REVISION OF THE NATIONAL AMBIENT AIR
QUALITY STANDARD FOR PHOTOCHEMICAL
OXIDANTS
January 6, 1978
STAFF SUMMARY PARE"
EXTERNAL REVIEW DRAFT
STRATEGIES AND AIR STANDARDS DIVISION
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
U. S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA
-------
Review of the National Ambient Air Quality Standard
for Photochemical Oxidants
This paper discusses the nature and extent of changes which might be
made to the existing air quality standard for photochemical oxidants. It
was prepared by the staff of the Strategies and Air Standards Division,
Office of Air and Waste Management, EPA, as part of the current review of
the oxidant national ambient air quality standard. It has not yet been
extensively reviewed within the Agency and therefore does not necessarily
reflect Agency policy or an Agency position. However, it is being made
available to solicit public review and comment which can be considered in
the development of an Agency position on the revision of the oxidant
standard.
We have classified the several decisions which must be made regarding
the standard into two categories:
1. Those areas where the Strategies and Air Standards Division
(SASD) has generated an informal staff position,
and
2. Those areas where this office is still considering a range of
alternatives. .
Background
On April 30, 1971, the Environmental Protection Agency published in
the Federal Register (36 FR 8186) National Ambient Air Quality Standards
for photochemical oxidants. The scientific, technical, and medical basis
for these standards is contained in the air quality criteria document for
photochemical oxidants published by the U.S. Department of Health,
Education, and Welfare in March, 1970. Pursuant to the provisions of
Sections 108 and 109 of the Clean Air Act, as amended, EPA is now in the
-------
process of reviewing, updating, and revising the AP-63 criteria document
as well as reviewing the need for modification to the existing photo-
chemical oxidant standard.
An external review draft of the oxidant criteria document was made
available to the public in September and was reviewed by EPA's Science
Advisory Board in a public meeting on November 10-11, 1977. Comments
from this review and meeting are being incorporated into a final draft
document scheduled for issuance in late February, 1978 (a second Science
Advisory Board public meeting to critique this draft may be held in late
January). At the time the revised criteria document is reissued, EPA
will propose a revised air quality standard for photochemical oxidants
or reaffirm the existing standard.
ISSUES WHERE EPA HAS DEVELOPED A STAFF POSITION
Several decisions must be made to explicitly define an ambient
air quality standard. OAQPS has reached a preliminary position on several
of these choices including: (1) designation of the chemical species for
the standard, (2) averaging time of the standard, (3) a determination on
the feasibility of promulgating a peroxyacetyl nitrate (PAN) standard,
i (i
(4) a standard with a deterministic rather than a statistical form.
Designation of the. Chemical Species for the Standard
The existing standard for photochemical oxidants was established
for the entire class of this complex mix of compounds. Unfortunately,
there are no satisfactory methods for accurately and reliably measuring
this collective class of pollutants. The method used to estimate ambient
-------
oxidant levels and to determine compliance measures only a single component
of the oxidant mix—ozone. The weakness of this approach is that the
chemical designation of the standard and the chemical composition of the
pollutant measured to determine compliance are not consistent. Ambient
ozone concentrations can range from approximately 65 percent to nearly
100 percent of the total photochemical oxidant concentration. There-
fore, ozone may be a poor indicator of the quantity and composition of
the non-ozone oxidant in the ambient air. Also of concern is that aside
from PAN, which is an important constituent of the photochemical oxidant
mix, the non-ozone oxidants remain essentially unidentified, cannot be
measured, and have not been uniquely associated with adverse effects.
The inconsistencies cited above argue for moving away from a total
photochemical oxidant standard to an ozone standard. Such a change also
appears reasonable from a health standpoint. Evidence in the revised
draft of the oxidant criteria document indicates that:
1. The majority of data presented in the revised criteria document
is based on ozone exposures. Nearly all of the clinical and toxicological
studies are based on effects from ozone. '
2. Some more recent epidemiological studies associated adverse
effects more closely with ozone than with total oxidants, and
3. Effects observed in clinical studies with ozone alone are
similar to those effects observed in epidemiological studies where
t
ozone occurs along with the complex mix of urban pollutants. These
findings from the health data further suggest that health effects observed
during periods of elevated photochemical oxidant concentrations are
-------
reasonably attributable primarily to ozone 1n the ambient air.
Because it is desirable to designate a standard which is consistent
with the material being measured to determine compliance, and the fact
that good data exist on the adverse effects of ozone, we plan to
propose to redesignate the photochemical oxidant standard as an ozone
standard.
