450582007
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/5-82-007
November 1982
Air
P/EPA
Review of the National
Ambient Air Quality
Standards for Sulfur Oxides:
Assessment of Scientific
and Technical Information
OAQPS Staff Paper
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S.Environmental Protection Agency
Research Triangle Park, N.C. 27711
November 1982
0*1-V 4
-------
-------
The cover illustration shows two lateral views, represented schematically and by
fluoroscope (x-ray), of the oral cavity during two breathing patterns. The top figures
depict the mouth during typical oronasal breathing with a small air passage between
the palate (P) and tongue (T). In this situation, much of the inhaled air enters through the
the nose. The bottom figures depict hyperventilation, characterized by rapid oronasal
breathing and an enlarged air space between the palate and tongue. Because the nose
is very efficient in removing S02 , the relative penetration of S02 to more sensitive
regions of the respiratory tract is largely dependent on the extent and character of the
oral component of breathing under various conditions. This is an important
consideration in evaluating controlled human exposure studies to SO 2 . which employ
various types of breathing and levels of exercise.
• Illustration courtesy of Dr. Donald Proctor.
This report has been reviewed by the Office of Air
Quality Planning and Standards, EPA, and approved
for publication. Mention of trade names or
commercial products Is not Intended to constitute
endorsement or recommendation for use.
-------
REVIEW OF THE NATIONAL AMBIENT AIR QUALITY STANDARDS FOR SULFUR OXIDES:
ASSESSMENT OF SCIENTIFIC AND TECHNICAL INFORMATION
OAQPS STAFF PAPER
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C.. 27711
November, 1982
-------
ii
ACKNOWLEDGEMENTS
This staff paper is the product of the Office of Air Quality Planning
and Standards (OAQPS). The principal authors include John Bachmann,
Jeff Cohen, Larry Zaragoza, and John Haines. The report includes comments
from OAQPS, the Office of Research and Development, Office of Planning and
Resource Management, and the Office of General Counsel within EPA and
was formally reviewed by the Clean Air Scientific Advisory Committee.
Helpful comments and suggestions were also submitted by a number of
independent scientists, by officials from the state agencies of
California and Maine, by the Department of the Interior, and the
Tennessee Valley Authority, and by environmental and industrial groups
including the National Resources Defense Council, the National Audubon
Society, the Environmental Defense Fund, the Non-Ferrous Smelter
Companies, the American Petroleum Institute, and the Utility Air
Regulatory Group.
-------
ill
TABLE OF CONTENTS
Page
List of Figures v
List of Tables vi
Executi ve Summary vi -j
I. Purpose i
II. Background f 1
III. Approach 3
IV. Air Quality Considerations 6
A. Chemical and Physical Characteristics 6
B. Ambient Concentrations 7
V. Critical Elements in the Review of the Primary Standards 14
A. Mechanisms 14
B. Effects of Concern 16
C. Sensitive Population Groups 29
D. Concentration/Response Information 34
VI. Factors to be Considered in Selecting Primary Standards
for Sulfur Oxides 53
A. Pollutant Indicator, Averaging Times, and Form of
the Standards 53
B. Level of the Standard 60
C. Summary of Staff Conclusions and Recommendations 83
VII. Critical Elements in the Review of the Secondary Standard 88
A. Vegetation Damage 88
B. Materials Damage 115
C. .Personal Comfort and Well-Being 124
-------
iv
Page
D. Acidic Deposition 125
E. Summary of Staff Conclusions and Recommendations 126
Appendix A. Factors that Influence Penetration and Deposition
of S02 A-l
A. Inhalation Patterns A-l
B. Absorption by Particles A-5
C. Mechanisms of Toxicity A-7
Appendix 8. Evaluation of Evidence for Effects on
Respiratory Mechanics and Symptoms B-l
A. Controlled Human Exposure Studies of S02 Alone B-l
B. Long-term Exposures to S02 Alone B-6
C. SOg in Combination with Laboratory Particles,
Other Gases B-6
D. Community Air Pollution B-10
Appendix C. Pulmonary Function Tests Used in Controlled
Human Exposure Studies of $03 C-l
A. Introduction C-l
B. Glossary of Terms C-2
Appendix D. Analysis of Alternative Averaging Times and
Exposure D-l
A. Analysis of SOg Air Quality Data D-l
B. Modeling Analysis D-7
C. Exposure Analysis D-ll
Appendix E. CASAC Closure Memorandum , E-l
References
-------
LIST OF'FIGURES
Number Title Page
4-1 Distribution of the Typical Hour of Maximum SOg among 153
Monitoring Locations, 1975-1978 12
4-2 Monthly Arithmetic Mean SC>2 Concentrations at Nonurban Sites
in the Northeast 12
6-1 Factors Needed to Render a Multiple-Exceedance Standard
Equivalent to a Single Exceedance Standard 59
6-2 Proportion of Sites Recording a Given Annual Average S02
Concentration, with 2 or more Short-Term S02 Concentrations
Exceeding 0.5 ppm for 1-Hour or 0.14 ppm for 24 Hours 80
6-3 Proportion of Sites Recording a Given Second Maximum 24-Hour
Average Concentration, with 2 or more 1-Hour S02 Concentrations
Exceeding 0.5 ppm 80
7-1 Concentration-Response Relationships for the Minimum Concen-
tration of S02 required to produce Foliar Injury 102
7-2 Corrosion Rate for Roofing and Siding (Large Sheets) 122
7-3 Corrosion Rate for Wire Fencing 122
D-l Basic Relationships Between the Second Highest 1-Hour Value
per Year and the Second Highest 24-Hour Value D-4
D-2 Location and Number of Expected Exceedances per Year of a
1-Hour Average Value of 0.5 ppm for a 1000 MW Power Plant
just Complying with a) the Current 24-Hour S02 Standard and
b) the Current 3-Hour S02 Standard D-8
D-3 Maximum 5 Concentrations for a 1000 MW Power Plant just
Complying with a) the Current 24-Hour S02 Standard and
b) the Current 3-Hour S02 Standard D-10
-------
vi
LIST OF TABLES
Number Title Page
1 Staff Assessment of Key Controlled Human Exposure Studies ...... xii
2 Staff Assessment of Short -Term Epidemiological Studies ......... xiv
4-la Characterization of S02 Levels (in ppm) at Population-
Oriented Monitoring Sites for Different Averaging Times ....... 9
4-lb Approximate Conversion of ppm to pg/m^ ........................ 9
4-2 Characterization of S02 Levels (in ppm) at Source-
Oriented Monitoring Sites for Different Averaging
Ti mes [[[ 11
• 5-1 Effects of S02 on Respiratory Mechanics and Symptoms ......... . 20
5-2 Major Sensitive Population Subgroups .............. ' ............ 30
5-3 Comparison of Responses in Normals, Atopies, and Asthmatics
to S02 [[[ 32
5-4 Selected Controlled Human Studies of Quantitative
Interest ..................... ................................. 38
5-5 Summary of Epidemiological Studies Providing Most Useful
Concentration/Response Information for Short-Term S02
Exposures [[[ 43
6-1 Staff Evaluation of Key Controlled Human Exposure Studies ..... 62
6-2 Triggers of Asthmatic Attacks ................................. 67
6-3 Staff Assessment of Short -Term Epidemiological Studies ........ 72
7-1 Controlled S02 Exposure Studies ............................... 93
7-2 Effects of Single 4-Hour S02 Exposure Mixtures on Foliar
Injury [[[ 95
7-3 Controlled Exposure Studies of Long-Term Exposures to
Pollutant Mixtures ............................................ 98
7-4 Field Studies of Chronic Ambient S02 Exposures ................ 106
7-5 Staff Assessment of Short-Term Vegetation Studies ............. Ill
A-l Oronasal Distribution of Inspired Air ......................... A-4
D-l Current Status of Second High 1-Hour Values Versus Second
-------
vi i
EXECUTIVE SUMMARY
This paper evaluates and interprets the available scientific and
technical information that the EPA staff believes is most relevant to
the review of primary (health) and secondary (welfare) National Ambient
Air Quality Standards (NAAQS) for sulfur oxides and presents staff
recommendations on alternative approaches to revising the standards.*
Review of the NAAQS is a periodic process instituted to ensure the
scientific adequacy of air quality standards and is required by Section
109 of the 1977 Clean Air Amendments. The assessment in this staff
paper is intended to help bridge the gap between the scientific review
contained in the EPA criteria document "Air Quality Criteria for
Particulate Matter and Sulfur Oxides" and the judgments required of the
Administrator in setting ambient standards for sulfur oxides. The staff
paper is, therefore, an important element in the standards review
process and provides an opportunity for public comment on proposed staff
recommendations before they are presented to the Administrator. The
focus of this paper is on sulfur dioxide (S02), alone and in combination
with other pollutants.
S02 is a rapidly diffusing reactive gas that i? quite soluble in
water. It is emitted principally from combustion or processing of
sulfur containing fossil fuels and ores. S02 occurs in the atmosphere
with a variety of particles and other gases, and undergoes chemical and
*The current standards for sulfur dioxide (SC>2) are: primary, 0.03 ppm
(80 Mg/m3) annual arithmetic mean and 0.14 ppm (365 yg/m3) 24-hour
average not to be exceeded more than once per year; and, secondary, 0.5
ppm (1300 ug/m3) 3-hour average not to be exceeded more than once per
year.
-------
VI11
physical interactions with them forming sulfates and other transformation
products.
At elevated concentrations, S02 can adversely affect human health,
vegetation, materials, economic values, and personal comfort and well-
being. S02 and its transformation products also are a major component
of pollution related acidic deposition and visibility degradation. Long-
term average SOg levels range from less than 0.004 ppm in remote rural
sites to over 0.03 ppm in the most polluted urban industrial areas. The
highest short-term values are found in the vicinity (< 20 km) of major
point sources. At such sites, maximum short-term levels for 24-hour, 3-
hour, and 1-hour averages can exceed 0.4 ppm, 1.4 ppm, and 2.3 ppm,
respectively.
Primary Standards
The staff has reviewed scientific and technical information on the
known and potential health effects of S02 cited in the criteria document.
The information includes studies of mechanisms of toxicity, effects of
high exposures to SOg, alone and in combination with particles and pollutant
gases, in controlled human and animal studies; epidemiological studies of
community air pollution; and air quality information. Based on this
review, the staff derives the following conclusions:
1) Due to its high solubility, $03 is readily removed in the moist
surfaces of the nose and other respiratory passages. With quiescent
nasal breathing, almost all inhaled SOg is removed in the
extrathoracic (head) region. This limits the potential for effects
on the more sensitive thoracic regions of the respiratory tract.
Factors that can increase penetration of S02 to these regions
-------
ix
include mouth and oronasal breathing, increased ventilation rates,
and the presence of high levels of particles or fog droplets that
may act as "carriers" for S02-
2) Although S(>2 may produce effects through several mechanisms, the
most striking acute effects observed appear to result from
stimulation of receptors in the tracheobronchial region leading to a
reflex bronchoconstriction mediated by the nervous system.
3) The major effects categories of concern associated with high
exposures to S02 include: (a) sensory and other non-respiratory
responses, (b) effects on respiratory mechanics and symptoms,
(c) aggravation of existing respiratory and cardiovascular disease,
(d) effects on clearance and other host defense mechanisms, and
(e) mortality.
4) The major subgroups of the population that appear likely to be most
sensitive to the effects of 862 include: (a) asthmatics,
(b) individuals not diagnosed as asthmatic but with atopic disorders
(e.g., allergies), and (c) individuals with chronic obstructive
pulmonary or cardiovascular disease. Other subgroups that may be
somewhat sensitive include the elderly and children.
5) Although a number of animal, controlled human, and community
epidemiological studies provide important qualitative information on
the range of possible responses to SOg, the most useful
concentration-response information comes from a few very recent
controlled human exposure studies and a limited set of
epidemiological studies reflecting British air pollution exposures
in the 1960's and early 1970's.
-------
Based on the scientific and technical review as well as policy
considerations, the staff makes the following recommendations and
conclusions with respect to primary S02 standards:
1) Laboratory studies show that peak ambient levels of SOg, acting
alone, can cause health effects in humans. Consequently, a separate
S02 standard is still appropriate. Because the range of possible
SOg pollution interactions may not be well captured by an
S02-particle index, and because separate particle and $63 standards
can be chosen with due consideration to potential interactive
effects, the additional complexities involved in specifying
combination S02/particle standards do not appear warranted in
terms of public health protection.
2) a) Retention of a standard with a 24-hour averaging time is
recommended.
b) Support for an annual standard at or near current levels is
largely qualitative. Nevertheless, because short-term standards
would not prevent increases in annual mean concentrations in some
heavily populated urban areas, consideration should be given to
retention of a primary annual standard for S02.
c) Based on a series of recent controlled human studies, considera-
tion of a new peak (1-hour) SC>2 standard is also recommended.
d) The 24-hour and potential annual and 1-hour standards should all
be expressed in statistical form; the decision on the allowable
number of exceedances for the 24-hour and potential 1-hour standard
should be made in conjunction with establishing levels for the
standards.
-------
XI
3) a) The staff assessment of key controlled human studies of peak
(minutes to an hour) SOg exposures is summarized in Table 1. The table
focuses on those studies involving unencumbered or free breathing with
exercise (chamber or facemask studies) and resting oral (mouthpiece)
exposures. In so doing, the staff recognizes that caution must be
used in extending the results of these laboratory exposures to ambient
conditions; attention must be paid to the differences in natural
oronasal and mouthpiece breathing, environmental conditions, and
changes in S02 concentration with time. The major effects observed
in these studies are increases in airway resistance and decreases in
other functional measures indicative of significant bronchoconstriction
in sensitive asthmatic or atopic subjects. At or above 0.5 ppm,
changes in functional measures are accompanied by perceptible symptoms
such as wheezing, shortness of breath, and coughing.
b) Based on this staff assessment, the range of 1-hour S02 levels of
interest is 0.25 to 0.75 ppm (650 to 2000 jig/m3). The lower bound
represents a 1-hour level for which the maximum 5 to 10 minute peak
exposures do not exceed 0.5 ppm, which is the lowest level where
potentially significant responses in free (oronasal) breathing asthmatics
have been reported in the published literature as of this writing. The
upper bound of the range represents concentrations at which the risk
of significant functional and symptomatic responses in exposed sensitive
asthmatics and atopies appears high. In evaluating these data in the
context of decision making on a possible 1-hour S02 standard, the following
considerations are important: (a) the significance of the observed
or anticipated responses to health, (b) the relative effect of S02
-------
TABLE 1. STAFF ASSESSMENT OF KEY CONTROLLED HUMAN EXPOSURE STUDIES
SO,
Concentration (5-60 minutes]
Observed Effects
Implications
1 ppm
0.75 ppm
0.5 ppm
0.25 ppm
Functional changes, possible
symptoms in resting asthmatics, oral
(facemask or mouthpiece) exposure.
Functional changes in free
breathing normal healthy
subjects, moderate to heavy „
exercise. No health effects.
Functional changes, symptoms in
free breathing (chamber) asthmatics,
moderate exercise.
Functional changes, symptoms
in oronasal (facemask) breathing,
asthmatics with moderate exercise
but not in asthmatics (chamber) with
light exercise.
No observed effect in free
breathing subjects.
Strong suggestion that at
this level even light exercise
for "mouth" breathing asthmatics
would result in comparable or
more marked changes.
Comparable oronasal exposures
in asthmatics, or atopies could
result in effects of significance.
Significant effects in mild
asthmatics with moderate exercise.
Lowest level of significant
response for free breathing.
Significant effects unlikely.
H
"J
Sheppard et al. (1980), Koenig et al. (1980). The second study used S0? in combination with saline
aerosol.
2Stacy et al. (1981); Bates and Hazucha (1973).
3Sheppard et al. (1981a); Linn et al. (1982a); Koenig et al. (1982a).
4Linn et al. (1982b).
5Kirkpatrick et al. (1982); Linn et al. (1982a).
-------
xiii
compared to normal day to day variations in asthmatics from exercise
and other stimuli, (c) the low probability of exposures of exercising
asthmatics to peak levels, and (d) five to ten minute peak exposures
may be a factor of two to three greater than hourly averages.
Independent of frequency of exposure considerations, the upper bound
of the range contains little or no margin of safety for exposed
sensitive individuals. However, the limited geographical areas
likely to be affected and low frequency of peak exposure to active
asthmatics if the standard is met add to the margin of safety.
c) The data do not suggest other groups that are more sensitive
than asthmatics to single peak exposures, but qualitative data
suggest tnat repeated peaks might produce effects of concern in
other sensitive individuals. Potential interactions of S02 and 03
have not been investigated in asthmatics and atopies. The qualitative
data, potential pollutant interactions, and other considerations
listed above should be considered in determining the need for and
evaluating the margin of safety provided by alternative 1-hour
standards.
4) a) The staff assessment of the short-term (24-hour) epidemiological
data is summarized in Table 2. The "effects likely" row denotes
concentration ranges derived from the criteria document at or above
which there appears greatest certainty that effects would occur.
The data do not, however, show evidence of clear population
thresholds, but suggest that effects may be possible at levels below
those listed in the "effects likely" row; the evidence and risks at
lower levels are, however, much less certain.
-------
xi v
TABLE 2. STAFF ASSESSMENT OF SHORT-TERM EPIDEMIOLOGICAL STUDIES
Effects/
Study
Effects Likely3
Effects Possible
Measured Sulfur Dioxide - ug/m3(ppm) -, 24 hour mean
Daily Mortality
in London1
500-1000(0.19-0.38)
Aggravation of
Bronchitis2
500-600(0.19-0.23)
< 500(0.19)
Combined Effects
Levels
500(0.19)
< 500(0.19)
^Deviations in daily mortality from mean levels examined in 3 studies encompassing
individual London winters of 1958-59 and 14 aggregate winters from 1958-72. Early
winters were dominated by high smoke and S02, principally from coal combustion
emissions, and with frequent fogs (Martin and Bradley, 1960; Ware et al., 1981;
Mazundar et al., 1981).
Examination of symptoms reported by bronchitics in London. Studies conducted
from the mid-1950's to the early 1970's (Lawther et al., 1970).
3CD, Table 14-8.
b) Based on this staff assessment, the range of 24-hour S02 levels
of interest are 0.14 to 0.19 ppm (365 to 500 ng/m3). Under the
conditions prevailing during the London studies (high particles,
frequent fogs, winter), the upper end of the range represents levels
at which effects may be likely according to the criteria document. The
risk of health effects should be lower when translating these results to
U.S. settings with particle levels at or below the ambient standards.
c) The uncertainties with respect to interactive effects with particles
or other pollutants and nature of effects are important margin of safety
considerations. In the absence of quantitative scientific evidence for
a well defined "no effects" level, the level of the current standard is
recommended as a lower bound. This level was previously judged to provide
an adequate margin of safety from the same effects under consideration
-------
XV
here. Qualitative data from animal toxicology, controlled human studies
and community epidemiology suggest risks of potential effects (e.g.,
slowed clearance) as well as the existence of sensitive groups (e.g.,
children) not evaluated in the more quantitative studies. These
factors, as well as potential pollutant interactions, exposure
characteristics, and whether the 24-hour standard is intended to
act as a surrogate for a 1-hour standard, should also be considered
in evaluating the margin of safety provided by 24-hour standards in
the range of 0.14 to 0.19 ppm.
5) Although the data are inconclusive and uncertain, the possibility of
effects from continuous lower level exposures to SOg cannot be ruled
out. Given the lack of epidemiological data suggesting long-term
effects of S02 at or near the levels of the current annual standard,
however, no quantitative rationale can be offered to support a
specific range of interest for an annual standard. Nevertheless,
air quality analyses summarized in Appendix D suggest that short-term
standards in the ranges recommended in this paper would not prevent
annual levels in excess of the current standard in a limited number
of heavily populated urban areas. Because of the possibility of
effects from a large increase in population exposure, consideration
should be given to maintaining a primary annual standard at or above
the level of the current standard.
6) Analysis of alternative averaging times to date suggest that while
any single standard might not be a suitable surrogate for other
averaging times, implementation of the current suite of primary and
secondary S02 standards (annual, 24-hour, 3-hour) provides substantial
protection against the direct health and welfare effects identified
-------
xvi
in the scientific literature as being associated with ground level
S02 air quality. This permits the consideration of reaffirming the
existing S02 standards as a reasonable policy option, following the
current criteria and standards review. Factors favoring such an
option include the substantial improvements likely in information on
1-hour effects over the next few years, the uncertainties in long-term
effects data, the possibility of substantial changes in S02 control
strategies prompted by regional effects, and the.practical advantages
of not requiring premature formulation and implementation of a new
S02 regulatory program associated with revising standards at this
time.
Secondary Standards
The staff examined information in the criteria document relevant to
the review of the secondary standards. Categories of welfare effects
examined include effects on vegetation, materials, personal comfort and
well-being, and acid deposition. Major staff conclusions and
recommendations are summarized below.
1) a) Damage to vegetation by S02, resulting in economic losses in
commercial crops, aesthetic damage to cultivated trees, shrubs and
other ornamentals, and reductions in productivity, species richness
and diversity in natural ecosystems, constitute effects on public
welfare in impacted areas. Such effects are associated with both
peak and short-term (minutes to hours) and long-term (weeks to
years) exposures to S02.
b) Given the available data on the acute effects of S02 on plants
(growth and yield and foliar injury), a 3-hour standard at or below
the level of the current secondary standard (0.5 ppm) may be needed
-------
xvii
to protect vegetation. If a 1-hour primary standard is chosen that
provides equivalent or better protection, then the averaging time
and level of the secondary standard can be made equal to the primary
standard. In the absence of a primary standard that provides adequate
protection for vegetation, a 3-hour secondary standard is recommended.
c) Available data on the effects of long-term S0£ exposures of
vascular plants (e.g., trees, shrubs, crops) suggest the possibility
of changes in species richness and diversity, reduced growth over
extended periods, and premature needle drop. However, these data
are weak and not developed well enough to provide the principal
basis for selecting the level of a long-term SO? standard. Existing
information, thus, cannot be used to show significant effects on
vascular plants at annual SOg levels below the current primary
annual standard, but does support the need to protect against the
effects of prolonged S02 exposure by limiting long-term S02
concentrations much above this level.
d) Current long-term S02 concentrations over large areas of the
northeast exceed levels that may be associated with effects on non-
vascular plants (e.g., lichens, mosses). Given uncertainties re-
garding the extent and importance of these potential effects on
natural ecosystems and the regional character of the exposures, the
staff recommends that the effects of SOg on non-vascular plants be
considered in the larger context of regional acid deposition -
visibility - fine particle strategies. As such, no separate long-
term secondary standard for non-vascular plants is recommended at
this time.
-------
XVTI1
2) a) Elevated long-term S02 concentrations in the presence of
moisture can damage a number of materials including: exposed
metals, paints, building materials, statuary, paper, leather, and
textiles. Control strategies have resulted in marked improvements
in long-term S02 levels over the past 13 years, which the criteria
document associates with substantial benefits. While the available
data are limited and do not permit definitive findings with respect
to the potential costs of S02 related material damage or provide
clear quantitative relationships for the full range of potentially
affected materials, they generally support the need for limiting
long-term SOg concentrations in urban areas.
b) Analysis of existing air quality data suggests that without
the primary annual standard, long-term urban air quality could
deteriorate and in a number of large urban areas might exceed the
current annual standard. Therefore, consideration should be given
to a long-term secondary SOg standard at or below the level of the
current annual primary standard of 0.03 ppm (80 ug/m3) to protect
against materials damage effects.
3) The staff concludes that a secondary S02 standard is not needed to
protect against effects on personal comfort and well-being.
4) The available scientific information indicates that the current 3-
hour and annual standards provide reasonable protection against the
direct welfare effects associated with ambient S02- In essence, the
data support maintenance of S02 standards at or below levels of the
current standards.
5) The acid deposition issue will not be addressed directly in this
review of the sulfur oxides standards.
-------
REVIEW OF THE NATIONAL AMBIENT AIR QUALITY STANDARDS FOR SULFUR OXIDES:
ASSESSMENT OF SCIENTIFIC AND TECHNICAL INFORMATION
OAQPS STAFF PAPER
I. PURPOSE
This paper evaluates and interprets the most relevant scientific
and technical information reviewed in the draft EPA document "Air
Quality Criteria for Particulate Matter and Sulfur Oxides" (EPA, 1982a)
in order to better specify the critical elements which EPA staff
believes should be considered in the possible revision of the primary
and secondary National Ambient Air Quality Standards (NAAQS) for sulfur
oxides. This assessment is intended to help bridge the gap between the
scientific review contained in the criteria document and the judgments
required of the Administrator in setting ambient standards for sulfur
oxides. As such, particular emphasis is placed on identifying those
conclusions and uncertainties in the available scientific literature
that the staff believes should be considered in selecting averaging
time, form, and level for the primary standards. While the paper
should be of use to all parties interested in the standards review,
it is written for those decision makers, scientists, and staff who
have some familiarity with the technical discussions contained in the
criteria document.
II. BACKGROUND
Since 1970 the Clean Air Act, as amended, has provided authority
and guidance for the listing of certain ambient air pollutants which may
endanger public health or welfare and the setting and revising of NAAQS
for those pollutants. Primary standards must be based on health effects
-------
criteria and provide an adequate margin of safety to ensure protection
of public health. As several recent judicial decisions have made clear,
the economic and technological feasibility of attaining primary
standards are not to be considered in setting them, although such
factors may be considered to a degree in the development of state plans
to implement the standards (D.C. Cir., 1980, 1981). Further guidance
provided in the legislative history of the Act indicates that the
standards should be set at "the maximum permissible ambient air level
. . . which will protect the health of any [sensitive] group of the
population." Also, margins of safety are to be provided such that the
standards will afford "a reasonable degree of protection . . . against
hazards which research has not yet identified" (Committee on Public
Works, 1974). In the final analysis, the EPA Administrator must make a
policy decision in setting the primary standard, based on his judgment
regarding the implications of all the health effects evidence and the
requirement that an adequate margin of safety be provided.
Secondary ambient air quality standards must be adequate to protect
the public welfare from any known or anticipated adverse effects
associated with the presence of a listed ambient air pollutant. Welfare
effects, which are defined in section 302(h) of the Act, include effects
on vegetation, visibility, water, crops, man-made materials, animals,
economic values and personal comfort and well-being. In specifying a
level or levels for secondary standards the Administrator must determine
at which point the effects become "adverse" and base his judgment on the
welfare effects criteria.
The current primary standards for sulfur oxides (to protect public
health) are 0.03 parts per million (ppm or 80 micrograms per cubic meter
-------
[yg/m3]) annual arithmetic mean, and 0.14 ppm (365 ug/m3), maximum 24-
hour concentration not to be exceeded more than once per year. The
current secondary standard for sulfur oxides (to protect public welfare)
is 0.5 ppm (1300 yg/m3), maximum 3-hour concentration, not to be
exceeded more than once per year. For both primary and secondary
standards, sulfur oxides are measured as sulfur dioxide (SOg). Thus,
S02 is the current indicator for the sulfur oxides standards.
Preliminary drafts of this paper were reviewed by the Clean Air
Scientific Advisory Committee (CASAC) in April and August, 1982. This
final product incorporates the suggestions and recommendations of the
CASAC as well as other appropriate comments received on the initial
drafts. The CASAC closure letter on the staff paper (Goldstein,
1983) is reproduced in Appendix E.
III. APPROACH
The approach used in this paper is to assess and integrate information
derived from the criteria review in the context of those critical ele-
ments which the staff believes should be considered in the review of the
primary and secondary standards. Particular attention is drawn to those
judgments that must be based on the careful interpretation of incomplete
or uncertain evidence. In such instances, the paper states the staff's
evaluation of the evidence as it relates to a specific judgment, sets
forth appropriate alternatives that should be considered, and recommends
a course of action.
Sections IV and V review and integrate important scientific and
technical information relevant to standard-setting. Because sulfur
oxides are often studied in combination with particulate matter, much of
the more important literature has already been assessed in the companion
-------
staff paper on particulate matter (EPA, 19825). Where possible,
pertinent references are made to the appendices of that paper with only
summaries in the main body of this paper. Section IV presents relevant
features of historical and current U.S. air quality to-support discussions
of both primary and secondary standards. Section V addresses the essen-
tial elements with regard to the primary standards; these include the
following:
1) identification of possible mechanisms of toxicity;
2) description of effects and judgments of critical effects
of concern for standard-setting;
3) identification of most sensitive population groups; and
4) discussion of controlled human and community studies re-
lating level (s) and duration(s) of exposure to indicators
of health effects.
Drawing from the discussion in Sections IV and V, Section VI
identifies and assesses the factors the staff believes should be
considered in selecting averaging times, form, and level of primary
standards. Preliminary staff recommendations on alternative policy
options in each of these areas are also presented.
Section VII examines information in the criteria document the staff
believes is most relevant with respect to secondary standards and
focuses on the direct effects of S02 on vegetation, man-made materials,
and personal comfort and well-being. Indirect effects on visibility and
climate and acid deposition are not discussed. The elements addressed
include:
1) description of effects and judgment of the critical
effects of concern for standard-setting;
-------
2) identification of causal mechanisms;
3) studies relating level(s) and duration(s) of exposure to
indicators of effects; and
4) factors to be considered in selecting averaging times, form,
and level of secondary standards.
Preliminary staff recommendations on policy options for secondary
standards are also presented.
The principal focus of this paper is on the effects of S02, alone
and in combination with other pollutants. Other sulfur oxide vapors
(e.g., $03) are not commonly found in the atmosphere. The effects of
the principal atmospheric transformation products of SOg (i.e., sulfuric
acid and sulfates) are discussed in the companion staff paper on particu-
late matter.
-------
IV. AIR QUALITY CONSIDERATIONS
This section summarizes the relevant chemical and physical
properties of S02 as it occurs in the ambient atmosphere and briefly
characterizes ambient S02 levels to provide perspective for subsequent
interpretation of health and welfare studies. More detailed discussion
of these areas is contained in Chapters 2 and 5 of the criteria document
("CD", EPA, 1982a); supplemental analyses of air quality relationships
are also outlined in Appendix D.
A. Chemical and Physical Characteristics
S02 is a rapidly diffusing reactive gas that is quite soluble in
water, readily dissolving to form sulfurous acid (H20'S02), a weak acid
that dissociates into hydrogen (H+), sulfite (SOj3), and bisulfite
(HS03") ions. The solubility of S02 and the ratio of sulfite to
bisulfite ions in aqueous solution increase significantly with
increasing pH (and local ammonia concentration). The sulfite/bisulfite
ratio is about 1:1000 at a pH of 7.4 (as in some body fluids) (CD, Figure 2-3).
Total dissolved S02 increases by a factor of about 104 from pH 3 to pH 7.
SOg occurs in the ambient air with a variety of particles and
other gases. The potential chemical and physical interactions among these
numerous substances are complex and incompletely characterized (CD, pp.
2-62 to 2-69). The following generalizations are of some significance
in evaluating the health and welfare effects of S02 in complex mixtures:
1) S02 is oxidized by a number of homogenous (gas-phase) and
heterogeneous (liquid or solid surface phase) mechanisms to form
sulfuric acid and, ultimately, other sulfates. S02 may also
reversibly dissolve in or attach to particles or may form stable
sulfite complexes (CD, pp. 2-38 to 2-40).
-------
2) At low relative humidities, ambient S02 may adsorb onto the
surfaces of dry participate matter, but the extent of adsorbtion is
limited by the available particle surface area. Because fine mode
(< 2.5 um) aerosols usually contain most of the available surface
area, adsorbed S02 would most likely be found in the fine mode
(CD, p. 2-71). Except in the case of particles of unusually high
surface/mass ratios, or in the presence of pollutants promoting
oxidation to sulfate, the amount of SOg adsorbed onto dry particles
may be only a small fraction of total fine particle mass (Schryer
et al., 1980).
3) At higher relative humidities, SOg will dissolve into available
droplet aerosol. S02 solubility in such droplets is greater with
lower temperature and lower droplet acidity (e.g., with elevated
ammonia levels). Unless converted to sulfates, the SOg may be
released from the droplet when it encounters lower gas phase S02
concentrations. At low ambient ammonia levels (< 1 part per billion
[ppbj), the ultimate capacity for droplet absorption of S(>2 is
limited to < 1% of droplet mass (Larson et al., 1978; Scott and
Hobbs, 1967). Because of their low pH, little interaction would be
expected between various ammonium sulfate or sulfuric acid aerosols
and S02-
B. Ambient Concentrations
S02 concentrations can be examined on three geographical scales:
1) the vicinity (< 20 km) of major point sources, 2) urban areas, and
3) multi-county to multi-state regions. As indicated by the criteria
document (CD, Chapter 5), S02 levels in urban areas and near many point
sources have been markedly reduced by control programs over the past 10
-------
to 15 years. Current (1979-80) S02 air quality data for approximately
900 population- and source-oriented sites with continuous monitors are
summarized in Tables 4-1 and 4-2. These sites reflect a more limited
number of counties and urban areas (approximately 500 total count'ies)
with multiple S02 monitors. Annual SOg levels are less than 0.03 ppm
(80 ug/m3) in 99% of both site categories. Although the distribution of
annual averages are similar in population- and source-oriented sites,
short-term concentrations (£24 hours) tend to be higher, at source
oriented sites. Less than 5% of population-oriented sites had second
maximum 24-hour levels in excess of 0.14 ppm (370 yg/m3); over 5% of 24-
hour second maxima at source-oriented sites exceeded this value.
Similarily, less than 1% of population-oriented sites had second maximum
3-hour averages in excess of 0.5 ppm, while over 5% of second 3-hour
maxima at source-oriented sites exceeded 0.5 ppm.
Tables 4-1 and 4-2 also contain two methods of displaying the 1-
hour average SOg concentrations observed during 1979-80 at these
sites. The first method consists of determining the distribution of
maximum and second highest 1-hour average concentrations. This
distribution corresponds with similar distributions at source- and
population-oriented sites for 24-hour and 3-hour averages. The second
approach gives the frequency distribution of the 99th percentile of all
hourly $03 values at these sites.
As indicated in the tables, in 50% of the sites, the annual maximum
1-hour S02 concentration would not exceed 0.16 ppm (population-oriented)
to 0.32 ppm (source-oriented). Maximum hourly levels at sites with the
highest concentrations can reach substantial values (0.7-2.3 ppm), but
such levels occur less than 1% of the time for 99% of all sites.
-------
TABLE 4-la. CHARACTERIZATION OF SO, LEVELS (IN PPM) AT POPULATION-ORIENTED
MONITORING SITES FOR DIFFERENT AVERAGING TIMES*
Distribution of the
Summary Statistic
Minimum
50th Percent! le
95th Percentile
99th Percentile
24-Hour Average
Annual
Average
0.00
0.01
0.02
0.03
Maximum
0.00
0.06
0.14
0.21
Second
Maximum
0.00
0.05
0.11
0.14
3-Hour Average
Maximum
0.00
0.12
0.34
0.45
Second
Maximum
0.00
0.11
0.28
0.40
Maximum
0.00
0.16
0.45
0.70
1-Hour Average
Second
Maximum
0.00
0.14
0.38
0.53
99th
Percentile
0.00
0.06
0.13
0.20
*Source: 1979-1980 SAROAD data base, 24- and 3-hour averages derived from running averages. Based on 761 site-years
at 521 sites.
TABLE 4-lb. APPROXIMATE CONVERSION OF PPM TO yg/nT FOR S02
PPM
0.01
0.02
0.03
0.04
0.05
0.1
0.5
0.75
1.0
Approximate yg/m
25
50
80
100
130
260
1300
2000
2600
-------
TABLE 4-2. CHARACTERIZATION OF SO, LEVELS (IN PPM) AT SOURCE-ORIENTED
MONITORING SITES FOR DIFFERENT AVERAGING TIMES*
24-Hour Average
Distribution of the
Summary Statistic
Minimum
50th Percentile
95 Percentile
99 percentile
Annual
Average
0.00
0.01
0.02
0.03
Maximum
0.00
0.07
0.23
0.41
Second
Maximum
0.00
0.06
0.18
0.28
3-Hour Average
Maximum
0.01
0.22
0.77
1.36
Second
Maximum
0.00
0.18
0.58
1.13
Maximum
0.01
0.32
1.33
2.30
1-Hour Average
Second
Maximum
0.00
0.27
1.00
1.62
99th
Percentile
0.00
0.08
0.27
0.49
*Source: 1979-1980 SAROAD data base, 24-and 3-hour averages derived from running averages. Based on'581 site-years
at 364 sites.
-------
11
The typical diurnal behavior of SOg is shown in Figure 4-1, which
represents a frequency distribution of maximum hourly S02 concentrations
at sites that are with levels among the nation's highest (Frank et al.,
1981). Maximum SO-2 levels most frequently occur in the late morning to
early afternoon during all seasons; this may be of some interest in
evaluating exposures for both humans and vegetation. The pattern is
consistent with elevated emission releases characteristic of major SOg
emitters in combination with the diurnal variance in atmospheric
turbulence. Substantial variance from this typical pattern can occur
for sites involving complex terrain, marked diurnal emissions trends, or
other exceptional atmospheric flows and source operating characteristics
(Frank et al., 1981). On a seasonal basis, monthly average S02 levels
tend to be higher in winter months in those areas affected by S02
emissions from space heating, but like diurnal patterns, seasonal
variation is dependent on site-specific source, terrain, and
meteorological characteristics (CD, p. 5-18).
Although non-urban levels of S02 are substantially lower than those
found in urban or source oriented sites, in some areas S02 levels appear
clearly in excess of those expected in most natural settings (< 0.004
ppm, CD, p. 5-5). Data from the 54 station non-urban SURE network
indicate that S02 concentrations, like those of its transformation
products (sulfates), are elevated on a regional scale in large portions
of the northeastern U.S. (Figure 4-2). .If the data from this network
are representative, ground level S02 concentrations in large portions of
the northeast (0.01-0.02 ppm, or 25-50 yg/m3) are substantially higher
than sulfate levels in the same region (6-12 yg/m3), on both a chemical
equivalent and mass concentration basis (Mueller et al., 1980). As
-------
TIME of OW, hr
om pm
6 9 NOON 3 6 9 12
I I I
1st QUARTER
2nd QUARTER
3rd QUARTER-
4th QUARTER
ANNUAL
I
I
6 9 NOON 369
am pm
TIME of DW. hr
12.
Figure 4-1. Distribution of the typical hour
of maximum SO, among 153 monitoring locations,
1975-1978 (Fr&nk et al., 1981). Sites from
EPA National Air Data Bank include all those
with 2:6000 hourly observations/year with annual
average for at least 1 hour 21.0.03 ppm S0?.
Figure 4-2. Monthly Arithmetic Mean Sulfur Dioxide
Concentrations at Nonurban Sites in the Northeast
(Mueller et al., 1980).
-------
13
evidenced by the same network, non-urban average and peak ozone levels
were also elevated in this region. Unlike S02, which on a regional
scale is highest in winter months, both ozone and sulfates tend to have
summer maxima.
-------
14
V. CRITICAL ELEMENTS IN THE REVIEW OF THE PRIMARY STANDARD
A. Mechanisms
This section discusses the factors that influence deposition and
clearance of S02 in the respiratory tract and outlines the mechanisms by
which S02, alone and in combination with other factors, may initiate
physiological and pathological responses. This discussion is intended
to aid interpretation of the role of S02 in producing responses observed
in humans.
1. Deposition and Clearance
An evaluation of the mechanisms by which S02, alone and in
combination with particles, may affect human health must recognize the
importance of deposition and clearance of S02 in the major regions of
the respiratory tract. As described in the criteria document, the
respiratory tract can be classified according to three major regions:
1) extrathoracic, including the passages of the nose, mouth, nasal
pharynx, oral pharynx, epiglottis, and larynx, 2) tracheobronchial,
including the ciliated airways from the trachea to the terminal
bronchioles, and 3) alveolar or pulmonary, including the respiratory
bronchioles, alveolar ducts and sacs, atria, and alveoli (CD, p. 1-52).
Due to its high solubility in aqueous solutions of near neutral
acidity (e.g., physiological fluids), SC>2 is readily absorbed upon
contact with the moist surfaces of the nose and other upper respiratory
passages (CD, p. 1-54). Deposition of S02 in the airways of the upper
extrathoracic region determines how much SOg is available for
penetration to the more sensitive larynx and the tracheobronchial and
alveolar regions. S02 (1 to 50 ppm) is almost completely absorbed
(_> 99%) by nasal removal under resting conditions in both man and
-------
15
laboratory animals (Frank et al., 1969; Speizer and Frank, 1966b;
Brain, 1970). No evidence for decreased nasal removal was found even
after six hours at 25 ppm (Andersen et al., 1974).
Factors that can increase penetration and deposition of S02 in the
respiratory tract over that observed for quiescent nasal breathing
include mouth and oronasal breathing, increased ventilation rates, and
the presence of airborne particles that may act as "carriers" for SOg.
These factors, critical in the interpretation of studies of the health
effects of SC>2, are summarized briefly in Appendix A.
Although most inspired S02 appears to be readily absorbed into
moist upper respiratory tract surfaces, about 15 and 30% of inhaled S02
is expired in resting human subjects for nose and mouth (only)
breathing, respectively (Melville, 1970). The results of Speizer and
Frank (1966b) suggest that the expiration may be the result of partial
desorption of inhaled S02- Absorbed S02 is rapidly transferred into the
circulatory system from all regions of the respiratory tract, a small
fraction of which may be desorbed into the alveolar region from the
blood (Frank et al., 1967).
In body fluids, S02 rapidly forms a solution of bisulfite and (in
smaller quantities) sulfite. Animal and in vitro studies indicate that
these substances can reversibly react with compounds containing
disulfide linkages (e.g., proteins) to form S-sulfonate (sulfonation
reaction) (Guunnison and Palmes, 1973). The predominant metabolic fate
of the sulfur/sulfonate complex appears to be oxidation to sulfate as
mediated by the enzyme sulfite oxidase and excretion in urine (Gunnison
and Palmes, 1973; Yokoyama et al., 1971).
-------
16
2. Mechanisms of Toxicity
SC>2 may produce physiological and, ultimately, pathological effects
by mechanisms that depend upon amount and site of deposition, reactions
with biological materials, and sensitivity of the affected region. The
major mechanisms of potential interest in S02 toxicity are discussed in
Appendix A. They can be briefly categorized as follows:
a) irritation of tissues or nerve receptors leading to
airway/functional changes;
b) alteration of clearance and other host defense mechanisms;
c) tissue irritation or damage leading to morphological
alterations; and
d) reactions with important cellular constituents.
