and thus, action was not warranted under the IL
program. However, they did decide to continue to review monitor
data quarterly to ensure that public health risk did not increase
significantly. While there are no costs associated with remedial
action for this case study, the regulatory authority would incur
a minimal cost associated with the risk assessment and to conduct
a quarterly review of data. The EPA estimates the cost to be
4-20
-------
minimal ranging between a tenth of a man-year and a fourth of a
man-year.
In the seventh case study, a local agency evaluates the
e
potential impact of a moderate size coal-fired utility power
plant on the local community with a population of less than
1,000. Over a 1-year period, monitor data indicate a total of 10
exceedances of the concern level, but information provided by the
source indicated that the peaks lasted less than 5 minutes in
duration because of the flat terrain around the source and the
quick dispersion of emissions. With this information, the
regulatory authority concluded that the risk to public health in
the area was low, no action would be taken under the IL program,
and that no further investigation of monitor data was necessary.
Other Studies
There are several other control techniques that could be
evaluated to measure the cost of an action under the IL program,
however, time and resource constraints preclude any additional
analysis. There are, however, two analyses that were conducted
for other programs that are worth noting in this document. The
first analysis is contained in a memo from the Office of Air
Quality Planning and Standards to the Office of Management and
Budget in response to comments on the 1995 proposal of the
regulatory option implemented under the authority of section 303
of the CAAA. In this analysis, various control techniques such
4-21
-------
as the installation of taller stacks, increasing the capacity of.
an existing Flue Gas Desulfurization unit (FGD scrubber),
installing a new FGD unit, and stack gas reheat. The analysis
considers controls for two sizes of utility sources (100 megawatt
and 1,000 megawatt boilers). The construction of new taller
stacks was found to be the least costly alternative, while the
installation of a new FGD unit was the most costly. The annual
cost for the small utility ranged from $0.2 to $6.9 million,
while the cost for the large utility ranged from $0.4 to $30
million.
In addition, SAIC conducted an analysis of fuel switching on
industrial boilers from fuel oil to natural gas. An example of
the analysis that is provided below concludes that with current
prices of fuel oil and natural gas, sources with the ability to
switch between these fuels could achieve incremental S02 emission
reductions at a cost savings. An example of a fuel switching
analysis is provided below.
Many oil fired boilers are equipped to burn both fuel oil
and natural gas through burners designed for use of both fuels.
For these facilities, it is often a matter of current fuel costs
that dictate the choice of fuels. In recent years, fuel oil
costs have remained slightly higher than natural gas costs making
natural gas the preferred fuel. As natural gas prices are
typically higher during the winter heating season, the ability to
4-22
-------
burn fuel oil during the winter is an advantage. Facilities that
have the ability to interrupt gas usage during peak consumption
periods and operate on a secondary fuel can typically purchase
gas on an interruptible basis at substantially reduced costs.
For facilities that are already equipped to switch fuels, this
change can be accomplished quickly and easily and can be done
based on current market fuel prices or fuel availability.
This example assumes a boiler at a source is currently
equipped to burn only fuel oil. Therefore, the cost of
converting a 165 mmBtu industrial boiler to dual-fuel (natural
gas and fuel oil) capability is examined. The boiler is
currently fueled by #4 fuel oil with a maximum sulfur content of
0.5 percent. For the purpose of this example, it is assumed that
the boiler is located in an ozone nonattainment area and that
low-NOx technology is required to comply with local air quality
regulations. As the cost of boiler replacement is relatively
low, this provides a reasonably high-end cost for conversion from
fuel oil to natural gas. The low-NOx technology employed in this
example is a low-NOx burner combined with flue gas recirculation
(FGR).
The initial capital cost of retrofitting the boiler of
$288,685 is based on vendor quotes. Assuming a 25-year equipment
life expectancy and 7 percent interest rate, the capital recovery
factor is 8.6 percent. This results in an annual cost for,
4-23
-------
retrofitting of $24,772.
The cost estimate was simplified by assuming that all
operational costs remained the same with the exception of fuel
costs. The boiler is assumed to run at approximately half
capacity for 8760 hours per year for both fuels. This resulted
in annual fuel cost of $5,911,578 when burning natural gas and
$6,193,894 for #4 fuel oil. An annual fuel cost savings of
$282,316 is produced by this switching of fuels. It should be
noted that the unit fuel costs ($3.25/cubic foot for natural gas
and $4.67/gallon for fuel oil) upon which these calculations are
based can vary with time and location. At the present time, fuel
oil prices are increasing making natural gas a relatively
inexpensive option.
A total of 378 tons per year of S02 were reduced by
switching to natural gas, based upon 8760 hours of annual
operation. Because there is an annual cost savings of $257,544
(i.e., fuel savings less equipment costs), there is a cost
savings per ton of S02 reduced of $681 resulting from switching
to natural gas fuel. As discussed in the introduction to this
example, this relatively low natural gas cost is based on
interruptible service; this could require periods of operation on
fuel oil that would decrease the overall savings by a minimal
amount. As stated above, variation in fuel costs will heavily
impact the actual cost per ton of S02 removed for a specifLic,
4-24
-------
facility.
4.5 Summary
This section provided information on the number of known
exceedances of the concern level of the IL program, and gave the
best estimate of the number of actions that will be taken upon
promulgation of the IL program. The occurrence of high 5-minute
peaks was demonstrated to be a unique scenario for each of the
case studies presented. Control alternatives for each case study
were evaluated in depth to determine the cost of implementing the
control at a source(s). The case studies indicate the range of
annualized cost for solutions to different 5-minute S02 problems
to be from approximately $300,000 to $2.2 million. In addition,
some case studies have no cost associated with the program since
action is not taken. Yet, other studies indicate the potential
for a cost savings of $257,544 or a total annualized cost of $30
million.
Because of the significant uncertainty surrounding the
determination of the total number of actions to be taken and due
to the wide range of potential costs associated with various
types of control alternatives, any attempt to present an estimate
of the total cost of the IL program in this analysis would be
meaningless. One could argue that an average cost of an action
could be determined based on the case studies provided, however,
this too would not provide a reliable estimate of total cesJ: of
4-25
-------
the IL program. While the case studies attempt to present
information on a representative sample of outcomes of an IL
action, time and resource constraints preclude the evaluation of
every variation of control alternatives.
This analysis demonstrates that the IL program provides a
significant amount of flexibility to regulatory authorities,
communities, and sources to achieve a reasonable solution to
short-term S02 problems at a substantially lower cost than other
potential regulatory vehicles. For example, the previously
proposed regulatory option of establishing a new short-term S02
NAAQS to eliminate exceedances of 0.60 ppm at any one time in a
given year was estimated to cost $1.75 billion. Several of the
sources assumed to incur costs under a NAAQS option would have
the potential to not have any regulatory action taken upon them
under the IL program and thus incur no compliance costs. Even if
all five of the actions predicted to occur under the IL program
have the highest end of costs estimated in the case studies of
this analysis ($2.2 million), the total cost of the IL program
would be $11 million, or $1.74 billion less than the NAAQS
option. Therefore, the IL program is a very cost-effective
solution to the public health risk associated with short-term
peaks of S02.
4-26
-------
REFERENCES
1. "Review of the Ambient Air Quality Standards for Sulfur
Dioxides: Updated Assessment of Scientific and Technical
Information, Supplement to the 1986 OAQPS Staff Paper
Addendum". U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards; RTP, N.C. Document No.
EPA/452/R-94-01, March 1994.
2. National Ambient Air Quality Standards for Sulfur Oxides
(S02)-Reproposal. U.S. Environmental Protection Agency.
(59FR58958-58980), November 1994.
3. Summary of 1988-1995 Ambient 5-Minute S02 Concentration
Data. Prepared by Systems Applications International under
subcontract to ICF Kaiser, Inc. for the U.S. Environmental
Protection Agency; RTP, N.C.; September 1995.
4. Regulatory Impact Analysis for the Proposed Regulatory
Options to Address Short-Term Peak Sulfur Dioxide Exposures.
U.S. Environmental Protection Agency; RTP, N.C.; November
1994.
4-27
-------
SECTION 5. ECONOMIC IMPACTS
5.0 Introduction
Analyzing the economic impacts of the alternative S02
regulatory options on non-utility sources is a very difficult
task. The set of sources potentially affected by the IL program
is quite broad, covering a wide variety of industry sectors in
the U.S. economy from Standard Industrial Classification (SIC)
codes 13 through 38, which encompasses any source that uses
fossil fuels such as coal, fuel oil, coke, or natural gas to
generate power in a boiler or to generate heat in a production
process. Several industrial sources and electric utilities use
boilers to generate power for process operations or for sale to
end-users. In addition, several of the metal industries (i.e.,
copper smelting, lead smelting, coke production) release S02
emissions during batch processes, such as when metal ores are
heated to very high temperatures to extract certain properties
from the ore. The Pulp and Paper industry is another industry
that emits S02 at the end of a batch process from the sulfite
pulping digestion cycle of the production process.
The breadth of industries potentially affected by the IL
program precludes the usual depth of coverage of a traditional
5-1
-------
economic analysis. In this analysis, market characteristics of a
sample of industries potentially affected by the program are
presented and discussed briefly. Then a qualitative discussion
of facility impacts is provided in Section 5.3.
5.1 Industry Background
Characteristics of a sample of the industries potentially
affected by the IL program are presented in Table 5-1. to
identify the magnitude of potential impacts on the affected
industries. Several factors such as the number of facilities and
companies in an industry, availability of product substitutes,
international competition, plus historical sales and employment
are used to evaluate each industry's level of competitiveness
(and thus industry structure), and stability to determine how
additional compliance costs will impact the industry3.
The last element provided in Table 5-1 to characterize an
industry is a determination of the existence of small entities.
Company employment levels indicate the existence or absence of
small entities in the industry. When data on company employment
levels are unavailable, a determination of the existence of any
a Other factors specific to a particular industry such as
expected future growth or expected new markets are also
considered in the determination of industry stability, but are
not included in Table VI-1. For more information on a particular
industry, refer to Reference 1: "Industry Profiles for
Term National Ambient Air Standard of Sulfur Dioxide."
5-2
-------
small entities is based on the percent of companies owning only
one establishment (versus companies owning multiple
establishments that tend to have higher employment levels) and
employment levels of the majority of facilities in the industry.
The information presented in Table 5-1 indicates that most
of the industries analyzed at the 2-digit SIC code level have few
substitutes available and international competition is prevalent,
limiting the ability of some industries to recover compliance
costs (and thus limit economic impacts) through increase pricesb.
In addition, practically all of the industries have small
entities.
Four industries, including food and kindred products,
chemicals and allied products, rubber and miscellaneous products,
and electronic equipment do not have close substitutes available,
which gives an indication of a minimal economic impact to be
anticipated. Although, these industries also face a competitive
international market that could increase their impacts0, the
markets of these industries are experiencing growth in the U.S.,
which limits the influence of foreign competition. Other
b The trends indicated in the table may occur because of
the broad scope of analysis of these industries at the 2-digit
SIC code level.
c The electronic equipment industry has international
competition, but the U.S. has strong foreign market •**—
opportunities.
5-3
-------
~«
tn
HI
W
«
o
DM
Q
Cd
EH
U
W
W
CO
^ CO
i 5
in i_i
TABLE
tNDUSTR!
ki
0
CO
U
LRACTERIST]
a
M
U
' 2 Q H
oo S W U
CM ft HI D
ug3§
CO £2 ftl
u
i «, w
S K H B
S r4 Q
O ^ f-1 O
HI ry rtj PH
CO fti
1 CO
W
CM 1-3
CM HI
H
U X
HI H
CO EH
Q
W
0£
1 Q W
2 HI
0 HI U
CM 5£ D
Q
U ^ O
H OS
CO Q (Xi
O
0
tj
SIC 13 -
OIL & GAS
EXTRACTION
H-t
O
EH
CO
H
W
H
U
2
jj
2
u
r>H
P£
H
CO
D
Q
•z
H
CO J2 ri M w
0 co i w S
*. A >C >H
csi '"
CN IT) rf}
CH rH O ^ O
CM c\j S; S S
«. ^
VD ^
VD 00 O ^ O
o en £ £ 2
CO CM
CD (T) -^ CO CO
LO VD ^ b3 [£J
O IT)
CM rH
0)
.Q
10
0 H
rH -rH n CO 0
cn (0 2 H g
> ^ >H ^
CM (fl
CM
JJ
O
c
EH
•3 H §
J2 fd
<0 C p. S
— H 0 C 0)
, < 14_( [ — I4_| , .p.) .^ ^3 ^3
coocoor~ nj+J So
O cn oo > (0 X OS
MMrHMcTi rt|c! rffi
EHll)^'O ftCO
WOS 30) 3W H4J C-H Mw
££ W 2 •*-! 2(U O3 H4-J ff
Oi EH 4-> -H J-> -H W CLi
2 O 1 -H 1C 1 -H 14-1
2 Kt r-H (0 4J 0)
a -H ft w a,
« 0 g X! g
• S r8 0 3 0
M U tu U CO U
2
H UD CM UD ><
CO ^ C^ *3* O
rt| oo "^ r^ ff,
til ^ - ^ U
K CM >£> CO EH
u m oo o w
2 CM CM ro
U r- cn cn >H
iJ CM rH O Q
co co ^r in c£
^ ^ v ^ H
EH 00 CTl CM EH
CO rH CM CO CO
rH rH rH
3 rH t- m H
^ co ^r in H
^ O" rH in U
u' r~ j~- co td
Q
15
^
M rH r- CM >H
DJ v ^ v H
a m co o EH
U rH 03 O CO
2 "3* CO "31
M
0)
rH
w -2 S
J r~ cxj _™ g
M ° °° IH 2
PH r^ "^* ^
5 TO HI
« « S 0
0 0 r-f « ^
4-J Q
o
c
m IH >n
o o o
0) 0) 0) 4->
3 3 3 C
H H iH 0)
fO fO fO E
>H > ^ > — > ^ &
CC H rH rH O
EH CM^-r-— rH— H
CO •• 00 00 C^ Q<
D>< cnw cnw cnw E
QEH rH4-> rH4-> rH4-> Cd
2 H C C C >i
H hq 1 tt) ID 10) 1 M
HI g g g 0
OQ ft fi fi 4J
• 5; -H -H -H w
HI EH JS XI .C -H
H CO CO 10 CO S
\
2
CO
S
•^
2
CO
u
><
CO
ca
>H
w
td
H
EH
M
E"
£5
C£l
J
i_J]
^
Jj?
CO
t*.^.^
»
M
M
M
LD
-------
CO
H
CO
><
os
O
Q
Cd
•a u
3 J
C Cd
•H CO
C CO
BLE 5-1 (co
F INDUSTRIE
ft. 0
ERISTICS
H
U
1
L)
2
O
H r.
1 EH g
r~ EH £
00 o a
U Cu t
w | I
f' f?
EH
*' 1 H &
^ 2 g
H W g
1
^i<
u 3 M
CO
SIC 30 -
RUBBER AND
ISCELLANEOU
PLASTICS
S
i S; J w
en « § U
CM ri U D
O Q
o g Q o
H g 2 DC
CO g fiS 04
O
H
EH
W
H
fC
Cd
EH
U
(i
fY
f£
3d
U
>H
PC
EH
W
J3
Q
2
H
^ CO
° rH 2 " "
*. ^
rH <*
CM OO
CM CM _. -. CO
en m g g H
in oo
rH 0 ,-
ID 0 O ^ O
^ ^ ^H
us m
en en
00 ^J1 o CO W
in rH g [d [d
*r CM
" ^ w W W
CM rH
CO
W W
C O
r^ — w o u X
co^Mco^tHr^ja -H CD
UOcnoco3 4J EnD
M rH en co (0 2 O
*H *H * M rH C^ fX PC
COO) (1)^ — '(1) MdOX
EnHflW^ W IDOHEH
HOC g(U gM OCI 4-1 -rH to
idCd 3-rH 3(U •— 1 i— 1 C4-1 H
OSEH 24-1 2-H UX! M--H 2
go -H C n) 4-) CD
5; CC 1 rH 1 (0 1 rH 1 0) H
OC -H ft -iH ft CO
g u g (0 g
• X (0 o > O
H u [« o cC u
fd en o rH >H
i-q en oo co C
CO rH ^T rH ft,
ft, •• W
EH CO r- T EH
M en en co co
CM 00 CO
i o
2 >T
H en o m ~
to o en vo i-^
< rH O U) *
DS cn U3 •* ri
O rH 0 0 X
2 CM CM CM g
0
W m en in 2
1-3 rH r- CO 5
CQ o ^D en 2
(^ - i
EH m ^ ^r H
co in co co g
rH rH rH Q
CD O
2 2
H „ «> r- H
n ;!; r- rn co
rf r^ o^ fi
W ° ^ Cd
OC ' 0 CO OS
o ^ o o u
H H
O
H •* VO O 2
i-3 rH ID en g
EH en CM ID f\
co o in in H
CO rH rH Q
JH
EH
H
H^ <4H <4H MH
MO 0 0
cq >,
< OJ 0) 0) M
EH 3 3 3 0
CO rH rH rH 4->
nj . — . ^> — . J> . — -H
(X, rH rH rH X
EH CM — f~ — rH —
co oo co en 4-)
p cnw cnn cnw c
Q I-I4-! H4-1 rH4-l (U
Z G C C g
H 1 01 1 0) 1 0) >i
e e go
ft ft ft rH
•H -H -H ft
H fl X! JC g
H CO CO CO Cd
/It
UJ
H
ja
TO
rH
••H
TO1
n3
0
c
CO
CO
w
^
CO
[d
W
>H
CO
Cd
H
EH
H
EH
J^
Cd
^q
^
J^
co
•
M
M
H
in
i
in
-------
co
H
co
s
2j
K
O
fa
Q
W
0) Cd
3 Hi
C W
•H CO
4J
C W
O W
O HI
f_.
rH CO
1 D
m Q
2
PO HI
CQ fa
i? O
CO
U
H
EH
CO
HI
f£
W
EH
O
2
I?
u
1 CO
rH 5 HI
rH B CC
cn <-3 a
CM O 2
CC HI
O EH fa
H Cd Ld
CO O, «
1
m 2 >,
en o ri
oo m y
u u ffl
CO
1 2
oo U 0
rr, M H
2 <* Q H
£ D H U
-53B
HS §
M Qj
CO W
1
^ cd
rH EH Q, CO
rH HI ^ rH
C^ H^ *r\ H
^ 5^"
U CO
HI
CO
1
ro Cd
S||3
CJ CO
(—1
CO
o
H
EH
co
HI
W
EH
U
rf
5
3C
u
^
c<
CO
£3
Q
2
HI
Cn O O r T r T
o o 2 5 5
CO CM •H •"
w co n
CM ^ W H X
CM ^ >H >H ^
rt\
UJ
<__i
CM t- 0 4J (0 CO
S 5< 2 O H Cd
<£> ^r C 'S ^
(fl
CM CM 2 CM CM
• --.^ ^< ^_* -»_'
cr, ^D ° w °
CO CM Z >H 2
CO
co w <
o 0
M — — < U
13
Cd ^r oo co 2
>-3 t— m en 2
CQ rH o in 3
< - 3
EH ^ ^ C— r~v
CO 00 CO CO H
rH rH rH Q
Cd >H
Hi CO Q
CQ ixi rH co a,
RC rH 00 Id
EH ^ U3 IJ3 EH
CO rH CO
x
w
CQ en o o §
< rH en t- 3
EH oo in in r;
Q
~ co oo rr ~
^-^ • - — * * — '
tH CM
Cd >H
i-q rH o en Q
§CO 'J' CM d,
VO CO CO Cd
EH - EH
co ro *3* in co
in m n
4-1 4-) 4->
dec
in tt) (U
g g g
• • Pi p! Pt ^
?H -rH -H -rH M
EH 43 £ A O
HI M CO CO 4J
i_3 to
H IH IH IH -H
mo o OK
t^ OJ H ITj (B (fl g
O5 > > > >i
H 0
CO CM I- H
CO
Cd
CO
U
CM
•~^-
CO
w
w
Cd
H
EH
H
r_H
2
W
h-3
H]
j^
CO
H
H
H
-------
w
H
W
>H
H
<
rf
PS
o
fa
Q
W
— EH
•o o
0) W
3 i-5
C W
•H M
4J
C OT
O W
U H
— OS
EH
rH C/3
1 D
m Q
2
Pd H
pn fa
5, O
EH
CO
U
EH
M
H
PS
W
U
2
2
p-J
u
. §
H O
H r3
PS W
o 2 s
H g CO
CO H
OS
Oi
2
0
(— |
1 EH
CM y
So
U ^
w g
o
u
o
M
H
OT
H
os
: w
EH
U
f3\
i
£n
o
>H
o^
EH
W
D
Q
2
H
(U
I — 1
-H W W n
^ ro W U ^
0
C
0)
rH
,Q
33
— (0 -rH EH
w n > 4-i
u a; n «; D w
H -H 0) O. W
EH 4-1 -H W 6 <
W -rl C 0> 0 Oj
H H fO 4-> U
os -H a 3 EH
W U g 4-1 rH W
EH (0 O -H rB O
U fa U 4-) C U
ft. (no
K MH MH J3 -iH EH
(I O O 3 4-> 2
U M M C U
tt) tt) 01 M H
H .Q .0 W tt) fa
W g g 0 4-> M
X 3 3 H G 2
OS 2 2 U H O
& fH
Millie/}
•
H
tt)
rH
J2 C5
rP cn o '"j tj
CO CTl CO "Jj ^
g " "t > ^
W rH rH M
4-1 Q
o
c
*•»
w
h4 cs] cji ,
>H -H -iH -rl M
EH .c ^ x: o
H CO CO C/3 4-*
p n
(—( MH MH MH -H
PQ O O O 33
2
EH tt) 0) 4) 4-1
W 3 3 3 C
H rH rH D
>H (0 rd a S
OS > > > >i
EH 0
W CM C- rH rH
^D 00 CD CT1 CX
Q Ol CJ1 Ol g
2 fH rH rH W
H
1 1 1 1
H
M
W
X
M
S
w
X
w
[V)
H
EH
H
EH
2
W
J
g
2
w
.
H
H
H
-H
e
3
ex
(0
MH
3
w
4-1
•H
T3
(1)
T3
H
O
C
-H
W
•H
W
rH
rH
• 3
W tt
n
(0 0)
rH 4-1
rH -H
O MH
•O rH
3
ro w
CD
cn c
tH C
o o
•rH
W 4-1
C IB
o e
•H M
rH O
rH MH
in
-------
industries, including textiles, paper and allied products and
primary metals also do not have close substitutes available, but
because the markets in these industries are mature and expect
little growth in the future, significant competition (both
domestically and internationally) influences the ability to raise
prices and minimize impacts. Therefore, producers in these
industries will probably absorb any compliance costs incurred and
experience significant economic impacts.
