High Altitude Vehicular Emission Control Program, Volume IV: Analysis of Experimental Results: Final Report


HIGH ALTITUDE VEHICULAR EMISSION CONTROL PROGRAM
VOLUME IV. ANALYSIS OF EXPERIMENTAL RESULTS
PREPARED FOR:
STATE OF COLORADO	ENVIRONMENTAL PROTECTION AGENCY
DEPARTMENT OF HEALTH	REGION VIII
DENVER, COLORADO 80220	DENVER, COLORADO 80203
TDiMf/ environmental

-------
6\^
c.\
DISCLAIMER
This report was developed for the Environmental Protection Agency
and the State of Colorado by TRW Transportation and Environmental
Operations under contract number 62-02-0048. The contents of this
report are reproduced herein as received from the contractor. The
opinions, findings, and conclusions are those of TRW and not necessarily
those of the sponsoring agencies. Mention of company or product names
does not constitute endorsement by either the Environmental Protection
Agency or the State of Colorado.
The results and conclusions presented are based on the data developed
from the experimental test program (conducted by Automotive Testing
Laboratories). The extent to which these data are not representative
of the vehicle population in the Denver area, however, could have a
significant impact on the resultant conclusions and recommendations.
^$hn,ca|RUKG,0PJ 8
999 18ttSteS^80C-L
Denver, CO eofof 500

-------
PREFACE
This report, "High Altitude Vehicular Emission Control Program,"
consists of seven volumes. Listed in the following are the subtitles
given for each volume:
o Volume I - Executive Suiunary, Final Report, January
1974.
e Volume II - Experimental Characterization of Idle
Inspection, Exhaust Control Retrofit and Mandatory
Engine Maintenance, Final Report, December 1973.
9 Volume III - Impact of Altitude on Vehicular Exhaust
Emissions, Final Report, December, 1973.
• Volume IV - Analysis of Experimental Results, Final
Report, December 1973.
o Volume V - Development of Techniques, Criteria and
Standards to Implement a Vehicle Inspection,
Maintenance and Modification Program, Final Report,
December 1973.
a Volume VI - The Data Base, Final Report, January 1974.
9 Volume VII - Experimental Characterization of Vehicular
Emission and Engine Deterioration, Final Report, June
1974.
The first volume summarizes the general objectives, approach and
results of the study. The second volume presents a detailed description
of the experimental programs conducted to define the data base. Volume
III reports the methods and analysis used in developing the basic
relationships between mass emissions and altitude. A quantitative
analysis of the results from the experimental program is presented in
Volume IV. The fifth volume provides an analysis of the techniques
and criteria required in estab 1ishing a vehicle emission control program
for the Denver area. The actual data base developed from the experimental

-------
program is given in Volume IV. Lastly, Volume VI reports the results
of the six month deterioration program.
The work presented herein is the product of a joint effort by
several consulting firms. Automotive Testing Laboratories (ATL) was
responsible for the design and implementation of the basic experiments.
TRW provided the data management and analysis of the experimental
results. Olsen Laboratories evaluated the feasibility of conducting

-------
ACKNOWLEDGEMENTS
TRW would like to acknowledge the efforts extended by the Colorado
State Department of Health and Region VIII of the Environmental Pro-
tection Agency.
The contribution of Messrs. Don Sorrels and Frank Taylor and
Ms. Lindsay Tipton of the Colorado State Department of Health were of
particular significance.
Mr. Dale M. Wells of Region VIII served as Project Officer.

-------
TABLE OF CONTENTS
Page
1.0 SUMMARY AND CONCLUSIONS	1-1
2.0 INTRODUCTION	2-1
3.0 METHODS OF ANALYSIS	3-1
3.1	Data Management	3-1
3.2	Statistical Analysis 		3-5
4.0 SURVEY OF ENGINE AND EMISSION STATES 		4-1
4.1	Engine Parameter Survey 		4-1
4.2	Key Mode Emission Survey	4-10
4.3	1975 CVS Mass Emissions Survey	4-21
5.0 IDLE INSPECTION PROGRAM
5.1	Idle Inspection Accuracy	5-1
5.2	Vehicular Exhaust Emission Standards 		5-8
5.3	Program Effectiveness 		5-27
6.0 EVALUATION OF M/NDATORY ENGINE MAINTENANCE 		6-1
7.0 EFFECT OF IDLE ENGINE ADJUSTMENTS ON EXHAUST EMISSIONS . .	7-1
8.0 ASSESSMENT OF SEA LEVEL AND HIGH ALTITUDE RETROFIT
DEVICES	8-1
8.1	Analysis of Sea Level Retrofit Devices 	 8-1
8.2	Analysis of High Altitude Devices	8-4
9.0 IMPACT OF TEST RESULTS ON TRANSPORTATION CONTROL PLAN . . 9-1
9.1	Emission Forecasts for 1977 	 9-1
9.2	Retrofit for Older Vehicles 	 9-4

-------
LIST OF TABLES
Table	Page
3-1	Data Management Approach	3-2
4-1	Summary of Engine Parameters and Emission Species
Surveyed	4-2
4-2 Survey of Engine States by Model Year	4-4
4-3 Survey of Engine States by Vehicle Manufacturer .	.	4-5
4-4 Survey of Engine States by Engine Displacement 		4-6
4-5 Survey of Engine States by Vehicle Weight 		4-7
4-6 Survey of Key Mode Emissions by Model Year	4-17
4-7 Survey of Key Mode Emissions by Vehicle Manufacturer . .	4-18
4-8 Survey of Key Mode Emissions By Engine Displacement . .	4-19
4-9 Survey of Key Mode Emissions by Vehicle Weight ....	4-20
4-10 Survey of CVS Mass Emissions by Model Year	4-45
4-11 Survey of CVS Mass Emissions by Vehicle Manufacturer . .	4-46
4-12 Survey of CVS Mass Emissions by Cubic Displacement . .	4-47
4-13	Survey of CVS Mass Emissions by Vehicle Weight ....	4-48
5-1	Idle Inspection Pass/Fail Criteria 		5-2
5-2 Analysis of Idle Inspection Accuracy for Pre-1968
Vehicles	5-4
5-3 Analysis of Idle Inspection Accuracy for Post-1967
Vehicles	5-5
5-4 Analaysis of Measurement Accuracy Between Laboratory
and Garage Equipment 		5-7
5-5 Rejection Fractions Versus Idle Mode Cutpoints
for Pre-Control1ed Vehicles 		5_9
5-6 Rejection Fractions Versus Idle Mode Cutpoints for
Controlled Vehicles 		5-10

-------
LIST OF TABLES - continued
Table	Page
5-7 Rejection Fractions Versus Multiple Idle Key Mode
Cutpoints for Pre-Controlled Vehicles 		5-13
5-8 Rejection Fractions Versus Multiple Idle Key Mode
Cutpoints for Controlled Vehicles 		5-14
5-9 Emission Cutpoints for Vehicles at 50°o Rejection Rate .	5-15
5-10 Analysis of Variance Results for an Idle HC Emission
Inspection	5-18
5-11 Analysis of Variance Results for an Idle CO Emission
Inspection	5-19
5-12 Analysis of Variance Results for an Idle NO Emission
Inspection	x .	5-20
5-13 Comparison of Passed and Failed Vehicular Emissions .	5-26
5-14 Comparison of Emissions for Failed Vehicles Before and
After Maintenance	5-27
5-15 Comparison of Emissions for Total Population Before and
After Maintenance	5-28
5-16	Idle Inspection Maintenance Costs 		5-29
6-1	Analysis of Mandatory Maintenance 		6-3
6-2	Mandatory Maintenance Costs 		6-4
7-1	Engine Adjustments Tested by the Experiment ....	7-4
7-2 Experimental Test Settings	7-5
7-3 Comparison of Adjustment Effects for HC	7-6
7-4 Comparison of Adjustment Effects for CO	7-8
7-5 Comparison of Adjustment Effects for N0x	7-10
7-6 Engine Adjustment Influence Coefficients for HC .	7-12
7-7 Engine Adjustment Influence Coefficients for CO . .	7-13
7-8 Engine Adjustment Influence Coefficients for NO . .	7-14
X

-------
LIST OF TABLES - continued
Table	Page
8-1 Comparison of Sea Level Retrofit Systems for Pre-1968
Vehicles	, .	8-5
8-2 Sea Level Retrofit Cost Analysis for Pre-1968 Vehicles . 8-6
8-3 Comparison of Sea Level Retrofit Systems for Post-1967
Vehicles	8-7
8-4 Sea Level Retrofit Cost Analysis for Post-1967
Vehicles	8-8
8-5 Sea Level Retrofit Cost-Effectiveness Ratios of the
Major Competing Devices for Post-1 967 Cars	8-9
8-6 High Altitude Retrofit Performance Data 	 8-10
8-7	High Altitude Retrofit Cost Analysis	8-11
9-1	Assumed and Measured Reductions for Transportation
Control Plan	9-2
9-2 Comparison of Effectiveness Between Original and Revised
Transportation Control Plan for 1 977 	 9-3

-------
LIST OF FIGURES
Fi qure	P_acje
4-1 Air/Fuel Ratio Frequency Distribution for Pre-Contro11ed
Vehicles	4-8
4-2 Air/Fuel Ratio Frequency Distribution for Controlled
Vehi cl es	.4-9
4-3	Frequency Plot of RPM for Pre-Control1ed Vehicles .	.4-11
4-4	Frequency Plot of RPM for Controlled Vehicles . .	. 4-12
4-5	Frequency Plot of Timing for Pre-Control1ed Vehicles	. 4-13
4-6	Frequency Plot of Timing for Controlled Vehicles . .	. 4-14
4-7	Frequency Plot of Dwell for Pre-Control1ed Vehicles .	. 4-15
4-8	Frequency Plot of Dwell for Controlled Vehicles . .	. 4-16
4-9 Idle HC Frequency Distribution for Pre-ControI led
Vehicles	4-23
4-10 Idle HC Frequency Distribution for Controlled
Vehicles	4-24
4-11 Idle CO Frequency Distribution for Pre-ControI 1ed
Vehicles	4-2b
4-12 Idle CO Frequency Distribution for Controlled
Vehicles	4-26
4-13 Idle NO Frequency Distribution for Pre-Control1ed
Vehicles	4-27
4-14 Idle NO Frequency Distribution for Controlled
Vehicles	4-2T.
4-15 Low Cruise IIC Frequency Distribution for Pre-Contro 1 1 ed
Vehicles	4-29
4-16 Low Cruise HC Frequency Distribution for Controlled
Vehicles	4-30
4-17 Low Cruise CO Frequency Distribution for Pre-Controlled
Vehicles	4-31

-------
LIST OF FIGURES - continued
Fi gure Page
4-18 Low Cruise CO Frequency Distribution for Controlled
Vehicles 4-32
4-19 Low Cruise NO Frequency Distribution for Pre-Controlled
Vehicles x 4-33
4-20 Low Cruise NO Frequency Distribution for Controlled
Vehicles * 4-3
-------
LIST OF FIGURES - continued
Figure Page
4-33 NO Mass Emissions Frequency Distribution for Total
Vehicle Population 4-51
5-1 Statistical Significance of Partitioning Vehicle Popu-
lation by Engine Displacement for Idle Emission Testing . 5-21
5-2 Statistical Significance of Partitioning Vehicle Popu-
lation by Engine Displacement for CVS Mass Emission
Testing 5-22
9-1 Predicted Emission Reduction Effectiveness for Pre-1968
Vehicle with Retrofit 9-6
-------
1.0 SUMMARY AND CONCLUSIONS
This document provides a quantitative assessment of the results of the
High Altitude Emission Test Program. The primary objectives of this pro-
gram were twofold. First, to determine the impact of a series of
experimental procedures on vehicular exhaust emissions, and second,
to identify those specific procedures that could prove cost effective
in reducing vehicular emissions for the Denver area. The basic experiments
performed during the course of the test program are listed below:
9 Idle Inspection and Maintenance
e Mandatory Engine Maintenance
• High Altitude Engine Adjustment
• Sea Level Retrofit
e High Altitude Retrofit
e Engine Deterioration
Embodied within this list are a large number of subexperiments which
were conducted as part of the overall research design. The analysis
presented herein focuses attention on those emission reduction pro-
cedures that tend to be most cost-effective. The resultant control
measures are then evaluated in light of the proposed Colorado Trans-
portation Control Plan.
The significant conclusions and recommendations that have emerged
from the study are given in the following:
Idle Inspection and Maintenance
e The vehicle survey showed that idle CO, timing, air
cleaner and ignition misfire are most in need of
repair. The inspection and repair of these engine
parameters should be included as part of the inspection/
maintenance program.
-------
• An annual idle inspection and maintenance program, con-
sisting of the measurement of idle HC and idle CO followed,
as necessary, by corrective maintenance, appears marginally
attractive as a control measure (6.6% reduction for HC,
4.6% reduction for CO and 0% for NO ). These estimates
include the effects of engine deterioration per the EPA
schedule. On a semi-annual basis the estimated emission
reduction effectiveness is 9.9% for HC, 6.9% for CO and
0% for NOx.
• An annual program of inspection and maintenance appears
to be most cost effective. The final determination on
the frequency of inspection should be based on the results
from the deterioration program.
o The costs associated with the idle program ($4.05 for
the inspection and $10.35 for maintenance) are well
within an acceptable range as revealed by a recent
survey of the motoring public.
o The two areas of greatest uncertainty in the program
involve the accuracy of the inspection process and the
effectiveness of engine repair. An increase in service
organization effectiveness could lead to significant
improvements in emission reductions.
9 Results from the experimental program indicate that a
state owned and operated inspection system may be
more effective than the corresponding licensed garage
approach.
• A minimum inspection program could form the basis for
insuring the operatabi1ity of advanced emission control
systems on post 1975 vehicles.
a An independent emissions surveillance system should be
developed as part of the inspection program. This
system would provide a feedback mechanism for measuring
the actual performance of the inspection and maintenance
program. The sample selected for this program should be
representative of the inuse vehicle population.
Mandatory Engine Maintenance
« A mandatory engine maintenance program does not appear
very cost-effective particularly in view of the high
cost of engine tune-ups (average cost of $49.10) and
the relatively low emission reduction performance (9.6%
for HC, 4.9% for CO, and 3.5% for N0V).
High Altitude Engine Adjustments
9 The adjustment of idle CO and timing were found to have
the largest influence on CO emission levels. (Statistically
significant at the 99% confidence level).
-------
e The adjustment of vacuum choke kick was found to have
a modest impact on CO emissions (statistically signifi-
cant at 90% confidence). The adjustment of rpm had
little effect on CO emissions (less than a 50% level of
confidence).
« None of the four adjustment parameters were found to
have a significant impact on either HC or NO emissions
(less than a 60% level of confidence in all cases).
a A program designed to adjust idle CO and timing for all
older vehicles in the population could yield significant
emission reductions at reasonable costs.
Sea Level Retrofit
e A combination of air bleed (AIR) and exhaust gas recir-
culation (EGR) retrofit systems for pre-control
vehicles yielded the greastest balanced emission
reductions for HC and CO (221 and 21%, respectively).
Similarly, a combination of vacuum spark advance dis-
connect and air bleed produced the greatest NO
emission reductions (47%).
e An AIR/EGR system for pre-control1ed vehicles would
yield a population weighted reduction of 3.3% for HC
and 6.1% for CO by 1977.
o The average installation cost for the AIR/EGR system was
approximately $37.00. It was found to have a negilible
effect on gasoline mileage.
e An oxidizing catalytic retrofit system for controlled
vehicles provided the largest emission reductions for
HC and CO (72% and 84%, respectively). The AIR/EGR
system yielded average emission reductions of 17% for
HC and 48% for CO. The impact of these systems on
gasoline mileage was insignificant.
o The high costs associated with the catalytic system
tends to preclude its application for the general
vehicle population. Consideration should be given,
however, to a catalytic retrofit program for fleet
vehicles.
e The AIR/EGR system appears more cost effective and
should be considered as a benchmark for comparing
other retrofit systems for controlled vehicles.
Additional experimental testing of the AIR/EGR is
necessary, however, because of the very small sample
sizes used in the present test plan (the average
number of vehicles tested was five (5)).
-------
High Altitude Retrofit
o The results from the high altitude retrofit experiments
revealed no statistically significant emission reductions
for either HC or CO. The one exception involved a subset
of cars belonging to the Chrysler family (Dodges). In
this one case the emission reductions for, HC and CO (26%
and 54%, respectively) were statistically significant.
o The installation of these systems had an adverse effect
of NO emissions, ranging from an increase of 16% for
Fords to an increase of 84% for Chryslers.
o A high altitude retrofit program does not appear as a
viable alternative because of the lower emission re-
duction potential for HC and CO and relative high costs
of installation.
Engine Deterioration
o Engine deterioration could have a significant impact
on the overall performance of each of the control
strategies studied (especially the idle inspection
program). The results from the ongoing deterioration
program should be considered before finalizing the
vehicular emission control plan for Denver.
Transportation Control Plan
0 The estimates used in the Colorado Transportation
Control Plan appear somewhat more optimistic than
those developed from the experimental test program
(particularly for the high altitude retrofit program).
o The application of the experimental data to the
existing Transportation Control Plan indicates
a 10% to 20% reduction in effectiveness for HC
and CO control, respectively.
• It appears, based on these findings, that additional
transportation controls may be necessary to meet the
minimum air quality standards. Additional emission
reductions for the vehicle population could be
achieved from one or more of the alternatives outlined
above.
0 The use of an AIR/EGR system instead of the high altitude
kit for 1968-1974 controlled vehicles could help achieve
the reductions required to meet the national standards in
the Denver AQCR.
• It is recommended that the existing Colorado Transportation
Control Plan be reviewed and updated in light of the evidence
gained from the experimental test program.
-------
2.0 INTRODUCTION
This report presents a detailed analysis on the results from the
experimental test program. The primary objective of this program was
to characterize the effectiveness and costs of several vehicular
emission control alternatives. The data and information derived from
these experiments is intended to hc.11 [j shape the final form of the
Colorado Transportation Control Plan. Particular program objectives
were the following:
9 Develop an emissions inventory baseline for vehicles
operating within the Denver AQCR
o Identify the most promising control alternatives for
additional testing and evaluation
o Evaluate the impact of the measured results on the
proposed transportation control plan.
The original transportation control plan, submitted by the State
to the EPA on May 1973, called for a three phase approach for controlling
emissions from light-duty vehicles. The data used in estimating the
effectiveness of the proposed plan was based on extrapolated performance
estimates developed at sea level. Recent evidence has indicated that
vehicles operating at high altitude (above 4000 feet) have substantially
different emission characteristics than vehicles operating at lower
altitudes. The EPA has acknowledged this fact and lias proposed rules
for developing certification procedures of new vehicles intended
for initial sale at high altitude.* This emphasis on the impact of
altitude on vehicular emissions has helped identify the need for a
Federal Register, Vol. 38, No. 197 (October 12, 1973).
-------
high altitude vehicular emissions data base.
The specific experiments performed as part of the current program
were designed to accomplish the following:
Engine and Emission States Survey
A survey was performed to ascertain the extent and fre-
quency of engine malfunctions and maladjustments occurring
in the vehicle population. The survey also included the
measurement of key mode and CVS emission levels. This
data served as a baseline for contrasting the effectiveness
of the various control alternatives, e.g. sea level retrofit.
The survey consisted of 300 vehicles which are representative
of the vehicle population distribution within the AQCR.
Results from the survey helped identify those particular
engine parameters that are cost-effective to repair.
/
Idle Inspection and Maintenance Program
This experimental program was designed to evaluate the
costs and effectiveness of a service garage administered
idle inspection and maintenance program. The program
consisted of inspecting the emission levels (HC and CO)
at idle and performing specific engine adjustments (air
fuel, rpm, timing) on those vehicles failing the inspection.
Additional engine repair was undertaken, as necessary, to
achieve the pre-established emission criterion. Again, a
fleet of 300 vehicles was used in measuring the effective-
ness of this program. Cost data were recorded on the
various phases of the inspection and maintenance process.
Mandatory Engine Maintenance Program
A program of mandatory engine maintenance involves the
periodic repair or replacement of specific emission
oriented engine components, e.g. air cleaner. Such an
approach avoids many of the problems associated with
inspection by concentrating exclusively on effective
engine repair. A mandatory program was simulated experi-
mentally by tuning up approximately 150 vehicles. Mass
emission measurements were performed both prior and
after maintenance. The costs associated with mandatory
repair were also collected and have been included in the
analysis.
Idle Engine Adjustment Program
An experimentally designed program was undertaken to
determine the influence of selected engine adjustments
on exhaust emissions. Four specific engine adjustments
were evaluated -- air/fuel ratio, rpm, basic timing and
-------
vacuum choke kick. These four adjustments were identified
as being inexpensive and easy to modify. The experiments
were performed using 25 cars which were chosen to represent
the vehicle population.
Sea Level and High Altitude Retrofit Program
A series of tests were undertaken to measure the relative
effectiveness of a number of leading retrofit systems. The
systems were partitioned in three categories: 1) sea level
systems for pre-1968 vehicles, 2) sea level systems for
1968-1974 vehicles, and 3) high altitude kits for 1968-1974
vehicles. Specific devices tested included air bleed,
exhaust gas recirculation, vacuum spark advance disconnect,
catalytic converter and various combinations of the above.
Additionally, the four major automobile manufacturers pro-
vided high altitude kits. The sample size for these
experiments range from 3 to 48. Cost data for each system
was also collected and evaluated.
The data developed from each of these experiments was processed using
TRW's Data Management System. The process results have been merged
into a single internally consistent data base. An analysis of these
results indicates that the assumed performance values used in the
proposed Transportation Control Plan are optimistic and; therefore,
additional vehicular control may be necessary. These results under-
score the impact of altitude on vehicular emissions.
The preceeding section (Section 1.0) has provided a summary of
the conclusions and recommendations synthesized from the study. This
section (Section 2.0) focuses on the background of the test program
and presents a summary of the basic experimental programs. Section 3.0
discusses the methods of analysis used in the evaluation including an
overview on TRW's Data Management System and statistical regression
methodologies. Section 4.0 summarizes the results from the vehicle
engine and emission survey. These results include frequency histogram
plots of the major variables that were measured. A detailed analysis
-------
of the idle inspection program is provided in Section 5.0. This
analysis examines the implications of inspection accuracy and main-
tenance effectiveness on the viability of this control approach.
Section 6.0 presents an evaluation of the cost effectiveness of
mandatory engine maintenance. In Section 7.0, the effects of idle
engine adjustments on exhaust emissions are studied and a set of
response influence coefficients are developed. Section 8.0 discusses
the effectiveness of both sea level and high altitude retrofit systems
as a means for reducing exhaust emissions from the inuse vehicle fleet.
Finally, in Section 9.0 the impact of the experimental test results
are evaluated with respect to the proposed transportation control plan.
Specific analyses are presented on the actual effectiveness of retro-
fitting the pre-controlled segment of the vehicle population.
-------
3.0 METHODS OF ANALYSIS
The principal reason for utilizing computer technology in this
effort was the need to process very large amounts of data and to perform
complex operations on this data.
With the use of computerized processing methods the data
generated by the numerous testing activities of this program could
be accessed and sorted, and concise and illuminating results prepared
rapidly and accurately.
Presented in this section are the data management and statistical
analysis techniques used in handling and developing trends from the
data base.
3.1 DATA MANAGEMENT
The nature and extent of the experimental test program required the
application of a total data management system. This section presents
a detailed description of the system utilized in support of the program.
The data management system, developed as part of the Coordinating Research
Council CAPE-13 Project, provides comprehensive capability for evaluating
multidimensional emission test.programs.
Figure 3-1 presents a schematic overview of the TRW Data Management
System as related to the current emission test program. This schematic
shows the division of work between TRW and ATL with respect to the
development and analysis of the basic data. ATL was responsible for
reducing the CVS mass emission data and preparing the raw data files
for the inspection, maintenance, and retrofit and adjustment experiments.
TRW performed the following basic data management activities:
-------
Table 3-1
Data Management Approach
-------
o edit test data
o sort test data
s store test data
e develop yraphic presentations
o retrieve test data
® statistical analysis
e compute basic statistics
These activities are also depicted in Figure 3-1. A brief description of
of each of these data management functions is presented in the subsequent
paragraphs.
Data Editing
A multi-purpose editing program (EDIT) was utilized to reduce potential
transportation and procedural errors between the test site (Denver)
and TRW's computing facilities (Space Park). Data listings obtained from
from the raw keypunch cards were analyzed and corrected using the remote
TRW timeshare system. This procedures allows review by botli the
testing engineers and data management specialists. The program EDIT was
utilized not only in examining specific data point entries , but also in
updating parameter values on the file record. This latter capability
was important as refinements to the raw data were implemented.
Data Storage
The total data has been stored on a permanent magnetic tape. This
approach serves two purposes: 1) provide a back-up for all of the data
developed during the experimental testing, and 2) permit a more cost
effective method of handling the data during the collection and editing
phase. A copy of the magnetic tape will be made available to the State
of Colorado and the EPA upon request. A working file will also be
maintained on the CDC 6500 disk pack for future analysis.
Graphic Presentation
A major part of the data analysis activity is to summarize
-------
information in the form of histograms. Two programs are used to perform
the retrieval and plotting functions. CETP is used to retrieve data
records for the requested plots, sort the data and write a data input
file for the plotting routine. HISTM provides histogram frequency
plots of the data. The plots are drawn by a Cal Comp plotter from a
tape written by HISTM.
Retrieving and Sorting Test Data
The principle data handling program in the data management system
(CETP-- Colorado Emission Test Program) serves as the basic
interface between the data base and the other software. This program
retrieves the selected data from disk storage and sorts it by a number
of classification systems. The data can be culled in the following ways:
• the total population • sort by weight group
e sort by manufacturer o sort by age group
t sort by make within a a sort by PASS/FAIL at idle
particular manufacturer inspection
• sort by engine size group
Additionally the data can also be sorted by:
o CVS
• key modes
t engine parameters
• costs
• vehicle characteristics , e.g. vehicle weight.
Basic Statistics
The program CETP also computes certain simple but essential
properties of the data including: standard deviations, t-values,
and degrees of freedom. This information provides a quick look capability for
-------
analyzing the raw data in addition to the more involved techniques
discussed in the next section.
3.2 STATISTICAL ANALYSIS
Several independent statistical packages were used in processing
the data relevant to this study.
A major computer program, one developed specifically for the Idle
Adjustment Program, is designated ANOV. It performs an analysis of
variance for an experimental design. In order to test the effects on
emissions produced by adjusting the settings of four engine parameters
and their interactions it was necessary to make several simplifying
assumptions. These are:
9 Choke kick has no second-order interactions with
idle CO, RPM and timing.
o Third and fourth order interactions are negligible.
e Parameter settings and emission levels are distributed
normally in the general population.
From these assumptions, a one-half fractional factorial experiment
involving only eight (8) tests per vehicle could be designed. Each
vehicle is subjected to all eight settings in random order and the
emissions at each setting are measured.
The ANOV program using this data, estimates two basic factors.
First, the variance of the emission levels between settings indicates
how much effect the different settings have on emission levels and,
second, the variance between vehicles within each setting provides a
standard of normalization for determining the net real effect. The
ratio of these two factors is compared to a critical "F-value". If the
"F-score" exceeds the F-value the effect is said to be statistically
significant, i.e. the effect produced by changing that setting is significant.
-------
Another kind of analysis of variance was also used. This one
(called Wilson's ANOVA) is a TRW system routine. It computes a confidence
level and probability that indicates whether or not a population divided
into several groups yields statistically different emissions.That is, the
resultant partition provides .1 measure of potential differences. This pro-
gram was used in determi tn rig I.tie mer'il. of separating t.he vehicle
copulation into sub-di vi s 1 oir> for key mode emission analysis. 'I he
results of this analysis are -j i von in sec t i on 1j . 0 .
Ttie third major statistical package was a program designed to
perform three main types of regression analysis; ordinary least squares,
two-stage least squares, and limited information single equation
analysis. Regression equations were derived with this program for the
high altitude experimental program ( see Volume III).
The major data handling program CLTP also computes such minor
statistics as means, standard deviations and t-scorcs. The t-scores
in turn are used to determine the significance of the given means and
standard deviations.
-------
4.0 SURVEY OF ENGINE AND EMISSION STATES
This section presents a survey on the existing state of engine
repair and emission levels for vehicles operating mi the Denver area.
The survey provides detailed diagnostic information on the extent of the engine
maladjustment and malfunction and on the corresponding emission levels
for HC, CO and NO . This information is of crucial importance in
establishing a benchmark for comparing the relative effectiveness of
alternative vehicular control strategies. Specifically, the survey
included the engine and emission parameters shown in Table 4-1.
4.1 ENGINE PARAMETER SURVEY
The surveyed engine parameters have been classified into three
sets to reflect diflerences in engine maladjustments and malfunctions:
1) Idle adjustments (idle fuel to air ratio, idle speed,
basi r. timing , and dwel 1) .
2) Ignition components affecting misfire or spark advance
(NO con trol ).
3) Induction system components affecting fuel to air ratio.
Table 4-2 shows the results of the survey partitioned by vehicle con-
trol type. The engine parameters are separated into two basic
categories -- distributed and nondistributed. Distributed parameters
are those of a continuous nature whereas nondistributed parameters
are either operating or failed. For example, the survey found 88
percent of all PCV values operating in the total population.
One interesting observation from this data is the relatively high
degree of incipient misfire found in the population. On the average
14 percent of the vehicles surveyed were found misfiring which is
-------
Table 4-1
SUMMARY OF ENGINE PARAMETERS AND
EMISSION SPECIES SURVEYED
Engine Parameters
Idle Subsystem Ignition Subsystem
o Idle Fuel/Air Ratio o Misfire
• Idle RPM o NO Control
/\
o Basic Timing
e Dwel1
Emission Species
Key Mode
Idle HC
Idle CO
Idle NO
x
HC @ 2500 RPM
CO @ 2500 RPM
HC @ Low Cruise
CO @ Low Cruise
NO @ Low Cruise
x
HC @ High Cruise
CO @ High Cruise
NO @ High Cruise
x
CVS-1975
HC Mass
CO Mass
NO Mass
X
Induction Subsystem
0 PCV Valve
8 Air Cleaner
s Air Pump
e Choke
-------
approximately four times larger than detected at sea level.* Since
misfire has a large effect on hydrocarbons, the incorporation of an
ignition tune-up as part of the inspection program may yield effective
results.
The results for basic timing are among the more interesting of
the distributed parameters in that the mean recorded value is near
manufacture specification and yet there is a large variance around the
mean value. This would tend to indicate that the adjustment of timing
to specification for a segment of tne population could produce significant
emission reduction results.
Tables 4-3 through 4-6 shows similar results with the data
partitioned by vehicle manufacturer, engine displacement and vehicle
weight, respectively. The variability in engine states revealed by
these different classification schemes indicates that these factors
must be taken into consideration as part of any inspection and maintenance
program.
Another perspective on the variability of engine states can be
gained from reviewing the histograms for the continuous parameters
shown in Figures 4-1 through 4-8. Figure 4-1 shows the relative dis-
tribution of idle CO for pre-controlled vehicles, i.e., pre 1968.
These results show a rather flat response between 2.4% and 9.6% with a
rapid drop off beyond this limit. The slight increase at 12%
reflects an accumulation for these recorded values beyond that point.
A similar trend is also shown for the control vehicles, i.e., post 1967.
* TRW, Inc., A Study of Mandatory Engine Maintenance for Reducing Vehicle
Exhaust Emissions, Vol. 4., July 1973.
-------
Table 4-2
SURVEY OF ENGINE STATES BY MODEL YEAR
DISTRIBUTED PARAMETERS
.p-
i
¦f*
PARAMETER
~TbLE"CO"(% Vol)"
IDLE RPj (RPM)
TIMING (DEG)
DWELL
SAMPLE SIZE!
(<195?>
MEAN SJ
5.533 2.75?
3 5.391 13 2. H27
-.300 -3 .JC9
-.7-91 ^ . 3 3 C
(110)
(1968*)
MEAN S D
<~.325 2.639
-11.195 130.^63
. 9 «<+ <+. 7 •?
- . 0 <~ 7 4 . j 9 •}
(19C)
AVERAGE
MEAN SD
4.767 2.737
5.887 132.737
0.513 6.230
-0.320 4.082
(300)
MONDISTRIBUTED PARAMETERS
(PERCENT)
Y'ES NO- " >ES MO YES NO
PCV 57 13 8 ^ 11 88 12
_SAMPL£ !IZJLi _ _ _ (91) (16 8) (259)
AIR CLNsT" """ ~ oK " i,6 63 37 59 41
SAMPLE SIZE! (110) (190) (300)
CHOKE l_ 100 G 97 3
SAMPLE SIZE: " (13 3) " ( 1 39) (297)
AIR PUMP 10 0 0 10G 0 100 0
SAMPLE SIZE: ( 1) ( 1^> (15)
MISFIRE " "" "" ' 18 " 3 2 " "" 19 81 19 81
SAMPLE SIZE*. (110) ( 1 9u ) (300)
NOX 0 0 8 0 2C 80 20
-------
PARAMETER
• IDLE CO (% Vol )
fQLE'S'PH (RPM)
TIMING (DEG)
DWELL
sample size?
PCV
SAMPLE size:
AIR~ CLN?
SAMPLE SIZE:
CHOKE
SAMPLE SIZE:
AIR PUM°
SAMPLE SIZE:
MISFIRE
SAMPLE size:
NOX
SAMPLE SIZE:
Table 4-3
SURVEY OF ENGINE STATES BY VEHICLE MANUFACTURER
DISTRIBUTED PARAMETERS
.U3 2.464
41 . 420 147. 8 7b
2 . 700 4. 3 15
-.640 4. 2 '~ 6
( 60)
( A MC )
MEAN SD
5.03 3 3 . 3 ?3
¦ 3 2 . 0 0 C 9 9.532
-.101 4 . 55 7
-.203 3.521
( 10 )
(FOREIGN)
MEAN SD
4.376 2.557
-1.515 198.165
1 .576 b.5 91
-1.727 5.611
( 3 3)
YE S
8 3
5 3
I
9 7
10 0
1 7
6 9
(120 )
(128 )
(128 )
( 10)
(128 )
NO
13
i*7
3
0
8 3
3 1
NONOISTRIB'JTEO PARAMETERS
(PERCENT)
YES
87
( 75)
( 79)
99
( 78)
10C
( 2)
( 73)
18
VO
13
33
1
0
SO
13
YES
96
50
94
100
16
ia j
( 49)
( 50 )
( 49)
( 3)
( 5 0)
NO
4
j
3'.
I
YrS
73
80
IOC
G
it]
IOC
( 13)
3)
( 6)
9 )
10 )
10 >
C )
1 0 )
11
NO
2?
? 1
7u
Y c S
3 3
35
103
0
24
s n
5 )
J 3 )
32 )
C>
33)
2)
NO
17
15
0
0
7 6
-------
HOOF
inLc no {% Vol)
TDLF RPM(RPM)
TIMTsr, (DEG)
DWELL
SflMPlr STZF:
"CV

SAMPLC
^IZEl
AIR CLNR

SAMPLE
SIZF!
CHOKF.

S A MPLr
^ I ZE!
A T? °UHP

S0M°LE
SI7FS
MTS^I^F

S A MDL E
SI7F :
NOX

?AMPLr
SI7FI
Table 4-4
SURVEY OF ENGINE STATES BY ENGINE DISPLACEMENT
DISTRIBUTED PARAMETERS
( <1C3)
MFAN
(100-199)
MF AN SO
¦5.
'.71'
k. ^23
3.17=;
182.997 -21.800 17?. 009
l.i»6H 6. 8
b.uza
( ?U)
.6*0 *.121
•1.2«f0 3.U79
( ?5)
(700-299)
M-flN SO
•5.23? 2.co2
21. 3 S 2 13l». 85U
-.I**7 6.967
-.70 6 3.6<*1
f 68 )
(300-399)
MF AN "O
<~.6*0 2.67^
9.363 116.?U i*
.36") 6.r N T )
YFS NO YES NO YFS MO YES NO YFS " ~ NO'
67 37 101 3 *«~ 16 99 11 91 9
( 3) ( 16) ( 61) (133) ( <~ 6)
8° 13 M "5? 61 t»9 r> 9 !fl 63 37
( 2«t) ( ?5> ( 6 73 IB T2 2T H 2
( ?L) ( ?5) ( 6«) (137) ( ^6)
P 100 100 " 10 ~i ^
-------
Table 4-5
SURVEY OF ENGINE STATES BY VEHICLE WEIGHT
DISTRIBUTED PARAMETERS
ldOO-2799
PARAMETER
IDLE CO (% Vol )
ni_£ 3PM (RPM)
"Timing (deg)
dwell
S^HPLC' S 17 E:
MEAN
k • 3 36
SO
2.793
- 3 . U 7 3 17 6. 779
.571+ 5.918
— 1 • 3 tJ 5.156
( t»6 )
210 0-37 99
1E4N
<~ . 778
sn
2.631
>. 3 5 (+ 132.3 69
.39^ 5.675
-.737 "3.522
(127)
3800-^799
MEAN SO
<+. 16 7 2. lol
1. 0 3 6 1GQ. 339
«(+6i+ 7.2^3
.'+ 6 <+ 3.136
(112)
t,8Q -:-57 99
SO
MEAN
3 . 6 U /
2. 5^6
-3U.be7 1^1.503
l.^OU 3.066
.13"? 5.203
( 15)
PCV
AIR CLN*
CHOKE
AIR PUMP
MISFIRE
NO X
sample size:
SAMPLE SIZF!
SIMPLE 3 IZr:
"sample "size:
SAMPLE SIZE:
SAMPLE SI7
-------
PARAMETER. SURVEY IDLE CO (< 1968 J
o
CvJ
o~
<£>
o"
-F*
I
CO
CM
>- •-
z
UJ
ZD
a
LU
OD
cc •.
u_o
o
o"
o
o
°Q .00
2 .40
4.80 1 .20
PERCENT
9.60
12.00
-------
PARAMETER SURVEY IDLE CO (1968+)
-p»
i
to
>-
(_)
UJ°
a
LlJ
cc
Li_
o
o
o
o
•-
^.00 2-20 4.40 6.60 8-80 1 1-00
PERCENT
-------
The results for idle speed (Figures 4-3 and 4-4), basic timing
(Figures 4-5 and 4-6) and dwell (Figures 4-7 and 4-8), show somewhat
different results. For the most part, each of these parameters are
distributed around specification each exhibiting a rather large
variability. The plots for these variables tend to resemble a Gaussian
or normal distribution which has been found typical in other studies.*
4.2 KEY MODE EMISSION SURVEY
A similar series of analyses were performed for the key mode data
collected in the survey. Tables 4-6 through 4-9 show means and standard
deviations for the recorded key modes by model year, vehicle manufacture,
engine displacement and vehicle weight, respectively. This data was
recorded using the laboratory measuring equipment and can serve as a
baseline for evaluating the data recorded at the maintenance garages.
The abbreviations HC 2500 and CO 2500 represent HC and CO measurements
taken at 2500 rpm. The histogram plots were developed herein using data
taken from the Sun laboratory equipment.
Section 5.4 will provide a more definitive analysis on the need
for developing individual idle emission standards based on partitioning
the population. In summary, the analysis shows that individual standards
should be established for idle HC and idle CO by vehicle age i.e.,
pre 1968 vehicles, 1968-1970 vehicles, and post 1970 vehicles. The
analysis produced inconclusive results with respect to the other
classification schemes, e.g., engine cubic displacment.
*
TRW, Inc., A Study of Mandatory Engine Maintenance for Reducing Vehicle
Exhaust Emissions, Vol. 4., July 1973.
-------
PflRRMETEF? SURVEY IDLE RPM ( < 1 968 J
-42 .10
-24 .36
-6 .62
11.12
. 86
4€ .60
DEVIATION FROM SPEC *10
-------
in
cn
o
CO
PARflMEXE-R SURVEY IDLE RPM (1968+)
in
CNI
o
o
CM
~Z.
LU
a-
Qd •
Li_ o"
o
o"
m
o
o"
o
o
-26 - 10
-9 .86
OEViq
6 .38
ION --FROM
•22 .62
\
33 .^6
o o
1 0
R i.
-------
PARAMETER SURVEY TIMING C< 1968 J
-p»
i
o
in
o
o
on
>- •-
U>°
z:
LU
ZD
C3 _
U_°
o
o"
o
o
"-33.00 -23.60 -14.20 -4-BO 4-60
DEVIATION FROM SPEC
14 .00
-------
PARAMETER SURVEY TIMING (1968+)
o
m
-P»
I
o
o
o
CO
>- •.
U°
ZD
C3.
UJ
CM
CtL
U.O
o
o
"V13.00 -5.40 2.20 9.00 H-4 0
DEVIATION FROM SPEC
=¥
25.00
-------
PARAMETER SURVEY DWELL (<1968)
-P=»
I
cn
O
o'
o
co
>-
O
ZO
ll ICM
=>o
a
LlJ
cc
u_
o
o
o
-16 .00
-11 .20
6 .40
-1 .60
3 .20
B .00
DEVIATION FROM SPEC
-------
PARAMETER S.L1RVE.Y. DWELL (1.968+ )
o
in
o
o'
o
en
>- •.
C_»°
LU
ZD
a_
LU
(Nl
cd
Li_o
o
o"
o
o
I
3-
VM.00 -0.20 -2.40 3.40 9.20 15-00
DEVIATION FROM SPEC
-------
Table 4-6
SURVEY OF KEY MODE EMISSIONS BY MODEL YEAR
Mode
HC Idle(ppm)
HC 2500 (ppm)
HC Low Cruise (ppm)
HC High Cruise(ppm)
CO Idle (%)
CO 2500 (%)
CO Low Cruise (%)
CO High Cruise (%)
N0X Idle (ppm)
N0x Low Cruise (ppm)
N0x High Cruise(ppm)
Sample Size
Pre 1968
Mean SD
925.818 506.133
556.596 572.596
667.809 354.670
621.445 360.146
6.287 2.934
5.071 2.558
4.430 2.561
5.525 3.006
65.736 56.649
1145.736 773.669
1192.591 916.762
110
Post 1968
Mean SD
612.979 451.036
318.189 491.139
512.611 266.441
491.826 292.372
4.815 2.902
2.524 3.021
2.171 1.972
2.967 2.412
97.705 191.558
1542.837 838.791
1870.237 1095.910
190
Average
Mean SD
727.687
494.754
405.453
534.037
569.517
310.391
539.353
324.395
5.355
2.995
3.458
3.10S
2.999
2.458
3.904
2.915
85.983
156.852
1396.500
836.689
1621.767
1082.845
-------
Table 4-7
SURVEY OF KEY MODE EMISSIONS BY VEHICLE MANUFACTURER
GM FORD CHRYSLER AMC FOREIGN

Mode
Mean
SD
Mean
SD
Mean
SO
Mean
SD
Mean
SD
HC Idle (ppm)
741.812
483.621
698.076
454.836
665.980
454.268
680.700
299.684
851.575
698.577
HC 2500(ppm)
420.742
573.634
436.532
579.924
378.800
518.615
426.500
426.413
305.758
247.570
HC Low Cruise(ppm)
598.648
391.254
592.468
287.475
530.780
167.424
498.300
181.002
481.848
150.093
HC High Cruise(ppm)
569.359
395.171
544.899
268.691
515.140
309.730
527.300
217.608
450.030
120.187
CO Idle (%)
5.397
3.063
4.953
2.915
5.889
2.782
6.169
4.147
5.097
2.833
CO 2500 (%)
3.490
2.780
3.135
2.830
3.457
3.241
3.085
2.750
4.222
4.585
CO Low Cruise{%)
3.295
2.671
2.395
2.000
3.293
2.551
1 .981
1.334
3.164
2 A'-/'.
CO High Cruise (%)
4.319
3.258
3.028
2.395
4.680
2.322
5.208
3.942
2.825
2.227
NOx Idle (ppm)
77.641
67.369
85.797
57.961
125.260
359.857
77.500
66.130
61.848
29.07;
N0x Low Cruise(ppm)l
192.953
726.212
1665.251
949.525
1525.740
929.850
1601.900
661.439
1284.606
638.0 Y)
N0x High Cruise(ppm)392.141
965.330
1861.139
1081.560
1241.200
811.354
1302.500
1302.186
2612.758
1139.
-------
Table 4-8
SURVEY OF KEY MODE EMISSIONS BY ENGINE DISPLACEMENT
<100
Mode
Mean
SD
100-199
Mean SD
200-299
Mean SD
300-399
Mean SD
400+
Mean SD
HC Idle(ppm)
1043.292
767.756
680.880
405.494
723.015
484.342
745.752
473.734
541.565
345.191
HC 2500(ppm)
356.875
242.051
300.400
233.871
491.618
535.231
419.387
601.791
319.022
536. Mi
HC Low Cruise(ppm)
508.208
153.038
514.160
187.199
607.485
305.610
607.453
372.934
462.478
167.679
HC High Cruise(ppm)
460.583
135.881
522.800
170.840
583.971
293.519
573.409
404.470
422.065
165.670
CO Idle (%)
5.941
2.889
5.167
3.287
5.893
2.968
5.228
2.895
4.733
3.167
CO 2500{%)
5.358
4.971
3.833
3.926
4.191
3.154
2.879
2.162
2.903
3.210
CO Low Cruise(%)
3.457
2.085
3.254
2.640
3.702
2.824
2.752
2.414
2.321
1. / A':
CO High Cruise(^)
2.827
1.703
5.153
3.592
5.342
3.045
3.534
2.798
2.767
2.070
NOx Idle(ppm)
54.417
22.587
164.640
507.042
73.603
54.207
82.949
63.731
87.093
72.21 fj
NOx Low Cruise(ppm)
1203.458
537.070
1287.080
749.247
1339.103
876.337
1493.599
910.878
1352.348
701.041
NO High Cruise
X t „\
2537.792
1005.319
1546.600
1392.380
1133.574
923.666
1664.372
1035.819
1779.478
941.161
Sample Size
24
25
68
137
-------
Table 4-9
SURVEY OF KEY MODE EMISSIONS BY VEHICLE WEIGHT
Mode
1800-
Mean
2799
SD
2800-
Mean
3799
SD
3800-
Mean
4799
SD
4800-
Mean
5799
SD
HC Idle(ppm)
846.478
651.434
686.331
420.566
764.411
506.208
439.333
280.167
HC 2500 (ppm)
368.587
393.992
365.551
455.696
495.277
671.217
185.667
125.668
HC Low Cruise (ppm)
513.739
165.022
562.654
255.461
615.304
407.217
456.800
164.105
HC High Cruise(ppm)
473.326
140.136
532.354
233.590
592.536
450.614
404.000
162.189
CO Idle (%)
5.478
2.963
5.332
2.858
5.506
3.156
4.039
2.685
CO 2500 (%)
4.435
4.568
3.139
2.700
3.548
2.858
2.489
1 .987
CO Low Cruise(%)
3.313
2.468
2.860
2.465
3.032
2.565
2.978
1 .440
CO High Cruise(%)
3.171
2.374
4.057
3.067
4.057
2.965
3.717
2.610
NO Idle (ppm)
60.043
28.749
97.724
229.559
82.920
72.061
89.000
51.586
NOx Low Cruise(ppm)1241.630
772.581
1478.835
874.562
1389.116
831.471
1229.467
704.826
NO High Cruise
2219.152
1190.991
1501.008
1013.391
1542.643
1070.079
1403.000
869.143
(ppm)
-------
Figures 4-9 through 4-30 show frequency distributions for the key
mode data for precontrolled and controlled vehicles. This information
is extremely important in developing effective exhaust emission standards.
With the exception of idle CO, these distributions tend to resemble the
standard log normal function (a distribution that is shoved to the left
with an extended tail to the right). Of particular significance is the
data recorded for cruise hydrocarbons (Figures 4-15, 4-16, 4-21 and
4-22). The bimodal distribution reveals the presence of
extremely high emitters which is normally caused by incipient misfire.
In fact, the percentage of vehicles exhibiting high hydrocarbon emission
corresponds reasonably well with that measured for vehicles with misfire.
One interesting note is that HC 2500 also tended to indicate the presence
of misfire. This ability to potentially detect misfire of 2500 rpm
clearly should be considered in the formulation of the idle inspection
program.
4.3 1975 CVS MASS EMISSIONS SURVEY
The measurement of CVS mass emissions for HC, CO and NO represented
X
the last major element of the survey program. Establishing the level of
mass emissions for the vehicle population is of critical importance for
determining actual atmospheric loadings. Tables 4-10 through 4-13
present recorded mass emissions data by model year, vehicle manufacturer,
engine cubic displacement and vehicle weight, respectively.
Estimates of CO^ emission levels are also given along with the
normal emission species (HC, CO and NO ). The substantial difference
in recorded values between precontrolled and controlled vehicles for
-------
HC and CO mass is significant and, therefore, the two subpopulations
should be treated separately.*
A set of HC, CO and NO mass emission frequency distributions for
A
the total population are given in Figures 31 through 33, respectively.
The distributions show a similar trend to those recorded for the key
mode emissions. These plots also tend to confirm the relatively small
number of high emitting vehicles. Repair of these vehicles on the
other hand will yield disproportionately larger emission
reductions due basically to their larger values. An effective program
of vehicle inspection/maintenance should be designed towards detecting
and repairing these high emitters.
~
A more detailed analysis on the statistical properties of these
measurements is given in Section 5.4.
-------
-1
! KEY r:.0Ci ;UflVLf idle HC (< 1 968)
!T> '
c:
o
CM
I
ro
OJ
>-
LJ°
z:
_u
ZD
c «
UJ2
C£ •.
Li-O
o
•—I
o"
LO
o
o
o
o
69-22 1 1 4.4 4 159.66
ppn
2 0 4 . £ 8 250 • 1C
* 1 C '
^4 .00
-------
KEY MODE SURVEY IDLE HC (1968+)
o
in
o
4*
I
ro
o
CO
>- '-I
21
IxJ
ZD
C3
m -j
Li_ °
o
.. 4
O
^.00
50 .02
100
PPM
0 4
150 .06
=V
200
10l
08
250.10
-------
KEY MODE SURVEY IDLE CO (-
o°
z:
LU
r\3
en
O
O
o
o
^.00 2.60 5.20 1.B0
PERCENT
10.40
13 -00
-------
KEY MODE SURVEY IDLE CO (1968+)
ro

O
CM
*—4
o'
CM
>- •.
z:
LU
Z)
a
LU
03
Cd
Ll°
o
o"
o
o
^•00 1.60 5.20 1-80
PERCENT
-n
10.40
13 .00
-------
KEY MODE SURVEY IDLE NQX C < 1968 J
o
o
o
cn
>- •-
o
lO
o
UJ
ro
O
O
O
O
—I < 1 \
89-60 174.20 258-80 343.40 428,00
PPM
-------
KEY MODE SURVEY IDLE NOX (1968+
o
o
o
00
o
o
UD
>- -A
<-J°
Z
LU
Z3
C3
cc: .J
U.o
o
CvJ
O
O
o
I .00
52 .62
104.24
PPM
155 .86
207
10l
,48
="1
259 .10
-------
o
ID
KEY MODE SURVEY LGH CR HC K1968)
ci
in
i
ro
O
TT
o"
>-
<_>
Zo
LU V~>
=>o*
a
LU
cr
o
CM
o
o"
o
o
?6•40 71.14 115-88
PPTl
160 .€2
T
205 .36
* 1 0
250 . 10
-------
% KEY MODE SURVEY LOW CR HC ( 1968+ !
o
o
m
3
c_J j
Zo !
Ljcn
=3 o
o
UJ
Cd
u_
o
CM
O
o
17.10
63-70
110 .30^
PPM

156.90
—i ^
203.50

250.10
*10'
-------
I
CO
SURVEY LOW CR CO U 1968)
^.00 2'.80 5'.60 8.40 l'l .20 14-00
PERCENT
-------
KEY MODE SURVEY LOW CR CO (1968+)
CO
I
U>
ro
>-
LJ
2IO
LuT00
—' o
o
UJ
en
o
o
1=1=1=
=)
12 .00
^.00
2-40
4.80 T
PERCENT
20
9 .60
-------
If
CM
KEY MODE SURVEY LOW CR NOX (< 1968 J
o
C\J
to
o"
>-
u
ZCM
LU *"*
=>o"
a
LU
en
u_
oo
o
o"
o
o"
o
o
°S . 30
81 -70
1
58.10 234-50 310.90 387-30
PPM
« 1 0
-------
5, KEY MODE SURVEY LOW. CR NOX (19,68+)
o
o
OsJ
-F*
I
OJ
-p*
CD
«
o'
(_)
ZCM
LoJ—'
=>o
Cg
LU
Cd
00
o
o
o
o
112.90
99.60
186-30
PPfl
273 .00
359.70
* 1 0
446 .40
-------
KEY MODE SURVEY HLQH CR HC K 1968 J
I
U>
cn
^2 .90
1 13 .78
PPM
159 .22
204 .66
» 1 0
250 • 10
-------
o
CD
CO
-------
KEY MODE SURVEY HIGH CR CO (< 1968 )
o
CM
O
ID
o
CM
>- •-

LU
ZD
U>
-~sl
o
o
o
o
^.00 2-60 5 • 2C T-BO
PERCENT
10.40
13 .00
-------
M
-P*
i
CO
00
CM
o
o
OJ
CO
I ¦<
>- •.
U°
Z
LU
ZD
~
uj^:
CH ..
U.O
00
O
a
«
o
o
o
KET
MODE SURVEY HIGH CR CO (1.968+)
^3.00 2-40 4.80 7.20 9-60 12.00
PERCENT
-------
° KEY MODE SURVEY HIGH CR NOX (<1968 )
o"l
ID
CM
O
O
CsJ
o
I
CO
-
<_)
Zl
LU'
a
LU
a:
Li_
:io
ID
c
o
o
o
^ .90
8
5.66 1
64.42 242-18 319.94 391.TO
PPM
*10
-------
KEY MODE SURVEY HIGH CR NQX (1968+)
CM
-Pi
I
-P»
o
(_)
Zl<
U_T
=5(
o
LU
en
:cm
ao
o
o
o
o
a
14.10
112.08 210.06 308.04 406-02 504-00
PPM
* 1 0
-------
KEY MODE SURVEY 2500 HC (< 1968 )
o
10
o
o
o
o
CO
>- •-
<_>°
Z
LjJ
o
o
o
o
o,
10-50
58 .42
106 .34
154-26 202.18 250.10
PPM
*10'
-------
KEY MODE SURVEY 2500 HC (1968+)
o
CO
o
o
to
o
I
ro
(_)
ZO
LxJ"""
=>o
a
LxJ
QL
o
CM
^ .00
-------
KEY MODE SURVEY 2500 CO K 1968)
i
CO
O
\n
o
o
o
CD
V *.
u°
ID
C3 _
CXL
Ll.O
o
o'
^0 .00
4 -20
8.4C 12.€0
PERCENT
16 .B0
3-
21 .00
-------
o
LO
KEY MODE SURVEY 2500 CO (1968+
o
LO
O
o
(_)
Zlo
LijOO
=>o"
o
LU
cd
o
CM
O
o"
^D.OO
4.20 8.40 12.60
PERCENT
16 .80
21 .00
-------
Table 4-10
SURVEY OF CVS MASS EMISSIONS BY MODEL YEAR
MODEL YEAR
i
-to
cn
I MODE
i
_cjLJ q7J
i CO 2 1975
I
NOfX 1 975
PRE 1968
^ ti A N S O
10. I 7\ 5#/o9
i 4 S . 1 C o ^ 2.-y's 2
3 9 5. 4 9 f> 9 ? . S 1 9
2.113 1 . .31 0
196H*
MCAN S )
6 . J b 1 *~ . 71J
91.33? 44.1'+'¦
47 6.671 122.612
2 . H63 1.19 0
AVERAGE
MEAN SD
7.978 5.541
110.317 53.488
446.91 119.019
2.593 1.406
SAMPLC 3IZ-U
(113)
(11C )
-------
Table 4-11
SURVEY OF CVS MASS EMISSIONS BY VEHICLE MANUFACTURER
VEHICLE MANUFACTURER
(GK) (PO',D) (C-RY.) (ftMC) (FOREIGN)
MODE MEAN ST "It AN SD NE4N SD MEAN! SO i
HC 1975 8.316 6.'338 7.31^ 3.933 h.610 6.331 7.129 2.375 5.552 2.^31
CO" 1975 " 123.333 59. 096 95. 737
-------
Table 4-12
SURVEY OF CVS MASS EMISSIONS BY CUBIC DISPLACEMENT
CUBIC DISPLACEMENT
! K10Q) (10 0-199 > (200-299) (300-39^ («»000
! iOOE HEM sn __ME A N _ SO MF A N SD I^AN °D ME AN SO
HC 1975 6.??? ?.i»93 5,. ^ 2.962 <5.811 5.R57 R.36? 5.935 7.765 5.913
J. CO 1975 7V.308 25751+3 90.163 37.^25 118.23V «*9. 009 113.6^3 57.655 117. «*0 0 57.171
I
4^
^ C02 1975 270.196 38, 8 03. 317. UkU <*6.535 380.161 67.869 <*86.015 "^.772 591.6?8 98.399
NOX 1975 ?. ? 1H .85? 2.185 1. 081 2. 26? 1.298 2.77
-------
Table 4-13
SURVEY OF CVS MASS EMISSIONS BY VEHICLE WEIGHT
VEHICLE WEIGHT
1803-2799 21G1-3799 3*00-4793 4800-5791
fiOiJr Si MEAN SO MEAN SD MEW S'.l
HC 1975 5.791 2.355 7.571 4.194 9.484 7.3*5 6.394 3.251
CO 1975 77.861 31. 73? 103 . 62?. 4 7. 1 07 129.046 59.7*6 126.633 51 . 327
-p»
-g I C32 1975 " 296_.539 5 3.274 424.40 3 15.239 511. 462 95.765 616.513 110.73^
¦ NOX 1975 2.358 1.257 2. 437 1.249 2. 853 1.518 2.699 1.1*3
-------
MASS EMISSIONS SURVEY HC (TOTAL)
4*
!¦
VD
o
in
o
0
b"
o
cn
L)u
Z
u
73
C3 _
UJ°
cn
u_o
^.00
0 .20
18-40 26.6U
OM/MI
34 .80
43 .00
-------
GO
N
MR
S3 EMISSIONS SURVEY CO (TOTAL J
CM
o"
o
CM
o
I
en
O
(0
UJ
Z)
a
oc: •.
u_o
QO
O
^7.00 86-20
Figure 4-32 CO Mass Emissions Frequency Distribution for Total Vehicle Population
145.40 204.60 263-80 323-00
-------
-F*
cn
1/5
O
CO
in
CM
O
CM
LU
=3
o
uj 2
cd ¦.
u_°
o
^•4
«
O
tfi
o
o
o
MASS EMISSION SURVEY NQX (TOTAL)
4.00 6.00
on/Mi
=*-
8 .00
^
10 .00
^.00 2-00
-------
5^0 IDLE INSPECTION PROGRAM
Automotive emission control through a program of inspection and
maintenance can be accomplished by employing any one of a number of
basic alternatives. The experimental program undertaken in this study
consisted of the inspecting idle HC and idle CO emissions and determining
whether they were in conformance with the prescribed standards.
Specific pass/fail standards for both engine parameters and idle
emissions were established for pre-controlled vehicles and controlled
vehicles as shown in Table 5-1.
The engine maintenance performed on those vehicles failing the idle
emission test involved the adjustment of the idle parameters (air/fuel
ratio, speed, and timing) followed, if necessary, by additional
maintenance in order to ensure compliance with the emission standards.*
In effect, the average maintenance treatment can be classified as a
"super" idle program. The following sections present an analysis on
the effectiveness of the various phases of the idle inspection maintenance
program.
5.1 IDLE INSPECTION ACCURACY
The use of emission tests to determine the extent of engine mal-
adjustment and malfunctions is desirable from two perspectives. First,
the approach tends to have substantially lower costs than the so-called
functional tests; and, second it provides some indication of the level
of vehicular emissions. Unfortunately, an emission inspection is
* A fifty dollar maximum was assigned as an upper limit on the amount of
maintenance that could be performed. A more detailed discussion
concerning the maintenance program can be found in Volume II.
-------
Table 5-1
Idle Inspection Pass/Fail Criteria
Measurement Criteria
Pre-Control 1ed Control led
1. Idle HC 800 ppm 330 ppm
2. Idle CO 6% 4%
3. Idle RPM -80 rpm* -80 rpm*
4. Basic Timing 1.5 deg* 1.5 deg*
~
Criteria established as part of the basic experiment design, but not
-------
confounded by two basic problems:
Errors of omission - Those vehicles passing the inspection that
have engine maladjustments or malfunctions.
Errors of commission - Those vehicles failing the inspection that
are in a good state of repair.
Errors of omission tend to reduce the effectiveness of the inspection
program whereas errors of commission tend to increase program costs.
Both effects are normally detected in most applications. The experimental
program for Denver was no exception.
Tables 5-2 and 5-3 show the impact of these errors on inspection
accuracy for pre-controlled and controlled vehicles, respectively. For
illustrative purposes consider the case for idle RPM in Table 5-2 . Out
of the 110 vehicles surveyed a total of 16 were found to be outside
the pass/fail criterion established for RPM (i.e. -80 rpm). Using an
idle HC cutpoint of 800 ppm a total of 46 vehicles were rejected
(approximately 42 percent of the total) and of the amount, seven were
found with rpm settings beyond the criterion (approximately 15 percent
of the total). Thus an idle HC inspection was able to detect nearly
44 percent of the total failures in the population. It should be noted
that the percentage failure rate from the inspection (ratio of failed
vehicles to rejected vehicles) must be greater than that occurring
in the population in order for the screening process to be effective.
Otherwise a random selection of vehicles would detect more vehicle
failures than the inspection procedures. This situation occurred when
using idle CO for detecting rpm maladjustments (13.46% from screening
versus 14.55% for actual). For this example, the number of commision errors
was 39(46-7) whereas the number of omission errors was 9 (16-7).
-------
Table 5-2
ANALYSIS OF IDLE INSPECTION ACCURACY FOR PRE 1968 VEHICLES
PARAMETER
DIAGNOSTIC MODE
IDLE RPM £DEV_.2
IDLE HC
IDLE CO
TIMING(DEV.)
IOLE" HC " ~
1 OLE!
IDLE CO
TOTAL
SET CUJ POINT REJECTED FAILED FAILED AS PERCENT OF
PERCENT FAILURES
NO. PERCENT NO. PERCENT OF TOTAL DETECTED
110
<-30 RPM
. .