EPA-650/2-75-068
June 1975
Environmental Protection Technology Series
A METHODOLOGY
FOR DETERMINING THE EFFECTS
OF FUELS AND ADDITIVES
ON ATMOSPHERIC VISIBILITY
U.S. Environmental Protection Agency
Office of Research and Development
Washington, 0. C. 20460
-------
EPA-650/2-75-068
A METHODOLOGY
FOR DETERMINING THE EFFECTS
OF FUELS AND ADDITIVES
ON ATMOSPHERIC VISIBILITY
by
W. C. Kocmond, J. Y. Yang, and J. A. Davis
Calspan Corporation
Buffalo, New York 14221
Contract No. 68-02-0698
ROAP No. 26AAE
Program Element No. 1A1002
EPA Project Officer: William D. Conner
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D. C. 20460
June 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environ men la I Research
Center - Research Triangle Park, Office of Research and Development.
EPA. and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:
]. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2. ENVIRONMENTAL PROTECTION TECHNOLOGY
3. ECOLOGICAL RESEARCH
4. ENVIRONMENTAL MONITORING
5. SOCIOECONOMIC ENVIRONMENTAL STUDIES
6. SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation , equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution. This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
This document is available to the public for sale through the National
Technical Information Service, Springfield. Virginia 22161.
Publication No. EPA-650/2-75-068
-------
ABSTRACT
A methodology for determining the effects of fuels and additives on
atmospheric visibility has been developed using the smog chamber approach. The
methodology involves measuring visibility in a 590m smog chamber after first
introducing auto exhaust at a 300:1 dilution ratio, adding 0.05 ppm S02 and
irradiating the sample for 23 hours. Three 5.7 liter 1972 Chevrolets and one
1973 catalyst-equipped 6.55 liter Ford Galaxie were used in the study. The
effects on test results of exhaust dilution ratio, relative humidity, added
SO , primary particulates, evaporative emissions and irradiation time are
discussed.
The tests show that using commercial grade indolene fuel, the effects
on visibility of the additives F-310 and CI-2 are small compared to the effects
brought about by variations in engine performance. The presence of primary
particulates plays an important role in the initial and final visibilities noted
in the smog chamber. Irradiations of particle-free HONO+S02 mixtures comparable
to those found in auto exhaust did not produce equivalent visibility losses.
The final visibilities noted in the smog chamber were found to be
closely correlated with the initial HC/NO ratio. The correlation for the
commercial grade indolene is so good that final visibilities can be predicted
from the initial measurement of HC and NO in the chamber. Data generated using
EPA reference fuel show a similar correlation of visibility with HC/NO ratio,
but the data lie on a different straight line (due to lower sulfur content of
fuel). For the fuels and additives tested at a given HC/NO ratio, the sulfur
content of the fuel appeared to have the most important effect on visibility.
11
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION 1
2.0 EXPERIMENTAL FACILITIES 3
2.1 The Calspan Smog Chamber 3
2.1.1 Instrumentation 7
2.2 Vehicle Emissions Research Laboratory (VERL) 9
2.2.1 Design of Vehicle Exhaust Sampling and
Introduction System 11
3.0 RESULTS AND DISCUSSION 12
3.1 Methodology Development - Phase 1 12
3.1.1 Vehicle Conditioning and Emission Analysis 14
3.1.2 Effects of Dilution Ratio, Relative Humidity,
Variations in Natural Air, Added 802 an<* Added
Hydrocarbon on Test Results 18
3.1.3 Repeatability Tests - Phase 1 25
3.1.4 Summary of Phase I Test Results 35
3.2 Methodology Tests: Phase II - EPA Reference Fuel 36
3.2.1 Emission Analyses 36
3.2.2 Phase II Repeatability Tests - EPA Reference Fuel. 41
4.0 SUMMARY AND CONCLUSIONS 47
iii
-------
LIST OF TABLES
Table No. Page
I Emissions Laboratory Support Instrumentation 11
II Methodology Repeatability Tests - Cars A, B 5 C 27
III Results of Methodology Repeatability Tests - Cars A, B, C,
and Catalytic Converter-Equipped Ford 42
IV
-------
LIST OF FIGURES
Figure No. Page
1 Inside View of Calspan's 590 m Photochemical Aerosol
Chamber 4
2 Inside View of Top Area of Calspan's 590 m Photochemical
Aerosol Chamber 5
3 View of Air Circulation System Containing Ductwork and Char-
coal Filtering Beds ' 6
4 Vehicle Exhaust Sampling System 10
5 Revised Vehicle Emission Test Schedule 13
6 Hydrocarbon Emissions vs. Kilometers (Miles) Accumulated 15
7 Carbon Monoxide Emissions vs. Kilometers (Miles) Accumulated.. 16
8 NO Emissions vs. Kilometers (Miles) Accumulated 17
A
9 Effect of Variations in Natural Air on Test Results 19
10 Effect of Relative Humidity and Added S02 on Test Results 21
11 Effect of Dilution Ratio and Added SO- on Test Results 23
12 Comparison of Hexene-1-SO- Irradiation With and Without the
Addition of NO 24
13 Final Visibility in Smog Chamber vs. HC/NO Ratio 28
14 Repeatability Tests for 300:1 Dilution Plus 0.05 ppm Added SO- 30
15 Repeatability Tests for 300:1 Dilution Plus 0.05 ppm Added SO 31
16 Repeatability Tests for 300:1 Dilution Plus 0.05 ppm Added S02 32
17 Effects of Primary Particulates on Test Results - Car C 34
18 Hydrocarbon Emissions vs. Kilometers (Miles) Accumulated 37
19 Carbon Monoxide Emissions vs. Kilometers (Miles) Accumulated.. 38
20 NO Emissions vs. Kilometers (Miles) Accumulated 39
A
21 Visibility Data for Catalyst Equipped Ford With and Without
the Addition of Thiophene 44
22 Final Visibility in Smog Chamber vs. Initial HC/NO Ratio 46
-------
Section 1
INTRODUCTION
One of the provisions of the 1970 amendment to the Clean Air Act
stipulates that the Administrator of the Environmental Protection Agency
(EPA) may require a fuel or fuel additive manufacturer to conduct specific
tests in accordance with accepted test methods and procedures to determine
the effect of such emissions on the public welfare or on the emission control
performance of a vehicle. The contribution of these emissions to reduced
visibility in the atmosphere is related to this problem and therefore must
also be considered. In order to anticipate possible regulatory measures, a
methodology is required for assessing the effects of fuels and/or fuel additive
combustion products on atmospheric visibility. It has been the objective of
this investigation to develop such a methodology.
The effect of automotive exhaust emissions on atmospheric visibility
is greatly magnified by processes of photochemical aerosol formation. The
methodology must therefore be based on a realistic simulation of normal atmos-
pheric irradiation conditions. While the objective is to assess variabilities
in the ultimate visibility effects corresponding to differences in the types
of fuels and additives, evaluation tests must be performed on exhaust products
of the respective fuels and additives. There is, therefore, a need to stan-
dardize the vehicle engine operating procedures to minimize inadvertent altera-
tions in vehicle emission characteristics. These factors have been taken into
consideration.
As part of the methodology development, three 1972 Chevrolets with
350 CID engines were operated on low-lead and non-leaded indolene fuel of the
same base stock while mileage accumulation, emission analyses, and smog cham-
ber tests were performed. Part of the testing included chamber experiments to
determine the effects of the additives F-310 and CI-2 on exhaust emissions and
atmospheric visibility. After the 16,100 km (10,000 mile) accumulation point
was reached, the engines of the test vehicles were cleaned, and additional experi-
-------
ments were performed on two of the cars using EPA reference fuel of low sulfur
content. A replacement third car, a 1973 Ford Galaxie 500 equipped with a
catalytic converter and operating on both the low sulfur and a 0.1% sulfurated
fuel, was also evaluated during the final phases of the study.
Within this report, descriptions of the experimental facilities used
in the investigation are provided and results of the methodology development,
interpretations of the data, and recommendations for implementation of the
methodology are given.
-------
Section 2
EXPERIMENTAL FACILITIES
2.1 The Calspan Smog Chamber
The main facility used in developing the methodology was the 590 m
(20,800 f t ) Calspan smog chamber. The chamber, shown in Figures 1 and 2, is
constructed of 1.3 cm (0.5 in.) thick steel walls and is 9.1 meters (30 feet) in
diameter and 9.1 meters (30 feet) high. The inner walls of the chamber are
coated with a specially formulated fluoroepoxy, which is similar to teflon in
surface characteristics.
Lighting within the chamber is provided by a combination of fluoro-
escent daylight, blacklamp and sunlights installed inside 24 modules and
arranged in 8 vertical channels attached to the wall of the chamber. Each
module contains two 215-watt F96PG17/D daylight lamps, eight 85-watt P72T12HO
blacklights, and two 40-watt FS40 sunlamps giving a total of 28.5 kw of power.
