of                         for
       D.       Harris and Foy G. King
    U. S, Environmental           Agency
                                  Laboratory
                          Park, NC

INTRODUCTION
       Ambient  air                have  not
          to control               in the U,S,
as     the             pollutants, {1,2} Sulfur
oxides           monoxide          markedly
reduced by strict emission          for mobile
and                   The formation of ambient
ozone  is         in the urban atmosphere with
either  volatile  organic chemicals  fVQCsJ  or
nitrogen        (NOx)      the rate controlling
         depending  upon  the  relative  ratio
present. Much of the ozone reduction effort has
been  focused on  VOCs through  hydrocarbon
emission limits  from the same  sourc.es,  NOx
control           centered on  spark  ignition
vehicles using engine modifications and catalytic
tailpipe treatment. Modifications to compression
ignition engines have         to reduce these
emissions, but work is ongoing to make similar
emissions reductions from compression ignition
engines,
       Current        models rely upon emission
inventories for             to estimate effects
that control strategies could have  on future
pollutant   reduction scenarios.  Most  of the
inventory data are from samples  taken from the
sources  during  normal  operations.   Mobile
inventory       have  relied  upon laboratory
dynamometer   data   using   a  variety   of
fundamentally   and   experimentally   derived
conversion factors  to estimate in-use emissions.
Thts link has not beer, extensively confirmed due
to problems          emissions from m-use cars
    trucks.  This study endeavors to confirm or
modify those conversion factors by  building a
facility which will allow  sampling from heavy-
duty   diesel  vehicles  (HDOVs!    and  the
development   of   appropriate    models   for
operations related emission factors.


       The objectives of this project are to: 1}
define on-road emissions from HDOVs; 21
agreement    among   engine   and   chassis
                 Acurex Environmental Corp,
                       P.O. Box
                          Triangle Park, NC

dynamometers and on-road emission factors; 3)
evaluate  current   conversion   factors   for
dynamometer         develop appropriate ones
if needed;      4}         a modal  emissions
model.  The  on-road      component       an
instrumented 45-foot* cargo van traiier to
various  tractors  during  on-road  operation
according to a modal matrix,  These results are
          to those obtained during actual runs
over        routes. A separate road version of
a certification cycle is     to provide a link
to the dynamometer       The tractors will be
tested using a chassis dynamometer to check
          with that testing method using  urban
driving cycles, modal cycles, and  an
of the certification cycle. The engine from the
     and        dynamometer       truck will
be        to EPA's Office of Mobile Sources for
testing on their engine dynamometer  to directly
compare with the on-road      *or the
certification  cycie and to check against  the
original        Test Procedure {FTP} certification
data.
       Our experimental  approach involves on-
road testing according to the test matrix shown
in table 1.  This test  sequence  develops  the
modal    necessary to model the emissions as
a  function  of  speed,   load,   grade,   and
acceleration. These      are  fitted to  each of
three  routes  representing;    11  trans-urban
interstate travel,  21  terminal entry'exit, and 3^
urban delivery. The actua! routes are run, and the
emission factors generated  are compared to the
       model  results,  A final test  sequence
utilizes  a certification cycie  that  has  been
        to suit actual truck operational capability
to develop data  for direct comparison  with the
dynamometer
       The  modal experimental  design utilizes
four-lane iimited access highways with  Song and
relatively straight sections at three grades. These
* Metric equivalents appear s>. the «nci of this paper

-------
                        Table 1:  General Test Matrix {Each Tractor)
                    Slow               15 MPH                   {constant speed)
Speed               Medium             35 MPH                   (constant speed)
                    Fast                55 MPH                   (constant speed)


Accleration                Normal                    Shift gears as governed
0 to 55 MPH               Short Shift                 Shift @ 80-90% governed engine speed


                    Level               0%                       U.S. 70 near New Bern, NC
Grade               Moderate            2-3%                      1-26 near Hendersonvitie
                    Steep               5-7%                      1-26 near Hendersonviile


Downhill             Steep               5-7%                      Load variation only


                    Light               Empty Trailer               (25000 - 30000 Ib GCW)
     (GCW)          Medium             % Max Cargo Load          (up to  55000 ib GCW)
                    Heavy              Max Cargo Load            (up to  80000 Ib GCW}
{3 Speeds + 2 Accelerations)*(3 Grades » 3 Loads) -f  Downhill = 46 Triplicate Tests Maximum
highways provide  a  safe operating  area for
sometimes very  slow moving vehicles  when
testing is confined to low traffic load times. The
extent to which the  matrix is completed depends
upon  the  vehicle power  available to pull some
loads up grades as steep as 7%. A coast-down
test per White and Korst {3} is  run for  each
vehicle  to develop the power requirements for
each configuration.


