United States
Environmental Protection
Agency
Air, Noise and Radiation Branch
Region 7
324 East Eleventh St.
Kansas City, Mo. 64106
EPA 907/9-81-005
August, 1981
Air
f/EPA
Background Reseach
For Missouri I/M
Regulations
P082-14254
8
-------
BACKGROUND RESEARCH FOR MISSOURI
INSPECTION/MAINTENANCE
REGULATIONS
by
PEDCo Environmental, Inc.
2420 Pershing Road
Kansas City, Missouri 64108
Contract No. 68-02-3512
Task Order No. 13
PN 3525-13
Project Officer
Michael T. Marshall
U.S. ENVIRONMENTAL PROTECTION AGENCY
REGION VII
324 EAST ELEVENTH STREET
KANSAS CITY, MISSOURI 64106
September 1981
-------
DISCLAIMER
This Final Report was furnished to the U.S. Environmental
Protection Agency by PEDCo Environmental, Inc., Cincinnati, Ohio
in fulfillment of Contract Number 68-02-3512, Assignment Number
13. The opinions, findings, and conclusions expressed are those
of the authors and not necessarily those of the Environmental
Protection Agency or of cooperating agencies. Mention of company
or product name is not to be considered as an endorsement by the
Environmental Protection Agency.
-------
EMISSION ANALYZER SPECIFICATIONS
-------
PREFACE
The Clean Air Act Amendments of 1977 defined inspection and
maintenance (I/M) as a reasonable technique for the control of
carbon monoxide and ozone and mandated its implementation in
those areas where the states cannot demonstrate attainment of the
air quality standards for these pollutants by December 31, 1982.
The State of Missouri is in the development phase for their
I/M program and certain aspects of this proposed program needed
to be evaluated prior to implementation. This report presents
the results of the background research performed for the State of
Missouri to assist in the development of their program. The
basic procedure was to review the experience obtained in other
I/M programs and to formulate a specific program for the State of
Missouri based on their unique needs. This document contains
four separate reports which summarize the results for the major
research topics:
1. Emission Analyzer Specifications
2. Quality Assurance Procedures
3. Inspection Station Requirements
4. Standardized Procedures for Emissions and Tampering
Inspections
The first three reports were prepared and distributed separately
in final form in June 1981 and are reprinted here. The final
form of the fourth report has been presented only in this docu-
ment.
m
-------
CONTENTS
Figures iv
Tables v
1.0 Introduction 1
1.1 Purpose 1
1.2 Existing emission analyzer specifications 1
1.3 Principles of emission analyzer operation 2
1.4 Report organization 6
1.5 Summary 6
References 7
2.0 Accuracy of Readings 8
2.1 Analyzer calibration accuracy 8
2.2 Zero and span drift 10
2.3 HC hangup 10
2.4 Gas and electrical interferences 12
2.5 Leaks 15
2.6 Repeatability 15
3.0 Analyzer Features Affecting Measured Values 19
3.1 Temperature and humidity operating range 19
3.2 Warmup time and lockout 21
3.3 Probe design 21
3.4 Sample line design 23
3.5 Water removal system 25
3.6 Conversion factor 25
3.7 Low flow indicator 29
3.8 Meter design 29
3.9 Instrument range 31
3.10 Response time 34
3.11 Internal or external gas for calibration 34
References 38
4.0 Construction, Materials, and Dependability 39
4.1 Materials and finish, vibration and shock 39
protection
4.2 Particulate filter 39
4.3 Pump 42
4.4 Dependability 42
5.0 Additional Analyzer Features for Decentralized 45
Programs Recommended by EPA
5.1 Technical overview 45
5.2 Evaluation 47
References 48
iii
-------
CONTENTS (continued)
6.0 Analyzer Availability 49
6.1 Range of analyzers available 49
6.2 Selecting the analyzer 49
6.3 Certification of units 52
7.0 Evaluation of Analyzer Specifications 53
7.1 Technical overview 53
7.2 Operator error 57
7.3 Change in I/M program costs related to emission 59
analyzer selected
7.4 Recommended analyzer specifications 60
Appendices
A. Analyzer Specification Evaluation 62
B. Change in I/M Program Costs Related to Emission 75
Analyzer Selected
IV
-------
FIGURES
Number Page
1-1 Schematic Diagram of Typical Emission Analyzer 4
v
-------
TABLES
Number Page
1-1 Emission Analyzer Specifications Adopted by
Other States 3
2-1 Comparison of Specifications for Analyzer
Accuracy 9
2-2 Comparison of Specifications for Drift 11
2-3 Comparison of Specifications for HC Hangup 13
2-4 Comparison of Specifications for Gas and
Electrical Interference 14
2-5 Comparison of Specifications for Leaks 16
2-6 Comparison of Specifications for Repeatability 18
3-1 Comparison of Specifications for Temperature
and Humidity 20
3-2 Comparison of Specifications for Warmup Time
and Lockout 22
3-3 Comparison of Specifications for Probe Design 24
3-4 Comparison of Specifications for Sample Line
Design 26
3-5 Comparison of Specifications for the Water
Removal System 27
3-6 Comparison of Specifications for the Propane to
Hexane Conversion Factor 28
3-7 Comparison of Specifications for Low Flow
Indicator 30
3-8 Comparison of Specifications for Meter Design 32
VI
-------
TABLES (continued)
Number Page
3-9 Comparison of Values for Instrument Range 33
3-10 Comparison of Values for Response Time 35
3-11 Comparison of Specifications for Internal/
External Gas Systems 36
4-1 Comparison of Specifications for Materials,
Finish, Vibration, and Shock Protection 40
4-2 Comparison of Specifications for Particulate
Filters 41
4-3 Comparison of Specifications for Pumps 43
6-1 Analyzer Availability 50
7-1 Cumulative Measurement Error 55
7-2 Summary of Results 56
7-3 Effect of Measurement Error on Program Stringency 58
vn
-------
SECTION 1.0
INTRODUCTION
1.1 PURPOSE
The purpose of this report is to provide technical informa-
tion to the State of Missouri on emission analyzers potentially
available for use in a mandatory decentralized inspection/mainte-
nance program. The report will serve as the technical basis for
the state in adopting regulations for emission analyzer specifica-
tions.
1.2 EXISTING EMISSION ANALYZER SPECIFICATIONS
1.2.1 Nationally Recognized Analyzer Standards
Automobile exhaust analyzers have been commercially avail-
able for several years prior to the concept of an auto inspection/
maintenance program and have been used at auto repair facilities
for tune-up and diagnostic work. Using the emission analyzer for
inspection program places a new burden of accuracy on the instru-
ment.
Several attempts to improve the emission analyzer have been
made by establishing minimum specifications. There are four such
sets of minimum specifications that have been circulated nationally.
In 1974, the California Bureau of Automotive Repair (BAR) published
minimum instrument specifications for garages participating in
its Blue Shield program. These are generally known as BAR 74
specifications (California Bureau of Automotive Repair 1974). In
1979, BAR upgraded these specifications, and they are now known
as the BAR 80 specifications (California Bureau of Automotive
Repair 1979). The Equipment and Tool Institute (ETI) has also
developed a set of specifications (Equipment and Tool Institute
1979) that contain many common elements with the BAR 74 speci-
fications. The most restrictive set of specifications are recom-
mendations (not requirements) developed by the Environmental
Protection Agency (U.S. EPA 1980). One subset applies to manual
operation analyzers recommended for centralized programs, while
the other subset references specifications for computerized units
recommended for decentralized programs. The EPA specifications
are under minor revision. All states with an existing or planned
-------
I/M program use one of these basic four sets of specifications or
a slight variation of them.
1.2.2 Section 207(b) Requirements
Section 207 of the Clean Air Act also contains emission
analyzer specifications. Section 207(b) is a consumer protection
measure that requires vehicle manufacturers to warrant that a
1981 or later vehicle will conform to emission standards through-
out its useful life providing that the vehicle is properly main-
tained and operated. If the vehicle fails an "approved short
test", the manufacturer must repair the vehicle free of charge to
the consumer. Subsequent EPA promulgations have defined useful
life of a vehicle as five years and have defined the vehicle
parts that vehicle manufacturers must warrant. For 5 years, all
emission control devices must be warranted including the catalyst,
air pump, and electronic controls. For the first 2 years/24,000
miles, parts of the ignition system (spark plugs, ignition wires,
and similar components) are also covered.
In order for a consumer to take advantage of the vehicle
warranty, the vehicle must be tested using an "approved short
test". An approved short test is an idle test with an emissions
analyzer meeting specifications listed in the regulations.
Missouri's proposed idle test would meet the 207(b) requirements.
However, for Missouri consumers to take advantage of 207(b)
warranty benefits, the emission analyzer specifications for the
Missouri I/M program must be as stringent as the 207(b) emission
analyzer specifications. Relating the 207(b) emission analyzer
specifications to the four national sets of analyzer specifica-
tions described previously, ETI and BAR 74 standards are not as
stringent as the 207(b) requirements whereas BAR 80 and EPA
specifications are as stringent.
1.2.3 Specifications Adopted for Other State I/M Programs
Table 1-1 shows which states are committed to the respective
sets of analyzer specifications. Only 21 states have decided
which set of specifications to use. Eleven states have chosen
the BAR 80 specifications. Indiana has chosen EPA specifications
while two other states have specified Section 207(b) specifica-
tions. The remaining states have specified BAR 74 or ETI speci-
fications.
1.3 PRINCIPLES OF EMISSION ANALYZER OPERATION
The principles of emission analyzer operation are reviewed
in the following paragraphs in order to aid in understanding the
operation of emission analyzers. A schematic diagram of a typical
emission analyzer and a basic operational description is included
in Figure 1-1.
-------
TABLE 1-1. EMISSION ANALYZER SPECIFICATIONS ADOPTED BY OTHER STATES
Ixl
Region
I
11
III
IV
V
VI
VII
VIII
IX
X
State
Conn.
Mass.
R.I.
N.J.
N.Y.
D.C.
Del.
Md.
Pa.
Va.
Ga.
Ky.
N.C.
Tenn.
111.
Ind.
Mich.
Ohio
Wis.
N. Mex.
Tex.
Kans.
Colo.
Utah
Ariz.
Nev.
Oreg.
Wash.
Calif.
ETI
•
certified
repa i r
stations
Garages
•
BAR 74
•
• with
modifications
• with grand
father clause
•
•
BAR 80
•
•
centralized
contractor,
fleet station
•
•
•
Fleets
•
• with
modifications
•
•
EPA
•
Other
207(b)
207(b)
Decision
not yet made
by state
•
•
•
•
•
•
•
•
•
-------
Exhaust gas is collected at the tailpipe (1) and drawn through the pick-up and
hose (2). Condensation flows into a water trap (3) and is removed by an aspirator
(4). The gas continues through the selector valve (5) and filter (6) which
removes solids. The vacuum switch (7) signals gas line restrictions. Gas
then flows into the pump (8) and is forced through the HC and CO sampling
tubes (9). The restrictor allows some gas flow through the aspirator venturi.
Infrared energy, generated by a hot filament (10), is projected through the
sample tubes. The infrared beam is chopped by a synchronized disc (11),
focused on a filter (12) and converted to an electrical signal by the detector
(13). The signal is coupled to an amplifier (14) which drives the meter (15).
Reference (16) is the d.c. power supply for the emissions analyzer package.
The p.c. board (17) provides calibration adjustment of HC/CO span and high
meter ranges.
Figure 1-1. Schematic Diagram of Typical Emission Analyzer.
-------
The analyzer samples exhaust gases and provides meter readings
for percent of carbon monoxide (CO) and parts per million hydro-
carbon (ppm HC) emissions. Quantitative analysis of CO and HC is
determined by the amount of infrared energy absorbed by the gas
sample. CO absorbs energy of 4.65 micron wavelength and HC
absorbs energy of 3.42 micron wavelength. Since the analysis
principle is the same for both gases, only CO detection will be
described.
An infrared beam is directed through a sample tube and
focused on a detector. In front of the detector is an optic
filter that permits 4.65 micron wavelength to pass. CO will
absorb energy of 4.65 micron wavelength. Other gases in the
sample tube absorb energy of different wavelengths; therefore,
the optic filter must pass only the wavelength in which CO absorbs
infrared energy. The detector senses the infrared energy as heat
and transforms it into electrical voltage.
With no CO in the sample tube, a voltage of about 60 micro-
volts is developed by the detector. This signal becomes weaker
as the amount of CO increases. Amplifying such a small d.c.
voltage is very difficult due to drift; so the signal is con-
verted to a.c. by chopping the infrared beam with a rotating disc
before it reaches the detector. Holes within this disc and the
speed it rotates allow 10 pulses of infrared energy to strike the
detector each second. The detector output is 10 hertz, the
amplitude changing inversely with the amount of infrared energy
passing through the sample tube.
The detector output is coupled to the preamp stage of the
amplifier p.c. board. The preamp is a high gain, low noise
amplifier. The preamp output is coupled to the electronic filter
stage which is sharply tuned to 10 Hz. In this stage, all unwanted
noise or random signals are blocked and do not appear in the
output. A test point is provided on the amplifier p.c. board for
measuring the a.c. output from this stage. It will be approxi-
mately 3.5 volts a.c. with no CO in the sample tube.
The a.c. output from the electronic filter stage is rectified
and filtered and coupled to the comparer stage. At this point an
equal d.c. voltage of opposite polarity is injected by the Air
Adjust pot. This cancels the signal voltage and gives a zero
meter reading. If CO is introduced into the sample tube, the
signal voltage is reduced, and the difference compared and ampli-
fied by the comparer stage. The output from this stage is applied
to the meter via the Span Adjust potentiometer, and provides an
accurate reading in percent of CO present (BAR Automotive Service
Equipment Company 1980).
-------
1.4 REPORT ORGANIZATION
The emission analyzer specifications deal with items of
analyzer construction and operation that determine the ability of
the analyzer to obtain valid emission measurements. These ele-
ments have been grouped into four categories and appear as Sec-
tions 2 through 5. Section 6 contains a description of all
analyzers now available, their cost, and which specifications
they meet. Findings are presented in Section 7.
1.5 SUMMARY
Based on the findings of this report, the BAR 80 specifica-
tions are recommended because of availability of the analyzers,
compatibility with Section 207(b) requirements, overall accuracy
with respect to meeting the I/M program air quality goals, and
protection against operator error.
-------
REFERENCES
Bear Automotive Service Equipment Company. 1980. Model 42-080
Fuel Economy Analyzer Operating Instructions. Milwaukee, Wisconsin.
California Bureau of Automotive Repair. 1974. Performance
Criteria, Design Guidelines, and Accreditation Procedures for
Hydrocarbon and Carbon Monoxide Analyzers Required in California
Official Motor Vehicle Pollution Control Systems. Sacramento,
California.
California Bureau of Automotive Repair. 1979. Exhaust Gas
Analyzer Specifications. Sacramento, California.
Equipment and Tool Institute. 1979. ETI Model Specifications
for Garage Type NDIR HC CO Measurement Instruments for Use in the
Garage Phase of State I&M Programs.
U.S. Environmental Protection Agency. 1980. Analysis of the
Emission Inspection Analyzer. EPA-AA-IMS-80-S-A/B.
7
-------
SECTION 2.0
ACCURACY OF READINGS
2.1 ANALYZER CALIBRATION ACCURACY
2.1.1 Technical Overview
Analyzer calibration accuracy refers to the degree of preci-
sion by which an instrument is able to determine the true con-
centration of a precision gas. While all components of the
analyzer potentially influence accuracy, the reference is pri-
marily to the bench, which is a term used for the main sample
processing assembly of the analyzer including detectors, sampling
tubes, processing boards, infrared sources, and power supply (see
Subsection 1.4 for a description of these components). Accuracy
is determined by the precision of the components in the bench.
Beyond the inherent precision of the components other factors
potentially affect the proper operation of this system. They
include moisture, dust, temperature, vibration, altitude, electro-
magnetic interferences, power fluctuations, and zero and span
adjustments.
2.1.2 Comparison of Specifications
A comparison of specifications for analyzer accuracy appears
in Table 2-1. The specifications are stated in different formats
making specifications more or less restrictive in different
measurement ranges. The ETI specification is the least restric-
tive with respect to temperature range. Since analyzers in
Missouri often would be operated outside of the 55° to 85°F
temperature range, the narrow temperature range is a severe
shortcoming. BAR 74 specifications are the least restrictive
over the low measurement range. BAR 80 specifications are the
most restrictive over the low measurement range and are the same
as BAR 74 in the high measurement range. Section 207(b) specifi-
cations are slightly less restrictive than BAR 80 and EPA but
more restrictive than BAR 74 or ETI. The impact of analyzer
accuracy on the effectiveness of the inspection/maintenance
program is quantitatively evaluated in Section 7.0.
8
-------
TABLE 2-1. COMPARISON OF SPECIFICATIONS FOR ANALYZER ACCURACY
Standard
Specification
ETI
BAR 74
BAR 80
EPA
207(b)
0-500 ppm HC: ±25 ppm.
500-2000 ppm HC: ±60 ppm, or ±5 percent of reading, whichever is
greater.
0-2.5 percent CO: 0.1 percent, or ±5 percent of reading, whichever
is greater.
2.5-10 percent CO: 0.3 percent, or ±5 percent of reading, whichever
is greater.
Temperature range: 55° to 85°F. Double values when outside of
temperature range.
0-1000 ppm HC: ±30 ppm.
1000-2000 ppm HC: ±60 ppm.
0-5 percent CO: ±0.15 percent.
5-10 percent CO: ±0.30 percent.
Temperature range: 35° to 110°F.
0-400 ppm HC: ±12 ppm.
400-1000 ppm HC: ±30 ppm.
1000-2000 ppm HC: ±60 ppm.
0-2 percent CO: ±0.06 percent.
2-5 percent CO: ±0.15 percent.
5-10 percent CO: ±0.30 percent.
Temperature range: 35° to 110°F.
±5 percent of point.
Temperature range: 35° to 110°F.
200-220 ppm HC: ±15 ppm.
1.0-1.2 percent CO: ±0.1 percent.
Temperature range: not given.
-------
2.2 ZERO AND SPAN DRIFT
2.2.1 Technical Overview
Drift refers to the amount of meter reading change over a
period of time. Zero drift refers to a change of zero reading.
Span drift refers to a change in reading of a calibration point
on the upper portion of each scale. The calibration points are
established by adjusting the unit to calibration gases of known
concentrations. Instrument zero drift can exceed ±3 percent of
full scale at normal operating temperatures over a period of four
hours. Temperature extremes generally serve further to degrade
zero drift performance.
Subsection 1.4 explains the process by which the infrared
beam is directed through the sample tube and focused on the
detector. The detector senses the infrared energy as heat and
transforms it into electrical voltage. Drift may occur due to
the amplification of this small voltage. Subsection 1.4 also
describes how this is minimized by converting the signal to a.c.
by chopping the infrared beam with a rotating disc.
2.2.2 Comparison of Specifications
A comparison of specifications for zero and span drift
appears in Table 2-2. The amount of meter reading change is
most restrictive for the EPA specifications. Next most restric-
tive are the 207(b) and BAR 80 specifications. BAR 74 specifi-
cations are the least restrictive. The impact of emission analyzer
drift specifications on I/M program effectiveness is quantitatively
evaluated in Section 7.0.
2.3 HC HANGUP
2.3.1 Technical Overview
Hangup refers to the process of hydrocarbon molecules being
absorbed, condensed, or by any other method removed from the
sample flow prior to reaching the analyzer detector. It also
refers to any subsequent desorbtion of the molecules into the
sample flow when they are assumed to be absent. The main com-
ponent of the analyzer that influences HC hangup is the flexible
sample line. The operating environment in which the analyzer is
to be used is important in keeping HC hangup to a minimum.
Temperature and humidity extremes may affect the sample lines
ability to transport the exhaust gas to the analyzer.
10
-------
TABLE 2-2. COMPARISON OF SPECIFICATIONS FOR DRIFT
Standard
Specification
ETI
BAR 74
BAR 80
EPA
207(b)
Zero: HC: <15 ppm in a 1-hour period.
CO: <0.6 percent CO in a 1-hour period.
Span: HC and CO: ±3 percent of reading for the first hour and ±2
percent; of reading for second hour.
Zero and span: 0-1000 ppm HC: ±30 ppm HC in a 1-hour period.
1000-2000 ppm HC: ±60 ppm HC in a 1-hour period.
0-5 percent CO: ±0.15 percent in a 1-hour period.
5-10 percent CO: ±0.30 percent in a 1-hour period.
Zero: Shall not exceed ±12 ppm HC and 0.06 percent CO for 1-hour
operation.
Span: Shall not exceed ±12 ppm HC and 0.06 percent CO for first hour
and shall not exceed ±8 ppm HC and 0.04 percent CO for the second
hour of continuous operation.
Zero: HC: ±8 ppm HC for 1 hour.
CO: ±0.04 percent CO for 1 hour.
Time reduced to 30 minutes
if analyzer is equipped with automatic zeroing system.
Span: HC: ± 8 ppm HC for 1 hour.
CO: ±0.04 percent CO for 1 hour.
Zero and span drift shall not exceed ±0.1 percent CO and ±15 ppm HC
on the lowest range capable of reading 1.0 percent CO or 200 ppm HC
during a 1-hour period.
11
-------
2.3.2 Comparison of Specifications
A comparison of specifications for HC hang-up appears in
Table 2-3. It should be noted that all HC concentrations are
measured as hexane. The ETI and BAR 74 specifications are iden-
tical while the BAR 80 specifications are more stringent. The
BAR 80 specifications require that the HC hangup check procedure
be completed at 15 and 30 seconds after a 2 minute sample has
been taken as opposed to 30 seconds after a 1 minute sample for
the ETI and BAR 74. The HC hangup must be less than 200 ppm HC
for the ETI and BAR 74 specifications. The EPA specifications
are the most restrictive in that the HC hangup must be less than
20 ppm HC before each test.
2.4 GAS AND ELECTRICAL INTERFERENCES
2.4.1 Technical Overview
Vehicle exhaust gases contain some components which are of
no interest but which may affect measured readings of HC and CO.
These can be defined as noninterest gases. The interference
effects of various noninterest gases can be quite significant
with respect to operating environment. Carbon dioxide and N02
are two of the most dominant interference gases. Subsection 1.4
describes how some of the interfering gases are treated at the
detector.
Errors may also be caused by responses to electromagnetic
sources and power supply variations. Common forms of electro-
magnetic sources are:
1. Radio frequency interference (RFI)
2. Very high frequency interference (VHF)
In addition to RFI and VHF, induction, line interference, line
voltage and frequency variation, and static electricity need to
be controlled. Interference is determined by exposing the analyzer
to some, or all, of these conditions, depending on the standards
to be adopted.
2.4.2 Comparison of Specifications
A comparison of specifications for gas and electrical inter-
ferences appears in Table 2-4. The specifications vary with
respect to the number of interfering items and also the percent
fluctuation that is allowed for the specific item being examined.
The ETI specifications are the least restrictive with no criteria
specified for electrical interference. The BAR 74 and BAR 80
specifications are the same for gaseous interferences but no
12
-------
Standard
TABLE 2-3. COMPARISON OF SPECIFICATIONS FOR HC HANGUP
Specification
ETI Less than 200 ppm HC (10 percent full scale) in 30 seconds after
1 minute sample.
BAR 74 Less than 200 ppm HC (10 percent full scale) in 30 seconds after
1 minute sample.
BAR 80 Less than 200 ppm HC (10 percent full scale) in 15 seconds after
2 minute sample. And 60 ppm HC in 30 seconds.
EPA Less than 20 ppm HC (5 percent full scale) before each test.
13
-------
TABLE 2-4. COMPARISON OF SPECIFICATIONS FOR GAS AND ELECTRICAL INTERFERENCE
Standard
Specification
ETI
BAR 74
BAR 80
EPA
Gaseous: 5 items: CO-0.05 percent CO interface limit. Interfering
gas - 15 percent COp in N2, 1600 ppm hexane in N2_ HC-10 ppm HC
interface limit. Interfering gas - 10 percent CO in N?, 3 percent
hLO in air, and 3000 ppm NO in Np.
Electrical: No criteria specified.
Gaseous: 6 items: Interference shall not exceed ±1 percent full
scale HC and CO for the following gases: 15 percent C0? in N?)
1600 ppm HC in N^, 10 percent CO in N2, 3 percent H^O vapor in air
(saturated), 3000 ppm NO in N2, and 10 percent Op in N,,.
Electrical: No criteria specified.
Gaseous: 6 items: Interference not to exceed ±10 ppm HC or 0.05
percent CO (as measured on the low scale) for the following gases:
15 percent C02 in N2, 1600 ppm HC in H2, 10 percent CO in N2, 3 per-
cent HpO vapor in afr (saturated)(bubbfe method), 3000 ppm NO in N?,
and 10 percent 0? in N?.
Electrical: 5 items: Interference not to exceed ±10 ppm HC or 0.05
percent CO (as measured on the low scale) for the following radiation
sources: RFI, VHP, induction, line interference and line voltage,
and frequency variation.
Gaseous: 3 items: Interference not to exceed 6 ppm HC or 0.02
percent CO (as measured on the low scale) for the following gases:
14 percent C02, saturated water at 40°C (101°F), and 100 ppm N02-
Electrical: 6 items: Interference not to exceed 4 ppm HC or 0.02
percent CO (as measured on the low scale) for the following electronic
interferences: RFI, VHF, induction, line interference, line voltage
and frequency variation, static electricity (analog meters only, 10
ppm HC and 0.05 percent CO).
14.
-------
criteria are specified for electrical interferences with BAR 74.
The EPA specifications are the most restrictive with respect to
the number of electrical items tested but require only three
items to be examined under gaseous interference.
2.5 LEAKS
2.5.1 Technical Overview
Leaks in the sample system are probably the source of the
largest and most frequent errors that occur in emission measure-
ment systems. This is because a leak is transformed directly
into an error (i.e., a 15 percent leak is a 15 percent error).
Leaks will normally occur in the vacuum side of the system and
can be found in many places. The most common areas for leaks to
occur are in the sample line and other sample transport components.
Leaks may also occur due to improper operating or maintenance
procedures while changing the particulate filter or maintaining
the water trap.