The modules are covered with 0.64 cm (0.25 in.) annealed Pyrex panels which are
used to isolate the lamps from the chamber atmosphere. The measured light
intensity of the system was found to be k,[NO.]-0.23 min" or approximately 50%
of noonday sun intensity. More recent modifications to the chamber lighting
system have resulted in a measured kd[N02]~0.35 min" ; however, in order to
maintain continuity throughout this investigation, all experiments were per-
formed at the former light level.
Air purification is achieved by recirculating the chamber air through
a series of absolute-plus-charcoal filters, Figure 3. Particle concentrations
of less than 100 cm"3 and levels of N(>x <0.01 ppm, 03 <0.001 ppm, and S02
<0.001 ppm are routinely obtained after eight hours of filtration.
-------
Figure 1 INSIDE VIEW OF CALSPAN'S 590m3 PHOTOCHEMICAL AEROSOL CHAMBER
-------
Figure 2 INSIDE VIEW OF TOP AREA OF CALSPAN'S 590m3 PHOTOCHEMICAL
AEROSOL CHAMBER
-------
Figure 3 VIEW OF AIR CIRCULATION SYSTEM CONTAINING DUCTWORK AND
CHARCOAL FILTERING BEDS
-------
2.1.1 Instrumentation
Gas and particle analysis equipment that was used includes the
following:
(1) Bendix Model 8002 Ozone Analyzer -- This instrument is based
on the principle of photometric detection of chemiluminescence resulting from
the reaction of ozone with ethylene. The minimum detectable sensitivity is
0.001 ppm.
(2) Bendix Model 8101-B Nitrogen Oxides Analyzer -- Detection is
based on chemiluminescent reaction between nitric oxide and ozone. The detec-
tion limit for each of the nitrogen oxides is 0.005 ppm.
(3) Bendix Model 8300 Sulfur Analyzer Operation of this instru-
ment is based on the photometric detection of sulfur atoms excited in a hydro-
gen-rich flame. A set of scrubbers is used for selective monitoring of sulfur
dioxide and hydrogen sulfide. The minimum detectable sensitivity is 0.005 ppm.
(4) Bendix Model 8201 Reactive Hydrocarbon Analyzer Flame ioniza-
tion detection provides analysis of methane (CH.), total hydrocarbons (THC),
and reactive hydrocarbons (THC-CH.).
(5) Hewlett-Packard Model 5750 Gas Chromatograph -- The chromato-
graph is equipped with dual column and dual flame ionization detectors.
Depending on the column in use, either total hydrocarbon or individual com-
ponents can be analyzed. Detection limits of 0.01 ppm are achievable.
(6) Meteorology Research Inc. Model 1550 Integrating Nephelometer --
In operation, an air sample is drawn through a chamber where it is illuminated
by light from a pulsed flash lamp. The scattered light is detected by a photo-
multiplier and compared with a reference voltage from another phototube. The
nephelometer provides a measure of the scattering coefficient of the aerosol
which, in turn, is related to meteorological visual range. Observations of
visibility can be made in the range of -0.5 km (-0.3 miles) to >160 km
(>100 miles).
-------
(6) Gardner Small Particle Detector -- Used to measure total particle
2 6 -^
concentration in the range of 10 to 10 cm . Particle concentration is deter-
mined by nucleation of particles at high supersaturation, followed by observa-
tion of light attenuation within a small cloud chamber. Minimum detectable
particle size is about 0.002 urn.
(7) Environment One Model 100 Condensation Nucleus Monitor -- This
instrument operates on the same principal as the Gardner Small Particle
Detector; however, automatic measurement and recording features are included.
The minimum particle size detected is claimed to be 0.0025 \im. A number of
problems were encountered during the operation of this instrument, and no data
were acquired during several of the experiments. The instrument was not used
for the Phase II methodology tests.
(8) Thermo Systems Inc. Electrical Aerosol Analyzer (EAA) -- Late
in the program an EAA was obtained and used for size distribution analysis of
the smog aerosol. Actual delivery of the instrument was delayed until March
1974 so that only limited use could be made of the instrument. Operation of
the EAA is based on the fact that the mobility of particles decreases with
increasing particle size for particles less than 1 ym. In operation, the
aerosol is charged with negative ions generated by a corona discharge in the
charger. The aerosol emerges from the charger and flows to the mobility
analyzer for analysis. From the mobility spectrum obtained, the particle
size distribution can be inferred. The range of particle size measurement
is from 0.004 pro to -0.4 pm.
In addition to these instruments, special provisions were made to
transfer the Calspan vehicle emissions laboratory from Buffalo, N.Y. to our
Ashford smog chamber site. The system, which includes a chassis dynamometer
and sampling analysis equipment, is described in the following section.
-------
2.2 Vehicle Emissions Research Laboratory (VERL)
One of the program requirements stipulates that all tests be per-
formed in accordance with procedures set forth in the Federal Register, Vol.
57, No. 10. In order to do so, the emissions laboratory and dynamometer wore
transferred to a specially built garage adjacent to the smog chamber.
In operation, vehicles were driven over the LA-4 driving cycle on a
chassis dynamometer while a given quantity of exhaust gas and ambient air was
collected with a positive displacement gas sampler and transferred to Tedlar
storage bags. (When smog chamber tests were being run, a fraction of the
exhaust gases were diverted into the chamber before entering the CVS system.)
The sampling system, shown in Figure 4, consists of exhaust tubing, a series
of ambient air filters, a mixing chamber, heat exchanger, positive displace-
ment pump, four-speed, seven h.p. motor, four Tedlar collection bags, final
exhaust dumping hose, and necessary controls. To ensure a constant volumetric
flow, the gas mixture was kept at constant temperature by passing through a
heat exchanger. After storing a portion of the exhaust gas/ambient air mix-
ture in inert bags, the contents of the bags were transferred to the gas
analyzers for determinations of carbon monoxide, carbon dioxide, total hydro-
carbons, and oxides of nitrogen.
The concentrations of HC, CO, C0_, and NO were measured by four
^ A
multirange analyzers. Two nondispersive infrared analyzers (NDIR) were used
to measure CO and C0_. Total HC were measured by a flame ionization detector,
and oxides of nitrogen were measured by a chemiluminescent analyzer. A com-
plete listing of the pertinent equipment used in the emissions laboratory on
this program is shown in Table I.
-------
SWITCH
COUNTERS
SWITCH
DILUTION
AIR INLET
AIR FILTER
ASSEMBLY
VEHICLE
TAILPIPE
DILUTION AIR
SAMPLE DUMP
SAMPLE BAO EXHAUST
"STABILIZED"
EXHAUST
SAMPLE BA3
V OPTIONAL CONTINUOUS
SAMPLING LINE
HEAT-
EXCHANOER
| PREHEATEB ]
4
COOLANT
REVOLUTION COUNTER PICKUP
POSITIVE
DISPLACEMENT
PUMP
> EXHAUST TO ATMOSPHERE
Flexible Tubing
To Photochemical Reaction
Chamber
FIGURE 4. VEHICLE EXHAUST SAMPLING SYSTEM
10
-------
Table I.
EMISSIONS LABORATORY SUPPORT INSTRUMENTATION
UNiT
MODEL
DESCRIPTION
CHASSIS DYNAMOMETER
DRIVER'S AID
ENGINE COOLING FAN
ENGINE ANALYZE.-.
POSITIVE DISPLACEMENT SAMPLER
CO 5A j ANALYZES
CO2 GAS ANALYZER
I HC DETECTOR
MOx CAS ANALYZ33
CLAYTON CT-50 (VIF)
VARIAN G1000
HARTZELL N24-DU
SUNEET-1160
OLSON CVS 45E-R3
BECKMAN 3:5 8
BECKMAN 315 B
BECKMAN 400
T.E.C.O. 'OA
S?LI7 ROLLERS; 50-hp ABSORPTION UNIT;
STRAIN GAGE TCP.C.UE BRIDGE; A:?.-ACTUATEO
WHEEL LIFTS; MANUALLY CONTROLLED VARIA3LE
INERTIA FLYWHEELS (VIF); INERTIA WEIGHTS
FROM 15CO TO 3CC3 Ib IN 250-lb INCREMENTS AND
FROM 3000 TO 5500 ll> IN 50D-!fa INCREMENTS.
SINGLE-PEN, FIXED-SPEED STRIP CHART RECORDER;
COXTROLS MOUNTED ON FACE COVER.
FIXED-SPEED. 52^,5 cfm. PORTABLE CHASSIS.
23-in. SCOPE; VOLT LEAK UNIT; TACH DWELL UNIT;
COV.3-VAC UNIT; FUEL ;'U!W? TESTER; TIMING
ADVANCE UNIT EXHAUST CONDENSER AND
WIRING JUNCTION UNIT.
