Office of Transportation EPA420-D-06-005
United States and Air Quality February 2006
Environmental Protection
Agency
California Heavy Heavy-Duty
Diesel Truck Emissions
Characterization for
Program E-55/59
Draft Report
-------
EPA420-D-06-005
February 2006
California Heavy Heavy-Duty Diesel Truck Emissions
Characterization for Program E-55/59
Draft Report
Assessment & Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
Prepared for EPA by:
Coordinating Research Council (CRC)
NOTICE
This Technical Report does not necessarily represent final EPA decisions or positions.
It is intended to present technical analysis of issues using data that are currently available.
The purpose in the release of such reports is to facilitate an exchange of
technical information and to inform the public of technical developments.
-------
TABLE OF CONTENTS
TABLE OF CONTENTS i
LIST OF FIGURES iii
LIST OF TABLES viii
LIST OF APPENDICES x
LIST OF APPENDICES x
EXECUTIVE SUMMARY 1
NOMENCLATURE 6
INTRODUCTION AND OBJECTIVES 7
VEHICLE PROCUREMENT 9
EQUIPMENT & METHODS - REGULATED SPECIES 15
Translab Description 15
Preparation of the Vehicle for Testing 15
Engine/Vehicle Preparation 16
Wheel-Hub Adapter Installation 17
Test Vehicle Mounting 17
Connections of Vehicle to Laboratory 17
Pre-test Vehicle Operation 17
Vehicle Driving Instructions 18
Vehicle Practice Run 18
Test Run Procedures 19
Gaseous andPM sampling system 19
Gas Analyzer Operation 22
Calibration Procedures 22
Particulate Matter Filter Conditioning and Weighing 23
Test Cycles 23
Test Weights 27
EQUIPMENT AND METHODS - NON-REGULATED SPECIES & PARTICLE
SIZING 31
DRI's Residence Time Dilution Tunnel 31
Mini Dilution System 33
Particle Sizing with the SMPS 37
Particle sizing with the DMS500 40
Unregulated Emissions 42
Speciation Fleet and Approach 44
QUALITY ASSURANCE AND QUALITY CONTROL 46
RESULTS AND DISCUSSION, REGULATED SPECIES 48
HHDT Data Gathered 48
MHDT Data Discussion 60
Continuous Data 69
HHDDT Effect of Test Weight 72
MHDT Effect of Test Weight 74
COMPARISON OF DATA BETWEEN PHASES 75
RESULTS AND DISCUSSION, NONREGULATED SPECIES 78
Semi-Volatile Organic Compounds 78
-------
Volatile Organic Compounds 82
SPECIATION AND SIZING RESULTS 87
CONCLUSIONS, NONREGULATED SPECIES 126
TAMPERING, MALMAINTENANCE AND REFLASH VEHICLES 131
T&M Criteria 131
Determination of Candidates from Vehicle Inspection 131
Determination of T&M Candidate from Emissions 132
Results: T&M by Inspection 135
Verifying the Criteria in Phase 1: NOX Analysis 136
Verifying the Criteria in Phase 1: PM Analysis 137
Criteria ratios for all phases 142
"Before and After" Emissions 151
T&M and Reflash Discussion 167
ACKNOWLEDGEMENTS 170
REFERENCES 171
11
-------
LIST OF FIGURES
Figure 1: WVU Sampling System Schematic 20
Figure 2: AC5080 Short Test (time not to scale) 26
Figure 3: The AC5080 Schedule as driven by a 1990 Kenworth road tractor with a 370 hp
Cummins Ml 1 engine and a 10-speed manual transmission, with a simulated vehicle
test weight of 56,000 Ibs 27
Figure 4: PM emissions for vehicle E55CRC-27 which was tested at four weights, before
66,000 Ibs. was adopted as the highest test weight 28
Figure 5: NOX emissions for vehicle E55CRC-27 which was tested at four weights, before
66,000 Ibs. was adopted as the highest test weight 28
Figure 6: PM emissions for vehicle E55CRC-53 tested at two weights 29
Figure 7: NOX emissions for vehicle E55CRC-53 tested at two weights 30
Figure 8: Schematic of DRI dilution tunnel sampler (Hildemann, et al.6) 32
Figure 9: Exhaust Coupler with Sampling Probe for mini dilution tunnel 33
Figure 10: West Virginia University-Mini Dilution System 34
Figure 11: Schematic Representation of the Damping System 36
Figure 12: Exhaust Pulsation Damper 36
Figure 13: Differential Mobility Analyzer 37
Figure 14: TSI Model 3936 SMPS System 39
Figure 15: Working principle of the DMS5008 41
Figure 16: NOX emissions for the Idle Mode, in units of g/minute 49
Figure 17: NOX emissions for the Creep mode (56,000 Ibs.) 49
Figure 18: NOX emissions for the Transient mode (56,000 Ibs.). One of the vehicles in the
1975-1976 MY bin emitted NOX at 60 g/mile 50
Figure 19: NOX emissions for the Cruise mode (56,000 Ibs.). One 1998 vehicle (E55CRC-
10) was identified as a high NOX emitter in Phase 1 51
Figure 20: NOX emissions for HHDDT_S mode (56,000 Ibs.). The high value for the
1975-1976 MY bin is due to a single truck with highNOx emissions 52
Figure 21: NOX emissions for UDDS mode (56,000 Ibs.) 53
Figure 22: PM emissions for Idle Mode. Truck E55CRC-45, in the 1991-1993 MY bin,
was a prodigious emitter and raised the average value for that bin substantially 53
Figure 23: PM emissions for Creep mode (56,000 Ibs.). One vehicle in the 1991-1993
MY bin was a very high emitter of both PM andHC at light load 54
Figure 24: PM Emissions for the Transient mode (56,000 Ibs.). Truck E55CRC-16, a high
emitter, caused the high average in the 1977-1979 MY bin 55
Figure 25: PM emissions for the Cruise mode (56,000 Ibs.). E55CRC-16 contributed to
the one very high MY bin average 55
Figure 26: PM emissions for HHDDT_S mode (56,000 Ibs.). Four MY groups show data
from only one vehicle 56
Figure 27: PM emissions for UDDS mode (56,000 Ibs.) 57
Figure 28: CO emissions for Transient mode (56,000 Ibs.) 57
Figure 29: CO emissions for Cruise mode (56,000 Ibs.) 58
Figure 30: HC emissions for Transient mode (56,000 Ibs.) 59
Figure 31: HC emissions for Cruise mode (56,000 Ibs.) 59
Figure 32: NOX emissions for AC5080 (both laden and unladen). Gasoline-fueled
vehicles are separated from diesel-fueled vehicles for comparison 60
in
-------
Figure 33: NOX emissions for the Lower Speed Transient mode (both laden and
unladen) 61
Figure 34: NOX emissions for the Higher Speed Transient mode (both laden and unladen).
62
Figure 35: NOX emissions for the MHDT Cruise mode (both laden and unladen) 62
Figure 36: NOX emissions for the HHDDT_S mode (both laden and unladen) 63
Figure 37: NOX emissions for theUDDS mode (both laden and unladen) 63
Figure 38: PM emissions for the AC5080 mode (both laden and unladen) 64
Figure 39: PM emissions for the Lower Speed Transient mode (both laden and unladen).
65
Figure 40: PM emissions for the Higher Speed Transient mode (both laden and unladen).
65
Figure 41: PM emissions for the Cruise mode (both laden and unladen) 66
Figure 42: PM emissions for the HHDDT_S mode (both laden and unladen) 66
Figure 43: PM emissions for UDDS mode (both laden and unladen) 67
Figure 44: CO emissions for MHDTHI mode (both laden and unladen) 67
Figure 45: CO emissions for the MHDTCRmode (both laden and unladen) 68
Figure 46: HC emissions for MHDTHI mode (both laden and unladen) 68
Figure 47: HC emissions for MHDTCR mode (both laden and unladen) 69
Figure 48: Example of a continuous NOX emissions plot in g/s for E55CRC-42 (56,000
Ibs.) following the HHDDT Transient Mode 70
Figure 49: Example of a continuous HC emissions plot in g/s for E55CRC-42 (56,000
Ibs.) following the HHDDT Transient Mode 70
Figure 50: Example of a continuous CO emissions plot in g/s for E55CRC-42 (56,000
Ibs.) following the HHDDT Transient Mode 71
Figure 51: Example of a continuous exhaust temperature reading for E55CRC-42
(56,000 Ibs.) following the Transient mode schedule 71
Figure 52: Transient mode NOX weight effects 72
Figure 53: Transient mode PM weight effect 73
Figure 54: Cruise modeNOx weight effect 73
Figure 55: Cruise mode PM weight effect 74
Figure 56: Variation of HHDDT CO2 emissions by vehicle model year and by phase of
the program for the UDDS (56,000 Ibs.) 75
Figure 57: Variation of HHDDT CO emissions by vehicle model year and by phase of the
program for the UDDS (56,000 Ibs.) 76
Figure 58: Variation of HHDDT oxides of NOX by vehicle model year and by phase of
the program for the UDDS (56,000 Ibs.) 77
Figure 59: Variation of PM emissions by vehicle model year and by phase of the program
for the UDDS (56,000 Ibs.) 77
Figure 60: PAH in fuels and oils 81
Figure 61: Concentrations of carbonyl compounds (in ppbv) measured during the runs. 84
Figure 62: SMPS Particle Size Distribution for E55CRC-39 Operating on an Idle Mode88
Figure 63: SMPS Particle Size Distribution for E55CRC-39 Operating on Various Steady
Cycles 89
Figure 64: PM Mass Emissions for E55CRC-39 89
Figure 65: Ion Composite Results for E55CRC-39 91
IV
-------
Figure 66: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-39 91
Figure 67: DMS500 Total Number of Particles and Vehicle Speed vs. Time during the
HDDS for E55CRC-39 tested at 56,000 Ibs 92
Figure 68: DMS500 Total Number of Particles and Engine Power vs. Time during the
HDDS for E55CRC-39 tested at 56,000 Ibs 93
Figure 69: DMS500 60 nm and 20 nm Particle Number vs. Time during the HDDS for
E55CRC-39 tested at 56,000 Ibs 93
Figure 70: DMS500 60 nm Particle Number vs. 20 nm Particle Number during the HDDS
for E55CRC-39 tested at 56,000 Ibs 94
Figure 71: DMS500 60 nm Particle Number vs. Total Hub Power during the HDDS for
E55CRC-39 tested at 56,000 Ibs 94
Figure 72: DMS500 20 nm Particle Number vs. Total Hub Power during the HDDS for
E55CRC-39 tested at 56,000 Ibs 95
Figure 73: DMS500 size distribution of particles during Idle (t = 230 s) of the HDDS for
E55CRC-39 tested at 56,000 Ibs 96
Figure 74: DMS500 size distribution of particles during acceleration (t = 530 s) of the
HDDS for E55CRC-39 tested at 56,000 Ibs 97
Figure 75: DMS500 size distribution of particles during deceleration (t = 770 s) of the
HDDS for E55CRC-39 tested at 56,000 Ibs 97
Figure 76: DMS500 size distribution of particles during HHDDT_S and Vehicle Speed
vs. Time for E55CRC-39 tested at 56,000 Ibs 98
Figure 77: DMS500 size distribution of particles during Transient and Vehicle Speed vs.
Time for E55CRC-39 tested at 56,000 Ibs 98
Figure 78: DMS500 size distribution of particles during CruiseS and Vehicle Speed vs.
Time for E55CRC-39 tested at 56,000 Ibs 99
Figure 79: SMPS Particle Size Distribution for E55CRC-40 Operating on an Idle Mode
100
Figure 80: SMPS Particle Size Distribution for E55CRC-40 Operating on Various Steady
Cycles 101
Figure 81: PM Mass Emissions for E55CRC-40 102
Figure 82: Ion Composite Results for E55CRC-40 102
Figure 83: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-40 103
Figure 84: SMPS Particle Size Distribution for E55CRC-41 Operating on an Idle Mode
104
Figure 85: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-41 105
Figure 86: Total PM Mass Emissions Rate (g/hour) for E55CRC-41 106
Figure 87: SMPS Particle Size Distribution for E55CRC-41 Operating under Cruise and
Steady Conditions 106
Figure 88: SMPS Particle Size Distribution for E55CRC-41 Operating on Cruise and
Steady Conditions (50% GVWR) 107
Figure 89: Ion Composite Results for E55CRC-41 107
Figure 90: SMPS Single Particle Trace (82 nm) for E55CRC-41 Operating on HDDS,
Low Transient and High Transient Cycles (75%GVWR) 108
Figure 91: SMPS Single Particle Trace (24 nm) for E55CRC-41 Operating on Low
Transient and High Transient Cycles (75% GVWR) 109
v
-------
Figure 92: SMPS Single Particle Trace (88 nm) for E55CRC-41 Operating on Low
Transient and High Transient Cycles (50% GVWR) 109
Figure 93: DMS500 Total Number of Particles and Vehicle Speed vs. Time during
HDDS for E55CRC-41 tested at 56,000 Ibs 110
Figure 94: DMS500 60 nm and 20 nm Particle Number vs. Time for E55CRC-41 tested
at 56,000 Ibs Ill
Figure 95: DMS500 Total Number of Particles and CO vs. Time for E55CRC-41 tested at
56,000 Ibs 112
Figure 96: DMS500 60 nm Particle Number vs. 20 nm Particle Number for E55CRC-41
tested at 56,000 Ibs 112
Figure 97: DMS500 20 nm Particle Number vs. Total Hub Power for E55CRC-41 tested
at 56,000 Ibs 113
Figure 98: DMS500 60 nm Particle Number vs. Total Hub Power for E55CRC-41 tested
at 56,000 Ibs 113
Figure 99: DMS500 size distribution of particles during acceleration and deceleration for
E55CRC-41 tested at 56,000 Ibs 114
Figure 100: SMPS Particle Size Distribution for E55CRC-42 Operating on an Idle Mode
115
Figure 101: SMPS Particle Size distribution for E55CRC-42 operating on various steady
cycles 115
Figure 102: PM Mass Emissions for E55CRC-42 116
Figure 103: Ion Composite Results for E55CRC-42 117
Figure 104: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-42 117
Figure 105: DMS500 Total Number of Particles and Vehicle Speed vs. Time on Transient
mode for E55CRC-42 tested at 56,000 Ibs 118
Figure 106: SMPS Particle Size Distribution for E55CRC-43 Operating on an Idle Mode
119
Figure 107: SMPS Particle Size Distribution for E55CRC-43 Operating on Various
Steady Cycles 119
Figure 108: PM Mass Emissions for E55CRC-43 120
Figure 109: Ion Composite Results for E55CRC-43 121
Figure 110: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-43 121
Figure 111: DMS500 Total Number concentration of Particles and Vehicle Speed vs.
Time for E55CRC-43 tested at 56,000 Ibs 122
Figure 112: SMPS Particle Size Distribution for E55CRC-44 Operating on an Idle Mode
123
Figure 113: SMPS Particle Size Distribution for E55CRC-44 Operating on Various
Steady Cycles 123
Figure 114: PM Mass Emissions for E55CRC-44 124
Figure 115: Ion Composite Results for E55CRC-44 125
Figure 116: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-44 125
Figure 117: Elemental Carbon Emissions (Percentage of Total Carbon) 127
Figure 118: Inorganic Ionic Species Emissions 128
Figure 119: Lubrication Oil Emissions 129
Figure 120: Engine Wear Emissions 130
VI
-------
Figure 121: NOX/CO2 ratios for each run on the 25 vehicles, both transient and cruise.
One vehicle was found to have high NOX emissions on two runs of the cruise mode
136
Figure 122: NOX emissions plotted against the dispersed and time aligned power for
vehicle E55CRC-10. Most emissions during the Cruise mode were at an off-cycle
level. The vehicle was targeted for Reflash as a result of these data 137
Figure 123: PM/CO2 ratios for Cruise Mode operation of average value for each vehicle.
There are two runs each for E55CRC-1 to -13, and three runs each for E55CRC-14
to-25 138
Figure 124: PM/CO2 ratios for the Transient Mode operation of average value for each
vehicle. There are 2 runs each for E55CRC-1 to -13, 3 runs each for E55CRC-14 to -
25. One run for E55CRC-20 was anomalous low, but no cause was evident 139
Figure 125: Occurrence of criterion ratios by bin for the Laden Transient Mode
(PM/CO2). E55CRC-16 was not included in this analysis, because it was an extreme
outlier 140
Figure 126: Occurrence of criterion ratios by bin for the Laden Cruise mode (PM/CO2).
E55CRC-16 was not included in this analysis because it was an extreme outliner.141
Figure 127: Transient Mode NOX emissions before and after reflashing E55CRC-10... 153
Figure 128: NOX emissions versus power for E55CRC-10 after reflashing. The mass rate
of NOX emissions was reduced in many parts of the Cruise mode (compared to
Figure 122). The highest mass rate emissions were also slightly lower than before
thereflash 154
Figure 129: NOX emissions on Cruise mode (56,000 Ibs.) for three vehicles before reflash.
NOX emissions and hub power were time-aligned before plotting 158
Figure 130: NOX emissions on the three reflash vehicles were reduced in Cruise mode
after reflash. Test weight was 56,000 Ibs 158
Figure 131: As-received and retest after reflash NOX data for E55CRC-26 (56,000 Ibs.).
159
Figure 132: As-received and retest after reflash PM data for E55CRC-26 (56,000 Ibs.).
159
Figure 133: PM emission for E55CRC-28 for the HDDS, Cruise and HHDDT_S Modes.
160
Figure 134: PM emissions for the Transient and Creep Modes for E55CRC-28 161
Figure 135: NOX emissions for E55CRC-28 at 56,000 Ibs. on the UDDS and the Cruise
and HHDDT_S Modes 162
Figure 136: NOX emissions for E55CRC-28 at 56,000 Ibs. on the Transient and Creep
Modes 162
Figure 137: As-received and reflash emissions forNOx emissions for E55CRC-31 tested
at 56,000 Ibs 164
Figure 138: As-received and reflash emissions for PM emissions for E55CRC-31 164
Figure 139: shows the inability of E55CRC-57 to meet the target trace 167
vn
-------
LIST OF TABLES
Table 1: Highest, lowest, and average emissions of NOX and PM for all HHDDT (56,000
Ibs.) 2
Table 2: Emissions of regulated species by MY for HHDDT tested on the UDDS (56,000
Ibs.) 3
Table 3: Highest, lowest, and average emissions of NOX and PM for all MHDDT tested at
75%GVWR 3
Table 4: The effect of changing test weight on the diesel-fueled MHDT, in terms of a
simple average of emissions for that fleet 4
Table 5: Vehicles planned for recruiting in the E-55/59 program 10
Table 6: Federal and ARB past and present emissions standards 11
Table 7: Vehicle Selection Matrix: NOX, PM, VMT population (as percentages of total).
11
Table 8: Basic Information on the 78 trucks actually recruited 12
Table 9: Sampling system description and media type 21
Table 10: Test Schedule Summary. AC5080 performance depends on truck power to
weight ratio 24
Table 11: AC5080 Methodology 26
Table 12: Test Vehicle Details 44
Table 13: List of Target Analytes in the Particulate and Semi-Volatile Fractions 78
Table 14: List of Gas-phase Compounds Quantified by GC/MS Method from Canisters.
82
Table 15: List of Target Analytes in the Gas-Phase Carbonyl Compound Fraction 83
Table 16: Nitrosoamines Targeted for Analysis in the Vehicle Exhaust 84
Table 17: List of Heavy Hydrocarbons for Tenax Samples 85
Table 18: Damaged Tenax Samples 86
Table 19: Emissions level criteria used to declare a truck to be a high NOX emitter 134
Table 20: Definition of criterion ratios by bin for the Laden Transient Mode: see Figure
125 140
Table 21: Definition of criterion by bin for the Laden Cruise Mode: See Figure 126. ..141
Table 22: T&M Determination Data for the HHDDT Vehicles over the Cruise Mode. 143
Table 23: T&M Determination Data for the HHDDT Vehicles over the Transient Mode
147
Table 24: E55CRC-3: UDDS (g/mile) 151
Table 25: E55CRC-3 retest UDDS (g/mile) 151
Table 26: E55CRC-3: 1828-1 Idle, 1828-2 Creep, 1828-3 Transient, 1828-4 Cruise.
(g/mile, exceptidle in g/cycle) 151
Table 27: E55CRC-3: 1829-1 Idle, 1829-2 Creep, 1829-3 Transient, 1829-4 Cruise.
(g/mile, exceptidle in g/cycle) 152
Table 28: E55CRC-3 retest: 1939-1 Idle, 1939-5 Creep, 1939-3 Transient, 1939-4 Cruise
(g/mile, exceptidle in g/cycle) 152
Table 29: E55CRC-3 retest: 1942-1 Idle, 1942-5 Creep, 1942-3 Transient, 1942-4 Cruise.
(g/mile, exceptidle in g/cycle) 152
Vlll
-------
Table 30: E55CRC-10 baseline emissions and "reflash" emissions. T&M runs are
denoted in bold. NOx1 and NOX2 show values from two similar analyzers in parallel.
It should be noted that Idle and Creep Modes often return highly variable emissions
due to changing auxiliary loads on the vehicle 154
Table 31: E55CRC-16 Emissions before and after T&M procedures. T&M 1 in bold
represents testing with a new fuel injection pump; T&M 2 in bold italics represents
testing with a new pump and EGR totally disabled 155
Table 32: E55CRC-21 emissions T&M results in bold 157
Table 33: Sequence of testing for E55CRC-31 163
Table 34: E55CRC-45 as received, after first repair (R), and after second repair (RR).166
IX
-------
LIST OF APPENDICES
Appendix A: Photograph and Description of Each Test Vehicle
Appendix B: WVU Test Vehicle Information Sheets
Appendix C: Description of the WVU Transportable Heavy Duty Vehicle Emissions
Testing Laboratory (Translab)
Appendix D: WVU Vehicle Inspection Sheets
Appendix E: WVU Tampering and Malmaintenance (T&M) Issues Sheets
Appendix F: Engine Control Unit Downloads
Appendix G: DRI Methods
Appendix H: Graphical Representation of Each Test Schedule
Appendix I: SMPS and CPC reduction program
Appendix J: Table of Test Runs
Appendix K: WVU Short Reports for Each Test
Appendix L: WVU Summary Data Tables
Appendix M: Graphical Representation of Data Found in the WVU Short Reports
Appendix N: Chemical Speciation Results
Appendix O: List of Target Chemical Species Analyzed
Appendix P: Quality Assurance and Control
x
-------
EXECUTIVE SUMMARY
Program E-55/59 had the objective of acquiring regulated emissions measurements (for
the whole test fleet) and non-regulated emissions measurements (on a subset of the test
fleet) from in-use trucks in southern California. The project was conducted in four Phases
(denoted 1, 1.5, 2 and 3). The Phase 1 test fleet consisted of 25 Heavy Heavy-Duty Diesel
Trucks (HHDDT), selected to match a model year (MY) distribution developed by the
sponsors and to reflect engines in common use in California. In Phase 1.5 an additional
twelve HHDDT were studied, with a thirteenth truck tested on idle alone. The Phase 2
test fleet consisted of ten HHDDT and nine Medium Heavy-Duty Trucks (MHDT),
which included seven diesel - fueled MHDT (MHDDT) and two gasoline-fueled MHDT
(MHDGT). Phase 3 gathered data from nine HHDDT, eight MHDDT, and two MHDGT.
The Phase 2 and 3 data were valuable in adding post-2002 MY (2.5 g/bhp-hr NOX
standard) HHDDT to the E-55/59 program.
Test cycles for all HHDDT were the Urban Dynamometer Driving Schedule (UDDS) and
the HHDDT schedule, and some trucks were tested using the AC50/80. The HHDDT
Schedule consisted of four modes, namely Idle, Creep, Transient and Cruise, and a high
speed" cruise mode (HHDDT_S) was added for phase 1.5, 2 and 3. The MHDT were all
tested using a Lower Speed Transient Mode (MHDTLO), a Higher Speed Transient
Mode (MHDTHI) and a Cruise Mode (MHDTCR) of a recently developed California Air
Resources Board (CARB) MHDT Schedule, and three MHDT were tested using the
High-speed HHDDT_S mode of the HHDDT Schedule as well.. All HHDDT were tested
at 56,000 Ibs weight, with subsets tested at 30,000 Ibs. and 66,000 Ibs. as well. MHDT
were tested at 50% and 75% of Gross Vehicle Weight Rating (GVWR). The first thirteen
HHDDT in Phase 1 were subjected to sampling for non-regulated emissions, and the
samples from the first three of these trucks were analyzed. Non-regulated species were
measured from five HHDDT and one MHDDT in the Phase 2 test fleet. The HHDDT
emissions were characterized using the West Virginia University (WVU) Transportable
Heavy-Duty Emissions Testing Laboratory (TransLab).
Emissions were diluted in an 18 inch diameter tunnel, with flow controlled by a critical
flow venturi. Gaseous emissions were measured with research grade analyzers, and total
PM was measured from dilute exhaust using 70mm filters. PM data were acquired in
Phases 1.5, 2 and 3 using both conventional filters and a Tapered Element Oscillating
Microbalance (TEOM). The TEOM correlated well with the filters in each phase. Two
chemiluminescent analyzers were used to assure quality of NOX measurements.
The regulated emissions data for HHDDT (56,000 Ibs. test weight) were compared
between phases. Carbon dioxide data agreed well between phases, suggesting that vehicle
loading was consistent from phase to phase. Data for CO, NOX and PM were compared
between phases. There were too few trucks, distributed over a wide range of model years,
to reach emissions conclusions based on data from a single phase. For example, the Phase
2 fleet included only one truck in the 1991-1993 model year range, and its NOX emissions
were substantially lower than the NOX emissions from the five trucks in the 1991-1993
-------
model year range tested in Phases 1 and 1.5. The value of using the combined emissions
data from all phases to reach conclusions on age or model year effects was evident.
Table 1 provides a summary of NOX and PM data averages and ranges for the HHDDT
fleet in all Phases, and for all MY, by test cycle or mode, at 56,000 Ibs. test weight.
Oxides of nitrogen (NOX) varied fourfold on the Transient Mode, but sevenfold on the
Cruise Mode, reflecting both off-cycle behavior and recent stringent NOX standards. The
high-speed HHDDT_S mode also showed a sevenfold range of NOX. Average values for
NOX, in distance specific units, declined with respect to average cycle or mode speed. PM
variation (geometrically) was greater than NOX variation, with trucks producing
emissions at the laboratory lower detection limit on all modes, but with the maximum
values between four and eighteen times the average value for each mode or cycle.
Table 1: Highest, lowest, and average emissions of NOX and PM for all HHDDT
(56,000 Ibs.).
Highest NOX
Average NOX
Lowest NOX
Highest PM
Average PM
Lowest PM
Idle
(g/min)
4.43
1.10
0.00
0.71
0.04
0.00
HHDDT
Creep
(g/mile)
162.94
56.68
7.77
67.77
5.25
0.00
HHDDT
Transient
(g/mile)
60.22
23.50
13.65
17.14
2.65
0.00
HHDDT
Cruise
(g/mile)
49.43
18.07
7.03
17.14
1.08
0.00
HHDDT
HHDDT S
(g/mile)
55.75
17.94
8.05
3.83
0.85
0.00
UDDS
(g/mile)
53.35
21.67
11.98
12.99
1.78
0.00
Emissions were influenced by MY, as shown in Table 2, for the UDDS. Fuel consumed
(by carbon balance from CO2) was highest for the oldest MY bin. For the remaining
vehicles the 1999 and later MY vehicles used more fuel than the pre-1999 MY vehicles.
2003 and later MY HHDDT used 16% more fuel on the UDDS than 1975-1998 MY
trucks, on average. HHDDT for the three oldest MY bins produced about 27 g/mile NOX
on the UDDS. NOX was lower for intermediate MY trucks, rising back to about 27 g/mile
in 1998 MY trucks. For 1999-2002 MY, NOX was reduced to 19.22 g/mile, and, for 2003
and later MY, NOX was reduced to 13.74 g/mile on the UDDS. There was a strong trend
showing that PM was reduced for the HHDDT on the UDDS as MY advanced, with the
1986 and older MY fleet at 2 to 5 g/mile, and 2003 and later MY trucks at 0.50 g/mile.
The trend for CO followed the PM trend. Fleet-averaged HC emissions on the UDDS
showed a substantial reduction in the mid '90s.
-------
Table 2: Emissions of regulated species by MY for HHDDT tested on the UDDS
(56,000 Ibs.).
Pre-1975
1975-1979
1980-1983
1984-1986
1987-1990
1991-1993
1994-1997
1998
1999-2002
2003+
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
Average
Std Dev.
CO
g/mile
12.21
1.82
33.53
29.19
21.31
11.60
16.40
8.36
13.50
6.11
8.45
4.90
7.59
4.25
7.05
2.36
9.89
12.74
2.33
2.44
CO2
g/mile
2766
21
2298
206
2225
243
2174
80
2111
183
2071
71
2142
218
2206
100
2419
314
2529
254
NOX
g/mile
28.46
3.53
26.41
10.98
27.40
2.99
18.67
8.71
19.42
3.67
18.00
2.21
23.77
6.23
26.87
12.36
19.22
5.54
13.74
1.36
HC
g/mile
6.11
1.94
1.98
0.47
2.11
0.06
2.44
1.37
1.63
1.82
2.64
3.67
1.24
1.85
0.69
0.48
0.88
0.74
0.54
0.24
PM
g/mile
3.47
0.55
4.94
5.33
2.18
1.11
2.30
1.03
1.86
0.86
1.10
0.50
0.76
0.56
0.75
0.22
1.04
1.30
0.50
0.47
Table 3 presents an overview of NOX and PM emissions for the MHDDT. NOX emissions
from MHDDT varied less than for the HHDDT fleet. For laden operation on the all
cycles and modes, the highest NOX emitter was about twice the level of the lowest
emitter. PM variation was more substantial for the MHDDT; the highest PM mass
emissions were four or five times the lowest emissions. MHDDT showed little change in
NOX between tests at laden and unladen weight, as shown in Table 4. The increase of
50% in test weight produced an average NOX increase of only 11% on the UDDS, and
16% on the MHDTLO. There was virtually no weight effect on NOX emissions for the
MHDTCR for the MHDDT. The mode with the highest weight influence on PM was the
MHDTLO (40% increase). The mode with the lowest weight influence on PM was the
MHDTCR (7% increase).
Table 3: Highest, lowest, and average emissions of NOX and PM for all MHDDT
tested at 75% GVWR.
Highest NOX
Average NOX
Lowest NOX
Highest PM
Average PM
Lowest PM
MHDT AC5080
(g/mile)
16.50
9.81
6.49
2.52
0.79
0.16
MHDTCR
(g/mile)
14.03
10.48
8.74
1.72
0.71
0.19
MHDTHI
(g/mile)
20.87
12.95
8.85
2.95
0.94
0.20
MHDTLO
(g/mile)
33.05
20.20
12.20
4.38
1.48
0.25
UDDS
(g/mile)
18.82
12.71
8.36
4.13
1.10
0.19
-------
Table 4: The effect of changing test weight on the diesel-fueled MHDT, in terms of a
simple average of emissions for that fleet.
Average Laden NOX
Average Unladen NOX
Average Laden PM
Average Unladen PM
MHDT AC5080
(g/mile)
9.81
8.85
0.79
0.62
MHDTCR
(g/mile)
10.48
10.13
0.71
0.66
MHDTHI
(g/mile)
12.95
11.20
0.94
0.72
MHDTLO
(g/mile)
20.20
17.40
1.48
1.06
UDDS
(g/mile)
12.71
11.44
1.10
0.82
A Tampering and Malmaintenance (T&M) program was initiated in Phase 1 to identify
the incidence of high emitters and to determine the effect of repair on high emitters.
Vehicles were selected based either on an inspection, or on the measurement of PM and
NOX emissions above thresholds that took into account MY emissions standards. A 1985
MY truck showed reduction of PM and increase in NOX after a mechanical timing fault
was repaired. A 1979 truck was a prodigious PM emitter, and was found to defy effective
repair. A 1990 truck with a plugged air filter showed 27% PM reduction on the UDDS
with a new air filter. A 1993 truck with high HC and PM was repaired twice but
emissions remained high. A defective temperature sensor in one MHDT caused emissions
effects, but it was not subjected to repair. A 1998 MY HHDDT with high NOX emissions
was reflashed in Phase 1, and NOX was reduced during cruising behavior. Reflash of three
additional HHDDT in Phase 1.5 showed NOX reductions of 13.3%, 21.5% and 30.7% on
the Cruise Mode of the HHDDT schedule. One of these vehicles was also repaired by
replacement of a manifold air pressure (MAP) sensor, which also affected NOX
emissions.
Size distributions of PM were characterized from six trucks that were selected for non-
regulated emissions measurement. These data were acquired using a Scanning Mobility
Particle Sizer (SMPS), as well as a Differential Mobility Spectrometer (DMS500), which
was a newly released instrument. A raw exhaust slipstream was diluted by a factor of
thirty in a mini-dilution tunnel, and this dilute stream was sampled by the SMPS. The
DMS used the main dilution tunnel of the WVU TransLab, which had a fixed flow and a
varying dilution ratio. The SMPS detected a bimodal distribution with a nuclei mode for
Idle operation of one 2004 MY truck (E55CRC-40), and the DMS 500 detected both the
nuclei mode and accumulation mode during deceleration on this truck. The nuclei mode
was not evident under load for E55CRC-40. The other 2004 MY truck had only one
mode, with a very low particle count on Idle. A 1989 MY truck had a bimodal Idle
distribution, but the remaining HHDDT were unimodal. Comprehensive data were
acquired for steady operation using the SMPS on all trucks, and the DMS500 acquired
transient distributions for four of the trucks.
Sampling for chemical speciation was performed on thirteen HHDDT in Phase 1 and on
five HHDDT and one MHDDT in Phase 2. In Phase 1, only three of the thirteen trucks
had the samples analyzed, and the remaining samples were archived. The exhaust from
the TransLab tunnel was fed to a residence time chamber of the Desert Research Institute
(DRI). Data were acquired for methane and volatile organic compounds using a canister
and a field gas chromatograph. Semi-volatile organic compounds were captured in
-------
PUF/XAD media and PM soluble fractions were captured on Teflon-Impregnated Glass
Fiber Filters (TIGF) filters, and these were extracted and analyzed at the DRI laboratory.
Carbonyls were captured using DNPH cartridges, and nitrosamines were captured in
Thermosorb cartridges. Ions and Elemental/Organic carbon (EC/OC) split were
determined from quartz filters. Results from the speciation data are legion, and examples
include the fact that the EC/OC split differed substantially on Idle between the two 2004
MY trucks equipped with EGR, and that the ion and metal analyses varied widely
between trucks. The total carbon emissions from cruise mode were found to be the
highest, while the total PM mass emissions rates were the highest from the transient
mode. All pre-1999 vehicles, which were subjected to speciation analysis, were found to
emit higher levels of engine wear elements, such as iron, than the newer vehicles.
Likewise, oil control, as demonstrated by the lubricating oil-based ash components, was
better in post-1999 vehicles.
-------
NOMENCLATURE
CA California
CARB California Air Resources Board
CFR Code of Federal Regulations
CO Carbon Monoxide
CC>2 Carbon Dioxide
CRC Coordinating Research Council
CSHVR City Suburban Heavy Vehicle Route
CTA California Trucking Association
DAS Data Acquisition System
DRI Desert Research Institute
ECU Engine Control Unit
EMFAC Emissions Inventory Model employed by CARB
GVWR Gross Vehicle Weight Rating
HC Hydrocarbons
HEPA High Efficiency Particulate Air
HHDDT Heavy Heavy-Duty Diesel Truck
HHDDT_S HHDDT Short (referring to a shortened high-speed cruise mode)
MEMS Mobile Emissions Measurement System
mg Milligrams
MHDDT Medium Heavy-Duty Diesel Truck
MHDGT Medium Heavy-Duty Gasoline Truck
MHDT Medium Heavy-Duty Truck
MHDTCR MHDT Cruise Mode
MHDTHI MHDT High-speed Transient Mode
MHDTLO MHDT Low Speed Transient Mode
MY Model Year
N2O Nitrogen Dioxide
NIST National Institute of Standards and Technology
NOx Oxides of Nitrogen
PA Power Absorber
PM Particulate Matter
PSVOC Particulate/Semi-Volatile Organic Compound
QA/QC Quality Assurance & Quality Control
RH Relative Humidity
SOP Standard Operating Procedures
T&M Tampering and Malmaintenance
TEOM Tapered Element Oscillating Microbalance
TRANS3 HHDDT Transient
Translab Transportable Heavy Duty Vehicle Emissions Testing Laboratory
UDDS Urban Dynamometer Driving Schedule
USDOE United States Department Of Energy
VMT Vehicle Miles Traveled
VOC Volatile Organic Compounds
WVU West Virginia University
-------
INTRODUCTION AND OBJECTIVES
Heavy-duty diesel vehicles are known to be substantial contributors to the inventory of
oxides of nitrogen (NOx) and particulate matter (PM), but the quantification of their real
world emissions may be imprecise. As a result, emissions inventory predictions may err.