The remaining industries analyzed at the 2-digit SIC code
level each have individual indications as to the level of impacts
to be expected. The oil and gas extraction industry faces heavy
international competition and a volatile U.S. market, which
increases the likelihood of significant economic impacts. The
petroleum refining industry faces alternative fuel substitutes
and international competition, however, because the U.S. market
is the largest demander for petroleum products in the world and
imports are only expected to fill the gap between domestic
production and consumer demand, the impact on industry is
expected to be minimal. The transportation equipment industry
does not have close substitutes, and has a strong position in the
international market (particularly in the aerospace sector).
Therefore, the EPA expects that any impacts on this, industry will
be minimal.
The majority of the industries analyzed at the 4-diga-fe-SIC
5-8
-------
code level show indications that significant impact could exist
due to the availability of close substitutes and/or international
competition. The sulfuric acid production industry and the
petroleum refining industry are the exceptions in that the
products of these industries are key inputs to most industries in
the U.S. and there are no close substitutes available. As a
result, these industries are expected to experience minimal
impacts.
Although the electric utility industry is not included in
Table 5-1, facilities in this industry are currently considered
regulated monopolies because power generation and distribution is
limited to a pre-determined area associated with a power plant.
Because they do not face competition, regulated monopolies will
pass on all costs of operation to consumers. If the industry is
deregulated to open competition at the wholesale level (as is
currently proposed by the Federal Energy Regulatory Commission),
then the industry would face competitive decisions for changes in
market price.
5.2 Facility Impacts
-------
a vast number of sources in any one industry discussed above will
be impacted. Although evaluating the industry's potential
response to additional costs of production is valuable, that
evaluation assumes that several producers in an industry face the
same unit compliance costs. With uniform implementation of the
cost on several producers, an industry's marginal producer is
more likely to be affected which causes the market supply curve
to shift. A shift in supply reduces the economic impact on
sources through an increase in prices that allows producers to
recover some of the compliance costs incurred. With the IL
program, there is a potential of only one or two sources of an
industry to incur additional control costs to resolve a 5-minute
S02 problem. If the sources affected by the program are the
marginal producers of an industry, the market supply curve is not
likely to shift and the source would not benefit from increased
prices. Rather, the source would absorb the compliance costs and
incorporate them into the cost of production to determine their
optimal level of operation.
Compliance costs that would be absorbed by a firm are
considered fixed costs because the cost is usually associated
with control equipment, so the level of the cost does not vary
with the level of production. However, in the short-run it is
assumed that because a firm could shut-down and never buy control
equipment, all costs are variable. This results in an upward
shift of the average variable cost (AVC) curve that measu]f«s the
5-10
-------
firms operating costs resulting from the imposition of control
cost. If the AVC is greater than market price (Pm) , then the
firm would decide to not purchase control equipment and
temporarily shut-down operations2.
~s
^ A firm's decision process is somewhat different in the long
run. In the long run, the firm considers the marginal cost (MC)
function, the average cost (AC) function (that incorporates both
variable and fixed costs) and the market price (Pm) , which is
equated to a horizontal demand curve for the firm. If Pm is
greater than the firm's AVC, then the firm would continue to
operate at the same level of production. The figures below
display three potential outcomes of the implementation of the IL
program and its effect on a firm's decision process. First,
panel (a) shows an industry in which the market price is above
the firm's average cost function, and the firm'.s optimal level of
production is at Qf, where marginal cost (MC) equals price (Pm) .
In this scenario, the firm is earning an economic profit because
price is higher than the costs the firm faces. This could occur
in an industry in which the majority of producers had already
incurred the cost of applying pollution control equipment to
their operations prior to the implementation of the IL program,
which resulted in an increase in the market price. , The producers
that did not install equipment have temporarily enjoyed the
benefit of a price that is higher than their cost of operations.
In the second panel, the firm is operating at a level that,
5-11
-------
(a)
(b)
(c)
Figure 5-1. Firm's Position in the Market
Before Control Costs
equates price, marginal cost, and average cost, so the firm is
earning zero economic profits'*. Finally, panel ® shows a firm
that is earning negative profits, but because of a short-run
decision that the firm could still cover obligations on average
variable costs, the firm has continued to operate temporarily.
d Earning an economic profit is not likely to continue
for an extended period of time in a competitive industry because
the existence of economic profits provides incentive for other
firms to enter the market to claim a portion of the profits.
When firms enter the market, the industry supply curve shifts out
and decreases market price and therefore profits. Firms will
continue to enter the market until all firms are earning aero
economic profit.
5-12
-------
MC
AC
(a.1)
(a.2)
(b.1)
(C.1)
Figure 5-2. Firm's Position in the Market
After Control Costs
5-13
-------
Figure 5-2 shows a firm's position in the market after
imposition of the IL program. There are two potential outcomes
of the firm portrayed in panel (a) of Figure 5-1. First, after
the imposition of control costs for the IL program on the firm,
the AC curve rises. If the increase in AC is just enough to
equate with the market price (thus bringing the firm closer to
the average cost in the industry), the firm would be operating at
it's optimal level and earn zero economic profit. Another
outcome, as shown in panel (a.2), could be that the firm's AC
increases to a level above the market price. This would result
in the firm earning negative profits and deciding to close
permanently.
The outcomes for the firm in panel (b) of Figure 5-1 are
limited since the firm is currently operating where AC equals
price and marginal cost. Panel (b.l) shows that the increase in
AC that results from the increase in control costs could cause
the firm to close permanently because average operating costs
(AC) are greater than market price. The firm could reduce AC in
other areas to counteract the additional compliance costs and
potentially remain in the market earning zero economic profits.
The firm in panel (c) of Figure 5-1 who was on the verge of
temporarily shutting-down due to AC exceeding price, would close
permanently according to panel (c.l) if additional costs were
imposed for the IL program.
5-14
-------
5.3 Summary of Economic Impacts
Overall, the impact of the IL program on the industries
described in Section 5.1 is dependent on the effects on the
marginal producer of the industry. It is assumed that the
marginal producer is operating at the competitive equilibrium
that determines market price. If the marginal producer is
impacted by the IL program, the following industries have market
characteristics that indicate a potential to minimize impacts by
recovering compliance costs through increased market prices:
• food and kindred products
* chemicals and allied products
• petroleum refining
• rubber and miscellaneous products
• electronic equipment, and
• transportation equipment.
Firms of an industry who are not considered "marginal" are
implied to operate above the marginal producer6. If the marginal
producer is not impacted by the IL program, then the impacted
firm must be operating at an average cost that is less than
market price and is earning an economic profit. Increased costs
of operation from the imposition of compliance costs will move
the firm's AC up closer to market price which lowers economic
e Firms operating below the marginal producer would
operating at a loss and therefore exit from the industry.
5-15
-------
profit closer to the industry's competitive equilibrium.
Section 5.1 also indicates that compliance costs would more
likely be absorbed by any producer in industries, such as:
textiles, pulp and paper products, metals, and oil and gas
extraction. Because market price would not rise to recover
compliance costs, producers would face operating decisions
similar to the examples provided in Figures 5-1 and 5-2.
5.4 Impacts on Small Entities
Given the uncertainties as to the distribution of cost
across affect sources in the industries, it is not feasible to
interpret the potential impacts on small entities. Table 5-1
indicates the presence of small entities in nearly all of the
industries potentially impacted by the IL program. However, the
cost analysis indicates that the IL program may. impact a total of
5 areas of the country, which lessens the likelihood of seeing an
impact on a small entity. If a action under the IL program is
taken on a small entity, the costs associated with the action can
be quite low if the state allows flexibility in compliance
methods for the program. If action is taken on a small entity in
a declining industry (as indicated in Table 5-1), the impact
could be significant.
5-16
-------
REFERENCES
1. Research Triangle Institute; "Industry Profiles for a Short-
Term National Ambient Air Quality Standa.rd of Sulfur
Dioxide", Prepared under Contract 68-Dl-0143, Work Assignment
number 72; December 1993.
2. Landsburg, Steven; Price Theory and Applications; The Dryden
Press, 1989; pg!42-196.
5-17
-------
SECTION 6. BENEFIT ANALYSIS
6.0 Introduction
The current primary National Ambient Air Quality Standard
(NAAQS) for S02 was implemented to protect against the adverse
health effects of long-term, chronic exposure to S02. As has
been discussed in previous sections, there are certain areas of
the country, where the current NAAQS does not -ensure human health
protection against short-term, acute exposure to S02. Short-term
exposure over a 5-minute period to elevated levels of S02 has
been shown to have adverse health effects, particularly to
exercising asthmatics.
Because short-term peaks of S02 concentrations appear to be
source specific and because of the lack of representative
national monitoring data on these short-term peaks, the benefit
analysis shall use the same model plant case studies as discussed
in the cost analysis. However, benefits are estimated only for
the case studies that indicated action to be taken under the IL
Brogram. Although seven case studies were originally presented
in the cost analysis, only five of the studies had action taken.
Therefore, the benefit analysis is conducted for these five case
study areas. The model plants within these case study areas
6-1
-------
reflect emission characteristics in industries where short-term
peaks of S02 are likely to be a problem. Each model plant is
designed to be a composite of several facilities and does not
necessarily reflect the process or ambient air quality of any
particular plant.
The purpose of this analysis is to outline the steps
required to calculate the health benefits associated with
attainment of a 5-minute S02 standard of 0.6 ppm for the case
study areas. The first section of the chapter contains an
overview of the benefit calculation procedures. Quantitative
estimates of benefits, along with the qualifications associated
with these estimates will conclude the chapter.
6.1 Benefit Calculation Procedures
The benefit calculations for the case studies in this
analysis follow a five step procedure. The first step is to
identify concentration- response functions that quantify the
relationship between short term exposure to S02 and human health
status. These functions can be used to estimate the improvement
in health that may result from a regulatory program designed to
reduce short term emissions of S02. The second step is to
estimate the magnitude of the ambient air quality improvement
associated with the IL Program. The third step is to determine
the population cohorts that will be affected by the improvement
in ambient S02 air quality. Asthmatics are considered tctJae the
6-2
-------
population cohort most likely to be affected by short-term peaks
in S02 levels. The fourth step is to impute an economic value to
the estimated changes in health status for the relevant
population. This step relies on existing studies that have
calculated the willingness to pay for improvements in health
status. And finally, the fifth step combines the information
obtained in the first four steps to estimate the health benefits
associated with the proposed program. Each of these steps is
discussed in detail below.
6.1.1 Step 1; Identify the Relevant Concentration-Response
Functions
Numerous clinical studies have shown that asthmatic
individuals are particularly sensitive to short term exposures to
S02. The SO2 Criteria Document, Staff Paper, and its Addendum
have summarized the results of these studies1'2'3'4'5'6. These
documents suggest that moderately exercising asthmatics are
particularly sensitive to short term exposure (i.e., 5 to 10
minutes) to S02 in the range of 0.6 to 1.0 parts per million
(ppm). (An example of moderate exercise is climbing one flight
of stairs.) Concentrations within this range are likely to
result in lung function changes along with respiratory symptoms
such as wheezing, chest tightness, and shortness of breath.
Although the severity of these symptoms is likely to vary among
the sensitive individuals, the intensity of the response is
likely to be perceived by the individual as a mild asthma^aitack.
6-3
-------
Although some individuals may reduce activity, most individuals
exposed at these levels do not feel such a need and can still
function normally. Medication may be used to mitigate the
effects at these levels.
At short-term S02 levels greater than 1.0 ppm, an
increasingly greater percentage of exercising asthmatics will be
adversely affected. In addition, the responses are likely to be
more severe than those experienced at lower S02 levels. At these
levels non-exercising asthmatics will also begin to be affected.
Effects at these levels are likely to cause overt symptoms and
will probably cause the asthmatic to temporarily cease activity
and/or use medication to alleviate the respiratory symptoms.
Beneath short term levels of 0.6 ppm, the studies suggest
that less than 10 to 20 percent of exercising asthmatics will
experience significant lung function changes and respiratory
symptoms. Although some exceptionally sensitive individuals may
experience effects at these levels, the health effects for the
majority of individuals are unlikely to be perceptible and are
therefore not considered to be of major concern. Based on the
results of these studies, the benefit analysis will use the
ambient concentration of 0.6 ppm as a threshold beneath which no
significant health effects are likely to occur.
6-4
-------
Unfortunately, there is no one clinical study that has
developed a continuous concentration-response function relating
S02 exposure to health status for the range of S02 levels
considered in this analysis. In addition, the concentration-
response functions developed for exercising asthmatics may not be
applicable to non-exercising asthmatics. It is therefore
necessary to select a number of concentration-response functions
to calculate effects for the numerous S02 levels under
consideration. Consistent with the underlying data contained in
the S02 clinical studies, the concentration-response functions
are divided into two categories:
S02 levels greater than 1.0 ppm, and
SO2 levels greater than 0.6 ppm and less than or equal
to 1.0 ppm.
SO2 Levels Greater Than 1.0 ppm
Horstman et al.7 examined the effect of 10-minute exposure
to S02 ranging from 0.25 to 2.0 ppm on the specific airway
resistance (SRaw) of exercising asthmatics. The subjects were
young adults who were classified as "mild1 asthmatics. They found
that the prevalence of a 100 percent increase in SRaw went from
about 56 percent of the subjects at 1.0 ppm to 85 percent of the
subjects at 2.0 ppm. For the remaining 15 percent of subjects
who did not exhibit any significant response at 2.0 ppm, the
6-5
-------
results were extrapolated to predict that all subjects were
adversely affected at 10.0 ppm.
The data contained in this study are used to develop the
following simple linear concentration-response function for
short-term S02 exposure over 1.0 ppm and less than or equal to
2.0 ppm:
% Response = 0.345 + 0.278 S02 (1)
% Response = the percentage of the exercising asthmatic
population experiencing a 100 percent or
greater increase in SRaw
SO2 = 10-minute exposure to sulfur dioxide measured
in parts per million
For S02 levels equal to or in excess of 2.0 ppm, the
following concentration-response function is developed from the
extrapolated Horstman et al. data:
% Response = 0.836 + 0.0169 S02 (2)
Table 6-1 displays the percentage of exercising asthmatics
predicted to have changes in SRaw greater than 100 percent at
various short-term S02 levels above 1.0 ppm.
As previously mentioned, non-exercising asthmatics are also
likely to be affected at short term S02 exposure levels equal
6-6
-------
to or greater than 1.0 ppm. Sheppard et al.8 provides data on
the responses of non-exercising asthmatics to 10 minute S02
exposures of 1.0, 3.0, and 5.0 ppm. At 1.0 ppm, two out of seven
asthmatics at rest developed symptoms (i.e. chest tightness and
wheezing). Five out of seven experienced symptoms at 5.0 ppm.
These data are used to estimate the following dose-response
function for non-exercising asthmatics exposed to 10 minute S02
levels in excess of 1.0 ppm:
% Symptoms = 0.1547 + 0.1071 S02
(3)
Additionally, table 6-2 reports the percentage of non-exercising
asthmatics experiencing symptoms at various S02 levels above 1.0
ppm.
TABLE 6-1
PREVALENCE OF SRAW > 100% ASSOCIATED WITH
SHORT-TERM S02 LEVELS FOR EXERCISING ASTHMATICS*
10 Minute SO, (PPM)
1.1
1.2
1.3
1.5
1.75
2.0
2.5
3.0
3.5
% RESPONSE
0.651
0.679
0.706
0.762
0.832
0.870
0.879
0.887
0.895
* Estimated from Horstman et al. (1986).
6-7
-------
TABLE 6-2
SYMPTOM PREVALENCE ASSOCIATED WITH
SHORT-TERM SO2 LEVELS FOR NON-EXERCISING ASTHMATICS*
10 Minute SO, (PPM)
1.1
1.2
1.3
1.5
1.75
2.0
2.5
3.0
3.5
% RESPONSE
0.273
0.283
0.294
0.315
0.342
0.369
0.423
0.476
0.530
* Estimated from Sheppaid et al. (1980).
SO2 Levels Greater Than 0.6 ppm And Less Than Or Equal To 1.0
ppm--
For this range of S02 levels, the concentration-response
function developed by Abt Associates in a study for the EPA9 will
be used to estimate the improvement in health status. This
concentration-response function was developed from four studies
that examined the effect of short-term exposure (5 to 75 minutes)
to S02 ranging from 0.0 to 1.0 ppm10'11'12'13 and Roger et al.14 The
groups studied were exercising young adults who were diagnosed
with mild or moderate asthma. The following concentration-
response function was estimated:
log odds (Symptom) = -5.65 + 5.89 S02 +1.10 Status
(4)
6-8
-------
where Symptom = any respiratory symptom
such as chest
tightness, shortness of
breath, wheezing,
coughing, etc.
S02 concentrations in ppm for various exposure
periods (5 to 75 minutes)
a dummy variable reflecting asthma
severity; 0 = mild, 1 = moderate.
S02
Status
* The probability of experiencing a symptom can be calculated
by transforming the log odds equation into a probability:
„ , ,„ . e**(log odds(Symptom)
Prob(Symptom) = - -
1+e**(log odds(Symptom
(5)
For a 5-minute S02 exposure level of 1.0 ppm, the
probability of experiencing a symptom is 0.56 for a mild
asthmatic and 0.79 for a moderate asthmatic. The probabilities
predicted for different S02 levels are reported in Table 6-3.
Table 6-3
SYMPTOM PROBABILITY ASSOCIATED WITH
SHORT-TERM S02 LEVELS FOR EXERCISING ASTHMATICS*
SO, CONCENTRATION (PPM)**
0.6
0.7
0.8
0.9
1.0
PROB (SYMPTOM)
MELD ASTHMATIC
0.108
0.179
0.218
0.414
0.560
PROB (SYMPTOM)
MODERATE ASTHMATIC
0.266
0.395
0.540
0.679
0.793
* Estimated by Abt Associates (1996)
** Exposure to the SO2 concentrations ranged from 5 to 75 minutes.
6-9
-------
Underlying Assumptions
A number of assumptions must be made before the above
functions can be used in a benefits analysis. First, it is
assumed that the exposure conditions created to measure effects
in the laboratory environment are similar to that experienced
under ambient conditions. This assumption is especially tenuous
with respect to the Sheppard et al. study because S02 exposure
occurred through a mouthpiece. The actual dose of S02 in that
study was probably higher than that which would occur under
ambient conditions.
Second, the concentration-response functions developed above
are based on exposure durations of at least 10 minutes. It is
assumed that these functions can be used to estimate the health
effects associated with a minimum of 5 minutes of exposure to
S02. The impact of this assumption is likely to be minimal since
studies reviewed in the Second Addendum to the Criteria
Document15 have found that the response to elevated levels of S02
has a. rapid onset and reaches a peak within 5 to 10 minutes of
exposure. Longer periods of exposure while exercising (e.g. 30
minutes) do not appear to significantly alter the initial
response. In addition there appears to be a "refractory period"
during which repeated exposures to S02 result in a period of
diminished responsiveness. The duration of this refractory
period is unclear, although it appears to last no longer than 5
6-10
-------
hours16. The issue of the refractory period will be addressed in
the next section when the air quality data are discussed.
Third, the Sheppard et al. concentration-response function
used to predict effects above 1.0 ppm is based on a sample of
only seven non-exercising asthmatics. In addition, the Horstman
et al. function for levels above 2.0 ppm is based on data
extrapolated from the actual responses observed at levels
beneath 2.0 ppm. The benefits calculated from these functions
need to be viewed in light of these limitatons.
Fourth, the above concentration-response functions were
developed from laboratory conditions that were designed to
examine the effect of exposure to short-term S02 in the absence
of pre-medication with common asthma medications such as cromolyn
sodium and various beta agonists. These medications have been
shown to inhibit responses to S02. If the asthmatics residing in
the case study areas typically pre-medicate to control their
asthma, then the concentration-response functions developed above
will overestimate effects. The extent of this overestimation may
be minimized, however, since evidence suggests that mild
asthmatics typically do not pre-medicate and only use their
medication on an as needed basis. In addition, only about 20
percent of moderate asthmatics use their medication on a regular
basis17.
6-11
-------
Finally, the effects predicted by the above equations are
based on studies that examine the effect of S02 on young adults
diagnosed with mild or moderate asthma. It is assumed that these
results can be applied to the entire asthmatic population. If
certain segments of the asthmatic population (i.e., children,
elderly, severe asthmatics) are more sensitive to S02 than these
equations indicate, then effects will be underestimated.
6.1.2 Step 2: Identify the Improvement in Ambient Air Quality
The regulatory option under consideration establishes a 5
minute S02 average of 0.6 ppm as the target level for control.
To calculate the benefits associated with the program, both
baseline (pre-control) and post-control air quality data are
needed. This step discusses the procedures and assumptions used
to develop these data.
The baseline scenario reflects ambient S02 in the presence
of the current NAAQS and other existing regulations. Pre-control
levels were estimated using monitoring data from ambient
monitoring sites located in areas near actual facilities. These
data were used to design a theoretical Gamma distribution for
each of the model plants. Each probability curve predicted the
likelihood of exceeding the 5-minute S02 concentration of 0.6 ppm
during a
1-year period. The curve was used to determine the number of
times during the year that 0.6 ppm was exceeded and the ^
6-12
-------
magnitude of each exceedance. The probability curve was then re-
estimated to allow only one exceedance of 0.6 ppm during the 11-
year period. The S02 levels associated with this curve represent
the post control scenario.
v Exceedance data were generated for two areas of impact around
SJl
each of the model plants. The first area was only a few
kilometers from the source and was identified as the primary area.
This area contained the highest S02 concentrations associated with.
the model plant emissions. The second area was further downwind
from the source and contained relatively lower S02 concentrations.
This area was identified as the secondary area. Table 4 reports
the number of exceedances in the baseline scenario for both the
primary and secondary areas for each of the five case study areas.
As expected, most of the exceedances occur within the primary area
and fall within the 0.6 to 1.0 ppm range. Across the five case
studies, a 5-minute S02 average of 3.0 ppm was exceeded only 20
times during the year.
In order to calculate the changes in health risk associated
with this regulatory program, it is first necessary to determine
how many of the exceedances are likely to have an impact on the
risk of an asthma attack. The values reported in Table 4 reflect
the number of times within a given year a specific S02 value is
exceeded for at least one 5-minute period during an hour.