B. Effects of Concern
This section identifies and describes the principal effects that
may be associated with S02, alone and in combination with other
pollutants. Evidence for such associations is drawn from animal
toxicology, controlled human exposure and community epidemiological
studies.* Based on these data, as summarized in the criteria document,
the following effect areas appear to be of most interest:
1) sensory and other non-respiratory responses;
2) respiratory mechanics and symptoms;
3) aggravation of existing respiratory and cardiovascular
disease;
4) clearance and other host defense mechanisms;
*Frequent reference is made to previous staff evaluations of these data
as presented in Appendix B of the companion staff paper on particulate matter
(EPA, 1982b).
-------
17
5) potential mutagenesis/carcinogenesis; and
6) mortality.
The major implications of the available literature related to each of
these effects areas are summarized below.
1. Sensory and Other Non-Respiratory Responses
Sensory and certain other non-respiratory responses to sulfur
oxides have been studied in controlled human exposures. In Soviet
tests, the average odor threshold for S02 was 0.8 to 1 ppm (Dubrovskaya,
1957). In controlled U.S. tests, the odor threshold was 0.47 ppm (A.D.
Little, Inc. 1968). Under more typical conditions, most individuals
would probably be less responsive to such levels (CD, p. 13-5). The
sensitivity of the eye to light during dark and light adaptation is
increased by short-term SOg exposures as low as 0.23 to 0.34 ppm
(Dubrovskaya, 1957; Shalamberidze, 1967). Electroencephalographic
measurements of alpha rhythms show an interruption of these brain waves
at 20-second SOg exposures of 0.3 to 1 ppm (Bushtueva, 1962).
These effects of sulfur oxides have no known implications for
health, and near these threshold levels bear little obvious relation to
personal comfort and well-being. Given the odor thresholds, it is
possible that subjects might detect the presence of S02 at levels on the
order of 1 ppm SOg in controlled human studies involving nose breathing,
making truly double blind experiments difficult.
2. Respiratory Mechanics and Symptoms
Effects on respiratory mechanics can range from mild transient
changes of little apparent direct health consequence to incapacitating
impairment of breathing. Symptomatic effects also vary in severity, but
at minimum, indicate a biological response.
-------
18
a) S02 Alone
Few animal studies have evaluated the acute effects of S02 exposure
on pulmonary mechanics (CD, Table 12-3). The most sensitive species
tested is the guinea pig; the lowest 1-hour exposure to produce an
increase in flow resistance in any animals tested was 0.16 ppm (Amdur
and Underhill, 1970). Sensitivity varies with experiment and later
studies in the same laboratory found no responses at higher levels (0.2
to 0.8 ppm) (Amdur et al., 1978b). The collective experience of guinea
• pig exposures to S02 alone suggests the following:
1) Response of individual animals varies widely, with on the order of
10% of animals tested from 1964-1974 considered as "susceptible"
(Amdur, 1964, 1973, 1974).
2) Negative findings below 1 ppm in later years may be due to
differences in strains of animals (CD, pp. 12-15 to 12-16).
3) The time course of guinea pig response (response increased with
exposure, slow to recover) appears different from the short-term
responses seen in resting humans and cats and dogs (Corn et al.,
1972; Frank and Speizer, 1965; Nadel et al., 1965). .
4) The lowest level at which response occurs is substantially below
those seen for other animals and humans where exposure was also
through predominantly quiescent nasal breathing.
5) Exposure conditions (intrapleural catheter, restraints, nasal
breathing, effects of prior anesthesia) may increase or otherwise
alter response (MAS, 1978, p. 7-14).
In essence, it is clear that the guinea pig results should be
interpreted cautiously when drawing qualitative conclusions with respect
to effects in humans.
-------
19
A number of controlled human exposure studies have examined the
acute effects of S02, alone and in combination with other pollutants, on
respiratory function and on subjective symptoms indicating discomfort.
The more important findings of a collection of such studies are summarized
1n Table 5-1. The studies vary in a number of characteristics (e.g.,
exposure mode, exercise levels) that are important when considering more
quantitative observations than are intended in this section. These as
well as other studies of interest are more fully discussed in Chapter 13
of the criteria document and in Appendix B of this paper. Those most
useful for quantitative conclusions are also discussed in Section V-D of
this paper.
Key conclusions with respect to the effects of SC>2 derived from the
evaluation in Appendix B include:
1) The major observed response to short-term exposures to S02 appears
to be bronchoconstriction, usually evidenced in increased airway
resistance and decreased expiratory flow rates. Asthmatics, and to
a lesser extent, atopic individuals with allergic responses but no
clinical manifestations of asthma, are substantially more sensitive
to these effects. Significant increases in airway resistance and
occurrence of symptoms such as shortness of breath and wheezing
have occurred in short-term mouthpiece, face mask and/or free
breathing exposures to 0.5 to 0.75 ppm S02 with moderate exercise
(Sheppard et al., 1981a; Kirkpatrick et al., 1982; Linn et al.,
1982a,b), but not at rest or with light exercise (Sheppard et al.,
1981a; Linn et al., 1982a).
2) The time course of S02 induced reflex bronchoconstriction varies
with exercise, method, and duration of exposure, and subject
-------
Exposure
Observations*
References
He'd 11 hy_Subjec ts
5 ppiu SO.,, 3-180 minutes,
mostly resting, various
routes
1-4 ppin SO., 3-30 minutes,
mostly resting, various
rou tes
0.37-0.75 ppm SO?, 15-180
minutes, various routes,
activities
Host studies report changes in function measurements,
bronchoconstriction. Increased symptoms: throat
irritation and dryness, cough.
Mixed results; some subjects appear more sensitive;
some changes in function measurements indicating
bronchoconstriction (e.g., increased airway resistance,
decreased HEFR, MMFR); some subjects report symptoms
such as throat irritation; usually greater response
with oral breathing, exercise.
5 studies - No effects on function, symptoms,
equal to or less than 0.5 ppm SO-.
2 studies - Decrease In MEFR, FVC. FEV, n, HMFR at
0.75 ppm, free breathing with moderate't« heavy exercise.
Amdur et al. (1953); Andersen et al. (1974); Frank et al.
(1962); Kreisman et al. (1976); Lawther et al. (1975);
Melville (1970); Nadel et al. (1965); Newhouse et al. (1978);
Sheppard et al. (1980); Sim and Pattle (1957); Snell and
Luchsinger (1969); Tonono (1961)
Amdur et al. (1953); Burton et al. (1969); Frank et al.
(1962); Kreisnan et al. (1976); Lawther et al. (1975);
Melville (1970); Nadel et al. (1965); Sim and Pattle (1957);
Sheppard et al. (1980); Snell and Luchsinger (1969);
Tomono (1961); Koenig et al. (1982b)
Bates and Hazucha (1973); Bedi et al. (1979); Bell et al.
(1977); Horvath and Folinsbee (1977); Jaeger et al. (1979);
Snell and Luchsinger (1969); Stacy et al. (1981)
Atopic Subjects
1 ppm SO,, with and without
1 Pig/or NaCl (0.9 MO) drop-
lets. Rest and exercise,
10 to 40 minutes
Asthmatic Subjects
1-5 ppm SO., 5-150 minutes,
mostly at rest, oral
exposures
1 ppm SO, + 1 mg/mj NaCl
(0.9 unO'droplets. RH 75S.
oral, rest and exercise,
40-60 minutes
0.25 to 0.75 ppm SO.. 5-60
minutes,light to moderate
exercise, free breathing
(chamber or face mask)
0.1-0.5 ppm SO., 5-180
minutes, various activity
levels, oral exposures
At rest, I of 7 subjects had increased airway resistance.
With exercise, decreased flow and other parameters
suggesting effects in central and small airways.
No difference between S0? + NaCl, S02 alone.
Sheppard et al. (1981a); Koenig et al. (1982a)
ro
o
Harked changes In respiratory function In most subjects
(airway resistance, others). Numbers reporting
wheezing, shortness of breath, dyspnea Increase with
dose, exercise; some require medication.
Decreased flow, other parameters suggesting effects In
central and small airways; with exercise 5 of B
experienced wheezing, 3 of 8 shortness of breath.
No effect at 0.25 ppm. At 0.5 ppn, no response at
lowest exercise rate; with higher exercise, oronasal
(face mask) exposure resulted In increased airway
resistance in all subjects, symptoms in 4 of 6.
At 0.75 ppm and moderate exercise, chamber exposure
resulted in significant increase In both symptoms
and airway resistance; functional effects
less than those for similar mouthpiece exposures.
Incremental increases in airway resistance in some
subjects at all levels, slight change in MMFR,
resting subjects; wheezing, shortness of breath in
some subjects; potentiated by exercise.
Sheppard et al. (1980); Sheppard et al. (1981a);
Kirkpatrick et al.. (1982)
Koenig et al. (1980. 1981)
Linn et al. (1982a.b); Klrkpatrick et al. (1982)
Jaeger et al. (1979); Sheppard et al. (1981a); Linn et al.
(1982a); Klrkpatrick et al. (1982)
*lun«j function tests listed here are briefly described In Appendix C.
-------
21
sensitivity. In normal resting subjects, maximal response is rapid
(5-10 minutes) and decreases quickly following exposures. Recovery
may be delayed after exercise and appears longer for sensitive
asthmatics and subjects with apparent bronchial obstruction
(Sheppard et al., 1981a; Gokemeijer et al., 1973).
3) Preliminary evidence suggests S02 may increase sensitivity to
subsequent challenge by other bronchoconstrictors (Islam et al.,
1972; Reichel, 1972). This might explain the as yet unreplicated
finding of delayed symptomatic response following S02 exposure
(Jaeger et al., 1979).
4) A gradual rise in SOg exposure appears less likely to result in
reflex bronchoconstriction than a rapid "step function" increase
(Andersen et al., 1974; Schachter, 1982). In the ambient
environment, such an increase might occur by moving from indoors to
outdoors or by a meandering source plume. Because the response is
probably more dependent on dose at sensitive receptors than on
ambient S02 levels, the onset of exercise may also produce a "step
function" increase in "effective" SOg levels at these receptors;
this is evidenced by the marked increase in airway resistance
following exercise, even in subjects who were just previously
exposed to the same level at rest for 30 minutes at levels (0.5 or
0.75 ppm) that did not produce functional changes (e.g., Stacy et
al., 1981; Koenig et al., 1982a).
5) Longer-term studies (6 hours to 6 days) have found no effects of
continuous exposure to 0.3 ppm S02 in apparently resting normal
subjects or subjects with prior airway impairment but provide some
evidence of changes in nasal cross-sectional area and pulmonary
-------
22
function changes at 1 ppm (Andersen et al., 1974; Weir and
Bromberg, 1972). Suggestion of a progressive decrease in functions
was observed at 3 ppm. In these extended studies, impaired or
bronchitic subjects did not appear more sensitive.than normals, but
this is confounded by large day-to-day variation in baseline lung
function (Weir and Bromberg, 1972) or use of medication and
incomplete reporting of techniques (Reichel, 1972).
6) Chronic animal studies provide no evidence supporting concern over
direct SOg effects on respiratory mechanics of long-term average
exposures to even high ambient levels (CD, Table 12-3). Repeated
peak exposures, however, have not been properly assessed.
7) Although particles may potentiate the effect of S02 by increased
penetration or chemical reaction, controlled human exposures have
found mixed results, with little convincing evidence that such
enhancement occurs for laboratory aerosol conditions at realistic
peak aerosol levels (Appendix B). Ambient conditions that might
tend to maximize interactions between S02 and particles (cold
temperatures, fog droplets, substantial NOg, NH3) have not been
systematically examined in the laboratory.
8) Combinations of SOg and other atmospheric gases (principally 03)
have also produced mixed results. One study (Bates and Hazucha,
1973) reported marked synergism for S02 plus 03, both at 0.37 ppm,
but follow-up work in several laboratories failed to confirm the
initial findings. The weight of evidence suggests some reason for
caution with respect to mixtures of S02, 03, and sulfates, but the
issue is unresolved (CD, p. 13-52).
-------
23
9) A number of long-term community air pollution studies (Table B-3,
EPA, 1982b) have found that populations living in areas with high
SOg and particles tend to have a higher prevalence of respiratory
symptoms and lower lung functions than those living in areas with
lower pollution levels. The data do not, however, permit clear
identification of the importance of S02 in the pollution mix. Some
cross-sectional (Neri et al., 1975; Becklake et al., 1978) and
longitudinal studies (Van der Lende, 1973, 1975, 1981) provide weak
qualitative evidence of effects associated with differences in SOg
levels, although in the presence of particles.
Two series of studies (Lawther et al., 1974a,b,c; Dockery et
al., 1981) provide qualitative evidence of lung function decrements
in adults and children in response to episodic increases in S02
levels in combination with particles.
3- Aggravation of Existing Respiratory and Cardiovascular Disease
SOg induced bronchoconstriction or other acute responses described
above clearly aggravate asthmatics in laboratory situations and might
also aggravate those with other respiratory ailments and cardiovascular
disease. A number of community observational studies (Table B-4, EPA,
1982b) of episodic as well as more moderate acute exposures to high
levels of SOg in combination with particulate matter qualitatively
suggest that these exposures aggravate the conditions of cardiovascular
patients and individuals with bronchitis, emphysema, pneumonia, and
influenza. The relative importance of S02 in these pollutant exposures
is difficult to specify.
Epidemiological support for aggravation of asthma by SC>2 is weak
(CD, p. 14-34). In addition to the usual problems associated with
-------
24
community epidemiological studies (CD, Section 14.1), studies of
asthmatics are confounded by limited numbers of subjects, use of
medication, and multiple stimuli other than air pollution that may
affect response. In addition, no studies have measured short-term
(<_ 1 hr) peak levels most likely to be responsible for any effects.
Because 1-hour maxima tend to be highest in the vicinity of strong point
sources, the study of asthmatics by Cohen et al. (1972) conducted near a
coal-fired power plant is of most direct relevance to the controlled
human results. Although the results do suggest an SOg (and particulate
matter) effect on the frequency of attacks, potential biases and other
methodological difficulties with the study (CD, p. 14-33) and the lack
of 1-hour S02 data make unequivocal conclusions impossible.
4. Clearance and Other Host Defense Mechanisms
Major host defense mechanisms potentially affected by S02 include
clearance of particles and other foreign matter from the respiratory
tract and other respiratory system related defenses against infectious
agents.
Few controlled human studies have examined the effects of S02 on
clearance. Extended nasal exposure (4 to 6 hours) to 5 and 25 ppm S02,
alone and in combination with (2 or 10 mg/m^) "inert" particles,
significantly reduced nasal mucous flow rates in resting healthy adult
subjects (Andersen et al., 1974, 1977, 1981). The effects of particles
were additive at most. Mucous flow was also reduced at 1 ppm S02, but
not significantly so (Andersen et al., 1974). An interesting
observation in the earlier study was an abnormally nigh (4 of 15)
incidence of colds within one week of S02 exposures of 1, 5, and 25 ppm
on successive days. The evidence did not relate this increase to S02
-------
25
exposures. This anecdotal result prompted a follow-up study (Andersen
et al., 1977) in which S02 exposed and control groups were subjected to
experimentally induced rhinovirus infection. The S02 exposed group had
no increase in the number of colds and had fewer symptoms. The
significance of this finding with regard to cold incidence is made
questionable by the fact that virus was innoculated into a region of the
nose previously known not to be affected by S02 induced reduction in
clearance. The innoculated region is out of the mainstream sites of
deposition for both airborne virus and S02 (Andersen et al., 1977). A
related anecdotal result of interest is the observation of Weir and
Bromberg (1972) that recovery from mild upper respiratory infections was
slower for subjects exposed to 5 ppm S02 for several days than for
control subjects. Statistical evaluation was not possible due to the
small number of subjects.
Wolff et al. (1975a,b) and Newhouse et al. (1978) have examined the
effects of single acute exposures to S02 on tracheobronchial mucociliary
clearance. At 5 ppm, S02 produced only a transient acceleration in
tracheobronchial clearance in normal, healthy subjects for oral, resting
exposures of 1 to 2 hours. Significantly increased clearance was found
for a 2.5 to 3 hour exposure that included a 30 minute exercise
period. Exercise alone increased clearance but the rate for those
undergoing exercise and SC>2 was significantly faster. Unfortunately, no
published studies have examined S02 induced effects on clearance in
sensitive subjects or the effects of repeated peak exposures in humans.
Animal studies have examined the effects of both short- and long-
term S02 exposures on clearance and other host defense mechanisms (CD,
Table 12-4). Unlike the animal-human comparisons observed for sulfuric
-------
26
acid (e.g., Leikauf et al., 1981), donkeys appear far less sensitive to
acute S02 exposures than are humans, with no response at 25 ppm
(Spiegelman et al., 1968). Rats may be more sensitive (Ferin and Leach,
1973). Of particular interest is the finding that prolonged, repeated
exposures (1.5 to 7 hours/day) to relatively low levels may result in
clearance effects (Ferin and Leach, 1973; Hirsch et al., 1975). Ferin
and Leach (1973) exposed rats to 0.1, 1 or 20 ppm S02 for 7 hr/day, 5
days/wk for 10 and 25 days as an indicator of "integrated alveolar
clearance." S02 at 0.1 ppm accelerated clearance at 10 and 23 days, but
1 ppm accelerated clearance at 10 days, had no.significant effect at 18
to 20 days, and depressed clearance at 25 days. Twenty ppm depressed
clearance after 11 days. The mechanisms by which S02 might affect
apparent alveolar clearance is unclear, but would not appear to be
related to penetration and direct effects on alveolar macrophages or
cilia beat frequency (Fraser et al., 1968). Hirsch et al. (1975) found
that tracheal mucous flow was slowed in beagles exposed to 1 ppm S02 for
1.5 hours twice a day, 5 days a week, for 1 year. Clearance was
examined only once (24 hours after exposure) and not measured prior to
exposure. Thus, the time to produce an effect is unclear, and the
extent of the depression is called into question. In essence, the
results of S02 on clearance suggest some concern over repeated peaks,
but confirmatory testing is needed.
As discussed in Section V-A and Tables 12-4 and 12-13 of the
criteria document, the effects of weekly to 3-month exposures to high
levels of S02 appear to affect antiviral defenses but not susceptibility
to bacterial infection. Some suggestion of effects of S02 and carbon on
pulmonary immune responses exists (Zarkower, 1972) but their mechanisms
were considered puzzling.
-------
27
Community epidemlological studies (Holland and Reid, 1965; Lambert
and Reid, 1970) of elevated SOg in combination with particles suggest
that long-term exposures may be associated with an increase in the
prevalence of bronchitis. Although the role of SOg versus sulfuric acid
or other particles is not clear, animal studies suggesting effects of
repeated S02 peaks on clearance provide some basis for S02
involvemen.t. Community studies (Table B-7, EPA, 1982b) also suggest
increased infectious disease during pollution episodes and in children
and adults living in areas of higher S02/particle pollution. Again, the
relative role of SOg in such responses is unclear.
5. Possible Mutagenesis/Carcinogenesis
S02 and bisulfite have been reported to be mutagenic in microbial
test systems (CD, Table 12-15) at acidic pH. Potential mechanisms
(e.g., reaction with nucleic acids, free radical production) are
outlined in Appendix A. The relevance of the microbial results to whole
animals is unclear. Negative results have been reported for mammalian
cells and insects (CD, Table 12-15). The criteria document concludes
that "On the basis of present evidence, one cannot decide whether or not
bisulfite, and hence S02, is a mutagen in mammals" (CD, p. 12-77).
Two studies have examined the potential carcinogenicity of S02
alone. Lifetime exposures of rats to 10 ppm S02 for 6 hr/day produced
no squamous cell carcinoma (Laskin et al., 1976). Peacock and Spence
(1967) as reanalyzed by EPA (CD, Appendix, Chapter 12), found a
significant increase in primary lung carcinoma in female mice and
primary lung adenomas in male and female mice. Unfortunately, the data
do not permit calculation of exposure level, which was in the nature of
intermittent peaks (high concentration unspecified, 5 minutes, 5
-------
28
days/wk) for 300 days. The combination of 10 mg/rn^ benzo(a)pyrene (BaP)
and SOg produced no lung tumors in hamsters but a significant increase
in lung carcinoma in rats (Laskin et al., 1970). A follow-up study in
rats found that lifetime exposure to S02 (10 ppm, 6 hr/day) alone and
BaP (10 mg/m3, 1 hr/day) alone produced no significant change in tumors,
but various combinations of S02 and BaP did result in increased lung
tumors (Laskin et al., 1976). The criteria document concludes that it
is difficult to interpret the studies of carcinogenicity, "However, SOg
must remain suspect as a carcinogen or cocarcinogen in view of these
reanalyses and the positive results of mutagenicity assays" (CD, p. 12-
78).
Retrospective community epidemiological studies of cancer (Appendix
B.5, EPA, 1982b) have focused on areas with high particles (especially
BaP) pollution. Nevertheless, S02 levels were also high in most areas
examined. In essence, most studies find that cigarette smoking is the
dominant cause of lung cancer, but some small portion of the observed
gradient in urban/rural cancer rates in smokers and non-smokers may have
been related to high historical levels of community air pollution.
Available studies neither prove nor negate the possibility that S02,
alone or acting with particulate carcinogens, may have contributed to
cancer.
6. Mortality
A number of epidemiological studies have demonstrated an
association between peak S02 and particulate pollution and increased
mortality rates (Table B-10, EPA, 1982b). Observations during the
severe pollution episodes in Europe and the U.S. suggest that most of
-------
29
the pollution-related deaths were among the elderly and those with
chronic cardiopulmonary disease.
Because of the high correlation between S02 and particle levels
during these study periods, the relative importance of these pollutants
are difficult to distinguish. The combination of elevated levels of S02
and particles with fog in the urban smog episodes likely resulted in
increased penetration of S02 or its irritant transformation products.
Whether S02, alone or in combination with particles, increased mortality
during "episodic" as well as "non-episodic" conditions is unknown.
Several analyses of daily mortality in London, New York City, and
Pittsburgh suggest, however, that associations between air pollution and
mortality are primarily due to particles and not S02 (Mazumdar et al.,
1981; Schimmel and Murawski, 1976; Mazumdar and Sussman, 1981). Although
these studies employed refined statistical techniques, the criteria
document indicates that they cannot clearly separate the effects of S02
and particles nor rule out possible interactive effects between the
pollutants, humidity, and other factors (CD, p. 14-53).
C. Sensitive Population Groups
This section identifies the major groups most likely to be among
the most sensitive to the effects of S02, based principally on material
presented in Sections A and B. Available estimates of the size of each
of these groups in the U.S. population are also given. Table 5-2
draws upon information from epidemiological, toxicological, and
controlled human research in summarizing the observations that have
identified various subgroups and possible explanations for their
sensitivity. Much individual variation exists among the subgroups.
-------
TABLE 5-2. MAJOR SENSITIVE POPULATION SUBGROUPS*
ubgroup/
isease Characteristics
sthmatics
recurrent episodes of
cough, wheezing, and
breathlessness result-
ing from reversible
airway obstruction
caused by bronchial
constriction, bronchial
wall swelling, and
accumulated secretions
(NIAID. 1979)
ndividuals with
non-asthmatic
atopic disorders.
g., hay fever, other
allergies
- clinical hyper-sensitivity
state, or allergy, to a
variety of environmental
antigens (NIAID. 1979)
Individuals with
chronic obstructive
pulmonary diseases
• Bronchitis - chronic
inflammation of
the trachea and
bronchus with recurrent
cough and sputum
(Fishman. 1980)
• Bronchiectasis - chroni
dilatation of bronchi
(Fishman, 1980)
• Emphysema - unevenly
distended lung
air-sacs, usually with
loss of lung tissue
(Fishman, 1980)
opulation Estimates
.300.000
derived from prevalence
stlmate of NIAID, 1979
nd applied to 1982
.5. population)
20.000.000
(derived from prevalence
estimates of Johnson,
1982b and applied to 1982
U.S. population)
7.800,000 (DHEM. 1973)
Rationale (or Criteria)
- Hyperreactlve airways
(Boushey et al.. 1980)
Hyperreactlve airways
(Boushey et al.. 1980)
Mucus hypersecution and
blocked airways may worsen
effects of S02-Induced
bronchospasm.
Enlarged airspaces Increase
blood flow resistance throug
the pulmonary capillary net-
work. Increasing cardiac
stress.
Observation/Associations Supporting Increased Sensitivity
Exaggerated bronchoconstrlctlon In asthmatics exposed briefly
to SO.. Effects at SO, levels up to an order of "?9n1tude
owerzthan In normal subjects (Sheppard et al.. 1980. 1981a;
Linn et al.. 1982a.b; Koenig et al., 1982a).
Longer recovery to normal lung function required In
exercising asthmatics following brief SO, exposure
(Sheppard et al., 1981a; Koenig et al., 1981).
Increased bronchoconstrlctlon after SO. In atopic
Individuals without asthma compared with normal subjects
(Stacy et al.. 1981; Koenig et al.. 1982a;
Sheppard et al., 1980)
Numerous Investigators have found Individuals (10-201 of
subjects) substantially more responsive to S02 than the rest
of the group (CD. p. 13-20).
GO
o
Many of the deaths and Illnesses during and after air
pollution episodes were among people with pre-existing
obstructive diseases (Martin. 1964; Martin and Bradley. 1960;
Ministry of Health. 1954; Lawther et al.. 1970).
Longer recovery to normal lung function required In
bronchltlcs following brief SO. exposure
(Gokemeljer et al.. 1973).
•Other Identifiable groups that may have enhanced sensitivity to S02 Include Individuals with cardiovascular disease, the elderly, and children
and adolescents (see text).
-------
31
Moreover, some of these groups may be at lower risks, as in the case of
patients confined to indoor environments with lower than ambient SOg
levels.
Asthmatics are substantially more sensitive to a brief exposure to
S02 than are normal subjects. Table 5-3 provides a quantitative
comparison of the relative sensitivity to S02 of asthmatics, atopies,
and normal individuals as indicated by pulmonary function tests. Asthma
is a common disease that affects approximately 9.3 million people in
the U.S. (NIAID, 1979), although higher estimates have been made (Dodge
and Burrows, 1981).
Increased sensitivity to S02, less marked than in asthmatics (Table
5-3), has also been noted among atopic individuals without any clinical
signs of asthma (Koenig et al., 1982a; Stacy et al., 1981; Sheppard et
al., 1980). Koenig et al. (1982a) suggest that hyperreactive atopies
with exercise induced bronchospasm (EIB), who, in one study, made up
about 41% of the allergic persons under age 18 (Kawabori et al., 1976),
should perhaps be categorized as having subclinical asthma. The
undetected presence of asymptomatic atopies in studies of presumed
"normal" subjects may account for the recurrent finding of subjects
"hyperreactive" to S02 who generally make up 10-20% of the study group
(CD, p. 13-20). A similar incidence of susceptible individuals has been
observed in animal studies (CD, pp. 12-15 to 12-16). Temporary but
marked airway hyperreactivity occurs in otherwise normal subjects with
viral respiratory infection (Boushey et al., 1980).
Other diseases, which may or may not be associated with asthma
appear to be important factors in the predisposition of certain
individuals to the harmful effects of high S02/particulate matter
-------
32
TABLE 5-3. COMPARISON OF RESPONSES IN NORMALS, ATOPICS, AND ASTHMATICS TO S02
a) SO, Alone1 (Sheppard et al., 1980; 1981a)
Concentration (ppm)
5
3
1
.5
% Change
Normals
18*
-3
-4
NA
in Specific Airway
Atopies
33*
1.5
2.5
NA-
Resistance (SRaw)
Asthmatics
126*
76*
25*
-vO
*Significantly different from baseline (p < 0.05). Asthmatics, but not
atopies, are significantly different from normals.
^Adult subjects at rest. Oral-mouthpiece exposure, 10 minutes.
b) 1 ppm S02 + l.mg/m3 NaCl1 (Koenig et al., 1982a)
Pulmonary
Functional
Value
RT (3Hz)
Vmax50
w
max75
FEV1.0
FRC
Percent
Normals
3.0
-8.0
-7.0
-6.0
10
Change in Functional Value
Atopies
with EIB
41*
-29*
-44*
-18*
0.3
Extrinsic
Asthmatics
67*
-44*
-50*
-23*
7.0
*Significantly different from baseline.
Adolescent subjects, measurements made after 10 minutes of moderate exercise.
Oral-mouthpiece exposure, 30 minutes at rest, followed by 5-7 minutes
break, followed by 10 minutes exercise.
-------
33
combinations. The history of catastrophic pollution episodes as well as
the lesser episodes in London have clearly indicated that chronically
ill individuals, especially those with cardiovascular and obstructive
respiratory diseases, and the elderly, were more severely affected than
the general population. Whether these groups would also be sensitive to
SOg alone is not clear. Bronchitics are reported to be among the most
sensitive to elevated S02 and particles (Lawther et al., 1970). In
limited multi-day clinical tests however, subjects with minor and more
serious obstructive disease were not clearly more sensitive to S02 than
normals (Weir and Bromberg, 1972; Reichel, 1972). Subjects with
obstructed airways (e.g., bronchitics) may, however, require a
substantially longer recovery time to baseline function following brief
exposure to S02 than normal subjects (Gokemeijer et al., 1973).
Children are generally more active outdoors and contain a somewhat
higher percentage of asthmatics and atopies (NIAID, 1979). Enhanced
sensitivity and prolonged effects have been noted among children exposed
to high levels of ambient SO? and particulate pollution (Lebowitz et
al., 1972; Douglas and Waller, 1966; Colley et al., 1973; Kiernan et
al., 1976; Becklake et al., 1978), but the effects cannot be attributed
to a single pollutant. Moderate episodes of SOg (with particles) have
been associated with depression of lung function in children (Dockery
et al., 1981).
Based on limited studies, about 15% of healthy young adults tested
appear to be habitual "mouth", or oronasal breathers under the
conditions of the tests (Saibene et al., 1978; Niinimaa et al., 1981).
Because the subjects and test procedures used could have affected the
results, extrapolation of this estimate to the general population,
-------
34
including sensitive subjects, is uncertain. Anyone, however, may
temporarily shift to oronasal breathing during exercise, conversation,
singing, or illness with related nasal congestion. The decreased nasal
cross sectional area caused by high S02 levels (Andersen et al., 1974)
might itself result in increased oronasal breathing. While all mouth
breathers may not have comparable sensitivity to S02, mouth breathers in
sensitive groups (e.g., asthmatics or atopies) are likely to be at
greater risk of effect. Because allergic subjects often have nasal
obstruction, the proportion of "mouth" breathers in such groups may be
of equal or greater magnitude to that observed in healthy subjects.
People with sulfite oxidase enzyme deficiency who are unable to
detoxify S02 and/or sulfite for subsequent excretion are very few in
the U.S. (Cohen et al., 1973). Because the condition has serious.
clinical implications regardless of SOg exposure (Mudd et al., 1967;
Irreverre et al., 1967; Shih et al., 1977) it may be irrelevant for
present consideration. Other individuals with an intermediate enzyme
deficiency (heterozygotes) may be at an increased risk, especially with
chronic exposure to S02- Although intermediate levels of sulfite
oxidase activity in the liver of humans has been reported (Shih et al.,
1977) the frequency of the heterozygote gene is not known.
D. Concentration/Response Information
As outlined in Section B, responses to S02, alone or in combination
with other pollutants have been examined in roughly three time scales:
1) peak exposures (minutes-hours), 2) short-term exposures (hours-days),
and 3) long-term exposures (months-years). Although a number of animal,
controlled human, and community studies provide important qualitative
information on the range of possible responses to S02 on these time-
-------
35
scales, only a more limited set of controlled human and epidemiological
studies are of principal interest in examining concentration-response
relationships in humans. The following review summarizes those studies
cited by the criteria document as providing the most reliable
quantitative information as well as a few other studies that provide
reasonable evidence of exposure-response relationships without allowing
derivation of specific levels. A further assessment of these studies as
applied to selecting alternative levels for air quality standards is
presented in Section VI.
1. Peak Exposures
The best information on the effects of relatively brief (minutes-
hours) peak exposures to S02 is derived from studies of exposure of
humans under controlled laboratory conditions. The importance and
limitations of controlled human exposure studies are discussed in
Section 13.1 of the criteria document. Such studies can provide
accurate measurement of responses, exposure levels and conditions for a
single pollutant or simple combinations of pollutants, and produce
useful concentration-response relationships for a variety of subjects.
Important limitations include: 1) ethical considerations reduce the
exposure regimes and types of sensitive subjects studied; 2) the number
of subjects per test is limited; 3) repeated peak exposures and other
longer-term exposures usually cannot be done; 4) only a small number of
possible combinations of environmental variables and subject-related
variables can be examined; 5) laboratory studies may induce physical or
psychological artifacts; 6) to date, only a limited number of endpoints
(usually respiratory mechanics and symptoms) have been measured; and, 7)
functional measurements used vary from study to study, making
quantitative comparisons difficult.
-------
36
a) Normal Subjects (< 5 ppm)
Concentration-response information for SOg exposures of 1 to 5 ppm
in resting healthy subjects is summarized in the criteria document and
in Table 5-1. Both oral (mouthpiece) and nasal exposures have been
studied. Although a few isolated cases of sensitivity to SOg have been
found, it is not clear that such subjects are not atopic and or
undiagnosed asthmatics. Accordingly, the criteria document concludes:
"...available evidence points to 5.0 ppm (13.1 mg/m3) as being the
most probable lowest observed effect level for induction of
bronchoconstriction effects in healthy adults exposed to S02 while at
rest" (CD, p. 13-48).
Exercise and oral breathing increases fractional penetration and
total dose of S02 (Section V.A.) and therefore results in effects at
lower ambient concentrations. Based on the work of Lawther et al.
(1975), Melville (1970), Snell and Luchsinger (1969), and Koenig et al.
(1982b), the criteria document concludes that induction of pulmonary
mechanical effects may occur at S02 concentrations _>_ 1 to 3 ppm in
exercising healthy subjects. Most of these studies involved forced oral
breathing, in some cases with light exercise.
Table 5-4 summarizes the results of a selected group of controlled
human studies involving peak (_<_ 1 to 3 hours) exposures to S02 levels at
or below 1.0 ppm. When comparing these results for quantitative
purposes, it is important to note the sensitivity of the subjects
(normal, atopic, asthmatic), the exposure mode (chamber, oral-mouthpiece
and/or nose clip, oronasal, or nasal) and exercise status. Most of the
studies involved a rather rapid (step function) increase in S02
exposure. The importance of each of these factors is discussed in the
-------
37
criteria document (CD, Chapters 11, 13) and in Section V-B and Appendix
A of this paper. The comments column of the table provides some
interpretation of the functional testing results, based upon the general
nature of the tests as summarized in Appendix C and discussions in the
criteria document (Section 13.6).
Studies involving exercising healthy adults and free breathing
(chamber) exposures are summarized in Table 5-4. Bates and Hazucha
(1973) found decreases in several parameters indicative of airway
constriction with 0.75 ppm exposure of adults under intermittent light
exercise (minute ventilation (Ve) ~ 25 1/min). Due to the limited
number of subjects, group mean changes were not statistically
significant and the results are preliminary. Stacy et al. (1981) used
the same exposure regime but with a single exercise period of heavy
exercise (Ve ~ 60 1/min) sufficient to induce oronasal breathing.
Indications of large airway constriction were significant, with a trend
toward changes in smaller airways. Responses were greater in subjects
reacting positively to allergen skin tests, suggesting that some of
these healthy subjects were atopic (but without asthma). The criteria
document concludes that these studies provide weak evidence of S02-
induced bronchospasm in free breathing exercising healthy normal
individuals at levels of 0.75 ppm (CD, p. 13-49). As noted by the
authors and by the criteria document (CD, p. 13-51), the relatively
small (8-20%) reversible changes in functional parameters observed in
these studies are not likely to be of any health significance in the
normal healthy adults tested. The principal significance of these
studies is to indicate levels and conditions at which comparable or more
marked functional changes might be expected in more sensitive
-------
TABLE 5-1. SELECTED CONTROLLED HUMAN STUDIES OF QUANTITATIVE INTEREST
(.onifntr.it ions
No. iiw1_ Snl. Jet U
0. f> (i|Mn
0.7S p| HII
17 c nil(1
. J / - . D ppm
Dui.ilii.il
1^0 minutt-s
l?(l minutes
lype Sulij4'f.ls
."> 1/min)
*'
15 minute exercise
period, tnd of first
hour V - 60 1/min
e
exercise
V • 30 \/min
e
Ui serv.it ions?
Trend toward decrease in WFR (8-10i), H[FR,n..(20i ),
FVC.FEV. n (8-101) No change in CC.
1 .U
Significant increase in KHW, (mean •*• I5'Z),
trend toward decreased FI.F FEV/FVC;
maximal after 1 hr. No change in other
parameters, symptoms. Positive allergen
skin test subjects more responsive.
mixed.
(..-.•m-m-.
S»n(t)fs t-> possible constriction in
l.iMl<» airways, constriction in
inlennedidte to small airways as
measured by HHFR but nut in small
•tirways as measured by CC. Effect
increased through 1.5 hours. No
health viunif icance to normal subjects.
Indicates large airway constriction.
possible small airway constriction.
Consistent with Bates and lla/ucha
findings. Effects not of health
significance to normal subjects.
Indicates no effects in large or
small airways. No detectable
health effects.
Ki'lnence
Dates and llazucha
(19/1)
Stacy et al. (1981)
Horvath «nd Folinsbee,
(1977); (tedl et al.
(19/9; 1982); Bates and
Hazucha (1973); Bell
et al. (1977 )
Asthmatics and Atopies
'«•
1 ppm
1 ppM SO,
'1 •«)/iii:"l«dCI
droplet aerosol
1 |i.» SO, i
1 imj/M1 (l,,f 1
droplet aerosol
1.0 ppm S0?
with and without
druplet aerosol
0.71 ppm
10 minutes
5 minutes
60 Minutes
10 minutes
at rest;
S-7 minute
break; 10
minutes
enerrise
30 minutes
al rest;5-7
10 minutes
exercise
10 minutes
plus3
7 normals
7 atopies
(seasonal rhinitis)
7 asthmatics
6 asthmatics
adults
9 Extrinsic
asthmatics
(adolescents)
R extrinsic
asthmotics
(adolescents)
8 a topic
adolescents
asthmatics)
23 asthmatics
Oral-
mouthpiece
Oral-
mouthpiece
Oral-
facemask
(no mouthpiece)
Oral-
mouthpiece
Oral-
mouthpiece
l)0ral-
mouthpiece
2) Chamber
Resting
f xercise.
V - 31 l/min-
and hyperventilation
Resting
Rest, then
txen ise; we- .10
V >. 5? 1/min1
Rest, then
exercise .
V ^24-56 1/min
Exert is ing
V;v 40 1/min
*•
asthmatics had chest tightening, wheezing.
Harked increase (mean > 3001) In SRaw In all
subjects. Dyspnea and wheezing in all subjects.
Time course of response differs from resting
• subjects.
Signi Meant decrease in V M V ,,
(mean ^ 8-141) after 30 mTflulifV. • hVWgnUicant
change In R,. FRC, FfV, Q. No symptoms.
Post exercise (4-5 minutes)--si
-------
TABLE 5-4. SELECTED CONTROLLED HUMAN STUDIES OF QUANTITATIVE INTEREST
(Continued)
ConLentrdtion-. Duration lype Subjects
Exposure
Nude
ExercKt?
Status
05 W* 10 minutes 7 asthmatics
0-5 PP™ 10 minutes 5 asthmatics
05 W» 60 minutes 24 asthmatics
S minutes 6 asthmatics
0.5 ppm
0.25 ppm
0.25 ppm
0.1 ppm
ISO minutes 40 normals,
40 mild
asthmatics
10 minutes 7 asthmatics
60 minutes 24 asthmatics
10 minutes
2 asthmatics
selected for
sensitivity at
0.25, 0.5 ppm
Oral-
mouthpiece
Oral-
mouthpiece
Chamber
(nasal)
Doral-
mouthpiece
2)oronasal-
facemask
3)nasal-
facemask
mouth occluded
Oral-
chamber «ith
nose clip
Oral-
mouthplece
Oral-
mouthpiece
OExercising ,
»e •«. 35 1/min5
2)Resting
Exercising
V. -v 27 l/«iin
Intermittent
exercise
alternating with
rest 110 Kin.)
V. -v 27 1/min
Exercising
V • 40 1/min
Resting
Exercising K
Ve ' 30 1/min5
Intermittent
exercise
alternating
with, rest
(10 minutes)
Ve -^ 27 l/mtn
Exercising ,.
V •>. 35 1/min
Significant Increase (mean = 1151) In SRaw
(all 7 subjects responded) with exercise.
Shortness of breath and wheezing reported in 3
subjects. No change in SRaw after resting exposure.
Increase in SRaw (mean = 401 over post
exercise control) in 4 subjects. One subject
(with largest Increase In SRaw) reported
increase in respiratory syi«ptoms(unspecif1ed).
No significant changes In SRaw or FVC. No
Increase In symptoms. No delayed symptoms.
Oral: Significant increase in SRaw (mean = 901)
all subjects responded. Five subjects reported
shortness of breath, the sixth reported throat
irritation; 3 subjects coughed. No symptoms during
sham (humidified air).
Orunasal: Significant increase in SRaw (mean -vSOS)
all subjects responded. Four subjects reported
shortness of breath, nose and throat irritation,
2 had no symptoms.
Nasal: Significant Increase in SRaw (mean i28Z)
5 subjects responded. One subject reported
shortness of breath, two reported nose and throat
irritation. Oral SRaw significantly greater than
nasat. Other differences between exposure routes
not statistically significant.
No functional changes in 39 of 40 normals.
One "normal" 1} yr. old had decreased MHFR
(181), Increased Raw (301). after exposure.
Small decrease (2.71) in MMFR In asthmatics.
Delayed symptoms (3 subjects).
Significant Increase In SRaw (mean • 301); 3 c
7 subjects showed Increase. No symptoms.
No significant change In SRaw or FVC. No
increase In symptoms during exposure.
No delayed symptoms.
Significant, but small (unqualified)
increase In SRaw in both subjects.
No symptoms.
Indicates constriction in large airways.
Symptoms suggestive of clinical
significance.
Statistics not given. Indicates
constriction in large airways.
Symptoms suggestive of clinical
significance. Generally consistent
with Sheppard et al., 1981 above.
Indicates no effects on large or
small airways, symptoms. No
detectable health effects. Suggests
decreased effects of nasal vs.
mouthpiece exposures, somewhat smaller
ventllation.
