REVIEW OF ANOMALOUS OZONE PHENOMENA
AND THEIR IMPLICATIONS FOR
AIR QUALITY CONTROL STRATEGIES
TEKNEKRONJNC.
-------
REVIEW OF ANOMALOUS OZONE PHENOMENA
AND THEIR IMPLICATIONS FOR
AIR QUALITY CONTROL STRATEGIES
Prepared by:
Teknekron, Inc.
1019 19th Street, N.W.
Washington, D.C. 20036
for:
Standards and Regulations Evaluation Division
Office of Planning and Evaluation
U.S. Environmental Protection Agency
Project Officer:
Joseph Kivel
BOA NO. 68-01-2433
Task Order No. 1
May 1975
Teknekron,Inc.
-------
REVIEW OF ANOMOLOUS RURAL OZONE PHENOMENA AND THEIR
IMPLICATIONS FOR AIR QUALITY CONTROL STRATEGIES
1.0 INTRODUCTION
Until about five years ago, ozone as.an atmospheric pollutant was considered
to be an essentially urban problem resulting primarily from photochemical
reactions involving mobile source emissions of oxides of nitrogen and re-
active hydrocarbons. Rural concentrations of ozone had not been systemati-
cally measured because they were assumed to be insignificant on the basis
that, during the transport of air masses from urban to non-urban art
*"' "*'
locally generated ozone content would be in part destroyed and in part diluted.
in fact, most rural ozone level measurements at ground level per-
formed prior to 1970 did not reflect concentrations high enough to represent
'a source of concern. The term "high" as used here references EPA's primary
and secondary National Ambient Air Quality Standards for photochemical oxi-
dants. Both of these are set at 0.08 ppm, which is a maximum one hour con-
centration not to be exceeded more than once annually. By a "high" concentra-
tion is meant one which exceeds this standard.
In 1970 EPA conducted an investigation of damage to Christmas trees in Mt.
Storm, West Virginia and Gorman, Maryland believed to be caused by S02 emis-
sions of local origin. In the course of air quality measurements performed
during this study, relatively high oxidant concentrations were encountered.
For example, of 1,782 hourly measurements made at one monitoring site, 60.8%
x^S
equalled or exceeded the £.06 ppgr standard and 5.7% .equalled or exceeded 0.1
ppm. It was noted that many of the highest values were observed at night,
rather than during periods of maximum solar radiation intensity.
Teknckron,Inc.
-------
-2-
The oxidant concentrations were measured during the period May 29, 1970
to September 9, 1970. Measurements of NO levels, performed during the
A ^ , n i -
same period at the same sites, indicated concentrations at about background
level and considered too low to account for local oxidant generation. One
.conclusion of the report was to the effect that the available data were in-
sufficient to provide a satisfactory explanation of the observed photo-oxidant
levels.
Additional measurements were made in Mt. Storm, West Virginia area by Research
2
Triangle Institute (RTI) during the period August 26 - September 9, 1970 , em-
«
ploying both colormetric techniques (for total oxidant) and a chemiluminescent
ozone meter. The data show a considerable average diurnal variation in oxidant
levels from about 0.04 ppm at 8:45 A.M. to 0.075 ppm during the afternoon and
evening hours (1:00 to 7:00 P.M.). Further comparisons of the data obtained
from the-two above techniques indicated that most of the observed oxidant
(about 90%) was ozone.
During a later (1972) study, Research Triangle Institute performed a study
in the areas of Garrett County, Maryland and Preston County, West Virginia
which overlapped much of the region in which EPA had conducted the earlier
investigations cited above. The key objectives of this study were:
verification of the high ozone concentrations previously
encountered in neighboring areas
delineation of horizontal and vertical distributions of
atmospheric regions exhibiting high concentrations of ozone
Teknekron,Inc.
