ENVIRONMENTAL PROTECTION

AGENCY
                     SPECIAL REPORT
               OFFICE OF MOBILE SOURCES
         Analysis of the Economic and
      Environmental  Effects oi Compressed
         Natural Gas as a Vehicle Fuel

                  Volume II

             Heavy-Duty  Vehicles


                  April 1990

-------
     This  report  addresses  the  economic   and   environmental
issues associated with the  use of compressed  natural gas  as  a
motor vehicle  fuel.   Volume  I analyzes  the use  of  compressed
natural  gas  as a  fuel   for  passenger  cars  and  light  trucks.
Volume  II  considers  the use  of  compressed natural  gas  as  a
heavy-duty vehicle fuel.

-------
Analysis of the Economic and Environmental Effects
    of Compressed Natural Gas  as a Vehicle Fuel
                    Volume  II
                Heavy-Duty Vehicles
                    April 1990

-------
                            Table of Contents
Chapter

  1       Introduction
          Uses of CNG in Heavy-Duty Vehicles
          I.    Historical Use of CNG in Heavy-Duty
                  Vehicles                                          2-1
          II.   Recent Progress in CNG Heavy-Duty
                  Vehicle Applications                              2-3
          III.  Potential Future Markets for CNG
                  Heavy-Duty Vehicles                               2-11

          CNG Heavy-Duty Vehicle Technology
          I.    Introduction                                        3-1
          II.   Fuel Properties of Natural Gas                      3-1
          III.  Current Heavy-Duty CNG Technology                   3-3
                A.    Engines                                       3-3
                B.    Fuel Storage                                  3-7
                C.    Emissions from Current Technology
                       CNG vehicles                                 3-9
                      1.    Dual Fuel Vehicles                      3-9
                      2.    Dedicated CNG Vehicles                  3-14
                D.    Engine Efficiency                             3-21
                E.    In-use Performance                            3-25
          IV.   Optimized Vehicle Projections                       3-26
                A.    Emissions                                  '   3-26
                B.    Future Optimized Engine Efficiency            3-30
          V.    Safety Issues  for CNG uses  in Heavy-Duty
                  Application                                       3-31
                A.    Introduction                                  3-31
                B.    Fuel Properties and General
                        Considerations                              3-32
                      1.    Toxicity                                3-32
                      2.    Flammability                            3-32
                      3.    Hazards Associated With a Fire          3-35
                      4.    Issues of Special Concern in
                               The Use of CNG                        3-36
                C.    Implications for Vehicle Safety               3-37
                      1.    Refueling                               3-37
                      2.    Vehicle Operation and Crashes           3-38
                      3.    Maintenance                             3-40
                D.    Summary                                       3-41

-------
                        Table  of  Contents  (cont'd)

Chapter                                                              Page

  4       Economics of Using CNG  in Heavy-Duty  Application
          I.    Introduction                                         4-1
          II.    Domestic Natural  Gas Supply and Price                 4-1
          III.   CNG Refueling  Station Cost                           4-2
                A.     CNG Refueling Station Hardware                  4-2
                B.     CNG Refueling Station Hardware Cost             4-5
                C.     Total CNG Refueling  Station Cost               4-8
          IV.    Heavy-Duty CNG, Gasoline and Diesel
                  Vehicle Fuel Cost                                  4-12
                A.     Basis of Comparison                             4-13
                B.     Compression and Station Maintenance  Costs      4-13
                C.     Capitalized Service  Station Cost               4-14
                D.     Relative Fuel Prices                           4-15
                E.     Relative Vehicle Fuel Costs                    4-15
          V.    Heavy-Duty CNG Engine and  Vehicle Costs              4-20
                A.     Engine Costs                                   4-20
                B.     Vehicle  Costs                                  4-20

  5       Air Quality Benefits
          I.    Introduction                                         5-1
          II.    Urban Ozone Level                                    5-1
          III.  Air Toxics                                           5-9
          IV.    Global Warming                                       5-15
          V.    Other Air Quality Impacts                             5-20

-------
                           CHAPTER 1

                          INTRODUCTION
     This report  is one  in a  series  of  reports which EPA  is
preparing in order  to  describe the  environmental and  economic
impacts of various  clean alternative fuels.  It  deals  with the
use of  compressed  natural gas  (CNG)  as  a  fuel  for  heavy-duty
vehicles.    CNG,   of  course,  also   is  used   in   light-duty
vehicles.   However,  the nature and issues  associated  with  these
two  areas  are  sufficiently different  that EPA  has  chosen  to
issue  separate  volumes  on  light-duty  and heavy-duty  uses  of
CNG.   Other  reports  in  this  series  already  issued or  being
prepared deal with  methanol,  ethanol,  liquified  petroleum gas,
electricity  and  reformulated  gasoline.   In  sum,   they  will
provide  a  comprehensive   review  of   the  choices  of   clean
alternative fuels  which might  be used  to move  America  toward
cleaner cars  and trucks and reduced dependence upon petroleum.

     This  report  was  provided in draft  form  for   review  and
comment to other  EPA offices  as  well as  a variety of  external
organizations.   A number  of comments  were received in  response
to  this  request  for  review.   To  the  extent  possible  these
comments have been  incorporated into  this final  version  of the
report

     When  considered  as  a  vehicle   fuel,  CNG  is  distinctly
different from conventional gasoline or  diesel  fuels  in that it
is  a  gas  at  all normal  temperatures  and pressures.    Thus,  it
requires different  approaches  to  vehicle refueling and  to fuel
storage on the  vehicle.   On the other hand, being gaseous means
that  the  entire  CNG  fuel  system  must  be   a  closed  one,
eliminating evaporative  emissions entirely.   Because  of  this,
and other  clean burning  characteristics  of CNG, it  offers the
potential for significant emission reductions in vehicle uses.

     Considerable  experience  with  CNG  use   in a   variety  of
heavy-duty   applications   already  exists.     CNG   has   been
successfully used to power  vehicles  ranging from light delivery
trucks  to  full-sized   urban   buses.   These  applications  have
generally been  based upon conversions  of existing truck engines
to  run  on  both  gasoline and CNG;   which  has allowed  the  use of
CNG  fuel  even  though  it  is  not widely available  at fueling
stations.  However, when considered from the  perspective of use
as  a  clean  alternative  fuel,  the use  of dedicated vehicles,
which  can run  only  on  CNG,  assumes  a  dominant  role.   Such
vehicles,  because  they  can  be  optimized to  make   use  of the
specific  combustion properties of CNG,  hold  promise  of much
greater  emission  reductions and  fuel  consumption gains than do
dual-fueled vehicles.

-------
                               1-2

     The  remainder  of this  report  examines  in greater  detail
the  use  of  CNG  in  heavy-duty  vehicles,  both  current  and
future.  It begins with  an  introductory overview of current and
historical    applications    for   CNG    vehicles,    and    the
identification of  types  of  uses where the  most  ready growth of
CNG  would  be   possible.    Chapter   3  describes   CNG  engine
technology   and   presents   emissions  data   for   CNG  fueled
heavy-duty  engines.   Based  upon  projections   of  potential
progress  in emissions control  and  fuel  consumption,  emissions
estimates  for  more  advanced  future  CNG  engines  are  also
developed.   Chapter   4  reviews  the  costs  associated  with CNG
use.   These  include  vehicle  costs  for  engine  and  fuel  tank
hardware,  fueling  station costs  and  fuel  costs.    Since  heavy
duty applications  for CNG represent  a relatively  small  impact
on  overall  natural  gas use,  fuel cost  estimates  are based upon
the  assumption  that  current  natural  gas   supplies  to  the
domestic  market  can  be  used,  with  negligible  impact on gas
prices.  Finally, Chapter 5  provides a comparative  analysis of
the   environmental   benefits   of   CNG   compared   to   both
gasoline-fueled  and  diesel  engines.    This  analysis  focuses
principally  on  the  areas  of  ozone,  air  toxics  and  global
warming  and shows that CNG  indeed has the  potential  to provide
significant emissions benefits.

     As  in  the other reports in this  series, this report deals
with  the  topics  of  CNG   technology  and  the  economic  and
environmental  impacts of heavy-duty CNG  use.  There  are other
matters  related  to the  introduction  of alternative  fuels which
are  not specifically addressed.   These include such  things as
establishing   emissions   standards    and    test   procedures,
overcoming  institutional barriers,  etc.   While  such questions
are important, they are outside the scope of  these reports.

-------
                           CHAPTER 2

               USES OF CNG IN HEAVY-DUTY VEHICLES
     The  purpose  of  this  chapter  is  to  provide  background
information  on  the  use  of  compressed  natural  gas  (CNG)  in
heavy-duty  vehicles  in  the   United  States,   and  to  discuss
generally the potential  for  future  growth in  this  area.   The
Chapter  will address  economic,  technological,  and  regulatory
factors  only  peripherally   as   needed   to  explain   certain
projections and conclusions;  later chapters  in  this report will
detail findings  in these areas.

     This chapter is organized into  three sections.   Section I
addresses  the  historical  use of  CNG  in heavy-duty vehicles,
Section  II  describes  recent   activities  in the  conversion  of
heavy-duty  vehicles  to  CNG,  and   Section  III  describes  the
potential   future   market  for   CNG   in   heavy-duty   vehicle
applications.

I.    Historical  Use of CNG in Heavy-Duty Vehicles

     Experience  with  and interest  in  CNG as  an  alternative
motor-vehicle fuel differs significantly around  the globe.  Use
of CNG  in heavy-duty vehicles throughout  the mid-1970s  in the
United States was  sporadic  and isolated  at  best,  while  use of
CNG  in  parts of Europe,  South America, and Asia  has been more
enthusiastic.   While  economics  was   the  primary  motivating
factor  for  CNG  use both  abroad  and  in  the  United  States,
environmental  and  political  concerns  have   also  played  an
important role in the United States.

     CNG as  a vehicular  fuel  has  been popular in other parts of
the world  since  the mid-1930s.  In  Italy, for example,  use of
CNG  as  an  alternative   fuel  began  in  1935   as   part   of the
National Economic Policy,  which  called for  self-sufficiency in
all   materials.    Gasoline   shortages   during  World   War  II
increased  the demand for CNG, which was  used  for  both public
transportation and  private vehicles.  At  its peak from 1945 to
1960,  CNG captured  about ten percent  of  the   Italian  vehicle
fuel  market.[1]   When gasoline became less expensive  and more
readily  available  in  the 1960s,  however, CNG's  share  of the
fuel  market  fell.   The international appeal of CNG continues to
be strong: New Zealand has 125,000 cars  and  trucks converted to
natural  gas;  Canada has  20,000 natural  gas  vehicles; Argentina
has  15,000  natural  gas  vehicles;   and  the Soviet  Union has
200,000  vehicles,  with plans  to convert  another  300,000 by the
end of 1990.[2]

-------
                               2-2

     Use of CNG  in  heavy-duty vehicles in the United States was
given a boost in the 1970s with the general  increase in concern
for the environment.  Auto  emissions  were identified as a major
source of air pollution,  spurring  some interest  in development
of  alternative   fuels.   During this  period,  several  companies
offered systems  to  convert vehicles  from traditional  fuel  use
to natural  gas.   The market  was  not well  defined or financed,
however,  and the  demand  for vehicles  fueled  by  natural  gas
remained  low.    An  estimated 15,000  CNG-fueled  vehicles  were
operating  at this  point   in time  in  the  United  States,  a
fraction of which were heavy-duty vehicles.

     The  "energy crises"  of  the  1970s generated  a  concern for
energy  and  oil  security in  the  United States,  and temporarily
increased  interest   in  alternative  fuels.    The  United  States
made a  concerted effort  to find substitutes for  foreign  oil in
order  to  weaken the  hold  of  the  OPEC  cartel  on world  oil
prices.   The  transportation sector appeared to  be the  key to
meaningful   reductions   in   oil  consumption.     Interest   in
alternative fuels waned, however,  as  significant investments in
energy   efficient  technologies   between  the   mid-1970s   and
mid-1980s, combined with an increase  in the world's oil supply,
eventually forced lower oil prices.

     Motivated  by  the  potential  for substitution of CNG for
gasoline  and  diesel   fuel   to   reduce  fuel   costs,   improve
declining  urban  air' quality, ' and  create  energy  security,  a
variety  of  interrelated  market  and  regulatory  forces  caused
renewed  interest in CNG use  in heavy-duty  vehicles during the
1980s.    This    renewed   interest   stimulated   CNG-related
development,  resulting  in   a  viable  CNG  heavy-duty  vehicle
market.   This market  currently includes  an  estimated  1,500 CNG
heavy-duty vehicles*  served  (along with  light-duty vehicles and
trucks) by  up to 300  refueling stations  (mostly owned  by gas
utilities),  and supported  by several utilities, approximately
80 CNG vendors,   and related gas interest associations.
     For  purposes  of this  report,  "heavy-duty"  vehicles  are
     defined to have  a gross vehicle weight  (GVW)  of  more than
     8,500 pounds.  There  is,  however,  a disparity between this
     definition  and the  definition used  by the  Department  of
     Transportation (DOT).   Dot defines  a  "heavy" truck as:  1)
     a  single-unit  truck with  GVW greater than  26,000 pounds,
     2)  a tractor-trailer  combination,  3)  a truck  with cargo
     trailers,  and  4) truck-tractor  pulling no  trailer.   Some
     trucking  associations use  alternative definitions, such as
     a   GVW   in  excess   of  20,000  pounds.    Differences  in
     definitions  may  make  comparisons   of   vehicle  statistics
     difficult.

-------
                              2-3

II.   Recent Progress in CNG Heavy-Duty Vehicle Applications

     As shown in Figure  2-1,  the number of heavy-duty  vehicles
in  the  United States,  which represents  the  maximum  potential
market  for  CNG-fueled  heavy-duty  vehicles,   is  quite  large.
Efforts  by  the  CNG  industry  to  penetrate  this  heavy-duty
vehicle market have met with mixed  success.   Figure 2-2  shows
the  CNG heavy-duty  vehicles that  exist  today  in the  United
States  by  application,  including dump trucks  and  heavy-duty
pickup  trucks,   school  buses,   transit  buses,  United  Parcel
Service (UPS) delivery trucks, and trash  collection trucks.   Of
the   estimated   1,500   heavy-duty   vehicles   that   have   been
converted  to  CNG,  about 43  percent  are heavy-duty  dump  trucks
and   pickup   trucks,   and   18   percent   are   school   buses.
Unidentified vehicles  in the Figure  represent that  fraction of
the  total  number  of CNG heavy-duty  vehicles  that  the  American
Gas  Association  estimates  exists  but  which  have  not  been
identified to date  in  their  surveys.   That number should be
considered only an approximate value.

     As  detailed   in  Table  2-1,   the   conversion  programs
completed  to date  were  primarily sponsored  (and predominantly
funded)  by  gas  utilities,  frequently  in conjunction  with   a
transit authority or  school  district,  and have been designed to
test  evolving  diesel   conversion  technology.    Most   of  the
conversion projects have  involved  only  one or  two vehicles.
There  are  a few   notable  exceptions  to  these  generalities,
including  a  cooperative effort  between two  Ohio  gas companies
and  Flxible  Corporation  to  develop  prototypes  for  use  in
transit  districts,  and  a  Garland, Texas school  district that
decided on its own to dedicate its fleet of buses to CNG.

     Texas  itself  is  worthy of  further  note.   This  state is
attempting to bring its four  non-attainment  areas (Dallas-Fort
Worth,  Houston-Galveston-Brazoria,   Beaumont-Port   Arthur,   El
Paso)  into  compliance  through   proactive legislative  action.
Texas  Senate  Bill  769   requires  the  use  of  CNG  or  other
alternative  fuels  that  reduce emissions  in rapid transit buses
and,  if  necessary,  certain  local government  and  private  fleet
vehicles   in  non-attainment  areas  for   either  ozone,   carbon
monoxide  (CO),  oxides of  nitrogen,  or particulates.   By  1994,
940   transit buses  are  to  be  converted  to  CNG  with  the
possibility  of  4,038  school  buses  to be  added to  that   figure
(122  of which  are  presently CNG  fueled).   Details  on   Texas'
legislative  implementation schedule are given in Figure 2-3.[4]

      The  use of  both prototypes  and entire  fleet conversions
over  the  last  decade  has  had   the  effect  of  chronicling the
evolving  technology of  CNG-conversion  vehicles,  beginning with

-------
                                    2-4
                               FIGURE 2-1

                     U.S. Heavy-Duty Vehicle Market
         Number of Vehicles (thousands)
 1600
 1400i
  1200^
  100
-------
                                    2-5
                               FIGURE 2-2

             Distribution of Existing CNG Heavy-Duty Vehicles
         Number of Vehicles
  7001
  600-
  500-
  400-
          Tr ucks     School     Transit      UPS       Trash    Unidentified
                     Buses      Buses                 Trucks     Vehicles

                CNG Heavy-Duty Vehicle Applications
Sources:     American Gas Association, the Natural Gas Vehicle Coalition, Representatives of
           Gas Utilities and Representatives of Vehicle Conversion Suppliers.

-------
                               2-6

                           TABLE 2-1

          Existing CNG Heavy-Duty Vehicle Applications
Type   of   vehicle:    Bi-fuel   four-stroke   diesel  bus   (80%
diesel/20% CNG)
Number of vehicles:   1
Location:  Phoenix,  AZ
Goals   of   application:   Emissions   reductions,   technology
demonstration
Participant:  City of Phoenix
Sponsor:  Southwest Gas Corporation
Source:  Southwest Gas Corporation
Type of vehicle:  CNG-dedicated two-stroke diesel transit bus
Number of vehicles:  1
Location:  Tucson, AZ
Goal of application:  Technology demonstration
Participant:  City of Tucson
Sponsor:  Southwest Gas Corporation
Source:  Southwest Gas Corporation
Type of vehicle:  Heavy-duty truck
Number of vehicles:  630
Location:  Arizona
Goal of application: Economics
Participant:  Southwest Gas Corporation
Sponsor:  Not Applicable
Source:  Southwest Gas Corporation
Type of vehicle:  Two-stroke diesel transit bus
Number of vehicles:  2
Location:  Tacoma, WA
Goal of application:  Demonstration project
Participant:  Pierce Transit
Sponsors:  Washington Natural Gas Company, Pierce Transit
Source:  American Gas Association
Type of vehicle:  Transit bus
Number of vehicles:  2
Location:  Los Angeles, California
Goal of Application:  UMTA demonstration project
Participant:  Southern California Regional Transit District
Source:  American Gas Association

-------
                              2-7

                           TABLE 2-1

      Existing CNG Heavy-Duty Vehicle Applications (con't)
Type of vehicle:   Transit bus
Number of vehicles:   2
Location:  New York City
Goal  of  application:   CNG  bus demonstration  program  begun  by
New   York  City's   Department   of    Transportation   and   the
Department of Environment Protection
Participant:   Command Bus Company
Sponsors:  Brooklyn Union Gas  Company  provided engines; partial
funding from   UMTA
Source:  American Gas Association, Natural Gas Vehicle Coalition
Type of vehicle:  Trash collection truck
Number of vehicles:  1
Location:  New York City
Goal of application:  Demonstration program
Participant:  Department of Sanitation
Sponsor:  Brooklyn Union Gas Company
Source: Natural Gas Vehicle Coalition
Type of vehicle:  Delivery trucks (UPS)
Number of vehicles:  10
Location:  New York City
Goals of application:  Demonstration program
Participant:  United Parcel Service
Sponsor:  Brooklyn Union Gas Company
Source:  Natural Gas Vehicle Coalition
Type of vehicle:  Dual-fueled trucks
Number of vehicles:  2
Location:  Minneapolis-St. Paul, MN
Goal of application:  Demonstration program
Participant:  Twin Cities Metropolitan Transit Authority
Sponsor:  Minnegasco
Source:  American Gas Association

-------
                               2-8

                       Table 2-1 (cont'd)

          Existing CNG Heavy-Duty Vehicle Applications
Type of vehicle:  Gasoline-fueled school bus
Number of vehicles:  122
Location:  Garland, TX
Goal of application:  Economics
Participant:  Garland,  TX School District
Sponsor:  Unknown
Source:   Transportation  Department   and  Garland  Independent
School  District,     "Compressed Natural  Gas  System,  Two-Year
Summary," June 1985.
Type of vehicle:  School bus
Number of vehicles:  24
Location:  Tulsa, OK
Goal of application:  Economics
Participant:  Tulsa, Oklahoma School District
Sponsor:  Oklahoma Natural Gas Company
Source:  American Gas Association
Type of vehicle: ' School bus
Number of vehicles:  90
Location:  Indiana
Goal of application:  Economics
Participant:  Evansville & Vanderburg School Corporation
Sponsor:  Not available
Type of vehicle:  School bus
Number of vehicles:  40
Location:  Erie, Pennsylvania
Goal of application:  Economics
Participant:  Harbor Creek School District
Sponsor:  Not available

-------
                              2-9

                           TABLE 2-1

      Existing CNG Heavy-Duty Vehicle Applications (con1t)
Type of vehicle:   Dual-fuel CNG/diesel dump-trucks
Number of vehicles:   18
Location:  Columbus, OH
Goals of application:  Economics
Participant:   Columbia Gas Corporation
Sponsors:  Not applicable
Source:  American Gas Association
Type of vehicle:   Garbage truck
Number of vehicles:  2
Location:  North Miami, FL
Goals of application:  Testing program
Participant:  City of North Miami
Source:  Natural Gas Vehicle Coalition
Type of vehicle:  Transit bus
Number of vehicles:  1
Location:  Cleveland, OH
Goals of application:  Environmental, Economic
Participant:  Consolidated Gas Corporation
Sponsors:  Flxible Corporation, Consolidated Gas Corporation
Source:  Flxible Corporation
Type of vehicle:   Transit bus
Number of vehicles:  1
Location:  Columbus, Ohio
Goals of application:  Environmental, Economic
Participant:  Central Ohio Transit Authority
Sponsors:  Columbia Gas Company; Flxible Corporation
Source:  Flxible Corporation

-------
                                     2-10


                                 FIGURE 2-3

                   Projected Use of CNG Buses  In Texas *

      Number of Vehicles (thousands)
       Transit  School Transit  School  Transit  School  Transit  School
       Buses     Buses   Buses     Buses   Buses     Buses   Buses     Buses
           1989           1994           1996           1998

                 Legislative  Implementation Schedule
Source:
Texas Senate Bill 763 requires the use of CNG, or other alternative fuels that reduce
emissions, to  be used  in  rapid transit buses and, if  necessary,  certain local
government and private fleet vehicles in non-attainment areas.  Conversions of school
buses as well as other fleets may not be necessary if Texas' 4 non-attainment areas
(Dallas-Fort Worth, Houston-Galveston-Brazoria, Beaumont-Port Arthur, and El Paso)
reach and maintain required attainment levels.

The Economic Costs and Benefits of Proposed Amendments to the Texax Clean Air
Act, the State  Purchasing and General Services Act, the Metropolitan Rapid Transit
Authority  Act,  the  Regional  Transportation Authority  Act  and  the Citv
Transportation Department Act, prepared for Texas General Land Office, March
1989.

-------
                              2-11

CNG conversions  of  gasoline or  diesel  fuel systems  (dual-fuel
vehicles*)  and  moving   into   CNG-specific  fuel   systems  and
engines  (i.e.,  dedicated  vehicles).    The  initial  conversions
were  from gasoline  vehicles to  dual-fuel  vehicles.   Gasoline
conversions to  dual-fueled  vehicles  are no  longer under  such
active  consideration,   however,  because  such  vehicles  do  not
offer the cleanest or  most fuel-efficient  alternative,  as  will
be discussed in  Chapter  3.   Most gasoline conversions today are
to dedicated CNG vehicles.

     Complications  with  the diesel  engine  have  historically
limited  diesel  conversions  to  primarily bi-fuel  applications.
Recent  efforts  by various  gas  utilities  (e.g.,  Columbia  Gas
Co., Brooklyn Union  Gas  Co., Southwest Gas  Corp.)  have resulted
in  the  successful   conversions  of  2-stroke  diesel  buses  to
bi-fueled  vehicles,   and  a  4-stroke   diesel   bus  to  bi-fuel.
These  prototypes have  proven  that the  conversions  could  be
accomplished and that  the characteristic  black smoke from older
diesel-fueled   vehicles    could   be    eliminated.**     Diesel
conversions to dedicated  vehicles,  however, look more promising
in the long run, with the advent of the Cummins NG  L 10 engine,
which  is  derived  from  an  existing  Cummins  diesel  engine.
Already  being  tested in  buses,  the Cummins  design is  able  to
meet the  1991 bus emissions  standards.[3]

III. Potential Future Market for CNG Heavy-Duty Vehicles

     The   heavy-duty  CNG   market   currently   appears   to  be
undergoing  a  resurgence  of interest.   There are  about  170
confirmed CNG vehicles  that have been contracted or planned for
the  near future.  Many  of these vehicles  involve  diesel buses
 *    The  term  "dual-fuel"  is  used  to  refer  to  both  a  fuel
     system  that  enables  the  user  to  switch  back  and forth
     between  CNG  and gasoline  use using  a manual  switch,  as
     well  as  to CNG/diesel engines  (sometimes  called bi-fuel),
     which  idle on  100  percent  diesel  and  run on  80 percent
     CNG/20  percent  diesel.   This  term should not  be confused
     with  what  are   commonly  called  "flexible  fuel" vehicles.
     These latter  vehicles,  which operate on liquid  alternative
     fuels,  have  the ability  to  inter-mix  fuels   in  the  same
     fuel  tank.   In  contrast  to  this,  CNG,  being  gaseous,
     requires  its  own distinct fuel  storage and delivery  system.
 **   Given  both the  difficulty  of  2-stroke  engine conversion
     and  the  presence  in  the   exhaust  of  increased unburned
     fuel,  the technological focus  currently is  on converting
     4-stroke  diesel  engines.

-------
                              2-12

that will be  converted to dedicated  CNG use using  the  Cummins
NG  L  10  engine.   Table  2-2  briefly  describes  the  future
projects which  have been  identified.   Some  of these  projects
are  partially  funded  by   UMTA.    UMTA  is  providing  grants
totaling  $35.1   million  to   purchase  buses   that   run   on
alternative  fuels;  approximately  20 percent  is  earmarked  for
CNG-fueled  buses  (although  80  percent  of  the  applications
received have been for CNG).

     Besides  basic  economic  factors, the future market  for  CNG
use in heavy-duty vehicles  is  also dependent on the parameters
of  the  regulatory  environment  surrounding  use of  alternative
fuels.   Depending on  the particular  requirements  in the  final
version  of  the  Clean Air Act  amendments, and  implementation of
the 1991 and  1994 emissions standards for buses  and trucks, the
potential size of the CNG market could grow significantly.

     The President's proposed  amendments also  include a plan to
require  the use  of  clean alternative  fuels  in  new  urban  buses
in  all  cities with a  1980  population  of over  1 million.   This
proposal would affect  over  80  percent of all new  bus purchases
by  the  mid-1990s.  Given the current high level of  interest in
CNG applications  for buses,  this would  appear  to be  a promising
market for CNG.

     More broadly,  CNG appears  to  be a  readily adaptable fuel
for   a   variety  of   heavy-duty  uses.   Heavy-duty  vehicles
generally have sufficient space available for  CNG fuel cylinder
placement and, as will be seen in subsequent portions  of this
report,  are  not  greatly  affected  by the added weight of  the
fuel  cylinders.    CNG  appears  especially   suited  to  those
heavy-duty  applications which  have  access  to  central  fueling
stations  and  would  not require development  of  a  significant
fuel   delivery   infrastructure.   Thus,  promising   areas  for
potential  increased  future use  of  dedicated  CNG  heavy-duty
vehicles include  many  municipal applications such as school and
transit  buses,  delivery  trucks, and  trash  collection  trucks.
Other  centrally   fueled applications  also exist in  the private
sector   for   such areas  as  utility  service   vehicles,  local
delivery trucks,   etc.   Dual-fueled  vehicles  may  see  even
broader  usage than  this because of their ability  to operate on
both  CNG and conventional  fuel.    However,  as noted earlier,
most current  interest  seems  to  be  directed  at  dedicated vehicle
development.

