Report No. SR00-03-01
Failure Rate Analyses and
Development of Fast-Pass,
Retest, and CPP Algorithms for
IM147 Max CO Cutpoints
prepared for:
U.S. Environmental Protection Agency
March 2000
prepared by:
Sierra Research, Inc.
1801 J Street
Sacramento, California 95814
(916) 444-6666
-------
Failure Rate Analyses and Development of
Fast-Pass, Retest, and CPP Algorithms
for IM147 Max CO Cutpoints
prepared for:
U.S. Environmental Protection Agency
Regional and State Programs Division
Office of Mobile Sources
Under Contract No. 68-C7-0051
Work Assignment No. 1-08
March 2000
prepared by:
Richard W. Joy
Garrett D; Torgerson
Thomas R. Carlson
Sierra Research, Inc.
1801 J Street
Sacramento, CA 95814
(916) 444-6666
-------
DISCLAIMER
Although the information described in this report has been funded wholly or in part
by the United States Environmental Protection Agency under Contract No. 68-C7-0051,
it has not been subjected to the Agency's peer and administrative review and is being
released for information purposes only. It therefore may not necessarily reflect the
views of the Agency and no official endorsement should be inferred.
-------
Failure Rate Analyses and Development of
Fast-Pass, Retest, and CPP Algorithms
for IM147 Max CO Cutpoints
Table of Contents
page
1. Summary 1
Scope of Work .2
Projected IM147 Failure Rates 3
Modal IM 147 Fast-Pass Standards 6
Predictive Retest Algorithms 6
Modal Fast-Fail Criteria 7
Integrated Fast-Pass, Retest, and Fast-Fail Algorithm Results 9
Segment 2 Revised Integrated Algorithms Results 9
Development of IM147 Driver Variation Standards 12
Integration of CPP Variation Limits 12
SIP Credit Analysis . 12
Need for Follow-Up Analysis ; 14
2. Introduction 15
Scope IB
Organization of the Report 20
3. Test Data 21
4. IM147 Cutpoint Analysis 26
IM147 Test Length ... ..;.... ;... 26
Cutpoint Analysis 27
Comparison of Failure Rates 30
5. Optimized IM147 Test Criteria 36
Modal Fast-Pass Standards 36
Predictive Retest Algorithms 40
Modal Fast-Fail Criteria 43
Integration of Fast-Pass, Retest, and Fast-Fail Algorithms 50
Segment 2 Revised Integrated Algorithms Results 53
SIP Credit Analysis 53
Need for Follow-Up Analysis 54
-------
Table of Contents (continued')
page
6. Development of IM147 Variation Limits 55
Existing Tolerance Limits 55
Previous Analysis of Speed Variation Limits 57
Development of Positive Power-Based Variation Limits 58
Evaluation of IM147 Variation Limits on Test Time 66
7. Integration of CPP Variation Limits 68
8. References 71
Appendix A - Startup, Intermediate and Final IM240 and IM147 Cutpoints
Appendix B - Max CO, Startup, Intermediate and Final IM147 Failure Rates
Appendix C - IM 147 Regression Coefficients
Appendix D - Excess Emissions Identification for Max CO Cutpoints
Appendix E - Second-by-Second CPP Variation Limits
Appendix F - Regression Summaries
Appendix G - Fleet Distribution Data
-ii-
-------
List of Tables
page
1-1 Failure Rates, Third IM147 3
1-2 Predicted Arizona IM147 Failure Rates with MaxCO Outpoints
and Integrated Algorithms 5
1-3 Modeled Fast-Pass Results Excess Emissions vs. Average Test Time 7
1-4 Retest Algorithm Results 8
1-5 Fast-Fail Algorithm 8
1-6 Comparison of Integrated Algorithms vs. Standard IM147 Impact on
Test Time and Excess IM240 Emissions (MaxCO Outpoints) Lost 11
1-7 Average Test Time by Test Time Reduction Methodology and CPP 13
1-8 Comparison of IM147 Max CO Outpoints to IM240 Final Standards
Impact on Excess Emissions Lost 13
4-1 IM240 to IM147 Composite Regression Equation Coefficients 28
4-2 IM147 Composite to Phase 2 Regression Equation Coefficients 29
4-3 Revised IM147 Max CO Outpoints (Composite/Phase 2) 30
4-4 Failure Rates, Third IM147 31
5-1 Revised IM147 Segments . . 38
5-2 Modeled Fast-Pass Results Excess Emissions vs. Average Test Time 40
5-3 Excess IM240 (Max CO Outpoints) Emissions Identified
. (With and Without Fast-Pass) 41
5-4 Retest Algorithm Results 48
5-5 Fast-Fail Algorithm 50
5-6 Comparison of Integrated Algorithms (No CPP) vs. Standard IM147
Impact on Test Time and Excess IM240 Emissions (Max CO Outpoints) Lost ... 51
5-7 Comparison of IM147 Max CO Cutpoints to IM240 Final Standards
Impact on Excess Emissions Lost 54
6-1 Preliminary CPP Cutpoints Based on Top-50% Drivers 63
6-2 Centered CPP Cutpoints and Resulting Test Alport Rates 64
7-1 End Test Decisions Affected by CPP Errors 68
-iii-
-------
List of Figures
page
1-1 Modal IM147 Fast-Pass Standards 4
1-2 Integrated Test Results Maximum CO Outpoints Without CPP Limits 10
3-1 Test Sequence Used to Investigate Triplicate
IM147 Tests in Arizona Test Lanes 21
3-2 Model Year Distribution of Sample Fleet and AZ Overall Fleet 24
4-1 IM147 Trace Phase 1 and Phase 2 27
4-2 Failure Rate by Consecutive IM147 Max CO Cutpoints 32
4-3 Failure Rate by Consecutive IM147 Startup Cutpoints 33
4-4 Failure.Rate by Consecutive IM147 Intermediate Cutpoints 34
4-5 Failure Rate by Consecutive IM147 Final Cutpoints .35
5-1 EM147 Test Segments Used for Fast-Pass Cutpoint Development 39
5-2 Retest Algorithm During First IM147-LDGV 44
5-3 Retest Algorithm During Second LM147-LDGV 45
5-4 Retest Algorithm During First IM147-LDGT1, LDGT2 46
5-5 Retest Algorithm During Second IM147-LDGT1, LDGT2 47
5-6 Integrated Test Results Max CO Cutpoints Without CPP Limits 52
6-1 Comparison of IM147 Variation Limit Metrics , 60
6-2 Distribution of Arizona IM147 Positive Power Differences 62
6-3 Illustration of Second-by-Second IM147 Cumulative Positive
Power Variation Limits 65
6-4 Effect of Initial CPP Limits Multiplier on Effective IM147 Abort Rate 66
7-1 Process for Integrating CPP Variation Limits 70
-iv-
-------
1. SUMMARY
Under the Clean Air Act Amendments of 1990, metropolitan areas with the most serious
air quality problems are required to implement so-called "enhanced" I/M programs. Two
different test procedures for exhaust emissions testing in enhanced programs have been
approved by EPA: the "IM240" test, and the "Acceleration Simulation Mode" (ASM)
test. With either procedure, the efficiency of the testing process depends on how quickly
accurate decisions can be made as to whether a vehicle should pass or fail.
Inadequate vehicle preconditioning has previously been identified as a cause of false
failures in I/M programs. In fact, previous EPA and DEQ analyses estimate that 25% of
the vehicles failing the final IM240 standards would pass with further preconditioning,
and that these vehicles can be identified through modal analysis. To address this
problem, Sierra suggested, in another study, that Phase 1 of the IM240 test be eliminated,
instead using only the second hill of the IM240 (i.e., the "IM147") up to three times in
succession to ensure adequate preconditioning. Based on this recommendation, and to
address the issue of inadequate preconditioning without compromising test throughput
excessively, Arizona therefore decided to change its test procedure to the IM147
beginning January 1, 2000. This program upgrade will also include implementation of
"Max CO" cutpoints previously developed by Sierra, which are designed to maximize the
CO benefits of the program.
Regarding this proposed alternative to the existing IM240 test, several concerns needed to
be addressed prior to implementation in Arizona. First of all, how would emissions
identification and credit change with the new procedure. Secondly, because the length of
the proposed test may be equivalent to that of three IM147 tests, it can use considerably
more dynamometer time than the other aforementioned tests, thus increasing the cost of
the program. Prior to this study, two other studies have already been conducted to start
addressing these issues.
The first study addressed, among other things, reducing test time and projected emission
credit levels for the IM147 test. Part of the data analyzed in this study consisted of 101
tests where vehicles were given three back-to-back IM147 tests. These data were used to
create "Phase 2b" cutpoints, which are analogous to phase 2 cutpoints for the IM240 test,
thus giving vehicles two ways to pass. In addition, fast-pass cutpoints for both the entire
IM147 as well as Phase 2b were also developed. The other data used in this study
consisted of 2% random sample IM240 test data collected in the Arizona IM240 program.
These data, in combination with the 101 vehicle data, were used to project emission
credits. Unfortunately, since this study did not include back-to-back IM147 to IM240
testing, excess emissions identification rates and SIP credit could not be conclusively
-1-
-------
established. In a follow-up work assignment in 1998,1* EPA asked Sierra to evaluate 304
triplicate IM147 tests followed by an IM240 test. These data were analyzed to verify the
preliminary excess emission identification rates and average test time estimates projected
for the Arizona IM program from the Phase 2 and IM240 data sets collected in previous
studies. In addition, improved fast-pass and retest algorithms were developed for the
IM147 test using the same approach used previously in developing similar IM240
algorithms; however, the cutpoints scenarios evaluated in the study did not include the
Max CO cutpoints previously developed for DEQ.
PKE Speed Variation Criteria - In the 1997 IM240-related evaluation for EPA (SR98-02-
01),2 Sierra developed improved speed variation criteria based on the total Positive
Kinetic Energy (PKE) change per mile traveled during the IM240 cycle. These criteria
were designed to minimize the variation in emissions while still being feasible for use by
minimally trained drivers with a reasonable aptitude for dynamometer driving. However,
only IM240 drive cycle criteria were developed in the 1997 study. Therefore, further
analysis was needed to develop similar speed variation criteria for the IM147.
Scope of Work
To aid in the Arizona IM147 implementation effort, EPA issued a work assignment
(#1-08) to Sierra to complete the following tasks:
1. Develop projected IM147 failure rates for the Arizona I/M program;
2. Develop modal IM147 fast-pass standards for the Max CO cutpoints;
3. Develop modal predictive IM147 retest algorithms for the Max CO cutpoints;
4. Develop modal IM147 fast-fail criteria; and
5. Develop fast-pass and full-duration PKE criteria for the IM147 test.
Three distinct data sets were used in this study. The first two sets were the 304 vehicle
study, collected for SR99-10-02, and a 543-vehicle sample collected for this study. For
both of these sets, randomly selected vehicles were given triplicate IM147 tests followed
by an IM240 test. The third data set comprised 2,518 vehicles given triplicate IM147
tests and, if they failed the third test, an IM240 test. After removing invalid tests, the test
data sets used for this study consisted of 300, 535, and 2,512 vehicles, respectively (i.e.,
3,347 vehicles total). All of the data were collected by Gordon-Darby I/M lanes in
Phoenix, Arizona.
Superscripts denote references listed in Section 8.
-2-
-------
Projected IM147 Failure Rates
Prior to projecting failure rates for the IM147 test using each of the four set of emissions
standards (Startup, Intermediate, Final, Max CO), Max CO cutpoints developed as part of
SR99-10-02 were revised using the combined 300- and 535-vehicle data sets. To this
end, IM147 scores were regressed against IM240 scores. The resulting regression
equations were then used to derive IM147 cutpoints from the IM240 cutpoints. IM147
phase 2 cutpoints were developed similarly by regressing IM147 composite scores
against IM147 phase 2 scores. A scaling factor of 0.9 was multiplied against the
predicted phase 2 cutpoint to make the phase 2 cutpoint slightly more stringent than the
composite cutpoint, as it is with the IM240 test. Table 4-4 in Section 4 shows the revised
Max CO cutpoints.
After revising the Max CO cutpoints, failure rates for the IM147 test using the Startup,
Intermediate, final, and Max CO cutpoints were determined using the 3,347-vehicle data
set. These failure rates, which are shown in Table 1-1, were based upon the results of the
third IM147 test.
Table 1-1
Failure Rates, Third IM147
HC
CO
NOx
OVERALL
Vehicle
Type
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Max CO Cutpoints
LDGV
114
1666
6.4%
250
1530
14.0%
155
1625
8.7%
390
1390
21.9%
LDGT1
49
915
5.1%
156
808
16.2%
70
894
7.3%
219
745
22.7%
LDGT2
25
578
4.1%
63
540
10.4%
30
573
5.0%
93
510
15.4%
All
188
3159
5.6%
469
2878
14.0%
255
3092
7.6%
702
2645
21.0%
Startup Cutpoints
LDGV
106
1674
6.0%
130
1650
7.3%
141
1639
7.9%
281
1499
15.8%
LDGT1
52
912
5.4%
33
931
3.4%
48
916
5.0%
108
856
11.2%
LDGT2
33
570
5.5%
30
573
5.0%
28
575
4.6%
67
536
11.1%
All
191
3156
5.7%
193
3154
5.8%
217
3130
6.5%
456
2891
13.6%
Intermediate Cutpoints
LDGV
157
1623
8.8%
161
1619
9.0%
196
1584
11.0%
367
1413
20.6%
LDGT1
80
884
8.3%
53
911
5.5%
73
891
7.6%
165
799
17.1%
LDGT2
40
563
6.6%
42
561
7.0%
42
561
7.0%
93
510
15.4%
All
277
3070
8.3%
256
3091
7.6%
311
3036
9.3%
625
2722
18.7%
Final Cutpoints
LDGV
280
1500
15.7%
240
1540
13.5%
277
1503
15.6%
507
1273
28.5%
LDGT1
120
844
12.4%
85
879
8.8%
127
837
13.2%
239
725
24.8%
LDGT2
73
530
12.1%
56
547
9.3%
82
521
13.6%
149
454
24.7%
All
473
2874
14.1%
381
2966
11.4%
486
2861
14.5%
895
2452
26.7%
-3-
-------
This table helps clarify the relationship between the four sets of outpoints. The Startup,
Intermediate, and Final outpoints result in nearly equal HC, CO, and NOx failure rates, by
vehicle category, which increase as the stringency of the cutpoints increase. The Max CO
standards, when compared to the final standards, result in lower HC and NOx failure
rates, and higher CO failure rates, especially for light-duty trucks.
.Figure 1-1 shows the failure rates for each of the three IM147s, based on the Max CO
standards. As the figure shows, most of the decrease in failure rates due to the use of
multiple test cycles (i.e., to address the lack of adequate preconditioning) occurs between
the first and second IM147s. As a result, it is reasonable to expect that algorithms
designed to shorten the test would often end the test prior to the third IM147. This is, in
fact, the case, as is shown later in this report.
Figure 1-1
Overall Failure Rate by Vehicle Type
Max CO Standards
LDT1 LDT2
Vehicle Type
~ 1st IM147
02nd IM147
03rdIM147
Later in this report, implementation of fast-pass, fast-fail, and retest algorithms will be
discussed. The addition of the algorithms has a significant impact on failure rate. In
predicting which vehicles will ultimately pass the test, minor predictive errors associated
with the fast-pass algorithm lead to a small fraction of false passes and in turn a reduction
in the failure rate. The fast-fail and retest algorithms, which are used to predict failing
vehicles, will err on the side of false failures, hence acting to increase the failure rate.
One of the goals in developing the revised Max CO cutpoints was to minimize the
projected false failure rate; however, false passes were treated somewhat differently.
Rather than attempting to keep the number of false passes to a minimum, the study
instead focused on maximizing excess emissions identification while also minimizing test
time. As a result, the fast-pass algorithms allow a number of marginal vehicles that have
little impact on the overall excess emissions to falsely pass. While relatively insignificant
-4-
-------
from an emissions perspective, this results in a fairly substantial decrease in the projected
failure rate.
There is also a second factor affecting the projected failure rates shown in Table 1-1.
Review of the individual results for each of the three sequential IM147s shows that
roughly 80 vehicles (i.e., about 2.5% of the total sample) failed the third IM147 after
having passed the first test cycle. The reason for this degradation in emissions over the
three IM147s is unknown, but appears to be an artifact of the test protocol. Since these
vehicles would have passed out as a result of the first EM 147, they are considered false
failures that would not occur with the addition of the integrated algorithms.
Table 1-2 shows the effect on projected failure rates of applying the algorithms to the
3,347-vehicle sample. The failure rates shown in the table have also been adjusted by
normalizing the results to the vehicle fleet distribution contained in 2% random sample
data collected in the Arizona program from July 1997 through March 1998. (These were
the most recent vehicle type-specific data readily available to Sierra.) This fleet
distribution is detailed in Appendix G.
Table 1-2
Predicted Arizona IM147 Failure Rates with
MaxCO Cutpoints and Integrated Algorithms
Vehicle Class
Predicted Failure Rate*
LDGV
16.7%
LDGT1
11.4%
LDGT2
10.4%
Overall
14.8%
Normalized to Arizona fleet distribution obtained from 2% random
sample data for the period July 1997 - March 1998.
Historical Arizona IM240 test results also show a significant seasonal effect. This is
believed to be caused by two primary factors: the impact of changes in ambient
temperatures on the purging of fuel vapors to the canister (in some older models these
vapors are emitted directly to the engine), and the wintertime use of oxygenated gasoline.
Both factors will act to reduce wintertime failure rates. At the high ambient temperatures
typically experienced in Phoenix in the summer and fall, high purge rates from vehicles
idling in the queue lead to canister overloading and breakthrough. This in turn results in
higher emissions and an increase in projected failure rates. Older-technology vehicles are
particularly susceptible to this phenomenon, due to the vehicles' poorer fuel delivery
capabilities. This effect will not occur nearly as often during colder temperatures, thus
-5-
-------
contributing to lower wintertime failure rates. The use of oxygenated gasoline reduces
CO emissions, leading to further decreases in failure rates.
A review of historic failure rate data for the Arizona program shows that this seasonal
effect appears to have a 2%-2.5% impact on failure rates. For example, the IM240 failure
rate in January 1999 was 14.0% versus a failure rate of roughly 16%-16.5% .during the
July-September 1999 period. Since most of the data analyzed in this study were collected
during late summer, the wintertime failure rate would be expected to be significantly
lower than the projections shown in Table 1-2.
Modal IM147 Fast-Pass Standards
Using the same methodology employed in SR99-10-02, this study developed fast-pass
regression standards for the Max CO cutpoints. This methodology requires dividing the
IM147 test into a series of short segments over which emission mass is accumulated. By
performing multivariate linear regressions of these incremental segments, modal fast-pass
coefficients were developed to predict when vehicles would pass without having to
complete the entire test. Unlike the previous study, which divided the IM147 test into 14
segments, this study divided the test into 20 segments, thus increasing the frequency of
opportunities for fast-pass.
While a fast-pass procedure of this nature has the ability to greatly reduce test time, this
reduction has to be balanced against false passes. False passes occur when vehicles that
would otherwise fail an inspection are fast-passed out because their emissions over the
drive cycle are not appropriately characterized by the regression. For this study, false
passes are quantified by measuring excess emissions, which are defined as emissions
collected during an IM240 test in excess of the applicable standard for a given vehicle.
The IM147 test receives credit for identifying excess emissions if it fails a vehicle that
had excess IM240 emissions when using the same emission cutpoints (e.g. Max CO).
In this study, both the regression of the IM147 segments and the analysis of that
regression were performed using the combined data sets (3,347 vehicles) and the Max CO
standards. Table 1-3 details the results of this analysis.
Predictive Retest Algorithms
The workplan for the study called for Sierra to refine algorithms originally developed for
SR99-10-02 and then to apply them to the total vehicle sample (3,347 vehicles) to
determine their net effect on test time. In contrast to the fast-pass algorithm, where
misidentification results in false-passing vehicles and a loss in excess emissions
identification, the retest algorithm errors result in false failures, which can lead to
consumer complaints. Like the fast-pass algorithm, decreases in test time need to be
weighed against false failures to determine a reasonable compromise.
-6-
-------
Table 1-3
Modeled Fast-Pass Results
Excess IM240 Emissions3 (w/ MaxCO Cutpoints) vs. Average Test Time
No Fast-Passb
Fast-Pass Enabled
Excess HC Identified
97.8%
95.7%
Excess CO Identified
96.3%
92.7%
Excess NOx Identified
95.4%
82.4%
Average Test Time0
217 seconds
125 seconds
a Excess emissions normalized to Arizona 2% random sample fleet distribution data, July 1997 to March
1998.
b While vehicles cannot terminate in the middle of an IM147 without the fast-pass algorithm, the test may
end prior to completing three IM147s if the emissions measured at the end of any one of the IM147s
meet the applicable standards.
c Test time refers to the time actually required to operate the vehicle on the dynamometer.
While the original retest procedure described in SR99-10-02 used a combination of mass
and concentration emissions measurements to anticipate whether a vehicle would benefit
from additional testing, concerns expressed by Gordon-Darby regarding the complexity
of this algorithm led to a different approach for this analysis. Instead, a variation of the
fast-pass regression calculation was developed and used to predict emissions
improvement over an IM147 test. In short, the emissions result predicted after Segment 7
is compared to the emissions result predicted after Segment 19 to determine whether the
vehicle emissions are converging on the applicable standard.
Using this procedure, test time was reduced from 125 seconds (with fast-pass enabled) to
96 seconds. Table 1-4 details further results of the retest analysis.
Modal Fast-Fail Criteria
One of the requirements of this work assignment was to develop modal fast-fail criteria.
Unlike the retest procedure, which can terminate the test at the ends of the individual
IM147s, the fast-fail algorithm can terminate tests during an IM147 test.
Like the retest algorithm, errors committed by the fast-fail criteria result in false failures.
As a result, decreases in test time have to be weighed against false failures. With this in
mind, fast-failures cannot be made during the first of the three IM147s. Analysis showed
that the results of the first IM147 were too unpredictable relative to the final result to risk
false-failing vehicles. The final two IM147s, however, would serve reasonably well for
this purpose.
-7-
-------
Table 1-4
Retest Algorithm Results
LDGV
LDGT1, LDGT2
Total Number of Complete
Tests
1567
1780
# of Failures Without
Retest Algorithm
327 (20.9% of 1567)
273 (15.3% of 1780)
# of Correctly Identified
Failures3
245 (74.9% of 327)b
180 (65.9% of 273)b
# Failing After 1 IM147
94 (38.4% of 245)c
97 (53.9% of 180)c
# Failing After 2 IM147s
151 (61.6% of 245)c
83 (46.1% of 180)°
# of Passing Vehicles
Falsely Failed by Retest
0(0% of 1567)
0(0% of 1780)
a "Correctly identified failures" refers to those vehicles that were still failing at the end of the third IM147.
bThe number shown in parentheses is the number of failures without the retest algorithm.
cThe number shown in parentheses is the total number of IM147 Cycle 2 and 3 failures.
There are two different fast-fail algorithms, one for each of the final two IM147 tests.
The fast-fail algorithm for the second IM147 fails vehicles with excessively high
predicted emissions after segment 7 of the second IM147. The fast-fail algorithm for the
third IM147 test uses a variation of the fast-pass algorithm to predict failing vehicles
throughout the test. Both of these algorithms are described more completely in the body
of this report.
Table 1-5 shows the results of the fast-fail algorithms when applied to the 3,347-vehicle
sample. The fast-fail algorithm reduces average test time an additional 2 seconds, from
96 seconds (with fast-pass and retest enabled) to 94 (with fast-pass, retest, and fast-fail
enabled).
Table 1-5
Fast-Fail Algorithm .
Vehicle
Class
Second IM147 Fast-
Failures
Third IM147 Fast-
Failures
False Failures
LDV
124
211
3
LDT
99
182
3
Total
223
393
6
-8-
-------
Integrated Fast-Pass. Retest. and Fast-Fail Algorithm Results
This portion of the study combined all of the optimized algorithms to determine their net
effect on test time, false failures, and excess emissions while using the Max CO standards
with the 3,347-vehicle sample. The flow chart contained in Figure 1-2 shows the point at
which vehicles concluded the test and the reason they passed or failed. It also indicates
the average dynamometer test time for each category of vehicles. Following the flow
chart, Table 1-6 shows the net effect of the procedures on test time and excess emissions.
The table shows that excess emission identification using the Max CO cutpoints with the
fast-pass, retest, or fail-fail algorithms enabled is 95.9%, 93.1%, and 97.0% for HC, CO,
and NOx, respectively. (For comparison, the respective identification rates are 97.8%,
96.3%, and 95.4% without the various test criteria enabled.) This is down from the
identification rates developed in SR99-10-02,1 which identified 99.6% of the HC, 98.2%
of the CO, and 99.9% of the NOx with the fast-pass and retest algorithms enabled (fast-
fail was not considered). However, direct comparison of the two sets of results may not
be relevant for several reasons. First, the previous study measured excess emissions
captured against the Final Cutpoints rather than the Max CO cutpoints use for this study.
Second, because fast-fail was created for this study, it was not included in the previous
study results. Third, the retest algorithm has been modified as part of this study and will
therefore have a different effect on the results. Finally, the majority of the data used in
this study were collected with the newer model year exemptions in place and normalized
to the vehicle inspection fleet distribution for the period July 1997 to March 1998. As a
result, the vehicle distribution was skewed toward older vehicles relative to that in the
previous study. Less rigorous test criteria were also evaluated as part of this latest study.
However, their use yielded relatively little improvement in identification rate at the cost
of a large increase in test time. It was therefore decided not to pursue this latter option.
Segment 2 Revised Integrated Algorithms Results
The original integrated algorithm results were initially determined assuming fast-pass and
fast-fail results could not be rendered prior to the fourth segment (i.e., no earlier than Test
Time = 28 seconds). This was consistent with the procedure established in SR98-02-01.
Gordon-Darby, wishing to further minimize test time, requested that Sierra explore the
feasibility of rendering fast-pass and third IM147 fast-fail decisions after earlier segments
without degrading excess emission identification. Further investigation found that
decisions could be made as early as the end of segment 2 (i.e., at Test Time = 16 seconds)
if the error multiplier used in the fast-pass decision was increased during segments 2 and
3. For segment 2, the error multiplier was 3, while it was 2.5 for segment 3. Using these
criteria, average test time was reduced to 91 seconds without sacrificing any excess
emissions identification.
-9-
-------
Figure 1-2
Integrated Test Results
Maximum CO Cutpoints Without CPP Limits
TOTAL PASS = 2,802 TOTAL FAIL = 545
TOTAL FALSE FAILURES = 3
-10-
-------
Table 1-6
Comparison of Integrated Algorithms vs. Standard IM147
Impact on Test Time and Excess IM240 Emissions (Max CO Cutpoints) Lost2
Class
Model
Year
Group
Sample
Sizeb
Mean Test
Time
Standardb
Mean Test
Time w/
Algorithmsb
% Excess Emissions Identified'
HC
CO
NOx
LDGV
1981-82
105
286.4
140.1
-
87.6%
100.0%
1983-85
228
311.2
169.4
100.0%
97.6%
100.0%
1986-89
425
248.4
112.0
99.2%
97.1%
100.0%
1990-95
952
184.8
78.3
96.8%
85.3%
91.2%
1996+
70
154.3
37.1
-
-
-
All
1780
221.0
100.0
98.2%
93.3%
96.8%
LDGT1
1981-85
260
306.6
158.7
78.5%
97.3%
99.4%
1986-89
222
230.8
101.6
93.8%
90.8%
99.6%
1990-95
450
173.3
59.2
0.0% .
0.0%
100.0%
1996+
32
155.1
31.6
-
-
100.0%
All
964
221.9
94.9
83.4%
92.6%
99.5%
LDGT2
1981-85
94
307.5
158.8
100.0%
100.0%
77.2%
1986-87
64
253.2
101.4
100.0%
100.0%
100.0%
1988-95
427
166.5
54.2
-
0%
44.4%
1996+
18
146.0
28.0
-
-
-
All
603
197.1
74.7
100.0%
93.8%
83.1%
Weighted Average
3,347
217.0
94.0
95.9%
93.1%
97.0%
3 Test time results do not include impact of driver variation limits.
b Mean test time standard refers to the average dynamometer test time without the algorithms enabled.
This was determined using the 3,347-vehicle sample.
c Percent of IM240 (Max CO) excess emissions identified with the integrated algorithms enabled. This
was determined using the 835-vehicle sample and normalized to the Arizona 2% random sample fleet
distribution data, July 1997 to March 1998.
-11-
-------
Development of IM147 Driver Variation Standards
As will be shown in the body of this report, previous work performed on driver variation
limits utilized Positive Kinetic Energy (PKE) limits to evaluate driver performance.
Analysis conducted for this study, however, revealed limitations with this metric, which
can result in inappropriate driver errors. As a result, a new statistic, Cumulative Positive
Power (CPP), was developed to remedy this problem. Designed to be used in
conjunction with EPA-specified absolute speed variation limits, CPP produces an
improved, more predictable, driver evaluation criteria than PKE.
When applied to the total vehicle sample for this study (minus vehicles with absolute
speed violations greater than ą 2 mph), the new CPP criteria produced a total abort rate of
3.4%. Since the IM147 test, even without fast-pass, fast-fail, or retest, allows for vehicles
to pass the test after a single passing IM147, many of the driver errors in the total sample
would not be experienced since they occurred after the vehicle had already passed.
Taking this into account, the effective abort rate was 2%.
Integration of CPP Variation Limits
The CPP analysis was conducted on a subset of the 3,347-vehicle population, with
absolute speed excursion violations (as defined in EPA's IM240 guidance) removed.
This resulted in an overall data set of 3,006 vehicles. Using the 3,006-vehicle sample, the
average test time, with the fast-pass, retest, and fast-fail criteria enabled but without the
CPP criteria applied, was 89 seconds. Once the CPP criteria were enabled, 87 tests (of
the 3,006 vehicles) were extended, increasing the average test time by 1 second to 90
seconds. This resulting increase of 1.1% is less than the 2% increase projected at the end
of Section 6, which makes sense given that the 2% projection was made without the fast-
pass/fail and retest algorithms in place. The overall test time reductions caused by the
fast-pass/fail and retest algorithms would mean that fewer errors would be committed.
Table 1-7 summarizes the change in dynamometer test time with each succeeding set of
enabled criteria. The overall impact of all the criteria is to reduce the test time by 58%,
from 217 to 90 seconds.
SIP Credit Analysis
The above comparison of excess emissions identification between the IM240 and IM147
is based on the use of CO Max standards for both test cycles. To develop an estimate of
the allowable SIP credit that should be allocated to the revised IM147 CO Max standards,
it is also necessary to compare excess emissions identification between this scenario and
the IM240 with EPA-recommended final cutpoints in place. This is due to the need to
establish a link to using MOBILE for SIP modeling purposes. Configuring MOBILE
with CO Max standards is not feasible; therefore, a better approach is to run the model
with final EPA standards in place and use the excess emissions identification rates
developed in this study to adjust the resulting model outputs.
-12-
-------
Table 1-7
Average Test Time (seconds)
by Test Time Reduction Methodology and CPP
Scenario
Dyno Test
Time
Test Time
Reduction (%)
Cutpoint only, two possible retests
217
Added fast-pass
125
42
Added retest
96
56
Added fast-fail
94
57
Allowed fast-pass at end of Segment 2
91
58
Removed speed excursion violations
89
59
Added CPP limits
90
58
Table 1-8 shows the excess emission identification rates when the IM147 Max CO
cutpoints are compared to the IM240 final standards. Pollutant-specific identification
rates are shown both without and with the fast-pass, retest, or fail-fail algorithms enabled.
(The latter scenario includes fast-passing vehicles as early as at the end of segment 2.)
Since Arizona will be implementing the IM147 test procedure with the algorithms
enabled, the identification rates for this scenario are the ones that should be used to adjust
the MOBILE modeling results (based on final IM240 standards) for SEP credit purposes.
Table 1-8
Comparison of IM147 Max CO Cutpoints to IM240 Final Standards
Impact on Excess Emissions Lost3
Class
% Excess Emission Identified
(Without Fast-Pass, Retest, Fast-Fail)
% Excess Emissions Identified
(With Integrated Algorithms)
HC
CO
NOx
HC
CO
NOx
LDGV
95.2%
96.2%
84.8%
91.9%
95.1%
85.7%
LDGT1
80.5%
100.0%
70.6%
68.4%
98.1%
81.2%
LDGT2
87.9%
100.0%
46.6%
98.5%
97.7%
73.0%
Weighted
Average
91.3%
97.3%
79.9%
86.7%
95.9%
81.6%
a Percent of IM240 (Final Standards) excess emissions identified was determined using the 835 vehicle
sample. Data normalized to Arizona 2% random sample fleet distribution data, July 1997 to March
1998.
-13-
-------
As expected, the table shows that HC and NOx identification rates are significantly lower
with the IM147 Max CO cutpoints relative to final IM240 standards. This is due to the
fact that the Max CO cutpoints are designed to maximize the CO benefits of the program
at the expense of HC and NOx benefits, while keeping maximum failure rates in each
cutpoint category to acceptable levels. The CO identification rate of 95.9% (with the
algorithms enabled) shows that the Arizona program will achieve nearly all of the
modeled benefit of the final IM240 standards. Note that this will be substantially more
effective than the current phase-in IM240 standards. The table also shows that the
addition of the integrated algorithms results in little more than a 1% reduction in the
excess emissions identification rate for CO. (As noted above, the addition of the
algorithms reduces dynamometer test time from 217 to 90 seconds.).
Need for Follow-Up Analysis
As discussed above, the analysis results presented in the report are based on a relatively
small sample of IM147 and IM240 data. While the available data are significantly more
robust than the previous sample of 300 vehicles, it is clear that these results should be
revisited with a much larger sample once IM147 testing is initiated in Arizona. We
therefore recommend that as soon as one to two months of IM147 data are collected in
the program, they should be used to verify the validity of the cutpoints and algorithms
developed in this study. This follow-up analysis would allow for any required fine-
tuning of the cutpoints and algorithms.
IT It it
-14-
-------
2. INTRODUCTION
Under the Clean Air Act Amendments of 1990, metropolitan areas with the most serious
air quality problems are required to implement so-called "enhanced" I/M programs. One
element of an enhanced program is a more effective test procedure, than the simple idle
tests used in "basic" I/M programs. Two different test procedures for exhaust emissions
testing in enhanced programs have been approved by EPA: the "IM240" test, and the
"Acceleration Simulation Mode" (ASM) test. Both of these procedures have been shown
to be capable of separating vehicles with excessive exhaust emissions from other
vehicles; however, the accuracy of the test depends on whether tested vehicles have been
adequately preconditioned and whether the speed-time profile associated with each test
procedure is closely followed. With either procedure, the efficiency of the testing process
depends on how quickly accurate decisions can be made as to whether a vehicle should
pass or fail.
Inadequate preconditioning of vehicles prior to testing is a potential cause of inaccurate
or inconsistent test results because exhaust emission levels depend on how thoroughly a
vehicle has been warmed up. Before the vehicle is thoroughly warmed up, high
emissions can be caused by air-fuel ratio enrichment or an inactive catalytic converter. In
addition, increased emissions due to purging of loaded canisters may also be an issue
associated with inadequate preconditioning prior to I/M testing.
Inadequate vehicle preconditioning has previously been identified as a cause of false
failures in I/M programs. Under current EPA guidance, IM240 preconditioning
procedures are woven into the "two-ways-to-pass" standards. Vehicles that exceed the
emissions standards established for the entire 239-second test are passed or failed based
on emissions occurring during the last 147 seconds of the test (also called Phase 2 or the
IM147). The separate set of standards that applies to Phase 2 is slightly more stringent.
For vehicles that initially demonstrate high emissions, the first 93 seconds (Phase 1) of
the test are used to precondition the vehicle for the second phase of the test. In addition,
EPA calls for a "second-chance" test whenever a vehicle fails the initial test by less than
50% of the standard and was in a queue for more than 20 minutes before being tested.
Previous EPA and DEO Analyses - Considerable data have already been collected
regarding the preconditioning requirements for IM240 testing. During 1996 and 1997,
Sierra conducted evaluations of this issue using data obtained from samples of vehicles
recruited from IM240 lanes in Phoenix, Arizona, and a laboratory test program at Sierra's
facilities in Sacramento. The results of the 1996 analysis were reported in SAE Paper
No. 962091.3 The 1997 evaluation also included an analysis of the effect on test duration
-------
of adopting EPA-recommended "final" IM240 outpoints. Preliminary conclusions from
the two evaluations are summarized below.
1. Using the current IM240 test procedures, it is estimated that 25% of the
vehicles failing the final IM240 standards would pass with further
preconditioning.
2. Vehicles that would benefit from further preconditioning can be identified
through modal analysis of the emissions recorded during the IM240 test.
3. Two possible approaches to modifying the current preconditioning procedures
would be to:
a. retain existing IM240 test procedure and two-ways-to-pass standards, with
the entire IM240 to be repeated if the Phase 2 emissions failure is marginal,
emissions near the end of Phase 2 are relatively low, or emissions during
Phase 2 are significantly lower than during Phase 1; or
b. eliminate Phase 1 and make the initial pass/fail decision based on.running
only the IM147, with a second-chance test (another IM147) for all vehicles
that initially fail, and a third-chance IM147 test if emissions during the
second-chance test are significantly lower than emissions during the initial
test.
4. Adoption of final cutpoints and more effective preconditioning procedures
involving a second full-IM240 (Option 3.a. above) will increase the portion of
the test involving dynamometer operation by more than 100%.
The 1997 evaluation also involved the development of improved IM240 fast-pass
cutpoints using a modal regression approach originally pioneered by the New York
Department of Environmental Conservation (NYDEC).4 This study also involved the
development of modal predictive IM240 retest algorithms designed to minimize the
fraction of vehicles either (1) identified as needing a retest when they would still fail, or
(2) not identified as needing a retest when they would have passed if retested.
As a follow-up to the 1997 evaluation for EPA, Sierra subsequently conducted an
analysis (SR98-05-01)5 for the Arizona Department of Environmental Quality (DEQ) of
the effect on failure rates, I/M program benefits, and test duration of the following
changes to the current IM240 procedure: (1) implementation of the Option 3.b
preconditioning procedures summarized above; (2) adoption of interim ("Max CO")
cutpoints designed to maximize the carbon monoxide emission reduction benefits being
achieved by the program; and (3) the exemption of either the first four or first five model
years from program requirements.
To analyze the IM147-only preconditioning option, Sierra used a combination of data
from the 2% random test sample (consisting entirely of full-duration tests) that is
-16-
-------
routinely collected in the Arizona IM240 program and a limited number (101 tests) of
triplicate (back-to-back-to-back) IM147 tests that were conducted as part of the 1997
EPA evaluation. A key element of the analysis methodology involved the development
of "Phase 2b" cutpoints to complement the full IM147-only cutpoints. (The Phase 2b
cutpoints were applied in a manner similar to the current IM240 procedure in which
vehicles passing in Phase 2 are considered passing for the entire test.) Fast-pass cutpoints
for both the entire IM147 and Phase 2b were also developed.
As noted above, while the 1997 EPA study involved the analysis of a considerable
amount of IM240 data, only a small subset (101 vehicles) was of use in projecting credit
levels and test times for an IM147 test program. An additional concern is that the 101-
vehicle study was not specifically designed to determine excess emission identification
rates and SIP credit levels. As a result, EPA issued a follow-up work assignment to
Sierra in 1998 that involved the collection of test data from triplicate IM147 tests
followed immediately by a full-duration IM240. Data were collected from 304 randomly
selected light-duty cars and trucks arriving at the test lane during normal queuing
conditions. These data were then analyzed to verify the preliminary excess emission
identification rates and average test time estimates projected for the Arizona IM program
from the Phase 2 and IM240 data sets collected in previous studies.
As part of the 1998 study for EPA, improved fast-pass and retest algorithms for the
IM147 were developed using the same approach used previously in developing similar
IM240 algorithms; however, the cutpoint scenarios evaluated in the study did not include
the Max CO cutpoints previously developed for DEQ. Given Arizona's need for the
maximum feasible CO reductions from its I/M program, DEQ has decided to implement
this set of cutpoints. A follow-up study is therefore needed to develop improved fast-pass
and retest algorithms for the Max CO IM147 cutpoints. Gordon-Darby collected
additional test data from roughly 3,000 vehicles in the Phoenix area that can be used in
this analysis. Of these vehicles, approximately 2,500 received triplicate IM147s only; the
remaining 500 vehicles will receive triplicate IM147s followed by a single full-duration
IM240.
PKE Speed Variation Criteria - An additional IM147 implementation issue was the lack
of allowable speed variation criteria for the shortened drive trace. In addition to the false
failures caused by inadequate preconditioning, inadeqjuate control over vehicle operation
during the IM240 test procedure can contribute to inaccurate results. The ability of a
driver to follow the IM240 speed-time trace has a significant effect on the emissions
recorded during the test. To limit this variation in test results, tolerances are applied to
driver performance.
In the 1997 IM240-related evaluation for EPA (SR98-02-01),2 Sierra developed improved
speed variation criteria based on the total Positive Kinetic Energy (PKE) change per mile
traveled during the IM240 cycle. These criteria were designed to minimize the variation
in emissions while still being feasible for use by minimally trained drivers with a
reasonable aptitude for dynamometer driving. However, only IM240 drive cycle criteria
were developed in the 1997 study. Therefore, further analysis was needed to develop
similar speed variation criteria for the IM147.
-17-
-------
Scope
To address the issue of inadequate preconditioning without compromising test throughput
excessively, Arizona decided to change its test procedure to the IM147 beginning
January 1, 2000. This program upgrade will also include implementation of the Max CO
cutpoints previously developed by Sierra, which are designed to maximize the CO
benefits of the program.
To aid in this implementation effort, EPA issued a work assignment (#1-08) to Sierra to
develop the necessary test criteria. Under Work Assignment 1-08, Sierra is to complete
the following tasks:
1. Develop projected IM147 failure rates for the Arizona I/M program using the
3,000-vehicle data set currently being collected by Gordon-Darby. This
evaluation is to include start-up, midpoint, Max CO, and final IM147 cutpoint
scenarios.
2. Develop modal IM147 fast-pass standards for the Max CO cutpoints using (a)
the modal regression technique used in the 1998 EPA study to develop IM147
fast-pass standards, and (b) the 3,000-vehicle data set and the 304-vehicle data
set collected in 1998.
3. Develop modal predictive IM147 retest algorithms for the Max CO cutpoints
using (a) the same technique used in the 1998 EPA study to develop IM147
retest algorithms, and (b) the 3,000-vehicle data set and the 304-vehicle data
set collected in 1998.
4. Develop modal IM147 fast-fail criteria that can be used to terminate retests if
emissions performance is not improving during the retest.
5. Develop fast-pass and full-duration PKE criteria for the IM147 start-up,
midpoint, Max CO, and final cutpoints using the 3,000-vehicle and 304-vehicle
data sets, as well as the 16,581-vehicle data set from the 1997 study of IM240
PKE limits for EPA.
Seven different tasks were proposed to accomplish these objectives.
Task 1. Test Plan Development and Data Collection Assistance - This task covered
working with Gordon-Darby in its efforts to collect the test data needed to complete the
remaining tasks. Data collection, driver participation incentives, and other program-
related details were performed under the guidance of DEQ and Gordon-Darby and were
not Sierra's responsibility. Sierra provided assistance on an as-needed basis to resolve
any problems or questions (e.g., regarding test protocols, data record format, etc.) that
developed during the data collection process in Arizona.
-18-
-------
Task 2. Failure Rates - After completion of the vehicle testing described in Task 1, Sierra
analyzed the resulting data. Using the same approach as utilized in the previous EPA and
DEQ studies, projected IM147 failure rates were developed for both passenger cars and
light-duty trucks. Separate projections were generated for the start-up, midpoint, Max
CO, and final IM147 cutpoints developed under the previous analyses.
Task 3. Fast-Pass Cutpoints - Data obtained in Task 1 as well as the 304-vehicle IM147
data set collected in 1998 were analyzed to develop fast-pass cutpoints and algorithms
associated with the Max CO cutpoints. The same approach previously used to develop
fast-pass cutpoints for the other cutpoint scenarios (i.e., start-up, midpoint, and final) was
followed in this analysis.
Sierra developed fast-pass cutpoints and algorithms for the Max CO cutpoints, with
minor adjustments to the model year groups when the data indicated that such a change
improved accurate emission identification. Separate sets of fast-pass cutpoints were
developed for these new model year groups and vehicle classes as contained in EPA's
IM240 test guidance.
The impact of the resulting fast-pass cutpoints on average dynamometer test time and
excess emissions identified was evaluated using the same techniques as in the previous
analyses. Excess emissions identified will be expressed as the percent of excess
emissions that are identified relative to those identified on the IM240 test.
Task 4. Retest Algorithms - The same data used in Task 3 were analyzed to develop retest
algorithms associated with the Max CO cutpoints. While this task originally charged
Sierra with utilizing the same approach previously used to develop retest algorithms for
the other cutpoint scenarios (i.e., start-up, midpoint, and final), subsequent comments
from Gordon-Darby resulted in alternate retest algorithms. The impact of the resulting
retest algorithms on average dynamometer test time was evaluated using the same
technique as in the previous analyses.
Task 5. Fast-Fail Criteria - The same data used in Task 3 were analyzed to develop
criteria for evaluating mid-test emissions during IM147s in order to determine whether
emissions performance is improving during the retest. The resulting criteria were
structured to "fast fail" vehicles that are not benefitting from such retesting. The impact
of the resulting fast-fail criteria on average dynamometer test time was evaluated using
the same technique as in the previous analyses.
Task 6. Driver Variation Criteria - The same data as used in Task 3 were analyzed to
develop fast-pass and full duration driver variation limits for the IM147 start-up,
midpoint, Max CO and final cutpoints.* The analytical approach used in the 1997
The original workplan called for this latest analysis to use the 16,581-vehicle data set from Sierra's 1997
PKE study for EPA (SR98-02-01); however, as explained later in this report, a "time realignment" (of
emissions versus vehicle speed) was incorporated into the analysis, which made use of the previous data
problematic. It was determined that the effort required to adjust the data could not be justified in terms of a
significant increase in the accuracy of the results; thus, it was decided not to use these previous data.
-19-
-------
analysis was initially followed; however, subsequent results led to the development of an
improved variation metric, Cumulative Positive Power (CPP). Consistent with the
approach used in the previous analysis, the evaluation was structured to develop CPP
limits designed to keep the effective abort rate due to drive trace violations to less than
3%. The impact of the resulting criteria on average dynamometer test time was also
evaluated.
Organization of the Report
Following this introduction, Section 3 describes data collection and the data sets used
throughout this study. Section 4 explains the cutpoint analysis including the revision of
the Max CO cutpoints using the larger data sample; it also details failure rates for the
Startup, Intermediate, Final, and Max CO standards. Section 5 describes the optimized
fast-pass and rete'st criteria. In addition, it details the new fast-fail algorithm and criteria
as well as the net results of optimized criteria when hin simultaneously. Section 6
describes the new CPP driver variation limits and Section 7 integrates the CPP limits with
the optimized IM147 to show the net effect on test time. Section 8 lists the references
cited in the report.
###
-20-
-------
3. TEST DATA
Data used in this study are divided into three distinct groups consisting of 304, 543, and
2,518 vehicles. Vehicles in each of the groups were given three consecutive IM147 tests
regardless of the result.
The 304-vehicle sample was collected by Gordon Darby for Task 1 of Work Assignment
SR99-10-021 at the Gordon Darby I/M lanes in Phoenix, Arizona, during March 1998.
The data included 193 cars and 111 light-duty trucks tested over triplicate IM147 tests
followed by a full IM240 test as illustrated in Figure 3-1.
Figure 3-1
Test Sequence Used to Investigate Triplicate IM147 Tests in Arizona Test Lanes
5-15 minutes First Second Third
in queue IM147 test IM147test IM147 test
Time
Hot
IM240 Test
-21-
-------
For this set, the study inspector would select vehicles by scanning the queue for the
closest white vehicle waiting in the lanes. If there were no white vehicles in the queue,
the inspector would look for the palest, closest vehicle waiting in the lanes. The inspector
approached the first 1981 or newer vehicle following that vehicle, checked to make sure
the vehicle had at least half a tank of gas, and asked the vehicle owner if he or she was
interested in participating in a study that would take approximately 30 minutes, for a
payment of $50. The selection process resulted in vehicles waiting in a queue for
approximately 5 to 15 minutes prior to testing. Most of the vehicles participating in the
program were receiving their initial test; however, 12 vehicles in the database were being
re-tested after an initial failing score.
As discussed in SR99-10-02, four vehicles were pulled out of the original 304 vehicle
data set due to anomalous results. For this study, while Sierra did find some vehicles with
anomalous results in the newer data samples (e.g., passing the initial IM147 yet grossly
failing the final one), these vehicles were not removed from the sample, with the data set
instead being viewed as representative of the in-use fleet. To remain consistent regarding
the treatment of data from the older sample, however, the same four vehicles were
removed for this analysis. The four vehicles are described below.
Record 14, a 1988 Pontiac Bonneville, had relatively low CO emissions during
the first and second IM147 test (1.35 and 3.42 g/mi, respectively). However,
CO emissions during the third IM147 increased substantially (to 55.72 g/mi)
and were higher still during the IM240 following the IM147 testing. It is
interesting to note that CO was emitted in measurable quantities throughout the
test, and the large increases are not attributable to a specific section of the trace.
It thus appears that the gradual emissions increase could be attributable to
excessive purge as the vehicle warmed up or to some kind of catalyst
protection scheme.
Record 15, a 1989 Dodge Dynasty, had moderate CO emissions during the first
three IM147 tests (14 to 18 g/mi), but emissions during the IM240 test were
excessive, particularly during the end portion of that test (106 g/mi).
Reviewing the modal CO emissions in Figure 3-5, one observes that the
vehicle appears to go into open-loop operation at the start of the large hill of
the end portion of the test (i.e., beginning at about second 160 of the IM240).
Although CO emissions accrue throughout this test, the period from 160 to 230
comprises the bulk of the emissions.
Record 23, a 1993 Ford Ranger, shows a very similar emissions response
throughout the three IM147 tests. As seen in Figure 3-6, most of the CO
emissions occur during seconds 62 to 75 of the IM147. During the end portion
of the IM240, a similar pattern is observed. In that test, however, substantial
CO is also emitted during the high-speed portion of the trace. It is not entirely
clear what has caused this, but it appears that the vehicle did not follow the
speed-time trace as smoothly during the end portion of the IM240 as it did
during the first three IM147 tests.
-22-
-------
Record 24, a 1995 Toyota 4Runner, had decreasing emissions throughout the
first three IM147 tests, emitting only 0.72 g/mi CO during the third IM147.
After these vehicles were removed, the remaining sample from the original data set
consists of 191 cars and 109 trucks, for a total of 300 vehicles.
The 2,518- and 543-vehicle data sets were collected for this work assignment by Gordon
Darby at the ten I/M lanes in Phoenix, Arizona during June, July, and August 1999. Like
the original data set, these tests were conducted at the I/M lanes in Phoenix, Arizona.
Unlike the original data set, however, all motorists were asked to participate in the study,
rather than simply those following a white-colored vehicle in the queue. The exception to
this occurred toward the end of the testing when Gordon Darby staff, based on direction
from Sierra, targeted certain vehicle model years and vehicle types to ensure that these
groups were adequately represented in the test data.
The main difference between the 2,518- and 543-vehicle samples was administration of
the IM240 test at the conclusion of the IM147 test. For the 2,518-vehicle sample, only
failing vehicles were given the IM240 test. In the 543-vehicle sample, all vehicles were
given the IM240 test regardless of their IM147 result.
Motorists were not required to participate in this testing. To encourage motorists to
participate, inspection fees were waived for these tests. Inspection fees amount to $25
per inspection. Statistics detailing the number of refusals were not kept.
As previously mentioned, anomalous vehicle test data were not thrown out of the latter
two samples. There was, however, one vehicle identification number (VIN), "123456,"
that appeared multiple times with different vehicles. Gordon-Darby staff confirmed that
this was a test VIN and should be excluded from analysis, which was done. Once this
VIN was removed, the larger sample comprised 2,512 vehicles (1,360 cars and 1,152
trucks) while the smaller sample comprised 535 vehicles (229 cars and 306 trucks).
The model year distribution for each of the samples is shown in Figure 3-2. The YTD
(year-to-date) October 1999 line represents the initial test for 2% random sample vehicles
tested in 1999 through October. Anomalous vehicles have been removed.
For the most part, the model year distribution of the data samples mirrors the random
sample distribution reasonably well. One notable exception can be seen with newer
model year vehicles for the 300-vehicle sample. When that data set was collected, model
year exemptions for the five newest model years were not in place. As a result, this
sample has greater representation throughout these years. The newer data sets were
collected with the model year exemptions in place; therefore, they follow the 1999 YTD
random sample more closely.
One important difference in this study versus previous studies was how time-alignment
was handled. For the previous studies, individual channels (HC, CO, etc.) were aligned
-23-
-------
Figure 3-2
Model Year Distribution of
Sample Fleet and AZ Overall Fleet
at
a>
T
I
00
o>
o>
c
0)
o
d)
a.
20%
15%
10%
5%
0%
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
Model Year
2512 Vehicle Sample Y//A 535 Vehirle Samplp [HI 300 Vehicle Sample YTD October 1999
according to their T90 response times. The T90 response time is the length of time
necessary for the analyzer to see 90% of a positive step change in the gas concentration
that was introduced at the exhaust collection cone. By correcting for response time,
emission events can be linked to the corresponding drive trace event.
The justification for T90 time-alignment centered on the response curve of the gas bench.
Typically, the response curve for a bench will appear somewhat asymptotic; as the
measured gas value closes in on the actual value, the absolute rate at which the measured
value approaches the target value decreases. As a result, the T90 time, which is relatively
short, becomes a good approximation to time an event.
/
Unfortunately, the response time measurement of the analyzer is composed of two
elements, gas bench response time and transport time,, which is the amount of time
necessary for the CVS (Constant Volume Sample) blower to transport the sample from
the collection cone to the gas analyzer bench. This second response time element,
transport time, obscures the gas bench response curve through gas mixing that occurs
during transport.
-24-
-------
Noting this effect, Gordon-Darby staff, after reviewing the raw data, suggested that the
T90 time alignment overcompensated and thus caused short-duration emission events
during a transient cycle to appear to precede the triggering speed event. Given the
interaction of the transport time with the bench response time, Gordon Darby suggested
that a more appropriate alignment measure is T50 response time, which would be more
conservative and alleviate the aforementioned problem.
To accommodate this change, all the existing data, which were previously aligned using
T90 alignment by Gordon Darby, had to be realigned to T50 response times. This
included data collected specifically for this study. Per Gordon-Darby staff, this change
required shifting data for each of the channels (HC, CO, NOx, and C02) four seconds
later relative to the speed signal. This was accomplished by adding four seconds of data
to the front of each test, in which it was assumed that the modal emissions for the entire
period were identical to those measured during the "previous first second" (i.e., now
second 5) of the test. This assumption is considered reasonable since the vehicle is at idle
during this entire period.
I! I! it
-25-
-------
4. IM147 CUTPOINT ANALYSIS
This section of the report discusses the application of the IM147 Max CO cutpoints to the
3,347-vehicle data set. Divided into three parts, the section first discusses changes in
overall IM147 test time. The second part of the section addresses revising the Max CO
cutpoints originally developed in SR99-10-021 based on the large data set now available.
The third and final part of the section details the failure rate when these revised cutpoints
are applied to the" 3,347-vehicle data set.
IM147 Test Length
Sierra originally developed the IM147 test cycle based on the last 147 seconds of data
from the IM240 drive trace. (Figure 4-1 shows the speed/time profile of the IM147 drive
trace.) Following the precedence EPA established with the IM240 drive trace, this meant
that the modal results of the IM147 test would have 147 seconds of data. In short, there
were no constraints regarding an odd versus even number of seconds in the overall test
cycle, nor in Phase 2 of the test cycle.
To simplify implementation of the IM147 test in Arizona, Gordon Darby, and DEQ
agreed that modal data would be recorded once every two seconds instead of second-by-
second as is done with the IM240. This allows the IM147 cycle data to fit into the same
size record format as is currently used for full duration IM240 tests. As a result, having
an odd number of seconds in the drive trace creates a problem of what to do with the odd
second.
To alleviate the problem, Gordon Darby suggested that the speed/time trace define the
boundaries for the test time instead of the actual number of data points reported. In other
words, assuming the first speed/time point is labeled zero seconds, the first modal data
result would be recorded for second 1 and the last for second 146. Phase 2 of the test
would also be revised to start at second 66 (first data reported for second 67) and extend
to the end of the test. The net effect of both of these changes is that both the composite
results and the Phase 2 results will contain an even number of seconds. Since this
addresses the issue of Gordon-Darby's two-second average data collection with no
apparent negative consequences, Sierra revised the test length accordingly. Additional
information on the drive trace is presented in Section 5.
-26-
-------
Figure 4-1
IM147 Trace
Phase 1 and Phase 2
Time (sec)
Cutpoint Analysis
Task 1 of the work assignment required projecting failure rates for the Arizona I/M
program using the combined (3,347 vehicles) data set collected by Gordon Darby. The
projection includes four sets of emission cutpoints: start-up, midpoint, Max CO, and final
for the IM147 test. Since the original Max CO cutpoints were developed using the 304-
vehicle sample compiled for SR99-10-02, Sierra first revisited these cutpoints and model
year groupings to ensure their accuracy against the larger data set.
Accordingly, the model year groupings were modified to avoid anomalously high or low
failure rates for any individual model year/vehicle type combination. A second objective
in establishing the endpoints of the model year groupings was to ensure that changes in
emissions control technology were properly reflected in the various model year
groupings. (There is an obvious and direct relationship between the control technology
installed on a vehicle and its ability to comply with a given set of cutpoints.)
-27-
-------
Once the model year groupings were revised, the IM147 Max CO outpoints were
developed by regressing the final emissions results from the IM240 test against the final
emission results from the third IM147 test for each of the three exhaust constituents. For
this regression, Sierra used only vehicles from the two of the three data sets where
vehicles were automatically given an IM240 test regardless of their IM147 test results
(the 300- and 535-vehicle samples). Because the third data set contained IM240 data
only for vehicles failing the IM147, this data set would have created a regression bias and
was therefore excluded from this part of the analysis. The resulting regression equations
were then used to extrapolate IM147 composite cutpoints from the IM240 composite
cutpoints developed previously in SR99-10-02. Three linear regression equations were
developed for each of three model year groupings. Equation 4-1 illustrates the
regression equation.
IM147-Cutpoint = slope * (IM240 Cutpoint) + Intercept [4-1]
Table 4-1 details the appropriate model year/emission constituent regression coefficients
to be used in the above equation. Coefficients of correlation (r2 values) are also shown
for each of the regression equations. As expected, the r2 values shown in the table
demonstrate good correlation between the IM240 and IM147 cutpoints. (More detailed
regression results, including graphical plots, are shown in Appendix F.) The model year
groupings identified in this table are not the same as the model year groupings used in the
emission cutpoint tables, which vary depending upon vehicle type.
IM240 to IM147 Com
Table 4-1
josite Regression Equation Coefficients
Model Year
Group
Emission
Constituent
Slope
Intercept
Correlation
Coefficient
(r2 value)
1981-1985
HC
0.896629
0.110694
0.963
CO
1.020463
0.858255
0.979
NOx
1.065128
0.085613
0.978
1986-1989
HC
0.933646
0.056509
0.976
CO
0.939067
1.679632
0.969
NOx
1.077932
0.058971
0.956
1990+
HC
0.963839
0.026672
0.949
CO
1.037836
0.392486
0.840
NOx
1.102698
0.048771
0.918
-28-
-------
Phase 2 cutpoints were developed by regressing the IM147 composite test scores against
the IM147 Phase 2 scores. Unlike the composite score regressions, however, the
equations had to be adjusted to preserve the relationship between composite versus
Phase 2 scores existing in the IM240 test. EM240 Phase 2 cutpoints are more rigorous
than the composite cutpoints, presumably to minimize falsely passing vehicles. After
studying the IM147 data, a multiplier of 0.9 was used with the regression since it
provided additional defense against false failures while maintaining the possibility of a
Phase 2 pass. The following Phase 2 regression equation (4-2) was used to extrapolate
IM147 Phase 2 cutpoints from the IM147 composite cutpoints.
IM147Phase 2 = 0.9 * (Slope * (IM147 Composite) + Intercept) [4-2]
Table 4-2 details-the appropriate regression coefficients to be used with the above
equation, as well as the resulting r2 values. The r2 values contained in the table
demonstrate excellent correlation between the composite and Phase 2 cutpoints. Unlike
the composite regression coefficients shown in Table 4-1, IM147 composite to Phase 2
coefficients were held constant across model years.
Table 4-2
IM147 Composite to Phase 2 Regression Equation Coefficients
Emission Constituent
Slope
Intercept
Correlation Coefficient
(r2 value)
HC
0.807408
0.012886
0.973
CO
0.881965
-0.569281
0.975
NOx
0.989412
-0.083696
0.981
Table 4-3 shows the revised IM147 Max CO cutpoints developed using both sets of
regression equations. In two places, the revised NOx cutpoints seem anomalous because
they are actually less stringent for newer vehicles (1989 to 1990 LDGV and 1987 to 1988
LDGT2). As it turns out, these anomalies occurred at or near the regression equation
breakpoints.* The slight discontinuity caused by this change created the apparent
anomaly.
To apply the regression equations to the model year groupings, which vary depending upon the vehicle
type, cutpoints for some model years were determined using the adjacent regression equation.
-29-
-------
Table 4-3
Revised IM147 Max CO Cutpoints
(Composite/Phase 2)
Vehicle Class
Model Years
HC
CO
NOx
LDGV
1981-82
2.80/2.05
26.37/20.42
3.28/2.85
1983-85
2.08/1.53
17.19/13.13
3.28/2.85
1986-89
1.46/1.07
15.77/12.00
2.75/2.38
1990-95
0.99/0.73
12.85/9.68
2.81/2.42
1996+
0.80/0.59
12.85/9.68
2.25/1.93
LDGT1 -
1981-85
3.70/2.70
31.47/24.47
5.41/4.74
1986-89
2.86/2.09
25.16/19.46
4.91/4.30
1990-95
1.95/1.43
21.15/16.28
4.46/3.90
1996+
1.57/1.15
21.15/16.28
3.36/2.91
LDGT2
1981-85
4.06/2.96
51.88/40.67
6.48/5.69
1986-87
3.79/2.77
39.24/30.64
5.99/5.26
1988-95
2.92/2.13
26.34/20.39
6.11/5.37
1996+
2.34/1.71
26.34/20.39
4.46/3.90
Comparison of Failure Rates
After revising the Max CO outpoints using the 835-vehicle sample, the failure rates were
evaluated using each of the four sets of IM147 outpoints (Startup, Intermediate, Final, and
Max CO) and the combined vehicle data set (3,347 vehicles). While the IM147 Max CO
cutpoints were revised for this study, the IM147 Startup, Intermediate, and Final
cutpoints were developed as part of SR99-10-02 and are shown in Appendix A.
Table 4-4 shows how the failure rate changed with the different cutpoints for the third
and final IM147. The overall failure rate will be slightly less when actually implemented
since some vehicles will pass and therefore complete the test after an earlier IM147 even
though they would go on to fail the third one if the test was continued. More detailed
information on the failure rates is provided in Appendix B.
This table helps to clarify the relationship between the four sets of cutpoints. The
Startup, Intermediate, and Final cutpoints result in nearly equal HC, CO, and NOx
pollutant-specific failure rates for each cutpoint category (e.g., the pollutant-specific
failure rates for the startup cutpoints range from 5.7% to 6.5%). Both the pollutant-
specific and overall failure rates increase with increasing cutpoint stringency. The Max
CO standards, when compared to the final standards, reduce the HC and NOx failure
rates, and instead increase the CO failures, especially for light-duty trucks.
-30-
-------
Table 4-4
Failure Rates, Third IM147
Vehicle
Type
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Max CO Cutpoints
LDGV
114
1666
6.4%
250
1530
14.0%
155
1625
8.7%
390
1390
21.9%
LDGT1
49
915
5.1%
156
808
16.2%
70
894
7.3%
219
745
22.7%
LDGT2
25
578
4.1%
63
540
10.4%
30
573
5.0%
93
510
15.4%
All
188
3159
5.6%
469
2878
14.0%
255
3092
7.6%
702
2645
21.0%
Startup Cutpoints
LDGV
106
1674
6.0%
130
1650
7.3%
141
1639
7.9%
281
1499
15.8%
LDGT1
52
912
5.4%
33
931
3.4%
. 48
916
5.0%
108
856
11.2%
LDGT2
33 ~
570
5.5%
30
573
5.0%
28
575
4.6%
67
536
11.1%
All
191
3156
5.7%
193
3154
5.8%
217
3130
6.5%
456
2891
13.6%
Intermediate Cutpoints
LDGV
157
1623
8.8%
161
1619
9.0%
196
1584
11.0%
367
1413
20.6%
LDGT1
80
884
8.3%
53
911
5.5%
73
891
7.6%
165
799
17.1%
LDGT2
40
563
6.6%
42
561
7.0%
42
561
7.0%
93
510
15.4%
All
277
3070
8.3%
256
3091
7.6%
311
3036
9.3%
625
2722
18.7%
Final Cutpoints
LDGV
280
1500
15.7%
240
1540
13.5%
277
1503
15.6%
507
1273
28.5%
LDGT1
120
844
12.4%
85
879
8.8%
127
837
13.2%
239
725
24.8%
LDGT2
73
530
12.1%
56
547
9.3%
82
521
13.6%
149
454
24.7%
All
473
2874
14.1%
381
2966
11.4%
486
2861
14.5%
895
2452
26.7%
Figures 4-2 through 4-5 show how the failure rate changes as vehicles progress through
the three IM147 tests. Since most of the increase in failures occurs between the first and
second IM147 tests, it is reasonable to expect that algorithms designed to shorten the test
would end the test prior to the third IM147. As will be shown later in the report, this is
indeed the case. Once the fast-pass, retest, and fast-fail criteria are applied, only 162
vehicles out of 3.347.(4.8%) are tested beyond the second EM147.
-------
Figure 4-2
Failure Rate by Consecutive IM147
Max CO Outpoints
HC Failure Rate by Vehicle Type
u
i
~ 1st 1M 147
Ś 2nd IM 147
03rd IM 147
PC
LDT1 LDT2
Vehicle T ype
All
NOxFailure Rate by Vehicle Type
O
z
ED 1st IM 147
Ś 2nd IM 147
~ 3rd IM 147
PC
LDTI LDT2
Vehicle Type
All
CO Failure Rate by Vehicle Type
0 1st IM 147
Ś 2nd IM 147
~ 3rd IM 147
PC
LDT 1
LDT2
All
Vehicle T ype
Overall Failure Rate by Vehicle Type
0 1st IM 147
Ś 2nd IM 147
03rd IM 147
PC
LDT1 LDT2
Vehicle Type
All
-------
I
u>
u>
Figure 4-3
Failure Rate by Consecutive IM147
Startup Outpoints
HC Failure Rate by Vehicle Type
u
X
[0 I st IM 147
Ś 2nd IM 147
~ 3rd IM 147
LDT1 LDT 2
Vehicle Type
NOx Failure Rate by Vehicle Type
o
z
0 1st IM 147
Ś 2nd IM 147
03rd IM 147
PC
LDT1 LDT2
Vehicle Type
All
CO Failure Rate by Vehicle Type
o
u
O 1st IM 147
Ś 2nd IM 147
03rd IM 147
PC
LDT I LDT2
Vehicle Type
All
Overall Failure Rate by Vehicle Type
25.0%
Ł 20.0%
a
S 15.0%
~ 1st IM 147
Ś 2nd IM 147
Q 3rd IM 147
PC
LDT1 LDT2
Vehicle Type
All
-------
Figure 4-4
Failure Rate by Consecutive IM147
Intermediate Cutpoints
HC Failure Rate by Vehicle Type
~ '1st IM 147
Ś 2ndIM 147
~ 3rd IM 147
PC LDTI LDT2 All
Vehicle Type
NOxFailure Rate by Vehicle Type
~ 1st IM 147
Ś 2nd IM 147
~ 3rd IM 147
PC
LDT I
LDT2
All
Vehicle Type
CO Failure Rate by Vehicle Type
o
CJ
ED 1st IM 147
|2nd IM147
EJ3rd IM 147
PC
LDT1 LDT2
Vehicle Type
All
Overall Failure Rate by Vehicle Type
~ 1st IM 147
Ś 2nd IM 147
~ 3rd IM 147
PC
LDTI LDT2
Vehicle Type
All
-------
Figure 4-5
Failure Rate by Consecutive IM147
Final Cutpoints
HC Failure Rate by Vehicle Type
El 1st 1M 147
Ś 2nd IM I 47
~ 3rd IM 147
PC
LDTI LDT2
Vehicle T ype
All
NOxFailure Rate by Vehicle Type
(011st IM 147
Ś 2nd IM 147
123rd IM 147
0.0%
PC
LDTI LDT2
Vehicle T ype
All
CO Failure Rate by Vehicle Type
o
o
EH 1st IM 147
Ś 2nd IM 147
~ 3rd IM 147
PC
LDTI LDT2
Vehicle Type
All
Overall Failure Rate by Vehicle Type
0 1st IM 147
Ś 2nd IM 147
03rd IM 147
LDTI LDT2
Vehicle Type
-------
5. OPTIMIZED IM147 TEST CRITERIA
This section presents the analysis methodology and results used to develop optimized
IM147 test criteria. Criteria developed in the study include fast-pass standards, retest
algorithms, and fast-fail criteria. As described below, each set of criteria was evaluated in
turn to determine its overall effect on test time and excess emissions identification.
Modal Fast-Pass Standards
IM147 fast-pass standards were originally developed for startup, midpoint, and final
standards as part of SR99-10-02.1 The current study furthered that work by developing
fast-pass standards for the Max CO cutpoints. The fast-pass regression coefficients
determined for both SR99-10-02 and the current study were developed using the
methodology described in Sierra Report No. SR98-02-01,2 "Additional Study of
Preconditioning Effects and Other IM240 Testing Issues." As detailed in that report, the
selected drive trace is divided into segments over which mass emissions for HC, CO, and
NOx are summed. By performing a multivariate linear regression of the modal mass
emissions against the composite emissions result, we can determine the coefficients
needed to predict final emissions at mode ends throughout the test. Equation 5-1
illustrates how these coefficients are used to predict emissions.
P240* = Cn + Ł {Snm X Xnm} + {Me X En) t5*1]
m=1
Where: P240n = Predicted emissions after completing n segments
Cn = Regression intercept for equation n
Snm = Regression coefficient for segment m in equation n
Xnm = Total emissions over a given segment m in equation n
Me = Error multiplier (usually 2 unless otherwise specified)
En = Error in regression equation n
n = Equation number (corresponds to the number of modal segments
completed)
m = Segment number
After reviewing the proposed fast-pass and retest procedures described in SR99-10-02,
Gordon-Darby staff expressed several concerns regarding actual in-use implementation of
that procedure. In response, the workplan developed for the current project indicated that
-36-
-------
Sierra would initiate further discussions with Gordon-Darby and evaluate possible
changes to the previously developed procedures to address these concerns. Specific
concerns that were voiced by Gordon-Darby staff include the following:
1. Insufficient number of fast-pass segments - SR99-10-02 divides the
IM147 drive trace into 14 segments. While the segments in the first half
of the test usually comprise 8 to 10 seconds, several segments in Phase 2
of the drive trace are considerably longer (up to 19 seconds in duration).
Gordon-Darby expressed concern that extended segments may force
vehicles to be tested longer than necessary prior to a fast-pass.
2. Fast-pass segments containing an odd number of seconds - To allow the
resulting modal test data up to three possible IM147 cycles to fit into the same
size record format as currently used for full duration IM240 tests, Gordon-
Darby -plans to record two-second averages, rather than one-second recordings
as is presently done. To simplify its lane software, Gordon-Darby requested
that Sierra realign the segments to agree with the planned frequency of data
storage.
3. Nonalignment of fast-pass segments with proposed retest modes - To further
simplify its programming process, Gordon-Darby asked that Sierra try to
coincide retest mode breaks with segment breaks.
In addition to simply developing updated fast-pass regression coefficients and retest
algorithms based on an expanded data set, this study addresses the above concerns. In
response to the first two concerns as well as the test length issue (odd vs. even number or
seconds), the fast-pass segments have been revised as shown in Table 5-1.
As the table shows, there are now 20 segments ranging from 4 to 10 seconds in duration.
Each segment contains an even number of seconds. Phase 2 of the IM147 now begins at
segment 11 and extends through the end of the test. The test includes 146 seconds of data
(collected starting at second 0 and ending at second 146).
When the IM240 segments were originally created as part of SR98-02-01, they were
divided in such a way that segments characterized different modes of vehicle operation.
Some segments were composed of hard accelerations, while others characterize cruises,
and everything between, including decelerations. While this study re-aligned the
segments with deference to ensuring 4- to 10-second segment lengths divided along even
increments, care was taken to preserve the original intent of the segmentation. Figure 5-1
shows the re-aligned segments positioned against the drive trace.
-37-
-------
Table 5-1
Revised IM147 Segments
Segment
Initial Second Data
Final Second Data
1
1
4
2
5
16
3
17
22
4
23
28
5
29
34
6
35
42
7
43
48
8
49
54
9
55
60
10
61
66
11
67
76
12
77
82
13
83
92
14
93
98
15
99
108
16
109
112
17
113
116
18
117
122
19
123
132
20
133
146
-38-
-------
Figure 5-1
IM147 Test Segments Used for Fast-Pass Cutpoint Development
Time (sec)
Using these revised test modes, Sierra applied the aforementioned regression method to
the 3,347-vehicle sample supplied by Gordon-Darby. Thirty-nine distinct regression
models, shown in Appendix C, were developed from these data. Models were created for
each vehicle type (LDGV, LDGT1, and LDGT2), emission constituent (HC, CO, and
NOx), and phase (composite and Phase 2 only). There are 20 equations for the composite
models and 10 equations for the Phase 2 models.
After revising the regression models, they were applied to the 835-vehicle sample
(unbiased IM240 sample) to determine how the models would affect excess emissions
identified by the IM147 test. Excess emissions are defined as emissions collected during
an IM240 test in excess of the applicable standard for a given vehicle. The IM147 test
receives credit for identifying excess emissions if it fails a vehicle that had excess IM240
emissions. Table 5-2 shows the excess emissions versus test time results when the
Fast-Pass algorithm was enabled. Max CO standards were used for both the IM147 and
the IM240.
As Table 5-2 shows, the IM147 test with the fast-pass criteria enabled still identifies over
91% of the excess emissions for each of the three exhaust constituents and still
-39-
-------
Table 5-2
Modeled Fast-Pass Results
Excess Emissions vs. Average Test Time
No Fast-Pass
Fast-Pass Enabled
Excess HC Identified
97.8%
95.7%
Excess CO Identified
96.3%
92.7%
Excess NOx Identified
95.4%
82.4%
Average Test Time
217 seconds
125 seconds
significantly reduces average test time.* Without fast-pass enabled, the pass/fail
evaluation occurs only after each IM147 is complete. If the vehicle passes the IM147, the
test would be complete at that point. If the vehicle fails, another IM147 would be run,
with up to three IM147s conducted on any one vehicle. If the vehicle is still failing at the
end of the third IM147, it fails the overall test. Since each IM147 lasts 146 seconds, the
maximum time a vehicle could be tested is 438 seconds.
Without fast-pass, the average test time for the 3,347-vehicle sample was 217 seconds.
Once the fast-pass algorithm was applied, the average test time dropped to 125 seconds,
for a reduction of 92 seconds (42%). This is similar to the test time estimate of 121
seconds shown in SR99-10-021 for the final standards with fast-pass.
Table 5-3 details excess emissions identification by vehicle type and model year groups.
Appendix D provides additional excess emission analysis results. As will be shown later
in the report, excess emissions identified increase with the addition of retest and fast-fail
algorithms.
Predictive Retest Algorithms
As with the fast-pass standards, the predictive retest algorithms were developed as part of
SR99-10-02. Retest algorithms are intended to predict whether a vehicle would benefit
from additional testing. If the algorithm determines that a vehicle that was failing at the
end of the first or second of three IM147 tests would benefit from additional testing, then
the next IM147 would begin. If, on the other hand, the algorithm determines that the
vehicle would not benefit from additional testing, the test would be terminated at that
point and the vehicle would fail the inspection.
*
Average test time refers to the estimated time for that portion of the test in which the vehicle is driven on
the dynamometer.
-40-
-------
Table 5-3
Excess IM240 (Max CO Cutpoints) Emissions Identified3
(With and Without Fast-Pass)
Vehicle
Class
Model
Year
No Fast-Pass
Fast-Pass Enabled
Excess HC
Identified
Excess CO
Identified
Excess NOx
Identified
Excess HC
Identified
Excess CO
Identified
Excess NOx
Identified
LDGV
81-82
N/A
87.6
100.0
N/A
87.6%
100.0%
83-85
100.0%
97.6
100.0
100.0%
97.6%
100.0%
86-89
100.0%
99.8
100.0
99.2%
97.1%
100.0%
90-95
99.0%
8.6.3
91.2
96.7%
85.3%
46.0%
96+
N/A
N/A
N/A
N/A
N/A
N/A
ALL
99.6%
94.8
96.8
98.2%
93.3%
80.6%
LDGT1
81-85
78.5
99.6
99.4
78.5%
92.0%
99.4%
88-89
100.0
99.0
71.3
93.8%
90.8%
71.3%
90-95
100.0
0.0
100.0
0.0%
0.0%
100.0%
96+
N/A
N/A
100.0
N/A
N/A
0.0%
ALL
90.0
99.0
94.6
83.4%
91.0%
92.6%
LDGT2
81-85
92.3
100.0
79.7
92.3%
100.0%
79.7%
86-87
100.0
100.0
53.1
100.0%
100.0%
53.1%
88-95
100.0
100.0
65.7
100.0%
36.2%
56.6%
96+
N/A
N/A
N/A
N/A
N/A
N/A
ALL
96.1
100.0
65.8
96.1%
95.2%
64.5%
Total
ALL
97.8
96.3
95,4
95.7%
92.7%
82.4%
° N/A means that no vehicles failed the applicable IM240 cutpoint within this vehicle class/model year
grouping.
In the current study, the workplan called for Sierra to refine the algorithms using the
3,347-vehicle sample and the Max CO cutpoints. The original algorithms predicted test
outcomes using a combination of mass and concentration readings during specific modes
of the IM147 test. In conversation with Gordon-Darby staff, however, Sierra learned that
the suggested retest logic would be difficult to implement. While gas concentrations
could be determined during the test, it would be far less work if all of the retest criteria
referred only to mass. Another concern, as previously mentioned, was that the modes did
not align with the fast-pass segments. With these issues in mind, Gordon-Darby asked if
Sierra could find a more user-friendly retest logic.
While Sierra agreed to explore alternatives to concentration measurement in the retest
procedure, it was not clear that any viable alternatives existed since the justification for
using both concentration and mass made sense technically. While mass emission
-41-
-------
measurements are useful to gauge emission performance relative to the actual cutpoint,
the concentration measurements provide a measure of emission performance less affected
by engine load than mass emissions. As a result, emissions concentration measurements
seemed especially relevant for measuring improvements in engine performance during a
transient test.
After some experimentation, Sierra settled on a procedure that utilizes the previously
determined fast-pass regression coefficients to predict whether a vehicle would benefit
from additional testing. In short, if the vehicle has not fast-passed the inspection by the
end of the 19th segment and the predicted emissions fall outside a tolerance level
allowing automatic retest, the emissions result predicted after segment 7 is compared to
the emissions result predicted after segment 19 to determine whether the vehicle
emissions are converging on the applicable cutpoint. Equation (5-2), named the
convergence ratio for this study, illustrates how emission convergence on the applicable
standard is determined.
Segment! - Comp Std
Convergence._Ra,io - SegmentIg_ Comp_Std [5-2]
Where: Segment7 = Predicted emissions after segment 7 (Composite Regression)
Segmentl9 = Predicted emissions after segment 19 (Composite Regression)
Comp_Std = Composite Cutpoint for the specific emission
Common sense would dictate that the convergence ratio would be greater than one if the
emissions are converging on the cutpoint for vehicles where the emissions are above the
cutpoint. A convergence ratio less than 1, on the other hand, might suggest that the
emissions are actually diverging from the cutpoint. With this in mind, it makes sense that
a conservative decision threshold for the retest algorithm would utilize a convergence
ratio greater than one for the HC and CO channels. In general, emissions of both of these
pollutants reduce as a vehicle warms up. NOx, on the other hand, may actually increase
the longer a vehicle operates, so the appropriate convergence ratio threshold may be
greater than one.
Iterative solutions to this equation using actual vehicle data suggest, however, that both of
these assumptions may result in overly stringent application of the retest algorithm, thus
creating unacceptable false failure rates. In short, some vehicles failing for either HC or
CO during one of the early IM147 tests may go on to pass a later IM147 in spite of the
fact that their emissions appeared to be diverging from the cutpoint during the earlier
IM147. Regarding NOx emissions, modal emissions are even more difficult to predict.
As a result, the convergence ratio was not applied to the NOx channel. Instead, the retest
algorithm predicts NOx failures by simply comparing segment 7 and 19 NOx readings to
the previously mentioned tolerance level. If the readings are greater than the prescribed
multiple of the cutpoint, then the vehicle does not receive a retest.
-42-
-------
Separate algorithms were developed for LDGVs and LDGTs based upon whether one or
two IM147 tests had been completed. In all cases, a predicted emissions score of less
than the standard (but that does not trigger a fast-pass) is treated as cause for a retest.
This approach avoids the need to deal with negative convergence ratios, while also
providing maximum potential for vehicles to pass the test. The flow charts shown in
Figures 5-2 through 5-5 show the specifics regarding how these algorithms work.
Unlike the fast-pass algorithms, the error term in the fast-pass regression equations is not
included when determining predicted emissions for the Segment 7 to Segment 19
comparison. If the error term (which varies between regression models) were included,
the retest procedure would need to be refined for specific regression models in addition to
vehicle types. Without including the error term, LDGVs can simply be separate from
LDGTs while still maintaining acceptable accuracy.
Table 5-4 presents the results of the retest algorithm independent of the fast-pass
algorithm. As shown in the table, the retest criteria eliminate more vehicles after the
second test than after the first test. This is in contrast to the fast-pass criteria, which pass
a disproportionate number of vehicles after the first IM147 when compared to the second
IM147. Given the difference between the two criteria, this makes sense. While vehicles
can pass the test after early IM147 tests without the fast-pass enabled, the only way a
failing vehicle can end the test without the retest algorithm enabled is to run the full
duration of the test. As a result, the retest criteria on the first IM147 are judged by their
ability to predict emissions on the third IM147 whereas the fast-pass criteria are judged
by their ability to predict emissions on the current IM147, an easier task. The retest
criteria must therefore be more conservative on the first IM147 than the fast-pass criteria.
In addition to listing the point at which vehicles were denied a retest, Table 5-4 shows the
number of false fails as a result of the retest procedure. The criterion for false failing is
quite simple: a false failure occurs any time a vehicle that would go on to pass one of the
subsequent IM147 tests is denied a retest. As shown, there were no false failures.
Despite the conservatism evidenced by this result, average test time dropped from 125
seconds to 96 seconds when the retest algorithm was added to the fast-pass criteria.
Modal Fast-Fail Criteria
One of the requirements of this work assignment was to develop modal fast-fail criteria.
Unlike the retest procedure, which can terminate the test at the ends of the individual
IM147 tests, the fast-fail algorithm can terminate tests during an IM147 test. Like the
retest algorithms, the fast-fail algorithm has the potential to falsely fail vehicles that
would otherwise pass the inspection in the algorithm's absence. After looking at the test
data, it was apparent that the data from the first IM147 were too unpredictable to fast-fail
any significant number of vehicles during the first IM147 without also significantly
increasing false failure levels. While the retest algorithm utilizes almost all of the first
IM147 data before it makes a decision not to retest a vehicle, the fast-fail algorithm must
make essentially the same decision in a smaller amount of time. For this reason, it was
decided that the fast-fail algorithm would not function before the second IM147.
-43-
-------
Figure 5-2
Retest Algorithm During First IM147 - LDGV
Max CO Outpoints
Note: STD = Composite Standard for the specific pollutant.
Mode 7 Predicted = Predicted emissions after 7 segments using the composite regression
equations without error term included.
Mode 19 Predicted = Predicted emissions after 19 segments using the composite
regression equations without error term included.
-44-
-------
Figure 5-3
Retest Algorithm During Second IM147 - LDGV
Max CO Outpoints
Note: STD = Composite Standard for the specific pollutant.
Mode 7 Predicted = Predicted emissions after 7 segments using the composite regression
equations without error term included.
Mode 19 Predicted = Predicted emissions after 19 segments using the composite
regression equations without error term included.
-45-
-------
Figure 5-4
Retest Algorithm During First IM147 - LDGT1, LDGT2
Max CO Cutpoints
Note: STD = Composite Standard for the specific pollutant.
Mode 7 Predicted = Predicted emissions after 7 segments using the composite regression
equations without error term included.
Mode 19 Predicted = Predicted emissions after 19 segments using the composite
regression equations without error term included.
-46-
-------
Figure 5-5
Retest Algorithm During Second IM147 - LDGT1, LDGT2
Max CO Cutpoints
Note: STD = Composite Standard for the specific pollutant.
Mode 7 Predicted = Predicted emissions after 7 segments using the composite regression
equations without error term included.
Mode 19 Predicted = Predicted emissions after 19 segments using the composite
regression equations without error term included.
-47-
-------
Table 5-4
Retest Algorithm Results
LDGV
LDGT1, LDGT2
Total Number of
Complete Tests
1567
1780
# of Failures Without
Retest Algorithm
327 (20.9% of 1567)
273 (15.3% of 1780)
# of Correctly Identified
Failures3
245 (74.9% of 327)b
.180 (65.9% of 273)b
# Failing After 1IM147
94 (38.4% of 245)c
97 (53.9% of 180)°
# Failing After 2 IM147s
151 (61.6% of 245)c
83 (46.1% of 180)c
# of Passing-Vehicles
Falsely Failed by Retest
0(0% of 1567)
0(0% of 1780)
a "Correctly identified failures" refers to those vehicles that were still failing at the end of the third IM147.
b The number shown in parentheses is the number of failures without the retest algorithm.
c The number shown in parentheses is the total number of IM147 Cycle 2 and 3 failures.
In order to build on work already completed for the retest procedure, the fast-fail
algorithm for the second IM147 trace evaluates predicted emissions after segment 7 of
the drive trace.* To maintain uniformity for lane software programmers, the error term
was not included in the prediction of the mode 7 emissions since it was not in the retest
algorithm.
The vehicle fast-fails the second IM147 if any of the following are true:
For LDVs:
Predicted HC after 7 segments > (1.5 x Composite HC standard)
Predicted CO after 7 segments > (2.2 x Composite CO standard)
Predicted NOx after 7 segments > (1.4 x Composite NOx standard)
~
It is theoretically possible to identify additional fast-fails by using the same or different predictive
algorithms at the end of subsequent segments. While Gordon-Darby has expressed interest in this
enhancement in order to further reduce average test time, the timing of the study did not allow this issue to
be evaluated.
-48-
-------
ForLDTs:
Predicted HC after 7 segments > (1.1 x Composite HC standard)
Predicted CO after 7 segments > (1.5 x Composite CO standard)
Predicted NOx after 7 segments > (1.5 x Composite NOx standard)
Using these criteria, 99 trucks and 124 cars are fast-failed after segment 7 of the second
IM147. Different fast-fail criteria were used during the third IM147, since the potential
for falsely failing vehicles in subsequent IM147 tests is eliminated. As a result, the
modal fast-pass algorithm, with a minor modification, can be implemented to predict
failing vehicles. Instead of the error term being added to the predicted score, as is done
with the fast-pass algorithm, the error term is subtracted. This adjustment ensures
conservative emission estimates, which will help to minimize false failures when using
the algorithm. Equation 5-3 illustrates the third IM147 fast-fail algorithm.
n
P240n = Cn + J] {iSnrn X Xnm} ( Me X En) o-i
m= 1 P"-5!
Where: P240n = Predicted emissions after completing n segments
Cn = Regression intercept for equation n
Snm = Regression coefficient for segment m in equation n
Xnm = Total emissions over a given segment m in equation n
Me = Error multiplier (usually 2 unless otherwise specified)
En = Error in regression equation n
n = Equation number (corresponds to the number of modal segments
completed)
m = Segment number
Table 5-5 details the results of the fast-fail algorithm for both the second and third IM147
tests. As the number of false failures indicates, the fast-fail criteria were developed with
the intention of minimizing false failures. The results shown in the table are also
deceptive, since they are based on an analysis of the impact of the fast-fail algorithm in
the absence of the other test criteria. While over 600 fast-failures are shown in the table,
many of these are also subject to the retest criteria, resulting in a much smaller fast-fail
impact when the criteria are combined. In this latter case, the number of fast-fails falls to
roughly 200 vehicles. This effect, combined with the fact that such fails occur relatively
late in the 3-IM147 test cycle, leads to a fairly small impact on average test time. Once
the fast-fail algorithm was enabled with the fast-pass and retest algorithm, average test
time was reduced from 96 seconds to 94 seconds.
-49-
-------
Table 5-5
Fast-Fail Algorithm
Second IM147
Third IM147
False
Vehicle Class
Fast-Failures
Fast-Failures
Failures
LDV
124
211
3
LDT
99
182
3
Total
223
393
6
Integration of Fast-Pass. Retest. and Fast-Fail Algorithms
The next step in the analysis was to integrate the Fast-Pass, Retest, and Fast-Fail criteria
to determine their net effect. Table 5-6 shows how average test times and excess
emissions identified vary by model year range and vehicle class using the integrated
criteria.
As you can see, once the retest and fast-fail algorithms were added to the fast-pass
algorithm, excess emissions identification improved. There are two reasons for this:
(1) vehicles that would be falsely passed later in the test are now failed prior to that
decision being made; and (2) some vehicles that passed according to the IM147 criteria
yet would have failed the IM140 are now falsely failed on the IM147, thus increasing
IM240 excess emissions identification.