7.5-hp, 4-SPEED. FLOW RATES OF 150. 225. 300. AND
400 cfm, 4 BLACK TEDLAR SAGS AND 40 FEET OF
4-:r..-DIA.V.STER FLEXIBLE LEAKPROOF STAINLESS
STSS- TUBING.
S INFRARED ANALYZER; STACKED
CELL; a .=iA.\G2S; ± \% f^'-i. SCALE ACCURACY.
NON-D!S?E3SIVE INFRARED ANALYZER; 3 RANGES;
± 1% FULL SCALE ACCURACY.
FLAMS :C.\':ZAT!ON DSTECTOR; 4 RANGES PLUS
CONTiN JOUS ELECTRICAL SPAN ADJUSTMENT;
± ',% FULL SCALE ACCURACY.
CHEM: LUMINESCENCE ANALYZER; s RANGES;
± 1% FULL SCALE ACCURACY.
2.2.1 Design of Vehicle Exhaust Sampling and Introduction System
An exhaust sampling system was designed for diverting a fixed amount
of auto exhaust into the smog chamber during a test cycle. As shown in Figure
4, a concentric sampling tube was added to the CVS system which allowed for a
fraction of the exhaust gas to be admitted into the smog chamber following
mixing with dilution air. A Sutorbilt 4LXB contamination-free positive dis-
placement gas pump driven by a 3 hp 1750 RPM motor was used to divert various
amounts of the exhaust into the chamber. Under normal operating conditions,
the residence time of the exhaust gases in the introduction system was approxi-
mately 0.01 second so that coagulation of the primary aerosol was minimized.
11
-------
Section 3
RESULTS AND DISCUSSION
3.1 Methodology Development - Phase I
The basic objective of the proposed program was to develop a methodo-
logy to determine the effects of fuels and fuel additive combustion products on
atmospheric visibility. During the methodology development, the three test
vehicles were independently operated on the Clayton chassis dynamometer while
a portion of the auto exhaust was introduced into the smog chamber. Samples
of the auto exhaust were irradiated with simulated sunlight while measurements
were made of visibility, aerosol behavior, and gaseous constituents in the
chamber. A wide range of experimental variables were tested including exhaust
dilution ratio, effects of relative humidity, added S02, added HC, and the
influence of natural nuclei on visibility. The test schedule for distance
accumulation and emissions analyses is shown in Figure 5. Note that after
the completion of Phase I tests, i.e., 16,100 km (10,000 miles), all cars began
using the EPA reference fuel; at that time, Car C was replaced by a catalyst-
equipped 1973 Ford Galaxie provided by the EPA.
The experiments and concomitant results which form the basis for
the methodology devised in this study are described in detail in this section.
12
-------
FIGURE 5. REVISED VEHICLE EMISSION TEST SCHEDULE
DISTANCE ACCUMULATION CXI000)
Kilometers
Miles
9.7 11,3 12.9 14.5
6789
_l__J I L
16.1 17.7 19.3
10 11 12
J L
20.9 22.5 24.1
13 14 15
Vehicle No. 1
Vehicle No. 2
LEGEND
1. O - CONDITIONING TEST (SINGLE TEST)
2. A - REPRODUCIBILITY DEMONSTRATION TESTS
(TRIPLICATE TESTS)
3. A INDOLENE 0
B - INDOLENE 0 + 0.5 GM/GAL TEL
C - EPA REFERENCE FUEL
4. a - CHEVRON F-310
J3 - ETHYL CI-2
Y = 0.1% SULFUR
NOTES:
1. AN UNSPECIFIED NUMBER OF
PRELIMINARY TESTS WILL BE
PERFORMED DURING THE FIRST
11,300 km (7,000 MILES).
2. EACH ENGINE WILL BE CLEANED
AND A NEW EXHAUST SYSTEM
INSTALLED AFTER THE FIRST
16,100 km (10,000 MILES).
3. OPERATION OF VEHICLE #3 WAS
TERMINATED AT 16,100 km
(10,000 MILES) AND REPLACED
WITH A CATALYST-EQUIPPED CAR.
-------
3.1.1 Vehicle Conditioning and Emission Analysis
After initial delivery of the three Chevrolet test vehicles, Cars A
and B were operated on Amoco Indolene fuel containing 13.2 mg lead per liter
(0.05 gm lead per gallon) while Car C was operated on the same base stock fuel
but with 132 mg lead per liter (0.5 gm lead per gallon). Approximately 1,600 km
(1,000 miles) were accumulated on each vehicle every three weeks. For the first
12,875 km (8,000 miles), methodology and exhaust emissions analyses were per-
formed at 1,600 km (1,000 mile) intervals using the basic fuels without
additivies. Following the emissions tests at 12,875 km (8,000 miles), Ethyl
CI-2 (132 rag/liter) and Chevron F-310 (2 ml/liter) were introduced into the
test fuels as additives, and mileage accumulation was continued. After the
16,100 km (10,000 miles) point was reached and the triplicate tests were com-
pleted, the engines were disassembled and cleaned, followed by a new conditioning
series using the EPA reference fuel. The later tests, which comprised Phase II
testing, are described in a subsequent section of this report.
Results of the emissions analyses for the three test vehicles during
the first 16,100 km (10,000 miles) are shown in Figures 6, 7, and 8. The data
show HC, CO, and NO emissions respectively as functions of mileage accumulation.
A
The data in Figure 6 indicate relatively stable HC emissions for Cars A and B
for the first 12,876 km (8,000 miles) followed by a slight upward trend in these
emissions. For the most part, however, the HC emissions for Car A remained
basically the same throughout the first 16,100 km (10,000 miles). Car C, on
the other hand, showed the highest level HC emissions and experienced a steady
rise in the HC level until reaching stability at the 12,875 km (8,000 miles)
point giving approximately 1.5 gm HC/km (2.4 gm HC/mile).
Carbon monoxide and oxides of nitrogen emissions, Figures 7 and 8,
show consistent trends for all three vehicles. The data do not appear to have
been affected by the introduction of additives. As with HC, Car C had the
highest level of CO emissions but slightly lower than average N0x levels.
After the first 11,260 km (7,000 miles), all three cars had relatively stable
NO emissions. Repeatability of test results was somewhat better for each
vehicle during the second triplicate emissions series.
14
-------
1.49
o
12
Figure 6 HYDROCARBON EMISSIONS VS KILOMETERS (MILES) ACCUMULATED
ON VEHICLES
15
-------
56
48
40
32
I "
16
4 6
MILES(x103)
10
12
Figure 7 CARBON MONOXIDE EMISSIONS MB KILOMETERS (MILES) ACCUMULATED
ON VEHICLES
16
-------
5.6
4.8
4.0
3.2
(9
2.4
1.6
0.8
A =
B =
3.2
CAR K
CAR I
worn
4.8
......
km U103)
9.7
12.9
16.1
REPLACE
EXHAUST SYSTEMS
C^EAN ENGINES ON CAR|S
BEGIN USING ADD TIVES
*
I
AND
A §
B
4 6
MILESU103)
10
2.98
2.48
t
1.99
4
1.49
1.00
.50
12
Figure 8 NOX EMISSIONS VS KILOMETERS (MILES) ACCUMULATED ON VEHICLES
17
-------
These data were particularly valuable in interpreting results during
the methodology repeatability tests. For example, it was later shown that
even small amounts of excess hydrocarbon had an adverse effect on the final
visibility in the smog chamber. Thus, the high HC and low NO levels for
A
Car C were found to be responsible for the low visibilities observed after
irradiation and not necessarily the fact that leaded fuel was being used. (It
is possible that the leaded fuel influenced the HC emissions for Car C and thus
indirectly affected visibility; however, this cannot be established without
comparative tests with the same car using non-leaded fuel.)
3.1.2 Effects of Dilution Ratio, Relative Humidity, Variations in Natural
Air, Added SO. and Added Hydrocarbon on Test Results
Smog chamber tests were designed to study the effects of background
air (i.e., filtered vs. natural air), dilution ratio, relative humidity, and
added S02 on chamber visibility. In most experiments filtering of the air
(absolute plus charcoal) was accomplished before introducing exhaust into the
chamber. On occasion, background air containing natural nuclei from the rural
environment was used for a test. During this stage of methodology development,
the first 10 minutes of the LA-4 cycle was driven on the chassis dynamometer,
while a fixed amount of auto exhaust was admitted into the chamber. Dilution
ratios of 200:1, 300:1, 500:1, and occasionally 1000:1 were tested.
Visibility traces from selected experiments showing the effect of
variations in natural air on test results are shown in Figure 9. For the
experiments shown in Figure 9, outside air was first admitted into the cham-
ber followed by the addition of auto exhaust at a dilution ratio of 200:1.
The sample was then irradiated for up to 23 hours, while observations were
made of visibility, NO , oxidant, total sulfur, and particle concentration.