Indicates constriction in large
airways Is diminished by nasal, oronasal
breathing, but under these conditions
bronchoconstriction remains. Symptoms
suggestive of clinical significance,
increase going from nasal tu oronasal
to oral.
Sheppard et al.
(1911 la)
Linn et al. (I982a
Linn et al. (I982<
Kirkpatrick et al.
(1982)
Small changes in NNFR less than dlumal
variation, no clinical significance. The
"normal- responder nay have been an un-
diagnosed asthmatic. Delayed nocturnal
responses not replicated In other studies.
Small change in large airways; health
significance In these subjects unclear.
Indicates no effects on large or small
airways, symptoms. No detectable
health effects. Suggests decreased
effect of nasal mouthpiece (Sheppard
et al., 1981) exposure, somewhat
smaller ventilation, high humidity.
Small change In large airways; health
significance In these subjects unclear.
Jaeger et al. (1979
Sheppard et al.
(I98la)
Linn et al. (I982a)
Sheppard et a).
(I98la)
Exercise i
otherwise
activities, corresponding to these rates. Unless
feiV" tCM? ",'• t0ta' mil"fte Vent11atfon <».)' s« Tal>^ A-' '»•• , e ra
fied, ventilation rates given represent approximate mean values derived from Information presented In the study.
See Appendix C for a brief description of various functional tests listed here. Percentage changes Hsted here represent changes in
cjroup mean over comparable clean air exposure situation. "siio nere represent cnanges in
Exposure continued while pulmonary function tests were administered.
Kcwnig (19B2).
-------
40
subjects. The remaining chamber studies of exercising healthy adults
indicate that functional changes are not expected in such subjects at
S02 levels £0.5 ppm (CD, p. 13-49).
b) Asthmatics and Atopies (1 ppm)
Asthmatics and, to a lesser extent, atopic subjects appear
significantly more sensitive than normals to SOg induced
bronchoconstriction. This is illustrated most clearly in the work of
Sheppard et al. (1980) for S02 alone and Koenig et al. (1982a) for S02
and salt aerosol, both of which provide comparative responses of the
three groups under comparable exposure conditions.
Five studies in Table 5-4 have examined the effect of S02 on
asthmatics and atopies. Activity levels include resting to moderate
levels of exercise and ventilation rates. All studies used mouthpieces,
which tend to result in greater effects than would be expected for
normal breathing and otherwise similar exposures. Thus, effects should
not be directly extrapolated to ambient settings without accounting for
differences in ventilation and oral configuration. The Koenig group
used S02 and 1 mg/m salt droplet aerosol, which complicated assessment
of the role of S02- Nevertheless, empirical results and physico-
chemical considerations suggest the effects of the combination are
largely, if not exclusively, due to S02 alone. Results in atopies
suggest no significant difference between S02 alone and the combination
aerosol (Koenig et al., 1982a). Moreover, for this NaCl/S02 exposure
combination, less than 1% of the S02 would be expected to dissolve in
the droplet (Section IV), severely limiting the potential for increased
S02 penetration by a "carrier" mechanism. No sulfuric acid formation is
likely to have occurred (McJilton et al., 1976).
-------
41
The five studies of 1 ppm S02 (by two separate groups) are
qualitatively consistent. Despite the somewhat large inter- and intra-
subject variability in functional parameters noted in these asthmatic
subjects, it is clear that all studies found statistically and possibly
clinically significant changes in respiratory mechanics. As expected,
increased ventilation through exercise (or hyperventilation) resulted in
marked increases in airway resistance and perceptible breathing
difficulties. Under the ventilation rates, humidity and temperatures
used in these studies, no evidence existed of exercise induced
bronchospasm (EIB) in clean air. The most notable changes were reported
by Sheppard et al. (1981a) for asthmatics after 10 minutes of moderate
exercise and mouthpiece exposure. Although direct, quantitative
comparisons are not possible due to different measurements, the
responses in asthmatics observed by Koenig et al. (1982a) appear to be
less severe (but still significant), even though higher exercise rates
and longer exposure durations were involved. The lower response might
be related to differences in subjects, measurement parameters, or the
fact that Koenig's subjects were exposed at rest for 30 minutes prior to
exercise exposure while Sheppard's subjects received S02 and exercise
simultaneously. Of note, however, is that Koenig still observed
significant functional changes and symptoms after exercise exposure,
even after 30 minutes of "conditioning" at this relatively high
concentration. Functional parameters had not returned to baseline
values even 20 minutes after exercise.
c) Asthmatics (0.75 ppm)
One recent study has examined the response of exercising asthmatics
to natural oronasal and mouthpiece exposures (10 minutes) to 0.75 ppm
-------
42
S02 (Linn et al., 1982b). Although exercise levels (Ve ~ 40 1/min) were
lower than used in the study of normal subjects by Stacy et al. (1981),
functional responses of free breathing asthmatics to 10 minute exposures
were about an order of magnitude or more greater. For mouthpiece
exposures, functional responses (SRaw, FEV^g. PEFR, V^^g, Vmx25)
were consistent with those reported in the 1.0 ppm S02 mouthpiece
studies of asthmatics discussed above. For unencumbered breathing,
specific airway resistance (SRaw) also increased substantially over
.clean air controls, but the increase was about 55% of that observed for
mouthpiece breathing, presumably reflecting a decrease in S02
penetration with oronasal breathing. Based on measured ventilation
patterns (Table A-l), under the conditions of this study, the oral
ventilation rate for oronasal breathing would be about 50-65% of that
for mouthpiece breathing. Decrements in other functional parameters
(e.g., FEVj^g, Vmax5o) were somewhat smaller for unencumbered than for
mouthpiece breathing, but the differences were not statistically
significant. This study found a substantial increase in SRaw (55%) in
clean, humidified air at room temperature. This increase at Ve ~ 40
1/min is not consistent with the much smaller exercise induced
bronchoconstriction reported by Kirkpatrick et al. (1982) and Deal et
al. (1979).
Coincident with changes in lung function following S02 exposure was
a marked increse in symptom scores based on 11 sympto.m types measured.
Unlike SRaw, symptom responses were not significantly mitigated by unencumbered
breathing relative to mouthpiece breathing. Symptoms or functional
measurements were not reported on an individual basis. The authors did
observe "excess" responders and noted the removal of one sensitive subject
-------
43
from the group average because his plethysmograph readings went off the
scale of the instrument following both mouthpiece and unencumbered exposures
to S02- Linn et al. (1982b) concluded.that "...our results are consistent
with the possibility of a 'high-risk' subpopulation of asthmatics in whom
upper respiratory defenses are ineffective, suggested by Sheppard et al.
Even with effective upper respiratory defenses, many young adult asthmatics
clearly can be expected to show clinically and physiologically significant
responses to S02 at concentrations of 0.75 ppm or less."
d) Asthmatics (0.5 ppm)
Four studies in Table 5-4 examined 7 different exposures to 0.5 ppm
S02. The Jaeger et al. (1979) work is somewhat unique in the number of
subjects, exercise status (all at rest), and duration of exposure. The
major finding is a very small, but statistically significant difference
in airway constriction in resting asthmatics, breathing orally with nose
clips, but no change in normals. The small change is of no known health
significance, but much larger functional responses in one, apparently
atopic normal adolescent, and delayed symptomatic responses in this
subject and two asthmatics were considered by the authors as indicative
of possible concern. Although delayed responses are mechanistically
plausible if SOg sensitizes asthmatics to subsequent bronchoconstrictive
challenge (Reichel, 1972), the issue has not been systematically
investigated and other studies have not yet confirmed unequivocal
delayed responses following S02 exposures (Linn et al., 1982a,b;
Koenig et al., 1981).
The other studies of 0.5 ppm SOg involve averaging times of 5 to
60 minutes, with varying exercise rates and exposure modes. The most
pronounced functional and symptomatic responses were observed by
-------
44
Sheppard et al. (1981a) for moderate exercise (35 1/nrin) and mouthpiece
exposure. The results of the pilot experiment by Linn et al. (1982a)
(exposure identical with somewhat lower exercise, Ve ~ 27 1/min) and of
the oral exposures of Kirkpatrick et al. (1982) (higher exercise,
shorter duration) are qualitatively similar to the Sheppard work, but
the responses appear somewhat smaller in the more recent studies.
Varying exercise regimes, subject sensitivity, and exposure duration
appear sufficient to account for the observed differences. These
studies indicate that oral (mouthpiece) exposure with light to moderate
exercise can result in effects of concern in asthmatics (CD, p. 13-
51). They also suggest the possibility of similar responses at exercise
levels high enough to result in comparable oral ventilation rates with
free breathing. Because use of a mouthpiece decreases oral resistance
to breathing, however (CD, p. 11-11; Cole et al., 1982), it has been
speculated that mouthpiece exposures tend to overstate S02 effects with
free breathing, even with comparable oral flow rates (Proctor, 1981).
Two studies (Linn et al., 1982a; Kirkpatrick et al., 1982) used
exposure systems that more reliably simulate natural breathing
conditions. Linn et al. (1982a) found no response to 0.5 ppm S02 in a
chamber exposure. At the exercise levels used by Linn et al., 85% of
typical populations would be exclusively nasal breathing, but 15% (mouth
breathers) would have oral ventilation rates on the order of 15 1/min,
or close to the rate that might be observed for resting mouthpiece
breathing (Niinimaa et al., 1981). Based on the negative results of
Sheppard et al. (1981a) in resting asthmatics, no substantial effects
would be expected under the exposure regime used by Linn et al.
(1982a). with a higher total ventilation rate (40 1/min), Kirkpatrick
-------
45
found marked responses after 5 minute face mask exposures to
0.5 ppm S02. Both studies found evidence of exercise induced
bronchoconstriction in clean air (SRaw increase ~ 15%). Given the
conditions and exercise levels, EIB of this magnitude or less is
consistent with the work of Deal et al. (1979).
Although not precisely duplicating ambient conditions, no evidence
suggests that face masks would result in marked increases in penetration
of S02 over that obtained for natural oronasal breathing. If the
asthmatics in the Kirkpatrick study were similar to the "mouth
breathing" subjects in the Niinimaa study, the oral component of
ventilation (27 1/min) would be virtually identical to that used in the
Linn pilot study. Perhaps fortuitously, the increases in airway
resistance over clean air/exercise control in these studies are also
quantitatively similar (50% versus 40%). In essence, these studies
indicate the possibility of significant increases in airway resistance
and symptoms in free breathing asthmatics after short (5 to 10 minute)
exposures to 0.5 ppm SOg, but only with moderate or higher exercise
levels. The Kirkpatrick study further permits direct comparisons of oral
only, oronasal (facemask), and nasal only exposures at similar exercise
levels. All three routes result in significant increases in
bronchoconstriction suggesting that at high enough ventilation rates,
the effective fraction of SOg removed by the nose decreases, permitting
increased penetration.
e) Asthmatics (< 0.25 ppm)
Two studies in Table 5-4 examined exposure of asthmatics to S02
levels of _< 0.25 ppm. Sheppard et al. (1981a) found evidence of
bronchoconstriction in 3 of 7 subjects with exercise and mouthpiece
-------
46
exposures at 0.25 ppm and in the two most sensitive subjects, small (as
indicated by graph) changes at 0.1 ppm. The significance of these small
changes, unaccompanied by symptoms, is unclear. Linn et al. (1982a)
found no response for free breathing asthmatics at 0.25 ppm S02.
2. Short-term Exposures
The few controlled human studies that have examined S02 exposures
over periods of several hours to days are of limited quantitative value
due to unrealistic exposure levels and patterns, limited or unreported
activity levels, restricted numbers of endpoints measured, and in one
case, confounding influence of medication and limited description of
procedures and results (Appendix B). Weir and Bromberg (1972) found no
evidence of long-term functional changes (tests of large and small
airways, gas exchange region) or symptoms in 12 healthy subjects and 7
smokers with early signs of small airway impairment following continuous 5
and 4 day exposures to 0.3 ppm S02. Exercise, if any, was not reported.
Effects on nasal cross sectional area and mucociliary clearance, suggested
by 6-hour exposures to higher levels (1 ppm, Andersen et al., 1974), were
not examined. Functional parameters in the smokers were highly variable
under sham conditions, limiting evaluation of potential S02 responses. The
study indicates that exposure to a constant level of S02 at 0.3 ppm results
in no pulmonary function changes in normal adults or smokers with signs of
impaired small airways. Effects of similar exposures to other sensitive
groups or effects of more realistic variations about the mean S02 levels
have not been assessed.
The principal basis for developing quantitative assessments of acute
effects of ambient exposures of S02 on a daily basis is community
epidemiological studies. Such studies can provide strong evidence for the
-------
47
existence of pollution effects resulting from community exposures. The
major limitations of epidemiological studies as discussed in Section 14.1.1
of the criteria document include: 1} inadequate and inconsistent
measurement of the exposure burden of individuals; 2) variability in the
measurement of health endpoints (e.g., lung function, hospital
admissions, frequency of symptoms) and in the sensitivity of populations
studied; 3) failure to fully control for confounding, or covarying
factors, such as cigarette smoking and socioeconomic status; 4)
difficulty in distinguishing the effects of SOg from particulate matter
and other pollutants, and; 5) inability to establish a causal
relationship, or negate one, based on statistical associations.
Recognizing these limitations, the following discussion outlines
those studies cited by the criteria document as providing the most
reliable quantitative information as well as other studies that provide
useful information on the relative importance of SOg without allowing
derivation of specific levels. Epidemiological evidence for
quantitative assessment of SC>2 is drawn from the same limited set of
British studies previously evaluated in the particulate matter staff
paper (EPA, 1982b). These studies are summarized in Table 5-5. The
description and evaluation contained in Section V-D of the particulate
matter staff paper will not be repeated here. The discussion will focus
on the relative importance of SOg in producing the observed effects.
The early London mortality studies (Martin and Bradley, 1960;
Martin, 1964; Ware et al., 1981) found marked associations among smoke,
S02, and mortality in the concentration ranges indicated in Table 5-5.
The presence of dense fog may have been particularly important in these
winters. These studies did not attempt to separate the effects of
-------
TABLE 5-5. SUMMARY OF EPIDEMIOLOGICAL STUDIES PROVIDING MOST USEFUL
CONCENTRATION/RESPONSE INFORMATION FOR SHORT-TERM S0? EXPOSURES
Observed
Effects
Clear increases
in daily mortal-
ity
Likely increases
in daily mortal-
ity
Daily worsening
of health status
in bronchi tics
Time
1958-60
winters
1958-60
winters
1958-72
winters
1954-68
winters
Population
Metropolitan
London
Metropolitan
London
180, and later
= 1000 chronic
bronchitic
patients,
London
24-Hr Levels Where Effects
Are Likely (CD, Table 14-7)
BS (as ug/m3)*
>. 1000
500-1000
250-500
S02 (pg/m3)
> 1000
TO. 38 ppm)
500-1000
(0.19-0.38 ppm)
500-600
(0.19-0.23 ppm)
Comments
1958-59 winter episodes. Unusually
foggy with peak pollution. (Lawther,
1963; Holland et al., 1979).
"Greatest certainty" of Increases with
BS, SO, > 750 gg/m , but Indications of
small Increases with BS < 500 ug/m
(CD, Table 14-7). Significant
correlation most consistent for smoke.
Well designed. Peak values above daily
average may have been Important. Mini-
mum levels for any "significant" re-
sponse were 250 ug/m BS, 500 pg/m
SO,; significant associations among
sensitive subgroup at lower levels
(Lawther et al., 1970).
Study
Martin and Bradley
(1960)
Martin (1964)
Martin and Bradley
(1960)
Martin (1964)
Ware et al. (1981)
Mazumdar et al .
(1981)
Lawther et al. (1970)
Lawther (1958)
*British Smoke - A pseudo-mass indicator related to small particle (^4.5 urn) darkness (i.e., carbon content). Mass relationships approximate.
-------
49
from particulate matter. In their analysis of 14 London winters,
Mazumdar et al. (1981) attempted to partition the effects of the two
pollutants using two approaches: 1) categorizing mortality by quartiles
of pollutant levels (e.g., 4 ranges of smoke levels within each of 4
ranges of S02); and 2) regression in models that considered smoke and
S02, their squares and their interactions separately and jointly.
Examination of the quartile analyses shows that for the limited
number of comparisons where concentration of one pollutant is held
relatively constant while the other varies, there is a clear tendency
for mortality to rise with increasing smoke, but no consistent rise in
mortality with increasing S02 levels. In the regression analysis, when
pollutants were considered separately, both S02 and smoke were
significantly associated with mortality. When the pollutants were
considered together, only smoke is consistently positively and
significantly associated with mortality. The coefficient for S02 is
positive (nonsignificant) for the episodic (> 500 yg/m3 smoke and S02)
period.
On the basis of their analysis, the authors conclude that smoke
(particles), but not S02, is influential in causing mortality and that
no evidence of synergism was found. They further developed dose-
response models for smoke, but not SOg, which extended to smoke levels
below 500 pg/m3. The criteria document cautions that "serious questions
can be raised regarding specific details concerning the quartile
analysis used and the validity of reported conclusions regarding the
separation of BS from S02 effects" (CD, p. 14-21). It seems reasonable
to conclude, however, that while the 14-winter analyses suggests the
possibility of small increases in the risk of mortality at smoke levels
-------
50
less than the "likely effects level" (500 iig/m3), the published analyses
provide no such suggestion for S02 exposures below the likely effects
levels listed in the criteria document. Further examination of the
London data set and available analyses appear warranted before more
definitive conclusions can be reached.
Other time-series analyses of daily variation in mortality and
pollution in New York City between 1963 and 1972 indicate weak but
positive associations between non-episodic mortality and-daily levels of
particles, but not S02 (Schimmel and Murawski, 1976; Schimtnel, 1978).
The criteria document states that these latter analyses "create serious
doubt regarding reported associations between mortality and non-episodic
SOg levels present in New York City during the 1963 to 1972 period" (CD,
p. 14-26). Unfortunately, reliance on a single, central Manhattan
monitoring station as an estimate of pollutant exposures for the entire
New York City area "precludes clear quantitative statements regarding
possible effect or no-effect levels based on these results" (CD, p. 14-
26).
• The work of Lawther et al. (1970) on bronchitics is widely regarded
as being among the most reliable investigations of the effects of
particles and S02 on morbidity. The levels indicated in Table 5-5 are,
in the authors' judgment, the lowest leading to any significant response.
They suggested that the reported aggravations in health may have
reflected the effects of brief exposures to the maximum concentration
occurring during the day. These peaks were several times the 24-hour
averages, but, because of the wide dispersion of subjects and the
variation in the magnitude and timing of such peaks across the study
areas, the impact of peak values could not be examined directly.
-------
51
Moreover, it is possible that repeated variations in exposures during
the day may have lead to effects not accounted for by single daily peak
exposures.
Of some interest in this regard is the study by the same group
(Lawther et al., 1973, 1974a,b,c) of acute decrements in ventilatory
function in four healthy adults and two bronchitics. Responses were
associated'with variations in pollutant levels measured at their place
of work or treatment (St. Bartholomew's Hospital). After multiple
regression to remove time trend effects, S02 concentrations explained
the largest proportion of residual variance in peak flow rates, with
clearest association shown after walking exercise in heavy pollution.
Quantitative conclusions from this study are not possible.
Two additional morbidity studies involved hospital and emergency
room visits in London (Martin, 1964) and Steubenville, Ohio (Samet et
al., 1981). Although both suggest associations between admissions and
exposures to particles and SOg, neither used as sensitive a health index
for quantitative purposes as did Lawther's group.
In summary, the more quantitative epidemiological studies suggest
that effects may occur at S02 levels at or above 0.19 ppm (500 ug/m3),
24-hour average, in combination with elevated particle levels. The
strongest evidence relates to effects on bronchitics. Whether the
effects are due (in part) to S02 alone, heterogeneous formation of
sulfuric acid or other irritant aerosol, particles alone, peak values,
or repeated variations around the daily mean cannot be unequivocally
determined.
-------
52
3. Chronic Exposures
The relationship between long-term exposure to air pollution and
health has been extensively studied, but few of the studies provide
sound data or consistent findings sufficient to make quantitative
conclusions, especially regarding SOg. Geographic comparisons of
morbidity (Appendix B.5, EPA, 1982b) and mortality rates among
populations suffer from many limitations (CD, Section 14.1.1.1 and
p. 14-35). There are no studies that are useful in delineating
quantitative relationships between chronic exposures to sulfur oxides
and mortality, while only one quantitative study of chronic morbidity
effects possibly involving SOg has been identified in the criteria
document. The results of this study are outlined below.
Lunn et al. (1967, 1970) studied respiratory disease and lung
function of school children in four areas of differing pollution levels
in Sheffield, England. At age 5, both chronic upper respiratory
infections and lower respiratory tract illnesses were associated with
residence in more polluted areas. The criteria document notes that the
effects were likely with annual S02 levels of 131-275 ug/m3 (0.07-0.1
ppm) S02 in combination with 230-301 ug/m3 smoke (CD, Table 14-8). The
criteria document concludes that "no-effect" annual levels of S02 or
smoke cannot be determined from a follow-up study conducted by the same
investigators (Lunn et al., 1970; CD, p. 14-49).
-------
53
VI. FACTORS TO BE CONSIDERED IN SELECTING"PRIMARY STANDARDS FOR SULFUR
OXIDES
This section, drawing upon the previous evaluation of scientific
information, outlines the key factors that should be considered by the
Administrator in deciding what pollutant indicator to use for sulfur
oxides, establishing the level of primary standards, and designating
appropriate averaging times and frequency criteria. Preliminary staff
recommendations on the most appropriate policy options in each of these
interrelated areas are presented.
A. Pollutant Indicator, Averaging Times, and Form of the Standards
1. Pollutant Indicator
Elevated levels of S02 associated with health effects in historical
episodes were always accompanied by high levels of particles. As
discussed in previous sections, it has been difficult to separate the
effects of S02 and particles in these cases, and either or both
pollutants might have been surrogates for some hitherto unidentified
"active agent." Based on laboratory studies, particles may increase
respiratory tract penetration of absorbed S02 or otherwise enhance
toxicity by chemical or physical transformations. Combined
S02/particulate matter indicators have been suggested to account for
these potential interactions.
In its previous assessment (EPA, 1982b), the staff with CASAC
concurrence, recommended a separate standard for particles but left open
the possibility of linking S02 standards to particle levels. In
evaluating this possibility, the following factors should be
considered:
1) Recent evidence from controlled human studies shows that S02 acting
alone can, at realistic peak concentrations, produce substantial
-------
54
decrements In pulmonary function and increase respiratory symptoms
in exercising asthmatics (Section V-B).
2) In contemporary U.S. atmospheres, TSP, PM^*, and fine particle
readings all may be poor indicators of those aerosols most likely
to interact with S02- The acidity of sulfuric acid and other
ambient sulfates, major components of fine particle mass, precludes
substantial additional adsorption of S02- The presence of fog, low
temperature, and high humidity, which are not necessarily
associated with high particle mass, are more likely to lead to
interactions (Section IV).
3) The particulate matter standard should ensure that particle levels
are lower than those likely to be associated with health effects,
alone or in the presence of SOg. The principal transformation
products of SOg/particle interactions (i.e., sulfates) form a major
component of PMio and, as such, are included in the recommended
particle standards (EPA, 1982b).
4) Although the issue is unresolved, SC>2 may interact in an additive
or synergistic manner with other criteria pollutants, notably
ozone.
5) Sources of high S02 levels often are well controlled with respect
to primary particulate matter, and many sources of particles emit
no S02.
Given the health effects data, it appears that a" separate S02
standard is appropriate, whether or not a combination standard is also
*Particles less than a nominal 10 ym, termed "Thoracic Particles" in the
particulate matter staff paper (EPA, 1982). PM10 is the indicator recommended
by the staff and by CASAC for use in revised standards for particulate matter.
-------
55
adopted. In view of the above considerations, it is not clear that a
combination standard would result in any improvement in health
protection over separate standards for S02 and particles, each chosen
with due consideration of the potential for interactive effects. Unless
health related advantages can be identified, the additional complexities
involved in specifying and implementing combination S02/particle
standards do not appear warranted.
2. Averaging Time(s)
The current averaging times for the sulfur oxides primary NAAQS are
annual and 24-hour; they were based on available epidemiological
studies. The studies outlined in Section V still provide qualitative
and quantitative support for short-term standards, but do not provide
clear quantitative support for the current annual standard. Although
some studies suggest the possibility of effects associated with long-
term exposures to community air pollution containing SOg and particles,
data on respiratory tract deposition (Appendix A), long-term animal, and
community studies (Appendix B) indicate that any such effects are more
likely to be the result of exposures to repeated short-term peaks than
of continuous lower level exposures. The possibility of effects from
such lower level exposures cannot, however, be ruled out. In the
absence of more quantitative data, it appears appropriate to base the
decision on whether to retain an annual average S02 standard on the
long-term air quality that is anticipated as a result of the
implementation of short-term standards. Under the current standards,
maintenance of the 24-hour standard results in attainment of the annual
standard in the vast majority of sites. Nevertheless, air quality
analyses indicate that the exceptions where the annual standard is
-------
56
"controlling" include heavily populated urban areas such as New York
City, Chicago, and Philadelphia (Frank and Thrall, 1982). Thus,
although the health data are limited, the potential for increased
exposure in populated areas suggests that it is appropriate to consider
retaining an annual average standard.
Qualitative and quantitative (Table 5-5) studies support a 24-hour
averaging time. Some epidemiological investigators (Lawther et al.,
1970) have speculated that the observed health effects might be largely
due to short-term peaks on the order of an hour. While controlled human
exposures to peak levels clearly indicate functional and symptomatic
responses, these studies cannot capture the full range of exposures,
potential effects, and sensitive groups examined in the acute
epidemiological studies. Therefore, retention of a 24-hour averaging
time is appropriate.
Both animal and controlled human studies (Table 5-1) have reported
respiratory responses to peak S02 exposures lasting minutes to hours.
The more quantitative controlled human studies (Table 5-4) provide
evidence sufficient to warrant consideration of an additional shorter
averaging time for SC>2 standards. Most of the studies indicating
effects of concern involve averaging times of 1-hour or less. Based on
practical considerations related to monitoring, modeling, data
manipulation and storage, and implementation, the staff recommends
consideration of a 1-hour averaging time in addition to the 24-hour
period. In considering the level of a 1-hour standard, the range of
shorter-term (5-10 minute) peak levels associated with 1-hour averages
should be recognized, because effects have been observed following
controlled exposures as brief as 5 to 10 minutes.
-------
57
3. Form of the Standard
The staff recommends that the 24-hour and possible 1-hour and
annual standards be stated in a statistical form rather-than a
deterministic form (the current 24-hour standard is not to be exceeded
more than once per year). For the short-term standards this could be
accomplished by either: 1) setting a standard where an allowable number
of exceedances of the standard level would be expressed as an average or
expected number per year, or, 2) setting a standard where a given
percent of-the daily maximum hourly values would be expected to be less
than or equal to the standard level. The emissions reductions to be
achieved in the required control program would be based on a statistical
analysis of the monitoring or modeling data over a multi-year period
(e.g., the preceding 3-year period). In the case of the possible annual
standard, this would be accomplished by determining an expected value
based on multi-year monitoring data.
The statistical form can offer a more stable target for control
programs and, with reasonably complete data, is less sensitive to truly
unusual meteorological conditions than the deterministic form. The
general limitations of the deterministic form are discussed more fully
elsewhere (Biller and Feagans, 1981). Recognition of these limitations
has led EPA to promulgate or propose statistical forms for the ozone and
carbon monoxide standards.
For the purposes of this paper, ranges for standards presented
later will assume a statistical form. Alternative numbers of
exceedances may be desirable and should be considered. Multiple
exceedances are helpful in that they permit more stable portions of the
air quality distribution to be used as implementation targets. In so
doing, the interaction between multiple exceedance alternatives and the
level of the standards can be specified (Frank and Thrall, 1982).
-------
58
Two general approaches to establishing multiple exceedance
standards include:
1) Establish a standard level and permit more exceedances of that
level. With this approach,' increasing the number of exceedances
increases the risk associated with just attaining the standard.
2) Allow more exceedances, but adjust the level of the standard
downward so as to provide health protection roughly equivalent to
that associated with a single exceedance.
Figure 6-1 illustrates the relationship between standard level and
exceedances for 24-hour S02 values based on the empirical data, and
several assumed log normal distributions (Frank and Thrall, 1982). For
example, the figure shows that a standard with five allowable
exceedances has an average ratio of 0.67. This means that on the
average, the level of a 24-hour standard with five allowable exceedances
should be 67 percent of the level associated with a standard with one
allowable exceedance in order to provide the same degree of protection.
If this average adjustment factor of 0.67 were employed, however,
approximately half of the locations that would just meet this standard
would exceed the level associated with the standard with one allowable
exceedance. If protection equal to or greater than that associated with
a single exceedance standard was desired at 90% of the sites, the figure
indicates that the appropriate ratio would be 0.44, resulting in a lower
standard level. In this case about 90% of the sites might be
controlling to a greater extent than might be required under a single
exceedance standard. Trade-offs among risks, degree of over-control,
and the robustness of air quality targets should be considered when
deciding upon the number of exceedances. Comparable analyses have also
been prepared for alternative 1-hour standards (Frank and Thrall, 1982).
-------
59
Number of allowable exceedances
34567
i.U
0.9
0.8
0.7
0.5
0.1
0.3
0.2
0.1
-
.
-
1. 1
V':->-A I 1
VN,*^VX- **~*J
N * «
i
Box Plot
r— 90th percentile
i
J** *".. + *
i n^
h— 75th percentile
•—Median
' — Arithmetic average
Reference:
[standard
geometric
deviation
* ~~ 2.1
"* 2.8
"" 3.6
1
:?5th percentile
10th percentile
Figure 6-1. Factors needed to render a Multiple-Exceedance
Standard Equivalent to a Single Exceedance Standard
(Frank and Thrall, 1982). The box plots are based on 130
site-years with maximum 24-hour value > 0.12 ppm; the dotted
lines are derived from lognormal distributions and are
presented for comparative purposes.
-------
60
B. Level of the Standard
1. General Considerations
The major scientific basis for selecting S02 standards that have an
adequate margin of safety comes from controlled human exposures and
community epidemiological studies, with mechanistic support from
toxicological, deposition, and air chemistry investigations. The
limitations of available controlled human studies for quantitative
evaluation of ambient exposures of populations are summarized in Section
V-D and in the criteria document (Section 13.1). Such studies, provide
accurate measurements of specific pollutant exposures, but are limited
in exposure regimes, numbers and sensitivity of subjects, and severity
of effects tested, and may involve artifacts not representative of
ambient exposures. Community epidemiological studies, while
representing real world conditions, can only provide associations
between a complex pollutant mix and a particular set of observable
health endpoints. It follows that, although the scientific literature
provides substantial information on the potential health risks
associated with various levels and exposure patterns of SC>2, selection
of appropriate levels, frequency criteria, and averaging times remains
largely a public health policy judgment.
The following sections present a brief staff assessment of the
concentration/response relationships suggested by the most significant
controlled human studies and epidemiological studies in the criteria
document and indicate how these studies may be applied in developing
ranges for decision-making on standards for S02. The presentation also
outlines a qualitative assessment of the key factors, based on
contemporary U.S. exposures, that affect the margin of safety associated
-------
61
with the ranges of standards derived from these studies. This includes
identification of those important aspects of the qualitative human and
animal studies as summarized in Section V that should be incorporated
into margin of safety considerations. Peak (< 1-hour), short-term
(£ 24-hour), and long-term (annual), exposures are discussed separately.
2. Peak (< 1-hour) Exposjres
a) Derivation of Ranges of Interest from Controlled Human Exposure
Studies ~~~~~
Controlled human studies providing useful quantitative information
on peak SOg exposures are summarized in Section V-D and Table 5-4.
Although all of these studies provide important information, many
involved use of a mouthpiece or nose clip. As the criteria document
points out with regard to quantitative results in these studies,
"caution should be employed in regard to any attempted extrapolation of
these observed quantitative exposure-effect relationships to what might
be expected under ambient conditions" (CD, p. 13-50). With due regard
to this guidance, Table 6-1 presents a preliminary staff assessment of
the controlled human studies most useful in developing a range of
interest for selecting a 1-hour SOg standard. The table focuses on
those studies involving free breathing (chamber or facemask) exposures
and resting oral exposures. The former studies provide the closest
approximation of free breathing; the latter involve low ( ~ 10 1/min)
oral ventilation rates which would be exceeded in "mouth" breathers by
oronasal breathing with even light to moderate exercise (Table A-l).
Although the oral component of breathing may be different for oronasal
and mouthpiece breathing at comparable flow rates (Cole et al., 1982),
it is both prudent and consistent with the available SOg studies to
assume that no substantial differences exist in 503 induced responses
-------
TABLE 6-1. STAFF ASSESSMENT OF KEY CONTROLLED HUMAN EXPOSURE STUDIES
so2
Concentration (5-60 minutes)
Observed Effects
Implications
1 ppm
0.75 ppm
0.5 ppm
0.25 ppm
Functional changes, possible
symptoms in resting asthmatics, oral
(facemask or mouthpiece) exposure.
Functional changes in free
breathing normal healthy
subjects, moderate to heavy ?
exercise. No health effects.
Functional changes, symptoms in
free breathing (chamber) asthmatics,
moderate exercise.
Functional changes, symptoms
in oronasal (facemask) breathing,
asthmatics with moderate exercise
but not in asthmatics (chamber) with
light exercise.
No observed effect in free
breathing subjects.
Strong suggestion that at
this level even light exercise
for "mouth" breathing asthmatics
would result in comparable or
more marked changes.
Comparable oronasal exposures
in asthmatics, or atopies could ,
result in effects of significance.
Significant effects in mild
asthmatics with moderate exercise.
Lowest level of significant
response for free breathing.
Significant effects unlikely.
en
ro
Sheppard et al. (1980), Koenig et al. (1980). The second study used SO, in combination with saline
aerosol. '
2
Stacy et al. (1981); Bates and Hazucha (1973).
Sheppard et al. (1981a); Linn et al. (1982a); Koenig et al. (1982a).
4Linn et al. (1982b).
5Kirkpatrick et al. (1982); Linn et al. (1982a).
-------
63
for these breathing modes. Inferences made in the "implications" column
are supported by those mouthpiece studies that demonstrate the high
sensitivity of asthmatics and atopies to S02 and the substantial
enhancement of response in these subjects with increased ventilation.
The table indicates that functional changes and symptoms are likely
in "mouth" breathing asthmatics exposed to short-term peaks of 1 ppm S02
while involved in everyday activities such as walking outdoors. The
finding of functional changes in free breathing normals with moderate to
heavy exercise (Ve ~ 60 1/min) at 0.75 ppm SOg suggests that asthmatics
and atopies, who are substantially more sensitive than normals (Table
5-3; CD, p. 13-50), would experience significant responses if engaged in
exercise producing similar ventilation rates (e.g., moderate jogging).
Indeed, even with a lower exercise rate (Ve ~ 40 1/min), Linn et al.
(1982b) found "clinically and physiologically significant" responses in
free breathing young adult asthmatics at 0.75 ppm SOg. Results at 0.5
ppm SOg are mixed, but do suggest effects in asthmatics with moderate
exercise (Ve ~ 42 1/min), such as light jogging. This is the lowest
level of response for free breathing reported in the published
literature as of this writing. Studies of oronasal breathing (Appendix
A) suggest that with higher exercise (50-60 1/min), oral ventilation
rates (35 1/min) would be equivalent to those in the mouthpiece
exposures of 0.25 ppm S02 which produced small increases in airway
resistance in some asthmatics; no perceptible symptoms would be expected
(Sheppard et al., 1981a).
In evaluating these results in the context of decision-making on
possible standards, the following considerations are important:
-------
64
1) The Significance of the Observed or Anticipated Effects to Health.
Little controversy exists that a full asthma attack
represents an adverse health effect. In the extreme case, status
asthmaticus. which occurs in as many as 10% of adults hospitalized
for asthma (Senior and Lefrak, 1980), the situation can be life
threatening. In less extreme cases, day to day activities must be
terminated until medication that relieves the symptoms and eases
breathing can be administered. Even though relief is-usually rapid
following medication, an insult that requires the use of medication
to permit routine functioning might well be considered an adverse
health effect.
The controlled human exposure studies of S02 discussed above
were designed to avoid precipitating serious asthma attacks or
irreversible effects in exposed subjects. The question arises,
then, as to whether the observed responses themselves represent
adverse effects or serve as indicators of potentially more serious
consequences for larger populations exposed in ambient settings.
The criteria document concludes that temporary small changes
in pulmonary function observed in normal adults are of less concern
than functional changes and symptoms observed in asthmatics. The
document summarizes the issue as follows (CD, p. 13-51):
Probably of most concern are marked increases (> 10 percent)
in airway resistance and symptomatic effects (wheezing,
dyspnea) observed by Sheppard et al. (1981a) in a group of
mild asthmatics with oral exposure via mouthpiece to 0.5 ppm
(1.3 mg/nv3) sulfur dioxide during exercise. A recent article
(Fischl et al., 1981) and accompanying editorial (Franklin,
1981) in the medical literature discuss the inclusion of
indices of airway obstruction and presenting symptoms such as
wheezing and dyspnea among factors to be considered in
attempting to predict the need for hospitalization of asthma
patients following initial emergency room treatment (e.g.,
bronchodilator therapy, etc.) for asthmatic attacks.
-------
65
Based on this criteria document discussion, it appears that the
results of studies listed in Table 6-1 begin to be of some concern to
health when bronchoconstriction is accompanied by noticable symptoms.
This occurs for exercising asthmatics exposed to S02 levels at or above
0.5 ppm. The scientific literature does not, however, provide
sufficient information to specify a concentration at which effects of
concern should be considered adverse. In making such a judgment, the
Administrator should consider the discussion above and the following
additional factors:
a) In all cases, the bronchoconstriction and symptoms were
transient and reversible. The data do not provide direct
evidence of long-term consequences from repeated peak
exposures, but the possibility of such effects cannot be ruled
out.
b) Particularly at lower concentrations with free breathing
functional changes were moderate to small (A SRaw ~ 0 to 50%)
and within the range of variability observed for day to day
changes in many asthmatics. At 0.75 ppm, effects were more
substantial (A SRaw ~ 180%) (Table 5-4).
c) Most studies utilized mild, young-adult or adolescent
asthmatics. Even with the limited number of subjects studied
to date, some individuals appear particularly sensitive,
exhibiting functional changes and symptoms markedly greater
than average. Even more sensitive individuals may exist in the
population of "mild" asthmatics; those with more severe asthma may
also be more sensitive to S02 induced bronchoconstriction.
-------
66
d) Although the reported responses cannot be interpreted as
overt asthma attacks, the combination of bronchospasm and
symptoms might be perceived by some subjects as a "mild"
attack; this could well result in curtailment of desired
physical activities.
e) It is unclear whether psychological factors can induce
asthma attacks (NIAID, 1979). If they can, however, perception
of symptoms and bronchospasm induced by SOg could trigger more
serious consequences in suggestible subjects.
f) Given the limited combination of concentrations, subjects,
activity levels, and other environmental conditions studied to
date, firm conclusions about situations in which SOg would
likely induce asthma attacks are not possible. Even if it is
judged that the functional and symptomatic effects seen in
available studies are not themselves adverse, they could be
considered as possible indicators of more serious attacks that
are clearly adverse.
2) Relative Effect of SOg Exposure Compared to Exercise, Other
Stimuli.
One of the tests used in characterizing asthmatics is evidence of
exercise induced bronchospasm (EIB). The phenomenon has been
extensively investigated (e.g., Deal et al., 1979) and appears to
be related to temperature and relative humidity.of inspired air.
Under the exercise and other test conditions ( ~ 20°C, 70-95% RH) of
most SOg studies, EIB is usually small (increase in SRaw <_ 15%)
(Linn et al., 1982a; Sheppard et al., 1981a; Deal et al., 1979);
Linn et al. (1982b) observed a somewhat larger group mean EIB
-------
67
(SRaw ~ 55%). At low temperatures and humidities (-12°C, 0% RH)
more substantial increases in airway resistance can occur (Deal et
al., 1979). Under these environmental conditions, the EIB found by
Deal et al. (1979) (A SRaw ~ 40%) with clean air mouthpiece
breathing is smaller than the increase in airway resistance (50%
over exercise control) seen by Kirkpatrick with facemask exposure
to 0.5 ppm S02 at the same exercise rate. Thus, although EIB alone
may limit the maximal exercise rate of asthmatics, S02 enhances the
effect, further limiting activity or increasing the risk of a more
serious response. The combination of SOg, exercise, and cold dry
air has not yet been tested.
A number of other stimuli may induce bronchoconstriction in
asthmatics (Table 6-2). This does not diminish the importance of
responses potentially induced by peak S02 levels.
TABLE 6-2. TRIGGERS OF ASTHMATIC ATTACKS (NIAID, 1979)
I. Known
Al Allergen
B) Drugs
1) Allergy
2) Aspirin intolerance
3) Pharmacologic effects
0 Exercise
D) Industrial (avocational and occupational exposure)
E) Infections
F) Reflexes
II. Probable
Al Air pollutants
Bl Chemical irritants
Cl Sinusitis
Dl Vasculitis (periarteritisl
III. Possible
Al Emotions
Bl Endocrinologic imbalance
0 Weather
-------
68
3) Probability of Most Sensitive Exposures
Peak 1-hour SOg levels in excess of 0.5 ppm are rare with current
U.S. air quality, and almost always occur only in the vicinity of
major point sources (Tables 4-1, 4-2; Appendix D). Shorter term (5
to 10 minute) peaks at these levels may be somewhat more common,
but no systematic data exist. Moreover, indoor S02 levels are
almost always substantially lower than outdoor levels (CD, p. 5-
117). Thus, effects appear likely only for situations involving
asthmatics undergoing light to moderate exercise outdoors relatively
near (< 10 km) major point sources in conditions resulting in peak
(> 0.5 ppm, 5 to 10 minutes) S02 levels. Preliminary analyses
(Appendix 0) suggest that several hundred thousand asthmatics and
atopies may live in the vicinity of such sources, but because the
frequency and extent of levels in excess of 0.5 ppm is low, the
probability that an exercising asthmatic will be located in the same
time and area with such peaks appears small. Quantitative analyses
of the frequency of occurrence of such exposures are not yet available.
. Although under current air quality conditions, such situations must
be infrequent, they present the possibility of significant health
effects.
