EPA-450/4-S6-011
Review Of Control Strategies For Ozone And
Their Effects On Other Environmental Issues
By
Edwin L Meyer, Jr.
Air Management Technology Branch
Monitoring and Data Analysis Division
Office of Air Quality Planning and Standards
November 1986
-------
This report has been reviewed by the Office Of Air Quality Planning And Standards, U.S. Environmental
Protection Agency, and approved for publication. Any mention of trade names or commercial products is
not intended to constitute endorsement or recommendation for use.
EPA-450/4-W-011
-------
TABLE OF CONTENTS
Executi ve. Summary v
1.0 Introduction 1
2.0 Maximum Ozone Concentrations: One nay Travel Time 5
2.1 Implications from Photochemical Grid Model Strategy Runs 10
2.2 Implications from EKMA 21
2.3 Implications from Trend Analyses 31
2.4 Summary 39
3.0 Ozone Concentrations in Rural and Remote Areas 43
3.1 Hypothesis 1: Transport of Fossil Ozone 48
3.2 Hypothesis 2: Transport of Urban Precursors or
Subsequent Products 50
3.3 Hypothesis 3: In-situ Emissions 52
3.4 Hypothesis 4: Natural Sources 54
3.5 Use of Ambient Data to Evaluate Hypotheses 55
3.6 Implications from Preliminary Applications of a Regional
Seale Model 61
3.7 Summary 63
4.0 Other Environmental Issues 64
4.1 Nitrogen Dioxide 65
4.2 Acid Deposition/Visibility Impairment 67
4.2.1 Acid Deposition 67
4.2.2 Visibility Impairment 76
5.0 Summary, Conclusions, Implications 79
5.1 One Day Ozone Phenomenon 7Q
5.2 Ozone in Rural/Remote Areas 82
5.3 N02 85
5.4 Acid Deposition and Visibility 85
i a
-------
5.5 Imp! i cat ions , 86
5.5.1 Implications for Further Studies , 86
5.5.2 Implications for Ozone Strategies 90
6.0 Acknowl edgments c 94
7.0 References 95
11
-------
LIST OF FIGURES
1. Conceptual Diagram Depicting Roles of NOX and Organics in
Formation and Accumulation of Ozone .................................. 7
2. Conceptual View of Three Areas Experiencing Impacts
from Control Strategies for Ozone ....................... ............. 9
3. Change in Maximum Hourly Ozone Concentrations in SCAB
in 1987 (Run 1 Minus Run 2) .............. .- .............. ............. 14
4. Changes of Basinwide Total Ozone Exposure from Implementing
Strategy 2 (ANMOC =-49%, ANOX =-37%) Vs. Strategy 1
(ANMOC =-33%, ANOX =-23%) ................................ ......... .... 16
5. The Effect of Changes in NOX Emissions on the Airshed HC/03
Response Curves for Three Days Generated Using a 42% Reduction
in HC Emissions ..........................................
6. Example Ozone Isopleth Diagram ......................... ............... 23
7. Concept Underlying Equal Control Ratio ............ *.... ............... 24
8. Concept Underlying Equal Control Cost Ratio ............ ............... 28
9. Daily Maximum Ozone as a Function of Applied Controls.. ............... 30
10. South Coast Air Duality Management District Air Monitoring
Station Locations ...................................... .............. 33
11. Trend in Spatial Pattern of 3-Year Average Yearly Maximum
Ozone in LA Basin .................................. .................. 34
12. Estimated Ozone Precursor Emission Trends, South Coast Air Basin ..... 36
13. Frequency of Maximum 03 at Various Sites ............................. 38
14. Ozone Air Quality Trends by Area in South Coast Air Basin3-Year
Mean Daily Maximum-Hourly Concentration for July - September ......... 40
-------
LIST OF TABLES
1. Changes in Peak Ozone Modeled in South Coast Air Basin.. 11
2. Sensitivity of Predicted Peak Ozone in the Bay Area to
Various Control Strategies 20
3. Base Level NMOC/NOX Ratios Leading to Equal NMOC and NOX
Controls Needed to Attain the Ozone NAAOS. 26
4. Non-SMSA Ozone Sites Not Clearly Associated with an SMSA, 45
5. Rural Sites/Occasions with 03 >_ 0.1? ppm 57
6. Annual Arithmetic Mean NO? Concentrations , 68
7. Changes in Sulfate and Inorganic Nitrate Concentrations Due
to Reductions in Precursor Concentrations 73
8. Percent Contributions of Various Species to Light Extinction
in "Typical" Urban and Nonurban Atmospheres in the Eastern
United States ; 77
-------
EXECUTIVE SUMMARY
The purpose of this report is to review scientific information in order
to identify implications of reducing emissions of Volatile Organic Compounds
(VOC) and/or Oxides of Nitrogen (NOX) to achieve National Ambient Air Quality
Standards (NAAOS) for ozone (03). Such a review is intended to serve as back-
ground information which can be utilized by U.S. Environmental Protection
Agency (EPA) policymakers. Therefore, the review is one source of information
for those charged with formulating appropriate strategies to reduce 03.
The primary strategy recommended by the U.S. EPA to attain the 03 NAAOS
is to reduce VOC emissions. A considerable amount of experimental/model ing/
monitoring information has become available since the strategy was first
formulated. This information sheds light not only on relationships between
VOC and NOX control and reductions in 03, but relationships between VOC and
NOX control strategies and emerging environmental issues like acid deposition
and visibility attenuation as well.
Consequences of reducing VOC versus NOX are examined with regard to five
environmental issues. These issues include:
(1) 03 concentrations in an urban plume within one day's (i.e., 10
hours) travel time (at prevailing wind velocities) of a city;
(2) 03 in rural/remote locations which may not be impacted by an
urban plume emanating from a city within 10 hours of travel time;
(3) peak and mean concentrations of Nitrogen Dioxide (NC^);
(4) formation of acid species [nitric acid (HN03) and sulfuric acid
in the ambient air; and
(5) deterioration of regional visibility.
-------
Each of these issues is first discussed at some length. The review concludes
by (a) identifying potential implications regarding the effects of VOC and NOX
control on each of the five issues identified above, and (b) identifying
information needs and appropriate studies for fulfilling them.
Issue (1): One Day Ozone Problem
Relative effectiveness of VOC versus NOX controls in reducing peak 63
concentrations in an urban plume depends on a number of factors. These
include the prevailing NMOC*/NOX ratio, severity of the existing 03 concen-
trations, reactivity of the NMOC mix and, in some cases, atmospheric dilution.
A VOC strategy is more efficient as the NMOC/NOX ratio decreases, reactivity
of the NMOC mix decreases, severity of an 03 problem increases, and dilution
increases. At ratios below about 10:1, for typical NMOC compositions and
under a wide variety of conditions, modeling studies indicate that relatively
greater reductions in NOX than in VOC are needed to attain the NAAOS.
A VOC strategy has other potential advantages over an NOX strategy as
well. One possible consequence of an NOX strategy is to cause peak 03 concen-
trations to occur closer to the most heavily populated areas. This occurs
primarily because nitric oxide (NO) emissions tend to occur in the most
heavily populated areas. The most immediate effect of NO emissions is to
quench 03. Only after a period of several hours are conditions favorable
for 03 accumulation. Hence, until the NAAQS is actually attained, an NOX
strategy carries with it the potential for increasing population exposure
to 03 concentrations above the level of the NAAOS. In contrast, a VOC
strategy should delay accumulation of 03. This occurs because NMOC provides
*Non-methane organic compounds. Ambient concentrations of VOC are commonly
referred to as "NMOC.11 This convention is followed in this review.
VI
-------
the principal means of oxidizing NO so as to promote buildup of 03. Lower
concentrations of NMOC delay this process. The result of this delay is a
lowering of 03 concentrations throughout the most heavily populated areas.
For cities with NMOC/NOX ratios about 10-20:1, we enter a gray area in
which sensitivity tests performed with the Empirical Kinetic Modeling Approach
(EKMA) model suggest that in some cases strategies incorporating VOC and NOX
reductions could be more effective in reducing peak 03 than a strategy emphasiz-
ing VOC reductions only. That is, under some conditions, such a strategy may
result in earlier reductions in 03 then would be true for a VOC only strategy,
with only small differences in the ultimate VOC control target needed to meet
the NAAOS. Which strategy is most effective depends on the specific situation.
In cases where the NMOC/NOX ratio is typically greater than about 20:1,
it is possible that a strategy emphasizing NOX control would be more effective
than a VOC strategy in attaining the NAAOS. One concern about such a NOX
control strategy however is that it may increase population exposure to high
03 before the NAAOS is attained. This latter feature may impose practical
limitations on the implementation of an NOX strategy, so that it would be
viable only for cities which already have low concentrations of NOX.
In the absence of a city-specific demonstration that NOX control is a
more viable approach for attaining the 03 NAAOS, a strategy emphasizing VOC
control should remain the strategy of choice. Whether or not some supplementary
control of NOX would be beneficial is best determined on a case-by-case
modeling investigation. This report does not adequately cover costs attendent
with VOC versus NOX control. However, it is generally believed that the cost
-------
associated with large NOX reductions is greater than that for major VOC
controls. Hence, the "equal control" NMOC/NOX ratios discussed later in this
report are likely to overestimate the attractiveness of NOX control strategies.
Further, a VOC strategy appears less risky, in view of modeling results which
indicate the potential for strategies emphasizing NOX control to increase
temporarily the population exposure to 03 levels above that specified in the
NAAOS.
Issue (2): Rural Ozone
It is concluded that observed 03 concentrations in rural/remote areas
which are greater than 0.12 ppm are most often caused by urban plumes emanat-
ing from upwind urban areas less than 10 hours of travel time away. The most
likely alternative explanations for high rural 03 levels are: (1) generation
from manmade precursors emitted in "rural" areas,* and (2) transport of urban
03 aloft overnight and subsequent fumigation to the earth's surface on the
second day. An urban NOX strategy may be slightly more effective in reducing
peak hourly 03 concentrations in rural areas 10 or more hours travel time
downwind from a city. This is likely because an NOX strategy should reduce
peak 03 occurring closer to its city of origin. The resulting peak 03
levels should, therefore, be subject to greater dilution before reaching
rural/remote areas. However, transported NMOC has also been estimated to
make the job of attaining the NAAOS in downwind cities more difficult under
some circumstances. It is concluded that high 03 in rural/remote areas
resulting from urban plumes would be reduced regardless of whether an NOX or
VOC strategy is pursued in the urban area, so long as the strategy effectively
'According to the Agency's existing 03 policy, a rural area
can include small cities/towns with populations below 200,000.
vi ii
-------
reduces downwind peak 03 concentrations. However, presently available infor-
mation does not allow us to make an informed judgment as to whether a VOC
strategy (with no NOX reduction) could significantly exacerbate rural 03
concentrations over longer averaging times or hourly 03 concentrations which
are presently well below the NAAOS.
The foregoing information indicates the Agency's current policy of
giving higher priority to VOC controls in most urban areas is fundamentally
sound. However, some flexibility could be maintained in deciding whether to
pursue VOC or NOX controls in each "rural" nonattainment area,, It is generally
believed that NMOC/NOX ratios are higher in rural areas and in small cities
than they are in urban areas. This should tend to make control of NOX in
such locations more attractive than it is for large urban areas. However,
relative effectiveness of VOC versus NOX strategies depends on reactivity of
the NMOC mix, as well as the local NMOC/NOX ratio. If the reactivity of a
rural NMOC mix is much lower than that of a typical urban mix, this could
favor a VOC strategy, all other factors being equal. Thus, the reactivity as
well as the NMOC/NOX ratio should be considered in assessing alternative
strategies for reducing emissions in any "rural" area determined to be causing
violations of the 03 NAAOS.
Issue (3): Nitrogen Dioxide
Peak and mean concentrations of N02 can be diminished by reducing NOX
emissions. In contrast, reducing VOC appears to have only a small effect on
peak N02 and little impact on mean concentrations. Thus, it appears that
reducing NOX is the only viable means for reducing N02 in areas which violate
the Federal NAAQS for that pollutant. In some cases, locations pursuing a
-------
VOC strategy for 03 which are also in violation of both NAAOS for 03 and
may need somewhat higher VOC controls than would otherwise be necessary to
attain the 03 NAAOS.
Issue (4): Acidic Species Formation
Acidic species formation and acid deposition have not been studied as
long nor as extensively as the 03 problem and major efforts to study the
problem are currently underway. Consequently, judgments concerning the
effect of VOC versus NOX control on acid species must be more qualitative and
tentative than those made in relation to the 03 problem.
This study suggests that reducing NMOC has a small, ambiguous effect on
the production of nitric acid (HN03) and sulfuric acid (^$04) However,
reducing NOX appears likely to lead to*an approximately proportional reduction
in the formation of HN03. The impact of NOX strategies on H2S04 formation is
mixed, according to the best available estimates. NOX reductions apparently
lead to reductions in gas phase production of H2S04 but to an increase in
liquid phase H2S04 formation. It is concluded that in the western M.S., where
S02/H2S04 is relatively small, NOX reduction should most likely reduce forma-
tion of acidic species. In the eastern U.S., the impact of NOX reduction on
the concentrations of acidic species is less clear. Using the assumptions and
models described in this paper, reducing NOX will probably reduce acidic species
formation in the East. The approximately proportional relationship with HN03
should outweigh any small increase in H2S04 which might occur under some
circumstances.
Issue (5): Visibility
Like the acid deposition problem, there remains considerable uncertainty
regarding the impact of VOC and NOX reductions on visibility. Conclusions
-------
are therefore not as strongly supported as those regarding ozone control.
Regional visibility attenuation is primarily due to the formation of secondary
particulates. In the eastern U.S., most of the reduction in regional visibil-
ity is thought to occur from the presence of sulfate (S04~) aerosol. Sulfate
aerosols occur primarily in two forms: as ^04 resulting from liquid phase
formation, and as ammonium sulfate ((NH4)2S04) occurring primarily from the
reaction of ^04 with naturally present ammonia (NH3). In the presence of
ample S02/ S04=, reducing NOX is likely to result in little improvement in
visibility, and may even cause a slight deterioration (due to increased
liquid phase formation of ^804). However, if vigorous programs are under-
taken to reduce S02 (and S04=), reduction in NOX may lead to some further
improvement in visibility. This latter improvement, if it occurs, would
result from less HN03 being available to react with NH3 to form nitrate
(NOo) aerosol. Using a similar argument, in the western U.S. (where there
is little S04=), reduction in NOX may improve regional visibility. NOX con-
trol would also improve visibility somewhat by reducing concentrations of
NO?. Control of certain species of NMOC should reduce the production of
secondary organic aerosol. However, secondary organic aerosol is not believed
to account for a major portion of regional visibility attenuation.
The review of acidic species formation and regional visibility attenuation
suggests that control of NOX may sometimes be beneficial. Hence, for some
locations, the best strategies for dealing with these issues and for reducing
03 may not always be compatible. In order to minimize any incompatibilities,
selective control of NOX to improve visibility or to reduce potential for
acid deposition is suggested. This selection process entails: (a) control
of elevated point sources of NOX first in rural areas and then in other urban
xi
-------
areas and (b) more general control of NOX only in those urban areas using a
strategy which features NOX control to help reduce 03 or N02.
Summary Of Strategy Implications
On the basis of the information reviewed in this report, the following
strategy implications are derived.
1. A VOC control strategy is the most viable approach for attaining
the 03 NAAQS in cities having NMOC/NOX ratios less than about 10:1.
2. For cities having typical mixes of NMOC and NMOC/NOX ratios in the
range of about 10-20:1, it is less certain that a VOC control strategy for
reducing peak 03 always will be superior to one which also includes some NOX
control. A strategy in which both VOC and NOX are reduced may,, in some cases,
result in more rapid initial reductions in peak 03 at the expense of slightly
increasing the VOC reduction ultimately needed to attain the NAAOS. In other
cases, VOC control needed to attain the NAAQS also may be reduced if there is
some accompanying NOX control. A case-by-case review is recommended to
select the most appropriate strategy.
3. For cities with typical NMOC mixes and NMOC/NOX ratios greater than
about 20:1, it is possible that strategies emphasizing NOX control may be
most effective in attaining the 03 NAAOS. However, the impact of such a
strategy on population exposed to 03, cost of controls and city-specific
factors regarding reactivity of the NMOC mix, air quality and meteorology
should be assessed prior to selecting a strategy in which NOX control is
emphasized (i.e., prior to relaxing any nationally applicable requirements
for VOC regulation).
xii
-------
4. In the absence of a demonstration that an NOX strategy is preferable,
strategies emphasizing VOC control (with or without accompanying reductions
in NOX) should remain the ones of choice for reducing urban peak 03 concentra-
tions to the level of the NAAOS.
5. This review suggests that most rural hourly 03 concentrations in
excess of the NAAOS result from urban plumes, therefore, strategies which
reduce peak 03 concentrations in urban plumes to the level of the NAAQS
should also be effective in reducing the large majority of violations in
rural areas.
6. Higher NMOC/NOX ratios in rural areas suggest that 03 formation -in
such areas is limited by the availability of NOX. Prior to considering an NOX
strategy for a rural area or small city believed to cause 03 in excess of the
NAAOS, however, attempts should be made to characterize the NMOC/NOX ratio
and reactivity of the rural NMOC mix.
7. Control of NOX emissions appears to be the only viable means for
bringing areas violating the Federal N02 NAAQS into compliance. Thus, for
cities violating the NAAOS or for cities subject to growth and having N02
concentrations near the NAAOS, some NOX control may be necessary, even though
in some cases it may increase VOC control requirements to meet the 03 NAAOS.
8. The following conclusions with respect to acidic species and visibility
are less certain and should be viewed more qualitatively.
a. Rest available estimates suggest reducing NOX will reduce HN03,
but may, under some circumstances, increase H2S04 somewhat. Thus, NOX re-
duction appears likely to reduce acidic species levels in the western U.S. The
picture is less clear in the East. For the given set of assumptions described
xm
-------
in this review, it appears that reducing NOX would lead to a reduction in
acidic species formation in the eastern U.S. as well.
b. Although formation of acidic species is linked to the NMOC/NOX/03
problem, current, but limited, studies available to us at this time suggest
that VOC reduction may have little net effect on acidic species formation.
c. NOX reduction appears likely to improve regional visibility in
the West. In the Fast, however, little improvement in visibility is likely
from NOX control, unless and until substantial reductions in SO^/SO^3 occur.
d. Optimum control strategies for 03 may not be consistent with the
most effective strategies for N0?, acid deposition and visibility attenuation.
Informational Needs
Many of the strategy implications identified in this review are based
on less than complete information. Therefore, a number of additional efforts
have been identified which, if implemented, would improve the basis on which
strategy/policy decisions concerning NOX and VOC controls are made. These
efforts include:
(a) a more complete and continuous characterization of NMOC/NOX ratios
prevailing in urban and rural areas;
(b) incorporation and testing of rural chemistry in chemical kinetics
mechanisms and in available models;
(c) more extensive simulations of VOC and NOX control strategies with
sophisticated photochemical models;
(d) more extensive evaluation and application of acidic species models,
and
(e) a systematic assessment of available technology and costs associated
with VOC and NOX control so that economics can be more readily factored into
decisions to choose a mix of VOC and NOX controls for reducing 03.
xiv
-------
1.0 Introduction
Control of Volatile Organic Compounds (VOC) has been the U.S.
