EPA/AA/TDG/93-06
                         Technical Report
  Evaluation of Three Catalysts Formulated for Methane Oxidation
                   on a  CNG-Fueled Pickup Truck
                                by
                      Gregory K. Piotrowski
                        Ronald M.  Schaefer
                          December 1993
                              NOTICE

     Technical  Reports do  not necessarily  represent  final  EPA
decisions or positions.   They are intended to present technical
analysis of issues using data which are currently available.  The
purpose  in the  release  of  such  reports is  to  facilitate  the
exchange of  technical  information and to  inform the  public of
technical developments  which  may  form the basis  for  a final EPA
decision, position, or regulatory action.

              U. S. Environmental Protection Agency
                   Office of Air and Radiation
                     Office  of Mobile  Sources
           Regulatory Programs and Technology Division
                   Technology  Development  Group
                        2565 Plymouth  Road
                       Ann Arbor,  MI   48105

-------
        UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

                     ANN ARBOR. MICHIGAN 48105
                            DEC 22 1993                     OFFICE OF
                                                      AIR AND RADIATION
MEMORANDUM


SUBJECT:  Exemption From Peer  and Administrative  Review


FROM:     Karl H. Hellman, Chief     l//^
          Technology Development Group


TO:       Charles L. Gray, Jr., Director
          Regulatory Programs  and Technology  Division


     The attached report  entitled "Evaluation of Three Catalysts
Formulated  for  Methane Oxidation on a  CNG-Fueled Pickup  Truck,"
(EPA/AA/TDG/93-06) describes the emission results obtained from the
evaluation of three specialized methane catalysts supplied by three
different catalyst  manufacturers.    The  catalysts were  evaluated
using a CNG-fueled Dodge Dakota pickup truck.

     Since this report is concerned  only with the presentation of
data and its analysis and does not  involve matters of  policy or
regulations, your concurrence is requested to  waive administrative
review according to the policy outlined in your directive of April
22, 1982.

Concurrence:
                                     Date:
Cjiarles L. Gray/ Jr. , Director, RPT

Attachment

-------
                        Table of contents
                                                             Page
                                                           Number
I.   Summary	1

II.  Introduction	2

III. Description of Catalytic Converters 	  2

IV.  Description of Test Vehicle	7

V.   Test Facilities and Analytical Methods	9

VI.  Test Procedures	10

VII. Discussion of Test Results	11

     A.   Electrically Heated Methane Catalyst 	  11

     B.   Main Underfloor Catalysts	15

VIII.Conclusions 	  21

IX.  Acknowledgements	21

X.   References	22

APPENDIX A - Fuel Properties/Test Data Sheet	A-l

-------
I.   summary

     Three fresh  catalysts  formulated for methane oxidation were
evaluated by U.S.  EPA  on  a  compressed natural gas (CNG) vehicle.
The  first catalyst  was an  electrically heated,  compact,  quick
lightoff converter, provided by W. R. Grace Company.  The other two
catalysts were  larger  in  volume and were similar  in size to the
stock main catalyst on the test vehicle.  These two  larger volume
catalysts  were  provided  by Kemira  Oy  and  AlliedSignal.    No
durability miles were accumulated on any of the catalysts.

     The  CNG-fueled  test  vehicle was a  1991  Dodge Dakota pickup
equipped  with  a 318-CID  engine  provided by  Stewart & Stevenson
Power, Inc.   Although the  truck was  equipped for dual fuel (either
gasoline  or  CNG)  operation,  all  the testing  described here was
performed with CNG fuel operation over the Federal Test Procedure
(FTP) driving cycle.

     The Grace quick lightoff catalyst was tested without the use
of a  larger  main catalyst downstream.   The  effect of electrical
heating and secondary air  injection  upstream of this  catalyst were
also  investigated.   The  two  larger  volume  catalysts  were also
tested in an underfloor  location.   The effect  of secondary air
assist was also evaluated on these two catalysts.

     The  low-volume Grace  catalyst caused  substantial emission
level  reductions  in  non-methane  hydrocarbons   (NMHC),  carbon
monoxide  (CO),  and oxides of nitrogen  (NOx).   With 20/40-second
heat assist (resistive heating applied for 20  seconds  prior to key-
on and 40 seconds after key-on in Bag 1 only) ,  NMHC  and CO emission
levels  were measured  at  0.10  grams/mile  and  11.6  grams/mile
respectively over  the  FTP cycle,  over 60 percent reductions from
engine-out levels.  The lowest NOx levels were obtained without any
catalyst  assist and were  measured at 0.8 grams/mile  over the FTP
cycle, over a 50 percent reduction from engine-out levels.

     The  two  larger  volume  main   catalysts  both  resulted  in
emissions over  the FTP cycle  that  were below the levels  of the
California ultra-low emission  vehicle (ULEV)  standards, with and
without secondary air assist.  The test  data taken  for this report
did not permit the calculation of  non-methane organic gases  (NMOG)
as required  by  the California regulations, nor  was  a Reactivity
Adjustment Factor  (RAF) applied to the data.   Nevertheless, it is
the opinion  of  the authors  that the  test results  are low enough
that a 0.04 grams/mile NMOG  level would be met at zero miles with
a  fresh  catalyst.   The  Kemira  Oy  catalyst  without  air  assist
resulted in NMHC/CO/NOx emission levels  of 0.02/0.4/0.2 grams/mile
over  the FTP.    These emission  levels  were  attained at zero
accumulated system  miles.   The AlliedSignal  catalyst (smaller in
volume than the Kemira Oy unit) resulted in similar  low levels of
0.01/0.8/0.2 grams/mile without air assist.  Secondary air assist
provided little additional emission  reductions from the unassisted
catalyst  levels for both units.

