The
previous 1-hour primary and secondary federal standards for ozone
were each set at a level of 0.12 ppm. The standards were attained
when the expected number of days per calendar year with maximum
hourly average concentrations above 0.12 ppm was equal to or
less than 1, averaged over 3 years (i.e., no more than 3 exceedances
in 3 years). Current emissions control strategies focus on reducing
the highest hourly average concentrations.
On September 16,
1997, EPA made the decision to replace the 1-hour standard with
a new human health (primary) and secondary (welfare/vegetation)
standard for ozone. The 8-hour primary and secondary standards
are set at 0.08 parts per million (ppm) with the fourth highest
daily maximum concentration averaged over 3 years.
For developing appropriate
emission control strategies for meeting the 8-hour standard,
it is important to identify the unique patterns of hourly average
concentrations that make up the 8-hour violations and to investigate
the importance of mid-range concentrations (i.e., 0.06-0.09 ppm)
in defining these violations. For those sites whose 8-hour violation
is mostly defined by hourly average concentrations below 0.09
ppm, control strategies will have to change from focusing on
the higher hourly average concentrations to the mid-level (i.e.,
0.06 to 0.09 ppm) hourly average values.
Using the EPA's
Aerometric Information Retrieval System (AIRS), renamed the Air
Quality System (AQS), A.S.L. & Associates found that for
the period 1993-1995, approximately 50% of the areas that violated
the 8-hour standard were influenced by 4 or more occurrences
of mid-level hourly average concentrations (i.e., less than or
equal to 0.09 ppm).
In addition, A.S.L.
& Associates identified those sites that demonstrated a significant
reduction in ozone levels for the period 1980-1995. Using the
data from the sites that experienced reduced ozone levels over
the period of time, A.S.L. & Associates investigated whether
the rate of decline of the mid-level hourly average concentrations
was similar to the rate experienced by the high hourly average
concentrations. The analysis indicated that there is a greater
resistance to reducing the hourly average concentrations in the
mid range than the hourly average concentrations above 0.09 ppm.
Figure 1 below is an example that shows that the higher hourly
average concentrations (i.e., above 90 ppb) decreased at a faster
rate (greater negative rate per year) than the hourly average
concentrations in the mid-level range. The numbers of hourly
average concentrations in the low end of the distribution also
decreased. Apparently, both the high and low ends of the distribution
are moving toward the center of the distribution.
As control strategies
are implemented, the resistance to lower the higher hourly average
concentrations will be low, but the resistance will increase
as one attempts to reduce the mid-level values. This experience
is similar to a piston compressing against a gas. The resistance
is initially low; however, the resistance increases as the piston
continues to compress against the gas.
Based on this analysis, which used empirical
data, it appears that when control strategies are implemented,
the "piston effect" will cause higher hourly average
concentrations to be reduced faster than the mid-level values.
Similar to the results obtained for the hourly values, the rate
of decline for the 8-hour daily maximum values in the mid range
will be much slower than the higher 8-hour values. Figure 2 illustrates
the slowing down process for Fairfield County, Connecticut. Note
the rapid decrease in the early years and then a "flattening"
of the curve in the later years. The summer 2004 in the eastern
United State was a fairly low ozone year due to meteorological
conditions. This is observed below in a drop in the 3-year average
of the 4th highest 8-hour daily maximum concentration for the
2002-2004, 2003-2005, and 2004-2006 periods. The 3-year average
of the 4th highest 8-hour daily maximum concentration increased
from 0.089 ppm for the 2004-2006 period to 0.092 ppm for the
2005-2007 period.
From 1980 through
1994, significant decreases occurred. However, since 1994, no
significant trend in decreasing ozone has occurred. When one
compares the Fairfield County figure with the results described
in EPA's latest trends report, the similarities are observed.
The slight downward pattern observed for the periods 2002-2004,
2003-2005, and 2004-2006 is a result of the 0.081 ppm 4th highest
annual 8-hour average value experienced in 2004. The year-by-year
figure below illustrates what is happening.
The 0.081 ppm value
that occurred in 2004 is the lowest 4th highest annual 8-hour
average value recorded between 1980 and 2007. The 4th highest
values recorded in the years 2005 and 2006 were 0.090 and 0.095
ppm, respectively. The effect of the 0.081 ppm value was to depress
the 3-year average of the 4th highest 8-hour concentration for
the 2004-2006 period. As noted in the previous figure, the 4th
highest 8-hour average increased from 0.089 for the 2004-2006
period to 0.092 for the 2005-2007 period. The elimination of
the 2004 contribution of the 0.081 ppm value (as a result of
including only the years 2005, 2006, and 2007), resulted in a
higher 3-year average for the 4th highest 8-hour concentration
for the 2005-2007 period. It appears that the "piston"
effect is affecting, in a large degree, the nation's ability
to improve its air quality.
