An Examination of Interim Emission Control
     Strategies for Heavy Duty Vehicles
       (A Regulatory Support Document)
                     by
  The Emission Control Technology Division
Office of Mobile Source Air Pollution Control
United States Environmental Protection Agency
               March 30,  1976

-------
Table of Contents
I.
II.
III.








IV.








V.
Tables
1
2 '

3
4
5

6

7
Introduction 	

Current Heavy Duty Test Procedures and Potential
Improvements 	 	
A. Test Procedures 	
B. The Relationship of Current Test Procedures
to Urban Mass Emissions 	 	
C. Gasoline Test Cycle Improvements 	
D. Instrumentation Problems and Improvements 	
E. Existing Standards 	
F. Summary and Conclusions. 	 	
Technology Assessment 	 	
A. .Current Heavy Duty Emission Performance 	

C. Industry Capability at Various Emission Levels .... • •
D. Development Required to Meet Future Standards
E, Control Technology - Summary 	
F. Evaporative Emission Control 	
G. Additional Technology Assessment Information ....

Emission Control Cost Effectiveness 	

Past and Proposed Heavy Duty Emission' Standards 	
Comparison of SARR with Preliminary New York City
Truck Data 	 	 	
Comparison of LDV Operation Data from 5 Cities 	
Relationship Between SARR Emissions and FTP Emissions . .
Expected Percent Reduction in Urban Road Emissions for
Proposed Percent Reductions Measured by the Current FTP .
Percent Reductions in Gasoline Heavy Duty Emissions
Measured by the FTP and Over the Modified UDDS. ..'...
Manifold Vacuum and Corresponding Load Levels 	
1
4

7
7

8
17
18
19
22
23
23
23
29
32
32
37
38
39
45

1

9
10
14

15

16
17

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

   8      Summary Chart - Heavy Duty Emissions. .  .	      25

   9      Development Effort to Meet Possible Future H0 Emission
          Standards	      33

.  IDA     Summary Chart Heavy Duty HC Control Technology. ......      34

  10B     Summary Chart Heavy Duty CO Control Technology	      35

  IOC     Summary Chart Heavy Duty NOx Control Technology 	      36

  11      Preliminary HDV Evaporative Emission Data	.  .      37

  12      Possible Future HD Emission.Standards 	 .  .      41

  13      Cost per Engine Over 1973 Base to Meet Possible Emission
          Standards	      42

  14      Background Information on Cost Effectiveness for Various
          Control Options	      47

  15      Cost Effectiveness ($/ton) for Various Control Options.  .      48


Figures'

   1      Relationships Between Urban and FTP Emissions for Gas Trucks  12

   2      Relationships Between Urban and FTP Emissions for Diesel
          Trucks	. . .	      13

   3      Schematic Diagram of Proposed Sample Handling and
          Analysis System	.	      20

   4      Fuel Consumption Penalties Estimated by Manufacturers at
          California 5 (HC + NOx) /25 (CO) Standard	 .  .      31

   5      Combination of HC and NOx Based on 1974 Gasoline
          Certification Performance	 ....    .43

   6      1974 Certification Sales and Diesel Smoke Emissions-Peak      44


References
                       /
          References. . 	 	.;....      49

-------
                               ABSTRACT

     The Emission Control Technology Division of the Office of Mobile
Source Air Pollution Control has examined the need for, and possible
strategies for, achieving reductions in gaseous and smoke emissions
from heavy duty vehicles and their powerplants.  The relationship of
existing test procedures to urban mass emissions, the technology
available to meet reduced emission levels, the cost of various con-
trols and their effectiveness are each examined in arriving at an
optimum strategy for near term application to heavy duty .vehicles.

-------
1.   Introduction
     Exhaust emissions  from heavy  duty vehicles  (HDV) were first
regulated in California beginning  with the  1969 model year.   The
standards applied only  to gasoline engine hydrocarbon  (HC)  and carbon
monoxide (CO) emissions.  EPA adopted the California standards and
test procedures in 1970 and added  a separate  smoke  procedure and
standards for Diesel engines covering all 50  states.  In 1972 Cali-
fornia lowered the HC and CO standards.  In 1973  California revised the
gasoline test procedure to a mass  measurement basis, added a standard
for oxides of nitrogen  (NOx), established a Diesel  engine procedure,
and extended applicability of revised standards to  all  types of heavy
duty engines.  In addition, evaporative HC  standards were established
for gasoline fueled vehicles.  EPA adopted  the 1973 California package
with the 1974 model year except for the evaporative HC  regulation.
EPA also added a peak smoke level  standard  to Diesel engines.   For
model year 1975, California has lowered gaseous emission levels and
has proposed very stringent standards for 1977.   Except for minor  tech-
nical amendments to the regulations, EPA has  not  proposed any further
changes through model year 1978.   Table 1 summarizes the past and
future standards applicable to heavy duty vehicles  through 1977.	•
                              Table_
             Past and Future Heavy Duty Emission Standards
                                                    (a)
Pollutant
Hydrocarbons

Oxides of
Nitrogen
Carbon
Monoxide
Smoke (d)
Opacity
Evaporative
Ir^drocarbons
(c)
1969
C
275

NR

1.5
NR
NS
F
(b)
NR

NR

NR •
NR
NR
(c)
1970-1
C
275

NR

1.5
F
275

NR

1.5
40.20.NR
NR I NR
(c)
1972
C
180

NR

1.0
F
275

NR

1.5

NR
NR
1973

-------
     The test procedures associated with heavy duty vehicles are
engine dynamometer exercises at steady state conditions.  Unique pro-
cedures are applied to the gasoline and Diesel engines.  These exercises
are described as "9" and "13 mode" tests respectively.  The Diesel
smoke test procedure is a unique and separate transient test procedure.

     The original heavy duty engine standards represented a very small
improvement over uncontrolled levels for HG and CO.  The 1974 California
and 1975 Federal standards at best represent a 55% reduction for gaso-
line engines and almost no reduction for Diesel engine gaseous emissions.
This contrasts with the 80 +. %* reduction in light duty vehicle emissions
represented by the 1975 interim standards.  In addition, long range
agency planning documents (1) have assumed equitable treatment of all
categories of mobile sources.  Although past assumptions need not
restrict future regulatory action, continuation of mobile source con-
trol strategies, which are less stringent than LDV strategies, affects
the air quality impact basis upon which previous decisions have been
made.  These factors, coupled with the fact that current and proposed
standards for control of other mobile sources are not sufficient to
allow achievement of national ambient air quality standards, suggest
a need to examine heavy duty vehicle control strategies for the years
to come.

     The information presented and analyzed in this report is aimed at
identification of a cost effective emissipn control strategy for short
term application ot heavy duty vehicles and their power plants.  The
analysis is based on the assumption that recently proposed revisions
to the definition of Light Duty Truck will be adopted such that heavy
duty will encompass primarily vehicles over 8500 pounds gross vehicle
weight rating (GVWR)(2).  Those trucks in the range of 6000 to 8500
pounds GVWR, formerly heavy duty, will become light duty.  Thus, in
terms of 1973 sales data the heavy duty class would be approximately
40% of its former size (3).  On a vehicle miles travelled (VMT) basis
the revised light duty truck class and heavy duty class represent 60
and 40% of total truck VMT respectively (4).

     The objective of the report is to provide answers to the following
broad questions and to adequately address the issues contained therein.

     1.   What level of control is technologically feasible for heavy
          duty gasoline and Diesel engines?

     2.   What costs are associated with the various technology options
          over the useful life of the vehicles?
*NOx excepted.

(1)  Numbers in parenthesis correspond to references listed on the
     last page of the report.

-------
     3.   To what extent are reductions measured by current test pro-
          cedures real reductions in urban emissions?

     4.   Should HC and NOx as well as gas and Diesel engine designs
          be treated on a separate standards basis in the short term?

     5.   What optimum regulatory approach is appropriate for applica-
          tion to heavy duty vehicles including standards, test pro-
          cedures and time phasing?

     The report presents conclusions and recommendations immediately
following the introduction.  Following the conclusions,  section III of
the report examines the current heavy duty vehicle control strategy.
Included are detailed discussions of test procedures, their relation-
ship to on the road emissions and needed improvements.  The existing
standards and the issue of separate versus combined standards are
treated.

     Section IV explores the availability of technology and the levels
of control achievable, and the impact of proposed California standards
is described.  The fifth section presents cost effectiveness calculations
and comparisons to other mobile source strategies.                    •,  •.

     The need for additional emission control of mobile sources as well
as the environmental and economic impact of the several control strategies
are found in the Environmental and Inflationary Impact Statement (Ref.
#16).

-------
II.  Suti.:nary of Conclusions and Recommendations

     The current Federal Heavy Duty engine dynamometer test procedures
(FTP) were carefully examined in terms of their inherent technical
weaknesses and resulting ability to estimate actual emissions of in use
trucks on an urban road route.  The result of this examination was
the identification of mathematical relationships for predicting road
route emissions from FTP measurements.  In general, there was good
statistical correlation between the FTP and road route for all consti-
tuents except Diesel CO emissions.  However, the existing tests were
found to be poor "predictors" of CO and NOx emissions in that a given
reduction in terms of FTP emissions results in a much smaller reduc-
tion in actual road emissions.  This is partially explained by the lack
of certain characteristic operating modes in the current procedures as
well as improper mathematical weighting of the existing modes.  Thus,
there is clearly a need to develop procedures and operating exercises
which correlate one for one with urban emissions.  However urban
emission reductions, especially hydrocarbons, can, with the current test
procedures, be estimated with reasonable confidence as an interim step.

     Several procedural and instrumentation improvements were also
identified as necessary to make it possible to measure advanced control
system engines with reasonable accuracy.  A revised sampling and analy-
tical system using chemiluminescence and flame ionization for measure-
ment of total NOx and hydrocarbon has been developed.  In addition, the
gasoline 9-mode test reference points were redefined in terms of percent
load instead of manifold vacuum.  Equivalence ratios for adjusting the
standards to account for changes in the procedure have been developed
from preliminary data.  The effect of substituting an'FID for NDIR on
HC measurement is a multiplicative increase of 1.3 to 1.4.  1.4 was used
in determining proposed standards.  The effect of substituting chemi-
luminescence and a converter for NDIR on NOx is a multiplicative factor
of .9 to 1.0.  Equivalence was assumed in determining proposed standards.

     The issues of having unique standards for gas and Diesel and
separation of currently combined HC and NOx constituents were examined.
Both types of powerplants potentially provide the same service function
and there are no data to quantify any potential difference in stringency
between the two test procedures.  Therefore, no justification exists
for having unique standards.

     The agency has determined that oxidant control is best approached
through control of hydrocarbons, while ambient NOx levels are'the major
justification for NOx control.  Therefore, there is no long term justi-
fication for continued combination of HC and NOx as a single standard.
However, tradeoffs have"already been made between the two constituents;
yet,  for the short term:, s~bme incentive can be given to control of
HC by stating a separate HC standard and continuing with a combined total
HC 4- NOx standard.  This approach is recommended for an interim step.

-------
     The availability of emission control technology to apply to heavy
duty engines, the level of control achievable, and the expected capa-
bility of the various manufacturers to apply such techniques to pro-
duction engines were examined.  Very low levels of emissions (90% from
a 1972 baseline) were achieved in laboratory experiments using combi-
nations of oxidation and reduction catalysts and EGR on several heavy
duty engines.  However, reduction catalysts have not been used in pro-
duction automobiles or trucks.  Therefore, of the three, only oxidation
catalysts and EGR are considered viable control techniques for an
interim strategy.

     Manufacturers' data were solicited in a series of individual meet-
ings to determine their positions regarding achievable future standards.
Although several have engines capable of meeting the 5 gram HC + NOx and
25 gram CO standards proposed by California for 1977, rather severe fuel
penalties for the gasoline engine of at least 7% and 3 to 5% for Diesel
as well as restricted availability and substantial first cost increase
can be expected.  Attention was focused on levels of emissions achiev-
able with no increase in brake specific fuel consumption (BSFC).  Levels
of control achievable both with and without catalyst technology were
initially determined.  Subsequently, as a result of certification of
engines to 1975 California standards and additional information supplied
to the California Air Resources Board in hearings held February 4, 1975,
these initial determinations were revised.  It has now been determined
that levels of 1.5 HC, 25 CO, and 10 HC + NOx (gm/BHPHR) are achievable
by both gasoline and Diesel engines with no increase in BSFC by -1978,
using "good" non catalyst technology.  This includes improved fuel
management, air pumps and some utilization of EGR.  Diesel engines, many
of which already meet the proposed HC and CO standards, are expected to
utilize several different approaches including EGR, improved fuel injec-
tion and turbochargers to meet the proposed NOx level.

     The ability to reach the previously listed gaseous emission stan-
dards with no fuel penalty was based on the assumption that no further
reductions in smoke levels would be required.  However, examination of
current (1975) certification performance exposed the fact that nearly
all Diesel engine families have demonstrated the ability to meet a peak
smoke level of 35% opacity.  The standard is 50% opacity.  Therefore, in
order to prevent manufacturers from actually increasing smoke levels as
a tradeoff against reaching the proposed interim gaseous standards, a
35% peak smoke standard is recommended.

     Heavy duty gasoline fueled vehicles are not currently covered by
any evaporative hydrocarbon standard except in California where they
must meet the light duty standard qf 2 grams/test.  California specifies
the federal light duty test procedure, but certifies the system by
design evaluation rather than requiring confirmatory testing of com-
pleted vehicles.  The degree of evaporative control achievable using
current automobile test methods was examined using preliminary data from
heavy duty trucks.  It appears that although some control of evaporative

-------
losses can be achieved, if EPA were to now adopt California practices
for heavy duty vehicles, much more cost effective control would be
possible with a properly designed enclosure based test procedure.
Therefore, it is recommended that control of heavy duty vehicle evapo-
rative losses be deferred until such a test procedure and appropriate
compliance method can be determined.

     The cost effectiveness of heavy duty vehicle interim control was
calculated and compared to other mobile source control alternatives.
An increased first cost of approximately $110 per gasoline fueled vehicle
can be associated with the proposed standards.  The $120 estimated in-
crease in Diesel engine cost amounts to less than 0.1 cent per mile •-
additional operating expense because of the limited first cost increase.
Lack of a fuel penalty for either gasoline or Diesel engines, results
in heavy duty interim emission control at least as cost effective as
control of LDVs and LDTs to interim levels, and more cost effective
than other anticipated mobile source emission control strategies.

     For the reasons outlined in this summary and others detailed in
the report to follow, it is recommended that the agency adopt an interim
control strategy for application to heavy duty vehicles beginning in
model year 1979.  Standards of 1.5 HC, 25 CO and 10 HC + NOx (gm/BHP-HR)
exhaust emissions with 35% peak smoke opacity should be proposed by
NPRM as soon as possible.

-------
III.  Current Heavy Duty Vehicle Test Procedures and Potential Improve-
      ments                                                  '•

     This section of the report examines the exiting test procedures,
their origins, limitations, and potential improvements.  In addition,
the current standards and the issues of equivalent gas and Diesel and
combinatorial HC and NOx standards are examined.

     A.   Test Procedures;

     1.   Current Methods and Their Weaknesses.

     Two test procedures currently exist for examination of heavy duty
engine gaseous emissions.  The "9-mode" and "13-mode" FTP are applied to
gasoline fueled and Diesel engines respectively.  The test procedures
are developments of industry, the 9-mode from MVMA and the 13-mode from
EMA, promulgated originally by the State of California and later adopted
by EPA without substantial modification.  Additionally, a separate smoke
test procedure, developed by EPA, which simulates worst case transient
manuevers for acceleration and engine lug down is applied to Diesel
engines.

     The existence of two separate gaseous emissions test procedures is
explained by an examination of two factors.  First, the gasoline engine
was the first of the two types of engines to be controlled in 1969 with
Diesel regulation deferred until 1973.  Therefore, the gasoline proce-
dure was developed before a need to control Diesels was recognized.
Secondly, a Diesel engine cannot be run on the gasoline test procedure
because manifold vacuum, the control parameter in the "9-mode" test does
not exist on the unthrottled Diesel.  In addition, the 2000 rpm constant '
speed of the "9-mode" test, although characteristic of gasoline engines,
is closer to maximum speed for most Diesel engines.

     Both gaseous emission test procedures have several inherent character-
istics which limit their usefulness for accurate assessment of urban
emissions.  These include:

     a.   Steady state operation at each of the mode test points.
          Actual road usage involves transient operation involving
          continuously changing acceleration and deceleration rates.

     b.   Single speed (2000 prm) for the gasoline 9-mbde and dual speed
          (60 and 100% of rated engine RPM) for the Diesel 13-mode.

     c.   No consideration of cold start influences, as both tests begin
          from a "warmed up" engine condition.

-------
                                   8

     d.   Lack of a wide open throttle (full load) test point for the
          gasoline 9 mode.

     e.  - Weighting factors for the various modes have questionable
          relationships to overall urban vehicle use patterns.

     f.   Specification of gasoline 9 mode reference points in terms of
          manifold vacuum.  The use of manifold vacuum is impossible for
          supercharged engines and the level of manifold vacuum is
          affected by application of advanced emission control technology
          such as EGR.                                          >

     g.   The gas analyses systems use NDIR for all gaseous emissions
          except Diesel HC measurement.  These methods are inaccurate
          for HC and NOx and are especially unsuitable for application
          to any lower standards than those that currently exist.

     Because of these primary problems, it was necessary to evaluate to
what extent the current procedures related to urban mass emissions and
also develop technical improvements in the test procedure and instrumen-
tation system.

     B.   The Relationship of Current Test Procedures to Urban Mass
          Emissions

     The extent to which changes in heavy duty vehicle standards measured
with the current test procedures relate to changes in actual urban on-
the-road emission was evaluated.  Southwest Research Institute  (SwRI)
has performed warmed-up emissions testing of • 'heavy duty gas and Diesel
trucks over the San Antonio Road Route (SARR), an actual urban road
network. (5,6)  At the present time, EPA is analyzing extensive HDV
operational data in New York and Los Angeles with a smaller data base
from St.- Louis.  These data will be used to develop representative HDV.
chassis driving cycles (speed vs. time) and representative HDV engine
driving cycles (RPM and load vs. time).  If necessary, distinct rep-
resentative driving cycles will be determined for buses, 2 axle gas
trucks, 3 axle gas trucks, tractor trailer gas trucks, 2 axle Diesel
trucks, 3 axle Diesel trucks, and tractor trailer Diesel trucks.  These
cycles will consider hot/cold weighting factors and will be generated
using MonteCarlo techniques from a data base consisting of over 350
truck days of operational data.  Until this analysis is complete, some-
time in 1976, EPA does not have definitive,information on how repre-
sentative a given driving cycle is.  Table 2 compares summary statistics
over the SARR with preliminary truck operational data from New York
City.

-------
                                       Table 2

              Comparison of-SARR with Preliminary New York City Truck Data
 Percent Time     Average    Speed    Percent of-time at various speeds (mph) j^:.
,at Idle Range    Speed      Range    Idle  
-------

10
Table 3
Comparison of LDV Operation Data from
Operation
Data N.Y.C.
Average Speed, 40.5
Freeway, mph
Average Speed, 17.8
Non-Freeway, mph
Average Speed, 21.6
Overall, mph
% Total Time, 17.5
idle
% Total Time, 26.5
Cruise
% Total Time, 29.1
Accel
% Total Time, 27.0
Chicago Cincinnati
46.5 54.6
21.3 22.4
24.5 25.9
14.1 11.3
30.9 30.7
28.3 30.9
26.8 27.0
5 Cities
Houston
50.3
23.0
27.7
11.3
36.8
27.4
24.6
L.A.
48.7
23.2
29.3
10.1
34.3
29.8
25.9
   Decel
                         wo.necessary to develop representative  urban
.                       .
                                                  s^a
 emissions.  It is an actual urban cycle and as- such,  *  is superior to

 whLh°a^rarient CyCle;  ThS SARR ±S the onl^ urba^  road network over
 which actual mass emissions of trucks have been measured.

     SwRI has performed emission testing of the same  heavy duty gas and
 Diesel trucks over the SARH as well as by the  appropriateiederal test
 procedure   For each truck tested,  measurements of grams of pollutant/
 pound ot fuel are obtained for both the road test and the engine dy^a-
mometer test.   Regression analysis  was performed in order to relatfthe
natS°ofr^e T,^ f namometer c?cle ^  the  road emission rate.   Exami
nation of the data ladxcatea that the  relationship was linear over the
region w.iere data existed.   This assumption was supported by linear
correlation coefficients of .8 or higher.  A graph^freprLent^on of

-------
                                   11

the regressions is given in Figures 1 and 2 (pages 12 and 13), for
gas and Diesel trucks respectively.  The solid line on each graph
represents the best fit linear regression line.  The dotted lines
represent the 90% confidence interval around the mean predicted road
route emissions for a given FTP emission result.  The dashed line pre-
sents the case where the FTP emissions equal the road emissions.  Thus,
the more closely the solid line approximates the dashed line, the
better the emissions measured over the FTP represent emissions measured
during actual urban road operation.  The mean of all observations for
each pollutant and powerplant (gas or Diesel) is notated with an x on
each of the graphs.

     The SwRI samples of gasoline and Diesel trucks were selected to
approximate the national distribution of trucks with respect to engines
and GVWR.  Although it is difficult to accurately stratify such small
samples (the gasoline sample consisted of 25 trucks and the Diesel
sample consisted of 10 trucks), the SwRI samples have the same average
GVWR as the national population:  approximately 23,000 pounds GVWR
for gasoline trucks and 56,000 pounds GVWR for Diesel trucks.  Addi-
tional variability could be explained for a given size truck by adding
GVWR as an independent variable in the regression equations (thus in-
creasing the correlation and decreasing the width of the confidence
interval).  However, the emission estimates which represent the national
population of trucks can be accurately assessed without incorporating
weight as a covariate since the input sample of trucks represents the
national distribution.

     From an engineering standpoint, an emission rate of zero on the
dynamometer should result in an emission rate of zero on the road
(unless emissions only result from operation over a mode which is
present in just one of the two tests).  However, forcing a linear
regression through the zero-zero point would distort the prediction
in the -region of actual data and the region of greatest interest.
Therefore, in order to accurately fit the data in the region of low
emissions, a higher order curve fit would be required.  For the purpose
of this analysis, the linear regression not forced through zero was
considered appropriate since the data appeared linear for emission
rates as low as those being considered for the interim regulations.
The resultant linear regressions are given in Table 4.

     Table 5 (page 15) displays the expected percent reduction in urban
road emissions for proposed percent reductions measured by the Federal
test procedures.  Table entties are obtained by converting grams/bhp-hr
to grams/pound of fuel (by using the brake specific fuel consumption)
and using the regression equations from Table 4.

-------
                                     12
                                  Figure 1



 RELATIONSHIPS  BETWEEN  URBAN  AND  FTP  EMISSIONS  FOR  GAS  TRUCKS
         HC  EMISSIONS
   18
E
O)
   15
   12
IQ
i  •
         3    6    9    12   15    18   21


                FTP (gm/lb)
   21



   18



£  «

E
0>
**  12
CO
OC  9
       NOX  EMISSIONS
             6    9    12   15

               FTP  (gm/lb)
                               18   21
                                          300
                                        E
                                        a> 200
                                       CD
                                       CC
                                       3
                                          100
                                                      CO  EMISSIONS
                                                     100       200


                                                         FTP (gm/lb)
                                                                      '300
                                                            KEY
                                               BEST FIT  LINEAR REGRESSION LINE.

                                               FTP EMISSIONS : URBAN EMISSIONS LINE.

                                               90% CONFIDENCE INTERVAL AROUND
                                               MEAN URBAN VALUE FOR A GIVEN
                                               FTP VALUE.

                                               SAMPLE MEAN OF ALL OBSERVATIONS.

-------
                                    Figure  2
   RELATIONSHIPS  BETWEEN  URBAN  AND  FTP  EMISSIONS  FOR  DIESEL TRUCKS
E
o>
   7.0
   5.0
   3.0
    1.0
        HC  EMISSIONS
10       3.0       50


       FTP (gm/lb)
    25
E   20

01
CD
tr.
    15
    10
          NOX EMISSIONS
          10   20   30   40    50


                FTP (gm/lb)
                                60
                                    7.0
                                             30




                                         3   25

                                         ~

                                         E
                                         O)
                                         CO
                                         oc
                                   20




                                   15




                                   10
                                                 CO EMISSIONS
                                                       6    9   12   15


                                                          FTP (gm/lb)
                                                                         18
                                                   KEY




                                    — BEST  FIT LINEAR REGRESSION LINE.

                                    — FTP EMISSIONS : URBAN EMISSIONS LINE.

                                    ••• 90°/0 CONFIDENCE INTERVAL AROUND

                                       MEAN URBAN VALUE FOR A GIVEN

                                       FTP VALUE.


                                     x SAMPLE MEAN OF ALL OBSERVATIONS.

-------
                               14

                            Table 4

      Relationship Between SARR Emissions and FTP Emissions (a)

                          Gas Trucks

          HCSARR   " 1'33  HCFTP ~    '44»  r "  '81
          COSARR   =   '689 COFTP +81.1,   I -  .90

          NOxSARR a   .591 NOXJYP+   2.57,  r -  .91

                          Diesel Trucks

          HCSARR   =   t769 HCFTP +    -861, r »  .88

          COSARR   =2.11  COppp -   7.28,  r «  .80

          NOxSARR •»   .271 NOxFTp+   7.53,  r -  .89


 (a)  Emissions expressed in grams/pound of fuel.

     Examination  of Table 5 indicates that for gasoline trucks, de-
 creases . in the HC and NOx emissions measured over the FTP accurately
 reflect the changes in HC and NOx road emissions which would occur if
 more stringent standards were adopted over current observed levels.  For
 Diesel trucks, the decrease in HC measured over the FTP accurately re-
 flects the decrease' in HC road emissions although Fhis cannot be seen from/
 Table 5 since the current range of operation is below the proposed
 standards.  However, decreases in gasoline truck CO emissions and Diesel
 truck CO and NOx  emissions measured by the FTP do not result in equiva^
 lent reductions of on the road emissions in the region of current emis-
 sion performance.  Lack of correspondence on the gasoline CO emission
 and Diesel NOx emission reductions from 1974 levels can be seen in Table
     _                                                      r>
 5.  The lack of correlation in Diesel CO emissions cannot be seen in
 Table 5 since Diesel vehicles currently are substantially below the
 proposed standards.  Lack of Diesel CO correlation can be seen in Figure
 2.

     For gasoline trucks, these conclusions are supported in a study
 independent of those discussed above, performed by SwRI. (8)  The study
 tested eight gasoline engines over the FTP, the UDDS (with modified
 accel/decel rates where necessary) and the EPA 23 mode cycles in order
 to investigate the reductions in exhaust emission levels attainable
using various control techniques.  Evaluation over the 23 -mode cycle and
 the UDDS assessed the emissions performance of various engines and
 control approaches during operation of the engine outside the range of
 the current test procedure.   The results are shown in Table 6.

-------
                                  15

                                  Table 5

Expected'Percent Reduction in Urban Road Emissions for Proposed Percent
              Reductions Measured by the Current FTP

             Percent Reduction from 1974 HDV Standards

                  Non-Catalyst Technology        Catalyst Technology

                                  SARR            FTP            SARR
FTP
HDV
                FTP
                HDV
Gas
    HC(a)
    CO
    NOx(a)
    HC + NOx
43
25
17
25
44
 8
13
27
71
63
17
34
74
21
13
40
Diesel
    HC(b)(c)
    CO
    NOx(b)
    HC + NOx
25
39
25
26
21
 2
63
39
34
65
21
16
   Percent Reduction from Actual 1974 Certification Performance

                  Non-Catalyst1 Technology         Catalyst Technology
Gas
    HC
    C0(c)
    NOx
    HC + NOx
                   FTP
                   HDV
38

 9
18
               SARR
39

 8
21
                FTP
                HDV
69
38
 9
29
               SARR
71
 9
 8
36
Diesel
    HC(c)
    C0(c)
    NOx
    HC + NOx
20
 1
                20
                13
                9
                2
(a) Assume a 1974 standard of HC = 5.22 gms/bhp-hr and NOx = 10.78  gms/
bhp-hr based on sales weighted estimate of the 1974 certification HC/NOx
split.  Standards are adjusted for HC measured by FID.

(b) Assume a 1974 standard of HC - 1.2 gms/bhp-hr and NOx = 14.8 gms/bhp-
hr based on sales weighted estimate of the 1974 HC/NOx certification split.

(c)  A dash indicates vehicles currently are substantially below 1974
and/or proposed standards - thus comparison of percent reduction mean-
ingless.

-------
                                  16

                                  Table 6

Percent Reductions in Gasoline Heavy Duty Emissions Measured by the
                    FTP and over the Modified UDDS

                   9-Mode FTP                   UDDS

     HC               85                         72
     CO               90                         51
     NOx              54                         49
     Examination of Table 6 supports the results given in Table 5;
 that is,  the  current gasoline FTP overestimates the CO air quality
 reductions achieved with a' change in control technology.  This is pri-
 marily  caused by the lack of a wide open throttle  (full load) opera-
 ting mode which is a characteristic mode for urban driving of heavy
 duty trucks.  Under full load operation, gasoline fueled engines have
 a  fuel  enrichment feature designed to provide optimum power.  This has
 the side  effect of significantly increasing the CO emission rate.

    - The  fact that changes in NOx emissions measured by the 13 mode FTP
 do not  correspond with equivalent changes measured over the SARR for
 Diesel  engines can also be explained.  In general, Diesel trucks emit
 considerably  more oxides of nitrogen under the 13-mode test than when
 driven  over the SARR.  This follows when the type of driving represented
 by .the  urban  traffic route is compared to the intent of the 13-mode
 test.   In the 13-mode test, the idle mode is given a 20 percent weighting
 factor  and each of the remaining speed and load points an equal 8 per-
 cent weighting factor.  Thus, the 13-mode test gives the same weight to
 100 and 75 percent load as it does to zero and 25 percent load.  That
 is, the JL3-mode test was developed so as to reflect all types of HD
 Diesel  vehicle operation, both urban and inter-city.  The SARR, in
 contrast, is  not a high load factor test since the road course did
 not stress the vehicle in the way sustained high-speed, high-power
 inter-city highway driving does.  However, it better estimates the
 emission  factors which result when large Diesel trucks operate in and
 around  urban  areas.

     Test procedure variability and absolute level of emissions explain
 why CO  emissions measured by th 13-mode FTP do not correspond with equi-
 valent  changes measured over the SARR for Diesel engines.  CO emissions
 from Diesel are very low and the measurement of these emissions using
 current instrumentation is extremely variable.  Therefore, the actual
 correlation between FTP and road emissions for Diesel CO is. less than
 any of  the other five correlations while the variation in the road route
 emission  levels is considerably greater.  These two effects combine to
 prevent accurate road route predictions from FTP predictions. .

-------
                                    17

     In conclusion, the air quality benefits attributed to various
control strategies will be computed using the air quality related HDV
emissions.  Thus, standard changes for gasoline truck CO and Diesel
truck CO and NOx will not result in equally large air quality improve-
ments.,

     C.   Gasoline Test Cycle Improvements

     The problem of specifying gasoline engine load points in terms of
manifold vacuum was evaluated in a separate technical report. (9)  The
report redefines the "9-mode" test points in terms of percent of maximum
torque as opposed to manifold vacuum.  Torque points were derived from
regression analysis of 19 heavy duty gasoline engines.  Table 7 lists
the current manifold vacuum levels, derived percent torque levels, and
range observed among the 19 engines.
                             Table 7
          Manifold Vacuum and Corresponding Load Levels

fa)
Mode^ ' .
5
2,4,6,8
3
7

Manifold
Vacuum
19 (in Hg.)
16 (in Hg.)
10 (in Hg.)
3 (in Hg.)

% Observed
Torque
10
25
55 /
90
Range
% Torque
19 Engines
5-18
22-32
51-65
78-100
       (a)
          Modes 1 and 9 are idle and closed throttle respectively.
     -The  fact  that  a  substantial variation  in  torque  exists  for  a  given
 level  of  manifold vacuum suggests  that  some engines tested by  the
 modified  procedure  may have  different emission results  although  on the
 average,  emission levels monitored by the two  methods will be  the  same.
 Fortunately, gasoline engine emissions  are  highly  sensitive  to load
 variation only near no load  and full load operation.  Data gathered on
 eight  engines  were  presented in the report  (9) to  verify this  fact.  The
 expected  impact of  a  change  in test points  from manifold vacuum  to
 percent load on an.individual manufacturer  is  minor because  the  no load
 and near  full  load  emission  rates  are functions of relatively  minor
 carburetor recalibrations.   Near no load, carburetor  calibration is
 affected  by "idle"  and off idle" fuel rates which  are easily adjusted by
 changing  fuel  and air bleed  openings.   Near full load,  the adjustment of
 the "power valve" calibration is usually a  change  in  spring  tension on
 the valve slide.

-------
                                   21
     From the environmental viewpoint there is little incentive to
provide separate standards for each type of power plant since such a
strategy would likely give economic advantage to the "dirty" power-plant
(for HC and CO the gasoline fueled engine and for NOx the Diesel).
Assuming both were controlled to technologically feasible limits of the
Diesel with standards set a such a level, the gasoline engine could be
ruled out of the marketplace.  Likewise if standards based on maximum
feasible control for gasoline NOx emissions were set, the Diesel engine
may not be able to comply.  A more rational approach seems to be to
define the need for additional control and the technologically feasible
limits for both powerplants, then set a common standard representing
strict control of the "dirtiest" source and let economic advantages of
each dictate their continued use.

     The combination of NOx and HC standards into a single requirement
was first proposed by the State of California and adopted with issuance
of the 1973 California standards.  The approach was adopted by EPA the
following year.  The California Air Resources Board position stemmed
from their belief that oxidant formation could be controlled by reductions
in HC or NOx.  Most recent research suggests that reductions in hydro-
carbons are the most effective means of controlling oxidant formation.
In addition, the agency has determined that control of NOx should be
based solely on the need to meet ambient air quality N0~ standards. (12)
Thus, there is little justification for continuing the practice of
combining HC and NOx as a single standard.  Separating the two con-
stituents and setting separate standards now posts a problems, however,
in that manufacturers have been trading off control of HC for NOx and
vice-versa in order to optimize performance and economy.  Therefore,
some manufacturers have relatively high NOx and others high HC.  The
extent to which this occurs is treated in the technology assessment.

     Based on input from the National Air Data Branch of EPA (12),
control of hydrocarbon emissions is first priority since many regions in
the US have ah oxidant problem and the problem is more widespread than
previously estimated.  Taking these facts into account along with the
existing tradeoff situation of HC and NOx, there are several options to
stating future standards.

     1.   Set an HC standard and NOx standard separately.
     2.   Set an HC standard and a maximum HC + NOx standard.
     3.   Set standards for HC and NOx separately and a third combined
          HC + NOx standard at a lower total level.

Option 1 is desirable but does not satisfy the problem of past tradeoffs
and could therefore.result in wasted development and possible elimination
of engines with unbalanced levels of either constituent.  Option 2 places
importance on control of HC, allows those who have traded off NOx control
;:or HC to continue development along such lines, and peanlizes those who
have traded off HC control for NOx control.  Option 3 treats both con-
stituents with equal importance, yet would force those who have
obc.j.jneu a hi^h degree of control for either constitutent to maintain
-ha . .-.octroi..

-------
                      Figure*  3
Sp*n and     Span  and
Zero Gases   Zero  Gases
                                                                                           •Legend
                                                                                            rgk,
                                                                                               ip^y"	""'"  Backf-res sure  Rcgula
                                                                                               an           Internal  Contra!  Lo<
  itor  with
Loop Show.t
                                                                                                            Lina  Pfessufe  Regulator  with
                                                                                                            Internal  Contrc-  Loop  Shown
                                                                                               l^d	  Ball  Valve or Equivalent
                                                                                                             (Oarkend Leg Indicates  Common

                                                                                                             Flow Control or  Needle  Valv-
                                                                                               JVC         Plug Valve  or  Equivalent
                                                                                               VV^          (Cvntar  Port is Common)

                                                                                          |\\\\\\[  """"-""
                                                                                                             Unspecified  Flow Controls
                        Heavy  Duty  Eihaust Gas
                      Sampling  and  Analytical Train

-------
                                    19
     E.   Existing Standards

     Standards for the 1975 model year are 16 gm/BHP-HR HC + NOx and 40
gm/BHP-HR CO in 49 states.  California standards are 10 gm/BHP-HR HC +
NOx and 30 gm/BHP-HR CO using identical test procedures except for the
lack of a durability requirement on Diesel engines.   Additionally, smoke
standards determined by federal procedures for 50 states are 20% accel,
15% lug down, and 50% peak opacity.  California also has an evaporative
HC standard of 2 gm/test. .

     The specification of heavy duty gaseous emission standards in
units of grams/brake horsepowerhour takes into account the degree of
useful work performed by the class of vehicles into  which these engines
are built.  That is, the level of pollutants in terms of mass allowable
is directly proportional to the energy output from the engine.

     Common numerical standards have been specified  for both gas
and Diesel engines even though there is a substantial difference in
the test procedure which introduces .uncertainty about the real degree
of equivalency which exists.  No data other than those already presented
in this section are available to determine equivalency of the two pro-
cedures.  It is difficult to run gasoline engines by 13 mode procedures
or Diesel engines by 9-mode procedure because of the differences in load
parameter control and instrumentation.  The long term objective is, of
course, to have common or at least equitable procedures.  However, in the
interim it will be necessary to assume equivalence lacking definitive
data.

     Assuming equivalence in procedure there is no basis for having
separate gas and Diesel standards since both engine types provide the
same function in commercial service.  The two engine types are in com-
petition and the factors which govern the choice of  one powerplant
over another include:

     1.   First cost - The Diesel engine is approximately 3 times
          the cost of an equivalent capacity gasoline powerplant.

     2.   Operating economy - The Diesel engine enjoys approximately
          25% better fuel economy when operated in the same size truck.

     3.   Maintenance - Diesel engines typically operate 200-250,000
          miles between major overhauls.  This is twice the interval
          for gasoline engines.

     Commercial truck managers carefully evaluate the economic factors
before choosing a powerplant.  It is extremely doubtful that emission
performance plays any role in their decisions.

-------
                                      18

     D.   Instrumentation Problems and Improvements

     The sampling system and instrumentation associated with current
test procedures have several deficiencies.  These include measurement of
NO only by NDIR when standards are specified for NOx.  Additionally,
chemical driers are used such that some effect on conversion of NO- to
NO has been identified.  On the gasoline "9-mode" analysis system, HC is
measured by NDIR and then adjusted by a constant which was determined to
be the relationship between NDIR response and total hydrocarbon measure-
ment using an FID.  FID is currently specified for analysis of HC on all
other mobile source emission test procedures.  The NDIR instrument has
severe limitations in accuracy when measuring low concentration levels.
Raw concentrations of undiluted exhaust are measured at ambient conditions
which results in condensation of water as well as heavy hydrocarbons
within sample handling arid instrumentation.

     An extensive problem analysis and development program aimed at
solving the majority of these instrumentation problems was conducted.
The effort, recently completed, is documented in two technical reports.
(10) (11).

     The first of these technical reports (10) describes a completely
redesigned sample handling and analysis system.  The schematic diagram
for the system hardware is shown in Figure 3 (page 19).  The major
changes include:  substitution of chemiluminescence analyzers for NDIR
to allow measurement of total NOx at low concentration levels; the NDIR
hydrocarbon instrument is replaced by FID for gasoline engines to allow
accurate "total" hydrocarbon measurement at low concentrations; the
need for chemical driers and the problems of hydrocarbon "hang-up" and
water condensation have been eliminated by a high bypass ratio heated
sample line and ice bath incorporated between the N0_ to NO thermal
converter and the reaction chamber.

     This redesigned system is expected to significantly improve the
accuracy of heavy duty eng-ine emission measurement especially at low
concentration levels while simultaneously reducing maintenance problems.

     The second technical report (11) examines the net impact of the
proposed measurement changes on the level of emissions measured and
proposes correction factors to be applied to current standards or pro-
posed standards if based on existing sampling methods.  Data from
several sources including EPA, Southwest Research Institute, TRW Systems,
Scott Research and others were analyzed.  The relationships are shown in
equations (1) and (2) below:

     (1)  NOx (CHEM) = NO (NDIR) x (.9 to 1.0)     , ,
     (2)  HC  (FID)  = HC (NDIR) x (1.27 to 1.42)  U;
     (a)
        Includes current test procedure ratio of 1.8.

-------
                                   22
     Option 1, totally independent and separate standards, should be
  K: long tc/Tii goal and will be proposed with adoption of advanced urban
operation based test procedures.  However, taking into account the facts
that control of NOx by existing procedures results in much smaller gains
in urban emissions reductions (see Table 5) and recognizing the need to
concentrate efforts on HC control which is well predicted by current
procedures, option 2 is more appropriate for an interim control strategy.
Option 3 would overemphasize control of NOx considering current uncer-
tainties about NOx ambient air quality.

     F.   Summary & Conclusions

     There are cnrea separate test procedures applied to heavy duty
engine emissions measurement.  These are the "13 mode" and transient
smoke test procedures for Diesel engines, and the "9-mode" test pro-
cedure for gasoline engines.  These procedures have several deficiencies
which have been examined in terms of possible limitations on their use-
fulness for estimating urban emissions, and relationships have been
developed to quantify current capabilities.  In addition, numerous in-
strumentation and measurement problems have been addressed through a
redesign of the sampling and analytical systems.

     Both the 9 and 13 mode tests are good statistical predictors of
urban HC and NOx emissions.  Also, the 9-mode is a good statistical
predictor of urban CO emissions whereas the 13 mode is not.  The problem
with both procedures is that for CO and NOx there is poor correspondence,
i.e., equivalent percent reductions are not attainable.  In the short
term, with not too stringent standards, significant reductions in
absolute levels are possible, but at more stringent standard levels, no
additional control may be obtainable.

     Test procedure improvements include the substitution of percent
load for manifold vacuum as a reference point in the "9-mode" test,
substitution of chemiluminescence and FID analyzers for gasoline HC and
all NOx measurements, as well as development of a high flow rate heated
sampling scheme to eliminate condensation and delay problems.  These
modifications if adopted will sufficiently improve test procedures to
allow reasonable confidence in measuring emission rates.

     The issues of separate gas and Diesel standards and separate HC and
NOx standards were examined.  There is no reasonable basis for having a
unique standard for each powerplant design.  Neither is there continuing
justification for combining HC and NOx as a single standard other than
the fact that manufacturers may be able to achieve a higher level of
control without loss in marketability of their complete product lines if
standards are specified in terms of EC, and HC plus NOx.

-------
                                    23
IV.  Technology Assessment
     The capability of heavy duty engine manufacturers to produce engines
laving lower emission levels was examined as a separate project within
EPA.  Experimental work was.conducted under contract to Southwest Research
Institute and the Bureau of Mines on several engines to identify the
lowest levels of emissions achievable.  In addition, meetings were held
with all the major manufacturers to determine the level of technology
each has developed, expected fuel penalties associated with various
standards, and lead times for various control hardware.  Finally, in-
house data and manufacturers' data were analyzed to determine current
emission performance and fuel consumption rates.

     A separate report of the technology assessment panel (K. Hellman
and R. Wagner, et. al.) was prepared (13).  The information presented
herein is limited to the key elements of the panel's efforts.  This
information included a summary of then current emission performance,
analysis of control techniques that could be used to meet various emis-
sion levels, fuel penalties and costs.  After the preparation of the
technology panel report, further information was obtained from manu-
facturers as a result of hearings held in connection with California's
consideration of adopting several alternative heavy duty standards.  A
summary of these findings is also included.  Finally, summary and con-
clusions of the overall technology report are presented.

     A.   Current Heavy Duty Emission Performance

     The status of current Heavy Duty engines as far as emission per-
formance is concerned, falls into the same two classes as the two engine
types, Diesel engines and gasoline engines.

     Current Diesel engines exhibit gaseous emissions performance not
too much different from uncontrolled Diesel engines.  The advances that
have been made in Diesel emission control have been primarily in the
area of smoke emissions.  Substantial reductions in smoke have been made.

     Current gasoline emission levels show substantial reductions in
HC and CO emissions compared to uncontrolled gasoline engines.  NOx
levels of current gasoline engines are higher than those of uncontrolled
gasoline engines.  In general, although reductions have been made in
gasoline engine HC and CO emissions, these current values are still not
as low as those from current Diesel engines.

     T'.ie current and uncontrolled levels for Heavy Duty engines as
nijtcur-ad by the HDV FTP are shown in Table 8.  Also in Table 8 are
show.! cha future California requirements.  All percent reductions are
baiiticl or. riDV ?TP measurements and are not necessarily percent reduc-
tions in urban emissions as discussed in Section III.

     B.   Control Technology for Heavy Duty Engines

     1.   rlC Covitrol Techniques

-------
                                    24

Diesel Engines

     The HC control techniques available for Diesel engines are primarily
those that involve modification of the combustion process.  These combus-
tion process modifications are generally obtained by changing the fuel
injection timing and/or rate, improving the fuel injector design, chang-
ing the air motion in the cylinder during combustion, changing the combus-
tion chamber design, and changing the amount of air available for HC
oxidation, for example, by turbocharging.  Of these methods, the approach
considered most attractive, strictly as an HC control technique, ,is the
approach of improving the fuel injection system.  The need for more
sophisticated fuel injection systems is apparent to most manufacturers,
and some have had active programs to develop improved systems underway
for some time.  Improved fuel injection systems are an attractive route
to improved HC control, since the improvements in the combustion process
obtained may well have beneficial effects in other areas; BSFC and smoke,
for example.

     Another approach toward control of HC emissions from Diesels is the
control of crankcase emissions.  Currently no requirement exists for
Diesel crankcase control.  Control of this type of HC emission is well-
known, being the positive crankcase ventilation (PCV) valve which has
been used on passenger car engines for many years.  Although some work
will have to be done to adapt the approach to turbocharged engines,
control of crankcase emissions from Diesels is feasible.  However, there
is a need to qualify and quantify Diesel crankcase emissions before
adopting'a regulation.                                           •

Gasoline Engines

     The HC control techniques available for gasoline engines parallel
those for gasoline LDV.  Basically, these techniques include:  a) im-
proved air/fuel management, b) changes to the combustion chamber to
reduce the surface to volume ratio, c) spark timing retarded away from
MBT, d) air injection, e) thermal reactors (both rich and lean), and
f) oxidation catalysts.

     Of the above control techniques, improved air/fuel management, air
injection and oxidation catalysts are the most attractive control
techniques.  The combustion chamber modifications, the spark retard and
the thermal reactors were judged not as attractive basically because
some of them (spark retard, rich thermal reactors) may have deleterious
effects on fuel consumption.  The combustion chamber modifications and
lean thermal reactors were also considered not as attractive because,
for most manufacturers, the gasoline engines are essentially derivations
of -passenger car engines and the trend in the passenger car engine
design appears to be toward combustion chambers with more mechanical
octane which tend to have higher surface-to-volume ratios and also
toward use of catalytic converters as opposed to lean thermal reactors.

    ' 2.   CO Control Techniques

'j-.esel Engiaas

-------
.1977  California
Table 8
FID
!iC
(Cin/lWF-hi-)
CO
(f;ru/!HIP-hr)
N0:<
(f.\m/BHP-hi-)
HC-WOx -
(i;m/l!lll'-lir)
Smoke
(% Opacity)
optional standard
•^—•—Uncontrolled
?}'- Diesel Avg.
2 3 4 IS
•lit Diesel Avg. '74 Gas
|'77
Uncontrolled [|'75
Dtcsol Av,> 1 r— 	 '
srt'i > "in rAi
10 20 I 30
1 '74 Diesel Avg. ' '74 Gasoline
Summary Chart - Heavy Duty Emissions
.Uncontrolled Gasoline Avg.
]f\ k . •
10
oline Avg.
California Std
California Std
1 	 '74 Standard
*, i I i
40 50 60 70
Avg.
Uncontrolled
Gasoline Avg.
i A i i i
1 2 3 4 5 6 7
,,,„., . „ . "75 Calif orni
'77 California Std
I .1 i 1 fTl ^
1 2345
Visible '74 Lug
Limit Standard
10
'74 Accel
Standard
r-l 1 . im r-l .
•' " ' A (V» ' Bl U y o Y Y 1
1 10 H 20 1 	 1 1 30
'74 Lug ' '74 Accel 'Uncontrolled Uncontrollet
Average Average Lug Average Accel Averaj
15 20
Uncontrolled Gasoline Avg. «
i , „ • [ i i /^
80 90 100 110 120 130
1977 California Uncontrollcd
j optional standard [oiesej AV&.
e£\ I I . . fVl „ , A fell 1*1 1 ii • 1
^8 9 I 10 1 	 11 12
Uncon- .Uncon- line Avg. Avg.
trolled trolled
a Std 1 Q^esel Avg . {Gasoline Avg.
fHLr is0 20
' 	 '74 Diesel Avg. '74 Standard
'74 Peak
Standard
1 A Cl - 1
40 50 1 60
1 Uncontrolled
;e Peak Average
• i ...
140
i
13


                                                                                                                  to
                                      '74 Peak
                                      Average

-------
                                   26

     The low CO emissions from Diesel engines are primarily a function
of the lean air/fuel ratio characteristic of Diesel combustion.
Techniques that supply more air or better combustion chamber mixing, will
lower CO emissions, although there is currently no concerted effort in
the Diesel engine field to lower CO emissions, since they are already
so low.

Gasoline Engines

     Control techniques for CO from gasoline engines are much the same as
those for HC emissions, namely, improved air/fuel management (including
lean air/fuel ratio operation), air injection, thermal reactors and oxida-
tion catalysts.  As is the case with HC control techniques, the most
attractive approaches involve improved air/fuel management, air injection,
and oxidation catalysts.

     The change in gasoline engine test procedure from manifold vacuum
to percent power is expected to have greatest impact on CO control.  Manu-
facturers will have basically two options:  to set the power enrichment
to come in at power levels greater than 90 percent power, if possible,
or to design the emission control system to be able to handle the amount
of power enrichment that occurs during the test.  The first approach,
setting the power enrichment in such a way that it is not in operation
during the test, may be difficult to do since the percent increase when
power enrichment is used is typically greater than 10 percent.  Since the
first approach involves relatively less effort than the second, the panel
considers it likely that some manufacturers may choose to derate their
engines in such a manner as to not have the power enrichment occur during
the cycle.  The possibilities for "false derating", i.e., derating to get
by the test with the engine actually producing more power than the manu-
facturer claims to EPA, is an area that must be considered closely in the
certification/enforcement process.  A true derating could have beneficial
effects on gasoline truck fuel consumption.  This is the case if one com-
pares the likely fuel consumption performance of two gasoline engines
that produce the same power, one engine (smaller in displacement) operat-
ing under power enrichment and the other (larger engine) operating
without power enrichment.

     3.   NOx Control Techniques

Diesel Engines

     The techniques available for the control of NOx from Diesel
engines are injection timing changes, injection system improvements,
catering of the combustion process by combustion .chamber design, and
addition of a diluent to the air/fuel mixture.

     Examples of modifications to the combustion chamber are the
changing of the air motion ("swirl" and "squish") during combustion and
changing the physical characteristics of the chamber from the common
opr.n chamber:. direct injection type to one that is more complicated in
al'£.js, such as a divided chamber ("swirl camber" or "prechamber), a
     iiiiiiOu of the _wo ("poker head"), or a modified bowl type ("squish

-------
                                   27


     Two main diluents have been studied for use in Diesel engines.  These
are water injection and EGR.

     The use of improved injection systems and EGR are believed to be the
approaches that will yield the best results in the short term.  The use of
water injection is not considered to be practical.  Changes to different
combustion chamber configurations on a widespread basis were not con-
sidered to be achievable in the short term because of lead time and
industry attitudes toward combustion chambers.  It appears that each manu-
facturer, in general, has his own preferred type of combustion chamber de-
sign, which he is somewhat reluctant to change.  This attitude coupled
with the common wisdom that the combustion chambers with low NOx cap-
ability (for example, the prechamber) have higher fuel consumption, has
also buttressed the reluctance to change.  Even some manufacturers who
currently make indirect inject ion:-engines are considering changing to
direct injection engines in the hopes of improved fuel consumption.  How-
ever, the common wisdom concerning the direct versus indirect injection
engines may not be correct, since the comparisons are rarely made at the
same NOx emission level, possibly leading to mistaken approaches by some
manufacturers.

     The Diesel industry also seems somewhat reluctant to use the EGR
approach.  Most of the arguments against EGR are not technically
valid, but are reflections of uncertainty based on the inexperience of
the industry with actual field experience on EGR-equipped trucks.  The
major technical problem with EGR on Diesels is the proper management and
control of the amount of EGR, because a poorly designed system may result
in excessive smoke emissions.

-------
                                    28
Gasoline Engines

     Although EGR is also applicable to gasoline engines and is an
effective control technique, other methods are applicable to gasoline
engines for NOx control.  Spark retard is an effective technique, but
massive retard was ruled out as a primary NOx control technique due to
its effect on fuel consumption.  Besides EGR and spark retard, the use
of NOx catalysts is possible on gasoline engines.  Recent developments
in the area of metallic NOx catalysts show much promise for application
in the automotive area and this technology is transferable to the Heavy
Duty gasoline engine application.  However, because of the time frame of
the LDV statutory NOx standards that will require use of NOx catalysts
(1978), it is likely that use of NOx catalysts on Heavy Duty gasoline
engines will lag this 1978 date by enough to put it out of the time
frame of this study.  Therefore the use of an EGR system of the "super"
proportional type was considered to be the most attractive NOx control
technique for gasoline engines.

     4.  Smoke Control Techniques

     For Diesel engines the smoke control techniques involve approaches
that reduce smoke under two general operating conditions, steady state
modes and transient modes.  The control of steady state smoke is achieved
by reducing or eliminating any pockets of unburned fuel in the cylinder
near the end of the combustion process, by changing the fuel injection
system, and by improving air/fuel mixing.  Turbocharging can help smoke
control in steady state conditions because more air is provided to
promote complete combustion.  Additionally turbocharged engines can be
calibrated under steady state conditions to be leaner overall than
naturally aspirated engines, yet deliver more power since they are
boosted.  Transient smoke is controlled by limiting the fuel available
during a transient, for example during vehicle acceleration.  Without
control over the fuel, initially there would be too much fuel and
excessive smoke -would result.  This problem is even more severe with
turbocharged engines since the maximum fuel rate is based on the amount
of air provided by the turbocharger and at low speed the turbocharger
cannot provide all the air that it can when it is up to speed.  Accord-
ing to the industry, the technique of limiting the fuel on initial
acceleration may have two drawbacks in field use.  First, without
changing the transmission, a vehicle may not be able to climb out of the
sub-surface loading ramps common in many areas.  Secondly, drastic
limitations on the initial fuel rate may encourage tampering in the
field if the drivers feel that their vehicles are significantly down on
acceleration power compared to vehicles without the tight controls.

     Because there was no firm evidence available to indicate that
current 1974 levels of Diesel smoke are objectionable, no further re-
ductions in Diesel smoke were considered.  However, the lowering of the
standards to current levels of control to prevent manufacturers from
increasing smoke emissions through trade-offs is proposed.

-------
                                    29
     5.   Emissions and Noise

     Much has been said about the "tradeoff" between emission control
and noise control, usually via the "increased retard, more cooling,
higher fan speed, more noise" argument.  There is not necessarily any
tradeoff between emissions and noise control for heavy duty vehicles.
The noise aspects of any vehicle/engine package will have to be
optimized in the future, along with the emissions, and control tech-
niques for noise will have to be implemented.  It is possible that the
noise control system may be different for engines meeting^different
emission levels, but this does not mean that there is some tradeoff and
one has to give up emission control to get noise control or vice versa.
It just means that the overall engine/vehicle system to control both
noise and emissions may be different from those designed to control just
noise or just emissions.  The more frank members of the Diesel industry
have indicated as much.

     C.   Industry Capability at Various Emission Levels

     The report team began this study, not by considering specific
emission levels per se', but by considering the answers to the
following questions:

     1.   What emission standards will result in the least penalty
in BSFC?

     2.   What emission standards represent the maximum control of
HC emissions feasible in the timeframe considered?

     3.   What is the current industry-wide capability to meet the
1977 California Heavy Duty standards?

     4.   What emission standards will be achievable by gasoline
engines using oxidation catalysts?


     1.   Techniques for Minimum Fuel Economy Loss

     As discussed earlier, techniques applicable to- the control of HC
and NOx emissions can have the most important influence on BSFC, for
a given engine system.  At levels of HC and NOx emissions below the
current requirements, the fuel consumption change depends to a .great
extent on what control systems are used to meet the lower levels.  The
use of injection or spark retard will involve fuel consumption penalties
over current BSFC values, for both Diesel and gasoline engines.  However,
the use of more sophisticated technology than just simple retard can
reduce or eliminate these potential losses.  Since most of the data
and development effort have been expanded toward control of NOx emissions,
NOx level was used as the indicator for determining the BSFC impact.
The report team concludes that a NOx level of 9 gm/BHP-HR can be met
with no significant increase in BSFC.  This NOx level will probably require

-------
                                      30
development of more sophisticated technology than exists on current
heavy duty engines.  Although there is enough time for the develop-
ment of this improved technology, manufacturers can also meet our
9 gram/BHP-HR level by recalibrating current engines, if their
development programs are not successful, and/or they will accept
the fuel consumption penalties of approximately 4 percent that may
result if no further development work is done.

     The fuel consumption impact of any degree of emission control is a
subject,.of much interest.  The industry estimates of these potential
losses varies widely as shown in Figure 4 (page 31).  The estimates of Cater
pillar, Cummins, and General Motors reflect recent development efforts
whereas those of Chrysler, Ford, and International Harvester represent
much older development work presented to the GARB in 1973.  The gross
overestimations of fuel penalty by Chrysler, Ford, and IH as compared
to the others clearly represent a lack of development effort by these
manufacturers in the opinion of the report team.  Furthermore, the
report team concludes that the large fuel consumption "penalties"
quoted by some manufacturers are more the result of questionable data
chosen to try to influence CARB's decision on their standards than they
are estimates representative of good faith efforts to achieve the
emission levels required with minimum or no BSFC penalty.

     2.  Maximum Control of HC Emissions

     Maximum control of HC emissions primarily involves investigation
of the HC control capability of the gasoline engine, since the Diesels
are already so low.  The report team estimates that a HC level of
3.0 gm/BHP-HR is close to the lowest level that can be achieved by
most gasoline engines without catalysts, although advanced air in-
jection systems, such as modulated air injection systems, with suf-
ficient air flow capability, may be able to achieve lower levels.
If gasoline engines use catalytic control of HC, levels of 1.5 gm/BHP-
HR are possible, allowing for a conservative  (i.e. large) catalyst HC
efficiency reduction on durability.  Diesel engines also have the
capability to meet the 1.5 gm/BHP-HR level.

     3.  California 1977 Standards

     The 1977 California standards of 5 (HC+NOx) and 25  (CO) determined
by California's certification procedure are approximately equal to
1 (HO/25 (CO)/5 (NOx) on a Federal basis.  This level of control
can be demonstrated in California by 1977, but the report team
estimates that this will involve fuel consumption penalties of
3 percent for Diesel engines and 7 percent for gasoline engines,
at a minimum, compared to 1974.  In the opinion of the report team,
the Heavy Duty industry will not have the capability to produce at a
full range (comparable to the 1974 lines) of  engines nationwide that
meet the 1 (HC)/25 (CO)/5 (NOx) levels until  after 1978.  Since the

-------
                                                31
                                  Figure 4


                  Fuel  Consumption Penalties Estimated by

          Manufacturers at  California 5(HC+N*Ox)/25 (CO)  Standard
       40
o
to
o
o
0)
s,
w
00
q!
41
O
M
0)
•a
a)
09
W
       30.
       20
10
                                                                                    U
                  n
                                                           P
                   V
                   a
                   0)
                   O.
                   14
                     (U
                     C
o
in
a

I
v
*H
in
O

C
                                               O
                                               os
                                               O
                                                  
-------
                                      32


time frame  considered  in  this  report  does not  include  che  post-1978
time frame, meeting  the "California"  .".evels  of 1  (HC)/25  'CO)/5 (NOx)
was not considered feasible, because  :'.r,.  tYls tirr.e  fram-?.  '1:78)  m-ir'cet
demand may  not be met.  Consideration of e^.r'.^erio-  .lev*1."..? like the
1  (HC)/25 (CO)/5  (NOx) should  be  re-exaained by Z7A ---he-  ;he nev
heavy duty  test procedures, now under development,  are promulgated.
This additional lead time will also give the r.anufacturers £:".?/». ".o
optimize and improve their systems, and  cout>led with the practical
experience  from field  use in California, systems  superior  to those
now planned for California in  1977 could be  available  on a national
basis in the post-1978 time frame, if air quality  considerations
require their use.

     4.  Gasoline Engines with Oxidation Catalysts

     As the requirements  for HC and CO emissions are reduced, the use  of
oxidation catalysts  for Heavy  Duty gasoline  engines becomes more
necessary.  For most gasoline  engines it has been  estimated that
HC levels below the  2  to  3 level  will require  the  use  of catalysts.
CO levels below the  18 to 20 range will  also be met most easily by
use of catalysts.

     D.  Development Required  to  Meet Future Standards

     Meeting the standards of  3.0 (HC)/30 (C0)/9.0  (NOx) or 1.5 (HC)/
15 (C0)/9.0 (NOx) considered in this  report  will require development
work to avoid the occurance of fuel consumption penalties.   Generally
speaking, these development programs  are underway  now, but they
must be focused toward these potential standards.

     The programs underway to  meet the 1977  California standards for Heavy
Duty engines will provide the  basis for  most of the work necessary
to meet the standards discussed in this  report.

     Listed in Table 9 (page 33)  are  estimates of where the development
effort could be concentrated to effectively  meet the standards.

     E.  Control Technology -  Summary

     Listed in table 10 are estimates of the ranges ir* gaseous  emissions
over which different types of  control techniques could be  used.   The   .
applicable range is  shown for  each control technique.  Some control techniques
overlap others since each type of technical  approach has somewhat different
effects on different engines.  The values for  the  1974 certification results
are indicated by the square-ended bars and the estimates for various control
techniques are shown with slant-ended bars.    Although the 1974 certification
results are not adjusted  for test procedure  differences, other  estimates are.

-------
                                            33.
Pollutant
    HC
    CO
    NOX
    Smoke
                                      Table 9

                           Development Effort to Meet
                    Possible Future H-D Emission Standards
                                          Engine Type
    Gasoline H-D Engine
Expanded work in the area
of air/fuel management, air
injection systems.  Serious
development work required
on oxidation catalysts for
the 1.5 HC level.  Work on
developing positive fuel
shut off during decelera-
tions should be expanded.

Expanded work in the area
of power enrichment versus
engine size tradeoffs.
More work on air/fuel manage-
ment and air injection.
Serious development work
required on oxidation
catalysts for the 15 CO
level.

More development work needed
on optimizing EGR rate and
control and engine calibra-
tions with EGR.  Close
monitoring of NOx catalyst
development in LDV area.
    Diesel H-D Engine

Expanded effort in the area
of improved fuel injection
systems.
Only work needed is monitoring
of levels to ensure that CO
problems do not occur.
Serious effort on EGR systems
and EGR control and engine
optimization with EGR required.
More work required on direct
injection versus indirect
injection engines as a function
of NOx level.
Development effort required to
produce advanced turbochargers
that are more controllable.
Evaluations of superchargers
with superior transient re-
sponse should be expanded.

-------
                                    Table lOA
                   Summary Chart Heavy Duty lie  Control  Technology
                        Ranges  for Various Control Techniques
^^Advonced F, I. system?^

- • -• __^"^

| ftnnjjc of J.974 Cert Rcsu
v 1974 Die


^x-^'AIR + VX. catalysVv.

_^-^"~ AIK + Dll +

^^"^^


\

Hydrocarbons,
UC, CM/BHP-1IR


Iraprbved Inlcctors ^^

ts-Diesel
cl Average



Improved Fuel Meter inff^-^

AfR plus Ueccll Modulator.

— AIR Inlcction.



l '

D
I
IE
s
E
L
G
A
S
0
(DM) ^--^ L
I
(AIR) "" ' 	 ^ N
E
Range of 197A Cert Results - Gasoline |
i
^ 1974 Gasoline Average
i
i ii .iii , \ i , . . „ , i
                                                                                                                           OJ
Possible
Standard
foBslble
Sfanaara

-------
                   Table lOP
Summary Chart-Heavy Duty CO Control Technology
     Ranges for Various Control Techniques

,,,.,Xt^At?vnnccd l\l, systems .,.,., 	 ..

I RanRC of 1974 Cert Resulte-DIesel
y 1974 Diesel Average
s\

/^ "AIR + OK catalyst \

^"^ AIR + in

_^^

1

Carbon Monoxide,
CO, CM/BHP-HR






proved Metering "^X^

Air Injection, (AIR) \^

Range of 1974 Cert Results - Gasoline
V 19ZA Gasoline Av

D
I
E
S
F
L

G
S
0
L
I
N
E
1
Tage
. . . __ .... 	 1
0 5 ^0 15 210 25 30 '
                                                                                                            OJ
           Possible
           Standard
Possible
Standard

-------
                                                         Table IOC
                                      Summary Chart-Heavy Duty NOx Control Technology
                                           Ranges for Various Control Techniques
                                         Proportional EGR Systems
V0

1 1 1974 Indirect Jniectio

1 Ranee of 1



^^ NOx Catalyst "^^

^^''''Proportional EGR Systems ^""^

f Range of 1974 Cert Resul

Oxiucs of Nitrogen,
NOx, CM/BHP-HR
D
I
v {IDI)_ Enclnes
S
374 qert Results-Diesel 1 E
L
\j 1974 Dloael Averaee
A .'" 	
G
A
S
0
. L
I
:s-Gasoline | N
E
« 1974 Gasoline Avcrone

0 1 2 3 4 ' 5* 6 7 8 9 I'O 11 12 13 14 ' 15
                                                                                                                                                   ON
                                                                              Possible
                                                                              Standard

-------
                                   37

     F. .  Evaporative HC Control

     Heavy duty gasoline fueled vehicles are not currently covered by
any evaporative hydrocarbon standard, except in California where they
must meet the light"duty standard of 2 grams/test.  California
specifies the Federal light duty test procedure, but certifies the
system by design evaluation rather than requiring confirmatory testing
of completed vehicles.  EPA has conducted limited tests on heavy duty
vehicles using the enclosure, or SHED, technique.  These data, sum-
marized and compared to LDV standards in Table 11, show that HDV
gasoline fueled vehicles are a significant source of evaporative HC
emissions, even when controlled to the current California standard.
                               Table 11

               Preliminary HDV Evaporative Emission Data

                                   Total SHED Evaporative        Equivalent HC
     Vehicle Description              Loss gms/test	           gins/mi.
1.  HDV Gasoline (uncontrolled)              30                    .2.8

2.  HDV Gasoline (Calif, controls)           18                       1.5

3.  HDV Diesel (uncontrolled)                 2                       0.1

4.  LDV (1978 statutory exhaust std.)         -                       0.41

5.  LDV (1979 Proposed Evap. std.)            2                       0.2

     The experimental data obtained on a large Diesel powered tractor with
multiple fuel tanks show this uncontrolled vehicle had evaporative losses
below 2 gms/test primarily because of the low volatility of Diesel fuel
coupled with the "sealed" nature of the Diesel fuel injection system.  The
data suggest that no evaporative standard would be required for Diesel
heavy duty vehicles unless gasoline fueled vehicles were required to go
below 2 gms/test or in the event of multi-fuel engines or use of broad
distillate fuels.

     It is probably feasible to control gasoline fueled HDVs to near the
levels of a controlled LDV.  This is because of the two primary sources
of evaporative loss:  carburetor and fuel tank.  The carburetor design
and environment for an HDV is not significantly different than for an
LDV.  Further, fuel tank (diurnal) losses may be much greater as a
result of large volume and multiple fuel tanks.  However, control by
carbon canister absorption can be effected by increasing the volume of
the storage device or using multiple devices.

     The specific hardware employed in such a low emission evaporat.'-va
control system includes a carbon vapor storage canister, fuel tank vapor

-------
                                   38


liquid separator, modified carburetor with trapped venting, and miscel-
laneous hoses and tubing to connect-.the several components.  The cost
of such a control system applied to light duty vehicles is aoproxr-nately
$15 per car.  For heavy duty vehicles, EPA estimates the per vehicle
cost of an effective control system capable of meeting the 2 gm/test
standard at $25.  This higher amount accounts for the larger canister
and additional plumbing of multiple fuel tanks when applied.

     The tradeoffs associated with evaporative control includes (in
addition to the higher first cost) , a slight weight penalty estimated
at 10 Ibs., and possible increases in exhaust hydrocarbon and carbon
monoxide emissions if the storage media is purged rapidly and in a mode
of relatively poor combustion efficiency.  This latter problem can be
minimized through experimental development.

     There is no fuel penalty associated with evaporative control,
and, in fact, theoretically, there is a slight savings since the fuel
previously lost is available for combustion.  Such a benefit, if
attainable, has not been quantified.

     A significant reduction in heavy duty evaporative emissions is
expected to be possible with adoption of the current California heavy
duty evaporative standard.  The magnitude of such a reduction, 1.3
gms/mile, is small in comparison to the reduction in average HC urban
emissions resulting from the proposed gaseous emission standard (7.9
gms/mile).  At an estimated cost of $25 per vehicle, such control would
still be cost effective, but not as cost effective as the control of
gaseous emissions to proposed levels  (Ref. Table 15).

     The primary reason for the limitation in ability to obtain greater
reductions in evaporative emissions is the lack of a comprehensive test
procedure.   The adoption of an enclosure (SHED) procedure is expected
to alleviate this constraint, but requires much development before it
can be applied to heavy duty.  An equally important problem is the
method of insuring compliance with evaporative standards because test-
ing requires a total vehicle.  EPA currently certifies engines only for
heavy duty and would have to begin dealing with total vehicle manu-
facturers in any compliance program.  California has sidestepped this
problem by not requiring a demonstration of compliance.  Engineering
judgment is substituted.  The inability to insure compliance by certi-
fication testing of all possible vehicle configurations could further
erode the estimates of. potential emissions reductions obtainable if
EPA were to merely adopt the California program.

     G.   Additional Technology Assessment Information

     Several months after the!technology assessment panel completed
their report, which is summarized in sections A through E, California
held hearings to consider the impact of imposing several alternative
standards more stringent than their current standards of HC + NOx =
10 gms/BHP-HR, CO - 30 gms/BHP-HR.  Specifically, California proposed
the following options for 1977:

-------
                                   39

                         HC + NOx       NOx       CO   (gms/BHP-HR)
     (1)       —             5         --        25
     (2)       1.5            7.5       —        25
     (3)       1.5.          10.0       —        25
     (4)       1.0            —        7.5       25
An "either or" standard consisting of options (1) and (4) was sub-
sequently adopted.  However, in the course of deliberations, manufac-
turers presented written statements on their ability to comply and the
technology and fuel consumption impacts of option (3) which is within
the range of standards previously considered by the EPA panel (Ref. 13).
The significant findings from these submissions related to the develop-
ment and demonstrated performance of gasoline engines.  All manufac-
turers stated that they could meet the levels of option (3) using
current California systems or minor recalibrations of same.  The systems
in use are primarily engine modifications (fuel management), AIR, and
some EGR.  No changes in fuel consumption from current California levels
were forecast.  GM certification data for 1975 show all California
engines currently meet the option (3) levels with slightly improved BSFC
over corresponding 49 state engines certified to higher emission levels.
Very little (one engine family), if any, catalyst usage by gasoline
engine manufacturers was forecast.

     Diesel manufacturers could all meet the option (3) levels with
little change from currently certified California configurations.
However, Caterpillar said they could not meet the 1.5 HC level with one
high volume engine family, a direct injection mid-range engine used in
intermediate sized vehicles.  Cummins forecast a 5% fuel efficiency im-
provement while other manufacturers cited no change from *7.5=-Califbrnia
levels.

     In summary, the development work forecast by the technology panel
has been continuing such that the breakpoint for non-catalyst techno-
logy capability with no fuel penalty appears to have already moved to
levels of HC = 1.5, CO = 25, HC + NOx = 10 gms/BHP-HR.  The use of
"good" technology as demonstrated by :GM and Cummins and available to all,
is expected to make such levels achievable in 1979 by all manufacturers
over a full product range even after accounting for the change in HC
measurement procedure.

     H.  • Standards and Recommendations

     Based on analysis of data currently available, it appears that
heavy duty engines can meet emission standards more stringent..than
the current Federal requirements.  These standards could be applicable
for model year 1979.  If not,jthe lead time necessary to develop and
tool up will require approximately two years from the time a NPRM is
issued.

     For HC and CO emissions, the emissions capability of gasoline
engines generally is not as good as Diesel engines.  Introduction

-------
                                   40

of catalytic control of HC and CO for gasoline engines would brinj;
their emission level capability close to that of Diesel engines.  The
NOx emissions from current Diesel engines, are approximately 10 per-
cent higher than gasoline engines.

     Several sets of Heavy Duty emission standards are possible :'.n
the near term.  The range encompassed is as follows:  1.5 <_ HC < 3.0,
15 £ CO <_ 30, NOx = 9 gms/BHP-HR.  Selecting different standards with-
in these ranges primarily impacts the technology expected to be utilized
which further impacts the complexity, cost and fuel efficiency of the
engines.  The standards, their relationship to current standards, and
uncontrolled emissions are shown in Table 12.

-------
41
Table 12
Possible Future H-D Emission Standards



Uncontrolled
Levels
Current 1974
Standard
% Reduction
from Uncontrolled
1974 Average
Performance
% Reduction
from Uncontrolled
Diesel
HC CO NOx HC+NOx


1.2 6.4 10.5 11.7

- 40 16

none - none

.9 5 11.2 12

25% 21% t-6%] [-3%]
(gms/BHP-hr)

Gasoline
Smoke
(A)

27

20

26%

13

52%
Smoke
(L)

19

15

21%

7

63%
Smoke
(P)

56

50

11%

22

61%
HC CO


10.2 130

40

70%

3.4* 24

(53%) 31%
NOx HC+NOx


4.9 15.1

16*

None

9.9 13*

[-102%] (3%)
Possible Future
Standard (No
catalyst for
gasoline engine)

% Reduction
from Uncontrolled

% Reduction from
1974 Standard
3.0    30     9
none  none
       25%
20     15     50    3.0   30    9
20%    21%    11%   71%   77%  None
 0%     0%     0%
25%
Possible Future
Standard (No
catalyst for
gasoline engine) 1.5 25
% Reduction
from Uncontrolled None none - •
% Reduction from
1974 Standard - 38%


10

15%

38%


20

20%

0%


15

21%

0%


35 1.5

38% 85%

30%


25

81%

38%


10

34%

38%
Possible Future
Standard (catalyst
on gasoline engine)   1.5   15
% Reduction
from Uncontrolled

% Reduction from
1974 Standard
None  none  14%
      63%
                             20      15      35     1.5    15
26%    21%    38%   85%   88%   None
 0%     0%    30%
63%
* NDIR HC
  Numbers in (  )  are corrected for instrument difference in sensitivity between FID and NDIR.
  Numbers in [  ]  are increases in emission rates.

-------
                                   42

     No fuel consumption penalty for either set of possible standards
is expected if the manufacturers optimize for both emissions and fuel
consumption, which is possible using technology wh5.ch can be available
for all engines by 1979.  In fact, based on recent light duty vehicle
performance, it is expected that a recovery of any existing fuel penalty
will be possible if catalyst technology is applied to gasoline engines.

     The first costs associated with the two possible sets of standards
depend on the technology used to meet the standards.  Some of the tech-
nical approaches that will be used to meet the standards will involve
costs that are not directly attributable to just the standards.  With
fuel consumption a major concern, the engine system choice will require
a more sophisticated system than one that would just meet the emission
standards with no regard for fuel consumption.  Therefore, the cost
attributable to just meeting the standards are difficult to quantify.
However, the cost for the entire system, optimized for both fuel eco-
nomy and emissions, can be estimated*  These estimates are shown below.

                               Table 13

  Cost Per Engine Over 1974 Base to Meet Possible Emission Standards

                              Diesel Engine       Gasoline Engine

          HC/CO/NOx
           3/30/9                  $120                $ 75
          1.5/25/9                 $120                $110
          1.5/15/9                 $120                $165

     The standards, which are now stated on a combined basis, can be
expressed separately.  However, for reasons discussed in Section III,
a more appropriate method for the interim period is to have a separate
HC standard and a combined HC + NOx standard.  The more stringent
individual standards presented in this technology assessment stated
on the combined basis would be as follows:

          separate:      HC  =1.5 gr/BHP-HR           HC  =1.5
                         NOx =9                 HC + NOx  =10
                         CO  =15                       CO  =15

     The fact that combining the HC and NOx is not a simple addition of
the two separate standards is explained by the fact that an individual
standard results in actual performance distributions below the standard
such that, based on past experience, manufacturers will likely have a
sales weighted average HC performance of 1 gm/BHP-HR when certifying
against a 1.5 gm/BHP-HR standard.  In .addition, there is an inverse
relationship between HC and NOx emissions when using advanced control
systems, such as EGR.  Therefore, the probability of having high HC and
NOx emissions simultaneously is very low.  Data from 1974 certifica-
tion engines were analyzed t;o demonstrate this phenomenon.  Figure 5
illustrates the standards for HC and NOx which would be appropriate
to the existing 15 gm/BHP-HR HC + NOx standard.

-------
          HC                     NOx
(a)  95% of engines certified below these  levels.

               Combination of HC  and KOX  Performance
                   To A Combined  Standards Level
HC+NOx (gr/BEP-HR)
      Thus,  in terms  of  current  gasoline performance, HC  levels  of  5  and
 NOx levels  of 12  correspond with  a  combined  level  of 15,  2  grams lower
 than a simple addition  ,of  the two performance  levels.  Standards of  ap-
 proximately HC =  6 and  NOx = 13 would be  appropriate if  a 16  gram
 standard were stated separately.

      Although a reduction  in.'smoke  levels from current performance was
 not considered in the technology  assessment  and all manufacturers
 indicated their development work  on gaseous  emissions was based on no
 further reductions in smoke levels,  there is a need to reexamine the
 smoke standards.  This  is  true  because current performance  is well below
 the standards for peak  smoke and  it  would be counter productive to allow
 manufacturers to  degrade smoke  performance in  achieving  reduced gaseous
 emission levels.

      The 1974 certification peak  smoke levels  were analyzed and are  pre-
 sented in Figure  6.   These data show that 99%  of all Diesel engines  are
 below 35% peak smoke opaeiiyT-  Furthermore, a more recent examination
 of  1975  Federal and  California  engine certification data  indicated that
 the  ability  to meet  a 35%  peak  smoke level has been demonstrated in  at
 least  one test engine for  every Diesel engine  family.  All Diesel  engines
 certified in California to gaseous standards similar to .those under  con-
 sideration here currently meet  a  35% peak value.   Therefore, a  change in
 the  peak smoke standard from 50%  opacity  to  35% opacity should  be  made as
 protection against future  degradation.

     Exhaust  gaseous  emission standards of HC  = 1.5, CO = 25, HC + NOx =
 10 gm/BHP-HR are  recommended for  the 'following reasons:   1)  Non-catalyst
 technology exists to meet these standards without  a fuel penalty;  2) The
 intermediate  CO level of 25 will  allow all manufacturers  to meet the
 standards with little or no catalyst usage.  Gasoline engine first cost
 increase is,  therefore, minimized.   For Diesel engines, the total  cost
 represents less than 0.1 cent per operating mile;  3)  The lower, standards
 represent approximately the same  relative  stringency as the interim LDV
 emission standards (80% reduction from uncontrolled), and 4) The degree
 of CO control  lost (25 versus 15  gms/BHP-HR) is not as great as the
numbers indicate because the 9 mode  test over  estimates urban emission
 reductions.   Diesel engines already meet the 15 level.

-------
                                         Figure 6
1=5   -  6-3
                                1974  CERTIFICATION

                  SOLES*flND  DIESEL SMOKE  EMISSIONS  - PERK

o
«J
8^
°3'
a
w
V
PROJECTEE
J <.- . W V


•1>
v»,-


'-S •<
^
t
.

t
- . - .


'


1.2fc%>-










"'*.•'".«.
4.fl%.
.-',,. '•'-' ;'

•*s» it.








-
.4.53%


. 1_ W 1 II IV
41.51%



-.


x





J U/ i t_»^





12 . 65^

-
. •• ' • '- •" •



L_i_ *J\ *\J



22.59%




• ••



l \l_ v_liJlx^^/.JL\Jliwr 1.





U77?

"




-25     26-30   31-35    36-4
PERK  (X^^pPBCjlTJ.    :  '-;••••
        ;>;;,;;-    FREPRRED  8Y?
'•  Projected;faMly sale?; arc equally split for each certification test.
11-15    16-20   21-25     26-30   .31-35    36-40   41-45
46-50   51 5:

-------
                                    45


V.   Emission Control Cost Effectiveness

     The cost effectiveness of an emission control option depends on
several factors whose combined costs, coupled with an estimate of
emission reductions from some baseline level, yield a statistic  (in
dollars per ton of pollutant or equivalent units) which can be com-
pared to other control options.  For  the purpose of this analysis,
the baseline levels of emissions are  those which existed immediately
prior to implementation of the control option of interest.  The  costs
to control from these levels to the levels under consideration include
development costs, manufacturing costs, operating costs, and main-
tenance costs.  Manufacturing costs have been estimated from manu-
facturer supplied data and technology assessment.  Operating costs
amount to the fuel penalty associated with each control option.

     Maintenance costs are those associated with the repair and  re-
placement of components which would not have been present in pre-control
'option vehicles or were adversely affected by the operation of the
given emission control option.

          Table 14 presents background information necessary to  calculate
cost effectiveness for the previously discussed options.  The cost ef-
fectiveness of implementing LDV interim standards is included for com-
parative purposes.  No attempt has been made to determine the cost ef-
fectiveness of a VMT reduction option.  Such an option results in large
fuel consumption decreases and, from  that standpoint, is attractive.
However, implementation of such an option involves determining costs
for the development and operation of effective mass transportation
systems along with public inconvenience.  These costs are extremely
difficult to assess.

     Cost effectiveness is calculated using the following equation:

                                   Control System Cost Over Useful Life
     Cost Effectiveness ($/ton) =  Emissions Reduction  (tons/mile) x
                                   Total Lifetime Miles

where

     Control System Cost Over Useful Life = Development Cost '+
     Manufacturing Cost + Maintenance Cost + Operating Cost

     Total lifetime miles have been calculated from References 14 and 15
by computing scrappage rates and average annual miles as a function of
vehicle age.  The lifetime miles used in the cost effectiveness  calcul-
ations are:

          LDVs                100,000 miles
          LDTs                100,000 miles
          HDVs-gas            160,000 miles
          HDVs-Diesel         436,000 miles

-------
                                   46

     Using the data presented in Table. 14 as inputs to the cost
effectiveness equation, Table .\3 was prepared showing cost effective-
ness in $/ton for each control option.

     An analysis of Table 15 indicates that the initial implementation
of catalyst control options on a giv^n mobile source is extremely cost
effective since such an option does not result in fuel consunption
penalties from uncontrolled levels.  Thus, control of LDVs, LDTrs, and
HDVs to interim levels which require the application of catalysts
actually results in a cost davings, on an individual vehicle basis,
over existing control options for these categories.  In contrast, control
from one catalyst option to another does not result in any cost savings.
Considering just a cost effectiveness argument, LDT and HDV interim
control is more cost effective than LDV control in excess of the
interim levels.

     Summary

     Given that additional air quality reductions are needed, as is the
case in many AQCRs, additional controls on HDVs and LDTs are more ef-
fective than transportation control plans for LDVs and LDTs in reducing
the mobile source emission contribution, and technologically more
feasible than reducing LDV standards below the statutory levels.  In
fact, additional HDV control is the most cost effective mobile, source
control option for reducing hydrocarbon emissions over current levels
and is more cost effective than going from LDV interim to LDV statutory
standards.  Although the proposed LDT and HDV strategies will result
in additional NOx and CO reduction, these reductions, based on a
national vehicle mix, are of a lesser magnitude than those which are
obtained by going from interim to statutory control levels for LDVs.
However, implementation of LDT standards and HDV standards is more
cost effective than going from LDV interim to statutory standards.
Inspection/maintenance for LDVs and LDTs, while providing additional
HC and CO control, is not as cost effective as stricter standards for
LDTs and HDVs.

-------
                                                   Table 14
Background Information
Option
LDV/N -*-LDV/I
LDV /I-*- LDV /S
(HC, CO)
LDT/N-*»LDT/I
HDV/N-«*HDV/I
gas
Diesel
HDV/N-»-HDV/I
gas
Diesel
HDV/N-«»-HDV/I
gas
Diesel
LDV, LDT/IM
LDV, LOT EVAP
(6gms/test)
on Cost Effectiveness for Various Control Options
Operating Cost-
Baseline Emission Levels Proposed Emission Levels Fuel
Base (Expressed in Urban Equivalent Values) Consumption
Year HC CO Nox HC CO Nox Change £/
1974
1976
1975
1974
1974
1974
1974
1974
1974

1978
3.
1.
3.
11.
3.
11.
3.
11.
3.
15%
1.
4
5
2
1
8
1
8
1
8

7
grams/mile
39.0
grams/mile
15.0
grams /mile
33
grams /mile
139
25.3
grams/mile
139
25.3
grams/mile
139
25.3
reduction from
HC and
—
3.1
2.
4.2
14.4
22.0
14.4
22.0
14.4
22.0
baseline
CO
—
1.5
.41
1.7
3.2
3.8
6.9
3.8
3.2
3-8
grams /mile
15.0 2.
grams /mile
3.4 2.
grams/mile
18 2.3
grams/mile
140 13.3
25.3 19.9
grams /mile
146 13.3
25.3 19.9
grams /mile
127 13.3
25.3 19.9
levels of
.6
— —
-13.8%
0%
0%
0%
0%
0%
0%
0% b/
0%
-8%
0%
Developmert a/
+ Manufacturing
+ Maintenance
Cost
$230
$125
$ 66
$110
$120
$ 75
$120
$165
$120

$ 7.30
a/   Control system costs over the useful life for each option are split among all pollutants  for  LDV,
     LOT, and HDV gas options, assigning cost to the pollutant each component is to control.   For  the
     HDV Diesel options, all costs are associated with NOx control since no additional HC  or CO  control
     is obtained.

b/   Fuel change information is based on General Motors estimates.  These estimates, along with  ECTD
     cost estimates, compiled from discussion with the HDV manufacturers.

£_/   A negative change implies a decrease in fuel consumed.
     I - Interim Strategy

     IM - Inspection Maintenance

     N - No Further Control

     S - Statutory Requirement

-------
                                                        Table 15

                           Cost Effectiveness of Proposed Federal Emissions Control Programs
Control Program

LOT Interim
  Standard
                   Baseline Emissions
               Program 	Emissions Level  I/
               1974 LDT
               Standard

               HDV
               Current
               Control
EPA Administrators  1975 Fed.
Proposed 1980-81    LDV
Fed.Standards 2/
LDV Fed.
Statutory Stds.
HDV Fed.
Interim
Standard

LDV/LDT IM   4/

LDV/LDT EVAP
               Standards

               1975 Fed.
               LDV
               Standards
2.0/20/3.1
                                   5.5/58/6.3
1.5/15/3.1
                                   1.5/15/3.1
                                        3/
                              1.7
                         Emissions After Control
                            Program Initiated
                                                                 1.7/18/2.3
.9/9.0/2.0
                              .41/3.4/2.0
                                        3/
                              .6
1''

21
All emissions factors in grams per mile HC/CO/NOx, except for HDV.
                                                                                          Cost of Control  ($/ton)
                        HC
                                                       202
303
                        437
                Gas     20-23
                Diesel
                        61-414

                        50
          CO
                                  14
30



41



0-3

4-29
          NOx
                     53
165
                    165
                    136-247
                     45-119
                                                                                                                        00
Decision of the Administrator, Application for Suspension of 1977 Motor Vehicle Exhaust Emission Standard,
March 1975.
_3/   Heavy Duty Vehicle  (HDV) emissions levels based on 1974 vs. proposed 1979 HDV emissions standards.

     1974 Standards:     16 grams/bhp-hr  (HC + NOx), 40-grams/bhp-hr .(CO).
     1979 Proposed Standards: 1.5 grams/bhp-hr (HC), 25.0 grams/bhp-hr CO, 10.0 grams/bhp-hr (NOx).
V   TM Cost Effectiveness calculations assume a 30% failure rate.

-------
                                    49

                              References
 1.  NOx Analysis Report, Vol.  I, Office  of  Air Quality Planning and
     Standards EPA, September 21, 1973.

 2.  Position Paper on A Revised Light Duty  Truck C-J.-ni-c  by ECTD.,
     EPA, September 18, 1974.                              V

 3.  "1973 Motor Truck Facts",  Motor Vehicle Manufacturers.Association
     Publication.
                                  •.;,«;'                  .
 4.  Highway Statistics/1972, U.S. Department of Transportation,
     Federal Highway Administration.                    ?i/;

 5.  "Mass Emissions from Diesel Trucks Operated Over a Road Course",
     EPA Report. No. 460/3-74-017, August 1974.

 6.  "In^-Use Heavy Duty Gasoline Truck Emissions:   Part I, Mass
  ;  '.Emissions from Trucks Operated Over  a Road Course",  EPA Report
 -  /No. 460/3-73-002-a, February 1973.

 7.  Vehicle Operations Survey  Prepared for  Coordinating Research .
     Council and EPA by Scott Research Laboratories,  Inc., December 17,
     1971.

 8.  "Emissions Control Technology Assessment of Heavy Duty Vehicle
     Engines", EPA Report No. 460/3-74-007,  December  1973'.

 9.  Substitution of Percent Load for Manifold Vacuum in the Heavy
     Duty Gasoline Test Procedure, C. J.  France, September,30,  1974.

10.  Proposed Interim'Regulation Instrument  Changes for Heavy Duty.
     Vehicle, Gaseous Emissions  Measurement,  W.  B.  Clemmens,.
     October 24, 1974.

11.  Impact on Existing Standards Due to  Proposed Instrumentation
     Changes in the Heavy Duty  FTP, W. B.  Clemmens, -September 18, 1974.
          \                      '                                     : .
12.  Letter from Robert E. Neligan to John P. DeKany, July 2.6,  1374.
     Subject:  Heavy Duty Vehicle Emission Control Strategies.
               :    :;$.                ...'       '       ••-•  • •   ••.  • •'  '
13.  "An Assessment of the Current State  of  the Art and .Future
     Possibilities for Emission Control for  Heavy Duty Engines",
     prepared, by EGTDiy, November 18, 1974.
            ' *      *-.  .;(•"(                   :     '      _    "" '
14..  "Truck Inventory and .Use., Survey", 1972  .Qettatis of- Transportation,
     U.S. Department of Commerce, October '1973V;;'
                                . •***,        .-..:-'••',•.,.
15.  "Automobile.Exhaust Emission Surveillance - Analysis of the
     FY 72 Emission Factor Program", EPA  Report No. 460/74-001,
     February 1974.

16.  Environmental and Inflationary Impact Statement, .Interim Heavy
     Duty Engine^ Regulations :;f or 1979 and Later Model .Years, '• 1976.

-------