Environmental Protection Agency's (EPA's) primary approach for reducing high
ambient concentrations of ozone (03) to comply with National Ambient Air
Quality Standards (NAAOS) for a number of years. Although Nitrogen Oxides
(NOX) are also widely recognized as precursors for 03, control of NOX has
been undertaken principally to reduce ambient levels of Nitrogen Dioxide
(N02) which exceed or approach NAAOS for N02. The purpose of this report is
to review monitoring/modeling/experimental studies to identify implications
of VOC versus NOX controls to reduce ambient 03. The report is intended to
serve as a source of information to help decision makers weigh certain (but
not all) environmental benefits/consequences of strategies which emphasize
reducing VOC versus those which emphasize reducing NOX. In addition, the
report is intended to identify certain information gaps which, if filled,
could provide a more informed basis for choosing the most desirable means to
reduce 03.
Implications of reducing VOC or NOX emissions extend to other environ-
mental concerns besides 03 and N02. Acid deposition and visibility attenua-
tion can be potentially affected by choice of strategies to reduce 03 as
well. This review, therefore, addresses the implications of VOC versus NOX
controls on the following five environmental issues:
1. Impact on 03 concentrations occurring in urban plumes within one
day's travel time (i.e., about 10 hours) downwind from cities;
2. Impact on 03 concentrations greater than the NAAOS occurring
in rural areas sometimes greater than 1 day's travel time downwind from
major cities;
-------
3. Impact on peak and mean concentrations of NC^;
4. Effect on the formation of acidic species in the atmosphere; and
5. Effect on regional visibility attenuation.
There are several reasons why the review described in this report is
timely. First, trends for 03 reported in the National Air Quality and
Emission Trends Report, 1984 (U.S. EPA, 1986) are not as pronounced as
they are for several other pollutants which have been regulated over
comparable time frames. It is appropriate to seek explanations for this
finding. This review is one attempt to do so.
Second, information obtained from networks of Nonmethane Organic Compound
(NMOC)* monitors deployed during the summers of 1984 and 1985 suggests higher
typical NMOC/NOX ratios than had previously been measured (Baugues (1986)).
The new measurements were obtained using a method believed to be considerably
more reliable than continuous techniques used in previous years. For example,
Richter et al. (1985) report highly reproducible results which are also
highly correlated with Gas Chromatographic (GC) sum-of-species results and
exhibit only a small (8%) positive bias with respect to the latter. The
median of median ratios measured in 23 cities was about 1?:1. In 19 of the
cities, median NMOC/NOX ratios ranged from about 9-14:1. Thus, this range of
ratios may be thought of as "typical" of median ratios obtained in the 1984-85
studies.1" Thus, potential presence of higher than expected NMOC/NO ratios in
*In this report, NMOC and VOC are sometimes used interchangeably
Usually, however, NMOC is used to denote ambient concentrations
while VOC is used to denote emissions of organic compounds. '
tSeveral small Gulf Coast cities were excluded in compiling these data.
Data from these cities may be dominated by petrochemical facilities
report very low concentrations of NOX and, therefore, higher NMOC/NOX ratios.
-------
several cities could make a review of the Agency's 03 reduction strategy
prudent.
Third, since the original VOC strategy was conceived, additional
information in the form of modeling results, field and experimental data has
become available. It is prudent to review this information in the context
of current 03 control strategies and other environmental concerns.
Decisions regarding appropriate strategies for reducing 63, like many
other environmental decisions, must be made in the presence of uncertainties.
In order for the strategy implications identified in this review to be weighed
properly by decision makers, it is appropriate to identify limitations in the
underlying data base used to draw conclusions. One such limitation is that
many of the cited references are not from the peer-reviewed open literature.
It was necessary to cite several draft or internal reports, because only
these have sufficient detail to allow one to draw conclusions regarding the
efficacy of VOC and NOX controls. Some of these papers describe work which
is too recent to have appeared in peer-reviewed journals. Others were in-
tended to serve as working level papers to define objectives for subsequent
work, rather than as end products prepared specifically for publication.
A second limitation of the review is that it relies on modeling results
which simulate relatively few days in relatively few cities. Particularly
heavy reliance is placed on modeling results obtained for the Los Angeles
area. Given the difficulty and expense of performing detailed modeling
simulations, it makes sense that the most detailed information is available
for the location where the 03 problem is most severe. Although several other
modeling studies appear to be in qualitative agreement with the Los Angeles
-------
results, the general applicability of the Los Angeles findings can be and has
been questioned.
A third source of uncertainty results from the observation that predicted
sensitivity of 03 to changes in NMOC and/or NOX depends on the chemistry
assumed in the models. In this review, heavy reliance is placed on predictions
made using carbon bond mechanisms (CB-2 and CB-3). While qualitative agreement
exists between the carbon bond mechanisms and other chemical mechanisms used
in the review (i.e., LIRA02 and Dodge), quantitative estimates are subject to
some uncertainty.
Uncertainties about relationships between VOC and NOX control strategies
and changes in urban 63 levels are smaller than those which relate to acid
deposition and visibility attenuation. These latter two problems are more
complex and have not been studied as extensively. As a result, conclusions
related to acid deposition and visibility attenuation must be viewed more
qualitatively and as preliminary. In reviewing potential consequences of VOC
and NOX control on acidic species formation and visibility, heavy reliance is
placed on results of models which have undergone limited validation. Further,
for acid deposition, model applications which are reviewed are limited in one
case to clear sky and non-precipitating stratus cloud environments, and in a
second case to one episode. Model inputs/assumptions appear reasonable in
light of present knowledge, and the models appear to perform adequately in
the limited tests to which they have been applied. Nevertheless, it is neces-
sary to make a number of assumptions concerning appropriate model inputs.
Further, it is not yet clear how important the situations simulated are rela-
tive to other environments in acidic species formation, nor is the relative
-------
importance of liquid vs. gas phase reactions in leading to acid deposition
entirely understood.
Finally, it is well known that many species of VOC are toxic and certain
compounds derived from NOX may be potentially toxic as well. The desirability
of controlling VOC or NOX in order to reduce toxic species has not been
addressed in this review.
This review is organized as follows.
Relationships between changes in NMOC and NOX concentrations and high
03 concentrations in an urban plume are discussed in Chapter 2.
Potential effects of control strategies on 03 concentrations in
excess of the NAAQS observed in rural areas are discussed in Chapter 3.
Chapter 4 addresses additional concerns, including the effect of VOC
and NOX controls on N02 and the role of NOX and VOC emissions in contributing
to acid deposition and visibility impairment. The review concludes by
identifying a series of implications attendant with strategies to reduce
03, as well as informational needs.
2.0 Maximum Ozone Concentrations: One Pay Travel Time
High concentrations of 03 observed in the vicinity of urban areas are a
'function of ambient concentrations of NMOC and NOX (resulting from emissions
occurring in the urban area and from transport of precursors from upwind
sources), transport of 03 from upwind areas, local meteorological conditions
and location of the ambient 03 monitors with respect to the city. NOX provides
the most important means by which 03 forms. This occurs when N02 is photolyzed
to form nitric oxide (NO) and an oxygen atom (0). The resulting oxygen atom
combines rapidly with molecular oxygen (03) to form 03. However, in the
absence of other factors, the resulting 03 disappears when it oxidizes the
-------
newly formed or fresh NO (the most abundant species of NOX emitted by manmade
sources) to reform N02» Hence, an equilibrium is established among NO, NO^
and 03 which prevents large buildup of 03. The role of NMOC is to provide an
alternate means (rather than n3) of oxidizing NO to N02. Organics in the
atmosphere undergo oxidation to form alkyperoxy free radicals (R02) which
then oxidize NO to NO;?. Some of the resulting reaction products are themselves
capable of oxidizing NO. The rate of oxidation of NO by some free radicals
is thought to occur over 1000 times faster than the rate of oxidation by 03.
The result of these interactions between NMOC and NOx is that the presence of
NMOC allows 03 to accumulate to higher concentrations, and one molecule of
NOX can lead to several molecules of 03 being formed as the result of repeated
occurrences of the photolysis/oxidation cycle. Figure 1 illustrates the
conceptual role of NMOC and NOX in the formation and accumulation of 03.
Meteorology affects the process described in the preceding paragraph in
several ways. First, the rates of photolysis for N02 and certain other
species are affected by sunlight intensity (e.g., latitude, time of year,
cloud cover). .Second, the rates for several key chemical reactions are also
affected by temperature. Third, since 03 levels depend on concentrations of
its precursors, dilution resulting from prevailing meteorology can affect 03
levels. Fourth, wind speed and direction determine whether a polluted air
mass affects a monitoring site as well as the amount of reaction time available
before the most heavily polluted air impacts a monitor.
Location of the ambient 03 monitors determines the levels of 03 which are
observed in or near a city. Because of the usual diurnal patterns in tempera-
ture, sunlight intensity and emissions, peak 03 concentrations are most
likely to occur in the afternoon. Therefore, ambient monitoring sites
-------
-------
observing highest levels of 03 are typically those which the prevailing winds
cause to be impacted by peak morning emissions during the afternoon.
Since the 03 NAAOS is concerned with daily maximum concentrations, we
often focus on sites which currently measure the daily maxima. However, in
order to assess the implications of NOX vs. VOC control strategies, it is
important to review potential consequences at a variety of sites. Figure 2
is a conceptual representation of geographic zones within and downwind of a
city. Area I is characterized by high VOC and NOX emission densities typi-
fied by large central urban commercial/industrial districts. Area I is also
often characterized by high population. Area II represents suburban or
surrounding rural countryside, typically within 4-10 hours travel time from
the center city at prevailing wind velocities. Area III depicts rural or
remote countryside greater than 10 hours travel time away from a city at
prevailing wind velocities.
With the exception of atmospheric stagnations, highest daily maxima 03
concentrations generally occur at sites in Area II. Maxima at Area I sites
tend to be somewhat lower for several reasons. Ozone levels in this region
are scavenged by fresh NO emissions which occur throughout the day. The 03
accumulation process, described previously, takes several hours to reach its
maximum. By that time, prevailing meteorological conditions have carried the
relatively great emissions from Area I into Area II. Because there are fewer
emissions in Area II, immediate scavenging by fresh sources of NO is reduced.
By the time an urban plume typically reaches Area III, it is late in the day,
further photolysis of N02 and certain other species is limited, and dilution
and dry deposition result in a lowering of 03 and its precursors.
-------
-------
It is now appropriate to discuss the effect reducing VOC and NOX
emissions might have on ambient 03 levels observed in Areas I and II. A
discussion of Area III will be deferred until Section 3.0. The effect of
controls can be assessed using models as well as a limited amount of trend
data.
2.1 Implications from Photochemical Grid Model Strategy Runs
Application of photochemical grid models is particularly useful, because
it most readily allows one to examine the impact of NOX/VOC control strategies
in Area I as well as in Area II. Results from a limited number of strategy
runs involving both VOC and NOX controls are available for Los Angeles (SCAOMD,
1982a), St. Louis (Cole, et al. 1982), the San Francisco Ray Area (ABAG,
1979, ABAG, 1982, Whitten, et al. 1981) and Denver (ABAG, 1979, Anderson, et
al. 1977).
Four separate investigations have been conducted which include the
impact of NOX controls on predicted 03 concentrations in the Los Angeles Basin
(SCAOMD, 1981), (SCAOMO, !P82a), and (SCAOMD,1982b). Of these, a study
conducted for the California Air Resources Board (SCAQMD, 1981) provides the
most unambiguous assessment of the effects of NOX control, so it will be
discussed in greatest detail.
Table 1 is based on information obtained in the California Air Resources
Board (CARB) study. In the CARB study, hypothetical control scenarios were
applied uniformly throughout the basin. For example, if NMOC was reduced by
50% in Los Angeles County, it was also reduced by 50% throughout the rest of
the Basin. As one can see from the Table, uniform NMOC reduction (columns
(9) and (12)) is predicted to result in reductions in daily maximum 03 at all
locations in the Los Angeles Basin. In contrast, the predicted results of
10
-------
g
O if)
N H
X
Mi
< <
»*
10
»*
I
g
»*
l*»
1-4
I
CM
CM
I
gg
en
i
n n
s
CM
»*
en
S»«
to
ui CM
i i
a
CM
g
t-«
»
g
*
i-*
<
»*«
O O
ID ID
I I
CM
CM
*7
vo
CM
00 ~«
*»
^.
f-4
I
en
00
r» CM
i i
H »
g
»*
en
CM
CM
r-t
I
»*
\f>
CO
I
g
CM
I
o
M
o
«
T
»
T
cn
CM
CO
UJ
i
z
o
o o
g
CM
««
i
*«
I
. CM
O
00
o e
41 S.
> a
c_ a
41
Sf
O
CM
CM
co
CM
I
CM
CO
I
CM
00
Ul C <
X X-l
. O «M O
ro u ai o c
CM
f+
I
r- 4>
CO *J
a» u
a*
en
CO
Ml
o
»- X
O 3
4- O
^J
C_ X
3 41 O C "~ 4>
« O'< COO X
41 > C C >
<«~o
IS 3 § ^ 5'a. 21 O C £ 2
a. M o «' co-^^cooooe
^41 O »
4J 4) «^ C
4) O U *» »-
11
-------
uniform NOX reductions (column (13)) are mixed. In Area I (refer to Figure
2) locations, the CARB study suggests that reducing NOX is counterproductive
in reducing 03. That is, for a given level of hydrocarbon control, the larger
the reduction in NOX, the smaller the reduction in 03 (e.g., columns (fi) -
(9)). Reduction in NOX alone leads to an increase in 03 (column (13)). In Area
II locations, moderate reductions in NOX (i.e., up to about 50%) appear to
lower the effectiveness of NMOC reductions (columns (7) - (12)). More drastic
reductions in NOX (i.e., column (6) vs. column (7)) lead to greater reductions
in daily maximum 03 levels for a given reduction in NMOC. Since highest 03
levels prevail in Area II locations, a strategy calling for drastic reductions
in NMOC and NOX leads to the greatest reduction in the Basinwide daily maximum
concentration of 03 (column (6), last row).
The strategy simulations presented in Table 1 were obtained by using
meteorological inputs believed appropriate for the second day of a 2-day
episode (June 26-27, 1974). As a result, relatively high boundary conditions
were assumed (e.g., NMOC = 0.216 ppmC, 03 = .06 ppm). The use of high boundary
conditions could affect the general applicability of the proceeding findings
as well as explain some of the peculiar looking results (e.g., increase in 03
at coastal sites accompanying 75% reductions in both NOX and NMOC emissions).
In addition, the results in Table 1 reflect emission levels and composition
characteristics of the 1975-76 emission inventory for the South Coast Air
Basin.
12
-------
Examination of three other studies in the South Coast Air Basin suggests
results which are, for the most part, qualitatively similar to those in the
CARB study. First, a study sponsored by the SCAQMD (SCAQMD, 1982b) assumed
lower boundary conditions (e.g., NMOC = .12 ppmC, 03 = .04 pprn) and used a
1979 emissions inventory. Two strategies were examined. Strategy 1 assumed
that VOC and NOX emissions would be reduced to the forecast 1987 emissions
(i.e., ANMOC = -33%, ANOX = -23% from a 1979 base period). Strategy 2
added the effects of various additional controls which were regarded as
feasible by 1987 by the SCAOMD (i.e., ANMOC = -49%, ANOX = -37% from a
1979 base period).
Figure 3 is a "deficit diagram", which has been excerpted from the
SCAOMD study (SCAOMD, 1982b). In the figure, cross hatched areas are those
in which the daily maximum 03 concentration at ground level increased as
the result of applying the more stringent Strategy 2 instead of Strategy 1.
Lined areas are those in which daily 03 maxima decreased. Although the
change in VOC emissions between the two strategies confounds the comparison
somewhat, the results are essentially similar to the ones described earlier.
That is, increased reductions in NOX emissions (1) lead to increased 03 levels
near most major sources of NOX and (2) result in lower concentrations of 03
further downwind. Additional studies conducted separately by Systems Applica-
tions, Inc. (SAI) and studies sponsored by the Electric Power Research Insti-
tute (EPRI) yield qualitatively similar results in most respects (SCAOMD,
1982a). An exception is that the EPRI study does not indicate reductions in
the basinwide daily maximum 03 concentrations occurring whereas all of the
13
-------
LU
O CD
M cn
O
>-
_l
DC
CO CXJ
DC
cn
x cn
< 2:
o
UJ
CD
DC
LLJ
< CJ
CJ O
CJ
on
LU
DC
ID
CD
ID
DC
U_
m en
cu i
^ cn
CD ^r
cn i
8
CJ
LU
UJ CD
CD Z
z <:
o
OJ
oc CE
JD
CXJ
GO
O)
o
o
CJ
o
cr
o
LU
i
CL
<
o
14
-------
other studies suggest that substantial reductions in NOX will yield some
reduction in the basinwide maximums (all of which are predicted to occur in
downwind areas).
One further point merits discussion before we leave the Los Angeles
modeling studies. As a rule, the majority of an urban area's population is
likely to live near the area having the highest "emissions of VOC and NOY
^
(i.e., Area I, Figure ?.}. Los Angeles is no exception to this rule.
"Exposure" can be defined as the time period during which the population is
exposed to a concentration above a specified level times the population
exposed. Figure 4 depicts change in exposure accompanying SCAOMD Strategy 2
vs. Strategy 1 as a function of specified 03 concentrations (SCAOMD, 1982b).
We see in this example that if we specify a relatively low 03 concentration
(e.g., £ 15 pphm), the reductions in NOX (and VOC) emissions actually lead to
an increase in population exposure. This follows from the fact that a re-
duction of NOX emissions of the magnitude associated with Strategy 2 leads to
increases in the relatively low 03 concentrations occurring near sources of
NOX, where most people in the Basin live. However, if our principal concern
is to lower exposure to very high concentrations (e.g., greater than 15
pphm), this example indicates a decrease in exposure. The explanation for
this latter finding is that such high 03 concentrations are primarily found
in downwind areas, and it is in these areas where the strategy (AHC=-49%,
ANOx=-37%) is most effective.
The Los Angeles modeling studies discussed in the preceding paragraphs
are subject to an important caveat. This is that the studies do not allow us
to infer the impact that NOX control has on VOC control levels needed to
15
-------
II
o
z
<
LU
LU
a:
g $
co
o Co*
X J^
z "
S r,X
I-; o
CT>
O
a
03
00
CsJ
00
CO
0>
16
O
CO
-------
attain the 03 NAAQS. It is this question which has been of greatest interest
for regulatory applications. Also, the general applicability of the Los
Angeles studies may be questioned because they were conducted for a limited
sets of meteorological conditions and boundary conditions, and with a chemical
mechanism which has since been updated. Further, prevailing concentrations
of 03 are much higher in Los Angeles than elsewhere. Grid Modeling studies
conducted elsewhere are discussed in the next several paragraphs.