-------
-3-
determination of ozone synthesis and destruction rates
and also transport patterns
identification of ozone source(s) (in the area studied)
The principal findings and conclusions were:
a) The occurrence of high concentrations of ozone was verified.
v/
About 11% of all one hour samples exceeded NAAQS (0.08 ppm)
b) The local atmosphere was considered to be well mixed, so that
it was not believed that the high ozone concentrations were >/
restricted to ground or near ground level
c) The area of high ozone concentration occurrence was relatively ^
wide spread
d) While ozone synthesis rates could not be determined, destruction
rates (normalized)were considered "well within the range expected
for naturally occurring'agents of local origin."
e) Local ozone formation was not considered to be a likely causal
factor. Data resulting from measurements near high voltage power
lines was inconclusive; local atmospheric concentrations of N02
and non-methane hydrocarbons were at background levels and thus
the possibility of significant ozone generation by photosynthesis
was ruled out.
f) The following hypothesis was advanced to explain the observed
ozone concentrations: Ozone precursors, generated at (unidenti-
fied) sites remote from the study area, enter the troposphere.
Teknekron,Inc.
-------
-4-
The air mass containing these is, because of 'appropriate
meterological conditions" transported to the area. Ozone
4£^syntnesized during transport. It is also assumed that
that the precursors (which are also ozone destructors) are
consumed in the process, leaving appreciable CL concentrations .
which reach lower atmospheric levels through turbulence.
During the summer of 1973, Research Triangle Institute conducted a four
4
state investigation for the purpose of determining the extensiveness of
nonurban distribution of high ozone concentrations. The study sites were
located at McHenry, Maryland; Kane, Pennsylvania; Coshocton, Ohio; and
Lewisburg, West Virginia. The major findings were:
a) NAAQS for photochemical oxidants were exceeded at these sites
with significant frequency (15% to 37% of the hours from June 26
to September 30, 1973 for the different sites).
b) Except at McHenry, Maryland, N0? concentrations were insignifi-
cant.
c) The pattern of diurnal variation, with the 0, concentrations
peaking late afternoon or early evening, was similar for all
four sites.
d) Changes in ozone concentration at the four monitoring sites
occurred with little, if any, lag.
While the above study was in progress, ozone measurements above ground level
were conducted via aircraft during.a few days in August and September 1973.
Tcknekron,Inc.
-------
-b-
The data indicated that the concentrations of 0^ observed at ground level
extended up to several thousand feet, but not above the inversion layer.
The basic conclusion was that the observed ozone concentrations^were gejierated
in the 1 ower atmojpMiie-.aci.cL.w.epe-nQ't^t-pa.n.s.ported.-v.er.t-ica-Lly..downward from
the stratosphere in consequence of turbulence.
The observation common to most of the above studies to the effect that ozone
concentration peaks generally occurred after the periods of maximum solar
radiation was specifically addressed only in the last report cited (refer-
ence 5). The opinion was expressed that this delayed peak was more suscept-
ible to explanation on the basis of NO oxidation occurring in association
with hydrocarbons of relatively low reactivity than through transport from
the stratosphere.
There have been several reports of high concentrations of ozone in western
non-urban areas, such as Pt. Magu, California where the high (L levels were
believed to be caused by precursors generated in or near Los Angeles. Other
instances of high non-urban ozone concentrations in California which are
attributed to transport (as opposed to local origin) occurred in Mineral
7 8
King Valley and in Indio which is 140 miles from Los Angeles. The Indio
observations are of particular interest in that the observed 03 levels ex-
ceeded the 0.08 standard with higher frequency than was true of any other
monitoring site within the State.
Teknekron,Inc.
-------
-6-
Other reported observations of high 0,. levels in non-urban areas include:
q
Miami, Florida (attributed to long range transport)
Several sites in Florida and Alabama . Jhe^datr NAAQS were
exceeded most frequently during winter and early spring. (M.
