EFFECTS OF REDUCED USE OF LEAD IN GASOLINE ON VEHICLE
EMISSIONS AND PHOTOCHEMICAL REACTIVITY
By
A. P- Altshuller
Director
Division of Chemistry and Physics
ENVIRONMENTAL PROTECTION AGENCY
National Environmental Research Center
Research Triangle Park, N. C.
February 1972
-------
EFFECTS OF REDUCED USE OF LEAD IN GASOLINE ON
VEHICLE EMISSIONS AND PHOTOCHEMICAL REACTIVITY
A. P. Altshuller
A program for reduction of lead from gasoline should have the following
positive environmental aspects:
1. Provide a lead-free gasoline for use with catalytic control
system on 1975 and subsequent model vehicles.
2. Provide a schedule for reduction of lead compatible with
achievina acceptable atmospheric levels of lead particulate
material in a reasonable time interval.
3. Minimize changes in fuel composition that may lead to undesirable
increases in other pollutants such as polynuclear aromatic hydro-
carbons and in photochemical reactivity.
These results should be accomplished with minimal adverse impact in the
following economic considerations:
1. Increase in price per gallon to consumer
2. Minimize and smooth out investment costs
3. Minimize crude oil increase
4. Minimize and smooth out process industry construction activity.
The requirement is for optimizing the environment impact at the minimal
economic cost to the community.
As indicated above, reduction in lead in fuel results in increases in
aromatic content to maintain octane number. Increases in aromatics in
fuel have been associated in some studies with increasing polynuclear
aromatic hydrocarbons in exhaust emissions and increased atmospheric
I
photochemical reactivity. Although increases in the aromatic content
-------
of gasoline are not the sole means of maintaining octane requirement, it
appears to be the only practical means of reaching this objective rapidly.
The rate and amount of increase in aromatic content will be determined by
the selection of a schedule for removal of lead additives from gasoline.
With respect to the effects of aromatics the rate of removal of lead additives
from each grade of gasoline is particularly significant as related to the
use of these grades in sub-groups of the vehicle population. These sub-
groups may be categorized as follows:
1. Uncontrolled vehicles - 1967 model year and earlier models
(1965 model year and earlier models in California).
2. 1968 to 1970 model year vehicles with control devices.
3. 1971 to 1974 model year vehicles requiring lower octane
gasoline and equipped with evaporative control systems.
4. 1975 model year and later vehicles.
Presently available experimental results on prototype vehicles indicate
that the hydrocarbon emissions including aromatic hydrocarbons will be
extremely low and little influenced by the aromatic content of the fuel
for 1975 model year and later vehicles. For 1971 through 1974 model year
vehicles the combination of evaporative control devices, efficient first
generation tailpipe emission control devices and lean mixture operation
should tend to minimize effects of aromatic content of fuels. The use
of tailpipe emission control devices provided significant reductions in
effects of aromatic content of emissions from 1968 to 1970 vehicles com-
pared to 1967 and earlier model vehicles.
-------
Fuel Composition and Vehicle Polynuclear Aromatic Hydrocarbon Emissions
Emission requirements for hydrocarbons based on Federal Standards if
normalized to 1.0 for 1967 and earlier models, result in emission factors
of 0.3 for 1968-1969 model vehicles, 0.2 for vehicles through 1974 and
0.02 for 1975 and later vehicles. The polynuclear aromatic hydrocarbon
emissions of 1968-1970 vehicles with emission controls using several fuels
average 30% of the emissions of uncontrolled 1964 to 1966 vehicles based
1 2
on the work of Hoffman, et al and Gross . The polynuclear aromatic hydro-
carbon (PNA) emissions of 1971 and later vehicles are assumed to follow the
control of hydrocarbons and carbon monoxide.
The influence of fuel composition particularly aromatic content, has been
1 2
investigated by both Hoffman, et al and Gross . Both studies agree on
an increase in PNA emissions in uncontrolled vehicles with increasing
aromatic content of fuel. Hoffman and co-workers reports a linear relation-
2
ship between aromatics in fuel and PNA in emissions. Gross indicates a
decreasing sensitivity to aromatics in fuel of PNA in exhaust comparing
2
results from 1966, 1968, and 1970 vehicles. In work reported by Gross
the PNA emissions from a 1970 vehicle actually decreased with increase in
fuel aromatics. This result was attributed to the adverse effect of leaded
deposits on PNA emissions. As a consequence, an unleaded high aromatic
fuel operated with a vehicle with unleaded deposits can lead to lower PNA
emissions than a lower aromatic leaded fuel operated in a vehicle with leaded
deposits. Thus, the sensitivity of 1968 and later model year vehicles PNA
emissions to fuel aromatic content is not resolved.
o
Computations have been carried out using the Bonner and Moore Study results
on the economic analysis of proposed schedules for removal of leaded additives
from gasoline. Schedules A, 0, and N were selected as the less severe
schedules considered which would result in lead reduction with lesser effects
on investment costs, added cost per gallon, crude oil demand, and process
industry construction activity. The lead reductions proposed for the three
schedules were as follows:
-------
Selected 1971 Bsnner and Moore Lead Reduction Schedules
Schedule
A
0
N
1971
1
5
0
0
(% Removal
1972
15
21
59
Over 1970-1
1973
15
33
59
971
1
974
15
45
73
Usage)
1
975
17
59
84
1976
20
63
88
1980
50
77
91
The computations consider the PNA emissions from the vehicle populations
as the aromatic contents of the 93 RON, and 100 RON blends varied with
schedule and year. Two sub-models were utilized, one assumed a linear
variation on PNA emissions with aromatic content of the fuel blends for
1974 and earlier model years. The second sub-model considered that a
siam'ficant aromatic sensitivity exists for only 1967 and earlier un-
controlled vehicles. The distribution of vehicle population was obtained
assuming the available U. S. distribution pattern from 1971 vehicles back
would hold throuohout the 1970's. The distribution would be as follows:
current year, 11$; 1 year old, 10%; 2 years old, 10%; 3 years old, 10%;
4 years old, 9%, 5 years old, 9%; 6 years old, 8%; 7 years old, 7%; 8
years old, 6%; 9 years old, 5%; 10 years old, 4%; older vehicles,11%.
