ISSUE PAPER
\
Estimated Public Health Impact as a Result of
Equippi ng Light-Duty Motor ye hic ieswi th
Oxidation Catalysts.
Office of Research and Development
Office of Air and Waste Management
Environmental Protection Agency
January 30, 1975
-------
11533
Table of Contents
Executi ve Summary , 1
Intrcducti on 5
Emission Factors' ..... 7
Exposure Estimates -. 14
Other Non-Regulated Emissions 20
Health Effects 22
Automoti ve Sulfate Control Options .' 25
Benefit-Risk Analysis 35
Conclusions 33
-------
Executive Summary
This report is intended to provide a timely update to the three
technical papers submitted to the Senate Public Works Committee on
January 11, 1974, in support of the Administrator's testimony before
that body on November 6, 1973, regarding the impact on public health
due to equipping light-duty motor vehicles with oxidation catalysts.
The report is based upon extensive data gathered by the; Office of
Research and Development and the Office of Air and Waste Management
as a result of an accelerated and expanded research program mandated
by the Administrator. The detailed technical findings of these
programs have been gathered into a single separate EPA report entitled '
"Annual Catalyst Research Program Report," dated November 1974.
One problem which complexed the Administrator's decision regarding
the use of oxidation catalysts and the related emission of sulfuric acid
was the widely different-results obtained from the two measurement
methods employed by EPA and industry laboratories. One such method
indicated high sulfuric acid emissions from catalyst-equipped vehicles
and easily measurable levels from non-catalyst vehicles. The other
method suggested moderate emission rates of sulfuric acid from catalyst-
equipped vehicles and, little, if any, from non-catalyst vehicles. This
issue has been resolved. £atalyst-equipped light duty motor vehicle,s_d.o.
ejrnt^significant quantities of sulfuric acid while non-catalyst.vehicles .
have only trace emissions of this substance. Additionally, we have had
TITS"opportunity t^etel^tne^ sulfuric acid emission rates from production
19,75 catalyst-equipped vehicles unlike the prototype vehicles in the
past. Our studies have shown that we cannot use the Federal Regulated
Emissions Test Procedure to obtain representative sulfuric acid emission
factors, however. This is due to a sulfate storage phenomenon associated
principally with the pelleted catalyst system. Thus, estimated emission
factors for sulfuric acid must employ additional driving cycles.
Our emission factor_estimate in November 1973 was 0.05 gm/mile sulfuric
add'Current data snow that different emission factors"'are needed to re-
flect different emission control designs and geographical differences in
gasoline sulfur levels. We now estimate that in highway driving, catalyst-
equipped vehicles designed to meet the 1975-76 Federal Interim Standards
outside California will average 0.03 gm/mile sulfuric acid. Estimates for
such vehicles designed to meet either the 1975-76 California Interim
Standards or the 1977 statutory standards are 0.05 gm/mile nationally, and
0.08 gm/mile in California. (Gasoline in California Contains more sulfur^
than the national average.)
Estimates of population incremental exposures to sulfuric acid from
oxidation catalysts have employed both physical models (CO dispersion,
assumed activity models, .etc.) and surrogate models (lead, carboxyhemo-
globin). As noted in last year's report to Congress, these models show
exposure levels which may vary by as much as.a factor of two. The physical .
models have been used in this.analvsis t.n generate the ?4 hnur and peak hourly
-------
exposure level esti-iites showt- HI fable 1. The ?4 hou" fe
represent projected exposure 'ievcVi for an assumed commuter1 i.i«o
lives within 150 feet of a major freeway anu travels that iVebv-.vy
to and from his job in a l.arge urban center. The peak hour estimates
represent-projected exposures for persons travelling the freeway
described in Table 1. It should .,he noted that the public heaU;!-.
benefit-risk analysis (pps. 35-37) uses exposure estimates for large-
urban cities based on the rarboxyhemoglobin surrogate method which
are independent of assumedJ.:;Jan activity patterns.
Table 1
Incremental Sulfuric Acid Peak Hourly Exposures for a
Pedestrian Near a Major Arterial Thoroughfare and 24 Hour Average
Exposures for a Commuter Living Near the Expressway for
Normal and Adverse Meteorological Conditions
Assumptions: CO dispersion model from a 10 lane expressway with
20,000 vehicles per hour at 30 mph average speed
Hydrocarbon/CO.
Emission Control
Scenario"
Continue 49 State
jnterini Standards
for 30 years.
Adverse .'let.
Normal -'-let.
Imole-.ent California
Interim Standards or
Statutory Standards
Nationally in 1977.
Adverse ;'.et.
Normal Met.
~f, i 1 form' a Interim or
Stdt'jtory- Standards in
Ca'i ifornic. Only.
Adverse "ot.
Normal Mot.
Pca: ib'jrlv Increrental,
Exposure to K9SO. (pcm/m )
aftlr ^
?. yrs.
52
3
52
3
140
" 8
4 yrs.
104
6
140
8
280
16
7 yrs.
156
9
227
13
.420
24
10 yrs.
208
12
315 /
18
560
32
Increrc:iL3 1 24-hour H,:-0.
.Expcr.;.re for Urbr.n Res'c'-i-h't
L i v i ;" ' ' i - " r E x ri r n. s s v: 3 v
2 yrs .
5
.8
5
.8
13
2
4 vrs . , 7 yrs .
10
1.6
13.8
2.1
25
4.3
15
2.1:.
22.6
3.4
' 40
6.4
'. U y r
20
3.2
31.4
4.7
54
3.5
'Assumes 2 hrs on expressway, 1 hr street canyon or complex source, 13 hours
home, 8 hours at work
^Different sulfuric acid emission .factors are employed. See Table 3 in body
of report.
Healt i effects thresholds -for exposures tQ sulfuric acid have not been
well defined. A useful ..index,; however, is the health effects threshold for
exposures to ambi.ent suspended .part'sculate sul; ic-s.. These data .suggest that
adverse effects occur in. susceptible persons i- iur population after exposures
to IQi! 'n?- suspended, su.lfates, for 24 hrs. or ,.-.! .-<>. Sulfurvc acid 1.-;. believed
to be m,.,.. irritating,.than the wa.ter soluble ^uS^vindod sulfate materv.l.
Short te.Ti peak exposure thresholds are not available. However1, tin.:
-------
- 3 -
literature suggests that healthy young adults exhibit changes in
respiratory"function after 15 minute exposures to concentrations of
350 ,-tgm/m3 of sulfuric acid. We believe that the 10 ^gm/rii3 for 24
hour exposure represents a level at which people with existing heart
and respiratory disorders .will be adversely affected. This, of course,
provides'no margin of. safety. ,-.'...
Several options have been considered for the control or -re.duction
of sulfurfc acjLd emissions from oxidation catalyst-equipped vehicles.
Tfiese fall into two basic classes':'^ limitation of the1 quantity of sul-
furic acid emission from new motor vehicles and limitations on the
suTfur 'contenf'of motor vehicle fuels. f!o regulatory approach could
be expected to have any impact on automotive sulfate emissions before
1977 and then only if regulations requiring reduced fuel sulfur levels
based upon fuel blending and allocation are found feasible. Gasol irie_
desulfurization, while technically.feasible, cannot be expected to have
much impact before 1980. Control of vehicle sulfate emissions through
TaTfate" emission standards or possible shifts by manufacturers to non-
catalyst technology would likely not have a significant imnarf
before toe '979 mode; y».-;-.
Other emission prcductb unique to oxidation catalyst-equipped vehic'les\ -
have been suggested and/or reported. These include hydrogen sulfide (H2$),
ghgsp'hine (PH3), platinum, palladium, alumina (Al?03) , nitric acid (HNQ3),
C3robn"~di-sulfide (C$2) , and carbonyl sulfide (COS). Consumer complaints
have been received regarding a "rotten-egg" odor from some 1975 catalyst-
equipped vehicles. This is likely H2S , but could also be COS or CS?. As
these compounds are reduced forms of sulfur (while sulfuric acid is an
oxidized form), it appears that they (and PH3) are generated during cold
start choked operation or in the event of certain types of carburetor
malfunction or misadjustment or air pump failure causing the catalyst to
lack sufficient oxygen to operate as an oxidizing medium. EPA is investi-
gating this problem in-house, under contract, and with the auto manufacturers.
While such emissions appear to occur rarely, it should be noted that H2^,
COS, CS?, and PH3 are very potent toxic materials if humans are exposed to
sufficient concentrations. Nitric acid emissions have been shown to be
unaffected by oxidation catalysts. Current production vehicles, unlike
some earlier prototypes, apparently do not emit platinum or palladium at
levels that can be detected. Emissions of alumina (the catalyst substrate)
also appear to be very low. /
Related research programs are underway to characterize non-regulated
emissions from potential alternative light-duty vehicle power plants. Most
emphasis has been focused on diesel engines. The work, to date, suggests
that while ciiesels can easily achieve the statutory HC and CO emission
targets, emissions of particulates, sulfur particulate, and organic
materials, especially aldehydes, may pose dis-benefits to their widespread
use unless such emissions are controlled.
' \
A detailed benefit-risk analysis was performed to estimate the trade-offs
to public health in using oxidation catalysts by comparing increased sulfuric.
acid exposure dis-benefits to benefits associated with reduced exposures co
carbon monoxide and oxidants (unburned hydrocarbons are the key precursors;.
-------
Although the comparision of health benefits and risks is difficult to
precisely quantify, the results of our recent analysis suggest that,
if sulfate exposures increase as our models project, the continued
use of oxidation catalysts on virtually all new vehicles to reduce
hydrocarbon and carbon monoxide emissions would result in-a net
public health risk from increased sulfate exposure after 4 model
years. It should be noted however, that this conclusion is based
upon assumptions about dose responses and human exposure about which
there still remain uncertainties.
-------
-5-
Introduction
In mid-1972, EPA research groups voiced concern about the potential
for hazardous non-regulated emission products from oxidation catalysts. -
planned for use on 1975 model year light-duty-motor vehicles. Data from
a number of sources began surfacing in early 1973, which suggested that
sulfuric acid, platinum and palladium were emitted from such systems.
As more data became available and the EPA laboratories in both Ann Arbor
and Research Triangle Park intensified their investigations, it became
apparent that sulfuric acid was indeed ,e;ni tted from oxidation catalysts.
The issue quickly focused upon the potential public health ir.pact of such
emissions, while the time available to make any decision regarding the
use of the devices was rapidly diminishing. The Senate Public Works
Committee held hearings addressing the issue specifically on November 5
and 6, 1973.
In testimony before the Senate on November 6th, the Administrator
voiced his decision not to interfere with planned use of oxidation catalysts
on 1975 model year vehicles. In addition, he mandated an accelerated and
expanded EPA research program to attack four primary .issues:
1. Accelerate work on development of a reliable test
procedure for automotive sulfate emission measurement.
2. Consider all feasible alternatives for automotive sul-
fate emission control.
3. Improve the Agency's ability to estimate the public
health impact of sulfate and other automotive emissions.
4. Improve understanding of the atmospheric chemistry
involved in these emissions and initiate an appropriate
air monitoring program.
This broad, interdisciplinary research program has been initiated,
It is being conducted in close cooperation by both the Office of Research
and Development and the Office of Air and Waste :-:anarenient. The detailed
technical results of the effort nave been prep-rrea in a single, separate
report entitled "Annual Catalyst Research Program Report''1 dated November
1974, which is currently being circulated within the Agency in draft t'crm
for comments and review.
This report draws upon the technical report for information, but is
not intended to be a technical report in itself. Rather, it is intended
to put the issue of the future use of oxidation catalysts on motor vehicles
-------
-0-
1n a perspective of public health benefits and risks and to suggest
what impact decisions related to future emissions standards will have
thereon. The specific topics covered are those which influenced the
decision regarding catalysts in late 1973 and, with a greatly expanded
research information base, attempts to project future impacts. It
must, however, be noted that most of the research programs were only
initiated in mid-1974 and are not now completed. On balance though,
we feel that a much better defined basis currently exists to assess
these complex issues than was available last year at this time.
-------
-7-
Emission Factors
At the' time EPA was' preparing testimony for the Senate in November
1973, regarding catalyst-generated sulfuric acid, two rather different
approaches were being'used'to ascertain the emission levels of sulfuric
acid by both EPA and industry. One, an adaption of a stationary source
SOZ/SOs method (referred to as Method 8) was principally used by the
EPA laboratory in Ann .Arbor and. by Chrysler Corporation. Data obtained
by this method indicated that substantial sulfuric'acid was emitted by
non-catalyst vehicles while catalyst-equipped vehicles emitted hic.-ier
levels. The otner method, used in minor .variations of design, was an
exhaust air-diluted particulate technique' pioneered by Habibi of DuPont.
This method indicated that little or no sulfuric acid was emitted by
non-catalyst vehicles but that easily measurable quantities were emitted
by oxidation catalyst-equipped vehicles. Extensive research has shown
that the air-dilution method is correct, while Method 8 generated arti-
fact sulfate. At the time the Administrator made his decision regarding
the use of catalysts on 1975 vehicles, it was not known which method was
correct.
The emission rate of sulfuric acid from'oxidation catalyst-equipped
vehicles, while clearly established, has'been somewhat more difficult to
quantify. Sulfuric acid emissions are dependent upon at least the following
variables: vehicle cycle, fuel sulfur level, catalyst type, degree of
HC/CO control desired (75 National Interim Standards vs 75 California
Interim Standards vs Statutory Standards), fuel economy, catalyst pre-
conditioning, catalyst age, catalyst operating temperature, exhaust oxygen
content, and catalyst space velocity.
Vehicle Cycle:
The relationship of automotive-generated pollutants to achieving Air
Quality Standards is generally predicted using the Federal Test Procedure
(FTP) for motor vehicle emissions. This procedure employs cold and hot
start cyclic tests separated by a 10-minute hot soak for a total of 41
minutes. Mass emission rates are determined for the complete cycle for
the pollutants which are regulated. Obviously, this test procedure was run
to determine sulfuric acid emission rates from catalyst-equipped vehicles.
In addition, various cruise modes were operated to better elucidate the
sulfuric acid emission rates from catalyst-equipped vehicles. It was 'found
that the mass emission rate of sulfuric acid was substantially different
for cold start, hot start, cyclic, and cruise operating modes. In addition,
the two basic catalyst systems (monolithic and pelleted) were dramatically
different in their sulfuric acid emission characteristics.
-------
-8-
Further investigation was undertaken to better identify the cycle
dependence of the sulfuric acid emission rate. These studies have
established that the pelleted catalyst system is a low emitter of sul-
furic acid under conditions of the cold-start Federal Test Procedure
while it is similar to the monolithic systems under cruise conditions.
Further investigation has clearly shown that the pelleted catalyst
stores sulfur compounds under cold -start conditions and releases them
as sulfuric acid and SOg.gas at higher, temperatures associated with
cruise modes. Ford has also conducted fairly extensive studies which
showed little storage in monolithic catalyst but significant storage'.
1n pelleted catalysts, particularly when new. . .
' ." ' '' ..*'
It now appears certain. that the Federal Test Procedure cannot be ,
solely used to obtain sulfuric acid emission factors. MOSt EPA research,
both in-house and under contract, has focused in recent months upon-.
development of a representative cycle to properly reflect real world :
sulfuric acid emission rates. A combination of the Federal Test Procedure',
the Highway Fuel Economy Test Procedure, and Highway Cruise Modes appears
needed to properly assess sulfuric acid emission rates for air quality >' '
projections.
j * c
Fuel Sul fur. Level : '
Sulfuric acid emissions from oxidation catalysts result from the "
oxidation of S02 gas which is emitted from the combustion chamber of the
engine as a result of the burning of sulfur compounds in the fuel. The
average gasoline in the U.S. contains 0.03 weight % sulfur while Southern.
California gasoline contains 0.05 to 0.07 weight % sulfur. Table 2 pre-
sents representative data relating sulfuric acid emissions to fuel sulfur
levels. .
-------
-9-
Table 2
Effect of Fuel Sulfur Level on Sulfuric Acid
Emissions from Oxidation Catalyst-Equipped Vehicles
Fuel Sulfur
Wt %
.019
.091
.no
.019
.032
.057
.082
.107
Cycle
40MPH
Cruise
40MPH
Cruise
40MPH
Cruise
FTP
FTP
FTP
FTP
FTP
Catalyst Type
Monolithic
Pelleted
Monolithic
Pelleted
Monolithic
Pelleted
Monolithic
Monolithic
Monolithic
Monolithic
Monolithic
H2S04 gms/mile
0.019
0.002
0.123
0.126
0.163
0.168
0.009
0.014
0.019
0.023
0.029
-------
-10-
Clearly, and n'ot unexpectedly, sulfuric acid emission rates
increase with increasing sulfur levels in the fuel. Some debate
exists as to whether such increases are linear with fuel sulfur
level , however, for all catalyst systems. It is normally assumed,
and is in this report, that fuel sulfur level and sulfuric add
emission rates are linearly related. It is essential, therefore,
to establish the fuel sulfur level of gasoline in estimating sulfuric
acid emission factors and the related exposure estimates. We have
assumed 0.03 weight % for the nation with the exception of Southern
California where 0.05 weight % sulfur has been assumed.
HC/CO Control Level;
The 1975-76 Interim Emissions Standards provide two emissions
targets for the auto manufacturer: reasonably stringent California
standards which almost all manufacturers haye met utilizing air-injected
oxidation catalyst technology, and less stringent national .standards
which have resulted in a majority of the manufacturers employing catalysts,
in many cases without air-injection. The oxidation catalyst is a
device which requires excess oxygen in the oxidation of HC and CO. The
message is quite simple to achieve greater oxidation (lower levels of HC
and CO) ni"J *r provide more excess oxygen. All -data available indicates that
increasing the excess oxygen (through air-injection) available to the
catalyst increases sulfuric acid emission rates. Data reported by GM
suggest that doubling the oxygen level doubles the sulfuric acid emission
rate. Therefore, quite apart from the fuel sulfur level factors discussed
earlier, we can expect California Interim Standard and Statutory Standard
vehicles to emit greater quantities of sulfuric acid. Should non-a.ir-
injection catalyst approaches be employed to achieve future emission stand-
ards, however, the emission of sulfuric acid would likely be substantially
reduced. Non-catalyst-equipped vehicles, on the other hand, do not show
changes in sulfuric acid emissions with increasing exhaust content .(air
injection). i
Fuel Economy;
Sulfuric acid emissions are directly related to the amount of sulfur
dioxide (SOg) passing over the catalyst per mile driven, all other factors
held constant. Thus, any shift toward a higher catalyst-equipped vehicle
population average fuel economy would proportionally reduce the emission
factor for sulfuric acid from such vehicles. Efforts to, in essence, drive
the catalyst harder by air-injecting and exhaust a dirtier product from the
engine to the catalyst in order to optimize engine fuel economy and drive-
ability would not likely provide such a benefit, however. The major benefit
would accrue from a shift to a lower average vehicle weight and lower
aerodynamic drag.
-------
-n-
Catalyst Pre-Conditiom'ng^:
Pre-conditioning factors are important in ascertaining reliable sul-
furic acid emission rates from some catalysts. Hhen pelleted catalysts
are used, significant sulfur storage occurs as discussed earlier. Often,
when testing new pelleted catalysts, no sulfuric acid emissions are
measured. This is likely due to the capability of the large amount of
catalyst substrate (5-7 pounds) to store large quantities of sulfur com-
pounds until saturated. The same storage capability, even with high
mileage pelleted catalysts, makes it essential that we know the immediate
pre-history of the vehicle before testing. For example, very different
sulfuric acid emission rates would be obtained if the vehicle had experienced.
several repetitive FTP tests as compared to a long high speed cruise. These
factors appear to be much less important when testing monolithic catalysts.
Catalyst Age:
In general, most current data indicates a trend for more raoid in-
creases in regulated emissions during early mileage accumulation and a less
rapid increase with higher mileages. It thus appears that catalyst deteriora-
tion is more severe during the first miles than the later ones. One would
expect that as the catalyst's efficiency in converting HC and CO decreased,
so would its efficiency in converting SOo to sulfuric acid. While only
limited experimentation has been dona on this subject, it tends to confirm
this anticipation. Work conducted under contract to our Ann Arbor labora-
tories suggest on the order of a 25;"c.; decrease in sulfuric acid emission rate
with a high mileage catalyst. Such factors are not, however, sufficiently
quantified and have not been .used in exposure estimates and emission factor
calculations.
Catalyst Operating Temperature:
In theory, catalyst temperature would be expected to have a significant
effect on sulfuric acid formation from SOo. Thenrodynairn'cally, one would
expect decreased conversion to sulfate as~catalyst temperature increased
assuming the oxygen content was held constant. Indeed, one tends to see
such effects with monolithic catalysts, principally in oench-scale tests.
Similar data from pelleted catalysts is simply not available due to the
overwhelming effect of sulfur storage. In addition one typically calculates
sulfur conversion percentages of 30:,; for catalysts operating in cruise modes
while the theoretical conversion efficiency is over twice that percentage.
Unfortunately, automotive catalysts operate most of the time at tem-
peratures where sulfur conversion is theoretically quite efficient. Sub-
stantially lower operating temperatures would result in less efficient
HC/CO conversion and much higher temperatures would result in catalyst
durability problems.
-------
-12-
Space Velocity;
It would be expected that the rate of exhaust product flowing across
the catalyst would have an effect on sulfur conversion efficiency due to
reaction kinetics effects due principally to a limitation of the reactant
trasldence time in the catalyst chamber. Only GM has reported results of
sych studies which indicated a 4% decrease in sulfuric acid formation with
a 4-fold increase in space velocity. The effect of space velocity therefore,
appears to be a very minor one.
Emission Factors;
The many factors which influence the emissions of sulfuric acid from
catalyst-equipped light duty motor vehicles have been reviewed. Most
of the data and results discussed have been obtained using prototype
vehicles. Only very recent and limited data was available to us from
actual 1975 production vehicles tailored to achieve the Federal and Califor-
nia Interim Standards. It is apparent from the foregoing that those
parameters most important in determining catalyst sulfuric acid emission
factors are fuel sulfur level, vehicle cycle, catalyst type, and the HC/CO
emission standard. Our ability to quantify the effects of these factors
is best for the first parameter and becomes progressively poorer.
Based upon the information available to us at the present time,
researchers in the Office of Research and Development and the Office of
Air and Waste Management concur with the following sulfuric acid
emission factors (Table 3).
Table 3
Sulfuric Acid Emissions Factors For 1975 and Subsequent
Model Year Light-Duty Motor Vehicles
(Assumes 100% Catalyst Usage in California 1975-76 and Nationally 1977-86)
Vehicle
Model Year
1975/76
1975/76
1977 - 1986
Fuel
Sulfur
0.03%
0.05%
0.03%
HC/CO
Emission Standard
Federal Interim
(49-state)
California Interim
Statutory
H2S04 Emission Factor (gms/mile
FTP
.01
.025
.02
;HFET
.03
.08
.05
'
^Cruise (MPH)
301
.03
.12
.07
40
03..
.11
.07
60
.04
.09
.06
FTP » Federal Test Procedure (Regulated Emissions)
HFET - EPA Highway Fuel Economy Test
Note: California Interim and Statutory Emission Standard Catalyst-
equipped vehicles are assumed to use air injection
-------
-13-
These emissions factors take into account fuel sulfur levels,
percentage of 1975-1975 vehicles equipped with catalysts, the per-
centage of vehicles equipped with a catalyst on one-half the exhaust
systems (75-76), and the percentage of vehicles equipped with
pelleted (40:J) and monolithic (60'.') catalysts. The 1975 National
Interim Standards average fuel economy is assumed for each model year.
Application of these emission factors to exposure estimates will be
discussed in a subsequent section of this report.
-------
-14-
Exposure Estimates
Estimates of incremental sulfuric acid exposures as a result of
the use of oxidation catalysts have employed b oth physical models,
(such as carbon monoxide dispersion) and surrogate models (such as
carbon monoxide, carboxyhemoglobin and lead). For purposes of this
review only physical models will be employed. A number of basic
assumptions are made: (1) sulfuric acid aerosols disperse like a
stable gas and (2) sulfuric acid emission factors are as shown in
the previous Table 3 for the conditions specified. Typical and ad-
verse meteorological conditions are examined and projections of peak
hour and 24-hour average exposures are made based upon the vehicular
source, vehicular traffic flow, vehicular average speed, and meteoro-
logical conditions specified for such exposure estimates. Estimates
are shown for a single major expressway, intersecting expressways, a
street canyon, and a complex source typified by a sports complex.
For selected cases, estimates for incremental catalyst generated
sulfuric acid exposures are made for the work place and at home for
an urban resident living near a major arterial throughway and working
in the downtown urban area. From such exposure estimates and an
assumed activity pattern, 24-hour incremental catalyst generated sul-
furic acid exposures are estimated.
CO Dispersion Model - Peak Hourly Exposures
Table 4 provides tabulated peak hourly incremental sulfuric
acid exposures based upon various emission scenarios for selected
receptors for a single, major arterial throughway, two intersecting
major arterial throughways, a street canyon, and a complex source
typified by a sporting event. The major arterial throughway assumed
is that used in the January 11, 1974 paper, "Estimated Changes in
Human Exposure to Suspended Sulfate Attributable to Equipping Light
Duty Motor Vehicles with Oxidation Catalysts." It is a heavily
travelled 10-lane expressway with a daily flow of 233,000 vehicles and
a peak flow of 20,000 vehicles per hour. (J.F. Kennedy in Chicago
carries 267,000 vehicles per day and the Santa Monica Freeway in Los
Angeles carries 226,000). The intersecting freeways are assumed to be
8 lanes, each with 2,000 vehicles per hour for 4 lanes each and 1,000
vehicles per hour for each of the other 4 lanes. Adverse meteorological
conditions, E stability (slightly stable), and 1 M per second wind
velocity are assumed. The expressway street canyon employs an 8 lane
expressway with 2,000 vehicles/lane/hour in 4 lanes and 1,000 vehicles/lane/
hour in the other 4 lanes. Vertical-walled, flat-topped buildings with a
2 M per second wind at 90° are assumed. The sporting event used to project
a complex source exposure assumes a parking area and sport center 600 M
by 700 M adjacent to two intersecting 4 lane streets. The parking lot is
assumed to contain 3,000 vehicles which exit via 3 exit roads onto the
two 4-lane streets in one hour. These vehicles are imposed upon a flow
of 100 vehicles/lane/hours on the two 4-lane streets. Light winds
(1 M/Sec.), neutral (D) stability, and a 1,000
-------
-15-
M mixing height are assumed. In all cases, except the complex source, sul-
furic acid emission factors based on the highway fuel economy cycle are
used. For traffic inside the sports complex,'the Federal Test Procedure
is considered more representative.
The above exposure estimates are, as noted, based upon dispersion
model estimates for localized concentrations. Perhaps a more meaningful
exposure assessment is the assumed activity model estimate presented in
the above referenced January 11, 1974, paper. This model assumes that a
commuter spends two hours daily on busy arterial throughfares, one hour
in a street canyon or complex source, thirteen hours at home, and eight
hours at work. Based upon the model parameters presented in that paper,
the following exposure projections (Table 5) were prepared assuming
sulfuric acid emissions factors represented by the Highway Fuel Economy
Test Procedure.