Multiple 5-minute exceedances occurring within a 1 hour pexiod
6-13
-------
are not reflected in the table. Since the refractory period
associated with exposure to S02 is at least an hour, the
incremental health effects from multiple exceedances that occur
during a 1-hour period are likely to be minimal. Exceedances
that occur over sequential hours need to be taken into account,
however, since these exceedances may also occur during the
refractory period and therefore may not cause any additional
health effects beyond those associated with the first exceedance.
In addition, exceedances that occur outside of waking hours
should not be considered to have an impact on the risk of an
asthma attack. Unfortunately, information on the sequence and
timing of exceedances is not available from the underlying
monitoring data used to estimate the data reported in Table 6-4.
However, an estimate of the average number of exceedance hours
within an exceedance day of 1.67 hours is available from a
Summary of 1988-1995 Ambient 5-Minute S02 Concentration Data18.
This estimate is used to adjust the exceedances reported in Table
4 to account for multiple exceedances during an exceedance day.
Although no specific adjustment can be made to account for
exceedances that occur outside of waking hours, anecdotal
evidence from one area suggests that the majority of the
exceedances under consideration occur during waking hours.
6-14
-------
Table 6-4
PREDICTED ANNUAL EXCEEDANCES OF ALTERNATIVE
5-MINUTE S02 CONCENTRATIONS*
5-Minute
S02
(ppm)
0.6
0.7
i.
0.8
0.9
1.0
1.1
1.2
1.3
1.5
1.75
2.0
2.5
3.0
3.5
TOTAL
Case Study 1
Primary
18
13
10
7
6
4
3
5
3
2
2
1
0
0
74
Secondary
4
2
1
0
0
0
0
0
0
0
0
0
0
0
7
Case Study 2
Primary
86
75
65
58
51
46
41
37
30
24
19
12
8
5
557
Secondary
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Case Study 3
Primary
20
16
13
11
9
7
6
10
9
6
7
4
2
3
123
Secondary
8
5
3
0
0
0
0
0
0
0
0
0
0
0
16
Case Study 4
Primary
30
23
18
14
12
9
8
11
9
6
7
3
1
1
152
Secondary
18
12
6
3
1
0
0
0
0
0
0
0
0
0
40
Case Study 5
Primary
30
22
16
12
9
7
5
4
3
1
1
0
0
0
110
Secondary
2
1
0
0
0
0
0
0
0
0
0
0
0
0
3
Exceedances reported are incremental exceedances. As an example, 0.7 ppm was exceeded 57 times during the
year for Case Study 1, but the exceedance fell between 0.7 ppm and 0.8 ppm only 13 times.
6.1.3 Step 3; Determine the Population Affected By the Change in
Air Quality
One of the necessary components for the benefit analysis is
information about the population exposed to the pre- and post-
control S02 concentrations estimated in Step 2. In this step,
information on the size and shape of the S02 plumes emittad_from
6-15
-------
the five model plants is combined with population data that
characterize the geographic areas with S02 emission sources to
develop estimates of the exposed population. This step discusses
the techniques and assumptions used to develop these estimates.
First, estimates of population densities that characterize areas
around S02 emission sources are developed. Then, the area of
land affected by modeled S02 plumes is determined. Finally, the
population densities and the land areas are combined to form
estimates of potentially exposed populations.
Demographic Characteristics--
Table 6-5 presents qualitative population characteristics
for the five areas near the S02 emission source as well as at a
distance away from the source. The meaning of the population
characteristic descriptors is discussed further below.
Table 6-5
POPULATION CHARACTERISTICS
Study Area
Areal
Area2
Area3
Area4
Area 5
Primary Area
Town
Town
Town
Town
Small Urban
Secondary Area
Rural
No impact
Town
Small Urban
Small Urban
6-16
-------
Population Density
As shown in Table 6-5, and described in the associated text,
the populations in the case study areas were characterized as
"small urban," "town," "rural" and "no impact" in decreasing order
of population density. These terms are not meant to reflect an
official definition, but are designed to be generally descriptive
of the study area. The following discussion describes the
determination of characteristic population densities.
When possible, 1986 population figures from the 1988 City-
County Data Book19 were used to determine population and area of
the case study area. In addition, similar data were collected
for the surrounding county and any nearby metropolitan areas. To
obtain population density, the ratio of population to land area
was computed. In those instances where no data was available in
the City-County Data Book, The Rand McNally Road Atlas was used
to obtain population information.
To determine population and area of the surrounding "rural"
county area, the population and area of any metropolitan area was
subtracted from county population and area within the study area
to form a "non-urban" population and land area value. Computing
the ratio of the "urban-excluded" population and area provides a
"rural-only" population density value.
6-17
-------
Table 6-6 presents the results of this data analysis. The
first column indicates the case study area. The second and third
columns show population (rounded to hundreds, as appropriate) and
area in square miles. The fourth column presents the population
density in persons per square mile. Finally, population density
in persons per square kilometer are presented in the fifth
column. The equivalence of 1 mile to 1.609 km was used to make
the conversion of 1 square mile to 2.589 square kilometers. The
fourth column is divided by 2.589 to form the fifth column which
presents population per square kilometer.
Selection of Representative Characteristics--
Based on the population densities displayed in Table 6-6,
"typical" numbers were chosen that appeared to reflect the
population density characteristics for the demographic
characteristics shown in Table 6-5. Table 6-7 .shows the
correspondence between the labels used and the densities chosen.
As can be seen from Table 6-7, the term "Small Urban"
reflects the population density characteristic of the downtown
areas of small metropolitan areas and is set at 1000/km2 (one
thousand people per square kilometer). "Town" reflects the
population density characteristic of the central areas of small
rural towns and is set at 500/km2. "Rural" reflects the popula-
tion density characteristic of the more rural areas outside
central areas of small towns and is set at 20/km2, which is_the
6-18
-------
average of the rural areas in Table 6-6. Finally, "No Impact"
reflects the population density characteristic of the open
countryside and is set at 0/km2.
6-19
-------
Parabolic Exhaust Plume--
The next step is to determine the land area impacted by S02
emissions. Once the area of. impact is determined, the population
density characteristics can be used to estimate the population
exposed to high short-term S02 concentrations.
Table 6-6
POPULATION DENSITY TABLE
Location
Area 1 - County
Area 1 - Town
Area 1 - Rural
Area 2 - County
Area 2 - Small Urban
Area 2 - Town
Area 2 - Rural
Area 3 - County
Area 3 - Town
Area 4 - County
Area 4 - Small Urban
Area 4 - Rural
Area 5 - County
Area 5 - Small Urban
Area 5 - Rural
Population
102,400
4,000*
98,400
112,500
32,200
3,300*
80,300
164,500
2,500*
104,700
59,300
45,400
120,100
80,300
39,800
Area (sqmi)
5300
3.0"
5300
1,600
14
2.0**
1,580
700
3.00*
300
17
300
2,600
30
2600
Density (Pop/sqmi)
19.3
1333.3
18.6
70.3
2,300.0
1,650.0
50.8
234.3
833.3
349.0
3488.2
151.3
46.2
2676.7
15.3
Density (Pop/sqkm)
7.5
515.0
7.2
27.2
888.4
637.3
19.6
90.7
322.0
134.8
1347.4
58.5
17.8
1033.9
5.9
*/
**/
Source: 1988 City-County Data Book, 1986 population figures, unless otherwise indicated (See text).
1994 population obtained from Rand McNally Road Atlas.
Area estimated by review of maps.
6-20
-------
Table 6-7
POPULATION DENSITY CHARACTERISTICS USED IN IMPACT ANALYSIS
Demographic
Characteristic
Small Urban
Town
Rural
No Impact
Population Density
(People/km)
1000
500
20
0
Based on concentration dispersion data for each of the case
study areas, estimates can be made of the size of the S02 plume
as a function of downwind distance. It is assumed that the
intersection of the ground with a plume is parabolic in shape
with the emission source located at the turning point in the
parabola. Plume widths for each case study area were estimated
at a distance from each emission source where no 5-minute S02
concentrations in excess of 0.6 ppm were predicted to occur.
Column two of Table 6-8 reports these distances to the source and
the width of the plume at this distance. With these data, it is
possible to compute the area within the parabolic plume at any
arbitrary distance from the emissions source.
Table 6-8
S02 PLUME CHARACTERISTICS
Study Area
Area 1
Area 2
Area 3
Area 4
Area 5
Distance to Source
14km
14km
14km
21km
14km
Width of Plume
2km
2km
2km
3km
2km
6-21
-------
Primary and Secondary Exposure --
The modeled S02 plume touches down and will expose any
person at ground level to possibly high values of S02 when that
individual is near the emission source. This short-range area of
impact is called the primary area. This distance is generally
only a few kilometers from the emission source. Farther downwind
the level of S02 is not as high but more people may be exposed
since a larger area is affected. This area is called the second-
ary area. The farthest distance for which exceedances are
observed is generally about 15 kilometers. Beyond this distance,
sufficient dissipation and dispersion occurs that little health
impact from high S02 concentrations is believed to occur from the
given emission source.
Using the case study models, the distances to the primary
and secondary exposure boundaries were identified. These
boundary values are reported in Table 6-9. Table 6-9 also
reports the area (km2) within each primary and secondary exposure
region based on the exposure boundaries and the parabola shape
characteristics reported in Table 6-8. In Area 2, it was
predicted that no short term exceedances would be observed in the
secondary area due to a relatively short stack height at the
model plant and the modeled dispersion characteristics.
Exposed Population--
6-22
-------
The data on exposed areas within the S02 plume shadow
presented in Table 6-9 are combined with the area characteristics
presented in Table 6-5 and the associated population densities
for these areas presented in Table 6-7 to form estimates of the
exposed population.
6-23
-------
Table 6-9
S02 PLUME AREAS
Study Area
Area 1
Area 2
Area 3
Area 4
Area 5
Primary Region
Distance (km)
3.0
3.5
3.0
3.5
3.5
Primary Region
Area (km2)
1.8516
2.3333
1.8516
2.8577
2.3333
Secondary Region
Distance (km)
14.0
N/A
14.0
21.0
14.0
Secondary Region
Area (km2)
16.8150
N/A
16.8150
27.1423
16.3334
Table 6-10 presents the estimates of the exposed population in
each of the case study areas. The population at risk from
exposure to short-term elevations in S02 are a subset of these
population estimates -- namely, exercising asthmatics and to a
lesser extent non-exercising asthmatics. The population
estimates are multiplied by the most recent national estimates of
the prevalence rate for asthma obtained from National Center for
Health Statistics20 to estimate the number of asthmatics at risk.
Since the prevalence rate is higher for children (7 percent) than
for adults (5 percent), the population estimates contained in
Table 6-10 are first broken down by age using information from
EPA's Environmental Justice Data Base on the percent of the
population under age 18. The resulting population estimates are
then multiplied by the relevant prevalence rates to obtain
estimates of the asthmatics exposed in the case study areas.
Estimates of the number of asthmatics exercising during any
waking hour range from 0.2 percent to 3.3 percent. A value"of
6-24
-------
1.7 percent is used in this analysis which is consistent with the
value used in Abt Associates21. Estimates of the number of
exercising and non-exercising asthmatics for the case study areas
are provided in Table 6-11.
6-25
-------
Table 6-10
EXPOSED POPULATION
Study Area
Area 1
Area 2
Area 3
Area 4
Area 5
Exposed
Population
1,262
1,167
9,333
28,571
18,667
Table 6-11
ASTHMATIC POPULATION AT RISK
Case Study
1
2
3
4
5
Primary Area
Exercising
<1
1
1
1
2
Non-Exercising
52
64
50
76
127
Secondary Area
Exercising
<1
no impact
8
25
15
Non-Exercising
19
no impact
465
1,445
887
Limi ta. t± ons - -
The exposed population values presented here reflect various
assumptions about the size and shape of S02 emission plumes, and
the density and distribution of populations exposed to the SO2
concentrations found within the emission plumes. It was assumed
that the emission plume was oriented in a specific direction over
population areas which reflected a "worst-case" scenario. Actual
wind always determines which direction a plume dissipates
6-26
-------
emissions from any source. Knowledge of prevailing winds could
better assist in ascertaining the likelihood that the case study
plumes actually shadow population areas. However, no account was
taken of local meteorology except for the stability class charac-
teristics which influence the size and shape of the emission
glumes.
Estimates of the asthmatic population are obtained using
data on national asthma prevalence rates. The EPA has recieved
comments on the 1994 proposal that indicates the prevalence of
asthma in areas that are known to have S02 problems can be
higher than the national average. If the population exposed in
these study areas have different prevalence rates due to SO2 or
other factors, then the estimates presented in Table 6-11 may be
biased. The use of one estimate to characterize the percentage
of waking hours devoted to exercise for all asthmatics may result
in biased estimates if activity patterns vary significantly among
asthmatics. For example, exercise may be encouraged for mild and
moderate asthmatics as a way of controlling their asthma. Severe
asthmatics, on the other hand, may be discouraged from
exercising. In addition, the exercise patterns of asthmatic
children and young adults are likely to be much different than
tlie exercise patterns of asthmatic adults.
6-27
-------
6.1.4
The previous three steps are required to estimate the
changes in health risk associated with the short-term S02 changes
under consideration. This step involves the economic valuation
of these health risk changes. The improvement in an asthmatic's
health resulting from a reduction in short-term exposure to S02
may manifest itself in a variety of ways. Certain improvements
in lung function may not be perceived by the individual and
therefore are very difficult to value. Others, such as changes
in symptoms like chest tightness and wheezing are likely to be
perceived by the individual. A decrease in symptoms is likely to
result in reductions in discomfort, the need to undertake
averting behavior, the loss of leisure, work or school time, and
medical expenditures. Members of the individual's family may
also experience a reduction in the emotional and financial costs
associated with coping with the individual's symptoms.
It is very difficult to determine the true economic value of
a reduction in symptoms. There are four valuation techniques
that have been used to estimate the economic value of air-
pollution induced changes in health: contingent valuation, cost
of illness, averting behavior, and hedonic valuation. Although
none of these techniques completely measures economic benefits,
they have been used as approximations. The advantages and
shortcomings of these approaches have been reviewed in IEC22,
U.S. Department of Commerce23, and elsewhere. -^—
6-28
-------
The results of two of the contingent valuation studies
reviewed by IEC are directly applicable to this analysis. Rowe
and Chestnut24'25 used a 1983 survey of 82 asthmatic individuals
living in Glendora, California to collect data on asthma
severity, medication use, and activities undertaken to mitigate
asthma. These data were used to obtain estimates of the
•«?
willingness to pay (WTP) to reduce the frequency of an asthma-
related illness. The results of the study indicate a WTP of
$43.53 to avoid one bad asthma day with mild symptoms and $59.48
to avoid one bad asthma day with moderate symptoms (1993
dollars).
Before using the results of the Rowe and Chestnut studies to
value the health risk changes estimated in the previous steps it
is necessary to make a few adjustments. First, the
concentration-response functions identified in Step 1 predict
two types of health risks: percentage changes in specific
airway resistance (SRaw) and percentage changes in symptoms. It
is difficult to determine the value of a change in SRaw because a
change in this health indicator may not be perceived by the
individual. For purposes of this analysis, it is assumed that
the percentage changes in SRaw estimated from the concentration-
response function are accompanied by asthma symptoms that the
individual can perceive.
6-29
-------
Second, symptom severity is not identified in the
concentration-response functions developed in Step 1. The EPA26
provides information on symptom severity based on percentage
changes in SRaw and percentage changes in forced expiratory
volume (FEV). Table 12 reports the gradation of response
severity for alternative SRaw and FEV percentage changes. Abt27
developed a regression equation from data contained in Linn et
al.28 and Roger et al.29 that relates the percentage change in
SRaw to alternative S02 levels:
%ASRaw = 201.03 AS02 (6)
where %ASRaw = the percentage change in specific airway
resistance
AS02 = the change in the 5-minute concentration of S02 in
ppm.
Equation 6 can be used to estimate the changes in S02 required to
produce a change in SRaw between 100 percent and 200 percent
(defined in Table 12 as a moderate effect) and greater than 200
percent (defined in Table 12 as a severe effect) for alternative
S02 changes. The equation can also be used to estimate the S02
change associated with a change in SRaw of less than 100 percent
(defined in Table 6-12 as a mild effect). Based on the above
information, the baseline S02 levels associated with symptom
severity can be obtained. These results are reported in Table 6-
13.
6-30
-------
Table 6-12
COMPARATIVE INDICES OF SEVERITY OF RESPIRATORY EFFECTS SYMPTOMS,
SPIRONETRY, AND RESISTANCE
Type of Response
AinSRAW
AinFEVl.OFVC
Duration of Effect/
Treatment
Symptoms
Mild
Increase < 100%
< 10%
Spontaneous recovery
< 30 minutes
Mild, no wheeze or chest
tightness
Moderate
Increases up to 200%
Decrease of 10 to 20%
Spontaneous recovery
< 1 hour
Some wheeze or chest
tightness
Severe
Increases more than 200%
Decrease > 20%
Bronchodilator required to
resolve symptoms
Obvious wheeze, marked
chest tightness, breathing
distress
Source: EPA (1994a).
Third, the WTP estimates from Rowe and Chestnut are limited
to mild and moderate symptoms. An estimate of $78.10 for the WTP
to avoid a severe asthma attack is obtained by assuming that a
severe attack differs from a moderate attack in that the
individual will cease activity for two as opposed to 1 hour and
there is some additional discomfort associated with a severe
attack.
Table 6-13
ASTHMA SYMPTOM SEVERITY RELATED TO
5 MINUTE S02 EXPOSURE
SO, (ppm)
1.0 and below
1.1 to 1.5
Above 1.5
Symptom Severity
Mild
Moderate
Severe
6-31
-------
Fourth, the WTP estimates from Rowe and Chestnut are for a
symptom day. It is unclear, however, whether a symptom day is
limited to just one asthma attack. Because information on the
timing of the 5-minute exceedances is not available for the five
case study areas, it is possible that multiple exceedances will
occur during a day. Although the impact of multiple exceedances
occurring during the refractory period has been addressed in Step
2, exceedances that occur outside the refractory period cannot
be addressed. If these types of multiple exceedances are
numerous, then multiple asthma attacks may occur during a one day
period and the use of the Rowe and Chestnut WTP estimates may
result in an overestimate of benefits.
Fifth, the WTP estimates are based on a survey undertaken in
1983. The WTP estimates have been adjusted to 1993 dollars using
the Consumer Price Index (CPI). The 1993 WTP estimates only
account for price changes that have occurred during this time
period and do not reflect any increases in WTP for symptom
reductions that may have occurred over the same time period due
to increases in real income or changes in preferences.
Consequently, the 1993 values may be underestimates of the true
WTP.
And sixth, the WTP estimates are taken from contingent value
studies that may not capture all the economic benefits associated
with reducing symptoms. For example, benefits to relativ&s-and
6-32
-------
employers are excluded. Also, medical costs not borne directly
by the individual with the symptoms (e.g. insurance) will not be
included. The contingent valuation itself may result in biased
estimates if participants in the survey give strategic responses.
The valuation of symptom reductions based on the contingent
^aluation approach will not reflect any improvements in
productivity or reductions in medical expenditures that are not
perceived by the individual.
One comment recieved by EPA indicates that "in 1990 the cost
of illness for asthma was $6.2 billion; the cost of school
absenteeism due to asthma was nearly $1 billion, and 43 percent
of the costs was associated with emergency room use,
hospitalization, and death30. Although only a portion of the
$6.2 billion estimate of cost of illness for asthma can be
attributed to short-term S02 peaks, the statement indicates that
the WTP measure obtained from the 1983 study may not have
captured all areas of cost associated with asthma.
Another factor to consider is that it is uncertain whether
the individuals surveyed for the WTP estimate incorporated a
value for side effects of asthma medication. While the Staff
^aper recognizes that the use of medication can mitigate some of
the effects of short-term S02 peaks, the EPA recieved comments
that state that the staff paper "failed to summarize the side
effects of these medications."31 The commenter indicates feioat
6-33
-------
brochodilators are used for acute need, however, the medication's
effect lasts for hours. Common side effects of some medications
include heart palpitations, tachycardia, nausea, muscle tremors,
diziness, weakness, restlessness, apprehension, and anxiety.
Individuals taking medication (such as steroids) for severe
asthma can experience side effects such as stunted growth and
osteoporosis32. Additional medication to combat some of the side
effects may be necessary, which adds to medical costs and can
create other additional side effects.
6.1.5 Step 5; Estimate Benefits
The fifth and final step of the benefits analysis combines
the information obtained in the previous steps to calculate
benefits. From the concentration-response functions, the change
in asthma symptom prevalence resulting from a change in short-
term exposure to S02 can be calculated for the relevant
population cohorts. These estimates are then multiplied by the
population estimates developed in Step 3 and the WTP estimates
reported in Step 4 to yield monetary benefit estimates. A sample
calculation for the inner area of Case Study 1 for the non-
exercising asthmatic cohort is provided below for illustrative
purposes.
Based on the S02 related symptom severity defined in Table
13 and the Sheppard et al. equation (Equation 3), the change in
6-34
-------
the prevalence of severe and moderate symptoms associated with a
change in S02 can be calculated from:
Aprevalence(severe) = 2^ 0.1071 (S02 (sevj - 1.0 ) (6)
i =1
m
APrevalence (mod) = £ 0.1071 (S02 (modi)-1.0) (7)
i =1
where:
APrevalence(severe) = change in the % of the non-exercising
asthmatic population experiencing a
severe symptom during a one year period,
APrevalence(mod) = change in the % of the non-exercising
asthmatic population experiencing a
moderate symptom during a 1-year period,
S02(seVi) = average S02 that exceeds 1.5 ppm during
the ith 5-minute period,
S02 (modi) = average S02 that exceeds 1.0 ppm and is
less than or equal to 1.5 ppm during the
ith 5-minute period,
s = the number of exceedances greater than
1.5 ppm during 1 year, and
m = the number of exceedances greater than
1.0 and less than or equal to 1.5 ppm
during 1 year.
Only changes in SO2 exposure above the 1.0 ppm level are
calculated for Equations 6 and 7, since non-exercising asthmatics
are not considered to be sensitive beneath this level. For Case
Study 1, APrevalence(sev) is 0.5355 and APrevalence(mod) is
O-.4284,
6-35
-------
Once the changes in prevalence rates are obtained , they are
divided by the factor of 1.67 to account for the possibility of
multiple exceedances occurring during the refractory period.