-------
                              2-13

                           TABLE  2-2

              New CNG Heavy-Duty  Vehicle Projects
Type of vehicle:   CNG-dedicated transit bus
Number of vehicles:   1
Location:  Phoenix,  AZ
Participant:   Phoenix Transit
Sponsor:  Southwest  Gas Corporation
Comments:  Southwest Gas  has  given  the  "okay"  to  order  the
Cummins  engine for  the  bus;  Southwest Gas will  lease  bus at no
cost to Phoenix Transit for one or two years.
Source:  Southwest Gas Corporation
Type of vehicle:  Trash collection trucks
Number of vehicles:  8
Location:  New York City/Staten Island
Participant:  Snug Harbor Cultural Center
Sponsor:  Brooklyn Union Gas Company
Comments:  None
Source:  Natural Gas Vehicle Coalition
Type of vehicle:  Heavy-duty trucks
Number of vehicles:  50
Location:  New York City
Participant:  NYC Department of Parks
Sponsor:  Brooklyn Union Gas Company
Comments:  None
Source:  Brooklyn Union Gas Company
Type of vehicle:  Postal vehicles
Number of vehicles:  50
Location:  New York City/Staten  Island
Participant:  Post Office
Sponsor:  Brooklyn Union Gas Company
Comments:  None
Source:  Brooklyn Union Gas Company

-------
                              2-14

                           TABLE  2-2

          New CNG Heavy-Duty Vehicle Projects  (con't)
Type of vehicle:   Transit bus
Number of vehicles:   10
Location:  Southern California
Participant:   Southern California Transit Districts
Sponsor:  Flxible Corporation
Comments:   Contract  is  on  back-order  because of  performance
testing  to  be conducted on  existing  converted  buses  and  the
volume of conversion contracts.
Source:  Flxible Corporation
Type of vehicle:  Transit bus
Number of vehicles:  3
Location:  Fort Worth, TX
Sponsor:  Flxible Corporation
Comments:   Contract  is  on  back-order because  of  performance
testing  to be  conducted  on existing  converted buses  and the
volume of conversion contracts
Source:  Flxible Corporation
Type of vehicle:  Transit bus
Number of vehicles:  2
Location:  Dallas, TX
Sponsor:  Not available
Comments:   Contract  is  on  back-order  because  of  performance
testing  to  be  conducted on  existing converted  buses  and  the
volume of conversion contracts.
Source:  Flxible Corporation
Type of vehicle:  Transit bus
Number of vehicles:  5
Location:  New Jersey
Sponsor:  Not available
Comments:   Contract  is  on  back-order  because  of  performance
testing  to  be  conducted  on existing  converted buses  and the
volume of conversion contracts
Source:  Flxible Corporation

-------
                               2-15

                            TABLE  2-2

           New CNG Heavy-Duty Vehicle Projects  (con't)
Type of vehicle:   Transit bus
Number of vehicles:   2
Location:  Los Angeles,  CA
Participant:   L.A. County Transportation Commission
Sponsor:  Southern California Gas Company
Comments:  None available
Source:  American Gas Association
Type of vehicle:  Transit bus
Number of vehicles:  5
Location:  Rochester, NY
Participant:  Rochester-Genesee Regional Transportation Authority
Sponsor:  UMTA (75% of cost)
Comments:  None available
Source:  Rochester Gas & Electric, Rochester, New York
Type of vehicle:  Transit bus
Number of vehicles:  23
Location:  Buffalo, Syracuse, and Long Island, NY
Participants:   Transit  authorities  in   Buffalo,  Syracuse,  and
Long Island, NY
Sponsor:  UMTA (75% of cost)
Comments:   Buffalo,  Syracuse,  Long Island,  and  Rochester   (see
previous  entry)  transit  authorities  are  involved  in  a group
purchase   of  28   natural  gas-fueled   transit  buses.     UMTA
assistance  totaling  $6.3 million will  pay for  75%  of the cost,
with state and local funds providing a 25% share.
Source:  Department of Transportation
Type of vehicle:  Transit bus
Number of Vehicles:
Location:  Tacoma, Washington
Participant:  Pierce Transit
Sponsor:  UMTA  ($8.3 million)
Source:  New York Times, March  1989

-------
                              2-16

                      References Chapter 2


     1.     Pietro    Magistris,     "Compressed    Natural    Gas
Distribution   System   in    Italy,"    in   Symposium   Papers:
Nonpetroleum   Vehicular   Fuels   II,    presented   in   Detroit,
Michigan, June 15-17, 1981.

     2.     American  Gas  Association,   "Natural   Gas  Vehicles
Bulletin," Cat. No. G92238,  1989.

     3.     Conversation  between   ICF   Incorporated   and   Jeff
Seisler, Natural  Gas Vehicle Coalition,  October  1989.   Phase I
engine  tested  at  Southwest  Research  Institute  with  oxidation
catalyst.

     4.     Gross   and   Weinstein,  "The   Economic   Costs   and
Benefits of Proposed Amendments to the  Texas  Clean Air Act,  the
State  Purchasing  and  General  Services  Act, the  Metropolitan
Rapid   Transit   Authority   Act,  the   Regional   Transportation
Authority Act,  and the City  Transportation  Department Act"  for
the Texas General Land Office, March 1989.

-------
                           CHAPTER 3

               CNG HEAVY-DUTY VEHICLE TECHNOLOGY
I.    Introduction

     In this chapter the hardware,  emissions  and performance of
heavy-duty  CNG  vehicles  will be  discussed.   First,  a  brief
discussion  of the  properties of natural  gas as  a  vehicle fuel
will be presented,  followed by a description of  the hardware on
current  dual-fuel   and  dedicated  CNG  vehicles.   Next,    the
emissions   and  performance   of   current  technology  will   be
presented.    Finally,   projections    of   the   emissions   and
performance of  future  optimized dedicated CNG  engines will  be
derived.  The emissions information  presented here  will  be used
in Chapter  5  for  a comparison of  the environmental  impacts of
heavy-duty   CNG  vehicles   and  their   gasoline   and   diesel
counterparts.
    k
II.   Fuel Properties of Natural Gas

     The  most  fundamental  difference between  natural gas  and
conventional motor  fuels  (i.e.,  gasoline, diesel fuel)  is that
natural  gas,  unless cryogenically  stored  (i.e.,  at  extremely
low  temperatures)  under  pressure,   is  in  gaseous  form  rather
than  a liquid.   As  can be seen from the  fuel properties  in
Table  3-1,  the  energy  density (Btu/gal)  of natural gas as it is
stored on a vehicle is very  low  compared to  liquid  fuels.   This
has  significant  impacts in the  area  of  onboard fuel storage as
will be discussed shortly. •

     The composition of natural gas varies,  as  can  be seen from
Table  3-1.   The  methane  content  of  natural  gas  is typically
over  90 percent  although  it  can be lower.   This  results  in
unburned  hydrocarbon  (HC)  emissions  that are  largely methane,
as  will be seen later.   In terms  of ozone  forming potential,
methane  is  essentially unreactive.   Thus, on a mass basis, CNG
HC  emissions  are much  less ozone forming that  the  HC emissions
from  gasoline and  diesel  engines.   This is discussed further in
Chapter  5.

      Two  of the more important properties of fuels in relation
to  engine design  are the octane and  cetane ratings.  The  octane
rating of  a fuel  is  a  measure  of  its resistance to  knock
 (spontaneous  combustion  away  from  the spark-initiated  flame
front).   Good  antiknock   properties are   important  in  spark
 ignited engines because  they  allow  for  increased  compression
ratios and a resultant  increase in  engine  efficiency.   As is
shown  in  Table  3-1  the   octane   rating  of   natural  gas  is
significantly higher than for gasoline.

-------
                                            3-2

                                         Table  3-1

                Properties of Natural Gas and Conventional Petroleum Fuels
Properties

Chemical
Constituents
Boiling Range
 (°F 9 1 atm)

Specific Gravity

Btu/ft3 of Mixture
  (LHV)

Btu/gal (LHV)
Btu/lb   (LHV)
                     Commercial
                     Unleaded
                     Gasoline

                     Mixture of
                      Hydrocarbons
                      (chiefly
                     80 to 420
No. 2 Diesel
    Fuel

Mixture of
 Hydrocarbons
 (chiefly
     -C20>
320 to 720
                     0.71  to 0.78[2]    0.79  to  0.88**

                     95.5****           96.9****
                      114,132
                      18,900
129,400
18,310
   Typical
     CNG
90-98% Methane
 Remainder,
 Ethane and
 Other
 Paraffins,
 C02,  H2,
 He,  N2

-259*
                           0.13***

                           87 .0****
19,760 @ 2400 psi
70°F

21,300
Octane Number
  (R+M)
    2

Cetane Number
  Range

Stoichiometric
  Air/Fuel Ratio
  (weight)

Pealc  Flame Temp
                      87-93
                      5-20
                      14.5  to  15.5
                      3900-4100
N/A
 38-51
 14.5  to  15.1
                                                                    120*****
N/A
17.2
                            3410
 **
 ***
       Pure Methane.   Other  minor constituents  (ethane,  propane,  etc.)  boil at
       higher temperatures.
       At 60°F with respect to water  at 60 F.
            o                             o
       At 80 F with respect to water  at 60 F.
****   At  Stoichiometric  gaseous  air/fuel ratio,  14.7  psia,  60°F,  lower  (net)
       heating value.
*****  Octane number  ratings  above  100  are correlated with  given concentration
       of tetraethyl lead in iso-octane.

Source:  EA-Mueller,  Inc.  and "Gaseous  Fuel  Safety  Assessment
         for Light-Duty Automotive Vehicles," M.C.  Krupka,  A.T.
         Peaslee,   and  H.L.   Laquer,    Las   Alamos   National
         Laboratory, November 1983.

-------
                              3-3

     The cetane rating of a fuel  is a measure of its  ability  to
autoignite when  compressed and heated.  A  high cetane fuel  is
essential for  the proper  operation of  a  compression  ignition
(i.e.,  diesel)   engine  where  no  external   ignition  source  is
used.   Although  no  actual  cetane  number  is  available  for
natural gas,  its cetane rating  is very low  compared to  diesel
fuel.   This  is  reasonable given  that  in  a sense   octane  and
cetane  are  opposites and  natural  gas  has   a very high  octane
rating.  For  this reason  natural  gas  generally cannot be used
in  an internal  combustion engine without  an external  ignition
source  (which  could be  a  spark  plug  or a  pilot  injection  of
diesel fuel).

Ill. Current Heavy-Duty CNG Technology

     A.    Engines

     As  noted  in Chapter  2,  CNG  has  been  used  on  a  limited
scale as a heavy-duty vehicle fuel  for many  years  in  the  United
States.  Generally,  a conversion kit consisting of fuel storage
cylinders, high pressure  fuel  lines,  fuel   pressure  regulation
equipment and  some type of  fuel-air mixer  has  been  retrofitted
onto  a vehicle  originally designed to  operate on   some  other
fuel,  usually gasoline.   Most  often the capability  to  operate
on  the  original  fuel was  retained,  giving  rise to  a dual-fuel
vehicle which was clearly  not optimized for  CNG use.

     The  operation  of  gasoline-derived dual-fuel vehicles  is
somewhat  different  than  for  diesel-derived versions.   Since
gasoline  engines  already  have   a spark  ignition system,  the
dual-fuel gasoline/CNG engines are  designed  for  the  use of only
one  fuel  at a  time and usually  have  a  switch  on  the dashboard
to  allow  the  driver  to  choose  which  fuel  to  run  on.   In
contrast   to   gasoline   engines,   diesel   engines   rely   on
compression  of   the  charge  to   autoignite  the  fuel.   Since
natural  gas  has low  cetane  ratings,  compression  ignition  is
difficult  to  initiate  and  control  and an external  ignition
source  is  required.   Generally,   in  conversion  situations,   a
pilot  injection of  diesel fuel is  used  to  initiate  combustion.
With this type  of conversion the engine retains the ability to
run on  pure  diesel  fuel  or  varying  amounts   of natural  gas
replacing  diesel fuel (up to 90  percent at high loads)  but not
on  pure natural gas.[l]   In  any  case,  as  will  be described
further  below,   to  date  very   little  effort  has  gone  into
accurately  characterizing  the   emissions,  performance,   or  the
long  term   durability   of  converted  heavy-duty  CNG  powered
vehicles.

      Only   recently  has   serious   effort    been   directed  at
developing  heavy-duty vehicle  engines dedicated  and optimized

-------
                              3-4

for CNG,  taking  advantage of  the unique  qualities of  natural
gas as  a  vehicle  fuel.   In this regard, there  are  two distinct
approaches being  used,  which this report  will characterize  as
"lean-burn   combustion"   and   "stoichiometric   combustion."
Lean-burn  combustion  engines  have been  derived  from  current
diesel  engines,   which  operate  with  a substantial  amount  of
excess  air  in  the  combustion  chamber,   giving  rise  to  the
lean-burn  designation.    The   stoichiometric   approach,   which
utilizes  a  chemically  correct fuel-air mixture with no  excess
air, has  been derived  from gasoline-fueled engine designs using
a three-way catalyst control system.   Such  engines  must operate
very close  to  ideal,  stoichiometric, air-fuel  mixtures for the
three-way catalyst to  perform properly.

     Each  of  these combustion  systems  has  its  advantages  and
disadvantages,  and both are  considered  good candidates for more
widespread  use  in  various  heavy-duty  vehicle  applications.
Although  both  are spark-ignited, homogeneous  charge  combustion
systems, their engine-out  emissions and catalyst  strategies are
different  and  result   in  characteristically different  tailpipe
emissions.  Since there are  characteristic differences  in both
engine  hardware and emissions performance,  the two designs will
be  evaluated  separately  in  this  report.   Also,  since  the
lean-burn  combustion  system  is  derived   from  and  will  most
likely   replace   diesels   in   future   uses,   diesel   engine
performance  will  be  the  baseline   for   comparison  with  CNG
lean-burn  combustion   engines.    Similarly,  evaluation of  CNG
stoichiometric  combustion  engines relative  to  current  fuels
will be- based  upon gasoline-fueled designs.   Even though this
convention will  be carried through the rest of this  report,  it
should  be born  in  mind  that there  is no fundamental  reason
preventing  the  eventual   crossover  of future applications  of
lean-burn and stoichiometric combustion designs.

     It  is  also  worth noting   at  this point that  there  are
different  emissions  testing  cycles  for  current  diesel  and
gasoline-fueled heavy-duty engines.   These cycles  are  used  to
reflect  the characteristic  differences between typical  usage
patterns  for  these engines.  When comparing  emissions between
lean-burn  and  stoichiometric CNG combustion designs,  a single
cycle will be used for both,  to  insure  compatible results.  The
diesel  cycle has  been  chosen for this purpose because  it is the
only  cycle  for  which  data  on both  designs  exists.   The
evaluation  of • environmental benefits  in  Chapter  5,  however,
will  be   based   upon   the  diesel  cycle  when   comparing  to
conventional  diesel   engines,   and   the   gasoline  cycle  when
comparing to conventional  gasoline-fueled engines.

     As   just  noted,   the  lean-burn  dedicated  CNG  engines
currently  being  developed  are  based   on   existing  diesel

-------
                              3-5

engines.   They are usually run  at  a relative fuel-air ratio  of
around 0.7,  although  fuel-air ratios well below this  have  been
demonstrated by Southwest Research  Institute using  a  prechamber
with  a  stoichiometric  spark-ignited  charge  and   a  very  lean
mixture  in  the main  chamber.[2]   This   latter  approach  holds
promise to  improve  the efficiency/NOx tradeoff  associated  with
lean-burn  engines.    In  general  the  open  combustion  chamber
approach  utilizes   intake  throttling  for power  control.   The
prechamber  approach  is unthrottled except at  idle and  relies
upon  control  of  the  amount of  fuel  being metered  into  the
chamber through check valves  for power control.  In either  case
the  low  cetane number of  natural   gas means  a strong spark  is
needed for ignition in dedicated CNG engines.

     Although  the  antiknock  gualities of  natural  gas are  very
good,  the   compression  ratios   of  current   diesel  engines
(approximately 19 to 1) are generally too  high  to  prevent knock
in  a  dedicated  CNG  engine.   Thus,  the  compression  ratios  in
lean-burn CNG  engines are generally  reduced  to below 15  to 1
and  can  be as low  as 10 to  1.   The lowest  compression ratios
are  conservative designs to  account for  the wide range  of  fuel
composition which the engine might see in-use.

     In  contrast  to  lean-burn   engines,  stoichiometric  CNG
engines  are  generally based  on   gasoline  engine  technology.
They have  used a 3-way  catalyst based control  system,  with or
without  the  closed  loop  controls characteristic  of  current
gasoline'engines.   The relative fuel-air  ratio is  ideally held
extremely   close  to   one   so  that  the  3-way   catalyst  can
simultaneously   oxidize   THC  and   CO   while  reducing  NOx.
Throttling  of  the  fuel-air  mixture is used  to control  power.
As  was previously  mentioned, the  antiknock  characteristics  of
natural  gas are much better than gasoline.   As  a  result the
compression  ratio  of a gasoline engine can  be boosted somewhat
when converted to dedicated CNG use.

     The  fuel-air  mixer  is  a  central   component   in any CNG
engine configuration.   The mixer  serves  the  same  function on  a
CNG  engine as a carburetor   or  fuel  injectors  on  conventional
engines,  i.e., metering  the fuel  into  the  air at  the proper
fuel-air  ratio for  proper  operation.  Precise control  of the
fuel-air  ratio  over  the  entire  speed  and  load  range  of the
engine is essential  in achieving good performance,  fuel economy
and  emission characteristics.

     Common mixer  designs  today   are  generally mechanical  in
nature  and meter  the  fuel   as   a   function  of  intake   air
pressure.   A common mixer design is shown  in Figure 3-1.  In
addition to this spring-loaded  diaphragm  variety,  venturi  based
mixers are also  available.   These  current technology  mixers  are

-------
                                3-6
                            Figure 3-1
                                        r=J3
             INTAKE (DOWN) STROKE
                 Diaphragm Operated Air-Gas Mixer
Source:   IMPCO Master Catalog,  1987

-------
                              3-7

not nearly  as  sophisticated as the  fuel  metering equipment  on
today's conventional engines.   The less developed nature  of  CNG
technology  in  conjunction  with  the  difficulty  in  general  of
precisely metering a gas combine to produce current  CNG engines
which do not have  the  precise  control of  fuel-air ratios needed
to  truly optimize  performance and  emissions.   Although  much
progress  is  being  made  in the  area of  fuel metering  for  CNG
engines,  including  the  introduction  of  simple  closed  loop
control systems, which utilize  an exhaust oxygen sensor (lambda
sensor)  and  feedback  controls  to   maintain  proper   fuel-air
ratios,  and  the  development   of  fuel  injection  to  replace
mixers, there is still a great  deal of work to be done to bring
the level  of  CNG fuel metering technology  in  line with current
gasoline and diesel technology.

     B.    Fuel Storage

     The   onboard   storage  of   fuel  for   CNG  vehicles   is
significantly  different  than for  vehicles operating  on liquid
fuels.   The natural  gas  is  stored  in  gaseous form in  high
pressure  cylinders at  pressures  up  to  3,000 psi.  The number
and size of the cylinders mounted on  the  vehicle determines the
amount  of fuel  stored  on the vehicle  and thus  its range.   On
heavy-duty  trucks  the cylinders  are  often mounted  underneath,
as shown in Figure 3-2.

     Currently,  the  weight and  bulk  of  CNG  cylinders  is  a
concern with CNG vehicles, both in terms  of  available  space for
storage   and   reduced   performance   and  fuel  economy  due  to
increased   weight.    However,   this   concern  appears   to   be
primarily  related  to  light-duty  applications  and  does  not
appear  to present  as  much difficulty  in the heavy-duty area.
Conventional  steel  cylinders   filled  with  natural   gas  weigh
roughly  five times more than diesel  or gasoline  tanks and fuel
on  an  energy  equivalent basis.   Also, a given volume  of natural
gas at  3000  psi contains only about  one  fifth of the  energy of
the  same volume of diesel fuel. [3]   The  effect  this has on the
weight and range of heavy-duty vehicles depends  a great deal on
the application.   As  will be shown in Chapter 5, putting enough
storage  capacity   on  a  heavy-duty  vehicle  for  it  to  have
equivalent  range on CNG  as  its gasoline  or  diesel  counterpart
increases vehicle  weight  6.5-9.4  percent.   In many applications
this  is not  a problem.   However,  for transit  buses   increased
weight  may  result in  a  de-rating  of  the  passenger carrying
capacity  of  the bus.   Thus, there is a trade-off between weight
penalty  and vehicle  range  which must  be  examined   for  each
application.   Finally,  CNG fuel tanks are presently constrained
to   cylindrical   shapes   and   do   not  offer   the  packaging
flexibility that is available with liquid fuels.

-------
                                3-8

                            Figure 3-2
                                  I  AAiOCIMf TANK
                         TAW MOUOTnM MACMT*
          Typical Mounting Locations For CNG  Cylinders
             on  School  Buses and Medium-Duty  Trucks
Source:  Nu-Fuels,  Inc.

-------
                              3-9

     Although most  cylinders  currently  being used  are of  the
plain   steel  variety,   lighter   weight   designs   including
fiber-wrapped steel  and fiber-wrapped aluminum are  beginning to
see some  commercial  use.   Advanced  all-composite  cylinders  are
also being developed.  These designs  offer  much improvement in
the weight  to energy  ratio over  plain steel  cylinders as  is
shown in Figure 3-3.   However,  Figure  3-3  shows that  even  the
best designs  still  have a  2:1  weight disadvantage  compared to
gasoline  and diesel fuel.   Since  this weight  penalty  arises
from the storage cylinders  and  not the fuel itself, the area of
fuel storage system weight  and  bulk reduction,  through  the  use
of   increasingly   lighter  materials   to  reduce   weight   and
adsorbent  technology  to  reduce needed storage  volume  and/or
pressure,  is  an  area where additional  work can yet  be done to
improve  the  attractiveness  of   CNG  as  an   alternative   to
conventional motor fuels.

     Natural  gas  may  also be  stored  onboard  as  a  cryogenic
liquid  (LNG).   The resultant  improvement  in  the  volumetric
energy content of LNG  as  compared  to CNG may result in improved
range or performance due to lower  storage weight and volume  per
unit of  energy  stored.  However,  because natural  gas liquifies
at -259°F, the onboard storage  vessels must be  well  insulated.
Even   then   an   unused   vehicle   must   vent   boiloff   gas
periodically.  While this has been  reported to be  as  frequent
as  every 7-8 days[4},  the American Gas Association indicates
that current LNG  vehicles can hold  LNG without boiloff  for up
to  three  weeks.   The   costs   associated   with  liquifaction,
refueling, 'and storage hardware for  LNG make  it  less attractive
at  present,  although  this   could  change  with  advances  in
technology and changes in fuel economics.   The vast majority of
natural   gas powered   vehicles   today  have  fuel  stored  in
compressed,  rather than liquified  form.   Therefore, this report
will only address CNG.

     C.    Emissions From Current Technology CNG Vehicles

     1.    Dual-Fuel Vehicles

     The  actual  emissions of  current technology  CNG-powered
heavy-duty engines are not well characterized and  there is  only
an  extremely limited amount of transient emission test results
available.*   Turning  first to  dual-fuel vehicles,  EPA  is  not
     The EPA  heavy-duty  transient  test (as defined  in the U.S.
     Code  of  Federal  Regulations,  Title 40,  part  86)  is the
     standard  engine  test   for   heavy-duty  engine  emissions
     certification.    The  engine   is  placed   on   an  engine
     dynamometer  and  run  through  a  standardized  test   cycle
     which simulates in-use engine operation.

-------
                                 3-10

                             Figure 3-3
            Comparison of Fuel and Storage System Weights
                           For Various Fuels
                          Compressed natural gas
           Plain steel       Wrapped aluminum       Methanol          Gasoline
                 Wrapped steel       All-composite          Diesel
                              Fuel
Storage
Source:     "Natural  Gas Vehicles:
            Art",  Sierra  Research
            1989.
 A Review  of the  State of  the
Report  No.  SR89-04-01,   April

-------
                              3-11

aware of  any  heavy-duty transient testing  ever performed  on  a
dual-fuel  retrofit  engine.    Presumably,   conclusions  can  be
drawn,  at  least  for  stoichiometric  heavy-duty  gasoline/CNG
dual-fuel  engines,   based  on   limited  testing   of   similar
light-duty dual-fuel  vehicles.   These  results show  that  the
vehicle's operation on CNG can yield large reductions in CO and
the  same  or  somewhat    lower   non-methane   HC   compared  to
operation  on  gasoline.    However,   CNG  also  yields  somewhat
higher  NOx  and  much  higher  methane  emissions"  compared  to
gasoline.[5]    Also,  the  emissions   can   vary  a  great  deal
depending  on  the  quality  of  the  conversion  and  the state  of
tune  the vehicle  is  in.   Reductions  in  CO  are  by no  means
guaranteed,  as   is  evidenced  by  testing  of   three  light-duty
dual-fuel  retrofit  vehicles  done   at  the  EPA  Motor  Vehicle
Emission Laboratory.   Two  of  the  three vehicles  tested  in  as
received condition  had CO emissions  two to eight  times higher
on natural gas than on gasoline.   Subsequent  recalibration and
maintenance,  however, yielded a substantial CO benefit.[5]

     Recently,  the United  Parcel Service  (UPS)  converted ten
parcel  delivery  vehicles  to  dual-fuel  gasoline/natural  gas
operation.  A diagram of  the vehicle  is  shown in Figure 3-4.
During  the two  year  UPS  program  the vehicles   will  be  run
exclusively  on  CNG  when  possible  and  information  on  fuel
economy and maintenance will be collected.

     Emission  testing   results   of   one  of   these  converted
vehicles  are shown in  Figure 3-5.   Although  the  data  shows
emissions  to be lower  on  natural   gas than gasoline,  these
results  are of  limited use in the context of  this  study for two
reasons.  First,  the  test  cycle used was a  chassis dynamometer
test  in  which the whole vehicle was run over  a cycle simulating
New York  City driving conditions.   This means that comparisons
between  the  UPS  vehicle  and  engines  tested  over  the  EPA
heavy-duty engine transient test cycle cannot be made.  Second,
no   baseline  data   of   the   vehicle  before   conversion  is
available.   Thus,   although  the data  shows  the  vehicle  to  be
cleaner  on CNG  than gasoline,  it  is  unknown what  effect,  if
any,  conversion had  on gasoline  emissions.   Past  testing  on
light-duty vehicles has shown that  the addition of a dual-fuel
CNG conversion  kit to a gasoline vehicle can degrade emissions
performance   on   gasoline.[6]    Nevertheless,   the  data  is
interesting  as   a  reference  and the UPS  program  promises  to
provide  some much  needed data  on  the  durability, maintenance
requirements  and emissions deterioration of CNG vehicles.