While these projections must make assumptions regarding meteorological
conditions, vehicle driving patterns, vehicle densities, and human activi-
ties, there dees exist a limited amount of data gathered on and near real-
world roadways that have been tabulated in Table 6. Sulfuric
acid exposures are projected based upon the CO-sulfuric acid emission
factor relationship. This, of course, assumes no CO contribution from
any but automotive sources. This is likely a reasonable estimate in
many of the study cases. Additionally, CO and S02 surrogates based upon
the Los Angeles Catalyst Program Freeway Sampling Station are included.
These values are corrected for backgrounds contributed by off-roadwav
sources.
It is evident that California vehicles emit greater amounts of
sulfuric acid and the resultant exposures, are, therefore, higher. This,
in part, is due to the fact that Southern California gasoline contains
more sulfur than the national average gasoline (Table 3). Should Califor-
nia gasolines be reduced in sulfur content to national averages, the
emission factors for the 1975/76 California Interim Standards would be
the same as the National Statutory Standards Emission factors. A comparison
of exposures resulting from maintenance of the National Interim Standard
through the 19S1 model year versus a movement to the statutory HC/CO
standards for the 1977 model year indicates that sulfuric acid exposures
would be increased about S0:i over the levels associated with the continuance
of the National Interim Standards if catalysts are continued in use without
efforts to reduce sulfuric acid emissions. We would expect; no difference
in sulfuric acid emissions from vehicles on a national basis whether at the
California Interim Standard or the Statutory HC/CO Standard.
-------
-16-
Table 4
Dispersion Model Projected Incremental Peak Hourly Exposures
to Catalyst Generated Sulfuric Acid for Specified Vehicular
Sources, Sulfuric Acid Emission Factors, and Receptor after
2, 4, 7, and 10 "ciel Years
Source
10 lane exoressway
Adverse reteorolooy.
Worst wind angle
10 lane expressway
Normal meteorology
Intersecting
8- lane expressways
Adverse meteorology:
Worst wind angle (40 MPH)
Same
Expressway
Street canyon
Sporting
Event
(one hour)
Sulfuric Acid
emission factor(HC/CO)
Emission Standard^
,03 (49 state)
.08 (Calif.)
.05 (Stat)
.03 (49 State)
.03 (Calif.)
.OS (Stat)
.03 (49 State)
.08 (Calif.)
.05 (Stat)
. Same
.03 (49 State)
.08 (Calif.)
.05 (Stat)
.01 (49 State)
.025(Calif.)
.02 (Stat)
above for
lot area
plus
.03 (49 State)
.08 (Calif.)
.05 (Stat)
(perimeter
streets)
Peak
hourly incremental
exposure - H2SQ4
after:2
2 years
5?
740
52
3
8
3
25
66
25
5.5
14
5.5
7
19
7
6.4
16
6.4
6.6
16.5
6.6
4 years
104
280
140
6
16
8
50
132
66
11
29
14
»
14
38
20
12.8
32
19.2
13.2
33
19.7
in ugm/m
7 vears 10 years
156
420
227
9
24
13
75
198
i
107
16.5
44
23
21
57
33
19.2
48
32
19.8
49.5
32.8
208
560
315
12
32
18
100
265
148
22
58
32
28- '
'76
46
25. -6
64
44.8
*
26.4
66
45.9
ReceotorS
Pedestrian
near
expressway
Same
Pedestrian
near
expressway
500 meters
downwind of
Intersection
Resident
100 meters
from each
expressway
downwi nd
Pedestrian
next to
expressway
Parkinq lot.
no perimeter
"sTreets
contribution
Edae of
parking lot
with
perimeter
streets
contribution
State - 1975/76 National Interim Standards
Calif - 1975/76 California Interim Standards
Stat ' - Statutory HC/CO Standards
1977 on as5~-.es 49 State far 1975-76
22.4J, and 10 years represent approximately 25,50,75 and 1001 vehicle miles traveled by vehicle age.
324-hour average exposures at this specified receptor would be approximately 0.125 of the peak hourly values
^California emission factors are approximately the same for Interim California and' Statutory Standards
5c.», r
-------
-17-
Figure 1
Dispersion Model Geometries
-^ A
IO LA us X'iv '-y
i
«
IOt>M
U.Mt r-r
f
/,
PI R-
S LA:JS
oKT COMPLEX
-------
Table 5
Predicted 24 Hour Incremental Exposures to Sulfuric Acid Emitted
from Catalyst-Equipped Vehicles Assuming Specified Human Activity
Model for an Urban Resident
Assumed <,
Receptor
Commuter
living near
arterial
throughway
Commuter
living away
from arterial
throughway
Sulfuric Acid
Emission
Factor^
,03 (49 State)
.08 (Calif.)
.05 (Stat)
.03 (49 State)
.08 (Calif.)
.05 (Stat)
3
Incremental 24 hour Sulfuric Acid Exposure In vgm/m
2 model vears
normal
met.
.8
2
.8
.6
1.6
.6
adverse
met.
5
13
5
4
10.7
4
4 model vears
nonral
meto
1.6
4.3
2.1
1.2
3
1.6
adverse
met.
10
26
13.8
8
21
10.7
7, model 1 wear?
normal
mete
2.4
6.4
3.4
1.8
4.8
2.6
adverse
met.
15
40
22.6
12
32
17.4
10 model vears
normal
met.
3.2
8.5
4.7
2.4
7
3.6
adverse
met.
20
54
31.4
16
43
24
co
^Assumes 2 hours on expressway, 1 hour street canyon or complex source, 13 hours at home, and 8 hours at work.
2Assume Table 3 emission factors. 49 State = 1975/76 National Interim Standards, Calif. = 1975/76 California
Interim Standards, and Stat = Statutory HC/CO Standards. Stat. assumes 49 State Interim for years 1975-76.
-------
Table fi
Ca
rbon Won oxide -- Sulfur'c Acid 5u.rron.ile C.ssed Upon Reported fio-xlway S-jinp-ling Studies.
Four Years of Catslyst-Djiiipped Cars Assumed (W, VMf)
Usino lli'ihv/av Fii"1 Eoonoiw Emission factor; and for ?. Pni'.sion -Scenarios
r97f-79 at: 49 State Interii'i/1975- 76-49 <^af.? 1977-7)! Statutnrv
Road Type/
Hare
Transrwniiatton
'Expressway - NYC
12 Lane-,
West 40th Street -
'IYC - 1 tone:
Citv Struct -'.
"4 L3i;°s
VidJ'JCt Street
5 twines
At Grade
' 8 Lanes
Snallow Cut
6 Lar>es
Coc-i- Cut
6 i.<^>
Dc'jljle Cantilever
6 Lans'»
Los Angeles
Catalyst Freeway
Site
Values shown r>
* interin itandar
2 vears & 49 St
Maximum
Vehicles
Per Hour
14.200
890
. 1 .500
7 ,000
4,700
6 .400
7.600
12,000
present 4 catal
K/2 years 49 i
ite (.03) and 2
Average
Vehicles
Per Day
160,000
)
8,640
28 .000
ion",ooo
55.030- -
107,000
1 1-8 .000
100,000
-I/O, QUO
/st model year:-
tate then 2 ycj
years at Statu
Ave rage
Speed
(MPH)
46
15
16
38
43
3!!
48
45
for two hvdr
rs statutory.
Lory (.05) -
Assumed CO
Emission
Factor (gin/mile)
21
75
75
.26
21
26
21
22
21
ocarbon/CO emission co
.04 average h>SO,, emis
Assumed
Sulfuric Acid
EinisMxpn factor
iS"'/M)r
.03/.04'
.03/.04
.03/.04
.037.04
'.fl3/.04
.037.04
.n?/.n4
.03/. 04
.us -
(Calif. IIFE7 Emis-
sion Factor-)
trol scenarios; i.e.
ion factor. (Re*"- T
C(|uivalcnt Sulfuric Acid
Exposure ugni/M3
Max.
liour
88/ 117
11.3715
14/IR.7
lli/22
27/.15
76/101
li6/74
33/44
30.4
, 4 years ;
able 3)
24-Hour
Avg.
21/27
2.4/3.2
3.6/4.8
6/7.6
6/8.3
IB/24
'4/19
12/16
t 49 state
Other
-
9.3/ITi max, hr.
2/3 24 -hr. avg.
3rd fl(X>r outside
2/3 24-hr, avy.
sidewalk
-- -- - -
. '
4/5 24-hr.
'.iny sidewalk
-
.
5 6 -month avg.
10 monthly avn. -
wax. lir. of day
12 24-hr, max.
based upon SO?
Remarks
A! r ri ohts
uui 1'Jing
Street
Canyon
-------
-20-
Other Non-Regulated Emissions
The issue of sulfuric acid emissions from oxidation catalysts
has triggered a far more sophisticated concern for the non-regulated
emissions issue than was evident even in the fairly recent past. While
such emissions from catalyst systems has, of course, been the princi- .
pal focus such studies have been expanded in the area of potential
alternate power plants to the current gasoline reciprocating engine.
Emissions of platinum, palladium, alumina, carbon di-sulfide, hydrogen
sulfide, phosphine, and carbonyl sulfide have been reported or strongly
suspected from various catalysts systems. Of greater past research
interest has been the emissions of particulates, lead, polynuclear
aromatic hydrocarbons, aldehydes, and phenols.
Platinum, palladium, and alumina (catalyst support material) were
identified by a few laboratories in exhaust particulate from earlier
prototype catalysts. While slight increases in alumina are seen with
current production systems, the emissions of platinum and palladium are
at least below detectable limits.
Hydrogen sulfide and phosphine have been identified in the exhaust
from oxidation catalyst-equipped vehicles. As these .ire reduced
species, they are apparently formed when insufficient oxygen is present
in the exhaust passing the catalyst. This can occur in catalyst-equipped
vehicles, primarily certain 49-State Interim Standard Vehicles which do
not use air injection when the vehicle is started cold and operating under
choked conditions. It could also occur in some vehicles if the air
injection pump fails. The principal concern has been hydrogen sulfide
emissions due primarily to its disagreeable odor which is easily detected
at very low concentrations. As a result of related consumer complaints,
EPA requested information of this subject from the auto manufacturers on
November 7., .1974. Responses varied, but none indicated surprise that under
certain conditions hydrogen sulfide would be exhausted. Corrective action
usually involved carburetor or air pump adjustments. The most extensive
response was from Toyota who had initiated an odor study in 1973 during
their catalyst development program. Their research suggests the possibility
that the odor is not uniquely hydrogen sulfide which is consistent with
some of the ORD emissions research personnel's suggestion of carbon di-sulfide
and/or carbonyl sulfide. Current data suggests that reduced species,
probably odorous, are likely from some catalyst-equipped vehicles under
certain operating conditions. Our research regarding such emissions is
underway and progressing. Fortunately, most of these compounds can be
detected via their odor at concentrations well below dangerous levels. How-
ever, all are highly toxic in sufficient concentrations. Reported emissions
rates of hydrogen sulfide (40 ppm maximum) and phosphine (1 ppm) are low
enough that exposures to levels sufficient to insure adverse health effects
is not likely except, potentially, in highly confined spaces in which case
-------
-21-
carbon monoxide levels would also pose a health threat.
Substantial decreases in the emissions of lead, parti cul ate,
aromatic Jivd_rocarbpiis , phenols, and~ aldehydes are achieved
caTalysTs" (lies«f"provi dlTpubl i c hTalth benefi ts .
Emissions of n on T regulated, pollutants from diesel powered passenger
vehicles has 'recently begun to receive increased attention. Preliminary
data from EPA programs suggests that certain non-regulated emissions from
diesel s should be viewed iii the cpntext of the catalyst-generated non-
regulated emissions which have been of principal concern herein. In-
summary, diesel powered cars have low regulated emissions and good fuel
economy. Hov
-------
-22-
Health Effects
Particulate Sulfates and Sulfuric Acid;
While significant progress -has been made in control of sulfur
dioxide ($03) emissions, recent health'studies suggest that the degrada-
tion products of 502, namely sulfuric acid aerosols and-participate
sulfates which are formed in the atmosphere, may be more potent irri-
tants than S02 itself. Animal studies have shown that sulfuric acid
and sulfate compounds were much, more-potent irritants than 502 gas alone.
Additionally, these studies suggest that the .mass weight of sulfates is
is an insufficient basis upon which to predict irritant p'otency; parti-
culate size, chemical composition,, and temperature determine the tox.ic
potential of particulate sul fates.. Effect on the respiratory system^in
animals has been observed at sulfuric acid concentrations of 30 ygm/M3.
Healthy human volunteers exhibited rapid increases in respiratory rate
when exposed for even short durations (15 minutes) to sulfuric acid aerosol
concentrations of 350 to 500 ygm/M3.and immediate irritation of the throat
and nose at concentrations of 1100 ugm/M3. S02.,.on the other hand,'will
not cause such symptoms in most humans even at concentrations exceeding
10,000 ygm/M3.
More recent reports based upon epidemiological studies carried out
as part of the EPA Community Health and Environmental Surveillance
System (CHESS) indicate that adverse health effects in communities may
be more closely associated with exposures to suspended particulate sulfate
than to other pollutants. A monograph summarizing the results.of CHESS
studies in New York City, Utah, and Chicago was published in June 1974.
A second series of individual research reports summarizing .results from
the eastern U.S. and an additional year's.effort'in the New York metropolitan
area is being prepared as a second monograph. In general, these studies
confirm the adverse effects of air pollution and particulate sul fates
previously reported.
The health parameters in the studies reported are chronic respira-
tory disease, lower respiratory disease, pulmonary function, acute
respiratory disease, irritation of asthmatics, and irritation of symp-
toms reported by the general population during an acute air pollution
episode. From these and other studies, the following best judgment
threshold concentrations for selected adverse health effects due to
suspended sulfate particulate exposures are projected (Table 7).
-------
-23-
Table 7
Threshold Concentrations of Suspended Sulfate Particulate .
Exposures for Selected Adverse Health Effects (Best Judgment)
Threshold Concentration of Suspended
Adverse Health Effect Sulfates and Exposure Duration
Increased Daily Mortality 25 ygm/M3 for 24 hrs, or longer
Aggravation of Heart and Lung 9 pgm/M3 for 24 hrs. or longer
Disease in the Elderly
Aggravation of Asthma 6-10 ygm/M3 for 24 hrs. or longer
Excess Acute Lower Respiratory 13 ygm/M3 for several years
Disease in Children
Excess Risk for Chronic 10 vgm/M3 for up to 10 years
Bronchitis (non-smokers)
PJatinum and Palladium:
The reported health effects studies, limited as they may be, have
focused upon human health effects due to exposures to soluble platinum
compounds. Such exposures occur principally in platinum refineries, for
which a TLV (threshold limit value for occupational exposure) of 2 ugm/M^
has been established. Inhalation exposures to this material, and other
soluble salts .of platinum, are reported to result in "platinosis ," a disease
the symptoms of which are similar to asthma. Exposures to insoluble plati-
num salts or the metal are generally considered to be safe, although only
extremely limited work in this area has been conducted. Related studies
of palladium are essentially nonexistent.
A substantial toxicological research program focusing on platinum
and palladium was initiated by EPA in early 1974 following preliminary
data from a number of laboratories which indicated that these metals
were emitted from prototype catalysts. Preliminary results from these
studies suggest the following:
1. Platinum, as a soluble salt, is a potent sensitizer and is
highly allergenic.
2. The population susceptible to platinum allergy is unknown.
3. Platinum and palladium can bo absorbed into biological tissues
of animals by various exposure routes.
4. Soluble platinum compounds are more toxic than lead when
administered orally while palladium is more so when adminis-
tered intravenously.
-------
-24-
5. Palladium apparently acts as a nonspecific cardiac irritant
as well as a peripheral vasoconstrictor in animals.
We have not, however, been able to detect platinum or palladium
emissions from production catalysts or to identify their compounds
emitted from earlier devices. Platinum has been shown .to methylate,
however, in preliminary studies. We are determining the stability
of such compounds at.the present .time.
While we do not now believe that platinum or palladium pose a
general emission - inhalation exposure potential, the longer term
potential risks associated with large scale introduction of these
metals into general use cannot now be assessed.
Hydrogen Sulfuie^ Phosphine., Carbon D1-sulfide, and Carbony.l Sulfide;
Hydrogen sulfide and phosphine have been identified in catalyst-
equipped vehicle exhausts. Carbon di-sulfide and carbonyl sulfide have
been suggested but not identified although they have been seen in exper-
imental reduction catalyst studies.
Hydrogen sulfide in high concentrations (500-1000 ppm) acts pri-
marily as a systemic poison causing unconsciousness and death through
respiratory paralysis. At.lower concentrations it acts as a respiratory
irritant (50-500 ppm), while at low concentrations it affects the eyes
(5-20 ppm). Low concentration effects also, include nervousness, cough,
nausea, headache, and insomnia, .it can be easily detected at very low
concentrations (<1 ppm) by the o.lefactory senses which, however, are
rapidly desensitized at high concentrations'. While HoS odor has been
identified in the exhaust of a limited number of catalyst-equipped cars,
concentrations in undiluted exhaust would be expected to be at or below
the TLV for normal fuel sulfur levels.
Phosphine is a toxic gas with a TLV of 0.3 ppm. Exposure symptoms
include diarrhea, nausea, vomiting, tightness of chest, cough, headache,
and dizziness when exposed.to levels averaging below 10 ppm. Death has
been reported after 1 to 2 hours exposure to concentrations of the
order of 8 ppm. There are apparently no cumulative effects except the
longer range possibility of chronic phosphorus poisoning. There are
scattered reports of qualitative identification of PH3 in catalyst car
exhaust but current regulations on phosphorous levels in unleaded gasoline
should prevent levels approaching the TLV even in undiluted exhaust.
-------
-25-
Automotive Sul fate . Control Options
introduction:
Approaches which could be used to limit sulfate emissions from
automobiles fall into three. basic. classes: limitations on the quantity
of sulfates emitted from individual vehicles, and limitations on .the
sulfur content of motor vehicle fuels and limitation of the percent of
automobiles employing catalysts. The EPA currently has regulatory
authority to carry out the first two types of approach. Section 202(a)
of the Clean Air Act permits the establishment of emission standards
for new vehicles, while Section 211 (c) permits the regulation of motor
vehicle fuel composition. The auto industry may choose the third ap-
proach in order to avoid sulfate emission regulation.
At present3 information on the feasibility, costs, and time re-
quirements of various approaches to reducing automotive sulfate emissions
is quite limited. Some such information was submitted to EPA In response
to the Federal Register Notice of March 1974 seeking information an auto-
motive sulfates. More recent submissions to EPA on the status of auto
manufacturers' emission control development programs indicate that their
efforts in the area of sulfate control are quite limited. EPA currently
has studies in progress which are relevant to both vehicle emission
limitations and fuel sulfur limitations.
Vehicle Emissions Limitations;
Automotive sulfate emissions could be limited through imposition of
a light duty vehicle emission standard for sulfates. Implementation of
suchyan emission standard would require the development of a standardized
test procedure which could be used on prototype vehicles on a routine
basis during annual emission' certification and which might also be used
in various field surveillance or enforcement audit tests of production
vehicles. This approach would also require the establishment of a
level for the emission standard below which automotive sulfate a&ilssions
are considered to be acceptable. Most importantly, implementation of
such an emission standard would require the availability of satisfactory
control technology to permit the simultaneous achievement of the sulfate
standard and the existing and future standards for HC, CO., and NQX,
-------
-26-
--- Approaches which might be taken by auto manufacturers to reduce
sulfate emissions can be categorized into three classes:
a) Adoption of non-catalyst emission control technology
sfor most models
i
b) Modification of catalyst system design to reduce
sulfate formation
c) Use of sulfate traps
Available information on the feasibility, cost, and time required for
implementation of each of these approaches is discussed briefly below.
Adoption of Non-Catalyst Emission Control Technology
Oxidation catalysts have been adopted by auto manufacturers because
they represent a. highly effective and relatively inexpensive approach to
reducing HC and CO emissions from gasoline-powered reciprocating internal
combustion engines. Use of catalyst technology to clean up engine
exhaust has also permitted manufacturers to reduce their reliance on
certain types of engine modifications previously used for HC and CO control
,and to tune engines for improved fuel economy.
Nevertheless, various non-catalyst approaches also have either
demonstrated capability or promise of meeting current interim and planned
statutory emission levels for HC and CO without the use of oxidation
"catalysts, assuming that a NOX "standard no more stringent than 1.5 - 2.0
c|/iui is requi/ed. The technological status of these non-catalyst alternatives
is discussed more fully in the annual "State of the Art" report on automotive
emission control technology prepared for use in the standards suspension
consideration. However, the principal alternatives and some of their more
important characteristics are summarized in Table 8.
the possible exception of the diesel engine, the limited data
able on sulfate emissions from non-catalyst vehicles, including
with CVCC stratified charge and rotary engines, indicate that
minimal formation of sulfate can be expected. Implementation of non-
catalyst technology on most new cars is not likely to be possible before
the 1979 model year if either the 1975 California interim standards or
1977 standards for HC and CO apply then. However, non-catalyst technology
need not be applied to all cars in order to avoid sulfate concentration
problems.
-------
TABLE c
Non-Cat a lyst Emission Control. Alternatives
Approach
"Lean Burn" Engine
Modifications
Conventional Engine
with Rich Thermal
Reactor
CVCC Stratified
Charge Engine
Diesel Engine
Emissions
(g/mi)
less than
0.001 ,
less than
0.001
less than
0.001
'
less than
0.001
Earliest
Widespread
Availability**
1979
1979
Beyond
1980
Beyond
1980
Beyond
1980
Fuel Economy
Impact
HC/CO » HC/CO «
0.9/9.0 0.41/3.4
None
10%
loss
None
20%
gain
15%
loss
10%
ioss
20%
loss
10%;
loss
20%
gain
20%
loss
First Cogt
Impact
HC/CO = HC/CO *>
0.9/9.0 0.41/3.4
$160
'less
None
$50
less
$125
more
None
$90
less
$90
less
$140
less
$35
more
$90
.'less
Other Remarks
Could use leaded
gasoline.
Could use leaded
gasoline. Possible
particulate emission
problem.
Could use leaded
gasoline.
Requires distillate
fuel; desulfurizatio
may be needed. Odor
emissions need more
study.
t
i
Could use loaded £
gasoline. PNA
Rotary Engine with
Thermal Reactor.
i , . - emissions need tnorc^
*Rei-atiye to conventional engine with oxidation catalyst in 1979 controlled to specified HC/CO leveV, V.e., diesel at*"
.41/3.4 riC/CO would cost $35 more than catalyst-equipped conventional engines at these levels
**EPA estimates.
-------
-28-
The situation for diesel engines is much less clear. One study
has reported sulfate emissions of about 0.02 g/mi from a light duty diesel
vehicle, comparable with the levels for some catalyst-equipped cars. Other
studies have not found such high levels. While the small number of light
duty diesel vehicles currently in use in the U.S. would not be expected
to have a substantial adverse impact on ambient sulfate levels even if
the higher value is correct, further quantification of light duty diesel
sulfate emissions (as well as other unregulated emissions from the diesel)
is necessary before any effort is made to encourage widespread adoption of
the diesel engine as an alternative to using oxidation catalysts.
Modification of Catalyst System Design to Reduce Sulfate Formation
The efficiency of a given oxidation catalyst in converting sulfur
dioxide in engine exhaust to sulfuric acid should depend upon a number
of design characteristics of the engine-catalyst system. The physical
and chemical characteristics of the catalyst material, the catalyst
operating temperature, the amount of oxygen in the exhaust stream enter-
ing the catalyst, and the amount of residence time which the exhaust gases
spend in contact with the catalyst should all influence the degree of sulfate
formation.
These same factors would also be expected to influence the extent
to which the catalyst performs its intended function of oxidation of HC
and CO. Changes made to reduce sulfate formation would generally be
expected to result in soins loss of HC and CO control, making it more
difficult for auto manufacturers to achieve required levels of HC and CO
with a possible adverse effect on fuel economy. Thus, possible modifications
to reduce sulfate formation need to be. considered in the context of their
interactions with HC and CO control and fuel economy. ,
t.' '
Catalyst type (pelleted vs. monolith) has a significant effecfron
sulfste emission rates during certain types of driving. The differences
appear tc result from the much greater ability of pelleted catalysts to
store and later release sulfates than on any difference in the ability
of the two catalyst types to form sulfates. While pelleted catalysts exhibit
much lower sulfate emission rates than monoliths during low speed driving,
such as the Federal emission test procedure (FTP), pelleted catalyses are
also characterized by much higher emission rates than monoliths during the
transition from intervals of low-speed driving to high-speed driving, when
previously stored sulfur compounds may be released. At this time, ;it does
not appear that pelleted catalysts hold any major overall advantage over
monolith catalysts in terms of reduced sulfate impact. \
-------
- 29 -
The limited experimental data presently available do not indicate that
changes in catalyst type (Pellet vs. Monolith), amount of noble metal used,
catalyst operating temperature, or exhaust residence time are likely to achieve
major reductions in sulfate emission rates while achieving reasonable levels of
HC and CO control with the catalyst.
However, there is very limited experimental data which suggests that
different catalyst formulations may reduce sulfate emissions at a given level
of HC and CO. The control of oxygen content of the exhaust also appears to
show significant potential as a sulfate control technique. Very limited
experiments have shown that elimination of injection of air.into the exhaust
stream except during engine warm-up can achieve reductions in sulfate emissions
of 50% or more. A substantial loss in CO control accompanied the sulfate
reduction, but the final HC and CO levels were at least still well below the
1975 California interim standard levels. More sophisticated types of air
injection modulation might allow substantial sulfate reductions with less
loss of CO control, but these approaches have not yet been investigated. If
modulated air injection is found to be a feasible sulfate control approach,
it should be relatively inexpensive (about $20) approach which would be avail-
able for implementation as early as the 1978 model year.
An alternative approach to controlling oxygen concentrations in the catalyst
would be the use of an oxygen sensor in the exhaust providing feedback control
of a fuel injection system. This technique, used in the 3-way catalyst, is
being looked at as a candidate for simultaneous control of HC, CO and NOX.
This system is receiving particular attention in Europe and was discussed in
detail in the recent NAS report on emissions control devices.
An important consideration in any further development of oxygen level
control techniques for sulfates should be thorough characterization of other
unregulated pollutants, particularly other sulfur compounds. The relatively
limited problems with hydrogen sulfide odor from catalyst-equipped cars now
being reported could be magnified greatly if cars were designed to operate
with very low excess oxygen levels at the catalyst.
In addition to the factors discussed above, reductions in vehicle weight
and aerodynamic drag implemented to improve fuel economy will also reduce
sulfate emission rates. For example, the joint DOT-EPA fuel economy study
concludes that an increase in automotive fuel economy from an average of
15.0 mpg in 1975 to 17.3 mpg in 1980 is feasible through shifts to smaller
cars. Such a shift would reduce average sulfate emissions per vehicle by
about 10%, independent of any changes in catalyst system design.
-------
-30-
Use of Sulfate Traps
Preliminary testing conducted under an EPA contract has shown that
traps which absorb sulfuric acid through chemical reaction with a solid
trapping material show promise as a sulfate control technique. A proto-
type trap Was tested over 25,000 miles of use and showed consistent
reductions in sulfate emissions of 90% or more during all types of driving.
Furthermore, particulate emissions resulting from attrition of the sorbent
material were found to be minimal.
: The prototype trap exhibited one major flaw: excessive pressure drop
which increased with use of the trap. The contractor is currently
screening other possible trap materials and investigating ways of re-
ducing the pressure drop while maintaining high trapping efficiency.
Unfortunately, the auto manufacturers' annual submissions on their progress
in light duty vehicle emission control indicate that no effort to develop sulfate
traps is being made by the industry.
If the manufacturers were to initiate a major effort to develop sulfate "
traps it is unlikely that they could be available for installation prior to
the 1979 model year, at the earliest. The cost of a sulfate trap should be
somewhat less than that for a pelleted catalyst, probably in the range of $50 :
to $100 per vehicle. If trap replacement is needed periodically (say at
25,000 mile intervals) the lifetime cost would be higher.
V Implementation of a Sulfate Emission Standard
As noted above, implementation cf non-catalyst technology at the 1975
California interim or 1977 statutory HC and CO control levels could probably
not be widespread prior to the 1979 model year. Implementation of a modulated
air-injection approach to controlling sulfate formation is not likely before
the 1979 model year at the earliest. All of these estimates reflect develop-
ment and production lead times only and assume that manufacturers begin a
substantial effort to reduce sulfate emissions immediately. In fact, manu-
facturers are likely not to begin major development efforts until after a
sulfate emission standard is promulgated.
Based upon the present limitations of automotive sulfate test procedures
and the minimum time required to develop and adopt regulations, a decision
now to proceed with a sulfate emission standard would not result in final
regulations being promulgated before early 1976. A delay to mid-1976 is
quite possible.
-------
-31-
Once regulations are promulgated, both EPA and industry laboratories
will require some period of time to procure and install the new test
equipment needed arid to develop competence in the more complex emission
tests involved. Independent of the availability of suitable control
technology, certification testing under a sulfate standard could not ...
carried out before the 1979 model year.
In summary, a major effort to reduce sulfate emissions voluntarily,
started now by auto manufacturers, might be capable of achieving
significant reductions in sulfate emissions beginning with the 1978 model
year. It is more likely, however, that significantly reduced sulfate
emissions will not be possible through vehicle emission, limitations prior to
the 1979 or 1980 model year.
Fuel Sulfur Limitations:
Available vehicle test data generally indicate that the rate of
sulfate emission from catalyst-equipped cars decreases linearly with
reductions in gasoline sulfur content. Therefore, a 50% reduction in
the average sulfur content of unleaded gasoline, for example, would
result in a corresponding 50% reduction in sulfate emission factors.
Reducing gasoline sulfur levels has the advantage over vehicle emission
.limitations of acting on the whole population of catalyst-equipped cars
oh the road at the time control is instituted, whereas vehicle emission
limitations will impact only on new catalyst-equipped cars sold after
the effective date of the limitation. Thus, for example, achieving a
50% reduction in fuel sulfur level would have a larger immediate effect
in reducing sulfate exposures than would achieving a 50% reduction in
sulfate emissions from new cars through changes in vehicle technology.
Section 211(c)(l)(A) of the Clean Air Act authorizes EPA to regulate
the composition of motor vehicle fuels if emission products resulting
from use of those fuels endanger the public health or welfare. This is
the same authority that has been used to adopt the lead phase-down re-
gulations. The Act. specifies that as a precondition to adopting such
regulations, the EPA must consider all relevant medical and scientific
evidence available, including consideration of other technologically or
economically feasible means of achieving the needed emission reductions.
Thus, prior to adopting regulations requiring a reduction in the sulfur
content of unleaded gasoline, EPA must establish that reduced sulfate
emissions are necessary to protect the public health or welfare, and that
control of such emissions through vehicle modifications is.either not
technically feasible or is substantially more costly than reducing fuel
sulfur levels.
-------
-32-
If faced with a requirement to reduce gasoline sulfur content,
petroleum refiners have two basic types of approaches available to
them. One is to blend the lower sulfur content components available
from the refinery into the unleaded gasoline v/hile using the higher
sulfur materials in other gasoline grades. The second is to install
additional processing steps to remove sulfur from various refinery
streams. These two approaches, which differ substantially in time
and extent of applicability, are discussed further below.
Blending and Allocation Approaches
As produced by a modern petroleum refinery, gasoline is not simply
a natural product separated from other components of .crude oil. Rather,
different grades of gasoline, are "assembled" by the refinery by bjending
together in suitable proportions the products of various chemical pro-
cessing steps. Some of these materials have a sulfur content of less
than 10 parts per million (ppm) while others may be as high,as 1500 ppm.
In most parts of the U.S., the final blend typically has a sulfur content
of about 300 ppm.
By selectively blending the lower sulfur content materials into the
unleaded gasoline product, a refiner should theoretically be able to
achieve major reductions of the sulfur content of that gasoline. Of
course, the sulfur level of leaded grades would increase accordingly,
but the minimal conversion of fuel sulfur to sulfates by non-catalyst
cars which would use the leaded grades makes this of no concern from an
air quality standpoint.
Practically, the refiner is limited in the extent to which he can
carry out this process. First, the components that are low in sulfur
are high in octane. These high octane components must be used to some extent
in the leaded gasolines to provide satisfactory octane quality. In fact,
the institution of the lead phase-down regulations would increase, the need
for these low sulfur components in leaded grades. This is further compounded
by the fact that high sulfur components do not achieve as large an octane
boost from adding lead as do low sulfur materials. Thus, there is an
incentive to the refiner to use the low sulfur components in the leaded,
rather than unleaded, grades. Finally, the refiner is Vimited in the extent
to which he could reduce unleaded gasoline sulfur levels through blending by
the relative demand for unleaded and leaded gasolines.. As more and more
cars on the road demand unleaded gasoline, and fewer leaded, the refiner
must begin to use more of the higher sulfur materials in the unleaded grade
in order to produce the needed amount of that product.
-------
-33-
< Based on comments received from refiners, it appears that a maximum
o.f about 20% to 30% of all gasoline produced could be unleaded of low
(100 ppm) sulfur content, but this would not necessarily be geographically
distributed according to the need for such fuel. Depending on the extent
to which sales of new cars pick up from their current low rate, and '
assuming that virtually all new cars sold beginning with the 1975 models
require unleaded gasoline, selective blending could provide an adequate
Supply of low sulfur unleaded gasoline approximately through calendar
year 1977. . The cost to refiners of implementing such a blending scheme
has not yet been quantified, although it would be quite small in com-
parison with installing new fuel desulfurization capacity.
An extension of the blending approach would be allocation of low
sulfur unleaded gasoline preferentially to those geographic areas where
automotive sulfate impacts would be largest. Some reallocation of low
sulfur products would probably be needed to provide areas such as
Southern California with adequate supplies of low sulfur gasoline even
through 1977; more extensive allocation efforts could keep up with new
car demand for unleaded gasoline in key areas somewhat longer. An EPA
review of the extent to which allocation would be feasible is in progress
and should be completed by May 1975. Contacts have also been made with FEA,
whose allocation authority would probably be required to implement such a
program.
The implementation of a blending and allocation scheme would involve
several steps. First, the details of the allocation plan would have to be
developed. This would rely on the EPA review now in progress and would
probably not be completed before late summer of 1975 at the earliest.
Then, some effort to convince refiners to implement the scheme would be
needed. If this v/ere to take the form of regulations under Section 211 of
the Clean Air Act and under suitable FEA authorities, promulgation of such
requirements before mid-1976 is unlikely. Alternatively, a voluntary
compliance approach might be considered, but at a minimum this would seem
to require pricing action by FEA to allow refiners to recover the added
costs of blending and shipping to target areas the low sulfur gasoline.
Finally, refiners would require some time to implement the new procedures.
Actual blending of low sulfur gasoline should be possible quite quickly
(probably less than 6 months), but procurement of additional storage and
transport facilities might be needed to distribute the fuel to the target
areas. Overall, it appears unlikely that significant quantities of low
sulfur unleaded gasoline would be available in key target areas before
1977, and possibly not until 1978.
Desulfurization
While blending and allocation approaches could potentially provide
low sulfur unleaded gasoline for a few years, if low sulfur gasoline
is needed on a continuing basis the only alternative is desulfurization.
-------
-34-
EPA has conducted one contract study of the feasibility and costs of
implementing gasoline desulfurization and a second study to refine
those estimates is in progress. Another study has been conducted for
General Motors and several petroleum companies have made their own
analyses of this approach.
In general, it is agreed that the technology needed to reduce the
sulfur level of unleaded gasoline (at least to 100 ppm) is presently
available, and that it will take a capital investment of about $2-4
billion and a lead-time of about 4 to 6 years to put the petroleum
industry in a position to meet the demand of a full population of cars
requiring unleaded low sulfur gasoline. A cost increase of about 1-2
cents per gallon would be required to provide low sulfur gasoline and
an energy penalty of about 1% is typical of the estimates.
The 4-6 year lead-time estimate given above is from the time that
refiners make a commitment to install desulfurization capacity. In
view of the large investments required, it must be assumed that such
commitments will not be made at least until after final promulgation
of appropriate EPA regulations, and possibly not until any legal
challenges to the regulations have been resolved. Thus, desulfurization
is a long-range alternative which would not be expected to have much
impact until after 1980.
SUMMARY OF CONTROL OPTIONS:
No regulatory approach could be expected to have much effect on
the projected growth in the impact of automotive sulfate emissions before
the end of the 1976 model year. An interim measure in the form of
regulations requiring reductions in the average sulfur level of unleaded
gasoline based on fuel blending and allocation of those low sulfur fuel
supplies to the most critically impacted areas might be put into effect during
1977 if vigorous efforts to develop the needed regulations are initiated
soon. If very limited preliminary indications regarding the feasibility
of reducing sulfate emissions through control of vehicle air injection are
proven out, and if manufacturers initiate efforts to develop and adopt
such techniques quickly, significant (at least as an interim measure)
reductions in sulfate emissions could be achieved beginning with 1978
models. However, if emission standards are required to force such develop-
ment and implementation, no impact before the 1979 models is likely. By
1979, auto manufacturers might also be able to utilize sulfate traps or
non-catalyst emission control technology to meet sulfate emission standards.
If vehicle control technology for sulfates is found infeasible, and if
continued reliance on oxidation catalysts is necessary to achieve other
goals, gasoline desulfurization will be required. However, desulfurization
cannot be expected to have much impact until after 1980. Once implemented,
though,, desulfurization could relatively quickly reduce automotive sulfate
emissions from the entire car population.
-------
-3b-
Beriefit -- Risk Analysis
The National Environmental Research Center, Research Triangle Park
has prepared a detailed paper entitled "Estimates of the Public Health
Benefits arid Risks Attributable to Equipping Light Duty Motor Vehicles
with Oxidation Catalysts." The paper attempts to put into perspective
the benefit/risk aspect of equipping light duty motor vehicles with
oxidation catalysts by examining the public health impact of changing
emissions and changing air quality on the urban population of our nation.
It does so by calculating emission factors for a changing vehicle popula-
tion, projecting the impact of emissions changes on ambient air quali^,
estimating probable changes in human exposures, constructing dose-
response functions for adverse effects using best-judgment health
intelligence assessments, and estimating the public health benefits and
risks. While there are a number of major difficulties customarily
encountered when one attempts to develop dose-response relationships
linking environmental agents to adverse human health effects, the paper
does represent a best, and largely original, attempt to provide at least
rough assessments of such benefits and risks.
Benefits are derived from the reduction in exposures to carbon
monoxide and oxidants for which hydrocarbons are the precursors. -Risks
are assessed due to the increased exposure to sulfuric acid resulting
from the use of oxidation catalysts. A number of adverse health effects
are examined. A brief review of the results of this analysis, by health
effect, follows:
MojtaVUyj
One expects that reducing carbon monoxide exposures will reduce
premature deaths from myocarclio.l infarctions and that increasing particu-
late sulfate-sulfuric acid exposures will increase the risk of premature
death in elderly persons already afflicted with chronic heart and lung
disorders. A benefit-risk comparison suggests that the use of oxidation
catalysts should, after 10 model years, prevent a modest number of
premature deaths. Initially, benefits can be expected in large cities
in-both the eastern and western United States. Later, one can maintain
net benefits in large western cities but a net rjj>k_ for excess mortality
will" be created in the east. Most of the benefits attributable to
catalysts occur in the first 4 to 6 years when uncontrolled vehicles are
replaced with new catalyst equipped vehicles, while the risk associated
with catalysts increases linearly over time.
Aggravation of Asthma:
Photochemical oxidants and particulate sulfate - sulfuric acid can
aggravate asthma. This benefit-risk analysis indicates that any reduction
in the aggravation of asthma attributed to reduced oxidants will be
overwhelmed in all geographic areas by an increased risk attributable to
catalyst generated particulate sulfate-sulfuric acid. Statistically,
one would expect increased risk for asthma to begin rather promptly with
measurable adverse effects seen in cities of over 100,000 population in
the second to fourth catalyst model year.
-------
Aggravation of Heart and Lung Disease;
Elevated ambient levels of oxidants and participate sulfate-
sulfuric acid are thought to aggravate the symptoms experienced by
elderly persons with chronic heart and lung disorders. The benefit-risk
approximation indicates a net increase in the aggravation of heart and
lung disorders when all vehicles are equipped with catalysts. However,
one would expect most of the benefits in the far west and most of the
increased risk in the eastern United States. Net benefits are expected
for 4 to 6 model years in Southern California while only net risks are
projected elsewhere at any time.
Acute Lower Respiratory Disease in Children:
While particulate sulfate-sulfuric acid exposures are thought to
increase the frequency of acute lower respiratory disease in children,
the existing epidemiological studies have not yet been able to disentangle
such oxidant effects. EPA scientists, however, feel that measurable benefits
would follow the reduction of peak oxidant exposures in Southern California.
The analysis projects little or no catalyst-associated risk among children
in Southern California, but substantial risks for children in the eastern
United States.
Chronic Respiratory Disease Symptoms:
At the present time there is not a substantial body of laboratory
or epidemiologic evidence suggesting that either oxidants or carbon
monoxide constitutes a risk factor for chronic respiratory disease.
Such is not the case for particulate sulfate-sulfuric acid. The analysis
suggests that the use of catalysts will result in a substantial increased
risk for chronic respiratory disease. Again, almost all of the project
risk occurs in the eastern United States and will begin to occur only
after the expected exposures projected for 2 to 4 model years of catalyst-
equipped vehicles are maintained for several additional years.
Prevalence of Irritation Symptoms Among Otherwise Healthy Adults:
Eye irritation, transient cough, ill defined chest discomfort, and
headache are considered. Increases in the frequency of such symptons
after exposures to elevated levels of oxidants are well documented.
While no dose-response function is available for exposures to particulate
sulfate-sulfuric acid although increases in irritation symptoms are
hypothesized. The use of oxidation catalysts thus provides a net benefit
in so far as irritation symptoms are concerned, particularly in Southern
California.
-------
-37-
While there are a number of important caveats repeated in this
benefit-risk analysis paper, the conclusions are:
1. The introduction and continued sales of light duty motor
vehicles equipped with oxidation catalysts will probably
result in a net public health risk if no control measures
are instituted.
2. A large portion of the benefits projected will occur in
Southern California.
3. A large portion of the risks will be concentrated in, but
not limited to, the eastern United States.
4. Benefits will be greater for the first few model years as
uncontrolled vehicles are replaced by stringently controlled
vehicles.
5. On balance, public health risks will exceed benefits in all
areas of the country after 4 model years are equipped with
catalysts.
-------
Conclusions
1) Exhaust sulfuric add collection and analytical techniques have
been refined and specific procedures documented. The Federal Emissions
Test Procedure for regulated pollutants cannot be used solely to ascertain
sulfuric acid emission factors. Sulfuric acid emission factors have been
estimated. Vehicles being built to meet the National Interim Emissions.
Standards have lower sulfuric acid emissions than do these designed to
meet either the California Interim Standards or the Statutory HC/CO
Emissions Standards, for which comparable sulfuric acid emission rates
are expected.
2) Physical model exposure estimates suggest that on and near major
arterial thoroughfares in major urban cities the incremental sulfuric
acid exposures will exceed the health effects threshold a few days of
the year after 2 model years of vehicles are equipped with catalysts
in Southern California and after 3 to 4 model years are so equipped
nationally.
3) Best judgment adverse health effects thresholds for particulate
sulfates arevlO ugm/M3 for 24 hour exposures. These threshold values
are for susceptible persons with existing heart and/or respiratory
disorders. Threshold effects levels for sulfuric acid aerosols
are probably somewhat lower than these values. Short term, (1-2 hours)
thresholds are not available, but easily measured effects are observed
in young, healthy people after 15 minute exposures to concentrations
of sulfuric acid of 350-500 ugm/N|3.
4) Three basic control approaches are available to limit sulfuric
acid emissions: increased use of non catalyst technology, modifications
to catalyst systems to reduce sulfate formation or trap sulfates
before they are emitted, and reductions in gasoline sulfur levels.
Gasoline desulfurization cannot be expected to have much impact
before 1980 although as an interim measure gasoline sulfur reductions
through blending and allocation procedures -probably on a selective
geographic basis- may be possible. Control of vehicle sulfate emissions
through an emission standard would likely have no impact before the
1979 model year. At 49 State or California Interim Standard levels, or
at Statutory Standard levels, manufacturers are unlikely to eliminate
catalyst usage on a major proportion of their production before the
1979 model year.
-------
- 39 -
5} A public health benefit/risk analysis which weighs the benefits
associated with decreased carbon monoxide and oxidant exposures
against,the risks due to increased sulfuric acid exposures, suggests
that on a national basis health risks exceed the benefits after
4 model years are equipped with catalysts. It should be noted,
however, that this conclusion is based upon assumptions about dose
responses, and human exposure about which there still remain
uncertainties.
6) A number of unique non-regulated emissions in addition to sulfuric
acid have been identified and/or suggested from oxidation catalyst-
equipped vehicles. None appear to pose a short term health risk
at this time; however, we lack detailed emissions information in
many cases. We are not able, at this time, to estimate the potential
long term chronic health risks. Research is accelerating on these
issues.
------- |