Finally, benefits for a 1-year period are calculated by
multiplying the change in prevalence rates by the non-exercising
asthmatic population (Nonexpop) and the appropriate estimate of
WTP:
n C-L. APrevalence (sev) .T _. , n
Benefits = * Nonexpop * 78.10 +
^ 1.67 )
I APrevalence (mod) XT rn „_)
* Nonexpop * 59.48
I 1.67 I
The benefits accruing to the non-exercising population in
the primary area of Case Study 1 are equal to approximately
$2,100.
Benefits for the other asthmatic cohorts and the other case
study areas can be calculated in a similar manner.
6.2 Quantification of Estimates
Benefits for the case study areas are reported in Tables 14
through 18 in 1993 dollars. Of the areas under consideration,
Case Study 2 has the largest benefits. This result is due
primarily to the fact that the 5-minute S02 average of 0.6 ppm
was exceeded 471 times during the 1-year period examined in the
6-36
-------
analysis. Case Study 4 has the second highest benefits. These
estimates appear to be driven by the large number of exceedances
and large population in the secondary area along with the
relatively large number of exceedances in the primary area. Case
Study 1 has the smallest benefits due to a combination of small
population and relatively few exceedances.
On the whole, the benefits reported in Tables 6-14 through
6-18 are relatively small. The predominant reason for this
result is that the short-term peaks in S02 under examination
impact a fairly small geographic area within the local vicinity
of the model plants. The small geographic area coupled with the
fraction of the local population assumed to be "exercising
asthmatics" significantly limits the number of people considered
to be at risk. Although non-exercising asthmatics are relatively
less sensitive to mild peaks in short-term exposure to S02 than
exercising asthmatics, the benefits accruing to this population
cohort drive the benefit estimates because there are so many of
them compared to the exercising asthmatics. Because of the
relatively large number of exceedances in both the primary and
secondary areas, a sensitivity analysis on the non-exercising
asthmatic population was done for Case Study 4 to see how
altering the assumptions regarding the underlying concentration-
response would impact the benefit estimates. In the sensitivity
analysis, the Horstman et al. concentration-response
6-37
-------
Table 6-14
BENEFITS FOR CASE STUDY 1
Primary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Secondary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Total
Severe
Incidents
$ 157.47
1,303.00
0
0
$1,460.47
Moderate
Incidents
$ 283.30
793.88
0
0
$1,077.18
Mild
Incidents
$ 197.20
0
3.24
0
$200.44
Total
$ 637.97
2,096.88
3.24
0
$2,738.09
Table 6-15
BENEFITS FOR CASE STUDY 2
Primary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Secondary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Total
Severe
Incidents
$ 2,639.71
26,626.02
0
0
$29,265.73
Moderate
Incidents
$ 3,535.33
9,446.90
0
0
$12,982.23
Mild
Incidents
$1,844.85
0
0
0
$1,844.85
Total
$ 8,019.89
36,072.92
0
0
$44,092.81
Table 6-16
BENEFITS FOR CASE STUDY 3
Primary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Secondary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Total
Severe
Incidents
$ 674.29
7,260.07
0
0
$7,934.36
Moderate
Incidents
$ 588.01
1,792.22
0
0
$2,380.23
Mild
Incidents
$272.57
0
657.73
0
$930.30
Total
$ 1,534.87
9,052.29
657.73
0
$11,244.89
6-38
-------
Table 6-17
BENEFITS FOR CASE STUDY 4
Primary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Secondary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Total
Severe
Incidents
$ 837.58
7,844.39
0
0
$8,681.97
Moderate
Incidents
$1,031.76
3001.67
0
0
$4,033.43
Mild
Incidents
$ 558.63
0
2,495.31
0
$3,053.94
Total
$ 2,427.97
10,846.06
2,495.31
0
$15,769.34
Table 6-18
BENEFITS FOR CASE STUDY 5
Primary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Secondary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Total
Severe
Incidents
$ 152.68
1,113.63
0
0
$1,266.31
Moderate
Incidents
$ 861.77
2,132.43
0
0
$2,994.20
Mild
Incidents
$ 779.04
0
36.1
0
$815.14
Total
$1,793.49
3,246.06
36.1
0
$5,075.65
Table 6-19
SENSITIVITY ANALYSIS OF BENEFITS FOR CASE STUDY 4
Primary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Secondary Area
Exercising Asthmatics
Non-Exercising Asthmatics
Total
Severe
Incidents
$ 837.58
24,225.52
0
0
$25,063.10
Moderate
Incidents
$ 1,031.76
14,921.00
0
0
15,952.76
Mild
Incidents
$ 558.63,
0
2,495.31
0
$3,053.94
Total
$ 2,427.97
39,146.52
2,495.31
0
$44,06930
6-39
-------
function was used along with the assumption that 25 percent of
the non-exercising asthmatics would respond like that predicted
by Horstman et al. for S02 levels equal to at least 1.0 ppm, and
50 percent would respond at levels above 1.5 ppm. The results of
this analysis are reported in Table 6-19. Benefits increased by
a factor of three, suggesting that the underlying concentration-
response function has a significant impact on benefits.
6.3 Limitations of Analysis
The benefit estimates provided in this analysis need to be
viewed in light of numerous qualifications. Although these
qualifications have been discussed in detail in the preceding
sections, they are briefly summarized here to conclude the
chapter.
Concentration-response functions--
The concentration-response functions used to calculate
benefits in this chapter are developed from data obtained in
controlled laboratory settings for small samples of mild and
moderate asthmatics. The applicability of these functions to the
five case study areas may be tenuous if the populations and
exposure conditions and relationships observed in the underlying
studies are not representative of the case study areas.
One example of a limitation with respect to these functions is
that the functions do not consider the effect that pre-meelication
may have on the relationship between S02 and exposure. Although
mild and moderate asthmatics are typically not known to
6-40
-------
premedicate, the omission of this possibility may bias the
benefit estimates.
Air quality data--
Pre-control data for each of the five case study areas were
^estimated using monitoring data from sites located in areas near
actual facilities. Data on meteorology were not available;
consequently the S02 plumes were oriented in a direction which
would impact the most people. Ideally, incorporation of local
meteorological conditions would have provided better estimates of
the impacted population.
Information of the number of exceedances occurring during the
refractory period (i.e., the period of reduced responsiveness to
short-term S02 exposure) could not be obtained from the
monitoring data. The exceedance data were adjusted using
national monitoring data from 1988 to 1995. If the national data
are unrepresentative of the conditions that would exist in the
five case study areas, then the resulting benefit estimates will
be biased.
Population data--
Again, national data were used to estimate the data required
for the five case study areas. In this case, national estimates
of the asthma prevalence rate and the percentage of waking hours
spent exercising were used to estimate the exercising and*-non-
exercising asthmatic population in the case study areas. The EPA
has received comments on the 1994 proposal that indicate that the
6-41
-------
prevalence of asthma in area that are known to have SO2 problems
can be much higher than the national average33. If the
asthmatic population in these areas differs significantly from
that estimated from the national data, the affected population
estimates used in this analysis may be inaccurate.
Willingness to pay--
The willingness to pay (WTP) estimates were taken from the
results of a contingent valuation survey undertaken in 1983. If
preferences have changed since the survey was undertaken, the
1983 WTP may be inaccurate. Also, the WTP estimates obtained
from the contingent valuation approach probably underestimate the
true economic benefit associated with air quality improvements
because they exclude the value of unperceived improvements in
health status, the benefits accruing to relatives and employers,
and costs not typically borne by the individual.
Other Benefit Categories -
The benefits reported in this analysis are limited to the
health benefits accruing to the asthmatic population from changes
in their short-term exposure to S02. The welfare benefits
associated with any visibility improvements and reductions in
materials and agricultural damage that may accompany the
implementation of a program designed to limit short-term S02
exceedances have not been evaluated in this analysis. Because
the control strategies chosen to resolve a short-term S02-problem
can subsequently achieve longer term S02 emission reductions
year-round, there are secondary benefits that can be achieved in
6-42
-------
these other benefit categories. In this sense the benefits
reported in this chapter are underestimates. The nature of these
unquantified benefit categories is described below.
Ecosystem Impacts
.. In addition to causing human health effects, sulfur dioxide
*4
can also impact vegetation and ecosystems. Low doses of S02 can
increase growth and yield in plants growing in sulfur-deficient
soils. However, if the rate of absorption of S02 is greater than
the plant's ability to metabolize S02, toxic metabolites can
reach sufficient concentrations within the plant to cause foliar
injury, reduction in growth and yield, and with acute exposures,
plant death. A number of studies have developed dose-response
functions for crop species such as soybeans, oats and wheat. The
revised EPA S02 staff paper34 indicates that in nonarid regions
where there is high temperature, high humidity, and abundant
sunlight -- conditions that increase plant responsiveness to S02
-- visible injury may develop in sensitive species to 5-minute
exposures of 1 to 2 ppm S02.
Responses to sulfur dioxide at the individual plant level can
have broader impacts at the community and ecosystem level.
Ijijury to vegetation can affect species composition and nutrient
cycling within terrestrial ecosystems. Studies on grassland
ecosystems have shown impacts of low ambient concentrations
(greater than 0.02 ppm) of S02 to different trophic levels—within
the ecosystem.
6-43
-------
Sulfur dioxide emissions have been implicated as a cause of
acid precipitation and the acidification of aquatic ecosystems.
Decreases in pH levels in streams, ponds, and lakes can affect
all trophic levels of the aquatic ecosystem, resulting in a loss
in species diversity of phytoplankton, zooplankton and various
fish species. The National Acid Precipitation Assessment Program
reports the loss of lake trout, rainbow trout and walleye and
smallmouth bass at pH levels below 5.5.
6-44
-------
Odors
The 1982 EPA S02 staff paper (4) reports that studies have
found the odor threshold for S02 to range from 0.47 ppm to 1 ppm.
Regarding this subject, one member of the Clean Air Scientific
Advisory Committee (CASAC) of EPA's Science Advisory Board wrote
.that he has "the strong impression that for a substantial portion
of the general public, likely a majority, experiencing
perceptible S02 (over, perhaps, .4 ppm) in ambient air degrades
the quality of life by making people perceive that they live in
polluted air" (5). To the extent that this regulatory program
reduces short-term S02 peaks below the odor threshold, positive
benefits would be achieved in terms of reduced occurrences of
noxious odor.
Ma. t eria. 1 s Damage
Much research has been conducted on the effects of sulfur
oxides on materials. Sulfur dioxide, specifically, can
accelerate the corrosion of metals such as iron, galvanized
steel, copper and aluminum-based metals. Additionally, S02 can
erode and soil stone and paints. Dose-response functions have
been developed that relate ambient S02 concentrations to physical
damage to a number of materials. For this analysis, however, it
i^ not possible to determine the level of reductions in short-
term S02 peaks in respect to materials damage, so this category
is not investigated.
Particulate Matter Benefits
6-45
-------
A portion of S02 emissions will be transformed in the
atmosphere to particulate sulfate. Epidemiology studies have
shown statistically significant associations between ambient
particulate matter concentrations and incidence of respiratory
symptoms, emergency room visits and hospital admissions for
respiratory conditions, exacerbation of chronic respiratory
disease and mortality. Additionally, ambient particulate matter
contributes to visibility impairment and soiling of materials.
To the extent that S02 emission reductions are achieved through
the IL program, it is expected that particulate matter benefits
would also be attributable to this action. Although the program
may also have an impact on the annual emissions of SO2 and
therefore on the production of sulfates, these impacts have not
been addressed. Consequently, any health and welfare benefits
that may result from reductions in sulfate levels are omitted
from this analysis.
6.4 Environmental Justice Considerations
Executive Order 12898, "Federal Actions to Address
Environmental Justice in Minority Populations and Low-Income
Populations", directs each Federal agency to "make achieving
environmental justice part of its mission by identifying and
addressing ... disproportionately high and adverse human health
or environmental effects of its programs, policies, and
activities on minority populations and low-income populations."
In order to comply with the provisions of this Order, a general
screening analysis has been conducted to examine the
6-46
-------
sociodemographic characteristics of the case study areas in which
controls could be necessary.
A number of factors indicate that asthma may pose more of a
^health problem among non-white and urban populations. As the SO2
Criteria document35 indicates, there is a higher prevalence of
asthma among African-American and urban populations.
Additionally, mortality rates due to asthma are at least 100
percent higher among non-whites than the national average,
although death due to asthma is a rare event (1 per 10,000
asthmatic individuals). In New York and Chicago for example,
non-white mortality rates from asthma may exceed the city average
by up to five-fold and exceed the national average by an even
.larger factor.
With respect to the effects of short-term S02 exposures on
asthmatics, the S02 criteria document36 indicates that controlled
human exposure studies have not systematically studied African-
American and Hispanic adolescents and young adults. Therefore,
it is not known the extent to which the controlled exposure
studies, as discussed in the criteria document and staff paper,
reflect accurately the responses to 5-minute S02 exposures among
minority populations. Additionally, one CASAC member stated in
his comments on the S02 staff paper that staff paper results may
be important in that moderate asthmatics from urban areas and
lower socioeconomic status may be at particularly high ri-sJc- if
exposed to S02 above 0.60 ppm while at a high level of activity.
Asthmatic individuals from these subpopulations in general may
6-47
-------
have inadequate medical follow-up, may have irregular medication
use and frequent lung function deterioration37.
Considering the above factors, a general screening analysis is
conducted to examine the sociodemographic characteristics of the
case study areas potentially impacted by short-term S02 peaks.
For each area, data from the EPA's Environmental Justice Database
is used to obtain population estimates that are disaggregated by
non-white population and non-white asthmatic population (using
the national average asthma rate of 5 percent). As has been
stated, research indicates that asthma prevalence among non-white
individuals is potentially higher than the national average, but
also children tend to have a higher prevalence rate than that of
adults. Although, there is insufficient information at this time
to define asthma prevalence among non-whites for specific
geographic areas, the National Center for Health Statistics has
provided a prevalence rate for children (of ages less than 18
years) at 7 percent. In addition to population data, information
on households below the poverty level3 (including minority
households below the poverty level) is also provided.
For the localized areas in which the S02 plumes of the case
studies are assumed to disseminate, the total population of each
area ranged from 7600 to nearly 100,000 people, with an average
population of 34,000. On average, 5 percent of the population in
the case study areas are non-white, and more than 25 percent are
a For the screening analysis, the poverty level has been
defined as any household income below $15,000 per year.
6-48
-------
children. Nationally, 16.5 percent of the population is non-
white and 11 percent is less than 18 years of age38. While the
percent on non-white individuals in the case study areas is below
the national average, the percentage of children residing in
these areas is more than double the national average.
Using the national average of asthma prevalence, a typical
area would have 1700 asthmatic individuals potentially impacted
by short-term burst of S02b, or 85 non-white asthmatic
individuals0, and 595 athmatic children6.
Additionally, the areas also have an average of 14,850
households. Twenty-seven percent of these households would be
classified below the poverty level and five percent of these
households below the poverty level are occupied by non-white
individuals. Nationally, only 12.7 percent of the total
households in the U.S. are below the poverty level, indicating
that the case study areas have twice as many households below the
poverty level.
b Calculation: 34,000 average population x 5 percent
asthma prevalence rate = 1700 asthmatics on average in the case
study areas.
c Calculation: 34,000 avg. population x 5 percent non-
white population on average x 5 percent ashtma prevalence rate =
85 non-white asthmatic individuals on average in the case study
areas. ^—
d Calculation: 34,000 avg. population x 25 percent
children in population on average x 7 percent asthma prevalence
rate = 595 asthmatic children on average in the cae study areas.
6-49
-------
Overall, the populations in the case study areas do not show
any indications that a disproportionate number of non-white
individuals would be impacted by short-term S02 ambient
concentrations greater than 0.60 ppm. This analysis, however,
does not cover all possible areas of the country with short-term
S02 peak concentrations greater than 0.60 ppm. Other areas of
the country may have a higher percentage of non-white citizens.
The analysis indicates that there are twice as many children
residing in the case study areas as compared to the national
average, and potentially 595 could have asthma and thus
experience health impacts during peak S02 concentrations. In
addition to the large number of children potentially exposed to
peak S02 concentrations, 27 percent of the households in the case
study areas are below the poverty level, which twice the national
average. It should be noted, however, that it is not known how
many of the households below the poverty level contain asthmatic
individuals. Given the available data, this analysis gives an
indication that a disproportionate number of children and
households below the poverty level are exposed to short-term S02
peaks. In general, children do not have the resources to
relocate or take action against sources of S02 emissions.
Similarly, households below the poverty level may be dependent on
the local industrial sources for employment. In addition to
having limited resources to relocate or take action against
sources of S02 emissions, they may be reluctant to do so if
action would detriment employment opportunities. -—
6-50
-------
During an evaluation 5-minute ambient concentrations for the
IL program, a regulatory authority would use information specific
to the area of analysis to determine if there were indications
that under the criteria for environmental justice action would be
warranted.
6-51
-------
REFERENCES
1. Air Quality Criteria for Particulate Matter and Sulfur
Oxides. U.S. Environmental Protection Agency; Office of
Health and Environmental Assessment; Research Triangle Park,
NC; Document no. EPA-600/8-82-029aF-CF.3V; 1982.
2. Air Quality Criteria for Particulate Matter and Sulfur
Oxides: V 1, Addendum. U.S. Environmental Protection
Agency; Research Triangle Park, NC. Document no. EPA-600/8-
82-029aF; 1982.
3. Review of the National Ambient Air Quality Standards for
Sulfur Oxides: Assessment of Scientific and Technical
Information, OAQPS Staff Paper. U.S. Environmental
Protection Agency; Office of Air Quality Planning and
Standards; Research Triangle Park, NC. Document no. EPA-
450/5-82-007; 1982.
4. Second Addendum to Air Quality Criteria for Particulate
Matter and Sulfur Oxides (1982): Assessment of Newly
Available Health Effects Information. U.S. Environmental
Protection Agency; Office of Health and Environmental
Assessment; Research Triangle Park, NC. Document no. EPA-
6€0/8-86-020F; 1986.
5. Supplement to the Second Addendum (1986) to Air Quality
Criteria for Particulate Matter and Sulfur Oxides (1982) :
Assessment of New Findings on Sulfur Dioxide Acute Exposure
Health Effects in Asthmatic Individuals. U.S. Environmental
Protection Agency; Office of Health and Environmental
Assessment; Research Triangle Park, NC. Document no. EPA-
600/FP-93/002; 1994.
6. Review of the National Ambient Air Quality Standards for
Sulfur Oxides: Assessment of Scientific and Technical
Information — Supplement to the 1986 OAQPS Staff Paper
Addendum. U.S. Environmental Protection Agency; Office of
Air Quality Planning and Standards; Research Triangle Park,
NC. Document no. EPA-452/R-94-013; 1994.
7. Airway Sensitivity of Asthmatics to Sulfur Dioxide;
Horstman, D. et al. Toxicology and Industrial Health,
2:289-298; 1986.
8. Lower Threshold and Greater Bronchomotor Responsiveness,
American Review of Respiratory Disease, of Asthmatic——
Subjects to Sulfur Dioxide. Sheppard, D. et al. American
Review of Respiratory Disease, 123:873-878; 1980.
6-52
-------
9. The Benefits and Costs of the Clean Air Act, 1970-1990,
Draft Report to Congress, Appendix D; Prepared by Abt
Associates for the U.S. Environmental Protection Agency,
Office of Air and Radiation; May 1996.
10. Asthmatics' Responses to 6-hr. Sulfur Dioxide Exposures on
Two Successive Days. Linn, W.S. et al. Archives of
Environmental Health, 39:313-319; 1984.
"11. Replicated Dose-Response Study of Sulfur Dioxide Effects in
„_ Normal, Atopic, and Asthmatic Volunteers. Linn, W.S. et al.
American Review of Respiratory Disease, 136:1127-1134;
1987.
12. Effect of Metaproteronol Sulfate on Mild Asthmatics'
Response to Sulfur Dioxide Exposure and Exercise. Linn,
W.S. et al. Archives of Environmental Health, 43:399-406;
1988.
13. Responses to Sulfur Dioxide and Exercise by Medication-
Dependent Asthmatics: Effect of Varying Medication Levels.
Linn, W.S. et al. Archives of Environmental Health, 45:24-
30; 1990.
14. Bronchoconstriction in Asthmatics Exposed to Sulfur Dioxide
During Repeated Exercise. Roger, L.J. et al. Journal of
Applied Physiology, 59:784-791; 1985.
15. Reference 4.
16. Reference 10.
17. Reference 5.
18. Summary of 1988-1995 Ambient 5-Minute S02 Concentration
Data, Draft Final Report prepared by Systems Applications
International (SAI)under subcontract to ICF Kaiser, Inc. for
U.S. EPA's Office of Air Quality Planning and Standards,
Research Triangle Park, September 1995.
19. County and City Data Book, 1988. U.S. Department of
Commerce, Bureau of Census. U.S. Government Printing
Office. Published in 1990.
20. 1994 National Prevalence Rates for Asthmatics. National
Center for Health Statistics. Telephone conversation,
March 1996.
21. Reference 9.
22. Memorandum to Jim De Mocker, U.S. EPA on Review of Existing
Value of Morbidity Avoidance Estimates: Draft Valuation
Document. Industrial Economics Incorporated; September
1993.
6-53
-------
23. Natural Resource Damage Assessments under the Oil Pollution
Act of 1990 — Appendix I — Report of the NOAA Panel on
Contingent Valuation. U.S. Department of Commerce;
58FR4601-4614, January 1993.
24. Oxidants and Asthmatics in Los Angeles: A Benefit Analysis.
Energy and Resource Consultants. Report prepared by Rowe,
R. and L. Chestnut for the U.S. Environmental Protection
Agency, Office of Policy Analysis; Document no. EPA-230-07-
85-010, Washington, DC, March 1985.
25. Addendum to Oxidants and Asthmatics in Los Angeles: A
Benefit Analysis. Prepard by Rowe, R. and L. Chestnut and
Energy and Resource Consultants, Inc. for the U.S.
Environmental Protection Agency, Office of Policy Analysis;
March 1986.
26. Reference 5.
27. Reference 9.
28. References 11, 12, 13.
29. Reference 14.
30. Docket submittal: A-84-25, VIII-D-11.
31. Docket submittal: A-84-25, VIII-D-11.
32. Docket submittal: A-84-25, VIII-D-90.
33. Docket submittal: A-84-25, VIII-D-88.
34. Review of National Ambient Air Quality Standards for Sulfur
Oxides: Assessment of Scientific and Technical Information -
OAQPS Staff Paper. U.S. Environmental Protection Agency,
document no. EPA-450/5-82-007; November 1982.