To remedy this uncertainty, Program E-55/59 was initiated. This program, consisting of
Phases 1, 1.5, 2, and Phase 3, was managed by the Coordinating Research Council
(CRC), and supported by the following sponsors:
Coordinating Research Council, Inc.
California Air Resources Board
United States Environmental Protection Agency
United States Department of Energy Office of FreedomCar & Vehicle Technologies
through the National Renewable Energy Laboratory
South Coast Air Quality Management District
Engine Manufacturers Association
The primary objective of the E-55/59 research program was to quantify regulated and
certain unregulated gaseous and PM emissions for heavy-duty (primarily diesel) vehicles
in the South Coast Air Basin to support emissions inventory development. Another
objective of the E-55/59 program was to quantify the influence of T&M on heavy-duty
vehicle emissions. WVU and DRI also teamed to address the measurement of emissions
in a comprehensive effort that carefully considered a representative fleet and
representative vehicle activity. The overall E-55/59 research program involved both the
medium and the heavy-duty chassis dynamometer operated together with a dilution and
sampling system to address both regulated and unregulated species.
WVU characterized exhaust emissions from a total of 25 Heavy Heavy-Duty Diesel
Trucks (HHDDT) in California in Phase 1 of the E-55/59 study. A T&M study was also
developed in Phase 1. The first three vehicles were so-called "overlap" vehicles and
were evaluated both under the USDOE "Gasoline/Diesel PM Split Study" and the E-
55/59 study. The overlap vehicles were sampled for both regulated and unregulated
emissions and the unregulated emissions samples were analyzed. The next ten vehicles in
Phase 1 underwent testing for regulated emissions but the extent of sampling for
chemical characterization was reduced and these samples were archived for possible
chemical analysis at a later stage. The remaining twelve vehicles in Phase 1 were tested
only for regulated emissions and the PMio fraction.
Phase 1 of the research program covered only heavy heavy-duty (gross vehicle weight
rating greater than 33,000 Ibs.) vehicles. Test cycles included the UDDS, and four modes
of the HHDDT schedule developed for the E-55/59 program and the AC5080 cycle.
Phase 1.5 of the program had the objective of acquiring regulated emissions
measurements from twelve in-use trucks in southern California, and supporting a third-
party characterization of certain non-regulated species from five vehicles in the Phase 1.5
-------
test fleet. Processing of these third-party data is not addressed in this report. In addition,
three vehicles were re-flashed with a post consent decree engine map, and re-tested. One
of the re-flash vehicles was also repaired and re-tested. A thirteenth vehicle was added to
the Phase 1.5 study when difficulties were encountered in testing a new 2004 model year
(MY) truck with intelligent traction control on the chassis dynamometer. The thirteenth
vehicle was also a 2004 MY truck selected to satisfy the requirements for MY
distribution of the test plan for Phase 1.5. A high speed mode was added to the HHDDT
schedule in Phase 1.5.
Phase 2 had the objective of acquiring regulated emissions measurements from nineteen
in-use trucks in southern California. Ten were HHDDT, seven were Medium Heavy-Duty
Diesel Trucks (MHDDT) and two were Medium Heavy-Duty Gasoline Trucks
(MHDGT). The Phase 2 effort supported a DRI characterization of non-regulated species
from five HHDDT and one MHDDT (total of 6) in the Phase 2 test fleet. Non-regulated
exhaust characterization included speciation of both particulate matter (PM) and volatile
organic compounds (VOC), and determination of particle size distribution. For the
MHDDT and MHDGT, the AC5080 schedule was combined with a three mode Medium
Heavy-Duty Truck (MHDT) schedule consisting of a low speed transient, a high-speed
transient, and a cruise cycle. A twentieth vehicle was added to the study when
difficulties were encountered in testing a 2002 MY truck on the WVU medium-duty
chassis dynamometer.
The final phase, Phase 3, had the objective of acquiring both regulated and unregulated
species from a selected fleet of California trucks. Nine trucks were HHDDT, eight were
MHDDT, and two were MHDGT. For the HHDDT, test cycles were the UDDS and the
HHDDT schedule. For the MHDT, the UDDS and MHDDT schedule were used.
The overall CRC E-55/59 program was planned to include sixty three HHDDT,
seventeen MHDDT, and four MHDGT for measurement of regulated emissions.
Regulated emissions were collected from all vehicles. In addition, non-regulated
emissions were collected and analyzed on eight HHDDT and one MHDDT.
-------
VEHICLE PROCUREMENT
Table 5 shows the combined truck recruiting objectives for all phases. This table lists the
vehicles planned for each phase of the E-55/59 program broken down by vehicle MY,
weight, fuel, and whether vehicles were to be subjected to non-regulated emissions
analysis.
These vehicles complied with a MY distribution determined by the Air Resources Board
and Coordinating Research Council. HHDDT were considered to be vehicles with a
gross vehicle weight rating (GVWR) over 33,000 Ibs., as well as any "full size" single
axle tractors, since these typically have a Gross Combination Weight of 52,000 Ibs. to
80,000 Ibs. The term "full size" was intended to exclude any medium heavy-duty "hot
shot" combinations that have become more common on the highways over the last
decade.
The rationale for basing recruitment on MY is discussed below. The most important
variable that influences emissions was believed to be the engine certification standard.
This is reflected in the engine MY and the vehicle MY corresponds closely to the engine
MY. The vehicle MY may not reflect the appropriate standard in the unusual case of a
vehicle re-power when an engine of a newer standard may be installed. Rebuilds will
normally return the engine to its original condition, whereas a re-power will usually
employ a newer technology engine. The engine MY influences the level of emissions for
certification, as shown in Table 6, which presents the Federal and California standards for
such engines. At time of manufacture, vehicle MY is either the same as engine MY, or
one year later.
The California Air Resources Board (CARB) used the EMFAC emissions inventory
program to generate data for vehicle population, VMT, NOx production and PM
production. These CARB data (each weighted 25%) were combined to arrive at a test
vehicle distribution by MY.
Table 7 shows the VMT, NOx, PM and population distributions. WVU was instructed by
ARB and CRC to employ these distributions in the execution of E-55/59. The Table 7
matrix was processed further to yield an integer number in each MY group. The target
MY distribution was altered in Phase 2 and 3 to allow the inclusion of 2002-2004 MY
trucks.
-------
Table 5: Vehicles planned for recruiting in the E-55/59 program.
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
HHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDDT
MHDGT
MHDGT
MHDGT
MHDGT
MHDGT
MHDGT
MHDGT
MHDGT
MHDGT
MHDGT
Overall
Total
pre-1975
1975-76
1977-79
1980-83
1984-86
Pre-1987
(Phase
1.5 only)
1987-90
1987-92
(Phase
1.5 only)
1991-93
1994-97
1998
99-02
03+
Subttl
pre-1975
1975-76
1977-79
80-83
84-86
87-90
1988-
1990
91-97
(Phase 3
Only)
91-93
94-97
98
99+
(Phase 2
Only)
99-02
(Phase 3
Only)
00+
Subttl
pre-1975
1975-76
1977-79
80-83
84-86
87-90
1988-
1990
91-97
(Phase 3
Only)
00+
Subttl
Phase 1
Reg
Species
Qty
1
1
1
2
3
0
4
0
3
5
2
3
0
25
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
25
Chemical
analysis
qty1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
0
Post-Oct '02
2.5g/bhp-hr
NOx2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
0
Phase 1.5
Reg Species
Qty
0
0
0
0
0
1
0
1
3
8
0
13
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
13
Chemical
analysis
qty1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
0
Post-Oct '02
2.5g/bhp-hr
NOx2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
0
Phase 2
Reg
Species
Qty
0
0
0
0
1
0
2
0
1
0
1
1
2
10
0
0
0
1
0
1
0
0
1
1
1
-)
0
0
7
0
0
0
0
1
0
0
0
1
2
19
Chemical
analysis
qty1
0
0
0
0
0
0
1
0
0
1
0
1
2
5
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
6
Post-Oct
'02
2.5g/bhp-hr
NOx2
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
2
2004
4.0g/bhp-
hr Nox
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Phase 3
Reg
Species
Qty
1
1
0
1
0
0
3
0
0
-)
0
0
1
9
1
0
0
0
1
1
0
0
1
2
1
0
1
0
8
0
0
0
0
0
0
1
1
0
2
19
Subttl
2
2
1
3
4
1
10
1
4
10
6
13
6
63
1
0
0
1
1
2
0
0
2
3
3
3
1
0
17
0
0
0
0
1
0
1
1
1
4
84
10
-------
Table 6: Federal and ARB past and present emissions standards.
FEDERAL HEAVY-DUTY TRUCK STANDARDS
MODEL
YEAR
1974-78
1979-83
1984-87
1988-90
1991-93
1994-97
1998-02
2003+
1 Note: the I
2 Assumes 2
HC1
CO
NOX
PM
HC+NOx
g/bhp-hr
—
1.5
1.3
1.3
1.3
1.3
1.3
0.52
40.0
25.0
15.5
15.5
15.5
15.5
15.5
15.5
—
—
10.7
10.7
5.0
5.0
4.0
2.0
—
—
—
0.60
0.25
0.10
0.10
0.10
16.0
10.0
—
—
—
—
—
—
CALIFORNIA HEAVY-DUTY TRUCK
STANDARDS
MODEL
YEAR
1975-76
1977-79
1980-83
1984-86
1987-90
1991-93
1994-97
1998-02
2003+
HC1
CO
NOX
PM
HC+NOx
g/bhp-hr
—
1.0
1.0
1.3
1.3
1.3
1.3
1.3
0.52
30.0
25.0
25.0
15.5
15.5
15.5
15.5
15.5
15.5
—
7.5
—
5.1
6.0
5.0
5.0
4.0
2.0
—
—
—
—
0.60
0.25
0.10
0.10
0.10
10.0
—
6.0
—
—
—
—
—
—
1C standards shown are total hydrocarbons except for model year 2003+ which is NMHC.
.5 g/bhp-hr (NOX+NMHC) with a 0.5 g/bhp-hr NMHC cap effective October 2002.
Table 7: Vehicle Selection Matrix: NOX, PM, VMT population (as percentages of
total).
pre-1975
1975-76
1977-79
1980-83
1984-86
1987-90
1991-93
1994-97
1998
1999-01
Pop
5.2
1.5
4.8
7.8
12
23.7
13.3
16.6
4.0
11.1
100
VMT
1.2
0.5
1.9
3.9
7.7
19.8
14.2
22.9
7.0
20.9
100
NOX
2.0
0.8
3.1
6.1
9.1
18.5
13.1
24.4
8.5
14.5
100.1
PM
3.4
1.3
5.3
10.7
13.7
21.8
14.1
18.1
4.2
7.4
100
Average
2.95
1.03
3.78
7.13
10.63
20.95
13.68
20.50
5.93
13.48
100.03
The California Trucking Association (CTA), through a survey, obtained information for
use in Phase 1 on the distribution of engine types by MY, and so by certification standard
(MY group). The CTA matrix was then reviewed to identify preferred engine
manufacturers in each MY group. The matrix did contain errors due to imprecise survey
replies (with similar engine models having multiple names, truck models substituted for
engines, some medium duty engines in the matrix and engines assigned to the wrong
manufacturer) and an effort was made to compensate for these errors. The test matrix
was selected considering engine count and in some cases the need to gather some engine
models together or represent a variety of engines.
Many of the vehicles recruited in Phase 1 were procured by CTA. To identify potential
11
-------
study participants, CTA first contacted the survey respondents that had the desired MY
trucks and engines in their fleets and indicated on the survey that they would be willing to
participate in the study. CTA also sent out several faxes soliciting participation and
phoned the entire membership in the Los Angeles/Orange/San Bernardino/Riverside
areas to complete the study. Since not all the members had participated in the survey, the
engine makeup of many fleets was unknown and the CTA staff were "cold calling" to
find the appropriate study engines. One truck, E55CRC-21, was procured from a rental
company by CTA.
A few trucks in Phase 1 were procured directly by WVU. In all subsequent phases of the
study, WVU recruited all of the test vehicles. These vehicles were procured from a
variety of sources including owner-operators, truck fleets, and rental companies. Many
of the trucks in Phase 1.5 and following were procured from used truck dealerships.
Table 8 lists the vehicles which were actually recruited under this program. The vehicles
are broken down by their reference number, weight class and fuel, vehicle MY, vehicle
manufacturer, horsepower, and engine manufacturer. Trucks were designated as
E55CRC-XX according to the order of recruitment. Details are provided in Table 8 for
78 trucks, but three of the trucks recruited were not characterized fully for emissions.
Truck E55CRC-37 was a 2004 Volvo HHDDT with traction control and the truck could
not be operated on the chassis dynamometer. Idle emissions only were obtained for this
truck. Truck E55CRC-67 was a 1975 Ford HHDDT procured with assistance from the
Gateway Cities Program. It was not functional beyond brief periods of idle. Truck
E55CRC-52 was a 2002 Isuzu MHDDT. Subsequent to testing, a detailed quality
assurance review of the data found that coastdown parameters had been incorrect and,
therefore, that the dynamometer loading was significantly in error. The vehicle had been
sold and the new owner would not allow it to be returned for a retest. A substitute
vehicle was found.
Table 8: Basic Information on the 78 trucks actually recruited.
E55CRC-
(truck)
E55CRC-1
E55CRC-2
E55CRC-3
E55CRC-4
E55CRC-5
E55CRC-6
E55CRC-7
E55CRC-8
E55CRC-9
E55CRC-10
E55CRC-11
E55CRC-12
E55CRC-13
E55CRC-14
M=MHDT
H=HHDT
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Vehicle
model
year
1994
1995
1985
2000
2000
1995
1990
1996
1998
1998
2000
1986
1978
1986
Vehicle
Manufacturer
Freightliner
Freightliner
International
International
Freightliner
Freightliner
Peterbilt
Kenworth
Peterbilt
Sterling
Freightliner
International
Freightliner
Freightliner
Engine Model
Series 60
3406B
NTCC-300
C-10
N14-435E1
Ml 1-370
Series 60
Ml 1-300
C12
Series 60
ISM- 11
300
350
LTA10
Engine
Power
(hp)
470
375
300
270
435
370
450
370
410
470
330
300
350
270
Engine
Manufacturer
Detroit
Caterpillar
Cummins
Caterpillar
Cummins
Cummins
Detroit
Cummins
Caterpillar
Detroit
Cummins
Cummins
Cummins
Cummins
12
-------
E55CRC-
(truck)
E55CRC-15
E55CRC-16
E55CRC-17
E55CRC-18
E55CRC-19
E55CRC-20
E55CRC-21
E55CRC-22
E55CRC-23
E55CRC-24
E55CRC-25
E55CRC-26
E55CRC-27
E55CRC-28
E55CRC-29
E55CRC-30
E55CRC-31
E55CRC-32
E55CRC-33
E55CRC-34
E55CRC-35
E55CRC-36
E55CRC-37
E55CRC-38
E55CRC-39
E55CRC-40
E55CRC-41
E55CRC-42
E55CRC-43
E55CRC-44
E55CRC-45
E55CRC-46
E55CRC-47
E55CRC-48
E55CRC-49
E55CRC-50
E55CRC-51
E55CRC-52
E55CRC-53
M=MHDT
H=HHDT
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
M
H
H
H
H
H
H
H
H
M
M
M
M
Vehicle
model
year
1973
1979
1993
1991
1987
1992
1990
1993
1983
1975
1983
1999
2000
1999
2000
1999
1998
1992
1985
2004
2001
2001
2004
2003
2004
2004
1998
2000
1995
1989
1993
1989
1986
1998
1994
2001
1994
2002
2001
Vehicle
Manufacturer
Kenworth
White
Freightliner
Ford
International
Peterbilt
Freightliner
Ford
Peterbilt
Kenworth
Freightliner
Freightliner
Freightliner
Freightliner
Volvo
Freightliner
Kenworth
Volvo
Freightliner
Freightliner
Sterling
Peterbilt
Volvo
Volvo
Volvo
Freightliner
Ford
Freightliner
Peterbilt
Volvo
Volvo GM
Volvo GM
Ford
Freightliner
International
International
International
Isuzu
CMC
Engine Model
NTC-350
3208
L-10
L-10
L-10
Series 60
3406B
L10-280
Plate Not
Available
NTCC-350
Plate Not
Available
C-10
Series 60
Series 60
1SX475ST2
Series 60
N14-460E+
3406B
3406
Series 60
Series 60
C-15
ISX
ISX
ISX
Series 60
B5.9
3406
Series 60
3406
L10-280
3176
6V92
N14 Plus
3406?
DT466 C-195
DT-408
A210F
4HE1-TCS
1GMXH08.15
12 Gas V8
Engine
Power
(hp)
350
200
330
300
300
450
400
280
UNK
350
UNK
270
500
500
450
500
460
280
310
500
470
475
500
530
530
500
210
435
470
300
(est.)
280
400
(est.)
350
447
300
(est.)
195
210
175
270
Engine
Manufacturer
Cummins
Caterpillar
Cummins
Cummins
Cummins
Detroit
Caterpillar
Cummins
Cummins
Cummins
Cummins
Caterpillar
Detroit
Detroit
Cummins
Detroit
Cummins
Caterpillar
Caterpillar
Detroit
Detroit
Caterpillar
Cummins
Cummins
Cummins
Detroit
Cummins
Caterpillar
Detroit
Caterpillar
Cummins
Caterpillar
Detroit
Cummins
Caterpillar
International
International
Isuzu
GM
13
-------
E55CRC-
(truck)
E55CRC-54
E55CRC-55
E55CRC-56
E55CRC-57
E55CRC-58
E55CRC-59
E55CRC-60
E55CRC-61
E55CRC-62
E55CRC-63
E55CRC-64
E55CRC-65
E55CRC-66
E55CRC-67
(not tested)
E55CRC-68
E55CRC-69
E55CRC-70
E55CRC-71
E55CRC-72
E55CRC-73
E55CRC-74
E55CRC-75
E55CRC-76
E55CRC-77
E55CRC-78
M=MHDT
H=HHDT
M
M
M
M
M
M
H
M
H
H
H
H
H
H
M
H
M
M
M(G)
M
H
M
M
M(G)
H
Vehicle
model
year
1983
1992
1988
2000
1982
1990
1995
2000
1983
2005
1994
1988
1989
1975
1995
1989
1998
1995
1992
1974
1969
1975
1993
1987
1975
Vehicle
Manufacturer
Ford
Ford
Ford
Freightliner
Ford
International
Ford
1999
Kenworth
Freightliner
Kenworth
International
International
Ford
International
Ford
Freightliner
Ford
Ford
Ford
Kenworth
Ford
Ford
Chevrolet
Peterbilt
Engine Model
370-2V Gas
V8
210
3208
3126
V8-8.2 4087-
7300 770
LTA10
L10
3126
3406
C15
N14
LTA10
LTA10
UNK
DT466 C-195
C215
LTA10
LTA10
7.0 EFI
1150
UNK
V8-8.2 4087-
7300 C59
8.3L
V8
UNK
Engine
Power
(hp)
296
210
215
330
160
(est.)
300
300
190
350
500
310
240
270
UNK
230
215
210
300
UNK
185
UNK
UNK
UNK
UNK
UNK
Engine
Manufacturer
Ford
Ford
Caterpillar
Caterpillar
Detroit
Cummins
Cummins
Caterpillar
Caterpillar
Caterpillar
Cummins
Cummins
Cummins
Cummins
International
Ford
Cummins
Cummins
Ford
Caterpillar
Cummins
Detroit Diesel
Ford
Chevrolet
Cummins
Once a vehicle was recruited, a WVU driver or WVU commercial driver (when required)
drove the vehicle to the WVU test site. The driver inspected the vehicle before moving
the vehicle: no vehicles were rejected on the basis of this preliminary inspection. A few
vehicles, not licensed at the time of testing, were towed to the test site.
Each vehicle was photographed at the laboratory site. These photographs are gathered in
Appendix A of this report. Basic vehicle information was logged on the Vehicle
Information Forms which are gathered in Appendix B.
14
-------
EQUIPMENT & METHODS - REGULATED SPECIES
Translab Description
The characterization of the emissions took place using a WVU Transportable Heavy Duty
Vehicle Emissions Testing Laboratory (Translab). The Translab incorporated a chassis
dynamometer testbed, full scale dilution tunnel with a critical flow venturi and blower,
70mm PM filtration system and research grade analyzers. For all of Phases 1 and 1.5
and for the first several vehicles of Phase 2, the Translab was located at Ralphs Grocery,
1500 Eastridge Ave., Riverside, CA. The latter part of the testing was completed at 1084
Columbia Ave., Riverside, CA on property leased from the University of California,
Riverside. The Translab had the same basic arrangement in all of phases of the E-55/59
CRC Study. WVU employed two different HHDDT chassis dynamometer test beds, one
for Phase 1 and the other for Phases 1.5, 2, and 3. These test beds were similar in design
and calibrated to provide the same inertia and road-load characteristics. Some MHDTs
were tested using the second HHDDT test bed, while others were tested using WVU's
medium-duty chassis dynamometer. The coastdown methodologies used to determine
vehicle inertial and drag characteristics assured that loading was properly applied
independent of which chassis dynamometer was utilized.
The laboratory equipment, including the gaseous emissions analyzer bench, PM system
and dilution system used, remained the same throughout the program. Procedures with
respect to PM filter processing, analyzer calibration and background correction were the
same for all four phases. For completeness of this report, a description of the WVU
Translab has been included in Appendix C.
Beginning with E55CRC-26, the laboratory used a Tapered Element Oscillating
Microbalance (TEOM) analyzer to quantify continuous PM mass. The TEOM sampled at
a rate of 2 liters/minute from the primary (full scale) dilution tunnel using the same port
as the PM filter.
In Phase 1.5, Phase 2, and Phase 3 research, exhaust temperature was measured in the
first section of test exhaust after the tailpipe using a J-type thermocouple. Exhaust
temperature was not measured in Phase 1 research.
Preparation of the Vehicle for Testing
Vehicle Inspection
Each truck, when received at the test site, was inspected for safety, tampering or
malmaintenance and engine control unit (ECU) status. The safety inspection was
conducted by the truck's test driver (who in most instances also drove the vehicle from its
source). The vehicle was inspected for:
• Exhaust leaks
15
-------
• Air leaks in the brake system
• Other visible brake problems (including frayed lines, damaged slack adjusters)
• Damaged drive tires
• Drivetrain damage (including worn universal joints)
• Loose fan bearing or worn engine drive belts
• Damaged vehicle controls
Vehicle information was collected on the WVU Test Vehicle Information Sheets and they
are included in Appendix B. The vehicle safety inspection was verified by the
completion of WVU Vehicle Inspection Report. These sheets appear in Appendix D.
Each vehicle was also visually inspected for T&M. This inspection procedure was
developed in Phase 1 of this program. Items for inspection appeared in the WVU T&M
Issues sheet for each vehicle. Copies of the WVU T&M Issues Sheets appear in
Appendix E of this report.
The engine control unit (ECU), in vehicles so equipped, was interrogated where possible.
An interface and software for this interrogation, developed for use by the WVU Mobile
Emissions Measurement System (MEMS) and used in other related programs for on-
board vehicle emissions measurement1 was used to log ECU data continuously for
vehicles equipped with a compatible ECU. These data were transmitted to WVU-
Morgantown electronically as soon as they were available. In several cases, relating both
to the system used to read the data and to the vehicle ECU, interrogation was not
possible. All available ECU downloads are presented in Appendix F. In cases where the
same vehicle ECU download was performed on two occasions (and was identical), only
the first download is shown in Appendix F.
Engine/Vehicle Preparation
Prior to being placed on the dynamometer, each vehicle was inspected and the following
inspections and checks were performed and recorded as the T&M report for that vehicle:
• Engine oil, power steering fluid, and coolant levels confirmed to be in operating range.
• Fuel tanks filled to provide sufficient fuel for the entire testing procedure (typically 3/4
tank for multiple weights with repeated runs or Va tank for testing at one or two
weights with single runs).
• Inspection for signs of fuel, oil, and coolant leaks from the vehicle.
• Exhaust after-treatment devices, when installed, inspected and model/type recorded
and compared to specifications provided on engine and vehicle data plates.
• Condition of the air filter checked air cleaner service indicator checked and, if the air
filter was severely blocked, project supervisors and sponsors contacted prior to fitting
with a new filter.
• Fuel sample (32 oz.) and oil sample (4 oz.) take for post testing analysis.
• Exhaust system inspected for leaks.
• Engine checked for excessive/irregular noise and malfunctions in the cooling system.
16
-------
Wheel-Hub Adapter Installation
In order to connect the drivetrain of the vehicle to the dynamometer components, the
outer wheels of the dual wheel set were removed on the forward-most drive axle and
replaced with rims designed to accept specially designed adapters. The inside tires were
checked for any cuts and bulges and tire pressure was verified/adjusted to be within +5
psi of the manufacturer's maximum tire pressure specifications.
For vehicles equipped with a power divider, the power divider was locked in.
Using the information stamped on the vehicle tires, the rolling diameter was determined
and recorded into the vehicles database to allow for the proper dynamometer flywheel
selection.
Certain MHDT were characterized using a medium-duty chassis dynamometer test bed.
In these cases the installation on the dynamometer was conventional. Power was taken
from the vehicle's wheels by the rollers, and no hub connection was employed.
Test Vehicle Mounting
In order to mount the vehicle onto the dynamometer, ramps were installed and the vehicle
was backed on to the test bed with the driven wheels (forward wheel set in the case of
dual axle vehicles) centered over the front set of rollers. The vehicle axles were then
attached to the dynamometer bed using chains to prevent both forward-aft and side-to-
side motion. The hub adapters were then connected to the dynamometer drivetrain and
the front of the vehicle was raised to level the vehicle.
If excess tire side wall deflection was observed (or later if tire heating became an issue)
on vehicles with single rear axles, the rear of the vehicle was supported partly using
jacks. Up to 50% of the normal weight was removed from the single wheel (which
normally would share the weight with another wheel). Scales were used beneath the
jacks to verify the amount of weight reduced.
Connections of Vehicle to Laboratory
A monitor, used by the driver to observe the target and actual vehicle speed, was placed
in the vehicle. The exhaust of the vehicle was routed to the dilution tunnel using
insulated, flexible exhaust tubing.
Pre-test Vehicle Operation
Prior to operating the vehicle, the test engineer estimated the dilution tunnel flow rate that
would provide the optimum balance between providing sufficient flow to maintain the
particulate filter face temperature below 125°F while not over diluting the exhaust such
that emissions concentrations could be accurately measured by the laboratory analyzers.
After starting the dilution tunnel blower, the vehicle engine was started and operated
17
-------
throughout its range while the safety officer inspected the exhaust system (including both
the vehicle exhaust and the exhaust transfer tube) for leaks.
In cases where the vehicle would not start after 10 seconds of cranking, technicians
investigated to determine the cause for the failure and attempted to affect repairs.
At all times, the vehicle was operated a manner representative of in-use operation, and
where appropriate and available, according to the manufacturer's recommendation. For
vehicles equipped with automatic transmissions, idle modes less than one minute in
duration were run with the transmission in "Drive" and the brakes applied. In the case of
vehicles equipped with manual transmissions, the vehicles were run in gear with the
clutch disengaged.
Vehicle Driving Instructions
During high acceleration portions of the trace, the vehicle was driven at maximum power
while in gear with the clutch engaged, but the gears were changed in the same manner
that they would be changed while driving on the road. The gears were not "crashed"
aggressively to keep up with the trace. Conversely, the vehicle was not driven with
casually slow shifting. The objective during acceleration was not solely to match the
trace as closely as possible, but also to mimic the way in which the subject vehicle would
actually accelerate on the road.
During braking events, where that portion of the trace ended at zero velocity (idle), the
driver would downshift during braking at speeds above 15 to 20 miles per hour. Below
15 to 20 miles per hour, brakes alone were used. In many 9 and 10 speed transmissions
with typical over-the-road gear ratios, this would typically mean that downshifting would
continue until the highest gear in low range (5th gear in a 10 speed, 4th gear in a "L plus
8" 9 speed configuration) was attained. When the deceleration did not reach zero, but
blended into another acceleration ramp, or a cruise section, gears were used throughout to
maintain normal on-road driving practice.
The driver selected gears based upon judgment of the gears that would be used on the
road. For example, the driver did not take off in the lowest gear if a higher gear would be
the norm in use. In driving 13 and 18 speed (ranged and split) transmissions, it was not
considered necessary to use all gears and typically splitting was not performed in this
case. However, 10 speed splitter transmissions were not driven as 5 speed units at high
inertial loads.
Vehicle Practice Run
For each vehicle, a practice test was performed. This served to help the driver familiarize
himself with the driving/shifting/braking characteristics of the vehicle while the ranges of
the gaseous emissions analyzers were set. The practice run also allowed the test engineer
to ensure that the dynamometer was loading the vehicle properly.
18
-------
If the driver did not follow the trace properly, variances were recorded in the comments
section of the test report and an explanation was offered. If any abnormality was
observed during this practice run, the supervising engineer was informed.
Test Run Procedures
Overall Vehicle Test Procedure
Background PM samples were taken at the beginning and end of each test day, unless
otherwise designated by program requirements.
To execute a hot test, the truck was warmed through operation on or off the dynamometer
until it had reached an operating temperature that caused the radiator to engage in
cooling. This may be judged by the heat of the radiator or opening of the thermostat.
Alternately, it was considered sufficient that the truck was operated for a cycle with a
minimum duration of 10 minutes with a minimum energy intensity of 7.5 x 1CT6 cycle
energy per mile per mass of vehicle (axle hp-hr/lb.-mile). Cycles meeting this criterion
include the WVU Smile route, the CSHVR, the HDDS, and the Transient and Cruise
modes of the California Heavy Heavy-Duty Diesel Vehicle Test (HHDDT) cycle. It is
acknowledged that when cold weather testing is performed, these cycles may not warm
the vehicle to the point that the radiator is in use but the vehicle was deemed to be at
operating temperature under those ambient conditions. No cold conditions were
encountered during the testing in Riverside. The truck was keyed off 30-60 seconds after
the end of this warm-up.
The test schedule commenced no later than 20 minutes after the warm-up but not before
10 minutes after the warm-up. If the20 minute period was exceeded, another warm-up
was required. These criteria were based on a prior study that evaluated HHDDT schedule
repeatability. Based upon measurements of emissions during the warm-up, analyzers
were re-ranged during this 10 to 20 minute period as required. The vehicle was started
and allowed to idle between 30 and 60 seconds prior to the beginning of the test and, after
the completion of the test, the vehicle was idled for 30 seconds and then keyed off.
During the 10 to 20 minute period between tests, integrated and background bags were
analyzed, particulate filters were changed out and analyzers were checked for drift by
checking their response to samples containing 0% and 100% of the respective analysis
gas.
Gaseous and PM sampling system
WVU's Total-Exhaust Double-Dilution Tunnel
19
-------
Figure 1: WVU Sampling System Schematic.
Figure 1 shows the WVU double dilution tunnel. A description of the various
components on WVU's sampling system is given in Table 9. This table includes details
of the DRI residence chamber and the sampling for non-regulated species from the
chamber.
20
-------
Table 9: Sampling system description and media type.
ID
1
2
o
J
4
5
6
7
8
9
10
Sample Description
HC
CO/C02
NOx
TEOM
TPM
PM10 Fraction for Gravimetric
Analysis
MOUDI
DRI Dilution Chamber.
The chamber diluted exhaust from
WVU's tunnel through two PM2 5
cyclones.
Volatile Organic Compounds (VOC)
Methane
Semi-volatile Organic Compounds
(VOC)
SVOC - Low Molecular weight
Nitro-PAH
Carbonyls
Nitrosamines
PM, SOF, Organic Compounds
Elemental Analysis and
Gravimetric Analysis (PM2 5)
Ammonium and Ions (Nitrate, Nitrite,
Chloride, Sulfate)
EC/OC
Particle Analysis by Mass
Spectroscopy
Real-time PM Size Distribution
Media Type
Phase 1
Heated FID Analyzer
NDIR Analyzers
Chemiluminescent Analyzer
TEOM Filter
70 mm T60A20
URG Model 2000-30 EA
Cyclone @ 28.3 1pm ; 47 mm
TX40HI20WW Filters
Greased Aluminum
Substrates; TX40HI20WW
Afterfilter
Model Bendix-Unico 240
Cyclones@113 1pm
Canister, Field GC
Canister
PUF/XAD and TIGF Filters
TENAX Tubes
PUF/XAD and TIGF Filters
DNPH Cartridges
Thermosorb Cartridges
TIGF Filters
TIGF Filters
Quartz Filters
Quartz Filters
ATOMFS
Phase 2
Heated FID Analyzer
NDIR Analyzers
Chemiluminescent Analyzer
TEOM Filter
70 mm T60A20
URG Model 2000-30 EA
Cyclone @ 28.3 1pm ; 47 mm
TX40HI20WW Filters
Not Applicable
Model Bendix-Unico 240
Cyclones@113 1pm
Canister, Field GC
Canister
PUF/XAD and TIGF Filters
TENAX Tubes
PUF/XAD and TIGF Filters
DNPH Cartridges
Thermosorb Cartridges
TIGF Filters
TIGF Filters
Quartz Filters
Quartz Filters
ATOMFS
DMS 500
Total Paniculate Matter
Total Particulate Matter (TPM), as defined by the US EPA, is sampled by filtering diluted
diesel exhaust at filter face temperatures of less than 52°C (125° F). TPM has been
reported as a regulated emission species. A 70mm filter holder was connected that
contained two 70mm T60A20 fiberglass filters, the primary and secondary filters to
capture the regulated TPM.
21
-------
Gas Analyzer Operation
Pre-Test Procedure
Prior to initiating each test, the background and integrated/dilute bags were evacuated to
ensure that no sample was left in them from the previous test. Since the HC and NOx
analyzers were sensitive to temperature changes, their sample pumps were started at least
1 hour prior to starting an emissions test. The particulate filter holder was loaded with
conditioned and weighed primary and secondary filters and connected to the secondary
dilution tunnel. With the sample pumps and analyzers operating and sampling from the
dilution tunnel, sample flow rates were adjusted to meet specifications and maintain test-
to-test consistency.
In-Test Procedure
At the onset of the data collection phase of each test, the gas operator ensured that the fill
rates for the dilute and background bags were sufficient to provide enough sample for
two analyses if required. The flow rate through the particulate filter was checked against
the setting for the test (typically 2-4 scfm depending on PM loading and outside
temperature). During a test, the engineer and gas operators monitored the responses of the
dynamometer controls and emissions analyzers to ensure that they were operating
properly. A safety operator, stationed at the dynamometer, watched for inconsistencies in
vehicle or dynamometer operation. The most common vehicle problem involved tire
degradation during the test due to overheating which was compensated for by the safety
operator applying a water spray to the tires. The test engineer also reviewed the data
collected during the previous test to ensure that the various lab analyzers and sensors had
operated correctly.
Post-Test Procedure
At the conclusion of the test, the particulate filter holder was disconnected from the
secondary dilution tunnel and the loaded particulate filters were placed stored in an
environmental chamber for conditioning. The integrated and background bags were
analyzed for HC, NOx, CO, CO2 concentration and the levels were recorded. The CO and
CO2 concentrations from integrated and continuous measurements were compared to
ensure that they agreed to within 5%.
Calibration Procedures
Sensors and analyzers were calibrated per the manufacturer's recommendation and per
Title 40, Part 86 of the CFR. Calibration procedures for the entire laboratory were
followed for each analyzer when the calibration gas source was changed or after analyzer
repair.
22
-------
Particulate Matter Filter Conditioning and Weighing
The relative humidity and temperature of the environmental chamber, used to condition
particulate filters before weighing, were maintained at 50% (+/-10%) and 70 °F (+/- 10
°F). Filters were stored in petri dishes that were cleaned using alcohol and lint free cloth
prior to use. Filters were conditioned in the environmental chamber for a minimum of 5
hours before weighing. A set of two particulate filters were kept in the chamber and
weighed each time filters were weighed to check for possible problems with either the
weighing instrument or the environmental chamber. Filter weight was determined using a
CAHN-32 microbalance which was calibrated using a NIST traceable 200 mg calibration
weight.
Speciated analysis of filters, performed by DRI, is discussed in Appendix G.
Forms
A set of forms, shown in the Quality Assurance Program section were filled out by the
field engineers/technicians during vehicle emissions evaluation tests:
Quality Assurance/Quality Control and Emissions Data Report
Field Custody Log
Analyzer Check-off Sheet
Analyzer Calibration Record
Test Vehicle Information Sheet
Vehicle Test Sequence
Vehicle Inspection Form
T&M Issues Form
The "Test Sequence" form was designed to enable WVU field engineers keep track of the
tests that have been completed. The "Test Sequence" depicts tests in the correct
chronological order.