      Two  preliminary  test  sequences  were
conducted to develop  familiarity with diesel
engine testing and  specific test procedures, A
diesel powered  generator set was  tested  to
familiarize personnel with the effects of load on
monitored parameters and establish the ranges of
these parameters, A bank of resistive loads could
be  switched  to  simulate the ioad-basad test
procedure, A more complete development phase
was conducted using a diesel powered  pickup
truck and a horse trailer  to simulate the HDDV
tractor-trailer.
      The first  step  was   to  Identify the
appropriate road  sections to conduct the test
matrix.  The  North  Carolina  Department  of
Transportation fNCDOT)  provided detailed site
plans of  sections  which met  our  grade and
distance  requirements  as  weil  as  practical
recommendations to guide our selection, Leve!
grades were easy to find near the Atlantic coast
of North  Carolina,  and a 0.0%,  20-mile long
section of U.S. 70 near New Bern was selected.
This has the added advantage of including a large
rest area  which we could use for  short term
staging. The 3 and 7% grades are numerous but
usually too short to  meet our requirement for at
least 60 seconds of  continuous data acquisition.
The NCDOT located a stretch  of  Interstate 26
south of  Hendersonviile which  includes both
slopes. It too has a rest area nearby for staging.
      The development of the sampling and
analysis system was based upon designs     by
the authors during stationary source monitoring
projects.  A  continuous  emission  monitoring
(CEM) bench was assembled from backup units
available from ongoing laboratory projects. This
short rack-mounted  unit was set into the horse
trailer on instrument shock mounts. A computer-
based data acquisition system operating labtech

-------
Notebook  software  was   mounted  on  the
instrument rack with       shock padding. The
pickup was wired  with sensors, and a heated
sampling  line  connected  the tailpipe  to the
sample conditioning system in the trailer storage
bay. Since these instruments operate on 110 volt
AC,  a gasoline-powered generator was mounted
in the pickup bed. The monitored parameters,

      Table 2:  Monitored Parameters
   Emissions Msamireaeats   Operational
  o, (*>

  CO, !t>

  CO (ppm)

  HO, fppm)

  WC fppai

  Exhaust Flow {scfffl)
                       Vehtcte Speed inpht
togiae $p«d Irp

Bat: G- Force        —-iiL
                                                            •
                                                X    JL
                                                , TWO i   |0»
                                                               !
                                                              KOia
    LI
                                           Computerized
                                           Daft; Acq
                                           System
                                                                 T      I
                                   Figure 1:  Pickup Test Facility Schematic

                          (smoke meter) at the end.  A sheath air system
                          was  added  to  the  smoke  rneter  to  limit
                          contamination on the lenses. The sample port is
                          closely coupled to an electrically controlled ball
                          valve which allows either exhaust or calibration
                          gas to flow through the entire  sampling system.
                          This provides total system quality assurance
                                  Stack Measurements
                                                 Hi
                                 Engine Measurements ,«•«« -^i"S?
                                               4*ara-:- £^BXiM-.?
                                                O         h=-     S^
                                        Operational yeasurements

                                        i  Ata»et*n,t.
                               Computerized
                               Data Acquisition
                               System
                                                  Exhaust Sample Measurements
— Mnc OO* by NOR, fpn
 - NOs bf OwAimhssoinM, ppm
                                Figure 2:  HDDV Facility Schematic

-------
          3: HDDV Sampling Trailer Layout

(QA)  checks in  addition to the individual CEM
calibration and QA tests. A heated Tefion-iined
sampling  line connects the sample  port to a
heated filter and pump mounted on the front of
the trailer. A short,       tine connects to the
instrument rack,  and the sample remains heated
to the total  hydrocarbon and chemiluminescent
NOx analyzers, while a parallel sample is cooled
and dried using a PerrnaPure dryer for the non-
dispersive infrared (NDIR) analyzers.
      The  rack-mounted industrial  computer
data acquisition system uses Labtech Notebook
and  Lotus  1-2-3  to  record   calibration  and
measurement data. Some time alignment of the
      from  the individual  analyzers  is  done
manually due to differing response times of the
instruments.  The  computer   was   recently
upgraded to a Pentium chip to handle the large
quantities of data generated by the test facility.
       A second trailer  is  used  as  a support
facility. It contains a smaii office/laboratory area
and  an      for storing the weights not being
used in the test facility. The two trailers are
parked back to back In the  staging area, and an
electric pallet jack is used to move the weights
between the two  trailers. Permanent staging
      have been set up near the remote test sites
with electric power to operate the facilities.