      Turning  to  lean-burn dual-fuel  CNG  engines  with diesel
fuel  pilot  ignition,  the  emission  benefits  of  this  type  of
system  are also unclear.   Steady-state testing of  a converted
Caterpillar  3406  large  truck engine  with  electronic injection

-------
                   Figure 3-4
UPS NATURAL GAS CONVERSION
    United Parcel Service
                                                     I
                                                     I—'
                                                     r-o
      (I) THREE NATURAL GAS CYLINDERS EQUALING 15 GALLONS
      (2) TWO REGULATORS TO REDUCE PRESSURE
      (3) NATURAL GAS-AIR MIXER
      (4) GASOLINE CARBURETOR
      (5) GASOLINE ENGINE
                           UNI II I) I'AIICI I SI HVIl.l ,'HIHK)KI YN UtJIOIJ t.»V, ( I )M|'AI| i

-------
                   Figure 3-5
EMISSIONS OF UPS PACKAGE CAR
             WHILE MOVING

               CARBON
               MONOXIDE
(grams per mile)

   NITROGEN
   OXIDE
               44.9
               62
               H
                        7.9
                  6.5
                          5.9
               REDUCED 85%
   REDUCED 25%
CARBON
DIOXIDE
792.8



                                  612.2
REDUCED 23%
HYDRO-
CARBONS
                                         2.6

                                            2.3
REDUCED 13%
                                     GASOLINE
                                           GAS
                                                   OJ
                                                   I
                       SOURCE UNITED PARCEL SERVICE/ nROOKLYN UNION GAS COMPANY

-------
                              3-14

of both CNG and the  diesel pilot  showed  mixed results,  with a
reduction in  NOx,  little  or no  effect  on HC  and  particulate,
and  a  significant  increase  in  C0.[l]    With  little  useful
emissions data on  these types  of conversions, it  is  difficult
to predict any emissions  benefit with any assurance and  it  is
clear  that  more  testing  of  heavy-duty  dual  fuel  conversions
must take place before any emission benefits  attributed to them
can be accurately quantified.

     2.    Dedicated CNG Vehicles

     As was  previously discussed,  the properties  of  CNG  as a
vehicle  fuel  are  quite  different from  those of  gasoline  or
diesel fuel.   Because of this, dual-fuel  CNG engines  cannot  be
fully optimized for  natural  gas  operation,  because compromises
in engine design  have to  be  made in  order to operate on both
fuels.  At  the same time, dual-fuel  vehicles clearly have  a
place in the  transition  to clean alternative fuels because they
offer  the  flexibility  to  operate  on  conventional  and  CNG
fuels.  This  ability will be especially  valuable  in easing the
transition to dedicated CNG use while  the CNG refueling station
infrastructure is  further developed.   However, due to  the lack
of substantial emissions data on  dual-fuel engines, this report
is  not in  a  position to  quantify the  benefits  of  dual-fuel
vehicles (other than to  say  that  they will be less  than those
of dedicated  vehicles) and will  focus principally on dedicated
vehicles.    Since  lean-burn  and  stoichiometric  engines  have
somewhat  different   emission  characteristics  they   will  be
treated separately, beginning with lean-burn.

     a.    Lean-Burn Combustion

     Cummins Engine  Company  is  presently  developing a  dedicated
CNG version of  its L-10 heavy-duty diesel engine for commercial
introduction  into  the bus engine market   in  1991.   To  date, two
different   configurations   have  been    tested.    The    first
configuration   utilizes    a    mechanical    diaphragm   mixer
manufactured  by  IMPCO  Carburetion,   Inc.  similar  to   the one
shown  in  Figure 3-1.   This  system has  separate  idle  and full
power adjustments  but is  an open-loop (i.e., one which does not
utilize feedback from  an  exhaust  sensor  for automatic  fuel-air
ratio  control)  mechanical  system.    The  second   configuration
utilizes  an  open-loop electronically  controlled  venturi-type
mixer  manufactured by  TNO Road Vehicles  Research  Institute  in
Holland.  A diagram  of a  similar system  is  shown  in Figure
3-6.   Both  configurations utilize intake  throttling  for  power
control  and an  oxidation catalyst  for   HC  control.    Transient
test  results  are  not publicly  available for  the  two  CNG L-10
configurations.   However, Cummins has  released  its   emissions
design targets, and  based upon EPA's review  of the confidential

-------
                               3-15
                           Figure 3-6
               Catalyst
V
Dj
d

Engine

r-— 	 '
Mixing unit






      Open Loop Electronically Controlled  Air/Fuel System
                     With Venturi-Type Mixer
Source:  TNO Road-Vehicles Research  Institute

-------
                              3-16

transient test  data  EPA is confident  these  targets  will likely
be  met  or  exceeded.   Cummins's  design  targets  are shown  in
Table 3-2.[7]

     Also  included  in  Table 3-2 are  emission  results for  a
natural gas-fueled Caterpillar 3406.   This  engine also utilized
an  open-loop  IMPCO   system  and  had  no  catalyst.[8]   It  was
developed for steady-state electrical  cogeneration purposes and
was  not optimized  for  transient operation.[9]   Nevertheless,
these data are useful in analyzing emission  trends of lean burn
heavy-duty CNG  engines,  especially given  the limited amount of
data available.

     Using  the   information  on   the   Cummins   L-10   and   the
Caterpillar  3406,  emission  levels  which,  for  the  purposes  of
this  report,   will   represent   current   lean-burn  combustion
technology  emissions  were  developed.    For  the  most  part,
Cummins'   design  targets   are   assumed   to    represent   the
capabilities  of  current technology.   Some modifications  and
additions were  made  to  the  design targets  as  Cummins  did not
specify  CC>2/  non-methane  HC  or  formaldehyde  levels  in  its
targets.

     The CC-2  and  formaldehyde  values  shown in  Table  3-2  for
current  technology  were  chosen  to   to   be representative  of
current  technology  based  on  energy   consumption data  on  the
Caterpillar  3406  as  well   as  confidential  data  on the  L-10
engine emissions  and  catalyst efficiencies.   The non-methane HC
was  assumed  to  be ten percent of total hydrocarbon, based upon
the  Caterpillar  data.   The  use  of a catalyst could  affect  this
number  somewhat,  but to  an unknown  degree.   Experience  with
conventional catalysts on light-duty CNG  vehicles suggests that
the  non-methane  fraction would  be decreased.  However, the L-10
CNG  catalyst has been optimized  for total  hydrocarbon control,
meaning  that  it  has significantly better  methane control  than
conventional   catalysts.    It    would   therefore   affect   the
non-methane fraction much less, if at all.

     Concerning   the  particulate  emission   level   chosen  to
represent  current technology,  the  Cummins  design   target  was
used  even  thought  the  Caterpillar  engine  showed  particulate
emissions  ten  times  higher  than this.   There  is  no   inherent
reason  why CNG  should  have high particulate  levels  in and of
itself,  given   that  fuel-derived  particulate  is  generally
attributed  to  heavy,  long  chain  hydrocarbons  which  are not
found  in  natural gas.   The  higher   particulate  rate  on  the
Caterpillar  engine   has  been  attributed   to   lubricating  oil
passing  the  piston  ring  pack  and  entering   the  combustion
chamber.   This   assumption was  supported  by the  fact  that the
particulate   was  90   percent   soluble   organics,   generally

-------
                                     3-17

                                  Table 3-2

                    Summary of Low Mileage Emissions From
                    Lean-Burn CNG Engines Operated Over the
                   EPA HDDS Transient Test  Cycle (q/BHP-hr)
Pollutant

THC
NMHC
NOx
Part.
CO
C02
Formaldehyde
Cummins CNG L-10
Desiqn Targets
0.9
	
4.5
0.06
4.0
• — •—
Cat. 3406
w/o catalyst
9.2
0.84
4.1
0.60
3.2
0.34
1991 Diesel
Standards
1.3
	
5.0
0.25/0.10*
15.5
_ _ _
  Assumed
Current CNG
Technology

     0.9
     0.09
     4.5
     0.06
     4.0
     575
     0.05
     The  0.10  standard  applies  to  urban  bus  engines  only.
     standard applies to all other heavy-duty diesel engines.
  The  0.25

-------
                              3-18

associated  with   lube  oil.    Improved  oil   control   is   a
significant  part  of  current   efforts   to   reduce  engine-out
particulate levels  for  1991  diesels, and  low oil  consumption
designs have been successfully  incorporated  into  diesel engines
in  these  attempts.   Thus,  there is  no  reason to  believe than
lean-burn  combustion CNG  engines  will  have  high  particulate
levels.

     b.     Stoichiometric Combustion

     The  only  available transient  engine emissions data  for  a
dedicated  Stoichiometric  heavy-duty  CNG  engine  is  shown  in
Table 3-3.  This  engine  was  built from a Chevrolet 454 gasoline
engine and was converted to dedicated  CNG use  by  Brooklyn Union
Gas  (BUG).   This engine was tested at  the  EPA Motor Vehicle
Emission  Laboratory (MVEL)  in  Ann  Arbor,  MI.   Although  the
results from  two  configurations are  shown here,  only the IMPCO
mixer configuration was introduced  into the two buses which are
presently  operating in  New  York  and,  thus,   this engine  was
chosen  to  represent  current   Stoichiometric  technology.   The
hardware  for  the  BUG engines  is similar to that described for
the Cummins L-10, with  the addition  of  a lambda sensor  in the
exhaust stream of the  TNO  configuration for closed loop control
of the Stoichiometric fuel-air mixtures, as shown in Figure 3-7.

     During the  testing of the BUG engine  at  MVEL  tests were
run  over  both the  diesel engine test  cycle  and  the gasoline
engine  test  cycle.   Although, these  cycles  are  different,  the
results from  both are useful  and are shown in Table 3-3.  The
results from the diesel test cycle will be used to  compare this
engine  to the  lean  burn  CNG  engine while  the gasoline test
cycle results  will  allow  a  direct  comparison  to  the gasoline
engine  from which it was  derived.   This  latter  comparison will
be  discussed  in   Chapter  5  while  the  comparison  of  the
Stoichiometric  and  lean  burn   CNG engines will  be  discussed
shortly.

     One  other  point that should be mentioned with  respect to
the BUG data  in  Table 3-3 is that the with-catalyst results for
the  gasoline  test   cycle  were  hot   start  only,  rather  than
composite   results.    Official   transient   test   results  are
weighted  six  hot  start tests  to  one  cold  start  test  and  a
composite  result  is  arrived  at.    No  cold  start  tests were
performed  on the  gasoline  test  cycle with a catalyst and only
the hot start  results  are  shown  in Table 3-3.   However, the hot
and cold  start  data for the without-catalyst tests  were  nearly
identical.  Given this  fact  together with the six to  one  hot to
cold  start ratio, the hot start results  shown  are  assumed to
represent  composite results.

-------
                                        3-19
                                     Table 3-3

                          Summary of Low Mileage Emissions
                         From the Brooklyn Union Gas (BUG)
                        Stoichiometric CNG Engine (q/BHR-hr)
                     IMPCO Mixer
TNO Mixer
Pollutant     w/o  Catalyst  w/catalyst    w/o catalyst  w/catalyst

                               Diesel Test Cycle
Applicable
Standards
THC
NMHC
NOx
Part.
CO
Formaldehyde
3.6
0.82
6.62
0.01
31.9
529
0.03
1.03
0.15
1.33
0.01
10.8
557
0.0008
3.57
0.83
6.87
0.01
23.3
520
0.02
1.01
0.17
1.16
0.01
6.64
540
0.0007
1.3
—
5.0
0.25/0.10
15.5
—
—
                           Gasoline  Test  Cycle**
THC
NMHC
NOx
Part.
CO





C02
Formaldehyde
*
**
***
1.7
0.40
5.75
0.01
28.5
0.72
0.09
0.51
0.01
10.6
1.47
0.34
6.57
0.05
8.16
0.84
0.12
1.46
0.01
4.54
474 500 471 504
0.02 0.0001 0.02 0.0007
The 0.10 standard applies to urban bus engines only.
applies to all other heavy-duty diesel engines.
The without catalyst results are hot start results only,
The less stringent standards apply to engines used in
14,000 Ib. GVWR.
1.1/1.9***
—
5.0
--
14.4/37.



1***
The 0.25 standard
not composite.
trucks greater than

-------
                                 3-20


                             Figure 3-7
                Catalyst
               X Sensor
                                                  Gas
                           Engine speed
                                      X Valve
^
q
d

Engine

r^- 	 	
Mixing unit






         Closed Loop Electronically Controlled Air/Fuel
                  System With Venturi-Type Mixer
Source:  TNO Road-Vehicles Research Institute

-------
                              3-21

     The  assumed emissions  from  current  technology  lean-burn
and stoichiometric combustion  engines  are  shown side by side in
Table 3-4  and graphically for  THC,  CO and  NOx in  Figure  3-8.
As can be  seen,  the  HC emissions of the two engines are similar
while the  NOx  and  formaldehyde levels for the  lean-burn engine
are  higher  than  the  stoichiometric  engine.    It  should  be
recognized that  data on  formaldehyde  emissions  from  lean-burn
CNG engines  is very preliminary, with  data  available  from only
one engine test.  The lean-burn  engine  particulate  emissions
are  also  higher  than those  from  the stoichiometric  engine,
although   both  are   very   low.    On  the   other  hand,   the
stoichiometric  engine  has  substantially  higher  CO  emissions
than the  lean-burn  engine.  Finally, the  stoichiometric engine
has  somewhat  lower  C02  emissions  due  primarily  to  a  higher
amount  of   the   fuel   carbon  being  emitted  as   CO  (energy
consumption  from the  two engines  being  nearly  equivalent,  as
will  be  discussed  below).    A   complete   discussion  of  the
environmental  impacts  of  CNG emissions including  a comparison
with gasoline  and diesel can be found in Chapter 5.

     D.    Engine Efficiency

     The  efficiency  of CNG  engines  is an  important discussion
as  it  relates to  both fuel  economy and cost,  as well as total
C02  emissions   and  relative  global   warming   effects.   Fuel
efficiency  in particular  is  very important in the  heavy truck
market,  especially  among  fleet  and  line-haul  operators,  as   a
large  truck  operator may  spend as  much as  $10,000 annually for
fuel alone.

     A  very  large  determinant  of  engine  efficiency  is  the
compression  ratio,  which  is  in a large part  determined by the
anti-knock  gxiality  (octane)  of  the  fuel  for  a  stoichiometric
engine.   For a  diesel  engine the compression  ratio is high to
initiate  spontaneous  (compression)  ignition.   Thus,   a diesel
engine  has  higher  thermal  efficiency than   a  stoichiometric
gasoline  engine.  When  deriving a  CNG  engine  from  a diesel
engine,  compression  ratio must  be  reduced to  control knock,
reducing  efficiency.   Conversely, when using a gasoline engine
base  for  a  dedicated  CNG engine, the  compression ratio can be
raised  due   to  the  superior  knock   characteristics   of  CNG,
raising   the  thermal  efficiency.   Experiments  with   a single
cylinder  spark  ignition  engine  showed  an  increase  in thermal
efficiency  of   22  percent through  both  increased compression
ratio optimization and a  lean fuel mixture.[10]  However, it is
difficult  to  predict from  this  type  of  work  exactly  what
efficency   improvements   can   be   made    on    a   full   size
multi-cylinder engine.

     The   energy  consumption  rates  for  the  various  engines
previously  discussed  are  shown  in Table  3-5.    The  current
technology lean-burn CNG engine  and the 1990 diesel L-10 values
were  estimated using a carbon  balance  procedure  on the exhaust

-------
                             3-22
                         Table 3-4

        Current Technology Heavy-Duty CNG Lean-Burn
       and Stoichiometric Engine Emissions (g/BHP-hr)
Pollutant             Lean Burn            Stoichiometric

  THC                    0.9                    1.03
  NMHC                   0.09                   0.15
  NOx                    4.5                    1.33
  Part.                  0.06                   0.01
  CO                     4.0                   10.8
  C02                      575                    557
  Formaldehyde           0.05                   0.0008

-------
                              3-23
                           Table 3-5

                 Energy Consumption of Current
              Heavv-Dutv CNG Engines  (3TU/BHP-hr)
	Engine	      Diesel Fuel        CNG          Gasoline

LlO/Current lean-burn    7180           10,000

Cat. 3406                7430           10,800

BUG (Chevy 454)
 diesel test cycle        —             9,950

Gasoline test cycle       —             8,950        10,036

-------
                FIGURE 3-8

 COMPARISON OF CURRENT LEAN BURN AND

     STOICHIOMETRIC THC, CO AND NOx
 EMISSIONS
               g/BHP-hr
           12



           10



            8
              y\
lean burn      2


stoichiometric

            0
                  THC
    CO

POLLUTANT
NOx
                                                    OJ
                                                    I
                                                    co

-------
                              3-25

components.    The  Caterpillar  CNG  3406  is  the  steady  state
cogeneration engine while the diesel  3406  value is typical of a
1988 3406 engine.[8]  The Chevrolet  454  gasoline value  is  also
typical of a current engine.

     As can  be seen  from Table 3-5  the lean-burn CNG  engines
consume  39  to  45  percent  more  energy  than  their   diesel
counterparts.    This  is  not surprising given  that  the  peak
thermal efficiency of the diesel engines is higher  than  the CNG
engines  and  the  presence of  throttling  losses  with the  CNG
engines at low loads.   Conversely,  as was expected, the  BUG CNG
engine  consumes  less  energy than  the  Chevrolet  454  gasoline
engine from which  it  was derived.   In  fact,  an  eleven  percent
reduction in  energy consumption was  achieved.   Over  the diesel
test  cycle,   however,  the  CNG  lean-burn  and  stoichiometric
engines consumed  almost the  same amount of energy.   As  will be
discussed  in  the  section  on  optimized   engine  efficiency,
optimized  lean-burn CNG  engines  are expected  to  consume less
energy than optimized stoichiometric CNG engines.

     E.    In-Use Performance

     The  in-use  emissions  performance  and durability  of  CNG
engines   is   an   area   where   little   information   has   been
collected.  Generally,  in-use  emissions deterioration would be
a  function  of engine-out emissions  deterioration  and catalyst
deterioration.  Since the engine technology used on CNG engines
is  not  fundamentally  different  in  characteristics  affecting
likely   durability   than   that   currently   being   used   on
conventional engines, there  is  no  reason to expect that, as CNG
technology matures, the  engine-out  emissions  deterioration of
CNG  engines  would  be significantly  different  than for gasoline
or  diesel  engines.    It  is  true  that  the  BUG  engine  has
experienced  some   difficulty  maintaining  proper  calibrations
during routine service, causing dramatic increases  in HC and CO
emissions.[11]    Valve  seat recession  has also  been a concern
on  some  configurations.   However,  EPA expects  such problems to
be  readily solvable on  future versions.

     As  for  in-use catalyst performance,   it  is again unlikely
that,  given  the   similarity  of  CNG  catalyst  technology  to
current   in-use   technology,   there  will  be  a   significant
difference  in deterioration.   Catalyst  use  on  future  diesel
engines may  present added durability issues due to the presence
of  particulate in diesel  exhaust and the potential for catalyst
plugging.   However,  the  actual  emissions  impact of catalyst
deterioration  or  of catalyst failure on a diesel  engine would
be  much  less  than  for the CNG  engine,  due  to  the fact that the
emission  reductions being provided by the  diesel catalyst would
only be  relatively modest to begin with,  in  the range of 20 to

-------
                              3-26

30 percent.   A failed catalyst  on a  lean-burn  CNG  engine,  on
the other hand, would  dramatically increase HC,  CO and aldehyde
emissions.

IV.  Optimized Vehicle Projections

     A.    Emissions

     Because   the   development   of  dedicated   heavy-duty  CNG
technology  is  still in  its  early stages there  is  every reason
to believe  that significant  improvements  can be  made  in both
emissions  and  efficiency.   Past  experience  with  gasoline  and
diesel  technology  would  support this.  The  use of  electronic
controls,    fast-burn    combustion    technology,     increased
compression ratios, and  general  engine optimization  would seem
likely  candidates  for improving CNG engines.   Since,  as noted
earlier,  dual-fuel  engines  offer  limited  opportunities  for
optimizing  CNG  combustion,   the projections  of optimized  CNG
engine  emissions  will  be  focused  on dedicated   engines.   It
should  be  noted that the projections  derived here  are based on
limited   current   data   and   assumptions   about   projected
improvements over current  designs.   Actual  improvements made in
optimizing  CNG engines may  yield different  results  than those
projected here.

     It  appears   that  little   of  the  work  on   CNG  engine
development  thus  far  has  been  directed  at  emissions,  as  is
evidenced by  the  lack of data.   In  this  section projections of
the potential  emissions  of optimized CNG heavy-duty  engines of
the mid-1990s  will  be  made.   It should be noted that no data on
advance  concepts  for  CNG  were  available   and  the  projected
improvements   were  based  largely  on  assumptions  about  the
potential   of   different  technologies  and   extrapolation  of
gasoline and diesel vehicle  experience.

     Based  on  confidential  prototype  test  data on the Cummins
CNG  L-10,  as well   as  evaluations  of  the   potential  for
improvements  in engine-out emissions and catalyst  efficiencies,
the  emissions  of  an optimized  dedicated  lean-burn  CNG  engine
were projected.  These projections are shown in Table 3-6 along
with the current technology emissions.  An  improvement  in fuel
consumption was also  projected, as will be  discussed  shortly.
This   improvement   results   primarily  in   CC-2    reductions,
although   its   effects  are  felt  on  all  emissions  through
engine-out  reductions.

     Given  the relative  infancy of  CNG technology as  compared
to diesel  technology  and  the limited amount of work which has
been   done  on catalyst   optimization  for   methane   and  other
hydrocarbon components of CNG  exhaust,   significant  reductions

-------
                               3-27


                           Table  3-6

               Estimates of Optimized Heavy-Duty
           CNG Lean-Burn Engine Emissions  (q/BHP-hr)


                                         1994
Diesel                Current           Diesel   "    Projected
Pollutant            Technology        Standards      Optimized

THC                     0.9               1-3             0.6
NMHC                    0.09              	             0.06
NOx                     4.5               5.0             4.0
Part.                   0.06              0.10            0.05
CO                      4.0               15.5            1.5
(X>2                     575               	             525
Formaldehyde            0.05              	             0.03

-------
                              3-28

in future HC emissions should be possible.  For  this  analysis a
one-third reduction  in total HC was  projected through improved
combustion  and   catalyst   optimization.    The   percentage  of
non-methane HC was  held unchanged at 10 percent  of total  HC as
with  the  current technology numbers.  A  similar reduction in
formaldehyde was projected.

     Smaller percentage  reductions in NOx  and particulate were
projected.   In the  case  of NOx,  there  is   little  basis  for
improvement and,  if anything, future efforts  to improve engine
efficiency and performance could put  upward pressure  on NOx.
In the  case of particulate, any reduction  would likely come as
a  result  of  improved  oil  controls.   Since  1991-type  diesel
engines such  as  that  from  which  the lean-burn  CNG  engine was
derived  have  already  introduced  many  advanced  oil  control
features,  there is little improvement projected.

     Finally,   there  are   reasons   to  believe that  future
lean-burn  CNG  engines  will  have  significantly   lower  CO.
Although the current L-10  CNG engine is meeting  a design target
of 4.0  g/BHP-hr,  there is  evidence  that  CNG  engines  can  reach
CO  levels much  closer to  current diesels (1.5-2.5  g/BHP-hr).
This  fact,  along with the  high  CO  conversion  rate  available
from  oxidation  catalysts,  should  allow  a significant reduction
in future engine CO  emissions.

     Turning to  stoichiometric  combustion engines,  engine-out
emission  and  catalyst efficiency  improvements  were  projected
separately  due to   the  availability  of   data  on  the  current
stoichiometric engine  both with  and  without  the catalyst.  The
current technology  and projected optimized emission  levels for
stoichiometric heavy-duty  CNG engines are  shown in  Figure 3-7.
Considering  first   engine-out  emissions,   for  all  emission
components  except  C02/  which  is largely  a  function  of fuel
consumption, it was assumed that levels could be reduced  by 20
percent   over    current    levels.    This   seems   a   reasonable
assumption  given  the  relatively  young  nature of  dedicated
heavy-duty  CNG   technology  in   comparison  to  the   level  of
sophistication of gasoline engines.   Such  things as  electronic
controls,   more  precise   fuel   metering,   and  spark   timing
optimization  offer   means  of  improvement  to  current  systems.
Additionally,  improved  fuel efficiency,  as  will be discussed
shortly, would also  serve to reduce engine-out emissions.

     For  the  stoichiometric three-way  catalyst  it was assumed
that  a  five percentage point  improvement  could  be  made for all
pollutant   conversion  efficiencies   except    particulate  and
gasoline  test   cycle  NOx.   No  improvement in  particulate was
projected because future  improvements in  catalyst HC efficiency
will  likely   not  affect   the  heavy  hydrocarbons  which  are

-------
                              3-29

characteristic of the particulate.   Also, no  improvement  in NOx
efficiency was projected  for the gasoline test  cycle  catalyst.
This  is   because the  catalyst  already  showed  a  91  percent
efficiency  and  an   efficiency  much  higher  than this  is  not
likely,  at  least in-use.   The  projected catalyst  efficiencies
for  the   future  optimized stoichiometric  engine  are  shown  in
Table 3-7, along with the resultant tailpipe emissions.

     B.     Future Optimized Engine  Efficiency

     The  last topic  of  discussion for  future  optimized  CNG
engines is  that  of  fuel  efficiency improvements.  there  are a
number of reasons to believe that  fuel efficiency of  dedicated
CNG  heavy-duty  engines   can  improve  over  current  technology.
Natural  gas  as   a   fuel  has  a  relatively  slow  flame  speed
compared  to  gasoline and diesel fuel.  The  development  of fast
burn combustion  chambers  would  result in an increase  in peak
pressure  and  higher  engine thermal  efficiency.   This is an area
that can apply equally to lean-burn and stoichiometric engines.

     For  lean-burn engines, the  current technology has a rather
low  compression  ratio compared to  the  BUG engine  and  is a
rather conservative  design.   Thus,  an increase  in  compression
ratio and a  resultant increase  in thermal efficiency could be
expected.   Also, the  prechamber approach previously  discussed
has  potential for  higher  cylinder pressures  and  better fuel
economy.  For these  reasons  it seems  reasonable  to  assume that
a  significant improvement  in fuel  efficiency  can  be made  for
lean-burn  engines.   Thus,  it was  assumed  that  a  ten  percent
decrease  in  brake-specific   fuel  consumption  (from  10,000  to
9,000  BTU/BHP-hr)  could  be   achieved  in  future  optimized lean
burn   CNG  engines   over  current   technology.    This   is    a
significant  improvement,  but  would still  result in  an energy
consumption  for  the optimized  lean-burn CNG  engine  about  25
percent higher than the diesel.

     For  stoichiometric  engines, the  compression ratio  of  the
current  technology CNG  engine  is  already  higher than  that of
the  current  lean-burn  and   there   is not   as  much  room   for
compression  ratio increase.   Also, the prechamber  approach is
not  likely  to  be  useful  in  stoichiometric  engines,   and  no
efficiency  increase  is expected here  for it.  For these reasons
it  is  reasonable to  assume  that,  while  further  improvement in
stoichiometric  CNG  engine efficiency  can be expected,  it will
not  be as .large as  that for  lean burn.   Thus,  it  was assumed
that  only  a  five  percent  reduction  in  brake-specific  energy
consumption   would  be  achieved on future   stoichiometric   CNG
engines  over  current technology.   This results in  a fifteen
percent   lower   fuel   consumption  rate   for    the  optimized
stoichiometric   engine  compared to the  current  Chevrolet  454

-------
                                            3-30
                                        Table  3-7
                             Estimates of Optimized  Heavy-duty
                      CNG Stoichiometric Engine Emissions (q/BHP-hr)
Pollutant
THC
NMHC
NOx
Part.
CO
C02
Formaldehyde
THC
NMHC
NOx
Part.
CO
C02
Formaldehyde

w/o
Catalyst
3.60
0.82
6.62
0.01
31.9
529
0.03
Current
Catalyst
Efficiency
72%
32%
80%
0%
66%
	
97%
Proiected Ootimized
With
Catalyst
Diesel
1.00
0.15
1.32
0.01
10.8
557
0.0008
W/O
Catalyst
Test Cycle
2.88
0.66
5.30
0.01
25.5
499
0.02
Catalyst
Efficiency
77%
87%
35%
0%
71%
	
97%
With
Catalyst
0.66
0.09
0.79
0.01
7.40
534
0.0006
Gasoline Test Cycle
1.7
0.40
5.75
0.01
28.5
474
0.02
58%
78%
91%
0%
63%
	
99.5%
0.72
0.09
0.51
0.01
10.6
500
0.0001
1.36
0.32
4.60
0.01
22.8
453
0.02
63%
83%
91%
0%
68%
	
99.5%
0.50
0.05
0.41
0.01
7.29
480
0.0001

-------
                              3-31

gasoline  engine,    A  similar  comparison  to   other   gasoline
engines would be expected.

     For  an  optimized  stoichiometric  combustion  CNG  engine
operating over  the  diesel  cycle  test cycle  this five  percent
reduction in  energy consumption  over current technology  would
result  in  a brake-specific  energy consumption  of  about  9,450
BTU/Bhp-hr.   Comparing this  value to  the optimized  lean-burn
combustion engine value of  9,000 BTU/Bhp-hr it can'be  seen that
optimized  lean-burn CNG  engines  are expected  to  be  somewhat
more   fuel-efficient   than   their   stoichiometric   combustion
counterparts.

V.   Safety Issues for CNG Use in HP Applications

     A.    Introduction

     For  certain  heavy-duty  applications,   CNG  is  likely  to
become a viable alternative in the near  future.   However,  there
are  several  issues  which should be analyzed  to  insure  that the
relative safety risks of using compressed natural  gas  as a fuel
for  heavy-duty  applications   are  known  and  addressed.   These
issues  are  discussed  in  these  sections.   The  specific issues
looked  at  include  flammability,   explosivity,   toxicity,  and
special safety  issues  of  concern to the use  of  CNG as  well as
the  implications  of these issues  when using CNG as a  fuel for
heavy-duty  vehicles.   The other   fuels  looked  at  are  gasoline
and diesel fuel for comparison purposes!

     For conventional  fuels,  there is a broad base of tabulated
in-use  experience and,  as a result,  detailed   comparisons  of
many  aspects  of  the  safety  issues  can  be  made  empirically.
However,  in the  case  of  CNG  used  as  a  motor  vehicle  fuel,
experience  is  only just  now being  built up  and,  therefore,
there  remains  an  insufficient supply of in-use  data to produce
meaningful   statistical   results.*    Therefore,   the   safety
analysis contained here will  focus mainly on  expert projections
of  the safety-related  issues involved in  the use of  CNG as  a
heavy-duty  motor   vehicle  fuel.     This   analysis   is  mainly
qualitative  in  nature,  although   a  limited  amount of actual
in-use  data is  included.
      CNG  fueled vehicles have been in use  for  as much as thirty
      years  in other countries with no reports of unusual safety
      problems, however.

-------
                              3-32

     B.     Fuel Properties and General Considerations

     1.     Toxicity

     The  issue  of   toxicity  will  receive   only  peripheral
attention  in  this report.   Natural gas,  being  mostly methane,
is not toxic.   However,  because of  its  gaseous nature,  it  can
act  as  an  asphyxiant.   Therefore,   from  a toxicity viewpoint,
the  only  concern  in handling  methane  is  to  ensure  proper
ventilation   so  that   concentrations   sufficient   to   cause
asphyxiation do not form.

     In  contrast  to   CNG,  diesel fuel  and gasoline  both  pose
significant toxicological  risks.   Gasoline and  diesel fuel  are
toxic if  ingested, inhaled,  or  even absorbed through  the skin.
The  potential  effects  of  acute   or  prolonged  exposures  are
nausea,   vomiting,   cramping,  liver  and  kidney  damage,  lung
irritation, and  central  nervous system depression  ranging from
mild  headaches  to  coma  or  even  death.[12,13,14,15,16]   In
addition,  both  gasoline  and  diesel  fuels  have  carcinogenic
risks associated with them.   While diesel  fuel vapors  are  not
known to be  carcinogenic,  skin tests  show diesel fuel  to  be
weakly to moderately  carcinogenic.   On the  other hand, gasoline
is believed to pose  carcinogenic  risks  from vapors as well as
from direct contact with the fuel.[17]

     2.     Flammability

     With   an   estimated  15,000  to  20,000   fires  resulting
annually  from  motor  vehicle  accidents,  the  risk  of  fire  is
probably  the  greatest safety risk  associated with  fuels.[18]
Associated  with these fires  are approximately 1700 deaths, 3700
serious  injuries,  3600  moderate injuries  and property damage
costs of  well over three  billion dollars.[18,19]   There are an
additional   2553   reported   service  station   fires   annually
resulting in approximately four deaths and 115 injuries.[18]

     While  all fuels  are  flammable, the conditions needed  for
ignition  and  the  severity of the results of an  ignition depend
on  the  properties   of  the  fuel.    Table  3-8  summarizes  the
properties  thought   to  be  most   important  in  assessing  the
hazards  of  different  fuels as well  as other key physicochemical
properties.

     Because  of the  varying  properties of  the  different fuels,
the  probability of an actual  ignition  is strongly dependent on
the conditions under  which a  fuel spill or  leakage occurs.  The
analysis  begins with a  look  at the hazards  associated with  a
fuel  escape under  conditions  with good  ventilation  (such  as
outdoors)  and then reviews conditions with poor ventilation and
the potential  results should  a fire  occur.

-------
                                            Table 3-8
                     Selected Physicochemical Properties of Automotive Fuels
               Property

Flanunability limits, vol % in air
Detonability limits, vol % in air
Minimum ignition energy in air, raJ
Autoignition temperature, K(°F)

Flash point, K(°F)
Flame temperature K(°F)
Energy content, lower heating value
     1.  Btu/gal.

     2.  Btu/lb

Diffusion coefficient in NTP air,*** cm/s
Buoyant velocity in NTP air,*** m/s
Density of liquid, g/cm-*

Density of gas relative to air = 1.00
Liquid/gas expansion ratio
Vapor pressure or equivalent,**** atm

Viscosity of liquid § NBP***, poise
Normal boiling point***, K(°F)

Threshold limiting value (TLV), ppm
Storage conditions
   CNG*

5.3 - 15.0
6.3 - 13.5
  0.29
813 (1004)

85 (-306)
2148 (3898)

19,760 @ 2400
psi, 294k (70°F)
21,300

0.16
0.8 - 6
                                     0.555

                                         1
                                     asphyxiant
                                     Compressed Gas,
                                     2400 - 3000 psig
 Gasoline

 1.0 - 7.6
 1.1 - 3.3
  0.24
 501 - 744
 (422-880)
 230 (-45)
 2470 (4478)

 114,132 (AVG)

 18,900 (AVG)
 (60° api)
 0.05
 nonbuoyant
 0.70 - 0.78 §
  1 atm
  3.4
  156
 0.54 - 1.0
e 311K (100°F)
   0.002
 310 - 478
 (100 - 400)
   500
Liquid @
ambient T&P
                                                                         Diesel Fuel

                                                                         0.5 - 4.1

                                                                         0.3 (est)
                                                                         533 (500)

                                                                         325 (125) min.
                                                                         129,400 (AVG)

                                                                         18,310 (AVg)
nonbuoyant
0.82 - 0.86

  >4.0 (est)

0.0005 § 311K
(100°F)(calc.)
    0.02
480 - 600
(405 - 620)
    500
Liquid @
ambient T&P
**
***
****
Source:
Properties are primarily  those of methane.   It is  recognized,  however, that  natural
gas sources  vary in composition.   Property values  will  therefore deviate  to  a small
extent from pure methane.
Properties refer to Grade  No.  2 diesel fuel.
NTP equals 293.15 K (68°F) and one atmosphere;  NBP  equals  normal  boiling point.
For  gaseous   fuels,  refers to "equivalent  vapor  pressure"  when  released  from  high
pressure storage  container (see  Sec.  VII),  or  maximum possible  pressure   in  ambient
environment.

   "Gaseous Fuel  Safety  Assessment for  Light-Duty Automotive Vehicles," M.C.  Krupka,
   A.T.  Peaslee,  and H.L.  Laquer,  Las  Alamos  National Laboratory,  November,  1983.
                                                                                                     I
                                                                                                     CO

-------
                              3-34

     a.     Ease of Ignition Under Conditions of Good Ventilation

     Under conditions of good ventilation,  accumulation  of fuel
vapor  in  flammable  concentrations  is   likely  only  in  close
proximity to the vapor source.  Any fuel  release  outdoors would
likely  be well-ventilated.   Available studies  also show that
even fuel releases  due  to collisions  in  areas such  as  tunnels
should  be sufficiently ventilated  to  prevent the  formation of
hazardous   concentrations   of    fuel   vapor.[20]"   In   these
situations,   the  volatility  of  the liquid  fuels combines with
the  lower flammability limits,  vapor densities  and  diffusion
coefficients of the  vapors  to form the most critical factors in
assessing   ignition   probabilities.   EPA's   methanol   safety
analysis  rated  diesel  fuel  as  the safest material  under these
conditions  and  gasoline  as  the  most dangerous  with  methanol
fuels being an  intermediate risk.[13]   These rankings  follow
the relative volatilities of the fuels.   A study of  the  safety
issues  of CNG use  done by  the Los Alamos National  Laboratory
for  DOE  in  1983  concluded  that  CNG fuels  were  safer  than
gasoline  due  to  the high rate  of  dispersion  and  the relatively
higher  lower  flammability  concentration.[21]   Diesel  fuel  was
once  again  considered  to  be the safest fuel due to  the fact
that  it  would take  a  relatively high temperature  to  allow an
ignitable mixture of vapor to form (flash point 125°F).

     b.     Ease of Ignition Under Conditions of Poor Ventilation

     Under  conditions  of poor  ventilation, the  probability of
an ignition  is greatly  different than under conditions  of good
ventilation.  Since  the  fuel  vapors  are confined to  a closed
space, the  likelihood  of  ignition  is  greater for  all  types of
fuels than  under  conditions where the  vapors  are  allowed to
disperse  readily.    These  are  the types  of  conditions  which
might be found  in  a  storage or  repair  garage or  in  a poorly
ventilated covered parking area.

     Again,  under  these conditions, diesel  fuel  is  the  safest
of   the   three  alternatives   due  its   very  low  volatility.
Temperatures  high   enough   to  generate   the  needed   vapor
concentrations  are  rarely  found  unless  an  ignition   source
strong  enough  to heat  the   fuel  is nearby.   Therefore,   diesel
fuel  under  most   conditions    is   relatively  safe   from  a
flammability standpoint.

     Gasoline  can  easily   form   flammable  or  even  explosive
mixtures  under poorly  ventilated  conditions  since  the  vapors
cannot  disperse  as readily.  As  with diesel  fuels,  vapors are
heavier  than  air  so that  the greatest risk of fire or explosion
would be  near the ground.   For  this reason,  existing repair and
indoor  refueling stations   often  put  electrical  equipment and
other possible sources of ignition up  in  the ceiling.

-------
                              3-35

     CNG,   as  a  lighter  than  air  gaseous  fuel  also  poses  a
serious flammability and  explosion risk under  conditions  which
do not allow the  vapors  to dissipate.   These  hazards would  be
similar to the  types  of  hazards often faced by  residential  and
commercial users  of  natural  gas  for  heating  or  other  uses.
Furthermore,  the  most likely place  for formation of  flammable
or  explosive  mixtures  of  natural  gas  would  be   near  the
ceiling.    Therefore,  facilities which  would  handle  repairs  or
refueling  of  CNG  vehicles  indoors  may   not  want   to  place
ignition sources  close to the  ceiling.  Proper  ventilation  and
gas sensors would likely be  utilized  in any  such  facility  to
eliminate  the possibility  of  a combustible mixture  collecting.
In  some   cases   seperate   facilities   for   CNG  vehicles  and
liquid-fueled vehicles may be  required.

     3.    Hazards Associated  With a Fire

     The  most  telling  distinctions between  liquid  fuel  fires
and natural gas fires  is  the  rate of combustion and  the amount
of smoke produced by  combustion.   Both conventional liquid fuel
vapors and natural gas  will  rapidly  combust   any  fuel  vapors
present  in flammable  concentrations  once  a source of ignition .
is   introduced.     The  major   differences   in   flammability
characteristics  are   in   sustained,   severe   fire   properties.
Pools  of  gasoline or  diesel  fuel  burn with a defined  rate  of
heat  release  based on the heat of  combustion and the  rate of
combustion of  the  fuel  and  produce a  large  amount  of smoke.
Rather than  burning from  a pool  at a  fairly  well  defined rate
natural gas  will  burn with a 'torch type flame at the  site of
the leak.  The  rate of  heat  release will be  controlled by the
rate of fuel  release  and various  safeguards can be implemented
to control the  fuel release  rate.   These safety strategies are
discussed  in  a   later   section.    Furthermore,   natural  gas
produces very little smoke while burning.

     Diesel  fuel  fires  tend  to  start  slowly  but  progress
violently.    High  heat   release   rates   result   in   a   high
probability  of  spreading  the  fire to  other  nearby flammable
materials  or causing serious  burns to exposed individuals.[13]
Gasoline,  due to  its  high volatility, immediately erupts  into  a
fully  developed fire.  Because  of  the  rapid burn rate, gasoline
has  an  even higher  heat  release  rate  than  diesel  fuel  and
therefore  poses an  even  greater risk for spreading the  fire or
for serious  burns to  individuals. [13]   Also,  both gasoline and
diesel will produce  a  large amount of  smoke  during combustion.
This  smoke itself can pose serious  risks of  injuries to nearby
individuals.

     Natural  gas,  on  the other  hand,  burns  at a  lower  flame
temperature  than  gasoline  or  diesel  fuel.   Since the  rate of
combustion would be defined by  the  rate  of  release  of the.fuel,
it  is more  difficult  to quantify  a rate  of  heat  release.   It

-------
                              3-36

is, however, estimated  by  the Los Alamos  report that  the rate
of heat  release  for CNG is  less than that of  burning  pools of
liquid fuel.[21]   A more recent  study by EBASCO  indicates that
the heat release rate would  be less  than 40 percent of  the heat
release rate of gasoline fires.[20]   Furthermore  the  burning is
more  likely  to  be  confined  to  a  small  area  immediately
surrounding  the  release  point  than  for  liquid  fuel  fires.
Consequently the likelihood  of spreading the fire is  less for a
CNG torch fire than  for a  liquid fuel pool fire.   In turn,  the
likelihood  of  an   individual  receiving  a  serious burn  would
probably be  less  than  for  liquid fuel  fires.    Finally,  since
natural  gas burns   without  producing   sizeable  quantities  of
smoke  and  toxic  materials,  toxic risks  from  exposure to  the
products of  combustion  would be  less compared to  gasoline  and
diesel fuel fires.

     4.    Issues of Special  Concern  in the Use of CNG

     Several  researchers  have  identified  issues  of   special
concern   to    CNG   and    pressurized   gaseous    fuels   in
general.[21,22]  First  of all, there is the inherent danger of
storing  and  handling a compressed gas.   If a  fuel line should
rupture,   particularly  at  a  refueling  station,  injury  could
result from the flailing hose.   Another concern with compressed
gases  is the  possibility  of  frostbite  resulting  from someone
being  exposed  to gases cooled  by rapid expansion or  fixtures
cooled  by  these  gases.   These  issues  are  dealt  with  more
thoroughly in section C below.

     Second, there  is  the  concern  of  the fuel  cylinder being
improperly   restrained.    Cylinders    generally  have   greater
structural integrity than the  surrounding vehicle and therefore
have  the potential  to  penetrate into  parts  of  the  vehicles
where they  were  not intended  to be  or  to  break  loose  from the
vehicle   and  become   a  potentially  dangerous   projectile,
especially  in  collisions.    However,  with proper placement  and
restraint  of  CNG   cylinders,  risks  posed  by   fuel  cylinders
breaking  loose should  be  no  greater compared to  conventional
fuel tanks.

     The  Department  of Transportation   (DOT)  has  established
standards  for  the   safe  transport  of  hazardous  materials.
Natural  gas  falls  into  this  designation.   These  standards
establish  both  maximum  operating  and   burst   pressure  for
cylinders,  the  type   of  testing  required   before  use  and
periodically   during   the   cylinder   lifetime,   and   allowable
contaminants   in   the   compresed   natural  gas  to   prevent
corrosion. [23]   While the cylinders  used on CNG  vehicles  in the
U.S.  meet  these specifications,  it   should be  noted  that  the
specifications were  designed for  the transport  of CNG and not

-------
                              3-37

specifically for  the use  of these  cylinders  as  vehicle  fuel
tanks.   DOT has not  yet  set  standards for cylinders used in CNG
vehicles.

     There is also concern about  these cylinders being  able  to
withstand corrosion  from external  sources  and from contaminants
in the gas itself.   Indeed, the National  Fire Protection Agency
has  established  purity  standards  for compressed  natural  gas
which  limit  the  water  content  to  less   than  0.5  pounds  per
million  cubic  feet,  0.1  grains  of  hydrogen  sulfide  per  100
cubic feet, and no  more than three  percent  carbon  dioxide for
gas to be  transported  in steel  cylinders.  Pipeline quality gas
specifications   are   generally   less   stringent.    However,
cylinders thus far have  shown excellent resistance to corrosion
from current gas supplies.

     Furthermore,   the  cylinders  must  be • able  to  withstand
physical strain.  Cylinders  will be  subjected to  the strain of
repeated  cycles  of pressurization   and  temperature  swings.
Rigorous  standards  have been  set for  structural  resistance to
such  conditions  and currently   available  CNG  cylinders  have
routinely been certified to meet these standards.

     Finally,  there  is  the  question of  the integrity  of the
fuel cylinders  and   their  ability  to  withstand  physical  abuse.
Severe  abuse tests  have  been performed  on  CNG  cylinders  to
demonstrate their fundamental durability.   These cylinders have
survived being  dropped in a  car from  a height  of  more than 60
feet,  having sticks of  dynamite  strapped to  the  side  of the
cylinder  and detonated,  and being  shot  at  with  small caliber
bullets.[24]

     C.    Implications for Vehicle Safety

     1.    Refueling

     The hazards posed in  refueling with  compressed natural gas
are  different  from  those  posed in refueling with conventional
liquid   fuels.    As   long   as   normal,   properly  functioning
equipment  is being used,  CNG refueling should be generally less
hazardous  than refueling  with  gasoline  or  diesel  fuel   since
there  will be  no toxic  or  flammable vapors  escaping  from CNG
refueling  equipment,  as  there  often  is  with  conventional
refueling  equipment.  In  the event  of equipment  failure, CNG
systems  would  offer   a  significant  advantage   compared  to
gasoline   or   diesel  systems  in  the  area  of   environmental
exposure.   While  gasoline  and  diesel  fuel storage  tank  or
dispensing  equipment leaks  could lead to  contamination of the
surrounding  environment  with toxics,  CNG  leaks  would  introduce
no  such  toxic  materials  into  the  environment.   On  the  other

-------
                              3-38

hand,  CNG  equipment  failure  could  pose  a  greater  risk  of
physical injury to the  operator  compared to gasoline and diesel
fuel systems.  These  injuries  could take the form  of cryogenic
burns from  gas  cooled by rapid  expansion  or  injuries resulting
from being  struck  by  a flailing  hose.[4]   Both of  these risks
can  be  minimized  by designing  the  equipment  to  both  resist
catastrophic  failures  and so  that  if  such  a  failure were  to
occur,  it  would occur  at a  point which  is already anchored,
contains a valve to shut  off  the flow of  fuel,  and/or  is in an
area where people are not likely to be exposed to the leak.

     In  cases  where  the  refueling  is  being  done  indoors,
flammable  concentrations  of  fuel  vapors   of   any  type   can
build-up.    For  CNG  systems,   very  little  vapor  should  be
released during  normal  operation.   However,  in the event  of
fuel leakage  from  malfunctioning equipment,  large quantities of
vapor  could  rapidly  escape.   The  risks  of  vapor  build-up,
however,  can be  minimized  by  enclosing  the  fuel  line in  a
ventilation  line.   Since fuel   leakage  cannot  be  completely
prevented,  however,  the best  strategy for minimizing  risks  is
building design.   Proper ventilation  and placement  of equipment
which could  serve  as  ignition sources should greatly reduce the
risks of fire or explosion posed by any fuel source.

     2.     Vehicle Operation and Crashes

     During   normal   operation  of   existing  fleets  of   CNG
vehicles,  it  appears  that small  leaks have  been observed 'with
greater frequency  than  in conventionally  fueled vehicles.[21]
It  would  be reasonable to conclude that the highly-pressurized
nature of the CNG fuel system could make it more prone to small
leaks and to  more  fuel  being released from a given leak.   Small
leaks  pose  concerns  about  vapors  accumulating  to  flammable
concentrations  either   in   vehicle   compartments   or  vehicle
storage enclosures.   Incorporation of  design features such  as
vents  in  the  vehicle  body  and  ventilation   of  garages  can
mitigate  the  risks  of  fuel  leakage.   Currently,  there  is
insufficient  data  to  determine to  what  extent  small fuel leaks
actually pose any hazards.

     In vehicle collision scenarios, CNG would  appear to  pose a
level of  risk somewhere between  diesel fuel and gasoline.   To
analyze collision hazards it is  necessary  to  evaluate the risks
of,  and the likely extent of, fuel leaks.   It is also necessary
to  examine  the ease with which such problems can be dealt  with
in  the event of combustion of leaked or leaking fuel.

     In  the  absence  of  extensive  data  on  the   use   of  CNG
vehicles,  it  is difficult to make  an  accurate assessment  of the
relative risks  of  fuel  release.   However,  some assessments can

-------
                              3-39

be made  as to  estimates of  relative  risks.   Because  of  the
structural integrity  needed by  the fuel  storage cylinders  to
hold  compressed  natural  gas,  these  cylinders  are  much  more
likely than gasoline or  diesel fuel tanks to survive collisions
without release  of  fuel from the storage tank.   On the  other
hand,  fuel  lines,  valves  and fittings would  be  more prone  to
severe leaks  than gasoline  or  diesel  systems because  of  the
pressurized nature  of the  fuel.   However, safety devices  such
as fuel release regulators and solenoid valves to-shut  off  fuel
flow  when  the  engine  stops  can  be built  onto  the  fuel  cylinder
to lessen the severity of any release of natural gas  from  a CNG
vehicle,  and minimize the significance of this risk.

      In the  event of  a fuel release  resulting  in a fire,  the
resulting problems from  a CNG  fire are  likely to be easier  to
deal  with.  First of  all,  fires  from liquid fuels are difficult
to  extinguish  and  difficult  to  control   if   they  are   not
extinguished.    On  the  other  hand,  natural  gas  torch  fires
(which  are the  only  kind  likely  to  be  sustained)   can  be
extinguished  by  shutting  off the fuel  source,   which  is  not
generally possible in liquid fuel  fires.   Some currently in-use
CNG  vehicles  have a  readily accessible  quarter   turn  shut-off
valve for this purpose.  Should it be difficult to shut  off the
flow  of CNG,  it  should still be  possible  to  control  the damage
from  the  fire  because  of  its  localized   nature.    Severe
explosions  are  unlikely in any  case  since  CNG  cylinders are
designed to handle conditions  likely to lead  to  explosions and
will  eventually  vent  off  gas which  will burn in a relatively
controlled manner rather than rupturing  to produce an explosive
release.

      Despite  the lack  of  a database  of CNG  use  sufficient  to
draw  definitive conclusions, there  are  some  surveys  and studies
which do  present   some  in-use  information.    In  1987,   the
American  Gas  Association  completed a  survey of  fleet  use  of
dual-fueled  vehicles.    Their  survey,   covering   434.1  million
miles  of  accumulated  use  with  dual-fueled   CNG/gasoline
vehicles,  claimed that  the  injury rate  for  CNG vehicles was
significantly  less  than for all  U.S.  vehicles and  fewer   fires
were  attributed  to  the CNG  fuel  system  than to  the  gasoline
fuel  system.[24]  Furthermore, data from  New  Zealand indicated
that  CNG  vehicles  caught  fire  much  less frequently  than did
conventionally  fueled vehicles.[20]   However, the   information
about vehicle  operation characteristics in these  reports lacked
information on number  of miles travelled or. CNG  and on injury
causes   was   therefore  insufficient   to  snable   any   solid
conclusions to be drawn  from the  data reported.

      In cases  where  the fuel release occurs before  a source of
ignition  is  available,  both gasoline and CNG have the potential

-------
                              3-40

to form a flammable cloud of vapor which  could burn explosively
when   an  ignition   source   is   introduced.   However,   such
conditions   are   only   likely   under   conditions   of   poor
ventilation.   Tunnels  and  lower  decks of  multi-decked bridges
have attracted attention as a  possible  location where explosive
conditions  might   occur  following  an   accident.    Both  the
Triborough Bridge  and  Tunnel  Authority  and the Port  Authority
of  New   York  and  New  Jersey,  as  part  of  their  general
restrictions on compressed  gas transportation  (stemming  from a
severe LP-gas accident  in the  first half of this century in the
Holland  Tunnel),  limit  the amount  of  compressed natural  gas
which can be  transported in these areas to less than 100 pounds
of gross weight with each  cylinder  being less  than  ten pounds
of  gross  weight.    This  same  restriction  applies  to  all
flammable compressed gases.[25,26]

     In spite of these  concerns,  available  studies suggest that
tunnels  generally have sufficient  ventilation to prevent  gas
vapors   from   building   up   large   volumes   of   flammable
concentrations,  so  that the relative  flammability risks of  the
fuels  should  be  similar  to  those  for  conditions  of  good
ventilation.[21]   Both the Los  Alamos report  and the recently
completed EBASCO  report for the  New York State Energy Research
and  Development  Authority,  Brooklyn  Union  Gas  Company  and
Consolidated  Edison Company on  the  safety of  CNG  in tunnels
concluded  that  CNG  is  safer  than  gasoline  for  use  as  a
transportation fuel  in  tunnels,  although neither  study claimed
CNG   would   be   safer   than   diesel    for   use    in   these
environments.[20,21]   The  EBASCO  report  in  particular  did  a
detailed  modeling study  of likely  fuel  concentrations  in  the
Holland  tunnel which would result from a rapid release of fuel
from  a  bus   and   concluded that  there  would be  only  a  very
limited  time and location  where there  would  be a  flammable
concentration of  natural gas,  and that  the only scenario under
which  natural gas  vehicles  might  pose  a  greater   risk  than
gasoline vehicles would be if the  fuel  line ruptured under  the
vehicle  and  the natural  gas  became  trapped under the vehicle
with no  safety  devices to stop the  flow  of gas or to ventilate
the undercarriage of the bus.

     3.    Maintenance

     The  issue  of  maintenance might  also  pose  some concerns.
On  CNG  equipped  vehicles,  it is  possible that  a maintenance
worker could release a  large  amount  of  fuel  by inadvertently
creating a vent  in the fuel system.  In  such  a  case,   the worker
could  also  face  the  risk of a  cryogenic  burn  as  described
earlier.   If  the  fuel  release  occurred  outdoors,   no  other
hazards  should  be  posed unless  there was  an  ignition source
immediately  present, due  to   the  rapid  dispersion of the gas.
On  the  other hand,  if  the  maintenance  were  being performed

-------
                              3-41

indoors,   there  is  a  potential   for  the rapid  formation of  a
flammable or explosive cloud.

     The  same suggestions  for garage  design described  earlier
would  reduce  the  hazards associated  with  a  fuel  leak  during
maintenance.   In   particular,  well  ventilated  buildings  will
greatly  reduce  the  risks of  a  flammable  or  explosive  cloud
forming.   Furthermore, it  should  be  pointed out that  a  vehicle
with a properly functioning  solenoid valve to shut off the fuel
flow  when  the  engine  is  not  running   could   minimize  the
possibility of  the  maintenance  crew  accidentally  releasing  a
large  volume  of   fuel.    With   the  establishment   of   proper
maintenance procedures (such  as  shutting  the  fuel  flow  valve
during  maintenance)  and  the  accumulation  of  experience  in
maintaining CNG vehicles,  the  risks  associated with maintaining
CNG  vehicles   should  not  necessarily  be  different  than  for
maintaining conventionally fueled vehicles.

     D.    Summary

     Evaluating   CNG  as  an   alternative  fuel  for   use   in
heavy-duty  vehicles  from a  safety  standpoint,  it can  be seen
that CNG poses  some  unique  safety  concerns due  to  its  gaseous
and  pressurized  properties.    On  the  other hand,  it has some
properties  which   are superior   (from  a  safety  viewpoint)  to
those of conventional  liquid fuels,  including  the fact that the
material  itself is  not toxic (a  significant safety benefit over
conventional  liquid  fuels).    Owing  to  the  lack of  extensive
experience  with this fuel  in the United States, it is difficult
to  draw  definitive  conclusions  regarding  the  safety  of this
fuel relative to  fuels with  a broader base of use.  However, it
would  appear  that   the  area  of  most  concern and  uncertainty
would  be flammability.   Studies  have suggested  that  CNG would
be  safer than  gasoline in well ventilated  areas  but  could pose
a greater risk of explosion  in areas with poor ventilation.  In
the  event of  a  fuel  leak or spill,  therefore,  the risk  of fire
from CNG  could  be  comparable to   the  risk from  gasoline  and
somewhat greater  than the risk from diesel  fuel.   However,  the
consequences  of  a  CNG  fire  should  be  less  severe than  for
gasoline  or   diesel  fuel   fires.     Although   the   relative
likelihood  of fuel  spills  or  leaks  from  CNG as  compared to
conventional  fuels  is yet to  be determined,  it would  appear
that CNG  would  not pose  a  greater  f lammability  risk than
gasoline.   Despite the need  for  further work  in ensuring that
fuel leaks are  minimized  and  controlled,  in  developing safe
maintenance practices and  in  assuring  that  refueling   can be
done safely,  it  would appear  that  there  are  no safety  issues
which cannot be dealt with which  would preclude the development
of  CNG  as  an alternative  fuel for  heavy-duty vehicles.    Taking
all  factors into  consideration,  it  would  appear  that   CNG is
certainly no more dangerous than  gasoline as a vehicle fuel.

-------
                              3-42

                      References Chapter 3
     1.     L.  E. Gettel, G. C.  Perry,  "Performance and Emission
Characteristics  of  a   Dual   Fuel  Caterpillar  3406  Engine,"
presented at  the Fifth  Windsor  Workshop on  Alternative  Fuels,
Ontario, Canada, June 13, 1989.

     2     G.   O'Neal,   Southwest  Research  Institute,  telephone
communication, November  1988.

     3.     "Natural Gas Vehicles:  A Review of  the State  of the
Art," prepared  by  Sierra  Research Inc.  for  the Natural  Gas
vehicle Coalition, Report No.  SR 89-04-01, April 1989.

     4.     M.  A. Deluchi, R.A.  Johnston,  D.  Sperling, "Methanol
vs.  Natural   Gas Vehicles:   A  Comparison  of  Resource  Supply,
Performance,    Emissions,   Fuel  Storage,  Safety,   Costs,   and
Transitions,"   SAE  Paper  No.  881656,  Society  of  Automotive
Engineers, Warrendale,  PA 1988.

     5.     R.    I.   Bruetsch,   "Emissions,  Fuel   Economy,   and
Performance   of  Light-Duty   CNG   and  Dual-Fuel   Vehicles,"
EPA/AA/CTAB-88-05, U.S. EPA MVEL  Ann Arbor, MI June 1988.

     6.     J.   Alson,  J.  Adler, T.  Baines,  U.S.  EPA.,  "Motor
Vehicle  Emission Characteristics  and  Air  Quality  Benefits  of
Methanol  and  Compressed  Natural  Gas,"   presented  at   the
Symposium  on   Transportation  Fuels  in  the  1990's  and  Beyond,
Monterey, CA.  July 1988.

     7.     Telefax from V. K.  Duggal,  Cummins  Engine Co.,  to J.
W. Mueller, U.S. EPA, October 12,  1989.


     8.     J.R.  Gladden,  J.C.  Kline,  "Development of Fast  Burn
Combustion With Elevated  Coolant  Temperatures  for  Natural Gas
Engines," report to the   Gas Research  Institute, Caterpillar,
Inc., Peoria,   IL, July  1988.

     9.     D.   J.  Waldman,  J.  R. Gladden, D. L. Endicott,  and B.
A.  Cull,  "Caterpillar  3406  Spark  Ignited  Natural   Gas  Engine
Emissions  on  EPA  Heavy-Duty  Transient  Test  Cycle,"  report  to
the  Gas  Research  Institute,  Caterpillar   Inc.,  Peoria,   IL,
January 1989.

     10.   R.    D.   Fleming,   G.   B.   O'Neal,   "Potential   for
Improving the Efficiency of a  Spark  Ignition  Engine for Natural
Gas  Fuel,"   SAE  Paper   No.   852073,  Society  of  Automotive
Engineers, Warrendale,  PA 1985.

     11.   C.    Spielberg,   "Compressed   Natural  Gas   Program
Monthly  Report 82,  January-May  1989," New York City Department
of Transportation, September  12, 1989.

-------
                              3-43

                 References Chapter 3 (cont'd)

     12.    "Material  Safety  Data  Sheet:  Diesel  Fuel  Oil  No.
2-D," Occupational  Health Services, Inc.,  September,  1985.

     13.    "A  Perspective  on  the Flammability,  Toxicity,  and
Environmental   Safety   Distinctions   Between   Methanol    and
Conventional  Fuels,"  Paul  A.   Machiele,  Standards  Development
and Support Branch, U.S. EPA,  August,  1989.

     14.    "Material  Safety  Data  Sheet:   Gasoline/Automotive,"
Occupational Health Services,  Inc., September, 1985.

     15.    "Hydrocarbon  Contact Injuries,"  J.F.  Hansbrough et.
al. ,  The  Journal  of  Trauma, Vol.  25, No.  3,  March,  1985.   As
cited in reference 13.

     16.    "Gasoline  Intoxication,"  W.  Machle,  J.  Amer.  Med.
Assoc., [1]:  1967- 1971,  1941.   Cited in "A Perspective on the
Flammability,  Toxicity, and  Environmental Safety  Distinctions
Between Methanol   and Conventional  Fuels," Paul  A.  Machiele,
Standards  Development  and Support  Branch,  U.S.  EPA,  August,
1989.

     17.    "Summary  and  Analysis  of  Comments  Regarding  the
Potential   Safety   Implications   of   Onboard  Vapor   Recovery
Systems," Office of Mobile Sources, U.S. EPA,   August 1988.

     18.    "Methanol Fuel Safety:  A  Comparative  study fo M100,
M85,   Gasoline, and Diesel  as  Motor  Vehicle  Fuels," Paul  A.
Machiele,  SDSB, U.S. EPA, August 1989.

     19.    Memorandum:   "Analysis  of Fuel  Tank-Related Fires,"
from  Kathleen  A.   Steilen,  Standards  Development   and Support
Branch, to Charles  L.   Gray,   Jr.,  Director,  Emission Control
Technology Division, U.S. EPA,  April, 1987.

-------
                              3-44

                  References  Chapter_3_icont'd)

     20.    "Safety Analysis  of  Natural Gas Vehicles  Transiting
Highway Tunnels,"  EBASCO  Services  Incorporated,  August,  1989.
Prepared  for  New  York  State Energy  Research and  Development
Authority,  The  Brooklyn  Union  Gas  Company,  and  Consolidated
Edison Company of New York, Inc.

     21.    "Gaseous   Fuel   Safety  Assessment  for   Light-Duty
Automotive  Vehicles,"  M.C.   Krupka,   A.T.   Peaslee,   and  H.L.
Laquer,   Los   Alamos   National   Laboratory,   November   1983.
Prepared for the Department of Energy.

     22.    "Identification   of   Safety  Related  Research   and
Development Needs  for CNG Vehicle Fuel•Systems,"  Gas  Research
Institute, March, 1983. (GRI-82/0061)

     23.    Code  of   Federal  Regulations,   Title  49,   Section
173.34.

     24.    "Severe Abuse Testing"  video tape distributed by the
CNG Cylinder Corporation

     25.    Rules  and  Regulations  Governing  the  Use  of  the
Triborough  Bridge  and  Tunnel  Authority  Facilities  and  the
Transportation of Hazardous Material as in  effect September 30,
1984.

     26.    Hazardous  Material:   Transportation  Regulations  at
Tunnel and  Bridge Facilities, The  Port Authority  of NY  & NJ,
revised June 1,  1976.

-------
                              3-45

      References For Further Information on CNG Technology

     "Assessment of  Methane-Related  Fuels for  Automotive Fleet
Vehicles," U.S.  Department  of Energy,  DOE/CE/50179-1,  February
1982.

     W. A.  Goetz,  D.  Petherick  and T.  Topaloglu,  "Performance
and Emissions of Propane, Natural Gas,  and Methanol  Fuelled Bus
Engines,"   SAE  paper   no.   880494,   Society   of   Automotive
Engineers, Warrandale, PA,  1988.

     R.   R.   Raine,   J.   Stephenson   and   S.   T.   Elder,
"Characteristics of  Diesel  Engines Converted  to Spark Ignition
Operation  Fuelled  with  Natural  Gas,"   SAE  paper  No.  880149,
Society of Automotive Engineers,  Warrendale,  PA, 1988.

     "Assessment   of  Costs   and  Benefits   of   Flexible  and
Alternative  Fuel  Use   in   the   U.S.   Transportation  Sector,"
Progress  Report  Number One:   Context  and Analytical Framework,
DOE/PE-0080, January  1988.

     "Dual-Fuel  School   Bus   Demonstration,"   New  York  State
Energy Research  and  Development  Authority,  Albany,  NY, October
1986.

     Vice President's Task  Force on Alternaitve Fuels," Report
of  the Alternative  Fuels Working  Group."   Office  of  the Vice
President, Washington, DC, July  1987.

     L.  E.  Gettel,  G.  L.   Perry,  J.   Boisvert  and  P.  J.
0'Sullivan,  "Microprocessor   Dual-Fuel  Diesel  Engine  Control
System," SAE Paper No.  861577, Society of Automotive Engineers,
Warrendale, PA,  1986.

     N.  J.  Beck,  et. al,  "Electronic  Fuel  Injection for Dual
Fuel   Diesel   Methane,"   SAE  Paper  No.  891652,   Society  of
Automotive Engineers, Warrendale, PA, 1989.

     J.   Heenan   and   L.   Gettel,   "Dual-Fueling   Diesel/NGV
Technology,"   SAE  paper No.   881655,   Society  of  Automotive
Engineers, Warrendale, PA,  1988.

     P.  C.  Few  and  P.  Sardari, "Dual  Fuel  Control  of a High
Speed  Turbocharged   Diesel   Engine,"   SAR  Paper  No.   871670,
Society of Automotive Engineers,  Warrendale, PA,  1987.

     P.  C.  Few and H.  A.Newlyn,   "Dual  Fuel  Combustion  in  a
Turbocharged  Diesel   Engine,"  SAE paper No.  871671,  Society of
Automotive Engineers, Warrendale, PA, 1987.

     T.  Adams,  'The Development of Ford's  Natural  Gas  Powered
Ranger,"  SAE paper  No.  852277, Society of Automotive Engineers,
Warrendale,  PA,  1985.

-------
                              3-46

    References For Further Information on CNG Safety (con't)

     American Gas Association,  "A.G.A.  Requirements  for Natural
Gas Vehicle (CNG) Conversion Kits," No.  1-895, August 1985.

     G. A.  Karim and  I.  Wierzba,  "Experimental  and Analytical
Studies of  the Lean Burn Operational Limits  in Methane Fuelled
Spark Ignition and Compression  Ignition Engines,"  SAE  paper No.
891637, Society of Automotive Engineers, Warrendale,  PA, 1989.

     D.  Micle,  T.  Krekee  and  T.  Giannacopoulos,  "Electronic
Injection System for Natural Gas  in  a Diesel Engine-Development
and  Testing,"  SAE  paper  No.   890852,  Society  of  Automotive
Engineers, Warrendale, PA, 1989.

     J. van der  Weide, et.  al.,  "Experiences with  CNG and LPG
Operated  Heavy Duty  Vehicles  With Emphasis  on U.S.  HD Diesel
Emission   Standards,"   SAE  paper   No.   881657,    Society  of
Automotive Engineers, Warrendale, PA, 1988.

     "Assessment  of  Methane-Related Fuels  for  Automotive Fleet
Vehicles,"  Volumes   1-3,   The   Aerospace   Corporation,  Energy
Conservation Directorate, DOE/CE/50179-l,  Office  of  Vehicle and
Engine  R  &  D,  U.S.   Department  of  Energy,  Washington,  D.C.,
February, 1982.

     "The  Practical  and  Economic Considerations  of Converting
Highway  Vehicles to  Use Natural  Gas  as   a  Fuel,"  Richard  L.
Bechtold, et al., SAE Technical Paper 831071, 1983.

     "Some Considerations of the Safety of Methane  (CNG),  as an
Automotive  Fuel  -   Comparison  with  Gasoline,   Propane,  and
Hydrogen  Operation,"  G.A.  Karim,  SAE  Technical  Paper 830267,
1983.

     "High  Speed  Collision   and   Severe  Abuse   Testing  of
Composite  Reinforced  Aluminum  CNG  Vehicle  Fuel  Cylinders,"
Norman  C.  Fawley,  Symposium  Papers,  Nonpetroleum   Vehicular
Fuels  IV,  Institute  of  Gas  Technology,  Chicago,   Illinois,
October,  1984.

     "Environmental  and  Safety Aspects  of  Natural Gas-Fueled
Vehicles,"   J.W.   Porter,   Symposium   Papers,    Nonpetroleum
Vehicular  Fuels  II,  Institute  of  Gas   Technology,  Detroit,
Michigan, June,  1981.

     "The  San  Antonio  Story,"  J.W.  Brooks,  Symposium Papers,
Nonpetroleum  Vehicular Fuels II,  Institute of  Gas  Technology,
Detroit, Michigan, June,  1981.

-------
                              3-47

    References For Further Information on CNG Safety (con't)

     "Safety  Testing  of  LPG  and Gas-Fueled  Vehicles," J.  van
der  Weide,   Symposium  Papers,   Nonpetroleum  Vehicular   Fuels
Symposium,  Institute  of  Gas  Technology,  Arlington,  Virginia,
February, 1980.

     "A  Fleet of  Twenty  CNG  Cars  and  Trucks  in  Upstate  New
York: Ten  Months Experience," D.J. Whitlock,  Symposium Papers,
Nonpetroleum Vehicular Fuels  III,  Institute of  Gas Technology,
Arlington, Virginia, October, 1982.

     "Analysis of  the Pneumatic  Burst of a Large Seamless Steel
Vessel in Natural Gas Service,"  B.W.  Christ,  U.S.  Department of
Transportation report DOT/MTB/OHMR-78-4,  March 1979.

     Gas-Powered  Vehicle  Evaluation Program,  TES,  Ltd.,  report
TES C372, for the Road and Motor Vehicle Traffic Safety Branch,
Transport Canada, March, 1982.

     "Dual-Fuel  Motor  Vehicle  Safety  Impact  Testing,"  U.S.
Department  of  Transportation report  DOT/HS-800622,  November,
1971.

     "Assessment  of  Research and Development  Needs for Methane
Fueled Engine Systems," T. J.  Joyce,  Final  Report,  Gas Research
Institute report GRI 81/0046, March 1982.

     "The  Benefits  and  Risks  Associated  with Gaseous-Fueled
Vehicles,"   D.    Shooter   and   A.  Kalelkar,    Report  to  the
Massachusetts Turnpike  Authority,  Case No.  74400-2, prepared by
A.D. Little,  Inc., May 1972.

     "CNG   Reports,  No.   1,"   T.J.   Joyce   Associates,   Inc.,
October,  1982.

     "Report   of  Overseas  Visit  to  Investigate  Compressed
Natural  Gas in  Italy,"  R.N.  Abrams,  A.L.  Titchener,  and J.P.
West,  Liquid  Fuels   Trust  Board,  Wellington,   New  Zealand,
February,  1980.

     "Preliminary Analysis of  the  Safety History of Natural Gas
Fueled  Transportation  Vehicles,"   J.  Winston  Porter,  American
Gas  Association,  Arlington, Virginia,  December,  1979.

     Gas  Powered Vehicle Evaluation  Program,  Transport Canada,
Road Safety  Branch,  TES Report CE 372, Ottawa,  Canada, March
1982.

-------
                              3-48

    References For Further  Information  on CNG  Safety  (cont'd)

     "The  Benefits   and Risks  Associated  with  Gaseous-Fueled
Vehicles," D. Shooter and A. Kalelkar,  Arthur D. Little Inc.,
May, 1972.

     "Installation of  Compressed  Natural Gas Fuel  Systems  and
Containers on  Highway  Vehicles  and Requirements  for  Refueling
Stations," Canadian Gas  Association,  CAN 1-B149.1-M80,  October,
1982.

     Code  of   Practice   for   CNG  Compressor   and  Refueling
Stations, NES 5425 Part 1,  1980.

-------
                           CHAPTER 4

                     ECONOMICS OF USING CNG
                   IN HEAVY-DUTY APPLICATIONS
I•    Introduction

     In this chapter the economics of using CNG  as  a heavy-duty
vehicle  fuel  will  be  presented.   First,  the domestic  natural
gas  supply will  be discussed followed  by  a  presentation  of
natural  gas  prices.    Next,   CNG  refueling  station costs  and
hardware will be discussed.  Following this,  overall fuel costs
for  current  and future natural gas, gasoline and diesel fueled
heavy-duty  vehicles will  be  compared.   Finally,  vehicle  and
engine   costs   associated  with  dedicated   CNG   use  will  be
discussed.

II.  Domestic Natural Gas Supply and Price

     The total United States proved  reserves  of  dry natural gas
in  1987  were  187.2 trillion  cubic  feet  (TCF).Cl]   At current
domestic usage rates this is enough  to supply the United States
for  over  nine  years.[2]   These  proved  reserves  only include
identified  sources whose quantity,  quality and  location  are
known  and  which  can  be  economically  extracted  with  existing
technology.   Addition   of estimates  of   conventional  resources
which  have been identified  and are  estimated to be potentially
recoverable  economically  bring  the  total  U.S.   conventional
resource  base  total  to  about 900  TCF.[3]   Large  amounts  of
natural  gas from unconventional  resources  such  as  coal seams,
Devonian  shale,  geopressurized brines and  tight gas reservoirs
are  also  available but would  only  be economical  to extract at
somewhat   higher,   though  unclear,  increases  in   natural  gas
prices.[4]  For  the purposes of this report  it will be assumed
that  the  use of  CNG  as a  heavy-duty  vehicle  fuel  will not
impact  the  domestic  demand  or  price  of  natural  gas  to  any
significant degree.  This is a reasonable assumption given that
a   significant  penetration  of  CNG into  the   total  domestic
heavy-duty  fleet   (i.e.,  ten  percent)  would  result  in  an
increase  of natural gas use  of just  two  to three percent, still
well below total domestic usage rates of  the  early  1980s.

     The   United  States  currently  has   a  massive  natural  gas
transmission  and distribution pipeline  network  in place which
serves  a  large  portion  of  the country.   Also,   in most major
cities  and many other   areas there is  an extensive  distribution
network  in place which can  be easily tapped.  For  the purposes
of this study  it  will be assumed  that  CNG  would  generally be
used for  heavy-duty vehicles in  areas which  already  have   a
distribution  infrastructure  in place and thus,  no new  capacity

-------
                               4-2
would need to be installed.  Also, it  is  likely that heavy-duty
CNG applications  would generally be centrally  fueled fleets in
urbanized  areas  and  would  have ready  access  to natural  gas
distribution lines.

     The  American   Gas  Association  has   recently  published
projected natural gas  prices  for vehicle refueling  stations to
the year  2005. [5]   These  projections  for 1995 (in- 1989 dollars)
range  from  $2.59/mmBTU to  $5.89/mmBTU   (lower  heating  value)
depending on the region of the  country  being considered.   This
range of  prices  will  be used in this  report  to  develop natural
gas vehicle fuel cost estimates.

     No  attempt  will  be made  in this report  to predict future
natural gas  price  trends.   Unlike some other  alternative  fuels
such as  methanol,  where a  whole new  market  must be developed,
the use of natural gas as a heavy-duty vehicle  fuel  is expected
to  result  in  little  perturbation  in the  natural  gas market.
Thus, future predictions  of  price  trends  are  not  as  critical
here as in the analysis of other alternative fuels.

Ill. CNG Refueling Station Cost

     In  this  section  the  cost  of CNG  refueling stations  for
heavy-duty applications will be  examined.   First, a description
of  CNG  refueling  station  operation,  hardware  and some factors
influencing  station  design  will  be  presented.   Next,' some
typical   current   prices   for   different    refueling   station
components will  be  shown.   Finally,  the range  of total station
costs will  be  discussed.   Due to the wide variety of heavy-duty
vehicle  fleet  sizes   and   applications,  the  purpose  of  this
section  is  to give the reader  some idea of  the range of costs
involved  in  a  CNG refueling  station  rather than  to  define  and
cost   out   a   "model   station"   to   represent   the   "typical"
heavy-duty station.

     A.    CNG Refueling Station Hardware

     In  contrast to  liquid fuels, CNG  is  a gas  and must  be
compressed  for  storage onboard  a vehicle.   Thus, the  refueling
station  equipment needed  for  CNG  is  different  than  that  for
gasoline  or  diesel fuel.   Generally,  there are  two  methods of
refueling  a  CNG vehicle,  slow-fill   and  fast-fill.   Although
these two methods are  similar in  some respects,  they are quite
different  in  others.    A  diagram  of  a  typical  CNG  fueling
station utilizing both  fill methods is shown in Figure  4-1.

-------
                                    4-3

                               Figure  4-1
         MIGN-P«SSU«
         PIPING
      COMPRESSOR
                                                      HOSE DROP ASSEMBLY
                                                       • TIME-FILL REFUELINO
                                                        («• m 14 (Muni
• QUICK-FILL REFUELINO
  «to 5 i
                                          STORAGE CASCADE


                                           CASCADE PANEL




                                         FOUNDATION
                                  FILL HOSE
                    Typical  CNG Refueling Station
Source:    "Assessment  of  Methane-Related  Fuels  for  Automotive
            Fleet  Vehicles",  DOE/CE/50179-1, February 1982.

-------
                               4-4
     With the  slow-fill  method  there  is  a  direct  connection
between  the vehicle   and the  natural  gas compressor.   Thus,
refueling time  is  directly related  to the total  fuel  capacity
of the  vehicle  or  vehicles being fueled  and the  discharge rate
of the  compressor,  which  is  generally given in  standard cubic
per minute  (SCFM).   Slow-fill can take  less than  one half hour
for a single vehicle  fueled  on a  larger compressor, as  is  the
case,  for  example,  with  the Brooklyn  Union  Gas  (BUG)  buses
discussed   in   Chapter   3. [6]    However,  slow-fill   stations
generally refuel more than one vehicle at  a time  and often  use
smaller  compressors  than the BUG   program station.   Thus,  a
typical  slow-fill  operation   will  generally take  several hours
and is usually performed overnight.

     In  contrast  to   slow-fill,  fast-fill  systems  utilize  a
large-volume  compressed   gas  storage  system,   or  "cascade",
between the compressor and the vehicle.   Generally,  the cascade
is divided  into three banks  of  cylinders, with  each bank at a
different pressure.  The  lowest pressure  bank is  used  first  and
an automatic or  manual sequential  valve system switches to the
higher  pressure  banks as  the  pressure between the  cascade  and
the  fuel  tank  is  equalized.    Through  the successive connection
with  higher pressure  banks  the vehicle can be  refueled in a
matter  of several  minutes with fast-fill, as opposed to several
hours with  slow-fill.   Similarly to fast-fill  vehicle  fueling,
the  fast-fill  cascade  banks  are   filled selectively  by  the
compressor  through  a  priority valving  system,  usually  with  the
highest pressure bank being filled first.

     Station  design   including   choice   of   slow-fill  versus
fast-fill,  compressor sizing,  cascade  sizing  (if  fast-fill),
and  the number  of fuel  hoses is  determined  by  the number of
vehicles to be fueled, their  onboard fuel storage capacity,  and
their demand pattern.  A small captive fleet may only require a
small   compressor   and  slow-fill   capability  to  refuel  all
vehicles  together   overnight.   A  larger  fleet  will require a
larger  compressor,  more  fuel  hoses and  may  even  utilize some
fast-fill   capabilities.   With   this   system  vehicles  could
generally  be  slow-filled  together  overnight,  but  fast-fill
would be  available for  special-purpose  filling during  the day.
Finally,  a  commercial fueling  station  such  as  a truck stop
would  be exclusively fast-fill  and  would have  a very large
compressor  and  cascade  capacity,   as  well  as  fuel  dispensers
with metering capability  rather than simple fuel hoses.

     Natural gas compressors  are  generally two  to four stage
compressors  with   a   discharge  pressure   of   3,600    psig.
Compressors  for  small  installations   such as  fleet  stations
generally have  a discharge rate of well below 100   SCFM.  The
gas  supply for the  compressor  is taken  from  an underground
natural  gas transmission line,  much  like  a   residential home

-------
                              4-5
hook-up.   The natural gas pressure  in  these lines  is  generally
one  to   five   psig  for  distribution   lines,   although  main
transmission lines can have  pressures  of several hundred psig.
The line  pressure to the compressor affects cost and  sizing as
less  compression   is  needed  when  starting  with a higher  gas
inlet pressure.

     In  addition   to  the refueling  hardware  itself  there  are
other components to the CNG  refueling  station.   These  include a
concrete  pad  to  mount  the  compressor on,  as   well  as  any
enclosure  that  may  be  used  to  protect  the  compressor  and
cascades  (if any)  from  the  elements.   There  may also be  a
significant  piping  link  to  the gas  source  depending   on  the
proximity of the  station  to  the underground gas line.   However,
for most applications in any sizeable  city  this  would  likely be
a fairly short line.

     B.    CNG Refueling Station Hardware Cost

     In  order to  understand  the total cost of  a CNG  refueling
station  it  is useful to first look briefly  at typical costs of
the  various  components of  a  station.   At  the heart   of  any
refueling  station  is  the   compressor.   This  is  usually  the
single most  expensive component of the  station.  The  cost of a
variety of compressors from  various sources  was  compiled and is
presented  in Table  4-1.   As  can  be  seen there  is a fairly
linear   relationship between  the  compressor   cost   and  its
capacity.   Generally,  a small  fleet would  use  a  compressor on
the lower  end of  the capacity scale while  a larger fleet would
likely  need a  compressor  on the higher end of  this  scale.  A
high  volume  commercial  truck   stop  would  likely   require   a
compressor sized  larger than anything  on this table.

     Table  4-2  shows  typical  costs  for  other CNG  refueling
station   components.    The  fueling   post   for   slow-fill   or
fast-fill  application would typically cost $500 to  $1,000 per
hose   (vehicle).    In  contrast,  a   two   hose   dispenser  for
commercial   fast-fill   use,    which    may   include   metering
capabilities  and   a sequential  valve  system,   can   cost   over
$35,000.

      If  fast-fill capability  is desired,  a small  cascade and
associated  priority  and  sequential   valving  will cost  under
$15,000.   This  type  of  cascade system  would  generally  be  used
to  fast-fill only a few trucks during  the day.   If  the fleet
were   large  or  utilized fast-fill  exclusively, several small
cascades  or  one or more  large  cascades may be required.

-------
                                4-6
                            Table 4-1

          Typical CNG Refueling Station Compressor Costs
Source

  A
  B
  A
  C
  B
  B
  A
  C
  C
  A
  D
  A

Capacity (SCFM)
4.5
25
30
30
30
50
57.8
63-85
100
130
130
155
Inlet
Pressure (psiq)
0-5
—
0-5
5-15
—
—
0-5
40-60
150
0-5
15
0-5

Cost
$ 12,800
30,000
37,000
39,500
40,000
46,000
50,000
41,500
45,000
86,000
126,000
90,000
Sources:  References 6 through 9.

-------
                                4-7

                            Table 4-2
              CNG Refueling Station Component Costs
  Component
Fuel Post
Dispenser (2 nozzle)
Cascade (20 cylinder-9,200 SCF)
Cascade (3 20" x 22'  tubes,  27,000 SCF)
Sequential & priority valve systems
   Cost
$500-1,000/Vehicle
 25,000-35,000
  8,500-9,800
   35,000
 2,600-5,000(for both)
Sources:  References 6 through 8.

-------
                               4-8
     C.    Total CNG Refueling Station Cost

     The  total  cost  of  a  CNG  refueling  station  is  obviously
dependent a great deal on its total capacity and  whether  or not
it is  a  fast-fill  system.   To give some idea of the price range
of CNG refueling  stations,   Table  4-3  shows  estimates for  the
total  cost of  a  small  light-duty  fleet  station  and a  large
truck  stop,  as  well  as  the  actual  costs  of  the  refueling
station  for  the  BUG  bus  program  in  Brooklyn,   NY,  and  a
refueling  station  used  for  a  small  (10  vehicle)  school  bus
fleet  in  Syracuse,  NY.  The  small fleet  station  estimate shows
that   it  is  possible  for  a  small  fleet  (17-35  light-duty
vehicles,  significantly  fewer  for   heavy-duty)  to  utilize  a
slow-fill   refueling    station   which    costs    well   under
$100,000. [10]   Conversely,  a commercial truck  stop which would
require  a large compressor  and cascade  capacity,   as well  as
meter-type  fuel  dispensers  is projected  to  cost  well  over
$600,000.[11]   Although this  report is  intended to  cover  issues
specific   to   CNG  use   in  heavy-duty   applications,   it  is
reasonable  to  assume  that  the  lightest  of  heavy-duty  CNG
vehicles  (pick-up and  delivery vans,  for  example) may refuel at
public    stations    intended    primarily    for    light-duty
applications.   Cost  estimates for these  types  of stations were
developed  in  Volume  I  of   this  report  and  are  shown  to  be
$225,000 to $396,000.

     Table  4-4  shows  the cost  breakdown of the CNG  refueling
station  used  for  the   school  bus   fleet  in.  'Syracuse,  New
York.[12]  This  station utilized two  30  SCFM  compressors for a
total  capacity  of  60  SCFM.   It also had  a small  cascade for
fast-fill  purposes.   Although  this  station was  only  used for
ten buses its capacity would  allow a  fleet significantly larger
than this,  as  the compressors  were  only  required  to  run about
five hours  a day.   This station is an  excellent  example  of the
type of station a small captive fleet could use.

     From  the  perspective  of a somewhat  larger  station,  Table
4-5  shows  the  cost  breakdown of the  BUG  bus  refueling station.
This   station  is  a  simple  slow-fill  system  with  a   130  SCFM
compressor  and  a single  two-hose  metered-type dispenser.   The
total  cost  for this  system  is rather  high  for   the initial
application (two  buses)  for  a couple  of  reasons.   First,  the
compressor is  a  very large one for a slow-fill system servicing
a  two  vehicle  fleet.   However,  this  allows slow-filling  of  a
single bus in  less  than  thirty  minutes.   In  fact,   when the
compressor is  operating 20 hours a  day this  system is capable
or  refueling 45  buses in a  day.  Second,  Brooklyn  Union  Gas is
authorized to sell CNG  from  this station  commercially, thus the
expensive  dispenser  rather  than an inexpensive fill post.  This
fact   would   also  justify   the  large   compressor  capacity.
Finally,  the  installation  and materials  cost  is somewhat high
as  the original  concrete slab and support  pilings  were placed,
unbeknownst  to  the  designers,  on  an abandoned landfill  with
poor   soil  conditions.   Thus,   following  initial  startup  the
system had to  be  disconnected so  that a new  concrete slab and
additional  support  could be  added,   adding  three weeks  to the
initial  five-week installation time.

-------
                                  4-9
                              Table 4-3
          Total Costs for Heavy-Duty CNG Refueling Stations

       Station                                       Cost
Small fleet station                                $ 81,500
Syracuse school bus station                         101,933
BUG bus station                                     278,108
Commercial truck stop                               641,000
Sources:  References, 6,10,11,12.

-------
                              4-10



                           Table 4-4


       Cost Breakdown for Syracuse Bus Refueling Station


           Equipment  (incl. two  30
            SCFM compressors)               $  78,133
           Compressor pad                       3,500
           Installation                         1,000

           Total (1982 dollars)              $  82,633

           Total (Current dollars)           $101,933
Source:     "Dual-Fuel  School Bus Demonstration", New  York State
           Energy Research  and Development  Authority,  Albany,
           NY,  October 1986.

-------
                              4-11


                           Table 4-5


           Cost Breakdown  for BUG Refueling Station
          Compressor  (130 SCFM)             $126,286
          Dispenser  (two hose)                25,000
          Other Materials                      9,653
          Installation                       117,169

          Total                             $278,108
Source:     C.    Spielberg,   "Compressed   Natural  Gas   Program
           Monthly Report #2, January-May 1989," New York City
           Department of Transportation,  September 12,  1989.

-------
                              4-12
     With this perspective  on the BUG station,  it  appears that
a  slow-fill  system  designed  for  overnight  fill  of  several
vehicles  could  be  substantially  cheaper.   Use  of   a  smaller
compressor  and  simple  fill  posts  rather  than  a  meter-type
dispenser, together with the  elimination  of  the unexpected soil
condition problem, could easily  cut  the  cost of this  system in
half.  A  larger  compressor  and  the  addition of  some fast-fill
capacity  (at  $15-20,000  total  additional  cost  for  cascade,
priority and  sequencing valves,  additional  fueling hoses, etc.)
would  still  likely  keep  total  system  cost  under  $200,000.
Thus, for  the purposes of  a  small to moderate  sized fleet the
total  refueling  station  cost  could  range  from  $80-200,000
depending  on  total   capacity  and  whether  or  not  fast-fill
capacity is included.    Conversely, the worst case would likely
be  the  large  fast-fill  truck  stop  at  over  $600,000.[11]
However, for the purpose of  this report it will  be assumed that
most heavy-duty  CNG applications will  utilize a smaller station
in a fleet setting.

     It should be noted  that  the cost of land  is  not included
in this  discussion  as part of  the total  CNG refueling station
cost.  For  the purposes of  this  report  it  is  assumed  that  a
fleet CNG refueling station would  be constructed on the site of
the  fleet's current  refueling  station and  no  additional  land
costs would be incurred.   However, EPA recognizes  that,  due to
the physical size of  the the compressor and  cascades, placing a
CNG  refueling station  within the  physical  confines  of  an area
designed for a diesel or gasoline  refueling  station may present
a problem at some facilities,  resulting in some added cost.

     Finally,  in the  case of  truck  stops, which  are generally
located away  from urban  areas, there  may be a  cost  associated
with  the  pipeline  required  to  connect the   station  to  the
natural   gas   distribution   pipeline.    This   connection  was
estimated by  the Department of Energy to  average  between l and
5  miles  long  depending  on the  area  of  the  country, and  was
estimated to  cost $200,000/mile.[11]   However,  for the purposes
of this report  it was  assumed the most heavy-duty applications
of  CNG,   at  least in  the  near term,   would  be  in centrally
fueled  fleet  settings in  urban areas,  which  would  not  incur
this cost.   Thus,  it was not included in the economic analysis
contained in this chapter.

IV.  Heavy-Duty CNG,  Gasoline and Diesel Vehicle Fuel Cost

     This section will present an  estimate of  the relative fuel
costs  for  both  current  and  future  CNG,  gasoline  and  diesel
fueled heavy-duty vehicles.   First,   an  "equivalent  gallon"  of
natural  gas  in  relation  to  gasoline  and diesel  fuel will  be
established  so the  fuel  costs  can  be projected  on  an  energy

-------
                              4-13
basis for  direct  comparison.   Second, the  cost of  compressing
the  natural  gas   for  refueling  will  be   calculated.    The
capitalized refueling station cost will  then be calculated  and
all  of  the factors  will be  combined for  a comparison of  per
equivalent gallon fuel costs.  Finally,  the total vehicle  fuel
costs  for both  current   and  future  CNG,   gasoline  and  diesel
fueled heavy-duty vehicles  will  be estimated  using these  fuel
prices, and factoring  in relative  engine efficiencies  and other
relevant factors.

     A.    Basis of  Comparison

     Although natural  gas  is  stored and  burned  in a  gaseous
form,  it  is  easiest  to  compare  natural  gas consumption  to
consumption  of  gasoline or  diesel   fuel   on   an  energy  basis
(i.e., "equivalent gallon").  The  energy density of natural gas
is  typically 1,030  BTU/SCF  higher  heating  value  (HHV).[13]
However,  for  purposes of  comparison with  gasoline and  diesel
fuel,  the lower heating value  (LHV)  of  natural  gas  must  be
used.   This  value  is  not  usually quoted, but an  earlier  EPA
technical  report  determined  that the   LHV of  natural  gas  is
typically  around  90 percent  of the  HHV.[14]    Thus,  an  energy
density  of 930  BTU/SCF  will  be used  for  natural  gas in  this
report.   Comparing  this  to the BTU/gal value  for  diesel  fuel
from  Table 3-1  and  114,132  BTU/gal  for  9   RVP  gasoline[13]
yields an  energy equivalence  of  122.7 SCF of natural gas to one
gallon of  gasoline,  and  139.6 SCF of natural  gas  to one  gallon
of  diesel  fuel.    These relationships  will  be  used for  the
comparison of relative fuel costs between the fuels.

     B.    Compression and Station Maintenance Costs

     The  natural  gas  prices  just presented  are  not  the  only
factor  in the  cost  of  natural gas to the heavy-duty CNG vehicle
operator.   As   will  be  seen, there  is  a  cost  of energy  to
compress  the natural  gas  during refueling, which  is  dependent
on  the  efficiency of the compressor  and the price  of  energy to
power  the  compressor  motor.  Also,  station  maintenance  costs
are  significant  enough to  consider.    These  costs  must  be
factored  into the cost of natural gas as a vehicle fuel.

      Data   on   the   energy   used  per  volume   of   natural  gas
compressed are  available for  four  different compressors  in both
public  and fleet use  in Canada.[15]  These  compressors  varied
in  output delivery   capacity from 20  to 178  SCFM.   The  energy
used to  power  the  compressors   ranged  from  0.0075  to  0.0099
KW-hr/SCF, depending on  compressor efficiency.

      Compressors  currently  used in  CNG  refueling  stations  are
generally powered  by  electric  motors.    It  is  reasonable  to
assume  that  in  the  future compressors may  be powered by natural

-------
                              4-14
gas  engines   at   a  significant   energy   cost   savings   over
electricity.   This  is  especially likely in  the case  of  large,
commercial   station  compressors.    However,   since   current
compressors  generally  use  electricity   and  most  heavy-duty
applications   of    CNG   are   expected   to   utilize   smaller
compressors,  electricity  costs will  be  used here  to  calculate
compression  costs  while recognizing  that  these  costs  may  be
significantly reduced in some future cases.

     The 1988 national  average commercial  price for electricity
was 7.Oltf/KW-hr.[13]  Using the  equivalent gallon relationships
previously derived along  with  this  current electricity  price
yields current natural  gas compression  costs of  6.4-9.0  cents
per  equivalent  gallon  of  gasoline,  and  7.3-9.7  cents  per
equivalent gallon of diesel fuel.

     Based on  actual  maintenance cost  data  from  a variety  of
actual  CNG  refueling   stations,  DeLuchi  et.al.   estimated  CNG
station    operation    and    maintenance   costs    to    be
$0.25-0.50/mmBtu.[16]      Using    the    energy    equivalences
established  earlier  results  in   CNG   station  operation  and
maintenance  costs  of  $0.03-0.06 per  equivalent gallon for both
gasoline and diesel fuel.

     C.    Capitalized Service Station Cost

     In order to include the capitalized refueling station cost
in  the  cost  of   natural  gas   a  typical  refueling  station
configuration  was  assumed.   In   general,   station  cost  and
station capacity  are  somewhat linear  [17],  and the capitalized
refueling  station  cost  is probably much more  dependent  on the
utilization  rate   of  the  station  than  on it's  capacity.   For
example,  an  urban  transit bus  fleet  would  have  a  fairly low
utilization  rate  as it's  vehicles would be on  the road all day
and could  only be  refueled  at night.   In contrast to this,  a
fleet of delivery  vehicles which are  in and out of the facility
throughout the day  could  much  more  effectively  utilize  some
fast-fill  capacity during  the day  in addition  to refueling at
night.  Thus,  for  the purposes of this report  a  single station
design  will  be  used  and  capitalized  refueling  station  costs
will be derived using high and low utilization  rate scenarios.
     The  refueling  station  chosen   is   similar   to  the  one
previously  discussed for  the  Syracuse  school  district.   This
station has  two cascade banks  for fast-fill  capability  and 60
SCFM compressor  capacity.   Further, this  station  is assumed to
cost about  $100,000.   The  per-equivalent-gallon costs will  be
calculated based  on a 10 year payback period  with  a 10 percent
rate of return.

-------
                              4-15
     For the  low  utilization rate scenario it was  assumed that
the vehicles  would  only be  available  for  refueling  8 hours  a
day.   In  addition to  operating the compressor  for 8 hours  in
slow-fill operation,  the cascades  could be  used  initially  to
fast-fill  additional  vehicles  during  this   8   hour  period,
utilizing the compressor  an  additional  4  hours  a  day  for
cascade   filling.   This  means   that  this  station  could  refuel
about 11 urban transit  buses,  each using  the equivalent  of  27
gallons  of diesel fuel  a day (similar to the BUG buses).   This
scenario  yields   capitalized   refueling   station  costs   of
12.80/gal for gasoline and 14.6#/gal for diesel  fuel.

     For the  high utilization  scenario  is was  assumed that the
compressor could  be  operated 20 hours  a day,  making efficient
use  of   both  slow-  and  fast-fill  capacity.    Although  this
scenario  is  probably not  representative  of  an  urban  transit
operation, for  purposes of  comparison  to  the  low utilization
scenario this type  of  operation  could  refuel  around 20  urban
transit   buses daily.   This  high  utilization  scenario  yields
capitalized  refueling  station  costs  of 7.50/gal  for  gasoline
and 8.20/gal for diesel fuel.

     D.     Relative Fuel Prices

     Tables  4-6  and  4-7 shows  the relative comparison  of fuel
prices  between  gasoline  and  CNG,  and diesel   fuel  and  CNG,
respectively.  The  gasoline prices  were taken from  a previous
EPA  report. [18]   The diesel cost  is  the  average of  the  first
seven months  of  1989,  without  taxes.[13]   Although state taxes
for  gasoline and diesel  fuel  are similar,  Federal   taxes  for
diesel  fuel  are  six cents/gallon  higher than for gasoline.  In
both  cases,  however,  the same  gasoline tax was  assumed  to  be
applied  to  CNG  on  an  energy-equivalent basis.   This  was done
because  it  is likely that a single taxation strategy would be
used  for  CNG,   and  given   the  potential  for  light-duty
applications,  equivalent  gasoline  taxes   seem   most  probable.
The  tax  on  a diesel-equivalent gallon of CNG is  somewhat higher
than  for  a  gasoline-equivalent   gallon of  gasoline,  due  to
diesel fuel's higher energy density.

      E.     Relative Vehicle Fuel Costs

      The  total   vehicle fuel  cost comparison  between  CNG and
gasoline  or  diesel  fuel must  take into consideration not only
relative  engine  energy use, but  also the  fuel  economy effects
of  increased fuel storage weight  and,  as  was  just  discussed,
the   cost  of natural  gas   compression.   As was  discussed  in
Chapter  3,  a current stoichiometric  CNG engine  uses almost 11
percent  less energy  than its  gasoline counterpart,  while the
future  optimized CNG  engine  uses  some  fifteen  percent  less

-------
                               4-16
                           Table  4-6
      Gasoline and CNG Energy  Equivalent  Price Comparison

Cost Classification                    Gasoline     Natural Gas
Extraction, refining, other             $0.69
Long-range and local distribution        0.06
Current natural gas end user
  delivered price range*                  —        $0.30-0.67
Service station markup**                 0.09
Capitalized refueling station cost        —        $0.08-0.13
Compression cost                          —        $0.06-0.09
Operation and maintenance cost            —        $0.03-0.06
Profit markup***                          —        $0.00-0.01
Taxes                                    0.24            0.24
Total                                  $ 1.08       $0.71-1.20
*    The  end user  price  range  for  natural  gas  includes  all
     distribution costs
**   The  service  station  markup  for  gasoline   includes  all
     overhead and operating costs as well as profit markup
***  The  $0.00   profit  markup  applies  to  fleet-owned stations
     while a profit of $0.01 was assumed for commercial stations

-------
                              4-17

                           Table 4-7

     Diesel  Fuel  and CNG Energy Equivalent  Price Comparison


Cost Classification                    Diesel  Fuel  Natural Gas

End user diesel fuel price*             $0.56

Current natural gas end user
  delivered price range                   —        $0.34-0.76

Service station markup**                $0.09

Capitalized refueling station cost        —        $0.08-0.15

Compression cost                          —        $0.07-0.10

Operation and maintenance cost            —        $0.03-0.06

Profit markup***           .               —        $0.00-0.01

Taxes                                    Q.30            0.27

Total                                  $ 0.95       $0.79-1.35
*    The  end user  diesel  fuel  price is  the price  charged to
     fleet   operators   rather   than   for   resale   (i.e.,   to
     commercial retail outlets)
**   The  service  station  markup for  diesel  fuel  includes  all
     overhead and operating costs as well  as  profit  markup
***  The  $0.00 profit  markup  applies to  fleet-owned stations
     while a profit of $0.01 was assumed for  commercial stations

-------
                              4-18
energy.   Conversely, a current  technology lean-burn engine uses
39 percent more  energy than  its  diesel  counterpart,  while the
future  optimized lean-burn  CNG  engine  is  assumed  to  use  25
percent more energy.

     Fuel  storage  weight  is  an  important  part  of  the  total
vehicle fuel cost as  it  affects total vehicle weight and,  thus,
fuel economy.  For  the purpose of this  study  it  will be  assumed
that  the vehicles  being  compared  will  have equivalent  range
(i.e.,  equivalent   fuel  capacity).    This assumption  is  being
made  in order to  allow a  direct comparison between the  fuel
types.  This may be somewhat of a worst  case  approach,  as  many
heavy-duty vehicles could  likely operate satisfactorily  with
less fuel storage capacity  than they currently  have.   However,
as  will  be   described,  below,   the  overall  impact  on  fuel
consumption of maintaining equivalent range is small.

     For the purpose of calculated fuel  storage  weight and fuel
economy   penalties   two  model   vehicles  were   chosen  which
represent  excellent  candidates  for  dedicated  CNG  use  in  a
captive  fleet  setting.   The  first   is  a UPS parcel  delivery
truck  similar  to the  ones  discussed  in Chapter  3,  with  the
exception that this vehicle will be assumed  to  be a dedicated
CNG vehicle  and  will  not  have the 30 gallon  gasoline tank and
fuel weights included  in the weight  calculation.  The second is
an urban  transit bus  with  the CNG fuel  storage  equivalent  of  a
100  gallon  diesel   fuel  tank.  The  actual weight calculations
and resultant  fuel  economy  penalties are derived  in Chapter  5.
However,  the results  of these calculations  show  that bringing
the UPS delivery truck to  the equivalent range  on CNG  as  it
would have as  a  gasoline  vehicle would result in a five percent
fuel  consumption   increase  on  CNG  compared   to  gasoline.
Similarly, the transit bus would have  a seven percent increase
in fuel consumption.

     Combining  the  relative   engine  efficiencies,  the  fuel
storage weight impacts  and  the fuel  prices previously presented
yields  the  relative  total  vehicle  fuel costs  shown in  Table
4-8.    For   the  reasons   given  in  Chapter  3,  the  gasoline
comparison  is based  upon   the  stoichiometric  combustion  CNG
engine  performance  while  the diesel  fuel comparison is  based
upon the  lean-burn  combustion CNG engine.  As can  be  seen from
the  table,  the  stoichiometric  combustion   CNG   engine  offers
significant  potential  fuel  cost savings  over  an  equivalent
gasoline  vehicle.   Conversely,  the  fuel  economics of replacing
a  diesel  engine  with  a  dedicated CNG  engine are not as  good,
especially  with  current   CNG  technology.    This  was  to  be
expected given the fact that  diesel  engines  already represent  a
very  efficient form of fuel  combustion  and  the  relatively low
cost of diesel fuel compared to gasoline.

-------
                              4-19
                           Table 4-8
             Vehicle Fuel Costs  (Gallon Equivalent)

                      Gasoline Comparison*
Gasoline
CNG - Stoichiometric Combustion
Current
  $1.08
$0.67-1.12
                       Diesel Comparison**
                                    Current
Diesel Fuel
CNG-Lean Burn Combustion
  $0.95
$1.18-2.01
                                                  Optimized
$0.63-1.07
              Optimized
$1.06-1.81
*    Cost per gallon, or equivalent gallon, of gasoline.
**   Cost per gallon, or equivalent gallon, of diesel fuel.

-------
                              4-20


V.   Heavy-Duty CNG Engine and Vehicle Costs

     A.    Engine Costs

     There  is  currently  very little  information available  on
the  cost differential  between  a heavy-duty  CNG engine  and  a
heavy-duty  gasoline  or diesel-fueled engine.   What  information
is available characterizes the  costs associated with converting
a diesel-fueled  engine to either  dual-fuel or  dedicated  spark
ignition  CNG  operation.   Estimates  show  that  conversions  of
diesel engines to  dedicated  spark ignition CNG operation  range
in cost  from $3,100 to  $5,600.  [11]  These  cost estimates are
for   engine  conversion   only   and  do   not  include   such
vehicle-related components as CNG fuel cylinders.

     Although  conversion  has historically been  the  method for
obtaining heavy-duty CNG  engines,   there  is  a  clear  movement
within  the  heavy-duty  industry  toward  the  introduction  of
dedicated   CNG   engines   by   original  equipment  manufacturers
(OEM).   This  is  evidenced by Cummins'  commitment to  offer its
CNG L-10  engine  starting  in  1991.  Also,  most  major  heavy-duty
diesel  manufacturers are currently  involved to  some  degree  in
the development and assessment of dedicated CNG engines.

     It  is difficult to predict with  any  accuracy what the cost
of an  OEM-supplied dedicated CNG engine will  be.   In general,
though,  the bulk  of the  costs  associated  with conversion  to
dedicated  CNG  operation  are for  new  CNG-optimized  parts  to
replace the diesel-optimized  parts,  such  as the pistons, piston
rings,  cylinder  heads, camshaft  and intake  manifold.   For OEM
engines  these parts  would likely be similar  in cost  to current
parts.   The addition of  a spark ignition system would be needed
at  some  cost.   Some  savings may  be  possible  with   the  fuel
system  depending on the type of  system  used (i.e.,  mixer, port
injection,  direct  injection,  prechamber)  since  the  diesel fuel
injection system which  it would replace  is  generally regarded
as one  of the  more expensive systems on a diesel engine.   Some
initial  recovery of research and development costs  is expected
to result in a  higher  engine introduction  cost than  would  be
expected  in the  long  run.   In  general,  however, EPA expects
that  an  OEM  mass-produced,  dedicated  CNG  heavy-duty  engine
would  have  at  most  only   a  modest  price   increase  over  a
comparable  diesel  or gasoline engine and that  this  cost  would
not be large in comparison with the total engine cost.

     B.     Vehicle Costs

     As  with CNG  engines, there  are some vehicle modifications
which  would require the  engineering and developments  costs  to
be recovered in initial  vehicle  offerings without resulting in
a  significant   long   term   price  increase  for  the  affected

-------
                              4-21
components.   These would primarily include modifications  to  the
frame  and  engine compartment  to  accomodate  the  CNG  fueling
system (i.e., storage cylinders,  pressure  regulators,  refueling
interface).    Also,  the  exhaust  catalysts  to  be  used  on  CNG
vehicles are not  expected  to  differ  significantly  in  cost  from
those currently  used  on gasoline vehicles or  those expected to
be used  on  diesels.   If CNG  were replacing  a diesel  equipped
with a  trap oxidizer for  particulate  control,  significant  cost
saving would result.  However,  the  degree  of trap  oxidizer  use
to be  expected  for diesels is unclear  and difficult  to  predict
at the present time.

     The only area which  is  expected  to  increase  the cost  of
heavy-duty  CNG  vehicles to  any  significant  degree is that  of
fuel  storage.   As was  discussed in  Chapter  3,  fiber-wrapped
steel  tanks  are  currently  the most  likely choice  for  CNG
vehicle use.  Fiber-wrapped  aluminum cylinders also  offer  some
weight advantages  at  an increased price.  The current prices of
some typical  cylinders  are  shown in  Table  4-9.[19,20]   It  is
apparent  from   the   table  that  several   cylinders  would  be
required to have any  reasonable  amount  of  storage onboard.   For
reference,  The  Flxible Corporation  (transit  bus  manufacturer)
quotes  a  typical transit  bus  diesel fuel tank at  $1,134.[21]
It  is  readily  apparent  that   the  fuel  storage  cost  would
increase several times when moving to CNG, as  is  shown in Table
4-10.   However,  this  is still a  modest price increase when one
considers  the  base  price  of   a  diesel-powered  Flxible  bus
($173,120)..   This cost,  however,  would be  more  difficult  to
absorb  in a much smaller  vehicle.  It  is  also  apparent  that in
some   applications   the   available   space   may   make   adding
sufficient  fuel  storage capacity for equivalent range difficult
or impractical.

-------
                                                Table 4-9

                                 Prices of Typical CNG Storage Cylinders
  Type
Fiber-wrapped
   Steel
Fiber-wrapped
  Aluminum
Size (inches)
14 x 43.6
                    14 x 53.6
                    16.3 x 31.6
                    16.3 x 53.6
13 x 57.4
Capacity (SCF)
750
950
720
1260
Equivalent Capacity (gal) Purchase
Gasoline Diesel Fuel Volume
6 5.5 1-49
200+
7.5 6.9 1-49
200+
5.7 5.3 1-49
200+
10 9.2 1-49
200+
Unit
Cost
$ 610
470
$ 727
561
$ 644
497
$1007
777



-p-
i
1-0
K3
825
6.6
50-100      $2,400*
*    Current price  for  low  volume  - may  eventually fall  below  $1500  with
     dedicated production equipment and large quantities.
Sources:  References 19 and 20.

-------
                               4-23

                           Table 4-10

                     CNG Fuel Storage  Costs


                             UPS Vehicle           Urban Bus

Number of CNG cylinders           3                     11
     (16.3x53.Sin)

CNG cylinder cost*           $2,330-3,000         $8,500-11,000

Fuel tank savings               $340                  $1,134

Net fuel storage cost**      $1,990-2,660         $7,366-9,866
*    Price range is result of purchase volume range.
**   This  is  a  worst  case  approach,  as  CNG  vehicles  may not
     require equivalent energy  storage  capacity as conventional
     vehicles   (for  example:    the   BUG  bus  uses  less  than
     one-third of the equivalent diesel capacity a day).

-------
                              4-24
                      Chapter 4 References

     1.    "Natural    Gas    Monthly,"    Energy    Information
Administration, DOE/EIA-0130(89/07).

     2.    "Assessment  of  Costs  and  Benefits  of Flexible  and
Alternative  Fuel   Use  in  the   U.S  Transportation  Sector",
Progress Report Number  One:  Context  and Analytical  Framework,
DOE/PE-0080, January 1988.

     3.    "Assessment  of  the  Natural Gas  Resource  Base of the
United States", DOE/W/31109-H1, May 1988.

     4.    "Unconventional  Gas  Sources",  Vols  I-V,  National
Petroleum Council, December 1980.

     5.    "Natural  Gas Prices  for  the  Vehicle  Market,"  Issue
Brief  89-19, American  Gas Association,  Arlington, VA,  November
22, 1989.

     6.    C.   Spielberg,    "Compressed   Natural  Gas   Program
Monthly Report H2, January-May  1989",  New  York City Department
of Transportation, September 1989.

     7.    G.   Barker,   Norwalk   Company,   Inc.,    telephone
communication, November 1989.

     8.    D.   Leivestad,    Carburetion   and  Turbo   Systems,
telephone communication, November  1989.

     9.    M.     Garret,     Bauer     Compressors,     telephone
communication, November 1989.

     10.   R.L.  Bechtold,  G.   Wilcox,   "An  Assessment  of  the
Infrastructure Required to  Refuel a  Large Population of Natural
Gas  Vehicles",  SAE  paper  No.   892066,   Society  of  Automotive
Engineers, Warrendale,  PA, 1989.

     11.   "Assessment  of  Costs   and  Benefits of Flexible and
Alternative  Fuel  Use  In   the  U.S.  Transportation  Sector,"
Progress   Report   Three,   Vehicle   and    Fuel   Distribution
Requirements   (draft),  U.S.  Department   of  Energy,  Office  of
Policy, Planning and Analysis,  July 1989.

     12.   "Dual-Fuel   School   Bus  Demonstration,"    New  York
State  Energy  Research  and Development  Authority,   Albany,  NY,
October 1986.

     13.   "Monthly   Energy   Review   -   July   1989",   Energy
Information Administration, DOE/EIA-0035  (89/07).

     14.   R.I.   Bruetsch,    "Emissions,   Fuel   Economy,   and
Performance   of   Light-Duty  CNG  and  Dual-Fuel   Vehicles",
EPA/AA/CTAB-88-05, U.S. EPA, MVEL, Ann Arbor, June 1988.

     15.   "CNG   Compressor   Operations   Monitoring",   Canada
Department of  Energy, Mines and Resources, January 1986.

-------
                              4-25
                 Chapter 4 References  (cont'd)

     16.   M.A.  DeLuchi, L.A.  Johnston,  D.  Sperling,  "Methanol
vs.  Natural  Gas  Vehicles:    A comparison  of Resource  Supply,
Performance,   Emissions,  Fuel  Storage,   Safety,   Costs,   and
Transitions",  SAE  Paper  No.  881656,  Society  of  Automotive
Engineers, Warrendale, PA,  1988.

     17.   "Assessment of Methane-Related  Fuels  for  Automotive
Fleet Vehicles," DOE/CE/50179-1, February 1982.

     18.   "Analysis  of  the  Economic  and  Environmental  Effects
of  Methanol  as  an  Automotive  Fuel," Special   Report  of  the
Office of Mobile Sources, OAR, EPA, September 1989.

     19.   Pressed Steel Tank Co.,  Inc.,  Price List, Milwaukee,
WI, October  1989.

     20.   S.   Anthony,   Structural   Composites   Industries,
personal communication, November 1989.

     21.   Letter from T. Leedy,  The Flxible Corporation, to P.
Como, Southern California Rapid Transit District, January 1989.

-------
                              4-26
       References For Further  Information On CNG Economics

     R.  L.  Bechtold,   et   al.,   "The  Practical  and  Economic
Considerations of Converting Highway Vehicles to Use  Natural Gas
as  a  Fuel,"   SAE   Paper   No.  831071,  Society  of  Automotive
Engineers, Warrendale,  PA,  1983.

     K.  G.  Darrow,   "Economic  Assessment of  Compressed  Natural
Gas  Vehicles for  fleet Applications,"  Gas Research Institute,
Chicago,  IL, September 1983.

     "The  Gas  Energy   Supply Outlook Through   2010,"   Policy
Evaluation   and  Analysis   Group,   American   Gas   Association,
Arlington, VA, October 1985.

     "The   Economics   of  Alternative  Fuels   and  Conventional
Fuels,"  SRI Internatinal,   presented  to the  Economics Board  on
Air Quality and Fuels,  February 1989.

     Vice  President's  Task Force  on  Alternative  Fuels,  "Report
of  the  Alternative  Fuels  Working Group,"  Office  of the  Vice
President, Washington,  DC,  July 1987.

     R.  D.  Fleming,  R.  L.  Bechtold,   "Natural  Gas,  Synthetic
Natural   Gas  and   Liquified   Petroleum  Gases   as   Fuels   For
Transportations," SAE  paper No.  820959, Society of Automotive
Engineers, Warrendale,  PA, 1982.

-------
                           CHAPTER 5

     AIR QUALITY  IMPACTS OF CNG USE IN HEAVY-DUTY VEHICLES


I.    Introduction.

     Compressed natural  gas (CNG) is  an alternative  fuel  that
has the potential to provide significant benefits  for  urban air
quality.  This  chapter  will provide an  analysis of  the overall
air  quality  impact  of  using  CNG as   a  fuel  f-or  heavy-duty
vehicles,   focusing especially  on its   impacts  on  urban  ozone
levels, air  toxics,, and global warming.   Emissions of  both CO
and NOx will  also be briefly touched upon.

     As was  noted earlier,  CNG technology is  still undergoing
development   and   optimization.    As   such,   it   would   be
inappropriate to  discuss the  impacts  of  CNG vehicles based only
on current emission levels.  Thus, this analysis  will  compare
both  current and advanced  CNG vehicles  to  the  corresponding
petroleum-fueled vehicles.   Also,  the analysis  is  structured to
allow  comparison  of  lean-burn CNG  technology  to  comparable
diesels,  and stoichiometric  CNG  emissions  to  gasoline-fueled
engines.   This  is appropriate  because the  lean-burn engine was
derived from a  diesel  engine  and will  likely replace diesel
engines, while  the  stoichiometric engine  was derived  from and
will   likely  be   used  primarily   as   a    replacement   for,
gasoline-fueled  engines.   As was noted  in Chapter  3,  there are
currently different  test  cycles for diesel  and gasoline-fueled
(or  Otto-cycle)  engines.   To  allow  for  proper  comparison to
gasoline-fueled   vehicles,   the stoichiometric  CNG  emissions
listed here  are from the Otto-cycle testing.

II.  Urban Ozone  Levels

     One of  the most notable air quality benefits of CNG use is
its  ability  to  reduce the contribution of mobile  sources to the
urban  ozone  problem.   This benefit  could  be  significant  with
respect  to  heavy-duty  vehicles,  since traditional heavy-duty
vehicles  account  for  about  one-fifth  of  mobile  source   ozone
precursor  emissions.   The  effect  of  mobile  source  control
programs on  urban ozone levels  is generally  estimated  from the
change  in  non-methane  hydrocarbon (NMHC) emissions  alone,  even
though  a  given  change  in  emissins  does  not  produce  a linear
change  in ozone  levels  because  urban  ozone  formation   is   a
complex  process  involving  the chemical  reactions  of  NMHC and
NOx.   (The  effect  of  CNG  use  on NOx  emissions   is  discussed
further    below.)    Futhermore,   when   comparing   completely
different  fuel-types (such  as  CNG and  petroleum)  the  issue of
photochemical  reactivity  can  become  equally  important.   Most
NMHC emissions  from CNG vehicles are very  light  paraffins  such
as   ethane   or  propane.[1]    These species  are generally   less

-------
                               5-2
reactive   than  the   NMHC   emissions   from   petroleum-fueled
vehicles.[2,3]  Thus,  the benefit  from  NMHC reductions achieved
by using CNG  would  be  expected to be enhanced by a reduction in
the reactivity of  the emissions  as  well.   On  the other  hand,
however, CNG  vehicles  also emit  such highly  reactive compounds
as formaldehyde  and propylene.[1]   These  emissions  may  offset
this  benefit  to some  extent.  Unfortunately,  due both  to the
complexity of  the photochemistry  of  urban  ozone  formation, and
the  fact  that  only  a  limited   amount  of  speciated  exhaust
hydrocarbon  data for  CNG  vehicles   is available,   it  is  not
possible to quantify this  additional effect at this time.  Thus
in this  analysis,  only the  relative NMHC  emissions  are  shown,
but the reader is  reminded that  the actual ozone benefit may be
be greater than that  estimated  simply from NMHC  reductions.
This  is  different  from  how  the  Agency has  been handling the
impacts of methanol-fueled  vehicles; which emit  primarily only
two components  (methanol  and  formaldehyde).  The photochemistry
of both of these compounds  is, however,  fairly well understood,
partly  because  they  are  both one  carbon  molecules  that have
simpler chemistry than  larger molecules.    It  is  clear,  though,
that  further   information  on  the  speciation  of  NMHC emissions
from  heavy-duty CNG vehicles is  needed  before  their relative
impacts can be fully quantified.

     The estimates  of  emissions  from petroleum-fueled vehicles
which will be used here  correspond  to engines   in  a condition
similar  to the  CNG engines  that were used  as the  basis for
estimates   of   the   emissions    from   CNG   vehicles    (i.e.,
well-maintained  low-mileage  test engines).   As  was noted  in
Chapter 3, this  condition was used  as  the basis for emissions
estimates  because it  is  the only condition for which  sufficient
data  are   available  for  CNG  vehicles.   Some  CNG engines have
shown a  tendency toward high in-use emissions,  but  EPA expects
these difficulties  to  be  dealt with as the technology matures.
Thus, it  must be emphasized  that the  actual  in-use effects of
CNG  vehicles  relative  to  petroleum-fueled  vehicles could  be
different  from  those  described  here,  either better  or  worse,
depending  on  in-use emission  variability and deterioration.

     The estimates  of  exhaust emissions from  1991  diesels have
been  selected   to  represent  expected   emissions of 1991  bus
engines.   They were calculated using manufacturers'  test data,
and  scaling  the emissions  to meet  newer standards  for  NOx and
particulate.   Bus  engines were used for the  current estimates
since urban  buses  represent  a prime application for lean-burn
CNG  engines   in  the near  term.    Farther   into  the  future CNG
engines    could   be   used   more   broadly  in   other    diesel
applications,  since all diesels will need  to  meet the stringent
particulate  standards  after   1994.   The  1994  diesel estimates
are   the   results   from   testing   of   a   prototype    of    a
advanced-technology  1994  diesel   (non-bus)  engine produced  by
Navistar.[4]   Diesel exhaust  emissions  are summarized in Table
5-1.

-------
                                   5-3


                               Table 5-1

          Petroleum-Fueled Diesel Exhaust Emissions (q/BHP-hr)


              Current Bus Engines
Pollutant     1989 6V92  1990 L10      1991 Projected  1994 Navistar1

CO            1.5         2.5          1.72                  1.4
CC>2           640         549          6223                  574
NOx           8.2         5.01         4.54                  4.44
PM            0.32        0.37         0.225                 0.08
NMHC          0.636       0.466        0.4Q6                 0.296
Total HC      0.66        0.48         0.427                 0.30
1.   Emissions after catalytic treatment.
2.   Average of 6V92  and L10 CO  emissions  weighted by  relative sales
     volumes (80%/20%).
3.   Average  of  6V92  and  L10  C02  emissions  weighted  by  relative
     sales volumes.
4.   Needed to meet 1991 standards.
5.   Needed to meet  1991 heavy-duty diesel  standards.   Bus standard is
     0.1 g/bhp-hr.
6.   95 percent of total HC for diesel engines (EPA-45012-88-003a) .
7.   Average  of  ratios  of  total HC  to  PM for  6V92  and  L10  engines
     weighted  by  relative  sales volumes and  multiplied  by PM level
     need to reach 1991  standard.

-------
                              5-4
     In   addition   to   exhaust,   estimates   of   evaporative
emissions,  running  losses  and  refueling  emission  are  also
needed.  To  date  there  is  no data suggesting  that  diesels have
significant evaporative  or  running loss emissions,  so  these are
assumed  to  be  zero.    Diesels  also  have  minimal  refueling
emissions  due   to  the   low   volatility   of   diesel   fuel.
Preliminary EPA data have  shown such emissions  from diesels to
be on  the order  of  0.05  g/gal, which  on  a  g/BHP-hr basis  is
somewhat less than 0.Olg/BHP-hr.[5]

     The  estimates  of  current gasoline  exhaust emissions  are
based  on  data   from  manufacturer   testing   of  certification
engines.   The  exhaust HC  and  CO  emission factors  (Table 5-2)
were  calculated  by  averaging  emissions  of  1989  heavy-duty
gasoline   engines   which   were   actually   certified   to  the
heavy-duty standards.  This means that all  engines  that  did not
meet    the    standards,    but   were   certified   by   paying
non-conformance penalties  or through use  of  the "five-percent
option,"*  were   excluded.    This  average   total   hydrocarbon
emission factor was  multiplied by 0.75 to  convert  it to NMHC.
(Heavy-Duty gasoline-fueled  engine emissions  are assumed to be
about  25  percent  methane.    There  is some  uncertainty  in this
figure, but  it represents  a  reasonable assumption,  and  using a
somewhat  different  value  would  not  significantly  impact  the
analysis.)   The  C02 emission  factor comes  from  data   on  the
base GM 454  engine from  which the CNG engine  was derived.  NOx
levels  are those needed to meet the 1991 NOx standard.

     Non-exhaust  NMHC emissions   (see  Table  5-4)   for  current
heavy-duty  gasoline engines  are   based   on   MOBILE4.    The
evaporative emission factor  came  directly from MOBILE4,  and the
running loss emission factor  is an average  of MOBILE4 estimates
of running losses at 87  and 95°F.  These temperatures were used
because they bracket the temperatures typical  of days with high
ozone  levels  in  major  cities.   The  refueling  emission  factor
comes  from MOBILE4,  assuming a brake specific fuel consumption
of 0.531 Ib/BHP-hr., as was done in Chapter 3.

     The   emissions   from   future  heavy-duty   gasoline-fueled
engines  also  need  to  be   adjusted  for  the   impacts  of  the
President's  proposed clean  air act  amendments.   The estimates
of  gasoline-fueled  vehicle  emissions  under  the  President's
proposal  are  the   current  emissions  corrected for   enhanced
     This  option allows  a manufacturer  to  certify  up  to five
     percent of  its engines to  the non-catalyst standards.  The
     same   option   would   be   available  to   CNG  engines.
     Non-conformance  penalties,  on  the other  hand,  are  not  a
     viable long term strategy for either engine type.

-------
                               5-5


                           Table 5-2

      Gasoline-Fueled Vehicle  Exhaust  Emissions  (g/BHP-hr)
                 CO
                 C02
                 NOx
                 NMHC
                 Total HC
                                   9.2*
                                   753**
                                   4.5***
                                   0.45****
                                   0.60*
                                                   for   engines
                                                 1  g/BHP-hr  HC
**
** *
** * *
Average   of   1989   certification   results
certified  to  the  14.4  g/BHP-hr  CO  and  1
standards.
Certification C02  emissions  from  GM 454  gasoline-fueled
engines.
Needed to meet 1991 standard.
75 percent of total HC.

-------
                               5-6
evaporative  emissions   controls,  lower   gasoline   volatility
limits,  and  for  the  implementation  of  Stage  II  refueling
controls in non-attainment  areas.   EPA estimates that the lower
volatility  limits  would  result  in a  48  percent  reduction  in
evaporative  emissions and  a  16  percent  reduction  in  running
loss  emissions.    In  its  analysis  supporting  its  evaporative
emissions     rulemaking,     EPA     estimated     that,     when
consideringnon-tampered     heavy-duty     vehicles-,     enhanced
evaporative controls  would result in  a  40  percent reduction in
evaporative  emissions and  an  80 percent  reduction  in  running
loss  emissions  when  using  a 9  psi  gasoline.[6]   Refueling
emissions would  be affected by  two aspects of  the  President's
program:   reduction  of  fuel  volatility  to  9  RVP  and  the
implementation of  Stage  II controls in  selected non-attainment
areas.  The exact  amount  of control to be realized by the State
II requirements  is  a  function  of the degree of  coverage  of  the
program  (station  size  exemption  level)  and  the  degree  of
enforcement exercised by the states.   Based upon the experience
of several  state programs,  EPA has  selected a station exemption
cutoff  of  10,000 gallons  per  month for these  estimates.   With
this  cutoff,  Stage II efficiency has  previously been estimated
to  lie  between  57  percent  and  79  percent.[7]   Using  the
mid-point  of  this range  (68  percent),  and  recognizing  that
Stage  II  controls  will not significantly control the  spillage
portion of  refueling  emissions,  an overall Stage  II efficiency
of 64 percent results.

     The estimated  emissions of  non-methane hydrocarbons (NMHC)
from  both  optimized  and  non-optimized  CNG  vehicles  (from
Chapter  3),  diesel  vehicles  and  gasoline-fueled  heavy-duty
vehicles are shown in Tables 5-3 and 5-4.  Table 5-3 shows that
CNG vehicle NMHC emissions  are  67-80  percent  less  than those
from   the   advanced  diesel.   While   this  is  a  significant
reduction  on a  percentage basis,  it  must be  noted that  the
absolute   reductions   are   small,   because  diesel  emissions
themselves  are only a few tenths of a  gram per  brake horsepower
hour.   It  should  also  be  re-emphasized  that  this comparison
does  not   account  for  in-use  performance  differences.   The
actual  reductions  will  likely  be  smaller  than  this  since
diesels do  not  rely  on  exhaust aftertreatment  for  significant
control  of NMHC emissions,  and thus  will probably have  less
in-use deterioration  than CNG vehicles.

     The    potential    reductions   are    much   greater   for
stoichiometric   CNG   vehicles   relative   to   gasoline-fueled
vehicles.    Table    5-4    shows    the   emissions   from   the
stoichiometric CNG engine  are 93-96 percent less  than  those of
gasoline-fueled  vehicles.   The larger  reduction  arises  because
of   the   fact  that,    while  the   advanced  lean-burn   and
stoichiometric CNG  engines  have  similar emissions, the gasoline
vehicle NMHC  emissions are roughly four  times  as  high  as those

-------
                                         5-7
                                     Table 5-3

               NMHC Emissions From Diesel and Lean-Burn CNG Engines*
1991
Diesel
Exhaust 0.40
(g/BHP-hr)
Evaporative 0
(g/MI)
1994 +
Diesel
0.29

0

Current
Lean-Burn CNG
0.09

0

Optimized
Lean-Burn CNG
0.06

0

Running Loss
  (g/MI)

Refueling
  (q/BHP-hr)
0.01**
Total (g/BHP-hr)    0.41

Corrected for
  range and
  performance

Percent Reduction
from 1994 + diesels
0.01**
             0.30
                 0.09
                              0.10
                              67%
0.06
                                     0.05***
                                     80S
*    Emission  factors  are  based  on testing  of well-maintained,  low-mileage
     test engines.  In-use emissions would be expected to be higher.
**   Unpublished EPA data.
***  The  effect  of correctio  for range  and  performance was  small  enough to
     dissappear in the  roundoff to two significant digits.

-------
                               5-8
                           Table 5-4

                      NMHC Emissions From
        Gasoline-Fueled and Stoichiometric CNG Engines*
               Gasoline
                           Gasoline
                            under
                          President1s
                           Proposal
 Current
Stoich CNG
Optimized
 Stoich
   CNG
Exhaust          0.45
  (g/BHP-hr)

Evaporative**    1.10
  (g/mi)

Running Loss**   2.04
  (g/mi)

Refueling        0.45
  (g/BHP-hr)     	

Total            3.70
  (g/BHP-hr)

Corrected for
  Range and
  Performance

Percent
  Reduction
  from Gasoline
  under
  President's
  Proposal
                              0.45
   0.09
0.34
0.34
0.18
1.24
0
0
0
0.09
                                            0.09***
  0.05


     0
                                                         0.05
                  .05***
                                            93%
                 96%
**
     Emission factors are  based  on testing of  well-maintained,
     low-mileage   test  engines.    In-use  emissions  would  be
     expected to  be higher.
     Evaporative  and  running  loss  emissions  are  reported  as
     g/mi   and converted  to  g/BHP-hr   in  the  total   by  the
     conversion factor .89.
***  The effect  of  correction for  range  and  performance  was
     small   enough  to  dissappear  in   the   roundoff   to   two
     significant  digits.

-------
                              5-9
from diesels.   As noted  above,  this  means  that  CNG  vehicles
would  be  expected  to have  a  positive impact  on urban  ozone
formation,   even  without  considering  the  possibility  of  any
reactivity benefit.

     It should  be  noted that  these  emission  factors  are  all
calculated  on  a  g/BHP-hr  basis.   Actual  on-road emissions  of
CNG  vehicles  relative  to current  vehicles  would be  slightly
higher due to an increase in  fuel tank weight needed  to provide
equivalent  range  and performance.   It has been  estimated that
to  achieve  equivalent range  CNG  vehicles would require  26.8
additional  pounds  (using wrapped  steel  tanks)  for each gallon
of  equivalent   petroleum-fueled  vehicle  tank  capacity.[8]   A
previous EPA analysis has  shown  that this weight increase would
be compounded by  a  factor of  approximately  1.3 to account  for
other  necessary  modifications  to  the vehicle  to  carry  the
additional  weight.[9]   For   a   16,000   pound   gasoline-fueled
vehicle with a  30 gallon fuel capacity, and a 37,000  diesel bus
with  a 100  gallon  fuel  capacity,  achieving  equivalent  range
with a CNG vehicle would  require  increasing vehicle  weight  by
6.5  and 9.4 percent  respectively.  For  this  analysis it  was
also   assumed   that   to   achieve   equivalent   performance,  the
horsepower  would  need to  be  increased by the same percentages.
EPA  has previously  derived sensitivity  factors  for  the percent
change  in  fuel economy  from  changes  in  vehicle  weight  and
horsepower.[9]   It  was  estimated that for  light duty vehicles,
fuel  economy decreased  by 0.329  percent and 0.454 percent for
each  percent change in weight  and horsepower  respectively.   In
the  absence of  factors  for  heavy-duty vehicles,  it  is assumed
that  these  factors  can  be  applied to  heavy-duty vehicles  as
well.   Thus, the  range  and performance  adjustment  factors used
in   this   analysis   are   1.05  for  the  stoichiometric  engine
emissions  and   1.07  for   the  lean-burn  engine   emissions.   The
relative  NMHC  emissions  of  CNG  vehicles  are  calculated  by
multiplying the ratio of  g/BHP-hr  emission factors  by 1.05  or
1.07.   These   relative   NMHC  emissions  are   summarized  in
Figure 5-1.

Ill. Air Toxics

     CNG   is  expected  to  provide  significant  benefits  with
respect to air  toxics.   EPA  has presented its  estimate of the
impact of  conventional  mobile sources  on air  toxic  in  great
detail previously.[10]   The  analysis  in this  chapter  uses the
estimates   from  that  work   of   cancer  incidences  caused  by
heavy-duty  vehicle  emissions  to provide  a perspective  on the
toxics impacts  due to heavy-duty  vehicles.   It  then develops an
estimate  of the  per-vehicle  reductions  expected to result from
CNG  use.

-------
                     5-10
                 FIGURE 5-1A
  RELATIVE DIESEL/CNG NMHC EMISSIONS
           Heavy-Duty Vehicles
          g/BHP-hr
     0.5
            1991
           Diesel
 1994     CURRENT   ADVANCED
Diesel   Lean-Burn CNQ  Lean-Burn CNQ
                 FIGURE 5-1B
RELATIVE GASOLINE/CNG NMHC EMISSIONS
             Heavy-Duty Vehicles

        g/BHP-hr
NMHC
         Current  President's Proposal  Current     Advanced
         Gasoline      Gasoline   Stolchlometrlc CNG Stolchiometrlc CNQ

-------
                              5-11
     The following mobile  source-related toxic pollutants  were
examined:    benzene  (including  exhaust,   evaporative,   running
loss,  and   refueling   benzene),   gasoline  refueling   vapors,
exhaust   1,3-butadiene,   polycyclic   organic   material   (POM)
absorbed   onto   gasoline-derived   particulate   matter,    and
formaldehyde.  These  pollutants  are emitted  by diesel  and/or
gasoline-fueled vehicles  and  are  classified by  EPA as  either
known  or   probable  human  carcinogens.    The   base  heavy-duty
vehicle cancer  cases in  the  year  2005   from  petroleum  fueled
vehicles are  shown in Table 5-5.   Included  are  cases predicted
for  the nine  cities affected by  the  light-duty  alternative
fuels  program included  as  part  of the  President's clean  air
program.[11]  These  nine  cities are expected  to be  those  most
likely  to  use alternative  fuels to  combat their  severe  ozone
problems.   This choice was  also made to  be consistent  with  the
other   EPA   reports,  on   alternative   fuels.    Since   these
projections are directly proportional to  population,  they could
easily be extrapolated to larger areas.

     Formaldehyde  in ambient   air  includes both "direct"  and
"indirect" formaldehyde.  Direct formaldehyde  is  emitted in the
exhaust of  vehicles, while  indirect formaldehyde is  formed in
the   atmosphere   from  the   reactions   of   various   reactive
hydrocarbons.   As   discussed   in  the   Final   Rulemaking  for
methanol-fueled vehicles,  indirect  formaldehyde  is  responsible
for  the  majority  of  the  formaldehyde   in  ambient  air.[12]
However, this  is  not true when dealing  with that  portion from
heavy-duty  vehicles,  since  they  emit   a higher  fraction  of
formaldehyde directly.

     The estimate of the  impact of  diesel particulate emissions
was  done  under the  assumption that all  diesels  will need trap
oxidizers   to   meet   the  stringent   post-1994   particulate
standards,  and includes  the  impacts  of  traps  failing in use.
More  recent developments,, however,  suggest that trap  use  may
not  be so pervasive; other,  more reliable  technologies  may be
used.   Thus  the  actual   total risk  from  diesel  particulate
emissions  is expected to be somewhat  lower  than this estimate.
In  addition,  since in-use deterioration  is  a  factor  in  all of
the  risk  estimates  in  Table  5-5,  the  per-vehicle reductions
developed  below cannot  properly be  applied to  these  numbers.
Comparisons  of  the two  sets of numbers can, however, be used to
provide a perspective on the potential benefits available.

     As  can be  seen from  Table 5-5,  heavy-duty vehicles  are
predicted  to contribute to about 35 annual cancer incidences in
these  nine  major  cities.    This  is approximately  one-half  the
number estimated  for  light-duty  vehicles  (69),[13]  and  when
diesel  particulate  effects   are   excluded    the   heavy-duty
contribution drops to about one-sixth (which is  expected, given
the fact  that heavy-duty  vehicles  account for  only one-fifth of
mobile source VOC emissions  in  general).  Nevertheless,  on  a
per-vehicle basis the  impact  of  CNG use  on air toxics  can be
significant.

-------
                                  5-12
                              Table  5-5

              Air Toxics From Traditional Mobile Sources
Toxic

Exhaust Benzene
Evaporative Ben
Running Loss Be
Refueling Benzene
Gasoline Refueling
 Vapors (W/0 Benze:
Exhaust 1,3-Butadi
Exhaust Gasoline P<
Direct Formaldehyd
Indirect Formaldeh
Diesel Particulate

Total
2005 Base
Heavy-Duty
Cancer Cases
In Nine Cities
e 1.47
nzene 0.20
enzene 0.05
ene 0.06
ling 0.35
enzene)
tadiene 4.29
ne POM 3.09
ehyde 0.08
Idehyde 0.66
late 24.10
35.15

Cases Due
to Diesel
Emissions
0.53
0.00
0.00
0.00
0.00
2.74
0.00
0.55
0.42
24.10
28.34
Cases Due
to Gasoline
Emissions

  0.94
  0.20
  0.05
  0 . 06
  0.35
  6.80

-------
                              5-13
     The per-vehicle air toxic reductions  for  CNG vehicles were
calculated by comparing the toxic  emissions  of CNG vehicles  to
those  of  petroleum-fueled vehicles.   This was  straightforward
for  the evaporative,  refueling,   and  running  loss  emissions,
since CNG vehicles would not  normally  have such emissions.  The
exhaust benzene  emissions  for CNG  were based  on very  limited
speciated  hydrocarbon  data   found  in  reference  1,  where  the
emissions of  benzene were estimated to  be one  percent of  the
NMHC  emissions.   (Benzene  emissions  from  gasoline  and  diesel
engines were  estimated  to be  3.5  percent and 1.1  percent  of  HC
emissions  respectively.)[10]   The  presence of  benzene in  CNG
exhaust  is  somewhat  surprising,  since  CNG  itself  does  not
contain  any  benzene,   and  it would  not  be  expected  to  form
during  combustion.   The only  remaining source of  benzene  would
be  the lubricating  oil.  Clearly,  more data are  needed in this
area.   The   reduction  of   1,3-butadiene  emissions   is   very
difficult to  estimate for CNG vehicles,  since it  is difficult
to   distinguish   1,3-butadiene   from  butane   when   speciating
hydrocarbons.  It seems unlikely  that CNG vehicles  would have
large  emissions   of   1,3-butadiene,  but  they  may  have  some.
Therefore, for this  analysis,  it  was assumed,  that CNG vehicles
would   result  in  a  near  total   (99   percent)   reduction  in
1,3-butadiene.   The   direct  formaldehyde  impact   was  estimated
from the  emission factors  listed  in Chapter 3, and  by  assuming
(as  was done  in reference 10)  that  formaldehyde accounts for
3.1  percent  of gasoline HC  emissions and 3.0 percent of  diesel
HC  emissions.   The   indirect  formaldehyde  impact was  assumed
proportional  to   NMHC reductions,  and the  diesel  particulate
reductions  were  based  on  the  particulate  emissions  factors
listed  earlier.

     Table 5-6 shows  the relative  reductions  in toxic emissions
from CNG  vehicles compared  to future petroleum-fueled vehicles,
for   both   current   technology   and    for   future   optimized
technology.   (These   estimates  were   adjusted  for  range  and
performance  in the  same  fashion  as  the  NMHC emissions.)   In
most  cases,  the CNG  vehicles  show  major  reductions in  air
toxics.   In  fact, the only toxics  in  which the  reductions are
below   70   percent   are  the   direct  formaldehyde  and  diesel
particulate  impacts   for  the  diesel/lean-burn  comparison.   The
direct   formaldehyde  emissions   may   very   well  be   lowered
significantly in the future  by  technological  advances  such as
catalysts  specifically designed  to control  formaldehyde.   The
reduction in  diesel  particulate is  below  70  percent because the
advanced  1994 diesel  engine  already  has  very  low particulate
emissions  so as  to  meet  the  stringent  particulate standards
that   take  effect   in  1994.   When   all  the  emissions  are
considered,  the  lean-burn CNG engines  are  expected to result in
approximately a  19-35 percent  reduction  in  per-vehicle  toxic
emissions  compared  to  diesel  engines.   The  stoichiometric CNG
engine would be expected  to  result  in  approximately   a  99

-------
                              5-14

                           Table 5-6a

                Per-Vehicle Air  Toxic Reductions
              of CNG Vehicles Compared to Diesels
Toxic

Exhaust Benzene
Evaporative Benzene
Running Loss Benzene
Refueling Benzene
Gasoline Refueling
  Vapors (W/0 Benzene)
Exhaust 1,3-Butadiene
Exhaust Gasoline POM
Direct Formaldehyde
Indirect Formaldehyde
Diesel Particulate

Weighted Total**
                               Percent
                              Reduction
                               Current
                              Lean-Burn
 Percent
Reduction
Optimized
Lean-Burn
                                 19
   35
**
Direct formaldehyde risk would increase,
Weighted according to Table 5-5 impacts,
                           Table 5-6b

                Per-Vehicle Air Toxic Reductions
              of CNG Vehicles Compared to Gasoline
Toxic

Exhaust Benzene
Evaporative Benzene
Running Loss Benzene
Refueling Benzene
Gasoline Refueling
  Vapors (W/0 Benzene)
Exhaust 1,3-Butadiene
Exhaust Gasoline POM
Direct Formaldehyde
Indirect Formaldehyde
Diesel Particulate

Weighted Total*
                                Percent        Percent
                               Reduction      Reduction
                                Current       Optimized
                             Stoichiometric Stoichiometric

                                   96             98
                                  100            100
                                  100            100
                                  100            100

                                  100            100
                                   99             99
                                  100            100
                                   99             99
                                   92             96
                                   NA             NA
                                   99
    99
     Weighted according to Table 5-5 impacts.

-------
                              5-15
percent  reduction  in  per-vehicle toxic  emissions compared  to
gasoline engines.*  These overall  reductions  are also  shown  in
Figure 5-2.

     One  might  expect  greater  overall  reductions  from  CNG
vehicles as compared to  diesels.   However, the  comparison here
is with  future diesels  (i.e.,  1994)  which will  have  extremely
low levels of  particulate  (see  Table 5-1) in order to  meet the
1994  diesel  particulate  standard.    Since  the  diesel  toxic
impact is  largely  dominated by particulate,  the  total  percent
reduction  in  air toxics  from  lean-burn CNG  primarily  reflects
the reduction in particulate.

     Of  course,  the   overall   impact  of   these  per-vehicle
reductions depends on  the  fraction of the heavy-duty fleet that
is eventually replaced by CNG.   Since, at  this time,  it is very
difficult  to  predict  overall  penetration,  this  analysis  is
limited  to  per  vehicle   reductions.    Also,    it   should  be
reemphasized  that  these predictions  are  based  on  low-mileage
emissions;  thus the  actual in-use  impacts  could be  somewhat
different.   It would  be reasonable  to  assume  that  since both
CNG and  gasoline-fueled vehicles  use  catalysts  as  the primary
means  of  emission  control,  the   in-use  deterioration  will  be
similar.  However, this  is not true  for diesels, which do not
rely  on  exhaust  aftertreatment  for  significant  hydrocarbon
control,  and would  be  less affected  by  in-use deterioration
than  CNG vehicles.   On the  other  hand,  diesels  are  expected.to
rely  on  aftertreatment  for   at  least  partial   control  of
particulate  emissions,  while CNG  vehicles will  not;  so in this
regard CNG would be  expected  to  have some  in-use  advantages.
Thus,  at  this time,  it is not   possible  to fully  assess the
overall  in-use impact.  It  is,  however,  possible to  say that
CNG vehicles  should  offer  significant reductions of  air  toxic
emissions from heavy-duty vehicles.

IV.   Global Warming

      Recently,  the  greenhouse  effect  (i.e.,  the  effect  of
emissions  of  certain  "greenhouse" gases,  most  notably C02/  on
global   temperatures)  has   been   receiving   a   great  deal  of
attention.   Since  combustion  of  different  fossil  fuels can
result  in different C02  emissions,  it is  appropriate  that the
analysis of the  environmental  impact  of CNG  vehicles  include
its C02  impact.  Because  CNG  has a  higher  energy  density per
carbon   atom   than  traditional  petroleum  fuels  (about   20-30
percent  more),  there is  a  potential  for  reductions  in the
     The  composite  weightings  are  based  upon  the  relative
     contributions  from  Table  5-5.    Based upon  the  earlier
     cautions  about comparing  those values  to the per-vehicle
     reductions,   these  composite   weightings  can   only  be
     considered to be approximate values.

-------
                         5-16
                   FIGURE 5-2A
   Relative Air Toxics Impacts Diesel/CNG
              Heavy-Duty Vehicles
     relative
   percentages
                   1994
                   Diesel
  Current
Lean-Burn CNG
 Advanced
Lean-Burn CNG
                   FIGURE 5-2B
  Relative Air  Toxics Impacts Gasoline/CNG
              Heavy-Duty Vehicles
        120 -r


        100-


         80 -
  relative

percentages60 _


         401


         20
               Gasoline       current       advanced
           President's Proposal Stoichiometric CNG  Stoichiometric CNG

-------
                              5-17
global  warming   impact   of  motor   vehicles.    However,   this
potential  benefit  can  be  offset  by  changes  in  the  energy
efficiency of the vehicle.

     The estimated  CC>2  emissions of  CNG vehicles used  in this
analysis are  those  described in  Chapter 3.  The  C02 emissions
of diesels  were  earlier estimated  to be 622  g/GHP-hr for 1991
vehicles,  and 574  g/BHP-hr  for  1994 vehicles  using the  same
approach used to calculate  the  NMHC emissions.  The emission
factor  (753  g/BHP-hr)  for  gasoline-fueled  vehicles  is  from
manufacturer test data. [14]

     In  addition   to  CC-2,  other   gases   can   have  a  very
significant   impact  on  global   warming as   well.    The  most
important   of these,  when  considering   the   effect  of  CNG
vehicles,  is methane.   Methane  is  much   more  effective  than
CC-2  at  absorbing  infrared  radiation;   in  fact  it has  been
estimated  that each molecule of  methane in  the atmosphere has
an  effect  equivalent  to  approximately 25  molecules  of  CC-2
(approximately a  70:1 ratio on  a weight basis). [15]   Thus any
C02 emission benefit  from CNG  vehicles  will be offset to some
extent by the increased methane  emissions,  even  if the increase
in  terms  of grams per  mile  is small.   Unfortunately though,
there  is still  much  debate about  other complicating factors.
Most  significant is  the  fact  that methane is  known  to  have  a
shorter  atmospheric  lifetime  than CC-2,  which  would  serve  to
decrease  its  impact  on   global temperatures  to some  extent.
There  is currently  no consensus on what the  decrease would be;
but  it clearly  cannot  be  ignored.   This  analysis   will  use  a
factor  from an analysis by  the University of  California which
accounted   for  both  the   absorption   and   atmospheric lifetime
effects  of methane. [16]   EPA  is not  intending  to endorse this
study,  which made   a  number  of  simplifying  assumptions  in
arriving  at  its  results.   Rather,  it  has  been chosen  as   a
conservative  approach  so  as  not to over  emphasize the still
uncertain  role of methane  in global  warming.  According  to the
University  of California  analysis,  each gram of methane emitted
can be considered equivalent to  11.6  grams  of
     Finally, the total effect of the use  of  any fuel on global
warming  also depends on  the secondary emissions,  both CC>2  and
methane,  that occur during  the  production and  distribution of
the  fuel.   The energy  consumption  at all  stages of production
and  distribution  can be  converted  to equivalent  CC-2  emissions
and  added   to  the  vehicular  emissions  of   C02  and  methane.
These   effects   were   analyzed   previously  in  a   draft   EPA
report. [9]    There   it  was   calculated   that  the  ratios  of
secondary  CC-2  and  methane  to  vehicular  CC>2  emissions   for
CNG   vehicles  using   domestic   gas   are   1:3.2   and   1:122
respectively.   A  similar   analysis  for   gasoline   showed   the
ratios  to  be 1:4.4  and 1:8720.   (The gasoline-based  ratios were

-------
                              5-18
assumed to  be valid  for  diesels as  well.)   The fact  that  the
ratios of secondary methane  emissions are so different  for  CNG
and  gasoline  is  not  surprising since  most  of  the  secondary
methane emissions for  CNG  occur  during the  distribution chain,
for which there are no comparable emissions with gasoline.

     When considering  all  of  these  factors,  as is done in Table
5-7,  the global  warming  impact  of  heavy-duty  CNG  vehicles
appears to  be  in  the  same  range of that  of  petroleum-fueled
vehicles.   Compared to diesels,  lean-burn  CNG  vehicles  would be
expected  to  result  in a  13-15  percent  increase in  the global
warming impact.  However,  compared to gasoline-fueled vehicles,
stoichiometric  CNG  vehicles   would   be  expected  to  achieve  a
19-23  percent  reduction  in  the  global  warming  impact.   This
latter benefit  is  largely  related to the improved efficiency of
the CNG engine noted in Chapter 3.

     The reader is cautioned that the two  sets  of figures shown
in  Tables 5-7a and  5-7b  are  not directly  comparable,  as  they
are  derived   from  different   heavy-duty  test   cycles.    As
discussed earlier  in this report, these  test  cycles  are based
on  the usage  patterns  of  the   engines   they  represent.   The
gasoline  test cycle  (Table 5-7b) is  based on the usage patterns
of  a  typical  gasoline  engine  whereas  the  diesel test  cycle
(Table  5-7a)  is  representative  of  the  heavier  loads  and  duty
cycles  a  diesel engine  encounters.   Earlier  discussions (see,
for  example, Table 3-7)  have shown that significantly different
emissions result from the change in test cycles.

     The  results   in  Table  5-7   show   that   vehicular   CC>2
accounts  for  about 70-80 percent of  the  global warming impact,
and  much  of the remainder  is due to  the  energy consumed during
the   production  and   distribution   of   the   fuels.    Methane
emissions play  a  relatively  small  role.   However, the  role of
methane  estimated here  is very  sensitive to  two  assumptions.
First,  it  is  obviously  dependent   on  the  assumed  per  gram
conversion  factor used to  convert   the  methane  to  equivalent
C02-   Second,  it  is  also  dependent  on  levels  of  methane
control assumed here for CNG  vehicles.   Both the lean-burn and
stoichiometric  engines  have  sufficient   catalytic  control  of
methane   to  be  able  to   meet  the  heavy-duty  engine  total
hydrocarbon standard.   This  would  not generally be  the case,
for  example, with current light-duty vehicle CNG technology. .

     It  should be noted that  this   analysis  was  done  assuming
that CNG  was  produced from current  domestic production sources
of  natural  gas.   The  conclusions   of  this  analysis  are  also
sensitive to  this  assumption.    If  the  CNG  was   imported  from
overseas, the C02 emitted during production  would be  doubled,
primarily due to  the energy  consumed during  liguifaction and
ocean  transport  of  the  natural gas.   On  the other  hand,  if
natural  gas that  is currently  being flared  or  vented  to  the
atmosphere  were  used,  then  clearly  the  CNG vehicles  would
provide a very  significant

-------
                                          5-19
                                       Table 5-7a

                        Global Warming Impact o£ Lean-Burn CNG*
Vehicle             C02    Methane

Current Diesel      622     0.02

Advanced Diesel     574     0.01

Lean-Burn CNG       575     0.81
 (Current)

Lean-Burn CNG       525     0.54
 (Optimized)
                    C02
                    From
                 Production

                    141

                    130

                    180


                    164
                        Methane
                          From
                       Production

                          0.09

                          0.07

                          4.71


                          4.30
                       Equivalent
                         C02 **

                          764

                          705

                          819


                          745
                        Range-
                       Adjusted
                         C02**

                          764

                          705

                          876


                          797
*    g/BHP-hr.
**   Total impact in equivalent CC>2 emissions,
                                         Table 5-7b
                        Global Warming Impact of Stoichiometric CNG*
Vehicle
Gasoline
Stoich. CNG
(Current)
C02
753
500

Methane
0.15
0.63

C02
From
Production
171
156

Methane
From
Production
0.09
4.10

Equivalent
C02 **
927
711

Range -
Adjusted
C02**
927
747

 Stoich.  CNG
  (Optimized)
480
0.45
150
3.93
681
715
 *     g/BHP-hr.
 **    Total  impact in equivalent C02  emissions.

 Note:       Due  to the fact  that  the lean-burn and Stoichiometric  analyses  are  based
            on different engine test cycles  the  results shown in Tables  5-7a  and 5-7b
            are  not directly comparable and  no comparisons  should be made between  the
            two.

-------
                              5-20
global warming  benefit,  since  the  emissions  from the  vehicle
would be replacing both the vented and  flared  emissions  and the
emissions  from  the  petroleum-fueled  vehicle  simultaneously.
Finally,  if natural gas  were  to be produced from  coal  (as  in a
large  scale  CNG program)  the  CC>2  emitted during  production
would more than double.

V.   Other Air Quality Impacts

     Emissions  of  CO  and  NOx  are  also  important  from  an
environmental  perspective;  both  can result in adverse  health
effects.   NOx also  plays  an important role  in the  formation of
ozone  in urban  areas  and  can  contribute to acid  rain.   Using
the CNG  emission projections from  Chapter 3 and  the petroleum
fuel numbers presented at the beginning of this chapter,  CO and
NOx  emissions of  the  various  engine  types  can  be  compared.
These data are given in Table 5-8.

     Turning  first  to  the  lean-burn engine,  Table  5-8a  shows
little variation  in CO or NOx amongst engine types,  except for
the current technology lean-burn  engine CO value.   As described
in  Chapter   3,   EPA   expects   this  value   to   be   reduced
significantly  in  future designs.   As  for the  stoichiometric
engine,  Table  5-8b  shows it  to have a  significant NOx benefit
compared  to  its  gasoline-fueled  counterpart.   The  CNG  engine
here shows  an  approximately 90  percent NOx  reduction.   Low NOx
is  a   characteristic  of  stoichiometric heavy-duty CNG  engines
with  a  three-way catalyst  as  shown in  Table 3-3.   Finally,
although  overall  CO emissions  are  similar,  stoichiometric CNG
engines may have some  CO  advantage  over gasoline engines during
cold  start as  the  gaseous  fuel  overcomes  the need  for  cold
start enrichment.

-------
                               5-21


                            Table  5-8a

                    CO and NOx Emissions of
            Heavy-Duty Diesel/Lean Burn CNG Engines*
                            CO (q/BHP-hr)          NOx (q/BHP-hr)

Current Diesel                  1.7                    4.50

Advanced Diesel                 1.4                    4.44

Lean-Burn CNG                   4.0                    4.50
  (Current)

Lean-Burn CNG                   1.5                    4.00
  (Optimized)
     Emissions  are  based  on  testing  of  well-maintained,  low
     mileage test engines.   In-use emissions would be higher.
                           Table 5-8b
               CO and NOx Emissions of Heavy-Duty
              Gasoline/Stoichiometric CNG Engines*
                            CO (g/BHP-hr)         NOx  (g/BHP-hr)

Gasoline                         9.2                   4.5

Stoichiometric CNG              10.6                   0.51
   (current)

Stoichiometric CNG               7.3                   0.41
   (Optimized)
      Emissions  are  based  on  testing  of  well-maintained,   low
      mileage test engines.  In-use emissions would be higher.

-------
                              5-22
                      References Chapter 5
     1.     "Definition  of  a  Low-Emission   Motor   Vehicle  in
Compliance with the Mandates  of  Health and Safety  Code  Section
39037.65  (Assembly Bill  234, Leonard,  1987)," California  Air
Resources Board,  May 19, 1989.

     2.     Atkinson, R. ,  "Kinetics  and  Mechanisms  of  the  Gas
Phase Reactions  of the  Hydroxy  Radical with  Organic  Compounds
under  Atmosoheric  Conditions,"   Chemical  Reviews,  1985,   85,
69-201.

     3.     "Reactivity/Volatility  Classification   of   Selected
Organic  Chemicals:   Existing Date,"  US EPA,  EPA-600/3-84-082,
August 1984.

     4.     Data supplied to EPA by Navistar,  September,  1989.

     5.     Hutchins,  P.P.,   "Gasoline,  Diesel   and   Methanol
Refueling Emissions  -  Data Collection,"  Memorandum to  Charles
L.  Gray,  Jr., Director,  Emission Control Technology  Division,
August 24, 1989.

     6.     Stout,  A.,  "Reductions in  Evaporative  Emissions and
Running  Losses  From  Enhanced  Vehicle-Based  Control,"  Draft
Memorandum to Charles L.  Gray,  Jr.,  Director,  Emission  Control
Tecnology Division, November 1989.

     7.     Draft    Regulatory   Impact   Analysis:     Proposed
Refueling  Emission  Regulations   for  Gasoline-Fueled   Motor
Vehicles-Volume  I  Analysis  of   Gasoline  Marketing  Regulatory
Strategies,"     EPA     Office     of     Air     and    Radiation,
EPA-450/3-87-001a, July 1987.

     8.     "Natural Gas Vehicles:  A Review  of the State of the
Art," Sierra Research,  NO SR89-04-01, April 13, 1989.

     9.     Sprik,  T.  L.,  "Alternative  Transportation  Fuels and
The Greenhouse Effect,  U.S.EPA, Draft Report.

     10.   "Air  Toxics  Emissions  and  Health  Risks From Mobile
Sources,"  Jonathan  M.   Adler   and   Penny  M.   Carey,   U.S.
Environmental Protection  Agency,  APCA  Paper  No., 89-34A.6, June
1989.

     11.   "Text  of the Remarks  by  the President  on  the Clean
Air  Act  Announcement,"  Office of  the  Press  Secretary,  June 12,
1989.

     12.   "Standards for Emissions  From  Methanol-Fueled Motor
Vehicles  and Motor Vehicle Engines;  Final Rule,"  54  FR 14426,
April 11, 1989.

-------
                              5-23
     13.    "Analysis of  the  Economic and  Environmental  Effects
of Methanol as  an  Automotive  Fuel,"  Special Report,  EPA Office
of Mobile Sources,  September  1989.

     14.    Data  from  General   Motors   certification  records,
October 1989.

     15.    Ramanathan,  V.,  R.  J.  Cicerone,  H.   3.   Singh,  and
j.T.Kishl,  "Trace  Gas  Trends  and  Their  Potential  Role  in
Climate  Chance,"   Journal  of  Geophysical   Research,  1985,  90,
5547-5555 .

     16.    Deluchi,  M.   A.,   R.  A.   Johnston,   D.   Sperling,
"Transportation Juels and the  Greenhouse  Effect,"  University of
California, UER-180, December 1987.

-------
                              5-24
           References For Further  Information  on CNG
     "National Air  Quality and Emissions  Trends  Report,  1987,"
EPA 450/4-89-001, March  1989,  U.  S.  EPA Office of  Air  Quality
Planning & Standards.

     Alternative  Transportation  Fuels:    An  Environmental  and
Energy  Solution,   Daniel   Sperling,   Editor,   Quorum/Greenwood
Press (Westport,  Connecticut), 1989.

     "The  Air  Quality  Benefits  of  Alternative  Fuels,"  EPA
Office  of  Mobile Sources,  prepared  for the Alternative  Fuels
Working  Group  of  the  President's  Task  Force  on  Regulatory
Relief, July 1987.

     "Guidance on Estimating Motor Vehicle  Emission Reductions
From the  Use of  Alternative Fuels  and Fuel  Blends," EPA Office
of  Mobile  Sources,  Technical  Report  No.  EPA-AA-TSS-PA-87-4,
January 29, 1988.

     "The  Emission  Characteristics  of  Methanol  and Compressed
Natural  Gas in  Light Vehicles,"  J.  A.  Alson,  APCA  paper  No.
88-993, June 1988.

     Carter,  W.P.L.,  R.  Atkinson,  "Computer Modeling Study of
Incremental Hydrocarbon  Reactivity," Environmental  Science  and
Technology, 1989, 23, 864-880.

     Moulis,  C.  E.  "Formaldehyde Emissions  from Mobile Sources
and  the Potential  Human Exposures,"  Air  and  Waste Management
Association Paper 89-34A.1, June 1989.

     "National Emission  Standards  for Hazardous Air Pollutants;
Benzene," Final Rule  and Proposed Rule,  54 FR 38044.

     "Mutagenicity    and    Carcinogenicity    Assessment    of
1,3-Butadiene,"   EPA  Office   of  Research   and  Development,
EPA/600/8-85/004F,  September 1985.

     "Diesel  Particulate  Study,"  EPA Office of  Mobile Sources,
November 1983 (Available in Public Docket  ttA-80-18.)

     Bolin,  B.,   Doos,  B., Jager,  J.,   and Warrick,  R. ,  eds. ,
SCOPE   29:    The   Greenhouse   Effect,   Climatic   Change,   and
Ecosystems, John  Wiley & Sons, New York, 1986.

     Eddy,   John   A.,    "The   Solar   Constant   and  Surface
Temperature,"    Interpretation  of   Climate   and  Photochemical
Models,  Ozone  and  Temperature  Measurements,   AIP  Conference
Proceedings  No.   82,  Ruth A.  Reck  and  John  R.  Hummel,  eds.,
American Institute  of Physics, New York, 1982,  pp. 247-262.

-------
                              5-25
       References For Further Information on CNG (cont'd)
     Ramanathan,V.,  Callis,  L.B.Jr.,  Cess,  R.O.,  Hansen,  J.E.,
Isaksen, I.S.A.,  Kuhn, W.R.,  Lacis,  A.,  Luther,  P.M.,  Mahlman,
J.D.,  Reck,   R.A.,  Schlesinger,  M.E.,  "Trace Gas  Effects  on
Climate,"   Atmospheric   Ozone   1985   World   Meteorological
Organization,  Global   Ozone  Research  and  Monitoring  Project,
Report No.  16.

     World Meteorological  Organization (WMO)   (1986)  "Report  of
the  International Conference on the  Assessment of the  Role  of
Carbon  Dioxide   and  of  Other  Greenhouse   Gases  in  Climate
Variations ' and  Associated  Impacts,"  Villach,  Austria,   9-15
October 1985, WMO No.  661.

     "Comparing the Impact  of  Different  Transportation Fuels  on
the   Greenhouse  Effect,"   prepared   for    California   Energy
Commission by Michael D.  Jackson,  Douglas D.  Lowell,  Carl  B.
Moyer, Stefan Unnasch, Acurex Corporation,  October 1987.

     "Transportation  Fuels  and the Greenhouse  Effect," Mark  A.
Deluchi,  Robert  A.   Johnson,  Daniel  Sperling,  University  of
California, Davis, October  1, 1987.

     "Summary  of  Available  Research  on  Natural Gas  Vehcile
Methane  Emissions  .Contribution  to   the   Greenhouse  Effect,"
Natural Gas Vehicle Coalition.

     "Natuarl  Gas and Climate Change:  The Greenmhouse Effect,"
American Gas Association, Issue Brief  1989-7, June 14,  1989

-------