EPA 600/3-83-087
PB83-262071
Characterization of Emissions and Fuel
Economy of In-Use Diesel Automobiles
New York State Dept. of Environmental
Conservation, Albany
Prepared for
Environmental Sciences Research Lab.
Research Triangle Park, NC
Sep 83
U.S. Department of Commerce
Rational Technical Information Service
-------
PB83-26207 1
EPA-600/3-83-087
September 1983
CHARACTERIZATION OF EMISSIONS AND FUEL ECONOMY OF
IN-USE DIESEL AUTOMOBILES
by
Richard E. Gibbs
James D. Hyde
Robert A. Wbitby
New York State
Department of Environmental Conservation
Albany, New York 12233
Delip R. Choudhury
New York State
Department of Health
Albany, New York 12201
EPA Grant R805934
Project Officer
Peter Gabele
Emissions 1 leasurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
TECHNICAL REPORT DATA
(Please read Inuructiom on the tei erse be/orr complain*!
1. REPORT NO.
EPA-600/5-85-087
3. RECIPIENT'S ACCESSION NO.
PB8 3 262071
4. TITLE AND SUBTITLE
CHARACTERIZATION CF EMISSIONS AND FUEL ECONOMY OF
IN-USE DIESEL AUTOMOBILES
6. REPORT DATE
September 1983
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
Richard E. Gibbs, James D. Kyde, Robert A. Whitby,
and Del ip R. Choudhury
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME ANO ADDRESS
New York State
Department of Environmental Conservation
Division of Air
Albany, N.Y. 12233
1O. PROGRAM ELEMENT NO.
C9YAlC/01-0458(FY-83)
11. CONTRACT/GRANT NO.
Grant R805934
12. SPONSORING AGENCY NAME ANO ADDRESS
Environmental Sciences Research Laboratory - RTP, N.C,
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
13. TYPE Of REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
6 ABSTRACT
Exhaust emissions from tv/enty 1977-1980 in-use light-duty diesel vehicles were
measured to determine the effects of driving cycle, mileage accumulation, and test
conditions. Hydrocarbons, CO, CO,,, NO and particulates were measured for the FTP,
HFET, CFDS, NYCC, 50 mph cruise(50t) an<3 idle cycles. Particulate extract was tested
for mutagenicity by the Ames method. Selected composite extracts were chemically
analyzed and bioassayed. Emissions (g/mi) and fuel consumption generally increased
in order 50C < HFET < CFDS < FTP < NYCC. GM vehicles generally had higher emissions
and more sensitivity to driving cycle than the Mercedes-Benz and VW vehicles.
Particulate extract emissions were not generally cycle dependent. NO emissions
decreased with mileage accumulation while other emissions increased or were
unaffected. Fuel economy was determined by the carbon balance method, by fuel meters
and by fueling records. Over-the-road fuel economy was always lower than carbon
balance fuel economy. A new method for real-time particulate measurement is de-
scribed using a Tapered -Element Oscillating Microbalance (TEOM). The TEOM mass was
within IQ% of the gravimetric mass with a response time of 8-15s. Effects of driving
cycle sequence, dilution tunnel, sub-FTP temperatures and mutagenic artifact forma-
tion were examined. Bulk extract samples were fractionated and analyzed by GC, GC/MS
and HPLC/UV. The acidic fraction had the highest specific activity, but most total
activity was in the neutral fraction which contained fluorenones and oxy-PAH's.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/CPEN ENDED TERMS
c. COSATI Field/Gio-jp
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS iTIiis Heporil
UNCLASSIFIED
21. NO. Of PAGES
20. SECURITY CLASS l't,,is puff/
UNCLASSIFIED
22. PRICE
EPA Fo"n 2220-1 (Rf. 4-77) PREVIOUS EDITION is OBSOLETE
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
PREFACE
Within the New York State Department of Environmental Conservation, the
Division of Air is charged with the responsibility of monitoring and regulating
atmospheric pollutants in the public interest to protect human health and
environmental quality. The scope of this task includes the investigation of specific
regulated and unregulated emission sources in order to assess environmental impacts.
Diesel powered automobiles are rapidly increasing as a percentage of the total vehicle
population and pose potential health and environmental quality risks which must be
understood and considered in air quality management.
The expertise in automotive emissions technology of the Automotive Emissions
Laboratory was employed to undertake this study of fuel economy and emissions
measurement from in-use diesel automobiles in order to determine typical emission
values and the effects of mileage accumulation on emissions. The results of this study
represent a valuable contribution to the development of air programs for the
protection of public health and the environment.
r\\. Hdvey /
Director, Division of Air
in
-------
ABSTRACT
A New York State research study on exhaust emissions tested twenty J977-1980
light duty diesel vehicles repeatedly over a two-year mileage accumulation period.
Measured emission parameters were hydrocarbons, carbon monoxide, carbon dioxide,
nitrogen oxides and particulates from the FTP, HFET, CFDS, NYCC, idle and 50 mph
driving cycles. Individual particulate samples were Soxhlet extracted with dichloro-
methane to partition the particulate into extract (soluble) and residue (insoluble). The
extracts were tested for mutagenicity by the Ames Salmonella
typhimurium/microsome method. Selected composite particulate samples were also
collected for detailed chemical analysis and subsequent bioassay.
Emissions (g/mi) ny driving cycle generally increased in the order 50C < HFET <
CFDS < FTP < NYCC and fuel economy decreased in that order. Vehicles in the
General Motors group generally had higher emissions than the Mercedes-Benz and
Volkswagen groups and were more sensitive to driving cycle. Particulate extract
emissions showed very little cycle dependence but residue emissions were very cycle
dependent. In general, emissions were unaffected by, or increased with, mileage
accumulation with the exception of NOX, which decreased.
Dynamometer fuel economy was determined by the carbon balance method.
Over-the-road fuel economy was determined by using fuel meters and vehicle
odometers and taking vehicle fueling records. Dynamometer fuel economy was always
higher than over-the-road fuel economy. The FTP best approximated average over-
the-road fuel economy.
A new method for the real-time measurement of diluted diesel particulate by a
Tapered Element Oscillating Microbalance (TEOM) is presented. The mean ratio of
TEOM results to those obtained by the FTP method using 47 mm filters was 0.96
(CV = 13%). Data for limited vehicle testing are presented.
Results of special experiments are presented for the effects of driving cycle
sequence on emissions, of the dilution tunnel on particulate measurement and
mutagenic activity, and of cold ambient temperatures. Particulate and extract were
re-exposed to diluted exhaust to determine the exposure-time effects on mutagenic
activity.
A discussion is presented of work conducted to isolate, identify and quantify the
chemical substances responsible for the mutagenicity of the extract. Bulk samples of
extract were fractionated and analyzed by GC, GC/MS, and HPLC/UV. The acidic
fraction had the highest specific activity, while most of the total activity was in the
neutral fraction due to its greater mass. Four-ring PAH's and their alkyl-substituted
homologues were predominant. Ketone, quinone, carboxaldehyde and hydroxy deriva-
tives and their alkyl homologues were characterized. The single largest family of
IV
-------
compounds detected was 9-fluorenone and its Cj-C^ alkyl homologues.
This report was funded by the U.S. Environmental Protection Agency under
Grant R805934. This is the final report for the period September 1, 1978 to March 31,
1982.
-------
CONTENTS
Preface jii
Abstract iv
Figures viii
Tables xiii
Abbreviations and Symbols xv
Acknowledgement - xvii
Project Collaboration xviii
1. Introduction 1
2. Conclusions 3
General Conclusions 3
Specific Conclusions ^
3. Experimental Approach 8
Introduction 8
Vehicle Sample Group 11
Mileage Frequency of Tests 11
4. Gaseous and Paniculate Emissions 16
Summary 16
Driving Cycle Effects 16
Mileage Accumulation Effects 25
5. Fuel Economy 50
Introduction 50
Cycle Variations of Fuel Economy 50
Mileage Accumulation Effects on Fuel Economy 50
Over-the-Road Fuel Economy 51
6. Bioassay Characterization 51
General Methodology 57
Bioassay Standardization 57
Factors Affecting Assay Results 59
Ames Test Results 63
7. Special Tests 75
Introduction 75
Effects of Driving Cycle Order 75
Dilution Tunnel 80
Exposure of Particulate and Extract to Dilute Filtered Exhaust. 82
Exposure of Particulate and Extract to Sunlight 87
Cold Ambient Particuiate Tests 89
8. Chemical Characterizaiton of Extracts 91
Introduction 91
Results and Discussion 91
9. Real-Time Particulate Measurements 104
References 112
Appendices
A. Tabular Emission Data 118
B. Typical Dose Response Curves for TA 98 (-) Ames Tests 131
C. Test Fuel Analysis 132
D. Experimental Procedures 135
E. Emission and Bioactivity Correlations m^
Vil
-------
FIGURES
Number
3.1 Mileage test point roster and mileage distribution of vehicle tests .... 11
3.2 Chronological distribution of vehicle tests 12
3.3 Monthly distribution of vehicle tests 12
4.1 Cycle variations of particulate, residue and extract for the General
Motors vehicle group 20
4.2 Cycle variations of particulate, residue and extract for the
Volkswagen vehicle group 20
4.3 Cycle variations of particulate, residue and extract for the
Mercedes-Benz vehicle group 20
4.4 Cycle variations for particulate by vehicle group: (a) actual
values, (b) normalized to the FTP 21
4.5 Cycle variations of residue by vehicle group: (a) actual
values, (b) normalized to the FTP 21
4.6 Cycle variations of extract by vehicle group: (a) actual
values, (b) normalized to the FTP 23
4.7 Cycle variations of % extractible by vehicle group: (a) actual
values, (b) normalized to the FTP 23
4.8 Cycle variations of hydrocarbons by vehicle group 24
4.9 Cycle variations of carbon monoxide by vehicle group 24'
4.10 Cycle variations of nitrogen oxides by vehicle group 24
4.11 Mileage accumulation effects for FTP particulate by vehicle group ... 27
4.12 Mileage accumulation effects for FTP residue bv vehicle group 28
4.13 Mileage accumulation effects for FTP extract by vehicle group 30
Vlll
-------
Number Page
4.14 Group emission envelopes for participate, residue and extract 31
4.15 Mileage accumulation effects for FTP hydrocarbon by vehicle
group - 32
4.16 Mileage accumulation effects for FTP carbon monoxide by vehicle
group 34
4.17 Mileage accumulation effects for FTP nitrogen oxides by vehicle
group 35
4.18 Mileage accumulation effects for FTP % extractible by vehicle
group 36
4.19 Cycle variations of fuel specific paniculate by vehicle groups 38
4.20 Cycle variations of fuel specific residue by vehicle groups 38
4.21 Cycle variations of fuel specific extract by vehicle groups 38
4.22 Frequency distributions of %V for FTP gaseous parameters 44
4.23 Frequency distributions of 96V for FTP participate parameters 44
4.24 Frequency distributions cf %V for HFET gaseous parameters 45
4.25 Frequency distributions of %V for HFET participate parameters 45
4.26 Cycle variations of non-volatile hydrocarbons by vehicle groups 48
4.27 Mileage accumulation effects for average of FTP, CFDS, HFET and
50C non-volatile hydrocarbons by vehicle group 49
5.1 Cycle variations of fuel economy, miles per gallon, by vehicle
group 51
5.2 Cycle variations of fuel economy, ton-mile/gallon, by vehicle
group 51
5.3 Mileage accumulation effects for FTP fuel economy, mpg, by vehicle
group 52
5.4 Over-the-road fuel economy vs. speed for Car //I 55
5.5 Over-the-road fuel economy vs. speed for Car //5 55
5.6 Over-the-road fuel economy vs. speed for General Motors group 56
6.1 Chronology of Ames activity of 5 ug dose of 2-NF(2-nitrofluorene) ... 58
IX
-------
Number Page
6.2 Chronology of Ames activity of 0.5 ng dose of NQNO
(4-nitroquinoline-N-oxide) ...... . ............................. 58
6.3 Chronology of Ames slope of diesel standard extract (25, 50, 75,
100 ug doses) .............................................. 59
6.4 Effects of long-term sample storage on mutagenic response -
storage as extract in DCM vs. storage as filter ................. 60
6.5 Effects of long-term sample storage on mutagenic response -
storage as extract vs. storage as filter ........................ 60
6.6 Effect of incubation period on mutagenic response .................. 62
6.7 Reproduce bility of mutagenic response for ten replicate plates
' of diesel standard extract ................................. ... 62
6.8 Reproducibility of mutagenic response for replicate 50C filter
samples [[[ 62
6.9 Mileage accumulation effects for averaged Ames activity ........... 64
6.10 Cycle variation of average Ames activity for all vehicle groups ...... 65
6.1 1 Mileage accumulation effects for extract and Ames activity ......... 72
6.12 Cycle variations of residue, extract and revertants/mile normalized
to the FTP for the General Motors group ....................... 74
6.13 Cycle variations of residue, extract and revertants/mile normalized
to the FTP for the Volkswagen group .......................... 74
6.14 Cycle variations of residue, extract and revertants/mile normalized
to the FTP for the Mercedes-Benz group ....................... 74
7.1 Variation of HFET and 50C particulate emissions for consecutive
runs [[[ 77
7.2 Variation of IDLE particulate emissions for consecutive runs ......... 77
7.3 Variation of FTP gaseous emissions and fuel economy on successive
days ................ . ...................................... ... 79
7.4 Locations of particulate samples removed from the dilution tunnel .... 81
-------
Number Page
8.1 (A) Fractionation of diesel emission particulate extracts
(B) Eiutant profile for chromatographic f ractionation of
neutral diesel particulate extract . . .............................. 93
8-2 Mutagenicity of diese! particulate extract sample S-l and
its fractions (Car //I) ........................................... 93
8.3 Silica gel f ractionation of neutral diesel particulate extract
8.4 Distribution of mutagen:city among sample S-l subtractions
of neutral diesel particulate extract of Car //I
8.5 Gas chromatogram of a mixture of PAH standards .................. 96
8.6 Gas chromatogram of (A) PAH fraction of ciesel particulate
' extract (S1-C2) and (B) its HPLC subtraction of C (S1-C2) .......... 97
8.7 Total ion chromatogram of the PAH fraction S 1-C2 ................. 98
8.8 Total ion chromatogram of the PAH fraction S2-C2 ........ ......... 98
8.9 HPLC profile of PAH fraction of extract sample S-l from Car //I ---- 98
8.10 Ultraviolet spectra of HPLC eluates of PAH fraction of extract
from Car //I ............. ...................................... 98
8.11 (A) HPLC profile of high-molecular-weight PAH fraction of
sample S-l extract of Car if I
-------
Number Page
9.6 Reproducibility of TEOM mass rate (ug/sec) for three consecutive
trials of the NYCC schedule by Car //5 (Oldsmobile 5.7 liter diesel) ... 109
9.7 Reproducibility of TEOM mass accumulation (10"^ grams) for
three consecutive trials of the NYCC schedule by Car //5
(Oldsmobile 5.7 liter diesel) 109
9.8 TEOM mass rate (ug/sec) and vehicle acceleration (mph/sec) for
Car #5 (Oldsmobile 5.7 liter diesel) over the FTP Bag 3 schedule 110
9.9 TEOM mass rate (ug/sec) vs. time at three dynamometer inertial
weight settings for Car //5~(bldsmobile 5.7 liter diesel) driven over
the FTP Bag 3 schedule Ill
9.10 Accumulated TEOM mass (10-'* grams) vs. time at three
• dynamometer inertia! weight settings for Car //5 (Oldsmobile 5.7
liter diesel) driven over the FTP Bag 3 schedule Ill
xu
-------
TABLES
Number Page
3.1 Vehicle Specifications and Dynamometer Test Conditions 10
3.2 Vehicle Sample Group by Model Year 12
3.3 Vehicle Test Driving Cycle Sequences 1*
4.1 Summary of FTP Particulate and Gaseous Emissions-Phase 3 17
4.2 Summary of CFDS Participate arid Gaseous Emissions-Phase 3 17
4.3 Summary of HFET Particulate and Gaseous Emissions-Phase 3 17
4.4 Summary of 50C Particulate and Gaseous Emissions-Phase 3 17
4.5 Summary of NYCC Particulate and Gaseous Emissions-Phase 3 18
4.6 Summary of IDLE Particulate and Gaseous Emissions-Phase 3 18
4.7 Mileage Statistics for Test Vehicles and Vehicle Groups 26
4.8 Mean Fuel Specific Emission Parameters by Vehicle Group
and Test Mileage 39
4.9 Significance Level (o Level) for Manufacturer Fuel Specific
Emission, All Mileage Means 37
4.10 Significance Level (a Level) for Mean Fuel Specific Emissions
by Test Mileage Range 40
4.11 Ratio of Phase 1 to Phase 3 Emissions 41
4.12 FTP %V by Vehicle Group and Test Mileage Range 43
4.13 Significance Level (o Level) for Differences in %V Means
by Mileage Range and Between Vehicle Groups 46
5.1 Fuel Economy and Vehicle Speed from Underhood Meters 54
6.1 Effect of High Doses on Slope of Dose-Response Curve 62
6.2 Ames Activity, Group-Phase Ratios 63
xiii
-------
Number Page
6.3 Extract/Residue Sample Correlation Coefficients by Driving
Cycle and Vehicle Group Over All Test Phases 67
6.4 Bioactivity/Emission Parameter Sample Correlation Coefficient
Signs and Significance Levels by Driving Cycle and Vehicle Group
Over All Test Phases . 69
6.5 Bioactivity/Emission Parameter Pairs with Sample Correlation
Coefficient Absolute Values in Excess of 0.60 70
7.1 Sequence of Driving Cycles for Schedule Change Experiment 76
7.2 Day-to-Day Variations of FTP Particulate Emissions 76
7.3 Particulate Emissions from HFET, 50C and IDLE - Day-to-Day
and Repetition Variation 77
7.it Reproducibility of 50C Particulate, g/mi 78
7.5 Day-to-Day Variations of FTP Gaseous Emissions 79
7.6 Day-to-Day Variations of HFET, 50C and IDLE Gaseous Emissions .... 80
7.7 Fuel Economy Variations for Duplication Test, mpg 80
7.8 Ames Activity of Dilution Tunnel Particulate 81
7.9 Exposure of Particulate to Dilute Exhaust 83
7.10 Emissions Data from Particulate Exposure Tests 8*
7.11 Exposure of Extract to Dilute Exhaust Gas - Car #5 85
7.12 Exposure of Extract to Dilute Exhaust - Car //I 86
7.13 VW Cold-Start Particulate Comparisons 90
8.1 Summary of Large Particulate Samples for Chemical and Bioassay
Characterization 92
8.2 Compounds Characterized by El GC/MS of Sample S1-C2 96
8.3 Compounds Characterized by GC/MS in Subtraction 4 102
8.4 Compounds Characterized by GC/MS in Subtraction 5 102
9.1 Comparison of Total Mass Emission Determinations by TEOM and
Standard 47 mm Filter Gravimetric Methods for Particulates from
Diesel Vehicles 106
xiv
-------
ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS OF UNITS
cfm cubic feet per minute
cm centimeter
g gram
in inch
Kg kilogram
Kw kilowatt
L, J liter
ib pound
m meter
mi mile
min minute
ml milliliter
mm millimeter
mph miles per hour
sec second
Mg microgram
ABBREVIATIONS OF THINGS
A Automatic
AEL Automotive Emissions Lab
BaP benzo(a)pyrene
C . carbon
CFDS Congested Freeway Driving Schedule
CI chemical ionization
CO carbon monoxide
CV coefficient of variation
CVS Constant Volume Sampler
DCM dichloromethane
DISP displacement
dm quantity of mass
DMSO dimethylsulfoxide
DPE diesel particulate extract
Ei - electron impact
EPA/RTP Environmental Protection Agency,
Research Triangle Park
EXT extract
f frequency
FTP Federal Test Procedure
xv
-------
GC
GC/MS
GM
HC
HFET
HFID
Hg
HP
HPLC
IDLE
I.W.
K
KgF
M
MB, M-B
MIDPT
MPG
MPH
MS
N
NBS
NOX
NQNO
NVHC
NYCC
OLDS
bxy-
P, Part, PT
PAH
POP
RES
REV
SOF
STD DEV
S9
TA98, TA100
TEOM
UV
VW
2-NF
50C
gas chromatography
gas chromatography/rnass spectrometry
General Motors
hydrocarbon (by HFID)
Highway Fuel Economy Test
hot flame ionization detector
mercury
horsepower
high pressure liquid chromatography
idle in neutral gear
inertia! weight
kilo, thousand
kilogram fuel
manual
Mercedes-Benz
midpoint
miles per gallon
miles per hour
mass spectroscopy
number of items
National Bureau of Standards
nitrogen oxides (as nitrogen dioxide)
^f-nitroquinoline-N-oxide
non-volatile hydrocarbon
New York City cycle
Oldsmobile
oxygenated
particulate
polynuclear aromatic hydrocarbons
positive displacement pump
residue
revertant bacterial colonies
soluble organic fraction
standard deviation
liver homogenate fraction
bacterial tester strains
Tapered Element Oscillating Microbalance
ultraviolet
Volkswagen
2-nitrofluorene
50 mph steady cruise
SYMBOLS
°C
n
_R2
x
a
%V
degree Celsius (centigrade)
number of items in group
correlation coefficient
mean (arithmetic)
standard deviation
percent emission change per 1000 miles
without S9
xvi
-------
ACKNOWLEDGEMENTS
The research team expresses appreciation to the many Air Division and Department
staff who, each in their own way, aided the success of this project over several years
time.
The stenographic/word processor services of Stephanie Liddle and Linda Stuart
throughout this report preparation, and the cartographic work of Gary Lanphear and
Carol Clas throughout the project, are sincerely appreciated.
The cooperation of individuals within the NYS Department of Health contributed
significantly and are gratefully acknowledged.
Staff support from the EPA Mobile Sources Research Branch was an invaluable aid in
the development and conduct of the project.
Cooperation from the NYS Thruway Authority for vehicle access was especially
valuable in achieving project goals. Finally, the private citizens who entrusted their
vehicles for iosting purposes are to be thanked for their unremunerated cooperation,
and making possible the repetating study of "in-use" diesel vehicles.
xvu
-------
PROJECT COLLABORATION
A project of this scale succeeds only by the cooperative effort and sharing of
ideas from a diverse group of dedicated people. Project accomplishments are here
properly attributed to a teamwork approach by the following project members:
Name
Richard Gibbs
James Hyde
Robert V/hitby
Robert Johnson
Paul Werner
Ben Hill
Stan Byer
Tom Hoffman
Delip Choudhury
Brian Bush
Charles Doudney
Mary Franke
Charlie Rinaldi
Organization
Environmental Conservation
Environmental Conservation
Environmental Conservation
Environmental Conservation
Environmental Conservation
Environmental Conservation
Environmental Conservation
Environmental Conservation
Health Department/Toxicology
Health Department/Toxicology
Health Dept./Env. Health
Health Dept./Env. Health
Health Dept./Env. Health
Area
Project Leader
Lab Data/Project Data Analysis
Computer Analysis/TEOM
Vehicle Testing/TEOM
Vehicle Testing
Computer Analysis
Project Logistics
Chemical Laboratory
Extract Chemical Analysis
Extract Chemical Analysis
Bioassay
Bioassay
Bioassay
A vast array of organizational support staff in both participating New York State
Departments played a major role in making the work possible.
XVlll
-------
SECTION 1
INTRODUCTION
The control of mobile source emissions to achieve air pollution goals has been
one of the significant technical/social challenges of recent years. The multitude and
diversity of these emission sources, the relative difficulty of reducing emissions of
some pollutants from in-service vehicles and the co-existence of vehicles and people in
urban settings all have impacted approaches taken to minimize the effect of mobile
source emissions on air quality.
Mobile source emission programs began essentially as a gasoline-powered vehicle
emissions control program. These efforts now must not only adapt to the increased
diversity of gasoline combustion technologies, but must also address the presence of a
significant diesel-powered component in the vehicle population. In New York State
the number of diesel-powered light-duty vehicles in 1975 was a mere 3,265, but by 1980
this number had increased at an average annual geometric rate of about 68% to
become 44,122, an increase from 0.05% to 0.64% of this total vehicle population in a
five year period (1). This increase has occurred during a period of sales depression in
the general new-car market. The significantly different emissions of diesels, coupled
with their recent accelerated penetration into all vehicle weight classes, thus
represents an emerging factor in the mobile source picture that demands adjustment
to the process of developing policies to optimize the nation's transportation/energy/air
quality balance.
How many diesels will there be? What will be their emissions impact? How will
diesel related emission technology evolve? Should diesels continue to be exempt from
inspection/maintenance requirements? What are the composite HC, CO, NOX, and
paniculate impacts on ambient air quality levels for various possible future diesel
scenarios? In concert with other emissions, wnat visibility and health impacts should
be expected?
These are not all the questions, but they typify the magnitude and uncertainty of
the impact that present events will make for some future time, a time when little can
be done but accept the vehicles and their elfects in retrospect. Thus the need to
develop a technical viewpoint to aid in the identification of important diesel issues
was an underlying motivation behind this joint New York State/Environmental Pro-
tection Agency research study described in this project report.
The New York State diesel study grew from the need for comprehensive
emissions test data on in-use diesel automobiles. Comprehensive, in this context,
refers to mass emissions testing for HC, CO, NOJ:, fuel economy, paniculate by the
identical tests used by vehicle manufacturers to certify new vehicles - the Federal
Test Procedure (FTP), and additional investigations regarding the characteristics of
the emitted particulate. Other types of vehicle operation were included to obtain the
1
-------
same types of emission data for conditions ranging from New York City driving to
highway operation. Thus tailpipe participate was measured for various vehicle types,
vehicle mileage/ages, commonly used fuel and lubricating oils, and vehicle operating
modes. Particulate samples collected for each of the many combinations of
parameters mentioned were individually solvent extracted to determine the
"extractible" and "residue" make-up of the total particulate. The extracts (or soluble
organic fractions) were then subjected to a bioassay screening test to determine the
relative mutagenic potency as functions of many of the test parameters. All data
were entered into a computer data base from which analyses, correlations and
tabulations were performed.
A parallel effort to the in-use vehicle study was directed toward chemical
identification of the mutagenic species present in the extract samples. The chemistry
of the extracts is analogous in complexity to cigarette smoke, and the mutagenic
activity is not attributable to the major components. Thus chemical identification of
the mutagenic species is a difficult task and only a few select samples were
extensively examined. Three project vehicles (Oldsmobile, Volkswagen, and Mercedes-
Benz) were operated on the dynamometer for a tota4"of 50 days to collect sufficient
particulate (50-100 g of particulate each) to obtain, after solvent extraction, the
quantity of extract needed for chemical characterization work.
The in-use vehicle regulated emissions testing, experimental developments in
obtaining particulate emissions data, chemical characterization of large samples,
bioassay results from both in-use vehicle tests and large particulate samples, and
analysis of project data base results all are significant aspects of this study. This
report summarizes the study and its findings shortly after completion of the data
collection effort. Further examination of the data base, chemical analyses on the
archive of project extract samples, etc. may well extend the application of these data.
The project was not intended to directly address the policy type of questions listed
above, but was designed to contribute technically to their understanding.
The analyses and discussions in this report use only a portion of the data
collected in this project. Many other worthwhile analyses could be made but were not
made due to time restraints. In general the analysis has been limited to Phase 3 data
(controlled conditions) and in many cases to only FTP Phase 3 data (e.g. in the mileage
accumulation section). In many cases a preliminary analysis of the data showed
extension to another phase or to other cycles would not significantly contribute to the
results of this study. Vehicle averages for 12 parameters for Phase 1 and Phase 3 are
ghen in Appendix A. In addition, this appendix includes data for six parameters from
ea::h of the three individual bags of the FTP. These data are not discussed in the
report.
Samples of all fuels used in the project were extensively analyzed. Although
many instances were found in which fuel effects were apparent, no detailed analysis of
fuel composition effects on emission parameters have been attempted.
Almost all Ames data discussed represent strain TA98 testing without activation
although many samples were analyzed with activation and with other strains. The
bioactivity-emissions correlation study lumped all Phases together in order to achieve
the largest possible data base. The study was limited, however, to determination ol
linear correlation coefficients. Non-linear or multivariable analysis of the data base;
must be left lor future work.
-------
SECTION 2
CONCLUSIONS
The New York State research team investigated the emission characteristics of
light-duty diesel passenger vehicles typical of the 197S-1979 period when the study
began. Many results apply to diesel emissions in general; but others are specific to
this vintage vehicle and should not be extrapolated to other cases. This section is
composed of those general conclusions felt to relate to the diesel emission field, and
specific conclusions from the experimental work performed.
GENERAL CONCLUSIONS
Particulate, CO, HC, and NOX emissions (g/mi) increased significantly for New
York City driving conditions relative to the certification driving cycle for the Federal
Test Procedure.
Particulate emissions for some vehicles exhibited significant increases with
mileage accumulation whereas other vehicles remained relatively constant (no deterio-
ration) throughout the period of the project.
When significant mileage accumulation increases in particulate mass emissions
were observed, the increase was predominantly due to increases in the extractible (not
residue) portion of the particulate.
Total hydrocarbon emissions are composed of both gas and condensed (as part of
particulate) phase material. The particulate-bound portion was 20-60% of the total
hydrocarbon emission with the 50C having the highest values and the IDLE the lowest.
Driving cycle, vehicle type, and vehicle age accounted for the variation range.
Measurement of diesel engine particulate mass emissions in real-time is now a
possibility. This development impacts several areas of diesel emissions understanding
previously limited by use of long-time filter sampling. Some areas wheie this
capability can be expected to have application: 1) focusing engine development work
on emission prone modes, 2) correlation of vehicle operation characteristics with
certification test cycles, and 3) developement of short vehicle tests suitable for on-
road vehicle emissions projections.
The bioassay potency of particulate extracts (by Ames tester strain TA98
without activation) exhibited considerable variation with project variables, but was
positive for all samples (over 1100) analyzed.
Direct correlation of bioassay activity with specific vehicle-test parameters was
not achieved, but a complete examination of the data base for possible correlations
-------
has not been performed.
For all General Motors vehicles as a group, Ames test activity correlated very
significantly with the residue (and not with the extractibla) portion of the particulate
samples. This finding was true to a lesser extent for other vehicle groups but was
emphasized in the General Motors case by the higher extractible content of the
particulate.
Mileage accumulation plots of % extractible and extract bioassay activity
(revertants/ ug extract) showed these two parameters to be inversely related for
virtually all vehicles.
The solvent extractible portion of the particulate is mainly derived from
absorption of hydrocarbons and hydrocarbon derived species onto a carbon core formed
in the combustion process. However, some extractible is certainly formed in the
combustion chamber along with the carbon core and this material may contain most of
the direct acting mutagens. The bulk of the extract is primarily a diluting medium for
the small amount of biologically active material.
The polynuclear aromatic hydrocarbon (PAH) emission products cannot account
for the observed direct acting mutagenicity of extracts, but the nitro-PAH and oxy-
PAH species likely combine to account for much of the Ames activity.
The oxy-PAH species present in typical samples from this project have been
characterized in new levels of detail, and thus contribute to the advancement of a
chemical basis for understanding diesel emissions. The reaction processes, whether in
combustion, emission, sampling, extraction, bioassay or atmospheric zones, leading to
nitro-PAH and oxy-PAH should be active topics for future diesel research .
Comparisons of over-the-road fuel economy to laboratory carbon balance fuel
economy showed the laboratory measure to be approximately 15-20% above the over-
the-road case at corresponding vehicle average speeds. Road grade, temperature,
snow, wind, etc. can easily account for this increment.
The FTP laboratory value corresponded most closely to the over-the-road value
even though the average vehicle speed in the FTP is much lower than for over-the-
road operation. This speed discrepancy thus approximately caused the same fuel
economy effect as the real-world factors not included in the laboratory test.
SPECIFIC CONCLUSIONS
Driving Cycle Effects
(1) For Car //5, a 1979 Oldsmobile Cutlass Cruiser with 5.7 L engine, it was
found that:
a) the cycle driven prior to the FTP cold soak had no effect on gaseous
emissions and had little effect on particulate emissions except when that cycle was an
IDLE;
b) previously driven IDLEs increased the particulate emissions of subsequent
driven cycles but did not affect gaseous emissions;
c) gaseous measurements were not aff'.cted by the previously driven cycles;
d) gaseous measurements when repeated on separate days had a coefficient of
-------
variation, CV, of less than 5% except for the IDLE where the CV's were several times
larger;
e) the fuel economy of a cycle increases with increased repetition of that
cycle during a day.
(2) For all vehicle groups, driving cycles affect emissions in a consistent
manner. The general trend of decreasing emissions was NYCC > FTP > CFDS > HFET>
50C.
(3) The Genera1. Motors group had emissions which were-much more sensitive to
driving cycles than were the Volkswagen and Mercedes-Benz groups.
(*) With the exception of residue for the HFET and 50C, the General Motors
group had greater g/mi emissions than the Volkswagen or Mercedes-Benz groups.
(5) Cycle variations of particulate, g/mi, were principally due to variations in
residue mass for all vehicle groups.
(6) Residue, g/mi, was very cycle dependent for the General Motors group but
only slightly cycle dependent for Volkswagen and Mercedes-Benz groups. It increased
in the order 50C < HFET < CFDS < FTP < NYCC.
(7) Extract, g/mi, showed very little cycle dependence except for a very large
increase for the NYCC for all vehicle groups. The CFDS, HFET and 50C had similar
extract, g/mi, emissions with a slight increase (0.02 g/mi overall) in the order CFDS <
HFET < 50C for all vehicle groups.
(8) Extract, %, showed no cycle variations for the Mercedes-Benz group, few
variations for the Volkswagen group and strong cycle dependence for General Motors
group.
(9) The HFET and 50C had very similar gaseous emissions except for hydro-
carbons from the General Motors group.
Mileage Accumulation Effects
(1) FTP particulate emissions did not show a mileage related deterioration
(increase) for the Volkswagen group and the Mercedes-Benz groups, but showed a large
deterioration for the General Motors group due to a large increase in extractible
emissions.
(2) FTP hydrocarbon emissions did not show a mileage related deterioration for
the Volkswagen and Mercedes-Benz groups, but showed a deterioration for the General
Motors group.
(3) FTP carbon monoxide emissions showed a deterioration of the General
Motors and Volkswagen groups but not for Mercedes-Benz group.
M FTP nitrogen oxides emissions showed a decrease with accumulated mileage
for the General Motors and Mercedes-Benz groups but no trend for the Volkswagen
group.
-------
Fuel Economy
0) Over-the-road fuel economy as measured by underhood meters between
vehicle tests was approximately 15-20% less than laboratory carbon-balance fuel
economy at comparable average vehicle speeds.
(2) Over-the-road fuel economy, while of higher average vehicle speed than the
FTP, was most closely approximated by FTP economy.
(3) Laboratory fuel economy increased v/ith increased average speed of the
driving cycle, i.e. NYCC
-------
activity parameters were generally very weak although often statistically significant.
A certain degree of consistency in the sign of the linear correlation coefficient was
observed over most vehicle types and driving schedules for certain parameter pairs.
Revertants per ug extract was the bioactivity parameter which most frequently
yielded a statistically significant correlation coefficient. The correlation coefficient
for revertants/ pg extract and emissions were typically negative for extract emission
parameters and positive for residue emission parameters.
Real-Time Particulate Measurement
(1) The first successful test of a prototype real-time diesel particulate mass-
measurement instrument was achieved. This instrument, a Tapered Element
Oscillating Mjcrobalance (TEOM) was used in dilution tunnel sampling and demon-
strated correspondence with standard filter collection while providing real-time
resolution of particulate emissions.
(2) The time resolution achieved during this test of the TEOM was 8 seconds,
and with further work this time constant may be decreased.
(3) TEOM data may be obtained either directly from the instrument micro-
processor or from data storage on a host computer. When the latter method is
employed, the TEOM raw data signal may be reanalyzed by user defined algorithms to
obtain new information from existing test data.
Fuel Specific Emissions
(1) Fuel specific particulate (g/kg Fuel) was fairly constant for all driving cycles
for Mercedes (s 3.5-^.0 g/kg Fuel).
(2) Fuel specific particulate for Volkswagen group showed an increasing trend
with higher speed cycles (3.5-5 g/kg Fuel); whereas General Motors group vehicles
showed a decreasing trend with higher speed cycles (5.5-4 g/kg Fuel).
(3) The main variations in fuel specific emissions were due to changes in residue
and not extract emissions.
Chemical Characterization of Extract
0) The acidic fraction of the extract had the highest specific activity but most
of the total activity was in the neutral fraction due to its greater mass.
(2) Four-ring PAH's and their alkyl-substituted homologues were the predomi-
nant constituents.
(3) 9-fluorenone and its Cj-C^ alkyl homclogues was the single largest family of
compounds.
-------
SECTION 3
EXPERIMENTAL APPROACH
INTRODUCTION
The technical literature contains numerous reports of diesel emissions investiga-
tions, and no attempt will be made here to provide a complete literature review. The
bulk of literature on this subject is found in the publications of the Society of
Automotive Engineers (SAE). Collections of SAE papers on diesel emissions are
available (2-10). Works are available on measurement of emissions (11-15), health
effects (16-24) and policy decision (25-27). Some historical perspective will be given to
provide the reader a sense of the context at the time this study was undertaken, and
the references listed for this section provide access to the general body of diesel
emissions literature.
Prior to the mid-1970's, diesel combustion was viewed primarily in relation to
heavy-duty vehicles. Research reports on emissions for these types of engines dealt
with participate emission measurement, emission effects of fuel properties, chemical
characterization of extracts, particulate size data, odor characterization, and opacity
measurements. Some detailed chemical characterization work was undertaken at this
time, but the emphasis tended to be on known toxic or carcinogenic species such as
benzo(a)pyrene. "Beginning in the rnid-1970's there was increased emphasis on light-
duty diesel emissions from passenger cars as well as continued heavy-djty character-
ization studies, measurement technique development, and physical characterization.
The first reports of positive bioassay indications from diesel particulate extract
by the Ames test were paralleled by projections of the increasing use of diesel engines
for light-duty vehicles. These combined factors directed much research into the
complex organic mixtures obtained from solvent extraction of particulate samples.
The chemical species responsible for these Ames test indications were soon found not
to be the compounds of historical attention, such as BaP, and the search for these
chemical mutagens has been a major theme in research efforts since the Ames test
was applied to diesel emissions. The validity of sampling procedures, the potential for
artifactual mutagen formation, the evolution of bioassay methods as research tools,
and the partial correlation of bioassay data with nitro-PAH species have all been
active topics since the chemical quest for the mutagen identification began. The
important questions about what happens to diesel particulate once.emitted to the
atmosphere (as contrasted to a dilution tunnel) form an entire field in complexity,
experimentation and interpretation.
Particulate emission standards and measurement protocol for certification of
1981 and later model year new cars were promulgated by EPA in 1979. While the
measurement protocol does not address sampling for either mutagens-or PAH's, the
mass emissions measurement protocol has come into general use by most researchers.
8
-------
The new car certification dat? have been complemented by very little in-use diesel
particulate data obtained by re-testing after mileage accumulation.
The present study was thus designed to explore several aspects of the diesel
emissions field by examining a diverse sample of consumer operated diesel vehicles
over a period of significant mileage accumulation. Information was needed, for
instance, on the following questions to furiner the understanding of diesel emissions:
How do particulate mass emissions vary among vehicles, with the age of vehicles,
among type of vehicle operations, and for fuel/oil types encountered among in-use
vehicles? Beyond mass emission questions, how does particulate character
(extractible, residue, bioassay activity) change with vehicle operation and fuel/oil
variations? Which chemical species in particulate extract are significant in the
extract bioassay results? How can vehicle particulate emissions be measured in real-
time rather than as integrated test-cycle averages from filter samples? How can the
accumulated data from in-use diesels be analyzed to reveal trends that connect
project data elements that would be isolated pieces of information if taken alone?
VEHICLE SAMPLE GROUP
The vehicle sample for this study included 21 -n-use diesels. Most of these
vehicles were tested repeatedly over a 28-month period of mileage accumulation.
Since ail vehicles in the sample group did not accumulate mileage at the same rate,
the total number of vehicle tests were not the same for all sample group vehicles. The
mileage accumulation intervals between vehicle tests averaged 12,000 miles, but no
strict mileage interval was used, in response to the varieties of driving encountered by
the vehicles. Since a goal of the study was to retest the same vehicles, the vehicles in
the study did not include new diesel engine types that were introduced to the market
during the course of the study.
Table 3.1 lists each sample group vehicle by a project "Car //" identifier which is
used in subsequent data presentations. Table 3.1 also lists the make, model, model-
year, engine displacement/cylinder configuration, transmission type, and emission
chassis dynamometer road-load and inertia settings used ior testing.
Referring to sample group vehicles by the Car // identifiers, some further
information on the background of the vehicles may provide useful information for
understanding project results: Cars //I and #5 were procured by New York State to
provide loan vehicles to private vehicle owners who permitted their vehicles to be
tested in the project. These two vehicles were also used extensively for project
testing needs for collection of large particulate samples, protocol development, etc.
These vehicles received normal maintenance by project staff.
Cars in. and #3 were operated by the New York State Thruway Authority as part
of their normal administrative fleet. Car #10, a Thruway Authority Dodge pickup
truck equipped with a Mitsubishi diesel engine, was included in the study to obtain
comparison data from an engine type not in widespread use.
Car //ll, a privately owned Volkswagen Rabbit, entered the project with 120,000
miles, and was last tested at 170,000 miles, and thus was the only project vehicle with
testing at such high mileage.
Car #18 was the only 1980 model-year 5.7 L Oldsmobile diesei in the sample
9
-------
group. This vehicle did not have emissions which were typical of the 1979
Oldsmobiles. Since it wai the only 1980 Oldsmobile in the group, and numerous design
changes were made between the 1978-79 and 1980 5.7 L engines, subsequent groupings
of General Motors 5.7 L engines in this paper exclude this vehicle. Thus the emissions
data presented for the GM 5.7 L diesels pertain only to 1979 model year designs.
Car 1/20 was a 1978 General Motors 5.7 L diesel tested only once to evaluate the
effect of an experimental crankcase oil additive, and thus the results from this test
are not included in any composite emission results presented for the General Motors
5.7 L diesel vehicles.
Car #21, a 1973 Dodge van retrofitte^ in 1975 with a Chrysler-Nissan CN6-33
diesel engine, was tested to obtain comparison data. This type of engine was used in
the New York City diesel taxicab study (28).
Where applicable throughout this report, the vehicles have been divided into
groups according to manufacturer. The groups are:
GM: Cars// 2, 3, 1, 5, 7 and 16;
VW: Cars// 1, 6, 8, 9, and 11;
MB: Cars// 12, 13,14, and 19;
OTHER: Cars//10,15,17,18, and 21.
The privately owned in-use vehicles were solicited by letter to diesel vehicle
registration holders in the Albany, New York area. In consideration for the use of
their vehicle for testing purposes, the owner received a loan vehicle during the test, an
oil/filter change as part of the test, a full fuel tank upon the return of their vehicle,
and two car wash tickets.
TABLE 3.1 - VEHICLE SPECIFICATIONS AND DYNAMOMETER TEST CONDITIONS
CAR tl YEAR MAKE
^^—^ — - — —
1
3
5
6
7
8
9
10
11
12
13
15
16
17
18
19
20
21
MODEL
"79-
79
79
79
79
80
79
78
79
78
77
77
78
79
79
79
79
80
79
78
78
VW
OLDS
OLDS
OLDS
OLDS
VW
CADILLAC
VW
VW
DODGE
VW
M-B
M-B
M-B
AUDI
OLDS
PEUGEOT
OLDS
M-B
OLDS
DODGE
Rabbit
Cutlass Cruiser
Cutlass Cruiser
98 Regency
Cutlass Cruiser
Rabbit
Eldorado
Rabbit
Rabbit
D-IO Mitsubishi
Rabbit
240-D
300-CD
240-D
5000
Delta 88
504
Cutlass Cruiser
300-SD (Turbo)
Delta 88
Tradesman 200
ENGINE
ISPLACMENT
1-4
V-8
V-8
V-8
V-8
1-4
V-8
-4
-4
-6
-4
.it
-5
-4
-5
V-8
1-4
V-8
1-5
V-8
1-6
1.5L
5.7 L
5.7 L
5.7 L
5.7 L
1.5L
5.7 L
1.5 L
1.5L
4.0 L
1.5 L
2.4 L
3.0 L
2.4 L
2.CL
5.7 L
2.3 L
5.7 L
3.0 L
5.7 L
3.3 L
TRANS- DYNAMOMETER
MISSION H.P. I.W.
M
-------
MILEAGE FREQUENCY OF TESTS
The total mileage accumulation by the sample group during the project was in
excess of 700,000 miles - an average of about 35,000 miles per vehicle. Figure 3.1 is a
mileage test point roster showing the odometer readings on each vehicle for its tests.
Figure 3.1 also shows how all 80 tests are distributed as a function of accumulated
mileage. The 10-50,000 mile band contains W (60%) of all project tests and the 50-
100,000 mile band contains 23 (30%) of the tests.
ML£ACE TEST POINT ROSTER
lOK>J04090eOfDae>90 100.
2
5
4
*
6
_ T
to
i t(
I
17
u
-------
TABLE 3.2- VEHICLE SAMPLE CROUP BY MODEL YEAR
MODEL YEAR
MANUFACTURER <1977 197; 1978 1979 1980
GENERAL MOTORS 0 0-1 .6 1
VOLKSWAGEN 01121
MERCEDES-BENZ 01120
AUDI 0 0 01 0
PEUGEOT 00010
MITSUBISHI 01000
NISSAN 1 0 0 0 0
TOTALS 1 2 <• 12 2
TOTAL TOTAL
VEHICLES TESTS
21
36
20
12
6
3
2
SO
r
">,
j/j*
a
—
CC*
*ja»*Do - M w £ 5» <
JAN FE3 MAR APS MAY ,"JfJ JUL AUC SE=> OCT \'OV DtC
Figure 3.3 Monthly distribution of vehicle tests.
12
-------
almost half of the tests were performed from December to March, when diesel fuel
commonly contains additives. Tests performed in this time period would have the
largest differences between "as-received" and control fuel.
Vehicle Test Protocol
The chemical and bioassay characterizations of diesel particulate are, at
present, fields of continuing research and development. At the initiation of this
project, the state of the art in these fields impacted upon the vehicle testing protocol.
Since little information was available on, for instance, how vehicle test conditions
affected bioassay results, a lengthy test protocol was adopted to comprehensively
measure the vehicle emissions beyond what would have been required if the only goal
of the study was to measure vehicle emissions.
The desire to test the vehicles in "as-received" condition, and the need for some
standardized test condition was met by doing both. Thus each complete "vehicle test"
was composed of repeated "test phases". Each test phase included five or six types of
vehicle operation to examine their differences among modes of operation. Since little
data were available on pre-conditioning effects, special tests were performed to
determine the impact of engine operation modes immediately prior to a vehicle test.
Vehicle Test Cycles
The vehicle test cycles used in this study were: the new vehicle emissions
certification driving schedule (Federal Test Procedure - FTP), the Congested Freeway
Driving Schedule - (CFDS), the Highway Fuel Economy Test -(HFET), a steady-speed
50 mph cruise (50C), idle operation at zero vehicle spee^ with vehicle in neutral or
"park" (IDLE), and the New York City Cycle (NYCC). Although NYCC has frequently
appeared in the technical literature, its origins have not been well-documented. The
NYCC was developed by Alfred DeFilippis and Doe Mariano of the New York City
Department of Environmental Protection to represent driving conditions in highly
congested urban traffic (39). Speed-time data from midtown Manhattan were used in a
stochastic model that examined mv. nents up to the third order to generate speed-time
tables for testing purposes. Vehicle acceleration and deceleration halves of the
velocity-acceleration plane were treated separately to reflect vehicle differences
implied by these mathematically inverse, but operationally distinct, modes. Specific
details about the length, average speed, number of vehicle stops/mile, etc. for the
other driving cycles are available in various reference sources and will not be
tabulated here (30).
The vehicle test protocol was designed to provide insight into several areas of
diesel emissions lor which little or no information was available. Among these were:
•the variation of particulate and gaseous emissions with vehicle age, fuel,
lubricating oil types, and driving cycle;
•the variation of particulate character (extract, residue and bioassay) with the
above parameters;
•the variation of emissions from day-to-day and the reproducibility of the results
of a driving cycle;
•the variations among vehicles of the same type and between different groups of
vehicles.
To anticipate these topics from in-use vehicles over a two-year testing period, a
test protocol employing three replicate driving cycle sequences was adopted, each
13
-------
with different fuel/lubricating oil test conditions. Within project documentation, each
of these conditions was labeled as a "Phase". Each of the three phases thus included
separate vehicle gaseous/particulate measurements for a sequence of driving cycles.
The three phases corresponded to the following vehicle fuel/lubricating oil conditions:
Phase 1 = vehicle tested "as-received".
Phase 2 = project control fuel, "as-received" oil.
Phase 3 = project control fuel, fresh oil of manufacturer speci'.ication.
When an in-use vehicle was procured for testing, the entire data accumulation
from that procurement was labeled a "Test".
TEST
Phase 1
Phase 2
Phase 3
Driving Cycle Sequence Driving Cycle Sequence Driving Cycle Sequence
The original driving cycle sequence is given in Table 3.3a as a list of the driving
cycles and vehicle soak conditions employed. After 3^ complete vehicle tests, the
project data were analyzed and reported (31), and at this point the test protocol was
modified. The changes were: 1) All Phase 2 testing was eliminated, leaving only
Phase 1 and Phase 3 as a complete vehicle test for tests beyond //34, and 2) The driving
cycle sequence was modified as shown in Table 3.3b.
TABLE 3.3. VEHICLE TEST DRIVING CYCLE SEQUENCES
A B
Vehicle Tests
1-3*
50C, 30 min*
50C, 30 min.