Several strategies involving NOX controls have been simulated for St.
Louis (Cole, et al. 1982). These studies utilize the 1975 emissions data
base compiled for the RAPS study, assume low boundary conditions which are
kept constant in the pre- and post-control simulations, and utilize meteoro-
logical data from 3 days on which high 03 was observed to occur during 1976.
The meteorological cases range from moderate wind flow (days 159, 195) to
severe stagnation (day 275). The modeling domain considered in the St.
Louis Study is smaller than that used in Los Angeles. That is, the maximum
downwind distance (from Center City) for which 03 is calculated is about 25
miles in St. Louis as opposed to about 75 miles for Los Angeles. Hence, we
might expect that the predicted effect of NOX controls would be less favorable
in St. Louis than in Los Angeles, all other factors being equal.
The results of the St. Louis studies reported herein dwell on the impact
of uniform reductions in NOX emissions on the basinwide predicted maximum 03
concentrations. The results of strategy runs for each of the three days
simulated are shown in Figure 5. The negative slopes of the three lines in
the figure indicate that, for the 42* level of NMOC control, reducing NOX
emissions results in higher peak 03 than would occur if NOX remained constant.
We also note that on the day for which five strategies were simulated (nay
17
-------
20
-t-
c;
cu
60
tt
C
0)
u
I*
ft)
ft*
20
-L.
AO
-L.
60
Percent change In NO
195
-40
-60
^**^ .275
FIGURE 5. The effect of changes In NO emissions on the Airshed
HC/03 response. Curves for three days generated using
a 42X reduction in BC emissions.
Source: Cole, etal. (1982)
18
-------
275), this slope reduces as changes in NOX become more drastic. In the
St. Louis studies, any benefits from NOX control are less apparent than is
the case for the Los Angeles results. This may be the result of the more
limited modeling domain, somewhat lower NMOC/NOX ratios derived from the
RAPS data, as well as differing meteorological and boundary conditions.
The Association of Bay Area Governments (AB'AG) has used the LIRAQ
photochemical grid model (with LIRA02 chemistry) to assess the impact of NMOC
and/or NOX control on predicted 03 concentrations in the San Francisco Bay
area. Table 2 depicts change in basinwide peak hourly 03 concentrations as a
function of various strategies. ABAG (1982) also notes that NOX control has
an adverse effect on exposure to 03 over the modeling domain (i.e., 100 km x
100 km). The results in Table 2 are qualitatively similar to those in Figure
5, and were obtained using a different grid model with a different chemical
mechanism.
More limited assessments of NMOC and NOX control are available for the
Denver area as well (Anderson, et al. 1Q77 and ABAG, 1979). Here, the
following strategies were compared: (1) a strategy reducing NMOC by 30%; (2)
a strategy in which both NMOC and NOX are reduced by 30%; and (3) a strategy
in which NOX was increased by 65%. It was found that the third strategy led
to the greatest reductions in 03 within the 30 mi x 30 mi modeling domain,
whereas Strategy 2 resulted in the least reduction in 03. It should be
pointed out that the modeling domain in the Denver study is the smallest of
those reviewed, and the chemistry used in the Urban Airshed Model (Carbon
Bond I) is now dated. Nevertheless, the results are qualitatively similar to
those obtained in San Francisco, St. Louis and Los Angeles.
19
-------
TABLE 2.
SENSITIVITY OF PREDICTED PEAK OZONE IN THE BAY
AREA TO VARIOUS CONTROL STRATEGIES
Results/Strategy ° '*> -40 -6° -80 '4° ^ 0
4N°xW 0 0 0 0 0 -20 -40 -80
Predicted Peak 03 .189 .140 .082 .069 .055 .119 .Ofi4 .233
t Change 1n Peak 03 . _26% _57% .63% _n% _m _gg%
Source: Whitten, et al., 1981
20
-------
2.2 Implications from EKMA
In Section 2.0, we noted that presence of both NMOC and NOX is necessary
for high concentrations of 03 to be formed. It follows that if the amount of
NMOC is great in relation to the available NOX, the formation of 03 will be
United by the available NOX, and vice versa. This relationship is usually
expressed quantitatively using the NMOC/NOX ratio which occurs in the morning
before the onset of photochemistry.
Relative amounts of NMOC and NOX are considered implicitly in the pre-
viously described photochemical grid modeling studies. For example, a review
of extensive NMOC and NOX data collected during the RAPS study suggests that
the prevailing NMOC/NOX ratio In St. Louis during 1976 was about 8:1 (Freas,
et al. 1978). More recent measurements made during summer 1985 suggest typical
ratios in St. Louis in the order of 9-10:1. Although there is considerable
uncertainty over what NMOC/NOX ratios exist in Los Angeles, the South Coast
Air Quality Management District used ratios ranging from 9-11:1 as best
estimates in a 1982 EKMA analysis for the South Coast Air Basin (SCAOMtl,
1982c). Los Angeles, St. Louis, San Francisco and Denver are large cities
with diverse mixtures of sources. fiipson (1984a,b) has reviewed NMOC species
data measured in large cities in the Northeast Corridor with diverse mixtures
of sources. This review suggests that the reactivity of the organic species
mix is similar in each of these cities, and can be characterized reasonably
well by a default mixture recommended by Killus, et al. (1984). Finally,
all four cities had severe 03 problems (Max. 03 > .18 ppm) during the periods
modeled with the grid models. Hence, the results reported in Section 2.1 are
probably most applicable for large cities with diverse sources of precursors,
severe 03 problems and prevailing NMOC/NOX ratios in the order of 7-12:1.
21
-------
What can be said about cities having less severe 03 problems, differing
NMOC/NOX ratios and/or mixes of organic pollutants with differing reactivities?
In this section, the EKMA model is used to address these questions.
Figure 6 is an example of an 03 isopleth diagram of the sort generated
by the OZIPM model underlying EKMA (Gipson, 1984c). In these diagrams,
maximum hourly 03 is plotted as an explicit function of initial NMOC and NOY
^
precursor concentrations, and as an implicit function of reactivity, meteoro-
logical assumptions, boundary conditions and diurnal emission patterns. The
solid diagonal line 1n Figure 6 is a line of constant slope, and depicts an
NMOC/NOX ratio. The dashed lines represent high and low NMOC/NOX ratios. If
we examine the "high ratio" line, we find that the shape of the Isopleth
diagram 1s such that reducing NOX levels will produce more apparent benefit
than reducing NMOC. Conversely, the "low ratio" line implies greater benefits
from reducing NMOC.
Isopleth diagrams like Figure 6 can be used to gain insight into the
relative efficacy of reducing NMOC vs. reducing NOX under various conditions.
To gain such insight we define the "Equal Control Ratio" (ECR).
The ECR is defined as that NMOC/NOX ratio at which equal relative
reductions in NMOC or in NOX would result in attainment of the NAAOS for
03. The ECR is illustrated conceptually in Figure 7 for an observed 03
concentration of 0.24 ppm and for the set of assumptions used to generate
the diagram in Figure 7. Interpretation of Figure 7 is relatively straight-
forward. In the example shown, for NMOC/NOX ratios which are greater than
(i.e., to the right and below) the ECR would require less relative reduction
In NOX than in NMOC to meet the NAAQS for 03. That Is,
22
-------
O4 0.9
lf* »,* 1>4 1.6
i..e
FIGURE 6. Example Ozone Isopleth Diagram
23
-------
NO*
0/4 0.6 0.8 1,0 i.g 1.4 1.6 f.
FIGURE 7. Concept Underlying "Equal Control Ratio"4
24
-------
< ANMOC .
NOxB NMOCB
The shape and spacing of an Isopleth diagram depends on a number of the
implicit meteorological, air quality and emission assumptions identified
previously. Therefore, one might expect that the value of the ECR depends on
these assumptions as well. To test this hypothesis, we have conducted a
series of sensitivity studies using EKMA with the Carbon Bond 3 (C8-3) chemical
mechanism. In these studies, recommended default values were assumed for
boundary conditions and diurnal emission patterns (Gipson, et al . 1983 and
Gipson, 1984d). Effects of reactivity, atmospheric dilution and severity of
an areas's 03 problem on the ECR were assessed. Table 3(a) summarizes the
results for "pure" (i.e., VOC reduction only or NOX reduction only) VOC and
NOX strategies. In Table 3(a), "high", "default" and "low" reactivities
correspond to three different NMOC mixtures investigated by Gipson (1984b).
"Low", "Medium" and "High" dilution are identified in Gipson, et al . 1983.
We see from Table 3(a), that the ECR increases with decreasing reactivity,
with increasing severity of the 03 problem and with increasing dilution. The '
"default" reactivities underlying the results in Table 3 were recommended by
"11 us. et al. (1984), and are based on empirical observations in a number of
cities. Further, as previously noted, Gipson (1984a) has found that the
default mix adequately describes reactivity in additional cities. Therefore,
If we assume that the "default reactivity" adequately characterizes the NMOc'
mix in most cities, we conclude that the ECR for "pure" VOC vs. "pure" NOX
strategies is probably between 10-20:1 over a range of 03 severity and dilution
conditions encompassing most cities in the U.S. having problems attaining the
03 NAAQS.
-------
TABLE 3.
BASE LEVEL NMOC/NOX RATIOS LEADING TO EQUAL NMOC AND NOY
CONTROLS NEEDED TO ATTAIN THE OZONE NAAQS
(a) NMOC-Only vs. N0x-0nly Strategies
Scenario
Low Dilution
Medium Dilution
High Dilution
Base Level
07
High Reactivity
Default Reactivity
Low Reactivity
High Reactivity
Default Reactivity
Low Reactivity
High Reactivity
Default Reactivity
Low Reactivity
0.
Equal
Control
Ratio
4.5:1
12:1
9:1
6:1
11:1
12:1
7:1
10:1
16:1
16
%
Control
65%
72%
68%
58%
70%
55%
65%
65%
60%
Equal
Control
Ratio
6.3:1
12:1
12:1
7:1
12:1
17:1
8:1
16:1
22:1
0.20
%
Control
76%
72%
76%
72%
70%
68%
75%
75%
75%
0.
Equal
Control
Ratio
7.5:1
14:1
16:1
9-1
15-1
22:1
10:1
20:1
26:1
24
%
Control
79%
77%
80%
77
-------
Table 3(b) compares a "pure" VOC reduction strategy vs. a strategy
calling for equal percent reductions in VOC and NOX. Here We see that the
ECR for these two types of strategies tends to be lower than is the case when
a "pure" VOC strategy is compared with a "pure" NOX strategy. Thus, for the
example shown, Table 3(b) indicates that if the base level 03 concentration
is 0.20 ppm and the NMOC/NOX ratio is greater than 10:1, a smaller percent
reduction in VOC is needed to attain the NAAOS if NOX is reduced by a compar-
able percentage than would be the case for a "pure" VOC strategy.
While the ECR is a useful concept, it does not take account of cost
and/or technological feasibility of VOC vs. NOX controls. In reality, these
latter factors may be of overwhelming importance in selecting an appropriate
control strategy for a city. To illustrate the potential importance of con-
trol cost/technical feasibility factors, we introduce a new term: the "equal
control cost ratio" or ECCR. The ECCR is defined as that NMOC/NOX ratio at
which the cost of reducing NMOC to attain the 03 NAAOS equals that of reducing
NOX. Borrowing an example from Bilger (1978), to illustrate, suppose for the
situation exemplified by Figure 7, it cost twice as much to reduce NOX as it
does to reduce NMOC by the same relative amount. This would mean that it
would be more economical to control NMOC unless the NMOC/NOX ratio were such
that:
>. (2) ANO
Using this example, the ECCR is illustrated further in Figure 8, assessing
a peak observed 03 concentration of 0.20 ppm.
-------
^o 0:2oV7
FIGURE 8. Concept Underlying "Equal Control
Cost Ratio"
28
-------
Clearly, a strategy requiring control of both VOC and NOX would be more
costly than one calling for VOC reductions only, m the preceding example, a
"pure" VOC strategy would be preferred unless
ANHOC
TWTOlnltlal
or, more generally,
pure
ANMOC
strategy
- C°St atta1en' strategy Involving
reductl'ons 1n voc and NO
The ECCR would be expected to exhibit sensitivities to reactivity,
dilution and 03 severity which are in the same direction as those affecting
the ECR. However, it is much more difficult to make general estimates of the
ECCR, because it requires knowing cost vs. control functions for each source
category of VOC and NOX as well as a knowledge of the mix of VOC and MOX
sources within a city. As Rilger (1978) notes however, there is a general
perception that large amounts of VOC control are less costly and more feasible
than large amounts of NOX control. If this is true, we are probably safe in
assuming that, for "pure" VOC and "pure" Nox strategies, the ECR represents a
lower estimate for the ECCR in urban areas subject to 03 levels well above the
NAAOS.
In choosing a strategy to reduce 03 in and downwind from urban areas, it
may be useful to consider how rapidly one might expect any reduction in 03 to
occur. Grid models or EKMA can be used to provide Insight into this issue.
Figure 9 was obtained using EKMA/CB3 with boundary conditions, emission
patterns, default reactivity and medium dilution as defined in several reports
»y Gipson (1983, i984b. 1984d). The curves In Figure 9 depict daily maximum
-------
o»
IS
o
B
o
^
£
CO
<5
w §
f* o
£>§
OQ
HH ^
a
o
«
o
o
fe
g
5=
P
«<
i
!3
53
Q
r-»
PH
Si
fe
o
a.
a.
en
o
X
>-
CO
til
GC
CO
CJ
o
2:
a
x
o
I- X
CJ O
o
UJ O
GC Z
CJ
o
UJ
L
o
CD
o
(O
o
in
o
m
cr
o
X
o
CJ
o
0 6
^» »^
o
en
o
CXI
CJ
a
UJ
GC
UJ
o
GC
UJ
a.
§
O
)=«
I-
<
CXI
<
CXI
o
01
CVJ
o
CXI
CO
(O
CXI
30
-------
03 as a function of relative reductions in VOC, NOX or both precursors, given
a base level 03 concentration of 0.24 ppm and an initial NMOC/NOX ratio of 12:1.
Since, under the assumed conditions, an NMOC/NOX ratio of 12:1 is less than
the ECR (15:1) shown in Tables 3(a) and (b), percent reduction in VOC needed
to attain the NAAQS in a "pure" VOC strategy is less than relative reductions
of NOX or VOC and NOX in the other strategies. However, for the conditions
simulated, we also see that large reductions in VOC are necessary before a
"pure" VOC strategy becomes as effective as the others in reducing 03.
Before we leave this Section, it is necessary to remind the reader that
the conclusions/illustrations obtained with EKMA may depend on the input
assumptions to the model as well as the starting point on the EKMA diagram.
Although the results reported herein will be used, together with other types
of analyses, to draw some general conclusions regarding strategies in subse-
quent Sections, city-specific case-by-case analyses are recommended.
2.3 Implications From Trend Analyses
It is next appropriate to examine available air quality and emission
trend data to determine the extent to which effects of controlling NOX and
VOC predicted by models have been borne out by actual experience. Although
many urban areas have ongoing regulatory programs to reduce 03, the most
extensive and reliable records are available for the South Coast Air Basin.
Since this is also the region where the most extensive modeling of NOX control
strategies has been performed, our review of trend information will be confined
to this area.
Trend data are difficult to interpret, because they are subject to a
number of confounding factors such as annual fluctuations in meteorology,
growth patterns which are not uniform, changes in air quality instrumentation/
31
-------
procedures, potential differences in a monitor's microscale environment and
possible differences in emission inventory techniques over a long period of
historical record. In addition, there may be considerable uncertainty over
key factors such as historic emission inventories and NMOC/NOX ratios. In
particular, ratios of VOC to NOX emissions often differ from measured ambient
ratios. There are few ambient NMOC data available against which to compare
emission trends. Therefore, the best we can hope to do is to examine avail-
able trend data to see whether they provide qualitative agreement with what
we have been able to conclude so far about the relative effects of reducing
NMOC and/or NOX.
Figure 10 indicates the location of the four counties comprising the
South Coast Air Basin (SCAB), and the location of key monitoring sites used
in the subsequent analyses (SCAQMD, 1982d). Figure 11 plots yearly maximum
03 concentration averaged over 3 years. For example, values shown for 1965
are the average of yearly maxima observed at a particular site during 1964,
1965 and 1966. Averaging yearly maxima in such a way is intended to reduce
confounding factors introduced by varying meteorology. Data from each of
the 7 sites considered are plotted in such a way that their horizontal
position is roughly proportional to their distance from downtown Los Angeles,
so as to provide a rough view of 03 levels along the prevailing wind direction
To show trends in the spatial distribution of 03, the process used in plotting
the curve for 1965 was repeated at 5 year intervals for 1970, 1975 and 1980.
We see from Figure 11 that there was a marked improvement in the trend para-
meter for West Los Angeles and for downtown Los Angeles between 1965 and
1970, with little improvement thereafter. Further downwind, we note rela-
tively little change in 03 at Pasadena, Azusa and Pomona between 1965-70,
32
-------
UJ
CO
2
g
i-
<
o
o
<
H-
co
o
2
o:
o
5
<
h-
o
o:
i-
co
o
H-
2
LJ
^
LU
O
<
>
H
^
<
ID
O
; ,^ ^: .ln^/1"""
. '-if/'" .".4^>. vi:"'-^?v'x"VA^: ' ^.
^iiiiiiiitiiiiiiiiirtiM'^.'7' ' «j|]iit£ ^*: .. . jfr **.*.
*> ' '^..' \
^.-1>Sfr +* i^.-c'. « -*:
O)
3
CD
o
r
CL
CX)
33
-------
in
in
55
a
on
cc
UJ
m
<
I
o
o
O
a.
03
Z)
LU
O
<
CO
<
Q.
O
a
i » i I . » . I . i . i
I I I I I II I
HHdd "ON03 3N020 HflHIXVH
09
CO
tu
09 »=!
oe
o
o
oc
Li.
02 5
^^
01
O
Q
X
O
01-
34
-------
marked improvement between 1970-75 and relatively little change thereafter.*
At the farthest downwind site (San Bernardino) we note a slightly increasing
trend in 03 between 1965 and 1980.
In order to provide a possible explanation for the observations in
Figure 11, it is necessary to review demographic and emissions trends in
the South Coast Air Basin. Between 1960 and 1980, the population of the
four counties comprising SCAB has increased about 25%. However, the
most dramatic growth has occurred in Orange, Riverside and San Bernardino
counties. In 1960, the population in Los Angeles County was about 4 times
that of the other three counties combined. In 1980, population in these
other three counties totals about half that in Los Angeles County. Hence,
there has been disproportionately large growth in population (and emissions)
further east (i.e. downwind). Nevertheless, Los Angeles County emissions
are still, by far, the greatest in the Basin (CARB (1986)).
