CbRPO«AT.ON                           EPA Report No. EPA 460/3-82-006
DCN 82-240-016-17-12
                         GUIDANCE ON QUALITY CONTROL

                                    FOR

                      INSPECTION/MAINTENANCE PROGRAMS
                                FINAL REPORT
                               Prepared by:
                            Radian Corporation
                               Prepared for:

                   U.S.  Environmental  Protection Agency
                       Inspection/Maintenance Staff
                            2565 Plymouth Road
                           Ann Arbor,  MI  48105
                               25 June 1982
            8501 Mo-Pac Blvd./P.O. Box 9948/Austin, Texas 78766 / (512)454-4797

-------
                    NOTICE

This report  was  prepared for  the  Environmental
Protection   Agency   by   Radian   Corporation,
Austin,  Texas,  in  fulfillment  of  EPA  Contract
No.  68-02-3513,   Task  Order  No.   17  for  the
purpose  of  assisting  state  and local  agencies
in  implementing  effective   vehicle   emissions
inspection   and   maintenance   programs.    The
contents  of  this report  are  reproduced  herein
as  received   from   Radian   Corporation.    The
opinions,  findings, and  conclusions  expressed
are those of the  author  and  are  not necessarily
those  of the  Environmental  Protection  Agency.
Mention  of company  or product  names  is  not  to
be  considered    as   an  endorsement   by   the
Environmental Protection Agency.

-------
                            TABLE OF CONTENTS
                                                                        Page

1.0       INTRODUCTION	     1
2.0       QUALITY CONTROL FOR EMISSIONS ANALYZERS 	     3
          2.1   Daily Quality Control Checks	     6
          2.2   Weekly Quality Control Checks 	     8
          2.3   Preventive Maintenance	    20
          2.4   Documentation	    22
          2.5   Audits	    25
3.0       QUALITY CONTROL FOR INSPECTION TEST PROCEDURES	    27
          3.1   Quality Control  Procedures Prior to the Test	    27
          3.2   Quality Control  Procedures During the Test	    31
          3.3   Quality Control  Procedures After the Test 	    33
          3.4   Inspection Checklist	    35
4.0       QUALITY CONTROL FOR INSPECTION STATION AUDITING 	    37
          4.1   Checking the Performance of the Equipment	    38
          4.2   Checking Procedures for Routine Tasks 	    40
          4.3   Other Items to Inspect During the Audit 	    45
          4.4   Taking Corrective Action	    45
          4.5   Documentation	    49
5.0       QUALITY ASSURANCE IN AN I/M PROGRAM	    54

-------
                                  LIST OF TABLES
                                                                       Page

1.     Emissions Performance Warranty Emissions Standards	    6
2.     Preventive Maintenance Schedule 	   21
3.     Audit Criteria for Record Checks and Personnel
       Proficiency Checks	47
4.     Quality Assurance Questions for an I/M Program	55
                                  LIST OF FIGURES

1.     Methods of Introducing Span Gas to Analyzer Probe	11
2.     Typical Analyzer Front Panel with Analog Meters 	   13
3.     Sample Analyzer Log Sheet	24
4.     Inspection Checklist	36
5.     Sample Audit Checklist. ... 	   50
                                     m

-------
1.0  INTRODUCTION

Inspection/Maintenance (I/M) programs reduce emissions from vehicles by requir-
ing the repair of those vehicles that fail an emissions test.  In order for
a program to operate-effectively and equitably, it is extremely important
that accurate tests be performed at all inspection stations.  Therefore, the
quality control of the emissions analyzers and inspection procedures play a
major role in the overall success of an I/M program.  Auditing of inspection
stations by state personnel is one means for assuring the quality of emissions
measurements.  However, to maintain the effectiveness and equity needed in the
auditing system, the quality control of the auditing process is equally impor-
tant.

The first step in designing an adequate quality control program for an I/M
program is to develop a quality control plan.  The quality control plan des-
cribes which items need to be maintained or evaluated and at what frequency
the maintenance (including adjustments) or evaluations need to be performed.
To implement the quality control plan, a quality control manual containing
specific quality control procedures must be developed to assure that all
analyzers are routinely checked and maintained in proper working order,
all records are completed and reported accurately, and all prescribed
regulations and procedures are being followed by inspection station personnel.
Without these assurances, I/M program officials cannot ensure that motorists
are receiving fair and accurate inspections.

Quality control in I/M measurements is also important because of the emissions
performance warranty.  The emissions performance warranty allows owners of
1981 and later model year vehicles to obtain warranty repairs if their vehicles
fail an I/M test.  In order for owners to get these repairs, however, the I/M
test must meet certain requirements and a minimum level of quality control
must be provided in the I/M program.  Thus, it is important that the quality
control program meet the requirements established in the emissions performance
warranty regulations, if warranty coverage is to be made available.

-------
This report provides guidance on the areas that need to be addressed
in a quality control plan and on recommended quality control procedures that
can be used in the quality control program.  This guidance is consistent
with the provisions of the emissions performance warranty regulations.  In
addition, the following general assumptions were made in developing this
report.

     1.   The I/M program will use a manually operated emissions analyzer
          which meets the general analyzer specifications of the emissions
          performance warranty regulations.

     2.   The emissions test will consist of a simple idle inspection test
          along with a brief 2500 rpm preconditioning mode for pre-1981 model
          year vehicles.

     3.   A two-speed idle test (where vehicles will be subjected to emissions
          standards for each speed) will be used for 1981 and later model
          year vehicles.

The following section discusses the quality control plan and procedures that
the individual inspection facilities should be required to use to insure that
analyzers maintain accuracy.  Section 3.0 of this report provides quality
control procedures to assure that the test procedures are performed correctly.
Section 4.0 describes quality control procedures that can be used during
auditing by administrating agency personnel to further assure that accurate
inspections are being performed.  Section 5.0 discusses the importance and
role of the quality control elements in providing overall quality assurance
in the I/M program.

-------
2.0  QUALITY CONTROL FOR EMISSIONS ANALYZERS

The nondispersive infrared (NDIR) analyzer is the primary inspection instru-
ment used  in I/M programs.  The  NDIR analyzer consists of a gas analyzer
which produces an electrical signal in response to exhaust gas con-
centrations, and a meter display (analog or digital)  that translates the
electrical signal  for the user.  The emissions analyzer is used to determine
the concentrations of carbon monoxide (CO) and hydrocarbons (HC)  in vehicle
exhaust.  In an I/M program, emissions analyzer accuracy must be  maintained
to assure that vehicles are properly passed or failed during inspections.
Therefore, periodic checks of a NDIR analyzer are needed to verify its
accuracy.

Several  different checks can be performed to verify the accuracy of a NDIR
analyzer as noted below.  The first two checks verify that the meter displays
correctly portray the electrical output from the gas  analyzer.  The third and
fifth checks verify that the gas analyzer is generating appropriate electrical
output in response to exhaust gas concentrations.  The purposes of the other
checks are described below.

     1.    Mechanical zero check of analog meter -- If the analyzer is equipped
          with analog meters, a check can be made to assure that the meter
          needles are aligned with the zero readings before the instrument
          is turned on.  If necessary, adjustments are usually made with a
          setscrew on the face of each meter.  See the diagram in Figure 2
          on page 13.

     2.    Electrical span -- On analyzers with analog meters, it is possible
          to perform an electrical span to make sure that the meter needles
          align with the span lines on the meters when the analyzer is in the
          "span" mode.  For analyzers with digital meters, when the analyzer
          is placed in  "span" mode, the analyzer should respond with predeter-
          mined meter readings.  If necessary, adjustments are made by turning
          the HC and CO span adjustment potentiometers.  On some analyzers
          electrical span adjustments are automatically performed as needed.

-------
3.   Zero check through the port -- This check is performed by putting
     the analyzer in "zero" mode to assure that the analyzer does not
     indicate concentrations of HC or CO when room air or zero gas is
     being sampled.  If necessary, adjustments are made by turning the
     zero adjustment potentiometer on each meter.  On some analyzers
     zero adjustments are automatically performed as needed.

4.   Hydrocarbon hang-up check -- This check is very similar to the zero
     check and it is performed to assure that the sampling system (i.e.,
     the probe, filters, and sampling lines) is not a source of hydro-
     carbon readings.  Hydrocarbon hang-up problems can usually be
     alleviated by purging the sampling system with clean dry air,
     changing filters, or cleaning the sample hose.

5.   Gas spanning -- Gas span checks consist of introducing a known con-
     centration of span gas into the analyzer and determining whether the
     analyzer response to the span gas is correct.  If the analyzer's
     response is outside the specified limit, adjustments need to be
     made to the analyzer so that the readings obtained are accurate.