The results obtained from this distribution did not differ significantly
from a California vehicle distribution pattern also available. (Table !)
Schedule A provides a lead-free fuel for 1975 vehicles, but does little
towards removal of lead from gasoline throuah 1976. The schedule is such
that a small initial increase of polynuclear aromatics occurs, but this is
adjusted so that there is no net change compared to the reference schedule
by 1973 followed by a small decrease in computed PNA emissions after 1974
compared to the reference schedule.
Schedule 0 provides a lead-free aasoline by 1975, but also provides for
increasinoly large removals of lead in the 1972 to 1975 period. This
schedule provides a net decrease in computed PN/1 emissions in 1972 to 1976,
laraely because of the reduction in aromatic content o^ the 100 RON fuel
used by the hiah-emission level older vehicles requirina this fuel.
-------
Schedule N provides for an almost immediate larae reduction in the lead
content of aasoline with additional almost complete removal by the late
1970's. The severity of schedule M is such as to increase aromatic content
in the three qrades of gasoline so as to cause a sizable increase in com-
puted PNA emissions over the reference schedule. The increase becomes
particularly large in 1975 and subsequent years.
In judging the overall impact of such schedules on atmospheric PNA levels,
the contribution of tailpipe emissions to total atmospheric loading must
be estimated. In cities such as Los Angeles and Washington, D. C. and
several other communities most of the atmospheric PNA. can be associated
with automobile emissions. The atmospheric levels in such communities
q A
averaae around 2 ua/m for benzo(a)pyrene. In highly industrialized communi-
ties in the eastern and southern U. S. with significant usage of coal,
o
benzo(a)pyrene levels averaoe from 4 to 10 ug/m with several communities
o
havinn levels averaoina at or above 20 ua/nf . Therefore, the automotive
contribution can ranae from 10 to 50% in such coal burning areas and higher
in non-coal burnina or non-industrialized communities. A 50% increase in
PNA emissions, assumino proportionality, may lead to an increase of about
o
1 ug/m of benzo(a)pyrene, compared with present overall community levels
of 2 to 20 ug/m3.
There is a compensating factor even when increases in PNA emissions occur.
As older higher-emission vehicles are removed from the roads, PNA emissions
will decrease as indicated by the decrease in the reference schedule. Even
with schedule N, this factor more than compensates for the increases in PNA
emissions compared to the reference schedule after 1972. Even a 60% in-
crease in 1980 in PNA emissions (schedule N, sub-model I) versus the reference
schedule still results in a PNA emission level which is only one-third of
the emission level in 1971 (reference schedule). The appropriate sub-
model becomes siam'ficant in such cases, since sub-model II predicts a
25% rather than 60% increase in 1980 for schedule N. A. resolution of the
differences in experimental results on aromatics in fuel sensitivity of
polynuclear aromatic hydrocarbons in exhaust is needed.
-------
p
Phenols. Gross reported on analyses for phenols in vehicle exhaust. Phenols
are linearly related to fuel aromatic content at least in the 11 to 46%
aromatic range. The phenol emissions from the 1970 vehicle tested were 70
to 75% as high as were the emissions from 1960 and 1968 vehicles. The con- '
centration levels of phenols in exhaust are not sufficient to lead to at-
mospheric levels of concern. Phenols as other organic species may effect
the carcinogenic potential of polynuclear aromatic hydrocarbons.
Photochemocal Reactivities. The term photochemical reactivity is a general
term which refers to a number of individual chemical, physical or biological
manifestations resulting from the reactions caused by ultraviolet irradiation
0
in the 3000-4000 A region of simple or complex mixtures of organic substances
with nitric oxide and/or nitrogen dioxide. The manifestations include ozone,
nitrogen dioxide, peroxyacyl nitrate (PAN), or aldehyde maximum concentrations or
dosages (ppm x time), eye irritation response, plant damage and aerosol
formation (visibility reduction). Other chemical effects such as rate of
hydrocarbon reaction or nitrogen dioxide formation from nitric oxide are
readily measured, but such effects are not of direct environmental concern
in themselves. The intensity of these individual manifestations are not
proportional to one another. For example, a single hydrocarbon or mixture
of hydrocarbons compared to another hydrocarbon or mixture with nitrogen oxides
when irradiated may produce large ozone yields but low eye irritation responses.
Most individual olefinic or aromatic hydrocarbons when irraidated with nitro-
gen oxides do produce substantial intensities for one or more manifestations.
Most paraffinic hydrocarbons and acetylene are much less reactive than olefins
or aromatic hydrocarbons.
The reactions of these irradiated systems are sensitive to temperature,
humidity, concentration, composition, and ratio of reactants, and added
substances such as sulfur dioxide or carbon monoxide. Changes in composition
of gasolines, emission mixtures or atmospheric samples do not produce as large
variations as does the adjustment of concentration and ratio of reactants.
Nevertheless, the effects of varying olefins or aromatic content for fuels
on photochemical reactivity has been the subject of much experimental investi-
gation.
-------
Concern has been expressed that the removal of lead from gasoline will
result in the need for higher aromatic content to which has been attributed
a possibly higher photochemical reactivity. In practice, the increase in
aromatics in gasoline must occur concurrently with reduction of olefins,
paraffins or both fractions in the fuel. The paraffins are in part converted
in the combustion process to move reactive olefins - ethylene, propylene,
butenes. The net result is a complex balance of composition and reactivity
in vehicle emissions which are not readily predicted from laboratory irra-
dations of simple mixtures. As a result, the present discussion will be
concerned with the reactivities resulting from irradiation of complex hydro-
carbon mixtures with nitrogen oxides, exhaust emissions and evaporative
emissions.
Several current and past studies (Hum and Davis^, Dishart6, Wigg, Campion
and Peterson? provide a good basis on fuel-exhaust hydrocarbon relationships.
Such relationships are needed in computation of photochemical reactivity.
The results can be summarized as follows:
1. For both uncontrolled and device controlled vehicles utilizing
leaded or non-leaded fuels, exhaust aromatic content is linearly
related to fuel aromatic content. No effect can be shown of fuel
olefin content on exhaust aromatic emissions.
2. Exhaust olefins decrease as fuel aromatics increase. The major
portion of exhaust olefins are the cracked products from combustion
(ethylene, propylene, butenes) not the unburned fuel olefins.
Exhaust olefin content is linearly related to the sum of fuel
olefin plus fuel paraffin content. Again the effects of control
devices versus uncontrolled exhaust emissions along with
leaded versus non-leaded fuels is minimal.
3. Increasing fuel aromatic content from 0 to 55% results in
a decrease in total molar aldehyde emissions of about 10%. The
-------
relationship between fuel aromatics and exhaust aromatic aldehydes
is linear.
The above results make it possible to assume linear relationships in
emissions within and among various vehicles types of 1967 and earlier, and
up to 1975 model year. Emission tests on prototype vehicles indicate that
the more reactive fuel and exhaust aromatics and olefins are removed in
greater relative amounts than cracked products such as methane., ethylene
and acetylene by catalytic devices and by thermal reactors.
Photochemical Reactivity - Oxidant - The Federal air quality standards are
based on ozone or corrected oxidant measurements. No standards exist related
to other photochemical reactivity manifestations such as eye irritation,
visibility reduction, plant damage nor the chemical rate of reaction parameters.
The corrected oxidant measurement involves correction for sulfur dioxide and
nitrogen oxides. The net oxidant values resulting have ozone as the major
constituent, but the smaller amounts of oxidants such as peroxyacyl nitrates
(PAN + PPN) would also be included in the measurement. Therefore, laboratory
smog chamber measurements of irradiated vehicle emissions or simulated vehicle
emissions for ozone and peroxyacyl nitrates as a function of hydrocarbon
emissions are most significant. Because of the differences in concentration -
time relations in the atmosphere compared to the smog chambers, dosage values -
PPM minute for the duration of the experiment usually 5 or 6 hours, are
utilized rather than the peak concentration in the smog chamber.
In work reported recently (Sturn and Dimitriades^) irradiationcchamber
measurements were made on emissions from a vehicle pperated on four different
leaded and non-leaded fuels of varying aromatic content. Results were ob-
tained on both exhaust plus evaporative emissions (1/1) and evaporative
emissions. The results are given in Table II. Fuel A, of 29% aromatic
content and 99.5 RON is least reactive. But fuel C (leaded) of 22% aromatic
content and 95.4 RON is not significantly less reactive than fuel B
(lead-free) 45% aromatic content and 101.9 RON. Fuel D (lead-
-------
free) of 31% aromatics is significantly more reactive than fuel B (lead-
free) of 45% aromatic content while fuel A is less reactive than fuel C
of lower aromatic content. Fuel D is more reactive than fuel C. There-
fore, no consistent relationship exists between fuel aromatics and
these measures of photochemical reactivity whether leaded or non-leaded
fuels are compared.
q
In a more recent set of unpublished measurements (Air Pollution Progress
Report, June 30, 1971) from Bureau of Mines, Bartlesville Research Center
reactivity measurements were obtained using 7 fuels in one vehicle to pro-
vide emissions irradiated in a smog chamber- The ozone, oxidant and
peroxyacyl nitrate (PAN) results available are summarized in Table III. The
reactivity values again do not vary with fuel aromatic content. The average
exhaust reactivity of the 52% aromatic content fuel is within 1% of the
average reactivity value for all 7 fuels. The reactivities of fuels 7003,
7005, and 7006 of 32, 28, 26% aromatic content are consistently higher
than fuel 7011 of the high aromatic fuel. The reactivities of the evaporative
emissions of the fuels containing 52% and 32% aromatics are at the low end
of the highest aromatic content fuel 7011 is lower than the reactivities
of all of the lower aromatic content fuels. The reactivities of the evapora-
tive emissions increase with fuel olefin content rather than fuel aromatics.
In a recent study in the EPA laboratories complex, hydrocarbon mixtures
were irradiated with nitrogen oxides. The mixture simulated atmospheric
hydrocarbon composition. The paraffin-acetylene fraction consisted of n-
pentane, isopentane, 2-menthyl pentane, 2, 4-dimethyl pentane, 2, 2, 4-
trimethyl pentane and acetylene. The olefin fraction consisted of ethylene,
propylene, butene^l, cis-butene-2, 2-methyl-m-xylene, r^-propylbenzene, see-
butyl benzene and 1,2, 4-trimethylbenzene. The proportions of each fraction
were varied to determine the reactivity effects of replacement or addition
of the three major hydrocarbon fractions. Part of the measurements are plotted
in figures 1 and 2 for oxidant (corrected) and peroxyacyl nitrate (PAN) as a
-------
10
function of olefin to aromatic content and of aromatic content and of
aromatic to paraffin content. Substitution of aromatics for olefins cause
consistent reduction in both oxidant and peroxyacyl nitrate dosages. A
further reduction occurs as aromatics are largely replaced by paraffins.
