vvEPA
United States
Environmental Protection
Agency
Environmental Sciences Research ««•
Laboratory ^
Research Triangle Park NC 27711
Research and Development
EPA-600/S3-81-030 July 1981
Project Summary
Historical Emission and
Ozone Trends in the
Houston Area
Marilyn Davis and John Trijonis
An analysis of historical trend data
for hydrocarbons, nitrogen oxides,
and ozone in the Houston region from
1974 to 1978 is conducted in order to
check the EKMA ozone model and to
obtain a better understanding of the
Houston ozone problem. In addition to
historical trend analyses, the study
includes a detailed investigation of
seasonal/diurnal/meteorological pat-
terns in ambient ozone data at two
Houston monitoring sites (Mae Drive
and Aldine).
Historical emission trends are com-
piled for 26 NMHC source categories,
for 15 NOx source categories, for each
year from 1974 to 1978, and for five
sub-areas of the Houston region. It is
found that estimated NMHC emissions
in the Houston region decreased 17%
from 1974. to 1976 but then increased
12% from 1976 to 1978, so that there
was only a slight net decrease over the
4-year period. Estimated NO, emis-
sions increased by 18% over the 4
years. These emission trend estimates
involve significant uncertainties.
Ambient precursor data indicate an
even more pessimistic picture of his-
torical trends (i.e. moderate increases
in NMHC and substantial increases in
NO, from 1974 to 1978).
The historical emission trends are
entered into the EKMA isopleth model
to predict historical ozone trends; the
predicted trends are then compared to
actual ozone trends. Predicted ozone
trends change very little from 1974 to
1978 (except for a slight increase
from 1976 to 1977). For most sites/
indices, actual ozone trends show
little net change from 1974 to 1978,
although there are significant fluctua-
tions within the 4-year period. Most of
the discrepancies between predicted
and actual trends can be explained by
the large error bounds in the analysis
(e.g. uncertainty in the emission trend
estimates and meteorological variance
in the actual ozone trends). It is con-
cluded that historical emission changes
in the Houston region were not large
enough from 1974 to 1978 to provide
a definitive test of the EKMA isopleth
method.
Although this study of historical
trends in Houston is unable to provide
a definitive test of the EKMA method,
it does provide a better understanding
of the emission and air quality effects
being produced by control programs
and source growth in the Houston
region. In particular, this study ex-
plains an apparent paradox noted by
previous researchers—the lack of an
ozone decrease from 1974 to 1978 in
light of expected sizeable reductions
in hydrocarbon emissions. There is no
paradox; hydrocarbon emissions did
not decrease significantly from 1974
to 1978 because rapid source growth
negated the effects of controls and
because many of the controls were
installed prior to the 1974-1978
period.
This Project Summary was developed
by EPA's Environmental Sciences He-
search Laboratory, Research Triangle
-------
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The Empirical Kinetic Modeling Ap-
proach (EKMA) has recently been pro-
posed by EPA as a method for evaluating
ozone control strategies. Before the
EKMA method is accepted as a reliable
technique for control strategy analysis,
it should be subjected to validation
studies. The purpose of this report is to
test the EKMA method using historical
trend data for the Houston area. Specifi-
cally, nonmethane hydrocarbon (NMHC)
and nitrogen oxide (N0«) emission
trends in Houston are estimated for the
years 1974 to 1978; these emission
trends are entered in the EKMA method
to predict historical ozone trends; the
predicted ozone trends are then com-
pared to actual ozone trends as a test of
the EKMA method.
The historical trend data presented
herein are not only of interest with
respect to testing EKMA but also by
themselves. By conducting a detailed
analysis of precursor (NMHC and NO,)
emission trends, we can show how
emissions for individual source catego-
ries have changed due to controls and
source growth, and how total emissions
have responded to trends for individual
source categories. By examining trends
in ambient NMHC and NO, data, we can
check our emission trend estimates
with an independent measure of pre-
cursor trends. By analyzing historical
ozone trends, we can investigate whether
or not the ozone trends make sense in
light of precursor changes and meteoro-
logical fluctuations. Thus, this study is
useful not only as a means to validate
EKMA, but also as a means to under-
stand the air quality effects actually
being produced by control programs and
source growth in the Houston region.
There are two basic reasons why
Houston was selected for the study
region. First, as evidenced by the Houston
Area Oxidants Study sponsored by the
Chamber of Commerce and by several
Houston programs sponsored by EPA, a
great deal of interest currently exists in
the Houston ozone problem. Second,
substantial changes in historical pre-
cursor emissions are required for an
adequate test of the EKMA method;
previous investigators in Texas have
suggested that large reductions in
NMHC emissions, on the order of 20 to
35%, have occurred in the Houston area
since 1974.
Figures 1 and 2 illustrate some of the
important characteristics of the study
area. In the study, actual ozone trends
are determined from data at two Houston
monitoring stations, Mae Drive and
Aldine. The investigation covers the
years 1974 to 1978 at Mae Drive and
1975 to 1978 at Aldine. These are the
only years for which complete smog
season ozone data are available at the
two locations.
Emission trends are compiled herein
for 26 NMHC source categories and 15
NO, source categories on a year by year
basis from 1974 to 1978. The emission
trend study region consists of Brazoria,
Galveston, and Harris counties. Emission
trends are determined for the entire
study region and individually for five
sub-areas of the region: Brazoria County, •
Galveston County, Ship Channel area of
Harris County, Houston city area of
Harris County, and remainder of Harris
County. Because our analyses of ambient
ozone data indicate that sources within
Harris County are mainly responsible
for ozone episodes at Mae Drive and
Aldine, emission changes for Harris
County only are used in the isopleth
validation tests.
Conclusions
The objectives of this study are (1) to
gain a fundamental understanding of
ozone phenomena at Mae Drive and
Aldine by examining temporal ozone
patterns and ozone/wind relationships;
(2) to determine historical trends in
NMHC and NO« emissions in Houston
resolved on a yearly basis, for various
HS Houston Ship
Channel
Heavy Traffic
Moderate Traffic
Light Traffic
Figure 1. Approximate spatial distribution of traffic density in the Houston area.
-------
ub-areas, and for all major source
categories; (3) to characterize historical
ozone trends at Mae Drive and Aldine;
and (4) to test the EKMA isopleth model
by comparing predictions of historical
ozone trends with actual ozone trends.
The following sections summarize our
findings and conclusions with respect to
each of the above four objectives.
Phenomenological Analysis of
Ozone and Wind Data
As expected, ozone levels tend to be
higher at Aldine (located asa downwind
receptor site for Houston air pollutants)
than at Mae Drive (located near the
major source areas). For both locations,
the seven months-April-October-comprise
a distinct "smog season" that is con-
ducive to elevated ozone concentrations.
The most salient characteristic of the
ozone/wind relationship at both Mae
Drive and Aldine during the smog
season is the association of high ozone
with calm winds. Another important
feature of the ozone/wind relationship
at both locations is the association of
high ozone with transport from either
the Houston City area or the Ship
Channel area.
The daily ozone maximum is most
likely to occur from 12 AM to 4 PM at
Mae Drive and 1 PM to 5 PM at Aldine.
Assuming that solar irradiation effec-
tively begins at 8:00 AM, the mean
irradiation times up to the ozone peak
during the smog season are 6V4 hours at
Mae Drive and 6% hours at Aldine. On
days of extreme ozone, the mean irrad-
iation time is approximately 7 hours at
both locations.
Historical Emission Trends
Emission trends are determined by
three factors: uncontrolled emission
Each • — 5000 tons/year
of nonmethane hydrocarbon
emissions from major
point sources.
^Figure 2. Approximate spatial distribution of hydrocarbon emissions from major
point sources in the Houston area.
levels for various source categories,
growth of source activity levels, and the
timing and effectiveness of controls.
Each of these three factors often involves
considerable uncertainty. To help min-
imize these uncertainties, we have
attempted to use the most accurate and
up-to-date information available from
industrial representatives, consulting
firms, trade associations, and local/
state/federal agencies. Despite our
efforts, certain information gaps persist
which lead to significant uncertainties
in our emission trend estimates.
Estimated emissions of NMHC in the
three-county study region decreased
17% from 1974 to 1976 and increased
12% from 1976 to 1978, so that only a
slight NMHC reduction (7%) occurred
from 1974 to 1978. The single major
source category undergoing a significant
reduction in NMHC from 1974 to 1978
was the chemical industry which, de-
spite extremely rapid growth in output,
managed to decrease NMHC emissions
by 34%. Although some degree of con-
trol was installed on all other major
NMHC source categories from 1974 to
1978, these controls were insufficient
to overcome growth rates; thus, NMHC
emissions from motor vehicles, the
petroleum industry, and other sources
all increased slightly from 1974 to
1978. The two major reasons for the
lack of significant NMHC reductions in
the Houston region from 1974 to 1978
were rapid source growth and the fact
that most of the stationary source
controls were already in place by 1974.
The regionwide 7% decrease in NMHC
emissions from 1974 to 1978 was
composed of a 34% decrease in Brazoria
County, a 13% decrease in Galveston
County, and a 1% increase in Harris
County. Brazoria County experienced a
significant decrease in NMHC because
the chemical industry (which was con-
trolled during the period) is of great
relative importance there.
Estimated emissions of NO, in the
three-county region rose by 18% from
1974 to 1978. The NO, increase was a
product of high growth and the absence
of NO, controls (exceptforautomobiles).
The predominant part of the NO, rise
resulted from a 43% increase in chemi-
cal industry NO, over the four years.
Increases in NO, from other sources
were as follows from 1974 to 1978:
petroleum industry—13%, power plants—
15%, motor vehicles—7%, and all other
sources—13%. The increase in NO,
emissions from the chemical and petro-
-------
leum industries would have been sub-
stantially greater were it not for a
voluntary fuel conservation program
that significantly decreased fuel use per
pound of product.
The regionwide 18% increase in NOx
consisted of a 20% rise in Brazoria
County, a 7% rise in Galveston County,
and a 22% rise in Harris County. The
Ship Channel area of Harris County
experienced an especially large in-
crease in NOx (45% over the four years)
due to growth in the chemical and
petroleum industries.
Trends in ambient precursor data at
Mae Drive and Aldine present a more
pessimistic picture than our emission
trend data. The ambient data indicate
that NMHC concentrations increased
somewhat (on the order of 7 to 23%) and
NOx concentrations increased substan-
tially (on the order of 16 to 48%) from
1974 to 1978. It should be noted,
however, that the ambient trend data
(like the emission trend data) also
involve significant uncertainties.
Historical Ozone Trends
Complete smog season ozone data
are available for 1974-1978 at Mae
Drive and 1975-1978 at Aldine. Con-
sidering the entire historical record, no
overall trends are evident in the ozone
data. The only exception is the yearly
second maximum hourly concentration
at Aldine which displays a significant
decrease from 1975 to 1978.
Meteorological normalization of the
ambient ozone trends should eliminate
weather biases in the trends and should
reduce the statistical error bounds on
the trends. In a companion study, how-
ever, our co-workers encountered sev-
eral fundamental difficulties with mete-
orological normalization techniques
that precluded their application to the
Houston ozone data. Further research
into meteorological adjustment proce-
dures is necessary before such proce-
dures can be used in practice.
Stratification of the historical ozone
trends by wind direction fails to reveal
any consistent patterns at Mae Drive or
Aldine. Because of the limited number
of data points for given wind directions,
the wind-stratified trend data appear to
be dominated by large stochastic fluctu-
ations from year to year.
Validation of the EKMA
Isopleth Method
In the validation studies, historical
emission trends and the standard EKMA
4
isopleth model are used to predict
historical ozone trends; these predicted
trends are then compared to actual
ozone trends. The tests are conducted
from 1974 to 1978 at Mae Drive and
1975 to 1978 at Aldine using two ozone
air quality indices: the 95th percentile of
daily maxima during the smog season,
and the yearly second maximum hourly
concentration.
Because the emission changes during
the study period were rather small, and
because NMHC and NO. emissions
underwent rather similar variations
from 1975 to 1978, the predicted ozone
trends are insensitive to the initial
NMHC/NO. ratio (ratios of 8.5:1, 17:1,
and 47:1 are used in the tests). The two
major sources of error in the analysis
are uncertainties in the emission trend
estimates and (especially) statistical
errors in the ambient ozone trends
caused by yearly weather fluctuations.
Predicted ozone trends show little
change from 1974 to 1976, a slight
increase from 1976 to 1977, and little
change from 1977 to 1978. For both air
quality indices at Mae Drive and for the
95th percentile of daily maxima at
Aldine, actual ozone trends increase
moderately in 1975 and/or 1976 and
then decrease somewhat in 1977 and
1978. The net result for these three
cases is that the isopleth model under-
predicts ozone somewhat in 1975 and/or
1976 but yields fairly good agreement in
1977 and 1978. The discrepancies
could easily be explained by potential
errors in the emission trends and by
meteorological variance in the actual
ozone trends. In the fourth case, yearly
second maximum ozone at Aldine,
actual ozone underwent a substantial
decrease from 1975 to 1978, producing
a large discrepancy with predicted
trends. The large discrepancy might be
explained by the veryanomalous behav-
ior or this particular air quality index at
Aldine (i.e. by an extreme statistical
error in this air quality index at Aldine).
Given the rather large error bounds in
our analysis, it is difficult to tell if there
is any fault with the EKMA model itself.
In fact, a major finding of this study is
that emission changes in Houston from
1974 to 1978 were not large enough to
provide an adequate test of the EKMA
model.
Previous investigators in Texas have
noted an apparent paradox in the sense
that ozone levels increased somewhat
at Mae Drive and Aldine from 1974 to
1976, a period in which hydrocarbon
emissions were supposedly reduced by
about 20 to 35%. Our analysis helps to
explain this paradox. For one, the ozone
increase from 1974 to 1976 seems to
have been a temporary fluctuation
(ozone levels came back down in 1977
and 1978). Also, the NMHC emission
decrease was not as much as expected
(12% in Harris County and 17% in the
three-county study region from 1974 to
1976).
Marilyn Davis and John Trijonis are with Technology Service Corporation, Santa
Monica, CA 90405.
Basil Dimitriades is the EPA Project Officer (see below).
The complete report, entitled "Historical Emission and Ozone Trends in the
Houston Area." (Order No. PB 81-184 574; Cost: $12.50, subject to change)
will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
« US GOVERNMENT PRINTING OFFICE. 1961 -757-01Z/7201
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS 0000329
230 S DEARBORN STREET
CHICAGO IL 60604
------- |