The excess emissions identification rate shown in Table 5-6 is down from the
identification of SR99-10-02,1 which identified 99.6% of the HC emissions, 98.2% of the
CO, and 99.9% of the NOx with the fast-pass and retest algorithms enabled. However,
direct comparison of these results may not be relevant for several reasons. First, the
previous study measured excess emissions captured against the Final Outpoints rather
than the Max CO cutpoints developed for this study. Second, because fast-fail was
created for this study, it was not included in the previous study results. Third, since the
retest algorithm has been modified as part of this study, it will have a different effect on
the results. In the previous study, the retest algorithm improved excess emission
identification by 2.7% versus 1.1% for this study. Lastly, most of the vehicle data used in
this study were collected with the model year exemptions in place and then normalized to
the inspection fleet, based on 2% random sample data collected from July 1997 to March
1998. This skews the model year distribution older when compared to SR99-10-02,
which used data unbiased by model year exemptions.
Figure 5-6 illustrates how the integrated criteria combined to produce emission results
and final test times. Note that only 7 of the 3,347 vehicles were tested over the full
duration of all three IM147 tests.
-50-
-------
Table 5-6
Comparison of Integrated Algorithms (No CPP) vs. Standard IM147
Impact on Test Time and Excess IM240 Emissions (Max CO Cutpoints) Lost
Class
Model
Year
Group
Sample
Size"
Mean Test
Time
Standard3
Mean Test
Time w/
Algorithms"
% Excess Emissions Identified1"
HC
CO
NOx
LDGV
1981-82
105
286.4
140.1
-
87.6%
100.0%
1983-85
228
311.2
169.4
100.0%
97.6%
100.0%
1986-89
425
248.4
112.0
99.2%
97.1%
100.0%
1990-95
952
184.8
78.3
Ś 96.8%
85.3%
91.2%
1996+
70
154.3
37.1
-
-
-
AH"
1780
221.0
100.0
98.2%
93.3%
96.8%
LDGT1
1981-85
260
306.6
158.7
78.5%
97.3%
99.4%
1986-89
222
230.8
101.6
93.8%
90.8%
99.6%
1990-95
450
173.3
59.2
0.0%
0.0%
100.0%
1996+
32
155.1
31.6
-
-
100.0%
All
964
221.9
94.9
83.4%
92.6%
99.5%
LDGT2
1981-85
94
307.5
158.8
100.0%
100.0%
77.2%
1986-87
64
253.2
101.4
100.0%
100.0%
100.0%
1988-95
427
166.5
54.2
-
0%
44.4%
1996+
18
146.0
28.0
-
-
-
All
603
197.1
74.7
100.0%
93.8%
83.1%
Weighted Average
3,347
217.0
94.0
95.9%
93.1%
97.0%
a Mean test time standard refers to the average dynamometer test time without the algorithms enabled.
This was determined using the 3,347-vehicle sample.
b Percent of IM240 (Max CO) excess emissions identified with the integrated algorithms enabled. This
was determined using the 835-vehicle sample and normalized to the Arizona 2% random sample fleet
distribution data, July 1997 to March 1998.
-51-
-------
Figure 5-6
Integrated Test Results
Maximum CO Outpoints Without CPP Limits
TOTAL PASS = 2,802 TOTAL FAIL = 545
TOTAL FALSE FAILURES = 3
-52-
-------
Segment 2 Revised Integrated Algorithms Results
The original integrated algorithm results were determined assuming fast-pass and fast fail
results could not be rendered prior to the fourth segment (i.e., no earlier than Test Time =
28 seconds). This was consistent with the procedure established in SR98-02-01.
Gordon-Darby, wishing to further minimize test time, requested that Sierra explore the
feasibility of rendering fast-pass and third IM147 fast-fail decisions after earlier segments
without degrading excess emission identification. Unfortunately, simply moving the
decision forward with the existing algorithms, while shortening test time, did degrade
excess emission identification. To adjust for this change, the error multiplier used in the
fast-pass decision needed to be increased to 3 during segment 2 and to 2.5 during segment
3. While the fast-fail algorithm used during the third EM 147 also uses an error multiplier,
it can be left at 2 during segments 2 and 3 without increasing false-failure incidence.
Using these criteria, the earliest possible fast-passes were moved to the end of segment 2
(i.e., Test Time = 16 seconds) and average test time was reduced to 90.5 seconds without
sacrificing any excess emissions identification. Excess emissions identification remained
at 94.4% for HC, 95.4% for CO, and 95.7% for NOx.
SIP Credit Analysis
The comparison of excess emissions identification between the IM240 and IM147 that is
presented above is based on the use of CO Max standards for both test cycles. However,
to develop an estimate of the allowable SIP credit that should be allocated to the revised
IM147 CO Max standards, it is also necessary to compare excess emissions identification
between this scenario and the IM240 with EPA-recommended final cutpoints in place.
This is due to the need to establish a link to using MOBILE for SIP modeling purposes.
Configuring MOBILE with CO Max standards is not feasible; therefore, a better
approach is to run the model with final EPA standards in place and use the excess
emissions identification rates developed in this study to adjust the resulting outputs.
Table 5-7 shows the excess emission identification rates when the IM147 Max CO
cutpoints are compared to the IM240 final standards. Pollutant-specific identification
rates are shown both without and with the fast-pass, retest, or fail-fail algorithms enabled.
(The latter scenario includes fast-passing vehicles as early as at the end of segment 2.)
Since Arizona will be implementing the IM147 test procedure with the algorithms
enabled, the identification rates for this scenario are the ones that should be used to adjust
the MOBILE modeling results (based on final IM240 standards) for SIP credit purposes.
As expected, the table shows that HC and NOx identification rates are significantly lower
with the IM147 Max CO cutpoints relative to final IM240 standards. This is due to the
fact that the Max CO cutpoints are designed to maximize the CO benefits of the program
at the expense of HC and NOx benefits, while keeping maximum failure rates in each
cutpoint category to acceptable levels. The CO identification rate of 97.9% (with the
algorithms enabled) shows that the Arizona program will achieve nearly all of the
modeled benefit of the final IM240 standards. Note that this will be substantially more
-53-
-------
Table 5-7
Comparison of IM147 Max CO Cutpoints to IM240 Final Standards
Impact on Excess Emissions Lost*
Class
% Excess Emission Identified
(Without Fast-Pass, Retest, Fast-Fail)
% Excess Emissions Identified
(With Integrated Algorithms)
HC
CO
NOx
HC
CO
NOx
LDGV
95.2%
96.2%
84.8%
91.9%
95.1%
85.7%
LDGT1
80.5%
100.0%
70.6%
68.4%
98.1%
81.2%
LDGT2
87.9%
100.0%
46.6%
98.5%
97.7%
73.0%
Weighted
Average
91.3%
97.3%
79.9%
86.7%
95.9%
81.6%
* Percent of IM240 (Final Standards) Excess Emissions Identified was determined using the 835-vehicle
sample.
effective than the current phase-in IM240 standards. The table also shows that the
addition of the integrated algorithms results in less than a 1% reduction in the excess
emissions identification rate for CO.
Need for Follow-Up Analysis
As discussed above, the analysis results presented in the report are based on a relatively
small sample of IM147 and IM240 data. While the available data are significantly more
robust than the previous sample of 300 vehicles, it is clear that these results should be
revisited with a much larger sample once IM147 testing is initiated in Arizona. We
therefore recommend that as soon as one to two months of IM147 data are collected in
the program, they should be used to verify the validity of the cutpoints and algorithms
developed in this study. This follow-up analysis would allow for any required fine-
tuning of the cutpoints and algorithms.
###
-54-
-------
6. DEVELOPMENT OF IM147 VARIATION LIMITS
In addition to fast-pass/fail cutpoints and retest criteria, trace variation limits were also
developed for the IM147 test. Under Task 6 of the Work Assignment, an analysis
methodology that was used to develop IM240 variation limits under an earlier EPA
study2 was applied to the pilot IM147 data to develop similar limits for the IM147 test.
During the course of the effort, an alternate statistical metric, Cumulative Positive
Specific Power (CPP), was identified that resulted in better second-by-second variation
limits than the Positive Kinetic Energy (PKE) metric employed in the previous EPA
study. This section of the report describes why the new statistic was selected, and how
IM147 CPP variation limits were developed and evaluated to ensure they do not produce
excessive test abort rates. It also assesses their individual impact on average
dynamometer test time. (The combined effect of fast-pass/fail cutpoints, retest criteria
and IM147 trace variation limits is discussed in Section 7.)
Before describing the effort performed under the current Work Assignment, a review of
the existing IM240 tolerance limits and a summary of the previous evaluation of those
limits are presented.
Existing Tolerance Limits
The prescribed driving cycles for the transient IM240 and IM147 tests consist of varying
second-by-second speeds ranging from zero (i.e., idle) to 56.7 mph, with maximum speed
changes of ą3.3 mph/sec. During actual I/M testing, the driver watches a graphical
display of the prescribed or "reference" speed/time trace overlayed with the actual
second-by-second trace as it is being driven as an aid to following the reference trace and
anticipating upcoming speed changes. (The visual display also indicates prescribed shift
points for manual transmission vehicles along the trace.)
Since each vehicle has different performance characteristics, it is impossible, even for
highly skilled drivers, to precisely follow the second-by-second reference trace speeds
during actual "one-time-only" testing. As a result, EPA originally developed a set of
speed-based tolerance limits for the IM240 test that defined the leeway allowed to the
driver in trying to follow the reference trace for the test to be considered valid. Those
tolerance limits consisted of two components: (1) speed excursion limits; and (2) speed
variation limits. Each of these criteria is described below.6
-55-
-------
Speed Excursion Limits f85.2221 (e) (4)1 - Speed excursion limits shall apply
as follows :
(i). The upper limit is 2 mph higher than the highest point on the trace within
1 second of the given time.
(ii). The lower limit is 2 mph lower than the lowest point on the trace within 1
second of the given time.
(iii). Speed variations greater than the tolerances (e.g., during gear changes) are
acceptable provided they occur for no more than 2 seconds on any occasion.
(iv). Speeds lower than those prescribed during accelerations are acceptable
provided they occur for no more than 2 seconds on any occasion.
Speed Variation Limits [85.2221 Ye) (5)1
(i). A linear regression of feedback value on reference value shall be performed on
each transient driving cycle for each speed using the method of least squares,
with the best fit equation having the form: y = mx + b, where:
(A), y = the feedback (actual) value of speed;
(B). m = the slope of the regression line;
(C). x = the reference value; and
(D). b = the y-intercept of the regression line.
(ii). The standard error of estimate (SE) ofy on x shall be calculated for each
regression line. A transient driving cycle lasting the full 240 seconds that
exceeds the following criteria shall be void and the test shall be repeated:
(A). SE = 2.0 mph maximum.
(B). m= 0.96 -1.01.
(C). r2 = 0.97 minimum.
(D). b = ą2.0 mph.
Simply stated, the speed excursion limits require that vehicles be driven within ą2 mph of
the reference trace, accommodating for gear changes and other momentary excursions.
-56-
-------
The speed variation limits, though a bit more difficult to comprehend, were intended to
ensure that speed differences from the reference trace within the ą2 mph excursion limits
"envelope" would not bias the resulting measured emissions.
However, EPA suspended the use of the speed variation limits on November 23, 1993,
pending further evaluation.
Previous Analysis of Speed Variation Limits
Inadequacy of Speed Variation Limits - In 1998, Sierra completed a study for EPA2 that
evaluated the ability of the linear regression-based speed variation limits to identify high
emissions-producing speed variations. It was found that the linear regression criteria
were inadequate in flagging high-emissions speed variations. The reason for this finding
was that the standard error (SE) statistic does not give "appropriate" higher weighting to
speed deviations occurring at the critical high-emission points along the IM240 trace.
Instead, it gives equal weighting to all speed variations and thus (along with the other
regression statistics used in the speed variation criteria) is ill-suited to identifying those
speed deviations that substantially affect IM240 emissions.
Evaluation of Alternative Statistics - During this study, two alternative statistical
measures were evaluated for their ability to better identify IM240 speed variations that
significantly affect measured emissions:
1. DPWRSUM7 - the sum of absolute changes in specific power; and
2. Positive Kinetic Energy fPKE) - the sum of positive differences in kinetic
energy per unit distance.
It was found that the PKE statistic provided a better measure than DPWRSUM for
identifying those speed variations from the reference trace that produce high emissions.
This finding was supported by analysis of modal (i.e., second-by-second) speed and
emissions data from a random sample of 16,581 full* IM240 tests from the Arizona I/M
program. It was determined that the high-emission portions of the IM240 test closely
corresponded with periods of acceleration. A examination of both the DPWRSUM and
PKE statistics found that DPWRSUM is increased during both decelerations and
accelerations. PKE, on the other hand, is increased only during acceleration periods.
Thus, the DPWRSUM statistic is "diluted" with speed variations during decelerations
that have little effect on emissions. As a result, the PKE statistic was reasoned to provide
a better measure of significant emissions-producing speed variations.
Development of PKE-Based IM240 Variation Limits - From this finding, PKE variation
limits were then developed as potential replacements to the original regression-based
"Full" tests refer to those run over the entire test duration, regardless of fast-pass or fast-fail status.
-57-
-------
speed variation limits. The basic approach used to develop PKE-based speed variation
criteria for the IM240 consisted of the following elements:
1. Establishing upper and lower "composite" PKE limits for full 240-second tests
from the Arizona data sample; and
2. Scaling these composite limits based on the cumulative PKE at each second of
the IM240 reference trace to produce second-bv-second PKE variation limits.
Second-by-second PKE variation limits were established to ensure that compliance with
the reference trace was maintained throughout the test to minimize emissions bias during
fast-pass and fast-fail determinations.
From analysis of-the Arizona data, the PKE variation limits were established over a range
expected to produce no more than a 3% increase in the test abort rate when applied as a
replacement for the speed variation criteria, in conjunction with the existing speed
excursion (i.e.. ą2 mph) criteria-
Development of Positive Power-Based Variation Limits
Under the current study, the randomly collected triplicate Arizona IM147 tests described
in Section 3 were also analyzed to develop IM147 trace variation limits. Similar to the
earlier IM240 study, upper and lower composite limits were first established over the full
duration of the IM147 test. The composite variation limits were then scaled at each
second of the reference trace to produce second-by-second variation limits.
Before the details of how these IM240-developed methodologies were adapted for the
IM147 test are discussed, an explanation of why positive power was used as a
replacement for PKE as the variation limits metric is provided.
Use of Positive Power Instead of PKE - During the course of developing the second-by-
second IM147 variation limits, a number of tests were identified for which PKE-based
variation limits were being exceeded under periods of deceleration. This was clearly a
problem. As discussed earlier, statistical metrics used to establish variation criteria were
selected based on their ability to identify high emissions-producing variations that
coincide with acceleration events. Although the composite PKE statistic does this well
on a cumulative basis over the entire test, it is less-suited when applied on a second-by-
second basis.
The reason for this can be seen by first considering how PKE is calculated. Over a
traveled driving cycle of distance x, cumulative PKE per unit distance is defined as
follows:
-------
Em
CPKE(x) = x [6-1]
jcbc
The term PP, is referred to as the positive specific power at time t and is given by the
following equation:
iV2 - (V<-i)2 whenV, >Vtml
f 0 whenV, ŁV,_,
PP' = i [6-2]
By definition, positive (specific) power* is non-zero during acceleration and zero during
cruise and deceleration events. Cumulative positive power (CPP) is defined as the sum of
positive power at each second t over a transient driving cycle of T seconds, or
[6-3]
CPP(T) = Y,PP,
(=0
Thus, CPP increases during accelerations over a transient driving cycle and remains
constant during cruise and deceleration.
Conversely, cumulative PKE decreases during cruise and deceleration because the
denominator in Equation [6-1], which represents the summed distance driven, still
increases while a vehicle is cruising or decelerating. Cumulative PKE becomes constant
only during periods of idle (i.e., zero speed).
Figure 6-1 illustrates the different behavior of each metric over the IM147 test. It shows
second-by-second speed, CPP, and cumulative PKE over the reference trace. Speed (in
mph) is plotted against the left axis. Cumulative positive power and cumulative PKE are
plotted against the right axis (in mph2/sec and miles/hr2, respectively).
As seen in Figure 6-1, CPP either increases or remains constant over the entire duration
of the transient IM147 test. On the other hand, cumulative PKE decreases during both
deceleration and cruise events, by definition. Furthermore, cumulative PKE can actually
decrease during modest accelerations after the initial portion of the IM147 test. For
Power is literally defined as the rate of change in kinetic energy (or work). Strictly speaking, specific
power (i.e., power per unit mass) would be calculated by the velocity times the acceleration at time t.
When the change in kinetic energy is evaluated on a second-by-second basis as defined in Equation 6-2,
instead of as a net change from the beginning to the end of the cycle, PP, as defined in that equation
approaches the strict definition of positive power.
-59-
-------
Figure 6-1
Comparison of IM147 Variation Limit Metrics
Time (seconds)
example, this phenomenon can be seen at Seconds 11-14 in Figure 6-1, where the
reference trace exhibits acceleration from Second 5 through Second 14 but cumulative
PKE begins dropping beyond Second 11.
Because of this behavior, it was difficult to establish reasonable second-by-second
variation limits based upon cumulative PKE differences between the reference and driven
traces using a scaling approach similar to that developed under the 1998 IM240 study.
That approach basically consisted of scaling the composite PKE interval limits
established at the end of the test by the percentage difference of these limits from the
composite reference value.
Since the magnitude of critical emissions-affecting deviations during accelerations is
difficult to distinguish from the magnitude of deviations that do not substantially affect
emissions during deceleration and cruise, this scaling approach fails when based upon
cumulative PKE. The result is that cumulative PKE-based second-by-second variation
limits of a specific scaled interval width will either falsely flag less important deviations
during decelerations or not identify deviations during accelerations that do affect
measured emissions.
Thus, two alternate approaches were considered for establishing reasonable second-by-
second IM147 variation limits:
1. Use of a better-behaved statistic, such as positive power, in conjunction with
the basic scaling approach; and
2. Development of separate variation limits for each second using deviation
distributions (reference vs. actual) at each second compiled from a full modal
analysis of the test data sample.
-60-
-------
Although conceptually appealing since variation limits are established independently for
each individual second, the latter approach would require a rigorous modal analysis just
to develop initial variation limits for each second. A complex iterative process would
then be required to evaluate these limits over the entire trace and "tune" them in a manner
that yields an acceptable overall abort rate. This timing step would also consider the
relative impact of trace variations at each second on measured emissions. For example, it
would be desirable to apply tighter trace variation limits during acceleration segments
that produce high emissions than during less emissions-significant idling segments if the
resulting overall abort rates (across all segments) can be kept at acceptable levels.
In addition, it is believed that imposing tighter, independently established second-by-
second trace variation limits would result in a much greater degree of "re-leaming" by the
lane inspectors as they adjust to the impact of these limits. This re-learning process and
any necessary re-tuning of the variation limits could best be evaluated through an initial
pilot study before being implemented on a program-wide basis. During this pilot study, it
is also envisioned that different approaches to providing visual feedback to drivers as
they attempt to follow the trace could be evaluated. For example, this could involve
providing dynamically updated forward-looking "trace envelopes" or speeds that guide
the driver back toward the reference trace when a nominal excursion begins in a manner
that complies with the second-by-second variation limits.
As a result of the scope of the latter approach, the first approach was selected because it
was believed to substantially overcome the shortcoming of the PKE-based metric while
being less resource-intensive to apply and test than limits developed from a full modal
analysis. It should also be noted that using CPP as a replacement for PKE assumes that
the existing ą2 mph criteria developed by EPA will be applied in conjunction with CPP-
based variation limits. This assumption is dictated by the use of a cumulative metric in
specifying second-by-second variation limits.
Development of Composite Variation Limits - Composite IM147 variation limits were
generated using a similar methodology to that employed in the 1998 IM240 study, except
cumulative positive power (CPP), rather than cumulative PKE, was used as the statistical
metric.
Figure 6-2 shows the distribution of composite (i.e., 147-second) CPP calculated from the
second-by-second actual speeds in the Arizona data sample, expressed as the percent
difference between actual and reference IM147 CPP.
As the figure shows, actual CPP appears normally distributed, although the median CPP
is approximately 1% higher than the reference value. To determine how far to go along
the "tails" of the CPP differences distribution to set composite limits, CPP differences
among the test lane drivers in the Arizona data were examined. The basic concept
applied in setting the CPP limits was to identify a significant fraction of drivers who,
historically, could always (or nearly always) run IM147 tests within the selected CPP
limits. Given a mixture of ability among individual drivers to follow the reference trace,
Sierra sought to identify the fraction of "competent" drivers who could follow the trace
-------
Figure 6-2
Distribution of Arizona IM147 Positive Power Differences
% Difference Between Reference and Actual Positive Power
(Sample Size = 9,306)
14%
g 12%
o
5 10%
O B/o
O
g. 6%
2%
0%
Ś
111 11
I III I I I
I I 111 I I I I I
III III 1111111
-10%-9% -8% -7% -6% -5% -4% -3% -2% -1% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15%
Percent Difference from Reference Value
more consistently than others when conducting IM147 tests for a range of vehicles. This
subset of competent drivers and tests was used to establish composite CPP limits.
From the Arizona data sample of over 10,000 valid triplicate IM147 tests, records in .
which speed excursions (over ą2 mph from the reference trace) occurred were discarded,
leaving a remaining sample of 9,306 tests. For the purpose of establishing composite
CPP variation limits based on the capabilities of "good" drivers, speed excursion tests
were removed from this portion of the analysis.
The remaining data sample was then grouped by individual drivers, one for each of the
231 drivers that were found in the sample. Driver groups containing fewer than 25 test
records were then discarded; this left a total of 112 driver groups, which encompassed
86% of the tests in the total sample (i.e., before discarding small-sample driver groups).
Composite CPP was then calculated for each test in the remaining driver groups. The
mean and standard deviation of CPP from the tests within each driver group were also
computed. The driver groups were then ranked by increasing CPP standard deviation and
the top 50% of the drivers (based on lowest CPP standard deviation) were used to
compute possible composite CPP outpoints.
Table 6-1 lists a series of possible CPP cutpoints computed from the percentile CPP
variance among the top 50% drivers. For example, the CPP cutpoints shown for the 2%
row under the "Top 50% Percentile" column (the third column in Table 6-1) indicate that
96% (100% - 2 x 2%) of the tests from the top 50% drivers had composite CPP within
4,437 and 4,949 mph2/sec.
-62-
-------
Table 6-1
Preliminary CPP Outpoints (mph2/sec) Based on Top-50% Drivers
(Sample Size = 7,962 Tests)
Top-50%
Driver
Lower Tail
Top-50%
Driver
Upper Tail
Top-50%
Driver
Percentile
Low-End
CPP
(mphVsec)
High-End
CPP
(mphVsec.)
Interval
Width
(mphVsec)
0.25%
99.75%
ą0.25%
4,348
5,077
728
0.5%
99.5%
ą0.5%
4,368
5,039
671
0.75%
99.25%
ą0.75%
4,391
5,013
622
1.0%
99.0%
ą1.0%
4,403
4,987
584
2.0%
98.0%
ą2.0%
4,437
4,949
512
3.0%
97.0%
ą3.0%
4,459
4,920
462
4.0%
96.0%
ą4.0%
4,475
4,903
428
5.0%
95.0%
ą5.0%
4,483
4,882
398
Note that these preliminary outpoints are not centered about the IM147 reference CPP
value of 4,617 mph2/sec (as evidenced by the shifted CPP distribution shown earlier in
Figure 6-2). To generate a series of "final" composite limits for evaluation, Sierra
applied the interval widths shown in Table 6-1 to the reference CPP value to produce
"centered" limits about the reference value.
Incremental abort test rates for each set of centered cutpoints were then calculated based
on both the entire 9,306 test Arizona data sample and the top-50% driver subset. The
results are presented in Table 6-2, which lists both "simple" and "effective" abort rates.
Simple abort rates represent the fraction of tests in the sample for which the composite
CPP cutpoints would have been exceeded. Effective abort rates were calculated from
simple rates by subtracting the fractions of emission-pass tests that exceeded the upper
CPP cutpoint and emission-fail tests that exceeded the lower CPP cutpoint. The idea is
that tests on vehicles that had passing emission scores but were driven with high CPP
should not be aborted. Similarly, tests that failed on emissions despite being driven
below the lower CPP cutpoint should also be considered valid tests and not aborted.
Thus, tests should be aborted only when the upper CPP variation limit is exceeded for an
emissions failure or the lower CPP variation limit is exceeded during a passing test. The
emissions pass/fail determinations used to calculate the effective abort rates shown in
Table 6-2 are based upon the "Max CO" cutpoints developed earlier in the study.
In addition to the calculated test abort rates, Table 6-2 also shows the percentage of
drivers who are always within the limits of each set of CPP cutpoints.
-63-
-------
Table 6-2
Centered CPP Cutpoints and Resulting Test Abort Rates
Top-50%
Driver
Percentile
Centered CPP Limits
(mph2/sec)
Abort Rates (%)
from All Tests
Abort Rates (%) from
Top-50% Driver Tests
Percentage of All
Drivers Within Limits
Lower
CPP
Limit
Upper
CPP
Limit
Interval
Half-
Width
Simple
Abort
Rate
Effective
Abort
Rate
Simple
Abort
Rate
Effective
Abort
Rate
> Lower
Limit
< Upper
Limit
ą0.25%
4,253
4,982
364.2
5.8%
1.6%
1.2%
0.2%
95.7%
53.9%
ą0.5%
.4,282 ; |
4,953_ \
335,6 '
JW]
' , 2:1%*
^ " 1.9%;
-0.4%.
ATA%
ą0.75%
4,307
4,928
310.8
8.7%
2.6%
2.7%
0.6%
91.8%
40.5%
ą1.0%
4,325
4,909
292.1
10.5%
3.1%
3.7%
0.9%
88.4%
37.1%
ą2.0%
4,361
4,873
255.9
14.1%
4.2%
6.3%
1.6%
81.5%
28.9%
ą3.0%
4,386
4,848
230.9
17.3%
5.3%
8.9%
2.6%
77.6%
26.3%
ą4.0%
4,403
4,831
214.2
19.9%
6.2%
10.9%
3.1%
74.1%
24.6%
ą5.0%
4,418
4,816
199.1
22.6%
7.2%
13.2%
3.9%
67.2%
22.4%
Sample
Size
-
-
-
9,306 Tests
7,962 Tests
231 Drivers
Note: Shading indicates the CPP limits proposed for use by Sierra, and corresponding data.
Based on the results given in Table 6-2, Sierra proposes the use of composite lower and
upper CPP variation limits of 4,282 and 4,953 mph2/sec, respectively (shown in the
shaded row in Table 6-2). As indicated in the table, these composite CPP limits are
expected to increase the (effective) abort test rate by 2.1% relative to that resulting from
EPA's recommended ą 2 mph limits based on available test data. Since the goal of the
analysis was to specify variation limits that kept the abort rate due to these variation limit
violations to 3%, composite CPP limits resulting in only a 2.1% incremental abort rate
were selected. This left some "room" below the 3% target for the impact of also
imposing second-bv-second CPP variation limits.
If drivers are selected based on their ability to perform as well as the best 50% of the
current drivers, then the abort rate would drop to just 0.4%. In practice, it is expected that
the abort rate will increase by less than this amount as drivers "adjust" to the new limits.
Development of Second-bv-Second Variation Limits - Using the recommended
composite CPP limits of 4,282 and 4,953 mph2/sec, second-by second CPP limits were
generated by scaling the percentage difference of these limits from the composite
reference value (7.3%) to the CPP calculated at each second from the reference trace.
This approach was further modified as described below.
1. To provide drivers with a short period to "learn" to drive each test vehicle,
second-by-second CPP limits were not imposed until t=30 seconds; and
-64-
-------
2. To further accommodate wider allowable variations (on a percentage basis) in
second-by-second CPP at the beginning of the transient IM147 test, a "CPP
Multiplier" factor was applied that widened the allowed CPP limits
progressively from the end of the test to where limits begin at t=30 seconds.
From its maximum value beginning at t=30 seconds, the CPP multiplier factor
was linearly decreased to a value of unity (i.e., 1.0) at t=146. In other words, at
the end of the test, the CPP limits were set equal to the composite CPP limits.
Furthermore, this linear narrowing was applied only over the acceleration
sections of the IM147 trace, during which the reference CPP is increasing.
During cruise and deceleration periods, the limit widths were held constant (as
the reference CPP also remains constant).
This latter improvement to the methodology employed in the 1998 IM240 study, in
conjunction with the use of CPP instead of CPKE, enabled second-by-second IM147 CPP
limits to be specified so variation limit aborts were not falsely triggered during
deceleration and cruise portions of the transient test. Note that second-by-second CPP
variation limits developed in this manner can still be exceeded during cruise and
deceleration events, signaling tests that should be aborted. However, falsely triggered
"anomalous" exceedances that occur from the use of a PKE-based metric are eliminated
under this modified approach.
Figure 6-3 illustrates this modified second-by-second CPP-based variation limit concept.
The thick solid line shows the reference CPP over the IM147 test; the thinner solid lines
represent the lower and upper CPP limits established as described above.
These CPP traces are plotted against the left axis. Speed, indicated by the dashed line, is
plotted against the right axis. Note that the CPP limits can be held constant during cruise
and deceleration periods. Thus, variation limit exceedances in these intervals are real,
rather than anomalous artifacts of the statistical metric.
Figure 6-3
Illustration of Second-by-Second IM147 Cumulative Positive Power Variation Limits
Time (sec)
-65-
-------
To establish second-by-second CPP variation limits that produced expected test aborts
near the 3% target rate, a range of initial CPP multipliers (from 2.0 to 6.0) were
evaluated. These initial multipliers specify the width of the variation limits at the starting
point (t=30 seconds) relative to the composite interval width at the end of the test. For
example, an initial CPP multiplier of 2.0 means that the starting interval width was 14.6%
(2.0 x 7.3% composite CPP interval width) of the reference CPP trace at that point.
Figure 6-4 shows the increase in effective abort rate as a function of varying initial CPP
multipliers. The diagonally striped region represents lower CPP variation limit aborts,
the shaded region above shows upper limit aborts. Since these abort events are mutually
exclusive, their sum represents the total expected effective abort rate from
implementation of the CPP variation limits.
Figure 6-4
Effect of Initial CPP Limits Multiplier on Effective IM147 Abort Rate
(Sample Size = 9,306)
2.0 2.5 3.0 3.5 4.0 4.5 5.0
Initial CPP Limits Multiplier
Based on the analysis results, an initial CPP multiplier of 3.5 is recommended for
implementation in Arizona. Second-by-second CPP variation limits based on the use of
this multiplier are shown in Appendix E.
Evaluation of IM147 Variation Limits on Test Time
Using the second-by-second CPP variation limits described in the preceding section, an
analysis was conducted of the impact of these variation limits on average dynamometer
test time. This was a simplistic analysis since it addressed only the singular impact of the
variation limits. (A more exhaustive analysis of the combined test time impact of CPP
variation limits in conjunction with fast-pass, fast-fail and retest decisions is presented in
Section 7.)
-66-
-------
In this simple analysis, the "without limits" or base average test time was assumed to be
146 seconds, the length of a full IM147 test. This assumption was necessitated by the
Arizona data sample, which contained only full tests. The recommended CPP limits were
found to produce a total effective abort rate of 3.4% based on the Arizona IM147 data.
The average time at which the aborts occurred under these limits was determined to be
101.6 seconds.
Thus, the "with limits" average test time was then calculated as follows:
With Limits Test Time = Base Test Time + (Abort Rate x Average Abort Time)
= 146 + (3.4% x 101.6 seconds) = 149.5 seconds
It should be noted that this approach also assumes that an aborted test is performed
successfully on the subsequent re-test. Under these assumed conditions, the CPP
variation limits will increase average test times by approximately 2% [(149.5-146) +
146],
II II II
mm
-67-
-------
7. INTEGRATION OF CPP VARIATION LIMITS
The final phase of this analysis involved integrating the CPP criteria with the other
algorithms included in this study to determine net test time.
As shown in Section 6, there are both high-end CPP errors and low-end CPP errors.
High-end CPP errors occur when the vehicle is driven too aggressively, whereas low-end
CPP errors occur when the vehicle is driven too smoothly, essentially minimizing the
peaks and valleys of the trace. Because high-end CPP errors will create additional
emissions and in turn make it more difficult for a vehicle to pass the test, a test where a
vehicle passed in spite of high-end CPP errors is considered a valid test. If the same
vehicle failed because of high emissions, the cause is assumed to be the high-end CPP
error; therefore, the test would need to be extended to ensure fairness. Low-end CPP
errors, however, would result in lower mass emissions and make it easier for vehicles to
pass falsely. In those cases where the vehicle passes with a low-end CPP error, the result
would be invalid and the test would need to be extended. If a vehicle fails with a low-end
CPP error, the result is valid and the test would terminate. Table 7-1 -details which CPP
violations affect which decisions.
Table 7-1
End Test Decisions Affected by CPP Errors
Decision Type
Prohibited by:
Fast-Pass
Low-end CPP error
Retest
(initiate another IM147 cycle)
High-end CPP error
Fast-Fail
High-end CPP error
End-of-Test Pass
Low-end CPP error
End-of-Test Fail
High-end CPP error
While the above decision types can be prohibited by the corresponding CPP errors, the
error must occur while data for that decision were being produced in order to affect the
decision. In other words, not all high-end CPP errors occurring during an IM147 test
-68-
-------
would necessarily invalidate failing results, nor would all low-end CPP errors invalidate
passing results. If an applicable power error occurred while data required to make a
particular decision were being collected, then the decision would be invalidated and the
test would continue. Specifically, this would apply to fast-pass and fast-fail decisions.
For example, if the vehicle's emission data were clean enough to permit a fast-pass at
second 40, but there was a low power violation at second 34, then the fast-pass decision
would be invalidated and the test would continue. On the other hand, if the first power
violation occurred after second 40 (e.g., at second 41), then the vehicle could be fast-
passed at second 40.
In a related issue, the CPP error is reset at the conclusion of each IM147. As a result,
low-end CPP errors occurring during the first IM147 do not prohibit pass-oriented
decisions in subsequent IM147 tests. The same is true for high-end CPP errors and fail-
oriented decisions.
Using this logical framework, the CPP variation limits were applied to the 3,347-vehicle
sample to determine how this algorithm would affect test time. Since the CPP variation
limits are designed to be imposed in concert with the ą2 mph speed limits detailed in the
IM240 guidance, vehicles already failing the ą2 mph speed limits were eliminated from
the sample since they would be aborted regardless of the CPP outcome. Of the original
3,347 vehicles, 3,006 remained after these vehicles were eliminated from the sample.
The flow chart detailed in Figure 7-1 shows how the CPP decision integrates with the
fast-pass/fail and retest algorithms previously discussed in this report. Note that in cases
where a CPP violation prevented a decision during the third IM147, the vehicle would,
after completing the third IM147, restart the third IM147 again. For the average test time
computation, it was assumed that CPP errors during the third IM147 extended the test
time 146 seconds.
Given the 3,006-vehicle sample, the average test time, with the fast-pass, retest, and fast-
fail criteria enabled but without the CPP criteria applied, was 89.07 seconds. Once the
CPP criteria were enabled, 87 vehicles' tests (of the 3,006 vehicles) were extended,
increasing the average test time by 0.98 seconds (or 1.1%) to 90.05 seconds. This is less
than the 2% increase projected at the end of Section 6, which makes sense given that the
2% projection was made without the fast-pass/fail and retest algorithms in place. The
overall test time reductions caused by the fast-pass/fail and retest algorithms would mean
that fewer errors would be committed.
-------
Figure 7-1
Process for Integrating CPP Variation Limits
###
-70-
-------
8. REFERENCES
"Determination of Emissions Credit and Average Test Times for IM147
Testing," Prepared by Sierra Research for the U.S. Environmental Protection
Agency, Report No. SR99-10-02, October 10, 1999.
"Additional Study of Preconditioning Effects and Other IM240 Testing
Issues," Prepared by Sierra Research for the U.S. environmental Protection
Agency, Report No. SR98-02-01, February 2, 1998.
Heirigs, P.L. and J. Gordon, "Preconditioning Effects on I/M Test Results
Using IM240 and ASM Procedures," SAE Paper No. 962091, 1996.
Whitby, R., C. Shih, W. Webster, and R. Card, "Enhanced Inspection and
Maintenance Program: An Investigation of Real Time Data Utility and
Dynamic Quality Assurance Instrumentation - Task 1: A Predictive Modal
Regression Methodology for IM240 Fast-Pass/Fail Decisions," New York
State Department of Environmental Conservation, EPA Cooperative
Agreement #X992008-01-0, November 29, 1996.
"Analysis of Alternate IM240 Cutpoints, Phase 2 Testing, and Exempting
New Vehicle Models on Test Duration and Projected I/M Benefits," Prepared
by Sierra Research for the Arizona Department of Environmental Quality,
Report No. SR98-05-01, May 12, 1998.
"High-Tech I/M Test Procedures, Emission Standards, Quality Control
Requirements, and Equipment Specifications: IM240 and Functional
Evaporative System Tests," (Revised Draft), U.S. Environmental Protection
Agency, EPA-AA-RSPD-IM-96-1, June 1996.
W. J. Webster and C. Shih, "A Statistically Derived Metric to Monitor Time-
Speed Variability in Practical Emissions Testing," New York State
Department of Environmental Conservation, presented at the 6th CRC On-
Road Vehicle Emissions Workshop, March 18-20, 1996.
###
-71-
-------
Appendix A
Startup, Intermediate, and Final IM240 and IM147 Cutpoints
-------
Startup IM240 Outpoints and IM147 Cutpoints*
(Composite/Phase 2 Cutpoints in g/mi, IM240 - IM147)
Model Years
HC
CO
NOx
LDGV
1981-82
2.00/1.25 -2.00/1.20
60.0/48.0 - 58.0/30.0
3.0-3.3/1.2
1983-85
2.00/1.25 -2.00/1.20
30.0/24.0-30.0/15.0
3.0-3.3/1.2
1986-90
2.00/1.25 -2.00/1.00
30.0/24.0 - 30.0/10.0
3.0-3.0/1.2
1991-93
1.20/0.75 - 1.30/0.60
20.0/16.0-21.0/10.0
2.5-2.9/1.0
1994-95
1.20/0.75 - 1.20/0.60
20.0/16.0-21.0/10.0
2.5-2.7/1.0
1996+ (Tier 1)
0.80/0.50 - 0.80/0.50
15.0/12.0-15.0/7.0
2.0-2.1/0.9
LDGT1
1981-83
7.50/5.00 - 6.70/4.70
100.0/80.0 - 95.0/50.0
7.0 - 7.6/2.9
1984-85
3.20/2.00 - 2.90/2.00
80.0/64.0 - 76.0/40.0
7.0 - 7.6/2.9
1986-87
3.20/2.00-2.90/1.60
80.0/64.0-76.0/31.0
7.0 - 7.0/2.7
1988-90
3.20/2.00-2.90/1.60
80.0/64.0-76.0/31.0
3.5-3.6/1.3
1991-93
2.40/1.50-2.60/1.20
60.0/48.0-61.0/31.0
3.0-3.4/1.1
1994-95
2.40/1.50-2.40/1.20
60.0/48.0-61.0/29.0
3.0-3.2/1.1
1996+(Tier 1)
1.00/0.63 - 1.0/0.60
20.0/16.0-21.0/10.0
2.5-2.7/1.1
LDGT2
1981-83
7.50/5.00 - 6.70/4.70
100.0/80.0-95.0/50.0
7.0 - 7.6/2.9
1984-86
3.20/2.00 - 2.90/2.00
80.0/64.0 - 76.0/40.0
7.0 - 7.6/2.9
1987
3.20/2.00-2.90/1.60
80.0/64.0-76.0/31.0
7.0 - 7.6/2.7
1988-90
3.20/2.00-2.90/1.60
80.0/64.0-76.0/31.0
5.0-5.1/1.9
1991-93
2.40/1.50-2.60/1.20
60.0/48.0-61.0/31.0
4.5-5.1/1.9
1994-95
2.40/1.50-2.40/1.20
60.0/48.0-61.0/29.0
4.5-4.8/1.9
1996+ (Tier 1)
2.40/1.50-2.40/1.20
60.0/48.0-61.0/29.0
4.0-4.3/1.7
Developed for SR99-10-02
A-l
-------
Intermediate IM240 Outpoints and IM147 Cutpoints Developed in This Study
(Composite/Phase 2 Cutpoints in g/mi, IM240 - IM147)
Model Years
HC
CO
NOx
LDGV
1981-82
1.40/0.88- 1.40/0.90
45.0/36.0 - 44.0/23.0
2.3-2.8/1.0
1983-85
1.40/0.88 - 1.40/0.90
23.0/18.0-23.0/12.0
2.3-2.8/1.0
1986-90
1.40/0.88 - 1.40/0.70
23.0/18.0-23.0/9.0
2.3-2.6/1.0
1991-93
1.00/0.63 - 1.10/0.50
18.0/14.0-18.0/9.0
2.3 - 2.6/0.9
1994-95
1.00/0.63 - 1.00/0.50
18.0/14.0- 18.0/9.0
2.3 - 2.5/0.9
1996+ (Tier 1)
0.70/0.45 - 0.80/0.40
13.0/10.0 - 15.0/6.0
1.8-2.2/0.8
LDGT1
1981-83
5.50/3.50-4.90/3.40
85.0/68.0-81.0/43.0 .
5.8 - 6.3/2.4
1984-85
2.40/1.50-2.30/1.50
60.0/48.0 - 60.0/30.0
5.8 - 6.3/2.4
1986-87
2.40/1.50-2.30/1.20
60.0/48.0 - 60.0/26.0
5.8 - 5.8/2.2
1988-89
2.40/1.50-2.30/1.20
60.0/48.0 - 60.0/26.0
3.0-3.3/1.2
1990
2.40/1.50-2.30/1.20
60.0/48.0 - 59.0/26.0
3.0-3.3/1.2
1991-93
2.00/1.25 - 2.10/1.00
50.0/40.0 -51.0/26.0
2.8-3.2/1.1
1994-95
2.00/1.25-2.00/1.00
50.0/40.0-51.0/25.0
2.8-3.0/1.1
1996+(Tier 1)
0.90/0.57- 1.30/0.60
17.0/13.0-31.0/8.0
2.2-2.7/1.0
LDGT2
1981-83
5.50/3.50 - 4.90/3.40
85.0/68.0-81.0/43.0
5.8 - 6.3/2.4
1984-86
2.40/1.50-2.30/1.50
60.0/48.0 - 60.0/30.0
5.8 - 6.3/2.4
1987
2.40/1.50-2.30/1.20
60.0/48.0 - 60.0/26.0
5.8 - 6.3/2.2
1988-90
2.00/1.50 - 2.30/1.20
60.0/48.0-60.0/26.0
4.3-4.6/1.6
1991-93
2.00/1.25 -2.20/1.00
50.0/40.0-51.0/26.0
4.0-4.6/1.6
1994-95
2.00/1.25 -2.00/1.00
50.0/40.0-51.0/25.0
4.0-4.3/1.6
1996+(Tier 1)
1.60/1.00-2.00/0.90
38.0/30.0-51.0/18.0
3.0-4.1/1.3
A-2
-------
Final IM240 Cutpoints and IM147 Cutpoints Developed in This Study
(Composite/Phase 2 Cutpoints in g/mi, IM240 - IM147)
Model Years
HC
CO
NOx
LDGV
1981-82
0.80/0.50 - 0.80/0.50
30.0/24.0-30.0/15.0
2.0 - 2.3/0.8
1983-85
0.80/0.50-0.80/0.50
15.0/12.0-16.0/8.0
2.0 - 2.3/0.8
1986-89
0.80/0.50 - 0.80/0.50
15.0/12.0 - 16.0/8.0
2.0 - 2.2/0.8
1990-93
0.80/0.50 - 0.80/0.50
15.0/12.0- 15.0/8.0
2.0 - 2.2/0.7
1994-95
0.80/0.50 - 0.80/0.50
15.0/12.0 - 15.0/7.0
2.0 - 2.2/0.7
1996+ (Tier 1)
0.60/0.40 -0.80/0.30
10.0/8.0- 15.0/5.0
1.5-2.2/0.6
LDGT1
1981-83
3.40/2.00-3.10/2.10
70.0/56.0 - 67.0/35.0
4.5-4.9/1.8
1984-85
1.60/1.00- 1.70/1.00
40.0/32.0-43.0/20.0
4.5-4.9/1.8
1986-87
1.60/1.00- 1.70/0.80
40.0/32.0-43.0/20.0
4.5-4.6/1.7
1988-89
1.60/1.00- 1.70/0.80
40.0/32.0-43.0/20.0
2.5-2.9/1.0
1990-93
1.60/1.00- 1.60/0.80
40.0/32.0-41.0/20.0
2.5-2.9/1.0
1994-95
1.60/1.00- 1.60/0.80
40.0/32.0-41.0/20.0
2.5-2.7/1.0
1996+ (Tier 1)
0.80/0.50- 1.60/0.50
13.0/10.0-41.0/6.0
1.8 - 2.7/0.8
LDGT2
1981-83
3.40/2.00-3.10/2.10
70.0/56.0 - 67.0/35.0
4.5-4.9/1.8
1984-86
1.60/1.00- 1.70/1.00
40.0/32.0-43.0/20.0
4.5-4.9/1.8
1987
1.60/1.00- 1.70/0.80
40.0/32.0 - 43.0/20.0
4.5-4.9/1.7
1988-91
1.60/1.00- 1.70/0.80
40.0/32.0 -43.0/20.0
3.5-4.0/1.3
1992-93
1.60/1.00- 1.70/0.80
40.0/32.0-41.0/20.0
3.5-4.0/1.3
1994-95
1.60/1.00- 1.70/0.80
40.0/32.0-41.0/20.0
3.5-3.8/1.3
1996+ (Tier 1)
0.80/0.50- 1.60/0.50
15.0/12.0-41.0/7.0
2.0-3.8/0.9
A-3
-------
Appendix B
Max CO, Startup, Intermediate, and Final IM147 Failure Rates
-------
Failure Rate
Max CO Cutpoints
HC
CO
NOx
OVERALL
V Type
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
First IM147
LDGV
207
1573
11.6%
352
1428
19.8%
236
1544
13.3%
543
1237
30.5%
LDGT1
76
888
7.9%
204
760
21.2%
83
881
8.6%
281
683
29.1%
LDGT2
31
572
5.1%
87
516
14.4%
44
559
7.3%
127
476
21.1%
All
314
3033
9.4%
643
2704
19.2%
363
2984
10.8%
951
2396
28.4%
Second IM147
LDGV
126
1654
7.1%
256
1524
14.4%
166
1614
9.3%
406
1374
22.8%
LDGT1
59
905
6.1%
167
797
17.3%
70
894
7.3%
231
733 .
24.0%
LDGT2
24
579
4.0%
61
542
10.1%
31
572
5.1%
91
512
15.1%
All
209
3138
6.2%
"484
2863
14.5%
267
3080
8.0%
728
2619
21.8%
Third IM147
LDGV
114
1666
6.4%
250
1530
14.0%
155
1625
8.7%
390
1390
21.9%
LDGT1
49
915
5.1%
156
808
16.2%
70
894
7.3%
219
745
22.7%
LDGT2
25
578
4.1%
63
540
10.4%
30
573
5.0%
93
510
15.4%
All
188
3159
5.6%
469
2878
14.0%
255
3092
7.6%
702
2645
21.0%
Failure Rate By Model Year Grouping - First IM147
Max CO Cutpoints
HC
CO
NOx
OVERALL
V Type
Year
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
19
86
18.1%
32
73
30.5%
24
81
22.9%
55
50
52.4%
83-85
48
180
21.1%
108
120
47.4%
53
175
23.2%
144
84
63.2%
86-89
85
340
20.0%
96
329
22.6%
87
338
20.5%
169
256
39.8%
90-95
55
897
5.8%
112
840
11.8%
72
880
7.6%
171
781
18.0%
96+
0
70
0.0%
4
66
5.7%
0
70
0.0%
4
66
5.7%
LDGT1
81-85
38
222
14.6%
109
151
41.9%
54
206
20.8%
152
108
58.5%
86-89
22
200
9.9%
57
165
25.7%
17
205
7.7%
73
149
32.9%
90-95
16
434
3.6%
38
412
8.4%
11
439
2.4%
55
395
12.2%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
18
76
19.1%
44
50
46.8%
16
78
17.0%
55
39
58.5%
86-89
8
56
12.5%
13
51
20.3%
13
51
20.3%
27
37
42.2%
88-95
5
422
1.2%
30
397
7.0%
15
412
3.5%
45
382
10.5%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
314
3033
9.4%
643
2704
19.2%
363
2984
10.8%
951
2396
28.4%
B-l
-------
Failure Rate By Model Year Grouping - Second IM147
Max CO Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
13
92
12.4%
29
76
27.6%
18
87
17.1%
49
'56
46.7%
83-85
32
196
14.0%
81
147
35.5%
43
185
18.9% '
116
112
50.9%
86-89
57
368
13.4%
82
343
19.3%
67
358
15.8%
141
284
33.2%
90-95
24
928
2.5%
64
888
6.7%
38
914
4.0%
100
852
10.5%
96+
0
70
0.0%
0
70
0.0%
0
70
0.0%
0
70
0.0%
LDGT1
81-85
33
227
12.7%
99
161
38.1%
48
212
18.5%
139
121
53.5%
86-89
19
203
8.6%
48
174
21.6%
10
212
4.5%
59
163
26.6%
90-95
7
443
1.6%
20
430
4.4%
11
439
2.4%
32
418
7.1%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
16
78
17.0%
40
54
42.6%
14
80
14.9%
53
41
56.4%
86-89
5
59
7.8%
11
53
17.2%
11
53
17.2%
22
42
34.4%
88-95
3
424
0.7%
10
417
2.3%
6
421
1.4%
16
411
3.7%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
209
3138
6.2%
484
2863
14.5%
267
3080
8.0%
728
2619
21.8%
Failure Rate By Model Year Grouping - Third IM147
Max CO Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
13
92
12.4%
29
76
27.6%
16
89
15.2%
46
59
43.8%
83-85
29
199
12.7%
79
149
34.6%
39
189
17.1%
110
118
48.2%
86-89
48
377
11.3%
77
348
18.1%
65
360
15.3%
137
288
32.2%
90-95
24
928
2.5%
64
888
6.7%
35
917
3.7%
96
856
10.1%
96+
0
70
0.0%
1
69
1.4%
0
70
0.0%
1
69
1.4%
LDGT1
81-85
31
229
11.9%
95
165
36.5%
47
213
18.1%
134
126
51.5%
86-89
15
207
6.8%
43
179
19.4%
12
210
5.4%
56
166
25.2%
90-95
3
447
0.7%
18
432
4.0%
10
440
2.2%
28
422
6.2%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
16
78
17.0%
40
54
42.6%
13
81
13.8%
53
41
56.4%
86-89
6
58
9.4%
13
51
20.3%
10
54
15.6%
23
41
35.9%
88-95
3
424
0.7%
10
417
2.3%
7
420
1.6%
17
410
4.0%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
188
3159
5.6%
469
2878
14.0%
255
3092
7.6%
702
2645
21.0%
B-2
-------
Failure Rate
Startup Cutpoints
V Type
HC
CO
NOx
OVERALL
Fail Pass % Fail
Fail | Pass | % Fail
Fail | Pass % Fail
Fail Pass % Fail
First 1M147
LDGV
180
1600
10.1%
179
1601
10.1%
216
1564
12.1%
416
1364
23.4%
LDGT1
75
889
7.8%
37
927
3.8%
84
880
8.7%
166
798
17.2%
LDGT2
56
547
9.3%
37
566
6.1%
45
558
7.5%
103
500
17.1%
All
311
3036
9.3%
253
3094
7.6%
345
3002
10.3%
685
2662
20.5%
Second IM147
LDGV
109
1671
6.1%
133
1647
7.5%
152
1628
8.5%
291
1489
16.3%
LDGT1
51
913
5.3%
40
924
4.1%
51
913
5.3%
115
849
11.9%
LDGT2
32
571
5.3%
31
572
5.1%
34
569
5.6%
74
529
12.3%
All
192
3155
5.7%
204
3143
6.1%
237
3110
7.1%
480
2867
14.3%
Third IM147
LDGV
106
1674
6.0%
130
1650
7.3%
141
1639
7.9%
281
1499
15.8%
LDGT1
52
912
5.4%
33
931
3.4%
48
916
5.0%
108
856
11.2%
LDGT2
33
570
5.5%
30
573
5.0%
28
575
4.6%
67
536
11.1%
All
191
3156
5.7%
193
3154
5.8%
217
3130
6.5%
456
2891
13.6%
Failure Rate By Model Year Grouping - First IM147
Startup Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
30
75
28.6%
13
92
12.4%
24
81
22.9%
48
57
45.7%
83-85
56
172
24.6%
75
153
32.9%
49
179
21.5%
122
106
53.5%
86-89
57
368
13.4%
50
375
11.8%
72
353
16.9%
128
297
30.1%
90-95
36
916
3.8%
37
915
3.9%
71
881
7.5%
114
838
12.0%
96+
1
69
1.4%
4
66
5.7%
0
70
0.0%
4
66
5.7%
LDGT1
81-85
44
216
16.9%
24
236
9.2%
21
239
8.1%
74
186
28.5%
86-89
24
198
10.8%
11
211
5.0%
29
193
13.1%
53
169
23.9%
90-95
7
443
1.6%
2
448
0.4%
32
418
7.1%
37
413
8.2%
96+
0
32
0.0%
0
32
0.0%
2
30
6.3%
2
30
6.3%
LDGT2
81-85
29
65
30.9%
26
68
27.7%
9
85
9.6%
41
53
43.6%
86-89
14
50
21.9%
6
58
9.4%
7
57
10.9%
20
44
31.3%
88-95
13
414
3.0%
5
422
1.2%
29
398
6.8%
42
385
9.8%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
311
3036
9.3%
253
3094
7.6%
345
3002
10.3%
685
2662
20.5%
B-3
-------
Failure Rate By Model Year Grouping - Second IM147
Startup Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
24
81
22.9%
13
92
12.4%
18
87
17.1%
38
67
36.2%
83-85
35
193
15.4%
53
175
23.2%
43
185
18.9%
92
136
40.4%
86-89
36
389
8.5%
42
383
9.9%
55
370
12.9%
101
324
23.8%
90-95
14
938
1.5%
25
927
2.6%
36
916
3.8%
60
892
6.3%
96+
0
70
0.0%
0
70
0.0%
0
70
0.0%
0
70
0.0%
LDGT1
81-85
31
229
11.9%
23
237
8.8%
18
242
6.9%
58
202
22.3%
86-89
18
204
8.1%
14
208
6.3%
18
204
8.1%
39
183
17.6%
90-95
2
448
0.4%
3
447
0.7%
14
436
3.1%
17
433
3.8%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
20
74
21.3%
23
71
24.5%
9
85
9.6%
35
59
37.2%
86-89
8
56
12.5%
5
59
7.8%
8
56
12.5%
16
48
25.0%
88-95
4
423
0.9%
3
424
0.7%
17
410
4.0%
23
404
5.4%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
192
3155
5.7%
204
3143
6.1%
237
3110
7.1%
480
2867
14.3%
Failure Rate By Model Year Grouping - Third IM147
Startup Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
23
82
21.9%
13
92
12.4%
14
91
13.3%
35
70
33.3%
83-85
33
195
14.5%
56
172
24.6%
39
189
17.1%
90
138
39.5%
86-89
35
390
8.2%
36
389
8.5%
56
369
13.2%
96
329
22.6%
90-95
15
937
1.6%
24
928
2.5%
32
920
3.4%
59
893
6.2%
96+
0
70
0.0%
1
69
1.4%
0
70
0.0%
1
69
1.4%
LDGT1
81-85
34
226
13.1%
20
240
7.7%
17
243
6.5%
56
204
21.5%
86-89
17
205
7.7%
12
210
5.4%
16
206
7.2%
35
187
15.8%
90-95
1
449
0.2%
1
449
0.2%
14
436
3.1%
16
434
3.6%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
21
73
22.3%
22
72
23.4%
7
87
7.4%
33
61
35.1%
86-89
9
55
14.1%
6
58
9.4%
7
57
10.9%
16
48
25.0%
88-95
3
424
0.7%
2
425
0.5%
14
413
3.3%
18
409
4.2%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
191
3156
5.7%
193
3154
5.8%
217
3130
6.5%
456
2891
13.6%
B-4
-------
Failure Rate
Intermediate Cut
>oints
V Type
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
First IM147
LDGV
286
1494
16.1%
238
1542
13.4%
281
1499
15.8%
533
1247
29.9%
LDGT1
121
843
12.6%
62
902
6.4%
122
842
12.7%
247
717
25.6%
LDGT2
84
519
13.9%
47
556
7.8%
71
532
11.8%
142
461
23.5%
All
491
2856
14.7%
347
3000
10.4%
474
2873
14.2%
922
2425
27.5%
Second EM147
LDGV
170
1610
9.6%
161
1619
9.0%
214
1566
12.0%
386
1394
21.7%
LDGT1
87
877
9.0%
57
907
5.9%
78
886
8.1%
176
788
18.3%
LDGT2
46
557
7.6%
40
563
6.6%
48
555
8.0%
100
503
16.6%
All
303
3044
9.1%
258
3089
7.7%
340
3007
10.2%
662
2685
19.8%
- Third IM147
LDGV
157
1623
8.8%
161
1619
9.0%
196
1584
11.0%
367
1413
20.6%
LDGT1
80
884
8.3%
53
911
5.5%
73
891
7.6%
165
799
17.1%
LDGT2
40
563
6.6%
42
561
7.0%
42
561
7.0%
93
510
15.4%
All
277
3070
8.3%
256
3091
7.6%
311
3036
9.3%
625
2722
18.7%
Failure Rate By Model Year Grouping - First IM147
Intermediate Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
52
53
49.5%.
20
85
19.0%
27
78
25.7%
64
41
61.0%
83-85
88
140
38.6%
91
137
39.9%
64
164
28.1%
146
82
64.0%
86-89
94
331
22.1%
65
360
15.3%
100
325
23.5%
169
256
39.8%
90-95
51
901
5.4%
58
894
6.1%
90
862
9.5%
150
802
15.8%
96+
1
69
1.4%
4
66
5.7%
0
70
0.0%
4
66
5.7%
LDGT1
81-85
71
189
27.3%
42
218
16.2%
37
223
14.2%
117
143
45.0%
86-89
36
186
16.2%
14
208
6.3%
39
183
17.6%
71
151
32.0%
90-95
14
436
3.1%
6
444
1.3%
44
406
9.8%
57
393
12.7%
96+
0
32
0.0%
0
32
0.0%
2
30
6.3%
2
30
6.3%
LDGT2
81-85
37
57
39.4%
32
62
34.0%
17
77
18.1%
53
41
56.4%
86-89
19
45
29.7%
9
55
14.1%
10
54
15.6%
24
40
37.5%
88-95
28
399
6.6%
6
421
1.4%
44
383
10.3%
65
362
15.2%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
491
2856
14.7%
347
3000
10.4%
474
2873
14:2%
922
2425
27.5%
B-5
-------
Failure Rate By Model Year Grouping - Second IM147
Intermediate Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
36
69
34.3%
15
90
14.3%
23
82
21.9%
49
56
46.7%
83-85
51
177
22.4%
66
162
28.9%
60
168
26.3%
118
110
51.8%
86-89
61
364
14.4%
50
375
11.8%
76
349
17.9%
134
291
31.5%
90-95
22
930
2.3%
30
922
3.2%
55
897
5.8%
85
867
8.9%
96+
0
70
0.0%
0
70
0.0%
0
70
0.0%
0
70
0.0%
LDGT1
81-85
57
203
21.9%
34
226
13.1%
31
229
11.9%
95
165
36.5%
86-89
24
198
10.8%
18
204
8.1%
25
197
11.3%
52
170
23.4%
90-95
6
444
1.3%
5
445
1.1%
21
429
4.7%
28
422
6.2%
96+
Q
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
29
65
30.9%
32
62
34.0%
15
79
16.0%
50
44
53.2%
86-89
10
54
15.6%
5
59
7.8%
11
53
17.2%
19
45
29.7%
88-95
7
420
1.6%
3
424
0.7%
22
405
5.2%
31
396
7.3%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
303
3044
9.1%
258
3089
7.7%
340
3007
10.2%
662
2685
19.8%
Failure Rate By Model Year Grouping - Third IM147
Intermediate Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
34
71
32.4%
16
89
15.2%
20
85
19.0%
48
57
45.7%
83-85
48
180
21.1%
66
162
28.9%
58
170
25.4%
113
115
49.6%
86-89
56
369
13.2%
49
376
11.5%
72
353
16.9%
128
297
30.1%
90-95
19
933
2.0%
29
923
3.0%
46
906
4.8%
77
875
8.1%
96+
0
70
0.0%
1
69
1.4%
0
70
0.0%
1
69
1.4%
LDGT1
81-85
56
204
21.5%
34
226
13.1%
29
231
11.2%
92
168
35.4%
86-89
21
201
9.5%
16
206
7.2%
25
197
11.3%
50
172
22.5%
90-95
3
447
0.7%
3
447
0.7%
18
432
4.0%
22
428
4.9%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
25
69
26.6%
31
63
33.0%
14
80
14.9%
47
47
50.0%
86-89
10
54
15.6%
8
56
12.5%
8
56
12.5%
19
45
29.7%
88-95
5
422
1.2%
3
424
0.7%
20
407
4.7%
27
400
6.3%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
277
3070
8.3%
256
3091
7.6%
311
3036
9.3%
625
2722
18.7%
B-6
-------
Failure Rate
Final Cutpoints
V Type
HC
CO
NOx
OVERALL
Fail Pass % Fail
Fail Pass | % Fail
Fail | Pass % Fail
Fail Pass % Fail
First IM147
LDGV
499
1281
28.0%
346
1434
19.4%
396
1384
22.2%
733
1047
41.2%
LDGT1
210
754
21.8%
109
855
11.3%
184
780
19.1%
348
616
36.1%
LDGT2
135
468
22.4%
66
537
10.9%
120
483
19.9%
219
384
36.3%
All
844
2503
25.2%
521
2826
15.6%
700
2647
20.9%
1300
2047
38.8%
Second IM147
LDGV
309
1471
17.4%
247
1533
13.9%
296
1484
16.6%
539
1241
30.3%
LDGT1
137
827
14.2%
91
873
9.4%
134
830
13.9%
256
708
26.6%
LDGT2
77
526
12.8%
56
547
9.3%
88
515
14.6%
157
446
26.0%
All
523
2824
15.6%
394
2953
11.8%
518
2829
15.5%
952
2395
28.4%
Third IM147
LDGV
280
1500
15.7%
240
1540
13.5%
277
1503
15.6%
507
1273
28.5%
LDGT1
120
844
12.4%
85
879
8.8%
127
837
13.2%
239
725
24.8%
LDGT2
73
530
12.1%
56
547
9.3%
82
521
13.6% ,
149
454
24.7%
All
473
2874
14.1%
381
2966
11.4%
486
2861
14.5%
895
2452
26.7%
Failure Rate By Model Year Grouping - First IM147
Final Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
70
35
66.7%
30
75
28.6%
39
66
37.1%
78
27
74.3%
83-85
148
80
64.9%
120
108
52.6%
81
147
35.5%
183
45
80.3%
86-89
179
246
42.1%
102
323
24.0%
144
281
33.9%
247
178
58.1%
90-95
100
852
10.5%
90
862
9.5%
132
820
13.9%
220
732
23.1%
96+
2
68
2.9%
4
66
5.7%
0
70
0.0%
5
65
7.1%
LDGT1
81-85
116
144
44.6%
69
191
26.5%
70
190
26.9%
165
95
63.5%
86-89
71
151
32.0%
29
193
13.1%
53
169
23.9%
104
118
46.8%
90-95
23
427
5.1%
11
439
2.4%
59
391
13.1%
77
373
17.1%
96+
0
32
0.0%
0
32
0.0%
2
30
6.3%
2
30
6.3%
LDGT2
81-85
50
44
53.2%
41
53
43.6%
32
62
34.0%
73
21
77.7%
86-89
29
35
45.3%
14
50
21.9%
20
44
31.3%
42
22
65.6%
88-95
56
371
13.1%
11
416
2.6%
68
359
15.9%
104
323
24.4%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
844
2503
25.2%
521
2826
15.6%
700
2647
20.9%
1300
2047
38.8%
B-7
-------
Failure Rate By Model Year Grouping - Second IM147
Final Outpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
56
49
53.3%
26
79
24.8%
37
68
35.2%
74
. 31
70.5%
83-85
104
124
45.6%
91
137
39.9%
79
149
34.6%
157
71
68.9%
86-89
110
315
25.9%
83
342
19.5%
101
324
23.8%
181
244
42.6%
90-95
39
913
4.1%
47
905
4.9%
79
873
8.3%
127
825
13.3%
96+
0
70
0.0%
0
70
0.0%
0
70
0.0%
0
70
0.0%
LDGT1
81-85
90
170
34.6%
63
197
24.2%
65
195
25.0%
145
115
55.8%
86-89
36
186
16.2%
22
200
9.9%
39
183
17.6%
69
153
31.1%
90-95
11
439
2.4%
6
444
1.3%
29
421
6.4%
41
409
9.1%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
46
48
48.9%
41
53
43.6%
31
63
33.0%
72
22
76.6%
86-89
14
50
21.9%
12
52
18.8%
18
46
28.1%
33
31
51.6%
88-95
17
410
4.0%
3
424
0.7%
39
388
9.1%
52
375
12.2%
96+
0
18
0.0%
0
18-
0.0%
0
18
0.0%
0
18
0.0%
ALL
523
2824
15.6%
394
2953
11.8%
518
2829
15.5%
952
2395
28.4%
Failure Rate By Model Year Grouping - Third IM147
Final Cutpoints
V Type
Year
HC
CO
NOx
OVERALL
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
Fail
Pass
% Fail
LDGV
81-82
53
52
50.5%
25
80
23.8%
32
73
30.5%
67
38
63.8%
83-85
91
137
39.9%
88
140
38.6%
81
147
35.5%
156
72
68.4%
86-89
100
325
23.5%
76
349
17.9%
96
329
22.6%
171
254
40.2%
90-95
36
916
3.8%
50
902
5.3%
68
884
7.1%
112
840
11.8%
96+
0
70
0.0%
1
69
1.4%
0
70
0.0%
1
69
1.4%
LDGT1
81-85
83
177
31.9%
57
203
21.9%
61
199
23.5%
137
123
52.7%
86-89
31
191
14.0%
24
198
10.8%
36
186
16.2%
65
157
29.3%
90-95
6
444
1.3%
4
446
0.9%
29
421
6.4%
36
414
8.0%
96+
0
32
0.0%
0
32
0.0%
1
31
3.1%
1
31
3.1%
LDGT2
81-85
44
50
46.8%
41
53
43.6%
31
63
33.0%
73
21
77.7%
86-89
16
48
25.0%
11
53
17.2%
17
47
26.6%
32
32
50.0%
88-95
13
414
3.0%
4
423
0.9%
34
393
8.0%
44
383
10.3%
96+
0
18
0.0%
0
18
0.0%
0
18
0.0%
0
18
0.0%
ALL
473
2874
14.1%
381
2966
11.4%
486
2861
14.5%
895
2452
26.7%
B-8
-------
Appendix C
IM147 Regression Coefficients
-------
APPENDIX C
IM147 Composite Regression Coefficients, HC, 1981 to 1985 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT1
81-85
0.73806
0.95039
13.5112
92
LDGT1
81-85
0.52814
0.34501
2.7265
5.5143
P3
LDGT1
81-85
0.40889
0.23152
0.5547
3.2006
6.2124
P4
LDGT1
81-85
0.38717
0.22535
1.0106
3.4095
3.2694
3.5979
P5
LDGT1
81-85
0.38872
0.23976
-0.8175
2.5639
3.1218
2.0148
3.4755
P6
LDGT1
81-85
0.34978
0.25562
-2.8842
2.7609
1.6633
1.5365
2.3446
3.2626
P7
LDGT1
81-85
0.33929
0.24395
-2.5969
2.4215
2.0512
1.644
0.5818
2.2896
2.9199
P8
LDGT1
81-85
0.30707
0.21404
-0.9821
1.8007
1.4784
1.745
0.0989
0.8955
2.4381
4.82
P9
LDGT1
81-85
0.30661
0.21485
0.9303
1.7873
1.3395
1.7754
0.0281
0.6238
2.4791
4.5115
0.8987
P10
U0GT1
81-85
0.30246
0.21468
1.189
1.8114
1.1577
2.1188
0.3265
0.5235
2.6637
4.255
-0.5878
2.1284
P11
LDGT1
81-85
0.25475
0.05761
-0.2293
0.7411
1.0904
1.5866
0.0582
1.3727
1.9974
2.4699
0.822
1.4541
1.8008
P12
LDGT1
81-85
0.24251
0.05503
-0.1605
0.7753
0.9307
1.635
0.0664
1.336
2.0956
2.1337
0.2754
2.1379
1.2063
1.1106
P13
LDGT1
81-85
0.19345
0.05879
1.3617
0.2166
0.9329
1.5508
-0.1316
1.2355
2.3977
1.6874
0.0101
0.9701
0.9488
0.4982
3.2351
P14
LDGT1
81-85
0.1631
0.04219
1.279
0.1384
1.1759
1.4082
0.0948
1.1156
2.0131
1.2938
0.729
1.1342
0.4477
-0.3274
1.562
4.0745
P15 .
LDGT1
81-85
0.13928
0.03346
1.8241
0.0799
1.2164
1.2299
0.2827
0.8749
2.2275
0.6681
0.8755
0.7239
0.5659
0.1723
0.7424
1.1987
2.7671
P16
LDGT1
81-85
0.13703
0.03021
1.8786
0.023
1.2823
1.1806
0.305
0.8514
2,2773
0.5758
0.8491
0.7433
0.5775
0.2633
0.6459
1.2728
2.0331
2.1897
P17
LDGT1
81-65
0.09132
0.00475
1.4179
0.434
0.6904
0.7838
1.0571
0.616
1.5938
0.7518
0.2064
0.8618
0.5896
0.4328
0.7951
1.3837
1.3066
0.6626
2.7633
P18
LDGT1
81-85
0.07314
-0.00235
1.5086
0.5696
0.7494
0.8068
0.8363
0.5804
1.3972
0.5907
0.4057
0.6221
0.5658
0.584
0.7038
0.8587
0.988
0.3277
1.9729
1.7883
P19
LDGT1
81-85
0.06979
-0.00621
1.375
0.6281
0.6996
0.852
0.9084
0.5029
1.4414
0.5441
0.3942
0.5659
0.6012
0.6081
0.7093
0.7313
0.6296
0.2443
1.7266
1.5869
0.729
IM147 Phase 2 Regression Coefficients, HC, 1981 to 1985 Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg-
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
81-85
0.5023
0.12381
4.4447
B12
L0GT1
81-85
0.48187
0.11013
3.3244
1.9633
B13
LDGT1
81-85
0.35264
0.07757
1.7904
-1.1796
5.9021
B14
LDGT1
81-85
0.32195
0.06184
0.9978
-1.0185
3.8122
5.5161
B15
LDGT1
81-85
0.2826
0.04064
1.0928
0.0499
1.7408
0.7784
4.7049
B16
LDGT1
81-85
0.27957
0.0348
1.0816
0.2105
1.5379
0.9523
3.4207
3.6479
B17
LDGT1
81-85
0.14692
-0.01627
1.0035
0.5208
1.1766
2.2402
1.4703
0.4909
4.6919
B18
L0GT1
81-85
0.11692
-0.02022
0.9177
0.8016
0.9222
1.5324
0.8911
0.0868
3.252
2.7658
B19
LDGT1
81-85
0.11286
-0.02299
0.9799
0.8216
0.9206
1.4305
0.4043
0.0294
2.8156
2.4898
0.9698
-------
IM147 Composite Regression Coefficients, HC, 1986 to 1989 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
PI
LDGT1
86-69
0.49429
0.5423
17.2823
P2
L0GT1
86-89
0.37436
0.2461
2.2738
4.5192
P3
LDGT1
86-89
0.33313
0.22224
0.5252
2.5885
4.6599
P4
LDGT1
86-69
0.29595
0.20426
0.5943
2.3293
2.3375
3.7634
P5
LDGT1
86-69
0.2936
0.19895
0.6763
2.0377
2.3025
2.9749
1.8038
P6
LOGT1
86-69
0.28058
0.20644
-0.6191
2.2969
1.2069
2.4549
-0.1084
4.2386
P7
LDGT1
86-69
0.27458
0.19491
-0.0117
2.0167
1.3562
2.7382
-1.5023
2.6456
2.8231
P8
LDGT1
86-89
0.26273
0.17858
0.9768
1.9562
1.0027
2.2073
-1.5778
1.5977
2.003
3.2606
P9
LOGT1
86-69
0.25683
0.17484
0.544
2.0132
0.9204
2.1628
-2.2132
0.7045
2.0543
2.3838
3.4294
P10
LDGT1
86-89
0.25153
0.17214
0.4457
2.0253
0.8623
1.9864
-2.2523
0.4119
2.2232
1.6726
1.6078
3.6006
P11
LDGT1
86-69
0.19659
0.07478
2.1606
0.6298
1.0463
1.2006
-1.6683
1.7941
1.4349
0.7533
1.4174
2.3957
1.957
P12
LDGT1
66-89
0.17271
0.05609
2.4645
0.4889
0.5378
1.0905
0.6636
1.2017
1.3698
1.1634
0.8772
3.332
1.2131
1.5195
P13
LDGT1
66-89
0.13816
0.05847
2.1195
0.4529
0.7286
1.2692
-0.7285
1.427
1.284
0.6442
0.4274
2.4657
0.9117
0.2322
2.6432
P14
LDGT1
86-69
0.12661
0.04656
2.2369
0.3811
0.6514
1.4033
-0.4687
1.3427
1.1456
0.9582
0.0864
2.2192
0.7164
0.4248
1.479
2.6071
P15
LDGT1
86-89
0.11362
0.03727
2.1223
0.3411
0.6537
1.4601
-0.202
1.1529
1.232
0.6887
-0.2172
2.1843
0.7658
0.6133
0.9282
0.7457
2.1404
P16
LDGT1
86-69
0.10769
0.03525
1.9852
0.3076
0.7867
1.4374
-0.022
1.1859
1.1329
0.515
-0.1541
2.0989
0.7572
0.7681
0.784
0.499
0.7808
4.7323
P17
LDGT1
86-89
0.05356
0.01318
1.0985
0.6236
0.7621
0.7653
0.5658
1.0327
0.7036
0.4129
0.7975
1.2149
0.7601
0.6019
0.8454
0.937
0.7894
0.9132
3.009
P18
LDGT1
86-89
0.04127
0.00269
1.2483
0.5985
0.676
0.9508
0.3871
1.095
0.7732
0.6052
0.8039
0.761
0.6892
0.7308
0.7205
0.82
0.9126
0.0726
2.0163
1.4844
P19
LDGT1
86-89
0.03395
0.00397
1.0933
0.6168
0.695
0.9026
0.4507
1.0319
0.9511
0.5369
0.7983
0.6339
0.6758
0.6909
0.7183
0.7503
0.7572
-0.3355
1.6677
1.2005
0.8458
IM147 Phase 2 Regression Coefficients,
HC,
1986 to 1989 Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant.
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
;>
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
86-69
0.29834
0.1158
3.3457
B12
LDGT1
86-69
0.26979
0.09053
2.3987
1.9564
B13
LDGT1
86-89
0.20617
0.08073
1.4658
0.0492
4.1671
B14
LDGT1
86-89
0.19052
0.06515
1.1799
0.3012
2.4794
3.6519
B15
LDGT1
86-89
0.16862
0.04542
1.1829
0.6531
1.4949
0.6859
3.2681
B16
LDGT1
86-69
0.16083
0.04255
1.1483
0.9014
1.2632
0.3783
1.3376
6.662
B17
LDGT1
86-89
0.07626
0.01566
1.0548
0.7801
1.1751
1.2709
1.2392
0.9683
4.2486
B18
L0GT1
86-89
0.05999
0.00012
0.9466
0.9733
1.0114
1.038
1.3674
-0.1144
2.9584
1.987
B19
LDGT1
86-89
0.04869
0.00402
0.9313
0.9308
0.9901
0.9768
1.1033
-0.6703
2.4005
1.5697
1.23
-------
IM147 Composite Regression Coefficients, HC, 1990 to 1995 Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C16
C19
PI
LDGT1
90-95
0.37406
0.22818
17.1854
P2
LDGT1
90-95
0.23507
0.05902
-2.145
5.1262
P3
LDGT1
90-95
0.21189
0.05784
-1.1729
3.3619
3.4597
P4
LDGT1
90-95
0.20276
0.05212
-2.7245
3.2374
1.6019
3.7987
P5
LDGT1
90-95
0.19505
0.05057
-3.0392
2.7597
1.1351
2.24
3.8993
P6
LDGT1
90-95
0.18478
0.05162
-3.3501
2,6388
0.6397
1.4629
1.9249
4.9984
P7
LDGT1
90-95
0.17282
0.04206
-2.9083
2.3805
0.6434
1.9124
-0.3621
3.0178
4.1577
P8
LDGT1
90-95
0.16217
0.03882
-1.4065
1.7566
1.0327
1.3221
-0.6046
1.7837
2.9574
4.5516
P9
LDGT1
90-95
0.15997
0.03954
-1.5284
1.8223
0.7356
1.6157
-0.8447
0.9682
2.5193
3.7565
3.0158
P10
LDGT1
90-95
0.15899
0.03962
-1.3222
1.7605
0.8841
1.526
-0.7219
0.6441
2.5385
3.3964
1.3921
2.2962
P11
LDGT1
30-95
0.1242
0.00888
0.4575
0.8942
1.2216
1.078
-0.2219
0.7626
1.9585
1.2513
1.7074
1.5915
1.5461
P12
LDGT1
90-95
0.108
0.00853
0.5566
0.7543
0.7122
1.489
-0.1897
0.6244
1.9355
0.7067
0.8076
2.4795
0.9978
1.8755
P13
LDGT1
90-95
0.08846
0.01744
0.8836
0.5089
0.6095
1.7266
0.0115
0.5537
1.4386
0.4008
0.5776
1.904
0.7549
0.8219
2.3703
P14
LDGT1
90-95
0.07835
0.01772
0.858
0.474
0.6951
1.5244
0.3476
0.7542
1.4394
0.4383
0.7477
2.2305
0.5943
0.9555
0.9132
2.1616
P15
LDGT1
90-95
0.06853
0.01326
0.7115
0.5394
0.7616
1.4283
0.4159
0.4769
1.3363
0.4071
0.6623
1.647
0.6826
0.8356
0.5762
0.7447
2.2358
P16
LDGT1
90-95
0.0593
0.01319
0.6956
0.6235
0.517
1.2912
0.6027
0.5332
1.4561
0.3365
0.4783
1.6692
0.6904
0.6659
0.7576
0.7049
0.1973
5.5813
P17
LDGT1
90-95
0.04054
0.0066
0.721
0.7143
0.508
0.9341
0.7735
0.3397
1.3289
0.8072
0.4656
0.9685
0.6693
0.7801
0.6168
0.8507
0.941
2.2139
2.3062
P18
LDGT1
90-95
0.03216
0.00497
0.6809
0.6934
0.5834
0.9437
0.759
0.6708
0.9442
0.9559
0.356
0.6599
0.6671
0.8103
0.5884
1.0269
0.3629
1.3443
1.5874
1.5359
P19
LDGT1
90-95
0.02955
0.00468
0.7103
0.6964
0.6123
0.9353
0.834
0.5442
0.9229
0.8944
0.4112
0.8891
0.6698
0.7038
0.6909
0.7838
0.3123
0.6669
1.5046
1.3036
0.7468
IM147 Phase 2 Regression Coefficients,
HC,
1990 to 1995 Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
811
LDGT1
90-95
0.19826
0.0335
3.1539
B12
LDGT1
90-95
0.16235
0.0226
1.7885
3.1368
B13
LDGT1
90-95
0.12683
0.02866
1.1305
1.1678
3.5627
B14
LDGT1
90-95
0.11708
0.02986
0.9821
1.3215
2.042
2.4823
B15
LOGT1
90-95
0.09769
0.0205
1.024
1.124
1.0503
0.5065
3.5529
B16
LDGT1
90-95
0.08453
0.02072
1.0421
1.0995
1.3341
0:4695
0.6556
7.7639
B17
LDGT1
90-95
0.05719
0.01154
1.0027
0.9977
1.0045
0.9826
1.4994
2.7725
3.1834
B18
LDGT1
90-95
0.04461
0.008
0.9558
1.0711
0.8811
1.2953
0.5862
1.7403
2.1672
2.144
B19
LDGT1
90-95
0.04108
0.00785
0.9823
0.9356
1.0185
0.9731
0.499
0.8303
2.0505
1.8394
1.0152
-------
IM147 Composite Regression Coefficients, HC, 1996 and Newer Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
CIS
C19
P1
LDGT1
96+
0.05316
0.03402
18.1995
P2
LDGT1
96*
0.04307
0.02411
0.2121
2.417
P3
LDGT1
96+
0.03308
0.02193
2.6279
1.1433
2.9158
P4
LDGT1
96+
0.0321
0.02215
3.4627
1.1835
1.5689
1.9266
P5
LDGT1
96+
0.0267
0.01928
1.088
0.5294
2.0247
-3.8687
8.4714
P6
LDGT1
96+
0.02641
0.01933
0.5547
0.4887
1.8979
-4,0225
7,8031
1,8996
P7
LDGT1
98+
0.02534
0.01654
0.953
0.3268
2.2496
-3.2672
4.7342
1.2515
3.3211
P8
LDGT1
96+
0.02524
0.01597
-1.5963
0.6417
2.0178
-4.3447
3.8465
1.9882
2.8209
1.9699
P9
LDGT1
96+
0.02522
0.0165
-0.9944
0.5289
2.1342
4.3968
4.6568
1.7412
2.8435
2.4683
-2.6396
P10
LDGT1
96+
0.02529
0.01642
-1.4673
0.5347
2.3503
-5.1056
4.7009
2.0105
2.7607
2.6563
-1.6108
-0.7547
P11
LDGT1
96+
0.01755
0.00739
3.4599
0.1081
1.652
-1.1947
3.4359
0.4622
0.6672
0.794
-0.5304
-0.2246
1.7144
P12
LDGT1
96+
0.0151
0.00659
3.8459
0.4104
0.2567
1.1211
0.4848
0.8471
1.2853
0.5367
1.1122
0.0179
1.1163
1.9287
P13
LDGT1
96+
Ś 0.01241
0.00389
1.7514
0.4378
1.7539
-0.5187
-0.8979
0.5987
2.4378
1.2817
1.9041
-1.2482
0.9844
0.1784
2.6171
P14
LDGT1
96+
0.01176
0.00367
1.875
0.4721
1.4005
0.0285
-0.4781
0.4911
2.4771
0.9339
1.5111
-0.5207
0.7769
0.1877
2.2359
1.1134
P15
LDGT1
96+
0.00769
0.00241
1.0482
0.6074
0.9552
-0.3978
1.4364
1.1394
0.7747
1.4416
0.0533
-0.1206
0.7233
0.4499
1.3519
0.2003
2.2561
P16
LDGT1
96+
0.00774
0.0024
1.0165
0.6108
0.9582
-0.4078
1.4359
1.1436
0.7653
1.4501
0.0568
-0.1213
0.7218
0:4484
1.3528
0.1992
2.2428
0.0735
P17
LDGT1
96+
0.00585
0.00306
1.3687
0.5499
1.0977
0.7636
0.4468
0.6346
0.7879
0.6314
0.3617
0.6995
0.7613
0.3632
1.0963
0.709
0.7912
1.2412
3.4916
P18
LDGT1
96+
0.00432
0.00121
1.569
0.5806
0.9368
0.6842
0.9482
0.5032
0.7086
0.6239
0.5212
0.4646
0.6817
0.6338
0.8277
0.6983
0.744
0.4028
2.2628
1.7838
P19
LDGT1
96+
0.00416
0.00114
1.2715
0.6216
0.8437
1.0272
0.6835
0.4345
0.8099
0.3982
0.7078
0.75
0.6649
0.7239
0.6761
0.8112
0.5347
0.2273
2.2233
1.3046
0.5973
IM147
Phase 2 Regression Coefficients, HC, 1996 and Newer Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5-
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
96+
0.02393
0.00865
2.4226
B12
L0GT1
96+
0.02095
0.00863
1.6528
1.9763
B13
LDGT1
96+
0.01824
0.00691
1.444
0.667
2.5512
B14
LDGT1
96+
0.01653
0.00607
1.2007
0.3963
2.3614
1.9099
B15
LDGT1
96+
0.01077
0.00256
1.0854
0.6091
1.8127
0.4665
3.018
B16
LDGT1
96+
0.01076
0.00241
1.0549
0.6405
1.7899
0.4709
2.8141
0.9908
B17
LDGT1
96+
0.00794
0.00414
0.8756
0.7439
1.2805
1.0073
1.458
1.4709
3.9228
B18
LDGT1
96+
0.00581
0.00161
0.8235
0.9534
1.0124
1.0609
1.0031
0.6433
3.018
2.4016
B19
LDGT1
96+
0.00549
0.00161
0.823
1.0394
0.8747
1.0551
0.9342
0.2958
2.5227
1.8686
0.7675
-------
0
1
U\
1
M147 Composite Repression Coefficients, HC, 1981 to 1985 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
CIS
C16
C17
C18
C19
P1
LDGT2
81-85
0.92059
1.08952
14.1316
92
LDGT2
81-65
0.49433
0.24032
2.1062
6.4379
P3
LDGT2
81*85
0.47471
0.22581
0.921
5.3132
2.7842
P4
L0GT2
81-85
0.43631
0.17782
-1.7872
5.4763
0.3261
3.8518
P5
LDGT2
81-85
0.42908
0.2114
1.5166
4.7372
0.0105
2.9709
2,5245
P6
LDGT2
81-85
0.41687
0.23473
-3.8569
4.5338
-0.9009
2.7375
1.9498
3.2209
P7
L0GT2
81-85
0.41066
0.23366
-3.2863
4.2499
-1.1231
2.3552
1.2812
2.4743
2.1991
P8
LDGT2
81-85
0.3672
0.16688
-1.6179
3.5835
-0.8816
1.6341
-0.5481
3.1967
0.3748
5.7221
P9
LDGT2
81-85
0.3566
0.18117
-1.9127
3.5055
-0.9418
2.6495
-0.9804
1.9891
0.0338
4.0901
3.7721
P10
L0GT2
81-65
0.34151
0.16655
-1.8705
3.3075
-0.9637
3.3159
-1.7891
1.43
-0.0942
4.2235
1.3603
3.899
P11
LDGT2
81-85
0.2931
-0.04548
1.0746
1.8029
-1.2122
2.864
-0.6499
0.7309
0.2679
3.2607
-0.113
4.0698
1.9325
P12
L0GT2
81-85
0.2758
-0.01823
0.7863
1.8538
-0.984
2.6077
-0.7763
0.9299
0.773
2.1956
-0.3985
4.4575
0.873
2.4503
P13
LDGT2
81-85
0.21405
-0.01142
0.4126
1.5075
0.6583
2.6141
-0.5084
0.917
1.449
-0.2757
1.0903
2.7846
0.8509
0.03
3.3522
P14
LDGT2
81-85
0.17574
-0.01255
0.2502
1.2847
-0.0742
1.733
0.4325
1.3417
1.2187
0.2611
1.1197
1.7719
0.3758
0.7006
0.8312
4.2084
P15
LDGT2
81-85
0.15062
-0.02221
0.8141
1.325
-0.0587
1.4355
0.2426
1.2996
1.3459
0.694
0.5531
1.836
0.507
0.8254
-0.1245
2.1165
2.3277
P16
LDGT2
81-85
0.14738
-0.02755
1.0121
1.3194
-0.0133
1.4206
0.2735
1.0742
1.3679
0.895
0.1188
2.0783
0.5591
0.9525
-0.3444
1.9718
1.5317
2.7401
P17
LDGT2
81-85
0.11644
-0.04735
1.4605
1.1844
0.1561
0.8181
0.2509
0.6122
1.2664
1.4097
-0.3806
2.1461
0.6223
0.997
-0.2513
1.5706
1.3565
1.6797
2.2361
P18
LDGT2
81-85
0.10411
-0.05875
1.9421
1.2153
0.2284
0.6248
0.1763
0.9731
0.9762
1.0664
0.3199
1.6679
0.7965
0.7901
0.01
1.0203
0.9553
1.6582
1.5344
1.4185
P19
LDGT2
81-85
0.08606
-0.02861
1.9188
1.1185
0.3074
0.5037
0.6515
0.9559
0.5634
1.1722
0.4972
1.0016
0.7179
0.584
0.4013
1.317
0.5551
-0.9186
1.2833
0.9697
1.3485
IM147 Phase 2 Regression Coefficients, HC, 1981 to 1985 Model Year LDGT2's
Repression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
81-85
0.53604
-0.06735
4.7635
B12
LDGT2
81-85
0.50867
-0.03629
2.9814
3.9143
813
LDGT2
81-85
0.37516
-0.03573
2.2833
-1.1494
5.7782
B14
LDGT2
81-85
0.31715
-0.05881
1.5613
0.0866
1.7691
6.5887
B15
LDGT2
81-85
0.29115
-0.05873
1.6798
0.2156
0.484
3.6579
3.2814
B16
LDGT2
81-85
0.28504
0.05551
1.6963
0.4373
0.1942
3.4658
1.7593
4.9957
B17
LDGT2
81-85
0.19206
-0.07062
1.2418
0.7375
0.4405
2.7184
1.6887
1.1505
4.1633
B18
LDGT2
81-65
0.17741
-0.07525
1.3222
0.5566
0.4852
2.1651
0.9953
1.6325
3.2854
1.6783
B19
LDGT2
81-65
0.15121
0.03936
1.1587
0.4814
1.1411
2.1958
0.3955
-1.7449
2.7898
1.0076
1.9981
-------
0
1
Os
1
M147 Composite Repression Coefficients, HC, 1986 to 1987 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
0
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT2
86-67
0.76767
0.79061
14.62
P2
LDGT2
86-67
0.53753
0.28237
1.2627
6.2363
. !l
P3
LDGT2
86-67
0.36485
0.18393
-0.7945
3.7354
5.7608
P4
LDGT2
66-67
0.34875
0.16586
-2.6149
3.7629
2.8334
4.1762
P5
LDGT2
86-67
0.34617
0.17953
-3.6873
3.212
2.761
3.4285
2.4018
P6
L0GT2
86-67
0.29325
0.2Q542
-3.5626
3.7025
1.7413
0.4872
1.0679
4.3236
. "
P7
LDGT2
86-87
0.2934
0.19684
-3.5264
3.6876
1.7296
0.6143
0.2317
4.2624
0.8683
pa
LDGT2
86-87
0.26213
0.17396
-1.8292
3.3926
1.1562
0.6702
0.0208
3.449
0.3557
2.6082
P9
L0GT2
86-67
0.27163
0.1983
-1.1117
3.3764
1.0296
0.699
0.2722
2.4556
0.6353
0.6
2.1369
P10
LDGT2
66-67
0.26655
0.20643
0.6291
3.3422
0.5757
0.6074
0.6213
2.1855
0.9202
0.0129
1.3566
1.7368
P11
LDGT2
86-87
0.23982
0.0819
0.4024
2.3049
0.6701
0.0523
1.2179
2.4136
0.3361
0.0779
0.9104
1.6901
1.4014
P12
LDGT2
86-87
0.21585
0.04628
0.8765
1.6118
0.7539
0.3407
0.9193
1.8565
0.6585
0.4701
0.4832
1.6797
1.0229
2.1489
P13
LDGT2
86-67
0.17953
0.04669
1.6471
0.9949
0.665
1.0029
0.4324
1.1786
0.7455
0.9956
-0.3607
1.9373
1.0646
0.024
2.6969
P14
LDGT2
66-67
0.16923
0.01715
2.1778
0.6591
1.0222
0.7528
0.0545
1.7038
0.8256
0.5799
0.1446
1.6426
1.1863
0.1661
1.154
2.4636
P15
LDGT2
86-67
0.16212
0.02841
2.045
0.4649
0.7435
.1.0759
0.3389
1.3816
1.0395
0.9294
0.0477
1.5028
1.0456
0.1781
0.9342
1.5234
1.2274
P16
LDGT2
86-67
0.15721
0.03244
2.1199
0.4169
0.672
0.9335
0.3563
1.582
1.3979
0.0804
0.2926
1.6162
1.1907
0.4448
0.6092
0.8887
0.3523
3.6537
P17
LDGT2
86-87
0.08118
0.01059
0.6472
0.7516
0.3931
0.9394
1.5168
0.882
0.3132
0.5162
0.8422
0.3101
0.8207
0.6645
0.7572
1.2216
0.5723
1.2293
3.38
P10
LDGT2
86-67
0.06876
-0.02245
0.747
0.8866
0.3561
0.7928
1.5497
0.9638
-0.0777
0.7405
0.6591
0.4782
0.7233
0.7721
0.8208
0.8129
0.6412
0.1007
2.9199
1.2143
P19
LDGT2
86-87
0.05744
-0.02534
0.8277
0.9324
0.4772
0.6834
1.4638
0.7072
0.6021
0.5295
0.8642
0.1132
0.7642
0.7717
0.5765
0.3706
0.4851
0.0561
2.1533
0.9384
1.4322
IM147 Phase 2 Regression Coefficients, HC, 1986 to 1987 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
"i
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
86-67
0.42555
0.04924
4.1247
B12
LOGT2
86-87
0.34262
0.01535
2.3444
4.5554
B13
LDGT2
86-87
0.26401
0.06727
1.8518
0.0098
4.3001
B14
LDGT2
86-67
0.26019
0.04305
1.8614
0.0618
3.3888
1.611
B15
LDGT2
86-67
0.24666
0.06334
1.6413
0.0316
2.594
0.5597
1.9278
816
LDGT2
86-67
0.24422
0.06216
1.731
0.2275
2.4543
' -0.1869
1.1753
3.2006
B17
LDGT2
86-87
0.11773
0.01838
1.1424
0.827
1.1111
1.9999
0.6793
1.223
4.3846
B18
LDGT2
86-67
0.10087
-0.03741
1.0229
1.0633
1.2587
1.1759
0.6269
0.1785
3.8251
1.6179
B19
LDGT2
86-87
0.09638
0.03963
1.0707
1.0296
1.0107
1.2016
0.7089
-0.201
3.2092
1.6709
0.8058
-------
o
ij
I
M147
Composite Repression Coefficients, HC, 1988 to 1995 Model Year LDGT2's
Reqression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
Cfi
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT2
88-95
0.48034
0.5235
11.3071
P2
LDGT2
88-95
0.33713
0.21906
-2.96
4.9422
P3
LDGT2
88-95
0.29651
0.19834
-1.7489
2.8353
4.9489
P4
L0GT2
88-95
0.28747
0.18089
-1.9477
2.7951
2.9733
3.2647
P5
LDGT2
88-95
0.27016
0.17705
2.2098
1.8312
2.2325
1.406
5.8661
P6
LDGT2
88-95
0.25421
0.19481
-2.5612
2.2441
0.5267
0.4177
3.0039
5.8248
P7
LDGT2
88-95
0.24477
0.17308
-2.1598
1.867
0.6968
0.5785
0.8635
4.352
4.2662
P8
LDGT2
88-95
0.22604
0.14292
-1.1587
1.5845
0.5456
0.1762
0.5776
1.9469
3.7472
4.4401
P9
LDGT2
88-95
0.22226
0.14551
-1.2733
1.703
0.284
0.1479
0.9179
0.6334
3.6501
3.448
2.3522
P10
LDGT2
88-95
0.21442
0.13617
-0.7091
1.5866
0.3214
0.3077
0.4365
0.1063
3.9046
2.8806
-0.9464
5.4516
P11
LDGT2
88-95
0.1654
. 0.04125
0.6538
0.773
0.5132
0.6898
0.6807
0.4255
2.0621
1.1813
0.4048
2.5588
2.1807
P12
LDGT2
88-95
0.14608
0.03783
1.0502
0.5966
0.307
0.8399
0.3903
0.6138
1.7604
1.3774
-0.14
2.8081
1.4104
2.3284
P13
L0GT2
88-95
0.08638
0.03352
1.3017
0.5004
0.8055
0.8896
0.2765
1.4656
0.5695
0.5182
0.3106
1.2433
0.9247
-0.2803
4.1519
P14
LDGT2
88-95
0.07556
0.02491
1.2979
0.524
0.7653
0.8982
0.3475
1.5181
0.5104
0.7567
0.2981
1.4094
0.6811
0.0851
2.7444
1.9985
P15
LDGT2
88-95
0.06718
0.02356
1.1938
0.5183
0.8188
0.8118
0.379
1.4304
0.7007
0.8512
0.4013
1.2797
0.6663
0.2996
1.6768
0.8504
1.7815
P16
LDGT2
88-95
0.06237
0.02258
1.1224
0.5458
0.8938
0.7295
0.4333
1.4798
0.6019
0.7965
0.4454
1.2725
0.6721
0.4389
1.2102
0.7846
1.1152
3.1105
P17
LDGT2
88-95
0.04421
0.01056
0.9076
0.7137
0.6003
0.6133
0.8089
0.9512
0.7459
0.7414
0.5887
0.9827
0.6061
0.6414
0.972
0.8458
1.1687
1.7368
1.9774
P18
LDGT2
88-95
0.03718
0.00462
0.9099
0.6818
0.6872
0.6822
0.7561
0.9525
0.6907
0.7852
0.6052
1.0177
0.6169
0.7309
0.6146
0.7739
1.0312
0.9695
1.4043
1.372
P19
LDGT2
88-95
0.03217
0.00297
0.8825
0.7085
0.7059
0.6951
0.7723
0.8271
0.7424
0.7667
0.703
0.8462
0.6346
0.6886
0.6361
0.6291
0.7448
0.7476
1.2353
1.0716
0.9076
IM147 Phase 2 Reqression Coefficients, HC, 1988 to 1995 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
88-95
0.23751
0.05786
3.8759
B12
LDGT2
86-95
0.2067
.0.04911
2.4274
3.3621
B13
LOGT2
88-95
0.1199
0.03897
1.2412
-0.4078
5.5892
B14
LDGT2
88-95
0.10757
0.02785
1.0283
0.0365
3.9664
2.4645
B15
LDGT2
88-95
0.09668
0.02657
1.0392
0.3282
2.5887
0.8805
2.4135
B16
LDGT2
88-95
0.09084
0.02471
1.0542
0.5363
1.9743
0.7835
1.5325
4.1018
B17
LDGT2
88-95
0.06106
0.01248
0.8883
0.8408
1.4659
1.0499
1.5851
2.3095
2.7427
B16
LDGT2
88-95
0.05212
0.00425
0.9142
0.9654
1.0256
0.9252
1.4081
1.2094
2.0957
1.7973
B19
LDGT2
88-95
0.0449
0.00374
0.9299
0.9243
1.0294
0.748
0.9948
0.9323
1.7813
1.4026
1.2714
-------
o
t
oo
IM147 Composite Regression Coefficients, HC, 1996 and Newer Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
015
C16
C17
C18
C19
P1
LDGT2
96+
0.07243
0.06002
10.2
P2
LDGT2
96+
0.04564
0.0237
-7.2812
4.6718
P3
LDGT2
96+
0.04313
0.03223
-2.4032
1.3094
6.1404
P4
LDGT2
96+
0.03729
0.03292
-1.4437
2.1679
-2.2327
9.7739
P5
LDGT2
96+
0.02867
0.03311
-7.1815
1.5592
3.8084
-6.997
8.5612
P6
LOGT2
96+
0.02468
0.03246
-5.2125
2.8958
-0.7992
-3.162
2.3757
7.8985
P7
LDGT2
96+
0.02471
0.03212
-5.6903
2.6269
-0.3486
-3.879
1.8571
6.769
2.1598
P8
LDGT2
96+
0.02458
0.03221
-5.7203
2.2642
1.1436
-7.0745
3.5654
5.3631
0.5718
2.5776
P9
LDGT2
96+
0.02153
0.02755
-5.6174
3.2239
-0.1186
-6.6505
2.7679
1.8663
-1.0747
0.5327
8.5733
P10
LDGT2
96+
0.02175
0.0276
-5.5557
3.1793
-0.113
6.663
2.9877
1.5294
-1.4181
0.9223
7.6307
0.9647
P11
LDGT2
96+
0.01561
0.01512
-0.1593
1.3225
0.8979
-3.5644
1.4176
3.2321
-2.661
2.4201
3.7658
1.2056
1.069
*
P12
LDGT2
96+
0.01535
0.01155
0.1184
1.3108
0.2317
1.9686
1.0103
3.2791
-1.9369
2.0526
2.5266
1.346
0.9439
0.9442
P13
LDGT2
96+
0.0116
0.00591
0.5316
0.6014
1.3453
-1.5835
1.0716
4.5702
-0.6276
1.6068
1.2052
-0.739
0.9742
0.8137
3.1462
P14
LDGT2
96+
0.00894
0.00333
0.9294
0.6687
0.8295
0.2036
0.5645
4.0094
0.2628
1.1846
-0.1226
0.2024
0.7931
0.6864
2.3879
2.3089
P15
LDGT2
96+
0.00519
0.00298
0.5728
0.3247
2.2863
-2.5268
2.2831
1.4541
-0.658
2.6037
-0.4889
1.5224
0.8251
-0.0309
1.4354
0.4248
1.7293
P16
LDGT2
96+
0.00341
0.00091
0.6938
0.6433
0.9494
-0.2435
1.2108
1.9848
0.1134
1.3832
0.4148
0.7129
0.7604
0.2037
1.1446
0.8148
0.7145
4.0429
P17
LDGT2
96+
0.003
0.0006
0.5483
0.8035
0.6446
0.0208
1.0382
2.0883
0.1055
1.0991
0.7076
0.6796
0.7787
0.2756
1.0048
0.7528
0.7994
3.2034
1.1992
P18
LDGT2
96+
0.00229
0.00034
0.8423
0.7701
0.5405
0.6178
0.7621
1.54
0.2461
0.9463
0.785
0.7135
0.7393
0.5553
0.7347
0.5376
0.9234
2.5586
0.3526
1.0986
P19
LDGT2
96+
0.00148
-0.00015
0.5489
0.7737
0.6253
0.5369
0.7319
1.0865
0.7396
0.777
0.8586
0.6472
0.714
0.5043
0.8109
0.8951
0.552
1.6026
0.688
0.6722
0.8733
IM147
Phase 2 Regression Coefficients, HC, 1996 and Newer Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
96+
0.02667
0.01375
1.9495
B12
LDGT2
96+
0.02266
0.00456
1.3011
3.138
B13
LDGT2
96+
0.01837
0.00086
1.3623
0.2518
4.0937
B14
LDGT2
96+
0.015
-0.00298
1.1501
0.0589
2.9165
3.3772
B15
LDGT2
96+
0.01018
-0.00336
1.1443
0.9415
1.7337
0.5943
2.5841
B16
LDGT2
96+
0.00753
0.00463
1.0828
1.0417
1.2129
0.903
1.0508
6.8124
B17
LDGT2
96+
0.00666
-0.00476
1.1331
1.0977
1.0456
0.7683
1.3074
4.7654
2.3941
B18
LDGT2
96+
0.00371
0.00109
0.9732
1.1893
0.7137
- 0.4116
1.4972
2.8064
0.0005
2.4412
B19
LDGT2
96+
0.00243
-0.00121
0.9785
0.9384
0.9374
1.0534
0.8374
1.4572
0.714
1.4339
1.3705
-------
M147 Composite Regression Coefficients, HC, 1981 to 1982 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
011
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGV
81-82
0.49218
0.57705
15.1354
P2
LDGV
81-82
0.33532
0.27587
2.7958
4.3107
P3
LDGV
81-82
0.28385
0.2387
3.237
1.8573
5.3824
P4
LDGV
81-82
0.26651
0.22287
1.6681
2.0142
3.3342
3.0522
P5
LDGV
81-82
0.25039
0.21722
-0.3542
1.7343
3.0584
-0.4158
5.9309
P6
LDGV
81-82
0.23492
0.22612
-1.1998
1.7931
2.0663
-0.0624
2.4987
4.8232
P7
LDGV
81-82
0.22127
0.21033
-1.4452
1.5589
1.4953
0.1839
1.2039
3.8754
3.5496
PS
LDGV
81-82
0.21552
0.19445
0.9213
1.4891
1.3035
-0.0672
1.0624
2.967
2.9559
2.5679
P9
LDGV
81-82
0.2067
0.18041
-1.0513
1.5796
1.3012
-0.0519
1.564
-0.0917
1.2979
2.0321
6.4454
PtO
LDGV
61-82
0.18913
0.16356
-1.3284
1.5978
1.0833
0.3527
0.1824
0.3708
1.1991
2.1788
0.884
7.9922
P11
LDGV
61-82
0.15452
0.0563
0.0087
0.5393
1.7119
0.1012
0.6423
0.9748
1.0961
0.177
1.9283
6.7153
1.4557
P12
LDGV
81-82
0.13848
0.03382
0.3059
0.4616
1.2496
0.2523
0.8742
0.2381
0.6512
1.0074
1.8257
7.1
0.6496
1.7116
P13
LDGV
81-82
0.11063
0.03458
1.1018
0.2871
1.8035
0.9765
-0.5314
0.4907
1.1534
0.6704
1.3477
4.1068
0.6637
0.0626
3.0307
P14
LDGV
81-82
0.10379
0.02583
0.937
0.3096
1.715
1.0249
-0.5161
0.4784
1.2933
0.6542
1.7892
3.4911
0.4746
0.3833
2.0844
1.8075
P15
LDGV
81-82
0.09549
0.02733
1.3271
0.2571
1.7659
0.7304
0.2147
0.1823
1.1956
0.9924
1.0719
2.8131
0.4955
0.6194
1.2124
0.4439
2.0986
P16
LDGV
61-82
0.07932
0.0312
1.0118
0.4766
1.3033
0.8794
0.0472
0.402
1.6192
1.449
-0.2368
2.1792
0.4806
0.6063
1.0279
0.7352
0.5135
5.3144
P17
LDGV
81-82
0.05521
0.00814
1.2935
0.6849
0.6167
0.5737
0.616
0.8743
0.8448
1.724
0.1427
0.6903
0.6243
0.6678
0.6972
1.4033
0.6874
2.3373
2.2858
P18
LDGV
81-82
0.04359
0.00235
1.4765
0.7063
0.5683
0.9993
-0.1434
1.1276
0.8748
1.3533
-0.3586
1.3373
0.6347
0.6998
0.796
0.716
0.6921
1.817
1.4148
1.4078
P19
LDGV
81-82
0.03591
0.00148
1.2991
0.7134
0.6447
0.8407
0.2792
0.9872
0.8646
0.9518
0.1384
1.1208
0.6929
0.6275
0.8021
0.5739
0.7875
0.2237
1.2967
0.9834
1.1392
IM14
7 Phase 2 Regression Coefficients,
HC,
1981 to 1982 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5 .
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
81-82
0.32813
0.13912
3.2894
B12
LDGV
81-82
0.29163
0.08753
1.8437
3.4587
B13
LDGV
81-82
0.18796
0.06684
0.9136
-0.314
6.0599
B14
LDGV
81-82
0.1799
0.05326
0.6172
0.1991
4.6274
2.6027
B15
LDGV
81-82
0.15358
0.05277
0.6903
0.7618
2.1672
-0.1274
4.2479
B16
LDGV
81-82
0.12174
0.04997
0.8196
0.7862
1.6353
0.667
1.0053
8.5142
B17
LDGV
81-82
0.08162
0.00913
1.0095
0.7175
1.2863
1.8724
0.9087
3.3981
3.148
B16
LDGV
81-82
0.06421
0.00073
0.9515
0.8623
1.1518
0.998
0.9777
2.5859
1.9343
1.9181
B19
LDGV
81-82
0.05142
0.00017
0.9854
0.8032
1.1865
0.7374
1.1
-0.3037
1.7543
1.3113
1.6987
-------
0
1
M147 Composite Repression Coefficients, HC, 1983 to 1985 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGV
83-65
0.40712
0.48122
11.2061
P2
LDGV
83-85
0.28564
0.23584
0.3793
4.0312
P3
LDGV
83-65
0.2251
0.18194
0.1504
1.6625
5.6839
P4
LDGV
63-85
0.20949
0.1636
-0.6566
1.89
3.4063
3.0941
P5
LDGV
83-85
0.19447
0.16094
-1.0863
1.2885
2.6869
2.1291
3.7766
P6
LDGV
83-65
0.19153
0.16531
-1.4624
1.3044
2.1907
1.8106
3.1299
2.0652
P7
LDGV
83-85
0.16355
0.1542
-1.1191
1.2382
2.1956
1.3759
2.2879
1.3053
2.1756
P8
LDGV
83-65
0.1742
0.14505
-1.1029
1.1144
1.7862
0.8889
2.4267
1.2368
1.6895
1.9949
P9
LDGV
83-85
0.16499
0.14615
-0.7229
1.073
1.8176
0.7769
1.4795
0.6194
1.4189
0.9084
4.175
P10
LDGV
83-65
0.16024
0.14347
-0.7243
1.1373
1.5682
0.8427
0.8657
0.8685
1.1993
0.834
2.645
2.9155
P11
LDGV
83-85
0.12238
0.06297
0.38
0.2293
1.7054
0.1936
1.2879
1.6893
0.5856
0.5369
2.6104
1.7875
1.5441
P12
LDGV
83-85
0.10785
0.05022
0.4878
0.4094
0.8846
0.5066
1.0774
1.8031
0.7369
0.5491
1.8603
2.2819
1.0196
1.6327
P13
LDGV
83-65
0.07961
0.0331
0.7767
0.4332
0.8179
0.9806
1.0015
1.2108
0.561
0.4915
1.5593
1.6272
0.8246
0.3997
2.5228
P14
LDGV
83-85
0.068
0.02445
0.861
0.3782
0.9405
0.9692
0.9471
1.0556
0.6664
0.4702
1.5692
1.559
0.6216
0.5139
1.483
2.5758
P15
LDGV
83-85
0.05617
0.01633
0.6641
0.3847
0.9494
0.9339
0.8878
1.1718
0.6879
0.3691
1.2553
1.3016
0.7499
0.5579
0.6795
1.1332
2.2871
P16
LDGV
83-85
0.05545
0.01675
0.7233
0.393
0.9564
0.9028
0.9463
1.1024
0.7194
0.3887
1.2701
1.1977
0.7436
0.5865
0.627
1.1212
1.6334
1.417
P17
LDGV
63-85
0.03679
0.00841
0.893
0.5126
0.7936
0.6136
1.0495
0.8218
0.8047
0.6079
0.984
0.9734
0.7209
0.7276
0.5005
1.2506
1.4269
0.4297
2.417
P18
LDGV
83-65
0.02939
-0.0015
0.9063
0.6276
0.7336
0.799
0.8401
0.7909
0.6547
0.6256
1.0592
0.5782
0.7013
0.749
0.5944
0.8615
1.0441
0.6308
1.7952
1.3705
P19
LDGV
83-65
0.0284
-0.00036
0.6248
0.6354
0.7403
0.7849
0.8685
0.8016
0.6772
0.6005
1.0662
0.4753
0.7239
0.7376
0.6183
0.7249
0.9182
0.3388
1.7224
1.1415
0.5047
\MU
7 Phase 2 Regression Coefficients, HC, 1983 to 1985 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C.1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
83-65
0.2057
0.10209
2.648
B12
LDGV
83-85
0.17708
0.0755
1.6316
2.7713
Q13
LDGV
63-65
0.12398
0.04472
1.2782
0.5298
4.0757
B14
LDGV
83-65
0.10965
0.03254
0.9666
0.713
2.5673
3.6248
B15
LDGV
83-85
0.08609
0.01872
1.0596
0.7937
1.0694
1.2338
3.6814
916
LDGV
83-85
0.08613
0.02003
1.0528
0.6555
0.9553
1.2083
2.7852
2.7867
B17
LDGV
83-65
0.05674
0.00853
0.9868
0.9946
0.7356
1.5594
2.1184
0.6029
3.6309
B18
LDGV
83-85
0.04106
-0.0038
0.9461
1.0589
0.7763
1.1008
1.4725
0.8392
2.602
1.9484
B19
LDGV
6345
0.03964
-0.00206
0.9706
1.0465
0.8189
0.9395
1.277
0.41
2.4945
1.6256
0.6907
-------
o
M147 Composite Repression Coefficients, HC, 1986 to 1989 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13 .
C14
C15
C16
C17
C18
C19
P1
LDGV
S6-89
0.31427
0.27681
15.5205
P2
LDGV
66-89
0.23056
0.12883
1.6181
4.2149
P3
LDGV
86-69
0.19626
0.11002
2.0358
2.3971
3.9903
P4
LDGV
86-69
0.18598
0.09989
0.6425
2.546
1.7035
3.3548
P5
LDGV
86-69
0.17584
0.09767
0.5819
1.852
1.5615
1.555
3.8429
P6
LDGV
86-89
0.16772
0.10204
-0.8376
1.9082
1.0801
1.0965
1.9459
3.8412
P7
LDGV
86-69
0.16444
0.09594
-0.6742
1.6725
0.9995
1.3082
0.7861
3.2631
2.3581
P8
LDGV
86-89
0.15806
0.08909
0.1422
1.5556
0.6844
0.9331
1.0208
2.2001
1.5379
2.5805
P9
LDGV
86-69
0.1519
0.08696
0.1948
1.5757
0.7441
1.0414
0.6635
1.0744
1.3289
1.468
3.5915
P10
LDGV
86-69
0.14467
0.0797
-0.4894
1.5393
0.9657
0.7863
0.5873
1.2938
1.3345
1.1773
0.2863
4.8853
P11
LDGV
86-69
0.11095
0.02972
0.2469
0.721
1.1728
0.8987
0.4133
1.9423
0.6973
0.4156
0.9936
2.4194
1.6263
P12
LDGV
86-89
0.09318
0.02564
0.8246
0.6275
0.7263
0.7966
0.542
1.6116
0.6663
0.672
0.4503
2.6883
1.1012
1.7602
P13
LDGV
66-69
0.06902
0.02184
1.3196
0.4061
0.8931
0.8959
0.5829
1.0567
0.7867
0.7859
0.5864
1.9088
0.9881
0.291
2.4099
P14
LDGV
66-89
0.06241
0.01809
1.4552
0.31
1.0344
0.9038
0.6073
1.022
0.7521
0.848
0.6425
1.6254
0.813
0.5647
1.1453
2.4731
P15
LDGV
66-89
0.05136
0.01585
1.2694
0.4095
0.9607
0.9108
0.6871
0.8398
0.7304
0.8585
0.5526
1.7815
0.6924
0.6284
0.7316
0.9325
2.1051
P16
LDGV
66-69
0.04861
0.01432
1.1846
0.4732
0.8917
0.9259
0.7055
0.8742
0.7619
0.7564
0.5578
1.5489
0.7021
0.6507
0.7862
0.9721
1.2367
2.5815
P17
LDGV
66-89
0.03332
0.00699
1.0682
0.673
0.7585
0.7037
0.8742
0.6284
0.7246
0.6208
0.7319
1.1727
0.7152
0.6517
0.7253
1.0692
1.1351
1.0816
2.2316
P18
LDGV
86-69
0.02685
0.0032
1.1443
0.6139
0.8508
0.7766
0.7017
0.7447
0.6277
0.623
0.6745
1.0469
0.7238
0.6588
0.7335
0.8575
0.8976
0.668
1.5797
1.3415
P19
LDGV
86-69
0.02176
0.00312
1.0887
0.6218
0.766
0.6703
0.7888
0.6802
0.7991
0.5738
0.7502
0.9406
0.7447
0.674
0.6947
0.6605
0.8184
0.1544
1.3072
0.9142
0.9739
IM14
^7 Phase 2 Regression Coefficients, HC, 1986 to 1989 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
811
LDGV
86-89
0.17572
0.05913
2.9828
B12
LDGV
86-89
0.14079
0.04187
1.8474
2.837
B13
LDGV
86-89
0.10263
0.03308
1.4761
0.4901
3.5664
B14
LDGV
66-89
0.09666
0.02888
1.2557
0.6541
2.0611
2.9206
B15
LDGV
86-89
0.07943
0.02543
1.0638
0.9286
1.331
0.6468
3.2335
B16
LDGV
86-89
0.07243
0.02198
1.0486
0.9587
1.3534
0.8765
1.5227
4.8276
B17
LDGV
86-89
0.04689
0.01029
1.0457
0.8812
1.097
1.3615
1.4856
1.7603
3.2281
B18
LDGV
86-89
0.03789
0.00418
0.9974
0.9025
1.0724
1.0277
1.1846
1.1391
2.3711
1.6028
B19
LDGV
86-89
0.03036
0.00369
1.0225
0.6952
0.9808
0.8231
1.1
0.3076
1.8483
1.2733
1.3385
-------
0
1
tť
K)
M147 Composite Repression Coefficients, HC, 1990 to 1995 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
05 .
C6
C7
C6
C9
C10
C11
C12
C13 .
C14
C15
C16
C17
C18
C19
P1
LDGV
90-95
0.2971
0.16705
15.7948
. if
92
LOGV
90-95
0.19863
0.05505
-1.9351
4.6837
P3
LDGV
90-95
0.16628
0.03974
-1.5889
2.5546
5.6414
P4
LDGV
90-95
0.1584
0.03534
-1.9531
2.652
3.0556
3.5632
P5
LDGV
90-95
0.14861
0.0377
-1.7859
2.1399
2.3054
1.5953
3.8796
P6
LDGV
90-95
0.1428
0.04157
-2.1978
2.3528
1.3397
1.1611
1.6243
4.1795
. i
P7
LDGV
90-95
0.13017
0:03906
-1.447
1.8386
1.3743
1.2125
-0.2829
2.1347
4.9626
L
P8
LDGV
90-95
0.12213
0.03722
-0.5405
1.6099
1.2031
0.3856
-0.1326
1.4892
3.3678
3.4765
P9
LOGV
90-95
0.11733
0.03754
-0.7341
1.7237
1.0051
0.4924
-0.9573
0.6324
3.2462
2.3004
3.9191
P10
LDGV
90-95
0.11526
0.0368
-0.6148
1.6847
1.071
0.425
1.176
0.676
2.845
2.502
1.4335
2.9572
P11
LDGV
90-95
0.08741
0.00996
0.6446
0.8311
1.0111
0.5067
-0.4113
1.2288
2.2201
1.2696
1.6991
1.3219
1.5051
P12
LDGV
90-95
0.07472
0.0105
0.6979
0.7429
0.706
0.5223
-0.0564
1.3101
1.9432
1.288
1.2908
1.3897
0.6836
1.8454
P13
LDGV
90-95
0.05916
0.0139
1.3045
0.4375
0.9242
0.6499
0.0689
1.2742
1.493
1.04
1.3582
0.4913
0.8544
0.3112
2.5569
P14
LDGV
90-95
0.05362
0.01549
1.386
0.3479
1.0903
0.5466
0.252
1.3522
1.424
0.6349
1.5658
0.679
0.6813
0.3943
1.6017
2.0376
P15
LDGV
90-95
0.04337
0.01025
1.2116
0.4485
0.9328
0.7036
0.5038
1.3664
1.1003
0.7903
1.3103
0.2473
0.6803
0.7135
0.7193
0.5976
2.3136
P16
LDGV
90-95
0.04152
0.00959
1.1338
0.4558
0.9892
0.8217
0.5423
1.2542
0.9416
0.7779
1.3289
0.1879
0.6808
0.7232
0.692
0.6981
1.451
2.8503
P17
LDGV
90-95
0.0313
0.00573
0.9123
0.6301
0.5621
0.8487
0.6701
1.1772
0.8709
0.7111
1.0362
0.425
0.686
0.7574
0.6663
0.8447
1.2224
1.3549
2.4338
P18
LDGV
90-95
0.02304
0.004
1.0116
0.6089
0.6873
0.6706
0.6772
1.1438
0.7472
0.6904
0.8929
0.5431
0.6656
0.7093
0.6728
0.682
1.042
0.6819
1.3735
1.591
P19
LDGV
90-95
0.01929
0.00299
0.874
0.6682
0.6242
0.6993
0.6995
1.2226
0.7716
0.6058
0.6359
0.606
0.6915
0.724
0.6677
0.6623
0.8334
0.1215
1.1562
1.238
0.9297
IM1A
7 Phase 2 Repression Coefficients, HC, 1990 to 1995 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
90-95
0.13911
0.0203
2.9778
B12
LDGV
90-95
0.11712
0.01783
1.7491
2.9654
B13
LDGV
90-95
0.08624
0.01945
1.2881
0.3333
3.9332
B14
LDGV
90-95
0.08064
0.02126
1.0669
0.4262
2.8321
2.3928
B15
LDGV
90-95
0.06433
0.0142
1.0083
0.9588
1.2508
0.4227
3.4379
B16
LDGV
90-95
0.06126
0.01375
0.9905
0.9989
1.1536
0.6257
2.0703
4.3157
B17
LDGV
90-95
0.04442
0.00749
0.9692
0.997
1.0495
1.0206
1.8198
2.0719
3.5584
B18
LOGV
90-95
0.03245
0.00457
0.9306
0.9602
0.9645
0.8007
1.3774
1.0699
1.9911
2.2266
B19
LDGV
90-95
0.02708
0.00366
0.9309
0.9723
0.9484
0.8169
1.0622
0.3307
1.6177
1.7198
1.2796
-------
IM147 Composite Regression Coefficients, HC. 1996 and Newer Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGV
96+
0.10038
0.04026
29.2462
P2
LDGV
96+
0.07371
0.01788
5.9162
4.0658
P3
LDGV
96+
0.06844
0.01874
2.9638
1.9564
5.5186
P4
LDGV
96+
0.06132
0.01973
2.7332
1.6433
3.0543
4.4161
P5
LDGV
96+
0.06046
0.02162
2.8327
1.3424
2.314
3.7839
1.694
P6
LDGV
96+
0.05435
0.02387
1.2378
1.2522
0.457
3.4931
-1.2143
7.4758
P7
LDGV
96+
0.05447
0.0241
1.0091
1.2788
0.4429
3.4792
-1.2783
7.8471
0.2285
P8
LDGV
96+
0.05359
0.02337
0.9941
0.7729
0.7794
3.2658
-2.1201
7.6602
-0.5933
3.1409
P9
LDGV
96+
0.05279
0.02482
2.6229
0.6273
0.6458
3.3195
-1.1123
4.6464
-0.1966
0.788
4.1995
P10
LDGV
96+
0.05143
0.02476
1.0359
0.4556
1.1796
3.0156
-1.3707
3.8218
0.108
0.823
0.5809
6.8862
P11
LDGV
96+
0.03867
0.00936
5.4115
-0.1934
1.9861
0.9553
-0.8572
0.8886
0.8704
0.1062
0.029
6.6498
1.6503
P12
LDGV
96+
0.02114
0.009
3.1864
0.1934
1.3914
1.1882
0.032
-0.1178
0.7255
1.0187
0.0701
4.5106
0.6495
2.1154
P13
LDGV
96+
0.01901
0.00675
3.9129
0.3208
1.0647
0.8952
0.4296
0.6086
0.9736
0.1471
-0.7025
3.5491
0.6163
1.6571
1.3153
P14
LDGV
96+
0.01746
0.00454
4.3463
0.3067
1.0782
1.0439
0.3355
1.3018
1.0886
-0.2129
-0.7694
2.9793
0.5207
1.3494
1.2261
1.5726
P15
LDGV
96+
0.01604
0.00595
4.6859
0.2669
0.8615
0.9562
0.5265
1.8859
0.9556
-0.0501
-1.2322
2.431
0.5884
1.2496
1.3707
-0.6015
1.5225
P16
LDGV
96+
0.01579
0.00648
4.6016
0.2309
0.8473
0.9826
0.3155
2.3189
0.8167
0.4521
-1.291
2.0816
0.5345
1.4982
1.0515
-0.1967
2.1338
-2.9193
P17
LDGV
96+
0.01236
0.00351
2.9817
0.5075
0.6011
1.1216
0.3319
1.6187
0.6518
0.5096
0.0325
1.425
0.5667
1.2205
0.6224
1.0067
1.5045
-2.3679
2.3031
P18
LDGV
96+
0.00892
0.00122
2.9957
0.5568
0.7792
0.788
0.2152
1.8089
0.6335
0.4301
0.2697
0.9495
0.5389
1.1609
0.7993
0.5202
1.4602
-1.9165
1.033
1.7225
P19
LDGV
96+
0.00643
0.00092
1.469
0.8076
0.4864
0.7865
0.6192
1.1938
0.534
0.2013
0.8084
1.3759
0.6416
0.824
0.4897
0.7037
0.7724
0.0797
0.3611
1.5812
0.8653
I
M147
Phase 2 Regression Coefficients, HC, 1996 and Newer Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
96+
0.06231
0.01367
2.7246
B12
LDGV
96+
0.03471
0.01289
1.0956
3.2189
B13
LDGV
96+
0.02909
0.00972
0.7958
2.4462
2.013
B14
LDGV
96+
0.02639
0.00805
0.7353
2.2615
1,8431
1.2834
B15
LDGV
96+
0.02692
0.00988
0.8024
2.2022
1.7552
-1.4225
1.8689
B16
LDGV
96+
0.02666
0.01042
0.7459
2.4894
1.4031
-0.8725
2.5744
-3.5714
B17
LDGV
96+
0.0192
0.00442
0.8885
1.711
0.9711
0.9804
1.8655
-2.5604
3.883
B18
LDGV
96+
0.01449
0.00165
0.7232
1.6635
1.0171
0.5616
1.7084
-1.9289
2.0901
2.3916
B19
LDGV
96+
0.01043
0.00074
0.92
1.1468
0.7967
0.6882
1.2025
-0.0572
0.6503
2.302
1.1734
-------
0
1
~*
-U
1
M147 Composite Repression Coefficients, CO, 1981 to 1985 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11'
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT1
81-85
10.1754
14.5276
12.5442
P2
LDGT1
81-65
7.53286
8.22078
5.3429
3.5563
P3
LDGT1
81-85
6.63268
6.60978
2.7905
2.1525
5.3312
P4
LDGT1
81-85
6.52317
6.63083
1.5125
2.2786
3.2782
3.8618
K
P5
LDGT1
81-85
6.508
6.69361
1.2349
2.1748
2.865
3.4811
1.0432
.
P6
LDGT1
81-85
6.43582
6.71221
0.1813
2.3646
2.9243
1.5307
0.1722
2.7427
P7
LDGT1
81-85
6.19445
6.37324
0.5635
1.9693
2.7172
0.6361
-1.561
3.2843
2.9955
P8
LDGT1
81-85
5.94027
5.98064
1.2119
1.6393
2.7167
0.5356
-1.4442
1.3759
1.7876
3.6004
. .H
P9
LDGT1
81-65
5.9353
5.97383
1.0059
1.6737
2.5314
0.6549
-1.7102
0.8508
1.7967
3.4358
1.1942
P10
LDGT1
81-85
5.85738
6.02848
1.177
1.7285
2.0533
0.7067
-2.0802
1.0202
1.2427
3.6953
-1.2444
3.1472
P11
LDGT1
81-85
4.17486
2.67355
0.0251
1.2531
0.5349
0.6814
-0.4647
2.1824
0.0334
2.27
1.2267
2.243
1.0397
P12
LDGT1
81-85
3.65635
1.97698
-0.5472
1.2512
-0.2635
2.0162
-0.5276
2.225
0.3841
1.8143
0.5359
2.6593
0.7892
1.1045
P13
LDGT1
81-85
2.30258
0.92158
0.7798
0.5868
0.3445
2.2218
0.3028
1.3536
0.523
1.1833
0.7632
1.2554
0.7979
0.6638
2.4441
P14
LDGT1
81-65
1.68789
0.51912
1.7468
0.429
0.6515
1.5816
0.7182
0.8416
0.48
1.2506
1.093
1.1172
0.6993
0.5859
1.7776
1.5319
P15
L0GT1
81-85
1.13559
0.36037
1.4986
0.481
0.5203
1.5061
0.696
0.646
0.8751
0.9462
0.7302
1.215
0.6963
0.7448
0.985
0.9478
1.4391
P16
LDGT1
81-85
1.03437
0.37532
1.4149
0.5205
0.4715
1.4258
0.7753
0.6646
0.8775
0.7646
0.5543
1.4051
0.6989
0.7734
0.9487
0.9689
0.8955
1.6125
P17
LDGT1
81-85
0.96568
0.20115
1.2423
0.5923
0.2634
1.3236
0.9825
0.7034
0.7693
0.7682
0.1979
1.3783
0.7136
0.7407
0.9773
0.9693
0.8911
1.2362
1.2653
P18
L0GT1
81-85
0.75949
0.06889
1.1529
0.5685
0.4385
1.1423
0.7942
0.9943
0.7711
0.6846
0.5815
1.031
0.7017
0.7499
0.8098
0.8097
0.647
0.846
1.2494
0.9829
P19
LDGT1
81-85
0.34973
-0.06051
1.1508
0.6559
0.4115
1.1472
0.7599
0.9619
0.799
0.6744
0.5327
0.985
0.7106
0.7294
0.7378
0.6891
0.6385
0.7063
1.0232
0.6907
0.8306
IM147 Phase 2 Regression Coefficients, CO, 1981 to 1985 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C16
C19
B11
LDGT1
81-85
9.2129
9.11082
1.5105
B12
LDGT1
81-85
5.96613
4.72873
1.1535
1.7914
B13
LDGT1
81-85
3.55873
1.89325
1.0491
0.955
3.6598
B14
LDGT1
81-85
2.92833
1.4436
0.9077
0.9194
2.7661
1.929
B15
L0GT1
81-65
1.93926
0.86159
0.9453
1.0943
1.32
1.0292
2.3486
B16
LDGT1
81-85
1.76086
0.83361
0.9517
1.1151
1.2896
1.1023
1.3595
2.6847
B17
LDGT1
81-65
1.44658
0.21699
0.9768
0.9803
1.3491
1.2314
1.27
1.5265
2.526
B18
LDGT1
81-65
1.172
0.01644
0.9407
1.0108
1.0591
1.0119
1.2102
1.0895
2.58
1.3347
B19
LOGT1
81-85
0.69787
-0.12536
0.9656
0.9797
0.9936
0.8578
0.9082
0.6798
2.1293
0.9645
1.1513
-------
o
~I
1
M147 Composite Regression Coefficients, CO, 1986 to 1989 Model Year LDGTVs
Repression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGTl
86-89
7.18532
10.8086
9.9102
P2
LDGT1
86-89
5.68789
7.0459
1.8741
2.6779
P3
LDGT1
86-89
5.04319
5.5558
1.7455
1.371
5.3393
P4
LDGT1
86-69
4.88921
5.13459
1.4696
1.5629
2.9026
5.3333
P5
LDGT1
86-89
4.78915
5.08892
1.6656
1.4035
2.0371
4.1723
2.0022
P6
LDGTl
86-89
4.74332
5.11226
1.0229
1.4755
1.9866
3.8059
1.0898
2.4375
P7
LDGT1
86-89
4.65225
4.91955
1.7462
1.3554
1.7097
3.8416
-0.0053
1.9673
1.6899
P8
LDGT1
66-89
4.48557
4.61047
2.0142
1.3439
1.7523
2.7365
0.09
0.1646
0.9526
2.9249
P9
LDGT1
86-89
4.44611
4.5639
2.1382
1.316
1.6099
2.7676
-0.2399
-0.7903
0.9655
2.4766
2.079
P10
LDGT1
86-69
4.41672
4.48084
2.0022
1.3046
1.5957
2.4537
-0.4021
.-0.7631
0.7899
2.5819
0.8104
2.0187
P11
LDGT1
86-89
3.05161
2.26609
2.8544
0.9359
0.8115
1.4536
0.41
0.3277
0.4047
2.1497
-0.2782
2.0562
0.9911
P12
LDGTl
86-89
2.51613
1.51088
2.8342
0.7464
0.9245
1.9441
0.5707
-0.2141
0.4114
2.0168
0.2073
1.9253
0.7436
1.2083
P13
LDGT1
86-89
1.68355
0.8064
2.1795
0.6233
0.6001
1.3103
0.4715
0.7177
0.6871
0.7545
0.7656
1.1501
0.7503
0.8943
2.2692
P14
LDGT1
66-89
1.31741
0.68081
2.1784
Ś 0.5849
0.6912
1.3598
0.4357
0.9373
0.6956
0.8278
0.6331
1.2087
0.7057
0.7686
1.824
1.1207
P15
LDGT1
86-89
0.93911
0.47
1.8779
0.6014
0.6301
1.3592
0.4395
0.6121
0.676
0.6638
0.7005
0.9787
0.7117
0.7437
1.0888
0.8731
1.3809
P16
LDGT1
86-89
0.88497
0.41787
1.6764
0.5971
0.6595
1.3638
0.5037
0.6323
0.8426
0.6198
0.6814
1.0802
0.7147
0.7671
1.022
0.8818
0.9654
1.3933
P17
LDGT1
86-89
0.76094
0.28167
1.1884
0.633
0.6369
1.0655
0.6729
0.7423
0.7527
0.7112
0.7965
0.8042
0.7055
0.7447
0.9245
0.8662
1.048
0.6815
1.7386
P18
LDGT1
86-89
0.51513
0.11517
1.107
0.6616
0.6251
1.1599
0.7126
0.7927
0.5228
0.8379
0.7639
0.7464
0.7046
0.7204
0.7302
0.7811
1.0809
0.4756
1.39
0.869
P19
LDGT1
86-89
0.22775
0.06824
0.9804
0.6792
0.6446
1.0212
0.6645
0.7468
0.6847
0.7006
0.7495
0.7734
0.7063
0.7111
0.7207
0.7067
0.7928
0.5475
0.9726
0.7062
0.7346
IM147 Phase 2 Regression Coefficients, CO, 1986 to 1989 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
86-89
5.73419
6.44585
1.4889
B12 .
LDGT1
86-89
3.83261
3.46613
1.1555
1.6694
B13
LDGTl
66-89
2.36328
1.33398
1.0419
1.1807
3.2911
B14
LDGTl
86-89
1.90917
1.13186
0.9737
1.0107
2.6834
1.4946
B15
LDGT1
86-89
1.36531
0.75413
0.976
0.9839
1.5355
1.1464
1.973
B16
LDGT1
86-89
1.28845
0.69885
0.9841
1.0114
1.4453
1.1588
1.3867
1.9667
B17
LDGT1
86-89
1.04812
0.39116
0.9599
0.9873
1.256
1.1763
1.4736
0.6337
2.6609
816
LDGT1
86-89
0.7463
0.14799
0.9544
0.9576
1.0029
1.0365
1.5243
0.6229
2.1332
1.1447
B19
LDGT1
86-89
0.33544
0.07788
0.9581
0.9522
0.9677
0.9398
1.0857
0.7154
1.4816
0.9344
1.0357
-------
0
1
)*
ON
I
M147
Composite Regression Coefficients
o
o
CD
CO
0 to 1995 Model Year LDGTI's
. Regress
on Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
Cl1i
C12
C13
C14
C15
C16
C17
C18
C19
Pi
LDGT1
90-95
7.96708
6.47319
8.6737
P2
LDGT1
90-95
5.93561
3.60059
0.3029
2.9243
P3
LDGT1
90-95
5.49867
2.58569
0.1967
1.6676
5.73
P4
LDGT1
90-95
5.47555
2.41839
0.0472
1.683
4.5333
2.9942
i
P5
LDGT1
90-95
5.34097
2.4681
0.3008
1.3842
2.8744
1.4949
3.7142
P6
LDGT1
90-95
5.23567
2.50459
-0.4457
1.3784
2.6883
0.8341
2.2673
4.9505
P7
LDGT1
90-95
5.12065
2.31146
-0.3312
1.2837
1.873
-0.1146
0.7803
4.7671
3.2802
P8
LDGT1
90-95
4.93002
2.15394
-0.1308
1.2614
1.6587
-0.7666
0.4112
1.9844
1.3783
5.3488
P9
LDGT1
90-95
4.92436
2.15209
-0.2521
1.2934
. 1.4663
-0.6767
0.1855
1.6357
1.4433
4.9503
1.3627
P10
LDGT1
90-95
4.83877
2.07448
0.0883
1.3418
1.2572
-0.6746
-0.0356
1.4432
1.0559
4.4422
-1.7507
4.57
. ..
P11
LDGT1
90-95
3.05447
0.67888
0.7448
0.9163
0.7304
-1.1973
0.6625
1.9991
1.1363
2.2194
-0.3005
2.3958
1.1499
P12
LDGT1
90-95
2.59434
0.35922
0.8716
0.834
-0.2278
-0.1746
0.685
1.603
1.4913
1.7526
-0.5786
2.4993
0.9124
1.5835
P13
LDGT1
90-95
1.69051
0.3538
0.7846
0.7978
0.2453
0.7291
0.7109
0.6836
0.9317
0.9732
0.7569
0.3609
0.8116
0.8909
2.4633
P14
LDGT1
90-95
1.25118
0.29289
0.7084
0.7156
0.4878
0.5609
0.6826
0.8354
0.9735
1.1145
0.55
0.9761
0.7373
0.7493
1.6233
1.1532
P15
LDGT1
90-95
0.88666
0.17653
0.6832
0.7443
0.6247
0.7585
0.551
0.8411
0.8994
0.6594
0.516
0.6847
0.7688
0.6805
1.0955
0.8841
1.4675
P16
LDGT1
90-95
0.83688
0.14342
0.6546
0.7325
0.6626
0.7804
0.6
0.754
0.9659
0.5296
0.6743
0.5828
0.7741
0.6855
0.9886
0.964
1.0266
1.788
P17
LDGT1
90-95
0.75939
0.03084
0.6751
0.7626
0.4984
0.7727
0.6819
0.6512
0.8746
0.5588
0.7025
0.4405
0.774
0.6594
0.9682
0.9724
1.0041
1.3254
2.054
P18
LDGT1
90-95
0.46016
0.02847
0.6665
0.7378
0.5678
0.7356
0.81
0.7878
0.6105
0.7172
0.7969
0.5663
0.7101
0.721
0.8057
0.915
0.6348
1.2367
1.3997
1.0199
P19
LDGT1
90-95
0.16891
0.00569
0.7401
0.7243
0.6471
0.7729
0.6775
0.7704
0.7643
0.6818
0.7573
0.6839
0.7157
0.7187
0.7375
0.6868
0.717
0.6511
1.1384
0.7248
0.7498
IM147 Phase 2 Regression Coefficients, CO, 1990 to 1995 Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
90-95
5.48884
2.92303
1.7403
B12
LDGT1
90-95
3.74619
0.83254
1.3383
2.3792
B13
LDGT1
90-95
2.29939
0.47273
1.1019
1.1903
3.3532
B14
LDGT1
90-95
1.7181
0.48708
1.0162
1.02
2.3871
1.5186
B15
LDGT1
90-95
1.20254
0.23609
1.0388
0.9035
1.4703
1.2205
1.9615
B16
LDGT1
90-95
1.13572
0.20105
1.0447
0.9125
1.3168
1.3287
1.3674
2.3837
B17
LDGT1
90-95
1.04495
0.01286
1.0374
0.8678
1.2457
1.3709
1.2968
1.8468
2.5009
B18
L0GT1
90-95
0.65189
0.0106
0.9551
0.9717
1.0606
1.2615
0.8351
1.7471
1.7371
1.3845
819
L0GT1
90-95
0.219
0.00146
0.968
0.9695
0.9969
0.9364
0.9681
0.9047
1.5264
0.9898
1.012
-------
IM147 Composite Regression Coefficients, CO, 1996 and Newer Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
PI
LDGT1
96+
1.82409
1.39419
5.531
P2
LDGT1
96+
1.63165
0.95561
-5.9413
3.0869
P3
LDGT1
96+
1.57733
0.85422
-1.5783
1.5971
3.8856
P4
LDGT1
96+
1.57696
0.84862
-1.5432
1.758
1.4232
4.1058
P5
LDGT1
96+
1.55655
0.77387
-1.8497
1.5252
1.4375
1.5669
3.3072
P6
LDGT1
96+
1.48358
0.83228
-5.5406
1.6396
-0.2011
-2.3153
2.1554
10.6111
P7
LDGT1
96+
1.47504
0.7148
-5.2472
1.8872
0.9914
-0.8568
0.4965
7.7306
2.5842
. _
P8
LDGT1
96+
1.48311
0.71859
-5.1507
1.8553
-0.9277
-0.9266
0.671
8.0227
2.621
-0.2793
P9
LDGT1
96+
1.49155
0.72084
-5.0954
1.8331
-1.0008
0.6946
0.7565
7.9469
2.63
0.1935
-0.3333
P10
LDGT1
96+
1.46528
0.70223
-3.9278
1.5364
-0.9574
-0.075
1.1515
8.3687
2.439
-0.1927
-2.1271
1.4542
P11
LDGT1
96+
0.71796
0.22242
-1.2984
1.4739
0.5359
-1.1903
0.4023
2.2155
2.7505
0.2649
0.6008
1.0433
0.8928
P12
LDGT1
96+
0.54429
0.13226
1.524
1.022
0.7868
-1.432
0.1372
2.482
1.6269
-0.3682
2.2202
0.4979
0.8781
1.6265
P13
LDGT1
96+
0.43766
0.08336
1.4971
0.7999
1.0816
-1.4102
0.8701
2.4778
1.5114
-0.0111
0.8915
0.5019
0.8812
0.3603
3.2131
P14
LDGT1
96+
0.27254
0.0707
1.3293
0.9437
0.7684
-0.6826
0.5074
1.6189
1.5858
0.2076
0.9221
0.5798
0.729
0.6284
2.4591
0.7018
P15
LDGT1
96+
0.19651
0.03007
1.3733
0.8799
0.3876
0.2025
0.6389
2.4824
0.9733
0.4929
0.3408
0.8102
0.7261
0.7297
1.4475
0.6703
1.466
P16
IDGT1
96+
0.17615
0.02979
1.0484
. 0.6101
0.7323
-0.205
0.5366
2.4872
0.7994
0.4735
0.5616
0.7865
0.7275
0.7712
1.4017
0.6818
0.9833
2.2195
P17
LDGT1
96+
0.13011
0.05988
1.062
0.6811
0.3816
0.7616
0.4795
2.2042
Ś 0.9584
0.5832
0.1536
0.8447
0.7233
0.7198
1.1616
0.6987
1.0372
0.2683
1.8782
P18
LDGT1
96+
0.08591
0.02837
1.3696
0.6559
0.7295
0.2618
0.8191
1.2599
0.5828
0.7744
0.6755
0.6858
0.712
0.7269
0.7236
0.7111
0.9386
0.4881
1.7121
1.0654
P19
LOGT1
96+
0.03308
0.00956
0.6181
0.7187
0.7481
0.81
0.6377
0.8223
0.7051
0.6935
0.8481
0.6648
0.7101
0.7232
0.6944
0.7169
0.8125
0.226
1.1367
0.7471
0.7019
IM147
Phase 2 Regression Coefficients, CO, 1996 and Newer Model Year LDGT1 's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
96+
1.23604
0.96823
0.9526
B12
LDGT1
964
0.74363
0.24386
1.1933
2.1843
B13
LDGT1
96+
0.60473
0.1527
1.2056
0.5562
4.3384
B14
LDGT1
96+
0.40405
0.15643
0.9921
0.9282
3.3758
0.9673
815
LOGT1
96+
0.30462
0.06538
0.9969
1.0801
1.9431
0.914
2.0105
B16
LOGT1
96+
0.26097
0.03614
0.9954
1.08
1.8842
0.9365
1.1866
3.7436
B17
LDGT1
96+
0.2048
0.0653
0.9914
1.0331
1.5352
0.9586
1.2475
1.026
2.448
B18
LDGT1
96+
0.12661
0.03391
0.9631
0.9575
0.9469
0.9698
1.2044
0.6083
2.4102
1.4239
B19
LDGT1
96+
0.04352
0.01126
0.9604
0.9595
0.9395
0.9708
1.1345
0.2607
1.4873
1.0202
0.9453
-------
0
1
>*
OO
I
M147
Composite Repression Coefficients, CO, 1981 to 1985 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
010
r
V
C11-
012
C13
C14
C15
C16
017
018
019
P1
LDGT2
81-85
18.0705
25.7173
7.5612
P2
LDGT2
81-85
12.3795
12.9451
3.4376
3.6374
.
P3
LDGT2
81-65
11.5598
10.6148
1.0212
2.3546
5.6508
*
P4
LDGT2
81-65
11.1259
9.72659
-0.5375
2.5842
2.5322
5.5716
P5
LDGT2
81-65
11.1623
9.79162
-0.516
2.5693
2.415
5.44
0.2344
P6
LDGT2
81-85
10.5815
10.5149
-5.1729
2.5377
2.4642
3.5854
-2.439
9.5373
i!
P7
LDGT2
81-85
10.5066
9.63447
5.1298
2.3487
2.1758
3.1663
-2.7919
8.8001
1.9586
. :i
pa
LDGT2
81-85
10.4986
9.46232
-4.5397
2.2916
2.2264
2.9027
2.935
8.1749
1.56
1.182
f
P9
LDGT2
81-85
10.534
9.45853
-4.6546
2.2745
2.3176
2.9007
-2.96
8.3683
1.5537
1.2902
-0.3956
P10
LDGT2
81-85
10.5104
9.14764
-4.0569
2.3269
1.7791
3.0871
-3.3367
7.9403
1.0377
1.7694
-1.8655
2.3687
P11
LOGT2
81-85
6.70683
1.00692
-0.9625
0.8295
2.6212
0.599
-1.3472
6.2535
1.0658
1.6592
-0.2765
1.4949
1.0962
P12
LDGT2
81-85
5.84441
1.64077
0.9579
0.8376
2.0719
1.2901
-1.7228
6.1373
1.3423
1.3993
-0.7286
1.7599
0.6956
1.5499
P13
LDGT2
81-65
3.26892
0.61651
1.3225
0.6113
0.8575
1.1276
0.3105
2.4597
0.7224
0.0984
0.8037
1.0488
0.8185
0.7462
2.8417
P14
LDGT2
81-85
2.22826
0.53462
1.2795
0.6654
0.9477
1.2845
0.5586
1.7615
0.7222
0.4197
1.0753
0.8413
0.6372
0.838
1.8815
1.2908
P15
LDGT2
81-85
1.43068
0.16253
1.4763
0.6916
0.8805
1.0498
0.7614
1.593
0.6161
0.2931
0.9481
0.6408
0.759
0.6649
1.0119
0.9471
1.4604
P16
LDGT2
81-85
1.39761
0.09202
1.4913
0.6618
0.9629
0.9832
0.8115
1.4561
0.5963
0.4052
0.7374
0.7242
0.7696
0.6414
0.9815
0.9546
1.1597
1.2077
P17
LDGT2
81-85
1.30226
-0.04105
1.2639
0.683
0.8813
0.8942
0.7966
1.641
0.6092
0.4611
0.6586
0.3214
0.791
0.5936
0.982
0.9715
0.9473
1.1888
1.5709
P18
LOGT2
81-85
0.98169
-0.23459
1.3573
0.7081
0.8658
0.8401
0.6551
1.4524
0.5751
0.6074
0.7206
0.4241
0.7757
0.6063
0.7425
0.8257
0.8673
1.0036
1.3323
0.8645
P19
LDGT2
81-85
0.50955
-0.1442
1.2761
0.7138
0.6615
0.8723
0.6415
1.4442
0.6884
0.5997
0.8506
0.4704
0.7392
0.6704
0.7048
0.6711
0.7256
0.4192
1.1487
0.8223
0.7477
IM147 Phase 2 Regression Coefficients, CO, 1981 to 1985 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
08
09
C10
C11
C12
C13
C14
C15
016
C17
018
019
811
LDGT2
81-85
13.6662
10.3119
1.5327
B12
LOGT2
81-85
9.5454
6.3219
0.9685
2.6838
B13
L0GT2
81-85
4.74433
1.25606
1.0757
0.9968
4.0692
B14
LDGT2
81-85
3.50162
1.57213
0.8395
1.1681
2.9369
1.7248
B15
LDGT2
81-85
2.41832
0.59238
1.0109
0.918
1.4255
1.2686
2.1564
B16
LDGT2
81-65
2.31489
0.51543
1.0284
0.8761
1.3639
1.2777
1.5125
2.5385
B17
LOGT2
81-85
2.02742
0.13816
1.0611
0.7865
1.3133
1.3648
1.0331
2.4287
2.7411
B18
LDGT2
81-65
1.59498
-0.3698
1.0435
0.6038
0.976
1.1376
0.9439
2:1064
2.2992
1.2758
B19
LDGT2
81-65
1.13333
-0.2565
, 1
0.8755
0.9246
0.9223
0.7526
1.3895
2.0596
1.205
1.0168
-------
0
1
VO
IM147 Composite Regression Coefficients, CO, 1986 to 1987 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT2
86-87
10.9444
12.9835
9.9921
P2
LOGT2
86-87
7.55854
7.97406
5.9552
2.7507
P3
LDGT2
86-87
6.99501
6.13276
4.0152
1.8868
4.4945
P4
LDGT2
86-87
6.97238
5.95324
3.4929
1.8483
3.4182
2.3575
P5
LDGT2
86-87
6.96833
6.05169
3.0277
1.6681
2.9579
1.9105
1.5754
P6
LDGT2
86-87
6.92215
5.98054
2.189
1.8432
2.2804
1.5316
0.9365
3.0663
P7
LDGT2
86-87
6.94695
5.98146
2.2344
1.7997
2.1735
1.5149
0.7182
3.2075
0.349
P8
LDGT2
86-87
6.86254
5.92929
3.5275
1.4609
2.7997
0.5921
0.3135
i 2.0879
-0.2127
2.0942
P9
LDGT2
86-87
6.87158
5.95327
3.6626
1.5007
2.5734
0.597
0.1327
1.6367
0.0355
1.5977
1.3024
P10
LDGT2
86-87
6.8465
5.84494
3.819
1.5958
2.1327
0.6932
0.2752
1.0075
-0.5524
1.789
-0.4389
2.5615
P11
LDGT2
86-87
4,00509
1.79665
1.9731
0.449
2.1242
-1.7461
2.0304
1.3946
-0.7103
1.246
1.7513
2.4635
1.0775
P12
LDGT2
86-87
3.20981
1.45021
1.8521
0.2977
1.3666
1.06
2.0745
-0.8491
0.2216
1.3232
1.1353
1.6136
0.6802
1.8785
P13
LDGT2
86-87
2.18993
0.93828
1.6364
0.6089
1.2938
1.9515
0.3665
-0.5328
-0.0587
1.2006
0.4256
1.2367
0.6922
1.2207
2.0554
P14
LDGT2
86-87
1.37534
0.49488
1.2757
0.6254
1.0665
0.9774
0.576
0.938
0.198
1.1141
0.5629
1.3587
0.6742
0.7466
1.5811
1.3798
P15
LDGT2
86-87
0.82014
0.15633
1.0946
0.6111
0.9745
0.7593
0.7578
1.1728
0.4226
1.0261
0.3048
1.0888
0.7145
0.6425
0.9896
1.02
1.5309
P16
LDGT2
86-87
0.69514
-0.03757
1.0745
0.6025
1.0009
0.7342
0.6475
1.2043
0.645
0.9395
0.3972
0.8668
0.7284
0.6071
0.9474
1.1426
0.9553
2.2245
P17
LDGT2
86-87
0.60214
-0.0821
1.1799
0.6279
0.8864
0.6465
0.8573
0.9847
0.4956
1.0034
0.4603
0.6907
0.7232
0.6711
0.8613
1.1323
0.9966
1.4224
1.7537
P18
LDGT2
66-87
0.44333
0.03364
1.1598
0.6606
0.7253
0.5353
0.9864
1.1754
0.4065
0.9965
0.5489
0.8032
0.7002
0.7326
0.6569
1.0226
0.9838
0.6221
1.7755
0.6944
P19
LDGT2
86-87
0.18065
-0.04526
1.0637
0.7123
0.7409
0.8109
0.662
0.9665
0.652
0.7498
0.6309
0.6545
0.715
0.7245
0.7123
0.6921
0.7437
0.6817
1.153
0.6878
0.7348
IM147 Phase 2 Repression Coefficients, CO, 1986 to 1987 Model Year LDGT2's
Repression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
86-87
7.78582
5.80047
1.4386
B12
LDGT2
86-87
4.68466
3.16394
0.9308
2.6872
B13
LDGT2
86-87
3.05474
1.67487
0.9606
1.5982
2.7862
B14
L0GT2
86-87
2.02059
1.1291
0.8908
1.1326
2.3405
1.8259
B15
LDGT2
66-87
1.24999
0.47151
0.9544
0.9442
1.4229
1.3063
2.1977
B16
LDGT2
86-87
1.06317
0.16747
0.9833
0.8605
1.3578
1.508
1.2803
3.3341
B17
LDGT2
86-87
0.90457
-0.06369
0.9757
0.9354
1.2019
1.5292
1.2917
2.1953
2.4754
B18
LDGT2
86-87
0.76129
-0.02736
0.9433
1.0434
0.9962
1.3721
1.2582
1.4319
2.4588
0.8181
B19
LDGT2
86-87
0.34719
-0.0285
0.9787
0.9885
0.9663
0.8916
0.9566
0.9606
1.5579
0.947
1.0918
-------
0
1
K>
O
1
M147 Composite Repression Coefficients, CO, 1988 to 1995 Model Year LDGT2's
Reqression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT2
68-95
7.29843
8.32241
9.6375
P2
LDGT2
88-95
5.72678
5.74122
-0.9189
2.3772
P3
LD6T2
88-95
5.24439
4.67052
-1.3811
1.1541
5.5635
P4
LDGT2
88-95
5.14626
4.26813
-1.2272
1.0883
3.6162
4.9841
P5
LDGT2
88-95
5.09801
4.25023
-1.2746
0.9561
3.0791
3.914
1.7536
P6
LDGT2
88-95
5.03227
4.30487
-2.116
1.045
2.75
2.5651
1.1445
4.8242
P7
LDGT2
88-95
4.94785
3.9903
-1.6246
0.9734
2.2915
2.261
0.0539
4.0718
2.4087
P8
LDGT2
88-95
4.69866
3.89264
-1.7403
1.0233
1.931
1.7839
0.2408
2.7207
1.5017
2.1926
P9
LDGT2
88-95
4.85108
3.81187
-1.2151
1.0106
1.7096
1.9917
-0.0267
1.4417
1.4322
1.6155
3.0089
P10
L0GT2
88-95
4.82605
3.68452
-1.0168
1.0461
1.7253
1.7644
-0.2401
1.2295
1.1312
1.6015
1.909
2.3455
P11
LDGT2
88-95
3.00665
1.74476
1.0382
0.6435
2.2076
0.8361
0.297
0.8814
1.2011
-0.12
2.4158
1.0538
1.0849
P12
L0GT2
88-95
2.71202
1.40261
1.3589
0.5437
1.8138
1.0668
0.3264
0.7761
1.4578
0.0633
1.6352
0.6125
0.9029
1.3647
P13
L0GT2
88-95
1.9294
0.60523
1.2096
0.6003
1.3401
0.909
0.5644
1.2411
0.8508
0.3162
0.575
0.1662
0.8851
0.7097
2.7113
P14
L0GT2
88-95
1.26478
0.42753
1.2507
0.6101
1.151
0.906
0.608
1.707
0.8161
0.593
0.3194
0.7475
0.7497
0.8072
1.7136
1.2385
P15
L0GT2
88-95
0.8779
0.26841
0.9697
0.6415
1.0695
0.7999
0.5865
1.284
0.8121
0.5778
0.5467
0.957
0.7335
0.8178
1.057
0.6384
1.2288
P16
LDGT2
68-95
0.79251
0.18425
1.0108
0.6516
1.0689
0.8061
0.5356
1.3495
0.8301
0.6371
0.4534
0.9417
0.7116
0.8113
1.0192
0.8943
0.9962
1.2531
P17
LDGT2
88-95
0.69731
0.07985
0.6509
0.7135
0.6792
0.7703
0.5294
1.3068
0.6755
0.6246
0.4121
0.958
0.7118
0.8086
0.9187
0.9121
0.9794
0.751
2.056
P18
LDGT2
88-95
0.52944
0.00314
0.7855
0.7161
0.8485
0.6381
0.64
1.2114
0.5735
0.7301
0.4666
0.6616
0.7036
0.7599
0.6897
0.6686
0.9015
0.6581
1.2682
1.0027
P19
LDGT2
86-95
0.18228
0.02029
1.0068
0.7103
0.7155
0.8232
0.7148
0.8349
0.6247
0.8302
0.6614
0.7366
0.7094
0.724
0.6921
0.7083
0.7232
0.6986
1.0875
0.7309
0.7311
1M14
7 Phase 2 Regression Coefficients,
o
o
1988 to 1995 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
05
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
CIS
B11
LDGT2
88-95
5.55431
4.86126
1.4596
B12
LDGT2
66-95
3.81315
2.54349
1.2451
2.0945
B13
LDGT2
88-95
2.63707
0.83295
1.1846
0.9524
3.6628
B14
LDGT2
88-95
1.77124
0.72633
1.0256
1.1004
2.4663
1.6377
B15
LDGT2
86-95
1.22283
0.50627
1.0014
1.1189
1.5439
1.0872
1.7128
816
LDGT2
88-95
1.11757
0.38075
0.9754
1.1094
1.4934
1.1553
1.4158
1.6429
B17
LOGT2
86-95
0.95396
0.13503
0.9673
1.1001
1.2482
1.2114
1.3643
0.9484
2.9337
B18
LDGT2
88-95
0.72512
-0.00576
0.9541
1.0331
0.9195
1.156
1.2511
1.1276
1.7748
1.3609
B19
LDGT2
88-95
0.28372
0.00704
0.9656
0.9834
0.9474
0.9412
1.0187
0.8827
1.65
0.9988
0.9646
-------
0
1
to
IM147 Composite Regression Coefficients, CO, 1996 and Newer Model Year LDGT2's
Reqression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C16
C19
P1
LDGT2
96+
1.3088
1.42622
17.1995
P2
LDGT2
96+
1.1479
1.00342
-1.078
3.0953
P3
LDGT2
96+
0.98949
0.98019
-2.6776
0.7119
4.3482
P4
LOGT2
96+
0.82952
0.97186
-4.3164
0.4586
-0.0693
9.4052
P5
LDGT2
96+
0.82272
0.92541
-6.1685
0.6234
0.0838
7.9052
1.1172
P6
LDGT2
96+
0.60854
0.93687
-2.5537
0.4345
0.9927
5.0074
0.2744
5.1496
P7
L0GT2
96+
0.78612
0.93219
0.8299
0.1902
0.867
2.2463
-0.6382
7.4747
1.8026
P8
L0GT2
96+
0.7922
0.9145
-0.5421
0.3098
1.1613
-0.1172
-0.5476
8.6914
1.4725
1.1768
P9
LDGT2
96+
0.79439
0.90764
0.4323
0.2804
1.57
1.9801
0.584
8.2734
1.281
1.4763
1.8463
P10
LDGT2
96+
0.80176
0.89411
0.6291
0.4422
1.1925
-2.1872
-0.6255
7.6689
1.3637
1.6415
0.9728
1.2038
P11
LDGT2
96+
0.45829
0.43587
2.7438
0.9615
1.1761
-4.6902
-0.051
7.5376
0.5427
2.6889
0.9224
1.6589
0.7053
P12
LOGT2
96+
0.43361
0.27262
1.6129
0.8974
1.1808
-3.2179
0.1827
6.7307
0.6851
2.4296
1.0559
0.404
0.7027
1.0882
P13
LDGT2
96+
0.35545
0.13528
0.1926
1.1224
-0.5669
2.979
0.8935
3.5682
0.4381
0.8157
0.0928
0.4589
0.7161
0.668
1.9197
P14
LDGT2
96+
0.30207
0.07328
2.5856
1.3224
-0.4318
1.9157
1.1728
1.7862
0.501
0.7493
0.097
1.2673
0.6916
0.617
1.1886
1.8756
P15
LDGT2
96+
0.11959
0.01946
0.6675
0.9676
-0.1805
1.423
0.6812
2.4687
0.9513
0.6769
0.0759
1.097
0.7126
0.6574
1.1486
0.3594
1.6893
P16
LDGT2
96+
0.09136
0.01277
-0.0995
0.9293
-0.1463
2.0115
0.7771
2.1495
0.9056
0.4462
0.4595
0.6343
0.7138
0.6364
1.0287
0.625
1.3436
1.1992
P17
LDGT2
96+
0.08571
0.00815
0.3122
0.914
0.0235
1.495
0.6462
2.28
0.6668
0.6935
0.6686
0.5081
0.7115
0.6609
0.9459
0.5995
1.3506
0.864
1.1952
P18
L0GT2
96+
0.07115
-0.00155
0.1596
0.9576
0.1434
1.207
0.6318
1.7452
0.5467
0.8902
0.2782
0.9571
0.7097
0.6328
0.8136
0.7233
1.1938
0.7903
0.6252
0.8195
P19
LDGT2
96+
0.02689
0.00249
-0.2617
0.9243
0.5171
0.6083
0.8825
1.183
0.5554
0.926
0.5734
0.7532
0.715
0.657
0.7458
0.7682
0.6645
0.7015
0.8289
0.6613
0.7521
IM147
Phase 2 Repression Coefficients, CO, 1996 and Newer Model Year LDGT2's
Reqression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C16
C19
B11
LDGT2
96+
1.20408
1.12226
0.7609
B12
LDGT2
96+
0.62558
0.40305
0.9685
2.1567
B13
LDGT2
96+
0.54186
0.23919
0.9716
1.5278
2.1489
B14
LDGT2
96+
0.44605
0.15037
0.9303
1.3986
1.0695
2.924
B15
LDGT2
96+
0.24307
0.03103
0.9646
1.2679
1.1889
0.9274
2.4187
B16-
LDGT2
96+
0.21899
0.02112
0.9694
1.1748
0.9741
1.2313
1.9455
1 1.756
B17
LDGT2
96+
0.19501
0.00313
0.9536
1.2274
0.7628
1.1104
1.9911
0.8779
2.8529
B18
LDGT2
96+
0.14188
-0.00797
0.9522
1.0935
0.6831
1.2308
1.578
1.0698
0.7732
1.802
B19
LDGT2
96+
0.06636
0.01816
0.9663
0.9491
0.8971
1.0203
0.7701
0.8901
1.1191
1.0268
1.2815
-------
0
1
to
to
IM147 Composite Repression Coefficients, CO, 1981 to 1982 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
V
ii
C11
C12
C13
C14
C15
C16
C17
C16
C19
P1
Lb6V
81-82
7.99839
11.8892
10.2939
P2
LDGV
81-82
6.58035
8.47137
3.9709
2.1482
. .I
P3
LDGV
81-82
5.95523
7.35499
2.3706
1.1885
4.3273
'
. ;Ł
P4
LDGV
81-82
5.84107
6.94384
0.9352
1.4
2.5147
4.4184
. 'ir
P5
LDGV
81-82
5.67429
6.92733
-0.9051
1.3052
1.0082
3.8467
3.2208
P6
LDGV
81-82
5.60045
7.2234
-1.4742
1.4463
0.1894
3.6997
2.3343
2.4304
. :p
P7
LDGV
81-82
5.42769
6.90375
-1.2377
1.0873
0.2076
2.5448
1.0667
3.0396
2.007
. .'I
P8
LDGV
81-82
5.17703
6.82475
-1.395
0.841
-0.567
2.793
1.0784
2.1889
0.6894
3.583
'L
P9
LDGV
81-82
5.11543
6.5602
-1.8042
0.9128
-0.049
1.6424
0.3327
1.1075
0.5581
3.277
3.6636
. 1!
P10
LDGV
81-82
5.09166
6.36664
-1.8837
0.9302
-0.4412
1.7282
0.1522
1.3964
0.1539
3.943
1.4796
2.6588
.
P11
LDGV
81-82
3.36176
2.42133
-1.7143
0.9431
0.3844
-1.067
0.3333
1.7489
0.1534
1.8436
3.6102
2.5886
0.995
P12
LDGV
81-82
2.45131
1.51191
-0.6053
0.8264
0.3739
0.1846
0.3965
1.281
0.4926
1.5431
2.6906
2.8681
0.6782
1.3748
P13
LDGV
81-82
1.64222
0.81857
0.547
0.7456
0.5366
1.486
-0.1283
1.4627
0.5971
1.0752
1.2221
2.1211
0.6873
0.8834
2.0965
P14
LDGV
81-82
1.30098
0.6002
0.761
0.7639
0.4632
0.8416
-0.3448
2.0537
0.7512
1.0435
0.8919
2.1412
0.6644
0.6957
1.6479
1.2133
P15
LOGV
81-82
0.99586
0.48274
1.4122
0.6691
0.5819
1.4446
0.1117
1.6713
0.9146
0.8677
0.8203
0.9725
0.6913
0.6885
1.0384
0.6235
1.3336
P16
LDGV
81-82
0.92334
0.37973
1.5086
0.6203
0.6059
1.5811
0.0371
1.4292
1.0143
0.8284
0.7229
0.8179
0.6963
0.6941
0.9588
0.9625
0.9091
1.8136
P17
LDGV
81-82
0.64728
0.19446
1.3677
0.6505
0.544
1.2293
0.3115
1.4563
0.9885
0.8554
0.4817
0.4999
0.7063
0.6726
0.9646
0.9625
0.9546
1.1516
2.0177
P18
LDGV
81-82
0.52289
0.04744
1.1672
0.7297
0.3484
1.2653
0.3087
1.1707
0.6643
0.8253
0.7848
Ś0.8903
0.6988
0.7311
0.7096
0.8592
0.8362
1.0156
1.7867
0.8424
P19
LDGV
81-82
0.31963
-0.00493
1.0652
0.7221
0.5675
1.1578
0.3797
0.9834
0.7238
0.6876
0.994
0.7578
0.7005
0.7001
0.7374
0.7792
0.7058
0.244
1.232
0.646
0.9011
IM14
7 Phase 2 Regression Coefficients, CO, 1981 to 1982 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C0
C7
C8
C9
C10
i;
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
81-82
7.87561
6.04896
1.2099
B12
LDGV
81-82
4.28681
4.12849
0.9421
1.067
B13
LDGV
81-82
2.91132
1.85708
0.9717
1.0199
3.5026
B14
LDGV
81-82
2.35783
1.43426
0.9278
0.7999
2.9182
1.5911
B15
LDGV
81-82
1.81322
0.97781
0.9942
0.8184
1.529
0.995
2.1475
B16
LDGV
81-82
1.64673
0.74808
1.0076
0.8215
1.3625
1.3316
1.1639
3.4285
B17
LDGV
81-82
1.35105
0.12268
1.0045
0.658
1.3822
1.29
1.2026
1.7464
4.1093
B18
LDGV
81-82
0.83894
0.09966
0.9679
0.9648
0.9202
1.1549
1.0845
1.5149
3.5169
1.2191
B19
LDGV
81-82
0.57427
-0.10409
0.9662
0.9148
0.9743
1.0447
0.9099
0.3112
2.66
0.9357
1.318
-------
0
1
K)
IM147 Composite Repression Coefficients, CO, 1983 to 1985 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
cie
C19
Pi
LDGV
83-85
5.65667
8.71977
10.0773
P2
LDGV
83-65
4.39498
6.05704
2.7834
2.4038
P3
LDGV
83-85
3.79704
5.05452
2.132
1.2291
4.6404
P4
LDGV
83-85
3.62371
4.67806
0.5952
1.3042
2.354
5.4862
P5
LDGV
83-85
3.59464
4.60345
0.0971
1.2134
2.0391
4.9125
1.269
P6
LDGV
83-85
3.59268
4.59476
0.3543
1.2106
2.0275
4.62
1.1146
1.0934
P7
LDGV
83-85
3.44031
4.32493
-0.5781
1.2446
0.8896
4.3893
0.6174
1.6598
1.5993
P6
LDGV
83-85
3.36343
4.2564
0.0459
1.093
0.9849
3.5699
0.5789
0.9952
1.2147
1.5455
P9
LDGV
83-85
3.33029
4.1852
-0.1659
1.1245
0.9751
3.3156
0.1104
0.6868
1.2398
1.3047
2.0264
P10
LDGV
83-85
3.29987
4.12599
0.0405
1.1476
0.8196
3.3153
-0.2824
0.8442
1.1246
1.22
1.1895
1.552
P11
LDGV
83-85
2.20291
2.25511
0.3918
0.6788
0.9189
1.6168
0.932
1.9937
0.5168
0.7853
1.9632
1.1107
0.7316
P12
LDGV
83-85
1.87972
1.58471
0.5439
0.7415
0.6536
1.6537
0.727
1.925
0.7033
0.7588
1.3654
1.0997
0.6415
1.1026
P13
LDGV
83-85
1.28242
0.91668
0.7111
0.7088
0.7627
1.8163
0.2956
1.6032
0.6711
0.6173
1.5002
0.6472
0.6692
0.8018
1.9269
P14
LDGV
83-65
1.03127
0.58797
0.6489
0.7423
0.3578
1.6308
0.4662
1.5147
0.8643
0.6344
1.5828
0.5435
0.6488
0.7632
1.5457
1.1794
P15
LDGV
63-85
0.69673
0.34058
0.7164
0.7324
0.354
1.9052
0.5776
1.1365
0.8592
0.5437
1.0426
0.8335
0.6765
0.708
0.9536
0.8489
1.4982
P16
LDGV
83-85
0.6836
0.3288
0.6397
0.7348
0.4045
1.8616
0.5991
1.1119
0.8675
0.5275
0.9881
0.8608
0.6754
0.7075
0.9317
0.628
1.269
0.931
P17
LDGV
83-65
0.61016
0.1403
0.7783
0.7181
0.5497
1.5209
0.4902
1.1225
0.8292
0.592
0.9313
0.8148
0.6905
0.7307
0.9393
0.8363
1.1503
0.6464
1.6992
P18
LDGV
83-85
0.35809
0.01542
1.0268
0.6988
0.5932
1.2395
0.5512
1.1677
0.6596
0.7122
0.6022
0.6293
0.6933
0.7668
0.6573
0.8294
1.0268
0.6176
1.4839
0.9529
P19
LDGV
83-85
0.25947
-0.01001
1.0127
0.7023
0.6463
1.1344
0.5259
1.0927
0.704
0.6265
0.7924
0.6521
0.7068
0.755
0.7028
0.7409
0.743
0.4739
1.3579
0.7294
0.7327
IM14
Y7 Phase 2 Regression Coefficients, CO, 1983 to 1985 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C18
C17
C18
C19
B11
LDGV
83-85
4.69212
5.7524
1.0243
B12
LDGV
83-85
2.83315
3.12116
0.9251
1.6564
013
LDGV
83-85
1.86354
1.56942
0.9295
1.1338
2.8182
B14
LDGV
83-85
1.55825
1.15742
0.9063
1.0736
2.325
1.5629
B15
LDGV
83-85
1.07233
0.69467
0.943
0.9688
1.4054
1.0868
2.1751
B16
LDGV
83-85
1.05389
0.68187
0.9437
0.9647
1.3864
1.0625
1.8313
'1.3471
B17
LDGV
83-85
0.89995
0.27297
0.9612
0.9964
1.3316
1.1059
1.5922
0.8992
2.8516
B18
LDGV
83-85
0.54856
0.0412
0.9415
1.0459
0.8624
1.115
1.3817
1.1042
2.1846
1.3008
B19
LDGV
83-85
0.4078
-0.005
0.9594
1.0223
0.9206
0.9952
0.9616
0.8242
1.9266
0.9809
1.0636
-------
o
t
Ni
4^
IM147 Composite Repression Coefficients, CO, 1986 to 1989 Model Year LDGV's
Regression Coeffiq
tients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
06
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGV
86-89
4.67926
5.71778
10.6405
r
P2
LDGV
86-89
3.51709
3.81768
0.1364
2.695
P3
LDGV
86-89
3.15842
3.11069
1.0124
1.6189
3.9459
. f
P4
LDGV
86-89
3.14894
3.04972
0.8613
1.649
3.2671
1.3569
. '
P5
LDGV
86-89
3.10204
3.03267
0.7486
1.4903
2.3894
0.6299
1.8322
S
P6
LDGV
86-89
3.09685
3.06465
0.6252
1.4634
2.3373
0.5195
1.5731
0.9056
P7
LDGV
86-89
3.02573
2.66107
1.1098
1.2652
1.9531
0.4921
0.7359
0.8227
2.1087
P8
LDGV
86-89
2.94297
2.66329
0.7801
1.2178
1.8801
0.6138
0.7593
-0.8902
1.5027
2.0203
. <
P9
LDGV
66-89
2.67652
2.54621
<0.0543
1.2924
1.523
1.0828
0.2369
-1.106
1.5137
1.3545
3.0417
P10
LDGV
86-89
2.74734
2.37808
-0.6515
1.4027
1.4393
1.1934
-0.1403
-0.5027
1.2073
0.9038
-0.333
4.4623
P11
LDGV
66-89
1.60566
1.04515
0.0874
0.8282
1.4645
1.1894
0.8248
-0.2838
0.6291
1.0042
1.4625
2.0759
0.8639
P12
LDGV
86-89
1.49356
0.71931
0.2804
0.7493
0.9793
1.513
0.8296
0.0469
0.689
1.0787
0.769
1.9598
0.7434
1.2423
P13
LDGV
66-89
1.06067
0.36461
0.8207
0.6529
0.9226
1.272
0.8321
0.1223
0.6425
0.6893
0.7679
1.6076
0.7379
0.7345
2.1093
P14
LDGV
86-89
0.82231
0.26307
0.8376
0.5953
0.933
1.0658
0.6038
0.4366
0.672
0.8766
0.8378
1.4217
0.7134
0.7172
1.4718
1.4518
P15
LDGV
86-89
0.57466
0.17901
0.6519
0.6644
0.7016
1.0893
0.6865
0.5537
0.8363
0.7644
0.7334
1.1584
0.7057
0.7004
0.9553
0.9039
1.5623
P16
LDGV
86-89
0.54701
0.18097
0.946
0.6657
0.6984
0.9951
0.7165
0.6267
0.8323
0.7529
0.6652
1.0979
0.7051
0.7127
0.9675
0.8974
1.1743
1.3482
P17
LDGV
86-89
0.49266
0.13124
0.9156
0.6933
0.6436
0.674
0.7016
0.686
0.7814
0.7572
0.5724
0.9245
0.7053
0.7153
0.9399
0.6866
1.0881
0.9489
1.6484
P18
LOGV
66-89
0.29132
0.05317
0.9002
0.685
0.7018
0.9872
0.6206
0.7756
0.6789
0.7453
0.7249
0.7506
0.7065
0.7322
0.7864
0.697
0.6566
1.1084
1.3762
0.8805
P19
LDGV
86-89
0.1491
0.02273
0.8668
0.6976
0.726
0.8665
0.6806
0.7684
0.7148
0.6669
0.6743
0.7484
0.7087
0.7177
0.7096
0.7249
0.7202
0.6852
0.9434
0.7022
0.8271
MU
17 Phase 2 Regression Coefficients, CO, 1986 to 1989 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
86-89
4.18172
3.9913
1.1684
B12
LDGV
86-89
2.21032
1.63478
1.0475
1.964
813
LDGV
86-89
1.5326
0.82575
1.0148
1.0859
3.0946
B14
LDGV
86-89
1.21271
0.6268
0.9705
1.0461
2.17
2.0218
B15
LDGV
86-89
0.82446
0.40601
0.9653
0.9766
1.3403
1.2022
2.2818
B16
LDGV
86-89
0.77323
0.3766
0.9636
0.9888
1.3495
1.1956
1.6328
2.1301
B17
LDGV
86-89
0.67252
0.23667
0.9624
0.9661
1.2888
1.1945
1.4722
1.3414
2.4799
B18
LDGV
86-89
0.40383
0.10653
0.9566
0.984
1.0723
0.9316
1.1674
1.5464
1.9865
1.1802
B19
LDGV
86-69
0.20724
0.05209
0.9611
0.965
0.9597
0.9776
0.9801
0.9182
1.3409
0.9424
1.1303
-------
o
to
IM147 Composite Regression Coefficients, CO, 1990 to 1995 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
ca
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGV
90-95
5.5829
4.52493
10.561
P2
LDGV
90-95
4.19016
2.55934
-2.4176
3.5176
P3
LDGV
90-95
3.59909
1.7089
-2.0167
1.8381
5.8658
P4
LDGV
90-95
3.52751
1.57279
-2.1878
1.7555
4.1493
4.0887
P5
LDGV
90-95
3.35967
1.52742
-1.8847
1.3658
3.0822
2.7595
3.0615
P6
LDGV
90-95
3.30911
1.53896
-2.1747
1.433
2.5763
2.1516
2.309
3.2307
P7
LDGV
90-95
3.20465
1.37753
-1.7351
1.2877
2.4869
1.7618
0.6153
2.8007
3.0019
P8
LDGV
90-95
3.08075
1.25584
-0.8703
1.3157
1.5808
1.0639
0.6544
1.6621
1.8801
3.4334
P9
LDGV
90-95
3.03308
1.28742
-0.0409
1.3472
1.2629
1.2814
0.1146
0.9574
2.0383
2.5679
2.6687
P10
LDGV
90-95
2.99284
1.29175
-0.6667
1.2906
1.1937
1.4946
-0.2007
0.8633
1.7797
2.4995
0.7394
2.5847
P11
LDGV
90-95
1.79334
0.39573
0.8907
0.875
0.9738
0.6525
0.6724
1.4441
0.9505
1.8704
1.0336
1.2774
0.946
P12
LDGV
90-95
1.48201
0.27031
0.8772
0.7905
0.9191
0.8699
0.6683
0.8286
0.9734
1.6847
0.9938
1.2059
0.7634
1.2265
P13
LDGV
90-95
1.22124
0.18956
0.9352
0.7049
1.1123
0.7943
0.6333
0.6174
0.819
1.3469
0.6633
0.9046
0.7518
0.8615
1.6924
P14
LDGV
90-95
0.96232
0.19909
1.2249
0.6231
0.6871
0.5243
0.6879
1.0339
0.9491
1.0285
1.2323
0.735
0.7153
0.7873
1.2566
1.1834
P15
LDGV
90-95
0.67059
0.09713
1.1865
0.6235
0.8229
0.6645
0.7023
0.9149
0.9983
0.9575
1.0487
0.5318
0.7254
0.7636
0.8921
0.7821
1.3034
P16
LDGV
90-95
0.61862
0.0845
1.1157
0.6363
0.8291
0.672
0.6393
0.9358
0.9492
0.8718
1.0229
0.6422
0.7279
0.7605
0.8704
0.8116
0.9084
1.6959
P17
LDGV
90-95
0.56789
0.04406
1.1034
0.6685
0.6934
0.7551
0.6966
0.8027
0.9189
0.7923
0.9078
0.6476
0.7299
0.7393
0.8716
0.8032
0.8822
1.1592
1.9586
P18
LDGV
90-95
0.38985
0.01218
1.0107
0.6685
0.7981
0.7109
0.6672
0.9253
0.6833
0.6064
0.9037
0.6447
0.7163
0.726
0.7487
0.7691
0.7468
1.0451
1.2236
0.9766
P19
LDGV
90-95
0.13949
0.01779
0.7784
0.7029
0.7108
0.7886
0.6798
0.8583
0.733
0.6745
0.7331
0.685
0.7139
0.7148
0.711
0.7002
0.7152
0.6362
0.9674
0.7322
0.7843
IM14
V7 Phase 2 Repression Coefficients, CO, 1990 to 1995 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
90-95
4.103
2.35754
1.4957
B12
LDGV
90-95
2.32229
0.91944
1.1523
2.0251
B13
LDGV
90-95
1.76794
0.5229
1.0723
. 1.2322
2.7642
B14
LDGV
90-95
1.40264
0.44364
0.9979
1.1201
2.0763
1.6438
B15
LDGV
90-95
0.97281
0.25022
1.0003
1.0639
1.4071
1.0512
1.8686
B16
LDGV
90-95
0.88873
0.20771
1.0023
1.0561
1.3449
1.0929
1.2656
2.5533
B17
LDGV
90-95
0.79003
0.10796
0.9981
1.009
1.2809
1.0795
1.1947
1.6393
3.0349
B18
LDGV
90-95
0.54771
0.04247
0.9734
0.9933
1.0674
1.0346
1.0249
1.4049
2.0212
1.3189
B19
LDGV
90-95
0.19036
0.03127
0.9671
0.9711
0.9708
0.9442
0.9644
0.8616
1.3588
0.9845
1.0769
-------
0
1
ro
ON
IM147 Composite Regression Coefficients, CO, 1996 and Newer Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
Pi
LDGV
96+
4.2502
1.81884
53.0314
. u
P2
LDGV
96+
2.79375
0.66889
-6.9133
5.3642
. i
P3
LDGV
96+
Ś2.31951
0.49886
-22.4698
2.9121
7.8787
P4
LDGV
96+
2.30066
0.48852
-20.2357
2.6582
5.7046
6.7489
. i
P5
LDGV
96+
2.29856
0.4874
-19.0504
2.2906
4.9636
7.0228
1.4214
. ;
P6
LDGV
96+
2.17485
0.36586
-18.1249
2.1863
5.4292
5.3931
-0.6432
6.8037
P7
LDGV
96+
2.17872
0.34724
-17.6869
2.1129
5.4618
5.4355
0.6956
6.7245
0.3983
P8
LDGV
96+
2.17729
0.32839
-15.8387
1.9759
5.2235
5.2769
-1.1683
6.4131
0.2201
1.4309
P9
LDGV
96+
2.16612
0.4024
-11.8473
1.4566
4.5337
6.4172
-1.2079
5.4459
0.3128
0.4287
3.2042
P10
LDGV
96+
2.17081
0.40258
-11.4985
1.3861
4.5521
6.4182
-1.1694
5.5108
0.2871
0.3698
2.6985
0.5987
P11
LOGV
96+
1.3457
0.04672
-1.5411
1.3212
1.6098
6.5926
-0.785
3.2891
0.7158
-0.2278
1.3044
1.2284
1.1171
P12
LDGV
96+
0.94601
0.03677
0.1569
1.1325
0.7478
7.1762
-2.7331
2.8294
0.8581
0.5253
-0.3667
2.4476
0.8375
1.6915
P13
LDGV
96+
0.83127
0.04161
6.5433
0.7725
0.7897
2.3429
-1.2999
2.5769
0.8026
1.0385
0.6814
0.7591
0.8719
1.3576
1.4743
P14
LDGV
96+
0.36237
0.07318
-1.7846
0.7477
1.0289
0.323
0.8116
2.2639
0.6876
0.4145
1.1104
0.4144
0.6973
0.7747
0.877
1.8748
P15
LDGV
96+
0.29315
0.05053
0.3484
0.5009
0.8684
2.0337
0.8189
2.1406
0.8321
0.2796
-0.0113
0.3684
0.7271
0.7433
1.1116
1.0359
0.8817
P16
LDGV
96+
0.29205
0.04924
0.0274
0.4967
0.8209
2.0574
0.6019
2.0106
0.6654
0.1479
0.1851
0.3987
0.7276
0.7477
1.1209
1.0535
0.7882
0.5132
P17
LDGV
96+
0.23667
0.02554
-0.2222
0.7177
0.3166
1.7918
0.9891
1.2672
0.7007
0.7511
0.1642
0.6983
0.7129
0.7751
0.8196
1.3698
0.5666
0.1481
2.2487
P18
LDGV
96+
0.211
0.02076
1.6186
0.5321
0.4849
0.6707
1.2541
1.116
0.6247
0.9096
0.3981
0.9301
0.7173
0.7777
0.8774
0.7931
1.0214
0.0389
1.7607
0.6392
P19
LDGV
96+
0.05419
0.00995
0.7814
0.7041
0.7989
0.4985
0.7204
0.9434
0.6972
0.728
0.6052
0.6829
0.7065
0.7486
0.7235
0.6863
0.7334
0.6843
0.7333
0.7391
0.7203
M147 Phase 2 Repression Coefficients, CO, 1996 and Newer Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
015
C16
C17
C18
C19
B11
LDGV
96+
2.8132
0.6365
1.9389
B12
LDGV
96+
1.55476
0.22139
1.2687
2.444
B13
LDGV
96+
1.21697
0.12297
1.1955
1.6069
1.8026
B14
LDGV
96+
0.54556
0.13004
0.9864
1.1698
1.3748
2.3969
B15
LDGV
96+
0.4606
0.1147
0.9893
1 0617
1.3564
1.8394
0.7573
B16
LDGV
96+
0.4633
0.09077
0.985
1.0604
1.4116
1.8485
0.4981
1.6802
B17
LDGV
96+
0.33866
0.04766
0.968
1.0626
1.1629
1.9325
0.5734
0.6152
3.1735
B18
LDGV
96+
0.30627
0.03898
0.9596
1.0743
1.0207
1.4501
0.9473
0.8378
2.7247
0.6646
B19
LDGV
96+
0.07437
0.01481
0.9598
1.0098
0.9451
0.986
0.9371
1.0864
1.159
0.9807
0.9897
-------
0
1
to
IM147 Composite Regression Coefficients, NOx, 1981 to 1985 Model Year LDGTI's
- Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT1
81-85
1.62529
2.96654
22.8751
P2
LDGT1
81-85
0.82038
0.84089
-0.5552
5.7627
P3
LDGT1
81-85
0.76521
0.73117
1.5629
4.5821
7.1986
P4
LDGT1
81-85
0.75286
0.70493
-1.1236
4.4969
4.9474
8.0174
P5
LDGT1
61-85
0.74049
0.73769
-3.3308
4.0049
3.8497
6.663
3.702
P6
LDGT1
81-85
0.73674
0.71486
-3.9127
4.0869
3.8046
3.6966
3.0011
4.8137
P7
LDGT1
81-85
0.71739
0.69385
-3.0212
3.5234
3.5081
3.9937
-0.1792
6.3496
3.9431
P8
LDGT1
61-85
0.70652
0.67353
2.6651
3.5435
2.1822
2.6706
0.0115
3.2904
2.6238
3.9683
P9
LDGTl
81-85
0.70505
0.67708
-1.8358
3.4819
2.4475
3.1809
0.6337
3.7784
2.952
4.1428
-2.2115
. :
P10
LDGT1
81-85
0.70227
0.66689
-2.4177
3.494
2.3056
3.6942
0.2591
3.8357
2.4151
4.2563
-4.019
2.4819
P11
LDGT1
81-85
0.61346
0.34129
0.06
2.6963
1.5535
2.4383
1.5201
5.655
0.7401
2.8173
-2.4441
1.0401
1.3101
P12
LDGT1
81-85
0.59022
0.30932
-0.6809
2.658
1.3697
1.2656
1.4801
3.8951
1.356
2.1592
-2.7018
1.3911
0.8585
1.6633
P13
LDGT1
81-85
0.27799
0.12998
-0.1294
0.9449
1.4979
2.9854
0.7785
0.1823
-0.313
0.9976
0.5738
0.4624
0.694
1.076
2.7355
P14
L0GT1
81-85
0.19309
0.05416
0.6065
0.856
0.9251
1.6344
0.8251
-0.094
0.4201
1.4703
0.998
0.7742
0.5699
0.79
1.6886
1.7582
P15
LDGT1
81-85
0.10867
0.02502
0.5161
0.6742
1.1768
1.4558
0.2902
1.3526
0.7996
0.8754
0.732
0.6248
0.6617
0.8183
0.6141
1.2315
1.3505
P16
LDGT1
81-85
0.09725
0.01363
0.4935
0.6784
1.1208
1.1307
0.3923
0.9511
0.8969
0.5086
0.7311
0.6773
0.699
0.7813
0.8355
1.1375
1.1283
1.0972
P17
L0GT1
81-85
0.09344
0.01174
0.5252
0.7346
0.9639
1.1015
0.6838
0.7005
0.7356
0.6406
0.6359
0.4841
0.6953
0.7762
0.8511
1.1597
1.0383
1.0222
1.0523
P18
LDGT1
81-85
0.05721
0.00436
0.6442
0.6817
1.0499
1.0725
0.6056
1.15
0.5722
0.7474
0.8465
0.5131
0.6842
0.8341
0.6429
0.922
0.8319
1.1396
0.7657
0.8484
R19
LDGT1
81-85
0.0165
0.00076
0.7099
0.6986
0.7653
0.7747
0.7168
0.9007
0.7285
0.7007
0.7342
0.6359
0.7093
0.736
0.7337
0.6964
0.7033
0.7238
0.7471
0.7212
0.7201
IM147 Phase 2 Regression Coefficients, NOx, 1981 to 1985 Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
019
B11
LDGTl
81-85
1.076
0.85959
3.4954
B12
LDGT1
81-85
1.00048
0.67351
2.4254
3.4463
B13
LDGTl
81-85
0.38989
0.20314
0.8889
1.8075
3.7573
B14
LDGTl
61-65
0.27037
0.0891
0.8213
1.2937
2.454
2.3704
B15
LDGTl
81-85
0.15297
0.04337
0.8745
1.292
1.0958
1.7178
1.8359
B16
LDGT1
81-85
0.13123
0.01686
0.9513
1.0861
1.1153
1.5763
1.4991
1.5837
B17
LDGT1
81-85
0.12525
0.01412
0.9423
1.0698
1.128
1.6173
1.4024
1.4526
1.3599
B16
L0GT1
81-85
0.08127
0.0094
0.8966
1.2134
0.8469
1.3163
1.0875
1.7522
1.1639
1.0764
B19
LDGT1
81-85
0.01759
0.00132
0.9552
1.0089
0.0919
0.9568
0.9493
1.0003
1.0401
0.9479
1
-------
o
K>
oo
IM147 Composite Regression Coefficients, NOx, 1986 to 1989 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11 i
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT1
86-89
1.23043
2.06694
23.5015
P2
LDGT1
66-89
0.70187
0.74907
1.2554
4.6182
P3
LDGT1
86-89
0.65788
0.66459
1.9511
3.4741
6.6849
P4
LDGT1
86-89
0.65847
0.6622
2.0121
3.4718
6.5638
0.4801
P5
LDGT1
86-89
0.60304
0.65314
0.8979
2.5886
3.7778
0.693
7.82
P6
LDGT1
86-89
0.60221
0.64909
0.8986
2.5939
3.729
-0.3487
7.391
2.7912
P7
LDGT1
86-89
0.58554
0.63588
-0.2658
2-2631
2.7626
0.2519
4.6311
4.814
3.3278
P0
LDGT1
86-89
0.57068
0.60583
0.1372
2.1042
2.0549
-1.463
4.181
1.0084
2.9915
4.4107
P9
LDGT1
86-89
0.57122
0.60597
0.1342
2.1057
2.0299
-1.4503
4.1619
0.9698
2.989
4.3952
0.1053
P10
LDGT1
86-89
0.56454
0.59282
0.9316
2.1076
1.5254
-0.8557
3.4898
1.6834
2.3002
4.575
-1.5419
3.1754
P11
LDGT1
86-89
0.44918
0.31619
0.8774
1.547
0.3721
-1.2258
3.7047
1.8503
0.299
3.4571
-0.7772
1.6757
1.4183
P12
LDGT1
86-89
0.39015
0.23116
2.1899
1.3771
-0.3977
-1.386
2.7841
0.6668
0.7794
1.8436
-0.1039
1.861
0.955
2.7651
P13
LDGT1
66-89
0.21107
0.11066
-0.9618
0.6673
1.0909
0.613
1.2202
0.9175
0.3457
1.1499
1.3016
-0.1198
0.8075
1.3691
2.4303
P14
L0GT1
86-89
0.15824
0.06623
-0.4774
0.6974
0.7145
0.8449
1.214
0.9814
0.6256
0.7295
1.5692
0.6474
0.63S9
0.8685
1.8063
1.4835
P15
LDGT1
66-89
0.10582
0.03533
0.0217
0.6721
0.9237
0.9185
0.462
1.2051
0.9519
0.6397
1.0663
0.5564
0.7033
0.8047
0.8872
1.1532
1.2727
P16
LDGT1
86-89
0.09624
0.03427
0.4994
0.6974
0.6351
1.2145
0.5601
0.5254
1.017
0.3771
0.9662
0.7977
0.7267
0.8275
0.9401
1.0394
0.9564
1.2214
P17
LDGT1
86-89
0.09019
0.02973
0.3688
0.7491
0.5027
1.1374
0.7202
0.7519
0.9628
0.4449
0.9374
0.4114
0.7103
0.847
0.8964
1.087
0.896
1.1054
1.2517
P18
LDGT1
88-89
0.05194
0.01155
0.561
0.6874
0.7247
0.8944
0.4555
1.1666
0.7413
1.03
0.6717
0.4644
0.7014
0.8099
0.637
0.8736
0.8444
0.9947
0.7536
0.9298
P19
LDGT1
86-89
0.01448
0.0007
0.522
0.7051
0.7802
0.7612
0.7043
0.8674
0.7473
0.7107
0.6734
0.7097
0.7081
0.7257
0.7131
0.7176
0.7067
0.7115
0.6881
0.7158
0.7428
IM147 Phase 2 Regression Coefficients, NOx, 1986 to 1989 Model Year LDGTTs
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11 '
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT1
86-89
0.71447
0.56214
2.9488
B12
LDGT1
86-89
0.58562
0.33666
1.7796
4.5068
B13
LDGT1
86-89
0.29253
0.15935
1.0449
2.1599
3.1651
B14
LDGT1
66-89
0.22004
0.08924
0.9001
1.3657
2.486
1.9841
B15
LDGT1
86-89
0.14504
0.05
0.9776
1.164
1.1562
1.5604
1.7468
B16
LDGT1
86-89
0.13142
0.04594
1.0201
1.0708
1.2975
1.3785
1.3303
1.5302
B17
LDGT1
86-89
0.12255
0.04125
0.977
1.1131
1.218
1.4703
1.2356
1.4639
1.602
B18
LDGT1
86-89
0.0715
0.02293
0.937
1.1335
0.8403
1.2196
1.0929
1.5766
0.9306
1.2015
B19
LDGT1
86-89
0.01559
0.00168
0.9648
0.9858
0.9551
0,9715
0.9562
1.0147
0.9809
0.9807
0.9956
-------
0
1
K>
VO
1
V1147 Composite Regression Coefficients, NOx, 1990 to 1995 Model Year LDGH's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
Cl6
C17
C18
C19
P1 .
LDGTl
90-95
1.15989
1.24128
32.9634
P2
LDGT1
90-95
0.72181
0.40146
0.679
4.5914
P3
LDGT1
90-95
0.69216
0.32475
2.1542
3.4912
6.7253
. i
P.4.
LDGTl
90-95
0.66749
0.3123
2.1371
3.4737
5.023
5.8563
P5
LDGTl
90-95
0.62455
0.33205
3.4043
2.3634
1.7239
5.0562
8.7446
. -
P6
LDGTl
90-95
0.61978
0.33046
2.2722
2.3005
1.5599
2.9584
8.2913
6.1514
P7
LDGT1
90-95
0.58792
0.29749
2.0622
1.8473
0.0555
3.6238
4.2618
7.9809
5.3164
P8
LDGT1
90-95
0.55746
0.26129
5.9179
1.7445
-0.5662
1.235
4.3563
0.782
3.7446
7.1976
P9
LDGT1
90-95
0.54008
0.24653
4.1085
1.8313
-1.1591
2.0187
2.9278
-0.7348
3.4121
5.4986
7.8688
P.10
LDGT1
90-95
0.53603
0.2342
2.929
1.8441
-1.4993
2.6412
2.6552
0.0076
2.7177
5.458
6.0386
2.7176
P11
LDGT1
90-95
0.45579
0.05778
3.2806
1.051
0.2096
0.7143
2.8133
1.4291
0.9563
2.9071
6.0179
0.9428
1.4469
P12
LDGT1
90-95
0.36036
0.0273
4.3485
0.9563
-0.635
0.3902
2.43
0.4675
1.7444
0.5941
4.1616
1.3325
0.6814
3.9706
P.13
LDGT1
90-95
0.17758
0.04529
0.6553
0.7331
0.4798
1.1424
1.2661
0.9419
0.3362
0.8
1.2658
0.277
0.7426
1.2309
2.6602
P14
LDGTl
90-95
0.12741
0.02345
0.6061
0.7033
0.6424
1.2708
1.124
1.4638
0.5682
0.7251
0.9404
0.6888
0.6431
0.6253
1.6054
1.8253
P15
LDGT1
90-95
0.07822
0.01761
1.1782
0.6831
0.7996
1.3552
0.7806
0.9482
0.6863
0.8429
1.0081
0.3774
0.6949
0.7428
0.8665
1.0954
1.3651
P16
LDGT1
90-95
0.06969
0.01628
1.367
0.7128
0.824
1.0691
0.8406
0.6506
0.711
0.7051
0.914
0.6359
0.7085
0.724
0.8452
1.0797
1.0435
1.2099
PI 7
LDGT1
90-95
0.06569
0.00712
1.0171
0.713
0.7449
0.9912
0.8383
0.5662
0.7047
0.7856
0.8682
0.5184
0.7126
0.7606
0.8503
1.0453
1.0145
1.1224
1.218
P18
LDGT1
90-95
0.03909
0.0008
0.7599
0.7108
0.6532
1.2612
0.7148
1.122
0.5737
0.837
0.7408
0.6827
0.6897
0.8094
0.6626
0.8525
0.8302
1.1087
0.8134
0.8578
P19
LDGT1
90-95
0.01086
0.00079
0.757
0.7
0.7412
0.7991
0.7237
0.854
0.7206
0.6726
0.7413
0.7075
0.7135
0.7264
0.7129
0.729
0.7202
0.7122
0.7434
0.6941
0.73
IM147 Phase 2 Regression Coefficients, NOx, 1990 to 1995 Model Year LDGH's
Regression Coefficients
Segment
dumber
RMS
Error
Reg. .
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
cii
C12
C13
C14
C15
C16
C17
C16
C19
B11
LDGT1
90-95
0.70725
0.11508
.
3.158
812
LDGT1
90-95
0.54083
0.02981
1.5474
5.8806
B13
LDGT1
90-95
0.24246
0.06072
0.9747
1.7935
3.6
B14
LDGT1
90-95
0.1753
0.03555
0.8927
1.2292
2.2351
2.4522
B15
LDGT1
90-95
0.10866
0.02903
0.9331
1.1278
1.1969
1.4839
1.8483
B16
LDGTl
90-95
0.09571
0.02747
0.9823
1.0076
1.1844
1.4428
1.4242
1.6458
B17
LDGTl
90-95
0.08918
0.01127
0.9705
1.0641
1.1573
1.415
1.3597
1.5734
1.7341
B18
LDGT1
90-95
0.05491
0.00454
0.9276
1.1839
0.9094
1.1571
1.0953
1.6597
1.2012
1.1201
B19
LDGT1
90-95
0.01259
0.0022
0.9623
0.9896
0.9651
0.9817
0.9732
0.9769
1.063
0.9406
0.9926
-------
n
i
U>
O
IM147 Composite Regression Coefficients, NOx, 1996 and Newer Model Year LDGTI's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT1
96+
0.75596
0.48362
23.6975
P2
LDGT1
96+
0.46673
0.18401
12.1233
3.0933
P3
LDGT1
96+
0.46374
0.16288
12.749
2.6461
3.1462
P4
LDGT1
96+
0.46514
0.16338
12.1028
2.534
2.3048
4.9422
P5
LDGT1
96+
0.3791
0.14765
18.1804
1.5554
-1.8687
0.8812
12.6955
P6
L0GT1
96+
0.38121
0.14804
18.193
1.5564
-1.9029
1.0072
12.8058
-0.5274
P7
LDGT1
96+
0.3264
0.12947
19.6606
1.0854
-0.0898
-4.9504
5.5367
-2.3077
9.8773
P8
LDGT1
96+
0.31962
0.14076
16.8826
0.9749
1.2061
-5.4908
2.7593
-4.5733
4.9464
8.5778
P9
LDGT1
96+
0.32062
0.15138
18.5988
0.8971
1.8707
-4.2693
2.4773
-3.9369
4.9682
9.0399
-2.7492
P10
LDGT1
96+
0.24391
0.13494
9.3292
0.7629
2.6156
-11.0979
1.3803
2.9882
1.7874
11.2329
-6.1993
6.6929
P11
LOGT1
96+
0.18099
0.10785
0.7068
0.5528
3.4731
8.9995
-0.2732
4.7615
0.0782
7.2458
-2.3664
2.4504
1.28
P12
LDGT1
96+
0.17159
0.09436
-0.6219
0.5365
2.5353
-3.0512
-0.8472
1.9087
-0.1596
6.9805
-2.7139
3.1761
0.9818
1.8863
P13.
LDGT1
96+
0.12899
0.07087
1.7268
0.3591
2.536
1.5628
0.5764
-2.7771
1.4626
2.9783
1.5186
1.4469
0.7662
1.0775
1.7081
PU
LOGT1
96+
0.10769
0.06233
-1.0403
0.4537
1.9505
0.0036
0.7496
1.6955
0.3857
3.6594
-0.8929
0.9085
0.7212
0.4779
1.0256
1.5265
P15
LDGT1
96+
0.06499
0.02733
-0.039
0.6102
1.5778
0.9818
0.0528
0.1089
1.1166
2.0576
0.9366
1.0631
0.7396
0.86
0.6974
0.7671
1.3416
P16
LDGT1
96+
0.05153
0.02163
0.4719
0.5741
2.1518
-0.4167
0.1408
-0.69
0.7472
2.1867
0.0185
0.9315
0.8149
0.7897
0.797
0.6681
0.6861
2.2482
P17
LOGT1
96+
0.04794
0.01483
0.175
0.6024
1.5796
0.5006
0.714
-1.1578
0.8449
2.0981
0.6903
1.1098
0.7149
0.9709
0.8286
0.7074
0.7195
1.7387
1.6876
P18
LDGT1
96+
0.02782
0.00211
0.6747
0.6416
1.5148
-0.1218
0.5752
0.4924
0.7534
1.8426
-0.2846
0.9695
0.7007
0.83
0.7552
0.7376
0.7664
1.1279
0.6032
0.758
P19
LDGT1
96+
0.00403
0.00098
0.7432
0.7168
0.7163
0.6233
0.7292
1.0286
0.6325
0.9521
0.6402
0.7217
0.7125
0.7112
0.7066
0.7343
0.6734
0.7676
0.7274
0.7239
0.7481
IM147
Phase 2 Regression Coefficients, NOx, 1996 and Newer Model Year LDGTVs
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGTl
96+
0.28558
0.12907
2.2171
B12
LDGT1
96+
0.24779
0.09313
1.6406
3.4138
B13
LDGT1
96+
0.18123
0.08816
1.252
1.6471
2.2975
B14
LDGTl
96+
0.15739
0.06574
0.9727
1.3922
1.7725
1.7148
B15
LDGTl
96+
0.09057
0.02458
1.0721
1.1839
0.9796
1.028
1.925
B16
LDGTl
96+
0.07613
0.02379
1.1061
1.2115
1.0191
0.9869
1.1519
2.6649
B17
LDGTl
96+
0.07139
0.00843
1.0352
1.3901
1.1173
0.956
1.1159
2.3796
2.0104
818
LDGT1
96+
0.04333
0.00582
0.9841
1.2998
1.0418
0.9729
1.1585
1.4094
0.9684
1.1153
B19
LDGT1
96+
0.00565
0.00178
0.9761
0.9749
0.9793
0.9737
0.9206
1.0687
0.9432
0.9717
1.0089
-------
IM147 Composite Regression Coefficients, NOx, 1981 to 1985 Model Year LDGT2's
<
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
PI
LDGT2
81-65
1.99804
3.69101
10.2569
P2
LDGT2
81-85
0.87409
0.91821
-7.9377
6.5305
P3
LDGT2
81-85
0.84122
0.83444
-8.613
5.4145
7.2247
P4
LDGT2
81-85
0.82892
0.89342
-10.2897
5.4633
2.5096
8.8657
P5
LDGT2
81-85
0.80333
0.89081
9.4068
4.7319
0.276
9.1558
5.3418
P6
LDGT2
81-85
0.7983
0.87735
-13.437
4.8675
0.2244
7.2171
4.8373
4.2644
P7
LDGT2
81-85
0.79972
0.89425
-13.7206
4.6923
-0.887
7.62
4.3577
4.6628
1.0456
P8
LDGT2
81-85
0.79473
0.84415
-11.4499
4.5789
-1.4261
5.7988
3.4916
2.7174
1.1653
3.2995
P9
LDGT2
81-85
0.79705
0.83069
-11.6615
4.6051
-1.4956
5.8933
3.2922
2.8161
1.1255
3.028
0.9261
i
P10
LDGT2
81-85
0.79648
0.81111
-11.8131
4.561
-1.2986
5.9935
2.4713
3.4666
0.516
3.1827
-0.4726
2.7254
P11
LDGT2
81-85
0.6859
0.51142
-7.4827
3.0898
-1.5249
6.9352
3.5295
2.9385
1.4316
3.7346
1.3238
2.5498
1.2169
P12
LDGT2
81-85
0.54957
0.33422
-8.1307
2.4318
-0.7549
0.1755
1.141
5.7388
3.1487
2.2295
0.7669
0.0157
0.6425
3.1201
P13
LOGT2
81-85
0.28368
0.1926
-3.0729
0.495
1.2524
2.9861
-0.2928
2.9259
1.4715
0.5835
2.0376
-0.4876
0.712
1.7639
2.5466
P14
LDGT2
81-85
0.17079
0.09831
-1.9366
0.7267
0.5707
1.4814
-0.6512
2.8236
1.4133
1.5048
1.8051
0.7881
0.6265
1.0991
1.5732
1.6465
R15
LDGT2
81-85
0.13538
0.05816
-1.2593
0.6843
0.9631
1.699
-1.1762
2.6604
1.5116
1.2555
1.0619
1.1308
0.6644
0.9993
0.9615
1.3864
0.9139
P16
LDGT2
81-85
0.12927
0.04855
-0.6587
0.6411
0.9878
1.2005
-0.6445
2.468
1.4956
0.99
1.0154
1.2544
0.6922
0.8896
1.0162
1.349
0.717
0.6881
P17
LDGT2
81-85
0.12798
0.04485
-0.3609
0.6627
0.916
1.0272
-0.4442
2.2519
1.3213
1.1276
1.0486
1.1459
0.6823
0.8485
0.9897
1.3779
0.7091
0.682
0.6099
P18
LDGT2
81-85
0.08606
0.02155
0.3328
0.587
1.0724
1.0072
0.5742
1.9792
0.4674
1.4066
1.3569
-0.0211
0.6682
0.8649
0.7604
0.9339
0.6473
0.8806
1.1799
0.9731
P19
LDGT2
81-85
0.02488
0.0013
1.1429
0.6401
0.9555
0.8446
0.5989
0.8719
0.8311
0.6923
0.5888
0.7318
0.7118
0.6787
0.7213
0.7279
0.7365
0.6649
0.7072
0.679
0.7799
IM147 Phase 2 Regression Coefficients, NOx, 1981 to 1985 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
81-85
1.42447
1.55851
3.2137
B12
LDGT2
81-85
1.03629
0.77984
1.5673
6.4951
B13
LDGT2
81-85
0.38433
0.27378
0.8325
2.718
3.5162
814
LOGT2
81-85
0.26921
0.18815
0.8123
1.6716
2.4085
2.1998
B15
LDGT2
81-85
0.22469
0.11816
0.8718
1.4697
1.4685
1.8766
1.2618
B16
LDGT2
81-85
0.19939
0.09578
0.9679
1.128
1.6012
1.7461
0.8574
1.4532
017
LDGT2
81-85
0.19039
0.09229
0.9221
1.0819
1.4847
1.7994
0.8974
1.351
1.5644
B18
LDGT2
81-85
0.15301
0.07095
0.8347
1.2468
1.0994
1.3645
0.8676
1.6334
1.4952
1.0081
B19
LDGT2
81-85
0.03621
0.00936
. 0.943
0.9237
0.9993
0.9988
0.978
0.8541
1.0342
0.8947
1.1014
-------
0
1
LO
K)
IM147 Composite Regression Coefficients, NOx, 1986 to 1987 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGT2
86-67
1.61179
3.13663
31.3453
P2
LDGT2
86-87
0.97795
1.10882
-3.0631
5.8632
P3
LDGT2
86-87
0.94502
0.9839
-5.0759
4.548
8.9192
P4
LOGT2
86-87
0.93501
0.93053
-3.3825
4.4591
5.2516
12.419
P5
LDGT2
86-87
0.85394
0.96285
-4.4203
3.6091
-1.7208
11.4179
9.1873
P6
LDGT2
86*87
0.84757
0.91259
-4.4095
3.6676
-1.5513
6.2687
8.3678
9.6507
P7
LDGT2
86-87
0.85029
0.91803
-4.0853
3.616
-2.1286
6.0176
7.6929
10.0876
0.7099
P8
LDGT2
86-87
0.85275
0.92194
-3.8044
3.5415
-2.4148
4.8872
7.7612
9.5053
0.6261
1.3726
P9
LDGT2
86-87
0.83178
0.86005
-3.778
4.3986
Ś6.3555
9.9437
4.6162
8.3808
-0.6741
-0.2851
9.9702
P10
LDGT2
86-87
0.80985
0.90764
-7.0196
4.1229
-5.9089
8.2497
2.6936
10.9511
-1.1815
0.3123
3.2549
6.8821
P11
LDGT2
86-87
0.76148
0.62189
-8.9566
3.094
-4.9268
6.2947
5.15
9.9243
-2.4765
-0.213
3.9969
6.4604
0.8832
P12
LDGT2
86-87
0.59003
0.3735
-2.4217
2.087
-2.6617
4.0616
2.224
1.9821
-1.3263
-2
4.7639
5.4191
0.2795
5.406
P13
LOGT2
86-87
0.33385
0.19007
0.2814
1.1386
0.5116
3.255
-0.1077
2.4058
0.5387
-1.7814
4.2622
-0.1798
0.5637
0.4606
3.2021
P14
LDGT2
86-87
0.20099
0.13811
1.8759
0.6935
0.5775
3.0569
0.9302
1.5194
0.7032
0.6974
2.2407
0.0121
0.4793
0.2025
1.9049
2.1646
P15
LDGT2
86-87
0.1036
0.0461
0.3765
0.6236
1.678
0.8805
1.0648
1.097
0.2424
-0.0264
1.7158
0.0601
0.6449
0.9194
1.0397
1.1878
1.3176
P16
LDGT2
86-87
0.09657
0.03819
0.6854
0.5604
1.7082
0.6146
0.6222
1.3221
0.5089
-0.0753
1.3097
0.8115
0.675
0.9002
0.9782
1.1823
1.0924
0.8604
P17
LDGT2
86-87
0.08202
0.02243
0.7341
0.6449
1.3141
-0.1109
0.6071
1.3461
0.4166
0.5635
1.6121
-0.0273
0.635
0.8997
0.9541
1.1481
1.0157
0.8336
1.7925
P18
LDGT2
86-87
0.05719
0.02206
0.9248
0.6705
1.2239
0.0494
0.6674
1.7312
0.3068
0.8156
1.3908
0.0199
0.6645
0.9323
0.8095
0.9223
0.7808
0.9324
1.241
. 0.7403
P19
LDGT2
86-87
0.01482
0.00786
0.6527
0.6745
0.6642
0.6799
0.7944
0.7706
0.6813
0.8969
0.6883
0.6733
on
0.7009
0.7385
0.7569
0.6679
0.7207
0.8026
0.7243
0.7208
IM147 Phase 2 Repression Coefficients, NOx, 1986 to 1987 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
68-87
1.44862
1.6677
2.507
B12
LDGT2
86-67
0.92821
0.61792
0.4543
9.6299
B13
LDGT2
66-87
0.45405
0.32299
0.78S2
0.5919
4.3588
Bl4
LDGT2
88-87
0.27745
0.22123
0.6178
0.2412
2.551
2.9603
B15
LDGT2
86-87
0.14699
0.0967
0.8032
1.2698
1.2876
1.6105
1.8179
016
LDGT2
86-87
0.1382
0.07565
0.8S9
1.1449
1.2732
1.5831
1.6009
0.9193
B17
LDGT2
86-87
0.1172
0.05283
0.8945
1.2281
1.135
1.5637
1.3545
1.3493
2.1364
B18
LDGT2
86-87
0.08824
0.05293
0.8887
1.3239
0.9624
1.2543
1.0381
1.5212
1.2998
0.9849
B19
LDGT2
86-87
0.0176
0.00876
0.958
0.9848
0.9632
0.9799
0.9647
0.9458
1.1217
0.9723
0.9846
-------
0
1
U>
U)
IM147 Composite Repression Coefficients, NOx, 1988 to 1995 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
Pi
LDGT2
88-95
1.24878
2.35108
23.5648
P2
L0GT2
88*95
0.82176
0.87411
0.1479
4.6355
P3
LDGT2
88-95
0.77845
0.71657
-1.7162
3.425
8.1364
P4
LDGT2
88-95
0.7742
0.70093
-1.8478
3.4367
6.2476
5.1308
.
P5
LDGT2
88-95
0.73556
0.71294
-1.4757
2.6692
3.7613
3.2899
6.2663
P6
LDGT2
88-95
0.72738
0.72398
-2.5889
2.6139
3.8661
-0.5835
5.6306
6.8957
P7
LDGT2
88-95
0.70212
0.63963
-1.6939
2.0359
3.1437
0.5064
1.7183
8.4446
5.0667
P8
LDGT2
88-95
0.68022
0.60531
-0.4848
1.8702
2.347
-1.174
1.1716
2.9882
4.0341
6.3227
P9
LDGT2
88-95
0.67704
0.59719
-1.6152
1.9533
1.922
0.993
0.5073
2.3082
3.7294
5.9607
3.1202
. f
P10
L0GT2
88-95
0.66525
0.59249
-1.2143
1.9942
1.1629
0.3777
0.1791
2.4339
2.4689
6.3272
-0.1222
4.5356
P11
LDGT2
88-95
0.49566
0.16456
-1.995
0.9652
1.6003
1.6143
0.2178
0.8508
-0.2399
3.1105
1.8601
2.7668
1.9044
P12
L0GT2
88-95
0.39011
0.13915
-0.5048
0.8942
0.6646
0.7011
0.5649
0.3036
0.9763
0.3149
0.4631
3.3575
0.9895
3.9381
P13
LDGT2
88-95
0.22071
0.13586
0.1295
0.5892
1.2326
1.6722
0.4435
0.7176
0.4615
0.651
1.5772
0.5428
0.7152
1.6198
2.5117
P14
L0GT2
88-95
0.16814
0.08085
0.4177
0.6242
1.2076
1.54
0.4306
1.0374
0.8469
0.7753
1.5915
0.5229
0.5567
1.0483
1.735
1.6545
P15
LDGT2
88-95
0.09792
0.04274
1.2914
0.6299
0.932
1.5301
0.7496
0.6925
0.7213
0.5876
.1.3142
0.4284
0.6686
0.9379
0.8221
1.0317
1.4681
P16
LDGT2
88-95
0.08515
0.04
1.3815
0.6274
0.8992
1.2443
0.777
0.5902
0.9112
0.3659
1.1635
0.6339
0.6996
0.6562
0.8431
0.9855
1.1161
1.2384
P17
LDGT2
88-95
0.0793
0.03
1.1604
0.6609
0.7332
1.166
0.8211
0.5607
0.8664
0.4764
1.1288
0.4527
0.7155
0.6591
0.8406
0.9915
1.0422
1.1105
1.2867
P18
L0GT2
88-95
0.04482
0.00528
1.0205
0.6891
0.8472
0.9762
0.6666
1.0589
0.5961
0.8411
0.7658
0.5158
0.703
0.6891
0.6603
0.6286
0.8216
1.0348
0.9346
0.8792
P19
LDGT2
88-95
0.0136
-0.00252
0.7213
0.711
0.747
0.8258
0.7089
0.7055
0.696
0.744
0.7329
0.7233
0.715
0.7382
0.7138
0.716
0.7117
0.7363
0.7333
0.6995
0.7243
IM147 Phase 2 Regression Coefficients, NOx, 1988 to 1995 Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
68-95
0.71642
0.34014
3.1658
B12
LDGT2
88-95
0.55127
0.22451
1.6945
5.1988
013
LDGT2
88-95
0.30245
0.20374
0.8854
2.4317
3.3283
B14
LDGT2
88-95
0.23377
0.14045
0.7529
1.6901
2.356
2.1936
B15
LDGT2
88-95
0.13663
0.07379
:. 0.881
1.4099
1.0647
1.3754
2.0165
B16
LDGT2
88-95
0.11912
0.06757
0.9498
1.1559
1.1515
1.31
1.5401
1.6409
B17
LDGT2
88-95
0.10918
0.04812
0.9645
1.1668
1.1296
1.3343
1.4075
1.494
1.8722
Q16
LDGT2
88-95
0.0635
0.01509
0.9291
1.3099
0.8516
1.1259
1.093
1.5612
1.2713
1.1432
019
LDGT2
88-95
0.01423
0.0016
0.9872
1.0052
0.9647
0.96
0.9599
1.0009
1.0227
0.9483
Ś1.0006
-------
n
t
U>
4^
IM147 Composite Regression Coefficients, NOx, 1996 and Newer Model Year LDGT2's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
013
014
C15
C16
C17
C18
C19
Pi
LDGT2
96+
0.578
0.71379
15.4654
P2
LDGT2
96+
0.33821
0.37061
-21.218
3.256
P3
LDGT2
96+
0.33307
0.33021
-17.4893
2.5344
3.6906
P4
LDGT2
96+
0.30416
0.30038
-14.5682
2.2242
11.6872
-12.2279
i
P5
LDGT2
96+
0.29072
0.28842
-17.6088
1.8185
9.3908
-10.8536
4.734
P6
LDGT2
96+
0.2883
0.30337
-11.6965
1.745
9.6903
-10.7171
5.8457
-7.1022
P7
LDGT2
96+
0.24718
0.25136
-20.0849
1.7221
5.3396
-2.4287
-0.5675
-9.9875
8.4455
t
P8
LDGT2
96+
0.23492
0.2519
-22.4763
1.5623
5.2794
-3.3995
-2.4191
-8.9711
6.6514
7.9536
P9
LDGT2
96+
0.23606
0.24316
-22.27
1.5055
5.3167
-3.4781
-3.0309
-9.6937
6.8514
8.1317
3.1882
<
P10
LOGT2
96+
0.2359
0.23077
. -20.9734
1.5584
5.6716
-3.4051
4.2894
-8.7832
6.8235
6.3126
1.8606
3.9221
N
P11
LDGT2
96+
0.18371
0.16444
-10.5542
0.5675
3.1636
-1.1457
-3.3023
0.5825
3.4287
3.1947
4.4137
2.9738
1.3085
P12
LDGT2
96+
0.18578
0.16354
9.9999
0.6169
2.8046
-0.7267
3.2312
0.1935
3.5982
3.0439
4.2462
2.9303
1.2191
0.2662
P13
LDGT2
96+
0.12127
0.05817
-4.2485
0.6116
4.4173
-3.0746
-1.9566
-5.1095
2.1593
3.236
5.8951
2.3221
1.0537
-0.1452
1.6328
P14
LOGT2
96+
0.06953
0.0183
4.3668
0.7373
2.7372
0.575
-1.3744
-3.22
2.644
-0.5548
1.4212
3.1253
0.6404
1.0543
0.791
1.4102
P15
LDGT2
96+
0.05854
0.01128
2.8175
0.7837
1.5604
1.7698
-0.6221
-2.7325
2.4484
0.2815
0.0066
2.3246
0.5592
1.0307
0.7388
1.1608
0.7986
P16
LDGT2
96+
0.05584
0.00873
1.9819
0.7799
1.6432
1.5678
0.4135
-2.4395
2.0557
0.0994
0.2982
1.9349
0.6182
0.9336
0.7441
1.0662
0.7148
1.3475
P17
LDGT2
96+
0.05251
0.00847
0.351
0.7816
1.4739
1.5467
-0.3431
-2.0093
2.0146
0.208
0.4812
1.4451
0.723
0.6611
0.7379
1.0331
0.6538
1.3684
2.1742
P18
LDGT2
96+
0.01918
0.00084
1.8364
0.7759
1.0399
1.1145
0.2474
-0.0511
0.6696
0.6814
0.6723
1.4583
0.5956
0.9049
0.7647
0.6482
0.6884
1.3503
-0.5973
0.8893
P19
LDGT2
96+
0.00325
0.00226
0.7097
0.7313
0.8044
0.6612
0.6632
0.833
0.6604
0.6611
0.5647
0.8016
0.72
0.6732
0.7118
0.7105
0.7272
0.7857
0.7169
0.7106
0.6983
IM147
Phase 2 Regression Coefficients, NOx, 1996 and Newer Model Year LDGT2's
Regression Coefficients 1
Segment
Number
RMS
Error
Reg.
Constant
CI *
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11 :
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGT2
96+
0.25624
0.28161
2.2649
B12
LDGT2
96+
0.25427
0.26879
1.8738
1.3389
B13
LDGT2
96+
0.18429
0.10134
1.6678
0.8038
2.0982
B14
LDGT2
96+
0.11213
0.05754
1.2535
1.3129
1,0561
1.9394
B15
LDGT2
96+
0.08972
0.02934
1.1523
0.9849
1.0364
1.4467
1.3152
B16
LDGT2
96+
0.08172
0.02161
1.1412
0.9687
1.0064
1.2969
1.0936
2.5063
B17
LDGT2
96+
0.07624
0.01744
1.1996
0.6743
0.9948
1.295
1.0153
2.0425
3.1694
B18
LDGT2
96+
0.03345
0.0095
0.929
1.0616
1.0355
1.1034
0.9077
1.821
-0.1023
1.2122
B19
LDGT2
96+
0.00415
0.0038
0.9821
0.9175
0.9698
0.9467
0.9808
1.0103
1.1121
0.9437
0.9983
-------
0
1
U>
Lh
IM147 Composite Regression Coefficients, NOx, 1981 to 1982 Model Year LDGV's
Repression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LDGV
B1-82
0.91799
1.65041
42.7692
P2.
LDGV
81-82
0.63601
0.72708
17.5251
4.1787
P3.
LDGV
81-82
0.52331
0.58513
-4.6789
2.4424
11.7817
P4
LDGV
81-82
0.52237
0.58386
-6.215
2.4752
10.4671
3.1134
P5
LDGV
81-82
0.50775
0.50405
-4.2723
2.2111
9.2772
2.5761
3.9991
P6
LDGV
81-82
0.50352
0.48347
-3.9075
2.1456
8.9952
1.1125
3.4211
5.3989
P7
LDGV
81-82
0.49839
0.48687
-4.3121
2.0136
6.6708
0.9449
1.0442
5.8614
2.4035
P8
LDGV
81-82
0.46669
0.46683
1.6928
1.6175
7.9174
-2.5315
1.6902
0.2702
0.6553
7.5772
P9
LDGV
81-82
0.46311
0.452
-0.6119
1.7211
7.5725
-2.1824
0.4138
-2.0068
0.7582
7.4243
3.9487
P10
LDGV
81-82
0.46376
0.44782
-1.0048
1.7433
7.4758
-2.0726
0.0943
-1.7963
0.4584
7.4711
3.3527
1.0886
P11
LDGV
81-82
0.35194
0.23815
-6.5826
0.7668
4.8431
-0.79
0.1979
2.4979
0.684
4.4696
3.7564
0.6973
1.4384
P12
LDGV
81-82
0.2927
0.17049
-9.8733
0.9681
2.776
2.0862
0.4024
1.4025
2.1207
3.7386
2.3379
1.5715
0.6911
2.5825
P13
LDGV
81-62
0.18109
0.07569
-4.5202
0.7632
1.5688
1.3953
0.0498
2.8038
-0.6444
2.0452
0.4039
1.7917
0.6671
1.3367
2.4626
Pt4
LDGV
81-82
0.14099
0.02431
-4.1674
0.7506
1.6956
1.9641
0.1583
1.4761
-0.6375
1.7248
-0.4851
2.7419
0.5236
0.9827
1.8418
1.5633
P15
LDGV
81-82
0.08737
0.01811
-1.382
0.8401
1.3947
1.6138
-0.2041
1.0019
0.7778
0.8836
1.2699
0.6037
0.6393
0.8763
0.6065
0.6832
1.7403
P16
LDGV
81-82
0.07312
0.0041
-1.1445
0.8151
1.058
1.5224
0.0582
0.8216
1.2094
0.5747
1.1231
0.8514
0.7173
0.8313
0.6082
0.7402
1.2902
1.5334
P17
LDGV
81-82
0.07264
0.00367
-0.9811
0.8032
1.0413
1.4118
0.1444
0.648
1.1681
0.6401
1.1443
0.7002
0.7202
0.816
0.6414
0.7463
1.2653
1.4284
0.5328
P18
LDGV
81-82
0.04243
0.00563
0.017
0.7339
0.8526
1.377
0.2716
1.0442
0.6919
0.8178
0.9258
0.5464
0.7223
0.7683
0.6426
0.8313
0.6107
1.4947
0.4033
0.8387
P19
LDGV
81-82
0.01342
0.00395
0.7869
0.7185
0.788
0.8447
0.6524
0.7441
0.6472
0.857
0.6045
0.7562
0.6997
0.7449
0.6929
0.7279
0.7185
0.7737
0.6197
0.7165
0.6907
IM147 Phase 2 Regression Coefficients, NOx, 1981 to 1982 Model Year LDGV's
1 Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
81-62
0.59924
0.60079
2.8438
B12
LDGV
81-62
0.50466
0.48014
1.6
3.6643
B13
LOGV
81-62
0.2646
0.11652
0.8595
2.316
3.4315
B14
LDGV
81-82
0.21268
0.04456
0.7104
1.676
2.6006
2.1072
B15
LDGV
81-82
0.13034
0.02244
0.8971
1.4034
0.8839
1.2572
2.3849
B16
LDGV
81-62
0.1083
0.01418
1.0311
1.1362
1.0575
0.9885
1.7362
2.0954
B17
LDGV
81-62
0.10478
0.0082
1.0248
1.1015
1.1113
1.0374
1.6264
1.8908
1.4012
B18
LDGV
81-82
0.06108
0.00004
0.9845
1.139
0.8671
1.145
0.9874
2.2913
0.654
1.1564
819
LDGV
81-82
0.01536
0.00328
0.9626
1.0202
0.9556
0.9706
0.9459
1.0635
0.9025
0.9816
0.9686
-------
0
1
u>
On
IM147 Composite Repression Coefficients, NOx, 1983 to 1985 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
PI
LDGV
83-85
0.96463
1.86521
18.7418
P2
LDGV
83-85
0.58177
0.74711
0.782
4.5182
P3
LDGV
83-85
0.57164
0.69924
1.7395
3.8215
3.9404
P4
LOGV
83-85
0.56866
0.69182
2.2485
3.7307
2.712
4.1135
Śr
P5
LDGV
83-85
0.54912
0.68329
0.0309
3.6538
-0.5552
2.7207
4.4539
P6
LDGV
83-85
0.54056
0.65766
-0.2605
3.6825
0.6893
0.2755
3.3123
6.9261
P7
LDGV
8345
0.52478
0.64771
-1.2424
3.0961
-0.7131
-0.3096
1.2743
8.0479
3.3901
P8
LDGV
83-85
0.49039
0.64355
-0.2025
2.5593
-0.8873
-2.2555
0.9327
0.9917
2.648
6.9554
P9
LDGV
83-85
0.49081
0.6366
-0.2479
2.5692
-0.9586
-2.1078
0.7709
0.6195
2.6479
6.804
0.8258
P10
LDGV
83-65
0.48198
0.62502
-0.6852
2.5126
-1.7481
0.7564
0.3327
1.3097
1.8572
6.9911
-1.6919
3.7996
P11
LDGV
83-85
0.43238
0.38649
0.6471
1.837
-1.1074
0.237
1.3594
1.4974
0.1122
6.1798
0.33
1.9052
1.0346
P12
LDGV
83-85
0.35372
0.24907
1.0365
1.599
-1.5529
1.0464
1.5953
-0.2612
1.2403
3.835
0.566
1.7663
0.4528
3.3405
P13
LDGV
83-65
0.22239
0.11604
1.4435
0.8789
. 1.1138
3.1005
0.6475
1.7479
0.451
1.0221
0.5773
0.455
0.4514
1.7854
2.2469
P14
LDGV
83-65
0.16685
0.04792
1.1199
0.8083
0.993
2.3512
0.3301
2.1868
0.9638
0.6226
1.2885
1.0545
0.5328
0.7862
1.2774
2.0142
P15
LDGV
83-85
0.09344
0.00442
0.6972
0.724
1.0561
1.1547
0.647
1.6301
0.8766
0.4133
1.7045
0.185
0.7132
0.6701
0.6719
1.0248
1.5987
P16
LDGV
83-85
0.07618
0.01155
1.4799
0.6974
0.9307
0.8525
0.7603
0.9115
1.0175
0.3596
1.3353
0.4874
0.7501
0.6536
0.7419
0.6969
1.1713
1.4857
P17
LDGV
83-85
0.07136
0.00873
1.0722
0.7157
0.74
1.0551
0.6004
0.9688
0.8579
0.6022
0.9517
0.6109
0.7611
0.5992
0.7747
0.9329
1.1217
1.2072
1.6127
P18
LDGV
83-85
0.03952
0.0055
0.6259
0.6692
0.8058
1.0136
0.6208
1.1765
0.512
0.9231
0.7662
0.5965
0.7047
0.7879
0.6504
0.8038
0.8335
1.1663
1.2587
0.9322
P19
LDGV
83-85
0.01224
-0.00128
0.8422
0.7007
0.7698
0.724
0.7372
0.8013
0.7199
0.7142
0.6943
0.7149
0.7141
0.7285
0.7042
0.7381
0.7145
0.6689
0.7795
0.691
0.7508
IM147 Phase 2 Regression Coefficients, NOx, 1983 to 1985 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
01
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C16
C19
B11
LDGV
83-85
0.73374
0.81776
.2.4382
B12
LDGV
83-85
0.54834
0.39309
1.1577
5.767
B13
LDGV
83-85
0.31629
0.20822
0.6375
2.8068
3.1803
814
LDGV
83-85
0.24412
0.1176
0.8023
1.3747
1.9218
2.6998
BIS
LDGV
83-85
0.13375
0.03665
0.9663
1.0509
0.9033
1.3245
2.2612
B16
LDGV
83-85
0.10708
0.03554
1.0501
0.6984
1.0487
1.1265
1.6402
2.0329
B17
LDGV
83-85
0.09639
0.02013
1.0496
0.8104
1.0655
1.2345
1.5252
1.6719
2.3989
B18
LDGV
83-85
0.05614
0.0163
0.9253
1.1246
0.8814
1.096
1.1113
1.8248
1.4559
1.1912
B19
LDGV
83-85
0.01369
0.0022
0.9666
0.9844
0.9642
0.9894
0.982
0.9116
1.1138
0.9386
1.0241
-------
0
1
U)
-J
IM147 Composite Repression Coefficients, NOx, 1986 to 1989 Model Year LDGV's
Repression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
P1
LOGV
86-89
0.78565
1.40846
22.9983
P2
LDGV
86-89
0.57844
0.77237
9.21
3.0925
P3
LDGV
66-89
0.53576
0.63136
9.1703
2.0307
7.0173
P4
LDGV
86-89
0.52985
0.61276
9.5621
2.0154
5.1612
5.5354
P5
LDGV
86-89
0.48733
0.58411
7.6422
1.5437
3.1981
4.1123
5.8762
P6
LDGV
86-89
0.469
0.58
4.0676
1.525
2.9935
0.4973
4.8986
10.1283
P7
LDGV
86-89
0.44452
0.51948
2.5722
1.2222
2.2925
1.1125
2.3363
11.7286
3.4574
P8
LDGV
86-89
0.42664
0.46953
3.5526
1.2073
1.7483
0.3084
2.0313
6.7259
2.6637
5.3318
P9
LDGV
86-89
0.42322
0.45426
3.3018
1.301
1.139
0.1369
1.5033
5.9904
2.5273
4.6613
3.2834
P10
LDGV
86-69
0.42046
0.44434
2.7682
1.279
0.8669
0.8265
1.1624
5.9839
1.9554
4.9824
2.1016
2.3949
P.11
LDGV
86-89
0.3165
0.16273
1.6969
0.9234
0.2926
1.5843
1.5047
5.7857
0.6453
3.2517
2.4719
1.4771
1.3713
P.12
LDGV
86-89
0.28211
0.13132
0.4472
0.9805
-0.3268
1.5058
1.3641
3.9271
1.2167
1.5887
1.8636
1.9637
0.9702
2.2297
PI 3
LDGV
66-89
0.17181
0.05351
1.5839
0.7587
0.7909
1.2166
0.7145
1.5691
0.9527
1.0643
1.0876
0.6012
0.7526
1.15
2.3208
P14
LDGV
86-89
0.13065
0.04173
0.3316
0.7154
1.0709
0.6518
0.7486
1.2646
0.8718
0.8749
0.8522
0.9529
0.6065
0.9816
1.5163
1.6439
P15
LDGV
86-89
0.08069
0.01822
0.1209
0.6879
0.733
1.0408
0.7212
1.5437
0.8545
0.957
0.2801
0.7471
0.6975
0.8112
0.8905
1.0278
1.3107
P16
LDGV
86-89
0.06908
0.01317
0.5535
0.7013
0.7623
0.918
0.6392
1.068
0.9735
0.7864
0.2568
0.9626
0.7416
0.7472
0.8532
0.8939
1.013
1.467
P17
LDGV
86-89
0.06367
0.01125
0.3901
0.7232
0.6166
0.8153
0.6598
1.0147
0.8687
0.8487
0.2077
0.8136
0.7444
0.7667
0.847
0.9405
0.9462
1.3241
1.4006
P18
LDGV
66-89
0.03823
0.00394
0.7351
0.6866
0.7439
1.0847
0.6838
1.2044
0.6292
0.8002
0.5291
0.672
0.6852
0.8818
0.7191
0.82
0.7757
1.2183
0.9059
0.8386
P19
LDGV
66-89
0.01005
0.00003
0.7593
0.7055
0.7238
0.7901
0.6814
0.9355
0.7036
0.7211
0.6966
0.7005
0.7134
0.7342
0.7231
0.7021
0.7072
0.7254
0.7597
0.7156
0.7326
IM147 Phase 2 Regression Coefficients, NOx, 1986 to 1989 Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
CI
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
86-69
0.52925
0.42098
2.5542
B12
LDGV
86-69
0.43224
0.27635
1.6545
3.9022
B13
LDGV
86-89
0.23852
0.0957
1.0646
1.7542
3.3487
B14
LDGV
86-89
0.16118
0.07257
0.867
1.4108
2.1959
2.261
B15
LDGV
66-89
0.11203
0.02849
0.9505
1.1788
1.2539
1.4295
1.7812
B16
LDGV-
66-49
0.0963
0.02351
1.0428
0.9898
1.2047
1.2246
1.4141
1.9452
B17
LDGV
86-89
0.08734
0.01558
1.0275
1.0114
1.167
1.2958
1.2756
1.7751
1.9705
B18
LDGV
86-89
0.05354
0.0073
0.9133
1.2514
0.967
1.1377
1.0194
1.8001
1.2499
1.0893
B19
LDGV
86-89
0.01153
0.00104
0.964
0.9967
0.973
0.9592
0.9531
0.9973
1.0544
0.9572
1.0152
-------
0
1
U>
00
IM147 Composite Repression Coefficients, NOx, 1990 to 1995 Model Year LDGV's
Reqression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11 '
C12
C13
C14
C15
016
C17
C18
C19
P1
LDGV
90-95
0.85265
0.98234
27.2602
P2
LDGV
90-95
0.55676
0.37911
3.4685
3.9198
. ' ii.
P3
LDGV
90-95
0.5187
0.31019
2.4287
2.702
6.9157
P4
LDGV
90-95
0.51364
0.29256
2.566
2.669
5.4346
4.5269
P5
LDGV
90-95
0.47026
0.27475
0.7261
2.0325
3.4764
2.5823
7.1639
P6
LDGV
90-95
0.46835
0.27776
0.0795
2.0153
3.3673
1.5718
6.6904
3.6255
P7
LDGV
90-95
0.43598
0.24653
-0.2787
1.6476
1.7967
2.607
2.986
4.5264
4.7362
P8
LDGV
90-95
0.42397
0.22613
0.8668
1.576
1.4199
1.0315
2.7698
1.4176
3.8634
4.4221
P9
LDGV
90-95
0.41414
0.2177S
0.4251
1.6155
0.9945
0.7619
2.1999
0.108
3.4947
3.5743
5.7316
ť
P10
LDGV
90-95
0.40712
0.2139
-0.4251
1.6685
0.371
1.1081
1.7587
0.4377
2.6567
3.7398
4.0116
3.1506
P11
LDGV
90-95
0.32541
0.07608
Q.4951
1.01
0.6419
0.3247
2.3515
0.3887
1.0667
2.0267
3.9523
1.1292
1.4038
P12
LDGV
90-95
0.27382
0.05866
1.1507
0.9457
0.7041
0.1603
1.9823
-0.7329
1.4374
0.757
2.5555
1.2857
0.8276
3.2666
P13
LDGV
90-95
0.1552
0.0465
1.245
0.7538
1.2793
0.8753
0.6342
0.2174
0.6015
0.6589
2.2733
-0.1902
0.7277
1.3434
2.4018
P14
LDGV
90-95
0.11701
0.02975
1.1371
0.7436
0.97
0.8665
0.7991
0.5198
0.7343
0.5517
1.7367
0,4654
0.6168
1.0241
1.4687
1.7969
P15
LDGV
90-95
0.07219
0.00927
0.8725
0.7153
0.8392
0.9635
0.7945
0.6765
0.7708
0.6798
1.0226
0.5631
0.7013
0.6864
0.8237
1.0439
1.3295
P16
LDGV
90-95
0.06398
0.00808
1.0798
0.7045
0.8792
0.8433
0.7229
0.5834
0.8931
0.5466
0.9541
0.7035
0.7254
0.6265
0.8134
0.9715
1.0589
1.191
P17
LDGV
90-95
0.05809
0.00266
0.7561
0.7292
0.6974
0.8682
0.8458
0.6179
0.6355
0.681
0.8312
0.5502
0.7239
0.822
0.6206
0.9771
0.9638
1.1024
1.5163
P18
LDGV
90-95
0.03251
0.00233
0.6937
0.7144
0.6856
0.9367
0.7529
0.9301
0.619
0.8318
0.7798
0.5858
0.6898
0.6309
0.7014
0.849
0.7786
1.1002
0.83
0.8865
P19
LDGV
90-95
0.00903
0.00112
0.8195
0.7065
0.733
0.7558
0.7072
0.8124
0.7212
0.7331
0.7165
0.7193
0.7121
0.7215
0.7187
0.7075
0.7149
0.7316
0.7374
0.7123
0.7172
IM147 Phase 2 Regression Coefficients, NOx, 1990 to 1995 Model Year LDGV's
Reqression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5
C8
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C16
C19
B11
LDGV
90-95
0.4955
0.16783
2.7648
B12
LDGV
90-95
0.39806
0.10265
1.5279
4.9874
B13
LDGV
90-95
0.21598
0.0764
0.9775
2.036
3.2274
B14
LDGV
90-95
0.16191
0.04667
0.8655
1.4947
2.0303
2.4727
B15
LDGV
90-95
0.09916
0.01709
0.9685
1.262
1.1221
1.4177
1.8321
816
LDGV
90-95
0.08828
0.01425
1.0137
1.1226
1.1413
1.3001
1.4762
1.565
B17
LDGV
90-95
0.07957
0.00619
0.9986
1.1256
1.132
1.322
1.3198
1.5162
2.0499
B18
LDGV
90-95
0.04511
0.00552
0.9273
1.1868
0.9219
1.1679
1.0538
1.5928
1.0608
1.1805
B19
LDGV
90-95
0.01076
0.00217
0.9663
0.9854
0.9735
0.9568
0.9719
1.0002
1.022
0.9682
0.9807
-------
0
1
u>
VO
IM
147 Composite Regression Coefficients, NOx, 1996 and Newer Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4
C5 .
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
PI
LDGV
96+
0.29865
0.37011
11.7738
P2
LDGV
96+
0.22677
0.24297
-3.0327
2.4026
P3
LDGV
96+
0.21591
0.222
0.4545
1.4421
5.448
P4
LDGV
96+
0.21456
0.2173
-0.7302
1.4606
4.3233
3.1071
P5
LDGV
96+
0.20702
0.21364
1.0698
1.2242
3.8494
-0.6109
4.3718
P6
LDGV
96+
0.20746
0.2141
0.6213
1.2154
3.841
0.6842
4.2145
0.9383
P7
LDGV
96+
0.20382
0.20285
2.5242
1.0207
3.6971
-1.4265
3.1047
0.3945
2.683
P8
LDGV
96+
0.19807
0.19482
4.1563
0.9074
3.4911
-2.0343
2.3803
-0.0536
2.1013
3.3361
P9
LDGV
96+
0.19619
0.19096
1.8739
1.0364
2.2721
-1.119
2.141
-0.5391
2.1297
2.5845
3.318
P10
LDGV
98+
0.1935
0.18673
1.493
0.998
2.3525
-1.4944
2.0734
0.242
1.6938
2.8999
1.2192
1.9845
P11
LDGV
96+
0.15206
0.11563
6.6686
0.3811
2.506
-1.1863
0.7057
-0.2193
2.4209
0.8267
1.5013
0.5498
1.1615
P12
LDGV
96+
0.14242
0.09025
7.3264
0.4696
2.0937
-0.2105
0.3797
-0.6436
2.1032
0.8737
0.8521
0.6422
0.984
1.8962
P13
LDGV
96+
0.1117
0.0621
4.8271
0.6696
2.0895
-2.2591
0.9164
Ś0.565
1.4492
-0.1321
0.1274
0.7773
0.8735
0.8257
2.0634
P14
LDGV
96+
0.0781
0.04141
3.4134
0.5916
1.3165
-0.2995
1.3675
-0.2069
1.4165
0.0989
0.2565
0.3993
0.7549
0.6841
1.288
1.631
P15
LDGV
96+
0.05148
0.02187
2.4596
0.6558
1.17
0.6931
0.7565
0.2435
0.7215
0.5341
0.5256
0.6843
0.7209
0.6775
1.1075
0.9069
1.122
P16
LDGV
96+
0.04778
0.01987
3.1395
0.6091
1.203
0.5425
0.9077
0.4201
0.7848
0.5764
0.1403
0.7957
0.7431
0.6929
1.0516
0.9276
0.7894
1.4029
P17
LDGV
96+
0.04525
0.01326
3.3943
0.6187
0.8612
0.7678
1.1673
0.4078
0.7682
0.5486
0.2441
0.6409
0.7404
0.726
0.935
1.0264
0.7278
1.3847
1.368
P18
LDGV
96+
0.02353
0.00022
1.5178
0.6979
0.7651
0.9288
0.6991
0.3426
0.7996
0.6796
0.6094
0.5847
0.7147
0.8429
0.6623
0.895
0.6746
0.9795
0.7813
1.0666
P19
LDGV
96+
0.004
0.00038
0.7646
0.7189
0.7176
0.7276
0.6851
0.7511
0.7279
0.6879
0.7681
0.6817
0.7174
0.7317
0.7291
0.7203
0.7125
0.7336
0.7584
0.7113
0.7181
I
M147 Phase 2 Repression Coefficients, NOx, 1996 and Newer Model Year LDGV's
Regression Coefficients
Segment
Number
RMS
Error
Reg.
Constant
C1
C2
C3
C4 .
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
B11
LDGV
96+
0.20953
0.18322
1.6724
B12
LDGV
96+
0.19448
0.14434
1.3687
2.7382
B13
LDGV
96+
0.155
0.09509
1.1262
1.2948
2.6037
B14
LDGV
96+
0.10659
0.06065
0.9665
0.9898
1.6977
2.2378
B15
LDGV
98+
0.06912
0.03298
0.9526
0.6932
1.4872
1.252
1.5305
816
LDGV
96+
0.0649
0.03049
0.9772
0.9069
1.4267
1.2614
1.1325
1.7334
B17
LDGV
96+
0.06203
0.02204
0.9694
0.9325
1.3118
1.3384
1.0793
1.6717
1.6273
B18
LDGV
96+
0.03172
0.00214
0.9508
1.1071
0.9089
1.1952
0.9329
1.2914
1.0022
1.4675
B19
LDGV
96+
0.00523
0.00111
0.9672
0.6874
0.9855
0.0701
0.8675
0.0838
1.0369
0.6689
0.9757
-------
Appendix D
Excess Emissions Identification for Max CO Cutpoints
-------
Excess Emissions with Fast-Pass, Retest, Fast-Fail Enabled
Max CO Outpoints
(Data Not Normalized for Model Year Distribution)
Excess
Excess HC
Excess
Excess CO
Excess
Excess NO
Vehicle
Model Year
IM240 HC
with fast-
Excess HC
IM240 CO
with fast-
Excess CO
IM240 NOx
with fast-
Excess NOx
Class
Range
(grams)
pass (grams)
Identified
(grams)
pass (grams)
Identified
(grams)
pass (grams)
Identified
LDGV
81-82
0
- 0
N/A
56.74
49.73
87.6%
3.61
3.61
100.0%
83-85
6.13
6.13
100.0%
243.24
237.02
97.4%
3.49
3.49
100.0%
86-89
10.72
10.65
99.3%
301.8
294.27
97.5%
9.92
9.92
100.0%
90-95
8.96
8.68
96.9%
185.26
157.96
85.3%
6.59
6.02
91.4%
96+
0
0
N/A
0
0
N/A
0
0
N/A
ALL
25.81
25.46
98.6%
787.04
738.98
93.9%
23.61
23.04
97.6%
LDGT1
81-85
10.99
8.32
75.7%
605.55
590.71
97.5%
19.02
18.72
98.4%
88-89
5.55
5.13
92.4%
533.37
485.61
91.0%
2.37
2.36
99.6%
90-95
0.42
0
0.0%
1.77
0
0.0%
1.65
1.65
100.0%
96+
0
0
N/A
0
0
N/A
0.43
0.43
100.0%
ALL
16.96
13.45
79.3%
1140.69
1076.32
94.4%
23.47
23.16
98.7%
LDGT2
81-85
15.62
15.62
100.0%
441.44
441.44
100.0%
6.69
6.69
100.0%
86-87
10.39
10.39
100.0%
285.54
285.54
100.0%
4.11
2.91
70.8%
88-95
0.21
0.21
100.0%
26.15
15.16
58.0%
0.68
0.27
39.7%
96+
0
0
N/A
0
0
N/A
0
0
N/A
ALL
26.22
26.22
100.0%
753.13
742.14
98.5%
11.48
9.87
86.0%
Total
ALL
68.99
65.13
94.4%
2680.86
2557.44
95.4%
58.56
56.07
95.7%
D-l
-------
Excess Emissions with Fast-pass Enabled
Max CO Cutpoints
(Data Not Normalized for Model Year Distribution)
Excess
Excess HC
Excess
Excess CO
Excess
Excess NO
Vehicle
Model Year
IM240 HC
with fast-
Excess HC
IM240 CO
with fast-
Excess CO
IM240 Nox
with fast-
Excess NOx
Class
Range
(grams)
pass (grams)
Identified
(grams)
pass (grams)
Identified
(grams)
pass (grams)
Identified
LDGV
81-82
0
0
N/A
56.74
49.73
87.6%
3.61
3.61
100.0%
83-85
6.13
6.13
100.0%
243.24
237.02
97.4%
3.49
3.49
100.0%
86-89
10.72
10.65
99.3%
301.8
294.27
97.5%
9.92
9.92
100.0%
90-95
8.96
8.68
96.9%
185.26
157.96
85.3%
6.59
6.02
91.4%
96+
0
0
N/A
0
0
N/A
0
0
N/A
ALL
25.81
25.46
98.6%
787.04
738.98
93.9%
23.61
23.04
97.6%
LDGT1
81-85
10.99
8.32
75.7%
605.55
560.92
92.6%
19.02
18.72
98.4%
88-89
5.55 .
5.13
92.4%
533.37
485.61
91.0%
2.37
1.69
71.3%
90-95
0.42
0
0.0%
1.77
0
0.0%
1.65
1.65
100.0%
96+
0
0
N/A
0
0
N/A
0.43
0
0.0%
ALL
16.96
13.45
79.3%
1140.69
1046.53
91.7%
23.47
22.06
94.0%
LDGT2
81-85
15.62
14.47
92.6%
441.44
441.44
100.0%
6.69
5.72
85.5%
86-87
10.39
10.39
100.0%
285.54
285.54
100.0%
4.11
2.31
56.2%
88-95
0.21
0.21
100.0%
26.15
15.16
58.0%
0.68
0.27
39.7%
96+
0
0
N/A
0
0
N/A
0
0
N/A
ALL
26.22
25.07
95.6%
753.13
742.14
98.5%
11.48
8.3
72.3%
Total
ALL
68.99
63.98
92.7%
2680.86
2527.65
94.3%
58.56
53.4
91.2%
D-2
-------
Excess Emissions with Fast-pass Disabled
Max CO Cutpoints
(Data Not Normalized for Model Year Distribution)
Excess
Excess HC
Excess
Excess CO
Excess
Excess NO
Vehicle
Model Year
IM240 HC
w/o fast-
Excess HC
IM240 CO
w/o fast-
Excess CO
IM240 Nox
w/o fast-
Excess NOx
Class
Range
(grams)
pass (grams)
Identified
(grams)
pass (grams)
Identified
(grams)
pass (grams)
Identified
LDGV
81-82
10.99
8.32
75.7%
605.55
603.26
99.6%
19.02
18.72
98.4%
83-85
5.55
5.55
100.0%
533.37
527.14
98.8%
2.37
1.69
71.3%
86-89
0.42
0.42
. 100.0%
1.77
0
0.0%
1.65
1.65
100.0%
90-95
0
0
N/A
0
0
N/A
0.43
0.43
100.0%
96+
16.96
14.29
84.3%
1140.69
1130.4
99.1%
23.47
22.49
95.8%
ALL
15.62
14.47
92.6%
441.44
441.44
100.0%
6.69
5.72
85.5%
LDGT1
81-85
10.39
10.39
100.0%
285.54
285.54
100.0%
4.11
2.31
56.2%
88-89
0.21
0.21
100.0%
26.15
26.15
100.0%
0.68
0.38
55.9%
90-95
0
0
N/A
0
0
N/A
0
0
N/A
96+
26.22
25.07
95.6%
753.13
753.13
100.0%
11.48
8.41
73.3%
ALL
0
0
N/A
56.74
49.73
87.6%
3.61
3.61
100.0%
LDGT2
81-85
6.13
6.13
100.0%
243.24
237.02
97.4%
3.49
3.49
100.0%
86-87
10.72
10.72
100.0%
301.8
301.22
99.8%
9.92
9.92
100.0%
88-95
8.96
8.87
99.0%
185.26
160
86.4%
6.59
6.02
91.4%
96+
0
0
N/A
0
0
N/A
0
0
N/A
ALL
25.81
25.72
99.7% Ś
787.04
747.97
95.0%
23.61
23.04
97.6%
Total
ALL
68.99
65.08
94.3%
2680.86
. 2631.5
98.2%
58.56
53.94
92.1%
D-3
-------
Appendix E
Second-by-Second CPP Variation Limits
-------
IM147 CUMULATIVE POSITIVE POWER (CPP) VARIATION OUTPOINTS
IM147 REFERENCE DATA
TIME SPEED CPP
(sec)
(mph)
fmph2/sec}
0
0.0
0.00
1
0.0
0.00
2
0.0
0.00
3
0.0
0.00
4
0.0
0.00
5
3.3
10.89
6
6.6
43.56
7
9.9
98.01
8
13.2
174.24
9
16.5
272.25
10
19.8
392.04
11
22.2
492.84
12
24.3
590.49
13
25.8
665.64
14
26.4
696.96
15
25.7
696.96
16
25.1
696.96
17
24.7
696.96
18
25.2
721.91
19
25.4
732.03
20
27.2
826.71
21
26.5
826.71
22
24.0
826.71
23
22.7
826.71
24
19.4
826.71
25
17.7
826.71
26
17.2
826.71
27
18.1
858.48
28
18.6
876.83
29
20.0
930.87
30
20.7
959.36
31
21.7
1001.76
32
22.4
1032.63
33
22.5
1037.12
34
22.1
1037.12
35
21.5
1037.12
36
20.9
1037.12
37
20.4
1037.12
38
19.8
1037.12
39
17.0
1037.12
40
17.1
1040.53
41
15.8
1040.53
42
15.8
1040.53
43
17.7
1104.18
44
19.8
1182.93
45
21.6
1257.45
POWER VARIATION OUTPOINTS (mph2/sec)
"BASE" MULT. VARYING CPP LIMITS
DELTA FACTOR DELTA LQM HIGH
69.7
3.500
244.0
715.31
1,203.41
72.8
3.424
249.3
752.44
1,251.08
75.1
3.386
254.2
778.47
1,286.79
75.4
3.348
252.4
784.71
1,289.53
75.4
3.348
252.4
784.71
1,289.53
75.4
3.348
252.4
784.71
1,289.53
75.4
3.348
252.4
784.71
1,289.53
75.4
3.348
252.4
784.71
1,289.53
75.4
3.348
252.4
784.71
1,289.53
75.4
3.348
252.4
784.71
1,289.53
75.6
3.311
250.4
790.16
1,290.90
75.6
3.311
250.4
790.16
1,290.90
75.6
3.311
250.4
790.16
1,290.90
80.3
3.273
262.7
841.53
1,366.83
86.0
3.235
278.1
904.80
1,461.06
91.4
3.197
292.2
965.27
1,549.63
E-1
-------
IM147 CUMULATIVE POSITIVE POWER (CPP) VARIATION OUTPOINTS
IM147 REFERENCE DATA POWER VARIATION OUTPOINTS (mph2/sec)
TIME
SPEED
CPP
"BASE-
MULT.
VARYING
CPP LIMITS
(sec)
(mph)
(mph2/sec')
DELTA
FACTOR
DELTA
LOW
HIGH
46
22.2
1283.73
93.3
3.159
294.8
988.97
1,578.49
47
24.5
1391.14
101.1
3.121
315.6
1,075.55
1,706.73
48
24.7
1400.98
101.8
3.083
314.0
1,087.02
1,714.94
49
24.8
1405.93
102.2
3.045
311.2
1,094.73
1,717.13
50
24.7
1405.93
102.2
3.045
311.2
1,094.73
1,717.13
51
24.6
1405.93
102.2
3.045
311.2
1,094.73
1,717.13
52
24.6
1405.93
102.2
3.045
311.2
1,094.73
1,717.13
53
25.1
1430.78
104.0
3.008
312.8
1,118.02
1,743.54
54
25.6
1456.13
105.8
2.970
314.3
1,141.83
1,770.43
55
25.7
1461.26
106.2
2.932
311.4
1,149.88
1,772.64
56
25.4
1461.26
106.2
2.932
311.4
1,149.88
1,772.64
57
24.9
1461.26
106.2
2.932
311.4
1,149.88
1,772.64
58
25.0
1466.25
106.6
2.894
308.4
1,157.84
1,774.66
59
25.4
1486.41
108.0
2.856
308.6
1,177.85
1,794.97
60
26.0
1517.25
110.3
2.818
310.8
1,206.47
1,828.03
61
26.0
1517.25
110.3
2.818
310.8
1,206.47
1,828.03
62
25.7
1517.25
110.3
2.818
310.8
1,206.47
1,828.03
63
26.1
1537.97
111.8
2.780
310.8
1,227.18
1,848.76
64
26.7
1569.65
114.1
2.742
312.9
1,256.78
1,882.52
65
27.3
1602.05
116.4
2.705
314.9
1,287.13
1,916.97
66
30.5
1787.01
129.9
2.667
346.4
1,440.65
2,133.37
67
33.5
1979.01
143.8
2.629
378.1
1,600.89
2,357.13
68
36.2
2167.20
157.5
2.591
408.1
1,759.09
2,575.31
69
37.3
2248.05
163.4
2.553
417.1
1,830.90
2,665.20
70
39.3
2401.25
174.5
2.515
439.0
1,962.29
2,840.21
71
40.5
2497.01
181.5
2.477
. 449.6
2,047.41
2,946.61
72
42.1
2629.17
191.1
2.439
466.2
2,163.02
3,095.32
73
43.5
2749.01
199.8
2.402
479.8
2,269.18
3,228.84
74
45.1
2890.77
210.1
2.364
496.6
2,394.15
3,387.39
75
46.0
2972.76
216.1
2.326
502.5
2,470.24
3,475.28
76
46.8
3047.00
221.5
2.288
506.7
2,540.32
3,553.68
77
47.5
3113.01
226.3
2.250
509.1
2,603.92
3,622.10
78
47.5
3113.01
226.3
2.250
509.1
2,603.92
3,622*10
79
47.3
3113.01
226.3
2.250
509.1
2,603.92
3,622.10
80
47.2
3113.01
226.3
2.250
509.1
2,603.92
3,622.10
81
47.2
3113.01
226.3
2.250
509.1
2,603.92
3,622.10
82
47.4
3131.93
227.6
2.212
503.6
2,628.37
3,635.49
83
47.9
3179.58
231.1
2.174
502.5
2,677.11
3,682.05
84
48.5
3237.42
235.3
2.136
502.7
2,734.73
3,740.11
85
49.1
3295.98
239.6
2.098
502.7
2,793.27
3,798.69
86
49.5
3335.42
242.4
2.061
499.5
2,835.88
3,834.96
87
50.0
3385.17
246.0
2.023
497.7
2,887.49
3,882.85
88
50.6
3445.53
250.4
1.985
497.1
2,948.47
3,942.59
89
51.0
3486.17
253.4
1.947
493.3
2,992.84
3,979.50
90
51.5
3537.42
257.1
1.909
490.8
3,046.58
4,028.26
91
52.2
3610.01
262.4
1.871
491.0
3,119.03
4,100.99
E-2
-------
IM147 CUMULATIVE POSITIVE POWER (CPP) VARIATION OUTPOINTS
IM147 REFERENCE DATA POWER VARIATION OUTPOINTS (mph2/sec)
TIME
SPEED
CPP
"BASE"
MULT.
VARYING
CPP LIMITS
(sec)
(mph)
(mph2/sec)
DELTA
FACTOR
DELTA
LOW
HIGH
92
53.2
3715.41
270.0
1.833
495.1
3,220.33
4,210.49
93
54.1
3811.98
277.1
1.795
497.5
3,314.53
4,309.43
94
54.6
3866.33
281.0
1.758
493.9
3,372.43
4,360.23
95
54.9
3899.18
283.4
1.720
487.4
3,411.82
4,386.54
96
55.0
3910.17
284.2
1.682
478.0
3,432.20
4,388.14
97
54.9
3910.17
284.2
1.682
478.0
3,432.20
4,388.14
98
54.6
3910.17
284.2
1.682
478.0
3,432.20
4,388.14
99
54.6
3910.17
284.2
1.682
478.0
3,432.20
4,388.14
100
54.8
3932.05
285.8
1.644
469.8
3,462.23
4,401.87
101
55.1
3965.02
288.2
1.606
462.8
3,502.17
4,427.87
102
55.5
4009.26
291.4
1.568
457.0
3,552.29
4,466.23
103
55.7
4031.50
293.0
1.530
448.4
3,583.09
4,479.91
104
56.1
4076.22
296.3
1.492
442.2
3,634.06
4,518.38
105
56.3
4098.70
297.9
1.455
433.3
3,665.39
4,532.01
106
56.6
4132.57
300.4
1.417
425.5
3,707.05
4,558.09
107
56.7
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
108
56.7
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
109
56.3
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
110
56.0
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
111
55.0
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
112
53.4
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
113
51.6
4143.90
301.2
1.379
415.3
3,728.63
4,559.17
114
51.8
4164.58
302.7
1.341
405.9
3,758.70
4,570.46
115
52.1
4195.75
305.0
1.303
397.4
3,798.38
4,593.12
116
52.5
4237.59
308.0
1.265
389.7
3,847.93
4,627.25
117
53.0
4290.34
311.8
1.227
382.7
3,907.64
4,673.04
118
53.5
4343.59
315.7
1.189
375.5
3,968.10
4,719.08
119
54.0
4397.34
319.6
1.152
368.0
4,029.31
4,765.37
120
54.9
4495.35
326.7
1.114
363.9
4,131.49
4,859.21
121
55.4
4550.50
330.7
1.076
355.8
4,194.70
4,906.30
122
55.6
4572.70
332.4
1.038
344.9
4,227.76
4,917.64
123
56.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
124
56.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
125
55.8
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
126
55.2
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
127
54.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
128
53.6
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
129
52.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
130
51.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
131
50.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
132
48.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
133
44.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
134
41.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
135
37.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
136
34.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
137
30.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
-------
IM147 CUMULATIVE POSITIVE POWER (CPP) VARIATION OUTPOINTS
IM147 REFERENCE DATA POWER VARIATION OUTPOINTS (mph2/sec)
TIME
SPEED
CPP
"BASE-
MULT.
VARYING
CPP LIMITS
(sec)
(mph)
(mph2/sec)
DELTA
FACTOR
DELTA
LOW
HIGH
138
27.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
139
23.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
140
20.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
141
16.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
142
13.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
143
9.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
144
6.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
145
2.5
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
146
0.0
4617.34
335.6
1.000
335.6
4,281.74
4,952.94
Cycle Sums
4617.34
1.000
4,281.74
4,952.94
E-4
-------
Appendix F
Regression Summaries
-------
Regression Summary - Composite HC, IM240 to IM147
Model Years 1981-1985
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.981278
R Square 0.962907
Adjusted R 0.962636
Square
Standard 0.331332
Error
Observations 139
ANOVA
df
SS
MS
F
Significance F
Regression
1
390.4283
390.4283
3556.441
6.9E-100
Residual
137
15.03995
0.109781
Total
138
405.4682
Coefficients
Standard
Error
tStat
P-vaiue
Lower 95% Upper 95%
Intercept
0.110694
0.036049
3.070666
0.002576
0.03941 0.181979
IM240
0.896629
0.015035
59.6359
6.9E-100
0.866899 0.92636
IM240 Line Fit Plot
~ IM147
Predicted IM147
F-l
-------
Regression Summary - Composite CO, IM240 to IM147
Model Years 1981-1985
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.989621
R Square 0.979349
Adjusted R 0.979198
Square
Standard 5.150236
Error
Observations 139
ANOVA
df
SS
MS
F
Significance F
Regression
1
172333.7
172333.7
6497.045
2.6E-117
Residual
137
3633.916
26.52493
Total
138
175967.6
Coefficients
Standard
Error
tStat
P-value
Lower 95% Upper 95%
Intercept
0.858255
0.536573
1.599511
0.112011
-0.20278 1.919292
IM240
1.020463
0.01266
80.60425
2.6E-117
0.995428 1.045497
IM240 Line Fit Plot
IM147
Predicted IM147
IM240
F-2
-------
Regression Summary - Composite NOx, IM240 to IM147
Model Years 1981-1985
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.988936
R Square 0.977995
Adjusted R 0.977835
Square
Standard 0.322696
Error
Observations 139
ANOVA
. df_ SS MS F Significance F
Regression 1 634.0613 634.0613 6088.969 2E-115
Residual 137 14.26619 0.104133
Total 138 648.3275
Coefficients Standard t Stat P-value Lower 95% Upper 95%
Error
Intercept 0.085613 0.045469 1.882899 0.061834 -0.0043 0.175525
IM240 1.065128 0.01365 78.03185 2E-115 1.038136 1.09212
F-3
-------
Regression Summary - Composite HC, IM240 to IM147
Model Years 1986-1989
All Vehicle Types
SUMMARY OUTPUT
Repression Statistics
Multiple R 0.987927
R Square 0.975999
Adjusted R 0.975876
Square
Standard 0.189037
Error
Observations 198
ANOVA
df
SS
MS
F
Significance F
Regression
1
284.8199
284.8199
7970.309
1.1E-160
Residual
196
7.004082
0.035735
Total
197
291.824
Coefficients
Standard
Error
tStat
P-value
Lower 95% Upper 95%
Intercept
0.056509
0.015868
3.561206
0.000463
0.025215 0,087802
IM240
0.933646
0.010458
89.27659
1.1E-160
0.913021 0.95427
IM240 Line Fit Plot
IM240
~ IM147
Predicted IM147
F-4
-------
Regression Summary - Composite CO, IM240 to IM147
Model Years 1986-1989
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.984568
R Square 0.969374
Adjusted R 0.969218
Square
Standard 5.123058
Error
Observations 198
ANOVA
df
SS
MS
F
Significance F
Regression
1
162823.5
162823.5
6203.812
2.5E-150
Residual
196
5144.161
26.24572
Total
197
167967.7
Coefficients
Standard
Error
tStat
P-value
Lower 95% Upper 95%
Intercept
1.679632
0.403976
4.157756
4.8E-05
0.882936 2.476328
IM240
0.939067
0.011922
78.76428
2.5E-150
0.915554 0.96258
IM240 Line Fit Plot
350
300
250
Ł 200
T
^ 150
100
50
0
IM240
F-5
-------
Regression Summary - Composite NOx, IM240 to IM147
Model Years 1986-1989
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.977553
R Square 0.955611
Adjusted R 0.955384
Square
Standard 0.298767
Error
Observations 198
ANOVA
df
SS
MS
F
Significance F
Regression
. 1
376.638
376.638
4219.491
1.6E-134
Residual
196
17.49525
0.089261
Total
197
394.1332
Coefficients
Standard
tStat
P-value
Lower 95% Upper 95%
Error
Intercept
0.058971
0.035732
1.650379
0.100467
-0.0115 0.129439
IM240
1.077932
0.016594
64.95761
1.6E-134
1.045206 1.110659
IM240 Line Fit Plot
~ IM147
Predicted IM147
F-6
-------
Regression Summary - Composite HC, IM240 to IM147
Model Years 1990-1995
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.974081
R Square 0.948834
Adjusted R 0.948723
Square
Standard 0.100407
Error
Observations 464
ANOVA
df_ SS MS F Significance F
Regression 1 86.37186 86.37186 8567.38 2.3E-300
Residual 462 4.657643 0.010081
Total 463 91.0295
Coefficients Standard t Stat P-value Lower 95% Upper 95%
Error
Intercept 0.026672 0.00526 5.071064 5.74E-07 0.016336 0.037007
IM240 0.963839 0.010413 92.56014 2.3E-300 0.943376 0.984302
IM240 Line Fit Plot
IM240
~ IM147
Predicted IM147
F-7
-------
Regression Summary - Composite CO, IM240 to IM147
Model Years 1990-1995
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.916596
R Square 0.840149
Adjusted R 0.839803
Square
Standard 3.35827
Error
Observations 464
ANOVA
df
SS
MS
F
Significance F
Regression
1
27384.98
27384.98
2428.183
4.6E-186
Residual
462
5210.424
11.27797
Total
463
32595.41
Coefficients
Standard
tStat
P-vaiue
Lower 95%
Upper 95%
Error
Intercept
0.392486
0.180702
2.172
0.030364
0.037385
0.747586
IM240
1.037836
0.021061
49.2766
4.6E-186
0.996448
1.079224
IM240 Line Fit Plot
140
0 50 100 150
IM240
F-8
-------
Regression Summary - Composite NOx, IM240 to IM147
Model Years 1990-1995
All Vehicle Types
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.958073
R Square 0.917905
Adjusted R 0.917727
Square
Standard 0.306825
Error
Observations 464
ANOVA
df
SS
MS
F
Significance F
Regression
1
486.2964
486.2964
5165.597
6.2E-253
Residual
462
43.49332
0.094141
Total
463
529.7897
Coefficients
Standard
tStat
P-value
Lower 95% Upper 95%
Error
Intercept
0.048771
0.020544
2.374
0.018004
0.0084 0.089143
IM240
1.102698
0.015343
71.87209
6.2E-253
1.072548 1.132848
IM240 Line Fit Plot
~ IM147
Predicted IM147
0 2 4 6 8
IM240
F-9
-------
Appendix G
Fleet Distribution Data
-------
Fleet Distribution Data
Vehicle Distribution Data - 9 months 2% random sample (initial test)
Data Collected between 7/1/97 and 3/31/98
LDGV LDT1 LDT2
Model
Year
# Tests
% of Fleet
# Tests
% of Fleet
# Tests
% of Fleet
1981
102
1.02%
31
0.31%
13
0.13%
1982
138
1.38%
33
0.33%
18
0.18%
1983
153
1.52%
52
0.52%
17
0.17%
1984
318
3.17%
80
0.80%
33
0.33%
1985
372
3.71%
129
1.29%
29
0.29%
1986
414
4.13%
172
1.71%
48
0.48%
1987
450
4.49%
167
1.66%
40
0.40%
1988
540
5.38%
205
2.04%
54
0.54%
1989
540
5.38%
235
2.34%
62
0.62%
1990
512
5.10%
183
1.82%
44
0.44%
1991
542
5.40%
229
2.28%
40
0.40%
1992
533
5.31%
201
2.00%
74
0.74%
1993
559
5.57%
270
2.69%
62
0.62%
1994
654
6.52%
306
3.05%
110
1.10%
1995
673
6.71%
310
3.09%
131
1.31%
1996
100
1.00%
45
0.45%
10
0.10%
1997
0
0.00%
0
0.00%
0
0.00%
1998
0
0.00%
0
0.00%
0
0.00%
1999
0
0.00%
0
0.00%
0
0.00%
2000
0
0.00%
0
0.00%
0
0.00%
Total 6600 2648 785
Distribution Between Vehicle Types
LDGV
LDT1
LDT2
Total 10033
# Tests % of Fleet
6600
65.8%
2648
26.4%
785
7.8%
G-l
------- |