35. Reference 4.
36. Reference 4.
37. Respiratory Therapy, Mount Sinai Medical Center. Schacter,
E. Neil letter to George Wolff, Chairperson of the Clean Air
Act Scientific Advisory Committee; May 2, 1994,
38. Statistical Abstract of the U.S., 1994.
6-54
-------
SECTION 7. BENEFIT-COST ANALYSIS
7.0 Net Benefit Analysis
This section provides comparisons of the estimated benefits
and costs associated with the IL program. Comparisons of the
benefits and costs are referred to as a benefit-cost analysis (or
net benefit analysis) and are presented here in response to
Executive Order 12866 and the Unfunded Mandates Reform Act of 1995
which require a qualitative and quantitative comparison of bene-
fits and costs of any regulatory action that is considered to be
"significant." While the EPA does not believe the IL program will
have a significant impact on the national economy, the IL program
evolved in part due to comments received on earlier proposed
implementation strategies, which were deemed to be significant.
Also, the characteristics of the IL program - local responsibil-
ity, flexibility, community involvement - represents a novel
regulatory approach. For these reasons, the EPA has judged the IL
program to be significant as defined by E.O. 12866 and thus
prepared the analyses of costs and benefits.
The implementation of the IL program will lead to favorable
health and other welfare effects that represent a clear improve-
ment in the economic well-being of some members of society?- At
7-1
-------
the same time, however, costs may be incurred as additional
resources are committed to reduce emissions to permissible
levels. These costs cause a reduction in the economic well-being
of some members of society. Given that these costs are generally
incurred as air quality is improved, an evaluation of the net
impact of an air quality improvement on society's economic well-
being requires an assessment of both benefits and costs.
Because of tremendous uncertainties in estimating the
circumstances when an action under the IL program will be imple-
mented, a national estimate of the cost and benefits of the
program is not feasible. Instead, the benefit and cost analyses
utilized case studies to evaluate a sample of potential actions.
The results of these analyses are displayed in Table 7-1 in
annualized 1993 dollars. As indicated by the table, costs exceed
benefits by a significant amount. The small magnitude of bene-
fits results from mainly two factors. First, the short-term
peaks in S02 under consideration impact a fairly small geographic
area within the local vicinity of the model plants. The small
geographic area leads to a relatively small number of people
being exposed to these short term peaks. Second, the benefit
estimates are limited to the health benefits accruing to asthmat-
ics. The welfare benefits associated with any ecosystem, visi-
bility, odor, materials damage, or particulate matter improve-
ments that may result from control of short-term peaks in S02
have not been considered. •**—
7-2
-------
Table 7-1
QUANTIFIED BENEFITS AND COSTS
OF SELECTED CASE STUDIES
(Annualized values in 1993 dollars)
Case Study
1
2
3
4
5*
Benefits
$ 2,700
44,100
11,200
15,800
5,100
Costs
$1.87 million
1.15 million
0.34 million
2 .24 million
0.27 million to
0.31 million
* Two cost estimates are provided for Case Study 5. The first one repre-
sents the costs associated with a 10% rollback in emissions while the
second assumes a 20% rollback. See Section for a discussion.
Although the cost that are determined for the case studies
exceed the quantifiable benefits, the IL program provides a
significant amount of flexibility to regulatory authorities,
communities and sources to achieve a reasonable solution to
short-term S02 problems at substantially lower cost than other
potential regulatory vehicles to address the problem. For
example, the previously proposed regulatory option of establish-
ing a new short-term SO2 NAAQS to eliminate exceedances of 0.60
ppm at any one time in a given year was estimated to cost $1.75
billion. Several of the sources assumed to incur costs under a
NAAQS option would have the potential to not have any regulatory
action taken upon them under the IL program and thus incur no
compliance costs. Even if all five of the actions predicted to
occur under the IL program have the highest end of costs -e&ti-
7-3
-------
mated in the case studies of this analysis ($2.2 million), the
total cost of the IL program would be approximately $11 million,
which is $1.739 billion less than the NAAQS option. Therefore,
the IL program is a very cost-effective solution to the public
health risk associated with short-term peaks of S02.
Additionally, Executive Order 12898 requires that each federal
agency shall make achieving environmental justice part of its
mission by identifying and addressing, as appropriate, dispropor-
tionately high and adverse human health or environmental effects
of its programs, policies, and activities on minority and low-
income populations. A screening analysis of the population
residing in the case study areas indicates
Overall, the IL program will be precipitated by community
involvement. The value a community places on resolving a 5-
minute problem could be substantially different than values used
to estimate benefits in this analysis. The willingness-to-pay to
avoid-experiencing any symptoms of an asthma attack, and/or the
population of asthmatics in the community could be higher than
the national average used in the benefit analysis. In addition,
communities may have an intrinsic value to place on environmental
justice, and have an indication of the level of benefits accrued
for visibility improvement, reduction in odor or materials
damage, or reduced level of particulate matter that can be
achieved by controls installed to resolve a short-term
7-4
-------
lem. Conversely, the control strategies chosen for the case
studies may not be the method of resolution chosen in actual IL
program actions. The flexibility provided by the program allows
for new and innovative methods of control that have the potential
of being less costly that some of the alternatives examined in
the case studies. It is anticipated that regulatory authorities,
citizens, and sources will use the IL program as guidance to
determine if the level of public health risk in the local area
warrants action to resolve a 5-minute S02 problem.
7-5
-------
APPENDIX A
CASE STUDIES OF
ALTERNATIVE CONTROL STRATEGIES
FOR THE INTERVENTION LEVEL PROGRAM
-------
APPENDIX A. CASE STUDIES OF ALTERNATIVE
CONTROL STRATEGIES FOR THE IL PROGRAM
As is discussed in Section 4, a cost analysis of all
possible outcomes of the implementation of the IL program is not
feasible. Alternatively, the cost of program is evaluated using
representative case study examples. The case studies are
intended to represent typical, real-world situations. The are
not intended to represent or prejudge any particular facility.
Actual short-term S02 monitoring data was available for a limited
number of sites. This monitoring data was used to develop
statistical profiles of ambient S02 concentrations typical of
areas near certain industries or groups of industries. Process
equipment described in the case studies is intended to be
representative of equipment found in facilities typical of the
particular case study. In order to avoid any appearance of
prejudging the real-world situations these cases were derived
from, changes were made to the process equipment, control
equipment or control strategies, and/or affected population
descriptions. Efforts were made to describe processes typical of
the represented industry to the extent that these modification
could still reasonably conform to the situations presented.
A.I Case Study It One Source Impacting a Local Community at the
Concern Level
There are currently eight operating copper smelters in the
U.S. Several of these smelters are located in rural Western U.S.
settings in towns with relatively small populations. This case
study represents a larger smelter located in a valley setting in
the Western U.S. As a larger smelter, it is assumed to have an
annual production capacity of 250,000 tons per year. Information
on the copper smelting industry was derived from the report for
the Primary Copper Smelters National Emission Standard for
Hazardous Air Pollutants (NESHAP)1.
For the purposes of this study, the model smelter was
assumed to be located in a valley setting. Although the facility
had installed air pollution controls sufficient to meet the
National Ambient Air Quality Standards (NAAQS) , stagnant weather
conditions and thermal inversions produced high short-term
ambient concentrations. For this reason, the use of tall stacks
with the ability to enhance dispersion under these weather
conditions was used to correct short-term problems. Based on the
limited short-term S02 monitoring information available,
monitoring indicates that copper smelters are in the middle range
for both number and severity of exceedances of the proposed
5-minute concern level of 0.6 ppm.
A-2
-------
Description of Model Plant
The model plant, depicted in Figure A-l.l., is a composite
of several typical facilities and does not necessarily reflect
the process of any one actual plant. Case studies from Primary
Copper Smelters NESHAP and the AWMA Air Pollution Engineering
Manual2 were used to develop the sample facility. The model
facility utilizes a flash furnace smelting system that combines
the roasting and smelting stages of the production process.
Through roasting and smelting, raw ore concentrate and silica
fluxes produce high-grade copper matte (largely cuprous sulfide
and ferrous sulfide). The remaining iron and sulfur are removed
from the matte in a converter to yield molten blister copper. A
Pierce-Smith converter, the most commonly used converter in the
U.S., is assumed to be used in this model plant. Finally, the
blister copper undergoes both fire and electrolytic refining to
eliminate any contaminants and produce copper that can be as much
as 99.97 percent pure. Because a flash furnace is used, an
electric arc furnace is employed to clean the flash furnace of
copper buildup.
Sulfur dioxide and particulate matter are produced during
the roasting/smelting, conversion, and fire refinement stages.
The converter is the most significant source of S02 gas. For
this reason, both primary hoods located at the furnace mouth, and
secondary hoods that vent the building containment area, are used
to control converter emissions. Primary capture hoods alone are
sufficient to control emissions from the flash furnace and the
fire refinement equipment. The gas effluent streams of fugitive
S02 emissions that are not captured by hoods are combined for S02
and particulate removal. Typically, double pass electrostatic
precipitators (ESP) and sulfuric acid plants are used to remove
particulate matter and S02, respectively. Once treated, the
effluent gases are released through the main common stack.
Emissions from the electric arc slag cleaning furnace are
captured and treated separately from effluent produced by the
other stages. The model plant employs a flue gas desulfurization
scrubber to remove S02 before it is vented out of its own stack.
Typical copper smelters have one stack (or more) for gases
treated by the acid plant and one stack (or more) for gas streams
with lower S02 concentrations. The model plant has a total of
two stacks.
Monitoring Data.
Using monitoring data obtained from the "Summary of 1988-
1995 Ambient 5-Minute S02 Concentration Data3," a statistical
distribution was developed to predict the number of annual
exceedances and severity (in concentration) at typical copper
smelting facilities. The probability curve was based on
A-3
-------
u
a
..•T
Cn
id
CO
O
O
O
h
Oi
4J
rt
id
i 0) o c_
3-..
13 C.Z
Oo"
la u <
4
0)
M
g,
•H
fc
"O
01
o
(O
J 3
en i*.
C
a;
u
C
o •
CJ
° s
c
c
t
cue
C
) *^*
J •"
=1
3 CO
e*. \
Flash
(Roasting
Con ver
CJ
S;
c
u
— **• 1_ — *"
u
en
/•^
^^•^^•^^n
U
Fire
liefinem
!
L.
0)
a.
o.
0
L-^^ —^
0)
•o
o
c
•<
U —I
"_;: e
Electorlyl
Refineme;
n.
fr<
m
CTI
di
01
i_
Ol
n.
Q.
o
u
-------
the Gamma distribution and predicted the likelihood of exceeding
the threshold concentrations ( 0.6, 1.0 and 2.0 ppm). The right
side tail of this distribution shows the estimated pre-control
concentration and is presented in Figure A-1.2. This figure
shows 74 annual exceedances of 0.6 ppm for a 5-minute period
within an hour. An annual average of 24 exceedances of 1.0 ppm
and three exceedances of 2.0 ppm are predicted (see Table
A-l.l.). This table shows the total number of annual exceedances
of the 5-minute S02 value as well as the number in the
concentration increment (i.e., 0.6 to 0.7 ppm). Since the IL
program is based on public endangerment, both the frequency and
'severity of the exceedances will be considered in determining
appropriate remedial action. As human health impacts are
dependant on ambient concentrations, this information is equally
important in determining regulatory benefits that will be derived
from reductions in ambient levels.
Baseline Conditions
The copper smelter example represents a scenario with a
fairly high number of exceedances, however, the potential to
exceed 2.0 ppm is fairly small. This represents a "middle of the
road" scenario based on currently known exceedance situations.
The IL program offers states flexibility in addressing this type
of situation. The burden on the affected facility can vary
substantially based on the severity of the problem and the
availability of remediation options that would be acceptable to
the state. Therefore, a discussion of the selected remedial plan
and its associated costs is provided below. This situation
represents a departure from traditional "end-of-pipe" controls in
that the selected plan contains intermittent controls and use of
stacks greater in height than GEP. Presently, compliance with
NAAQS cannot consider stack heights greater than GE (213 ft. Or
65m.). As this short-term S02 is proposed under section 303,
current SIP preclusions against intermittent controls and stack
heights greater than GEP would not apply to control strategies
under this program, as long as the source continuous to comply
with ambient air requirements as would occur at permitted stack
heights.
1) Existing monitoring data shows a high number of
exceedances of the 5-minute level of concern. Analysis
of 3 years of data showed exceedances of the 0.6 ppm
level of concern occurred an average of 74 hours per
year. 3 years of exceedance data are presented in
Table A-1.2.
2) Upon review of NESHAP Primary Copper Smelter Final
Report, it was determined that the majority of copper
smelters were located in small, rural towns with
populations of 2,000 or less. For this reason a^small,
rural town scenario was chosen for this study. Given
A-5
-------
(0
H
I
u
c
o
U
CQ
•si
•&
O
C/3
C
'i
wi
•c
3
"j
o
X
Ul
~
C
<
•9
^J
'J
•J
"J1
£
u
•J
•>
3
3
o
U
H
i
o
h
g,
•H
h
(uidd
-H
Q
o
oo
o
in
O
(N
-------
0)
in
ro
o
m
in
•
CM
O
CN
in
r-
•
H
in
«
H
m
i-i
OJ
»
H
H
•
H
O
H
O\
O
00
O
[>
O
VD
O
rf
O
cs
o
S02
Concentration3
O
O
H
m
«*
00
(N
H
VD
H
O
CM
VD
OJ
m
ro
m
«*
V£l
in
^
C^
# Annual
Exceedances
O
0
rH
M
OJ
m
in
m
<*
W3
>
0
H
m
H
00
H
# Exceedances
in Increment
o
o
o
o
0
o
o
o
o
o
o
o
H
H
in
c\
H
# Exceedances
After Controlb
a. Source: Gamma distribution of monitor data from ICF Kaiser, 1995
b. Control strategy: stack height of 1040 ft., and 800 ft. for the acid plant.
-------
the large number of violations indicated by the
distribution function, it is likely that the local
community around the model facility is exposed to
levels of S02 greater than 0.60 ppm, and as much as 3
annual exceedances of 2.0 ppm. This analysis assumes
the state has decided to act on the 0.6 ppm level of
concern and that the state operates two air quality
monitors around the model plant in which the smelter is
located.
3) Additional investigation of the monitoring data shows a
seasonal nature in the exceedances, indicating that the
exceedances could more closely be related to thermal
inversions and stagnant weather conditions than to high
emissions or upset conditions occurring at the
facility. It is also assumed that the facility has
installed adequate controls that demonstrate attainment
of the current S02 NAAQS. As the process is already
controlled, most S02 is already removed from the
emission stream. In negotiation with the state to
implement the IL program, the facility asserts that the
location of the town and the smelter in a valley would
require the addition of redundant controls to ensure
against future exceedances of the concern level. If the
facility was required to install additional add-on air
pollution controls to streams with existing controls,
these additional controls would have costs comparable
to the original control equipment yet remove relatively
less S02, yielding low cost effectiveness. Due to the
presence of thermal inversion the facility asserts that
if the stack is built tall enough to overcome the
"inversion cap" then the stack emissions will be
released outside of the valley and away from the town.
The state agrees to consider review of such an
alternative pending dispersion modeling conducted by
the facility as a basis to compare control
alternatives.
4) The facility submitted roll back modeling to evaluate
options for additional controls and to evaluate the use
of taller stacks to improve dispersion during periods
of poor atmospheric mixing. The facility used 1991
data (as the worst year from Table A-1.2.) and modeled
those hourly conditions that had produced exceedances
during that year. Three major emissions .points were
modeled, these were the main stack, the slag stack, and
fugitive process emissions. Most emissions came from
the main stack. Increased stack height was
demonstrated to elevate the plume out of the valley and
result in substantial improvement in ambient air
quality. In order to achieve similar improveme»fee in
ambient air quality, additional controls would be
A-8
-------
needed. The facility estimated improvements to capture
efficiency for converter hoods (initial S02 capture)
and improvements to the acid plant SO2 removal
efficiency would cost between $120 million and $185
million (based on information submitted to EPA during
public comment), respectively, in order to achieve the
same benefit as the stack height increase. Addition of
two new stacks of 1,040 and 800 feet would result in
equivalent ambient air quality improvement during
stagnant air conditions at an estimated cost of $13.9
~ million. Based on this information, an agreement to
proceed with a plan for the use of taller stacks was
approved.
5) Although the state and the facility agreed to the
principal of using tall stacks, one major point of
contention remained. The state asserted that use of
the tall stacks would only be permitted on an
intermittent basis. The facility asserted that brief
(i.e., less than 1 month) and intermittent use of the
stacks would cause chemical and thermal damage to the
system because repeated cycling between stacks would
elevate acid formation in the stack leading to chemical
damage and increased stress on refractory due to
increased temperature changes. Specifically,
intermittent switching to and from the taller stack
could cause condensation of acid vapors on the cool
surfaces of the stack and promote corrosion. Another
concern is that repeated heating and cooling of the
refractory lining of the ducts and stacks would cause
damage due to the expansion and contraction of the
heating and cooling bricks. The state expressed concern
that prolonged use of taller stacks might have long-
range adverse impacts on the environment, such as
decreased visibility, increased ecological damages, or
an increased contribution to acid rain concentrations.
The facility performed visibility modeling and
predicted no adverse impacts. A compromise allowed for
use of the tall stacks constantly from September
through February. During warm weather months, the tall
stacks would be used during stagnant conditions only, a
decision would be made on a daily basis based on pre-
established meteorological and stack dispersion
characteristics determined by the modeling effort as to
whether use of the tall stacks would be required.
Cost of Control
The capital and annual costs for remediation of short-term
S02 concentrations through the use of tall stacks are presented
below. For the purposes of this case study, it was assumed-that
two tall stacks would be needed to control two major gas streams
A-9
-------
Table A-1.2. Model Copper Smelter-3 Years of Exceedance
Data
A-10
-------
at the facility. This study also assumes that it was not
economical to try to extend the stacks and that two new stacks
were required. A summary of these estimates is presented in
Table A-1.3., while Tables A-1.4 and A.5 present detailed
information on the cost calculation using a format derived from
the OAQPS Control Cost Manual4. The stack 's useful life was
assumed to be 30 years and the annualized costs were calculated
using an interest rate of 7.0 percent for the payment of the
capital cost of the stack. Overall stack costs were taken from a
memorandum regarding the Supplemental Section 303 Cost Analysis
for the Regulatory Impact Analysis for the Proposed Regulatory
©ptions to Address Short-term Peak Sulfur Dioxide Exposures5.
Maintenance costs and costs of other items such as electricity
for elevators and airplane warning lights were estimated based on
the engineering judgement.
Additional capital costs included money necessary to perform
air dispersion modeling to determine adequate stack heights. The
modeling is assumed to cost $100,000 and will be redone every 5
years. The capital recovery factor for the modeling is 4.39
percent for 5 years. The annualized modeling cost is $24,390 per
year. The annual costs for operation and maintenance of the tall
stacks were adjusted upward to reflect higher anticipated costs
resulting from elevated levels of acid gases in the stacks in
addition to heating and cooling stress on the refractory lining
in the duct work and stack. The costs are summarized below.
Table A- 1.3
Case 1: Cost of Control (1993 dollars)
Affected Unit
Main Stack
(1040 ft)
Slag Stack (800
ft)
Dispersion
Modeling
Total
Capital Cost
$7.3 million
$6.6 million
$0.1 million
$14 million
Annualized Cost
$0.97 million
$0.88 million
$0.024 million
$1.87 million
A-ll
-------
Summary
For the copper smelter case study, the majority of the
concentrations of S02 measured over a 5-minute period occur
around the concern level of the IL program. The frequency of
these occurrences, the geography of the area, and the historical
weather patterns give evidence that the risk is great enough for
the State to require action to protect the health of the at-risk
population surrounding the model plant.
Because of the existence of seasonal thermal inversions
around the facility, the State and source negotiated a plan of
intermittent control for 6 months of the year. It was determined
that add-on controls would be costly (capital costs of
approximately $120 to $185 million). Such controls would result
in minimal annual emissions reductions because they would be
redundant to existing controls. The intermittent use of taller
stacks to push plumes above thermal inversion layers would
reduce the risk of exposure to the local population at
significantly lower costs than air pollution control equipment.
The source provided dispersion modeling to demonstrate that the
use of taller stacks would not affect attainment of the NAAQS,
adversely impact ecological and agricultural species, or
contribute to increased acid rain. The State allowed the use of
taller stacks to address the 5-minute problem only if current
requirements in the source's permit to attain the current NAAQS
were not violated. The resulting cost of the resolution is
approximately $2 million per year, which is substantially less
that other options.
A-12
-------
Table A-1.4(a) Cost of New Intermittent Main Stack
CAPITAL COST DETERMINATION
Construction of Tall Stacks for Copper Smelter
Main Stack (1040 feet)
irect Costs
'urchased equipment costs $425,000.00
Additional ID Fan, Elevator
Aircraft Avoidance Lights
Ductwork, Ash Hopper
Freight (.05 of EC) $21,250.00
Purchased eqpmt. cost, PEC $446,250.00
irect installation costs
Foundations & supports $500,000.00
Construction & Materials $5,900,000.00
Electrical (. 10 of PEC) $44,625.00
Ductwork $320,000.00
Insulation for ductwork $17,850.00
Direct installation cost $6,782,475.00
ite preparation $100,000.00
uildings N.A.
Total Direct Cost, DC $6,882,475.00
irect Costs
Engineering (. 10 of PEC) $44,625.00
Field expenses (.05 of PEC) $22,312.50
Contractor fees (.10 of PEC) $44,625.00
Start-up (.02 of PEC) $8,925.00
Performance test (.01 of PEC) $4,462.50
Contingencies (.03 of PEC) $13,387.50
Total Indirect Cost, 1C $71,400.00
"AL CAPITAL INVESTMENT = DC + 1C $7,300,125.00
A-13
-------
Table A-1.4(b) Cost of New Intermittent Main Stack (cont.)
COPPER SMELTER ANNUALIZED COST SHEET (Main Stack)
Direct Annual Costs
Factor
Unit Cost
Total
Operating Labor
Operator
Supervisor
Maintenance
Labor
Material
Utilities
Natural Gas
Electricity
Total DC
4 hrs/stack changeover 15.77/hr *
.15 of operator -
Repair stack & ductwork 17.35/hr *
Replace refractory, dampers, -
expansion joints,
fan blades, ect.
Aircraft Lights,
Elevators, Ash
Removal
$3.50/kft~3
$0.08/kWhr
$1,892.40
$1,182.75
$112,500.00
$75,000.00
N.A.
$8,608.00
$199,183.15
Indirect Annual Costs
Overhead
Administrative charges
Property Taxes
Insurance
Capital Recovery
Total 1C
Factor
.60 of operating,
supv., & maint.
labor & materials
.005 of TCI
.01 of TCI
.01 of TCI
TCI x CRF ****
Unit Cost
Total
$69,345.09
$36,500.63
N.A.