For most sites that violate the 8-hour
primary standard, the attainment of this standard may be extremely
difficult. A peer-reviewed paper describing the piston effect
and its consequences was published in the June 1998 issue of
Environmental Science & Technology (Lefohn et al., 32(11):276A-282A).
In addition, in 1997, the "Piston Effect" was discussed
in an Atmospheric Environment New
Directions Column.
EPA has commented on the effects of the
"piston effect". In its October 5, 1998 report "Use
of Models and Other Analyses in Attainment Demonstrations for
the 8-Hour ozone NAAQS," the Agency noted
"Analyses of regional modeling results
suggest that relative reduction factors tend to have lower values
(i.e., greater reductions occur) if the starting concentrations
are higher (ECR, 1998; Meyer et al., 1997). Lefohn, et al., (1998)
report similar findings in their review of recent trend data."
Using models, several investigators have
commented on the difficulty in reducing the mid- level hourly
average concentrations while reducing the fourth highest 8-hour
average daily maximum concentration. Saylor et al. (1999)
noted that for the Atlanta, Georgia area, NOx emissions reductions
greater than 60 – 75% would be required to reduce the mid-level
hourly average concentrations. Winner and Cass (2000) noted that
the higher hourly average concentrations were reduced much faster
than the mid-level values during simulation modeling for the
Los Angeles area. Reynolds et al. (2003) analyzed ambient
ozone concentrations used in conjunction with the application
of photochemical modeling to determine the technical feasibility
of reducing hourly average concentrations in central California.
Various combinations of volatile organic compounds and oxides
of nitrogen emission reductions were effective in lowering modeled
peak 1-hour ozone concentrations. However, VOC emissions reductions
were found to have only a modest impact on modeled peak 8-hour
ozone concentrations. Reynolds et al. (2003) reported
that 70 – 90% NOx emissions reductions were required to
reduce peak 8-hour ozone concentrations to the desired level
in central California.
Reynolds et al. (2004) used ozone
measurements in conjunction with photochemical modeling to assess
the feasibility of reducing hourly average concentrations in
the eastern United States. The authors reported that various
combinations of volatile organic compound and oxides of nitrogen
emission reductions were effective in lowering modeled peak 1-hour
O3 concentrations. VOC emission reductions alone had only a modest
impact on modeled peak 8-hour ozone concentrations. Anthropogenic
NOx emissions reductions of 46-86% of 1996 base case values were
needed to reach the desired level of the 8-hour value in some
areas.
Reynolds et al. (2003, 2004) have
commented on possible chemical explanations for the observation
that more prominent trends in peak 1-hour ozone levels than in
peak 8-hour ozone concentrations or in occurrences of mid-level
(i.e., 0.06 – 0.09 ppm) have been reported. The authors
noted that when anthropogenic VOC and NOx emissions are reduced
significantly, the primary sources of ozone precursors are biogenic
emissions and CO from anthropogenic sources. Chemical process
analysis results indicated that slowly reacting pollutant such
as CO could be contributing on the order of 10 – 20% of
the ozone produced. The authors recommended that further work
focus on the need to confirm that biogenic emissions have not
been significantly overestimated in the most recent emission
inventories and on the examination of the effects of CO reductions.
If you would like
to read a summary of our May 1997 report describing our findings,
please click here.
References
Reynolds, S. D.; Blanchard, C. L.; Ziman,
S. D. (2003) Understanding the effectiveness of precursor reductions
in lowering 8-hr ozone concentrations. J. Air & Waste Manage.
Assoc. 53: 195-205.
Reynolds, S. D.; Blanchard, C. L.; Ziman,
S. D. (2004) Understanding the effectiveness of precursor reductions
in lowering 8-hr ozone concentrations - Part II. The Eastern
United States. J. Air & Waste Manage. Assoc. 54: 1452-1470.
Saylor, R. D.; Chameides, W. L.; Change,
E. C. (1999) Demonstrating attainment in Atlanta using urban
airshed model simulations: impact of boundary conditions and
alternative forms of the NAAQS. Atmos. Environ. 33: 1057-1064.
Winner, D. A.; Cass, G. R. (2000) Effect
of emissions control on the long-term frequency distribution
of regional ozone concentrations. Environmental Science &
Technology. 34: 2612-2617.