-------
                                -2-
II.  Introduction

     The  current U.S.  highway vehicle  fleet is  almost totally
dependent on  petroleum-based  (i.e., gasoline  and Diesel)  fuels.
Alternatives  to  petroleum-based  fuels  may  become  increasingly
important in future years because of their potential contribution
to a solution for air quality problems as well as  a means  to lessen
the demand for imported oil  in the U.S.  [1,2]


     One  candidate  for serious  consideration as  an alternative
motor  vehicle fuel  is compressed  natural  gas   (CNG).   CNG  is
composed primarily of  methane  (CH4) , but it  may  contain  up  to 10
percent higher  weight  hydrocarbons  (mostly ethane,  propane and
butane). [3]  Ozone-forming  photochemically reactive fuel-related
emissions  from  CNG  vehicles  consist  primarily  of  non-methane
hydrocarbons (NMHC).  CNG-fueled vehicle exhaust is 90-95 percent
methane,  a  relatively non-reactive   hydrocarbon  (HC)  species.
Gasoline vehicle exhaust HC consists of 65-95 percent more reactive
non-methane hydrocarbon species, however.  [1]  It  is estimated that
CNG-fueled vehicles may have 36-93 percent lower volatile organic
compound  (VOC)  emissions,  determined  on a reactivity-equivalent
basis,   than   gasoline-fueled   vehicles   (depending   on   the
configuration of the CNG vehicles  as either dual-fuel or  optimized
for CNG).  [l]

     The large methane  fraction of  the HC exhaust emissions from
CNG-fueled vehicles  is a problem for engine  designers, because
methane is the most difficult  HC  to oxidize catalytically.   Some
vehicle  emission tests  conducted  at  General  Motors  Research
Laboratories showed that poor methane conversion on lean calibrated
CNG-fueled vehicles occurred over the FTP  cycle when commercial
three-way catalysts were used.  [4]  Recently further General Motors
research  indicated  good  methane  conversion  occurred  when  a
palladium/alumina catalyst  was used together  with slightly rich
feedstream conditions.  [5]  Other catalyst development work to date
suggests  that with  proper  engine controls,  selected  catalyst
formulations can  effectively control  methane emissions from CNG-
fueled vehicles.  [6]

     EPA's Technology  Development  Group routinely  conducts and
publishes the results  from emission control technology evaluations
to spur  further interest in new technologies.   Recently,  three
catalyst companies  furnished  EPA with samples  of fresh,  unaged
catalysts specifically formulated for use on CNG-fueled vehicles.
These catalysts were evaluated on  a  CNG-fueled  pickup truck at the
EPA National Vehicle and  Fuel  Emissions  Laboratory.   The results
from this evaluation are presented in this report.

III. Description of Catalytic Converters

     The  first   converter  evaluated  here was  a  quick  lightoff
electrically  heated  catalyst  supplied by  Grace  Company.    This

-------
                              -3-
single-segment,  metallic foil substrate had a  total  volume of 0.22
liters (Figure 1) .   The  energy required for electrically hearing
this catalyst was supplied from a dedicated 12-volt, 115 amp-hour,
deep-cycle battery.   The current delivered to the catalyst averaged
about 500 amps during the time of resistive heating.  A switch and
engine starter motor relay were used  to  supply  energy  from the
dedicated battery when  desired.   Table  1  contains  more detailed
specifications of  the Grace  electrically  heated catalyst.   The
catalyse  pictured   here   is   much   smaller  in   volume  than  a
conventional main converter and was  not designed  as a substitute
for a main underfloor catalyst.
                             Figure  l
               Grace  Electrically Heated  Catalyst

-------
                               -4-
                             Table 1
            Grace Electrically  Heated Methane Catalyst
                 List of Detailed Specifications
Catalyst diameter

Catalyse length

Catalyst volume

Total substrate weight

Cells per square inch

Noble metal loading

Ratio of noble metals
 Specifications

    2.51 inches

    2.75 inches

      13.24  in'

      335 grams

            160

        30 g/ftj

?t:Pd:Rh  0:1:0
     The second converter  evaluated  was provided by Kemira Oy of
Finland.  This unit was  cylindrical  in shape and had the  largest
volume (4.71 dmj)  of the three converters evaluated.  This converter
was designed for  heavy-duty truck applications utilizing CNG fuel.
This converter utilized  a  single metallic foil  substrate and was
mounted in the underfloor  location for our testing.  A picture of
this  unit   is  provided  as Figure  2;  a  more  detailed  list  of
specifications for the Kemira Oy catalyst is provided in Table 2.
                             Figure 2
                    Kemira Ov Methane  Catalyst

-------
                              -5-
                             Table 2
                    Kemira Oy Methane  Catalyst
                 List of Detailed Specificationa
Prototype identification number
  Total weight of converter
Substrate;
  Diameter
  Length
  Volume
  Cells per square inch
  Total surface area
  Cross section
  Precious metal loading
  Total precious metal
  Precious metal ratio
  Foil material
  Foil thickness
  Weight of substrate
  Shell material
  Shell thickness
  Specification
           5382
          7300g

          200mm
          150mm
       4.71 dm3
            600
         18.8m2
       314.2cm2
      1.77g/dm3
          8.34g
1:1:0  Pt:Pd:Rh
       W 1.4767
         0.05mm
          4118g
       W 1.4512
          1.5mm

-------
                              -6-
     -he last converter evaluated in this program was supplied by
AlliedSignal  Inc.   This  catalyst had  a total  substrate volume
between the two other units (2.78dnr) and was similar in size to the
stock catalyst on the  truck.   This  catalyst utilized two  ceramic
substrates with similar sizes and loadings and was also evaluated
in  the  underfloor  location.    A  picture of  this  converter  is
provided  in Figure  3   below,  and  a  detailed  list  of  catalyst
specifications is provided in Table 2.
                             Figure 3
                  AlliedSianal Methane catalyst
                             Table 3
                  AlliedSignal Methane Catalyst
                 List of Detailed Specifications
  Material

  Total catalyst volume

  Cross section (oval)