Martinez of EPA's Monitoring and Data Analysis Division has
noted that these seasons are preceeded by continental air
mass movements and that this suggests long-range transport
of 03.)
New York State . Correlation of ozone levels observed at
urban and non-urban sites suggested a common source which is
believed to be external to the State. The report questions
whether the observed high levels of intrastate ozone can be
attributed primarily to manmade sources as opposed to the
contributions of naturally occurring precursors and transport
from the stratosphere.
12
Wisconsin . During the summer of 1973, observed concentra-
tions of ozone exceeded NAAQS at seven sites, one of which
was a rural area in the south-central portion of the State.
Wind transport was believed to be the causal factor.
The above investigations by no means exhaust the list of rural ozone con-
centrations studies performed to date. They do, however, suffice to indi
cate the prevalence of occurrence of high non-urban levels of CL.
Teknekron,Inc.
-------
-7-
2.0 .DISCUSSION
A variety of mechanisms has been suggested for accounting for observed rural
ozone concentrations. High non-urban levels (that is, levels exceeding
NAAQS) are of particular concern because of their possible air pollution
control implications. The term "possible" is used in this context because
these implications cannot be considered independently of the causal mechan-
isms, some of which may lie beyond the reach of control measure capabilities.
The principal mechanisms which have been considered to date include:
a) Meteorological transport of anthropogenic ozone to non-urban
s i tes.
b) Meteorological transport of anthropogenic ozone precursors
to non-urban sites, with ozone production occurring either
en route or on.arrival.
c) Meteorological transport of naturally occurring statospheric
ozone downward to ground level.
Local formation by high voltage power lines.
e) Local photochemical formation due to natural precursors. '
/f)/ Local formation caused by lightening.
^/
At least two of the above mechanisms can obviously be eliminated as signifi-
cant causes, namely (d) and (f). As indicated earlier, measurements of ozone
formed in the vicinity of power lines did not show conclusive results. Fur-
ther, 0- formation by this means would be local rather than pervasive. While
lightning discharges may generate appreciable quantities of ozone, such
generation i s sporadic_and_i,n_n.Q_y/ay_ ciLrrelaias. w.lth-.tJbe,,observed data.
Teknekron,Inc.
-------
-8-
The possibility of appreciable rural ozone formation caused by solar irradi-
ation of naturally occurring precursors (e) has been considered. R.A. Rasmussen
- =_=^CT~~=>
and K.H. Rasmu^s_en___jstimajbeJtjiat the total world-wide emission rates of re-
active hydrocarbons of botanical origin may exceed total antropogenic HC
'
emission rates by a factor of six. The generation of hydrocarbons by plant
species is, from a global viewpoint, a widely diffused process so that con-
centrations of HC ini^urJMjasnay be less than those observed in urban
'~~ ' ~ -
and near-urban areas characterized by large volumes of veh
.......,JEg,3r~E.-a: ft i-T^ar- 1 g-r~r**-^r **-£!' " CT-^TT: -r- **-£ «ST-.»-.XJT y^- -r*fc»-?^*"s»«».J'.^«*i^ ' '-
and/or the presence of stationary emitting sources. Further, in some rural
settings, remote from man-made emissions, it was noted that prevailing NO
/\
concentrations were at or near background levels. This, however, is not
intended to imply that the role of naturally occurring precursors in rural
ozone formation is wholly dismissed as a serious possibility. EPA reaction
chamber studies conducted at RTP indicate that the ratio of CL molecules
formed per reacting carbon atom is about 4/1. Non-methane hydrocarbon con-
15 ^
centration.measurements performed in rural areas of Ohio in 1974 indicated
average values ranging from 0.09 ppm to 0.33 ppm (as carbon) for three study
sites. Such concentrations of hydrocarbons, if fully reacted,could potentially
generate ozone levels substantially above NAAQS (0.08 ppm). (In this context,
note that the national primary and secondary ambient air quality standard
for hydrocarbons is 0.24 ppm as a maximum 3 hour concentration not to be
exceeded more than once annually. A concentration, of _on1.y onejtenth of jthis !
standard represents a potential ozone level of approximately 0.096 ppm.)