Some data suggest that rapid rises in S02 levels, such as those
involved in many of the controlled studies, are more likely to
produce effects than are more gradual rises. As discussed in
Section V-B, however, the rapid rise might result from a) movement
from indoors to outdoors, b) onset of exercise resulting in a rapid
rise in SOg at sensitive respiratory tract receptors, c) movement
into an area of peak levels (by vehicle or otherwise), as well as,
d) an actual rapid increase in ambient levels at a point.
-------
69
4) Variance about the 1-hour average.
The controlled studies in Table 5-4 indicate that effects occur
within 5-10 minutes but do not necessarily worsen with continued
exposure over the course of an hour. Five and ten minute averages
will vary about the 1-hour mean. Thus, for an area just attaining
a 1-hour standard of 0.5 ppm, 5 or 10 minute peaks will be
higher. Based on typical distributions, the peak is likely to be
within a factor of 1.4 to 2.4, or less than 1.2 ppm (Larsen, 1968;
Burton and Thrall, 1982). As the ratio of 5-minute to 1-hour
concentrations increases, the geographical extent of the peak tends
to decrease.
Based on the above evaluation of the studies and related factors,
the staff believes that an appropriate range of interest for a possible
1-hour S02 standard is 0.25 to 0.75 ppm. The lower bound of 0.25 ppm
represents a 1-hour level for which maximum 5 to 10 minute peak exposures
are not likely to exceed 0.5 ppm, the lowest level where potentially
significant responses in asthmatics have been observed with free breathing,
but with a facemask. Although mouthpiece experiments suggest the possibility
of modest functional changes produced by 10 minute exposures to 0.25 ppm
S02, with moderate exercise, the effects themselves are not clearly of
health significance. Furthermore, heavy exercise rates are required to
produce comparable or greater oral ventilation for free breathing, reducing
the probability of exposure to this condition. The upper bound of the
range (0.75 ppm) represents a level at which (based on the results in
free breathing asthmatics and normals, the increased sensitivity of
asthmatics and atopies in mouthpiece experiments, and the results of
-------
70
oronasal exposures at 0.5 ppm), the risk of effects appears high. Based
on normal air quality variations a 1-hour standard of 0.75 ppm would
permit 5-10 minute peaks in excess of 1 to 2 ppm during the peak hours,
and would permit multiple hours in which the 5-10 minute peak would
exceed 0.5 to 0.75 ppm, even when the 1-hour average is within this
range.
Independent of frequency of exposure considerations, 1-hour
concentrations at the high end of the above range would provide little
margin of safety for exercising asthmatics. The low frequency with
which such peak values would occur in the presence of active sensitive
subjects tends to enhance the margin of safety.
b) Additional factors to be considered in evaluating margin of
safety and risks - peak exposures
Additional factors that should be considered in evaluating a margin
of safety for a 1-hour standard include:
i) Interaction with Other Pollutants
The -studies in Table 6-1 were done with S02 alone. As discussed in
Section V, additive or greater than additive responses might occur where
peak S02 levels occur in combination with 03 or particles. The
available data do not provide clear evidence for synergism in either
case, but controlled exposures have not examined such combinations in
sensitive asthmatics. Because peak S02 exposures occur during morning
to early afternoon hours (Figure 4-1), combinations with 03 at or near
the standard level (0.12 ppm) are possible. Because peak S02 levels
near combustion sources will be accompanied by high NOX levels,
simultaneous peaks of S02 and 03 near these sources are not likely.
Significant health related interactions between S02 and N02 have not
been demonstrated.
-------
71
ii) Risks for Other Sensitive Groups. Effects Not Evaluated
Based on the available data, asthmatics and atopies appear to be
the most sensitive segments of the population with respect to peak S02
induced functional changes. To the extent the suggested range is
protective of these groups, the risk of functional effects in other
groups appears small. The other major effects of concern (changes in
clearance, aggravation of bronchitis, genetic toxicity, and mortality)
have not been evaluated adequately in controlled human studies, but may
result from repeated peak exposures over longer time periods. These are
addressed in the following sections.
3. Short-term (24-hour) Exposures
a) Derivation of Ranges of Interest from Epidemiological Studies
Although a number of epidemiological investigations provide
qualitative evidence for the effects of short-term exposure to S02 in
combination with particles, the criteria document indicates that those
most useful for developing quantitative conclusions include a series of
studies and reanalyses of daily mortality in London (Martin and Bradley,
1960; Martin, 1964; Ware et al., 1981; Mazumdar et al., 1981; CD, p. 14-
16 to 14-24) and studies of bronchitis patients in London (Lawther et
al., 1970). These studies are briefly evaluated in Section V-D (Table
5-5). The staff assessment of these studies for deriving ranges of
interest for 24-hour S02 standards is summarized in Table 6-3 and
discussed below.
The London mortality studies have been characterized by the
criteria document as suggesting that notable increases in excess
mortality ocurred in the range of 500-1000 pg/m3 British Smoke (BS) and
S02 (0.19-0.38 ppm) and are most likely when both pollutants exceeded
-------
72
TABLE 6-3. STAFF ASSESSMENT OF SHORT-TERM EPIDEMIOLOGICAL STUDIES
Effects/
Study
Effects Likely3
Effects Possible
Measured Sulfur Dioxide - u9/m3(ppni) - 24 hour mean
Daily Mortality
in London 1
500-1000(0.19-0.38)
Aggravation of
Bronchitis2
500-600(0.19-0.23)
< 500(0.19)
Combined Effects
Levels
500(0.19)
< 500(0.19)
Deviations in daily mortality from mean levels examined in 3 studies encompassing
individual London winters of 1958-59 and 14 aggregate winters from 1958-72. Early
winters were dominated by high smoke and S02, principally from coal combustion
emissions, and with frequent fogs (Martin and Bradley, 1960; Ware et al., 1981;
Mazundar et al., 1981).
Examination of symptoms reported by bronchitics in London. Studies conducted
from the mid-1950's to the early 1970's (Lawther et al., 1970).
3CD, Table 14-8.
about 750 pg/m3 (0.29 ppm S02). As indicated in Table 6-3, these
estimates represent judgments of the most scientifically reliable
"effects likely levels" for daily S02 (and BS) and mortality, at least
in the context of historical London pollution exposures. Because of the
severity of the health endpoint and the need to provide an adequate
margin of safety in standard setting, it is important to determine
whether the data support the possibility of health risks below these
"effects likely levels." As discussed in the criteria document (CD,
Section 14.3.1.2) and the particulate matter staff pa'per (EPA, 1982b),
the London mortality studies and reanalyses do support the possibility
of a monotonic dose-response for particles, with no obvious "threshold"
at 500 ug/m3.
-------
73
The situation with respect to S02, however, is much less clear.
When S02 alone is compared with daily mortality deviations, an apparent
association persists at levels below 500 yg/m3 or 0.19 ppm (Ware et al.,
1981; Mazumdar et al., 1981). However, if attempts are made to account
for particles and weather variables, the association between mortality
and pollution on non-episodic days (BS, S02 < 500 yg/m3) persists for
smoke but not for S02 {Mazumdar et al., 1981). This analysis also
questions the role of higher S02 levels (> 500 yg/m3 or 0.19 ppm) with
respect to mortality effects. The criteria document concludes, however,
that the effects of S02 and BS at these higher concentrations cannot
clearly be separated (CD, p. 14-24). Table 6-3 reflects the staff
conclusion of Section V-D, namely that the published evidence does not
suggest a significant risk of increased mortality for exposures to S02
alone at concentrations below the likely effects levels listed in the
criteria document.
Lawther's studies of bronchitic patients began during periods of
high pollution (1954) and continued through the time when levels were
considerably lower (e.g., 1967-68; mean winter S02, 204 yg/m3, 0.08 ppm;
BS, 68 yg/m3). Lawther et al. (1970) found an association between peak
pollution and health status of bronchitics and that responses declined
as controls reduced pollutant levels. Because of the nature of the
study, effects were related to peak daily concentrations, but the
authors felt the effects were more likely the result of brief exposures
to short-term peaks "several-times the 24-hour average." While this
suggestion is plausible, the available data do not permit any quantitative
evaluation. The mechanisms of aggravation of bronchitis might involve
responses (e.g., slowed clearance) other than the functional changes
-------
74
observed in asthmatics. Thus, 1-hour standards based on protecting
asthmatics might not be adequate for bronchitics. Therefore, 24-hour, or to
a lesser extent, longer-term pollutant averages, should be used to indicate
the possibility of repeated short-term levels of concern. Lawtheir's
suggestion that 24-hour averages of 250 yg/m3 smoke and 500 yg/m3 (0.19 ppm)
SOg represent "the minimum pollution leading to any significant reponse"
appears reasonable, although the possibility that short-term peaks of
concern may occur at lower 24-hour levels cannot be discounted. The
criteria document notes that a summary of results for selected patients for
the winter 1967-68 (Table III, Lawther et al., 1970) suggests a statistically
significant correlation between S02 (as well as smoke, sulfuric acid, and
temperature) and symptom scores for a winter with only one day of S02 > 500
yg/m3 (0.19 ppm) and BS >_ 250 yg/m3. While such simple correlations
do not demonstrate effects at lower levels, they do suggest that 500 yg/m3
(0.19 ppm) S02 (and 250 yg/m3 BS) may not be absolute thresholds for
the most sensitive bronchitis patients (CD, p. 14-53).
Both the mortality and bronchitic studies represent population
groups among the most sensitive to pollutant effects. Taken together,
as in Table 6-3, the studies indicate that effects are likely with S02
levels at or above 500 yg/m3 (0.19 ppm) in the presence of smoke levels
of 250 yg/m3 to 500 yg/m3. Thus, 500 yg/m3 (0.19 ppm) represents an
upper bound for the range of interest for 24-hour S02 standards.
Although the data suggest that effects may be possible at lower levels
and, therefore, a need to include an adequate margin of safety below the
more certain effects levels, the studies provide no scientific basis for
establishing a lower bound for a range of interest for 24-hour S02
standards. In the absence of such data, it seems reasonable to
-------
75
recommend the level of the current standard (365 pg/m^ or 0.14 ppm) as a
lower bound. This level was previously judged to provide an adequate
margin of safety from the effects under consideration here. Although
the effects of S02 cannot be unequivocally separated from particles, for
reasons specified in Section VI-A, a separate SOg standard is
desirable. For days in which particulate matter approaches the high end
of the recommended range (350 yg/m3 as PM10; EpA, 1982b), simultaneous
SOg levels'at the high end of the above range (0.19 ppm) would provide no
clear margin of safety. With lower particle levels, the margin of safety
increases.
b) Additional Factors to be Considered in Evaluating Margin of
Safety — Short-term Exposures
The range of interest was derived from British data representing
two effects categories and London pollution from 1958 to 1972. In
considering standards within this range the following additional factors
should be considered:
i) Interactions with Other Pollutants or Conditions
As noted above, interactions with expected maximum particle
concentrations have been taken into account to establish the range of
interest. Interactions of S02 with ozone and other photochemical
oxidants have not been evaluated on a daily basis. The results of
short-term (2-hour) studies at high SOg and 03 levels do not permit
clear conclusions (Appendix B).
Based on the experiences in London and other areas and solubility .
considerations, conditions involving low temperature, high humidity fogs
with substantial ammonia, with or without high primary particles, might
exacerbate the effects of S02-
-------
76
ii) Relative Exposure
Based on measurement of indoor exposures and ventilation reported
by Lawther et al. (1970), indoor pollutant exposures in British
residences more closely tracked outdoor levels than is-the case in more
tightly sealed U.S. residences. Thus, for comparable outdoor concentrations,
the overall exposure to maximum 24-hour outdoor pollution was likely to
have been greater in urban areas of Great Britain than in contemporary
U.S. exposure situations. Since many of the more sensitive individuals
may be confined indoors, the extent of increased mortality or symptoms
would tend to be lower in the U.S. than observed at comparable levels in
British studies.
i ii) Risks for Other Sensitive Groups, Effects not Evaluated
Based on the evaluation of toxicological, controlled human, and
qualitative epidemiological data, Section V-B and Table 5-2 identify a
number of effects that would be observed or anticipated to occur in
sensitive groups as a result of such exposures. The studies used to
derive the range addressed: 1) premature mortality in very sensitive
individuals with chronic respiratory and cardiovascular diseases,
individuals with influenza, and the elderly; and 2) morbidity (aggravation
of disease) in bronchitic patients. Because it is reasonable to expect
morbidity at or below levels at which mortality occurs in these sensitive
groups, the London mortality studies, which involved a large population,
also may be considered as a relatively sensitive indicator of morbidity
risks. Thus, to the extent that the London data suggest a risk of small
increases in mortality at levels as low as 0.19 ppm, some risk of morbidity
is possible at or somewhat below this level. This risk should be considered
in evaluating potential standards.
-------
77
Other groups not expressly addressed by the short-term British
studies are children and asthmatics. Daily exposures to particles and
S02 appear to be associated with increased symptoms of respiratory
disease, particularly in sensitive children (Lebowitz et al., 1972).
Repeated peak S(>2 exposures might reduce resistance to infection by
slowing nasal or other mucociliary clearance rates (Section V-B). Other
qualitative studies suggest possible small effects of repeated peak or
24-hour exposures to SOg on lung function in children in contemporary
North American cities (Becklake et al., 1978; Dockery et al., 1981).
The significance of these preliminary findings is unclear at this time,
and no reason exists to suggest important effects below the range of
interest. Potential effects on sensitive asthmatics are most likely
related to peak exposures and are addressed by the recommendations with
regard to a 1-hour standard.
The above discussion addresses, in part, several of the major
effects categories of expected effects of SOg outlined in Section V-B.
With respect to those not addressed: 1) Qualitative evidence summarized
in that section suggests that repeated acute exposures to high levels of
S02 and community air pollution may affect clearance and other host-
defense mechanisms, possibly resulting in increased infections and
disease. 2) The question of potential mutagenesis/carcinogenesis of SOg
alone or in combination with carcinogens (BaP) is unresolved. The
available toxicological data mostly involve extended repeated exposures
to very high exposures, but do not permit quantitative evaluation.
Epidemiological studies indicate that cigarette smoking is the major
determiniant of lung cancer, but do not rule out the possibility of a
-------
78
small effect of high historical levels of air pollution. Both peak (1-
hour) and short-term (24-hour) ranges would maintain these as well as
annual SOg concentrations in U.S. cities to levels well below these
historical values, but would not prevent long-term air quality in some
heavily populated areas from deteriorating to levels above those allowed
under current standards (Frank and Thrall, 1982).
Studies of hospital admissions or emergency room visits, although a
crude indicator of morbidity, sometimes can provide some suggestion of
the effects of pollution on several of the above categories. The
results of Samet et al. (1981), although essentially negative, provide a
suggestion of a very small, but statistically significant, association
between emergency room visits for respiratory diseases and SOg levels
between 2 to 369 yg/m3 (0.001-0.14 ppm) and TSP levels in the range of
14 to 700 yg/m3.
4. Long-Term (Annual) Exposures
As discussed in Section V, despite a number of studies of long-term
community exposures of S02 in combination with particles and other
pollutants, the criteria document concludes that there is little
quantitative data useful in developing long-term concentration response
relationships for SOg. Based on the work of Lunn et al. (1967, 1970)
long-term S02 levels should be maintained below 0.07 ppm (180 ug/m3), a
level not approached in contemporary U.S. atmospheres. The health basis
for limiting annual SOg levels beyond those permitted by shorter-term
standards is suggested by the more qualitative evidence summarized in
Section V-B and also outlined in the previous section. These potential
effects of interest for which quantitative data are not available
include effects on clearance and other host defense systems, and, to a
-------
79
lesser extent, potential mutagenicity or co-carcinogenicity of S02. The
major concern is whether repeated S02 peaks permitted by 24-hour or 1-
hour ranges in area-source dominated populations centers might, after
some long time period, result in increased risk of such effects. While
the available data do not provide strong evidence for substantial risks
on an individual basis, the analysis of Frank and Thrall (1982) suggests
that elimination or substantial relaxation of the current annual
standard could result in increased exposures to large numbers of people
in several heavily populated urban centers. The potentially large
increase in exposed individuals and the qualitative evidence for effects
should be considered in deciding whether to maintain an annual standard
at or above the current level of 0.03 ppm (80 ug/m3).
5. Relationships Among Averaging Times
An important consideration in evaluating potential standards are
the interrelationships among 1-hour, 24-hour, and annual averaging
times. Although several recent analyses are available (Frank and
Thrall, 1982; Burton et al., 1982; MES, 1982) information in this area
remains incomplete. Data from the approximately 900 sites used in
Tables 4-1 and 4-2 show that attainment of the current annual standard
(0.03 ppm S02) would not prevent multiple exceedances of the lower bound
of the recommended ranges for both 1-hour and 24-hour standards (Figure
6-2). This figure indicates that a national annual standard would
probably not be a useful surrogate for shorter term standards without
being unnecessarily restrictive in some areas. Conversely, as noted
above and in Appendix D, current and alternative shorter term standards
(24-hour, 3-hour, 1-hour) would not prevent long-term levels from
exceeding the current annual standard in some population-oriented sites
-------
- t.si
2
a
UJ
a «•«
••"
J
-24-HOUR DATA
-I-HOUR DATA
A
./' °
/'
*sS*j
~^r<* "
r-z
-------
81
which are few in number, but represent substantial .numbers of people
potentially exposed (Frank and Thrall, 1982).
It is also apparent that a 1-hour standard would not guarantee
attainment of the 24-hour standard at all site types. For example,
analysis of continuous S02 data at population oriented sites in 4 major
urban areas suggests that attainment of a 0.5 ppm 1-hour maximum
standard in these areas would result in a 24-hour maximum of 0.21 ppm
(550 yg/m3) (Johnson, 1982a). The situation is reversed at many sites
dominated by large point sources (Frank and Thrall, 1982).
The question of whether a 24-hour standard is a useful surrogate
for 1-hour effects has been examined by air quality and modeling
analyses (Appendix D). The adequacy of a 24-hour surrogate depends on
site type, level and number of exceedances considered adequate for
protecting health, and the degree of over-control required vs a 1-hour
standard. Figure 6-3 shows the proportion of sites with second 1-hour
maxima in excess of 0.5 ppm as a function of second 24-hour maxima for
the 900 sites noted above. These data indicate that with attainment of the
current 24-hour standard, over 40% of the sites would have second hourly
maxima in excess of 0.5 ppm. An even higher proportion of sites can be
expected to have second hourly maxima in excess of 0.25 ppm. The air
quality and modeling analyses summarized in Appendix D suggest that the
current 24-hour standard effectively limits 1-hour maxima at population-
oriented sites. At many sites dominated by strong point sources, however,
the current 24-hour standard would be a useful surrogate for a 1-hour
standard only if multiple (2 to over 20/yr) exceedances of SOg levels in
the range of 0.5 to 0.75 ppm were considered acceptable. Even a more
stringent 24-hour standard of 0.12 ppm would not substantially improve
-------
82
1-hour air quality at many sites with high 1-hour values. Still tighter
24-hour surrogates would be overly stringent at sites with currently
acceptable air quality. Relaxing the 24-hour standard to a level in the
upper portion of the range of interest (0.16 to 0.19 ppm) could permit a
substantially larger number of sites with second maximum hourly values in
excess of 0.75 ppm.
The simulation modeling of large point sources of SOg (1000 megawatt
power plant) suggests that the current secondary 3-hour standard provides
substantially better protection against 1-hour peaks than does the 24-hour
standard, and might itself be a useful surrogate. In this simulation,
the worst site is predicted to experience 3 to 4 expected exceedances/year
of 0.5 ppm and only 1 expected exceedance per year of 0.75 ppm. The air
quality analysis of Frank and Thrall (1982) does not suggest substantially
improved protection with the 3-hour standard, but because it is based on
only 1 or 2 monitors per source, the air quality analysis is less reliable
on this point.
The .analysis of air quality relationships suggests that no single
averaging time will provide the same degree of protection and control
afforded by the other averaging times in all situations. The current
24-hour standard would prevent 1-hour peaks in the range of interest
from occurring in most population oriented sites, but would allow
multiple exceedances of these values in many source oriented sites.
Similarly, the 24-hour standard limits high annual values in most, but
not all sites of interest. The current 3-hour secondary standard limits
1-hour peaks even more than the 24-hour standard, but does not
materially affect long-term urban values. In essence, while additional
analyses of alternative averaging times are needed, the work to date
-------
83
suggests that implementation of the current suite of SOg standards
(annual, 24-hour, 3-hour) provides substantial protection against the
direct effects of S02 identified in the scientific literature. This
permits consideration of reaffirming the existing S02 standards as a
reasonable policy option following the current criteria and standards
review.
C. Summary of Staff Conclusions and Recommendations
The major staff conclusions and recommendations made in Section VI,
A-B are briefly summarized below:
1) Laboratory studies show that peak ambient levels of SOg, acting
alone, can cause health effects in humans. Consequently, a
separate S02 standard is still appropriate. The additional
complexities involved in specifying combination S02/particle
standards do not appear warranted in terms of public health
protection.
2) Support for an annual standard at or near current levels is largely
qualitative. Nevertheless, because short-term standards alone would
not prevent increases in annual mean concentrations in some heavily
populated urban areas, consideration should be given to retention
of a primary annual standard for S02- Based on a series of recent
controlled human studies, consideration of a new peak (1-hour) S02
standard is also recommended. The 24-hour and potential annual and
1-hour standards should all be expressed in statistical form; the
decision on the allowable number of exceedances for the 24-hour and
potential 1-hour standard should be made in conjunction with
establishing a level for the standards.
-------
84
3) Based on a staff assessment of controlled human exposures to peak
(minutes to hours) S02 concentrations, the range of 1-hour SOg
levels of interest is 0.25 to 0.75 ppm (650 to 2000 yg/m3).
The lower bound represents a 1-hour level for which the maximum
5 to 10 minute peak exposures do not exceed 0.5 ppm, which is
the lowest level where potentially significant responses in free
(oronasal) breathing asthmatics have been reported in the published
literature as of this writing. The upper bound of the range represents
concentrations at which the risk of significant functional and
symptomatic responses in exposed sensitive asthmatics and atopies
appears high. In evaluating these laboratory data in the context
of decision making on possible 1-hour standards, the following
considerations are important: (a) the significance of the observed
or anticipated responses to health, (b) the relative effect of S02
compared to normal day to day variations in asthmatics from exercise
and other stimuli, (c) the low probability of exposures of exercising
asthmatics to peak levels, and (d) five to ten minute peak exposures
may be a factor of two to three greater than hourly averages.
Independent of frequency of exposure consideration, the upper
bound of the range contains little or no margin of safety for
exposed sensitive individuals. The limited geographical areas
likely to be affected and low frequency of peak exposure to active
asthmatics if the standard is met add to the margin of safety. The
data do not suggest other groups that are more sensitive than
asthmatics to single peak exposures, but qualitative data suggest
repeated peaks might produce effects of concern in other sensitive
-------
85
individuals. Potential interactions of S02 and 03 have not been
investigated in asthmatics. The qualitative data, potential
pollution interactions, and other considerations listed above
should be considered in determining the need for and evaluating the
margin of safety provided by alternative 1-hour standards.
4) Based on a staff assessment of the short-term epidemiological data,
the range of 24-hour S02 levels of interest are 0.14 to 0.19 ppm
(365 to 500 yg/m3). Under the conditions prevailing during the
London studies (high particles, frequent fogs, winter), the upper
end of the range represents levels at which effects may be likely
according to the criteria document. The risk of health effects
should be lower when translating these results to U.S. settings
with particle levels at or below the ambient standards.
The uncertainties with respect to interactive effects with
particles or other pollutants and the nature of effects are
important margin of safety considerations. In the absence of
quantitative scientific evidence for a level where effects are
unlikely, the level of the current standard is recommended as a
lower bound. This level was previously judged to provide an
adequate margin of safety from the same effect: under consideration
here. Qualitative data from animal toxicology, controlled human
studies and community epidemiology suggest risks of potential
effects (e.g., slowed clearance) as well as the existence of
sensitive groups (e.g., children) not evaluated in the more
quantitative studies. These factors as well as potential pollutant
interactions, exposure characteristics and whether the 24-hour
standard is intended to act as a surrogate for a 1-hour standard,
-------
86
should also be considered in evaluating the margin of safety
associated with alternative standards in the range of 0.14 to 0.19
ppm.
5) Although the data are inconclusive and uncertain, the possibility
of effects from continuous lower level exposures to SOg cannot be
ruled out. Given the lack of epidemiological data suggesting long-
term effects of S02 at or near the levels of the current annual
standard, however, no quantitative rationale can be offered to
support a specific range of interest for. an annual standard.
Nevertheless, air quality analyses summarized in Appendix D suggest
that short-term standards in the ranges recommended in this paper
would not prevent annual levels in excess of the current standard
in a limited number of heavily populated urban areas. Because of
the possibility of effects from a large increase in population
exposure, consideration should be given to maintaining a primary
annual standard at or above the level of the current standard.
6) Analyses of alternative averaging times to date suggest that
while any single standard might not be a suitable surrogate for
other averaging times, implementation of the current suite of
primary and secondary S02 standards (annual, 24-hour, 3-hour)
provides substantial protection against the direct health and
welfare (see Section VII) effects identified in the scientific
literature as being associated with ground level S02 air quality.
This permits the consideration of reaffirming the existing S02
standards as a reasonable policy option, following the current
criteria and standards review. Factors favoring such an option
include the substantial improvements likely in information on
1-hour effects over the next few years, the uncertainties in
-------
87
long-term effects data, the possibility of substantial changes in S02
control strategies prompted by regional effects, and the practical
advantages of not requiring premature formulation and implementation
of a new S02 regulatory program associated with revising standards
at this time.
-------
88
VII. CRITICAL ELEMENTS IN THE REVIEW OF THE SECONDARY STANDARD
This section discusses information drawn from the criteria document
that appears most relevant in the review and possible revision of
secondary standards for sulfur oxides. Three major categories of
welfare effects are examined. Within each category, the paper presents
1) a brief summary of relevant scientific information, 2) an evaluation
of potential quantitative relationships between S02 and effects, and
3) a staff assessment of whether the available information suggests
consideration of secondary standards that differ from the recommended
primary standards. Preliminary staff recommendations on policy options
for secondary standards are also presented.
A. Vegetation Damage
1. Nature of Effects and Factors Affecting Plant Sensitivity
Plant responses to S02 exposure may result in physiological and
biochemical effects, foliar injury, and reductions in growth and
yield. The specific response varies with the pollutant concentration,
duration of exposure, and is enhanced or mitigated by certain biological
and environmental factors.
Information on the mechanisms of toxicity of S02 suggests that
plant responses are strongly linked to the penetration of SO? and
accumulation of sulfite and bisulfite within plant cells. Since sulfate
is considered to be much less toxic than sulfite and bisulfite, toxic
effects are believed to occur when the rate of accumulation exceeds the
rate of conversion of bisulfite and sulfite to sulfate (Thomas, 1951).
In general, there is a progression of effects that can be expected
following either short-term or long-term (repeated peak or continuous)
exposure. At the lowest reported response levels, short-term exposures
-------
89
tend to be associated with potentially reversible physiological effects
(e.g., gas exchange rates, enzyme activities). As exposure levels
increase, foliar injury and reductions in growth and yield may result.
However, reductions in growth and yield are not always accompanied by
foliar injury, and conversely, foliar injury is not always a reliable
indicator of growth and yield (CD, p. 8-11).
More specifically, the types of effects can be summarized as
follows:
a) Physiological and Biochemical Effects
Physiological and biochemical changes may be expected to precede
the development of foliar injury and changes in growth and yield.
Although foliar injury is the most obvious effect of S02 on plants, it
is only the end result of a series of events that have occurred at the
sub-cellular level of biological organization. Initial changes in
photosynthetic and transpiration rates and biochemical status (e.g.,
enzyme activities, metabolite concentrations) are subtle responses that
may be followed by more severe measures of plant injury (CD, p. 8-5).
b) Foliar Injury
Foliar injury appears to be associated with the penetration of SOg
to substomatal cavities where damage to mesophyll cells may lead to
collapse of tissues and changes in leaf color (NAS, 1978; Mclaughlin et
al., 1979; Evans and Miller, 1975). Short-term exposure to SOg produces
foliar injury, which varies in appearance from ivory to white in most
broadleaf plants to darker colors (e.g., brown, red) in other plants
such as conifers (CD, p. 8-10). Long-term exposure to S02 is
characterized by chlorotic spots or mottling. Leaves are most sensitive
to S02 exposure just after growth to full development (NAS, 1978, p.
105).
-------
90
c) Growth and Yield
Effects of S02 exposure on growth and yield have been examined, to
some extent, in both short-term and long-term exposure studies.
Although increases in foliar injury and decreases in growth and yield
tend to occur simultaneously when S02 exposures are sufficiently high,
foliar injury is an imprecise measure of the effect of S02 on growth and
yield parameters. For example, studies using soybeans report that
growth and yield reductions occur with minimal or no accompanying foliar
injury (Reinert and Weber, 1980) and it is possible to have foliar
injury with no apparent effect on crop yield (Heagle et al., 1974).
Although plant damage may follow long-term exposures, enhanced
growth has also been reported when soil sulfur is not sufficient to meet
plant needs and atmospheric S02 levels are low (CD, pp. 8-7 to 8-10).
Noggle and Jones (1979) conclude that low to moderate levels of
atmospheric S02 can be beneficial for agricultural crops in areas with
sulfur deficient soils. In natural ecosystems, however, S02 induced
increases (or decreases) in plant growth can affect species richness and
diversity and might reduce the stability and productivity of such
ecosystems (CD, pp. 7-9 to 7-13). Thus, increased yield in natural
ecosystems is not always beneficial (CD, pp. 8-9 to 8-10).
The extent, severity, and type of plant response to S02 exposure
varies with plant genetics, the condition of the plant itself, and
environmental conditions. The inherent susceptibility of plants to S02
varies greatly among plant species and even among cultivars, varieties,
or clones of the same species (CD, pp. 8-12 to 8-14). In general, the
influence of genetics on plant susceptibility is stable and does not
change unless selection, mutation, or hybridization occurs (-NAS, 1978,
-------
91
pp. 101-103; Wilson and Bossert, 1971). Other important factors
affecting the severity of damage which should be considered include:
1) The stage of growth at the time of plant exposure (NAS, 1978, pp.
104-105);
2) The plant's nutrient and water status; the later is particularly
important because plants are less physiologically active and more
resistant to gas exchanges (including pollutant gases) when
stressed for water; and
3) Environmental conditions such as the presence or absence of other
pollutants, soil moisture (Hill et al., 1974; Zimmerman and
Crocker, 1934), wind speed (Ashenden and Mansfield, 1977),
temperature (Heck and Dunning, 1978), light (Thomas and Hendricks,
1956) and relative humidity (Mclaughlin and Taylor, 1981).
2. Quantitative Relationships
The following general criteria were used to select those studies
most relevant to evaluating quantitative relationships:
1) Exposure regimes should be at or near realistic levels anticipated
in the ambient environment. The following levels were used as a
guide: short-term exposures (minutes to hours) < 1 ppm (2600
yg/m3); and, long-term exposures (weeks to years) < 0.1 ppm (260
ug/m3).
2) Exposure and growth conditions must be adequately defined.
3) Plant response should be expressed so that effects on exposed
plants can be compared to control plants.
Finally, exposure studies not meeting the above criteria may be included
if discussion of their validity and significance is an important part of
assessing evidence.
-------
92
Quantitative information on plant response to S02 exposure is drawn
from controlled and ambient exposure studies. The strengths and
weaknesses of both are outlined below.
a) Controlled Exposure Studies
Controlled exposure studies are comprised of greenhouse, environmental
chamber, and field chamber studies. Although a common characteristic of
these studies is to control plant exposure to pollutants, controlled
exposure studies also allow for the control of other environmental variables.
Since field chambers allow plants to be grown and exposed under conditions
that approximate actual field conditions, experiments employing field
chambers are generally considered to reasonably simulate plant growth in
the field. Controlled exposure systems have some inherent problems that
complicate extrapolation of experimental results to actual field conditions.
For example, most of these studies have exposed plants at wind speeds
that are lower than those experienced under field conditions. This is
expected to reduce pollutant uptake and plant sensitivity (Ashenden and
Mansfield, 1977; Nobel, 1974). On the other hand, investigators typically
attempt to optimize growth conditions (e.g., soil fertilization, moisture),
which should generally increase plant sensitivity to pollution.
i) SO? Alone
The most relevant short-term and long-term exposure studies of the
effects of SOg alone are listed in Table 7-1. The short-term studies
suggest that plants'are affected by S02 exposures of 1.5 to 4 hours
duration. Although some early studies reported foliar injury in
agronomic crops following single 1-hour exposures to less than 1 ppm
(NAS, 1978; Zimmerman and Crocker, 1934), more recent studies have not
evaluated the effect of single hour exposures on other biological
-------
93
TABLE 7-1.
Concentration
ppm (»g/m )
Short-tern
0.2S (660)
0.50 (1300)
0.50 (1300)
.
0.54 (1915)
0.75 (1965)
0.79 (2070)
1.0 (2600)
1.0 (2620)
Long-term
0.05 (130)
0.06 (160)
(see text)
0.067 (190)
0.068 (190)
0.05 (131)
to 0.2 (524)
Exposure Duration/
Growth Conditions*
2 Hours
1.5 Noun
2 Hours
4 Hours
1 .5 Hours
2 Exposures
Environmental Chamber
4 Hours
2 Hours
1.5 Hours
2 Exposures
Environmental Chamber
8 Hours/day
5 Days/week
5 weeks
68 days continuously
Plants were grown 1n
closed top field
chambers and potted
In an artificial
soil mixture.
26 weeks continuous
closed outdoor chambers
winter growth
open sided greenhouse
Exposures.of 0.11
(290 ug/nr) for 4 to
20 weeks from 9:30 a.m.
Monday through
5:00 p.m. Friday
10 weeks continuously
Closed top field
chambers
Results
Foliar Injury In 7 pine species.
No significant change 1n fresh shoot weight or
Increase In foliar Injury (soybean).
No effects In 3 begonia cultlvars; 2 snap-
dragon cultlvars.
Reductions 1n shoot Might In 1 of 3 petunia
cultlvars; 1 of 2 coleus cultlvars.
Reduction In flower number In 1 of 3 petunia
cultlvars.
Foliar Injury 1n 50% of the leaves exposed 1n
Red Kidney Beans. Concentrations es low as
0.4 ppm produced 4i foliar Injury and 100:
foliar Injury was observed at 1.6 ppm sulfur
dioxide.
Slight but not statistically significant
reduction 1n dry root weight of oats. Authors
suggest "threshold."
Foliar Injury 1n 501 of the leaves exposed 1n
tempo beans. The lowest concentration producing
foliar Injury was 0.6 ppm. At 1.6 ppm where
82: of the leaf surface area was Injured.
Reductions 1n shoot weight 1n 1 of 3 petunia
cultlvars; 2 of 2 coleus cultlvars; 1 of 2
snapdragon cultlvars.
Reductions 1n flower number 1n 2 of 3 petunia
cultlvars; 1 of 2 coleus cultlvars.
No effect 1n 3 begonia cultlvars.
191 reduction 1n top dry might for oat
plants harvested at 28 days.
Reductions 1n root fresh and dry weights
of radish plants (Cherry Bel1e--a sensitive
cultlvar).
SOS reduction 1n foliar weight of alfalfa.
Other measures of plant growth were also
depressed.
Decreases In the dry weight of living leaves
(SOS), dry weight of stubble (551) and
other growth Indices 1n S23 ryegrass.
No effect In wild clones.
Reductions In dry weight of green leaves of
2 grasses ranging from about 25 to 40*4 at 20
weeks and other reductions 1n growth were also
observed between 4 and 20 weeks.
Potted grafts of old spruce showed a
reduction 1n CO, uptake which was well
correlated with decreases 1n both annual
growth ring width and wood density. Beech
seedlings exposed during winter to SO.
displayed progressive decreases 1n terminal
bud viability the following spring that
were attributed to S02 exposure.
References
Berry (1971, 1974)
Heagle and Johnson (1979)
Aded-ipe et al. (1972)
Jacobson and Colavlto
(1976)
Heck and Dunning (1978)
Jacobson and Colavlto
(1976)
Adedlpe et al. (1972)
Heck and Dunning (1978)
Mngey et al. (1973b)
Neely et al. (1977)
Bell and Clough (1973)
Ashenden (1.979)
Keller (1978, 1980)
'Plants nere grown under greenhouse conditions unless otherwise Indicated.
-------
94
endpoints. Studies in Table 7-1 indicate that conifer seedlings are
particularly sensitive to SOg exposure—a factor which may affect seedling
survival. In addition to those studies reporting foliar injury, Adedipe
et al. (1972) found that reductions in flower growth and development may
also occur at relatively low exposure regimes (e.g., 0.5 ppm for 2 hours).
The long-term exposure studies in Table 7-1 do not necessarily
utilize realistic fluctuations in concentration during long-term
exposures. The studies on tree species (Keller, 1978, 1980), ryegrass
(Bell and Clough, 1973), and radish (Tingey et al., 1971b) were designed
to maintain continuous uniform exposure to S02- The studies by Keller
(1978, 1980) are regarded as careful work demonstrating how extended
exposure to S02, in sufficient amounts to depress photosynthesis, can
reduce plant growth. The study by Tingey and coworkers (1971b)
reporting reductions in radish harvest parameters is of particular
interest because it found that plants may be affected at relatively low
exposure .levels. The median concentration (0.06 ppm) in the study by
Neely and coworkers (1977) apparently included repeated peak
exposures. In this study, the highest single hour daylight
concentration during the growing season reached 0.60 ppm S02 and the
second and third highest peak 1-hour daylight concentrations were 0.42
and 0.40 ppm S02, respectively (Neely, 1982). Consequently, in this
study, it is not possible to determine the relative importance of
repeated peak versus long-term low-level exposure.
ii) Pollutant Mixtures
Studies of the effects of pollutant mixtures show that plants may
exhibit different types of responses which are influenced by several
-------
95
variables and are difficult to predict. Three general kinds of
responses may follow plant exposure to mixtures of S02 and one or more
other pollutants.
1) Additive responses. An additive response is a response that would
be predicted from the effects of individual pollutants alone. For
example, if a 2-hour exposure to either of two pollutants each at
0.10 ppm caused 10 percent foliar injury, 20 percent foliar injury
would be expected following simultaneous exposure to both
pollutants.
2) Greater than additive or "synergistic" responses. One example of a
"synergistic response" is one in which there is no foliar injury
from exposure to either pollutant alone and foliar injury is
observed following simultaneous exposure to both pollutants.
Within the context of plant responses to pollutant exposure a
"synergistic" response refers only to the degree of response and
does not imply anything about potential mechanism(s) of injury.
3) Less than additive or "antagonistic" responses. Antagonistic
responses are expected when exposures are high and effects are
relatively pronounced.
Two of these three kinds of responses are found to some degree in the
experimental results using various combinations of S02 and 03 or NOg
shown in Table 7-2. At lower concentrations (e.g., 0.10 ppm NOg and
0.05 ppm S02), little if any increase in foliar injury is observed. As
concentrations are raised (e.g., 0.10 ppm N02 and 0.10 ppm $02), the
increase in foliar injury is greater than additive. At higher
concentrations, foliar injury increases for the S02 and 03 but not for
the S02 and N02 exposure mixtures. The variability of response at
-------
96
TABLE 7-2. EFFECTS OF SINGLE 4-HOUR S0?
EXPOSURE MIXTURES ON FOLIAR^INJURY
a)
0
Species
Alfalfa6
Broccol i
Cabbage
Radish
Tomato
Tobacco
(Bell W3)
b)
2
Species
Pinto Beans
Oats
Radish
Soybean
Tobacco
(Bell W3)
Tomato
Percent
Ozone/Sulfur Dioxide
0/0.10 0.10/0.10 0
2 24
1 34
0 22
0 50
0 50
0 95
1 n
Leaf Injury '
Concentration
.10/0.25 0.
21
11
14
22
10
88
(ppm/ppm)
10/0.50
60
19
54
46
13
96
0/0.50
19
0
0
1
0
0
Percent Foliar Injury '
Nitrogen Dioxide/Sulfur Dioxide (ppm/ppm)
0.10/0.05 0.10/0.10 0.
0 11
0 27
1 27
1 35
9 11
0 1
15/0.10 0.
24
12
24
20
18
17
20/0.20
16
10
6
9
4
0
0.15/0.20
4
0
4
1
6
0
]Source: Tingey et al. (1973a).
2Alfalfa (Medicago sativa. L. cv. Vernal), broccoli (Brassica oleracea botrytis.
L., cv. Calabrese), cabbage (Brassica oleracea capitata. L. cv. All Season), radish
(Raphnus sativus, L. cv. Cherry Belle), tomato Lycopersicon esculentum, Mill, cv.
Roma VF), and tobacco (Nicotiana tabacum, L. cv. Bell VL). Pinto beans (Phaseolus
vulgaris L., cv Pinto), oats (Avena sativa L., cv. Clintland 64), soybeanTGIycine
max^L., Merr., cv. Hark), tobacco (Nicotiana tabacum L., cv. Bel W3).
3Source: Tingey et al. (1971a).
Average foliar injury resulting from 0.10 ppm 03 only was 2% alfala, 7% radish,
40% tobacco, and no foliar injury for the other three species.
Concentrations below either 2 ppm nitrogen dioxide or 0.50 ppm sulfur dioxide alone
produced no foliar injury.
Sensitive cultivar.
5
-------
97
higher concentrations of N02 and S02 is not readily explained. This
reduced response at higher exposures is unlike that observed for the 03
and S02 mixtures and may be related to.stomatal closure, which would
reduce pollutant influx into leaves. Although both of these studies
reported pronounced synergism at concentrations below the thresholds for
single pollutants, some other studies examining plant response to
similar pollutant mixtures have not found synergistic effects at levels
significantly below the thresholds for the single pollutants (e.g.,
Bennett et al., 1975; Heagle and Johnson, 1979).
An additional complication arises when assessing controlled
exposure studies of pollutant mixtures in that exposure regimes employed
are not necessarily representative of those occurring in the ambient
environment. For example, peak 03 concentrations are unlikely to occur
simultaneously with peak N02 and S02 concentrations near combustion
sources because NO, which is released along with N02, is effective at
scavenging 63.