Averaging Time of the Standard and Exposure Durations of Concern
Clinical studies clearly show impairment of lung function in moderately
exercising healthy subjects exposed to ozone for two hours. Since the
impact of ozone is related to the total dose delivered to the respiratory
tract and since more intense exercise would shorten the time required to
deliver an equivalent dose, exposure durations of less than two hours
are of concern for protection of individuals engaged in intense exercise.
A recent clinical study published this year by DeLucia, and not available
for inclusion in the current draft of the criteria document, appears to
confirm this thesis as it shows lung function changes in exercising
subjects after one hour exposure to relatively low ozone levels. Based
on these considerations, OAQPS does not propose a change in the current
one hour averaging time; of'the standard.
Promulgation of a Primary Standard for PAN
Ozone and PAN are the primary chemical species in the oxidant
mix for which health effects data have been documented. The key
effect associated with exposure to PAN is eye irritation. However,
at the present time the health data upon which to base a separate
PAN standard is inadequate and routine PAN measurement methods are not
available. Most of the studies which have documented the effects of
-------
PAN have used ozone or total oxidants as a surrogate for the material
causing the adverse effect.
Ozone is not a reliable indicator of PAN. Recorded data shows
ozone/PAN ratios ranging from 3 to 150. This variation in the ratio
of ozone to PAN makes it extremely difficult to correlate the eye
irritation effects of PAN with specific ozone values. However, it has
been shown that at ozone levels of about 0.1 parts per million (ppm),
PAN concentrations will be at a level below those associated with
perceptible eye irritation effects. This is true even for pessimistic
(low ozone/PAN ratios) assumptions regarding the ozone/PAN ratios.
Despite the lack of a separate PAN standard, those measures taken
to reduce oxidant/ozone precursor emissions will also reduce PAN levels.
In fact, the revised draft criteria document reports that smog chamber
studies show control of oxidant precursor emissions have a greater im-
pact on PAN levels than on ozone/oxidant levels.
A Standard with a Deterministic vs. Statistical Form
The current oxidant standard is stated in a deterministic form,
that is, a standard violation is determined by the second highest
ambient concentration as measured by a single ambient monitor. A
similar standard expressed in a statistical form would allow the expected
number of exceedances of the standard to be one. The original purpose
of permitting a single exceedance of the existing standard was to allow
for unusual meteorological conditions that were unrepresentative
of air quality in a given area. Unfortunately, this objective is not
achieved by the current standard since the once-a-year approach
-------
specifies in effect that there be zero probability that the second high
concentration measured in a year exceed a given value. That is, when a
single exceedance of the standard is permitted, a second or third
exceedance is also likely to occur. If this probability is zero, then
pollutant emissions must be sufficiently low to prevent the standard
concentration from being exceeded even in years in which very rare
adverse weather conditions pccur.
Another fundamental problem with the current deterministic standard
is that it focuses on a single measured value, the second high observa-
tion. This value is subject to instrument error, is not a stable
statistic, and also will vary in any given area over a period of time.
Use of such a random statistic to determine compliance and levels of
control can lead to values that are unrepresentative of the true air
quality problems in an area.
An approach to avoiding the problems of a deterministic air quality
standard is to redefine the standard in a way that does not require us
to rely on a single measured value. This can be done, while maintaining
the focus of the standard on peak concentrations, by defining the standard
in terms of probability distributions or expected values. For example,
the standard could read 0.08 ppm one hour average with an expected
number of exceedances of one per year. We plan to propose a change from
; ,'
the current deterministic standard to one stated in statistical terms.
-------
ISSUES WHERE EPA IS CONSIDERING A RANGE OF ALTERNATIVES
Preliminary reviews of the health evidence presented in the draft
revised criteria document suggest that the standard level be set somewhere
below 0.15 ppm and that the choice within that range be based on
the level and nature of health risk tolerated. At this time, we are not
prepared to accept recommendations made to us to set a standard at a
t
higher level, but we are ready to listen to these and other proposals
during the public review process. Recommendations which we are not
prepared to accept at this time include: (1) Ford Motor Company's
recommendation for a 0.24 ppm max one hour concentration not to be
exceeded more than once in a 90 day period. (2) The recommendation of
the city of Houston, Texas for a standard in the range of 0.2 ppm to 0.4
ppm and, (3) The Manufacturing Chemists Association recommendation of a
0.25 ppm 4 hour average standard. We are also in disagreement with
several other groups (American Petroleum Institute, General Motors Co.,
and Shell Oil Co.) who fell short of recommending a specific standard
level but concluded that there are no significant effects due to ozone
at levels below the 0.25 - 0.37 ppm range. We disagree with the estimated
effects level suggested by these proposals (particularly for sensitive
segments of the population) and with the implication that a standard
should be set at a relatively high level.