CFDS
HFET X 3
SOAK, overnight
FTP
CFDS
HFET X 3
IDLE, 30 min
Repeated for
each of 3 fuel/
oil combinations^
Phase I, 2. 3
Vehicle Tests
2J-80
50C, 15 min*
HFET X 3
SOAK, overnight
FTP
CFDS
HFET
NYCC
50C, 15 min
IDLE, 15 min
Repeated for
each of 2 fuel/
oil combinations;
Phase 1, 3
•Pre-Test conditioning, no data taken.
-------
Both driving cycle sequences given in Table 3.3 began with an afternoon portion,
followed by an overnight soak at laboratory conditions, and concluding with a morning
portion that started with an FTP. Sufficient duplicate testing of the same driving
cycle was intentionally included to document particulate and paniculate character
changes due to changes during a test phase. Af';er test 3*», some of this emphasis was
dropped, and the New York City Cycle added to examine slow speed vehicle operation.
The test sequence spanned vehicle modes from idle to 50 mph, and driving cycles with
a wide range of speed variability.
Each complete vehicle test, with its multiple phases and cycle sequence in each
phase, required almost one week of work. The project included over 3,200 filter
measurements of particulate and 1,800 50 cm x 50 cm filter/solvent extractions.
Project archives contain individual extract, fuel, oil, and particulate residue samples
from the many test conditions. This report, assembled a few months after the
completion of vehicle testing, summarizes and analyzes data to the extent possible
within that time frame. However, the archival samples represent a potentially fertile
area for future studies.
15
-------
SECTION 4
GASEOUS AND PARTICULATE EMISSIONS
SUMMARY
This section presents Phase 3 emissions and fuel economy data by vehicle group
for the FTP (Table 4.1), CFDS (Table 4.2), HFET (Table 4.3), 50C (Table 4.4), NYCC
(Table 4.5) and IDLE (Table 4.6). Each table entry consists of the mean value of a
parameter for all tests for the given vehicle group, the standard deviation, a, of that
set and the coefficient of variation, CV, % (a/mean x 100). This form of presentation
gives both absolute emissions and an indication of the observed "spread" in the data so
that graphical differences shown in subsequent discussions will be more meaningful.
The coefficients of variation in Tables 4.1-4.6 show that the heterogeneity of a
vehicle group depends very strongly on the parameter of interest. Gaseous emissions
for these cycles have CV's in the range of 25% to 67% for hydrocarbons but only in the
range of 11% to 26% for carbon monoxide and nitrogen oxides. Particulate and residue
CV's range from 10% to 48% while extract has very high CV's, ranging from 35% to
93%. The factors causing high coefficients of variation for extract are likely the same
as those which affect the hydrocarbons. Fuel economy is the most consistent
parameter having a CV range of only 4% to 9% except for the IDLE where fuel
economy in minutes per gallon has a CV of 8-16%.
Mean values for these parameters and several other parameters on an individual
vehicle basis are presented in Appendix A for both Phase 3 and Phase 1.
DRIVING CYCLE EFFECTS
Introduction
This section presents the data with respect to effects of the various driving
cycles. All data in this section are averages of all available data for a given cycle.
For the first 34 vetvcle tests particulate data were collected from the FTP, CFDS,
HFET, 50C and IDLE cycles but gaseous data were collected only from the FTP,
CFDS, and HFET, For the last 46 tests both particulate and gaseous data were
collected from these five cycles as well as from the NYCC. Although the figures
which follow present averages for specific parameters and cycles, some averages are
necessarily based on fewer tests due to the absence of gaseous or particulate data.
Parf;culate, Residue and Extract
The definition of diesel "particulate" is operational as opposed to exact. !t is
operational because it is based on the conditions of measurement rather than the
16
-------
TABLE 0.1. - SUMMARY OF FTP PARTICIPATE AND CASEOUS EMISSIONS - PHASE 3
CM
Emission Mean a CV,%
PARTICIPATE, g/mi 0.89 0.19 22
RESIDUE, g/mi 0.6) O.I) 20
EXTRACT, g/mi 0.25 0.16 63
% EXTRACTIBUE 26.3 11.0 42
HYDROCARBONS, g/mi 0.65 • 0.21 J2
CARBON MONOXIDE, g/mi 1.69 0.22 13
NITFOGEN OXIDES, g/mi 1.71 0.19 II
FUEL ECONOMY, mpg 20.0 1.0 J
Mean
CV.%
Mean
CV.%
TABLE 0.2. SUMMARY OF CFDS PARTICULATE AND CASEOUS EMISSIONS - PHASE 3
CM VW
Emission Mean o CV,% Mean a CV.%
PARTICULATE, g/mi 0.62 0.20 33 0.3d 0.08 24
RESIDUE, g/mi 0.39 O.I I 27 0.26 0.07 27
EXTRACT, g/nv 0.22 0.16 73 0.08 0.03 40
%EXTRACTIBLE 33 14 »l 23 8 35
HYDROCARBONS, g/mi 0.4! 0.16 38 0.26 0.08 30
CARBON MONOXIDE, g/mi 1.12 0.14 13 0.92 0.21 23
NITROGEN OXIDES, g/mi 1.38 0.15 II 0.83 0.20 24
FUEL ECONOMY, mpg 27.* I.I 4 53.6 2.6 )
Mean
0.43
0.37
0.06
15
0.17
0.90
1.35
31.1
MB
o
0.06
0.07
0.03
8
0.08
0.13
0.19
2.7
CV.%
11
19
49
51
45
14
14
9
TABLE 4.3. SUMMARY OF HFET PARTICULATE AND CASEOUS EMISSIONS - PHASE 3
Emission
PARTICULATE, g/mi
RESIDUE, g/mi
EXTRACT, g/mi
* EXTKACTIBLE
HYDROCARBONS, g/mi
CARBON MONOXIDE, g/mi
NITROGEN OXIDES, g/mi
FUEL ECONOMY, mpg
Mean
0.49
0.29
0.20
37.7
0.33
0.95
1.36
30.1
CM
a
0.20
0.11
0.16
15.7
0.14
0.10
0.16
1.27
CV.%
42
40
78
42
42
II
II
«
Mean
0.30
0.24
0.07
23.2
0.20
o.;7
0.85
.V.2
VW
o CV,%
0.07
0.07
0.03
8.6
0.06
0.19
0.22
3.22
2)
29
41
37
30
24
26
6
Mean
0.39
0.34
0.06
14.6
0.13
0.8)
1.34
37.7
MB
0
0.06
0.06
0.03
8.6
0.06
0.12
0.20
3.0
CV,%
14
19
58
58
49
14
15
9
TABLE 4.4. SUMMARY OF 50C PARTICULATE AND CASEOUS EMISSIONS - PHASE 3
Emission, Mean
PAUTICULATE, g/mi 0.41 0.20 48
RESIDUE, g/mi 0.21 0.09 40
EXTRACT, g/rni 0.20 0.16 82
% EXTRACTIBLE 43 17 40
HYDROCARBONS, g/mi 0.41 0.18 43
CARBON MONOXIDE, g/mi 0.94 0.11 12
NITROGEN OXIDES, g/mi 1.30 0.15 12
FUEL ECONOMY, mpg 32.2 I.) )
Mean
CV.%
Mean
0.37
0.31
0.06
I)
0.14
0.82
.'.2
34.4
MB
o
0.07
0.07
0.03
8
O.D6
0.15
O.I I
2.4
CV,%
20
21
57
50
45
18
9
7
17
-------
TABLE 4.). SUMMARY OF NYCC PARTICULATE AND CASEOUS EMISSIONS - PHASE 3
CM
Emission Mean o CV.%
PARTICULATE, g/mi I.SS 0.23 12
RESIDUE, g/mi I.3J 0.14 10
EXTRACT, g/mi 0-)) O.l» 3)
% EXTRACTIVE 27-' ••0 W
HYDROCARBONS, g/mi 1.69 O.)l 30
CARBON MONOXIDE, g/mi ».02 O.J3 13
NITROGEN OXIDES, g/mi 2.82 0.37 13
FUEL ECONOMY, mpg II.I 0.8 7
Mean
0.4*
0.30
0.14
3I.S
0.49
2.06
l.)3
29.4
vw
o
0.0)
0.06
0.06
12.)
0.33
0.)3
0.20
1.1
CVj%
10
19
• 5
39
67
26
13
>
Mean
0.17
0.7)
0.13
D.I
0.07
2.27
2.32
17.0
MB
o CV.%
0.12
0.12
0.0)
J.I3
0.2)
O.X>
0.27
1.3
Id
16
38
3*
)2
22
12
t
TABLE 4.6. SUMMARY OF IDLE PARTICULATE AND CASEOUS EMISSIONS - PHASE 3
Emission
PARTICULATE, g/mi
RESIDUE, g/mi
EXTRACT, g/mi
%EXTRACTIBLE
HYDROCARBONS, g/mi
CARBON MONOXIDE, g/mi
NITROGEN OXIDES, g/mi
FUEL ECONOMY, mm/gal
Mean
O.IS8
0.14)
0.0*7
22
0.230
O.M9
0.160
163
CM
o
0.028
0.0 IS
0.021
8
O.I II
0.103
0.033
13
CV.%
1)
13
48
3)
08
19
21
8
Mean
VW
o CV.%
0.017 0.011
0.007 0.003
0.011
)3
0.0)8
0.186
0.099
0.009
21
63
93
«0
0.0)9 103
0.121. 67
0.026 26
561
Mean
0.0)3
0.04)
0.008
I)
0.040
0.1)3
0.128
346
MB
o
0.014
0.013
0.003
)
0.030
0.027
0.032
)7
CV.%
26
28
36
3)
74
18
2)
16
properties of the substance. The mass of particulate is determined from the weight
gain of a specified filter at specified temperature and flow rate ranges. Anything
collected by the filter (except uncombined water - for which no correction is made) is
called particulate.
The particulate itself can be further fractionated by means of solvent extraction.
Again, an operational definition applies. The "extract" is that material which is
removed by a particular solvent under the specified conditions. In addition, the
material which is not extracted can be called the "residue". Because of the
complimentary nature of extract and residue and the existence of possible carcinogens
in the extract, it has been customary to use only the extract and to express it as a
percentage of the particulate, i.e., percent extractible rather than as a mass emission
in itself.
With proper choice of solvent and extraction conditions almost all organic (as well
as some inorganic) compounds can be removed from the particulate. The remaining
"residue" is primarily carbon with insoluble inorganics and organics. Production of the
carbon (frequently called "soot") in the diesel engine has been extensively studied.
Although the extract has received much attention in the form of chemical analysis and
biological hazard assessment, only limited work has been conducted to determine the
source of the extract and the mechanism by which it becomes "particulate".
There are four locations at which particulate matter could be formed:
(1) the engine cylinders,
(2) the exhaust system,
(3) the dilution tunnel, and
W the filter.
18
-------
The carbon portion of the residue is formed in the engine cylinders as a
combustion product. After production, it undergoes physical changes before collection
but is probably changed very little in mass. Some of the material produced during
combustion is probably solvent extractible hydrocarbon derived material. In the post-
combustion section of the sampling system, additional extractible material can be
formed by:
(1) absorption/adsorption,
(2) mass diffusion,
(3) condensation, and
(4) chemical reaction.
Often it is the cooling of the exhaust (caused by dilution and heat loss) which promotes
the conversion of gaseous material to particulate material via a mechanism involving
the above processes. The mass, particle size, shape and surface morphology of the
carbon portion could affect the quantity of gaseous material converted into
particulate.
In the course of this study we observed that the residue mass was a function of
the engine family, driving cycle and operating temperature and did not appear to be
influenced by fuel composition, lubricating oil or engine condition (mileage
accumulation). We wiii, therefore, consider residue to be a measurable, reproducible
and independent parameter with a physical significance equal to that of the extract.
The effects of driving cycles on the particulate, extract and residue are shown in
Figures 4.1, 4.2, and 4.3 for the General Motors, Volkswagen, and Mercedes-Benz
groups, respectively. The cycles have been arranged on the x-axis such that the
average speed of the cycle (except the ID~E) increases from left to right and the
speed variability of the cycle increases from right to left. The General Motors group
data show that the main contributor to the cycle-to-cycle difference observed in the
particulate was the residue. The extract contribution was relatively constant except
for a large increase for the NYCC. The IDLE data (expressed in units of
grams/minute) shows a large decrease from the 50C data (expressed in grams/minute)
in all three categories. (For comparisons with IDLE data, the 50C data can be easily
converted from gram/mile to grams/minute by multiplying by 50/60 or 0.83.)
The Volkswagen group data. Figure 4.2, are very similar to the General Motors
group data except that the residue is much less influenced by driving cycle and,
therefore, the particulate shows less cycle dependence than did the General Motors
group. The IDLE data are very different rhan for the General Motors group in that
both residue and extract are very low both in absolute emission rate and relative to
the 50C. This is the only cycle for which the extract is greater than the residue. The
very low emissions for this cycle made measurement difficult and the accuracy and
precision of IDLE data for Volkswagen is much lower than for any other cycles/makes.
The Mercedes-Benz group data, Figure 4.3, show trends very similar to both the
General Motors and Volkswagen groups. The large increase in residue for the NYCC is
very similar to that observed for the General Motors group. The large decrease in
residue and extract shown for the IDLE is much like that exhioited by the Volkswagen
group; but here the residue is clearly the main component of particulate. The extract
is very similar to that of Volkswagen in absolute terms for all six cycles.
Figures 4.4, 4.5, and 4.6 compare the vehicle groups for emissions of particulate,
residue and extract respectively. Figures 4.4b, 4.5b and 4.6b show the same data as
the corresponding "a" figures but normalized to the FTP to better illustrate the cycle
19
-------
S3-
O WWTICULBTE
* RESIDUE
•f EXTRACT
O PWMICULfiTE
RESIDUE
+ EXTRfiCT
NTCC FTP CFCS rTET SO C IDLE
WCC FTP CFOS MFET 50 C IDLE*
Figure 0.1 Cycle variations of paniculate, residue
and extract for the General Motors vehicle group.
•IDLE units are g/minute
Figure <».2 Cycle variations of paniculate, residue
and extract for the Volkswagen vehicle group.
•IDLE units are g/minute
E
O PfKTlCULRTE
» RESIDUE
EXTRBCT
NTCC FTP CFDS rTET 50 C IDLE
Figure 4.3 Cycle variations of paniculate, residue
and extract for the Mercedes-Benz vehicle group.
•IDLE units are g/minute
20
-------
L
» r*
tu
5
O GENERAL MOTORS
» VGHSWRSEN
+ MERCEDES-BENZ
NTCC FTP cros WET so c IDLE
NICC FTP CFOS MFET SO C JOLE
Figure 4.4 Cycle variations of paniculate by vehicle group: (a) actual values,
(b) normalized to the FTP.
*1DLE units are g/minute
O KNERBl MOTORS
» VOLKSWBStN
+ HERCEDES-etNZ
b
O CENERBL MOTORS
* VOUSWRGEN
+ MERCEDES-BENZ
NTCt FTP CFD5 MFET SC t IDLE*
NTCC flF CFOS HFE1 SO C JOLF.
Figure 4.5 Cycle variations of residue by vehicle group: (a) actual values,
(b) normalized to the FTP.
*1DLE units are g/minute
21
-------
sensitivities of the three groups. Figure 4.4 shows trends of increasing particulate
with driving cycle: 50 C < HFET < CFDS < FTP < NYCC and by group: VW < MB <
GM. The General Motors group shows a strong cycle sensitivity which is much greater
than that of the Volkswagen and Mercedes-Benz groups. For the General Motors
group, the NYCC paniculate is more than double the FTP and the 50C particulate is
about one-half the FTP. For the Volkswagen group, in contrast has values of 1.1 and
0.8 times the FTP. The Volkswagen group shows very little cycle sensitivity with only
about a 50% increase in particulate from the 50C to the NYCC.
Figure 4.5 shows the cycle-to-cycle trends for residue. The General Motors curve
is very different from those for the Volkswagen and Mercedes-Benz groups. The
residue for the HFET is lower than that for the Mercedes-Benz group; and the 50C
residue is lower than for both the Mercedes-Benz and Volkswagen groups. This trend
reversal was not observed for any other emission. Figure 4.5b shows that the General
Motors curve is fundamentally different. There is no leveling off of the residue for
the CFDS, HFET and 50C to the IDLE. The Volkswagen curve shows almost no
difference between cycles except for a large decrease at the IDLE.
Figure 4.6 shows that the extract for the General Motors group is two to three
times greater than for the Volkswagen and Mercedes-Benz groups. Cycle variations
are relatively small and constant for all groups except for the NYCC which has
extract about twice that of the FTP for all groups.
Figures 4.7a and 4.7b show the trends for the percent extractible,
(extract/particulate)xlOO. Comparison of Figures 4.7a and 4.4a shows that the General
Motors group percent extractives are inversely related to the total particulate except
for the IDLE which has the lowest value. The Volkswagen group data show a large
increase for the NYCC and a very large increase for the IDLE. The IDLE value may
be biased to the high side due to difficulties in measuring the small quantities of
particulate emitted by the Volkswagens and inclusion of the normally relatively insig-
nificant amount of extract from the filter material itself. The Mercedes-Benz group
data showed a very low and very consistent percent extractibles and almost no cycle
variation except for a small decrease for the FTP.
Gaseous Emissions
Hydrocarbons-
Average hydrocarbon emissions, g/mi, are shown in Figure 4.8 for all vehicle
groups. The most noticeable aspect of this data is the large increase in HC for the
NYCC for all vehicle groups. The NYCC HC averages 2.5 times higher than the FTP
for the General Motors group and 1.7 times higher than the FTP for the Volkswagen
and Mercedes-Benz groups. In all instances the HFET has lower HC than the other
driven cycles and the trends of increasing emissions are: HFET < 50C
-------
I-
^_ «« *
O GfNfRPL MOTORS
» VDHSWRKN
•» MERCEDES-BENZ
r
^~
Sr "'
B...
O CCNERPL MOTORS
* VOH.SWRMN
4 IV.Kl.Df.i-Bf.H2
NTCC FTP CFDS MFE1 50 C JDlE
NTCC FTP CfDS HFET SO C JDLE
Figure 4.6 Cycle variations of extract by vehicle group: (a) actual values,
(b) normalized to the FTP.
*1DLE units are g/minute
g
O KNERfiL MOTORS
VOUSHfl&tN
**«"
O CfNERSL MOTORS
A VOlKSMQGCN
+ «ERCEOES-BtNZ
NTCC FTP CFOS hTET 50 C IDLE
nrCC FTP CFDS HFET 50 C JDLE
Figure ^.7 Cycle variations of % extractible by vehicle group: (a) actual values,
(b) normalized to the FTP.
23
-------
O KNERBl MOTORS
* VOUSMR«N
* MERCEDES-BENZ
O KNERW. "OTORS
• VOLISW«C£N
ICRCEDES-BENZ
NICC FTP CFDS HfET 50 C IDLE
NICC F1P CFOS Mff.1 SO C IDLE*
Figure 4.8 Cycle variations of hydrocarbons by
vehicle group.
•IDLE units are g/minute
Figure 4.9 Cycle variations of carbon monoxide by
vehicle group.
•IDLE units are g/minute
3 GENERSL HOTC«S
A vatSwSOEN
•f MERCEDES-BENZ
NICC FTP CFDS MFEI 50 C IDLE
Figure 4.10 Cycle variations of nitrogen oxides by
vehicle group.
*IDLE units are g/minute
-------
which is similar to that of HC for the General Motors and Volkswagen groups.
Nitrogen Oxides-
Figure it. 10 shows that NOX emissions for the General Motors and Mercedes-Benz
groups are very similar, with the General Motors group higher for all cycles but
deviating significantly only for the FTP and NYCC. The values for the Volkswagen
group are roughly 60% of the comparable values for the General Motors or Mercedes-
Benz groups. There is a difference among group cycle variations for the CFDS, HFET
and 50C. For all groups the IDLE shows a large decrease.
MILEAGE ACCUMULATION EFFECTS
Introduction
This section discusses changes which occurred in emission parameters as the
vehicles accumulated mileage. In this study the miieage differences between initial
and final tests on an individual vehicle basis ranged from 16,000 to 73,COO miles with
an average of 34,600 miles. Annual vehicle mileage accumulation ranged from 8,400
to 36,100 miles with an average of 22,300 miles. It is recognized that the effects of
mileage accumulation (and the coincident ageing of the vehicle) can be affected by
many parameters which are uncontrolled and largely unknown in this real-world in-use
study, such as engine mechanical work, injector timing, adjustments, etc. Table 4.7
gives information on the odometer mileage, test intervals and annual usage of the test
vehicles.
The data in this section are from the FTP only. In general the mileage
accumulation trends for a given emission parameter did not vary greatly from cycle to
cycle.
Participate Emissions
Emissions collected by filtration on teflon coated glass fiber filters (Pallflex
T60A20) under the conditionr prescribed by the Federal Test Procedure for light duty
diesel vehicles are discussed in this section. The emissions are divided into three
categories: (a) particuiate, which is the regulated and defined parameter, and its two
constituents (b) residue, the remainder after solvent extraction and (c) extract, the
soluble organic f raction-SOF.
Particuiate —
The effects of mileage accumulation on FTP particuiate, g/mi, are shown in
Figure 4.11a for the General Motors group, Figure 4.1 Ib for the Volkswagen group and
Figure 4.lie for the Mercedes-Benz group. Figure 4.1 Id shows all three vehicle groups
on common axes to allow better visual comparison. The irregular enclosures around
the group names are envelopes which contain all values for the respective group.
Vclues for Car //& and Car //ll are presented separately. All the General Motors
vehicles (Figure 4.1 la), except Car #16, showed increases in particuiate with mileage
ranging from slight (Car #3) to very large (Car //4 and Car #2). Increases as great as
50% occurred over the mileage accumulation period. The particuiate emissions
generally ranged from 0.65 to 0.9 g/mi except for tests on Car //2 after about 30,000
miles and all tests on Car //4.
25
-------
TABLE 4.7 - MILEAGE STATISTICS FOR TEST VEHICLES AND VEHICLE CROUPS
Car 8
2
3
4
5
7
16
CM
Odometer Mileage
Initial
11,069
Final
76,318
79,223
40,744
30,68$
50,683
67,266
57,487
4
Miles
73,326
73,063
34,015
27,068
33,568
35,468
46,418
No.
Tests
8
8
4
4
(I
3
31
Average Mileage
Between
Tests
10,475
10,438
11,338
9,023
11,858
17,731
11,240
Per
Year
32,600
32,500
17,700
14,800
29,400
22,400
23,200
1
6
8
9
11
VW
3,576
4,279
48,075
16,340
119,196
38,293
19,753
33,825
80,920
47,764
171,199
70,690
16,177
29,546
32,835
31,424
52,003
32,397
20
5,392
9,489
10,945
10,475
17,334
10,799
8,400
16,900
17,100
18,000
32,800
18,700
12
13
14
19
MB
43,444
26,006
22,3)7
23,043
28,703
75,195
42,520
41,964
44,846
51,131
31,751
16,514
19,647
21,803
22,429
3
3
3
3
12
15,876
8,257
9,824
10,902
11,215
18,100
9,000
11,200
15,400
13,400
10
15
17
18
ALL
34,165
22,754
34,941
4,341
64,846
82,956
77,251
32,^92
30,681
60,202
42,310
28,451
34,593
2
6
3
4
78
30,681
12,040
21,155
9,484
12,580
33,500
36,100
26,700
19,000
22,300
The Volkswagen group, Figure 4.lib, exhibited particulate in the range of 0.25 to
0.55 g/mi. All Volkswagens had particulate lower than the lowest General Motors
vehicle. No overall mi.'.eage accumulation trend was apparent, but individual vehicles
showed large relative variations in particulate. Car //I and Car //9 showed large
increases while Car if6 had a large consistent decrease. Car //ll, a very high mileage
vehicle, had particulate emissions which were average for its group. For Mercedes-
Benz vehicles, Figure 4.lie, particulate emissions were in the range of 0.4 to 0.7 g/mi
with no apparent mileage trend. Car #12 and Car #13 showed decreases, Car #14
increased, and Car #19 first decreased but then increased.
Residue-
Residue, g/mi, is shown for each vehicle group in Figures 4.12a, 4.12b, and 4.12c
and for all groups together in Figure 4.12d. In Figure 4.12a for General Motors, there
is a very tight cluster of data in the 0.5 to 0.75 g/mi range with the obvious exception
of Car ilk which was much higher. Some vehicles showed slight mileage effects such
as the increase noted for Car //5 and the decrease exhibited by Car #16, but in general
no overall trend was indicated. The Volkswagen group, Figure 4.12b, had residue which
tanged from 0.2 to 0.5 g/mi but was generally below 0.4 g/rni. The Volkswagen data
varied more car-to-car than the General Motors group data (with Car #4 excluded). A
26
-------
510
So
IS
30 MS 6D 75
1ES7 fULERM. lOOD'S
SO
So"
o toe 01
* CM 06
+ C<* 08
x c«e cs
« Cf* II
30 60 SO 120
TEST HJLEBGC. IDOO'S
ISO 1BC
o
O3-
.0'
UJ
tr
L
O CW 12
A CAR 13
•f C«? 1M
X CW 19
IS
30 MS 60 75 90
TEST HJLEWt- lODO'S
0 30 60 90 120 ISO 1BO
TEST MILERGC. 1000'S
Figure ^.11 Mileage accumulation effects for FTP particulate by vehicle groups: (a)
GenerJ Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three
groups.
27
-------
w»
go
O CflR 07
• CIW 03
* C*K 04
X CfiR OS
• OK 07
» CfK 16
IS 30 MS 60 75 90
TEST HJLERM. 1000'S
e c«? 01
* CflR 06
+ c«e ce
X CM 09
e CW 11
D 30 60 90 1?0 ISO 180
1ES1 NJLER&C. 1000'S
SS-
o cr# i?
A Cfue 13
<*• cw in
X CfK 19
IS 30 MS 60 75 90
TEST HJLEAGC. IDOO'S
"0 30 60 90 1?0 I5D )BO
US! HILEBCe. 1000'S
Figure 4.12 Mileage accumulation effects for FTP residue by vehicle groups: (a)
General Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three
groups.
28
-------
possible cause of this may have been the combination of four model years and two
transmission types in the Volkswagen group. The Mercedes-Benz group residue,
Figure 4.12c, was in ihe region of 0.35 to 0.5 g/mi. The trends in residue were
identical to those observed for the participate as the extract was very small and
constant.
Extract-
Figures 4.13a, 4.13b, and 4.13c show the variation of extract, g/mi, with accumu-
lated mileage for rhe General Motors, Volkswagen and Mercedes-Benz groups
respectively. Figure 4.13d shows all vehicle groups on common axes. Several General
Motors vehicles exhibited very large (3x to 5x) increases in extract during the mileage
accumulation period. In three cases (Cars #2, //4 and //7), there were very large
increases in the 20,000 to 30,000 mile range. All vehicles exhibited some increase in
extract during the mileage accumulation period. Below about 15,000 miles, all values
were below C.2 g/mi but by 40,000 miles the upper range of the General Motors group
extract had increased to over 0.6 g/mi. Extract values for the Volkswagen group were
very much lower than for the General Motors group, ranging from about 0.05 to 0.15
g/mi. Car //6 showed a large decrease (in a relative sense) after its first test. The
other vehicles showed increases after their first tests. Overall, there was no apparent
mileage accumulation effect, and the individual extract changes were all less than O.i
g/mi for the mileage accumulation period. The Mercedes-Benz group, Figure 4.13c,
extracts were generally in the 0.04 to 0.07 g/mi range except for Car //12 which had
about 0.14 g/mi extract. All Mercedes-Benz vehicles showed high, consistency over the
mileage accumulation period, and there was no overall mileage effect.
Figure 4.14 combines the emission envelopes of Figure 4.11, 4.12, 4.13 with
common axes (Car #11 has been removed for simplicity). From this figure, it is
apparent that the increases in particulate for the General Motors group are primarily
due to extract. The particulate of the Volkswagen and Mercedes-Benz groups was due
primarily to the residue and the extract was small and constant.
Gaseous Emissions
Hydrocarbons—
HC emissions for the FTP are shown in Figure 4J5a for the General Motors group,
in Figure 4.15b for the Volkswagen group, in Figure 4.15c for the Mercedes-Benz group
and in Figure 4.15d for all groups on common axes. General Motors vehicles generally
exhibited significant increases in FTP hydrocarbon emissions, g/mi, with increased
mileage. Data on six vehicles for the FTP are shown in Figure 4.15a. Th^se data
generally cluster in the 0.45 to 0.7 g/mi range with several notable exceptions. Car //2
showed a doubling of HC between tests at about 30,000 and *-5,000 miles. During this
period the engine heads were replaced. The changing of the heads or some other
alteration (deliberate or accidental) may account for the increased HC. Car #7
showeo a similar increase in HC in about the same mileage range.
The Volkswagen group had HC emissions that generally ranged from 0.2 to 0.4
g/mi as shown in Figure 4.15b. Individual vehicle variations with mileage were
generally not large or consistent, with both increasing and decreasing trend' observed.
No overall group trend with mileage was apparent.
The Mercedes-Benz group HC emissions were similar to those of the Volkswagen
29
-------
e cw 02
« car 03
« CRR CM
X CfK OS
« ewe o?
» CfK 16
25
d
8
8
O CPU 01
A CAR 06
•* CAR 08
X CAR 09
* CAR II
IS 30 >45 60 75
TEST MILEPvX. 1000'S
SO
30 60 90 120 ISO 180
TEST H;UAG£. 1000'S
o cm 12
A cm )3
+ CAR m
X CRR 19
IS 30 MS 60 75
TEST KILERGE. 1000'S
90
CM 12
30 60 90 120 ISO 180
TEST MILEAGE. 1000'S
Figure 4.13 MiJeage accumulation effects for FTP extract by vehicle groups: (a)
General Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three
groups.
30
-------
PRRTICULflTE
RESIDUE
EXTRRC7
30 60 90
CAR 4
0 30 60 90
TEST MJLEflGE. JOOO'S
CAR 12
30 60 90
Figure 4.14 Group emission envelopes for: (a) particulate, (b) residue, (c) extract.
31
-------
mo
g
&
O CRR 0?
A CM 03
•f CM (W
X CM OS
# CM 07
* CM 16
30 45 60 75 90
TEST MJLERGE. 1000'S
o cm 01
A CM 06
+ CM OB
X CM 09
» CM II
y>s«-*
0 30 60 90 1?0 150 160
TEST HJURGE. 1000'S
in
o
go
O CRP IZ
* cae is
+ CfiR 1M
X CfiR 19
s.
cnc>
g
IS 33 US 60 75 90
TEST H1LERCE. 1000'S
0 30 60 9: 120 ISO ISO
TEST M1LEHK. 1000'S
Figure 4.15 Mileage accumulation effects for FTP hydrocarbons by vehicle groups: (a)
General Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three
groups.
32
-------
"1
,
group and were in the OJ5 to 0.5 g/mi range as shown in Figure 4.15c. No clear
mileage accumulation effects were apparent for the group, and individual vehicles
showed different mileage effects.
Carbon Monoxide—
CO emissions for the FTP are shown in Figure 4.l6a for the General Motors group,
Figure 4.16b for the Volkswagen group, Figure 4.16c for the Mercedes-Benz group, and
Figure 4.16d for all groups on common axes. CO emissions for the General Motors
group generally ranged from 1.4 to 2.2 g/mi. Mileage accumulation trends varied from
almost neutral to about a 30% increase over the course of the tests. Most of the CO
mileage trends were very similar to those observed for HC, but the increases were
smaller. The Volkswagen group had CO, Figure 4.16b, which ranged from 0.8 to 1.5
g/mi. The CO increased with mileage accumulation for all vehicles; but there was
much scatter in the data and no overall mileage trend was apparent. The Mercedes-
Benz group, Figure 4.16c, had CO emissions in the 0.9 to 1.6 g/mi range. Unlike the
General Motors and Volkswagen groups, the Mercedes-Benz group generally decreased
in CO with increasing mileage, but there were insufficient data to establish a definite
trend.
Nitrogen Oxides—
The variation of FTP nitrogen oxides with accumulated mileage is shown for the
General Motors group in Figure 4J7a, for the Volkswagen group in Figure 4.17b, for the
Mercedes-Benz group in Figure 4.17c, and for all groups on common axes in
Figure V.17d. Except for Car //4 (which consistently displayed unusual behavior) all '.he
General Motors vehicles showed decreases in NOX in the range of 5-20% with accumu-
lated mileage. Car #4 showed a very large and unexplained increase. The General
Motors group as a whole showed a decrease in NOX with accumulated mi!ea£ which
averaged about 0.2 g/mi per 50,000 miles. The Volkswagen group data were very
consistent except for Car #9 and were in the 0.8 to 1.1 g/mi range with no apparent
overall trend. All Mercedes-Benz vehicles exhibited decreased NOX at higher mileage.
The rate of decrease was much higher than for the General Motors grcup, and an
overall group decrease with increasing mileage was apparent.
% Extractive—
Extractible data are shown for individual vehicles by vehicle groups in Figures
4.18a, b and c and for the groups as a composite in Figure 4.18d. These figures are
very similar to those for extract, g/mi, (Figure 4.1-3); The General Motors group data
are very similar to the extract, g/mi, data except that Car //4 has a % extracfible that
is more typical of the group. The group as a whole has Jess scatter than it did for
extract. The Volkswagen and Mercedes-Benz groups have % extractible data which
are very similar to their extract, g/mi, data.
The General Motors and Volkswagen group show considerable overlap particularly
up to about 30,000 miles. The Mercedes-Benz group, however, (with the exception of
Car //12) has % extractives which were lower than all other vehicles. The difference
between Car #12 and the other three Mercedes-Benz vehicles may be related to its
greater age and higher mileage.
33
-------
O CM 02
* CM 03
•» C(* CM
X CAR 05
0 CM 07
* CRR IB
15
30 MS 60 75 90
TEST MJLEflK. 1000'S
1C
O CWc Cl
A CM 06
+ C«R 08
X CBR 09
« CRK 11
30 60 90 1?0 ISO IB3
TEST HILER&e. 1000'S
o
UJ
OfM
I"
O CRR 1?
» CBR 13
•f CfiP 1M
X CRI? 19
15 30 MS 60 75
1ES1 nlLEHCE. 1000'S
90
30 60 90 )?0 150 1BC
1ES1 MJLERGE. 1000'S
Figure 4.16 Mileage accumulation effects for FTP. carbon monoxide by vehicle groups:
(a) General Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three
groups.
-------
30
-------
O CRR 82
• CRK 03
+ CRK 0<<
X CM OS
* CRR 07
» CRR 16
O CRR 01
* CM 06
+ tm os
X CM 09
« CRR II
15 30 MS 60 75 90
TEST fllLESGE. 1000'S
30 60 90 120 ISO 160
TEST M1LERCE. 10:0'S
O CBR 12
* CfiR 13
+ CRR m
X CflR 19
15 39 US 60 75
TEST MRERGE, 1000'S
SO
s
30 60 90 120 153 JB3
TEST MJLERCE. 1033'S
Figure 4.18 Mileage accumulation effects for FTP % extractive by vehicle group:
(a) General Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three
groups.
36
-------
Fuel Specific Emissions--
Previous discussions have expressed emissions in terms of mass per unit distance
traveled, i.e., grams per mile. This section will discuss particulate, residue and
extract in terms of mass per unit fuel consumed, i.e., grams per kilogram fuel (g/KgF).
Fuel consumption was computed by the carbon balance method, and the weight of fuel
derived from the measured specific gravity of the fuel. This method of expressing
emissions permits comparisons of all groups and cycles on a common basis.
The data for fuel specific particulate, residue and extract are shown in Figures
4.19, 4.20, and 4.21 respectively. Several aspects of these figures are interesting. In
Figure 4.19 there is a very large diversity in the IDLE particulate of the three vehicle
groups. Figures 4.20 and 4.21 show that these differences are due mainly to changes in
the residue. In fact, almost all of the variations of the particulate were due to
variations of the residue. The IDLE residue for the General Motors and Mercedes-
Benz groups showed a large increase relative to the other cycles but the Volkswagen
group had a large decrease.
The fuel specific residue data showed the NYCC to be more comparable to IDLE
than other cycles. This correspondence between the IDLE and NYCC would be
expected because 40% of the NYCC time is at idle. The connection between the
NYCC and IDLE has not been apparent in earlier comparisons because of the
incompatibility of units.
Statistical Comparisons of Fuel Specific Emissions--
The mean fuel specific emission parameters for the FTP are segregated by group
and test mileage in Table 4.8. Using the Smith-Satterthwaite test (32) for pairs of
mean values, a t-statistic may be generated and tested against the null hypothesis that
the data pair means were sampled from populations of equal means. For most
parameters the t-statistic achieved a high significance (low a) level so that the null
hypothesis could be rejected with a high degree of confidence (90% confidence or
greater; o=0.100 or less). In these cases, differences observed between means were
judged statistically significant. As the data were normalized to a fuei specific basis,
these observations may represent real differences in fuel combustion processes among
the engine types. The significance levels achieved by the t-statistic for pairs of means
are given in Table 4.9.
TABLE <(.9. SIGNIFICANCE LEVEL (a LEVEL) FOR VEHICLE GROUP FUEL SPECIFIC EMISSIONS
ALL MILEAGE MEANS
GM/VW
CM/MB
VW/MB
HC
-
.OPr-
.005
NO*
.005
.005
.100
CO
.005
.100
.005
PART
.100
.005
.005
EXT
.005
.005
.005
P.ES
-
.025
.050
REV
.005
.100
.005
The differences in General Motors and Volkswagen mean fuel specific hydrocarbon
emissions, over all test mileages, were not statistically significant; however, compari-
sons of the General Motors and Volkswagen means to the Mercedes-Benz hydrocarbon
mean (considerably lower) were statistically significant. In like manner General
Motors appeared significantly lower than Volkswagen and Mercedes-Benz in terms of
37
-------
a
•jo
O GENERRL MOTORS
A VOtKSKSGEN
•f MERCEDES-BENZ
O GENERRL MOTORS
* VOUSHB&EN
+ HERCEDES-BEN2
NTCC FTP CFDS WET 50 C IDLE
Figure 4.19 Cycle variations of fuel specific
paniculate by vehicle groups.
NTCC
FTP CFDS MFET 50 C :OLE
Figure 4.20 Cycle variations of fuel specific residue
by vehicle groups.
O GENERRL MOTORS
A vouswac^N
+ MERCEDES-BENZ
NTCC FTP CFDS HFET SO C IDLE
Figure 4.21 Cycle variations of fuel specific extract
by vehicle groups.
38
-------
TABLE 4.8. MEAN FUEL SPECIFIC EMISSION PARAMETERS BY VEHICLE GROUP AND TEST MILEAGE
CM
TEST FTP HC FTP NOX
MILEAGE G/KGF G/KGF
FTP CO FTP PART FTP EXT FTP RES FTP REV
G/KGF G/KGF G/KGF G/KGF 105/KGF
&-20K
20-40K
40-60K
ALL
VW
0-20K
20-40K
40-60K
ALL
MB
0-20K
20-40K
40-60K
ALL
•Ames
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
N
MEAN
STD DEV
10
10637.
5484.
8
26921.
4833.
7
46714.
5666.
31
34725.
23175.
7
11894.
6110.
4
28756.
6664.
3
50638.
4713.
20
53773.
51272.
-
6
29418.
6385.
4
43193.
1259.
12
40622.
15619.
08
945
95
605
99
203
19
59
82
996
64
363
71
270
35
52
33
199
50
243
00
25
9
2.8332
.3393
8
3.0040
.6831
7
4.7219
1.3664
30
3.5959
1.1653
7
3.1086
.7716
4
2.8436
.4488
3
3.0633
.7219
20
3.4004
.8116
-
6
1.2881
.4006
4
2.3C91
1.0732
12
1.9258
.9506
9
9.8803
.7694
8
9.1525
.5831
7
8.9880
.9993
30
9.2242
.8738
7
10.3448
.4796
4
12.7144
3.1388
3
12.4374
2.8300
20
11.2190
1.9612
-
6
10.9305
.6130
4
10.1387
.6253
12
10.4410
.7642
9
8.9688
.8399
8
8.5552
.9686
7
9.5386
1.5307
30
9.1318
1.0982
7
12.2463
2.6798
4
12.0794
1 . 3967
3
13.3187
1.1859
20
12,8973
2.0090
-
6
7.9555
1.6227
4
8.6032
2.2906
12
8.3430
1.6790
4
1
4
5
1
4
1
4
1
3
4
4
3
4
3
9
.6329
.0069
8
.5289
.3536
7
.4713
.4510
30
.8930
.0216
7
.8342
.2060
4
.8587
.3203
3
.6813
.3149
19
.4814
.8474
-
6
.4617
.4669
4
.0389
.3223
12
.6802
.4505
9
.8375
.1794
8
1.0336
.3850
7
2.0042
.9532
30
1.3539
.8286
7
.8941
.3989
4
.6097
.1183
3
.9004
.3096
19
.9004
.3459
-
6
.3396
.0837
4
.5568
.2857
12
.5139
.2787
9
3.7954
.9272
8
3.4953
.2192
7
3.4672
.7158
30
3.5391
.6282
7
3.9401
1.0560
4
3.2490
.3410
3
3.7809
.4650
19
3.5810
.7821
-
6
3.1221
.4109
4
3.4822
.1554
12
3.1663
.4083
9
32.9287
18.4596
8
28.2102
9.0524
6
34.9440
10.7958
28
32.4003
14.1037
7
113.0282
106.6898
4
96.0641
41.0499
2
145.4095
28.9461
17
100.3466
73.3652
-
6
26.7239
14.3987
1
15.4572
8
24.7175
12.8410
test with TA98(-).
39
-------
fuel specific NOX, while the Volkswagen group was significantly higher in fuel specific
CO. For fuel specific particulate, Mercedes-Benz was significantly lower in mean
value over all mileages. The fuel specific extract means were each significsntly
different from each other. The Mercedes-Benz group was significantly lower in
residue, and the Volkswagen group was significantly higher in TA98 Ames activity,
each on a fuel specific basis, over all mileages.
Comparisons of means at different mileage intervals did not demonstrate many
statistically significant differences among the mean values, Table 4.10. General
Motors group fuel specific HC emissions appeared higher in the 40-60K mile range
while fuel specific NOX appeared to decrease after 20K miles. Fuel specific extract
for the General Motors group was significantly higher in the 40-60K mile range. Fuel
specific particulate for the Volkswagen group was .significantly lower in the 20-40K
mile range. For the Mercedes-Benz group, fuel specific NOX decreased while fuel
specific particulate and residue increased significantly in the 40-60K mileage range.
TABLE 4.10. SIGNIFICANCE LEVEL (a LEVEL) FOR MEAN FUEL SPECIFIC EMISSIONS
BY TEST MILEAGE RANGE
MILEAGE RANCE(Kmi) HC NO^ CO PART EXT RES REV
0-20/20-40 - 0.025 - ...
0-20/40-60 0.010 0.050 - - 0.010
20-40/40-60 0.010 - .100 .100 .025
VW
0-20/20-40 ... 0.050 .100 .100
0-20/40-60 - ... ...
20-40/40-60 ... 0.025 - .100 .100
MB
0-20/20-40 Mri p..-.
0-20/40-60 NU UAIA
20-40/40-60 ' .100 .050 - .025 - .050
As-Received versus Control Fuel/Oil Emissions
Previous results and discussions have dealt exclusively with Phase 3, where
control diesel fuel and fresh manufacturer specified lubricating oil were used in all
vehicles. The "as-received" fuel test condition was usually similar to the control
condition, with the notable exception of the winter season when gasoline, kerosene and
other additives were commonly combined with diesel fuel. Analyses of as-received
and control (AEL) fuels are given in Appendix C. While the difference between Phase 1
and Phase 3 might be expected to be quite small for regulated emissions, the effects
on particulate, residue, extract and bioassay characterization were unknown and con-
sidered to be potentially greater.
A first approach to examining the data base for emission differences between
phases was to form a ratio of emission results from Phase 1 to those of Phase 3 for
each specific vehicle test cycle and emissions parameter. These rati'bs were then
grouped by manufacturer, and Table 4.11 gives the average of these phase ratios and
the associated coefficients of variation (CV). Bioassay data are found in Table 6.2.
The phase ratios are, with few exceptions, very close to unity. This suggests little
40
-------
MEAN
CV%
MEAN
CV%
MEAN
CV%
MEAN
CV%
.9369
54.85
1.0402
30.08
.8309
25.09
1.0726
30.84
.9741
8.71
1.0412
16.03
.9455
10.62
1.0313
H. 79
—i -— *» ii
1.0238
5.95
.9901
6.13
1.0000
7.17
.9888
9.90
1.0014
4.61
.9933
5.10
1.0080
6.9J
.9992
4.74
.9888
17.56
1.0477
19.90
.9884
13.69
1.0287
20.75
.9386
28.88
1.1464
73.25
1.0459
19.14
1.1158
28.45
1.0343
18.32
1.0538
42.67
.9828
14.41
1.0244
28.15
average effect of the fuel/oil changes between Phases J and 3.
'TABLE 4JI. RATIO OF PHASE 1 EMISSIONS TO PHASE 3 EMISSIONS
HC CO NOX MFC PART EXT RES
CM
VW
MB
OTHER
The Mercedes-Benz group hydrocarbon data and the Volkswagen and "other"
group extract data exhibit the most pronounced Phase 1 to Phase 3 differences. No
explanation can be given for these findings until further analysis of the collected fuel
and oil samples are performed.
The CV values for HC were much higher than any other gaseous parameter, and
at least some of this resulted from the seasonal fuel variation mentioned above. The
CV values for particulate, extract and residue were in the 13-29% ''ange, except for
the Volkswagen group with significantly higher CV values for extract and residue. The
magnitude of these CV values suggests that there are significant individual changes
between phases. The CV values found for repeated particulate measurements
reported elsewhere in this report are about 6%, and were much Jess than the phase
ratio values. Thus some, but not all, of the phase ratio CV was due to measurement
uncertainty. A more detailed fuel/oil correlation analysis will be required to further
define the basis for the observed emission changes between "as-received" and "control"
test conditions.
Emission Changes Between Successive Tests
This section describes the changes observed in various emission parameters
between successive tests. The average interval between tests was about 12,000 miles,
with a standard deviation of +5,000 miles for all tests. Also the mileages bounded by
the first and last test of the sample group vehicles were not uniform. These factors
make the computation of emissions change with accumulated miles a difficult concept
to establish for ail cars on a common basis. The new-car certification definition of
emissions deterioration could thus not be applied to these vehicles. Furthermore, the
emission changes between successive tests on these in-use vehicles frequently changed
in response to factors other than mileage accumulation per se, and using mileage
accumulation as a singular regression variable was not found to give an accurate
representation of the emission changes observed. Engine mechanical work, injection
timing, adjustments, etc. were often more dominant uncontrolled factors causing
larger emission changes than those caused by accumulated mileage between successive
tests. It was recognized, however, that the in-use data base could be used to provide
indications of emissions changes between successive tests on in-use vehicles over
periods of mileage accumulation.
A simple method of representing emission-mileage effects was to present the
HI
-------
absolute value of the emission parameters as measured at increasing mileage for each
vehicle. This was an obviously useful approach for certain applications, and tables and
figures of this type are presented in the Mileage Accumulation Effects portion of this
section. An alternate method of comparison was developed which describes the
change in emissions as a percentage of the emission at the first test of any two-test
interval. Expressing the emission change as a percentage of the previous test value,
the absolute emissions level was removed, making direct comparisons among high and
low emission vehicles possible. To account for the variation in test interval mileage,
the change in c.'-issions between any two successive tests was normalized to a 10,000
mile basis. The factor which was adopted for the following comparisons was thus the
change in emissions between any two successive tests, expressed as a percentage of
the emissions level of the first of the two tests in question, nor talized to a 10,000
mile basis, or %V. Thus we define:
(Xj - X0) 106
%V — ___—^——
(x0) (MI-MO)
where: Xj = emission value at current test
Xo = emission value at previous test
MI = vehicle mileage at current test
Mo = vehicle mileage at previous test
The project data base was processed to provide these %V values for the FTP and HFET
cycles. For vehicles with n total tests at different mileage accumulation points, this
resulted in n-1 matrices of emisJon change values. The FTP %V data are given in
Table 4.12.
While these computations could be carried out for all specific test cycles in each
test, the summary results in the form of distribution curves are shown for only the
FTP and HFET cycles in Figures 4.22-4.25. The %V values are both positive and
negative, indicating increases and decreases in emissions between successive tests.
When all change values arc averaged, the mean provides a rough indication of long-
term changes with accumulated miles.
This computation and data presentation framework does not represent aJl that
could be done with project data to address the question of changes with mileage, but
does provide a first view of the data from the project in terms of changes as they were
observed.
Figure 4.22 gives the FTP Phase 3 frequency distributions for HC, CO, NOX, and
MPG. The %V mean value and CV are given for each distribution. Smoothed
distributions are shown rather than discrete data to facilitate visual comparisons
among parameters. Figure 4.23 gives the same type of frequency distribution for
particulate, residue, and extract for the FTP. Figures 4.24 and 4.25 give" the same
data p-esentations as Figures 4.22 and 4.23, except the results are for the HFET. Both
HFET and FTP plots for corresponding emissions parameters are very similar. In
Figures 4.23 and 4.25, the extract exhibits a larger degree of variation than the
residue, with the total particulate between these values being a composite of both.
As a result of the smoothing of discrete data in Figures 4.22-4.25, the
distribution curve mean given in the figure may be somewhat different than the true
data mean given in the upper left of each figure. Furthermore, the data sets from
which the distributions were obtained included all project test vehicles and all test
intervals. Segregation of the data by vehicle manufacturer type could reveal
42
-------
TABLE ».I2. FTP 94V BY VEHICLE CROUP AND TEST MILEAGE RANGE
CAR
CROUP
CM
Vf
MB
ALL
MIDPT
MILEAGE
RANGE*
0-20 K
20-40 K
40-60 K
ALL
0-20 K
20-40 K
40-60 K
ML
0-20 K
20-40 K
40-60 K •
ALL
0-20K
2O-40K
40-60 K
ALL
n
X
0
n
X
0
n
X
0
n
X
0
n
X
0
n
X
0
n
X
0
n
X
o
n
X
0
n
X
0
n
X
0
n
X
<}
n
X
o
n
X
o
n
X
o
n
X
0
HC
1.4
18.)