Teknekron,Inc.
-------
-9-
The possibility of vertical transport of naturally occurring stratospheric
ozone to ground levels (c) during frontal passages (or through other meteoro
logical mechanisms) has often been suggested as an explanation for high ob-
served rural concentrations of oxidant. Rasmussen considers that ozone
transported from the stratosphere may typically account for about 0.03 to
0-. 05 ppm of the total observed ground levels. He also estimates that pre-
cursors of plant origin may contribute 0.02 ppm. This would imply a "natural"
background of up to 0.07 ppm, although more usual background levels are said
to be of the order of 0.04 to 0.05 ppm. /
Meteorological transport of anthropogenic ozone (a) and ozone precursors (b)
from urban areas and from sites containing stationary sources of HC and NO
X
emissions is regarded as a key, if not the prime, causal agent accounting for
high rural oxidant concentrations. Wind transport is believed to be a signi-
ficant fa'ctor in the transfer of ozone and/or precursors from specific sites
of origin (such as urban areas) to downwind rural locales where high local
concentrations are observed. This mechanism appears to be substantiated for
relatively short transport ranges of the order of 50 to 100 miles both for
ozone and other pollutants (such as particulates and SO-)- Long range trans-
port is attributed to large air mass movements and this view is consistent
with observed correlations of non-urban ozone measurements made at sites sep-
arated by great distances within the eastern portion of the U.S. Because of
the large size and slow movement of these air masses, the transported pollut-
ants (0., and/or precursors) may be picked up from several different sites and
Teknekron,Inc.
-------
-10-
therefore may be more appropriately considered regional in origin rather than
as attributable to a specific urban area.
Considerations of the above kind, coupled with the fact that hydrocarbons
previously regarded as non-reactive are now know to be "slow" (L precursors,
influenced the-orientation underlying EPA's recent draft paper entitled
"National Strategy for Control of Photochemical Oxidants," January 15, 1975.
In essence, this paper recommends an areawide or regional approach to HC emis
sion control and stresses the importance, in this regard, of stationary as
well as of mobile sources.
The prospects for achieving and maintaining NAAQS for rural ozone are not
regarded with uniform optimism throughout EPA, even assuming that HC controls
were to be fully effective on a regionwide basis. One view, expressed by
the Gas Kinetics and Photochemistry, Section, Atmospheric Chemistry and
Physics Branch of the Chemistry and Physics Laboratory at NERC/RTP, is based
on the following: Under given conditions of irradiation, the quantity of
ozone generated through precursor reaction is a function of the HC/NO ratio
A
of the precursor mix. At small ratios, ozone formation tends to be limited.
As the ratio increases, the rate of production rises (up to some optimum
ratio). Also, high ratios tend to favor the effectiveness^ of hydrocarbons
of low reactivity in ozone_generation, although, as stated earlier, the pro-
cess is slower. It is argued that in those cities, such as Los Angeles and
Philadelphia, where urban ozone has been reduced through hydrocarbon emis-
sion control measures, this reduction largely reflects the attainment of a
Teknekron,Inc.
-------
-11-
lower HC/NO ratio which is inherently less favorable to ozone generation.
/\
However, through the meterological transport of urban air containing residual
oxides of nitrogen to rural areas, the quantitative relationships between HC
levels in these areas and the transported oxides establish HC/NO ratios which
X
are conducive to 03 formation, even though the absolute concentrations may not
be high. This argument implies that even if rural ozone generation arising
wholly from natural precursors can be shown not to be a significant causal
factor (with respect to high concentrations of O-J, the role of natural hydro-
carbons (in association with ^n^hrp5ogjnjc_oxides ^f nitrogen transported to
rural areas) may be of considerable importance. It also implies that control
_^rrm-_ j.-Bqi.:^^s^^gi-~ _-^^_ii- ^iagaafe'^"Ji*P''i'<*r'g- "*" -~-"T--" *!*s """ - "*" - " -"" "*T "*"' ""r
programs directed to the reduction of HC emissions on a regional basis may,
in the absence of related measures addressed to NO , prove insufficient to
X
ensure the attainment and maintenance of NAAQS for 0, in some rural areas.