18
-------
AUTO EXHAUST + NATURAL AIR (200:1)
AUTO EXHAUST + NATURAL AIR (200:1)
AUTO EXHAUST + NATURAL AIR (200:1)
A AUTO EXHAUST + NATURAL AIR (200:1)
AUTO EXHAUST+ NATURAL AIR (200:1)
NATURAL AIR ONLY
RH
35%
35%
35%
80%
80%
CARC
CARC
CARC
CAR A
CAR A
V)
uj
S
»-
Hi
m
v>
8 10 12 14
IRRADIATION TIME (HRS)
16
18
20
22
t
m
>
Figure 9
EFFECT OF VARIATIONS IN NATURAL AIR ON TEST RESULTS
19
-------
The data in Figure 9 show that variations in the nucleus concentra-
tion and background gaseous contaminant level in the natural air produce widely
different initial visibilities as well as important differences in test results.
For example, a comparison of seemingly identical experiments, such as 2, 5, and
6 or 7 and 8, show poor repeatability. Despite the fact that added realism is
provided by using natural air in the smog chamber, the wide variability caused
in test results makes this procedure unacceptable in terms of methodology
development. In all subsequent experiments, therefore, chamber air was normally
filtered for several hours before the start of an experiment in order to remove
particulates and trace gaseous contaminants from the atmosphere.
In Figure 10 the effect of differences in relative humidity and added
SO- on test results is shown. Air which was first filtered thoroughly was used
in these experiments and all dilution ratios were the same (i.e., 200:1). As
shown by the data for run no. 4, auto exhaust plus filtered air which was
irradiated for 20 hours at low humidity (~35%) resulted in only a small restric-
tion in visibility. On the other hand, experiment 10,which represents the same
test condition but high humidity (-80%), shows a much larger visibility loss
with time. After 18 hours of irradiation in that experiment, visibility
decreased from 80+ km (50+ miles) to approximately 29 km (18 miles). The effect
of high humidity can be anticipated, of course, since any hygroscopic component
of the nuclei in the chamber will deliquesce and enlarge if there is sufficient
moisture available.
The combined effects of high humidity and added S02 can be seen by
comparing experiments 11 and 12. Here, the same dilution ratios were used,
but approximately 0.8 ppm S0_ was added to each system shortly before intro-
ducing auto exhaust into the chamber. As shown, the addition of sulfur dioxide
had a substantial effect on the visibility within the test environment. In
the high humidity case (experiment 12), there was an effect even before irra-
diation was started. This probably was the result of a dark reaction of the
SO followed by enlargement of the H2S04 aerosol at high humidity. Once the
lights were turned on, increased photooxidation of S02 occurred causing the
production of significant additional light scattering aerosol. After 22 hours of
irradiation, the visual range had dropped to 3.9 km (2,4 miles) in this experimert,
20
-------
NO
4n.
r-
L-h i
10 O C
11 A .... ^
11 ZA
12 A_
50 1
40
30
20
OT
UI
_l
I
> 10
j
5
to
5
4
3
2
1
(
0
,m. ,....,..,
I AUTO EXHAUST + FILTERED AIR (200:1)
) AUTO EXHAUST + FILTERED AIR (200:1)
i AUTO EXHAUST + FILTERED AIR (200:1) +
AUTO EXHAUST + FILTERED AIR (200:1) +
4 ....
^ ^Sj. *v *
r ^x°\
f '^0. , NA ~A
l
\
" A
i
1
r "
I
I
| |
>x
1 A
J
\
\
4 ^6
; >
1
1
r ~t~~
TiJ
1
1
!
V
X,
\2
L -
1
.... '.Xt--i
N
\,
X
X
0.8 ppm S02
0.9ppmSO2
t i
RH
35% CAR A
80% CAR A
35% CAR A
80% CAR A
t T
i 4 '
i
l^°~l
1
I
A^
A....
v\
>o.
VN..
i
r 1 1
^0^
1
_ .,. T
~
i
, _
4
rr=r*
i
i
....10
1
i
i
i
t
\
)
i
ii^A
""
,
L
12
1
) 2 4 6 8 10 12 14 16 18 20 22
80.5
64.4
48.3
32.2
16.1
8.1
6.4
4.8
3.2
>
J
5
IRRADIATION TIME (HRS)
Figure 10
EFFECT OF RELATIVE HUMIDITY AND ADDED SO2 ON TEST RESULTS
21
-------
Because of the dominant effect which high humidity had on test results, it was
decided that subsequent experiments would be performed at lower humidities in
the range of 30+10% RH. Smaller dosages of SO , more typical of levels found
in natural urban atmospheres, were also considered desirable.
Results showing the effect of differences in dilution ratio and
additional comparisons of the effect of added S0_ on visibility are shown
in Figure 11. These experiments show that in the absence of any added SO-
a dilution ratio not greater than 200:1 is required to produce even slight
visibility losses after irradiation (e.g., experiments 4 and 13). With the
addition of S02 (i.e., 0.12 ppm) to the system, dilution ratios of approxi-
mately 500:1 produce some visibility restriction. At a dilution ratio of 1000:1,
no visibility loss at all is observed even in the presence of added S0_.
During the course of these experiments, it was noted that aerosol
formation was closely tied to the oxidation and disappearance of nitric
oxide (NO) from the system. This observation, while important in terms of
understanding the mechanisms of particle formation, was a complicating factor
in the methodology development. For example, auto exhaust from the test
vehicles, when operated in accordance with the Federal test schedule, is
inherently high in nitrogen oxides and relatively low in reactive hydrocar-
bons. Under these conditions, the oxidation of NO proceeds slowly and aero-
sol formation is suppressed. On the other hand, in the presence of larger
concentrations of reactive hydrocarbon, the oxidation of NO and the formation
of ozone is greatly accelerated and particle formation mechanisms are favored.
This effect is illustrated in Figure 12 where visibility data for several
experiments are shown. In the one case, 2 cc (0.7 ppm) of hexene-1 was added
to a system containing 2.1 ppm NO and a trace amount of SO . The large inhi-
biting effect of nitric oxide on the formation of light scattering aerosol is
obvious from these results in which visibility never dropped below 80+ km (50+
miles) in the case where NO was present. The data are compared with a similar test
in which only background levels of NO were present. In this case, appreciable light
scattering aerosol was formed, and substantial visibility losses were observed
during the course of the experiment. A third case is illustrated in which
22
-------
NO
4 ) D D AU'
11 1 fll 1
11 \ O O AU-
1 ll 1 A ^ * ' "
1** / w
19 A-
50
40
30
20
. j
f
:; i ~ i
0
1 !
2
..
k
!
!
j
tit!
4 6
i
1
i
3)F 4
i
i
I I
i
I
1
J
|
I
1
1
|
"^"^i
l-*v
! T
.
i
' """
L_ 4
^V^ _ ,1 /»/| A
^V>N^>4L14
i ^^^.^ AO O
_ oo 9 A
1 *" A
| 1
1
11 1 T "~ 'o-1 >
* c t ' r ^ H
*O i i -» 13
j 1 s
l
i r
i J CO
f 1 >
i i 8.1
;
I
1 JAR
j j
i. J ^ t
i 1 3-2
| j
1 i
! i
1
il i i
8 10 12 14 16 18 20 22
IRRADIATION TIME (HRS)
Figure 11 EFFECT OF DILUTION RATIO AND ADDED SO2 ON TEST RESULTS
23
-------
O O HEXENE-1 (0.7 ppm) + 0.07 S02
HEXEIME-1 (0.7 ppm) +0.035 S02 + 2.1 NO
50
40
30
20
CO
LU
I
>
t 10
m
CO
5
4
3
2
1
i _ i i
' ! t
i :
i i
i
j
i
I
1 ! 1
;
\
\
i V
i
i
i
1
i
1
1
__ I
i
1
!
1 m T
N3 1
\
<
\
I
i
!
ii nt/\ci\ic
4
T ; T
] 1 i
"] !
l ^^*ti rt rti i
^^
V
1
T ^s
1 ^
N.Q
^".
SJ^
1
1
i
] r*] eft r\
\L
B
\
*.^Q
^^
| j
0 2 4 6 8 10 12 14 16 18
TIME (MRS) *-
i
^
\
\
h T
K "
pr
i
t "
i |
20 22
180.5
64.4
48.3
32.2
16.1
8.1
6.4
4.8
3.2
t
Figure 12 COMPARISON OF HEXENE-1«SO2 IRRADIATION WITH AND WITHOUT
THE ADDITION OF NO
24
-------
2 cc (0.7 ppm) hexene was added to a 500:1 auto exhaust + filtered air system
with 0.04 ppm added SO-. Here approximately 13 hours elapsed while the NO
was being oxidized out of the system. After that time there was rapid forma-
tion of ozone and of light scattering aerosol as indicated by substantial visi-
bility losses. It-was determined following these tests that as part of the
methodology repeatability demonstration, an additional experiment would be
performed for each car. In these tests, 3 cc of the low-lead or non-leaded
fuel would be added to the exhaust gas + S0_ system as an evaporative emission.