*
29.9
26.)
5
-0.7
12.7
24
It.)
25.7
0.0
48.6
-6.5
5.8
2
-14.9
10.0
15
-1.4
28.6
NO,
-0.6
15.4
1.)
5.5
5
-).6
6.1
20
-1.0
9.5
-5.2
5.0
17.5
36.)
2
6.0
9.4
15
).?
16.6
CO
4.7
9.4
0.9
8.0
5
7.9
7.5
24
J.8
8.7
12.8
57.5
2.9
20.7
2
2.6
5.)
15
6.3
32.2
FUEL
MPG
-1.5
6.0
y
0.6
1.6
5
2.0
4.1
24
0.4
4.5
).7
5.1
0.8
5.1
2
0.2
5.)
15
1.8
5.2
PT
11.4
18.4
9.0
10.0
5
).o
f.7
25
7.9
14.0
5.8
55.8
16.4
4).)
2
15.2
9.5
14
6.7
37.0
EXT
35.1
25.4
)2.0
41.0
5
7.0
12.6
25
27.4
34.4
24.7
111.7
9.3
31.J
2
31.4
14.6
14
16.9
65.4
RES
6.4
24.0
4.)
17.6
5
2.5
10.3
25
3.0
17.4
8.1
69.6
20.8
58.2
2
11.2
15.4
14
7.4
46.8
»EXT
24.9
36.8
21.1
)9.0
5
3.6
8.8
25
18.8
)2.1
26.0
98.2
4.6
56.1
2
14.6
21.7
15
12.8
59.4
REV REV
OTPTTjfTXT
7.5 -0.)
88.9 94.6
8.0 -6.8
56.9 38.6
4 4
34.6 31.5
54.0 43.2
23 23
9.2 0.)
64.8 62.4
195.5 277.5
179.0 298.2
297.9 562.4
586.4 1065.0
1 1
20.8 21.7
-
13 13
J4I.O 232.3
291.3 521.3
. __-NO DATA
19.9
)2.0
1 .
-9.0
-
S
13.5
29.6
14
5.J
30.4
19
1<|.9
29.8
12
1.7
17.0
58
S.S
26.0
-11.2
9.5
1
-04
-
8
-9.1
9.1
U
-2.4
11.8
19
-0.2
18.2
12
0.3
6.5
58
-1.0
12.3
-7.5
10.2
|
-5.1
-
8
-7.7
R.8
id
7.j
32.9
19
-1.7
13.1
12
6.9
11.8
58
2.9
19.1
8.7
11.0
1
-0.7
-
8
7.5
9.9
1"
0.0
6.1
19
4.)
7.5
12
0.5
4.1
58
2.1
6.2
-0.7
22.2
1
-9.5
-
8
-2.1
U.I
14
9.8
33.9
20
6.8
23.9
12
6.4
11.3
58
5.9
22.9
7.7
14. 1
1
-2.1
-
8
5.3
12.7
;«
30.8
64.9
20
17.?
35.4
12
11.4
20.2
58
19.2
42.1
-1.5
23.1
1
-11.6
-
8
-3.1
19.9
14
6.1
42.5
20
5.8
26.5
12
5.2
13.7
58
3.4
27.0
10.6
14.2
1
9.1
-
8
9.3
12.4
14
24.1
60.9
20
11.6
32.9
12
4.8
15.0
59
13.7
37.8
36.4 17.2
63.6 60.9
1 I
39.2 25.6
-
4 4
37.1 19.3
51.9 49.9
14 14
71.2 95.7
153.3 228.2
16 16
78.0 121.1
249.7 452.3
10 10
20.6 17.5
40.1 30.2
49 48
47.7 67.5
166.5 288.5
•The midpoint odometer mileage between successive vehicle tests.
-------
FTP-GASEOUS EMISSIONS
PROJECT FREQUENCY DISTRIBUTIONS Of %V
XV< XtMISSION CHANGE BETWEEN 10.000 MIlC TESTS
Figure 4.22 Frequency distributions of %V for FTP
gaseous parameters.
FTP-PfiRTlCULATE EMISSIONS
PROJECT FREQUENCY DISTRIBUTIONS Of %V
XVI % EMISSION CHANCE BETWEEN «,OCOMn.C TESTS
RESIDUE (g/mi)
EXTRACT (j/m.)
WlT[»v«l«> »
MM X V • t*.f X
EXTRACT (X)
Figure *.23 Frequency distributions of %V for FTP
particulate parameters.
-------
HFET-GASEOUS EMISSIONS
PROJECT FREQUENCY DISTRIBUTIONS OF %V
MFET-PARTICIAATE EMlSSONS
UtOJECT FREQUENCY1 DISTRIBUTIONS OF %v
-P-
Vi
XV « XEMISSIOH CHANGE BETWEEN 10,000 MltC TESTS
HYDROCARBONS
INT[»*is> 5«
MttX
FUEL ECONOMY (m/gol)
•«0 -40 -K 0
(%V)
•*o *•<-»
XV X EMISSION CM4NOE BETWEEN 10.000 MLC TESTS
PARTICUIATCS (g/
RESIDUE Ig/im)
MTfKvAlS • S?
W( AM X V i 4.1 X
C v < IUX
EXTRACT (g/iru)
-to -«n -«o
(XV)
Figure 4.24 Frequency distributions of %V for
HFET gaseous parameters.
Figure 4.25 Frequency distributions of %V for
HFET particulate parameters.
-------
differences not apparent in the overall grouping presented here.
Statistical Tests on %V—
The general parameter, %V, has been previously defined as the change in a
vehicle emission parameter between two successive tests, calculated on a 10,000 mile
basis. The %V is, effectively, a deterioration rate (assuming that emissions are
increasing) in relative (taken as a percentage increase from the last test) rather than
absolute terms. These parameters may, therefore, be used in a comparative manner to
examine mileage accumulation effects or vehicle group differences.
One method to compare deterioration rates against mileage accumulation levels
was to test for statistically significant differences among the means at c. v> mileage
level. From Table 4.12, means were tested pairwise using the Smith-Satterv.c. test
(32) and an "equal population means" null hypothesis. For example, within the u ~ral
Motors group, the mean deterioration rate, %V, for HC is highest in the 20.. ; to
40,000 (20-40K) mile range. The difference between the %V mean in the 0-20- ™d
20-40K ranges was judged statistically significant since the t-statistic for th it-i
pair leads to the rejection of the null hypothesis at the 95% confidence (o=0.0:
Similarly the General Motors group %V HC means at 20-40K and 40-60K mil, *ve
judged to be unequal at the 97.5% confidence (o = 0.025) level. The difference *.V
HC means at the 0-20K and 40-60K levels was not judged to be statistically sign^.cant
as the t-statistic o-level was greater than 0.100 (less than 90% confidence reject..in of
null hypothesis). Table 4.13 gives the a-levels which were achieved by the t-stv..-.:.-
for pairs of means: o-levels in excess of 0.100 (o=0.100 was considered marginal;^
significant) are not listed individually in the table but are represented by a oas:-..
There were insufficient data for statistical comparison of %V means at mileage -ns
in the Mercedes-Benz group.
TABLE 1».I3. SIGNIFICANCE LE^EI. (a LEVEL) FOR DIFFERENCES IN %V MEA':C
BY MILEAGE RANGE AND BETWEEN VEHICLE GROUPS
VEHICLE GROUP &
MILEAGE LEVEL HC NOx CO FUEL PART EXT RES % EXT REV/ REV
(1000 MILES)- MPG wgPpgEXT
CM
0-20/20-ftO .050 -. ... .....
0-20/KO-60 ..... .025 - .100
20-40/00-60 .025 .100 .IOJ - - .100 ... .JQO
VW
0-20/20-itO - - - - -- . - ..
0-20/40-60 --..- .....
20-
-------
In Table 4.12, standard deviations on the order of or in excess of mean values are
frequently observed. This degree of scatter in the data was a major factor in the low
level of statistical significance among data pairs reported in Table 4.13, and is
especially evident in the VW %V data. Nonetheless, some differences appeared to be
significant: GM %V for HC at 0-20/20-40K and 20-40/40-60K miles; GM %V for
extract at 0-20/40-60K miles; GM/VW %V for HC; GM and VW/MB %V for NCX;
GM/MB %V for CO; GM/MB %V for fuel economy; and GM/MB %V for extract.
This analysis has relevance in estimation of emission deterioration factors and
the rate of deterioration with mileage accumulation. Comparison among vehicle
groups may also be made. The basis of comparison, however, is the % change in
successive tests, a relative measure, rather than absolute changes such as presented in
Figures 4.11-4.18.
Non-Volatile Hydrocarbons (NVHC)—
The HFID (heated flame ionization detector) measurement of gaseous hydro-
carbon emissions includes any hydrocarbons associated with the particulate matter
that volatilize in the hot filter (190°C) in the HFID probe system. The particle-bound
HC may not all volatilize at 190°C, but work by Cuthbertson trt. aJ. (33) confirms that
most of these HC species will not remain in the hot filter under these conditions.
The solvent extraction of particulale laden filters also provides a measure of the
particle-bound hydrocarbons, and although this measure may not represent an absolute
measure of particle-bound HC, a carefully performed extract determination is the
most consistent indicator in this study for particle-bound HC mass.
The solvent derived extract provides, in comparison with the HFID measure-
ments, an approximate indication of how much of the HFID mass could be resulting
from the volatilization of particle-bound hydrocarbons in the 190°C hot filter. The
non-volatile hydrocarbons (NVHC) term is thus defined as:
Solvent Derived Extract HC x 100
NVHC(%) =
NVHC is the percentage of the HFID value that could reasonably be expected to be
resulting from particle-bound HC volatilization as compared to dilute exhaust (gas-
phase) HC components.
Manufacturer group average results for the various test cycles are shown in
Figure 4.26. For all groups the higher speed cycles tend to exhibit higher percentage
values of non-vciatile hydrocarbons. Mileage accumulation effects on NVHC are
shown for individual vehicles within each manufacturer group in Figure 4.27 a-d. The
data from four test cycles were averaged at each vehicle test to obtain a four-cycle
average (FTP, CFDS, HFET, 50C) parameter for mileage accumulation plots. The
mileage accumulation trend for the General Motors group is more consistent than for
the Volkswagen group. The Volkswagen group exhibits both increasing and decreasing
trends, whereas most other sample group vehicles showed an increase in NVHC with
increased mileage accumulation. This indicates that increases in total hydrocarbon
emissions with vehicle age were due more to increases in particulate extract than to
gas phase hydrocarbons. However, these results must be considered in the context of
the dilution tur.nel where they were measured, and the interaction of particulate/gas
47
-------
phase hydrocarbons may not be the same as for direct atmospheric emission.
O GENERAL MOTORS
A VOLKSHRKN
+ MERCEDES-BENZ
g, X OTHER VEHICLES
IDLE NTCC FTP CFDS HFET 50 C
Figi re <(.26 Cycle variations of non-volatile hydro-
carbons by vehicle groups.
-------
IS 30 US 60
TEST MRER5E IN IO:D-S
75
S3
e CM 01
» LWt 06
+ CWt 08
X CM 09
# CM 11
30 60 93 120
TEST MRERGE iN :000 S
ISO 163
S
I
*m.
O CRR 12
A CM 13
+ CAR IV
X CRR 19
IS 33 < CflR 21
d
IS 30 IS
TEST HILERCE IN
60
Figure '».27 Mileage accumulation effects for average of FTP, CFDS, HFET and 50C
non-volatile hydrocarbons by vehicle groups: (a) General Motors group, (b) Volkswagen
group, (c) Mercedes-Benz group, (d) other vehicles.
-------
SECTION 5
FUEL ECONOMY
INTRODUCTION
The project was designed to permit up to three independent measures of vehicJe
fuel economy. As part of the laboratory emissions testing carbon balance fuel
consumption, in accordance with EPA certification procedure, was measured for each
test cycle. In addition, each vehicle was outfitted with under-hood flow totalizing
meters and an engine hour meter. Two fuel meters were used on each vehicle to
measure fuel flow into and from the injection system, thus permitting calculation of
fuel consumed between tests by subtraction. For the two loan vehicles (Car //I and
Car //5), fuel and engine hour meter readings were recorded at each fueling to obtain
fuel economy and average speed data for each tank fill.
Fuel economy by the carbon balance method was determined for the FTP, CFDS
and HFET for the first 34 tests and for the FTP, CFDS, HFET, 50C, NYCC and IDLE
for the last 46 tests. The data presented in this section are for Phase 3 only. In
addition to the data presented here, group averages with standard deviations can also
be found in Tables 4.1 through 4.6. Table A-18 in Appendix A gives individual vehicle
averages and the corresponding Phase 1 data are given in Table A-17. These tables also
give the fuel economy of the individual bags of the FTF.
CYCLE VARIATIONS OF FUEL ECONOMY
In Figure 5.1 the carbon balance fuel economy of the General Motors,
Volkswagen, and Mercedes-Benz groups are presented for the various test cycles.
Figure 5.2 gives the same data in units of ton-mile/gallon to normalize fuel
consumption by vehicle test weight. In terms of ton-mile/gallon the General Motors
group had slightly higher fuel economy than the Volkswagen and Mercedes-Benz groups
at the 50C, but exhibited a significant penalty for low speed test cycles with large
speed variability. All three vehicle groups exhibit essentially two distinct regions on
either side of the CFDS test cycle, which may be related to lower speed operation at
lower transmission gear ratios.
MILEAGE ACCUMULATION EFFECTS ON FUEL ECONOMY
Fuel economy in miles per gallon as determined by the carbon balance method is
shown for the individual vehicles in Figures 5.3a, b, and c and for the groups together
in Figure 5.3d. All data are the Phase 3 FTP. The General Motors group shows no
mileage accumulation effects either for single vehicles or for the group as a whole.
The 30 tests on six vehicles over a 75,000 mile interval range from 18.2 to 21.8 mpg
50
-------
le-
g-
s'
go.
fi«H
5
8
GENERftL MOTORS
VOUSHflCEN
MERCEDES-BENZ
• GENERAL MOTORS
A VOUSHflGEN
+ MERCEDES-BENZ
NTCC
FTP CFOS HFET 50 C
NTCC FTP CFOS HFET SO C
Figure 5J Cycle variations of fuel economy,
miles/gallon, by vehicle group.
Figure 52 Cycle variations of fuel economy, ton-
mjle/gallon, by vehicle group.
with an average of 20.0+1.0 mpg. The Volkswagen group has fuel economy substan-
tially higher than either the General Motors or Mercedes-Benz group. Or. a mpg
(absolute) basis these variations are similar to those of Mercedes-Benz but on a
relative scale they show less scatter because of the higher fuel economies. All five
vehicles have higher fuel economy than any other vehicles. Some vehicles show fuel
economy increases with increased mileage but no overall trend is apparent. Car #6,
the only vehicle in the group with a 5-speed transmission had an average fuel economy
of 43.7 mpg compared to 43.0 mpg for the 4-speed models. (The advantage of the 5-
speed transmission was more apparent in the HFET data where the 5-speed average
fuel economy was 61.8 mpg versus 56.4 for the 4-speed transmission models. See
Appendix A, Table A-18.)
The Mercedes-Benz group of Figure 5.3c have a range of fuel economies from
20.9 to 28.9 mpg with an average of 24.7+2.2 mpg. Variations for individual venicles
are larger than for the General Motors groups and display a trend of increased fuel
economy at higher mileages.
The composite of the vehicle groups in Figure 5.3d shows the data sets to be
almost mutually exclusive and following the order GM < MB « VW. Part of the
variation in the Volkswagen group is probably due to the mix of model years and
transmissions while the Mercedes-Benz group contains vehicles of different years,
models, engines and transmissions.
OVER-THE-ROAD FUEL ECONOMY
The fuel meters were positive displacement type totalizing meters (Servis
Recorder Company) with a volumetric accuracy of +1%. Fuel measurements were not
compensated for temperature variation, and distance traveled was determined from
51
-------
g
§
O CRR 02
* CHR 03
+ CRR OM
X CRR 05
O CfiR 07
* CAR 16
33 «5 60 75
1ES1 MRERGE. JOOO'S
90
IK
to.
O CAR 0)
A cm? 06
+ CRI? 08
X CBf? 09
« CRR II
30
60 90 120 ISO
TEST HILER&E. 1000'S
183
O CHR 12
A CRR 13
•f CHR U
X CRR 19
15
3f US 63 75
TEST MILERGE. 1003'S
90
8'
0 33 B3 93 )2P 153 ]E:
HS1 MJLERGE. JGSS'S
Figure 5.3 Mileage accumulation effects for FTP fuel economy, mpg, for (a) General
Motors group, (b) Volkswagen group, (c) Mercedes-Benz group, (d) all three groups.
-------
the vehicle odometer. Therefore, the overall fuel economy and vehicle speed data
were less accurate than would have been possible with research grade instrumentation.
The basic intent of these meter installations was to obtain data that corresponded to
those which a vehicle owner would obtain based on tank-fill records. For this purpose
the meter installations were very satisfactory. This point is demonstrated by the
meter records for Car //I and Car //5 as compared to fuel log book data. These data,
collected over approximately 50,000 total miles of vehicle usage differed by 1.6%
between the meter and tank-fill data. Table 5.1 gives the over-the-road fuel economy
and vehicle speed data obtained between vehicle tests by the underhood meters for,
each vehicle in the sample group.
Figure 5.4 gives the over-the-road fuel economy and vehicle speed results for
individual tank-fills as determined from the underhood meters for Car #1. The smooth
line connects the laboratory carbon-balance fuel economy results for this vehicle. At
all average speeds, the carbon balance fuel economy was higher than meter fuel
economy. Figure 5.4 also contains the frequency distributions of fuel economy and
average speed from underhood meters and vehicle odometer/hour meter. Car //I was
used frequently for short trips and around-the-town driving. Its average speed was
about 30 mph. The 32 data points are evenly distributed about the FTP laboratory fuel
economy value. The CFDS laboratory value is an upper limit for the fuel economy.
The carbon balance fuel economy for the FTP, a cycle,with a lower average speed than
most of the over-the-road data, best approximated the average over-the-road fuel
economy. The FTP was, therefore, the best laboratory fuel economy indicator of in-
use fuel economy for this vehicle. A similar conclusion was reached in an earlier in-
use study of 56 catalyst-equipped cars from this geographic area (34). Of course, the
over-the-road data include many factors not specifically accounted for in controlled
laboratory testing, such as road variations, hills, snow and slush, payload, etc.
Figure 5.5 gives results for over-the-road and carbon balance fuel economy
measurements for Car //5 in the format described above. Car //5 was frequently used
for long, high-speed trips as evidenced by the number of average speeds above 40 mph.
The average overall speed was about 35 mph. In this case about 75% of the over-the-
road fuel economy exceeds the FTP but only one value exceeds the CFDS. The FTP
was still the best laboratory fuel economy indicator for actual in-use economy.
Figure 5.6 gives results in the same formates Figures 5.4 and 5.5, except that
the over-the-road data presented are for the General Motors group between tests in
contrast to the tank-fill data of Figures 5.4 and 5.5. Each data point now represents
5-10,000 miles of vehicle use. The carbon balance results are now the average of all
tests for all General Motors group vehicles. The same general conclusions were
evident for these data as from the previous results for individual tank fills on single
vehicles.
53
-------
TABLE 5.1. FUEL ECONOMY AND VEHICLE SPEED FROM UNDERHOOD METERS
ENGINE
CAR MAKE
6 YEAR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
VW
79
OLDS
79
OLDS
79
OLDS
79
OLDS
. 79
VW
80
CADILLAC
79
VW
78
VW
79
DODGE
78
VW
77
MB
77
MB
78
MB
79
AUDI
79
OLDS
79
PEUGEOT
79
OLDS
80
DISP.
MODEL Oilers)
RABBIT
CUTLASS
CRUISER
CUTLASS
CRUISER
98
REGENCY
CUTLASS
CRUISER
RABBIT
ELDORADO
RABBIT
RABBIT
D-IO
RABBIT
240D
300CD
240D
5000
DELTA
88
504
CUTLASS
CRUISER
1.5
5.7
5.7
5.7
5.7
1.5
5.7
1.5
1.5
4.0
1.5
2.4
3.0
2.4
2.0
5.7
2.3
5.7
TRANS-
MISSION
M4
A3
A3
A3
A3
M5
A3
M4
M4
A3
M4
M4
A4
M4
M5
A3
M4
A3
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
MPG
MPH
43.2
31.3
42.9
23.8
38.8
17.1
16.9
24.3
35.1
46.4
33.9
20.2
28.3
36.8
42.2
31.5
17.0
28.2
42.6
45.0
36.8
23.0
29.4
33.8
34.6
38.0
32.2
31.7
40.8
59.8
DATA FOR TEST INTERVALS
-
26.5
44.8
24.4
41.8
14.2
13.7
23.1
35.3
42.9
32.7
28.5
35.6
37.2
29.6
40.0
34.9
38.1
26.6
280
29.4
23.6
31.0
33.6
21.5
32.9
30.0
40.2
33.7
42.7
28.2
44.3
25.8
43.6
21.1
39.5
34.7
46.5
32.8
19.5
27.0
37.6
41.1
32.3
41.0
-
31.3
30.7
32.4 15.0 19.2 17.9
44.8 45.1 45.5 50.3
23.0 - 23.4 16.8
39.0 40.1 39.8 29.7
39.7 36.5
Note: No meteri were installed on Cars I) 19, 20, 21; missing data implies defective meter operation.
5«f
-------
50
£40
*3O J
""20J
10
60i
10-
0 10 20 30 4b SO 6O 0
8*
10
•»•> DATA TAKEN FROM UNDERHOOO METERS AT TANK FILLS
• FTPC I LABORATORY CARBON BALANCE FUEL ECONOMY DATA
IcFDS / AVERAGE OF ALL TEST5; CT=2MPG
THFETl
10 (5 20 25 30 35 40 45 50 55
SPEED- MPH
^> 10 15 20 25 30 35 40 45 50 5i-
SPEED-MPH
Figure 5.4 Over-the-road fuel economy vs. speed for Car //I.
so
« 40
|
§20
10
50
W2O
10-
•f t DATA FROM UNDERHOOD METERS TAKEN AT TANK FILLS
• NYCC "I
«FTP I LABORATORY CARBON BALANCE FUEL ECONOMY DATA
ACFDS /AVERAGE OF ALL TESTS. C' = 1MPC
Tj
0 10 20 30 40 50 60 0 5 <0 15 20 25 3O 35 40 45 50 55
fREOuEr»CY DiS'BiB-iTiOM uprrn MPM
OF oes£RvAT,ONS:%i SPEED-MPM
5* »•
10-
10 15 20 25 30 35 40
SPEED-MPH
Figure 5.5 Over-the-road fuel economy vs. speed for Car //5.
55
-------
50
f
E 30
"20
50
20
10'
+ DATA TAKEN FROM UNDERMOOD METERS BETWEEN VEHICLE TESTS
•> NYCC "I
• fTP I LABORATORY CARBON BALANCE FUEL ECONOMY DATA
A CFDS / AVERAGE Of ALL TESTS; O-« I 3 MPG
0 10 20 30 4C 50 60 0
=. 60
40
•f-f
4?*
5 10 15
20 25 30
SPEEO-MPM
35 40 45 5C 55
5 10 15 20 25 3G 35 4C 45 50 55
SPEED-MPM
Figure 5.6 Over-the-road fuel economy vs. speed for General Motors group.
-------
SECTION 6
BIOASSAY CHARACTERIZATION
GENERAL METHODOLOGY
Extract samples from the 50 cm x 50 cm filters were individually tested for
extract dose/response potency by the Ames Salmonella typhimurium/microsome
method with tester strain TA98, without metabolic activation (TA98-). Tester strains
other than TA98 were used, and assays with metabolic activation were also performed.
However, these assays were mainly applied to those special samples subjected to
detailed chemical analysis and these data are included in Section 8 dealing with
chemical characterization of extracts. A detailed description of the entire assay
protocol will not be presented here, except to note that protocol guidelines specified
by Ames (35) and modified by deSerres and Shelly (36) were followed unless
specifically noted. Certain other procedural aspects of the assay will be described
since these may provide insight to the TA98(-) assay as applied to the over 1100 extract
samples of the in-use diesel portion of the project.
The solvent-free extracts which were stored at -80°C were redissolved in 75%
dichloromethane, 25% acetone and split into two portions. One portion was redried
and returned to freezer archive storage. The other portion was redried, weighed and
redissolved in DMSO for the bioassay. Duplicate plates were run at extract doses of 0,
10, 20, 30, 40, 50, 75, 100, and 200 ug. The raw data from each plate count (Biotran
Model C-lll) were entered into the computer from which a data plot was obtained. The
data were fitted to several straight line regressions utilizing first, all the data, and
subsequently with the 200 ug dose data eliminated from the regression, and then with
the 200 and 100 ug doses eliminated, etc. Summary results for each regression were
printed on the display plot. Three typical examples of these regression/plot outputs
are shown in Appendix B. Judgement was frequently needed to select the "best" slope
of the dose/response curve by examination of the raw data, comparison of different
regression fits, R^ values, intercept of the fitted line, etc. This judgement was not
absolute or unequivocal but was applied consistently to obtain a slope (revertants/Mg
dose of extract). R^ values were typically greater than 0.90 and usually the highest
dose of the regression line was 50, 75 or 100 ug. 12-16 individual raw plate counts
were typically used to determine the slope.
BIOASSAY STANDARDIZATION
Application of the Ames bioassay to a mileage accumulation study of diesel
emissions was an extension of the assay beyond past applications. In this study, we
attempted to compare assay data in a semi-quantitative manner from the same
vehicles over a 2-year period. To make thp data more consistent all extracts from a
given vehicle test were assayed as a batch. Thus 12 individual extracts were assayed in
parallel to remove as many biological variables as possible in making comparisons
57
-------
within each vehicle test.
In eddition to the normal positive co.itrols to verify the activity of the tester
strain, a single bulk diesel paniculate extract sample was used to verify the activity
of the tester strain. Figures 6J and 6.2 show the project chronology of the 5 yg dose
of 2-NF (2-nitrofluorene) and the 0.5 yg dose of NQNO (4-nitroquinoline-N-oxide).
The average of 74 measurements for those two positive controls was 736^11% and
563^11%, respectively.
h(jAW
15 30 45 61 75
CrlRONOLOOJCfli. RUN OIQ£R
93
15 30 US 63 75
CHRONOLOGIC^. RUN ORDER
Figure 6.1 Chronology of Ames activity of 5 ug dose
of 2-NF (2-nitrofJuorene).
Figure 6.2 Chronology cf Ames activity ol 0.5 ug
dose of NQNO Ct-nitroquinoline-N-oxide).
The paniculate extract control sample used throughout the assay originated from
Car //5 as part of a particulate collection effort to produce a large extract sample for
chemical analysis. Aliquots of this extract were kept in freezer storage and included
in the assay of each batch of 12 extracts from a single vehicle test. Rather than use a
single dose of this control extract in an assay wherein the real use of the data was
slope determination, a 4 dose control extract protocol was adopted. Control extract
doses of 25, 50, 75 and 100 yg were used as part of each batch of extracts and the
slope of the control standard was computed. These results are presented in Figure 6.3
showing chronologically the 74 slope values computed from the single extract sample
used throughout the project. The mean slope throughout the project was 2.13
revertants/yg extract with a coefficient of variation of 13% which compares favorably
with the 11% found for the pure chemical controls.
The consistency of the diesel extract control activity throughout the project
provides an interesting measure of sample degradation, a topic of considerable concern
and experimental attention. In future examination of the project data the control
extract slope value may be used to normalize project data to a constant control slope
value, but this step was not possible in the preparation of the report.
58
-------
IS 30 MS 60 75
ChKOWLOSJCRL RUN ORDER
90
Figure 6.3 Chronology of Ames slope ol diesel
standard extract (25, 50, 75,100 ug doses).
FACTORS AFFECTING ASSAY RESULTS
Filter and Extract Storage
The relative consistency of the TA98(-) response to the diesel extract control
sample over the time period of the project was encouraging. The extract sample in
this case was kept in freezer storage. Additional data were collected to examine
storage effects on extract as filter bound particulate, dichloromethane (DCM)
solution, and dried extract. These investigations used the 50 cm x 50 cm filters from
the duplicate driving cycles of a given phase in the first 3tt vehicle tests. During this
portion of the project HFET and CFDS cycles were run in the afternoon and repeated
after the FTP of the next day. The filters from these duplicate cycles were
considered as identical, and this assumption was undoubtedly the limiting factor in
interpreting the experiments described below. Some of these filters were extracted
and assayed, some were kept as filters, and some were extracted with the extract kept
in DCM solution, while others were kept as solvent-free extract. All samples were
kept in freezer storage at -80°. Figures 6.^ and 6.5 show the number of days each
filter was kept in any of the above storage conditions, and also give the specific
TA98(-) activity.
In several cases it can be seen that the method of storage had little effect on the
mutagenic activity, while in others a significant change would be projected. Given the
uncertainty in what these "matched" filters would have given at time zero, it is
impossible to quantify the changes in these samples. The relative similarity regardless
of storage history does suggest that the direct acting mutagenic agents do persist in
these samples over an extended period of time.
59
-------
STORED AS
STOWED AS ExTRiCT
STORED AS DOM SOLUTION OF EXTRACT
SAMPLE
195 [7"
SLOPE
4T
vvr
207
698 \////////S////////\
|66
59
2 1
1 9
2 <
26
33
51
23
sr
(00 200 300
DAYS FROM FILTER COLLECTION
Figure 6A Effects of long-term sample storage on mutagenic response - storage as
extract in DCM vs_. storage as filter.
S'c«EO as
M'E-'THACT
STORED «S
XM SOLUTION
Of CKTRACT
SLOPE
24
I 0
1 3
I 9
*«» V//////////A
< 6
20
620 V//////A
V///////7/7///7A
1OO ZOO
DAYS FROM FILTER COLLECTION
300
SO
«0
Figure 6.5 Effects of long-term sample storage on mutagenic response - storage as
extract vs. storage as filter.
60
-------
Effects of Incubation Period
The length of incubation time for the assay protocol was 72 hours for most of the
work reported here. Figure 6.6 shows the results of an experiment to determine the
effect of incubation time on the assay response. Revertant counts are shown for
several chemical positive controls and the diesel particulate extract (DPE) control for
incubation periods from 48-120 hours. The relative consistency of all the results
indicates that the extract incubation time was not an important variable in the
observed revertant counts, and that the 72-hour protocol value was acceptable.
Plate Count Reproducibility
The reproducibility of the assay protocol within a given day was measured by
assaying 10 plates containing a constant 50 ug DPE dose. The results are shown in
Figure 6.7, snowing a 6% coefficent of variation within this particular test.
Effects of High Doses on Slope
The highest extract dose used in a linear regression will change the computed
slope even at low doses where the R^ values would indicate a satisfactory fit for both
choices. Table 6.1 shows a random sample of 10 extract samples wherein different high
doses of 30, 50, 100 and 200 ug were used to compute the slope. From the normalized
results at the bottom of Table 6.1 a rough indication of how the high dose choice
affects the slope can be seen. When the 30 ug dose is considered as the base, the 50,
100, and 200 ug high doses gave a 5%, 15%, and 42% reduction in slope respectively.
Activity Measurement from "Duplicate" Samples
As part of an experiment for another purpose, a number of 50 cm x 50 cm filters
were collected from Cars //I and //5 (Volkswagen and OJdsmobile loan cars) operated
repeatedly on the 50C cycle. A sideline of the experiment gave individual extract
samples for each filter. Unfortunately, the program did not include detailed extract
emissions data (% extract or extract g/mi). The individual extract samples were
assayed, and the slopes from each consecutive 50C are shown in chronological run
order in Figure 6.8. The Oldsmobile data are very uniform whereas the Volkswagen
data exhibit a marked increase throughout the test period. The magnitude of the
slopes for the Volkswagen in the third and subsequent runs v.-ere significantly higher
than for any other 50C values for the vehicle. No obvious vehicular reason can be
given for these results.
These data suggest that at the current state of capability the Ames TA98(-) assay
as a particulate characterization tool is subject to variability that impacts its
interpretation. One cannot distinguish slopes that differ by only a few percent.
Differences of 100% are likely to t/e indicative of real sample effects but could easily
result from unknown parameters in the experiment from which the samples originate.
The overall uncertainty in quantitative Ames data is the combination of the
uncertainties in the sampling and the lab analysis. It has become evident that control
of test parameters as usually practiced may not always provide reproducible Ames
data. Our knowledge of the test parameters affecting Ames data is very limited and
therefore a large uncertainty exists in its quantification. We have therefore analyzed
a large number of samples and grouped the data in the hope of reducing random error.
Remedies to systematic errors such as artifact production, sa.nple degredaiion and
6i
-------
KXX>
800
in
z
t 600
u
UJ
a;
400
2OO
Q
TA98(-S9)
OMBA
20 K9
o—o-—° °--~^
^^°
NONO
0»|.a
BoP
*2^-o °
OPE
0 20 40 60 80 <00 <2O
INCUBATION PERIOD (HOURS)
TABLE 6.1
EFFECT OF HIGH DOSES ON SLOPE OF DOSE-RESPONSE CURV
SLOPE (Reverunti/ng)
HUhett £»«« Uird In Fit (,.R)
S.ople 30 50 100 200
886 4.1 3.6 4.1 3.8
921 1.3 1.4 1.3 1.0
931 1.7 2.1 1.8 1.4
939 2.0 1.9 1.4 0.7*
965 2.3 1.8 1.5 1.8
987 2.1 1.9 1.4 1.0
1009 15.3 15.7 12.1 8.3
1033 4.2 4.1 4.5 3.7
1056 3.8 3.5 3.3 2.»
1068 3.2 2.3 2.5 2-2
Average 4.0 3.8 3.4 2.6
No realized 1 0.95 0.85 0.58
*ttot on linear portion of curve.
Figure 6.6 Effect of incubation period on rnutagenic
response.
in _
180
160
(4O
£ 130
<
c 10O
£ so
60
4O
20
TA98 (-S9I | jo-
DOSE: 50pg.RSO'6 GIO.
|^
0—^y'°~~ 0 0—0 5
°~">/ ss-
s
to JJ.
I
S0
ul"
ox
in"
O CfW 01 ft
A ^AA 4*C I \
• CBK 05 £•>_ / \
/ V
1 ^®s. /
/ ^°
/
/
/
I
Ox,,/
^^^ i A .
23456789 10
DETERMINATION
50 100 ISO 200 ?SO
TOTRt TIME OF SOC . MINUTES
300
Figure 6.7 Reproducibility of rnutagenic response
for ten replicate plates of Jiesel standard extract.
Figure 6.8 Reproducibility of rnutagenic response
lor replicate 50C filter samples.
62
-------
matrix interferences are generally beyond the scope of this work. The useful
application of the Ames test to the assessment of the bio-hazard of diesel participate
extracf has not been suficiently refined to permit unequivocal interpretation of the
quantitative data.
AMES TEST RESULTS
Fuel and Lubricating Oil Effects
Phase 1 and Phase 3 differ in the fuel and lubricating oil used. In Phase 1 the car
is tested with fuel and oil which were in the vehicle when received, while in Phase 3
control fuel and oil are used. The combined effect of this fuel and oil change was
investigated by comparison of TA98(-) activities of the two phases. For five driving
cycles (IDLE excluded) the ratio of activity of Phase 1 to the activity of Phase 3 was
calculated and called the 'cycle-phase-ratio". The average of the five cycle-phase-
ratios for a given test wao designated the "test-phase-ratio". The average of all test-
phase-ratios within a vehicle groi-p was defined as the "group-phase-ratio". Table 6.2
gives the group pha^c ratios data and the coefficients of variation for bio.^ctivity
parameters.
TABLE 6.2. AMES ACTIVITY, GROUP-PHASE RATICS
CM VW MB
// test-phase ratios 29 18 8
revertants Ratio 1.2'* 1.10 0.97
ug extract CV,% 42 70 28
revertants Ratio 1.13 1.26 1.09
ug particulate CV,% 39 7
-------
Mileage Accumulation Effects
In order to examine mileage accumulation effects on Ames activity, the Phase 1
and Phase 3 test cycle data from a given vehicle test (IDLE assays not included) were
averaged to obtain a single Ames activity parameter for each vehicle test. Figure 6.9
shows these data connected point-to-point for each car. While Ames activity does
exhibit large fluctuations with accumulated miles, Figure 6.9 suggests that mileage
accumulation is not the primary cause of the observed changes. This was supported by
examination of the data for the high-mileage Car #11 (120,000-170,000 miles). This car
exhibited Ames activity essentially similar to the average of other project cars. All
mileage accumulation effects are difficult to assess due to the problems of obtaining
consistent quantitative Ames activities over long time periods.
vt
15
60
75 90 105 120
TEST MILERGE / 1000
135
150 IBS
1BC
Figure 6.9 Mileage accumulation effects for averaged Ames activity.
Vehicle Group Averages
Figures 6JOa, b, c, d give the average values of An-.es activity for each cycle by
vehicle group. Four Ames activity characterization parameters are shown: specific
activity of the extract (revertants/yg extract); activity expressed in terms of a
common amount of emitted participate (revertants/ug particulate); activity expressed
per vehicle mile (revertants/mile); and fuel specific activity (revertants/kg fuel
consumed).
Some effects of test cycle are evident in Figures 6JOa, b, c, d especially for the
Volkswagen group and for the NYCC for all groups. The General Motors group
exhibited essentially unchanged extract and particulate based activity over all test
cycles.
The FTP and CFDS generally had the highest activity measure regardless of
vehicle group, test cycle, or activity parameter. No explanation is offered to account
for this result at present. Since the FTP is the only cycle with a vehicle cold-start
included, it is qualitatively different than other test cycles, as opposed to quantitative
variations in such parameters as speed, acceleration, idle time, etc. Early in the
-------
H
i
REVERTRNTS/UG SOF
O GtNERflL MOTORS M
A VOLISMAGEN
•» MERCEDES-BENZ
X OTH» CARS
REVERTRNTS/UG PRRTJCULRTE
O CENERfll MOTORS
A VOLKSMflGEN k
4 MERCEDES-BENZ •*
X OTHER CARS
NTCC nr cros trn so c IDLE
HTCC nr t/os WET se c IDLE
REVERTRNTS / MILE
O GENERAL HOTORS
A VOLHSMflGEN
+ MERCEOES-BOC
X OTf« UKS
§
R-
REVERTRNTS/KG FUEL
O KNEKft. MOTORS
X OTHER CARS
NTCC FTP CfOS MFET SO C
NTCC FTP CFOS MFEV SO C J&E
Figure 6.10 Cycle variations of average Ames activity for all vehicle groups in units
of: (a)revertants/yg extract, (b) revertants/ ug particulate, (c) revertants/mile, and
(d/ revertants/kg fuel.
-------
project it was noticed that FTP Ames activities were frequently higher than for other
cycles of the same day. A short series of special tests were conducted on Cars tfl and
//5 to examine the effect of cold-start operation on participate, extract, residue, and
Ames activity in the winter of 1981. These tests revealed the Bag 1 (FTP cold-start)
Ames activity (revertants/mi) to be 40-80% higher than the Bag 3 (FTP hot-start)
values (37). While this finding may contribute to an explanation of the relative Ames
activity results for the FTP, the Ames activity result for the CFDS will require a
more fundamental understanding of the interaction of various test cycles and complex
emission products.
The coefficients of variation associated with the group averages of Figure 6.10 are
quite large, and range from a low of 30-50% for the Mercedes-Benz group, 50-70% for
General Motors, to 70-100% for the Volkswagen group. While vehicle group and test
cycle differences are apparent, there are major unexplained variations for groups and
different activity parameters.
Bioactivity and Emissions Correlations
Linear correlation coefficients and the significance level of the coefficients were
computed for bioactivity and extract-residue data. Bioactivity was expressed in five
parameters: revertants/mile, revertants/kg-fuel, revertants/ug extract,
revertants/Ug residue, and revertants/ug particulate. Emission extract and residue
parameters were: grams extract/mile, grams residue/mile, grams extract/kg fuel,
grams residue/kg fuel, % extract, and the grams residue/grams extract ratio. Three
threshold significance levels (SL) were selected: SL=1 at alpha=0.05, SL=2 at
alpha=0.01, and SL=3 at alpha=0.005. The determination of the significance level was
made by testing a population zero correlation null hypothesis (H0: p=0) against the
population non-zero correlation alternative hypothesis (Ha: P^O) using the test
statistic z:
z = (/S3V 2) In [(1+r) / (1-r j]
where z is approximately normally'distributed, n is the number of samples, and r is the
sample correlation coefficient (32). From this test it was determined if the sample
correlation coefficients were statistically significant and, if so, the degree of signifi-
cance.
The data were stratified by driving cycle, test phase, and vehicle type to yield
individual sample data sets. Data were also segregated simply by: driving cycle and
test phase without regard to vehicle type; driving cycle and vehicle type without
regard to test phase; and driving cycle without regard to either vehicle type or test
phase. Five driving cycles were studied: FTP, CFDS, HFET, NYCC, and IDLE.
Correlation coefficients and signficance levels for the data segregated by driving
cycle and vehicle type (but not by test phase) are presented in Appendix E. By lumping
the test phases together, a greater number of statistically significant sample correla-
tion coefficients were observed than for segregated phase data sets. The results
showed the same general features, however, for both individual phase and the lumped
phase data sets. The large quantities of data generated by the correlation exercise
make it necessary to restrict this discussion to the more universal case of lumped
phases.
It is recognized that linear correlation may not appropriately describe all
parameter relationships, and this investigation, therefore, must be considered pre-
liminary in nature rather than comprehensive. The results are suggestive of possible
66
-------
parameter associations which should receive additional statistical attention in the
form of non-linear correlation and multiple parameter correlation studies.
Of primary consideration in this analysis was the determination of the degree and
significance of the correlation between emission(extract and residue) parameters and
the bioactivity parameters. The first item of importance, therefore, was to determine
the sample correlation coefficients and levels of significance between the emission
parameters themselves on grams/mile and fuel-specific (grams/kg fuel) bases, segre-
gated by driving cycle and vehicle type. These results are given in Table 6.3.
Approximately half of the observed correlation coefficients in Table 6.3 are not
statistically significant and only 20% are significant at the alpha-0.005 level. Thus
extract and residue do not, in general, exhibit strong linear correlation within the
context of these data.
TABLE 6.3 EXTRACT/RESIDUE SAMPLE CORRELATION COEFFICIENTS
DRIVING CYCLE AND VEHICLE GROUP OVER ALL TEST PHASES
BY
Driving
Cycle
FTP
HFET
CFDS
NYCC
Idle
Vehicle
Croup
CM
VW
MB
Other
All
-M
VW
MB
Other
All
CM
VW
KB
Other
All
CM
VW
MB
Other
All
CM
VW
MB
Other
All
Extract (9/Bilel*
Residue (g/nilel*
n i Si.
74 .011 0
47 .052 0
21 -.448 1
35 -.236 0
177 .218 3
73 .089 0
46 -.038 0
22 -.549 3
37 -.492 3
178 -.0(2 0
73 .216 1
48 .078 0
22 .648 3
36 .581 3
179 .164 1
29 .019 0
16 .602 2
9 .252 0
22 -.407 1
76 .378 3
72 .089 0
46 .223 0
22 .068 0
37 .260 0
177 .300 3
Extract (g/kg-fuel)
Residue (g/kg-fuel I
n r £L
72 -.040 0
47 .060 0
20 -.179 0
34 -.366 1
173 -.223 3
73 .049 0
46 -.033 0
22 -.191 0
37 -.473 3
178 -.214 3
73 .113 0
47 .073 0
22 -.295 0
36 -.502 3
178 -.141 1
29 -.140 0
If -.676 3
9 .131 0
22 -.079 0
76 -.184 0
29 -.199 0
17 .107 0
10 .460 0
22 .055 0
78 -.098 0
' \~
n • number of samples
r • sample correlation coefficient
St. • significance level where :
61 • 1 for 0.05 > a > 0.01
EL - 2 for 0.01 £ o > 0.005
EL • 3 for 0.005 £ o
• change units to g/ninute for
Idle driving rycle
In Table 6.3 it is observed that for all cars over the FTP cycle, extract and
residue on a grams/mile basis show a very weak (r=.218) positive association but show a
similarly weak (r= -.223) negative association on a fuel specific basis. This observation
may be explained by the moderate negative correlations observed between. fuel
economy (miles per gallon) and both extract grams/mile (r=-.485; SL=3) and residue
grams/mile (r= -.686; SL=3) for the FTP cycle.
67
-------
A very weak but statistically significant negative correlation for extract and
residue on a fuel specific basis (r= -.214; 5L=3) was observed for all vehicles over the
HFET cycle, and weak positive correlations were observed for all vehicles over the
NYCC (r= .378; SL=3) and IDLE cycle (r= .300; SL=3) on grams/mile and grams/minute
bases, respectively. Somewhat higher correlation coefficients were observed for
specific vehicle groups; however, it is cautioned that the sample sizes for these data
were small. Six data sets were observed to have correlation coefficients with absolute
values greater than or equal to 0.549 which represents 30% or more explained variance
as a result of a linear association.
It is not established here that extract and residue are independent parameters. It
is observed, however, that for most of the vehicle groups and driving cycles, the linear
correlation between extract and residue is either not statistically significant or, if
significant, is not strong.
Table 6.4 lists the sign of the sample correlation coefficient and the significance
level for each bioactivity-emission parameter pair by driving cycle and vehicle group.
Sample correlation coefficients with absolute values of 0.60 or greater are indicated in
Table 6.4 and listed specifically in Table 6.5. The general level of correlation between
parameters in Table 6.4 is very weak. Typically significant r values range from 0.2
to 0.4. It is observed, however, that, in spite of the weak degree of correlation,
independent data sets for driving cycles and vehicle types show considerable consis-
tency for similar driving cycles in terms of the sign of the sample correlation
coefficient for specific parameter pairs. For example, the revertants per ug extract
parameter demonstrates such sign consistency for all vehicle groups over the FTP,
HFET, and CFDS cycles when sample correlation coefficients with extract grams per
mile, fuel specific extract, % extract, and the residue to extract ratio are considered.
For other parameters, effects are observed which are vehicle group specific. A
consistency in the sign of the sample correlation coefficient is observed for the
General Motors group over the higher speed cycles for the revertant per ug residue vs.
residue gram per mi.e and fuel specific residue parameter pairs. Similarly, for the
Volkswagen group over the higher speed cycles, a consistent, weak, positive sample
correlation is observed for the revertant per mile, fuel specific revertants, and
revertant per ug part-culate vs. residue to extract ratio parameter pairs.
The lower speed cycles, NYCC and IDLE, frequently do not exhibit the same level
of significance as the higher speed cycles for specific vehicle groups.
One of the most notable observations is that revertants per ug extract and the
extract emission parameters are negatively correlated, albeit weakly, possibly indi-
cating that as greater quantities of extractibles are emitted by a vehicle the
proportion of bioactive organics diminishes. The bioactivity per ug extract sample
correlation coefficients with fuel specific residue and the residue to extract ratio
parameters are positive. Furthermore, the revertants per mile and fuel specific
revertant parameters tend to show a more general pattern of statistical significance
for correlation vs. residue gram per mile and fuel specific residue than vs. the extract
parameters. The question posed by these observations is that of the interaction of
extractible emission organics with inextractible carbonaceous and inorganic emission
matter which together make up the total emission particulate matter, and the
consequences of any such interaction with respect to the biological activity of the
extractible organics. It may be postulated that since revertants per yg particulate and
emission parameter sample correlation coefficients exhibit generally low levels of
68
-------
TABLE 6.4 BIOACTIVITY/EMISSION PARAMETER SAMPLF. CORRELATION
COEFFICIENT SIGNS AND SIGNIFICANCE LEVELS BY DRIVING CYCLE
AND VEHICLE GROUP OVER ALL TEST PHASES
Extract
grans
per mile'
Residue
grans
per mile*
Extract
grans
per kg
fuel
Residue
grams
per kg
fuel
t
Extract
Residue
to
Extract
Ratio
Cycle"
CM
vw
MB
Other
All
Cycle"
CM
MB
Other
All
Cycle"
VW
:IB
Other
All
Cycle**
CM
VW
MB
Other
All
Cycle*'
CM
VW
MB
Other
All
Cycle**
G.I
VW
MB
other
All
Revertants
per mile *
f 1! C N I
41 43
43
•1 43
r H c N i
+1
• 1 43
•».? 43 41 4_3
«1 43 p_3j
F K c N i
• i 4;
1 *3 41
r H c N i
«i 41
3 43 « 3
4] O
QGO^
1 O «3 43 £)
F H C N 1
43
-3 -1 43
-2 -1 43
-1 -3
r H c N :
-3
1 43 41
••2 42
Revertants
per kg fuel
F H C N I
41
r H c N i
+1
41 +2
+ 1 -»3 +3 *1
F H C N I
41
41 4J
F H C S 1
4] 42
«3 43 43
4} O
«i OOO
<2 430
r H c N i
-3 -1 41
-3 -2
-3 -1
F H C N I
41 43 41
43 43
41 -1
Revertants
per ug
extract
F H C N 1
-3 -3 -3 -3 -1
cid^3
-2 -3 -3 -1 -3
F H C N 1
41
41 $ 8*?
F H C N I
-3 -3 -3 -3 -1
F H C N T
42 41 41 @ -1
42 Q(JG +1
43 4343
F H C N I
-3 -3 -3 -3 -3
-3 -3 -3 -1
$ § £f O O
F H C N I
43 43 43 43 42
HH:23°
S89:!S
Kevertantc
per vg
residue
F H C i: 1
43
<1
43 42 42
F H C N I
-3 -3 -3
F K C N I
_.)
