EPA-AA-SDSB-80-
Technical Report
February, 1980
Recommended Specifications
for I/M Type Analysis Systems
Used to Pass or Fail Vehicles
by
William B. Clemmens
Merrill W. Korth
Gordon J. Kennedv
NOTICE
Technical Reports do not necessarily represent final EPA decisions or posi-
tions. They are intended to present technical analyses of issues using data
which are currently available. The purpose in the release of such reports
is to facilitate the exchange of technical information and to inform the
public of technical developments which nay form the basis for a final EPA
decision, position, or regulatory action.
Standards Development and Support Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
U.S. Evnironmental Protection Agency
-------
Table of Contents
I. Introduction
II. Discussion of Specifications
A. Methodology 6
B. Comparison to Other Specifications 6
C. Costs 10
III. How to Use the Specifications
A. Overview 13
B. Definitions and Abbreviations 14
C. Analyzer Technology 18
IV. I/M Analysis System Specifications
A. Gases IP
B. Gas Cylinders 21
C. Durability Criteria 22
D. Design Requirements 23
E. Analyzer Performance 30
F. Sample System Performance 33
G. Operating Environment 34
W. Fail-Safe Features - 35
I. System Correlation to Laboratory Analyzers 42
J. Manuals 43
V. Test Procedures 45
A. Traceability of Analytical Gases 46
B. Gas Cylinder Specifications 47
C. Durability Test Procedures
1. Vibration and Shock 48
2. Sample Line Crush 50
3. Sample Handling Temperature Effect 51
4. Filter Check and Hang-up 53
D. Design Requirement Inspection and Test Procedures
1. Useful Life 56
2. Name Plate 56
3. Sample System " 56
4. Sample Pump 56
5. Sample Probe 56
6. Sample Line 56
-------
Table of Contents (continued)
Page
D. Design Requirement Inspection and Test Procedures
(continued)
7. Analyzer Spanning System 57
8. Analyzer Ranges 59
9. System Grounding 59
10. System Vents 59
E. Analyzer Performance Test Procedures
1. Calibration Curve 60
2. Resolution 62
3. Compensation
a) Altitude 63
b) Pressure and Temperature 66
c) Non-Compensated Systems 69
4. Zero and Span Drift 70
5. Span Drift (see E.4.) 73
6. Noise (see E.8.) 73
7. Sample Cell Temperature 74
8. Gaseous Interference and Noise 76
9. Electrical Interference 79
10. Propane to Hexane Conversion Factor 83
F. Sample System Test Procedure
1. Sample Cell Pressure Variation, Low Flow and 86
Response Time
2. (See F.I.) 90
3. (See F.I.) 90
4. (See F.I.) ' 90
5. System Leakage 91
G. Operating Environment Test Procedure 93
H. Fail-Safe Systems
1. Test Procedures for All Systems
a) Warm-up Lock-out 95
b) Low Flow 97
c) Leak Check 97
2. Test Procedures for Decentralized Systems
a) Automatic zero/span 98
b) Automatic Read 98
c) Printer 98
d) Vehicle Diagnosis 98
e) Anti-Tampering 98
I. System Correlation Test Procedures
1. NDIR Correlation 99
2. FID Correlation . 103
-------
I. Introduction
It is doubtful that anyone would disagree with the statement that automo-
biles (and trucks) are part of the air pollution problem in the country. In
the past. Congress has specified, through the Clean Air Act and various
amendments to the Act that new cars must meet certain emission performance
standards prior to introduction of those vehicles into commerce. These
performance standards have always been checked under closely controlled
laboratory conditions with sophisticated equipment. The required accuracy
and performance of this equipment has always been specified within the
context of the laboratory situation.
Recently, there has been more emphasis on checking the performance of in-use
vehicles. This is occurring through the implementation of state inspection
and maintenance programs as well as the forthcoming emission repair warranty
regulations (207(b)j authorized by the Clean Air Act. Further, the new 1984
Heavy-Duty (HD) Truck Federal Regulations specify an idle standard as well
as a driving cycle standard. Both the 207(b) warranty emissions test for
hydrocarbons (HC) and carbon monoxide (CO), and the Heavy-Duty idle emission
test for CO are expected to be conducted on in-use vehicles with data gener-
ated mostly by state-run I/v programs.
The economic implications about the accuracy of the data from these I/M
programs is of concern. The 207(b) emission repair warrantv provisions and
the Heavy-Duty idle test data may be used to determine who pays for emission
repair maintenance on failed vehicles - the individual consumer, or the
vehicle manufacturer. Additionally, the Heavy-Duty idle data may serve as a
basis for recall action since failure of the wn idle test is the same as
failure of any other Federal emission standard. For those failed vehicles
not covered by 207(b) or HD standards, it is possible that the amount of
repair maintenance performed will have some relationship to the degree of
failure - (i.e. only those repairs that are needed to bring the vehicle into
compliance will be performed). If this minimal maintenance scenario comes
to pass, then the accuracy and variability of the measurements system will
play an important role in the dollar amount of the maintenance sold.
Because the I/M emission check (as well as 207(b) and HD idle) is less
sophisticated than the Federal driving cycle compliance checks, it might be
expected that the requirements necessary to achieve acceptable accuracy and
^variability with I/M analysis systems would be less sophisticated than those
used in the laboratory. To some extend this statement is true, but it is
not totally true. Consider for instance, the laboratory analyzers are
operated by trained and skilled technicians. These personnel can spot
problems in analysis sometimes even before they happen. Complete engineer-
ing departments are constantly checking the laboratory emission values
against past values and design goals. Under these conditions, errors in
measurement can be detected and eliminated. In a sense, the human- mind is
adding considerable sophistication to the equipment. While not saying that
I/M analyzer operators do not have the mental capacity to comprehend the
variety of emission analysis problems, it can be said that the typical I/M
-------
analyzer operator generally has not had the training or experience that the
laboratory technicians and engineers have had. This training problem is
compounded even further in decentralized I/M programs. Mechanics in a
repair shop just cannot be required to understand the gamut of measurement
problems. Even though mechanics must by necessity become familiar with the
I/M equipment, there is an economic incentive to direct their technical
skill toward repairing vehicles.
Another significant difference between the laboratory analyzer, and the I/M
type analyzer is the environmental operating conditions. Laboratory analy-
zers are never operated in hostile environments. They are generally oper-
ated in heated and air conditioned buildings with humidity control. I/M
programs, on the other hand, expect analyzers to operate from California to
Maine, from Texas to Minnesota, and from summer to winter in a variety of
enclosures ranging from rain and wind shelters to permanent structures.
This variety of hostile environmental conditions places a much greater
burden on an I/M analyzer than a laboratory analyzer ever encounters.
The lack of measurement savvy by the I/M analyzer operator, and the signi-
ficant variation in environmental operating conditions are just two among
many reasons that would suggest an I/M analyzer should be more sophisticated
than a laboratory instrument. However, the goals of the I/M program, and
the cost of the I/M analyzer to the user must be kept in mind.
With these ideals in mind, the important parameters in the emission measure-
ment process were evaluated. A balance among capability, cost, complexity,
and operator skill has hopefully been achieved in the recommended I/1* anal-
yzer specifications presented in this document.
-------
II. Discussion of Specifications
A. Methodology
To those familiar with I/M analyzer specifications publicized by other
organizations (State of California Bureau of Auto Repair, Equipment and Tool
Institute, etc.), the specifications included in this document will appear
somewhat different even though they parallel the 1980 California specifica-
tions in some areas. To the user not fully acclimated to the emission
measurement process, the publishment of yet another specification that is
different from previous publications could cause confusion about which
document should be used. To avoid confusion about the source of the speci-
fications in this document, a brief description of the procedure used to
generate the specifications is presented, followed by a description of the
significant differences between this document and other specifications.
Prior to assigning any values to any measurement parameters (accuracy,
drift, etc), a review of all of the parameters affecting the I/M measurement
process was made. A model of error propagation was then formed to show the
interaction of the various parameters. Approximate ranges of errors were
assigned to each parameter based on test data, experience, and/or engineer-
ing judgment. Previously published specifications also provided many ini-
tial values. Using classical statistical approaches, this model could then
evaluate the impact of each specification on the accuracy of the final
emission value.
The important aspect of the model is that it focuses on the real issue
involved in I/M measurements. That issue is: What is the acceptable accur-
acy of the I/M measurement process? After the answer to this question is
known, the parameter specifications can be adjusted to insure the desired
result.
There is a wide variety of opinion about what is accurate enough. A prelim-
inary evaluation of I/M analyzers with the model indicate the error in
current I/M systems could be on the order of + 25 percent (50 percent range)
or worse. EPA laboratory tests have already shown a 25-30 percent variation
between analyzers, even though the analyzers were set up and operated by
highly skilled technicians under favorable environmental conditions.
The staff at the EPA laboratory in Ann Arbor feels that the real variability
in the field among I/M analysis systems is unknown at the current time. The
staff also feels that the preliminary data of 25 to 50 percent variability
is too large to meet I/M, 207(b), and HD goals while maintaining consumer
acceptance of the impartiality and fairness of the I/M program.
B. Comparison to Other Specifications
As stated earlier, this specification roughly parallels the technical por-
tions of the 1980 State of California specifications. No attempt is made to
recommend accrediting procedures for analyzers (other than a brief discus-
sion in Chapter III Section A). However, there are differences in content
as well as differences in techniques for specifying the same or similar
parameters between this document and the 1980 California specifications.
-------
These differences are the result of applying the methodology in addition to
the conceptualization of the actual step by step procedures that could be
used to verify compliance to the recommended specifications. Some specifi-
cations in other documents, although historically specified, almost defy any
type of verification measurements. Foremost, it was considered that any
specification that was recommended should be capable of convenient measure-
ments (although possibly repetitious and dull) by any emissions laboratory
with typical test equipment (i.e. digital voltmeters, chart recorders,
analytical gases, etc.).
The concept of parameter verification lead to the inclusion of step by step
procedures that can be used for verification by anyone. No other specifi-
cation document provides such procedures. While some may claim that such
procedures are not necessary or that they reduce flexibility, experience in
emission analysis has shown that there are many ways to interpret even what
appears to be straight forward specifications with strikingly divergent
results. The recommended verification procedures, therefore, provide a
framework for interpreting the specifications in a consistent manner.
Another area of conceptual differences in the verification process is the
use of statistical techniques on the verification data. The use of the
statistical techniques are necessary to insure that the verification data
actually represents the true performance of the analyzer under test. The
use and extent of statistics is not overly done, and the tendency is to use
them in the more important areas (e.g. accuracy) to reduce the measurement
burden.
A final area of conceptual difference between this specification and others
is the inclusion of specific quality control features into the basic analy-
sis system specifications. Generally, these .features are classified under
the category of "fail-safe" sub-systems. In addition, many of the other
analysis system requirements contained in this document are specified in a
manner that reinforces the fail-safe concept. The degree of sophistication
of the built-in quality control features is determined by the end use of the
system. Analyzers intended for use in decentralized systems generally
require more sophisticated features (than centralized systems), due to the
more hostile environment and less sophisticated operators.
The highly desireable aspect of these built-in features is that they tend to
improve the overall credibility of the I/M program. The improved credibil-
ity is obtained because the analysis system requires the operator to comply
with good measurement practices as well as preventing the operator from
committing improper practices. Both of these actions improve the overall
accuracy of the I/M results.
A hidden advantage that makes these fail-safe features even more desireable
is that the forced improvement of operator practices relieves some of the
burden on state audit teams. Because of these features, it may be possible
to perform less frequent audits, and still obtain better quality I/M results
than with more frequent audits of systems without the built-in checks.
-------
Traversing from conceptual differences to more specific differences, the
following items represent the more significant areas of difference.
1. Analyzer Calibration Curve; The analyzer calibration curve re-
quirements are more statistically sound than other publish specifica-
tions. Since the calibration curve sets the basic accuracy of the
analyzer, it is important to measure the true performance when attemp-
ting to verify compliance to the specifications. Additionally these
recommendations require better accuracy than other specifications at
the projected 207(b) emission warranty levels of 220 ppmh WC and 1.2%
CO, and the Heavy-Duty idle standard of 0.47% CO.
2. Analyzer Fail-Safe Systems; These recommended specifications
include requirements for certain lock-out features in order to prevent
improper operation of the analyzer. These fail-safe systems are segre-
gated into systems applicable to all analyzers, and those systems
addressing the unique problems associated with decentralized programs.
The decentralized fail-safe systems will most likely require an on-
board microprocessor.
3. Analytical Gases; All analytical gases must be traceable to Na-
tional Bureau of Standards, Standard Reference Materials (SRMs), or an
EPA Office of Mobile Source's approved standard. A procedure is refer-
enced for determining traceability.
4. Analyzer Measurements; A policy decision was made that all I/M
measurements would be considered on a "dry-basis". This decision is
consistent with the 1984 Federal P.D. Truck idle standard, the pro-
jected 207(b) requirements, and the effective policy of many state I/M
programs (due to the act of using a refrigeration water trap). The
effect of this decision would be to consider all current and future
data as dry-basis data. However, the recommendation to implement a
recommended specification for dry measurement hardware should be de-
layed until January 1986. For analyzer manufacturers who wish to
implement dry measurement systems sooner, they may do so by either of
the two techniques permitted - electronic compensation or physical
water removal. The effect of implementing a dry-basis measurement is
to remove a potential 7-10% variability in measurement.
,5. Analyzer Spanning Concepts; The recommended specifications require
all analyzers to be gas spanned at least once a week and electrical
spanned at least once an hour. More frequent gas span checks are
recommended every 4 hours. Two alternatives to the 4 hour gas span
check are possible which would reduce the use of analytical gas. One
allows the use of sample cell pressure and temperature compensation.
The other allows analyzers with a basic accuracy significantly better
than that required in these specifications to demonstrate that even
considering the degradation of accuracy (due to external effects of
pressure and temperature), the analyzer would still be within the
recommended accuracy specifications.
-------
The weekly gas and hourly electrical spanning concept is consistent
with projected requirements for 207(b) and the H.D. idle standard. The
frequency of gas spanning is less than emissions testing experience and
engineering judgment would dictate, but more frequent than other speci-
fications which generally call for a 30 day gas span check.
6. HC Measurements; Most I/M specifications require reporting of HC
as ppm hexane. Since the span gas is propane, and the analyzer re-
sponds differently to propane than hexane, a propane to hexane factor
is required. Additional verification of the relation of this factor to
the composition of exhaust hydrocarbons is generally required by com-
paring the I/M analyzer performance to the performance of a laboratory
NDIR HC analyzer.
These general requirements are included in this document but with more
definitive procedures for developing the factor and conducting the
comparison. But, there is a more fundamental question involved in
determining HC measurement criteria. The point is that the Non-disper-
sive Infrared (NDIR) HC analyzer is limited in its ability to measure
all of the hydrocarbons in the exhaust. Current NDIR analyzers tend to
measure only straight chain paraffinic_ hydrocarbons, although they do
have some response to aromatics and aldehydes. The lack of response to
these non-paraffinic compounds is the primary reason why the Federal
Test procedure changed from NDIR HC measurement to FID (flame ioniza-
tion detector) HC measurement 8 years ago (1972).
Even though catalyst vehicles tend to produce more paraffinic hydro-
carbons than non-catalyst vehicles, chromatographic analysis of
gasoline-fueled engine exhaust indicates significant quantities of the
non-paraffinic hydrocarbons even in catalyst vehicles. The very same
compounds that current NDIR HC analyzer-s have difficulty in measuring.
Two questions arise from this discussion - "Why measure HC at all if
the only measurement technique is NDIR?", and "Why can't the FID be
used in I/M work?". First, although the NDIR does not respond well to
all hydrocarbons, a reduction in hydrocarbons that it can measure still
tends to reduce the overall HC output from a vehicle. Further, in many
cases reduction of paraffinic HC can also reduce other classes of
hydrocarbons.
