Environmental Protection Technology Series
LIGHT-DUTY DIESEL EMISSION CORRECTION
FACTORS FOR AMBIENT CONDITIONS
Environmental Sciences Research Laboratoi
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-116
July 1977
LIGHT-DUTY DIESEL EMISSION CORRECTION
FACTORS FOR AMBIENT CONDITIONS
by
Charles T. Hare
Southwest Research Institute
San Antonio, Texas 78284
Task 5 Interim Report
Contract No. 68-02-1777
'Project Officer
Ronald L. Bradow
Emissions Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for pub-
lication. Approval does not signify that the contents necessarily re-
flect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommedation for use.
-------
FORWARD
This document presents completed work on one phase (out of five)
of a large contract effort characterizing diesel engine emissions. This
particular segment is, therefore, an interim report of findings by
Southwest Research Institute relative to the value of humidity cor-
rection factors needed for testing diesel-powered passenger cars for
NO emissions.
x
Ambient air temperature, humidity, and barometric pressure in-
fluence the emission rates of pollutants from passenger cars. For
example, in cold weather gasoline engines are slow to warm up, carbur-
etor chokes remain closed longer, and hydrocarbon emissions are ele-
vated. The emission rate of NO from a passenger car is especially
sensitive to humidity. The reason is well known; namely, the higher
the water vapor concentration in the engine charge, the lower the ef-
fective fuel-air mixture density must be. Thus, high humidity produces
low rates of heat release, low cylinder gas temperatures, and hence
low NO .
x
In emissions certification, ambient conditions can be held con-
stant and, thus, all cars
-------
ABSTRACT
Since emission measurements from passenger cars are performed at
one standard set of ambient conditions and since emission rates of HC,
CO, and NO are sensitive to temperature and humidity, it is necessary
to determine the influence of ambient conditions on emissions from
major classes of vehicles. Although such information has been available
for gasoline engine powered cars for sometime, no such data were avail-
able for diesel powered passenger cars.
This report indicates that diesel HC and CO emissions are rela-
tively insensitive to ambient conditions. Diesel NO emissions, how-
ever, are sensitive to humidity but to a smaller extint than gasoline
engines. Humidity correction factors for NO emissions also appear to
vary with vehicle power-to-weight ratios and are greater for higher
powered vehicles.
This interim report was submitted in partial fulfillment of Con-
tract No. No. 68-02-1777 by Southwest Research Institute under the
sponsorship of the U.S. Environmental Protection Agency. This report
covers a period from November 1975 to August 1976.
-------
ratio vehicles such as the new V-8 engines due in the fall of 1977.
In fact, the authors point out that the one six-cylinder vehicle tested
was much more humidity-sensitive than the four-cylinder models, and
plausible reasons are given for this effect.
No significant temperature or humidity effects for hydrocarbon or
CO emissions were found. This is probably due to the quick warmup of
diesels relative to gasoline engines. Since the range of barometric
pressures available in San Antonio was small, the current results are
not necessarily applicable to high altitude, low station pressure areas
such as the Rocky Mountain States.
Dr. Ronald L. Bradow
Project Officer
IV
-------
CONTENTS
Forward ill
Abstract v
Figures viii
Tables x
1. Introduction 1
2. Conclusions 2
3. Vehicles, Fuel, and Test Instrumentation 4
Test vehicles 4
Test fuel properties 4
Instrumentation and test equipment 8
4. Experimental Plan and Test Program Details 12
\
Experimental plan 12
Details of the test program 14
5. Results, Analysis, and Correction Factor Computations . 17
Results and statistical analysis 17
Computation of correction factors 22
6. Comparison with Other Correction Factors 33
References 36
Appendices
A. Tabular Data 37
B. Coefficients of equations relating dependent and
and independent variables 43
C. Stepwise multiple regression computer outputs 47
-------
FIGURES
Number Page
1 Datsun 220C 5
2 International 100 with Perkins 6.247 engine 5
3 Mercedes 240D 5
4 Peugeot 504D 5
5 Datsun 220C test vehicle on dynamometer, with humidity
and temperature measuring equipment 9
6 Details of air sampling and measurement points for
humidity and temperature 9
7 Second view of air sampling and measurement points 10
*
8 Constant-Volume Sampler (CVS) used for light-duty
diesel exhaust sampling 11
9 Details of heated hydrocarbon analysis equipment 11
10 Instrumentation for measurement of dilute (bag)
emission concentrations 11
lla Planned and actual ambient conditions, Datsun 220C 15
lib Planned and actual ambient conditions, Perkins 6.247 15
lie Planned and actual ambient conditions, Mercedes 240D 15
lid Planned and actual ambient conditions, Peugeot 504D 15
12 NOX emissions as a function of humidity for a Datsun
220C diesel sedan 23
13 NOX emissions as a function of humidity for an
International pickup truck with Perkins 6.247
diesel engine 24
14 NOX emissions as a function of humidity for a
Mercedes 240D diesel sedan 25
viii
-------
FIGURES (continued)
Number Page
15 NOX emissions as a function of humidity for a
Peugeot 504D diesel sedan 26
16 Humidity correction factor K^ for NOX emissions
as a function of humidity, average of results
given in Table 15 32
17 Humidity correction factor comparison 34
IX
-------
TABLES
Table Page
1 Description of Test Vehicles 6
2 Properties of Test Fuel, Federal Specifications, and
"National Averages" for Comparison 7
3 Contract "Test Schedule" 13
4 Revised Temperature and Relative Humidity Values 13
5 Results of NOX Calibration Gas Cross-check 16
6 Correlation Coefficients (r) Between Variables of Record. ... 18
7 Correlation Comparison for Humidity 19
8 Correlation Comparison for Temperature.1 20
9 Correlation Comparison for Humidity and Temperature 21
10 Coefficients of NOX versus Humidity Equations 27
11 Separate Vehicle Correction Factors 28
12 Common Slope Equation Coefficients and Values of L 29
13 Standardized NOX Correction Factors 30
14 Coefficients of Normalized NOX versus Humidity
Equations, All Vehicles Combined 31
15 Summary of Humidity Correction Factor Results 31
16 Humidity Correction Factors for NOX in Tabular Form 35
-------
SECTION 1
INTRODUCTION
Light-duty, diesel-powered vehicles were included under Federal exhaust
emission standards beginning with the 1975 model year,*1) recognizing that
their U.S. sales volume was likely to become appreciable in the near future.
At that time-, as is still the case today, all diesel-powered, light-duty ve-
hicles available to the consumer were either of foreign manufacture or were
equipped with engines of foreign manufacture. These vehicles have tradition-
ally been powered by relatively small engines, having displacements of 3 liters
(183 in3) or less. Emission test procedures for light-duty diesels have been
as similar as possible to those for gasoline-powered vehicles, and emission
standards for the two engine types used in light-duty vehicles have been (and
probably will remain) the same.
At this writing, it appears that the diesel-powered automobile is on
the threshold of a relative "population explosion" in the United States. Con-
cern over fuel economy is one of the driving forces behind this predicted ex-
pansion, but another is certainly a desire by auto manufacturers to secure a
competitive advantage by offering the consumer something novel. The vehicles
which will create the boom, if it comes within the next two years or so, will
be the Volkswagen and Oldsmobile diesels. Anticipating this situation, it
becomes more important to refine existing emission test procedures, providing
additional assurance that present and future emission standards will in fact
achieve air quality goals.
To date, with the exception of continuous HC sampling and integration,
calculation procedures for light-duty diesel FTP'S have been the same as those
for light-duty gasoline FTP's. EPA recognized in the regulations for light-
duty diesels,*1) however, that the NOX correction factor for intake air humi-
dity (Kn) could be modified as necessary pending the availability of test data.
This report contains the information required to make decisions on factors for
correction of light-duty, diesel-powered vehicle emissions to standardized
ambient conditions. These decisions should help to place measured diesel
emissions values on a firmer base, thereby providing greater accuracy in com-
parison of environmental hazards associated with gasoline- and diesel-powered
vehicles.
-------
SECTION 2
CONCLUSIONS
1. combined data from the four light-duty diesel test vehicles, using
a linear model, yielded the following humidity (H) correction factor for NOX
(same for two equal slope/equal intercept variations):
1 - 0.00217 (H-75)
Kh -
where H = humidity in grains H2O/lbm dry air. The equations on which the fac-
tor is based displayed correlation coefficients (r2) of 0.569 ("normalized"
data) and 0.566 ("standardized" data) with the combined emissions data. This
factor is very similar to that originally used for correction of NOX emissions
from heavy-duty diesels.^
2. On the average, a quadratic equation in H (one containing H and H2
terms) correlated better with NOX data than di'd either a linear equation in
H alone or one linear in H and temperature (T). For data on individual ve-
hicles, correlation coefficients (r2) for the quadratic averaged 0.653, those
for the linear in H averaged 0.570, and those for the linear in H and T aver-
aged 0.594. The factor computed using quadratics in H (average of coefficients
for equations using both "standardized" and "normalized" data) is
h = 1 - 0.00228 (H-75) + (1.86 x 10~5) (H-75)^ '
and the r2 of the quadratic factor is 0.616 for "normalized" data and 0.613
for "standardized" data.
3. In addition to the combinations of "independent" variables already
mentioned (H alone, H with T, and H with H2) , emissions were also regressed
against: T alone; T and T2; H, T, and HT; and H, T, H2, and T2. For NOX,
correlations were either worse than for the linear and/or quadratic in H, or
else the additional complication of introducing more variables could not be
justified in terms of improved correlation. For the other emissions (HC and
CO) , results were too mixed and/or correlations were too poor to justify com
putation of correction factors from the equations.
4. Emissions of HC and CO from the International 100 pickup truck
equipped with Perkins 6.247 engine were more strongly dependent on humidity
and temperature than those from the other vehicles. No facts are available
to explain this result, but it may be related to the much lower specific
loading (kg vehicle mass per available engine kW) of the Perkins engine as
-------
compared to the others (lower specific loading would mean lower overall F/A
ratios for this vehicle)-
5. Combination of NOX emission values for the four vehicles by "nor-
malizing" them appeared to yield good data for computation of a final correc-
tion factor. This process eliminated the effect of differing NOX emission
magnitudes among the test vehicles by transforming the data to ratios of "as
measured" values versus "best predicted" values at a standard humidity among
the four test vehicles.
.6. Use of a greater number of test vehicles would be desirable for any
future research aimed at improving the statistical basis for light-duty diesel
emission correction factors.
-------
SECTION 3
VEHICLES, FUEL, AND TEST INSTRUMENTATION
Each topic of this section is treated in a separate subsection for
clarity. Vehicle parameters and specifications are outlined first, followed
by test fuel specifications and requirements. Instrumentation used for test-
ing and analysis is discussed to conclude the section.
TEST VEHICLES
The four light-duty, diesel-powered vehicles used for test purposes
were a Datsun 220C, an International pickup with Perkins 6.247 engine, a Mer-
cedes 240D, and a Peugeot 504D. These vehicles are shown in Figures 1 through
4 for documentation, and descriptions of them are given in Table 1. It was
planned initially to use five test vehicles, but the fifth one was not avail-
able when needed. The decision to proceed with only four vehicles, but to
conduct more tests per vehicle than had been planned, was approved by the Pro-
ject Officer. '
The particular test vehicles used reflected availability of vehicles
for EPA programs at the time testing began more strongly than they reflected
the population of diesel-powered, light-duty vehicles. The Mercedes 240 and
Peugeot 504 were the only diesel automobiles on the U.S. consumer market when
testing began, so to that extent they could be considered representative.
The other two vehicles, however, were research prototypes as far as the U.S.
market was concerned. Loaded vehicle weights ranged from about 1400 to 2000 kg,
and engine size ranged from 2.1 to 4.1 liters. All the engines were of the in-
direct injection, naturally-aspirated type, with similar injection systems
(Bosch and Bosch-liscensed) and compression ratios between 21.0 and 22.2.
Each vehicle was equipped with a 4-speed manual-shift transmission.
TEST FUEL PROPERTIES
All four vehicles were operated on Type 2-D emissions test fuel as
specified in Federal regulations.^ Inspection results on the particular
fuel batch used, EM-238-F, are given in Table 2 along with required specifi-
cations and "national average" properties for comparison. The test fuel was
well within Federal specifications for all properties except end point, at
which it was coincident with the upper limit. As compared to a "national
average" No. 2 fuel, the test fuel contained more sulfur and somewhat more
high-boiling material. Although no hydrocarbon composition data were avail-
able in the survey data, ^ it is likely that the test fuel contained more
aromatics than an average No. 2 fuel.
-------
Figure 1. Datsun 220C.
Figure 2. International 100 with
Perkins 6.247 engine.
-
Figure 3. Mercedes 240D.
Figure 4. Peugeot 504D.
-------
TABLE 1. DESCRIPTION OF TEST VEHICLES
Vehicle Model
Engine Model (if different)
Datsun 220C
Nissan SD22
International 100
Perkins 6.247
Mercedes 240D
OM616
Peugeot 504D
XD90
\
V.I.N.
Engine No. (if different)
Body Type.
Loaded Weight, kg (lbm)a
Inertia Equivalent, kg (lbm)
V
Transmission
Displacement, 1
Cylinders
Power, kW (hp) @ rpm
Injection System
Combustion Chamber
Compression Ratio
Distance on Vehicle, kmb
QL230-103467
SD22-116440
4 door sedan
1551 (3419)
1588 (3500)
4 speed manual
2.16 (132.1)
4
52.2 (70) @ 4000
Kiki
prechamber
22.0
19,861
4H1CODHB23906
247J1042
pickup truck
1982 (4370)
2041 (4500)
4 speed manual
4.06 (247.7)
6
91.0 (122) @ 4000
Kiki
prechamber
21.1
17,830
11511710066208
616916-10-052895
4 door sedan
1492 (3289)
1588 (3500)
4 speed manual
2.40 (146.7)
4
46.2 (62) @ 4350
Bosch
prechamber
21.0
4,677
504A90-2034350
X203043508
4 door sedan
1402 (3091)
1361 (3000)
4 speed manual
2.11 (128.9)
4
48.5 (65) @ 4500
Bosch
prechamber
22.2
4,694
a curb weight plus 136 kg (300
b at end of tests
lbm)
-------
TABLE 2. PROPERTIES OF TEST FUEL, FEDERAL SPECIFICATIONS, AND
"NATIONAL AVERAGES" FOR COMPARISON
Fuel Type
Fuel Code
Density, g/ml
Gravity , °API
Cetane (D976)
Viscosity, CS (D445)
Flash Point, °C (°F)
Sulfur, wt. % (D1266)
FIA:
aroma tics , %
olefins, %
saturates , %
Distillation (D86) :
IBP, °C (°F)
10% pt., °C (°F)
50% pt., °C (°F)
90% pt.f °C (°F)
EP, °C (°F)
Carbon, wt. %
Hydrogen, wt. %
Nitrogen, wt. %
2D Test Fuel
EM-238-F
0.845
36.0
48.6
2.65
94. (202)
0.35
29.8
1.6
68.6
192 (378)
213 (415)
257 (495)
312 (593)
349 (660)
86.8
12.9
0.005
Federal 2D
Specification
b
33 - 37
42 - 50
2.0 - 3.2
54 (130) minimum
0.2 - 0.5
27 (minimum)
b
b
171-204 (340-400)
204-238 (400-460)
243-282 (470-540)
288-321 (550-610)
304-349 (580-660)
b
b
b
"National Average"
No. 2a
b
35.7
49.3
2.71
b
0.249
b
b
b
190 (374)
221 (430)
261 (502)
307 (585)
333 (632)
- b
b
b
a average of five regional averages, 1976 ERDA Diesel Fuel Survey(2),
not sales-weighted
k no specification or no data
-------
A Type 1-D diesel fuel is specified alongside the Type 2-D fuel in the
heavy-duty and light-duty emission regulations,(3»D in case a given manufac-
turer requires No. 1 to be used in its engines. For the test vehicles and
for other market entries anticipated, however, No. 2 diesel fuel will prob-
ably continue to be recommended. The main reason for the more widespread
use of No. 2 fuel is economy. Its price per unit volume is equal to or lower
than No. 1 fuel, while having considerably greater density (and proportion-
ately higher energy content) per unit volume. The only foreseeable circum-
stance which would move fuel usage for diesel cars toward No. 1 fuel would
be a dramatic increase in urban diesel smoke and/or odor complaints as the
light-duty diesel population increases.
INSTRUMENTATION AND TEST EQUIPMENT
The four diesel-powered vehicles used for test purposes were operated
on a standard 2-roll chassis dynamometer, in this instance a Clayton Model
CT-200 which had been modified to EC-50 configuration. This dynamometer used
a 37.3 kW (50 hp) water brake absorber and a belt-driven variable inertia
system to simulate road operation. Inertia and power settings were based
on vehicle weight -and were set according to Federal procedure.^ For test
purposes, the rear tires of the vehicles were inflated to 3.16 kg/cm2 (45
psig) to minimize deflection on the rolls. The Datsun 220C vehicle shown in
Figure 5 was operating on the chassis dynamometer.
Figure 5 also shows the position of the auxiliary cooling fan in front
of the vehicle, producing an air flow of apprdximately 2.36 m^/sec (5000 ft3/
min) . Sampling or measurement points for all the air analysis instrumentation
were located within 0.6 m (2 ft) of the inlet plane of this fan. The air in-
strumentation included two air (dry bulb) temperature thermocouples, one forced
air psychrometer, one electronic hygrometer, and a dewpoint-measuring device.
Figure 6 shows another view of the instruments and sampling/measurement points
with the psychrometer at upper left, perforated relative humidity/dry bulb
temperature sensor for the electronic hygrometer at center, bare-tip thermo-
couple at bottom center, and dewpoint instrument at bottom right. The white
object near top center is a small funnel to which the dewpoint instrument's
sample line was attached. Another view of the area behind the fan is given
by Figure 7.
Of the humidity- and temperature-measuring instruments noted above, only
the electronic hygrometer output and the two dry-bulb temperatures were re-
corded on a continuous basis. The other instruments were monitored manually,
and readings were taken from them at intervals of 2 to 5 minutes during each
test. Yet another source of data was the National Weather Service, from
which humidity data were obtained on an hourly basis during the days and
times when tests were being conducted. The Weather Service data were not
intended as primary information to be used in a statistical sense, but rather
as corroboration of data obtained by our direct measurements. Accuracy of
all the measurements and correlations between systems will be discussed later
in the report.
Measurement of CO, NOX/ and CO2 gaseous emissions was accomplished us-
ing a constant-volume sampler (CVS) and a set of low-concentration gas analyzers
8
-------
Figure 5. Datsun 220C test vehicle on dynamometer,
with humidity and temperature measuring equipment.
Figure 6. Details of air sampling and measurement
points for humidity and temperature.
-------
to read the diluted (bag) emission concentrations. Hydrocarbon emissions were
sampled just after dilution occurred (prior to entry into the CVS), and were
analyzed by a heated FID on a continuous basis. An electronic integrator pro-
vided the means of extracting an average value from the FID output. The CVS
used is shown in Figure 8; and the heated FID detector/oven, control unit,
chart recorder, and integrator are shown in Figure 9. Instruments used for
measurement of bag concentrations (called the "bag cart") are shown in Figure
10. This cart contains a chemiluminescent NOX analyzer, an NDIR CO2 analyzer,
and two low-range NDIR CO analyzers (one long-path and one standard).
Figure 7. Second view of air
sampling and measurement points.
10
-------
Figure 8. Constant-Volume Sampler
(CVS) used for light-duty
diesel exhaust sampling.
Figure 9. Details of heated
hydrocarbon analysis equipment.
Figure 10. Instrumentation for measurement of
dilute (bag) emission concentrations.
11
-------
SECTION 4
EXPERIMENTAL PLAN AND TEST PROGRAM DETAILS
This section deals first with the experimental plan designed to gather
meaningful data about effects of ambient conditions on light-duty diesel
emissions. The second subsection covers the details of the test program as
it actually occurred and the effects of deviations from the original plans.
All this information constitutes the foundation on which the results and con-
clusions of the program are based.
EXPERIMENTAL PLAN
Following submittal of the Contract Work Plan early in the program, ef-
forts began to assemble the apparatus required for control of intake air tem-
perature and humidity. A problem was discovered with the planned approach,
however, because it was to control properties of the engine intake air only
rather than the ambient. The fallacy in the original line of thought was
that "ambient" data were to be taken in a controlled airstream leading to the
vehicle air intake, rather than in a totally-controlled ambient (the room).
Once this problem had been thoroughly discussed, it was decided to use nat-
urally-occurring humidity conditions with control on room temperature only.
A revised Work Plan was submitted to document this change. Original plans
also called for a five-vehicle test program, but it was later agreed to uti-
lize four vehicles (with a greater number of tests per vehicle) because the
fifth vehicle could not be supplied for the program. It was recognized that
a greater number of test vehicles would produce more representative statistics
on light-duty diesels, but other vehicles were simply not available for test
purposes.
The Contract "TEST SCHEDULE", with computational corrections as neces-
sary, is reproduced in Table 3. The specified tolerance on relative humidity
for each test was +_ 2 percent. It was decided that for test purposes, a
slightly different set of temperatures would be employed, namely 68, 77, and
86°F (20, 25, and 30°C). The reason for this change was that light-duty FTP
regulations call for test temperatures from 68 to 86°F (20 to 30°C). With
this minor modification, the relative humidity portion of Table 3 was recom-
puted and now appears as Table 4. These conditions in Table 4 were those
sought (or an approximation thereof) during the test program. Variables ac-
tually used to decide on the worth of running at a given set of ambient con-
ditions were (or were calculated from) original independent variables; and
they were temperature and specific humidity expressed in grains H2O/lbm dry
air.
12
-------
TABLE 3. CONTRACT "TEST SCHEDULE"
Note; Replications in parentheses
wt. % H20
OC(->
0 7*5
1.00
1 9^
1 en
i 71;
•? nn
Humidity expi
vol. % H2O
Opno
1 9O1
1.598
1 qqc
9 •sqn
9 TQC
T T7P
•essed as
grains H20/
lbm drv air
•jc: o
JD . Z
co q
70.7
pp £
i nfi fi
1 tA. 1
i A9 q
Pd/Pw
OQQO
. yyz
OQQP
0.984
Oqor»
Oq"7R
OqTO
.;7 /O
n QCQ
Relative
65°F
•JT Q /O\
J / . O \f.)
CC C /O\
75.3(2)
2 humidity
75°F
oc Q / o ^
.y v^-/
An o r '^^
53.5(3)
CC Q / 0\
Qn n ^*5^
(%) at
85 °F
38.6(2)
AQ ") O\
C.-1 1 O\
D / . / \£.)
(L-J *) O\
if. 1 1 o ^
TABLE 4. REVISED TEMPERATURE AND RELATIVE HUMIDITY VALUES
Humidity as
grains H20/lbm dry air
35.2 ( 5.03)a
52.9 ( 7.56)
70.7 (10.1 )
88.6 (12.7 )
106.6 (15.2 )
124.7 (17.8 )
142.9 (20.4 )
Relative humidity (%) at
68°F (20°C)
34.0
51.0
67.8
77°F (25°C)
-25.1
37.6
50.1
62.5
74.9
86°F (30°C)
____
37.4
46.6
55.9
65.1
74.3
a values in parentheses in g ^O/kg dry air
13
-------
Noting the number of replications specified for each set of conditions
given in Table 3, the original test plan called for 29 FTP's per vehicle (145
FTP's total). It was also requested that the order of the tests be randomized
on a daily basis, so that the vehicles would not be tested in the same order
all the time. This request was complied with by preparing a randomized daily
sequence based on a table of random digits. It does not seem necessary to
reproduce this sequence as a separate item, since it can be deduced readily
from general data tabulations (including dates and run numbers) which are pre-
sented later in the report.
DETAILS OF THE TEST PROGRAM
The decision to use naturally-occurring humidity conditions so that room
ambients could be the variables of record resulted in a somewhat different set
of ambient conditions for tests on each vehicle. Room temperatures were some-
times increased by adding heat, but decreases in temperatures were not attempted
due to the probability of moisture removal in the air conditioners. In a few
cases, tests were conducted using a steam generator in the room to maintain
humidity above ambient. Moisture addition was necessary only for some of the
higher specific humidity conditions. Both humidity and temperature remained
essentially constant during individual tests.
than refer again to general cjata tabulations for actual tes;t am-
bient conditions, these data are plotted in Figures lla through lid as compared
to the planned set of ambient conditions. These graphs should make visuali-
zation of the comparison easier than would a simple tabulation. Although the
humidity points did not fall exactly on the planned values in most cases, the
range and distribution of points achieved should be satisfactory from a statis-
tical standpoint. Since humidity points had to be accepted essentially as
they occurred naturally, a larger number of tests had to be performed than
was planned initially. A total of 174 valid tests were conducted on the four
vehicles, which compares to a total of 145 planned tests on five vehicles.
The tests were split quite evenly among the vehicles; with 44 being conducted
on the Datsun 220C, 45 on the International/Perkins 6.247, 43 on the Mercedes
240D, and 42 on the Peugeot 504D.
Measures used to help control data quality throughout the test program
included CVS propane checks, dynamometer calibrations, and NOX converter checks.
Data from each day of testing were tabulated and graphed to determine whether
or not any investigation should be conducted for processing errors. The ana-
lysis instrumentation was fully calibrated on a monthly basis with gases named
by EPA's Ann Arbor laboratory or with gases traceable to them. This instru-
mentation and these gases were further used to cross-check four NOX calibration
gases sent to SwRI by the Project Officer. The results of this cross-check
are given below in Table 5; and it is apparent that agreement is quite good
in the lower concentrations, but somewhat less satisfactory as the concentra-
tions increase (disagreements up to about 4 percent) . Almost all the NOX con-
centrations analyzed during this program were in the range of 30 to 50 ppm
(dilute sample bags) . No reasons have been identified as yet for the apparent
calibration differences indicated in Table 5. While absolute accuracy is im-
portant for all the emissions data, it probably is just as important for the
ambient data.
14
-------
KEY LI planned
D 1 run
150-
.i= 125-
<0
t
E
^ 100-
o
CM
tn __
c 75-
S
0)
^,
2 50-
E
3
25-
o
[0)0
a o a
o
>•
" \3 S" ~°-
c
D a a -°
a Diii °
O D X
a o a a v,
D EO D -E
a <2
a a>
*
an ^
a a ^
00 E
n 5 x
o g 3-
a a
a
o a oo
1 1 1 1
70 75 80 85
Temperature, " F
FIGURE 11a. PLANNED AND ACTUAL AMBIENT
CONDITIONS, DATSUN 220C
150-
.h 125-
.
•o
J 100-
o
CM
X
£ 75-
0)
=6 50-
'E
3
X
25-
o
[G~|
a @>
a
n .-
L^J to
D ^
a a- '
a c
o @ -°
a s* B.
o X
D u>
fa| [D|P [o] .E
a £
a a>
n D . o a 1
a D £
a « n ° =
o
O DO
a o o
D
1 1 1 1
70 75 80 85
@ 2 runs
9 3 runs
150-
125-
100-
75-
50-
25-
fl—
o
o
n-
0 0
o
0
o
0
o Q [o"|n
D O
a o a a
n a-
D
o a
a
Q
a
0 n °
a a
o O
a o O
a
° I I I I
70 75 80 85
Temperature, " F
FIGURE 11b. PLANNED AND ACTUAL AMBIENT
CONDITIONS, PERKINS 6.247
150-
125-
100-
75-
50-
25-
A
a
D 0
D
a
|O |
a
n 0
D
a a o o
A o
0 D
a a
n 0 n
a
0 0
n - n
a a a
D 0
o
o o
a
a a o o
o
0 1 1 1 1
70 75 80 85
Temperature, °F
FIGURE 11c. PLANNED AND ACTUAL AMBIENT
CONDITIONS, MERCEDES 240D
Temperature, F
FIGURE 11d. PLANNED AND ACTUAL AMBIENT
CONDITIONS, PEUGEOT 504D
15
-------
TABLE 5.
RESULTS OF NOX CALIBRATION GAS CROSS-CHECK
Cylinder
number
MM-2784
MM-2892
MM-2930
MM-2890
Concentration in ppm by analyzing laboratory
EPA-Research
Triangle Park
24.0
89.0
279.5
534.
SwRI -
first check
23.6
87.5
271.
516.
SwRI -a
second check
273.
511.
EPA-Ann
Arbor
24.07
88.87
265.7
505.2
checked against NBS calibration gases
Data on ambient conditions were recorded continuously near the inlet
of the vehicle cooling fan, which is the recommended location for such meas-
urements. These data (dry-bulb temperature and relative humidity) were also
integrated electronically, and they were read manually at intervals of approx-
imately 2 mirifetes. Ambient temperature and^dewpoint temperature were measured
manually at somewhat longer intervals (about 5 minutes) , and both wet- and dry-
bulb temperatures were recorded on a similar schedule. Relative humidity data
from the electronic hygrometer proved to be most reliable of the three humidity-
measuring measurements. This instrument was also calibrated periodically ac-
cording to ASTM recommended practice E104-51. Tl^a dewpoint instrument was used
for all tests during which it operated properly, but it had to be repaired t
several times during the program. It was also 'discovered that data taken using
the psychrometer were inaccurate unless an inordinate amount of time was de-
voted to its care and maintenance; so it was eliminated early in the program,
and no psychrometer data appear in this report. The care necessary to obtain
accurate psychrometer data has been discussed in the literature in some detail. I
It has already been noted that variation in humidity during each parti-
cular run was requested to be + 2 percent relative humidity or less, as re-
ferred to the mean. This tolerance is, therefore, a function of several vari-
ables. It ranges from + 2.1 grains H2O/lbni dry air (+0.29 g H2O/kg dry air)
at 68°F (20°C> and 28.80 in Hg to +_ 4.0 grains H2O/lbm dry air (+_ 0.57 g H2O/
kg dry air) at 86°F (30°C) and 29.78 in Hg. These acceptance bands are quite
reasonable in most cases, especially for the situation in which humidity is
being controlled. To determine the acceptance band for a given test, the mean
relative humidity was first computed from mean specific humidity, temperature,
and atmospheric pressure. Relative humidity was then permitted to vary + 2
percent, and specific humidity was calculated for the extremes. Some runs
were included in the data base which did not quite meet the + 2 percent R.H.
criterion. The additional criteria which these latter runs did meet included:
- relative humidity range within approximately + 5 percent of mean,
- absence of variations which would invalidate time-averaged mean,
- absence of anomalous emissions data.
16
-------
SECTION 5
RESULTS, ANALYSIS, AND CORRECTION FACTOR COMPUTATIONS
The first part of this section is devoted to presentation of the re-
sults in summary form and to statistical analysis of the emission and ambient
data. The second and last subsection covers computation of correction factors
for NOX at non-standard humidity conditions.
RESULTS AND STATISTICAL ANALYSIS
The results of this Task were those which might have been expected
after a review of previous work on studies of the relationship between emis-
sions and ambient conditions. (•>,&) some of the trends that were observed
are summarized below:
- increases in humidity were associated with substantial decreases in
NOX for all jfour vehicles
- changes in temperature were associated with relatively minor changes
in NOX for all four vehicles
- changes in humidity and temperature were associated with relatively
minor changes in production of hydrocarbons and CO from three of the
four test vehicles.
1
The main objective of this task, in addition to confirming existence of the
above trends, was to compute the correction factors needed to correct measured
emission values to those that would be expected at standard ambient conditions.
All the emission correction factors which have been adopted for use in Federal
emission regulations in the past are different from one another^1'3'7', so it
could not be assumed, without testing, that the .light-duty diesel could legit-
imately use one of these other factors.
Emission and ambient data, contained in Appendix A, were gathered on 174
valid FTP runs. In order to determine if any linear relationships existed be-
tween the supposed "independent" variables (humidity by three methods, temper-
ature, and atmospheric pressure), correlation coefficients were calculated
for each pair of variables. It was expected that high correlations would
exist among the humidity values determined by the three methods, that temper-
ature would be only weakly dependent on atmospheric pressure and humidity,
and that humidity (regardless of method) would be essentially independent of
atmospheric pressure. The correlation coefficients (r) actually calculated
are presented in Table 6 and indicate the strengths of the relationships be-
tween the variables.
17
-------
TABLE 6. CORRELATION COEFFICIENTS (r) BETWEEN VARIABLES OF RECORD
Independent variables
dewpoint humidity,
grains I^O/lbm dry air
weather service humidity,
grains I^O/lbm dry air
temperature , ° F
atmospheric pressure,
in Eg
Hygrometer
humidity ,
gr/ibm
0.965
0.925
0.492
-0.746
Dewpoint
humidity ,
gr/lbm
0.893
0.552
-0.778
Weather service
humidity ,
gr/lbm
0.251
-0.710
Temp.
°F
-0.348
Although all correlations were statistically significant (p < 0.05 using
"t" statistics) , those between "independent" variables were in accord with ex-
pectations except the relationships between humidity and atmospheric pressure.
All three humidity values correlated more strongly with atmospheric pressure
than expected. This result was created by a local situation, namely that de-
creased in atmospheric pressure are frequently followed by southerly winds
carrying humid air from the Gulf of Mexico. Different correlations probably
exist in .."other areas due to differences in topography, latitude, and proximity
to large bodies of water.
Since all the humidity values were highly correlated (r >. 0.89) and
since humidity, as measured with the electronic hygrometer was the most reli-
able result from the methods employed, the hygrometer values alone were used
in the statistical analysis. Temperature was the second variable chosen due
to its weak relationship with humidity and atmospheric pressure. Finally,
atmospheric pressure was not considered as an important ambient variable due
to its high correlation (r <. -0.71) with humidity and the expected overlap of
information that would result if both these variables were included. Thus,
only temperature and hygrometer humidity were chosen to be used as the inde-
pendent variables in the analyses described below.
Generalized linear equations were determined utilizing emissions (HC,
CO, or NOX) as the dependent variable and humidity (H) and temperature (T)
as the independent variables. Linear as well as polynomial regressions were
calculated; for each vehicle and each pollutant. The equations were of the
following forms:
(1) E = b0 + bx X
and (2) E = b + bx X +
X
where E = emission value predicted by the regression equation
= constant term
18
-------
b^ = regression coefficient for linear effect of variable X
bll = regression coefficient for quadratic effect of variable
X = independent variable (humidity or temperature).
X
The coefficients obtained using equations (1) and (2) with humidity and tem-
perature as the independent variables are contained in Appendix B. Appendix
C contains computer printouts of stepwise multiple regressions conducted on
individual vehicles, including analysis of variance and summary tables.
Table 7 consists of correlation comparisons between the linear and qua-
dratic fits of humidity to emissions. Using the coefficients of determination
(r2) as a criterion, the best fits are between NOX and humidity (average r2 =
0.653 for quadratic fit) , although all emissions are strongly associated with
humidity for the In ternational -Perkins .
TABLE 7. CORRELATION COMPARISON FOR HUMIDITY
Dependent
Variable
HC
•at-
CO
NOX
Independent
Variable (s)
H
H, H2
Improvement
H
H, H2
Improvement
H
H, H2
Improvement
Coefficients of determination (r2)by vehicle
Datsun
0.020
0.085
0.065
0.050
0.051
0.001'
0.572
0.772
0.200a
Int'l.
Perkins
0.576
0.584
0.008*
0.697
0.724
0.027a
0.486
0.568
0.082*
Mercedes
0.000
0.012
0.012
0.002
0.054
0.052
0.609
0.658
0.049*
Peugeot
0.078
0.127
0.049
0.000
0.002
0.002
0.612
0.615
0.003
Average
0.168
0*202
0.034
0.187
0.208
0.021
0.570
0.653
0.083
indicates improvement was significant at the 0.05 level
Temperature was not as strongly correlated with emissions as humidity,
and this weaker correlation is indicated by the coefficients of determination
in Table 8. Temperature had higher correlations with CO and HC than with NOX/
particularly for the International-Perkins. Utilizing a quadratic fit, the
average r2 was 0.224 for CO, 0.176 for HC and 0.075 for NOX. The addition of
the T2 term to the linear fit of T resulted in an average increase in r2 of
0.074 for HC, but only 0.018 for CO and 0.015 for NOX- Significant increases
Xp < 0.05) in r2 occurred only when adding the T2 term to the linear fit of
T to HC for the Datsun and Mercedes.
*.
Additional regression equations were generated to determine whether or
not humidity and temperature (together) predicted emissions better than these
same independent variables fit separately, as in Tables 7 and 8. Linear as
19
-------
well as stepwise polynomial regressions were calculated for each vehicle and
each pollutant. The forms of generalized equations utilized are given as
follows:
(3) E = b0 + t^H + b2T
(4) E = b0 + bxH + b2T + b12HT
and (5) E = bQ + ^H + b2T + b1;LH2 + b22T2
/\
where E, bo, b]^, and bi:L are as defined in equations (1) and (2) and
b2 = regression coefficient for linear effect of T
kl2 = regression coefficient for the interaction effect of H and T
b22 = regression coefficient for quadratic effect of T
H = humidity (independent) variable .
and T = temperature (independent) variable.
The coefficients obtained in using equations (3) , (4) , and (5) with H and T
are contained in Appendix B. Note that equations (4) and (5) were formed
from a stepwise regression procedure in which H and T were forced into the
equation. Consequently, at times only one of'the quadratic terms (H2 or T2)
entered the equation (w4-th H and T) due to a low tolerance level on the other
quadratic term.
TABLE 8. CORRELATION COMPARISON FOR TEMPERATURE
Dependent
Variable
HC
CO
NOX
Independent
Variable (s)
T
T, T2
Improvement
T2
T, T*
Improvement
T
T, T2
Improvement
Coefficients of determination (r2) by vehicle
Datsun
0.046
0.131
0.085a
0.003
0.024
0.021
0.055
0.070
0.015
Int'l.
Perkins
0.235
0.295
0.060
0.424
0.471
0.047
0.037
0.051
0.014
Mercedes
0.012
0.115
0.103
0.128
0.131
0.003
0.077
0.099
0.022
Peugeot
0.113
0.165
0.052
0.267
0.268
0.001
0.069
0.079
0.010
Average
0.102
0.176
0.074
0.206
0.224
0.018
0.060
0.075
0.015
a indicates improvement was significant at the 0.05 level
20
-------
Table 9 contains, the correlation comparisons for equation (3) with
equations (1) and (2), and for equations (4) and (5) with equation (3)- Coef-
ficients of determination (r2) are again used as the criterion for determining
the best fits. The average r2 values for a fit linear in H and T were 0.194
for HC, 0.342 for CO, and 0.594 for NOX. For a fit including the cross product
term (HT), the average r2 values were 0.254 for HC, 0.365 for CO, and 0.671 for
NOX. For the quadratic fit, they were 0.272 for HC, 0.378 for CO, and 0.673
for NOX. Except for CO, the linear fit in H and T showed no significant im-
provement over the linear fit in H. This same fit was significantly (p < 0.05)
better than the linear fit in T for NOX on all four vehicles; for CO on the
Peugeot; and for CO and HC on the International-Perkins. The results thus
indicate the strong relationship that exists between NOX and humidity for all
vehicles and between all emissions and ambient data for the International-
Perkins .
TABLE 9. CORRELATION COMPARISON FOR HUMIDITY AND TEMPERATURE
Dependent
Variable
HC
CO
NOV
A
HC
CO
NOX
HC
CO
NOX
Independent
Variable
H,T
Improv. over H
Improv. over T
H,T
Improv. over H
Improv. over T
H,T
Improv. over H
Improv. over T
H,T, HT
Improv. over H,T
H,T, HT
Improv. over H,T
H,T HT
Improv. over H,T
H,T,H2,T2b
Improv. over H,T
H,T,H2,T2b
Improv. over H,T
Improv. over H,T
Coefficient of determination (r ) by vehicle
Datsun
0.047
0.027
0.001
0.090
0.040
0.066
0.601
0.029
0.546a
0.063
0.016
0.095
0.005
0.717
0.116
0.153
0.106a
0.102
0.012
0.787
0 . 186a
Int'l
Perkins
0.588
0.012
0.353a
0.764
0.067a
0.340a
0.523
0.037
0.486a
0.591
0.003
0.781
0.017
0.601
0.078
0.621
0.033
0.793
0.029a
0.604
0.081a
Mercedes
0.013
0.013
0.001
0.144
0.142a
0.016
0.614
0.005
0.537a
0.031
0.018
0.207
0.063
0.713
0.099
0.119
0 . 106a
0.242
0.098a
0.659
0.045a
Peugeot
0.128
0.050
0.015
0.372
0.372a
0.105a
0.638
0.026
0.569a
0.330
0.201
0.378
0.006
0.654
0.016
0.194
0.066
0.376
0.004
0.643
0.005
Average
r2
0.194
0.026
0.093
0.342
0.155
0.132
0.594
0.024
0.535
0.254
0.060
0.365
0.023
0.671
0.077
0.272
0.078
0.378
0.036
0.673
0.079
a indicates significant improvement at 0.05 level
b H and T forced into equation, others entered by significance
21
-------
Including the interaction term (HT) with the linear fit in H and T
yielded an average increase in r2 of 0.060 for HC, 0.023 for CO, and 0.077
for NOX. These increases were significant (p < 0.05) for HC from the Peugeot
and for NOX from the other three vehicles.
The use of a quadratic equation in H and T instead of a linear fit
yielded an average increase in r2 of 0.078 for HC, 0.036 for CO, and 0.079
for NOX. These increases were significant (p < 0.05) for all emissions from
the Mercedes, for CO and NOX from the International-Perkins, and for HC and
NOX from the Datsun. For the NOX emissions, the H2 term was the main vari-
able influencing the significant increases.
With the exception of results for the International-Perkins, low cor-
relations existed between CO or HC and each of the independent variables in
Tables 7, 8, and 9. Both CO and HC emissions from the International-Perkins
were much more sensitive to humidity and temperature than expected, and no
reason is known for this anomaly. In terms of engine design, the Perkins is
not considered to be so grossly different than the others as to cause such
results. One statistic which does set the International-Perkins apart from
the other vehicles, however, is its weight-to-power ratio (all vehicles as-
sumed loaded as light-duty vehicles). This value, in kgAW, is 21.8 for the
International-Perkins, 29.7 for the Datsun, 32.3 for the Mercedes, and 28.9
for the Peugeot. It is not known if the weight/power statistic is related to
the anomaly noted above, but it does indicate that the Perkins engine was prob-
ably operating at a lower fraction of available power than the other engines
(i.e., at lower F/A). In summary, due to the overall relatively low corre-
lations associated with HC or CO and the ambient data, and because of the
unusual results of these emissions in the International-Perkins vehicle, HC
and CO corrections will not be discussed further in the text. The overall
results from Tables 7, 8, and 9 confirm the existence of strong and consistent
associations between NOX and humidity for all four vehicles, and it is this
relationship that will be explored. Equations with H2 and T2 terms have also
been compared to those with an HT interaction term, indicating that the qua-
dratic form is a better predictor.
To examine the NOx-humidity relationships in more detail, scatter plots
have been constructed for each vehicle and are presented as Figures 12 through
15. The linear and quadratic equations in H have been plotted for each vehicle,
and the coefficients of these equations are contained in Table 10. On the
average, addition of the H2 term increased r2 from 0.570 to 0.653, which is
a. substantial improvement. Most of this improvement was made in fitting the
curve for the Datsun 220C, which seemed to have a strong tendency toward
"leveling off" of NOX emissions as humidity increased above about 120 grains
B2O/Ibni dry air. Emissions of NOX from the Datsun, International -Perkins, and
Mercedes all tended to become less sensitive to humidity above some humidity
level. Taking all four vehicles together, however, can still result in com-
putation of a relationship which is useful for most ambient humidity values.
COMPUTATION OF CORRECTION FACTORS
The regression equations generated in the previous subsection (RESULTS
AND STATISTICAL ANALYSIS) were restricted to data on the individual vehicles.
22
-------
1.20
1.10
1.00
0.90
QJ
a
CD
0.80
NOX = 1.161-0.00643H
+ 2.92X 10-5 H2
NOX = 1.039-0.00195H
0.70
0.60
0.50
20 40 60 80 100 120
Humidity (H), grains H2O/lbm dry air
140
160
FIGURE 12. NOX EMISSIONS AS A FUNCTION OF HUMIDITY FOR
A OATSUN 220C DIESEL SEDAN
23
-------
1.20
1.10
1.00
0.90
_o
1:2
^
d>
Q.
CA
n>
O 0.80
represents two runs
NOX= 1.126-0.00428H
+ 1.68X 10-5 H2
NOX = 1.057-0.00173H
0.70
0.60
0.50
20 40 60 80 100 120
Humidity (H), grains HjO/lb,,, dry air
140 160
FIGURE 13. NOX EMISSIONS AS A FUNCTION OF HUMIDITY FOR AN INTERNATIONAL
PICKUP TRUCK WITH PERKINS 6.247 DIESEL ENGINE
24
-------
1.20.
1.10
1.00
0.90
_o
12
i
x 0.80
O
0.70
0.60
0.50
NOX = 1.044-0.00423H
+ 1.46X
NOX = 0.983-0.00199H
CO
o
20 40 60 80 100 120
Humidity (H), grains H2O/lbm dry air
140
160
FIGURE 14. NOX EMISSIONS AS A FUNCTION OF HUMIDITY FOR
A MERCEDES 240D DIESEL SEDAN
25
-------
1.20
1.10
1.00
S
I 0.90
o
09
Q.
(A
CO
0.70
0.60
o o
NOX = 0.859-0.00208H
+ 2.86X 10-6 H2
= 0.848-0.00165H
represents two runs
0.50 ,
20 40 60 80 100 120
Humidity (H), grains H2O/lbm dry air
140 160
FIGURE 15. NOX EMISSIONS AS A FUNCTION OF HUMIDITY FOR
A PEUGEOT S04D DIESEL SEDAN
26
-------
but the desired end result of this project is to obtain a single humidity
correction factor for NOX applicable to the whole class of light-duty, diesel-
powered vehicles.
TABLE 10. COEFFICIENTS OF NOX VERSUS 'HUMIDITY EQUATIONS
Equation
type3
Quadratic
. Vehicle
TH— DoT-lr-i «a
Datsun
IH-Perkins
Mercedes
Peugeot
(ho>
Constant
1r\to
1 nR7
OOP'S
OPAQ
1.161
1.126
1.044
0.859
(bi)
H coefficient
_r\ nm QP
— n nm *?^
— n nm QQ
_ ,n nm AR
-0.00643
-0.00428
-0.00423
-0.00208
7
H^ coefficient
2.92 x 10~5
1.68 x 10~5
1.46 x 10~5
2.86 x 10~6
r2
.572
OAQfi
Ocrto
.ouy
0/:i *>
. D-L
-------
where
and
bi + 2bn (75)
b0 + bi (75) + bn (75)2
(75)
(75)2 •
Several methods were considered for combining data from the four vehicles to
obtain the correction factors, including:
1. Obtaining a humidity correction factor for each vehicle separately
and then averaging the results, i.e., the regression equations for the four
vehicles would have different intercepts and difference slopes.
2. Establishing statistically a commonality of "H" coefficients in the
linear NOx-humidity equations, i.e., determining a common-slope, different-
intercept model for the combined four vehicles.
3. Treating all data as if it were for a single vehicle, i.e., ignoring
any vehicle-to-vehicle differences in magnitude of NOX emissions and generating
a common slope and common intercept, model.
The first method, unequal slopes and intercepts, yielded four regression
equations, relating NOX to H with an average r2 of 0.570, and four equations
relating NOX to H and H2 with an average r2 of 0.653. The resulting coeffi-
cients were presented in Table 10. Notice from Table 7 that for three of the
four vehicles the quadratic fit in H is significantly better than the linear
fit.
Humidity correction factors for NOX were calculated for each vehicle
for both the linear and quadratic equations. The results are given in Table 11
along with the average values of L and Q that would be utilized in establishing
an overall humidity correction factor.
TABLE 11. SEPARATE VEHICLE CORRECTION FACTORS3
Equation Type
and coefficient (s)
Linear
L
Quadratic
L
Q
Datsun
-0.00218
-0.00244
3.46 x 10~5
IH-Perkins
-0.00187
-0.00197
1-.87 x 10-5
Mercedes
-0.00239
-0.00251
1.81 x 10~5
Peugeot
-0.00228
-0.00230
0.40 x 10~5
Average
-0.00218
-0.00230
1.88 x 10~5
Kt, =
1 + L (75) in linear case
K , 1 ^
h 1 + L (H-75) + Q (H-75)2 in quadratic case
28
-------
The second method, common slope and unequal intercepts, produces some
useful results. It is better than the first method in that it requires only
one regression equation, but it also necessitates the separate computation
of four correction factors. The common slope, unequal intercept method shows
that the slopes of the NOx-humidity equations for the four vehicles are close
enough together to be considered equivalent statistically. The hypothesis
that the four slopes are unequal was tested by comparing the mean squared
errors of the regression equations calculated with and without a parallel
line assumption. The resultant F statistic was not significant at the 0.05
level, so the common slope model was not rejected. The combined regression
equation, which had an r of 0.722, is given after Table 12.
TABLE 12. COMMON SLOPE EQUATION COEFFICIENTS AND VALUES OF L
Vehicle
Datsun
Int1 1.— Perkins
Mercedes
Peugeot
Intercept
1.029
1.064
0.972
0.862
Slope
0.00183
0.00183
0.00183
0.00183
Average
L
-0.00205
-0.00197
-0.00219
-0.00252
-0.00218
Equation: NOX = bo + ToiXi H-bj^H, i = 2, 3, 4
0 if "i th" vehicle data not used
where X^ =
if "i th" vehicle date used
Note: The intercept of the Datsun was bo; for the IH-Perkins,
b0 + b2; for the Mercedes, b0 + b3; and for the Peugeot, bo + b4.
Since each equation has a different intercept, the values of L in the
linear K^ factors are different for each vehicle. The average value, -0.00218,
is the same as the average value obtained from Method 1. Due to the difficulty
in employing this method on quadratics, quadratic factors were not obtained.
The third method, common slopes and common intercepts, consisted of a
simple combination of all 174 data points into a single regression equation
relating NO to humidity. This approach was rejected due to the scatter and
poor correlations resulting from combination of all the raw data. Two varia-
tions on this technique were then tried in an attempt to eliminate the adverse
effects of differing NOX magnitudes among the test vehicles.
The first variation consisted of standardizing the observed NOX values
from each vehicle by subtracting the vehicle mean NOX and dividing by the
vehicle NO,, standard deviation as follows:
29
-------
observed NOX - mean NOX
standardized NOX = standard deviation NO*
The standardized values for all four vehicles were then combined and fit to
humidity in a single regression equation. The results for the linear and
quadratic fits are given below:
standardized NO,, = 1.362 - 0.01794 H, r2 = 0.566
J^
standardized NOV = 1.976 - 0.04080 H + 0.000150 H2, r2 = 0.631.
Ji
There was a significant increase (p < 0.0001) in the goodness of fit utilizing
the quadratic equation in H as compared to the linear model.
To obtain the humidity correction factors for NOX (unstandardized), the
values of L and Q were calculated, adjusting for the different means and
standard deviations of each vehicle's data. These values are given in Table
13 along with their averages. The average results are again in good agree-
ment with Methods 1 and 2.
TABLE 13. STANDARDIZED NOX CORRECTION FACTORS
Equation Type
Linear
L
Quadratic
L
Q
Datsun
-0.00218
-0.00229
1.88 x 10~5
IH-Perkins
-0.00200
-0.00210
1.72 x 10~5
Mercedes
-0.00226
-0.00239
1.95 x 10-5
Peugeot
-0.00223
-0.00235
1.92 x 10-5
Average
-0.00217
-0.00228
1.87 x 10~5
The second variation on Method 3 consisted of computing the ratios of
the observed NOX emissions data to the predicted NOX values at an arbitrary
humidity point for each vehicle. These "normalized" data were then combined
to derive a common regression equation. The equations given in Table 10,
both linear and quadratic, were used to compute a predicted value for NOX at
H = 75 for each vehicle. Normalized NOX values were then obtained (separate
for quadratic and linear models) using the following definition:
measured NOX
normalized NOX = predicted NOX at H = 75
Combined equations relating normalized NOX to humidity were then gener-
ated, and the resulting coefficients are given in Table 14. The quadratic
fit yielded a significant (p < 0.001) improvement in correlation over the
linear equation, supporting similar findings from the other methods utilized
in this project. The definition of normalized NOX forces the normalized NOX
values calculated by both the linear and quadratic equations to be 1.0 at the
value H = 75. This second variation, using normalized NOX values, is the
30
-------
only method evaluated which yields a single factor for correction of NOX with-
out averaging values of L and Q. The computed coefficients for the two cases
are
L = -0.00217 for the linear case,
and L = -0.00228, Q = 1.85 x 10~5 for the quadratic case.
TABLE 14. COEFFICIENTS OF NORMALIZED NOX VERSUS HUMIDITY EQUATIONS,
ALL VEHICLES COMBINED
Equation typea
Linear
Quadratic
(bo) constant
11 CO
. J-D J
1.274
(bl) H coefficient
A AAO 1 "7
— U .UUZJ. /
-0.00504
(bn) H2 coefficient
1.84 x 10~5
r2
OR£Q
0.616
a linear form, normalized NOX = bo +
quadratic form, normalized NOX = j^ +
These results are very similar to those obtained using the other methods, as
shown in Table 15.
TABLE 15. SUMMARY OF HUMIDITY CORRECTION FACTOR RESULTS3
Linear Case
Quadratic Case
Method
Unequal slopes
and unequal
intercepts
Equal slopes
but unequal
intercepts
Equal slopes
and equal
intercepts
a) Std. NOX
b) Nor. NO,.
A
Average
r2
0.570
0.722
0.566
0.569
Average
L
-0.00218
-0.00218
-0.00217
-0.00217
Average
r2
0.653
0.631
0.616
Average
L
-0.00230
-0.00228
-0.00228
Average
Q
1.88 x 10~5
1.87 x 10~5
1.85 x 10~5
linear case Kh =
+ L (H.75)
quadratic case Kh = ^ + ^
+ Q (H_?5)2
Correction factors representing the averages of the linear and quadratic re-
sults given in Table 15 are shown in Figure 16.
31
-------
1.3 f-
1.2
o
•S1-1
X
O
o
CO
"- 1.0
o
6
§ 0.9
I
0.8
linear
quadratic
average of
linear
equations
average of
quadratic equations
0.7*-
20 40 60 80 100 120
Humidity (H), grains H2O/lbm dry air
140
160
FIGURE 16. HUMIDITY CORRECTION FACTOR Kh FOR NOX EMISSIONS
AS A FUNCTION OF HUMIDITY, AVERAGE OF
RESULTS GIVEN IN TABLE 15
32
-------
SECTION 6
COMPARISON WITH OTHER CORRECTION FACTORS
Studies were conducted in the past on other classes of vehicles and en-
gines with the aim of determining applicable factors for the correction of
measured emissions to "standard" ambient conditions. The classes studied
were light-duty, gasoline-fueled vehicles^); heavy-duty diesel engines^);
and heavy-duty, gasoline-fueled vehicles. The results of these studies have
appeared in corresponding Federal Emission Regulations^1'3'7'8^ as correction
factors for NOX emissions, other ambient effects having been considered negli-
gible by EPA.
The existing correction factors for NOX are:
light-duty gasoline; Kh = j, . 0.0047 (H_75)
heavy-duty diesel; Kh = i+(0.044 F/A-0.0038)(H-75)+(-0.116 F/A+0.0053)(T-85)
heavy-duty gasoline; Kh = 0.634 + 0.00654H - 0.0000222H2.
In addition, a simpler correction factor was used for heavy-duty diesels
through about mid-1974; and it was ~*
n 1 - 0.0025 (H-75)
*'s *
These factors can be compared to those generated by this project (results of
equal slope/intercept-normalized NOX method shown for example), which are:
based on linear equation; Kh = ± _ ^^ (H_?5)
based on quadratic equation; Kh = nnooo m ?R\ ^ n QC: v 1n-5^ iv
i ~ u• uu^zts \n~~/o) T I.L.OD x J.u / \r.
The most striking comparison which can be made, of course, is that the
light-duty diesel factor based on a linear NOx-humidity relationship is very
similar to the original (and since replaced) factor for heavy-duty diesels.
This light-duty diesel factor shows less sensitivity to humidity than the
light-duty gasoline factor. These relationships are given in tabular form
in Table 11 and in graphical form in Figure 17. The range shown in Figure 17
for the current heavy-duty diesel factor incorporates all the expected variation
33
-------
1.4 r
1.3
1.2
o
X
O
o 1.1
i
ID
1.0
3
0.9
0.8
light-duty
gasoline
original heavy-
duty diesel
light-duty diesel
(linear)
light-duty diesel
(quadratic)
current range for
heavy-duty diesel,
depends on temp, and F/A
0.7
heavy-duty gasoline
20 40 60 80 100 120 140 160
Humidity (H), grains H20/lbm dry air
FIGURE 17. HUMIDITY CORRECTION FACTOR COMPARISON
34
-------
in both test temperature (70°F to 100°F) and F/A ratio (0.005 to 0.07). This
factor, by using F/A as a variable, is restricted to modal steady-state engine
operation.
The heavy-duty gasoline factor and the quadratic-based light-duty diesel
factor are qualitatively similar, with the gasoline factor being a stronger
function of humidity. Application of either the linear-based or the qua-
dratic-based factor computed from data acquired in this task would be a rel-
atively simple matter, and it remains for the sponsor to decide whether or not
the significantly better fit of the quadratic form is sufficient reason to
deviate from the customary linear-based form for the light-duty diesel. Rea-
sons for the differences between correction factors discussed here include
not only type and size of engine, but also the differing duty cycles required
to perform the various test procedures (e.g., difference between heavy- and
light-duty gasoline correction factors).
TABLE 16.
HUMIDITY CORRECTION FACTORS FOR NOX IN TABULAR FORM
imidity (H) ,
jrains H2O/
Lbm/dry air
0
20
40
60
80
100
120
140
160
Light-duty diesel
Linear
0.860
0.893
0.929
0.968
1.011
1.057
1.108
1.164
1.226
Quadratic
0.785
0.847
0.907
0.963
1.011
1.048
1.070
1.076
1.065
Heavy-duty diesel ( 3 » 8 )
Current , range
0.738 to 0.825
0.782 to 0.880
0.831 to 0.942
0.887 to 1.014
0.951 to 1.099
1.024 to 1.198
1.110 to 1.317
1.212 to 1.463
1.334 to 1.645
Original
0.842
0.879
0.920
0.964
1.013
1.067
1.127
1.194
1.270
Light-duty
gasoline (D
0.739
0.795
0.859
0.934
1.024
1.133
1.268
1.440
1.665
Heavy-duty
gasoline (7)
0.634
0.756
0.860
0.946
1.015
1.066
1.099
1.114
1.112
35
-------
REFERENCES
1. Federal Register, Volume 38, No. 151, August 7, 1973.
2. "Diesel Fuel Oils, 1976", Technical Information Center, U.S. Energy Re-
search and Development Administration, November 1976.
3. Federal Register, Volume 37, No. 221 (Subpart J), November 15, 1972.
4. A. Wexler and W. G. Brombacher, "Methods of Measuring Humidity and Test-
ing Hygrometers," National Bureau of Standards Circular 512, September 28,
1951.
5. M. J. Manos, J. W. Bozek, and T. A. Huls, "Effect of Laboratory Ambient
Conditions on Exhaust Emissions." Paper 720124 presented at SAE Meeting,
Detroit, January 10-14, 1972.
6. S. R. Krause, D. F. Merrion, and G. L. Green, "Effect of Inlet Air Humidity
and Temperature on Diesel Exhaust Emissions." Paper 730213 presented at
SAE Meeting, Detroit, January 8-12, 1973.
7. Federal Register, Volume 37, No. 221 (Subpart H), November 15, 1972.
8. Federal Register, Volume 38, No. 124, June 28, 1973.
36
-------
APPENDIX A
TABULAR DATA
37
-------
TABLE A-l. TABULAR DATA BY RUN
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Date
12/12/75
12/12/75
12/12/75
12/15/75
12/15/75
12/15/75
12/15/75
12/16/75
12/16/75
12/16/75
12/16/75
12/17/75
12/17/75
12/17/75
12/17/75
12/18/75
12/18/75
12/18/75
12/18/75
1/7/76
1/7/76
1/7/76
1/7/76
1/8/76
1/8/76
1/8/76
1/8/76
1/9/76
1/9/76
1/9/76
1/9/76
1/13/76
1/13/76
1/13/76
1/13/76
1/14/76
1/14/76
1/14/76
1/14/76
1/16/76
Vehicle
I . H . -Perkins
Mercedes
Datsun
Mercedes
Datsun
Peugeot
I. H. -Perkins
Mercedes
Datsun
Peugeot
I. H. -Perkins
Mercedes
Datsun
Peugeot
I. H. -Perkins
Mercedes
Datsun
Peugeot
I . H . -Perkins
Mercedes
Datsun
Peugeot
I . H . -Perkins
I . H . -Perkins
Peugeot
Datsun
Mercedes
Mercedes
Datsun
I. H. -Perkins
Peugeot
I. H. -Perkins
Datsun
Peugeot
Mercedes
Mercedes
I. H. -Perkins
Peugeot
Datsun
Datsun
Emissions, grams/km
HC
0.47
0.18
0.17
0.13
0.16
0.34
0.62
0.14
0.12
0.44
0.59
0.11
0.08
0.42
0.63
0.15
0.14
0.37
0.58
0.10
0.15
0.39
0.54
0.54
0.34
0.15
0.19
0.49
0.59
0.18
0.51
0.23
0.17
0.55
0.39
0.15
0.07
CO
1.95
0.63
0.80
0.63
0.75
1.01
1.46
0.65
0.76
1.03
2.01
0.61
0.77
1.09
1.75
0.61
0.69
0.95
1.71
0.69
0.74
1.01
1.91
1.96
1.10
0.67
r\a-(- a
0.83
1.87
r\a J-«
2.12
0.84
1.00
0.70
0.69
1.94
1.03
0.86
0.81
NOX
0.90
0.77
0.81
0.90
0.96
0.70
0.91
0.99
0.96
0.72
0.96
0.85
0.94
0.71
0.89
0.90
0.99
0.84
1.00
0.99
1.11
0.85
1.10
1.15
0.87
Discar
1.04
Discart
1.09
1.03
Discar
0.77
0.77
0.67
0.91
0.93
1.07
0.76
1.01
1.03
CO2
244.86
238.15
222.85
248.44
231.17
229.47
232.97
261.35
231.24
234.61
247.37
235.48
228.91
232.56
219.50
242.40
218.01
230.91
232.55
255.01
250.82
255.42
258.16
258.77
257.67
237.25
irfl
233.43
245.01
j_ j
246.72
239.08
240.72
256.44
256.81
259.60
235.07
242.44
241.13
Fuel,
km/&
10.8
11.2
12.0
10.8
11.5
11.6
11.3
10.2
11.5
11.3
10.7
11.4
11.7
11.4
12.0
11.0
12.2
11.5
11.3
10.5
10.6
10.4
10.2
10.2
10.3
11.3
11.4
10.8
10.7
11.2
11.0
10.4
10.4
10.2
11.3
11.0
11.1
gr H20/
lbm air
72.4
74.1
74.9
50.4
44.6
40.9
41.6
35.8
34.3
35.2
34.3
45.9
42.4
41.3
35.1
11.6
10.1
10.4
10.7
15.0
13.2
13.0
12.9
8.3
9.1
9.8
17.1
20.1
72.2
73.7
70.8
63,9
24.6
24.0
21.8
21.5
33.9
Temp . ,
oF
76.3
77.6
78.4
73.1
72.5
71.2
70.3
75.8
75.2
77.8
77.1
70.2
67.3
68.0
68.0
76.9
76.4
76.0
79.0
67.1
67.1
67.2
67.4
66.8
67.4
68.4
77.8
76.9
77.7
77.4
77.2
77.2
68.1
67.7
68.0
67.6
77.2
pa,
in Hg
29.18
29.08
29.08
29.28
29.37
29.37
29.35
29.28
29.28
29.27
29.24
29.28
29.29
29.30
29.27
29.78
29.78
29.78
29.78
29.40
29.40
29.39
29.39
29.69
29.69
29.60
29.51
29.48
29.09
29.14
29.12
29.09
29.60
29.60
29.60
29.50
29.36
(continued)
38
-------
TABLE A-l (continued)
Run
No.
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Date
1/16/76
1/16/76
1/16/76
1/19/76
1/19/76
1/19/76
1/19/76
1/20/76
1/20/76
1/20/76
1/20/76
1/21/76
1/21/76
1/21/76
1/21/76
1/22/76
1/22/76
1/22/76
1/22/76
1/23/76
2/11/76
2/11/76
2/11/76
2/11/76
2/13/76
2/13/76
2/13/76
2/13/76
2/16/76
2/16/76
2/16/76
2/16/76
2/17/76
2/17/76
2/17/76
2/17/76
2/26/76
2/26/76
2/26/76
2/26/76
Vehicle
Mercedes
I. H. -Perkins
Peugeot
Datsun
I . H . -Perkins
Mercedes
Peugeot
I . H . -Perkins
Mercedes
Datsun
Peugeot
Datsun
Peugeot
I. H. -Perkins
Mercedes
Peugeot
Datsun
I . H . -Perkins
Mercedes
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
I . H . -Perkins
Datsun
Mercedes
Datsun
Mercedes
I. H. -Perkins
Peugeot
Datsun
I. H. -Perkins
Peugeot
Mercedes
I . H . -Perkins
Mercedes
Datsun
Peugeot
Emissions , grams/km
HC
0.13
0.50
0.42
0.14
0.50
0.15
0.39
0.52
0.12
0.12
0.40
0.16
0.43
0.51
0.48
0.07
0.59
0.12
0.13
0.44
0.15
0.60
0.15
0.48
0.60
0.13
0.14
0.12
0.63
0.42
0.16
0.75
0.35
0.10
0.58
0.08
CO
0.69
2.04
0.99
0.83
2.13
0.66
1.04
1.99
0.66
0.82
1.07
0.80
0.95
1.98
T-\~J_-.
0.88
0.77
1.93
0.61
0.59
1.06
0.83
2.11
0.69
1.04
2.13
0.79
0.64
0.76
in T^Tf-n
2.20
1.02
0.92
2.79
1.04
0.67
2.16
0.68
r\-x j_
n-i-i-
NOX
1.02
1.07
0.86
0.96
0.94
0.94
0.77
1.10
0.97
1.12
0.93
1.01
0.79
1.10
Discar
0.81
1.10
1.10
0.95
0.95
0.67
0.86
0.91
0.84
0.71
0.87
0.85
0.82
0.70
Discar
0.78
0.63
0.79
0.82
0.79
0.95
0.95
0.98
i Discs
a Disca
i
C02 .
261.14
252.37
237.09
250.13
260.42
266.22
249.75
273.05
257.48
252.65
257.28
228.68
233.46
259.26
flrfl
227.68
229.04
266.63
238.41
247.04
251.86
259.23
267.66
256.56
244.35
264.66
235.51
250.85
236.62
rlr~fl
260.02
232.27
253.55
279.71
247.52
265.76
265.50
246.98
Fuel,
km/5,
10.2
10.5
11.2
10.7
10.1
10.0
10.6
9.7
10.4
10.6
10.3
11.7
11.4
10.2
11.7
11.7
9.9
11.2
10.8
10.5
10.3
9.9
10.4
10.9
10.0
11.3
10.7
11.3
10.1
11.4
10.5
9.4
10.7
10.1
9.9
10.8
gr H20/
lbm air
26.3
23.2
21.6
47.8
47.4
44.5
44.1
31.5
29.1
21.9
19.2
13.1
11.3
10.6
11.6
11.1
11.2
10.9
22.3
89.6
78.4
73.3
74.8
62.8
75.4
69.1
68.4
96.5
96.6
98.1
108.7
107.4
63.1
41.1
54.9
57.7
Temp . ,
OF
77.2
76.8
77.6
86.6
86.6
86.7
86.2
76.9
77.7
78.1
76,9
86.5
86.7
86.3
87.7
88.1
86.0
87.6
87.3
77.3
78.1
76.3
78.3
68.9
77.9
76.7
78.2
78.7
78.7
78.7
87.9
86.8
85.3
87.2
70.9
69.0
pa,
in Hg
29.38
29.38
29.38
29.35
29.35
29.35
29.34
29.63
29.65
29.63
29.59
29.65
29.60
29.56
29.53
29.65
29.48
29.41
29.20
29.44
29.43
29.40
29.38
29.27
29.28
29.28
29.28
29.08
29.29
29.25
28.98
29.03
28.94
28.88
29.37
29.35
(continued)
39
-------
TABLE A-l (continued)
Run
No.
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
Date
2/27/76
2/27/76
*•/ £• 1 / 1 \J
2/27/76
2/27/76
3/1/76
3/1/76
3/1/76
3/1/76
3/2/76
3/2/76
3/2/76
3/2/76
3/3/76
3/3/76
3/3/76
3/3/76
3/4/76
3/4/76
3/4/76
3/4/76
3/8/76
3/11/76
3/11/76
3/11/76
3/12/76
3/12/76
3/12/76
3/12/76
3/19/76
3/19/76
3/19/76
3/19/76
3/22/76
3/22/76
3/22/76
3/22/76
3/23/76
3/23/76
3/23/76
3/23/76
Vehicle
I. H. -Perkins
Wt^yf of^oc
i. 1C^ (_»C?UC A
Datsun
Peugeot
Mercedes
Datsun
I. H. -Perkins
Peugeot
Datsun
Peugeot
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Mercedes
Peugeot
Datsun
I. H. -Perkins
Datsun
I. H. -Perkins
Mercedes
Peugeot
Peugeot
Mercedes
I. H. -Perkins
Datsun
Datsun
Peugeot
I. H. -Perkins
Mercedes
Peugeot
Datsun
Mercedes
I. H. -Perkins
Mercedes
Peugeot
I. H. -Perkins
Datsun
Emissions , grams/km
HC
0.65
0.20
0.14
0.13
0.71
0.49
0.14
0.41
0.68
0.11
0.47
0.15
0.67
0.08
0.11
0.44
0.20
0.72
0.15
0.64
0.12
0.40
0.49
0.39
0.77
0.18
0.21
0.30
0.71
0.13
0.36
0.19
0.12
0.75
0.12
0.29
0.65
0.29
CO
1.97
0.87
Data
0.64
0.92
2.52
1.19
0.83
1.09
2.29
0.61
1.03
0.79
2.07
0.62
0.64
1.06
0.84
1.97
0.86
1.82
0.67
1.03
1.03
0.66
2.27
0.71
0.80
0.94
2.12
0.60
0.92
0.77
0.59
2.14
0.59
0.83
2.15
0.79
NOX
0.96
C02
256.82
JL/XO^rdX.
0.95
Discar
0.78
0.85
0.87
0,66
0.79
0.63
0.84
0.64
0.64
0.77
0.77
0,64
0.62
0.61
0.76
0.83
0.79
0.85
0.85
0.70
0.63
0.79
0.86
0.77
0.82
0.79
0.83
0.74
0.83
1.01
0.90
1.01
0.81
0.83
1.00
0.94
236.82
(3r<3
238.18
229.33
251.94
241.70
231.78
238.19
254.02
212.53
231.95
228.34
252.52
233.37
225.43
251.63
236.91
273.91
230.62
236.79
259.20
238.32
230.91
251.34
255.72
224.62
246.41
247.13
238.37
238.35
229.84
238.24
236.94
253.96
227.24
238.23
258.82
235.13
Fuel,
km/A
10.3
11.3
11.2
11.6
10.4
11.0
11.5
11.1
10.3
12.6
11.4
11.7
10.4
11.5
11.9
10.6
11.3
9.6
11.6
11.1
10.3
11.1
11.5
10.6
10.3
11.9
10.8
10.8
11.0
11.2
11.6
11.2
11.3
10.4
11.8
11.2
10.2
11.3
gr H20/
Ibm air
58.2
61.4
92.0
91.3
90.3
92.0
96.8
95.7
95.0
97.6
98.3
95i9
94.6
96.0
95.0
88.6
83.6
74.5
75.9
69.4
73.3
75.4
100.8
104; 7
105.9
108.7
78.6
75.1
71.2
70.1
34.7
33.9
36.4
36.3
50.2
50.6
52.7
49.2
Tenrp. ,
op
68.9
68.3
71.1
70.6
70.1
70.9
71.0
70.7
70.2
70.6
76.5
76.1
76.3
76.7
76.8
79.1
78.3
77.3
69.4
68.4
68.3
68.0
77.1
77.8
75.1
75.4
85.7
85.2
85.8
86.1
76.8
77.3
86.5
87.3
75.4
75.8
77.7
77.3
pa,
in Hg
29.35
29.35
29.03
29.03
29.02
29.01
29.05
29.07
29.05
29.04
29.06
29.04
29.04
29.02
28.94
28.80
28.87
28.83
29.07
29.14
29.13
29.09
28.96
28.99
28.98
28.99
29.02
29.02
29.02
28.96
29.41
29.43
29.41
29.37
29.44
29.42
29.43
29.43
(continued)
40
-------
TABLE A-l (continued)
Run
No.
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
Date
3/24/76
3/24/76
3/24/76
3/24/76
3/29/76
3/29/76
3/29/76
4/6/76
4/6/76
4/6/76
4/6/76
4/14/76
4/14/76 •
4/14/76
4/14/76
5/5/76
5/5/76
5/5/76
5/5/76
5/6/76
5/6/76
5/6/76
5/6/76
5/25/76
5/25/76
5/25/76
5/25/76
5/27/76
5/27/76
5/27/76
5/27/76
5/28/76
5/28/76
5/28/76
5/28/76
5/31/76
5/31/76
5/31/76
5/31/76
6/1/76
Vehicle
Peugeot
Mercedes
I. H. -Perkins
Da t sun
Datsun
Peugeot
Mercedes
I. H. -Perkins
Datsun
Peugeot
Mercedes
I. H. -Perkins
Mercedes
Datsun
Peugeot
I . H . -Perkins
Mercedes
Datsun
Peugeot
Mercedes
Datsun
Peugeot
I . H . -Perkins
Datsun
Mercedes
I . H . -Perkins
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
I. H. -Perkins
Mercedes
Datsun
Peugeot
Datsun
Peugeot
I. H. -Perkins
Mercedes
Peugeot
Emissions, grams/km
HC
0.38
0.15
0.75
0.14
0.16
0.40
0.11
0.67
0.11
0.34
0.15
0.66
0.13
0.12
0.30
0.69
0.14
0.04
0.31
0.13
0.09
0.34
0.76
0.15
0.10
0.83
0.26
0.17
0.76
0.11
0.25
0.62
0.13
0.06
0.33
0.09
0.09
0.64
0.05
0.28
CO
0.86
0.63
2.49
0.81
0.80
0.97
0.62
2.27
0.65
0.95
0.50
2.59
0.59
0.80
1.00
2.41
0.63
0.79
0.92
0.61
0.77
0.94
2.55
0.86
0.71
2.83
0.98
0.76
2.40
0.61
0.91
2.31
0.62
0.74
0.94
0.80
1.00
2.60
0.63
0.98
NOX
0.75
0.81
0.85
0.86
0.79
0.69
0.89
0.98
0.81
0.77
0.78
0.97
0.78
0.87
0.65
1.01
0.84
0.86
0.70
0.80
0.82
0.72
0.92
0.91
0.80
0.96
0.69
0.89
0.93
0.78
0.65
0.97
0.81
0.87
0.71
0.85
0.63
0.78
0.72
0.65
CO2
240.96
252.13
254.38
242.65
246.31
233.01
249.12
263.84
226.82
239.27
227.21
278.80
253.08
256.75
226.82
274.62
270.13
250.30
235.26
268.73
253.11
238.22
270.55
272.27
268.54
286.52
246.39
248.63
271.77
267.83
225.99
260.27
248.16
240.38
235.80
243.03
243.00
254.23
253.24
237.69
Fuel,
km/ Jl
11.0
10.6
10.3
11.0
10.8
11.4
10.7
10.0
11.8
11.1
11.8
9.4
10.6
10.4
11.7
9.6
9.9
10.7
11.3
10.0
10.6
11.2
9.7
9.8
10.0
9.2
10.8
10.7
9.7
10.0
11.8
10.1
10.8
11.1
11.3
11.0
11.0
10.3
10.6
11.2
gr H20/
Ibm air
69.9
76.3
78.7
75.7
128.8
94.3
21.1
83.1
82.4
80.8
81.3
120.7
122.7
122.6
121.1
118.1
120.6
124.7
127.4
119.7
120.0
115.7
114.6
139.5
139.6
135.2
135.4
105.4
106.8
108.5
105.1
90.3
92.1
90.7
95.2
142.2
145.8
147,2
132.8
139.3
Temp. ,
op
87.1
86.9
86.8
86.4
86.8
87.0
86.9
87.8
87.1
85.8
87.4
86.8
86.7
86.2
86.4
86.4
87.1
87.2
86.0
87.3
87.5
86.3
86.8
87.7
88.0
87.0
88.1
86.1
86.9
87.4
86.6
84.8
87.4
87.0
88.2
87.1
87.1
88.6
87.3
85.7
pa,
in Hg
29.19
29.22
29.09
29.06
28.96
28.94
28.85
29.12
29.12
29.10
29.08
29.15
29.15
29.15
29.14
29.08
29.09
29.09
29.04
29.08
29.09
29.11
29.12
28.99
28.97
28.95
28.91
29.21
29.21
29.21
29.23
29.17
29.18
29.16
29.14
28.99
28.97
28.95
28.94
29.00
(continued)
41
-------
TABLE A-l (continued)
Run
No.
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
Date
6/1/76
6/1/76
6/1/76
6/2/76
6/2/76
6/2/76
6/2/76
6/3/76
6/3/76
6/3/76
6/3/76
6/4/76
6/4/76
6/4/76
6/4/76
6/4/76
6/14/76
6/14/76
6/14/76
6/15/76
6/15/76
6/15/76
6/15/76
Vehicle
Datsun
I . H . -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
I. H. -Perkins
Datsun
Mercedes
Datsun
Mercedes
I. H. -Perkins
Peugeot
Datsun
I. H. -Perkins
Peugeot
Mercedes
I. H. -Perkins
Mercedes
Datsun
Peugeot
Emissions , grams/km
HC
0.12
0.79
0.14
0.47
0.12
1.00
0.10
0.35
0.80
0.16
0.08 .
0.18
0.12
0.78
0.41
0.13
0.90
0.40
0.10
0.91
0.17
0.15
0.43
CO
0.83
2.77
0.61
1.01
0.83
3.03
0.67
1.00
2.54
0.78
0.61
0.83
0.60
2.68
0.96
0.83
2.93
1.03
0.66
2.77
0.67
0.85
0.56
NOX
0.85
0.82
0.69
0.56
0.70
0.76
0.70
0.67
0..86
0.82
0.78
0.84
0.80
0.94
0.71
0.84
0.86
0.63
0.75
0.81
0.76
0.85
0.56
C02
248.71
269.05
252.23
252.85
262.95
270.22
268.86
240.36
264.11
245.21
259.77
248.57
257.21
263.48
241.12
258.02
278.67
246.72
271.50
276.41
281.50
268.25
245.40
Fuel,
km/A
10.7
9.7
10.6
10.5
10.2
9.7
9.9
11.1
9.9
10.9
10.3
10.7
10.4
10.0
11.0
10.4
9.4
10.8
9.9
9.5
9.5
10.0
10.8
gr H20/
Ibm air
136.5
137.6
140.0
146.5
138.9
134.9
137.9
95.9
90.4
89.7
86.7
92.6
94.2
88.7
77.3
147.6
140.5
130.7
136.1
158.1
150.9
141.4
147.5
Temp. ,
op
85.0
86.1
86.6
87.8
85.9
86.4
86.5
87.4
87.4
86.7
86.3
85.3
87.7
87.4
85.1
87.3
87.5
85.8
88.3
87.0
88.2
86.3
87.1
pa,
in Hg
29.09
29.16
29.15
29.14
29.16
29.16
29.16
29.15
29.15
29.15
29.15
29.18
29.12
29.13
29.13
29.10
29.07
29.07
29.05
29.06
29.06
29.06
29.07
42
-------
APPENDIX B
COEFFICIENTS OF EQUATIONS RELATING DEPENDENT
AND INDEPENDENT VARIABLES
43
-------
TABLE B-l. COEFFICIENTS OF HYDROCARBON (HC) EQUATIONS
"Independent"
variable (s)
H
T
H,H2
T,T2
H,T
H,H2,T,T2
H,T, HT
Vehicle
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I . H . -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I . H . -Perkins
Mercedes
Peugeot
Datsun
I . H . -Perkins
Mercedes
Peugeot
Constant
0.155
0.503
0.134
0.421
0.255
0.037
0.193
0.679
0.126
0.529
0.120
0.381
-1.821
4.537
-2.505
-2.219
0.246
0.351
0.200
0.634
-1.577
3.570
-2.569
-1.711
0.124
0.466
0.074
-0.022
Coefficient by "independent" variable
H
-1.53E-4
2.16E-3
-1.46E-5
-5.26E-4
9.14E-4
191 T?_Q
5.10E-4
1.02E-3
— A Q*7T7_£
— t . y ici~ D
1QTI?_'J
. y ICi—J
5 . 36E-5
-2 . 73E-4
6.37E-4
1.71E-3
_a
6.93E-4
1.85E-3
4.57E-5
2.20E-3
1.08E-2
T
11Q1?_7
7.86E-3
-7.51E-4
1 "71T?_'J
5.18E-2
— 1 HOT? — 1
6.82E-2
7rtQ7?_O
. uy£i""^i
-1.24E-3
2.09E-3
-8.85E-4
-2.89E-3
4.48E-2
-8.06E-2
7.00E-2
5.66E-2
2.76E-4
6.66E-4
6.53E-4
5.27E-3
H^
-6.95E-6
6.22E-6
-3.43E-6
— i n^p— R
•
-4.40E-6
1.29E-6
5.51E-7
-6.39E-6
____*. __
T;
-3 . 38E-4
7.39E-4
-4.37E-4
-4.76E-4
-2.91E-4
5.28E-4
-4.50E-4
-3.79E-4
HT
-2.31E-5
2.33E-5
2^"7T? — ^
-1.34E-4
r2
0.020
0.576
0.000
0.078
0.046
0.235
0.012
0.113
0.085
0.584
0.012
0.127
0.031
0.295
0.115
0.165
0.047
0.588
0.013
0.128
0.153
0.621
0.119
0.194
0.063
0.591
0.031
0.330
F-level (significance) insufficient for inclusion
-------
TABLE B-2. COEFFICIENTS OF CARBON MONOXIDE (CO) EQUATIONS
"Independent"
variable (s)
H
T
H,H2
T,T2
H,T
H,H2,T,T2
H,T, HT
Vehicle
Datsun
I . H . -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Constant
0.779
1.708
0.639
0.996
0.838
-0.281
0.795
1.385
0.777
1.843
0.663
0.990
2.029
11.693
1.132
1.775
0.910
0.638
0.812
1.485
1.815
7.954
0.876
1.501
0.828
1.506
1.000
1.391
Coefficient by "independent" variable
H
2.84E-4
7.08E-3
-4.23E-5
2.20E-5
3.71E-4
2.07E-3
-9.15E-4
2.64E-4
4.32E-4
5.77E-3
1.36E-4
6.03E-4
4.12E-4
3.34E-3
-1.05E-3
2.77E-4
1.70E-3
-8.74E-3
-3.05E-3
2.18E-3
T
3.15E-2
—A RAT? — T
-3.10E-2
-2.77E-1
-1.06E-2
-1.49E-2
-1.77E-3
1.46E-2
-2.31E-3
-6.65E-3
-2.50E-2
-1.71E-1
-2.71E-3
-6.75E-3
-7.60E-4
3.86E-3
-4.59E-3
-5.48E-3
H^
•
-5.66E-7
3.29E-5
5 . 70E-6
-1.61E-6
a
1.55E-5
7.97E-6
2.22E-6
T2
1 . 94E-4
1.97E-3
5.46E-5
6.41E-5
1.48E-4
1.18E-3
a
a
HT
-1.53E-5
1.76E-4
3.82E-5
-1.92E-5
r2
0.050
0.697
0.002
0.000
0.003
0.424
0.128
0.267
0.051
0.724
0.054
0.002
0.024
0.471
0.131
0.268
0.090
0.764
0.144
0.372
0.102
0.793
0.242
0.376
0.095
0.781
0.207
0.378
F-level (significance) insufficient for inclusion
-------
TABLE B-3. COEFFICIENTS OF NITROGEN OXIDES (NOX) EQUATIONS
cr>
"Independent"
variable (s)
H
T
H,H2
T,T
H,T
H,H2,T,T2
H,T, HT
Vehicle
Da t sun
I .H. -Perkins
Mercedes
Peugeot
Datsun
I .H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Datsun
I. H. -Perkins
Mercedes
Peugeot
Constant
1.039
1.057
0.983
0.848
1.180
1.145
1.160
0.982
1.161
1.126
1.044
0.859
3.247
3.078
3.797
-0.457
0.812
0.824
0.895
0.676
2.998
3.065
1.023
-0.233
1.612
1.364
1.512
0.885
Coefficient by "independent" variable
H
-1.95E-3
-1.73E-3
-1.99E-3
-1.65E-3
-6.43E-3
-4.28E-3
-4.23E-3
-2.08E-3
-2.21E-3
-2.01E-3
-2.08E-3
-1.85E-3
-6.12E-3
-3.93E-3
-4.20E-3
-2.22E-3
-1.46E-2
-1.10E-2
-1.26E-2
-3.38E-3
T
•5 COI7_ ~i '
-2.72E-3
-4.01E-3
-3.24E-3
-5.67E-2
-5.25E-2
-7.14E-2
3.38E-2
3.08E-3
3.19E-3
1.18E-3
2.34E-3
-4.84E-2
-5.26E-2
a
2.61E-2
-6.82E-3
-3.52E-3
-6.35E-3
-2.64E-4
H^
2Q9T? ^
1£QT? £
. OOEi D
1.46E-5
2.86E-6
2.62E-5
1.27E-5
1.42E-5
2.57E-6
T2
3.37E-4
3.17E-4
4.27E-4
-2.37E-4
3.16E-4
3.52E-4
3.41E-6
-1.53E-4
HT
1.50E-4
1.09E-4
1.26E-4
4.28E-5
r2
0.572
0.486
0.609
0.612
0.055
0.037
0.077
0.069
0.772
0.568
0.658
0.615
0.070
0.051
0.099
0.079
0.601
0.523
0.614
0.638
0.787
0.604
0.659
0.643
0.717
0.601
0.713
0.654
F-level (significance) insufficient for inclusion
-------
APPENDIX C
STEPWISE MULTIPLE REGRESSION COMPUTER OUTPUTS
47
-------
7b/07/?7.
PACE
00
FILF HARE (CHF.ATION OA1F. = 7h/07/?7.)
3UHFH.F DATSUN
UEPENUFNT VARIABLE.. HC
VAHIAHLfcO) ENTERED UN STtP N'.IMHF.H
3..
* • MIILTIFIE REGRESSION
H80
MULTIPLE H
R SQUAHt
ST0 DEVIATION
VARIABLE
TF.MP
GEH
TSO
H30
(CONSTANT) -1
.311711 AN4LVS1S
.15313 HHIiKFSSIl
OF VAHIANCE
)N
.111311 RF.SIl'HAL
H 3TO ERPOR fl f
•»«t77B17ht-01 .J'flJlS^^F-Ol
bS'UStsE-OI ,t.5bBSM3llF.-03
e«11lf.l?nE-ll3 .l77bl3BbE-03
13'I71713E-OS ,H30'»775bE-OR
,S7?37Bb 1.UBQ3570
SIGNIFICANCE
i>.S<«IISl1»
.111.
.RHI1831H3
.33H
8. 11878825
. 10*1
i.ntinbes
.311
2.1317551
.15?
OF SUM OF SUUARES MEAN SQUARE
1. .013)1 .110333
31. .0735R .U01HH
BETA VARIABLE PAMIAL
ELASTICITY
b.HSBHPSl
2b. 11275
.SlnlOb?
.3S113
-7.oai72?1
-13.21bbl
-.btbSObl
-.21311
SIGNIFICANCE
ALL VARIABLES ARE IN THE EQUATION.
VAHIANcE/cOVARIANCF. MATRIX OF THE UNNORMALIZFTO REGRESSION COEFFICIENTS.
TEMP
GEH
TSO
HSO
.00077
-.00(100
-.ooono
.00000
TEHP
.onouo
.00000 .0(1000
-.noouo -.00000
GEH
TSO
.oonoo
HSQ
TOLtHANCE f
SIGN1F1CANC.F.
-------
?b/Q7/2?.
PAGE as
FILE HARE (CREATION DATE = 7b/07/27.)
SUBFILE DATSUN
DEPENDENT VARIABLE.. HC
MULTIPLE REGRESSION
STEP VARIABLE
ENTERED REMOVED
i TEMP
CEH
e TSU
3 HSO
TO
OR REMOVE
1.15711
.obsua
SIGNIFICANCE
.388
,7Sb
.057
MULTIPLE «
.81330
I3bl?0
.3S17D
R SQUARE
.0*550
. 0*708
.13083
.153*3
R SQUARE
CHANGE
.QH5SO
.00158
.08375
.oaabo
SIMPLE R
-.21330
-.1H207
-!aoioi
OVERALL f SIGNIFICANCE
1.0127b ,3Tt
VO
-------
7b/07/?7.
PAGt 31
FILE HARE . (CREATION DATF = 7b/07/?7.)
SUBFILE OATSUN
MULTIPLE REGRESSION
OfcPF.NOENl VARIABLE.. CD
VARIAHIE(S) ENTERED ON STEP NIIMHEK
c!..
TSO
Mill TIPLE R
R SdllARE
310 OF VI A 1
VAHIAHLE
TEMP
GEH
TSO
(CONSTANT)
,311lb ANALYSIS OF VARIANCE
.lOIHt. HEGHfSSlUN
IDN .05I7<< RESIDUAL
B STU EHROR B F
SIGN1F ICANCE
-,2f qSMbRE-Ul .321bBS57E-dl .h0371bl>5
.112
,'»)23tS22E-03 . 22 1 8 72hlF-U 3 3.473'537n
.070
.l*ROI)03>»E-n3 ,?H'«77713F-03 .5223*70?
.171
1.R153172 1.2570737 2.0853721
OF SUM OF SQUARES MF.AN SQUARE
3. .01217 .OU»Uh
10. .10721 .(IDPbH
BETA VARIABLE PARTIAL
ELASTIC11Y
-3.2l-S717b HSU .0010H
-2.502f 9
.32SH779
.010S2
3.0f4S3b3
l.H77b
F SIGNIFICANCE
1.M21H .?2b
TOLERANCE f
SIUN1FILANCF
F-LEVEL OR TOLERANCE-LEVEL INSUFFICIENT FOR FURTHER COMPUTATION.
VAHIANCE/COVARlANCE MATRIX OF THE UNNOHMALI/EO REGRESSION COEFFICIENTS.
TEHp
GF.H
TSQ
.00103
.00000 .onoon
-.oonui -.nunnu
TEMP
GEH
T3U
-------
76/07/27.
PAGE 32
FILE HARE (CREATION DATE = 7b/07/27.)
SUBFILE DATSUN
DEPENDENT VARIABLE.. CO
MULTIPLE REGRESSION
STEP
VARIABLE
ENTERED REMOVED
1 TEMP
GEH
a TSO
SUMMARY TABLE
F TO SIGNIFICANCE
ENTER OR REMOVE
SIMPLE R
1.78832
3.S075S
.52235
.IBS
.055
MULTIPLE R R SQUARE R SUUAHE
CHANGE
.OStJtb .003*2 .00312
.30022 .01013 .U8b?2
.31
-------
PAGE
to
f II.E HARE (CREATION DATF = 7h/n7/?7.)
8USFILF OATSUN
MULTIPLE REGHESSION «
VARIABLE.. Nn»
VAHIABUECS) F.NTEMEU ON STEP NIJMBEH 3.. 'SO
HIJLTIPU a
3JD DEVIATION
.8B710
.7Rf,MH
.05211
HESIOllAL
VAHIArtlE
TEMP
GEH
TSO
(CONSTANT)
VARIABLES IN THE EWIATtON
B STI) EHROW B
-.•»8H<»B7e7E-(ll
OF VARIANCE
IN
F
SIGNIFICANCE
.155
(t
.nno
.Kb
OF SUM OF SUUARF.3 MEAN SHUARE F SIGNIFICANCE
<», .SMll1* .OH78U 3b.f>ia>v» .0(10
31. .ins«u .nu??e
BETA VARIABLE PARTIAL TOLtHANCE F
ELASTICITY SIGNIFICANCE
"*li?S5is
" '"M"S
3.1B75115
2.3177S
ALL VARIABLES ARE IN THE EOOATION.
VARIANCE/COVARIANCE MATRIX uF THE UNNORMALIZED REGRESSION
TEMP
GEH
TSO
HSQ
.00111
-.oooni
-.00001
.onoon
.00000
.00000
-.uonoo
.00000
-.ooono
.nnnon
TEMP
GEH
T8Q
H80
-------
7b/07/27.
PAGE 3b
U1
U)
FILE HARE (CREATION DATE = 7b/07/27.)
SUBFILE OATSUN
DEPENDENT VARIABLE.. NOX
STEP VARIABLE
ENTERED REMOVED
I TEHP
6EH
2 HSO
3 TSQ
F TO
EN TEH OR REMOVE
2.17058
Sb.leBbl
30.8Sb7b
£.20051
3 U M f A R Y
.o « t s a i
B L E
R SUUARE
.055*7
.b0130
,7?»'»2
,?8fat*
R SQUARE
CHANGE
. 055*7
.5*582
.I73b3
.01202
SIMPLE R
-.23553
-,75bSB
-,b2138
-.23102
OVERALL F SIGNIFICANCE
30. lib** .000
*S,'»OS71 .000
3b,0127S .000
-------
7h/D7/27.
PAGE 10
FILE HARt (CREATION DATE = 7b/l)7/?7.)
SUHFILE PERKINS
MULTIPLE REGRESSION
OF.PENDFNT VARIABLE.. HC
VARIArtLEO) FNTERF.O ON STEP NUMHFR
3..
MULTIPLE «
M SQUARE
3TO DEVIATION
.7R780
,b?l)b3
ANALYSIS OF VARIANCE OF
REGRESSION H.
RESiniJAL »0.
SUM OF SUUAHES
.3RSH7
MEAN SQUARE
.(IHh3?
.OOSBI
F SIGNIFICANCE
Ib.SSSM .HUH
VARIABLES IN THE EQUATION
VARIABLE
TEMP
GEH
TSO
HSO
(CONSTANT) 3.Sb18Blfl
SID ERROR B
.SSH37S31E-03 .3115Bnh7E-03
1.MRI1H17B
F BETA
SIGNIFICANCE ELASTICITY
. ins
11b>»
.11R
.ass
3.bl)3B10<*
.ObS
,bn?B778
S.llltHOS
S.lUBBb
,071C1SC»S
VARIABLES NOT IN THE EQUATION
VARIABLE PARTIAL TOLtKANCE F
SIUN1FICANCE
ALL VARIABLES ARE IN THE EQUATION.
VARIANCE/COVARIANCE MATRIX OF THE UNNORMALI2ED REGRESSION COEFFICIENTS.
TEHP
GEH
TSU
HSU
.00237
-.00001
-.00002
.noooo
.00000
.00000 .00000
-.noooo -.00000 .nnonn
TEMP
GEH
TSU
HSO
-------
7b/07/?7.
PAGE 41
FILE HARE (CREATION DATE = 7b/07/87.)
SUBFILE PERKINS
Ul
Ul
» * * t **********
DEPENDENT VARIABLE.. HC
STEP VARIABLE
ENTERED REMOVED
1 TEMP
GEH
8 TSU
3 HSO
f TO
ENTER OR REMOVE
1.84S71
3b.014Sb
3.H83ba
.03404
SUMMARY
.870
.000
.Ob9
.855
R E G H
T A B L
IPLE R H
.4B4Sb
.7bb84
.787bO
.78780
E S S I
E
SQUARE
.83480
.58805
.b8031
R SQUARE
CHANGE
.83480
.35385
.0388b
.00032
SIMPLE R
.484Sb
.75881
.4S8b4
.7SSS5
OVERALL F SIGNIFICANCE
2S.q7bBS .000
88.38754 0
lb.3SSb4 .UOO
-------
7b/07/27.
PAGE
FILE HARE (CREATION DATE * 7b/U7/S7.)
SUBFILE PERKINS
******************
DEPENDENT VARIABLE.. CO
VAHIABLE(S) ENTERED ON STEP NUMBER
MULTIPLE REGRESSION
3..
HSU
MULTIPLE R
R SQUARE
STp DEVIATION
.71355
.IblBS
ANALYSIS OF VARIANCE DF
REGRESSION V.
RESIDUAL »0.
SUM OF SQUARES
MEAN SQUARE
1.01110
F SIGNIFICANCE
38.3b7H7 .000
•-- VARIABLES IN THE EQUATION
B STD ERROR B
.10731915
.118V08S7E-Oa .bB7!1017E-03
VARIABLE
TEMP
GEH
TSQ
HSQ
(CONSTANT)
ALL VARIABLES ARE IN THE EQUATION.
VARIANCE/COVARIANCE MATRIX OF THE UNNQRHALIZEO REGRESSION COEFFICIENTS.
F
SIGNIFICANCE ;
B.5Sa»»75
.118
1.1133501
i .ibb
a.lblbllB
.013
1.0013858
.381
3.b711H8
.oba
BETA
1 ELASTICITY
-3.S3B1110
-b . 11811
.3131507
.mas
3.83t5758
3.<<003n
.asiatBt
.05043
TEMP
GEH
TSQ
HSQ
.01152
-.00007
-.00007
.00000
TEMP
.00001
.00000 .00000
-.00000 -.00000
GEH
TSQ
.oonno
HSQ
...—.__._ VARIABLES NOT IN 1HE EQUATION
VARIABLE PARTIAL TOLERANCE F
SIGNIFICANCE
-------
7b/07/B7.
f*AGE
FILE HARE (CREATION DATE
SUBFILE PEKKINS
7b/07/27.)
DEPENDENT VARIABLE.. CO
STEP VARIABLE
ENTERED REMOVED
1 TEMP
GEH
2 TSQ
3 HSU
F TO
ENTER OR REMOVE
11.S7V84
bO.bbBb?
H.58385
1.00S3S
SUMMARY
.001
.000
.038
.381
R E
T A
IPLE R
.faSlOS
.B7»2b
.88771
.BSObS
G K t S S I
B L E
R SQUARE
. 12312
,7b»33
.78803
.71385
R SQUARE SIMPLE R
CHANGE
.H2312 .bSIOS
.3tO*e .B3*S5
.02370 .b5»15
.00522 .B18H1
OVERALL F
bB.lOSIS
SO. BOBBY
38.3b?H7
SIGNIFICANCE
.000
0
.000
Ul
-------
?b/07/P7.
PAGt
FILE HARE (CREATION DATE = 7b/07/P7.)
SUHFILF PERKINS
MULTIPLE REGRESSION
DEPENDENT VARIABLE.. NOX
VARIAi)LF.(S) ENTERED ON SfEP NUMBER
3..
TSO
MULTIPLE R
R SOUARfc
STO DEVIATION
.77710
.ObBYl
ANALYSIS OF VARIANCE OF
REGRESSION ».
HESIDUAL >»U.
SUM OF SQUARES
.8H535
.1P717
MEAN SQUARE
.0713*
.OOHfaH
F SIGNIFICANCE
15.PH5H7 .I10U
Ul
03
VARIABLE
TEMP
GEH
HSQ
TSO
(CONSTANT)
VARIABLES IN THE EQUATION
STD ERROR »
-.9SblS7'*3E-01
.S55200H2E-03
.352272bbE-03 .27770173E-03
SIGNIFICANCE
.532
Ib. 1Sn?53
.1)00
».]35511b
.ntt
l.bOllbBb
.eie
3.3H53U72
.075
HETA
ELASTICITY
-3.7191th?
-f.SSHbb
-1.5B51S11
-,31S7b
,81038b8
3.1070Sb7
VARIABLFS NOT IN THE EQUATION
VArtlABLE PARTIAL TOLtRANCE F
SIGNIF1LANCE
ALL VARIABLES ARE IN THE EQUATION.
VAHIANCE/COVARIANCE MATRIX OF THE UNNORMALTZED REGRESSION COEFFICIENTS.
TEMP
GEH
TSQ
HSU
.001B8
-.00001
-.00001
.00000
TEMP
.00000
.00000 .00000
-.00000 -.00000
GEH
TSO
.OOOUO
HSQ
-------
7b/07/a?. . PAGE »<»
FILE HARE (CREATION DATE a 7b/07/a7.J
SUBFILE PERKINS
U1
VO
DEPENDENT VARIABLE.. NO
STEP VARIABLE
ENTERED REMOVED
1 TEMP
GEH
a HSU
3 TSO
X
SUMMARY T A
F TO SIGNIFICANCE MULTIPLE R
ENTER OR REMOVE
ta. 87088 .000 .7a3»»
b.tabiq .015 ,?bb?8
l.bon? .812 .77710
B L E
H SQUARE R SQUARE SIMPLE R
CHANGE
.03bBH .03fa8* -.1S1SS
.Sa33b ,tBbS2 -,faS712
.58715 .OfatSS -.594Sb
.b03B1 .01SSH -.18772
OVERALL,? SIGNIFICANCE
^3.0587^ .000
is.atst? .000
-------
?b/07/27.
PAGE
FILE HAME (CREATION OATE = 7b/07/37.)
3UBFILF HEHCEOES
OEPENOFNT VAHIAHLE.. HC
VARtABIt(S) fNlFREI) ON SIEP NUMBER 3.. HSO
MIJLUPIF. R .B'l'iaS ANALYSIS OF VAKIANCE
R SOUAME .11920 HEGRESSION
STO DEVIATION .lift?? RESIDUAL
VAHIABLt B STO EHHOR B " F
SIGNIFICANCE
TEMP . ?092b;i4E-lll . 34 H 3 174 hF-(l 1 4 . 1 »51n1»fl
GEH -.bS^hMiqHt-OH .RHl'«S2B2E-n3 . Kb23h!»SnE-0?
.lib
TSQ -.»5'.b5m7F.-m . 22 1 32 7 1 1 £-03 ». 2383373
.0*1.
OF SUM OF SQUARES MEAN SUUAHE F SIGNIFICANCE
t . . l)l?7» .OU31H 1.2HShJ .?HJ
3B. . 09113 .OU?HR
BETA VARIABLE PARTIAL TOLERANCE F
ELASTICITY SIGNIFICANCE
lO.RHBHbO'*
-.OS3b8«)H
-.03b37
-10.3H7S'»t7
-22.50923
HSU
(CONSTANT) -2.b02U5IR
.855
.051R3
ALL VAHIABLES ARE IN THE EQUATION.
VAHIANCE/COVARlANCE MAFHIX OF THE UNNOHMALI2FJJ REGRESSION COEFFICIENTS.
TEMP
GEH
TSO
H3Q
.00121
-.01)001
-.nnooi
.00000
TEMP
.oonoo
.noooo
-.onooo
GEH
.onono
-.onnnu
TSU
.ounno
HSfJ
-------
7b/o7/27. PAGE is
FILE HARE (CREATION DATE = 7b/0?/27.)
SUBFILE MERCEDES
A********************** MULTIPLE REGRESSION ***********************
DEPENDENT VARIABLE.. HC
SUMMARY TABLE
STEP VARIABLE f TO SIGNIFICANCE MULTIPLE R R SQUARE R SQUARE SIMPLE R OVERALL F SIGNIFICANCE
ENTEKED REMOVED ENTER OR REMOVE CHANGE
I TEMP .S3i»Bl ,4bS .10847 .01177 .01177 -.10847 .27034 .7bS
GEH .Ob3faB .802 .11549 .01334 .00157 -.01204
2 TSU 4.b4834 .037 .34411 .11841 ,10507 -.11878 l,74bll .173
3 H3Q .033^3 .855 .34525 ,11120 .0007S -,03B3b
-------
PAGE 10
FILE HARE (CREATION DATE « 7h/07/i7.)
SUBFILE HRRCEOE3
MULTIPLE REGRESSION
DEPENDENT VARIABLE.. CO
VAHIAHLECS) ENTERED ON STEP NUMBER 2.. MSQ
MULTIPLE R
R SQUARE
STl) DEVIATION
.5*200
ANALYSIS UF VARIANCE OF
REGRESSION 3.
RESIDUAL 31.
SUM Of SQUARES
.Ulb31
.05101
MEAN SQUARE
.001*1
.00131
F SIGNIFICANCE
H.15.033 .018
to
.....—.....—..._».. VARIABLES !N THE EllUATION -—---- — - — -—.- —
VARUHLE H 3TO ERROR B f BETA
SIGNIFICANCE ELASTICITY
TEMP
GEH
HSQ
(CONSTANT) .B7S8ie
-------
7b/07/27. PAGE 81
FILE HARE (CREATION DATE » 7b/07/27.)
SUBFILE MERCEDES
DEPENDENT VARIABLE.. Co
****** MULTIPLE REGRESSION
STEP VARIABLE
ENTERED REMOVED
1 TEMP
GEH
a HSO
SUMMARY TABLE
F TO SIGNIFICANCE MULTIPLE R R SQUARE R SQUARE SIMPLE R
ENTER OR REMOVE CHANGE
b.bSBbS
.7*718
5.02378
.01*
.3S3
.031
.35839 .12837 .1SB37 -.35821
.37SSH .14435 .01SS8 -.0437S
.H
-------
7b/07/27.
PAGE S3
FILE HARE (CREATION DATE » 7b/07/?7.)
SUBFILE MERCEDES
MULTIPLE REGRESSION
DEPENDENT VARIABLE.. M<)X
VARIAHLE(S) ENTERED ON STEP NUMBER
MULTIPLE R
R SQUARE
STD DEVIATION
.HUH
.bBIBO
.ObtOd
2.. HSQ
ANALYSIS UF VARIANCE
REGRESSION
RESIDUAL
DF SUM UF SQUARES
3. .301(11
31. .15177
MEAN SQUARE
.10301
.UU110
F SIGNIFICANCE
eS.llbUl U
. . VARIABLES IN THE EQUATION
VARIABLE B STD ERROR B F BF.TA
SIGNIFICANCE ELASTICITY
.«bh7lhSOE-03 .1S-UOS72E-03
-.•»20bSBblE-02
TEMP
GEH
HSU
(CONSTANT) 1.007Bb07
13 18,bOb7ll
.000
.b2RbSSb3E-OS 5.1*30081
.125837*4
.f]33il?b
.o^soa
-.3788*
.BB1SHB7
.19B04
.......... VARIARLFS NOT IN THE EQUATION —
VARIABLE PARTIAL TOLERANCF F
SIGNIFICANCE
TSQ .r-UBO .OOOM1 1.7BH7b01
F-LEVEL OR TOLERANCE-LEVEL INSUFFICIENT FOR FURTHEtt COHPUTAtlON.
VARIANCE/COVARIANCE MATRIX OF THE UNNORMAHZED REGRESSION COEFFICIENTS.
TEMP
GEH
HSQ
.00000
.ooono .noooo
-.00000 -.00000
.onooo
TEMP
GEH
HSQ
-------
7b/o7/27. PAGE af
FILE .HARE (CREATION DATE = 7b/07/87.)
SUBFILE MERCEDES
DEPENDENT VARIABLE.. NQX
MULTIPLE REGRESSION
STEP VARIABLE
ENTERED REMOVED
1 TEMP
GEH
e HSO
SUMMARY TABLE
F TO SIGNIFICANCE MULTIPLE X R SQUARE R SQUARE SIMPLE R
ENTER OH REMOVE: CHANGE
.SSBOb
5S.7S08H
S.IH301
.000
.OSS
.78375
.0?bh3 .07bb3
.J>37b3 -.78031
-,7020b
OVERALL f SIGNIFICANCE
J1.8H87b
es.nboi
.DUO
0
U1
-------
7b/07/87.
CAGE
FILE HARE (CREATION DATE => 7b/07/27..)
SUBFILE PEUGEOT
MULTIPLE REGRESSION
OEPENOFNI VARIABLE.. H.C
VAHIABLE(S) ENTERED ON STEP NUMBER
3.
MSB
MULTIPLE R
R SQUARh
3TD DEVIATION
VARIABLE
TEMP
GEH
TSQ
HSQ
(CONSTANT)' -1
.tHObB
. ISY20
.0758?
RECESSION
RESIDUAL
VARIABLES IN THE EQUATION
STO ERROR B
.S027330HE-01
.lln
-------
7b/07/27.
PAGE
FILE HARE (CREATION DATE = 7b/Q7/27.)
SUBFILE PEUGEOT
DEPENDENT VARIABLE.. HC
STEP
i
e
3
VARIABLE
ENTERED REMOVED
TEMP
GEH
TSU
HSQ
f TO
ENTER OR REMOVE
2.25bBb
,b11?8
!7B101
» H U L T I P
3 U M M
SIGNIFICANCE
.m
.HU
'.38J
LE R E G R t 3 S I
* R Y TABLE
MULTIPLE H K SQUARE
,33581 .11202
.3SH17 .1282H
.^aOI* .17711
R SQUARE
CHANGE
,01!>*b
.OHU10
.01701
SIMPLE R
-.33SBS
-.S7HQO
-.32723
OVERALL F SIGNIFICANCE
2.8b170 ,|)bS
2.727b7 .057
2.gii1eiH .085
-------
7b/07/27.
PACE
00
FILE HARE (CREATION DATE = 7b/07/27.)
SUBFILE PEUGEOT
DEPENDENT VARIABLE.. CO
VAHIABLE(S) ENTERED ON STEP
MULTIPLE R
R SQUARE
STD DEVIATION
VARIABLE
TEMP
GEH
HSO
(CONSTANT)
NUMBER 2.. HSO
,bl?88 ANALYSIS OF VARIANCE
.37Sb3 REGRESSION
.05575 RESIDUAL
B STD ERROR B F
.b7SUSllE-02
.27bB7<*b7E-03
.2222BVBOE-OS
1.500fa7»0
SIGNIFICANCE
.mS51R3E-02 28.717028
.000
.781S1551E-03 .IBISbOll
.72b
,SO
-------
7b/07/S7.
PAGE
FILE HARE (CREATION DATE =
SUBFILE PEUGEOT
?b/07/a?.)
DEPENDENT VARIABLE..
CO
MULTIPLE RE6HES3ION
STEP VARIABLE
ENTERED REMOVED
1 TEMP
GEH
I HSO
SUMMARY
TABLE
F TO SIGNIFICANCE MULTIPLE R R SQUARE K SQUARE SIMPLE R
ENTER OR REMOVE CHANGE
83.13880
b.SSttb
.000
.Ol*
.bfaS
.37Sb3
,2b703 -.Slb?b
.iOSIb .01380
.U0313
OVERALL K SIGNIFICANCE
U.S7S3£ .IIUU
7.b203i .nnn
-------
7b/07/27.
PAGE
12
FILE HARE (CREATION DATE = 7b/07/27.)
SUBFILE PEUGF.OT
DEPENDENT VARIABLE.. NOX
VAKIABLE(S) ENTERED ON STEP NUMBER 3.. HSQ
MULTIPLE R
R SQUARE
STD DEVIATION
VARIABLE
TEMP
GEH
TSQ
HSQ
(CONSTANT) -
ALL VARIABLES
.80181 ANALYSIS OF VARIANCE
.bH2B1 REGRESSION
.OSbSH RESIDUAL
B STD ERROR H F
SIGNIFICANCE
.2bl«b703E-01 ,372
.23331713 l.»«H2»3» ^SISbHIBE-Ol
.873
ARE IN THE EQUATION.
VARIANCE/COVARIANCfc MATRIX OF THE UNNORMALIlED REGRESSION
TEMP
GEH
TSO
HSQ
.001H1
-.00001 .00000
-.OOOD1 .00000 .0000(1
.00000 -.00000 -.00000 .noono
OF SUM OF SQUARES MEAN SQUARE f SIGNIFICANCE
<*. .21210 .05323 lb.fa5c>SS .100
37. .1182b .110320
BETA VARIABLE PARTIAL TOLERANCE t
ELASTICITY SIGNIFICANCE
2.nnean2
-1.05272HB
-. 23*53
-1.13H817H
-1.37003
.HOObOB
.08b11
COEFFICIENTS.
TEMP GEH TSO MSO
-------
7b/07/a?. PAGE 13
FILE HARE (CREATION DATE = 7b/07/S7.)
SUBFILE PEUGEOT
it********************** MULTIPLE REGRESSION ***********************
DEPENDENT VARIABLE.. NOX
SUMMARY TABLE
STEP VARIABLE f TO SIGNIFICANCE MULTIPLE ft R SQUARE R SUUARI- SIMPLE R OVERALL F SIGNIFICANCE:
ENTERED REMOVED ENTER OR REMOVE CHANiifc
1 TEMP 2.831*2 .100 ,8blSS .ObBbS .ObBbe -.2blSS i1.3?H77 .UUO
GEH bl.3b3*3 U .7RBBO .b380H .bbH-»b -.78218
2 TSQ ,277bB .bOl .BQUtH .bt070 .U02b3 -,2bS2H e!2.bH72b .OOU
3 H30 .aabSS .b37 .80181 .bH28S .OU21S -,7tll7S Ib.bSSSS .UUU
-------
FILE HARE (CREATION OATF = 77/m/P*.)
8UBFRF
PARE ?•»
A********************** MULTIPLE RERHE38ION A**********************
OEPFNDFNT VARIABLE.. HC
VAHtAHI.F(S) FN1ERED ON fl|FP NUMBER ?.. HT
MULTIPLE R »p5M3B ANALYSTS OF VARIANCE OF RUM OF SQUARES MFAN SQUARE f SIGNIFICANCE
R SQUARE .Ob3b» REGRESSION -3. - iOOSHH .;. .nnifll .B1101 ."IS*
ADJUSTFD R SQUARE -.nn?bh RESIDUAL m i.aeme
TEMP .37SbU788E-ll3-.aiqihl1bF.-n3 jlSBlf»07E-01 -- -.OH331P1
slot .15*75- -
HT - -5a30SH37RE-nH -• -.-SBPlnaiaF.-n* - ibb'iniaiB -l.B«»»7h7ll—
L:M
r--(CONSTANT) - ,1331«HH.S ,173»110b i,5in3B1Jl
I- ... - ... - j»71
-.tpo- - -J*nHebi-
p ALL VARIABLES ARE IN THE-EOUATION.
rrv
.3
(-—•-• • --•_ • - •• - ; ;.; -7... '..".'_ r™.".'..l
r -.- • • ~ ~ ' - "'" : "'TH
-------
77/M3/S1. PAGF.
FILE HARE CCREATION DATF. = 77/03/S1.)
I SUBFILE OATSUN
I ni***-***«•***»*»*»****<
-DEPENDENT VARIABLE.. HC
MULTIPLE REGRESSION
SUMHAKf-TA-BLE ...
STEP VARIABLE F TO SIGNIFICANCE MULTIPLE n R SQUARE R SCJUARF SIMPLE f OVERALL F
ENTERED REMOVED ENTER OR REMOVE CHANGE
i GEH
TEMP
- a MI
.nbfin?
1.157)]
.bb-HlS
.iH?o7 .neniB .oania -.item
.Mt?flB .02bBS -.J1330
.llbjbl .UJSSb -.IhbS?
J.
W
L
-------
77/03/?*. PAGE ,??
FILE HARE (CREATION DATE = 77/M3/PH.)
I 8IIMFT1.F DATSHN • •
| •*•««******••««»••«»««» MULTIPLE REGRESSION *««*««•«*«*****»***»*«*
I OFPFNOFNT VARIABLE.. CO .. .
| VARIAHIE(S) ENTERED ON 3TFP NUMBER ?.. HT - -
1
I • ..,..._..._„-.._ . .
MULTIPLE R .30IH3 ANALYSIS OF VARIANCE OF SUM OK SQUARES MEAN SQUARE f SIGNIFICANCE
I R SQUARE it)15l3 REGRESSION 3. .nil3b .110379 I.ftll75 .?Sb
AHJUSTFO R SQUARE .fl?7ab RF.SiniJAL Hfl; ilfiuoq .110270
I STO nEVIATION .05H8 COFFF OF VARIABILITY b.S PCT
i_i VARIABLES IN THE EQUATION ---—ii-.-i.-=.i^i^.i—j- •— -•sbea&<.a>«B VARIABLES NOT IN THE EQUATinw —
B - STD ERROR ft F BF.TA - -VARIABLE PARTIAL • -TOLERANCE F
• ---- --- -------- -- —*.*..
SIGNIFICANCE ELASTICITY - - .-...- SIGNIFICANCE
.533 .Ibb58
"TEMP ".7S«»S98H8E-n3 ' .aSa'tS^nBE-O?
(CONSTANT) .B888»Hn7 — -;1S<»83360- 17.17887B—
" " .000 "
I ... .... I.. . . . . .... - - .-.- .- .. I
[—»LL VARIABLES ARE IN-THE EQUATION;— — -< 1
LT7
~:..".: :~i
"".".;:_"::
-------
77/ni/an. PAGE as
FILE HARE (CREATION DATE = 7?/D3/?».)
I SUBFILE OATSUN .. - . -. .
, »********«*****(.******* MULTIPLE « F- G HESSION *****«***«******««*.**« i
p- DEPENDENT VARIABLE.. CO |
I - -SUMMARY-- T A- B I. E ' - '
. -- !
'•STEP VARIABLE f TO SIGNIFICANCE HIILTIPLE " -H SUUARE R sotiA«e SIHPLE R OVERALL f
r ENTERED REMOVED ENTER OR REMOVE cHANGt
-o r-
en
i GEH 1.1H7S1; .09S .aatbn - .nsotb .IISOHS .aatbo a.nsn?? .itt
TEMP i.?8B3a .lat --- • .3onaa— -.nmus ----.D-tHhi -.nSH'tb-- -
a • HT • .23(181 .fa*! - -.aoB'tS .01513 •-•.HDSili) .11713 1.H017S .aSb
-------
77/03/?H. PAGE 311
FILF HAUt (CREATION OATF = 77/OV?1*.)
; SUHFIl.F OATSUN ,
i I
, ******************** * * * MULTIPLE REGHLSSIOI ***** * * **************** .
, DEPENDFNI VARIABLE.. NOX .
i VARIAHIE(S) FNTFRED ON STEP NUMBER ?.. HT
' MULTIPLE R iflfhqo ANALYSIS OF VARIANCE OF SUM OF SQUARES MEAN SQUARE F SIGNIFICANCE '
I
I
i R 8QHARF. ;7178H REGRESSION 3. - ,3SbSt .11885 33.8?OhS .nod
1 AOJUSTFU R SQUARE .b^hllj RESIDUAL mi.
I STD DEVIATION .05R88 COFFF OF VARIABILITY b.7 PCT
VARIABLES IN THE EQUATION —.->-.--....--...«.---. ......... ---- ............ VARIABLES "NOT IN -THE EQUATION
I VARIABLE B STO ERROR H F BETA - -VKRIABLE PARTIAL-- TOLERANCE F
•jl i - SIGNIFICANCE ELA3TICIIY --.--- - SIGNIFICANCE
i- 6EH - -4l»b?5153E-ol . 3n7bPS3oE-n? ??.b()5»?8 - --S-.bb7-teb5 -
1 •• .000 -U30POV • - — -—• -- -•- - --
i - TEMP -.bBEEbnntlF-ll? - ;?B7HbBScIE.OP -Sihl7nBRS -^r*
HT •-- • il50»"»o1?E-nS .-...3?15««3H8E-n»>- ' lh;"»01S»H - — S. 1717887- ..... --------------------------- ..... ----------------- ....... -•: ......... - ----- i
.... .. . ...... -' --------- i --------- ---------- ........ - ........... ----- — - ..... — . . . i
(CONSTANT) libll^mi— ------ i?P78Bnon -•- -5n.03b»33
r
1 ...... • • ...... ••• - - - ;000- ------
-ALL- VARIABLES ARE- IN THE- EQUATION;
L77777777
r 7
I7~.77777'7
.7" ...... , ................ '""" ' ' ..... " ......... 7" ........ """"". ......... ."". ............. ~".""~7..77:77" 77]
r"~.~7~"7"" ............ ~ ..... . ---..- ........... ...... ••"_." ....... ~" "~ ............. ~'~7 "7"7 771
r~'~ ............ • :"" ' .......... ~" •• ..... -- --- ...... -- ........ ; ...... _— ...... -777"]
r~ .— -77-- ......................... - .": -.7.7 .. ": "7.77 7; -7-7:7:70
L7777-": - ......... • " • "' . •---—-- .................. - " ....... - ......... ........ ;; . - ..... - " ~ ........ •"~'7T~1
-------
77/03/2H. PAGE 31
FILE HARE (CREATION OATE = 77/03/24.)
r--* »*** ******* MULTIPLE REGRESSION *
r SUBFILF OATSUN
r- DEPENDENT- VARIABLE;. - NOX -• - - ........ -- ^ . .... ._ - _ ... .
I _ .. - -. - s U M M A R Y-~ T A-b-L"E •••-" ~ ' '" '" -"' '
I- ST1:p • VARIABLE F TO SIGNIFICANCE MULTIPLE R R SQUARE R SQUARE SIMPLE R- OVERALL f SIGNIFICANCE
(--• -- ENTERED REMOVED ENTER OR REMOVE .- CHAN«E
r 1 GEH • Sb.lSSbl U .7ShS8 .57ai»l .5781*! • -.75bSB 3l),SlbHS .IKIU
TEMP — a.s7nsB .nss
r - 2 • HT " ~ ' lb 4015* .000 .BHbln .71724 .1151H -.71(128 33.H2nbS .IKIO
I. - - " - - - -..-....
-------
77/03/?f.
PARE
00
FILE HARE (CREATION DATE = 77/03/Jt.)
3U8FILF —PERKINS ... . .,
MULT I-P- L E -- R E G-R E 8 8 I 0 N *
DEPENDENT VARIABLE.. HC --
VARIABLE(S) ENTERED ON STEP NUMBER ?.
HT
-MULTIPLE R 47bR51
•R -SQUARE — ..... - .5SII73
ADJU8TFD R SQUARE .5bil7S
3TD DEVIATION .0787*
ANALYSIS OF VARIANCE DF 8UM-OF SQUARES
REGRESSION --- 3, --------------- , 3^,9(1
RESIDUAL - H18 ....... - -
COE.FF OF VARIABILITY 11.1 PCT - — -
MEAN 3IJUARE
— ---- *i2?30
F BIGNIF ICANCt
lS.7ab3b 0
-VARIABL-E
6E-H ---------
VARIABLES IN THE EQUATION I VARIABLES NOT- IN--THE EOUAT ION
g STD ERROR B - -F- BET* J-VARlAHLE PARTIAL T-OkERANCF ^ •- ••-•
SIGNIFICANCE- ELASTICITY SIGNIFICANCE
(-CONSTANT-) ---- r
;b07
•*kt-VARUBL-ES-ARE-lN THE EOUAT I ON; -
-------
77/03/?». PAGE 35
FILE MARE (tREATION DATE » 77/03/gH.)
SU6FILF PERKINS
* « « . «..«««««««..«**«.»« MULTIPLE REGKtSSION
-DEPENDENT VARIABLE.. HC
S'U H M A R'Y - T -A B IT E '-
STEP VARIABLE f TO SIGNIFICANCE MULTIPLE "U" H -SQUAHf- R 30ll»i*t S1HPLE R ------- OVERALL F -SIGNIFICANCE'
FNtEHEO REHOVEO 'ENTER OR REHOVE " ~ ------ ' " " ' CHANGE ^" ----- - ' " ' " --------- """"
(-fH " " 3b.0l»^b .000 ---- .75881' -- .57571 .S?*?^ .^5881 dS.S7h8S — 70UII - '
-• TEMP ~ l.Z»171 '- .870 ----- .7bh8t ----- .58805 --- idlBab .IBHSb- ------ - ..... - -------- ---------
HT '-" .SbStH .b07 • " -.7b85S ---- .51073 .00?b8 - .Tb?*S - — !H,7eb3b' ------- 0"
to
-------
77/03/24. PAGE 37
FILE HARE (CREATION DATE = 77/03/8H.)
L-f-- PERKINS
•»—*—«-»—»-* -»-•» * «» «-« * * *«» » * »*« M U L- T- I P L -E R E-G RE 3'3 10 N
-DEPENDENT VARIABLE;-. CO - . . . _
-V*HIABLE(»> ENTERED ON-3TEP NUHBER- ?„. HT
MULTIPLE R .B8»0a ANALYSIS OF VARIANCE DF SUM OF SQUARES MEAN SQUARE '" F SIGNIFICANCE
•R-SOMAHC- i78l»t REGRESSION Ji "»T33120 — I;
-------
PAGE 3H
FILE HAKE (CREATION DATE = 77/»3/?f.)
I SUBFILE PERKINS
• - *.••*««•**************** MULTIPLE REGRESSION
DEPENDFNT VARIABLE.. CO
SUMMARY- TAHIE
IFICANCE MULTIPLE H H SO
ENTERED REMOVED ENTER OR REMOVE CHANGt
STEP VARIABLE F TO SIGNIFICANCE MULTIPLE H » SQUARE R SQUARE SIMPLE R OVEKALL f SIBNIFICANCF '
oo
H. L
I
i REH bO.hhflb? .01111 .83415 .bl?!* .bS?14 .H3HHS bB.luSis
TEMp 11.R7181* .001 .B7tEb- .7bH33 .Ob?11 .bSlOl -
a HI a.aiasn ,nsn .HSHOS .78141 .oi7is .8b4b|» 4B.B77as
-------
77/D3/?-*. PARE
FILE HARF . (CREATION DATE = 77/OVJ").)
SUHFTLF PERKINS
*»*»
OFPF.NOFNT VARIABLE.. NOX
VARIABIK(S) ENTEWFO ON STEP NIIHBFH
MW.MPU » .7751S
R SQUARE .hOHS
AOJURTFn H SOIJAWE ,S7?15
STO HEvIAriON .tlb7>8
?.. HT
ANALYSIS OF VARIANCE
REGRESSION
RESIDUAL
COFFF (IF VARIABILITY
OF
3.
»l.
7.3 PCT
3IIM OF
MEAN SQUARE
.n^Ji
.notss
ool
to
--VARIABLES IN THE EQUATION — ,—*..^—..-.-..
B STO ERROR B F BETA
SIGNIFICANCE ELASTICITY
,llO»";57hE-nt . 32ne*b1HF-ll? ll.81lil.S1 • •'
.001
VARIABLE
GF.H
TEMP
HT
(CONSTANT)
ALL VARIABLES ARE IN THE EQUATION-.
3;12b<5001
.007 .7?f]0n
»..«««»«.«-VARIABLE3-NOT IN THE EQUATION
VARIABLE - - PARTIAL TOLtRANCE F
• • • SIGNIFICANCE
-------
77/n3/a'». PAGE ti
FILE HARE (CREATION OATF = 77/03/'?t.)
- SUBFILE - PERKINS ........
r — »• * * *' * ' * * * ' * * * * *' * '
—OEPENOENT VARIABLE.. "NOX
n .:.:;; _::
MUITIP'LE REGHES3ION «****
"]
SUMMARY' —T A fl-'L E ~
STEP VARIABLE ' ' F TO ' SIGNIFICANCE MULTIPLF R K SQUARE R SOUAHE S1MPL£-R ..... OVERALL F SIGNIFICANCE
• ..... ENTERED REMOVED ENTER OH HEMOVE . .. . ..... _ - ... -CHAN6(: ....
I HI
H2.870HB
3.2
-------
77/03/PH.
PAGE 11
FILE HARE (CREATION DATE = 77/cn/?'(.)
SUBFILE MF.RCFDF.S
OEPENUFNT VARIABLE.. HC
VARIARIE(S) ENTERED ON STEP NUMBER 3.. HT
MULTIPLE REGRESSION *
MULTIPLE R .177flb
R SQUARE- --- .03135
AOJUSTFO R SQUARE -»nH31b
STO DEVIATION .05153
ANALYSIS OF VARIANCE OF
REGRESSION - 3. -
RESIDUAL 3*1.
COFFF OF VARIABILITY sc.c PCT
SUM OF SQUARES
.-00335
MEAN SQUARE
SIGNIFICANCE
.731
;HU?b5
VARIABLE
g 6EH-—
-TEMP -
— VARIABLES IN THE EQUATION
B STD ERROR H F — -HETA
..-.-. SIGNIFICANCE- ELASTICITY
13 - ,?SS772q7F-fl? . 7ShS'f 1P1-- —-
13 - ,317SB73bE-n3- .'
_HT -.35713R71F.-01* - . 3niq<*abflE-D*
-.(.CONSTANT) - .7H3bOq7qE-n I ,17t3<
-------
77/03/24. PAGE IS
FILE HARE (CREATION DATE = 77/03/24.)
SUHFILF MERCEDES - ...
-••oft******************** MULTIPLE NEGritSSION *****»**********»<.»„••*
DEPENDENT VAHIAHLE.. HC
• • 8 U M M A R Y 1 A B I. K - - - - I
STEP VARIABLE • F TO SIGNIFICANCE MULTIPLE H H bQUAKE N SUIIAHE SIMPLE « OVERALL h
F.NTEKEO REMOVED ENTER OR REMOVE CHANGE
,.- - 1 GEH .OhShR .81)? .01204 .OJUlt .(IIJIU4
I T£Mp .53fBl .4h^ .11541 .111334 .tlMl1}
g HT .72525 .4110 .l?7(lb .03135 .II1HHJ -.(13374
00 f
Ul l
-------
77/03/SH.
PAGE 1 7
FILE »UPE (CREATION DATE = 77/03/St.)
SIIPFILF MFRCEOES
MULTIPLE REGRESSION
OFPFMDFNT VARIABLE.. CO
VAR1M»o
. iHbt3
.01701
ANALYSTS OF VARI4NCE OF
REGRESSION 3. -
RESIOUAl. 3S.
COEFF OF VARIABILITY 5.B PCT
SUM OF SOUAHF3
.05313
MEAN SOUAHE
.noHbb
.no 137
SIGNIFICANCE
CO
CT*
| VARIABLE
— VARIABLES IN THE EQUATION —-•
B 8TO ERROR B F
BETA
SIGNIFICANCE ELASTICITY
..iiii^s-i VARIABLES NOT IN THE EOI'AMON
VARIABLE PARTIAL•••- TOLERANCE f
SIGNIFICANCE
-.3nsuo38E*nS
? ?.fl23nnn?
.101
-.'»59377?bE»na ;19K31b31F.»nP 8.b3b363
-------
77/03/3-*. PAGE 1H
FILE HARE (CREATION DATE = 77/03/PH.)
I SUBFILE MERCEDES
t" *•-***«*••********••
I DEPENDENT VARIABLE.. CO
* MULTIPLE REGRESSION *
SUMMARY TAH'LE
STF.P VARIABLE f TO SIGNIFICANCE MULTIPLE M rt 30UARF H SQU»«E SIMPLE R
ENTERED REMOVED EN1FR OR REMOVE CHANGE
1 fiEH
TEMP
? HT
.7H7I8
h.bSBbt
.n-»3?s .noise
.14H35
.(IHb
F STGNTKICANCE
3.37t)«
3 . »II1 b H
-------
FILE HAHF. (CMfAIKIN DAIF = 7?/(M/?1.)
SURF]IF HfRCFOHt
[>FPF|g|l*Nl VARIARLF.. NO*
VARIAHIF(R) FNTEWED ON SIFP NUMRF-R f.. HT
HI. I I P I. F P F r, H t 9 S J M N
MULT I PI F R
R SQIUWF,
Ar)J'KTrn W SOUARE
STD DEVIATION
.?t8«3
«N»LYStS OF VARIANCF OF
SSION 1.
iAL H".
HP VARIABILITY ?.n prr
SUM OF" SQIUHfS
.35HJB
..HIM
MEAN 3QUARR
.IJ13R
d
VAHIABIF.
• GEH
I TEMP
r-HT
I (CONSTANT)
.— VARIABLES IN m EUIIATION ——
H STD ERROR H F HFFU
SIGNIFICANCE ELASTICITY
tun
S';?
oni
9
II
I.J J311
VARIARUF.S NOT IN I Hk EQUATION
PARTIAL rottRANce F
ALL VARIABLES ARE IN-THE EQUATION; -
r 7
-------
77/03/2H. PAGE 21
FILE HARE (CREATION DATE = 77/OV24 .)
f-~ SUBFILF MERCEDES
i::*.*.* ************* * * * *
r DEPENDENT VARIABLE.. NOX
* MULTIPLE HFGktSSION **
- SUMMARY -TAR-LE
STEP VARIABLE F TO SIGNIFICANCE MULTIPLE R H SOUAKE h SQUARE SIMPLE R
ENTERED REMOVED ENTFR OR REMOVE CHANGF.
I 1 REH
1 TEMP
i e HT
.01)0
.H5S
.noi
.VHU31 .buHfla .bOHHH -.7BU31
.7B37S .bJHab .UD538 -.?7bH3
.71283 .O^BSb -.7<*3fat
OVf-RALL F SIGNIFICANCE
.unu
ii
00 I
-------
FILE HAUE (CREATION OATF = 77/m/PH.)
91WFII.F PRUKFOl
REGRESSION
>fNT VARIABLE.. HC
VA»T*tMFI9) fNTFHED ON STEP NIIMHFH P.. HT
HIM. IT P|. E R
H SlJUARF
AOJIIRTFr> R
STO (1EV1ATK1M
.57111?
ANALYSIS OF VARIANCE OF
RFKRF8RION 1.
RESIIWAL 3«.
COfFH OF VARIABILITY 17.'
SUM OF 3NUARFS
.OBblB
MEAN SMUARF.
SK-NIF ICAkK.t
. nili'
I
I
Q
__ VARIABLES IN THE EttllATION -.--. — .-. ——....... .......... -VAHIAHLF3 NOT IN 1HE EQUATION |
B 9TH ERROR 8 F BETA VARIABLE PARTIAL TOLhHANCE F ;
j SIRNIFICANCE FtASTICITY- 8IGNIFICANCF !
'-• " .one
r TEHP .Se7H0381E.|.ie ;?"»NeblnHE-OS 3.lb11ini
1 .I1B3 1.10118 ........
' .ring - -g.gnnii^ --• •-• '
r- (CONSTANT) -iSltOS-fiaE-nl ;S3?300»0 ;BH9tinbf.OF-08- «- — — ]
I . . . _9p(. ....... .... . . ....._... 1
p-ALU VARIABLES ARE IN THE EQUATION; - -— - j
r~~:~~ ". 7".~ " "~ " """"i
r.
(
r.t~
51.
-------
PAGt S
FILE HARE (CREATION DATE = 77/(13/St.)
SUBFILE PEUGEOT
*********************** MULTIFIE HEGKtSSION ***********************
DEPENDENT VARIABLE.. HC
S U h M A K Y TABLE-
STEP VARIABLE F TO SIGNIFICANCE MULTIPLE H w SQUARE « SOIIAHE- SIMPLE R OVERALL F SIUNIUCANCH
REMOVED ENTEK OR REMOVE • CHANGE
1 GEH .b117g .til .£7411(1 .I1V7HH .077MH -.?74|)(1 £.8hH7() .lihM
TEMP a.esbflb .1*1 ,isBi7 • .jees^ - .osrm -.sssaq I
2 HT H.tfOb? .nOS .S7t|)S .38l5u .Pnic1) - 3311*1' b S£f?u imd
-. . I
-------
PARE ;
FILF HAKE (CREATION OAIF = 77/in/pn.)
PMIRFOT
**»*»******««,.*.* MULTIPLE REGRESSION
VAUIAHLE.. TO
VARIAHIF(S) FNTFRFD ON 3IF.P NUMBER ?.. HT
Mill T I PI K R .hlsnp ANALYSIS (IF VARIANTS OF SUM OF SQUARES MFAN SQUARE F SIRNIF1CANCH
R Sf)i|AI»F .S'HH* REGRESSION J. .07153 .OS3B5 ;
AnjllSTFI) R SOUARF. .SJIlh nFMHIIAL. IH. ,117hl .OOIID
STD OFVIATION .(iSShT CIIEFf OF VARIABILITY S.h PCT
| ik 1- VARIABLES IN THE EOUATK1N «.bi.«.i._-*uj. . - i.^.--..-..- VARIAHLF3 NOT" IN fME EQUATION ----—---.. ,
I VARIABLE B STD ERRnR B F liFTA ' - VARIABLE PARTIAL -TOLERANCE - f i
VD , ' -ii • • ---«. — '
M f 3IGNIFICANCF ELASTICITY 91GN1HCAN(.E j
.H?n " ,lb70n ' ' - . . . . .
I TEMP «-.5')81»1l>t33E«>n;? . 21* 15n7H SEMI? 5.157i|Bpn -;SRSR41»B - — - • I
.557 ».1?07H - - '
I- (CONSTANT) l.SSO'fflbn .laiab^h"! -SS.ISSJSB - - - - -- "I
.onn .. - . . ._. . . •
r ' ...--- - -. — -•• .-..---.._ ..j
i- ALL VARIABLES ARE IN THE EOUATTON. - - ---..- - -j
r ."" "" " " .'""'.." ""."""".." ~..'.~'~.~:"'.."~:". ~ """' ~~~i:~~ n
r
r
f.._.
L"l.
r:'
-------
PAUE 8
FILE HARE (CREATION OATF = 77/03/?H.)
SUBFILF PEUtiEOT
** MULTIPLE wE<;ntssioN
OEPENOFNT VARIABLE.. CO
SUMMARY Tfihlt'
STEP VARIABLE F TO SIGNIFICANCE MULTIPLE H N 3UUARE R SQUARE SIMPLE H
FNTEHED REMOVED ENTF.R o« REMOVE CHANGE
i r.FH
TEMP
? HT
10
U)
b.SSHHh
aa.nofln
.3SlhB
.D138II .
.nj38n
.557
OVEHALL F SlliNIi- ICANCK
7.
-------
°»r,t in
FII.F MAHF. (CREATION OATF = 77/m/Pt.)
3IIHFH.F PFIIUF'll
• *»»»...».,«**»»****«*« fULIIPLF R F. G H h ' 3 3 I II N t
OFPFNOfNl VARIARLF... NOX
VAHTAfMK(8> FNTKHF.O ON RTF.P NUMBER P.. HI
MIILTIPIE R
I R SnUAWF.
' AH-'IISTFO R SO.IJAHE .h?7l)fl
, STD OFylArlnN .054HB
ANALYSIS OF VARIANCF DF SUM OF 3QUARFS
HEGHI-SSION 3, *eib7(l
RF.SIDIIAL JR. . tilth
COFFF OF VARIABILITY 7.h PCI
MEAN SQUARE
.07??3
.01)301
f Sl«M[F ICAnCK
33. 181*3 .fidl)
1C
VARIAHIE
GFH
TEMP
- HT
(CONSTANT)
VARIABLES IN IMF EOIMTION —
B RTI) ERROR R F HET»
SIGNIFICANCE ELASTICITY
. ...._ VAUIAHLFS NOT IN 1HE EOUATJim
VAHJAHLE PARTIAL TOLF.HANCF f
SlUWlFICANtF.
-.PhtfOBltF-lia iP3H3tB5nF-PP. 4 1 P. -llhOIOF-n 1
.18b9lP33
3?.,HMIS7b -
.371P.H
i:
ALL VARIABLES ARE IN THE EQUATION:
-------
PAGE 11
FILE HARE (CREATION DATE = 77/r)3/ai.)
SUBFILF PEUGEOT
DEPENDENT VARIABLE.. NfjX
MULTIPLE
STEP
VARIABLE
HEHOVF.O
fiEH
TEMP
HT
f TO
OH HENOVt"
S.B31H?
J .
SUMMARY TAHIf
SIGNIFICANCE MULTIPLE M R SQUARE
Ci
. llHI
.782IH .SllSfi
.7HB8fi • .bJHtia
.KOHH3
SUHARE SIMPLE
CHANGE
-,7H?lfl
-.etilSS
-.7SH7S
OVERALL
SIGNlKICAHCt
31. J?H7? .(Hill
cii.HHl»J .null
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing)
i. REPORT NO.
EPA 600/2-77-116
2.
4. TITLE ANDSUBTITLE
LIGHT-DUTY DIESEL EMISSION CORRECTION FACTORS FOR
AMBIENT CONDITIONS
3. RECIPIENT'S ACCESSIOWNO.
5. REPORT DATE
July 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Charles T. Hare
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southwest Research Institute
8500 Culebra Road
San Antonio, Texas 78228
10. PROGRAM ELEMENT NO.
1AA601 BC-09 (FY-77)
11. CONTRACT/GRANT NO.
68-02-1777
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory-RTF, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Interim
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Since emission measurements from passenger cars are performed at one
standard set of ambient conditions and since emission rates of HC, CO, and
NO are sensitive to temperature and humidity, it is necessary to determine
the influence of ambient conditions on emissions from major classes of
vehicles. Although such information has been available for gasoline engine
powered cars for sometime, no such data were available for diesel powered
passenger cars.
This report indicates that diesel HC and CO emissions are relatively
insensitive to ambient conditions. Diesel NO emissions , however, are
sensitive to humidity but to a smaller extent than gasoline engines.
Humidity correction factors for NO emissions also appear to vary with
vehicle power-to-weight ratios andXare greater for higher powered vehicles.
17.
a.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
*Air pollution
*Automobiles
*Diesel Engines
*Exhaust emissions
*Correction
*Temperature
*Humidity
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
13B
13F
21G
2 IB
04B
3. ^iSTSiii'J^lo:. STATEVf NT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
21. NO. OF PAGES
106
20 SECURITY CLASS (Thispage\
UNCLASSIFIED
22. PRICE
EPA for IT. .-220-1 I9-/J)
96
------- |