The added HC was expected to provide a more realistic HC/NOx ratio of approxi-
mately 3:1 instead of the 1:1 ratio that normally was observed.
In summary, then, the following test conditions were chosen as a
result of the methodology development phase of the program.
Dilution Ratio: 300:1
Relative Humidity: 30±10%
Added SO : 0.05 ppm
Irradiation Time: 23 hours
Temperature: 24°±5°C (75±10°F)
One additional test was performed for each of the vehicles in which 3 cc of
the test fuel was added to the chamber as an evaporative emission.
3.1.3 Repeatability Tests - Phase I
Triplicate repeatability tests of the Phase I methodology were per-
formed on each test vehicle after the 12,900 and 16,100 km (8,000 and 10,000
mile) accumulation points. Following tests at the 12,900 km (8,000 mile) point,
the additives CI-2 and Chevron F-310 were introduced into the test fuels, and
distance accumulation was continued. The repeatability tests involved admitting
auto exhaust into the large smog chamber at a 300:1 dilution with air, adding
O.OS ppm S0_ and irradiating the sample for 23 hours.
25
-------
Results of methodology repeatability tests are summarized in Table II
and compared with test data acquired prior to the addition of additives to the
test fuels. Good reproducibility of test results was achieved for Cars B and C,
but not for Car A. In fact, during the performance of the second test series at
16,100 km (10,000 miles), the variability in Car A test data was particularly
large. Careful inspection of the data showed that variations in engine per-
formance were causing large differences in the HC level within the chamber which,
in turn, produced significant changes in the final observed visibility. After
completing several experiments, measurements of initial HC in the chamber were
made using a Beckman flame ionization detector from the vehicle emissions labora-
tory. These measurements allowed us to establish the HC/NO ratio in the chamber
at the start of each experiment and to correlate this ratio with the final visi-
bility in the chamber. The close correlation is illustrated in Figure 13,
showing essentially an exponential dependence of final visibility on the initial
HC/NO ratio inherent in the vehicle exhaust. The dashed line shows the effects
of further increases in the initial HC/NO ratio as a result of introducing
evaporative emissions into the chamber.
In terms of final visibility within the smog chamber, the excellent
repeatability of test results for Cars B and C (although fortuitous) was the
result of nearly identical concentrations of exhaust emissions being intro-
duced into the chamber during the respective test runs. In all cases, varia-
tions in emissions, especially for NO, were observed to correlate closely
with extreme changes in relative humidity. For Car A tests, the exhaust
HC/NO ratio varied from 0.61 to 1.13; consequently, the observed final visi-
bility for these tests also varied over a wide range. On warm, humid days
the NO emissions were predictably somewhat less giving a high HC/NO ratio
and lower final visibilities in the chamber. During the winter, weather was
less of a factor since the relative humidity was always low once the outside
air reached room temperature. (Because of these observed changes in engine
emissions as a function of relative humidity, special provisions were made
to control the carburetor air intake for the Phase II tests involving the
EPA reference fuel. These tests are discussed in the next Section.)
26
-------
TABLE II.
METHODOLOGY REPEATABILITY TESTS - CARS A, B § C
Visibility in Kilometers (miles)
Test
Car A
non-
lead
fuel
*
F-310
Car B
non-
lead
fuel
+
CI-2
Car C
low-
lead
fuel
+ CI-2
#18
#19
#20
#21
#22
#11
#13
#15
#16
#17
#23
#24
#25
#26
Condition
300:1 + 0.05 SO-
ii 2
n
n
300:1 + 0.05 SO,
+ 3 cc HC *
300:1 + 0.05 S02
n
300:1 + 0.05 S0_
I
+ 3 cc IIC
300:1 + 0.05 S02
n
300:1 + 0.05 SO,
+ 3 cc HC *
T
°C
25.5
25.0
25.0
26.1
26.7
19.4
21.7
23.3
26.7
25.5
21.1
23.9
24.4
28.3
RH
%
46%
32%
41%
55%
43%
39%
31%
40%
36%
40%
36%
48%
49%
51%
NO-
ppm
2.94
4.10
3.78
2.91
3.58
3.45
3.55
3.47
3.30
3.27
3.68
3.20
3.32
2.97
N0f
ppm
1.30
2.20
1.90
1.21
1.28
1.64
1.78
1.62
1.35
0.90
1.57
1.13
1.29
--
HC
ppmC
3.15
2.50
2.95
3.30
8.05
3.60
3.55
3.90
3.75
8.85
5.25
5.25
5.00
10.8
Initial
HC/NO
1.07
0.61
0.79
1.13
2.25 '
1.04
0.97
1.13
1.13
2.70
1.43
1.64
1.51
3.63
after 23 hours
Anril 16,100 km
Apri1 (10,000 mi)
35.1^
61.1 (42.5+19
46.7 f (26. 4+12)
27.3.X
11.3
35.1^
36.8 1 31.9+5.0
27.0 f(19.8±3.1)
28.5-'
9.6
I9'a*\ 14.311.9
15 gJC 8.9±1.2)
s.'e
irradiation
Feb. ..12. 870 km
(8,000 mi)
17.2
29.8-1 30.9^2.
33.0 i(19.2±l.
12.7
30 . 6 ^
36.7 ( 33.5+3.
33. 0-J (20.812.
12.9
28.2*1 26 5+4
292'.°5J(16-5*2'
1
3)
2
0)
0
5)
-------
0.6
1.0
HC/NO
Figure 13 FINAL VISIBILITY IN SMOG CHAMBER VS INITIAL HC/NO RATIO OF EXHAUST
EMISSIONS (TEST NUMBER SHOWN ADJACENT TO EACH DATA POINT)
28
-------
As a result of the aforementioned variability in engine performance,
no meaningful comparisons of additive and non-additive tests could be made for
the Car A data. Results for Car B during the two test series were essentially
unchanged leading to the conclusion that there was no effect of the additive
within the experimental variability of the data. Final visibility data for
Car C were significantly lower during the additive test series. To conclude
that there was an adverse effect due to the additive is not warranted, however,
since the observed low visibility could have been predicted based on the Car C
HC/NO data alone. It is not obvious, however, whether the comparatively high
HC emissions for Car C were affected by the additive or were inherent in the
engine performance. It may be noted that even in the absence of CI-2, Car C
with low-lead fuel gave somewhat lower visibilities during the additive-free
test series than Cars A and B which were operated on lead-free fuel.
One observed effect of the additive Ethyl CI-2 in combination with
the low-lead fuel (Car C) was that the emission of primary particulates from
Car C produced an immediate visibility loss in the smog chamber, even before
the lights were turned on. This can be seen in Figure 14 in which visibility
in the smog chamber is plotted against irradiation time in hours. Note that
the initial visibilities lie in the range of about 50 to 70 km (32 to 45 miles)
after introducing exhaust into the chamber. All additive-free tests for Car C and
all tests for Cars A and B show no initial visibility losses for a 300:1 dilution
ratio. The data for Cars A and B showing visibility vs. time are plotted in
Figures 15 and 16. The very pronounced effect of the added evaporative emis-
sion on final visibility can also be noted in these figures.
Several additional experiments were performed to determine the
effects of primary particulates on visiblity within the smog chamber after
irradiation. Three types of experiments were performed. In the initial test,
auto exhaust (dilution ratio 300:1) plus 0.05 ppm S02 was irradiated for 23
hours. In the second experiment, all primary particulates in the auto exhaust
were removed by absolute filtration before turning the lights on. Exactly
0.05 ppm SCL was then added to the system, and the exhaust gas plus S02 mix-
ture was irradiated for 23 hours. A final experiment involved the introduction
29
-------
CAR C LOW LEAD FUEL WITH ADDITIVE ETHYL CI-2
NO.
1
2
3
-O
D D
O O
EXPERIMENT NO. 23
EXPERIMENT NO. 24
EXPERIMENT NO. 25
EXPERIMENT NO. 26
CAR C
300:1 +0.05 S02
300:1 +SO2 + HC
V)
HI
I
H
_i
m
T
E
£
_i
5
8 10 12 14 16
IRRADIATION TIME (HRS)
20
22
Figure 14 REPEATABILITY TESTS FOR 300:1 DILUTION PLUS 0.05 ppm ADDED SO2
30
-------
CAR A NON-LEADED FUEL WITH THE ADDITIVE CHEVRON F-310
tn
ui
NO
1 )
2 I
3
4 )
K
O---
D-- -
A---
n__
-- O
-
-D
r>
---O EXPERIMENT NO. 18 CAR A
EXPERIMENT NO. 19 CAR A \ 300:1 + 0.05 S02
EXPERIMENT NO. 20 CAR A
EXPERIMENT NO. 21 CAR A
O EXPERIMENT NO. 22 CAR A 300:1 + S02 + HC
3.2
8 10
IRRADIATION
12 14
TIME (HRS)
16
18
20
22
t
>
00
Figure 15 REPEATABILITY TESTS FOR 300:1 DILUTION PLUS 0.05 ppm SO2
31
-------
CAR B NON-LEADED GAS WITH ADDITIVE ETHYL CI-2
NO
LU
m
t/3
O O EXPERIMENT NO. 11 CAR B
-- EXPERIMENT NO. 13 CAR B
D D EXPERIMENT NO. 15 CAR B
£. A EXPERIMENT NO. 16 CAR B
300:1 + 0.05 SO
O
-O EXPERIMENT NO. 17 CAR B 300:1 + S02 + HC
!