Beyond this, a question arises with respect to the adequacy of NAAQS for HC
and NO . It is noted that the standard for the latter of 0.05 ppm represents
X , _ . -- -- . i »IP "'"* ' ' " II II ' I "'I"" "''T -"""" ' -- "
an annual arithmetic mean only and .not_-a.n_UPBer. limit to short period (one to
-^ , -
three hour) concentrations.
Teknekron,Inc.
-------
-12-
3.0 .CONCLUSIONS
Some of the following comments are presented as tentative inferences because
it is not considered that information and data available to date are sufficient
to warrant firm conclusions. This view is essentially consistent with that
of the Agency which is inherent in the fact that EPA is now supporting and,
proposes to support in the future, technical research programs directed to the
elucidation of the rural ozone problem, among other oxidant related questions.
The following views are oriented to the consideration.of high rural 0, levels
within the reference framework of interests of the Standards and Regulations
Evaluation Division rather than from a purely technical standpoint.
Certain presumptive, or at least potential, causes of high
concentrations of rural ozone are, by their nature, beyond
EPA's control.
These include:
a) lightning
b) solar irradiation of natural precursors
c) downward movements of stratospheric ozone
It is considered by some investigators that these mech-
anisms do not substantially contribute to observed high ozone
concentrations in rural areas. This very likely is true of (a).
It may be true of (b) because, in the absence of introduced pollut-
ants, rural air is fairly free of NO so that the local HC/NO
ratio may be well beyond the peak of the curve at which 0 genera-
tion is maximum. The question of stratosphere to troposphere
ozone movement (c) is not regarded as wholly resolved in a manner
Teknekron,Inc.
-------
-13-
which would definitely eliminate this mechanism as a possible
significant cause of high rural ozone levels in all cases.
Meteorological transport of anthropogenic ozone and precursors
to rural areas, whether by relatively local winds or by exten-
sive air mass movements, appears to be not only a plausible ex-
planation, but one which is consistent with a considerable body
of observed data. Assuming that thisis the major mechanism,
it would appear that HC control strategies which effectively
reduced ozone formation in urban areas and in the vicinity of
TrvvoTved'"stationary emission sources would obviously also re-
-_
jch. They would
duQgthe quantities of ozone transport
not, however, necessarily prevent the local rural formation
of 03 brought about by the interaction of transported
oxides of nitrogen with naturally occurring local HC.
From the above, it follows that effective NO controls on a
A
regional basis may be no less important as an 0^ control
strategy than is EPA's proposed HC control program.
The NAAQS for HC (3 hour annual maximum of 0.24 ppm) may be
/
high in terms of the implications for ozone formation, since
this corresponds to a potential 03 level of 0.96 ppm.
Even assuming that NAAQS for. ozone precursors were achieved
and maintained uniformly throughout the U.S., it may not follow
that NAAQS for 0^ would also be attained uniformly.
Tcknekron,Inc.
-------
-14-
An ozone standard of 0.1 ppm may be more realistic than the
present level of 0.08 ppm. (In this connection, it was noted
that in many instances in which rural levels exceeded standards,
these levels fall between 0.08 and 0.1 ppm).
Teknckron,Inc.
-------
-15-
4.0 -REFERENCES
1. Environmental Protection Agency. Mt. Storm, Hest Virginia-Gorman,
Maryland-Keyser, West Virginia, Air Pollution Abatement Activity,
Air Pollution Control Office Publ. APTD-0656, Research Triangle
Park, N.C., April 1971.