  Length

  Cells per square inch

  Noble metal loading

  Ratio of noble metals
        Ceramic

      169.66  in3

   4.5 x 7.0 in

         3.4 in

            400

       100 g/ft3

0:10:1 Pt:Pd:Rh

-------
                               -7-
IV.  Description of Test Vehicle

     The test  vehicle used for this  evaluation  was a 1991  Dodge
Dakota pickup truck.  The truck was converted to CNG  operation  by
Stewart & Stevenson Power, Inc. of Commerce City, Colorado,  using
their patented Gaseous Fuel Injection (GFI)  system.  The  truck had
dual-fuel (gasoline or CNG) capability; fuel selection  was governed
by a switch inside  the passenger compartment.  Figure 4 depicts the
GFI system.
                             Figure 4
                  Gaseous Fuel Injection System
                TYPICAL INSTALLATION
         Battery Power —

        Switched Power —

         Fuel Injector —

        Starter Solenoid —

         Colt Negative —

         Knock Sensor —

          Fuel Gauge —

      TDC or Tachometer —

        Intake Air Temp —

      Manifold Skin Temp —

        Selector Switch —

        Oxygen Sensor —
Figure A1: NGV System Schematic

   High Pessure Fuel Line
                      Metering Valve
                      and Computer
                                    Pressure
                                    Regulator
    Fuel
    Nozzles
MAP
Vacuum
Line Tee
                         GFL
                               Engine
                               Coolant
     High pressure  CNG was supplied  from  storage cylinders to  a
single-stage regulator where the pressure was reduced to 100 psig.
Two CNG storage tanks were located inside the passenger compartment
and were capable of filling to approximately  3,500 psig pressure.
EPA, however,  filled  these  tanks only  to a maximum 2,500 psig
pressure when refueling.  With this fill level,  it was possible to
conduct approximately  four tests  over the FTP  cycle before the
truck was refueled.  EPA also installed two excess flow valves at
the outlet of each storage tank  for safety purposes.

-------
     One hundred psig fuel passes from the pressure  regulator to a
primary fuel filter and then to a regulator referred to as  a compu-
valve.   The compu-valve commands  the appropriate combination of
solenoids and injectors to operate for the required  period of time
necessary to meter  the  desired  amount of fuel.   The metered fuel
then passes from the metering  valve to spray  nozzles in the intake
system of the engine,  in this  case the air cleaner.   These nozzles
introduce fuel into the  intake air stream and promote nixing of the
inlet air and fuel.

     The GFI system used speed density calculations  for both fuel
flow and air flow and solved for the correct air/fuel mixture based
on  a fixed  (software) stoichiometric value of the  fuel consumption
in the area of operation.   Speed  density  calculations are  based on
engine speed and the temperature and pressure of the inlet air and
fuel.  The GFI system senses manifold absolute pressure, barometric
absolute  pressure,  and fuel  absolute  pressure.   Temperatures
monitored include intake air temperature,  manifold  skin temperature
(air  temperature at  engine  intake  valve),  and   fuel  regulated
temperature.  The first two sensors are remotely  located,  and the
last is embedded in the metering valve.

     Figure  5  is a picture of  the test  vehicle.    The  truck was
tested  at  4,750  Ibs   equivalent  test  weight  and  13.0 actual
dynamometer  horsepower.   The vehicle was loaned  to U.S. EPA by
Stewart & Stevenson Power, Inc.
                             Figure 5
               CNG-Fueled  Dodge Dakota Test Vehicle

-------
                              -9-
V.   Teat Facilities and Analytical Methods

     Emissions -testing at EPA was conducted on a Clayton model ECE-
50 double-roll chassis dynamometer, using a direct-drive variable
inertia flywheel unit and road load power control unit. The Philco
Ford constant volume sampler has a nominal capacity  of 600 standard
cubic feet per minute.   NOx emissions were determined by a Beckman
Model 951A chemi luminescent NOx analyzer.  Methane  (CH4) emissions
were  measured  with   a   Model  8205   Bendix  methane  analyzer.
Hydrocarbon  emissions  were determined using a  Beckman  Model 400
flame ionization detector.  CO  and C02  were measured using a Bendix
Model 8501-5CA infrared analyzer.

     Procedures were developed for the  determination of both total
and  non-methane hydrocarbons  and fuel  economy for natural  gas
vehicles based  on the properties  of  the natural  gas  fuel.   The
natural gas  used  to refuel the test vehicle was obtained from a
commercial pipeline located near the EPA motor vehicle laboratory.
This fuel was not specifically analyzed,  but it was assumed to be
typical of natural  gas used in the Detroit  area for home heating
applications.  The properties used in these calculations (Appendix
A) were provided to EPA  by a local gas company.  Appendix A also
presents natural gas fuel economy and  a  gasoline equivalent fuel
economy value.

     The properties of the natural gas fuel included mole percent
compositions  of  nitrogen, carbon dioxide, helium, hydrogen,  and
hydrocarbons.  These values were used to calculate specific gravity
and  net  (lower)  heating  value.    This   data  was  then  used  to
determine the properties seen  in Appendix A such as:

     1.   Composite H/C ratio for the total hydrocarbon components
in the fuel, H/CTHC;

     2.   Composite  H/C  ratio for all  non-methane  hydrocarbon
components in the fuel,
     3.   Non-methane carbon weight fraction (CWFNMHC)  of the fuel,
grams of carbon per gram of non-methane hydrocarbon, excluding C02
and inert gases not consumed in the combustion process;

     4.   Mass fraction,  grams  of  fuel per gram of carbon, where
carbon is based  on carbon in hydrocarbon components  (and C02)  in
the fuel, g NGV fuel/g C  in NGV fuel;  and

     5.   Energy density  of the fuel  in BTU's  per gram of fuel,
expressed as net (lower)  heating value, BTU/g NGV  fuel.

     From  these  properties,  corrected  total  and  non-methane
hydrocarbon values were obtained based on the fuel's components in
their proper  mole  fractions.  Calculations  for CO,  C02,  and NOx
exhaust  emissions  were  unchanged  from  procedures described  in
Section 86.144-78, Title  40, of the Code of  Federal Regulations.