Table 7-3 summarizes plant responses following long-term exposure
to pollutant mixtures. The lack of an effect of S02 on soybean yield in
the higher exposure regime in the Heagle et al. (1974) study is
unexpected in view of the decrease in growth reported by Tingey and
coworkers. This may have occurred because different measures of
productivity were used and because the measurements were not taken at
the same stage of growth. The Red Kidney Bean study was designed to
introduce an increment of 0.10 ppm (260 u9/m3) S02 with 50 percent non-
filtered air in the South Coast Air Basin. Oshima (1978) reports that
the 03 concentrations in the exposure chambers with 50 percent non-
filtered air probably reached 0.15 ppm (294 ug/m3), 0.12 (235 yg/m3),
-------
TABLE 7-3. CONTROLLED EXPOSURE STUDIES OF LONG-TERM EXPOSURES TO POLLUTANT MIXTURES
Concentration
ppm (ug/m )
u^ rr \ r" Jl ; 1 ,_
0.05 (98) 0,+
0.05 (131) SO,
\ / 2
0.05 (98) 0,+
0.05 (131) SO,
c.
0.10 (262) S02+
partially
f i 1 tered
ambient air
containing 0,*
0.10 (200 0,+
0.10 (260) 502
Exposure
Period
8 hrs/day
5 days/week
18 days
8 hrs/day
5 days/week
5 weeks
6 hrs/day
on 45 days
over 11
weeks
6 hrs/day
133 days
(continuous-
ly)
Response
Synergistic
Additive/
Antagonistic
Synergistic
Additive
(No SO,)
i
Comments
Greenhouse grown soybeans displayed reductions
in top weight (21%), root fresh (24%) and dry
(20%) weights and shoot-root fresh and dry weight
ratios. No effect on top dry weight or plant
height. Plant growth was not affected by either
pollutant alone.
Greenhouse grown radishes displayed less than
additive reductions in plant fresh weight, and
root dry weights. Additive reductions were
observed for leaf fresh and dry weights. Exposure
to either S02 or 0, alone affected plant growth.
Approximately a 33% reduction in total seed weight
of Red Kidney Bean when compared with ambient air
having 50% of the ambient 0, removed. Plants
were grown in pots placed in the field inside
special design closed top field chambers.
No effect in soybeans from sulfur dioxide alone or
combination with 0, (different from the effects of
Oo alone) plant growth, and foliar injury. Plants
were grown in the ground and exposed in closed
field chambers.
Reference
Tingey et al.
(1973b)
Tingey et al.
(1971b)
Oshima (1978)
Heagle et al.
(1974)
ID
O>
*See text.
-------
99
and 0.12 (235 ug/m3) ppm on the highest, second highest, and third
highest 03 days, respectively. Exposure regimes reported in the Oshima
study would violate the current 03 standard (0.12 ppm, 1-hour average,
not to be exceeded more than once per year, statistical form) and the
exposure regime used in the study by Heagle and coworkers probably would
have violated the standard.
In summary, short-term and long-term exposure studies of pollutant
mixtures provide mixed results that are difficult to interpret.
Although these studies indicate that pollutant mixtures can produce
effects that are greater than additive (especially at low exposure
levels), additional research is needed to resolve reported differences.
i1i) Zonal Air Pollution System Exposure Studies
The Zonal Air Pollution System (ZAPS) is a relatively recent
development in controlled exposure systems that allows plants to be
grown without the use of chambers. The ZAPS system releases controlled
amounts of S02 from a pipe upwind of the vegetation. Plant response to
different exposure levels are examined by placing plants and S02 monitors
at different distances downwind of the ZAPS release pipe. It is difficult
to characterize exposure regimes in ZAPS studies for two reasons:
1) Estimates of plant exposure are subject to uncertainty. The number
of monitors are limited and concentrations at each location are not
continuously monitored. Furthermore, the close proximity of the
vegetation to the release point of S02 and variations in
meteorological conditions and canopy structure also increase the
uncertainties associated with plant exposure (Miller et al.,
1980).
-------
100
2) Other pollutants are not excluded in this exposure system and
could, if present, influence plant response.
The most extensive ZAPS S02 exposure studies have been conducted by
Miller and coworkers. In one study using repeated exposures Tasting 4.5
hours, significant reductions in the yield of soybeans ranged from 6.4
percent at 0.09 ppm SOg to 45 percent at 0.79 ppm (Sprugel et al.,
1980). Although ambient 03 levels may have exceeded 0.10 ppm on more
than one occasion, it is not possible to attribute any effect on
soybeans to 03 exposure alone. Within this study, exposure to 5
different mean S02 concentrations between 0.09 (240 ug/m3) and 0.36 ppm
(940 yg/m3) for 18 repeated 4.5 hour exposures reduced yield from 6.4 to
19 percent. No single 3-hour average S02 concentration in these studies
exceeded 0.5 ppm. The consistency of results over a two year period at
different geographical locations supports these findings.
b) Ambient Exposure Studies
Field observations of plant response in agronomic settings and
natural ecosystems show that S02 exposure can increase foliar injury,
decrease plant growth and yield, and decrease the richness and diversity
of plants. Ambient exposure studies have three major advantages over
controlled exposure studies: 1) plants are exposed and grown under
conditions that occur in the ambient air; 2) large sample sizes can be
collected; and 3) only field observations can demonstrate long-term
changes in plant communities resulting from SOg exposures. On the other
hand, it is more difficult to define actual exposure conditions (i.e.,
frequency, concentration, duration), and to ascribe the observed plant
responses to S02 alone.
-------
-------
101
i) Short-Term Exposures
The criteria document cites ambient exposure studies. These
studies are important because tfrey provide:
1) quantitative estimates of short-term SOg exposure and observations of
plant response,
2) the only available data on plant responses to short-term exposures
in the field observed near point emission sources, and
3) observations on larger numbers of species than would be practical
in laboratory settings.
The ambient exposure studies of Dreisinger and McGovern (1970) and
Mclaughlin and Lee (1974) provide the most extensive observations of
foliar injury in the vicinity of major point sources of S02 (CD, pp. 8-
19 to 8-24). Observations of minimum concentrations required to produce
foliar injury in these studies are limited by the number of S02
monitors, location of individual plant species, and distribution of S02
exposures (i.e., concentration, frequency, exposure duration).
Figure 7-1 shows concentration-response curves for foliar injury
that have been developed from observations on a variety of agricultural
crops, trees, and native vegetation from the 1970 to 1973 growing
seasons. This figure does not represent the percent of foliar injury,
but the cumulative percent of affected species (i.e., those species in
which some foliar injury was observed in conjunction with air quality
measurements) for a given concentration of SOg. Plotted lines were
fitted by eye.
The data shown in Figure 7-la were collected in the vicinity of a
coal fired power plant in the southeastern U.S. from 1970 to 1973.
Following review of these data, Mclaughlin and Lee concluded that most
species in the area "are not visibly affected by sulfur dioxide when
-------
CUMULATIVE PERCENT OF AFFECTEO'SPECIES
o
-h -"• fD b O -••
T3 O O c-t- ~O C
—' fD fD 3- ~S 1
Cu 3 fD O fD
3 Cu c-t- Cu CL
ft- —t CU !/} CZ *—J
CO —IUD c-t- O I
fD fD fD t—'
Cu LO in -5
-h fD O O
fD O -h CZ —' O
O -i. . _-. O
c-t- fD C/) OO CU 3
fD in ~o • ~5 o
CL. fD fD
(DO -J. 3
CL X —'• CT 3 e-t-
C CU fD -!_i. -s
-S =3 co c: cu
UD fD -h (D • O
CU -h -S 3
Cu fD —'• I
3 -i. n < o> -s
^ 3 c-t- fD -—- fD
fD Q. CO
O fD Q_ O"U
3CU -h fD O
fD O '—* -S -S 3
3" <"* O —J- trt
-tl fD 3 < fD
c in < fD
3 c-t- fD OJ Q. -5
IQ CL O -h —•
Cu *•< "O Crt ~5 CU
r+- -J-T3 O r+
-•• S 3 fD 3 -••
O fD to O O
3 -i -i. CO 3
-*• O CO 3"
OO 3 —i in -J.
O ->• 00 T3 T3
—' c-t- G) —i. fD CO
fD "^ e-t- O
CO fD d —<• -t>
CO X —'• Cu fD O
T3 3 c-t- (/> -^
c-f O tj. fD
3- CO C CL 3 c-t-
fti fD "5 fD 3"
3 CL^ 3 Cu fD
c-t- CL CU 3
3~ C Cu -S Cu —»•
ro -s c-t- 3
Cu O —'•
O 3
CU O
in cu i
10 3 —. ;
fD -•• fD
Cu O fD rt- —'• O
~ - fD -I 3
-s ro o
a. fD
3
< -h n
fD C O
_ 3 3 '
-a -•• n b -5
fD CO fD O € Cu
-S CU 3 CU fD c-t-
O c-t rt- 3 -S -J.
fD -". -5 ai O
3 O CU CLTD 3
rt- 3 c-t- Cu —'
* _i. . cu O
in o 3 -h
3" rt- 3 c-f
O 3" O OO
S fD 01 C -i. O
3 -S3 3TO
Cu fD C
CU O —> rt- -S
O" c-t- t/l Qj 3- fD
O c: 3- <-t- (D J3
< cu o -J- c:
fD —i S < -J.
3 fD -5
3 . fD
i
cr
fD
-5
S '"
n
§
a
I s
CUMULATIVE!PERCENT OF AFFECTED 'SPECIES
«» 3 S 8 8
~T
O
i
i
3
o
i
§
m
|
201
-------
103
peak (about a 3-minute average), 1-hour average, and 3-hour average
sulfur dioxide concentrations do not exceed 1.0, 0.50 and 0.20 ppm,
respectively." Similarity Dreisinger and McGovern concluded that S02
concentrations should not exceed 0.70 ppm for 1-hour, 0.40 ppm for 2
hours, 0.26 ppm for 4 hours and 0.18 ppm for 8 hours under relatively
sensitive conditions. However, review of their data indicates at least
several species developed foliar injury following exposure to
concentrations below those reported above.
For both studies, all species were not exposed and injured during
any specific fumigation event, and plants were'not always in the most
sensitive condition during each exposure (e.g., exposures when stomates
were open). In the Mclaughlin and Lee study, the greatest percentage of
affected species following a single exposure is about 20 percent.
The data plotted in Figure 7-1 provide relatively consistent
concentration-response relationships indicating that the number of
plants showing first signs of foliar injury affected within a given
concentration range approximates a normal distribution. The slope of
the lines for the Dreisinger and McGovern data appear to be steeper than
the slope of the lines for the Mclaughlin and Lee data, suggesting that
differences in peak to mean S02 ratios, plants exposed, and
environmental conditions may have affected plant response in different
situations. Reports of increased foliar injury when peak to mean ratios
are high suggest that peak rather than mean concentrations are probably
of greatest importance. Dreisinger and McGovern report that plants are
most sensitive during periods of warm, humid weather that are favorable
for plant growth.
-------
104
Several limitations that should be considered when evaluating these
data are listed below.
1) Although these plots show that the concentration required to
produce foliar injury decreases with Increasing duration, it is not
possible to establish definitively which averaging period is best
linked to the production of foliar injury.
2) It is not possible from these data to derive specific estimates of
the amount of leaf surface affected or reductions in growth and
yield for individual species.
3) The data shown in Figure 7-1 have only incorporated results from
species that were affected by known S02 concentrations. In the
Mclaughlin and Lee study SOg related foliar injury was observed in
a total of 196 species over a 20 year period. No foliar injury was
reported in 23 additional species. Figure 7-1 does not reflect
data on the other plants affected by S02 in the Mclaughlin and Lee
study because the specific SOg exposures listed are only associated
with 84 of the 196 affected species. Of a total .of 39 species
observed in the Dreisinger and McGovern study, only 2 species
exhibited no foliar injury at the observed concentrations.
ii) Long-Term Exposures
Most field studies of long-term S02 exposures have focused on
natural ecosystems. Unlike short-term exposures studies, observational
studies of long-term SOg exposure typically focus on different biological
endpoints including the richness and diversity of species in an area,
seedling survival, premature leaf drop, and reduced growth (Linzon, 1978;
Rosenberg'et al., 1979; Winner and Bewley, 1978a,b).
-------
105
Although the importance of short-term peaks relative to long-term
averages in producing effects remains unclear, observational studies
provide strong support that long-term S02 exposures can affect the
diversity and richness of species. A study of species composition in
the vicinity of a 25-year old coal fired power plant demonstrates
reductions in the number of vascular* plant per unit area and species
diversity (Rosenberg et al., 1979). Similarly, Winner and Bewley
(1978a,b) report that both vascular plants and mosses show a marked
increase in understory species diversity with increasing distance from
the source. The criteria document notes "changes in moss communities
were conspicuous and included decreasing values for moss canopy
coverage, moss carpet depth, dry weight, capsule number, and for the
frequency of physiologically active versus inactive moss plants" (CD, p.
8-47).
Table 7-4 summarizes results of long-term ambient S02 exposures. A
difficulty in interpreting long-term ambient exposure studies is
determining whether short-term peaks or long-term concentrations are
most closely associated with the effects observed. In the Canadian
studies on White Pine, Linzon indicates that the maximum 30 minute S02
concentrations are estimated to be about 2.5 ppm at the West Bay site
and 1.0 ppm at the Portage Bay site. Since concentrations even below
these levels may cause foliar injury and other effects following short-
term exposures, the effects observed at the West Bay site may be
associated with repeated peak rather than long-term low-level
exposures. However, both the author of the study and the criteria
document associate the effects observed with long-term exposure average
concentrations of 0.045 ppm S02 for the West Bay site and 0.017 ppm S02
*Vascular plants (e.g., agricultural crops, trees, shrubs) are characterized
by a vascular system that is used in the movement of water and other materials
to different parts of the plant. Conversely, non-vascular plants (e.g.,
lichens, mosses) lack a vascular system.
-------
TABLE 7-4. FIELD STUDIES OF CHRONIC AMBIENT S02 EXPOSURES
Source/Location
Vascular Plants:
Nickel-Copper Smelters/
Sudbury District
Ontario, Canada
(study on White Pine)
West Bay
(19 miles N.E.)
Portage Bay
(25 miles N.E.)
Grassy to Emerald Lake
(40-43 miles N.E.)
Lake Hatimenda
(93 miles W)
Non-Vascular Plants:
Coal Fired Power Plant/
Ohio
Industrial Pollution/
England and Males
Exposure, ppm
(Duration)
0.045*
(10 year average)
•
0.017*
(10 year average)
0.008*
(10 year average)
0.001
(5 month growth
season- 1971)
control for above
0.020
(Annual Average)
0.03 to 0.06
0.02 to 0.03
0.01 to 0.02
0.005 to 0.01
<0.005
[Annual Averages)
Effects
1.3% reduction in net tree volume and
a 2.6% annual average mortality rate,
foliar injury was observed on foliage
from both the current year and
previous years.
0.5% loss in net tree volume and a 2.5%
annual average mortality rate, foliar
injury on current year needles developed
slowly and injury from the previous
year was apparent.
A 1.8* increase in net annual tree
volume and a 1.4% annual average
mortality rate, little foliar injury
was apparent.
A 2.1% increase in net annual tree
volume and a 0.5% annual average mortality
rate.
Two lichen species absent. The
distribution of more resistant became
apparent at annual average concentrations
of 0.025 ppm S02.
% reduction in the number of tolerant
lichen epiphytes:
95
62
24
13
0 .
References
Linzon (1971, 1978)
Showman (1975)
Hawksworth and Rose
(1970)
LeBlanc and Rao (1975)
•Exposures probably included some heavy metals (e.g., As, Cu) that are found in smelter emissions (Chapter 4).
o
cr>
-------
107
for the Portage Bay site. As shown in Table 7-4, Linzon (1978)
indicates that there are differences in the effects observed as
distances are increased from the smelter and thus, peak concentrations
are reduced. The nature of the effects observed suggests that they may
be associated with either continuous or repeated peak S02 exposures
(Linzon, 1978). The presence of metals may have affected plant response
or sensitivity in this study.
Long-term ambient exposure studies show that the community
composition of non-vascular plants* can be affected by exposure to long-
term low-levels of S02- Results of these studies are summarized in
Table 7-4. Similarly, non-vascular plants are also reported to have
responded to long-term S02 levels between 0.005 and 0.03 ppm near
Sudbury, Canada (LeBlanc and Rao, 1975). In contrast to vascular
plants, there are several reasons that non-vascular plants are more
sensitive to S02 and that long-term and not short-term exposures are
probably associated with the effects observed at low long-term S02
exposure levels, including:
1) Sulfur metabolites cannot be eliminated through translocation
(Nieboer et al., 1976). Lichen mapping studies have repeatedly
shown associations between changes in plant communities and the
accumulation of sulfur metabolites (CD, p. 8-53).
2) Pollutants cannot be excluded during periods of high pollutant
levels because they lack both epidermis and stomata (Nieboer et
al., 1976).
3) Non-vascular plants probably have less buffering capacity than
vascular plants (Nieboer et al., 1976).
*See footnote on page 105.
-------
108
For these reasons, non-vascular plant communities appear to be more
sensitive to long-term low-level S02 exposures than vascular plants.
Although investigators have associated effects in lichens with long-term
low-level exposures, the relative importance of repeated pollution
episodes as compared with continuous exposure to low S02 levels cannot
be determined at this time.
Non-vascular plants vary widely in their distribution across the
U.S. and serve a variety of functions in a number of different
environments from the desert to coniferous forests. Non-vascular plants
affect survival of other plant species (CD, pp. 7-9 to 7-13). In some
areas, non-vascular plants participate in the process of soil formation,
prevent erosion, play a role in succession, and can be an important food
source for foraging animals when other vegetation is scarce (Linzon,
1978; Winner et al., 1978). Non-vascular plants can also play an
important role in influencing the movement of materials within the
ecosystem. For example, lichens fix 5 to 20 percent of the total
nitrogen requirement (2 to 11 kg nitrogen/ha) for Douglas fir, the
dominant producer in northwestern coniferous forests (Denison, 1973).
3. Staff Recommendations
S02 induced damage to commercial crops, cultivated ornamentals and
marked changes in natural ecosystems clearly may affect public
welfare. If the recommended ranges of the primary standards do not
provide adequate protection, the staff recommends consideration of
vegetation effects in evaluating the need for both short- and long-term
secondary standards.
a) Short-Term Exposures
Controlled exposure and field observation studies of the responses
of vascular plants to S02 exposure provide support for continuation of a
-------
109
short-term standard to protect vegetation. These studies show a
progression of effects following S(>2 exposure from physiological and
biochemical changes to foliar injury and decreases in growth and
yield. The original 3-hour S02 standard (0.50 ppm), promulgated in
1970, is apparently based on foliar injury in Mountain Ash following
a single controlled exposure to 0.54 ppm (1400 ug/m3) (DHEW, 1970). This
standard was maintained in 1973 following a review of the secondary SOg
standards where additional foliar injury data provided added support for
plant injury to SOa exposure (38 FR 25678; EPA, 1973). '
Available evidence shows that plant responses vary with both
concentration and time. Controlled exposure studies examining plant
response to S02 show that concentration is generally more important in
the production of plant responses which emphasizes the importance of
peak exposures (CD, p. 8-15; Mclaughlin et al., 1979). However, as
exposure periods are extended, concentrations required to produce
effects diminish. Comparison of levels associated with plant injury and
ambient variations in SOg concentrations with time indicate levels that
are reported to injure vegetation are only expected in the vicinity of
point sources. The occurrence of peak S02 levels between mid-morning
and early afternoon, a time when stomates are typically open, increases
the likelihood that plants will be injured by ambient fumigations near
point sources. Because meteorological conditions and emission rates
vary with time, maximum S02 concentrations producing effects following
short-term exposures should be associated with short-term fumigation
periods from less than an hour up to several hours. Controlled exposure
studies (Table 7-5) demonstrate that plants can be affected by S02
exposure for 2 to 4 hours. Thus an averaging period of 3 hours would be
a reasonable choice.
-------
110
The level of a revised standard would depend on the extent to which
anticipated effects would be considered to be adverse. ZAPS field
studies on soybeans suggest multiple peak SOg exposures that do not
exceed the level of the current 3-hour secondary standard can result in
reduced growth and yield. Although few controlled exposure studies at
low exposure regimes have been performed to date, these studies indicate
that single exposures that would not violate the current S02 standard
can increase foliar injury (See Table 7-5) and affect growth and yield
(i.e., 2 hours at 0.5 ppm). Foliar injury data from ambient exposure
studies show greater continuity of response concentrations because of
the larger number of observations. The concentrations associated with
foliar injury for the ambient exposure studies in Table 7-5 were derived
from the observations plotted in Figure 7-1. In each case, the
concentration associated with foliar injury corresponds with the
concentration associated with producing foliar injury in about 20
percent of the species affected. Although foliar injury may be caused
by natural phenomena (e.g., insect damage, unfavorable weather, and
nutrient stress), increases in foliar injury reduces the aesthetic value
of natural vegetation and ornamental plants, as well as, the commercial
value of some leafy agricultural crops (e.g., spinach).
While the acute studies by Tingey and co-workers (1971a,b)
on pollutant mixtures of S02, NOg, and 03 suggest that exposures below
current standard levels may affect foliar injury, the results of these.
studies do not agree with some other work reporting that foliar injury
occurs only at higher concentrations (e.g., Bennett et al., 1975; Heagle
and Johnson, 1979). Ambient exposure studies near point sources shown
in Figure 7-1, where mixtures of NOg and 03 might be expected to be
-------
TABLE 7-5. STAFF ASSESSMENT OF SHORT-TERM VEGETATION STUDIES
Averaging
Period
Controlled Exposure
Results L
Ambient?Exposure
Results^
1-Hour
2-Hour
3-Hour
4-Hour
0.25 ppm - Foliar Injury
0.50 ppm - Reduced Growth and
Productivity
0.40 ppm - Foliar Injury
0.50 ppm - 20% Foliar Injury
0.40 ppm - 20% Foliar Injury
0.32 ppm - 20% Foliar Injury
Derived from Adedipe et al. (1972), Berry (1971, 1974), Jacobson and Colavito, (1976) and Heck
and Dunning (1979). All results from single exposures at stated level and averaging time (See Table 7-1).
2Derived from Dreisinger and McGovern (1970) and Mclaughlin and Lee (19.74). Values provided in
this column are for concentrations that are associated with foliar injury following a single exposure.
Values shown have been interpolated from Figure 7-1.
-------
112
present with S02 in low ambient concentrations, suggest that plant
response is not much different from the response that would be expected
for S02 exposure alone. In essence, these data do not readily support a
combination standard at this time.
Based on the available data, a 3-hour secondary standard at or
below the level of the current standard may be needed to protect
vegetation. A range of 0.4 to 0.5 ppm (1050 to 1300 pg/m3) would be
roughly equivalent to the lowest reported exposure regime affecting
growth and productivity. Such a standard would not be expected to
prevent against all foliar injury but should protect over half of
exposed species against foliar injury even during "sensitive" exposure
periods. Available air quality information indicates that such
concentrations would only be expected to occur in the immediate vicinity
of point sources (Burton et al., 1982; Frank and Thrall, 1982).
Although cultivated crops may be equally sensitive in different regions
if plant water status- is not stressed, natural ecosystems would be
expected to be less sensitive in arid regions (e.g., southwestern U.S.)
where soil moisture and relative humidity tend to be lower. Given the
available information, 1-hour standards in the middle and lower portions
of the range of interest for a primary standard (0.25 to 0.75 ppm) may
provide equivalent or better protection for vegetation than the current
3-hour standard; in this case, the secondary standard may be set equivalent
to the primary standard. In the absence of a primary standard that
provides adequate protection the staff recommends a 3-hour secondary
standard.
-------
113
b) Long-Term Exposures
Available evidence suggests that long-term exposure to S02
may affect foliar injury, growth, productivity, and the number's and
density of species present in plant communities.
Studies on the effects of vascular plants suggest that long-term
exposures may reduce productivity, cause premature leaf drop, and
otherwise affect plants. Although the investigations by Tingey and co-
workers (1971b) and Keller (1978, 1980) suggest decreases in measures of
productivity following exposure to 0.05 ppm (131 pg/m ) S02, other
studies have not reported similar effects until slightly higher
levels. Several ambient exposure studies examining effects on vascular
plants have effectively demonstrated changes in species richness and
diversity, but do not provide specific levels that can be associated
with such changes (e.g., Rosenberg et al., 1979). However, reductions
in tree volume and increased mortality in White Pine have been
associated with 10 year average SOg concentrations as low as 0.017 ppm
(45 yg/m^). Nevertheless, it is unclear if these effects were most
clearly associated with repeated peak or the long-term average exposures
(CD, p. 8-19). The presence of metals may have affected plant response
in this study. The reported damage at the Portage Bay site showed
little injury associated with acute exposure. Moreover, peak half hour
concentrations (< 1 ppm, 2600 ug/m^) reported by Linzon near the Portage
Bay site indicate that the second highest 3-hour concentration probably
did not exceed the current secondary standard level of 0.5 ppm (1300
ug/m3). Thus, it is unlikely that effects observed in this study can be
attributed to a short-term exposure. Although these data suggest that
long-term S02 levels may affect vascular plants, the data are not well
-------
114
enough developed to provide the principal basis for selecting a level
for a long-term secondary standard but do suggest the need to limit
long-term S02 levels.
Mapping studies of non-vascular plant responses provide
convincing evidence that non-vascular plants may be affected by S02 at
low levels. Although non-vascular plants display both visible injury
and changes in community composition, available studies suggest that
species diversity appears to be the most sensitive index of non-vascular
plant response (LeBlanc and Rao, 1975). As Table 7-4 shows, non-
vascular plants may be injured by exposure to annual arithmetic mean
concentrations of S02 from as low as 0.005 ppm (13 ug/m3) to about 0.06
ppm (160 ug/m3); the later concentration can eliminate most species of
epiphitic lichens (LeBlanc and Rao, 1975).
If the data of Mueller et al. (1980) (See Figure 4-2) are indeed
representative of regional S02 levels in the northeast quadrant, then
existing studies suggest that current non-urban S02 levels may have
effects on the richness and diversity of non-vascular plants primarily
in the northeast where rainfall, relative humidity, and soil moisture
are relatively high. Assuming that the deposition and accumulation of
sulfur in non-vascular plants is an important parameter that affects
plant response, the deposition of S02 (northeast seasonal concentrations
of ~ 25 to 50 yg/m3) may be even more important than sulfate (northeast
seasonal concentrations of ~ 6 to 12 yg/m3) deposition in determining
the response of non-vascular plants to sulfur oxides. To the extent
that the contribution of non-vascular plants to the stability and
productivity is considered to be important, reductions in regional S02
-------
115
levels in the northeast may be warranted. Other uncertainties in the
exposure-response relationships of non-vascular plants that should be
recognized include:
1) Since many non-vascular plants live deep within the canopy of their
environment, current monitoring efforts may overestimate actual
concentrations experienced by non-vascular plants.
2) Although existing studies associate effects with long-term averages
and indicate that the mechanism of toxicity is associated with
accumulation of sulfur in tissues, the importance of repeated
episodes on plant response cannot be eliminated at this time.
Since regional loadings of atmospheric sulfur is a problem that will be
addressed during future actions on visibility (EPA, 1982b) and acidic
deposition, the staff recommends that any action on the regional effects
of S02 on non-vascular plants be deferred until that time.
B. Materials Damage
1. Nature of Effects
S02 alone or in conjunction with hygroscopic and other active
components of particulate matter contributes to the physical damage of a
wide range of materials. SOg has been associated with the corrosion of
ferrous and non-ferrous metals, degradation of zinc and other protective
coatings, and with the deterioration of inorganic building materials
(e.g., concrete and limestone), as well as paper, leather goods, works of
historical interest, and certain textiles. SOg is corrosive in the
presence of moisture because of its electrolytic properties that enhance
electrochemical corrosion and because of its acidity. In addition to
moisture, other factors that affect the extent of S02 induced damage
include temperature, sunlight, wind speed, protective measures taken, as
well as the level of SOg to which the material is exposed.
-------
116
Sulfur dioxide-induced materials damage becomes economically
significant, and thus is most likely to affect public welfare only when
one or more of the following conditions are met:
1) the service life of the material is impaired or prematurely ended;
2) the frequency of maintenance tasks must be Increased; or
3) the quantity or quality of the service rendered by the affected
object is diminished (Gillette, 1975).
As discussed in the criteria document (CD, Chapter 10), the evidence on
the effects of SOg on materials is drawn from field and laboratory
studies. In addition, estimates of the economic loss associated with
SOg damage have been made. The key findings from these assessments are
presented below:
1) The evidence presented in the criteria document clearly establishes
that SC>2 (at humidities commonly exceeded in large regions of the
U.S.) can accelerate the corrosion of exposed ferrous metals (Upham
1967; Haynie and Upham, 1971; Haynie and Upham, 1974). With the
exception of weathering steel, these findings have little practical
significance however, since ferrous metals are normally
protected by paint, zinc, or by addition of alloys when exposed
outdoors. Such measures reduce corrosion and in the case of paints
enhance the aesthetic appearance of materials. Certain non-ferrous
metals such as aluminum and copper, however, are relatively
unaffected by levels of S02 currently observed in the ambient air.
(Fink et al., 1971; Abe et al., 1971).
2) Both field and laboratory studies have shown that zinc corrosion is
accelerated by SOg and that time of wetness, SOg concentration, and
surface geometry are determining factors (Guttman, 1968; Haynie et
-------
117
al., 1976; Haynie, 1980). Such findings are potentially
significant because of the use of zinc as protective coatings for
ferrous metals.
3) Controlled exposure chamber'studies by Campbell et al. (1974) and
Spence et al. (1975) have examined the effects of SO2 on several
exterior paints. The results from these studies suggest that
coatings, which do not contain aluminum or magnesium silicates or
calcium carbonate or combinations thereof, are relatively
unaffected by S02 at concentrations up to 1.0 ppm after exposure
periods of 400, 700, and 1000 hours. For an oil base house paint
which contained magnesium silicate, Spence et al. (1975) reported
statistically significant errosion at 0.03 ppm S02 after exposure
periods of nominally 250, 500 and 1000 hours which they attributed
to the effects of S02 and relative humidity. Unfortunately, their
exposure study of acrylic latex house paint had to be terminated
before completion of the exposure cycle. As a result, data on this
widely used coating are not available at S02 concentrations (< 1.0
ppm) that are more relevant to current ambient conditions. The
investigators cautioned that their findings for one coating should
not be generalized to an entire class of paints because of
differences in paint formulations.
4) The studies cited in the criteria document (CD, pp. 10-34 to 10-37)
primarily show that S02 can react with calcareous materials and,
thus, cause deterioration of inorganic materials such as concrete,
marble, and limestone used in buildings, monuments, and some
statuary. These studies have not fully investigated the relative
-------
118
contribution of S02 as compared to other components of acidic
deposition (wet and dry) as well as other physical and biological
agents.
5) The studies of the effects of SOg on textile, leather and paper
goods are less relevant since these materials are not typically
exposed for long periods in the ambient environment. The preservation
of documents and old books is, however, of concern to libraries and
museums (CD, p. 10-37).
6) The criteria document concludes that the dollar value of SOg
associated material damage in 1970 ranged from $450 million to $1.4
billion in 1978 dollars. It was also estimated that the reduction
in the average annual S02 ambient levels in U.S. urban areas from
1970 to 1978 resulted in a total U.S. annual benefit for 1978 of
approximately $0.4 billion dollars (CD, p. 10-71). These
estimates, however, must be viewed with caution since they provide
only a crude indication of the costs associated with S02 materials
damage. Some of the limitations inherent in the underlying
estimates include:
a) the lack of reliable damage functions for the full range of
materials that may be affected by S02;
b) uncertainty as to quantities and distribution of materials in
place, and the level of S02 as well as the environmental
conditions (i.e., relative humidity) to which they may have
been exposed;
c) uncertainty as to whether the useful life of the materials was
actually impaired; and
d) inadequate consideration of substitution possibilities.
-------
119
2. Mechanisms and Quantitative Relationships
Of the materials potentially affected by SOg, the most accurate
damage functions have been developed for zinc coated materials-. Based
on these functions, it is possible to relate the corrosion of zinc to
varying concentrations of SOg under different environmental conditions.
S02 induced damage to zinc coatings may also be economically significant
because zinc is used as a coating for ferrous metals that are to be
exposed in the ambient air. Haynie (1974) attributed 90 percent of the
'Fink et al. (1971) nationwide estimate of exterior metal corrosion costs
(some $1.4 billion) to the corrosion of galvanized (zinc coated) steel.
On the other hand, Stankunas et al. (1981) in their survey of Boston,
Massachusetts, found far less than the average per capita amount of bare
galvanized metal exposed. They reported an annual cost due to SOg corrosion
of only $335,000 for the Boston metropolitan area. No attempt was made
to extrapolate this to a national estimate. They attributed their findings
to development of new coatings that adhere to galvanized metals and which
are aesthetically more pleasing and afford additional protection.
Zinc is anodic with respect to steel and, thus, when .zinc and steel
are in contact with an electrolyte, the electrolytic cell provides
current to protect the steel from corrosion while some oxidation of the
zinc occurs (CD, p. 10-23). In addition, zinc naturally forms a
protective film of zinc carbonate which reduces the rate of corrosion.
SOg or other acidic gases, in the presence of moistupe, however,
destroys or prevents the film from forming, thus resulting in the
acceleration of the corrosion process.
Reporting on the work of a number of investigators, the criteria
document identified several factors that influence the corrosion rate of
-------
120
zinc coated materials. These include: the time the material surface is
wet, SC>2 concentration, wind speed, and geometry of the exposed
surface. The latter factor controls delivery of S02 to the material
surface and, in part, explains why certain surfaces such as zinc coated
chain link fences exhibit higher corrosion rates than zinc coated siding
and roofing materials.
Haynie (1980) incorporated these factors into the following
mathematical expression to determine the corrosion rate for galvanized
materials (zinc coated) under various conditions:
Cz = 2.32 tw + 0.0134 V0-781 (S02) tw, for small sheets (7-1)
Cz = 2.32 tw + 0.0082 V0-781 (S02) tw, for large sheets (7-2)
Cz = 2.32 tw + 0.0183 V0-781 (S02) tw> for wire fence (7-3)
where:
Cz = corrosion (ym)
tw = time of wetness (years)
V = wind velocity (m/s)
o
S02 = sulfur dioxide concentration (yg/nr)
To determine the time of wetness when the average relative humidity is
known, Haynie, 1980 developed the following relationship:
f = (1-k) RH
100-k (RH) (7-4)
where:
f = fraction of time relative humidity exceeds the
critical value (90% for zinc sulfate)
RH = average relative humidity
k = an empirical constant less than unity (k = 0.86)
Based on equations 7-2, 7-3, and 7-4, the annual rate of corrosion
for zinc coated roofing, siding, and wire fence can be plotted as shown
-------
121
in Figures 7-2 and 7-3. This assumes a typical relative humidity of 70%
for inland areas east of the Mississippi River, 80% for the seacoast,
and 30% for desert areas as in California and Arizona. Under these
conditions, the fractional time of wetness is 0.25 for inland eastern
states, 0.35 for the seacoast, and 0.057 for desert areas. A typical
wind speed of 4 meters per second (9 mph) was used.
It is evident from Figures 7-2 and 7-3 that relative humidity as
well as SOg can significantly effect the corrosion of zinc coated
materials. This must be considered in estimating potential nationwide
damages to these materials.
An additional consideration in making such estimates is whether the
service life of the zinc coated material have been impaired or
prematurely ended. This is a function of the coating thickness and
prevailing maintenance practice. Gillette (1975) reported that coating
thickness varies with intended use, ranging from 26 pm for building
accessories and wire fencing to 85 pm for pole line hardware. The work
of Bird (1977) and Haynie (1980), however, suggest coating thickness on
a given piece of material can vary as much as +_ 40 percent. For wire
fencing, the product expected to be most sensitive to corrosion, this
means the first rust could occur after 16 urn of corrosion. Using Figure
7-3, it can be estimated that first rust may occur in wire fencing after
10, 13, or 17 years of exposure at 70 percent relative humidity and
long-term mean SOg concentrations of 0.03, 0.02, and 0.01 ppm (80, 50,
25 pg/m3), respectively. In the absence of S02, first rust would not
occur until 28 years of exposure at 70 percent relative humidity. Thus,
it can be concluded that S02-related damage for this material can occur
-------
I
t 2
100
(0.04)
200
(0.08)
300
(0.11)
400
(0.15)
Annual Mean SO, Concentrations, pg/m (ppm)
100
(0.04)
200
(0.08)
300
(0.11)
400
(0.15)
Annual Mean SO^ Concentrations, ug/m (ppm)
Figure 7-2. Corrosion Rate for Roofing and Siding (Large Sheets).
Relationships based on Equations 7-2 and 7-4 (Haynie, 1980).
Figure 7-3. Corrosion Rate for Wire Fencing.
Relationship's based on Equations 7-3 and 7-4
(Haynie, 1980).
-------
123
down to and at background SOg concentrations found in the eastern United
States. For other zinc coated products, with different surface
geometries and/or coating thicknesses, the time period required for
first rust will be longer.
Depending on the prevailing maintenance practices, the material
could be painted when rust first appears to provide extended protection
or be allowed to deteriorate further until the substrate is damaged to
the point that replacement is necessary. Both approaches are commonly
employed (Stankunas et al., 1981). This uncertainty with respect to
response coupled with the absence of accurate determinations of the
amount of materials actually exposed under varying conditions has
precluded development of complete and accurate national cost estimates.
The work of Stankunas et al. (1981), however, suggests that such costs
may be significant assuming other major metropolitan areas in the country
experience similar costs ($335,000 per year) as estimated for Boston.
The 1976 annual S02 values in Boston ranged from 15 to 48 ug/m3
(Stankunas et al., 1980).
3. Staff Recommendations
The association of SOg with damage to a number of materials suggests
that economically important impacts may result at elevated long-term SOg
concentrations. Control strategies have resulted in marked improvements
in long-term S02 levels over the past 13 years, which the criteria document
associates with substantial benefits (CD, pp. 5-16, 10-71). While the
available data are limited and do not permit definitive findings with
respect to the potential costs or provide clear quantitative relationships
for the full range of potentially affected materials, they generally
support the need for limiting long-term $03 concentrations in urban
areas.
-------
124
Analysis of existing air quality data (Frank and Thrall, 1982)
suggests that without a primary annual standard long-term urban air
quality could deteriorate and in a number of large urban areas might
exceed the current annual standard. Therefore, consideration should be
given to a long-term secondary SO2 standard at or below the level of the
current annual primary standard of 0.03 ppm (80 ug/m^) to protect
against materials damage effects.
C. Personal Comfort and Well-Being
1. Description of Effects, Quantitative Data
In addition to the health effects discussed in Sections V-B, short-
term exposures to SOg may also be associated with other perceived
responses that may not be related to health; examples include odor
perception, eye, nose, or throat irritation, and other non-respiratory
sensory responses. As summarized in Section V-B.l, the available data
indicate that S02 is not detected by human senses at levels below 1 ppm
except under controlled conditions that either tend to underestimate the
threshold of detection, or use indicators of unknown relevance to
personal comfort and well-being (e.g., brain alpha rhythms, eye
sensitivity to light).
A number of investigators have noted subjective perceptions of
irritative or painful effects such as eye, nose, or throat irritation
and cough associated with short-term SOg exposures above 5 ppm (Lehmann,
1893; Yamada, 1905; Greenwald, 1954; Amdur et al., 1953) and as low as
0.5 ppm among asthmatics (Kirkpatrick et al., 1982). Healthy subjects
exposed for 6 hours to S02 had a "threshold of discomfort" at 1 ppm
(Andersen et al., 1974). The addition of 2 mg/m3 "inert" plastic dust
-------
125
to this S02 level increased the level of discomfort and the incidence of
throat and nose irritation (Andersen et al., 1981). Symptoms such as
wheezing and shortness of breath have been observed among asthmatics
after short-term exposures to 0.5 ppm SOg (Table 5-1). These symptoms
are clearly indications of personal discomfort but they are accompanied
by substantial functional responses and are of apparent relevance to
health (CD, p. 13-51). As such they are not considered here.
2. Staff Recommendations
Because the effects relevant solely to personal comfort and well-
being are generally associated with short-term S02 levels above the
range of interest for the primary standard (0.25-0.75 ppm), there is no
support for a secondary standard based on personal comfort and well-
being.
D. Acidic Deposition
On August 20-21, 1981, the Clean Air Scientific Advisory Committee
(CASAC) concluded that the issue of acidic deposition was so complex and
important that a significantly expanded and separate document would be
necessary if NAAQS were to be selected as a regulatory mechanism for
control of acidic deposition. CASAC noted that a fundamental problem of
addresing acidic deposition in a criteria document is that it is
produced by several pollutants (including oxides of nitrogen, oxides of
sulfur, and acidic fine particles). Consequently, a document on acidic
deposition would include various pollutants contributing to wet and dry
deposition. The Committee also recommended that a revised version of
the acidic deposition chapter be retained in the particulate
matter/sulfur oxides and nitrogen oxides criteria documents. In
-------
126
response to these recommendations, EPA is in the process of developing
an acidic deposition document that will provide a more comprehensive
treatment of this subject. Thus, the Issue will not be addressed
directly in this staff paper.
E. Summary of Staff Conclusions and Recommendations
Major staff conclusions and recommendations with respect to
secondary standards (Section VII.A-D) are summarized below:
1) a) Damage to vegetation by 302 resulting in economic losses in
commercial crops, aesthetic damage to cultivated trees, shrubs and
other ornamentals, and reductions in productivity, species richness
and diversity in natural ecosystems constitute effects on public
welfare in impacted areas. Such effects are associated with both
short-term (minutes to hours) and long-term (weeks to years)
exposures to SOg.
b) Given the available data on the acute effects of S02 on plants
(growth and yield and foliar injury), a 3-hour standard at or below
the level of the current secondary standard (0.5 ppm) may be needed
to protect vegetation. If a 1-hour primary standard is chosen that
provides equivalent or better protection, then the averaging time
and level of the secondary standard can be made equal to the
primary standard. In the absence of a primary standard that
provides adequate protection for vegetation, a 3-hour secondary
standard is recommended.
c) Available data on the effects of long-term S02 exposures of
vascular plants (e.g., trees, shrubs) suggest the possibility of
changes in species richness and diversity, reduced growth over
extended periods, and premature needle drop. However, these data
-------
127
are weak and not developed well enough to provide the principal
basis for selecting the level of a long-term SOg standard. Thus,
existing information cannot be used to show significant effects at
annual S02 levels below the current primary annual standard, but
does support the need to protect against the effects of prolonged
S02 exposure by limiting long-term S02 concentrations much above
this level.
d) Current long-term S02 concentrations over large'areas of the
northeast exceed levels that may be associated with effects on non-
vascular plants (e.g., lichens, mosses). Given uncertainties
regarding the extent and importance of these potential effects on
natural ecosystems and the regional character of the exposures, the
staff recommends that the effects of S02 on non-vascular plants be
considered in the larger context of regional acidic deposition -
visibility - fine particle strategies. As such, no separate long-
term secondary standard for non-vascular plants is recommended at
this time.
