EFFECTIVENESS OF ORGANIC EMISSION CONTROL PROGRAMS
AS A FUNCTION OF GEOGRAPHIC LOCATION
April 1977
Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
-------
Contents
Abstract
Contents
1.1 Introduction
2.0 Rationale for Differentiating Needed Controls of Organic Emissions
in Urban and Rural Areas
2.1 Theoretical
2.2 Experimental
2.3 Ambient Data
3.0 Differentiating Urban and Rural Areas with Regard to Control
Effectiveness and Emission Offsets for Sources of Organic Emissions
3.1 Relating Levels of NO to Population
A
3.2 Determining Areas Where Control of Existing Sources Will Re
Most Effective
3.3 Multi-day Stagnation Episodes
3.4 Emission Offset Requirements for New Sources
4.0 References
-------
Abstract
Photochemical oxidants result from a complex interaction amonq
organic and NO precursors and prevailing meteorological conditions.
/\
This paper reviews recent information in an attempt to see how control
of one of these factors (organic precursors) might be most effectively
applied to diminish the oxidant problem. Theoretical and exoerimental
evidence suggest ozone forming potential is most sensitive to organic
precursors at low non-methane hydrocarbon to NO (NMHC/NO ) ratios.
r\ X
Ambient data indicate low ratios primarily occur in urban or suburban
areas. Ratios are high in rural areas as a result of low levels of NO .
Therefore, control of organic emissions is likely to be most effective
ds a means of reducing high ambient levels of oxidant which are formed
in urban areas. Control of NO emissions may ultimately be needed to
X
reduce the amount of oxidant synthesized in rural areas, although the
magnitude of the rural synthesis problem is much less than the oxidant
generated in urban areas. Ambient N0~ and population data are used to
define those areas where organic emission control programs would be
most effective.
The Clean Air Act requires stringent limitations on new sources
that would exacerbate existing violations of a national ambient air
quality standard (NAAQS). Two types of regions are identified for the
purpose of establishing emission offset requirements for new sources of
organics affecting areas that exceed the NAAQS 'or oxidant. The most
stringent requirements are necessary for new sources locating in the
* See EPA "emission offset" policy published in the Federal Register of
December 21, 1976.
-------
2
broad vicinity of urbanized areas which are not in compliance with the
NAAQS for oxidant. Somewhat less restrictive requirements are needed
for sources which are far from major urbanized areas.
1.0 Introduction
Numerous ambient ozone data collected in recent years provide amole
indication that widespread violations of the primary National Ambient
Air Quality Standard (NAAQS) for photochemical oxidants occur in urban
and rural areas alike. There are three key elements in the formation
of ozone:- non-methane hydrocarbon (NMHC) precursors, oxides of nitrogen
(NO ) and appropriate meteorological conditions. A number of statistical
A
studies have shown ambient levels of ozone to be most sensitive to
meteorological parameters (e.g., temperature). ~ Therefore, the ensuing
discussion of relationships among NMHC and NO precursors presupposes
A
meteorological conditions which are favorable for accumulation of high
levels of ozone. The purpose of this paper is to look at the effectiveness
for ozone reduction of controlling volatile organic compound (VOC)
2/
emissions— in different areas.
1. "Ozone" will generally be used in this paper rather than the more
general term oxidant, since ozone is by far the largest component of
photochemical oxidant and most monitoring is specifically for ozone.
2. Due to the ambient measurement technique the term non-methane
hydrocarbons (NMHC) will generally be used when referring to air quality
levels. When referring to emissions, the term volatile organic compounds
will be used, since the other classes of organics in addition to hydro-
carbons play key roles in the synthesis of ozone.
-------
3
Control strategies to reduce ambient levels of ozone have tradi-
tionally focused on the control of VOC emissions. This is still the
most realistic strategy for the reduction of ozone levels. The impact
of such strategies will vary, however, according to prevailing levels of
NO and meteorological conditions. Theoretical, experimental and
A
ambient evidence are examined to determine where VOC emission control
programs are likely to be most effective. The information which is
derived can then be used in two ways. First, it can be used to estab-
lish priorities for controlling existing sources of VOC located in dif-
ferent geographical areas. The second application concerns the review
of new sources of VOC emissions. Since the Clean Air Act requires that
i
new sources not be allowed to exacerbate existing violations of an air
quality standard, EPA has developed an emission offset policy for new
o
sources affecting non-attainment. The information derived in this
paper will be used to show that it is reasonable for these offset require-
ments to be applied differently in rural areas than in urban areas.
2.0 Rationale for Differentiating Needed Controls of VOC Emissions
In Urban and Rural Areas
2.1 Theoretical
The basic mechanism leading to the buildup of ozone is described by
g
ions (1) through (4).
NO + 03 > N02 + 0? (1)
-------
4
N02 hv> NO + 0 (2)
0 + 02 + M > 03 + M (3)
NO • RO- i N0? (4)
Where "M" in Equation (3) represents an energy accepting third body
(e.g., a molecule, particle, etc.) and ROA represents an organic peroxy
radical. Organic peroxy radicals can be easily formed by the oxidation
of organic compounds such as hydrocarbons (organic molecules containing
only carbon and hydrogen) and compounds containing carbon and hydrogen
plus additional elements such as oxygen. Equations (5) and (6) are
several different means by which peroxy radicals can be formed.
02 .
RH + OH' > R02 + RCHO (5)
RCHO + OH" * RC002 + H20 (6)
Where RH is an organic compound
RCHO is an aldehyde
OH' is an hydroxyl radical
Hydroxyl radicals can be formed naturally by the reaction of
atomic oxygen in an excited state with water vapor. Additional hydroxyl
radicals are formed as shown in Equations (7) and (8).