i ^
Primary Standard Level, Risk, and Margin of Safety
The Clean Air Act requires that the EPA Administrator establish
ambient air quality standards that protect sensitive segments of the
population, with an adequate margin of safety, from the adverse effects
-------
8
of specified pollutants. This is an extremely difficult task and
requires that health judgments be made as well as reasoned assessments
of health criteria and information. Uncertainties exist with regard to
the way studies were performed, the lack of experimental data for
sensitive individuals, the synergistic and exacerbating effects of other
pollutants in the ambient air, and the lack of conclusive experimental
or epidemiological evidence to support an exact effects level. In an
attempt to deal with these uncertainties in a reasonable and logical way
for arriving at a decision on the standard, two complementary
approaches were used to assess the significance of the health effects
information presented in the oxidant criteria document. One is a
qualitative consensus judgment of leading health experts regarding
effects levels, risks, and safety factors. The second approach is more
quantitative and separates effects judgment from risk judgments, but
also calls upon health experts for health related judgments.
Qualitative Approach
This input to the standard setting process provided an in-depth
interpretation by a panel of leading health experts of the effects data
provided in the criteria document. Experts were selected based on their
contribution to the key health effects studies cited in the criteria
document and their general medical background. Included in
the group were both those researchers who had performed these studies
as well as those EPA health experts responsible for preparation of
the criteria document. At a two-day meeting on June 7 and 8, 1977,
panel members addressed issues regarding effects level, margins of
-------
safety, risks, sensitive population, seriousness of effects, exposure
durations, and the technical merit of individual studies.* Key findings
from these discussions are summarized below.
With regard to health studies, the panel reached consensus that
short-term exposure to ozone, in the range of 0.15 ppm to 0.25 ppm may
impair mechanical function of the lung, and may induce respiratory and
related symptoms in sensitive segments of the population. These symptoms
and effects will be more readily induced in exercising subjects, particu-
larly in a complex urban atmosphere environment in which ozone can
' ,'
interact with other pollutants.
The panel judged that the occurence of respiratory symptoms and
alterations of mechanical function of the lung have important public
health implications, particularly for the developing lungs of young
children. Although such effects appear to be reversible in exposed
young adults, they represent a potentially serious risk for asthmatics
and other individuals with ariway disease. In the population of
individuals with varying states of biological adaptability, exposures
which produced the abovejdescribed effects may at*times overwhelm the
biological defense of some1 persons. Thus,.the reversibility of effects
in experimentally exposed healthy subjects should not be generalized to
the entire population.
In reviewing the Schoettlin and Landau asthma study used totsupport
the original standard, the panel agreed that evidence supports the
statement that a portion of asthmatics will be affected by maximum
*Panel report available as part of support material for this paper.
-------
10
hourly concentrations of 0.2 ppm* and that this effect is likely to
occur at concentrations in the range of 0.15 to 0.25 ppm in some
asthmatics or other persons with sensitive airways. More recent
Japanese epidemiological studies by Kagawa and Toyama were also cited
as showing a decrease in ventilatory function of school children at
ambient ozone concentrations from 0.1 to 0.3 ppm. The panel concluded
that these studies further supported the evidence for an increased
health risk from ozone exposure over the range of 0.15 to 0.25 ppm
and that a likelihood of lesser but real health risks exists at even
lower concentrations.
Of key importance in the animal studies are findings of increased
susceptibility to bacterial infection following ozone exposures of 0.1
ppm. The panel agreed that these studies have definite human health
implications. Although an exposure level associated with such effects
in humans may be different, these reactions in laboratory animals
represent basic biological responses to infectious agents, and there is
no reason to believe that the pollutant induced alterations of basic
defense mechanisms in experimental animals would not occur in similarly
exposed and challenged humans. However, the panel was not aware, of
any epidemiological evidence that susceptibility to infection increases
in persons exposed to ozone and other photochemical materials. The
panel also agreed that the margin of difference between ozone concen-
trations that produce serious toxicological effects in animals (as
*This number is at variance with the draft criteria document value of
0.25 ppm since the panel estimate represents a medical judgment of the
most likely effect level for'-this study and not simply the concentration
recorded by investigators.
-------
11
well as symptomatic and lung function changes in humans) and ambient
levels of ozone is much smaller than for any other atmospheric pollutant.
In reviewing the body of evidence on health effects, the health
panel concluded that there is no reason to suggest a change from the
concentration defined by the existing primary air quality standard,
namely, 0.08 ppm. This conclusion was based upon the panel consensus
that a variety of adverse effects are likely to occur in some segments
of the population from short-term ozone exposures of 0.15 to 0.25 ppm
and upon other evidence that suggests, though less conclusively, a
possibility of effects at concentrations as low as 0.1 ppm.* The panel
recognized that this standard provides a very small margin of safety,
as noted above. The issues of how many times the 0.08 ppm one-hour level
could be exceeded without increased health risks was also addressed.