Although not plotted, replacement of olefins with paraffins also causes
a substantial reduction in oxidant and peroxyacyl nitrate dosages.
All of these results clearly show that increase in aromatics in fuel should
not be expected to lead to increases in oxidant or peroxyacyl nitrate
dosages. Although not listed, the maximum oxidant concentrations also are
insensitive to aromatic content of fuels.
Formaldehyde Dosage - Formaldehyde is one of the eye irritants associated
with photochemical smog. Formaldehyde dosage decreased by 20% as aromatic
content of the fuels utilized increased from 23 to 52% in the Bureau of
g
Mines irradiations of vehicle exhaust emissions mixtures.
Eye Irritation Response - Eye irritation responses of human panelists have
been reported in several studies involving irradiation of exhaust emissions
from vehicles utilizing fuels of varying hydrocarbon composition.
A significant reactivity parameter utilized for over 15 years is the eye
irritation response of human panelists to irradiated auto exhaust or hydro-
carbon-nitrogen oxide mixtures. One of the studies seeking to quantitate
eye irritation response to fuel composition effects on vehicle exhaust were
conducted by Schuck and co-workers in 1958 utilizing four vehicles and
five test fuels. The hydrocarbon to nitrogen oxide ratios and chamber
residence times were varied by changes in engine operating conditions. The
highest level of eye irritation was obtained by use of the olefinic fuel
(52% olefin, 24% aromatic). The basis mix fuel (16% olefin, 22% aromatic)
produced an exhaust which when irradiated caused less eye irritation than
produced from the olefinic fuel, but more eye irritation than produced
from the exhaust formed by use of an aromatic fuel (32% aromatic, 0.0%
-------
11
olefin; 33.5% aromatic, 0.5% olefin) or a paraffinic fuel (97% paraffin,
3% aromatic). In several of the experiments only the higher eye irritation
response of the olefinic fuel could be distinguished statistically from
that produced from the other fuels. The eye irritation response produced
from the aromatic and paraffin fuels could not be distinguished. During
1 o
the same period Hamming, et al. measured fuel composition effects on the
reactivity of irradiated exhaust. The influence of olefin content of fuel
was emphasized and a 22% increase in eye irritation index was reported to
occur with an increase of olefin content by 25% (6 to 31% olefin content).
An increase of the aromatic content of fuel 36% (2% to 38%) was required
to produce 8% increase in eye irritation index.
13
Work done by Wilson and Co-workers relating several photochemical re-
activity parameters to fuel composition effects on exhaust including some
eye irritation panel response measurements. Irradiated exhaust from use
of a 55% aromatic fuel caused eye irritation with 10% greater severity than
from a 25% aromatic fuel. When 0.75 ppm of sulfur dioxide was added to the
exhaust, the irradiated exhaust from use of a 55% aromatic fuel caused eye
irritation 90% more severe than from 25% aromatic fuel. This result was
attributed to a reduction in eye irritation in the exhaust with higher
olefin content (25% fuel aromatic content). Other measurements on this
fuel with 0.1 ppm S02 added resulted in a 30% reduction in eye irritation.
However, measurements with a 60% olefin - 40% aromatic-6 component mixture
showed no effect of increasing S02 from 0.00 to 0.3 ppm. A mixture con-
sisting of 7.5% aromatics, 0.5% olefins and the rest paraffinic upon
irradiation with nitrogen oxide produced 30% less eye irritation with no
S02.
While S02 does appear to have an effect on eye irritation of some hydro-
carbon mixtures, it is not at all clear at present whether S02 at atmospheric
levels around 0.1 ppm or less would interact with exhaust hydrocarbon
mixtures having smaller shifts in composition that used in the work of
Wilson, et al. to produce significant changes in eye irritation.
-------
12
Eye irritation responses also were obtained in the work in EPA laboratories^
on mixtures of the hydrocarbon fraction described earlier in this paper.
The eye irritation indices of the panel responses are tabulated in Table
IV- These mixtures represent the types of atmospheric composition which
would result from very large changes in fuel usage. Therefore, the changes
in eye irritation indices will be greater than those resulting from the
smaller shifts in fuel composition involved in the Bonner and Moore fuel
schedules.
A shift from a purely aromatic exhaust mixtures through an equal carbon ppm
mixture (26 mole% aromatic, 74% mole% olefin) to a purely olefinic mixture
causes no significant change upon irradiation in eye irritation index. Only
when an appreciable fraction of paraffinic hydrocarbons are substituted for
olefins and aromatics does a really significant shift in eye irritation in-
dex occur. Thus a mixture of the same carbon ppm level but containing a
molar composition of 25 mole% aromatics, 27 mole% olefins and 48 mole%
paraffins (mixture 4) does result upon irritation in 35 to 40% less eye
irritation than mixtures of olefins and/or aromatics only. Based on the
compositional relationships obtained by Wigg, Campion and Peterson,? this
type of exhaust composition (mixture 4) would be expected from a fuel con-
taining 45 to 50% aromatics. Changes in composition resulting in large
variations in the proportions of olefins and aromatics (mixtures 5 and 6)
with elimination of paraffins cause small downward shifts in eye irritation.
If the contribution of the amount of paraffin present is considered
additive, mixtures of 10 carbon ppm of olefin with 10 carbon ppm of paraffin
(63 mole% olefin, 37 mole% paraffin) would have an eye irritation index of
1.40 (1.10 + 0.30). A mixture of 5 carbon ppm olefin, 5 carbon ppm aro-
matic and 10 carbon ppm paraffin (39% olefin, 14% aromatic, 47% paraffin)
would have an eye irritation index of 1.35 (1.05 + 0.30). So shifts in
which olefin ranges from 27 mole% to 63 mole% and aromatic from 0 mole% to
25 mole% compute to cause a +_ 8% variation in eye irritation response.
-------
13
A significant downward shift occurs again when the 10 carbon ppm total
mixtures of olefins or olefins plus aromatics are substituted for by paraffins
as in mixture 7. Substitution of 5 carbon ppm of olefin in mixture 6 by
almost an equal amount of paraffin in mixture 7 with a small increase in
aromatics causes a decrease in index of 25% in eye irritation index. The
complete replacement of olefins or aromatics with paraffins in mixture 8
results in a mixture which upon irradiation causes one-quarter or less of
eye irritation from mixtures 4, 5 or 6.
Similar results were obtained in mixtures irradiated with 0.5 ppm
of nitrogen oxide. Again partial replacement of aromatic or olefin with
paraffin causes a significant decrease in eye irritation response. How-
ever, a mixture of a high aromatic content (mixture 10) has the same eye
irritation index, 0.80 units, as does mixture 11 with much less aromatic
and a large increase in olefin content.
These results with simulated atmospheric mixtures fail to show: substantial
effects on eye irritation of substituting aromatics for olefins. Only sub-
stitution of paraffins for olefins and aromatics can be predicted to cause
large decreases in eye irritation in irradiated hydrocarbon - nitrogen oxide
mixtures. Shifts of aromatics and olefins over ranges which are at the ex-
tremes for realistic gasoline composition changes are likely to cause eye
irritation responses of about +10%. This result is consistant with the
results of Hamming, et alj2 and Wayne, 1962,14 in which large increases
in fuel olefins from 7 to 31% and fuel aromatics from 2% to 72% caused their
eye irritation index to vary from 64 to 79 units or +_7.5 units in 70 units.
Overall the results do suggest that extremely large increases in olefins
or aromatics in gasoline so that lower reactivity paraffins and acetylene
were substantially reduced could cause modest increases in eye irritation.
However, the Wigg, et alJ relations clearly show within the range of 10
to 55% aromatics in fuel, increases in exhaust aromatics are largely com-
pensated by a reduction in exhaust olefins. For example, at 20% fuel
-------
14
aromatics, the exhaust mixture contains 10% aromatic and 33% olefins or
a total of 43% olefins and aromatics., while for 30% fuel aromatics the
exhaust mixture contains 15% aromatic 30% olefin or a total of 45% olefin
and aromatics. Thus a 10% increase in fuel aromatics results in only 2%
increase in exhaust aromatics plus olefins, resulting in only a small change
in eye irritation potential.
Plant Damage - Damage to various species, particularly petunia, pinto bean,
tobacco wrapper, and bluegrass has been obtained by irradiation of hydrocarbon-
nitrogen oxide mixtures. The chemical species causing the damage are ozone,
peroxyacyl nitrates and possibly alkyl hydroperoxides from certain aldehydes.
The available measurements suggest the following order of decreasing plant
damage potential by class of organic substance olefins aromatics aldehydes
paraffins-acetylene (Altshuller^). The trend should correlate with the oxi-
dant or ozone and peroxyacyl nitrate dosages discussed. As indicated, in-
crease in aromatics with reduction of olefins in fuel result in no significant
adverse effect of increased aromatics on these reactivity parameters. The
reactivity of evaporative emissions is more depended on olefin than on
aromatic content. The aliphatic aldehydes that produce plant damage upon
irradiation are the products from the olefin not the aromatic content of
exhaust. These comments are fully consistent with the one study in which
fuel composition was considered as a parameter related to the plant damage
potential of exhaust products. In the work done by Noble, et al^6 and Wayne,14
much greater plant damage to bluegrass was obtained by irradiation of the
exhaust from a highly olefinic fuel than from a highly aromatic fuel.
Based on the above considerations, an increase in the aromatic content of
the fuel would not be expected to increase the plant damage potential of
irradiated exhaust. Depending in part on the detailed composition of the
aromatic fraction of the fuel, increase in aromatics in fuels could lead to
a reduction in the plant damage potential of the irradiated exhaust and
evaporative emissions.
-------
15
Aerosol Formation - Olefins and alkylbenzenes when irradiated with nitrogen
oxides and sulfur dioxide produce aerosols. Four carbon olefins and smaller
produce only small amounts of aerosol compared to higher molecular weight
olefins and alkylbenzenes. Aerosol formation is greatly increased with
addition of sulfur dioxide to irradiated olefin-nitrogen oxide mixtures,
but not to aromatic-nitrogen oxide mixtures. Paraffinic hydrocarbons and
acetylene when irradiated with nitrogen oxides and sulfur dioxide produce
very small amounts of aerosols. Wilson, Miller, and Levy, demonstrated
that the degree of stirring is critical. They also showed that binary hydro-
carbon mixtures in which one or both hydrocarbons are aromatic produce much
less aerosol than would be predicted from respective single-component
experiments.
Earlier work by Schuck, et al, on irradiated exhaust indicated olefinic
fuels produced the greatest amount of aerosols. The work reported by
12
Hamming and co-workers suggested an increase in aerosols with increase
in fuel aromatics. Wilson, et al, investigated this problem using exhaust
or exhaust plus evaporative losses with 0.1 ppm of added sulfur dioxide.
Five leaded and unleaded fuels were used with an engine having leaded
deposits. The primary aerosol formation was not related to fuel aromatics.
Photochemically aerosols (secondary) formed in somewhat greater quantities
as measured by light-scattering coefficient with increasing fuel aromatics.
These results are complicated by the results for a fuel with 25% fuel
aromatics which produced much more secondary aerosols than fuels containing
15 and 21% aromatics and as much aerosol as produced from fuels containing
48% and 55% aromatics. The secondary aerosols from irradiation of the two
fuels averaging 18% aromatics produced 40% less aerosol than the two fuels
averaging 52% aromatics. For total aerosols (primary plus secondary) the
two low aromatic fuels produced 30% less aerosol than formed from the two
high aromatic fuels. The results for fuels with the 25, 48, 55% aromatic
contents could be interpreted as indicated in leveling off in aerosol for-
mation between 25 and 55% fuel aromatics. In view of the difference in
57-951 O - 72 - 2
-------
16
sensitivity of olefins and aromatics to sulfur dioxide the use of a range
of sulfur dioxide contents might lead to differences in fuel composition
effects also.
Computed Photochemical Reactivities - The previous discussion of photo-
chemical reactivities has been based on experimental results on irradiated
mixtures in smog chambers. Another approach is to compute reactivities
from exhaust composition measurements and from a reactivity scale which
assigns a reactivity value to each hydrocarbon class or sub-classes.
In the paper by Wigg, Campion and Peterson7 the exhaust composition measure-
ments were used to calculate photochemical reactivity by the composite re-
activity scale developed by Altshuller^5' 18, which averages over a number
of chemical, physical and biological smog manifestations. Using this scale,
class rankings were obtained in which olefins, aromatics, and paraffins -
acetylene-benzene were weighted 5.5, 4.5, and 0.5, respectively. The exhaust
composition in change with fuel aromatic content was calculated as follows:
Exhaust Alkylbenzenes (mole % in exhaust) =0.49 Aromatic Content (Mole %
in fuel)
Exhaust Olefins (mole % in exhaust) = 39 - 0.3 x Aromatic Content (Mole %
in fuel)
or
= 13 + 0.23 (Olefin + Paraffin content)
(Mole % in fuel)
The reactivities are a linear summation of the product of exhaust composition
multiplied by the reactivity factor. Based on this approach, the increase in
reactivity with an increase in fuel aromatic content from 20 to 40% is 4%.
-------
17
A similar computation was made which used the hydrocarbon subclasses. The
result was an increase in specific exhaust reactivity of 7% for an increase
in fuel aromatic content from 20 to 40%. The specific reactivity levels off
and decreases above 40% because of the shift in internal composition of the
aromatic fraction. The authors utilize a 7% decrease in hydrocarbon mass
emissions as representing the effect of shifting from a leaded to unleaded
fuel. Based on this correction, the two reactivity computations give a net
change of -3% and 0%. Wiggs, et alJ conclude that the predicted impact
of increased fuel aromaticity on photochemical reactivity is very small.
Dimitrades, Ecceston, and Hum, computed eye irritation from exhaust
composition and eye irritation scales developed by Heuss and Glasson^O and
Altshuller^8. Based on the two scales used to obtain comparing a 23% aromatic
content fuel with a 50% aromatic content fuel, + 10%, + 1% increase;
comparing another 23% aromatic content fuel with a 43% aromatic content
fuel, 0%, + 11%; comparing the same 23% aromatic content fuel with a 36%
aromatic content fuel, + 7% and 10%. The increases for the three sets
of fuels compared averages overall for the two scales at 7% for fuel
aromatic content increase of 20% (23% to 43%).
Summary
The results discussed in this paper are summarized in Table V in which
either the quantity of the pollutant emitted or the intensity of the
photochemical effect observed with solar irradiation is related to an in-
crease in fuel aromatic content on a per vehicle basis. These estimates
do not include the compensation in effects of fuel aromatic increases
resulting from the decrease in overall emissions reported to occur from
substitution of unleaded deposits from lead-free fuel usage for the present
lead deposits. This unleaded deposit effect has been found in most studies
to reduce gaseous emissions and polynuclear aromatic hydrocarbon emissions.
The cause of this deposit effect has not been isolated with respect to lead,
phosphorus and base fuel characteristics.
-------
18
As indicated in the previous discussion, the emissions for the vehicle
population rather than single vehicles are of importance. Since emission
factors vary with model years, the overall effect of a change in fuel aro-
matics depends on the aromatic content of each grade of fuel and the vehicle
sub-population utilizing that grade of fuel. The emission factors decrease
from 1.0 to approximately 0.02 for exhaust aromatics and olefins and probably
for polynuclear aromatics in comparing controlled 1967 model year and earlier
vehicles with 1975 model year vehicles. Even where positive effects occur
(in some vehicle sub-populations) the overall effects can actually be negative
or only slightly positive if the fuel aromatics are decreased in high emitter
sub-populations and increased in low emitter sub-populations. As a result,
schedule 0 in the Bonner and Moore Study would lead to a small net decrease
or no effect for polynuclear aromatic hydrocarbons and phenols through 1967
and no effect for photochemical reactivity. Schedule A results in no
photochemical effect and a small change of ± 10% or less in emissions depending
on the year by year schedule. For photochemical reactivity effects even a
change of 5% in average aromatic hydrocarbon emissions averaged over the
entire vehicle population which would correspond to an overall increase of
10% in fuel aromatics could barely be detected for any manifestation with
the possible exception of aerosol formation. Both schedules A and 0 require
increases of less than 10% in the pool fuel aromatic content. Additional
compensation results from distribution of the increases in fuel aromatics
to the low emitting vehicles.
3
If an ambient air standard at or near 2 yg/m annual or seasonal average
is selected, schedule A will not provide adequate reductions of lead until
into the 1980's. Schedule 0 will provide reductions so that cities at or
below 4 yg/m will be at or within the Standard by 1974, cities at or below
3
5 yg/m will be at or within the Standard by 1975 and cities at or below
3
6 yg/m will be at or within the Standard by 1977. Schedule N would achieve
needed ambient air standards by 1974. However, this more severe schedule
would lend to appreciable increases in atmospheric levels of polynuclear
-------
19
aromatic hydrocarbons, phenols and perhaps organic aerosols. Schedule N
is somewhat more costly in terms of cost per gallon to the motorist and
does induce a business cycle in the construction industry.
Recommendation
Either schedule A or 0 in the Banner and Moore study would achieve most
of the objectives listed at the beginning of this discussion. Schedule 0,
but not schedule A would come close to achieving acceptable atmospheric
levels of lead particulate material in a reasonable time interval. Neither
schedule A or 0 should result in any significant adverse effects on
emissions or photochemical reactivity because of changes in fuel composition,
Other schedules similar to A and 0 could be selected which should be equally
acceptable.
-------
20
References
1. Hoffman, Jr., C. S., Willis., R. L., Patterson, G. H., and Jacobs,
E. S. "Polynuclear Aromatic Hydrocarbon Emissions from Vehicles,"
presented to Div. Petroleum Chem, Amer Chem Soc., Los Angeles,
March 28 - April 2, 1971
2. Gross, 6. P., Second Annual Report on Gasoline Composition and Vehicle
Exhaust Gas Polynuclear Aromatic Content. CRC-APRAC Project No. CAPE-6-
68. APCO/EPA Contract CPA-70-104. Period ending April 15, 1971. Wayne,
L. G., The Chemistry of Urban Atmospheres, Technical Progress Report
Vol. III. Los Angeles County Air Pollution Control District, Dec 1962
3. Bonner and Moore, Assoc., Inc. "An Economic Analysis of Proposed
Schedules for Removal of Lead Additives from Gasoline," Prepared for
EPA under control No. 68-02-0050, June 1971
4. Air Quality Data from the National Air Surveillance Networks and Contributing
State and Local Networks, 1966 and 1967 editions, U. S. Dept. of Health,
Education, and Welfare, Public Health Service. Nat!. Air Pollution Control
Administration Publications No. APTD
5. Hum, R. W., Davis, T. C. "Gas Chromatographic Analysis Shows Influence of
Fuel on Composition of Automotive Engine Exhaust" Presented at 23rd mid-year
meeting, American Petroleum Institute, Div. of Refining, Los Angeles, Calif.,
May 12, 1958
6. Dishart, K. T., "Exhaust Hydrocarbon Composition, its Relation to Gasoline
Composition" Presented at 35th mid-year meeting, American Petroleum
Institute, Div. of Refining, Houston, Texas, May 14, 1970
-------
21
7. Wigg. E. E., Campion, R. J. and Peterson, W. L., "The Effect of Fuel
Hydrocarbon Composition on Exhaust Hydrocarbon and Oxygenate Emissions"
Unpublished report, Esso Research and Engr., Dec 1971
8. Sturn, G. P., Jr., and Dimitriades, B., "Reactivity of Emissions from
Leaded and Lead-free Fuels" Presented before Div. of Water, Air and
Waste Chemistry, American Chemical Society, Los Angeles, Calif.
March 28 - April 2, 1971
9. Air Pollution Progress Report. From Bureau of Mines, Bartlesville Energy
Ressarch Center to Environmental Protection Agency for Quarter Ending
June 30, 1971
10. Division of Chemistry and Physics, Natl. Environ. Research Center, RTP,
Environmental Protection Agency, unpublished results
11. Schuck, £. A., For
-------
22
15. AHshuller, A. P., "Reactivity of Organic Substances in Atmospheric
Photooxidation Reactions" Int. J. Air & Mater Poll. 10. 713 (1966)
16. Noble, W. M., Pelle, W., Wright, L. A., and Mader, P. P., "A Pre-
liminary Report on the Use of Plants ad Indicators for a Comparison
of the Smog Forming Properties of the Exhaust Gases from Two Types
of Motor Fuels." Los Angeles Air Pollution Control District, Sept. 1959
17. Wilson, W. E., Jr., Miller, D. F., Levy, A., "The Effect of Fuel
Composition on Engine-Exhaust Aerosols. API Project EF-2, Nov 1971
18. Altshuller, A. P., "An Evaluation of Techniques for the Determination of
the Photochemical Reactivity of Organic Emissions," J. Air Pollution
Control Assoc. ]6^, 257 (1966)
19. Dimitriades, B., Ecceleston, B. H., and Hum, R. W. "An Evaluation of
the Fuel Factor Through the Direct Measurement of Photochemical Reactivity
of Emissions" J_. Air Poll. Control Assoc. 20_, 150 (1970)
20. Heuss, J. M. and Glasson, W. A., "Hydrocarbon Reactivity and Eye
Irritation" Environ. Sci and Tech. 2^, 1109 (1968)
-------
23
Table I. CALCULATION OF IMPACT OF BONNER AND MOORE (1971 STUDY)
LEAD-REDUCTION SCHEDULE ON POLYNUCLEAR AROMATIC TAILPIPE
EMISSIONS FROM PASSENGER VEHICLES
Schedule
0 - Submodel Ib
0 - Submodel IIC
N - Submodel Ib
N - Submodel II c
A - Submodel Ib
A - Submodel IIC
Year
1971
1972
1973
1974
1975
1976
1980
Baseline3
71.0
63.7
56.4
49.8
42.3
35.5
15.0
Change from baseline, %
-
-
-
-
+10
+ 9
-13
-13
+16
+12
-
-
- 7
- 9
+15
+10
_
-
-12
-11
+22
+12
-
-
- 2
- 8
+56
+38
-
-
- 1
- 6
+44
+26
-14
-14
+22
+13
+60
+25
- 6
- 9
Baseline reference number normalized to 100% vehicles as per U.S. vehicle
distribution and to relative emission factors.
Linear aromatic fuel content effect on PNA emissions for 1974 and earlier
models.
£
Linear aromatic fuel content effect on PNA emissions for 1967 and earlier
models and no aromatic fuel content effect on PNA emissions for 1968 and
later models.
-------
24
Table II. REACTIVITY OF FUELS WITH DIFFERENT AROMATIC CONTENTS - IRRADIATED
EXHAUST PLUS EVAPORATIVE EMISSIONS AND EVAPORATIVE EMISSIONS SEPARATELY
Concentration of reactants
HC/NOX: 6 ppm C/l ppm
1. Oxidant (KI), ppm-min.
2. PAN + PPN, ppm-min.
1. Oxidant (KI), ppm-min.
2. PAN + PPN, ppm-min.
HC/NOX: 3 ppm C/0.5 ppm
1. Oxidant (KI), ppm-min.
2. PAN + PPN, ppm-min.
1. Oxidant (KI), ppm-min.
2. PAN + PPN, ppm-min.
Type of
sample
Exhaust
+ evaporation
Evaporation
Exhaust
+ evaporation
Evaporation
Fuel designation/% aromatic content
Fuel C/
22%
60
8.0
5.6
2.1
65
5.4
19.6
1.7
Fuel A/
29%
28
5.7
-
1.7
45
4.7
-
1.6
Fuel D/
31%
70
11.0
13.8
4.0
75
8.3
27.1
3.0
Fuel B/
45%
61
9.4
3.1
1.7
71
7.6
15.5
1.4
-------
25
Table III. REACTIVITY OF FUELS WITH DIFFERENT AROMATIC CONTENTS -
IRRADIATED EXHAUST AND EVAPORATIVE EMISSIONS
Concentration
of reactants
HC/NOX: 6/pptn C/l ppm
1. Oxidant (KI),
ppm-min.
2. Ozone (chemi-
luminescence) ,
ppm-nnn.
3. PAN + PPNS
ppm-min.
1. Oxidant (KI),
ppm-min.
2. PAN, ppm-min.
HC/NOX: 3/ppm C/l ppm
1. Oxidant (KI),
ppm-min.
2. Ozone (chemi-
luminescence) ,
ppm-min.
3. PAN + PPN,
ppm-min.
Type of
sample
Exhaust
Evapora-
tion
Exhaust
Fuel designation/% aromatic content
7002/
23% A
152
145
19.2
11.9
3.7
96
93
8.6
7004/
24% A
162
152
19.8
5.4
2.1
99
100
9.2
7006/
26% A
163
160
21.3
10.6
3.6
107
104
9.8
7007/
27% A
138
140
16.0
5.5
2.0
91
95
8.3
7005/
28% A
174
157
23.2
17.7
4.5
110
107
10.4
7003/
32% A
165
160
22.1
4.8
2.5
98
97
10.0
7011/
52% A
163
157
20.7
1.7
1.6
103
98
9.2
-------
26
Table IV. EYE IRRITATION INDEX FOR VARIOUS MIXTURES OF IRRADIATED
HYDROCARBONS AND NOX
Mixture
number
1
2
3
4
5
6
7
8
9
10
11
12
Aroma tics,
Olefin,
Paraffin,
ppm C/mole %
20 /1 00
0/0
10 / 26
8 / 25
0/0
5 / 26
5.5 / 41
0/0
5 / 26
5.25/ 41
2 / 11
0/0
0 / 0
20 /TOO
10/74
3/27
10 /TOO
5/74
0/0
0/0
5/74
0/0
1.5/ 25
0/0
0/0
0/0
0/0
9/48
0/0
0/0
4.5/ 59
10 /TOO
0/0
4.5/ 59
6.5/ 64
10 /1 00
NOX,
ppm
1
1
1
1
1
1
1
1
0.5
0.5
0.5
0.5
Eye irritation
i ndex
1.95
1.90
1.90
1.20
1.10
1.05
0.89
0.30
1.02
0.80
0.78
0.32
-------
27
Table V. POLLUTION EFFECTS RESULTING FROM 10% INCREASE IN FUEL AROMATIC
CONTENT PER VEHICLE
Substance emitted or effect observed
Probable result of 10% increase in fuel
aromatic content per vehicle
Tailpipe emissions
Polynuclear aromatic hydrocarbons
Phenols
Aromatic hydrocarbons (gasoline
fraction)
Olefins
Photochemical reactivity
Ozone or oxidant
Peroxyacyl nitrate (PAN)
Formaldehyde and other aldehydes
Eye irritation
Aerosol formation
Plant damage
Averaged overall of above
manifestations
Linear increase up to 1968 (2) model or
1975 model vehicles
Linear increase up to 1970 model
vehicles-unknown on later models
Linear increase up to 1975 model
vehicles
Approximately linear decrease above 10%
fuel aromatic content
No increase overall and decrease for
evaporative loss contribution
No increase overall and decrease for
evaporative loss continuation
Decreases with increase in fuel aromatic
2 to 5% increase
10% increase (perhaps occurs only below
25% fuel aromatic)
No increase or possible decrease
Very small or none (within ±5% change)
-------
28
250
TOTAL HC/NOX
• 10 ppm C/l ppm
A 20 ppm C/l ppm
• 10 ppm C/0.5 ppm
100% 50% PARAFFIN 100% 50% OLEFIN 100%
PARAFFIN 50% AROMATIC AROMATIC 50% AROMATIC OLEFIN
MIXTURE COMPOSITION
Figure 1. Effect of paraffin, olefin, aromatic mixture concentration on oxidant dosage.
100%
PARAFFIN
50% PARAFFIN
50% AROMATIC
TOTAL HC/NOX
• 10 ppm C/l ppm
A 20 ppm C/l ppm
• 10 ppm C/ 0.5 ppm
50% OLEFIN
50% AROMATIC
OLEFIN
AROMATIC
MIXTURE COMPOSITION
Figure 2. Effect of paraffin, olefin, aromatic mixture composition on oeroxvacvl nitrate /PANU dosage.
------- |