RO" + 02 -* H02 + RCHO (7)
H02 + NO > N02 + OH' (8)
Where HOA is a hydroperoxy radical
-------
5
Thus, it can be seen that the role of urban organic emissions is to
provide a large, concentrated source of peroxy radicals (Equations (5),
(6) and (7)). This process is enhanced somewhat by the presence of NO
(Equations (4), (7) and (8)). The role of oeroxy radicals is to provide
another means (rather than ozone) for oxidizing NO and N02 (Equation
(4)). While both ozone and the peroxy radicals are likely to oxidize
the NO, the role of the organic radicals (and emissions) in the synthesis
of ozone can best be pictured by considering only Equations (2) through
(4) as shown in Figure l(a). At low NMHC/NO ratios, the rate by which
NO is converted to N02 (Equation (4)) is influenced by the availability
of organic compounds. Therefore, the effects of controlling orqanic
compounds are to slow the conversion of NO to N02, thereby lowering the
NOp/NO ratio, and to insure that a larger portion of the NO which is
converted to N0« is done through the destruction of ozone (Equation
(1)). If the oxidation of NO by organics (Equation (4)) is delayed
sufficiently so that the sun has passed its zenith before significant
amounts of N02 are created, photodissociation of NO- (Equation (2)) will
be reduced, and lebs ozone will be allowed to accumulate (Equation (3)).
At moderately high NMHC/NO ratios, the greater availability of
X
organic radicals means that all of these radicals are not consumed as
rapidly in reactions with NO (Equation (4)). As a result, more reactions
between the radicals and N02 are able to occur as shown in Equation (9).
R02 + N02 -»• Organic Nitrates (9)
-------
Figure 1: Schematic Relationships Among Oxidant and Precursors Under
Low, Moderate, and High NMJIC/NOX Ratios
(a) Low NMHC/NOX Ratios
fl.O)
(Higher concentrations realize
as a result of higher N02/N0
ratios)
ed
(b) Moderately High NMHC/NOx Ratios
Eq.(9)
ROj
( Serves as a sink for NO )
/\
Eq.(4)
Eq.(l)
NO
M
M
(Ozone concentration limited
by available NO )
A
RNO.
Eq.(9)
ROJ +
(c) Very High NMHC/NOV Ratios
(Limits available NO
M
Eq.(lO)
+ RH—*• Oxidation
Products
M
(Ozone limited by avail
able NO , and scaveng-
ing by organic
species)
-------
7
Reactions such as Equation (9) remove N02 from the system. Thus,
the amount of ozone formed and accumulated as a result of Equations (2),
(3) and (4) begins to become limited by the availability of NO , and
A
becomes less sensitive to additional organic precursors. The situation
at moderately high NMHC/NO ratios, in which reactions between radicals
and N0? begin to limit the sensitivity of ozone to additional organic
precursors, is depicted in Figure l(b).
At very high NMHC/NOX ratios, it is possible that the excess
organic precursors can react to such an extent with ozone that the
addition of still further organic precursors has little effect, or may
even result in slightly lower levels of ozone.
RH + 0^ > More highly oxidized compounds (10)
The case of very high NMHC/NO ratios is depicted schematically in
Figure l(c).
The point of the rather involved preceding discussion is that the
present theoretical understanding of smog formation indicates that the
sensitivity of ozone-forming potential to changes in NMHC precursor
levels decreases as the NMHC/NOX ratio increases.
2.2 Experimental
The definition of "low", "moderately high" and "very high" NMHC/NOX
ratios described in the previous section, as well as the exact functional
relationship among organic and NO precursors and ozone, depend on a
number of factors. Some of the more important factors include reactivity
of the organic precursor mix, lighting intensity and temperature. The
-------
8
sensitivity of oxidant to precursor levels can best be demonstrated with
smog chamber experiments in which data points are obtained under similar
or identical temperatures, lighting conditions, organic reactivities and
dilution rates. In such smog chamber experiments, specified levels of
organic and NO precursors are injected into a large transparent chamber.
X
This mix is then irradiated and the maximum amount of ozone formed
within several hours is noted. If this experimental procedure is
repeated a sufficient number of times, a set of ozone isopleths can be
derived which express maximum 0~ concentration as a function of initial
organic and NO precursor concentrations. Figures 2 through 5 depict
/\
ozone isopleths obtained from different smog chamber experiments.
Figure 6 was derived from a chemical kinetic model in which artifacts
introduced by chamber wall effects have been deleted. '
The conditions accompanying each set of smog chamber experiments
have been noted in Figures 2 through 6. These figures depict a wide
range of experimental conditions. For example, reactivity of the organic
mixtures vary from that associated with auto exhaust (Figures 2 and 6)
to an urban synthetic mixture of compounds (Figures 3 and 4) to a
single relatively unreactive paraffin (Figure 5). Some experiments were
conducted under constant lighting conditions (Figure 2 and 3) while
others (Figures 4 through 6) incorporated diurnally varying lighting
intensity. Experiments were conducted within a range of maximum tem-
peratures (70 to 100°F) and in chambers of varying size and composition.