$73,001.25
$588,290.82
$767,137.78
TOTALANNUAL COST
$966,320.93
n = 30 yr equipment life
i = 7% interest rate
8.0586%
A-14
-------
Table A-1.5(a> Cost of New Intermittent Slag stack
CAPITAL COST DETERMINATION
Construction of Tall Stacks for Copper Smelter
Slag Stack (800 feet)
irect Costs
Purchased equipment costs $400,000.00
Additional ID Fan, Elevator
Aircraft Avoidance Lights
Ductwork, Ash Hopper
Freight (.05 of EC) $20,000.00
Purchased eqpmt. cost, PEC $420,000.00
Direct installation costs
Foundations & supports $390,000.00
Construction & Materials $5,350,000.00
Electrical (.10 of PEC) $42,000.00
Ductwork $320,000.00
Insulation for ductwork $16,800.00
Direct installation cost $6,118,800.00
Site preparation $100,000.00
Suildings N.A.
Total Direct Cost, DC $6,218,800.00
direct Costs
Engineering (.10 of PEC) $42,000.00
Field expenses (.05 of PEC) $21,000.00
Contractor fees (.10 of PEC) $42,000.00
Start-up (.02 of PEC) $8,400.00
Performance test (.01 of PEC) $4,200.00
Contingencies (.03 of PEC) $12,600.00
Total Indirect Cost, 1C $67,200.00
DTAL CAPITAL INVESTMENT = DC + 1C $6,606,000.00
A-15
-------
Table A-1.5(b) Cost of New Intermittent Slag Stack (cont.)
COPPER SMELTER ANNUALIZED COST SHEET (Slag Stack)
Direct Annual Costs Factor Unit Cost
Total
Operating Labor
Operator
Supervisor
Maintenance
Labor
Material
Utilities
Natural Gas
Electricity
Total DC
4 hrs/stack changeover 15.77/hr *
.15 of operator
Repair stack & ductwork 17.35/hr *
Replace refractory, dampers, -
expansion joints,
fan blades, ect.
Aircraft Lights,
Elevators, Ash
Removal
$3.50/kft~3
$0.08/kWhr
$1,892.40
$1,182.75
$105,000.00
$70,000.00
N.A.
$8,608.00
$186,683.15
Indirect Annual Costs
Overhead
Administrative charges
Property Taxes
Insurance
Capital Recovery
Total 1C
Factor
.60 of operating,
supv., & maint.
labor & materials
.005 of TCI
.01 of TCI
.01 of TCI
TCI x CRF ****
Unit Cost
Total
$64,845.09
$33,030.00
N.A.
$66,060.00
$532,353.78
$696,288.87
TOTALANNUAL COST
$882,972.02
n = 30 yr equipment life
i = 7% interest rate
8.0586%
A-16
-------
A. 2 Case Study 2 i One Source Exceeding the Endancrerment Level
This case study evaluates control for a model primary lead
smelting facility. There are approximately four primary lead
smelters in operation in the United States. Each facility can
produce between 150 and 550 tons of lead per day. Particulate
matter and S02 are the pollutants of most concern to primary lead
smelters.
The lead smelter model scenario is one which represents a
great potential to exceed the 2.0 ppm endangerment level. This
scenario also shows a situation where there are large numbers of
exceedances of the 5-minute concern level of 0.6 ppm and few
exceedances of the 3-hour and 24-hour ambient standards. Figure
A-2.1 presents the modeled exceedances predicted for each S02
concentration level. The example illustrates situations where
tighter adherence to current SIP requirements will not provide
adequate protection of short-term ambient conditions. It is
assumed that this model scenario is located in hilly terrain
which causes complex weather conditions such as thermal
inversions that tend to increase exposure problems. Due to the
frequency of endangerment level concentrations, EPA and the State
are requiring immediate installation of additional air pollution
controls. This case study presents costs for installation of a
wet scrubber to control the blast furnace gas stream.
Description of Model Plant
The model plant, depicted in Figure A-1.2, is a composite of
several typical facilities and does not necessarily reflect the
process of any one actual plant, however, the model is
representative to actual facilities. Primary lead smelting
industry profiles found in the AWMA Air Pollution Engineering
Manual and AP-426 were used to develop the sample facility.
The process of smelting lead involves three primary steps -
sintering, reduction, and refinement. The sintering stage
converts sulfide ore concentrate (containing trace amounts of
copper, iron, and zinc) into sinter (PbO) via a sinter machine.
When the sulfur content of the sinter charge is between 5-7
percent, system operation and product quality are optimized.
This optimal sulfur content is maintained by adding silica and
limestone (sulfide-free fluxes) and large amounts of recycled
sinter and smelter residues to the mix. The sinter machines
continuous conveyor of perforated or slotted grates can be
ventilated by either an updraft or downdraft system. The updraft
design was chosen because it permits greater production rates,
has a lower pressure drop (i.e. requires less blower capacity),
requires less maintenance, and, perhaps most importantly, *«uLlows
the use of a weak gas recirculation methodology. This
A-17
-------
4J
a
H
a)
•o
&
0)
4J
rH
JJ
O
c
'i
a) S
- I
| J
ft *«
•8
u <
•••5 S
£ ^
e. 73
ts
a.
o
I
- U
03
H
•
C
3
o
CO
o
-------
Figure A-2.2. Lead Smelter Model Plant Process Diagram
Concentrate
Single-Stage
Acid Plant
Y
SulfbricAcid
Sintering Machine
Baghonse
A
Blast Furnace
Baghonse
Slag
Dross Furnace
Lead
A-19
-------
recirculation design permits more efficient and more economical
use of control methods, such as sulfuric-acid-recovery devices.
Reduction of lead is done in a blast furnace. The feed
material to the blast furnace is sinter (80-90 percent of charge)
and metallurgical coke (8-14 percent of charge) which is reduced
to lead bullion and slag. S02 emissions from the blast furnace
are a function of residual lead sulfide and lead sulfate content
as well as the amount of sulfur captured by constituents of the
slag, particularly copper.
A preliminary refining process of cooling and heating the
rough lead bullion in kettles (dressing) is performed to remove
metals from the lead. Sulfur-bearing materials, zinc, and/or
aluminum are mixed with the rough lead bullion to facilitate the
removal of copper. The final refinement of the lead consists of
a series of five steps that further remove metals from the lead.
The result is 99.990 - 99.999 percent pure lead that is cast into
10 Ibs "pigs."
The model plant utilizes a dual-gas-stream system to capture
a highly concentrated (5-7 percent S02) , or strong stream, from
the feed end of the machine, and a weak stream (< 2 percent)
pulled from the discharge end of the machine. The weak stream is
recirculated back through the feed bed where it might otherwise
be vented to the atmosphere after particulate removal.
Recirculation will reduce the production capacity, allow for more
convenient and cost-effective S02 removal and recovery, and
increase particulate generation at the discharge end of the
machine.
Because the smelting process produces such a concentrated
form of S02 from the sintering machine, many lead smelters find
it profitable to produce sulfuric acid from the off gas and sell
the acid commercially. Sulfuric acid is produced through a
contact process which uses a vanadium-based catalyst to turn the
S02 into S03. The S03 is then reacted with water to produce the
sulfuric acid. Before the S02 from the sintering process can be
converted to S03/ the off gases must be cleaned and dried using
electrostatic precipitators and a drying tower. The S03 is sent
to an absorbing tower where the S03 is absorbed by strong
sulfuric acid and water is added to keep the acid concentration
at 98 to 99 percent. The model facility uses a single stage
sulfuric acid plant to remove S02 from the sintering machine.
The removal efficiency is estimated at 96.5 percent.
The reduction stage will eliminate approximately 15 percent
of the total amount of sulfur found in the original ore
concentrate (versus 85 percent eliminated by sintering) with one
half in the form of S02 and the other half in the slag. The
concentration of S02 is dependent upon the amount of dilufrien air
introduced into the effluent gas stream. The S02 emissions from
A-20
-------
the reduction process consist of emissions from the blast
furnace. The reduction and dressing process emissions are ducted
to a baghouse and vented to a stack. It is assumed that there is
no S02 control is used for this stream. Table A-2.1 summarizes
the S02 emissions from the model facility.
Costs of Control
Exceedances of the endangerment level promoted quick action
by EPA and the State. Although there are exceedances of the 3
and 24-hour standards, compliance with the standards would occur
by other methods that would not solve the short-term problem.
For this reason, the entire cost of control has been attributed
to the short-term rule. However, the additional controls will
also reduce ambient concentrations of S02 to a level that should
eliminate 3 and 24-hour exceedances.
Table A-2.1.
Estimated S02 Emissions From Model Primary Lead Smelter
Process
Sintering
Machine
Blast
Furnace
Throughput*
(tons
produced)
100,000
100,000
AP-42
Emission
Factor
(Ib/ton
lead)
550
45
Control**
Efficiency
( percent)
96.5
0
Totals
Emissions
(tons per
year)
962.5
2,250.0
3,212.5
* A rough average of facility throughput data presented in
"Background Information for New Source Performance Standards:
Primary Copper, Zinc, and Lead Smelters, Volume I."
** From AP-42
The proposed method of S02 emission control is the addition
of a packed bed scrubber to the blast furnace exhaust. Applying
scrubbers to a previously uncontrolled source is much more cost
effective than adding equipment to processes which have controls.
Scrubbers are commonly used for S02 control and can achieve a
control efficiency of 95 percent. For a throughput of 100,000
tons per year, this added control would remove 2,137.5 tons of
S02 being emitted from the reduction process. This reduction in
emissions could potentially result in the control of two thirds
of the S02 previously emitted, a reduction that will was *"^~~
A-21
-------
demonstrated to eliminate 0.6 ppm exceedances. The costs of
adding this scrubber and of performing the rollback modeling are
detailed in Table A-2.2. Modeling costs are based on an assumed
initial cost of $100,000 that would not need to be repeated for a
5-year period annualized (24.39 percent capital recovery factor)
to $24,390.
The capital cost of installing a double contact sulfuric
acid scrubber is estimated to be $181,652. Cost estimates were
done using one of the twenty "CO$T-AIR" spreadsheet developed by
EPA's Office of Air Quality Planning and Standards to estimate
the costs of installing and operating several different types of
air pollution control equipment. Annualized costs were estimated
to be $317,819 for a primary lead smelting facility with an
average throughput of 100,000 tons per year. Assuming a $317,819
annualized cost and an added removal of 2137.5 tons of S02 per
year, the cost of S02 emission reduction is $149.16 per ton. The
detailed capital and annualized costs estimated by the "CO$T-AIR"
program are presented in Table A-2.2
Table A-2.2.
Case 2: Cost of Control (1993 dollars)
Affected Unit
Double Contact
Scrubber
Rollback
Modeling
Total
Capital Cost
$0.18 million
$0.1 million
$0.28 million
Annualized Cost
$0.32 million
$0.024 million
$0.344 million
Summary
Lead smelters have a large number of instances where they
exceed the 0.6 ppm ambient concentration level that is considered
a "level of concern". In order to prevent these short-term
exceedances, it is proposed that a packed bed scrubber will be
installed on the blast furnace. The reduction stage is often not
controlled for S02 emissions and it is believed that this would
have a significant positive impact on local short-term S02
levels. Emission modeling, costing approximately $60,000 per
year per facility is necessary to verify any impacts of this
strategy. It is estimated that this option would have a saved
cost, per ton of S02/ of $149.16.
A-22
-------
Table A-2.3 (a) Cost of Wet Scrubber System
TOTAL ANNUAL COST SPREADSHEET PROGRAM-WET IMPINGEMENT SCRUBBERS [1]
(Total flowrates > 77,000 acfm)
COST BASE DATE: June 1988 [2]
VAPCCI (Third Quarter 1995): [3] 115
INPUT PARAMETERS
-- Inlet stream flowrate (acfm): 51260
- Inlet flowrate/unit (acfm): 25630
- ' ' ' ' -2nd iteration: 25630
— Number of units: 2
— Inlet stream temperature (oF): 135
-- Inlet moisture content (fractional): 0.20
-- Inlet absolute humidity (Ib/lb b.d.a.): [4] 0.155
-- Inlet water flowrate (Ib/min): 212.4
-- Saturation formula parameters: [5]
Slope, B: 3.335
Intercept,A: 9.405000E-09
-- Saturation absolute humidity (Ib/lb b.d.a.): 0.1520
-- Saturation enthalpy temperature term (oF):[6) 144.9
-- Saturation temperature (oF): 145.0
-- Inlet dust loading (gr/dscf): 3.00
-- Overall control efficiency (fractional): 1
- Overall penetration (fractional): 0
— Number of stages (trays): 3
-- Scrubber liquid solids content (Ib/lb H2O): 0.11
-- Liquid/gas (L/C) ratio (gpm/1000 acfm): 2.5
-- Material of construction (see list below):[7] 1
DESIGN PARAMETERS
— Scrubber pressure drop (in. w.c.): 4.50
-- Inlet air flowrate (dscfm): [8] 18264
- Inlet (= outlet) air flowrate (Ib/min): 1369.0
- Outlet water flowrate (Ib/min): 208.1
-- Outlet total stream flowrate (acfm): 25956
-— Scrubber liquid bleed rate (gpm): 8.112
— Scrubber evaporation rate (gpm): -0.52
-- Scrubber liquid makeup rate (gpm): 7.59
A-23
-------
Table A-2.3(b) Cost of Wet Scrubber System
CAPITAL COSTS
Equipment Costs ($):
-- Scrubber, one-stage: [9]
two-stage:
three-stage:
-- Total scrubber (base):
(escalated):
-- Other (auxiliaries, e.g.):
— Total equipment:
Purchased Equipment Cost ($):
Total Capital Investment ($):
0
0
61,618
61,618
80.598
0
80,598
95,106
181,652
ANNUAL COST INPUTS
Operating factor (hr/yr):
Operating labor rate ($/hr):
Maintenance labor rate ($/hr):
Operating labor factor (hr/sh):
Maintenance labor factor (hr/sh):
Electricity price (S/kWhr):
Chemicals price ($/ton):
Process water pnce ($/1000 gal):
Wastewater treatment (S/1000 gal):
Overhead rate (fractional):
Annual interest rate (fractional):
Control system life (years):
Capital recovery factor (system):
Taxes, insurance, admin, factor:
Item
8000
12.96
14.26
8
1.50
0.059
0
0.20
3.80
0.60
1?0
10
0.1424
0.04
ANNUAL COSTS
Cost (S/vr)
Operating labor
Supervisory labor
Maintenance labor
Maintenance materials
Electricity
Chemicals
Process water
Wastewater treatment
Overhead
Taxes, insurance, administrative
Capital recovery
Total Annual Cost ($/yr)
103.680
15.552
21.384
21,384
9,965
0
729
14.796
97,200
7.266
25,863
317,819
Wt. Fact.
0.326
0.049
0.067
0.067
0.031
0.000
0.002
0.047
0.306
0.023
0.081
1.000
W.F.(cond-)
0.816
0.104
1.000
A-24
-------
Table A-2.3(c) Cost of Wet Scrubber System
Notes:
[1] Data used to develop this program were taken from 'Estimating Costs
of Air Pollution Control' (CRC Press/Lewis Publishers, 1990).
[2] Base equipment costs reflect this date.
[3] VAPCCI = Vatavuk Air Pollution Control Cost Index (for wet
scrubbers) corresponding to year and quarter shown. Base equipment cost,
purchased equipment cost, and total capital investment have been
escalated to this date via the VAPCCI and control equipment vendor data.
[4] Program calculates from the inlet moisture content.
[5] By assumption, the saturation humidity (hs)-temperature (ts) curve
is a power function, of" the form: hs = A*(ts)*B.
[6] To obtain the saturation temperature, iterate on the saturation
humidity. Continue iterating until the saturation temperature and
the saturation enthalpy term are approximately equal.
[7] Enter one of the following numbers: carbon steel--' I'; coated
carbon steel—'1.25': fiberglass-reinforced plastic (FRP) or
polyvmyl chloride (PVC)--'2.0\
[8] Measured at 70 oF and 1 atmosphere.
[9] Equipment cost is a function of the number of scrubber stages.
Cost does NOT include fans, pumps, or other auxiliaries.
A-25
-------
school-aged, children
The fifth case study looks at control of elevated emissions
of S02 resulting from pulping activities at a paper mill. At
this paper mill, pulp digestion produces S02 that is emitted at
the end of the batch operation. This venting of S02 lasts only a
short duration, but can produce substantial concentrations of
S02. The model facility for this case study is located in a
small town in the Midwest. The location of this facility in a
populated area produced complaints of shortness of breath during
these emissions periods. EPA and the State reacted to these
complaints with a monitoring program. As the results of this
monitoring effort confirmed the relationship between the digester
operation and elevated S02 concentrations, the monitoring effort
was followed by requirements for control of the S02 emissions.
This case study evaluates the costs of using scrubbers to control
these emissions. Although the scrubbers have a high capital
cost, modifications to the process, coupled with installation of
the scrubbers effectively eliminated these conditions of elevated
S02 concentrations.
Description of Source
The main source of short-term S02 emissions is an acid
sulfite pulping digester. In the typical acid sulfite process,
sulfurous acid is used along with a bisulfite such as ammonium,
magnesium, calcium, or ammonium to digest wood fiber under
elevated temperature and pressure. For the purposes of this case
study, the facility is assumed to be similar to the magnesium
process unit shown in Figure A-3.1. While other pulping
activities (such as Kraft pulping) emit reduced sulfur products,
sulfite pulping produces substantial amounts of S02. The major
source of S02 is venting of the digester at the end of the batch
operation when process gas is vented to the atmosphere. This
venting occurs through a blow tank. Process chemicals are
recovered at the end of the digester batch; brown stock washers
associated with recovery of process chemicals also produce some
emissions. At the model plant, approximately 2,100 pounds of
S02 is normally vented in three cycles taking approximately 15
minutes each. This venting takes place at the end of a digestion
cycle that takes approximately every 10 hours. There are three
digesters that operate in a sequential batch mode so that a
venting occurs approximately every 3.5 hours.
This concentrated venting over a short period of time has
resulted in frequent complaints from the public and has resulted
in complaints of shortness of breath from a nearby elementary
school. During the period of 2 years the state agency recorded
complaints on several different occasions of strong odors and
A-26
-------
£
I
i
<
0)
M»
u
-------
symptoms of painful breathing. The strong odors are likely to
result from reduced sulfur compounds and are not associated with
the short-term health impacts of. S02 under investigation in this
study. These complaints were reported by several residents and
school officials regarding exposure in close proximity to the
mill.
In response to these complaints, the EPA initiated a special
monitoring program. The average annual number of S02
concentrations over values associated with short-term S02 health
impacts are resulting from this monitoring program are summarized
in Figure A-3.2. Due to the high number of exceedances of both
the 0.6 ppm level of concern and the 2.0 ppm endangerment level,
EPA and the State determined to take prompt action.
As EPA, the State, and the facility all agreed that the only
substantial source of S02 was the venting from the digester, it
was determined that these emissions would need to be controlled.
A limited set of screening level dispersion modeling runs were
performed by EPA to determine the effect of add-on controls.
Using an estimated 95 percent control effectiveness for the S02
reduction resulting from addition of a scrubber, this screening
modeling showed that air quality would meet the 0.6 ppm level of
concern. The addition of air pollution control equipment was
determined to be the only alternative that would eliminate air
quality problems.
The addition of air pollution controls resulted in a
substantial decrease in S02 emissions. The new emissions
limitations negotiated between the facility and the regulators
established 35 Ibs./hour and 50 Ibs. over any 2-hour period as
the new emissions limits after installation of controls. As the
digester tank blows can cover more than 1-hour period, the
emission reduction was calculated over the 2-hour period as a
97.6 percent reduction in S02 emissions. As this was better than
the 95 percent emissions reduction calculated by the model, this
level of control was deemed to be adequate The overall
emissions reduction resulting from control was calculated as
2,050 Ibs. per blow. Averaging this over the 3.5 hour batch
cycle, the average emissions reduction is 585 Ibs. per hour.
Using an estimated 7,000 hours per year of operation, an annual
emissions reduction of 2,048 tons per year of S02 is achieved.
Costs of Control
As none of the monitored exceedances of the 5-minute
standard resulted in exceedances of NAAQS, the entire cost of
additional controls will be attributed to the IL program. The
costs of the screening modeling performed by EPA were minimal and
are not detailed in the overall costs. This leaves the costs of
the scrubber as the only costs for correction of this problem.
In order to effectively install controls, the digester needed to
A-28
-------
1
3
W3
e
'i
««
o
8
1
X
Ed
1 O
W M
<
o
S I1
cs
Q.
O
I
« §
- U
cr\
CN
in
C5
O
oo
O
t-
-------
be substantially rebuilt. The reconstruction of the digester and
installation of a wet scrubber to control S02 had a capital cost
of $9.2 million. This control device resulted in an emissions
reduction of over 97.6 percent of the S02 previously emitted.
The annualized costs of the entire project are estimated to be
$1.15 million. The capital and annualized costs of this project
are presented in Table A-3.1.
Table A-3.1.