In Figure 12 we see that Rasinwide emissions of VOC have been decreasing
fairly steadily between 1966 and 1984. In contrast, NOX emissions increased
from the mid '60's to about 1978 and have decreased thereafter at a slightly
greater relative rate than the VOC reductions.
In terms of air quality trends in the Basin, we see that there are
essentially 2 periods of note: 1965-1975, where there was a pronounced
decrease in .aximun, 03 in Los Angeles County as well as basinwide, and 1975-80
where there is little change. During the former period VOC emissions decreased
by about 12% while NOX increased about !9%, leading to an estimated 26%
rT
overstated. urlng °-75 ay, for unknown reasons, be
35
-------
FIGURE 12
2200
750 .
500 «
250 _
000 _" ..----"
'50 .
500 .
250 _
ESTIMATED OZONE PRECURSOR EMISSION TRENDS
SOUTH COAST AIR 3ASIN
reactive hydrocarbons
oxides of nitrogen
966 1969
972
975
YEAR
978
981
984
SOURCE: CARB (1986)
36
-------
decrease In the ratio of VOC to NOX emissions. One would expect to see a
moderate decrease in maximum 03 within a county having ambient NMOC/NOX
ratios in the order of 10:1. We see from Figure 11 that between 1965-1980
Basinwide maxims were reduced by about 21% and Los Angeles countywide
maxims were reduced by about 32%. This reduction is somewhat greater than
implied from Airshed Modeling results where (I)-SCAOMD (mi, est1mated a m
reduction in the Basinwide maxima and a 25% reduction in the countywide
maxima accompanying a 50% reduction in the VOC/NOX emissions ratio (i e
ANMOC . .50%, ANOX=0) and (2) Cole et al. (1982) estimated a 20-35%
decrease of ozone maxima accompanying a 42% reduction in the ratio in the
smaller modeling domain at St. Louis. However, CARB (1986) has cautioned
that the apparent reduction in observed peak 03 may be overstated.
nuring the latter period (1975-80), VOC emissions were reduced by about
17%, NOX was about constant and the resulting ratio of VOC to NOX emissions
decreased by about 17%. nuring 1975-80, we observe a small (-5%) decrease
in the Basinwide maximum but an 11% increase in the Los Angeles County maximum.
The basinwide maximum during this latter period appears to have shifted
further downwind, but the trend in its value is small. A recent analysis of
Los Angeles trends by Kumar, et al. (1984, also notes a downwind shift in
Basinwide maxima as a result of the long-term strategy of emphasizing VOC
controls. ngure 13 1f reproduced from tMf ^ ^ ^ ^ ^^
the relative frequency with which the Basinwide daily maximum 0, concentration
occurs far downwind (San Bernandino) vs. closer to downtown Los Angeles (Azusa,
We note an increase in the percentage of Basinwide daily maxima occurring further
downwind, (at San Bernardino) between ,971-81 with a corresponding decrease at
37
-------
CO
CO
CO
fD
O
ss
w
oo
§g
{X4
O
O
^
w
or
w
CO
LU
O
GC
LU
0.
38
-------
Azusa. Thus, the observed shifts in the magnitude and location of observed
peak 03 levels are in qualitative agreement with model predictions.
In a recent report (CARB (1986)), CARB has reviewed the effect of the
more vigorous NOX control program begun in the late 1970's on 03 trends
observed in the South Coast Air Rasin between 1978-1983. Figure 14 is
excerpted from that report. Figure 14 uses a more robust trend parameter
(mean daily maximum 03 concentration) than shown in Figure 11. Further,
the trends shown have been spatially averaged for a number of sites. Never-
theless, trends shown in Figures 11 and 14 are similar. Both figures show a
pronounced decrease in Basinwide 03 maximum between 1965-75, with little
change between 1975-80. Both figures show a shift in the Basinwide maximum
further downwind. However, between 1978-83, Figure 14 shows a decrease in
both Basinwide and Los Angeles County maximas. In addition there is a small
Increase in 03 in the western (upwind) part of the Basin. CARB attributes
the foregoing observations to the more vigorous NOX reduction programs which
began to accompany the VOC control program in the late 1970's. The more
limited amount of the 03 increase in the NOX source-intensive area than that
anticipated from modeling results may be due to an emphasis on NOX controls
from elevated sources.
2.4 Summary
Modeling and trend information have been reviewed to estimate effects of
controlling NOX and NMOC on daily maximum 03 concentrations within one day's
travel time of a city. Our findings suggest that if the prevailing NMOC/NOX
ratio is less than about 10:1 and the reactivity of a city's NMOC mix is
typical, VOC controls are the most effective means for attaining the NAAQS
for 03. For cities in which NMOC/NOX ratios are about 10:1 or greater, the
39
-------
30
FIGURE 14
TRENDS BY AREA
C
0 25 .
N
C
E
N
T 20 .
R
A
T
o '5
0 1 «J -
N
I
o 10
p
P i
h
m 5 J
MHIB^_M^^MHH_
"*-
^
NE Los Ang«i«s Co.
I""H" San S«rnardfno/R I versfde
SW South Coast Air Basin
s .\.Tce les iDii---/ see -.e.xr
966 1968 1970 1972
974 1976 1978 1980 1982 1984
YEAR
SOURCE: CARB (1986)
40
-------
Picture is less clear. The most effective strategy (VOC, NOX or reductions
in both) depends on a number of factors including severity of the 03 levels
reactivity, dilution and cost. Generally, the Equal Control Ratio (ECR,
increases with increasing severity of the 03 problem, decreasing reactivity
and increasing atmospheric dilution. Because cost information is likely to
depend on a city-specific mix of sources and because of the other factors
mentioned above, case-specific evaluations of VOC vs. NOX strategies would be
necessary to determine the most appropriate strategy.
Modeling results from locations having NMOC/NOX ratios in the order of
10:1 suggest the following:
(a) Controlling NOX may increase 03 levels near sources of NO, unless
VOC emissions are reduced substantially (see Table 1). This undesirable
feature of NOX control can perhaps be reduced by emphasizing control on
elevated sources.
(b) In areas up to about 20-30 niles downwind of NOX source-intensive
locations subject to severe (e.g., > 0.1* ppm) 03 levels, moderate reductions
in HOX (e.g., ~25%) may reduce effectiveness of VOC reduction strategies;
benefits in these areas appear only after drastic curtailment of NOX;
(O Because of the distribution of urban populations, in areas'subject
to very high 03 concentrations (e.g., ,.os Angeles), exposure to concentrations
of 03 above the level of the NAAQS (e.g., Ml.lZppm) may increase as the
result of vigorous strategies to reduce NOX; however, exposure to higher
concentrations (e.g., >_ 0.16 ppm) may decrease.
(d) In contrast, VOC control appears to be uniformly beneficial with
greatest benefits occurring near locations where NOX levels are relatively
high;
41
-------
(e) In areas subject to severe 03 concentrations (e.g., >_ 0.18 ppm),
moderate control of both VOC and NOX ordinarily appears to be less effective
in reducing daily maximum levels of 03 within 20-30 miles of NOY source-
^
intensive areas than are strategies to control VOC alone by moderate to drastic
amounts. However, controlling both VOC and NOX may be more effective in
reducing 03 (sometimes the peak areawide 03) further downwind (e.g., > ~ 30
miles) than is controlling VOC only.
(f) Areas subject to less severe (£ 0.17 ppm) 03 may sometimes be able to
meet NAAQS more readily by reducing both VOC and NOX. A case-by-case modeling
analysis is necessary to prescribe the most effective strategy.
(g) Under some conditions, reductions in peak 03 may occur most rapidly
if both VOC and NOX are reduced. However in some cases, controlling NOX may
increase the amount of VOC control ultimately needed to attain the NAAflS.
Air quality trend data from Los Angeles appear reasonably consistent
(although not entirely so) with modeling results. During a period (1965-75)
in which VOC emissions decreased and NOX emissions increased (resulting in a
moderate decrease in VOC/NOX emission ratio), we see some reduction in the
Basinwide daily maximum 03 and an even greater decrease at sites nearer to
the major area of precursor emissions. During a second period (1975-80), in
which NOX emissions were about constant and reduction in the VOC/NO₯ ratio
^
was smaller, smaller changes in 03 were observed. During a third period
(1978-83) where NOX reductions accompanied VOC reductions, the 03 trend far
downwind reverses and begins to decrease. In addition, there is a small
increase in 03 near source-intensive areas. Finally, the Los Angeles trend
data suggest that one net effect of control programs (where selective reduction
42
-------
in VOC has been greater than for NOX) has been to shift the Rasinwide peak 03
further downwind in addition to reducing the severity of the concentrations.
The reader should be cautioned once again that our review is based
on analysis of relatively few locations using a limited number of modeling/
chemical assumptions. A case-by-case review is urged for specific regulatory
applications. Nevertheless, the results which are obtained appear, on the
whole, to be consistent with one another.
3.0 Ozone Concentrations In Rural Or Remote Areas
We now turn to the implications of NOX and VOC control on 03 observed
in rural/remote locations greater than a single day's (e.g., 10-hours) travel
time downwind from urban areas at prevailing wind velocities. Unlike the
"single day" problem discussed in Section 2, models to quantify effects of
controls on multiday or rural 03 levels are still under development. Hence,
judgments concerning the effects of controls must be based on more qualitative
information. Several years ago, the Office of Air Quality Planning and
Standards (OAQPS) asked the Atmospheric Sciences Research Laboratory (ASRL)
to prepare a "state of the science" assessment concerning the role of NOX in
rural 03 formation. The resulting effort has led to a review and assessment
by Altshuller (1986). Much of the information presented in this section is
based on material contained in this report. An additional earlier review
conducted for OAOPS (Martinez et al. 1979) has also been used liberally.
Four hypotheses can be advanced to explain high concentrations of
03 in rural/remote areas.
1. High 03 concentrations occur as the result of transport of
"fossil"* 03 from urban plumes.
'Fossil 03 is defined as 03 which is formed on the preceding day, remains
aloft overnight and is fumigated downward to the earth's surface the next
day.
43
-------
2. High 03 concentrations occur as the result of transport of
"partially spent" precursors from urban areas, or ozone which results from
subsequent reactions of these partially spent precursors.
3. High 03 concentrations occur as the result of in-situ emissions
of fresh precursors in nearby relatively rural locations.
4. High 03 concentrations occur naturally."
Each of these will be discussed in turn. To help us draw some conclusions
regarding the relative importance of each hypothesis, 1981-83 n3 data
reported to SAROAD have been examined for sites which are not: located in
Metropolitan Statistical Areas (MSA's). There are some 136 such sites. Of
these, 58 sites are either just outside of an MSA or can be readily associated
with an MSA less than 10 hours travel time away under prevailing wind conditions.
Twenty-one (21) of these 58 sites have expected exceedance rates greater than
1.0. Information concerning the remaining 78 of the 139 sites is contained
in Table 4. From the information in Table 4, we see that it is quite unusual
for concentrations greater than 0.12 ppm to occur at sites in SAROAD which
are not obviously impacted by emissions from urban plumes less than a day's
travel time away. In fact, only 13 of the 78 sites have observed any 03
concentrations greater than the level of the NAAOS during 1981-83. Only four
sites have an expected exceedance rate greater than 1.0 per year. Further
implications arising from the information in Table 4 will be identified in
Section 3.5. Finally, results from an application of a preliminary version of
the EPA's Regional Oxidant Model will be summarized, and possible implications
from these results will be identified.
44
-------
TABLE 4.
NON-MSA OZONE SITES NOT CLEARLY ASSOCIATED WITH AN MSAA
CA
IL
IN
IA
KY
Location
+ Valdez-Cordova
o Apache Sit.
Flagstaff
Prescott
Yuma
+ Bishop
+ Clearlake
+ Colusa
El Centre
Grover City
+ Lakeport
+ Mono Co.
Morro Bay
Ni porno
San Luis Obispo
San Luis Ob. Co
S. Lake Tahoe
Ukiah
+ Willows
+ Yreka
Effingham Co.
La Salle
Marion
Quincy
Kasciusko Co.
Madison Co.
Call away Co.
Carter Co.
Clay Co.
Falmouth
Fulton Co.
Hopkins Co.
Livingston Co.
Metcalfe Co.
Muhlenberg Co.
Ohio Co.
Paducah
Site
020565004J02
0300501 10A08
030280004F01
030660002F01
030960003F01
050780001101
051 390001 F05
051 520001 F01
052240002101
053040001101
053665001101
054760001101
054940001101
055150001101
057040002F01
057060001 F01
057830006F01
058400005101
058760001 F01
058860001 101
1 42220001 F01
1 40800001 F01
144720001 F01
146440005F01
1 52300001 F01
162360003G01
Years
1981
1981-83-
1981-83
1981-83
1981-83
1981
1981
1981-83
1981-83
1982-83
1981-83
1982-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981
1981-83
Top 3 03 Cone.
.111,
.075,
.095,
.099,
.112,
.080,
.070,
.110,
.180,
.110,
.080,
.090,
.090,
.110,
.090,
.110,
.100,
.080,
.110,
.070,
.109,
.128,
.114,
.098,
.113,
.095,
.096,
.075,
.092,
.080,
.101,
.080,
.050,
.100,
.120,
.100,
.080,
.090,
.090,
.100,
.090,
.100,
.100,
.080,
.110,
.070,
.102,
.127,
.111,
.093,
.100,
.095,
.094
.070
.092
.077
.100
.080
.050
.100
.120
.100
.080
.090
.080
.100
.OQO
.100
.090
.080
.110
.070
.100
.118
.106
.091
.097
.094
180500004F05
180620500F05
180740005F05
181140002F05
181360001F05
181840001F05
182320003F01
182760001F05
182960005N02
183080031F05
183180024F01
1981-83
1983
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
.143,
.114,
.107,
.119,
.105,
.135,
.133,
.109,
.129,
.129,
.115,
.129,
.110,
.105,
.113,
.101,
.121,
.122,
.107,
.126,
.122,
.114,
.116
.109
.094
.113
.100
.114
.122
.104
.106
.115
.111
Avg.
Exp Exc.*
0
0
0
0
0
0
0
0
0.4
0
0
0
0
0
0
0
0
0
0
0
Q
0.7
0
0
0
0
0.8
0
0
0
0
0.5
0.4
0
0.7
0.4
0
45
-------
State Location
KY
LA
ME
MI
MN
MS
MO
MT
NH
NY
NC
OH
OR
TN
UT
Prestonburg
Pulaski Co.
Russell ville
Trigg Co.
De Ridder
f\ T T
Galliano
Morgan City
+ Acadia Natl . Park
Caribou
+ Hancock Co.
+ Oxford Co.
Marquette Co.
Mankato
Madison Co.
+ Mark Twain Nat'l.
For.
o Custer Nat'l. For.
Berlin
Keene
o Essex Co.
+ Croatan Nat'l.
For.
Edgecombe Co.
Farmville
Lenior
Martin Co.
Robeson Co.
Dunn Co.
Conneaut
Logan
o Crook Co.
Riles Co.*
Logan
o Uintah Co.
Site
183400002F05
183460002F05
183600003F05
183860001 N03
1 90800001 F01
191025001F01
191940002F01
20001 0003F05
200260001 FOB
200495002F05
200885006J02
233280003F01
242100002F01
251 720001 F03
262950001A08
270310101A08
30004001 4F01
300340006F01
332020002F03
3409451 01 A08
341300099F05
341400099F05
342300003F03
342560099F05
343380099F05
350340003F03
361 480001 F01
363560002F01
3804201 11A08
4411 40001 N03
460500001 F01
461 200001 F01
Years
1981-83
1981-83
1981-83
1981-83
1983
1983
1983
1982-83
1981
1982-83
1982
1981
1981-82
1981-82
1981-83
1981-83
1981-83
1981-83
1981-83
1981-83
1983
1982-83
1981-83
1983
1982-83
1981-83
1981-83
1981-82
1981-83
1981-83
1982
1983
Top 3 0^ Cone.
.106, .106, .101
.109, .107, .102
.114, .110, .107
.129, .118, .116
.116, .081, .078
.114, .114, .103
.101, .098, .096
.138, .135, .128
.070, .065, .057
.142, .125, .115
.113, .108, .105
.095, .085, .085
.077, .072, .070
.123, .118, .113
.115, .115, .110
.085, .085, .085
.098, .097, .094
.135, .131, .107
.115, .114, .110
.095, .090, .090
.123, .116, .108
.108, .105, .099
.118, .117, .113
102, .097, .094
106, .101, .099
.081, .080, .079
.158, .154, .150
.110, .105, .105
.075, .065, .065
.012, .011, .011
.075, .072, .072
.070, .068, .065
Avg.
Exp Exc.*
0
0
0
0.4
n
w
0
0
3.2
0
1.6
0
0
0
0
0
0
0
1.7
0
0
0
0
0
0
0
0
2.1
0
0
0
0
46
-------
Avg.
State Location Site Years Top 3 0^ Cone. Exp Exc.*
VT Brattleboro 470120002F01 1982-83 .121, .109, .102 0
+ Green Mtn. Nat'l. 470265101A08 1981-82 .105, .100, .095 0
For.
+ Windsor 470600001F05 1981 .093, .091, .080 0
VA Fauquier Co. 481120002F03 1982-83 .129, .120, .114 0.7
Marion 481920005F01 1981-83 .110, .110, .100 0
WI o Chequamegon Nat'l. 510490001A08 1981-82 .090, .080, .080 0
For.
Marathon Co. 511920991F01 1983 .101, .092, .092 0
Platteville 512720007F01 1982-83 .107, .104, .102 0
o Vilas Co. 513640002F05 1981-82 .088, .084, .079 0
*Average expected exceedance calculations are estimated by averaging the computed
expected exceedances for each year that a monitor was operating. It is recognized
that the procedure of giving each year's calculated expected exceedance equal weight
is not strictly correct in cases where sample sizes vary greatly from year to year.
For the sites having non-zero expected exceedances, however, the variability in the
sample size does not appear to be great.
x Decimal error?
o Remote Site
+ Site not near towns >_ 5000 pop.
A All sites have at least one year in which the sample size was greater than 100.
47
-------
3-1 Hypothesis 1: Transport of Fossil Ozone
This hypothesis is that relatively high concentrations of 03 in
rural/remote areas are caused primarily by 03 transported aloft overnight and
fumigated downward after sunrise on the second day. This hypothesis would
also be appropriate for explaining relatively high 03 levels occasionally
seen during nighttime hours if, for some reason, the nocturnal inversion is
temporarily disrupted. Ozone in urban plumes carried aloft overnight is
chemically stable and not subject to dry deposition removal processes. How-
ever, it may be subject to dilution (and spreading) resulting from wind shear.
Certain portions of an urban 03 plume could be transported over large distances
by "nocturnal jet" winds aloft. Because nighttime 03 aloft is chemically
stable, subject to some dilution and capable of being transported long dis-
tances, one would expect it to result in broader (geographical) peaks with
lower concentrations than present In an urban plume during the first day.