6.   Low scale/high scale correlation check -- This check assures that
     the analyzer's response to a span gas is the same on both the low
     and high scales (if applicable).  Most analyzers have setscrews
     that can be used to adjust the low scale/high scale correlation,
     if necessary.

7.   Leak check -- This check is performed to assure that there are no
     excessive leaks in the sampling system of the analyzer.   Leaks cause
     sample dilution resulting in potentially large measurement errors
     and are one of the most common sources of error in I/M measurements.

-------
In addition to the above checks, some simple, common sense quality control
checks need to be included in the quality control manual.  These checks
should include such things as a reminder to periodically check the water
trap and sample filters and the importance of properly warming up the
analyzer before testing.

The emissions performance warranty regulations contain certain minimum quality
control  requirements for emissions analyzers.  In order for a vehicle owner
to receive warranty repair, the following quality control provisions must be
incorporated into the I/M program.

     1.    Within one hour prior to a test, the analyzer must have been
          zeroed and electrically spanned.

     2.    Within one week of a test the analyzer shall have been checked
          with a span gas that is traceable to NBS standards (±2%).   The
          span gas used shall have concentrations either between the
          standards specified by the warranty regulations (see Table 1)
          and the local jurisdiction's inspection standards for 1981 or
          later model year vehicles, or be within -50% to +100% of the
          standards shown in Table 1.

     3.    Within one week of a test, a leak check must have been made.
          If the analyzer has a separate calibration/span port, the CO
          readings using span gas through the probe and through the port
          shall be made.  Differences of over 3% between the port and
          probe readings shall require the repair of leaks.  (The
          warranty regulations do not expressly identify how to leak-
          check analyzers that do not have separate span ports.)

-------
        Table 1.  Emissions Performance Warranty Emissions Standards
     Test and Mode                         ,HC\           ,  C0  ^
                                           (ppm)           (percent)

 1.  Idle:Idle                              220               1.2
 2.  2-speed idle:
       2500 rpm                             200               1.0
       Idle2                                200               1.0
 3.  Two mode loaded:
       30 mph loaded                        220               1.2
       Idle                                 220               1.2

1 ppm as hexane.
2 Lowest individual  HC and CO readings observed at the two idle modes.

By incorporating the quality control requirements of the emissions performance
warranty regulations along with other provisions that are necessary for good
quality control, it is possible to develop a standard quality control  program
for emissions analyzers.  This program should address daily quality control
items, weekly quality control items, preventive maintenance,and audits.

2.1  Daily Quality Control Checks

Certain analyzer functions need to be checked daily to assure accuracy of the
emission measurements.  These checks are usually an integral  part of the
inspection procedure; therefore, the recommended procedures to be followed are
contained in Section 3.0 of this report.  However, these daily checks  will be
described briefly in this section.  They mainly cover the following areas:
zero checks, electrical  spans, filter and water trap checks, hydrocarbon
hang-up checks, and adequate warm-up of the analyzer.

Zero checks and electrical span checks are conducted in order to identify
and eliminate electronic errors of zero and span drift, respectively,  in the

-------
analyzer.  Most analyzers are designed to drift less than 2 to 3 percent
per hour for both zero and span.  Therefore, zero and span drift usually
represent only small errors in the emissions measurement.  Even though
these errors are small, they are additive to other errors in the measurement
system.  Moreover, they can be easily eliminated by simple adjustments
which take only seconds.

Filter and water trap checks are conducted as additional safeguards against
damage to the analyzer.  These devices are designed to collect particulates
and water that are in the exhaust sample to keep them out of the sample cell
where they cause soiling, possible damage and improper readings.  To make
sure that these devices stay in proper operating condition, the operators
should visually inspect them regularly.

Hydrocarbon hang-up checks are performed to make sure that the analyzer's
hydrocarbon readings are not being biased by hydrocarbon adsorption and de-
adsorption within the sampling system.  In a decentralized I/M program the
analyzer is not only used for emissions inspections but also for diagnosing
and repairing failed vehicles and for tune-up adjustments.  In these repair
modes, the analyzer may be subjected to extended exposure to potentially
high concentrations of hydrocarbon emissions, thus increasing the likelihood
of a hydrocarbon hang-up problem.  If hydrocarbon hang-up is allowed to go
unchecked and uncorrected, related errors of 10 to 30 percent are typical.
On the other hand, if a hydrocarbon hang-up check is used routinely, and
any detected problems are corrected,  measurement errors due to hydrocarbon
hang-up can be virtually eliminated.

Adequately warming up the analyzer before use is critical to obtaining
accurate emissions measurements.  If an analyzer is not fully warmed up,
the-analyzer electronic components are unstable, causing potentially severe
zero and span drift and erratic readings.  In addition, an analyzer that  is
not fully warmed up may not be very sensitive to changes in pollutant
concentrations.  Because of these influences, measurements made with an

-------
analyzer that is not adequately warmed up could have very large errors.
Available data indicate that these influences can cause errors from 0 to 200
percent in individual cases.  Since analyzer warm-up can have such a major
influence on the accuracy of emissions measurements, the importance of
adequate analyzer warm-up should be stressed in the quality control manual.

2.2  Weekly Quality Control Checks

The weekly quality control checks consist mainly of spanning the analyzer
with a span gas of known concentration (preferably the span gas should
meet the requirements in the emissions performance warranty regulations).  In
addition, a leak .check should be performed once a week.  The procedures for
recommended weekly quality control checks are presented later in this sec-
tion.

Gas spanning needs to be conducted periodically to identify and correct
errors caused by shifts in the sensitivity of the instrument.  These shifts
occur mainly because of changes in environmental conditions, such as
barometric pressure and ambient temperature.  However, internal changes in
the analyzer can also affect the analyzer's accuracy.   These influences do
not usually cause abrupt changes in analyzer accuracy; they cause gradual
change from day to day.  From week to week, however, the total change can be
significant (greater than +5 percent error).  Therefore, weekly gas spanning
is recommended as a precaution (and required for compliance with the emission
performance warranty regulations).

Leaks in the sampling system are probably the source of the largest and most
frequent errors that occur in emissions measurements.   This is because a
leak is transformed directly into a measurement error (i.e., a 15 percent
leak results in a 15 percent error).   Because analyzers are moved around in
the inspection stations and the sampling hose and probe have a high
potential for damage, the potential for system leaks is high.  Therefore,
leak checks are recommended as a weekly quality control check.  Leak checks

-------
should also be performed whenever analyzer maintenance (such as, filter re-
placement or hose replacement) is performed that could cause a leak.

A combination gas span/leak check can be performed in a few minutes.  Properly
done, these checks ensure that the accuracy of the analyzer is maintained
within an acceptable range of +5 percent.

2.2.1     Procedures for Weekly Quality Control Checks

As mentioned, the weekly quality control checks consist primarily of low
scale gas spanning and leak checking.

2.2.1.1   Equipment Required

1.   Span gas and related equipment -- In order to perform gas spanning, it
     is necessary for the station to have a supply of span gas.  For efficiency
     it is recommended that the station purchase a full-size (I-A) cylinder
     that contains a triblend of carbon monoxide (CO) and hydrocarbons (HC)
     (as propane) in nitrogen.  The concentrations in the cylinder should be
     approximately 1.5% CO and 600 ppm propane.  This gas should be traceable
     to +2% of NBS standards.  These concentrations are used to assure that
     the analyzer is accurate near the emissions standards specified, in the
     emissions performance warranty regulations.  It should be noted that
     most analyzer manufacturers recommend gas spanning on the high scale
     (usually 80 to 90 percent of full scale on the high range).  Gas
     spanning on the low range, as recommended here, does not decrease the
     overall accuracy of the analyzer; it just assures that the accuracy is
     obtained on the low range where most vehicles are passed and failed in
     an I/M program.  An important issue regarding span gas is the accuracy
     (or traceability) of the span gas.  It is beyond the scope of this
     report to cover the details of this issue.  However, EPA and the gas
     industry are jointly investigating the current practices of gas manu-
     facturers in naming span gases in order to determine the proper naming
     procedures.  This will be the subject of a future EPA report.

-------
2.   Pressure regulators -- The span gas cylinder must have a pressure
     regulator attached to it.  This regulator should be set to the pressure
     that is specified in the analyzer operating manual.  In the absence of
     a specification, the pressure should be set at 5 to 6 psi (gauge
     pressure).  Downstream of the pressure regulator there should also be
     another valve to regulate the flow of span gas.  In addition, some manu-
     facturers, recommend that a restricting orifice be placed in the line to
     limit the maximum flow rate.