Case 3: Cost of Control (1993 dollars)
Affected Unit
Double Contact
Scrubber Costs
Overall Project
Costs Including
Digester
Retrofit &
Scrubber
Total
Capital Cost
$1.99 million
$9.45 million
$9.45 million
Annualized Cost
$0.68 million
$1.15 million
$1.15 million
The annualized cost of operation is combined with the
estimated emissions reductions to provide a cost effectiveness
estimate for control. This cost effectiveness estimate is
provided in Table A-3.2. OAQPS control equipment cost models
were used to estimate an approximate capital and annual cost of
control, this information is presented in Tables A-3.3 and A-3.4
Table A-3.2. Cost Effectiveness of Control
Tons of S02
Emissions Reduced
per Year
2,048 tons
Annualized Cost of
Control in Thousands
of Dollars
$ 1,150
Cost Per Ton
Reduced (dollars
per ton)
$562
A-30
-------
Table A-3.3(a) Scrubber Costs
TOTAL ANNUAL COST SPREADSHEET PROGRAM-WET IMPINGEMENT SCRUBBERS [1]
(Total flowrates > 77,000 acfm)
COST BASE DATE: June 1988 [2]
VAPCCI (ThircJ Quarter 1995): [3] 115
INPUT PARAMETERS
-- Inlet stream flowrate (acfm): 275000
- Inlet flowrate/unit (acfm): 91667
- ' • • ' -2nd iteration: 68750
-- Number of units: 6
- Inlet stream temperature (oF): 300
— Inlet moisture content (fractional): 0.20
-- Inlet absolute humidity (Ib/lb b.d.a.): [4] 0.155
- Inlet water flowrate (Ib/min): 446.1
- Saturation formula parameters: [5]
Slope, B: 3.335
Intercept.A: 9.405000E-09
-- Saturation absolute humidity (Ib/lb b.d.a.): 0.2006
- Saturation enthalpy temperature term (oF):[6] 157.8
- Saturation temperature (oF): 157.6
-- Inlet dust loading (gr/dscf): 3.00
-- Overall control efficiency (fractional): 1
- Overall penetration (fractional): 0
-- Number of stages (trays): 3
- Scrubber liquid solids content (Ib/lb H2O): 0.11
-- Liquid/gas (L/G) ratio (gpm/1000 acfm): 2.5
- Material of construction (see list below):[7] 2
DESIGN PARAMETERS
- Scrubber pressure drop (in. w.c.): 4.50
-- Inlet air flowrate (dscfm): [8] 38355
- Inlet (= outlet) air flowrate (Ib/min): 2874.9
-- Outlet water flowrate (Ib/min): 576.7
-- Outlet total stream flowrate (acfm): 59141
- Scrubber liquid bleed rate (gpm): 17.036
- Scrubber evaporation rate (gpm): 15.68
- Scrubber liquid makeup rate (gpm): 32.71
A-31
-------
Table A-3.3(b) Scrubber Costs
CAPITAL COSTS
Equipment Costs ($):
- Scrubber, one-stage: [9]
two-stage:
" three-stage:
- Total scrubber (base):
(escalated):
- Other (auxiliaries, e.g.):
-- Total equipment:
Purchased Equipment Cost ($):
Total Capital Investment ($):
0
0
674,929
674,929
882,828
0
882,828
1,041,737
1,989,718
ANNUAL COST INPUTS
Operating factor (hr/yr):
Operating labor rate ($/hr):
Maintenance labor rate ($/hr):
Operating labor factor (hr/sh):
Maintenance labor factor (hr/sh):
Electricity price ($/kWhr):
Chemicals price ($/ton):
Process water price ($/1000 gal):
Wastewater treatment ($71000 gal):
Overhead rate (fractional):
Annual interest rate (fractional):
Control system life (years):
Capital recovery factor (system):
Taxes, insurance, admin, factor:
8000
13
14.26
8
2
0
0
0.20
3.80
0.60
0
10
0.1424
0
ANNUAL COSTS
Item
Cost ($/yr)
Wt. Fact.
Operating labor
Supervisory labor
Maintenance labor
Maintenance materials
Electricity
Chemicals
Process water
Wastewater treatment
Overhead
Taxes, insurance, administrative
Capital recovery
103,680
15,552
21,384
21,384
22,706
0
3,141
31,073
97,200
79,589
283,291
0.153
0.023
0.031
0.031
0.033
0.000
0.005
0.046
0.143
0.117
0.417
W.F.(cond.)
Total Annual Cost ($/yr)
678,999
1.000
0.382
0.534
1.000
A-32
-------
Table A-3.4(a) Project Costs
CAPITAL COST SHEET
jet Costs
jrchased equipment costs $2,000,000.00
Additional ID Fan, Elevator
Aircraft Avoidance Lights
Ductwork, Ash Hopper
Freight (.05 of EC) $100,000.00
Purchased eqpmt. cost, PEC $2,100,000.00
irect installation costs
Foundations & supports $500,000.00
Construction & Materials $5,900,000.00
Electrical (. 10 of PEC) $210,000.00
Ductwork $320,000.00
Insulation for ductwork $84,000.00
Direct installation cost $7,014,000.00
lite preparation $100,000.00
uildings N.A.
Total Direct Cost, DC $7,114,000.00
direct Costs
Engineering (.10 of PEC) $210,000.00
Field expenses (.05 of PEC) $105,000.00
Contractor fees (.10 of PEC) $210,000.00
Start-up (.02 of PEC) $42,000.00
Performance test (.01 of PEC) $21,000.00
Contingencies (.03 of PEC) $63,000.00
Total Indirect Cost, 1C $336,000.00
1C + SP + Bldg.
OTAL CAPITAL INVESTMENT = DC + 1C $9,450,000.00
A-33
-------
Table A-3.4(b) Project Costs
Direct Annual Costs
Factor
Unit Cost
Total
Operating Labor
Operator
Supervisor
Maintenance
Labor
Material
6hrs/day,360days/year $12/hr
.15 of operator -
3hrs/day;360days/yr $13.20/hr
same as labor costs -
$25,920.00
$3,888.00
$14,256.00
$14,256.00
Utilities
Natural Gas
Electricity
Total DC
$3.50/kft ~ 3
$0.08/kWhr
N.A.
$148,614.72
$206,934.72
Indirect Annual Costs
Overhead
Administrative charges
Property Taxes
Insurance
Capital Recovery
Total 1C
Factor
.60 of operating,
supv., & maint.
labor & materials
.02 of TCI
.01 of TCI
.01 of TCI
TCI x CRF ****
Unit Cost
Total
$34,992.00
$47,250.00
N.A.
$94,500.00
$761,541.51
$938,283.51
TOTALANNUAL COST
$1,145,218.23
**** CRF = i(1+i)~n/(1+i)~n-
n = 30 yr equipment life
i = 7% interest rate
8.0586%
A-34
-------
A.4 Case Study 4; One Source Impacting Multiple Communities
This case study involves a hypothetical area with a
petroleum refinery that has historical S02 emissions problems.
Numerous complaints have been received from several citizens from
"the area around the refinery of health problems including asthma
and respiratory difficulties, burning eyes, and throats. In the
past getting action from such a source has been complex and
difficult. Large facilities often have massive political clout
and impressive legal forces behind them making them extremely
powerful against the state. Sometimes citizens from several
states are affected by the emissions of one facility therefore
requiring the cooperation and coordination among numerous states.
Several facilities, including the model facility in this study,
have long histories of conflict with state regulations. This
case study will investigate a facility's use of an Operations and
Maintenance (O&M) approach coupled with Continuous Emissions
Monitoring (CEMs) as described in Case Study 2, to reduce
excessive short-term S02 emissions. In addition to O&M
activities to eliminate current problems with excess emissions,
additional controls will be needed to ensure public protection
from elevated short-term S02 concentrations. This case study
includes installation of controls on a catalytic cracking unit
that currently is uncontrolled.
Description of Source
This case study involves a localized area that intersects
several states and EPA regions. It is a highly industrialized
area with mostly "smoke stack" type sources. It contains an oil
refinery which emits large quantities of S02. In the
hypothetical situation a combination of frequent air inversions
and the existence of a downwind population makes it imperative
that the facility minimize emissions to avoid adverse impacts on
the local population.
The model refinery described in this case study converts
crude oil into various combustion fuels. The difficulty resides
in the fact that crude oil contains sulfur that has the potential
to contaminate the environment. The facility process over
200,000 barrels of crude oil per day which is considered fairly
large.
The refining process requires the use of several pieces of
equipment that produce S02. The primary source of S02 emissions
is from a large Fluid Catalytic Cracking Unit (FCC) used to
convert complex hydrocarbons into blending stocks and fuel oils.
The large FCC unit was built before the implementation of an NSPS
and therefore was allowed to emit a much higher limit of -SO^ than '
would be allowable today, as well as operate without a pollution
A-35
-------
control device. Annual allowable emissions from this unit are
2300 Ib/hr. This "grandfathered" unit has an emission limit of
2000 ppm (see Table A-4.1) whereas, the NSPS for newer catalytic
cracking units limits emission is 250 ppm S02. As a result, this
unit is the largest single source of S02 in the area.
The refinery operates a CO boiler that (as mentioned in Case
Study 2) takes flue gas from the FCC units that are steeped
Table A-4.1.
Emissions for Refinery in Case Study 4
Unit
Sulfur Recovery
Unit 1 + Sulfur
Recovery Unit 2
(SRU)
Reduced Crude
Conversion
System (RCC)
Fluid Catalytic
Cracking Unit
(FCC)
NSPS
Emission
(ppm)
250
250
2000
Annual
Allowable
SO2
Emission
(Ib/hr)
86
(50+35)
1200
2300
Annual
Allowabl
e S02
Emission
(TPY)
376.68
5256
10074
with S02 and CO. The CO boiler takes the offgases, combusts them
to produce process steam. A fluidized bed boiler into which
limestone is injected removes sulfur oxides.
Fuel oils must be hydrotreated in order to remove sulfur in
order to be sold commercially. Desulfurizing feed is mixed with
hydrogen in a catalyst reactor. In this process sulfur is
reduced to H2S. The H2S is then sent through a sulfur recovery
unit where it is turned into commercial grade sulfur that sold to
recover some of the operating costs.
The model refinery also operates two Claus Sulfur Recovery
Units (SRUs). The SRUs take off-gases put out by the FCC units
and convert it into elemental sulfur which can then be sold
commercially (this process is presented in detail in Case_jStudy
2). Operation of this equipment results in emissions of S02,
A-36
-------
H2S, and other VOCs. The SRUs spike H2S levels in the
atmosphere. The SRUs at the facility have a 99.7 removal
efficiency rate for sulfur.
The facility also employs a Shell Glaus Offgas Treating
(SCOT) unit. This equipment allows for the removal of sulfur
compounds from the SRU tail gas before its incineration. The
compounds are then converted to H2S, which is then recycled back
into the SRU.
A Reduced Crude Conversion System (RCC) that is used is
similar in function as the FCC unit, but it is designed to
process heavier feeds using a mixture of fresh catalyst and
equilibrium catalyst from the FCC. The unit processes heavy
vacuum gas oils, #3 crude unit bottoms, and #1 refinery lube
plant vacuum unit bottoms into light gases, gasoline, and cyclic
oils. As with the FCC unit, the catalyst is coked in the
cracking reaction. Sulfur is released in this reaction but as in
the FCC unit it is removed with limestone in fluidizing beds.
Baseline Conditions
The current 24-hour and 3-hour S02 NAAQS are usually met
using existing control equipment. The facility is required to
notify state officials when these levels are exceeded. As of 1993
the facility has been required to purchase and operate a 24-hour
video system which allows regulators to constantly monitor its
happenings. Ongoing exceedances of the current NAAQS are
believed to result from process upsets. The facility has
installed Continuous Emissions Monitors (CEMs) on all four of the
major S02 emitting units. The state has required an effective
notification procedure to be implemented at the refinery which
would result in swift notification of federal, state, and local
agencies in the occurrence of an emissions release. Under the
requirements of the facility's permit, the State has required a
maintenance plan to eliminate excess emissions. Although this
additional O&M should reduce short-term S02 problems, it will not
be sufficient enough to eliminate them. As this O&M plan has
been required under the existing SIP to meet NAAQS, its costs are
not part of the IL program burden. It is assumed that
improvements in O&M will, however, provide benefit by reducing
the frequency of short-term exceedances.
Monitoring Data
Presently, the model facility is assumed to have drastically
reduced 3-hour and 24-hour S02 standard violations. The facility
has improved control equipment designed to lower S02 pollution
emitted from their processes. Specifically, the addition of the
SRUs has decreased NAAQS violations drastically. The use of the
SCOT unit which is 99.7 percent effective at removing S02-4ias
aided the facility's attempts of lowering S02 emissions. The
A-37
-------
facility is currently hydro-treating off gases with desulfurizing
feed which greatly decreases emissions of sulfurous compounds.
Violations of both 3-hour and 24-hour S02 standards still exist,
but they are very infrequent.
The area surrounding the model plant still suffers from
excessive short-term concentrations of S02 resulting from process
upsets and generally high emissions from one uncontrolled unit.
Figure A-4.1 shows the cumulative annual predicted exceedances of
5-minute S02 levels of the IL program. In order to rectify the
situation a plan is developed which results in the establishment
of 5-minute emission limits for S02 sources used by the refinery
including:
1) In depth modeling of short-term (5-minute) average data
which would be used in establishing a 5-minute emissions
limit. The collection of such data will require the
construction of a meteorological station in the vicinity of
the refinery. Obtain at least 2 years of meteorological
data for comparison to monitoring data and use in dispersion
models to show post-SIP emissions.
2) Use any existing monitoring stations around the refinery
to also obtain 2 years of short-term monitoring data along
with meteorological data.
3) Appropriate 5-minute emissions limits should be
determined for the refinery from the rollback modeling using
the meteorological data. The modeled concentrations will be
compared to actual measured concentrations to ensure
validity in the modeling process.
4) Modeling will be used to determine if there is a need
for dry scrubbing of the FCC unit. A dry scrubber device
added to the unit would possibly reduce S02 emissions by 95
percent thereby eliminating short-term NAAQS violations (3
hrs) .
Costs Associated with the IL Progarm
After the 5-minute emissions are established, O&M
(Operations and Maintenance) techniques will be required to
eliminate exceedances of the IL program. As mentioned in Case
Study A-2, O&M practices are relatively cost effective. The
refinery is obligated to report any exceedances of the 5-minute
standard, explain why it occurred, and describe what they plan to
do in the future to be sure it does not happen again. This type
of system will require extra costs in monitoring and recording
these costs are presented in Table A-4.2. and are developed on
the same basis as those in Table A-2.4.
A-38
-------
Pi
w
o
4J
H
O
Pi
to
at
o
0
*d
0)
0)
o
X
H
0)
4J
0
•S"
0fcj A
1 y*
in H
<*T
°U
M
»
o
c
•J
rr
-- fN ^-,
Q.
o
m
_L
CU
u
o
U
g
5
0)
M
g,
•H
(mdd
o
r-i
O
oo
O
(N
O
-------
Currently the hypothetical refinery is using a CEMS to
monitor source emissions hourly. This means that no new
additional equipment will be necessary to comply with the IL
program requirements. The present monitors are capable of
sampling continuously and producing concentrations as needed.
These concentrations are taken from the monitor and recorded by
strip charts or digital data loggers. The charts would not need
to be modified, but they would need to be read more often. This
would lead to increased labor expenses for the facility. The
digital loggers could still be used, but they would need to be
reprogrammed to take 5-minute average readings along with the
current 3-hour and 24-hour readings.
Five-minute averages for stack flow, stack temperature,
stack concentrations, and calculated emissions requirements would
increase the facilities burden, as well as the increased
monitoring, reading, and validating the additional data gathered.
The addition of the dry scrubber device onto the
"grandfathered" FCC unit will be a substantial cost burden.
Table A-4.2. details the capital and annual costs for the
purchasing and installation of such equipment for the refinery.
For a predicted emission reduction of 95 percent, or 9,570 tons
per year S02, the annual cost of $2,189,021.76 results in a cost
of $228.74 per ton of S02 emission reduction. The additional
costs -for improved O&M are needed to meet existing SIP
requirements and are, therefore, not attributed to the IL
program. It should be noted that the FCC unit addition is the
only item presented in the costs in this case study and that ICR
provides national O&M costs and effort is not duplicated for this
report. A modeling demonstration has been included in the costs
to represent dispersion modeling conducted to demonstrate the
adequacy of the proposed controls to reduce levels to no more
that 1 exceedance hour of 0.6 ppm over any 5-minute period. The
modeling was estimated to cost $100,000 and was assumed to not
need to be repeated any more frequently than 5 years. Therefore
this cost was capitalized over 5 years using a capital recovery
factor of 24.39 percent for an annual cost of $24,390.
A-40
-------
Table A- 4. 2.
Case 4: Cost of Control (1993 dollars)
Affected Unit
Dry Scrubber
Additional CEM
Activity for
One Unit (see
Table B.4. )
Dispersion
Modeling
Total
Capital Cost
$20.4 million
none
$0.1 million
$20.5 million
Annual! zed Cost
$2.19 million
$0.01 million
(290 hours ©33.75)
$0.024 million
$2.224 million
Summary
Though the facility in this case study is fictitious, it is
intended to be representative of actual facilities. As with
actual facilities, if proper equipment controls are placed on
emission sources then subsequent control costs, operations and
maintenance, can be at a minimal. The total estimated costs to
the model source is estimated to be $228.74 per ton of S02
emission reduction. For a 95 percent reduction of S02 emissions
an annual cost of $2,189,021.76 cost of air pollution control
would be incurred for the addition of the dry scrubber to the FCC
unit. The additional costs resulting from modeling and reporting
do not provide emissions reductions and were not included in the
cost per ton reduced. However, their cost is relatively small
and including these costs makes the overall annual cost of
elimination of short-term S02 problems $2.224 millions. The new
unit addition would provide some benefit in reducing the number
of 3- and 24-hour standard exceedances for the model facility.
Coupled with the required O&M improvements to minimize excess
emissions, modeling demonstrated compliance with the NAAQS and
the IL program.
A-41
-------
Table A-4.3(a)
Costs for Addition of Dry Scrubber Device
Direct Costs
Purchased equipment costs 310,000,000.00
Additional ID Fan, Elevator
Aircraft Avoidance Lights
Ductwork, Ash Hopper
Freight (.05 of EC) $500,000.00
Purchased eqpmt. cost, PEC $10,500,000.00
Direct installation costs
Foundations & supports $500,000.00
Construction & Materials $5,900,000.00
Electrical (. 10 of PEC) $1,050,000.00
Ductwork $320,000.00
Insulation for ductwork $420,000.00
Direct installation cost $8,190,000.00
Site preparation $100,000.00
Buildings N.A.
Total Direct Cost, DC $8,290,000.00
Indirect Costs
Engineering (.10 of PEC) $1,050,000.00
Field expenses (.05 of PEC) $525,000.00
Contractor fees (.10 of PEC) $1,050,000.00
Start-up (.02 of PEC) $210,000.00
Performance test (.01 of PEC) $105,000.00
Contingencies (.03 of PEC) $315,000.00
Total Indirect Cost, 1C $1,680,000.00
TOTAL CAPITAL INVESTMENT = DC + 1C $20,370,000.00
A-42
-------
ct Annual Costs
Table A-4.3(b)
Costs for Addition of Dry Scrubber Device
Factor
Unit Cost
Total
erating Labor
Operator
Supervisor
intenance
Labor
Material
6hrs/day;360days/year S12/hr
.15 of operator -
3hrs/day;360days/yr
same as labor costs
$13.20/hr
$25,920.00
$3,888.00
$14,256.00
$14,256.00
ities
Natural-Gas
Electricity
Total DC
SS.SO/kft^S
S0.08/kWhr
N.A.
$148,614.72
$206,934.72
ect Annual Costs
erhead
ministrative charges
)perty Taxes
urance
pital Recovery
Total 1C
Factor
.60 of operating,
supv., & mamt.
labor & materials
.02 of TCI
.01 of TCI
.01 of TCI
TCI x CRF ****
Unit Cost
Total
$34,992.00
$101,850.00
N.A.
$203,700.00
$1,641,545.04
$1,982,087.04
AL ANNUAL COST
$2,189,021.76
**** CRF = i(1+i)~n/(1+i)~n-
n = 30 yr equipment life
i = 7% interest rate
8.0586%
A-43
-------
A.5. Case Study 5; Multiple Sources Affecting a Local
Community
The next case study involves a hypothetical area containing
several S02 sources that when combined, contribute to
concentrations greater than the 5-minute concern level of 0.6
ppm. In the majority of the case studies, little or no overlap
of the current 3-hour and 24-hour National Ambient Air Quality
Standards (NAAQS) with 5-minute exceedances is anticipated.
However, in this case study current State Implementation Plan
(SIP) efforts to meet the NAAQS will form the basis of the effort
to control 5-minute concentrations. In order to meet the
existing NAAQS, O&M activities and emissions controls coupled
with CEMs limits have been put in place. The CEMs will be used
to demonstrate continuous compliance with emissions standards.
While the current effort was designed to meet current NAAQS, this
effort will assist in lowering short-term emission peaks and
reducing the frequency of exceedances of 5-minute S02
concentrations.
Description of Sources
The facilities that are located in close proximity and
contribute to the 5-minute problem, include two oil refineries,
two sulfur recovery facilities, and a coal burning power plant.
A description of the model plants associated with these
facilities is provided below.
Oil Refineries: The two refineries in the area refine
approximately 50,000 barrels of crude oil per day, which
classifies them as medium in size. The refining process turns
crude oil into various petroleum products including liquefied
petroleum gas, gasoline, kerosene, jet fuel, diesel fuel, fuel
oils and lubricating oils. Refining operations consist of
separation operations, conversion processes and petroleum product
treatments. These three processes are responsible for splitting
crude oil into its various components, recombining them into
useful fuels and lubricants and then treating and blending them
to stabilize and enhance the performance characteristics of the
final petroleum products.
Separation processes involve atmospheric and vacuum
distillation of the crude oil into various petroleum fractions.
These fractions are then broken down or combined to form various
products through the conversion process. Some conversion
processes are cracking and coking to break large molecules into
smaller ones; isomerization and reforming to rearrange moiecular
A-44
-------
structures of compounds; and polymerization and alkylation to
combine smaller molecules into larger ones. Treatment processes
are used to remove impurities from the petroleum products.
Treatments include deasphalting, hydrodesulfurization,
hydrotreating, chemical sweetening and acid gas removal.
Sources of emissions from petroleum refineries are generally
the volatile petroleum products or the combustion sources, which
results from sulfur existing as an impurity in the crude oil.
Process heaters and boilers at the two facilities run off of oil
and fuel gas and as such are the largest contributor to S02
emissions at these sources. The sulfur content of the fuel oils
for the two refineries varies between 3 and 6 percent by weight.
Primary sources of S02 emissions from the refineries are the
fluid catalytic cracking (FCC) unit, carbon monoxide (CO) boilers
and the coker CO boilers.
The FCC units use catalysts and high temperatures to convert
complex hydrocarbons into blending stocks and fuel oils. A
gas/oil stream entering the FCC is heated and fed into a
catalytic reactor, where the cracking reaction occurs. The
reactor vapors are then sent to fractionation columns where they
continue to be processed. The pores of the catalyst are covered
with hydrocarbons, sulfur, nitrogen, and trace metals. The spent
catalyst is then stripped of the entrained impurities in a
catalyst regenerator. Hot combustion air and steam strippers
remove the hydrocarbons and the off-gases are removed from the
regenerator by a flue. The flue gas which is rich in CO and S02
is then sent to a CO boiler which combusts the reactor off gases
to generate process steam. S02 and other pollutants are vented
from the CO boiler stack to the atmosphere.
The coking unit uses high temperature treatments of heavy
residual oils to produce light hydrocarbon products and petroleum
coke. The coke particles are burned in a combustor. The off-
gases which contain S02 from the combustor are sent to the CO
boiler where they are burned along with supplemental fuel oil.
As part of the fuel treatment process, fuel oils are
hydrotreated to remove sulfur before they are marketed. The
hydrotreatment process mixes the oil feed with hydrogen in a
catalyst reactor. A byproduct of the hydrotreatment is sulfuric
acid (H2S), ammonia, nickel and volatile organic constituents
(VOCs). The H2S is sent to a local sulfur recovery facility to
turn it into elemental sulfur to be sold commercially.
Sulfur Recovery Facilities: The sulfur recovery facility uses a
Glaus sulfur recovery unit (SRU) to extract the sulfur from the
refinery off-gases. The sulfur recovery facility is located
adjacent to one of the refineries and handles only waste gas from
A-45
-------
this refinery, and is classified as medium in size. Within the
SRU, part of the H2S is oxidized to form S02. The remaining part
of the H2S reacts with the S02 to form elemental sulfur as shown
below.
2H2S -i- 302 -> 2S02 + 2H20 + Heat
2H2S + S02 ~> 3S + 2H20 + Heat
The product gases of the reaction are sent to a waste heat
boiler. The gases then pass through a condenser where elemental
sulfur is removed. The gas stream is then reheated and sent to a
catalytic reactor where the H2S and S02 reaction continues. The
reactor condenser process is repeated twice to remove as much of
the sulfur as possible from the off gases. Off-gases of this
process contain S02, H2S, other reduced sulfur compounds and
VOCs.
Coal-Fired Power Plant; The power plant in this area uses a 165
MW coal-fired utility boiler, which is classified as small in
size. The power plant burns an average of 88 tons of coal per
hour with a peak of 110 tons per hour during peak winter months.
As the coal is combusted S02 is emitted from a 350 foot tall
stack.
Cogeneration Facility: Within the area is also a cogeneration
facility that burns petroleum coke from one of the refineries and
produces process steam for use within the refinery to replace
high sulfur fuel oil. The cogeneration facility produces S02
from the coke burning operations.
Baseline Conditions
Existing emission limits for S02 are intended to prevent
exceedances of the 3-hour and 24-hour SO2 NAAQS. Major sources
of emissions from each facility have specific emission limits,
while the remaining sources, such as valves, vents and flanges,
are required to use appropriate maintenance, repair and operating
practices.
Along with the emission limits are additional requirements
for facilities during meteorological situations that prevent
dispersion of S02 emissions. These requirements have
supplemental emission limitations when calculated buoyancy fluxes
are below a certain level. When these meteorological conditions
exist, emissions of S02 are required to be reduced so that ground
level concentrations do not exceed the standards due to reduced
pollutant dispersion from the lower buoyancy fluxes.
In order to assure compliance with the emission limifeations
A-46
-------
for the sources in the area, the regulatory agency instituted
several monitoring and reporting requirements for the local S02
sources.
Each source which has a 3 or 24-hour emission limit is
required to install and maintain a CEM. The GEM must record at
least 90 percent of the operating time for that source.
Monitoring Data
The close proximity of the sources has produced continued
violations of the 24-hour S02 standard. The frequency of these
violations has been low, with an average of about 3 violations
per year. Violations of the 3-hour standard have occurred,
however, they are less frequent. Previous SIP efforts have not
eliminated violations, so the regulatory agency required the
installation of CEMs on all of the major S02 sources coupled with
emissions limitations tied to 3-hour and 24-hour emissions
standards. Dispersion modeling was used to establish new
emissions limitations for each facility that will be verified
through CEM monitoring. These new emissions limits are predicted
to eliminate NAAQS violations.
Prior to the installation of CEMs, monitoring data indicate
an average of 32 exceedances of the short-term standard of 0.6
ppm for 5 minutes out of an hour. After installation of the
CEMs, data show twelve violations of the concern level. Thus, it
has been assumed that the current SIP strategy will not eliminate
5-minute exceedances.
To address the 5-minute problem, a working group is
established representing the State and the regulated facilities.
A plan is developed in response to the State's prediction of
continued exceedances of 5-minute levels. The major elements of
the plan are:
1) Recalibrate the meteorological station to obtain 2 years
of 5-minute average data for use in establishing short-term
emissions limits.
2) Use existing monitoring stations to obtain 2 years of
short-term monitoring data concurrent with meteorological
data, to help identify those conditions producing high
short-term S02 concentrations.
3) Current information indicates that short-term events
result from stagnant conditions occurring during winter
weather. These conditions are fairly rare; typical high
wind speeds in the area are sufficient to produce adequate
dispersion for avoidance of short-term exceedances. The
design plan will, therefore, focus on establishing ^_
A-47
-------
V
o =
c/3 .=
e |
°S i
1/1 3
i o
I I
I 3
c3 i
y
'•3
&• ?
~ <£
* -J-
I
at
I
CJ
oo
£
U
(ludd Q-Q
in o m
r-i r-i —
-------
intermittent emissions limits.
4) An analysis will be performed to determine total S02
reduction required in aggregate for the area to determine
suitable emission limits for the sources.
5) The Acid Rain Emissions Trading Program is being used as
a model for achieving desired emissions reductions during
these stagnant conditions. The State will allocate
allowances to each source to achieve the aggregate S02
reductions necessary to eliminate the 5-minute problem.
Overall, the State will use the meteorological and
monitoring data to establish periods of the year when
intermittent control are required to prevent 5-minute bursts of
S02. During these periods of the year, the sources can use a
trading program to reduct the combined effect of 5-minute S02
emissions at the lowest cost. In the program, sources use a
variety of control methods that are viable under the IL program.
Activities such as temporarily scaling back production rates at
sulfur recovery facilities, or use of lower sulfur coal at the
power plant represent substantially more cost effective
approaches to control than requiring additional controls at the
refineries. It is assumed that the refineries are business
competitors, and therefore, unlikely to trade. However, the
sulfur recovery facilities are dependant on refinery operation
and would be likely to scale back operation provided that they
were compensated (allowances purchased) for their efforts. The
power plant could also conceivably sell allowances based on
demand on the grid (ability to lower production) and on their
ability to utilize cleaner fuels.
Costs Associated with the IL Program
Costs to Regulated Facilities: To meet the SIP requirements,
facilities within the area are already monitoring their emissions
on an hourly basis. For the IL program, monitors will sample
continuously and are capable of producing the concentrations as
needed. The concentration data from the monitors is recorded by
either strip charts or by digital data loggers. Strip charts
would not have to be modified but would have to be read more
often. The digital data loggers would have to be reprogrammed to
report 5-minute averages in addition to the 3 and 24-hour
averages. Because, no new or additional equipment is anticipated
to comply with the monitoring requirements of the IL program.
All costs related to purchase of GEMS and the activities to meet
the SIP are assumed to be baseline costs and are not included as
additional costs imposed by the IL program.
Facilities would incur an increased burden due to the*-—
A-49
-------
additional amount of reporting due to the 5-minute continuous
monitoring requirements. Five minute averages for stack flow,
stack temperature, stack concentrations, and calculated emissions
would be required in addition to the current reporting
requirements. Additional effort will be required to validate the
added amount of monitoring data that would be collected. Once
collected, this information will need to be reported to the
State. The costs of this additional reporting are assumed to be
similar to the incremental reporting burden costs developed for
the continuous air monitoring (CAM) Rule7,8. These costs assume
0.5 hours of new record keeping per pollutant point for 260 days
per year; an additional 0.5 hours twice a year is the assumed
burden for transmitting this information to the State. As three
parameters are being recorded (flow, temperature, and
concentration), these burdens are multiplied by three assuming
the burden is similar to tracking three different pollutants.
This produces an annual incremental burden of 393 hours per year
for each of the process units identified in Table A-5.1. The CAM
Rule regulatory impact assessment (RIA) identifies this task as
being carried out by an employee with a burdened hourly pay scale
of $40.00. The annualized cost of the incremental burden of the
5-minute monitoring effort is calculated to be $15,720 per
regulated stack. With 11 stacks in the area, this totals to
$172,920 per year.
In this case study, it is assumed that the final benefit of
implementing the current SIP efforts to achieve the 3-hour and
24-hour standards has not yet been fully determined. Sufficient
information is not available at the onset of the program to fully
define the emissions reductions needed to meet short-term ambient
concentration goals. The final benefit of implementation of the
current SIP efforts to achieve the 3-hour and 24-hour standards
has not yet been fully determined. However, two roll-back
scenarios have been developed to account for the costs of
reducing emissions beyond the current SIP efforts to meet 5-
minute standards. These rollback scenarios assume that either a
10 percent or 20 percent increase in emissions reduction will be
required beyond the SIP required 3-hour limits in Table A-5.1.
The costs of a linear reduction (rollback) of 10 percent and 20
percent of the remaining emissions were estimated based on the
market based cost per ton of emissions reductions is indicated by
the SO2 Allowance Trading Program. A cost of $270 per ton
reduced was used to estimate the cost of control9. As previously
mentioned, the number of exceedances of the 5-minute standard has
decreased to approximately 12 per year based on current
information. As this data represents a limited set of
observations, 25 days per year has been used in the roll-back
model to represent the number of annual days during which
emissions will need to be lowered and trading will occur.
Costs to Regulatory Agencies: Costs for increased ambient and
A-50
-------
meteorological monitoring efforts are not included in the case
studies, these costs are documented nationally in the Information
Collection Request (ICR) memo that is submitted to the docket in
support of the modeling requirements contained in the proposal of
CFR Part 58. The cost estimates in this case study reflect the
additional reporting requirements the regulatory agency will have
to review and expend added effort to insure compliance with the
new requirements associated with expending man-hours for
compliance audits of the facilities, reviews of source tests and
-compliance reports and for the preparation and review on
monitoring protocols. Manpower hours were determined on a per
source unit, a per facility and an overall fixed cost basis for
the various tasks necessary to carry out the enforcement of the
short-term emission limit. Table A-5.2 demonstrates that the
added work loads for the local regulatory agency totals to 953.3
hours per year and $32,173.88.
Summary
The total estimated cost to the affected sources was
estimated to be between $210,855 and $248,890 per year depending
on the amount of emissions reduction required to eliminate the
few remaining elevated short-term S02values. The cost of
monitoring and reporting may somewhat overestimate the reporting
burden due to the fact that the facilities are already required
to report on longer term basis. However, the overall cost of
reporting developed for the Enhanced Monitoring Rule RIA was
considered the best estimate available. The overall burden of
this case study on both the affected sources and the permitting
agency combined is estimated to be between $243,029 and $280,964
per year.
A-51
-------
In
id
o
•0
m
at
at
H
H i-l
• O
in M
i 4->
< a
O
0) U
H
n IH
a) o
EH
!••
JJ
a
o
L>
^
in
0)
ra
id
03
4J
CO
0
CJ
rd
4-)
O
4-J ^
CO O
UJ O rd
U 4J
CD CQ 03
c CD^; ^
•H £ U 0
>-l -H rd CN
rH O 4J 4-> >
rd 4-> M co •>
3 -H o in
c c a j-i H
5 o o o to-
rt, S PH M-I ^"
HH CD
0 C
-H rH
M MO
QJ -H M
3 rd (U O
S Q & CJ
£-4
0) T)
OH CU
O
CO C T)
O O CD
U EH OH
O
rd —
A a >,
•H O rd
rH -H T3
CM o co
3 fi
<&> "O O
O CD 4J
CN CE! — '
0
£} C >>
rH o rd
rH -H TJ
O 4-) ^.
OS U CQ
3 a
«*p 73 O
0 (U 4J
H 04 —
in
m
CO
o
H
CN
o
CN
CT>
»
CN
[^
H
in
o
CN
CN
VD
•
m
CQ
-H 0
03 H
03 •—
•H
S CQ
O
t| | .J
0 4-)
U
4J ^J
03 -a
O 0)
U ptj
0
CO
CO
^f
CN
0
CN
(Ti
»
CN
f»-
H
in
CN
0
CN
<*
CN
H
H
03
•H O
03 CN
CQ —
-H
S 03
W JH
o
IH -H
O 4J
o
4J 3
CQ T3
O (U
U OJ
0)
U
a
id
a
m
™!
0)
U
d
id
-H
rH
Q.
e
U
•O
d
id
in "H
«*
n
o
jj
a
0
u
d
•0
£j
(Q
rH
in
c>»
i
n
m
•M-
U)
•
CN
in
H
H
o
1
U)
CO
H
^
CQ
(U
(U S
U Q)
0">
O QJ
CQ «
kO
•
•<*
m
v>
in
>
,
m
m
H
•
O
H
in
H
H
O
1
H
U)
Q)
U
a
rd CQ
-H 4J
rH H
ao
6 a
O Q)
CJ PH
o
•
0
£ °
«-
in
c^
«
m
m
•w-
o
o
CN
m
H
H
1
0
1
a
o
•H
rH 4J
0 rd
o M
0 rd
O CD
V-i M
CM Oi
,
m
f»
H 00
- 00
CN
m
-------
A. 6 Case Study 6; Several Sources with Exceedances at Night
The case studies presented to this point have demonstrated
some situations in which there was a need for implementation of
the IL program and have discussed the costs associated with
implementation. In the two case studies that follow, the
regulatory authorities (State or local agencies) investigate
situations where exceedances of the concern level are known, but
as a result of a simplified risk assessment, they have determined
that the risks associated with the violations were not
significant enough to warrant action under the IL program.
In this case study, monitor data is evaluated for an area
with three coke oven facilities in close proximity to each other.
The area is part of a metropolitan statistical area (MSA) and is
officially designated as a nonattainment area for the S02 NAAQS.
However, due to a consistent record of attaining the 3-hour,
24-hour, and annual NAAQS, the regulatory authority has requested
the EPA to redesignate the area to attainment. Because of the
potential to be redesignated to attainment, the regulatory
authority has devoted 10 monitors to evaluate S02 emissions in
the area. Eight of the monitors are located around the coke oven
facilities, five of which have reported measurements in excess of
0.60 ppm over a 5-minute period.
Coke Oven Production Process and Emissions
Iron and steel are refined metals used for making several
various products. In a series of processes, refined iron ore is
manipulated in blast furnaces to produce iron metals, which are
then used in the production of steel. Coke is the chief fuel
used in blast furnaces for the conversion of iron ore into iron
metals. Coke is a metallurgical coal that has been baked into a
charcoal-like substance that burns more evenly and has more
structural strength than coal.
The coking procedure is performed in ovens that are
constructed in groups with common side-walls, called batteries.
During the coking process, coal is fed into the coke oven battery
through ports at the top of the oven, which are then covered with
lids. The coal is then heated in the absence of air in specially
designed refractory chambers. Volatile material is driven off in
the form of raw coke oven gas and then piped through an offtake
system (for distillation and separation), where valuable by-
products such as phenols, naphthalene, benzene, toluene, and
ammonia are recovered as part of the production process10.
After valuable by-products are removed from the coke oven
gas, the remaining products could be fed to a desulfuriza£ipn
A-53
-------
plant to reduce the sulfur contained in the gas to levels
acceptable for fuel use. The sulfur in coke oven gas exists as
H2S and organic sulfur compounds (primarily carbon disulfide,
CS2, and carbonyl sulfide, COS). A fairly typical coking coal
might contain about 1 percent sulfur, and about half of the
sulfur remains in the coke after carbonization. Perhaps 95
percent by volume of the sulfur in the coke oven gas is in the
form of H2S; of the remainder, CS2 accounts for 3.5 percent and
COS for 1.5 percent11. When coke oven gas that has not been
desulfurized is burned as a fuel, sulfur is emitted as S02.
Desulfurization has a long history, as sulfur was once removed
from gas for residential fuel use by contact with iron oxide, or
through the absorption of acidic gases in a basic solution or
oxidizing solution. With the advent of natural gas in the
1950's, desulfurization became much less common and is now only
practiced at larger facilities.
Monitoring Data.
In the analysis that follows, ten monitors in the area are
used to evaluate 5-minute ambient concentrations of S02 around
three coke oven facilities. The first facility is a small entity
that operates one by-product recovery coke battery with less than
75 ovens, and produces approximately 20 MMCF of coke oven gas
(COG) per day. The second facility, which is moderate in size,
has five coke batteries with more than 300 ovens, and produces
approximately 60 MMCF of COG per day. The third facility is
considered large in the industry because it has 10 coke oven
batteries with more than 800 ovens that produce approximately 200
MMCF of COG per day. All of the facilities have desulfurization
plants for the recovery of H2S for processing and future sale.
Although there is only one recorded violation of the NAAQS
in the past 4 years, the regulatory authority is aware of the
potential to exceed 0.60 ppm SO2 over a 5-minute period around
these sources, especially if the desulfurization plant has a
malfunction or is shut-down. For example, one facility reported
that malfunctions caused the desulfurization plant to shut-down
for 251 days in 1 year3, which increases S02 emissions by 10
times during periods of plant operation. In an evaluation of the
5-minute problem, the regulatory authority collected monitored
data for the years of 1993 and 1994, which is summarized in Table
A-6.1 below. Overall, the concentrations varied in severity from
0.60 to 1.0 ppm (with a majority of exceedances occurring around
Note that data does not exist to indicate the number of
days of operation of the facility. It is possible that
the source shut-down all operations including the
desulfurization plant for long periods of time during
the year. "*
A-54
-------
Table A-6.1: Summary of Monitored
5-Minute Data: 1993-1994
Monitor
_1
2
3
4
5
Totals
No. of
5-min.
blocks
with
exceedanc
es*
27
29
7
3
2
68
Number of
hours
with
exceedanc
es
13
8
3
3
2
29
Number of
hours w/o
desulfuriz
ation
6
3
3
2
0
14
Max 5-
min.
Value
(ppm)
0.84
1.00
0.87
0.77
0.80
Time of
max value
reading
4-5 a.m.
4-5 a.m.
4-5 a.m.
llpm- 12am
12am- lam
*Note: Although violation of the IL program is measured as an exceedance
occurring in any 5-minute block of an hour and only 1 hour can be recorded as
a violation, this column provides information on the number of multiple
exceedances of 0.60 ppm that could occur in an hour.
0.80 ppm). As the table demonstrates, over a 2-year period there
were 68 measurements that exceeded the 0.60 ppm concern level.
These measurements occurred in a total of 29 hours. Fourteen of
the hours that recorded exceedances did not have desulfurization
plants in operation at the facilities which would impact the
monitor.
With this information and the knowledge that a significant
portion of the 1.3 million people living in the area could be
affected by the 5-minute exceedances, the regulatory authority
decided examine the data more closely to determine if the level
of risk to the population warranted action under the IL program.
As part of the investigation, the regulatory authority discovered
that approximately 55 percent of the hours with exceedances (16
out of 29) were between the times of 11:00 p.m. and 5:00 a.m.,
while nearly 80 percent of the exceedance hours (23 out of 29)
occurred between 9:00 p.m. and 6:00 a.m. In addition, the
regulatory authority was not aware of any citizen complaints
pertaining to a short-term exposure of S02. As a result, the
regulatory authority concluded that the public health risk from
exposures of concern, that an asthmatic individual would
encounter peak S02 concentrations while outside and engaged in
moderate exercise, is very low due to time of day when these
peaks occur. This assessment was reinforced by the absence of
complaints about pollution-related breathing difficulties. Based
on this assessment of very low public health risk, the regulatory
authority concluded that action under the IL program was noT
A-55
-------
warranted. However, the monitored data would continue to be
evaluated quarterly to ensure that public health risk did not
increase significantly. Copies of the evaluation were provided
to the three sources, as well as community environmental groups.
While there would not be any control costs associated with
this case study because no remedial action was taken, there would
be a minimal cost to the regulatory authority associated with
conducting the risk analysis and the quarterly review of the
monitoring data. The EPA estimates these costs to range between
a tenth of a man year to a fourth of a man year. In addition, it
is assumed that the monitoring costs are negligible since the
monitoring sites and operations where established to support
redesignation of the area to attainment for the NAAQS.
A-56
-------
A.7 Case Study 7; One Source Impacting a Rural Community!
This case study evaluates the impact of a single source on a
rural community with a total population of less than 1,000. The
source is a coal-fired utility power plant that is moderate in
size (500 GW) which is located on flat terrain. The State
contacted the source regarding the potential for the regulatory
authority to begin a risk assessment of 5-minute peaks of S02.
In response to the notice, the source provided monitor data to
the State that indicated the existence of 5-minute peaks that
exceed 0.60 ppm. Over a 1-year period the source's monitor that
was located 3.8 km from the plant measured a total of 10 short-
term peaks, with the majority of the peaks at concentrations near
0.60, but a few of the exceedances reached 0.80 ppm. Additional
information provided by the source indicated that the duration of
the peaks lasted less than 5-minutes (i.e., 1 to 2 minutes)
because of the flat terrain of the area and the quick dispersion
of emissions.
Based on this information and the history of the source
being in compliance with the S02 NAAQS, the State decided that
the risk to public health in the area was low and that no further
investigation was necessary.
A-57
-------
REFERENCES
1. Primary Copper Smelters National Emissions Standard for
Hazardous Air Pollutants. Emissions Standards Division,
OAQPS/USEPA, Research Triangle Park, July 1995.
2. A. Buonicore and W. Davis, Air Pollution Engineering Manual.
Air and Waste Management Association, Van Nostrand Reinhold,
New York 1992.
3. Summary of 1988-1995 Ambient 5-Minute SO2 Concentration
Data. ICF Kaiser, Systems Application International,
Research Triangle Park, September 1995.
4. OAQPS Control Cost Manual; U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards;
Document no. EPA 450/3-90-006. January 1990 and April 1991.
5. Memorandum: Supplemental Section 303 Cost Analysis for
Regulatory Impact Analysis for the Proposed Regulatory
Options to Address Short-term Peak Sulfur Dioxide Exposures.
Prepared by William Vatavuk et al. for Ronald Evans,
February 27, 1995.
6. Compilation of Air Pollutant Emission Factors, Volume I:
Stationary Point and Area Sources, AP-42, U.S. Environmental
Protection Agency 1985.
7. Technical Support Document for the Regulatory Impact
Analysis of the Enhanced Monitoring rule. Mathtech Inc.,
Princeton, N.J., September 1993.
8. Benefits and Costs of Enhanced Monitoring (40 CFR parts 51,
52, 60, 61, amended; 40 CFR part 64, added): An impact
Analysis Conducted in Response to Executive Order 12866.
Mathtech Inc., March 1995.
9. Reference 7.
10. Environmental Assessment Coke By-Product Recovery Plants,
U.S. EPA, Industrial Environmental Research Laboratory, RTF,
N.C. 27711; EPA Document No. 600/2-79-016; January 1979.
A-58
------- |