Further, one would expect to observe the maximum effect from "fossil" 03 to
occur in the morning after the breakup of the nocturnal Inversion. Thereafter,
dry deposition and possible chemical decay due to reaction with vegetative
VOC emissions should reduce the impact of fossil 03. Fossil 03 would therefore
not be expected to be of significance for more than about 24-36 hours. An
exception to this generalization Is possible when a plume is transported over
water. In this case, the smooth surface of the water and the daytime strati-
fication of the atmosphere resulting from the flow of warmer air over cooler
water inhibit dry deposition. Altshuller (1986) cites European studies where
transport of elevated 03 concentrations over water is apparent for up to 48
hours. Because of 03 stability over water, coastal sites might be subject to
high levels of fossil 03 at anytime of day.
48
-------
Reduction of fossil 03 from urban plumes should be subject to the same
conclusions as those pertaining to the control of downwind 03 in "one day"
applications. Recall that for urban areas having NMOC/NOX ratios about 10:1
or less, photochemical grid modeling applications suggest that controlling
NMOC was uniformly beneficial, and controlling NOX is less effective than
controlling NMOC. Minor reductions in NOX in concert with NMOC control could
actually diminish the reduction in downwind 03 arising from NMOC control in
some cases.
However, in other cases having larger modeling domains, reduction of
NOX in addition to VOC appears to have beneficial effects in reducing 03 far
downwind. In the absence of regional scale model applications, the effect of
NMOC and NOX controls on fossil 03 is best estimated on a case-by-case basis
using urban scale models. In the case of grid model applications, the most
useful indicator would usually be the effect of controls on 03 predicted
later in the afternoon near the downwind boundary of the grid. In the case
of EKMA applications, the most useful indicator would be 03 predicted downwind
at the end of the simulation period (i.e., usually 5-6 pm).
As urban NMOC/NOX ratios increase, the shape of the EKMA/CB3 03 isopleths
suggests that less drastic reductions in NOX would reduce 03 levels downwind
at the end of the first day. EKMA results suggest that even for high urban
NMOC/NOX ratios (e.g., >15:1), strategies emphasizing NOX reduction would
often require substantial NOX controls (e.g., see Table 3) for an urban plume
including high 03 concentrations (i.e., _> 0.16 ppm) to be reduced below the
level of the NAAOS. However, somewhat less reduction in NOX might be neces-
sary to prevent an exceedance of the NAAOS due to fossil 03, if the fossil 03
is subject to significant dilution overnight.
49
-------
3'2 HyP°thes1s 2: Transport of Urban Precursors or Subsequent Prnri,,rt<
This hypothesis Is similar to the first one, except that In this case it
1s supposed that 2nd day elevated 03 concentrations result not primarily from
fossil 03. but from 2nd day reactions of transported 03 precursors. Overcast
weather on the first day might result in relatively small amounts of fossil
03 being transported. Ozone resulting from hypothesis 2 most likely would
not be high in the mid-morning, as would be the case with fossil 03. Instead,
one would expect high 03 to occur later in the day to allow transported NOX
and/or NMOC to react with each other or with locally generated precursors.
To evaluate the viability of this hypothesis, it. Is useful to review
information regarding the chemical lifetimes of NMOC and NOX. A review by
Altshuller (1986) indicates that the chemical lifetime* of NOX within an
urban plume varies inversely with sunlight intensity as well as with the
degree to which the urban plume is polluted. NOX reacts with hydroxyl (OH)
radicals or (1n the absence of S02) with NH3 to form nitric acid (HNn3) or
ammonium (NH4+) salts. These products are then subject to dry deposition or
absorption by water droplets so that the NOX from which they have formed Is
no longer available to generate additional 03. N02 Is also subject to reac-
tions with the intermediate products of reactions between aldehydes and OH
radicals. The result of this latter set of reactions is peroxyacetyl nitrate
(PAN). An equilibrium is established between PAN and N02. During the night-
time, this equilibrium favors PAN whereas during the subsequent day, conditions
.ore favorable to the regeneration of NOX occur. Altshuller (1986) indicates
1m1t f°r N°x I1f<"1me «ith1n an urban plume to be about 8 hours.
-
effects of dilution". a concentration, discounting the
50
-------
It is actually likely to be considerably less in downwind areas. Spicer
(1983) has shown that lifetime of NOX decreases rapidly at high NMOC/NOX
ratios such as those in rural areas. However, it is possible for some NOX,
particularly that emitted in the late afternoon before a nocturnal inversion
forms but after the period of maximum solar intensity (and OH radical concen-
trations), to be carried aloft overnight in the form of PAN. Nevertheless, the
NOX available in the urban plume at the end of the first day is expected to
be limited (Altshuller, 19846). Any available NOX would be expected to
rapidly decay subsequently on the second day.
Lifetimes of different NMOC species vary considerably. For example,
most olefins have lifetimes shorter than NOX, whereas the estimated lifetimes
of alkanes range from 2-3 times that of NOX to several days and longer
(Altshuller, 19846). Since the most reactive organic species do not constitute
a major fraction of NMOC, we would expect to find decreasingly reactive mixes
of NMOC and higher NMOC/NOX ratios as we move into more and more remote areas.
Indeed, reviews of available collocated NMOC and NOX data indicate a tendency
for NMOC/NOX ratios to increase as one moves into more rural locations (OAQPS
(1977)). Urban applications of EKMA/CB3 suggest that, under some circum-
stances, low upwind residual NMOC levels transported into cities can
significantly affect VOC controls estimated as necessary to attain the 03
NAAQS.
The foregoing information suggests that, in the absence of fresh
precursor emissions, the ability of an urban plume to generate additional 03
will eventually be limited by a lack of NOX. Further, strategies which
reduce NMOC (and as a result reduce concentrations of OH radicals) could have
the potential for prolonging the chemical lifetime of NOX. However, presence
51
-------
of residual NMOC In an aged urban plume could cause difficulty in meeting
NAAOS for 03 downwind of cities (Rraverman, et al. 1985), (Ripson, l
-------
throughout the day. Observed peaks would most likely occur in the afternoon.
Observed high 03 concentrations resulting from the plumes of individual point
sources would most likely be characterized by sudden increases in observed 03
levels. Impacts from such sources are likely to be highly transitory at any
fixed monitor location.
Observations as well as some modeling studies suggest that buildup of
03 within NOX plumes from large point sources is possible. Roth types of
studies, however, indicate that composition of the air with which such a plume
is diluted is of critical importance. For example, Altshuller (1986) cites
a number of field studies in the eastern U.S. where net increases in 03 of
20-50 ppb above background levels occur several hours-downwind from the point
source. Such a buildup has not been observed in rural western areas. These
findings are consistent with modeling studies which predict 03 buildups
within eastern NOX plumes which are qualitatively similar to those observed,
and which predict no buildup in western areas characterized by very low
background NMOC levels.
The potential for some 03 formation by emissions from small cities/
towns can be discerned from Table 4. For example, consider the rural sites
reporting within the State of Arizona. With the possible exception of Yuma,
these sites should be subject to similar meteorology. We note that the
Flagstaff, Prescott and Yuma sites are all subject to impacts from nearby
small cities/towns, whereas the Apache Sitegreaves site (Greenlee County) is
more remote. We see that the 03 levels at the remote site are lower than at
the other sites.
Kelly et al. (1984) have also recently tried to assess how much 03
observed in rural areas is potentially attributable to local emissions, as
53
-------
opposed to transport. These Investigators conclude that most of the diurnal
increases In rural 03 which they observed were attributable to transport of
03. but that in-situ formation from locally generated precursors Is responsible
for part of the Increase. Interestingly, Kelly et al. (1984) conclude that
In-situ formation of 03 In rural areas can be decreased by reducing rural VOC
emissions as well as NOX emissions. This result is surprising, because one
would expect NOX control to be more effective than VOC controls in reducing
03 generated in rural areas (where NMOC/NOX ratios are high). If this conclu-
sion is correct, one possible explanation may lie in the reactivity of rural
nixes of NMOC. Recall from Table 3 that the "equal control" NMOC/NOX ratio
Increases with decreasing reactivity of the NMOC mix. The range of reactivi-
ties considered in Section 2.0 is appropriate for urban areas. The developers
of the CB-3 mechanism caution against considering reactivities outside this
range with the mechanism as written (Killus et al. 1984). In order to be
able to more precisely assess the relative effects of NOX vs. VOC controls
on rural sources, a more versatile mechanism is needed, and local estimates
of prevailing NMOC/NOX ratios are desirable. It may be possible to better
resolve this question when more advanced mechanisms become available. In our
judgment, presently available information suggests that reducing rural NO
emissions should reduce 03 resulting from in-situ rural reactions of NOX a'nd
NMOC. In our opinion, the Kelly (1984) study does not represent a sufficient
basis for presuming reduction in rural VOC will be beneficial in reducing 03.
3-4 Hypothesis 4: Natural Sources
This hypothesis is that high concentrations of 03 in rural/remote areas
are attributable to natural sources. Two principal natural sources have been
discussed at length in the literature: stratospheric 03 and 03 formed from
54
-------
biogenic or geogenic precursors. In a recent review, Altshuller (1983) has
concluded that natural hydrocarbons do not contribute substantially to the
formation of 03 in the ambient air. Typical surface background concentrations
of 03 observed in rural areas within the U.S. during the late 1970's are in
the order of 0.02-.05 ppm (Altshuller 1984a). These concentrations would
constitute an upper limit for long-term surface concentrations due to strato-
spheric sources. However, surface concentrations substantially greater than
these levels are possible for short periods. These short term excursions are
not necessarily associated with meteorological conditions which ordinarily
correspond with high 03. Instead stratospheric intrusion is likely to be
accompanied by strong frontal passages and "tropopause folding" events.
Although these events are most likely in the spring in most parts of the
continental U.S., they can occur at other times. Such episodes would be
characterized by a high hourly concentration of 03 with relatively low con-
centrations before and after, and would be accompanied by the aforementioned
meteorological conditions.
Controlling NOX or VOC emissions would have little effect on high surface
03 concentrations attributable to stratospheric sources. The most appropriate
strategy would be to regard stratospheric intrusions as exceptional events,
and address the remaining incidences of high 03 with one's control strategy
(U.S. EPA 1986b).
3'5 IJse of Ambient Data to Evaluate Hypotheses
As described earlier, 03 data reported to SAROAn for non-MSA sites
during 1981-83 have been reviewed for the purpose of evaluating each of the
four hypotheses described in the preceding sections. This review can only be
55
-------
regarded as superficial, because meteorological data on the individual occa-
sions have not been reviewed. However, the air quality data themselves are
useful indicators of which hypotheses are most likely. In addition, Altshuller
(1984a) and Martinez et al. (1979) have reviewed earlier data from special
study sites. Martinez et al. (1979) conclude that all occurrences of 03 >
0.10 ppm observed during 1977 at 9 rural sites in the Sulfate Regional Exper-
iment (SURE) network are attributable to "long range transport" of 03 from
urban areas. Altshuller (1986) observes that fumigation by an urban plume
is the most likely and frequent explanation for high 03 observed in rural
areas. The results of the recent review are consistent with these earlier
conclusions. That is, based on observed diurnal 03 patterns and the location
of rural 03 monitors with respect to major urban centers, we believe that one
day transport of 03 from urban areas 1s the most common explanation for 03 in
excess of 0.12 ppm observed at rural sites. Therefore, the four hypotheses
described above are useful only as possible explanations for infrequent,
marginal excursions above the level of the primary NAAQS.
Table 5 identifies sites at which it is not obvious which, if any, large
urban areas may cause an Impact. As already noted, problems complying with
the primary NAAOS are rare at such sites. Thus, it 1s possible to focus on
the few occasions and locations for which daily maxima greater than 0.12 ppm
were observed. Table 5 summarizes available information concerning instances
in Table 4 where daily maximum 03 exceeds 0.12 ppm. Each site will be briefly
discussed.
The Conneaut, Ohio site's proximity to Cleveland makes it a good candidate
for single day impact from a large urban area. The site's proximity to Lake
Erie suggests the posslbnity of a lake effect, similar to that documented
56
-------
TABLE 5.
RURAL SITES/OCCASIONS WITH 03 _> .12 ppm
Location Description
Conneaut, OH NE corner of
Ohio, near Lake
Erie, about
60 mi . NE of
Cleveland
Pop. -15,000
:alloway Small County SW
County, KY corner of KY.
Located in Murray
(pop. -15,000.
i*uun uy pop *
-28,000
Located -35 mi .
SE of Paducah
(30,000); 40 mi.
W. of Clarksville,
TN (45,000)
eene, NH S. Central NH
~15 mi. N. of
MA line. ~~
Pop. -21,000.
40 mi. N. of
Springfield, MA,
50 mi . NW of Boston
)pkins, Co. Small County
'Y -40,000 in
NW KY -30 mi .
Hates
5/22/81
5/23/81
7/16/82
6/11/83
8/19/83
7/11/81
6/8/83
"*
8/11/81
7/7/82
i
9/9/81
Hours >
.12 ppm
1500-1900
1 200-1 800
1500-1600
1300-1700
1600-1700
1300
(>.100ppm,
1300-1600)
1500-2000
'
1600-1700
2000-2100
.
1500
1400-1500)
Peak
Hour
1800
1600
1600
1600
1600
i ____ ____
1300
1500
1700
____,
2000
.
1500
Peak Most Likely
Value^ppm Explanation*
139 u, 3
150 u, 1
131 u, 3
.154 u, 3
.158 u
.129 3
143 3,1,2
.
131 u, 3
.
.135 u, 1
_
135 u, 3, 4
-" " i-»v*iuviiic,
IN, (133,000) and
25 mi. SE of
Owensboro (56,000)
80 mi. NW of
Nashville (469,000)
-------
Rural Sites/Occasions with 03 _> .12
ppm
Location
El Centro,
CA
Description
Town of 21 ,000
in SE CA. 15 mi
N of Mexicali,
MEX (264,000);
^80 mi. E
of San Diego W.
intervening Mtn.
range.
Dates
12/20/81
Hours ^
.12 ppm
1100-1300
Peak
Hour
1100,
1200
Peak
Value,ppm
.180
Most Likely
Explanation*
u, 1, 4
Hancock
County, ME
Large, rural
county (-35,000)
on coast in
eastern Maine.
Surrounds
Acadia National
Park
7/8/82 1800
>.100 ppm
1600-1900
8/17/83
1800
1700-1900 1800
.125
.142
u, 1
u, 1
Fauquier Rural (-26,000) 6/23/83
County, VA county in N. Virginia
about 30 mi. SW of
DC, 30 mi. NW of
Fredericksburg and
60 mi. N. of
Richmond. Largest
town: Warrenton
(-4000).
1100-1700 1100
.129
u, 1, 3
Acadia
National
Park, ME
Muhlenberg
County, KY
Coastal
Park in
Maine
National
eastern
7/4/83
County in W. KY.
Largest towns
Greenville (4000),
Central City (3500).
County Pop. -27,000.
Site 4.3 mi. SSE from
Paradise Steam Gen.
Plant. 60 mi. NNW of
Nashville
7/28/83
7/10/83
1600-1700 1700
1500 1500
(>.100 1400-
1600)
1600-1700 1600
(>.100
1500-1800)
.129
u, 3
58
-------
Rural Sites/Occasions with 03 _> .12
ppm
Location
Ohio County,
KY
Description
County in W.KY
(-19,000) Site
in Dundee, about
20 mi. SE of
Owensboro and
40 mi. NW of
Bowling Green
(40,000). 75 mi,
N of Nashville.
Dates
7/11/83
Hours >_
.12 ppm
1400-1700
Peak
Hour
1600
Peak
Value,ppm
.129
Most Likely
Explanation*
3, u
Livingston
County, KY
Rural County
-5000) in NW
KY. Approx. 20 mi.
NE of Paducah
(30,000).
7/20/83
1600-1700
1600
.133
Trigg County
KY
Rural county
(-4000) at
land between
the Lakes Park
in SW KY. 20 mi.
NW of Clarksville,
TN (45,000)
8/6/83
1600
1600
.129
Transported "fossil" 03
2Transported, partially reacted precursors
3In-situ formation involving local emissions
4Natural sources
u-Impact from urban source area less than one day's travel time away
-------
for SE Wisconsin transporting an urban plume over the lake and then onshore
(Cole, et al. 1977). Secondary explanations are buildups in the late after-
noon due to reactions of precursors emitted by small cities/towns. In one
case (5/23/81, the 2nd day of a 2 day "episode") transport of fossil 03 may
be a v1ab!e explanation. The most likely explanation for the Callaway, Ohio,
Livingston and Trigg County sites (all in Kentucky) is that generation of 03
occurs from precursors emitted in small cities/towns. In most cases, the 03
is relatively high for prolonged periods in the late afternoon, and there is
no nearby major city. Local generation is also possible for the Muhlenberg
County (KY) incident, although the relatively short distance to Nashville
(^0 mi) makes a brush with that cUy's plume a possibility. The geograph-
ical location, as well as the timing of the Keene, NH incidents, make single
day impacts from one or more large urban areas the nost likely explanation
for high 03 there. The timing and short duration, as well as the location of
Hopkins County, KY, suggest several possible explanations for high 03 there
including a single day urban plume impact, stratospheric intrusion accompanying
a weather front or impact fro. an isolated point source plume. The most
likely explanation for high 03 at El Centro, California is the site's prox-
imity to a large city. However, the infrequency with which high 03 is seen
at El Centro as well as the timing and time of year of the incident suggest
stratospheric intrusion, transport of fossil 03 or spurious data as possible
alternative explanations. The timing and duration of the incidents of Hancock
County, Maine suggest single or multiday transport of 03 from urban areas as
likely causes. These explanations are possible at Acadla as we!!. However
the facts that high concentrations occur somewhat earlier at Acadia and
happen exclusively on holidays or weekends suggest that we cannot ru!e out
60
-------
local generation at Acadia. The location of Fauquier County, Virginia as
well as the length and timing of the incident there suggest the indicated
three explanations as most likely.
3'6 ^"cations from Preliminary Applications of a Regional Scale Model
Application of the first generation of the EPA's Regional Oxidant Model
(ROM 1) assessing the effect of VOC and NOX controls over a large region has
been described in a recent report by Lamb (1986). The ROM 1 application
simulates emissions and meteorological conditions occurring over a 9-day
period in July 1980. The modeling domain encompasses roughly a 1080 KM (E-W)
X 774 KM (N-S) area in the northeastern US, centered just to the NW of
Scranton (PA) and encompassing the entire northeast corridor from northern
Virginia to Maine. In this study, VOC and NOX projections between 1980-87, were *
made utilizing 1982 03 State Implementation Plans (SIP's). Although there is
spatial variability in the 1980-87 projections, over the entire grid VOC
emissions were reduced by 32% and NOX emissions were diminished by 8%. Thus,
the strategy could be characterized as one emphasizing VOC reductions with
some supplementary reductions in NOX. Both initial and boundary concentrations
of 03, NMOC and NOX were assumed to be irreducible and at concentrations
characteristic of natural background.
Because of uncertainty in many of the meteorological inputs (e.g., wind-
fields) to such a model, Lamb (1986) aggregates receptor locations according
to four groupings: urban, suburban, rural (agricultural) and wilderness
(natural). Generally, "rural" locations are defined as being 50-100 km from
an urban location and characterized by non-urban (e.g., agricultural) land
use. A wilderness (natural) location is greater than 100 km from urban areas.
Predicted changes in ozone characteristic of each receptor category are
reported. These include changes in peak and median hourly 03 concentrations,
61
-------
as well as changes in peak and median daily daylight (i.e., 7 hours: 0900-
1600 Local Standard Time) average 03 concentations*. The latter averaging
time may be of interest in assessing potential effects of ozone and control
programs on vegetation.
For purposes of this review, the following results of the ROM 1 application
are the most pertinent:
(1) The bulk of ozone generated by VOC and NOX emissions is produced
within about 100 km of major urban areas in the northeast;
(2) In general, at any given location and hour the 03 concentration in
the control case is less than or equal to that in the base case;
(3) Within each receptor group, peak concentrations are reduced by
larger percentages than the median values are reduced;
(4) Ozone is reduced more at sites near major VOC and NOX sources than
at locations far away.
(5) Similar patterns were observed for both 1- and 7-hour averaging times;
(6) Overall, the simulated strategy appears to have two effects: (a) it '
causes a delay in 03 formation, and (b) it reduces the total quantity
of 03 produced;
(7) In rural and wilderness areas, maximum ozone occurred later in the
simulation as a result of the strategy. Conversely, in urban and
suburban areas, the strategy seemed somewhat more effective midway
through the simulation.
in the highest houry e^ refers to the change
the 9-day simulation. 3 ^served anywhere in wilderness areas during
-------
As Lamb (1986) points out, results from the first generation ROM runs
should be regarded as preliminary and interpreted with care. For example,
ROM 1 uses an early chemical mechanism (Demerjian, et al., 1979), which may
or may not be suited for rural applications. Second, several simplifications
are made regarding windfields and vertical mixing. Further, biogenic emissions
were not considered. In locations where NOX is"increased, this could have an
effect on estimated changes in 03 concentration. Many of these shortcomings
are expected to be reduced or eliminated when the second generation ROM
becomes available.
Keeping the preceding caveats in mind, the preliminary results from ROM
app.ear consistent with many of the conclusions drawn earlier from our review
of air quality data and results of urban scale model applications. For
example, conclusions concerning preeminence of urban emissions in causing
high rural/remote 03 levels and the tendency of VOC reduction strategies to
delay 03 formation and shift (reduced) peaks further downwind, as well as the
assertion that strategies which reduce peak 03 downwind from cities should
also have some (reduced) beneficial effect in more remote areas, are all
reinforced by these preliminary findings.
3.7 Summary
It is concluded that single day transport from urban areas Is frequently
the most likely explanation for 03 levels above 0.12 ppm observed at rural
sites. As such, the same conclusions regarding NOX and VOC control which
were drawn in Section 2.0 for downwind areas subject to single day impacts
can be drawn here. Four hypotheses have been suggested as explanations for
the relatively infrequent occasions when 03 greater than 0.12 ppm occurs at
sites where single day transport from urban areas is not the obvious cause.
63
-------
Even for these latter Incidences, single day transport is often the most
lik-ly explanation. Of the regaining explanations, generation of 03 fro.
locally emitted precursors and overnight transport of fossil 03 appear the
next most likely. Presence of high NMOC/NO, ratios in rura! areas suggests
NOX control has potential for reducing 03 generated by emissions in rural
areas. A .ore quantitative assessment awaits characterization of reactivity
of rural nixes of organics, better documentation of rural NMOC/NOX ratios and
applications of a chemical mechanism more suited for simulating rural 03
formation, m cases where overnight transport of urban fossil 03 is the most
mely source of high 03, strategies to reduce single day downwind 03 should
>e effective in reducing 03 in a rura! area as well. lt 1s not poss1ble tQ
na*e a definitive Judgment regarding whether reliance on urban strategies
^sizing VOC control may exacerbate rural 03 in areas not presently obser-
-9 Nigh o,. However, preliminary applications of the EPA Regional Oxidant
Model suggest that strategies proposed in 1982 SIP's (i.e., emphasizing VOC
Auction in urban areas, may, on average, result in percent reductions in
Pea* and typical hourly and 7-hour average 03 concentrations in rura! and
««. locations which are directionally similar, but diminished from those
occurring at urban/surburban sites.
4*° °ther Environmental
« f,r, ,
3 «»»,tr.,,m , raau im<> ^
j(
»
or ^
-. .. . «-. »,. ,. ,ppropr,,t, a,tcm
64
-------
we identify three such issues: exposure to N02, acid deposition, and visibility
impairment.
4.1 Nitrogen Dioxide
As discussed in Section 2.0, N02 is formed from the oxidation of NO.
There are two major pathways for this oxidation to occur: oxidation by free
radicals formed from organic pollutants (i.e., photochemical synthesis) and
oxidation by 03 (i.e., 03 titration). We have also seen that, during periods
or locations where free radicals are plentiful, photochemical synthesis is
the predominant pathway. Implications regarding reduction of N02 associated
with each of these pathways were reviewed several years ago (Meyer et al.
1980). That review reached the following conclusions.
(a) Smog chamber data, empirical anfl chemical kinetics models all
suggest that reductions in NOX levels will lead to approximately proportional
reductions in both peak and mean concentrations of N02. An exception to this
may be the case in which peak N02 occurs solely as the result of titration
reactions with 03 that are limited by the amount of 03 available.
(b) The proportionality constant between peak (and mean) N02 and NOX
appears to remain approximately the same for a wide variety of NMOC/NO
^
ratios, emission patterns and dilution conditions. This implies that the
linear rollback model has the potential to serve as a useful means for
approximating the impact of NOX controls on ambient N02.
(c) The Information presented suggests that programs to reduce NMOC
levels will not have an appreciable impact on mean N02. There may, however,
be a small reduction in peak N02 accompanying reduction in NMOC for NMOC/NOX
ratios in the order of 10:1. Smog chamber data and modeling results examined
-------
1-ply a reduction In pea. N0? fro. 0.20 percent accompanying a 50 percent
reduction in NMOC.
(d) Because the proportionality constant between peak N02 and NMOC
reduction appears to be sensitive to a number of factors, the relationship Is
probably not a linear one.
(e) The preceding concisions appear vail* under NMOC/NO, ratios in the
of 10:1. However, chemical kinetics model stations (i.e., using the
«MA/nonfiE model, suggest that, if tht prevail1ng ^^ ^ ^ ^
to very low values (as a result of 03 SIPs), the relative effectiveness of
Nnx and VOC controls in reducing N02 is different. Por example, for ratios of
less than about 2-3U, reduct1on 1n peak ^ ^ ^ ^ ^^ ^
further VOC controls than to further NOX controls.
More recent modeling exercises conducted by SCAOMD (19825) with the
-ban Airshed .odel (Carbon Bond 2 .echanis.) 1n LoS Angeles also indicate
«« reducing NOX missions w1l, reduce short-term N02 concentrations by vary-
i". -unts throughout the Los Angeles Basin. These results are comp,icated
^ consideration of simultaneous VOC control strategies, non-unifo spatial
application of the VOC and N0X strategies, and the 1mpact of the N0x ^ ^
-tegies on the geographical location of the N02 peak. 61vM tht preced{ng
compilations, the analysis suggests that reductions in NOX may result in
Auctions in peak N02, which are less than proportional. Similar results
--een obtained in preliminary analyses available for Philadelphia (Burton
Presently, the NAA0S for N02 1s one which ^ ^ ^
l»t of locations with observed, annual average N02 values greater than
-------
0.035 ppm. This is shown in Table 6. As Table 6 indicates, there are
relatively few locations with annual N02 levels approaching the NAAns, and
only one (Los Angeles) where the NAAOS was exceeded. Meyer et al. (1MO),
have reviewed high hourly N02 data reported to SAROAD in the late 1970's.'
This review implies that, unless a short term NAAOS less than about 0.30'ppm
is promulgated, the need to meet N02 NAAOS is not going to provide a strong
rationale for additional national emission control programs to reduce NOX
unless or until long-term growth threatens to result in increased emissions.
4'2 Acid Deposition/Visibility Impairment
Acid deposition and visibility impairment have been lumped together in
this discussion for several reasons. First, with the exception of some
regulations to protect visibility in Class I (primarily rural) areas, there
are no major Federal regulatory programs currently in effect which specific.!!,
deal with these issues. Nevertheless, both issues are currently being studied
by the Agency, with a view towards the need for possib!e future regulatory
programs. Thus, strategies for dealing with NOX and MMOC to reduce 03 should
try to avoid exacerbating these other problems. Second, many of the same
processes thought to result in acid deposition are also thought to be Important
contributors to visibility impairment. Third, the ability to quantify the
effects of VOC and/or NOX controls on visibility and acid deposition is more
uncertain then is the case for 03 or N02. As a result, conclusions in this
section should be viewed as more qualitative and preliminary than others
presented In this report.
4.2.1 Acid Deposition
The two major acidic species contributing to acid deposition are sulfuric
acid (H2S04, and nitric acid (HN03) (NRC, 1983. NCAR, 1983). Acidic species
67
-------
TABLE 6
ANNUAL ARITHMETIC MEAN N02 CONCENTRATIONS*
MSA
Boston, MA
Nassau-Suffolk, NY
New York, NY-NJ
Baltimore, MD
Philadelphia, PA-NJ
Chicago, IL
Nashville-Davidson, TN
St. Louis, MO-IL
Denver-Boulder, CO
Salt Lake City-Ogden, UT
Anaheim-Santa Anita-
Garden Grove, CA
Los Angeles-
Long Beach, CA
Riverside-
San Bernardino, CA
*Values shown are the
Site
220240002F01
332900005F01
334680010F01
210120040F01
397140047H01
141340001G01
442540011G01
264280072H01
060580002F01
460920001F01
053620001101
050900002101
056680003101
Concentration, ppm
.044
.035
.041
.036?
.040
.052
.053?
.035?
.052
.037
.048
Year
1984
1984
1984
1984
1984
1982
1982
1984
1983
1984
1982
.062
.043
1982
1982
annual mean
°f Mm Potential number of
68
-------
may be formed by one of two methods: (1) through gas phase oxidation, or
(2) through absorption of precursor(s) by liquids and subsequent liquid phase
oxidation.
Gas phase reactions of importance include the following:
H20
S02 + OH »H2S04 + H02 (hydroperoxy radical) (1)
N02 + OH *HN03 ^
H20
£3 c. 0 3 V «J /
Liquid phase reactions thought to be significant are:
S02 + H202 (hydrogen peroxide) ^H2S04 + 02 (4)
S02 + 03 + H20 * H2S04 + 02
and, under some circumstances
Fe+3, Mn+2, H20
S02 + 1/2 02 . ^ H2S04
Nitric acid formation occurs about 10 times faster than sulfuric acid
formation, and results almost exclusively from gas phase oxidation. Reaction
(2) is thought to be the principal pathway for HN03 formation. Sulfuric acid
results both from gas and liquid phase oxidation, with reactions (1), (4) and
(5) thought to be the most significant. Above pH of about 5.5, reaction (5)
1s probably of greatest Importance among the liquid phase reactions (Scire,
et al. 19856). However, reactions (5) and (6) are self-limiting. For example,
for pH less than 4.0, the H2S04 production rate from reaction (5) 1s only
about 1 percent that of reaction (4). Thus, the reaction of S02. with hydrogen
69
-------
Peroxide (H202, (reaction 4) may be , ^
Actions »,. (2), (4) and (5)§
which are formed in the sequence Qf photolyi1f/oxidit1on ^^ ^
OC/NOX/03. For example, formation of H202 results fro. hydroperoxy
radicals, which themselves result from photolys1s of aldehydes. Hydrogen
Peroxide Is also ll.lted by compet1ng reactions Involvlno fo.atlon of N02
're. NO. Hydroxy! (OH, radical are fo.ed In the oxidation of NO to N02 by
HO* among other species. Presence of NO, N02 and 03 Is also noted. Thus
strategies Involving changes In NOX and/or NMOC levels could have a potentially
""inc.* l^pact on for.atlon of acidic species, m the following paragraphs,
Pertlnent resuUs of recent experimental and modeling studies are Identified
Splcer (1983, has studied formation of NOX oxidation products (1 .
HN03 and PAN, 1n s.og chamber experiments. ln these experiments a 17 "
component synthetic urhan mix was Irradiated In an Indoor chamoer under low
^dlty and little or no S02. NMOC/N0X ratios In the 18 experiments varied
fro. about 4.63:1. As the ,,., ^^ ^ ^ ^^ ^ ^ ^
NOX fon^ed oxidation products Increased. Thus, the I1fet1me Qf
-------
of H202 greater than that predicted by many photochemical models is required
to account for sulfate formation (i.e., H202 concentrations of roughly 4 ppb
would be necessary to account for S04= observed in rainwater, whereas only
about 1 ppb H202 is believed available). If it is true that S02 to sulfate
conversion at low pH's (i.e., pH < 4-5) is limited by processes in the
NMnc/NOx/03 photolysis oxidation cycle which affect the production and accumu-
lation of H202, control of NOX may have an effect on liquid phase formation of
H2S04. As we have previously noted, H202 is formed by combination of two
hydroperoxy (H02) radicals. However, H02 also oxidizes NO to N02. Thus,
reducing NOX could increase the availability of H02 to form H202. Increased
production of H202 could increase liquid phase formation of H2S04 (reaction
(4)).
The issue of liquid phase conversion of S02 to S04= has also been
investigated by Scire et al. (I985b). These authors use a chemical model to
hypothesize that, except at very low S02 concentrations, liquid phase S02 to
S04= conversion is first importantly Influenced by availability of oxidant
(i.e., H202) but is ultimately limited by pH. That is, after the small amount
of available H202 (e.g., 1 ppb) 1s exhausted, conversion takes place largely
as the result of reaction with n3 via reaction (5). Since background levels
of 03 would be more than enough to convert all available S02, conversion is
ultimately limited by PH. Availability of NH3 can increase SO, to S04=
conversion by raising PH. However, the acidity of the liquid droplet is
lifted by the fact that reaction (5) is pH llMlted. A key ins1ght Qffered
by Sdre et al. (i9«5b) is the timing or sequence with which various liquid
Phase reactions converting SO, to S04= come into play. If available H.O,
is eliminated before pH becomes a limiting factor in reaction (5), then the
71
-------
role of NOX control on liquid phase formation of H2S04 nay be less 1mportant
than would otherwise be the case.
Seigneur et .,. (1M4t.b) have also deve1oped an
of a .ode! ,.P acidic species fornatlon. The nodel 1s . box mode1 wh1ch stapts
with the CB-3 mechamsn and adds several reactions to take more detailed
account of S04=, NOj- and H202 chemistry (Seigneur et ,1 . (I984a)). m the
.ode,, transfer between gas and liquid phases 1s United by 11qu1d pnase
reaction rates, and, like the Sclre et a!, .ode! , the ultimate anount of
gaseous naterlal absorbed Into the !1qu1d Is ,1mlted by Henry's Law. Model
predictions are qualitatively consistent with compos1t1on of rain/cloud water
at three locations (Seigneur et al . 1984a).
The Seigneur mode1 has been used to note the 1n,pact of SO,, NOX and NHOC
on sulfate (H?S04> and nitrate (HN03) formation under clear skies and within
stratus clouds, during both sunner and winter. Table 7 has been adapted fron
one appearing 1n Seigneur et al . (1M4b,. Me see th§t§ m^ ^ ^
(dry, conditions, reductions 1n S02 and NOX lead to approx1mately proportional
reductions In S04= and N03- respectively, according to the .ode,. Reductions
i" NOX also result In snail reductions In S04=. Reduct1ons 1n NHOC ^ 1n
sn.aH, counterproductive changes In sunder and s.al, reductions In SO/ during
«1nter. Seigneur (1983) has exp,a1ned these findings In terns of countering
effects resulting fron reductions of OH and H02 radical production versus
longer Hfetlnes and reaction cycles for those OH radicals which renaln
Table 7 Indicates greater non-linearities 1n the relationships between
S04 , N03- and their precursors when cloud chenlstry ,s Introduced. We see for
exanple, that reduction of S02 and N0X continue to be beneficial In reducing
Predicted levels of SO,' and ,-. respectively. ^^ ^ ^
72
-------
TABLE 7.
CHANGES IN SULFATE AND INORGANIC NITRATE CONCENTRATIONS DUE
TO REDUCTIONS IN PRECURSOR CONCENTRATIONS*
Variable
Conditions
Clear Sky
Sulfate
Nitrate
ASO?
June
-48%
+1%
= -50%
December
-48%
0
ANOX
June
-18%
-55%
= -50%
December
-9%
-58%
ANMOC =
June r
+9%
+11%
-50%
)ppprnKop
-9%
+4%
Stratus Clouds
Sulfate
Nitrate
-22%
0
-26%
+2%
+30%
-41%
+33%
-32%
-7%
-5%
-9%
Estimated Overall
Effect
Sulfate
Nitrate
-35%
+1%
-37%
0
+6%
-55%
+12%
-58%
+1%
+11%
-7%
+4%
*After Seigneur, et al. (1984b)
73
-------
cloud water, reduction in NOX apparently has a fairly substantial counterpro-
ductive effect on S04= production. Seigneur et al. (19R4 a) attribute this
finding to lower reactivity of the cloud system which leads to a net increase
m OH radicals in the cloud environment. The "estimated overall effect-
numbers in Table 7 have been derived for illustrative purposes, assuming that
HN03 production results almost entirely from gaS phase oxidation, whereas gas
and liquid phase oxidation are roughly of eoual significance in H2S04 production
U-ing these assumptions, the Seigneur et al. (1984 b) modeling results imply
that NOX control wi,, lead to approximately proportional reductions in HN03
However, some .small increase in H2S04 may result. The model predicts that
reducing NHOC w1l, have re,ative,y ,1ttl. effect on HN03 or H2S04 concentrations.
S«re, et a,. (1985.) have utilized a regional scale grid model to
simulate the effect of 50, reductions in VOC, NOX and S02 on wet, dry and
total deposition of sulfur averaged over a single 7-day period 1n July !980
The model adds wet chemistry to a gas-phase chemical mechanism described by
L-ann, et al. (1984). The modeling domain in this study covers roughly the
eastern one-half of the United States as well as southern Ontario and Ouebec
The results are qual1tat1vely similar to those reported by Seigneur for acidic
species formation. For example, for the modeled period, Scire et al. found a
50, reduction in VOC resulted in little change in wet, dry or total sulfur
deposition and only a slight increase in average S04= concentrations (2-4,)
HN03 concentrations typically increased by 4, or less. A 50, reduction in
NOX emissions led to a typical net change in total sulfur deposition of +5,
while S04- concentrations were reduced 12-14, over a large portion of the
9rid. As with the Seigneur model, a 50, reduction in NOX emissions led to a
nearly proportional reduction in HN03 over large areas of the grid
74
-------
Several caveats with regard to the preceding discussion are in order.
First, the Seigneur and Scire (1985b) studies address formation of acidic
species rather than deposition. Formation of acidic species per se may not
completely describe the potential for acid deposition. For example, It is
believed that the dry deposition velocity for S02 is relatively high. Once
removed from the atmosphere, S02 may behave as an acid. Preliminary, 11mited
regional scale studies by Scire et al. (I985a) suggest that neither VOC nor
NOX control are mely to have major effects on S02 concentrations averaged
over several days and distances of 80 km or more. Further, once S02 remaining
in the atmosphere is converted to S04= by gas phase reactions, little deposition
is likely to occur until the S04= is absorbed by droplets and precipitation
occurs. Second, the Seigneur model has not as yet been applied to simulate
environments other than clear skies or stratus clouds. For example, results
simulating acidic species formation in stratocumulous clouds or in precipitating
clouds have not been reported, if these latter environments are relatively
important in contributing to add deposition (compared to clear sky or stratus
cloud environments), the generality of conclusions regarding NOX and NMOC vs.
acid species formation may be questioned. Similarly, the regional scale model
application described in Scire et al. (I985a) represents a single incident
and this model is still evolving. Finally, In the preceding discussion we
have ignored seasonal effects. It has been suggested, for example, that HN03
formation may be of greatest concern in the winter and spring due to the
effect of "acid shock" accompanying snow melt, if this is Indeed the case,
mechanisms which are effective in forming HN03 under low temperatures or
limited sunlight may be of somewhat greater importance than ImpHed in our
discussion.
75
-------
4-2*2 Visibility Impairment
This discussion is confined largely to the "regional haze" phenomenon as
opposed to i^acts fro. individual plu.es ("plu.e blight"). Secondary parti-
culate .atter, formed as the result of che.ical reactions in the atmosphere
i» generally believed to be an extre.ely important cause of regional visibility
extinction. Table 8 provides estimates of the relative instance of various
elects leading to visibility attenuation in the eastern U.S. (SAI, 1M5)
in the eastern U.S., sulfate aerosol is the single most i.portant cause
050%) of regional visibility extinction. Secondary organic aerosol is also
believed to be of sone .inor significance. Refining causes indude prinary
fin. partlculate (.ostly carbon particles) and N02. m addition to sulfates
(S04 ) forced as the result of reactions (4), (5, and (6) and reaction (!)
followed by absorption by water droplets, reaction of S02 or H2S04 vapor with
1. (NH3) is an extre.ely i.portant source of so,- aerosol resulting in
visibility attenuation in the East.
The consensus at present is that nitrate (NO,-, aerosol is a relatively
u-1.port.nt source of visibility inpair.ent in the East. For example, li.it-
ed .easurenents in the Snofcy Mountains indicate particulate N03- levels are
bout !-» of .easured fine particulate .atter whereas S04= constitutes about
50% of the mass of fine (< 2.5 m} part1culate mattep ^^ ^ ^
'984,. One reason for the H.ited influence of particulate N03- on visibility
Caster than the reaction of NH3 with NO," (SAI, 1985). In the East, where
-ere is plenty of sulfate, there is insufficient NH3 to convert gaseous HN03
3 aerosol. Recent .easurenents conducted in Uarren, Michigan by Cadle
(1985) indicate particulate N
-------
TABLE 8
PERCENT CONTRIBUTIONS OF VARIOUS SPFCIES TO
"TYPICAL" URBAN AND NONURBAN
FYTTMr-rrru. TM
EASTER S
Species
^-
Blue-sky Rayleigh scatter
Fine Particles (< 2.5 urn)
S04= (+H?0)
N03- (H26)
Elemental carbon
Primary Organic Carbon
Secondary Organic Carbon
Other
Subtotal
Coarse Particles
(2.5 m < d < 10 pm)
Soil dust ~~
Anthropogenic
Subtotal
Other TSP (> 10 pm)
Nitrogen Dioxide Gas (N0?)
Potal
Percent of
Urban
____
5.3
14.8
6.8
34.2
14.2
3.2
9.5
82.3
0.6
1.9
2.5
1.1
9.0
100.0
lotal txtinction
Nonurban
i,
45.9
18.3
0
6.0
2.8
13.8
1.4
42.3
1.8
5.5
7.3
3.7
0.9
100.0
urban
.
0
16.0
6.9
37.2
15.4
1.7
10.3
87.5
0
2.1
2.1
0.7
9.7
100.0
Nonurban
" ..
0
52.6
0
17.1
7.9
0
3.9
81.5
0
15.8
15.8
0
2.6
100.0
Source: SAI (1985)
77
-------
Th1S is approximately 20-5n* of the S04= levels at the sa.e site, Spending
on the season of the year. Thus, it 1s poss1ble that particulate ^ ^
somewhat .ore Important In reducing visibility In urban areas, m the Western
U.S., where there are 9eneral,y very low levels of S02 or sulfate, HN03 vapor
*ay be neutralized by ammonia to form am.onium nitrate (NH4N03, aeroso!
Hence, reduction of HN03 in the West .ay improve visibility there. We might
also speculate that if vigorous programs to reduce S02 are undertaken in the
East and NOX emissions are not reduced, visibility improvement might be less
than anticipated. This would presumably follow from greater availability of
NH3 to form N03" aerosol.
The Seigneur model, described previously, has been applied to estimate
the effect on visibility of reducing NOX by w. The results have been sum_
.arized by SAI (1985). Accordl.ng to th- ,,.,
due to reduction in N02 and (1n the western U.S. only, as a result of
reduced HN03. Increases in H202 production resulting from NOX control should
-crease liquid phase S04= formation and thereby reduce visibility. Ambivalent
-pacts on OH radical concentrations will ,ead to ambivalent impacts on
-s^lity. m summary, SAI (1985, estates that a 9% reduction in NOX will
lead to a ±« change in visibility in the ,ast and a ^ to -4* change in the
denotes an improvement in visibility in this context
78
-------
5-° Summary. Conclusions And Implications
In this review, we have considered potential consequences of controlling
VOC and NOX with respect to:
(1) 03 concentrations in urban plumes within one day's travel time of
the urban source of precursors;
(?) 03 concentrations observed 1n rural/remote areas;
(3) peak and mean concentrations of N0;>;
(4) formation of acidic species which may ultimately lead to acid
deposition, and;
(5) formation of aerosols which may lead to visibility impairment.
5-1 One-Pay Ozone Phenomenon
Our review suggests that controlling VOC and controlling NOX both have
the potential to reduce peak 03 concentrations in urban plumes. Whether or
not reductions will actually occur depends on a host of factors, including
prevailing NMOC/NOX ratios, reactivity of the NMOC mix and severity of the
city's 03 problem. EKMA/CB3 analyses reported herein indicate that VOC
control strategies are likely to be most effective if the prevailing morning
NHOC/NOx ratios are less than about 10:1. Preferences for a VOC strategy
should increase as the reactivity of an area's NMOC mix decreases and as the
severity of a city's 03 problem and the dilution to which an urban plume is
subject, increase. Thus, a city may well have an "Equal Control Ratio"
(ECU) greater than 10:1 if any of the following occur: a relatively unreactive
mix of NMOC, high dilution, severe 03 problems (e.g., peak 03 > 0.18 ppm).
Further, if one assumes that control costs associated with major NOX reductions
are greater than those associated with major VOC controls, in areas with severe
03 concentrations a VOC control strategy may be preferable to an NOX control
79
-------
strategy at NMOC/NOX ratios greater than the ECR's reported In this review.
Under the most likely conditions in urbans areas with 03 > 0.14 ppm, it is
Improbable that a strategy emphasizing NOX control would be preferable to a
VOC strategy unless the prevailing NMOC/NO, ratio were greater than about 20-1
The preceding EKMA analyses address the effects of VOC and NOX control
strategies on peak 03. However, the change 1n the magnitude of the peak
03 concentration does not tell the whole story. It 1s informative to
compare columns (12) and (13) in Table 1. The prevailing morning NMOC/NOX
ratios in Los Angeles are thought to be about 10-11:1, and we see that 50*
reductions in NMOC and NOX result in comparable reductions in predicted
Baslnwide peak 03 (-19%). However, the NOX strategy, while reducing the
peak, shifts it in, closer to the sources of NO (and the population centers)
Although not apparent from Table 1 (because the base peak is predicted to
occur at the furthest location downwind already), the Kumar et al. (1984)
trend analysis suggests that an NMOC strategy reduces the peak and tends to
shift It out, away from the population centers. This outward shift is consis-
tent with findings reported from modeling studies in St. Louis (Gipson 1982)
Furthermore, Table 1 Indicates that there is a marked adverse effect
associated with the NOX strategy which occurs in heavily populated areas
closer in to the city than where the predicted peak lies (at the DOLA and
Anaheim sites, for example). However, trend data from the South Coast
Air Basin suggest that this adverse effect can be diminished by selectively
controlling sources of NOX and by concurrently controlling VOC emissions. In
short, in a strategy emphasizing NOX control, things could get worse in the
-st heavily populated areas before they get better. In contrast, improvement
'» 03 levels closer 1n than the peak are noted with the VOC strategy. Since
80
-------
an NOX strategy may likely lead to a temporary deterioration in 03 concentra-
tions in the most populated areas, it is, perhaps, even more important that
the prescribed NOX reductions (made with little accompanying VOC reductions)
be implemented expeditiously than is the case for VOC controls under a VOC
control strategy.
Trend data and modeling results suggest that strategies in which some
NOX reduction accompanies VOC reductions may be more effective than VOC-only
strategies in reducing peak 03 to the level of the NAAOS under some conditions
if NMOC/NOX ratios are about 10:1 or greater. Such a strategy could also
have the potential for reducing peak 03 more rapidly than a VOC-only strategy
in some cases.
Although we have used "Equal Control Ratios", which we have derived from
our analyses, we have also noted that they can be sensitive to factors which
may vary from location to location. Further, there are risks associated with
Increased population exposure to 03 greater than 0.1? ppm associated with an
NOX strategy. Therefore, it is advisable for each individual city considering
a strategy incorporating some NOX in addition to VOC controls to generate a
city-specific EKMA isopleth diagram reflecting appropriate assumptions concerning
dilution, post-SAM emissions and, if possible, reactivity. A relatively sim-
ple check to insure that populated areas will be adequately protected during
the period before the full effect of the NOX strategy is realized, would be
to review present NOX and 03 concentrations during periods of high photochemical
activity (e.g.,~10AM - 4PM). If the sun, of the (03 + NO * N0?) concentrations
for each hour at each site in the urban core (i.e., where the potential adverse
effect of NOX control may be greatest) is less than 0.12 ppm, reduction in
NOX should not lead to an excursion above the level of the NAAOS where none
81
-------
existed before. Alternatively, a .ore sophisticated analysis with an Eulerian
photochemical grid model could be performed to insure the strategy is not
likely to exacerbate 03 concentrations above the level of the NAAQS in populated
areas. A third alternative might be to selectively reduce NOX only at those
sources not likely to impact nearby ground level concentrations of 03.
Barring such an individual analysis, we conclude that strategies emphasizing
VOC reductions (with or without small relative reductions in NOX) remain the
safest, most effective means for reducing urban 03 levels.
5-2 Ozone in Rural/Remote Areas
From our review of 03 observed in rural areas, we conclude that the
great majority of observed concentrations > 0.12 Ppm result from urban plumes
originating from large cities less than a day's travel time away, m our
discussion of the "single day phenomenon" above, we note that an urban VOC
control strategy, while reducing peak 03, may shift the peak further downwind
Hence, the possibility exists that such a strategy could exacerbate 03 levels
far downwind but still within a single day's travel time. We also note that
strategies to reduce urban NOX drastically may realize their greatest potential
benefits far downwind. In our opinion, the VOC strategy is not likely to lead
to significant increases in 03 levels further downwind than the site of the
Peak 03. This opinion is based on two findings, both of which have to do
with timing. First, it takes several hours (e.g., up to 10) for an urban
Plume to reach the locations we are concerned about. However, we have seen
that NOX has a relatively short lifetime, even at low NMOC/NO, ratios Past
reviews (OAOPS, 1,77) have indicated NMOC/N0X ratios increase further downwind
This, by Itself, would tend to decrease the lifetime of any remaining NOX
till further. Second, by the time a plume reaches these far downwind areas
82
-------
the sun has passed its zenith by several hours. Hence, the photolysis/oxidation
cycle has begun to slow down and is less likely to be able to compensate for
the effects of dilution.
Our conclusions are consistent with the 1965-80 03 trend data from San
Bernardino. We note that the 03 trend in San Bernardino increases slightly
over this period coincident with larger reductions in VOC than NOX emissions
upwind. This small upward trend could be reflective of the migration of the
population and emissions further east (i.e., downwind). In contrast, the
Urban Airshed Modeling analyses, in which uniform control strategies are
simulated for Los Angeles, predict reductions in 03 at the farthest reaches
of the modeling domain (-75 miles) accompanying all simulated VOC reduction
strategies. We therefore conclude that if a city's VOC strategy is sufficient
to reduce peak downwind concentrations of 03, 03 concentrations even further
downwind but within one day's travel time should also ordinarily be reduced,
but possibly by a smaller percentage. For a given reduction in peak 03
concentration, benefits of NOX control may exceed those of VOC control at
such downwind sites. If such an additional benefit occurs, 1t would do so
because the peak 03 concentration is likely to occur further upwind and
sooner under an NOX control strategy. Hence, these high concentrations will
be subject to further dilution before reaching remote/rural areas.
Our review of high 03 in rural/remote areas suggests that other potential
explanations result in high 03 much less frequently and in more moderate
levels than does the single day phenomenon. We have examined four hypotheses-
(1) transport of fossil 03; (2, transport of precursors; (3) in-situ formation
due to "rural" emissions; and (4) natural causes. Of these, natural causes
(I.e., stratospheric 03) should be distinctive by its accompanying meteorology,
83
-------
and .ay further be Identified by a distinctive diurnal pattern and time of
year in which it occurs. Hypothesis 3 appears to be the most likely of the
remaining three explanations. Reduction of NOX enissions occurring in rural
areas appears to have potential for reducing rural 03. Wore quantitative
estimates would require one to characterize NMOC/NOX ratios, estimate the
reactivity of the rural mix of NMOC and apply a model incorporating a chemical
mechanism adapted for use for a rural mix of NMOC. Reduction in fossil 03
should accompany urban reductions of NMOC or NOX providing these controls are
successful in reducing first day peaks. We believe hypothesis t (with regard
to NOX) is the most difficult of the hypotheses to evaluate. Unless NOX is
replenished by fresh emissions toward the end of the first day, NOX levels in
the plume at the end of the first day are likely to be very low. Even 1f the
fresh emissions occur and the NOX survives as PAN overnight, the next day it
is likely to decay rapidly (with the high rural NMOC/NOX ratios). However it
.ay take only very small increases in NOX to Increase ozone forming capacity
m rural/remote areas somewhat. Hence, some Increase in rural 03 due to in-
creased transport of N0X cannot be ruled out. W1th regard to NMOC transport
we have seen that transport of NMOC into N0x-rich areas (i.e., cities) can be
an important factor in urban attainment demonstrations under some circumstances
We expect upwind strategies to control NMOC will reduce some of this transported
precursor. Preliminary results obtained 1n an application of a regional
scale model are consistent with expectations derived from our analysis of air
quality data. That is, most 03 appears to result from urban emissions of VOC
and NOX, and strategies emphasizing VOC reduction appear to reduce 03 in remote
areas though by a smaller percentage than is true in urban/suburban areas
84
-------
5.3 NO?
Our review indicates that reducing NOX should also reduce peak and mean
levels of N02, whereas reducing NMOC has a relatively small, uncertain impact
on peak N02 only. Most recent analyses suggest that reduction in peak short-
term N02 may be slightly less than proportional to reductions in NOV. Our
/\
review of current data indicates that existing NAAOS for N02 are threatened
in only a few locations. Therefore, from a national perspective, immediate
concern for reducing current levels of N02 may deserve a lower priority than
concern for reducing current levels of 03.
5-4 Acid Deposition and Visibility
Two major acid species are of concern: HN03 and H2S04. HN03 is thought
to form almost entirely from gas phase reactions, and appears to be directly
proportional to available NOX. Hence, strategies to reduce NOX should be
effective in reducing HN03 vapor. Gas and liquid phase oxidation both are of
potential importance in the formation of H2S04. Reduction of NOX appears to
have a small benefit in reducing gas phase formation of H2S04. However, in a
stratus cloud environment, competing processes appear to cause NOX control to
lead to fairly substantial increases in H2S04 production. These observations
lead us to conclude that NOX reduction in the west (or in other areas where
there are low S02/S04= levels) should reduce the concentration of acid species
and, most likely, the potential for acid deposition. In the east, where
there is ample S04=, the picture is less clear. To get a more quantitative
assessment, it would be necessary (1) to estimate the relative importance of
liquid phase vs. gas phase formation of H?.S04, (?) to estimate the relative
abundance of H2S04 vs. HMO,, and (3) to estimate the relative importance of
dry vs. wet deposition. If we assume that (1) gas and liquid phase oxidation
85
-------
are roughly of equal importance for H2S04, (2) H2S04 and HN03 are about
equally abundant and (3) wet and dry deposition are about equally lmportant
then reducing NOX should reduce the presence of acid species and deposition
In the eastern U.S. according to available models. The impact of reducing
NMOC on concentrations of acidic species appears to be small and ambiguous,
according to the same models.
Secondary aerosols are the primary cause of regional visibility attenu-
ation. Sulfate particles are of particular importance. Sulfate aerosol
consists primarily of ( H2S04 droplets, formed by liquid phase oxidation or
by gas phase oxidation followed by absorption and (2, (NH4)2so4 particles
formed from reactions between SO, or H2S04 and NH3. In locations where S04=
is sparse, NO," aerosol may be of some significance in reducing visibility as
a result of reactions betweeen HN03 vapor and NH3. In addition, N02 may lead
to some visibility attenuation in urban areas. Hence, in the western U.S
reduction of NOX could Improve regional visibility, m the eastern U.S., the
effect of NOX control on visibility would depend on the relative importance
of aerosols formed by reaction of H2S04 vapor with NH3 vs. aerosols which
depend on liquid phase reactions to form S04=. The possibility exists that
-jor programs to reduce S04* (to improve visibility, may be less than success-
ful unless NOX emissions are also reduced.
5.5 Implications
5-5.1 Jmplications for Further Studies
1. Efforts should be undertaken to better define NMOC/NOX ratios as
well as ambient NMOC and NOX concentrations prevaiUng m urban areas.
The effectiveness of NMOC vs. NOX controls for a particular city depends
on the relative amounts of NMOC and NOX present. Potential detrimental
86
-------
impacts of NOX control near populated areas should depend on present levels
of NOX as well. Recent development of the cryogenic preconcentration approach
for measuring NMOC makes these measurements more reliable than formerly.
Monitoring these ratios may lead to more enlightening 03 trend analyses. The
following might be considered as follow-up activities:
(a) Select a limited number of pilot cities to begin measurement of
NMOC with collocated MOX in at least two locations per city.
(b) Scrutinize NOX measurements more closely than before using methods
such as those proposed by Richter et al. (1979) and closer checks on equipment
and procedures. Such efforts are needed to more completely ensure that high
quality data underlie calculation of NMOC/NOX ratios.
(c) Re-examine reasons for apparent discrepancies between ambient
NMOC/NOX ratios and those derived from emission inventories by reviewing
appropriate files from past runs of the Urban Airshed Model and by comparing
measured 6-9AM ratios with those derived using other procedures such as the
one recommended by Kill us et al. (1984).
(d) Continue efforts to evaluate the NMOC cryogenic preconcentration
technique and to make it more useable by State/local agencies.
(e) Conduct a limited number of special studies to characterize NMOC/NO
X
and reactivity in small cities/towns and surrounding rural countryside.
(f) Evaluate alternative means for estimating NMOC/NOX ratios, such as
measurement of NMOC/(NOX * PAN + HN03 + N03') ratios at sites observing high 03,
(g) Investigate potential usefulness of individual NMOC species as
tracers for source categories or as indicators of reactivity. Perform species
measurements periodically in conjunction with NMOC measurements.
87
-------
2. Efforts should be continued to develop and evaluate chemical
mechanisms which are readily adaptable for considering both urban and rural
situations.
To pursue this Implication, we suggest consideration of the following
activities.
(a) Continued support should be given for development and sensitivity
testing of the CB-X and ALW chemical mechanisms.
(b) Appropriate default assumptions (e.g., reactivity, levels and
composition of transported precursors, etc.) should be developed for use with
these mechanisms.
(O Incorporation of these mechanisms (or simplified versions) into the
computer model used in EKMA and into nore sophisticated photochemical dispersion
models should be undertaken. Further assessment of the effects of using dif-
ferent chemical mechanisms on the evaluation of control strategies should be
made using a standardized testing protocol.
3. Additional sophisticated modeling of implications arising from VOC
vs. NOX control should be undertaken. These efforts are needed to test the
generality of conclusions drawn from the Los Angeles, San Francisco and St
Louis modeling studies. They are also needed to more fully substantiate and
quantify conclusions which we have drawn concerning effects of controls on
high 03 levels in rural/re-note areas, as well as on 03 levels lower than 0.12
ppm which may be of concern for a secondary NAAQS or for future primary
NAAOS.
To support such modeling work, we suggest consideration of the following
items.
88
-------
(a) Continue development of the Regional Oxidant Model (ROM) and
carefully select strategy scenarios involving VOC control, NOX control and
measures to control both precursors. These scenarios should be used to
develop more quantifiable impacts on 03 in rural/remote areas.
(b) Perform an urban scale modeling study in which the modeling domain
extends further downwind than was the case in the St. Louis Urban Airshed
Model Studies, and perform a more extensive set of strategy simulations to
assess relative merits of VOC vs. NOX control.
(c) Perform Urban Airshed Model strategy simulations for a smaller
city, having potentially higher NMOC/NOX ratios and an NMOC mix having some
potential for differing from the default mix. Since the model has already
been run and evaluated in Tulsa, Oklahoma, this might be a logical location
for such work.
4. The Seigneur and Scire (1985b) models for acid species and secondary
aerosol formation need to be evaluated more extensively and reconciled, as does
the model described by Scire et al (1985a). In addition, efforts should be
made to investigate the suitability of models to predict N02, 03, acid species
and aerosol formation in and on the fringes of NOX and NMOC point source
plumes.
Seigneur et al. (1984a) make a number of suggestions for improvement/
evaluation for their model. In addition to these, sensitivity tests in which
NMOC/NOX ratios, reactivity, diurnal mixing height and emission patterns are
varied would appear to be useful. Evaluation and (if necessary) development
of a point source model is of interest for two reasons: (a) to help test the
extent to which conclusions arising from the Seigneur and Scire (1985b) models
can be generalized, and (b) to have a more specific treatment of specific
89
-------
sources which may be subject to requirements under 03> acid deposition or
visibility regulatory programs.
5. Available VOC and NOX control technology and the associated costs
should be reviewed systematically for various source categories. If strate-
gies incorporating NOX reductions gain more widespread use, the number and
diversity of potential solutions to the 03 probtem will increase. Greater
attention needs to be paid to the economic implications of different strategies
to help select the most cost effective one for an urban area.
5'5'2 Implications for Ozone Strategies
1. Our review suggests that VOC control continues to be the preferable
strategy for reducing peak 03 levels to the NAAOS downwind of cities having
NMOC/NOX ratios below about 10:1.
2. Cities having NMOC/NOX ratios about 10:! or greater could, in some
cases, reduce their downwind peak 03 concentrations to leve!s below the NAAOS
*ost effectively by using some MOX reductions In addition to VOC control
strategies. To determine whether a VOC-only or a VOC strategy with some
NOX control is most effective or feasible, a case-by-case Investigation would
be required. The most expeditious strategy depends on a number of factors
including atmospheric dilution, reactivity of the NMOC mix, severity of the
"3 levels and cost/feasibility of controls.
The information we have reviewed implies that benefits of NOX control
may increase (a) as NMOC/NOX ratios increase; (b, as NMOC reactivity increases-
(O as NOX concentrations during the daylight hours decrease and (d) as the
severity of the city's peak downwind 03 concentrations decreases. However a
strategy incorporating substantial NOX controls has higher risks associated
«th u than does one emphasizing VOC controls with or without small reductions
90
-------
in NOX. That is, in an area with a moderate to severe 03 problem, a strategy
incorporating large NOX reductions, may result in conditions in the most
heavily populated areas deteriorating before they improve. Further, such a
strategy may have several practical problems such as those described in
paragraph 4 below. Thus, city-specific analyses (i.e., grid modeling) con-
sidering the above factors would be necessary to justify adoption of a strategy
emphasizing NOX controls. In particular, such analyses should include some
means of assurance that the NOX strategy will not cause violations of the
NAAOS in heavily populated areas where no violations presently occur. In the
absence of a city-specific assessment of a strategy emphasizing NOX control
vs. one emphasizing VOC control, a VOC reduction strategy would appear to be
the approach of choice for reducing 03. Whether or not a VOC strategy might
best include ^some NOX reductions should be considered on a case-by-case basis
using city-specific EKMA or the Urban Airshed photochemical grid model.
3. Strategies which encourage increases in NOX emissions should not be
permitted, since such strategies may increase the potential for acid deposition,
visibility degradation and threaten NAAQS for N02.
4. Our review suggests that in some cities NOX control will not assist
in reducing 03 levels, whereas in others it may have merit. Consequences
of NOX control with regard to 03 appear to be much more of a "mixed bag" than
VOC control. Thus, in contrast to VOC controls, NOX control programs which
are consistent with flexible city-specific strategies would appear most
promising.
(a) For example, mobile source programs, such as I/M, which can be
applied selectively should prove useful.
91
-------
(b) Efforts to develop NSPS, define BACT for new sources and RACT for
existing sources of NOX should be useful. However, application of BACT/RACT
to sources of NOX as an 03 control strategy should be more selective than is
the case for sources of VOC.
5. On the basis of our review, we conclude that high 03 concentrations
(1 0.12 ppm, observed in rural areas occur almost entirely as the result of
impacts from urban plumes. We further conclude that strategies which are
effective in reducing peak 03 in urban plumes should generally also reduce
high 03 concentrations further downwind. Therefore, we believe the existing
strategy of giving higher priority to controlling VOC emissions in urban
areas is fundamentally sound. It may be desirable, however, to reconsider
the 200,000 population cutoff currently used to distinguish "urban" from
"rural". Our review suggests that, although not a widespread major problem,
emissions from population centers smaller than 200,000 can result in per-
ceptible increases in 03 levels. Therefore, it would be prudent to conduct
studies to see whether particular localities «200,000 pop.) are causing
violations of the 03 NAAOS and, if so, consider the need for additional
emission reductions. Determination of whether NOX or VOC strategies in small
cities or "rural" areas would be preferable would be most appropriately made
on a case-by-case basis.
6. We do not believe that the available evidence concerning overnight
transport of N02/PAN and subsequent reaction of N02 leading to second day rural
03 formation to levels above the NAAQS is particularly convincing. Neverthe-
less some increase in rural NOX, and hence, rural 03, due to urban VOC control
strategies cannot be ruled out at this time. In cases where NMOC and NOX
reduction result in equal reductions in first day peak 03 levels, the NOX
^
92
-------
strategy would probably result in greater reductions in 03 further downwind.
Counterbalancing this however, are findings which suggest that (1) transported
NMOC can last in measurable quantities for more than a day and (2) transported
NMOC can make the task of attaining the NAAQS in downwind NOX rich areas
(i.e., cities) more difficult if those cities employ VOC control strategies.
On balance, we do not believe that concern over 03 far downwind is a sufficient
basis for choosing an urban NOX strategy over a VOC strategy. However, it
nay help justify incorporation of some selective NOX controls in strategies
emphasizing VOC reduction.
7. If there is a local problem in complying with a present or future
N02 NAAQS, reduction in NOX emissions should be beneficial, unless the re-
sulting N02 is formed by reaction with 03 and is limited by the available 03.
NMOC reductions appear to have small, ambiguous benefits on peak N02 and no
apparent benefit on mean N02.
8. The limited information available to us implies that VOC control has
a small, ambiguous and nonlinear effect on formation of acidic species. Con-
trolling VOC may have some beneficial effects so far as improving regional
visibility is concerned. However, such improvements are likely to be small,
because only certain NMOC species form aerosol and such secondary organic
aerosol is not thought to lead to major regional visibility attenuation.
9. NOX control appears likely to reduce HN03 concentrations. Reducing
NOX has an ambiguous effect on H2S04 formation according to best available
estimates. In the absence of high S02 or S04=, reduction of NOX may improve
regional visibility. Thus, we believe that NOX control could be potentially
beneficial in the western U.S., both with regard to reduction of ambient acid
species concentrations and improvement in regional visibility. In the eastern
93
-------
U.S., reduction in NOX will most likely result in little improvement in region-
al visibility unless substantial reductions in S04= first occur. In the East,
some reduction in acidic species concentrations is most likely if N0X reductions
are implemented, provided the assumptions we have made are appropriate.
The preceding findings suggest that strategies to reduce 03 may not be
entirely compatible with strategies to reduce add deposition and/or to improve
visibility. In order to minimize adverse Impacts on cities with strategies
emphasizing VOC control to reduce 03, the following set of selective controls
should be considered for controlling sources of NOX in strategies to reduce
acid deposition or to improve visibility:
(a) NOX sources in cities opting for a strategy emphasizing NOX reduction
to reduce 03;
(b) elevated point sources of NOX in western rural areas;
(c) elevated point sources of NOX in eastern rural areas;
(d) elevated point sources of NOX in western cities; and
(e) elevated point sources of Nf)x in eastern cities.
6.0 Acknowledgments
A number of comments/suggestions were made by staff of the Monitoring
and nata Analysis division as well as the Atmospheric Sc1ences ^^
Laboratory. These comments and discussions were particularly helpful as were
more limited suggestions by a large number of peer reviewers. Finally, the
splendid clerical support provided by Mrs. Catherine Coats and Mrs. Josephine
Harris in typing and assembling this document is gratefully acknowledged.
94
-------
7.0 References Cited
A. P. Altshuller, "Review: Natural Volatile Organic Substances and Their
Effect on Air Quality in the United States", Atmospheric Environment 17 (11)
pp. 2131-2165 (1983). L '
A. P. Altshuller, Assessment of the Contribution of Stratospheric Ozone To
Ground Level Ozone"l:oncent rat Ions, EPA-60Q/S-64-144 (August 1984a).
A. P. Altshuller, Assessment of the Role of Nitrogen Oxides in Nonurban Ozone
Formation, ASRL, ORB, U.S. EPA, Research Triangle frark, NC 27711 (I984b).
A. P. Altshuller, "The Role of Nitrogen Oxides in Nonurban Ozone Formation in
the Planetary Boundary Layer Over N. America, W. Europe and Adjacent Areas of
Ocean", Atmospheric Environment 20 (2) pp. 245-268, (1986).
G. E. Anderson, S. R. Hayes, M. J. Hillyer, J. P. Killus, and P. V. Mundker,
Air Quality in the Denver Metropolitan Region, 1974-2000, EPA-908/1-77-002
(May 1977).
Association of Bay Area Governments (ABAG), Application of Photochemical
Models, Volume 1, The Use of Photochemical Models in Urban Ozone Studies
prepared under M.S. EPA Contract 68-02-3046 (December 1979). '
Association of Ray Area Governments (ARAG), 1982 Bay Area Air Quality Plan
(December 1982). *
K. A. Baugues, A Review of NMOC, NOY and NMOC/NOV Ratios Measured In 1984
a"d 1985» EPA-4bU/4-8b-OlS, fn Preparation,II.S. EPA. Office of Air Quality
Planning and Standards, Monitoring and Data Analysis Division, Research
Triangle Park, NC (1986).
R. W. Bilger, "Optimum Control Strategy for Photochemical Oxidants,"
Environmental Science and Technology 12 (8), pp. 937-940 (August 1978).
T. N Braverman and J. L. Haney, Evaluation and Application of the Urban
Airshed Model in the Philadelphia Air Quality Control Region. EPA 45Q/4-8R-nn3
(June 1985). '
C. S. Burton, Personal Communication, Agenda for Progress Meeting EPA
Contract 68-01-7033, Work Assignment 25, "Environmental Need for Heavy Duty
Truck NOX Controls", Systems Applications, Inc. (January 1985).
S.H Cadie, "Seasonal Variations in Nitric Acid, Nitrate Strong Aerosol
Acidity and Ammonia in an Urban Area," Atmospheric Environment. 19 (1)
pp. 181-188 (1985). : '*
California Air Resources Board, Technical Support Division, The Effects of
Nltrogen on California Air Quality. Report * TSD-8b-01, (March
-------
Model fof S Control ?n St
APCA Meeting, New Orleans, LA (June 1982).'
OzonLe
P"' ^I^' "APP"«tl-on of
Paper 82-20.1 , 75th Annual
Model for
'"c EKMA in the Ozone RIA,"
PapeV (Ser"l98lb)!Vlty °f EKMA/CB'3 t0 °rgan1c Activity," Internal
AMTB
<.-..
°f EK"A/CB3 to
Aloft," Internal AMTB
^c^l^^^^^^\^^^ -I--, - ^hr of
the Carbon Bond ^^-^^^^f^
96
-------
R. G. Lamb, Numerical Simulations of Photochemical Air Pollution in the
Northeastern United States: ROM1 Applications, EPA-600/3-86-033. (July 1986).
F. W. Lurmann, A. C. Lloyd and R. Atkinson, ADOM/TADAP Model Development Pro-
gram. Volume 6: Gas Phase Chemistry ERT Document P-B980-530* ERT, Newburv
Park, CA, (1984).
J. R. Martinez and H. B. Singh, Survey of the Role of NOX in Nonurban Ozone
Formation, EPA-450/4-79-035 (September 1575).
E. L. Meyer, n. H. Sennett, H. S. Cole and H. G." Richter, Technical Basis for
Developing Control Strategies for High Ambient Concentrations of Nitrogen
Dioxide, EPA-450/4-8Q-Q17 (September 1980K "
National Center for Atmospheric Research (NCAR), Regional Acid Deposition:
Models and Physical Processes, prepared for U.S. EPA under Interagency
Agreement No. AD49F2A203 (1983).
National Research Council (NRC), Committee on Atmospheric Transport and
Chemical Transformation in Acid Precipitation, Acid Deposition, Atmospheric
Processes in Eastern North America, National Academy Press, Washington. OC
(1983). ~ y
OAQPS, Effectiveness of Organic Emission Control Programs as a Function of
Geographic Location, Internal Report (April 1977K "
OAQPS, MDAD, Standard Metropolitan Statistical Areas (SMSA) Regulatory Analysis
Air Quality Data Base, 1982-84, Internal Report (February iQftfiV.
H. G. Richter, E. L. Meyer and D. H. Sennett, "A Graphical Procedure for
Screening and Analyzing High N02 Concentrations," Paper 79-12.4 Presented
at 72nd Annual APCA Meeting, Cincinnati, OH (June 197$).
H. G. Richter, F. F. McElroy, V. L. Thomson, "Measurement of Ambient
NMOC Concentrations in 22 Cities During 1984," Paper 85-22.7, to be
presented at 78th Annual APCA meeting, Detroit, MI, (June 1985).
J. S. Scire, A. Venkatrum, F. W. Lurmann, R. Yamartino, Summary of Recent
Sensitivity Analyses with the ADOM/TADAP Prototype Model (J)rat't). Document
P-D205-280, ERT, Inc., Concord, MA (hebruary 19&5a).
J. S. Scire, A. Venkatram, "The Contribution of In-Cloud Oxidation of SO?
to Wet Scavenging of Sulfur," Atmospheric Environment 19 (4) p. 637 (19856)
C. Seigneur, Personal Communication, Memo to E. L. Meyer (August 16, 1983).
C. Seigneur, P. Saxena, M. Dudik, G. Z. Whitten and P. M. Roth, Modeling
Studies of Sulfate and Nitrate Chemistry: The Effect of ChangesTfr^TTfur
Dioxide Nitrogen oxide and Reactive Hydrocarbon Levels. Systems A^TTTTTTnnc
^'hdation Coal Company and K.body Coal
97
-------
ct Results of SAI Alr.h.H
"
nuamv Tren" in
1sr1ct' -A-1r ua1ity
fui
,oa,
. 2 PP. 2*1-272 (1984)
Areas," Atmospheric Envlron-nt
ecnoogy. asadena
Vis1bimy Ejects
. Lees Emissions and Air n...m.,
6JCTOrandum "" I". C-litorma Institute
98
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. " [T
4. TITLE AND SUBTITLE
Review of Control Strategies for Ozone and Their
Effects on Other Environmental Issues
7. AUTHOR(S)
Edwin L. Meyer, Jr.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. EPA, Office of Air Quality Planning and Standards,
Monitoring and Data Analysis Division
Research Triangle Park, NC 27711
12. SPONSORING AGENCY NAME AND ADDRESS
Same
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
November 1906
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
35
EPA/450/4-ST-011
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
This review summarizes theoretical, experimental, field and modeling data related
to effects of reducing volatile organic compounds (VOC), oxides of nitrogen (NO ) -
or both for meeting the ambient air quality standard for ozone (O.J. Implications
are reviewed for several environmental concerns. These include 0~ levels within
and immediately downwind of major sources of precursors (i.e., cities), 0? con-
centrations in rural/remote areas, ambient levels of nitrogen dioxide (N0«),
acidic species formation and visibility attenuation.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
ozone
VOC control strategies
NO control strategies
rural ozone
acid deposition
visibility attenuation
18. DISTRIBUTION STATEMENT
Unlimited
EPA Form 2220-1 (Rev. 4-77) PREV.OUS ED.T.ON is OBSOLETE
b.IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report)'
20. SECURITY CLASS (This page)
c. COSATi Field/Group
21. NO. OF PAGES
114
22. PRICE
-------
------- |