United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S3-83-025 June 1983
Project Summary
Ambient Hydrocarbon  and Ozone
Concentrations   Near  a  Refinery
Ken Sexton and Halvor H. Westberg
  Atmospheric emissions from the
Marathon Oil  Company refinery at
Robinson, Illinois were investigated
during June and July 1977. Surface
and aerial measurements were used in
an integrated, three-dimensional mon-
itoring network The study focused on
three major areas: (1) characterization
of gaseous components within the re-
finery effluent especially non-methane
hydrocarbons (NMHCs)  and nitrogen
oxides (NOX); (2) sunlight bag-irradia-
tion experiments to examine ozone-
forming potential of refinery emissions;
and (3) aerial measurements of changes
in plume chemistry during the first six
to eight hours of transport. Concen-
trations of ozone, nitrogen oxides, sul-
fur dioxide, methane, carbon dioxide,
individual nonmethane hydrocarbons,
and halocarbons were recorded on a
routine basis. In addition, meteorologi-
cal  parameters such as wind speed,
wind direction, solar radiation, and
mixing height  were also measured.
Results indicate that levels of hydro-
carbons and nitrogen oxides were ele-
vated downwind of the refinery. In the
effluent, concentrations of  hydrocar-
bons and of nitrogen oxides exceeded
background values by as much as 300-
fold and 10-fold, respectively. Irradia-
tions of captured  refinery emissions
suggest that ozone can be produced
photochemically in amounts that vary
according to NMHC/NOX ratios and
initial NMHC concentrations. Contin-
uous measurements made from an air-
craft documented instances of ozone
buildup in the refinery plume  as  it
drifted downwind.

  This Project Summary was developed
by EPA's Environmental Sciences Re-
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
  Pollution caused by photochemical oxi-
dant production is becoming an increas-
ingly widespread  problem in the United
States, despite regulatory actions to limit
emissions of precursors tooxidants. Once
thought to be exclusively an urban phe-
nomenon, ozone (03)  levels in many
suburban and rural areas routinely violate
the current 120-ppb National Ambient
Air Standard.  Recognition of the impor-
tance of long-range multi-day transport of
airborne pollutants has directed attention
to the  regional nature of ozone/oxidant
pollution episodes and underscored the
need for effective control strategies. Re-
quirements for retrofit emissions control
measures on major stationary sources are
an integral part of contemporary regula-
tory policy. Before the efficacy of these
control measures can be evaluated, the
relative contribution of specific  point-
source categories to observed downwind
03 buildup must be  known.
  Petroleum refineries are major sources
of several primary pollutants, including
hydrocarbons, nitrogen oxides (NOX), car-
bon monoxide (CO), and paniculate matter.
Concerns about refinery emissions have
focused primarily on nonmethane hydro-
carbon (NMHC)  releases,  and  control
measures, such as floating  roof storage
tanks, have been instituted at most sites.
Because airborne emissions from refining
operations contain significant quantities
of hydrocarbons and nitrogen oxides, it is
reasonable to assume that ozone is formed
within the plume during appropriate mete-
orological conditions.

-------
  This report summarizes the results of an
ambient air monitoring study conducted
during June  and July 1977 near  the
Marathon Oil Company refinery at Robinson,
IL The refinery which processes approxi-
mately 200,000 barrels of crude oil  per
day is located in a rural region away from
other major pollution sources.  Aerial and
ground-level measurements were designed
to examine chemical transformation within
the plume  during the first  six to eight
hours of transport. Results of continuous
monitoring at the field laboratory, captive
air irradiation studies of refinery emissions,
and  aerial measurements in the plume
during transport are presented.  Discus-
sion focuses on the evaluation of local 03
concentrations due to photochemical pro-
duction of ozone in the refinery emissions.

Procedure
  Both surface and aerial monitoring sta-
tions were used to investigate the refinery
emissions.  The focal  point for  ground-
level operations was a 7-m instrumented
trailer located approximately 5 km east of
the refinery.  Pollutant parameters mea-
sured at the field laboratory included con-
centrations of 03, NOX, methane (CH4),
CO, total hydrocarbons, individual C2 - C-\ Q
NMHC and halocarbons; wind speed and
direction; turbulence;  dew  point; solar
radiation; and temperature.  In addition,
vertical temperature discontinuity was re-
corded continuously using a monostatic
acoustical sounder.  Natural sunlight ir-
radiation experiments were also conducted
at this site using 500-1 Teflon bags filled
with either refinery emissions or back-
ground air.
  A twin-engine Rockwell Aero Comman-
der served as an airborne monitoring plat-
form. Instruments on board the aircraft
allowed for continuous measurements of
03, NOX, sulfur dioxide (S02), and conden-
sation nuclei (CN) visual range, and flight
parameters.  Six-liter grab samples were
collected in stainless steel  canisters or
Teflon bags  and  returned  to the field
laboratory for gas-chromatographic analy-
sis.