-------
                              -10-
Specific  calculations for  determining adjusted  total  and  non-
methane  hydrocarbon emissions  were detailed  in  a  previous  EPA
technical  report  describing  testing  of  several  natural  gas
vehicles. [7]

VI.  Test Procedures

     The goal of this program was the evaluation of three catalytic
converters on a natural gas fueled truck.  Three separate catalyst
manufacturers   supplied   EPA   with  prototype  converters   for
evaluation.

     The first catalyst evaluated was an electrically-heated quick
lightoff  converter supplied  by the W.  R. Grace  Company.   EPA
conducted this  evaluation with  and without catalyst heat and air
assist.   Resistive heating  was applied to the catalyst  for  20
seconds prior to key-on in the Bag 1 portion of  the FTP and for 40
seconds after key-on  in Bag  1.   No resistive  heat  was applied to
the catalyst for the remainder of the FTP.   The effect of secondary
air  assist  when combined with the  resistive heating  strategy
described previously was also investigated.  This air  assist period
was 60 seconds after key-on in the Bag 1 portion of the FTP only.
The  air  assist time  was  limited  to  60  seconds  so  that  the
conversion of NOx  emissions would not  be  inhibited.   Air flow to
the catalyst was kept  constant at 5 standard cubic feet per minute
and was provided by a shop air hose.

     The  two  additional catalysts did not arrive at  EPA until
approximately six months after the testing with  the Grace catalyst
was complete.   No emission tests  were performed with the Dakota
truck  during  this  six-month  period; the  truck was  started  and
driven periodically to keep the battery charged.

     The second catalyst evaluated was provided to EPA by  Kemira Oy
of Finland.   Three different configurations of this catalyst were
evaluated.   A straight pipe was inserted  in place  of the stock
underfloor  catalyst to  enable  the determination  of engine-out
levels.   The  engine-out  levels  measured  during this testing were
significantly different than  those measured prior  to testing the
Grace  catalyst,  so they are reported separately.   The Kemira Oy
catalyst  was  then installed on the  truck  and tested without any
assist.   The last configuration utilized  the Kemira Oy catalyst
with 60 seconds of air assist after start in Bag 1.  Again, the air
assist rate was kept constant at 5 standard cubic feet per minute.

     The  third catalyst evaluated was supplied by AlliedSignal
Inc.,  located  in Tulsa,  Oklahoma.  This catalyst was tested using
procedures  similar to those  used  for  the  Kemira  Oy  catalyst
evaluation.      The   engine-out  emissions   measured   prior   to
installation  of the AlliedSignal  catalyst  were similar to those
obtained  before  the Kemira Oy testing.

-------
                               -11-
VII. Discussion of Test Results

     A.   Electrically Heated Methane Catalyst

     The results from the  EHC evaluation are reported separately
because of its much smaller catalyst volume than the others tested.
The engine-out emissions measured  here differed  significantly to
those reported in the next  section.  There was a substantial delay
between this testing  and the testing described in the next section.
The direct  effect of  this  time delay  on  vehicle hardware  and
emissions is not  known.  It is also possible that the difference in
elevation at  the Stewart &  Stevenson  facility in  Colorado (where
the vehicle was converted)  and the EPA laboratory in Michigan may
have  contributed to  the  change in  engine-out  emission  levels
experienced here over time.

     Resistive heating and  air assist were used only during the
cold-start portion of Bag 1 of the FTP.  The first 505 seconds of
the FTP are referred to as Bag 1; the cold-start portion consists
of the initial minutes of Bag 1 during which the engine and exhaust
system  heat  to   a   relatively  steady-state  temperature.    The
following discussion comments  on differences  in  exhaust emission
levels which may  be related to catalyst reactions, heat assist, or
air assist.   All testing was conducted at an ambient temperature of
728-73°F.  Bag 1 emission levels are given in grams over the test
segment (Bag 1) .  Composite FTP emissions are reported  in grams per
mile.

     Figure  6 presents Bag  1 total hydrocarbon emission levels
obtained during  testing with  the  Grace catalyst.    "Engine-out"
levels  were  obtained  with no  catalyst present  in  the  exhaust
system.    "Stock Catalyst"  represents  testing  with the  stock
catalyst in  the  exhaust  system.    "No  assist"  represents  testing
with only the low-volume  Grace electrically  heated catalyst in the
underfloor  location  without any catalyst assist.   "Heat  assist"
utilized a 20/40-second resistive heat assist.  "Heat & air assist"
utilized both a 20/40-second resistive heat  scheme and a 60-second
air assist after start in Bag 1 only.

     The stock catalyst  caused  a  42  percent reduction in  Bag l
total hydrocarbons from engine-out levels, from 7.44 grams  to 4.34
grams.  The small-volume  Grace catalyst without any assist lowered
total hydrocarbons to 5.36  grams  over  Bag 1.  When heat assist was
applied  to  the Grace  converter, Bag  1 total hydrocarbon  (THC)
levels were similar to those measured with the larger  volume stock
catalyst at  4.53 grams,  a  40 percent reduction from engine-out
levels.

     Combined catalyst resistive  heating/air assist caused  a slight
increase in  Bag  1 THC to  4.66 grams.    This  was unexpected;  the
addition of air over the catalyst may have caused the catalyst to
cool slightly, delaying lightoff.  It  is  not known how significant
any cooling effect may have been, as catalyst skin temperature and
inlet/outlet  air temperatures were not monitored.