43 43 42 42
41 41
F H C N I
-2 -3 -3
-1
41
-1
-2 -3 -3
f H C N 1
41 41 43
41 43
41 41 42 43
43 43 43 43 43
F H C N I
-3 -1 -3
O-i
-1 -1
-1 -1 -1 -3 -3
Revertants
per ug
particular;
F H C K I
-3 -2 -1 4l
-1
I.i -1 -3 -1
F H C N 1
-1 -2 -3
41 -1
41 41 -1
t H C JV I
-3~-2~-2
-1
-1
-1 -1
F H C .; 1
-1 -2 -3 41
-1
41
42 41 41
43 42 -1
F H C N I
-1 41
-1 -1
-1 -1 43
F H C N J
-1
4l 41 41
41 41
-3
• Nurabers in table refer to the significance level (a level) of the
sample correlation coefficient for the given parameter pair and the
vehicle group and driving cycle indicated. The sign preceding the
significance level is the sign of the sample correlation coefficient.
The significance levels and corresponding a levels are :
SL-1 for 0.05 £ O > 0.01 i
for 0.01 £ 0 > 0.005 ; SL-3 for 0.005 £ a
•• Driving cycle code : F-FTP, H«HFET, C-CFDS, N=NYCC, 1-ldle
* For Idle cycle change units to per minute basis.
Circled values (eg. £3) ) indicate thrt tnu absolute value
of the sample correlation coefficient was greater than C.600.
L'r circled values indicate sample correlation coefficients
which, although statistically significant (i.e. non-zero),
had absolute values jess than 0.600. Correlations wore
frequently very weak ( |r|< 0.3 ! and complete listings of
values may be found in the Appendices.
69
-------
TABLE 6.J BIOACTW1TY/EMISS1ON PARAMETER PAIRS V1TH SAMPLE
CORRE1 ATION COEFFICIENT ABSOLUTE VALUES IN EXCESS OF 0.60
Bioactivity
Parameter
per mil*
Revertants
per kg fuel
per
Revertants
p«r
119 reviJue
Bmittion
Parameter
9 /mile
Residue
9/k9-fuel
Residue
9/k9-fuel
9 /mile
Residue
f /mile
Extract
9/kc-fuel
Residue
9/k9-iuel
% Extract
Residue
to
Extract
Ratio
Residue
to
Extract
Ratio
Drivin9 _ Vehicle
Cycle ~ Croup
__
HFET - Othe
CFDS - Othe
NYCC - Othe
Idie • Ml
HFET - Othe
CFDS - Othe
NYCC - Othe
HFET - MB
HFET Other
eras Other
CFDS KB
FTP KB
HFET «B
HFET Other
CFDS Other
NYCC Other
NYCC - VW
Idle - CM
FTP - MB
HFET - MB
BFET - Other
CFDS - Other
K?cc - other
Idle - Other
FTP - MB
HFET MB
FTP Other
HFE1 Other
CFDS Other
Idle Other
KfCC VW
NYCC - KB
n
33
3}
IB
67
33
32
ie
20
33
32
20
IB
20
33
32
IB
14
26
19
20
33
33
IB
33
19
20
32
33
32
33
14
7
r
' .609
.622
.638
.660
.654
.664
.631
-.611
.631
.629
.6;-J
-.647
-.636
.714
.693
.689
.612(SL'2I
.639
-.64B
-.659
-.654
-.60S
-.605
-.623
.719
• 627
. *05 ,',
.-07 •'
.731
.646
.665
-.72HSL.1I
Significance Level (SL) is 3
(a £. 0.005) unless otherwise indicated
statistical significance, there may be an interaction of extract and residue parameters
which determines bioactivity rather than one or the other independently determining
the bioactiviry. This hypothesis is by r.o means proven in these data; the question is
merely raised lor future research consideration.
In certain cases a sample correlation was noted within specific vehicle groups, but
the sign of the correlation changed for the "All" group (unsegregated data for all cars
without regard to vehicle type). This is most pronounced for the revertants per minute
and % extract parameters for IDLE. Explanation is made by considering two factors:
first, the correlations are very weak, and, second, the pairwise data points may fall
into approximately discrete regions. Thus within a vshicle xype, the limited data could
exhibit a weak positive association but the location of grouped data regions could show
an overall weak negative association as observed by the "All" group.
70
-------
Finally, Table 6.5 presents the strongest (in a relative sense) sample correlation
coefficients observed for the bioactivity-emission parameter pairs. The absolute
values of the correlation coefficients in Table 6.5 lie in the 0.60 to 0.73 range.
Two points are noted from Table 6.5; the "Other" vehicle group is more frequently
observed with relatively high sample correlation than are the specific vehicle groups,
and, while % extract and residue to extract ratio are inverse expressions for the same
physical parameter, the mathematical formulation of the ratio leads to a statistically
significant increase in the absolute value of the sample correlation coefficient for the
HFET and CFDS "Other" data and a marginally significant increase for the FTP
Mercedes-Benz data. One explanation for the higher sample correlation observed for
the "Other" group is that data points for Car //15, an Audi, tended to be variable but
generally higher in both emission and bioactivity parameter values than data for other
cars in this group. The higher correlation is thus chiefly due to the effect of this
vehicle.
Additional work with these data using non-linear correlation and multiple para-
meter correlation techniques is suggested.
Correlations of Ames Activity with Extract/Residue
Included in the project data base are many parameters which could correlate with
Ames activity changes. Many pair-wise correlation coefficients were calculated but
none was found to adequately explain the observed changes in Ames activity. Earlier
researchers have found that fuel system parameters strongly affect Ames results. An
EPA study (38) showed that injector malfunction can dramatically alter Ames activity.
Work conducted at Nissan Motor Co. Ltd. (39,
-------
CflR 2 OLDSMOB1LE
O I EXTRACT
A Rtv/uc EXTINCT x
CRR >i OLDSMOB1LE
O X EXTRACT
A REV/UC EXTRfCT X 8
•8
M
Is"
£
0 10 K 30
HJIEKK / 1000
SO 69
CRI? 9 VW
O X OtTRflCT
A REV/UC EXTRfCT
tt
iO tt 30 US
MJLEKZ / 1000
50 EO
CRR 15 RUDI
O X ETTRHCT
A KCV/UO CXTRflCT
5s
15 30 MS SO 75 93
HlLERCE / 1000
Figure 6J1 Mileage accumulation effects for extract and Ames activity,
revertants/ yg extract for (a) Car 112, (b) Car it it., (c) Car f>9, and (d) Car //15.
72
-------
Figures 6J2, 6J3 and 6.14 show the vehicle group-average data for residue,
extract, and revertants (per mile basis), each normalized to its FTP value, and plotted
by test cycle. In Figure 6.12 the Ames activity and residue for the General Motors
group are seen to have almost identical changes between cycles, except the NYCC.
The similarity between these normalized residue/Ames activity data, as compared to
the extract data, suggests the possibility that residue and not extract (at least extract
by itself) be considered a probable original source of the Ames active material that
ends up in the extract due to the extraction process.
Examination of Figures 6J3 and 6.14 does not show the same residue/revertant
trend for the Volkswagen and Mercedes-Benz groups. However, in these groups the
residue, extract and revertant data are all in general proximity, suggesting either
could provide a basis for correlation.
The comparatively low values of the NYCC Ames activity values for all three
groups in Figures 6.13 to 6.14 are unexplained. While both residue and extract are at
their maxima for the NYCC, the Ames activity for the NYCC is distinctly lower,
especially for the Volkswagen and Mercedes-Benz groups. This may be due to the
large amount of time spent at idle, changes in particulate character due to the low
average speed, stop-start driving, or other unknown factors.
73
-------
Off,
-'
O RESIDUE IC/HI)
A EXTRRCT (C/MII
+ REVERTflNTS OOVHI)
NTCC FTP CFDS HFET 50 C
Figure 6.12 Cycle variation of residue, extract and
revertants/mile normalized to the FTP lor the
General Motors group.
OCD
O RESIDUE IG/MI)
* EXTRflCT (G/MH
+ REVERTHNTS (lOVrlJ)
L -•
O RESIDUE (C/MJ)
» EXTRRCT IC/MI)
REVERTflffrS I10VMI)
NTCC FTP CFOS !FET SO C
NTCC FTP CFOS HFET SO C
Figure 6J3 Cycle variations of residue, extract and
revertants/mile normalized to the FTP for the
Volkswagen group.
Figure 6J* Cycle variations of residue, extract and
revertants/mile normalized to the FTP for the
Mercedes-Benz group.
-------
SECTION 7
SPECIAL TESTS
INTRODUCTION
This section discusses several ad hoc experiments conducted during this project to
investig^t" s>,:.t;c ^ situations related to interpreting project data. While the results
presented here may impact the emission results and their interpretation, they are
discussed separately within this section. The topics discussed include the effects of
driving cycle sequence on emission results, characterization of particulate adhered to
dilution tunnel wal'.s, the exposure of particulate and extract to filtered exhaust gas,
and effects of cold-ambient vehicle soak on emissions.
EFFECTS OF DRIVING CYCLE ORDER
The use history of a vehicle "immediately" prior to emission testing can have an
effect on test results. The Federal Test Procedure therefore requires that certain
procedures be followed in the half-day period before actual emission testing. In this
study vehicles have been operated in a variety of test cycles after completing the
three-bag FTP cycle. After completing 31 in-use vehicle tests the schedule of driving
cycles was altered. Concern was raised that data collected before and after this
"schedule change" would not be comparable beca ise the vehicle's driving history
immediately prior to a particular cycle would be different. Therefore an experiment
was designed to determine if the order in which cycles were driven affected measured
emissions. These experimental data can also be used to make emissions data
reproducibility estimates.
Experimental Design
The "schedule change" test was conducted with a 1979 OJdsmobile (Car ft5) with
dynamometer conditions of 12.5 hp at 50 mph and 4000 Ib I.W. and control fuel and lube
(Phase 3 conditions). The test ran for seven days and was preceded by a complete
normal vehicle test, and followed by an additional short special test. Thus data for
more than a two week period of vehicle testing were available. Table 7J shows the
days of testing and the order in which cycles were driven. Five FTP cycles were run
over a seven day period employing five different "pre-conditioning" schedules. Three
of these "pre-conditioning" schedules consisted of the same driving cycles driven in
different sequences. On three consecutive days an FTP Bag 3 (only) was run after the
vehicle was operated for about two hours on a repeated cycle. Particulate was
collected from all cycles; gases vere analyzed for all FTP bags and many ether cycles.
75
-------
1 ABLE 7.1 - SEQUENCE OF DRIVING CYCLES FOR SCHEDULE CHANGE EXPCRIMENT
August
7 t » io ii a
FTP
BAG I
BAG 2
BAG I
CFDS
NYCC
FTP
BAG 1
BAG 2
BAG )
JOC
JOC
JOC
JOC
JOC
JOC
FTP
BAG 1
BAG;
BAG )
IDLE
IDLE
IDLE
IDLE
IDLE
IDLE
FTP
RAG 1
BAG 2
BAG >
HFFT
MFET
HFtT
HFET
HFET
HFET
MFET
FTP
BAG '
BAG 2
BAG )
JOC
IDLE
HFET
FTP
BAo I
BAG 2
BAG)
SOAK SOAK
(JOCi«
HFET
HFET
HFET
SOAK
SOAK
(MC) JOC IDLE HFET (JCCI
HFET SOAr: SOAK SOAK HFET
MFET FTP 8/ 0 3 FTP BAG J FTP BAG J HFfcT
MFET IDLE HFtT JOC HFET
HFET JOC IDLE
SOAK SOAK SOAK SOAK
(JOC)
JOC-i
JOC-5
JOC-IO
JOC-5
JOC
(JOC- XI)
(JOC)
JOC JOC
vc Joe
JOC JOC
JOG-JO joe
JOC-M
JOC
JOT-JO
'OC-JO
• I )For Equilibration only-no , 5,
and 7 (first run) being distinctly higner thar: other values. Table 7.1 shows that these
cycles were driven either :mmediat^Iy af;er an IDLE or shortly after a long period of
iDLE'3. Note also in Table 7.2 th-»t the FTP Bag 3 driven after an IDLE end 1C minute
hot soak gave very high results. It is therefore evident that periods ol idling substan-
tially increase part'cuiate observed in subsequent drive.i cycles. This .Tiay be caused
by deposition of partirulate in the exhaust system, by a d-^r.ge in the fuel system or
some other unknown {actor. The August 7 r>'ata arc interesting in thst the HFET
driven immediately after tiie IDLE wes very high but the HFET driven in the afternoon
(following J5 minutes of 50C,' was normal. Project test schedules always performed
the iDLE as the last cycle of a phase, and th° first cycle of the next phase was always
preceded by 15 minutes of 50C bsiore taking data. Therefore, these results were
probably unaffected by this 1 OLE-effect phenomenon.
76
-------
TABLE 7.J- PARTICIPATE EMISSIONS FROM HFI-T, XC AND IDLE - DAY-TO-DAY AND REPETITION VARIATION
O.y
Filter
HFET-I
HFET-2
HFET-J
HFET-4
HFET-J
HFET-6
MFET-7
HFET-i
D»y
Filter
JOT.-l
»: 2
XK.-J
JOC-4
JOC-J
JOC-6
JOC-7
50C-8
Day
Filter
IDLE-1
IOLE-2
IDLF.-3
IDLE-4
IPLE-5
IDLE-6
IDLE-7
0.419
0.393
July
29
B
0.39*
30
A
3
A
*
A
Augmt
) 6
A A
7
A
7
B
0.426 0.392 O.M4 0.500 0.458 0.464 . •« 0.551
0.3S9
July
30
A
G.34J
July
30
A
0.3J5
0.341
0.34)
0.32/
0.329
f>.'26
O.ii)
4
A
4
B
0.3))
0.327
0.330
0.393
) 6
A A
0.3)) 0.3U
0.422
0.431
0.42)
0.427
0.430
0.41)
0.4)1
'ugust
7 10
A A
I/.362 0.360
0.45J
0.416
II
.-*-
0.35S
0.367
0.3)2
0.36)
0.347
0.363
0.3)2
0.3)4
0.42S
0.3SO
II
B
0.364
0.371
12
A
0.337
0.335
0.33)
0.333
0.17) 0.17)
0.171
0.)7)
O.I7«
0.173
0.174
O.IM
August
B
0.172
0.174
0.171
6
A
7
A
O.ISO 0.179
Cycle Repetition Effects—
The effects of cycle repetition were studied by repeating a given cycle six or
seven times over a period of about 2 hours. The only interruption between cycles was
the time necessary to change filters. The cycles tested were the HFET, 50C and
IDLE. Other cycles were not included because of the burden on drivers and vehicles.
The data appear as p?rt of Table 7.3 and are shown graphically in Figures 7.1 an 7,2.
046!
O VflkTIR
• '»' ru.Tt»
OI9l
Oi6
0 16
0-32
123456
RUN NUMBER
Figure ?J Variation of HFET and 50C paniculate
tmi'iions for conseojtive runs.
127-456
RUN NUMBER
Figure 7.2 Variation of IDLE paniculate emissions
lor onsecurve runs.
77
-------
The HFET shows a large decrease in participate after the first cycle followed by a
leveling off. Tests conducted after a lunch break yielded particulate levels similar to
the initial value. The initial and final values in Figure 7J are thought to be greater
than normal, possibly caused by the vehicle not being fully warmed up. I or the first
HFET, the preceeding FTP run of about 11 miles did not fully warm up the vehicle.
For the final HFET, the soak period during lunch cooled off the vehicle. A similar
trend appeared for the 50C as shown in the lower curves of Figure 7.1. The data given
in Table 7.it allow an assessment of the reproducibility of particulate measurements
taken at different times:
Table 7.* - Reproducibility of 50C Particulate, g/mi
Data From Average ±1 o CV,%
August 4* 0.33710.012 3.5
August 11 0.35810.008 2J
August 12 0.33410.002 0.6
All Dates 0.3*710.01* 3.9
*Excluding "cool-start" (after lunch) values.
The ratio of the August 11 average to the August 4 average is 1.062 and the
comparable ratio for the August 12 average is 0.991. The 18 measurements have a
range of 0.326 to 0.369 g/mi (average +6%).
Idle particulate emissions are shown in Figure 7.2. The entire range of values for
both fiiter probes is less than 0.003 g/minute. The absence of a driver for the IDLE
most likely accounts for this high reproduciL-ility. IDLE particulate measured on four
other days had a range of 0.1750 to 0.1802 g/minute.
Gaseous Emission
FTP hydrocarbon, carbon monoxide and nitrogen oxide emissions are given in
Table 7.5 and are shown graphically in Figure 7.3. Gaseous measurements were very
reproducible with coefficients of variation increasing in the order: NOX
-------
TABLE 7.5 - DAY-TO-DAY VARIATIONS OF FTP GASEOUS EMISSIONS
30
July
31
HC
g/tni
CO
g/mj
NO,
g/mi
BAG 1 C
BAG 2 C
BAG 3 (
FTP (
BAG 1
BAG 2
BAG 3
FTP
BAG I
BAG 2
BAG 3
FTP
1.89
1.61
).50
).64
.84
.80
.46
.71
.65
.94
.68
.81
August
4 _5_
O.SI 0.88
0.54 0.53
0.44 0.45
0.57 0.58
1.73
1.70
1.36
1.61
1.68
2.00
1.69
1.8)
.112
.76
.39
.67
.70
.95
.66
.82
6 7 !0
0.82 0.83 0.79
0.59 0.54 0.52
0.45 0.44 0.45
0.60 0.57 0.55
1.77
-
1.37
-
1.69
.
1.69
-
.82
.74
.36
.65
.64
.93
.65
.80
.74
.71
.37
.62
.70
.97
.70
.84
Mean
CV.*
0.84
0.56
0.46
0.58
.79
.74
.38
.65
.68
.96
.68
.82
0.04
0.04
0.02
0.03
0.05
0.04
0.04
0.04
0.03
0.03
0.02
0.02
4.8
6.5
5.0
5.4
2.6
2.3
2.8
2.4
1.5
1.4
1.2
I.I
Fuel Economy
20
1.8
16
1.4
1.2
1 UH
o>
O.8
0.6
0.4-1
Q2
MPG —
HC
293OJ11
JULY
^3456769 10
AUGUST
V
22.05
O
•21.08
UJ
u.
Figure 7.3 Variation of FTP gaseous emissions and
fuel economy on successive days.
ruel economy data are shown in Table 7.7. Two features of this data are readily
noticeable: (a) the fuel economy changes very little from day-to-day and (b) fuel
economy increases for consecutive repetitions of the same cycle. The data in
Table 7.7 show relatively minor changes in fuel economy from day to day with
coefficients of variation in the range 0.9 to 2.1%. On days when a cycle was repeated,
however, the fuel economy increased at a rate of about 5-6% per hour for the HFET
and 50C and 3% per hour for the IDLE. Since fuel economy was determined by the
carbon balance method, this increase was due mainly to decreased CO2 emissions as
the HC and CO emissions changed very little. This may indicate that several hours of
79
-------
driving are required before maximum fuel economy is achieved. Therefore cycles
driven at the end of a sequence may have better fuel economy than they would
otherwise.
DILUTION TUNNEL
The use of a dilution tunnel for diesel particulate sampling may have unforeseen
and non-predictible consequences. During the diesel testing program, it became
necessary to test a catalyst (non-project) vehicle with the 50 cm x 50 cm filter in
operation. The filter was found to collect an even light gray coating of particulate
plus large aggregates of dark particulate which were obviously of diesel origin,
TABLE 7.6 - DAY-TO-DAY VARIATIONS OF HFET, 50C AND IDLE GASEOUS EMISSIONS
?u!y August
30 3456667 Mean a CV.%
HFET
g/mi
HC
CO
NO,
0.30
0.94
1.33
0.32
0.92
1.38
0.31
0.95
1.39
0.30
0.96
1.37
0.29
0.94
1.35
0.28
0.92
!.28
0.32
0.96
1.36
0.30
0.91
1.35
0.02
0.02
0.04
4.9
1.8
2.8
50C
g/n.i
HC
CO
NO.
30
0.28
0.93
1.38
0.27
0.87
1.35
0.26
0.83
1.33
0.26
0.87
1.28
0.26
0.89
1.39
0.27 0.28
0.94 0.93
1.37 1.39
0.27
0.89
1.36
0.01
0.04
0.04
3.4
4.6
3.0
IDLE
g/min
HC
CO
NO,
30
0.18
0.41
0.18
0.17
0.43
0.18
0.15
0.42
0.17
0.11
0.36
0.17
0.11
0.35
0.17
0.18
0.47
O.J7
0.17
0.43
0.17
0.15
0.41
0.17
0.03
0.04
0.00
20.2
10.2
2.8
TABLE 7.7- FUEL ECONOMY VARIATIONS FOR DUPLICATE TEST, MPG1
BAG 1
BAG 2
BAG 3
FTP
BAG 3
ALONE -
50C
50C
50C
HFET
HFET
HFET
IDLE'
IDLE
IDLE
3uly
30
19.9
20.3
22.9
20.9
33.1
31.8
166.0
20.3
21.2
23.6
21.6
23.9
33.5
31.9
36.0
173.6
20.0
20.5
23.2
21.1
23.2
33.6
August
6
20.1
23.4
23.3
33.6
19.6
21.2
23.5
21.4
33.0
31.2 31.1 31.5
32.0
33.4
170.7 172.1 170.0
175.3
178.0
10
19.7
20.5
23.4
21.1
Mean
19.93
20.74
23.33
21.11
23.47
33.36
0.26
0.43
0.25
0.28
0.38
0.29
CV,%
1.30
2.06
1.07
1.31
1.61
0.86
31.40 0.32 1.01
170.48 2.86 1.68
'IDLE fuel economy in units of minutes idle per gallon.
80
-------
indicating the potential for re-entrainment of participate adhered to tunnel surfaces.
Disassembly of the tunnel and probes showed all surfaces to be completely covered
with a velvet black deposit of particulate approximately 1/16 in thick. Samples were
collected from three tunnel locations and the tunnel was chen thoroughly cleaned. A
sample from the bottom of the tunnel midsection was analyzed for Ames activity and
the results are shown in Table 7.8 identified as sample //i.
TABLE 7.8 - AMES ACTIVITY OF DILUTION TUNNEL PARTICULATE
rev %
Strain S9 ue ext ext
1 TA98 + 21 10.8 2
1 TA100 - 50 10.8 5
1 TA100 + 22 10.8 2
4 TA98 - 17 20.0 3
5 TA98 - 25 16.0 H
6 TA98 - 27* 25.3 7*
10 TA98 - 24 12.1 3
11 TA98 - 200** 2.7 6**
*Slope difficult to determine - approximate value.
**A11 doses gave responses of 2100 rev/plate or greater.
About nine months later the tunnel was again disassembled and eight samples
were collected at various points in the tunnel and probes as identified in Figure 7A.
Portions of five of these samples were extracted with dichloromethane in the same
manner as regular particulate filters. These extracts were analyzed for Ames activity
and the results appear in Table 7.8 as Samples //4, 5, 6, 10 and 11. All activities were
unusually high, particularly when expressed as revertants/ug extract.
DILUTION
AIR
EXHAUST
Figure 7.4 Locations of particulate samples removed from the dilution tunnel.
The Ames activities were considerably more uniform when expressed in
particulate mass basis. The extractive (%) and the specific activity combined in an
inverse fashion to give this result. This finding also suggests that the direct-acting
mutagens present may be more closely related in origin to the residue than to the
extract portion of the particulate.
81
-------
The physical character of the participate changed noticeably moving from the
tailpipe collecting tube (Sample //ll) to the inlet of the CVS (Sample //9). The
paniculate deposited in the connecting tube (which was used primarily for
Volkswagens) was composed of small, compacted particles which adhered tightly to the
tube and had very little extractible matter. Progressing down the tunnel from the
area where dilution air enters and the mixture cools, the particulate grows progres-
sively lighter (lower bulk density) and has higher percent extractibles. The sample
from the inlet to the CVS contained much material of low density which became
airborne when disrupted.
Some of these differences are probably due to temperature and velocity changes
occurring within the system. It is also reasonable to assume that erosion of old
particulate from the wall occurs concurrently with deposition of fresh particulate. In
addition, volatile material in the deposited particulate would tend to move toward the
cooler end of the tunnel. This is possibly indicated by. the increase in % extractibles
noted at the cooler end of the tunnel.
EXPOSURE OF PARTICULATE AND EXTRACT TO DILUTE FILTERED EXHAUST
Introduction
There is a concern in the diesel emission field, about the effects (if any) on
particulate collected on a filter from exposure to the gareous components of the
exhaust. Two experiments were conducted to investigate the effects of exposure to
dilute exhaust gases on mutagenecity by the Ames TA9S(-) assay. In the first
experiment, particulate was collected from a 50C driving cycle for varied times and
then re-exposed to filtered dilute exhaust gas for varied times. In the second
experiment a composite sample of dichloromethane extract was prepared and applied
to. clean filters, which were then exposed to filtered dilute exhaust gas. Collection of
particulate for the composite extract sample for this second experiment entailed
repeated vehicle runs throughout a single day, and these individual particulate samples
provided an opportunity to determine the variability in specific mutagenesis activity
from successive test cycles. These results, while peripheral to the main objective of
the study, provide a backdrop against which to consider the results of the exposure
tests and are discussed in Section 6.
Vehicle Test Conditions
All tests were performed at a steady 50 mph cruise mode, except that the
acceleration and deceleration portions to and from the 50 cruise condition could not be
excluded from the particulate collection due to sampling equipment constraints.
Dynamometer conditions were as given in Table 3.1.
Exposure of Particulate to Filtered Exhaust
This portion of the test program was performed with Car #5. Particulate was
collected during a 5CC for 5, .10, 15, and 30 min duration. After completion of
particulate collection, the filter was retained in its holder and covered by two unused
filters and replaced in the particulate sampling apparatus. A second 50C was then run ,
using this package of three filters. The first filter now collected particulate while the
middle filter acted as a back-up to avoid further collection of particulate on the last
(original) filter. Table 7.9 shows the time of particulate collection, time of added
82
-------
exhaust exposure, particulate, extract, and Ames results.
TABLE 7.9. EXPOSURE OF PARTICULATE TO DILUTE EXHAUST
Particulate Additional Part Part Extract Ames Activity TA98(-)
Collection Exhaust 47mm 502 % g/mi REV REV REV
Min Exposure g/mi g/mi pgE ygP mi
min m '
5 0 0 0.40 0.58 30.7 0.122 4.5 1.4 5.4
5 5 12.2 0.35 0.41 34.3 0.120 3.9 1.3 4.7
5 10 24.1 0.38 0.48 34.2 0.130 3.8 1.3 5.0
5 15 35.7 0.38 0.43 34.7 0.134 3.7 1.3 5.0
10 0 0 0.35 0.40 3.59 0.127 3.4 1.2 4.4
15 0 0 0.36 0.37 39.9 0.144 2.5 1.0 3.6
15 0 0 0.36 0.39 37.1 0.133 2.4 0.9 3.2
15 0 0 0.36 0.37 36.5 0.133 2.3 0.8 3.0
15 15 34.3 0.37 0.38 39.3 0.145 3.0 1.2 4.4
15 30 66.6 0.35 0.36 39.3 0.138 3.1 1.2 4.3
30 0 0 0.37 0.35 40. 0.147 2.5 1.0 3.6
30 0 0 0.35 0.35 36.8 O.UO 2.4 0.9 3.1
30 15 31.0 0.35 0.34 37.9 0.132 4.3 1.6 5.7
30 30 59.5 0.35 0.34 37.8 0.133 2.8 1.0 3.7
In a 50C the 50x50 cm filter must be operated during the acceleration to 50 mph
and the deceleration from 50 mph. Therefore, the filter collects particulate from the
acceleration and deceleration modes as well as from the 50 mph cruise. This is
apparent in Table 7.9 where the 50x50 cm particulate values (g/mi) are significantly
higher than the <*7 mm particulate values (which do not include the acceleration/-
deceleration modes) for the 5 min collection times. The effect diminishes for longer
times.
For none of the three particulate collection times was there a significant or
consistent change in the extract itself or in its Ames activity. For the 5 min
particulate collection case there was a consistent decrease in the activity with
exposure. This difference was not large enough to be considered significant. This
difference, as well as the lower % extract for 5 min filters, may be attributed to the
relatively large effect exerted on these parameters by the particulate collected during
the acceleration/deceleration associated with this cycle. The high particulate and
Ames activity and low % extractive of the filter used to collect particulate for 5 min
with no subsequent exposure may indicate that the vehicle was not completely
equilibrated when this sample was collected, since effects of this type have been
observed in several cold-start/hot-start comparison experiments reported previously
(37, t2).
The 15 and 30 min particulate collection cases show small but non-significant
increases in Ames activity which did not relate to exhaust exposure time. Both the
mass and percentage of extract show no significant changes as a result of re-exposure
to dilute exhaust. Changes in extract can be attributed to the previously mentioned
characteristic of the driving cycle.
Gaseous measurements were made on three cycles during the two days that the
test was run. These results are given in Table 7.10.
83
-------
TABLE 7.10. EMISSIONS DATA FROM PARTICULATE EXPOSURE TESTS
PARAMETER MEAN ±1 a
HC 0.28+0.01 g/mi
CO 0.91+0.01 g/mi
NOX 1.29+0.03 g/mi
CO2 296±6 g/mi
Fuel Economy 34.1+0.7mpg
The consistency of the gaseous data above and the 47 mm participate data in
Table 7JO indicate that particulate collection and subsequent exhaust exposure were
performed under consistent dilution tunnel conditions.
Conclusions
The experimental results do not support the hypothesis that the gaseous com-
ponents of diluted exhaust affected the mutagenic activity of the extract of
particulate collected on Pallflex T60A20 media under the conditions of the
experiment.
Exposure of Extract Alone to Filtered Exhaust
The extractible material bound to the carbonaceous portion of diesel particulate
may be only partially available for reaction with gas phase components while in
particulate form. To test for potential reactions between the extract and dilute
exhaust, a series of experiments were performed wherein extract obtained from the
consecutive tests described above was re-deposited with a solvent carrier on a clean
filter. The solvent was dried, leaving the extract laden filter for subsequent exposure
to dilute filtered exhaust gas. Approximately 100 mg aliquots of extract were applied
to the central portion of a 50x50 cm filter over an area of about 1250 cm2. As in the
particulate exposure tests, these extract carrying filters were covered with two blank
filters to isolate the extract filter while exposing the extract to dilute exhaust gas and
collecting a corresponding particulate sample. Results for Car //5 and Car //I are
presented separately below.
Car #5 Results—
The three filters containing approximately 100 mg of extract were exposed to
dilute filtered exhaust for 5, 15, and 30 min respectively. Normal particulate samples
were also collected for 5, 15, and 30 min, and a 15 min particulate sample was also
exposed to exhaust for 15 min. The results are shown in Table 7.11.
The three filters with applied extract had Ames responses greater than the three
corresponding particulate filters which were exposed with them. The Ames activities
of the extracts were, however, lower than the 3.7 revertants/ ug average activity of
the extracts used to produce the composite extract. Analysis is complicated by the
lack of a zero time exhaust exposure filter with applied extract. Thus, it is not
possible to determine what the activity of the extract was immediately prior to
exposure or if the activity of the applied extracts was affected by re-extraction.
It is clear that filters 1664, 1663, 1662 do not show a dose-response relationship
which would indicate an exhaust exposure effect. Any effect, however, may be
84
-------
TABLE 7.11. EXPOSURE OF EXTRACT TO DILUTE EXHAUST GAS - CAR 05
Extract Paniculate Paniculate Exhaust exposure Extractives* Ames
Applied Collected Collection rnin m* mg % rev/pg
Filter mg mg Time(min)
1660 101 5 16.1 93.9 93.0 2.5
1663 107 1} 39.3 102.2 96.0 I.!
1662 99 30 76.5 95.2 96.2 3.0
Pooled extract used to prepare filters 1660, 1663, 1662 3.7
1666 0 I3& O* SO 131.9 10.9 2.4
1671 0 227 50 37 .* 1.6
1670 0 530 15 0 10.\ 1.3
1667 0 1129 30 0 36.7 • 1.5
1668 0 5S2 15 15 39.1 37.5 3.6
a. A blank filter will contribue about 2 mg of extract itself. Therefore, extract weights and percentages are higher than actual.
b. Filter 1666 was run for 50 minutes as a backup filter.
complete by the 5 min exposure time. Filter 1668, which had 15 min exhaust exposure
to 15 min of collected particulate, shows enhanced activity as compared to filters 1671,
1670 and 1667 which would seem to indicate that exhaust exposure increases Ames
activity. This result does not agree with data shown in Table 7.9.
It is of further interest that filter 1666, which was run for 50 minutes as a backup
filter and showed a 13 mg weight increase, had an Ames response greater than that of
the extract from the filter which preceeded it in the gas stream. If one multiplies the
Ames activity (2A revertants/ pg) by the extract weight (1.09x10^ pg) the result is
total revertant count attributable to the filter. For filter 1666 this value is 2.6x10^
revertants. If a filter contained 100 mg (JO5 pg) extract, the increase in Ames activity
from particulate passing through the primary filter and collecting on the extract-
exposure filter would be less than 2.6xloVlO* = 0.3 rev/pg. This indicates that the
filter exposed to filtered exhaust did not change appreciably due to additional material
passing through the primary filters.
The data do not indicate that filtered diluted exhaust gas has a substantial effect
on the Ames activity of extracted material. The data do not, however, prove that
there is no effect. Defects in the experimental design and the small data set do not
allow the experiment to give a conclusive answer.
Car //I Results-
Similar to the Oldsmobile experiments, filters with approximately 100 mg applied
extract were exposed to filtered dilute exhaust for 5, 10, and 20 min respectively.
Normal particulate filters of 10,15 and 20 min length were also collected. The results
are shown in Table 7.12.
The results for filters 1653, 1652 and 1651 indicate a decrease in Ames activity
with increased exhaust exposure. Although the changes may appear significantly
Jarge, past experience with this vehicle has indicated that differences of this
magnitude occur in the Ames analysis of triplicate runs. Therefore, the three data
values alone are not sufficient to substantiate an exhaust exposure effect.
Furthermore, all of the Ames values determined in the experiment were significantly
lower than the values measured for the extracts comprising the composite (Ave +lo =
2
-------
TABLE 7.12. EXPOSURE OF EXTRACT TO DILUTE EXHAUST - CAR 01
Filter
16)3
1652
16)1
Extract
Applied
mg
99
98
99
Paniculate
Collected
mg
0
0
0
Particulate
Collected
Time(min)
0
0
0
Pooled extract used to prepare filters 16)3, 16)2, 16)1.
1660
1661
1658
16)9
16)6
16)7
16)0
0
0
0
0
0
0
81
7
10
601
1009
1271
699
0
0
10
I)
20
10
Exhaust erposure*
Time (min) m'
10
20
10
I)
0
0
0
10
15.6
26.)
16.2
22.8
38.)
22.0
Extractives'*
mg %
90.8
92.1
92.6
6.0
6.9
93.8
126.2
186.)
108.6
79.1
9).S
9«.0
93.)
15.6
12.)
l«.7
D.6
97.7
Ames
rev/ug
10.0
8.4
6.9
20.8
1.7
1.6
7.1
7.3
8.3
8.4
9.3
a. Volume corrected to 20°C.
b. A blank filter will contribute abut 2 mg of extract itself. Therefore, extract weights and percentage are higner than actua!.
c. Room air (filtered) drawn through filter overnight at about 20°C.
Ames results with consecutive tests showing large variations in response. It also
appears that when this vehicle is operated under 50C conditions, the Ames activity of
filters collected consecutively shows a significant increase. This might contribute to
the trend observed with filters 1653, 1652 and 1651, as these filters were run in
numerical order.
Filters 1658, 1659 and 1656 represent increased paniculate collection times. The
slight increase in Ames activity with increased collection time is not considered
significant. The significant fact is that the average activity of these extracts is 7.6
rev/Mg vs. 24.8 rev/ug for filters collected a week earlier. This difference must be
predominantly real as the reproducibility of the Ames test is much better than the
difference would suggest. Filter 1657 represents 10 min of exhaust exposure to a filter
with 10 min of particulate collection. The change in Ames activity is not significant.
Filter 1660 was run as a blank, i.e., covered by two filters and exposed as were the
filters with extract. Filter 1661 was run as a back-up to filter 1659. Both filters
showed a small increase in weight which was mainly extractible. The extracts showed
Ames results of 1.7 and 1.6 rev/ug respectively. Tr.sse values are much lower than
those of primary filters and indicate that material of higher Ames activity is
associated with particles collected on the primary filter. This is the reverse of the
results for the Oldsmobile tests in which the primary filters had lower activities than
the backups.
Although an apparent dose-response decrease in Ames activity for increase
exposure to dilute filtered exhaust has been observed, the inherent Ames activity
variability of extract from this vehicle makes the result very uncertain.
Discussion
The results do not constitute a definitive experiment on the reactivity of extract,
on or off particulate. Variations in Ames mutagenicity arising from vehicle para-
meters that are only poorly understood complicate attempts to isolate any interactions
between dilute exhaust gases and extract. If results for a single experiment were all
86
-------
that one had, a dramatic effect could be reported. When a battery of tests are
considered, the picture is not always consistent. The fact that major increases in
Ames mutagencity were not generally observed with increased particulate collection
time, particulate re-exposure, or extract re-exposure suggests that the particulate
sample conditions are not the main cause of the observed activities. The several cases
of observed decreases in Ames activities with sample exposure is deserving of further
investigation.
An additional consideration in t!i° ctudy of sampling effects on mutagenic activity
is the presence of dilution tunnel cebris. We have assayed extract from tunnel
scrapings, and found, like ethers, that this material exhibits elevated activity - about
30 rev/ug in our case. Thus we have a sampling system that is coated with material
of considerably higher activity than the samples we wish to measure. If this debris re-
entrains (arid it does; the activity will be increased. For instance, a filter with extract
exhibiting 4.3 rev/ug started with a total particulate mass of 1.2!3g. Assuming a
tunnel debris activity of 30 rev/ug, and an activity of 2.5 rev/ug for particulate
uncontaminated by tunnel debris, a contaminated sample of 4.3 rev/ ug would require
the following amount of debris:
4.3 = 2.5 (1.213-Tunnel Particulate) + 30 'Tunnel Particulate)
Tunnel Particulate = 0.05 g
Thus only 50 mg of tunnel debris could raise the activity from 2.5 to 4.3 rev/ ug for a
total sample size of over 1200 mg particulate. The exhaust tubing connecting the
vehicle to the dilution tunnel contains debris of significantly higher activity than the
tunnel itself, although the exlractible content from this source is much lower. Thus a
significantly smaller debris contribution from this source would cause the same change
in observed specific activity.
Thus, mutagen sampling artifacts can occur by means other than exhaust gas
exposure during the time scale of particulate collection. Elevated specific activities
of tunnel and connector pipe debris suggest a long-term change in extract character
not readily seen in short-term collection tests.
EXPOSURE OF PARTICULATE AND EXTRACT TO SUNLIGHT
In the late summer of 1981, three 50 x 50 particulate filters from Car //5 were
collected from successive 15 min 50C tests. One filter (//1561) was kept at room
conditions for two days (no light) and extracted. A second filter (#1560) was left in the
filter holder. The third filter (//1562) was extracted immediately after collection and
the extract alone redeposited on a fresh 50 x 50 filter as previously described. The
particulate filter and extract laden filter were then set outside in bright sunlight for
portions of two days (8 hours total exposure). Following this time, all three filters
were extracted and assayed with tester strain TA98(-). The dose/response plots for
these samples, and those from a comparable 50C sampie from an in-use test on this
vehicle near the same time are shown for two dose ranges in Figure 7.5. The specific
activities of these samples are fairly low, and this is consistent with the in-use test
results for this vehicle.
Examination of the three plots for samples 1560, 1561, and 1562, shows a response
below 25 ug dose. For the sample of extract exposed to sunlight (1562) this initial
response is followed by a general lack of consistent response at higher doses. This
87
-------
200
loo
100
100
*• SOC fro* nornal in-uu test, 11517
B» particulate exposed to ambient, 11560
O paniculate kept indoor•, I1E61
extract, aabienL exposure*
1 1S62
.^#^'
••****
e • * *
.-•«••'
"1"
V*
£
.D..
20 30
EXTRACT DOSt ug
A- SOC from normal ln-iu* test, 11517
Inarticulate exposed to anbient, tl560
C"T»articulat* kept indoors, I1S61
D*redttpoBitad vttract« «Kbi«nt exposure.
11562
b
100
T
s
100
!
* .••"""
..-• ...» e
. ».•* .••"' ..-•
.'...:i-- c..--- c
«.-• ^'' » ..••c
i-:=:?:;;:!-V-.c---"" . ..--» •
/.«• .-•* c •••*.. •
' ..* 8 u
2^ so ?s ibo i2s no
err PACT DOSE uq
..••••
• »•" ^
B
175 JC
Figure 7.5 Dose-response curves for exposure of particulate and extract to sunlight:
(a) 0-50 fjg dose range, (b) 0-200 fjg dose range
88
-------
may be due to the loss of direct acting mutagens or the formation of toxic agents
during exposure. The participate exposed to sunlight exhibits an extract dose/response
plot not dissimilar to the comparison particulate samples unexposed to sunlight. The
slightly higher extract response for the sunlight exposed particulate may be accounted
for by the slight decrease in extractible % for the sunlight exposed sample.
Although these results from a single experiment must be viewed cautiously, they
do indicate the relati/e sunlight exposure changes between particle-bound and
extracted direct-acting mutagens ior this one test condition.
COLD AMBIENT PARTICULATE TESTS
Mid-project review of the Ames mutigenicity data from various vehicle test
cycles showed the FTP to be frequently somewhat elevated in comparison to other
cycles. Since the FTP was the only test cycle with an overnight vehicle soak as part
of the test, this aspect of vehicle conditioning was considered a possible factor. A
series of particulate emission, extract and bioassay tests were performed in the winter
of 1981 on Car //1(VW Rabbit) to evaluate the effect of vehicle soak temperatures.
In project FTP testing particulate from the three FTP bags was collected on a
single filter. Bag I and Bag HI have an identical 505 second driving scheduie, Bag I
beginning the FTP after overnight laboratory vehicle soak, and Bag III commencing
after a 10 min hct-soak subsequent to Bag II. For the special tests reported in this
section the Bag I (or Bag III) driving schedule was used *o collect separate partiojlate
samples ior only the appropriate 505 second test, in contrast to a normal FTP wnich
provided only a composite particulate sample from Bags I, II, and III. Two additional
samples were obtained by soaking the vehicle overnight at winter ambient (outdoor)
conditions and then testing at laboratory conditions to obtain cold-ambient soak Bag I
and Bag 111 samples. Thus the four vehicle conditions tested were: DBag I after
overnight ambient cold soak; ?'Bag III hot start following condition (1); 3)Bag I after
normal overnight vehicle soak in laboratory; . 4)Bag III following condition (3).
Continuous temperature monitoring was set up for a>overnight ambient, b)vehicle
injection fuel line, and c)vehicle crankcase lubricating oil.
Table 7J3 gives the results of duojicate tests at each of the four conditions. The
test result ior condition (it) above, normal hot-start (Bag 111), was used as the base
condition for comparison results. The emission parameters listed are all given as
ratios to the base condition results. The mean temperature vilues given result from
fairly large variations in temperature between the beginning and end of the particular
test, and thus should be considered 35 o ~.!y rough guides to the temperature hictories
encountered.
Examination of Table 7.13 shows the following for successively lower tempera-
tures (lower rows in the Table 7.13): Dtotal particulate increased by 12-74% over the
base, 2)extractible expressed as a % of total particulate dropped by 13-36%, 3)extract
as a mass emission showed little change _+13%, 4)residue accounted for the majority of
increase in total particulate, showing an increase from 24-92% over the base
condition, 5Hhe various bioassay measures show increased bioactivity ranging from
230% to 400%, depending on computation choice. Braddock (42) in contrast found the
bioactivity to decrease v/ith decreasing temperature.
The increase in residue corresponds more closely to the increases in bioactivity
89
-------
TABLE 7.13. VW COLD-START PARTICIPATE AND AMES ACTIVITY COMPARISONS*
Vehicle Test Condition
« : Base Condiiion-Normal
FTP Bdg III Hoi Start
2 = Bag III Hot Start
following cold ambient
lest
} : Normal FTP Bag I utter
overnight Ub soak test
1 = FTP Bag I alter over.
night ambient soak
Mean Temperature.- °C
Overnight Injector Crankcase Paniculate Extract Residue Revertanls - TA98I-)
Soak Fuel Line Lube (g/ini) (*) (g/mi) (g/ini) Rev/p Eilract Kev/pgParl. IO'Kev/mi
20
20
0
2J
IS
21
5
90
9»
01
0.3» 2».S 0.081 0.26 5.8 1.4* ».9
Results below arc ratioed values to the base condition in each column.
1.0
I.IS
1.74
0.87 0.87
0.79 0.93
0.64 1.11
1.03
1.24
1.92
1.5
I.S
3.6
1.3
1.1
2.3
1.3
1.7
4.0
•Average results for two duplicate tests at each condition.
than any other emissions parameter measured. The injector fuel line temperature
corresponds to the observed emission changes more closely than other temperature
measurements. These results suggest the need for a definitive experiment with
controlled variation in fuel temperature and corresponding measurement of
particulate, extract, residue, and bioactivity. If fuel temperature does effect the
combustion/formation of trace bioactive species, the unknown variation in injector
fuel line temperature during laboratory vehicle testing could account for some of the
apparently random bioactivity results, since the fuel temperature could be affected by
fuel recirculation to the fuel tank, and fuel volume in the tank. The correspondence of
residue increase to the increased bioactivity is similar to findings reported elsewhere
in this report connecting residue and not extract to the bioactivity of extract samples.
90
-------
SECTION 8
CHEMICAL CHARACTERIZATION OF EXTRACTS
INTRODUCTION
The objective of this work was to isolate, identify, and quantify the chemical
substances responsible for mutagenic activity of the organic extracts of diesel
emission participates.
Diesel participates are highly resyiiable (43) and may constitute a significant
inhalation health hazard to the human population. The chemical composition of the
organic extractible matter of the particulates is extremely complex. When this work
was initiated in 1979 there was very little published information about the identity of
possible mutagens in diesel particulate extracts. ; Polynuclear aromatic hydrocarbons
(PAH) were the major class of organic carcinogens/mutagens whose presence had been
established (44-46) in extract samples from diesel particulates. Both parent PAHs and
their alkyl derivatives were detected in these studies. Subsequent studies utilizing the
Ames Salmonella mutagenesis assay and other short-term bioassay methods have
indicated that diesel particulate extracts possess significant mutagenic activity even
in the absence of mammalian enzymes (47, 48), suggesting mutagenic contributions
from compounds other than PAHs, which require metabolic activation to exhibit
mutagenicity.
The approach adopted for this study was to integrate the Ames microbial
mutagenesis assay as a biological monitor with chemical fractionation, identify the
mutagenic fractions for detailed structural characterization of the constituents by
several complementary techniques, and assess the mutagenic contribution of the PAHs
present in the extract samples.
This section is a brief account of the significant parts of the work carried out
during 1979-3une 1981. The report was not written in a descriptive manner, since more
detailed accounts of the results obtained are summarized in publications that have
originated from this work (References 48 to 52).
RESULTS AND DISCUSSION
Particulate Sample Collection
The project was initiated by generating a few large project-reference samples for
characterization and bioassay work to proceed in parallel with subsequent in-use
vehicle testing. The three vehicles used to generate these samples were a Mercedes-
Benz 300-D obtained from EPA in Ann Arbor, a aiesel Rabbit (Car //I), and a diesel
Oldsmobile (Car #5). Particulate samples for this characterization/bioassay effort
91
-------
were collected by the sample equipment and general procedures as described earlier.
The daily testing sequence was not the same as used for the in-use vehicle study, and
involved only two driving cycles. An FTP began each day of sample collection during
which one 50 cm x 50 cm filter was collected. With the FTP as a warm-up the rest of
the testing day consisted of 2*f HFET cycles broken into 8 groups of three HFET cycles
per single 50 cm x 50 cm filter. One day of testing thus generated one FTP and eight
HFET 50 cm x 50 cm filters.
Two fuels were used for sample collection from the Mercedes 300-D. A 750 liter
lot of EPA control fuel from EPA/RTP was used first to generate filters to be shipped
to EPA/RTP for subsequent extraction and analysis. All other large particulate
samples in these three vehicles were obtained using a 7500 liter lot of project
reference fuel in underground storage at AEL. At the beginning of each large sample
collection each vehicle was serviced with fresh oil and oil filter. Oil samples were
collected from the vehicle at the conclusion of the sample collection.
Gaseous emissions were measured for one FTP and one HFET approximately every
fifth day of sample collection. Particulate emissions were likewise measured by the
k7 mm probe filters described earlier. The individual 50 cm x 50 cm filters were
extracted by the same procedure as previously described and the extracts from all
filters from each cycle pooled to yield one composite sample. The FTP and HFET
filters were treated separately to provide a comparison between vehicle test cycles.
Table 8.1 contains the summary information on these large samples to the point of
their bioassay and chemical characterization analyses. Table 8.1 also identifies each
of these samples by codes which reference these samples in the chemical analysis
portions of this report. In total these samples represent over 13,000 dynamometer
miles performed on 50 days of laboratory testing and using in excess of 450, 50 cm x
50 cm filters.
Vehicle
TABLE 8.1. SUMMARY OF LARGE PAR-.CULATE SAMPLES FOR CHEMICAL AND BIOASSAY CHARACTERIZATION
Emissions (g/mi)
Comments
Fuel
Oil
Sample
Identi-
fication
.Mercedes-Benz EPA Castrol
300-D Reference 20VV-50
SE-CC
Mercedes-Benz AEL Castrol
300-D Reference 20K-50
SE-CC
Rabbit
Oldsmobile
AEL Castrol
Reference 20W-50
SE-CC
AEL Castrol
Reference 10-30
SE-CC
Large Particulate Sample
Partic-
Driving ulate Extract % HC CO NOX Part MPG
Cycle (g) (g) Extract
S2
L2
FTP
HFET
23.1
2J5.00
2.1
19.49°
L3
L4
L5
SI
L7
S3
FTP
HFET
FTP
HFET
FTP
HFET
5.9
53.2
4.3
53.76
11.52
115.1
0.54
6.1
0.98
15.58
1.99
27.5
9.1 0.22 1.11 1.14 0.63 21.7
8.5° 0.12 0.82 1.52 0.39 28.0
9.2 0.22 1.08 1.88 .64 21.8
11.5 0.11 0.79 1.57 0.36 30.0
22.8 0.37 0.93 0.95 0.32 41.0
29.0 0.18 0.74 0.90 0.28 S1.9
17.3 0.48 1.60 1.83 0.79 21.1
23.9 0.23 0.88 1.44 0.31 31.3
Vehicle shipped from
EPA, Ann Arbor
Filters shipped to
EPA/RTP for extraction
Car HI
Car 15
Sample also used for
project bioassay sld.