2.5.2 Comparison of Specifications
A comparison of specifications for leak detection appears in
Table 2-5. There is no value specified for leaks in the ETI or
BAR 74 specifications. The BAR 80 specifications state that the
sample system must be provided with a means of testing the por-
tion of the sample system which is normally exposed to less than
atmospheric pressure. This capability must be designed into the
analyzer. Pass/fail criteria are specified. The EPA specifica-
tion is more restrictive in that it requires a comparison of the
span gas response introduced through the span network to the
response of the same gas introduced through the probe and sample
line. Weekly checks are required by EPA specifications while the
frequency of field checking is not specified for BAR 80 specifi-
cations .
2.6 REPEATABILITY
2.6.1 Technical Overview
Repeatability refers to the instruments capability to pro-
vide the same value for successive measurements of the same
sample. Analyzer repeatability is highly influenced by the
accuracy of the unit. If the degree of precision by which an
instrument is able to determine the true concentration of a gas
is low, then the chances of repeatability are also diminished.
Some of the same factors which may affect accuracy also affect
repeatability. These include moisture, dust, temperature, vibra-
tion, electromagnetic interference, and power fluctuations.
15
-------
TABLE 2-5. COMPARISON OF SPECIFICATIONS FOR LEAKS
Standard
Specification
ETI No value specified.
BAR 74 No value specified.
BAR 80 The pass/fail criteria is based on dilution introduced by a leak
causing an error of 10 ppm HC or 0.05 percent CO when 80 percent low
range gas is being introduced at the probe. Frequency not specified.
EPA 3 percent of comparative gas readings, weekly.
207(b) 3 percent of comparative gas readings, weekly.
16
-------
2.6.2 Comparison of Specifications
A comparison of specifications for repeatability appears in
Table 2-6. The ETI, BAR 74, and BAR 80 specifications are iden-
tical and allow for a repeatability error of ±2 percent of full
scale. EPA specifications are the least restrictive over higher
readings allowing for an error of ±5 percent of point.
17
-------
TABLE 2-6. COMPARISON OF SPECIFICATIONS FOR REPEATABILITY
Standard
Specification
ETI
BAR 74
BAR 80
EPA
Analyzer repeatability shall be within ±2 percent of full scale during
five successive samples.
Analyzer repeatability shall be within ±2 percent of full scale during
five successive samples.
Analyzer repeatability shall be within ±2 percent of full scale during
five successive samples.
Analyzer repeatability shall be within ±5 percent of point.
18
-------
SECTION 3.0
ANALYZER FEATURES AFFECTING MEASURED VALUES
3.1 TEMPERATURE AND HUMIDITY OPERATING RANGE
3.1.1 Technical Overview
Variations in temperature and humidity affect analyzer
accuracy by affecting the performance of various components in
the bench. Most serious is exposure of the analyzer to high
temperature environments and to a lesser extent, low temperatures.
Various components in the bench are enclosed in insulated
housings to keep the temperature of the component at a constant
temperature to avoid temperature induced-changes in filter wave-
length. The range in ambient temperatures affect the ability of
the analyzer to maintain constant component temperatures. Minor
changes in temperature of the components can be corrected with
the span or zero adjustment. Greater change in temperature
cannot be adjusted and cause inaccurate readings.
Excess water vapor in the exhaust gas, primarily caused by
high humidity, tends to cause hydrocarbon deposits on the sampling
tube walls causing internal light reflections which introduce
substantial inaccuracy to the system. Most analyzer units have a
water removal system (see Subsection 3.5).
3.1.2 Comparison of Specifications
A comparison of standards for temperature and humidity
appears in Table 3-1. All four specifications are different with
the ETI specification the least restrictive and the EPA being
most restrictive. With regard to temperatures, ETI specifica-
tions list a relatively narrow temperature range of 55° to 85°F
while the other specifications specify a range of 35° to 110°F.
The BAR 80 specifications include an exposure to a 10 mph wind.
The relative humidity range specification is 0 to 85 percent for
ETI, BAR 74, and BAR 80, while EPA specifications require 0 to
100 percent.
Ambient temperatures in St. Louis often are outside of the
55° to 85°F range even inside auto repair facilities. This would
limit the usefulness of ETI specification analyzers. Humidity
19
-------
TABLE 3-1. COMPARISON OF SPECIFICATIONS FOR TEMPERATURE AND HUMIDITY
Standard
Specification
ETI
BAR 74
BAR 80
EPA
Temperature 55° to 85°F. For temperatures between 35° to 55°
and 85° to 105° multiply the accuracy limits by 2.0.
Relative humidity 0 to 85 percent, while exposed to winds up to
10 mph.
Temperature range 35° to 110°F, while exposed to 10 mph winds.
Relative humidity 0 to 85 percent.
Temperature range 35° to 110°F, while exposed to 10 mph winds.
Relative humidity 0 to 85 percent.
Temperature range 35° to 110°F.
Relative humidity 0 to 100 percent (raining).
20
-------
over 85 percent can affect analyzer accuracy during periods when
temperatures exceed 80°F (Olson Laboratories 1977; Clemens 1981).
Highly variable readings or machine failure result. An examina-
tion of two years of meteorological data for St. Louis (National
Oceanic and Atmospheric Administration 1976-1977) revealed that
there were no occurrences of 85 percent or greater humidity that
coincided with temperatures in excess of 80°F. Therefore, the 0
to 85 percent humidity range appears adequate.
3.2 WARMUP TIME AND LOCKOUT
3.2.1 Technical Overview
Warmup time refers to the period from power-on to stabilized
operation for an emissions analyzer. As noted in previous sec-
tions, certain components in the bench portion of the emissions
analyzer are very temperature sensitive and must be maintained at
a constant temperature to perform accurately. These components
must be heated or cooled to specified temperatures. The warmup
time encompasses the heating or cooling period.
Analyzer designs vary with respect to indicators that the
analyzer is properly warmed-up. Some analyzers have no indica-
tion at all. Other analyzers contain indicator lights. In both
these cases, emission measurements can be made before the warmup
period is complete. A third category of analyzer contains a
lockout system which prohibits premature operation usually by
locking the metering device to zero or full scale.
3.2.2 Comparison of Specifications
A comparison of specifications for warmup time is shown in
Table 3-2. ETI and BAR 74 specifications are very similar with a
slight difference in ambient temperature range. BAR 80 require-
ments change the warmup time from 30 to 15 minutes over a slightly
different temperature range and also require a lockout feature.
The EPA specification does not use the time parameter. The
system must have a lockout device until certain component tem-
peratures are achieved within the analyzer. The lockout feature
effectively prohibits premature analyzer operation.
3.3 PROBE DESIGN
3.3.1 Technical Overview
The probe is the device inserted into the tailpipe to draw
the exhaust sample. There are several issues regarding probe
design, most centering around sample dilution. If the probe is
not inserted far enough, part of the gas sample drawn may be
21
-------
TABLE 3-2. COMPARISON OF SPECIFICATIONS FOR WARMUP TIME AND LOCKOUT
Standard
Specification
ETI
BAR 74
BAR 80
EPA
The analyzer shall reach stabilized operation within 30 minutes from
power on over the temperature range of 55° to 105°F.
The analyzer shall reach stabilized operation within 30 minutes from
power on over the temperature range of 70° to 110°F.
The analyzer shall reach stabilized operation within 15 minutes from
power on over the temperature range of 75° to 105°F. Must have lock-
out capability.
The analyzer must have a warmup lockout feature with indicators. No
specific warmup time is specified. The lockout feature shall prevent
operation of the printer and readout meter until the system is warmed
up.
22
-------
ambient air as opposed to exhaust gas. The amount of dilution is
directly transmitted into measurement error, i.e., 15 percent
dilution will cause the reading to be 15 percent too low. There
is considerable controversy in the technical literature about how
far a probe must be inserted to avoid sample dilution.
The first issue is the requirement for a device to secure
the probe to the tailpipe. When the analyzer is in use, it is
common practice for the analyzer operator to insert the probe and
return to the front of the car to make adjustments and/or record
test results. Exhaust pressure, vibration and the weight of the
sample line act to push the probe out of the tailpipe causing
possible sample dilution. Many tailpipes are bent or have a
screen very near the opening, limiting the depth of insertion
possible with a rigid probe. A flexible portion on the end of
the probe allows deeper insertion. Even with a flexible end,
some tailpipes will not allow adequate insertion without a tail-
pipe extender (a straight tube which fits tightly over the end of
a tailpipe allowing the gas sample to be drawn from within the
extender). The requirement for a handgrip is a user convenience
necessitated by the heat of the exhaust and the metal probe.
Other specifications for the probe tip relate to calibration
procedures and leak check procedures.
3.3.2 Comparison of Specifications
A comparison of specifications for probe design is shown in
Table 3-3. ETI specifications are the least restrictive in
general and no requirement for tailpipe insertion is given. BAR
74 specifications are similar except for the 6 inch probe inser-
tion requirement. BAR 80 and EPA specifications are very similar
in nature and are more restrictive than ETI or BAR 74. BAR 80
specifies a 12 inch insertion and EPA specifies a 16 inch inser-
tion. Both contain similar requirements for retention, hand
grips, serviceability, tailpipe extenders and provisions for leak
checks.
Research regarding the required depth of probe insertion to
insure an accurate reading gives different results." In general,
a 12 inch minimum is suggested to insure against sample dilution.
3.4 SAMPLE LINE DESIGN
3,4.1 Technical Overview
Sample line variables deal with resistance of the line to
oil which is commonly found around testing locations, composition
of the line such that the hangup is minimized (see subsection
2,3), flexibility at low temperatures, and length.
23
-------
Standard
TABLE 3-3. COMPARISON OF SPECIFICATIONS FOR PROBE DESIGN
Specification
ETI The probe should not slip out of a vehicle tailpipe when in use for
analysis. A thermally insulated comfortable handgrip shall be pro-
vided on the sample probe handle. Probe design should prevent errors
due to dilution.
BAR 74 Same as ETI, plus the probe should be flexible enough to extend into
a curved tailpipe 6 inches.
BAR 80 Retention - probe must incorporate a position means of retention to
prevent its slipping out of tailpipes when in use.
Handgrip - Same as above.
Flexibility - probe shall be flexible enough to extend into a 1-1/2
inch diameter tailpipe having a 3 inch centerline 90° radius bend.
The probe insertion depth must be 12 inches or greater.
Probe serviceability - probe tip and flex section must be replaceable
as a unit.
Calibration gas introduction - probe tip must be designed to permit
the introduction of calibration gas from a 1/2 inch I.D. flexible
rubber hose inserted over the tip.
Anti-dilution - the device may be supplied by the manufacturer to
prevent sample dilution.
Probe cap - cap must be provided for performing a system leak check.
EPA Very similar to BAR 80 except the probe insertion depth is 16 inches.
24
-------
3.4.2 Comparison of Specifications
Specifications for sample line design are shown in Table
3-4. The four standards are not radically different. All specify
a composition minimizing HC hangup and resistance to oil. Speci-
fications for length vary with BAR 80 requiring the longest line.
3.5 WATER REMOVAL SYSTEM
3.5.1 Technical Overview
Most nondispersive infrared analyzers (NDIR) type emission
analyzers cannot accept moisture-laden exhaust and operate properly.
Accuracy can vary by 10 percent depending on the amount of mois-
ture removed. Almost all analyzers use a momentum type water
trap. Specifications deal primarily with serviceability to the
analyzer operator.
3.5.2 Comparison of Specifications
Specifications are compared in Table 3-5. They are quite
similar. All require that the device be self draining. BAR 74
and EPA require that the collection vessel be visible.
3.6 CONVERSION FACTOR
3.6.1 Technical Overview
The term conversion factor refers to the propane to hexane
conversion factor that accounts for the difference in analyzer
response between propane and hexane. It is also referred to as a
"C" factor or a P.E.F. (propane equivalence factor). Emission
analyzers measure hexane (c6H-,4) in exhaust. Calibration gas is
typically propane. During instrument gas calibration, the HC
value is determined by multiplying the propane concentration by
the conversion factor.
The conversion factor varies by individual instrument, even
when the instruments are of the same design. Therefore, a factor
is usually marked on each analyzer.
3.6.2 Comparison of Specifications
Specifications for the conversion factor are compared in
Table 3-6. With the exception of BAR 74 specifications which do
not regulate the factor, the other specifications are similar.
BAR 80 specifications require that each instrument be permanently
labeled as opposed to writing the factor on a sticker which is
the common practice. This is a protective measure to minimize
25
-------
TABLE 3-4. COMPARISON OF SPECIFICATIONS FOR SAMPLE LINE DESIGN
Standard
Specification
ETI
BAR 74
BAR 80
EPA
The flexible line shall be oil resistant and of such composition that
HC hangup will be minimized.
Same as ETI, plus the line length shall be at least 20 feet if the
instrument is placed at the front of the vehicle. The line may be
shorter if the unit is placed at the rear of the vehicle, provided a
remote meter readout is included.
Flexible line shall be oil resistant and of such composition that HC
hangup shall be minimized. The sample line shall resist permanent
crushing and kinking and will be unaffected by exhaust gas tem-
peratures. The sample line shall remain flexible enough to coil the
hose into at least a 24 inch diameter roll at 60°F. The line length
shall be at least 25 feet.
Sample line shall not alter the exhaust sample and shall minimize HC
hangup. Sample line shall be sufficiently flexible at the tempera-
tures to be encountered during vehicle testing to allow normal inspec-
tion operations. Sample line length is 15-25 feet.
26
-------
TABLE 3-5. COMPARISON OF SPECIFICATIONS FOR THE WATER REMOVAL SYSTEM
Standard
Specification
ETI
BAR 74
BAR 80
EPA
207(b)
The water removal system shall either be continually selfdraining or
not require emptying more than once during the normal work day or be
provided with an alarm to indicate needed draining.
The water removal collection vessel shall be visible to the operator.
The water removal system shall be located as low as possible relative
to the analyzer and shall either be continually selfdraining or not
require emptying more than once during the normal work day.
The water removal system shall be continually selfdraining.
A water trap shall be included in the sample system. The trap shall
be selfdraining and visible to the operator. The sample system
shall be designed to prevent condensable water from occurring in the
sample system downstream of the water trap.
The exhaust sampling system shall include a moisture separator.
27
-------
TABLE 3-6. COMPARISON OF SPECIFICATIONS FOR THE
PROPANE TO HEXANE CONVERSION FACTOR
Standard
Specification
ETI
BAR 74
BAR 80
EPA
For all instruments this factor will fall within the range 0.46 to
0.58. This number must be marked on the exterior surface of the
instrument.
Not specified.
Each instrument shall be permanently labeled with its correction
factor carried to three places. Filter factor range for HC shall be
limited between 0.490 and 0.540.
The mean propane to hexane conversion factor for each analyzer sold
as a pass/fail inspection analyzer shall be between 0.48 and 0.56
for each test point.
28
-------
changing the conversion factor to make an analyzer calibrate
within specified levels.
3.7 LOW FLOW INDICATOR
3.7.1 Technical Overview
Low flow refers to a condition wherein an adequate amount of
sample gas is not being drawn into the analyzer to obtain an
adequate reading. A low flow condition could result in an abnor-
mally low measured concentration. Low flow can be caused by many
factors including the water removal system, the particulate
removal system, the sample line, or a component malfunction. On
most analyzers, a low flow indicator light indicates a low flow
condition.
3.7.2 Comparison of Specifications
Specifications for low flow indicator systems are compared
in Table 3-7. The specifications for low flow indicators are
very similar. All require a flow meter (or equivalent) or an
indicator that is activated due to poor response time and/or
meter error.
3.8 METER DESIGN
3.8.1 Technical Overview
Variables regarding meter design are:
1. Analog or digital
2. For analog meters, increments on meter
3. Distance at which the meter can be read
4. For analog meters, scale indicator lamps
Metering devices can be analog (meters with a printed scale
and indicator needle) or digital. Analog devices are potentially
less accurate since they require the operator to read a value off
a scale leading to a possible error due to parallax or misinter-
pretation of the numerical values. Two ranges are usually required
so that readings on the low end of the scale can be more accurately
made. Digital devices, while more costly, display a measured
value and require no interpolation by the operator. Two ranges
are not required.
29
-------
TABLE 3-7. COMPARISON OF SPECIFICATIONS FOR LOW FLOW INDICATOR
Standard
Specification
ETI
BAR 74
BAR 80
EPA
The sample system will be equipped with a flow indicator. The indica-
tor should give a warning indication before either the response time
has exceeded 15 seconds or the flow reduction has resulted in errors
3 percent of full scale on the high range.
The sampling system shall be equipped with a flow meter (or equivalent)
which will indicate sample flow degradation sufficient to cause an error
greater than ±3 percent full scale.
The sampling system shall be equipped with a flow meter (or equivalent)
which shall indicate sample flow degradation when meter error exceeds
3 percent of full scale, or causes system response time to exceed 13
seconds to reach a 90 percent value, whichever is less.
The analyzer must have a low sample flow indicator. The low flow
indicator shall be activated when the sample flow rate is decreased
to a point that would not allow the analysis system to meet the
response time specifications.
30
-------
Analog meters must be graduated in increments to facilitate
observing the measured values. Different increments are specified
for the high and low scale.
Some of the specifications require that the display be of
sufficient size to be read from a specified distance. This
eliminates problems of operators misreading the value from a
distance or having to approach the analyzer to obtain the reading.
Certain specifications also require scale indicator lights.
This is critical on analog devices where two scales are printed
on the same meter. Misinterpretation of the scale being used can
cause a significantly different value to be recorded.
3.8.2 Comparison of Specifications
Specifications for meter design are shown in Table 3-8.
ETI, BAR 74, and BAR 80 specifications are similar. All speci-
fications allow analog or digital meters. On the high scale (see
Subsection 3.9), analog meters must be graduated in increments of
50 ppm of HC or less, and 0.2 percent of CO or less. For the low
scale, CO measurement requirements vary from 0.05 percent to 0.20
percent, while HC requirements vary from 10 ppm to 25 ppm. EPA
specifications state only that meters be appropriately scaled.
The distance at which the reading must be discernable varies from
8 feet (ETI) to 10 feet (BAR 74 and BAR 80) to 15 feet (EPA).
BAR 80 and EPA specifications require range indicator lights.
3.9 INSTRUMENT RANGE
3.9.1 Technical Overview
Instrument range refers to the range of values that can be
displayed on an analyzer. For analog devices, (see Subsection
3.8) two scales are usually required, a low scale and a high
scale. A low scale is required so that more precise readings can
be taken at low measured values. The high scale spans the whole
range of values. For digital devices, a specified range of
values must be available as opposed to a high scale and low
scale.
3.9.2 Comparison of Specifications
Specifications for instrument range are compared in Table
3-9. High scale ranges of 0 to 10 percent CO and 0 to 2000 ppm
HC are required in all specifications for analog devices. A
greater range of CO values is specified for the low scale ranging
from 0 to 2 percent CO (BAR 80 and EPA), 0 to 2.5 percent CO
(ETI), and 0 to 5.0 percent CO (BAR 74). For HC, low scale value
requirements range from 0 to 400 ppm HC (BAR 80 and EPA), 0 to
500 ppm HC (ETI), and 0 to 1000 ppm HC (BAR 74).
31
-------
Standard
TABLE 3-8. COMPARISON OF SPECIFICATIONS FOR METER DESIGN
Specification
ETI Reading must be discernable at 8 feet for metered type devices. The
HC scale having a 2000 ppm range will be graduated in increments of
50 ppm or less. The low HC scale will be graduated in increments of
25 ppm or less. The 0-10 percent scale will be graduated in incre-
ments of 0.2 percent CO or less. The low CO scale will be graduated
in increments of 0.1 percent CO or less below 2 percent and 0.2 per-
cent CO or less above 2 percent. Digital readout systems are accep-
table provided the resolution equals or exceeds the scale graduation
indicated previously.
BAR 74 Reading must be discernable at 10 feet. High scale shall be graduated
in increments of 0.2 percent CO and 50 ppm HC or less. The meter face
of the low scale shall be graduated in increments of 0.1 percent CO
and 25 ppm HC or perhaps less if the scale covers less than 5 percent
CO and 1000 ppm HC.
BAR 80 Reading must be discernable at 10 feet with 0.250 inch high numerals
of bold design. The meter face of the high scale shall be graduated
in increments of 0.2 percent CO and 50 ppm HC or less. The meter
face of the low scale shall be graduated in increments of 0.05 percent
CO and 10 ppm HC or less. Scale indicator lamps are required. Digital
meter increments shall not be greater than 0.05 percent for CO and
10 ppm for HC or less than 0.01 percent for CO and 1 ppm for HC.
EPA Reading must be discernable at 15 feet. Readout devices (digital or
analog) shall be scaled and capable of reading negative values up to
-5 percent of full scale for each range regardless of operational mode.
All multi-range analyzers shall have range indicating lights.
32
-------
TABLE 3-9. COMPARISON OF VALUES FOR INSTRUMENT RANGE
Standard
Specification
ETI
BAR 74
BAR 80
EPA
Dual scale coverage is prescribed.
HC 0 to 2000 ppm CO 0 to 10%
0 to 500 ppm 0 to 2.5%
Digital readout devices can have a single scale provided they meet
the resolution requirements.
Dual scale coverage is prescribed.
HC 0 to 2000 ppm CO 0 to 10%
0 to 1000 ppm 0 to 5%
Digital readout devices can have a single scale provided they meet
the resolution requirements.
Dual scale coverage is prescribed.
HC 0 to 2000 ppm CO 0 to 10%
0 to 400 ppm 0 to 2%
If digital panel meters are used, the respective ranges shall be 0 to
10 percent for CO and 0 to 2000 ppm for HC.
Dual scale coverage is prescribed.
HC -50 to 1000 ppm CO -0.5 to 10%
-20 to 400 ppm -0.1 to 2%
Digital readout devices can have a single scale provided they meet
the resolution requirements.
33
-------
3.10 RESPONSE TIME
3.10.1 Technical Overview
Response time is defined as the reaction time between a
change in concentration at the inlet to the sample system and the
time the analyzer indicates a given percentage of that change.
The basic problem associated with response time is the operator
recording the concentration too soon after inserting the probe or
making an engine adjustment when the probe is already inserted.
3.10.2 Comparison of Specifications
Specifications for response time are shown in Table 3-10.
ETI and BAR 74 specifications are the same with BAR 80 specifica-
tions being slightly more restrictive. EPA specifications require
that 95 percent, as opposed to 90 percent, of the final reading
be reached within a minimum time, but the minimum time of 14
seconds is 4 to 6 seconds longer than the ETI, BAR 74, and BAR 80
specifications.
3.11 INTERNAL OR EXTERNAL GAS FOR CALIBRATION
3.11.1 Technical Overview
Analytical gases that are used to adjust or return the
analyzer response characteristics to those determined by the
calibration gases are designated as span gases. Zero gas is an
analytical gas that is used to set the analyzer response at zero.
3.11.2 Comparison of Specifications
A comparison of specifications for analyzer gases appears in
Table 3-11. The ETI and BAR 74 specifications are similar in
that they only require the analyzer to have the external fittings
which are necessary for calibration. The calibration containers
are not required to be an integral part of the unit. In addition
to the external calibration gas port, the BAR 80 and EPA speci-
fications require the span gas container to be an integral part
of the analyzer unit.
There are advantages and disadvantages to both external and
internal gas systems. If the system is external, the calibration
can be performed through the probe thus testing the probe, sample
line and analyzer. The disadvantage is that for each calibration,
the equipment must be assembled. This is a deterant because of
the time it takes and because of potential inaccuracies if the
equipment is not assembled correctly. With an internal system,
the disadvantages of an external system are overcome; the proce-
dure is easy to perform and no equipment assembly is required.
34
-------
TABLE 3-10. COMPARISON OF VALUES FOR RESPONSE TIME
Standard
ETI
BAR 74
BAR 80
EPA
207(b)
Specification
Concentration indication must reach 90 percent of the final
lized reading within 10 seconds.
Concentration indication must reach 90 percent of the final
lized reading within 10 seconds.
Concentration indication must reach 90 percent of the final
lized reading within 8 seconds.
Concentration indication must reach 95 percent of the final
lized reading within 14 seconds.
Concentration indication must reach 95 percent of the final
lized reading within 15 seconds.
stabi-
stabi-
stabi-
stabi-
stabi-
35
-------
TABLE 3-11. COMPARISON OF SPECIFICATIONS FOR INTERNAL/EXTERNAL GAS SYSTEMS
Standard
Specification
ETI
BAR 74
BAR 80
EPA
Instruments requiring gas calibration shall be equipped with a
suitable calibration gas port.
Fittings for external gas connection are required. Gas calibration
containers may or may not be an integral part of the instrument system.
Fittings for external gas connection are required. In addition, the
analyzer shall incorporate an integral gas system for calibrating
high range scales of HC and CO. Probe tip must be designed to permit
the introduction of calibration gas.
Fittings for external gas connection are required.
incorporate an integral gas system for calibration,
percent CO and 600 ppm HC.
The analyzer shall
preferably at 1.5
36
-------
Conversely, the calibration is performed internally and does not
involve the probe on sample line. Thus, an internal calibration
system should be coupled with a leak check procedure (Subsection
2.5).
Quality assurance procedures are examined in depth in another
PEDCo report (PEDCo 1981).
37
-------
REFERENCES
Clemens, W. 1981. Telephone conversation with W. Clemens, U.S
Environmental Protection Agency, Inspection and Maintenance
Staff, Emission Control Technology Division, Office of Mobile
Source Air Pollution Control. Ann Arbor, Michigan.
National Oceanic and Atmospheric Administration. January 1976-
December 1977. -Local Climatological Data, Monthly Summary—St.
Louis. Asheville, North Carolina.
Olson Laboratories, Inc., Ann Arbor, Michigan. 1977. Vehicle
Exhaust Emission Instruments Evaluation. Prepared for U.S.
Environmental Protection Agency, Ann Arbor, Michigan. Publica-
tion No. EPA-460/3-77-014.
38
-------
SECTION 4.0
CONSTRUCTION, MATERIALS, AND DEPENDABILITY
4.1 MATERIALS AND FINISH, VIBRATION AND SHOCK PROTECTION
4.1.1 Technical Overview
Most analyzers are used in automotive repair facilities and
can be subject to certain abuses such as exposure to corrosive
materials, vibrations, and shock.
4.1.2 Comparison of Specifications
Specifications for materials and finish (Table 4-1) are
similar for all four sets of specifications. Materials and
finish must be resistant to corrosive materials likely to be
found in an auto repair facility.
Specifications for vibration and shock are also very similar.
EPA specifications could be slightly more restrictive since they
define "unaffected".
4.2 PARTICULATE FILTER
4.2.1 Technical Overview
All analyzers require a particulate removal device to remove
particulate from the exhaust gas sample. Often the particulate
filter is combined with the water removal system. The particu-
late is removed with a filter that requires periodic changing.
Failure to change the filter would result in a low flow condition
(Subsection 3.7). A potential problem with the system can result
if the operator improperly installs the filter resulting in
potential leaks (Subsection 2.5).
4.2.2 Comparison of Specifications
A comparison of specifications for particulate filters
(Table 4-2) indicates that the ETI and BAR 74 specifications are
the same requiring only that there be a replaceable particulate
filter. BAR 80 specifications further require that the filter be
adequate to operate for a 2-hour period measuring gas levels of
39
-------
TABLE 4-1. COMPARISON OF SPECIFICATIONS FOR MATERIALS, FINISH,
VIBRATION, AND SHOCK PROTECTION
Standard
Specification
ETI.
BAR 74
BAR 80
EPA
Materials and finish resistant to corrosive type materials found in
automobile service centers and sufficiently durable to withstand
environmental conditions there. Vibration and shock - system opera-
tion shall be unaffected by the vibration and shock encountered under
normal operating conditions in an automotive service center.
Material and finish - same as ETI. Vibration and shock - same as
ETI, plus instruments, motors, and pumps shall be shock mounted as
necessary to absorb any vibrations which might effect the system
operation.
Materials and finish - same as ETI.
BAR 74.
Vibration and shock - same as
Materials and finish - same as ETI. Vibration and shock - all ana-
lyzer and system performance checks must be met during vibration and
shock test and the span shift cannot be more than 2 percent full
scale of low scale.
40
-------
TABLE 4-2. COMPARISON OF SPECIFICATIONS FOR PARTICULATE FILTERS
Standard
Specification
ETI
BAR 74
BAR 80
EPA
The sampling system shall be equipped with a replaceable participate
filter(s) as required to maintain the operating performance of the
analytical and sampling system.
Same as ET'I.
The sampling system shall be equipped with reusable or inexpensive
replaceable filter elements of adequate size to permit uninterrupted
use for at least a 2-hour period while continually measuring exhaust
gas at 800 ppm HC level. Filter bowels shall be transparent. Filter
elements shall be visible from the exterior of the instrument.
Particulate filters shall be included in a sample system. If the
filter is on a vacuum side of the system, it should be leak checked
every time a filter is changed. The filter element and filter system
shall be designed to prevent particulates larger than 5 microns from
entering the sample cell of the analyzer.
41
-------
800 ppm HC and that the system be visible from the exterior of
the instrument. The EPA specification is the most restrictive.
While there is no visual requirement, the system must prevent
particulates larger than 5 microns from entering the sample cell
and the instrument manual must indicate that the system be leak
checked every time a filter element is changed.
4.3 PUMP
4.3.1 Technical Overview
An exhaust gas sample pump is required to create the air
flow through the analyzer. The main area of concern about pumps
is useful life. Beyond durability, the pump should be located
upstream of the optical bench to help prevent the possibility of
dillution air leaks and should have a pump control switch to
enable analyzer standby. The last feature is required so that
the analyzer can be left on all day (avoiding warmup time) with-
out running the motor and drawing gas through the analyzer.
4.3.2 Comparison of Specifications
Pump specifications (Table 4-3) are very similar. All
require a design which will enable 2000 to 2500 hours of opera-
tion or about 1 year of operation (about 2000 hours assuming 8
hours of operation each weekday). Only BAR 80 specifications
require the pump control switch and the pump location upstream
from the bench.
4.4 DEPENDABILITY
4.4.1 Technical Overview
There are no specifications for dependability in any of the
four national sets of specifications other than the pump require-
ment. No reliable studies are available which compare the dependa-
bility of various analyzers under long-term field operating
conditions.
Some preliminary conclusions can be made based on conversa-
tions with analyzer operators. The analyzer is a sophisticated
instrument that requires specialized skills to repair. Gas
station personnel are seldom capable of analyzer repair. The
number of organizations that can provide repair services are few.
In larger cities with I/M programs, most major analyzer manufac-
turers have a service system. However, the number of analyzers
in operation in one area may not justify, to the manufacturer, a
large service department. Some independent organizations do
42
-------
TABLE 4-3. COMPARISON OF SPECIFICATIONS FOR PUMPS
Standard
'Specification
ETI
BAR 74
BAR 80
EPA
The pump should be designed to operate for 2500 hours while sampling
air in the operating condition encountered in the instrument. It
should also be able to survive 2500 test exposures of 2 minutes each
in normal exhaust gas while operating in an instrument.
Minimum operational life of 1 year with no mechanical or electrical
failure. Manufacturer shall furnish pump life test data.
Exhaust gas sample pump shall be shock mounted. A minimum operational
life of 2000 hours is required. A pump control switch shall be
incorporated to enable analyzer standby and gas checks with the pump
off. The pump should be downstream of the bench.
The sample pump shall be designed for at least a 2000 hour life of
continuous duty.
43
-------
repairs. In general, analyzer operators feel that repair costs
are not very competitive, thereby leading to unduly high repair
charges. Sufficient data not are available to validate or
invalidate this claim.
Analyzer technology is relatively new and few NDIR analyzers
have been on the market more than a short period of time. Manu-
facturers have discovered design flaws in some earlier instru-
ments and have often corrected the design problem in new units
and in some cases retrofitted old units. BAR 80 units are just
becoming available, while EPA units are still in development.
Some problems with newly designed equipment would seem inevitable.
Instrument warranties on existing machines are usually a 1-year
manufacturers warranty for parts and labor.
44
-------
SECTION 5.0
ADDITIONAL ANALYZER FEATURES FOR DECENTRALIZED
PROGRAMS RECOMMENDED BY EPA
The EPA has recommended that states implementing a decen-
tralized I/M program add computer capability to the basic units
described in preceding sections. In general, these are fail-safe
features that automatically control functions that would normally
be required of the operator. These functions include automatic
zero/span check (electrical and gas), automatic leak check, and
automatic HC hangup check. These checks are made automatically
at prescribed intervals. If the machine fails any test, it
becomes inoperative. In contrast, these checks may or may not be
performed on a noncomputerized model, by an operator who may or
may not correctly respond to the results of the check. In addi-
tion to these computerized checks, an automatic read system, dual
tailpipe system, automatic test sequence, printer, vehicle diag-
nosis, anti-tampering, and system diagnostic testing capabilities
are recommended.
These features require an onboard microprocessor. Only one
such unit with some of the features described above is commercially
available at this time. The State of New York contracted with
Hamilton Test Systems to produce a computerized unit. Similar
arrangements could probably be made with this manufacturer and/or
others given sufficient lead time.
Because these automated features all require a microproces-
sor they are evaluated as a single option.
5.1 TECHNICAL OVERVIEW
5.1.1 Automatic Zero/Span Check
A timer is installed in the analyzer and at prescribed
intervals, electrical span checks and gas span checks are required.
At the prescribed interval, readout devices are driven to full
scale until appropriate procedures are followed to complete the
check. If the instrument passes the check, the instrument returns
to normal operation. If it does not, it remains inoperative and
an indication of analyzer malfunction is displayed.
45
-------
5.1.2 Automatic Leak Check
This system is similar to the zero/span check. A leak check
is automatically performed at specified intervals. Failure
causes the readout to be driven to 100 percent of full scale and
a fail indicator to be displayed.
5.1.3 Automatic Hang/up Check
An HC check will be automatically be made before each test.
Failure to pass the test will cause the unit to become inoperative.
5.1.4 Automatic Read System
Without this system, slight variations in displayed concen-
trations will result as long as the probe is in place. This
system will begin to integrate or average the values 17 seconds
after the test has begun and will continue to do so for 15 seconds.
The analyzer displays the integrated value.
5.1.5 Dual Tailpipes
This system allows values from two tailpipes on the same car
to be averaged.
5.1.6 Automatic Test Sequence
For tests with more than one mode (e.g., idle and 2500 RPM),
the system identifies the integrated value for each mode, makes a
pass/fail decision, and either prints or stores the result until
the end of the test sequence.
5.1.7 Printer
Following each test the following information is printed:
date, inspection facility number, instrument serial number,
inspection test number, HC and CO standards, integrated HC and CO
values, and pass/fail indication.
5.1.8 Vehicle Diagnosis
The vehicle diagnosis system allows readings to be displayed
continuously (as in a manual system) to allow for vehicle diag-
nosis. During this period, the printer, the automatic data
collection system, and the automatic read system are inoperative.
5.1.9 AntiTampering and System Diagnostic Testing
All analyzer adjustments (zero/span check, antidilution
limits, span gas concentrations, etc.) are put in a tamper proof
box sealed with a gummed lable, braided wire and crimpted lead
seal, or similar device.
46
-------
5.2 EVALUATION
Most of the additional analyzer features for decentralized
programs recommended by EPA are to protect against improper
analyzer operation, i-e., failure to perform quality assurance
procedures. A quantitative evaluation of the improvement in the
quality of the data generated with the EPA recommended features
would require long-term evaluation of I/M programs with and
without the added features. No such long-term evaluations have
been performed. In fact, no analyzer is commercially available
with all these features with which to perform the evaluation.
While no such data is available, it is reasonable to assume
that all analyzer operators would not perform all quality assurance
procedures. However, a quantitative evaluation comparing incre-
mental cost versus incremental data validity cannot be made.
Quality assurance procedures are evaluated more fully in another
PEDCo report (PEDCo 1981).
47
-------
REFERENCES
PEDCo Environmental, Inc. 1981. An Evaluation of Quality
Assurance Procedures for the Missouri Inspection/Maintenance
Program.
48
-------
SECTION 6.0
ANALYZER AVAILABILITY
6.1 RANGE OF ANALYZERS AVAILABLE
Units suitable for an inspection/ maintenance program are
generally referred to as nondispersive infrared analyzers (NDIR).
Manufacturers typically produce a freestanding unit with CO and
HC measurement capability, as well as units wherein the emission
analyzer is part of a unit which contains other automotive diag-
riositc equipment such as scopes, tachometers, voltmeters, etc.
Emission analyzers are listed in Table 6-1 by manufacturer,
specifications they meet and price. Only the basic unit is
listed (without other diagnostic equipment). In certain cases,
the same unit is marketed under different trade names such as
Allen and Rotunda.
Most units that meet BAR 74 specifications would also meet
ETI specifications. Most BAR 74 units appear to also meet Section
207(b) requirements while others do not meet accuracy requirements.
Analyzer response time is also a question (see note under BAR 80
unit). The range of costs for ETI/BAR 74 units is $2500 to
$4170.
The range of costs for BAR 80 units is $4995 to $6950. All
BAR 80 units appear to meet 207(b) requirements. The only reser-
vation concerns analyzer response. Section 207(b) requires a
response of 95 percent of the full reading within 15 seconds.
Most manufacturers report response tests in terms of 90 percent
of reading in 6 to 10 seconds, depending on the manufacturer.
With the exception of the Marquette 42-090, no independent test
data is available to verify adherance to the 95 percent of reading
in 15 seconds, 207(b) specification.
6.2 SELECTING THE ANALYZER
States with decentralized programs have taken one of two
approaches in selecting analyzers to be used in their I/M program.
The first approach is to adopt a national set of specifica-
tions and allow any manufacturer to produce units that will meet
49
-------
TABLE 6-1. ANALYZER AVAILABILITY
Manufacturer
Allen Group Test Products
Allen3
23-360
23-390
Bear Automotive
Marquette
42-080
42-090
Environmental Tectonics
me
FMC Corporation
Autoscan
805
Hamilton Test Systems
Autosense
600 £.2 gas)
CVISa
KAL-EQUIP
5011
Peerless
675
Snap-on-Tools
MT497AS
Sun
EPA- 75
MGA-90
Cost, dollars
ETI/
BAR 74
3675
3061
2500
3550
4170
2995
3592
4045
BAR 80
5500
5500
4995
5850
n.a.f
6950
EPA
EPA
recommended
decentralized
features
5
7, 8
Appears to
meet 207(b)
requirements**
X
X
X
X
X
X
X
X
X
X
X
X
Allen Testproducts also distributes analyzers under the AMSERV, MTSE and
. Rotunda logo.
FMC Corporation also distributes analyzers under the Rotunda and Atlas logo.
c Bear Automotive also distributes analyzers under the Atlas and Rotunda logo.
(continued)
50
-------
TABLE 6-1 (continued)
Automatic data collection and storage realized.
Sun also manufacturers analyzers for Ford Motor Company (Rotunda) and Atlas
Supply Corporation. The units sold under "Rotunda" and "Atlas" are identical
f to those sold by Sun Electric.
KAL-EQUIP also distributed analyzers under the NAPA BALKAMP logo.
9 Section 207(b) response requirement of 95 percent of reading within 15
seconds is not definitely met. Manufacturers specifications indicate a
response of 90 percent of reading in 6 to 10 seconds, depending on
manufacturer.
EPA RECOMMENDED DECENTRALIZED FEATURES
1. Automatic zero span check. 6. Dual tailpipes.
2. Automatic leak check. 7. Printer.
3. Automatic hangup check. 8. Vehicle diagnosis.
4. Automatic read system. 9. Antitampering.
5. Automatic test sequence. 10. System diagnostic testing.
51
-------
that set of specifications. Since the manufacturers design and
market their analyzer toward a specific set of specifications, a
variety of units is immediately available to the I/M program.
Examples of states taking this approach are Nevada and Rhode
Island.
The second approach is for a state to compile their own
specifications such as in New York. Because of the uniqueness of
the specifications, no analyzers were being manufactured to meet
the specifications. To insure that an adequate number of units
would be available, it was necessary to contract with one manu-
facturer to build the units and only the one type of analyzer can
be used in the New York program. While this approach allows the
state to customize their specifications, it requires a significant
lead time (to contract with a manufacturer and to have the units
designed and produced), and eliminates other analyzer manufacturers
from supplying units for the program.
Some modification from the first approach is possible if
slight changes to the national specifications are made to which
manufacturers can respond immediately. An example would be
modifying BAR 74 specifications for the probe (6 inches in length)
to the BAR 80 specifications for the probe (12 inches). Because
many manufacturers market both a BAR 74 and a BAR 80 unit, the
manufacturer can make production modifications that allow BAR 80
probes to be included with BAR 74 units.
6.3 CERTIFICATION OF UNITS '
States that allow any manufacturer to participate in the I/M
program (first approach in Subsection 6.2) have a certification
process to insure that each analyzer used in the I/M program
meets the adopted specifications. The certification processes
vary slightly but generally require at least three elements:
1. Data by independent testing labs to prove that the
instrument meets specifications mandated by the state.
2. Instrument description including operation, models and
prices, schematics and photographs, and instruction
manual.
3. Business status verifying that the manufacturer is
viable to manufacture, warrant, and repair the analyzer.
Certification is given to a particular unit for a specified
time and can be revoked by the state under certain conditions. A
list of certified units is published by the state and distributed
to all prospective testing locations. The testing locations are
free to choose from all certified units.
52
-------
SECTION 7.0
EVALUATION OF ANALYZER SPECIFICATIONS
7.1 EFFECT OF MEASUREMENT ERROR ON PROGRAM STRINGENCY
7.1.1 Technical Overview
The four sets of analyzer specifications were used to deter-
mine estimates in the following three areas:
1. The range of measurement error associated with each set
of specifications.
2. The effect these errors could have on the number of
cars passing or failing in the proposed I/M program.
3. The effect that the change in the pass/fail distribu-
tion of motor vehicles would have on the overall I/M
program effectiveness in meeting air quality goals.
Each set of specfications was evaluated to calculate the
maximum measurement error that could be expected, based strictly
on the specifications. The calculated error values do not include
any errors due to the operator. This limitation is critical
since operator error could double or triple the calculated error
values (see Subsection 7.2).
Using the calculated measurement errors, an approximate
three standard deviation error band (3a) was assigned to each set
of specifications. These error bands were used to determine the
effect of measurement error on the distribution of passing or
failing vehicles during the proposed Missouri I/M program.
Using these results, it was then possible to estimate the
measurement error impact on the proposed program stringency, and
thus program effectiveness in meeting air quality goals.
The following subsections present the results of the evalua-
tions and a brief discussion of the methodology. A complete
description of the methodology used in the evaluations can be
found in Appendix A.
53
-------
7.1.2 Cumulative Measurement Error
Only certain of the specifications listed in the previous
sections could be assigned error values. These specifications
were evaluated at the pass/fail limit proposed by the State of
Missouri, (20 percent failure rate), for each of two model-year
groups for both HC and CO. The 20 percent failure rate has been
specified to achieve the emission reductions called for in the
Missouri SIP.
Each specification was assigned either a positive or nega-
tive error value. The cumulative signed error was then calculated
for each set of specifications using an error propagation formula.
These cumulative errors are shown in Table 7-1 and are expressed
as a percentage of the specific emission value. It should be
noted that these are estimates of measurement error at fixed
concentrations. Also, as the emission levels decrease, the
cumulative error increases.
As seen in the table, the ETI and BAR 74 specifications
yield similar cumulative errors. Both sets of specifications
have cumulative CO positive errors that are smaller than the
negative CO errors and the cumulative HC positive errors are
greater than the negative HC errors. The magnitude of the BAR 74
CO errors are smaller than those for ETI, and the BAR 74 HC
errors are approximately equal to the ETI.
The BAR 80 and EPA cumulative errors are smaller in magni-
tude than both the ETI and BAR 74. Also, the cumulative positive
errors are strictly less than the cumulative negative error for
both pollutants.
7.1.3 Effect of Measurement Error on Pass/Fail Determination
An average error, without regard to sign, was calculated for
each set of specifications shown in Table 7-1. Due to the simi-
larties in the ETI and BAR 74 results, one average error was
assumed to be representative of both. Using the data obtained
from the Missouri pilot I/M program and standard probability
theory, it was possible to calculate the number of vehicles that
would incorrectly pass or fail during the proposed program.
These totals were subdivided into two categories:
Category 1 - Vehicles that passed when they should have
failed.
Category 2 - Vehicles that failed when they should have
passed.
These results are shown in Table 7-2. It is evident from
the data that relatively small changes in error have a great
impact on the number of incorrect determinations.
54
-------
TABLE 7-1. CUMULATIVE MEASUREMENT ERROR
Specification
ETI
BAR 74
BAR 80
EPA
Sign
+
+
+
+
Cumulative error, % of emission value
CO
6%
11.7
16.2
9.2
14.5
5.5
11.5
6.0
8.3
4%
17.7
21.5
11.2
16.8
8.2
13.0
6.0
8.2
HC
450 ppm
29.6
25.6
30.0
26.0
11.8
14.0
6.9
8.2
400 ppm
33.2
28.5
33.8
29.0
13.2
15.0
7.2
8.8
55
-------
TABLE 7-2. SUMMARY OF RESULTS
Pollutant
CO
HC
Specification
EPA
BAR 80
ETI/BAR 74
EPA
BAR 80
ETI/BAR 74
Number of vehicles
Incorrectly
passing
Category 1
658
2437
6039
787
5014
14299
Incorrectly
failing
Category 2
827
3085
7814
320
3509
11745
Total
incorrect
1485
13853^
1107
8523
26044 M
56
-------
7.1.4 Change in Program Stringency
The number of vehicles that fall into Category 1 have the
effect of lowering the actual I/M stringency. Using the data in
Table 7-2 this effect was quantified for each set of specifica-
tions as demonstrated in Table 7-3. It is evident from these
data that as the measurement error increases the effective pro-
gram stringency decreases.
If any one of the analyzers were used in the planned pro-
gram, the emission reductions specified in the SIP would not be
achieved unless the program stringency were increased. However,
as shown in Table 7-3, the increased stringency would probably be
required only for HC and only if either a BAR 74 or ETI analyzer
were used in the program.
7.2 OPERATOR ERROR
Many of the changes made between the BAR 74 and BAR 80
specifications relate to features intended to minimize analyzer
operator error. Primary among these features are:
0 Warmup lockout
0 Integral gas spanning system
0 Automatic electrical span
0 Leak check capability
0 12-inch probe to eliminate sample dilution
0 Self-draining water removal system
0 Changes in instrument meter ranges to lessen errors in
reading test values
The EPA unit has additional features to minimize operator error
as compared to the BAR 80 features. These items were listed in
Section 5.
Unlike the specifications evaluated in Section 7, these
features to minimize operator error cannot be quantitatively
evaluated since a probability term is required that would repre-
sent the likelihood of an operator improperly using an analyzer
because of ignorance or neglect. If an operator using a BAR 74
unit were to use the analyzer properly, the BAR 80 features to
minimize error would not produce an increase in accuracy. How-
ever, in a decentralized program, some operator error is inevi-
table. The potential magnitude of error from improper analyzer
operation is equal to or greater than the quantifiable specifica-
tion evaluations in Subsection 7.1.
In order to illustrate this point, one aspect of operator
error, leak checking, was evaluated for its effect on the cumula-
tive error. In the error analysis described in Appendix A, it
57
-------
TABLE 7-3. EFFECT OF MEASUREMENT ERROR ON PROGRAM STRINGENCY
Pollutant
CO
HC
Specification
EPA
BAR 80
ETI/BAR 74
EPA
BAR 80
ETI/BAR 74
Error
7
10
15
8
14
29
Number of
incorrectly
passing
vehicles
658
2437
6039
787
5014
14299
Actual
stringency, %a
19.9
19.8
19.5
19.9
19.6
18.8
Planned stringency is 20 percent.
58
-------
was assumed that the error due to flow leaks would be no greater
than 0.3 percent CO for both BAR 74 and ETI analyzers. At a 6
percent CO emission level, this value equates to a 5 percent
leak. If all other individual errors were to remain constant and
an additional 15 percent flow leak were to be introduced into the
system, the resulting cumulative negative errors in Table 7-1
would increase by 9.0 percent for the ETI analyzers and 9.7
percent for the BAR 74 analyzers.
This conservative example demonstrates the potentially large
effect that operator error can have on measured emission values.
Consequently, it is important to establish appropriate quality
assurance requirements for a decentralized program to minimize
these errors. Alternatively, these requirements can be minimized
by using more sophisticated analyzers such as the BAR 80 or EPA
specification units that have features to reduce operator error.
7.3 CHANGE IN I/M PROGRAM COSTS RELATED TO EMISSION ANALYZER
SELECTED
I/M program costs are borne by three groups: the state, the
decentralized station owner, and the vehicle owner. Certain
program costs are changed or redistributed depending on the
emission analyzer selected. These items and their related cost
estimates are presented briefly in this subsection. A full
discussion of the calculation methodology is presented in Appen-
dix B.
Significant changes in program cost have been identified in
four major areas. Each of these areas is discussed below and the
affected groups are identified.
7.3.1 Gas Calibration
If an automatic self-gas calibration specification is included
in the analyzer requirements for the I/M program, then both the
state and the decentralized station owner would realize a cost
saving. During the first year, the state could reduce manpower
costs for station audits by $59,400. The 1000 stations would
each save $260 on the labor cost for weekly calibrations. Weekly
calibration is the minimum frequency recommended by the EPA for
decentralized programs. No unit is presently available with
ciutomatic gas calibration features. EPA estimates that this
self-calibration feature could be included in a new production
analyzer for around $150.
7.3.2 Data Processing
The state could achieve additional cost savings if the EPA
option for automatic data collection was included in the analyzer
59
-------
requirements. By utilizing this option, the state would reduce
data handling costs by approximately $16,950 the first year of
the program.
There are two drawbacks to this option. First, no analyzer
is presently available that contains this option. Second, the
option would only be possible with a computerized analyzer.
7.3.3 Instrument Costs
The presently available analyzers were listed in Table 6-1.
The range of analyzer costs was found to be as follows:
Specification Range of costs
ETI/BAR 74 $2500-4170
BAR 80 $4995-6950
Based on the range of costs, each station would spend between
$2500 and $7000 for its inspection analyzer, depending on the
type of unit purchased.
7.3.4 Repair Costs
Varying levels of instrument measurement error were pre-
sented in Subsection 7.1. Based on the HC error for each set of
specifications it was shown that the overall consumer costs
decrease with increasing measurement error. However, program
stringency and associated emission reduction also decrease. With
a perfectly accurate analyzer, annual consumer repair costs were
estimated to be $8.4 million. The estimated costs for each set
of analyzer specifications was determined to be:
Annual reduction in
Specification consumer repair costs
ETI/BAR 74 $8,310,610
BAR 80 $8,347,325
EPA $8,383,655
It should be reiterated that these are only estimated values and
should only be used to compare the relative values for each
analyzer. The analysis methodology used to arrive at these num-
bers included no estimate of additional measurement error due to
the operator.
7.4 RECOMMENDED ANALYZER SPECIFICATIONS
It is recommended that the State of Missouri adopt the BAR
80, or some similar set of specifications, for their I/M program.
60
-------
The rationale for this recommendation is based on a number of
considerations. First, while the EPA specifications are more
accurate than any of the others, there are no analyzers on the
market that meet these specifications. Also, EPA may make changes
in the specifications in the near future. Second, the BAR 80
specifications analyzers meet the 207(b) warranty requirements.
These warranty benefits will be important to the motor vehicle
owners as a means of lowering repair costs. The third considera-
tion is the overall measurement accuracy of available analyzers.
As shown in Table 7-3, the larger HC measurement error in the
ETI/BAR 74 specifications yields a lower effective program strin-
gency than is seen for the BAR 80 specifications. The final
consideration is the BAR 80 analyzer features that protect against
operator error. Operator errors are potentially greater than
inherent analyzer errors, leading to larger measurement inac-
curacies .
These four considerations—availability, 207(b) compati-
bility, accuracy, and protection against operator error, suggest
that the BAR 80 specifications are the best choice for the Missouri
I/M program.
61
-------
APPENDIX A
ANALYZER SPECIFICATION EVALUATION
METHODOLOGY
Four sets of analyzer specifications were used to determine
estimates in the following three areas:
1. The range of measurement error associated with each set
of specifications.