The answer to the second question is that an FID is a complex piece of
laboratory equipment that requires a mixture of hydrogen and helium for
fuel, and purified air for an oxidizer. At this point, the staff is
somewhat skeptical about the compatability of an FID in an I/M test
environment.
The response characteristics of current NDIR analyzers could, however,
be improved, and new less complex techniques could be developed. In
order to provide an incentive for this advancement in technology,
beginning in January 1986, it is recommended that all I/M analyzers
correlate to an FID HC analyzer by the procedures provided. To assist
-------
10
in the degree of correlation that is technically feasible, it is recom-
mended during the verification of current analyzer performance that an
additional test be performed that compares__the I/M HC analyzer to^ an
_FID. Review of the NDIR/FID correlation datageneTaTe3™B7™th>e—1accredi-
tation procedures is recommended prior to the implementation of the
1986 specifications.
7. System Leak Checks; Leaks in the sample system are probably the
source of the largest and most frequent errors that occur in practi-
cally all emission measurement systems. This is because a leak is
transformed directly into an error (i.e. a 15% leak is a 15% error).
Most laboratories have rigid procedures for leak checking of analysis
systems, and the process of searching for a leak can be very time
consuming.
I/M analysis systems have special problems that the laboratory systems
do not. First, laboratory systems are generally not moved around; I/M
systems in decentralized systems are. Two, laboratories have special
equipment to identify leaks, the expertise to use the equipment, and
the knowledge to repair leaks inside the systems; the I/M operator
generally does not have this knowledge available. Three, the labora-
tories generally have a strong committment to prevent and repair leaks,
and subsequently provide resources of time and money for this commit-
ment; the independent decentralized system operator may not have this
commitment, and may not be able to apply the resources to it. Four,
the flow rates used in I/M systems are so small that even a 10% leak is
difficult to measure without laboratory equipment, let alone a 2 or 3%
leak.
Most other specifications do not provide for a routine leak check, and
if they do provide for equipment, the equipment is usually a tapered
tube flow meter. A tapered tube flow meter is not practical for field
use. They are a high maintenance item if built into the system, and
tend to stick from a combination of hydrocarbons and water. This is
true even in a laboratory situation with overkill on filter changing
frequency. If the flow meter is used to monitor system response time
as well, the capacity of the flow meter is too large to read leaks
without a 15-20% error in the reading. Primarily for these reasons a
flowing span gas leak check (through the sample line) is recommended on
a weekly basis.
C. Costs
In all likelihood, the additional features and specifications will increase
the retail cost of the I/M analysis system. Developing incremental costs
for each change in specification or additional feature may not be possible
(at least not without extensive analysis) due to the degree of interaction
between the features. Another aspect of cost, however, is the overall cost
to the economy. More precise and fail-safe analyzers may well decrease the
cost of performance checks on the analyzers conducted by state audit teams.
Furthermore, more accurate and less variable systems will tend to improve
-------
11
allocation of the repair costs and the need for maintenance to those vehi-
cles that really need it. To the consumer with the marginally passing
vehicle, that could get by with less maintenance because of the better
discrimination, it would also be a cost savings. Further the more accurate
fail-safe analyzer makes it considerably more difficult for an unscrupulous
operator to defraud a consumer, or for a well meaning operator to make an
error in analysis. The prevention of fraud or accidental errors may also
lead to cost sayings.
When costs are considered not only the incremental cost of the analyzer to
the user should be considered, but the total cost and impact on society must
be considered as well. The more accurate system has the potential for a net
savings to the national economy plus having positive environmental benefit.
Even though the benefits to society do exist, the cost of the better system
to the system user must be reasonable. The EPA staff estimates that approx-
imately 50,000 pass/fail I/M units will be sold to support the I/M programs.
Assuming 10 analyzer manufacturers _!_/ actively market this type of analyzer,
and assuming equal market share, each manufacturer would be able to spread
the development costs across approximately 5000 analyzers.
Development and modification costs could be as high as 5100,000 2f on some
systems. Spread across 5000 analyzers this would be $20 per analyzer.
Using a 3 to 1 mark-up which is standard industry practice _2_/, would cause
the retail price to increase $60 per analyzer due solely to development
costs. Since current analyzers are in the $2500-$3500 price range, it is
apparent that development costs are not an overwhelming burden.
A list of assumed system changes is provided in Table 1 along with estimated
retail cost increase. These costs represe.nt an average estimate, some
systems may require more improvement, while others may be less costly to im-
prove. The basic group of improvements ($690) are designed for well-
trained, competent analyzer operators. These improvements are expected to
reduce current measurement errors of 25 to 50 percent down to the 10- to 15
percent range. The cost for these improvements are estimated at $17 to $69
for each percent of improvement in measurement accuracy and variability.
For decentralized programs, accuracy and repeatability of emission measure-
ments by less sophisticated analyzer operators can be compensated for by
adding a microprocessor to the analysis system. The microprocessor allows
simpler external system controls and automatic operation. The simpler
operation does increase the cost of the system. However, considering the
cost to the consumer, and the great potential to inhibit improper system
operation and fraud, the increase in cost is reasonable and well worth it.
\j Eight(8) Manufacturers participated in the December 20, 1979 Equipment
and Tool Institute development of performance test specifications for
garage type analyzers.
2j Based on informal conversations with industry representatives.
-------
12
Table 1
Item Retail Cost Increase
1. Basic Improvements
a) Development $ 60
b) Gas Spanning System 5150
c) Improved Detector $210
d) Improved Signal Conditioning $ 80
e) Leak Check System $120
f) Sample Cell Heater $ 70
$690
2. Decentralized System Safe Guards
a) Microprocessor Based System 700
$1390
Options
a) Anti-dilution $600
b) Printer $250
c) Data Collection and Storage $300
-------
13
III. How to Use the Specifications
A. Overview
As they are stated, the analyzer specifications listed in Chapter TV have
meaning only when properly defined. The test procedures presented in Chap-
ter V provide this context. These test procedures are intended to provide a
consistent technique to verify analyzer performance. As such, these tests
should only be conducted by a laboratory with sufficient facilities, and
personnel experienced in emission measurement.
The test procedures should be performed in the following order.
1. Follow manufacturer's initial start-up and pre-test procedure as
listed in the manufacturer's manual(s).
2. Durability Tests
1st - Vibration and Shock Test
2nd - Sample Line Crush Test
3rd - Temperature Effect
4th - Filter Check
3. Inspect Design Requirements
4. Analyzer Performance
5. Sample System Performance
6. Operating Environment Test
7. Fail-Safe Features
8. Correlation Tests
After the analyzer passes these tests, the user can be fairly confident of
the in-use emission test results. However, the confidence applies only to
the analyzer tested.
As stated before, this document is not an accreditation procedure. This
document presents specifications and procedures for determining the perfor-
mance of an individual analyzer, not a product line.
In order to develop confidence in the manufacturer's product line as a
whole, additional analyzers must be tested. The number of additional analy-
zers tested determine the confidence level of the product line. For in-
stance, every single laboratory analyzer used at the EPA Laboratory in Ann
Arbor, and at most auto manufacturers' facilities are thoroughly checked out
before they are put into service. In many instances, the analyzer is re-
quired to meet performance checks even before it is paid for. This measure
is probably impractical for the large volume of I/M instruments.
-------
14
One practical solution would be to perform a selective enforcement audit
(SEA) as part of the analyzer manufacturer's QA/QC procedures. EPA uses
this SEA principle to evaluate emission compliance of automotive manufac-
turers' product lines. The SEA sample plan is a sequential sample plan, and
relies on separate criteria (number of failed units) to determine a pass
decision versus a fail decision. Such a sample plan can determine with a
reasonable level of confidence compliance of a product from test data on
less than 1 percent of the product line.
Sample selection, procedures, and test plans can be found in 40CFR 86 Sub-
part K and Appendix X (both in the Federal Register, January 21, 1980). An
inspection plan incorporating an audit quality level (AOL) of 10% would
insure with 95 percent confidence level that no more than 10 percent of the
production would fail to comply with the specifications.
A sampling plan for determining product line compliance (over 400 units of
production) is presented in Table 2. If the market share is 5000 I/M units
per analyzer manufacturer, as suggested by the cost section, the upper limit
of 50 units to be tested represents only 1 percent of production. In auto-
motive SEA testing, the determination of pass/fail is usually determined
with much less testing - normally around 12 to 17 units. Twelve I/M analy-
sis systems tested would represent only 0.24 percent of production which
should not be a burden to the manufacturer.
There are other types of production quality checks that could be used.
However, the advantages of the SEA type check of a small sample size as well
as a sequential nature may outweight the possible improvements of other
checks. In fact, if tigher quality is required, the SEA sampling plan could
be adjusted for yearly production rates (see 40CFR 86, Appendix X, Federal
Register, January 21, 1980).
Once the analysis systems are delivered, periodic performance checks of
certain variables can insure proper operation. However, these periodic
checks are useless if the analyzer owner cannot obtain effective repair
service and parts. Further, performance checks after maintenance are neces-
sary to insure that the analyzer was repaired properly.
The previously mentioned items are beyond the scope of this report, but it
was felt that they should be mentioned in order to make the reader aware of
some of the additional factors involved in obtaining reasonable measurement
results.
B. Definitions and Abbreviations
The following definitions are somewhat general but are provided for refer-
ence. For many definitions, the specific test procedures used for verti-
fication of the specifications provide the final interpretation.
-------
15
10% AOL
Table 2
SEA Sample Plan
Over 400 Units Produced
No. of
Failed Units
For
Pass Decision
(1)
(1)
(1)
(1)
(1)
(1)
0
0
0
0
0
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
4
4
4
4
No. of
Failed Units
For
Fail Decision
(2)
(2)
3
3
3
4
4
4
4
4
5
5
5
5
5
6
6
6
6
6
7
7
7
7
7
8
8
8
8
8
No. of
Units
Tested
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
46
47
48
49
50
No. of
Failed Units
For
Pass Decision
4
5
5
5
5
6
6
6
6
6
7
7
7
7
7
8
8
8
8
8
9
9
9
9
9
10
10
10
10
12
No. of
Failed Units
For
Fail Decision
9
q
9
9
9
10
10
10
10
11
11
11
11
11
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
No. of
Units
Tested
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
(1) Test sample passing not permitted at this stage.
(2) Test sample failure not permitted at this stage.
-------
16
1. Accreditation: Certification that an analysis system has been
tested by the required procedures, and it has been verified that the
analysis system complies with those specifications. Further, accredi-
tation implies that the analyzer manufacturer follows proper procedures
to insure that subsequent production systems are also in compliance.
2. Accuracy; The combination of bias and precision errors that quan-
tify the difference between the analyzer reading and the true value
(see equivalency).
3. Analytical Gases: Gases of known concentration used in the analy-
tical process as a reference. The three basic categories are:
a) Calibration Gas
b) Span Gas
c) Zero Gas
4. Analysis System; A system that incorporates an analyzer(s) and
sampling components for the purpose of exhaust gas analysis.
5. Analyzer: A device that has the capability to identify unknown
concentrations of particular constituents in automobile or truck ex-
haust gases by comparison to analytical gases. Commonly used inter-
changeably with "instrument".
6. Calibration Gases: Analytical gases that are used to determine the
accuracy of an analyzer calibration curve.
7. CO: carbon monoxide
8. C0?; carbon dioxide
9. Detector; The portion of the analyzer that detects the constituent
of interest, and provides the original signal proportional to the con-
centration of the constituent.
10. Drift; The amount of change with time of analyzer reading. Two
components of drift are:
a) Zero drift - change in zero reading
b) Span drift - change in the difference between zero and span
readings
11. Dry-Basis Concentration; The resultant concentration after the
water has been removed from the sample either physically or by elec-
tronic simulation techniques.
12. Equivalency; A statistical comparison of a candidate analysis
system performance versus the reference analysis system performance on
exhaust gas in order to determine the acceptability of the candidate
system.
-------
17
13. FID; Flame ionization detector. Most common laboratory analyzer
used for determination of hydrocarbon (HC) concentrations in exhaust
samples.
14. fs; Full scale of the analyzer.
15. Hang-up; Hang-up refers to the process of hydrocarbon molecules
being absorbed, adsorbed, 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.
16. HC; hydrocarbons
17. Instrument; see analyzer
18. Interference (electronic); Analyzer read-out errors caused by
instrument response to electromagnetic sources and power supply varia-
tions. Common forms of electromagnetic sources are:
a) Radio Frequency Interference (RFI)
b) Very High Frequency Interference (VHF)
19. Interference (gases); Analyzer read-out errors caused by instru-
ment response to non-interest gases typically occuring in vehicle
exhaust.
20. L.S.; Low scale or range of the analyzer.
21. NDIP; Non-dispersive Infrared Analyzers
22. Optical Bench; The portion of the analyzer that consists of the
main sample processing and detecting assembly. Generally included are
the detector, optical filters, sample tubes, infra red source, and
chopper systems.
23. Precision; Statistical quantification of random measurement
errors.
24. ppm; parts per million by volume
25. ppm C; ppm by Carbon atom
26. ppm C3 or ppmp; ppm propane (C H )
J_L______^___t_l_ • j J^
27. ppm C6 or ppmh: ppm n-hexane
28. Propane to Hexane Conversion Factor; A factor that accounts for
the difference in analyzer response (relative response) between propane
and n-hexane. Sometimes referred to as a "C" factor, or a propane
equivalence factor (P.E.F.).
-------
29. Response; Analyzer indication to a gas.
30. Response Time; The reaction time between a change in concentra-
tion at the inlet to the sample system and the time the analyzer indi-
cates a given percentage of that change.
31. Sample System; The portion of the analysis system that is respon-
sible for delivering an unaltered sample to the analyzer.
32. Span Gases; Analytical gases that are used to adjust or return
the analyzer response characteristics to those determined by the cali-
bration gases.
33. Zero Gas: An analytical gas that is used to set the analyzer
response at zero.
C. Analyzer Technology
These specifications were determined based on current I/M practice of using
non-dispersive infrared (NDIR) analyzers for HC and CO measurements.
Nothing in these specifications should be contrued as prohibiting other
analysis techniques. Potential improvements in technology should be consi-
dered on a case by case basis. To that extent, many of the concepts ex-
pressed by these specifications can be applied.
-------
IV. I/M Analysis System Specifications
A. Gases
1. Accuracy
a) Calibration gases: + 2% of True Value
+ 1% of NBS Standard Reference
" Material (SUM)
b) Span gases: + 3% of True Value
+ 2% of NBS Standard Reference Material
c) Zero gases: less than 10 ppm C hydrocarbon
less than 50 ppm CO
less than 1000 ppm CO
2. Composition
a) CO and HC (propane) with N or air diluent
b) CO with N or air diluent
c) Zero gas may be bottled gas or chemically purified room air
such as with an activated charcoal trap on the analysis system.
d) Fexane to propane conversion factor checking gases shall
consist of n-hexane in nitrogen, gravimetrically blended to the
accuracy of SRM's. The gravimetric analysis is only valid for 1
year from the date of analysis unless historical data and corre-
lation checks can verify stability erf the gas concentration. The
gravimetric analysis is void if the gas cylinder temperature drops
below 20°C (68°F) for any reason including shipping and storage.
The required concentrations are:
i) 200 ppmh ( +_ 15 ppmh)
ii) 1800 ppmh ( +_ 150 ppmh)
3. Recommended number of gases
a) for new instrument check out
i) 7 equally spaced concentration values of calibration
gases per range
ii) bottled zero gas
iii) 2 concentrations of hexane/propane conversion factor
checking gas
-------
20
b) for in-use systems
i) one span gas concentration between 70-95% of fs on low
range for each HC and CO analyzer
ii) room air zero gas
c) for periodic check (i.e. state inspection) of in-use systems
i) a minimum of 3 concentrations per range for each analyzer,
or
ii) one concentration for each cutpoint with a concentration
value within 10 percent of the cutpoint (need not exceed 5
concentration levels), and
iii) bottled zero gas
-------
21
B. Gas cylinder specifications
1. Span gases
a) Gas cylinder should be of a 1A size.
b) All cylinders shall meet DOT specifications for 1A cylinders.