-------
9
Despite the diverse sets of conditions under which the 0^ curves in
Figures 2 through 6 were obtained, each set of isopleths has several
attributes in common. For example, the shape of the curves is fundamen-
tally similar. It can be seen that as NMHC/NO ratios increase, the
A
maximum ozone concentration formed several hours later becomes less and
less sensitive to additional increases in initial concentrations of
organic precursors. The lines emanating from the origin in Figures 2
through 6 represent constant NMHC/NO ratios. Under many of the experi-
mental conditions utilized, the ozone-forming potential within several
hours becomes relatively insensitive to changes in organic precursor
concentrations at NMHC/NO ratios exceeding 20:1. Under all conditions
/\
tested, ozone-forming potential within one solar day does not appear to
be sensitive to organic precursors at NMHC/NO ratios exceeding 30:1.
J\
Thus, prevailing ratios of 30:1 or better might well be used as a rule-
of-thumb for establishing whether a strategy of reducing organic emissions
is likely to be effective under most meteorological conditions.
2.3 Ambient Data
In Sections 2.1 and 2.2, it has been concluded that the bulk of
theoretical and experimental evidence suggests that under most meteoro-
logical conditions, ozone-forming potential is likely to be relatively
insensitive to organic precursors at NMHC/NO ratios exceeding 30:1. In
this section, ambient air quality data are reviewed to determine where
such high ratios are most likely to prevail.
-------
Figure 2: Ozone As A Function of Initial Precursors -- Bureau of Mines Experiments
10
0.6-
0.5-
0.4-
1.0.3.
a.
0.2.
0.1.
°3 =
.08
.15 .20
30
r
i.o
.4J3 ppm
Experimental Conditions
- Constant Light Intensity
- Automotive Exhaust
- Temp, approx. 92s F
- Irradiation Time : 6 hours
- 100 ft netal & glass chamber
- Static Experiments
Extrapolated Data
NMHC, ppmC
2.0
3.0
-------
12
Figure 3: Ozone versus Initial Precursor Levels -- University of California (Riverside)
.20
Experimental Conditions
- Constant Light Intensity
- 6-hour Irradiation
- Synthetic Urban Mix
- Glass Chamber (5800 liters)
- Temp. 24 to 32°C
- Static Experiments
- - - Extrapolated Data
.30
.40PPm
0:1
1.0 NMHC, ppmC
-------
0.61
0.5"
0.4.
0.3.
0.2.
0.1-
Figure 4: Ozone versus Initial Precursor Levels -- UNC Outdoor Chamber
13
Experimental Conditions
- Diurnal Light Variat-: . (Suiry ".ay
- Synthetic Urban Mix
- Static Exps.
- Maximum Temp approx 110°F
- 148m3 Teflon Chamber
Extrapolated Data
£ *&.'•&&
3.0
NMHC, ppmC
-------
Figure 5: Ozone versus Initial Precursor Levels — "Aged" Precursor Mix
14
10:
(fl CL2 0*3
"oT4(£5oTeoT?Is7.9 1.
NMHC, ppmC
Experimental Conditions
_ Outdoor Diurnai Light Variation
- Isopentane + N02
- Static Experiments
- Max Terno 73 to 92°F
- April to May Experiments
- 100 to 150 liters teflon bags
30:1
-------
Figure 6: Ozone versus Initial Precursor Levels — Modeled Results
16
.24J
.22
.20-
.18-
.16-
.14 -
.12
.10 -
.08 -
.06 -
.04 -
.02
32 .34 .36 ppm
Experimental Conditions
- Diurnal Light Intensity (0700-1600)
- Propylene + n-Butane Combination
of Equivalent Reactivity to Auto
Exhaust
- Dilution = 3%/hour
- Temperature approx. 80°F
0, = .08 .12 .16 .20
NMHC, ppmC
-------
15
Table 1 presents available NMHC and NO data from urban, suburban
and rural sites. As Column (5) In Table 1 shows, there is a tendency
for NMHC/NOX ratios to increse as one moves from urban to suburban to
rural areas. Looking at Column (4) in Table 1, it appears as though
high ratios in rural areas are primarily due to low ambient levels of
N0y. With the exception of the Lawrenceville site, which is located
J\
near a source of NO , measured levels of NO in rural areas are near the
A n
detection limit of the monitoring instrument (.005 ppm = 10 yg/m ).
Table 2 provides further indication of exceedingly low mean levels of
2
NO which appear to exist in rural areas. The low mean values in Table
A
2 reflect high incidence of undetectably low concentrations and a few
concentrations above the detectable limit. Table 3 reveals just how
infrequent detectable concentrations of NO were at the rural sites
A
presented in Table 2. Non-methane hydrocarbon concentrations were not
measured at the four sites in Tables 2 and 3. However, individual
species of light hydrocarbons were measured. As shown in Table 4, the
sum of these species generally totaled about 0.2 ppmC. The information
in Tables 2 and 4 is combined in Table 5 to provide a rough estimate of
NMHC/NO ratios at four additional rural sites. The information in
A
Tables 2 through 5 can be used to qualitively support the conclusions
drawn from Table 1 that:
(1) NMHC/NO ratios tend to be high (greater than 30:1) in rural
areas and low (less than 30:1) in urban and suburban areas,
and,
(2) The principal reason for high NMHC/NO ratios in rural areas
is the extremely low prevailing concentrations of NO.
-------
TABLE 1: Ambient NMHC, NOX and NMHC/NOX Levels at Selected Monitoring Sites
(1)
Location
Austin, Tx.
Corpus Christi, Tx.
Dallas, Tx.
Los Angeles, Ca.
Canton, Oh.
San Antonio, Tx.
El Paso, Tx.
St. Louis, Mo.
Houston, Tx.
Phoenix, Az.
Azusa, Ca.
Canton, Oh.
Aldine, Tx.
Clute, Tx.
El Paso. Tx.
Texas City, Tx.
Nederland, Tx.
West Orange, Tx.