       The first tractor being tested is a 1989
Ford L9000 Cab-over with  a 315 HP Cummins
NTC engine. Modal testing has been completed.
The  results  for the NOx  emissions  for  each
constant speed, load,  and  grade condition are
presented in figure  4. Little  effect  is noticed
except for the low speed runs where significantly
elevated levels are found. The  grams/brake
horsepower-hour  {g/BHP-hr) data  are plotted  in
figure 5 as  a function of the estimated engine
    m, Hwwswwi   m, i^wns     3®    1$, **w*s   s«, {^WM
                                                       SS^H^KP**  «,Lo-W%*   M    «,*«»^ IS.isM^W
                                                                    M    tft, HM%Hi*  1ft .js-^I^M

           C«i®»M Sp«»d C««*m®«*, mph                               C«anM»iS Sp®®d Coml»m* mph
                Figure 4; Average Nitrogen Oxides Emissions by Operating Mode

                                               4

-------
                              — » — *
               •M  «V» — —-—•""
                               1WS    120G    1*W
     Figure 5; Nitrogen Oxides Emissions Versus Engine Torque
torque  required  by the  truck. The
certification  level  of  7.88  g/BHP-hr  for  this
engine is indicated on this  figure with most of
the data  below that level  except for the  low
torque operating conditions. Analysis of these
     is continuing     will be reported in a later
paper.
       As part of a cooperative agreement with
the North Carolina Truck Driver Training Program
at Johnston Community College (JCC), we have
conducted acceleration studies to determine the
effect, if any, progressive shifting has on vehicle
emissions.   Progressive  shifting   limits  the
maximum  engine  speed  (in  revolutions  per
minute,  rpm)  reached  in  each  gear constant
before shifting to the next successive gear. The
normal  procedure,  acceleration   to  maximum
governed rpm in each gear, is  compared to the
progressive method in figure 6, No definite trend
can be identified in this initial test sequence. The
small circle plotted on this graph  represents the
total emissions from that vehicle at a constant
55 mph for the distance it took to accelerate it
to 55  mph.  This  suggests that  limited access
bypasses reduce emissions when  compared to a
bypass with  stoplight traffic controls.
       figures 7 and 8 present examples of the
data traces from  actual runs.  The acceleration
run in figure  7 shows that the CEIVls and engine
sensors  readily follow the changes in engine
operating conditions as the vehicle goes through
the gears. A section of urban delivery route is
shown in figure 8. The bottom line displays the
total  g-load  which is  the sum of the forward
acceleration  vectors due to gravity and velocity
change. This line shows the spikes as the vehicle
accelerates in each gear, and the corresponding
                                                 Zero to 55 MPH Acceleration Emissions
                                                      Sm»t» E
-------
                TnckSpatf
          Figure ?: Fully Loaded Truck Acceleration
   60

   SB

   id
 !-•   ..!/•
       Pollution,   National  Academy  Press,
       Washington, DC, 1991.

2,     Southern Oxidants  Study 1995.  The
       State  of The Southern Oxidants Study:
       Policy-Relevant  Findings   in   Ozone
       Pollution Research, 1988-1994. Prepared
       by the Southern Oxidants Study, North
       Caroiina State University, Raleigh, NC.
       April 1995,

3.     White,  R,   A.,  Korst,   H,   H.,   "The
       Determination    of    Vehicle    Drag
       Contributions from Coast-Down Tests/*
       SAE 720099,  Society  of Automotive
       Engineers, Warrendaie, PA, 1992.


       Readers more familiar  with  the metric
system may use the following  equivalents  to
convert to those units:
       5/9(°F-32} = 1°C
       0.305 ft =  1  m
       0,0283 ft3 = 1 m3
       746 hp = 1  W
       0,454 tb =  1  kg
       1.609 mi = 1 km
     F:gur* 8: Test Data from S«ction of Urban Delivery Routs
this program. The results presented are specific
to the  engine/tractor  combination  that was
tested, and may not be typical of the fleet as a
whole.   Nonetheless,  the  remaining units  will
cover all major engine manufacturers as well as
represent the  range of emission standards the
engines were designed to meet through the 1994
standard. This project is expected to continue for
2 more years.

REFERENCES
1.     National Research Council, Rethinking the
       Ozone Problem In Urban and Regional Air

-------
                           OF THE


                LaJolla, California



                 July 24-27, 1995
                                   NO
                            BC
     or
                              NO,
O
 O
    by:



of

            for



          of

-------