E
it
ffi
>
8 10 12 14
IRRADIATION TIME (HRS)
Figure 16 REPEATABILITY TESTS FOR 300:1 DILUTION PLUS 0.05 ppm ADDED SO2
32
-------
of equivalent amounts of HC plus NO into clean filtered chamber air in
order to simulate auto exhaust pollutant levels comparable to those used in
previous tests. The results are shown in Figure 17 in which visibility is
plotted as a function of irradiation time for each of the tests.
In test number 1 (auto exhaust with primary particulates +0.05 ppm
SO-), there was an immediate reduction in visibility after introducing the
exhaust products into the chamber. This effect is typical for Car C which
was being operated on low-lead fuel with the additive Ethyl CI-2. After irradi-
ation for 21 hours, the smog chamber visibility dropped to 5.5 km (3.4 miles),
a value which is consistent with the methodology tests. In experiment 2 (i.e.,
filtered auto exhaust + S0_), a much lesser visibility loss was noted. In fact,
the minimum visibility of about 29 km (18 miles) was observed after 17 hours of
irradiation, followed by a slow improvement to 40 km (25 miles) during the next six
hours of the experiment. Aerosol size distribution analysis for these two
experiments show that the surface concentration of particulates at the end
of experiment 1 was 2.8 x 10 ym /cm and at the completion of experiment 2
about 1.2 x 104 nm2/cm3 or less than half as much. This difference in surface
concentration is sufficient to account for the observed differences in visi-
bility. (The amount of scattered light is directly proportional to the cross
sectional area of the particles.) Visibility data for a third test involving
HC + NO + S02 gases is also shown in Figure 17. For this case, only a modest
reduction in visibility was noted after lengthy irradiation.
These test results demonstrate that primary particulates play an
important role in the initial and final visibilities noted in the smog chamber.
Furthermore, the tests show that irradiations of HC+NO*S02 mixtures comparable
to those found in auto exhaust do not produce equivalent visibility losses.
It is concluded that in order to examine the effects of fuels and fuel additive
combustion products on atmosphere visibility natural auto exhaust containing
primary particulates must be used.
33
-------
EXPERIMENT
NO.
TEST CONDITION
1 -
2 O-
3 A-
AUTO EXHAUST WITH PRIMARY PARTICLES (300: 1) + 0.05SO2 HC/NO = 4.74
AUTO EXHAUST - FILTERED - (300:1) + 0.05 SO,
-A HC + NO + SO2 (0.05 ppm)
HC/NOX = 2.50
HC/NOX = 2.92
FILTERED AUTO EXHAUST
8 10 12 14
IRRADIATION TIME (MRS)
16
18 20
22
Figure 17 EFFECTS OF PRIMARY PARTICIPATES ON TEST RESULTS - CAR C
34
-------
3.1.4 Summary of Phase I Test Results
From the analysis of data acquired during the first phase methodology
development and repeatability tests, the following conclusions can be drawn:
(1) The effects oto visibility in a 590 m3 (20,800 ft3) smog chamber
of the additives Ethyl CI-2 and Chevron F-310 are small compared to the effects
brought about by natural variations in engine performance.
(2) The combined effect of Ethyl CI-2 with a low-lead fuel had a
marked influence on the initial visibility observed in the smog chamber. No
other combination of fuels or additives showed this effect.
(3) The final visibility achieved after irradiation of a 300:1
exhaust gas-filtered air mixture with 0.05 ppm added SO- is closely correlated
with the initial HC/NO ratio in the chamber. In fact, the correlation is so
good that final visibilities can be accurately predicted from an initial
measurement of total HC and nitrogen oxide in the chamber.
(4) Besides having the expected effect of reducing initial visi-
bility in the smog chamber, primary particulates in auto exhaust were found
to play a dominant role in the long-term visibility losses as well (i.e.,
23 hours irradiation).
It should be noted that the basic objective of this program was
methodology development. No attempts were made to perform a sufficient num-
ber of tests on any single system to permit rigorous statistical assessment
of the effects of specific fuels or additives on visibility. The above sum-
mary of test observations serve only to demonstrate the efficacy of the
methodology derived. Some test procedure shortcomings are quite evident.
Methodology modifications were therefore again emphasized in Phase II of the
program. Humidity control on test vehicle carburetor intake air to minimize
HC/NO variability on a given vehicle-fuel system was identified as an essen-
tial requirement from Phase I test results.
35
-------
3.2 Methodology Tests: Phase II - EPA Reference Fuel
After completion of the Phase I triplicate emissions tests and cham-
ber experiments at the 16,100 km (10,000 mile) point, the engine heads from the
three test vehicles were removed and cleaned, and new exhuast systems were
installed. By mutual agreement, Car C was dropped from the testing program
and replaced by a catalyst-equipped 1973 6.55 liter (400 CID) Ford Galaxie 500
supplied by the EPA. Distance accumulation was again started on each of the
cars using lead-free, low sulfur (-124 ppm) reference fuel. After 4,830 km
(3,000 miles), a triplicate emission and smog chamber test series was performed
on all cars followed by the addition of the additives Chevron F-310 and Ethyl
CI-2 to the fuel of Cars A and B, respectively. At the same time, fuel used
in the catalyst Ford was doped with sufficient thiophene to give a sulfur
content of -1,000 ppm.
Earlier difficulties in maintaining stable temperature and humidity
conditions in the emissions laboratory were rectified by using a refrigeration
system with reheat coils to supply controlled air to the cars during a test.
In operation, refrigerated air from a one-ton air conditioner was directed
through ductwork and over a series of heaters slightly ahead-of the air intake
of the carburetor. By regulating the amount of heat supplied to the refriger-
ated air within the ductwork, humidities between 30% and 80% could reliably
be obtained. For our tests, the relative humidity was kept at approximately
40%.
3.2.1 Emission Analyses
Complete results of the emissions analysis for each of the test cars
for Phases I and II are shown in Figures 18, 19, and 20. The tests were per-
formed every 1,610 km (1.000 miles) with the Phase II triplicate emissions tests
performed at the 20,920 and 24,140 km (13,000 and 15,000 mile) points.
In general, the Phase II hydrocarbon emissions for Cars A and B are com-
parable to Phase I tests and average approximately 0.9 g/km (1.5 g/mile). The
36
-------
2.8
2.4
-,$
2.0
1.6
1.2
0.8
0.4
A
KI LOITERS
9.7 12 9
u:
I-H |
"a.;"
N-4 i
K\
16.1
1
.
A.
BEGIN" USING ADD If VES
INDOLENE FUEL
.Si\.
b I
REPLACE:EXHAU?T
SYSTEM ^ND CL$AN
_EIJGINE_Si- CARi A § I
"lij;"
H;
u;
Oil
VF
,/
1.49
.75
.50
.25
BEGIN USING ADDITIVES
REFERENCE FUEL
i * ' h
S 8
MILES (x 103)
10
12
14
FIGURE 18. HYDROCARBON EMISSIONS VS KILOMETERS CMILES) ACCUMULATED ON VEHICLES
37
-------
56
3.2
4.8
KILOMETERS
7 12.9
16.1
19.5
2S.
REPL CE EXFJAUST S.YSTEM iAND
CLEA ENGINES - CJARS A |§ B
j..
48
7
...> i
\
x
40
32
24
16
1
= C R
= C R
= C R
A
B
C
F = C TALYt C FOR
BEGIN USING ADDITIVES
INDOLENE FUEL
-p!-
BEGIN USING ADDITIVES
REFERENCE FUEL
cc.
10
12
14
2468
MILES (x 103)
FIGURE 19. CARBON MONOXIDE EMISSIONS VS KILOMETERS (MILES) ACCUMULATED ON VEHICLES
38
-------
3.2
KILOMETERS
12.9
19.3
25.7
4.8
BEGIN USING
INDOLENE FUEL
BEGIN USING ADDITIVES
REFERENCE FUEL
MILES (X 10 )
FIGURE 20. OXIDES OF NITROGEN EMISSIONS VS KILOMETERS (MILES ACCUMULATED ON VEHICLES)
39
-------
catalyst-equipped Ford is predictably much lower in HC emissions, generally
averaging about 0.9 g/km (0.5 g/mile). Repeatability of test results at the
20,920 and 24,140 km (13,000 and 15,000 mile) points are improved over earlier
triplicate emissions tests, probably because of the controlled humidity air
supply system.