3.   Manifold and equipment to introduce span gas into analyzer probe --
     Several different designs for a manifold can be used to introduce the
     span gas into the analyzer probe.  The main point behind each method is
     to assure that the analyzer is not being overpressurized or starved of
     span gas while introducing the span gas through the probe.   Following
     are descriptions of different methods that can be used.

Method A -- With this method, a tee connection is placed into the line that
connects the span gas cylinder to the analyzer.  A balloon is placed over
the open end of the tee (see Figure 1).  When the gas is introduced to the
analyzer, the flow valve is adjusted such that the balloon is inflated and
then closed just enough to allow the balloon to maintain a constant inflation
during spanning.  This assures that the analyzer is obtaining adequate
quantities of span gas while, at the same time, not being overpressurized.  A
screen in the tee can be used to prevent the balloon from being sucked into
it in the event that the span gas flow is too low.

Method B — With this method, a rotometer is used instead of the balloon in
Method A.  In this case, the flow valve is adjusted until the rotometer indi-
cates a slight amount of flow through the rotometer.  This method should be
used only in well ventilated areas in order to guard against CO poisoning.

Method C -- In this method, which was developed by EPA, an "audit box" is
placed in the span gas line.  This box contains a vacuum gauge that indicates
when the sample line is under vacuum.  When span gas is introduced into the
                                     10

-------
analyzer,  there should  be  close to zero vacuum in the lines.  Thus the flow
valve is adjusted  until  the audit box indicates that there is zero vacuum
in the line.   More details on  the audit box are available from EPA.

        Figure 1.   Methods of  Introducing Span Gas to Analyzer Probe
           Pressure
           Regulator
           n-cc—A
        Span Gas
        Cylinder
                                            Balloon
            Flow to Analyzer Probe
Flow Control Valve
                               Method A
 Method 8 -  Rotometer  instead of balloon.
 Method C -  "Audit box" (or vacuum gauge)  instead of balloon.
                                     11

-------
4.   Connectors to span port and probes -- Inspection stations which have
     analyzers with separate span ports will  need connectors that will allow
     them to introduce the span gas into the port and probe.  If the analyzer
     does not have a separate span port, then the station will need only a
     connector for the probe.  Generally most analyzers have serrated male
     nipples for the span ports; thus the only equipment that is needed is a
     piece of properly sized, flexible tubing attached to the manifold to
     get a leak-tight fit.  Alternatively, "quick-connect" couplings may
     be used, but they must be kept in good repair.   In order to introduce
     span gas into the probe, it is necessary that the connector be leak-
     tight.  Again, some form of flexible tubing with a hose clamp should
     provide a leak-tight connection.

2.2.1.2   Procedures to Conduct a Combination Gas Span/Leak Check

A combination gas span/leak check provides an easy and efficient way of
evaluating an analyzer's ability to properly respond to a known span gas
concentration and, at the same time, to assure that there are no excessive
leaks in the sampling system.  The procedures listed below should be utilized
to perform a combination gas span/leak check.

A typical control panel and analog meter display is  shown in Figure 2.  The
actual  location of various adjustment mechanisms varies considerably among
different makes and models of analyzers.  Therefore, the analyzer manual
should be consulted to determine the exact location  of the applicable
adjustment mechanisms.

Gas spanning is affected, to an extent, by the room  temperature of the
station.  Therefore, it is recommended that gas spanning be conducted at a
time during the day when the room temperature is at  a normal level.

Note:  For analyzers with analog meters, care must be taken in setting the
meter or taking readings in order to avoid errors due to parallax.  Analog
                                     12

-------
HC Ze
Po

Figure 2: TYPICAL ANALYZER FRONT PANEL WITH ANALOG M
HC Meter
(
HC High Scale-
HC Low Scale
Mechanical
Zero Setscre
HC Scale
selector
o Adjustment
entiometer

.. — •
i


V
\
\
--- ,0o lT-J5°o ,
>-^ -^-v^f,
°y\o<> ,3ii3L c%>
\s \ ^>- — <•/ O ^*
-*s ^s>
ZERO SET SPAN SET

HYDROCARBON
rt)
w
pHYOROCARBON-i r— MODE (
1 HI LO ' 'OFF TEST
DD DL
ZERO SPAH
O/^\
y EXHAUST
_L
HC Span Adjustment
rotentiometer


IETERS
CO Meter
— 	 	 ... —

< J— LI - -
>>>>^""^
0 / $ I0
V / • 1 1
^ ^>-J — J-.
"v^^^
ZERO SET

"*-
^
SPAN SET
CARBON -MONOXIDE
 	 ___
:ONTROL 	 1 rCARBC
rZERO SPAN1 '
1DD
Zf
ANALYZER. C


N MONOXIDE,
Hi LO '
DD — •
:RO SPAN
) 0-

CO Zero Ad Dustmen.
^
- r
c
. M<
z<
MM^I^
"•""—
J
) Hieh Sralf
I Low Scale
chanical
ro Setscrew
r^-_
CO Scale
Selector
__CO Span Adjustm
Potentiomete

Potentiometer
Mode Selection
Buttons
nt


-------
meters are properly read from directly in front of the meter.  Standing off
to one side or above the meter when settings are made or readings are taken
can introduce errors because of an improper line of sight.

                 Combination Gas Span/Leak Check Procedures
                         (from analyzer "off" mode)

1.   If the analyzer has analog meters, then prior to turning on the analyzer
     the mechanical zero should be checked.  If necessary,  the mechanical
     zero should be adjusted by means of the setscrew on the face of each
     meter.  (Note that this setscrew is not the zero adjustment potentiometer.)

2.   If the analyzer has analog meters, static electricity  can build up on
     the meters and affect their operation.  The meters can be checked for
     static electricity build-up by the operator simply touching each meter
     with his/her hand while the analyzer is warming up.  If excessive
     static electricity is present, the meter needle will waver.  If static
     electricity is present, it can be removed by spraying  the meter face
     with a static removal spray and wiping off the meter face with a clean
     cloth.  Alternatively, the meter face can usually be freed of static
     electricity by wiping it with a clean, damp cloth.

3.   The analyzer must be warmed up as required by the manufacturer (or 30
     minutes in absence of manufacturer's recommendation).   The analyzer
     should be warmed up in the standby mode (unless the manufacturer
     recommends otherwise) to avoid unnecessary operation of the analyzer
     pump.

4.   A visual  check should be made of the water trap and filters to make sure
     that drain lines are not clogged and/or the filters are not dirty.
     Repairs should be made if necessary.
                                      14

-------
5.   After the analyzer is warmed up, turn on the analyzer pump and check
     the zero setting by activating the zero switch (if analyzer is so
     equipped) on the face of the analyzer.  If necessary, the HC and CO
     meters can be adjusted to zero using the zero adjustment potentiometer
     on each meter (see note below).  Zero checks and adjustments should
     be made on the low scale unless the manufacturer recommends otherwise.

6.   Check electrical span by activating the electrical span switch.
     Electrical spanning and adjustments must be made on the scale specified
     by the manufacturer in the analyzer manual.  The needle should align
     with the span line on the meter.  If necessary, adjustments can  be made
     by turning the HC and/or CO span adjustment potentiometers.

7.   The zero and electrical span should be alternately checked until both
     adjustments are correct.  Some analyzers have automatic zeroing  and/or
     electrical spanning systems that automatically adjust these parameters.

Note:  If the meters will not zero by adjusting the zero adjustment poten-
tiometer, do not use the mechanical zero setscrew to zero the meters.  Call
a qualified service technician or the instrument manufacturer to have the
necessary repairs made.

Note:  Altitude adjustment -- Some analyzers have an expanded span setting
with different span lines corresponding to different altitudes.  If the
analyzer is so equipped, the altitude of the locale in question should be
a consideration in setting or checking the electrical span.   The manufacturer's
recommended procedure should be followed.

8.   Perform HC hang-up check.  With the analyzer on the low range (analog
     meters), allow the probe to sample zero or room air for 30 seconds.  If
     the hydrocarbon reading is greater than 20 ppm after 30 seconds, then
     the system should be allowed to purge itself by sampling room air.  If
     after 2 minutes of purging, the HC reading is still greater than 20 ppm
                                     15

-------
     HC, the probe and line should be disconnected from the analyzer and
     cleaned (usually by blowing clean, dry compressed air through it).   In
     some cases, it may be necessary to change the filter to alleviate the HC
     hang-up problem.

9.   After making sure the analyzer pump is turned on, connect the span  gas
     cylinder to the probe using one of the recommended configurations.   (See
     Section 2.2.1.1)  Slowly open the span gas cylinder valve.   Make sure
     there is proper flow by observing the rotometer ball  position, balloon,
     or vacuum gauge.  Allow the analyzer to sample the span gas for 30
     seconds.  Record the meter reading for carbon monoxide (CO).   Then,
     immediately close the cylinder valve.