8 Only 229.4Sg of paniculate were extracted.
b Average of eleven extractions of 15 filters (s 20g) each, a - 0.74.
92
-------
Fractionation of Particulate Extracts
The extracts were fractionated by a two-step procedure. The first step was a
liquid-liquid partitioning by which acidic, basic and neutral compounds were separated
as shown in Figure 8.1a. The acidic fraction showed the highest specific activity
profile of the three fractions, as shown in Figure 8.2 and is discussed in Reference 48.
The neutral fraction comprised the largest fraction (90-95%) by extract mass for all
three composite samples studied. This neutral fraction was subfractionated by silica
gel preparative adsorption chromatography described in Reference 48 and Figure 8.1b.
O.CI,
PM 10-11
'V'l
100
80
-01,11,-C,«»-
11,0
4 3 • r
"i 2 -
I
In!
1
•
D-S9
J
200 *OO 6OO BOO
ELUTION VOLUMEw>
Figure 8.1 (A) Fractionation of diesel emission
paniculate extracts. (B) Elutant profile for chro-
matographic Jractionation of neutral diesel particu-
late extract. (48)
Figure 8.2 Mutagenicity of diesel particulate
extract sample S-l and its fractions (Car ffl). (<»8)
Several changes were subsequently introduced to the procedure. The samples
were introduced to the top of the column coated on prewashed and dried sea sand. The
column was sequentially eluted with solvents of increasing polarity as shown in
Figure 8.3. All fractions were concentrated in Kuderna-Danish evaporators to about
1 ml and transferred to tared test tubes. The remaining solvent was evaporated on an
"Evapo-mix" under gentle vacuum at 30°C and weighed. In previous work, we used a
small amount of silica gel to introduce the sample but this resulted in poor recovery of
certain compounds. Use of sand improved recovery of these compounds. The elution
volume of the PAH fraction, designated as subtraction C2, was determined by the
elution volumes of fluor .ie and coronene under identical chromatographic conditions.
93
-------
Neutral Fraction
Sample on Sand
Si 1 ic.i Ool Column
2flxlcm
I'l silica C.cl 60
Hexane. 70 ml
I SIIRKHACTION
Cl
5t C1I3C12 in h<-x,inc, 200 m
jKinii
C2
11 8» CHjCMj in hex.inc, RO ml
2) 10* CII,C1, in hox.inp, BO pil | KimrpAtTloHJ
C.I
3) 20» C1I2C12 in hcxanc, 80 ml
S0% CHjClj in hexano, 80 ml
ISIIIII-PACTIONI
" -J
I" """I
rtljCl^. RO ml
SMIirRAITIlilJ
50X Ether in CH2Cl2. <0 "1
Molh.inol. 40 ml
::"nrRA(Ti'>ij
Figure 8.3 Silica gel fractionation of neutral diesel
particulate extract.
16-
14
12
10
>
v. 6
u. 0
85 8
J
D-S9
D+S9
fil
_rll
I 234567
SUBFRACTION
Figure 8.4 Distribution o( mutagenicity among
sample S-l subtractions of neutral diesel particulate
extract of Car HI. (48)
-------
Final elution with methanol was essential to obtain the last traces of some polar
components, not all of which eluted completely with ether.
All subtractions (C1-C7) were bioassayed with the tester strains TA98 and TA100
at six or more dose levels. Mutagenicity distributions among subfractions of the
neutral fraction of sample SI are shown in Figure 8.4. The trend in mutagenic potency
of the subtractions did not change as a result of modifications in subfractionaticn
procedure. The PAH subfraction C2 showed only marginally positive activity in the
presence of S-9 for all three composite samples. Subfraction C4 had the highest
specific activity for all three composite' samples. The subfraction C5 also had
considerably high mutagenic activity. Subfractions C3, C6, and C7 showed some
mutagenic activity. The chemical characterization work emp'.asized subfractions C2
(PAH subfraction), C3, C4, and C5.
Characterization of the PAH subfraction C2: PAHs are a well-recognized family
Of carcinogens and mutagens that require metabolic activation for expression of
mutagenicity. Many members of this family have long been considered responsible for
carcinogenicity of combustion effluents. It was shown in this work that PAHs (parent
and alkyl substituted) have only marginal contribution to the mutagenicity of diesel
particulate extract samples.
The major problem in PAH analysis is separation and conclusive identification of
individual isomeric compounds, since carcinogenic properties of PAHs are isomer-
specific. Many PAH reference standards are not available, making conclusive identifi-
cation of the isomers especially difficult. In the present work, glass capillary gas
chromatography, alone and in conjunction with mass spectrome^ry, and high per-
formance liquid chromatography coupled with rapid scanning ultraviolet spectrometery
were utilized for PAH identification in diesel emission particulate extract samples.
Much of the work on sample SI has been described in References 49, 50 and 51.
Gas chromatographic method using an SE-54 coated glass capillary column gave
excellent separation of many isomeric PAHs, e.g., benzo(a)anthracene and chrysene,
benzo(b)fluoranthene and benzo(k)fluoranthene, and benzo(e)pyrene and
benzo(a)pyrene. Separation of these groups of PAHs is important since some are
moderate to strong carcinogens, whereas others are relatively innocuous. A chromato-
gram of 21 PAHs is shown in Figure 8.5. A gas chromatographic profile of the PAH
subfraction S1-C2, typical of the PAH subfractions from diesel extract samples, is
shown in Figure 8.6A. The major peaks identifiable by comparing their retention times
with standard PAHs are phenanthrene, anthracene, fluoranthene and pyrene. Minor
components could not be identified with confidence from their retention time alone.
Gas chromatography/mass spectrometry of the same sample enabled characterization
of over forty PAHs as tabulated in Table 8.2. Total ion chromatograms (GC/MS) of
this and another sample are given in Figures 8.7 and 8.8.
An HPLC method was developed to preconcentrate the lexicologically significant,
higher-molecular weight PAHs, which were present in low concentration only. A gas
chromatographic profile of an HPLC fraction enriched with low molecular weight
PAHs by this method is shown in Figure 8.6b. Additional confirmation for identity of
the major PAHs was obtained by measuring a complete UV spectrum of each HPLC
separated PAH. Presence of anthracene, phenanthrene, fluoranthene, and pyrene were
unequivocally confirmed this way. The spectra and HPLC chromatograms are shown in
Figures 8.9 and 8.10. The constituents of the HPLC fraction of S1-C2 were also
confirmed; the HPLC profile and a number of UV spectra are given in Figures 8.1 la
95
-------
468 910
T
It 12
v_
IS 1517
\
10
20
30
MINUTES
Figure 8.5 Cos chromatogram of a mixture of PAH standards. CC conditions: 40-m X 0.35-mm
id SE54 glass capillary column: flame ionizalion detector: temperature /00°C for 2 mm. programmed
10 I70'C at lO'/min. to 209° at j'/mm. to 26i°C at 7"/mm, to 290*C at 8"/mm. Peak iden-
tities: I, fluorene (3 ng); 2. phenanthrene (3 ng); 3. anthracene (6 ngl; 4. 2-methylphenanthrrne (6
ng); S, 2-methylamhracene (6 ng): 6. 3.6-dimethylphenanthrene (6 ngl: 7. fluoranthene 13.6 ngl: 8.
pyrene f6 ng); 9. benzofalfluorene 112 ng); 10. benzolbjftuorene n
885
911
91'
9*5
980-
1032
lOtl
1057
106k
1077
lllk
11<3
117'
1181.
13*5
1353
Iks*
17JI
1775
Cat pound
Cp-Alliylphenanthrene/-anthraeene
C?-Allt/lphenanthrene/-anthracene
riuorvnthene
Acephenanthrylene/aceanthrylene
Pyrene
Cj-Alkylphenanthrene/ -anthracene
Cj-Alkylphenanthrene/-anthraeene
C^-Alkylpheuanthrene/-anthracene
Methyl pyrene/-fluoranthene
Beoiofa_)fluorene
Benxo[b]rluorene
Methylpyrene/-fluoranthene
Methylpyrene/>fluoranthene
BeniotgM If luoranthent
Acepyreoe
Chryaene. b-nxola_]anthracene( trlphenylene
tenlolb.^.tklriuoralithene
Bentola^lpyrene. benxo|^)pyrene
96
-------
5!£2
10
20
JO
4O
SIC2C
I '
1
22
30
40
MINUTES
Figure 8.6 Gas chromaiogram of A. PAH fraction of diesel paniculate extract ISI-
C2) and a. its HPLC subtraction C (SI-C2). GC conditions 45-n x 0.35-nm id
SES4 glass capillary column; flame ionitation detector: temperature. 110-C tor 2
min. programmed to 170~C at I0°/min. to 2l2iC at 3°/mm. to 278°C at 8'fmin.
Peak identities: I, phenanllirene; 2. anthracene; 3-6, meihylanlhracene/-phenan-
Ifirene; 7. 2-phenylnaplnl:alene; 8-10. dimelhylanlhracene/-phenanihrene; 11, flu-
oranthene; 12. aceantlir)lene/acephenonthrylene; 13, pyrene; 14-15, trimethylan-
thracene'-phenanthrene; 16, bemolghijftuoranlhene; 17, benzojajanthracene; 18,
triphenylene; 19, thryiene; 20, bcnio'tb)fluoranllienc; 21. bcniol\]fluoranlhene;
22, benzolklfluoranlhene;23, beniofelpyrene; 24, ben;o[a]pyrene. (50)
97
-------
OO.O
EJ ••>«
tCM 600 600 1000 I2OO MOO ItOO COO 1OOO
T.-r 1000 IJM »:« IOOO ZJ» M 40 SO 00 «:20
SCAN 400 6OO BOO IOOO I2OO MOO 1600 1600 20OO
T1«6.4O OOO IMO «:*, ZOOO Z>*0 t».«> 3O-OO W2O
Figure 8.7 Total ion chroinatogram oi the PAH
fraction SI-C2. (50)
Figure 8.8 Total ion chromatogram of the PAH
fraction S2-Ci. («8)
"
Figure 8.9 HPLC profile of PAH fraction of extract
sample S-l from Car ttl. (49)
HPLC condition:
-------
and 8.lib.
to
MINUTtS
B
jiOO
Minn*
Figure 8.11 (A) HPLC profile of high-molecular-
weight PAH fraction of sample S-l extract of
Car //I. (B) UV spectra of HPLC eluates. (51)
Column: HC-ODS (25 cm x 2.6 mm id); mobile
phase, 60% MeCN-H2O for 5 min, linear gradient to
100% MeCN in iO min.; flow rate, i.O ml/min.
PAH subtractions of diesel particulate extract samples (S2 and S3) from two other
vehicles gave similar profiles. Characterization was performed by capillary GC,
capillary GC/MS and by HPLC/UV spectrometry.
Characterization of Mutagenic Subtractions
Subtraction C4, representing 2.3-3.8% by mass of the neutral fraction, had in
genera) the highest specific mutagenic activity, accounting for ^2-52% of the direct-
99
-------
activity mutagenicity of the neutral fractions. Subtraction C5 comprised 3.7-^.8% of
the neutral fraction by mass and had the second highest specific activity, accounting
for 13-20% of the mutagenicity of the neutral fractions. The GC/MS characterization
of subtractions C4 and C5 of sample S\ are discussed, and details are described in
Reference 52.
Bioassay results clearly indicated the presence of direct-acting mutagens in these
subtractions, although promutagens were presumably present as well. The chromato-
graphic isolation procedure suggested that these fractions contained compounds more
polar than parent and alkyl substituted PAHs. The chemical manipulations involved in
the preparation of the subtractions also suggested that the mutagens were stable
chemicals and not reactive alkylating agents.
Glass capillary GC with flame ionization detection demonstrated the enormous
complexity of these samples. Since possible identities of the constituents were
unknown, chromatographic conditions could not be optimized, and structural informa-
tion in the constituents of the subtractions were thus obtained by capillary gas
chromatography-mass spectrometry sample examination under El and CI conditions.
Thorough searches,pi the reconstructed ion chromatograms were made to identify the
mass spectra of the maximum number of compounds possible. Interference-free
spectra were obtained by using computer-assisted techniques.
The total ion chromatogram of subtraction C4 is shown in Figure 8.12. A number
of alkyl fluorenones were identified in this subtraction. These included isomeric
methyl fluorenones, C2-alkylfluorenones, C3-alkylfluorenones, and C/^-alkyl-
fluorenones. Mass spectral characteristics of these compounds are discussed in
Reference 52. While other structural possibilities cannot be totally discounted for
these compounds, evidence strongly indicates that most of these compounds are indeed
alkylfluorenones. Additional evidence for these structural assignments have since
been obtained by HPLC/UV spectrometry. Benzo(a)fluorenone was also a major
constituent of this fraction. A list of compounds characterized in this fraction is
given in Table 8.3.
Figure 8.13 shows a total ion chromatogram of subtraction C5 of sample SI. Some
important constituents of this fraction are <*H-cyclopenta(def)phenanthrene-fr-one,
several isomers of 6h[-benzo(cd)-pyrene-6-one, methylanthraquinone/phenanthrene-
quinone, anthracene-and phenanthrene-9-carboxaldehyde, 9-fluorenone, and 7H-
benz(de)anthracene-7-one. A complete list of compounds identified in this subfraction
by GC?MS is given in Table 8.*f. Mass spectral characteristics in support of the
structural assignments are also discussed in Reference 52.
Recently the possible presence of mutagenic nitro-PAHs, particularly nitropyrene,
has received a great deal of attention. We were not able to detect any nitropyrene in
subtractions C
-------
TOTAL
SCAN 200
TIME 6:40
300
10=00
400
13:20
600
20=00
700
23:20
Figure 8.12 Total ion chromatogram (El) of subtraction b. (52)
100 -
TOTAL
SCAN 200 300 400 500 600 700 800 900
TIME 6--40 10:00 13=20 16=40 20:00 23=20 26=40 30=00
Figure 8-13 Total ion chromatogram (El) of sample S-l subtraction 5. (52)
101
-------
TABLE 8.3^ COMPOUNDS CHARACTERIZED BY
GC/MS IN SUBFRACTION <* (52)
compd
C,-»lkyl-9 fluorenon*
C,-«lkyl-9 Huorrnonr
C,-»lkyl-9-fluor*none
4W-ryclop*nti[d.r/)ph*ninihren-4-on*
C,-tlkyl-9-fluorenonf
TABLE8.lt- COMPOUNDS CHARACTERIZED BY
GC/MS IN SUBFRACT10N 5 (53)
acan
no.
308
338
363
368
394
400
415
422
431
403
438
440
452
473
475
488
492
497
El
parent
peak
194
208
208
208
222
204
222
222
222
212
206
218
236
226
236
226
236
220
C,-alkyi-9-fluor*non*
9W-thiox§nthen-9-one
»nlhnc*n*/pht»nanthr*n*-9-
c&rboxaldehyde
hydroxypyrencffluoranthrne
C.-*Ikyl-9-nuorenone
j
C4-»lkyl-9-nuor*none
unidentified, hydroxy - m/t 210
C«-aJkyl-9-fiuortnon*
m*thyl»nthrtc«ne/phen»nthr»ne-
539' 230 b«nto| a} Ruorenone
S94 342 unidentified
El
par-
•can ent
no. peak
245 170
276 180
264 164
326 194
332 194
365 196
370 208
400 204
436 206
441 206
421 222
431 222
439 222
4M 218
462 216
474 236
492 236
Ml 220
carboxatdehyde
549 25P unidentified (an oxy-PAH)
553 230 b«nz|GV)inthracenone/
benzof)uor*none
567 230 b*nz(df ]*nlhr»cenone/
henzofluorenone
604 230 7//-benz[d>]anthracen-7-one
633 278 dibutyl phthalate
64£ 258 benz|o)anthracenedione
663 244 hydroxychrysene/benz(ojanthracene/
tnphenylene
710 254 brnzo[rd)pyrenone
717 254 benzolrdjpyrenone
754 254 bcnzojrdjpyrenone
rompd
methylnaphlhaldehyde
9-flu ore none
Cj-a'kylnaphthfcldehyde
methyl-9-f.uorenone
ahthrone/phenanthrone
9//-xanthen-9-one
•nthraqumone
4//-cyclopenta|de/)phenanthren-4-one
anlhracene/phenanthrene-9-
rirboxaJdehydr
•nlhrarene/phenanihrtne-9-
ctrboxaidehydr
C,-alkyl 9-fluor*none
methylanthraqumor.e/phenanthrenequinone
methylanthraqu'none/phenanthrenequinone
hydroxypyrene/fluoranthene
Kydroxypyrmr/fluorarithene
C,-alkyl-9-fluorenone
C,-alkylanthraquinone
Most compounds detected in the mutagenic subtractions C3, C*t, and C5 of the
diesel paniculate extracts were oxy-PAH derivations. These compounds could con-
ceivably be formed by oxidation of the corresponding parent PAHs during sample
collection. Nitro-PAHs could be formed by reaction of nitrogen oxides with parent
PAHs on the filter. Some evidence to this effect has been presented recently (53, 54).
However, it appears likely that at least some oxy-PAHs are formed prior to collection
on the filters. Since a wide variety of PAHs are frequently detected in airborne
particulates collected usually by somewhat analogous method, it is less likely that the
PAHs adsorbed to the diesel particulates will be extensively oxidized during the
relatively short sample collection period. Carefully designed studies need to be
performed to determine if any of the mutagenicity of diesei particulate extract
samples is an artifact of sample collection procedure.
During the course of this investigation and since its completion in June 1981,
several papers on the isolation of mutagenic fractions of di°sel particulate extracts
and characterization of the constitutents present in theue fractions have been
published (53-58). Schuetzle, e_t. ah, (55) have characterized several PAHs and oxy-
PAHs in two diesel particulate extract samples by low and high resolution GC/MS.
This group has also characterized several nitro-PAHs including 1-nitropyrene in four .
extract samples by GC/MS, high resolution MS and MS/MS technique.
The mutagenicity of many oxy-PAH compounds detected in the mutagenic
fractions of diesel particulate extract samples by us and others, is undetermined at
this time. Several nitropyrenes have been shown to be direct-acting mutagens as
determined by the Ames Salmonella assay. However, there is no evidence from this
study at this time to suggest that mutagenicity of the diesel particulate extracts is
102
-------
primarily due to the nitro-PAHs. Various oxy-PAHs are likely to make a considerable
contribution to the mutagenicity of the diesel particulate extracts. A few benzo-
pyrene ketones have been reported to be direct-acting mutagens (59) as are oxidation
products of pyrene and some hydroxy-PAHs(60). Additional information regarding
mutagenicity of polycyclic carbonyl compounds is needed to assess possible mutagenic
contributions of such compounds. Research should be directed to such studies, as well
as to investigations on confirmed identification of constituents in mutagenic fractions
of different extract samples from vehicles under real-world driving conditions.
In the time since the completion of chemical analysis work by Dr. Choudhury on
this project, 10 in-use vehicle particuiate extract samples have been examined by the
chemical research group of Karasek and Sweetman, e_t. ah of the University of
Waterloo (Ontario). A report on their work characterizing the PAHs in these 10
samples selected from the in-use vehicle portion of this project has been submitted to
the Ontario Ministry of the Environment (61).
Samples were analyzed by GC and GC/MS after HPLC fractionation. 29 PAH's
were quantified. 23 ketone, quinone, carboxaldehyde and nitro-PAH compounds were
identified in the moderately polar fraction. Benzo(c)cinnoline and methyl-benzo(c)
cinnoline were identified in the polar fractions. The few most abundant PAH's
appeared to account for a significant portion of the mutagenicity (with activation).
The highest nitro-pyrene concentration coincided with the greatest direct-acting
mutagenicity but was not high enough to explain the activity. Benz(cd)pyrenone was
tentatively identified in another sample with high direct-acting mutagenicity.
103
-------
SECTION 9
REAL TIME PARTICIPATE MASS MONITORING
A portion of this project was directed toward the first time use of a tapered
element oscillating microbalance (TEOM) for real-time diesel participate mass
monitoring. The full results of this application have been reported previously (62), and
this section will only summarize the instrumentation and study results.
The TEOM is a hollow glass rod, fixed at a wide base, with a removable filter
element attached to the narrow top, and oscillating in an electric field (Figure 9.1).
The TEOM has been shown to behave as a harmonic oscillator (63) with a frequency
dependence upon the mass collected by the filter element. If the TEOM is oscillating
at frequency fa at time - ta and between time = ta and time = t& collects a quantity of
mass, dm, the frequency of oscillation will be reduced to frequency f& at time = tt>.
The functional relationship is given by:
dm = K
- f
-2
where K is a determinable calibration constant.
SCE VE*
TOP viivv
ES-^0-j©[-5
/BT-*- P«070 T*A*J
SAMPLE FLO*
TEOf.1 C^E^A
1 E'ect-fC '*'d is se* UD DC'^een f»e'd p'o*es
2 imcae c< topered e'er^en? .$ c^ojecTca c^ c^o'o"Q^ss'or
3 OsCit'3*'Ori of 6l6;rv€riT tliTioTeS t^CTCO!ty 0" rT^C^Q^CC''* i
01 ** C vC'Toge OJ'CJ* tr&r\ D^'O'C-'fO^S S'D'
^ £C vC^'O^e iS 0?TiCi'fted Ontf CCS-^5 *C CC*CjC'-ve 03"" C"
5 Ffec-e~cy o* ostii!3f-on 3"-o ^e'-ce "^osscx" fii'ef e'e^c^-* s oe*e—-~-*;
t)> f'CUJtnCy C0un**r
Figure 9.1. Schematic representation of 7EOK instrurer.taticn.
-------
Thus the TEOM is shown to be a mass monitor. The TEOM, as tested (Figure 9.2),
could respond to dilution tunnel concentrations as low as 1 to 2 mg/m^ with a response
time on the order of 8 to 15 seconds. Increased sensitivities could be achieved by
longer response times or increasing the sample flow rate (2 to 2.5 1/min in this study).
Longer response time compromises the objective of rr.' ximizing the real time
capabilities of the instrument, while increasing the sample flow rate may introduce
more signal noise. In this work, the instrument flow rate was actually reduced from
the 51/min manufacturer design value as the loading of the filter during a driving
cycle could become excessive. The resultant pressure drop across the filter could
cause it to collapse, dramatically changing K in the mass frequency equation, thus
invalidating a measurement.
95->m 00 SS fUBE
Figure 9.2. TEOM evaluation experimental design.
The TEOM signal was digitally filtered using a moving linear regression algorithm
described in more detail elsewhere (62). As a consequence of this necessary signal
filtering, the instrument, at the state of development for this study, was determined
to have minimum response time frames on the order of 8 to 15 seconds.
Study vehicles (Car //I and Car #5) were driven over the FTP Bag 3 and NYCC
driving schedules using both standard ^7 mm filter collection and TEOM monitoring for
a total of 29 mass emission tests which are reported in Table 9J. This table compares
the total mass results determined by: (1) the TEOM using the fundamental mass-
frequency equation; (2) a gravimetric determination of the mass collected by the
TEOM filter; and (3) gravimetric determination of the standard procedure kl mm filter
mass. In the mean, the gravimetric determinations were within 10% of the TEOM mass
determination; however, some large deviations were observed (tests 8 and 22). The
TEOM filter is not designed for gravimetric analysis, and removal for weighing can
result in the dislodging of particulate matter. Thus, filter handling may be the cause
of extreme differences between the frequency and gravimetric determined masses.
105
-------
Other errors may have been encountered as a result of the adsorption of water on the
filter and paniculate and differences in filter media. The TEOM filters were Ballston
type CQ or CH material and the 47 mm filters were Pailf lex type T60A20. The TEOM
sample flow rate was 2 to 2.5 1/min while the 47 mm flow rate was 9 1/min. The
47 mm gravimetric values in Table 9J were adjusted to account for the difference in
flows.
TABLE 9.1 COMPARISON OF TOTAL MASS EMISSION DETERMINATIONS BY TEOM AND
STANDARD 47mm FILTER GRAVIMETRIC METHODS FOR PARTICIPATES FROM DIESEL
VEHICLES
Rur. •
1
2
3
t
S
t
t
e
9
ID
11
12
13
14
15
ie
17
16
19
20
21
22
23
24
2S
26
J7
28
29
1
TEQK
KiCTCproeeueff
Outpjt
«J
.505
.357
.36*
.384
.423
.354
.305
.363
.333
.327
.332
.303
.291
.219
.269
.309
.330
.311
.281
.343
.266
.297
.274
.253
.365
.430
.395
.576
.340
2
TOOK
Grr/u»tric
Filter
Elrsnt
nq
.S39
.390
.J74
.414
.«"6
.387
.289
.216
.296
.284
.264
.326
.259
.255
.261
.272
.311
.327
.274
.353
.293
.206
.273
.227
.347
.375
.333
.523
.346
3
Mlutisn
Tur.-»l
47 rr,
Filur
«>•
.400
.399
.393
.419
.475
.366
.356
.416
.336
.325
.335
.327
.326
.324
.341
.321
.319
.3C2
.300
.356
.327
.324
.34)
.342
.346
.360
.341
.542
.392
Mean
C.V. (%)
IKVIO
1
J
0.94
0.90
0.99
0.93
0.93
0.91
.06
.62
.13
.15
.17
.93
.12
.09
.03
.14
1.10
0.95
1.03
0.97
0.91
1.43
l.OD
1.11
1.11
1.15
1.19
1.10
0.96
1.07
151
1
J
1.26
O.CE
0.95
0.62
1.14
0.96
0.65
0.92
0.9S
i.c:
C.99
0.83
c.ts
c.ee
0.7!.
c.9e
1.C3
1.C3
0.94
0.96
i.f.
0.92
c.e:
0.74
1.11
1.13
1.16
1.06
C.f
C.96
131
collected ness, bag)*
innadittely after run. .
aled
. flow, weights drterauwd
Figures 9.3-9.5 present the real time data obtained for Cars //I and //5 over the
NYCC apd FT? Bag 3 driving schedules. The CVS flows were 9 m3/min for Car //I and
14 m'/min for Car //5. Therefore, although the TEOM mass accumulation for each car
over the FTP Bag 3 schedule was nearly the same, the mass emission from Car //5 was
greater than for Car //i (Figures 9.4 and 9.5).
Figures 9.6 and 9.7 present a three run test of driving cycle real time mass
emission rate reproducibility using the TEOM. Variation in the actual vehicle
emissions and the driver ability to reproduce a driving schedule were factors in
addition to instrument variability which affect the overall reproducibility presented.
The TEOM frequency signal was digitally filtered using a 15 sec moving (one
second units) linear regression algorithm. Applying the same filtering to the vehicle
one second acceleration data for comparison to the TEOM mass rate data emphasized
106
-------
VEHICLE SPEED VS TIME
Ch. 100. TOO. 300, MOO. 500. COO.
ENGINE RPM VS TIME
JOO. 200. 300. MOO. 500. 600.
MRSS RRTE VS TIME
0. JOO. 2C&. 300. UDO. 500. EOO.
RCCUMULR7ED MRSS
JOO.
300. 300.
1J* ISECOCS)
tiOO. 500. ECO.
Figure 9.3. Vehicle speed, engine rpm, TEOM mass rate (ug/sec),
and total TEOM accumulated mass vs. time for
Oldsmobile 5.7 liter diesel driven over the NYCC
schedule.
107
-------
o
00
»
1
go.
K
I*
VEHICLE SPEED VS TIME
100. JOC. 300. MOO.
| ENGINE RPM VS TIME
"b. . 100. 700. 300. HOC.
MHSS RflTE VS TIME
0. 100. 700. 300. UDO.
RCCUMULRTED MflSS
100.
700. 300.
TIHC litCONDSl
»OC.
MO.
(00.
SOC.
(00.
sac.
(oc.
SOD.
(00.
si-
VEHICLE SPEED VS TJME
ItC. 70C. 300. HX. SOO.
HO.
ENGINE RPM VS TIME
"fc. IOC. 700. 38C.
* MRS5 (5RTE VS TIME
SOC. (00.
0. IOC. 700. 300. HOG.
fiCCUMULRTED MR5S
SOB. (00.
100.
?M. 300.
UnC I5CCVOS)
HOC.
SOC.
tOO.
Figure 9.4. Vehicle speed, engine rpm, TEOM mass rate (jg/sec),
and total TEOM accumulated mass vs. time for
Oldsmobile 5.7 liter diesel driven over the FTP
Bag 3 schedule.
Figure 9.5. Vehicle speed, engine rpm, TEOM mass rate (ug/see),
and total TEOM accumulated mass vs. time for
Volkswagen Rabbit 1.5 liter diesel driven over
the FTP Bag 3 schedule.
-------
>
10
s
0. 300. 200. 300.
1JME (SECONDS)
1400.
SOD.
EOO.
Figure 9.6. Reproducibility of TEOM mass rate (yg/sec) for
three consecutive trials of the NYCC schedule
by Car #5 (Oldsmobile 5.7 liter diesel).
V. 1*3
CC
3 DO. 200. 300.
TIME ISECC1NDS)
VJDO.
SOD.
BOD.
Figure 9.7. Reproducibility of TEOM mass accumulation (10 grams)
for three consecutive trials of the NYCC schedule by
Car #5 (Oldsmobile 5.7 liter diesel).
109
-------
the strong relationship between vehicle acceleration and mass rate as shown in
Figure 9.8. The negative mass rates which appear to be associated with periods of
deceleration, are of particular interest. Although no experimental data concerning the
source of this observation were obtcdned, it was hypothesized that the effect was
caused by the desorption of water from the filter element and collected particulate
matter during the drier and clearer exhaust conditions of deceleration. It is
considered probable that this same phenomenon also occurs in the standard particulate
collection procedures but is unobserved as the gravimetric filter method is not a real
time technique.
33.
100.
113.
I6J.
XD.
233.
267.
300. 333.
TIC ISKONDSI
367.
MOO.
133.
HO.
SOD.
Figure 9.8 TEOM mass rate (ug/sec) and vehicle acceleration (mph/sec) for Car #5 (Oldsmobile 5.7
liter diesel) over the FTP Bag 3 schedule. (TEOM mass rate computed by least squares regression
smoothing of TEOM frequencies using 15 second smoothing window; vehicle acceleration obtained by
identical smoothing algorithm applied to one second average vehicle speeds.)
In a final test of the instrument, the inertial load of a vehicle was varied for three
trials of the FTP Bag 3 schedule. The real time results in Figures 9.9 and 9.10 show
that mass rate peak values were greater for higher inertial loading but that the total
accumulation was signficantly greater only for inertial loading in excess of the 4000 Ib
inertial loading specified by the vehicle manufacturer.
This short study to evaluate the TEOM as a real time monitor for diesel
particulate mass emissions concluded that the instrument, as tested, was capable of
determining mass rates over time frames as short as eight seconds arid total accumula-
tions in good agreement, in genera], with standard gravimetric filter procedures. The
TEOM was considered to give reproducible results to the extent that TEOM results
were within the range of variation considered reasonable for repetitive vehicle tests.
110
-------
MO
g-
SbOOfc,
SO.
ISO.
200.
fa. 300. SbO.
lint ISCUMOSI
HOD.
»so.
Figure 9.9 TEOM mass rate (ug/sec) vs. time at three dynamometer inertial weight settings for
Car *5 (Oldsmobile 5.7 liter diesel) driven over the FTP Bag 3 schedule. (The manufacturer's
specified inertial setting was 4000 Ib.)
5500 Ib
SO.
100.
ISO.
200.
2so. inn.
I1IC (SECONDS!
3so.
MOO.
MSO.
SOO.
550.
Figure 9.10 Accumulated TEOM mass (10~^ grams) vs. time at three dynamometer inertial weight
settings for Car US (Oldsmobile S.T liter diesel) driven ovor the FTP Bag 3 schedule. (The
manufacturer's specified inertial setting was 4000 Ib.)
Ill
-------
REFERENCES
1. D.R. Cahill, State of New York, Department of Motor Vehicles, persona) com-
munication to Richard Gibbs, April 27, 1982.
2. Society of Automotive Engineers, Inc., "The Measurement and Control of Diesel
Particulate Emissions," SAE Progress in Technology Series, \7_i 1979-
3. Society of Automotive Engineers, Inc., "Diesel Combustion and Emissions," SAE
Proceedings, 86, 1980.
ft. Society of Automotive Engineers, Inc. "New Diesel Engines, Combustion, and
Emissions Research in Japan," SAE Special Publications, 468, 1980.
5. Society of Automotive Engineers, Inc., "Diesel Combustion and Emissions, Part
II," SAE Special Publications. fr8fr, 1981.
6. Society of Automotive Engineers, Inc., "Diesel Combustion and Emissions, Part
III," SAE Special Publications, 495, 1981.
7. Society of Automotive Engineers, Inc., "Fuel & Combustion Effects on Particulate
Emissions," SAE Special Publications, 502, 1981.
- . • '
8. Society of Automotive Engineers, Inc., "Diesel Engine Combustion, Emissions and
Particulate," SAE Proceedings, 107, 1982.
9. Society of Automotive Engineers, Inc., "Diesel Combustion and Emissions
Research Report," SAE Special Publications, 525, 1982.
10. Society of Automotive Engineers, Inc., "Passenger Car Diesels," SAE Progress in
Technology Series, 24, 1982.
11. \V. Lipkea, J. Johnson, and C. Vuk, "The Physical and Chemical Character of
Diesel Particulate Emissions - Measurement Techniques and Fundamental Con-
siderations," SAE Special Publications, fr30, 1978.
12. C. Hare, "Methodology for Determining Fuel Effects on Diesel Particulate
Emissions," Environmental Protection Agency Report, EPA-650/2-75-056, March
1975.
13. J.M. Perez, F.3. Hills, D. Schuetzle and R.L. Williams (Eds.), "Informational
Report on the Measurement and Characterization of Diesel Exhaust Emissions,"
Coordinating Research Council Report, 517, December 1980.
112
-------
1*. Coordinating Research Council, "Informational Report: 1979 Progress of the
Chemical Characterization Panel of the Composition of Diesel Exhaust Project
and Results of Participate Extraction Round-Robin," Coordinating Research
Council Report, 316, March 1980.
15. F. Black and L. Doberstein, "Filter Media for Collecting Diesel Participate
Matter," Environmental Protection Agency Report, EPA-600/2-81-071, May 1981.
16. 3. Santodonato, D. Basu and P. Howard, "Health Effects Associated with Diesel
Exhaust Emissions: Literature Review and Evaluation," Environmental Protection
Agency Report, EPA-600/1-78-063, November 1978.
17. R. McClellan (Ed.), "Diesel Exhaust Emissions Toxicology Program - Status
Report," Lovelace Biomedical & Environmental Research Institute Report, LMF-
73, January 1980.
18. R. McClellan (Ed.), "Diesel Exhaust Emissions Toxicology Program, Status Report,
July 1980," Lovelace Biomedical & Environmental Research Institute Report,
LMF-82, July 1980.
19. R.G. Cuddihy et. aJL, "Potential Health and Environmental Effects of Diesel Light
Duty Vehicles," Lovelace Biomedical & Environmental Research Institute Report,
LMF-82, October 1980.
20. R.G. Cuddihy et. ah, "Potential Health and Environmental Effects of Light Duty
Diesel Vehicle II," Lovelace Biomedical & Environmental Research Institute
Report, LMF-89, October 1981.
21. W.E. Pepelko, R.M. Danner and N.A. Clarke, "Health Effects of Diesel Engine
Emissions, Proceedings of an International Symposium," U.S. Environmental
Protection Agency Reports, EPA-600/9-80-057a and EPA-600/9-80-057b,
November 1980.
22. A.M. Shefner, "Respiratory Carcinogenicity of Diesel Fuel Emissions," U.S.
Environmental Protection Agency Report, EPA-600/l-81-05f, July 1981.
23. J. Harris, "Potential Risk of Lung Cancer from Diesel Engine Emissions," National
Academy Press, Washington, D.C., 1981.
2*. National Research Council, "Health Effects of Exposure to Diesel Exhaust,"
National Academy Press, Washington, D.C., 1981.
25. W. Hogan, "Decision Analaysis of Regulatory Diesel Cars," National Academy
Press, 1982.
26, R. McDonald, "Diesel Car Regulation and Traffic Casualties," National Academy
Press, 1982.
27. National Research Council, "Diesel Cars: Benefits, Risks and Public Policy,"
National Academy Press, 1982.
113
-------
28. K. Hergenrother, "The Performance of Diesel Taxicabs in New York City fdxicab
Fleets," SAE Technical Paper Series. 780629, June 1978.
29. (a) J. Mariano, The City of New York, Department of Environmental Protection,
personal communication, March 24, 1982; (b) 3. Mariano and A. DeFilippis,
"Emission Inventory Projects," The City of New York Intradepartmental
Memorandum to 3. Sontov.-ski, April 22, 1975; (c) A. DeFilippis, The City of New
York, Department of Environmental Protection, personal communication to
Marcia Wi:i:-.n, EPA Ann Arbor, April 11,1975.
30. (a) J.Somers, et. al.f "Automotive Sulfate Emissions-A Baseline Study," SAE
Technical Paper Series 770166, February 1977; (b) T. Austin, K. Hellman,
C. Pa-Jlsell, SAE Technical Paper Series, 740592, August 1974; (c) C. Hare,
"Characterization of Gaseous and Particulate Emissions from Light Duty Diesels
Operated on Various Fuels," Report EPA-460/3-79-008, June 1979, pp. 37-39.
31. R.E. Gibbs, 3.D. Hyde and S.M. Byer, "Characterization of Particulate Emissions
from In-Use Diesel Vehicles", Sfl.E Technical Paper Series, 801372, October 1980,
Report EPA/J-80-209.
32. 1. Miller and 3.E. Freund, "Probability and Statistics for Engineers," Prentice-Hall,
Inc., Englewood Cliffs, N.J., 1965, p. 174.
33. R.D. Cuthbertson, H.C. Stinton, and R.W. Wheeler, "The Use of a Thermo-
gravimetric Analyser for the Investigation of Particulates and Hydrocarbons in
Diesel Engine Exhaust," SAE Technical Paper Series, 790814, September 1979.
34. R.E. Gibbs, et. aL, "Sulfate and Particulate Emissions from In-Use Catalyst
Vehicles: Regulated/Unregulated Emissions and Fuel Economy," Report EPA-
600/9-79-047, December 1979.
35. B. Ames, 3. McCann ano E. Yamascki, Mutation Research, 3[, 347-364,1975.
36. M.3. DeSerres and M.D. Shelby, "The Salmonella Mutagenicity Assay: Recom-
mendations," Science, 203, 9 February 1979, 563-565.
37. R. Gibbs, 3. Hyde and R. Whitby, "Particuiate Emission Characterization Studies
of In-use Diesel Automobiles," presented at EPA 1981 Diesel Emissions Symposium,
Raleigh, N.C., October 5-7,1981.
38. P.M. Black, et. ah, "Exhaust Emission Patterns for Two Light-Duty Diesel
Automobiles," SAE Technical Paper Series, 810081, February 1981.
39. Nissan Motor Co., Ltd. "1-Nitropyrene Emissions from Five Production Model
Diesel Vehicles and the Effect of Damping Valve on the Emission," October 1981,
4 p.
40. Nissan Motor Co., Ltd., and Nissan Diesel Motor Co., Ltd., "Analysis of the
Factors Affecting Unusually High BaP Emissions from a Nissan SD-22 Diesel
Engine Vehicle Observed by EPA," October 1981,15 p.
114
-------
41. C.R. Clark, A.L. Brooks, et. aL, "Influence of Driving Cycle and Car Type on the
Muta&enicity of Diesel Exhaust Particle Extracts," 1981 Diesel Emissions
Symposium, Raleigh, N.C., October 5-7,1981.
42. 3. Braddock, "Emission of Diesel Participate Mutagens at Low Ambient
Temperature," 1981 Diesel Emissiosn Symposium, Raleigh, N.C., October 5-7,1981.
43. D.S. Barth and S.M. Blacker, 3. Air Pollution Control Assoc., 28, 769, (1978).
44. F.W. Karasek, RJ. Smythe, RJ. Lauh, 3. Chromatography, KM, 125 0574).
45. B.A. Bricklemyer and R.S. Spindt, SAE Technical Paper Series 1978, 780115.
46. C.F. Rodriguez, "Draft Report on Characterization of Organic Constituents in
Diesel Exhaust Particulates," June 1978.
47. J. Huisingh, R. Bradow, R. Jungers, L. Claxton, R. Zweidinger, S. Tejada, J.
Bumgarner, F. Duffield, M. Wates, V.F. Simon, C. Hare, C. Rodriguez, L. Snow,
In "Application of Short-Term Bioassays in the Fractionation and Analysis of
Complex Environment! Mixtures;" M.D. Waters, S. Nesnow, 3.L. Huisingh, S.S.
Sandhu, L. Claxton, Eds.; Plenum Press: New York, 1979; p. 381-418.
48. D.R. Choudhjry and C.O. Doudney, "Mutagenic Activity of Diesel Emission
Particulate Extracts and isolation of the Mctagenic Fractions," Environment
International, 5, 249-253 0 81);
49. D.R. Choudhury and B. Bush, "Polynuclear Aromatic Hydrocarbons in Diesel
Emission Particulate," Environment International, 5, 229-233 (1981);
50. D.R. Choudhury and B. Bush, "Gas Chromatography-Mass Spectrometric Char-
acterization of Poiynuclear Aromatic Hydrocarbons in Particulate Diesel
Emissions," ACS Symposium Series, 149, 357-368 (1981);
51. D.R. Choudhury, "Application of On-line High Performance Liquid Chromato-
graphy/Rapid Scanning Ultraviolet Spectroscopy to Characterisation of Poly-
nuclear Aromatic Hydrocarbons in Complex Mixtures," In: Polynuclear Aromatic
Hydrocarbons; Chemical Biological Fate, Ed. by M. Cooke and A.J. Dennis, p.
265-276, Battelle Press, Columbus, OH, 1981;
52. D.R. Choudhury, "Characterization of Polycyclic Ketones and Quinones in Diesel
Emission Particulates by Gas Chromatography/Mass Spectrometry," Environ. Sci.
and Technol., 16, J02-106 0982).
53. T.L. Gibson, A.I. Ricci, R.L. Williams, In "Polynuclear Aromatic Hydrocarbons:
Chemical Analysis and Biological Fate;" M. Cooke, AJ. Dennis, Eds.; Battelle
Press, Columbus, OH, 1981; p. 707.
54. J.A. Yergey, T.H. Risby, and S.S. Lestz, Environ. Sci. and Technol., 54, 354,1982
55. D. Schuetzle, F.S.-C. Lee, T. Prater, S. Tejada, Int. 3. Environ. Anal-Chem., 9,
93-144,1981. ~
115
-------
56. M.L. Yu and R.A. Hites, Anal. Chem., 53, 951, (1981).
57. M.D. Erickscn, D.L. Neuton, E.D. Pellizzari, K.B. Tomer, 3. Chrorr.atoR. Sci., 17,
449, 0979).
58. D. Schuetzle, T.L. Riley, TJ. Prater, T.M. Harray, D.F. Hunt, Anal. Chem., _54,
265, (1982).
59. M.F. Salamone, 3.A. Meddle, M. Katz, Environ. Int., 2, 37, (1979).
60. J.N. Pitts, Philos. Trans. R. Soc. London, Ser. A 1979, 290, 551.
61. F. Karasek, 3. Sweetman, R. Clement, F. Vance, H-Y Tong, "Chemical Character-
ization of Organic Compounds Associated with Diesel Engine Participate
Emissions," Fina' Contract Report to Ontario Ministry of Environment, April 1982.
62. R. Whitby, R. Gibbs, R. Johnson, B. Hill, S. Shimpi, and R. 3orgenson, "Real-Time
Diesel Paniculate Measurement Using a Tapered Element Oscillating
Microbalance," SAE Technical Paper Series, 820463, February 19S2.
63. H. Pataschnick and G. Rupprecht, "The Tapered Element Oscillating
Microbalance, A Monitor for Short-Term Measurement of Fine Aerosol MC.SS
Concentration," Environmental Protection Agency Report, EPA 600/2-81-146,
August 1981.
64. 3.N. Pitts, Jr., K.A. VanCauwenberghe, D. Grosjean, eJU ah, "Atmospheric
Reactions o* Poiycyclic Aromatic Hydrocarbons: Facile Formation of Mutagenic
Nitro Derivatives," Science, 202, 515-519,1978.
65. F.S.-C. Lee, T.M. Harvey, TJ. Prater, M.C. Paputa, and D. Schuetzle, "Chemical
Analysis of Diesel Particulate Matter and an Evaluation of Artifact Formation,"
Sampling and Analysis of Toxic Organics in the Atmosphere, ASTM STP 721,
American Society for Testing and Materials, 1980, pp. 92-110.
66. C.R. Clark, TJ. Truex, F.S.-C. Lee, and I.T. Salmeen, "Influence of Sampling
Filter Type on the Mutagenicity of Diesel Exhaust Particulate Extracts," Atmos.
Environ.' 15_, 397-402,1981.
67. Dennis Schuetzle, "Sampling of Vehicle Emissions for demical Analysis and
Biological Testing," Karolinska Institute Symposium" Biological Tests in the
Evaluation of Mutagenicity and Carcinogenicity of Air Pollutants," Stockholm,
Sweden, February 8,1982, Env. Health Perspective 3. (In press).
68. R.L. Bradow, R.B. Zweidinger, F.M. Black, and H.M, Dietzman, "Sampling Diesel
Engine Particle and Artifacts from Nitrogen Oxide Interactions," SAE Technical
Paper Series, 820182, February 1982.
69. Federal Register, 40CFR86, p. 14496 f.f., March 5,1980.
70. T. Penninga, "Evaluation of Exhaust Collection Configuration on Diesel
Particulate Measurement," U.S. Environmental Protection Agency Report, EPA-
AA-TAEB-79-8, Ma;' 1979.
116
-------
71. K. Stulp, "Particulate Measurement-Efficiency of PalJflex T60A20 Filter Media,'1
U.S. Environmental .Protection Agency Report, EPA-AA-SDSB-79-22, Duly 1979.
71. J. Braddock and P. Gabele, "Emission Patterns of Diesel-Powered Passenger Curs
- Part II," SAE Technical Paper Series, 770168, February 1977.
117
-------
APPENDIX A
VEHICLE AVERAGE EMISSION RESULTS
A-l Paniculate, Phase 1 (g/mi)
A-2 Participate, Phase 3 (g/mi)
A-3 Residue, Phase 1 (g/mi)
A-<» Residue, Phase 3 (g/mi)
A-5 Extract, Phase I (g/mi)
A-6 Extract, Phase 3 (g/mi)
A-7 Extract, Phase 1 (%)
A-8 Extract, Phase 3 (%)
A-9 Non-Volatile Hydrocarbons, Phase I, (% of Total HC)
A-10 Non-Volatile Hydrocarbons, Phase 3, (% of Total HC)
A-ll Total Hydrocarbons, Phase 1, (g/mi)
A-l2 Total Hydrocarbons, Phase 3, (g/mi)
A-l3 Carbon Monoxide, Phase 1, (g/mi)
A-l* Carbon Monoxide, Phase 3, (g/mi)
A-15 Nitrogen Oxides, Phase 1, (g/mi)
A-I6 Nitrogen Oxides, Phase 3, (g/mi)
A-17 Fuel Economy, Phase 1, (MPG)
VIS Fuel Economy, Phase 3, (MPG)
A-19 Revertants/yg Particulate, Phase 1
A-20 Revertants/ug Particulate, Phase 3
A-21 Revertants/ug Extract, Phase I
A-22 Revertants/ug Extract, Phase 3
A-23 Revertants/mile, Phase 1
A-24 Revertants/mile, Phase 3
118
-------
Table A-l:
F-AR
CAR
1
2.
3
4
5
6
7
B
9
10
11
12
13
14
16
17
18
IV
70
21
Table A-2:
ULA1E. G/fll, PHASf 1
FTP
0.38
O.V2
O.V4
1.20
O.HI
0.33
O.VH
0.42
0.33
0.59
0.36
0.47
0.62
0.50
0.64
0.82
0.45
0.81
O.4H
O.93
0.47
CFDS
0.35
0.73
0.4U
O.87
0.52
0.25
0.56
0.41
0.33
0.53
O.33
0.41
0.43
O.40
0.51
0.53
0.41
O.39
O.43
0.61
0.35
HR T
O.3H
O.SO
0.31
O.6V
0.38
0.22
0.3/
0.36
0.34
O.50
O.3O
0.37
0.42
0.38
0.44
0.3V
O.35
0.23
0.43
0.44
0.42
50C
0.3V
0.4V
0.26
O.64
0.2V
O.2O
0.37
0.30
0.31
0.38
0.28
0.3o
0.40
0.35
0.36
O.32
0.27
O.19
0.37
0.91
0.37
HYCC
0.51
2.02
1.75
2.37
1.74
0.40
2.2H
0.50
0.42
0.68
0.53
0.91
O.V2
0.94
1.1H
1.68
0.84
1.03
0.84
2.10
0.57
bAGl
0.52
1.04
O.VO
1.52
1.03
0.51
1.03
O.55
0.53
1.12
O.48
0.52
0.76
0.56
0.70
l.OO
0.5V
1.66
0.60
1.06
0.69
I
-------
Table A-3:
CAR
. G/nl, PHASF 1
FTt> CFlib HFET
50C
NYCC
1
2
3
4
5
6
7
B
9
10
11
12
13
14
IS
16
17
IB
19
20
21
0.32
0.59
0.60
0.98
0.65
0.28
0.72
0.34
0.?6
0.45
O.24
0.34
0.56
0.44
O.44
0.69
0.20
0.41
0.43
0.73
0.33
O.29
0.3V
0.33
0.67
0.39
0.20
O.37
0 . 3.1
0.25
0.39
0.21
0.29
0.38
0.34
0.35
0.41
0.15
0.14
0.39
0.44
0.13
0.32
0.23
0 . 22
0.52
0.27
0.1H
O.23
O./H
0.2H
0.33
0.19
0.27
0.3H
O.J3
0.33
o.:-9
0.11
0.09
0.40
0.28
0.12
0.33
o.:-o
O.I/
0.46
O.l'O
0.17
0.19
0.24
0.25
0.20
0.18
0.25
0.35
0.28
0.27
0.22
O.O8
O.O6
0.33
0.52
0.07
0.41
1.24
1.42
1.B9
1.37
0.2V
1.5H
0.3V
0.28
0.46
0.25
0.70
0.83
O.bO
0.74
1.37
0.30
0.31
0.74
1.45
0.46
Table A-4:
RrSIDUK, G/H1, PHASE 3
CAR FTP CFUS HFF.T
7
8
9
10
11
12
13
14
15
16
17
IB
19
20
0.37
0.5V
0.60
0.95
0.63
0.31
0.59
0.33
0.26
O.bH
0.25
0.40
0.55
0.49
0.46
0.67
0.22
0.32
0.4H
0.23
0.29
0.3?.