Groton, Ct.
Portland, Me.
Lawrenceville, 111.
Pownal , Me.
Simsburgy, Ct.
McConnelsville, Oh.
Wilmington, Oh.
Wooster, Oh.
(2)
Type of Site
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Urban
Suburban
Suburban
Suburban
Suburban
Suburban
Suburban
Suburban
Suburban
Suburban
Suburban
Rural
Rural
Rural
Rural
Rural
Rural
(3)
Distance to Nearest Population
Center > 50, 000, Miles
20 mi ENE of LA CBD
2 mi N of Downtown
12 mi NNE of Houston
40 mi S of Houston
5 mi ESE of Downtown
10 mi NU of Galveston
7 mi NW of Port Arthur
20 mi E of Beaumont
2 mi S of New London-Groton ,
40 mi ENE of New Haven
2 mi NU of Downtown
t 50 mi N of Evansville, Ind.
2 mi ENE of Large Refinery
i- 17 mi NE of Portland, Me.
10 mi NU of Hartford, Ct.
63 mi SE of Columbus, Oh.
27 mi E of Dayton, 40 mi NE
of Cincinnati, Oh.
30 mi U of Akron and Canton, Oh.
(4)
6-9a.m. Mean
NMHC/NO
ppmC/ppffl
.28/.025
.65/. 035
.137. 038
1.47/.147
.99/.090
.21/. 022
.56/. 064
1.14/.135
2.84/.110
1.23/.146
.65/.073
.53/.050
.53/. 025
.487. 027
.91/.017
.667. 034
.35/.023
.247. 029
.32/.034
.22/.024
.42/.005
.44/.006
.30/.006
.12/.007
.337. 008
(5)
6-9a.m. Median
NMHC/NOX Ratio
10.6:1
18.5:1
5.0:1
9.8:1
9.0:1
11.2:1
9.2:1
8.0:1 A
7.7:1
21.9:
8.5:
9.4:
10.0:
22.5:
11.5:
40.0:
21.7:
18.0:
9.5:1
11.4:1 +
8.6:1
95.0:
47.0:
54.0: *
17.5: *
40.0: *
(6)
Sampling Period
(Days)
7/1-9/30/75 (43 days)
7/1-9/30/75 (44 days)
7/1-9/30/75 (44 days)
7/1-9/30/75
7/1-7/25/74
58 days)
14 days)
7/1-9/30/75 (37 days)
7/1-9/30/75 (42 days)
July-Oct'75/June-Oct'76
7/1-9/30/75 (43 days)
9/24-10/1/73
7/1-9/30/75
7/1-7/24/74
7/1-9/30/75
7/1-9/30/75
7/1-9/30/75
7/1-9/30/75
7/1-9/30/75
7/1-3/30/75
7/15-8/21/75
7/29-8/27/75
6/13-6/23/74
7/29-8/27/75
7/15-8/22/75
6/14-8/31/74
6/14-8/31/74
6/14-8/31/74
(7 days)
86 days)
17 days)
21 days)
49 days)
30 days)
23 days)
36 days)
44 days)
(15 days^
(5 days)
(10 days)
(22 days)
(14 days!
* Mean rather than median ratios
+ On days when the site was downwind from Portland. Uhen the site was upwind, the median ratio was 31:1.
A Mean median value for 2 summer's data at 6 sites in the City of St. Louis
-------
TABLE 2: Statistical Summary of Hourly Oxides of Nitrogen
Concentration Measurements - Rural Stations
(June 27 - September 30, 1975)
Station
Bradford, Pennsylvania
Creston, Iowa
Wolf Point, Montana
DeRidder, Louisiana *
Mean Hourly Mean Hourly
Concentration Concentration Case Count
(yg/m ) (ppm )
NO N02 NO N02 NOX NO N02
2.4 5.1 .002 .003 .005 2265 2259
4.7 4.3 .004 .002 .006 2162 2162
<1.0 1.5 <.001 .001 .001 1318 2136
1.9 4.9 .002 .002 .004 2444 2444
* June 27 - October 31, 1975.
-------
TABLE 3: Cumulative Frequency Distribution of Hourly Concentrations of Oxides of Nitrogen -
oo
Rural Stations (June 27 - September 1975)
Concentration
(wg/m )
0
10
20
30
40
50
60
Percent of
Bradford ,
NO
100.0
50.6
4.8
0.5
0.0
0.0
0.0
0.0
Pa.
NOo
100.0
73.9
15.1
2.1
0.5
0.1
.0.0
0.0
Hourly Averages Greater
Creston ,
NO
100.0
75.0
6.1
3.0
0.0
0.0
0.0
0.0
la.
NOo
100.0
91.8
1.6
0.2
0.0
0.0
0.0
0.0
Than Stated Concentration
Wolf Point,
NO
100.0
9.7
0.1
0.1
0.0
0.0
0.0
0.0
Mt.
NOo
100.0
33.2
3.9
0.4
0.0
0.0
0.0
0.0
DeRidder,
NO
100.0
19.2
9.1
0.4
0.1
0.1
0.0
0.0
, **
La.
NOo
100.0
48.0
13.1
4.8
1.0
0.3
0.2
0.0
** June 27 - October 31, 1975.
Detection Limit = 10 yg/m
-------
TABLE 4: Statistical Summary of Selected Hydrocarbon Analyses at
Rural Sites During Summer
1975 *
Compound
Ethane & Ethyl ene
Propane
Propylene
Acetyl ene
n-Butane
1 -Butane
Isobutane
Isopentane
Cyclopentane
n-Pentane
Toluene
o-Xylene
TOTAL
Of Above
Wolf Point, Mt.