10.   Introduce span gas through the calibration/span port (if the  analyzer is
     so equipped—see note below)  as follows.

     a.   Connect the hose from the span gas cylinder to the span  port in  a
          leak-tight manner.

     b.   Turn off the analyzer pump (unless the manufacturer recommends
          otherwise).

     c.   Set the control  switch on the analyzer to calibrate (if  so equipped)

     d.   Open the span gas valve  to introduce the span gas into the analyzer.
          The meter response should be almost instantaneous.  As soon as the
          meter responses peak (approximately 5-10 seconds), record the  meter
          responses for both HC and CO.   Then,  immediately turn  off the  span
          gas valve.   The  CO meter response should be within +5% of the  span
          gas labeled CO value.  The HC  meter  response should  be within  +5%
          of the value obtained by multiplying  the span gas labeled propane
          value times the  analyzer's hexane/propane equivalency  factor.  If
          the CO and/or HC responses are not within +5%,  adjustments  are
          necessary.

                                     16

-------
Note:  If the analyzer does not have a span port, the span adjustment must
be checked while span gas is being introduced through the probe in step 9
of these procedures.  In this case, however, both CO and HC readings would
be recorded during step 9.  Analyzers which do not have span ports cannot
be leak checked using the method in these procedures (the flowing span gas
leak check).  Instead the leak checking method recommended in the analyzer
service manual should be used to make sure that no gross leaks are present
in the sampling system.  Some small leaks, which would ordinarily be detected
using the flowing span gas leak check, may not be detected by the alternative
methods recommended in the service manual (these alternatives include mostly
vacuum based checks, such as vacuum decay checks, maximum vacuum checks, or
low flow indicator checks).  However, checking and adjusting the gas span
through the probe will compensate for any errors caused by such small leaks.

11.  Determine the percent difference between the CO response obtained through
     the span port and the CO response obtained through the sampling probe as
     follows:

          % Difference = Response PO - Response PR   X 100
                              Response PO

          Where:

          Response PR = CO response with span gas sampled through probe.

          Response PO = CO response with span gas sampled through port.

     The percent difference must not exceed 3% to comply with the emissions
     performance warranty regulations.  If it does, the leak should be
     located and fixed before further vehicle testing is conducted with the
     analyzer.

 12.  Adjust span, if necessary, as described below.
                                      17

-------
Note:  Span adjustments should be made with the span gas being introduced
through the span port.  If the analyzer has no span port, then the span gas
must be introduced through the probe while span adjustments are made.  Steps a
and b should be performed while span gas is flowing through the sample cell
(pump off, if introducing span gas through the port; pump on, if introducing
span gas through the probe).  When the adjustments are completed, the span gas
cylinder valve should be immediately closed in order to limit the amount of
CO being released into the room as well as to conserve span gas.

     a.    While span gas is  flowing through the sample cell,  using the HC
          span adjustment potentiometer, adjust the HC meter  response to
          read the correct value (cylinder propane concentration  times the
          hexane/propane equivalency factor).   (The HC span adjustment
          potentiometer is sometimes located on the back of the analyzer
          rather than as shown in Figure 2.)

     b.    While span gas is  flowing through the sample cell,  using the CO
          span adjustment potentiometer, adjust the CO meter  to read  the
          correct value (cylinder CO concentration).   (The  CO span adjustment
          potentiometer is sometimes located on the back of the analyzer
          rather than as shown in Figure 2.)

     c.    Turn off the span  gas cylinder.

     d.    Purge the sample cell by turning on the pump and  allowing the
          analyzer to sample room air for a few seconds.

     e.    Put the analyzer in "zero" mode (pump on) and recheck zero  on each
          meter.  If necessary, adjust meters to zero with  zero adjustment
          potentiometers.  With the pump off (unless  spanning through the
          probe), reintroduce the span gas.   The meter responses  should be
          within +2% of the  readings in steps  a and b.   If  not,  repeat steps a,
          b,  c,  d and e until  the meter readings are  repeatable within +2%.
                                     18

-------
     f.   Repeat steps c and d; then activate the electrical span and adjust
          the electrical  span setscrew (usually internal) on each meter so
          that the meter needles align with the span line (analog meters)
          or the meters obtain the appropriate value (digital meters).

Note:  The electrical span setscrew is not the span adjustment potentiometer.

Note:  Step f is a most critical step in gas spanning the analyzer.   The
electrical span setscrew sets the electrical reference point vis-a-vis the
span gas for the analyzer bench.  Failure to make this final adjustment
correctly would defeat the purpose of gas spanning.  The electrical  span
setscrew should never be adjusted except in conjunction with the gas spanning
procedure.

13.  While gas spanning,  perform a low scale/high scale correlation check
     (analog meters only).  The low scale/high scale correlation check is
     performed by reading the analyzer response during gas spanning on both
     the low and high scales and comparing the two readings.  The readings
     should be equal.  If they are not, an adjustment should be made if
     possible.  There is usually an internal setscrew to adjust the scale
     correlation.

Note:  In order to comply with the emissions performance warranty regulations,
it is necessary that the analyzer be checked for leaks (and gas spanned) at
least once a week.  In addition, leak checks should be performed after all
maintenance that could create a leak in the sampling system (for example,
after filter replacement).

2.2.1.3   Quality Control Checks for Analyzers with Automatic Quality Control
          Systems

Some analyzers, particularly those which meet the EPA computerized analyzer
specifications, have automatic quality control systems.  These analyzers have
                                     19

-------
automatic gas spanning, leak checking, and hydrocarbon hang-up checking
capabilities as well as other automatic functions, such as automatic testing
and automatic data collection.   When analyzers with these capabilities are
used, the inspection station and the auditor are, to an extent, relieved of
their responsibilities to perform the routine analyzer quality control  checks.
However, these automatic systems are not totally foolproof.  The analyzer
manufacturer should be required to provide information to the administrating
agency on how the auditors can assure that the automatic systems are operating
properly.  In addition, the inspection station should be required to have
certified service technicians periodically service the analyzer to assure
that it is maintained in proper operating condition.

2.3  Preventive Maintenance

Another very important part of the quality control for emission analyzers is
preventive maintenance.  As defined in the EPA Quality Assurance Handbook
for Air Pollution Measurement Systems, Volume 1, preventive maintenance is an
orderly program of equipment cleaning, lubrication, reconditioning, adjusting
and/or testing in order to prevent a failure of a system or parts thereof
during use.   Most analyzer operating and/or service manuals include specific
preventive maintenance schedules for particular makes and models of analyzers
which should be followed if available.  A schedule for preventive maintenance
which can be used when the manufacturer's recommended schedule is not
available is described in Table 2.  Table 2 is broken down by weekly and
monthly tasks.  It should be noted that some of the items listed in the
table should be replaced as needed.  This is determined by the inspection
that is performed at periodic intervals.  The inspector must review the
operating manual  for the analyzer to determine if other preventive maintenance
is needed.

Emissions analyzers are generally very durable and usually need minimal
maintenance, if properly used and given proper care.  The most frequent
maintenance items are the sample filter and water trap.  Failure to properly
                                     20

-------
service the analyzer will lead  not only to  inaccurate  emissions  measurements
but also to premature failure of the analyzer.  The frequency of replacement
or cleaning of parts is directly affected by the outside environment.

              TABLE 2.   PREVENTIVE MAINTENANCE  SCHEDULE
 Equipment
   Part
Maintenance
 CO, HC Analyzer
 CO, HC  Analyzer
 CO, HC  Analyzer
 CO, HC  Analyzer
 Pressure  Regulator
Weekly  Items

Sample  Filter  .    Inspect (Replace as
                    needed)
Whole  System
Leak  check (also leak
check after any
analyzer maintenance)
Monthly Items

Sample Lines


Probe

Line
Remove  and Blow-out,
Replace as needed

Inspect and Clean

Leak-check with soap
solution
Note:  The water trap should  be inspected  at least daily for water build-up.
This  is especially true for those analyzers which do  not have a  continuous
drain  line.  But even if drain lines are used, the associated check valve
can become plugged with exhaust deposits.
                                  21

-------
By observing certain operating practices, inspection station personnel can
extend some of the variable maintenance intervals.  For instance, filter
life can be extended if the inspector will remember the following.

     1.   Do not start or warm up a vehicle with the probe inserted into
          the tailpipe.  (Each time a vehicle is started loose deposits are
          blown out the tailpipe.)

     2.   Purge the analyzer sampling system after use by allowing the
          analyzer to sample fresh air away from the tailpipe.