0.34
0.64
0.38
0.19
0.36
0.35
0.26
0.37
0.22
0.30
0.42
0.34
0.38
0.42
0.12
0.15
0.41
0.14
0.29
0.22
0.22
0.50
0.26
0.15
0.25
O.2B
0.25
0.24
0.19
0.27
0.41
0.31
0.34
0.42
0.10
0.10
0.35
0.12
50C
0.30
0.18
0.16
0.40
0.21
0.16
0.19
0.24
0.25
0.16
0.19
0.27
0.40
0.27
0.29
0.23
0.09
0.07
0.29
0.07
NYCC
0.37
1.26
1.38
1.67
1.34
0.29
1.21
0.32
0.29
0.23
0.68
0.89
0.76
0.62
1.32
0.30
0.39
0.63
0.43
120
-------
Table A-5:
Table A-6:
EXTRACT, G/MI, PHASE 1
CAR FTP CPUS Ht-T.T
EXTRACT, G/rtl, PHASE 3
CAR FTP CFUS HFET
•JOC
50C
NVCC
1
2
3
4
5
6
7
B
9
10
11
12
13
14
15
16
17
IB
19
20
21
0.06
0.33
0.14
0.22
0.15
o.ot,
0.26
O.OB
0.07
0.15
0.11
0.12
0.06
0.06
0.20
0.13
0..?6
0.45
0.05
0.20
0.14
O.06
0.34
O.I?
0.19
0.12
0.05
0.19
O.OB
O.OB
0.16
0.12
0.11
0.05
0.05
0.15
0.11
0.27
0.25
0.04
0.16
0.22
0.05
0.27
0.09
0.17
0.11
0.04
0.14
0.07
0.06
0.1B
0.11
0.10
0.05
0.05
0.11
0.10
0.?4
0.14
0.04
0.16
0.30
O.O6
O.29
O.O9
0.1B
O.OV
0.03
0.17
0.06
0.06
0.18
0.10
0.11
0.05
0.06
0.10
0.10
C. 19
0.13
0.03
0.3U
0.30
0.10
0.78
0.33
0.4V
0.3H
O.OB
0.70
O.ll
0.14
0.23
0.2B
0.21
0.10
0.13
0.44
0.32
O.54
0.72
0.10
0.65
0.11
NYCC
1
2
3
4
5
6
7
B
9
10
11
12
13
14
IS
16
17
IB
19
20
0.08
0.38
0.17
0.29
0.19
0.07
0.23
0.07
0.07
0.1V
0.11
0.13
0.06
0.07
0.16
0.14
0.29
0.49
0.04
0.10
0.06
0.36
O.14
0.27
0.15
O.Ob
0.19
0.0?
0.08
0.17
0.12
0.11
0.05
0.05
0.12
o.ii
0.24
0.26
0.05
0.21
0.05
0.32
0.12
0.25
0.13
0.05
0.16
0.07
0.07
0.17
0.10
0.11
0.05
0.04
0.09
0.15
0.24
0.17
0.03
0.30
0.05
0.33
0.13
0.24
0.12
0.03
0.17
0.07
0.06
0.15
0.09
0.10
0.05
0.05
0.08
0.10
0.21
0.16
0.03
0.25
0.10
0.76
0.42
0.82
0.3B
0.09
0.62
0.12
O.13
0.25
0.25
0.23
0.10
0.13
0.33
0.34
0.63
0.70
0.09
0.13
121
-------
Table A-7:
PffcCENT OOKACT. PHASE I
CAft
CPUS
HH r
NYCC
t
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
IB
IV
20
21
16. 0
34. 0
IB. A
17. 7
19. O
16.1
26.0
IV. 1
20.4
24.7
33.1
26.0
V.7
12.2
30.1,
15. V
5ft. 2
55.9
V.V
21.5
30.5
16.4
44.1
26.0
21. H
23.7
IV. 5
33.0
20.1
22.7
29.2
37.8
28.3
10.8
13.4
30.0
21. S
64.4
63. S
10.6
27.0
62.7
14.4
M.4
1>V. 4
24.1
27.4
IV. 4
37.5
:?o.3
IV. 3
33.4
36.3
27.3
10.9
12.2
24. U
25.3
67.9
61.4
9.2
36. O
71.3
1',.6
V/j . 7
34.7
V7.1
29.6
16.4
46.2
21 .V
19.5
4X.9
35.3
29.6
11.8
17.2
26.5
29.7
68.4
66.2
9.1
42.2
HO. 4
1B.V
38.7
1H.3
20.3
21.5
22.7
30.7
21.9
33.8
33.1
53.6
23.1
10.6
14.2
37.6
18. U
64.2
70.5
12.5
31.0
?0.1
Table A-8:
PERCENT EXTKACI, F'HASE 3
CAR-
CFDS
HFfcT
iOC
NYCC
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
17.9
37.1
22.4
21.7
??.V
I/. 4
27.1
17.6
20.1
23.2
32.0
24. V
9.9
11.7
24.9
17.0
56.6
58.6
B.6
30.8
16.7
47.2
29.2
2/.5
28.5
20.4
33.1
20.7
24.0
31.7
34.9
27.4
10.2
12. B
24.4
20.6
63.0
62.1
10.6
60.1
15.7
53.4
34.0
30.3
32.0
24.5
36.6
20.0
21.9
U.7
34.3
28.1
1O.O
11.9
20.6
24.6
69.0
63.6
8.5
71.0
15.5
5B.7
43.3
31.8
34.5
17.1
42.5
22.4
20.0
48.0
31 .7
27.3
11.1
14.7
20.2
29.5
67.5
68.9
8.6
77.0
21.3
37.5
23.5
24.8
22. J
23.3
33.7
27.9
30.0
30.9
52.2
22.5
10.3
14.8
34.6
20.3
67.2
66.0
12.7
24.0
122
-------
Table A-9: .
NON-VOLITILE HYDROCARBONS, PHASE 1. (» of Total HC)
CAR
FTP
CFDS
HFf.T
50C
NVCC
BAG1
Table A-10:
NON-VOLATILE HYDROCARBONS, PHASE 3, <* of Total HC)
FTP
CPUS
HKET
50C
NYCC
BAG1
bAl.2
!7.2
66.4
64.1
39.7
4U.5
88.7
44.3
56.1
98.4
12.1
20.9
36. V
24.8
38.8
35.1
53.9
40.2
20.2
51.7
58.1
25.8
37.2
129.7
29.3
48.8
35.4
20.7
16.9
46.4
35.9
29.9
17.8
32.8
26.3
31.5
25.4
23.8
31.3
21.9
32.9
59.1
41.1
22.9
32.0
21.9
44.6
24.9
40.1
27.5
33.3
19.5
18.6
19.5
42.7
30.9
38.9
30.0
23.8
39.7
25.0
29.9
26.4
30.9
38.0
30.2
26.4
39.7
27.1
33.1
15.1
27.8
29.5
29.2
21.8
48.8
31.8
48.9
35.3
19.6
37.6
22.7
33.2
45.4
40.7
47.3
36.5
36.9
40.0
35.2
38.4
24.1
47.2
32.5
12.9
BAG3
1
2
3
4
5
6
7 '
8
9
10
11
12
13
14
15
16
17
18
19
20
27.0
43.0
31.1
40.4
31.9
31.3
32.0
21.5
32.0
45.0
30.7
29.5
29.2
22.0
36.0
27.6
34.^
19.3
26.1
12.2
23.4
61.5
42.7
63.7
42.5
22.0
43.2
28.0
41.0
67.2
41.3
41.1
38.8
28.0
36.8
34.5
47.9
24.9
45.0
12.8
26.?
64. H
43.4
74.5
44.1
30.6
45.1
35.0
37.6
83.9
55.8
46.6
47.2
34.2
33.5
55.0
70. C
41.4
4 '.i.3
12.5
24.3
79.1
56.4
67.2
52.9
29.2
53.9
32.7
47.7
94. O
57.1
53.4
55.3
34.8
3H.5
46.4
76.5
50.0
42.7
9.6
24.4
34.0
31.3
45.4
30.5
35.5
28.3
22.7
87.3
46.8
23.8
28.4
48.7
24.0
48.5
28.5
22 . 2
17.2
32.2
34.1
30.1
38.7
26.0
36.7
23^7
25.6
19.9
31.3
44.3
37.4
22.7
23.6
19.1
41.6
25.7
33.1
24.7 •
30.6
16.2
25.4
42.4
32.1
37.3
31.7
39.7
33.6
21.2
32.5
44.4
23.2
30.9
29.6
21.9
34.8
25.7
34.3
15.5
19.9
18.6
26.6
49.4
35.8
53.6
38.5
23.1
35.0
22.2
32.5
48.4
35.4
36.0
36.8
26.7
34.3
35.1
36,1
25.2
35.9
6.0
123
-------
Table A-ll:
GASEOUS HYDROCARBONS, G/ttI, PHASE 1
CAR
FTP
CfDS
HFET
50C
NYCC
BAG1
Table A-12:
GASEOUS HYDROCARBONS, G/M1 , PHASE 3
CAR
FTP
CFHS
HFET
50C
NYCC
BAG1
BAG2
BAG2
BAG 3
i
2
3
A
5
6
7
a
V
10
11
12
13
14
15
16
17
ie
19
20
21
0.33
0.71
0.48
0.55
0.52
0.24
0.69
0.34
0.22
0.36
0.32
0.35
0.19
0.23
0.49
0.47
0.74
2.18
0.14
0.74
0.73
0.28
0.50
0.2V
0.31
0.31
0.22
0.42
0.33
0.19
0.31
0.25
0.23
0.11
0.14
0.39
0.27
0.47
O.VO
O.OB
0.42
1.87
0.29
0.41
O.24
0.25
O.24
0.20
0.37
0.2)
0.21
O.29
0.17
0.21
O.09
0.11
0.33
0.21
0.33
0.40
o.oa
0.36
2.32
0.37
0.55
0.24
0.28
0.24
0.13
0.45
0.22
0.19
0.21
0.15
0.22
0.08
0.12
0.30
0.22
0.27
0.36
0.06
0.39
2.47
0.47
2.13
1.31
1.25
1.08
0.12
1.73
0.55
0.31
0.39
1.11
0.56
O.12
0.45
0.88
0.93
2.61
4.63
0.23
1.B1
0.38
0.48
0.9V
0.67
0.85
0.76
0.36
0.86
0.4H
0.33
0.47
0.37
0.58
0.25
0.33
0.50
0.67
0.85
3.26
0.18
1.18
1.14
0.29
0.72
0.47
0.50
0.50
0.18
0.70
0.26
0.17
0.30
0.31
0.31
0.20
0.22
0.48
0.47
0.74
2.27
0.15
0.67
0.42
0.29
0.60
0.37
0.41
0.38
0.25
0.55
0.40
0.22
0.37
0.32
0.25
0.13
0.16
0.50
0.32
0.64
1.20
0.11
0.54
1.02
BAG3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
0.32
0.8-5
O.LJ
0.66
0.5V
0.24
0.77
0.33
0.21
0.39
0.36
0.45
0.21
0.30
0.43
0.51
0.88
2.53
0.17
0.82
0.26
0.55
0.33
0.40
0.36
0.23
0.41
0.34
0.20
0.26
0.28
0.28
0.12
0.18
0.33
0.31
0.50
1.13
0.10
1.67
0.22
0.4A
0.27
0.32
0 . 28
0.20
0.33
0.21
0.20
0.21
0.18
0.23
0.10
O.12
0.27
0.25
0.35
0.51
0.07
2.43
0.25
0.64
0.29
0.44
0.29
0.14
0.44
0.29
0.19
0.16
0.19
O.22
O.tl
0.15
0.25
0.29
0.33
0.45
0.07
2.60
0.41
2.28
1.36
1.52
1.25
0.26
2.21
0.55
0.19
0.53
1.05
0.80
0.25
0.55
0.67
1.27
2.90
4.26
0.28
0.39
0.51
i.oa
0.76
0.90
0.82
0.32
1.02
0.45
0.32
0.75
0.42
0.65
0.29
0.40
0.46
0.64
1.03
3.26
0.22
1.09
0.25
0.83
0.54
0.65
0.58
0.20
0.66
0.26
0.16
0.28
0.35
0.43
0.21
0.30
0.42
0.53
0.8B
2.85
0.17
0.44
0.29
0.69
0.43
0.49
0.45
0.26
0.58
0.39
0.23
0.31
0.36
0.33
0.15
0.22
0.43
0.38
0.77
U37
0.13
1.36
124
-------
Table A-13:
CARBON MONOXIDE. G/hl, PHASE 1
CAR FTP CPUS HKET HOC
NYCC
BAG1
BAG2
Table A-14:
CARBON MDNOXIUK', ti/MJ, PHASE 3
CAR FTP CFKS HFET 50C.
NYCC
BAG1
BAG2
BAG3
1 1
2
3
A
5
6
7
8
9
L.17
.74
.48
.97
.61
.04
.71
.26
.05
10 1.97
11 1.16
12 1.21
13 1.19
14 1.27
15 1.85
16 1.57
17 1.55
18 2.67
1# 0.97
2O J . 75
21 1.81
1.05
l.?O
O.VV
1.27
1.02
0.74
1.11
1.16
0.81
1.60
0.82
O.VO
O.H4
1.02
1.43
1.01
O.V6
1.40
0.74
1.11
1.56
1
0
0
1
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
O
1
.12
.98
.83
.06
.91
.69
.93
.94
.83
.29
.66
.86
.77
.95
.23
.87
.Bl
.97
.71
.97
.46
1.42
0.98
0.82
1.07
0.93
0.60
0.90
0.98
0.79
1.09
0.65
0.77
0.81
1.00
1.15
0.83
0.78
0.86
0.64
0.92
1.04
2.
4.
3.
4.
3.
1.
4.
1.
1.
2.
3.
2.
1.
2.
3.
3.
4.
6.
1.
4.
1.
07
22
68
42
54
44
03
92
62
63
23
09
72
43
49
32
06
00
72
47
88
1.45
1.87
1.63
2.12
1.76
1.22
1.77
1.58
1.82
2.34
1.27
1.54
1.30
1.53
2.11
1.76
1.61
3.16
1.13
2.01
2.21
1
1
1
.04
.87
.54
2.00
1.69
1.00
1
1
0
1
1
1
1
1
1
1
1
2
0
1
1
.83
.04
.82
.80
.15
.18
.20
.26
.75
.64
.74
.83
.95
.81
.43
1.21
1.50
1.26
1.65
1.35
0.97
1.44
1.44
0.90
2.03
1.07
1.01
O.98
1.10
1.82
1.27
1.13
1.99
0.90
1.43
2.24
BAG3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1.18
1.83
1.53
1.97
1.64
1.05
1.62
1.27
0.92
1.92
1.14
1.30
1.23
1.47
1.67
1.56
1.5V
2.63
0.99
1.79
1.00
1.20
1.02
1.32
1.09
0.76
1.09
1.17
0.81
1.43
O.87
0.93
0.91
1.03
1.29
1.02
0.93
1.51
0.73
1.72
0.94
1.02
0.87
1.07
0.92
0.65
0.91
0.88
0.75
1.10
0.67
0.87
0.8V
0.94
1. 10
0.87
0.85
1.03
0.6H
1.56
1.15
1.02
0.84
1.14
0.94
0.61
0.94
1.05
0.75
0.99
0.60
0.80
0.89
0.99
1.01
0.83
O.79
0.93
0.62
1.10
2.01
4.47
3.73
4.67
3.31
1.79
3.99
1.87
1.66
3.45
2.99
2.78
2.00
2.55
2.98
3.60
4.31
6.41
1.76
1.97
1.47
1.92
1.68
2.10
1.79
1.20
1.48
1.55
1.19
2.43
1.28
1.53
1.37
1.67
1.98
1.67
1.63
2.89
1.14
2.19
1.09
1.93
1.58
2.07
1.72
1.04
1.78
1.06
0.81
1.75
1.11
1.33
1.27
1.50
1.55
1.65
1.74
2.88
0.99
1.4S
1.15
1.55
1.32
1.69
1.37
0.95
1.40
1.44
0.91
1.88
1.10
1.08
1.06
1.25
1.66
1.32
1.26
1.94
O.90
2.12
125
-------
Table A-15:
NTTROGfcN COUDKS, ti/hl, PHASE 1
CAR FTH CPUS HFET 50C
NYCC
bAGl
Table A-16:
NITROGF..N OXlIiF.ii, li/MI, PHASE 3
CAR FTP CPUS HFET 50C
NYCC
BAG!
BAG2
BA62
BAG3
1
2
3
A
5
6
7 .
8
9
10
11
12
13
14
15
16
17
18
19
20
21
0.90
l.KO
1 .57
2.03
1.8V
O.B9
l.Hl
0.95
1.20
3.30
0.93
1.29
1.76
1.39
1.90
1.61
1.03
1.10
1.70
1.54
1.79
0.81
1.47
1.29
1.62
1.3V
0.62
1.3 A
o.as
1.09
3.14
0.76
1.13
1.4H
1.27
1.43
1.19
0.84
O.79
1.55
1.22
1.50
O.B1
1.43
1.29
1.62
1.41
0.57
1.27
0.89
1.13
3.39
0.75
1.O6
1 , 45
1.27
1.40
1.11
0.79
0.74
1.55
1.22
1.45
0.82
1.38
1.21
1.63
1.34
0.56
1.25
0.96
1.23
3.44
0.76
1.08
1.33
1.26
1.41
1.11
0.76
0.77
1.36
1.25
1.48
1.26
3.19
2.41
3.47
3.00
1.67
2.82
1.45
1.64
4.04
1.45
1.93
2.51
2.17
2.79
2.83
1.51
2.17
2.54
2.39
2.98
0.92
1.70
1.52
1.91
1.74
0.90
1.64
0.97
1.19
3.66
0.96
1.36
1.85
1.41
1.81
1.56
1.10
0.99
1.77
1.51
1.71
0.90
1.86
1.62
2.12
2.02
0.96
1.95
0.94
1.23
3.07
0.95
1.27
1.77
1.42
2.05
1.71
1.07
1.21
1.65
1.57
1.87
C.89
1.72
1.50
1.97
1.77
0.75
1.66
0.95
1.17
3.44
0.89
1.28
1.72
1.35
1.69
1.46
0.91
0.97
1.74
1.51
1.71
BftG3
1
2
3
4
5
6
7
a
9
10
11
12
13
14
15
16
17
18
19
20
0.93
1.73 .
1.53
1 .99
1.82
0.84
1.74
0.96
1.21
3.15
0.88
1.32
1.81
1 . 50
2.05
1.6O
1.02
1.10
1.71
1.73
O.B2
1 . 47
1.25
1.64
1.45
0.59
1.34
O.B7
1.12
3.01
0.76
1.11
1.49
1.26
1.49
1.21
0.79
0.81
1.53
1.42
0.85
1.41
1.25
1.59
1.41
0.55
1.2.B
0.91
1.15
3.19
0.78
1.09
1.48
1.26
1.45
1.14
0.8C
0.69
1.52
1.42
O.H5
1.35
1 . 20
1.59
1.36
0.56
1.21
0.96
1.30
3.07
0.82
1.12
1.31
1.23
1.45
1.12
0.77
0.80
1.32
1.42
1.35
2.97
2.44
3.42
2.84
1.63
2.62
1.37
1.83
4.42
1.44
2.05
2.63
2.?3
2.93
2.89
1.60
3.20
2.37
2.87
0.99
1.62
1.47
1.85
1.69
0.82
1.60
0.97
1.18
3.44
0.90
1.38
1.91
1.56
1.99
1.45
1.04
1.01
1.95
1.74
0.91
1.81
1.58
2.07
1.92
0.90
1.88
0.96
1.24
2.94
0.87
1.30
1.80
1.51
2.20
1.72
1.05
1.20
1.61
1.77
0.92
1.68
1.50
1.96
1.72
0.74
1.61
0.96
1.19
3.33
0.88
1.30
1.76
1.43
1.79
1.49
0.95
0.98
1.72
1.63
126
-------
Table A-17:
FUEL KCIINOrtY, MPfi, PHASE 1
CAR FTP CFBS HFfc'T
50C
NYCC
HAG1
BAG2
BAG3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
42. V
20.0
20.7
19.4
20.7
41 .7
18.7
42. A
43.8
1.8.5
41.7
27. A
22.9
27.8
27.2
19.3
27.4
20.2
23.3
19.2
23.1
53.4
27.1
28. 0
27.1
30.3
55.7
26.7
52.9
54.8
24.9
53.4
34.9
28.7
32.4
34.7
26.9
34.2
28.8
28.9
26.3
28.3
55 . 7
30.1
30.4
2V . 4
31.3
58.9
30.2
54.0
56.1
26.0
54.9
36.4
30.4
34.6
36.8
29.3
33.7
32.0
30.3
28.7
29.9
S7.9
32.0
32.7
33. H
33.2
64.0
32.4
54.4
60.0
27.2
57.3
37.5
31.6
35.6
38.7
31.4
37.1
34.1
31.6
30.8
32.3
29.9
10.6
11.6
10.3
11.4
28.6
10.9
28.3
30.8
13.7
26.7
18.6
17.1
18.5
19.1
10.9
18.9
11.5
15.5
11.2
15.9
38.5
18.3
18.8
18.2
19.0
37.0
17.9
39.1
39.8
15.7
38.5
25.1
20.0
25.2
25.6
17.0
24.5
17.9
20.5
17.0
20.3
44.1
19.7
20.5
18.8
20.3
41.7
18.1
43.3
44.2
1°.2
41.7
27.8
23.5
28.1
27.1
19.5
27.3
20.1
24.0
19.6
23.7
44.7
21.8
22.7
21.7
23.2
46.0
20.8
44.4
46.6
19.9
44.6
29.4
24.3
29.3
29.0
21.3
30.7
22.4
24.4
20.4
24.6
Table A-18:
FUEL FCllNOhY, MPG, PHASE 3
CAR FTP CFliS HFE.T
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
41.7
J9.9
2O.7
19.4
20.9
43.7
19.0
43.2
44.0
19.7
43.0
27.0
23.2
25.6
26.5
19.1
26.7
19.9
23.1
24.5
52.0
27.2
28.2
26.4
28.1
57.1
26.H
52.7
53.6
25.8
52.5
34.9
29.4
31.8
34.5
26.4
35.9
28.3
28.3
29.0
55.1
29.8
30.H
29.3
31. .2
61.3
29.9
54.7
58.4
27.4
56.6
36.6
30.2
33. fl
36.7
28.9
35.0
31.5
30.3
29.8
bOC
58. S
32.0
33.6
30.3
33.1
65.3
32.5
56.6
59.0
29.1
57.8
37.5
33.0
34.8
39.5
30.8
36.6
34.0
32.4
31.3
NYCC
29.3
10.9
12.0
10.2
11.8
29.6
11.0
31.0
29.6
12.3
27.5
17.4
16.7
18.0
19.3
10.3
18.3
11.3
15.8
16.2
BAG1
36.7
18.5
19.3
18.2
19.5
40.5
17.8
39.5
39.8
17.
39.
24.
20.2
22.8
24.7
18.0
24.1
18.0
20.4
20.9
6AG2
42.6
19.6
20.3
19.0
20.6
43.6
18.1
43.9
44.5
20.2
43.8
27.1
23.9
25.8
26.3
18.9
26.9
19.8
23.8
25.9
HAG3
«,4.4
21.8
22.8
21.1
22.9
46.7
2) .0
45.0
46.6
21.1
44.9
29.1
24.4
27.5
28.5
20.6
28.5
21.9
24.3
25.2
127
-------
Table A-19: REUEKTAMTS PER on PARTICULATE, PHASE i
CAR FTP l>"titi HFt-T SOC NYCC
1
2
3
A
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
1.2
0.8
0.9
0.5
0.6
2. A
O.B
1.3
1.0
0.9
1.7
0.5
0.5
0.8
1.2
0.8
O.V
0.8
0.6
1.0
l.V
1.5
0.9
O.B
O.4
0.5
3.3
0.9
1.6
1.2
0.9
1.4
0.5
0.5
0.9
1.4
0.8
0.9
0.9
0.9
0.8
2.3
1.6
0.9
0.7
0.3
0.6
3.3
1.0
1.1
1.2
0.6
1.0
0.3
0.5
0.6
1.7
0.9
0.4
0.9
0.6
1.0
2.4
1.5
0.7
0.7
0.3
0.6
1.7
1.3
0.7
0.9
0.3
1.2
0.1
0.3
0.7
1.7
0.9
0.4
1.0
0.8
1.5
1.4
1.3
O.3
0.4
0.4
0.2
1.1
0.4
0.2
0.6
0.4
1.3
0.4
0.1
0.4
0.9
0.5
0.3
0.3
0.3
0.9
0.5
Table A-20: KEVERTANTS PKK U(i PAKTfCUl.ATK, PHASF-3
CAR FTP CFMS HFKT 50C NYCC
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
3.0
0.7
0.8
O.S
0.6
2.2
O.H
2.ti
1.4
0.9
1.6
0.5
0.4
0.9
1.4
1.0
0.3
1.8
0.9
2.5
2.3
O.H
0.7
0.4
0.6
3.3
0.7
2.7
1.3
1.0
l.B
O.4
0.4
1.0
l.B
1.0
0.7
1.1
1.6
2.2
2.3
0.7
0.8
0.3
0.6
2.3
0.7
2.4
1.3
0.7
1.1
0.3
0.3
0.7
1.8
1.0
0.5
1.1
0.8
2.2
1.7
0.9
0.9
*0.3
0.6
1.6
O.B
1.0
1.1
0.5
1.0
O.3
0.2
0.8
1.2
1.0
0.5
1.1
O.8
1.7
1.6
0.3
0.5
O.S
0.2
l.V
0.6
0.2
0.6
0.6
l.B
6.4
0.1
0.5
1.0
0.6
0.2
0.4
0.3
0.4
128
-------
Table A-21: REVERTANTS PER UG EXTRACT, PHASE 1
CAR FTP CPUS HFET soc NYCC
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
7.V
2.9
4.9
3.4
3.7
16.8
3.5
6.7
6.7
3.4
5.2
1.9
5.3
5.9
4.3
4.9
1.6
1.6
5.5
4.5
6.2
9.7
2.4
3.5
2.3
2.5
17.6
3.4
7.5
6.9
3.3
3.9
2.1
5.1
6.6
4.6
3.9
1.4
1.4
7.7
3.1
3.7
12.3
2.2
2.9
1.1
2.6
18.6
3.3
5.5
8.0
1.7
3.1
1.1
5.4
4.6
7.1
3.5
0.5
1.4
5.9
2.8
3.4
9.7
1.6
2.2
1.3
2.2
11.4
3.8
3.0
6.1
0.6
3.4
0.5
2.3
3.5
7.1
3.2
0.6
1.6
8.2
3.5
1.7
7.4
0.9
2.4
1.8
1.1
8.6
1.2
0.9
1.8
1.3
2.4
1.6
1.4
2.6
2.4
2.9
0.4
0.4
2.2
2.8
2.5
Table A-22: REVERTANTb PtK UG KXTRAC1, PHASE 3
CAR FTP CFDS HFET 50C NYCC
1
2
3
4
6
7
8
9
10
11
12
13
'.4
15
16
17
18
19
2O
18.1
2.8
3.9
2.7
2.6
14.6
3.8
15.4
9.7
3.8
5.0
2.0
4.6
7,9
6.0
5.3
0.6
3.0
11.6
8.0
13.9
2.5
2.7
1.8
2.3
17.7
2.0
14.0
8.2
3.3
5.3
1.4
4.3
8.4
8.3
4.6
1.2
1.9
18.2
3.6
14.9
1.9
2.7
1.4
1.9
10.3
2.6
14.1
8.0
1.6
3.3
1.0
3.6
6.4
8.7
3.V
0.8
1.8
10.0
3.1
11.0
2.1
2.0
1.3
1.8
10.2
2.3
4.4
8.4
0.9
3.3
1.1
2.1
5.3
6.3
3.3
0.8
1.7
8.9
2.2
7.6
0.9
2.4
1.6
1.0
20.2
1.8
0.7
2.3
1.8
3.2
2.0
1.3
3.8
2.8
2.8
0.4
0.6
2.3
1.8
129
-------
Table A-23: REVERTANTS PER MILE x io5, PHASE i
CAR FTP CFDK HFKT 50C NYCC
1
2
3
4
5
6
'/
e
V
10
11
12
13
14
15
16
17
18
19
20
21
4.3
6.6
6.7
6.5
4.5
8.3
7.4
5.6
3.2
5.8
6.0
2.3
3.4
3.8
7. B
6.3
4.2
6.7
2.8
9.0
8.8
o.5
5.8
3.9
3.7
2.5
8.2
4.V
6.7
3.9
5.3
4.8
2.2
2.3
3.8
7.2
4.2
3.8
3.2
3.6
5.1
8.2
6.6
3.6
2.1
l.tt
2.0
7.7
3.7
3.8
4.2
3.1
3.3
1.2
2.3
2.5
7.6
3.5
1.3
1.8
2.2
4.4
10.2
6.6
3.1
2.0
2.2
1.6
3.5
4.4
2.0
2.9
1.1
3.3
0.5
1.0
2.3
6.3
2.9
1.1
1.7
2.8
13.4
b.l
6.9
6.V
8.9
9.6:
4.2
4.1
10.5
1.2
2.4
2.9
6.7
3.0
1.6
3.7
11.0
9.1
2.1
2.8
2.4
18.2
2.9
Table A-24: REVERTANTS PER MILE x io5, PHASE 3
CAR FTP CFDS HFET 50C NYCC
1 13.9 8.2 8.4 6.4 7.7
2 5.4 4.4 3.1 3.1 6.8
3 5.4 3.3 2.7 2.6 9.9
4 5.6 3.4 2.0 1.5
5 a.6 3.3 2.2 1.9 3.8
6 7.8 8.3 4.2 3.3 7.1
7 6.2 3.6 2.5 2.2 9.6
8 10.4 11.8 7.6 2.6 1.0
9 4.9 4.6 4.2 3.5 2.5
10 6.7 5.6 2.8 1.4 4.4
11 5.0 6.2 3.1 2.8 8.1
12 2.7 1.6 1.0 1.1
13 2.7 1.9 1.6 1.0 1.3
14 5.? 4.1 2.5 2.5 4.7
15 9.2 9.7 7.7 4.8 9.8
16 7.7 5.1 6.6 3.1 9.5
17 1.7 2.7 1.6 1.2 2.2
18 16.4 4.7 2.7 2.1 4.7
19 4.H 8.0 2.9 2.3 1.9
20 8.1 7.7 V.4 5.5 2.4
130
-------
APPENDIX B TYPICAL DOSE RESPONSE CURVES FOR TA98(-) AMES TESTS
I.It ft.41
I.M M.M
].*( »».!•
I.I* **.«•
t.tl »«.»*
rrcu nil"
IH
in
if*
1*1
m
HI
• 01
«».*«
«*.«!
¥l,«.»
• ••BO
»J
...*—*
MI
in
Ml
131
-------
APPENDIX C
VEHICLE TEST FUEL ANALYSES
Table C-l: Test Fuel Analysis.
Data for as-received test fuel for all tests are lh ted by test number. Average values
and statistics are presented for all as-received fuel!.. "AEL" denotes Project Reference
Fuel Analysis. Columns are from left to right: Test number; API gravity at 60°F by
ASTM D287-67; specific gravity at 60°F by ASTM D2S7-67; Heating Value (BTU/lb) by
ASTM D2<*0-6f; Ash Content (%) by ASTM D^82-7<»; Sultur (S) by ASTM D1552-60;
Carbon (C) and Hydrogen (H) by ASTM D-3178-73 reapproved 1979 - modified (oxidant
purified air instead of oxygen); Nitrogen (N) by ASTM D3I79-73 reapproved 1979;
Oxygen (O) by ASTM "Ultimate Analysis Method" D3176-7'* reapproved 1979 (mathe-
matical differencing method); Percent Saturates (% SAT), % Olefins (% OLEF), and %
Aromatics (% AROM) by ASTM DI319; and calculated Cetane Index (CETANE INDEX)
by ASTM D976-66 reapproved 1971. After test 3<» ASTM D1319 analyses not available
from EPA/RTP.
Table C-2: Test Fuel Distillation Profile by ASTM D86-67.
Data for as-received test fuel for all tests are listed by test number. Average values
and statistics are presented for all as-received fuels. "AEL" denotes Project Reference
Fuel analysis. Columns are from left to right: Test number, initial boiling point, 5%
recovery, 10% recovery, 20% recovery, 30% recovery,
-------
TABLE C-l ANALYSIS OF TEST FUELS*
u»i to* inrrric
at e»»«Tt tMvin
14.
14.
19.
15.
14.
15.
tl.
11.
17.
10 15.
11 It.
12 18.
11 15.
14 17.
15 It.
14 It.
IT 15.
11 11.
14 !«.
20 15.
21 )».
22 It.
2) It.
24 15.
25 15.
2t 14.
27 15.
21 14.
2t 15.
!C )4.
11 19.
12 !9.
!) 19.
14 )«.
)5 15.
It 14.
17 17.
It 14.
«C J7.
41 * .
42 2.
4 )
14 .
45 .
41
47
• t
«:
•?
« \ f
•i i .
5) ) .
M ) .
• • ) .
it n.
*7 )4 .
51 )7.
'.- 14.
(0 17.
•1 17.
i: 34.
4 ) 15.
44 15.
45 )5.
(t 15.
«,7 )5.
(« 34.
€ > 15.
?C 14.
71 14.
72 )5.
7) 11.
74 14.
75 )4 .
76 41.
It 4; _
7« )e.
74 )'.
1C It.
VC It.
C » .
:v 4 .
0.852*
0.852*
0.8471
0.8441
0 . 4 • 1 4
0.8458
0.4)41
0.1571
0.117)
0.14(1
C.tlll
0.4)18
0.44ft
0.1)7)
0.841)
0.8411
0.849)
C.4448
0.8441
0.847)
0.4421
0.141)
0.811)
0.8478
9.1471
0.8441
0.'441
'C. • 5 "4
0.844)
C.I45I
t.t4SI
0.144)
0.847)
O.l))8
C.tlit
0. 1448
0.8)4)
0.8112
0.8514
0.1)44
c . • ; -,4
c. »:
l.C 159
IL It. 2 0.8418
im/u
1942t
14110
1944*
1427*
19421
19)41
14449
19129
1197*
19910
19542
19521
11512
19542
19971
11411
1451)
14449
14104
19)72
19745
19411
19780
197 10
144C2
19449
19429
19722
19741
14494
141)4
19117
197)9
14815
14455
14459
144*1
147C5
14241
19681
14427
147M
14;:5
1 14 14
14481
19542
115:2
mi*
14414
15424
J9It4
19!9"
1 T 515
19K2
19)17
1954 1
19552
mi;
115)9
144?)
1>442
14545
19444
14524
14!I5
14537
14IC)
115C8
114 73
inj c
1«492
14712
11485
1-147
19" )1
19!) 5
I"47
1«455
19444
1452f
1)7
5.7
14454
•UV/Oftb
1)7971
1)7494
117)54
0
1)7(75
1)1249
1)1127
ll'IO)
1)1445
1)7(21
1)4119
1)95)4
117911
1)4X7
117045
11(101
1 171)2
1)4(04
1)4959
1144(9
118951
!)())(
119274
1)6(11
Ut72t
1K4K
119(51
119129
1)8499
11(1(4
1187)1
114299
11792)
1>7GC2
114875
11(217
1)112)
114724
1171)1
4 %* * d 4
1 Jv v * i
I)442(
1)5425
11'bSf
11(124
136)34
1 17? 10
1J744C
1)(«52
11(104
1)7212
1 36C94
1)71)7
1)7055
115944
11M44
1 ) • 8 44
1 )C8 24
1KC14
1)745)
1 1?) 9t
1)7(44
117(30
117(22
117455
117)75
1)6152
1 11224
117281
1 )!5CO
1 16(48
1 35' 15
1)474)
114IC8
1 )6C 10
1 )* 1 *4
1)4205
1)I«9I
1141
C.9
1)4717
t M
0.0005
0.0005
0.0005
0.0010
0.0009
0.0009
O.OC09
0.0009
0.0005
0.0009
0.0005
0.0009
0.0005
0.0009
0.0010
0.0010
0.0005
O.C005
C.OC09
O.C009
0.0005
0.0009
O.OCC5
0.0005
O.OC05
0.0005
C.0009
O.OOC5
C.CC05
O.CC05
0.0005
O.OCC5
O.COC5
O.OC50
O.CC05
O.CC05
0.0005
O.CCC5
O.CCC5
0. CCC5
C^ f f I
• 3 C C 3
C.CCC5
c.:cc5
O.CCC9
C. CC50
C.OC35
C.CC05
C.CC05
C.f 005
C.CC05
C.CC05
c.oecs
C.OC05
;.:CM
~ . rCG5
C . - CC5
0. C005
O.CCC5
C.30C5
O.OCC5
O.COC5
O.OOC5
C.OTC5
C.COC5
O.COC5
C.OC20
O.CCC5
O.C005
O.OC05
O.CC50
P. C 04 3
C. CC05
O.CC05
c.c:c5
c.::o5
C. CCC5
C.CCC5
o . c o •: 5
f.OC55
C.CO'.S
O.ICC9
c'.:cio
124.4
C.0005
*l
0.11
0.22
0.10
o.ot
0.11
0.11
0.0*
0.11
0.10
0.21
O.Jt
O.li
0. 11
C.24
0.21
0.11
0.11
0.11
O.li
O.lt
0.11
0.25
C.ll
O.lt
0.14
0.11
0.14
0.2t
0.24
0.11
0.11
0.10
0..0
0.10
0.15
0.15
O.Jt
0.01
0.20
0.11
C5*v
• 47. V
0.07
O.lf
0. 22
C.C7
0.15
0.01
0.22
0.19
C.15
C.15
0.14
0.15
O.lt
C.ll
C.17
1. J«
C.22
0.11
0.12
0.15
0.19
C.19
0.24
O.lt
0.18
0.24
C.ll
C.12
0.19
0.12
0.11
0. 25
i.)2
C.C9
0.07
C.C9
0.2)
0.21
0.11
c.ct
41. :a
0.14
V
at. ft
11.14
11.11
11.12
tt.lt
15.42
• 1.42
ll.lt
lt.lt
17.24
19.20
lt.lt
Bt.lt
11.11
85. *0
at. it
at. tt
It. 21
If. 91
17.24
• f .91
• 5.35
It. 02
Ct.ff
11.70
17.19
85.54
at.it
• 7. It
87.14
• 7. It
• 7.48
8'>. 87
84. ft
tf.f 7
14.44
8f.ll
15.94
86.16
• (.21
(6.44
B5.2(
85.59
at. 20
88.11
(5.74
86.26
at. it
81.14
at. 47
• 5.99
86. ;6
• f . ( 7
Bf .( 2
8(. 49
85.94
• 6.44
(4.28
8(.42
8f.22
86.39
8f .44
84.40
((.15
(t.55
It. 41
16.16
• (.51
at. 5!
8(.52
at. »]
55.0;
8(.C2
85. 51
85.01
84.09
9f .CO
85. 8(
B(.51
44.31
o!58
l(.ll
"**
12.**
12.7}
11.14
12. It
12.51
12.14
11.22
12. fl
11.1*
12.11
11.54
11.29
11.14
11.21
11.00
11.04
12.95
12.lt
12.94
12.27
12.74
12.51
12.10
11.72
12.91
11.01
12. f}
12.51
11.72
12.28
Ii.t5
11.15
12.09
11.14
12.74
11.14
1J.12
11.17
1..14
11.21
'• » **
i J • J9
11.11
14. Cl
11.21
11.28
11.59
ll.lt
11.12
12.98
12.94
12.87
11.20
11.45
12.89
13. CJ
12. (J
11.40
11.21
12.95
11.11
11.4'
11. 21
11. C4
ll.lt
11.19
11.12
11.00
12.18
12.72
11.21
11.21
11.51
11.10
11. 51
11.14
11.81
11. (1
11.18
11.15
11.18
12. «f
C . * 1
4.C7
ll.CO
"t*
0.04
0.04
0.04
0.04
0.01
0.04
0.04
0.04
0.04
0.04
O.Of
0.04
O.Of
O.Of
0.04
O.Of
0.01
0.01
0.0)
0.04
0.01
0.04
0.01
0.01
0.04
0.04
0.04
0.0]
0.04
0.01
0.01
0.04
0.04
0.01
0.01
0.01
0.01
C.01
0.01
0.01
0 • 04
0.01
0.07
C.10
0.1C
0.01
0.01
C.C4
C.C6
0.01
C.C4
C.C1
O.C4
C.C4
0.-.3
0.01
O.C1
0.01
C.01
C.C1
C.C]
0.0)
0.01
O.C1
O.C1
O.C1
0.01
C.C1
0.01
C.01
O.C1
O.C1
C.04
C.C4
O.C1
0.0]
0.01
C.04
O.C1
O.C4
e.:2
48.7:
O.C1
to • UT.
0.20 45. ',
O.fl fl.
0.41 45.
O.ft 64.
0.44 42.
t.Ot ft.
0.21 15.
0.1} t*.
0.11 tt.
0.28 72.
0.42 75.
O.lf 59.
0.12 tl.
0.11 58.
0.85 12.
0.58 6i.
O.ff 61.
O.lt tO.
0.2f 42.
0.15 t2.
1.12 ft.
0.54 (8.
2.18 (2.
1.51 tl.
1.00 (0.
0.91 99.
0.84 59.
C.21 fO.
C.75 64.0
1.0) (1.4
0.90 99.0
O.Cf 61.9
0.41 0.0
0.02 0.0
0.10 0.0
0.11 0.0
0.6', C.O
0.42 0.0
017 A A
• i it U • V
C.CC C.O
0.12 0.0
C.?2 C.O
C.ll C.C
C.72 0.0
0.59 0.0
O.tO O.C
C.47 C.C
0.48 C.C
0.62 0.0
0.12 O.C
C.21 -.0
o.ie c.o
0.75 ".0
0.48 C.3
C.12 C.O
0.52 O.C
C.09 0.0
O.lt 0.0
C.lt O.C
0.1C C.O
C.2C C.C
C.lt O.C
C.15 O.C
o.ie o.c
0.29 C.C
0.19 O.C
C.ll C.O
C.C5 0.0
C.19 0.0
1.54 0.0
O.ft C.3
0.91 C.O
1.C1 C.C
C.12 C.C
C.5C C.O
C.72 0.0
0.01 C.O
0.5C 6).«
".44 ) . 8
88.54 5.9
0.47 (7.1
4OtmnS 4 UKM.
anui
IKKI
O.f 11. » 45.2
C.4 36.2 1
O.t 11. t 1
O.f 15.1 1
1.1 15.1 4
0.5 15.1 <
l.C 11. t 4
1.1 29. t I
C.f 11. t 4
l.t 25. t (
1.1 21.4 1
1.9 ie.4
0.8 18.2 1
0.9 40. t 4
0.4 If. 9
1.2 if. a i
1.2
4.1
4.2
7.2
5.2
4.t
2.1
4.t
B.7
5.f
SI. 2
8.4
5.5
>C.O
t.t
C.f It.f 46.6
c.c ir.i i
C.f If. 5 <
O.t 17.0
5.6
17.5
if .e
0.8 Jl.C 45.!
C.4 17.1 47.2
l.C 15.4
1.0 18.9
l.C J?.4
C.5 4C.O
0.4 38. t
1.0 15.0
C.5 38.1
C.4 43. f
l.C lt.1
C.C 0.0
C.C 0.0
C.O 0.0
C.O 0.0
C.O 0.0
16. C
16.9
BC.l
15.5
16. t
14.9
15.2
If .7
56.4
18. f
It.l
If. 2
• 9.7
43.6
o.c o.o <«.:
C.C 0.0 5C.C
C.C C.O 52. t
c.c c.c
IT.9
c.c c.o ;i.j
c.c c.c
c.c c.c
c.: c.c
c.c c.c
c.c r.c
C.O C.O
c.r r.o
r.c o.o
c.c o.c
c.c c.c
c.c r..c
c.c r.c
c.c o.c
C.O l.C
c.r r.c
c.c o.o
c.c c.c
C.C C.3
c.c c.c
C.C 0.0
C. C.O
C. C.O
c. c.c
C. 0.0
c. o.c
c. c.c
C. 0.0
c. c.c
C. C.O
c. c.c
c. c.c
C. 0.0 4
C . C.O (
ie.:
15.4
((.<
15. «
If. 2
17. t
5C.4
15.2
15.2
i . \
?.:
1.2
e.i
2.2
1.2
C.2
e.c
7.T
c.C
1.4
4.2
!••
7. ;
i.l
3.5
6.1
1.1
6.(
6.4
e.4
«.r
5.7
'. .C
C. 0.0 44.7
C. 15.5 4T.3
0. 4.0
(3.) 11.2
2, (
5.6
C.I 12. ( 45.5
Tests 1 to 80 are fuels used for Phase 1 only.
AEL designates the control fuel used in Phase 1
and Phase 3.
133
-------
TABLE C-2: Test Fuel Distillation Profile by ASTM D86-67.
TEST
10
11
12
11
14
15
16
17
18
10
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
K
36
37
38
30
40
41
42
41
44
45
46
47
4r
4«
50
51
52
53
54
55
56
57
56
Kit
60
61
(2
(3
64
C
((
(7
(8
60
70
71
72
73
74
75
76
77
7P
7?
PO
• vc
tc
cv
M 1
•OIL rr.
260
244
2P3
279
272
295
291
233
332
348
356
376
350
35E
365
3«!6
358
374
378
164
357
334
383
36 «
381
370
Itl
154
350
381
378
3(4
370
376
3(5
358
338
352
354
328
341
363
165
154
350
34P
350
354
3*3
31 1
357
375
318
264
358
366
362
356
365
372
366
358
360
398
358
358
356
36 C
346
3(6
334
360
352
346
362
346
342
342
343
46
13
336
t ft
368
366
3(C
370
355
3(2
366
395
291
198
382
388
414
0
402
399
400
3«9
4C6
412
396
385
386
406
390
405
394
410
392
394
406
406
392
400
411
394
380
355
389
386
384
172
370
360
320
400
379
>7<
383
391
306
392
393
40(
384
366
400
382
404
390
404
406
406
398
)98
420
392
358
390
390
376
392
370
314
376
370
380
364
360
380
365
31
8
0
10 ft
414
400
357
J97
306
3D4
390
422
361
421
309
403
433
302
420
414
416
415
420
430
414
412
412
421
416
421
414
437
417
420
421
426
418
426
429
412
418
375
412
405
406
248
387
407
370
430
397
350
395
413
416
411
411
425
'10
400
424
41P
420
412
422
424
424
410
420
44(
420
416
412
410
400
412
396
402
398
386
392
300
398
396
406
24
6
398
JO ft
44P
44P
421
424
437
426
416
450
407
451
416
420
455
416
448
434
435
436
440
450
416
442
434
437
444
439
441
478
430
442
•31
442
444
456
452
416
429
390
438
424
430
370
402
429
410
458
420
420
42C
442
445
438
434
447
444
438
446
444
444
440
4<6
450
450
444
448
4(2
446
442
438
436
426
436
416
428
424
402
410
410
41P
418
434
17
4
423
10 ft
46:
45!>
445
446
470
446
436
464
425
472
435
436
476
433
465
450
449
456
456
466
452
462
451
457
464
.55
4tO
495
457
460
455
4(0
463
460
470
455
444
406
456
440
450
406
416
446
0
480
440
441
447
Iti
466
458
453
4f '.
4(3
4C2
466
470
1(1
462
470
469
470
468
468
462
465
462
460
454
446
454
434
446
442
416
426
428
436
436
454
17
4
442
40 ft
485
4(4
463
464
494
467
452
478
443
490
454
451
495
450
483
4C9
464
472
472
486
470
481
468
473
48 (
471
480
507
472
(76
469
476
482
500
489
477
460
420
477
456
472
435
lit
466
0
500
455
• CC,
467
421
4E4
476
472
4(2
464
486
484
4PB
482
482
490
488
468
490
48t
510
490
460
480
470
466
474
452
466
462
426
440
444
454
454
472
18
4
4C1
SO ft
St4
401
463
481
511
486
466
496
462
509
472
4(8
514
467
504
465
479
469
490
504
488
SCO
484
4P7
502
487
496
520
469
401
464
495
499
524
5C7
495
47(
432
495
471
466
454
440
483
472
520
460
461
488
5(10
503
4C6
401
499
504
506
502
5(18
502
502
508
5C4
SC6
504
506
526
50C
496
500
490
484
404
470
4(6
482
442
456
462
47:
4)2
490
19
4
463
(0 ft
522
510
SCI
SCO
510
506
486
511
464
528
492
464
534
466
521
503
495
507
508
524
506
520
506
504
522
504
519
54«
507
507
502
509
517
S42
526
513
406
444
513
480
506
472
452
504
500
540
5CC
5CT
SIC
51°
522
515
Sjo
51"
524
52P
522
530
520
522
530
512
526
524
528
548
520
516
520
SC6
506
514
490
504
504
456
476
4PO
4«4
454
SO?