Results

Ground-Level Measurements
  Continuous ground -level monitoring was
conducted at the field laboratory 5 km east
of the refinery complex. Measu rements at
this site during periods of fumigation pro-
vided an indication of pollutant concentra-
tions associated with the refinery effluent
Nonmethane hydrocarbons and NOX were
the best indicators of refinery emissions at
the surface, and CO and CH4 also exhibited
higher values during fumigations. Nitrogen
oxides and NMHC concentrations typically
jumped from background levels of < 10
ppb  and  < 0.1  ppmC,  respectively, to
approximately 20 ppb and 5 ppmC as
emissions passed over the trailer.   In-
plume CO concentrations commonly in-
creased about 0.2 ppm and CH4 values
exceeded background levels by 0.4 ppm.
  Because NMHC are a major component
of refinery emissions, considerable effort
was devoted to determining hydrocarbon
concentrations and  composition  in the
plume.  Hydrocarbon concentrations  in
morning samples 0600-0800 CDT col-
lected at the fenceline commonly exceeded
1000 /ig/m3 (1,500 ppbC), with values
varying from 200 -18,000 /xg/m3 (300 -
27,000 ppbC). Observed in-plume hydro-
carbon values at the surface were found to
be negatively correlated with early morning
mixing height as determined by the acous-
tical-sounder records. The proportion of
olefins, aromatics, and paraffins measured
in refinery emissions remained relatively
constant throughout the  monitoring pro-
gram. Olefins routinely comprised 1 - 5%
of each plume sample, aromatics 5 -1 5%,
and paraffins 80  - 95%.
  More detailed  gas-chromatographic a-
nalysis was performed to investigate iden-
tities of C2 - C10 NMHCs present in back-
ground and plume samples. Ground level
samples  from  outside the  plume were
composed primarily of the paraffinic spe-
cies ethane, propane, i-butane, n-butane, i-
pentane, and n-pentane.  Refinery emis-
sions were also composed principally of
paraffins, with ethane, propane, i-butane,
n-butane,  i-pentane, and n-pentane fre-
quently accounting for more than 50% of
the total.  Other compounds  present at
relatively high  concentrations  within the
plume included 2-methylpentane, 3-methyl-
pentane,  n-hexane,  methylcyclopentane,
2,4-dimethylpentane,  benzene,  2,3-di-
methylpentane,  3-methylhexane,  2,2,3-
trimethylpentane, n-heptane, methylcyclo-
hexane, toluene, and the isomeric xylenes.

Captive-Air Irradiations
  Natural sunlight irradiations of 500-1
Teflon bags filled with petroleum refinery
emissions or background air were used to
assess ozone-forming potential.  Three
bags (2 refinery and 1 background) were
collected between 0600-0800 CDT and
initial NMHC, 03, and NOX  concentration
measurements were performed at the field
laboratory before exposing the chambers
to sunlight  Additional nitric oxide (NO)
was added to selected refinery samples in
order to examine the effects  of varying
NMHC/NOX ratios.  Changes in 03, NO,
and  nitrogen dioxide (N02)  were moni-
tored periodically throughout each  run.
Ozone production in plume and non-plume
samples  was  compared to  provide an
indication of ozone-forming potential as-
sociated with refinery emissions.
  Maximum amounts of 03 produced in
background samples were less than 80
ppb in all cases, while refinery samples to
which no NO  had been added had 03
concentrations as high as 220 ppb. Plume
samples to which additional NO had been
added showed  maximum 03 generation
on the order  of 500 ppb.   Irradiation
experiments revealed  that  NMHC/NOX
ratios and  absolute amounts of NMHC
were important factors in determining 03
concentrations. Optimum conditions for
photochemical  ozone production occurred
at NMHC/NOX ratios between 10 and 20,
and  the  amount of 03  generated was
proportional to initial NMHC levels.

Aerial Measurements
  The instrumented aircraft made several
flights downwind of the refinery to investi-
gate in-plume  polluant levels.  Elevated
hydrocarbon concentrations  were  docu-
mented in the plume, with peak values in
the range  of 200-300  jug/m3 (300 -
400) ppbC) at 8 km, decreasing to 75 -
100 jug/m3 (100 - 150 ppbC) at 15 km.
Levels of NOx(+5 to + 1 5 ppb), S02 (+25
to +30  ppb),  and  condensation  nuclei
were also higher in the plume than outside
the plume.  On two occasions, July 8 and
July 21, 1977, 03 in the plume exceeded
background concentrations. Peak in-plume
03 enhancement was approximately 30
ppb  on July 8  and  15  ppb on July 21.
Background 03 levels for the two days
were 85 ppb  and  70 ppb respectively.
Both instances of ozone buildup occurred
when prevailing southerly winds carried
emissions to the north of the plant
  Data from two flights on July 21, 1977
are typical of pollutant levels measured in
the plume at various distances downwind.
The initial flight was conducted between
0630 - 0745 CDT before sufficient solar
radiation was available for maximum photo-
chemical activity. During this period, pre-
vailing winds carried Marathon emissions
to the northeast of the refinery. A*ter a
vertical spiral to 4300 m over the ^lant,
the aircraft made several passes through
the area of plume drift and then flew to 19
km from the refinery.
  Temperature variations  with  altitude,
recorded during the spiral at 0640 CDT,
showed a distinct temperature inversion
at approximately 350 m. Above the inver-
sion, 03 values were relatively constant at,
70 ppb, while beneath this level concentra-'

-------
tions decreased to less than 30 ppb near
the surface. The July 21 acoustical sounder
tracing revealed the presence of a low-
level radiation inversion at about 50  m
from 0300-0600  CDT, with a gradual
weakening and lifting between 0600 -
0930 CDT. Sounder records indicate a
breakup of the inversion by0930CDTand
well-mixed conditions prevailing up to at
least 500 m.
  Downwind  of the refinery,  the aircraft
made a number of passes through the
Marathon plume.   In-plume S02 values
were about 100 ppb higher than back-
ground levels at 8 kmand35 ppbhigherat
16  km  Condensation nuclei  levels on
both passes showed a ten-fold increase
over background levels. Ozone concentra-
tions, on the other hand, exhibited marked
depletion due to scavenging by NO in the
plume.  In-plume ozone values were de-
pleted  below background  levels  by ap-
proxi mately 3 5 ppb at 8 km and 10 ppb at
16 km.
  Hydrocarbon grab samples collected by
the  aircraft revealed elevated C2 - C10
NMHC concentrations in the plume. The
NMHC concentration in background sample
taken outside the plume was less than 20
jug/m3 (30 ppbC).  The  NMHC  level
exceeded 1000 jug/m3 (1500 ppbC) in a
plume sample collected 0 - 3.5 km down-
wind, while at 8 -11 km the concentration
was greater than 200 ju,g/m3 (300 ppbC).
  In order to investigate the possibility that
03  is produced within  the  effluent,   a
second flight was conducted  on July 21
from 1005 - 1 200 CDT.  Data collected
during  this flight demonstrate that 03
buildup does occur in the refinery's dis-
charges.  Emissions were observed to the
northeast of the plant and plume bound-
aries were well defined out to 26 km
downwind.  Temperature  data indicate
that the inversion observed on the earlier
flight had completely dissipated by 101 5
CDT, thus confirming the acoustical-
sounder tracings.  Ozone concentrations
were relatively constant up to 1200m,
suggesting that the mixing layer extended
to at least that altitude.  Measurements at
19 km from the plant tend to corroborate
this observation since  emissions were
detected at 1000 m.
  Cross-plume passes at 8, 13, 19, and
26 km downwind of the refinery revealed
that S02, CN, and 03 concentrations were
higher inside the plume boundaries than
outside.  The buildup of 03  within the
plume amounted to 7 ppb at 8 km, 8 ppb
at 13 km, 15 ppb at 19 kmandS ppbat26
km.  Plume traverses at several altitudes
19 km from the refinery suggested that
refinery emissions were stratified,  with
maximum 03  enhancement  and plume
width occurring near the surface. Greatest
ozone concentrations inside plume bound-
aries on this flight were measured at 250
m, 19 km from the plant


Conclusions and
Recommendations
  Photochemical  03 formation from re-
finery emissions is clearly a complex phe-
nomenon, depending on many meteoro-
logical variables and emission character-
istics. The fact that in-plume 03 buildup
was recorded in the effluent on two out of
six flights shows that 03 can be generated
solely from petroleum refinery discharges.
Meteorological and plume composition
data do not account for the absence of 03
enhancement on two-thirds of the sampling
days. Further investigations of the reasons
for  sporadic 03 formation may provide
valuable information about photochemical
activity in ambient atmospheres.
  Data reported here concentrated on 03
production  during the first six to eight
hours of transport It may be that maximum
03 levels occur much later, because 80 -
90% of NMHC emissions from refineries
are in a relatively slow-reacting classifica-
tion, the paraffinic compounds. Although
the contribution of low molecular weight
hydrocarbons  to  long-range  multi-day
transport of ozone is not well understood,
the potential impact of these emissions on
regional pollution levels should not be
ignored.
  Future captive air studies should involve
the use of N02 instead of NO. Too much
time is required to oxidize the NO to N02
by the NO addition. Also, the initial 03 is
titrated with the NO addition.  Thus, the
NO spiked air samples do not have suffi-
cient daylight  to both oxidize NO and
generate maximum 03.
  K. Sexton andH. H, Westberg are with Washington State University, Pullman, WA
     99164
  Joseph J. Bufalini is the EPA Project Officer (see below).
  The complete report, entitled "Ambient Hydrocarbon and Ozone Concentrations
     Near a Refinery,"  (Order No. PB 83-195 958; Cost: $23.50, 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

-------
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
        U S  ENV1R  PROTECTION  AGENCY
        REGION  5  LIBRARY
        230  S  DEARBORN  STREET
        CHICAGO  IL 60604

-------