     3.  Remove the analyzer probe as soon as possible after the emissions
         readings are taken.

These same three operating practices will also keep the sampling hose cleaner
and reduce the likelihood of hydrocarbon hang-up.  Some other common sense
operating practices are listed below.

     1.   Do not drive over the sampling hose.

     2.   Do not unnecessarily expose the sampling hose to exhaust heat.

     3.   Always coil  and properly store the sampling hose and probe after use.

     4.   Keep the analyzer in a location that is relatively free of tempera-
          ture changes and drafts.

     5.   Do not store materials (such as manuals or tools)  on top of the
          analyzer.

     6.   Keep the analyzer as free as possible of dirt and grease.

2.4  Documentation

Another very important function is to document all of the weekly quality
control checks, maintenance,and audits that are performed on the analyzer.
                                      22

-------
An instrument log should be kept that documents the occurrence and results of
these activities.  This logbook should provide a continuous chronological
record of the condition of the analyzer.  Ideally, the logbook should be
physically attached to the analyzer in order to prevent its loss as well
as to serve as a reminder to the inspectors and auditors to make the appro-
priate entries.

The responsibility for making analyzer quality control checks and for ensuring
proper analyzer maintenance rests with the inspection station personnel.  To
insure compliance in frequency and procedure, as well as to identify trends
in analyzer operation, it is important that major analyzer quality control
checks (such as gas spans and leak checks), analyzer maintenance, and analyzer
audits be recorded in the analyzer logbook.  Such information is not only
useful to ensure that required quality control checks and maintenance are
performed, but also provides the administrating agency with a way to evaluate
the relative in-field performance by different makes and models of analyzers.

To minimize the time required to maintain analyzer logbooks, the log sheets
should be designed to be as simple as possible.  If this is done, only a few
minutes a week would be required to keep these analyzer records.

Figure 3 is an example of a log sheet that can be used to record the quality
control checks and maintenance that are performed.  All weekly quality control
checks and any maintenance that is performed should be recorded on this log
sheet.  The following gneeral  instructions should be followed when making
an entry on the log sheet:

     1.   The specific procedure should be clearly identified.  At a minimum,
          the following procedures should be recorded.

          a.   Leak check
          b.   Gas span check
          c.   Filter replacement
                                     23

-------
                    Figure 3.  Sample Analyzer Log Sheet
Site:     5uJ*nson's   Gutf
Analy
Model
Hexan
Equiv
Span
Co
Type
zer Manufacturer, -.
and Serial No.: HLUn) rr>ooeL & J339O , S££/4L # fl£ I+3W
e/propane
alency Factor: O- 6<3
Gas Cyli
nder No. : X40J
ncentration: /6? % CO &3£ ppm Propane
of zero
DATE



4-/1. to
2-43. ?o
air: *>oo*> A/4.

PROCEDURE
9^r/4"*
o>ed<
V** P,oWsr
"•JCtf-
Autrr
S^)AAJ 9 AS
RESULTS/COMMENTS
Co />«00fi - /.64%
Co fb*r . /.67%
/« MAS - oK C« S^AA; -OK
WC spA/v; • OK.
Co P/«06e - /. t>8%
H£ P»AT - 3SOpfm
J\OAKS - OK ^0 Sf/»A>- OK
We. SP/*A» - Qonff"
Co PoAT • 1-69%
Co PM8C - /• 67%
1^/4 QtoA^ • 4 ?X) J^A ^^
rT W rV^t 1 w W t/ 00 /•»
A
-------
          d.   Sample line maintenance
          e.   Sample line replacement
          f.   Probe maintenance
          g.   Maintenance to pressure regulator

     2.   The date should be indicated for each procedure.

     3.   The results should be specific.  If the system is okay, state so.
          If a defect or discrepancy is found, state the nature and degree
          of the defect, and the corrective action taken.

     4.   Operator or other person performing the procedure should be indi-
          cated on the form (signature or initials).
2.5  Audits
If performed correctly, the weekly quality control checks that are performed
by the inspection station personnel should be adequate, in most cases, to
maintain the accuracy of the analyzer.  Nevertheless, it is still important
for personnel from the administrating agency to periodically visit (audit)
the stations and independently verify the accuracy of the analyzer.   Recom-
mended procedures to use in conducting the audits are presented in Section
4.0 of this report.

Checking the analyzer and analyzer logbook are very important aspects of the
audit.  Checking the analyzer and the logbook allows the auditor to assure
that the analyzer is being properly maintained in good operating condition.
Audits by the administrating agency can also satisfy several other purposes,
such as, reviewing inspection records and verifying that the station complies
with applicable licensing requirements.

Internal audits conducted periodically by the shop foreman, service manager,
or another appropriate supervisor within the larger inspection stations are
                                     25

-------
useful to ensure analyzer accuracy, inspector proficiency, and data quality.
Internal audits enhance the quality in the I/M program by improving inspector
accountability.  In addition, stations which routinely conduct internal audits
will undoubtedly have fewer problems passing audits by the administrating
agency.
                                     26

-------
3.0       QUALITY CONTROL FOR INSPECTION TEST PROCEDURES

The previous section described methods to assure that the analyzer is
accurately measuring the emissions from a vehicle.  However, a key aspect in
the overall accuracy of the emissions measurement is the inspection test
procedures and how well they are performed.  Quality control is important in
every aspect of the emissions inspection.

There are three basic parts of the inspection process.  Prior to the actual
inspection, the inspector must go through pre-test procedures to prepare the
analyzer and the vehicle for the inspection.  The next step is the inspection
itself.  Following the inspection, certain post-test activities, including
the documentation of the inspection results, must be conducted.  It is very
important that the documentation be performed carefully, since the data
collected in an I/M program play an important part in the evaluation of
program effectiveness.

This section provides specific quality control procedures to integrate into
the three parts of the emissions inspection process.  These procedures should
be emphasized in training programs that are conducted by the administrating
agency prior to the start and during the operation of an I/M program.  In
addition, the administrating agency should stress these procedures when it
is supervising the emissions inspectors, especially in those stations which
have apparent problems with quality control.

3.1       Quality Control Procedures Prior to the Test

The pretest procedures consist primarily of the preparation of the analyzer
for the inspection along with the preparation of the vehicle.

3.1.1     Procedures for Analyzer Preparation

Prior to performing an emissions inspection, the analyzer must be prepared
for testing.  The following procedures are recommended for analyzer preparation,
                                     27

-------
Note:  See the reference to errors caused by parallax on page 12.

1.     Before the analyzer is turned on,  perform a mechanical  zero check
       (analog meters).   See Section 2.2.1.2,  Step 1.   If the meter needle
       is not aligned with the zero reading, it is necessary to  adjust  the
       setscrew on the face of the meter.

2.     Before inspections may be performed,  the analyzer must be adequately
       warmed up.  While the analyzer is  warming up,  the meters  (analog)
       should be checked for excessive static  build-up.   See Section 2.2.1.2,
       Step 2.  Several  methods can be used  to insure that the analyzer is
       warmed up.

       a.    It is recommended to turn the analyzer on when the  first
             person arrives at the inspection  station.   Having the analyzer
             ready for testing as early as possible in the work  day may reduce
             customer inconvenience by reducing waiting  time.

        b.    The analyzer should be on for a minimum  of 30 minutes prior
             to testing, unless the manufacturer recommends a shorter
             period.   The analyzer should be in the standby mode
             (pump off)  during warm-up in order to conserve the  pump,
             conserve electricity, and reduce  noise.   If there are
             doubts whether the analyzer  is  warmed up, the operator
             could check it with span gas.  If the meter reading
             increases slowly, the analyzer is not warmed up.  Or, if
             an unstable reading is obtained,  the analyzer may not
             be warmed up.

       c.    Some analyzers, in particular those meeting the EPA
             computerized analyzer specifications, will  not sample
             unless they are warmed up.  Thus, if the station has one
             of these analyzers, other provisions to  assure that the
             analyzer is warmed up may not be  necessary.
                                     28

-------
3.     The operator should check the analyzer log sheets  to determine when
       the last gas span and leak check were performed.   If a week or more
       has elapsed, then the analyzer needs to be gas spanned and leak
       checked (see Section 2.2).

4.     Several times during the day, the operator should  visually check  the
       water trap and empty it as required.  Even if equipped with a  contin-
       uous drain line, the water trap should be checked  anyway,  because the
       check valve can become easily clogged.  Filters should also be
       visually checked at least daily and changed as necessary.

5.     Immediately after warm-up and before each test the operator should
       check and adjust the zero and electrical  span.

6.     Before each test, a hydrocarbon hang-up check should be performed.
       See Section 2.2.1.2, Step 6.

7.     The inspectors should check any warning signals on the analyzer to
       make sure that they are not illuminated.   These signals include
       the low flow indicator and maintenance warning lights.

It should be noted that the first two steps in the analyzer preparation  do
not need to be repeated for each test, unless the analyzer is turned  off
between emissions tests.  The third check (checking the analyzer  log  sheets)
need only be performed once a day.  However, the last three checks should be
performed before each test.

3.1.2     Procedures for Vehicle Preparation

Prior to performing an emissions test, the vehicles should be prepared for
testing as follows.

1.     All accessories, such as, air conditioning, heater, lights, etc.,
       should be turned off.

-------
The vehicle should be checked to make sure it is adequately warmed
up.  The recommended method to assure that the vehicle is at normal
operating temperature is to check the vehicle's temperature gauge
(if so equipped) to make sure it is in the warmed-up position,  or
to feel the radiator or top radiator hose to determine if it is warm.
(Caution:  Be careful to avoid contact with the engine fan.)  One
or more of the following additional simple checks can also be used
for this purpose, but they are not considered to be as reliable
as the above method.

a.    If the vehicle has been idling for greater than five minutes,
      then most likely it is adequately warmed up.   The vehicle
      may not be warmed up if it is still  on fast idle.

b.    If the vehicle was just driven into the inspection station,
      ask the vehicle operator how far it was driven.  If driven
      over three miles in warm weather or over five to six miles
      in cold weather, then the vehicle should be adequately
      warmed up.  This check is not always conclusive and is
      especially variable in cold weather.

c.    During cold weather if water vapor is coming out of the
      exhaust, this may indicate that the vehicle is not adequately
      warmed up.  However, this type of check is not always
      conclusive.

d.    If the inspection also includes a safety inspection, perform
      the emissions test after all the safety checks have been
      made.  During the safety inspection, allow the vehicle to
      idle whenever possible.  (Make sure that the exhaust is
      adequately vented.)  If the safety program requires a road
      test, the emissions test could be performed after this
      road test.
                               30

-------
Note:  If the vehicle has to be warmed up, the inspector can make efficient
       use of his or her time by filling in the vehicle information portion
       of the inspection record and checking the analyzer while the
       vehicle is warming up.

3.     The inspector should determine if the vehicle has exhaust leaks since
       they may cause dilution of the test sample.   Most leaks should be
       audible to the inspector; however,  a visual  check should always be made
       to ensure that the exhaust system is intact.   If doubt exists, the in-
       spector could momentarily restrict  the tailpipe and listen for leaks.
       Most safety programs require a leak-free exhaust.  In this case, the
       emissions could be checked after the exhaust system has been checked
       for leaks and passed.  If leaks cannot be repaired or the test sample
       taken upstream of the leak, the test must be voided.   (Caution:  Exhaust
       leaks are safety hazards to both the driver  and to the inspector.)

3.2       Quality Control Procedures During the Test

The following general quality control procedures should be followed during
the emissions test.

1.     Switch the analyzer to sample mode  (it should be on standby between
       emission tests).

2.     When the idle test is being used, precondition the vehicle by
       operating the engine at 2,500 (± 300) rpm for up to 30 seconds
       (15 seconds is usually sufficient).  Preconditioning is not
       necessary when the two-speed idle test is being used.  (Note:
       Preconditioning removes any pollutants which may build up in the
       exhaust system while  the vehicle is idling and/or warming up.  Although
       preconditioning is included as a recommendation of this report, the
       idle test without preconditioning  is an acceptable and effective short
       test for I/M  programs.
                                     31

-------
3.     With the vehicle running,  insert the probe into the exhaust  pipe—If
       the probe cannot be inserted 16 inches  into the tailpipe,  an extender
       (leak tight fit) should be placed on the  end of the tailpipe to  meet
       this criterion.   The 16 inch insertion  depth is recommended  to
       avoid dilution.   A clip or other devices  on the probe can  be used  to
       prevent the probe from falling out during the inspection.

4.     Read the analyzer according to the following instructions:

       a.     Wait a minimum of 30 seconds or until a stable reading is
             obtained on the analyzer.  (Most  analyzers have response
             times of 15 seconds  or less.)  On some Ford vehicles,
             erroneous  readings may result if  the vehicle has been  idling
             for very long (as a  result of the air injection system
             diverting  to the atmosphere).  The air system can be
             reactivated by turning off and restarting the engine.
             (Remember  to remove  the probe while restarting.)  The  emissions
             test should be conducted immediately after restarting.

       b.     If the analyzer has  analog meters,  make sure that errors due
             to parallax are avoided.  See reference on page 12.  Analog
             meters typically have both the high and the low scales on
             the same meter; therefore, the operator must take care to
             make sure  that the proper scale is  read.

       c.     If the emission levels fluctuate  during the test, it is
             recommended to use the average of the low and high
             readings.   However,  individual states may have different
             procedures to handle this situation.

       d.     The emissions results should be recorded immediately on
             the form.   However,  if machine readable forms are used, the
             blanks can be filled in later.
                                     32

-------
        e.    If  a  two-speed  idle  test  is  performed, a tachometer must
             be  connected  to  the  engine.  The analyzer can be positioned
             beside  the  front fender so the  inspector can control the
             throttle  and  take readings simultaneously.  Alternatively,
             an  idle speed adjustment  tool  can  be affixed to  the
             throttle  to maintain the  proper high idle speed.

5.      Additional  items  to be aware of during  the inspection.

       a.     Of  foremost importance is that the inspector uses  the  correct
             emission  standards in testing  a vehicle.  All  I/M  programs
             have  specific standards  for  different types of vehicles,
             generally,  for vehicles  of similar model years.  Thus  it
             is  very important that the inspector correctly identifies
             the model year of the vehicle.   If there  is any  doubt  about
             the model year, the  inspector  should check  the vehicle
             registration card.  Some  I/M programs may classify emission
             standards in other ways,  e.g., by  engine  size  or number of
             cylinders.   Whatever the  case, inspection station  personnel
             must  be very careful to  use  the proper emission  standards
             in  testing  a vehicle.

       b.     The inspector must make  sure that  vehicles  exempt  from the
             inspection  requirements  are  properly identified.   For
             example,  in some I/M programs  vehicles older than  15 years
             are not required to  be inspected.   In addition,  vehicles
             above certain gross  vehicle  weights are  exempt from  I/M
             tests in  many programs.

       c.     The inspector must make  sure that  there  are no kinks  in
             the sampling hose and that the vehicle is not  parked on the
             hose.
                                     33

-------
3.3       Quality Control  Procedures After the Test

The following general  quality control  procedures  should  be  followed  after
the emissions inspection is completed.

1.     Put the analyzer back on standby—After the test  is  completed the
       inspector should place the analyzer back in the standby mode.   This
       is done to reduce the contamination of the sampling  line with water,
       pollutants and  dirt, and it also adds to the life of the pump
       and other related equipment.   In addition  to putting the analyzer
       back on standby, the hose and probe should be properly stored
       after each test.

2.     Complete data recordings—After the test is finished, the inspector
       should complete the data forms.   As mentioned earlier, most of the
       vehicle data should be recorded prior to the inspection, while the
       vehicle is being warmed up.  Afterwards the inspector needs to make
       sure that all appropriate items have been  filled  in, and, if  the form
       is machine readable, the appropriate slots need to be shaded.   Care
       must be taken to keep the form neat and clean.   This is especially
       important when  machine readable forms are  used.  The inspector
       must carefully  determine the pass/fail  status of  the vehicle.   In
       addition, if the vehicle is receiving a waiver, the  inspector must
       also make sure  that all appropriate data items  have  been completed
       on the waiver form.

3.     Optional;  Provide  evaluation forms to the motorist—The administrating
       agency should consider the merits of requiring  the inspection station
       to provide a brief  inspection station evaluation  form to each
       motorist after  the  emissions inspection has been  completed.  After
       completing the  form, the motorist would mail  it directly to the agency.
       The evaluation  forms would be designed to  document the motorist's
       perceptions of  the  performance of the inspection  station (both
       positive and negative).  It is recognized  that  motorists would be more
                                    34

-------
       likely to report inferior performance through these evaluations; thus,
       they would provide an additional  means to identify and correct
       quality control  problems.

3.4       Inspection Checklist

In order to assure that the inspection is performed properly, an inspection
checklist, which outlines all  inspection procedures, should be provided to
the station.  This list could  be posted  on the analyzer to prompt or remind
the inspector.  Figure 4 is an example of an inspection checklist.
                                     35

-------
                       Figure 4.   Inspection Checklist

PRETEST
Analyzer Preparation
       Daily Items
         •   Check mechanical zero.
         •   Warm up analyzer in  standby mode.
         •   Check analyzer log.   Does analyzer need to be spanned or
             leak checked?
         •   Check water trap and filter.  (Check several  times a  day.)
       Before Each Test
         •   Warm up analyzer (if turned off after previous test).
         •   Check zero and electrical span.
         •   Check warning indicators.
       Vehicle Preparation
         •   Turn off all  accessories.
         •   Warm up vehicle.
         •   Check for exhaust leaks.