The panel agreed that the level of health risks increased: (1) in
proportion to the hourly concentration above 0.08 ppm, (2) in proportion
to the number of hours in one day above 0.08 ppm, and (3) in proportion
to the frequency of days in which hourly averages exceed 0.08 ppm, though
the latter conclusion was recognized to be quite judgmental and generally
*
lacking in confirmatory studies. Nevertheless, the panel could not cite
a medical reason to suggest that an exceedance of the standard were
witohut health risks.
*While we respect the opinion of the panel members, we do not necessarily
agree that: (1) inadequate,data exist to select a standard level (be it
the same or different from the existing standard), (2) that any possible
decision to reaffirm the existing standard be based on a lack of data.
Conversely, such a choice should be based on health criteria and evidence
in the criteria document.
-------
12
Quantitative Approach
In an attempt to provide a more rigorous approach to the standard
setting process, we have developed a new methodology for systematically
collecting and assessing the data important for a standard setting deci-
sion. This methodology has the following characteristics: 1) it considers
all available data and where critical data is lacking, the best possible
judgment of medical experts is used, 2) the uncertainty associated with
the judgment of medical experts is quantified, and 3) quantitative
decision analysis techniques are used to calculate the risks associated
with alternative options regarding the standard. The output of this
approach estimates the risk that sensitive segments of the population
will be subjected to identified adverse health impacts at various
alternative standard levels.
The preliminary step in the quantitative approach is to develop
numerical estimates regarding the judgments of medical experts on
their confidence that an effect level exists at a specified pollutant
concentration. A number of such estimates were developed through the
cooperation of medical experts with substantial recognition in the
area of oxidant/ozone effects research. These individuals include
Drs. Timothy Crocker, Jack Hackney, Bernard Goldstein, David Bates,
Richard Ehrlich, and Carl Shy. The key effects categories identified
by these investigators to describe the health impact of ozone and
photochemical oxidants are: 1) alteration of pulmonary function,
2) cough and chest discomfort, 3} reduced resistance to bacterial
infection in man, and 4) aggravation of asthma, emphysema and chronic
bronchitis.
-------
13
Risk values were generated by combining the judgment on effects
levels by medical experts with our best assessment of the expected
distribution of peak ozone values at various standard levels. Pre-
liminary results from this analysis suggest that health effects can be
observed in selected populations at standard levels of about 0.1 ppm
ozone. For example, the risk that at least a small percentage of in-
dividuals in sensitive segments of the population will suffer any one or
a combination of the above mentioned effects is in the range of 70
percent to 90 percent for an ozone standard of 0.1 ppm.
Number of Exceedances of the Standard
The existing standard permits a single exceedance of the standard
level each year. The rationale for providing this flexibility was not
based on health criteria but on the need to allow for unusual meteorological
conditions. A move to an increased number of exceedances of the standard
cannot be supported based on existing health data. Rather, the health
data indicate that the insult to the respiratory tract is related to
the total dose of ozone delivered, i.e., it is related to the exposure
concentration, frequency of exposure and duration of exposures. Thus,
the health risk to an individual increases not only with the ozone
concentration but also with the number of exposures. Furthermore, while
healthy individuals may not be at serious risk from a single one-hour
exposure to ozone, similar exposures of sensitive persons such as
asthmatics may induce a serious health effect, and repeated exposures of
even healthy persons may lead to increased risks of respiratory impair-
ment. Consequently, we question whether it is prudent to significantly
modify the number of exceedances Cor expected exceedances) of the standard.
-------
14
Because of our concerns regarding unusual meteorological events, we
are considering changing the form of the standard to allow one (or more)
days in which the hourly standard is exceeded. Permitting a day of
hourly values above the standard would more reasonably account for rare
meteorological events. Unusual weather conditions are likely to result
in two or three consecutive hours of high ozone values. A standard
which allows a full day of hourly exceedances would permit the unusual
day (with several high ozone values) to occur an average of once a year;
the present standard would have registered a violation for this day.
Unfortunately, this approach has the same limitation as the existing
rule for exceedances (one hour per year permitted) in that it cannot be
based on health criteria. There will undoubtly be some finite increase
in health risk associated with moving from a single hourly exceedance to
a single day of hourly exceedance(s). We are not prepared at this time to
estimate that incremental risk or make a judgment on whether such an
incremental risk would be acceptable. Our staff position on this issue
is open and we welcome comment on how to best account for meteorological
concerns without impairing health protection when defining allowable
exceedances of the standard.
------- |