19
4
501
70 •
541
530
523
520
549
527
506
532
504
55C
514
502
555
510
546
522
515
52 o
526
544
528
542
525
523
544
525
541
572
527
510
522
511
537
5(5
547
536
518
460
516
510
510
• ft
467
526
526
5(4
525
5:6
534
540 •
544
538
534
536
547
550
544
55:
544
54k
522
540
546
544
S50
570
540
536
542
526
524
536
510
530
526
472
516
502
516
516
530
20
4
522
M ft
565
555
539
545
571
544
532
558
538
578
541
526
577
53,6
573
540
539
555
552
570
556
570
551
sso
574
SSO
569
(03
553
557
550
55E
555
591
572
561
546
4PO
56 0
537
556
526
487
555
557
590
55C
5 50
563
56 6
568
see
5f2
56?
5-3
SPC
570
5EO
574
S7P
582
5f2
562
570
578
564
567
562
56E
554
SS2
562
542
556
556
494
524
52f
546
546
556
21
4
546
to t
554
5»9
585
581
601
594
570
594
562
(K
5(2
569
608
576
609
589
574
593
590
(08
598
(08
592
588
(06
5E6
(07
(22
594
590
586
59C
596
625
607
603
592
512
597
581
596
575
518
593
606
(29
(02
557
(01
60!
605
611
604
592
(OP
616
610
622
616
620
624
598
608
(06
(18
(22
(05
(04
(06
596
594
600
592
596
596
528
562
572
59f
598
595
21
3
565
tS ft
(16
(06
(07
(10
(27
(24
(01
(22
(19
(49
(K
(09
(34
(10
(42
624
(04
(28
(18
(36
(30
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
c
0
(38
0
(34
CI7
(35
(44
«1
631
640
654
P
(50
(42
656
0
618
(42
(36
(56
(60
(46
(4E
(42
(42
626
630
(16
632
624
566
CC
616
628
628
(30
17
3
(00
BID IT.
(IS
61C
(07
(1!
(27
628
(07
(10
(30
((8
(51
(44
(56
(45
(6C
(53
641
65(1
650
664
(60
646
(38
(51
(49
(51
(49
(56
626
(41
(48
651
638
654
(54
0
(40
570
(45
(30
(40
(42
(00
(56
(56
674
656
(50
654
(54
655
650
654
650
(57
(56
632
650
662
658
6(5
646
662
656
666
(60
tie
(52
(54
(54
(40
(46
0
654
(50
S-.6
(14
(10
636
636
(44
ie
3
(15
Reproduced from
best available copy.
ft UC.
16. »
91.6
»6. C
9(. 1
96.1
96.7
97.1
«7.t
97.2
97.6
99.0
96.0
96. (
96.2
96.1
96.4
99. C
98.4
96.0
• 8.0
97.8
95.5
96.0
96.0
96.5
sc. s
96.5
96.0
96.0
96.0
98.0
98.0
97.5
96.5
97.4
97.2
97.4
87.5
97.5
97.5
97.0
97.4
96.5
96.6
96.0
E7.0
97.6
?E.O
57.2
57.1
97.4
*P.O
97.1
97l9
07. f
95.2
96.2
97.5
96.5
95.5
97.0
98.0
96.0
97.0
96.5
96.5
97.0
07.0
97.5
97.2
97.5
97. 5
97.5
97.5
97.0
97.2
97.5
97. C
97.0
97.2
1.4
1.4
96.8
€3
ft UI. ft LOS
3.C 0.
3.3 0.
3.0 C.
3.7 0.
2.9 1.
2.4 0.
2.8 0.
2.3 0.
2.7 0.
2.0' 0.
1.0 0.0
1.7 0.3
1.2 0.2
1.8 0.0
1.5 O.C
1.5 0.1
1.0 O.C
1.5 0.1
1.8 0.2
1.5 0.5
1.3 0.5
4.5 0.0
4.0 0.0
2.0 0.0
3.5 O.C
1.5 0.0
3.5 O.C
1.5 0.0
4.0 O.C
2.0 0.0
2.0 0.0
2.0 0.0
2.5 0.0
3.5 0.0
2.5 0.1
2.6 0.2
2.4 0.2
2.5 0.0
2.5 0.0
2.5 0.0.
3.0 O.C
2.5 0.1
1.5 0.0
1.4 0.0
4.C 0.0
3.0 0.0
2.1 0.3
1.6 0.2
1.9 0.?
1.7 1.0
l.» 0.8
C.6 .4
1.5 .4
1.3 .P
1 • V * ^
1.5 .5
2.8 .0
2.5 .3
i.: .0
2.5 .0
2.C .5
2.C .C
1.5 .5
1.5 .5
2.0 .0
2.0 . 5
1.5 .0
2.0 .0
2.0 .0
1. i .0
1.6 .0
1.5 .0
1.5 -C
1.5 . C
1.5 .0
1.8 .2
1.8 . C
1.5 .0
2.0 .0
2.C .0
2.2 0.8
0.6 0.6
37.3 71.2
1.1 0.1
134
-------
APPENDIX D
EXPERIMENTAL PROCEDURES
This Section describes the experimental, analytical, and calibration procedures
used to obtain the emissions data. Procedures for the extract chemical identification
work are presented in Section 8.
VEHICLE EMISSIONS TESTING
Gaseous emissions and fuel consumption were measured by the 1979 EPA proce-
dures simultaneously with particulate sampling by the 1981 EPA procedures. The
vehicle test bay was maintained at constant- temperature by two air conditioning/-
heating/humidification/dehumidification units with a combined capacity of 106 kw and
circulating ^25 m^/min conditioned air onto the vehicle positioned on the
dynamometer.
Dilution tunnel sampling has come into general use for measurement of diesel
paniculate mass emissions, and particulate samples collected by this method have also
been widely used for further particulate characterization. While the potential for
chemically reactive species to contribute artifacts in dilution tunnel filter collection
has been noted (64-67), dilution tunnel sample collection has become the accepted
choice in lieu of any proven sampling protedure. Recent work (68)-has provided some
• guidelines for dilution tunnel sampling to minimize the potential artifactual formation
of direct-acting mutagens during the sampling process due to the presence of NO2 in
the exhaust gas. Thus while all reasonable experimental care was exercised, it must
be noted that sampling procedures valid for emissions...measurement cannot be
extrapolated to be quantitative sampling procedures for all chemical species present.
Since the chemical identity of all emission products are not known, the degree to
which the emission chemistry interacts with sampling procedures cannot be quantita-
tively evaluated.
Particulate Emissions Measurement
Collection Apparatus —
Particulate emissions were measured according to 1981 EPA procedures (69) for
new vehicle certification using a positive displacement pump (POP) for air dilution of
the vehicle exhaust, followed by particulate sampling in a dilution tunnel. To obtain
samples of sufficient size for subsequent chemical and bioassay analyses from short
vehicle tests, a scaled-up particulate sampling system was added to the dilution
tunnel. Figure D-l is a schematic of the exhaust dilution and particulate sampling
equipment. Participate mass emissions for vehicle tests were determined .by propor-
tional sampling of dilute exhaust by either of two identical it? mm filter systems.
135
-------
CAS HITCH
PUMP
2 SYSTEMS
BOOTS
FLOW METE*
DILUTION AIR FROM
AiR CONDITIONER
Y
DISPLAYS
ACCUMULATION
AND RATE
VARIABLE
SPEED DRlvt
28'ZB"'/-."
VEHICLE EXHAUST
INLET
\
SOLENOIDS ACTUATED BT
CVS BAG SWITCHES
~~ DUMyv
FILTER
MIXING BAFFLE
PARTICULATE DILUTION TUNNEL
Figure D-l. Schematic of Particulate Collection Apparatus.
Each filter could be isolated from the probe and dilution tunnel by smooth-bore ball
valves to permit changing filters during vehicle testing. Each kl mm filter was
equipped with an identical 47 mm backup filter immediately downstream of the
primary filter. Constant sampling flow was maintained as particulate was progres-
sively deposited on the filter by means of mass flow controllers (Tylart Model FC-202)
capable of maintaining constant mass flow over a wide range of gas temperatures and
pressures. These flow controllers operated from a common power supply and display
panel, and had individually adjustable mass flow rate set-points. These independent
probe systems were flow calibrated each week with a Meriam laminar flow element
(//50 M310).
The particulate sampling system for obtaining bulk samples for extraction used a
filter assembly developed by EPA employing a 50 cm x 50 cm (20 in x 20 in) square
filter. A nominal flow of 2.8 m^/min (100 cfm) was drawn through this filter to yield
0.5 to 2 g of particulate from typical vehicle tests. The dilute exhaust passing through
the 50 cm x 50 cm filter was measured by a Roots-type totalizing flow meter (Dresser
Industries Model 5M125) mounted on the vent side of the blower used to draw sample
through the filter (Figure D-l). The volume filtered through the 50 cm x 50 cm system
during emissions testing was obtained from a digital accumulator on the Roots meter
actuated in synchronization with other sampling devices by the CVS bag fill switches.
Continuously recorded temperatures at the Roots meter exit were used to correct the
volumetric data to standard conditions. Thus, routine mass emission tests could be
performed simultaneously with bulk particulate collection for further chemical
characterization. Propane tests with both CVS and 50 cm x 50 cm filter operating
verified the flow measurement calibration of the complete system.
136
-------
The filter used in both the 47 mm and 50 cm x 50 cm filter systems were Pallflex
Products Corporation Type T60A20 teflon coated glass fiber. The nominal flows of the
47 mm and 50 cm x 50 cm systems were 9 1/min and 2.26 x 10^ 1/min respectively,
referenced to 0°C. Dilute exhaust flows per area of filter exposure were thus 0.014
l/sec/cm2 and 0.018 l/sec/cm^ for the 47 mm and 50cm x 50 cm filters respectively;
both within the EPA specified range for particulate measurement (0.00951 > flow >
0.0272 l/sec/cm2 at 0°C). Operation of the 50 cm x 50 cm filter in conjuction with
the CVS required the 50 cm x 50cm system to be started before the CVS to avoid
filter tearing. This operational constraint made it impossible to select only a portion
of any given driving cycle for bulk sample collection since the filter flow had to begin
before CVS start-up and extend past CVS shut-down.
The maximum allowable temperature of dilute exhaust for particulate sampling
by EPA procedure is 52°C; and this constraint directly affects the volume of dilution
air necessary to cool the vehicle exhaust prior to sampling. Sampling requirements for
the variety of vehicles, engine sizes, inertia weights, and vehicle operating modes
encountered in this study were best met by using a range of dilution air flows. The
need to obtain measurable levels of gaseous species simultaneously with particulate
sampling prevented setting the dilution air flow at one high value to satisfy only the
particulate temperature limit sampling requirements. To accommodate these needs,
an existing POP - CVS was modified by addition of a 7.5 kw variable frequency three-
phase motor controller to give a variable speed motor drive providing continuous
adjustment of CVS flow between 1.5-15 m^/min (50-525 cfm). In combination with the
2.8 m^/min (100 cfm) bulk particulate sampler flow and the conditioned-air ducted to
the dilution air filter box, the range of CVS flows available satisfied the 52°C
maximum temperature requirement for all vehicle test conditions. Combined CVS and
50 cm x 50 cm filter flows have been consistently set to give a total dilute exhaust
flow of 11.4 m^/min (400 cfm) for Volkswagen and 17 m^/min (600 cfm) for Mercedes-
Benz and General Motors vehicles respectively. Strip' chart recordings -of dilute
exhaust temperatures were obtained for all tests^ -The PDP-CVS blower and Roots
meter for measurement of 50 cm x 50 cm filter flow were mass flow calibrated as a
function of blower speed, using turbine flow meter standards from Autotronics Corp.
(Models 100P-750-MNYNV and 100-450-SFN6-8, respectively.
Connection between the vehicle tailpipe and the dilution tunnel was either by a
short length (approximately 1 m) of 7.6 cm (3 in) uninsulated flexible exhaust tube for
vehicles with tailpipe exhaust near the tunnel (passenger side), or by a longer
conniictor to vehicles with a tailpipe connector on the side of the vehicle opposite the
tunnel location (driver side). Two types of connector were utilized in this latter case.
Through the first 17 vehicle tests, this connector was of the same type as above, but of
sufficient length (4 m) to reach the tailpipe. For subsequent vehicle tests a smooth
stainless steel tube (7.6 cm x 3 m), lagged with 5 cm of fiberglass insulation, was used
to link the tunnel to the tailpipe with short flexible uninsulated sections to make the
connections. Particulate emission tests with the insulated and uninsulated systems
showed approximately a 5% mass increase for the insulated case. EPA has also
reported (70) on the effect of tailpipe hookup on observed particulate emissions.
After nine months of daily testing the dilution tunnel was carefully cleaned.
From the weight of tunnel scrapings and an estimate of the total mass of particulate
entering the tunnel throughout the course of testing, the particulate losses in the
tunnel were estimated to be 3%.
137
-------
Weighing of Filters—
-i'-»r
The 47 mm diameter Pallflex filters were placed in a constant humidity chamber
for at least 24 hours before each weighing. Chamber temperature was maintained at
18 to 21°C with a relative humidity range of 55 to 60%. The equilibrated filters
weights, in micrograms, were obtained on a Mettler ME 30 microbalance. The 50 cm x
50 cm filters were not environmentally equilibrated, but were stored in areas where
the temperature was usually about 20°C and the relative humidity was in the 30 to
50% range. Tare weights were taken by carefully rolling the filters into approximately
15 cm cylinders and placing them on a semicircular weighing pan in a Mettler H10T
balance. The balance doors were kept open to accommodate the length of the rolled
filters. The balance itself was in a large five-sided box to reduce air currents and the
weighings were performed in a balance room without air circulation. Gross weights
were taken after the filters had been folded into quarters and were performed in a
normal manner. Weights were taken to the nearest milligram with the uncertainty in
the net weight estimated to be about + 5 mg.
Extraction of Filters—
The 50 cm x 50 cm filters used for bulk particle collection were extracted with
dichloromethane to obtain the extract as a percentage of the particulate mass on the
50 cm x 50 cm filters. The filters were folded and placed in 50 mm Soxhlet extractors
without thimbles. Extraction was carried out with approximately 300 ml of dichloro-
methane (Burdick and Jackson Laboratories, Distilled in Glass™ grade) for 24 hours
at three to four cycles per hour.
The extracts were vacuum filtered into 500 ml flasks through an 0.2 pm
Fluoropore filter (Millipore FGLP) to remove any particles which may have carried
over. The solution was heated to boiling before filtering, and the filtering apparatus
was kept hot during filtration to prevent precipitation of any sparingly soluble
material on the filter. The 500 ml flasks were rotary evaporated to about 75 ml
volume and the sample transferred to acid and solvent washed, dried and tared 125 ml
borosilicate glass bottles with Teflon lined caps. These bottles were then partially
immersed in a warm water bath while attached to the rotary evaporator to remove the
remaining solvent to yield the dried extract. The bottles were maintained overnight at
room temperature and a partial vacuum of about 10 in Hg in order to insure complete
solvent removal. Final weighing provided, by difference, tlie mass of extract. Bottles
with dried extract were then stored at -80°C until needed for bioassay.
For each of tests 1 to 34 there were twenty-one 50 cm x 50 cm filters, 15 of
which were extracted and the remainder kept in -80°C storage. The six unextracted
filters were the CFDS and HFET filters from the afternoon portion of each phase. For
tests 35 to 80 all 12 filters were extracted.
Particulate Backup Filters-
Twelve vehicle tests have been conducted with backup filters for every 47 mm
filter collected to investigate their importance for the various vehicles, fuels, and
driving schedules. Average backup filter contribution to the combined mass (primary
plus backup) was 1.65% (o = 0.95) for 295 individual test.cycles (i.e. 295 separate
filters). The backup contribution for FTP testing was 2.66%~Ta = 1.6) for 96
observations. The group of filters with the highest backup contribution came from the
IDLE test mode, where 35 tests gave backup contributions that averaged 5.68%
138
-------
(a = 5.75). Prior to the time of this finding for the IDLE test, the CVS flow setting had
been left at a constant value for all cycles for a given vehicle. This resulted in the
IDLE test sampling being conducted at a significantly higher exhaust dilution ratio
than the other cycles used. Backup filters irom IDLEs under these conditions
frequently gave higher absolute mass loadings thun for those for other (higher speed)
driving cycles. The collection efficiency for filter media is known to be increased by
previously deposited paniculate on the filter. With the very dilute exhaust of the
IDLE the primary filter never received sufficient mass loading for this efficiency
enhancement to occur. A subsequent decision to decrease the dilution ratio for IDLEs
increased the primary filter loadings and decreased the backup filter loading.
In contrast to the general findings that backup filters collected a small per-
centage of the total paniculate, tests on a Peugeot (Car #17) exhibited distinctly
different results. Particulate from this vehicle did not mat on the face of the filter
media but penetrated to give a grey appearance on both the front and back of the
filter. Backup filters on 30 Peugeot test cycles averaged 9.4% of the total participate
mass. In 7 Peugeot tests, two 50 cm x 50 cm sheets were used in series in the same
holder to effect a filter backup for the bulk sampler. The second 50 cm x 5C cm filter
from these tests contained an average of 9.5% of the total. This observation parallels
that of the EPA for a Peugeot vehicle (71, 72). Backup filters have been discontinued
for all vehicles except the Peugeot, but variations in sampling conditions and vehicle
type can impact the need for backup filters.
Particulate Emissions Reproducibility—
Within each test phase of the first 34 tests, the duplicate CFDS and HFET cycles
performed in the afternoon and again the following morning, have been compared to
determine the stability of particulate emissions during the given phase. Table (D-l)
shows the reproducibility for these two driving cycles for each vehicle test phase for
all 34 tests and also by vehicle make subcategory. On the average, the reproducibility
is in the 1-3% range, but the standard deviations for the sample are in the 5-15% range.
The Volkswagens show a larger standard deviation than other groups, possibly due to
their manual transmissions in contrast to the automatic transmissions on most other
sample group vehicles.
TABLE D-l. PARTICULATE EMISSIONS REPRODUCIBILITY FOR DUPLICATE CYCLES*
Ditterence Between Duplicate Cycles «1 a
OBS. PHASE 1 PHASEl PHASE 3 THREE PHASE
IN EACH AVERAGE
AVERAGE CFDS HFET CFDS HFET CFDS HFET CFDS HFET
ALL TESTS 34 1.0%» 8.8 1.2%«I0.6 O.l%.10.2 1.8%±8.l I.2%« 9.2 0.8%.S.O 0.7% 1.3%
GENERAL MOTORS 16 0.5%* 73 2.7%. 9.0 -1.2%. 7.5 0.0%±6.3 0.6%. 5.8 l.3%.5.5 0.8% 1.7*
VOLKSWAGEN 10 I.0%«t2.3 I.8%.I«.9 3.5%«I5.7 0.6%«/.2 2.9%«I3.« 3.7%«9.2 2.5% 2.0%
MERCEDES-BENZ » 1.8%. 8.3 -5.7%* 7.9 -0.3%. 5.6 -3.6%»3.« 3.1%. 8.8 0.2%.2.0 1.7% -3.2%
•First 31 tests only.
Note: % > 0 indicates first cycle > second cycle.
% < 0 indicates first cycle < second cycle.
139
-------
Filter Pair Correspondence—
The two independent 47 mm diameter filter collection systems were routinely
operated simultaneously during all CFDS cycles after vehicle test 14 as an ongoing
check of correspondence and data integrity for these systems. For 126 CFDS tests
with duplicate particulate measurement by both 47 mm systems, the average ratio of
the two results was 0.997+5.4%, indicating acceptable performance of this system.
The 50 cm x 50 cm bulk particulate collection system was designed to obtain
significant quantities of particulate for chemical and biological analyses, and as such
was not designed to insure strict gravimetric correspondence .with the 47 mm diameter
filter system. However, the measurements taken to provide a particulate emission
value can be compared to the 47 mm value used for reported vehicle emissions. While
this comparison could be made on all test cycles, here we present the comparison for
only the CFDS tests. The mean absolute difference of the 50 cm x 50 cm particulate
emission rate referenced to the 47 mm system for 157 CFDS tests was -0;04 g/mi,
indicating a small but systematic lower collection rate by the bulk particulate system.
Expressed in terms of a ratio of the 50 cm x 50 cm to 47 mm system, this becomes
0.912±6.5%. _
Gaseous Emissions Measurement
Analytical Instrumentation-
Analytical instruments for measurement of gaseous emission components were:
Gas Method Instrument
CO2 NDIR Horiba A1A21
CO NDIR Horiba AIA21AS
NOX Chemiluminescence Beckman 951
HC-background Unheated FID Scott 215
HC-continuous dilute Heated FID Horiba IX
Zero and span checks were performed before each test cycle with gases which were
referenced to primary standards five times during the course of this project.
The heated FID for dilute exhaust HC measurements used a 1/4" diameter
stainless steel heated probe, followed by a heated filter and heated transfer line to the
detector unit. The probe system was fitted with an "overflow zero gas/span gas"
system to permit instrument calibration through the heated line input. The sample
pump and FID burner assembly were located in an oven. All sample lines, filter, and
oven were maintained at 191°C. Thermocouples with digital readouts were used to
monitor all of these temperatures. The sample probe was located just off the
centerline of the dilution tumel (near to the particulate probes) approximately 8
tunnel diameters downstream from the exhaust inlet. Data from the heated FID were
recorded on an integrating strip chart for visual display and electronically integrated
and printed for each vehicle test cycle through, circuit controls on the CVS console.
Other gaseous analyses were performed ;n a console, which contained sample
handling, detectors, strip chart recorders and operator controlled zero/span devices.
All sample handling equipment was either Teflon or stainless steel. Individual
instrument sample flew was monitored by a flowmeter and regulated in the instrument
140
-------
inlet line. All instruments operated with positive pressure at the inlet ports.
Instruments were each individually vented to a common non-pressurized exhaust
system. Dilute exhaust from the CVS reported teflon bag samples was dried and
filtered upon entry to the analytical bench, and filtered once more upon entry to each
particular instrument.
All zero, span, sample, and calibration outputs from the instruments were
continuously recorded for each instrument. The daily schedule of instrument operation
began with a zero and span check before Bag 1 of the FTP and again in Bag 2. They
were subsequently zeroed/spanned before alternate test cycles throughout the day.
Each instrument main span check was followed by a second span gas check on a
different instrument range to check range-to-range correspondence.
Instrument Calibration
Gaseous instrumentation was calibrated on a monthly basis using laboratory gas
standards referenced to NBS. A Horiba gas divider (Model S-TEC, SGD-78) was used
to provide various calibration gas blends on which the instrument calibration was
fixed. The span gases for daily instrument checks were then named as part of the
calibration process.
Analysis instrumentation was also checked by participation in a subscription
cross-reference service of Scott Research. Seven cross-reference tests were per-
formed during the project, and the results of AEL determinations compared to results
reported by Scott Research are tabulated below as full-scale percentage error:
Cross-reference, % Error
HC 0.17
CO -1.02
C02 -1.19
NO -2.1
-------
Table D-2. Propane Injection Test Results
LOW SPEED HIGH SPEED
Classification (315cfm) (508 cfm) Composite
Observed Percentages
Total No. of Tests 28 72 100
% Error > 0% 53 34 39
% Error £0% 47 66 61
% Error > +2% 7 66
% Error < -2% 1 23 17
-2% < % Error < +2% 91 71 77
Thus 71% and 91% of the respective high and low blower speed propane tests were
within ^2% recovery limits giving a composite project value of 77%.
DYNAMOMETER
The chassis dynamometer used throughout the project was a Clayton CTE50,
equipped with automatic road-load and direct-drive variable inertia (250 Ib increments
to 3,000 Ib, 500 Ib increments above 3,000 Ib) features. The dynamometer was warmed
up each day before testing with a 15 min 50 mph cruise, during which time the road-
load setting for that test was established and the drivers aid recorder calibration
checked. Three periodic complete dynamometer calibrations were performed during
this project by a coast-down procedure. Coast-downs were recorded on a strip-chart
recorder and timed to give four load points for each inertia setting. This calibration
was stored as a linear regression to permit load/inertia specifications to be readily
selected and checked.
1N-USE VEHICLE HANDLING
Upon delivery to AE1, for testing, an in-use vehicle was first checked for
sufficient tank fuel to complete the "as-received" portion of testing. If insufficient,
additional commercial diesel fuel was added, usually 5-10 gallons from a local vendor.
A one liter sample of tank fuel sample was then pumped from the tank and sent out for
fuel analysis. An additional fuel sample was also taken for project archives.
Engine oil and transmission level were then checked to verify acceptance for
testing. The drive tires were removed and replaced with AEL tires of vehicle
manufacturer specification inflated to 45 psig. This procedure guarded against
unnecessary wear on vehicle owners tires during the extensive dynamometer test
(about 200 miles) and aided project results by providing a uniform test tire throughout
the length of the program.
After completion of Phase 1 the engine oil was drained and a sample taken for
project reference. The oil filter was changed ar,*. the crankcase refilled with oil
(Castrol) of proper specifications. The oils used we.
142
-------
Summer Winter
General Motors 30W 10W30
Volkswagen 20W50 10W30
Mercedes-Benz 20W50 10W30
At the time of oil change, the fuel was also changed to AEL control fuel. A portable
auxiliary fuel tank was connected to the injector pump inlet. The vehicle was then run
for 15 rnin at 50 mph to purge the fuel injection system back into the vehicle tank
through the return line. The return line was then connected tc the auxiliary tank to
complete the switch over to control fuel before the next portion of testing.
-------
APPENDIX E
EMISSION AND BIOACTIVITY CORRELATIONS
TABLE E-l. FTP Cycle - GM Vehicle Group - All Phases
TABLE E-2. FTP Cycle - VW Vehicle Group - All Phases
TABLE E-3. FTP Cycle - MB Vehicle Group - All Phases
TABLE E-*. FTP Cycle - "Other" Vehicle Group - All Phases
TABLE E-5. FTP Cycle - "All" Vehicle Group - All Phases
TABLE E-6. HFET Cycle - GM Vehicle Group - All Phases
TABLE E-7. HFET Cycle - VW Vehicle Group - All Phases
TABLE E-8. HFET Cycle - MB Vehicle Group - All Phases
TABLE E-9. HFET Cycle - "Other" Vehicle Group - All Phases
TABLE E-10. HFET Cycl 2-"All" Vehicle Group-All Phases
TABLE E-ll. CFDS Cycle - GM Vehicle Group - All Phases
TABLE E-l2. CFDS Cycle - VW Vehicle Group - All Phases
TABLE E-l3. CFDS Cycle - MB Vehicle Group - All Phases
TABLE E-l*. CFDS Cycle - "Oth-jr" Vehicle Group - All Phases
TABLE E-l5. CFDS Cycle - "All" Vehicle Group - All Phases
TABLE E-l6. New York City Cycle - GM Vehicle Group - All Phases
TABLE E-17. New York City Cycle - VW Vehicle Group - All Phases
TABLE E-l8. New York City Cycle - MB Vehicle Group - All Phases - "
TABLE E-19. New York City Cycle -"Other" Vehicle Group - All Phases
TABLE E-20. New York City Cycle - "All" Vehicle Group - All Phases
TABLE E-21. IDLE Cycle - GM Vehicle Group - All Phases
TABLE E-22. IDLE Cycle - VW Vehicle Group - All Phases .
TABLE E-23. IDLE Cycle - MB Vehicle Group - All Phases
TABLE E-2*. IDLE Cycle - "Other" Vehicle Group - All Phases
TABLE E-25. IDLE Cycle - "All Vehicle Group - All Phases
144
-------
TABLE E-l. Emission and Bioactivity Correlations
FTP Cycle - GM Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
E/MLE G/HILE G/KB-F Z G/KG-F RATIO PER HI U6 PART UG EXT U6 RES PER K6-F
N 74 74 72 74 72 74 72 72 72 72 70
HEAN .2170 .6707 1.142? 23.5129 3.6168 3.9188 6.0317 .7111 3.6226 .9251 32.5424
STU IiEV < .1376 .1367 .6779 10.0183 .6769 1.7440 2.5353 .3461 2.2806 .«171 14.1181
KSD Z 61.0821 20.3855 59.5171 42.6078 13.7156 M.5023 42.0335 48.6719 62.9533 45.0816 43.3569
CORRELATION HATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT lOCr REV REV PER REV PER REV PER 100K PtV
6/KILE G/KG-F Z G/KG-F RATIO PER HI UG PART US EXT UG RES PER KG-F
FXTRAfT H 72 74
e/Hiir .011 .997 .938
G/H"C SL= 0 Sf.« 3 SL= 3
RESIDUE . 72 . 74
G/HILE .UUi •A7o
SL= 0 SL= 2
EXTRACT Q7i
6/KG-f SLJ?3J
KEY:
EXTRACT
«. OF DATA PAIRS Z
SAfPLE CORR. COEFF.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL = 1 FOR 0.05>ALP«AX).01
SL = 2 FOR 0.01iALPHA>0.005
SL = 3 FOR 0.005}ALPHA
72 74
-.022 -.788
SL= 0 SL= 3
72 74
.971 .395
SL= 3 SL= 3
72 72
-.040 -.793
SL= 0 SL= 3
72 74
-.322 -.908
SL= 3 SL= 3
RESIDUE 470|
e/Kf^C »*vO
U/Aw • Cl ™ T
SL- 0
RES/EXT
RATIO
72
-.121
SL= 0
72
.052
SL= 0
70
-.122
SL= 0
72
-.125
SL= 0
70
.059
SL= 0
72
.063
SL= 0
100K REV
PER MI
72
-.332
SL= 3
72
-.236
SL= 1
70
-.318
SL= 3
72
-.246
SL= 1
70
-.210
SL= 1
72
.129
SL= 0
72
.926
SL= 3
REV PER
UG PART
72
-.567
SL= 3
72
-.011
SL= 0
70
-.554
SL= 3
72
-.568
SL= 3
70
.023
SL= 0
72
.545
SL= 3
72
.795
SL= 3
72
.877
SL= 3
REV PER
IIC CVT
Ifl tAI
72
-.124
SL= 0
72
-.318
SL= 3
70
-.122
SL= 0
72
-.013
SL= 0
70
-.294
SL= 2
72
-.089
SL= 0
72
.924
SL= 3
72
.970
SL= 3
72
.752
SL= 3
REV PER
UGttS
70
-.139
SL= 0
70
.008
SL= 0
70
-.139
SL= 0
70
-.130
SL= 0
70
.045
SL= C
70
.062
SL= 0
70
.99?
SL= 3
70
.941
Sl= 3
70
.853
SL= 3
70
SL= 3
145
-------
TABLE E-2. Emission and Bioactivity Correlations
FTP Cycle - VW Vehicle Group - All Phases
EXTRACT PES;DVE
{/«ILt
EXTRACT EXTRACT RESjtuE RES/EXI ioo« «v REV PEI RJV n* REV PEI ipo* «i
I 6A&-T RATIO KR«l iX PMT US ElT US RES «» l6-
«i
47 47 47 4B 47 47 45 «6 47 45 4)
.0741 .2957 .8*^5 20.6*72 3.3890 4.4715 4.6690 1.TO1 10.39?! 2.2?47 76.54:9
L -027} .Oe«» .3150 7.JJJ4 .7*17 J.73 0
45
.03?
SI* 0
45
-.020
Si* 0
45
.125
SI* 0
45
.106
Si* 0
45
.922
SI* 3
45
.985
SI* 3
45
.813
SI* 3
G/nliE G/KG-f
EXTRACT 4? 47
6/nllE
Si* 3 Si'-* 1 SO
t 6/KG-f
47 48
666 .758
« 3 Sl« 3
RES'VJE 47 47 47
G/*iif .«? .994 .783
KEY:
NO. OF
SAifiE
SI* 0 SL«
EXTR6CT
G/KG-f
EXTRACT
CATA r"F.
3 SL« 3
47 47
040 -.534
< 0 Si* 3
«7
.735
SL» 3
SIGNIFICANCE LC.tL
: - i
Si * J
IC0,0!
FOR o'.c:>iPto'.oo5
FOR O.OC5.;Ai.r*A
RESIDtC
RATIO
47
-.279
Sl« 1
47
SL« '0
4/
.987
Si* 3
47
.060
Sl« 0
47
-.529
fi •
RES/ET7
RATIO
100K
PER
PER "I
47
-.644
Si* 3
47
-.768
Si* 3
47
.543
Sl« 3
47
-.770
SI* 3
47
-.944
SI* 3
^47
Sl*"°3
REV
HI
REV
UG
UG PAST
»5
-.038
SL* c
45
.017
Sl« 0
45
.407
SI* 3
45
.033
SI* 0
45
-.225
Si* 0
45
.433
SI* 3
45
.279
SI* 1
PER
PAkT
REV
U6 1
UGEXT
46
.018
Si* 0
45
-.105
Sl« 0
45
.175
Si* 0
45
..JCJ
SI* 0
46
-.154
Si* 0
45
.K3
Sl« 0
45
.SO
SI* 1
45
.945
51* 3
PER
mi
REV
ItC
16 RES
47
-.265
Si* 1
45
-.356
Sl« 3
45
.331
Sl« 1
45
-.373
SI* 2
46
-.507
Si* 3
45
.315
SL« 2
45
.533
Si* 3
45
.871
SI* 3
46
.954
SI* 3
occ
-------
TABLE E-3. Emission and Bioactivity Correlations
FTP Cycle - MB Vehicle Group - All Phases
&f iwtti ijiwct Rtsjwt ws/jrt 100* SEV «fv Ptt «v KB «w f«
L£ &/H&-7 z s/ntrT KM fa ftfcuf udfA«i us tit U&HS
i/
» 2J 2» 29 21 20 21 19 19 20 19 18
*£*" .0790 .4549 ,55<» 15.1317 2.9909 4.9574 3.3034 .6159 5.33V5 .7111 2t.8!24
SU> !>£« .03«6 .0392 .28C,8 7.2J81 .4068 3.0782 1.6441 .i»l3 3.«r20 .3264 11.27tt
dSD Z 4). 8464 t9.6Ci» 51.5051 47.834] 13.6025 44.2441 50.3720 47.3033 45.4543 45.f320 51.4572
COMCLATION MTIIX
ecsipoe t*TR»ct nwftct «sitwe itts/tn io*lPvA>0.01 R£S/W 1* 19 i9 19 58
SL • 2 & 0 O^^ 005 MHO -3W .330 .719 .193 .242
a«3 FW o.'oc'5^h« a* o a* o a* 3 a* o a* o
100H ftftf A9 19 19 19
p£R Kl .954 .853 .932 .978
a* 3 a* 3 a« 3 a* 3
KV PfR 19 19 '8
UG PAil .SSI .987 .964
a* 3 a* 3 a* 3
StV P£R 19 18
US EXT -76* .7*8
a* 3 a* 3
U£ RtS .965
: SL» 3
147
-------
TABLE E-4. .Emission and Bioactivity Correlations
FTP Cycle - "Other" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100* REV REV PE« REV PER REV PER 100K REV
«/«Lc 6/MLT 6/K6-7 Z 6/K6-T RATIO KR M UG PART UG EXT U6 RES PER K6-f
N 35 35 34 34 34 35 32 33 33 32 31
•£* .2549 .3741 1.4254 39.4125 2.4355 1.9545 7.8413 1.1849 3.7545 2.1107 50.8091
ST[> LEV .1504 .1484 .7581 14.56C8 .9893 1.1892 5.3490 .7015 2.7582 1.5152 33.1915
fcSD Z 5816080 39.4561 46.4372 41.8069 40.6202 40.8442 48.4708 59.1058 73.4423 71.7881 45.3259
CORRELATION MTKIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100« REV REV PER REV PER REV PER ION REV
6/nlLE B/Mrf Z 6/*6-f RATIO PEK M UG PART UG EXT UG RES PER «H
rvrtATT 35 3*
tX'K'ftCl ^ ^_g -^
G/MLE s^* 0 ^ j
RESIDUE _^j4
L SL* 1
EXTRACT
KET:
NO. OF WTA PAIRS
SA4\E COS*. COIFF.
SIGNIFICANCE LEVEL •
SIGNIFICANCE LEI.B.S:
SL « 1 FOR 0.05>ALP»«X).01
SL * 2 FOR O.OI/ACP*A>O.OW
SL * 3 FOR 0.005JAOHA
35
.621
SL* 3
35
-.485
SL* 3
34
.847
SL* 3
EXTRACT
34
-.394
SL« I
34
.920
SL* 3
34
-.368
Si* 1
34
-.751
SL* 3
RESIDUE
6/K&-f
35
-.725
SL« 3
35
.497
SL* 3
34
-.808
SL* 3
35
-.952
SL* 3
34
.729
SL* 3
RES/EX1
> RATIO
32
.396
SL« 1
32
.229
SI* 0
31
.314
SL* 1
32
.074
SL* 0
31
.196
SL* 0
32
-.013
SL* o
10
-------
TABLE E-5. Emission and Bioactivity Correlations
FTP Cycle - "All" Vehicle Group - All Phases
SSGF KEF
EKF *W
177 173 179 173 177 168 170 172 168 164
.4873 1.0903 24.9990 3.2503 4.0381 6.2385 1.0848 5.6968 1.4748 46.9007
.6655 .13.2587 _ .873? 2.3097 4.0300 .. .8575 .5.6958 .1.2135 39.5901
7 »1
177
.1706
STD IiEVi .1337 .2018 .6655 13.2587 .8739 2.3097 4.0300 .8575 5.6?5B 1.2U5 1V.3V01
fcSD: 78.3824 41.4117 61.0331 53.0370 26.8871 57.1986 64.5*85 79.0421 99.9833 82.2820 84.4126
CORRELATION HATR1X
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
t/MLE 6/HG-f Z 6/KS-f RATIO PER HI U6 PAR! U6 EXT U6 RES PER K6-F
EXTRACT 177 173 177
S/niLE .218 .940 .808
SL* 3 SL* 3 SL* 3
RESIWE 173 177
G/HILE -.007 -.292
SL« 0 SL= 3
EXTRACT 173
6/KG-f "889
Sl= 3
KEY:
EXTRACT
«. OF DATA PAIRS x
SAtftE CORR. COEFF.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL « 1 FOR 0.05>AIPHA>0.01
Si * 2 FOR 0.01 JALPHA>0.005
SL « 3 FOR 0.005>Af'HA
173 177
-.177 -.656
SL« 2 SL« 3
173 177
.686 .223
SL« 3 SL« 3
173 173
-.21'3 -.762
SL* 3 SL* 3
173 177
-.587 -.839
SL* 3 SL* 3
RESIDUE 173
6AG-F .432
SL* 3
RES/EXT
RATIO
168
.159
SI* 1
168
.026
SL* 0
164
.175
SL* 1
168
.094
SL* 0
164
.152
SL* ,
1 *'.'*
-.df.v
SL- 0
100K REV
.
PER HI
168
-.219
SL* 3
168
-.383
SL* 3
164
-.102
SL* 0
170
-.067
SL* 0
164
.007
•> o
269
.050
£!.« 0
168
.793
SL* 3
REV PER
UG PAH
168
-.427
SL* 3
168
-.246
SC« 3
144
-.369
SL* 3
170
-.376
SL* 3
164
.206
SL* 3
168
.431
SL* 3
168
.604
SL* 3
170
.870
SL* 3
REV PER
U6OI
168
.062
SL* 0
168
-.430
SL* 3
164
.174
SL* 1
169
.238
SL» 3
164
-.169
SL* 2
168
-.166
SL* 1
U8
.838
SL* 3
168
.914
SL* 3
168
.661
SL* 3
REV PER
U6 RES
164
-.121
SI* 0
164
-.248
SL* 3
164
.001
SL* 0
164
-.058
SL* 0
164
.Ib9
SL* 2
164
.036
SL* 0
164
.K4
SL* 3
164
.952
SL* 3
164
.832
SL* 3
164
.873
bL* t
149
-------
TABLE E-6. Emission and Bioactivity Correlations
HFET Cycle - GM Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT
J/WLE 6/ML£ 6/KG-F
73
.1667
SU> KVi .1263
KSD I 75.7550
73
.2839
.1069
37.6568
73
1.3464
1.0164
75.4913
EXTRACT RESIDUE
I 6/KG-f
73
34.5636
1M505
41.6065
77
2.2925
.3228
35.6894
RES/EXT 10W REV REV PER REV PER REV PER 100K REV
RATIO PER HI UG PART U6 EXT U6 RES PER K6-f
73
2.3535
1.2355
52.4966
70
2.8685
1.7010
59.3005
70
.7049
.3807
54.0088
70
2.4129
1.5962
66.1529
70
1.0950
.5563
50.9866
70
23.1416
13.3765
57.8116
CORRELATION HATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100* REV RFV PER REV PER REV PER 100K REV
6/HILE 6/K6-F Z 6/KG-f RATIO PER HI US PART UG EXT UG RES PER KG-f
EX f ((ACT 73 73 73 73
G/MLE -604 -'Ml -415 '623
b lLt SL* 3 SL« 0 SL* 3 SL« 3
G/«LE SLs'(
EXTRACT
6/KG-f
KEY:
NO. OF DATA PAIRS
SAfPLE CORK. COEFF.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
73 73 73
)89 .998 .878
0 SL* 3 SL* 3
73 73
.064 -.318
SL* 0 SL* 3
"TT .865
1 SL* 3
RESIDUE
6/KG-F
73
-.083
SL* o
73
.068
SL* 0
73
.995
SL* 3
73
.049
SL* 0
73
-.337
SL= 3
SL « 1 FOR 0.05>ALPHA>0.01 RES/EXT
SL = 2 FOR O.CmiPwA.^.005 RATIO
SL * 3 FOR O-OOiiftLf-iiA
100K
PER
73
-.573
SL* 3
73
-.704
SL* 3
73
.365
SL* 3
73
-.707
SL* 3
73
-.903
SL* 3
73
.388
SL* 3
REV
HI
REV
UG
70
.168
SL* 0
70
.096
SL* 0
70
.147
SL* 0
70
.078
SL* 0
70
.077
SL* 0
70
.122
SL* 0
70
-.150
SL* 0
PER
PAkT
REV
lit
. Uv
70
-.150
SL= o
70
-.296
SL* 2
70
-.281
SL* 2
70
-.302
SL* 2
70
-.150
SL* 0
70
-.295
SL* 2
70
.008
SL* 0
70
.760
SL* 3
PER
CVT
tAI
REV
uG
70
-.331
SL* 3
70
-.514
SL* 3
70
-.117
SL* 0
70
-.515
SL* 3
-.489
SL* 3
70
-.117
SL* 0
70
.428
SL* 3
70
.576
SL* 3
70
ft"l*%
.872
SL* 3
PER
RES
70
.070
SI* 0
70
.089
SL* 0
70
-.397
SL* 3
70
.079
SL* 0
70
.269
SL* 1
70
.-.422
SL* 3
70
-.350
SL* 3
70
.779
SL* 3
70
ft JL ft
.901
SL* 3
70
g . ^
.617
SL* 3
150
-------
TABLE E-7. Emission and Bioactivity Correlations
HFET Cycle - VW Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER RJV Ptt 100K REV
6/ftlUE G/HlLT 6/KG-F I 6/K6-F RAffd PER HI UG PART US EXT U6 RES PERKG-F
N 46 46 46 47 46 46 45 45 46 45 45
KAN .0647 .2348 .9787 22.1757 3.5455 4.1434 5.0425 1.6962 9.6739 2.1392 76.2151
STD DEV I .0268 .0670 .4102 8.4923 1.0020 1.7616 4.0125 1.3367 8.8747 1.5760 59.830?
SSbX 41.4699 28.5412 41.9131 38.2954 28.2621 42.5157 79.5726 78.8057 91.7387 73.7682 78.5027
CORRELATION MATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/HILE 6/K6-F Z G/KG-f RATIO PER HI UG PART UG EXT UG RES PER HH
EXTRACT 46 46
G/HILE -.036 .992
SL= 0 Sl= 3
RESIDUE <*
G/HILE -.054
SL= 0
EXTRACT
6/K6-F
KEY:
46 46
.788 -.038
SL= 3 SL= 0
46 46
-.603 .983
SL= 3 SL= 3
4t 46
.789 -.033
SL= 3 SL= 0
EXTRACT 46
NO. OF DATA PAIRS
WfLE COftR. COEFc.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVaS:
SL * 1 FOR 0.05>ALPHA>0.01
SL = 2 FOR 0.01>ALPHA>0.005
St = 3 FOR 0.005>ALPHA
Z -.596
SL= 3
RESIDUE
**fc. v A Wk
46
-.719
Sl= 3
46
.653
SL= 3
46
-.721
SL= 3
46
-.938
SL= 3
46
.646
SL= 3
RES/EXT
RATIO
iftf
45
-.207
SL= 0
45
.414
SL= 3
45
-.212
SL= 0
45
-.391
SL= 3
45
.414
SL= 3
45
.414
SL= 3
x cm
AVWM llfc.*
PER HI
• ti* ni
45
-.299
SL= 1
45
.114
SL= 0
45
-.290
SL= 1
45
-.308
SL= 1
45
.129
SL= 0
45
.291
SL= 1
45
.922
SL= 3
REV PFR
IVtV TLH
IK: pan
w
< wtr\ >
REV
IIC
Uv
45
-.431
SL= 3
45
.313
SL= 1
45
-.428
SL= 3
45
-.513
SL= 3
45
.325
SL= 1
45
.537
SL= 3
45
.948
SL= 3
45
.954
SL= 5
PER
EXT
Wn 1
REV
UG
45
-.219
SL= 0
45
.038
SL* 0
45
-.208
SL= 0
45
-.201
SL= 0
45
.055
SL= 0
45
.189
SL= 0
45
.897
SL= 3
45
.993
SL= 3
45
t\ A ^
.917
SL= 3
PER
RES
45
-.221
SL= 0
45
.397
SL= 3
45
-.219
SL= 0
45
-.393
SL= 3
45
.408
SL= 3
45
.415
SL= 3
45
.997
SL= 3
45
.930
SL= 3
45
f^C 1
.952
SL= 3
45
.906
SL= 3
151
-------
TABLE E-8. Emission and Bioactivity Correlations
HFET Cycle - MB Vehicle Group - All Phases
N
HE AN
>Tli KV
f 1 1* 1 W V
RSDZ
EXTRACT
S/HILE
22
.0573
« .0320
55.8909
RESIDUE
6/H1LE
22
.3299
.0569
17.2457
EXTRACT
G/KG-F
22
.5298
.3419
64.5285
EXTRACT
22
14.9793
8.5686
57.2030
RESIDUE
6/KG-F
22
2.8884
.3517
12.1747
RES/EXT
RATIO
22
7.5872
3.7352
49.2306
100K REV
PER HI
20
1.9613
.9656
49.2323
REV PER
U6 PART
20
.4988
.2294
45.9930
KP
20
4.7150
2.9410
62.3765
REV PER
UG RES
20
.5716
.2494
43.6342
100KREV
PER K6-f
20
16.9145
8.3363
49.2849
CORRELATION HATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/HILE G/KG-F I G/KG-F RATIO PER HI UG PART UG EXT UG RES PER KG-F
EXTRACT 22 22 22
E/HILE
KEY:
NO. OF
SAim
-.549 .996 .976
SL: 3 SL= 3 SL= 3
RESIDUE 22 22
G/HILE Sl=593 Sl=^
EXTRACT 22
6/KG-F .V84
SL= 3
EXTRACT
DATA PAIRS Z
CORR. COtfF.
22
-.169
SL= 0
22
SL«
22
-.;»3
SL= 0
22
-.339
SL= 0
22
-.919
SL= 3
22
SL="93
22
-.928
SL= 3
22
-.937
SL= 3.
20
-.207
SL: 0
20
.468
Sl= 1
20
-.237
SL= 0
20
-.316
.51= 0
20
-.271
SL= 0
20
.309
SL= 0
20
-.289
SL: o
20
-.339
SL= 0
20
-.617
SL= 3
20
.412
SL= 1
20
-.636
SL= 3
20
-.659
SL= 3
20
-.132
SL= 0
20
.246
SL= 0
20
-.150
SL= 0
20
-.205
St.= 0
20
-.094
SL= 0
20
.381
SL= 1
20
-.113
SL= 0
20
-.198
SL= 0
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL= 1
SL = 2
SL = 3
FOR 0.05>ALPHA>0.01
FOR 0.01')ALPHA>0.005
FOR 0.005>ALPHA
RESIDUE
G/KG-F
22
.230
SL= 0
RES/EXT
RATIO
20
.455
SL= 1
20
.186
SL: 0
IOOKREV
PER HI
20
.265
SL= 0
20
.215
SL= 0
20
.972
SL= 3
REV PER
UG
PART
20
.124
SL= 0
20
.627
SL= 3
20
.794
SL= 3
20
SL=653
REV PER
U6
EXI
20
.266
SL= 0
20
.085
SL= 0
20
.968
SL= 3
20
SL=9893
20
.779
SL= 3
REV PER
U6
RES
20
.478
SL= 1
20
.053
SL: o
20
.952
SL= 3
20
.953
Sl= 3
20
.699
SL: 3
20
.969
CI 5 1
«JL"
-------
TABLE E-9. Emission and Bioactivity Correlations
HFET Cycle - "Other" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE
G/«IL£ 6/H1LE 6/HG-f X G/KG-F
RES/EXT 100K REV REV PER REV PEP. RfV PER 1QCK REV
RATIO PER Nl U6 PART UG EXT UG m PER OH
" 37 37 37 37 37 37 33 33 33 33 33
«AN .1586 .2108 1.4139 45.6840 1.9410 1.8384 4.9655 1.2695 4.2182 2.4515 46.3749
STJ rev .0723 .1206 .6372 21.0105 1.1951 1.5448 3.6992 .8516 4.1441 1.8843 39.40s8
RSD X "45.5895 57.2121 45.0684 45.9908 61.5696 84.0298 78.5262 67.0806 98.2434 76.8630 84.9787
CORRELATION MATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
8/MILE G/KG-F X 6/KG-F RATIO PER HI UG PART UG EXT UG RES PER KG-F
EXTRACT 37 37
G/MLE -.W .961
SL* 3 SL* 3
RESIDUE 37
G/hlLE -.478
SL= 3
EXTRACT
6/KG-F
37
.795
SL= 3
37
-.888
SL= 3
37
.765
SL= 3
KEY:
EXTRAC.T
NO. Cf DATA PAIRS
SAfftE CORR. COEFF.
