EPA-AA-EOD-80-1
Correlation Report
1979 EPA-Volkswagen
Light Duty Diesel Correlation Study
by
Carl Scarbro
Correlation Group
Testing Programs Branch
Engineering Operation Division
Office of Mobile Source air Pollution
Ann Arbor, Michigan 48105
-------
Introduction
In response to a meeting with representatives of Volkswagen on September 14, 1979,
a correlation study was organized to characterize some of the factors causing
hydrocarbon offsets during the 1980 diesel certification program. (Volkswagen read
40% lower hydrocarbon emissions than EPA) The hypotheses presented at this time by
Volkswagen were: (1) offsets do not occur because of sample line variations from
lab to lab, and (2) offsets are caused by diesel fuel composition differences
between EPA's fuel and Volkswagen-Wolfsburg's fuel. These two hypotheses were
tested in two separate experiments at EPA as part of the correlation study.
After completing the testing for the above two studies, Volkswagen returned the
test vehicle to their Wolfsburg facility in Germany and performed an independent
study with larger vehicle cooling fans. VW also began a second, but similar, vehi-
cle in an interlaboratory round robin with the first vehicle. The purpose of the
fourth study using the two vehicles was to identify test cell environment effects
on vehicle warmup during testing.
In November Volkswagen brought both vehicles to the United States and requested EPA
to do a series of six Hot LA-4s on each vehicle at the EPA facility. We agreed to
do the test cell environment effects study if they would test the vehicles in as
many non-Volkswagen laboratories as they could. They accepted. These tests were
performed on Volkswagen's test fuel used in this country and performed in an iden-
tical manner at the six facilities participating in the round robin.
This report is the summary of the three different studies performed in cooperation
with EPA plus the results of the independent Volkswagen cooling fan experiments.
These studies included similar vehicles and facilities, but addressed different
correlation cause/effect relationships.
Conclusions
Factors which can cause a hydrocarbon offset between EPA and VW's development
facility in Wolfsburg have been indentified as cooling fan capacity, test cell
environment's effect on vehicle and/or sample system temperature stabilization,
and possibly an effect due to fuel characteristics and/or a fuel change effect.
The offsets between Volkswagen-Wolfsburg, Chrysler Proving Grounds, and EPA-Ann
Arbor on a large number of LA-4s on two Rabbit diesels is below 6%. VW-Toledo and
VW-Westmoreland are 25% below EPA and FIAT is 50%. lower.
Recommendations
It is unfortunate that a better real world simulation of air movement over the test
vehicle is not readily adaptable to present certification testing, but the use of a
standardized single speed fan would facilitate more comparable results between test
facilities. It is also desirable that test personnel have adequate direction to
duplicate the fan to vehicle placement that is desired by the manufacturer, within
the constraints of the Federal Register.
The soak control vs. oil temperature controlled vehicle tests indicate cell envi-
ronment does effect the vehicle and/or sampling system thermal state before startup.
-------
Since test cells are not easily redesigned to reflect one another one must
therefore perform hot temperature stabilized LA-4s vs. hot soak controlled LA-4s to
quantify offsets due to the cell environment.
The fuel controversy is best resolved by a standard test fuel that can be bought by
all manufacturers from a single lot.
Vehicles and Test Plans
The chronology of events, test plans, test fuels, and locations are summarized
below:
Sample Line
Test Site Vehicle Fuel n/Vehicle Fan Temp. °F
Location
VW-Toledo 1
VW-Westmoreland 2
EPA-Ann Arbor 6
EPA-Ann Arbor 6
VW-Toledo 1
EPA-Ann Arbor 6
EPA-Ann Arbor 6
EPA-Ann Arbor 6
EPA-Ann Arbor 6
VW-Toledo 1
1
1
1
1
1
1
1
1
1
1
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
Wolfsburg
EPA
AMOCO-50
AMOCO-50
Vehicle was shipped to Wolfsburg, Germany.
VW-Wolfsburg
VW-Wolfsburg
VW-Wolfsburg
VW-Wolfsburg
VW-Wolfsburg
VW-Wolfsburg
6
6
Inspection
4,6,8
Inspection
.4,6,8
6
6
1
1
1
2
2
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
Vehicles were shipped to the United States.
VW-Westmoreland 2
VW-Toledo 1
VW-Toledo 1
FIAT-Dearborn 1
Chrysler 5
EPA-Ann Arbor 6
1&2
1
1&2
1&2
1&2
1&2
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
AMOCO-50
6
6
6
6
6
6
6
6
6
6
11
18
40
34
6
6
6
6
6
6
6
6
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
SCFM
SCFM
SCFM
SCFM
SCFM
SCFM
SCFM
SCFM
SCFM
SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
5000 SCFM
375
375
375
322
375
375
375
375
375
375
375
375
375
5000 SCFM 375
375
375
375
375
375
375
375
375
All test sequences were 6 Hot LA-4s or 1 FTP and 5 Hot LA-4s to maximize the amount
of data that could be collected in the allotted test time and equipment availabil-
ity. The specifics on vehicle preps and mileage accumulation for the three studies
are outlined in Tables 1, 3, and 5 in the appendix.
Vehicle *1 was a 1978 production VW diesel Rabbit and is referred to as the soak
controlled vehicle in the test cell environment study. Vehicle #2 was another
-------
1978 production VW diesel Rabbit but with temperature monitoring of engine oil,
engine intake air, and fuel in the fuel tank . This #2 vehicle was the tempera-
ture controlled vehicle which, after the 10 minute soak, was driven at 30 mph until
the engine oil temperature was 200°F. The vehicle then returned to idle and the
hot LA-4 began.
Discussion of Results
Test results are presented in the appendix for each specific vehicle/test procedure
combinations.
Volkswagen's claim of hydrocarbon measurement insensitivity to sample line tempera-
ture on the Rabbit diesel was validated at EPA. It is interesting to note that
most diesel vehicles including the VW powered Volvo and VW built Audi 5000 are sen-
sitive to sample system temperature in the continuous diesel hydrocarbon system.
This has been verified by EPA-Ann Arbor, VW-Wolfsburg, and Volvo-Goteborg. Figure
1 in the appendix demonstrates it is characteristic for diesel hydrocarbon measure-
ments to be sensitive to temperatures, but this sensitivity was not observed for
the VW Rabbit that was tested at EPA.
Volkswagen's hypothesis that different fuels cause offsets on hydrocarbon emissions
was not proved or disproved in our test procedure. Statistically there are emis-
sion offsets for both HC and C02 between EPA's diesel #2 and Wolfsburg fuel,
AMOCO-50 and Wolfsburg fuel, but not between AMOCO-50 and EPA's diesel #2. When
viewing data chronologically, however, the total variability of the vehicle ranged
from .55gram/mile to a low of .34 gram/mile on the baseline fuel. This contained
the sample means of the other test fuels. Experience with Volvo certification
vehicles:.using Volkswagen technology indicates that something during a fuel switch
may cause a large hydrocarbon offset that becomes smaller as the vehicle accummu-
lates mileage. This report therefore does not resolve the fuel switch contro-
versy. (See Fig. 2, 3, 4 for chronological test data.)
The analysis of the fuels used in the study are shown in the appendix. The differ-
ences in Volkswagen's and EPA's analysis were dramatic. Using EPA's analysis, the
AMOCO-50 was within Federal Register specifications, except for the initial boiling
point and 10% point in the distillation. The Wolfsburg fuel was out of specifica-
tion for aromatics, API Gravity, viscosity, flash point, initial boiling point,
10%, and 50% in the distillation.
Hypothesis testing (T-stat) accepted the hypothesis that the cold start hydrocarbon
data mean .was equivalent to the hot start data, on all three fuels overall during
the fuel switch test procedure.
The result of the two vehicle test procedures, soak time vs. soak time plus oil
temperature stabilization, did indicate the degree of offset between EPA and the
other labs due to cell environment differences. The vehicle that was temperature
stabilized had approximately the same test to test variability as the soak time
stabilized vehicle, but had a lower lab to lab offset for hydrocarbon emissions,
indicating ambient cooling conditions during phase 1 and the 10 minute soak causes
some of the offset found between manufacturers and EPA.
Variability over time can be seen in Figure 4, 5, 6, and 7 for both test vehicles.
The soak controlled test vehicle displays more variability in hydrocarbon emis-
sions, even site to site within Volkswagen-Wolf sburg, than the temperature con-
trolled vehicle. Whether the temperature of importance is vehicle intake air,
-------
engine oil or sample exhaust system through to the continuous hydrocarbon sample
point was not determined by this study.
Volkswagen performed tests on both vehicles using a different capacity cooling fan,
both in Germany and Toledo on vehicle #1 and in Germany only for vehicle #2
using the vehicle handling procedure outlined in Table 5. The larger cooling fans
caused an approximate 20% increase in HC. The small capacity fans were the origi-
nal cooling fans used in Wolfsburg for Volkswagens certification work. At that
time Volkswagen was 40% lower than EPA in diesel hydrocarbon results in the FTP.
The critical temperature in this study, as measured by vehicle #2 temperature
recorder, was engine intake air. The cooler the intake air the higher the hydro-
carbon emissions. (See Figure 8 and 9).
The tests conducted by Fiat in Dearborn and Volkswagen Toledo indicated the impor-
tance of cooling fan location. LA-4s 1 to 3 in the Fiat tests had a different fan
location than Run's 4 to 6 (see Figure 10) on the temperature controlled vehicle.
The lower temperature traces for Run's 4 to 6 resulted in slightly higher HC
results. Toledo tests for the same vehicle was an attempt to better simulate air
intake temperature traces generated at EPA. The result of the experiment did
bringToledo's HC emissions closer to the level of hydrocarbon emissions measured at
EPA on the same vehicle over the same test procedure. (Figure 11 and 12)
-------
APPENDIX
Table 1
Table 2
Table 3
Table A
Table 5
Table 6
Table 7
Table 8
Table 9
Table 101"
Table 11
Figure 1
Figure 2
Figure 3
Figure A
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Vehicle Test Plan for Volkswagen HFID Sample System Temperature Study.
HFID Sample System Temperature Study Gaseous Emissions from VW Rabbit
*1 for Hot Start LA-4s.
Vehicle Test Plan for Volkswagen Fuel Effects Study: EPA Tests.
Hot LA-A Emission Results for Volkswagen Fuel Effects Study: Three
Fuels, Vehicle #1, Sample Line Temp, at 375°F, EPA Tests.
Vehicle Test Plan for Volkswagen Test Cell Effects Study at Six Facili-
ties.
Vehicle Coast Down Times (In Seconds)
Gaseous Emission from Vehicle *2, Temperature Controled Vehicle, at
Six Facilities for Volkswagen Test Cell Effects Study.
Gaseous Emission from Vehicle #1, Soak Controled Vehicle, at Six
Facilities for Volkswagen Test Cell Effects Study.
Offset on Two Volkswagen Rabbits at Five Facilities for Hot Start
LA-As: Fuel AMOCO 50.
Cooling Fan Capacity Study on VW Rabbits #1 and #2: Fuel AMOCO 50,
Sample Lines at 375°F.
Analysis of Diesel Test Fuels.
Hydrocarbon Response vs. Sample System Temperature for Vehicles Driving
Hot Start LA-As.
HC and NOX vs. Test Sets in Chronological Order on Vehicle
September 1979.
in
CO and C02 vs. Test Sets in Chronological Order or Vehicle #1 on
September 1979.
Hydrocarbon vs. Time: Vehicles #1 and #2.
CO vs. Time: Vehicles *1 and #2.
NOX vs. Time: Vehicles #1 and #2.
C02 vs. Time: Vehicles #1 and #2.
Site 6 VW Wolfsburg, Hot LA-As, Fan-cap. < 5000 cfm.
Site 6 VW Wolfsburg, Hot LA-As, Fan-cap. 5000 cfm.
Fiat-Dearborn, Hot LA-As, Two Fan Locations.
-------
Figure 11 VW-Mobile Lab Toledo, Fan Location #1 for Six LA-4s.
Figure 12 VW-Mobile Lab Toledo, Fan Location #2 for Six LA-4s.
Calculations
-------
Table 1 Vehicle Test Plan for Volkswagen
HFID Sample System Temperature Study
Vehicle *1 Fuel Used: AMOCO 50 EPA Tests
Procedure
LA-4 prep, no fuel drain, 12 hr. soak
FTP + 10 min. soak
LA-4 + 10 min. soak
LA-4 + 10 min. soak
LA-4 + 10 min. soak
LA-4 + 10 min. soak
LA-4 + 10 min. soak
LA-4 prep, no fuel drain, 12 hr. soak
FTP + 10 min. soak
LA-4 + 10 min. soak
10 min. soak
LA-4
LA-4
LA-4
LA-4
+ 10 min. soak
+ 10 min. soak
+ 10 min. soak
Sample System Temp in °F
375°F
375°F
322°F
322°F
322°F
375°F
375°F
375°F
322°F
322°F
322°F
375°F
Table 2 HFID Sample System Temperature Study
Gaseous Emissions from VW Rabbit #1 for Hot Start LA-4s
Facility
EPA-
Ann Arbor
n=5
EPA-
Ann Arbor
n=4
HC
Statistic g/mile
!<
S
%c.v.
><
s
%c.v.
.50
.02
4
.49
.04
8
CO
g/mile
.98
.02
2
.98
.05
5
NOX
g/mile
.96
.02
2
.94
.02
2
C02 Sample
g/mile Fuel Filter Temp.
232.4 AMOCO-50
1.3
1
231. AMOCO-50
1.4
1
375°F
322°F
-------
Table 3 Vehicle Test Plan for Volkswagen
Fuel Effects Study: EPA Tests
Vehicle #1 Sample Line 375°F
Procedure
LA-4 prep
FTP, 5 Hot LA-4s, 10-12 min. soaks between tests
Fuel drained and filled to 50%
40 road miles and LA-A prep, 12 hr. soak
FTP, 5 Hot LA-As, 10-12 min. between tests
Fuel drained and filled 50%
AO road miles and LA-A prep, 12 hr. soak
FTP, 5 Hot LA-As, 10-12 min. between tests
Fuel drained and filled 50%
AO road miles and LA-A prep, 12 hr. soak
FTP, 5 not LA-A's, 10-12 min. between tests
Fuel
AMOCO 50
AMOCO 50
Wolfsburg fuel
Wolfsburg fuel
Wolfsburg fuel
EPA's diesel #2
EPA's diesel #2
EPA's diesel #2
AMOCO 50
AMOCO 50
AMOCO 50
Table A Hot LA-A Emission Results for Volkswagen Fuel Effects Study:
Three Fuels, Vehicle #1, Sample Line at 375°F, EPA Tests
Facility
EPA
Ann Arbor"'
n=A
FTP
EPA
Ann Arbor
n=5
FTP
EPA
Ann Arbor
n=6
FTP
EPA
Ann Arbor
n=6
FTP
EPA*
Ann Arbor
n=21
FTP
Test
Date
Sept. 25,
1979 "
(n=l)
Sept. 27,
1979
(n=l)
Sept. 29,
1979
(n=l)
Sept. 28,
1979
(n=l)
Sept
1979
(n=4)
Statistic
f
s
°/r \t
/OO. V .
>r
s
%c.v.
7
s
%c.v.
>r
s
%c.v.
s
%c.v.
HC
g/mile
.AO
.01
2.5
.A3
.36
.OA
11
.32
.A5
.02
A
.AA
.A3
.05
12
.Al
.41
.05
12
.40
CO
g/mile
.95
.01
1
1.0
.94
.06
6
.90
.93
.05
5
.90
.94
.05
5
1.0
.94
.03
3
.952
NOX
g/mile
.96
.01
1
.98
.94
.02
2
.94
.91
.03
3
.93
.93
.02
2
.96
.93
.03
3
.95
C02
g/mile
234.0
1.5
1
239
221.0
1.6
2
224
224.6
2.2
1.0
229
226.0
2.2
1
235
226.1
5.2
2
231.75
Fuel
AMOCO-50
Wolfsburg
EPA's
#2 diesel
AMOCO-50
%
AMOCO
Wolfsburg
EPA's #2
diesel
*The mean and standard deviation of all tests in the fuel effect study at EPA on
Vehicle #1
-------
Table 5 Vehicle Test Plan for Volkswagen
Test Cell Effects Study at Six Facilities
Fuel used: Toledo AMOCO Sample lines at 375°F
Testing Sequence Vehicle 2 Vehicle 1
1 LA-4 prep LA-4 prep
2 12-hour soak 12-hour soak
3 FTP FTP
4 10-min. soak+ 10-min. soak
5 30mph until oil temp=100 C
stop vehicle
6 Hot LA-4 Hot LA-4
7 Repeat sequence 4 thru 6 Repeat sequence 4 thru
until 5 Hot LA-4's are 6 until 5 Hot LA-4's
completed are completed
8* (3) 55 to 45mph coast downs (3) 55 to 45mph coast
downs
*At EPA and VW-Westmoreland only.
Table 6 Vehicle Coast Down Times (In Seconds)
Vehicle #1 Vehicle *2
Facility Date 5>5-5mptv 5£-45mph Date 55-5roph 55-45mph
VW-Westmoreland 9/79 96 11.6 12/79 91 12.1
VW-Toledo 9/79 82 10.2
EPA-Ann Arbor 9/79 95 11.3 12/79 11.1
EPA-Ann Arbor 1/80 11.2
-------
Table 7 Gaseous Emissions from Vehicle #2 Temperature Controlled Vehicle
At Six Facilities for Volkswagen Test Cell Effects Study
Facility
VW-
Wolfsburg
Cells 4,6,
8, Inspection
n=34
VW-
Westmoreland
n=14
VW-
Toledo
n=18
FIAT
Ann Arbor
n=6 ---
Chrysler
n=6
EPA
Ann Arbor
n=6
Test
Date
Nov
1979
Dec
1979
Nov-Dec
1979
Dec
1979
Dec
1979
Dec
1979
Statistic
~x
s
SflC.V.
~x
s
%C.V.
x"
s
%c.v.
><
s
%c.v.
a
s
%c.v.
><
s
%c.v.
HC
g/mile
.30
.045
15
.27
.03
11
.30
.04
13
.23
.02
9
.30
.03
10
.30
.02
7
CO
g/mile
.85
.085
10
.85
.05
6
.98
.07
7
.87
.03
3
.94
.04
4
.801
.036
4
NOX
g/mile
.90
.08
8
.93
.04
4
.92
.025
3
1.03
.09
9
1.04
.035
3
.996
.012
1
C02
g/mile.
244.8
7.3
3
239.5
3.5
1
247.8
3.3
1
255.6
9.8
4
247.4
2.15
1
237.4
3.3
1
Fuel
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
-------
Table 8 Gaseous Emissions from Vehicle ffl Soak Controlled Vehicle
At Six Facilities for Volkswagen Test Cell Effects Study
Facility
VW-
Wolfsburg
Cells 4,6,
8, Inspection
n=30
VW-
Westmoreland
PA
n=6
VW-
Toledo
n=6
FTP
-j"
FIAT
n=6
Chrysler
n=6
EPA
Ann Arbor
n=6
FTP
Test
Date
Nov
1979
Nov-Dec
1979
Nov-Dec
1979
(n=l)
Dec
1979
Dec
1979
Oan
1979
(n=l)
Statistic
7
s
%c.v.
7
s
%c.v.
>r
s
%C.V.
X
s
%C.V.
x"
s
%C.V.
>T
s
%C.V.
HC
g/mile
.40
.08
20
.30
.018
6
.30
.005
2
6
.35
.20
.02
10
.41
.015
4
.39
.04
10
.35
CO
g/mile
.99
.16
16
.96
.035
4
.93
.067
7
1.0
.87
.03
3
1.01
.025
2
.96
.03
3
1.0
NOX
g/mile
.92
.08
9
.97
.05
5
.91
.03
3
1.09
1.00
.03
3
.966
.015
2
1.06
.02
2
1.09
C02
g/mile
247.1
19.8
8
239.7
2.5
1
242.2
5.26
2
246
247.8
3.1
1
234.0
1.7
7
232.9
3.4
1
246
Fuel
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
Toledo
AMOCO
-------
Table 9 Offset1 on Two Volkswagen Rabbits
at Five Facilities for Hot Start LA-4's: Fuel AMOCO-50
Sample Lines at 375°F
Offsets on
Vehicle #1
6 tests at EPA
HC'
CO
NOX
C02
Offsets on
Vehicle #2
6 Tests at EPA
HC
CO
NOX
C02
!% offset = (Facility
VW-
Wolfsburg
n=30
2.6%
3.1%
-13.2%
6.1%
n,34
0.0%
6.1%
-9.6%
3.1%
- EPA) * 100
VW-
Toledo
n=21
-23%
-3.1%
14.2%
4.0%
n=18
0.0%
22.3%
-7.6%
4.4%
VW-
Westmoreland
n=6
-23%
0%
-8.4%
2.9%
n=!4
-10.0%
6.1%
- 6.6%
0.9%
Fiat
n=6
-48.7%
- 9.4%
5.6%
6.4%
n=6
-23.0%
8.6%
3.4%
7.7%
Chrysler
n=6
5.1%
5.2%
-8.9%
4.7%
n=6
0.0%
17.4%
4.4%
4.2%
EPA
-------
Facility
Table 10 Cooling Fan Capacity Study on VW Rabbits
*1 and #2: Fuel AMOCO-50, Sample Lines at 375°F
HC
CO
NOV
Statistic g/mile g/mile G/mile
C02
g/mile
Fan
Capacity
Vehicle *1
VW-
Wolfsburg
Cell #6
n=ll
VW-
Wolfsburg
Cell #6
VW-
Toledo
n=6
VW-
Toledo
1=6 -^
Vehicle *2
VW-
Wolfsburg
Cell #6
n=6
VW-
Wolfsburg
Cell #6
n=6
10-79
10-79
12-79
12-79
8-79
9-79
10-16
11-8
X
s
%C.V.
x~
s
%c.v.
x
s
%C.V.
X
s
%C.V.
£
s
%C.V.
7
s
%C.V.
.42
.055
13
.33
.04
12
.30
.005
2
.25
.023
9
.24
.015
6
.30
.015
5
1.03
.115
11
.92
.04
4
.93
.067
7
.93
.09
10
.82
.01
1
.82
.02
2
.88
.033
4
.89
.04
4
.91
.028
3
.97
.02
2
.83
.012
1
.90
.037
5
"239.0
8.84
4
237.3
5.9
2
242.2
5.3
2
241.5
3.6
1
243
3.2
1
241.9
3.7
2
approx.
5000cfm
Less than
5000cfm
high
capacity
low capacity
tan
Less than
5000cfm
approx .
5000cfm
-------
Table 11 Analysis of Diesel Test Fuels
Cetane Number
Aromatics
Olefins
Saturates
Distillation
Data in °F
IBP
5%
10%
20%
30%
40%
50%
60%
70%
80%
90%
95%
EP
AMOCO 50
's Analysis
Wolfsburg
diesel #1
Wolfsburg
diesel #1
i9.2
Federal
AMOCO 50 diesel #1 diesel #1 XEPA's Register
EPA's Analysis1 VW's Analysis EPA's Analysiyaiesel #2 Specification^
44.0
A95
43.3
29.0%
.5%
70.5%
0.0
54.5
368
35.6
41.2
1.87cst
118°F
.276wt%
249
345
370
391
408
430
452
475
501
530
573
607
627
98.1
lOOvp
o.o\
34.4
2.61cst
138°F
.38wt%
360
420
510
596
627
42-50
27% min.
33-37
2.0-3.2cst
130°F min.
0.2-0.5
340-400
400-460
470-540
550-610
580-660
6.80
'The averageyof two samples drawn from the same fuel batch.
-------
15
Table 11 Analysis of Diesel Test Fuels
Wolfsburg Wolfsburg Federal
AMOCO 50 AMOCO 50 diesel #1 diesel #1 EPA's Register
VW's Analysis EPA's Analysis1 VW's Analysis EPA's Analysis diesel #2 Specification;
Cetane Number
Cetane Index 48.5
Aroniatics '"-^^ -- '''.'
. Olefins . '^' : " ;
Saturates
API Grav. 36.5
Vis at 40 C 2.0cst
Flash Point 153°F
Sulfur
Distillation
Data in °F
IBP 341
5%
10% 402
20%
30% -.--..
40%
50% 495
60%
70%
80%
90% 601
95%
EP 615
Recovery
Residue
Total Recovery
Loss
43.3
45.6 50.0
.:;29.0% '.; jU-V;^.'- -...,..-,:
- ;' -.5% :'""~;---" ' ' '
70.5%
35.4 42.34
2.21cst 1.7cst
131°F 137°F
.271wt%
276 318
379
396 368
425
446
467
486 455
506
526
548
580 582
610
626 623
98.3
1.7
100.0
0.0
48.2 44.0
50.0 47,4
20.0% 35.6
.5%
79.5%
41.2 34.4
1.87cst 2.61cst
118°F 138°F
.276wt% '.38wt%
249 360
345
370 420
391
408
430
452 510
475
501
530
573 596
607
627 627
98.1
1.9
100.0
0.0
42-50 .
27% min.
33-37
2.0-3. 2csi
130°F min
0.2-0.5
340-400
400-460
470-540
550-610
580T660
C/H Ratio
6.80
-'The average of two samples drawn from the same fuel batch.
-------
i
<;
-------
---T--:-:--: -i -t
! ! I i ! . ! ! ' : < ' , ! I ' : ! _! :
j~~i'~ 7r~F~' !~~1I I ~j j
t-._ufl Krv'HieWt? L^_rToVeAc> 4..*v tnocci. JIEl
'-r-'-~-/- ' t--i-J. ^~~-4.-iIj.-l
I ^r-0^ A' ; gftft ; C^rt'flrW ; ; u;tjfe i«L-T
-iTbS-isM--
ijipzi^izn.
n~~v>-'
(:QO y^-
_i i_> C/
1 ! _l_^J_:_:
-UjJ-:-L
I ; :_^_. i_ ;._, j_^.;___L.. .:_;._J_.
rr
: ^
feS>::
l.*^'-
_.iq:
L'_
! -r-
i/'/JKL
-------
-, ___,_,_ rr.,_._ ._^_rr^_1 ^ . I ' / ~-^~-r~J~-
~c~ o^fe^^ :cte/e/
l"--y
. I !_;_' i
,-a _ ._ ; i ...._._: -...L,.. i . _^
^... 4_1_;......1_.;±:: ._ J _:._:_:_. t; i;.
. ___ _ _j_. _, i. _L_.,__ j i
-------
j un: mot fiNv 11101 im
HDMI i 01 o; >: M -IWMJ
-------
KDV AMMV.1S'>-',1.| »ui I1IO.'. MNV II10I 'li
' M.'illl I l»l Oi X 0'. IW-
-------
mn Hint am HIOI 'HI:
II3NI I Ol 05 X Ol-IW-U(
-------
s =
-------
Test #
?/J2^^
1U^5"
HUl-
HUZ
It WO
UMI _
Date
S~£3
J-^3
S-^9
S-A9
7-6
?-.«
X
s
2S
t'S
*?
HC (g/mi)
0-2S
0' M .
0- AS
O- £ ( '"
0- £2, -
V- A3-
0>JH
0. 0 i$
0.03
o.o&
CO" (g/ml)
0'?£
0.S&.
0- Sf
0*3 I
0-83
0.30
0-2Z
0- 0 I
O.Q3L
0-0 (
NOX (g/mi)
O.^^i
o.w
0.8S
OJX \
o.SZ- -
0.-i3
' o. 92,
Q.oOL
G.02
(9-0 /
C02 (g/mi)
£42.6
W3..8
JWrt
XHS.I
A HO. 6
2 16. 3
<2H3
1.?.
6.1
J. £ ' !
-------
J.
Test §
72 Jtff
IbHbl
IZWS
13031
1303J
nom
Date
]0-l(,
10- ft
to-30
If- 1
.11-4
l(- 12
X
s
^3
^ -f
T^"
IIC (y/mi)
^.j; .
^J2
9.3.8
JW6.S
<2<13.3
ZHIA-
3.70
l.H
3.7
-------
Fv«v, tO
I/Ill/
Hot
Test *
/
^
J .
^
5
(,
Date
UL'-3
IS.- 3
12-3
IS.- 3
tt-3
«*3
X
3
2s
*-$
VTTI
HC (g/rr.i)
^-,23 - -
l?-^/
G.ZJ:"
0.
3.63
/?-3
' 9.< '
-------
: 1113354(13
Test 1
1
*
3
Y
5
*
Date
IhW
//-In
11-30
If -30
11*30
X
3
Is
t -5
"j/rft"
HC (g/mi)
^?.52^ - -
'0-2*
0.21
o-
-------
\fW -
Test §
a-
I
ii-
12-1 JJ
a- 1
0.0 (
y/'-'.i )
o .
?
?' .
i
(5
3
r
r~ "
CO (o/:..i)
\J & G
O ^7
0.0%
NOX (g/mi)
^. ^^
<9. y?
^.^ ;
o.i i
0-3H
0.91
0,9,
-------
Calculations
The results of the test plan can be analyzed statistically to determine if the off-
sets between the various test groups are real. The variation between control
groups vs. experimental groups must be statistically similar, caused by similar
effects in both groups. This allows us to ascribe the offset in group means to the
experimental parameter changed in the test procedure.
In the case of the fuel study the values of interest are summarized in the follow-
ing table:
standard
Fuel HC(g/mile) sample size degrees of freedom deviation
ni ni-1 Si
6 5 .02
10 9 .02
5 4 .04
o
The variance, s, are equal for fuels 1 and 2 and the hypothesis that the means
of these fuel are equal is accepted with the T-stat test.
EPA
AMOCO-50
Wolfsburg
1
2
3
'.45
.42
.36
xl ~ X2
t =
sp2
nl
where, SP^, the pooled variance is
SP2 = (nrl)(S1)2 + (n2-l)(S2)2
(n1-l) + (n2-l)
then,
t = 2.90.
Now the critical value of t where the probability of not accepting a true hypothe-
sis is 1 out of 100 times ( = .01) is ^ 2.9768. Our calculated t lies within the
bounds of our critical value. We accept the hypothesis that the hydrocarbon emis-
sions generated or AMOCO-50 and EPA fuels are statistically similar.
The variances between fuels 1 and 3 are not equal but may be similar. The F sta-
tistic will help us determine this by knowing the variance, the degrees of freedom
for each sample, and the probability of not accepting a true hypothesis we are
willing to accept.
-------
If the hypothesis is correct that the variances are equal, then the above calcu-
lated F will be between .0644 and 15.52. They are similar statistically. The cal-
culated T statistic therefore can be calculated with the pooled variance sp2.
t = 4.865
The critical t value for = .01 is +_ 3.2498. The means are not equal. The
hydrocarbon emissions are different because of the difference in fuels, EPA vs.
Wolfsburg diesel fuel.
The hydrocarbon emission test variances for fuel 2 and 3 are again tested for the
hypothesis of equality with the F statistic,
S2
2
F - - - 95
h9,4 ~ 2 ~ ' 5
3
where the critical values of F are .069 and 14.55 when =.01. The variances are
statistically similar. The calculated t statistic is -3.95 and the critical value
for t when =.01 is / 3.0123. The hydrocaron emissions are not equal between the
AMOCO-50 and Wolfsburg diesel fuels.
Similar F and t stat tests indicate several important facts; that the variance for
the temp controlled vs. soak controlled vehicle are equal, the variance between Hot
LA-4s and FTPs on both vehicles are equal and that hydrocarbon emission results
between Hot LA-4s and FTPs are similar statistically.
The use" of an analysis of variance (ANOVA in MIDAS the MTS statistical service)
using fuel as the method of stratifying the data groups, indicates that there is a
difference in the emission results between the groups for HC and
------- |