Test sequence numbers are employed by WVU to keep track of all tests in its database.
This is WVU's primary means of accessing test results for any vehicle evaluated in the
past 12 years. Each test sequence number has several run numbers associated with it.
Test Cycles
Each vehicle was tested using a suite of test schedules which varied between phases, and
between HHDDT and MHDT. The target cycles and modes are shown graphically in
Appendix H of this report. The HDDS was used for all classes of vehicles. For the
HHDDT, the HHDDT Schedule2, consisting of four modes (Idle, Creep, Transient and
Cruise), was also used. The trucks were tested at 56,000 Ibs. simulated weight on the
UDDS and the HHDDT Schedule, plus a high-speed cruise mode created for Phase 1.5.
The high-speed cruise mode caused high dilution tunnel and tire temperatures during
testing and was shortened in duration at the beginning of Phase 1.5, after which it was
23
-------
termed the "HHDDT Short" (HHDDT_S). In Phase 1.5, the five modes of the HHDDT
schedule were also used to characterize emissions at 30,000 Ibs. and 66,000 Ibs. test
weights, after preliminary work suggested that the original 75,000 Ibs. maximum target
weight placed too much stress on both truck brakes and the dynamometer. This mode
was intended for use in testing all HHDDT. HHDDT_S has a maximum speed of 67 mph
and an average speed of slightly over 50 mph. In some cases trucks were governed at a
sufficiently low road speed so that the HHDDT_S mode could not be executed
reasonably. In these cases, as judged by the WVU field engineer, no HHDDT_S mode
was attempted.
Table 10 shows the time duration, average speed, and distance covered for each cycle or
mode, and the computer code used to describe that cycle or mode.
Table 10: Test Schedule Summary. AC5080 performance depends on truck power to
weight ratio.
Code
TEST D
IDLE32
CREEP34
TRANS3
CRUISES
HHDDT_S
MHDTLO
MHDTHI
MHDTCR
AC5080
Schedule
UDDS
HHDDT
idle
HHDDT
creep
HHDDT
Transient
HHDDT
cruise
HHDDT
short
MHDT
low speed
transient
MHDT
high-speed
transient
MHDT
cruise
AC5080
Time(sec)
1060
1800
1032
688
2083
760
370
1190
1910
idle 10 sec
50kph 60 sec
SOkph 80 sec
Ave
Speed
18.8
0
1.7
14.9
39.9
49.9
9.4
21.6
40.1
N/A
Distance
(miles)
5.5
0
0.5
2.9
23.1
10.5
1.0
7.1
21.3
N/A
For those HHDDT subject to further chemical analysis, the Idle, Creep, Transient, and
Cruise and higher speed modes were each repeated. In order to collect sufficient mass for
speciation, it was necessary to increase the length of some of the modes. For example,
the mode denoted creep34 consisted of four HHDDT creep (creepS) modes in a sequence
sampled as a single cycle. Similarly idle32 consisted of two idle runs together (i.e. the
idle time was doubled).
24
-------
For the MHDT not subject to speciation, the AC5080 and the MHDT Schedule were
added to the UDDS. The MHDT schedule consisted of three modes: lower speed
transient (MHDTLO), higher speed transient (MHDTHI), and cruise (MHDTCR)3'4 In
Phase 2, for three of these vehicles, the HHDDT_S mode was included in the MHDT
schedule. In Phase 3, all MHDT judged capable of adequately maintaining the necessary
speed ran the HHDDT_S. For the one medium-heavy-duty vehicle subject to further
chemical analysis in Phase 2, idle (not included on the other MHDT vehicles), lower
speed transient, higher speed transient, and cruise modes were each repeated to create
paired data sets at the laden weight.
The AC5080 is a short test proposed as an Inspection and Maintenance (I/M) cycle5. It is
a mixed-mode test having two full-load accelerations and two steady-state cruises at 50
km/hr (31.1 mph) and 80 km/hr (49.7 mph). It is less aggressive than the DT80, but
according to its creators it may be more representative of on-road driving. It requires the
use of an inertia-simulating dynamometer. WVU developed a specific protocol for
implementing the AC5080 Short Test because the previous descriptions of the cycle were
ambiguous in the cycle execution.
The AC5080 test has two full-power accelerations and two steady-state cruises as shown
in Table 11 and Figure 2. Since the time of occurrence of points "C" and "E" are vehicle
and load dependent, the scheduled speed trace (the speed-time line that the driver is
expected to follow) was generated in real time. When the program had finished
initializing, the instruction "Start the Engine" appeared on the screen accompanied by
three horn blasts. The driver then started the engine immediately and allowed it to idle.
After sixty seconds, data capture commenced. This corresponds to point "A" on the Test
Schematic. Ten seconds later, a ramp was drawn on the driver's display. This ramp was
generated to reach point "C" in around six seconds. The driver was expected to treat this
as a "free" acceleration (maximum acceleration, while driving in a normal style), and to
reach the steady state speed of 50 Km/hr (31.1 mph) as quickly as possible. Once the
vehicle reached 48 Km/hr (30 mph), point "D" is defined, at exactly sixty seconds in the
future. When this time had elapsed, the second rapid acceleration ramp was generated.
This ramp was drawn to reach point "E" in about 4.5 seconds. When the vehicle speed
reached 78.5 Km/hr (49 mph), point "F" was defined to be 80 seconds later than this
time. Upon reaching point "F," the data acquisition was stopped and the driver was free
to slow the vehicle speed in the safest manner. The 49 mph value was determined to be a
lower bound of approximation to reaching 50 mph. An actual test is shown in Figure 3.
25
-------
Table 11: AC5080 Methodology.
Path
A-B
B-C
C-D
D-E
E-F
F-G
Description
Idle. Select low gear.
Rapidly accelerate to 50 km/hr (31.1 mph)
Maintain 50 km/hr. (31.1 mph)
Rapidly accelerate to 80 km/hr (49.7 mph)
Maintain 80 km/hr. (49.7 mph)
Return vehicle to stop with engine at idle.
Duration
10 seconds
Vehicle Dependant
60 seconds
Vehicle Dependant
80 seconds
Vehicle Dependant
60
50
40
J3
a
a 30
o
13
>
20
10
A B
C
D
G
Figure 2: AC5080 Short Test (time not to scale).
26
-------
60
55
50
45
•5 40
1)
1)
35
30
•3 25
20
15
10
j
/
/
\
50
100 150
Time (s)
200
250
Figure 3: The AC5080 Schedule as driven by a 1990 Kenworth road tractor with a
370 hp Cummins Mil engine and a 10-speed manual transmission, with a simulated
vehicle test weight of 56,000 Ibs.
Test Weights
All of the HHDDT tested in this study were deemed to have gross vehicle weights of
80,000 pounds, in that they were all typical over-the-road tractors. The GVWR displayed
in the cab of such tractors may vary from 30,000 Ibs. (for single axle trailers) to over
52,000 Ibs. (for tandem axle trailers), but trailer weight is not considered in these GVWR
values. For tractor-trailers, it is the combination weight that is of interest, rather than the
tractor weight. The test weight used for all vehicles was planned to be 56,000 Ibs.,
representative of laden use. The test weights for Phase 1.5 were planned to be 30,000
Ibs., 56,000 Ibs. and 75,000 Ibs. However, it was evident that the 75,000 Ibs. operation
proved stressful to the truck brakes because only one or two axles were being used to
slow an inertia that would be slowed with five axles in normal on-road use. Also, the
accelerations and decelerations at 75,000 Ibs. showed potential to cause dynamometer
damage. Some preliminary test runs, presented in Figure 4 and Figure 5, showed that a
difference still existed in emissions between 56,000 Ibs. and 66,000 Ibs. operating weight,
and so the 66,000 Ibs. weight was used as a maximum weight test weight instead of the
75,000 Ibs. weight for the remainder of the program. PM increased monotonically with
test weight for the Transient Mode, but did not change appreciably between the 56,000,
66,000, and 75,000 Ibs. runs on the F£HDDT_S, where wind drag becomes more
important. For the case of NOX (see Figure 5), the Transient mode NOX emissions at
66,000 Ibs. were substantially higher than at 75,000 Ibs.
27
-------
30,000
56,000 66,000
Simulated Test Weight
75,000
Figure 4: PM emissions for vehicle E55CRC-27 which was tested at four weights,
before 66,000 Ibs. was adopted as the highest test weight.
35.0
30.0
25.0 -
20.0 -
15.0
10.0
30,000
56,000 66,000
Simulated Test Weight
75,000
Figure 5: NOX emissions for vehicle E55CRC-27 which was tested at four weights,
before 66,000 Ibs. was adopted as the highest test weight.
28
-------
The MHDT were tested at both laden and unladen weights. Laden weight was set to be
75% of the gross vehicle weight rating (GVWR). The unladen weight was set at 50% of
the GVWR.
Figure 6 shows an example of the monotonic increase in distance-specific emissions of
PM as test weight increases. Figure 7 shows that, with the exception of the AC5080
driving schedule, NOx emissions increased with test weight as well. These emissions
differences vindicated the effort associated with testing at two weights.
0.025
• 75% GVWR
D 50% GVWR
Figure 6: PM emissions for vehicle E55CRC-53 tested at two weights
29
-------
75%GVWR
D 50% GVWR
Figure 7: NOX emissions for vehicle E55CRC-53 tested at two weights.
30
-------
EQUIPMENT AND METHODS - NON-REGULATED SPECIES & PARTICLE
SIZING
DRI's Residence Time Dilution Tunnel
DRI provided a "residence time dilution sampling system" for sample collection during
Phase 1 and Phase 2 of this study. The dilution tunnel was identical to those used during
the Northern Front Range Air Quality Study and it was based on a dilution stack sampler
designed and tested by a group led by Dr. G.R. Cass6. Figure 8 shows a schematic of the
DRI sampler. Because of the short sampling time and generally low PM emission rates
expected for certain cycles (particularly the Idle mode), the sampler was used without
dilution as a residence time chamber only. The dilution sampler was interfaced with the
WVU dynamometer dilution system. The samples were drawn through cyclone
separators with a cut-off diameter of 2.5 mm, operating at 113 1pm and collected using
the DRI Sequential Filter Sampler (for inorganic species) and the DRI Sequential Fine
Particulate/Semi-Volatile Organic Compound (PSVOC) Samplers for organic species. In
addition, a separate sampling line (without cyclone) was used for collecting canisters,
DNPH cartridges and Tenax samples. The details of sample collection techniques are
described in the appropriate standard operating procedures (SOPs) and are available upon
request from the Organic Analytical Laboratory (OAL) of DRI.
31
-------
SAMPLE IHLET AND_
TEMPERATURE SENSOR
STACK-
S TVPE PITOT TUBE-
SAMPLE PORT (TVP)
0-F IN-HjO
MAGNEHLIC GAGE
ACTIVATED
CHARCOAL
FILTER
W^GLASS
WOOL
BACK-UP
HEATED YENTURI
SENSOR
. HEATED SAMPLE LIME
W HEMP SENSOR
.50" K 2M LONG SS
SOLPMMAK
LPM MIH
HEAVV DUTY
CLAMP W^TFE GASKET
BUTT-WELD FERRULE
TIG WELD CNLV
TURBULENT DILUTION TUNNEL
RESIDENCE TIME 2.4 SEC * 1200 LPM
DILUTION RATIO 2511 - SOU
Ot" 11 ¥ EFFECTIVE LENGTH
TO SAMPLER*
(113 LPM)
RESIDENCE
CHAMBER
40 SEC RES TIME
* 22t LPM
2.5uMCVLCiHES
-TO SAMPLER
(11JLPM)
Figure 8: Schematic of DRI dilution tunnel sampler (Hildemann, et al. )
32
-------
Mini Dilution System
The Scanning Mobility Particle Sizer (SMPS) employed a mini dilution tunnel to dilute
the raw exhaust for sampling. The DMS 500 did not employ this mini dilution tunnel, but
was used to sample directly from the main CVS tunnel.
The mini dilution system shown in Figure 10 incorporated a tunnel with its length
maintained at ten times its diameter to facilitate good mixing and uniform distribution at
the sampling zone. An orifice plate was provided at the inlet of the dilution system to
ensure proper turbulence. The tunnel was wrapped with heating tape and insulation to
maintain a temperature of 46° C (115° F) to prevent water condensation. A dilution ratio
of 1:30 was maintained throughout this study. A 12 cm (5 inch) exhaust coupling with a 2
cm (% inch) probe welded on its body was used to draw a partial sample from the raw
exhausts shown in Figure 9. One end of the probe had a quick disconnect fitting for ease
of coupling the probe to the mini dilution tunnel. The sampling was not isokinetic. The
sample line from the probe to the mini dilution tunnel was kept as short as possible and
well insulated. Dilution air was supplied immediately before the mixing orifice by a
separate pump. The dilution air was passed through a refrigerated dryer and a HEPA
filter to provide particle and moisture free air. A separate vacuum pump was used to draw
in the exhaust sample.
Raw Exhaust
T
Figure 9: Exhaust Coupler with Sampling Probe for mini dilution tunnel
33
-------
Mini Dilution Tunnel
70 mm Filter Holder
Diluted Exhaust Co2 Sample
Temperature Controllers
Refrigerated Dryer
(Dilution Air)
Heat Exchanger
(Dilution Air)
Mini Tunnel Vacuum
Pump (Dilution Air)
Dilution Air Pump
Heater Insulation
Dilution Air Manifold
*—Raw Exhaust
Raw Exhaust Co2 Sample
Dual Co2 Analyzers (Raw & Dilute)
Co2 Sample Bypass Valves
Thermo-Electric Chiller (Co2 Sample
Conditioning)
Data Acquisition System
Co2 Analyzer Calibration Ports
Mass Flow Controller
Power Supply
Co2 Bypass Pumps
Figure 10: West Virginia University-Mini Dilution System
34
-------
Both the dilution air and total air flow rates were controlled by two Sierra mass flow
controllers calibrated from 0 to 200 slpm (0 to 7 scfm). Sampling probes were installed
ten diameters downstream of the mixing orifice to ensure uniform concentration. A dual
range CO2 analyzer was used to measure the raw and dilute CO2 concentrations. A 10-
point analyzer calibration was employed using 15% and 1% CC>2 bottles. The analyzer
was checked everyday for a zero and span using nitrogen as zero gas and 15% and 1%
CO2 bottles as span gas for the analyzers. The raw and dilute exhaust gases were filtered
by inline filters and chilled by thermoelectric chillers for removal of particulates and
moisture, respectively. The filters were checked every day and replaced if necessary. A
purge air supply for the CC>2 analyzers was provided by an external oil free compressor
through a HEPA filter. This purge air was useful in flushing the analyzer of any residual
gas present, which may provide erroneous readings.
The mass flow controllers and the CO2 analyzers were connected to a data acquisition
system (National Instruments DAS 6020-E BNC). A program in Visual Basic was
developed by WVU to control the data acquisition system (DAS) pad and to save
continuous data from the analyzer and the mass flow controller. Other parameters such as
tunnel pressure, and temperature were also recorded on a continuous basis. The DAS pad
also maintained the user defined dilution ratio. The desired dilution ratios were achieved
in two ways, first a flow-based system and the other is a CO2 based system. With the
flow based system, the dilution ratio was calculated using the following equation:
DR = total / (total - dilute)
The dilution ratio was maintained by adjusting both the total and dilute flow rates. The
total and dilute flow rates were measured and set by the mass flow controllers. The CO2-
based system measured the raw and dilute CC>2 concentrations and the dilution ratio was
the simple ratio of raw to dilute concentrations. The dilution ratio was maintained by
controlling the mass flow controllers based on the CC>2 readings. Through experience it
had been determined by WVU that the flow based system was more accurate in
controlling the dilution ratio and the CC>2 readings were used to verify the set dilution
ratio. It may be argued that the dilution ratios may be better controlled using CC>2
concentrations than using mass flow controllers. WVU has both systems on the mini-
dilution tunnel that is used for PM sizing work. Problems associated with deskew times,
gas dispersion in the tunnel, and chillers complicate the CCVbased control method,
especially during transient tests. Over steady state tests, the CO2-based dilution ratio
control can be very accurate. One probe was installed next to the dilute CO2 probe for
particle size distribution measurements using the SMPS. Carbon impregnated electrically
conductive Tygon tubing was used to transfer the sample from the tunnel to the
instrument. This tubing was used to prevent any particle losses due to electrostatic
deposition on the walls of the sampling tube.
Exhaust pulsations from the engine were encountered during this study and an exhaust
pulsation damper was devised, as shown in Figure 12. The damper consisted of a 5 gallon
tank with one end cut off and sealed with a rubber diaphragm. The rubber diaphragm was
35
-------
helpful in reducing the exhaust pulsations. The top of the tank had a tee connection which
incorporated a straight probe as seen in Figure 11, to minimize particle losses into the
damper. This system provided a pulsation free stream of sample with minimal particle
loss. The damper was flushed with particle free air and evacuated after every test to
remove any remaining particles or even volatile compounds, which may have found their
way into the damper. This was done because the volatile compounds may adhere to the
walls of the damper and subsequently form particles and create sampling artifacts.
Sample
from Mini
Tunnel
Straight probe
Pulsation free sample to SMPS
Pulsation
pressure
acting on
rubber
Figure 11: Schematic Representation of the Damping System
Figure 12: Exhaust Pulsation Damper
36
-------
Particle Sizing with the SMPS
The SMPS consists of two basic components, the Electrostatic Classifier or the Mobility
Analyzer and a Condensation Particle Counter. The use of the classifier to produce
monodisperse aerosols is illustrated in Figure 13. The classifier or mobility analyzer
separates particles based upon their electrical mobility and the resulting particles are
transported to the particle counter to obtain number concentrations. In mobility analyzers
particles are first charged, and then the aerosol is classified in a high electric field
according to the electrical mobility of the particles. The particle size distribution is
obtained on the basis of the relationship between mobility and sizes. The ultimate size
parameter determined from electrical mobility measurements is the equivalent mobility
diameter.
Poly disperse
aerosol
Imp actor
^
Kr-85
Bipolar
Charger
Sheath Air
•*•
i '
II
Particles of same
mobility
diameter passing
through te slit
Monodisperse
aerosol
Figure 13: Differential Mobility Analyzer
The inlet of the classifier has an impactor which removes all particles above 1 jim since
large particles tend to have multiple charges and stripping these charges is very difficult.
The impactor has a nozzle and an impactor plate and different size orifices (0.071 and
0.0475 cm) are used depending on the flow rate. The polydisperse aerosol passes through
37
-------
the bipolar charger containing a radioactive source (Krypton), which exposes the aerosol
particles to high concentrations of bipolar ions. The particles and ions undergo frequent
collisions due to random thermal motion of the ions. After reaching a state of
equilibrium, the particles carry a bipolar charge distribution. The polydisperse aerosol
passes into the DMA between the sheath flow, which creates an air curtain, and the outer
cylinder. The central rod is charged with a high negative voltage and the positive
particles are attracted to the rod. Particles with high mobility precipitate at the top end of
the rod and particles with low mobility exit the DMA along with the bypass air. Only
particles with a narrow mobility range exit the DMA through the slit as monodisperse
aerosol, which pass on to the Condensation Particle Counter for concentration
measurements.
To obtain a full scan of the particle sizes, the voltage in the central rod of the classifier is
increased exponentially with respect to time which is termed the up-scan of the DMA and
the voltage is dropped to zero which is termed as the down-scan. The up-scan and down-
scan times are typically 120 and 15 seconds, respectively. These times can be increased
by the user to obtain a better resolution if the particle concentrations are steady.
The Condensation Particle Counter (CPC), also called the condensation nucleus counter
(CNC), is the most common instrument used to determine number concentrations of
diesel particles. Upon entering the CPC the aerosol stream is saturated with alcohol
(typically butanol) vapor. As the mixture is cooled in the condenser tube, the vapor
becomes supersaturated and condenses on particles. As a result, the particles grow to a
diameter of about 10 jim, allowing for optical detection. Particle size detection limit in
the CPC is related to the increasing saturation ratio which is required with decreasing
particle diameters. Modern CPCs have detection limits of around 3-10 nm.
CPCs can be operated in two modes: (1) the counting mode and (2) the opacity mode. In
the counting mode, pulses of scattered light from individual particles are counted. This
mode provides the most accurate measurements, but can be used only at low particle
concentrations. In the opacity mode, used for concentrations above 104/cm3, number
concentrations are determined from the total scattering intensity. This mode, generally
subject to a larger error, requires that all particles grow to the same diameter and that the
optical system be frequently calibrated. The CPC instrument is very sensitive to the
ambient temperature, which affects the degree of supersaturation, as well as to
positioning and vibrations. For these reasons, it is suitable primarily for laboratory
measurements and its use for in-field measurements is more challenging. The CPC is
used for particle detection in many aerosol size distribution measurement instruments.
Both the CPC and the Electrostatic Classifier make up the SMPS system. The CPC and
the classifier are interfaced with an analog BNC connector cable. A personal computer
with custom software provided by TSI was used to record data and control the classifier.
A TSI model 3936 SMPS system was used for this study (Figure 14). DMA models 3081
(Long DMA) and 3085 (Nano DMA) were used interchangeably to obtain a full particle
size distribution down to 4 nm and up to 800 nm. A Model 3025A CPC was used for
38
-------
particle counting since it could detect particles with diameters as low as 3 nm and had a
response time of one second.
The SMPS is basically designed for steady state operations, where the exhaust
characteristics are invariant with time; that is, the engine is at a constant load and RPM or
when the vehicle is operated at a constant speed. During these steady states,
concentrations of all particles are almost constant, but the concentrations vary whenever
the load or speed of the engine/vehicle varies. For transient measurements of the
engine/vehicle, the engine/vehicle was first operated on a steady state mode and a full
size scan was obtained. Particle sizes were chosen from the full scan and the number of
particles chosen was dependent upon the number of times the vehicle was operated
through the transient cycle. For transient tests, the classifier operated in manual mode,
which allowed the user to enter a particular particle size and track the concentration of
that particular size for the entire test.
Figure 14: TSI Model 3936 SMPS System
To assure accurate measurements, the SMPS was flow calibrated before the start of the
project. An electrospray aerosol generator was used to calibrate the SMPS. The generator
was capable of producing 15, 30 and 70 nm sucrose particles. Leak checks were
performed every day to ensure that the SMPS had no leaks. Two types of leak checks
were performed, a HEP A filter check and a zero voltage check. Since HEP A filters have
high filtration efficiencies, a HEPA filter was connected to the inlet of the SMPS and a
full size scan was performed. If any particles were detected above the noise level of the
instrument, which was typically IxlO3 particles/cm3, then a thorough leak check was
performed according to the instrument instruction manual. The second method required
the voltage on the collector rod to be set at zero and checked for any particles being
measured by the CPC. Since the voltage on the collector rod determines the particle size
that exits the classifier, a zero voltage on the rod should ensure no particles exit the
classifier. Both these methods were used everyday to make sure the SMPS had no leaks.
If any leak was detected in the system, the manufacturer's leak check procedure was used
to track and solve the problem. The inlet nozzle on the classifier was cleaned everyday
with alcohol to remove any particle deposits, which may block the flow. The impactor
plate was also cleaned with alcohol and a thin layer of vacuum grease was applied to
prevent particle bounce.
39
-------
Particles are classified in the DMA by increasing the voltage exponentially and the
particles leaving the classifier are increasing in size. The particles are then counted by the
CPC and the values for concentrations for each particle size are stored in the PC as raw
counts. The particle size obtained from the raw data assumes that each particle has a
single charge on it after passing through a bipolar charger, which is not actually true.
Multiple charges on a particle increase its mobility and can be incorrectly binned into the
smaller diameter range. TSI uses an inbuilt algorithm in its SMPS software (Aerosol
Instrument Manager-SMPS) to correct for multiple charges based on Fuchs-Gunn aerosol
charging theory and a truncated triangular transfer function based on Knutson and
Whitby7.
The software provided by TSI to record only CPC concentrations (Aerosol Instrument
Manager-CPC) does not provide any algorithm for the reduction of data. The single
particle charge correction, transfer function and efficiencies of the CPC and impactor
have to be applied to the raw data. Two programs were developed in MathCAD, one to
reduce the CPC data and the other to correct the SMPS data for the dilution ratio
(Appendix I). The final data for the SMPS were presented as particle size based
concentration variations with concentrations in normalized units of particles/cm3
(dN/dlogDp). A log normal distribution was found to fit the data from the SMPS very
well. The final data from the CPC were presented as time-based concentration variations
of a selected particle size with concentrations in normalized particles/ cm3 (dN/dlogDp).
Particle sizing with the DMS500
A second instrument used in Phase 2 to measure the size and number distributions of PM
was a differential mobility spectrometer, DMS500 (Figure 15). The DMS500 was
directly connected to the secondary dilution tunnel and was used to sample at a point
close to the standard 70mm filter of the laboratory. Only primary dilution in the full-scale
CVS tunnel was employed. Secondary dilution was not needed in the tests because the
filter face temperature remained below 52°C with just primary dilution. The working
principle of the DMS500 is that it draws sample flow from the dilution tunnel containing
aerosols/ particles. It is capable of measuring 41 particle size bins (3.16 nm - 1000 nm)
as a continuous number count at a maximum frequency of lOHz. The size bins start at a
diameter of 3.16 nm and increase in the geometric ratio of 1.15 for each next size bin up
to the end limit of 1000 nm. The impactor at the inlet removes particles greater than 1000
nm and then the particles are charged by a corona charger. The charging of the particles
is argued to depend only on size and not on material composition. The dilute exhaust then
flows as a uniform laminar sheath through an annular geometry between the inner and
outer electrodes, as shown in Figure 15, and the charged particles are carried in this dilute
exhaust flow. The particles are then deflected towards grounded electrometer rings by
their repulsion from a central high voltage rod. Their landing position on the electrometer
rings is a function of the particle's charge and momentum. The particles give up their
charge to the electrometer amplifiers through the rings and the resulting currents are
output as particle number and size distribution through the user interface. There are
40
-------
studies showing the good accuracy of the DMS500 when compared with the SMPS and
ELPI7.
Sample Aerosol Inlot
Aerosol Charger
Virtual Earth Electrode Rings
Sheath airflow
Charged Particle Trajectories
High voltage
electrode
Figure 15: Working principle of the DMS500 .
Micro-Orifice Uniform Deposit Impactor (MOUDI)
WVU used the Micro-Orifice Uniform Deposit Impactor (MOUDI) for gravimetric
analysis of size-segregated PMo.i emissions from 10 vehicles (E55CRC-2 - E55CRC-11)
in Phase 1. The MOUDI, MSP Model 110 is a cascade impactor that classifies particles
by their aerodynamic diameter in the range of 18 |j,m to 0.056 |j,m. Model 110 has ten
stages with nominal 50% efficiency curve aerodynamic diameters of 0.056, 0.100, 0.180,
0.320, 0.560, 1.0, 1.8, 3.2, 5.6, 10 and 18.0 urn. The MOUDI moderates the pressure
drop needed to size submicron aerosols by using nozzles of very small diameter (2000
nozzles of 52 |j,m in diameter in the final stage). It operates at a rate of 30 1pm. The flow
rate is monitored by measuring the pressure drop between the first and the fifth stage with
a differential pressure gage. The differential pressure was adjusted by a needle valve to a
pre-calibrated pressure drop corresponding to a flow rate of 30 1pm. The advantage that
the MOUDI has over other cascade type impactors is its ability to collect ultra fine
particles with a moderate pressure drop and a uniform deposit.
Samples from the MOUDI were collected on greased aluminum substrates and one 37
mm Gelman Sciences TX40 after filter (for PM size fraction less than 0.056 |j,m).
41
-------
Gelman Sciences 47 mm Analyslides were used for transport and storage of the
substrates. Several researchers have espoused the necessity of applying a thin layer of
grease to the substrate surface in order to minimize particle bounce (Baron and Willeke,
1993; Marple et al., 1991). According to the majority of researchers, particle bounce is
not an important issue in diesel particulate matter sampling. This claim was found to be
in error by the Principal Investigators in an NREL-funded study (Gautam et al., 2002).
Particle bounce is a major problem in diesel particulate matter sampling. Upon impact,
some of the particles may bounce off the substrates and get re-entrained in the sample
stream where they will pass to successive stages. This will distort the size distribution
toward the smaller diameter regions, with no definitive method to predict or correct for
this phenomenon. Substrates, used in this study were greased by Ms. Sue Stine, who had
formerly worked for the US Bureau of Mines with Dr. Bruce Cantrell, and later for MSP,
Inc. (manufacturer of the MOUDI).
Cyclones
Size-selective cyclone samplers (URG Model 2000-30 EA, 28.3 1pm, 5odae=10|j,m) were
used to collect samples of PMio of exhaust PM. The PMio cyclone is an inertial particle
separator that uses centrifugal force to remove heavier particles from a gas stream (>10
|im and >2.5 jim, respectively). The diluted exhaust sample enters the cyclone
tangentially, and is drawn through the cyclone body and the in-line filter by a vacuum
pump. Maintenance of this exact flow rate can be difficult if the conditions at which the
sample is being collected are not near standard. Cyclone use in a dilution tunnel requires
a feedback control system to adjust the flow to account for variations in temperature and
pressure. This signal is used to control a mass flow rate controller that will be
responsible for maintaining the required constant actual flow rate.
DRI used a Model Bendix-Unico 240 PM2.5 cyclone to draw dilute exhaust samples from
WVU's dilution tunnel into the DRI's residence time dilution chamber. The cyclones
were an integral part of the residence time chamber. The cyclone in the DRI residence
time chamber operated at a volumetric flow rate of 113 liters/minute.
Unregulated Emissions
WVU and DRI (subcontractor to WVU) were jointly responsible for unregulated
emissions measurement, and WVU also provided support for Dr. Kim Prather's Aerosol
Time-of-Flight Mass Spectrometer (ATOFMS) data collection. ATOFMS data are not
discussed in this final report. Listed below are the specific tasks/activities conducted by
WVU and DRI.
42
-------
In addition to the gaseous sampling probes and a secondary dilution tunnel, WVU
installed the following on the primary dilution tunnel:
1. PMio cyclone with TX40 filters for gravimetric analysis (PM25 sampling was
conducted by DRI using the DRI residence time chamber).
2. Micro-Orifice Uniform Deposit Impactor (MOUDI) probe for gravimetric
analysis of size-selective PMo.i samples. Greased aluminum substrates were
employed for the purpose (for E55CRC-2 to E55CRC-11).
3. Probe for DRI's residence time dilution chamber.
4. Probe for Dr. Kim Prather's ATOFMS sampling in Phase 1.
Samples were collected by WVU using DRI's residence time dilution chamber (with
dilution) for the first two runs (1773-2 and 1773-3) of E55CRC-01. These runs are not
considered for the study since the filters and other media did not collect sufficient
quantity of samples. For all subsequent runs in Phase 1, and for the rest of the study (6
vehicles in Phase 2), the DRI's residence time dilution chamber was used without any
secondary dilution to increase the captured mass.
Tunnel blanks and tunnel backgrounds were also collected, as listed in the Test Sequence,
for the three overlap vehicles.
1. Volatile organic compounds (VOC) - Canisters; field GC
2. Methane - Canisters
3. Semi-volatile organic compounds (SVOC) - PUF/XAD and TIGF filters
4. SVOC, low molecular weight compounds - Tenax tubes
5. Nitro-PAHs - PUF/XAD and TIGF filters
6. Carbonyls - DNPH cartridges
7. Nitrosamines - Thermosorb cartridges
8. PM soluble organic fraction (SOF): organic compounds - TIGF filters
9. Elemental analysis and gravimetric analysis (PM2.s) - TIGF filters
10. Ammonium and Ions (Nitrate, Nitrite, Chloride, Sulfate) - Quartz filters
11. Elemental Carbon/Organic Carbon (EC/OC) - Quartz filters
12. PM2.5 fraction for gravimetric analysis
WVU collected the PMio fraction using a cyclone for all 25 vehicles in Phase 1, and 6
'speciation vehicles' in Phase 2. The PM0.i fraction was sampled in Phase 1 using the
MOUDI for E55CRC-2 to E55CRC-11 only.
Chemical speciation samples were collected by DRI, using the residence time dilution
chamber. DRI used a Model Bendix-Unico 240 PM2.5 cyclone to draw dilute exhaust
samples from WVU's dilution tunnel into the DRI's residence time dilution chamber.
The cyclones were an integral part of the residence time chamber. The cyclone in the DRI
residence time chamber operated at a volumetric flow rate of 113 1pm (4 cfm). However,
there was no additional dilution performed within the DRI chamber. The list of chemical
compounds analyzed and the media used to collect them are given below:
43
-------
Details of the sample collection and the analytical procedures followed by DRI are
provided in Appendix G.
Speciation Fleet and Approach
The sampling of non-regulated species occurred at the same time that regulated species
were sampled. Five HHDDT with a Gross Vehicle Weight Rating (GVWR) of 80,000
Ibs. were selected specifically for non-regulated sampling. These were E55CRC-39,
E55CRC-40, E55CRC-42, E55CRC-43 and E55CRC-44. A sixth vehicle, E55CRC-41
was a MHDT and hence cannot be used for comparison purposes directly with the
HHDDT in this study. The details of the test vehicles, engines, model years and the test
fuel used is given below in Table 12.
Table 12: Test Vehicle Details
Vehicle ID
E55CRC- 39
E55CRC- 40
E55CRC-41
E55CRC- 42
E55CRC- 43
E55CRC- 44
Vehicle
Manufacturer
Volvo
Freightliner
Ford
Freightliner
Peterbilt
Volvo
Vehicle
Model
Year
2004
2004
1998
2000
1995
1989
Engine
Manufacturer
Cummins
Detroit Diesel
Cummins
Caterpillar
Detroit Diesel
Caterpillar
Engine
Model
ISX 530
Series 60
B5.9
3406
Series 60
3406
Engine
Model
Year
2004
2003
1997
1999
1994
1989
Primary
Fuel
CARB
CARB
CARB
CARB
CARB
CARB
No aftertreatment devices were used on any vehicle. The vehicles were tested in an "as
received tank fuel" condition in the same way as for regulated emissions characterization.
Fuel samples for every vehicle were collected at the end of the test period and were sent
for chemical analysis to Core Laboratories for analysis of cetane number, total sulfur
content, aromatic content and viscosity. Another sample of the fuel was sent to DRI for
analysis of inorganic and organic compounds. Oil samples were collected for every
vehicle and were sent to oil sciences laboratory for 27-point analysis and sulfur content.
Another sample of the oil was sent to DRI for organic and inorganic analysis.
This non-regulated emissions study employed the same test schedules that were used for
the regulated emissions measurement.
The Idle mode was extended to run three times longer than the normal Idle mode to
facilitate sufficient sample collection on the speciation media. No speciation or particle
sizing data were recorded for the Creep mode. The Transient was run normally for 688
seconds. The Cruise mode was run normally for 2083 seconds. Chemical speciation and
particle sizing data were recorded for the Idle, Transient, Cruise and High-speed Cruise
Modes.
Each mode was repeated twice for every vehicle. Similar multiple modes were
composited by DRI on the same media to ensure sufficient sample collection. The SMPS
operated in a full-scan mode during the HHDDT schedule. For the Transient Mode, the
SMPS was locked onto a particle size of interest and concentration variation for that
44
-------
mode was measured. Particle sizes were selected based on the steady state size
distribution. A steady state run at 40 kph (25 mph) mode, not required for regulated
emissions testing, was executed prior to any other tests for the sole purpose of obtaining a
full particle size distribution by the SMPS. The steady state speed of 40 kph (25 mph)
was selected based on the average speed of the transient cycle. The DMS500 was
operated during the modes because it was not restricted to steady-state measurements.
45
-------
QUALITY ASSURANCE AND QUALITY CONTROL
The WVU Transportable Laboratory (TransLab) evaluated the emissions from vehicles
operating in Southern California fueled with diesel or gasoline between August 2001 and
May 2005. Testing was conducted in Riverside, CA. The independent Quality Assurance
Quality Control (QA/QC) review of the regulated emissions is summarized below.
The procedure used in reviewing the data involved four levels of checks. First, the field
engineer would make an assessment of the quality of the data as it was generated. If a
test was deemed to be invalid, the field engineer would mark that test as being invalid,
and provide the reason(s) why the test was invalid, and a new test would be run in its
place. Additionally, if a test was performed with an atypical result, but there were no
known problems, the field engineer would annotate the dataset in which this occurred and
would list potential reasons for the result. For example, if the vehicle was underpowered
and could not meet the desired target distance then this comment would be added to the
test record. At the end of the test day, the field engineer electronically transferred the
data to the project engineer at WVU in Morgantown, WV.
The second level of assessment of the data quality was performed at WVU by the project
engineer assigned to maintain the data. The project engineer received the data collected
the previous day and reviewed the dataset to insure that the correct vehicle/engine model,
test weights, and test cycles were performed. In instances when there were problems or
when atypical results were obtained, the field engineer would interact with the project
engineer during testing via email and/or phone conservations to make an assessment of
the actions, if any, to take. Additionally, the project engineer would review the integrated
mass emissions data and the pertinent analog and digital signals required to calculate the
emissions for potential problems. The project engineer maintained a database of the
results in tabular and graphical format to be able to identify tampered and or mal-
maintained (T/M) engines and or vehicles based on criterion developed during this
project. If a potential T/M vehicle was identified, this information was passed on to the
Principal Investigators, and this information was then communicated to the sponsors. If a
problem was identified with the data by the project engineer, then the project engineer
would interact with the field engineer to rectify the problem.
The third level of assessment was done by the Principal Investigators. The Principal
Investigators examined the data and provided feedback to the project engineer of
potential problems, who was separated from the data gathering process.
The fourth level of data quality assessment was performed by the QA/QC officer. The
QA/QC officer received the integrated mass emissions data in the form of single page
short reports from the project engineer in hardcopy or electronic (Microsoft Word)
format. Short reports are presented in Appendix K. The QA/QC officer would, at times,
examine raw and processed data with the project engineer as a spot check throughout this
test program. The test, vehicle, and emissions data summarized in each short report were
reviewed on a per test basis and per vehicle basis for inconsistencies in the information
46
-------
and emissions data. Inconsistencies in the information were brought to the project
engineer's attention and corrected. These corrections included misspellings, incorrect
vehicle description, incorrect engine description, lack of comments, and other minor
corrections. These corrections are not summarized in the table below. Questions,
concerns, and inconsistencies in the emissions data that were directed to the project
engineer are summarized in the table below. The corrective action taken is also listed.
The majority of the inconsistencies in the emissions data can be attributed to trying to
measure very low emissions levels (such as low engine power Idle and Creep modes) and
higher level emissions (high engine power transient and cruise modes) with one set of
emissions analyzers and one calibration range. There was a two order of magnitude
difference in the emissions concentration levels between the low and high engine power
modes, so that low power modes to have a much greater variability in this data.
There were difficulties in comparing the results from two different NOX analyzers during
the first half of the test program. The difference in the two NOX analyzers was addressed
throughout this test program by improving the test procedure to adjust the second NOX
analyzer flow and the changing of NOX analyzers to newer technology.
47
-------
RESULTS AND DISCUSSION, REGULATED SPECIES
HHDTData Gathered
Test Runs
Each execution of a cycle or mode was assigned a sequence number and a run number.
Appendix J presents a listing of sequence and run numbers, with the corresponding name
of the cycle, vehicle number and test weight. Some sequence and run numbers are
omitted in Table 3 because they are associated with background tests or with rejected
runs. In Table 3 the test modes or cycles are designated by the actual file names used.
"Test D" corresponds to the UDDS, trans3 and cruise3 refer to the three point smoothed
versions of the Transient and Cruise Modes that are customarily used. Some modes were
lengthened by being repeated to collect sufficient PM mass during the mode. Idle32
refers to a double length idle (1,800 seconds instead of 900 seconds), and creep34 refers
to four repeats of a creep run as a single mode. Table 3 corresponds to the sequence and
run numbers in the short reports appearing in Appendix K. MHDT are shown with 50%
and 75% test weights. The actual weights used appear in Appendix J.
Data Gathered
The cycle or mode-averaged data, in units of g/mile (except for idle, which is in time-
specific units) are all presented in Appendix K in the form of "short reports." These data
have also been translated into graphical representations in Appendix H. Data have been
gathered into summary tables in Appendix L. The full database containing continuous
vehicle operating data and gaseous emissions rates has been made available to the
sponsors separately in electronic form.
Discussion of individual truck emissions within the body of this report emphasizes two
species, NOX and PM, since NOX and PM data are assumed to be of greatest interest for
truck operation. However, CO and HC have also been presented with respect to select
cycles below and for all cycles in the appendices. The ARB cycles (Idle, Creep, Trans,
Cruise, and HHDDT_S) demonstrate a variety of driving conditions. Additional, more
comprehensive plots of the data appear in Appendix M. Some vehicles were "speed
limited" to the point where they were unable to follow the schedule trace and were not
tested for the HHDDT_S. HHDDT in Phase 1 were not exercised through the
HHDDT_S.
Figure 16 presents the NOX emissions for all of the HHDDT on the Idle mode of the
HHDDT schedule. The range of years appearing in Figure 16 corresponds to the CRC
grouping of the years of engine certification. The number below indicates the number of
vehicles tested in that MY range for which data are included in the chart. The 1998
vehicles show the highest NOX emissions of all MY bins. Later MY trucks produce
higher NOX than early MY trucks because the injection timing may be advanced on later
model trucks to reduce white smoking. The reader is cautioned that small number counts
of vehicles in some MY bins may cause the program data in some represent the whole
48
-------
California fleet faithful. This is especially true if one of the vehicles in the group is an
outlier, or high emitter.
o.o
Figure 16: NOX emissions for the Idle Mode, in units of g/minute.
9)
"
Figure 17: NOX emissions for the Creep mode (56,000 Ibs.).
49
-------
Figure 17 shows NOX emissions for all HHDDT trucks operated through the Creep mode
at 56,000 Ibs. test weight. This mode is primarily low speed with repeated upward and
downward speed ramps. The small distance traveled and high idle content lead to high
distance-specific emissions values. Results are less variable over the MY bins than for
idle, but idle and low load injection timing advance is also the most likely cause for the
high NOX emissions over the 1987 to 2002 period.
Figure 18 shows NOX emissions for all FHDDT trucks operated through the Transient
mode at 56,000 Ibs. test weight. The Transient mode represents typical urban driving
activity. Figure 18 shows that the NOX emissions for 1998 were higher than for
neighboring MY bins, but the highest NOX arose from the oldest vehicles. The 1975 to
1976 MY bin was the highest at 47 g/mile. This was primarily due to their being only
two sample vehicles, one of which emitted 60 g/mile while the other was 34 g/mile. The
1980 to 1983 MY bin contained vehicles consistently emitting high NOX. The emissions
of NOX prior to 2002 were all higher than 20 g/mile. The NOX emissions of the five 2003
and newer MY vehicles (equipped with exhaust gas recirculation for Cummins and
Detroit Diesel, or bridging/ACERT technology for Caterpillar) differ in that they
averaged less than 17 g/mile.
Figure 18: NOX emissions for the Transient mode (56,000 Ibs.). One of the vehicles in
the 1975-1976 MY bin emitted NOX at 60 g/mile.
50
-------
Figure 19 shows NOX emissions for all HHDDT trucks operated through the Cruise mode
at 56,000 Ibs. test weight. Figure 19 shows that the NOX emissions for 1998 were higher
than all but the 1975-1976 group of trucks. The 1998 trucks were also relatively higher
than they were for the Transient mode. This suggests that off-cycle injection timing was
enacted more on the Cruise mode than the Transient mode, which would be expected.
This issue is discussed in more detail in the section on T&M and reflash. The NOX
emissions of the five 2003 and newer MY engines (equipped with exhaust gas
recirculation or bridging/ACERT technology) differ in that they averaged to about 9
g/mile.
Prior studies with fewer HHDDT1'10'11'12 and earlier reports in the E-55/59 program have
suggested that NOX did not decline with MY. Figure 19 shows a pattern of high NOX on
the oldest trucks, as well as trucks in the 90's that have off-cycle injection timing
advance, with a decline in NOX for the latest MY trucks. However, there is no clear trend
in NOX with respect to MY from 1977 to 1997 for trucks exercised through the Cruise
mode at 56,000 Ibs. test weight.
40.0
35.0
Figure 19: NOX emissions for the Cruise mode (56,000 Ibs.). One 1998 vehicle
(E55CRC-10) was identified as a high NOX emitter in Phase 1.
Figure 20 shows NOX emissions for the HHDDT_S mode at 56,000 Ibs. test weight. This
sample of trucks excludes the Phase 1 HHDDT because they were not tested on the
51
-------
HHDDT_S. This mode has a high content of near-steady high-speed freeway operation.
A 2003 MY truck exhibited lowest NOX, at 8.05 g/mile. The 1969, 1975 and 1998 engine
years exhibited the highest NOX at approximately 29, 56, and 22 g/mile respectively. A
1975 Cummins-powered truck had the highest with 56 g/mile, as the lone truck in the
1975-1976 category that completed the HHDDT_S mode.
60.0
50.0
Figure 20: NOX emissions for HHDDT_S mode (56,000 Ibs.). The high value for the
1975-1976 MY bin is due to a single truck with high NOX emissions.
Figure 21 shows the NOX emissions for the UDDS. The distribution of emission levels of
this urban driving cycle is similar to that found in the Transient mode of Figure 18 above.
Figure 22 shows the PM emissions for the Idle mode. E55CRC-37 was tested in this
mode only since it could not operate on the dynamometer due to traction control
technology. For this reason there is an additional 2003+ data point on the Idle graph
(Figure 22) relative to the active modes. Some MY bins were high due to outlying
performance by individual vehicles. For example, E55CRC-45, discussed in the T&M
section, was a prodigious emitter of both PM and HC and biased the data for the 1991-
1993 MY group. PM was lowest in the 2003 and later MY group.
52
-------
Figure 21: NOX emissions for UDDS mode (56,000 Ibs.).
Figure 22: PM emissions for Idle Mode. Truck E55CRC-45, in the 1991-1993 MY
bin, was a prodigious emitter and raised the average value for that bin substantially.
53
-------
Figure 23 shows PM emissions for HHDDT for the Creep mode. The two highest bins
for the Creep mode are the same as for the Idle mode. The 2003+ group has a
significantly lower output than even the second lowest group (1998).
Figure 23: PM emissions for Creep mode (56,000 Ibs.). One vehicle in the 1991-1993
MY bin was a very high emitter of both PM and HC at light load.
Figure 24 shows PM emissions for all HHDDT trucks operated through the Transient
mode at 56,000 Ibs. test weight. E55CRC-45, a 1993 Cummins-powered tractor
demonstrated significantly higher PM emissions than the other test vehicles at low load,
but E55CRC-45 biased the data far less in the Transient mode. In the highest emitting
group, E55CRC-16 emitted 15 grams per mile of PM and the other vehicle in the 1977-
1979 group emitted 2 grams per mile. E55CRC-16 is discussed in detail in the T & M
section. There is a general downward trend in PM emissions on the Transient mode with
respect to MY.
Figure 25 shows the PM emissions for the Cruise mode. Unlike in most other modes, the
2003+ group does not have the lowest emissions. E55CRC-16, a 1979 Caterpillar-
powered HHDDT, had the highest PM emissions with 16 g/mile, and led to the high
value for the 1977 to 1979 MY group. Even without this MY group, it is evident that PM
has declined over three decades for cruising operation. PM emissions after the
introduction of electronic engines in the 1991-1993 period are lower than the emissions
from earlier mechanically controlled engines. This can be ascribed to improved fuel and
boost management, and higher injection pressures for later MY regulations.
54
-------
Figure 24: PM Emissions for the Transient mode (56,000 Ibs.). Truck E55CRC-16, a
high emitter, caused the high average in the 1977-1979 MY bin
Figure 25: PM emissions for the Cruise mode (56,000 Ibs.). E55CRC-16 contributed
to the one very high MY bin average.
55
-------
Figure 26 shows PM emissions for all HHDDT trucks operated through the HHDDT_S
mode at 56,000 Ibs. test weight. E55CRC-27 was the first vehicle tested using the
HHDDT_S Mode. E55CRC-69 was unable to attain sufficient speed to provide
meaningful data in the HHDDT_S mode. It was governed to 55 mph maximum speed.
The first three groups in Figure 26 had only one vehicle in each group, each of which had
significantly higher emissions than the newer vehicles. There appears to have been a step
change in PM emissions beginning with the 1991-1993 group. E55CRC-29, a 1999
Cummins, escalated the average for the 1999-2002 group by emitting 2.52 g/mile of PM,
a value more than twice as high as for any other vehicle in the group.
Figure 26: PM emissions for HHDDT S mode (56,000 Ibs.). Four MY groups show
data from only one vehicle.
Figure 27 shows the PM emissions for the UDDS. As was the case with the NOX
emissions, the UDDS and Transient modes show a similar distribution pattern.
Figure 28 shows the CO emissions for the TransS Mode. While there were a handful of
vehicles with emissions higher than other trucks in the MY bin, the highest was
E55CRC-23 (1983 Cummins) with 40 g/mile. Also, the 2003 Detroit Diesel engine of
E55CRC-44 was significantly (two to four times) higher in both CO and PM than the
other four vehicles in the 2003+ group.
56
-------
Figure 27: PM emissions for UDDS mode (56,000 Ibs.).
30.0
25.0
Figure 28: CO emissions for Transient mode (56,000 Ibs.).
57
-------
Figure 29 shows the CO emissions for the Cruise mode. In sympathy with the data for
the Transient mode, E55CRC-23 with its 1983 Cummins was nearly the highest emitter.
E55CRC-20 with a 1992 Detroit Diesel engine also had about 14 g/mile of CO emissions.
If it were not for the 2003 Detroit Diesel engine of E55CRC-34, the 2003+ group would
have represented an even more dramatic drop from previous model years.
Figure 29: CO emissions for Cruise mode (56,000 Ibs.).
HC emissions from vehicles built over the last decade are lower than from older vehicles
in the older vehicles in the test fleet. Figure 30 shows the HC emissions for the Transient
mode. E55CRC-45 had a 1993 Cummins engine. It emitted the highest level of HC at
5.43 g/mile. E55CRC-45 was a high emitter of both PM and HC at light loads, and has
been discussed in the T & M section. The 1994 Cummins of E55CRC-60 emitted 4.18
g/mile. Both of these trucks showed far higher emissions than were measured from the
other vehicles in each of their respective MY bins.
Figure 31 shows the HC emissions for the Cruise mode. E55CRC-45 (1993 Cummins)
again ranked as one of the top HC emitting vehicles, and biased the 1991-1993 MY bin.
The five most recent MY bins show lower emissions, on average than earlier MY bins,
but there is no clear, monotonic, pattern to HC reduction with respect to MY.
Additional plots showing the HHDDT emissions at test weights of 30,000 and 66,000 Ibs.
are available in Appendix M.
58
-------
Figure 30: HC emissions for Transient mode (56,000 Ibs.).
Figure 31: HC emissions for Cruise mode (56,000 Ibs.).
59
-------
MHDT Data Discussion
This section discusses results of the testing of the MHDT. The MHDT Schedule includes
a lower speed transient mode (MHDTLO), a higher speed transient mode (MHDTHI), a
cruise mode (MHDTCR), and the HHDDT_S. Figure 32 shows emissions from the
AC5080, which was also used. Figure 33 presents NOX emissions for the Lower Speed
Transient mode from the MHDT. In these, and all other figures in this section, 'laden'
refers to 75% of the gross vehicle weight rating (GVWR) and 'unladen' refers to 50% of
the GVWR. NOX emissions from these MHDT are similar to those of the HHDDT
transient mode.
E55CRC-57, in the 1999-2002 group, showed an unusually high unladen NOX emissions
value. This was ascribed to a failed temperature sensor, which caused a limp-home
mode. E55CRC-57 has been discussed in detail in the T & M section.
The three early model gasoline engines yielded NOX emissions on the Lower Speed
Transient mode that were similar in emissions from the diesel vehicles, but the one newer
gasoline vehicle, E55CRC-53, produced the lowest NOX relative to other categories of
truck.
Figure 32: NOX emissions for AC5080 (both laden and unladen). Gasoline-fueled
vehicles are separated from diesel-fueled vehicles for comparison.
60
-------
—
1
35.0
30.0
Figure 33: NOX emissions for the Lower Speed Transient mode (both laden and
unladen).
Figure 34 shows that the NOX emissions for the Higher Speed Transient mode were
consistently lower than for the Lower Speed Transient Mode. As in the previous mode,
E55CRC-53, with a gasoline engine, produced the lowest NOX emissions at fewer than 4
g/mile. The older gasoline vehicle, E55CRC-54, emitted NOX at a level characteristic of
the diesel trucks. From Figure 32 to Figure 47 it is important to realize that varying test
weights from truck to truck influences distance-specific emission levels.
As seen in Figure 35, NOX emissions during the MHDT Cruise mode were slightly lower
than for either of the Transient modes and were close to those of the high-speed cruise
(HHDDT_S) shown in Figure 36. The older gasoline fueled trucks on average emitted
NOX at a higher rate than the diesel fleet. Note that Figure 36 is for a small subset of
vehicles. Figure 37 shows NOX on the HDDS for MHDT.
61
-------
25.0
20.0
15.0
10.0
Figure 34: NOX emissions for the Higher Speed Transient mode (both laden and
unladen).
30.0
Figure 35: NOX emissions for the MHDT Cruise mode (both laden and unladen).
62
-------
—
'a
35.0
30.0
25.0
20.0
15.0
10.0
Figure 36: NOX emissions for the HHDDT_S mode (both laden and unladen).
"Sfc
Figure 37: NOX emissions for the UDDS mode (both laden and unladen).
63
-------
Figure 38, Figure 39, Figure 40, Figure 41, and Figure 42 show PM emissions for the
AC5080, MHDT Lower and Higher Speed transients, the MHDT Cruise mode, and the
HHDDT_S mode. As expected, PM values from the gasoline trucks were consistently
well below those of the diesel vehicles. The influence of test weight (laden vs. unladen)
on PM was noticeably greater than for NOX. For the limited number of trucks tested, one
may conclude that PM levels have not changed from MY 1991 to MY 2002.
Figure 38: PM emissions for the AC5080 mode (both laden and unladen).
For the HHDDT_S (Figure 42), only selected trucks were used from the MHDT fleet,
leading to relatively higher MY 1991-1993 group emissions. E55CRC-76 contributed to
the high emissions in the 1991-1993 MY group.
Figure 43 shows emissions of PM for the UDDS. The four gasoline vehicles emitted, on
average, an order of magnitude lower PM than the diesel fleet.
Figure 44 and Figure 45 present CO emissions data for the MHDT on the Higher Speed
Transient and MHDT Cruise modes. In contrast to their PM contribution, CO from the
gasoline vehicles was substantially higher than from the diesel vehicles.
64
-------
—
'a
Figure 39: PM emissions for the Lower Speed Transient mode (both laden and
unladen).
Figure 40: PM emissions for the Higher Speed Transient mode (both laden and
unladen).
65
-------
Figure 41: PM emissions for the Cruise mode (both laden and unladen).
—
a
0.4
0.2
0.0
Figure 42: PM emissions for the HHDDT_S mode (both laden and unladen).
66
-------
Figure 43: PM emissions for UDDS mode (both laden and unladen).
Figure 44: CO emissions for MHDTHI mode (both laden and unladen).
67
-------
Figure 45: CO emissions for the MHDTCR mode (both laden and unladen).
"Sfc
14.0
12.0
10.0
Figure 46: HC emissions for MHDTHI mode (both laden and unladen).
68
-------
Figure 47: HC emissions for MHDTCR mode (both laden and unladen).
Figure 46 and Figure 47 present HC emissions for the MHDT on the Higher Speed
Transient and the MHDT Cruise modes. The oldest MHDDT produced substantially
higher emissions of HC than the remaining MHDDT on the MHDT Cruise mode. The
newest MHDGT produced no more HC on cruise than the diesel truck of similar MY.
Additional plots for MHDT emissions on other modes and cycles are presented in
Appendix M.
Continuous Data
All dynamometer speeds and torques, all regulated gaseous emissions, exhaust and tunnel
temperatures, and TEOM data are available on a continuous basis. These data are
available to the sponsors separately from this report. Figure 48, Figure 49, and Figure 50
present examples of continuous NOX, HC and CO emissions from E55CRC-42 on the
Transient mode at 56,000 Ibs. test weight and Figure 51 shows the exhaust temperature of
this vehicle.
69
-------
"si 0.2
100
200
300 400
Time (s)
500
600
700
Figure 48: Example of a continuous NOX emissions plot in g/s for E55CRC-42
(56,000 Ibs.) following the HHDDT Transient Mode.
0.035
0.030
0.025
0.020
"Si
0.015
100
200
300 400
Time (s)
500
600
700
Figure 49: Example of a continuous HC emissions plot in g/s for E55CRC-42
(56,000 Ibs.) following the HHDDT Transient Mode.
70
-------
0.06
0.05
0.04
0.03
0.02
0.01
0.00
100 200 300 400
Time (s)
500
600
700
Figure 50: Example of a continuous CO emissions plot in g/s for E55CRC-42 (56,000
Ibs.) following the HHDDT Transient Mode.
a.
e
100
200
300 400
Time (s)
500
600
700
Figure 51: Example of a continuous exhaust temperature reading for E55CRC-42
(56,000 Ibs.) following the Transient mode schedule.
71
-------
HHDDT Effect of Test Weight
HHDDT data have been provided above only for the 56,000 Ibs. test weight, which was
used for all trucks. Data for fleet subsets tested at 30,000 Ibs. and 66,000 Ibs. appear in
Appendix N. Figure 52, Figure 53, Figure 54 and Figure 55 show the effect of test
weight for Transient and Cruise modes, for PM and NOX, for all trucks that were tested at
all three weights. NOX rose in all cases as the weight rose from 30,000 to 56,000 Ibs., but
in a few cases, NOX was lower when the weight was increased to 66,000 Ibs. This may
be ascribed to variability in injection timing over the engine map, and to different gear
selection at the higher weight.
Figure 52: Transient mode NOX weight effects
72
-------
"Sfc
o.w
7 f»
£> C\
^ n
A n
^ n
0 n
i r^
l.U
n n
f
f
J~l
1 1
H
^
^
XXXX
'XXXX
1
'XXXX
/xxx>
'XXXX
XXXX*
'XXXX
I
R
^
n
[
f
1
'XXXX
/xxx>
'XXXX
XXXX*
'XXXX
i
•
0
• 30,000 Ibs.
D 56,000 Ibs.
H 66.000 Ibs.
i tfl Hg
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
s
& .^r
Figure 53: Transient mode PM weight effect
1
i
rxxxxxxxxxxxxxxxxxxxy
J— 1
I
xxxxxxxxxxxxxxxxxxxx>
\
\
1-1
a
^xxxxxxxxxxxxxxxxxxxxxxxxx
\\
R
!
'XXXXXXXXXXXXXXX
• 30,000 Ibs.
• 56,000 Ibs.
E3 66,000 Ibs.
r-B
I^XXXXXX,
^xxxxxxxxxxx^
^xxxxxxxxxxxxxxy
Jl
y
-,C'
.O'
^
4?
Figure 54: Cruise mode NOX weight effect
73
-------
For PM, the weight effect was far less clear. No simple conclusions can be presented
except that PM rose as test weight was raised from 30,000 to 56,000 Ibs. for almost all
cases for the Transient mode.
-f
-s*
Figure 55: Cruise mode PM weight effect
MHDT Effect of Test Weight
Figure 34 and Figure 35 in the section above provide a partial view of the effect of test
weight on NOX emissions. While there were substantial truck-to-truck variations, the
average effect of weight across the MHDT fleet was a rise in NOX of 9% for the Higher
Speed Transient mode and 8% for the cruise mode as the weight increased from 50
percent to 75 percent of the GVWR. The research effort to establish the MHDT schedule3
found, in agreement with the present study, that effect of weight on NOX was smaller than
previously established for HHDDT. The average effect on PM seen in Figure 40 and
Figure 41 indicated an increase of 50% for the Higher Speed Transient mode and 47% for
the MHDT Cruise mode as weight was increased from unladen to laden.
74
-------
COMPARISON OF DATA BETWEEN PHASES
Data for the same MY bins were compared between phases to assess phase to phase
repeatability. Figure 56 shows data for CO2 emissions, in units of g/mile, for all HHDDT
operated through the HDDS at 56,000 Ibs. test weight.
3500
3000
2500
2000
1500
1000
500
• Phase 1
D Phase 1.5
EJ Phase 2
• PhaseS
g
V
V
V
V
V
V
V
*L
*;
V
V
1
V
V
V
V
V
V
V
V
V
V
\
1-1
^
V
V
V
V
V
V
V
V
V
V
V
V
PI
s
V
V
V
V
V
V
V
V
V
V
V
V
V
\
-
Pre- 1975- 1977- 1980- 1984- 1987- 1991- 1994- 1998 1999- 2003+
1975 1976 1979 1983 1986 1990 1993 1997 2002
Figure 56: Variation of HHDDT COi emissions by vehicle model year and by phase
of the program for the UDDS (56,000 Ibs.).
There is little variation in the CC>2 emissions values over MY or by phase, because CO2
values are governed by the engine efficiency and cycle energy demands. (The standard
deviation of the data below is 302 g/mile, approximately 13 percent of the average.)
These data build confidence in the repeatability of data between phases, and suggest that
both the dilution tunnel flowrate calibration and the CC>2 analyzer response remained
consistent between phases. In contrast, Figure 57 shows that CO varied substantially
within MY group. (The standard deviation of the data in Figure 57 is 8.7 g/mile, or 84%
of the average.) Since CO is known to be a species that varies widely from truck to truck,
this suggests that too few vehicles were selected in each phase to form a reliable opinion
on average CO emissions for each MY bin. NOX emissions are less variable over the fleet
than CO emissions, as shown in Figure 58. However, Figure 58 still shows variation
between phases. The NOX data were guaranteed to have reasonable accuracy by
employing two separate NOX analyzers for most of the program. PM data also varied
between phases, as shown in Figure 59. (The standard deviation of the data in Figure 59
75
-------
is 9.7 g/mile, or 41% of the average.) A correspondence of Figure 57 and Figure 59 show
a correspondence of high CO and PM by phase in sympathy with one another. Examples
are the Phase 1.5 data in the 1999-2002 and 2003+ MY bins. One must conclude that the
fleets in each individual phase were also too small for use in reaching conclusions on CO,
NOX or PM averages.
"Sfc
Pre- 1975- 1977- 1980- 1984- 1987- 1991- 1994- 1998 1999- 2003+
1975 1976 1979 1983 1986 1990 1993 1997 2002
Figure 57: Variation of HHDDT CO emissions by vehicle model year and by phase
of the program for the UDDS (56,000 Ibs.).
76
-------
60
50
40
30
20
10
Jl
Pre- 1975- 1977- 1980- 1984- 1987- 1991- 1994- 1998 1999- 2003+
1975 1976 1979 1983 1986 1990 1993 1997 2002
Figure 58: Variation of HHDDT oxides of NOX by vehicle model year and by phase
of the program for the UDDS (56,000 Ibs.).
• Phase 1
D Phase 1.5
IS Phase 2
• PhaseS
Pre- 1975- 1977- 1980- 1984- 1987- 1991- 1994- 1998 1999- 2003+
1975 1976 1979 1983 1986 1990 1993 1997 2002
Figure 59: Variation of PM emissions by vehicle model year and by phase of the
program for the UDDS (56,000 Ibs.).
77
-------
RESULTS AND DISCUSSION, NONREGULATED SPECIES
Semi-Volatile Organic Compounds
Exhaust Samples
Table 13 lists PAH, oxy-PAH, nitro-PAH, hopanes, steranes, organic acids and other
compounds of interest (with their mnemonics) quantified for this study. The minimum
detection limit (MDL) for PAH, oxy-PAH and other organic compounds was compound-
dependent, and in general, in the range of 20-50 pg/|il. For nitro-PAH analyzed by
negative ion CI method, the MDL was in the range of lpg/nl
Table 13: List of Target Analytes in the Particulate and Semi-Volatile Fractions.
Mnemonic
Polycyclic Aromatic
NAPHTH
MNAPH2
MNAPH1
BIPHEN
ENAP12
DMN267
DM1367
D14523
DMN12
M_2BPH
M_3BPH
M_4BPH
DBZFUR
ATMNAP
BTMNAP
CTMNAP
ETMNAP
FTMNAP
TMI235N
TM245N
JTMNAP
TM145N
ACNAPY
ACNAPE
FLUORE
PHENAN
A_MFLU
M_1FLU
B_MFLU
FL9ONE
XANONE
ACQUONE
PNAPONE
A MPHT
Compound Mnemonic
Hydrocarbons (PAH)and oxy-PAH
Naphthalene ANTHONE
2-methylnaphthalene ANRQUONE
1-methylnaphthalene DM3 6PH
Biphenyl A_DMPH
l+2ethylnaphthalene B_DMPH
2,6+2,7-dimethylnaphthalene C_DMPH
l,3+l,6+l,7dimethylnaphth DM17PH
1,4+1,5+2,3 -dimethylnaphth D_DMPH
1,2-dimethylnaphthalene E_DMPH
2-Methylbiphenyl ANTHRA
3-Methylbiphenyl M_9ANT
4-Methylbiphenyl FLUORA
Dibenzofaran PYRENE
A-trimethylnaphthalene ANTAL9
B-trimethylnaphthalene RETENE
C-trimethylnaphthalene BNTIOP
E-trimethylnaphthalene C1MFLPY
F-trimethylnaphthalene BMPYFL
2,3,5+I-trimethylnaphthalene CMPYFL
2,4,5-trimethylnaphthalene DMPYFL
J-trimethylnaphthalene M_4PYR
1,4,5-trimethylnaphthalene M_1PYR
Acenaphthylene BZCPHEN
Acenaphthene BAANTH
Fluorene M_7BAA
Phenanthrene CHRYSN
A-methylfluorene BZANTHR
1-methyffluorene BAA7_12
B-methylfluorene CHRY56M
9-fluorenone BBJKFL
Xanthone M_7BPY
Acenaphthenequinone BEPYRN
Perinaphthenone PERYLE
A-methylphenanthrene BAPYRN
Compound
Anthrone
Anthraquinone
3,6-dimethylphenanthrene
A-dimethylphenanthrene
B -dimethy Iphenanthrene
C-dimethylphenanthrene
1,7-dimethylphenanthrene
D-dimethylphenanthrene
E-dimethy Iphenanthrene
Anthracene
9-methylanthracene
Fluoranthene
Pyrene
9-Anthraaldehyde
Retene
Benzonaphthothiophene
1-MeFl+C-MeFl/Py
B-MePy/MeFl
C-MePy/MeFl
D-MePy/MeFl
4-methylpyrene
1-methylpyrene
Benzo(c)phenanthrene
Benz(a)anthracene
7-methylbenz(a)anthracene
Chrysene
Benzanthrone
Benz(a)anthracene-7,12-dione
5+6-methylchrysene
Benzo(b+j+k)fluoranthene
7-methylbenzo(a)pyrene
BeP
Perylene
BaP
78
-------
Mnemonic
M_2PHT
B_MPHT
C_MPHT
M_1PHT
Hopanes/Steranes
STER35
Compound Mnemonic
2-methylphenanthrene INCDPY
B-methylphenanthrene BGHIPE
C-methylphenanthrene DBANTH
1-methylphenanthrene CORONE
20S-13B(H), 17a(H)-diasterane STER52
STER36 20R-13B(H),17a(H)-diasterane HOP13
STER37 20S-13a(H),17B(H)-diasterane HOP14
STER38 20R-13a(H),17B(H)-diasterane STER53
STER39 20S-13B(H),17a(H)-diasterane HOP15
STER42 C27-20S5a(H),14a(H)- HOP 16
cholestane
STER43 20R5a(H),14B(H)-cholestane HOP17
STER44 C27-20S5a(H),14B(H),17B(H)- HOP 18
cholestane
STER45_40 20R5a(H),14a(H),17a(H)- HOP 19
cholestane&C29-
20S13B(H),17a(H)-diasterane
STER46 20S5a(H),14a(H),17a(H)- HOP20
ergostane
STER47 20R5a(H),14B(H),17B(H)- HOP21
ergostane
STER48 20S5a(H),14B(H),17B(H)- HOP22
ergostane
STER41 20R-13a(H),17B(H)-diasterane HOP23
HOP9 C27-tetracyclic terpane HOP24
STER49 20R5a(H),14a(H),17a(H)- HOP25
ergostane
HOP10 C27-tetracyclic terpane HOP26
HOP 11 C28-tetracyclic terpane HOP27
STER50 20S5a(H),14a(H),17a(H)- STER51
stigmastane
HOP12 C28-tetracyclic terpane
Polar Organic Compounds
HEXAC hexanoic acid HEPDAC
HEPTAC heptanoic acid ACVAN
BENAC benzoic acid PHTHAC
OCTANAC octanoic acid LEVG
MALEAC maleic acid TDECAC
SUCAC succinic acid ISPHAC
MEGUA4 4-me-guaiacol AZEAC
MESUCAC me-succinic acid MYRAC
NONAC nonanoic acid PDECAC
ETGUA4 4-ethyl-guaiacol PALAC
GUAC glutaric acid HEPTAD
Compound
Indeno [123 -cd]py rene
Benzo(ghi)perylene
Dibenzo(ah+ac)anthracene
Coronene
20S5a(H),14B(H),17B(H)-
stigmastane
18a(H),21B(H)-22,29,30-
Trisnorhopane
17a(H),18a(H),21B(H)-25,28,30-
Trisnorhopane
20R5a(H),14a(H),17a(H)-
stigmastane
17a(H),21B(H)-22,29,30-
Trisnorhopane
17a(H),18a(H),21B(H)-28,30-
Bisnorhopane
17a(H),21B(H)-30-Norhopane
18a(H),21B(H)-30-Norneohopane
17a(H),21B(H)-Hopane
17B(H),21a(H)-hopane
22S-17a(H),21B(H)-30-
Homohopane
22R-17a(H),21B(H)-30-
Homohopane
17B(H),21B(H)-Hopane
22S-17a(H),21B(H)-30,31-
Bishomohopane
22R-17a(H),21B(H)-30,31-
Bishomohopane
22S-17a(H),21B(H)-30,31,32-
Trisomohopane
22R-17a(H),21B(H)-30,31,32-
Trishomohopane
20R5a(H),14B(H),17B(H)-
stigmastane
heptanedioic (pimelic) acid
acetovanillone
phthalic acid
levoglucosan
tridecanoic acid
isophthalic acid
azelaic acid
myristic acid
pentadecanoic acid
palmitic acid
heptadecanoic acid
79
-------
Mnemonic Compound Mnemonic
SYRI syringol OLAC
DIMEB25 2,5-dimethylbenzoic acid ELAC
DIMEB24 2,4-dimethylbenzoic acid STEAC
DIME235 2,3- and 3,5- dimethylbenzoic NDECAC
acid
DECAC decanoic acid DHABAC
ALGUAI4 4-allyl-guaiacol (eugenol) ECOSAC
MESYR4 4-methyl-syringol ABAC
DIMEB34 3,4-dimethylbenzoic acid HCOSAC
HEXDAC hexanedioic (adipic) acid DOCOSA
CPINAC cis-pinonic acid TRICOSA
SALCYL salcylic acid TETRACO
FGUAI4 4-formyl-guaiacol (vanillin) CHOL
UNDEC undecanoic acid BSIT
ISEUG isoeugenol
Alkanes and Cycloalkanes
NOYCYHX Nonylcyclohexane EICOSA
HPYCYHX Heptylcyclohexane HENEIC
OCYCYHX Octylcyclohexane DOCOSA
NORFARN Norfarnesane TRICOSA
FARNES Farnesane DEC5YHX
TEDRAD Tetradecane DEC6YHX
PENTAD Pentadecane DEC7YHX
HEXAD Hexadecane_Norpristane DEC8YHX
HEPTAD Heptadecane_Pristane DEC9YHX
OCTAD Octadecane COSAN4
NONAD Nonadecane COSAN5
PHYTAN Phytane COSAN6
DECYHX Decylcyclohexane COSAN7
DEC1YHX Undecylcyclohexane CYHXEIC
DEC2YHX Dodecylcyclohexane CYHXHEN
DEC3YHX Tridecylcyclohexane COSAN8
DEC4YHX Tetradecylcyclohexane
Nitro-PAH
NI1NAPTH 1-nitronaphthalene NI2FLUOR
NI2NAPTH 2-nitronaphthalene NI3FLUOR
NI2BIPH 2-nitrobiphenyl NI1PYRE
NI3BIPH 3-nitrobiphenyl NI27FLUO
NI4BPH 4-nitrobiphenyl NI27FL9ON
NI13NAP 1,3-dinitronaphthalene NI7BZANTH
NI15NAP 1,5-dinitronaphthalene NI6CHRY
NI5ACEN 5-nitroacenaphthene NI13PYR
NI9ANTHR 9-nitroanthracene NI16PYR
NI4PHEN 4-nitrophenanthrene NI18PYR
NI9PHEN 9-nitrophenanthrene NI910ANTH
NI3PHEN 3-nitrophenanthrene NI6BAP
Nil SNAP 1,8-dinitronaphthalene
Compound
oleic acid
elaidic acid
stearic acid
nonadecanoic acid
dehydroabietic acid
eicosanoic acid
abietic acid
heneicosanoic acid
docosanoic acid
tricosanoic acid
tetracosanoic acid
cholesterol
sitosterol
Eicosane
Heneicosane
Docosane
Tricosane
Pentadecylcyclohexane
Hexadecylcyclohexane
Heptadecylcyclohexane
Octadecylcyclohexane
Nonadecylcyclohexane
Tetracosane
Pentacosane
Hexacosane
Heptacosane
Eicosylcyclohexane
Heneicosylcyclohexane
Octacosane
2-nitrofluoranthene
3 -nitrofluoranthene
1-nitropyrene
2,7-dinitrofluorene
2,7-dinitrofluoren-9-one
7-nitrobenz(a)anthracene
6-nitrochrysene
1,3-dinitropyrene
1,6-dinitropyrene
1,8-dinitropyrene
9,10-dinitroanthracene
6-nitrobenz[a]pyrene
The data were delivered in electronic format in units of ng/m . No unusual events for this
group of samples were noticed.
80
-------
Fuel and Oil Samples
Six fuel and five oil samples were analyzed for PAH and hopanes/steranes. The data
were presented in units of ug/ml and ug/g. Figure 60 shows the PAH concentrations in
diesel fuel and oil. For simplicity, the methylated isomeric PAH were added together
(i.e. dimethylnaphthalenes, trimethylnaphthalenes, etc). As seen from this figure, the
lower molecular weight (MW) PAHs were much more abundant in diesel fuel than in the
oil. Methylated PAH account for the majority of the PAH.
Lower MW PAH
30000
CRC44 CRC43 CRC42 CRC41
FUEL FUEL FUEL FUEL
CRC40
FUEL
CRC39 CRC44
FUEL OIL
CRC43
OIL
CRC42 CRC40
OIL OIL
CRC39
OIL
D naphth
• acnapy
S mphenan
D mnaph2
D acnape
D anthone
S mnaphl
D fluore
D anrquone
D biphen
• phenan
Sdmphenan
D dmnaphth
3 mflu
D mbiphen
• f!9one
• dbzfur
• xanone
D bibenz
D acquone
Htmnaphth
D pnapone
Higher MW PAH
CRC44
FUEL
CRC43
FUEL
CRC42
FUEL
CRC40
OIL
CRC39
OIL
Danthra •m_9ant Dfluora n pyrene HantalQ D retene •bntiop 0 mflu/mpy • bzcphen • baanth D m_7baa Dchrysn
• bzanthr • baa7_12 • chry56m • bbjkfl D m_7bpy D bepyrn Dperyle D bapyrn D incdpy Dbghipe Ddbanth Dcorone
Figure 60: PAH in fuels and oils.
81
-------
Volatile Organic Compounds
Canister Samples
Table 14 lists the compounds quantified on-site from the canister samples by the GC/MS
method. The MDL was 0.1 ppbv for all compounds.
Table 14: List of Gas-phase Compounds Quantified by GC/MS Method from
Canisters.
Mnemonic
ethane
ethene
acetyl
Ipropa
Iprope
Ipropy
ibuta
butle_ibute
bud!3
butan
t2but
c2but
bud!2
ipent
pentel
ble2m
n_pent
i_pren
t2pene
c2pene
b2e2m
bu22dm
cpente
cpenta
bu23dm
pena2m
penaSm
ple2me
n_hex
t2hexe
p2e2me
p2e3mt
c2hexe
p2e3mc
hxdilS
mcypna
pen24m
cpenel
Compound
ethane
ethene
acetylene
propane
propene
propyne
iso-butane
1-butene + isobutene
1,3-butadiene
n-butane
t-2-butene
c-2-butene
1,2-butadiene
iso-pentane
1-pentene
2-methyl-1 -butene
n-pentane
isoprene
t-2-pentene
c-2-pentene
2-methyl-2-butene
2,2-dimethylbutane
cyclopentene
cyclopentane
2,3 -dimethylbutane
2-methylpentane
3-methylpentane
2-methyl-1 -pentene
n-hexane
t-2-hexene
2-methyl-2-pentene
3 -methy 1-2-pentene
c-2-hexene
cis-3-methyl-2-pentene
1,3-hexadiene (trans)
methylcyclopentane
2,4-dimethylpentane
cyclopentene
Mnemonic
pa224m
n_hept
p2e23m
t!3dcp
mecyhx
pa234m
tolue
hx23dm
hep2me
hep4me
hep3me
hex225
n_oct
etbz
mp_xyl
oct3me
styr
o_xyl
n_non
iprbz
ipcyhex
a_pine
n_prbz
m_etol
p_etol
bz!35m
o_etol
b_pine
bz!24m_tbutbz
n_dec
bz!23m
limon
indan
detbzlS
detbz!4
n_bubz
prtol
iprtol
Compound
2,2,4-trimethylpentane
n-heptane
2,3 -dimethyl-2-pentene
t-1,3 -dichloropropene
methylcyclohexane
2,3,4-trimethylpentane
toluene
2,3 -dimethylhexane
2-methylheptane
4-methylheptane
3-methylheptane
2,2,5-trimethylhexane
n-octane
ethylbenzene
m/p-xylene
3-methyloctane
styrene
o-xylene
n-nonane
i-propylbenzene
isopropylcyclohexane
alpha-pinene
propylbenzene
3-ethyltoluene
4-ethyltoluene
1,3,5-trimethylbenzene
2-ethyltoluene
beta-pinene
1,2,4-trimethylbenzene+t-
butylbenzene
n-decane
1,2,3 -trimethylbenzene
limonene
indan
1,3 -diethylbenzene
1,4-diethylbenzene
butylbenzene
propyltoluene
isopropyltoluene
82
-------
Mnemonic Compound Mnemonic Compound
benze benzene n_unde n-undecane
cyhexa cyclohexane bz!245 1,2,4,5-tetramethylbenzene
hexa2m 2-methylhexane bz!235 1,2,3,5-tetramethylbenzene
pen23m 2,3-dimethylpentane ind_2m 2-methylindan
hexa3m 3-methylhexane ind_lm 1-methylindan
cyhexe cyclohexene
cpal3m 1,3-dimethylcyclopentane (cis)
heple 1-heptene
After the first dynamic system background run, a very high acetone peak that obscured
the iso-pentane peak was noticed. This peak persisted for a few subsequent runs but
gradually diminished. The system (i.e. dilution tunnel and all sampling lines) was
flushed overnight with ambient air on January 13, and the peak was nearly gone.
Unfortunately it reappeared again on January 21. In addition, some other peaks identified
as n-pentane, n-hexane, and 2-methyl-l-pentene were unusually high for some runs.
Clearly, this appears to be a contamination with some solvent, or mixture of solvents, but
the origin of the contamination was not determined.
Carbonyl Compounds
Table 15 lists the carbonyl compounds collected on DNPH cartridges and quantified by
the OTLC method. The MDL for carbonyl compounds was approximately 0.1 ppbv.
Table 15: List of Target Analytes in the Gas-Phase Carbonyl Compound Fraction
Mnemonic Compound
FORMAL Formaldehyde
ACETAL Acetaldehyde
ACETO Acetone
ACROLN Acrolein
PROAL Propionaldehyde
CROTON Crotonaldehyde
MEK Methyl ethyl ketone
MACROL Methacrolein
BUTAL Butyraldehyde
BENZAL Benzaldehyde
GLYOXL Glyoxal
VALAL Valeraldehyde
TOLUAL M-tolualdehyde
HEXAL Hexanaldehyde
The acetone contamination is clearly visible in these samples as well. Figure 61 shows
the concentrations of carbonyl compounds (in ppbv) measured during various runs
(shown in chronological sequence).
83
-------
Carbonyls
\
K
*>?
$
-------
Tenax Samples
Table 17 lists compounds quantified from Tenax cartridges. The MDL for these
compounds was in the range of 0.1 ppbv per sample.
Table 17: List of Heavy Hydrocarbons for Tenax Samples
Mnemonic Compound Mnemonic
HEPAL Heptanal IPRXYL_5
N_NON Nonane BZ1245
IPRBZ Isopropylbenzene BZ1235
N_PRBZ Propylbenzene IAMBZ
M_ETOL m-ethyltoluene IND_2M
P_ETOL p-ethyltoluene IND_1M
BZ135M 1,3,5-trimethylbenzene BZ1234
PHENOL Phenol DIPRB_13
O_ETOL o-ethyltoluene C5BZ_3
FURBZ 2,3-benzofuran THNAPH
FURPEN 2-pentylfuran DHNAPH
T_BUBZ t-butylbenzene DIPRB_14
OCTAL Octanal NAPHTH
BZ124M 1,2,4-trimethylbenzene INDDMA
MESTYR 4-methylstyrene INDDMB
MPCBZ 1,3-dichlorobenzene INDDMC
DEC IE 1-decene INDDMD
I_BUBZ Isobutylbenzene DECONE2
N_DEC Decane DECAL
S_BUBZ Sec-butylbenzene DODE1E
BZ123M 1,2,3-trimethylbenzene N_DODE
MJPRTOL m-isopropyltoluene PMEBZ
PJPRTOL p-isopropyltoluene NAP_2M
ODCBZ 1,2-dichlorobenzene NAP_1M
INDAN Indan N_TRID
INDENE Indene BIPHEN
OJPRTOL o-isopropyltoluene EN API 2
O_MEPHOL o-methylphenol DMN267
DETBZ1 1,3-diethylbenzene N_TETD
M_TOLALD m-tolualdehyde DM1367
TOL4PR 4-n-propyltoluene + 1,4-diethylbenzene D14523
BUTBZ Butylbenzene ACENAP
M_XYLET5 5-ethyl-m-xylene DMN12
DETBZ3 1,2-diethylbenzene ACENPE
MP_MEPHO m/p-methylphenol N_PEND
TOL2PR 2-n-propyltoluene FLUORE
P_XYLET2 2-ethyl-p-xylene N_HEXD
O_XYLET4 4-ethyl-o-xylene N_HEPD
TBUTOL_4 4-tert-butyltoluene PHENA
NONAL Nonanal N_OCTD
UNDE1E 1-undecene N NOND
Compound
5 -isopropyl-m-xylene
1,2,4,5-tetramethylbenzene
1,2,3,5-tetramethylbenzene
Isoamylbenzene
2-methylindan
1-methylindan
1,2,3,4-tetramethylbenzene
1,3 -diisopropy Ibenzene
Pentylbenzene
1,2,3,4-tetrahydronaphthalene
1,2-dihydronaphthalene
1,4-diisopropy Ibenzene
Naphthalene
A-dimethylindane
B-dimethylindane
C-dimethylindane
D-dimethylindan
2-decanone
Decanal
Dodecene
Dodecane
Pentamethy Ibenzene
2-methylnaphthalene
1 -methylnaphthalene
Tridecane
Biphenyl
1+2-ethylnaphthalene
2,6+2,7-dimethylnaphthalene
Tetradecane
1,6+1,3+1,7-dimethy Inaphthalene
2,3+1,5+1,4-dimethylnaphthalene
Acenaphthylene
1,2-dimethy Inaphthalene
Acenaphthene
Pentadecane
Fluorene
Hexadecane
Heptadecane
Phenanthrene
Octadecane
Nonadecane
85
-------
Mnemonic
FUBZ2ME
N UNDE
Compound
2-methylbenzofuran
Undecane
Mnemonic
N EICO
Compound
Eicosane
Since Tenax samples cannot be combined for analyses, each replicate run was analyzed
separately. Nine samples were invalidated due to problems with the B sampling channel
and damage to some of the A cartridges during transport. These samples are shown in
Table 18.
Table 18: Damaged Tenax Samples.
Cycle
CruiseS
CruiseS
MHDTLO
MHDTLO
MHDTCR
MHDTCR
IdleS 3
IdleS 3
IdleS 3
TransS
CruiseS
CruiseS
Test ID
2863
3863
2879
2879
2879
2879
2907
2913
2913
2921
2921
2921
Test Run
ID
09
10
06
07
08
09
03
01
02
01
05
06
86
-------
SPECIATION AND SIZING RESULTS
This section discusses exhaust particle size distributions (SMPS and DMS500) and
chemical speciation results for five heavy heavy-duty diesel trucks operating on various
modes of the California HHDDT cycle. Particle size distribution results from the SMPS
were corrected for dilution ratios. Since dilution ratio on the main tunnel was not
accurately known, DMS500 data presented in this report are for dilute flow in the main
CVS tunnel. The raw uncorrected results obtained from DRI's analytical laboratory for
chemical speciation were presented in ng/m3. All speciation data were reduced by
correcting for flow rates in the sampling system. The mass values for chemical speciation
were found using the equation
g/cycle of chemical species= (Vtunnel* Speciated sample in ng/ m3)/ l.OOE+09.
Vtunnel = Vmix + Total PM volume + PMi0 volume + DRI volume (m3).
Vmix was the total integrated volume sampled from the venturi for a particular cycle.
Total PM volume was the integrated volume sampled from the secondary dilution tunnel
for a particular cycle. PMi0 volume was the integrated volume sampled from the PMi0
cyclone for a particular cycle. DRI operated three PM2.5 cyclones at 113 1pm each; e.g.,
0.34 m3/min total. To obtain the total DRI sample volume, 0.34 m3/min was multiplied by
the sample time. The PMio and DRI flows were small with respect to tunnel flow.
Chemical speciation results are usually reported in g/mile, but since Idle mode was also
considered for comparison purposes, all speciation data are presented in g/hour. Only for
chemical speciation, both the Cruise Modes were integrated and results reflect the
combination of both the normal and High-speed Cruise Modes.
The data are depicted as "Tunnel Background" and "Test, Uncorrected" to illustrate the
magnitude of reductions, while being mindful of the role the tunnel background plays
when the primary emissions are so low.
Results presented for chemical speciation were not corrected for background values,
which is a departure from regulatory compliance reporting, and is different from the
treatment of regulatory emissions presented in this report. Background data are provided
separately. There are two barriers to representing data with background correction. First,
the exhaust dilution ratio in the main CVS was not known. The background correction for
regulated species was performed using a dilution factor, prescribed in the Code of Federal
Regulations (CFR), and this factor is not equal to the ratio. Second, background levels of
each species are known to vary with time, so that accurate quantitative correction is not
possible without parallel sampling trains. Finally, the role played by the tunnel
background, and tunnel history effects in speciation results is an unknown.
87
-------
E55CRC-39
E55CRC-39 was powered with a Cummins ISX 530 MY 2004 engine. Particle sizing and
chemical speciation results are presented below for various modes of the HHDDT
schedule. Figure 62 and Figure 63 present the SMPS particle size distribution for Idle and
steady modes.
A very distinct clear bimodal distribution was noted for the Idle mode with a nuclei mode
in the particle size range of 18 nm and an accumulation mode of 157 nm with
concentrations in the range of 3.5xl05 and l.SxlO5, respectively. The nuclei mode
formation is attributed to low engine load at Idle conditions. Exhaust temperatures are
low under idling conditions, which lead to formation of higher levels of volatile organic
compounds.
These volatile compounds are usually in the gas phase in the tailpipe. As the exhaust is
diluted and cooled in the atmosphere or the dilution system, these volatile compounds
may either adsorb the nucleated sulfuric acid and/or undergo homogenous nucleation to
form particles. Some semi-volatile particles adsorb onto existing soot particles and
increase their size and these particles are measured as the agglomeration mode particles.
Nuclei mode particles do not contribute much towards Idle mass emissions as seen in
Figure 64, but their number concentrations are high.
4.00E+05
1.00E+01
10
100
Particle Diameter (nm)
1000
Figure 62: SMPS Particle Size Distribution for E55CRC-39 Operating on an Idle
Mode
-------
4.00E+06
—•—Steady State
CruiseS
HHDDT S
1.00E+01
10 Particle Diameter (nm) 10°
1000
Figure 63: SMPS Particle Size Distribution for E55CRC-39 Operating on Various
Steady Cycles
14
12
10
Idle
CruiseS
HHDDT S
Figure 64: PM Mass Emissions for E55CRC-39
Chemical speciation of exhaust from E55CRC-39 showed that volatile organic compound
emissions were very high in the Idle mode at 0.68 g/hour when compared to the Cruise
mode at 0.21g/hour (Appendix O: Figure Ol to O5). In the semi-volatile group, polar
compound emissions for the Cruise mode were very similar to Idle mode emissions
89
-------
(Appendix O: Figure O9). The elemental carbon emissions for the Cruise mode were very
high at 8.3 g/hour when compared with the Idle mode at 0.21 g/hour (Figure 66). At idle,
the volatile compounds experience lower temperatures after being emitted or in the tail
pipe. The saturation ratios (partial pressure/saturation pressure) are also higher in light of
the fact that elemental carbon is present in very low quantities; hence, idle operation PM
emissions are driven by the lower exhaust temperatures and higher saturation ratios. The
lower concentration of elemental carbon contributes to nucleation of nano-particle
formation. These analyses clearly indicate that, for this vehicle, the nano-particles or
nuclei mode particles which are measured during the Idle Mode, are mostly made up of
volatile compounds. The heavier organic species undergo condensation onto the hetro-
molecular complex of water, sulfuric acid and lubricating oil based nano-sized ash
particles. For the steady state mode at 40 kmph (25mph), the Count Median Diameter
(CMD) of the particle size distribution was 48.89 nm with a maximum concentration
level of 1x106 particles/cm3. For the High-speed Cruise and normal Cruise Mode, the
particle size distribution was very similar with a CMD of 75.91 and 71.84 nm,
respectively. The concentration levels for the Cruise Modes were almost twice that of the
steady state mode and almost ten times that of the Idle Mode.
The geometric standard deviation (GSD) for the steady state, Cruise and High-speed
Cruise Modes were 1.83, 1.84 and 1.79, respectively. The geometric standard deviation
represents the distance of shift from the mean of the whole distribution. GSD values close
to 1.8 represent a good distribution about the mean for diesel exhaust. For the Idle mode
the GSD was 2.9 which indicates that the distribution was skewed and this can also be
inferred from Figure 62. The shift towards a higher particle size in the Cruise Modes was
likely due to a higher concentration of carbonaceous agglomerates in the exhaust, which
also tend to suppress nano-particle formation (see Figure 63). Sulfate emissions were
almost 20 times higher in the Cruise mode than in the Idle mode (Figure 66). Acetone
emissions were unusually high for the Transient mode at 212.4 g/hour (Appendix O:
Figure Oil).
On a mass basis, the contribution of nitro-PAH and hopanes & steranes compounds
(Appendix O: Figure O8 and O10) were not significant. Sulfates along with semi-volatile
and carbonyl compounds usually adsorb onto carbon particles and form a cluster-like
structure which increases the size of the particles, and the size spectrum is shifted
towards the larger accumulation mode which is evident from Figure 63. These analyses
show that accumulation mode particles formed during the Cruise mode are mostly made
up of semi-volatile compounds with a solid carbonaceous core.
90
-------
Figure 65: Ion Composite Results for E55CRC- 39
0 4
Trans
Cruise
Figure 66: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-39
91
-------
2.50E+06
O.OOE+00
-100
200 400 600 800 1000
Time (s)
•Total Number (N/cc) Vehicle Speed (mph)
Figure 67: DMS500 Total Number of Particles and Vehicle Speed vs. Time during
the UDDS for E55CRC-39 tested at 56,000 Ibs.
Comprehensive DMS500 data are available for E55CRC-39. Figure 67 shows the total
number of particles and vehicle speed versus time during the UDDS for E55CRC-39. The
total number goes as high as l.SxlO6 particles/cm3 during steep accelerations and goes
down to l.SxlO4 during idle conditions. There is no strong correlation between the total
number and engine speed, but as can be observed from Figure 67, during accelerations,
the total number was high and when the high speed was continued for a period, the total
number started to drop, e.g., between t = 600 and 700s. During decelerations, the total
number dropped to a minimum and continued that way during the idle period.
Figure 68 shows the total number of particles and engine power versus time. Figure 69
shows the distribution of the total number concentration of particles in the 60 and 20 nm
bins versus time. The 60 nm bin contains particles of diameters between approximately
55 and 65 nm, and the diameters for the 20 nm bin range between 17.5 and 22.5 nm. As
can be observed from the plot, the overall number of 60 nm particles, which represent the
accumulation mode particles, are almost an order of magnitude higher than the total
number of particles in the 20 nm bin. Figure 70 shows the total number of particles in the
60 nm bin versus the total number of particles in the 20 nm bin. There is a much sharper
increase in the total number of particles in the 20 nm bin during decelerations and a
sharper increase in the total number of particles in the 60 nm bin during accelerations. As
can be seen, there is no clear correlation between the two size bins.
92
-------
2.50E+06
- 500
O.OOE+00
-2000
200 400 600
Time(s)
800
1000
•Total Number of Particles (N/cc) Engine Power (hp)
Figure 68: DMS500 Total Number of Particles and Engine Power vs. Time during
the UDDS for E55CRC-39 tested at 56,000 Ibs.
u
JU
•J
u
ju
S
=
®
200
400 600
Time (s)
800
1000
-60 nm 20 nm
Figure 69: DMS500 60 nm and 20 nm Particle Number vs. Time during the UDDS
for E55CRC-39 tested at 56,000 Ibs.
93
-------
u
u
a
o
1.40E+05
1.20E+05
l.OOE+05
8.00E+04
6.00E+04
• y**
» »/• .«
= 1.2398x+12597
2 = 0.0365
O.OOE+00
O.OOE+00 5.00E+03 l.OOE+04 1.50E+04 2.00E+04 2.50E+04
20 nm bin (N/cc)
Figure 70: DMS500 60 nm Particle Number vs. 20 nm Particle Number during the
UDDS for E55CRC-39 tested at 56,000 Ibs.
1
I
S
=
-800
600
Total Hub Power (hp)
Figure 71: DMS500 60 nm Particle Number vs. Total Hub Power during the UDDS
for E55CRC-39 tested at 56,000 Ibs.
94
-------
Chart Title
3.EI04
u
u
-Q
s
s
z
e
M
S
o
>.Ef04
.E;04
-800 -600 -400 -200 0
Total Hub Power (hp)
200
400
600
Figure 72: DMS500 20 nm Particle Number vs. Total Hub Power during the UDDS
for E55CRC-39 tested at 56,000 Ibs.
Figure 71 shows the total number concentration of particles in the 60 nm bin versus the
total hub power. There is no strong correlation, but the total number shows an upward
trend with the positive hub power range. Since 60 nm particles are in the accumulation
mode, one would expect their count to rise when higher power operation demands lower
air to fuel ratio. Figure 72 shows the total number concentration of particles in the 20 nm
bin versus the total hub power. Again, there is no strong correlation between the two
series but in this case as one can observe from the plot, there are more particles in the
negative hub power periods, that is during decelerations. This is something to be
expected since nuclei mode particles are dominant during the deceleration phases.
The following plots show the size distribution of particles during idle, acceleration and
deceleration, respectively for E55CRC-39. As can be observed from Figure 73, particles
of around 150 nm are mostly observed. This peak size agrees well with the 157 nm peak
for the SMPS data, shown in Figure 62, but the nuclei mode peak evident in the SMPS
plot is absent. One must note that the dilution system used for the two devices differed.
During acceleration, as shown in Figure 74, a bimodal distribution is observed where
there are two peaks centered at around 58 and 135 nm, respectively. This plot shows the
transition from the Idle Mode, where particles of diameter 150 nm are dominant, to the
acceleration mode, where particles of diameter around 60 nm are observed, which are the
accumulation mode particles. Strong modes in the 50 to 100 nm range were observed for
the SMPS data shown in Figure 63. Figure 75 shows the distribution during the
deceleration region of the HDDS. Again, a bimodal distribution is seen. This time, the
95
-------
transition is from deceleration to Idle Mode. The first peak is centered around 15 nm,
which are mostly nuclei mode particles, and the second peak is centered around 175 nm,
which is the Idle region distribution.
Idle (t = 230 s)
u
JU
4500
4000
3500
3000
'•£ 2500
2000
1500
1000
500
100
200 300
400 500 600
Diameter (nm)
700
800 900
1000
Figure 73: DMS500 size distribution of particles during Idle (t = 230 s) of the UDDS
for E55CRC-39 tested at 56,000 Ibs.
96
-------
Acceleration (t = 530 s)
o
o
18000
16000
14000
12000
'• 10000
8000
-Q
S
6000
4000
2000
0 100 200 300 400 500 600 700 800 900 1000
Diameter (nm)
Figure 74: DMS500 size distribution of particles during acceleration (t = 530 s) of
the UDDS for E55CRC-39 tested at 56,000 Ibs.
Deceleration (t = 770 s)
4500
4000
3500
3000
'•£ 2500
£
2000
1500
1000
500
0 100 200 300 400 500 600 700 800 900 1000
Diameter (nm)
Figure 75: DMS500 size distribution of particles during deceleration (t = 770 s) of
the UDDS for E55CRC-39 tested at 56,000 Ibs.
97
-------
l.E+07
9.E+06
•HHDDT_S (1)
•HHDDT S (2)
•Vehicle Speed
100
200
300 400
Time(s)
500
600
700
Figure 76: DMS500 size distribution of particles during HHDDT_S and Vehicle
Speed vs. Time for E55CRC-39 tested at 56,000 Ibs.
2.E+06 -rf
-TRANS3(1)
TRANS3 (2)
-2 4.E+05
2.E+05
O.E+00
Figure
-80
100
200
300 400
Time (s)
500
600
77: DMS500 size distribution of particles during Transient and Vehicle
Speed vs. Time for E55CRC-39 tested at 56,000 Ibs.
98
-------
Figure 76 shows the total number of particles, determined using the DMS500 and vehicle
speed versus time during HHDDT_S for E55CRC-39. As can be observed, the total
number of particles is as high as 9xl06 during steep accelerations, and drops to 9xl04
particles/cm3 during idle regions. Looking at the data profiles for the two tests, the
problem may be that the DMS500 inlet probe was not properly mounted on the second
HHDDT_S test, thus sucking in air from the ambient and causing a decrease in the
particle number concentration readings."
Figure 77 shows the total number of particles, measured by the DMS500, and vehicle
speed versus time for two different Transient runs. Again, as it is observed from previous
plots, the total number of particles increases during accelerations and decreases during
decelerations. During the High-speed Cruise section, the total number starts to decrease
due to the decrease in the total number of accumulation mode particles. Repeatability is
good for the two traces in Figure 77.
r 60
3.E+06
CRUISE3 (1)
CRUISE3 (2)
Vehicle Speed
O.E+00
-80
500
1000
Time (s)
1500
2000
Figure 78: DMS500 size distribution of particles during CruiseS and Vehicle Speed
vs. Time for E55CRC-39 tested at 56,000 Ibs.
Figure 78 shows the total number of particles and vehicle speed versus time for two
different Cruise runs. During acceleration, the total number of particles increases and
during decelerations, the total number decreases. Repeatability is not as good as
illustrated in Figure 77, and the researchers speculate that this may have been due to the
accumulation of PM in the instrument. This was the first operation of the instrument in
the field by the researchers, and protocols for cleaning the instrument were not well
established at time of testing.
99
-------
E55CRC-40
E55CRC-40 was a powered by a Detroit Diesel Series 60, MY 2003 engine. Particle
sizing and chemical speciation results are presented below for various modes of the
HHDDT schedule. Figure 79 and Figure 80 present the SMPS particle size distributions
for Idle and steady modes.
A nuclei mode dominated distribution was seen for the Idle mode with a count median
diameter (CMD) of 25.88 nm with a concentration of 8xl04 particles/cc and a geometric
standard deviation of 1.64. These concentration levels were almost four times lower than
those observed for E55CRC-39. This drop in concentration could be attributed to
difference in the emissions of volatile organic compounds. The total VOC's emitted for
E55CRC-40 at 2.01 g/hour during the Idle mode was almost eight times lower in
magnitude as compared to E55CRC-39 at 15.51 g/hour. The VOC mass emission rates
reported here include contributions from unburnt fuel and oil. It should be noted that a
recent study by Sakurai et al.13 indicated that the volatile component of diesel particles
was comprised primarily of unburnt lubricating oil, and that contributions from unburnt
fuel, oxidized organic combustion products, and sulfuric acid were small. Sakurai et al.13
also reported that the organic component of diesel nanoparticles was comprised of
compounds with carbon numbers in the C24 - C32 range, which are almost entirely
derived from unburnt oil.
9.00E+04
to
<
o 8.00E+04
%
Q 7.00E+04
I
1 6.00E+04
° 5.00E+04
13
o
o
4.00E+04
3.00E+04
z
I 2.00E+04
| 1.00E+04
ro
o.
1.00E+01
10
100
Particle Diameter (nm)
1000
Figure 79: SMPS Particle Size Distribution for E55CRC-40 Operating on an Idle
Mode
100
-------
1.00E+01
10
100
Particle Diameter (nm)
1000
Figure 80: SMPS Particle Size Distribution for E55CRC-40 Operating on Various
Steady Cycles
VOC results were very similar for E55CRC-40 when Idle and Cruise mode were
compared (Appendix O: Figure O13 to O17). Emissions of n-hexane and 2-methyl-l-
pentene were high during the Transient mode when compared with the Idle and Cruise
Modes (Appendix O: Figure O15 and O16). For the steady state mode, the CMD for the
particle size distribution was 84.96 nm with a concentration of 4.5xl05 particles/cm3 and
a GSD of 1.75. Particle size distributions were markedly different for the two Cruise
Modes with a CMD of 45.1 nm for the normal Cruise mode and 62.24 nm for the High-
speed Cruise Mode. SMPS size distributions observed in the Cruise Modes were
characterized by the accumulation mode particles.
The shift of the size spectrum towards a higher accumulation range during the Cruise
Modes when compared to the Idle mode could be attributed to higher concentrations of
carbonaceous agglomerates (EC emissions) and higher sulfate emissions during the
Cruise Mode. Sulfate emissions were very high during the Cruise Mode at 0.27 g/hour
when compared with the Idle mode at 0.015g/hour (Figure 83). Sulfate emissions serve as
the precursors of particle formation seen in the nuclei mode. Sulfates would undergo a
hetero-molecular nucleation with water molecules to form complex clusters. Subsequent
adsorption of volatile organics and agglomeration with the elemental carbon results in the
distribution seen in Figure 80. Elemental carbon emissions for the Cruise Mode were
high at 5.5 g/hour when compared to the Idle mode at 0.07 g/hour (Figure 83), which also
contributed to most of the mass emissions (Figure 82). Semi volatile emissions were
similar for Idle and Cruise Modes except for alkanes and hopanes & steranes which were
higher during the Cruise Mode (Appendix O: Figure O18 to O22).
101
-------
8
7
6
,- 5
o
.c
o>4
3
2
1
Idle
CruiseS
HHDDT S
Figure 81: PM Mass Emissions for E55CRC-40
0.50
0.45
0.40
0.35
0.30
0.25
)
0.20
0.15
0.10
0.05
0.00
Idle
Trans
CRC40
Cruise
Figure 82: Ion Composite Results for E55CRC-40
102
-------
Figure 83: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-40
For E55CRC-40, DMS500 data proved unreliable, and were characterized by spikes in
the output data that did not represent credible particle counts.
E55CRC-41
E55CRC-41 was the only MHDT subjected to speciation.
CRC41 was a Medium Heavy-duty Vehicle (MHDT) powered by a MY 1997 Cummins
B5.9, 210 hp engine. The vehicle was tested over a lower speed transient, a higher speed
Transient, and the Cruise Mode. Additionally, PM concentration and size distribution
data was collected over the MHDT schedule that comprised of driving modes mentioned
above. Figure 84 shows the PM size distribution over the idle mode.
103
-------
Particle Concentration at Idle (using LDMA)
1.00E+06
1.00E+01
1.00E+00
1000
Figure 84: SMPS Particle Size Distribution for E55CRC-41 Operating on an Idle
Mode
The size distribution during idle operation (Figure 84) showed a nuclei mode centered at
40 nm and an accumulation mode of 136 nm with concentrations in the range of 1.4xl05
and 1.3xl05, particles/cc, respectively. Again, the lower exhaust temperatures under
idling conditions, and the relatively low carbonaceous soot (resulting in higher saturation
ratios) leads to formation of higher levels of nuclei mode particles, primarily from the
organics compounds originating from the lubricating oil. Recent studies by Gautam et
al.14 on natural gas-fueled heavy-duty vehicles have shown evidence that supports the
argument that lubricating oil significantly contributes to nuclei mode particles. When
total surface area provided by soot particles is reduced the nucleation process is
advanced. Hence, in driving modes where the elemental carbon component of the total
PM emissions is lower, the nuclei mode particles are emitted in higher concentrations,
provided the exhaust temperatures and relative humidity are conducive to nucleation15'16.
The binary nucleation theory of water and sulfuric acid has been shown to underestimate
the formation of nucleation mode particles by several orders of magnitude17. However,
organic compounds have been considered as a key species to control the growth of
nucleation mode particles18.
Chemical speciation of exhaust from CRC41 showed that most volatile organic
compounds, except for €2 to Cs alkanes, were higher in the Transient and Cruise Modes
than in the idle mode of operation (Appendix O: Figure O25 to Figure O36). The polar
compound emissions for the transients were the highest (Appendix O: Figure O33). The
elemental carbon emissions for the Cruise Mode were nearly seven times higher than the
Idle emissions (Figure 85). Figure 86 shows that total PM mass emission rates in the
104
-------
cruise mode are dominated by the elemental carbon emissions. However, the organic
carbon emissions are also significantly higher under cruise mode than under idle and
transient modes. This breakdown is reflected in the PM concentrations and size
distributions, as seen in Figure 84, Figure 87 and Figure 88). Cruise operation at a vehicle
inertia of 75% GVWR shows a nuclei mode peak at 26.9 nm with a concentration of 3.9
x 105 particles/cc when measured with a nano-DMA. The accumulation mode, with a
CMD of 76.4 nm is similar for the cruise and steady state (25 mph) operation. Particle
concentrations at 50% GVWR at cruising conditions did not produce any significant
concentrations of nano-particles. This may be attributed to lower emissions of volatile
species. Compared to the Idle and Transient Modes, emissions of sulfate emissions were
significantly higher in the Cruise Mode (Figure 89). As noted earlier, binary nucleation of
sulfuric acid and water do contribute to formation of nucleation mode particles, which
may or may not survive in the exhaust depending upon levels of carbonaceous soot
present in the exhaust stream.
MDTransI MDTrans2
Runs
Figure 85: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-41
105
-------
25
20
15
10
MD Idlel
MD Trans!
MD Cruise 1
Figure 86: Total PM Mass Emissions Rate (g/hour) for E55CRC-41
Particle Concentration at 75% GVWR
Particle Dia(nm)
Cruise-NDMA
Cruise-LDMA
Steady State at 25mph
Figure 87: SMPS Particle Size Distribution for E55CRC-41 Operating under Cruise
and Steady Conditions
106
-------
Paricle Concentration at 50% GVWR
Particle Dia(nm)
Figure 88: SMPS Particle Size Distribution for E55CRC-41 Operating on Cruise
and Steady Conditions (50% GVWR)
2.50
2.00 -
1.50
1.00
0.50
0.00
• Soluble Calcium
DSolu
• Solu
DSolu
• Amm
DSulfa
DNitra
• Nitrit
DChlo
ule Potassium
ule Magnesium
ule Sodium
onium
te
e
3
ide
r^
MDIdle
__
MD Trans
Runs
^^^™
MD Cruise
Figure 89: Ion Composite Results for E55CRC-41
107
-------
Figure 90 to Figure 92 show traces of single particle tracking (82 nm and 24 nm) under
transient operations. At 75% GVWR, the 82 nm particles are emitted at higher
concentrations than the 24 nm particles. The concentration of 82 nm particles decreases
by an order or magnitude at lower inertia weight (50% GVWR).
CRC 41 MHDT at 75% GVWR
Tracking 82nm Particle
200 250 300
Time (sees)
Figure 90: SMPS Single Particle Trace (82 nm) for E55CRC-41 Operating on
UDDS, Low Transient and High Transient Cycles (75%GVWR)
108
-------
CRC 41 Transient Cycle at 75% GVWR
Tracking 24nm particle
150 200
Time(secs)
Figure 91: SMPS Single Particle Trace (24 nm) for E55CRC-41 Operating on Low
Transient and High Transient Cycles (75% GVWR)
1.00E+08
1.00E+07
£• -i.ooE+06
| 1.00E+04
u
c
O
O
IS 1.00E+03
1.00E+02
CRC 41 Transient Cycle at 50% GVWR
Tracking 88.2nm particle
•AC5080
-Test D
Low Transient
-High Transient
150 200
Time(secs)
Figure 92: SMPS Single Particle Trace (88 nm) for E55CRC-41 Operating on Low
Transient and High Transient Cycles (50% GVWR)
109
-------
Figure 93 shows the DMS500 total number concentration of particles and vehicle speed
versus time during the UDDS for E55CRC-41. The total number is as high as 1.7xl06
during steep accelerations and drops to l.SxlO4 during Idle conditions. (There was an
error on the readings of the instrument until t = 370s in Figure 93) In Figure 93 the high
count between 620 and 750 seconds is due to a high count of 20 nm particles over that
time. The overall count versus time differs in behavior between this vehicle and heavy-
duty vehicles, such as E55CRC-39 (see Figure 67 for HDDS data) and E55CRC-42 and
E55CRC-43 (Figure 105 and Figure 111 for transient behavior). E55CRC-41 did not
show a high count associated with acceleration, while the HHDDT showed this behavior
clearly. E55CRC-41 had a smaller displacement than the HHDDT, higher engine speed,
lower boost, two valves per cylinder (versus four for the HHDDT) and an in-line
injection system. However, it should be speculative to associate any of these factors
directly with the differing behavior.
..E+06
JJ2.E+06
S2.E+06
I
9i
-- 100
E+06
El.
«
0-
"eS
+j
o
H5.E+05 -
O.E+00 4
0
-100
200
400 600
Time (s)
800
1000
Figure 93: DMS500 Total Number of Particles and Vehicle Speed vs. Time during
UDDS for E55CRC-41 tested at 56,000 Ibs.
Figure 94 shows the distribution of the total number of particles in the 60 and 20 nm bins
versus time. The 60 nm bin contains particles of diameters between approximately 55 and
65 nm, and the diameters for the 20 nm bin range between 17.5 and 22.5 nm. There is a
sharp increase in the total number of particles in the 20 nm bin during decelerations and a
sharp increase in the total number of particles in the 60 nm bin during accelerations,
sharp meaning a greater percentage of decrease/increase.
110
-------
O.E+00
470 570 670 770 870 970
Time (s)
-20nm 60nm
Figure 94: DMS500 60 nm and 20 nm Particle Number vs. Time for E55CRC-41
tested at 56,000 Ibs.
Figure 95 shows the total number of particles and CO versus time. This plot shows the
correlation between the two series. This kind of a relation is actually expected due to the
fact that a large portion of particulate matter is elemental carbon (EC).
Figure 96 shows the total number of particles in the 60 nm bin versus the total number of
particles in the 20 nm bin. As can be seen, there is no clear correlation between the two
size bins, although the envelope of the points is well defined.
Figure 97 shows the total number of particles in the 20 nm bin versus the total hub power.
There is no strong correlation between the two series. Figure 98 shows the total number
of particles in the 60 nm bin versus the total hub power. Again, there is no strong
correlation, suggesting that the particle formation cannot be attributed to engine load
alone.
Ill
-------
6.E+06
5.E-02
O.E+00
370
470
570
670 770
Time (s)
870
970
•Total Particle Number (N/cc) CO (g/s)
Figure 95: DMS500 Total Number of Particles and CO vs. Time for E55CRC-41
tested at 56,000 Ibs.
I
l.E+05
l.E+05
8.E+04
6.E+04
4.E+04
2.E+04
O.E+00
O.E+00
5.E+04 l.E+05 2.E+05
20nm (N/cc)
2.E+05
3.E+05
Figure 96: DMS500 60 nm Particle Number vs. 20 nm Particle Number for
E55CRC-41 tested at 56,000 Ibs.
112
-------
3.EI05
u
u
-350
50 100 150 200
Total Hub Power (hp)
Figure 97: DMS500 20 nm Particle Number vs. Total Hub Power for E55CRC-41
tested at 56,000 Ibs.
1.EI05
-350 -300
200
-4.EI04
Total Hub Power (hp)
Figure 98: DMS500 60 nm Particle Number vs. Total Hub Power for E55CRC-41
tested at 56,000 Ibs.
113
-------
Figure 99 shows the size distribution of particles during acceleration and deceleration
periods. During acceleration, a single mode distribution is observed where the peak is
centered at around 100 nm. This is expected since during accelerations, accumulation
mode particles are formed. For the deceleration region, a single mode distribution is seen.
The peak is centered around 20 nm, and these are the nuclei mode particles. This mode is
observed during deceleration periods, as the plot also shows.
3.E+04
g 2.E+04
g,
!/)
"3 2.E+04
*-C
'o l.E+04
£
S
Z 5.E+03
O.E+00
•Acceleration
• Deceleration
100 200 300 400 500 600
Diameter (nm)
700
800
900 1000
Figure 99: DMS500 size distribution of particles during acceleration and
deceleration for E55CRC-41 tested at 56,000 Ibs.
E55CRC-42
E55CRC-42 was a powered by a Caterpillar 3406, MY 1999 engine. Particle sizing and
chemical speciation results are presented below for various modes of the HHDDT
schedule. Figure 100 and Figure 101 present the particle size distribution for Idle and
steady modes.
A typical uni-modal size distribution was noted for the Idle mode for E55CRC-42 (Figure
101) with a CMD of 40 nm and concentration of 1.3xl05 particles/cc. In comparison to
E55CRC-39 and E55CRC-40, the size distribution shifted towards a larger particle size
for E55CRC-42 during the Idle Mode. In addition to the nuclei mode, the size distribution
is also characterized by a mode with a CMD of 100 nm.
114
-------
1.40E+05
1.00E+01
10
100
Particle Diameter (nm)
1000
Figure 100: SMPS Particle Size Distribution for E55CRC-42 Operating on an Idle
Mode
1.20E+06
1.00E+01
10
100
Particle Diameter (nm)
2.00E+05
-- 1.80E+05
(O
<
1.60E+05 I
1.40E+05 ra
5
-- 1.20E+05 1
c
~ 0)
1 .OOE+05 1 5
§0
+ 8.00E+04 §
Z
- 6.00E+04 |
z
.2!
+ 4.00E+04 ^
£
-- 2.00E+04
1.00E+01
1000
Figure 101: SMPS Particle Size distribution for E55CRC-42 operating on various
steady cycles
115
-------
The accumulation mode observed during the Idle could be due to an increase in elemental
carbon emission when compared to the previous vehicles tested. During the Idle Mode,
the elemental carbon emitted by E55CRC-39 and E55CRC-40 were 0.21 g/hour and 0.07
g/hour, respectively and for E55CRC-42, the values were higher at 0.55 g/hour (Figure
104). This increase in elemental carbon also increased the mass emission for the Idle
mode at 1.34 g/hour (Figure 102). Mass emissions during Idle for E55CRC-39 and 40
were 0.35 g/hour and 0.14 g/hour, respectively. Total volatile organic compounds emitted
by E55CRC-42 during the Idle mode at 2.99 g/hour were much lower when compared
with the Cruise mode at 6.04 g/hour (Appendix O: Figure O25 to O29).
18
16
14
12
5 10
o
.c
"5) s
6
4
Idle
CruiseS
HHDDT S
Figure 102: PM Mass Emissions for E55CRC-42
Particle sizes measured during the steady state and Cruise mode were similar with a
CMD of 93.6 nm, 98.39 nm and 93.71 nm for the steady state, normal Cruise and High-
speed Cruise Modes, respectively. Concentrations for the steady state mode were 6 times
higher in magnitude as compared with both the Cruise Modes. Concentration drop for the
Cruise mode could be due to a higher exhaust temperature. Sulfate emissions were higher
during the Cruise mode at 0.54 g/hour when compared to Idle mode at 0.06 g/hour
(Figure 104). Carbonaceous particles along with sulfates, and adsorbed of volatile and
semi-volatile particles mostly constitute the nature of the particles emitted during the
steady cycles. The total carbon in the idle mode PM is approximately evenly split
between organic and elemental carbon. While the Idle Mode emissions are characterized
by a distinct nuclei mode, the contribution of this fraction to the total PM mass emission
rate is very small.
116
-------
Figure 103: Ion Composite Results for E55CRC-42
Figure 104: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-42
117
-------
For E55CRC-42, DMS500 data on the HDDS, Cruise mode and HHDDT_S showed
anomalous behavior (output spikes) but two repeat runs on the Transient mode agreed
acceptably well, as shown in Figure 105.
..E+06
O.E+00
-100
100
200
300 400
Time (s)
500
600
•TRANS3 (1) TRANS3 (2) Vehicle Speed
Figure 105: DMS500 Total Number of Particles and Vehicle Speed vs. Time on
Transient mode for E55CRC-42 tested at 56,000 Ibs.
E55CRC-43
E55CRC-43 was a powered by a Detroit Diesel Series 60, MY 1994 engine. Particle
sizing and chemical speciation results are presented below for various modes of the
HHDDT schedule. Figure 106 and Figure 107 present the particle size distribution for
Idle and steady modes.
Particle size distribution for Idle Mode was unimodal with a CMD of 51.43 nm and a
GSD of 1.44 which indicates a good distribution about the mean. Maximum
concentration level was around 2.25xl05 particles/cm3. These particles are mostly
accumulation mode particles as observed for E55CRC-42. Reasons for this accumulation
mode are similar to those presented above for E55CRC-42.
118
-------
2.50E+05
E
o
2.00E+05
f
1.50E+05
o
*j
ro
o
O
<5
A
E
1.00E+05
5.00E+04
ro
o.
1.00E+01
10
100
Particle Diameter (nm)
1000
Figure 106: SMPS Particle Size Distribution for E55CRC-43 Operating on an Idle
Mode
1.80E+05
•o
"S 't/i
o
O
4.00E+06
-- 3.50E+06
-- 3.00E+06 *
-- 2.50E+06 | £
Ji
-- 2.00E+06 1 *
ss
u w
-- 1.50E+06
-- 1.00E+06 01
u
2.00E+04
i- 5.00E+05
1.00E+01
O.OOE+00
10
100
Particle Diameter (nm)
1000
Figure 107: SMPS Particle Size Distribution for E55CRC-43 Operating on Various
Steady Cycles
119
-------
The total volatile organic compounds emitted by this particular vehicle were 2.28 g/hour
for Idle and 3.39 g/hour for Cruise. Volatile compound emissions at Idle were 20% lower
than E55CRC-42. A shift in the particle size was noted for the steady cycles. For the
steady state and High-speed Cruise Modes, the GSD were 1.88 and 2.6, respectively,
which is higher than the normal GSD which denotes a random distribution about the
mean which can be clearly observed from Figure 107. The CMD noted for the steady
state, normal Cruise and High-speed Cruise Modes were 54.32, 68.5 and 69.56 nm,
respectively. It should be noted that in the case of a HHDDT_S or Cruise Mode scan,
there is no assurance that the driver might not move the pedal and disturb the steady
operation during SMPS scan. Elemental carbon emissions were very high during the
Cruise Mode at 3.17 g/hour when compared to Idle Mode emissions at 0.25 g/hour.
Organic carbon emissions were also higher during the Cruise mode at 6.15 g/hour when
compared to Idle Mode emissions at 1.03 g/hour.
o
Idle
CruiseS
HHDDT S
Figure 108: PM Mass Emissions for E55CRC-43
The increase in the particle size and concentration in the Cruise Modes is attributed to the
very high increase in the elemental carbon emissions. A slight increase by 12% in the
VOC emission was noted for the Cruise mode (Appendix O: Figure O37 to O41). An
85% increase in the sulfate emission was noted for the Cruise mode (Figure 109).
Calcium was also a major contributor to the inorganic species with an increase of 91%
for the Cruise Mode. Calcium emissions are attributed to lubrication oil, since calcium is
a major constituent in lube oil. Particles measured during the Cruise mode of this
particular vehicle were primarily carbonaceous soot with heavy hydrocarbons adsorbed
onto the surface.
120
-------
Figure 109: Ion Composite Results for E55CRC-43
9
8
7 -
6
3
o c;
£ s
0)
4 -
3
2 -
1 -
0
• Elemental Carbon
DOrganic Carbon
Idle
Trans Cruise
CRC43
Figure 110: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-43
121
-------
DMS500 data were acquired for the Transient Mode, and are shown in Figure 111.
l.E+06
9.E+05
100
200
300 400
Time(s)
500
600
TRANS3 (1) Vehicle Speed
Figure 111: DMS500 Total Number concentration of Particles and Vehicle Speed vs.
Time for E55CRC-43 tested at 56,000 Ibs.
E55CRC-44
E55CRC-44 was a powered by a Caterpillar 3406, MY 1989 engine. Particle sizing and
chemical speciation results are presented below for various modes of the HHDDT
schedule. Figure 112 and Figure 113 present the particle size distribution for Idle and
steady modes.
A bimodal distribution was noted for this particular vehicle during the Idle Mode, with
one nano-particle peak at 26.9 nm and another accumulation mode peak at 122 nm. It
should be noted that the Idle Mode is characterized by relatively lower PM number
concentrations. The only other vehicle where a bimodal distribution was observed was
E55CRC-39 which was a 2004 year model engine. The reason for the bimodal
distribution for E55CRC-39 was due to high levels of volatile compounds and elemental
carbon during the Idle Mode. The elemental carbon contributed to the accumulation
mode.
122
-------
4.50E+04
o 4.00E+04
Q 3.50E+04
O)
3.00E+04
.2 2.50E+04
I
S 2.00E+04
^ 1.50E+04
01
&
I 1.00E+04
5.00E+03
O.OOE+OO
100
Particle Diameter (nm)
1000
Figure 112: SMPS Particle Size Distribution for E55CRC-44 Operating on an Idle
Mode
1.00E+05
9.00E+04
g- 8.00E+04
O)
I 7.00E+04
^T 6.00E+04
o
1 5.00E+04
O
fc
E
ro
0.
4.00E+04
3.00E+04
2.00E+04
1.00E+04
1.00E+01
10
100
Particle Diameter (nm)
1000
Figure 113: SMPS Particle Size Distribution for E55CRC-44 Operating on Various
Steady Cycles
123
-------
The precursors of the nano-particle emissions could be traced back to lube oil based
volatile hydro-carbons, and the lower exhaust temperature during exhaust mode. Again,
nano-particle peak could be due to high emissions of semi-volatile compounds (Appendix
O: Figure O54 and O55). The total volatile compound emissions for this vehicle were
25% higher than E55CRC-43 and 35% higher than E55CRC-40. Further analysis yields
that aromatic emissions during the Idle Mode, especially benzene were prominent for
E55CRC-39 and E55CRC-44 when compared with the rest of the vehicles in the study
(Appendix O: Figure O53). The accumulation mode in the Idle was attributed to the
presence of carbonaceous agglomerates, with elemental carbon emission during the Idle
mode being the highest among all the vehicles at 1.22 g/hour.
30
25
20
15
10
Idle
CruiseS
HHDDT S
Figure 114: PM Mass Emissions for E55CRC-44
The steady state particle size distributions were compatible with vehicle speed. Particle
concentrations were increasing, though minimally, with increasing speed without much
variation in the particle size. The count median diameter for the steady state, normal
Cruise and High-speed Cruise were 173, 166 and 178 nm, respectively. Mass emissions
were highest during the Cruise and High-speed Cruise Modes when compared to other
vehicles (Figure 114). Elemental carbon which is basically soot was noted to be the
highest among all the vehicles during the Cruise mode (Figure 116) was also a major
contributor to the mass emission. This would also contribute to low saturation ratios;
hence, the absence of nano-particles. Sulfate emissions were among the lowest when
compared with all other vehicles (Figure 115). Results from these data suggest that the
fine particles measured during the Cruise mode were mostly solid carbonaceous particles.
124
-------
0.30
0.25
0.00
Figure 115: Ion Composite Results for E55CRC-44
90
80
60
10
• Elemental Carbon
DOrganic Carbon
Idle
Trans
CRC44
Cruise
Figure 116: Elemental Carbon and Organic Carbon PM Analysis for E55CRC-44
No DMS500 data are available for E55CRC-44.
125
-------
CONCLUSIONS, NONREGULATED SPECIES
Size distributions of PM from the six trucks were selected for non-regulated emissions
measurement. These were acquired using a Scanning Mobility Particle Sizer (SMPS), as
well as a Differential Mobility Spectrometer, Model 500 (DMS500), which was a newly
released instrument. The SMPS was fed dilute exhaust from a mini-dilution tunnel, using
a dilution ratio of thirty, while the DMS used the main dilution tunnel of the TransLab.
The SMPS detected a bimodal distribution with a nuclei mode for Idle operation of one
2004 MY truck (E55CRC-40), and the DMS 500 detected both the nuclei mode and
accumulation mode during deceleration on this truck. The nuclei mode was not evident
under load for E55CRC-40. The other 2004 MY truck had only one mode, with a very
low particle count on Idle. A 1989 MY truck had a bimodal Idle distribution, but the
remaining HHDDT were unimodal. Comprehensive data were acquired for steady
operation using the SMPS on all trucks, and the DMS500 acquired transient distributions
for four of the trucks.
Chemical speciation was performed on the six trucks, with exhaust from the TransLab
tunnel fed to a residence time chamber of the DRI. Data were acquired for methane and
volatile organic compounds using a canister and a field gas chromatograph. Semi-volatile
organic compounds were captured in PUF/XAD media and PM soluble fraction was
captured on TIGF filters, and these were extracted and analyzed at the DRI laboratory.
Carbonyls were captured using DNPH cartridges, nitrosamines in Thermosorb cartridges.
Ions and Elemental/Organic carbon (EC/OC) split were determined from quartz filters.
Results from the speciation data are legion, and examples include the fact that the EC/OC
split differed substantially on Idle between the two 2004 MY trucks equipped with EGR,
and that the ion and metal analyses varied widely between trucks.
Summarized below in a graphical form are the emissions of elemental carbon (as a
percentage of total), inorganic ionic species, lubrication oil-based elemental emissions,
and engine wear emissions in Figure 117 through Figure 120, respectively. Figure 117
shows that except for idle operation, the elemental carbon constitutes a significant portion
of total carbon emissions. Transient operation, in particular, yields in excess of 60% of
total carbon as elemental carbon. Total PM mass emissions, as discussed in the text, are
the highest for the Transient mode, followed by Cruise, and then Idle. Total PM mass
emissions under Idle are significantly lower than the Transient and Cruise modes.
However, in terms of total carbon mass emission rates, the Cruise mode emits the highest
levels, followed by the Transient mode of operation. The Idle mode is characterized by
significantly lower rates of total carbon mass emissions.
The mass emissions rates of sulfates dominate the ionic species in the cruise and transient
modes. The only other species of significance were nitrates and ammonium.
Total particulate matter comprises of elemental carbons, organic carbon, engine wear
elements, lube oil based ash components, water, and ionic species. Most of these
emissions are a result of un-burnt fuel and lubricating oil. Apart from severe engine wear
(abrasive wear) on boundary lubricated surfaces (cam lobe wear, tappet polishing,
126
-------
rocker/crosshead wear, ring wear at top and bottom reversal locations, etc.), oil control is
extremely critical from an emissions point of view. Based upon prior studies by Lapin et
al.19 and Gautam et al.14, it may be hypothesized that products of lubricating oil
combustion (partial) result in higher potential mutagenicity of exhaust emissions, and that
lubricating oil contributes to high concentrations of nanoparticles in exhaust emissions1.
Lubricating oil enters the combustion chamber through the following routes: as blow-by
gas, leakage past valve stem seals, piston rings and vaporisation from the cylinder liner
walls and combustion space20.
100%
90%
80%
u
2
o
€
re
O
5
c
0)
0)
LJ
70%
60%
50%
40%
30%
20%
10%
0%
nr
nK3
m
\
I
I
:
;
I
• Idle
m Transient
H Cruise
r
I
-
I
CRC39
CRC40
CRC41
CRC42
CRC43
CRC44
Figure 117: Elemental Carbon Emissions (Percentage of Total Carbon)
127
-------
2.50
2.00
1.50
O)
ta
ta
ta
£ 1.00
0.50
0.00
J .111
I
I
EINH4+
S SO4-2
ENO3-
• NO2-
DCI-
!
i i
^_ra H_l_^_B_fe_Bi 8 BB
O(OOO(OOO(OOO(OOO
i 5 -1 i 5 -1 i g -1 i 5 -1 i
^ ^ ^ ^ 1_ ^ ^ ^
>- 0 "~ " i Q " "~ "
CRC39 CRC40 CRC 41 CRC42 C
(0 O O (0 O
C <0 = C <0
5 Z - 5 3
" 0 " 0
RC 43 CRC 44
Figure 118: Inorganic Ionic Species Emissions
Figure 119 shows the mass emission rates of lubricating oil based elements from the test
engines. These elements contribute to the total PM mass. Again, sulfur and calcium are
the principal pollutants that can be traced back to the lubricating oil. In the older engines
(CRC-42 and CRC-43) phosphorus emissions were significantly higher than those from
other newer engines.
Figure 120 shows a summary of mass emissions rates of engine wear elements. Among
these elements, iron contributed the most to the total PM emissions under all modes of
operation. Of course, in terms of mass emission rates, iron emissions were the higher in
the Transient and Cruise Modes of operation.
128
-------
0.70
0.60 -
0.50
£ 0.40
"ft
in
I
n
.2) 0.30
UJ
0.20 H
0.10 -
0.00
DPb
• Mo
DCa
ns
• P
DMg
CRC39
CRC43
Figure 119: Lubrication Oil Emissions
129
-------
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
1— 1
HI
2
(
R
c
2
H
3RC3S
y
Cruise
n
HI
2
C
n
w
c
2
H
5RC4(
u
Cruise
n
MD Idle
n
I/)
c
2
H
Q
3RC4
DZn
• Cu
DFe
HI
w
I
O
Q
1
H
HI
2
(
:F
—
I/)
C
2
i-
tc
42
>
"
Cruise
1
HI
2
(
:F
=
I/)
C
2
i-
JC
4:
!
— 1
-
Cruise
Figure 120: Engine Wear Emissions
130
-------
TAMPERING, MALMAINTENANCE AND REFLASH VEHICLES
The E-55/59 program sought to identify high emitters, or vehicles that showed evidence
of T&M. Vehicles were identified as T&M candidates either from a pretest inspection or
from measurement of emissions that were high relative to developed threshold criteria.
The criteria for declaring NOX to be high were developed using NOX/CO2 ratios from
certification data, and the PM criteria were developed as PM/CC^ ratios considering
Phase 1 fleet data statistically. The program plan provided that high emitters could be
subjected to repair, followed by re-testing of emissions. In addition, during Phase 1.5 of
the program, three truck engine controllers were reflashed. These trucks had MY that
suggested that they may have off-cycle control strategies in the controller, and the effect
of reflashing to a lower NOX strategy was assessed.
Diesel vehicle deterioration and inspection and maintenance have been discussed
previously by Weaver and Klausmeier21 and Weaver et al.22, but few data exist either on
the incidence of high emitters or the effect of repair on truck emissions. The website
http://www.driveclean.com/downloads/2003 HDV Report.pdf provides modeling
information as part of the Ontario, Canada, Drive Clean program. Yanowitz and co-
workersl'23 have presented the change of emissions with truck MY, but this cannot be
uniquely ascribed to deterioration of emissions.
T&M Criteria
An objective of the E55/59 program was to quantify the influence of T&M on heavy-duty
vehicle emissions. This was achieved by identifying test vehicles that qualified as
candidates for the study, using two separate pathways:
1. Vehicle Inspection: Vehicles were considered candidates if they showed evidence
of T&M that may affect emissions. The evidence might be found at the time of
vehicle inspection, or revealed during the test procedure. A vehicle might qualify
for repair and retest as a result of inspection, but might prove to yield emissions at
a level below that considered high or gross.
2. High Emissions Levels: Vehicles were considered as candidates if they showed
emissions levels that were high, or gross, relative to expected levels for those
vehicles. It was not required that visual T&M was evident for these high emitting
vehicles.
Determination of Candidates from Vehicle Inspection
The vehicles from the program in Phase 1 of E-55/59 were recruited either directly by
WVU or with the aid of the California Trucking Association. In subsequent phases the
vehicles were recruited by WVU alone. Vehicles were transported to the test site in
Riverside, C A. Each vehicle was inspected to identify any of the following tampering or
malmaintenance items.
131
-------
1. Cleanliness of air filter.
2. Integrity of lines between turbocharger an intercooler.
3. Integrity of lines between intercooler and engine intake.
4. Presence of manifold air pressure sensor (MAP) and integrity of wiring
connection to MAP (This applies to most electronic engines).
5. Presence of manifold air temperature sensor (MAT) and integrity of wiring
connection to MAT (This applies to most electronic engines).
6. Integrity of wiring harness and connections to engine controller (This applies to
all electronic engines where the controller is visible).
7. Integrity of wastegate/boost control on turbocharger (This applies to many newer
electronic engines).
8. Condition of intercooler, though in many cases the core will not be visible
because it is sandwiched between the air conditioner and radiator.
9. Integrity of linkages and fuel lines surrounding fuel injection pump, or breakage
of seal on fuel injection pump (This applies to older mechanical engines).
10. Correctness of fuel (not dyed off-road fuel) in tank, where practical.
If inspection identified a technical or maintenance concern that might affect emissions,
the vehicle immediately became a "T&M candidate," which could potentially be
subjected to repair and retest.
Determination of T&M Candidate from Emissions
In the study, vehicles were potential candidates for repair and retest if they were
identified as high emitters. The term "high emitters" was intended to represent vehicles
emitting PM or NOX at gross levels; that is, levels, which are substantially higher than
would be expected from a statistical sample of similar, correctly operating vehicles. In E-
55/59 the criteria for high emissions were developed during Phase 1, and were applied in
this form to all phases of the program.
High HHDDT emitters of NOX were identified in the following fashion. NOX emissions
were known to be dependent on engine power (and hence fuel consumed), but relatively
insensitive to the transient nature of tests unless the engine control included off-cycle
strategies. In many cases NOX varies in a linear fashion with brake power and in direct
proportion to fuel consumed24'25. It is therefore reasonable to compare NOX/CO2 ratios for
different tests, although the investigators were aware of the way in which off-cycle
engine timing could elevate NOX levels. Cycle effects on NOX emissions have recently
been discussed by Clark et al.26 and are addressed in a paper by Yanowitz et al.1'
X
Table 6 provides the U.S. certification levels for engines. The "2003+" value for NO
recognizes that the 2004 regulations were brought forward from 2004 for the
manufacturer according to the Consent Decree, and approximates the NOX level at 2.0
g/bhp-hr. Ratios by mass of NOX/CO2 implied by the certification can be determined if a
132
-------
typical Federal Test Procedure certification level of 550 g/bhp-hr (738 g/kw-hr) of CC>2 is
assumed. When a normally operating engine is exercised through cycles other than the
certification cycle as in the case of the E-55/59 program, NOX/CO2 ratios may deviate
from the certification NOX/CO2 ratios for two reasons. Firstly, the engine timing is not
uniform over the whole operating envelope, and may differ slightly between the
certification envelope and remainder of the operation envelope. This may cause modest
increases (or decreases) in the NOX/CO2 ratio. This is likely to have a greater effect in
electronic engines or in mechanical engines without variable injection timing. Secondly,
many electronic engines have off-cycle NOX missions as a result of employing a separate
timing map during some instances of real-world operation. This is known to raise NOX by
a factor of up to 2.527, corresponding to 10 to 12 g/bhp-hr (13 to 16 g/kw-hr) NOX for
engines certified to 5 g/bhp-hr (6.7 g/kw-hr), even at high load conditions. Engines
operating off-cycle are not considered to fit the T&M criteria because they are in original
condition and need not be reconfigured except at the time of rebuild, under rules
prevailing during the time of the E-55/59 program. Off-cycle emissions of this kind occur
solely with electronic engines.
In the E-55/59 program, emissions values used to determine whether a vehicle was a high
NOX or PM emitter were taken from the transient and cruise modes of the HUDDT
schedule. This schedule was created and evaluated by the California Air Resources Board
and WVU in a previous cooperative effort28'29. There were at least two repeat runs of the
transient and cruise modes for each vehicle. Data were acquired at a test weight of 56,000
Ibs. (25,402 kg). A vehicle (1) had to show high NOX in two transient runs to be
considered a high NOX emitter on the transient mode, or (2) had to show high NOX in two
cruise runs to be considered a high NOX emitter on the cruise mode. Similarly, the vehicle
(3) had to emit high PM in two transient runs conducted to be considered a high PM
emitter on the Transient or (4) produce high PM in two cruise runs conducted to be
considered a high PM emitter on the transient mode. The vehicle was considered a high
emitter and as a candidate for repair and retest under any one of these four circumstances.
In order to avoid capturing "off-cycle" emissions vehicles as T&M vehicles, the
NOX/CO2 ratio for identifying a candidate vehicle was set at three times the certification
NOX/CO2 ratio for electronic engines. For mechanically injected engines, the criterion
NOX/CO2 ratio was set at twice the certification NOX/CO2 ratio.
Table 19 shows the resulting emissions level criteria used to declare a truck to be a high
NOX emitter. Table 19 uses California certification values and vehicle MY as a basis.
In cases where vehicles pre-dated the advent of brake-specific emissions standards for
values of NOX emissions, a value of lOg/bhp-hr (13.4 g/kw-hr) was used as if it were a
certification standard in identifying high NOX emitters. Early vehicles all have
mechanically controlled engines, so that the failure criteria shown below were based on a
NOX/CO2 ratio for these early vehicles using twice this assumed level; i.e., 20 g/bhp-hr
(26.8 g/kw-hr).
133
-------
Table 19: Emissions level criteria used to declare a truck to be a high NOX emitter.
California
1987-90
1991-93
1994+
1994-95
1996+
NOX
g/bhp-hr
(g/kw-hr)
6(8)
5 (6.7)
5 (6.7)
5 (6.7)
4 (5.4)
C02
g/bhp-hr
(g/kw-hr)
550 (738)
550 (738)
550 (738)
550 (738)
550 (738)
NOX / CO2
Certification
Ratio
0.0109
0.0091
0.0091
0.0091
0.0073
Mechanical
High Emitter
Criterion
(Cert x 2)
NOX/CO2
0.0218
0.0182
0.0182
0.0182
0.0145
Electronic
High Emitter
Criterion
(Cert x 3)
NOx/CO2
0.0327
0.0273
0.0273
0.0273
0.0218
No detailed criteria were developed in Phases 2 and 3 for the MHDT: the researchers
used their own judgment in reviewing the NOX emissions data to determine if a truck
should be considered as an MHDT T&M candidate. One MHDT, CRC-57, revealed a
cooling malfunction on the dynamometer and was the only MHDT that was identified as
a T&M candidate. CRC-57 was not repaired, and is discussed in a separate section below.
Determining high HHDDT PM emitters was more difficult than determining high NOX
emitters because PM is highly sensitive to the transient content of a cycle and therefore
cannot be projected for different cycles using certification data26'30'31. PM emissions drop
only very slightly as a result of off-cycle operation, so off-cycle operation need not be
considered with regard to PM. To determine PM high emitters, the data for each of the 25
trucks (all HHDDT) in Phase 1 of the program in the transient and cruise modes were
expressed as ratios of PM/CO2. This ratio was then divided by the PM/CO2 ratio for the
certification standards for that engine MY, assuming a level of 550g/bhp-hr (738 g/kw-
hr) of CC>2 for certification. This "criterion ratio" calculation is shown below. The
criterion ratio for vehicles that pre-date the PM transient test (FTP) standards in their MY
was based on a PM pseudo-standard of Ig/bhp-hr (1.3 g/kw-hr) PM.
Consider an example using actual data from a vehicle with an engine manufactured in
1999 or 2000 during the Transient mode of the HHDDT schedule:
Actual data for Transient Mode
PM = 0.4435 g/mile (0.2756 g/km)
CO2= 2599 g/mile (1615 g/km)
(PM/CO2)TranS = 0.000171
134
-------
1999 - 2000 Certification
PM = 0.1 g/bhp-hr (0.134 g/kw-hr)
CO2 = 550 g/bhp-hr (738 g/kw-hr)
(PM/CO2)cert = 0.000182
Criterion Ratio = (PM/C°2)Trans =0.939
(PM/C02)Cert
For a vehicle to be identified as a high PM emitter, this criterion ratio described above
had to be significantly higher than the criterion ratios for typical vehicles in the fleet. In
this case the term "typical vehicles" refers to both normal and high emitters. Note that the
inclusion of certification levels in the criterion ratio does allow for influence of changing
technology in identifying high PM emitters. To achieve this identification goal, it was
necessary to build a database of PM criterion ratios from the initial vehicles that were
tested, and determine the average and standard deviation of those criterion ratios. Any
truck that was two standard deviations above the average on two runs of the Transient
mode was deemed a high PM emitter for the transient mode. Any truck that was two
standard deviations above the average on two runs of the Cruise mode was deemed a high
PM emitter for the cruise mode. Since the database of vehicles tested evolved over the
duration of the program, it was possible that the statistics would change over time, but no
remedy for this change was possible. It was noted that in this methodology vehicles with
high emissions on the Transient mode but not on the Cruise mode would most likely be
"puffers," caused by turbocharger damage or fuelling limit tampering, while those that
were found to be high using the cruise criterion (whether they are high on the transient or
not) were likely to be continuous smokers at high load. The statistical criteria for high
PM emitters were finalized in Phase 1 and used to identify for high emitters of PM in
subsequent phases.
No detailed PM high emitter criteria were developed in Phases 2 and 3 for the MHDT:
the researchers used their own judgment in reviewing the PM emissions data to determine
if a truck should be considered as an MHDT T&M candidate.
Once a vehicle was identified as a high NOX or PM emitter, it was treated in the same
way as a vehicle that was found to have evidence of T&M upon inspection.
Results: T&M by Inspection
All vehicles were subjected to both inspection and emissions characterization. One
vehicle, E55CRC-21, was found to have a dirty air filter by inspection, and became a
T&M candidate. The repair and retest data of E55CRC-21 are presented in a separate
section below. Two other vehicles, E55CRC-16 and E55CRC-45, were visually identified
as prodigious smokers. E55CRC-16 subsequently was found to have emissions
management components removed, but it was treated in the program as having high PM
emissions by measurement. No other vehicles were considered as T&M vehicles by
inspection. However, in Phase 3 one HHDDT brought to the site was visibly in poor
repair, ceased running after arrival, and was not tested.
135
-------
Verifying the Criteria in Phase 1: NOx Analysis
Figure 121 shows the NOX/CO2 ratios for each run on the twenty-five Phase 1 vehicles,
for both Transient and Cruise Modes. Vehicle E55CRC-10 was the only vehicle to be
identified as a "high" NOX emitter according to the T&M plan.
0.020
0.018 -
fS
O
u
~s
o
5,
(U
•a
o
.«
'
-------
(5.36 g/kw-hr). The investigators believed that the high NOX was simply an original off-
cycle phenomenon, and would therefore not be remedied by any repair. This vehicle was
therefore not selected as a T&M vehicle, but was selected for reflash of its controller and
termed a Reflash Vehicle.
The finding that the high NOX emissions of E55CRC-10 were not due to a T&M cause
suggested that it would be difficult to identify high NOX emitters as T&M candidates if
they had electronic engine controllers that caused off-cycle emissions. In Phase 1.5 of the
program, rather than seeking high NOX emitters by the criterion, three trucks with
electronic engine control were selected directly for reflash and retest. The results for
these three trucks, which coincidentally included a T&M vehicle (E55CRC-28), are
treated in a separate section below.
1.4
*V V » ». •
*' ^ t *• - *
50 100 150 200 250 300
Dispersed and Time Aligned Axle Horsepower (ahp)
350
Figure 122: NOX emissions plotted against the dispersed and time aligned power for
vehicle E55CRC-10. Most emissions during the Cruise mode were at an off-cycle
level. The vehicle was targeted for Reflash as a result of these data.
Verifying the Criteria in Phase 1: PMAnalysis
Figure 123 presents the average value of PM/CO2 ratios for Cruise mode operation of
each Phase 1 HHDDT, and Figure 124 presents the average value of PM/CO2 ratios for
the Transient Mode operation of each vehicle. These values were then converted to
criterion ratios according to the T&M methodology described above. It was found that
the average criterion ratio for the Transient Mode was 1.2736 with a standard deviation
of 1.1486. It was found that the average criterion ratio for the cruise cycle was 0.5207
with a standard deviation of 0.3567. These data suggest that if trucks were as "clean" as
137
-------
when they were certified using the FTP, the Transient Cycle elicited higher brake-
specific PM than the FTP, while the cruise cycle exhibited lower PM than the FTP. This
is to be expected noting the relative transient content of the Transient and Cruise
Modes29'33 and the fact that transient behavior usually elicits higher PM production.
§
i
n nno
n nns
n nn^,
n nn^
n nn/i
n nn'j
n nno
n nm
n nnn
__n^,,n.n__ nnH
-r,n^w«t~ooovo-r,n^w«
nnnnnnnnn
5i>xo\O'-Hnt^^i>ft
UUUUUUUUU
uuuuuuuuuu
&&'&'&'&'&'&'&'&'&
uu
'&'&
uu
'&'&
Figure 123: PM/COi ratios for Cruise Mode operation of average value for each
vehicle. There are two runs each for E55CRC-1 to -13, and three runs each for
E55CRC-14 to -25.
138
-------
I
N
8
jj
0.005
0.004
0.003
0.002
0.001
n n
n nil n linn n
in
H
in
H
in
H
Figure 124: PM/CO2 ratios for the Transient Mode operation of average value for
each vehicle. There are 2 runs each for E55CRC-1 to -13, 3 runs each for E55CRC-
14 to -25. One run for E55CRC-20 was anomalous low, but no cause was evident.
less than
1 stdev
-1 to -0.5
stdev
-0.5 stdev
to avg
avg to 0.5
stdev
0.5 to 1
stdev
1 to 1.5
stdev
1.5 to 2
stdev
greater
than 2
stdev
I Occurences
21
19
7
1
1
Figure 125 shows the occurrence of criterion ratios by bin for the Transient Mode, where
the bins are described in
139
-------
Table 20. The resulting PM/CC^ ratios for each transient run identify E55CRC-6 with
high emissions on one run of the transient test.
Figure 126 shows the occurrence of criterion ratios by bin for the Cruise Mode, where the
bins are described in Table 21. Only one run for vehicle E55CRC-6, for the transient
case, fell outside of the two standard deviation limits so that this run represented high
emissions. The other run for the Transient Mode case for E55CRC-6 lay between 1.5 and
2 standard deviations above the average and so was not statistically high at the 95% level.
Therefore, since both runs did not indicate high emissions, E55CRC-6 was not designated
as a high emitter based on population statistics. E55CRC-8 and E55CRC-15 were
similarly behaved.
less than •
1 stdev
-1 to -0.5
stdev
-0.5 stdev
to avg
avg to 0.5
stdev
0.5 to 1
stdev
1 to 1.5
stdev
1.5 to 2
stdev
greater
than 2
stdev
D Occurences
21
19
1
1
Figure 125: Occurrence of criterion ratios by bin for the Laden Transient Mode
(PM/COi). E55CRC-16 was not included in this analysis, because it was an extreme
outlier.
Table 20: Definition of criterion ratios by bin for the Laden Transient Mode: see
Figure 125.
140
-------
Bin description
less than -1 stdev
-0.5 to -1 stdev
-0.5 stdev to avg
avg to 0.5 stdev
0.5 to 1 stdev
1 to 1.5 stdev
1.5 to 2 stdev
greater than 2 stdev
Bin Range
less than 0.125
0.125 to 0.699
0.699 to 1.274
1.274 to 1.848
1.848 to 2.422
2.422 to 2.997
2.997 to 3.571
greater than 3. 571
less than •
1 stdev
-1 to -0.5
stdev
-0.5 stdev
to avg
avg to 0.5
stdev
E55CRC-9 high both tests
E55CRC-15 high on one test
E55CRC-20 high on one test
0.5 to 1
stdev
1 to 1.5
stdev
1.5 to 2
stdev
greater
than 2
stdev
D Occurences
3
23
11
12
Figure 126: Occurrence of criterion ratios by bin for the Laden Cruise mode
(PM/COi). E55CRC-16 was not included in this analysis because it was an extreme
outliner.
Table 21: Definition of criterion by bin for the Laden Cruise Mode: See Figure 126.
141
-------
Bin Description
less than -1 stdev
-0.5 to -1 stdev
-0.5 stdev to avg
avg to 0.5 stdev
0.5 to 1 stdev
1 to 1.5 stdev
1.5 to 2 stdev
greater than 2 stdev
Bin Range
less than 0.164
0.164 to 0.342
0.343 to 0.521
0.521 to 0.699
0.699 to 0.878
0.878 to 1.056
1.056 to 1.234
greater than 1.234
The resulting PM/CC^ ratios for each cruise run identified E55CRC-9 as having high
emissions on two cruise runs. E55CRC-9 was therefore designated as a high emitter and
was a potential T&M candidate. The cause of the high PM emissions was not obvious,
but may have been due to engine wear. Repairs in this case would be costly. There were
several potential candidate T&M vehicles that had arisen at the time the E55CRC-9 was
discussed by the researchers, and insufficient support to address them all. A sponsor
decision was not to pursue E55CRC-9 for T&M repair and retest. E55CRC-20 showed
high PM emissions on only one Cruise mode run and was therefore not designated as a
high emitter.
The PM emissions of E55CRC-16 were prodigious and it exhibited high PM emissions
by all criteria on both Transient and Cruise Modes. It was selected as a T&M vehicle, and
is discussed in a separate subsection below. Because the values were so high that they
would skew the results, the E55CRC-16 results were not considered when calculating the
"expected'V'high" statistical criteria.
It is of interest to note that E55CRC-16 was most likely puffing because its Transient
Mode PM emissions were high but Cruise Mode PM emissions were modest in contrast
to E55CRC-9, which showed the opposite trend. Data also confirmed the visual
observation of high PM output from vehicle E55CRC-16. E55CRC-16 was designated as
a T&M candidate and is discussed in a separate subsection below. E55CRC-21 became a
T&M vehicle by inspection of its dirty air filter, but was not a high PM emitter by test
data criteria.
The criteria for PM high emitters was finalized in Phase 1, and applied to HHDDT in
subsequent phases. After Phase 1, some HHDDT in the program were only subjected to
one run of each mode: the criterion that the vehicle must have two high values in the
Transient mode or two high values in the Cruise mode could not be applied to those
vehicles. It revealed in Phase 1.5 that E55CRC-28 was a high PM emitter in Phase 1.5
and that E55CRC-45 was a high PM emitter in Phase 2.
Criteria ratios for all phases
Show the T&M criteria for all of the HHDDT vehicles examined in this study where
"high" NOX and PM emitters, as identified by the previously discussed methodology, are
highlighted in these tables. Only initial test data, not including any re-test data, were used
142
-------
in determining the average and standard deviations in the paniculate matter calculations.
CRC-16, a gross PM emitter, was not included in the statistical calculations either.
143
-------
Table 22: T&M Determination Data for the HHDDT Vehicles over the Cruise Mode
Vehicle
E55CRC-1
E55CRC-1
E55CRC-2
E55CRC-2
E55CRC-3
E55CRC-3
E55CRC-4
E55CRC-4
E55CRC-5
E55CRC-5
E55CRC-6
E55CRC-6
E55CRC-7
E55CRC-7
E55CRC-8
E55CRC-8
E55CRC-9
E55CRC-9
E55CRC-10
E55CRC-10
E55CRC-11
E55CRC-11
E55CRC-12
E55CRC-12
E55CRC-13
E55CRC-13
E55CRC-14
Model
Year
1994
1994
1995
1995
1985
1985
1999
1999
2000
2000
1995
1995
1990
1990
1996
1996
1998
1998
1998
1998
2000
2000
1986
1986
1978
1978
1985
(NOX/C02)
Actual
0.0184
0.0204
0.0129
0.0132
0.0057
0.0050
0.0107
0.0114
0.0120
0.0147
0.0188
0.0158
0.0107
0.0118
0.0177
0.0210
0.0155
0.0157
0.0329
0.0305
0.0107
0.0109
0.0087
0.0078
0.0134
0.0132
0.0095
(NOX/C02)
Certification
0.0091
0.0091
0.0091
0.0091
0.0093
0.0093
0.0073
0.0073
0.0073
0.0073
0.0091
0.0091
0.0109
0.0109
0.0091
0.0091
0.0073
0.0073
0.0073
0.0073
0.0073
0.0093
0.0093
0.0093
0.0093
0.0093
0.0093
Multiplier
o
J
3
o
J
3
2
2
o
J
3
o
J
3
3
o
J
3
o
J
3
3
o
J
3
o
J
3
3
2
2
2
2
2
2
Actual /
(Cert*
Multiplier)
0.68
0.75
0.47
0.48
0.31
0.27
0.49
0.52
0.55
0.67
0.69
0.58
0.33
0.36
0.65
0.77
0.71
0.72
1.51
1.40
0.49
0.59
0.47
0.42
0.72
0.71
0.51
(PM/CO2)
Actual
0.000075
0.000081
0.000099
0.000093
0.000434
0.000419
0.000131
0.000150
0.000149
0.000146
0.000172
0.000181
0.000272
0.000254
0.000144
0.000148
0.000278
0.000262
0.000081
0.000096
0.000111
0.000092
0.001187
0.000843
0.000455
0.000426
0.000390
(PM/CO2)
Certification
0.00018
0.00018
0.00018
0.00018
0.00182
0.00182
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00109
0.00109
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00182
0.00182
0.00182
0.00182
0.00182
0.00182
Actual /
Certification
0.4128
0.4441
0.5470
0.5124
0.2389
0.2302
0.7191
0.8250
0.8213
0.8048
0.9452
0.9951
0.2497
0.2332
0.7916
0.8128
1.5302
1.4402
0.4476
0.5306
0.6101
0.0504
0.6527
0.4638
0.2501
0.2345
0.2144
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.13
0.14
0.17
0.16
0.07
0.07
0.22
0.25
0.25
0.25
0.29
0.31
0.08
0.07
0.24
0.25
0.47
0.44
0.14
0.16
0.19
0.02
0.20
0.14
0.08
0.07
0.07
144
-------
Vehicle
E55CRC-14
E55CRC-14
E55CRC-15
E55CRC-15
E55CRC-15
E55CRC-16
E55CRC-16
E55CRC-16
E55CRC-17
E55CRC-17
E55CRC-17
E55CRC-18
E55CRC-18
E55CRC-18
E55CRC-19
E55CRC-19
E55CRC-19
E55CRC-20
E55CRC-20
E55CRC-20
E55CRC-21
E55CRC-21
E55CRC-21
E55CRC-22
E55CRC-22
E55CRC-22
E55CRC-23
E55CRC-23
E55CRC-23
Model
Year
1985
1985
1986
1986
1986
1979
1979
1979
1993
1993
1993
1991
1991
1991
1987
1987
1987
1992
1992
1992
1990
1990
1990
1993
1993
1993
1983
1983
1983
(NOX/C02)
Actual
0.0096
0.0096
0.0091
0.0090
0.0090
0.0056
0.0053
0.0058
0.0120
0.0113
0.0115
0.0085
0.0085
0.0088
0.0073
0.0075
0.0075
0.0141
0.0142
0.0135
0.0116
0.0115
0.0118
0.0109
0.0116
0.0114
0.0129
0.0134
0.0134
(NOX/C02)
Certification
0.0093
0.0093
0.0093
0.0093
0.0091
0.0091
0.0091
0.0091
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0091
0.0091
0.0091
0.0109
0.0109
0.0109
0.0091
0.0091
0.0091
0.0109
0.0109
0.0109
Multiplier
2
2
2
2
2
2
2
2
o
J
3
o
5
2
3
o
3
2
2
2
3
o
J
3
2
3
3
o
J
3
o
J
2
2
2
Actual /
(Cert*
Multiplier)
0.52
0.52
0.49
0.48
0.50
0.31
0.29
0.32
0.37
0.35
0.35
0.39
0.26
0.27
0.34
0.34
0.34
0.52
0.52
0.49
0.53
0.35
0.36
0.40
0.43
0.42
0.59
0.62
0.61
(PM/C02)
Actual
0.000376
0.000368
0.000810
0.000728
0.000778
0.009448
0.008619
0.008698
0.000365
0.000429
0.000310
0.000276
0.000263
0.000267
0.000603
0.000633
0.000577
0.000520
0.000534
0.000567
0.000643
0.000623
0.000540
0.000285
0.000268
0.000259
0.000545
0.000577
0.000581
(PM/C02)
Certification
0.00182
0.00182
0.00182
0.00182
0.00045
0.00045
0.00045
0.00045
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00045
0.00045
0.00045
0.00109
0.00109
0.00109
0.00045
0.00045
0.00045
0.00182
0.00182
0.00182
Actual /
Certification
0.2068
0.2022
0.4455
0.4002
1.7109
20.7851
18.9624
19.1364
0.3349
0.3930
0.2841
0.2529
0.2408
0.2449
0.5529
0.5801
0.5290
1.1443
1.1751
1.2475
0.5891
0.5714
0.4951
0.6262
0.5891
0.5697
0.2997
0.3171
0.3194
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.06
0.06
0.14
0.12
0.52
6.38
5.82
5.87
0.10
0.12
0.09
0.08
0.07
0.08
0.17
0.18
0.16
0.35
0.36
0.38
0.18
0.18
0.15
0.19
0.18
0.17
0.09
0.10
0.10
145
-------
Vehicle
E55CRC-24
E55CRC-24
E55CRC-24
E55CRC-25
E55CRC-25
E55CRC-25
E55CRC-26
E55CRC-27
E55CRC-28
E55CRC-29
E55CRC-30
E55CRC-31
E55CRC-32
E55CRC-33
E55CRC-34
E55CRC-35
E55CRC-36
E55CRC-38
E55CRC-38
E55CRC-38
E55CRC-39
E55CRC-39
E55CRC-40
E55CRC-40
E55CRC-42
E55CRC-42
E55CRC-43
E55CRC-43
E55CRC-44
Model
Year
1975
1975
1975
1983
1983
1983
1998
1999
1998
1999
1998
1997
1991
1984
2003
2000
2001
2004
2004
2004
2004
2004
2004
2004
2000
2000
1995
1995
1989
(NOX/C02)
Actual
0.0153
0.0172
0.0162
0.0138
0.0156
0.0156
0.0140
0.0118
0.0155
0.0097
0.0160
0.0178
0.0068
0.0174
0.0040
0.0099
0.0087
0.0049
0.0045
0.0049
0.0047
0.0044
0.0074
0.0082
0.0075
0.0077
0.0138
0.0123
0.0090
(NOX/C02)
Certification
0.0182
0.0182
0.0182
0.0109
0.0109
0.0109
0.0073
0.0073
0.0073
0.0073
0.0073
0.0091
0.0091
0.0093
0.0036
0.0073
0.0073
0.0036
0.0036
0.0036
0.0036
0.0036
0.0036
0.0036
0.0073
0.0073
0.0091
0.0091
0.0109
Multiplier
2
2
2
2
2
2
3
3
o
J
3
o
J
3
2
2
3
o
J
3
3
o
J
3
o
J
3
3
o
J
3
o
J
o
J
3
2
Actual /
(Cert*
Multiplier)
0.42
0.47
0.44
0.63
0.71
0.72
0.64
0.54
0.71
0.44
0.73
0.65
0.37
0.94
0.36
0.45
0.40
0.45
0.41
0.45
0.43
0.41
0.68
0.76
0.34
0.35
0.51
0.45
0.41
(PM/C02)
Actual
0.000360
0.000324
0.000295
0.000492
0.000464
0.000428
0.000219
0.000148
0.000377
0.001838
0.000148
0.000140
0.000355
0.000471
0.000829
0.000240
0.000211
0.000114
0.000108
0.000134
0.000104
0.000113
0.000107
0.000086
0.000173
0.000178
0.000044
0.000040
0.000227
(PM/C02)
Certification
0.00182
0.00182
0.00182
0.00182
0.00182
0.00182
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00045
0.00182
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00109
Actual /
Certification
0.1977
0.1784
0.1625
0.2704
0.2553
0.2355
1.2024
0.8125
2.0711
10.1100
0.8114
0.7707
0.7809
0.2593
4.5598
1.3176
1.1610
0.63
0.60
0.75
0.58
0.63
0.60
0.48
0.96
0.99
0.24
0.22
0.21
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.06
0.05
0.05
0.08
0.08
0.07
0.37
0.25
0.64
3.10
0.25
0.24
0.24
0.08
1.40
0.40
0.36
0.19
0.18
0.23
0.18
0.19
0.18
0.15
0.30
0.30
0.07
0.07
0.06
146
-------
Vehicle
E55CRC-44
E55CRC-45
E55CRC-46
E55CRC-47
E55CRC-48
E55CRC-49
E55CRC-60
E55CRC-62
E55CRC-63
E55CRC-64
E55CRC-65
E55CRC-66
E55CRC-69
E55CRC-74
E55CRC-78
Model
Year
1989
1993
1989
1986
1998
1994
1995
1983
2005
1994
1994
1989
1989
1969
1975
(NOX/C02)
Actual
0.0088
0.0060
0.0104
0.0056
0.0162
0.0080
0.0098
0.0090
0.0074
0.0154
0.0132
0.0109
0.0083
0.0135
0.0249
(NOX/C02)
Certification
0.0109
0.0091
0.0109
0.0093
0.0073
0.0091
0.0091
0.0109
0.0036
0.0091
0.0091
0.0109
0.0109
0.0182
0.0182
Multiplier
2
3
2
2
3
o
J
3
2
o
J
3
o
J
2
2
2
2
Actual /
(Cert*
Multiplier)
0.40
0.22
0.48
0.30
0.74
0.29
0.36
0.41
0.69
0.57
0.48
0.50
0.38
0.37
0.68
(PM/C02)
Actual
0.000224
0.000762
0.000506
0.000674
0.000261
0.000605
0.000789
0.001317
0.000214
0.000000
0.000420
0.000417
0.000895
0.001053
0.001652
(PM/C02)
Certification
0.00109
0.00045
0.00109
0.00182
0.00018
0.00018
0.00018
0.00182
0.00018
0.00018
0.00018
0.00109
0.00109
0.00182
0.00182
Actual /
Certification
0.21
1.69
0.46
0.37
1.45
3.36
4.38
0.72
1.19
0.00
2.33
0.38
0.82
0.58
0.91
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.06
0.52
0.14
0.11
0.44
1.03
1.34
0.22
0.36
0.00
0.72
0.12
0.25
0.18
0.28
147
-------
Table 23: T&M Determination Data for the HHDDT Vehicles over the Transient Mode
Vehicle
E55CRC-1
E55CRC-1
E55CRC-2
E55CRC-2
E55CRC-3
E55CRC-3
E55CRC-4
E55CRC-4
E55CRC-5
E55CRC-5
E55CRC-6
E55CRC-6
E55CRC-7
E55CRC-7
E55CRC-8
E55CRC-8
E55CRC-9
E55CRC-9
E55CRC-10
E55CRC-10
E55CRC-11
E55CRC-11
E55CRC-12
E55CRC-12
E55CRC-13
E55CRC-13
Model
Year
1994
1994
1995
1995
1985
1985
1999
1999
2000
2000
1995
1995
1990
1990
1996
1996
1998
1998
1998
1998
2000
2000
1986
1986
1978
1978
(NOX / CO2)
Actual
0.0130
0.0179
0.0070
0.0073
0.0059
0.0056
0.0078
0.0107
0.0100
0.0118
0.0109
0.0085
0.0129
0.0129
0.0089
0.0113
0.0063
0.0074
0.0132
0.0123
0.0076
0.0074
0.0076
0.0075
0.0109
0.0105
(NOX/C02)
Certification
0.0091
0.0091
0.0091
0.0091
0.0093
0.0093
0.0073
0.0073
0.0073
0.0073
0.0091
0.0091
0.0109
0.0109
0.0091
0.0091
0.0073
0.0073
0.0073
0.0073
0.0073
0.0093
0.0093
0.0093
0.0093
0.0093
Multiplier
3
o
J
o
J
3
2
2
o
J
o
J
3
o
J
3
o
J
o
J
3
o
J
3
o
J
o
J
3
o
J
3
2
2
2
2
2
Actual /
(Cert*
Multiplier)
0.48
0.66
0.26
0.27
0.32
0.30
0.36
0.49
0.46
0.54
0.40
0.31
0.39
0.40
0.33
0.41
0.29
0.34
0.60
0.56
0.35
0.40
0.41
0.41
0.59
0.57
(PM/CO2)
Actual
0.00019
0.00025
0.00026
0.00034
0.00105
0.00124
0.00017
0.00032
0.00029
0.00029
0.00081
0.00062
0.00045
0.00041
0.00037
0.00116
0.00039
0.00036
0.00031
0.00027
0.00017
0.00020
0.00189
0.00188
0.00087
0.00082
(PM/CO2)
Certification
0.00018
0.00018
0.00018
0.00018
0.00182
0.00182
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00109
0.00109
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00182
0.00182
0.00182
0.00182
0.00182
Actual /
Certification
1.07
1.36
1.41
1.85
0.58
0.68
0.94
1.74
1.57
1.57
4.46
3.40
0.41
0.37
2.03
6.36
2.13
2.00
1.71
1.49
0.92
0.11
1.04
1.03
0.48
0.45
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.12
0.15
0.15
0.20
0.06
0.07
0.10
0.19
0.17
0.17
0.48
0.37
0.04
0.04
0.22
0.69
0.23
0.22
0.18
0.16
0.10
0.01
0.11
0.11
0.05
0.05
148
-------
Vehicle
E55CRC-14
E55CRC-14
E55CRC-14
E55CRC-15
E55CRC-15
E55CRC-15
E55CRC-16
E55CRC-16
E55CRC-16
E55CRC-17
E55CRC-17
E55CRC-17
E55CRC-18
E55CRC-18
E55CRC-18
E55CRC-19
E55CRC-19
E55CRC-19
E55CRC-20
E55CRC-20
E55CRC-20
E55CRC-21
E55CRC-21
E55CRC-21
E55CRC-22
E55CRC-22
E55CRC-22
E55CRC-23
E55CRC-23
Model
Year
1985
1985
1985
1986
1986
1986
1979
1979
1979
1993
1993
1993
1991
1991
1991
1987
1987
1987
1992
1992
1992
1990
1990
1990
1993
1993
1993
1983
1983
(NOX/C02)
Actual
0.0078
0.0079
0.0079
0.0085
0.0088
0.0080
0.0069
0.0067
0.0067
0.0079
0.0078
0.0073
0.0083
0.0087
0.0085
0.0062
0.0065
0.0063
0.0126
0.0122
0.0108
0.0107
0.0110
0.0111
0.0092
0.0093
0.0098
0.0099
0.0092
(NOX/C02)
Certification
0.0093
0.0093
0.0093
0.0093
0.0093
0.0091
0.0091
0.0091
0.0091
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0109
0.0091
0.0091
0.0091
0.0109
0.0109
0.0109
0.0091
0.0091
0.0091
0.0109
0.0109
Multiplier
2
2
2
2
2
2
2
2
2
3
o
J
3
2
o
J
3
2
2
2
o
J
3
o
J
2
3
o
J
3
o
J
o
J
2
2
Actual /
(Cert*
Multiplier)
0.42
0.43
0.42
0.46
0.47
0.44
0.38
0.37
0.37
0.24
0.24
0.22
0.38
0.26
0.26
0.29
0.30
0.29
0.46
0.45
0.40
0.49
0.34
0.34
0.34
0.34
0.36
0.45
0.42
(PM/C02)
Actual
0.00080
0.00089
0.00081
0.00152
0.00148
0.00235
0.00529
0.00607
0.00570
0.00075
0.00084
0.00080
0.00066
0.00106
0.00106
0.00256
0.00210
0.00271
0.00004
0.00056
0.00057
0.00101
0.00106
0.00102
0.00099
0.00090
0.00099
0.00120
0.00125
(PM/C02)
Certification
0.00182
0.00182
0.00182
0.00182
0.00182
0.00045
0.00045
0.00045
0.00045
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00109
0.00045
0.00045
0.00045
0.00109
0.00109
0.00109
0.00045
0.00045
0.00045
0.00182
0.00182
Actual /
Certification
0.44
0.49
0.45
0.84
0.82
5.17
11.63
13.35
12.53
0.69
0.77
0.73
0.60
0.97
0.97
2.34
1.93
2.48
0.09
1.23
1.26
0.92
0.97
0.93
2.17
1.97
2.17
0.66
0.69
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.05
0.05
0.05
0.09
0.09
0.56
1.26
1.44
1.35
0.07
0.08
0.08
0.06
0.10
0.11
0.25
0.21
0.27
0.01
0.13
0.14
0.10
0.10
0.10
0.23
0.21
0.23
0.07
0.07
149
-------
Vehicle
E55CRC-23
E55CRC-24
E55CRC-24
E55CRC-24
E55CRC-25
E55CRC-25
E55CRC-25
E55CRC-26
E55CRC-27
E55CRC-28
E55CRC-29
E55CRC-30
E55CRC-31
E55CRC-32
E55CRC-33
E55CRC-34
E55CRC-35
E55CRC-36
E55CRC-38
E55CRC-38
E55CRC-38
E55CRC-39
E55CRC-39
E55CRC-40
E55CRC-40
E55CRC-42
E55CRC-42
E55CRC-43
E55CRC-43
Model
Year
1983
1975
1975
1975
1983
1983
1983
1998
1999
1998
1999
1998
1997
1991
1984
2003
2000
2001
2004
2004
2004
2004
2004
2004
2004
2000
2000
1995
1995
(NOX/C02)
Actual
0.0094
0.0131
0.0143
0.0135
0.0112
0.0118
0.0125
0.0064
0.0072
0.0087
0.0060
0.0157
0.0093
0.0068
0.0139
0.0055
0.0066
0.0059
0.0060
0.0052
0.0059
0.0058
0.0056
0.0063
0.0061
0.0054
0.0052
0.0105
0.0086
(NOX/C02)
Certification
0.0109
0.0182
0.0182
0.0182
0.0109
0.0109
0.0109
0.0073
0.0073
0.0073
0.0073
0.0073
0.0091
0.0091
0.0093
0.0036
0.0073
0.0073
0.0036
0.0036
0.0036
0.0036
0.0036
0.0036
0.0036
0.0073
0.0073
0.0091
0.0091
Multiplier
2
2
2
2
2
2
2
3
o
J
3
o
J
3
3
2
2
o
J
3
3
o
J
3
o
J
3
3
o
J
3
o
J
o
J
3
o
J
Actual /
(Cert*
Multiplier)
0.43
0.36
0.39
0.37
0.51
0.54
0.57
0.30
0.33
0.40
0.28
0.72
0.34
0.37
0.75
0.51
0.30
0.27
0.55
0.49
0.54
0.53
0.52
0.59
0.57
0.25
0.24
0.38
0.31
(PM/C02)
Actual
0.00122
0.00078
0.00169
0.00071
0.00087
0.00081
0.00071
0.00029
0.00044
0.00222
0.00220
0.00025
0.00040
0.00044
0.00150
0.00041
0.00053
0.00031
0.00021
0.00013
0.00018
0.00014
0.00014
0.00010
0.00010
0.00038
0.00035
0.00038
0.00011
(PM/C02)
Certification
0.00182
0.00182
0.00182
0.00182
0.00182
0.00182
0.00182
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00045
0.00182
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
0.00018
Actual /
Certification
0.67
0.43
0.93
0.39
0.48
0.45
0.39
1.60
2.39
12.23
12.12
1.39
2.22
0.97
0.83
2.23
2.90
1.72
1.14
0.72
1.00
0.78
0.80
0.53
0.54
2.09
1.94
2.09
0.62
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.07
0.05
0.10
0.04
0.05
0.05
0.04
0.17
0.26
1.32
1.31
0.15
0.24
0.10
0.09
0.24
0.31
0.19
0.12
0.08
0.11
0.08
0.09
0.06
0.06
0.23
0.21
0.23
0.07
150
-------
Vehicle
E55CRC-44
E55CRC-44
E55CRC-45
E55CRC-46
E55CRC-47
E55CRC-48
E55CRC-49
E55CRC-60
E55CRC-62
E55CRC-63
E55CRC-64
E55CRC-65
E55CRC-66
E55CRC-69
E55CRC-74
E55CRC-78
Model
Year
1989
1989
1993
1989
1986
1998
1994
1995
1983
2005
1994
1994
1989
1989
1969
1975
(NOX/C02)
Actual
0.0081
0.0081
0.0058
0.0080
0.0062
0.0113
0.0069
0.0098
0.0082
0.0053
0.0090
0.0105
0.0084
0.0087
0.0113
0.0190
(NOX/C02)
Certification
0.0109
0.0109
0.0091
0.0109
0.0093
0.0073
0.0091
0.0091
0.0109
0.0036
0.0091
0.0091
0.0109
0.0109
0.0182
0.0182
Multiplier
2
2
3
2
2
o
J
3
3
2
3
o
J
3
2
2
2
2
Actual /
(Cert*
Multiplier)
0.37
0.37
0.21
0.37
0.33
0.52
0.25
0.36
0.38
0.49
0.33
0.38
0.38
0.40
0.31
0.52
(PM/C02)
Actual
0.00044
0.00037
0.00723
0.00076
0.00131
0.00050
0.00125
0.00503
0.00204
0.00022
0.00000
0.00092
0.00095
0.00117
0.00159
0.00062
(PM/C02)
Certification
0.00109
0.00109
0.00045
0.00109
0.00182
0.00018
0.00018
0.00018
0.00182
0.00018
0.00018
0.00018
0.00109
0.00109
0.00182
0.00182
Actual /
Certification
0.41
0.34
16.08
0.70
0.72
2.79
6.97
27.93
1.12
1.24
0.00
5.12
0.87
1.07
0.87
0.34
(Actual /
Cert)
/ Average
(Actual/Cert
+ 2 Sigma)
0.04
0.04
1.73
0.08
0.08
0.30
0.75
3.01
0.12
0.13
0.00
0.55
0.09
0.12
0.09
0.04
151
-------
"Before and After" Emissions
Each vehicle that was identified as a T&M vehicle, repaired and retested, or reflashed and
retested, is presented below in order of occurrence in the program.
E55CRC-3: Injection Timing Change, Phase 1.
E55CRC-3, a single axle full size road tractor with a 1985 Cummins 14 liter engine,
became a T&M subject fortuitously. It was tested twice, because it was mistakenly
believed that data from the first set of tests were in error and that retest was necessary,
and the vehicle owner happened to effect repairs between the two test periods. E55CRC-3
did not meet the high PM emissions criteria but it was observed during a later data
analysis that PM was higher than would be expected for a vehicle of its MY and NOX was
lower. Data from the first round of testing suggested that the engine might be operating at
NOX levels as low as 3g/bhp-hr (4 g/kw-hr) and the NOX data for the retest corresponded
to about 4.5g/bhp-hr (6 g/kw-hr) on a certified engine test (FTP), based on NOX/CO2
ratio. The certification level for that engine was 5.1g/bhp-hr (6.8 g/kw-hr) so 4.5g/bhp-hr
(6 g/kw-hr) would correspond to certification level less a 10% "safety margin". Fuel
specifications in CA. had also changed since 1985 and recent CA fuel is argued to cause
lower NOX. It was evident from the retest data that the PM, CO and HC were lower in the
retest than in the original test. Table 24 and Table 25 summarize the data taken from the
test and retest for laden UDDS.
Table 26 to Table 29 summarize the data taken from the test and retest for laden
HHDDT.
Table 24: E55CRC-3: UDDS (g/mile)
Run Seq. No.
1826-3
1826-4
CO
14
12.4
NO/
12.5
12.3
NOX2
12.2
12.6
HC
2.7
2.78
PM
1.76
1.76
C02
2155
2255
mile/gal
4.6
4.41
BTU/mile
28238
29450
Miles
5.52
5.51
Table 25: E55CRC-3 retest UDDS (g/mile)
Run Seq. No.
1936-2
1936-3
CO
8.22
8.53
NO/
17.9
17.1
NOX2
18.8
16.9
HC
1.76
1.66
PM
1.15
1.57
C02
2413
2408
mile/gal
4.13
4.14
BTU/mile
31421
31359
Miles
5.2
5.26
Table 26: E55CRC-3: 1828-1 Idle, 1828-2 Creep, 1828-3 Transient, 1828-4 Cruise.
(g/mile, except idle in g/cycle)
Run Seq. No.
1828-1
1828-2
1828-3
1828-4
CO
10.4
26.48
8.58
2.30
NO/
8.7
14.42
9.07
5.53
NO/
8.8
13.24
8.89
5.47
HC
7.2
12.49
2.98
1.12
PM
1.84
5.10
1.60
0.42
C02
1088
2740
965
549
mile/gal
0.02
0.94
2.69
4.77
BTU/mile
9052
36451
12768
7188
Miles
0.01
0.13
2.73
23.02
152
-------
Table 27: E55CRC-3: 1829-1 Idle, 1829-2 Creep, 1829-3 Transient, 1829-4 Cruise.
(g/mile, except idle in g/cycle)
Run Seq. No.
1829-1
1829-2
1829-3
1829-4
CO
29.6
48
12.6
3.3
NO/
1.8
25.6
13.7
7.8
NOX2
2
25
14.4
8.1
HC
2.4
23.4
5.6
1.7
PM
2.37
6.05
3.02
0.65
C02
1167
5896
1884
851
mile/gal
<0
1.66
5.23
11.66
BTU/mile
15780
78121
24826
11137
Miles
-0.01
0.13
2.78
23
Table 28: E55CRC-3 retest: 1939-1 Idle, 1939-5 Creep, 1939-3 Transient, 1939-4
Cruise (g/mile, except idle in g/cycle)
Run Seq. No.
1939-1
1939-5
1939-3
1939-4
CO
5.87
20.7
9.47
2.65
NO/
9.2
44.3
20.4
12.4
NOX2
9.2
45.5
20
12.5
HC
4.19
10.19
2.38
0.99
PM
0.75
2.53
1.56
0.45
C02
1406
4878
2743
1579
mile/gal
<0
2.03
3.63
6.34
BTU/mile
18462
63862
35731
20492
Miles
-0.01
0.12
2.65
22.04
Table 29: E55CRC-3 retest: 1942-1 Idle, 1942-5 Creep, 1942-3 Transient, 1942-4
Cruise, (g/mile, except idle in g/cycle)
Run Seq. No.
1942-1
1942-5
1942-3
1942-4
CO
6.8
25.4
11.1
3.1
NO/
12.5
31.5
18.6
10.7
NO/
12.3
25.7
19
10.8
HC
4.28
9.01
2.95
1.03
PM
0.75
3.55
1.66
0.49
C02
1480
5593
3313
1603
mile/gal
<0
1.77
3.01
6.24
BTU/mile
80732
73153
43158
20812
Miles
0
0.11
2.63
22.02
The low NOX in the first round was due to substantially retarded timing on the truck. This
engine employed a mechanical variable injection timing system, which included an air-
operated cylinder that varied the position of the cam followers. It is likely that this unit
was worn, because during preventive maintenance between the original test and retest,
the vehicle owner effected repairs to the truck that included a rebuild of this air cylinder
with new seals. In other words, the timing was corrected between the original testing and
retesting, resulting in raised NOX, and lower PM, CO and HC, according to the well-
known NOX-PM tradeoff.
E55CRC-10: Reflash Vehicle in Phase 1
E55CRC-10, a 1998 MY Sterling tractor with a DDC Series 60 engine, exhibited high
NOX emissions. The emissions, as tested on the Heavy-Duty UDDS and the HHDDT
Schedule at 56,000 Ibs. (25,402 kg), are shown in Table 22. It was evident that the major
cause of the elevated emissions was "off-cycle" operation, as shown in Figure 122 above.
153
-------
After the vehicle was reflashed, the Transient mode NOX emissions for E55CRC-10 did
not change, but the UDDS NOX emissions dropped by about 20% as shown in
—
!/)
c
^o
*s
I/I
H
O
z
E55CRC-10
E55CRC-10 reflashed
UDDS
Transient
Cruise
Figure 127 and the reduction on cruise was over 30%. This is to be expected keeping in
mind the relative amounts of steady operation in each of these test modes and the fact that
steady operation encourages "off-cycle" behavior. Figure 128 shows the NOX emissions
versus power for E55CRC-10 after reflash, and this figure may be compared with Figure
122. The investigators concede that Idle and Creep data were too variable for "before and
after" comparisons because auxiliary engine loads affect idle and creep modes.
154
-------
a
_o
*S
VI
H
O
z
E55CRC-10
E55CRC-10 reflashed
UDDS
Transient
Cruise
Figure 127: Transient Mode NOX emissions before and after reflashing E55CRC-10.
155
-------
Table 30: E55CRC-10 baseline emissions and "reflash" emissions. T&M runs are
denoted in bold. NOXJ and NOX2 show values from two similar analyzers in parallel.
It should be noted that Idle and Creep Modes often return highly variable emissions
due to changing auxiliary loads on the vehicle.
Test ID
1963 Before
1963 Before
2195 Reflash
1966 Before
1969 Before
2197 Reflash
1966 Before
1969 Before
2197 Reflash
1966 Before
1969 Before
2197 Reflash
1966 Before
1969 Before
2197 Reflash
Test
Run
ID
2
3
5
1
1
2
2
2
3
o
J
3
4
4
4
5
Driving Schedule
UDDS
UDDS
UDDS
Idle
Idle
Idle
Creep
Creep
Creep
Transient
Transient
Transient
Cruise
Cruise
Cruise
CO
(g/mile)
8.37
9.16
10.02
6.45
7.65
4.47
30.07
40.23
33.31
11.77
10.91
9.53
2.78
3.05
2.42
NOx1
(g/mile)
38.68
40.57
32.71
21.36
17.81
27.78
54.9
60.71
75.66
32.99
30.48
31.71
49.43
46.24
33.93
NOX2
(g/mile)
38.58
40.25
32.79
22.71
17.66
29.73
50.52
62.35
75.1
33.13
31.09
31.54
49.11
46.42
33.58
HC
(g/mile)
0.22
0.22
0.27
1.49
1.15
0.9
4.43
5.26
5.4
0.46
0.49
0.66
0.13
0.14
0.15
PM
(g/mile)
0.53
0.53
0.41
0.16
0.13
0.07
1.07
1.49
1.09
0.78
0.67
0.59
0.12
0.15
0.12
C02
(g/mile)
2190
2088
2188
1176
923
1057
3834
4415
4697
2506
2485
2666
1501
1514
1513
o
W
6
y = 2E-06x" + 0.002x + 0.2032
R2 = 0.4803
y = 2E-06x + O.OOlx + 0.0749
R2 = 0.5935
0 50 100 150 200 250 300 350 400
Axle Horsepower (ahp)
Figure 128: NOX emissions versus power for E55CRC-10 after reflashing. The mass
rate of NOX emissions was reduced in many parts of the Cruise mode (compared to
Figure 122). The highest mass rate emissions were also slightly lower than before the
reflash.
156
-------
E55CRC-16: High PM T&M Vehicle in Phase 1
E55CRC-16 was a 1979 White tractor with a 3208 Caterpillar engine. It was identified
visually as a prodigious smoker and returned data as a high emitter. It was chosen as a
T&M candidate. Emissions data for this vehicle are presented in Table 31. The vehicle
was naturally aspirated, so the investigators believed that the high PM was due to over
fueling or substantially retarded timing. It was suggested34 that cam wear also might be
an issue. The vehicle was sent to a Caterpillar dealership with directions to inspect the
engine. They found no substantial wear or indication of poor injection timing. The
dealership determined that the engine had originally been equipped with exhaust gas
recirculation (EGR) for NOX reduction, but that EGR components had been removed and
that the system was not functioning.
The WVU investigators and the program sponsors determined that the fuel pump flow
should be measured, and corrected if necessary, or the fuel pump should be replaced with
a rebuilt unit. The dealership determined that they were unable to quantify the flow and
so the pump was replaced.
Table 31: E55CRC-16 Emissions before and after T&M procedures. T&M 1 in bold
represents testing with a new fuel injection pump; T&M 2 in bold italics represents
testing with a new pump and EGR totally disabled.
Test ID
2088 Before
2088 Before
2088 Before
2180 T&M 1
21 82 T&M 2
2181 Before
2183 Before
2093 Before
2094 Before
2095 Before
2181 T&M 1
21 83 T&M 2
2093 Before
2094 Before
2095 Before
2181 T&M 1
21 83 T&M 2
2093 Before
2094 Before
2095 Before
2181 T&M 1
21 83 T&M 2
Test Run
ID
3
4
5
3
2
1
1
2
2
2
2
2
3
3
3
3
3
4
4
4
4
4
Driving
Schedule
UDDS
UDDS
UDDS
UDDS
UDDS
Idle
Idle
Creep
Creep
Creep
Creep
Creep
Transient
Transient
Transient
Transient
Transient
Cruise
Cruise
Cruise
Cruise
Cruise
CO
(g/mile)
20.54
20.62
20.44
109.16
60.59
9.53
8.28
33.1
37.92
34.52
26.89
28.67
27.13
26.97
27.22
181.23
119.05
9.07
8.99
8.99
51.26
19.95
NOx1
(g/mile)
14.69
15.29
14.55
13.78
14.79
10.23
6.58
60.74
52.56
56.75
51.21
53.77
17.97
16.95
17.01
16.09
20.88
10.17
9.5
10.26
10.07
73.2
NOX2
(g/mile)
14.83
15.35
14.19
13.87
75.39
12.5
6.24
59.82
50.75
56.86
53.17
54.25
17.67
16.95
16.82
15.89
21.03
10.25
9.99
10.04
10.23
73.7
HC
(g/mile)
1.61
2.08
1.52
4.4
7.37
1.05
0.19
8.59
7.25
7.46
2.64
3.29
4.08
4.08
3.74
5.48
6.7
0.34
0.3
0.33
1.01
7.56
PM
(g/mile)
10.89
12.99
12.5
24.23
16.31
0.54
0.04
6.05
8.66
6.23
6.23
4.23
13.82
15.42
14.39
25.97
17.48
17.14
15.37
15.48
9.31
4.6
CO2
(g/mile)
2181
2316
2192
2065
1821
992
889
3867
3410
3736
3644
3419
2615
2541
2526
2321
2538
1814
1784
1779
1476
1504
157
-------
The vehicle was retested with the new fuel pump, and data are shown in denoted as T&M
1. There was virtually no change in the NOX emissions with the new pump on the UDDS,
the Cruise mode and the Transient mode. PM rose with the new pump for the Transient
cycle and HDDS, but was lowered for the Cruise mode.
A final inspection of the vehicle found soot in the intake manifold, which suggested that
the EGR system was still adding some EGR into the intake. The EGR feed was then
blocked and this eliminated EGR entirely, and the vehicle was tested again. This was not
a repair, but a diagnostic move. The elimination of the EGR (with the new fuel injection
pump) raised the NOX on the Transient and Cruise modes and on the UDDS as shown in
Table 31 denoted as T&M 2. PM was reduced relative to the PM with the new fuel pump
prior to disabling the EGR (i.e. PM for T&M 2 was lower than PM for T&M 1), for the
UDDS, and the Transient and Cruise modes. The Cruise mode reduction was substantial.
However, it was not clear that the cause of the high PM emissions was completely
resolved. The final PM levels for E55CRC-16 after both pump replacement and disabling
of EGR were higher for the UDDS and Transient modes than for the vehicle as received.
The T&M exercise with E55CRC-16 was therefore not fully successful and emphasized
the difficulty in attacking an old, potentially worn engine with no single, clear
malfunction.
E55CRC-21: T&M Vehicle by Inspection in Phase 1
E55CRC-21, a 1990 Caterpillar powered tractor, became a T&M vehicle by inspection.
The air filter was visibly dirty. Table 32 presents the emissions from E55CRC-21. It was
found using a manometer that the pressure drop across the filter at high air flow (during
full power acceleration) into the engine was 39 inches of water (9,714 Pa). This exceeded
the typical recommended intake depression limit of 18 inches of water (4,483 Pa). Intake
restriction reduces engine air intake flow, increases internal exhaust gas retained in the
cylinder, lowers the air to fuel ratio, and promotes formation of products of incomplete
combustion. The filter was replaced with a new filter. The intake depression dropped to
5.5 inches of water. The vehicle was retested and data are shown in Table 32. PM for the
vehicle on the UDDS dropped by 27%, while it dropped 18% for the transient mode and
25% for the Cruise mode. The CO dropped by over 20% for the UDDS and Cruise
modes, while the Transient mode showed a CO decrease of only 2 percent. The data
suggest that CO and PM may be reduced by replacing air filters regularly.
158
-------
Table 32: E55CRC-21 emissions T&M results in bold.
Test ID
2156
2156
2156
2168 T&M
2157
2158
2159
2169 T&M
2157
2158
2159
2169 T&M
2157
2158
2159
2169 T&M
2157
2158
2159
2169 T&M
2160
2160
Run
ID
o
J
4
5
3
1
1
1
1
2
2
2
2
3
o
J
3
3
4
4
4
4
1
2
Driving Schedule
UDDS
UDDS
UDDS
UDDS
CARB- idle3. eye
CARB- idleS. eye
CARB- idleS. eye
CARB - idle3.cyc
CARB - creepS. eye
CARB - creepS. eye
CARB - creepS. eye
CARB - creep3.cyc
CARB-transS.cyc
CARB-transS.cyc
CARB-transS.cyc
CARB - trans3.cyc
CARB - cruiseS.cyc
CARB - cruiseS.cyc
CARB - cruiseS.cyc
CARB - cruiseS.cyc
AC5080
AC5080
CO
(g/mile)
19.79
19.16
22.45
16.19
9.58
9.81
13.09
16.93
42.37
47.09
56.18
56.73
18.62
20.01
19.08
18.92
7.81
9.42
8.05
6.58
8.36
8.14
NOxi
(g/mile)
23.79
24.27
23.64
25.39
25.49
25.84
12.5
24.16
82.16
84.99
87.99
79.73
28.53
30.03
29.67
28.64
18.81
18.8
19.27
19.76
19.49
19.33
NOX2
(g/mile)
24.37
24.94
24.22
25.95
26.49
26.4
14.18
23.62
82.7
87.7
86.1
77.26
28.97
30.01
28.69
29.33
18.26
18.56
18.94
19.45
19.16
20.87
HC
(g/mile)
0.41
0.49
0.4
0.43
1.02
1.85
1.63
2.13
5.76
6.92
9.11
7.75
0.96
1.38
1.13
1.1
0.24
0.26
0.26
0.21
0.31
0.32
PM
(g/mile)
2.89
2.93
3.09
2.18
0.27
0.36
0.38
0.49
2.29
2.68
3.48
3.59
2.68
2.89
2.72
2.25
1.05
1.02
0.89
0.74
1.41
1.33
C02
(g/mile)
2389
2364
2400
2430
971
1055
967
1102
4067
4220
4537
4417
2664
2733
2668
2621
1627
1632
1640
1618
1863
1852
CRC-26. CRC-28 and CRC-31. Reflash Vehicles in Phase 1.5
CRC26 was a 1999 Freightliner tractor with a 1998 Caterpillar C-10 Engine. CRC28 was
a 1999 Freightliner with a 1998 Series 60 Detroit Diesel Engine. CRC31 was a 1998
Kenworth with a 1997 N14 Cummins engine. E55CRC-26, E55CRC-28 and E55CRC-31
were tested using the full baseline test program in Phase 1.5. At the end of the baseline
test program, all three vehicles were reflashed to a "low-NOx" ECU map by their
respective dealerships, and re-tested. After the reflash, the vehicle was retested using a
limited set of schedules, as follows. After warming the vehicle and dynamometer, the
researchers conducted a coastdown, a UDDS at 56,000 Ibs., a HHDDT Idle, a HHDDT
56,000 Ibs. Creep Mode, a HHDDT 56,000 Ibs. Transient Mode and a HHDDT 56,000
Ibs. Cruise Mode.
E55CRC-28 was also identified as a high PM emitter, and was tested (i) as received, (ii)
without repair but with reflash, (iii) with repair and reflash, and (iv) without repair but
with reflash. E55CRC-31 was a vehicle that required mechanical attention during testing,
as discussed in a section below.
Each of the three reflash vehicles is discussed separately below. Figure 129 (as received)
and Figure 130 (after reflash) show the NOX versus hub power plots for all three reflash
159
-------
vehicles on the Cruise Mode. Reduction of NOX due to reflash is evident. E55CRC-31
still shows a bifurcation in its emissions rate versus power output, but values over 0.8
g/sec have been eliminated after reflash.
i
100
200 300
Hub Power (Hp)
400
500
600
Figure 129: NOX emissions on Cruise mode (56,000 Ibs.) for three vehicles before
reflash. NOX emissions and hub power were time-aligned before plotting.
1.2 -r
• CRC 26 R
• CRC 28 R
A CRC 31 R
0.0
100
200 300 400
Hub Power (Hp)
500
600
Figure 130: NOX emissions on the three reflash vehicles were reduced in Cruise
mode after reflash. Test weight was 56,000 Ibs.
160
-------
As-received and reflash data for NOX for E55CRC-26 are presented in Figure 131. As
might be expected, transient operation was not affected by the reflash, and the NOX data
for the UDDS, Creep mode and Transient mode did not change appreciably due to
reflash.
Figure 131: As-received and retest after reflash NOX data for E55CRC-26 (56,000
Ibs.).
Figure 132: As-received and retest after reflash PM data for E55CRC-26 (56,000
Ibs.).
161
-------
The Cruise Mode, however, exhibited a reduction in NOX from 24.19 to 16.77 g/mile due
to the reflash. PM data for the Creep Mode, and Transient mode were not changed
appreciably by the reflash, but the PM for the UDDS was 27% lower and the PM for the
Cruise mode dropped from 0.38 to 0.21 g/mile after the reflash. One might expect PM
values to rise slightly after a reflash as part of the NOX -PM tradeoff, but TEOM data
(0.37 g/mile as-received, 0.20 g/mile after reflash) were consulted and found to support
the substantial reduction in PM filter mass on the Cruise mode after reflash.
E55CRC-28, selected as a reflash vehicle, was also identified as a high PM emitter. The
researchers suggested that the cause might be a failed manifold air pressure (MAP)
sensor. The MAP sensor is used in the engine control strategy to limit fueling in
sympathy with rising turbocharger boost during transients that demand rise in torque.
Without fueling limitation, the combustion becomes insufficiently lean and elemental
carbon is produced in rich zones during combustion. High CO is also produced by failure
to limit fueling during transients, and E55CRC-28 was also observed to have high CO
emissions. Valley Detroit Diesel, a dealership in Fontana, CA, confirmed that the MAP
sensor had failed, and were able to provide a new sensor. These sensors are readily
changed.
PM was the species of interest for the repair. Figure 133 and Figure 134 show the PM
levels produced by E55CRC-28 on the UDDS and HHDDT (four modes) at 56,000 Ibs.
test weight with and without reflash and with and without MAP sensor replacement
(repair).
6.0
"Sfc
o
O
z
E55CRC-28
E55CRC-28 Repair
without reflash
E55CRC-28 Repair
& reflash
E55CRC-28
Reflash without
repair
I UDDS
5.24
0.81
1.11
2.69
I Cruise
0.65
0.19
0.37
0.50
] HHDDT S
1.13
0.32
0.49
0.75
Figure 133: PM emission for E55CRC-28 for the UDDS, Cruise and HHDDT S
Modes.
162
-------
14.0
"Si
e
o
£
W
S
PH
I Creep
I Transient
I Creep
I Transient
6.10
12.17
E55CRC-28 Repair
without reflash
1.80
1.86
E55CRC-28 Repair
& reflash
1.64
1.71
E55CRC-28 Reflash
without repair
3.01
1.39
Figure 134: PM emissions for the Transient and Creep Modes for E55CRC-28.
In all cases, repair lowered PM substantially, whether with or without reflash of the ECU.
However, PM was also lowered in general for the reflash without repair. In the case of
the Creep Mode, the PM from reflash without repair was below even the values for PM
with repair. However, PM is known to be variable on the Creep Mode. It is possible to
curb PM production even if the MAP sensor fails by adding a time-based fueling
limitation to the ECU strategy. However, the authors have no direct information as to
whether the reflash code may have contained this added strategy.
The reflash would be expected to influence NOX more than other species. Figure 135 and
Figure 136 show the NOX levels before and after reflash. The repair without the reflash
raised NOX values by about 30% for the UDDS and the four HHDDT modes. The reflash
did little to abate NOX for the Transient and Creep Modes, but offered modest reductions
for the Cruise and HHDDT_S Modes and the UDDS for the repaired condition. NOX on
the UDDS was actually found to rise slightly on the UDDS for the case without repair.
163
-------
i
E55CRC-28
E55CRC-28 Repair
without reflash
E55CRC-28 Repair
& reflash
E55CRC-28
Reflash without
repair
IUDDS
25.73
33.57
29.43
29.22
I Cruise
26.63
33.57
26.36
18.95
DHHDDT s
32.45
42.46
27.88
21.33
Figure 135: NOX emissions for E55CRC-28 at 56,000 Ibs. on the UDDS and the
Cruise and HHDDT S Modes.
e
o
w
H
O
z
E55CRC-28
E55CRC-28 Repair
without reflash
E55CRC-28 Repair
& reflash
E55CRC-28
Reflash without
repair
I Transient
23.79
33.11
31.89
24.31
I Creep
54.34
76.36
93.63
81.28
Figure 136: NOX emissions for E55CRC-28 at 56,000 Ibs. on the Transient and
Creep Modes.
164
-------
E55CRC-31 was subjected to as-received testing and testing after reflash. The vehicle
suffered from a clutch failure, an exhaust leak, and a fuel injector failure during testing.
The initial testing of vehicle E55CRC-31 was conducted on 7/2/03 during the time of the
speciation sampling. HHDDT and HDDS tests were performed at 56,000 Ibs. and an
HHDDT at 66,000 Ibs. The vehicle was re-procured on 7/15/03 for the HHDDT 30,000
Ibs. test to complete the as-received testing.
The vehicle was then sent to a local dealership to be "reflashed" for a low NOX
calibration. On 7/16/03, the vehicle was retested with the reflash. One HHDDT at 56,000
Ibs. was completed but the required UDDS was not performed.
The vehicle was brought back on 7/28/03 to complete the reflash testing. While
conducting the retest with the low NOX reflash of vehicle E55CRC-31, a significant
exhaust leak developed. The vehicle was removed from the test bed to repair the exhaust
leak and testing continued with other vehicles. When vehicle E55CRC-31 was procured
again, a problem with the clutch prevented testing. WVU staff adjusted the clutch but this
did not solve the problem. A technician from the rental company was called to service the
vehicle. Both the clutch-brake and the clutch were found to need replacement. While
testing the vehicle after changing the clutch, an engine mis-fire was detected that was not
present during prior testing. The technician determined the problem to be a faulty
injector. The fuel injector was replaced. Testing was completed. Table 33 shows the
sequences of repairs and retests. These repairs were not treated as T & M repairs.
Table 33: Sequence of testing for E55CRC-31
Date
7/2/03
7/15/03
7/16/03
7/28/03
8/19/03
Tests performed
56,000 HHDDT
56,000 UDDS
66,000 HHDDT
30,000 HHDDT
56,000 HHDDT reflash
did not do the 56,000 UDDS
56,000 HHDDT reflash (TransS, CruiseS) to verify
data from 7/16/03
56,000 UDDS reflash (significant exhaust leak
developed)
Found problems with the clutch, clutch brake, and
fuel injector
56,000 HHDDT reflash
56,000 UDDS reflash
Seq. No.
2532
2534
2535
2573
2577/78
2625
2626
2685
2684
Action
Tests reported
Tests reported
Tests not reported
Tests not reported
Truck repaired
Tests reported
As-received and reflash emissions are shown in Figure 137 and Figure 138. The reflash
reduced NOX emissions (in g/mile) for the UDDS and the Creep, Transient and Cruise
modes of the HHDDT. PM emissions (in g/mile) rose slightly on the UDDS, and dropped
from 1.24 g/mile to 0.76 g/mile on the Transient Mode, but were otherwise changed little.
However, it was found that CO2 emissions were also reduced, (as shown in the short
reports in Appendix K) which is not expected, so that g/gallon values are not aligned with
the distance specific values.
165
-------
Figure 137: As-received and reflash emissions for NOX emissions for E55CRC-31
tested at 56,000 Ibs.
Figure 138: As-received and reflash emissions for PM emissions for E55CRC-31
166
-------
CRC-45 T&M High PM Vehicle in Phase 2
The Phase 2 E-55/59 program test plan included a provision for repair and retest of one
vehicle that was identified during the T&M inspection or determined otherwise to have
abnormally high emissions. Vehicle E55CRC-45 was noted visually to be a prodigious
smoker. It was not a high emitter on the Cruise mode but was a high emitter (on only one
test run) on the Transient mode. This vehicle emitted PM at an impressively high level at
low engine load. On the Creep mode, the PM mass (67.8 g/mile) exceeded the NOX mass
substantially, and PM exceeded NOX also on the Transient mode. The high PM mass was
accompanied by exceptionally high HC emissions. In some cases, the HC emissions were
of the order of one tenth of the fuel burned to form CC>2, and it was realized that raw fuel
was most likely being admitted to the exhaust stream.
The sponsors were notified of the high emissions and they elected to authorize repair and
retest. A repair was effected: it was determined at the dealership that the No. 1 injector
timing was out of specification, and the injector was replaced. The vehicle was retested as
E55CRC-45R. PM remained high after the repair. Both PM and NOX were reduced on the
Creep mode, but NOX and PM mass were still similar to one another (38.2 and 37.5
g/mile respectively). On Transient mode PM was virtually unchanged. PM rose from 1.1
g/mile on the Cruise mode before repair to 4.5 g/mile on the Cruise mode after repair, and
a similar increase was seen on the HHDDT_S. Clearly the emissions from the vehicle had
changed, but the cause of the high PM at low load had not been addressed.
The vehicle was returned to the dealer for additional diagnosis and repair, and a
reconditioned turbocharger was fitted. Data collected after this second repair were
recorded under the vehicle designation E55CRC-45RR. The emissions changed, but the
high PM was not cured. Once again, on Creep and Transient modes, the PM mass was
similar to the NOX mass.
Table 34 shows the emissions from E55CRC-45 as received, after the first repair, and
after the second repair. It is evident that the PM emissions were associated with high HC
emissions and that the fault was not diagnosed. The WVU investigators suspect that a
leak from an injector into the cylinder might have been the cause, but no further repairs
were authorized by the sponsors.
167
-------
Table 34: E55CRC-45 as received, after first repair (R), and after second repair (RR).
WVURefNum
E55CRC-45
E55CRC-45 R
E55CRC-45 RR
E55CRC-45
E55CRC-45 R
E55CRC-45 RR
E55CRC-45
E55CRC-45 R
E55CRC-45 RR
E55CRC-45
E55CRC-45 R
E55CRC-45 RR
E55CRC-45
E55CRC-45 R
E55CRC-45 RR
Test ID
2967
2985
20008
2968
2986
20009
2968
2986
20009
2968
2986
20009
2968
2986
20009
Test Run
ID
2
1
1
3
3
3
2
2
2
4
7
6
5
5
7
Cycle
TEST D
TEST D
TEST D
TRANS3
TRANS3
TRANS3
CREEP34
CREEP34
CREEP34
CRUISES
CRUISES
CRUISES
HHDDT S
HHDDT S
HHDDT S
CO g/mile
3.47
5.42
9.70
4.99
7.14
15.56
32.95
27.66
47.56
1.53
2.69
5.39
1.18
2.69
4.39
CO2 g/mile
2029
2230
2164
2369
2502
2513
3806
4061
3857
1468
1447
1372
1612
1611
1580
NOX g/mile
11.98
11.11
11.78
13.68
12.72
13.61
44.97
38.38
40.80
8.81
7.13
10.33
9.62
8.30
10.54
NOX2 g/mile
12.36
11.20
12.04
13.63
12.70
13.42
44.14
37.95
38.14
8.77
7.33
10.33
9.46
8.41
10.42
HC g/mile
15.76
31.70
29.24
29.46
47.03
40.96
222.73
158.52
134.90
5.43
17.40
14.54
3.09
19.78
17.09
PM g/mile
3.04
4.73
4.54
17.14
16.59
13.27
67.77
37.51
29.59
1.12
4.52
2.52
0.86
2.93
2.99
168
-------
E55CRC57: MHDT Temperature Sensor Failure
There were no formal criteria for identifying high MHDT emissions. Vehicle CRC57, a
MHDDT powered by a Caterpillar engine, showed unusually high unladen NOX
emissions. This vehicles apparently had an issue with a temperature sensor that indicated
an overheated engine to the ECU. The ECU would then command a limp home mode
resulting in the vehicle's inability to follow the test target speed (See Figure 139). The
clutch actuated cooling fan would only kick in after the engine actually was warm. At
that point the driver could return the vehicle speed to one consistent with the target trace.
Conversation with Robert Graze, of Caterpillar, confirmed that the WVU diagnosis of a
temperature sensor misrepresenting the temperature to the ECU was reasonable.
JS
a
•a
u
u
a
70.0
60.0
50.0
— Actual (mph)
— Target (mph)
500
1000 1500
Time (sec)
2000
2500
Figure 139: shows the inability of E55CRC-57 to meet the target trace.
T&M and Reflash Discussion
Identification of vehicles for T&M by inspection proved successful insofar as it targeted
E55-21, with a dirty air filter. Also, the excessive smoking of E55CRC-16 and E55CRC-
45 provided visual indications of T&M vehicles. However, inspections are difficult to
perform, and some forms of tampering with sensors may be difficult to identify. In this
way, visual inspection is not foolproof, but will detect some emissions-related problems.
The method of identifying high PM emitters also appeared to be sound, and used the best
approach that could be employed for rapid processing of field data. Moreover, the
identification of high PM emitters using data from two different HHDDT Schedule
Modes (Transient and Cruise) allowed the researchers to infer the cause of the smoking.
169
-------
High PM emissions on the Transient Mode, without high emissions on the Cruise Mode,
might indicate poor puff control.
The criterion for identifying high NOX emitters was based on a multiple of the
certification NOX to CC>2 ratio. NOX tracks CC>2 closely for a vehicle with a timing map
that keeps location of peak in-cylinder pressure at the same crank angle over a broad
operating range. In this way, the NOX to CC>2 ratio should be largely cycle independent,
and should rise with advanced timing and fall with retarded timing. Unfortunately this
approach is confused by the timing strategy that causes "off-cycle" NOX emissions in
many electronic engines. Off-cycle strategies may be based on two maps, one with
advanced timing (high NOX), and one with retarded timing (low NOX), and these maps
may be invoked during different sections of a cycle. Off-cycle strategies cannot be
regarded as a T&M target because the engine is behaving in the way that it was originally
sold, and any high NOX emissions arising during off-cycle use are neither due to
tampering nor due to malmaintenance.
Being cognizant of the "off-cycle" phenomenon, the researchers set generous T&M
margins to detect high NOX emitters. The intent was not to snare off-cycle trucks, but to
detect only extreme (uncharacteristically high) NOX emitters. In execution of this
strategy, only one high NOX emitter was detected, and this was determined through
further study to be rooted in an off-cycle issue. An important point was that the emissions
for this vehicle, E55CRC-10, were extreme for the Cruise Mode: it is known that steady
state cruising tends to elicit off-cycle behavior in many cases. The investigators also
admit that E55CRC-10 was found to be high because the criterion was developed with 5
g/bhp-hr (6.7 g/kw-hr) NOX engines in mind, while E55CRC-10 was a 4 g/bhp-hr (5.4
g/kw-hr) engine. The criterion should perhaps have been a NOX to CO2 ratio that was
fixed in value (say 0.03 by mass) rather than "three times the certification ratio".
However, E55CRC-10 would still have just been high on Cruise mode with a 0.03
criterion. It would have had "expected" emissions levels after reflashing if a 0.03
criterion were used. Otherwise, more sophisticated strategies, using plots such as those
shown in Figure 122, could be used to distinguish high NOX T&M trucks from "off-
cycle" behavior. The E55CRC-10 findings, in part, led to the reflash research in Phase
1.5 of the program, where reflash NOX reductions were verified.
As part of an inventory program in southern California, trucks that were identified as
high emitters were repaired and retested for emissions. A 1990 Caterpillar-powered
tractor was identified as having a dirty air filter, and air filter replacement did reduce CO
and PM emissions. A 1979 Caterpillar-powered truck was identified as a prodigious
smoker, but the cause of the high emissions was not satisfactorily identified. A 1998
Detroit Diesel-powered tractor was identified as having high NOX emissions, but they
were attributed to "off-cycle" timing strategies rather than T&M issues. The vehicle was
"reflashed" with a more retarded timing strategy and lower NOX emissions were seen
under cruise conditions. In the case of a 1985 Cummins-powered tractor, repairs to a
failed variable mechanical injection timing system raised NOX to an expected level and
reduced other regulated emissions. For a 1993 Cummins-powered truck, the cause of
excessive PM production was not cured, despite two rounds of repair: it was evident that
170
-------
the high PM was associated with exceptionally high HC emissions. A MHDDT
malfunctioned during testing, and was unable to follow the trace due to a false high
temperature warning. This was not an emissions fault, but did affect distance-specific
emissions by not allowing the vehicle to follow the cycle closely. It is concluded that
trucks that are poorly maintained or have failures are contributing additional emissions to
the inventory, but that diagnosis and repair is not necessarily straightforward.
171
-------
ACKNOWLEDGEMENTS
The WVU researchers are grateful to the sponsors listed below, for support of this
program.
• Coordinating Research Council, Inc.
• California Air Resources Board
• U.S. Environmental Protection Agency
• U.S. Department of Energy Office of FreedomCAR & Vehicle Technologies
through the National Renewable Energy Laboratory
• South Coast Air Quality Management District
• Engine Manufacturers Association
The dedication of the WVU field research team is gratefully acknowledged.
Ralphs Grocery was kind to provide a location for the WVU laboratory during part of this
research program.
The contributions of Thomas R. Long, Jr., Sairam Thiagarajan, Shuhong (Susan) Xu,
David McKain and Kuntal Vora in preparing and editing sections of this report are
acknowledged.
172
-------
REFERENCES
1. Yanowitz, J., McCormick, R.L., and Graboski, M.S., "Critical Review: In-Use
Emissions from Heavy-Duty Diesel Vehicles", Environmental Science &
Technology. Vol.334:729-740.
2. Gautam, M., Clark, N.N., Riddle, W., Nine, R., Wayne, W.S., Maldonado, H.,
Agrawal, A. and Carlock, M., "Development and Initial Use of a Heavy-Duty Diesel
Truck Test Schedule for Emissions Characterization," 2002 SAE Transactions:
Journal of Fuels & Lubricants Vol. Ill, pp. 812-825.
3. Clark, N.N., Gautam, M., Wayne, W.S., Nine, R.D., Thompson, G.J., Lyons, D.W.,
Maldonado, H., Carlock, M. & Agrawal, A., "Creation and Evaluation of a Medium
Heavy-Duty Truck Test Cycle" SAE Powertrain Conf, Pittsburgh, Oct. 2003, SAE
Paper 2003-01-3284.
4. Clark, N.N., Gautam, M., Riddle, W., Nine, R.D., and Wayne, W.S., "Examination of
a Heavy-Duty Diesel Truck Chassis Dynamometer Schedule," SAE Powertrain
Conference, Tampa, Fla., Oct 2004, SAE Paper 2004 -01-2904.
5. Anyon, P., Brown, S., Pattison, D., Beville-Anderson, J., Walls, G., Mowle, M.,
"Proposed Diesel Vehicle Emissions National Environment Protection Measure
Preparatory Work: In-Service Emissions Performance - Phase 2: Vehicle Testing",
National Environment Protection Council, November 2000
6. Hildemann, L. M., Cass, G. R., and Markowski, G. R., (1989) "A dilution stack
sampler for collection of organic aerosol emissions: design, characterization and field
tests," Journal of Aerosol Science and Technology , Vol. 10, pp 193-204
7. TSI model 3936 SMPS Instruction Manual, 1999.
8. Reavell, K.S., Hands, T., Collings, N., "Determination of Real Time Particulate Size
Spectra and Emission Parameters with a Differential Mobility Spectrometer," 6th
International ETH-Conference on Nanoparticle Measurement, August 2002.
9. Collings, N., Reavell, K., Hands, T., Tate, J., "Roadside Aerosol Measurements with
a Fast Particulate Spectrometer," JSAE Paper 20035407, 2003.
10. Clark, N.N., Wayne, W.S., Buffamonte, T., Hall, T., Lyons,D.W. and Lawson, D.,
"The Gasoline/Diesel PM Split Study: Heavy Duty Vehicle Regulated Emissions"
Coordinating Research Council On-Road Emissions Meeting, San Diego, Ca., May
2003
173
-------
11. Clark, N.N., Wayne, W.S., Nine, R.D., Buffamonte, T. M., Hall, T., Rapp, B.L.,
Thompson, G., and Lyons, D.W., "Emissions from diesel-fueled heavy-duty vehicles
in Southern California," SAE/JSAE Spring Fuels & Lubricants Meeting , Yokohama,
Japan 2003, JSAE Paper 20030232, SAE Paper 2003-01-1901
12. Yanowitz, J., McCormick, R.L. and Graboski, M.S., "In-Use Emissions from
Heavy-Duty Vehicles," Environmental Science & Technology, 2000, Vol 34 pp.
729-740.
13. Sakurai, H., Tobias, H.J., Park K., Zarling, D., Docherty, K.S., Kittelson, D.B.,
McMurry, P.H., and Ziemann, P.J., "On-line Measurements of Diesel Nanoparticle
Composition and Volatility," Atmospheric Environment, Vol. 37, pp. 1199-1210,
2004.
14. Gautam, M, Thiagarajan, S., Burlingame, T., Wayne, W.S., Clark, N., Carder, D.,
(2004), "Characterization of Exhaust Emissions from a Catalyzed Trap Equipped
Natural Gas Fueled Transit Bus", 14th CRC On-Road Vehicle Emissions Workshop,
March 29-31, 2004, San Diego.
15. Kim, D., Gautam, M., and Gera, D., "Parametric Studies on the Formation of Diesel
Particulate Matter via Nucleation and Coagulation Modes," Journal of Aerosol
Science, Vol 33, pp. 1609-1621, 2002.
16. Mathis, U., Mohr, M., and Zenobi, R., "Effect of Organic Compounds on
Nanoparticle Formation in Diluted Diesel Exhaust," Atmospheric Chemistry and
Physics, 4, pp. 609-620, 2004.
17. Mohr, M., Jaeger, L. W., Boulouchos, K., "The Influence of Engine Parameters on
Particulate Emissions," MTZ Worldwide, 62, pp. 686-692, 2001.
18. Mathis, U., Mohr, M., and Zenobi, R., "Effect of Organic Compounds on
Nanoparticle Formation in Diluted Diesel Exhaust," Atmospheric Chemistry and
Physics, 4, pp. 609-620, 2004.
19. Lapin, C.A., Gautam, M., Zielinska, B, Wagner, V.O., McClellan, R.O., (2002),
"Mutagenicity of Emissions from a Natural Gas Fueled Truck", Mutation Research,
Vol 519, pp. 205-209.
20. Matz, G."Massenspektrometrische Bestimmung der Olemission im Abgas von Otto-
und Dieselmotoren," Technische Universitat Hamburg-Harburg, Zwischenbericht
iiber das Vorhaben Nr. 758 (FVV-Nr. 67580).
21. Weaver, C.S. and Klausmeier, R.F., "Heavy-Duty Diesel Vehicle Inspection and
Maintenance Study", California Air Resources Board Contract A4-151-32, Final
Report, May 16, 1988.
174
-------
22. Weaver, C.S., Balam, M., Mrodick, C.J., Iran, C., Heiken, J., and Pollack, A.,
"Modeling Deterioration in Heavy-Duty Diesel Particulate Emissions", EPA Contract
8C-S112-NTSX, Report, 1998.
23. Yanowitz, J; Graboski, MS; Ryan, LBA; et al. (1999) Chassis dynamometer study of
emissions from 21 in-use HD diesel vehicles. Environ Sci Technol 33:209-216.
24. Ramamurthy, R. and Clark, N.N., "Atmospheric Emissions Inventory Data for Heavy
Duty Vehicles", Environmental Science & Technology, Vol. 33, Pp. 55-62, 1999
25. Clark, N.N, Atkinson, C.M., Lyons, D.W., and Ramamurthy, R., "Models for
Predicting Transient Heavy Duty Vehicle Emissions", SAE Fall Fuels & Lubricants
Meeting, San Francisco, October, 1998, SAE Paper 982652.
26. Clark, N.N., Tehranian, A., Nine, R.D., and Jarrett, R., "Translation of Distance-
Specific Emissions Rates between Different Chassis Test Cycles using Artificial
Neural Networks", SAE Spring 2002 Fuels & Lubricants Meeting, Reno, NV, SAE
Paper 2002-01-1754
27. Kern, J., Clark, N.N., Nine, R. and Atkinson, C.M., "Factors Affecting Heavy-Duty
Diesel Vehicle Emissions", Jour, of the Air & Waste Management Assoc., Vol. 52,
Pp. 84-94, 2002.
28. Clark, N.N., Lyons, D.W., Rapp, B.L, Gautam, M., Wang, W.G., Norton, P., White,
C. and Chandler, C., "Emissions from Trucks and Buses Powered by Cummins L-10
Natural Gas Engines", SAE Spring Fuels and Lubricants Meeting, Dearborn, MI,
May 1998, SAE Paper 981393.
29. Gautam, M., Clark, N.N., Riddle, W., Nine, R., Wayne, W.S., Maldonado, H.,
Agrawal, A. and Carlock, M, "Development and Initial Use of a Heavy Duty Diesel
Truck Test Schedule for Emissions Characterization", SAE Spring 2002 Fuels &
Lubricants Meeting, Reno, NV, SAE Paper 2002-01-1753
30. Clark, N.N., Jarrett, R.P. and Atkinson, C.M., "Field Measurements of Particulate
Matter Emissions and Exhaust Opacity from Heavy Duty Vehicles", Journal of the
Air & Waste Management Assoc., Vol. 49, 1999, pp. 76-84.
31. Jarrett, R. P., 2000, "Evaluation of Opacity, Particulate Matter, and Carbon Monoxide
from Heavy-Duty Diesel Transient Chassis Tests", M.S. Thesis, Dept. of Mech. and
Aero. Eng., West Virginia University, Morgantown, WV.
32. Ganesan, B., and Clark, N.N., "Relationships between Instantaneous and Measured
Emissions in Heavy Duty Applications", SAE Transactions 2001, Journal of Fuels
and Lubricants, Vol. 110, Section 4, pp. 1798-1806
175
-------
33. Agrawal, A., Carlock, M., and Maldonado, H., "Development of Heavy-Duty Vehicle
Chassis Dynamometer Driving Cycles", 12th Coordinating Research Council On-
Road Emissions Conference, San Diego, CA, April 2002.
34. Personal communications on E55CRC-16 vehicle with Robert Graze, of Caterpillar
(Peoria, 111).
176
------- |