SIGNIFICANCE LEVEL
X
37
-.542
SL= 3
37
.973
SL= 3
37
-.473
SL= 3
37
-.894
SL= 3
DFCTIdlF
•*W V • V^f t*
G/KG-F
SIGNIFICANCE LEVaS:
SL = 1 FOR 0.05>ALPHA>0.01
SL = 2 FOR O.Or.ALPHA>0.005
SL = 3 FOR O.OOSiALPHA
37
-.743
SL= 3
37
.840
SL= 3
37
-.687
SL= 3
37
-.940
SL= 3
37
.875
SL' 3
RES/EXT
RATIO
33
-.076
SL= 0
33
.542
SL= 3
33
.007
SL* 0
33
-.407
SL* 2
33
.609
SL= 3
33
.434
SL= 2
100K RCV
PER m
33
-.116
SL= 0
33
.315
SL= 1
33
-.041
SL= 0
33
-.274
SL= 0
33
.405
SL= 2
33
.311
SL= 1
33
.944
SL= 3
RF.U PFR
UG
PAW
33
-.472
SL= 3
33
.631
SL= 3
33
-.381
SL= t
33
-.654
SL= 3
33
.714
SL= 3
33
.707
SL= 3
33
.863
SL= 3
33
.838
SL= 3
REV PER
U6
EXT
33
.436
SL* 2
33
-.233
SL= 0
33
.442
SL= 3
33
.361
SL= 1
33
-.178
SL= 0
33
-.249
SL= 0
33
.608
SL= 3
33
.719
SL= 3
33
.279
SL= 0
REV PER
UG
RES
33
-.163
SL= 0
33
.566
SL= 3
33
-.052
SL= 0
33
-.463
SL= 3
33
.654
SL= 3
33
.480
SL= 3
33
.969
SL= 3
33
.934
SL= 3
33
.894
SL= 3
33
.531
SL* 3
153
-------
TABLE E-10. Emission and Bioactivity Correlations
HFET Cycle - "All" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100* REV REV PER REV PER REV PER 1 OOK REV
S/H1LC 6/HILE G/K6-F I G/K6-F RATIO PER HI US PART U8 EXT U6 RES PER KG-F
178 V3 178 179 178 178 168 168 169 168 168
->617 1 1645 31 2021 2.6168 3.3558 3.7547 1.0568 5.0141 1.5788 41.1794
If
HATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/WLE G/KG-F I 6/KG-f RATIO PER HI UG PART UG EXT UG RES PER HH
EXTRACT *7B 178 178 178 178 168
MULE '•°62 -960 -797 --307 --599 --049
wniLt SL* 0 SL= 3 SL* 3 SL= 3 SL* 3 SL* 0
RESIDUE J7B I78 "8 178 168
6/HlLE •-138 ••546 -731 -4*3 -155
b/niu SI* 1 SL* 3 SL* 3 SL* 3 SL* 1
EXTRACT 1$
Ql £ *i
wU. *
KEY:
EXTRACT
NO. OF tlATA PAIRS t
SAIfLE CORK. COtFF.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL = 1 FOR O.Oj}ALPHA>0.01
SL = ? FOR O.OnALPHA^.OOS
SL = 3 FOR 0.005)ALfHA
! 178 !78 168
5 -.214 -.618 .012
1 SL= 3 SL* 3 SL* 0
178 178 168
-.670 -.795 -.129
SL= 3 SL* 3 SL* 1
RESIDUE 178 168
6/KG-F .521 .387
SL* 3 SL* 3
RES/EXT 168
RATIO „ -061
SL* 0
100K REV
PER HI
168
-.216
SL* 3
168
-.103
SL* 0
163
-.130
SL* 1
168
-.162
SL* 1
168
.249
SL* 3
168
.075
SL* 0
168
.897
SL* 3
REV PER
UG
PAKT
168
-.374
SL* 3
168
.097
SL* 0
168
-.300
SL* 3
168
-.433
SL* 3
168
.464
SL* 3
168
.404
SL* 3
168
.805
SL* 3
168
.902
SL* 3
REV PER
US
EXT
168
.054
SL* 0
168
-.304
SL* 3
168
.116
SL* 0
168
.217
SL* 3
16*)
-.040
SL* 0
168
-.160
SL* 1
168
.797
SL* 3
168
.866
SL* 3
168
.638
SL* 3
REV PER
I'1
? RES
168
-.1/5
SL* t
168
.081
SL* 0
-it!
SL* "O
168
-.227
SL* 3
168
SL* 3
168
.141
SL* 1
168
.940
SL* 3
168
.936
SL* 3
168
.9v)8
SL* 3
168
.759
SL* 3
154
-------
TABLE E-ll. Emission and Bioactivity Correlations
CFDS Cycle - GM Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 1MK REV REV PER REV PER REV PER 1QOK RE\
6/hlL£ G/hlll P/KG-f I 6/KG-f RATIO PER hi UB PAST UG EXT UD RES PER KCH
73 73 73 74 73 73 71 72 72 71 71
.1922 .4057 1.4059 30.3232 2.9879 2.8229 4.0074 .7175 2.7917 1.0361 29.4410
DEV .:«10 .1174 .9p;9 12.7269 .8112 1.3455 1.8604 .3528 1.7873 .4941 13.1542
Zr73.3cj05 28.9346 69.6407 40.6583 27.1487 43.3719 46.4241 49.1701 64.0237 47.6899 44.6797
CORRELATION MATRIX
EXTRACT
6/MLE
RESIDUE EXTRACT
8/hlLE G/KG-F
73 73
.555 -.012
Sl= 3 SL= 0
RESIDUE •>'*
G/HILE SL= 1
KEY:
HO. CF DATA
SAlfLE CORR.
SIGNIFICANCE
SIGNIFICANCE
SL = 1 FOR 0
SL = 2 FOR 0
"JL - 3 FOR 0
EXTRACT
f* J1SP- f
6/KG-F
EXTRACT
Z
73
.560
SL= 3
73
.994
SL= 3
73
.172
SL= 0
EXTRACT
PAIRS Z
COEFF.
LEVEL
LEVaS:
,05>ALFHA>0.01
.01>^LPHA>0.005
RESIDUE
G/KG-F
74
.555
Si= 3
73
.884
SL= 3
73
-.193
Sl= 0
73
.908
Sl= 3
RESIDUE
G/KG-F
RES/EXT
RATIO
73
-.055
SL= 0
73
.139
SL= 0
73
.976
SL= 3
73
> .113
SL= 0
73
-.242
SL= i
RES/EXT
RATIO
1COK
nm
PER
100K REV
PER MI
73
-.503
SL= 3
73
-.721
SL= 3
71
.27C
£1 = 2
/ j
-.742
SL= 3
73
-.9U
SL= 3
73
.319
SL= 3
REV
HI
nl
REV
US
REV PER REV f ER
UGPART U6EXT
71 72
.165 -.092
SL» 0 BL= 0
71 71
.224 -.265
fjL- \ S'.= 1
71
.113 •
7;
-.324
SL= 0 51= 3
71 71
.182 -.282
SL- 0 SL= 2
71 72
.UC -.143
SL= 0 SL= 0
71
•}\
.046 -.350
SL= 0 SL= 3
71
-.140
SL= o .
71
.066
= 0
7t
pER 787
PART SL='7837
REV PER
1)6 EXT
REV PF.R
U6RES
72
-.280
St= 2
71
-.515
SL= 3
71
-.133
SL= 0
71
-.535
SL- 3
72
-.539
SL= 3
71
' A
-.135
SL= 0
7j
.545
SL= 3
71
.563
SL= 3
72
.843
SL= 3
REV PER
UB
RES
100K REV
PERK6-F
71
.068
SL= o
71
.050
SL= 0
•jj
-.367
SL= 3
71
.078
SL= 0
71
.209
SL= 1
71
-.4:2
SL= 3
71
-.252
SL= 1
71
' 4
.854
SL= 3
71
.928
SL= 3
71
i 1 O
.619
Sl= 3
Reproduced from $9
best available copy &Z&
155
-------
TABLE E-12. Emission and Bioactivity Correlations
CFDS Cycle - VW Vehicle Group - All Phases
"
HEAN
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
I/NILE G/KILE G/KG-F Z 6/KG-F RATIO PER HI US PART U6 EXT US RES PER K6-F
48
.075?
48 4? 48 47 48 46
.3540 1.0829 23.2287 3.4438 3.7813 6.1676
46 47 46 45
1.8827 9.7768 2.4250 89.6097
CORRELATION HATRIX
' RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 10« REV
6/NILE G/KG-F Z G/KG-F RATIO PER HI UG PART UG EXT UG RES PER KG-F
EXTRACT 48 47 48
5/KILE
KEY:
NO. OF
SAIfLE
.078 .994 .792
SL* 0 SL= 3 SL* 3
RESIDUE 47 48
G/HILE .049 -.513
SL* 0 SL* 3
EXTRACT <7
6/KG-F .785
SL= 3
EXTRACT
DATA PAIRS Z
CORR. COEFF.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL* 1
SL = 2
FOR 0.05iALPHA>0.01
FOR 0.01*AlPHA>0.005
FOR O.OOS^LPKA
47
.043
SL* 0
47
.980
SL* 3
47
.073
SL* 0
47
-.527
SL= 3
RESIDUE
G/KG-F
48
-.774
SL* 3
48
.501
SL= 3
47
-.767
SL= 3
46
-.949
SL* 3
47
.519
SL= 3
RES/EXT
RATIO
46
-.061
SL* 0
46
.388
SL* 3
45
-.047
SL= 0
46
-.285
45
.414
SL* 3
46
.325
SL* 1
1MK fiPU
PER M
1
46
-.209
SL= o
46
.098
SL* 0
45
-.177
SL= 0
46
:._250
45
.154
SL= 0
46
.230
SL* 1
46
.932
SL= 3
REV PER
UG PAkT
l^J I Fin 1
46
-.416
SL* 3
46
.235
SL* 0
45
-.387
SL= 3
46
-.495
SL= 3
45
.291
SL= 1
46
.542
SL* 3
46
.907
SL* 3
46
.952
SL= 3
REV PER
ne cxi
UW tAt
46
-.084
SL* 0
46
.023
SL* 0
45
-.052
SL* •)
46
-.101
SL* 0
45
.077
SL* 0
46
.144
SL* 0
46
.915
SL* 3
46
.937
SL* 3
46
f\f\f
.895
SL* 3
REV PER
U6RES
45
-.077
SL* 0
45
.370
SL= 2
45
-.055
SL* 0
45
-.288
SL* 1
45
.401
SL* 3
45
.324
SL* 1
45
.996
SL* 3
45
.948
SL* 3
45
H«\J
.921
SL* 3
45
A^ A
.930
SL* 3
156
-------
TABLE E-13. Emission and Bioactivity Correlations
CFDS Cycle - MB Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/HILE 6/NlLr 6/K6-F X 6/K6-F RATIO PER HI U6 PART US EXT U6 RES PER K6-F
22 22 22 22 22 22 20 20 20 20 20
.0643 .3539 .5823 16.0082 2.9613 6.6285 3.25*7 .7*80 6.3900 .8681 26.6824
' -0323 .0606 .3374 8.07)9 .3362 3.0424 2.6235 .4778 5.8607 .5204 19.82i9
RSD I «.8104 17.1347 57.9327 50.4234 11.3539 45.8988 80.6063 63.8795 91.7161 59.9459 74.2957
CORRELATION HATRIX
, RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/HILE 6/KG-F Z G/KG-F RATIO PER HI US PART 'JG EXT UG RES PER MH
EXTRACT 22
6/HILE
-.648
SL* 3 SL*
RESIDUE
6/HILE
SL*
22
997
3
22
665
3
EXTRACT
fi/KG-F
KEV:
HO. OF
SAMPLE
DATA PAIRS
CORR. COEFF.
22
.990
SL* 3
22
SL= 3
22
.990
SL* 3
EXTRACT
I
22
-.296
SL* 0
22
.858
SL* 3
22
-.295
SL* 0
22
-.413
SL* 1
22
-.907
SL= 3
22
.844
SL* 3
22
-.917
SL* 3
22
-.944
SI* 3
20
-.191
SL= 0
20
.595
SL* 3
20
-.212
SL* 0
20
-.282
SL* 0
~ •
SL*
*
20
212
0
20
481
SL* 1
"•
SL*
™ «
20
230
0
20
280
SL* 0
20
-.395
SL* 1
20
.682
SL* 3
20
-.415
Si.* 1
20
-.476
SL* 1
20
-.111
SL* 0
20
.411
SL* 1
20
-.J2B
SL* 0
20
-.179
SL= 0
20
-.120
SL= 0
JO
.536
SL* 2
20
-.138
SL* 0
20
-.209
SL* 0
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL* I
SL = 2
FOR 0.05>ALPKA>0
.01
RESIDUE"
6/K6-F
FOR 0.01)ALPHA>0.005
FOR O.OOS^ALPHA
22
.492
SL= 2
RES/EXT
RATIO
20
.5P5
SL* 3
20
.328
SL* 0
•
SL*
20
455
1
20
.278
SL=
0
20
.565
SL* 3
20
.515
SL* 2
20
.427
SL* 1
20
.183
SL* 0
20
.576
SL* 3
20
.247
SL* 0
100K REV .20 .2.0 .20 .2.0
PER HI
.
v//
.7/0
.764
.774
SL* 3 SL* 3 SL= 3 SL* 3
REV PER
UG PAH
20
.942
SL* 3
REV PER
U6EXI
20
.994
SL* 3
20
.905
SL* 3
REV PER
U6 RES
20
.982
SL* 3
20
.941
SL* 3
.980
SL* 3
157
-------
TABLE E-14. Emission and Bioactivity Correlations
CFDS Cycle - "Other" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRA
§/«!.£ 6/HILE 6/KGH
37
.1944
36
.2562
37
1.6373
CT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
• Z 6/K6-F RATIO PER HI U6 PART UG EXT U6 RES PER KG-F
37
45.2053
36
2.1958
36
1.7078
33
6.0494
33
1.2956
33
3.9242
32
2.4335
33
52.4203
N
BEAN
STP LEV .0721 .1225 .6053 18.3256 1.1180 1.2532 4.0479 ".6990 3163)5 V.3676 38.
RSD Z 37'. 1071 47.8131 36.9674 40.5386 50.9154 73.3774 66.9149 53.9501 92.5391 56.1991 72.7903
CORRELATION HATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER IOOK REV
6/HILE 6/KG-F Z 6/KG-F RATIO PER HI U6 PART UG EXT UG RES PER KG-F
EXTRACT 36 37 37 36 36 33
6/HILE --581 .943 .872 -.610 -.821 -.128
SL= 3 SL= 3 SL= 3 SL= 3 SL^ 3 SL= 0
RESIDUE 36 36 36 36 32
6/HILE '-553 -.690 .959 .853 .569
SL= 3 SL= 3 SL= 3 Sl= 3 SL* 3
EXTRACT 37 36 36 33
B/iw .823 -.502 -.771 -.048
KEY:
NO. Or DATA
SAtfLE CORR.
SIGNIFICANCE
SIGNIFICANCE
SL = 1 FOR 0
SL = 2 FOR 0
SL - 3 FOR 0
SL: 3
EXTRACT
^o i i\nv i
PAIRS Z
COEFF.
LEVEL
LEVELS:
.05>ALPHA>0.01
.01>ALPHA>0.005
.OOSJALPHA
SL= 3 SL= 3 SL= 0
36 36 33
-.884 -.942 -.371
SL= 3 SL= 3 SL= 1
RESIDUE 36 32
"WW A A^/fe fc » . • BhH
G/KG-F -866 -i22
SL= 3 SL= 3
RES/EXT 32
RATIO ._««
100K REV
PER HI
-
33
-.134
SL= 0
32
.308
Sl= 1
33
-.062
SL= 0
33
-.223
SL= 0
32
.391
SL= 1
32
.306
SL= 1
33
.918
SL» 3
REV PER
UG PART
33
-.480
SL= 3
32
.629
SL= 3
33
-.406
SL= 2
33
-.605
SL= 3
32
.693
SL= 3
32
.731
SL= 3
33
.877
SL= 3
33
.829
SL- 3
REV PER
UB EXT
32
.423
SL= 2
32
-.287
SL= 0
32
.428
SL-- ?
32
.443
SL= 2
32
-.227
SL= 0
32
-,312
SL= 1
32
.543
SL= 3
32
.747
SL= 3
32
.299
SL= 1
REV PER
U6RES
33
-.174
SL= 0
32
.573
SL= 3
33
-.047
SL= 0
33
-.405
SL= 2
32
.664
SL= 3
32
.471
SL= 3
33
.9b7
SL= 3
33
.907
SL= 3
33
.886
SL= 3
32
.495
SL= 3
158
-------
TABLE E-15. Emission and Bioactivity Correlations
CFDS Cycle - "All" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV m REV PER 100K REV
6/HILE 6/HlLE 6/KG-F Z 6/KG-F RATIO PER HI UG PART yg EXT U6 RES PER KG-F
« 180 179 179 181 178 179 170 171 172 169 'J69
hEAN .1462 .3286 1.267? 29.7437 2.9975 3.3233 4.3997 1.146J 5.3360 1.6588 49.6251
STD I|E« .1134 .1230 .7931 15.4190 .9979 2.1786 3.4156 .9493 5.8708 1.3092 46.3K.7
USD Z 7p.5317 37.4253 62.5662 51.8394 33.2892 65.5552 69.7105 82.8740 110.0215 78.9212 93.*7BO
*
CORRELATICX HATRIX
' RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/HILE G/X6-F Z G/KG-F RATIO PER HI UG PART UG EXT UG RES PER KG-F
EXTRACT l7' 179
6/HILE -164 -9*A
KEY:
HO. OF
SAtfLE
SL= 1 SL= 3
RESIDUE 1'B
AA1
Cls A
EXTRACT
G/KG-F
vf VXU *
£'
IlATA PAIRS
CORR. COEFK
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVELS:
SL = 1
a = 2
SL = 3
FOR 0.05>ALPHA>0.01
FOR 0.01W.PHAX>.W5
FOR O.OOS^ALNA
180 178
.770 -.201
SL= 3 SL» 3
179 178
-.394 .644
SL= 3 SL= 3
179 178
.816 -.141
SL" 3 SL= 1
KTRATT 1^8
I -.607
1 SL= 3
RESIDUE
l»t w 4 vk/L
G/KG-F
Uf nw '
•
179
-.631
SL= 3
179
.311
SL= 3
178
-.697
SL= 3
179
-.828
SL= 3
178
.440
SL= 3
RES/EXT
RATIO
170
-.004
SL* 0
169
.065
SL= 0
169
.063
SL= 0
170
-.065
SL^- 0
168
.336
SL= 3
169
.049
SL= 0
100K REV
PCB MI
i kt« iii
170
-.243
SL= 3
169
-.254
SL= 2
169
-.128
SL= 1
171
-.137
SL= 1
168
.193
SL= 2
169
.104
Sl= 0
170
.870
SL= 3
REV PER
f\Lv 1 LI\
UG PART
w rim i
170
-.396
SL= 3
169
-.075
SL= 0
169
-.332
SL= 3
171
-.431
SL= 3
168
.364
SL= 3
'.69
.456
SL= 3
170
.775
SL= 3
171
.896
SL= 3
REV PER
UG EXT
WW tM
169
-.038
SL= 0
169
-.397
SL« 3
168
.084
SL= 0
169
.206
SL= 3
168
-.049
SL= 0
169
-.136
SL= 1
169
.807
SL= 3
169
.917
SL= 3
169
. A A
.698
SL= 3
REV PER
UGRES
169
-.170
SL* 1
168
-.106
SL= 0
169
-.028
SL= 0
169
-.163
SL= 1
168
.410
SL= 3
166
.112
SL= 0
169
.914
SL= 3
169
.949
SL= 3
169
A*?A
.870
SL= 3
168
flT^
.837
SL= 3
159
-------
TABLE E-16. Emission and Bioactivity Correlations
NY City Cycle - CM Vehicle - All Phases
fiESIM ETTPtft UT84C1 BESJItt WS/EXT JW BW BfV HI KV Pfl JJW HI JW« «V
6VMLT U&-f t 6/*6-f fUUQ K»xl i*r**l U6 Ut UG US K» Wrf
• 2* 2* 2t 30 2t 29 24 P 27 26 24
*{M .54?: 1.4047 1.-.7S4 27.6444 4.1800 3.0044 7.8294 .3998 1.48-..2 .5557 23.3*J7
THt«t .2072 .:435 .V21 B.676» .7*04 1.2274 3.6Ui .1817
itSI' t 33.2187 17.3U7 15.7900 32.0^ 17.7147
CESIDUC CITKACT UTUH XSIDUC Rts/tit IOOK rev KV r» «v K« KV KI iooc etv
ftR Ml U6 FMT (A El! U6 RtS KB
">9 "f ^9 29 29 *4 ^4 24 ""i ^4
-.0*19 .992 .925 -.149 -.913 .048 -.ill -.524 .&>! -.0$
».« 0 SL« 3 Sl« 3 Si* 0 SL» I Sl» 0 Si.* C SI* 3 Sl« 0 Si* 0
29 29 29 29 24 24 24 24 24
-.030 -.360 .937 .2*3 .421 .134 .252 .519 .414
Si* 0 Sl« 1 Si* 3 £.« 0 Si* 1 Si.* 0 Si" 0 Si* 0 Si* 1
EXTRACT 29 29 29 24 24 24 24 2»
6/lW -'25 -.1*0 -.''-8 .040 -.'.41 -.528 .043 .OOi
* "^ SI* 3 St* 0 SI* 3 Sl« 0 k« 0 Sl« 3 k« 0 Sl« 0
EXTRACT 29 29 24 27 27 24 24
"'r*-1 -.4*2 -.945 -.030 -.127 -.54? .093 -.122
Si* 3 SL« 3 SI* 0 Si* 0 Si.* 3 Si.* 0 Si* 0
LEV£L KSlh« "* '4 24 24 24 24
6/!Sp .392 .349 .128 .264 -.0:5 .404
Sl« 1 SL« I SI* 0 31* 0 51* 0 SI* I
SI * 1 TOft 0.05>ALPv*>0.01 8ES/EXT ?» 2* 26 26 ?6
Si - 2 t"* 0 OliAipl^k>0 005 kATlQ -.003 .101 .532 -.148 .030
Si « 3 fOR 0.'o05$*Lfv,A ' a* 0 Sl« 0 Si* 3 Sl« 0 Si* 0
100K REV 26 24 24 24
PERRI .»2» -741 .V57 .991
Si* 3 SI* 3 Sl« 3 Si* 3
ȣV W 27 24 24
UG PAST -847 .942 .931
SI* 3 SI* 3 Sl» 3
StVPER 24 24
U6EXT ^.717 ^.7S2
KV K, ! 26
, • « « ' s,;90!
-------
TABLE E-17. Emission and Bioactivity Correlations
NY City Cycle - VW Vehicle Croup - All Phases
ess1 ssa? EBP "^ SB* w ws*
14 14 16 14 14 14 14 14 14 J« 14
* .113 .33M 1.17*,4 32.830* 7.4071 2.4«<>9 5.4910 1.-.905 4.14*2 !.«:«» 4?.rvO
«U i VOW .1*1* .iW 13.0*13 .!«? 1.IM7 3.56*8 .7*3) 7.7eJ4 1/412 2>.4>CO
X 47.9114 71.1172 42.8V.,? J».7V«a 24.1138 10.77U 44.6401 42.4190 tll.«744 60.2712 44.7414
flTtACt 14 14
-.602 ."4
14
6/Kllt -.^i*
tlTSACT
S1G-JFIC4XE LEVEL
uvasi
SL « i ryt
SL t 2 F0« 0.0';'A.^<*>«.005
Si. • 3 FGA
>tft KSIWC KS/Efl
14 14 14
17 -.477 -.6*2
14 14 14
R4 .980 .6*0
3 k« 3 k» 3
14 14 '.4
>64 -.474 -.erO
3 Sl» 3 k* 3
, . .^
-.e:7 -.'«3
k» 3 >.« 3
KSIDUC H
Sl« '1
KS/EH
RATIO
100K «V
ft* HI
14
.184
Si1 0
14
.037
k« 0
14
Sl«" 0
34
.'.73
S.« 0
1^4
SL« 0
14
.013
s»» c
1W 6EV
Pf» Ml
ttv
UG
In*
«V K> KV KK KV KR 16M KV
14
.101
k* o s-
•Ml*
Si* 0 S
14
.::«
Sl* 0 9
14
k«"o £
14
k» "o s
u
.050
^.* 0 ?
14
14 14
-.455 .413
.- 0 Sl« 0
14 14
.554 -.409
-1 5 5l* v'
14 '.4
-.449 .4'.8
l« 0 k« 0
14 '4
-.104 .1-8
.* 1 Sl* 1
14 14
.t;T -.l"j
t« 2 SI* 0
.4*5 -.115
K« 3 S'.« 0
14 i4
.735 .£75
Sl« 3 il« 3 Si.* 3
KR
P6kT
>
14 i4
^4 9'^
Jl'**3 Sc'*' 3
REV P€R 14
116 DT . -273
SI* 0
KU KD
U6"R£S'
?eprodacec! f
b^s! ^v^t!,* s'fl
14
SL«'' o
14
.104
s»« o
14
.118
Si* o
14
.'.00
Si* o
i4
Si* 6
•4
.135
S;» o
14
.*<•»
Sl* 3
:4
• ^64
St* 3
1 4
.7^5
S^* 3
•4
.539
?<.- 3
iom f ^
copy. V^ ^
161
-------
TAL~E E-18. Emission and Bioactivity Correlations
NY City Cycle - MB Vehicle Group - All Phases
B?ff BfflP BBS1 "V" I
• 9 9 9 10
•»"*• . .'.4*4 ,?«4J .7010 14.5107 3
•I- tfJ ' .0154 ,;23» .K34 5.ei»3
»iil 37.3«4? 16.;»70 43.^75 34.51M 16
a
'HV "^^P 1K*0
/li-* IMIU Kft
9 9
.5?32 5.9194 2.6
.i'.8 2.2J36 I.:
.44?; 37./J40 4*.i
MSIATIQN MTtll
RESIWt ETTR«CT EITR«CT RESIDUE RES/EIT 100* «W
I/MLC 6/nfr-f t 6/16-1
El'^ATl 9999
G/niil -.'.'^ .965 .9*0 .092
k« o k« 3 k« 3 k» o
RESIDUE 999
6/VlU -.:»« -.509 .870
6 IU k« 0 k« 0 k« 3
o o
nfTftACT
C/K6-* -'5' -J3*
k» 3 k« 0
Eimci . ?
•0. OF »*T« Pft!»S . I '•'-«
SA*VE co"tt. co£fF. k« o
SIS«I;FICA«CE LEVEL
6A5-f
S15«IFICW(CE LEVELS:
: RATIO PER HI
9 7
-.913 .490
k« 3 k* o
9 7
.579 -.020
k« 0 k« 0
9 7
-.9:0 .459
k* 3 Sl« 0
9 7
-.974 .402
k« 3 k« 0
9 7
.207 .231
k* 0 k» 0
k « 1 FOR 0.05>«.P**>0.01 RES/ETT ejj
k « 2 FO« o.o;»ALp"«>o.oo5 RATIO "•544
k « 3 FOR o.ooSyyj** *•• °
IOC* REV
FIR RI
* ^** *••
WW
7
467 .31
427 .i;
598 39.91
REV PER
JSPMT
7
.421
k* o
7
-.316
k« 0
7
.4'^
k« o
6
.550
k« 0
7
-.OM
cl* 0
7
-.645
k* o
7
.953
k* 3
REV PER
IK DALT
w
* n>% •
REV
116
ffl OT SW W*
8
0« 2.1!
;4l .fe(
,97 37. V
REV PEI
U6EJT
7
-.026
k« 0
7
k'» *0
7
-.053
k* o
8
-.045
S.« 0
7
-.092
k« o
7
-.199
Sl« 0
*'
3L c
8
.606
k- 2
PER
EXI
REV
US
8
IV V f
/ ' p •
'00 44.
REV PEI
U6RE?
/
.509
Sl« 0
7
-.347
k« 0
7
.4»5
k» o
7
.575
k» 0
7
-.055
Sl« 0
7
-.721
k« 1
7
.913
Sl« 3
7
.994
SI* 3
7
.6)3
-------
TABLE E-19.
Emission and Bioactivity Correlations
NY City Cycle - "Other" Vehicle Group - All Phases
£/«lL£
22
.47{7
..::««
RESIDUE
6/Hlir
21'
.41)39
EXTRACT EXTRACT RESIDUE «s/m iwa REV REV PER REV PCS REV PER
t/nfirf t O/n&T fcillO KB M U6 PAST UG ElT U6 RIS
2? "*?
1.9B-.94 48.606*6
:».6;.3
2.1875
1.2:43
21
1.4522*
I.OKJ
6.380*
5.3:40
IB
.4402
.450-3
'9
1.7444
18
J.2299 30.9402
..::«« .21:2 .MT? :».6;.3 1.2:43 I.OKJ 5.3:40 .450-3 1.2:18 .^735 3o.67«.5
45.550* 45.7034 42.7479 40.3471 55.9672 74.7312 83.4416 70.4156 69.4675 71.0195 99.7910
1/nltC
OKKIAUON MTRIX
EITftACT EXTRACT RESIDUE RES/EH 1 OCX REV REV PER REV PER REV PEft 100* REV
G/KC-F I 6/nfr-f RATIO Ptt HI US PAfiT UG EXT U6 R£S PER K6-f
EYT8ACT ""
6/alLE -.407
SI* 1 SI
RESIDUE
EXTRA
6M6-
V* **W
l£T:
"0. or DATA PAIRS
SA*LE cow. CCCFF.
SlWIFICAnCE LEVEL
£IWIFICA»C£ 1 EVaS i
k ° 2 ^o^ o.oi iitP^^o
St « 3 FOR 0.005>At>HA
•
ty It It
.7V9 /14 -.434
.« 3 SL* 3 SL* l
-,•» V) •»•»
-.:« -.7:4 .944'
.« b s,« ; k* 3
CT 22 22
f .7;3 -.079
iL* 3 v.* 0
EXTRACT ::
1 -.638
Si* 3
RESIDUE
22
-.903
Sl« 3
22
S^7<3
•)-
-.767
v.* 3
22
-.945
SL* 3
~t
.543
SL* 3
.0! RES/nT
.005 RATIO
-
100K
PER
18
.096
SL* o
;e
.5:9
Sl* 1
IB
SL* 1
/:8
Sl*' 0
•B
.679
SL* 3
-.C39
SL« o
REV
HI
REV
. UG
:B
-.075
SL* o
)8
.355
SL* o
18
.297
k* 0
18
SJ*"*i
13
.515
* ^
•8
.075
Si* 0
-.8
.546
SL* 3
PER
PACT
REV
US
IB
-,460
SL* 1
18
sJ?1
18
- ' j7
SL* o
IP
-.605
SL* 3
'.fi
.499
SL* 3
18
.531
SL* \
'•6
.750
SL= 3
1«
.615
St.= 3
PER
EKl
REV
U6
18
.351
SL* 0
18
-.037
Sk* 0
:8
•*M
JC* 4
.18
SL** 0
IB
.131
SL* 0
-'I
k'^'o
\3
.bi3
SL* 3
:3
.849
si* 3
18
.5:6
SL* i
KR
RES
13
.0*?
k* o
18
.476
SL* l
•*
.451
k« 1
•0
-.159
SL* 0
$
k* 3
•-8
.004
k* o
18
SL* "3
•8
k* M3
IS
B- .
v4
k* 3
• a
S^ *
.0
SL* 3
Reproduced from ~f ^\
best available copy. k-J
163
-------
TABLE E-20. Emission and Bioactivity Correlations
NY City Cycle - 'All" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 10* REV REV PER REV PtR REV PER 100K REV
6/HllE G/AllE G/KG-f Z 6/KG-f RATIO K& Ml UG PART uG EXT U6 R£S PER WH
N 76 76 76 78 76 76 65 67 69 65 65
«EA* .3929 .8310 1.5143 32.9703 3.1600 2.792? 6.3689 .6190 2.8817 1.0149 28.4697
STPDEV .2458 .5103 .74*6 16.9977 1.2467 1.8659 4.2334 ,5257 4.2793 1.0201 23.3070
llSD Z 6^.5636 61.4096 49.1698 51.5545 39.4533 66.6239 66.4*97 64.9349 1(8.4991 100.5141 81.8659
CORRELATION MATRIX :
, RESIDUE EXTRACT EXTRACT RESIDUE RES/EH 10<* KV REV PER REV PER REV PER 100X REV
6/nlLE 6/KG-f Z 6/KG-f RATIO PER M UG PART UG EXT UG RES PER WH
EXTRACT 76 76 76
t/HlLE
KET:
NO. OF
.378 .624 .566
SL* 3 SL* 3 SL* 3
RESIDUE 76 76
6/HlLE '.005 -.461
• ft + ••• \
SL* 0 SL* 3
EXTRACT 76
6/KG-f ."4
SL* 3
EXTRACT
DATA PAIRS Z
CORR. COtfF.
SIGNIFICANCE LEVEL
SIGNIFICANCE LEVaS:
SL* 1
SL « 2
SL* 3
FOR 0.05>ALPHA>0.01
FOR 0.01>AlPHA>0.005
FOR O.W5>A^HA
76
.050
SL* 0
76
.874
SL* 3
76
-.1R4
SL* 0
76
-.688
SL* 3
RESIDUE
" W*f 9 • V^
6/K&-F
V ' "V w
76
. -.588
SL* 3
76
.309
SL* 3
76
-.779
SL* 3
76
-.844
SL* 3
76
.505
SL* 3
RES/EH
RATIO
65
.284
SL* 1
65
.359
SL* 3
65
.365
SL* 3
65
.019
SL- 0
65
.412
SL* 3
65
-.179
SL* 0
100* REV
m W^ ™-L v
PCb lit
1 fc."^ |1t
65
-.269
SL* 1
65
-.388
SL* 3
65
.056
SL* 0
67
.133
SL* 0
65
-.197
SL* 0
65
-.186
SL* 0
65
.563
SI* 3
REV PER
^ f ~ t •*
\f. PM.-T
VTO • H* 1
65
-.485
SL* 3
65
-.225
SL* 1
65
-.349
GL* 3
67
-.308
SL* 2
65
.050
SL* 0
65
.248
SL* I
65
.412
SL* 3
67
.794
SL* 3
(ni pro
™ t » * t ^
tf£ PTfl
UV LA*
65
-.085
SL* 0
65
-.457
SL* 3
65
.269
SL* 1
65
.420
SL* 3
65
-.371
SL* 3
65
-.372
SL* 3
65
.491
SL* 3
65
.916
SL* 3
65
.526
SL* 3
REV PER
US RES
65
-.069
SL* 0
65
-.135
SL* 0
65
.2AB
SL* 1
65
.089
SL* 0
65
.086
SL* 0
65
-.190
SL* 0
65
.807
SL* 3
65
.904
SL* 3
65
.707
SL* 3
65
& i jh
.810
St.* 3
164
-------
TABLE E-21. Emission and Bioactivity Correlations
IDLE Cycle - GM Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 10JX REV
(/MILE 6/MLE 6/Mi-f I 6/K6-T RATIO PER Ml US PART US EXT U6 RES PER Kfr-F
72 72 29 73 29 72 69 70 70 69 24
.0393 .1455 2.2700 20.7310 6.4888 4.4698 .8245 .4412 2.3286 .5658 38.9337
STIi ItV .0200 .0224 .9293 7.6753 1.0045 1.8689 .4789 .2342 1.3078 .3130 17.4601
RSD Z 50.9708 15.3906 40.9606 37.0236 15.4806 41.8113 58.0893 53.0813 56.1652 55.3239 44.6458
CORRELATION HATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 100K REV
6/ftILE 6/KG-f Z 6/KG-f RATIO PER «I UG PART UG EXT UG RES PER HH
EXTRACT 72 29
6/MLE .089 .993
SL« 0 SL* 3
RESIDUE 29
G/K1LE .014
SL* 0
EXTRACT
6/KG-f
KET:
.940
SL* 3
72
-.223
SL* 1
29
.947
SL* 3
EXTRACT
NO. OF DATA PAIRS
SAiflE COKR. COE'F.
SIGNIFICANCE LEVEL
.
SIGNIFICANCE LEVELS:
SL * 1 FOR 0.05>ftLfHAM>.01
SL * 2 FOR 0.01 JALMAX1.005
SL « 3 FOR 0.005*LPKA
Z
29
-.247
SL* 0
29
.872
SL* 3
29
-.199
SL* 0
29
-.488
SL* 3
RESIDUE
G/KG-f
-.8$
SL» 3
72
.263
SL* 1
29
-.870
SL* 3
72
-.921
SL« 3
29
.469
SL* 3
RES/EXT
RATIO
69
.405
SL* 3
69
.274
SL* 1
26
.509
SL* 3
69
.318
SL* 3
26
.424
SL* 1
69
-.333
SI* 3
69
.218
SL* 1
69
.079
SL* 0
26
.272
SL* 0
70
.205
SL= 1
26
.377
SL* 1
69
-.268
SL= 1
100K REV 69
PER HI
.954
SL* 3
REV PER
UG PAH
6?
-.270
SL* 1
69
.227
SL* 1
26
-.457
SL* 2
70
-.337
SL= 3
26
.639
EL* 3
69
.287
SL* 2
49
.729
SL* 3
70
.821
SL* 3
REV PER
U6EXT
69
.400
SL= 3
69
.039
SL* 0
26
.562
SL* 3
69
.393
SL* 3
26
.169
SL* 0
69
-.416
SL* 3
69
.963
SL* 3
69
.977
SL* 3
69
.696
SL* 3
fcrii fc*"f>
us" RES"
.3?t
SL* 1
?6
.432
3L* 1
26
.406
SL* 1
n6
.203
SL* 0
26
.4ti5
SL* 2
26
-.138
SL* 0
26
,9s4
SL* 3
26
.969
SL* 3
26
.575
SL* 3
11
» O
.938
SL* 3
165
-------
TABLE E-22. Emission and Bioactivity Correlations
IDLE Cycle - VW Vehicle Group - All Phases
PfJ?*£T KSiy£ "TRACT- EXTRACT RESIDUE RES/EXT 100K REV REV PER RFV PER REV PER 100K REV
S/H1LE G/HILE 6/MK I 6AG-F HATIO PER HI lE
SL* 0
EXTRACT
6/K6-F
46 17
.•»55 .115
SL» 3 SL* 0
46 17
-.260 .977
SL* 1 SL* 3
17 17
.788 .107
SL* 3 SL* 0
EXTRACT 17
DATA PAIRS
CORR.
SIGNIFICANCE
SIGNIFICANCE
SL« 1
SL = 2
SL* 3
COEFF.
LEVEL
LEVELS:
FOR 0.05}ALPHA>0.01
FOR O.OUAlfHA>0.005
FOR 0.
005} ALPHA
Z -.414
SL* 1
RESIDUE
G/KG-F
46
-.353
SL* 2
46
.159
SL* 0
17
-.398
SL* 0
46
-.595
SL= 3
17
.268
SL* 0
RES/EXT
RATIO
43
.513
SL* 3
43
.240
SL* 0
15
.395
SL* 0
43
.412
SL* 3
-.281
SL* 0
43
-.222
SL* 0
100K REV
PER HI
43
.025
SL= 0
43
-.081
SL* 0
15
-.114
SL* 0
43
.106
SL* 0
-.506
SL* 1
43
-.119
SL= 0
43
.702
SL= 3
RtV PFR
UG
PART
43
-.124
SL* 0
43
-.078
SL* 0
15
-.239
SL* 0
43
-.082
SL* 0
-.476
SL* 1
43
.012
SL* 0
43
.574
SL* 3
43
.974
SL* 3
REV PER
US ntT
~
43
.266
SL* 1
43
-.190
SL* 0
15
.114
SL' 0
43
.433
SL* 3
15
-.494
SL* 1
-.204
SL= 0
43
.758
SL* 3
43
.808
SL* *
43
.706
SL= 3
REV PER
US
RES
15
.411
SL= 0
15
-.217
SL* 0
15
.396
SL* 0
15
.521
SL* 1
-.276
SL* 0
15
-.300
SL* 0
15
.9*8
SL* 3
15
.741
SL* 3
.653
SL* 3
15
.888
SI* 3
166
-------
TABLE E-23. Emission and Bioactivity Correlations
IDLE Cycle - MB Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/PCT IOOK REV REV PER REV PER REV PER iwc REV
«/«lLE G/MLT 6A6-7 Z 6/K6-T RATIO PER «1 US PART UG EXT US RES P«K6-F
u 22 22 10 22 10 22 20 20 20 20 8
h£AH .0071 .0403 .7946 15.3691 4.0652 6.0468 .1823 .4105 2.8800 .4825 17.8519
STD DEV .0022 .0103 .3517 4.6047 .8976 2.0176 .1125 .2807 1.8875 .1309 8.7418 .
RSti X Jl.1607 25.6867 44.1536 29.9611 22.0812 33.3660 61.7247 68.3790 65.5392 68.S926 48.9684
CORRELATION MATRIX
•RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER IOOK REV
6/ftIlE 6/KG-f Z 6/K6-f RATIO PER HI UG PART UG EXT UG RES PER KH
EXTRACT 21 10 22
6/NILE .068 .951 .769
SL* 0 SL= 3 SL* 3
RESIDUE 10 22
6/HILE -.100 -.566
SL* 0 SL* 3
KEY:
W.
Of
SAMPLE
EXTRACT 10
6/KG-f .896
SL* 3
EXTRACT
DATA PAIRS z
CORR.
SIGNIFICANCE
SIGNIFICANCE
SL«
SL «
SL*
1
'2
3
FORO.
COEFF.
LEVEL
LEVELS:
05>ALP«A>0.01
FOR 0.01>ALPrtA>0.005
FOR 0.
005)M-fHA
10 22
.390 -.621
SL* 0 SL* 3
10 22
.649 .681
SL* 1 SL* 3
10 10
.460 -.837
SL* 0 SL* 3
10 22
.022 -,936
SL* 0 SL* 3
RESIDUE 10
6/KG-F .002
SL* 0
RES/ETt
RATIO
20
-.079
SL* 0
20
-.118
SL* 0
8
.284
SL* 0
20
-.020
SL* 0
8
.283
SL* 0
20
-.110
SL* 0
20
-.229
SL* 0
20
-.385
SL* 1
8
.064
SL* 0
20
.104
Si* 0
8
-.088
SL* 0
20
-.230
SL* 0
IOOK REV ?0
PER HI
SL=^
REV PER
UG PART
20
-.371
SL* 0
20
-.230
SL* 0
8
-.041
SL^ 0
20
-.141
SL* 0
8
-.049
SL* 0
20
.011
SL* 0
20
.949
SL* 3
20
...'41
20
-.205
SL* 0
20
-.408
SL* I
8
.080
SL* 0
20
.144
SL* 0
8
-.091
SL* 0
20
-.267
SL* 0
20
.943
SL* 3
20
..999
B
.107
SL* 0
8
-.136
SL* 0
8
.437
SL* 0
8
.354
SL* 0
8
.369
SL* 0
8
-.325
SL* 0
8
.961
SL* 3
8
.. -888
REV PER 20 8
U6 EXF .949 .872
SL* 3 SL* 3
REV PtR g
UE KES .895
SL* 3
167
-------
TABLE E-24,
Emission and Bioactivity Correlations
IDLE Cycle - "Other" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV .REV PER KVJP REVJf* JSOK.REV
f/«LE 6/N1LE G/K6-F
IKftU KlblWfc KVJ/III 1WK KIV KtV KIK KtV re* KIV rtR 1VW Ktv
I 6/K6-T RATIO PER HI U8 PART jo EXT % RES PER K6-f
N 37 37 22 37 22 37 33 33 33 33 18
H[W .0431 .0160 3.4040 60.7690 1.2593 .9148 .2417 .6292 1.3192 1.7427 21.4140
STC DEV , .0440 .OOV4 3.0905 20.3477 .6961 .9380 .1547 .4378 1.147? 1.1737 12.4736
RSD 1 101.9776 58.6159 90.7895 33.5166 53.2808 102.5426 63.9899 69.5827 87.0860 67.3488 57.7076
CORRELATION MATRIX
RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT iOOK REV REV PER REV PER REV PER 100K REV
6/H'.LE 6/K6-F X 6/KG-F RATIO PER HI U6 PART DC EXT UG RES PER KG-f
EXTRACT 37 22
6/HILE
KEY:
W. OF
SAifLE
.260 .828
SL* 0 SL* 3
RESIDUE 22
G/HILE -050
SL* 0
EXTRACT
6/K6-F
37
.837
SL* 3
37
-.058
SL* 0
22
.758
SL* 3
EXTRACT
DATA PAIRS
CORR. CCEFF.
Z
22
-.171
SL* 0
22
.804
SL* 3
22
.055
SL* 0
22
-.486
SL* 1
SIGNIFICANCE LEVEL
RESIDUE
G/KG-F
SIGNIFICANCE LEVELS:
SL* 1
SL = 2
SL = 3
FOR 0.05>ALPHAX).01
FOR O.OUALPHA>0.005
37
-.584
SL* 3
37
.334
SL* 1
22
-.532
SL* 3
37
-.880
SL* 3
22
.694
SL* 3
RES/EXT
RATIO
33
.548
SL* 3
33
.518
SL* 3
IB
.265
SL* 0
33
.499
SL* 3
IB
.198
SL* 0
33
-.256
SL* 0
IOOKREV
'
PER HI
33
-.456
SL* 3
33
-.307
SL* 1
18
-.499
SL* 1
33
-.259
SL* 0
IB
-.167
SL= 0
33
.150
SL= 0
33
.119
SL* 0
REV PER
UGPAKT
33
-.573
SL* 3
33
-.02?
SL* 0
18
-.586
SL* 3
33
-.623
SL* 3
IB
.408
SL* 1
33
.646
SL* 3
33
-.026
SL* 0
3S
.814
SL* 3
REV PER
U6EXI
33
.174
SL* 0
33
-.292
SL* 1
18
.086
SL* 0
33
.441
SL* 3
18
-.408
SL* 1
33
-.385
SL* 1
33
.516
SL* 3
33
.675
SL* 3
33
.277
SL* 0
REV PF.R
U6R£S
.8
.100
SL* 0
18
.113
SL= 0
18
.304
SL* 0
IB
.232
SL» 0
18
.258
SL* 0
18
-.107
SL* 0
IB
.658
SL* 3
18
.497
SL* 1
IB
.1V8
SL= 0
18
.677
sL= I
168
-------
TABLE E-25. Emission and Bioactivity Correlations
IDLE Cycle - "All" Vehicle Group - All Phases
EXTRACT RESIDUE EXTRACT EXTRACT RESIDUE RES/fifT 100K REV REV KR REV PER REV PER 100* REV
6/K1LE 6/MILE 6/KG-f I 6/K6-F RATIO PER Ml US PART U6 EXT US RES PER KG-f
N 177 177 78 178 78 177 165 164 168 165 67
Kt(W .0289 .0673 2.2666 37.2272 3.5312 3.1263 .4732 .6733 2.2386 1.6023 32.9041
STD DEW .0285 .0659 2.0197 24.5281 2.5950 2.7032 .4634 1.0967 2.2594 3.3034 28.0038
RSDZ 98.6226 95.1333 89.1091 65.8875 73.4879 86.4659 97.9312 162.8839 100.V273 206.1694 85.1074
CORRELATION HATRIX
, RESIDUE EXTRACT EXTRACT RESIDUE RES/EXT 100K REV REV PER REV PER REV PER 10WC REV
6/MLE G/K6-F I 6/K5-F RATIO P£R HI US PART UG EXT UO RES PER KH
EXTRACT 177 78
6/MLE -300 '812
SL* 3 SL* 3
RESIDUE 7B
6/hILE "•«"
SL* 0
EXTRACT
6/KG-F
KFY:
177
.344
SL* 3
177
-.627
SL* 3
.652
SL= 3
EXTRACT
NO. OF DATA PAIRS
SAIfLE CORR. COEFF.
SIGNIFICANCE LEVEL
Z
78
.179
Sl= 0
78
.957
SL* 3
78
-.098
SL* 0
78
-.693
SL* 3
RESIDUE
6/KG-F
SIGNIFICANCE LEVELS:
SL = 1 FOR O.OSiALPHAXJ.Ol
SL * 2 FOR O.C1>4LPHA>0.005
SL * 3 FOR O.OC5WHA
177
-.276
SL* 3
177
.513
SL* 3
78
-.465
• SL- 3
177
-.816
SL* 3
78
.440
SL* 3
RES/EXT
RATIO
165
.483
SL* 3
165
.661
SL= 3
67
.236
SL* 1
165
-.210
SL= 3
47
.660
SL= 3
165
.101
SL= 0
100K REV
PFR HI
165
-.129
SL= 0
165
-.207
SL* 3
67 ,
-.104
SL= 0
166
.206
SL= 3
67
-.260
SL= 1
165
-.221
SL= 3
165
.201
SL= 3
REV PER
f\L v » W*
IB P&CfT
w^r
165
-.218
SL= 3
165
.071
SL= 0
67
-.261
SL= 1
166
-.220
SL= 3
67
-.069
SL= 0
145
.180
SL= 1
145
.386
SL= 3
144
.842
SL= 3
165
-.040
SL= 0
145
-.303
SL= 3
47
.044
SL= 0
165
.461
SL= 3
67
-.360
SL= 3
165
-.335
SL* 3
165
.114
SL= 0
145
.820
SL= 3
67
.146
SL= 0
67
.161
SL= 0
67
.211
SL= 1
6?
.177
SL= 0
67
.141
SL= 0
67
-.208
SL= 1
67
.689
SL= 3
67
.660
SL= 3
REV PER U5 4.7
U6
EXT
Ml _ ^
Mi _ ^
REV PER 47
UB RES ^38,
169
------- |