2. The effect these errors could have on the number of
cars passing or failing in the proposed I/M program.
3. The effect that the change in the pass/fail distribu-
tion of motor vehicles would have on the overall I/M
program effectiveness in meeting air quality goals.
The evaluation methodology used for these three areas is
discussed below. Included in the discussions are the assumptions
and mathematical procedures required to perform the evaluation.
ESTIMATION OF MEASUREMENT ERROR
Each of the analyzer specifications was evaluated to deter-
mine its effect on in-use measurement accuracy. The end result
was to calculate the cumulative measurement error, assuming that
the analyzers would not exceed the specification values stated by
the manufacturer. The error analysis was based totally on the
published analyzer specifications and it was assumed that the
analyzer operator would maintain the analyzer within the specifi-
cations. In other words, the calculated errors reflect only
those errors inherent to the specifications and do not include
any additional error due to improper operation and maintenance.
This additonal error is potentially larger than that inherent to
the specifications.
Some of the specifications are not expressed as percent CO
or ppm HC but rather as percent of scale or point. Consequently,
each specification was evaluated at specific pollutant levels.
These values were determined from data obtained during Missouri's
pilot I/M program.
62
-------
The data that were obtained included the distribution of
measured emission levels by vehicle model year. To simplify the
analysis, the ,data for individual model years were combined into
two model year groups—1969-1974 vehicles and 1975 and newer
vehicles. The resulting distributions are shown in Tables A-l
and A-2. These groups are almost identical to those associated
with the voluntary emission standards used in the pilot program.
Each specification was evaluated at the approximate 80
percentile emission value determined from the distributions in
Table A-l and A-2. These values are shown below:
80 percentile value
CO HC
1969-1974 vehicles 6 percent 450 ppm
1975 and newer 4 percent 400 ppm
The 80 percentile level was chosen, since Missouri has
specified a 20 percent stringency level for its I/M program.
Also, the error associated with very large or very small emission
values is of lesser importance in the determination of whether a
vehicle passes or fails the I/M test. Consequently, analyzer
errors were calculated only for the pass/fail emission values.
Each individual specification has a resultant measurement
error. However, only certain of these errors can be quantified.
The magnitude of the error is directly related to the stringency
of the specification. For some specifications there is an equal
chance that the measurement error will be positive or negative
(i.e., random error). For others, the measurement error will be
definitely of one sign or the other. Certain specifications are
interrelated and the total error is contained within one specifi-
cation.
Each of the analyzer specifications shown in Sections 2 and
3 were evaluated for resultant measurement error. These are
briefly discussed below. Included in the discussion are the
assumptions used and an indication of the sign of the error.
Specifications in Section 4 (construction materials and dependa-
bility) and Section 5 (decentralized program recommendations)
were not included in the error analysis. They were excluded to
limit the analysis to accuracy specifications only, as opposed to
dependability or operation error (see Subsection 7.2 for an
explanation of problems associated with analyzing operator error).
Table A-3 lists the specifications for which a related
measurement error could be determined. Also included in the
table is the indication of the sign of the error and the magni-
tude of the error at the given idle emission levels.
63
-------
TABLE A-l. DISTRIBUTION OF MEASURED CO VALUES
Range,
% CO
0- 0.5
0.51- 1.0
1.01- 1.5
1.51- 2.0
2.01- 2.5
2.51- 3.0
3.01- 3.5
3.51- 4.0
4.01- 4.5
4.51- 5.0
5.01- 5.5
5.51- 6.0
6.01- 6.5
6.51- 7.0
7.01- 7.5
7.51- 8.0
8.01- 8.5
8.51- 9.0
9.01- 9.5
9.51-10.0
1969-1974 vehicles
Frequency
106
67
42
53
45
64
52
57
42
71
38
56
33
31
24
31
14
27
7
21
Cumulative
frequency
106
173
215
268
313
377
429
486
528
599
637
693
726
757
781
812
826
853
860
881
Cumulative
percent
12.0
19.6
24.4
30.4
35.5
42.8
48.7
55.2
59.9
68.0
72.3
78.7
82.4
85.9
88.6
92.2
93.8
96.8
97.6
100.0
1975 and newer vehicles
Frequency
1052
155
106
92
106
98
65
92
72
70
50
63
47
31
22
26
16
19
13
9
Cumulative
frequency
1052
1207
1313
1405
1511
1609
1674
1766
1838
1908
1958
2021
2068
2099
2121
2147
2163
2182
2195
2204
Cumulative
percent
47.4
54.8
59.6
63.7
68.6
73.0
76.0
80.1
83.4
86.6
88.8
91.7
93.8
95.2
96.2
97.4
98.1
99.0
99.6
100.0
64
-------
TABLE A-2. DISTRIBUTION OF MEASURED HC VALUES
Range,
% HC
0- 49
50- 99
100- 149
150- 199
200- 249
250- 299
300- 349
350- 399
400- 449
450- 499
500- 549
550- 599
600- 649
650- 699
700- 749
750- 799
800- 849
850- 899
900- 949
950- 999
1000-1049
1050-1099
1100-1149
1150-1199
1200-1249
1250-1299
1300-1349
1350-1399
1400-1449
1450-1499
1500-1549
1550-1599
1600-1649
1650-1699
1700-1749
1750-1700
1800-1849
1850-1899
1900-1949
1950-1999
2000-2050
1969-1974 vehicles
Frequency
16
105
141
152
90
64
62
46
31
25
21
9
8
6
17
4
11
5
8
2
8
0
7
1
7
0
0
1
1
0
3
1
4
1
0
1
2
0
0
0
21
Cumulative
frequency
16
121
262
414
504
568
630
676
707
732
753
762
770
776
793
797
808
813
821
823
831
831
838
839
846
846
846
847
848
848
851
852
856
857
857
858
860
860
860
860
881
Cumulative
percent
1.8
13.7
29.7
46.9
57.2
64.5
71.5
76.7
80.2
83.1
85.5
86.5
87.4
88.1
90.0
90.5
91.7
92.3
93.2
93.4
94.3
94.3
95.1
95.2
96.0
96.0
96.0
96.1
96.3
96.3
96.6
96.7
97.2
97.3
97.3
97.4
97.7
97.7
97.7
97.7
100.0
1975 and newer vehicles
Frequency
307
438
266
215
192
141
124
83
88
60
57
28
32
24
17
11
17
14
16
4
15
2
11
5
5
1
3
0
6
1
2
0
2
2
2
2
0
0
0
0
11
Cumulative
frequency
307
745
1011
1226
1418
1559
1683
1766
1854
1914
1971
1999
2031
2055
2072
2083
2100
2114
2130
2134
2149
2151
2162
2167
2172
2173
2176
2176
2182
2183
2185
2185
2187
2189
2191
2193
2193
2193
2193
2193
2204
Cumulative
percent
13.9
33.8
45.9
55.6
64.3
70.7
76.4
80.1
84.1
86.8
89.4
90.7
92.2
93.2
94.0
94.5
95.3
95.9
96.6
96.8
97.5
97.6
98.1
98.3
98.5
98.6
98.7
98.7
99.0
99.0
99.1
99.1
99.2
99.3
99.4
99.5
99.5
99.5
99.5
99.5
100.0
65
-------
TABLE A-3. MEASUREMENT ERROR OF ANALYZER SPECIFICATIONS
Specification
Calibration accuracy
Zero dri ft
Span drift
HC hangup
Interference-gas (3)
Interference-electrical (5)
Leaks
Low flow
Response time
Cumulative error
Percent of emission value
Sign
+
±
±
+
±
±
-
±
-
4-
+
ETI
% CO
6 4
0.6
0.03
0.14
0.05
0.05
0.3
0.3
0.6
0.70
0.97
11.7
16.2
0.6
0.03
0.14
0.05
0.05
0.3
0.3
0.4
0.70
0.86
17.5
21.5
ppm HC
450 400
50
7.5
27
100
10
10
60
60
45
133
115
29.6
25.6
50
7.5
27
100
10
10
60
60
40
133
114
33.2
28.5
BAR 74
% CO
6 4
0.3
0.15
0.15
0.1
0.1
0.3
0.3
0.6
0.55
0.87
9.2
14.5
0.15
0.075
0.075
0.1
0.1
0.3
0.3
0.4
0.45
0.67
11.2
16.8
ppm HC
450 400
30
15
15
100
20
20
60
60
45
135
117
30.0
26.0
30
15
15
100
20
20
60
60
40
135
116
33.8
29.0
BAR 80
% CO
6 4
0.3
0.03
0.03
0.05
0.05
0.05
0.05
0.6
0.33
0.69
5.5
11.5
0.15
0.03
0.03
0.05
0.05
0.05
0.05
0.4
0.33
0.52
8.2
13.0
ppm HC
450 400
30
6
6
30
10
10
10
10
45
53
63
11.8
14.0
30
6
6
30
10
10
10
10
40
53
60
13.2
15.0
EPA
% CO
6 4
0.3
0.02
0.02
0.02
0.02
0.18
0.18
0.3
0.36
0.50
6.0
8.3
0.2
0.02
0.02
0.02
0.02
0.12
0.12
0.2
0.24
0.33
6.0
8.2
ppm HC
450 400
22.5
4
4
10
4
4
13.5
13.5
22.5
31
37
6.9
8.2
20
4
4
10
4
4
12
12
20
29
35
7.2
8.8
CTl
O1
-------
Analyzer Calibration Accuracy—Subsection 2.1
The maximum error was assumed to be the calibration limits
measured at 50PF and 85 percent humidity. A positive or negative
error is equally likely.
Zero and Span Drift—Subsection 2.2
It was assumed that a maximum of 1/2 of the zero and span
values would be present at the specific pollutant levels. Errors
could be positive or negative.
HC Hangup—Subsection 2.3
One-half of the allowable HC hangup values was used as the
maximum value that would be encountered in actual use. Error is
definitely positive.
Interferences—Subsection 2.4
Errors are specified for each gaseous interference. However,
EPA recommendations include only three interferents. Therefore,
only three individual error values were included. No specifica-
tions were included in the ETI and BAR 74 specifications for
electrical interferences. For the BAR 80 and EPA analyzers the
allowable error values were used, but only for the five items
required under BAR 80. ETI and BAR 74 error levels, for the same
five items, were conservatively assumed to be equal to the gaseous
specifications. Errors for both gaseous and electrical interfer-
ences could be positive or negative.
Leaks--Subsection 2.5
The EPA specifications were the only one of the four that
included leak checks. This is a potentially large source of
error. Negative errors due to leaks were conservatively esti-
mated to be the value which triggers the low flow indicator.
Repeatability—Subsection 2.6
These specifications are directly related to analyzer cali-
bration accuracy. No additional error is expected due to these
specifications.
Temperature, Humidity Operating Range—Subsection 3.1
Included in assumptions for calibration accuracy.
67
-------
Warmup Time and Lockout—Subsection 3.2
It was assumed that the analyzers would be operated only
after recommended warmup time.
Probe Design, Sample Line Design, Water Removal System, Conversion
Factor—Subsection 3.3-3.6
No errors can be assigned to these specifications when
analyzers are operated according to manufacturer's recommenda-
tions .
Low Flow Indicator—Subsection 3.7
Maximum error was assumed to be low flow indicator critical
value. Error could be positive or negative.
Meter Design—Subsection 3.8
Meter design is only critical for analog meters. The scale
divisions determine the potential error. The measurement error
was assumed to be one-half of the scale interval for the lowest
scale capable of reading the specified emission levels. Error
could be positive or negative.
Instrument Range—Subsection 3.9
Instrument range has been taken into consideration for
several other specifications.
Response Time—Subsection 3.10
The maximum error associated with the specification was
assumed to be 100 minus specified percent of final reading.
Error is definitely negative.
Cumulative errors were calculated from the specifications
shown in Table A-3. Both cumulative positive and cumulative
negatives errors were calculated using the following equation:
E0 = E,2 + E,2... + E 2 (Eg. 1)
C .L ^ Xx
where E = cumulative error (positive or negative)
E,E = individual error components
The resulting values are shown in Table A-3. As seen in the
table, the ETI and BAR 74 specifications yield similar cumulative
errors. Both sets of specifications have cumulative CO positive
errors that are smaller than the negative errors and cumulative
HC positive errors that are greater than the negative errors.
68
-------
The magnitude of the BAR 74 CO errors are smaller than those for
ETI, but the BAR 74 HC errors are almost identical to the ETI.
The BAR 80 errors are smaller than the ETI/BAR 74 and the EPA
errors are even lower.
EFFECT OF MEASUREMENT ERROR ON PASS/FAIL DETERMINATION
In order to determine the effect of cumulative measurement
error on the number of cars that pass or fail the idle test, it
was necessary to make a number of simplifying assumptions. These
assumptions were necessary so that standard probability relation-
ships could be used. These assumptions are discussed below.
Assumption 1
The measurement errors are normally distributed. As seen in
Table A-3, the errors actually have a positive or negative bias
depending on the pollutant. However, the use of probability
relationships requires that the error around a given measured
value have an equal likelihood of being positive or negative.
Assumption 2
The average percent cumulative error positive or negative
for a specific pollutant yields a reasonable estimate of a three
standard deviation (3a) error about the measured emission value.
As shown in Table A-3, the cumulative error varies with the
magnitude of the emission value. In addition, these are esti-
mates of the measurement error at fixed emisson levels. In order
to use a probability function the error must be expressed in
terms of the standard deviation. Also, expressing the error as a
percent of the measured value, approximates the results shown in
Table A-3.
Assumption 3
The vehicle distributions determined from the Missouri pilot
program are an unbiased sample of the vehicle population. This
was the most critical assumption, since there exists some level
of measurement error in the sample data. However, it was impos-
sible to isolate the measurement error inherent to the data. In
order to make any inferences about the effect of measurement
error, it had to be assumed that the data were obtained with a
perfect (no error) analyzer.
Assumption 4
The average combined error (percent), for the ETI and BAR 74
specifications was a reasonable estimate for the error in both
systems.
69
-------
Given the above assumptions, it was then possible to calculate
the numbers of vehicles that fell into the following two categories:
Category 1 - Vehicles that passed when they should have failed
(pass-fail)
Category 2 - Vehicles that failed when they should have passed
(fail-pass)
The number of vehicles in these categories was determined for
three different error levels for each pollutant. These were:
Percent error
BAR 74/ETI BAR 80 EPA
CO 15 10 7
HC 29 14 8
The error analysis was performed for three error levels for
each of two pollutants (CO and HC) based on vehicle distributions
for two model year groups (1969-1974, 1975 and newer). The
calculation procedures were focused on the pass/fail value for
each pollutant and model-year group. For vehicles whose true
emission value is greater than the limit, only negative errors
will affect the pass/fail distribution, i.e., vehicles that
passed when they should have failed. The number of vehicles in
this category can be determined from the left-tailed probability
of the standard normal distribution. Conversely, the right-tailed
probability can be used to determine the number of vehicles that
fail when they should have passed. The basic calculation steps
are outlined below. An illustrative example is shown in Table
A-4.
Step 1. Determine the midpoint value for each emission level
range for each pollutant.
Step 2. Identify fail point (20 percent stringency) for each
model year group and each pollutant within that group.
Step 3. Choose 3 a error for evalution—percent.
Step 4. Calculate 1 a error—3a error -r3.
Step 5. Calculate the standard deviation (s.d.) about each
midpoint value—s.d. = midpoint x % error.
Step 6. Calculate the right- or left-tailed probability value
(z) that with a given error, a vehicle, whose true
emission value is in the specified range, will fall
into Category 1 (pass-fail) or 2 (fail-pass). Repeat
for each interval.
fail point - midpoint
z ~ s.d.
70
-------
TABLE A-4. EXAMPLE CALCULATION
Pollutant: CO
Emission level: 5.51 - 6.0%
Midpoint: 5.755 Fail point: 6%
Error (3a): 15% Error (la): 5%
One standard deviation about midpoint: 5.755 x 0.05 = 0.2878
_ _ 6.0 - 5.755 _ n ft,
Z~ 0.2878 -- °'85
Tabled right-tail probability: 1-0.8223 = 0.1977
Sample population: 3,085 vehicles
Sample frequency of vehicles in 5.51-6.0% CO range = 56
Fraction of sample population: = °-0182
Number of fail-pass vehicles: 0.0182 x 0.1977 xl.2 x 106 = 4318
[Jased on the 15 % error 4.318 vehicles will fail the idle test when they
should have passed.
71
-------
Step 7. Look up the standard normal probability for calculated
z values in a statistic table.
Step 8. Record right (+z) or left (-z) tailed probability for
each value.
-z = table value
+z = 1- table value
Step 9. Calculate fraction of total sample population contained
in each emission level range. Fraction = total sample size
Step 10. Calculate the number of vehicles that fall into fail-
pass or pass-fail category for each interval, based on
an estimated population of 1.2 million vehicles in
proposed I/M program—Number = fraction x probability x
1.2 x 106.
Step 11. Sum the number of pass-fail and fail-pass vehicles and
total for each pollutant.
The overall results of the calcuations are shown in Table
A-5. The numbers shown in the table demonstrate the relative
effect of measurement error for the proposed I/M program. It is
evident that as the measurement error increases, the number of
incorrect pass/fail determinations escalates rapidly. The effect
that this has on the overall I/M program is evaluated in the next
Subsection.
Effect of Improper Pass/Fail Determination on Overall I/M Program
There are different consequences resulting from an improper
pass/fail determination depending on the category that is con-
sidered. For Category 2 vehicles, those that failed when they
should have passed, the consumers are affected in two ways.
Those consumers are out both the time and money involved in
needless repairs on the vehicles as well as the time required for
a retest. The I/M program is also affected. Little or no emis-
sion reduction will be realized by repairing these vehicles.
For Category 1 vehicles, those that pass when they should
fail, there are other consumer and program effects. These con-
sumers do not have the required repair or maintenance performed.
Consequently, the consumers do not realize potential fuel economy
savings and maintenance costs may be higher in the long run.
These vehicles have a significant effect on the I/M program. The
number of vehicles in this category actually reduce the program
stringency and the needed emission reductions are not realized.
72
-------
TABLE A-5. SUMMARY OF RESULTS
Pollutant
CO
HC
Specification
EPA
BAR 80
ETI/BAR 74
EPA
BAR 80
ETI/BAR 74
3a
error
7
10
15
8
14
29
Number of vehicles
Incorrectly
passing
Category 1
658
2437
6039
787
5014
14299
Incorrectly
failing
Category 2
827
3085
7814
320
3509
11745
Total
incorrect
1485
5522
13853
1107
8523
26044
73
-------
It is possible to calculate the effect on the program strin-
gency from these results. If it is assumed that the Category 2
vehicles will achieve an insignificant emission reduction, since
they are already clean; then, only the number of Category 1
vehicles will affect the program stringency. These vehicles
should achieve an emission reduction with a perfect measurement
system, but do not when error is introduced into the measurement.
Based on a population of 1.2 million vehicles, a 20 percent
program stringency is designed to fail 240,000 vehicles. The
actual program stringency with a given measurement error can be
calculated as follows:
Actual stringency = (240,000,- number of Category 1 vehicles)/
1.2 x 10°
This calculation was performed on the data shown in Table A-5
with the following results:
Number of
Category 1 Actual
Pollutant Error Specification vehicles stringency
CO 7 EPA 658 19.9
10 BAR 80 2437 19.8
15 ETI/BAR 74 6039 19.5
HC 8 EPA 787 19.9
14 BAR 80 5014 19.6
29 ETI/BAR 74 14299 18.8
From the above, it is evident that the HC errors have a
greater impact than do the CO errors. First of all, the magnitude
of the HC errors is greater than the associated CO error for each
set of specifications. Also, the distribution of measured HC
data, as shown in Table A-2, is skewed more towards low HC readings
than was seen in the CO data. In order to achieve an actual 20
percent stringency the pass/fail cut point would have to be
lowered to fail more of the population. The magnitude of this
adjustment would depend on which set of specifications were
selected.
74
-------
APPENDIX B
CHANGE IN I/M PROGRAM COSTS RELATED TO
EMISSION ANALYZER SELECTED
INTRODUCTION
The purpose of this appendix is to describe the framework
for evaluation of I/M program costs related to the emission
analyzer selected. I/M program costs are borne by three groups,
i.e., the state, the decentralized station owner, and the vehicle
owner.
COST CHANGES TO THE STATE
Cost changes to the state varying with emission analyzer
selection are shown in Table B-l. Most costs have no relation-
ship to the emissions analyzer. The five cost items that will
change with analyzer selection are related to two factors. These
factors are program audit functions, and data processing.
Gas Calibration
The EPA has specified that a gas calibration audit be per-
formed on all analyzers at least monthly in a decentralized
program. The only waiver from this EPA mandate is if computer
controlled emission analyzers with self-gas calibration features
are used. Then, gas calibration audits by state personnel need
only be performed quarterly. Only the EPA specifications (decen-
tralized program) require a computerized self-gas calibration and
would qualify as only requiring quarterly surveillance.
The need for state personnel to perform gas calibrations
only one-third as often, when using computerized self-calibrating
analyzers, impacts four of the five state costs varying by the
type of emission analyzer. The total number of examiners required
for program surveillance would be only one-third of the level
required for monthly gas calibrations. This would decrease costs
of hiring and training examiners, audit related equipment, examiner
wages, and surveillance costs (travel and expendables).
The Missouri Highway Patrol has estimated the need for an
additional eight motor vehicle inspectors of which seven would be
75
-------
TABLE B-l. CHANGE IN I/M PROGRAM COST RELATED TO EMISSION
ANALYZER SELECTED-STATE COSTS
Cost does
not vary
with analyzer
selected
Cost does
vary with
analyzer
selected
A. Start-up costs
1. Administrative
Wages
Hiring and training examiners
Initial public information
Initial program design
Mechanic training (if state sponsored)
Office equipment
Audit related equipment
2. State challenge lanes
Land acquisition
Building
Hiring and training personnel
Office supplies
B. Recurring costs
1. Administrative
Wages (direct and overhead)
Surveillance costs (travel, expendables)
Public information
Computer processing
Office equipment
2. State challenge lanes
Amortized capitol cost
Wages (direct and overhead)
Equipment maintenance
76
-------
involved in the audit function at any one time (Missouri Air
Conservation Commission 1980). Self-calibration would allow the
number to be decreased to three at an annual savings to the state
of $59,433 in the first year or $348,670 over five years, assuming
8 percent wage inflation each year. This cost would be transferred
to the station owner in the form of additional analyzer cost, but
the station will also have decreased labor costs. However, no
gas self-calibrating analyzer is commercially available at this
time. The subject is evaluated in greater depth in another
report (PEDCo Environmental 1981a).
DATA PROCESSING
The only other analyzer specification that would impact I/M
costs for the state is the EPA option for automative data collec-
tion. With this option (see Subsection 7.1), data is stored on a
magnetic cassette tape or other storage medium in a tamper proof
container with an antitampering seal. The state examiner retrieves
the cassette tape while at the station doing other audit functions.
The state can then machine read the data. This system contrasts
to the more usual situation wherein the station inspector fills
out a form, the forms are collected by the state examiner, and
later keypunched to cards or tape for analysis. With the auto-
matic data collection system, no manual data entry is required by
state personnel.
The Missouri Highway Patrol has estimated that two clerks
would be required for data inventory at an annual cost of $22,595.
Assuming 1.5 less clerks would be required, the cost savings to
the state over five years would be $99,419 (at 8 percent wage
inflation each year). These cost savings would be transferred to
station owners in increased analyzer cost but decreased labor
cost for record keeping. This subject is treated in greater
depth in another report (PEDCo Environmental 1981b).
COST CHANGES TO DECENTRALIZED STATION OWNERS
Cost changes to the decentralized station owner vary with
emission analyzer selection are shown in Table B-2. Four cost
items vary with the analyzer selected, all related to analyzer
specifications.
Analyzer Cost—Basic Units
The first item is analyzer cost. Analyzer costs are shown
in previously cited Table 6-1.
77
-------
TABLE B-2. CHANGE IN I/M PROGRAM COST RELATED TO EMISSION ANALYZER
SELECTED-DECENTRALIZED STATION COST
Cost does
not vary
with analyzer
selected
Cost does
vary with
analyzer
selected
A. Start-up costs
Analyzer and related equipment
Manuals and required equipment
Mechanic training
B. Recurring costs
Equipment repair
Calibration gas and expendables
Mechanic wages (direct and overhead)
Testing
Recordkeeping
Calibration
78
-------
Analyzer Costs—Optional Features
Beyond the cost of the analyzer, reoccurring costs vary with
the analyzer specified. These are labor costs associated with
recordkeeping and gas calibration. With the automatic data
collection system, time required to record the test results and
store the forms would not be required. However, since the I/M
program is projected to be combined with the vehicle safety
inspection which requires filling out a form, cost savings would
be minimal if the recordkeeping from both programs were combined
into the automatic data collection system. Savings in labor cost
in any event would probably be relatively slight amounting to
about two to four minutes per test over the long run. In the
short run, confusion about the automatic data collection keyboard
would probably obliterate any cost savings.
Some programs require that station operators gas calibrate
their analyzers weekly. Other programs do not require any gas
calibration by the station operator relying on the monthly state
gas calibration only. Self-gas calibrating units also eliminate
the need for station operator calibration. When gas calibrations
are required, the procedure would require no more than 15 minutes
weekly. Over a 5 year program, labor time would be less than 65
hours per station. As an average of $20 per hour, increasing at
8 percent per year, costs per station over the five year period
would be $1525.
The automatic gas calibration feature for an emissions
analyzer is not commercially available. Units are under develop-
ment. EPA (EPA 1980) estimates the cost of this feature to be
$150 per analyzer. However, it is likely that it would not be
available without the other features recommended by the EPA
specifications for decentralized programs. The EPA analyzers are
estimated to cost $6820, about $2470 more than a BAR 80 and $3445
more than a BAR 74 unit (EPA 1980).
Using a simplistic analysis, assuming a five-year analyzer
life and 1000 inspection stations with analyzers, and salary
increases of 8 percent per year, cost savings to the state would
be $348,670. The increased cost of the analyzer to the stations
(assuming the total EPA specification analyzer) would be $2,470,000
over a BAR 80 unit and $3,445,000 over a BAR 74 unit. If it is
assumed that only the self-gas calibration option could be obtained
for $150 per unit (this is unlikely), the added analyzer cost
would be $150,000. To this analysis must be added the economic
costs to the consumer relating to analyzer accuracy. These costs
are described below.
79
-------
COST CHANGES TO THE VEHICLE OPERATOR
Costs to the vehicle operator from an I/M program are the
cost of the inspection and cost of repairs. Time costs are also
involved but not considered, since they are not a direct cash
outlay.
Inspection Cost
The inspection cost increase to vehicle owners is currently
estimated to be;$3.50 over the cost of the present safety inspec-
tion. The type of analyzer would influence the inspection fee
insofar as analyzer costs change inspection costs per car for the
station.
Repair cost
The repair cost to the consumer will vary depending on the
analyzer selected. As seen in Appendix A the number of incorrect
pass/fail determinatons increases with a decrease in analyzer
accuracy. Since the program stringency also decreases with
decreasing accuracy, the overall repair cost to the consumers
also decreases. This seeming contradition is due to the fact
that fewer cars are failed with an instrument of lower accuracy.
There is related cost increase to the consumer when a low
accuracy instrument is used. Those vehicles that are failed
incorrectly will require unneeded repairs.
The difference between these two cost items represents the
overall reduction in repair cost to the consumer. Using the HC
data generated in Appendix A, the example in Table B-3 demonstrates
these points. The cost data are based on an average repair cost
of $35/vehicle. (Missouri Air Conservation Commission, 1980).
One other aspect of the results shown in Table B-3 should be
noted. First, the overall cost reductions represent repair
dollars that will not be spent. At least a portion of these
monies would represent lost profits for the decentralized stations.
80
-------
TABLE B-3. CHANGE IN I/M PROGRAM COST RELATED TO EMISSION
ANALYZER SELECTED-CONSUMER COSTS FOR REPAIRS
i Specification
EPA
BAR 80
ETI/BAR 74
HC error,
%
8
14
29
Number -of
incorrectly
passing
vehicles
787
3014
14299
Repair cost
reduction, $
27,545
175,490
500,465
Number of-
incorrectly
failing
vehicles
320
3509
11745
Repair cost
increase, $
11,200
122,815
411,075
Overall cost
reduction, $
16,345
52,675
89,390
-------
REFERENCES
Missouri Air Conservation Commission. 1980. Proposed Inspection
Maintenance Program for Missouri.
U.S. Environmental Protection Agency. 1980. Analysis of the
Emission Inspection Analyzer. EPA-AA-IMS-80-5-A.
PEDCo Environmental, Inc. 1981a. Support Document for Develop-
ment of the Missouri Inspection/Maintenance Program—Quality
Assurance Procedures.
PEDCo Environmental, Inc. 1981b. Support Document for Develop-
ment of the Missouri Inspection/Maintenance Program—System
Considerations.
82
-------
QUALITY ASSURANCE PROCEDURES
-------
CONTENTS
Tables
1.0 Introduction 1
1.1 Purpose 1
1.2 Environmental Protection Agency requirements 1
1.3 Report organization • 3
References 4
2.0 Elements of a Quality Assurance Program 5
2.1 Analyzer related 5
2.2 Ongoing audit functions 12
References 15
3.0 Findings 16
3.1 Recommended quality assurance program elements 16
3.2 State personnel and equipment costs 16
References 20
Appendices
A Overview of Quality Assurance Procedures for Existing 21
I/M Programs
B California Certified Gas Blender Specifications 27
C Sample State Inspection Report Forms from Nevada 45
and California
iii
-------
TABLES
Number
2-1 Possible Elements in a Quality Assurance
Program 6
3-1 Recommended Elements of a Quality Assurance
Program 17
3-2 Estimated First Year Cost of Quality Assurance
Procedures to the State 18
IV
-------
SECTION 1.0
INTRODUCTION
1.1 PURPOSE
The purpose of this report is to provide technical informa-
tion to the State of Missouri on the range of quality assurance
(QA) procedures available for use in a mandatory decentralized
inspection/maintenance (I/M) program. This report will serve as
the technical basis for the state in adopting regulations for
quality assurance procedures.
The principal objectives of a quality assurance program in a
decentralized I/M program are:
0 To check the condition and accuracy of emissions testing
equipment in the participating repair facilities.
0 To assure that factual test results are reported on all
vehicle emission levels.
1.2 ENVIRONMENTAL PROTECTION AGENCY (EPA) REQUIREMENTS
Most of the Environmental Protection Agency's requirements
for an I/M program are contained in a 1978 EPA policy memorandum
(Hawkins 1978). Requirements relating to quality assurance for
the governing agency are that they:
0 Provide for quality control regulations and procedures
for the inspection system including required calibra-
tions of all types and minimum recordkeeping.
0 Maintain records on the calibration of testing equip-
ment.
0 Inspect each facility to check facility records, check
the calibration of the testing equipment, and observe
that proper test procedures are followed.
0 Conduct an effective program of unannounced/unscheduled
inspections both as a routine measure and as a complaint
investigation measure.
-------
0 Operate a referee station where vehicle owners may
obtain a valid test to compare to a test from a li-
censed station.
In a separate action, EPA has also suggested a 30 day minimum
frequency for gas calibration by the governing agency. EPA is
preparing guidance on gas naming but it is not available at this
time.
Section 207 of the Clean Air Act also contains elements that
relate to quality assurance. Section 207(b) is a consumer protec-
tion measure that requires vehicle manufacturers to warrant that
a 1981 or later vehicle will conform to emission standards through-
out its useful life providing that the vehicle is properly main-
tained and operated. For 5 years, all emission control devices
must be warranted including the catalyst, air pump, and electronic
controls. For the first 2 years/24,000 miles, parts of the
ignition system (spark plugs, ignition wires, and similar com-
ponents) are also covered. For Missouri consumers to take ad-
vantage of 207(b) warranty benefits, the Missouri program must
meet 207(b) program specifications. Quality assurance related
procedures included in Section 207(b) are:
0 "Equipment shall be calibrated in accordance with the
manufacturers instructions."
0 "Within one hour prior to a test, the analyzers shall
be zeroed and spanned... An electrical span check is
acceptable."
0 "The analyzers shall have been spanned and adjusted if
necessary, using gas traceable to NBS [National Bureau
of Standards] ±2 percent within one week of the test."
0 "For analyzers with a separate calibration or span
port, CO readings using calibration gas through the
probe and through the calibration port shall be made
(Federal Register 1980)."
The EPA's recommended specifications for inspection analyzers
(EPA 1980) also contain recommendations (not requirements) from
which certain quality assurance procedures can be inferred. EPA
recommends that each state implementing a decentralized I/M
program adopt the EPA specification for a computer operated
analyzer. A computer operated analyzer automatically performs
several quality assurance functions at preset intervals. If the
analyzer fails the check, it becomes inoperable. The quality
assurance checks include an electrical zero/span check (hourly),
a gas calibration (4 hours recommended, weekly minimum), leak
check (weekly), and an HC hangup check (before each test).
Because of the instruments self-calibration procedure, EPA requires
-------
only quarterly state audits with this type of analyzer as compared
to the required monthly audits. No unit meeting these recommenda-
tions is commercially available.
1.3 REPORT ORGANIZATION
Section 2 contains a description of the range of activities
that can be incorporated into a quality assurance program and
these activities are evaluated for the Missouri I/M program.
Findings are presented in Section 3.
-------
REFERENCES
Federal Register. May 22, 1980. Rules and Regulations. Section
85.2217, Vol. 45, No. 101.
Hawkins, D.G. July 17, 1978. Memorandum to EPA Regional
Administrators. Washington, D.C.
U.S. Environmental Protection Agency. September 1980. Recom-
mended Specifications for Emission Inspection Analyzers.
EPA-AA-IMS-80-5-B.
-------
SECTION 2.0
ELEMENTS OF A QUALITY ASSURANCE PROGRAM
Quality assurance procedures are necessary at both the
decentralized testing location and the state audit levels.
Possible elements are included in Table 2-1. Most quality assur-
ance procedures are performed regularly by the decentralized
station operator and audited less frequently by state personnel.
The following subsections describe the range of activities pos-
sible for each major quality assurance element.
2.1 ANALYZER RELATED
2.1.1 Electrical Zero/Span
2.1.1.1 Overview—
An electrical zero and span check is a calibration function
to assure analyzer accuracy. The only regulations applying to
this procedure are Section 207(b) requirements that stipulate
that the analyzer be zero/spanned checked within one hour prior
to each test. On most manually operated analyzers, the operator
need only press a button to instigate the check. If the analyzer
fails the check, the operator must remove it from service. On
computer operated analyzers, the check is made automatically each
hour and if the analyzer fails the check it becomes inoperative.
Examination of existing programs (Arizona, California,
Cincinnati, Nevada, New Jersey, Oregon, and Rhode Island) indi-
cates an electrical zero/span check by the operator ranging in
frequency from twice per day, to hourly, to before each test.
2.1.1.2 Evaluation—
To meet Section 207(b) requirements and EPA recommended
procedures, an electrical zero/span must be completed within one
hour prior to each test. Because of this frequency, the proce-
dure must be performed by the analyzer operator. Since the
procedure is quick and easy to perform without error, this would
be no hardship.
Many decentralized stations will not be inspecting cars
every hour. Therefore, it is recommended that Missouri adopt
regulations requiring an electrical zero/span check immediately
-------
TABLE 2-1. POSSIBLE ELEMENTS IN A QUALITY ASSURANCE PROGRAM
Element
Performed by
decentralized station
operator
Performed or
checked by
state personnel
Analyzer related
Electrical zero/span
Gas calibration
Leak check
HC hangup check
Gas naming
Internal log
Ongoing audit functions
Data audit
Undercover vehicles
Citizen complaint investigations
Challenge stations
-------
prior to each test by the analyzer operator. This regulation
would meet 207(b) requirements and eliminate the need to zero/
span check the machine during the hours when it was not in use.
If the analyzer does not pass the check, it should be taken out
of service by the operator. The state auditor should perform a
zero/span check immediately prior to the gas calibration audit.
2.1.2 Gas Calibration
2.1.2.1 Overview—
In a gas calibration procedure, gas of a known concentration
is drawn into the emissions analyzer to check analyzer accuracy.
The concentration indicated by the analyzer must be within a
specified interval of the known concentration. Regulations
applying to this function are the EPA policy requiring monthly
gas calibration, and Section 207(b) requirements that stipulate
weekly gas calibration. On ETI and BAR 74 specification analyzers,
the operator must connect hoses to the analyzer and to the span
gas or draw span gas through the probe. On BAR 80 and EPA spe-
cification analyzers, no connections are required; only a button
must be pressed. On EPA specification analyzers, the interval is
automatically set for every power on and every four hours of
testing.
Examination of gas calibration practices for existing pro-
grams show a wide range of participation and frequency. With
centralized programs, the frequency varies from five times per
day by the lead inspector (Oregon) to weekly by garage personnel
(California). In some of the centralized programs, span gases at
several concentrations are used to check calibration over a range
of expected levels. In decentralized programs, a state audit is
performed once a month in most programs. The major difference is
whether the operator must perform gas calibrations or not. In
Nevada, for example, no gas calibrations by the operator are
required, meaning that the analyzer is gas calibrated only once a
month.
2.1.2.2 Evaluation—
To meet Section 207(b) requirements, weekly gas calibrations
are required. As a practical matter, it is recommended that the
gas calibrations be performed by the operator as opposed to state
personnel. To conform with EPA policy, a monthly state audit is
required.
The EPA recommendation that gas calibrations be performed
every 4 hours would add considerable expense for the station
operator for both labor (15 to 45 minutes/day) and for calibra-
tion gas. Also, many stations may not even perform inspections
as often as every 4 hours. No quantifiable data have been pre-
sented by EPA that support a significantly greater level of
analyzer accuracy with 4 hour calibrations as opposed to weekly
-------
calibrations. Therefore, it is recommended that Missouri adopt
regulations requiring weekly gas calibrations by the decentralized
station operator and monthly gas calibrations by state personnel.
2.1.3 Leak Checks
2.1.3.1 Overview—
A leak check is a procedure where the negative pressure
portions of the analyzer are checked for air leaks which would
allow an excess of ambient air into the sample gas and therefore
result in a lower concentration reading. The only regulations
applying to the procedure are Section 207(b) requirements that
stipulate weekly leak checks. The leak check is to be made by
comparing test results from a calibration gas drawn through the
probe to calibration gas through the calibration port; discrep-
ancies of over 3 percent require repair of leaks.
All analyzers have the capability to be leak checked in the
207(b) stipulated manner although leak checks have seldom been
performed in this manner. On BAR 74 analyzers used in the Cali-
fornia program, the leak check was actually part of the accuracy
check. Calibration gas was drawn through the probe and if the
machine was not within accuracy specifications, the analyzer was
checked for leaks as one of the possible causes for inaccuracy.
There is no BAR 74 specification for leaks. It is also possible
to perform a leak check on a BAR 74 analyzer by sealing the probe
tip and turning the motor on. If low flow is indicated, there
are probably no major leaks. Machine damage can sometimes result.
All BAR 80 analyzers have a vacuum gauge to be used to check
for leaks. The probe must be designed so that a cap or tube can
be placed over the probe tip. For weekly leak checks in California
by Blue Shield station operators, the probe is sealed with the
pump on. Vacuum rises to 12 to 25 inches of mercury and the pump
is turned off. Vacuum decay time is then analyzed. If the
vacuum drops more than 10 inches of mercury in one minute, it is
evidence of a leak. State auditors on monthly inspections do not
use this procedure, however. As with BAR 74 analyzers, a calibra-
tion gas is drawn through the probe. A discrepancy of greater
than 3 percent results in a check for leaks as one of the possible
causes for inaccuracy.
EPA specification analyzers are not required to have a
vacuum gauge but are required to have inherent leak check capa-
bility. Preliminary designs have a recepticle to put the probe
in. Circuitry then performs the comparison of measured values
from the probe and calibration point.
2.1.3.2 Evaluation—
All analyzers have the capability to be leak checked accord-
ing to the 207(b) requirements. No existing programs require
leak checks performed in this manner.
8
-------
To meet the 207(b) requirements, a weekly comparison of
measured values through the calibration port and through the
probe will be necessary. This procedure, however, can be com-
bined with the gas calibration procedure described in Section
2.1.2. The calibration gas drawn through the probe can serve as
the gas calibration procedure and half of the leak check procedure.
The state auditor should also perform the 207(b) required proce-
dure on a monthly basis.
2.1.4 HC Hangup
2.1.4.1 Overview—
HC hangup refers to the process of hydrocarbon molecules
being absorbed or condensed prior to reaching the analyzer de-
tector, primarily in the sample line. There are no regulations
requiring an HC hangup check. EPA recommends use of a comput-
erized analyzer that performs an HC hangup check before every
test. On ETI, BAR 74, and BAR 80 specification units, the only
way to check for HC hangup is to read the HC concentration after
the electrical zero/span check is made. The HC concentration is
usually between 0 and 15 parts per million (ppm) before the probe
is inserted into the tailpipe. The low concentration is usually
due to permitted zero drift and hexane in the ambient air. If
the HC concentration is 20 ppm or greater and the analyzer is
correctly electrically zeroed, HC hangup has occured. On EPA
specification units, an HC hangup button must be pressed before
each test or the analyzer will not enter the testing mode. No
existing decentralized programs require HC hangup checks.
2.1.4.2 Evaluation—
HC hangup is most often a problem at high-volume testing
locations where tests are performed at frequent time intervals.
This situation will not usually be the case at a decentralized
location. The only possible regulation with an ETI, BAR 74, or
BAR 80 analyzer is to require that the operator check the HC
concentration after the electrical zero/span check but before the
test. If the value is greater than 20 ppm HC, the test procedure
would not be commenced. This regulation would be even more
difficult to enforce than the quality assurance procedures detailed
above, but is probably worthwhile because of the minimal cost to
implement.
2.1.5 Gas Naming
2.1.5.1 Overview—
Gas naming refers to specifying the gas to be used in gas
calibration. Calibration gas is required for use by state audi-
tors, and if gas calibration is performed by the decentralized
station operator, by all stations. It is also desirable to have
local repair organizations have the same gas. If the gas is not
the exact concentration specified on the container, the gas
-------
calibration cannot be performed accurately; the difference between
analyzer accuracy and improperly named gas cannot be distinguished.
The only regulation applicable to gas naming is contained in
Section 207(b) wherein it states that the span gas be traceable
to NBS gas ±2 percent of the label value.
The basic problem is that NBS gas is prohibitively expensive
for widespread use. Practices in existing programs vary widely
but can be delineated into three basic procedures:
1. Where only state auditors need gas, the state can buy
NBS gas, blend their own gas, or buy gas from a gas
blender and check it against the NBS gas (Nevada).
2. In a variation of the first group, the state can blend
gas or buy gas in sufficient quantities, check it
against NBS gas, and sell calibration gas to decentral-
ized station operators and service organizations.
3. The state publishes specifications for span gas blending.
Any gas blender who will abide by the specifications
can supply gas to the program. The state buys NBS gas
and provides a split of the NBS gas to all certified
gas blenders. With this procedure, all gas is traceable
to the same NBS gas source although it is blended by
more than one gas blender (California).
2.1.5.2 Evaluation—
It is essential that all calibration gas be traceable to a
single gas to maintain consistency between state auditors, an
estimated 1000 decentralized testing locations, and service
organizations. Section 207(b) requires that the gas be traceable
to NBS gas. The first procedure is not applicable since decen-
tralized stations will require span gas. With regard to the
second procedure, the state has no gas blending capability at
this time. Development of this capability, which would duplicate
private sector capability, would be quite costly. Because of the
large number of potential span gas users, the certified blender
approach (third procedure) would place the least burden on the
state. The role of the state would be to promulgate gas blending
specifications, buy a small quantity of NBS gas to provide splits
of certified gas, certify gas blenders, and monitor gas blender
performance with regard to quality of span gas sold and price.
The State of California has used the certified gas blender
approach for several years and is satisfied with the system.
Their gas blender specifications are included as Appendix B. The
basic procedure is as follows:
1. Obtain NBS gas certified to be within 1 percent of the
label value.
10
-------
2. Contract with an independent testing laboratory to
verify 1 percent NBS gas. Have laboratory manufacture
additional gas to within 1 percent of label value. The
testing laboratory should use a Beckman 315 instrument
or equivalent to measure the gas concentration. The
calculation is as follows:
°/ deviation = master std< value - blender reported value
where master _ value from laboratory instrumentation
standard value calibrated with NBS gas
3. Certify precision certified blenders (see Appendix B
for requirements).
4. Provide split of I percent gas to all blenders.
5. Require, through blender specifications, the quality of
gas required (see Appendix B).
6. Perform periodic spot checks of span gas blends. The
checks would be performed by an independent testing
laboratory.
7. Require the gas blender to recall, correct or replace
at their expense any gas or container out of specifica-•
tion.
2.1.6 Internal Log
2.1.6.1 Overview—
An internal log of quality assurance and maintenance activi-
ties is a useful device for the station operator and for the
state personnel auditing the program. The station operator can
use the log as an internal device to schedule and document quality
assurance and maintenance activities. State personnel can inspect
the log to determine if all required activities are being carried
out in a timely manner. Suitable activities to be recorded in
the log, by date and person performing the activity involve:
Gas calibration
Leak check
Water trap check
Particulate filter replacement
Analyzer repairs
The 1978 EPA policy memorandum requires maintaining records
on the calibration of testing equipment. All existing I/M pro-
grams require some kind of recordkeeping.
11
-------
2.1.6.2 Evaluation—
Cost to the station operator to keep an internal log for
each analyzer would be quite low. All log activities could
probably be accomplished in ten minutes per week or less. While
the log is certainly not tamperproof, it could be inspected by
state personnel to monitor quality assurance and maintenance
activities. It is recommended that the state develop and print
standard log forms and require their use at all testing locations.
2.2 ONGOING AUDIT FUNCTIONS
2.2.1 Data Audit
2.2.1.1 Overview—
One of the major shortcomings of a decentralized I/M program
as opposed to a centralized program is the inability of the state
to continuously monitor all locations. This shortcoming is
partially mitigated by careful analysis of data. The basic
procedure is to examine test result data to determine irregular-
ities by inspection station or inspector. A matrix similar to
the following could be compiled.
Average cost
% passing of repair Repair type
Station x x x
Inspector x x x
Inspector x x x
Station xxx
Inspector xxx
Inspector xxx
Insepctor xxx
Etc. xxx
Program average xxx
There are no federal regulations regarding data audit.
However, all existing I/M programs have data audit systems that
vary widely with individual program goals and constraints.
2.2.1.2 Evaluation—
This type of data analysis is extremely effective in iden-
tifying abuses of the program. Stations or individual inspectors
passing an abnormally high number of vehicles, charging an abnor-
mally high amount for repairs, and performing the same repairs on
all vehicles can be readily identified. While occasional viola-
tions would not be spotted, chronic violations could be delineated.
Missouri already uses a similar data audit program in connection
with the vehicle safety inspection program. Pass/fail data by
station are analyzed. The chief drawback to enlarging such a
system to include the emissions test data is the computer cost to
12
-------
compile the data in the matrix format. This is particularly true
of the repair type indicator that requires a normalization of
values. It is recommended that the state compile emissions test
data on percent passing and average repair costs and investigate
costs of compiling repair type data.
2.2.2 Undercover Vehicles
2.2.2.1 Overview—
The concept of using undercover vehicles involves a state
employee taking unmarked vehicles with known mechanical problems
and emission levels to a testing location for an inspection and
maintenance test. A comparison of actual and reported results
can be used to determine the integrity of the testing location
and inspector. There are no federal regulations governing this
practice. Several existing I/M programs use this technique.
2.2.2.2 Evaluation—
Use of undercover vehicles can be effective in determining
violations of the program. The state already uses this enforce-
ment tool in connection with the safety inspection program. The
main drawback to expanding the program to include emissions tests
is expense. Labor cost is involved in modifying the vehicle,
taking the vehicle to the testing location, repairing the vehicle,
filing reports, etc. A pool of vehicles that will not be rec-
ognized is also required. Because of the costs involved, the
technique is usually restricted to a followup procedure for
citizen complaints and to irregularities discovered in the data
audit. The technique can also be used for random spot checks
although probably with less cost effectiveness.
2.2.3 Citizen Complaint Investigations
2.2.3.1 Overview—
Initiation of the I/M program will doubtless cause many real
and imagined consumer transgressions. To effectively followup
these consumer complaints, personnel need to be designated to
perform the function. There are no federal regulations regarding
handling of citizen complaints. All programs have some provision
for investigating citizen complaints but the emphasis on the
activity varies widely. Most responsive of existing programs is
California where the program is under the jurisdiction of the
Consumer Affairs Bureau.
2.2.3.2 Evaluation—
Interviews with persons at existing programs indicated that
there were large numbers of consumer calls during the first year
of the program. In Phoenix (annual testing) and in Los Angeles
(change of ownership testing), the number of calls was well over
100 a day through the first year. The calls concerned questions
13
-------
about how the system worked, testing locations, hours, and com-
plaints. Since the emissions testing in Missouri will be combined
with the existing mechanical inspection, some consumer education
has already occured. However, it is reasonable to assume a
significant number of consumer calls at least during the first
year. It is recommended that at least two personnel be added to
the program to handle citizen complaint investigations. One
person would be available by phone during business hours while
the second person could assist the first with calls and also
perform field investigations.
2.2.4 Challenge Station
2.2.4.1 Overview—
A challenge station is a referee station operated by the
governing body where a motorist can obtain an emissions test to
compare to a test from a licensed station. It serves as a quality
assurance measure because the motorist can compare readings taken
by an accurate state operated facility with readings taken at a
decentralized testing location. The EPA 1978 policy memorandum
requires a challenge station for all decentralized programs.
While the primary purpose of the challenge station is to act
as a referee, the challenge station can be used for additional
tasks. In Nevada for example, challenge station personnel also
grant waivers, license inspection stations, give mechanics tests,
license mechanics, and license fleet test locations.
2.2.4.2 Evaluation—
A challenge station is an expensive quality assurance mea-
sure. However, EPA will not approve an I/M program without one.
Addition of several other functions to the station, such as in
Nevada, helps to lower the cost of this quality assurance pro-
gram.
14
-------
REFERENCES
PEDCo Environmental, Inc. 1981. Support Document for Develop-
ment of the Missouri Inspection/Maintenance Program—Emission
Analyzer Specifications. Kansas City, Missouri.
15
-------
SECTION 3.0
FINDINGS
3.1 RECOMMENDED QUALITY ASSURANCE PROGRAM ELEMENTS
Recommended elements of a quality assurance program are
shown on Table 3-1. Active participation is required by testing
location personnel and state personnel.
For the station operator, an electrical zero/span check and
HC hangup check are recommended to be performed before each test.
Weekly single point gas calibrations, leak checks, and log entries
are also recommended. Span gas should be purchased by each
station from a gas blender certified by the state.
It is recommended that state inspectors inspect each loca-
tion monthly to perform or verify analyzer related, gas naming,
and internal log quality assurance procedures. Several ongoing
audit functions are also recommended for state consideration.
These functions are data audit, undercover vehicles, citizen
complaint investigations and challenge stations. The first three
of these four functions would represent expansions of audit
programs included in the existing safety inspection program.
3.2 STATE PERSONNEL AND EQUIPMENT COSTS
Estimated first year costs, for quality assurance, accrueable
to the state are shown in Table 3-2. The Missouri Highway Patrol
estimates the need for eight motor vehicle inspectors, two data
clerks and one programmer to support the program (Missouri Air
Conservation Commission 1980).
The estimate of eight additional motor vehicle inspectors is
based on the assumption that an additional 15 minutes inspection
time, over the inspection time for the safety inspection program,
would be required to perform emission test related inspections
for quality assurance. Fifteen minutes is an adequate period of
time to perform the quality assurance procedures recommended.
First year costs for the additional motor vehicle inspections was
$150,500 excluding capital costs.
16
-------
TABLE 3-1. RECOMMENDED ELEMENTS OF A QUALITY ASSURANCE PROGRAM
Element
Frequency
Performed by
decentralized
station operator
Performed or
checked by
state personnel
Analyzer related
Electrical zero/span
Gas calibration
Leak check
HC hangup check
Gas naming
Internal log
Continuous audit functions
Data audit
Undercover vehicles
Citizen complaint investigations
Challenge station
Before each test
Weekly
Weekly
Before each test
As required
Weekly minimum
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Continuously
Continuously
Continuously
Continuously
17
-------
TABLE 3-2. ESTIMATED FIRST YEAR COST OF QUALITY ASSURANCE
PROCEDURES TO THE STATE
Salaries
Inspectors (8 persons) $158,488
Data clerks (10% of time) 2,260
Programmer (10% of time) 2,200
Citizen complaints (2 persons) 39,622
Challenge station operators (4 persons) • 49,244
Equipment and Expendables
Emission analyzers for challenge station
(3 analyzers amortized over 5 years) 3,300
Calibration gas for inspection and challenge station
(80 bottles at $70/bottle) 5,600
Calibration related equipment (24 sets for inspectors + 2 sets for
challenge station + 4 spares = 30 sets amortized over 5 years) 900
Challenge station costs:
First year lease cost 14,000
Improvements (amortized over 5 years) 7,000
Utilities, insurance 3,000
CPU time for data analysis 4,000
$319,614
Sources: Missouri Highway Patrol 1981.
Missouri Air Conservation Comission 1980.
Missouri Department of Natural Resources 1981.
Environmental Tectonics Corporation 1980.
18
-------
The two clerks and programmer would have functions that
included other activities as well as quality assurance related
functions. Their quality assurance related duties would chiefly
be limited to the data audit functions. Since no additional data
would be gathered for the audit check beyond that already being
collected for program evaluation, their time would be limited to
running the audit computer program and analysis of results.
It is recommended that at least two additional personnel be
retained to handle citizen complaints. One of these persons
would be responsible for handling all citizen complaints sub-
mitted by phone, mail, or in person. The second person would be
responsible for assisting the first person and for field investi-
gations of citizen complaints, including the use of undercover
vehicles. This recommendation is based on reports from existing
programs, particularly in Los Angeles (change of ownership testing)
and Phoenix (annual testing), that recorded well over 100 phone
calls a day during the first year of program implementation.
The Missouri Department of Natural Resources (Missouri
Department of Natural Resources 1981) estimates the need for four
persons to service the challenge station. Two persons would be
at the challenge station continuously while the other two persons
would perform quality assurance checks at the testing locations.
Equipment and expendables include costs related to the
challenge station (rent, improvements, utilities, emission ana-
lyzers, and calibration equipment) and to the inspectors (cali-
bration gas and related equipment).
Based on these data, estimated costs to the state in the
first year for quality assurance would be $319,600, or about 32C
per car assuming 1,000,000 vehicles are inspected each year.
19
-------
REFERENCES
Environmental Tectonics Corporation. 1980. Spare Parts List.
Missouri Air Conservation Commission and Missouri State Highway
Patrol. 1980. Proposed Inspection/Maintenance Program for
Missouri. Final Report.
Missouri Department of Natural Resources, Air Pollution Control
Program. Telephone conservation with C. Lamb.
Missouri Highway Patrol. 1981. Telephone conversation with
Sergeant L. Walker.
20
-------
APPENDIX A
OVERVIEW OF QUALITY ASSURANCE
PROCEDURES FOR EXISTING
I/M PROGRAMS
21
-------
Following are brief summarizations of the QA tasks that are
followed by the existing I/M programs.
Each month, state officials in New Jersey calibrate the
analyzers at the state inspection lanes. In addition, New Jersey
officials visit each certified reinspection station at least once
every 2 months in order to verify the accuracy of the analyzers
and to inspect records. The officials look at both the recorded
emission levels and the charges to the customer in order to
determine if proper repairs are being performed. In some cases
the officials will reinspect vehicles with unusual or question-
able repairs. New Jersey also independently surveys about 12,000
vehicles to gather additional data about the program. Some of
these data can be used for QA analysis. To aid in the quality of
inspections and repairs, New Jersey also provides garages with
specifications for portable analyzers as well as a list of ana-
lyzers which comply with these requirements.
Arizona performs several types of quality assurance tasks.
Test data are analyzed to verify the contractors' charges to the
state. Officials also survey the stations and verify the accuracy
of the analyzers used at the inspection lanes at least once every
two weeks. The contractor calibrates the equipment on a weekly
basis. Officials verify analyzer accuracy at the fleet inspec-
tion stations at least once every 90 days. In addition, in
Arizona a repair facility may voluntarily register an emission
analyzer with the state. These registered analyzers are checked
for accuracy initially upon registration and at least once each
90 days thereafter.
22
-------
Like Arizona, California also performs surveillance of the
contractor and fleet testing facilities. The contractor stations
are inspected every 2 weeks for analyzer and inspection accuracy
as well as for proper housekeeping. In addition, other equipment
such as the report printers, data entry terminals, and ambient
carbon monoxide (CO) monitors are checked for calibration and
correct system operation. A leak check of the entire sampling
system is also conducted. Fleets are inspected every two months
for analyzer accuracy and proper completion of forms. In addi-
tion, the fleets may be asked to demonstrate certain repair and
diagnostic procedures. California also submits selected inspected
vehicles for reinspection at the fleet stations. The contractor
and the fleets are required to calibrate the analyzers weekly.
Nevada also devotes considerable time to quality assurance.
Officials visit each inspection station at least once per month
in order to verify the accuracy of the emission analyzer and to
collect records. Some of the records are then examined in order
to determine the reasonableness of the charges and repairs.
During the visits, the officials check to see that the station
has current service manuals with correct tune-up specifications.
Nevada also performs spot checks on some of the inspection stations,
An unidentified person will have an inspection performed on an
incorrectly operating car, such as a vehicle with a spark plug
wire removed. Garages will usually be investigated in this
manner as a result of complaints or challenge station checks,
although Nevada tries to spotcheck each garage at least twice a
23
-------
year. Additionally, Nevada requires that all waiver cases first
be checked by an official at the challenge station before approval.
This requirement helps to identify garages that need to be inves-
tigated. At the end of 1979, Nevada had revoked the licenses of
6 stations because of failure to perform the inspections correctly.
In addition, 15 to 20 percent of the analyzers were red-tagged
each month. When an analyzer is red-tagged, the state confiscates
the forms and the analyzer must be repaired (or calibrated)
before the station may resume inspections. (To overcome the
possibility of not being able to conduct inspections, many garages
have more than one analyzer). On the whole, Nevada officials
feel that the garages are doing a good job.
Oregon has always been concerned about good quality assurance.
For the first few years of the program, analyzers were calibrated
hourly and the stations were visited frequently by DEQ inspectors.
Now the lead inspector calibrates the analyzer hourly during the
morning when they are warming up and then every three hours after
that, or more frequently if they seem to require it (a minimum of
five times per day). It was reported that the analyzers hold
calibration very well. Each station has at least one extra
analyzer if difficulty arises. All stations are visited at least
once a week by a DEQ engineer/supervisor and the 50 fleet inspec-
tion stations are visited at least once a month. In addition, an
unannounced calibration visit is made to the stations monthly,
featuring cross reference testing of analyzers from different
24
-------
stations. As a result of these precautions, Oregon has had very
few quality assurance problems at the lanes. However, Oregon
officials are still concerned over the quality of the repairs and
feel that the program may benefit from closer control of the
retests. (Oregon has unlimited retests.)
Rhode Island officials make monthly visits to the licensed
garages and have the station personnel demonstrate a calibration
of the analyzers. While they are at the garages, they check the
calibration records and collect the emission test reporting
forms. (The garages must calibrate the analyzers weekly.) In
addition, some state vehicles are equipped with emission analyzers
which can be used in the roadside safety checks. In 1979 emis-
sions were checked in approximately 5,000 of the 26,000 roadside
checks.
In 1979 Rhode Island suspended the licenses of 13 garages
for violating the inspection requirements. However, all of these
suspensions were for improper safety inspections and not speci-
fically emission inspections. Officials in Rhode Island's inspec-
tion department report few emission related problems, although
there is little accurate data on the emission failure rate.
However, in the monthly garage inspections, officials note that
14 percent of the analyzers are initially out of calibration.
After the garages demonstrate a calibration, about 3 percent of
the analyzers are still out of specification (plus or minus 5
percent).
25
-------
In Cincinnati, the analyzers are calibrated every month as
part of a service agreement with the manufacturer. Cincinnati
performs few additional quality assurance tasks and has experi-
enced problems with large fluctuations in failure rates. How-
ever, these problems are minor in comparison with the enforcement
problems in Cincinnati.
In New York, the analyzers are calibrated every month by the
contractor. The contractor also picks up the data which is
recorded on a cassette tape which is located in the analyzer.
The data is then further reduced by the contractor. The state
inspects the stations bi-monthly and examines the analyzer and
its calibration records along with records identifing the vehicles
that have been inspected and the readings of hydrocarbon (HC) and
carbon monoxide (CO) from the emissions tests.
26
-------
APPENDIX B
CALIFORNIA CERTIFIED GAS
BLENDER SPECIFICATIONS
27
-------
OF CAlitO»NIA--AG»ICUlTUil AN0 SCIVICES AOENCY
BJMUNO a MOWII ML
^ BUREAU OF AUTOMOTIVE REPAIR
3116 MADSHAW tOAO. SACKAMtNTO, CAUF »M27
PHONE: (916) 322-2MO
Distribution: PRECISION GAS BLENDERS
Subject: REQUESTS FOR GAS BLENDER CERTIFICATION
Gentlemen:
Based on pending regulations, all precision gas blenders desiring
to supply exhaust analyzer calibration gas to approximately 10,000
official stations in California must be certified by the Bureau of
Automotive Repair (BAR).
Your response to the proposed California gas standard for use in
calibration of HC/CO infra-red emission analyzers ha* been very
encouraging. We have consolidated the comments and *ttgge*ti«>n*
from the many interested people in order to draft the finalised
version that is attached. This resulting standard, which ha* bean
adopted by the Bureau, represents the requirement* for a uoifora)
containerized gas blend to be authorized for use in Class A Official
Pollution Control Stations in California.
In accordance with the standard, your procedure for obtaining BAR
certification is as follows:
1. Submit documentation required by the attached standard
to the Bureau of Automotive Repair, 3116 Bradshaw Road,
Sacramento, California 95827.
2. After receiving notification of acceptance of this
material by BAR, obtain a sample of the master gas
blend from a laboratory to be designated by the
Bureau.
3. Submit two production cylinders of your calibration gas
blended in accordance with the standard to the laboratory
supplying the master gas sample or other test facility
indicated by the Bureau. The analysis of your gas must
be within 2 percent of the label values with the maater
gas as a reference.
4. Forward to the Bureau a copy of the laboratory analysis
indicating that your blend conforms to the standard.
28
-------
PRECISION GAS BLENDERS
March 5, 1975
Page 2
5. Providing all documentation and test data is in con-
formance with the standard, the BAR may isaue your
certification and include your organization on the
approved blender list. , •
We urge you to request certification of your product by submitting
the required documentation as soon as possible. Subsequently, in-
dustry and calibration gas users will be informed of all approved
gas blenders on a continuing basis.
Sincerely,
-^XC''
ROBERT C. ALEXANDER
Chief
RCA/DEG/jml
Attachment - Calibration Gas Standard
29
-------
STANDARD FOR CALIBRATION GAS
FOR HC/CO EMISSION TESTERS
SECTION 1
GENERAL PROVISIONS
Infra-red hydrocarbon and carbon monoxide emission testers require calibration
at designated intervals to maintain validity of spark ignition engige emission
test results. This standard applies to the calibration gas and portable
container to be used for calibrating HC/CO emission testers in the state of
California Class A Pollution Control Inspection Stations.
1.1 Scope
The scope of this standard encompasses:
Section 1 - General Provisions and Definitions
Section 2 - Requirements
Section 3 - Gas Blender Certification
Section 4 - Quality Assurance
Appendix A - Reference Documents List
Appendix B - Blended Gas Label
Appendix C - Application For Certification
1.2 Definitions
1.2.1 Blend Tolerance - the maximum percent deviated from the desired
concentration of the gas (blend of HC and CO) components.
1.2.2 Analytical Accuracy - the relative percent deviation of the blender's
reported concentrations of HC and CO gas components from the BAR
master standards.
Example: % Deviation = *Master Std. Value - Blender Reported Value Y 1nn
*Master Std. Value* 'uu
*Value from laboratory instrumentation calibrated with BAR master
standard.
30
-------
1.2.3 Calibration span gas - a blend of hydrocarbon (HC) and carbon monoxide
(CO) gases with nitrogen (N2) as the balance or carrier gas used for
adjusting HC/CO exhaust analyzer test instruments to a predetermined
calibration value for the measurement scale used.
1.2.4 Chemically pure gas - a high quality gas known as CP grade in the
industry. Typical minimum purity shall be: propane, 99.5%; carbon
monoxide, 99.8*,; nitrogen, 99.997%.
1.2.5 Qualified gas blender - a precision gas blender approved by the
Bureau of Automotive Repair capable of providing continuous quality
gas blends within the limits of this standard.
1.2.6 Authorized gas distributor - a company authorized by a qualified gas
blender to distribute (in approved containers) gas blends produced by
the qualified gas blender meeting all requirements of this standard.
1.2.6.1 Repackager - an approved gas blender or his authorized representative
responsible to assure by laboratory analysis that the accuracy of the
gas blend in repackaged or refilled containers complies with the
requirements of Section 2 of this standard.
1.2.7 BAR - Bureau of Automotive Repair (approving agency), State of
California, 3116 Bradshaw Road, Sacramento, California, 95827.
1.2.8 California BAR master standard - a primary gas blend standard
established by BAR for use in California and traceable to the NBS
through ARB standards.
1.2.9 Traceable - to be able to compare on a laboratory quality gas analyzer,
calibrated with a master standard blend, a sample gas blend to the
master or primary standard gas blend within stated tolerance limits.
31
-------
SECTION 2
2.1 Requirements
21.1 The. calibration span gas blend and portable container units shall
comprise the following items:
a. Gas blend - calibration span gas shall have a blend tolerance
not exceeding 5% and shall consist of a blend of 3,000 ppm
9
+J50 ppm of propane (HC) and 8% +_.4% carbon monoxide (CO) in an
inert balance gas of vaporized liquid nitrogen (Ng). The ana-
lytical accuracy of the blend shall be certified to be within
+2% of label values. Direct comparison shall be made to the
California BAR master standard blend traceable through Air
Resources Board (ARB) as outlined in paragraph 4.1.1 of this
standard.
a.I Component gas (HC and CO) purity shall be stated in the
analysis and have a guaranteed chemically pure quality as
defined in paragraph 1.2.4.
a.2 Balance gas purity - the Np balance gas shall be chemically
pure grade as defined in paragraph 1.2.4.
b. Containers (standard type) - Requirements for the recommended
gas blend container are as follows:
b.l Nonrefill able - a low pressure (260 psi nominal) portable
container approved by BAR.
b.2 ASME Unfired Pressure Vessel Code, Section VIII (260 psi
service, 325 psi proof, 500 psi burst) minimums.
b.3 Federal requirements - the applicable Department of Trans-
portation (DOT) specification for shipping containers of
compressed gas shall be complied with. See Appendix A.
32
-------
b.4 Nominal size - 750 cubic inches +5% (approximately 9 inches
inside diameter by 16 inches high overall, providing an
equivalent water capacity not to exceed 55 Ibs.).
b.5 Shutoff valve approved by CGA for gas service at pressure
specified in 2.1.1.b.l above with threaded male outlet or
adaptor to accept a 1/4" AN fitting to connect gauge,
0
regulator and discharge hose assembly.
b.6 Safety shield on container top to protect valve.
b.7 Container color - white enamel, exterior grade compatible
with all types of automotive service operating environments.
b.S Container material and compatibility - any container material,
weld flux, antiseize compound, paint, etc., shall not be used
in container assembly or charging equipment which could cause
contamination or degradation to or is incompatible with the
gas blend. All containers must comply with the applicable
California and Federal DOT, OSHA, ASME requirements and
California Division of Industrial Safety and must be approved
by the BAR.
b.9 Container precharging preparation - prior to blended gas
charging, each container will be purged with vaporized liquid
nitrogen or the blended gas for a minimum of two minutes
followed by evacuation to industry accepted conditions.
b.10 Stability - container shall be designed to be stored in an
upright position.
c. Containers - special types - Gas blenders or suppliers
organized to provide calibration gas in higher pressure
rechargeable containers shall submit a request to the BAR for
approval. The request shall include notation of the applicable
DOT and ASME code compliances, details of recommended regulator
33
-------
compatible with approved emission testers, CGA valve, con-
tainer size and weight, and quality assurance tests conducted.
d. Safety legend - caution notes shall comply with the applicable
DOT and OSHA regulations.
e. Identification - a gummed label acceptable to BAR, indicating
in ink, gas blender, DOT #, gas mixture analysis, certified
?
analytical accuracy tolerance (%), date filled and lot number,
shall be securely affixed to upper side of container. NOTE:
Gummed labels (see Appendix B) may be procured from BAR, 3116
Bradshaw Road, Sacramento, California, at time of gas blender
certification.
34
-------
SECTION 3
3.1 Gas Blender Certification
a. Gas blenders supplying calibration gas for use 1n the state of
California emissions testing program must be certified by the BAR.
The gas blender's capability to produce precision gas blends
and maintain analytical accuracy within +2% of the California
~~ 9
BAR master standard must be demonstrated to BAR in the appli-
cation for certification (Section 3.2).
3.2 Application for Certification
3.2.1 General Data and Product Sample Required
a. Submittal package: application forms shall be completed and
submitted to the Department of Consumer Affairs, Bureau of
Automotive Repair, 3116 Bradshaw Road, Sacramento, California,
95827. The application must be signed by an officer of the
gas blending company.
a.l Descriptive information:
1. Application form
2. Gas blenders facility description
3. Instrumentation utilized
4. Production capabilities
5. Estimated selling price (to user) per container of
blended gas in lots of 1-10, 1-20, 100 up, etc.)
6. Business status report: Gas blenders and distributors
shall submit the following information with their request
for certification:
a. Indication that the applicant is a bona fide gas
•blender or distributor of precisions gas blends.
35
-------
b. Evidence that the applicant is either a California
corporation or a registered out-of-state corporation.
c. Annual sales volume during the previous fiscal year.
7. Marketing plan: A marketing plan shall be submitted
which includes the following explicit information as a
minimum:
f
a. Statewide distribution methods preferably including
toll free telephone systems for emergency conditions.
b. A clearly defined method of disposition and replace-
ment of nonconforming products.
c. Gas filled container lot control details.
d. Provisions for product delivery to local distributors
within 30 days of certification.
8. Liability bonding information ($300,000 minimum per
occurrence): A copy of product liability insurance which
shows adequate protection in catastrophic failure con-
ditions (container or valve rupture, noxious gas leakage,
etc.) shall be submitted. Evidence of distributor bonding
coverage throughout the state shall be provided.
9. Container information and manufacturer's test results
(source, type, handling procedure, and lot control of
containerized gas units for statewide traceability).
10. Assurance that compliance with all applicable OSHA/Cal
OSHA standards have been made.
11. Brief description of facility quality control program.
a.2 Product sample: Two samples of blended gas (in approved
containers) from two lots shall be submitted to a laboratory
designated by the bureau for comparison with the California BAR
36
-------
HC/CO blend in accordance with Section 4; • Comparison of the
blended samples to the master standard gas on BAR designated
laboratory equipment must be within +2% of hte label value to
satisfy the analytical accuracy and traceability requirements
of this standard.
37
-------
SECTION 4
4.1 Quality Assurance - Initial certification of gas blenders and all gas
blend procurement will be dependent on assurance that requirements of
Section 2 of this standard have been satisfied. Test data noted below
shall be provided by the gas blender.
a. Sample span gas blends analysis test results compared to the
*
California BAR master standard gas as a reference as tested on
a laboratory quality gas analyzer by the ARB laboratory or a
laboratory specified by the BAR. Component and balance gas
manufacturer's purity certification shall demonstrate compliance
with the applicable paragraphs of Section 2.
b. Container manufacturers' data indicating compliance with 2.1.1.b
through 2.1.1.b.lO shall be submitted. Pressure test data and
assurance of compliance to applicable DOT, ASME, and OSHA
specifications shall be provided.
NOTE: Periodic spot retesting of span gas blends after acceptance may
be made by BAR. It will be the gas blender's responsibility
to recall, correct, or replace at their expense any blended
gas or container assembly found by the bureau to be deteriorated
or out of specification for any reason within one year after
purchase. The BAR standards for judgements in these matters
will be traceable to the Air Resources Board master standards.
38
-------
REFERENCE DOCUMENTS LIST
CALIBRATION GAS STANDARD
Documents listed for reference and additional sources of information.
Calibration Exhaust Emission Standards and Test Procedures for 1971
and Subsequent Model Gasoline Powered Motor Vehicles Under 6001 Pounds
Gross Vehicle Weight: California Air Resources Board, Sacramento,
California, November 20, 1968.
2. Control of Air Pollution from New Motor Vehicles and New Motor
Vehicle Engines: Federal Register Volume 33, Number 108, Part II
July 4, 1968.
3. Federal Register Volume 36, Number 105, May 29, 1971.
4. Code of Federal Regulations, Department of Transportation.
5. ASME Unfired Pressure Vessel Code, Section VIII.
39
-------
BLENDED GAS LABEL
CERTIFIED
No 01)053
DOT-NO.
77M-20 (3-75)
40
-------
APPLICATION FOR CERTIFICATION
1. General: Tab A is a sampbe certification application form. Broad
guidelines are contained in the following paragraphs. Tab B is a
summary checklist to be completed and certified by the applicant
and/or a laboratory acceptable to the bureau.
a. Bureau Policy: The bureau does not intend to undertake to
sponsor development or research testing of gas blending
operations but instead will confirm the blenders' test data
according to the evaluation procedures outlined in Sections 3
and 4.
b. Proprietary Information: The bureau will handle all pro-
prietary technical data contained in the application as
confidential.
c. Additional Information: If space provided on the application
is insufficient, enter "See attached sheet #1 or #2, etc." and
use a separate sheet for each item.
d. Completeness: The form should be filled in as completely as
possible. However, if information on some items is not avail-
able, so indicate and state the reasons therefore.
e. Cost Estimates: Cost estimates should be based on a quantity
production type of operation. Wholesale and retail cost
estimates are costs to the user rather than manufacturing costs.
f. Sale of Company Assets: The bureau shall be notified 1n writing
90 days prior to the sale of a company. The new owners must
submit a new application .for certification and obtain written
bureau approval.
41
-------
SAMPLE APPLICATION FORM FOR GAS BLENDER CERTIFICATE
1. Date of Application
2. The authorized representative of
(company, corporation, or subsidiary) hereby applies for certification
as an approved calibration gas blender for the state of California
Auto Emissions Control Program.
Authorized Signature
Full Company Name
Address
Phone Number
3. Location of major distribution centers (in California),
4. Estimated date containerized blended gas will be available
'42
-------
EVALUATION DATA CHECKLIST
(Submit in Order Shown)
1. APPLICATION FOR CERTIFICATION
2. DESCRIPTIVE INFORMATION
A. The gas blender's facility description;
B. Instrumentation utilized;
C. Production capabilities;
D. Estimated retail cost per container to user (in lots of
1-10, 10-20, 100-up, etc.) (may be handled by letter included
within the submittal package);
E. Business status report;
F. Marketing plan information in accordance with paragraph
3.2.1.a.1.7;
G. Copies of insurance documents;
H. Container information and manufacturer's test results;
I. OSHA compliance statement; and
J. Quality control plan.
3. TEST RESULTS - Comparative analysis of samples by number to the
California master standard with percent deviation should be tabulated.
NOTE: Include test conditions and instruments utilized by model and
manufacturer.
43
-------
ADDRESSES OF CERTIFIED CALIBRATION GAS BLENDERS
The following are approved suppliers of calibration gases for exhaust
analyzers as of January 20, 1980. Use this list to update page 48 of
your MVPC handbook:
Airco Industrial Gases Division
1588 Doolittle Drive
San Leandro, California 94577
Liquid Carbonic Corporation
5700 South Alameda Street
Los Angeles, California 90058
Horiba Instruments, Inc.
Irvine Industrial Complex
Santa Ana, California 92705
Union Carbide (Linde Division)
19200 Hawthorne Blvd.
Torrance, California 90503
Scott Environmental Tech., Inc.
2600 Cajon Blvd.
San Bernardino, California 92411
Matheson Gas Products
8800 Utica Street
Cucamonga, California 91730
Air Products and Chemicals, Inc.
23320 So. Alameda Street
Long Beach, California 90810
44
-------
APPENDIX C
SAMPLE STATE INSPECTION REPORT FORMS
FROM NEVADA AND CALIFORNIA
45
-------
AUDIT FORM FOR NEVADA
Procedures for Conducting an Authorized Station Inspection
1. Use authorized station checklist as follows:
a. Check if all licenses are displayed in a conspicuous
place, under glass or other transparent material.
b. Check if a sign indicating the set fee or hourly rate
is posted in a conspicuous place and meets the require-
ments listed on the checklist.
c. Record the number of inspectors.
d. Inspect the available manuals for adequate information
and list the specification manuals.
e. Inspect the tune-up equipment for condition and operating
capability.
f. Check the emission reports records for completeness
and availability.
g. Check for availability and proper use of compliance
certificates.
2. Use the following procedure to check the calibration of the
exhaust gas analyzer.
a. Have operator warm up and calibrate the equipment.
b. Using the plastic bottle with holes in it, insert the
sample probe through the bottom of the bottle and the
gas sample valve stem in the top. Release just enough
gas to pressurize the bottle for 3 or 4 seconds.
Record readings on check sheet. Transfer gauge and
valve to the second gas sample and repeat above.
Compare recorded readings with the sample gas values.
Readings should be within ±10%; 5% high or 5% low.
46
-------
B
DATE.
NO. OF I:;£?if:C7CRS
Yes
I
f~
Tar,
Station
Inspectors
\: 1st ion si on posted
erenres available
State exhaust emission standards
Specification manual
Titles - :
e-uo ecuiu-ent
the ir:r~ection records available and coiriolete
-..he Z-rrtif icate of Ccroliance forms available
'.re *-r:ev filled out orcserlv
^r/; s
|
1
I
._ .. __j
p-^M-e of infra red ecrui-.rrr.ent
" "
p-^M
I
So
Kunier
*•****************;
Tolerances
(•rrelation factor_
Propane
PPM
*. ********************** A********** ****************
Calibration Test
1st test reading: CO
% HC
Hi
Standard
Low
2nd test reading: CO_
3rd test reading: CO
4th test readina: CO
% HC
PPM
% HC
% HC
PPM
Hi
Standard
Low
I
I
I
OUT OF SERVICE
Yes
;ASON: Failed to j^ass calibration test
Approved Inspector not employed
Failure to keep bond in force
:Licer.se under • revocation_
rFailure to renew license_
:Other
-.: r. * * * * i * i * * * i * * i * * * * * A * * * * * ;
*****************************************************
1
st Certificate of CoTTtnliance issued on
19
Tnis equipment cannot be used for the issuance of a certificate of compliance until released
an agent of the Emission Control Section of the Department of Motor Vehicles. 386-5356.
I
:ation Authorized Representative_
•nission Control Officer
Back in Service: Date
19
Time
PM
mission Ccntrol Officer
47
-------
AUDIT FORM FOR CALIFORNIA
M.V.I..P. INSPECTION REPORT
FACILITY NAME
TYPE INSPECTION
- CH INITIAL
CU FOLLOW UP
HH PERIODIC
n
n
D
COMMERCIAL FLEET
DEALER FLEET
COLLECTIVE FLEET
QUALIFIED
FLEET NO.
MVPC NO.
ADDRESS
ARD NO. DATE
CITY ZIP
PHONE NO. NO. OF CERTIFICATES ISSUED PER WEEK
OWNERS NAME OR RESPONSI BLE MANAGING EMPLOYEE (PRJNT>
ANALYZER CALIBRATION INFORMATION
BRAND NAME
MODEL
SERIAL NO.
C3/C6 FACTOR
DATE LAST CALIBRATED
CORRECTED CAL GAS VALUES
HC
±
PPM PPDP v FA<"rr>p . PP
100 PPM AfTFPTARI F PANIRF TO PP
M HEX
M
ANALYZER ACCURACY CHECK
METER READINGS ACCEPTABLE
HC
HC
CO
CO
THRU CAL PORT PPM [ | YES
THRU PRORF PPM | | YF<
THRU CAL PORT % | ] YE!
THRU PROBE , " I I YE
GENERAL INSPECTION ITEMS
i.
2.
3.
4.
5.
6.
7.
8.
OFFICIAL SIGN DISPLAYED
CURRENT STA. LIC/ REG POSTED
CURRENT EMPLOYED LIC. INST./ADS.
CURRENT EMP. INST./ ADS. LIC. (s) DISPLAYED
PRICES POSTED
INSPECTION PROCEDURES POSTED
REG. INSPECTION STEPS FOLLOWED
CERTIFICATES ISSUED CORRECTLY
r~| NO
r~i NO_
> 1 1 NO
5 1 | NO
YES
NO
rn RDTTI p VAI IIP (>/„
+ A% RANRF TO %
REMARKS I-R ANALYZER ONLY
9. RECORD OF NOx STICKER MAINTAINED
10. RECORD OF CERT/ WORK ORDERS MAINTAINED
11. REQ. TOOLS & EQUIP. AVAIL. & SERVICEABLE
12. BAR BULLETINS, MECH. HANDBOOK CURRENT
13. TUNE UP SPECS. & SERV. DATA CURRENT
14. HAVE RECEIVED FLEET TRAINING
15.
16.
YES
NO
COMMENTS:
QUALI RED/
QUALIFIED /
INSPECTOR'S
CLASS 'A' MECH. QUALIFICATION NO. CLASS 'A' NO.
CLASS 'A' MECH QUALIFICATION NO. CLASS 'A' NO.
NAME AND I.D. NO.
QUALIFIED / CLASS 'A' MECH. QUALIFICATION NO. CLASS 'A' NO.
QUALIFIED / CLASS 'A' MECH QUALIFICATION NO. CLASS 'A* NO.
OWNER /MANAGER DATE
78M-27 (1/79)
48
TMP -OSP
-------
INSPECTION STATION REQUIREMENTS
-------
CONTENTS
1.0 Introduction 1
1.1 Purpose 1
1.2 Requirements 1
References 3
2.0 Inspection Station Requirements 4
2.1 Space requirements 4
2.2 Equipment requirements 5
2.3 Personnel requirements 6
References 8
iii
-------
SECTION 1.0
INTRODUCTION
1.1 PURPOSE
The purpose of this report is to provide technical informa-
tion to the State of Missouri on inspection station requirements
for a mandatory decentralized inspection/maintenance (I/M) program.
This report will serve as the technical basis for the state in
adopting regulations for inspection station requirements relative
to spatial, equipment, and personnel needs.
1.2 REQUIREMENTS
1.2.1 Environmental Protection Agency Requirements
Environmental Protection Agency (EPA) requirements per-
taining to inspection station requirements are contained in a
1978 EPA policy memorandum (Hawkins 1978) and later reiterated in
a 1982 document (Gray 1981). Requirements are that a private
inspection facility must:
0 Be licensed
0 Have an emission analyzer that meets state/local spe-
cifications
0 Employ an inspector with demonstrated proficiency in
I/M rules and regulations, test procedures, analyzer
use, quality assurance procedures and recordkeeping
0 Maintain and submit records related to vehicle inspec-
tions
0 Submit to unscheduled/unannounced audits
There are no Section 207(b) provisions relative to inspection
station requirements.
-------
1.2.2 State of Missouri Requirements
The State of Missouri has inspection station requirements
for its vehicle safety inspection program (Missouri State Highway
Patrol 1980). Because the I/M program will be coupled with the
safety inspection program, requirements for that program are
listed below. Minimum requirements are that a facility must:
0 Have an inspection area within an enclosed building of
sufficient length and width to accommodate a full size
domestically made passenger vehicle.
0 Be sufficiently lit, adequately heated, and properly
ventilated.
0 Have a floor that is substantially level and constructed
of a hard material, be kept clean, and be free from
excessive dirt, grease, and loose material.
0 Do no major mechanical repair work in the inspection
area during normal business hours if the station has
only one inspection area.
0 Have the following equipment - brake performance test
equipment, brake lining gauge, brake pad gauge, ball
joint gauge, lift or jack, scraper, measuring device,
and punch.
0 Employ a licensed inspection/mechanic. To obtain a
license, an applicant must; have one year practical
experience as an automotive applicant or have completed
a course of vocational instruction in automotive me-
chanics from a generally recognized educational insti-
tution; pass a written test; and demonstrate practical
knowledge by inspecting a vehicle.
-------
REFERENCES
Gray, C. L. 1981. Inspection/Maintenance Program 1982 SIP
Processing. Ann Arbor, Michigan.
Hawkins, D. G. 1978. Memorandum to Environmental Protection
Agency Regional Administrators. Washington, D.C.
Missouri State Highway Patrol. 1980. Missouri Motor Vehicle
Inspection Regulations. Jefferson City, Missouri.
-------
SECTION 2.0
INSPECTION STATION REQUIREMENTS
2.1 SPACE REQUIREMENTS
2.1.1 Overview
Spatial requirements for inspection stations vary signifi-
cantly. For Blue Shield stations in California, no physical
dimensions are required but the work area must be "approved by
the bureau". Other states such as Nevada give no physical di-
mensions but rely on descriptive terms such as "sufficient space
to test and inspect". As noted in Section 1.0, Missouri now uses
a similar approach in the safety inspection program specifying
that the test area by "of sufficient length and width to accom-
modate a full size domestic vehicle". Of existing programs, only
Rhode Island specifies minimum dimensions of 13 feet by 25 feet
with mechanical headlight aimers or 13 feet by 45 feet without
mechanical headlight aimers.
Spatial requirements also vary with respect to specification
that the testing area be in an enclosed building, specification
of floor material, and specification of heating and ventilation.
2.1.2 Evaluation
The Missouri space requirements for the safety program are
as stringent as any found in existing safety or I/M programs
except that the size of the testing area is not dimensioned in
feet. However the existing regulation requiring that the inspec-
tion area be "of sufficient length and width to accommodate a
full size domestic made passenger vehicle" could be changed to
"of sufficient length and width to test and inspect a full size
domestic made passenger vehicle". This wording would require not
only sufficient space to accommodate a full size vehicle but also
would insure access to the tailpipe for insertion of the probe,
and access to the hood area for needed adjustments or repairs.
It is recommended that Missouri retain their existing require-
ments with regard to enclosure, floor material, and heating and
ventilation. Beyond the existing needs of the safety program,
emission analyzers are sensitive to dust, temperature variation,
and extreme temperatures (PEDCo 1981). The existing requirements
-------
would minimize the effects of these factors that are detrimental
to analyzer accuracy.
2.2 EQUIPMENT REQUIREMENTS
2.2.1 Overview
Most existing decentralized programs have a list of required
equipment. Rhode Island requires an emission analyzer only. The
items on the list vary principally by whether a safety inspection
is included with the emissions test. Emissions related equipment
requirements are related to adjustments or repairs to carburetion
and ignition systems, i.e., tuneup functions. Items included
are:
0 Ignition analyzer-oscilloscope
0 Ammeter
0 Voltmeter
0 Tachometer
0 Vacuum gauge
0 Pressure gauge
0 Cam angle dwell meter
0 Ignition timing light
0 Compression tester
0 Distributor advance tester
0 Approved emissions analyzer and related calibration
equipment
0 Signs
0 Manuals
The requirement for an ammeter and a pressure gauge are unique to
California.
All programs specify that a sign must be posted. Some
programs, such as in Nevada, specify only a minimum size. Cali-
fornia and the existing Missouri safety program require a spe-
cific sign.
Only Nevada requires that the testing location keep service
manuals. The requirement reads:
Each authorized station shall have adequate information
available for the inspection to determine: what state
or federal emission control devices are required for
specific motor vehicles; and what the motor vehicle
manufacturer's emission control performance specifica-
tions are for the specific motor vehicle (Nevada Envi-
ronmental Commission 1979).
-------
The rationale for this requirement is that maintenance of vehi-
cles should consist of returning the vehicle to manufacturers
specifications. The station must have these specifications to
perform the required adjustments and repairs.
2.2.2 Evaluation
The only equipment necessary to perform an idle vehicle
emissions test is an approved emissions analyzer. Other equip-
ment is required only for adjustments or repairs. In the existing
Missouri safety inspection program, the only equipment required
is for the actual safety inspection and not for repairs to cor-
rect safety defects. Therefore, if the philosophy of the safety
inspection program was transferred over to the emission testing
program, the only equipment required would be the emissions
analyzer, such as in Rhode Island. Involved in the question of
requiring repair equipment is the judgment as to whether a station
operator's ability to perform an objective evaluation of vehicle
safety and emission levels is compromised by the ability and
desire to perform repairs. This judgment is beyond the scope of
this report. If the decision is made to require repair equip-
ment, the list in Section 2.2.1 is appropriate.
It is recommended that Missouri require posting of a specific
sign as opposed to setting minimum standards as in Nevada. In
Nevada, signs came in all sizes and shapes with some 30 feet or
more in length. The result is aesthetically unappealing.
It is further recommended that if repair equipment is required,
that service manuals be required by adopting wording similar to
the Nevada regulations. While not specifying specific publications,
all stations would be required to have all manufacturers' speci-
fications relating to vehicle emission levels.
2.3 PERSONNEL REQUIREMENTS
2.3.1 Overview
The 1978 EPA policy memorandum (see Section 1.2.1) requires
that the inspection facility "employ an inspector with demonstrated
proficiency in I/M rules and regulations, test procedures, analyzer
use, quality assurance procedures and recordkeeping". Existing
programs have responded to this requirement in different ways but
all require inspectors to be licensed.
For example, in Nevada, an applicant must submit a certifi-
cate of competence that indicates ability to perform major motor
vehicle tuneups, submit a certificate of competence issued by the
manufacturer of the exhaust gas analyzer to be used at the testing
location, pass a written test and perform a practical demonstra-
tion. At California Blue Shield stations, an applicant must have
-------
a certificate of competence to perform a major automotive tuneup
to the satisfaction of his employer; or a certificate of compe-
tence from a vehicle, device, or tuneup equipment manufacturer,
an industrial or trade school, or a public high school or junior
college. With such a certificate, the applicant can take the
required written test. In Rhode Island, an applicant must have
passed a training program approved by the state. In New York, a
single specified training course must be completed. In this
phased program, the course will be taught by a contractor, later
by the state, and later by certified private organizations. The
training is based on Colorado State University materials. In
Arizona, inspectors at the centralized locations are trained by
the contractor and fleet operators are instructed by the state in
ci seven-hour course.
2.3.2 Evaluation
All programs require licensing of inspectors and a written
test. Some programs require taking a specific training program
(New York) while most allow different training programs. If the
state test is an adequate test of required knowledge, the spe-
cification of a specific training program appears unnecessary.
Some programs also require a practical test. This further eli-
minates the need for a specific training program.
It is recommended that the state (or designee) offer a
training course consisting of instruction on program procedures,
how to interpret emission test readings, and how to perform
repairs. Additional instruction will be necessary on how to
operate the emissions analyzer and analyzer quality assurance
procedures. Because several brands of analyzers will be used in
the program, instruction on these procedures can best be provided
by the analyzer manufacturer.
Missouri now has a group of licensed inspector/mechanics but
that licensing procedure did not include any requirement for
instruction or testing relative to emission testing and related
adjustments and repairs. To fulfill EPA requirements and to meet
program objectives, it is recommended that all inspector/mechanics
be required to pass another written test for competancy relative
to emissions testing and repair, and pass a practical test on
emission analyzer use and quality assurance procedures. Require-
ments to take the test should be a certificate that the state
training course has been completed, and a certificate from the
analyzer manufacturer that the applicant is competent to operate
the emissions analyzer.
-------
REFERENCES
PEDCo Environmental, Inc. 1981. Support Document for Develop-
ment of the Missouri Inspection/Maintenance Program—Emission
Analyzer Specifications. Kansas City, Missouri.
Nevada Environmental Commission and Department of Motor Vehicles,
1979. State of Nevada Air Quality Regulations for Mobile Equip-
ment. Carson City, Nevada.
-------
STANDARDIZED PROCEDURES FOR
EMISSIONS AND TAMPERING INSPECTIONS
-------
CONTENTS
Figures iii
Tables iv
1.0 Introduction 1
1.1 Purpose 1
1.2 Requirements, emissions test procedure 1
1.3 Requirements, tampering inspections 4
References 6
2.0 Emissions Test Procedure 7
2.1 Comparison of alternate short test procedures 7
2.2 Check of idle speed 9
2.3 Dilution check 13
2.4 Tire pressure check 15
2.5 Inspection procedures for vehicles modified 16
to use an alternative fuel
2.6 Special problem areas 17
References 19
3.0 Tampering Inspections 20
3.1 Comparison of procedures to inspect for 20
tampering
3.2 Recommended procedures for tampering 23
inspections
3.3 Time and costs estimates for tampering 26
inspections
References 27
Appendix A Sources for Manufacturer's Emissions 29
Control Manuals
111
-------
FIGURES
Number Page
2-1 Excess Emissions Identified versus Failure Rate
for Short Tests 8
2-2 Data from Eight 1970-1976 Model Year Vehicles 12
IV
-------
TABLES
Number Page
2-1 Vehicle Emission Inspection Modes for
"Existing I/M Programs 10
3-1 . Tampering Inspections in States with
Existing I/M Programs 22
v
-------
SECTION 1.0
INTRODUCTION
1.1 PURPOSE
The purpose of this report is to provide technical informa-
tion to the State of Missouri on standardized procedures for
emissions testing and tampering inspections for a mandatory
decentralized inspection/maintenance (I/M) program. This report
will serve as the technical basis for the state in adopting
regulations relative to the test procedure.
1.2 REQUIREMENTS, EMISSIONS TEST PROCEDURE
1.2.1 U.S. Environmental Protection Agency (EPA) Requirements
Section 207(b) related rules and regulations (U.S. Environ-
mental Protection Agency 1980) contains requirements for the
testing procedure in the description of an approved short test.
To satisfy Section 207(b) requirements, any of three tests can be
used. These tests are as follows:
1. Idle test
0 Engine at normal operating temperature with all acces-
sories off
0 Analyzer warmed up in stabilized operating conditions
0 (Optional) engine may be preconditioned by operating it
at 2500 ±300 RPM for up to 30 seconds
0 With engine idling and transmission in neutral, insert
probe. Record exhaust concentrations after stabilized
readings are obtained or at the end of 30 seconds,
whichever occurs first
0 Repeat for multiple tailpipes, numerically average
results
-------
Two-speed idle test
0 Engine at normal operating temperature with all acces-
sories off
0 Analyzer warmed up in stabilized operating conditions
0 Attach tachometer pick up
0 With engine idling and transmission in neutral, insert
probe. Record exhaust concentrations after stabilized
readings are obtained or at the end of 30 seconds,
whichever occurs first
0 Repeat for multiple tailpipes, numerically average
results
0 Increase engine speed to 2500 ±300 RPM, repeat test
0 Allow engine speed to be reduced to free idle, repeat
test
0 Final results will be the lowest HC and CO readings
from either idle test, and/or from the high speed idle
test
0 The test can be terminated after the first idle mea-
surement if measured exhaust concentrations are below
the standard
Loaded test
0 Engine at normal operating temperature with all acces-
sories off
0 Analyzer warmed up in stabilized operating conditions
0 Place vehicle on dynamometer
0 Insert probe
0 (Optional) High speed mode, maximum 50 mph and 30
second duration for preconditioning
0 Drive for automatic or third gear for manual transmis-
sions shall be used. Operate vehicle at 30 ±1 mph roll
speed while measuring exhaust HC and CO. Record con-
centrations after stabilized readings are obtained or
at the end of 30 seconds, whichever occurs first.
-------
0 Repeat for multiple tailpipes, numerically average
results.
0 With engine idling and transmission in neutral, record
exhaust concentrations after stabilized readings occur
or at the end of 30 seconds, whichever comes first
0 Repeat for multiple tailpipes, numerically average
results
0 Results of either the loaded mode or the idle mode can
be used according to the choice of the operating juris-
diction
Section 207(b) requirements do not require any tampering or
equipment checks.
In addition to the Section 207(b) requirements, recommenda-
tions for the test procedure are contained in an EPA draft report
(U.S. Environmental Protection Agency 1981a). These recommenda-
tions are:
0 Pre-1981 vehicle, idle test with 2500 RPM preconditioning
0 1981 and later vehicle, pass/fail determination at both
idle and 2500 RPM
0 Randomly sample vehicles for idle speed to determine if
motorists are increasing idle speed to lower test
results
0 Consider a dilution check
0 Check tire pressure
The EPA has no specific requirements for the testing pro-
cedure in the 1978 EPA policy memorandum (Hawkins 1978). The
1982 SIP review checklist (Gray 1981) requires only that the test
procedure be described and that the test procedure include "an
exhaust or other test which demonstrates that proper maintenance
has been performed".
1.2.2 State of Missouri Requirements
Since there is no emissions test currently required in
Missouri, there are no procedures for emission testing.
-------
1.3 REQUIREMENTS, TAMPERING INSPECTIONS
1.3.1 U.S. EPA Requirements
Section 203(a)(3), in part, "prohibits manufacturers, dealers,
fleet operators, and commercial service facilities from knowingly
removing or rendering inoperative any device or element of design
installed on or in a motor vehicle or motor vehicle engine in
compliance with regulations following its sale and delivery to
the ultimate purchaser" (U.S. EPA 1981b). Prior to vehicle sale,
the act need not be done knowingly. Of note is that the law does
not apply to tampering done by an individual to his personal car.
A draft Field Office Tampering Manual (U.S. EPA 1981b) has
been prepared as a reference document for EPA enforcement per-
sonnel who conduct automotive testing inspections. Procedures
are included for inspecting eleven systems or devices. These
are:
1. Catalytic converter
2. Exhaust gas recirculation (EGR) system
3. Air injector reactor (AIR) system
4. Heated air intake system
5. Idle stop solenoid
6. Idle limiter caps
7. Vacuum spark retard system
8. Positive crankcase ventilation (PCV) system
9. Evaporative emission control system
10. Fuel tank cap
11. Filler tank restrictor
1.3.2 State of Missouri Requirements
The State of Missouri has requirements for an inspection of
air pollution control devices (Missouri State Highway Patrol
1980). Existing state requirements apply to 1968 and later model
years. Diesel fuel vehicles or vehicles operating exclusively on
propane fuel or compressed gas are exempted. Requirements are:
0 Check positive crank case ventilation (PCV) system for
integrity and vacuum suction
-------
Check air injection system for integrity and belt
slippage
Check for modifications to the air injector system.
Reject vehicle if there are loose wires or if any part
of the air pollution control system is missing.
Check catalytic converter. For vehicles manufactured
between 1975 and 1980, if equipped, reject if the
catalytic converter is bypassed, modified, or has
leakage. For 1981 or later model vehicles, reject if
vehicle does not have a converter and apply criteria
for 1975 to 1980 models.
-------
REFERENCES
Gray, C. L. 1981. Inspection/Maintenance Program 1982 SIP
Processing. Ann Arbor, Michigan.
Hawkins, D. C. 1978. Memorandum to Environmental Protection
Agency Regional Administrators. Washington, D.C.
Missouri State Highway Patrol. 1980. Missouri Motor Vehicle
Inspection Regulations. Jefferson City, Missouri.
U.S. Environmental Protection Agency. 1980. Motor Vehicles;
Emission Control System Performance Warranty Short Tests. 40 CFR
Part 85.
U.S. Environmental Protection Agency. 1981a. EPA Recommendations
Concerning Emission Inspection Procedures for Motor Vehicle
Inspection and Maintenance Programs—Draft. Ann Arbor, Michigan.
U.S. Environmental Protection Agency. 1981b. Field Office
Tampering Inspection Manual—Draft. Washington, D.C.
-------
SECTION 2.0
EMISSIONS TEST PROCEDURE
2.1 COMPARISON OF ALTERNATE SHORT TEST PROCEDURES
2.1.1 Overview
The idle test is the simplest of all 207(b) approved I/M
emission short tests. A preconditioning mode (2500 RPM) is
optional. The two-speed idle test includes emissions testing
during three distinct sampling modes; idle, 2500 RPM, and idle.
Emissions from any mode can be evaluated against the standard.
The loaded test requires emission measurement at 30 mph cruise
speed on a dynamometer and at idle. Emissions from either mode
can be evaluated against the standard. EPA has evaluated the
effectiveness of each type of test relative to its ability to
identify excess emissions.
Preconditioning the engine at ±2500 RPM before the idle test
is an option with the idle test procedure. The purpose of the
preconditioning is to reduce test variability by cleaning out an
engine that has been idling for an extensive period. While
engineering judgment indicates preconditioning may be desirable,
data to prove its effectiveness are not available (U.S. EPA
1981).
A comparison of the two-speed idle test with compliance
based only on the idle mode, and a loaded test is shown in Figure
2-1. The two tests are compared to the Federal Test Procedure
(FTP) which is the benchmark. The data are for pre-1981 vehicles.
For practical purposes, the short tests are about equal in their
ability to identify high emitting pre-1981 vehicles (U.S. EPA
1981).
For 1981 and later vehicles, EPA compared the effectiveness
of the two-speed idle test, with compliance required at idle and
2500 RPM, with the loaded test. Again, the two techniques were
comparable. However, the idle test was only about 71 percent as
effective as the other two test procedures.
-------
EXCESS
CO
EMISSIONS
IDENTIFIED,
100
90
80
70
60
50
40
30
20
10
0
I
I
I
I
FTP
TWO-SPEED IDLE
LOADED TEST
I I I
10 15 20 25 30 35
FAILURE RATE, %
40 45
EXCESS
HC
EMISSIONS
IDENTIFIED,
Figure 2-1.
100
90
80
70
60
50
40
30
20
10
0
I
I
I
FTP
TWO-SPEED IDLE
LOADED TEST
till
0 5 10
15 20 25 30
FAILURE RATE, %
35 40 45
Excess emissions identified versus failure rate
for short tests.
8
-------
2.1.2 Evaluation
A loaded test can be eliminated from consideration as im-
practical in a decentralized program. Significant capital ex-
penditures would be required at over 1000 stations. In addition,
the area with the dynamometer must be dedicated to emissions
testing, thereby eliminating much of the vehicle repair area at
many facilities.
The decision to be made is between an idle test and a two-
speed idle test. Related decisions are whether to allow precon-
ditioning and whether to have a different test for pre-1981 and
1981 and later vehicles.
Personnel from all existing I/M programs (Arizona, California,
New Jersey, New York, Nevada, Ohio, Oregon, and Rhode Island)
were contacted to determine why they picked a specific type of
test (idle, two-speed idle, loaded, or other). These data are
shown in Table 2-1. Specifically, any other test data, other
than the EPA data referenced in Subsection 2.1.1, was sought. No
other test data was available. Existing programs picked their
test method on the basis of cost to implement, time to perform
the test, what the emission analyzer manufacturer recommended,
and political expediency. Based on this survey of existing
programs, it is concluded that the EPA data relative to the
effectiveness of alternate testing procedures is the best avail-
able .
Based on the EPA data, it is recommended that the idle test
with preconditioning be used for pre-1981 vehicles and that the
two-speed idle test with compliance required at idle and 2500 RPM
be used for 1981 and later vehicles.
If it is not politically expedient to adopt different test
procedures for different year vehicles, it is recommended that
the two-speed idle test with compliance required at both levels
be used. This alternate recommendation is based on the EPA
finding that the idle test is only about 71 percent as effective
as the two-speed idle test for 1981 and later vehicles. Adoption
of the idle test for the newer vehicles would significantly
impact effective program stringency at an increasing rate through-
out the program.
2.2 CHECK OF IDLE SPEED
2.2.1 Overview
It is possible for vehicles that would normally fail an idle
test to pass the test by simply raising the idle speed of the
engine. An unusually high idle speed lowers the idle emission
-------
TABLE 2-1. VEHICLE EMISSION INSPECTION MODES FOR EXISTING I/M PROGRAMS
Inspection mode
Idle HC and CO
2500 RPM HC and
CO
Loaded HC and CO
Exhaust dilution
co2
Idle speed
State implementation
New York
Pass/fail
Pass/fail
4%
Pass/fail
New Jersey
Pass/fail
Ohio
Pass/fail3
Oregon
Pass/fail
Condition
veh/data
collection
Pass/fail
3%
Pass/fail
Arizona
Pass/fail
Condition
veh/data
collection
Pass/fail
4.5%
Nevada
Pass/fail
Condition
veh/data
collection
2250 RPM
Check and
adjust
Rhode Island
Pass/fail
California
Pass/fail
Condition
veh/data
collection
Data
collection
Pass/fail
4.5%
Pass/fail
.Transmission in drive - automatic transmission.
Transmission in neutral - manual transmission.
-------
concentration. Figure 2-2 shows this effect on eight cars tested
by EPA. The higher idle speed does not significantly reduce
overall FTP emissions of the vehicle measured in grams/mile.
Therefore, a vehicle with an increased idle speed over the manu-
facturers specifications has a better chance to pass the design
emission limits while being dirtier than some other vehicles
which fail the idle test. This has the effect of reducing pro-
gram effectiveness.
Vehicles are set at a high idle speed for a variety of
reasons. These reasons include simple tune-up errors, a pref-
erence for a faster and smoother idle and a quick-fix for a
stalling vehicle with another malfunction. Conceivably, the idle
could also be increased to pass the emissions test.
EPA suggests that if an idle speed inspection is instituted,
1200 RPM should be the lowest cutpoint used for the inspection.
Below this value, too many clean cars would fail the idle speed
check. Four cylinder engines tend to have high idle speeds more
frequently than six and eight cylinder engines. Therefore, EPA
recommends separate cutpoints to obtain equal idle speed failure
rates in each category. The recommended cutpoints are 1500 RPM
for four-cylinder engines and 1200 RPM for six- and eight-cylinder
engines. The idle speed check should only be used as a criterion
for a valid emission test. If a car has a high idle speed and
passing idle HC and CO levels the test should be considered
invalid. The idle speed should be corrected and the vehicle
retested to determine the emission levels.
For the opposite case, high idle speed and failing idle HC
and CO levels, the test should be considered valid, since lowering
the idle speed would not correct the cause of the high emission
levels.
As an option to checking the idle speed of all vehicles, EPA
suggests initial collection of idle speed data for all vehicles
during early phases of the program to determine if high idle
speeds are a problem. If they are not, the idle speed check can
be eliminated.
2.2.2 Evaluation
In Subsection 2.1, an idle test with preconditioning for
pre-1981 vehicles and a two-speed idle test for later vehicles
was recommended. Both test procedures require the use of a
tachometer (for preconditioning or the 2500 RPM speed). There-
fore, it would be very easy to implement the idle speed check.
It is recommended that a regulation for idle speed checks be
adopted.
11
-------
EFFECT OF IDLE SPEED ON IDLE EMISSIONS
o
t—*
oo
£ 4.0
o 3.0
o
0.0
IDLE CO, %
IDLE HC, ppm
2.0 -
1.0 _
400
1200 1800 2000
IDLE SPEED, NUETRAL GEAR
2400 2800
Figure 2-2. Data from eight 1970-1976 model year vehicles
12
-------
The regulation should include the outpoints recommended by
EPA: 1500 RPM for four-cylinder engines and 1200 RPM for six-and
eight-cylinder engines. Any test of a vehicle with a high idle
speed and passing emission levels should be considered invalid.
The vehicle should be repaired and retested before a determina-
tion of its emission levels is made.
Since a tachometer is necessary for determining engine
speed, several questions may arise pertaining to the accuracy and
reliability of these devices. According to one instrument manu-
facturer, few problems are experienced with tachometers, and
these generally occur after the fifth or sixth year of use (D.
Miller 1981).
Service personnel for this manufacturer carry a calibration
standard and check the calibration of the units during each
service call. These checks have shown that the units are usually
within 1 to 2 percent of the standard. From the conversation it
appears that only minimal problems can be expected with the use
of the tachometer. However, the state may want to be assured
that the tachometers used are in calibration. Therefore, the
state inspectors could be provided with a calibration standard
and the tachometer used for emissions testing audited during each
monthly audit. These 110 volt, 60 cycle standards could be
purchased for less than $100 apiece.
There are two basic types of tachometers that would be used
for evaluating engine speed. The simplest type attaches to a
spark plug wire with a spring-loaded clip. The other requires
that the tachometer be wired in series between a spark plug and
the distributor. Attaching the first type would require little
more time than opening the hood of the vehicle. The second type
would probably require no more than a minute of additional time
for a trained mechanic.
2.3 DILUTION CHECK
2.3.1 Overview
Dilution of engine exhaust before it reaches the working
part of the emission analyzer will cause emission readings to be
lower than the true values. Dilution can occur either from a
leaking vehicle exhaust system, inadequate sampling of the exhaust
(probe not fully inserted into tailpipe) or a leaking analyzer
sampling system. Of the three causes, the first cause is of
primary concern here. Design specifications for long probe
lengths minimize the second cause. Analyzer leak checks minimize
the third cause. In any event, outside air mixed with the exhaust
cause lower measured values allowing some vehicles to falsely
pass the emissions test. Dilution can be checked by using an
13
-------
analyzer with C02 measurement capability or less accurately, by
visual inspection. In existing programs with C02 measurement
capability, 1 to 3 percent of all vehicles are failed for dilu-
tion.
EPA makes no recommendation concerning dilution checks for
decentralized programs.
2.3.2 Evaluation
As noted above, there are two ways to check for dilution
caused by a leaking vehicle exhaust system, i.e., C02 measurement
and visual inspection.
CO- measurement capability on a new analyzer costs appro-
ximately $600. Retrofit of the capability to a CO/EC analyzer is
not usually practical. A separate C02 measurement instrument
costs approximately $1500 (U.S. EPA 1981). Since some decentral-
ized testing stations already have an adequate CO/HC analyzer
without the C02 capability, the cost of adding C02 measurement
capability to the program would range from $600 to $1500 per
station for 1000 stations.
The less effective visual check would be free of capital
cost but would require a minute or two of labor time. There is
no data to indicate how less effective the visual test is. Since
only 1 to 3 percent of all vehicles are failed due to dilution,
with a C02 analyzer, (U.S. EPA 1981) if the visual check iden-
tified half of the dilution occurrences, only 0.5 to 1.5 percent
would pass when they should fail due to dilution.
In quantitatively evaluating the benefit of C02 measurement
capability, the following assumptions were made:
0 1000 test locations, of which 800 are buying new ana-
lyzers and 200 must purchase a separate C02 measurement
instrument
0 A 1980 specification HC/CO analyzer cost of $5500 and a
HC/CO/CO2 analyzer cost of $6100 amortized over five
years
With these assumptions, C02 measurement capital cost for the
entire program would be $780,000, or $780 per station. Added to
the cost would be the direct expense and labor cost for measure-
ment and quality assurance.
No method could be derived for accurately comparing costs
and benefits from adding C02 measurement capability to the program.
However, assuming that 0.75 percent more vehicles would be failed
with C02 measurement capability than would be failed with a
14
-------
visual dilution check, about 1800 more vehicles would fail per
year. Assuming the failures were equally distributed among the
1000 stations, and amortizing the average $780 capital cost over
5 years, the cost for an additional 1.8 failures per station per
year would be about $156 per station per failure per year. This
would seem to be excessive. It is recommended that a visual
check be made of exhaust system integrity to identify dilution.
2.4 TIRE PRESSURE CHECK
2.4.1 Overview
The lower the pressure in a car tire, the more the tire
bends as it rolls. The bending causes increased fuel consump-
tion. Underinflation by 5 pounds causes a 2 percent decrease in
fuel economy (U.S. EPA 1981).
While the primary purpose of an I/M program is to reduce air
emissions, a secondary purpose is increased fuel economy. A tire
pressure check would contribute to the second purpose. EPA
recommends that tire pressure checks be required or encouraged
and that an on-premise air hose be required for tire inflation
(U.S. EPA 1981). It is recommended that the tire pressure be
determined and inflated at ambient conditions (when the tire
temperature is not increased due to travel). The pressure should
be set at a level about three pounds per square inch (PSI) above
the recommended comfort driving pressure (generally 26 to 28
PSI). It is recommended that the tires not be inflated above the
maximum inflation pressure shown on the sidewall of the tire
(generally 32 PSI).
Missouri's existing safety inspection program requires a
visual tire check for wear but not for inflation level.
2.4.2 Evaluation
EPA reports (U.S. EPA 1981) that 63 percent of all tires are
underinflated by an average of 4.5 pounds per square inch below
the recommended comfort pressure. Therefore, on a fleet basis,
fuel economy would be increased by 1.1 percent if the tires were
inflated to the recommended comfort pressure. The fuel economy
benefit can be increased beyond the 1.1 percent value. If all
tires at less than 28 PSI are inflated to exactly 28 PSI, the
fuel savings increases to 1.5 percent. If 32 PSI is used, the
savings is almost 3 percent.
The tire pressure check could be easily added to the inspec-
tion procedure. Checking the pressure in four tires would require
a. maximum of three minutes labor. Assuming a burdened mechanic
labor cost of $20.00 per hour, that all testing locations have a
15
-------
compressor, and that 1.2 million vehicles will be tested each
year, the cost of the tire pressure checks would be $1.00 per car
or $1.2 million annually.
EPA estimates that tire underinflation of 5 pounds cost a
motorist about $13 per year in added fuel cost. Using the con-
servative parameters noted above (63 percent underinflated by 4.5
pounds), average savings per vehicle would be $7.37 or $8.8
million for a 1.2 million vehicle population.
Even allowing for the conservative cost saving assumptions,
the cost saving is obvious. However, the tire pressure check
will increase the inspection cost to the station operator (proba-
bly without an increase in inspection fee) and is outside of the
legal intent of an I/M program. Therefore, the decision to check
tire pressure is arguable.
The purpose of I/M is to reduce emissions by keeping in-use
vehicles properly tuned. A properly tuned vehicle has both lower
emissions and improved fuel economy. Maintaining correctly
inflated tires also improves fuel economy and overall vehicle
emissions should also be reduced, since smaller quantities of
fuel are burned. This is an area where inspections could be
construed to include fuel economy considerations along with
emissions and safety.
It was conservatively assumed that $8.8 million could be
saved by the motorists in the I/M program if the tire pressure
checks were implemented. The cost to the station operator would
be about $1 per vehicle or $1.2 million annually for the program.
These figures imply an overall savings of about $7.6 million
which is probably near the total cost of the I/M program. Given
these estimated cost figures, the tire pressure check should be
seriously considered for inclusion in the program.
2.5 INSPECTION PROCEDURES FOR VEHICLES MODIFIED TO USE AN
ALTERNATIVE FUEL
2.5.1 Overview
Gasoline-powered vehicles may be converted to use a fuel
other than gasoline. The fuels normally used are liquefied
petroleum gas (LPG) and either compressed (CNG) or liquid natural
gas (LNG). Some vehicles are converted so that gasoline can be
used at times and the alternate fuel at other times (dual fuel
system). Complete fuel conversion systems exclusively use the
alternate fuel.
There are several options available for emission inspection
of these modified vehicles. The existing Missouri safety inspec-
tion regulations specifically exempt vehicles operating exclusively
16
-------
on propane or compressed gas from inspection of emission control
devices. This exemption could be expanded to include those
vehicles with dual fuel systems.
A second option would be to retain the exemption for com-
plete fuel conversions and test the dual-fuel systems for gaso-
line.
The third option would be to test the vehicles at their
regular emission standards regardless of the fuel used. Those
vehicles with dual-fuel systems would be tested for both fuels.
2.5.2 Evaluation
EPA makes no specific recommendation for testing modified
vehicles. Rather they recommend that gasoline-powered vehicles
be included in an I/M program. Only California specifically
exempts dual-fuel and permanently converted vehicles from the
inspection. Other states refer only to gasoline-powered vehicles.
It is generally assumed that excess CO and HC emissions are not a
problem for vehicles using an alternate fuel, since these fuels
burn cleaner than gasoline.
The proposed I/M program for Missouri refers to both gaso-
line and dual-fuel system vehicles since complete conversion
vehicles are presently exempted under the safety inspection
regulations. Consequently, it is recommended that dual fuel
systems be included in the I/M program. However, these vehicles
need only be tested for gasoline. The test procedures for these
vehicles will be identical to those for vehicles that are not
modified.
2.6 SPECIAL PROBLEM AREAS
2.6.1 Overview
The only special problem that has not been considered is the
testing procedure for Ford motor vehicles equipped with an air-
pump diverter valve. While these vehicles are idling, this valve
causes air from the air pump to be vented to the atmosphere
rather than to the exhaust manifold and through the calatylic
converter. The primary reason for diverting the air is to pre-
vent the catalytic converter from overheating.
The implication of this valve for the idle emission test is
that the resulting emission value will be artificially high unless
certain precautions are taken. The first is that the inspector
be aware of the problem. The second precaution is to precondition
the vehicle at 2500 RPM with the probe in the tailpipe for 30
seconds. Then, the engine should be allowed to reach curb idle
and the emission levels should be read within 30 seconds.
17
-------
2.6.2 Evaluation
The inspection procedures recommended in Subsection 2.1.2
should be sufficient to address this problem either preconditioning
the engine at 2500 RPM or the two stage idle test will take care
of the problem. The only specific requirement needed for the
idle test is that all vehicles, not just Fords, be accelerated
for 30 seconds at 2500 RPM with the probe in the tailpipe and the
emission levels read within 30 seconds of the vehicle reaching
curb idle.
18
-------
REFERENCES
U.S. Environmental Protection Agency. 1981. EPA Recommendations
Concerning Emission Inspection Procedures for Motor Vehicle
Inspection and Maintenance Programs—Draft. Ann Arbor, Michigan.
Miller D. 1981. Conversation with Regional Sales Manager, Sun
Electric Corporation, Olathe, Kansas.
19
-------
SECTION 3.0
TAMPERING INSPECTIONS
3.1 COMPARISON OF PROCEDURES TO INSPECT FOR TAMPERING
3.1.1 Overview
Tampering, as referred to in this section, is the removal,
disconnection, or making inoperative, any device that is an
operational part of the air pollution control system of a motor
vehicle as required by state or federal law. Eleven devices or
systems are on most recent model cars except in California where
retrofit devices have increased the number. The 11 devices or
systems are:
1. Catalytic converter
2. Exhaust gas recirculation (EGR) system
3. Air injection reaction (AIR) system
4. Heated air intake system
5. Idle stop solenoid
6. Idle limiter caps
7. Vacuum spark retard system
8. Positive crankcase ventilation (PCV) system
9. Evaporative emission control system
10. Fuel tank cap
11. Filler tank restrictor
The basic problem with tampering inspections is the diffi-
culty in performing one. Each car model has different emission
control devices and systems. In a few instances, systems have
changed in the same model year on the same model. With such
variation, the tampering inspector must know: (1) what systems
each model was manufactured with; and (2) how to recognize the
20
-------
system. The first problem requires a set of several manuals. No
single manual is available that lists all required equipment on
each model. The second problem requires mechanic training.
State programs have reacted to tampering inspections in
different ways. In California, which is a centralized program
and where there are retrofit devices besides the basic systems,
the inspector enters the make and model of car to a computer
terminal which in turn lists the required devices. This takes
care of the first problem listed in the previous paragraph but
not the second. In Nevada, a tampering inspection is required.
However, the requirement is not being enforced because mechanics
did not have the requisite skills to perform the inspections.
States not having a tampering inspection requirement generally do
not because of the difficulty in performing the inspections.
Table 3-1 lists the vehicle components inspected for tampering by
the three states that require the inspection for their existing
I/M program.
Missouri's existing safety inspection program regulations
require an inspection of the catalytic converter, the PCV system,
and the air injection system.
3.1.2 Evaluation
Missouri's experience in implementing its regulations re-
quiring inspection of the catalytic converter, PCV system, and
air injection system has been similar to other states efforts to
require tampering inspections. While some inspection failures
have been caused by tampering, enforcement has been difficult
because of at least two factors. First, no single manual is
available that lists all required equipment for every model car.
Therefore, garage station personnel do not know what devices were
manufactured on the car. The second problem appears to be one of
attitude. Since the primary thrust of the existing program is
safety, garage station personnel place less emphasis on inspection
for air pollution related equipment. About 1 percent of all
vehicles are failed for air pollution equipment (Missouri State
Highway Patrol 1981).
EPA, at the national level, acknowledges the difficulty in
completing a tampering inspection for all 11 items (U.S. Environ-
mental Protection Agency 1981). Personnel have unofficially
suggested a check that includes at least the following items:
catalytic converter, EGR system, missing belt from air pump,
heated air intake system, and filler neck restrictor.
It is recommended that the Missouri I/M program include
tampering inspections for the following devices or systems:
21
-------
TABLE 3-1. TAMPERING INSPECTIONS IN STATES WITH
EXISTING I/M PROGRAMS
Vehicle component
Catalytic converter
EGR valve
Air injection system
PCV valve
< Thermostatic air cleaner
Fuel fillerneck
Exhaust system modifications
Engine modifications
Components inspected
Oregon
va
V
V
V
V
V
V
V
Nevada
V
V
V
V
V
California
Fb
V
V
V
V
V
V = visual check.
F = functional check.
22
-------
0 Catalytic converter
0 EGR system
0 Air injection reaction system
0 Heated air intake system
0 PCV system
0 Fuel tank cap
0 Filler neck restrictor
Selection of the items to be examined was based on ease of
checking the device or system, practicality of checking the
device or system, and time required to check the device or system.
The check for the catalytic converter will require the use of a
lift or jack to check underneath the vehicle. However, it is
already necessary to hoist the vehicle as part of the safety
inspection to check ball joints, suspension linkage and wheel
play. The EGR, air injection, heated air intake, and PCV systems
can be visually checked by lifting the hood. No removal of the
air cleaner or other equipment would be required. The fuel tank
cap and filler neck restrictor can be easily visually checked.
Inspection of the idle stop solenoid and the idle limiter
caps was rejected because of the need to remove the air cleaner
and because of the difficulty of determining what equipment
belongs on different models. Inspection of the vacuum spark
retard system was rejected for similar reasons. Inspection of
the evaporative control system was rejected because of model
variations and the need to make a complete underbody check not
possible with a car on a jack.
3.2 RECOMMENDED PROCEDURES FOR TAMPERING INSPECTIONS
3.2.1 Catalytic Converter
Inspection Procedure;
1. Check for the presence of the catalytic converter. A
missing catalytic converter is tampering. If there is
any uncertainty as to whether a vehicle should have a
catalytic converter, the underhood vehicle emission
control information label should be checked. On most
vehicles, the emission control systems are listed on
this label.
2. On General Motors and American Motor models check the
plug in the body of the canister. Both companies now
use a pressed-in plug in most post-1976 models. If the
plug is nonstandard or is damaged, the catalyst material
may have been drained. The plug cannot be removed
readily and would probably be damaged if removed.
Removal of the catalyst material is tampering.
23
-------
3.2.2 EGR System
Inspection Procedure;
Inspect the system to determine if the valve, sensor(s)
and hoses are in place. All EGR disconnections and mis-
routings are considered tampering. Note that this would
include smashed valves, holes in valves, and vacuum hoses
plugged with bolts, fuses, or SB's. All vacuum hoses in the
system should be physically checked at connection points by
squeezing to determine if they have been physically obstructed.
The EGR system is an important control system for reducing
nitrogen oxides (NO ). NO cannot be detected by existing emis-
sion analyzers. The checks listed above will ensure that the
system is intact and that the lines running through the system
are not plugged. The procedure to determine if the system has
been rendered inoperable would be costly. Consequently, the
emission control benefit of inspecting this system may not be
fully realized by only a visual check.
3.2.3 Air Injection Reaction System
Inspection Procedure:
If the system is an air pump system, check for the
presence of the pump and the belt. Inspect all hoses for
obstructions by squeezing near connection parts.
If the system is not an air pump system, no inspection
for this system is required.
3.2.4 Heated Air Intake System
Inspection Procedure;
1. Check the vacuum line from the carburetor to the thermal
vacuum valve and from the thermal vacuum valve to the
vacuum motor on the air cleaner horn. If either line
is missing, plugged, or disconnected it is considered
tampering. It normally requires a deliberate act by a
mechanic or owner to disconnect these lines.
2. Check for the presence of the "stove pipe" which is the
black paper/foil or metal connection between the exhaust
manifold shroud and the air cleaner horn. If it is
present but not properly connected, or if it is torn or
deteriorated, it is considered disconnected.
3. Check the air cleaner top. If it is inverted, raised
due to nonstock air filter, or holes are punched in the
24
-------
air cleaner, the heated air intake is considered dis-
connected. This is considered tampering because the
heated air intake is deliberately defeated and because
it may render the PCV system, some evaporative control
systems, and other thermal vacuum valves inoperative.
3.2.5 PCV System
Inspection Procedure;
Check the PCV line for proper routing and connections.
Verify that the PCV valve is present and that there is no
relief device in the hose from the crankcase to the air
filter. If there is such a device or if either the line
between the PCV valve and carburetor or the "fresh air tube"
from the air cleaner to the crankcase are disconnected, it
is considered tampering.
3.2.6 Fuel Tank Cap
Inspection Procedure:
1. Check for the presence of the cap. If it is missing,
it is considered cause for failure of the vehicle.
2. Check the condition of the cap. If there are holes in
the cap, it is considered tampering. If the gasket
does not seal properly, it is considered malfunctioning,
Either condition is cause for failure.
3.2.7 Filler Neck Restrictor
Inspection Procedure;
Check the filler neck inlet for the presence of a
restrictor if the car requires unleaded fuel. Also check
that the decal label for "unleaded fuel only" is still
applied near the gasoline inlet. If there are any doubts as
to whether a vehicle requires unleaded fuel, check the dash-
board for an "unleaded fuel only" label near the fuel gauge.
The state could greatly assist garage station personnel in
performing tampering inspections by compiling a list of devices
or systems required on each vehicle model. This information is
available (Appendix A and other sources) but is contained in
several manuals that most service stations could not be expected
to have. Compilation of these resources by the state, with
periodic updates, would significantly improve the quality of
tampering inspections.
25
-------
3.3 TIME AND COSTS ESTIMATES FOR TAMPERING INSPECTIONS
In the previous subsections seven items were recommended for
inspection for tampering. It is difficult to estimate the amount
of time needed to check each item, especially since these items
would be only a small part of the entire inspection process. If
the tampering inspection were to be done completely independent
of all other parts of the inspection, the time to perform the
tampering inspection of the seven items would probably be on the
order of 7 to 10 minutes. However, if the tampering inspection
was integrated into the overall inspection process, the additional
time required to do the inspection would probably be less than
five minutes.
If it is assumed that it would take a trained mechanic 5 to
10 minutes to inspect the seven devices, and that his salary is
$20 per hour, then the additional cost for the tampering inspec-
tion would be $1.67 to $2.33 per vehicle. The annual cost for
the entire 1.2 million vehicles would be $2.0 to $2.8 million.
26
-------
REFERENCES
Colorado State University. 1980. Students Guide for the Federal
Course on Tampering Detection. Ft. Collins, Colorado.
Missouri State Highway Patrol. 1981. Telephone conversation
with Sargeant L. Walker. Jefferson City, Missouri.
U.S. Environmental Protection Agency. 1981a. Field Office
Tampering Inspection Manual—Draft. Washington, D.C.
U.S. Environmental Protection Agency. 1981b. Telephone conversa-
tion with S. Albrinck. Washington, D.C.
27
-------
APPENDIX A
SOURCES FOR MANUFACTURER'S EMISSIONS
CONTROL MANUALS
28
-------
Availability and cost of manufacturers emissions control manuals
can be obtained by written request to the following addresses.
Check with local dealers for vehicles not listed. They can
provide the addresses for obtaining manuals.
1. American Motors Corporation
Write to: American Motors Corporation
14250 Plymouth Road
Detroit, Michigan 48232
2. Chrysler Corporation
Write to: Chrysler Corporation
Service Department
Post Office Box 40
Detroit, Michigan
Datsun
Write to:
Parker Industries, Inc.
609 Deep Valley Drive
Rolling Hill Estates, California
90274
Ford Motor Company
Write to: Helm Incorporated
Post Office Box 07150
Detroit, Michigan 48207
General Motors Corporation
Write to: CMC Truck & Coach
Division Printing, Inc.
Department CMC
1179 Sylvertis Road
Pontiac, Michigan 48054
Honda
Write to:
Toyota
Write to:
8
Volkswagen
Write to;
American Honda Motor Company Inc.
Automobile Customer Service Department
100 West Alondra Boulevard
Gardena, California 90247 (213)327-8280
Toyota Motor Sales USA, Inc.
2055 West 190th Street
Torrance, California 90504
Customer Relations (312)532-5010
Robert Bently Incorporated
872 Massachusetts Avenue
Cambridge, Massachusetts 02139
29
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 907/9-81-005
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Background Research for Missouri I/M Regulations
5. REPORT DATE
September 1981
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
PN 3525-13
9. PERFORMING ORGANIZATION NAME AND ADDRESS
PEDCo Environmental, Inc.
2420 Pershing Road
Suite 300, Crown Center
Kansas City, Missouri 64108
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3512
Task Order No. 13
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Region VII
324 East Eleventh Street
Kansas Citv. Missouri 64106
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The 1977 Clean Air Act Amendments defined inspection and maintenance as a reasonable
technique for the control of CO and 0., and mandated its implementation in those areas
where the states cannot demonstrate attainment of the standards by December 31, 1982.
This document presents the results of the background research performed for the State
of Missouri to assist in formulating their program and is a compilation of four
separate reports: (1) Emission Analyzer Specifications; (2) Quality Assurance Proce-
dures; (3) Inspection Station Requirements-; and (4) Standardized Procedures for Emis-
sions and Tampering Inspections. The basic procedure was to review the experience
obtained in other I/M programs and to formulate a specific program for the State of
Missouri based on their unique needs. In addition to summarizing the experience of
others., the evaluation of the costs of the various aspects of the proposed program
are included.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
I/M Inspection/Maintenance
Missouri I/M
Emission analyzers
Quality assurance
Tampering inspections
Inspection procedures
Station requirements
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |