United States
Environmental Protection
Agency
Office ol Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-452/R-94-013
September 1994
Air
~Review of the National
Ambient Air Quality
Standards for Sulfur Oxides:
m, _ _ _ _ _ f - ¦ f»
Assessment ot bcientitic
and Technical Information
Supplement to the 1986
OAQPS Staff Paper Addendum
X
x
X
X
X
X
X
X
o
X
X
X
Q£ 0.75 1.0 2.0
PC (S02) (ppm)
slo 16.0
Air Quality Management Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
September 1994
-------
Acknowledgments
This staff paper is the product of the Office of Air Quality
Planning and Standards (OAQPS). The principal authors are Eric
Smith, John Haines, and Susan Lyon Stone. The authors worked in
close consultation with Dr. Lester Grant of the Environmental
Criteria and Assessment Office, and Dr. Larry Folinsbee and Dr.
Howard Kehrl of the Environmental Protection Agency's (EPA)
Health Effects Research Laboratory. The authors wish to extend
their appreciation to the many individuals whose combined efforts
helped provide much of the information contained within.
For information on health effects observed in controlled
human studies, the authors acknowledge the expert assistance of
William Linn and Deborah Shamoo of the Rancho Los Amigos Medical
Center. Dr. Jane Koenig, of the University of Washington,
provided a helpful perspective on the recent S02 literature.
The following people assisted in the development of the
chapter on air quality considerations. For air quality data
support, the authors acknowledge the assistance provided by Rick
Taylor of the Missouri Department of Natural Resources, Don
Hudnall of the West Virginia Division of Environmental
Protection, Office of Air Quality, Larry Butts of the Texas Air
Quality Control Board and Stan Skiba of the Allegheny County
Health Department, Bureau of Air Quality. George Duggan, Warren
Peters and Robert Stallings of OAQPS provided support with
computer analyses. Other technical support was provided by Chris
Knudson of EPA Region 8, as well as Lee Ann Byrd and David Lutz
of OAQPS's Technical Support Division, Monitoring and Reports
Branch. Data were also provided by representatives for several
industrial sources. Air quality modeling and exposure analysis
support was provided by Till Stoeckenius of Systems Applications
International, and John Irwin of OAQPS.
Finally, the clerical and support services provided by
Barbara Miles and Patricia R. Crabtree are acknowledged and
greatly appreciated.
-------
ii
TABLE OF CONTENTS
Page
List of Tables iii
List of Figures iii
I. introduction 1
A. Purpose 1
B. Background 2
C. Approach 8
II. Assessment of Health Effects 10
A. Sensitive Population Groups 10
B. Asthma 10
C. Medication Use 15
D. Nature and Time Course of Response 16
E. Concentration-Response Information 20
F. Other Considerations 23
G. Conclusions 29
III. Air Quality and Exposure Considerations 31
A. Occurrence of 5-Minute Peaks of 31
S02 in the Ambient Air
B. Peak-to-Mean Ratios 33
C. Nationwide Estimates of Short-Term 38
Peak S02 Levels
D. Nationwide Estimates of Exposure 43
E. Conclusions 52
IV. Staff Conclusions and Recommendations 54
Appendix A
Appendix B
References
-------
Number
ill
LIST OF TABLES
Page
2-1
2-2
3-1
3-2
3-3
3-4
3-5
B-l
B-2
Classification of Asthma by Severity of
Disease
Lung Function Changes in Response to 0.6
and 1.0 ppm S02 Compared to Typical Daily
Change and Responses to Exercise
Number of Ambient 5-Minute Averages > 0.75
and > 0.50 ppm S02, selected Sites 1989-93
Peak-to-Mean Ratios
Analysis of Hourly Averages Nationwide
Some Important Sources of Uncertainty in
Exposure Calculations
SOj Exposure Analysis Results (0.5 ppm)
Nationwide
Summary of Estimates of Expected Number of
Exposures of Exercising Asthmatics to
Elevated 5-Minute Average S02 Concentrations
for Utilities
Non-Utility Source S02 Exposure Analysis
Results
12
28
34
37
40
48
49
B-16
B-17
LIST OF FIGURES
Number
Page
2-1
Distribution of Individual Airway
Sensitivity to S02
21
-------
REVIEW OF THE NATIONAL AMBIENT AIR QUALITY STANDARDS FOR SULFUR
OXIDES: UPDATED ASSESSMENT OF SCIENTIFIC AND TECHNICAL
INFORMATION
SUPPLEMENT TO THE 1986 OAQPS STAFF PAPER ADDENDUM
I. INTRODUCTION
A. Purpose
This paper presents a summary of the evaluation and
interpretation of key new studies on the health effects
associated with short-term sulfur dioxide (S02) exposures
examined in the draft Environmental Protection Agency (EPA)
document, 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 Asthmatics (EPA-, 1994) and represents an update
of similar material in the 1986 sulfur oxides (SOJ staff paper
addendum (EPA, 1986a). Because the recently available health
effects information on S02 is related to short-term (5- to 10-
minute) exposures, this paper also updates available information
on the occurrence of short-term (5-minute) peaks of S02 in the
ambient air and on the likelihood that the at-risk population
will be exposed."
This staff paper supplement is intended to help bridge the
gap between the scientific review of recent health effects
information contained in the 1994 S02 criteria document addendum
supplement (subsequently referred to as "CD supplement" or "CDS,"
EPA, 1994) and the judgments required of the Administrator in
determining whether new regulatory initiatives are needed to
-------
2
provide increased protection to asthmatic individuals whose
health could be compromised if exposed to high 5- to io-minute
peak S02 levels. Factors relevant to this evaluation, as well as
staff conclusions and recommendations on alternative regulatory
cipp^ QaCnSS ulTc 6S6il w>6Gi XTl uUXS pu,p6IT *
B. Background
1. Legislative Requirements
Two sections of the Act govern the establishment and
revision of national ambient air quality standards (NAAQS).
Section 108 (42 U.S.C. 7408) directs the Administrator to
identify pollutants which "may reasonably be ^anticipated to *
endanger public health and welfare" and to issue air quality
criteria for them. These air quality criteria are to "accurately
reflect the latest scientific knowledge useful in indicating the
kind and extent of all identifiable effects on public health or
welfare which may be expected from the presence of [a] pollutant
in the ambient air . . ."
Section 109 (42 U.S.C. 7409) directs the Administrator to
propose and promulgate "primary" and "secondary" NAAQS for
pollutants identified under section 108. Section 109(b)(1)
defines a primary standard as one "the attainment and maintenance
of which, in the judgment of the Administrator, based on the
criteria and allowing an adequate margin of safety, [is]
-------
3
requisite to protect the public health."1 A secondary standard,
as defined in section 109(b)(2), must "specify a level of air
quality the attainment and maintenance of which, in the judgment
of the Administrator, based on [the] criteria, is requisite to
protect the public welfare from any known or anticipated adverse
effects associated with the presence of [the] pollutant in the
ambient air." Welfare effects as defined in section 302(h) [42
U.S.C. 7602(h)] include, but are not limited to, "effects on
c«/*s -J 1 e r.ra nyoncs i Jr4a>rT£»+" i aw ma nwa ma4"ov*i ale an i wa 1 e
aUll S / WdLer, LlOpo / VCyctd tluri/ ludillilauc lild uci ldxb ^ ailHUdXD /
wildlife, weather, visibility and climate, damage to and
deterioration of property, and hazards to transportation, as well
sc pets on 6foncmi c* vs Tu@cs and on D6i"son31 cnrnfOT*t" 3nr3 ijp 11 —¦
Q w w V JL C w w i9 W J i C wVJ I Will JL W v U X UC S3 U 1 Iv4 w i 1 ^CXi St w i 1Q JL wvill 1. WJL w U 1 4 W4 W C J. JL
being."
The U.S. Court of Appeals for the District of Columbia
Circuit has held that the requirement for an adequate margin of
safety for primary standards was intended to address
uncertainties associated with inconclusive scientific and
technical information available at the time of standard setting.
It was also intended to provide a reasonable degree of protection
against hazards that research has not yet identified. Lead
Industries Association v. EPA. 647 F.2d 1130, 1154 (D.C. Cir.
'The legislative history of section 109 indicates that a
primary standard is to be set at "the maximum permissible ambient
air level . . . which will protect the health of any [sensitive]
group of the population," and that for this purpose "reference
should be made to a representative sample of persons comprising
M mi ai m*
the sensitive group rather than to a single person in such a
group." S. Rep. No. 91-1196, 91st Cong., 2d Sess. 10 (1970).
The legislative history specifically identifies bronchial
asthmatics as a sensitive group to be protected. Id.
-------
4
1980), cert, denied, 101 S. Ct. 621 (1980); American Petroleum
Institute v. Costle, 665 P.2d 1176, 1177 (D.C. Cir. 1981), cert,
denied, 102 S. Ct. 1737 (1982). Both kinds of uncertainties are
nnmnnnanf c n4-Vi a v" *i +• 1 aifq 1 e Ka 1 nr.r
vUluPUJlcilUo Ui vile I JLSsiv dosUvlciLeU WlUil puXXULlUil dL IcVcXa* wciUW
those at which human health effects can be said to occur with
reasonable scientific certainty. Thus, by selecting primary
standards that provide an adequate margin of safety, the
Administrator is seeking not only to prevent pollution levels
that have been demonstrated to be harmful but also to prevent
lower pollutant levels that she finds may pose an unacceptable
risk of harm, even if the risk is not precisely identified as to
nature or degree.
In selecting a margin of safety, the EPA considers such
factors as the nature and severity of the health effects
involved, the size of the sensitive population(s) at risk, and
the kind and degree of the uncertainties that must be addressed.
Given that the "margin of safety1' requirement, by definition,
only comes into play where no conclusive showing of adverse
effects exists, such factors, which involve unknown or only
partially quantified risks, have their inherent limits as guides
to action. The selection of any numerical value to provide an
adequate margin of safety is a policy choice left specifically to
the Administrator's judgment. Lead Industries Association v.
EPA, supra. 647 F.2d at 1161-62.
Section 109(d)(1) of the Act requires that "not later than
December 31, 1980, and at 5-year intervals thereafter, the
-------
5
Administrator shall complete a thorough review of the criteria
published under section 108 and the national ambient air quality
standards , . . and shall make such revisions in such criteria
and standards ... as may be appropriate . . . ." Section
109(d)(2)(A) and (B) require that a scientific review committee
be appointed and provide that the committee "shall complete a
review of the criteria . . . and the national primary and
secondary ambient air quality standards . . . and shall recommend
to the Administrator any . . . revisions of existing criteria and
standards as may be appropriate . . . ."
2. Existing Sulfur Oxides Standards and Review to Date
The current primary standards for SO,, established in 1971,
are 80 micrograms per cubic meter (jLtg/ra3) [0.03 parts per million
(ppm) ] annual arithmetic mean, and 365 jixg/m3 (0.14 ppm) , maximum
24-hour concentration not to be exceeded more than once per year.
The current secondary standard for SO, (to protect public
welfare) is 1, 300 fxq/m3 (0.5 ppm) , maximum 3-hour concentration,
not to be exceeded more than once per year. For both primary and
secondary standards, SOx are measured as S02. Thus, S02 is the
current indicator for the SOx standards.
Review of the original S02 criteria and standards was
initiated in 1978, The Clean Air Scientific Advisory Committee
(CASAC) closed on the revised criteria document (which also
addressed particulate matter) in January 1982. An addendum to
the CD, which summarized recent controlled human studies on the
health effects of S02, was issued the same year. A staff paper,
-------
6
which identified critical issues and summarized the staff's
interpretation of key studies, received verbal closure at a CASAC
meeting xn August 1032 and formal written closure in August 1983«
In 1986, in response to the publication in the scientific
literature of a number of new studies on health effects of
particulate matter and S02, a second addendum to the criteria
document and a corresponding addendum to the SO, staff paper were
prepared. The CASAC sent the Administrator closure letters on
the criteria document addendum, dated December 15, 198 6, and on
the staff paper addendum, dated February 19, 1987, In the
closure letter on the staff paper addendum,"the majority of the
CASAC recommended consideration of a 1-hour standard in the range
of 0.2 to 0.5 ppm S02 to protect against 5-minute peaks of 0.4 to
1.0 ppm S02. The closure letter on the staff paper addendum is
reprinted in Appendix A.
On April 26, 1988 (53 FR 14926}, the EPA announced its
9 cu U c! w JL o JL O n n Q w uQ XT6 V X S 6 wil 6 clX JL w L Jl lly p XT X fuel XTy ct nO
secondary SOx standards (measured as S02) . In reaching the
provisional conclusion that the current standards provide
adequate protection against the health and welfare effects
associated with S02, the EPA was particularly mindful of
uncertainties in the available evidence concerning the possible
need for a new l-hour standard to protect against health effects
associated with 5- to 10-minute S02 exposures. Therefore, the
EPA specifically requested broad public comment on the
alternative of adding a new 1-hour primary standard of 0.4 ppm
-------
7
and making related changes to the existing standards. The IPA's
consideration of short-term health effects of S02 as well as its
rationale for other proposed changes are set forth in the
April 26, 1988 notice.
The EPA took final action on the secondary standard portion
of the 1988 proposal on April 15, 1993. The rationale for the
decision is presented in detail m the April 21, 1993 Federa1
Register notice that announced the decision (58 FR 21351).
With respect to the primary standards portion of the 1988
ea 1 4*Vi a PDfi hae -i a nri'nc!ai,,i+" /-3 q cs 4"* Vi a 4* vartu i r-oc
prupubal ^ LIlc Li n IldS? cJluCrcU in Lu a wDilocn w ucwlcc Llldu Zrc(jU XI65
by November l, 1994, either: l) final "action on the 1988
proposed decision not to revise the primary standards; or 2)
reproposal. The EPA is to take final action on a reproposal
1 year after completion of the public comment period.
The principal question to be resolved with respect to the
primary standards is whether a new short-term standard is needed
to protect asthmatics at elevated ventilation levels from 5- to
10-minute peak S02 levels. During the comment period on the 1988
proposal, a number of issues were raised concerning the possible
need for such a standard. These included: l) the health
significance of the responses reported in controlled human
studies to 5- to 10-minute S02 exposures, particularly at levels
below 0.75 ppm; 2) the possibility that moderate to severe
asthmatics may experience greater responses than the primarily
mild asthmatics studied to date; 3) whether asthmatics already
medicated to protect against other environmental stimuli would
-------
8
also be protected against S02 exposures; 4) whether a 1-hour
standard based on a typical peak-to-mean ratio of 2 to l will
provide appropriate protection from the full range of sources
that have the potential to emit high peak S02 levels2,* and 5) the
adequacy of the exposure analysis, which focused only on
asthmatics living near power plants.
In order to be better able to address these and other
issues, the EPA concluded that the 1986 addendum to the criteria
document and the associated S02 staff paper addendum should be
.updated to take into account more recent information.
C- Approach
The approach in this paper is to draw from the criteria
document supplement's (EPA, 1994) evaluation and interpretation
of the newly available health effects information on short-term
S02 exposures and to integrate that information with the
available information on the occurrence of 5- to 10-minute peak
CO 1 1 e i t"» q a i *¦ a n/4 aefl rt/-* i i vn /¦*
XC Vela XII wile: d alJJlci 11* dXJL uiiu asauuiaucu cs wliuclwcp w i.
potential exposures. Particular attention is drawn to judgments
related to determining an appropriate regulatory response given
the nature of the reported effects and the likelihood of exposure
to short-term peak S02 levels. Previous staff conclusions
2For present purposes, the peak-to-mean ratio of interest is
the ratio of the maximum 5-minute concentration for an hour
divided by the hourly average (thus a peak-to-mean ratio of 2 to
1 indicates for that hour the maximum 5-minute average was twice
the concentration of the hourly average).
-------
related to the existing primary standards or the secondary
standard will not be addressed here.
Section II provides a concise summary of key findings
presented in the criteria document supplement on health
significance of the effects of brief, concentrated exposures to
S02 on asthmatics at elevated ventilation. Emphasis is placed on
r A <2 £| i 3 rt'KC T™ n D 7* O n 1 I HO /"'•rtM e 1 H O 1^ Ckfi 1 T"\ 3 C G D C C 1 D f*t T* r"l 0 fS 11 V*! I 1
Ih** ~ Vi# IMP ^ CL JLk 1 J> Gl %¦» mSh JL VJ Imr» Jm* XS* Vb* JL i *mI JL> JL* «1» J> £¦) lESS iSi «m9 JL a 4 Vm !»¦* JLJ JL, «JL» V*
health significance of the reported effects. Section III focuses
on the available air quality and exposure information to support
discussions on the possible need for new regulatory initiatives
to address short-term peak levels of S02. Drawing from the
discussion in Sections II and III, Section IV identifies
alternative regulatory options and those factors EPA staff
believe should be considered in selecting among the alternatives.
-------
10
II. ASSESSMENT OF HEALTH EFFECTS
A. Sensitive Population Groups
Based on the assessment in the criteria document supplement,
the staff concludes that mild and moderate asthmatic children,
adolescents, and adults that are physically active outdoors
represent the population segments at most risk for acute S02
induced respiratory affects. Individuals with more severe
asthmatic conditions have poor exercise tolerance and, therefore,
are less likely to engage in sufficiently intense outdoor
activity to achieve the requisite breathing rates for notable
S02-induced respiratory effects to occur (EPA, 1994, p. 48).
Healthy nonasthmatic individuals are essentially unaffected
by acute exposures to S02 at concentrations below 2 ppm. It has
been suggested that nonasthmatic atopic3 individuals may be at
increased risk (EPA, 1986a, pg. 59; 53 FR 14932, April 26, 1988).
However, questions have been raised concerning whether the
subjects referred to as atopics in one set of studies (e.g.,
Koenig et al., 1987; Koenig et al.f 1988a,b) might be more
appropriately considered very mild asthmatics. Another recent
study (Linn et al., 1987), that compared the response of atopics
and mild asthmatics, found that the atopic group was not
3 "Atopic" is a term used to indicate individuals, not
diagnosed as asthmatics, with disorders manifested as
hypersensitivity to environmental antigens. Examples include hay
fever and other allergies. Approximately 8 percent of the U.S.
population is estimated to be atopic. Some additional percentage
of the population not diagnosed as atopic or asthmatic may also
display hyperreactive airway responses to S02.
-------
11
particularly responsive to S02. The difference in the incidence
of bronchoconstriction in atopics between the different studies
is most likely due to criteria used for diagnostic
classification, rather than real population differences. As
noted in the CDS (EPA, 1994, p. 52), there may be a significant
number of undiagnosed asthmatics and a number of subjects without
asthma who have exercise-induced bronchospasm. Xn the process of
estimating the number of individuals who are likely to be
affected by environmental S02 exposure, this uncertainty
regarding the incidence of S02 sensitivity in the population
should be considered.
B. Asthma
In assessing the significance of the S02-induced respiratory
effects in asthmatic individuals, it is important to have an
understanding of asthma as a disease in order to place the
findings from the controlled human exposure studies in
perspective. The Expert Panel Report from the National Asthma
Education Program of the National Heart, Lung and Blood Institute
(NIH, 1991) has recently defined asthma as:
Asthma is a lung disease with the following
characteristics: 1) airway obstruction that is
reversible (but not completely so in some patients)
either spontaneously or with treatment, 2) airway
inflammation, and 3) increased airway responsiveness to
a variety of stimuli.
-------
12
As indicated in Table 2-1, there is a broad range of severity of
asthma ranging from mild to severe.
Drawing from the discussion in the criteria document
supplement, the key information about the disease is presented
below:
1) About 10 million people or 4 percent of the population
of the United States are estimated to have asthma (NIH,
1991). The true prevalence may be somewhat higher.
Some researchers have estimated that 7 to 10 percent of
the United States population may be asthmatic (Evans et
al., 1987), because some individuals with mild asthma*
may be unaware that they have the disease and thus go
unreported. The prevalence is higher among African-
Americans, older (8- to 11- year-old) children, and
urban residents (Schwartz et al., 1990).
2) Common symptoms include cough, wheezing, shortness of
breath, chest tightness, and sputum production.
3) Asthma is characterized by an exaggerated
bronchoconstrictor response to many physical challenges
(e.g., cold or dry air, exercise) and chemical and
pharmacologic agents (e.g., histamine or methacholine).
4) Daily variability in lung function measurements is a
typical feature of asthma, with the poorest function
(i.e., lowest forced expiratory volume in 1 second
(FEV[) and highest specific airway resistance (SRaw)
being experienced in the early morning hours and the
-------
13
TABLE 2-1. CLASSIFICATION OF ASTHMA BY SEVERITY OF DISEASE"
Characteristics
Mild
Moderate
Severe
A. PfWreatment
Frequency of
exacerbations
Exacerbations of cough and
wheezing no more often than
1-2 limes/week.
Frequency of
symptoms
Degree of exercise
tolerance
Frequency of
nocturnal asthma
School or work
attendance
Pulmonary function
• Peak Expiratory
Flow Rate (PEFR)
• Spirometry
1 Melhocholine
sensitivity
Few clinical signs or
symptoms of asthma between
exacerbations.
Good exercise tolerance but
may not tolerate vigorous
exercise, especially prolonged
running.
Symptoms of nocturnal
asthma occur no more often
than 1-2 limes/month.
Good school or work
attendance.
PEFR >80® predicted.
Variability* <20%.
Minimal or no evidence of
¦irway obstruction on
spirometry. Normal
expiratory (low volume
curve; lung volumes not
increased. Usually a >15%
response io acute aerosol
bronchodilator administration,
even though baseline near
normal,
Melhocholine PQ,,
> 20 mg/mL.c
Exacerbation of cough and
wheezing on a more frequent basis
than 1 *2 times/week. Could have
history of severe exacerbations, but
infrequent. Urgent care treatment
in hospital emergency department
or doctor's office <3 times/year.
Cough and low grade wheeling
between acute exacerbations often
present.
Exercise tolerance diminished.
Virtually daily wheezing. Exacerbations
frequent, often severe. Tendency to have
sudden severe exacerbations. Urgent visits to
hospital emergency departments or doctor's
office >3 times/year. Hospitalization
>2 limes/year, perhaps with respiratory
insufficiency or, rarely, respiratory failure and
history of intubation. May have had cough
syncope or hypoxic seizures.
Continuous albeit low-grade cough and
wheezing almost always present-
Very poor exercise tolerance with marked
limitation of activity.
Symptoms of nocturnal asthma
present 2-3 times/week.
School or work attendance may be
affected.
PEFR 60-80% predicted.
Variability 20-30®.
Signs of airway obstruction on
spirometry are evident. Flow
volume curve shows reduced
expiratory flow al low lung
volumes. Lung volumes often
increased. Usually a >15%
response to acute aerosol
bronchodi lator admi nisi ration,
20 mg/niL.
Considerable, almost nightly sleep Interruption
due to asthma. Chest tight in early morning.
Poor school or work attendance.
PEFR <60® predicted.
Variability >30%.
Substantial degree of airway obstruction on
spirometry. Fiow volume curve shows marked
concavity. Spirometry may not be normalized
even with high dose steroids. May have
substantial increase in lung volumes and marked
unevennessof ventilation. Incomplete
reversibility lo acute aerosol bronchodilator
administration.
Mcthacholine PCM <2 mg'mL.
B. After optimal treatment is established
Response to and Exacerbations respond to Periodic use of bronchodilators Requires continuous, multiple around-the-clock
duration of therapy broncodilatora without the use required during exacerbations for drug therapy including daily corticosteroids,
of systemic corticosteroids in a week or more. Systemic steroids either aerosol or systemic, ollen in high doses.
12-24 h. Regular drug usually required for exacerbations
therapy not usually required as well. Continuous around-the-
except for short periods of clock drug therapy required,
time. Regular use of anti-inflammatory
agents may be required for
prolonged periods of time,
'Characteristics are general; because asthma is highly variable, these characteristics may overlap. Furthermore, an individual may switch
jnco different categories over time.
Variability means the difference either between a morning and evening measure or among morning peak flow measurements each day for a
week.
Although the degree of methacholine/histamine sensitivity generally correlates with severity of symptoms and medication requirements,
there are exceptions.
Source: National Institutes of Health (1991),
-------
14
best function (i.e., highest FEV, and lowest SRaw) occurring in
thfi mid-afternoon*
5) The degree of exercise tolerance varies with the
severity of disease. Mild asthmatic individuals have
good exercise tolerance but may not tolerate vigorous
ovovrii aa cn/^K sic nvrtl nn/taH vimrn nrt a 4"Vrma4~ i r»
cacJuCxsc b utn do pjr u JLon^cu xruiirixri^* nou sir a a s wrunci t>xc
individuals have diminished exercise tolerance and
individuals with severe disease have very poor exercise
tolerance that markedly limits physical activity.
6) Exercise-induced bronchoconstriction is followed by a
refractory period of several hours during which an
asthmatic individual is less susceptible to
bronchoconstriction (Edmunds et al., 1978), This
refractory period may alter an asthmatic individual's
responsiveness to S02 or other inhaled substances.
7} Asthma attacks can result in hospitalization or
emergency room treatment. It is estimated that
incidence of hospitalization for all asthmatic
individuals in the United States is about 45 per 1,000
asthmatics per year (NIH, 1991). Attendance at
emergency rooms for asthma in Vancouver, Canada was
uiiuducu to av(#ounu 1 ojl jl * £* percenx: ul snicir^Brrcy
room visits.
8) Data on asthma attack rates in the United Kingdom
suggest an incidence of asthma attacks requiring
medical attention, of <1 asthmatic patient-year (Ayres,
-------
1986; Nevill et al., 1993). A similar attack incidence
was estimated for the United States patients (Lebowitz
et al, 1985; Van Issen-Zandoliet et al., 1992).
9) In assessing the rate of incidence, it should be noted
that based on the Los Angeles asthma panel data (EPRI,
1988), only 15 percent of mild asthmatic individuals
see a physician annually for their asthma compared to
about 67 percent of the moderate asthmatics.
10) Death due to asthma is a rare event; about one per
10,000 asthmatic individuals. Mortality rates are
higher among males and aboat 100 percent higher among
non-whites. It has been reported that in two large
urban centers (New York and Chicago) mortality rates
from asthma among non-whites exceed the city average by
up to five-fold and exceed the national average by an
even larger factor (Sly, 1988; Evans et al., 1987; NIH,
1991; Weiss and Wagener, 1990; Carr et al., 1992).
There may be several possible explanations for this,
but the cause of these higher mortality rates has not
been explained.
In assessing the results from the controlled human exposure
studies discussed below, it should be noted that the individuals
who participate in such studies may not be representative of the
entire population of individuals with asthma. The subjects of
controlled exposure studies typically have mild allergic asthma.
In many cases, these individuals can go without medication
-------
16
altogether or can discontinue medication for brief periods of
time if exposures are conducted outside their normal allergy
season. In addition, African-American and Hispanic adolescents
and young adults have not been studied systematically. Subjects
who participate in controlled exposure studies are also generally
self-selected and this may introduce some bias. Thus, the extent
to which the participants in the studies reflect the
characteristics of the asthmatic population at large is not
known. Nevertheless, the high degree of consistency among
studies suggests either that the subjects are generally
%
representative of the population at risk or that any selection
bias is consistently present across a diverse group of
laboratories.
C. Medication Use
Many asthmatic individuals 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 responses to S02. This has led to
suggestions that asthmatic individuals may be protected from
v>AC riAFl eae 4* C A Kq a H cq f'Koif l r»a4"ia *¦» v» "i /"% v» +• a a i c>a
4 UO iDwj ucLdUSc wficy lUculva wc prlQi LQ SXcTClSc«
However, as discussed in the CD supplement (EPA, 1994), the
available data suggest that probably a substantial proportion of
asthmatic individuals would not be "protected" by medication use.
Most mild asthmatic individuals use medication only when symptoms
arise. Roth Associates (1988) reported that out of a panel of 52
asthmatic subjects, whose exercise patterns showed a wide range
-------
17
of variability, one third of the mild asthmatic subjects studied
had not used any asthma medication within the past year, and that
fewer than half used an inhaled bronchodilator at least once
during the past year. Only 20 percent of the moderate asthmatics
subjects studied use an inhaled bronchodilator on a regular
basis. Marks et al., (1992) also reported that beta-agonist use
was infrequent.
Even medication compliance for those on regular medication
varies considerably among asthmatic individuals (from none to
full compliance). Average compliance figures range from 50 to 70
percent (Smith et al., Weinstein and Cuskey, 1985; Smith et al.,
1986; Partridge, 1992). Given the relatively low medication use
ari/1 fnwn 1 l vaf"oe f" /msym *m anu Tn -i T /"-i a n/4 w aTa4"o a c^hwa"!" i ne
uflu JL xdI*wvi XT
-------
18
at increased risk from S02 exposure because of their potentially
poorer baseline level of lung function. Exposure of unraedicated
moderate asthmatics to S02 could cause additional deterioration
of lung function that could be cause for medical concern (EPA,
1994, p. 51).
D. Nature and Time Course of Response
The most stt*i k i no acute resnonse to SO« for asthmatics and
•L * lllw»3 v WJL J- A Jm 4 4VjJ U V J- C i3J^ V i WW WVJ JLW4L UU UlUllU V. 1 W£3> U i 1 V4
others with hyperactive airways is bronchoconstriction (airway
narrowing), usually evidenced as increased airway resistance,
decreased FEV\, or decreased peak flow, and the occurrence of
symptoms such as wheezing, chest tightness, and shortness of
breath (EPA, 1982a; EPA 1986a). This bronchoconstriction
response occurs quickly (within 5- to 10-minutes of exposure),
with two recent studies showing that the response can begin in as
little as 2-3 minutes, although the response does not reach
maximal levels until the exposure lasts five or more minutes
(Balmes et al., 1987; Horstman et al., 1988). The response is
also generally brief in duration; numerous studies have shown
that lung function typically returns to normal for most subjects
within an hour of exposure. This duration is similar to that
experienced in response to exercise and somewhat less than
experienced in response to allergens (EPA, 1994). Even if
exposure continues beyond the initial 5-10 minutes, lung function
may still return to normal as long as the subject ceases to
exercise and their ventilation rate decreases to resting levels
(Hackney, et al., 1984; Schatcher et al., 1984).
-------
19
A mild "refractory period" seems to exist in which
diminished responsiveness is seen when an individual is re-
exposed to S02 while at exercise. Lung function responses of
approximately 75 percent of those observed after an initial
exposure to S02 are observed after a second exposure ten to
fifteen minutes later (Roger et al., 1985; Kehrl et al., 1987).
The response diminishes further with subsequent exposures.
However, a few individuals may experience a worsening of response
upon re-exposure (Roger et al., 1985}. The duration of this
refractory period is uncertain, although it does not appear to
last longer than 5 hours on average (Linn et al., 1984).
Furthermore, longer periods of exposure while at exercise (i.e.,
30 minutes) do not lead to a statistically significant worsening
of the initial response (Kehrl et al., 1987, p. 352).
An important distinction between the response of asthmatic
individuals to S02 as compared to their response to allergens is
that no evidence indicates that the S02 response is accompanied
by any "late response," such as that often seen 4 to 8 hours
after allergen exposure.
The effects of S02 increase with both increased overall
ventilation rates and an increased proportion of oral ventilation
in relation to total ventilation (EPA, "1986a, p. 10) . Oral
ventilation is thought to accentuate the response because the
scrubbing of S02 by the nasal passageways is bypassed. For this
reason, in most clinical studies which have observed effects from
S02, the subjects have been exercising at ventilation rates of 35
-------
20
to 50 L/min, which equal or exceed the "switching point" (35.3
L/min) from exclusively nasal breathing to oronasal breathing
found on average for the general population by Niinimaa et al.
(1980).
Ventilation rates in the range of 35-40 L/min are comparable
to ventilation rates induced by climbing 3 flights of stairs,
light cycling, shoveling snow, light jogging, or playing tennis
(Cohen, 1983), and can be induced in the laboratory by walking at
3.5 mph up a 4 percent grade (Kehrl et al., 1987; Folinsbee,
personal communication). Ventilation rates in the range of 45-50
L/min are equivalent to moderate cycling, chopping wood, or light
uphill running, and can be induced by walking at 3.5 mph up an 8
percent grade (Folinsbee, personal communication). Even though
such exercise is not strenuous per se (in that it does not
approach an individual1s maximum oxygen consumption or the
ventilation rates of moderate jogging, heavy cycling, playing
basketball, or running), activity and ventilation data indicate
that individuals engage in outdoor activities at these
ventilation rates only a small percentage of the time (see
Section III.D.1).
Since oronasal scrubbing of S02 is important in mitigating
the effects of S02 (EPA, 1986b,' p. 4-26), asthmatic individuals
who are obligate mouthbreathers, or who are breathing through the
mouth due to some temporary condition, may be at greater risk of
experiencing responses to S02 (since their nasal scrubbing may be
bypassed at lower ventilation rates and to a greater extent than
-------
21
for those individuals capable of typical nasal breathing).
Several studies have estimated mouthbreathers to constitute
approximately 15 percent of the general population (Saibene et
al. , 1978; Niinima et al., 1980; EPA, 1986b, p. 4-26).
Bronchoconstriction effects may also be exacerbated by cold,
dry air and diminished under warm, humid conditions (EPA, 1986b,
pp• 4-35 to 4—37). As discussed ln the criterxa document
addendum (EPA, 1986b), Bethel et al. (1984) reported a
significant interaction between oral hyperventilation of cold dry
airland 0.5 ppm S07 via mouthpiece that resulted in a >200
percent increase in SRaw, whereas breathing S02 in warm humid air
or breathing cold dry air alone resulted in a <40 percent change
in SRaw. It has been well documented in numerous studies that
S02 may interact with weather factors (e.g., cold/dry air) and/or
exercise to cause exaggerated bronchoconstriction. This suggests
that airway cooling and drying may exacerbate S02-induced airway
constriction in hyperventilating asthmatic subjects, but
insufficient data exist by which to estimate the magnitude of any
combined effects of joint S02 and cold, dry air exposure under
more natural free-breathing conditions during exercise (EPA,
1994, p. 31).
Many features of the S02-induced bronchoconstriction
response resemble those of exercise-induced bronchoconstriction,
including the duration of the effect and the absence of a
substantial late response. However, it should be noted that
above a sufficient concentration, the response to S02 clearly
-------
22
4*ha vaenrinea si HK"H i** i Kii+'tiKI a +¦ a avav/^i *r« 1 ea
UL SUiJJ cCuS Cafl cApci lcllvc all CLicvw irolu Dvj Wucn a L cXciCliSe
while experiencing little or no effect from exercise in clean air
(Linn et al. , 1987).
E. Concentration-Response Information
The CD Supplement extensively reviewed several recent,
large-scale chamber studies with the aim of further investigating
the concentration where clinically significant responses began.
Because of the well-documented range in sensitivity to S02 among
asthmatic persons (e.g., Figure 2-1), variability in an asthmatic
individual's day-to-day responsiveness, and the nature of the
response itself, it was judged that neither simple group mean
statistics nor the responses of particularly sensitive
individuals were an appropriate focus. Rather, attention should
be focused on the concentrations where a significant proportion
of asthmatic individuals tested began to experience effects of
concern. Assessing effects of concern involved comparing the
responses experienced to S02 with those typically experienced in
response to typical daily variation in lung function, and to
other frequently experienced stimuli, such as exercise or
cold/dry air, and noting the frequency with which subjects felt
compelled to take medication or diminish workload. The CD
Supplement (EPA, 1994) summarized its evaluation of the recent
a 4* S ai a iP/%1 *1 Af.rcii •
Uw WQ QS X OX J.UWS m
a) At most, only about 10 to 20 percent of mild and
moderate asthmatic individuals exposed to 0.2 to 0.5 ppm S02
-------
100-
75
it 50
ja
3
E
3
o
25
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
HJ-
0.75 llo
Ti.o
0.25
0.5
2.0
5.0
PC(S02) (ppm)
Figure 2-1. Distribution of individual airway sensitivity to SO: (llorstman et al., 1986). PC (S02) represents concentration of S02 that, after
correction for exercise (V, - 42 l/min), resulted in a 100 percent increase in SRaw. Cumulative percentage of subjects is plotted as a function of
PC (S02) and each data point represents PC (SO,) for an individual subject. These data show substantial variability in sensitivity among mild
asthmatic volunteers.
Source: Horstman et al. (1986)
-------
2 4
during moderate exercise are likely to experience lung function
changes distinctly larger than those they typically experience.
Furthermore, only exceptionally sensitive responders might
experience sufficiently large lung function changes and/or
respiratory symptoms of such severity to be a potential health
concern, leading to the disruption of ongoing activities, the
need for bronchodilator medication, or seeking of medical
attention.
b) In contrast to the above projected likely consequences
of ambient exposures to 0.2 to 0.5 ppm S02 of mild and moderate
asthmatic persons, considerably larger lung function changes and
respiratory symptoms of notably greater severity would be
expected to occur due to exposure of such individuals to S02
concentrations of 0.6 to 1.0 ppm S02, That is, substantial
percentages (>20 to 2 5 percent) of mild or moderate asthmatic
individuals exposed to 0.6 to 1.0 ppm S02 while physically active
mai* 1 A Ka us ^ a Vi a a i<»ae ys i y-a 4" r» v»tf
-------
25
S02 effects, although it would still appear relatively unlikely
that the short-lived symptoms would be sufficient to cause many
to seek emergency medical attention.
The CD supplement (EPA, 1994) concludes that while the
relative health significance of the responses seen to S02 are
difficult to judge (see further discussion below), more concern
c ItMi % 1 j-I Via Hp a r~* i icf£a^ An ^ Vi q 4" A sA ¦ £ i""* CH n ^
SflwU 1Q yc iQQUocu OA tile rcSpunsc LO -2U • u JJjpni DUj Uildn to
concentrations of S02 <0.5 ppm (EPA, 1994, p. 46).
F. Other Considerations
In addition to information on the nature and severity of
pffpct" as ind i ratfd hv clinical DaT"aTne>te>T"<3 fcVi©T*p are spvera 1
C *» JL w W w VA fiS JLI i v* W Q wwVA W J w X X I I JL. w W JL M U JL W <5Xr w f Wil CS J. w t* X C vl V vi u JL
other factors that the Administrator may wish to consider:
1. SO-, Responsiveness and Asthma Severity
One concern voiced in the last review was whether more
severe asthmatic individuals than those studied to date might be
more responsive or experience more severe effects from SO,. At
that time, the evidence was judged insufficient to answer that
question (Appendix A).
Several of the more recent studies reviewed in the CD
supplement (Linn et al., 1987, 1990; McManus et al., 1989)
provide information on this question by reporting the responses
of asthmatic individuals with moderate to severe disease,
medication-dependent disease, or older individuals with
"intrinsic" asthma. When airway resistance was examined, the
Pi<5thmatia subiuayo tr* h^vis* «H?ni 1 ar
mXk\mr X25 JL V* V \mZ €¦!> tmm Ji V* *4L IMP Jhp J Saw Km* 1m* Bmb inr Vwp JL \m/ M-J ImmP *sS» * XSm Vm Vb' • 4 V* w 22# X JiliI X C31 X
relative changes but larger absolute changes to those observed
-------
26
for mild asthmatic individuals (Linn et al.f 1987). As the CD
supplement suggests (EFA, 1994, pp. 21-24), similar function
declines may have a greater impact on individuals with lower
baseline lung function, a situation more typical of moderate or
severe asthmatics.
In addition, a recent study suggests that older "intrinsic"
asthmatic subjects (McManus et al., 1989) may experience
bronchoconstriction, albeit from a mouthpiece exposure, even
while resting. The CD supplement concludes that while the data
is suggestive of greater responsiveness among those with more
caifav A i eoaca 4»K m aJ ¦» *-% —» 1
j.uri^ xuriv wXon towd4 q xsvsxs luq u moy Dcconis vatusc iox itisqxc^x
concern (EPA, 1994, p. 44).
The CD supplement also notes that severe asthmatics are less
likely to be sufficiently physically active, because of low
exercise tolerance, to be frequently at risk from peak
concentrations of S02. In addition, this segment of the
asthmatic population would be most likely to premedicate prior to
engaging in substantial outdoor activity.
2 . Effects of Asthma Medications on the SO., Response
Interest has been expressed concerning the ability of
typical asthma medications to protect against the effects of S02.
An argument can be made that if medications routinely used by an
-------
27
asthmatic, for reasons separate from the pollutant itself, also
confer protection against the effects of the pollutant, then this
consideration should be factored into the evaluation of risk. It
now appears that most regularly administered medications, such as
1 nViiS 1 fiiH rjc an/4 w* £2» 4~ Vi \/l vanf ^ i no i i r>we f enr«Vi » c
Xltli d X 6 U 3 U 6 Jl w X U El C¦* i is* HI C Uil V X A u 11 wilX* its ILiwU i i wl * S> I £!> 4 d 23
theophylline) appear relatively ineffective in protecting against
the S02 response (EPA, 1994, p. 34-41). In contrast, inhaled
beta-agonist bronchodilators are highly effective in reducing or
eliminating the lung function responses to S02 (EPA, 1994, p.
38). Since bronchodilators are most effective in preventing
effects if taken relatively shortly before exposure, the
frequency with which asthmatic individuals premedicate prior to
exercise is of interest.
As poxnted out xn Section C above, many asthmatics do not
use bronchodilators at all or do not use them with a frequency to
suggest that they consistently premedicate prior to exercise. In
fact, as pointed out above (Section E), many of the mild
asthmatic individuals, including those responsive to S02, have
little or no exercise-induced bronchoconstriction at the exercise
levels examined here, and thus would probably not feel a
compelling need to premedicate prior to exercise. Data on the
medication use of some of subjects in the clinical studies bear
out the conclusions that in general, mild asthmatics use
bronchodilators infrequently, as do some moderate asthmatics;
although a substantial portion of moderate asthmatic may use
bronchodilators frequently (EPA, 1994, Appendix B memo).
4
-------
28
& in a4* Viai" # a r"+" av +¦ a nnne i ¦! e 4»Ka4" ~•ViAVja 1 e enmA eni,tr,fae+" i nr»
Aliouiiexr idctor wo cunsiaer jls uuai wnexre JLa sonie suggestion
that excessive use of beta-agonist bronchodilators leads to a
worsening of asthma status (EPA, 1994, p. 41).
3. Effect of Other Air Pollutants on SO, Responsiveness
Koenig et a1« (1990) reported that response to S02 in
adolescent asthmatic subjects was potentiated by prior exposure
to 0,12 ppm ozone (03) . After 45 minutes of 03 exposure by
mouthpiece, a 15-minute mouthpiece exposure to low concentrations
of S02 (0.1 ppm) produced statistically significant decreases in
FEV, (8 percent total change, versus a 3 percent change without
prior Oj exposure). Symptoms scores did not change significantly
(although an increase in symptoms was reported for the combined
Oj and S02 exposure). Because of the reliance on mouthpiece
exposures at single concentrations for both pollutants, it is
difficult to fully evaluate the potential implications of this
experiment for ambient exposures to S02, but it gives suggestive
evidence that brief S02 exposures encountered against a
background of elevated 03 levels may lead to greater effects than
those seen in the controlled human exposure studies that examined
S02 alone.
The effects of prior N02 exposure on S02~induced
bronchoconstriction has been examined in two other studies
(Jorres and Magnussen, 1990; Rubinstein et al., 1990). One
mouthpiece study indicates that a 30-minute peak of N02 at 0.25
to 0.30 ppm increased airway responsiveness to S02 among
asthmatic individuals (probably due to a nonspecific increase in
-------
29
bronchial responsiveness), while a chamber study found no change
in responsiveness except for one subject (EPA, 1994, pp. 42-43).
4. Effects of SO? In the Context of the Typical Experience of
Asthmatic Individuals
Another factor that might be considered in assessing the
severity of S02 effects is the frequency with which the sensitive
population experiences similar effects as a result of normal
variation and reactions to other stimuli. As indicated above,
asthmatic "episodes," as indicated by self-reported asthma
attacks, self-reported symptoms (EPA, 1994, Appendix B memo), or
visits to the physician or emergency room (EPA, 1994, pp. 7-8),
seem to be a relatively infrequent occurrence for many or most
adult asthmatics. While it is uncertain how individuals would
perceive their responses from S02, those experiencing pronounced
responses to 0.6 to 1.0 ppm may perceive these events to be
asthma attacks (although it is judged relatively unlikely that
the effects would cause many to seek emergency medical attention)
(EPA, 1994, p. 50). In addition, the symptoms suffered by those
responding to S02 may attain levels of severity greater than
experienced on a typical day-to-day basis, especially among mild
asthmatics (EPA, 1994, Appendix B memo).
Table 2-2 shows that, for the indicator of lung function as
well, the effects seen in response to SO-, in the more sensitive
* * i
asthmatic individuals (especially the most sensitive 25 percent)
-------
Table 2-2. LUNG FUNCTION CHANGES IN RESPONSE TO 0.6 AND 1 PPM S02 COMPARED TO
TYPICAL DAILY CHANGE AND RESPONSES TO EXERCISE
ASTHMATIC
DAILY
PERCENTILE
MODERATE
S02
TOTAL
SEVERITY
CHANGE
OF TEST
EXERCISE
(corr. for
CHANGE
SUBJECTS
exc.)
MILD
-8%
50th
-2%
-21%
-21%
FEV,
75 th
-7%
-26%
-30%
MODERATE
-13%
50th
-8%
-10%
-25%
FEV,
75th
-14%
-31%
-39%
MILD (1985)
7
50th
+ 46%
+ 118%
+ 164%
SRaw
75th
+ 59%
+ 230%
+ 249%
I
Modified from Table 2 of CDS Appendix B memo. Table shows that the response due to S02 alone
(corrected for exercise) or the total response (considering the combined effects of S02 and exercise)
qonsiderably exceeds the change due to exercise or the typical daily change in most cases, especially for
the most sensitive 25% of responders (the 75th percentile group). The exercise and S03 numbers should
not be expected to sum to equal "Total Change" (see CDS, Appendix B memo).
-------
31
considerably exceeds the change in lung function due to exercise
or daily variability. A second comparison to exercise showed
that, when the symptom and lung function responses were examined
in combination (along with, in some cases, medication use), the
total effect of S02 combined with exercise on asthmatic
individuals clearly exceeded the effects of exercise alone. For
example, approximately 6-43 percent of asthmatic subjects
experienced what were classified as severe lung function changes
and moderate symptoms in response to S02, while no subjects did
so after exercise alone (EPA, 1994, Appendix B memo).
In summary, present data suggests that the effects
experienced by those asthmatic individuals responding to 0.6 to
1.0 ppm S02 are likely to be perceived as distinctive, notable
events outside the range of responses frequently experienced.
However, perception of symptoms is not necessarily a good index
of functional status. Some patients with near-fatal asthma
attacks had a poor perception of their breathing difficulty and
were thus unable to perceive an attack of severe bronchospasm
(EPA, 1994, p. 30).
G. Conclusions
In conclusion, the primary reasons for concern over the
effects of S02 in the range of 0.6 - 1.0 ppm are that a
substantial percentage of asthmatic individuals (>20 to 25
percent) experience pronounced changes in lung function that may
be viewed as a mild asthma attack, cause discomfort, prompt self-
administration of medication, and cause some individuals to alter
-------
32
their activity (even from a 10-minute exposure). Most adult
asthmatic individuals do not seem to experience asthmatic
episodes of similar magnitude with great frequency. Most
regularly administered medications are not very effective in
blocking the SOj response, and to obtain protection from the most
r">s*MTiTn/*M-i 1 if ncQi*i f t ifa 'misH i na ^ i rtn /Voi4>aa.a rrnn i cs 4* e ^ 4"Ha ae4"hma4* i r<
vunuuunjLy iis?evi ci icluivc in^ujLvaLiiUii |QcLa a^unioUo j f wiis ab wniud vJL\#
individual has to anticipate the need to premedicate prior to
exposure. (Although some asthmatics premedicate routinely before
exercise, such premedication is likely to be infrequently
practiced for much of the sensitive population). Lastly, some
conditions, such as prior exposure to 03, may exacerbate-the
response.
Factors that serve to mitigate, to some degree, concern over
S02 effects are that the response, like most asthma responses,
resolves over time; in most cases, the response has run its
course within an hour, with no evidence of later heightened
sensitivity such as is seen in a "late response." In addition,
while some individuals may reduce activity, most of the subjects
exposed at 0.6 to 1.0 ppm do not feel such a need and can still
function effectively despite whatever effects they perceive from
the SOj exposure. Fina1ly, medication does exist (primarily
beta-agonists) that can ameliorate the responses, either if taken
shortly before exposure or after the response has begun.
Given the above information, the staff agrees with the
recommendation of the CD supplement (EPA, 1994) that the likely
frequency of occurrence of such S02-induced effects is a factor
-------
33
to be considered in assessing the degree of public health concern
posed from exposures to peaks of S02.
-------
34
III. AIR QUALITY AND EXPOSURE CPUSTDERATIONS
Because the most recent health effects information on so2 is
related to short-term (5- to 10-minute) exposures, this section
summarizes recent information on the occurrence of monitored
high, 5- to 10-minute concentrations of S02 in the ambient air.
New information is presented on the variability of 5- to 10-
minute peak S02 concentrations within particular hourly periods,
which relates to the averaging time necessary for any effective
short-term standard. Estimates of the nationwide prevalence of
these short-term peaks of S02 are given.
A. Occurrence of 5-Minute Peaks of SO., in the Ambient Air
A central issue raised during the comment period on the 1988
proposal concerned whether the staff underestimated the
prevalence of short-term, 5- to 10-minute peaks of S02. Concern
focused on two issues: 1) whether nonutility sources, which were
qualitatively but not quantitatively considered m staff
6sxxnci wcw u l cxpusuzr6 f inxynt> con wjt x jju ws u si*ds laxt t xu x nmriw61 u x.
5-minute peaks of S02, and 2) whether a 1-hour standard of 0.4
ppm (based on a typical peak-to-mean ratio of approximately
2 to 1 derived principally from utility data) would provide
adeauate protection from hicih 5-minute peak SO, levels near
IT ** ¦ ««- "— w w im T mm mm mm -mmm • - .a. ^ * —- — mm • * —. — — jmr mm m m mm -mm £ .a. — 0.75 ppm S02 for 5 minutes or more, because
-------
35
this benchmark is approximately equal to the levels that would be
protected against by the 1-hour, 0.4 ppm standard advanced for
comment in 1988. For comparison purposes, the prevalence of
peaks >0.5 ppm was also determined if available data allowed.
Obtaining 5-minute data has proved difficult because the
shortest averaging period typically retained in monitoring data
K *a fil/* e 1 V"* /"M i v* 1VT 4-K q qv l ct* 1 'ft/"'* inr*r» 1 n>*e a e i +*ea/^ i n
ya,nj\D is x nour • iiurcQvcr ^ inc caxd n ny luwnx uuio aire o iwcu xn
locations that are designed to be representative of air quality
levels associated with 24-hour, annual, and 3-hour
concentrations, rather than to detect short-term peaks.
Despite these problems, data-gathering efforts to date
indicate that peak 5-minute levels of S02 >0.75 ppm can occur
around a number of different sources.4 While the data from these
ambient monitoring sites cannot always be attributed solely to a
single source, 5-minute concentrations of S02 in excess of 0.75
ppm have been recorded by a number of ambient air monitors sited
primarily to detect S02 emitted from distinct point sources.
These include one or more sources in the following source types:
utility boilers, industrial boilers, refineries,.pulp and paper
mills, copper smelters, primary lead smelters, sulfuric acid
plants, and steel mills (coke ovens), For those sources for
which the data were available, the number of peaks >0.50 ppm was
also calculated (Stone, 1994).
4In this paper, information on ambient 5-minute
concentrations of S02 refers to the highest of the 12 block
averages (12:00 to 12:05, 12:06 to 12:10, etc.) possible during a
clock hour.
-------
36
Data collected from monitors located near these source types
are summarized in Table 3-1. The S02 peak concentrations
enumerated in Table 3-1 were measured in the ambient air during
the years 1988 to 1993. Seven of the 12 sites listed recorded
hic[h 5-minute pea)cs xn the 1993 calendar year» These data
suggest that around some sources, numerous 5-minute peaks of S02
>0,75 ppm can occur. However, in some cases, fewer peaks have
been recorded around other sources of the same general type.
A few of the sources listed in Table 3-1 have recently
installed improved pollution control equipment which would be
expected to reduce the occurrence of S02 peaks. Thus, l:he data
in Table 3-1 are not intended to represent "typical" frequencies
of 5-minute peaks of S02 around the different source types
listed. They do illustrate that ambient peaks of S02 >0.75 ppm
can occur near a variety of sources.
Finally, it should be noted that high peaks did occur on
days when the existing 24-hour or 3-hour standards were exceeded.
In general, however, these data suggest that the current NAAQS
may offer less protection against brief, concentrated peaks of
S02 than previous staff analyses indicated.
B. Peak-to-Mean Ratios
The 1982 staff paper and the 1986 addendum summarized the
available information on the variance of 5- to 10-minute peak
concentrations within particular hourly periods. Based on its
assessment of the available data (Larsen, 1968; Burton and
Thrall, 1982; Thrall et al., 1982; Rote and Lee, 1983; Armstrong,
-------
TABLE 3-1. Number of Ambient 5-minute Averages >0.75 and >0.50 ppm S02
Selected Sites, 1989-93
Source
Approximate # of Hours With 1 or More
5-min Peaks / Period of Time
>0.75 ppm >0•50 ppm
Sulfuric Acid Plant
18/0.05 yr.
38/0.05 yr.
Petroleum Refinery/Industrial Complex2
56/0.38 yr.
114/0.38 yr.
Sulfite Paper Mill
83/1.0 yr.1
-
Allegheny County, PA2
35/0.92 yr.
-
Copper Smelter2
73/2.5 yr.
-
Primary Lead Smelter
72/1.15 yr.
125/1.15 yr.
Copper Smelter
14/1.0 yr.
51/1.0 yr.
Steel Mill
32/2.15 yr.
74/2.15 yr.
Utility/Industrial Complex
15/5.16 yr.
88/5.16 yr.
Industrial Boiler/Kraft Paper Mill
1/0.31 yr.
2/0.31 yr.
Petroleum Refinery
i
0/1.0 yr.
0/1.0 yr.
Petroleum Refinery
0/1.0 yr.
6/1.0 yr.
'Actually indicates instantaneous peak concentrations >1.0 ppm
2These sources had more than one monitor in their proximity. Data used from all monitors, but
hours with peaks only counted once, regardless of how many of the monitors recorded a peak for
that hour.
-------
38
1985, 1986) and relying on the premise that utilities would be
the dominant source of 5-minute exposures, the staff concluded
that 5-minute peak values were typically twice that of the
associated l-hour value. Thus, it was thought an hourly standard
of 0.4 ppra would protect against 5-minute peaks of approximately
0.8 ppm or higher.
The use of a 2 to 1 peak-to-mean ratio was questioned during
the public comment period on the 1988 proposal. One commenter
(Environmental Defense Fund, item IV-D-72, Docket A-84-25)
submitted data collected near three sulfite paper mills
indicating that high 5-minute peak S02 levels could occur that
were associated with very low hourly averages (i.e., peak-to-mean
ratios in excess of 2 to 1). While these data are limited to one
source type (and one of the sources had no controls on pertinent
orfn 1 Ti'mol'if" 4- H -a +- voen 1 4"o/^ i m t/q v-ir k i mif'a naaVe\ f*hQ\7
wtj y J. *1 w wii0.75 ppm. Therefore a peak-to-
-------
39
mean ratio was derived only for hours containing high 5-minute
peaks, but these are precisely the events any new standard would
be designed to guard against. Because of this restriction the
number of observations in these data sets is far less than those
examined in the 1982 staff paper and the 1986 addendum.
All of the mean and median peak-to-mean ratios for each of
these data sets are in excess of 2 to 1 (Table 3-2). The range
of hourly averages associated with 5-minute peaks >0.7 5 is very
broad, and in isolated instances peaks >0.7 5 ppm were observed
during hours in which the hourly average did not exceed 0.2 ppm.
While much of the variability in these peak-to-mean ratios
likely results from emission-rate variability, and sources with
better controlled, more uniform emissions may have fewer peaks
and fewer hours with high peak-to-mean ratios, Table 3-2 suggests
that no "typical" peak-to-mean ratio exists that can be used to
determine a uniformly-applicable hourly standard. Given the
broad range in hourly values associated with concentrated
5-minute peaks of S03, it appears that reliance on any single
hourly peak-to-mean ratio will risk over-controlling some sources
(if a high peak-to-mean ratio is assumed and a low hourly
standard chosen) or under-controlling other sources (if a low
peak-to-mean ratio is assumed and a high hourly standard chosen).
For example, among Allegheny County monitors, 84 hours had
average concentrations above 0.25 ppm, yet only 19 of these hours
(2 3 percent) had 5-minute peaks above 0.75 ppm. During the same
time period, peaks were recorded in 22 hours with average
-------
40
TABLE 3-2. Peak-to-Mean Ratios
Source
Copper Smelter
Allegheny County, Pa
Refinery/Industrial Complex
Primary Lead Smelter
Peak-to-Mean
3.5
3.6
7.5-1.2
0.17
90
4.0
3.7
10.9-1.4
0.07
39
2.9
2.4
7.3-1.1
0.11
23
4.0
3.22
10.37-1.68
0. 09
22
Average
Median
Range
MinHour (ppm)1
No. of Observations
Average
Median
Range
MinHour (ppm)
No. of Observations
Average
Median
Range
MinHour (ppm)
No. of Observations2
Average
Median
Range
MinHour (ppm)
No. of Observations
1 "MinHour" refers to the minimum hourly average associated with
minute peak of >0.75 ppm.
2 The refinery/industrial complex data contains fewer
observations than indicated in Table 3-1 because hourly averages wer
not available for all the hours recording high 5-minute peaks.
-------
41
concentrations below 0.25 ppm (Smith, 1993). If, for the
purposes of illustration, we assume an hourly standard of 0.25
ppm was in place, some of the 5-minute peaks would have been
restricted. However, many hourly concentrations without peaks
would be controlled, and the majority of the 5-minute peaks still
could have occurred, since the associated hourly concentrations
would be permissible.
C. Nationwide Estimates of Short-Term Peak SO-, Levels
The staff attempted to estimate the nationwide
prevalence of 5-rainute peaks > 0.50 and > 0.7 5 ppm. Because
5-rainute S02 data are not readily available, it was necessary to
rely on hourly data to generate more comprehensive estimates of
the likelihood of high short-term S02 peaks than those presented
in Table 3-1. The use of hourly data requires employing peak-to-
mean ratios to obtain estimates of 5-minute concentrations;
however, as pointed out above, peak-to-mean ratios may not give a
reliable indication of high short-term peak S02 levels. To
address this problem, staff assumed an upper bound peak-to-mean
ratio of 3-to-l (5-minute concentration to hourly average) and a
lower bound peak-to-mean ratio of 2-to-l.
For example, to obtain lower bound estimates of exposure to
5-minute, 0.7 5 ppm concentrations using the 2-to-l peak-to-mean
ratio assumption, the staff examined all hourly averages reported
in the AIRS database for the year 1992 that exceeded 0.38 ppm.
An hourly average of 0.3 8 ppm is the approximate value at which a
typical peak-to-mean ratio of 2 to 1 would predict on average a
-------
42
5-minute value £0.75 ppm.5 Fifty monitors (out of 721 monitors,
or approximately 7 percent) recorded at least one hourly average
as high as 0.38 ppm (Table 3-3A). At these monitors, only two
values greater than the level of the 24-hour primary standard and
t/a 1 noe '~"Of" ~•ho t —h ni iv* canATiHaTV c4"ja,n/4aT,H
liaWvii' \ru XUi6S (>£UcL Lt*JET wXaCIJi w*16 J JTavJIJIJl wCwU*iUG%Jl j & wuliUuia U Wei. C
recorded in 1992 (excluding a monitor based at the Hawaiian
Volcano).
Because several monitors can be located around a single
source, the number of counties (38) that had recorded hourly
averages >0.38 ppm may provide a better indication of the number
of distinct sources or sites. This represents"approximately a 50
percent reduction in the number of counties reporting 1-hour
averages >0.38 ppm since 1978. Much of this reduction has
occurred since 1989 (Smith, 1993). While estimating potential
population exposure is difficult, especially since the geographic
extent of the area affected by any short-term peaks is uncertain,
18 of these 38 counties contained urban populations (cities or
towns).
For the upper bound estimate of exposure, all hourly
averages >0.25 ppm were also examined assuming a peak-to—mean
ratio of 3 to 1 to predict the potential for high 5-minute
values. Based on the available data, the assumption that all
5The use of hourly averages to estimate high 5-minute peaks
miKa ir i s e afM^irAVi wisn+'A V"iar>aiiea e#"Mma A'f tnnn i f A>*e
UIU 9 JUcif V Ju c W £101 €¦ m Jl w X JL lllcS w c £¦} 6 c) uHS 6 His willVip O JL wllw Jill UII J. win w Jl
recording high hours will not have associated 5-minute peaks 2
(or 3) times as high; on the other hand, some monitors with low
hourly averages that therefore do not appear on Table 3-3 may
have high 5-minute peaks.
-------
TABLE 3-3. Analysis of Hourly Averages Nationwide1
A. Sites Recording High Hourly Averages - 1992
Sites Recording Hourlv Averacres
> 0.38 ppm
> 0.25 ppm
> 0.17 ppm
50 total sites
132 total sites'
247 total sites
(7%)
(18%)
(34%)
38 counties/18 cities
91 counties/65 cities
148 counties/124 cities
B. Sites Recording Multiple High Hourly Averages - 1992
Number of
Readings > the Hourly
Hourly Avg.
Case
Location
Average
(ppm)
(ppm, Peak-to-Mean)
1
3
5
0.38
0.75, 2 to 1
Sites
50
16
9
Counties
38
12
7
0.25
0.75, 3 to 1
Sites
132
74
52
0.5, 2 to 1
Counties
91
56
39
0.17
0.5, 3 to 1
Sites
247
164
119
Counties
148
107
82
C. Sites Recording High Readings in 1990, 1991/ & 1992
0.38 ppm
0.25 ppm
0.17 ppm
19 sites
72 sites
156 sites
16 counties
57 counties
106 counties
'For this table, all site counts exclude the Hawaii Volcano, which is a nonanthropogenic source
-------
44
hourly averages £0.25 ppm may have 5-minute peaks of 0.75 ppm or
rrrofl i pt sccnr11 srfin ufi t*n T*notti AnnAAT'>c! t*a ncjpy\/pit" i \/o Tho
%m ~i€i v>Q4 Ci*SI9 WVC* W0U ww Jt» Va>4J. v<|iG>4U u WMC5Gi5 JL v u l» J> V Q> m X *«>!<
numbers of both monitors and counties with at least one hourly
average >0.25 ppm are significantly greater than those for hourly
averages >0.38 ppm (Table 3-3A}. At this bound, 132 monitoring
sites and 91 counties, 65 of which contain urban populations,
potentially could experience 5-minute peak S02 levels >0.75 ppm.
For comparison, the staff also assessed the number of sites
that potentially could have 5-minute S02 levels >0.5 ppm. The
number of sites recording at least one hourly average >0.25 ppm
(132 sites, 91 counties, 65 urban -areas) serves as an estimate of
the number of sites that might experience 5-minute S02 level
>0.5 ppm, assuming a peak-to-mean ratio of 2 to i (lower bound).
In that same year, 247 sites, located in 148 counties with 124
urban areas, recorded at least one hourly value >0,17 ppm and
potentially could experience 5-minute peaks >0.5 ppm, assuming a
peak-to-mean ratio of 3 to 1 (upper bound).
The staff next examined how many of the sites and counties
experienced multiple hourly averages >0.38 ppm, 0.25 ppm, and
0.17 ppm during 1992. The results for the number of sites
recording 1, 3, and 5 hourly averages greater than or equal to
the three cutpoints are presented in Table 3-3B. The number of
sites recording multiple hourly averages >0.38 ppm decline much
more sharply than those recording hourly averages >0.25 ppm or
0.17 ppm. Only nine sites recorded five hourly averages >0.38
-------
45
ppm while 52 sites recorded five hourly averages >0.25 ppm and
119 sites recorded five hourly averages >0.17 ppm.
The staff also examined data from 1990 and 1991 to determine
how many of the sites that recorded high hourly averages in 1992
also had high hourly averages in the preceding 2 years (Table
3-3C). Of the 50 sites that recorded at least 1 hourly average
SO.38 in 1992, only 19 record those values in all 3 years. Of
the 132 sites recording hourly averages >0.25 ppm, only 72 of
those sites recorded hourly averages of >0.25 ppm in all of the 3
years examined. Similarly, of the 247 sites recording hourly
averages of >0.17 ppm in 1992, 157 recorded high hourly averages
in all 3 years. This information suggests that the occurrence of
monitored high hourly averages at a given site is variable.
The use of existing hourly data to assess the potential
prevalence of 5-minute peak S02 levels has other limitations
beyond those introduced by the use of peak-to-mean ratios. The
existing monitoring network is designed to accurately
characterize ambient air quality associated with 3-hour, 24-hour,
and annual S02 concentrations rather than to detect short-term
peak S02 levels. As a result, the EPA1s monitoring guidance on
siting criteria, the spanning of S02 instruments, and instrument
response time (Eaton et al., 1991) could lead to underestimates
of high 5-minute peaks and thus 1-hour averages for hours
containing those peaks. Such underestimates would lead to
underestimates of the number of nationwide sites recording high
hourly values in the results given above.
-------
Monitor siting constraints may be the biggest potential
source of underestimation of the occurrence of S02 peaks. In
1992, approximately 700 monitors reported data. This contrasts
to the more than 6,000 sources that may produce high peak S02
levels (Appendix B, Table B-2). Therefore, it is likely that
changes in monitoring siting and density in the proximity of S02
sources would increase the number of high 5-minute and associated
1-hour averages recorded.
D. Nationwide Estimates of Exposure
Another approach to estimating the frequency of short-term
peaks of S02 is through exposure'analysis, a technique that has
^ho aH W5I n'h arra "F l rir'AVTi i n it 1 i Iro 1 i ViaaH -Hh art
wills dUvlcU aUVantayc UI inturporatiny tuc XlJ\611XlUUu Lua v. afl
asthmatic individual may experience a response to that peak.
Exposure analysis predicts both the frequency that a concentrated
peak of S02 will occur (through air quality modeling) and the
probability that an asthmatic individual will be outdoors at
sufficient ventilation to be at risk from that peak. In the
analyses discussed below the probability of an "air quality
event" of a 5-minute peak >0.5 ppm S03 (or >0.75 ppm) is
determined and combined with the probability that an asthmatic
individual will be outdoors at sufficient ventilation (>3 5
L/min).
Since both the existence of concentrated peaks of S02 and
episodes of breathing at elevated ventilation outdoors are
relatively infrequent occurrences, the combined probability of
these events occurring simultaneously is relatively low.
-------
47
However, when a source produces numerous concentrated peaks or
31 r r ^ 42 3 I 3 Y*/"f £J n /"Si *1 #"f n 1"S 11 I 3 t* 1 "f* 1h| o I "I Q I I h rt rtn i n 9 C O C x* H 3 T"*
€1 ^ L *5> «•» w9 d X CI4 y C wX lv Ui«j 11 pUMU JL Q. t> JL U* I f «¦»* 16 JL JL Ai t» JL> JL11UUU JL1 1 lm CQ.»3CS3 UiiCt w
at least some asthmatic individuals in the vicinity will
encounter a peak while at sufficiently high ventilation outdoors.
The following discussion briefly reviews activity data used
for these analyses, and presents the results of two analyses
evaluating the probability that an asthmatic individual will be
outdoors at elevated ventilation and be exposed to a short—term
peak of S02 as a result of emissions from either utility or
nonutility sources. The utility exposure analysis was performed
by a contractor, System Applications, Inc., for the Utility Air
Regulatory Group. The nonutility analysis was performed by the
same contractor using a similar methodology for the Environmental
Protection Agency.
1. Activity patterns
Both exposure analyses used activity data derived from a
diary study of the general population carried out in Cincinnati,
Ohio. When this data was aggregated into hour blocks, from 0-3.5
percent of the people-hours were spent outdoors exercising at a
"high" activity level (Stoeckenius et al., 1990, p.8 and
Fig. 2-2).
For these analyses, individuals at a "high" activity level
were considered to be ventilating on average >35 L/min, the point
at which a majority of the general population begins oronasal
breathing (breathing through both mouth and nose). This is the
point at which nasal scrubbing of S02 begins to be bypassed and
-------
48
an asthmatic person is at greater risk of experiencing a
response. This is probably a reasonable approximation; however,
further work has shown some individuals at medium or moderate
activity may ventilate at >35 L/min, while some individuals do
not ventilate at that level, on average, even during what they
describe as "high" or "fast" activity.
Comparing the activity patterns for the general population
with the activity patterns of asthmatic persons is difficult.
Many mild asthmatic individuals, who constitute the majority of
asthmatic persons, and also some moderate asthmatics, are
encouraged, as part of their therapy, to exercise to maintain
lung function. Thus, some asthmatic individuals may be more
active than the general population. However, approximately 2 0
percent of people with asthma report at least some activity
limitation from their disease (NCHS, 1993), and it is reasonable
to expect that many of these individuals (particularly those with
severe disease) would be less active than the general population.
The only activity study which attempted to obtain a
representative sample of asthmatics found their activity levels
to be comparable to, or slightly greater than the general
population estimates (Roth Associates, 1988). Other studies
(Linn, 1991), composed primarily of individuals with moderate
and/or severe disease, have found comparable or lower activity
patterns (Appendix B),
-------
49
2. Exposure Analysis Results
The exposure analyses combine the probability of being at
elevated ventilation with the probability of encountering a peak
of S02. The probability of occurrence for a peak of S02 is
determined by using an air quality model to predict the number of
peaks occurring in an area within a year. The precision of the
air quality model estimate depends greatly on the quality of the
emissions data. For the utility analysis, detailed information
on actual emissions was available on a plant-by-plant basis. For
the nonutility analysis, actual data were not available. As
discussed in Appendix B, the following assumptions were made: for
many sources constant operation at the maximum hourly design rate
was assumed (a very high rate of operation), while for other
sources constant operation at the annual average emission rate
was assumed (a rate lower than that attained approximately half
the hours for the year, and probably lower than many hours with
high peaks). Neither approach to nonutility emissions provides
what would be most desirable, estimates of the frequency and the
geographic extent of concentrated peaks resulting in part from
emissions fluctuations of less than one hour duration.
The lack of emissions data means that the nonutility
anslvsi? a 1 a y-ci& ^onrrp of uncerta i ntv not hv +"Ha
U i 1 U JL j w Jk <13 i £ U i3 U JL d JL M C W U 4 v C w J* V* 1 1 K* C JL V# O X i 1 Vjr i IV v i9l I Q J. C U W Jf Wil C
utility analysis. Because of this, the nonuti1ity estimates of
exposure events, and number of asthmatic individuals exposed, are
-------
50
given as ranges that depend on some of the modeling assumptions.
Both analyses have a number of additional uncertainties that are
listed in Table 3-4, and described in Appendix B. Given these
uncertainties, these analyses were not intended to generate
precise estimates of the number of asthmatic individuals exposed.
However, these analyses do provide estimates of the relative size
of the potentially exposed population.
The analyses (Table 3—5J indicate that numerous exposures of
asthmatic individuals at exercise outdoors to concentrations >0.5
ppm may occur nationwide (180-395,000 events). (Throughout the
following text and tables,, all references to "exposures," "SO,
exposures," or "asthmatic individuals exposed" refer to exposures
of asthmatic individuals to S02 while at exercise outdoors).
However, relative to the total population of asthmatic
individuals, short-term S02 exposures do not appear to be a
pervasive problem. The 68,000 - 166,000 asthmatic individuals
estimated to be exposed 1 or more times per year to
concentrations >0.5 ppm S02 comprise approximately 0.7-1.8
percent of the total asthmatic population. Because the
population of asthmatic individuals 1iving in the vicinity of SO,
sources (and thus having the potential to be exposed to SO,) is
smaller than the total asthmatic population, it follows that more;
-------
TABLE 3-4. SOME IMPORTANT SOURCES OF UNCERTAINTY IN EXPOSURE CALCULATIONS
Source
Likely Magnitude & Direction
of Error on Exposure Estimate
Modeling Uncertainties
Prototype selection/binning
Meteorological modeling uncertainties
Peak-to-Mean Ratio
Representativeness
Weather & ratio uncoupling
Exposure Modeling
Activity pattern update
(ventilation rates, timing of exercise)
Asthmatic activity patterns
Uniform population assumptions (around utilities)
Emissions
Nonutility emission estimates
Non-included sources
Estimates of affected areas
Complex terrain
Overlapping sources !
Multiple peaks in an hour
Unknown
Unknown
Small to moderate, unknown
Small to moderate, unknown
Small to moderate, unknown
Small to moderate, over
Small, unknown
Large, over for some sources,
under for some sources
Small to moderate, under
Small, under
Unknown, under
Small to moderate, under
Small, under
-------
52
TABLE 3-5
^YpACT TO T? A MAT VCTC U17QTTT TQ ((I C PPM\
3U] UdwJCUli Al1OaO a\J—rOiJJL^ 1.0 \U*<3 aJrlYjy
NATIONWIDE
Total
Exposure
Events
180,000-395,000
No. of Asthmatic
Persons Exposed
IX or More
68,000-166,000
Percent of
Total Asthmatic
Population
0.7-1.8%
SECTOR-SPECIFIC
UTILITIES
•
NON-UTILITIES
Exposure Events
68,000
Exposure Events 114,000-325,000
Full Load
Exposure Events
— 118.0001
No. of Asthmatic
Persons Exposed
IX or More 24,000-122,000
Post-Title IV
Exposure Events
40,000
Post-Title IV
0.75 ppm Exposure
Events
9,000
Industrial Boilers
Exposure Events 56,000-201,000
'Estimated from Table B-l, Appendix B, applying the 5% correction (Rosenbaum et al., 1992, p.2).
-------
than 0.7-1.8 percent of this subset would be exposed.6 Because
the total number of exposure events exceeds the estimated number
of asthmatic individuals exposed by approximately 2-to-3-fold,
asthmatics exposed to SO^ at exerci.se are being exposed two to
three times a year on average, with more frequent exposures
possible for a substantial fraction.
The analyses indicate that asthmatic individuals are more
likely to be exposed multiple times during a year around
nonutility sources than utility sources. The 114,000 to 326,000
estimated exposures around nonutility sources are estimated to
affect 24,000 to 122,000 asthmatic persons, implying that exposed
individuals may be exposed more than four times a year, on
average.
This is in contrast to the utility situation, in which
68,000 exposures are estimated to affect approximately 44,000
asthmatic persons. However, the utility analysis did not take
into account the potential concentrating effects of terrain for
the estimated 25 percent or more of power plants estimated to be
located in complex terrain, which might be expected to increase
the chance that a proportion of asthmatic individuals living in
6 The utility analysis in Table B-l of Appendix B does
generate estimates of the number of asthmatics exposed as a
percentage of the number living in the vicinity of power plants,
but the non-utility analysis acknowledges that it cannot
discriminate amongst individuals living in proximity of more than
one source. For example, an asthmatic individual is counted twice
if living in vicinity of two sources. Thus, double counting and
an overestimate of the asthmatic population with the potential to
be exposed would be expected to occur.
-------
54
the vicinity of those plants would be exposed multiple times per
year.
The utility sector accounts for about 17-37 percent of the
total exposures. If the full load emissions allowable under
f hpi r nprin i f q uarp a^suinpfi rathpr than Actual pm i i nn^ thp
tllClii All JL. Wwi. C uOOUIUCVa 4 u WliCi V**Gli I QWUUuX CtllASOi VllQ f W4iC
total exposure events from utilities increases approximately 75
percent. Under full implementation of the restrictions being put
into place under the Title IV program to address acid deposition,
by the year 2015, exposures to emissions from utility boilers are
estimated to drop to about 58 percent of current levels,
contingent on trading decisions. An analysis of estimated
exposure events at 0.75 ppm S02 after the Title IV program shows
that exposures for utility sources at this higher concentration
are less than one-fourth of those at 0.5 ppm.
Among the nonutility sources, industrial boilers are the
source category most responsible for potential exposures,
accounting for approximately half the total exposure events from
this sector. Other categories that may result in a substantial
number of exposures include petroleum refineries, pulp and paper
mills, sulfuric acid plants, and aluminum smelters (not included
in the analysis were lead smelters, steel mills, cement plants,
and other potential sources of exposures). Among certain source
categories, such as aluminum smelters, copper smelters and
sulfite mills, estimates indicate that from 1.5 to 3 percent to
as much as 10 to 30 percent of the asthmatic individuals living
in the vicinity may be exposed at least once per year, depending
-------
55
on assumptions made in the air quality modeling (Appendix B,
Table B-2 and Notes).
When individual source categories could be examined more
extensively, the risk of exposures was very unevenly distributed
across the sources in the category. For instance, approximately
75 percent of the utility sector's post Title IV exposures were
estimated to result from less than 10 percent of the power plants
(Burton et al., 1987; Rosenbaum, 1992, Table 3). Similarly,
approximately half of the total industrial boiler exposures can
be attributed to a small proportion (1.5 percent) of the total
*
population of industrial boilers analyzed (Stoekenius et al.,
1990, Table 2-14).
For other source categories, the same "clustering of risk"
phenomenon may also be evident: for instance, sulfite paper
mills account for twice as many estimated exposures as kraft
mills, but represent only a sixth of the total paper mills.
Information on a source's mode of operation, control equipment,
and types of raw materials or fuel used may help in developing
focused, efficient implementation efforts.
E. Conclusions
The available air quality and exposure data provides a
strong indication that the likelihood that asthmatic individuals
will be exposed to 5- to 10-minute peak S02 concentrations is
quite low when viewed from a national perspective. The data also
indicate, however, that high peak S02 concentrations can occur
around certain sources or source types with some frequency. This
-------
56
suggests that asthmatic individuals that reside in the vicinity
of such sources or source types may be at greater risk than that
indicated for the asthmatic population as a whole. Because of
this, the staff recommends that the Administrator consider
targeted strategies when assessing approaches for reducing
potential peak S02 exposures.
-------
57
IV. STAFF CONCLUSIONS AND RECOMMENDATIONS
Based on the assessment and interpretation of the health
effects information presented in the criteria document supplement
and summarized above, the staff concludes:
1) About 10 million people, or 4 percent of the population
of the United States, are estimated to have asthma.
The prevalence is higher among African-Americans, older
children (8 to 11 years old), and urban residents.
Common symptoms include cough, wheezing, shortness of
breath, chest tightness, and sputum production. Asthma
is characterized by an exaggerated bronchoconstrictor
response to many physical challenges {e.g., cold or dry
air, exercise, specific stimuli such as pollen) and
chemical and pharmacological agents. Daily variability
in lung function measurements is also a typical feature
of asthma. Asthma attacks can result in
hospitalization or emergency room treatment. Death due
to asthma is, however, a rare event. Many asthmatic
individuals take medication to relieve symptoms and
functional responses associated with exacerbation of
this disease. One of the most commonly used asthma
medications (beta-agonist) also inhibits or ameliorates
responses to S02. Available data suggest, however,
generally low medication use and compliance rates for
many mild and moderate asthmatic individuals.
-------
58
Mild and moderate asthmatic children, adolescents and
adults represent the population groups most at risk for
short-term peak S02 induced effects. More severe
asthmatic individuals have very poor exercise tolerance
and therefore are less likely to engage in sufficiently
intense exercise to permit notable S02-induced effects
•I-a nrriir
Ww VvvUl •
A substantial percentage (>20 to 25 percent) of mild to
moderate asthmatic individuals exposed for 5 to 10
minutes^ to 0.6 to 1.0 ppm S02 during moderate exercise
would be expected to have respiratory 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).
After the initial 5 minutes of exposure to 0.6-1.0 ppm
S02 the severity of effects for many of the responders
is likely to be sufficient to be of concern, i.e., to
cause disruption of ongoing activities, use of
bronchodilator medication, and/or possible seeking of
medical attention. At S02 concentrations in this range
the intensity of distress is much more likely to be
perceived as an "asthma attack" than would be the case
at exposures below 0.5 ppm S02.
The effects observed after exposure to 0.6 to 1.0 ppm
S02 are relatively transient (not lasting more than a
-------
59
few hours) and are not likely to worsen or to reoccur
with the same magnitude of response if re-exposure to
another S02 peak occurred within the next several hours
after the initial exposure, should they choose to
resume physical exertion after amelioration or
cessation of any initial SOj^mduced distress.
6) At S02 concentrations at or below 0.5 ppm, only a
relatively small percentage (<10 to 20 percent) of mild
and moderate asthmatic individuals exposed to 0.2 to
0.5 ppm S02 during moderate exercise are likely to
experience lung function changes distinctly larger th'an
those they typically experience. Furthermore, compared
to the response at 0.6 to 1.0 ppm S02, the response at
or below 0.5 ppm S02 is less likely to be perceptible
and of immediate health concern.
In assessing the public health significance of the effects
reported at 0,6 ppm S03 or above, the Administrator should also
consider the following factors: l) the effects reported for mild
or moderate asthmatic individuals are likely to be more
pronounced if that individual is at higher than moderate
ventilation; 2) the degree of concern or perceived significance
of the response would likely increase with increased frequency of
exposure over the course of the year; 3) while prophylactic
bronchodilator medication use prior to exercise might protect
against S02-induced effects, the relatively low medication
compliance rates indicate that many mild and moderate asthmatic
-------
60
individuals may be unprotected, of particular concern are those
individuals of lower socioeconomic status with limited access to
health care; 4) the available epidemiological data do not
provide a basis for concluding that S02 contributes to excess
asthma mortality rates observed among non-white population groups
in large urban areas; and 5) the available air quality and
exposure data provides a strong indication that the likelihood
that asthmatic individuals will be exposed to 5- to 10-minute
peak S02 level is quite low when viewed from a national
perspective. Yet, the data also indicate that peak S02
concentrations do occur and suggest that asthmatic individuals
that reside in the vicinity of certain sources or source types
will be at increased risk.
Based on its assessment of the available health, air quality
and exposure data, the staff recommends that the Administrator
consider three possible regulatory alternatives:
1) Establish a new 5~minute NAAQS in the range of 0.6 to
1.0 ppm S02 expressed as the maximum 5-minute block
average in 1 hour. In view of the nature of the
response and the low probability that a given asthmatic
individual will be exposed while at elevated
ventilation, consideration should also be given to
permitting multiple exceedances (e.g., up to 5) during
a year. If the Administrator determines that a new 5-
minute NAAQS is needed, the staff also recommends that
it be implemented through a risk-based, targeted
-------
61
approach focusing on those sources that are roost likely
to produce repeated high 5-minute peaks during the
course of a year.
2) Establish a new regulatory program under the general
authority of section 303 of the Clean Air Act. Such a
new program should establish a target level for control
in the range of 0.6 to 1.0 ppm SOi expressed as the
maximum 5-minute block average in l hour. In
establishing the target level, the staff recommends
that multiple exceedances (e.g., up to 5) be permitted
during a year and that the program be implemented
through a risk-based, targeted strategy. This approach
would be designed to supplement the existing NAAQS by
placing, in effect, a cap on short-term peak S02
ambient levels, the exceedance of which would result in
enforceable action against the source(s) causing or
contributing to the exceedance. Thus, the program
would provide additional protection for asthmatic
individuals, without many of the burdens that
implementation of a new 5-minute NAAQS would impose
upon the states.
3) Retain the existing suite of standards but augment
their implementation by focusing on those sources that
are likely to produce high 5-minute peak S02 levels.
This approach would be aimed at assuring that the
existing standards are met through more targeted
-------
62
monitoring and adherence to existing regulatory
provisions governing good operating practice and upset
and malfunctions, thereby providing some additional
protection against short-term peaks.
In selecting among these alternatives, the staff recommends
that the Administrator consider the nature and significance of
the health effects associated with short-term peak S02 levels and
the size of"the mild and moderate asthmatic population
potentially at risk. Given the available scientific and
analytical data, the staff recognize that the ultimate decision
of the Administrator will be based in part on policy/legal
considerations.
-------
A-l
APPENDIX A
SAB—CASAC-8 7—22
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. C C. 20460
February 19, 1987
The Honorable Lee M, Thomas
«r»ici o*
**€ *QMiftiSTa«T3ii
Administrator
U.S. Environmental Protection
Agency
Washington, DC 20460
Dear Mr. Thmas:
The Clean Air Scientific Advisory Canaittee (CASAC) has completed
its review of the 1386 Addendum to the 1382 Staff Paper on Sulfur CKides
(Review of the National Acblent Air'-CXallty -Standards for Sulfur Qtldes;
Opdated Aaseaament of Scientific ai>d T&chiucai Infornatlon) prepared by
the Agency's Office of Air Quality. Planning arid Standards (QAQPS).
Hie Conroittee unanimously concludes that this doctsnent is consistent
in all significant respects with the scientific evidence presented and
Interpreted in the combined Air Quality Criteria Doctirent for Particulate
Katter/Sulfur Oxides (1982) and its 1986 Addendun, on which CASAC issued
its closure letter on Deoesfcer 15, 1986. She Canaittee believes that the
1986 Addendum to the 1982 Staff Paper on Sulfur GKides provides you with
the kind and amount of technical guidance that will be needed to make
appropriate decisions with respect to the standards. The Committee's
major findings and conclusions concerning the various scientific Issues
and studies discussed in the Staff Paper Addendum are contained in the
attached report.
Thank you fbr the opportunity to present the Coranittee's views on
this inportant public health apd welfare issue.
cc: A, James Barnes
Gerald Enison
Lester Grant
Vaun New ill
John O'Connor
Craig Potter
Tterxy Yosie
Sincerely,
Norton
Chairman
Clean Air Scientific Mvisory
Gaanittee
-------
A-2
SAB—CfiSAC-8 7-l
SUMfftRY OF MWCR SCIENTIFIC ISSUES AND CASAC
CONCLUSIONS CN THE 1986 DRAFT ADCOJDCJK
TO THE 1982 SULFUR OXIDES STAFF PAPER
The Conmictee found the technical discussions contained in the Staff
Paper Addendum to be scientifically thorough and acceptable, subject to
minor editorial revisions. This document is cots is tent in all significant
respects with the scientific evidence presented in the 1982 combined Air
Quality Criteria Document for Particulate Matter/Sulfur Oxides and its 1986
Addendum, on which the Conmlttee issued its closure letter on Deoeitfcer IS#
1986.
Scientific Basis for Primary Standards
The Conraittee addressed the scientific basis for a 1-hour, 24-hour,
and annual primary standards at aafljfe"*lengtfa-' in its August 26, 1983 closure
letter on the 1982 Sulfur Oxides Staff-Paper. That letter was based on
the scientific literature which had- been published up to 1982. The present
review has examined the more recently published studies.
It is clear that no single study of SO? can fully address the range of
public health issues that arise during the standard setting process. The
Agency has completed a thorough analysis of the strengths and teaknesses of
various studies and has deriwd its tecooraended ranges of interest by
evaluating the weight of the evidence* The Committee endorses this approach.
The CaiSRittee wishes bo consent on several major issues concerning the
scientific data that" are available. These Irwaire include:
• Recent studies more clearly inplicate particulate matter than S02
as a longer-tera public health concern at low exposure levels.
• A majority off Conreittee mesfeers belief that the effects reported
in the clinical studies of asthmatics represent effects of
significant public health concern.
• The exposure uncertainties associated with a 1-hour standard are
quite large. The relationship between the frequency of start-term
peak exposures and various soenarios of asthmatic responses is not
wen understood. Both EPA and the electric power industry are
conducting further analyses of a series of exposure assessment
issues. Such analyses haw the potential to increase the collectiw
understanding of the relationship between SO2 exposures and responses
obserrod in subgroups of the general peculation.
• The nunfeer of asthmatics vulnerable to peak exposures near electric
power plants, giwn the protection afforded by the current standards,
represents a snail mottoes of people. Although the Clean Air Act
requires that sensitive population groups receive protection, the
size of such groups has not been defined. CASAC believes that this
issue represents a legal/fcolicy matter and has no specific scientific
advice to provide on it.
-------
CASAC's advice on primary standards for three averaging tiroes is presented
belcw:
1-Hour Standard - It is our conclusion that a large, consistent
data base exists to document the bronchoconstrictius response in mild
to moderate asthmatics subjected in clinical chanters to short-term,
low lewis of sulfur dioxide while exercising, there is, however, no
scientific basis at present* to support or dispute the hypothesis that
individuals participating in the SO2 clinical studies are surrogates
for more sensitii« asthmatics. Estimates of the size of the asthmatic
population that experience exposures to short-term peaks of SDj
(0.2 - 0.5 parts per million Ipps) SO2 for 5-10 minutes) during light
to moderate exercise, and that can be expected to exhibit a broncho-
oo«stricti« response, varies from S,000 to 50,000. ¦
The majority of the Conraittee beliefs that the scientific evidence
supporting the establishment of.a new 1-hour standard is stronger than
it was in 1983. As a result# and in view of the significance of the
efSects reported in these clinical stbdies, there is strong, but not
unanimous support for the reccpraendation that the Administrator consider
establishing a new 1-hour standard for SO2 exposures. Hie Committee
agrees that the range suggested by EPA stiff (0.2 - 0.5 ppn) is
appropriate# with several reenters of the Coraaittee suggesting a standard
from the middle of this range. The Coomittee concludes that there is
not a scientifically demonstrated need for a wide margin of safety for a
1-hour standard.
24-Hour Standard - The more recent studies presented and analyzed
in the 1986 Staff Paper Addendum# ¦ in particular, the episodic lung
function studies in diildren (Docfcery et al., and Dassen et ai.) serve
to strengthen our previous conclusion that the rationale for reaffirming
the 24-hour standard is appropriate.
Annual Standard - the* Conraittee reaffirms its conclusion, usiced in
its 1983 closure letter, that- there is no quantitative basis for retaining,
the current annual standard. However, a decision to abolish the annual
standard nust be considered in the light of the total protection that
is to be offered by the suite of standards that will be established.
the abo« recommendations reflect the consensus position of CASAC. Not
all CASAC reviewers agree with each position adopted because of the uncertainties
associated with the existing scientific data. Howewr, a strong majority
supports eash of the specific recommendations presented above, and. the entire
Cortraittee agrees that this letter represents the oonsensus position.
Secondary Standards
The 3-hour secondary standard was not addressed at this review.
-------
J2- #
f-D-
UN1TE0 STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D.C. 20460
June 1, 1994
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
EPA-SA8-CASAC-LTR-94-Q07
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M St, S.W.
Washington, D.C. 20460
Subject: Clean Air Scientific Advisory Committee Closure on the
Supplements to Criteria Document and Staff Position
Papers for SO2
Dear Ms. Browner:
The Clean Air Scientific Advisory Committee (CASAC) at a meeting
on April 12, 1994, completed its review of the documents: Supplement to
the Second Addendum (1986) to Air Quality Criteria for Particulate Matter
and Sulfur Oxides; Assessment of New Findings on Sulfur Dioxide and
Acute Exposure Health Effects in Asthmatics; and Review of the National
Ambient Air Quality Standard^ for Sulfur Oxides: Updated Assessment of
Scientific and Technical Information, Supplement to the 1986 OAQPS
Staff Paper Addendum. The Committee notes, with satisfaction, the
improvements made in the scientific quality and completeness of the
documents.
With the changes recommended at our March 12 session,-written
comments submitted to the Agency subsequent to the meeting, and the
major points provided below, the documents are consistent with the
scientific evidence available for sulfur dioxide. They have been organized
in a logical fashion and should provide an adequate basis for a regulatory
decision. Nevertheless, there are four major points which should be called
to your attention while reviewing these materials:
m RocyciwfcfteeydBCiio
pTOiMawtthSoy/CiAWBimoncaciaftnai
-------
A-5
-2-
1. A wide spectrum of views exists among the asthma specialists
regarding the clinical and public health significance of the effects of 5 to
10 minute concentrations of sulfur dioxide on asthmatics engaged in
exercise. On one end of the spectrum is the view that spirometric test
responses can be observed following such short-term exposures and they
are a surrogate for significant health effects. Also, there is some concern
that the effects are underestimated because moderate asthmatics, not
severe asthmatics, were used in the clinical tests*
At the other end of the spectrum, the significance of the spirometric test
results are questioned because the response is similar to that evoked by
other commonly encountered, non-specific stimuli such as exercise alone,
cold, dry air inhalation, vigorous coughing, psychological stress, or even
fatigue. Typically, the bronchoconstriction reverses itself within one or
two hours, is not accompanied by a late-phase response (often more
severe and potentially dangerous than the immediate response), and shows
no evidence of cumulative or long-term effects. Instead, it is
characterized by a short-term period of bronchoconstriction, and can be
prevented or ameliorated by beta-agonist aerosol inhalation.
2. It was the consensus of CASAC that the exposure scenario of concern is
a rare event. The sensitive population in this case is an unmedicated
asthmatic engaged in moderate exercise who happens to be near one of the
several hundred sulfur dioxide sources that have the potential to produce
high ground-level sulfur dioxide concentrations over a small geographical
area under rare adverse metfeorological conditions. In addition, CASAC
pointed out that sulfur dioxide emissions have been significantly reduced
since EPA conducted its exposure analysis and emissions will be further
reduced as the 1990 Clean Air Act Amendments are implemented.
Consequently, such exposures will become even rarer in the future.
3. It was the consensus of CASAC that any regulatory strategy to
ameliorate such exposures be risk-based - targeted on the most likely
-------
A-6
-3 -
sources of short-term sulfur dioxide spikes rather than imposing short-
term standards on all sources. All of the nine CASAC Panel members
recommended that Option 1, the establishment of a new 5-minutes
standard, not be adopted. Reasons cited for this recommendation included:
the clinical experiences of many ozone experts which suggest that the
effects are short-term, readily reversible, and typical of response seen
with other stimuli. Further, the committee viewed such exposures as rare
events which will even become rarer as sulfur dioxide emissions stg
further reduced as the 1990 amendments are implemented. In addition,
the committee pointed out that enforcement of a short-term NAAGS would
require substantial technical resources. Furthermore, the committee did
not think that such a standard would be enforceable (see below).
4, CASAC questioned the enforceability of a 5-minute NAAQS or "target
level." Although the Agency has not proposed an air monitoring strategy,
to ensure that such a standard or "target level" would not be exceeded, we
infer that potential sources would have to be surrounded by concentric
circles of monitors. The operation and maintenance of such monitoring
networks would be extremely resource intensive. Furthermore, current
instrumentation used to routinely monitor sulfur dioxide does not respond
quickly enough to accurately characterize 5-minute spikes.
The Committee appreciates the opportunity to participate in this
review and looks forward to receiving notice of your decision on the
standard. Please do not hesitate to contact me if CASAC can Jbe of further
assistance on this matter.
Sincerely,
George T. Wolff, Ph.D.
Chair, Clean Air Scientific
Advisory Committee
-------
B-l
APPENDIX B
A. Additional Information About the Exposure Analyses
Tables B-l and B-2, placed at the end of this appendix,
provide a more full description of the results of the utility
analysis (Rosenbaum et al., 1992) and non-utility analysis
(S to^clcsn ius ct al. j 1990) . Footnotss to Tafol© B—2 cf i vb mo its
details about some of the assumptions used to generate certain
calculations presented in the main text.
B. Important Uncertainties Involved in the Exposure Analysis
Any nationwide exposure analysis must contain numerous
assumptions and simplifications. These can result in either
overstating or understating the estimates of exposures. A brief
summary of major sources of uncertainty is presented in Table 3-4
and discussed below. The major sources of uncertainty are
estimates of activity patterns, emission and dispersion modeling,
and the use of peak-to-mean ratios to estimate 5-minute peak S02
concentrations. For a more complete treatment, see EPA
(1986c,d), Burton et al. (1987), Stoeckenius et al. (1990),
Burton and Stoeckenius (1988), and Rosenbaum et al. (1992).
1. Activity Pattern Uncertainties
The activity pattern data used in these analyses has
undergone numerous revisions. The current analyses reflect the
best information available at the time regarding the number of
individuals estimated to be at high ventilation (Rosenbaum et
al., 1992; Stoeckenius et al., 1990). However, more recent work
has shown that some individuals at moderate activity may
-------
B-2
ventilate at > 35 L/mimite, while others may not ventilate at
that level, on average, even during "high" activity (Linn, 1991).
In addition, minute-by-minute ventilation estimates are now
available. The Human Exposure Model used for these analyses
assumed that an individual with asthma was at sufficient elevated
ventilation for the hour if she had at least 10 minutes of high
activity. This would lead to overestimates in hours where the 5-
minute peak did not coincide with the elevated ventilation.
However, some underestimation might occur during hours in which
individuals might briefly be at elevated ventilation {< 10
minutes) -and experience a high ambient S02 peak.
The activity data used in the exposure analyses is not
specific for individuals with asthma, which could affect the
exposure estimates. The general population activity data used
relied on a diary study of greater than 900 Cincinnati residents,
who were followed over 3-day intervals in cool or warm seasons
(see Johnson et al., 1993 for more information). The data as
used in Stoeckenius et al. (1990) and Rosenbaum (1992) indicate
that the general population spends 1.7% of waking hours at
strenuous, exercise, with peak hourly activity rates of
approximately 3.5% (Stoeckenius et al., 1990, Fig. 2-2, exact
numbers provided by Stoeckenius, personal communication).
In contrast, a study of a population-based sample of 136
Cincinnati residents with asthma followed over a 3-day period
reported the prevalence of outdoor exercise was greater than for
the general population: 3.3% of waking hours at strenuous
-------
B-3
exercise, with peak hourly activity rates of 7-9%, depending on
day of week (Roth Associates, 1992, p.3-7, 3-12). In a smaller,
3-day study of 52 asthmatic residents of Los Angeles involved in
the clinical studies conducted by the Linn group, survey
participants spent 2.4% of their waking hours, on average,
exercising outdoors (Roth Associates, 1988).
Finally, a more recent 7-day survey of 49 asthmatic
residents of Los Angeles who had been clinical subjects for the
Linn group, found that a group of individuals with primarily
moderate to severe disease spent only 0.2% of their total waking
time exercising outdoors at a fast breathing rate [or possibly
0.2% of hours, but it is unclear how much activity was needed to
classify an hour at high activity (Linn, 1991, p. 22, 37, and
Figure 4-1)].
One reason that the general population activity estimates
are lower than some of the activity estimates for the asthmatic
population might be because mild asthmatic individuals are
encouraged to exercise as part of their therapy. This factor may
not have been reflected in the Los Angeles surveys, because the
sample groups for both these studies probably overrepresented the
proportion of moderate to severe asthmatic individuals relative
to that asthmatic population as a whole. For the 1991 study, in
which high activity rates were extremely low, individuals with
moderate to severe disease comprised approximately 70% of that
group, in roughly equal proportions. Approximately 50% of the
-------
B-4
1988 group was composed of individuals with moderate to severe
asthma (Roth Associates, 1988, p.1-2).
The 1991 study (Linn, 1991) seems to suggest that a group
cairroased of more severe asthmatics is aenerallv less active
VWi it W W W V W U ^ Wt 14liU W> X Wta? w M 4 I W J* U Jk J *¦»
-------
B-5
analyses the models are used to provide estimates, specific to
time and location, of the ambient concentration of S02 for each
hour in a year. The accuracy of these models when used for this
purpose is not fully established, although an evaluation by Moore
et al. (1988) reported that the dispersion model used in the
utility analysis tended to over-predict slightly the average of
the highest hourly concentrations (e.g., 25 highest) relative to
those observed, when observations and predictions are allowed to
be unpaired in time and location.
In addition, tests of Gaussian dispersion models indicate
that stability classes (derived by averaging meteorological data
over a year or more) often fail to capture much of the
variability in meteorological parameters. For unstable weather
classes this may result in differences of up to 40 percent in
predictions of maximum concentration when compared to
measurements (Irwin, 1987). Whether such meteorological
variability would have much effect on predictions of exposures is
unclear, given the findings of Moore et al. (1988) and the fact
the exposure analyses did not rely solely upon stability classes,
but also used actual meteorological parameters (from historical
data) in estimating dispersion.
However, an important point to keep in mind when considering
the impact of air quality emission and dispersion uncertainties
is that the exposure estimates from the earliest EPA exposure
analysis indicated that exposures (around utilities) were
apparently the result of comparatively few ambient peaks of S02
-------
B-6
[on the order of 10-20 expected exceedances for any given ring
(EPA, 1986, Figure 3-3 and 3-4)]. If estimated exposures in the
subsequent analyses also result from relatively few ambient peaKs
(data comparable to these reports are not currently available),
then estimating the impacts of uncertainties on the exposure
estimates will be more difficult. It is conceivable that small
changes in the treatment of meteorological uncertainties, or
other uncertainties (such as the peak-to-mean and emission
modeling uncertainties discussed below) could lead to relatively
large changes in exposure estimates.
Additional meteorological uncertainty is introduced when
dispersion analysis is performed on a prototype source (utility
analysis), or meteorological records from one particular area are
applied nationwide (non-utility analysis). The utility analysis
used meteorological data specific to the prototypical plant's
location to model dispersion for all the sources in each of its
24 bins (a bin is a subset of sources modeled as resembling a
prototype source). Such steps are necessary to reduce the
computational complexity, but simplify meteorology by applying
meteorological data from one source to the modeling of many
sources. $
Due to the large number of sources, in the non-utility
analysis meteorological data from only one particular region was
used to model all sources from a source category. However,
efforts were made to diminish this uncertainty by choosing
conditions applicable to the region expected to account for the
-------
B-7
largest proportion of exposures (e.g., meteorological data from
the Pacific Northwest was used to model pulp and paper mills).
These simplified treatments of meteorology could result in either
over- or under-estimates of exposures in other regions of the
country.
3. Peak-to-Mean Ratio Uncertainties
Another potential source of uncertainty is introduced
through the use of peak-to-mean ratios. For present purposes,
the peak-to-mean ratio is the ratio of the maximum 5-minute
concentration for an hour divided by the hourly average (a peak-
to-mean ratio of 2 indicates for that hour the maximum 5-minute
concentration was twice the concentration of the hourly average).
Peak-to-mean ratios for these analyses were chosen using a Monte
Carlo simulation based on a frequency distribution derived from a
collection of observed peak-to-mean ratios.
Both analyses rely heavily (non-utility analysis) or
exclusively (utility analysis) on a distribution of peak-to-mean
ratios derived from 18 months of monitoring around the Kincaid
power plant, a tall, isolated coal-fired plant in Illinois.
This distribution has an average ratio of 2.2, and, although 88
percent of the values are peak-to-mean ratios of 3.5 or less
(Stoeckenius, 1990, Table 2-18), it does contain peak-to-mean
ratios up to 11 to 1. The Kincaid power plant is in the 80th
percentile of stack height (Burton and Stoeckenius, 1988). Thus,
on theoretical grounds, use of this ratio would be expected to be
conservative for the majority of the power plants in the nation.
-------
B-8
However, using a single peak-to-mean ratio distribution
introduces several uncertainties of undetermined magnitude. The
data from one source is essentially generalized to all sources.
The 18 months of data used to generate the ratio contained
apparently only three observations of a 5-minute concentration
above 0.5 ppm (Thrall et al., 1982, Figure 6), and the maximum
concentration observed equaled 0.5 6 ppm; both of these factors
might affect the applicability of the ratio to sources with more
numerous high 5-minute peaks. However, the Kincaid analysis
(Thrall et al., 1982, p. 27) noted a small but statistically
significant tendency for peak-to-mean ratios to decrease with
increasing hourly concentration. Thus, sources producing higher
ambient concentrations might have lower peak-to-mean ratios;
sufficient data is not currently available to test this
hypothesis.
Peak-to-mean ratios are also highly sensitive to weather
conditions. Thus, using a single peak-to-mean ratio from one
1 t* 1 AH woarie Via ^ ^V»o aecnwnf1 i An i c 4"Vi a +¦ k a
«L vjt# d X y 11 111 c a 11 w «¦» I Id w iail 1 w u o O1 Ul Xti£/ w JL JTT JLI9 IUu U6 ivitu w wtl c
meteorological conditions of that area apply to all areas.
Furthermore, choosing the peak-to-mean ratio from a distribution
of ratios from all hours, rather than hours segregated by
stability classes or other meteorological parameters, essentially
uncouples the choice of the peak-to-mean ratio from the weather
conditions for the hour used in the dispersion modeling.
It is difficult to determine what bias, if any, this would
bring to the analysis. Unfortunately, the Kincaid data was not
-------
analyzed in relation to meteorological conditions. Certain
weather conditions would be expected to result in high hourly
concentrations and low peak-to-mean ratios, while others result
in low hourly concentrations and high peak-to-mean ratios. Use
of a distribution from all hours could therefore overstate
exposures in some cases. However, if some sources have many
hours with moderate hourly concentrations, and if typically
moderate to high peak-to-mean ratios were observed, exposures for
these sources could be underestimated.
Other expected uncertainties resulting from use of a single
peak-to-mean distribution (e.g., whether monitor placement and
instrument response time understated peak-to-mean ratios) are
discussed in Burton and Stoeckenius (1988) .
The same uncertainties listed above apply to the application
of the Kincaid distribution to non-utility sources. However in
these cases, which typically involve much lower emission release
heights, use of the distribution is much more likely to overstate
the probability of high exposures. The higher peak-to-mean
ratios reported in Section III of this paper can probably be
explained by two factors. First, the ratios reported in Section
III do not examine all hours, but rather only those hours with
high (> 0.75 ppm) 5-minute peaks. Second, the sources examined
in Section III probably experienced substantial increases in
their emissions within the hour, which contributed to the high
observed peak-to-mean ratio. This second factor was partially
-------
B-10
taken into account for many sources in the non-utility analysis
by assuming constant operation at maximum design rate.
The estimates for non-utility sources resulting from use of
the Kincaid distribution were viewed as representing an upper-
bound on possible exposures. A different peak-to-mean
distribution with fewer extreme values (Stoeckenius et al., 1990,
Table 2-18) was used to generate lower-bound estimates.
The uncertainties surrounding the use of dispersion modeling
and peak-to-mean ratios in these analyses add considerable
uncertainty to the final estimates. However, refining the
methodology of these areas (for example, by using"5- to lQ-minute
rather than hourly meteorology data) would involve intensive
remodeling and other efforts.
4« Emission modeling Uncertainties
It should be noted that the recent uti1ity exposure analysis
(Rosenbaum et al., 1992) and elements of the non-utility analysis
(Stoeckenius et al., 1990) (refineries and some other sources,
see below) estimated actual exposures, rather than potential
exposures that could result if the source operates at its higher
permitted emission limit. Potential emissions were evaluated in
the previous IPA analysis (EPA, 1986). Thus, exposures could
increase for many of the sources analyzed if they decided to
increase emissions to the permitted limit (as was shown by the
full-load figures for utilitites given in Section III.D.2).
However, the objective of these analyses was to attempt to
-------
B-ll
estimate the magnitude of current exposures to S02, not to
predict the number of possible exposures.
Probably the largest single source of uncertainty in these
analyses is in the emissions estimates used for the non-utility
sources. For most non-utility sources, constant operation at the
maximum hourly design rate was assumed, which almost certainly
overstates actual emissions and exposures (and likely even
overstates potential emissions as well). For batch processes of
great variability, however (e.g., sulfite pulp and paper mills,
copper smelters), it is conceivable that peak emissions within an
hour may exceed the hourly design rates. These brief episodes
may be very important in terms of actual exposures, but
additional analysis would be required to determine whether the
frequency and magnitude of such episodes for these sources would
result in a greater number of exposures than the assumption of
constant operation at the hourly design rate.
For other sources (refineries and additional sources with
incomplete emissions information), annual emissions data was
used. This would be expected to understate emissions and
exposures, since these sources would undoubtedly emit at rates
above their annual emissions rate for a substantial number of
hours in the year, and probably these would be the hours
contributing most to exposures.
In contrast, the emission estimates for utilities, which
consisted of a Monte Carlo simulation using distribution of power
plant loads specific to season and time of day would be expected
-------
B-12
to produce estimates of actual exposures with much less
uncertainty. Some minor uncertainties about how fuel use and
consumption are handled with this approach are discussed in EPA
(198 6d).
5. Other Uncertainties
Another important uncertainty affecting these results is the
niTTpnt* inahil i tv of the tnodels used to address the effects of
vUii Ji CI I L XI lUik/ IXa UJr UUw iiiVVAv JL w W«Jv n4 WW CIV* v4Jr WiiC CJU i, CW Uw WJ.
complex terrain (which may affect the estimates for the more than
25 percent of U.S. power plants located in complex terrain, and
to a lesser extent the estimates for non-utility sources located
in complex terrain). In addition, the analyses did not attempt
to consider the effects of overlapping sources, occurrence of
multiple peaks in an hour, or some source types that might
contribute some additional exposures (lead smelters, coke ovens,
and possibly some small, < 25 MW, power plants). Each of these
factors might increase exposure estimates by small amounts.
Some understating of exposures might have also occurred in
the procedure used to estimate the affected area around different
sources. Exposures may occur beyond the 20 km, the furthest
distance typically modeled in the utility analysis (EPA, I986d,
p. 2-15), and within the distance (i.e., three building heights)
that could not be modelled in non-utility analysis (Stoeckenius
et al., 1990).
C. Conclus ions
Some of the uncertainties discussed above might be
relatively easy to address, while others might require intensive
-------
B-13
remodelling or data that is not readily available. Some
assumptions or simplifications would have to be retained if the
development of nationwide exposure estimates is to be a
manageable task. These uncertainties make it difficult to arrive
at precise estimates of the number of asthmatic individuals
exposed to SO,, and so the estimates of the total number of
annual exposures to high peaks of S02 should be viewed with
caution,
Nevertheless, despite the limitations, the recent exposure
analyses have provided better insight into the potential
magnitude of exposure to concentrated ambient peaks of SO, and
the sources most likely responsible for such peaks than was
previously available. The basic findings of these, analyses
(i.e., that exposures to high concentrations of SO, are
restricted to the vicinity of certain SO, sources and that these
exposures do not affect a large proportion of the nationwide
asthmatic population in any given year, although a greater
proportion of asthmatic individuals living close to certain
sources may be exposed} would probably not change even if
exposure estimates were to increase several-fold. Furthermore,
improved treatment of some of these uncertainties, such as those
resulting from the assumption of constant operation at maximum
design rates used for many non-utility sources, could
substantially decrease estimates of actual exposures, although
not necessarily potential exposures.
Refinements of these analyses or additional ambient
monitoring data could possibly indicate the need to reevaluate
-------
B-14
the relative importance of one source category versus another in
accounting for high 5-mimite peak ambient exposures. However/ it
is not expected that these refinements would alter the basic
thrust of the assessment that exposures to high 5-minute peaks of
SO-j are likely to be experienced almost exclusively by asthmatics
who are in the vicinity of a subset of SO, sources.
D. Notes on Calculations Performed for the Text (Section III.D.2)
All figures given in the text were obtained from Tables B-l
and B-2, with the exception of the number of asthmatic
individuals exposed 1 or more times. For the utility study, this
number was calculated from the number of asthmatic individuals
exposed listed in Table B-2, which does not contain an
approximately 5% correction described in Rosenbaum et al. (1992,
p. 2). When this correction is applied to the Number of
Asthmatics Exposed IX or more times, 46, 000 (the actual load
figures) becomes approximately 43,700. This number, when divided
by the total asthmatic population listed as in the vicinity of
utilities (3,896,000), yields 1.12% of the asthmatic population
in the vicinity being exposed l or more times, which is precisely
the figure given in Table B-l, which is taken from Rosenbaum et
al. (1992).
For the non-utility analysis, to obtain the percentage of
the asthmatic population exposed l or more times, the range of
numbers in the column listed "Expected No. Asthmatics Exposed At
Least Once Per Year" in Table B-2 is divided by the asthmatic
population column. For example, the 2,000 - 22,000 estimated
-------
1-15
exposures for sulfite paper mills divided by the asthmatic
population of 74,000, leads to estimates of 2.7 to 29.7 percent
of the asthmatic population being exposed 1 or more times.
For the calculations concerning the concentration of risk
within different "bins" of the utility analysis, bin Base 3A can
be seen to account for roughly 75% of the total utility
exposures, depending on tha scenario chosen (Table 3 of
Rosenbaum et al., 1992)• Table 3—3 of Burton et al., 1987 shows
that this bin accounts for 64 out of the 1034 (726 + 308) total
utility point sources considered.
For the non-utility analysis of industrial boilers, Table 2-
14 of Stoeckenius et al. (1990) indicates that 3 bins, E-7, E-10,
and E-12 (Table 2-14), contribute more than half of the total
exposures from the industrial boilers that were analyzed.
-------
Table B~l. Summary of Estimates of Expected Number of Exposures of Exercising Asthmatics
to Elevated 5-minute Average S02 Concentrations for Utilities
S02 CONCENTRATION
THRESHOLD
EMISSION RATE ESTIMATES
0.5 ppm
0.75 ppm
19871
68,335
(1.12)*
Not analyzed
Title IV2
39,587
(0.65)*
8,970
(0.15)*
Lowest of:
Title IV
compliance with current stds.,
compliance with 5-min 5xx std.
of 0.7 5 ppm
24,745
(0.41)*
3 ,903
(0.06)*
Lowest of:
Title IV
compliance with current stds.,
compliance with 5-min lxx std.
of 0.75 ppm
19,006
(0.31)*
2, 571
(0.04)*
* Percentage of asthmatics in vicinity exposed 1 or more times.
'Burton et al. (1987); updated in Rosenbaum et al. (1992).
2Based on acid rain Regulatory Impact Analysis.
(Derived from Rosenbaum et al., 1992)
-------
TABLE B-2. Non-Utility Source SO, Exposure Analysis Results
Source
Category
Number of
Sources
Total
Emissions
(103 tons/yr)
Total
Population
(thousands)
Asthmatic
Population
(thousands)2
Expected
Exposure
Events/Yr
Expected
Exposures Per
100
Asthmatics
Expected No.
Asthmatics
Exposed at
Least Once
Per Year5
Industrial
Boilers'
3,108
1,725
48,7026
2,0286
56, 0003-
201,000''
2.8-9.9
12,000-42,000
Petroleum
Refineries
187
639
35,208
1,457
27,0007
1.8
6,000-17,000
Pulp/Paper
Mills:
Sulfite
Kraft
23
118
60
402
2,230
9,110
74
340
10,000"-
35,000''
5, 0008-
18,0004
13-47
1.5-5.3
2,000-22,000
1,000-12,000
Copper
Smelters
5
319
469
18
2,000-'5
5,000®
11.1-28
400-3,000
Sulfuric Acid
Plants'3
74
152
27,418
990
6,000'5
-18,000*
0.61-1.8
1,000-12,000
Aluminum
Smelters
21
96
3,042
127
8,000-15
22,000®
6.3-17
2,000-14,000
Utility
Boilers10
2,700
16,524
98,793
3,896
72,000"-
125,000i:
1.9-3.2
46,000-80,000
TOTAL
6,236
19,917
224,976
8,930
186,000-
451,000
2.0"-5.014
70,400-
202,000
Footnotes reprinted in Appendix B.
Exposure estimates for utility boilers do not reflect the 5 percent downward adjustment reported in Rosenbaum et al. (1992).
(Source: Stoeckenius et al., 1990)
-------
B-18
Footnotes to Table B-2
Includes coal and oil fired industrial, commercial and
institutional boilers.
Estimated from regional and metropolitan area asthmatic
prevalence rates as reported in the 1983 National Health
Interview Survey (NHIS, 1985).
Lower bound derived by assuming building downwash effects
are negligible and that peak-to-mean concentration ratios
are similar to those observed at the Scottish Rites monitor
near downtown Billings, MT (site in urban area not directly
influenced by any single major source). This value is based
on extrapolation of sensitivity analysis results for one
dispersion prototype (Bin E) to all other prototypes.
Based on stack height/building height = 1.5 assumption and
j i c? a i ac rtHavar"^o,K'ii <2^ "i ¦F 'f-Sill i c a 1 a
U Jjj (¦* y £ EJtSE d IV Hits! Gl 11 £ C1 L X v «¦* m 11 ~ X» cl w L c X. X13 L JL w v JL v Ct JL JL» f «L >3 U JL ~ L c. U
point sources as observed by Thrall et al. (1982).
Derived from expected exposure event results by assuming
that ratio of number of individuals exposed one or more
times per year to the number of exposure events is equal to;
0.21 for lower bound, based on sensitivity results for one
industrial boiler prototype bin (Bin E); 0.64 for upper
bound (except 0.21 used upper bound of industrial boiler
category) based on calculations performed for the UARG
utility boiler analysis. Lower bound for utility boilers is
also based on 0.64. As pointed out in the UARG analysis,
not all exposed individuals will experience the same health
effect.
Individuals living within the vicinity of more than one
source represented by different prototypes are counted once
for each prototype. Thus,this total overestimates the
actual number of people 1iving within the vicinity of one or
more boilers.
Assumes continuous operation at an emission rate equal to
the annual average. Thus, may underestimate actual
exposures resulting from periods of operation at elevated
emissions. Exposures are based on use of peak-to-mean
ratios developed by Thrall et al. (1982).
Based on extrapolation of sensitivity of industrial boiler
exposure estimates to building downwash and peak-to—mean
ratios as in (3) above to this source category.
Based on prototype source/building configurations and on
peak-to-mean ratio distribution characteristic of tall,
isolated point sources as observed by Thrall et al. (1982).
-------
B-19
10- Includes all coal- and oil-fired utility boilers greater
than 25 MW. For additional information concerning these
results, consult the UARG S02 exposure analysis (Burton et
al., 1987).
11. Result from UARG S02 exposure analysis adjusted to account
for revised population activity profile. Based on estimates
of actual plant load.
12. Based on comparison of exposures calculated under actual
load vs. constant, full-load operation for three prototype
plants.
13.' Does not include plants associated with refineries (these
are incorporated into the refinery category estimates).
14. Population weighted average.
15. As in (8) above but without adjustment for building
downwash.
-------
REFERENCES
Armstrong, B. (1985) Peak/Mean Ratios for Phelps Dodge Douglas,
Inter-Office Memorandum thru David 0. chelgren, Manager,
Compliance Unit, Arizona Department of Health Services,
December 4, 1985.
Armstrong, B, (1986) Peak/Mean Ratios for Magma, Inter-Office
Memorandum thru David 0. Chelgren, Manager, Compliance Unit,
Arizona Department of Health Services, January 20, 1986.
Ayres, J. G. (1986) Trends in asthma and hay fever in general
practice in the United Kingdom 1976-83. Thorax 41: 111-116.
Balraes, J.R, ; Fine, J.M.; Sheppard, D. (1987) Symptomatic
bronchoconstriction after short-term inhalation of sulfur
dioxide. Am. Rev. Respir. Dis. 136: 1117-1121.
Burton, C.S.; Thrall, A.D. (1982) A Brief Review of Peak to One-
Hour Average Ratios Observed in the Vicinity of Power
Plants. Systems Applications, Inc., San-Rafael, CA.
Burton, C.S.; Stockenius, T.E.; Stocking, T.S.; Carr, E.L.;
Austin, B.S.; Roberson, R.L. (1987) Assessment of exposures
of exercising asthmatics to short-term S02 levels as a
result of emissions from U.S. fossil-fueled power plants.
Systems Applications, Inc., San Rafael, CA. Pub. No.
87/176, September 23, 1987.
Burton, C.S.; Stoeckenius, T.E. (1988) Uncertainties associated
with use of peak-to-mean ratios to estimate peak 5-minute
concentrations in the vicinity of power plants. Systems
Applications, Inc., San Rafael, CA, Pub. No. 88/191,
November 29, 1988.
Carr, W.; Zeitel, L.; Weiss, K. (1992) Variations in asthma
hospitalizations and deaths in New York City. Am. J. Public
Health 82: 59-65.
Cohen, J. et al. (1983) Analysis of S02 Controlled Human Data.
Memorandum to John Haines, Ambient Standards Branch, Office
of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, N.C., September 6, 1983.
Eaton, W. C.; Parkes, C.D.; Rickman, E.E. (1991) Evaluation of
the adequacy of current S02 measurement methods to measure
5-minute average concentrations. Research Triangle
Institute. Pub. No. 4699-061-01F. February 1991.
Edmunds, A.T.; Tooley, M.; Godfrey, S. (1978) The refractory
period afater exercise-induced asthma: its duration and
relation to the severity of exercise. Am. Rev. Resp. Dis.
117:247-254.
-------
2
EPA (U.S. Environmental Protection Agency) (1982a) Review of the
National Ambient Air Quality Standards for sulfur Oxides:
Assessment of scientific and Technical Information (OAQPS
Staff Paper) Office of Air Quality Planning and Standards,
Research Triangle Park, N.C. EPA-4 50/5-82-0 07. Available
from NTIS, Springfield, VA; PB 84-102920.
EPA (U.S. Environmental Protection Agency] (1982b) Air Quality
Criteria for Particulate Matter and Sulfur Oxides.
Environmental Criteria and Assessment Office. Research
Triangle Park, N.C. EPA-600/8-82-029. Available from NTIS,
Springfield, VA,* PB 54-156801/REB,
EPA (U.S. Environmental Protection Agency) (1986a) Review of the
National Ambient Air Quality Standards for Particulate
Matter; Updated Assessment of Scientific and Technical
Information (Addendum to the 1982 OAQPS Staff Paper).
Office of Air Quality Planning and Standards. Research
Triangle Park, N.C., December 1986.
EPA (U.S. Environmental Protection Agency) (1986b) Second
Addendum to Air Quality Criteria for Particulate Matter and
Sulfur Oxides (1982): Assessment of Newly Available Health
Effects Information. Environmental Criteria and Assessment
Office, Office of Research and Development, Research
Triangle Park, N.C. EPA-6 00/8-8 6-02 0F.
EPA (U.S. Environmental Protection Agency) (1986c) Staff
Assessment of Recent Comments on the Short-term Sulfur
Dioxide Exposure Analysis, Strategies and Air Standards
Division, Office of Air Quality Planning and Standards,
Research Triangle Park, N.C., July 1986.
EPA (U.S. Environmental Protection Agency) (1986d) An Analysis
of Short-term Sulfur Dioxide Population Exposures in the
Vicinity of Utility Power Plants. Strategies and Air
Standards Division, Office of Air Quality Planning and
Standards, Research Triangle Park, N.C., September 12, 1986.
EPA (U.S. Environmental Protection Agency) (1994) 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 Asthmatics. Environmental Criteria and Assessment
Office. Research Triangle Park, N.C.
Ivans, R., III; Mullally, D. I.; Wilson, R. W.; Gergen, P. J.;
Rosenberg, H. M. ; Graurr.an, J. S. ? Chevarley, F. M.;
Feinleib, M. (1987) National trends in the morbidity and
mortality of asthma in the US. Prevalence, hospitalization
and death from asthma over two decades: 1965-1984. Chest
91(suppl.): 65S-74S.
-------
3
Hackney, J. D.; Linn, W. s.; Bailey, R. M.; Spier, C. E.;
Valencia, L. M. (1984) Time course of exercise-induced
bronchoconstriction in asthmatics exposed to sulfur dioxide.
Environ. Res. 34: 321-327.
Horstman, D. H.; Roger, L. J.; Kehrl, H. R.; Hazucha, M. J.
(1986) Airway sensitivity of asthmatics to sulfur dioxide.
Toxicol. Ind. Health. 2:289-298.
Horstman, D. H.; Seal, E., Jr.; Folinsbee, L.J.; Ives, P.; Roger,
L.J. (1988) The relationship between exposure duration and
sulfur dioxide-induced bronchoconstriction in asthmatic
subjects. Am. Ind. Hyg. Assoc. J. 49: 38-47.
Irwin, J.S.; Rao, S.T.; Petersen, W.B.; Turner, D.B. (1987)
Relating error bounds for maximum concentration estimates to
diffusion meteorology uncertainty. Atmos. Envir. 21*. 1927-
1937.
Johnson, T.; Capel, J.; McCoy, M. (1993) Estimation of ozone
exposures experienced by urban residents using a
probabilistic version of NEM and 1990 population data. U.S.
Environmental Protection Agency, Research Triangle Park, NC.
Jorres, R.; Magnussen, H. (1990) Airways response of asthmatics
after a 30 min exposure, at resting ventilation, to 0.25 ppm
N02 or 0.5 ppm S02. Eur. Respir. J. 3: 132-137.
Kehrl, H.R.; Roger, L.J.; Hazucha, M.J.; Horstman, D.H. (1987)
Differing response of asthmatics to sulfur dioxide exposure
with continuous and intermittent exercise. Am. Rev. Respir.
Dis. 135: 350-355.
Koenig, J. Q.; Marshall, S. G.; Horike, M.; Shapiro, G. G.;
Furukawa, C. T.; Bierman, C. W.; Pierson, W. E. (1987) The
effects of albuterol on sulfur dioxide-induced
bronchoconstriction in allergic adolescents. J. Allergy
Clin. Immunol. 79: 54-58.
Koenig, J. Q.; Marshall, S. G.; van Belle, G.; McManus, M. S.;
Bierman, C. W.; Shapiro, G. G.; Furukawa, C. T.; Pierson, w.
E. (1988a) Therapeutic range cromolyn dose-response
inhibition and complete obliteration of S02-induced
bronchoconstriction in atopic adolescents. J. Allergy Clin.
Immunol. 81: 897-901.
Koenig, J. Q.; Mc Manus, M. S.; Bierman, C. W. ,* Shapiro, C. G.;
Furukawa, C. T.) Pierson, W. E. (1988b) Chlorpheniramine-
sulfur dioxide interactions on lung and nasal function in
allergic adolescents. Ped. Asthma Allergy Immunol. 2:
199205.
-------
4
Koenig, J. Q.; Covert, D.S.; Hanley, Q.S.; van Belle, G.;
Pierson, W.E. (1990) Prior exposure to ozone potentiates
subsequent response to sulfur dioxide in adolescent
asthmatic subjects. Am. Rev. Respir. Dis. 141: 377-380.
Larsen, R.I. (1968) A New Mathematical Model of Air Pollution
Concentration Averaging Time and Frequency (Presented at
61st Annual Meeting of the Air Pollution Control
Association. St. Paul, Minnesota.
Lebowitz, M. D.; Holberg, C. J.; Boyer, B.; Hayes, C. (1985)
Respiratory symptoms and peak flow associated with indoor
and outdoor air pollutants in the southwest. J. Air Pollut.
Control Assoc. 35:.. 1154-1158.
Linn, W. S.; Avol, I. L.; Shamoo, D. A.; Venet, T. G.; Anderson,
K. R.; Whynot, J. D.,* Hackney, J. D. (1984) Asthmatics'
responses to 6-hr sulfur dioxide exposures on two successive
days. Arch. Environ. Health 39: 313-319.
Linn, W.S.; Avol, E.L.; Peng, R.C.; Shamoo, D.A.; Hakney, J.D.
(1987) Replicated dose-response study of sulfur dioxide
effects in normal, atopic, and asthmatic volunteers. Am.
Rev. Respir. Dis. 136: 1127-1134.
Linn, W.S.; Avol, E.L.; Shamoo, D.A.; Peng, R.C.; Spier, C.E.;
Smith, M.N.; Hackney, J.D. (1988) Effect of metaproterenol
sulfate on mild asthmatics' response to sulfur dioxide
exposure and exercise. Arch. Environ. Health 43: 399-406.
Linn, W.S.; Shamoo, D.A.; Peng, R.C.; Clark, K.W.; Avol. E.L.;
Hackney, J.D. (1990) Responses to sulfur dioxide and
exercise by medication-dependent asthmatics: effect of
varying medication levels. Arch. Environ. Health 45: 24-30.
Linn, W. S. (1991) Short-term patterns of activity and
respiratory status in adult asthmatics: their relationship
to health risks from community air pollution. Electric
Power Research Institute, Project RP3215-01, Los Amigos
Research and Education Institute, Inc., Downey, California.
Marks, G,; Mellis, C.; Peat, J.; Woolcock, A. (1992) 0 agonist
usuage among adults with asthma in an Australian provincial
town [abstract]. Am. Rev. Respir. Dis. 145: A693.
McManus, M.S.; Koenig, J.Q.; Altman, L.C.; Pierson, W.E. (1989)
Pulmonary effects of sulfur dioxide exposure and ipratropium
bromide pretreatment in adults with nonallergic asthma. J.
Allergy Clin. Immunol. 83: 619-626.
-------
5
Moore, G.E.; Stoeckenius, T.£. ; Hanna, S.; Strimitas, D. ,* (1988)
Air Quality Simulation Model Performance for One-Hour
Averages. Sigma Research Corp., Lexington, MA, Pub. No.
SYSAPP-88/003, May 23, 1988.
National Institutes of Health (1991) Guidelines for the
Diagnosis and Management of Asthma. National Asthma
Education Program: Expert Panel Report. Pub. No. 91-3042.
Neville, R. G.; Clark, R. C.; Hoskins, G.; Smith, B. (1993)
Nationa1 asthma attack audit 1991-2. Br. Med. J. 306:
559-562.
NCHS (1993) Prevalence of Selected Chronic Conditions: United
States, 1986-88. National Health Survey, 1993. U.S.
National Center for Health Statistics, Hyattsville,
Maryland, February, 1993:DHHS Publication No. (PHS) 93-1510.
NHIS (1985) National Health Interview Survey. 1983. U.S.
National Center for Health Statistics, Hyattsville,
Maryland, October 1985.
Niinimaa, V. Cole, P.; Mint2, S.; Shephard, R. J. (1980) The
switching point from nasal to oronasal breathing. Respir.
Physiol. 42: 61-71
Partridge, M. R. (1992) Education and compliance. In: Barnes, P.
J.; Rodger, I. W.; Thomson, N.C., eds. Asthma: basic
mechanisms and clinical management. New York, NY: Academic
Press; pp. 723-727.
Roger, L. J.; Kehrl, H. R.; Hazucha, M.; Horstman, D. H. (1985)
Bronchoconstriction in asthmatics exposed to sulfur dioxide
during repeated exercise. J. Appl. Physiol. 59: 784-791.
Rosenbaum, A.S.; Hudischewskyj, A.B.; Roberson, R.L.; Burton,
C.S. (1992) Estimates of Future Exposures of Exercising
Asthmatics to Short-term Elevated S02 Concentrations
Resulting from Emissions of U.S. Fossil-fueled Power Plants:
Effects of the 1990 Amendments to the Clean Air Act and a 5-
minute Average Ambient S07 Standard. Pub. No. SYSAPP-92/016,
April 23, 1992.
-------
6
Rote, D.M.; Lee, J.L. (1983) Discussion and Interpretation of
Results of the Analysis of the St. Louis S02 Data: Final
Report, Energy and Environmental Systems Division, Argonne
National- Laboratory, Argonne, IL.
Roth Associates, Inc. (1988) A study of activity patterns among
a group of Los Angeles asthmatics. EPRI Research Project
940-5, November 1988.
Roth Associates, Inc. (1992) A Survey of Daily Asthmatic
Activity Patterns in Cincinnati. EPRI Research Project 940-
05, November, 1992.
Rubinstein, I. } Bigby, B. G. 'r Reiss, T. F.; Boushey, H. A., Jr.
(1990) Short-term exposure to 0.3 ppm nitrogen dioxide does
not potentiate airway responsiveness to sulfur dioxide in
asthmatic subjects. Am. Rev. Respir. Dis. 141:381-385.
Saibene, F.; Mognoni, P.; Lafortuna, L.? Mostardi, R. (1978)
Oronasal breathing during exercise. Pflugers Archiv. 378:
65-69
Schachter, 1. N.; Witek, T. J., Jr. ; Beck, G. J.f* Hosein, H. R.;
Colice, G.; Leaderer, B. P.; Cain, W. (1984) Airway effects
of low concentrations of sulfur dioxide: dose-response
characteristics. Arch. Environ. Health 39: 34-42.
Schwartz, J.; Gold, D.; Dockery, D. W.; Weiss, S. T.; Speizer, F.
E, (1990) Predictors of asthma and persistent wheeze in a
national sample of children in the United States:
association with social class, perinatal events, and race.
Am. Rev. Respir. Dis. 142; 555-562.
Sly, M. D. (1988) Mortality from asthma in children 1979-1984.
Ann. Allergy 60: 433-443.
Smith, E. (1993) Analysis of Short-Term Air Quality Data.
Memorandum to John Haines, Ambient Standards Branch, Office
of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC. July 6, 1993. Docket # A-84-25, IV-B-6.
Smith, N. A.; Seale, J. P.; Shaw, J. (1984) Medication
compliance in children with asthma. Aust. Paediatr. J. 20:
47-51.
Smith, N. A.; Seale, J. P.; Ley, p.; Shaw, J.,* Braes, P. U.
(1986) Effects of intervention on medication compliance in
children with asthma. Med. J. Aust. 144: 119-122.
-------
7
Stoeckenius, T.E.; Garelick, B.; Austin, 8.S.; O'Connor, K.;
Pehling, J.R. (1990) Estimates of Nationwide Asthmatic
Exposures to Short—term Sulfur Dioxide Concentrations in the
Vicinity of Non-Utility Sources. Systems Applications Inc.,
San Rafael, CA, Pub. No. SYSAPP-90/129, December 6, 1990.
Stone, S.L. (1994) Analysis of Short-Term Air Quality Data.
Memorandum to John Haines, Ambient Standards Branch, Office
of Air Quality Planning and Standards, U.S. EPA, Research
Triangle Park, NC. March 15, 1994. Docket # A-84-25,
IV-B-10.
Thomas, B.C. (1987) U.S. EPA, Strategies and Air Standards
Division. Memorandum to John Haines, Strategies and Air
Standards Division. September 17, 1987. Docket # A-84-25,
II-B-4.
Thrall, A.D.; Langstaff, J.E.; Liu, M.K.; Burton, C.S. (1982) On
the Variability in Peak Sulfur Dioxide Concentrations
Contained in Longer Term Averages. An Empirical Study of an
Isolated Power Plant. System Applications, Inc., San
Rafael, CA, Pub. No. 82262, October, 1982.
Van Essen-Zandvliet, E. E.; Hughes, M. D.; Waaklens, H. J.;
Duiverman, E. J.; Pocock, S. J.; Kerrebijn, K.F. (1992)
Effects of 22 months of treatment with inhaled
corticosteroids and/or beta-2-agonists on lung function,
airway responsiveness, and symptoms in children with asthma.
Am. Rev. Respir. Dis. 146; 547-554.
Weinstein, A. G.; Cuskey, W. (1985) Theophylline compliance in
asthmatic children. Ann. Allergy 54: 19-24.
Weiss, K. B.; Wagener, D. K. (1990) Changing patterns of asthma
mortality: identifying target populations at high risk. JAMA
J. Am, Med. Assoc. 264: 1683-1687.
-------
TECHNICAL REPORT DATA
Please read Instructions on reverse before completing}
1. REPORT NO. 2.
EPA-452/R-94-013
3. RECIPIENT'S ACCESSION NO.
4. TTTLEAND SUBTITLE
Review of the National Ambient Air Quality Standards for Sulfur
Oxides: Updated Assessment of Scientific and Technical
Information. Supplement to the 1986 OAQPS Staff Paper
Addendum
5. REPORT DATE
September 1994
6. PERFORMING ORGANIZATION CODE
7. authqr(s) Eric G. Smith
John H. Haines
Susan Lyon Stone
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Ambient Standards Branch (MD-12)
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11, CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF RETORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY notes
is, abstract This paper presents a summary of the evaluation and interpretation of key new studies on the health effects
associated with short-term sulfur dioxide (SO,) exposures and also updates available information on the occurrence of short-term (5-
minute) peaks of S02 in the ambient air and on the likelihood that the at-risk population will be exposed. This staff paper
supplement is intended to help bridge the gap between the scientific review of recent health effects information contained in the
1994 SO, criteria document addendum supplement and the judgments required of the Administrator in determining whether new
regulatory initiatives are needed to provide increased protection to asthmatic individuals whose health could be compromised if
exposed to high 5- to 10 minute peak S02 levels. Factors relevant to this evaluation, as well as staff conclusions and
recommendations on alternative regulatory approaches are presented in this paper.
The staff recommends that the Administrator consider three possible regulatory alternatives: 1) establish a new 5-minute
NAAQS in the range of 0.6 to 1.0 ppin S02 expressed as the maximum 5-minute block average in 1 hour, with 1 to 5 expected
exceedances; 2) establish a new regulatory program under section 303 of the Clean Air Act, with a target level in the range of 0.6
to 1.0 ppm SOj expressed as the maximum 5-ininute block average in 1 hour, with 1 to 5 expected exceedances; and 3) retain the
existing suite of standards but augment their implementation by focusing on those sources likely to produce high 5-minute peaks of
SOj,
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Sulfur Oxides
Air Quality Standards
Sulfur Dioxide
Air Pollution
IS. DISTRIBUTION STATEMENT
19. SECURITY CLASS (Report)
21. NO. OF PACES
Release Unlimited
Unclassified
99
2U. SECURITY CLASS (Page)
22. PRICE
Unclassified
iPA Form 2220-1 (Rev. 4*7t> PREVIOUS EDITION IS OBSOLETE
------- |