2} a) Elevated long-term S02 concentrations in the presence of
moisture can damage a number of materials including: exposed
metals, paints, building materials, statuary, paper, leather, and
textiles. Control strategies have resulted in marked improvements
in long-term S02 levels over the past 13 years, which the criteria
document associates with substantial benefits. While the available
data are limited and do not permit definitive findings with respect
to the potential costs of S02 related material damage or provide
clear quantitative relationships for the full range of potentially
-------
128
affected materials, they generally support the need for limiting
long-term S02 concentrations in urban areas.
b) Analysis of existing air quality data suggests that without the
primary annual standard, long-term urban air quality could
deteriorate and in a number of large urban areas might exceed the
current annual standard. Therefore, consideration should be given
to a -long-term secondary S02 standard at or below the level of the
current annual primary standard of 0.03 ppm (80 ug/m3) to protect
against material damage effects.
3) The staff concludes that a secondary SOg standard is not needed to
protect against effects on personal comfort and well-being.
4) The available scientific information indicates that the current 3-
hour and annual standards provide reasonable protection against the
direct welfare effects associated with ambient S02 in general. In
essence, the data support maintenance of SOg standards at or below
levels of the current standards.
5) The acidic deposition issue will not be addressed directly in the
review of the sulfur oxides standards.
-------
APPENDIX A. FACTORS THAT INFLUENCE PENETRATION AND DEPOSITION OF SO?
AND MECHANISMS OF TOXICITY
This material briefly summarizes subject-related and environmental
factors that affect penetration and deposition of S02,.,and mechanisms of
toxicity. This material supports the discussion in Section V.A. of this
paper.
A. Inhalation Patterns
SOg (1 to 50 ppm) is almost completely absorbed.^ 99%) by the
nasal passages under resting conditions in both man and laboratory
animals (Frank et al., 1969; Speizer and Frank, 1966b; Andersen et al.,
1974; Brain, 1970). One study in rabbits also found nearly complete
nasal absorption (95-99%) for 10 to 100 ppm SOe, but reported a decrease
in nasal removal to as little as 40% at lower S02 levels (0.1 ppm)
(Strandberg, 1964). The reason for the apparent difference in absorption
rate at 0.1 ppm S02 is not clear, although methodological difficulties
may have influenced the results. Comparable measurements (< 0.1 ppm)
have not been made in humans, but the criteria document concludes that
nasal removal at lower S02 levels should be similar to that at 21 1 ppm
(CD, p. 1-54).
Subject-related factors that can increase penetration of S02 over
that observed for quiescent nasal breathing include mouth and oronasal
breathing and increased ventilation rates. Forced mouth breathing has
been shown to increase S02 penetration significantly in animals (Frank
et al., 1969). Although direct measurements of S02 penetration for
mouth or oronasal breathing are not available in humans, measurements of
increases in expired S02 (Melville, 1970) and bronchoconstriction
(Speizer and Frank, 1966a; Melville, 1970; Snell and Luchsinger, 1969)
-------
A-2
for mouth only versus nasal breathing provide strong indirect evidence
of increased penetration with oral breathing through a mouthpiece with
the nose occluded. In these studies, the increased bronchoconstrictive
responses for resting oral versus nasal exposure are most marked at high
concentrations (15 to 28 ppm), and not always statistically significant
at lower levels (0.5 to 10 ppm). Comparisons between mouthpiece and
nasal or oronasal breathing in exercising asthmatics (0.5 ppm SOg)
indicate that bronchoconstriction increases when going from nasal to
oronasal to mouthpiece breathing (Kirkpatrick et al., 1982; Linn et al.,
1982a,b). These results indicate a parallel increase in S02 penetration
with route of inhalation, and also suggest nasal removal is less than
complete at high ventilation rates. There remains, however, some question
whether studies using mouthpiece or nose clips overstate oral penetration
of S02 under free breathing conditions and comparable oral ventilation
rates (Cole et al., 1982). Direct measurements of S02 penetration under
these conditions are not available in animals or humans. These matters
are of some interest in quantitative evaluations of the numerous controlled
human S02 experiments that involve mouth only exposures and the use of
nose clips.
Except in the case of blockage of the nasal passages by mucus or
other obstruction, typical "mouth breathing" is better characterized as
oronasal breathing, since the larger portion of the inspired air does
pass through the nose. Under test conditions, about 15% of subjects studied
to date appear to be habitual oronasal or mouth only breathers (Saibene et al.,
1978; Niinimaa et al., 1981). Anyone may shift to oronasal breathing during
conversation, singing, illness with nasal congestion, or exercise. High levels
-------
A-3
of S02 may increase nasal airflow resistance to the extent that the oral
portion of breathing is increased (Andersen et al., 1974). Most
individuals shift to oronasal breathing at flow rates greater than 30-35
1/min (Niinimaa et al., 1981), which is roughly equivalent to moderate
exercise as in walking briskly or bicycling (Cotes, 1979). At this
ventilation rate, about 43% of the inspired air bypasses the nose
(Niinimaa et al., 1981). Table A-l summarizes ventilation patterns
observed in "mouth" breathers and normal subjects.
Exercise or its analogue, deep breathing, can increase penetration
of S02 both through increased ventilation and by causing a shift to
oronasal breathing. Increasing nasal flow rate from 3.5 1/min to 35
1/min increased tracheal penetration of SOg in dogs from 0.01% to about
3.2% (Frank et al., 1969). In the same experiments, the combination of
changing from nose to mouth only breathing and increasing ventilation
rate from 3.5 to 35 1/min increased tracheal penetration of SOg from
0.01% to about 70%. Theoretical (Aharonson, 1976) and empirical (Brain,
1970) evidence indicates that besides increasing penetration of SOg,
higher ventilation rates result in increased deposition of S02 into the
mucosal lining of the upper airways. A number of human studies generally
support the notion that increased ventilation results in increased
penetration and deposition as indicated by enhanced functional responses
and/or increased symptoms (Kreisman et al., 1976; Lawther et al., 1975;
Sheppard et al., 1981a; Stacy et al., 1981). Of particular interest in
this regard is the similarity in both the time course and magnitude of
S02-induced bronchoconstriction observed under modified hyperventilation
and under exercise (Sheppard et al., 1981a).
-------
TABLE A-l. ORONASAL DISTRIBUTION OF INSPIRED AIR
Ventilation
Rate (1/min)
5
10
15
20
30
35.3*
40
45
50
60
80
90
Normal Breathers
Nasal Volume (1/min) Mouth Volume (l/m1n)
5 0
10 0
15 0
20 0
30 0
19.8 15.5
21 19
22.5 22.5
23 27
28 32
32 48
36.5 53.5
"Mouth-Only" Breathers
Nasal Volume (l/m1n) Mouth Volume (l/m1n)
4 1
6 4
8 7
9 11
12 18
13 22
14 26
15 30
17 33
18 42
23 57
~ ~
Representative
Activity2
Sleep
Standing
Walking normally
(2 mph)
Walking slow (1 mph),
carrying 10 Ib. load
Walking quickly (4 mph)
Climbing 3 flights of
stairs, light cycling
and snow shoveling
Light jogging, tennis
(singles)
Climbing 4 Tights of
stairs, moderate cycling,
chopping wood
Light uphill running
(8.6% grade), gymnastics
Moderate Jogging (7 mph)
Heavy cycling, climbing
7 flights of stairs
Heavy exercise,
e.g., basketball, running
(11 mph)
From Nlinlmaa et al. (1981)
From NAS (1958); Morehouse and Miller (1948); Karpovlch (1953); Rossler et al. (1960); Zenz (1975); and Cotes (1979). Activities are derived from
average maximal oxygen consumption and a range of ventllatory rates at each activity for healthy men. Because there 1s great variation In breathing
requirements among people, these can only be considered as approximations.
*Point at which normal breathers switch from nasal only breathing to oronasal breathing.
-------
A-5
Most studies investigating the effects of exercise or hyperventi-
lation used mouth only exposure to S02; the penetration of SOg and
resultant effects for exercise related oronasal breathing are, as expected,
somewhat lower than that reported for mouth only breathing, at comparable
total ventilation rates (Kirkpatrlck et al., 1982; Linn et al., 1982a).
B. Absorption by Particles
An additional factor hypothesized to increase penetration of S02 is
absorption onto particles capable of penetrating to the thoracic regions
(CD, p. 11-37 to 11-38). Direct measurements of the extent to which
penetration might increase are not available. -Theoretical considerations
summarized in Section IV-A suggest that at realistic peak levels of
particle/S02 combinations (_< 1000 ug/m3), only a relatively small amount
of S02 can be attached to or dissolved in particles. Nevertheless, both
animal and human studies have sometimes reported apparent "synergistic"
increases in responses to exposures to S(>2 and certain aerosols (CD,
Tables 12-11 and 13-14). In some cases, the .increased response is
attributed to the formation of sulfuric acid by catalytic metals (Amdur
and Underhill, 1968), but the combinations of S02 and water droplets or
NaCl aerosols used in most animal and human studies are unlikely to
produce an enhanced response through an oxidation mechanism (McJilton et
al., 1976). In this case, droplet aerosols containing H+, HS03", S02,
NH3, and dissolved electrolyte penetrate the thoracic region and result
in discrete "hot spot" deposition characteristic of particles as well as
the more diffuse gas deposition from S02 released from the droplet in
deeper regions.
The mixed results observed with regard to potentiation of S02
effects by non-catalytic dry particles and droplet aerosols appear
-------
A-6
consistent with the extent to which those particles can increase S02
penetration. S02 absorption by dry aerosols is limited by particle
surface area and the amount absorbed is likely to be a small fraction of
particle mass, particularly for coarse particles. High concentrations
(60 mg/m3) of coarse (6 to 8 ym) .NaCl aerosols did not increase response
to 5 ppm S02 in humans (Snell and Luchsinger, 1969). In animal experiments,
10 mg/m3 of dry submicrometer NaCl potentiated response to 2 ppm S02,
but 4 mg/m3 of the same aerosol was insufficient to enhance effects (Amdur, 1961).
In human studies, high levels of submicrometer NaCl (> 7 mg/m3) have either
enhanced (Toyama, 1964) or failed to enhance (Frank et al., 1964) S02 effects.
Humidity was not reported in these studies.
An equivalent mass of neutral pH droplet aerosols can absorb more
S02 than can dry particles and thus increase response. This was apparently
shown in animal exposures to dry and droplet NaCl aerosols by McJilton
et al. (1976), and suggested in human studies by the distilled water-S02
combinations used by Snell and Luchsinger (1969), Conditions that favor
increased S02 penetration through dissolution in droplets include:
1) Higher droplet pH resulting from increased ammonia levels in chambers
'(as in animal studies) or in ambient air can dramatically increase
dissolved S02. Higher ammonia concentrations in the mouth (Larson et
al., 1982) may thus actually increase effective penetration of S02
levels by increasing S02 solubility (CD, p. 2-67). Conversely,
sulfuric acid droplets absorb little $02; 2) High humidity - this
favors formation of hygroscopic aerosols or fog droplets; 3) Low temperature -
S02 is more soluble at colder temperatures; and 4) High droplet aerosol
concentration.
-------
A-7
C. Mechanisms of Toxicity
The major mechanisms of potential interest in S02 toxicity can be
categorized as outlined below:
1. Irritation of tissues or nerve receptors leading to airway
functional changes
Rapid bronchoconstriction (airway narrowing) is the major response
to short-term (< 1-hour) exposures to S02 at realistic peak ambient
levels (CD, p. 1-57). In both human and animal studies, it appears that
this response results from stimulation of irritant or "cough" receptors
located in the larynx, trachea, and more central bronchii and a reflex
contraction of bronchial smooth muscles, as mediated by neural pathways
involving the vagus nerve (Widdicombe, 1954; Nadel et al., 1965; Sheppard
et al., 1981a). The relative importance of the potentially responsible
agent(s) (H20'S02, sulfite, bisulfite, or hydrogen ion) is not known with
certainty. The mechanism(s) of action may include reaction of dissolved
SOg or one of the products of hydration with nerve endings (NAS, 1978, p.
7-10), or irritation of surrounding tissues leading to release of agents
that act on nearby receptors or are transported by the blood (CD, p. 12-14).
The mechanism by which airway constriction occurs for a particular
species may be influenced by the underlying state of health, time course
of exposure, and concentration, which may be more important than duration
of exposure (CD, p. 12-76). Direct action by S02 or locally released
agents is most likely involved in the reflex bronchoconstriction that
develops within minutes and tends to subside after 10 to 15 minutes or
at the end of exposure (Corn et al., 1972; Frank and Speizer, 1965;
Frank et al., 1964, Melville, 1970). Recent evidence suggests that SOg
-------
A-8
might trigger reflex bronchospasm in asthmatics by stimulating the
release of mediators (e.g., histamine) from mast cells (Sheppard et al.,
1981b). Induced release of humoral agents with delayed action or increased
secretion or pooling of mucus might be responsible for the progressive
airway narrowing associated with continued exposure in guinea pigs over
a period of hours (Amdur, 1959) and, after intermediate time periods
(hours - days), in humans (Andersen et al., 1974; Weir and Bromberg,
1972). In contrast to rapid recovery of normal humans to short-term
peaks noted above, responses in guinea pigs may require an hour or more
to return to normal values (Amdur, 1959).
Although evidence is not conclusive, peak or prolonged S02 exposure
may also increase sensitivity of bronchial receptors to subsequent
irritation by other bronchoconstrictive agents, as suggested by small to
moderate increases in reactivity to acetylcholine in dogs (Islam et al.,
1972) and humans (Reichel, 1972) following S02 exposures. At high
concentrations (17 ppm), S(>2 produces a transient depression in respiratory
rates in mice, possibly mediated through a neural reflex following
stimulation of receptors in the nasal region (Alarie, 1973). Available
data do not suggest this is an important mechanism in human responses to
peak ambient levels.
Because the smooth muscles involved in reflex bronchoconstriction
fatigue or become adjusted to altered tone over time and because long-
term penetration of S02 to the tracheobronchial region is limited by
upper airway removal, chronic exposure to SC>2 alone is not likely to
cause permanent changes in bronchial tone (CD, p. 12-19). Indeed, long-
term exposures, alone (< 5 ppm S02) and in combination with dry fly ash
-------
A-9
or sulfuric acid in animals have not produced notable changes in respiratory
function (Alarie et al., 1970, 1972, 1973, 1975). Repeated peak exposures
have not been adequately tested.
2. Alteration of clearance and other host defense systems
This may result in increased susceptibility to infection or contribute
to chronic respiratory disease. The effect of SOg on clearance apparently
varies with region of the respiratory tract, species, and nature of
exposure. SOg may affect clearance rates by affecting physicochemical
'properties of the mucus, increasing mucous secretion, damage to cilia,
or altering deposition patterns through bronchoconstriction. S02 (5
ppm) decreased nasal mucous flow rates in 3^ to 6-hour human exposures
(Andersen et al., 1974). The implications of this response for susceptibility
to infection are, however, unclear (Andersen et al., 1977). Short-term
oral exposure to 5 ppm S02 during exercise enhances exercise induced
increases in tracheobronchial clearance in healthy adults (Wolff et al.,
1975a; Newhouse et al., 1978), but no effects are seen at rest (Wolff et
al., 1975b). Effects of repeated peak and long-term exposures in animals
suggest a biphasic response; slightly accelerated clearance and no
change after several days followed by increasingly decreased clearance
with time or higher SO^ levels (Spiegelman et al., 1968; Ferin and
Leach, 1973; Hirsch et al., 1975). Similar biphasic responses have been
noted for sulfuric acid (at lower concentration than SOg) and cigarette
smoke (Leikauf et al., 1981). Because SOg effects orr clearance may also
be mediated through a reflex response (Wolff et al., 1975a), clearance
may be affected in airways distal to the site of S02 deposition.
Because little S02 penetrates to the pulmonary region, substantial
direct effects on pulmonary macrophages are not expected; none have been
-------
A-10
observed (Katz and Laskin, 1976). From limited infectivity work in
animals, it appears that susceptibility to bacterial infection is not
affected by high S02 concentrations (5 ppm for up to 3 months) (Ehrlich
et al., 1978). Antiviral defenses were Impaired by S02 in mice, but
• •"»»-»
only at high levels (7 to 10 ppm) for 7 days (Fairchild et al., 1972;
Lebowitz and Fairchild, 1973). Long-term exposure to 2 ppm alone and in
combination with 560 pg/m3 of carbon dust for 192 days altered pulmonary
and systemic immune systems (Zarkower, 1972). Effects of the combined
exposure were at most additive and dominated by carbon.
3. Tissue irritation or damage leading to morphological alterations
Because most SC>2 is deposited in the upper.airways, the potential
for damage is greatest in the upper respiratory passages (Giddens and
Fairchild, 1972; Hirsch et al., 1975). High concentrations (repeated
peaks of 50 ppm) produced damage to the mucous secreting cells in the
bronchial airways of rats (Reid, 1970). At lower long-term exposures
(0.14 to 5.1 ppm), S02 alone and in combination with dry fly ash particles
produced no significant morphological alterations in bronchial airways
or alveolar regions of monkeys (Alarie et al., 1972).
4. Reactions with important cellular constituents
The three major reactions of bisulfite (e.g., from dissolved S02)
with biological materials include sulfonation, auto-oxidation, and
addition to cytosine (CD, p. 1-57). A number of in vitro studies have
suggested that these reactions may result in effects on enzyme systems,
but, because major in vivo studies of these endpoints have not been
conducted, no evidence exists indicating effects in whole animals or
humans (CD, p. 12-7). Production of free radicals during auto-oxidation
(CD, p. 12-4) and the reaction with cytosine in DNA are mechanisms by
-------
A-ll
which bisulfite might induce mutation (Shapiro, 1977). SC>2 and bisulfite
are mutagenic in microbial systems at acid pH, but cytotoxicity, rather
than mutagenicity, is the most common response of cultured animal or
human cells (Thompson and Pace, 1962; Kikigawa and lizuka, 1972).,
Animal studies using high concentrations provide some suggestion that
S02 might be a carcinogen or co-carcinogen with benzo-a-pyrene. (Peacock
and Spence, 1967; Laskin et al., 1970, 1976). The evidence is, however,
inconclusive (CD, p. 12-78).
-------
-------
APPENDIX B. EVALUATION OF EVIDENCE FOR EFFECTS ON RESPIRATORY MECHANICS
AND SYMPTOMS
This section discusses and evaluates the key studies providing
information on the effects of S02, alone and in combination with other
pollutants, on respiratory mechanics and symptoms.
A. Controlled Human Studies of S02 Alone
Although differences in aspects such as respiratory function
indicators, subject activity levels, and routes of exposure make direct
comparisons difficult, the major observed response to short-term exposures
of S02 appears to be bronchoconstriction, usually evidenced in increased
pulmonary flow resistance or impaired forced expiratory flow rates. The
data indicate that most healthy adult subjects respond to short-term S02
exposures of 5 ppm or more (Table 5-1, Section V; CD, p. 13-48). At
lower levels, (1-4 ppm) results are more mixed, with a number of studies
reporting otherwise normal individual "hyperreactors." At 1 ppm, Frank
et al. (1962) reported a significant increase in pulmonary flow resistance
in one of 11 subjects during exposures lasting 15 minutes. Also at 1
ppm, Andersen et al. (1974) found a small but progressive decrease in
forced expiratory flow (FEF25_75%) and forced expiratory volume (FEV^g)
and an increase in nasal flow resistance in 15 adults exposed for 6
hours. With intermittent exercise, levels of 0.75 ppm S02 resulted in
decreases in several functional parameters in apparently healthy adults
(Bates and Hazucha, 1973; Stacy et al., 1981). The major responders in
the Stacy work might have been defined as mild atopies. Other work
indicates that such individuals are substantially more responsive to S02
than normal subjects (Koenig et al., 1982a,b; Sheppard et al., 1981a).
Asthmatic subjects appear to respond to S02 at levels an order of
magnitude lower than do healthy adults. Several studies in Table 5-1
-------
B-2
(Section V-B of main text) report marked changes in functional measurements
after short-term (_< 1 hour) oral (mouthpiece) exposure at rest and
during exercise at levels of 1-5 ppm with symptoms such as wheezing and
shortness of breath increasing with concentration. A number of subjects
were reportedly unable to tolerate these exposures. Symptomatic responses
are reported at levels as low as 0.5 ppm for 10 minutes (Sheppard et
al., 1981a; Kirkpatrick et al., 1982). Decreased functional measurements
can occur for oral (mouthpiece) exposures of 0.25 to 0.5 ppm with exercise
(Sheppard et al., 1981a). In this study, the two most sensitive asthmatic
subjects exibited increased airway resistance (without symptoms) for
mouthpiece breathing and exercise at levels of 0.1 ppm.
A number of studies suggest that acute reflex respiratory mechanical
changes induced by S02 take place within a relatively short period of
time, on the order of 1 to 5 minutes (e.g., Frank et al., 1964; Melville,
1970; Lawther et al., 1975). Deep breathing accompanying exercise can
temporarily reverse this bronchoconstriction (Nadel and Tierny, 1961)
and can delay the onset of respiratory changes (Sheppard et al., 1981a)
and otherwise lengthen the period in which decrements occur (Bates and
Hazucha, 1973). Responses are usually maximal within 10 minutes, and may
decrease somewhat and/or remain about constant for exposures of up to
one hour (Melville, 1970). This is in contrast to the steady increase
in resistance reported for guinea pigs over similar periods (Amdur,
1959).
After short-term exposures, recovery to normal functional values
can occur rapidly, in as little as 5 minutes in normal resting subjects
(Lawther et al., 1975), but may take 30-60 minutes for exposures involving
exercise (Bates and Hazucha, 1973), asthmatic subjects with exercise
-------
B-3
(Sheppard et al., 1981a; Koenig et al., 1981), or other sensitive subjects
(Lawther et al., 1975; Gokemeijer et al., 1973). In one study, (Jaeger
et al., 1979) symptomatic responses (wheezing and shortness of breath)
observed in three sensitive subjects (out of 80) did not occur until the
evening following a quiescent oral (nose clips) 3-hour exposure to 0.5
ppm S02. In this study, only very small decreases in one functional
response were seen immediately following exposure in most subjects and
it is not clear how S02, acting alone, could have produced such a delayed
response. Nevertheless, of the three reactors, two were asthmatics who
had been free of wheezing for several months before the experiment and
one (non-diagnosed) had never experienced wheezing previously, suggesting
some involvement of the exposures. One plausible explanation is suggested
by the observations of Islam et al. (1972) and Reichel (1972), outlined
in Section V-A, namely that S02 may sensitize airways to subsequent
bronchoconstrictive challenge. Thus, in this case, the attacks might
have been proximally caused by post-experimental exposures to other
substances. Additional work is needed to determine the extent to which
such S02 levels may increase airway sensitivity.
Three controlled studies of longer S02 exposures (> 3 hrs to days)
suggest the possibility of somewhat different respiratory responses.
Andersen et al. (1974) found that chamber (mostly at rest, nasal breathing)
exposure to 1 ppm S02 produced a progressive decrease in forced expiratory
flow (FEF25_75^) and to a lesser (nonsignificant) extent FEV-| ^ over a
six hour period. Changes were slight after 1 to 3 hours and were small but
significantly larger for one parameter (FEF25_75%), after 6 hours. There was
some suggestion of a carry over of functional decrement after exposure
into the next day ( = 18 hours), although the carry over was more convincing
-------
B-4
after a similar 5 ppm, 6 hour exposure. Some slight discomfort and
symptoms were noted at this higher concentration.
Weir and Bromberg (1972) exposed 12 healthy subjects and a selected
group of seven otherwise healthy smokers with evidence of early small
t >•
airway functional impairment for 5 and 4 days respectively to 0.3, 1,
and 3 ppm of SOg. Subjects were free breathing and may have exercised,
but such occurrences were not reported. No effects were noted at the
0.3 ppm level in either group. In the healthy group, small but significant
decreases in compliance were noted at 1 and 3 ppm. These changes were
maximal at the first measurement (24 hours), decreased to insignificant
levels over the next several days, and finally increased again after 120
hours in the 3 ppm exposure group. Recovery to normal values was complete
48 hours after exposure. The authors suggest that the early changes
were the result of the reflex bronchoconstriction observed in short-term
studies, an effect which can diminish with continued exposure. The
later effects, they argue, may represent a different mechanism of damage.
Responses in the impaired group were reported to be so variable under
both controlled and exposure conditions that consistent responses
(mechanics or symptoms) to SOg were not detectable. Under the conditions
of study (steady multiday exposures, mostly resting) the impaired group
was not more sensitive to the effects of SOg.
Reichel (1972) reported a gradual, but not statistically significant
increase in intrathoracic gas volume in normal subjects exposed for 6
days to 7.7 ppm S02- No trends in airway resistance were noted. All
subjects complained of various symptoms such as rhinitis, conjunctivitis
and throat irritation. In this study, the frequency of increased obstructive
bronchial reactions to acetylcholine increased as a result of the exposure.
-------
B-5
Bronchitic patients were also exposed but to lower levels (1.5 to 3.8 ppm
S02, 4 or 6 days). One of the groups with more serious obstructive disease
appeared to show a decrease in total resistance but no statistics are given
and the discussion suggests that no responses were seen. The results on
bronchitics are confounded because the patients had to be treated by
medicative therapy during the exposure; thus conclusions about their relative
sensitivity at these exposures are not possible.
In these long-term studies, S02 exposures were increased gradually
from control to maximum levels over a period of hours. This apparently
reduces the chance of more severe bronchoconstriction observed in short-
term studies after abrupt increases in concentration over a period of
seconds. Andersen et al. (1974), for example, observed that subjects slowly
brought to 5 ppm reported only mild discomfort while the investigators
entering the 5 ppm chamber from clean air felt "strong discomfort and a
cough which was difficult to suppress but disappeared after a few minutes."
The time course of flow reduction during free breathing chamber exposures
to 0.75 ppm SOg observed by Bates and Hazucha (1973) and Stacy et al.
(1981) suggest that the rapid rise in ventilation brought on by the
onset of intermittent exercise may be equivalent to a sharp increase in
concentration. In the Bates study, MEFR5Q% continued to decrease throughout
a 2-hour exposure with periodic exercise. In the Stacy study, subjects
were exposed at rest for 45 minutes to 0.75 ppm S02 before beginning
exercise. Based orl previous work on resting healthy adults, it is
highly unlikely that constrictive effects would have been observed
during the resting exposure. Despite this extended exposure prior to
exercise, airway resistance still increased significantly following a
15-minute exercise period. This is consistent with expectations in that
-------
B-6
even though external S02 levels remained constant, the effective concentration
at sensitive receptors increased rapidly with the increased oronasal
ventilation associated with exercise.
B. Long-term Exposures to S02 Alone
As discussed in Section V-A, long-term exposures of laboratory
animals to moderate SOg levels have not generally been observed to
produce decrements in respiratory mechanics (CD, Table 12-3). The only
exception is the work of Lewis et al. (1969, 1973), who found increased
pulmonary flow resistance and decreased lung compliance in dogs exposed
to 5.1 ppm SOg, 21 hours a day for 225 days. After 620 days, nitrogen
washout increased, but the other functional parameters were no longer
significantly different. This high level exposure of apparently resting
animals used in this and most animal studies may not adequately assess
the potential long-term effects associated with repeated peak exposure
and intermittent exercise characteristic of human exposures.
C. S02 in Combination with Laboratory Particles, Other Gases
The interaction of S02 with particles and with pollutant gases has
been variously reported to produce responses on respiratory mechanics
and symptoms sometimes exceeding those attributable to the two agents
administered separately, and sometimes not. Enhanced responses may
arise from physical sorption of S02 on or in particles permitting enhanced
respiratory penetration of the gas (Appendix A), chemical reactions to
form irritant pollutants (e.g., stable sulfites or sulfuric acid), and
deposition and combined actions of S02 and particles or pollutant gases
in various sites in the respiratory tract.
Amdur and coworkers studied combinations of S02 and particles in
the guinea pig model already shown to be quite sensitive to both S02 and
-------
B-7
sulfates. The results of fine aerosol/502 mixtures at low humidity,
typically administered for an hour, can be summarized as follows:
1) soluble catalytic aerosols (containing manganese, ferrous iron, or
vanadium) potentiated the response to S02 by reactions in the chamber or
in the animal that apparently formed a more irritant aerosol (e.g.,
sulfuric acid or sulfite complex); 2) certain salt aerosols (NaCl, KC1,
NH4$CN) potentiated SOg response in order of increasing S02 solubility,
but only at unrealistic levels (10 mg/m3) (Amdur and Underhill, 1968);
3) combinations of S02 and the most common atmospheric sulfate species
(ammonium sulfates, sulfuric acid) produced additive responses (Amdur et
al., 1978a), and; 4) S02 and insoluble dry aerosols (carbon, activated
charcoal, fly ash, ferric oxide, manganese dioxide, and motor oil)
produced, at most, additive responses (Amdur and Underhill, 1968; Costa
and Amdur, 1979).
As noted, the Amdur studies were apparently conducted at low
relative humidity (< 70%). Work on guinea pigs by McJilton et al. (1976)
found no response to a one hour exposure to S02 (1 ppm) and sodium
chloride (1 mg/m3) aerosol at low relative humidity (RH), but a marked
potentiation at high RH (80%) that allowed droplet formation. Droplet
pH prior to exposure was 3.8 with no detectable sulfate formation.
Chamber ammonia, which might have increased S02 solubility in the droplets,
was not measured.
A few controlled human studies have examined combinations of 862
and particulate matter, chiefly sodium chloride, water droplet aerosols,
and carbon particles. Dry salt/S02 combinations support the McJilton et
al. (1973) results in animals, that is, moderate levels (_< 1000 pg/m3)
of dry salt particles do not potentiate the effect of S02 (Frank et al.,
-------
B-8
1964; Burton et al., 1969; Snell and Luchsinger, 1969). Some Japanese
work suggest a possible potentiation by dry salt (Toyama, 1962; Nakamura,
1964), but these studies used high salt (7 mg/m3) or S02 (9 ppm) levels,
and, according to the criteria document (p. 13-27) and MAS (1978) used
inadequate functional measurements. In contrast to McJilton and Frank,
however, Koenig et al. (1982a,b) found no evidence of potentiation in
humans using virtually the same S02/NaCl droplet aerosol. Possibly,
this disparity is due to the presence of ammonia in the animal chamber,
but not in the human exposure.
The results of Snell and Luchsinger (1969), provide some controlled
human evidence that water droplet aerosols may potentiate the effect of
S02 by increased penetration and/or hot spot deposition. Increased
effects were suggested for S02/water droplet aerosol combinations, but
the resultant decrease in pulmonary function was independent of S02
concentration; thus enhancement occurred only at lower S02 levels (<_ 1 ppm),
Unfortunately, water droplet levels could not be estimated, ammonia
levels are not known, and the functional measurement techniques used are
of questionable validity (CD, p. 13-51; Kreisman et al., 1976).
Andersen et al. (1981) examined nasal exposure to combinations of
S02 (1 or 5 ppm) and a coarse (2-15 ym) plastic dust (2 or 10 mg/m3)
with and without a surface coating of vanadium oxide. Effects on nasal
mucous flows, nasal air flow resistance, forced expiratory flow
(FEF25-75%) and symptomatic discomfort were, at most, additive. This is
not suprising since most of the coarse aerosol did not penetrate the
nose and adsorption of S02 on dry insoluble aerosols (e.g., carbon) is
probably minimal (Schryrer et al., 1980).
-------
B-9
In summary, although particles may potentiate the effect of S02 by
increased penetration or chemical reaction, controlled human exposures
have found mixed results, with little convincing evidence that-such
enhancement occurs for laboratory aerosol conditions at realistic peak
aerosol levels. Ambient conditions that might tend to maximize interactions
between SOg and particles (cold temperatures, fog droplets, substantial
N02, NH3) have not been systematically examined in the laboratory.
Long-term exposures to S02 in combination with particulate pollutants
have been conducted only in animals, and are evaluated in Appendix B
(Table B-8) of the particulate matter staff paper (EPA, 1982b). These
studies in general have either found no lasting mechanical responses or
the functional decrements appear less important than underlying damage;
deep lung damage in such cases (Gillespie et al., 1980) does not appear
related to the S02 component of the mixtures (EPA, 1982b).
Combinations of S02 and other atmospheric gases have produced mixed
results. Bates and Hazucha (1973) first reported synergistic functional
responses to S02 + 03 (0.37 ppm each), but follow up studies in other
laboratories and with other investigators (Bell et al., 19.77; Horvath
and Folinsbee, 1977; Bedi et al., 1979) did not confirm these results.
Follow-up measurements by Bell et al. (1977) suggested that the system
used in the original study may have produced ultrafine sulfuric acid and
other sulfate aerosol formation (total sulfates as high as 200 yg/m3)
and may have permitted intrusion of ambient particles. The result of a
follow up study by Kleinman et al. (1981) with 0.4 ppm SOg, 03 and 100
yg/m3 sulfuric acid/ammonium sulfate aerosol found, however, only a
small incremental response over ozone alone; the increment was much
smaller than observed by Hazucha and Bates. A recent Japanese study
-------
B-10
(Kagawa and Tsuru, 1979) reports that exposures to S02 + 03, (0.15 ppm
each) results in a greater than additive increase in airway resistance,
but not to the extent seen by Hazucha and Bates. The finding of small
significant responses in this study may be related to differences in
methods of statistical analyses (Bedi et al., 1982). In sunwary, the
available evidence on synergism between 03 and S02 or 03, S02, and
sulfates suggests reason for caution, but the question is unresolved.
Combinations of S02 with N02 and 03 at both high and low levels (CD,
Table 13-5) resulted in no enhancement of functional changes (Linn et
al., 1980; von Nieding et al., 1979).
D. Community Air Pollution
Qualitative examination of short-term respiratory mechanical changes
in people exposed to fluctuating ambient levels of S02 and particles are
limited to two epidemiological investigations. Lawther et al. (1974a,b,c)
made daily measurements of lung function on four normal subjects for five
years, and two bronchitics for one winter in London. Daily variations in
lung function were small and mostly affected by respiratory infections,
although some direct effects of pollution were detected. After multiple
regression analysis, S02 concentrations explained the largest proportion of
variance in peak flow rates (PEFR) and airway resistance (indicated by
MMFR), with clearest associations shown after walking exercise during
episodic periods of heavy pollution. These results are limited because of
the small study group.
Small, but persistent declines in children's lung function (FEV^ Q)
were observed after a pollution episode in Steubenville, Ohio in which
S02 reached high 24-hour levels along with a somewhat lower TSP level
(Dockery et al., 1981). Similar effects were noted following an episode
-------
B-ll
dominated by high TSP levels and moderately high S02. The authors
conclude that this study provides suggestive but inconclusive evidence
regarding short-term (24-hour) changes in air pollution and FEVj^Q.
long-term significance of these changes has not been assessed.
A number of long-term epidemiological studies of chronic effects
have found that populations living in areas characterized by high particulate
matter and S02 tend to have a higher prevalence of respiratory illnesses
and symptoms and lower lung function capability compared with other
groups living in areas with lower pollution levels (see Table B-3, EPA,
1982b). Other qualitative studies discussed below have detected differences
that might possibly be explained by contrasting exposures to S02 rather
than other pollutants. As in most epidemiological studies, however, the
influence of particles and other factors confound interpretation.
Neri et al. (1975) compared adolescents and adults in a large city
with a similar group living near a large smelting operation (Sudbury,
Ontario) that required frequent shutdowns because 24-hour pollution
levels in town often reached "maximum permissible values" (_>_ 790 yg/m3
[0.3 ppm] S02 or _> 400 ug/m3 suspended particulate matter). After
controlling for smoking and occupational differences, residents of the
smelter town had significantly lower lung function than the city dwellers.
The criteria document concludes that very high periodic peak (_< 1 hr)
S02 exposure levels likely accounted more for any pollutant effects than
long-term exposures-to relatively low annual average levels of S02 or
annual mean particulate levels (which did not vary by much between the
two areas) (CD, p. 14-100).
The possibility that the lung function decrements as noted by Neri
et al. may reflect the cumulative effects of repeated peak exposures has
-------
B-12
been examined in other Canadian cross-sectional studies. Comparisons
were made between children (Becklake et al., 1978) and adults (Aubrey
et al., 1979) living near point sources and exposed to high SOg peaks
with similar groups in communities with little S02 pollution. Annual
particle levels (TSP) were moderately high in all areas, with one of the peak
S02 areas having frequent peaks of TSP as well. No significant differences
in respiratory symptom prevalence were noted for either group nor in
pulmonary function among the adults. However, among the children in the
high pollution areas, small airway function (as measured by closing
volume) was impaired, and a trend towards reduced lung function as the
winter progressed was noted. The authors felt that these subtle effects
might reflect the early stages of airway disease in the children exposed
to high S02 peaks with or without accompanying high particulate matter peaks.
Because of the preliminary nature of the study, the results can only suggest
minor effects that might be weakly associated with repeated exposures to
S02.
Van der Lende et al. (1973, 1975, 1981) followed a group of Dutch
adults living in an industrialized town along with a similar group of
rural residents for nine years. The initial cross-sectional comparison
revealed no difference in lung function, although the prevalence of
chronic cough and phlegm, and sputum production, was higher among the
urban residents. Repeated follow-up studies were done at three year
intervals over a period when the urban S02 levels decreased considerably
from high (with frequent peaks) to more moderate levels, and particles
(British Smoke) remained at relatively low levels. After controls for
smoking arid occupational differences, the urban residents showed an
-------
B-13
accelerated decline in lung function over that observed for rural residents.
Difficulties in assessing exposure to particles (calibration of smoke
data) and SOg (representativeness of monitor) limit interpretations of
these results.
-------
-------
APPENDIX C. PULMONARY FUNCTION TESTS USED IN CONTROLLED HUMAN STUDIES
OF SOg1
A. Introduction
Pulmonary function tests can provide objective information to
assist clinical judgment and help in the evaluation of respiratory
impairment caused by air pollutants such as S02. The tests and terms
used to describe various aspects of respiratory function referred to in
this paper are defined in the glossary that follows this section.
These tests principally measure respiratory mechanics, which includes
the combined effects of gravity, elastic recoil, and smooth muscle tone
upon airway caliber, lung volumes and airflow patterns. Several tests
have been developed in an attempt to differentiate among 1) kinds of
effects; for instance airway narrowing or obstruction as in asthma,
bronchitis, and emphysema, and restriction of the thorax, as in fibrosis
or skeletal and chest muscle disorders, and 2) region or site of action;
for instance the large airways (bronchi), small airways (bronchioles),
and gas exchange regions.
General references for this Appendix include: Cotes (1979), Fishman
(1976), Macklem and Mead (1967), Mead et al. (1967), and Wanner (1980).
-------
C-2
B. Glossary of Terms used to Describe Respiratory Mechanics and
its Tests
Airway conductance (Gaw): The reciprocal of airway resistance.
Airway resistance (Raw, pulmonary flow resistance): Resistance to the
flow of gas in the airways (one component of the total respiratory
resistance) measured by body plethysmography. Average values for
normal healthy individuals rtnge from 0.6-2.4 cm H20/l/sec. A
normal resting subject in whoiif bronchial obstruction is induced
does not usually experience symptoms until Raw is increased 3-fold
or more and severe dyspnea may not result until Raw is increased 5-
to 15-fold. Raw and its derivatives (Gaw, SGaw, §Raw) are mainly
affected by alterations in the resistance characteristics of the
large airways.
Alveolar oxygen tension (PAQ2): Partial pressure of oxygen in the alveolar
airspace (reflects amount of oxygen in alveolar air).
Arterial oxygen tension (PAoa): Partial pressure of oxygen in arterial
blood (reflects amount of oxygen in arterial blood).
Bronchoconstriction: Constriction of the airways, may be caused by
neural reflexes or direct effect on smooth muscle.
Bronchospasm: See bronchoconstriction.
Closing capacity (CC): The volume at which, during expiration, the
closure and trapping of gas in the alveolus (caused by increase in
applied pressure and reduction in the diameter of the small airways)
first occurs; CC = RV + CV. The assumption is that with narrowing
of the small airways, closure will occur at a higher lung volume.
Closing volume (CV): Volume of the lungs above residual volume where
the closure and trapping of gas in the alveolus occur (detected by
the change in expired nitrogen gas concentration); CV = CC - RV.
CV is higher with narrowed small airways, as in smokers.
CC: See closing capacity.
CV: See closing volume.
Flow rate from 25-75% expired vital capacity (FEF25_75%): Flow rate
measured from 25% to 75% of expired vital capacity.
Flow rate at 50% forced vital capacity (FEF5Q%): Flow rate measured
with 50% expired vital capacity remaining. Same as MEFR.
Primarily measures intermediate to large airway function.
Forced expiratory volume (FEV): Same as vital capacity.
-------
C-3
Forced expiratory volume at 1.0 sec (FEV^g)1 Fraction of vital capacity
expired in 1.0 second. Reduced in subjects with obstructive respiratory
disease but not in those who have restricted expansion without
obstruction. Mainly affected by alterations in the large airways.
Forced vital capacity (FVC): Same as vital capacity.
Functional residual capacity (FRC): The volume of air remaining in the
lungs at the end-expiratory position. Measures small airway function.
FEV]>o: See forced expiratory volume at 1.0 sec.
FEF25_75%: See flow rate 25-75% expired vital capacity.
FEF50%: See flow rate at 50% forced vital capacity.
FR: See respiratory frequency.
FRC: See functional residual capacity.
FVC: See vital capacity.
Maximum expiratory flow from 40% forced vital capacity (MEF4Q%):
Maximum rate of flow during forced expiration at 40% vital
capacity.
Maximum expiratory flow from 50% forced vital capacity (MEF5Q%):
Maximum rate of flow during forced expiration at 50% vital ca-
pacity. Same as FEF5Q%. Primarily measures intermediate to
large airway function.
Mid-maximal expiratory flow rate (MMFR, MMF,
See maximum expiratory flow from 50% forced vital capacity.
Same as FEF5p%.
MEF 40%: See maximum expiratory flow from 40% forced vital capacity.
MEF 50%: See maximum expiratory flow from 50% forced vital capacity.
MEFR 50%: See maximum expiratory flow from 50% forced vital capacity.
MMFR: See mid-maximal expiratory flow rate.
Nitrogen washout test: Method to assess the uniformity of distribution of
ventilation throughout the lungs; measures residual volume, functional
residual capacity, and closing volume.
PAQ2: See alveolar oxygen tension.
PaQ2: See arterial oxygen tension.
Residual volume (RV): Volume of a gas that remains in the lung
at the end of a complete expiration.
-------
C-4
Raw: See airway resistance.
RV: See residual volume.
Rj: See total respiratory resistance.
Specific airway conductance (SGaw): Extent to which the airways
conduct the flow of gas normalized to the individual's thoracic
gas volume:
ssaw
{jaw
TGV
Specific airway resistance (SRaw): Resistance to flow in the air
ways 'normali zed to the individual's thoracic gas volume: SRau =
Raw x TGV. aw
SRaw: See specific airway resistance
SGaw: See specific airway conductance.
Thoracic gas volume (TGV): Volume of gas in the thorax, whether in
free communication with the airways or not.
Total respiratory resistance (Rj): Total resistance of the respiratory
system consists of three components, airway resistance, lung tissue
resistance, and thoracic cage resistance. Primarily measures large
airway function.
TGV: See thoracic gas volume.
Vital capacity (VC): Maximum volume of gas that can be expelled
from the lungs by forceful effort following a maximal inspiration.
Reduction of VC may be due to loss of distensible lung tissue or
some type of limitation of respiratory movement not related to
disease of lung tissue.
Vmax50%: Maximum flow calculated at 50% of expired vital capacity from
a partial flow volume curve begun from approximately 60% of inspired
vital capacity. Primarily reflects changes in the caliber of the
small airways.
Vmax75%= Maximum flow calculated at 75% of expired vital capacity from
a partial flow volume curve begun from approximately 60% of inspired
vital capacity. Primarily reflects changes in the caliber of the
small airways.
-------
APPENDIX D. ANALYSIS OF ALTERNATIVE AVERAGING TIMES AND EXPOSURE
This appendix summarizes important aspects of several air quality,
modeling, and exposure analyses (Frank and Thrall, 1982; Burton et al.,
1982; Anderson, 1982) that were conducted in direct support of this
staff paper. The information is used to support discussions of alternative
averaging times and population exposure in Sections VI and VII. The
major questions examined include: 1) the usefulness of a 24-hour standard
as a surrogate for a 1-hour standard; 2) the relative protection afforded
by current and alternative standards in the ranges of interest specf^fid
in Section VI; 3) the impact of eliminating the annual S02 standard; and
4) how many people are potentially exposed to 1-hour SOg values of 0.5 to
0.75 ppm, the range of interest used for the August 1982 draft of this
staff paper. The reader is directed to the complete reports for a more
detailed discussion of the issues addressed, methodology, and results.
Further work, particularly in specifying population exposures and examining
the lower bound of 1-hour S02 values recommended by CASAC (0.25 ppm) following
their review of the August draft, is underway in support of decision-making
on the review of the SOg standards.
A. Analysis of SOg Air Quality Data
Frank and Thrall (1982) examined SOg monitoring data collected
during 1979 and 1980 representing 11 million hourly values at approximately
900 monitoring sites. Summary statistics for this data set are given in
Tables 4-1 and 4-2 in Section IV. Sites were classified according to
three types: population-oriented, source-oriented (excluding smelters),
and smelter-oriented. Because of the limited and variable number of
sites per area (typically 1 or 2 for source-oriented sites), differences
in site location criteria and data completeness, and assumptions needed
in scaling data, generalizations based on these analyses should be
derived and used with caution.
-------
D-2
1. 24-hour Standard as a Surrogate for 1-Hour Average
The relationship among various measures of 1-hour and 24-hour standards
is examined in a series of tables and figures (Frank and Thrall, 1982).
The usefulness of a 24-hour standard as a surrogate for a 1-hour standard
depends on the extent to which the 24-hour standard limits 1-hour values to
acceptable levels without causing additional controls in areas where both
24-hour and 1-hour air quality is already acceptable. Table D-l shows
that current control programs already limit the 2nd maximum 1-hour values
in most site-years examined. The distribution of 1-hour values versus 24-hour
values shows that tightening of the 24-hour standard to 0.12 ppm would not
change the 1-hour status of over 60% of the sites with second maximum 1-hour
values in excess of 0.5 ppm. Relaxing the 24-hour standard to over 0.16 ppm
could permit a substantially larger number of sites with second maximum 1-hour
values above 0.75 ppm.
Figure D-l is a graphic display of the same data. The figure indicates
that a 24-hour standard in the 0.12 to 0.14 range would control maximum 1-hour
values at virtually all population-oriented sites. At many smelter and other
source-oriented sites, the current 24-hour standard would be a useful
surrogate only if current air quality were considered acceptable. As indicated
in Figure D-l and in Tables 7 and 8 of Frank and Thrall (1982), current air
quality at many of these sites involves multiple hourly exceedances of 0.5
to 0.75 ppm S02- Reducing the 24-hour standard to 0.12 ppm would not result
in improvements in most of the sites with high 1-hour values, but would call
for additional controls at a number of sites meeting the current 24-hour
standard with no second maximum 1-hour values higher than 0.5 ppm. Thus,
-------
D-3
TABLE D-l. CURRENT STATUS OF SECOND HIGH 1-HOUR VALUES VERSUS SECOND
HIGHEST 24-HOUR VALUES IN TERMS OF SITE-YEARS (FRANK AND
THRALL, 1982).
2nd Highest 1-Hour
Values (ppm)
*
Ol
1/1
5
Population
Source
Ol
J
< 0.25 '
0.26-0.50
0.51-0.75
> 0.75
< 0.25
0.26-0.50
0.51-0.75
> 0.75
<_ 0.25
0.26-0.50
0.51-0.75
> 0.75
< 0.25
0.26-0.50
0.51-0.75
> 0.75
Number of
Site-Years
938
316
73
50
634
109
10
0
264
181
48
19
5
15
14
30
2nd Highest 24-Hour Values(ppm)
<0.12 0.13-.14 0. 15-. 16 .$0.16
938 000
305 533
54 11 4 4
19 7 3 21
634 000
107 2' 00
6 202
0 000
264 000
176 122
39 4 32
10 315
5 000
13 110
9 '5 00
9 4 2 15
* "All Sites" category includes data which could not be classified as
population, source, or smelter.
-------
D-4
I/I
LU
ID
Q
O
2.50-
A = population-oriented sites
B = source-oriented sites
• smelter-oriented sites
• "all sites"
.06
.33
.42
SECOND HIGHEST 24-HOUR VALUE
Figure D-l. Basic relationships between the second highest
1-hour value per year and the second highest 24-hour value
(After Frank and Thrall, 1982). The initial staff ranges of
interest for 1-hour standard (0.5 to 0.75 ppm) and possible
24-hour surrogates (0.12 to 0.16 ppm) are indicated by shading.
Sites with data above the diagonal line(s) would be "controlled"
by 1-hour standards, while those below the diagonals would be
controlled by 24-hour standards. In the hatched area, the
controlling standard would vary. Based on the data in this
figure, the current 24-hour standard would keep 1-hour values
below 0.5 ppm in almost all of the population oriented sites,
but would not prevent many of the source and most of the smelter
sites from having at least two 1-hour values per year in excess
of the indicated range.
-------
D-5
a 24-hour standard that would markedly improve maximum 1-hour air quality at
most sites would be unnecessarily restrictive at sites with acceptable air
quality.
2. Impact of Eliminating the Annual Standard
Table D-2 shows the current status of annual average S02 values versus
the second highest 24-hour values. Few sites exceed the current annual
standard (0.03 ppm). For source (including smelter) sites, attaining
the 24-hour standard would result in attaining the annual standard
(Frank and Thrall, 1982). For population-oriented sites, however, most
of the sites exceeding the current annual standard already meet the 24-
hour standard; therefore, without an annual standard, further improvement
would not be expected.
To further examine the consequences of eliminating an annual standard,
Frank and Thrall (1982) examined the distribution of annual averages
after: 1) adjusting levels downward to reflect attainment of the 24-hour
standard, by itself or in conjunction with the current 3-hour standard
(0.5 ppm) and/or the midpoint of the range of alternative 1-hour standards
(0.5 ppm); and 2) allowing for limited growth up to the level of the 24-hour
standard. Following these adjustments, fewer smelter sites would exceed
the annual standard, but "a substantially larger number of cases exceeding
the annual standard would occur, primarily among the population-oriented
locations" (Frank and Thrall, 1982). The 38 population-oriented sites
that would exceed t'he annual standard are identified in Table 18 of
Frank and Thrall (1982) and include a number of heavily populated cities
and counties, including New York City, Philadelphia, Cook County, IL
(Chicago), and Jefferson, KY (Louisville).
-------
TABLE D-2. CURRENT STATUS OF ANNUAL AVERAGES VERSUS SECOND-HIGHEST
24-HOUR VALUES (ppm) BY SITE TYPE (FRANK AND THRALL, 1982)
Annual Average
ug/m3 (ppm)
< 60 (< 0.023)
61-80 (0.024-0.031)
81-100 (0.032-0.038)
> 100 (> 0.038)
Second Highest 24-Hour Value (ppm)
Population
Total < 0.14 > 0.14
711 710 1
34 34 0
7 6 1
1 10
Source
Total < 0.14 > 0.14
498 487 11
12 8 4
1 1 0
1 1 0
Smelter
Total < 0.14 > 0.14
44 39 5
844
8 3 5 )!
404
-------
D-7
B. Modeling Analysis
Burton et al. (1982) analyzed the relationships among alternative
averaging times through air quality modeling of representative large
point sources. This analysis is.an important supplement to air quality
data analysis because: 1) most emissions limitations for major S0£
sources are based on modeling, rather than monitoring data; 2) monitoring
sites around point sources are generally too limited to capture maximum
short-term values; and 3) models can give estimates of the spatial
extent of potential exposures.
The modeling analysis simulated several hypothetical power plant
situations involving unscrubbed 1000 megawatt (MW) power plants with 400
foot stacks. Flat, rolling, and complex terrain settings were examined.
The CRSTER and COMPLEX I models were used in modes that simulate the
temporal variability in S02 emissions. Details of the analysis are
included in the report (Burton et al., 1982). The preliminary findings
are subject to uncertainties inherent in air quality modeling, apply
strictly .only to the few cases examined, and are sensitive to assumptions
made about implementation approach and background levels. Nevertheless,
they should be illustrative of relationships among averaging times for
large point sources.
The major questions concern the extent to which current S02 standards
limit peak 1-hour concentrations around major point sources. Because
the annual standard is rarely controlling near large point sources
(Frank and Thrall, 1982), only the 24-hour and 3-hour standards were examined
in the modeling analyses. Figures D-2a and b show the geographical extent
and number of expected exceedances of a 1-hour concentration of 0.5 ppm
-------
b)
a
i
oo
Figure D-2. Location and Number of Expected Exceedances per Year of a 1-Hour Average Value of 0.5 ppm for
a 1000 MW Power Plant just complying with, a) the Current 24-Hour SOp Standard, and b) the Current 3-Hour
S02 Standard (Burton et al., 1982)
-------
D-9
when the current standards are just met. Figure D-2a indicates that the
current 24-hour standard would permit yearly second hourly maxima in
excess of 0.5 ppm as far as 12 km (7.2 miles) from the source.' Over 20
exceedances of this concentration occur in areas within 2 to 5 km of the
source. Figure D-2b shows that the 3-hour standard is substantially
more protective in this modeled case. Second maxima in excess of 0.5
ppm are confined to areas less than 5 km from the source with no more
than 3-4 expected exceedances at any site. Figures D-3a and b show the
maximum expected 5 highest concentrations for the same cases. The 24-
hour standard would permit 4 exceedances of a 1-hour concentration of
0.75 ppm, while the 3-hour standard would permit only 1. The second
maximum hourly concentration associated with the 24-hour standard would
exceed 1 ppm, with correspondingly higher peak (5-10 minutes) values possible.
Based on this analysis, the current 24-hour standard would not prevent
multiple exceedances of 1-hour values in the range of 0.25 to 0.75 ppm
at distances of up to 12 km from major point sources. The current 3-hour
standard provides substantially better protection against such excursions
and could itself be a useful surrogate.
The above analysis was for the flat-terrain case. Available data do
not suggest rolling terrain would involve substantially different
conclusions. Results of complex terrain modeling are limited and less
reliable. As noted previously, the results are also sensitive to
a number of modeling assumptions used. In a number Of routine applications,
somewhat greater protection may be provided by the 24-hour standard
than in this analysis.
-------
D-10
a)
b)
EKFECTEO BXFECTEO EIFECTED EIFICTED EIFECTED
HIGHEST SECOND HIGHEST THUD HIGHEST FOMTH HIGHEST FIFTH HIGHEST
HI M-HOUR BVEKKCE
1-HOWt RVERBGE
3-HOUR IWERRGE
EXFECTED EKFECTED CKFECTEO EXFECTED EIFECTED
HIGHEST SECOND HIGHEST THIHO HIGHEST FOURTH HIGHEST FIFTH HIGHEST
24-HOUR flVERKE
3-HOUR RYEIIRCE
1-HOUR dVERRCE
Figure D-3. Maximum 5 Concentrations for a 1000 MW Power Plant just
complying with, a) the Current 24-Hour S0? Standard and b) the Current
3-Hour S02 Standard (Burton et al., 1932).
-------
D-ll
C. Exposure Analysis
A complete analysis of exposures of sensitive populations to SOg
would involve extensive air quality simulations, analyses and modeling
of activity patterns, and improved information on location and activities
of sensitive groups. Available techniques for exposure analysis are
currently being applied to SOg, but no results are available. To provide
some preliminary indication of exposure, the initial results of the
modeling described above were used to estimate the number of individuals
living in areas with potential for at least one hour per year in excess
of 0.5 ppm (Anderson, 1982).
As a first approximation, a screening algorithm was used to identify
those power plants that under several scenarios, might exceed 0.5 ppm
for 1-hour. The census tract data was used to calculate the number of
individuals living within various radii (up to 20 km) of the plants.
Based on the analysis of Figure D-2, most of the exceedances of 1-hour
values of interest will occur within 10 km of the source; hence, potentially
exposed populations will include those living within this distance.
Assuming "current" emission rates, the algorithm indicated 77 power
plants that potentially might cause exceedances of 0.5 ppm. This represents
a total potential exposure of 7 to 8 million people (one person exposed
to two plants is counted twice) who live within 10 km of these sources.
If asthmatics and atopies comprise the same proportion as they do in the
U.S. population at large (about 5 to 10%), then several hundred thousand
sensitive individuals live in locations where, with current emissions they
might be exposed. Because a number of sources actually emit lower amounts
than needed to meet regulations, exposures would increase if each source
just met its emission limits. Assuming all plants just met current S02
-------
D-12
emission limitations, between 7 and 21 million people (up to 10% of the
U.S. population) may live within 10 km of sources capable of causing
hourly exceedances of 0.5 ppm. Thus, assuming asthmatics and atopies
are distributed in geographical patterns similar to the population at
large, up to 10% of the sensitive population may live 1n areas that
would be permitted to exceed 0.5 ppm at some location, at least once per
year.
The above estimates are crude and tend to overstate exposures
because of 1) potential multiple counting, 2) conservative assumptions
in the screening algorithm, 3) at any given plant/time, the area in
excess of 0.5 ppm will be much smaller than the area of the 10 km radius
circle, and 4) indoor/outdoor differences, activity patterns and the
like substantially reduce the numbers of exercising asthmatics that
actually encounter a peak S02 value.
-------
APPENDIX E. CASAC CLOSURE LETTER
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
August 26, 1983
OFFICE OF
THE ADMINISTRATOR
Honorable William D. Ruckelshaus
Administrator
Environmental Protection Agency
Washington, D.C. 20460
Dear Mr. Ruckelshaus:
The Clean Air Scientific Advisory Committee (CASAC) has
completed its second and final review of the revised draft Office
of Air Quality Planning and Standards (OAQPS) Staff Paper entitled
Review of the National Ambient Air Quality Standards for Sulfur
Oxides; Assessment of Scientific and Technical Information.
The document is consistent in all important aspects with
the scientific evidence presented and interpreted in the combined
criteria document for sulfur oxides and particulate matter. It
has organized the data relevant to the establishment of sulfur
dioxide primary and secondary ambient air quality standards in a
logical and compelling way, and the Committee believes that it
provides you with the kind and amount of technical guidance that
will be needed to make appropriate decisions about revisions to
the standards.
During the course of the Committee's review of the Staff
Paper for Sulfur Oxides a number of significant scientific issues
related to the establishment of primary and secondary standards
were addressed. A Teview of the existing data base for this
pollutant led the Committee to conclude that there are two scienti-
fically supportable options for revising the existing standards.
One option for which there is strong but not unanimous support
on CASAC includes the following: establishment of a new 1-hour
primary standard in the range between .25-.75 parts per million,
retention of a 24-hour primary standard, conversion of the
current .03 ppm annual primary standard to an annual secondary
standard at or below that level, and selection of a revised 3-hour
secondary standard between a range of .40-.50 ppm. The other
option for which there is some support on the Committee is to
retain the existing primary and secondary standards, while
providing some additional public health protection by converting
the existing 3-hour secondary standard into a primary standard.
The choice between these options is a policy decision which is
not within the scope of the Committee's mission. CASAC's wishes
to inform you that either of these options would be supported by
the available scientific evidence.
-------
-2-
Other scientific issues and studies of interest-to the
review and possible revision of the primary and secondary
standards are reviewed in the attached report. In addition,
I have attached a recent CASAC report on research needs for
the gases and particles program within the Agency. It is
clear that there are major gaps in our understanding of
these pollutants and that the Agency should develop a more
balanced and more adequately funded research program.
I hope the CASAC1s findings and recommendations prove
useful to you as you review and consider revisions to the
sulfur dioxide standards. The Committee appreciates the
opportunity to advise you on this important issue, and it
will provide further review and comment to you during the
public comment period that follows the proposal of revised
standards in the Federal Register.
Sincerely,
Bernard D. Goldstein, Chairman
Clean Air Scientific Advisory
Committee
Attachment
cc: Alvin Aim
Charles Elkins
Terry F. Yosie
-------
- f
Findings, Recommendations and Comments
of the Clean Air Scientific Advisory Committee on the
OAQPS Revised Draft Staff Paper for Sulfur Oxides
CASAC's evaluation of the scientific basis for a review
and possible revision of the ambient air quality standards
for sulfur dioxide began with its recommendation in November
1978 that the Agency evaluate the joint interaction of sulfur
oxides and particulate matter on human health and the
environment by the development of a joint criteria document
for these pollutants. Following three public reviews of the
criteria document and its subsequent revision by Agency staff,
the Committee concluded in a letter to the Administrator
dated January 29, 1982 that the Agency's assessment of the
existing literature for these pollutants was scientifically
adequate. This report addresses the OAQPS staff's interpretation
of the criteria document and the scientific rationale that is
develped to support their proposals for reviewing and revising the
SO2 standards. ~
The Scientific Basis for Primary SOj Standards
1. A major OAQPS conclusion of the criteria document
review process was that sulfur dioxide continued to pose a
serious health problem to important subgroups of the population
which warranted its continued separate control. Thus, OAQPS
does not recommend a joint SO2/particles primary standard,
believing that current information on health effects and U.S.
exposures to these two pollutant categories warrants a
continuation of separate controls.
-------
-2-
CASAC concludes that separate SC>2 and particles standards,
each set with appropriate consideration for potential interactions,
does appear to protect public health. Furthermore, the complexities
of setting and implementing a joint SO2/particles standards
through monitoring and other requirements create numerous uncertain-
ties which the available scientific evidence is ill-equipped to
resolve. CASAC concurs with the OAQPS position and its supporting
rationale and recommends that you retain the current approach of
setting separate primary and secondary standards for sulfur dioxide
and particulate matter.
2. The scientific basis for a 24-hour standard stems primarily
from epidemiological studies. These studies (Lawther et al. 1970
[analysis of bronchitics]; Martin and Bradley, 1960, Mazumdar et
al., 1981, and Ware et al., 1981 [analysis of mortality]) do not
show evidence of clear thresholds, but they suggest that risk to
public health increases as concentration levels increase. The Air
Quality Criteria Document for Sulfur Oxides/Particulate Matter and
the SC>2 staff paper interpret these studies as suggesting that
increases in excess mortality occurred in the range of 500-1000
ug/m3 British Smoke and .19-.38 ppm SO2, and that such effects
are most likely when both pollutants exceeded 750 ug/m3 (.29 ppm
SO2>. Lawther"s study of reported symptoms among bronchitics
also suggests that this population group experiences significant
responses associated with 24-hour averages of .19 ppm SO2- Based
-------
-3-
upon these studies and the need for a margin of safety the
staff paper developed a range of interest between .14 to .19
ppm in recommending a revised 24-hour primary SO2 standard.
The upper end of the recommended range of .14 to .19 ppm
represents a level at which effects are identified in the
criteria document and for which there is little or no margin
of safety for exposed sensitive individuals. You should be aware
that the ranges of interest developed in the staff paper for the
24-hour standard were based on epidemiological studies which
provided quantitative concentration/response data of the
populations studied. A final decision on whether or not to
revise the 24-hour standard should also incoporate information
generated through controlled human, animal toxicology and the
less quantitative epidemiology studies discussed in the criteria
document and staff paper. In view of all of the above, CASAC
recommends that you consider "selecting a value at the lower
end of the range for the 24-hour standard, taking into account
whether a separate 1-hour primary standard is also established.
3. CASAC's review of the scientific evidence related to
the annual primary standard presents a dilemma because the
Committee could find no real quanitative basis for retaining
this standard. This is a troublesome issue because there is
the possibility that repeated S02 peaks of 1-hour and 24-hour
exposures might lead to effects on human respiratory systems
-------
-4-
over the long-term. Second, an annual primary standard
affords protection against health effects that can't be
measured well in short-term controlled human studies. Third,
air quality analysis conducted by OAQPS staff suggests that
1-hour and 24-hour primary standards in the range stated in
the staff paper would not prevent SO2 concentrations from
exceeding the current annual primary standard in some heavily
populated areas of the country. Fourth, as pointed out in
the discussion of secondary standards, there is a scientific
basis for a secondary standard at the level of the annual
current primary standard. Following extended discussion the
Committee concluded that some protection against chronic SC>2
exposures is needed, but that the most persuasive scientific
basis for an annual standard is found in the effects on welfare.
4. The scientific basts for the development of a 1-hour
primary standard rests largely on several major controlled
human clinical studies conducted by three separate laboratories
that were published in the peer reviewed literature in 1981
and 1982. These studies documented measurable changes in
respiratory function of exercising asthmatics exposed for short
periods at or below concentration levels of .50 parts per
million (ppm). The studies (Kirkpatrick et al. 1982; Koenig
et al. 1982; Linn et al. 1982; and Sheppard et al. 1981) raise
-------
-5-
the issue of how adequately the existing primary standards are
protecting public health and provide a scientific basis for a
1-hour primary standard that provides additional protection
against such reported short-term effects.
The OAQPS staff, after reviewing this data, proposed
consideration of a 1-hour primary standard in the range
between .50 to .75 ppm. The staff noted that the lower end of
the range represented the lowest level where potentially
significant responses in asthmatics have been observed with
oronasal breathing, and that the upper bound of the range
represented levels at which the risk of significant functional
and symptomatic responses in exposed asthmatics and other
sensitive groups appeared high.
CASAC has evaluated the OAQPS staff position that resulted
in the establishment of the range of interest at .50-.75 ppm.
The staff suggest that there may be little or no margin of
safety at the upper bound of the range. Air quality analyses
conducted by OAQPS also indicate that a 1-hour standard selected
from within the range would still permit exposures in excess
of one to two ppm during the peak five or ten minute intervals.
A related point is that establishment of a 24-hour standard
in the range of .14-.19 ppm would not necessarily protect
against shorter term peaks above the proposed 1-hour range
of .50-.75 ppm. This information suggests that a 1-hour
primary standard selected between .50-.75 ppm range might
-------
-6-
not adequately protect sensitive populations with an adequate
margin of safety from the effects acknowledged in the staff
paper that would occur as a result of brief peak exposures
to concentrations greater than the .50-.75 ppm hourly average
that a 1-hour standard would permit. Because five to ten
minute peaks can reach levels as much as two or more times
.the 1-hour average, CASAC recommends that the range be modified
to state the lower bound at .25 ppm.
In reviewing the issue of whether to establish a 1-hour
primary standard between .25-.75 ppm several additional
factors should be considered. These include 1) it is not
clear that the reported effects experienced at or below .50
ppm are significant. The functional changes and symptoms
reported in the .50-.75 ppm range appear to be reversible.
You will need to determine which effects you consider to be
adverse; 2) it is probable that some asthmatics are more
sensitive than those who took part in the studies; 3) given
current air quality conditions there is a low probability of
exposure to exercising asthmatics at peak concentration levels;
and 4) as the staff paper suggests, other stimuli interacting
with SC>2, such as temperature and humidity, may increase the
risk of an attack to exercising asthmatics more than either
of these factors acting alone.
-------
-7-
The Scientific Basis for Secondary SO? Standards
The kinds of effects reviewed by CASAC in relation to
the establishment of secondary ambient air quality standards
include those on vegetation, materials, and acidic deposition.
1. Current scientific information documents effects on
vegetation resulting from both short-term and long-term
exposures to SC>2 and/or SO2 in combination with other
pollutants. One should keep in mind that there is no single
concentration at which all species of plants are injured,
just as there is not single point or threshold at which all
humans suffer significant effects from SO2- What is at issue
in the development of secondary standards is the need to
protect sensitive vegetative species from effects such as
physiological and biochemical changes, foliar injury, and
reduced growth and yield. The available studies of SC>2
effects on vegetation represent approximately one percent of
total plant species, but they include such important species
as soybeans, barley, and white pine, to name a few.
An issue of increasing concern in the protection of
vegetation is that SC>2 is not present alone in the ambient
air except at a few isolated point sources. It almost
invariably occurs in the presence of other pollutants,
primarily nitrogen oxides and ozone. The scientific evidence
is conclusive that the combination of such pollutants is more
damaging to vegetation than the presence of SC>2 a.lone.
-------
-8-
The staff paper recommends consideration of a 3-hour
standard at or below the current secondary standard
level of .50 ppm to protect vegetation. Although there are
reports in the literature concerning plant injury at .10 to
.20 ppm averaged over several hours, there are great
uncertainties associated with the effects of the exposures
at these lower levels. The existing data on the acute effects
of SO2 on vegetation suggest to CASAC that a concentration
limit selected within a range of .40 to .50 ppm for a 3-hour
period would provide adequate protection to sensitive
vegetative species.
The review of longer term effects on plants was hampered
by a very limited data base, thus making it difficult to
distinguish whether such effects resulted from -chronic lower-
level exposures or a series of shorter-term peak exposures.
Available data do suggest, however, that changes in species
diversity and reduced growth in vascular plants are effects
that may occur over the long term. In addition, non-vascular
plants, particularly lichens and mosses, are affected by SC>2
during prolonged periods of exposure. On the basis of
scientific work conducted to date, CASAC concurs with the
OAQPS staff recommendation that an annual secondary standard
at or below .03 ppm (a level equivalent to the existing annual
primary SOj standard) would afford adequate protection to
vascular plant vegetation. The basis for concern over effects
-------
-9-
in non-vascular plants at lower levels needs to be strengthened.
CASAC also agrees with the staff proposal to address this
issue in the context of later action on fine particles and
acidic deposition.
2. The action of SC>2 alone or in combination with other
pollutants has been associated with a number of damages to
building materials, corrosion of ferrous and non-ferrous
structures, and impairment of other goods and materials.
OAQPS staff have reviewed the evidence documenting
materials damage from SC>2. These effects are responsible for
economically significant losses which have been adequately
summarized in both the criteria document and the staff paper.
Analyses of existing air quality data by OAQPS indicate that
continued protection against SO2~induced materials damage is
-~y
needed, and toward that end, the staff paper recommends
consideration of a long-term "SO2 standard at or below the
level of the existing annual primary standard (.03 ppm).
CASAC concurs with the staff recommendation.
3. Throughout its review of both the Air Quality Criteria
Document for Sulfur Oxides/Particulate Matter and the Staff
Paper for Sulfur Oxides, CASAC has recognized the complexity
of the acidic deposition problem. Since SO2 is only one of the
-------
-10-
precursor pollutants that lead to the formation of acidic
deposition, CASAC recommended in August 1980 that EPA prepare
a separate Critical Assessment Document that recognizes and
incorporates information on causes, effects and data bases
for all of the various pollutants relevant to acidic deposition.
This CASAC recommendation was accepted by two previous
Administrators, Douglas Costle and Anne Burford, and the
assessment document should be available for CASAC review in
the near future. At that time the Committee will be in a
position to provide a more comprehensive and critical
assessment of the acidic deposition problem.
Re-affirmation of the Existing Primary and Secondary Standards
Throughout its review of the staff paper, CASAC
recognizes that large uncertainties exist in Jthe data that
support.development of the options for setting the standards
discussed in the previous pages. Given these uncertainties
CASAC discussed the extent to which the existing standards
provide adequate protection to the public health. The
Committee recognizes the substantial improvements in air
quality that have occurred since the 1971 promulgation of the
primary SC>2 standards. In addition, more information on the
effects of the short-term SC>2 exposures should become available
in the peer reviewed literature in the next few years. Air
quality modeling analyses also suggest that attainment of the
proposed 24-hour and annual standards would not ensure complete
-------
-11-
attainment of the proposed 1-hour primary standard at all sites
within the ranges of interest stated. The reverse also
appears to be true.
CASAC's evaluation of the scientific evidence -associated
with existing averaging times in the staff paper leads the
Committee to conclude that continuation of the existing
primary and secondary standards also provides protection
aganist the effects identified in the criteria document and
staff paper from SC>2 at ground level. If you choose to
follow this option some CASAC members suggest that additional
health protection can be obtained by converting the existing
3-hour secondary standard into a primary standard. A principal
argument supporting the latter is that since the states are
already implementing a 3-hour secondary standard, conversion
---/
to a 3-hour primary standard would not be impractical. In
summary, in view of the many'uncertainties that pertain to
the review of the SC>2 standards, retention of the existing set
of primary and secondary SO2 standards is an option that you
ought to seriously consider at the present time.
Conclusion
CASAC recognizes that your statutory responsibility to
set standards requires public health policy judgments in
addition to determinations of a strictly scientific nature.
The submission of this closure letter completes the Committee's
-------
-12-
scientific assessment of this pollutant and we see no need to
provide any additional formal comnTents on the standards prior to
their proposal in the Federal Register. The public comment
period will then provide sufficient opportunity for the
Committee to provide any additional comment or review that
may be necessary.
-------
-------
REFERENCES
Abe, H., Y. Ishii, and H. Kato (1971). Evaluation of atmospheric factors
by analyses of corrosion products and surface deposits on copper
plates. Rail. Tech. Res. Inst. 12:170-174.
A.D. Little Inc. (1968). Research on Chemical Odors. Part I. Determination
of Odor Thresholds for 53 Commercial Chemicals. The Manufacturing
Chemists' Association, Washington, DC.
Adepipe, N.O., R.E. Barrett, and D.P. Ormrod (1972). Phototoxiclty and
growth response of ornamental bedding plants to ozone and sulfur
dioxide. J. Am. Soc. Hortic. Sci. 97:341-345.
Aharonson, E.F. (1976). Deposition and Retention of Inhaled Gases and
Vapors. In; Air Pollution and the Lung. E.F. Aharonson, A. Sen-David,
and M.A. Klingberg, eds., John Wiley and Sons, New York. pp. 13-24
Alarie, Y. (1973). Sensory irritation by airborne chemicals. CRC Crit.
Rev. Toxicol. 2:299-363.
Alarie, Y., W.M. Busey, A.A. Krurrnn, and C.E. Ulrich (1973). Long-term
continuous exposure to sulfuric acid mist in cynomolgus monkeys and
guinea pigs. Arch. Environ. Health 27:16-24.
Alarie, Y.C., A.A. Krumm, W.M. Busey, C.E. Ulrich, and R.J. Kantz, II. (1975).
Long-term exposure to sulfur dioxide, sulfuric acid mist, fly ash,
and their mixtures. Results of studies in monkeys and guinea pigs.
Arch. Environ. Health 30:254-262.
Alarie, Y., C.E. Ulrich, W.M. Busey, A.A. Krum, and H.N. MacFarland (1972).
Long term continuous exposure to sulfur dioxide in cynomolgus
monkeys. Arch. Environ. Health 24:115-128.
Alarie, Y., C.E. Ulrich, W.M. Busey, H.E. Swann, Jr., and H.N. MacFarland
(1970). Long-term continuous exposure of guinea pigs to sulfur
dioxide. Arch. Environ. Health 21:769-777.
Amdur, M.O. (1973). Animal studies. In: Proceedings of the Conference
on Health Effects of Air Pollutants, Washington, D.C. October 3-5,
1973. A report prepared for the Committee on Public Works, United
States Senate. Serial No. 93-15. U.S. Government Printing Office,
Washington, D.C. pp. 175-205.
Amdur, M.O. (1961).- The effect of aerosols on the response to irritant
gases. _!TK Inhaled Particles and Vapors. C.N. Davies, ed., Pergamon
Press, Oxford, England, pp. 281-294.
Amdur, M.O. (1964). The effect of high flow-resistance on the response
of guinea pigs to irritants. Am. Ind. Hyg. Assoc. J. 25:564-568.
Amdur, M.O. (1959). The physiological response of guinea pigs to
atmospheric pollutants. Int. J. Air Pollut. 1:170-183.
-------
Amdur, M.O. (1974). The long road from Donora. 1974 Cummings Memorial
Lecture. Am. Ind. Hyg. Assoc. J. 35:589-597.
Amdur, M.O., W.W. Melvin, Jr., and P. Drinker (1953). Effects of Inhalation
of sulfur dioxide by man. Lancet 2:758-759.
Amdur, M.O., and D.W. Underhill (1970). Response of guinea pigs to a
combination of sulfur dioxide and open hearth dust. J. A1r Pollut.
Control Assoc. 20:31-34. *>
Amdur, M.O., and D. Underhill (1968). The effect of various aerosols on
the response of guinea pigs to sulfur dioxide. Arch. Environ. Health
16:460-468.
Amdur, M.O., J. Bayles, V. Ugro, and D.W. Underhill (1978a). Comparative
irritant potential of sulfate salts. Environ. Res. 16:1-8.
Amdur, M.O., V. Ugro, and D.W. Underfill 1 (1978b). Respiratory response of
guinea pigs to ozone alone and with sulfur dioxide. Am. Ind. Hyq
Assoc. 39:958-961.
Andersen, I., P.L. Jensen, S.E. Reed, J.W. Craig, D.F. Proctor, and
G.K. Adams (1977). Induced rhinovirus infection under controlled
exposure to sulfur dioxide. Arch. Environ. Health 32:120-126.
Andersen, I., G.R. Lundqvist, P.L. Jensen, and D.F. Proctor (1974). Human
response to controlled levels of sulfur dioxide. Arch. Env. Health
Andersen, I., L. Mohave, and D.F. Proctor (1981). The human response
to controlled levels of sulfur dioxide and inert dust. Scand J
Work Environ. Health 7:1-7.
Anderson, G.E. (1982) Systems Applications, Inc., San Rafael , CA. Personal
communications with Henry Thomas (July, 1982).
Ashenden, T.W. (1979). The effects of long-term exposures to SO? and
N02 pollution on the growth of Dactyl is glomerata L. and Poa
pratensis L. Environ. Pollut. 18:249-258.
Ashenden, T.W., and T.A. Mansfield (1977). Influence of wind speed on
the sensitivity of ryegrass to S02. J. Exp. Bot. 28:729-735.
Aubrey, F., G.W. Gibbs, and M.R. Becklake (1979). Air pollution and
health in three urban communities. Arch. Environ. Health
34:340-368.
Bates, D.V., and M. Hazucha (1973). The short-term effects of ozone on
the lung. _[n: Proceedings of the Conference on Health Effects of
Air Pollutants, National Academy of Sciences, Washington, DC
October 35, 1973. Serial No. 93-15, U.S. Senate Committee on Public
Works, Washington, DC, pp. 507-540.
-------
Becklake, M.R., J. Soucie, G.W. Gibbs, and H. Ghezzo (1978). Respiratory
health status of children 1n three Quebec urban communities. An
Epidemiologic Study. Bull. Europ. Physiopath. Resp. 14:205-221.
Bedi , J.F., L.J. Folinsbee, S.M. Horvath, and R.S. Ebenstein (1979).
Human exposure to sulfur dioxide and ozone: absence of a synergistlc
effect. Arch. Environ. Health 34:233-239.
Bedi, J.F., S.M. Horvath, and L.J. Folinsbee (1982). Human exposure to
sulfur dioxide and ozone in a high temperature-humidity environment.
Am. Ind. Hyg. Assoc. J. 43:26-30.
Bell, J.N.B., and W.S. Clough (1973). Depression of yield in ryegrass
exposed to sulfur dioxide. Nature 241:47-49.
Bell, K.A., W.S. Linn, M. Hazucha, J.D. Hackney, and D.V. Bates (1977).
Respiratory effects of exposure to ozone plus sulfur dioxide in
Southern Californians and Eastern Canadians. Am. Ind. Hyg. Assoc. J.
38:696-706.
Bennett, J.H., A.C. Hill, A. Soleimani , and W.H. Edwards (1975). Acute
effects of combinations of sulfur dioxide and nitrogen dioxide
on plants. Environ. Pollut. 9:127-132.
Berry, C.R. (1974). Age of pine seedlings with primary needles affects
sensitivity to ozone and sulfur dioxide. Phytopathology 64:207-209.
Berry, C.R. (1971). Relative sensitivity of red, jack, and white pine
seedlings to ozone and sulfur dioxide. Phytopathology 61:231-232.
Biller, W. F., and T. B. Feagans (1981). Statistical forms of National
Ambient Air Quality Standards. Presented at the Envi ronmetrics
1981 Conference (April 8-10), Alexandria, Va.
Bird, C.E. (1977). Corrosion behavior of galvanized sheet in relation to
variation in coating thickness. Matter Prot. and Perform. 16:14-16.
Boushey, H.A. , M.J. Holtzman, J.R. Sheller, and J.A. Nadel (1980).
Bronchial hyperreactivity. Am. Rev. Respir. Dis. 121:389-413.
Brain, J.D. (1970). The uptake of inhaled gases by the nose. Ann. Otol.
Rhinol. Laryngol 79:529-539.
Burton, C.S., J. Nordin, T. Stoeckenius (1982). On a New Short-Term
Standard for Sulfur Dioxide: The Protection from Calculated Peak
1-Hour Concentrations Provided by the Existing 3- and 24-Hour
NAAQS in the Vicinity of a Hypothetical Power Plant. Systems
Applications, Inc. San Rafael, CA.
Burton, C.S., and A.D. Thrall (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, 6.G. , M. Corn, J.B.L. Gee, C. Vasallo, and A. P. Thomas (1969).
Response of healthy men to inhaled low concentrations of gas-aerosol
mixtures. Arch. Environ. Health 18:681-692.
Bushtueva, K.A. (1962). New Studies of the Effect of Sulfur Dioxide and
of Sulfuric Acid Aerosol on Reflex Activity in Man. In: Limits of
Allowable Concentrations of Atmospheric Pollutants. Book 5, B.S. Levine,
translator, U.S. Department of Commerce, Office of Technical Services,
Washington, D.C. pp. 86-92.
Campbell, 6.6., G.G. Shurr, D.E. Slawikowski, and J.W. Spence (1974).
Assessing air pollution damage to coatings. J. Paint Technol .
46:59-71.
Cohen, A. A., S. Bromberg, R.W. Buechley, L.T. Heiderescheit , and C.M. Shy
(1972). Asthma and air pollution from a coal fueled power plant
Am. Rev. J. Pub. Health 62:1181-1188.
Cohen, H.J., R.T. Drew, J.L. Johnson, and K.V. Rajagopalan (1973).
Molecular basis of biological function of molybdenum. The relationship
between sulfite oxidase and the acute toxicity of bisulfite and SO?
Proc. Natl. Acad. Sci . 70:3655-3659.
Cole, P., R. Forsyth, and J.S.J. Haight (1982). Respiratory resistance of
the oral airway. Am. Rev. Respir. Dis. 125:363-365.
Colley, J.R.T., J.W.B. Douglas, and D.D. Reid (1973). Respiratory disease
in young adults: Influence of early childhood lower respiratory
tract illness, social class, air pollution, and smoking. Br. Med. J.
0 «
Committee on Public Works, U.S. Senate (1974). A Legislative History of the
Clean Air Amendments. Volume 1. Serial no. 93-18. U.S. Government
Printing Office, Washington, D.C. prepared by the Environmental Policy
Division of the Congressional Research Service of the Library of Congress.
Corn, M. , N. Kotsko, D. Stanton, W. Bell, and A. P. Thomas (1972). Response
of rats to inhaled mixture of SO? and SO?-NaCl aerosol in air. Arch
Environ. Health 24:248-256.
Costa, D.L., and M.O. Amdur (1979). Effect of oil mists on the irritancy
of sulfur dioxide. I. Mineral oils and light lubricating oil. Am.
Ind. Hyg. Assoc. J. 40:680-685.
Cotes, J.E. (1979). Lung Function Assessment and Application in Medicine.
Blackwell Scientific Publications, London, England, pp. 266-267.
D.C. Cir. (1981). American Petroleum Institute v. Costle, Nos. 79-1104
et al . (D.C. Cir.) September 3, 1981.
D.C. Cir. (1980). Lead Industries Association, Inc. v. EPA. F. 2d, 14 ERC
1906 (D.C. Cir.) Cert. Denied 49 U.S.L.W. 3428 December 8, 1980.
-------
Deal, E. C. Jr., E. R. McFadden, Jr., R. H. Ingram, Jr., and J. J. Jaeger
(1979). Hyperpnea and heat flux: initial reaction sequence in
exercise-induced asthma. J. Appl. Physiol: Respirat. Environ.
Exercise Physiol. 46:476-483.
Denison, W.E. (1973). Life in tall trees. Sci. Am. 228-74.
DHEW [U.S. Department of Health, Education, and Welfare] (1970). 'Air
Quality Criteria for Sulfur Oxides. National Air Pollution Control
Association Publication #AP-50.
Dockery, D. W., N.R. Cook, B.G. Ferris, Jr., F.E. Speizer, J.D. Spengler,
and J.H. Ware (1981). Change in pulmonary function in children
associated with air pollution episodes. Presented at the 74th
Annual Meeting of the Air Pollution Control Association, Philadelphia,
PA (June 21-26, 1981). Paper #81-11.1.
Dodge, R.R., and B. Burrows (1980). The prevalence and incidence of asthma and
asthma-like symptoms in a general population sample. Am. Rev. Respir.
Dis. 122:567-575.
Douglas, J.W.B., and R.E. Waller (1966). Air pollution and respiratory
infection in children. Br. J. Prev. Soc. Med. 21:1-8.
Dreisinger, B.E., and P.C. McGovern (1970). Monitoring Atmospheric
Sulfur Dioxide and Correlating Its Effects on Crops and Forests
in the Sudbury Area. In: Impact of Air Pollution on Vegetation
S.N. Linzon, ed., OntalTo Department of Energy and Resource
Management, Toronto.
Dubrovskaya, F.I. (1957). Hygienic Evaluation of Pollution of Atmospheric
Air of a Large City with Sulfur Dioxide Gas. lr±: Limits of
Allowable Concentrations of Atmospheric Pollutants. Book 3.
B.S. Levine, translator, U.S. Department of Commerce, Office of
Technical Services, Washington, D.C. pp. 37-51.
Ehrlich, R., J.C. Findlay, and D.E. Gardner (1978). Susceptibility to
bacterial pneumonia in animals exposed to sulfates. Tox. Lett.
1:325-330.
Eisen, H.N. (1976). Immunology. Harper and Row, Hagerstown, Md. pp. 410.
EPA [U.S. Environmental Protection Agency] (1982a). Air Quality Criteria for
Particulate Matter and Sulfur Oxides. Environmental Criteria and
Assessment Office, Office of Research and Development, Research
Triangle Park, N.C. EPA 600/8-82-029a-c (December, 1982).
EPA [U.S. Environmental Protection Agency] (1973). Effects of Sulfur
Oxides in the Atmosphere on Vegetation; Revised Chapter 5 for Air
Quality Criteria for Sulfur Oxides. Office of Research and
Development, Research Triangle Park, N.C. EPA-R 3-73-030.
-------
EPA [U.S. Environmental Protection Agency] (1982a). Review of the National
Ambient A1r Quality Standards for Particulate Matter: Assessment of
Scientific and Technical Information (OAQPS Staff Paper). Office of
Air Quality Planning and Standards. Research Triangle Park, N.C.
EPA-450/5-82-001.
Evans, L.S. and P.R. Miller (1975). H1stolog1cal comparison of single and
additive 03 and SO2 injuries to elongating ponderosa pine needles.
Am. J. Bot. 62:416-421.
Fairchild, G.A., J. Roan, and J. McCarroll (1972). Atmospheric pollutants
and the pathogenesis of viral respiratory infection. Arch. Environ.
Health 25:174-182.
Ferin, J., and L.J. Leach (1973). The effect of S02 on lung clearance of
T102 particles in rats. J. Am. Ind. Hyg. Assoc. 3:260-263.
Fink, F.W., F.H. Buttner, and W.K. Boyd (1971). Technical-Economic
Evaluation of Air Pollution Corrosion Costs on Metals in the U.S.
APTD-0654. U.S. Environmental Protection Agency, Research Triangle
Park, N.C.
Fischl, M. A., A. Pitchenik, and L. B. Gardner (1981). An index predicting
relapse and need for hospitalizatlon in patients with acute bronchial
asthma. N. Engl. J. Med. 305:783-789.
Fishman, A.P. (1976). Chronic cor-pulmonale. Am. Rev. Respir. Dis. 114-
775-794.
Frank, N.H., and A. D. Thrall (1982). Relationships Among S02 Averaging
Times and Ambient Standards: Monitoring and Data Analysis Division
Report. Office of Air Quality Planning and Standards, Research Triangle
Park, N.C.
Frank, N.H., B.E. Blagun, and H. Slater (1981). Diurnal Patterns of Sulfur
Dioxide Presented at the 74th Annual Meeting of the Air Pollution
Control Association, Philadelphia, Pennsylvania. Paper #81-33.3.
Frank, N.R. (1964). Studies on the effects of acute exposure to sulfur
dioxide in human subjects. Proc. R. Soc. Med. 57:1029-1033.
Frank, N.R., M.O. Amdur, and J.L. Whittenberg (1964). A comparison
of the acute effects of S02 administered alone or in combination with
Nad particles on the respiratory mechanics of healthy adults.
Int. J. Air Water Pollut. 8:125-1 33.
Frank, N.R., M.O. Amdur, J. Worcester, and J.L. Whittenberger (1962).
Effects of acute controlled exposure to S02 on respiratory mechanics
in healthy male adults. J. Appl. Physiol. 17:252-258.
Frank, N.R. and F.E. Spei zer (1965). SOg effects on the respiratory
systems in dogs. Changes in mechanical behavior at different levels
of the respiratory system during acute exposure to the gas. Arch.
Environ. Health 11:624-634.
-------
Frank, N.R., R.E. Yoder, J.D. Brain, and E. Yokoyama (1969). S02 (35S labeled)
absorption by the nose and mouth under conditions of varying
concentration and flow. Arch. Environ. Health 18:315-322.
Frank, N.R., R.E. Yoder, E. Yokoyama, and F.E. Speizer (1967). The
diffusion of SO? from tissue fluids into the lungs following
exposure of dogs to 35S02.' Health Phys. 13:31-38.
Franklin, W. (1981). Asthma in the emergency room: Assessment and Treatment.
N. Engl. J. Med. 305:826-827.
Fraser, D.A., M.C. Battigelli, and H.M. Cole (1968). Ciliary activity and
pulmonary retention of inhaled dust in rats exposed to sulfur dioxide.
J. Air Pollut. Control Assoc. 18:821-823.
Giddens, W.E., and G.A. Fairchild (1972). Effects of'sulfur dioxide on
the nasal mucosa of mice. Arch. Environ. Health 25:166-173.
Gillespie, J.R. (1980). Review of the cardiovascular and pulmonary
function studies on beagles exposed for 68 months to auto exhaust
and other air pollutants. In: Long-term Effects of Air Pollutants
in Canine Species. J.F. Stara, D.L. Dungworth, J.C. Orthoefer, and
W.S. Tyler, eds. EPA Report #600/8-80-014, pp. 115-153.
Gillette, D.G. (1975). S02 and material damage. J. Air Pollut. Control
Assoc. 25:1238-1243.
Gokemeijer, J.D.M., K. DeVries, and N.G.M. Orie (1973). Response of the
bronchial tree to chemical stimuli. Rev. Inst. Hyg. Mines (Hasselt)
28:195-197.
Goldstein, B. (1983). CASAC Review and Closure of the OAQPS Staff Paper for
Sulfur Oxides. Closure letter to William 0. Ruckelshaus (August, 1983).
Greenwald, I. (1954). Effects of inhalation of low concentrations of
sulfur dioxide upon man and other mammals. AMA Arch. Ind. Hyg. Occup.
Med. 10:455-475.
Gunnison, A.F., and E.D. Palmes (1973). Persistence of plasma S-sulfonates
following exposure of rabbits to sulfite and sulfur dioxide. Toxicol.
Appl. Pharmacol. 24:266-278.
Guttman, H. (1968). Effects of atmospheric factors on the corrosion of
rolled zinc. In: Metal Corrosion in the Atmosphere, a Symposium
Presented at the Seventieth Annual Meeting, American Society for
Testing and Materials, Boston Massachusetts, June 25-39, 1967.
ASTM Special Technical Publication no. 435, American Society for
Testing and Materials, Philadelphia, PA. pp. 223-239.
Haynie, F.H. (1974). The economics of clean air. Mater. Prot. Perf.
13:33-38.
-------
Haynie, F.H. (1980). Theoretical air pollution and climate effects on
materials confirmed by zinc corrosion data, in: Durability of
Building Materials and Components. P.J. Sereda and G.G. Levitan,
eds., ASTM Special Technical Publication 691, American Society for
Testing and Materials, Philadelphia, PA, pp. 157-175.
Haynie, F.H., J.W. Spence, and J.B. Upham (1976). Effects of Gaseous
Pollutants on Materials—A Chamber Study. EPA-600/3-76-015, U.S.
Environmental Protection Agency, Research Triangle Park, NC.
Haynie, F.H., and J.B. Upham (1974). Correlation between corrosion
behavior of steel and atmospheric pollution data. In: Corrosion
1n Natural Environments. ASTM Special Technical PubTlcation 558,
American Society for Testing and Materials, Philadelphia, PA, pp.
33-51.
Haynie, F.H., and J.B. Upham (1971). Effects of atmospheric pollutants
on the corrosion behavior of steels. Mater. Prot. Perform 10:18-21.
Hawksworth, D.L. and F. Rose (1970). Qualitative scale for estimating
sulfur dioxide air pollution in England and Wales using epiphytic
lichens. Nature (London) 227:145-148.
Heagle, A.S., D.E. Body, and G.E. Neely (1974). Injury and yield responses
of soybean to chronic doses of ozone and sulfur dioxide in the field.
Phytopathology 64:132-136.
Heagle, A.S., and J.W. Johnston (1979). Variable responses of soybeans
to mixtures of ozone and sulfur dioxide. J. Air Pollut. Control
Assoc. 29:729-732.
Heck, W.W., and J.A.-Dunning (1978). Response of oats to sulfur dioxide:
Interactions of growth temperature with exposure temperature or
humidity. J. Air Pollut. Control Assoc. 28:241-246. '
Hill, A.C., S. Hill, C. Lamb, and T.W. Barrett (1974). Sensitivity of
native desert vegetation to SOg and to SOg and NO? combined. J. Air
Pollut. Control Assoc. 24: 153-157.
Hirsch, J.A., E.W. Swenson, and A. Wanner (1975). Tracheal mucous transport
in beagles after long term exposure to 1 ppm sulfur dioxide. Arch.
Environ. Health 30:249-253.
Holland, W.W., and D.D. Reid (1965). The urban factor in chronic bronchitis.
Lancet 1:445-448.
Horvath, S.M., and L.J. Folinsbee (1977). Interactions of Two Air Pollutants,
Sulfur Dioxide and Ozone, on Lung Functions. Grant ARB-4-1266,
California Air Resources Board, Sacremento, CA.
-------
Irreverre, F., S.H. Mudd, W.D. Heizer, and L. Laster (1967). Sulfite
oxidase deficiency: studies of a patient with mental retardation,
dislocated ocular lenses, and abnormal urinary excretion of S-sulfo-
L-cysteine, sulfite, and thiosulfate. Biochem. Med. 1:187-217.
Islam, M.S., E. Vastag, and W.T. Ulmer (1972). Sulfur-dioxide induced
bronchial hyperreactivity against acetylcholine. Int. Arch. Arbeitsmed.
29:221-232.
Jacobson, J.S. and L.J. Colavito (1976). The combined effect of sulfur
dioxide and ozone on bean and tobacco plants. Environ. Exp. Bot.
16:277-285.
Jaeger, M.J., D. Tribble, and H.J. Wittig (1979). Effect of 0.5 ppm sulfur
dioxide on the respiratory function of normal and asthmatic subjects.
Lung 156:119-127.
Johnson, T. (1982a). PEDCo Environmental memorandum (March 5,.1982), to
John Bachmann titled "Equivalent 1-hour and 24-hour S02 Standards."
Johnson, T. (1982b). PEDCo Environmental memorandum (September 10, 1982),
to Tom Feagans and Henry Thomas titled "Population Estimates of
Asthmatics and Atopies."
Kagawa, J., and K. Tsuru (1979). Respiratory effect of 2-hour exposure
with intermittent exercise to ozone and sulfur dioxide alone and in
combination in normal subjects. Jap. J. Hyg. 34:690-696.
Karpovich, P.V. (1953). Physiology of muscular activity. W.B. Saunders
Co., Philadelphia, p. 114.
Katz, G.V., and S. Laskin (1976). Pulmonary Macrophage Response to
Irritant Gases. In: Air Pollution and the Lung. E.F. Aharonson,
A. Ben-David, and M.A. Klingberg, eds., John Wiley and Sons, New York,
pp. 83-100.
Kawabori, I., W.E. Pierspn, L.L. Conquest, and C.W. Bierman (1976).
Incidence of exercise-induced asthma in children. J. Allergy
Clin. Immunol. 58:447-455.
Keller, T. (1980). The effect of a continuous springtime fumigation with SOg
and C02 uptake and structure of the annual ring in spruce. Can. J. For.
Res. 10:1-6.
Keller, T. (1978). Wintertime atmospheric pollutants—do they effect the
performance of deciduous trees in the ensuing growing season. Environ.
Pollut. 16:243-247.
Kiernan, K.E., J.R.T. Colley, J.W.B. Douglas, and D.D. Reid (1976).
Chronic cough in young adults in relation to smoking habits, childhood
environment and chest illness. Respiration 33:236-244.
Kikigawa, K., and K. lizuka (1972). Inhibition of platelet aggregation
by bisulfite-sulfite. J. Pharm. Sci. 61:1904-1907.
-------
10
Kirkpatrick, M.B., D. Sheppard, J.A. Nadel , and H.A. Boushey (1982).
Effect of oronasal breathing route on sulfur dioxide-Induced
bronchoconstriction in exercising asthmatic subjects. Am. Rev. Respir.
Dis. 125:627-631.
Kleinman, M.T., R.M. Bailey, Y.C. Chang, K.W. Clark, M.P. Jones, W.S. Linn,
and J.D. Hackney (1981). Exposures of human volunteers to a controlled
atmospheric mixture of ozone, sulfur dioxide and sulfuric acid. Am.
Indust. Hyg. Assoc. J. 42:61-69.
Koenig, J.Q. (1982). University of Washington, Seattle, Washington.
Personal communication (letter) with John Bachmann (March 29, 1982).
Koenig, J.Q., W.E. Pierson, and R. Frank (1980). Acute effects of inhaled
S02 plus Nad droplet aerosol on pulmonary function in asthmatic
adolescents. Environ. Res. 22:145-153.
Koenig, J.Q., W.E. Pierson, M. Horike, and R. Frank (1982a). Bronchoconstrictor
responses to sulfur dioxide or sulfur dioxide plus sodium chloride
droplets in allergic, nonasthmatic adolescents. J. Allergy Clin. Immunol
69:339-344.
Koenig, J.Q., W.E. Pierson, M. Horike, and R. Frank (1982b). Effects of
inhaled sulfur dioxide ($02) on pulmonary function in healthy adolescents-
exposure to SO? alone on S02 + sodium chloride droplet aerosol during
rest and exercise. Arch. Environ. Health 37:5-9.
Koenig, J.Q., W.E. Pierson, M. Horike, and R. Frank (1981). Effects of
SOg plus NaCl aerosol combined with moderate exercise on pulmonary
function in asthmatic adolescents. Environ. Res. 25:340-348.
Kreisman, H., C.A. Mitchell, H.R. Hosein, and A. Bouhuys (1976). Effect
of low concentrations of sulfur dioxide on respiratory function in
man. Lung 154:25-34.
Lambert, P.M., and D.D. Reid (1970). Smoking, air pollution and bronchitis
in Britain. Lancet 1:853-857.
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).
Larson, T.V., R. Frank, D.S. Covert, D. Holub, and M.S. Morgan (1982).
Measurements of respiratory ammonia and the chemical neutralization
of inhaled sulfuric acid aerosol in anesthetized dogs. Am. Rev.
Resp. Dis. (In Press).
Larson, T.V., N.R. Horike, and H. Harrison (1978). Oxidation of sulfur
dioxide by oxygen and ozone in aqueous solution: A kinetic study with
significance to atmospheric rate processes. Atmos. Environ. 12:1597-1611.
-------
11
Laskin, S., M. Kuschner, and R.T. Drew (1970). Studies in pulmonary
carcinogenesis. In: Inhalation Carci nogenesis. AEC Symposium
Series 18. M.G. Hanna, Jr., P. Nettesheim, and J.R. Gilbert, eds.,
U.S. Atomic Energy Commission, Oak Ridge, TN. pp. 321-351.
Laskin, S., M. Kuschner, A. Sellakumar, and G.V. Katz (1976). Combined
Carcinogen-irritant animal inhalation studies. In: Air Pollution
and the Lung. E.F. Aharonson, A. Ben-David, and M.A. Klingberg,
eds. John Wiley and Sons, New York. pp. 190-213.
Lawther, P.J. (1958). Climate, air pollution and chronic bronchitis.
Proc. R. Soc. Med. 51:262-264.
Lawther, P.J., A.G.F. Brooks, P.W. Lord, and R.E. Waller (1974a). Day-to-day
changes in ventilatory function in relation to the environment.
Part I. Spirometric values. Environ. Res. 7:24-40.
Lawther, P.J., A.G.F. Brooks, P.W. Lord, and R.E. Waller (1974b). Day-to-day
changes in ventilatory function in relation to the environment.
Part II. Peak expiratory flow values. Environ. Res. 7:41-53.
Lawther, P.J., A.G.F. Brooks, P.W. Lord, and R.E. Waller (1974c). Day-to-day
changes in ventilatory function in relation to the environment. Part
III. Frequent measurements of peak flow. Environ. Res. 8:119-130.
Lawther, P.J., P.W. Lord, A.G.F. Brooks, and R.E. Waller (1973). Air
pollution and pulmonary airway resistance: A pilot study. Environ.
Res. 6:424-435.
Lawther, P.J., A.J. MacFarlane, R.E. Waller, and A.G.F. Brooks (1975).
Pulmonary function and sulphur dioxide, some preliminary findings.
Environ. Res. 10:355-367.
Lawther, P.J., R.E. Waller, and M. Henderson (1970). Air pollution and
exacerbations of bronchitis. Thorax 25:525-539.
LeBlanc, F., and D.N. Rao (1975). Effects of air pollutants on lichens
and bryophytes. In: Responses of plants to Air Pollution. J.B.
Mudd and T.T. Kozlowski , eds., Academic Press, New York, NY, pp.
237-272.
Lebowitz, M.D., E.J. Cassell , and J.R. McCarroll (1972). Health and the
urban environment. XV. Acute respiratory episodes as reactions by
sensitive individuals to air pollution and weather. Environ. Res.
5:135-141.
Lebowitz, M.n., and G.A. Fairchild (1973). The effects of sulfur dioxide
on A2 influenza virus on pneumonia and weight reduction in mice: An
analysis of stimulus - response relationship. Chem. Biol. Interact.
7:317-326.
-------
12
Lehmann, K.B. (1893). Experimental!e Studien uber den Einfluss technisch
und hygienisch wlchtlger Gase und Dampfe auf den Organismus. VI.
Schwefliger Saure. [Experimental studies on the effects on the system
of technically and hygienically important gases and fumes. VI. Sulfuric
Acid.] Arch. Hyg. 18:180-191.
Leikauf, 6., D.B. Yeates, K.A. Wales, D. Spektor, R.E. Albert, and
H. Lippmann (1981). Effects of sulfuric acid aerosol on respiratory
mechanics and mucociliary particle clearance in healthy nonsmoking
adults. Am. Ind. Hyg. Ass. J. 42:273-282.
Lewis, T.R., D.E. Campbell, and T.R. Vaught, Jr. (1969). Effects on
canine pulmonary function via induced NO-2 impairment, particulate
interaction and subsequent SOX. Arch. Environ. Health 18:596-601.
Lewis, T.R., W.J. Moorman, W.F. Ludmann, and K.I. Campbell (1973). Toxicity
of long-term exposure to oxides of sulfur. Arch. Environ. Health
26:16-21.
Linn, U.S., R.M. Bailey, D.A. Shamoo, T.G. Venet, L.H. Wightman and
J.D. Hackney (1982a). Respiratory responses of young adult asthmatics
to sulfur dioxide exposure near simulated ambient exposure conditions
Environ. Res. 29:220-232.
Linn, W.S., D.A. Shamoo, C.E. Spier, L.M. Valencia, U.T. Anzar, T.G. Venet,
and J.D. Hackney (1982b). Respiratory effects of 0.75 ppm sulfur
dioxide in exercising asthmatics: influence of upper-respiratory defenses
Environ. Res. (In Press).
Linn, W.S., M.P. Jones, R.M. Bailey, M.T. Kleinman, C.E. Spier, R.D. Fischer
and J.D. Hackney (1980). Respiratory effects of mixed nitrogen
dioxide and sulfur dioxide in human volunteers under simulated
ambient exposure conditions. Environ. Res. 22:431-438.
Linzon, S.N. (1971). Economic effects of sulphur dioxide on forest
growth. J. Air Pollut. Control Assoc. 21:81-86.
Linzon, S.N. (1978). Effects of airborne sulfur pollutants on plants.
_In: Sulfur in the Environment: Part II, Ecological Impacts, J.O
Nriagu, ed., John Wiley and Sons, New York, NY. pp. 109-162.
Lunn, J.E., J. Knowelden, and A.J. Handyside (1967). Patterns of
respiratory illness in Sheffield infant schoolchildren. Br. J. Prev
Soc. Med. 21:7-16.
Lunn, J.E., J. Knowelden, and J.R. Roe (1970). Patterns of respiratory
illness in Sheffield junior schoolchildren. A follow-up study
Br. J. Prev. Soc. Med. 24:223-228.
Macklem, P.T., and J. Mead (1967). Resistance of central and peripheral
airways measured by a retrograde catheter. J. Appl. Physiol. 22:395-401.
-------
13
Martin, A.E. (1964). Mortality and morbidity statistics and air pollution.
Proc. R. Soc. Med. 57:969-975.
Martin, A.E., and W.H. Bradley (1960). Mortality, fog and atmospheric
pollution—An investigation during the winter of 1958-59. Mon. Bull.
Minist. Health Lab. Serv. .19:56-72.
Mazumdar, S., H. Schinmel , and I. Higgins (1981). Daily Mortality, Smoke
and S02 in London, England 1959-1972. Proceedings of the Proposed
SOX and Particulate Standard Specialty Conference. Air Pollution
Control Association, Atlanta, Georgia, pp. 219-239.
Mazumdar, S. and N. Sussman (1981). Relationships of air pollution to
Health: Results from the Pittsburgh study. Proceedings of the 74th
Annual Meeting, Air Pollution Control Association, Philadelphia, PA.
(June 21-26, 1981).
McJilton, C.E., R. Frank, and R.J. Charlson (1976). Influence of relative
humidity on functional effects of an inhaled S02-aerosol mixture. Am.
Rev. Respir. Dis. 113:163-169.
McJilton, C., R. Frank, and R. Charlson (1973). Role of relative humidity
in the synergistic effect of a sulfur dioxide-aerosol on the lunq.
Science 182:503-504.
McLaughlin, S.B., and N.T. Lee (1974). Botanical studies in the vicinity
of Widows Creek Steam Plant: Review of air pollution effects studies,
1952-1972 and results of 1973 surveys. Division of Environmental
Planning, Tennessee Valley Authority, Mussel Shoals, Alabama.
McLaughlin, S.B., D.S. Shriner, R. K. M'Conathy, and L.K. Mann (1979). The
effects of S02 dosage kinetics and exposure frequency on photosynthesis
and transpiration of kidney beans (Phaseolus vulgaris L.). Environ.
Exp. Bot. 19:179-191.
McLaughlin, S.B., and G.E. Taylor (1981). Relative humidity: Important
modifier of pollutant uptake by plants. Science 211:167-169.
Mead, J., J.H. Turner, P.T. Macklem, and J.B. Little (1967). Significance
of the relationship between lung recoil and maximum expiratory flow.
J. Appl. Physiol. 22:95-108.
Melville, 6.N. (1970). Changes in specific airways conductance in healthy
volunteers following nasal and oral inhalation of SO?. West Indian
Med. J. 19:231-235.
MES [Meteorological Evaluation Services, Inc.] (1982). Comments on Draft
OAQPS Staff Paper Entitled Review of the National Ambient Air Quality
Standards for Sulfur Oxides: Assessment of Scientific and Technical
Information. Prepared for American Electric Power Service Corporation,
Canton, Ohio.
-------
14
Miller, J.E., D.G. Sprugel , R.N. Mullen, H.J. Smith, and P.B. Xerikos
(1980). Open-air fumigation system for Investigating sulfur dioxide
effects on crops. Phytopathology 70:1124-1128.
Ministry of Health (1954). Mortality and morbidity during the London
Fog of December 1952. London, Her Majesty's Stationary Office.
Morehouse, L.E., and A.T. Miller (1948). Physiology of exercise. The C.V.
Mosby Co., St. Louis, pp. 246-259.
Mudd, S.H., F. Irreverre, and L. Laster (1967). Sulfite oxldase deficiency
1n man: demonstration of the enzymatic defect. Science 156:1599-1602.
Mueller, P.K., G.M. Hidy, K. Harren, T.F. Lavery, and R.L. Baskett (1980).
The occurrence of Atmospheric Aerosols in the Northeastern United States,
JJK Aerosols Anthropogenic and Natural, Sources and Transport, T.J.
Kneip and P.O. Lioy, eds. Ann. N.Y. Acad. Sci. 338:463-482.
Nadel , J.A., and D.F. Tierney (1961). Effect of a previous deep inspiration
on airway resistance 1n man. J. Appl. Physiol. 16:717-719.
Nadel, J.A., H. Salem, B. Tamplln, and Y. Tokiwa (1965). Mechanism of
bronchoconstriction during inhalation of sulfur dioxide. J. Aopl.
Physiol. 20:164-167.
Nakamura, K. (1964). Response of pulmonary airway resistance by interaction
of aerosols and gases in different physical and chemical nature,
Nippon Eiseigaka Zasshi 19:38-50.
NAS [National Academy of Sciences] (1958). Handbook of respiration. National
Research Council. W.B. Saunders Co., Philadelphia, pp. 147-149.
NAS [National Academy of Sciences] (1978). Sulfur Oxides. National
Academy of Sciences, Washington, D.C.
Neely, G.E. (1982). Research agronomist, Corvallis Environmental Research
Laboratory (U.S. Environmental Protection Agency, Corvallis, OR).
Personal Communication with Larry Zaragoza (April 2, 1982).
Neely, G.E., D.T. Tingey, and R.G. Wilhour (1977). Effects of ozone and
sulfur dioxide singly and in combination on yield, quality and N-
fixation of alfalfa. Proceedings of the International Conference
on Photochemical Oxidant Pollution and Its Control. EPA-600/3-77-001b.
Neri , L.C., J.S. Mandel , D. Hewitt, and D. Jurkowski (1975). Chronic
obstructive pulmonary disease in two cities of contrasting air
quality. Can. Med. Assoc. J. 113:1043-1046.
Newhouse, M.T., M. Dolovich, G. Obminski , and R.K. Wolff (1978). Effect
of TLV levels of SOg and H2$04 on bronchial clearance in exercising
man. Arch. Environ. Health 33:24-32.
-------
15
NIAID (1979). Asthma and other allergic diseases. U.S. Department of
Health, Education, and Welfare. NIH Publication 79-387. Washington,
U * I* •
Nieboer, E., D.H.S. Richardson, K.J. Puckett, and F.D. Tomassini (1976).
The phytotoxicity of sulfur dioxide in relation to measurable responses
in lichens. In: Effects of Air Pollutants on Plants. T.A. Mansfield,
ed. Cambridge University Press, New York, pp. 61-114.
Niinimaa, V., P. Cole, S. Mintz, and R.J. Shephard (1981). Oral nasal
distribution of respiratory airflow. Resp. Physiol. 43:69-75.
Nobel, P.S. (1974). Introduction to Plant Biophysical Physiology.
W.H. Freeman and Co., San Francisco, CA.
Noggle, J.C., and H.C. Jones (1979). Accumulation of Atmospheric Sulfur
by Plants and Sulfur-Supplying Capacity of Soils. EPA-600/7-79-109.
Oshima, R.J. (1978). The impact of sulfur dioxide on vegetation: a sulfur
dioxide-ozone response model. Final Report to the California Air
Resources Board Agree. No. A6-162-30.
Peacock, P.R., and J.B. Spence (1967). Incidence of lung tumors in LX
mice exposed to (1) free radicals; (2) S02. Br. J. Cancer 21:606-
618.
Proctor, D.F. (1981). Oronasal breathing and studies of effects of air
pollutants on the lungs (letter). Am. Rev. Respir. Dis. 123:242-243.
Reichel, 6. (1972). The Effect of Sulfur Dioxide on the Airway Resistance
of Man. Annual meeting of the German Society for Industrial Medicine.
Reid, L. (1970). Evaluation of model systems for study of airway
epithelium cilia, and mucus. Arch. Intern. Med. 126:428-434.
Reinert, R.A., and D.E. Weber (1980). Ozone and sulfur dioxide-induced
changes in soybean growth. Phytopathology 70:914-916.
Rosenberg, C.R., R.J. Hutnik, and D.D. David (1979). Forest Composition
of varying distances from a coal-burning power plant. Environ
Pollut 14:307-317.
Rossier, P.H., A.A. Buhlmann, and K. Wiesinger (1960). Respiration-
Physiologic Principles and their Clinical Applications. The C V
Mosby Co., St. Louis, p. 402. ' ' '
Saibene, p.. P. Mognoni , C.L. Lafortuna, and R. Mostardi (1978). Oronasal
breathing during exercise. Pfuegers Arch. 378:65-69.
Samet, J.M., Y. Bishop, F.E. Speizer, Y. Bishop, J.D. Spengler, and B G
Ferris, Jr. (1981). The relationship between air pollution and emergency
^J'oin5 in an industrial community. J. Air Pollut, Control Assoc.
—
-------
16
Schachter, E.N. (1982). Effects of sulfur dioxide on airway function
(letter to the editor). Am. Rev. Respir. Dis. 125:125.
Schimmel, H. (1978). Evidence for possible acute health effects of ambient
air pollution from time series analysis: methodological questions and
some new results based on New York City daily mortality, 1963-1976.
Bull N.Y. Acad. 54:1052-1108.
Schimmel, H., and T.J. Murawski (1976). The relation of air pollution to
mortality. J. Occup. Med. 18:316-333.
Schryer, D.R., W.R. Cofer, III, and R.S. Rogowski (1980). Synergistic
effects in trace gases—aerosol interactions. Scince 209:723.
Scott, W.D., and P.V. Hobbs (1967). The formation of sulfate in water
drops. J. Atmos. Sci. 24:54-57.
Senior, R.M., and S.S. Lefrak (1980). Status Asthmaticus. Chapter 49
Jn: Pulmonary Diseases and Disorders. A. P. Fishman, ed. McGraw-
Hill Book Co. pp. 593-599.
Shalamberidze, O.P. (1967). Reflex effects of mixtures of sulfur and
nitrogen dioxides. Hyg. Sanit. 32:10-15.
Shapiro, R. (1977). Genetic effects of bisulfite (sulfur dioxide).
Mutat. Res. 39:149-176.
Sheppard, D. (1982). University of California at San Francisco
Medical Center. Personal Communication with John Bachmann (April 15,
1982).
Sheppard, D.A., A. Saisho, J.A. Nadel , and H.A. Boushey (1981a). Exercise
increases sulfur dioxide induced bronchoconstriction in asthmatic
subjects. Am. Rev. Respir. Dis. 123:486-491.
Sheppard, D., J.A. Nadel, and H.A. Boushey (1981b). Inhibition of sulfur
dioxide-induced bronchoconstriction by disodium cromoglycate in
asthmatic subjects. Am. Rev. Respir. Dis. 124:257-259.
Sheppard, D., W.S. Wong, C.F. Uehara, J.A. Nadel, and H.A. Boushey (1980).
Lower threshold and greater bronchomotor responsiveness of asthmatic
subjects to sulfur dioxide. Am. Rev. Respir. Dis. 122:873-878.
Shih, V.E., I.F. Abrams, J.L. Johnson, M. Carney, R. Mandells, R.M. Robb
J.P. Cloherty, and K.V. Rajagopalan (1977). Sulfite oxidase
deficiency. Biochemical and clinical investigations of a hereditary
metabolic disorder in sulfur metabolism. N. Eng. J. Med. 297-1022-
1028.
Shownan, R.E. (1975). Lichens as indicators of air quality around a
coal-fired power generating plant. Bryologist 78:1-6.
-------
17
Sim, V.M., and R.E. Rattle (1957). Effect of possible smog irritants on
human subjects. J. Am. Med. Assoc. 165:1908-1913.
Snell, R.E., and P.C. Luchsinger (1969). Effects of sulfur dioxide on
expiratory flow rates and total respitatory resistance in normal
human subjects. Arch. Environ. Health 18:693-698.
Speizer, F.E., and N.R. Frank (1966a). A comparison of changes in pulmonary
flow resistance in health volunteers acutely exposed to SO2 by mouth
and by nose. Br. J. Ind. Med. 23:75-79.
Speizer, F.E., and N.R. Frank (1966b). The uptake and release of S02
by the human nose. Arch. Environ. Health 12:725-728.
Spence, J.W., F.H. Haynie, and J.B. Upham (1975). Effects of gaseous
pollutants on paints: A chamber study. J. Paint Technol. 47:57-63.
Spiegleman, J.R., G.D. Hanson, A. Lazarus, R.J. Bennett, M. Lippmann, and
R.D. Albert (1968). Effect of acute sulfur dioxide exposure on
bronchial clearance in the donkey. Arch. Environ. Health 17:321-326.
Sprugel, D.G., J.E. Miller, R.N. Muller, H.J. Smith, and P.B. Xerikos
(1980). Sulfur dioxide effects on yield and seed quality in field
grown soybeans. Phytopathology 70:1129-1133.
Stacy, R.W., M.M. Friedman, J. Green, D.E. House, L. Raggio, L.J. Roger, and
M. Hazucha (1981). Effects of 0.75 ppm sulfur dioxide on pulmonary
function parameters of normal human subjects. Arch. Environ. Health
36:172-178.
Strandberg, L.G. (1964). SOg absorption in the respiratory tract.
Arch. Environ. Health 9:160-166.
Stankunas, A.R., D.F. Unites, and E.F. McCarthy (1980). Air Pollution
Damage to Man-Made Materials: Physical and Economic Estimates. TRC
Environmental Consultants Inc. Wethersfield, Connecticut, p. II-5.
Stankunas, A.R., D.F. Unites, and E.F. McCarthy (1981). Air pollution
damage to manmade materials. Presented at the 74th Annual Meeting,
Air Pollution Control Association, Philadelphia, Pennsylvania, Paper
no. 81-65.3.
Thompson, J.R., and D.M. Pace (1962). Effects of SOg on established cell
lines cultivated in vitro. Can. J. Biochem. Physiol. 40:207-217.
Thomas, M.D. (1951). Gas damage to plants. Ann. Rev. Plant Physiol.
2:293-322.
Thomas, M.D., and R.H. Hendricks (1956). Effect of air pollution on plants.
In: Air Pollution Handbook. P.L. Magill, F.R. Holden, and C. Ackley,
eds. McGraw-Hill, New York, pp. 1-44.
-------
18
Thrall, A. D. (1982). Using Logistic Regression to Estimate Short-Term
Exceedance Probabilities from Annual Averages. Memorandum to N.H. Frank,
Monitoring and Data Analysis Division, Office of Air Quality Planning
and Standards, Research Triangle Park, N.C.
Tingey, D.T., R.A. Reinert, J.A. Dunning, and W.E. Heck (1973a). Foliar
injury responses of 11 plant species to ozone/sulfur dioxide mixtures.
Atmos. Environ. 7:201-208.
•-.!-../•
Tingey, D.T., R.A. Reinert, C. Hickliff, and W.W. Heck (1973b). Chronic
ozone or sulfur dioxide exposures or both affect the early vegetative
growth of soybean. Can. J. Plant Sci. 53:875-879.
Tingey, D.T., R.A. Reinert, J.A. Dunning, and W.W. Heck (1971a). Vegetation
injury from the interaction of nitrogen dioxide and sulfur dioxide.
Phytopathology 61:1506-1511.
Tingey, D.T., W.W. Heck, and R.A. Reinert (1971b). Effect of low
concentrations of ozone and sulfur dioxide on foliage, growth, and
yield of radish. J. Am. Soc. Hort. Sci. 96:369-371.
Tomono, Y. (1961). Effects of SO? on human pulmonary functions. Sangyo Igaku
3:77-85.
Toyama, T. (1964). Air pollution and its health effects in Japan. Arch.
Environ. Health 8:153-173.
Toyama, T. (1962). Studies on aerosols. Synergistic response of the
pulmonary airway resistance of inhaling sodium chloride aerosols
and SOg in man. Sangyo Igaku 4:82-92.
Upham, J.B. (1967). Atmospheric corrosion studies in two metropolitan
areas. J. Air Pollut. Control Assoc. 17:398-402.
Van der Lende, R., C. Huygen, E.J. Jansen-Koster, S. Knijpstra, R. Peset,
B.F. Visser, E.H.E. Wolfs, and N.G.M. Orie (1975). A temporary
decrease in ventilatory function of an urban population during an
acute increase in air pollution. Bull. Physiopathol. Respir.
11:31-43.
Van der Lende, R., T.J. Kok, R.P. Reig, J.P. Schouten, N.G.M. Orie, and
P.H. Quanjer (1981). The relationship between air pollution and
diminished lung function values. Results of a longitudinal
epidemiological study in the Netherlands. Proceedings of 74th
Annual APCA Meeting, Philadelphia, PA.
Van der Lende, R., G.J. Temmeling, B.F. Visser, K. de Vries, J. Wever-Hess,
and N.G.M. Orie (1973). Epidemiological investigations in the
Netherlands into the influence of smoking and atmospheric pollution
on respiratory symptoms and lung function disturbances. Pneumonologie
149:119-126.
-------
19
von Nieding, H.M. Wagner, 6., H. Krekeler, H. Lollgen, W. Fries, and
A. Beuthan (1979). Controlled studies of human exposure to single and
combined action of N02, 03 and SOg. Int. Arch. Occup. Environ. Health
43:195-210.
Wanner, A. (1980). Interpretation of pulmonary function tests. In:
Diagnostic Techniques in Pulmonary Disease, M.A. JSackner (ed.).
Marcel Dekker, Inc., New York, pp. 353-426.
Ware, J., L.A. Thibodeau, F.E. Speizer, S. Colome, and E.G. Ferris, Jr.
(1981). Assessment of the health effects of sulfur oxides and
particulate matter: analysis of the exposure-response relationship.
Environ. Health Persp. 41:255-276.
Weir, F.W., and P.A. Bromberg (1973). Effects of sulfur dioxide on human
subjects exhibiting peripheral airway impairment. Project No.
CAWC S-15, American Petroleum Institute, Washington, D.C., pp. 1-40.
Weir, F.W., and P.A. Bromberg (1972). Further investigation of the effects of
sulfur dioxide on human subjects. Annual Report Project No. CAWC
S-15, American Petroleum Institute, Washington, D.C., pp. 1-74.
Widdicombe, J.G. (1954). Respiratory reflexes from the trachea and bronchi
of the cat. J. Physiol. 123:55-70.
Wilson, E.O., and W.H. Bossert (1971). A Primer of Population Biology.
Sinauer Associates, Inc., Stanford, Connecticut, p. 47.
Winner, W.E., and J.D. Bewley (1978a). Contrasts between bryophyte and
vascular plant synecological responses in an S02-stressed White
Spruce association in Central Alberta. Oecologia 33:311-325.
Winner, W.E., and J.D. Bewley (1978b). Terrestrial mosses as bioindicators
of S02 pollution stress. Synecological analysis and the index of
atmospheric purity. Oecologia 35:221-230.
Winner, W.E., J.D. Bewley, H.R. Krouse, and H.M. Brown (1978). Stable
sulfur isotope analysis of SOg pollution impact on vegetation.
Oecologia 36:351-361.
Wolff, R.K., M. Dolovich, C.M. Rossman, and M.T. Newhouse (1975a).
Sulphur dioxide and tracheo-bronchial clearance in man. Arch. Environ.
Health 30:521-527.
Wolff, R.K., M. Dolovich, G. Obminski, and M.T. Newhouse (1975b). Effect
of sulfur dioxide on tracheobronchial clearance at rest and during
exercise. JJK Inhaled Particles IV. W.H. Walton, ed. Pergamon
Press, London, pp. 321-332.
Yamada, J. (1905). Untersuchunger fiber die quantative Absorption der Da'mpfe
eingen Saiiren durch Tier and Mensch. [Investigations on the quantitative
absorption of some acid fumes in animals and man.J PhD. Thesis, Wiirzburg
(See Lehrmann, K.B., 1908, Arch. Hyg. 67:57-98.)
-------
20
Yokoyama, E., R.E. Yoder, and N.R. Frank (1971). Distribution of 35S in
the blood and its excretion in urine of dogs exposed to SOo. Arch.
Environ. Health 22:389-395. i
Zarkower, A. (1972). Alterations in antibody response induced by chronic
inhalation of S02 and carbon. Arch. Environ. Health 25:45-50.
>**. ":
Zenz, C. (1975). Occupational Medicine: principles and practical
applications. Year Book Publishers, Inc., Chicago, pp. 113-114.
Zimmerman, P.W., and W. Crocker (1934). Toxicity of air containing
sulfur dioxide gas. Contrib. Boyce Thompson Instit. 6:455-470.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/5-82-007
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Review of the National Ambient Air Quality Standards
for Sulfur Oxides: Assessment of Scientific and
Technical Information OAQPS Staff Paper
5. REPORT DATE
November 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air, Noise and Radiation
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT : ~~~~ ~~— " ~
This paper evaluates and interprets the available scientific and technical infor-
mation that EPA staff believes is most relevant to the review of primary (health)
and secondary (welfare) National Ambient Air Quality Standards for sulfur oxides
(SO ) and presents staff recommendations on alternative approaches to reaffirming or
revising the standards. The assessment is intended to bridge the gap between the
scientific review in the EPA criteria document for particulate matter and sulfur
oxides and the judgments required of the Administrator in setting ambient air
quality standards for sulfur oxides.
The major recommendations of the staff paper include the following:
1) that the health and welfare data support the need for sulfur dioxide
standards;
2) that the reaffirmation of the existing suite of primary and secondary S02
standards (annual, 24-hour, 3-hour) remains a reasonable policy choice;
3) that new data from controlled human exposure studies warrants consideration of
a short-term (1-hour) standard;
4) that a short-term standard (<_ 4-hours) to protect vegetation should be
retained.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Sulfur Oxides
Sulfur Dioxide
Air Pollution
Particulate Matter
Air Quality Standards
I. DISTRIBUTION STATEMEN
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21 ^NO. OF PAGES
20. SECURITY CLASS (Thispage!
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |