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 125ฐF 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 52ฐC  (125ฐ  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 ฐF (+/- 10
ฐF). 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 FฃHDDT_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 ฅ 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 46ฐ C (115ฐ 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 52ฐC 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 fป
ฃ> 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) 10ฐ
                                                                       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
                                                                          =
                                                                          ฎ
                     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**
ป ป/• .ซ
                                                    = 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
   '•ฃ  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
 '•ฃ  2500
 ฃ
    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
    ฐ 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
 ฃ• -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.
ซ
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

ฃ
 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
                                            ง0

                                   + 8.00E+04 ง

                                            Z
                                   - 6.00E+04 |
                                            z
                                            .2!
                                   + 4.00E+04 ^
                                            ฃ

                                   -- 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 | ฃ



                                           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;
ฃ 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
    ฃ  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
ฃ 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/Cฐ2)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
   .ซ
   '
-------
(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  ป   ป.   •
                                      *'  ^  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.
  ง
  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^wซt~ooovo-r,n^wซ









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
     ฃ
    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

-------