-------
                              -12-
                              Figure 6
             Grace Methane Catalyst Evaluation
         Bag 1  Total Hydrocarbon  Emission Levels
Catalyst Configuration

            Engine-out

         Stock Catalyst

         Grace Catalyst
             No assists^
            Heat assist^
        Heat & air assist^
            7.44
4.34
    5.36
 4.53
 4.66
                             2468

                              Total Hydrocarbons (grams)
                     10
     Figure 7 presents Bag 1 carbon monoxide  (CO) emission levels
using  the Grace  catalyst  for the  same catalyst configurations
described above.  The stock catalyst was again very effective at
reducing Bag 1 CO levels from engine-out levels,  from 145.0 grams
to 32.9 grams.   Bag 1 CO levels were measured at 80.3  grams, almost
a 45 percent reduction from engine-out  levels  with the unassisted
Grace  EHC.  Resistive heating  further lowered Bag 1 CO levels to
51.0 grams, a 65 percent reduction  from engine-out levels.  Again,
adding air assist to resistive  heating had a detrimental effect on
Bag  1  CO  emissions.   In this  configuration, Bag  1 CO levels were
measured at 57.6  grams,  13 percent higher  than resistive heating
alone.    Each  configuration evaluated,  however, had  higher  CO
emissions than the larger volume stock  converter.

-------
                               -13-
                              Figure 7
             Grace Methane Catalyst Evaluation
         Bag  1  Carbon Monoxide Emission Levels


Catalyst Configuration
            Engine-out

         Stock Catalyst
         Grace Catalyst
                            145
             No assist
            Heat assisl
        Heat & air assist^
32.9
            80.3
      :57.6 i
                     0   20   40   60   80   100  120  140  160

                              Carbon Monoxide (grams)
      Table  4 presents  individual  bag emission levels  with  the
  small-volume Grace EHC.  Methane (CH4) and non-methane hydrocarbons
  (NMHC) followed a trend similar to those discussed  in Figure 6 for
  total hydrocarbons.  Oxides of nitrogen  (NOx) emissions were also
  reduced  significantly  over  Bag  1  with  the  unassisted  Grace
  catalyst, from 6.4 grams engine-out to  2.6 grams. Resistive heating
  slightly increased Bag  l  NOx  levels to  3.1 grams.   There  was no
  noticeable change when heat and air assist were combined, compared
  to  heat-assist-only levels when considering NOx.

-------
                               -14-
                             Table 4
. Grace Methane Catalyst Evaluation
Individual Bag Emission Levels (grams)
Catalyst
Configuration
THC
CH4
NMHC
CO
C02
NOX
Bag 1:
Engine-out
Stock catalyst
No assist
Heat assist
Heat & air
assist
7.44
4.34
5.36
4.53
4.66
6.18
4.34
4.73
4.11
4.22
1.26
**
0.62
0.43
0.44
145.0
32.9
80.3
51.0
57.6
1632
1774
1686
1718
1744
6.4
1.7
2.6
3.1
3.0
Bag 2:
Engine-out
Stock catalyst
No assist
Heat assist
Heat & air
assist
6.43
2.57
5.04
4.52
4.62
5.18
2.50
4.56
4.16
4.20
1.25
0.08
0.48
0.37
0.42
119.2
11.3
71.2
42.0
43.8
1758
1944
1848
1855
1876
5.8
2.0
1.8
2.5
2.4
Bag 3:
Engine-out
Stock catalyst
No assist
Heat assist
Heat & air
assist
5.62
3.38
4.45
4.03
4.08
4.55
3.38
3.94
3.64
3.62
1.08
**
0.50
0.40
0.46
90.2
14.9
64.8
39.8
41.2
1388
1518
1430
1448
1469
8.4
2.7
3.1
4.0
3.9
** Less than 0.005 grams measured.
     Table 5 presents  composite  FTP  emission levels as well as a
computed natural  gas fuel economy  (CNG MPG) value  based on the
properties of the fuel used with the Grace EHC.

-------
                              -15-
                             Table 5
Grace Methane Catalyst Evaluation
FTP Emission Levels (grams/mile)
Catalyst
Configuration
Engine-out
Stock catalyst
No assist
Heat assist
Heat & air
assist
THC
1.72
0.86
1.32
1.18
1.20
CH4
1.40
0.85
1.18
1.07
1.08
NMHC
0.32
0.01
0.14
0.10
0.12
CO
31.2
4.6
19.0
11.6
12.3
C02
436
480
453
459
464
NOX
1.8
0.5
0.6
0.8
0.8
CNG
MPG
13.7
13.7
13.8
14.0
13.8
     The lowest composite emission levels of THC and CO using the
Grace catalyst occurred with the resistive heat assist only.  THC
were measured at 1.18 grams/mile in this configuration, which also
resulted  in the  lowest  non-methane  hydrocarbon  level  at 0.10
grams/mile.   This  configuration resulted  in  a  11.6  grams/mile
composite level for CO.  Not unexpectedly,  the lowest composite NOx
level measured with the low-volume Grace catalyst occurred without
any heat or air assist at 0.6 grams/mile.   Natural gas fuel economy
was not appreciably influenced by the use of the Grace catalyst.
     B.
Main Underfloor Catalysts
     Two larger volume main  catalysts were also evaluated; these
were  supplied by  Kemira  Oy  and by AlliedSignal  Inc.    These
catalysts  were evaluated  using three  different configurations:
"engine-out"  (no  catalyst present in  the exhaust  system),  "no
assist"  (catalyst without  supplemental  air  assist),  and  "air
assist" (60 seconds of air assist to  the catalyst after key-on in
Bag  1  only).   Engine-out  emission levels  here,  measured before
testing each of these catalysts, were very similar and are averaged
in the figures below.  They differed substantially,  however, from
engine-out  levels  measured  prior  to  the  Grace  catalyst  testing
described in the previous section.
     Figure 8 presents Bag 1 emission levels of total hydrocarbons
when utilizing these two larger volume catalysts.