.051
.010
.004
.005
.008
0
.003
.009
.001
.010
.062
.063
.23
Creston, la.
.086
.010
.005
.007
.005
.002
.002
.005
.001
.016
.041
.013
.19
Bradford, Pa.
.056
.012
.004
.007
.018
.001
.009
.013
.001
.014
.037
.015
.19
DeRidder, La.
.057
.022
.006
.006
.012
.001
.003
.003
.001
.015
.056
.009
.20
* Concentrations expressed In ppmC
-------
TABLE 5; Ratio of the Mean Sum of Measured Hydrocarbon Species to Mean NO
Levels at Several Additional Rural Sites
Station Ratio of Mean Concentrations
Bradford, Pa. > 38:1
Creston, la. > 32:1
Wolf Point, Mt. > 230:1
DeRidder, La. > 50:1
ro
O
-------
21
3.0 Differentiating Urban and Rural Areas with Regard to Control
Effectiveness and Emission Offsets for Sources of Organic Emissions
3.1 Relating Levels of NOX to Population
The information in the preceding sections has shown that high
NMHC/NO ratios are rare in urban areas and are the rule in rural areas.
A
Further, under most meteorological conditions ozone-forming potential is
not likely to be very sensitive to changes in ambient levels of organic
precursors if the NMHC/NO ratio is high. Unfortunately, ambient organic
3\
concentrations are not monitored routinely, and there are relatively few
rural NMHC data. However, from the previous section, it would appear
that ambient levels of NO measured in rural areas might serve as a
A
suitable substitute for the NMHC/NOy ratios. If the ambient N0y
n A
levels were typically near the detectable range of available instrumenta-
tion, the data presented in Tables 1 and 4 suggest that the NMHC/NO
A
ratios would be high.
Since available data from rural sites (other than those already
presented) are measured as N02> these data were used as a surrogate for
NMHC/NO¥ ratio rather than NOV data.17 Nitrogen dioxide data collected
rt /\
during the third quarters of 1973-1975 at 86 locations classified as
"rural" or "remote" in EPA's data bank (SAROAD data) were examined.
Almost all of the ambient N02 data available in rural areas are obtained
through the use of 24-hour gas bubblers which are run about once every 6
o
days. Instrument sensitivity is about 10 wg/m (.005 ppm). A typical
sample size was about 15 readings per site per quarter. Only third
quarter data were considered, because meteorological conditions are
generally most conducive to ozone formation during this period.
-------
22
Figures 7 and 8 are log-log scatter diagrams in which county wide
NO and HC emission density (obtained from EPA's National Emission Data
A
1ft
System) are plotted against county population density for all counties
in the United States. According to the information in these Figures,
population may serve as a good surrogate for NO emissions in many
A
cases. Since NO has been observed to reach very low levels fairly
A
2
rapidly, if the NO emission density is relatively low, low levels of
A
N0? should predominate. However, in areas where total NO emissions are
t A
large and relatively concentrated (i.e., urban areas) NO is being
A
continually replenished. In addition, the removal of NOX by chemical
reactions should be much slower (i.e., this is the low NMHC/NO ratio
case discussed in Section 2.1). Thus, proximity to large and concentrated
"sources" of NOX (e.g., urban areas) should be an indicator of ambient
levels of NO and the sensitivity of ozone-forming potential to changes
A
in ambient levels of organic precursors. Furthermore, NO levels should
A
be high at greater distances from areas having higher NO emissions.
A
In Appendix A, distance between monitoring sites and urbanized
areas of specified size ranges is plotted against mean 24-hour NO?
concentrations observed at each monitoring site. Although there is some
scatter on these graphs (possibly reflecting the Impact of very localized
sources of NOX), a definite inverse relationship between NCL levels and
distances from urban areas can be discerned. Using the information in
Appendix A, it is possible to derive a rough graph of the distance at
which N02 approaches undetectable levels as a function of urbanized
-------
Figure 7.
VE5IU3
3EYSZTY
•
•
•
,
.'CSft'C. •
N
0
V •
• "* ~ 11 * Cl *
t
M •
I •
S . "*Cr- °*"* *
^ *
c
\ •
7 • !"«"•
r •
•' "
*" •
!t
1
V t
.
C
1 .
S
C •
R •
r .
*-
1 -.^Sf0*^ »
I
L 1
•
•
-.i:;ic«c: •
•
• i,
..L
SUSPF = -a* . -
NIJM3ER CF POINTS - 7
-------
Figure 8 . uvr
orisirv VEPSUS POPULATIGI. OEI.STT»
.
•
•
•
•
T
i. ~
T 1
: i i 111
r i 111
11 : i i
: i i
- i :
i i i
i 11 1 . i 11 1
i ^ 1 1 11 11 1 1
. i i 1111
i 11111
i »: 11 1111111
.: _ .. . i. .: i.. i
11 1 * A L 1*.« 1111 11
... ; : l. .11.
1 -- ......til.. 1
ii~l.
Rr .---I
M-i"3r!f or FOIV
-------
25
area size.— Figure 9, which is derived from the data in Appendix A,
depicts such a relationship.
The curve in Figure 9 represents the distance at which the mean 24-
hour NCL concentration becomes virtually undetectable for at least 90%
of the monitoring sites in the vicinity of urbanized areas of various
sizes. An N02 level (.007 ppm) slightly above the actual detectable
limit of the measurement method (.005 ppm) was used in deriving Figure 9
from the data in Appendix A. This reduces the errors caused by the
large uncertainty as to the true readings at the detection limit of the
measurement method and variations in the way data are reported at levels
below the detection limit (e.g., at levels below .005 ppm, readings may
be reported as .005 ppm).
It appears that for smaller urbanized areas, the urban plume is not
large enough to maintain its integrity and high precursor concentrations
for the period of time it takes for the photochemical reaction to cause
measurable increases in ozone levels. The data available to define the
minimum size of an urbanized area necessary to produce significant ozone
concentrations are quite limited, although some preliminary estimates
can be made.
Little ozone gradient was observed upwind and downwind from Portland,
24
Maine (population approximately 100,000) while a detectable gradient
was noted for several cities in Ohio ranging from 250,000 to 600,000
population. Table 6 is a tabulation of several cities ranging in
urbanized area population from 20,000 to nearly 300,000. Each of the
3. As defined by the U.S. Bureau of Census, urbanized areas generally
include core cities with a population greater than 50,000 plus any
closely settled suburban areas.
-------
Figure 9
Distance from Urbanized Areas Where NO
Approaches Undetectable Levels x
4,000,000 r
3,')00,000
3,000,000
o
2,500,000
Q.
O
10
S 2,000,000
1,500,000
1,000,000
500,000
200,000
Above 4,000,000 Use 85 Miles
20
40
60
Estimated Radius (Miles) Where NO
Approaches Undetectable Levels
80
100
26
-------
27
urban areas shown in Table 6 can be considered "Isolated" in that they
all are at least 40 miles from a larger neighboring city. Also shown in
Table 6 is the percent of time the selected cities reported oxidant
levels in excess of .08 ppm, as taken from the 1974 Monitoring Trends
Report. From the table it appears that the frequency of violations
generally becomes much smaller as the size of the urban area decreases.
23
Finally, upwind-downwind measurements at an isolated point source of
both VOC and NO (an oil refinery at Lawrenceville, Illinois) indicated
that there was no significant build-up of ozone as the plume was tracked
downwi nd.
The above studies suggest that for urbanized areas less than about
200,000 or for isolated point sources or clusters of point sources that
have emissions that are less than' an urbanized area of 200,000, there is
uncertainty whether they contribute significantly to local observed
ozone concentrations and whether VOC control programs in these areas
would be effective. This does not mean that some urbanized areas of
less.than 200,000 should not have an oxidant control strategy. In such
areas which are highly industrialized, and have greater emissions oer
capita than the average, oxidant control programs may be effective in
reducing ozone levels. However, on a nationwide basis, indications are
that significantly more benefits would be gained by focusing VOC control
programs on the larger urbanized areas (greater than 200,000). Further
work is planned to better resolve this issue during the summer of 1977.
-------
Table 6
Reported frequency in Which
Ozone Levels Exceed .08 PPM
for Various Size "Isolated" Urban Areas
Population
of
Urbanized
Area
294,000
279,000
268,000
266,000
257,000
255,000
254,000
241 ,000
236,000
223,000
180,000
159,000
149,000
139,000
132,000
132,000
104,000
99,000
60,000
45,000
40,000
20,000
Percent of
1974 Reported
Oxidant Values
Above the
Area Standard
Tucson, Arizona
Charlotte, North Carolina
Newport News /Hampton, Virginia
Davenport, Illinois
Mobile, Alabama
Des Moines, Iowa
Austin, Texas
Columbia, South Carolina
Las Vegas, Nevada
Chattanooga, Tennessee
Utica, New York
Lexington, Kentucky
Ogden, Utah
Eugene, Oregon
Topeka, Kansas
Cedar Rapids, Iowa
Provo, Utah
Reno, Nevada
Kingsport/ Johnson City,
Tennessee
Elmira, New York
Santa Maria, California
Carson City, Nevada
2
1
2
0
1
0.1
2
0.6
5
5
0.8
0
0.2
0.2
0
0.2
0
0.1
0.8
0.1
0.1
0.1
28
-------
29
3.2 Determining Areas Where Control of Existing Sources Will Be
Most Effective
Using the curve in Figure 9 and the location of each urbanized area
10
exceeding 200,000 population, the map in Figure 10 was prepared. The
shaded areas in Figure 10 represent geographical locations where it is
most likely that ambient levels of NOX will be high enough so that ozone
forming potential will be sensitive to changes in ambient NMHC levels or
VOC emissions. In addition to NMHC/NOY ratios that are conducive to
A
ozone formation, the shaded areas generally correspond to high absolute
levels of both precursors, further enhancing ozone-forming potential.
These arguments support the position that controlling organic emissions
from existing sources within the shaded areas in Figure 10, will be most
effective in reducing ozone concentrations.
As noted above, the shaded areas in Figure 1 were drawn for all
urbanized areas with a population greater than 200,000 and no specific
determination of non-attainment or that the exsiting SIP is substantially
inadequate was made for each area; therefore, if no SIP revision has
been requested, the existence of a circle in Figure 1 does not mean that
a SIP control strategy must be developed for such areas. Also as noted
below, there is a secondary concern over emissions outside the shaded
areas. Therefore, States should consider the adoption of Statewide
control measures for existing sources, perhaps time-ohasing requirements
into rural areas at a somewhat slower rate than in urban areas.
-------
Figure 10: Estimate of Geographical Areas Where Oxidant Control Strategies
Will Initially Be Most Effective in Reducing High Oxidant Levels.
-------
31
3.3 Multi-day Stagnation Episodes
In the preceding sections, a rationale for determining where VOC
control programs would be most effective was developed. However, there
is concern that under some very adverse meteorological conditions
(e.g., multi-day stagnation episodes with warm temperatures) sources of
organic emissions in unshaded areas would contribute to the overall
oxidant problem. Figures 2 through 6, for example, were derived from
experiments lasting less than one solar day.
To date, there have been only a limited number of multi-day smog
20 21
chamber experiments. ' Interpretation of the results of these experi-
ments is complicated by the very low precursor levels present on the
second (and sometimes) third days of the experiments. Consequently, it
is difficult to determine the extent to which observations are dominated
by artificially-induced chamber effects. Two observations are of
interest. First, ozone synthesis on subsequent days may be NO limited.
Hence, if rural organic emissions are poorly diluted, they may come into
contact with sufficient NO to form ozone on subsequent days. Second,
A
one of the effects of controlling organic precursors in urban areas
20
appears to be increasing the nighttime concentrations of N02- If
this effect occurs in the atmosphere, it may result in increased NO
A
concentrations occurring somewhat further from cities than at present.
Thus, if dilution is poor, sources built outside the circles in Figure
10 could, at sometime in the future, participate in the synthesis of
additional ozone unless NO is reduced.
A
-------
32
3.4 Emission Offset Requirements for New Sources
The situation involving multi-day stagnation episodes raises the
Q
question as to how EPA's emission offset policy should be applied to
major new VOC sources in urban and rural areas in various parts of the
country.
1) Shaded areas in Figure 10.
Based on the preceeding discussion, all sources in a given shaded
area in Figure 10 can be assumed to contribute to the ambient levels
that are conducive to ozone formation in that area. Therefore, a major
new VOC source locating within one of the shaded areas should be required
to obtain emission offsets from existing sources within that same area.
This would hold true even if the areas overlap. Only if the new source
were in the overlapping area of two circles could the offsets be obtained
from either of the two circles.
For administrative simplicity, it may be appropriate to alter the
shaded areas to conform to county or other political boundaries. In
addition, the boundaries of the shaded areas could be altered to account
for local meteorologial or topographical conditions, such as where
topography may make winds from a given direction highly unlikely.
However, since the oxidant standard is expressed in terms of a maximum
one-hour concentration, which in most locations could occur with any
wind direction, it is generally not appropriate to alter the area bound-
aries based on the shape of the wind rose.
-------
33
As noted above, Figure 10 was drawn for all urbanized areas with a
population greater than 200,000; no specific determination of non-
attainment was made for each of the areas in Figure 10. Documentation
of non-attainment should be done to confirm the need for control of VOC
emissions as well as emission offsets in each area. Lack of air quality
data should not be interpreted as attainment, but should reinforce the
need to obtain the necessary air quality data to document the attainment
status.
2) Unshaded areas in Figure 10.
1 2
Field studies ' have indicated that violations of the oxidant
standard can be widespread during June to October when atmospheric
stagnation periods of several days occur. As noted above, under such
meteorological conditions, it may be possible for precursor levels to
build up to a degree so that sources of organic emissions which may
normally not be significant contributors to the oxidant problem may
become so. Meteorological regimes which result in poor dispersion and
multi-day irradiation of oxidant precursors occur in a number of other
22
areas of the country. However, although there is concern over new
sources locating outside the shaded areas of Figure 10, the data at this
time do not appear to support the very rigorous requirements of EPA's
emission offset policy. A more appropriate requirement at this time
would be to minimize new source emissions as much as possible through
the use of best available control technology (BACT).
-------
34
4.0 References
1. Research Triangle Institute; "Investigation of Rural Oxidant Levels
as Related to Urban Hydrocarbon Control Strategies"; EPA-450/3-75-036;
March 1975.
2. Decker, C. E., et. al.; "Formation and Transport of Oxidants Along
Gulf Coast and in Northern United States"; EPA-450/3-76-033; August 1976.
3. Monitoring and Data Analysis Division, Monitoring and Reports
Branch; "Monitoring and Air Quality Trends Report, 1974"; EPA-450/1-76-001;
February 1976.
4. Wolff, 6. T., et. al., "An Investigation of Long Range Transport of
Ozone Across the Midwestern and Eastern United States"; Proceedings of
the International Conference of Photochemical Oxidant Pollution and Its
Control; Raleigh. N.C.; September 1976.
5. Hartwell, T. D., Hamilton, H. L.; "Examination of the Relationships
Between Atmospheric Oxidant and Various Pollutant and Meteorological
Variables"; EPA Contract 68-02-1096; December 1975; William Hunt, EPA
Project Officer.
6. Meyer, E. L., Jr., Freas, W., Summerhays, J. E., and Youngblood, P. L.;
"The Use of Trajectory Analysis for Determining Empirical Relationships
Between Ambient Ozone Levels and Meteorological and Emissions Variables";
Proceedings of the International Conference on Photochemical Oxidant Pollu-
tion and Its Control; Raleigh. N.C.; September 1976.
7. Ludwig, F. K., Johnson, W. B., Ruff, R. E., and Singh, H. B.; "Impor-
tant Factors Affecting Rural Ozone Concentrations"; Proceedings of the
International Conference on Photochemical Oxidant Pollution and Its
Control; Raleigh,
N.C.; September 1976.
8. Federal Register. 40 C.F.R. 51; Volume 41, Number 246; "Requirements
for Preparation, Adoption, and Submittal of Implementation Plans;" pages
55524 to 55530, and "Review of New Sources and Modifications;" pages
55558 to 55560; December 21, 1976.
9. United States Environmental Protection Agency, Office of Air Quality
Planning and Standards; "Control of Photochemical Oxidants - Technical
Basis and Implication of Recent Findings"; EPA-450/2-75-005; July 1975.
10. Dimitriades, B.; "Effects of Hydrocarbon and Nitrogen Oxides on
Photochemical Smog Formation"; Environmental Science and Technology,
Volume 6, page 253; March 1972.
-------
35
11. Dimitriades, B.; "Oxidant Control Strategies, Part 1, An Urban
Oxidant Control Strategy Derived from Existing Smog Chamber Data";
to be published in Environmental Science and Technology.
12. Pitts, J. N., Or; "California Experience in the Control of Photo-
chemical Smog"; Lecture presented at Smog 1976--Occurrence and Control
of Photochemical Pollution Symposium; Macquarie University, Sydney,
Australia; February 1976.
13. Jeffries, H. E., Kamens, R., Fox, D.; "Outdoor Simulation of Air
Pollution Control Strategies"; Final Progress Report EPA Grant Number
800914-04; September 1976.
14. Sickles, J. E.; "Ozone Precursor Relationships of Nitrogen Dixoide,
Isopentane and Sunlight Under Selected Conditions"; Doctor of Philosophy
Dissertation, University of North Carolina; 1976.
15. Whitten, G. Z., and Hogo, H.; "Mathematical Modeling of Simulated
Photochemical Smog"; EPA 600/3-77-001; January 1977.
16. Dodge, M. C.; "Combined Use of Modeling Techniques and Smog Chamber
Data to Derive Ozone-Precursor Relationships"; Proceedings of the
International Conference on Photochemical Oxidant Pollution and Its
Control; Raleigh, North Carolina; September 1976.
17. United States Environmental Protection Agency, Office of Air Quality
Planning and Standards, National Air Data Branch; SAROAD data.
18. United States Environmental Protection Agency, Office of Air Quality
Planning and Standards, National Air Data Branch; National Emissions
Data System.
19. United States Census Bureau; "1970 Census Users' Guide"; Part 1;
United States Government Printing Office; Washington, D.C.; 1970.
20. Jeffries, H., Fox, D., and Kamens, R.; "Outdoor Smog Chamber Studies
Effect of Hydrocarbon Reduction Nitrogen Dioxide"; EPA 650/3-75-011;
June 1975.
21. Sickles, J. E., Ripperton, L. A., and Eaton, W. C.; "Oxidant and
Precursor Transport Simulation in the Research Triangle Institute Smog
Chambers"; Proceedings of the International Conference on Photochemical
Oxidant Pollution and Its Control; EPA 600/3-77-001 a; January 1977.
22. Holzworth, G. C., "Mixing Heights, Wind Speeds and Potential
for Urban Air Pollution Throughout the Continuous United States,
AP-101, U.S. Environmental Protection Agency, January 1972.
-------
36
23. Westburg, H. H., and Rasmussen, R. A.; "Measurement of Light
Hydrocarbon In the Field and Studies of Transport of Oxldant Beyond
an Urban Area"; Technical Progress Reports, Contract 68-02-1232; J.
Bufalini EPA Project Officer; October 1973, July to August 1974,
September to December 1974.
24. Londergen, R. J., and Polgar, L. G.; "Measurement Program for
Ambient 03, NO and NMHC at Portland, Maine—Summer 1974"; prepared
pursuant to EPA Contract 68-02-1302; November 1975.
-------
Appendix A
Data for Relating NCk Concentration to
Distance from Various Size Urban Areas
-------
Figure A-l: N(L Concentration versus Distance from Urban Areas
0)
o
t_3
60
50
3( -
20
10
URBAN AREAS WITH POPULATIONS
50.000 - 100,000
(N02 Concentration - .0075 ppm)
20
40
60
80 100 120 140
DISTANCE (MILES)
160
180
200
-------
Figure A-2: N02 Concentration versus Distance from Urban Area
60
r* •
50
URBAN AREAS WITH POPULATIONS
100,000 - 500,000
e
CD
40
0)
u
o
0
CM
o
(NO, Concentration = .0075 ppm)
CO
10
20
40
60
80 100 120 140
DISTANCE (MILES)
160
180
200
-------
Figure A-3: N02 Concentration versus Distance from Urban Area
6C I—
URBAN AREAS WITH POPULATIONS
500,000 - 1,000,000
s 30
o
IB
I
-------
Figure A-4: N02 Concentration versus Distance from Urban Area
50
URBAN AREAS WITH POPULATIONS
GREATER THAN 1,000,000
tv,
c
40
Oi.1
20
Concentration = .0075 ppm)
\
20
40
60
80 100
DISTANCE (MILES)
120
140
160
180
200
-------
Figure A-5: NOo Concentration versus Distance from Urban Area
60
URBAN AREAS WITH POPULATIONS
250.000 - 750,000
o>
50
C
o
'J
20
(NO- Concentration = .0075 ppm)
e\j
10
I
20
40
60
80 100 120
DISTANCE (MILES)
140
160
180
200
-------
Figure A-6: NCL Concentration versus Distance from Urban Area
60 r-
50
40
URBAN AREAS WITH POPULATIONS
750,000 - 1,750,000
30
CO
o
IB
-t->
c
QJ
u
o
o
C\J
o
20
10
(NO- Concentration = .0075 ppm)
I
20
40
60
80 100 120
DISTANCE (MILES)
140
160
180
200
-------
Figure A-7:
Concentration versus Distance from Urban Area
URBAN AREAS WITH POPULATIONS
GREATER THAN 1,750,000
0)
u
20
(NO- Concentration = .0075 ppm)
CM
O
10
20
40
60
80 100
DISTANCE (MILES)
120
140
160
180
200
------- |