2. Richter, H.G., "Special Ozone and Oxidant Measurements in Vicinity
of Mt. Storm, West Virginia," Research Triangle Institute, Research
Triangle Park, N.C. 27709, October 1970.
3. Johnston, D., "Investigation of High Ozone Concentration in Vicinity
of Garret County, Maryland and Preston County, West Virginia, Final
Report, EPa Contract No. 68-02-0624, Research Triangle Park, N.C.
27707, January 1973.
4. Johnston, D., et. al., "Investigation of Ozone and Ozone Precursor
Concentrations at Nonurban Locations in Eastern United States,
Phase I, Final Report, EPA Contract No. 68-02-1077, Research Tri-
angle Park, N.C. 27709, May 1974.
5. Ripperton, L.A., et.al., "Airborne Ozone Measurement Study,"
Research Triangle Institute, Research Triangle Park, N.C., Presented
at 67th Annual Meeting of APCA, Denver, Colorado, June 9-13, 1974.
6. Lea, D.A., "Vertical Ozone Distribution in the Lower Troposphere
Near an Urban Pollution Cen-ter," J. Appl. Meteorology, 7, 252-267,
1-968.
7. Miller, P.R., et.al., "Oxidant Air Pollution in the Central Valley,
Sierra Nevada Foothills, and Mineral King Valley of California,"
Atmos. Environ., 6, 623-633, 1972.
8. Maga, J., Personal Communication to L.A. Ripperton.
9. Baljet, P.J., "Local Air Pollution Episode," Metropolitan Dade County
Pollution control, Miami, Fla., June 1972.
10. Nagler, Lewis, EPA Region IV, Personal Communication to E.L. Martinez.
11. Stasiuk, W.N., and Coffey, Peter, E., "Rural and Urban Ozone Relation-
ships in New York State," Journal of APCA, Vol. 24, No. 6, June 1974.
12. Wisconsin Department of Natural Resources, "A Report to the Board
DNR Concerning High Oxidant Levels in Wisconsin, Summer 1973,"
February 1974.
Teknekron,Inc.
-------
-16-
13. -Rasmussen, R.A., "What do the Hydrocarbons from Trees Contribute to
Air Pollution," J. Air Pollution Control Association, 22_ (7): 537-543,
14. Rasmussen, K.H., et. al., "Sources and Natural Removal Processes for
Some Atmospheric Pollutants," EPA-650/4-74-032, June 1974.
15. Preliminary Report, "Investigation of Rural Oxidant Levels as Related
to Urban Hydrocarbon Control Strategies," EPA - BOA Contract No. 68-
02-1386, December 1974.
16. Rasmussen, R.A., "Background Hydrocarbon Air Pollution Measurements
(5-6 Compounds)," Presented at American Chem. Soc., S.E. Regional
Meeting, October 1974.
Teknekron,Inc.
-------
Teknekron, Inc. L
4701 SANGAMORE ROAD
WASHINGTON, D.C. 20016 (202) 320-5600
January 18, 1977
Mr. E. Lillis
Manager; Air Management
Technology Branch
Office of Air Quality Planning
and Standards
Research Triangle Park, N.C. 27711
Dear Mr. Lillis:
I enjoyed meeting you last week, and appreciate having had the opportunity
to discuss your program. During the conversation, I mentioned our oxidant
study and you requested a copy of the report. This work was performed
about two years ago for EPA/OPE on a very limited budget. Should you have
any questions regarding the report, please feel free to contact me.
As we discussed, we are currently involved in a number of studies for EPA
in Washington, and are looking forward to future involvement in OAQPS pro-
grams.
Sincerely,
Sanford Cohen
Manager,
Washington Office
SC:mbs
Enclosure
WASHINGTON, D.C.
BERKELEY, CALIFORNIA
------- |