-------
                              -16-
                              Figure 8
          Larger Volume Main Catalyst Evaluation
         Bag 1 Total Hydrocarbon Emission Levels
Catalyst Configuration

            Engine-out

            Kemira Oy
             No assist
              Air assist
           AlliedSignal
              No assist

              Air Assist
                       5.88
1.82
 2.06
                     01      23456

                              Total Hydrocarbons (grams)
      Both catalysts were very effective at  reducing  Bag  1 total
 hydrocarbons.   With  the  Kemira  Oy catalyst,  Bag  1 hydrocarbons
 levels were reduced  from the averaged engine-out  level  of  5.88
 grams to  1.82  grams without any  catalyst assist, a  69  percent
 reduction.   Similarly,  the AlliedSignal  catalyst  reduced Bag  1
 hydrocarbon levels to 1.79 grams,  almost  a  70 percent reduction
 from engine-out levels.   Adding secondary air assist upstream of
 each  catalyst  did not  additionally  decrease Bag  l  hydrocarbon
 emissions;   in  fact,  these  levels  slightly  increased  above
 unassisted catalyst levels.

      Similar trends were  noted in Bag 1 CO  emission levels as with
 Bag 1 hydrocarbons, as seen in Figure  9 below.  The larger volume
 Kemira Oy  catalyst reduced Bag 1  CO  by almost 96 percent  from
 engine-out levels,  even without air assist.   (Engine-out CO levels
 were measured  at  55.1 grams,  the unassisted Kemira Oy catalyst at
 2.5 grams.)  As with Bag 1 HC,  when 60 seconds of air assist was
 added to  the  Kemira Oy catalyst, Bag  1  CO  levels  increased
 slightly.

-------
                              -17-
                              Figure 9
          Larger Volume Main Catalyst Evaluation
         Bag  1  Carbon Monoxide Emission Levels
Catalyst Configuration

            Engine-out

            Kemira Oy
             No assist
              Air assist
           AlliedSignal
             No assist

             Air Assist
2.5
3.1
  5.1:

 4.5:
                                55.1
                           10     20    30    40    50

                              Carbon Monoxide (grams)
                                 60
      The fresh AlliedSignal catalyst also effectively reduced Bag
 1 CO levels from the 55.1-gram engine-out level.   Without catalyst
 assist,  Bag 1 CO  levels were measured at 5.1  grams when using the
 AlliedSignal catalyst, almost a 91 percent reduction from engine-
 out levels.   When air assist  was  applied  to  this catalyst during
 the first minute  of Bag  1,  CO levels  were further reduced to 4.5
 grams.

      Tables  6  and 7 present individual  Bag emission levels using
 both the Kemira Oy and AlliedSignal catalysts.   Different engine-
 out levels are presented in  each table here and were obtained prior
 to  initiating testing with either  catalyst.   These values were
 averaged in the previous two  figures for simplicity.

-------
 -18-
Table 6
Kemira Oy Methane Catalyst Evaluation
Individual Bag Results (grams)
Catalyst
Configuration
EC
NOX
C02
CO
CH4
NMHC
Bag 1:
Engine-out
No assist
Air assist
5.79
1.82
2.06
7.8
0.6
0.6
1678
1773
1758
54.0
2.5
3.1
5.40
1.67
1.98
0.39
0.16
0.08
Bag 2:
Engine-out
No assist
Air assist
5.25
0.51
0.56
7.1
0.5
0.3
1814
1905
1900
44.0
1.3
1.6
5.03
0.49
0.55
0.44
0.02
0.01
Bag 3:
Engine-out
No assist
Air assist
4.95
1.39
1.46
9.3
1.7
1.6
1423
1504
1498
39.9
0.7
0.8
4.53
1.36
1.44
0.42
0.04
0.02

-------
 -19-
Table 7
AlliedSignal Methane Catalyst Evaluation
Individual Bag Results (grams)
Catalyst
Configuration
EC
NOX
CO2
CO
CH4
NMHC
Bag i:
Engine-out
No assist
Air assist
5.96
1.79
1.85
8.4
0.5
0.7
1673
1852
1765
56.2
5.1
4.5
5.54
1.66
1.65
0.39
0.13
0.19
Bag 2:
Engine-out
No assist
Air assist
5.59
0.57
0.56
7.7
0.2
0.2
1797
1907
1914
Bag 3:
Engine -out
No assist
Air assist
5.03
1.29
1.39
10.3
1.4
1.6
1419
1524
1495
49.8
2.6
2.5
5.26
0.56
0.56
0.32
0.01
**

41.2
1.7
1.5
4.70
1.28
1.39
0.33
0.01
**
** Less than 0.005 grains measured.

-------
                                -20-
     Tables  8  and 9  present  composite FTP  emission levels when
using both the Kemira Oy and AlliedSignal catalysts.
                             Table 8
Kemira Oy Methane Catalyst Evaluation
FTP Composite Emission Levels (grams/mile)
catalyst
Configuration
Engine-out
No assist
Air assist
EC
1.45
0.28
0.30
NOX
2.1
0.2
0.2
C02
451
472
470
CO
12.2
0.4
0.5
CH4
1.33
0.26
0.29
NMHC
0.12
0.02
0.01
CNG
MPG
14.0
13.9
14.0
                             Table 9
AlliedSignal Methane Catalyst Evaluation
FTP Composite Emission Levels (grams/mile)
Catalyst
Configuration
Engine-out
No assist
Air assist
HC
1.48
0.28
0.29
NOX
2.3
0.2
0.2
COj
446
479
473
CO
13.0
0.8
0.7
CH4
1.39
0.27
0.28
NMHC
0.09
0.01
0.01
CNG
MPG
14.0
13.7
13.9
     In the  absence of  any catalyst  assist,  use of  either the
Kemira Oy or AlliedSignal catalyst resulted in  low emission levels
over the FTP; these levels were below the  levels of the California
ULEV  Standards   of  0.04/1.7/0.2   grams/mile  for  NMHC/CO/NOx
respectively.  The test data taken for this report did not permit
the calculation of non-methane organic gases (NMOG) as required by
the California regulations, nor was a Reactivity Adjustment Factor
(RAF) applied to the data.  Nevertheless,  it is the opinion of the
authors that the test results are low enough that a 0.04 grams/mile
NMOG level would be met at zero miles with a fresh catalyst.  The
bulk of the  total  hydrocarbons are methane emissions, therefore,
non-methane hydrocarbon values are very low  (0.02 grams/mile with
the Kemira Oy catalyst  and 0.01 grams/mile with the AlliedSignal
unit).  CO levels were very low with each catalyst and well below
the ULEV  CO  level of 1.7 grams/mile.   NOx emissions with either
catalyst was measured  at 0.2 grams/mile.   Again,  these emission
levels were attained at zero accumulated system miles.