2. Concentration Label
A semi-permanent label shall be affixed or attached by the Certifying
Laboratory to each gas cylinder with the following information.
a) Name of the Gas Blender
b) Name of the Laboratory Analyzing the Gas Blend
c) Cylinder I.ID. number
d) Date of Analysis
e) Traceability to NBS or to other certified EPA Mobile Source
Standard
f) Statement of Impurities
g) Gas Concentration
-------
22
C. Durability Criteria
1. Vibration and Shock Test
(See Test Procedure)
2. Sample Line (see Test Procedure)
a) Line crush
3. Sample Handling System Temperature
Effect (20 minutes (? 600°F sample
inlet, see Test Procedure)
4. Filter Check (see Test Procedure)
a) 2 hr sample time
b) Sample until "low flow"
indication
a) zero shift: 2% fs LS
b) span shift: 2% fs LS
c) All analyzer and system
performance checks must
be met.
i) No visible failure or
deterioration
ii) Meet leak check
specifications
iii) Meet response time
specifications
a) No visible failure
or deterioration.
b) Meet leak check
specification.
c) Meet response time
specification.
d) Meet HC hang-up
specification.
i) Low Flow Not Activated
ii) Meet Leak Check
specifications
iii) Meet Response Time
Check
iv) Meet HC hang-up check
i) Time until "low flow"
system activates
ii) Until the low flow
system activates the
system response time
must be met.
-------
23
D. Design Requirements
1. The analysis system shall be designed for a minimum of a 5 year
useful life. Useful life has terminated when the analysis system can
no longer be repaired to meet the specifications in this document at a
cost of less than 60 percent of the replacement cost at the time of
repair.
2. NAME PLATE Permanently Located and Readable.
a) Analyzer System Manufacturer
i) Name
ii) Address
iii) Phone number (customer service)
b) Analyzer Detector or Optical Bench Manufacturer
i) If the same manufacturer as the analyzer manu-
facturer item b) may be deleted.
ii) Name
iii) Addresss
iv) Phone number (customer service)
c) Analyzer Model Number and Serial Number
d) Detector Model Number and Serial Number
e) Date of Analyzer System Assembly
3. Sample System
a) Sample system components should be designed for intermittent
sampling (20 minutes out of 10 hr.) of 600°F inlet (to probe)
exhaust temperature. Such systems shall be designated "Unloaded"
analysis systems. Systems designed for continuous duty with a
minimum of 1000°F inlet (to probe) exhaust temperature, and proper
sample handling equipment (i.e. water removal and/or heated lines)
may be called "Loaded" analysis systems.
b) The type of system (Loaded or Unloaded) shall be permanently
attached and prominently displayed in large letters.
c) The materials used in the sample handing system shall. not alter
the exhaust sample. Some examples of non-reactive materials are
teflon , viton~, stainless steel, silicone rubber (red), and in
some areas nylon . Soie examples of reactive materials are,
brass, copper, and tygon".
d) Water Trap:
-------
24
i) A water trap shall be included in the sample system. The
trap shall be self draining, visible to the operator, and
shall prevent condensable water from occurring in the sample
system downstream of the water trap.
ii) Alternative 1: Electronic correction to a "dry-basis"
reading is permitted if the temperature of the sample after
.the water trap is measured and the sample is assumed to be
saturated at the measured temperature. The water correction
system must be deactivated for leak checks.
iii) Alternative 2: A dry-basis measurement is permitted if
the absolute moisture content entering the analyzer is less
than the saturation moisture content at a sample gas temper-
ature of 7°C (45°F). Ice traps or refrigerators are per-
mitted.
e) Particulate Filter
i) A particulate filter shall be included in the sample
system.
ii) The filter shall be located in a manner that allows
convenient filter element replacement.
iii) The proper direction of flow for the filter body(s) and
element(s) (if applicable) shall be indicated in a manner
that is easily discernible.
iv) The filter body shall be- designed to provide leak free
operation with normal filter element changing frequency for
the lifetime of the analyzer, or a parts list of specific
replaceable parts shall be listed in the maintenance manual
as parts that may contribute to sample system leaks.
v) All manuals and filter element replacement instructions
shall indicate that the system should be leak checked every
time a filter element is changed. A similar message should
be located on or near the filter body(s).
vi) The filter element and filter system shall be designed
to prevent particulates larger in size than 5 microns from
entering the analyzer. Verification of filter particulate
size removal is permitted to be determined by the filter
manufacturer using standardized ASTM or Filter Industry
procedures.
4. Sample Pump
The sample pump shall be designed for at least a 2000 hour life of
continuous duty.
-------
25
5. Sample Probe
a) The probe must sample at least 16 inches from the end of the
tailpipe or dilution adaptor.
b) The probe shall have a flexible portion that will allow the
probe to be inserted in a 1M inch O.D. tailpipe. Assume the
tailpipe has a 3 inch radius 90° bend beginning 4 inches from the
end of the tailpipe.
c) The probe shall be designed for easy servicing and/ or replace-
ment.
d) A universal leak proof dilution adapter (tailpipe extender)
shall be provided.
e) The tailpipe extender shall be able to be attached to the
vehicle within 60 seconds.
f) The tailpipe extender shall be designed to allow the attachment
of standard service center building exhaust evacuation systems.
g) The tailpipe extender shall not alter the sample and the mater-
ial shall conform to D.3.C.) of this chapter.
h) The probe and tailpipe extender shall have sufficient hardware
(insulated handles, etc.) that will allow the user to insert,
attach, or remove the probe or the dilution adapter safely and
conveniently.
i) The probe or tailpipe extender shall be designed in a manner
that will prevent the probe or extender from being removed from
the vehicle unintentionally.
6. Sample Line
a) The sample line shall be flexible at the temperatures to be
encountered during vehicle testing (See Section G)).
b) The sample line shall not be longer than 35 feet nor shorter
than 10 feet (excluding the probe).
c) The portion of the sample line in contact with the exhaust
gases shall not alter the exhaust sample, and shall minimize HC
hangup, due to absorption, adsorption, desorption, outgassing,
etc.
d) The manufacturer must state the estimated useful life of the
sample line in the owners manual or on the name plate.
-------
26
7. Analyzer Spanning System
a) Overview and Frequency: All analyzers shall follow the follow-
ing spanning frequencies:
i) Weekly - gas span
ii) Daily
1) Recommended Basic System - gas span every 4 hours,
electric span every hour
2) Alternative 1 - electric span every hour
(pressure and
temperature
compensation)
3) Alternative 2 - electric span every hour
(extra basic
analyzer accuracy)
b) The analyzer shall be designed for routine gas spanning every
180 hours (once per week).
c) The gas spanning operation must automatically correct for
changes between the gas span point gain and the electrical span
point gain. In all cases the electric span point gain is the
parameter that must be changed such that the new electric span
point gain setting will cause the analyzer to read the span gas
properly.
d) Unless the other spanning alternatives are exercised the basic
analyzer shall be designed for routine gas spanning after:
i) every "power on" and warm-up sequence, and
ii) every 4 hours of "power on" condition.
e) Alternative Spanning System 1: . The analysis system may provide
temperature and pressure compensation to the analyzer output and
spanning system. The compensation shall be based on sample cell
pressure and inlet temperature. If such a system is used, and
verified by subsequent checkout, an electrical span check may be
substituted for the "power on" and 4 hour gas span checks, but not
for the weekly gas span checks. During analyzer checkout, only
one gas span check is allowed as part of the initial analyzer
set-up prior to the initiation of the entire check out procedure.
An electric span shall be substituted for all gas spanning opera-
tions in Chapter V (unless otherwise noted). Additional tests
will be required in order to verify the accuracy and linearity of
the compensation network at the various atmospheric pressures and
temperatures required by the performance specifications (Chapter
IV, Secion E).
-------
27
f) Alternate Spanning System 2: The basic accuracy and precision
of the analysis system may be significantly better than the speci-
fications put forth in this document. In such a case, the analy-
sis system may be capable of meeting the specifications of this
document without any pressure and/or temperature compensation of
the analyzer output and spanning system. If such a claim is made,
and verified by subsequent checkout, an electrical span check may
be substituted for the "power on" and 4 hour gas span checks, but
not for the weekly gas span checks. During analyzer check out,
only one gas span check is allowed as part of the initial analzer
set-up prior to the initiation of the entire checkout procedure.
An electric span shall be substituted for all gas spanning opera-
tions in Chapter V (unless otherwise noted). Additional tests
will be required in order to verify the accuracy of the analysis
system at the various atmospheric pressures and temperatures
required by the performance specifications (Chapter IV, Section
E).
g) All analyzers shall be electrically spanned every hour after
"power on" or 4 hour gas span checks.
h) The analysis system shall include suitable timers to insure
that the spanning frequencies are met. The timing systems shall
also prevent use of the analysis system by driving the readout
devices to full scale and prevent any printer that may be used
from printing. Performing the appropriate spanning operation
shall automatically reset the timer for that specific type of
operation regardless of the time elapsed since that operation was
last performed.
i) The analyzer shall be spanned wi'th flowing zero and span gases.
j) The analyzer system shall calculate and make available to the
operator the span gas set point as ppmh (hexane) based on the span
gas cylinder concentration in propane, and the analyzer's propane/
hexane conversion factor.
k) The span point for all analyzers shall be between 70 and 100
percent of full scale on the lowest range.
1) Appropriate valves, switches, and electrical controls shall be
provided that will allow the operator to conveniently select zero,
span, or sample gases, and zero and span the analyzer.
m) The analysis system shall include a structure for safely secur-
ing two 1A size cylinders.
8. Analyzer Ranges
a) Low-Range : i) 0-400 ppmh HC
ii) 0-2% CO
-------
28
b) High Range : i) 0-2000 ppmh HC
ii) 0-10% CO
c) An analyzer may use one range for HC or CO corresponding to the
high range values provided that the low range portion (0-400,
0-2%) meets all of the analyzer specifications as if that portion
were a separate range. During check out of the system the 0-400
ppmh and 0-2% portions of the high scale shall be treated as a
separate range.
d) All analyzer read out devices (digital or analog) shall be
appropriately scaled and capable of reading negative values up to
-10 percent of full scale for each range regardless of operational
mode (sample, span, leak check, etc.).
e) All HC analyzers and manufacturer's literature shall indicate
hydrocarbons as ppmh or ppmC6. The symbol ppm should never be
used for these analyzers.
f) All analyzers shall have range indicating lights that clearly
indicate which meter scale is to be read. The lights shall be of
sufficient intensity and color that will allow an operator to
identify the range selection in sunlight from a distance of 15
feet.
g) Additional analyzer ranges may be used provided that they meet
all specifications pertaining to the recommended ranges.
9. System Grounding
a) All systems shall have a 3 wire power cord with a 3 prong
grounded plug.
b) All systems shall:
i) have either a ground fault circuit that prevents the
analyzer from performing analysis functions (spanning, samp-
ling, etc.), or
ii) be tested with a 2 wire (non-grounded) extension cord or
non-grounded adapter with the power wires (hot and neutral)
in both positions.
c) If a ground fault circuit is used, the system shall have an
indicator light that activates when a ground fault is determined.
10. System Vents
a) No restrictions such as flowmeters may be placed downstream of
any analyzer vent (a series analyzer flow path is permitted)
unless the system can detect potential changes in restriction
(i.e. sticking flowmeter), and
-------
29
i) alert the operator of the problem which would require a
new gas span and/or repair of the component causing the
restriction, or
ii) use automatic compensation of the analyzer readout device
for the change in restriction.
b) A change in restriction that will cause a 3 percent of point
change in the analyzer response shall activate the alert system.
-------
30
E. Analyzer Performance Specifications
1. Calibration Curve Uncertainty including Bias errors, Precision
errors, and Historesis (Per test procedure) :
a) 5% of point
b) Below 60 ppmh HC or 0.30% CO the uncertainty specification need
not apply.
c) If the analyzer prevents the operator from reading below 10
percent of full scale on the high scale, then the uncertainty
specification need not apply to that portion of the high scale.
2. Resolution:
a) analog meters - 2.5% fs on each range
b) digital meters
i) xxxx ppmh
ii) xx.xx% CO
3. Compensation:
a) Altitude compensation: The analyzer shall have sufficient zero
and span adjustment to allow the spanning of the analyzer at any
altitude between 0 and 7000 feet. The external span adjustment may
be limited in range as long as a clearly marked internal adjustment
will make up the difference.
b) Compensated Systems: (Pressure and temperature compensation of
analyzer read-out and span system)
i) The temperature compensation network shall provide accurate
results over the ambient temperature range specified in Section
G of this chapter as well as exhaust gas temperatures up to
55°C
ii) The pressure compensation network shall provide accurate
and linear results (analyzer read-out) over a pressure range
of + 2 inches HgG from the local barometric pressure. The
system shall operate between 24 and 31 inches HgA. Test
points about which the pressure shall be varied in order to
ascertain the accuracy and linearity of the compensation
network are 24.5, 28.5, and 30.0 inches HgA.
c) Non- compensated Systems: For those analysis systems that claim
pressure and temperature compensation is not necessary, the analy-
sis system shall be tested to specification outlined in step b) for
pressure and temperature compensated systems.
-------
31
4. Zero Drift: + 2% fs L.S. for 1 hour
5. Span Drift: + 2% fs L.S. for 1 hour
6. Noise (clean environment): 0.50% fs L.S.
7. Minimum Sample Cell Temperature: 55°C (131°F)
8. Interferences
a) Gases
i) 14% CO,
ii) Saturated Water
-------
32
9. Propane to Hexane conversion factor: The nean 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. The mean
value shall be known at the 90% confidence level to two significant
figures. The confidence interval shall not exceed a 0.01 increment in
the correction factor. Due to the exceptional hang-up characteristics
of hexane, all components that come in contact with the hexane gas used
for determination^ of the factor shall be either clean stainless steel
or teflon" (viton valve seats are permitted).
-------
33
F. Sample System Performance Specifications
1. Maximum Sample cell pressure variation
between gas spanning and sampling : 4" F20
2. Maximum sample cell
Pressure variation during sampling : 6" HO
3. Maximum sample cell
Pressure variation between normal
flow and low flow indication : 4" HO
4. Response Time : 12 sec
Inlet of probe to 95% fs L.S. ((? low flow indication)
(See Test Procedure)
5. System Leakage Rate : a) sample side 3% of
(See Test Procedure) point on L.S.
b) span gas side, None
6. HC Hangup (see Test Procedure) : 5% fs L.S. in 20 sec.
-------
34
G. Operating Environment
1. The analysis system shall meet the analyzer performance and the
sample system performance specifications under the following conditions
(see Test Procedure):
a) Ambient Temperature: Between 35°F to 110°F
b) Relative Humidity: Between 10% to 99%
2. The analysis system shall be able to be stored at any temperature
between -20°F to 130°F with no adverse effects on subsequent analysis
system performance.
-------
35
H. Fail-Safe Features
1. Required Features For All Pass/Fail Systems
a) Warm-Up
i) The analyzer must have a warm-up lock-out feature with
indicators.
ii) The lock-out feature shall prevent operation of the
printer (if used) and read-out meter until the system is
warmed up.
iii) The lock-out feature shall be activated when:
1) the sample cell is less than the temperature speci-
fied in Chapter IV section E, and
2) When system power is first turned on. The lock-out
shall stay on until the zero drift is stabilized.
Stabilization is determined by observing the zero drift
over a 5 minute period after the lock-out feature deac-
tivates. The zero drift during this 5 minute period may
not exceed one-half of the zero drift specifications in
Chapter IV Section E.
b) Low-Flow
i) The analyzer must have a low sample flow indicator. If
the indicator is activated, the analyzer read-out shall be
driven to 100% of full scale and the printer (if used) shall
be prevented from printing.
ii) 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.
iii) The low flow indicator shall be prominently displayed,
and shall be observable from at least 15 feet away.
c) Leak-Check
i) The sampling system shall have an automatic leak checking
system for the sample side of the system.
ii) The analyzer shall have a timer that will allow the
analyzer to operate for 180 hours (once per week) between
leak checks. If after 180 hours the system is not leak
checked, or the system fails a leak check, the analyzer
readout shall be electronically driven to 100 percent of full
scale, and the printer (if used) shall be prevented from
printing.
-------
36
ill) The leak check shall compare the analyzer's response to
span gas with the analyzer's response to that same span gas
introduced through the probe and sample line. The span gas
flow to the probe shall be restricted such that the pressure
in the sample line during leak checking is approximately
equal to or slightly below the pressure occurring during
sampling.
.iv) The system shall automatically compare the two analyzer
responses, and make a determination of pass or fail.
v) A leak check pass or fail indicator shall be prominently
displayed.
vi) The analysis system shall provide a leak tight receptacle
for the probe on the system structure for the purpose of leak
checking the system.
2. Required Features for All Decentralized Pass/Fail Systems
The following fail-safe features are specified in order to prevent
improper operation of analysis systems used for decentralized systems.
However, they could be useful in centralized systems as well. In
general these features may require an on-board microprocessor.
a) Automatic zero/span check
i) The analyzer shall not have any adjustments available to
the operator for adjusting zero or span point for either the
gas spanning operation or the electrical spanning operation.
These adjustments shall be available in the tamper-proof box
described in Section H.2.d) of this chapter.
ii) The analyzer shall have a selector switch or button (with
indicator light) labled "gas span". Activation of the switch
shall cause the analyzer to automatically perform the gas
spanning sequence with flowing zero and span gas consistent
with the requirements for the manual analyzer spanning system
specified in Section D.7. of this chapter.
iii) The analyzer shall have a selector switch or button with
indicator light labeled "electrical span". Activation of the
switch will cause the analyzer to automatically perform an
electrical zero and span operation consistent with the re-
quirements for the analyzer manual spanning system specified
in Section D.7. of this chapter.
iv) The automatic gas spanning operation shall no't require
more than 90 seconds to complete once the "gas spanning"
switch is activated.
v) The gas span values shall be entered via switches or other
convenient means to the following resolution:
HC = XXXX ppm propane
CO = XX.XX% CO
-------
37
The switches shall he in a tamper proof hox as described in
Section H.2.d) of this chapter.
vi) If the adjusted span voltage changes by more than 20
percent of the previous span voltage, the analyzer read-out
shall be electronically driven to 100% of full scale, the
printer if used shall be prevented from printing, and an
indication of analyzer malfunction shall be prominently
displayed.
vii) The automatic checking feature shall compute the equiva-
lent hexane (HC) span set point based on the input propane
value, and the analyzer propane/hexane conversion factor.
b) Automatic Read System
i) The analyzer shall have a selector switch or button (with
indicator light) labeled "Sample" or "Test".
ii) Activation of the "sample" switch shall cause the analy-
zer system to begin integrating or averaging the analyzer re-
sponse 15 seconds after the switch is activated, and continue
integrating the analyzer response to a flowing sample for the
next 15 seconds.
iii) The analyzer read-out device shall display the inte-
grated value, and hold the display until reset. An indicator
light shall signal the operator when the integrated value is
displayed.
iv) The analyzer shall be prevented from printing the inte-
grated value until the "sample" switch is activated and the
"sample" cycle is completed.
c) Printer
i) The analysis system shall have a printer that provides the
consumer a receipt with the following information:
1) Applicable cutpoints or standards for HC and CO.
2) Integrated vehicle test values for HC and CO.
3) A pass or fail indication.
ii) The print system shall have appropriate means (switches,
etc.) to enter the cutpoints.
iii) The system shall determine items i)2) and i)3).
-------
38
d) Vehicle Diagnosis
i) For the purpose of vehicle diagnosis and/or repairs, the
analyzer shall have a selector switch or button (with indi-
cator light) labeled "Vehicle Diagnosis" or "Vehicle Repair".
ii) Activation of the "Vehicle Diagnosis" switch shall allow
the analyzer to continuously monitor the vehicle exhaust.
iii) The printer, or any automatic data collection system,
shall be prevented from operating anytime the analysis system
is in a "Vehicle Diagnosis" status.
e) Anti Tampering
i) The anti-tampering feature shall be designed to prevent
intentional tampering with the analysis system.
ii) All switches or entry access for automatic zero/span
check adjustments, anti-dilution limits, etc. shall be con-
tained in a tamper-proof box or other tamper-proof mechanism
with provisions for an inspector's seal.
iii) The tamper-proof system must allow convenient access by
an inspector.
3. Anti Dilution (Optional for all systems)
The anti-dilution feature can perform three useful functions: 1) most
important, it can serve as an indicator of vehicle exhaust system leaks
that would cause the measured emission values to be lower; 2) it can be
used to identify dilution of the sample due to either probe placement
or tailpipe extender leaks; and 3) it could potentially be used to
insure decentralized operators actually insert the sample probe (or
attach the extender), and actually measure vehicle exhaust samples.
i) The anti-dilution feature shall identify vehicle exhaust system
leaks and sample dilution.
ii) The preferred technique for identifying leaks is monitoring
the CO levels in the exhaust.
iii) At least three lower limit CO. values shall be capable of
being used.
1) no air pump
2) air pump
3) spare channel
iv) The resolution of the CO. span gas values entered shall be,
XX.X% C0-
-------
v) The CO. values shall be entered by switches or other convenient
means.
vi) The C0? analyzer shall meet all applicable analyzer design and
performance specifications (Chapter IV) between CO- values of 8%
and 14%.
vii) If the C0? is less than the lower limit, the analyzer output
shall be electronically driven to 100% of full scale, the printer
(if used) shall be prevented from printing, and an indication of
exhaust system dilution shall be prominently displayed.
viii) The analyzer operator shall be able to select one of the
three lower limits.
ix) The analyzer shall be prevented from reading auto exhaust
until one of the three limits is selected.
x) If a printer is used, the CO limit for the test shall be
printed.
4. Printer (Optional features for decentralized systems, Print system
optional for centralized systems)
i) The printer shall print the following for each vehicle tested
1) Date
2) Vehicle license plate number
3) Vehicle model year
4) Applicable cutpoints or standards for HC and CO (standard
feature for decentralized systems)
5) Integrated vehicle test values for HC and CO (standard
feature for decentralized systems)
6) A pass or fail indication (standard feature for decentra-
lized systems)
ii) Items 1) to 3) in 4. i) may be deleted if each vehicle tested
is assigned a sequential number by the printing system that can be
transferred to the inspector's invoice or testing form.
iii) Appropriate means (switches etc.) shall be provided to enter
the data required in 4.i) 1) through 4).
iv) The system shall determine items 4.i) 5) and 6).
-------
40
5. Automatic Data Collection (optional for all systems)
i) The automatic data collection feature shall store all of the
pertinent data on a magnetic cassette tape.
ii) The tape shall contain data on at least 150 vehicles.
iii) The tape shall be available only to an appropriate inspector,
and shall be protected by a seal.
iv) A keyboard shall be available to allow the following types of
data to be entered. As indicated, the system may automatically
enter certain data:
1) Date (Auto)
2) Vehicle license plate number
3) VIN
4) Vehicle make
5) Vehicle model
6) Vehicle Model Year
7) Emission Family Number (from emission label)
8) Odometer
9) Test Facility I.D. (Auto)
10) Facility Test Number (Auto)
11) Applicable cutpoints for HC and CO (cutpoints may be
automatically determined and entered based on Vehicle Model
Year selection)
12) Anti-dilution decision criteria (ie. air pump, no air
pump, etc.)
v) The processor shall then enter the following data on the tape.
1) Integrated vehicle test values for HC and CO.
2) A pass or fail indication for HC and CO.
-------
41
System Diagnostic Testing
a) Switches or other devices for rendering any fail-safe or auto-
matic feature inoperative for the purpose of diagnostic or perfor-
mance checking of the analyzer are permitted.
b) These switches (or devices) must be contained in a tamper proof
box(es) or other tamper-proof mechanism with provisions for a
seal.
c) All analyzer systems must be shipped with all fail-safe and
automatic features operating, and the defeat systems sealed.
-------
42
I. System Correlation (on Raw Exhaust) to Laboratory Analyzers
1. All analyzers (comparison to NDIR Analyzer per Test Procedure)
a) Precision Test: A P 1 5%
b) Slope Test: 0.95 < m < 1.05
c) Ratio of Nodal Ave: 0.90 7R 7 1.10
2. HC analyzer (comparison to FID per Test Procedure)
a) Analyzers produced through December 1985:
i) During analyzer check out determine the following para-
meters between the candidate instrument and an FID (See Test
Procedure)
1) Precision
2) Slope comparison
3) Patio of Modal Averages
ii) Each analyzer manufacturer shall use the data to develop
a historical comparison to FID analyzers that can be used to
evaluate, prior to implementation, the correlation parameters
specified for analyzers produced after January 1, 1986.
b) Analyzers produced after January 1, 1986 (see Test Procedure):
i) Precision Test: A P 1 5%
ii) Slope Test: 0.317 1 m ± 0.350
iv) Ratio of Modal Ave: 1.80 1 R 1 2.20
-------
43
J. Manuals
1. Each analyzer shall be delivered with one each of the following
manuals.
a) Operating Instructions
b) Maintenance Instructions
c) Initial Start-up Instructions
2. The manuals shall be constructed of durable materials, and shall
not deteriorate as a result of normal use over a five year period.
3. The operating instructions must contain :
a) The analyzer model and serial number
b) The propane/hexane conversion factor
c) A step by step sequence of pre-test procedures (span, leak
check, probe insertion, etc.)
d) Sampling procedures
e) A step by step sequence of post-test procedures (hang-up etc.)
4. The maintenance manual must contain:
a) Name, address, and phone number of the manufacturer's customer
service and maintenance center at the home office and nearest
field office.
b) Name, address, and phone number of the nearest service center
authorized to make warranty adjustments.
c) A technical description of the system.
d) A separate section that clearly outlines the required or anti-
cipated maintenance schedule. The schedule shall be broken down
into maintenance intervals such as weekly, monthly, etc.
e) A separate section that provides a step by step sequence for
each maintenance requirement.
f) A list of replaceable items such as filters, probes, etc. with
part numbers, and the estimated service life of each component.
g) A list of recommended spare parts that the user should main-
tain.
h) Functional mechanical and electrical schematics.
i) The manufactuer's warranty provisions.
-------
44
i. Each manual shall he attached to the analyzer in a manner that will:
a) allow convenient storage,
b) allow easy use, and
c) prevent accidental loss or destruction.
-------
45
V. Test Procedures
The test procedures in this chapter are presented to provide a consistent
interpretation of the specifications. Although the procedures do provide a
step by step sequence, and it is expected that the evaluator follow that
sequence, they are not a substitute for common sense or good engineering
judgment. Some analysis system designs may simply not be amenable to cer-
tain portions of certain test procedures. Further, there may be ways to
simplify or combine certain test procedures. In those cases, the evaluator
should review the procedure as written, determine the important conceptual
aspects and parameters of the procedure, and then use engineering judgment
in testing the analyzer.
The basic test procedure should be performed in a laboratory that has exper-
ience in automotive emission testing. The general testing should be con-
ducted at an ambient temperature between 68°F and 86°F. Section G requires
certain test procedures to be conducted at other ambient temperatures.
If either of the non-gas spanning alternatives are used (See Chapter IV,
Section D.7.), only one gas span operation is permitted for the entire
Chapter V procedure, unless otherwise noted. This single gas spanning
operation shall take place during the initial start-up procedure. Elec-
trical span will be substituted for all other gas spanning operations.
Analytical gases, of course, will be required for check out of the system.
-------
46
A. Traceability of Analytical Gases
The traceability of analytical gases may or may not be part of the analyzer
manufacturer's responsibility. The choice belongs to the state. However,
we would recommend that the analyzer manufacturers not be held responsible.
We suggest that gas manufacturers meeting certain performance standards be
accredited as acceptable vendors, or that the state provide distribution
service.
The traceability of all gases used in the I/M program is of course impor-
tant. Traceability implies not only blending, but analyzing the gases pro-
perly. Although operating the analysis equipment properly seems like a
trivial task, some very prominent laboratories have sometimes been embar-
rassed by subsequent review of their procedures and analysis of their
blended gases.
In order to provide the automobile manufacturers with what EPA/OMSAPC con-
siders proper traceability procedures, the Emission Control Technology
Division (ECTD) of EPA is publishing a Recommended Practice for Assuring Gas
Traceability to NBS. This procedure, which is a compilation of internal EPA
procedures, is applicable to I/M analytical gases.
The entire traceability procedure may involve more equipment than a state
may wish to become initially involved with. A minimum recommendation for
I/M programs is that each gas blender or certifying laboratory that is
allowed to supply gases for pass/fail I/M systems should follow the ECTD
procedure. The next step would be a continuing quality-audit of each sup-
plier by the state using the correlation portion of the traceability proce-
dure.
-------
B. Gas Cylinder Specifications
No test procedures required.
-------
C. Durability Test Procedures
1. Vibration and Shock Test Procedure
The vibration and shock test procedure is designed to evaluate the
ability of the analyzer to tolerate moderate use in a service facility.
Although a simple span test is performed, the real test of the analyzer
is the ability of the analyzer to meet the remaining performance checks
to follow this test.
a) Equipment Required
i) Candidate instrument.
ii) One span gas between 70 and 90 percent of full scale on
the low range, and bottled zero gas.
iii) A special test floor 6 feet by 10 feet. The top of the
floor shall be elevated 2 inches off the test facility floor.
The floor shall consist of an expanded metal grating with
diamond shape openings measuring 1 x 3.7 inches or equiva-
lent. The length of the floor in the direction of the "short
way of the diamond" shall be 10 feet.
b) Test Sequence
i) Warm up the analyzer.
ii) Span the analyzer on the low range with the span gas per
the manufacturer's recommendations (on the test floor).
iii) Record the zero response and the span response.
iv) Roll the instrument the entire length of the test floor
in the direction of the "short way of the diamond", and off
the end of the test floor onto the facility floor.
v) Pull the analyzer back onto the test floor.
vi) Repeat steps iv) and v) a total of six times.
vii) Check the zero response and span response (do not
adjust the analyzer).
viii) Record the zero response and the span response.
c) Calculations
i) Subtract the zero response from the span response in step
b) iii) (span before).
ii) Subtract the zero response from the span response in step
b) viii) (span after).
-------
49
iii) Subtract the zero response in step b viii) from the zero
response in b) iii) (zero shift).
iv) Subtract the value calculated in step c) ii) from the
value calculated in step c) i) (span shift).
d) Acceptance Criteria
i) If the values calculated in step c iii) and c iv) are less
than or equal to the vibration and shock zero and span shift
specifications, then the vibration and shock performance is
acceptable.
-------
50
2. Sample Line Crush Test Procedure
The sample line crush test is to be repeated under the "operating
environment" testing. If space is a problem in the environmental
chamber, after stabilization, the sample line can be removed from the
chamber to perform the crush test. If the line is removed from the
chamber, the crush test must be performed and the line returned to the
chamber within 3 minutes.
a) Equipment Required
i) Candidate instrument.
ii) One span gas between 70 and 90 percent of full scale on
the low range, and bottled zero gas.
iii) One vehicle with at least 4000 pound curb weight.
b) Test Sequence
i) Stabilize the sample line at the prevailing ambient
temperature.
ii) Warm up the analyzer. Do not turn on the sample pump.
iii) Span the analyzer with the analytical gases.
iv) Leak-check the system.
v) Stretch out the sample line across a solid (concrete etc.)
floor.
vi) Drive across the sample line so that at least one front
and one rear vehicle wheel passes over the sample line.
vii) Back over the sample line so that at least one front and
one rear vehicle wheel passes over the sample line.
viii) Repeat steps vi) and vii) twice.
ix) Leak-check the system.
x) Check for "low flow" indication.
xi) Check for visible failures, kinks, .deterioration, etc.
c) Calculations
(none)
d) Acceptance Criteria
i) If the system passes the leak check (step ix), does not
indicate a "low flow" condition (step x), and shows no sign
of damage (step xi), then the crushability performance of the
sample line is acceptable.
-------
51
3. Sample Handling Temperature Effect
a) Equipment Required
i) Candidate instrument.
ii) Catalyst equipped vehicle.
iii) Thermocouple read-out device.
iv) Type "J" or type "T" thermocouple (1/16 inch diameter
MgO shielded and grounded).
v) One HC span gas between 70 and 90 percent of full scale
on the low range, and bottled zero gas.
vi) Associated fittings.
b) Test Sequence
i) Locate a point approximately 1 inch upstream from the end
of the probe, and weld a thermocouple boss on the extender
(top side when installed on vehicle).
ii) Install the thermocouple into the extender to a distance
approximately half-way between the extender wall and the
sample probe.
iii) Warm up the analyzer.
iv) Span the analyzer with the analytical gases.
v) Leak check the system.
vi) With the vehicle running, locate a position away from
the vehicle to avoid vehicle contamination, and measure
the background HC levels with the sample system. Record
the background HC levels.
vii) Install the extender on the vehicle.
viii) Adjust the vehicle to elevate the exhaust gas tempera-
ture (as measured by the thermocouple) to at least 300°C
(572°F)and not more than 360°C (680°F).
ix) As soon as the exhaust gas has stabilized between 300°C
and 360°C, start a timer and begin sampling.
x) After twenty (20) minutes at the temperature in step
viii), remove the probe or extender, and simultaneously start
another timer.
-------
52
xi) Immediately move the probe or extender to the location
previously used to measure background HC, while continuing to
sample room air.
xii) Twenty (20) seconds after the probe or extender is
removed from the tailpipe, record the analyzer's response.
xiii) Introduce zero gas into the analyzer through the gas
' spanning system. Record the response.
xiv) Introduce span gas into the analyzer through the gas
spanning system. Record the response.
xv) Leak check the system.
xvi) Check for "low flow" indication.
xvii) Check for visible failure, melted or deformed parts,
deterioration etc.
xviii) When the testing is completed, remove the thermocouple
and cap the boss with a leak proof cap.
c) Calculations
i) Compute the difference in zero response between step b)
xiii) and b) xiv) (hang-up HC plus background HC and zero
drift).
ii) Compute the difference in zero response between step b)
xiii) and b) vi) (hang-up HC minus background).
iii) Compute the difference in span response between step b)
xv) and b) iv) (span drift).
d) Acceptance Criteria
i) The value computed in c) i) must be less than the hang-up
specifications plus the zero drift specifications.
ii) The value computed in c) ii) must be less than the hang-
up specifications.
iii) The value computed in c) iii) must be less than the span
shift specifications.
iv) The system must pass the leak check.
v) A low flow condition shall not be indicated.
vi) No portion of the system shall show signs of heat damage.
vii) If the above criteria are met, then the high temperature
performance of the system is acceptable.
-------
53
4. Filter Check and Hang-up Test Procedure
The filter check procedure consists of two phases - a 2 hour check, and
a check to determine the useful life of the filter. It is intended
that all subsequent system check procedures are to be conducted with a
filter and sample system that have experienced at least 2 hours of
exhaust sample as defined in this procedure. Therefore, the second
part of the filter test may be performed after the other tests are
complete if a log of sampling time during the other tests is main-
tained. If during subsequent testing, a low flow indication occurs due
to filter loading, the filter may be changed, the filter life recorded,
and testing may continue with a new filter.
a) Equipment Required
i) Candidate instrument.
ii) Test vehicle that can create a hydrocarbon sample of at
least 1500 ppmh with a spark plug wire removed or the choke
partially closed.
iii) Clean sample filter.
iv) One span gas between 70 and 90 percent of full scale on
the low range, and bottled zero gas.
b) Test Sequence
i) Install a clean sample filter.
ii) Warm up the analyzer.
iii) Span the analyzer with the analytical gases, and leak
check the system.
iv) With the vehicle running, locate a position away from
the vehicle to avoid vehicle contamination, and measure
the background HC levels with the sample system. Record the
background levels.
v) Attach the tailpipe extender to the vehicle.
vi) Insert the probe into the extender and begin sampling.
vii) Begin the vehicle malfunction.
viii) As soon as the emissions exceed 1500 ppmh, begin a timer.
ix) Maintain the HC level above 1500 ppmh, and sample the
vehicle for two (2) hours.
-------
54
x) During the two hours monitor the "low flow" indication.
xi) After two hours, remove the probe or extender, and
simultaneously start another timer.
xii) Immediately move the probe or extender to the location
previously used to measure background HC, while continuing to
sample room air.
xiii) Twenty (20) seconds after the probe or extender is
removed from the tailpipe, record the analyzer's response.
xiv) Introduce zero gas into the analyzer through the gas
spanning system. Record the response.
xv) Introduce span gas into the analyzer through the gas
spanning system. Record the response.
xvi) Leak check the system.
xvii) Alternatives: 1) at this point the filter testing
may be suspended in order to complete the other tests. A
log book of sampling time must be kept for the other tests,
or 2) reinsert the probe into the extender, and continue
testing until the "low flow" system activates, or 3) repeat
the 2 hour filter check at the end of the performance test-
ing of the analyzer, then continue testing until the "low
flow" system activates.
xviii) Record the elapsed time to "low flow" indication.
c) Calculations
i) Compute the difference in zero response between step b)
xiii) and b) xiv) (hang-up HC plus background HC and zero
drift).
ii) Compute the difference in zero response between step b)
xiii) and b) vi) (hang-up HC minus background).
iii) Compute the difference in span response between step b)
xv) and b) iv) (span drift).
d) Acceptance Criteria
i) The value computed in c) i) must be less than the hang-up
specifications plus the zero drift specifications.
ii) The value computed in c) ii) must be less than the hang-
up specifications.
iii) The value computed in c) iii) must be less than the span
shift specifications.
-------
55
iv) A low flow condition shall not occur during the 2 hour
check.
v) If the above criteria are met, then the performance of
the filter is acceptable.
vi) If the total elapsed time to "low flow" activation is
not less than 70 percent of the analyzer manufacturer's
estimate of sample life-time of the filter, the manufac-
turer's estimate is acceptable. If the elapsed time is less
than 70 percent of the estimate, the manufacturer's estimate
shall be revised downward.
-------
56
D. Design Requirement Inspection and Test Procedures
1. Useful Life; No test procedure.
2. Name Plate; Visual observation.
3. Sample System;
a) Visual observation.
b) Visual observation.
c) Visual observation.
d) See "Analyzer Gaseous Interference and Noise Test Procedure,"
Chapter V, Section E.8.
e) i) Statement by analyzer manufacturer on partical size and
element lifetime.
ii) See "Filter Check and Hang-Up Test Procedure."
4. Sample Pump: No test procedure.
5. Sample Probe; Test procedures self-explanatory or features can
be determined by visual observation.
6. Sample Line;
a) Flexibility Test Procedure: Perform the normal motions re-
quired to use the sample line for testing vehicles. Perform this
test for all environmental conditions used for check out under
Chapter V procedures. Make a determination about whether the
sample line can be used without a great deal of difficulty at the
conditions tested. If it is determined that the sample line is
not sufficiently flexible, then the manufacturer may suggest more
objective test procedures and/or more data to demonstrate compli-
ance.
b) Self-explanatory Test Procedure.
c) Visual observation.
d) Statement by analyzer manufacturer.
-------
57
7. Analyzer Spanning System Test Procedure
a) The performance and acceptance of the spanning system will
generally be determined by the system's performance on the other
test procedures conducted during the evaluation procedures, and
visual observation of the equipment features. In other cases
(such as event timers), the test procedures are self-explanatory,
and are not listed. Guidance for the one exception, gas span
versus - electrical span gain adjustments is provided by the fol-
lowing procedure.
b) Equipment Required
i) Candidate instrument.
ii) One span gas 70 to 90 percent of full scale on the low
range, and zero gas.
c) Test Sequence
i) Warm up the analyzer.
ii) Span the analyzer with the analytical gases. Record
the span value.
iii) Perform an electrical span check. Do not adjust
the analyzer. Record the electrical span point.
iv) Reintroduce the span gas. Do not adjust the analyzer.
Note the span response.
v) While the span gas is flowing through the analyzer,
arbitrarily increase the analyzer response (gain setting)
to a value approximately 10% to 15% greater than the span
value.
vi) Perform an electrical span check. Do not adjust the
analyzer. Record the electrical span point.
vii) Reintroduce the span gas. Arbitrarily reduce the
analyzer response (gain setting) to a value approximately
10% to 15% less than the span value.
viii) Perform an electrical span check. Do not adjust the
analyzer. Record the electrical span point.
ix) Repeat steps ii) through viii) a total of three times.
The increased and decreased span adjustments are to be
random settings.
-------
58
d) Calculations
i) Compute the standard deviation(s) and mean(x) of the
electrical responses recorded in steps c) iii) , vi), and
viii).
ii) Multiply the standard deviation(s) by a K factor of 2.6.
iii) Determine the following:
y = x + Ks
y2 = x - Ks
iv) Compute the uncertainty (U) of the electrical spanning
system by:
TT = sPan Point - y .ftn
Electrical Span Point
where, 1=1,2
e) Acceptance Criteria
i) If the uncertainty is less than or equal to the calibration
curve uncertainty (Chapter IV, Section E) , the electrical
span system is acceptable.
-------
59
8. Analyzer Ranges: Visual observation.
9. System Grounding; See Test Procedure E.9. (Analyzer Electrical
Interferences).
10. System Vents; If restrictions downstream of the analyzer exit are
apparent, a test procedure shall be devised that evaluates the analyzer
performance-under restricted conditions. Test Procedures E.3.a) (Alti-
tude Compensation) and F.I. (Sample Cell Pressure Variation, Low Flow,
and Response Time) shall be used for guidance.
-------
60
E. Analyzer Performance Test Procedures
1. Calibration Curve Test Procedure
a) Equipment Required
i) Candidate instrument
ii) Seven or more calibration gases for each range of each
analyzer roughly equally spaced over each range.
iii) Zero gas
iv) Associated valves and fittings
b) Test Sequence
i) If necessary, follow the manufacturer's instructions for
initial start-up and basic operating adjustments.
ii) Warm up the analyzer
iii) Zero the analyzer with the zero gas
iv) Span the analyzer with one of the calibration gases. The
span point should be approximately 80 to 90 percent of full
scale of the low range
v) Recheck the zero. If the zero has shifted, repeat steps
iii), iv), and v) a maximum of one more time.
vi) Do not adjust the zero or span controls on the analyzer
for the remainder of the test.
vii) Introduce the calibration gases in ascending order of
concentrations beginning with the zero gas. Record the
analyzer response to each concentration value.
viii) After the highest concentration has been introduced and
recorded, introduce the same calibration gases to the analy-
zer in a descending order. Include the zero gas. Record the
response of the analyzer to each gas. Record negative zero
responses (if any) as they occur as negative values.
ix) Repeat steps vii), and viii) a total of five times.
c) Calculations
i) For hydrocarbon analyzers, compute the hexane equivalent
(ppmh) of each calibration gas by multiplying the concen-
tration value in ppm propane (ppmp) by the propane/hexane
conversion factor listed on the analyzer.
-------
61
ii) Compute the standard deviation(s) and mean(x) of the
analyzers response for each concentration of calibration
gases. Include both upscale responses and downscale re-
sponses to the same calibration gas. Zero may be a special
case and not amenable to the calculations.
iii) For all concentration values except the highest value,
multiply the standard deviation(s) by a factor (K) of 2.5.
iv) Multiply the standard deviation(s) of the analyzers
response to the highest concentration by a K factor of 3.5.
v) Compute the following for each concentration
1) y = x + Ks
2) y2 = x - Ks
vi) Compute the uncertainty(U) of the calibration curve for
each concentration by:
concentration value-y. n
U — . , i x lUu
concentration value
where i = 1, 2
d) Acceptance Criteria
i) Identify the maximum uncertainty for each range.
ii) If the maximum uncertainty is less than or equal to the
uncertainty specification (Chapter IV, Section E), (plus or
minus), the calibration curve is acceptable. If the uncer-
tainty is greater than specification, the calibration curve
is not acceptable.
iii) If the calibration curve is not acceptable, then the
instrument manufacturer should undertake an engineering study
to identify the cause of the problem prior to continued
testing or introduction of the analyzer to the commercial
market.
iv) After the cause of the problem is identified and the
analyzer is repaired or adjusted this test should be re-
peated.
e) Repeat steps b), c), and d) for each range of the analyzer.
-------
62
2. Analyzer Resolution
a) Analog Meters: The resolution is determined by the interval of
the smallest graduation of the meter face for each analyzer range.
b) Digital Meters: The resolution is determined by the increment
of the least significant digit of the meter readout for each
range.
-------
63
3. Compensation Test Procedures
a) Altitude Compensation Test Procedure (not required for analyzers
with Spanning Alternative, 1 or 2)
i) Eauipment
1) Candidate instrument.
2) One pressure gauge or manometer that reads in absolute
pressure (e.g. 20 to 32 in. HgA) with 1% accuracy.
3) One differential pressure gauge (0 to 15 inches of
water).
4) One flowmeter (0-20 cubic feet per hour (CFH)).
5) Two needle valves.
6) One vacuum pump (greater than 20 CFH'? 24 in. HgA).
7) One span gas 70 to 90 percent of full scale on the low
range.
8) One Tedlar sample bag.
9) Associated lines and fittings (non-reactive).
ii) Test Sequence
This procedure is written around parallel HC and CO analy-
zers, and should be performed on each analyzer. If the
analyzers are in a series configuration, then an additional
pressure gauge(s) will be required, but the test can be
performed on both analyzers at the same time.
If the analyzer manufacturer states in writing that the
structural integrity of the sample cell will not withstand
the pressure differential associated with this test procedure
at 24 inches of HgA, the altitude check must be made in an
altitude chamber. In an altitude chamber only steps 8), 9),
10), 14), 15), 16) and 20) need be performed at the pressures
specified in step 13) and 18).
1) Identify the sample line entering the sample cell of
each analyzer.
2) Install a tee fitting in the sample line immediately
upstream of the sample cell (as close as practical).
Install the tee with branch pointing up.
3) Immediately upstream of the tee install one of the
needle valves.
-------
64
4) Identify the exhaust line from the sample cell and
install two tees or a 4-way cross fitting in the line
with the branches pointing up.
5) Downstream of the tees or cross fitting install the
flow meter.
6) Attach a differential pressure gauge to the tee
upstream of the sample cell and one of the exhaust tees.
Attach one gauge across each analyzer in a series sys-
tem.
7) Attach the absolute pressure gauge to the second
exhaust tee.
8) Warm up the analyzer.
9) Check for leaks.
10) Sample room air through the sample line with the
needle valve wide open.
11) Record the sample cell exhaust flow rate, the dif-
ferential pressure across each analyzer, and the abso-
lute pressure.
12) Attach the second needle valve after the flowtneter.
13) Adjust both the needle valves to obtain an absolute
pressure of 31 inches of HgA, the sample differential
pressure recorded in step ii), and a flow-rate of not
less than 2 cfh.
14) Fill the sample bag with the span gas, and introduce
the span gas from the bag into the sample line.
15) Use the analytical gases to set the zero and span of
the analyzer. Note if internal adjustments were re-
quired to span the analyzer.
16) Return to sampling room air.
17) Attach the vacuum pump to the exhaust line after the
second needle valve with an appropriate length of line,
and open both needle valves.
18) Turn the vacuum pump on, and adjust both needle
valves to obtain an absolute pressure of 24 inches of
HgA, the same differential pressure recorded in step
11), and a flow rate approximately the same as recorded
in step 11) .
-------
65
19) Refill the sample bag if necessary, and introduce
the span gas into the sample line from the bag.
20) Use the analytical gases to set the zero and span of
the analyzer. Note if internal adjustments were re-
quired to span the analyzer.
iii) Calculations (none)
iv) Acceptance Criteria
1) If the analyzer can be spanned properly in steps ii)
15), and ii) 20, and the technique required to span the
analyzer under these conditions is identified in the
owner's manual, then the altitude compensation network
is acceptable.
-------
66
b) Pressure and Temperature Compensated Analyzer Test Procedure
i) Testing concepts: This test procedure is to be performed
in order to identify the performance of any pressure or
temperature compensation systems under the various environ-
mental conditions that may be encountered during vehicle
inspection testing. In general, temperature compensation
will be evaluated during the basic check-out (68°F to 86°F)
and during the more hostile environmental temperature tests
specified in Section G. No other special testing would
normally be necessary.
In order to evaluate pressure compensation systems, addi-
tional testing is necessary. If the analyzer manufacturer
can make a case that testing the pressure compensation system
in a manner similar to the procedure specified in Chapter V
Section E.S.a) (altitude compensation) will represent actual
analysis system operating conditions in the field, then that
procedure (E.3.a) may be used for check-out. If a sufficient
case cannot be made, and a suitable alternative test proce-
dure cannot be determined, then performance evaluations of
the pressure compensation system must be carried out in an
altitude chamber. The pressure compensation test shall be
conducted at each environmental temperature condition speci-
fied in Chapter V. The tests shall be performed on each
range of each analyzer.
ii) Test conditions: The test conditions shall consist of
three basic barometric pressures about which excursions in
pressure shall be made. The values are:
Basic Test Excursion
Point (inches HgA) Points (inches HgA)
24.5 24.0, 26.5
28.5 26.5, 29.5
30.0 28.5, 31
iii) Test Sequence
1) Identify the concentration value that provides the
greatest uncertainty in measurement as determined by
Test Procedure E.I). Obtain a calibration gas at that
level (same bottle if possible).
2) Set up the test equipment for one of the three test
points.
3) Warm up the analyzer.
4) Gas span the analyzer with a different calibration
gas between 80 and 90 percent of full scale on the low
range and bottled zero gas.
-------
67
5) Leak check the system.
6) Alternately introduce zero gas and the calibration
gas identified in step 1) through the sample prohe until
a total of three readings are obtained and recorded for
each test step.
7) The test steps for the 24.5 in HgA test point are:
Test Step Action
24.5 gas span
24.0 3 readings
24.5 3 readings
26.5 3 readings
24.0 3 readings
26.5 3 readings
24.5 3 readings
24.0 3 readings
24.5 3 readings
26.5 3 readings
8) Repeat steps 4) through 6) at the 28.5 inch HgA test
point. The test steps for this test point are:
Test Step Action
28.5 gas span
26.5 3 readings
28.5 3 readings
29.5 3 readings
26.5 3 readings
29.5 3 readings
28.5 3 readings
26.5 3 readings
28.5 3 readings
29.5 3 readings
9) Repeat steps 4) through 6) at the 30.0 in. HgA test
point. The test steps for this test point are:
Test Step Action
30.0 gas span
28.5 3 readings
30.0 3 readings
31.0 3 readings
28.5 3 readings
31.0 3 readings
30.0 3 readings
28.5 3 readings
30.0 3 readings
31.0 3 readings
-------
68
iv) Calculations
1) For each test point, compute the standard devia-
tion (s) and mean (x) for the 9 readings obtained at each
pressure level.
2) Multiply each standard deviation by a K factor of
2.6.
3) For each pressure level at each test point, compute
the following:
y. = x + Ks
y2 = x - Ks
4) For each pressure level at each test point, compute
the uncertainty (U) of the measurements based on the
calibration gas used in step iii) 6) from:
concentration value-v. ,nr,
U = : : "-f x 100
concentration value
where i = 1, 2
v) Acceptance Criteria
1) Identify the maximum uncertainty
2) If the maximum uncertainty is less than or equal to
the uncertainty specification (Chapter IV, Section E),
(plus or minus), for the basic analyzer calibration
curve, the pressure compensation is acceptable. If the
uncertainty is greater than specification, the pressure
compensation is not acceptable.
3) If the pressure compensation is not acceptable, then
the instrument manufacturer should undertake an engin-
eering study to identify the cause of the problem prior
to continued testing or introduction of the analyzer to
the commercial market.
4) After the cause of the problem is identified and the
analyzer is repaired or adjusted this test should be
repeated.
vi) Repeat steps iii), iv), and v) for each range of the
analyzer.
-------
69
c) Test Procedure for Analyzers with Improved Accuracy
i) If an analyzer manufacturer claims that the measurement
uncertainty specifications (Chapter IV, Section H.) can he
met without pressure and/or temperature compensation, this
claim can be verified by the test procedure in Section K.
3-6) of this Chapter.
-------
70
4. Analyzer Zero and Span Drift Test Procedure
(See Test Procedure Section F.I.a), Warm-up Lock-out Test Procedure)
a) Equipment Required
i) Candidate instrument.
•ii) One span gas between 70 to 90 percent of full scale on
the low range.
iii) One cylinder of zero gas.
iv) One chart recorder (one megohm impedance or greater) with
approximately a 10 to 12 inch wide chart.
b) Test Sequence
i) Begin the test sequence with an analyzer that is turned
off, and has stabilized at the prevailing ambient temperature
for at least 3 hours.
ii) Remove the analyzer's protective cover.
iii) Locate the meter readout and attach leads from the
readout terminals to the chart recorder.
iv) Reinstall the protective cover.
v) Turn on the analyzer.
vi) Select a voltage range on the chart recorder so that full
scale of chart recorder equals full scale voltage of the
analyzer meter on the range under test.
vii) The chart recorder must indicate both negative and
positive zero drift. If necessary, offset the chart recorder
zero 5 units up scale.
viii) As soon as the warm-up lock-out feature deactivates,
zero and span the analyzer with the analytical gases per
manufacturer's operating instructions, and start the chart
recorder at a minimum chart speed of 0.5 inches per minute.
ix) Do not adjust the analyzer or recorder controls (zero or
span) for the remainder of the test.
x) Reintroduce zero gas (if not done already) and start the
test.
xi) Mark the chart paper indicating the zero response, the
span response, and the start of the test.
-------
71
xii) Flow zero gas through the analyzer for one hour.
xiii) Introduce the span gas to the analyzer (do not adjust
the equipment).
xiv) Reintroduce the zero gas to the analyzer (do not adjust
the equipment).
xv) Reintroduce the span gas to the analyzer (do not adjust
the analyzer).
xvi) Flow span gas through the analyzer for one hour.
xvii) Reintroduce zero gas to the analyzer.
c) Calculations
i) Compute the difference (as a percent of full scale chart
deflection) between the analyzer zero response for the first
span check (step b)viii)) and the zero response after the 1
hour span check (step b) xiv)) (zero drift).
ii) Locate on the chart the maximum and minimum analyzer zero
response during the first 1 hour period (step b) xii)).
iii) Compute the difference (as a percentage of full scale
chart deflection) between the analyzer zero-response for the
first span check (step b) viii)), and the maximum zero re-
sponse and then the minimum zero-response as identified in
step c) ii) (zero drift).
iv) Compute the analyzer span-response at the 1 hour span
check as the difference between the chart reading of the span
gas (step b) xiii) and the chart reading of the zero gas
(step b) xiv)) (span response).
v) Locate on the chart the maximum and minimum chart reading
during the one hour span period (step b) xvi)).
vi) Compute the difference (as a percentage of full scale
chart deflection) between the analyzer zero-response immedi-
ately prior to the one hour span period (step b) xiv)), and
the maximum and minimum span chart readings identified in
step c) v) (span response).
vii) Compute the difference (as a percentage of full scale
chart deflection) between the span chart deflection at the
end of the 1 hour span, and the zero-response after the 1
hour span (span response).
-------
72
viii) Compute the difference (as a percentage of full scale
chart deflection) between the span chart deflection immedi-
ately prior to the start of the one hour zero test (step b)
viii)) and the zero-response just prior to that span
check (span response).
ix) Compute the difference (as a percent of full scale chart
deflection) between the zero-response just prior to the start
-of the 1 hour span (step b) xiv)), and the zero-response just
after the 1 hour span (step b) xvii)) (zero drift).
x) Compute difference (as a percentage of full scale chart
deflection) between the following span responses in step c) :
(vi max) to (iv) span drift
(vi min) to (iv) span drift
(vii) to (iv) span drift
(viii) to (iv) span drift
d) Acceptance Criteria
i) If each value computed in steps i), iii) max, iii) min,
and ix) is less than or equal to the specifications listed
for zero drift, the zero drift of the analyzer is acceptable.
ii) If each value computed in step x) is less than or equal
to the specifications for span drift, the span drift of the
analyzer is acceptable.
-------
73
5. Analyzer Span Drift Test Procedure
(See Analyzer Zero and Span Drift Test Procedure, E.4.)
6. Analyzer Noise Test Procedure
(See Analyzer Gaseous Interference and Noise Test Procedure, E.8.)
-------
74
7. Analyzer Sample Cell Temperature Test Procedure
(See Test Procedure Section H.I.a), Warm-up Lock-out Test Procedure).
a) Equipment Required
i) Candidate instrument.
ii) Thermocouple readout device.
iii) Type J or type T thermocouple.
b) Test Considerations
i) The sample cell test procedure may be run concurrently
with the zero/span drift test or any other test that begins
with an analyzer that is not warmed up.
ii) It is recommended that the test equipment remain hooked
up to the analyzer for the duration of the analyzer check
out, and the temperature of the sample cell should be moni-
tored from time to time.
c) Test Sequence
i) Locate on the sample cell a point, based on engineering
judgement, that would be the coldest point.
ii) Attach a thermocouple to the sample cell at the location
identified in step i).
iii) Conductive, convective, and radiation losses must be
considered in the selection of the location, and in the
manner of thermocouple attachment.
iv) Stabilize the analyzer at the prevailing ambient temper-
ature for at least 3 hours.
v) Record the sample cell temperature.
vi) Turn on the analyzer.
vii) As soon as the warm-up lock out feature is deactivated,
record the temperature of the sample cell.
viii) Immediately switch to sample ambient air (or analytical
gas at the ambient temperature) through the sample line.
ix) Monitor the sample cell temperature over the next 5
minutes.
x) Record the lowest sample cell temperature during the 5
minute period.
-------
75
d) Calculations
(None)
e) Acceptance Criteria
i) If the temperature recorded in step c) v) is approximately
the same as the ambient temperature, the analyzer is properly
stabilized.
ii) If the temperatures recorded in step c) vii) and step c)
x) are equal to or greater than the specifications for sample
cell temperature, the sample cell temperature is acceptable.
iii) While monitoring the sample cell temperature during the
analyzer check-out procedure, if the temperature is less than
the specifications when concentration data (span or sample)
is read, the acceptance in step ii) is void.
-------
76
8. Analyzer Gaseous Interference and Noise Test Procedure
a) Equipment Required
i) Candidate instrument.
ii) One high impedence (1 megohm or greater) digital volt-
meter with at least 4% digit resolution. Generally, a 2 volt
,(1.9999 v) and a 20 volt (19.999 v) scale will be adequate.
A chart recorder of similar impedence, resolution, and scal-
ing may be used in place of a voltmeter.
iii) One calibration gas with approximately 14% CO and the
balance N
iv) One steam generator.
v) One mixing chamber with probe attachment, dilution valve,
and chamber air temperature readout system.
vi) One calibration gas with approximately 100 ppm NO . The
NO value must have been checked within 48 hours of the start
of the testing sequence with a chemiluminescent analyzer
meeting the specifications of 40 CFP 86 Subpart B, D, or N.
vii) One span gas between 70 and 90 percent of full scale on
the low range, and bottled zero gas.
viii) Tedlar sample bags.
ix) Associated lines and fittings (non-reactive).
b) Test Sequence
i) Remove the analyzer's protective cover.
ii) Locate the meter readout and attach the voltmeter or
chart recorder leads to the meter terminals.
iii) Reinstall the analyzer's protective cover.
iv) Temporarily bypass the water trap.
v) Warm up the analyzer.
vi) Introduce the span gas to the analyzer through the span
port.
vii) Adjust the span so that the analyzer reads 100 percent
of low scale.
-------
77
viii) Observe the voltmeter reading at 100 percent full scale
of the analyzer and record the voltage value. Alternatively,
adjust the chart recorder so that full scale is equivalent to
the analyzer full scale.
ix) Zero and span the analyzer with the analytical gases to
the proper values.
x) Fill a sample bag with span gas.
xi) Introduce the span gas from the sample bag to the analy-
zer through the probe.
xii) After the reading has stabilized, observe the voltmeter
or chart recorder. Record the highest value and the lowest
value over a 3 minute time span. The scale can be changed
for better resolution.
xiii) Fill another sample bag with CO .
xiv) Introduce the CO through the sample probe.
xv) Record the average voltmeter or chart recorder reading.
The scale can be changed for better resolution.
xvi) Fill a sample bag (after purging with N ) with NO .
xvii) Introduce the NO through the sample probe.
xviii) Record the average voltmeter or chart recorder read-
ing. The scale can be changed-for better resolution.
xix) Start the steam generator.
xx) Attach the probe to the mixing chamber and adjust the
dilution valve to obtain a 40°C (101°F) temperature in the
dilution box. To prevent condensation in the analysis system
this test should be performed with an ambient temperature
between 30°C and AO°C. This test cannot be performed cor-
rectly when testing at the lower ambient temperature condi-
tions specified in Section G of this chapter, and should
therefore, be omitted when testing under Section G.
xxi) Sample from the dilution box.
xxii) Record the average voltmeter or chart recorder reading.
The scale can be changed for better resolution.
xxiii) Reconnect the water trap.
-------
78
xxiv) Repeat steps xix) through xxii) with an amhient temp-
erature at the same value as in step xx) and 40°C in the
dilution box. When repeating this test at the operating
environmental conditions specified in Section G of this
chapter, perform this test with a 40°C dilution box tempera-
ture and the analysis system at the prevailing ambient temp-
erature.
c) Calculations
i) Compute the difference between the high and low reading
recorded in step b) xii), then divide that result by the full
scale voltage value recorded in step b) vii). This value,
expressed as a percentage and divided by 2, is defined as
noise of the analyzer.
ii) Divide the value recorded in step b) xv) by the full
scale voltage value recorded in step b) vii). The result,
expressed as a percentage, is the CC> interference of the
analyzer.
iii) Divide the value recorded in step b)xviii) by the full
scale voltage value recorded in step b)vii). The result,
expressed as a percentage, is the NO interference of the
analyzer.
iv) Divide the value recorded in step b)xxii) by the full
scale voltage value recorded in step b)vii). The result,
expressed as a percentage, is the water interference of the
analyzer.
v) Divide the value recorded in step b)xxiv) by the full
scale voltage value recorded in step b)vii). The result,
expressed as a percentage, is the water interference of the
system.
d) Acceptance Criteria
i) If the percentages calculated in paragraph c) for the
analyzer are equal to or less than the specifications for
noise and gaseous interferences (Chapter V, Section E.), then
the noise and gaseous interferences are acceptable.
ii) Inspect the analysis system downstream for condensed
water. If water is found, the effectiveness of the water
trap is not acceptable.
-------
79
9. Analyzer Electrical Interference Test Procedure
a) Equipment Required
i) Candidate instrument.
ii) One high impedence (Imegohm or greater) digital volt
meter with at least 4i$ digit resolution. Generally, a 2 volt
(1.9999 v) and a 20 volt (19.999 v) scale will be adequate.
'A chart recorder of similar impedence, resolution, and scal-
ing may be used in place of a voltmeter.
iii) One vehicle with high energy ignition system, and solid
core ignition and coil wires.
iv) One 3 amp or more variable speed (commutator type) hand
drill with a plastic handle.
v) One 20 foot extension cord with 3 sixteen gauge wires, and
a 2 plug outlet. For systems without ground fault circuits,
a 2 wire extention cord or a non-grounded adaptor will be
required.
vi) A CB transmitter at or near FCC legal maximum power with
a matching antenna.
vii) One variable voltage transformer (90v to 130vAC).
viii) One span gas between 70 and 90 percent of full scale on
the low range.
ix) One dry flannel cloth.
b) Test Sequence
i) Remove the analyzer's protective cover.
ii) Locate the meter readout, and attach the voltmeter or
chart recorder leads to the meter terminals.
iii) Peinstall the analyzer's protective cover.
iv) Plug the analyzer's power cord into a 2-outlet grounded
electrical outlet box and warm up the analyzer.
v) Use a non-grounded adapter to check any ground fault
circuit.
vi) If the system does not have a ground fault circuit,
perform the remainder of the test sequence with the non-
grounded adapter.
-------
80
vii) Mark the adapter and the outlet box, and perform the
testing sequence with the adapter in that orientation. After
completion of the sequence, reverse the adapter orientation
(hot lead versus neutral lead) and repeat the sequence.
viii) Introduce the span gas through the span port, and
adjust the span so that the analyzer reads 100 percent of low
scale.
-ix) Observe the voltmeter or chart recorder reading at 100
percent full scale of the analyzer, and record the voltage
value.
x) Zero and span the analyzer with the analytical gases to
the proper values. Record the values.
xi) Introduce span gas to the analyzer. Record the average
span gas voltage levels over a 5 minute time span. The scale
can be changed for better resolution.
xii) Either move the analyzer or the vehicle (engine compart-
ment) to within 2 feet of each other.
xiii) Introduce the span gas to the analyzer.
xiv) Open the hood and start the vehicle.
xv) Record the average span gas voltage level during a 5
minute time span.
xvi) Stop the vehicle engine.
xvii) Plug the electric drill into the other outlet of the
same receptacle that the analyzer is connected to.
xviii) Move the drill at approximately 3 to 4 feet high to
within 12 inches of the analyzer. At four locations around
the analyzer vary the drill speed from minimum to maximum
speed. Attempt to locate the positions that provide the
greatest interference.
xix) Record the average span gas voltage level at each posi-
tion.
xx) Key the CB radio within 50 feet of the analyzer.
xxi) Record the average span gas voltage level when the CB
radio is transmitting.
xxi) Turn the analyzer off.
xxii) Plug both the analyzer and the drill into the extension
cord. For systems without a ground fault circuit use either
a 2 wire extention cord, or a non-grounded adapter.
-------
81
xxiii) Plug the extension cord into the variable voltage
transformer and set the transformer output to 110 volts.
xxiv) Turn the analyzer on, and warm it up.
xxv) Check the zero and span. If necessary reset to the
exact voltage level recorded in step b) x).
- xxvi) Repeat step b) xii) through xxi).
xxvii) Record the average span gas voltage level as indicated
in steps b) xv), xix), and xxi).
xxviii) Reduce the transformer output voltage to 90 volts.
Let analyzer stabilize for 2 minutes.
xxix) Repeat steps b) viii) through xxi) with the extention
cord and 90 volt set-up. Non-relevant set-up steps may be
omitted.
xxx) Increase the transformer output to 130 volts. Let the
analyzer stabilize for 2 minutes.
xxxi) Repeat steps b) viii) through xxi) with the extension
cord and 130 volt set-up. Non-relevant set-up steps may be
omitted.
xxxii) If the unit has an analog meter, observe and record
the meter reading with span gas.
xxxiii) Rub the meter face 10 times with the dry flannel
cloth in the up-scale direction. Record the meter reading.
c) Calculations
i) Compute the difference between each set of voltage levels
recorded in step b) x) and b) xv). The maximum difference
expressed as a percentage of the voltage recorded in step b)
ix) is defined as the RF interference (RFI).
ii) Compute the difference between each set of voltage levels
recorded in step b) x) and b) xx). The maximum difference
expressed as a percentage of the voltage recorded in step
,b)ix) is defined as the induction interference.
iii) Compute the difference between each set of voltage
levels recorded in step b)x) and b)xxi). The maximum differ-
ence expressed as a percentage of the voltage recorded in
step b)ix) is defined as the VHP interference.
iv) Compute the difference between the voltage levels re-
corded in step b) x) and step b) xxvii). The largest result
expressed as a percentage of the voltage recorded in step b)
ix) is defined as the line interference.
-------
82
v) Compute the difference between the voltage levels recorded
in step b) x) and b) xxix). Compute the difference between
the voltage levels recorded in step b) x) and b) xxxi). The
largest value of the results expressed as a percentage of the
voltage recorded in step b) ix) is defined as the line vol-
tage interference.
vi) Compute the difference between the meter reading recorded
'in step b) xxxii) and b) xxxiii). Also compute the difference
as a percentage of full scale. The difference is defined as
the static electricity interference.
d) Acceptance Criterion
i) If the percentages calculated in paragraph c) are equal to
or less than the electronic interference specifications
(Chapter TV, Section E.), then the electrical interferences
are acceptable.
-------
83
LO. Propane to Hexane Conversion Factor Test Procedure
a) Testing Concepts
i) The conversion factor shall be evaluated at two hexane
concentration levels.
1) 200 ppmh ( _+ 15 ppmh) on the low scale, and
2) 1800 ppmh ( +_ 150 ppmh) on the high scale.
ii) A minimum of three samples of the hexane gas at each
concentration level shall be performed.
iii) Because of the exceptional hang-up characteristics of
hexane, an auxilliary sample/spanning system that is specially
constructed to be essentially hang-up free may be used. The
auxilliary system should be of laboratory quality, and may
bypass the candidate analysis system controls.
iv) The test should always be conducted at an ambient temper-
ature greater than 20°C (68°F).
v) A flame ionization detector (FID) properly optimized and
calibrated . (see 40 CFR 86 Subpart D or N) may be used to
provide additional quality control on the conversion factor
determination procedure.
b) Test Sequence
i) Set up the test equipment.
ii) Warm-up the candidate analyzer.
iii) Leak check the system used.
iv) Span the analyzer on the low range with a propane cali-
bration gas near ( _+ 5%) the expected low concentration hexane
response, and with bottled zero gas. Use the propane to
hexane conversion factor indicated on the analyzer.
v) Alternately cycle low concentration hexane gas and the
calibration gas through the analyzer a total of 3 or more
times.
vi) Record each response for each gas.
vii) Span the analyzer on the high range with a propane
calibration gas near ( +_ 5%) the expected high concentration
response, and bottled zero gas. Use the propane to hexane
conversion factor indicated on the analyzer.
-------
84
viii) Alternately cycle the high concentration hexane gas and
the calibration gas through the analyzer a total of three
or more times.
ix) Fecord each response for each gas.
c) Calculations
i) For the values recorded in step b) vi), and b) ix), compute
the mean (x) and standard deviation(s) for the hexane response
and the propane response.
ii) Use an aof 0.05 and the student's "t" test to determine
the confidence interval of the population mean at the 90%
confidence level based on the values computed in step c) i)
for each range.
iii) Use the following equation to determine the mean propane
to hexane conversion factor (CF) for each range:
(.5) + (.5)(x - Hex Cal)
CF = hex
Hex Cal
where:
x hex = the mean hexane response determined in
step c) i).
Hex Cal = concentration of hexane calibration gas.
iv) Determine if sufficient number of cycles (steps b) v)
and b) viii)) were run on each range by the following equation:
(.5)(x, - xr_) + (.5) Hex Cal
o.oi <_ hex CI ~
Hex Cal
where:
x hex = the mean hexane response determined in step c) i).
xp = the confidence interval determined in step c) ii).
*-• L
Hex Cal = concentration of hexane calibration gas.
v) For each range, multiply the mean propane response (x)
determined in step c) i) by the CF computed in step c) iii).
vi) Determine the difference between the CFs for each range,
and determine the mean CF.
-------
85
d) Acceptance Criteria
i) If the value computed in step c) iv) is greater than 0.01,
perform additional cycles (see steps b) v) and b) viii)) until
the computed value is less than 0.01.
ii) If the difference between the CFs computed for each range
(step c) vi)) is greater than .030 a sperate CF must be used
, for each range. If the difference is less than .030, a mean
CF may be used for both ranges.
iii) The CF value(s) determined by step d) iii) must agree
exactly with the CF posted on the analyzer. ASTM round-off
shall be used.
iv) The mean propane value determined in step c) v) must agree
with the propane calibration gas within 1 percent of the propan
calibration gas concentration.
-------
86
F. Sample System Test Procedures
1. Sample Cell Pressure Variation, Low Flow, and Response Time
a) Test Equipment
i) Candidate instrument system.
- ii) One span gas cylinder with a concentration between 70 and
90 percent of full scale, and bottled zero gas.
iii) One gauge (0-30 inches of water).
iv) One pinch clamp.
v) One Tedlar sample bag (approximately 3 cubic foot capacity)
vi) One 3-way ball valve, minimum orifice size is .180 inches.
vii) One stop watch or timer.
ix) Associated fittings and lines (non-reactive material).
b) Testing Sequence
This test procedure is written around parallel HC and CO analy-
zers, and should be performed on each analyzer. If the analyzers
are in a series configuration then an additional pressure gauge
will be required, but the test can be performed on both analyzers
at the same time.
i) Identify the sample line entering the sample cell.
ii) Install a tee fitting in the sample line immediately
upstream of the sample cell (in as close as practical).
Install the tee with the branch pointing up.
iii) Connect the gauge to the tee with the suitable length of
flexible tubing.
iv) Warm up the analyzer.
v) Zero the analyzer.
vi) Span the analyzer with the span gas through the spanning
network. Record the pressure gauge reading, and the span
response.
vii) Fecheck the zero. If the zero has shifted repeat steps
v), vi), and vii).
viii) Switch the analyzer from span gas flow to sample flow
(pump on).
-------
87
ix) Record the maximum and minimum fluctuations in the gauge
reading.
x) Attach the pinch clamp to the pressure gauge line. Slowly
pinch the line until the gauge fluctuations are minimized.
xi) The gauge reading should be approximately halfway between
the readings recorded in step ix) . Record the stabilized
gauge reading in step x) .
xii) Attach the common port of the 3-way valve to the probe
with a leak free adapter.
xiii) Attach a short length of tube to one of the remaining
two ports of the three-way valve, and sample room air through
the tube.
xiv) Fill the sample bag with span gas.
xv) Attach the sample bag to the remaining port of the 3-way
valve.
xvi) The 3— way valve should be sampling room air.
xvii) switch the 3-way valve to the sample bag, compare the
stabilized concentration reading of the bag to the reading in
step vi) . A difference of more than 1 percent from the con-
centration reading in step vi) could indicate a leak in the
system. Repair any leaks, and restart the procedure at step
xviii) Recheck the zero by switching the 3 way valve back to
room air.
xix) Identify the concentration value corresponding to 95% of
the span gas in the bag. Record the 95% value.
xx) Switch the 3-way valve to the sample bag and simultane-
ously start a timer.
xxi) When the analyzer response reaches the 95% value stop
the timer. Record the elapsed time.
xxii) Record the sample cell pressure after the analyzer
reading has stabilized. Record the stabilized analyzer
concentration reading.
xxiii) Identify the concentration value (5%) corresponding to
the stabilized reading in step xxii) (100%) minus the 95%
value identified in step xix) .
xxiv) switch the 3-way value back to room air and simul-
taneously start the timer.
-------
xxv) When the analyzer reaches the 5% value, stop the timer.
Record the elapsed time.
xxvi) Connect the needle valve between the probe and the
3-way valve.
xxvii) Adjust the needle valve until the low flow indicator
is just barely activated.
xxviii) Repeat steps xvi) through xxv).
xxix) Use the same 95% and 5% values identified in step xix)
and xxiii).
xxx) Record the 5% and 95% response times.
xxxi) Record the stabilized concentration reading, and sample
cell pressure in the same manner used in step xxii).
c) Calculations
i) Compute the difference in pressure between step vi) and
step xi), and step vi) and step xxii). The values will be
the pressure difference between spanning and sampling.
ii) Compute the difference between maximum and minimum pres-
sure readings in step b) ix). The value will be the pressure
variation during sampling.
iii) Compute the difference in pressure between step xi) and
step b) xxxi). The value- is the difference in pressure
variation between normal flow and low flow conditions.
iv) Compute the percentage change in analyzer gas response
between step b) xxii) and step b) xxxi).
d) Acceptance Criteria
i) If the calculated value in steps c)i, and c)iii are less
than 4 inches of water, and less then 6 inches of water for
step c)ii, the pressure variations in the analyzer flow
system under these test conditions are acceptable.
ii) If the change in analyzer response computed in step c)iv
is less than 1.5 percent and the elapsed times recorded in
step b) xxx) are less than 12 seconds, the low flow indicator
system is acceptable as well as the system response time.
iii) If the elapsed time in steps b) xxi) and b) xxv) is less
then 12 seconds, the system response time is acceptable under
normal conditions.
-------
89
iv) It is recommended that the sample cell pressure gauge(s)
remain hooked up for the duration of the analyzer check out.
If the sample pressure varies by more then 4 inches of water
from the most recent span pressure, then the acceptance in
d)i is void.
-------
2. Maximum Sample Cell Pressure Variation during Sampling Test Procedure
(See Test Procedure F.I)
3. Maximum Sample Cell Pressure Variation Between Normal Flow
and Low Flow Indication Test Procedure
(See Test Procedure F.I)
4. Response Time Test Procedure
(See Test Procedure F.I)
-------
91
5. System Leakage Test Procedures
a) Equipment Required
i) Candidate instrument.
ii) One tee fitting.
iii) One needle valve.
iv) One span gas 70 to 90 percent of full scale on the low
range, and bottled zero gas.
v) Associate lines and fittings (non-reactive).
b) Test Sequence
This sequence is written assuming that the analyzer incorporates
the automatic leak checking feature specified.
i) Install the tee fitting between the probe and the sample
line.
ii) Attach the needle valve to the remaining port of the tee
and close the valve.
iii) Warm up the analyzer.
iv) Zero and span the analyzer with the analytical gases.
v) Record the span value.
vi) Place the probe in the leak check receptacle.
vii) Record the analyzer response to span gas introduced
through the leak check recepticle, probe, and sample
line.
viii) If the analyzer response in step vii) differs from the
response in step v) by more than 1 percent of the value
recorded in step v), check the system for leaks or other
problems. Restart the procedure at step iii).
ix) Gradually open the needle valve until the leak check fail
light just comes on.
x) Record the analyzer response.
xi) Use the bubble check method to check for leaks in the
spanning system.
-------
92
c) Calculations
i) Compute the percent leakage by subtracting the response
recorded in step b) x) from the response in step b) v), and
then dividing result by the value recorded in step b) v).
Multiply the result by 100.
d) Acceptance Criteria
i) If the value calculated in step c) i) is equal to or less
than the leakage rate specification (Chapter IV, Section
F.5.), the system leakage indication system is acceptable.
ii) If the spanning system shows significant leaks at fit-
tings, attempt to repair the leaks by tightening or replacing
the fittings. If leaks occur in other locations of the
spanning system, or appear to be the result of system design.
an engineering report must be submitted by the analyzer
manufacturer describing the causes and preventive remedies
for the leak prior final acceptance of the system.
-------
93
G. Operating Environmental Test Procedure
1. Test Procedure
a) Equipment required
i) Candidate instrument
ii) Environmental chamber with temperature capability between
35°F and 110°F, and humidity capability between 10 and 9P
percent relative humidity.
b) Test Conditions
i) 105°F (+ 5%°F) with a relative humidity between 96 and 100
percent (non-condensing).
ii) 40°F (+ 5°F) with a relative humidity between 75 and 80
percent with a 10 mph wind.
iii) 35°F (+ 5°F) with a relative humidity between 15 and 25
percent.
c) Required Tests of each test condition (Test Procedure numbers
are given in parentheses).
i) System Warm-up (H.I.a))
ii) System Leakage (F.5.)
iii) Sample Line Crush Test '(C.2.)
iv) Analyzer Calibration Curve (E.I.)
v) Pressure and Temperature Compensation (E.3. b) and c))
vi) Analyzer Zero and Span Drift (E.4.)
vii) Sample Cell Temperature (as applicable, see E.7. and H.I.)
viii) Analyzer Water Interference (E.8.)
ix) Sample Cell Pressure Variation, Low Flow, and Response
Time (F.I.)
x) HC hang-up (C.4)
xi) If used, automatic zero/span system (H.2. a))
xii) If used, automatic read system (H.2. b) i) 1))
xiii) If used, anti-dilution system
-------
xiv) If used, printer
xv) If used, automatic data collection system
d) Test Sequence
i) The analyzer shall be turned off and allowed to stabilize
at least 3 hours at the test condition prior to beginning the
performance check.
ii) The warm-up test followed by the leak check test shall be
the first two tests performed at each test condition.
iii) The remaining tests may be performed in any convenient
order.
iv) An additional leak check shall be performed at the com-
pletion of the required tests for each test condition.
v) The first test condition shall be the 35°F condition
followed by the 40°F and the 105°F conditions.
e) Acceptance Criteria
i) If the analysis system passes all of the individual test
requirements as specified by each test in c) and d), then the
environmental operating characteristics of the system are
acceptable.
-------
95
H. Fail-Safe System(s)
1. Test Procedures for Features Required for All Systems
a) Warm-up Lock-out Test Procedure
i) Equipment Required
1) Candidate analyzer.
2) The equipment required to perform the Analyzer Sample
Cell Temperature Test Procedure (see Section E.7.).
3) The equipment required to Perform the Analyzer Zero
and Span Drift Test Procedure (see Section E.4.).
4) A timer.
ii) Test Sequence
1) Follow the basic measurement preparations indicated
in Test Procedures E.4. and E.7.
2) After the analyzer has stabilized at the ambient
conditions as determined by Test Procedure E.7. Turn on
the analyzer power and simultaneously start a timer. (A
chart recorder may be used).
3) The chart recorder must indicate both negative and
positive zero drift. If necessary, offset the chart
recorder zero 5 units up scale.
4) As soon as the warm-up lock-out feature deactivates,
record the elapsed time from power on, the sample cell
temperature, and immediately zero and span the analyzer
with analytical gases per manufacturer's operating
instructions. Simultaneously, start the chart recorder
(in already started, mark the chart) at a minimum chart
speed of 0.5 inches per minute.
5) Immediately after spanning, begin sampling ambient
air (or analytical zero gas at the ambient temperature)
through the sample line.
6) Monitor the zero drift and the sample cell temperature
for 5 minutes.
7) Record the lowest sample cell temperature during the
5 minute period.
-------
96
iii) Calculations
1) Locate on the chart the maximum and minimum analyzer
zero-response during the 5 minute period.
2) Compute the difference (as a percentage of full scale
chart deflection) between the analyzer zero-response
determined by the span check in step b) 4) and the
maximum zero-response, and then the minimum zero-response
as identified in step iii) 1).
3) Determine the largest difference in step c) 2) and
multiply it by 2 (zero drift).
iv) Acceptance Criteria
1) If the temperature observed in step ii) 2) is approx-
imately the same as the ambient temperature, the analyzer
is properly stabilized.
2) If the temperatures recorded in step ii) 4) and step
ii) 7) are equal to or greater than the specifications
for sample cell temperature, the sample cell temperature
is acceptable.
3) While monitoring the sample cell temperature during
the analyzer check-out procedure, if the temperature is
less than the specifications when concentration data
(span or sample) is read, the acceptance in step 2) is
void.
4) If the zero drift, as calculated in step iii) 3) is
less than the specifications for zero drift (Chapter IV,
Section E.), then the zero drift after warm up is
acceptable.
5) Acceptance of the above criteria constitutes accep-
tance of the Warm-up Lock-out system.
6) If the analyzer manufacturer indicates a typical
lock-out elapsed time to the ultimate user, then the
manufacturer must show that the elapsed time supplied
to the user is truly typical if the time recorded in
step ii) 4) is more than 30 percent longer than the
time indicated to the user.
-------
97
b) Analyzer Low Flow Test Procedure
i) See Test Procedure F.I., Sample Cell Pressure Variation,
Low Flow, and Response Time.
ii) Visual Observation of Features
c) Analyzer Leak Check Test Procedure
i) See Test Procedure F.5., System Leakage Test Procedure
ii) Visual observation of features
iii) Self explanatory test procedure for timer check-out.
-------
98
2. Test Procedures for Features Required on Decentralized Systems
a) Automatic Zero/Span Check Test Procedure
1) If the analysis system is equipped with the automatic gas
span feature, the gas spanning feature must be used throughout
the entire evaluation procedure unless noted otherwise (e.g.
pressure and temperature compensation, Section E.3. b)).
ii) The automatic span system shall be used with calibration
gas (instead of span gas) for the Calibration Curve Test
Procedure (E.I.).
iii) Self-explanatory Test Procedures for the other require-
ments.
b) Automatic Read Feature Test Procedure
i) If the analysis system is equipped with the automatic
read feature, the system shall be deactivated for all
evaluation testing except:
1) The response time portion of test procedure F.I.
shall be rerun with the automatic read feature opera-
tional.
2) All system correlation tests (Procedures in Section I)
shall be run with the automatic read feature operational.
3) Other tests self explanatory.
c) Printer Feature Test Procedure
If the analysis system is equipped with a printer, the printer
shall be checked for proper operation. The printer shall provide
the official system results to be used in the correlation proce-
dures (Section I).
d) Vehicle Diagnosis Feature Test Procedure
Check for proper operation. No other test procedures are
required.
e) Anti-Tampering Feature Test Procedure
i) Visual observation
-------
99
I. System Correlation Test Procedures
The correlation procedure is designed to evaluate the performance of the I/M
analyzers when sampling auto exhaust. The NDIR procedure is slightly dif-
ferent from the FID HC correlation procedure. However, it is expected the
data collection phase of each procedure can occur simultaneously. For
clarity they will be presented as separate procedures.
1. NDIR Analyzer Correlation Test Procedure
a) Equipment Required
i) Candidate instrument.
ii) A raw exhaust CO analysis system meeting the requirements
of 40CFR 86, Subpart D, for gasoline-fueled engines.
iii) A laboratory grade NDIR HC analyzer substantially simi-
lar in quality to the laboratory CO analyzer, and operated
according to the requirements for general NDIR analysis of
gasoline-fueled engines contained in 40 CFR 86, Subpart D.
iv) A tailpipe extension that meets the probe location re-
quirements in 40CFR 86 Section 312-79 (c) (v).
v) Two test vehicles.
vi) Calibration gases for each range used on both the candi-
date analyzers and the reference analyzer. In this test
procedure, calibration gase.s will be used instead of span
gases to span both analysis systems. The calibration gases
should be between 70 and 90 percent of full scale on each
range used. Calibration gases for the Subpart D analysis
system shall meet or exceed the requirements specified in
Subpart D.
vii) Option: An adjustable dilution box and mixing chamber
may be used to obtain different exhaust concentration levels
from the vehicle. The dilution box and mixing chamber would
be installed between the tailpipe and the analysis system
probes.
viii) Option: A chassis dynamometer would be useful for
loading the vehicle to obtain different emission levels, but
is not required.
b) Test vehicles
i) One 1975 or later non-catalyst light-duty vehicle (LDV),
LOT, or HDG is acceptable.
ii) One 1978 or later oxidation catalyst equipped vehicle
with air injection.
-------
100
ill) All vehicles must be in good operating condition with
all emission control systems functional. A 1975 FTP type
test may be performed on the vehicle to verify functional
operation of the emission controls.
c) Test Procedure Overview
The correlation procedure consists of testing the candidate analy- a^^
zer at several points on each range. The range scale of the ^^^
candidate analyzer determines the approximate test points. The 5JS
exact test points are then determined by the concentration levels *^j^*i
observed by the reference system. The exact test points are then iLt^m
replicated several times (minimum of 6) based on the reference £MB£
system response values. This replicate data is then analyzed to
determine correlation between the candidate system and the refer-
ence system.
d) Test Sequence
i) Select a test vehicle, and warm up that vehicle with at
least 30 minutes of hard load. (freeway operation, hard
accelerations, etc.
ii) Prepare the analysis systems for measurement (i.e. warm-
up, spanning etc.) operate the reference system according to
the provisions in Subpart D where applicable.
iii) Insert the probe of the candidate instrument into the
tailpipe approximately 16 inches.
iv) Operate the vehicle (or adjust the dilution box) to
obtain stable concentration readings at approximately 20, 50,
70 and 90 percent of full scale concentration value on the
low range of the candidate analyzer.
v) Select the lowest useable range on the Subpart D system
for each test point in d)iv) (see § 86.338).
vi) Sample for approximately 1 minute at each test point.
vii) Record the average emission value of the candidate j^^^
instrument and the average chart deflection from the refer- _~Sr~
ence system over the last 10 seconds for each test point. "™^t
The two readings recorded at each test point should be re- CSS
corded over the same time frame, and are defined as a CMMS
"paired" data point.
viii) Select the high range of the candidate instrument.
ix) Operate the vehicle (or adjust the dilution box) to
obtain stable readings at approximately 20, 50, 70, and 90
percent of full scale on the high range of the candidate
-------
mi
instrument. If the higher emission values require impracti-
cal vehicle operation, cease data generation and continue
with the test procedure. At least one vehicle should be able
to achieve the higher values.
x) For the attainable test points in the preceding step,
repeat step d)vii).
xi) The chart deflections recorded in steps d)vii) and d)x)
for the Subpart D system are now defined as "reference set
points" that will be used for the remainder of the correla-
tion test. When repeating the test sequence with the same
vehicle or a different vehicle, the vehicle or dilution box
should be adjusted to obtain, as close as possible, the exact
"reference" chart deflection. For the test points not at-
tainable in step d)ix), the chart deflection of the reference
system becomes the set point the first time a data pair is
recorded for that test point.
xii) Repeat steps d) iv) through d) x) at the set points de-
scribed in step d)xi). Record the emission values from the
candidate system, and the chart deflections from the refer-
ence system.
xiii) Ground or short two plug wires on V8 engines (opposite
sides of intake manifold), or one plug wire for 6 cylinder
and 4 cylinder engines.
xiv) Perform steps d)iv) through d)x) at the set points de-
scribed in step d)xi) a minimum of three times. Record the
emission values from the candidate system, and the chart
deflections from the reference system.
xv) Select the 2nd test vehicle, and repeat steps d)iv)
through d)xiv) at the set points described in step d)xi).
Record the emission values from the candidate system, and the
chart deflections from the reference system.
xvi) Repeat the procedure as necessary to obtain a minimum of
6 replicate responses at each test point.
e) Calculations
i) For the candidate system compute the mean (x) and standard
deviation(s) of the emission values for each test point.
ii) For the reference system convert each chart deflection to
a concentration value from the analyzer calibration curve.
iii) For the reference system compute the mean (x) and stan-
dard deviation(s) concentration values for each test point.
-------
102
iv) Compute the normalized precision difference ( A P) for
each test point by:
A?
where:
((Ks/ x) candidate) - ((Ks/ x) reference)
Sample Size
5
6
7
8
9
JC
3.5
3.1
2.9
2.7
2.6
Sample Size
10
11
12
13
14
2.5
46
40
36
2.31
v) For each range of the candidate analyzer perform a linear
regression on all of the paired data that were measured on
that range. Force the regression through zero. The refer-
ence system is the independent variable. Identify the
slope(m) of the regression line.
vi) For each range of the candidate analyzer compute the
ratio(R) of the analyzer mean concentrations for each test
point corrected for the slope(m) identified in the preceding
step by:
R = (x reference/x candidate)m'
f) Acceptance Criteria
i) Identify the largest A P value. If the largest A? value
is less than or equal to the specification for A P. then the
in-use precision of the candidate system is acceptable.
ii) If the slope (m) for each range of the candidate analyzer
is within the limits for slope, then the slope test results
are acceptable.
iii) Identify the minimum and the maximum ratio(R) of slope
corrected mean concentration values. If the minimum and
maximum ratios are within the range specified, then the mean
concentration ratio test results are acceptable.
-------
103
2. FID Analyzer Correlation Test Procedure
a) Equipment required
i) Candidate instrument.
ii) An HC analysis system meeting the requirements of 40CFR86,
Subpart D for gasoline-fueled engines.
iii) A tailpipe extension that meets the probe location
requirements in 40CFR86 Section 312(c)(v).
iv) Three test vehicles.
v) Calibration gases for each range used by both the candi-
date analyzers and the reference analyzer. In this test
procedure, calibration gases will be used instead of span
gases to span both analysis systems. The gases should be
between 60 and 90 percent of full scale on each range used.
vi) Calibration gases for the Subpart D analysis system shall
meet or exceed the requirements specified in Subpart D.
vii) Option: A dilution system that allows the auto exhaust
to be diluted in a controlled manner with a suitable mixing
chamber, may be used to obtain different exhaust concentra-
tion levels from the vehicle.
viii) Option: A chassis dynamometer may be used with or
without the dilution system to obtain different concentration
levels.
b) Test Vehicles
i) The type of test vehicle required for the FID correlation
test are the same as the type required for the NDIR correla-
tion. Generally it is preferred that the same vehicles be
used for both correlations.
c) Test Procedure Overview
The correlation procedure consists of testing the candidate anal-
yzer at several points on each range. The range scale of the
candidate analyzer determines the approximate test points. The
exact test points are then determined by the concentration levels
observed by the reference system. The exact test points are then
replicated several times (minimum of 6) based on the reference
system response values. This replicate data is then analyzed to
determine correlation between the candidate system and the refer-
ence system.
-------
104
d) Test Sequence
The test sequence is identical to CO correlation test required
with the following exception.
i) Stabilized concentration readings of approximately 10, 20,
35, 50, 70, and 90 percent of full scale concentration values
should be substituted for the value in step l.d)iv) and
l.d)ix). Record the HC concentration value from the candi-
date analyzer as ppmh or ppmC6.
e) Calculations
i) For the candidate system compute the mean (x) and standard
deviation(s) of the emission values for each test point.
ii) For the reference system convert each chart deflection to
a concentration value in ppm C3 (propane) from the analyzer
calibration curve.
iii) For the reference system compute the mean (x) and stan-
dard deviation(s) concentration values for each test point.
iv) Compute the normalized precision difference ( AP) for
each test point by:
A P = ((Ks/ x) candidate) - ((Ks/ x) reference FID)
where:
Sample Size JC Sample Size _K
5 3.5 10 2.5
6 3.1 11 2.46
7 2.9 12 2.40
8 2.7 13 2.36
9 2.6 14 2.31
v) For each range of the candidate analyzer perofrm a linear
regression on all of the paired data that were measured on
that range. Force the regression through zero. The refer-
ence system is the independent variable. Identify the
slope(m) of the regression line.
vi) For each range of the candidate analyzer compute the
ratio(R) of the analyzer mean concentrations for each test
point corrected for the slope(m) identified in the preceeding
step by:
R = (x reference FID/x candidate)m
f) Acceptance Criteria
-------
105
i) For HC analyzers manufactured before January 1986, the
following data shall be available:
1) Precision
2) Slope Comparison
3) Ratio of Modal Averages
ii) For HC analyzers manufacturered after January 1986, the
analyzer shall meet the following criteria:
1) Identify the largest A p value. If the largest /\ P
value is less than or equal to the specification for A
then the in-use precision of the candidate system is
acceptable.
2) If the slope(m) for each range of the candidate
analyzer is within the limits for slope, then the slope
test results are acceptable.
3) Identify the minimum and the maximura ratio(R) of
slope corrected mean concentration values. If the
minimum and maximum ratios are within the range speci-
fied, then the mean concentration ratio test results
are acceptable.
------- |