SDSB 79-04
                           Technical  Report
                           January, 1979
           Light-Duty Diesel  Gaseous  Emissions Measurement
             Comparison of Dilution Tunnel  Test Results  to
                    Certification  Cell  Test Results
                                  by
                              Jeff  Alson
                                NOTICE
Technical Reports do not necessarily represent final EPA decisions
or positions.  They  are  intended  to present technical analysis of
issues using data  which  are currently available.   The purpose in
the  release of  such reports  is  to  facilitate  the exchange of
technical information and to inform the public of technical devel-
opments which may form the  basis  for a  final EPA decision, position
or regulatory action.
               Standards  Development  and  Support Branch
                 Emission Control  Technology  Division
             Office of Mobile  Source  Air  Pollution Control
                  Office  of Air, Noise  and  Radiation
                 U.S.  Environmental Protection Agency

-------
Ab s t r ac t

     This  report summarizes gaseous emissions  data for  eleven
light-duty  diesel  vehicles using both  the  standard certification
test procedure and the dilution tunnel test procedure which incor-
porates  particulate  measurement.    The' greatest  variability  was
found  in the HC measurements.  This  is partially attributed to the
variability  inherent  in measuring  HC and,  in  the extreme  cases,
also to equipment inconsistencies  between  the test  cells.  CO, NOx,
and  CC>2  data  from  the dilution  tunnel  test  procedure  were  gen-
erally in good agreement with .the certification data with the. only
trend  being  slightly  lower NOx  and C02 values  from  the dilution
tunnel.  This is hypothesized as  a possible dynamometer effect.   It
is concluded that there are  no  significant  differences  in the  two
test procedures  with  regards to  the  measurement  of gaseous emis-
sions.

Introduct ion

     During the  past  six  months  eleven light-duty diesel vehicles
have been tested for gaseous emissions using the light-duty diesel
test procedure which  has  been developed to facilitate particulate
measurement.  This testing was conducted as the final phase of the
light-duty diesel particulate  baseline study.   Emissions  test
results  have  also  been  obtained  for  these vehicles using  the
current  light-duty diesel  certification  test  procedure.   A direct
comparison can be made between the  proposed test procedure and  the
present  certification procedure  regarding  the  measurement  of
gaseous  pollutants.   Obviously the viability  of the  new  test
procedure will rest heavily on its ability to closely duplicate the
results  of  the  present  certification system.   This  report  will
assess the ability of  the  new test procedure to do  this.

Test Procedure

     The certification emissions  test  results were  obtained accord-
ing to the procedure outlined  in  the Federal Register and the tests
were performed in light-duty diesel certification test  cells 5  and
6.  The new test procedure was described in the "Draft Test Proce-
dure  for Measurement   of  Gaseous  and Particulate  Emissions  from
Light-Duty Diesel Vehicles,"  September,  1978,  and was implemented
in  test  cell  508.   With  regards to gaseous pollutants  the  only
major difference between the procedures is that  the new one incor-
porates  a  dilution  tunnel rather than  a large baffle box  as  the
mixing medium.   Henceforth,  the new  test procedure will be referred
to as the dilution tunnel  system.

Results

     The results  of  the  dilution tunnel-certification  comparison

-------
                                 -2-
are given  in  Tables  I and II and in Graphs I through IV.  Table  I
gives  the actual  HC, CO,  NOx, CC^,  and  fuel  economy  data for
the eleven vehicles tested by both  the  dilution tunnel and  certifi-
cation  test  procedures.    Both  Federal Test Procedure  (FTP) and
Highway Fuel  Economy  (HWFE)  tests  we.re  conducted.   Table II  pre-
sents the same data with  the difference between the  dilution  tunnel
and certification  values  expressed  as  a percentage of the latter.
It also  gives  recent  Repca (correlation vehicle) values which are
indications of the variability present  among the  various  test  cells
at EPA at. this time.   It  should  be  noted,  however,  that  Repca is  a
gasoline-fueled vehic.le which limits its applicability with regards
to diesel test cells.

     Graphs I, II,  III, and IV give  a Cartesian presentation  of the
same data for each pollutant.   The X's  represent  data  from  the HWFE
tests and  the  solid  points represent  FTP  data.   Each graph  shows
the "ideal"  fit  which would result from perfect duplicity of the
data  from the  two test  procedures,  as well as the "best" fit
(least-squares regression)  lines for  both  the FTP  and  HWFE  data.
In addition,  Graph I  includes the "best" fit lines for HC excluding
the one pair  of  data  from the Peugeot   504D which greatly  deviates
from the rest of the  HC data.

Analysis of Results

     As  is  often  the case  with diesel- vehicles, and which is
indicated in Table II, HC was the  pollutant most sensitive  to the
change in  test  cells.   The  trend was  most pronounced in  the  HWFE
data with  nine  of  the eleven vehicles registering relative  dif-
ferences of  8.6  percent  or  more but  was  also evident  in the FTP
data where seven vehicles  experienced differences of  7.3  percent or
more.

     There are several factors which help  to explain or  ameliorate
this apparent  problem, however.   The  first  is   the  fairly  large
Repca value of 14  percent which  indicates  the inherent variability
problems with  HC measurements.  Admittedly,  Repca  is a gasoline-
fueled vehicle  but the same  variability problems with  regards to
diesel HC and the heated  flame ionization  detector  (HFID)  are  well
known.   Assuming the  Repca number to be some measure  of statistical
acceptability, we  find that  just three of the relative FTP  values
and four of the relative  HWFE values exceed this  figure.

     A second factor  to be considered is the general  scatter  of the
FTP data about the. ideal  fit  line as shown  in Graph  I.   Except for
the Peugeot,  Mercedes,  and  International  Scout vehicles  (the
highest  and  two lowest  FTP values),   the  FTP data  all  lies  very
close to the  ideal fit line with no trend  indicated.  This  latter
fact indicates that the effect, if any,  of  the dilution tunnel  upon
the majority  of HC measurements is.a random one.

-------
                                 -3-
     As mentioned  above and  indicated in Table  II,  the  Peugeot,
Mercedes,  and International Scout vehicles  experienced the  greatest
relative HC  differences.   The  latter two were the lowest  HC emit-
ters with concentrations in the  0.3  gram per nile  (gpm) range over
the  FTP.   The high  relative  values  for  these two vehicles are
clearly due  in  part to  their  small absolute  magnitudes.   The
Mercedes,  for  instance,  experienced  a 0.05 gpm difference between
the two test procedures.  This  doesn't appear  to be a problem until
it is  realized that  it  represents a  16.7  percent relative change.
Thus the small magnitudes of the  values contribute to the  apparent
correlation problem.

     Still,  the absolute differences  of  0.05  and  0.09  gpm for the
Mercedes and Scout vehicles have another  possible  explanation.  The
HFID HC analyzers in the certification test cells utilize  only one
range  (0 to 100 ppm HC)  for all HC measurements.   The HFID  analyzer
in the dilution tunnel  test cell utilizes three different ranges  (0
to 100, 0  to 50,  0 to 25 ppm HC), the lowest of which allows much
greater sensitivity  as  compared to the certification range.   Thus
when testing  cars  with  fairly  high  HC emissions  the  same ranges
were used  for both  the dilution  tunnel  and  certification tests.
But when testing low emitters  like the Mercedes and Scout  vehicles
the  lowest,  most  sensitive range was  used  in the dilution tunnel
test cell while the  higher range was still used  in the certifica-
tion test cells.   To further compound the problem  all background  HC
readings in  the certification  test  cells were taken  at the higher
range while all such readings  in the  dilution  tunnel test cell were
taken  from the most  sensitive scale.   Thus this factor may explain
why  the Mercedes  and Scout vehicles  had greater  absolute  HC dif-
ferences than most vehicles which emitted  far more HC.  This could
also account for the wider  differences in  the HWFE data base since
most of it is clustered around  the 0.2 to  0.3 gpm range, where the
dilution  tunnel analyzer  was  set as its most  sensitive scale.

     The most baffling  HC  data  are  those  of  the Peugeot  vehicle.
The dilution tunnel  measurements  were 0.22 and  0.47  gpm less than
the  certification measurements  for  the FTP and HWFE  tests and  20
and  69  percent less  on  a relative basis.  These data have such  an
effect   on  the least-squares regression  technique  (since  they are
the highest values)  that Graph  I  includes  regression lines exclud-
ing the Peugeot data.

     One possible  explanation has been suggested.  Peugeot  vehicles
are certified using  a special  high  volume  cooling fan which has a
much larger  capacity than  the   standard certification cooling fan.
Thus the  Peugeot  certification  results were  taken using  the high
volume  fan.   When  tested in the dilution  tunnel test cell,  however,
the  standard  cooling fan was  used.   The  effects, if  any, of the
special cooling fan  have not been quantified.   One possibility  is
that the  higher  volume   fan results   in a  decrease of  the coolant

-------
                                 -4-
 temperature  which  accordingly  decreases  the  combustion  chamber
 surface  temperatures.    Lower  combustion chamber  surface  tempera-
 tures  are known  to  increase HC and  CO  emissions due  to  a  larger
 quenching zone and the resultant increase in incomplete combustion.
 NOx,  on  the other hand,  would  be  expected to be  lower  due  to the
 greater  heat  losses  to the cylinder walls  resulting  in lower  peak
 temperatures in the cylinders.  The data  in Tables I and II support
 this  theory completely  as the  certification  cell using  the  high
 volume  fan  recorded higher  HC and CO  values and  lower NOx emis-
 sions.   As  further support of  this theory  it  should  be noted  that
 the Peugeot was the only vehicle to have  lower FTP NOx emissions in
 the certification test cell.  Thus it is  likely that much of the HC
 differences for the Peugeot  can be attributed  to  the  difference in
 cooling  fans.

     As  Tables I  and  II  and  Graphs II,  III, and IV indicate, there
 do not seem to be any significant  problems  in sampling CO, NOx, or
 C02 with the  dilution tunnel  test procedure.   The CO values  were
 in very  good agreement,  especially in view of  the 11% Repca number
 indicating  the  variability among  EPA test cells  in  measuring CO.
 As Graph II indicates,  the  FTP data  was well  scattered  about the
 ideal fit line showing no  trend whatsoever.  The HWFE  data from the
 dilution tunnel  did  prove to  be somewhat  lower,   but  this doesn't
 seem to  be a serious concern at this time.

     The NOx relative  values  were  generally within the Repca value
 of nine percent  as  well.   The dilution tunnel  FTP values  were
 consistently  lower  than  the  certification  values  (except  for the
 Peugeot  discussed earlier) while the HWFE values were  more scatter-
 ed.   The C02  relative values  were  rarely much  greater  than  five
 percent  although  they were often  greater than the Repca  value of
 three percent.  As with  NOx,  there was  a clear indication that the
 dilution tunnel  test   cell measured  slightly  lower C02 emissions.
 This  combination  of  lower NOx  and COj  emissions  points to  a  pos-
sible  dynamometer effect since  these  two pollutants are  most
 affected by dynamometer-road load considerations.   As  noted earlier
 there were  no  clear trends  with  regards to higher or  lower HC or
 CO measurements from  the  dilution  tunnel.   This  latter fact, along
 with  the lack of constant  differences   in  the NOx and C02 data,
 makes it unlikely that a CVS problem was a  factor.

 Conclusion

     Based on the above  results and discussion and the realization
 of the  variability inherent  in emissions testing,  it  is  concluded
 that there are no significant differences between  gaseous emissions
 measurements made with the dilution tunnel test procedure and those
 made using  the current light-duty  diesel certification test  proce-
 dure.

-------
                                                                     TABLE I

                            LIGHT-DUTY DIESEL COMPARISON OF DILUTION TUNNEL TEST  RKSULTS TO  CERTIFICATION  CELL  TEST  RESULTS
Vehicle & ID


Oldsmobile 350
93JF127
'79 Cert Vehicle
Oldsmobile 260
93HFI.49
'79 Cert Vehicle
Oldsmobilc 260
93HF-84482F
'79 F E Vehicle
Ohlsmobile 260
93HFI81
'79 Cert Vehicle
(5 Epd)
Chcv. Pickup 350
8TJ9-168K
' 79 C;:rt Vehicle
(3-hole injection)
Dutige Truck
B356
'79 Cert Vehicle
(4 spd)
Mercedes 300D
123.130-12017745
Special Uuild/Turbo-
Ch.irged with ECR
Volksv.igt'n Dasher
406 2 :4t>6
'79 Ccrl Vehi.-le
(4 spd)
Volkswagen Rabbit
iOb Z 2465
'79 Cert Vehicle
(5 spd)
Peugeot 504D***
622
' 79 Cert Vehicle
(4 spd)
Int'l Harvest Scout
300
' 79 Cert Vehicle
(4 spd)
Total HC -v
Dilution
Tunnel*
0.59
0.32

0.58
0.20

0.60
0.22

0.69
0.19


0.78
0.64


0.54
0.35


0.25
0.08


0.52
0.30


0.51
0.20


0.87
0.21


0.27
0.16


,pm/mi
Cert
Cell**
0.55
0.35

0.53
0.24

0.61
0.27

0.66
0.21


0.76
0.57


0.51
0.36


0.30
0.14


0.55
0.37


0.55
0.22


1.09
0.68


0.36
0.16


CO igi
Dilution
Tunnel*
1.51
0.92

1.35
0.75

1.49
0.89

1.91
0.72


1.58
1.16


2.61
1.37


1.35
0.73


1.19
0.64


1.01
0.41


1.69
0.69


1.40
0.64


m/mi
Cert
Cell**
1.60
1.00

1.32
0.78

1.52
0.92

1.90
0.73


1.63
1.28


2.50
1.47


1.41
0.79


1.21
0.72


1.00
0.43


2.01
0.76


1.42
0.58


NOx .vj
Dilution
Tunnel*
1.49
1.20

1.67
i.48

1.56
1.24

1.62
1.22


1.52
1.41


1.82
1.57


1.36
1.25


0.98
0.78


0.87
0.58


1.16
1.12


1.40
1.55


^m/mi
Cert
Cell**
1.63
1.28

1.84
1.54

1.58
1.28

1.73
1.26


1.52
1.34


1.86
1.71


1.40
1.01


1.05
0.83


0.95
0'.65


1.05
1.03


1.43
1.34


CO.,
Dilution
Tunnel*
458
334

409
312

412
304

391
270


501
398


469
403


463
363


262
198


238
172


362
307


407
383


^gm/mi
Cert
Cell**
486
362

459
329

422
315

396
283


538
427


489
443 '


480
394


277
220


246
185


361
307


441
367


Fuel Economy
Dilution
Tunnel*
22.1
30.3

24.6
32.9

24.4
33.2

25.7
37.5


20.2
25.3


21.4
25.1


21.8
27.9


38.3
50.9


42.4
58.9


27.3
32.6


24.8
26.5


mpg
Cert
Cell**
20.8
27.9

22.0
30.7

23.9
32.1

25.4
35.7


18.7
23.7


20.6
22.8


21.1
25.8


36.3
45. S


40.8
54.6


27.7
•32.8


22.9
27.7


Cycle


FTP
HWFE

FTP
HWFE

FTP
HWFE

FTP
HWFE


FTP
KWFE


FTP
HWFE


FTP .
HWFE


FTP
HWFE


FTP
HWFE


FTP
HWFE


FTP
HWFE


Dilution tunnel measurements were made in cell 508.
Cc r11f teat Ion measurements were made in cells 5 and 6.
Died special high-flow cooling fan for certification testing.

-------
                                          TABLE II






LIGHT DUTY DIESEL COMPARISON OF DILUTION TUNNEL TEST RESULTS TO CERTIFICATION CELL TEST RESULTS




                                       Relative Values*

Vehicle & VID
Oldsmobile 350
93J7127
'79 Cert Vehicle
Oldsmobile 260
93HF149
'79 Cert Vehicle
Oldsmobile 260
93HF-84482F
'79 FE Vehicle
Oldsmobile 260
931IF18J.
'79 Cert Vehicle (5 spd)
Chevrolet Pickup
8TJ9-168F
'79 Cert Vehicle
(2-hole injection)
Dodge Truck
B356
'79 Curt Vehicle (4 spd)
Mercedes 300D
123.130-12017745
Special Build/Turbo-
Charged with EGR
Volkswagen Dasher
406 Z 2466
"79 Cert Vehicle (4 spd)
Volkswagen Rabbit
406 Z 2465
'79 Cert Vehicle (5 spd)
Pc-Jgcot 5041)**
6*1-
'73 Cert Vehicle (4 spd)
Int'l. Harvester Scovit 300
'79 Cert Vehicle (4 spd)
Repca - one std. dev.
as 2 of mean

Total HC
7,
-7.3
8.6

-9.4
16.7

1.6
18.5

-4.6
9.5

-2.6
-12.3


-5.9
2.8

16.7
42.9


7.3
9.1

7.3
9.1

20.2
f9.1

25.0
0
14%


Total CO
%
5.6
8.0

-2.3
3.8

2.0
3.3

-0.5
1.4

3.1
9.4


-4.4
6.8

4.3
7.6


1.6
11.1

-1.0
4.6

15.9
9.2

1.4
-10.3
112

* Rcl;
Total

8.6
6.2

9.2
3.9

1.3
3.1

6.4
3.2

0.0
-5.2


2.2
3.2

2.9
-23.8


6.7
6.0

8.4
10.8

-10.5
-8.7

2.1
-15.7
9%

itive Value =
NOx Total CO
; %
5.8
7.7

10.9
5.2

2.4
3.5

1.3
4.6

6.9
6.8


4.1
9.0

3.5
7.9


5.4
10.0

3.2
7.0

-".3
0.0

7.7
-4.4
3%

Cert Cell — Dilution Tunnel
Fuel Economy
%
-6.2
-8.6

-11.8
-7.2

-2.1
-3.4

-1.2
-5.0

-8.Q
-6.7


-3.9
-10.1

-3.3
-8.1


-5.5
-11.1

-3.9
-7.9

1.4
0.6 '

-8.3
4.3
3X



Cycle
FTP
HWFE

FTP
HWFE

FTP
HWFE

FTP
HWFE

FTP
HWFE


FTP
HWFE

FTP
HKFE


FTP
HWFE

FTP
HWFE

FTP
i.i i ~

FTP
HWFE
FTP


                                                             Cert Cell




                                 ** Used special high-flow cooling fan for cert cell testing

-------
                     GR.APH I- DILUTION TUNNEL.
                                                      .  CERTIFICATION CELL  HC
        • FTP DATA

        X HW FED ATA
   U
  .1.0
J
U.1
•z


\

o
   0.8
  0.1
o
CD



5
o
Oi
k
o
h
  Qfe
  0,5
  0.3
  o.\
                                                         HVOFE
                                                                                        FTP
                          0.3     o.M     o.s     o.b     0.1     0,6

                             TOTAL UMPROCARBOM- ceftT^FicA.TioM CELL
                                                                                              i. a

-------
                                      TUUUEL CO vs. CERT I F 1CA.TIO t\ CELL CO
9-6
     • FTP
     X HNWFE DATA,
a.o
0.3.     O.M
                                0.8      »-O     (.a     |,^      l.b
                                  tAcmoxtPE-CERTIFICATION
                                                                     a.o

-------
               GRAPH'S!- "DILUTION TUWE.L UOX «. CERTIFICATION! CELL NOX
     • Rf> "DATA
     X HWE DATA
a.o
                                                                               FTP
0.4
0,b     0.6     1.0     LSI      l.M     »,t
   OF >4|TRO&EK> CERTlFIC^T^O^i CELL
                                                                      1,8

-------