-------
                               • -21-
     Adding  air  assist  provided  little  benefit  in  reducing
composite FTP emission  levels.   The only noticeable benefit from
air assist  resulted  with the Kemira  Oy  catalyst and non-methane
hydrocarbon  emissions.    When  air  assist  was  used  with  this
catalyst, non-methane emissions were reduced from the already low
0.02 grams/mile to 0.01 grams/mile.

VIII.Conclusions

     1.   It  is  difficult  to  compare  the  results  from  the
evaluation  of the  Grace EHC with those  from  the other catalysts
tested because  of  the operation and  smaller  volume  of the Grace
converter.    Despite its  small  size,  the  unassisted  EHC  was
effective at reducing Bag 1 THC, CO and NOx emission levels.

     Resistive heating in Bag 1  further lowered Bag 1 emissions of
THC and CO.   Emissions over  the  Bag  2  and 3  segments of this heat-
assist testing  were  also lowered.   The  addition of  catalyst air
assist did  not  appear to  further  lower emission levels.   Testing
using air assist without resistive heating was not conducted.

     2.   The Kemira  Oy catalyst was  very  effective  in reducing
emission  levels of  NMHC,  CO and  even  NOx.    Engine out  NMHC
emissions were  reduced  to extremely  low levels, as low  as 0.02
grams/mile over the FTP.  CO emissions were also very  low.  The use
of catalyst air assist did not further lower CO emissions, to our
surprise.   NOx  emissions were also reduced to  approximately 0.2
grams/mile.   NOx reduction activity was not greatly affected by the
use of catalyst  air assist  during  the  Bag 1  segment  of  this
testing.

     3.   The emissions  levels with  the AlliedSignal catalyst were
very  similar to  those  from the  evaluation  of the  Kemira  Oy
catalyst.

IX.  Acknowledgements

     The first catalyst  evaluated in this test program was supplied
to EPA  by the W.R. Grace and Company.   The  second  catalyst was
supplied by Kemira Oy,  located  in Vihtavuori,  Finland.   The last
catalyst used in this program was  provided by AlliedSignal Inc.,
located in  Tulsa, Oklahoma.   The  natural gas fueled Dodge Dakota
truck was loaned to EPA by Stewart &  Stevenson Power, Inc., located
in Commerce City,  Colorado.   The authors  would like to thank these
companies for their cooperation and support.

     The  authors  also  appreciate  the efforts  of James  Garvey,
Robert Moss,  and  Ray Ouillette of  the Technology Evaluation and
Testing Support Branch  who  conducted  the driving cycle tests and
emission  sampling.   The word processing and editing  efforts of
Jennifer Criss  and Lillian  Johnson  of the Technology Development
Group are also appreciated.

-------
                              -22-
X.   References

     1.   "Analysis of  the  Economic and Environmental Effects of
Compressed Natural Gas  as a Vehicle Fuel," Special Report, Office
of Mobile Sources, U.S. EPA, April  1990.

     2.   "Assessment of  the Costs and  Benefits  of  Flexible and
Alternative Fuel  Use  In The U.S. Transportation Sector," DOE/PE-
0080, January 1988.

     3.   Gas Engineers Handbook. First Edition, Industrial Press,
New York, NY, 1965.

     4.   "Exhaust  Emissions   from   Dual-Fuel   Vehicles  Using
Compressed Natural Gas  and Gasoline,"  Cadle et al., Air and Water
Management Association, Pittsburgh, PA, June 1990.

     5.   "Methane Oxidation over Noble Metal Catalysts as Related
to Controlling Natural  Gas Vehicle Exhaust Emissions," Oh, S., P.
Mitchell and R.  Siewert, Catalytic Control of Air Pollution, pp!2-
25, American Chemical Society, 1992.

     6.   Letter   from  B.   Bertelsen,   Executive   Director,
Manufacturers  of  Emission  Controls  Association  to W.  Reilly,
Administrator, U.S. EPA, January 4, 1993.

     7.   "1992 Natural Gas Vehicle Challenge:  EPA Emissions and
Fuel Economy," Breutsch, R. I. and M.  E. Reineman, EPA/AA/TDG/92-
05, June 1992.

-------
APPENDIX A
NATURAL GAS VEHICLE TEST ANALYSIS
Dyno: D209 Test No: 93-1543
Input Data for 1B7GL23Y8NS529176
"
Tost No. MFR Vehicle ID Veh Version
93-1548 20 1B7GL23Y8NS529176 0
Identification
REQIO MFR Initials Driver ID Oper ID LA4 Prep ID
22024 0 54768 54721 54768
Flags
EVAP Rag PARTIC RETEST RFCC
0 000
RUNCHQ
0
Disposition and Accounting
Veh Disp Test Olsp Void Cod* TPS
0 000


Dynamometer and Analyzer Site
Dyno IW Set TWHP Coast Down
D209 4750 13 0:00
Odometer
1766

Dyno/CVS Distance VMIX Seconds
Bag 1
Bag 2
Bag 3
Units
8378 4983 505.00
8999 8512 864.00
8375 4964 505.00
Ft
Exhaust
HCRD
Bag 1
Bag 2
Bag 3
CO
Bag 1
Bag 2
Bag 3
C02
Bag 1
Bag 2
Bag 3
NOx
Bag 1
Bag 2
Bag 3
Methane
Bag 1
Bag 2
Bag 3
NGFuel
Properties

COMMENTS:
THIS TEST WAS CO
Range Meter
14 86.3
14 49.7
1 4 74.2

18 71
18 38.1
18 56.5

22 69.9
22 47
22 81

15 64.2
15 3S.3
^M ' '•'
>•* 10.4
Fuel Meter
Units

page 1/3
Processed: 06/25/93 11:18

Test Type
05




Tire
0

9 Preps
0
Procedure
02
Test No.
93-1548

Test Date Key Start
6/10/93 9:53

Prep Date Prep Key Off
6/9/93 16:21

Ace Code Ace Code
0 0

CVS
27C

Reading
0
0
0
0
Analyzer HF1D
A203 A 2 03

Ambient Conditions
Barometer 29.05
Ambient 76.5
Dew Point 47.8
Temp Units D
•Background
Range
14
14
14
Meter
3.4
3.2
3.4



18
18
18
0.2
0.3
0.3


22
22
22
4
4.1
4.1


15
15
15
0.2
0.2
0.2


18
18
18
THC . NMHC NQ CNGTHC
H/C Ratio H/C Ratio NQ/C Ratio CWF
3.886 2.891 1.441 0.667
0
0
o
suika hU AT THE US. EPA NATIONAL VEHICLE AND FUEL EMISS
0.7
0.7
0.7
C02
WF
0.090
IONS LABORATC

HCRD
Bag 1
Bag 2
Bag 3
CO
Bag 1
Bag 2
Bag 3
CO2
Bagi
Bag 2
Bag 3
NOX
Bag 1
Bag 2
Bag3
Methane
Bag 1
Bag 2
Bag 3
GHV Specific
BTU/ft*3 Gravity
JOOO 0.593

RY - BOD. ANN ARBOR, MICHIGAJ

-------
APPENDIX A (CONT'D)
NATURAL GAS VEHICLE TEST ANALYSIS page 2/3
Dyno: 0209 Test No: 93-1548 Processed: 06/25/93 11:18
Raw Emission Determination for 1B7QL23Y8NS529176
Ambient
Conditions

HC
Bag 1
Bag 2
Bag 3
CO
Bagi
Bag 2
Bag 3
CO2
Bag 1
Bag 2
Bag 3
NOx
Bag 1
Bag 2
Bag 3
Methane
Bag 1
Bag 2
Bag 3
NMHCo
Bag 1
Bag 2
Bag 3
THC Reso Adlus*
Bagi
Bag 2
Bag3
COMMENTS:
THISfESfWA$
-------
APPENDIX A (CONT'D)
NATURAL GAS VEHICLE TEST ANALYSIS pag0 3/3
Oyno: 0209 Test No: 93-1548 Processed: 06/25/93 11:18
Emission. Mass, and Fuel Economy for 1B7GL23Y8NS529176
Natural Gas
Properties
Natural Gas
Properties
Dvno/CVS
Bag 1
Bag 2
Bag 3
CVS Corrected
Concentrations
Bag 1
Bag 2
Bag 3
CVS Mas*
Emissions
Bag 1
Bag 2
Bag 3
Total Mass
Emissions
Bag 1
Bag 2
Bag 3
Mass
Emissions
Bag 1
Bag 2
Bag 3
Composite
Emissions
Unrounded
Rounded
Natural Gas
Fuel Economy

THIS TbAT (JlJWC
Natural Gas
Fuel Economy

THIS IHSTCONC
THC
H/C Ratio
3.8860
NG
NG/C Ratio
1.4410
VMIX
4983
8512
4964
CH4C
oom
57.30
31.40
48.33
CH4
grams
5.39
5.05
4.53
CH4
grams
5.39
5.05
4.53
CH4
fl/Rll
1.501
iJSi
1 .33408
1.33S
gram* C
per mil*
129.349
THC
Density
18.749
NG
CWF
0.694
Roll Revs
8378
8999
8375
NMHCc
ppm
0.00
0.00
0.00
MAJLIS
run iw
gram*
0.00
0.00
0.00
UAJLJO
ranmw
gram*
0.00
0.00
0.00
NMHC
a/ml
0.000
0.000
0.000
NMMC
a/ml
0.00000
0.000
gram* NO
per grim* C
1.441
THC
CWF
0.754
CNGTHC
CWF
0.667
Mile*
3.593
3.860
3.592
TOTAL HC
ppm
57.30
31.40
48.33
TOTAL HC
gram*
5.39
5.05
4.53
TOTAL HC
gram*
5.39
5.05
4.53
TOTAL HC
a/ml
1.501
1.308
1.261
TOTAL HC
a/mi
1 .33498
1.335
Ijftm^
IwVK*
H/C Ratio
2.8910
Specific
Gravity
0.593
Dllut Factor
NMHC NMHC
Density CWF
17.568 0.805
CO2
WF
0.096
GHV NHV NG
BTU/ft*3 BTU/a Density
7000 43.967 20.470
Numerator Dilut Factor Correct
9.656 13.374 0.9252272
20.608 0.9514749
15.543 0.9356608
COc
ppm
325.85
159.11
247.83
CO
gram*
53.53
44.65
40.56
CO
gram*
53.53
44.65
40.56
CO
g/ml
14.898
11.569
11.292
CO
a/mi
12.183
12.2


JOTEu AT THE U.S. EPA NATIONAL VEHlCLb AMD filSL EMLSSIOr
FCng
Numerator
129.349
FCng
Denominator
14.205
X'I'HU AT THE U.S. EPA NATION
FCna
9.106
AL VEHICLE AN1
CO2na
1 7.894
CO2c NOxe
% pom
0.650 31.79
0.414 17.35
0.556 38.33
coa NO*
grams grams
1677.33 7.70
1824.15 7.18
1430.91 9.25
CO2 NOX
gram* gram*
1677.33 7.70
1824.15 7.18
1430.91 9.25
C02 NOX
g/ml g/ml
466.795 2.144
472.625 1.861
398.361 2.576
C02 NOX
g/ml g/ml
451.01 2.1159
451 2.12

UTG BTU'* NG BTU'S NQ
per gallon per mil* MPG
114132 8195.182 13.93
S LABORATORY. ANN ARBOR. MICHIGAN
gr C / mil* Proposed
Numerator Denominator CFR MPG
1658.878 124.464
13.33


-------