The CO emissions for Cars A and B averaged about 20-g/km (32 g/mile)
and showed a downward trend during the last 8,045 km (see Figure 19). The
final average values are comparable to those observed during the Phase I por-
tion of the program. The catalyst Ford showed a marked increase in CO emis-
sions throughtout most of the distance accumulation period. When the Ford was
first received, its CO level was 2.6 g/km (4.2 g/mile), in good agreement with
previous EPA measurements of 2.7 g/km (4.34 g/mile). Since that time, the CO
level increased to a high of almost 13 g/km (21 g/mile) at the 22,530 km
(14,000 mile) point. Decreased catalyst efficiency or a richer air/fuel ratio
are possible causes; however, these conjectures are not consistent with the
relatively stable HC data for this vehicle over the same period. It is worth
noting that substantial deposits were found in the carburetor of the Ford at
the 19,310 km (12,000 mile) point, causing it to run excessively rich; at that
point, the carburetor was cleaned and rebuilt, and the idle air/fuel ratio
reset to specifications. Because of the steady increase in CO emissions, oper-
ation of the catalyst-equipped Ford was terminated shortly after reaching the
22,530 km (14,000 mile) point. As shown, excellent repeatability of CO test
results was observed in both sets of triplicate tests on the Ford.
The NO emissions showed the greatest variability as a function of
J\
distance accumulation for all three cars (see Figure 20). The catalyst Ford had
lower NO emissions than Cars A and B, which showed sharp increases in NO
X «
emissions between 19,310 and 22,530 km (12,000 and 14,000 miles). Complete
system leak checks and analytical calibration showed no abnormalities. The fact
that the NO levels of Cars A and B 4,830km (3,000 miles) after engine cleaning
A
were consistent with those obtained after the initial 4,830 km (3,000 miles) accumu-
lation suggests that the cars were behaving normally throughout the testing period.
Reproducibility of NO test results in both triplicate emissions series was excellent-
for all cars.
40
-------
3.2.2 Phase II Repeatability Tests - EPA Reference Fuel
Triplicate repeatability tests using the recommended methodology were
performed in the Calspan smog chamber at the 20,920 and 24,140 km (13,000 and
15,000 mile) accumulation points. As before, the tests involved driving the
first 23 minutes of the LA-4 driving cycle on a chassis dynamometer, while a
fixed portion (300:1 dilution ratio) of the auto exhaust was admitted into the
smog chamber. After adding 0.05 ppm SO to the chamber air and regulating the
temperature to ~26.5°C (~80°F), the sample was irradiated for 23 hours while
the reactants and visibility were continuously monitored.
All three cars were operated on the identical, additive-free EPA
reference fuel (sulfur content 124 ppm) until reaching the 20,920 km (13,000 mile)
point. After performing triplicate emissions and visibility tests, the addi-
tives F-310 and CI-2 were introduced into the fuels of Cars A and B and thiophene
to the fuel of the catalyst Ford (sulfur content approximately 1,000 ppm). A
similar set of tests were conducted at the 24,140 km (15,000 mile) point. Results
of the visibility analysis are summarized in Table III.
f
The data in Table III show that the final visibilities for Cars A
and B at the 20,920 km (13,000 mile) point are dramatically improved over
earlier test results. Although nearly identical test procedures were used, the
final visibilities using the non-leaded, low sulfur reference fuel, were found
to be much better than in tests using commercial grade non-leaded and low-lead
indolene fuel. The low sulfur content of the reference fuel is the main
difference in the fuels and is thought to be responsible for the observed
differences in results. As before, the greatest scatter in the data was
observed for Car A. The catalyst-equipped Ford was the cleanest of the test
vehicles at this point, both in terms of HC and NO emissions and final
visibility after irradiation.
After introducing additives into the fuels of Cars A and B and thio-
phene into the fuel of the Ford, triplicate emissions tests were again per-
formed at 24,140 km (15,000 miles). Here, the greatest difference in test
results can be found in the data for the catalyst Ford. The addition of sulfur
41
-------
TABLE III.
RESULTS OF VISIBILITY ANALYSIS FROM METHODOLOGY REPEATABILITY TESTS
CARS A, B, C, AND CATALYTIC CONVERTER EQUIPPED FORD
Test Condition
CAR A 300:1 + 0.05 SO-
non- leaded fuel
+ F-310
CAR B 300:1 + 0.05 S02
non- leaded fuel
+ CI-2
CAR C 300:1 + 0.05 S02
low lead fuel
+ + CI-2
FORD
EPA Reference Fuel
VISIBILITY IN KILOMETERS CMILES) AFTER 23 HOURS IRRADIATION
12,900 km
C8.000 mi.)
17.2^
1 30.9±2.1
^'8 J C19.2±1.3)
33.0JI
36'6 J 33.5±3.2
36.7 (20.8±2.0)
33. OJ
28'2) 26.5±4.0
29. 0( (16.5±2.5)
22.5}
--
*
16,100 km
CIO, 000 mi.)
35.11
,, , / 42.5±19
01
u.
u
a
U3
U,
a
CL
W
£2
i i
C/3
^
20,920km
(13,OCC mi.)
77 >
1 nt / 87il6
J.UD i
J (54±10)
88 y
72 )
63 )
fifi ( 68±8
66 \ (42±5)
76 \
64 f
--
85 ^
/ 93±8
95 > (58±5)
98 )
24,140 km
CIS, 000 mi.)
93-x
92 ( 100±16
f C62±10)
116J
Reference fuel
plus F310
42TN
. I 47±5
> C29±3)
45j
Reference fuel
plus CI-2
-
47 }
/ 42±5
42 > C26±3)
\
37 J
Reference fuel
plus thiophene
-------
to the fuel also resulted in substantial lowering of visibility after irradi-
ation. For example, the average visibility after irradiation for the Ford
without sulfur was 93±8 km (58±5 miles), whereas with the added thiophene, the .
average measured final visibility was 42±5 km (26±3 miles). Visibility data
showing these differences can be seen in Figure 21. The actual concentrations
of sulfur-containing compounds observed in the smog chamber were -0.01 ppm for
tests performed at 20,920 km (13,000 miles) and 0.15 ppm (expected to be mainly
in the form of H2S04 mist) for the tests at 24,140 km (15,000 miles). This
large difference can easily account for the observed lowering in final visibility.
Another important difference in test results can be found in the data
for Car B. After introducing the additive CI-2 into the fuel at 13,000 miles,
lower visibilities were observed. The additive CI-2 is an organomanganese
compound (Methylcyclopentadienylmanganese Tricarbonyl). This additive is
known to be converted almost completely to a manganese oxide form, mainly Mn.^,
in the automobile exhaust. Even the small fraction of unburned CI-2 in the
exhaust is likewise decomposed rapidly to manganese oxides and carbonates in
the atmosphere by photolytic processes. Manganese species in the form of fine
particulate aerosols, as generated from automotive exhaust emissions, are known
to be effective in catalyzing the oxidation of SO to sulfate (Matteson et al.,
1969). It appears that in the case of low sulfur EPA reference fuel, the en-
hanced S02 photooxidation due to manganese aerosol catalysis may be responsible
for the observed lower visibility. In the earlier series of tests with the
higher sulfur content fuel at the 16,100 km (10,000 mile) point, an effect
of CI-2 was to cause a marked influence on the initial visibility in the smog
chamber. Other contributions may have been masked by appreciable S02 to sul-
fate conversion even in the absence of a catalyst.
For this particular example, the full range of visibilities is shown. In all
previous examples, the common practice of designating 80+ km (50+ miles) as
unrestricted visibility has been followed.
M.J. Matteson, W. Stoober, and H. Luther, Ind. Eng. Chem. Fundamentals, 8, 677
(1969).
43
-------
100
90
80
70
60
50
40
30
20
ffl
10
FORD NO. 1
FORD NO. 2
FORD NO. 3
THIOPHENE
WITH
A
WITHOUT
O
D
A
8 10 12 14 16
IRRADIATION TIME (HOURS)
18
20
22
Figure 21 VISIBILITY DATA FOR CATALYST EQUIPPED FORD WITH AND WITHOUT
THE ADDITION OF THIOPHENE
44
-------
Further insight into the relevance of these differences with respect
to earlier tests can be found from the data in Figure 22 showing final visibi-
lity after 23 hours irradiation vs. initial HC/NO ratio for all tests. The most
recent data generated using the reference fuel show a similar correlation of
visibility with initial HC/NO ratio, but the data points lie along a different
straight line. Although only two types of fuel were tested on this program,
it seems possible that a family of curves could be established for represen-
tative fuel types, those with high sulfur content lying lower on the scale
than low sulfur-bearing fuels. Thus, all three cars using the low sulfur
reference fuel were found to lie along the uppermost line. Although only
three tests were performed with the catalyst Ford using the sulfurated fuel,
the data show that the addition of thiophene to the fuel produced more reac-
tive emissions and visibilities which were substantially lower than in the
absence of the thiophene additive. The introduction of H«SO. droplets
and the photooxidation of SO. to form acid particles can be expected to pro-
duce substantial light scattering aerosol when photo-irradiated. The conclu-
sion to be drawn is that high sulfur-bearing fuels, when used even in catalyst-
equipped cars, can produce lower visibilities than non-catalyst cars being
operated on sulfur-free fuel.
A final point worth mentioning is that the addition of the additive
CI-2 to the reference fuel of Car B produced substantially lower visibilities
than without the additive. The suspected reasons for the catalytic effect of
the additive on the final observed visibility have already been discussed.
The data in Table III and those plotted in Figure 22 show that the points do not
lie along either of the straight lines for the fuels used in that vehicle.
There is insufficient data to state that the effect of the additive is conclu-
sive; however, based on the repeatability of the triplicate emissions and
methodology tests, there appears to be an effect which can be measured. As
in any experimental program involving a large number of variables, a greater
number of tests is needed to statistically validate these initial findings.
The recommended methodology appears capable of identifying the effects of indi-
vidual test variables. Large numbers of tests encompassing a sufficient matrix
of test variables would have to be performed if the nature and extent of the
visibility effect of an individual fuel or additive is to be ascertained.
45
-------
EPA REFERENCE FUEL -124 ppm SULFUR
CATALYST CAR WITH
1000 ppm SULFUR FUEL
OMMERCIAL INDOLENE FUEL
1.2
INITIAL HC/NO RATIO
Figure 22 FINAL VISIBILITY IN SMOG CHAMBER VS INITIAL HC/NO RATIO - LETTERS
DESIGNATE DATA (CARS A, B, AND F) FROM FINAL TEST SERIES USING
ADDITIVES IN EPA REFERENCE FUEL
46
-------
Section 4
SUMMARY AND CONCLUSIONS
It has been the objective of this program to develop a methodology
for determining the effect of fuel and fuel additive combustion products on
atmospheric visibility. Three 1972 5.73 liter (350 CID) Chevrolets and one
1973 catalyst-equipped 6.55 liter (400 CID) Ford Galaxie were used in the
study. As part of the methodology development, a wide range of experimental
conditions were tested in Calspan's 590 m (20,800 ft ) smog chamber. The
effects on test results of exhaust dilution ratio, relative humidity, added
SO-, primary particulates, evaporative emissions, and irradiation time were
investigated. From these investigations, a methodology was developed which
involves measuring visibility in a large smog chamber after first introducing
auto exhaust at a 300:1 dilution ratio, adding 0.05 ppm SO- and irradiating
the sample for 23 hours. In each case, vehicles were operated in an identical
manner on a chassis dynamometer while a fixed portion of the exhaust was diverted
to the chamber. From the triplicate emissions and smog chamber tests, the
following conclusions can be drawn:
(1) Using commercial grade indolene fuel, the effects on visibility
of the additives F-310 and CI-2 are small compared to the effects brought about
'by natural variations in engine performance.
(2) The combined effect of Ethyl CI-2 with low-lead indolene fuel
had a marked influence on the initial visibility lowering in the smog chamber.
No other combination of fuels or additives showed this effect.
(3) The additive CI-2 in combination with the lead-free, low sulfur
reference fuel produced substantially lower visiblities than without the addi-
tive. Catalytic effects of the additive on the photooxidation of SO- are sus-
pected.
(4) The tests show that primary particulates play an important role
in the initial and final visibilities noted in the smog chamber. Irradiations
of HC + NO + SO. mixtures comparable to those found in auto exhaust do not pro-
duce equivalent visibility losses.
47
-------
(5) The final visibility achieved after irradiation of a 300:1
exhaust plus filtered air mixture with 0.05 ppm added SO- is closely correlated
with the initial HC/NO ratio in the chamber. The correlation for the commer-
cial grade indolene is so good that final visibilities can be predicted from
an initial measurement of total HC and nitric oxide in the chamber. Data
generated using EPA reference fuel show a similar correlation of visibility
with HC/NO ratio, but the data lie on a different straight line. Although
only two fuel types were investigated in this study, the correlation seems
to be sufficiently high so that a family of lines could be established for
representative fuel types, those fuels with high sulfur content lying lowest
on the visibility scale.
(6) The catalyst-equipped vehicle operated on reference fuel
(124 ppm) sulfur) produced the least emissions but not necessarily the best
visibility in the smog chamber; i.e., the initial HC/NO ratio was still
the most important factor in determining the final visibility after irradia-
tion. There are limits to which this relationship can be carried, obviously,
since reductions in both the total HC and oxides of nitrogen in the natural
atmosphere would favor less severe smog.
(7) Operation of the catalyst car on high sulfur content (-1,000
ppm) reference fuel had a large effect on lowering the final visibilities
observed in the smog chamber. In fact, for the fuels and additives tested
at a given HC/NO ratio, the sulfur content of the fuel appeared to have the
most important effect on final visibility.
(8) In addition to having the expected initial effect on lowering
visibility, primary particulates in auto exhaust were found to play an impor-
tant role in long-term visibility reduction (i.e., after 23 hours irradiation).
From the tests performed on this program, it appears that a suffi-
ciently sensitive methodology has been developed to assess the effects of
fuels and additives on atmospheric visibility. While the tests were performed
exclusively in a smog chamber, it is believed that the uniquely large facility
provides sufficient realism to allow assessment of fuel and additive effects.
48
-------
The methodology appears well suited to testing the effect of variations in
catalyst-type, sulfur content of fuels, fuel composition and engine modifica-
tions. A baseline of data points has been established for comparison with
future experiments. Additional tests are needed to identify the most important
factors which influence total aerosol formation. With a sufficient data base,
it may be possible to develop a simple tailpipe measurement which could be
used to predict the effects of different fuel types on atmospheric visibility.
It is recommended that additional experimentation with fuels, engines
and emission controls be run to allow full understanding of the potential effects
of proposed improvements in auto emission control technology.
49
-------
TECHNICAL REPORT DATA
(Please read Instntctiont on the nvene before completing)
I. REPORT NO.
EPA-650/2-75-068
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
A Methodology for Determining the Effects of Fuels and
Additives on Atmospheric Visibility
5 REPORT OATE
June 1975
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
W.C. Kocmond, J.Y. Yang, and J.A. Davis
PERFORMING ORGANIZATION .REPORT NO
NA-5300-M-1
9. PERFORMING ORGANIZATION NAME AND ADDRESS
.Calspan Corporation
P^.O. Box 235
Buffalo, New York 14221
10 PROGRAM ELEMENT NO
1A1002; ROAP No. 26AAE
11. CONTRACT/GRANT NO
68-02-0698
12. SPONSORING AGENCY NAME AND ADDRESS
Chemistry and Physics Laboratory
Environmental Protection Agency
National Environmental Research Center
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 4/73 to 4/75
14 SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT ^~^~~~~~~~~~~~"~~"~"^
A methodology for determining the effects of fuels and additives on atmospheric visibility
has been developed using the smog chamber approach. The methodology involves measuring visi-
bility in a 590 m3 smog chamber after first introducing auto exhaust at a 300:1 detection ratio
adding 0.05 ppm S02 and irradiating the sample for 23 hours. Three 5.7 liter 1972 Chevrolets
and one 1973 catalyst-equipped 6.55 liter Ford Galaxie were used in the study. The effects on
test results of exhaust dilution ratio, relative humidity, added SC^, primary particulates,
evaporative emissions and irradiation time are discussed. The tests show that using commercial
grade indolene fuel, the effects on visibility of the additives F-310 and CI-2 are small com-
pared to the effects brought about by variations in engine performance. The presence of primar
particulates play an important role in the initial and final visibility noted in the smog cham-
ber. Irradiations of particle-free HC + NO + S0_ mixtures comparable to those found in auto
exhaust did not produce equivalent visibility losses. The final visibilities noted in the
smog chamber were found to be closely correlated with the initial HC/NO ratio. The correlation
for the commercial grade indolene is so good that final visibilities can be predicted from
the initial measurement of HC and NO in the chamber. Data generated using EPA reference fuel
show a similar correlation of visibility with HC/NO ratio, but the data lie on a different
straight line (due to lower sulfur content of fuel). For the fuels and additives tested at
a given HC/NO ratio, the sulfur content of the fuel appeared to have the most important effect
on visibility.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Visibility
Fuel and Fuel Additives
Exhaust Emissions
Smog Chamber
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
50
20. SECURITY CLASS (Thispage}
Unclassified
22. PRICE
EPA Form 2220-1 19-73}
------- |