TEST
         •   Switch to test mode.
         •   Precondition vehicle (if necessary).
         •   Insert probe.
         •   Connect tachometer lead for 2-speed test.
         •   Read analyzer after  30 seconds or until levels stabilize.
         •   Do not let Ford vehicles idle for more than 1 minute
             before taking readings.
         •   Record emission readings.
POST TEST
         •   Switch analyzer to standby.
         •   Complete data recordings.

-------
4.0       QUALITY CONTROL FOR INSPECTION STATION AUDITING

Periodic audits play an important part in the quality control  of inspection/
maintenance programs, especially decentralized ones.   By randomly visiting
inspection stations, auditors for the administrating  agency can help assure
that each inspection station is performing in accordance with the state's
licensing requirements.

There are two basic purposes for audits.  One purpose is to check for the
presence of the required equipment and supplies (i.e., analyzer, span gases,
tachometer, spare parts, inspection forms, stickers,  manuals, etc.) and to
check the performance of the emissions analyzers and  related equipment.  The
other purpose is to check the performance of the inspectors (or other
personnel) in conducting inspections, calibrating and maintaining the analyzer,
and keeping records.

Generally, for decentralized programs only one person would be needed to audit
a station, although a two-person team would be able to conduct the audits in
less time by dividing the individual audit tasks.  For example, while one
auditor is checking the analyzer, the other auditor could be reviewing
records.  For centralized programs, a two-person audit team may be needed
to inspect the station in a reasonable time period.

Auditors play a key role in the I/M program.  The auditors are the communica-
tion link between the inspection stations and the program managers.  The
auditors provide supervision and guidance to the inspection stations.  In
addition, the auditors directly control much of the feedback which the program
managers receive about the I/M program.  The two primary mechanisms for
feedback are the inspection records and the audit reports.  The auditors are
responsible for assuring the quality of the former during audits, and they
directly generate the latter.  Because of the significant role that auditors
play, especially in a decentralized I/M program, it is critical to the
success of the program that the administrating agency develop an effective and
comprehensive auditing program.
                                      37

-------
This section presents procedures to use in performing audits of the inspection
stations.  The sequencing of checks during the audit depends upon several
factors and should be left to the discretion of the auditor.  Normally, the
inspection station should be ready to conduct an inspection when the auditor
arrives.  In this case, it is usually easier for the auditor to perform the
analyzer checks first, and then check the proficiency of the inspection
station personnel (as needed) and check the records.  However, if the analyzer
is not turned on when the auditor arrives, the auditor should check the
records and other items first, while the analyzer is warming up.  In addition,
if a vehicle is being inspected when the auditor arrives, or at any other
time during the audit, the auditor should take the opportunity to observe
the inspection and evaluate the inspector.  A checklist (Figure 5) is provided
at the end of this section that can be used to assure that the auditor
adequately assesses all areas of importance during the audit.

4.1       Checking the Performance of the Equipment

The following procedures can be used to check the performance of the test
equipment.  If the analyzer is already turned on, these checks can usually
be performed in 30 minutes or less (per analyzer).

4.1.1      Checking Equipment Readiness

When the auditors arrive at the station, they should inquire as to whether or
not the station is ready to perform an inspection.  If so, the audit team
needs to check the condition and state of the analyzer.  They need to determine
if it is turned on and properly warmed up.  Generally, the analyzer should
be left on in the standby mode throughout the day.  However, if the audit
team arrives early in the day, it is reasonable to expect that the analyzer
has just been turned on or it may not be on at all.  If the audit team does
arrive before the analyzer is turned on, the auditors should take the oppor-
tunity to observe a normal power-on routine.  This routine is described in
Section 3.1.1.
                                     38

-------
4.1.2     Checking the Performance of the Analyzer

4.1.2.1   Equipment Required

The equipment required to check the performance of the analyzer is similar
to the equipment described in Section 2.2.1.1 (Equipment Required for the
Inspection Station).  However, the following additional or different equip-
ment will be needed in order to conduct an audit.

1.   Zero gas -- A bottle of zero gas that is hydrocarbon free (99.9% nitro-
     gen) or an air supply with an activated charcoal filter and a water
     trap.  (Using zero gas for zero checks should ensure that background
     pollutants in the station's room air do not bias the zero settings.)

2.   Audit span gas -- As mentioned earlier, the station should have a bottle
     of span gas that is close to the emissions standards of the emissions
     performance warranty regulations, (i.e., 1.5% CO, 600 ppm propane).  The
     auditors should also have a bottle of span gas with similar concentra-
     tions.  Since the inspection station's span gases are accurate to +2% NBS,
     it is recommended that audit span gases be accurate to +1% NBS.

4.1.2.2   Procedures

1.   Check the zero and electrical span, adjusting as necessary.

2.   Perform a HC hang-up check, preferably using a zero gas.  If the HC
     reading through the probe is greater than 20 ppm after 30 seconds while
     the reading through the port is 0 ppm, the appropriate maintenance needs
     to be recommended to the inspection station.  See Section 2.2.1.2,
     Step 6.  If the analyzer has no span port, this step must be omitted.

3.   Using the audit span gas, perform a combination gas span/leak check per
     the procedures in Section 2.2.1.2.  The analyzer should be within +5% of

-------
     the audit gas concentrations for both HC and CO and should not have any
     leak greater than 3%.  If analyzer accuracy is not within the +5% limit,
     the auditors should attempt to determine if there are any overt reasons
     for the discrepancy (other than the need for span adjustments).
     The check should include a comparison of the operator's span gas with the
     auditor's span gas to determine if the span gas concentration is dif-
     ferent from the value on the bottle.  If the analyzer does not have a
     separate calibration port (i.e., the span gas has to be entered through
     the probe), it is possible that a leak could be responsible for a low
     reading.  See Section 2.2.1.2 for description of methods to leak check
     analyzers that do not have span ports.  Any span adjustment or leak
     problems identified in an audit should be immediately corrected by the
     inspection station, or the analyzer in question should be immediately
     removed from service until it is repaired.   See Section 2.2.1.2, Step 10
     for recommended span adjustment procedures.

4.   Since gas spanning is being done on the low range, the auditor should
     check the low scale/high scale correlation  to make sure that the analyzer
     is accurate on the high range.

5.   Check other equipment on the analyzer -- In addition to the span checks,
     the auditors should check other items related to the analyzer.  These
     include filters and water drains.  In addition, the auditor should check
     the function of any of the automated checks if applicable.

4.2       Checking Procedures for Routine Tasks

In addition to checking the accuracy of the equipment that is used for the
emissions test, the auditors should also check to see that the inspector is
proficient in performing different tasks.  These checks do not need to be
performed during all audits; however, if there is evidence from previous audits
and/or from data analysis that the inspection station may not be performing
properly, then the proficiency of inspectors should be checked.  In some
cases, the inspectors may be performing inspections while the auditor is on
                                     40

-------
the premises.  In this case the auditor should always take the opportunity
to observe the inspections.

4.2.1     Observing Weekly Quality Control Checks

If the analyzer is found to be out of calibration, the auditor should have
the inspector attempt to adjust it.  The auditor should note if the inspector
does not perform the span adjustment properly (as described in Section 2.0 of
this report) and initiate appropriate actions to correct the problem (i.e.,
demonstrate the proper method to the inspector, or, if necessary, as in re-
curring cases, require the inspector to go through total retraining and
recertification.

4.2.2     Observing Inspections

As already mentioned, if possible the auditors should observe an inspection
while they are on the premises.  If there are no vehicles on the premises that
need inspection, the station could perform an inspection on the auditor's
vehicle.  When the inspection is being performed, the auditor should first
note whether the vehicle was adequately prepared.  That is, the vehicle
should be correctly warmed up and have no exhaust leaks.  In addition, the
auditor should note whether or not the analyzer was properly prepared for
the inspection.

The auditors need to answer the following questions to determine if the in-
spection is correctly performed.

1.   Is probe inserted correctly?

2.   Has the analyzer stabilized?

3.   Is the proper scale used (if dual scale)?

4.   Are the data recorded correctly?  During the inspection the auditors
                                     41

-------
     should record the values and compare these values with the values the
     inspector reports.  The auditor should also check to see that other
     data on the form are correctly completed.

5.   Were the correct emissions standards used?

4.2.3     Checking Documentation

Another important aspect of the performance of the inspection station is
maintaining adequate documentation for both analyzer records and inspection
records.  At least 10 percent (or 10-15 records, whichever is greater) of the
inspection records should be checked during each audit.  However, if problems
are found, or if there are recurring problems with particular stations, the
auditors may need to check a larger percentage of the inspection records.
Approximately 30 minutes should be allocated to checking documentation
unless automated data collection is used for inspection records.  In this
case, documentation checks should take approximately 10 minutes, since only
the analyzer records will need to be reviewed during the audit.

4.2.3.1   Checking Analyzer Records

The auditors should inspect the records that are kept in the log book to make
sure that the weekly quality control checks are being performed.  In addition,
the auditors should note whether preventive maintenance is performed per the
manufacturer's schedule or the schedule discussed in Section 2.3.  Finally,
the auditors should note the pressure in the span cylinder, record it in the
analyzer logbook,and compare this pressure with the pressure recorded during
the last audit.  If weekly span and leak checks are performed as required by
the emissions performance warranty regulations, then cylinder pressure
(assuming full size I-A cylinders) should drop approximately 50 to 100 psi
per month.  If a lower drop is noted, then the station may not be performing
weekly gas span/leak checks.  If other than size I-A span gas cylinders are
used, the amount of pressure drop may differ from the 50 to 100 psi noted
above.
                                     42

-------
The auditor should also qualitatively review the analyzer records.  They
should note whether all pertinent data are included in entries and if
reasonable values are entered.  In addition, they should also note if the
records are being completed as the analyzer is maintained.  (That is, are
the records current?)  At the same time, the auditors could note any indi-
cations of fraudulent record keeping.

It should be noted that the use of standard forms such as the log sheet dis-
cussed in Section 2.4 will greatly aid in the proper maintenance and review
of analyzer records.

4.2.3.2   Reviewing Inspection Records

The auditors can make several random checks on the inspection records.

1.   Are the records properly completed?  Are values entered for all fields?
     Are they legible?  The use of standard inspection forms that are easy
     to read and easy to understand will greatly aid record keeping; however,
     care must still be taken by the inspector in order to insure that data
     are recorded properly.

2-   Verify the garage's emissions test summaries, monthly reports, etc.
     (when applicable).  It is recommended that the garages prepare summaries
     or reports of their inspection activities on a monthly basis.  Doing so
     not only assists the state in reviewing the station's records, but also
     helps to ensure that the inspection station operator periodically
     reviews his/her station's performance.  One method to verify the accuracy
     of these summaries would be for the auditor to tabulate the overall
     failure rate for the preceding month.  The failure rate that the auditors
     arrive at could then be compared with the garage's tabulation.

3.   Randomly review 15 or 20 forms to answer the following questions.
                                     43

-------
    a.    Are proper standards being used?

    b.    Are vehicles being properly passed or failed?

    c.    Are recorded emissions readings reasonable?

    d.    Is there any evidence of improper record keeping?  For instance,
         is there a recognizable pattern to emissions readings (e.g., all
         CO readings are "	.0", etc.)?

    e.    Are the waivers legitimate as indicated by cost, type of
         repair, and the emissions reductions?

    f.    Is the repair data reasonable (if available)?  The auditors could
         look at the cost versus type of repair to determine if the garage
         is charging excessive costs for simple repairs.  The auditors
         could also question extremely high repair cost, especially if
         there is a very low emissions reduction.  At the same time, the
         auditors can note the general distribution of the repair costs.
         That is, are most less than $10, between $10 and $20 and so forth?
         These general distributions could then be compared with program
         trends to date.  If the station appears to have questionable
         repair data, then those stations could be audited more frequently.

4.  It is still necessary to manually review the data even if they are
    recorded on machine readable forms.  These forms must be checked to make
    sure that the data are entered in the proper fields and that they are
    coded properly.  In addition, it must be stressed that the forms must be
    clean and undamaged, and all the appropriate fields are completed.

4.2.3.3   Checking Sticker Records

While the auditors are at the garages, they should also review the records
kept on the inspection stickers.  Since most programs that use stickers
stagger inspections over the calendar year, the auditors could collect the
                                   44

-------
unused stickers from the previous month.  The auditors could note the serial
numbers of these stickers (or the number of stickers, if not serially
numbered), and with the aid of past data on the serial numbers of stickers
(or the number of stickers) issued to the station,the auditors could determine
the number of inspections that should have been performed.  This number
could then be compared with the number of inspections as noted by inspection
records for the same time period.  The auditors could also be responsible
for the delivery of stickers which would aid in the accounting of them.

4.3       Other Items to Inspect During the Audit

Aside from checking the accuracy of the analyzer and the procedures used by
the inspectors, the auditors could perform other checks while they are at
the station.  First, they could determine whether all licensing requirements
are being met.  In many cases, for example, inspection stations must have
current repair manuals or certain equipment in addition to the emissions
analyzer on the premises in order to comply with state licensing requirements.
Furthermore, some state specifications require that certain documents such
as the station's license, the inspection procedure, emission standards,
etc. be posted in the station.  Depending on the particular state's rules
and regulations, there may be other licensing requirements which apply to
the stations.

It is also important for the auditors to review complaints with the inspec-
tor and/or the station manager.  Complaints regarding repair practices and
emission challenge checks should be carefully reviewed.

4.4       Taking Corrective Action

All of the audit functions discussed thus far have dealt with the things that
auditors evaluate during an audit—the analyzer, the records, the proficiency
of the station personnel, etc.  In making these evaluations, the auditor in
each case should have predetermined limits (or ranges) of acceptability in
order to be able to rate the performance of the station.  Suggested limits of

-------
acceptability for emissions analyzers were discussed in Section 2.0; i.e., an
analyzer should be accurate within +5% during gas spanning and should have
no leaks which cause errors of more than 3%.  The quantification of performance
and the limits of acceptability are less precise for record checks and
personnel proficiency checks.  In these cases, more variability in performance
must be expected because of the differences in the abilities and attitudes
of the people involved directly in these areas.  In contrast, the analyzer
is a machine built to specific tolerances and, therefore, can be expected to
perform to tight and well-defined limits.  Table 3 lists some examples of
the criteria that can be used in developing limits of acceptability for
record checks and personnel proficiency checks; however, the reader should
not construe Table 3 as an exhaustive listing of these criteria.

In some cases (hopefully a small minority), during the conduct of audits of
inspection stations, the auditor will inevitably identify problems which need
to be corrected.  The problem may be as simple as untidy record keeping or as
serious as a flagrant disregard of proper testing procedures by the inspec-
tor.  Depending on the severity of the identified problem(s) and on the system
of penalties for ma 1 performance in the particular state involved, the auditor
may have various courses of corrective action open to him/her.  Many problems
can be corrected cooperatively simply by identifying and discussing them
with the appropriate station personnel.  If this is not successful, the
auditor may be able to correct the less severe problems, such as the case
of untidy record keeping, by giving the station personnel an oral or written
warning and then following up during the next audit to make sure the problem
was corrected.  In more serious cases or in cases where there are recurring
problems, the auditor may need to take harsher action, such as, suspending
an inspector's license and/or the station's license or requiring that an
inspector go through retraining and recertification.  Whatever the case, the
auditor must be prepared to take appropriate corrective actions as problems
are identified.  At the same time, auditors must be careful to document all
the facts before making a judgment about the performance of an inspector
or inspection station.  In some cases, a definite problem may not be identi-
fied conclusively in a single audit and must be followed up in a series of
                                     46

-------
TABLE 3:  AUDIT CRITERIA FOR RECORD CHECKS AND PERSONNEL PROFICIENCY CHECKS

Record Checks

     1.   Is an inspection record being completed for each inspection (i.e.,
          are there an appropriate number of inspection records as compared
          to the number of stickers issued)?

     2.   Is each record completed properly?  All fields completed?
          Reasonable entries?  Any indications of incorrect or fraudulent
          entries?

     3.   Are the inspection records legible?

     4.   Are vehicles being properly passed or failed?

     5.   Are the correct emission standards being used?

     6.   Is the failure rate realistic?  Too high?  Too low?  Are suspected
          inconsistencies explainable?

     7.   Are repairs consistent with the failure mode?  In other words, are
          the problems with failed vehicles being properly diagnosed and
          repaired?  (This information may not always be available.)

     8.   Is the repair cost consistent with the repair?  (This information
          may not always be available.)

     9.   Are there variations in performance (failure rates, repair costs,
          etc.) between inspectors within the same inspection station?

     10.  Are any required monthly summaries or reports being done properly
          and on time?
                                 (Continued)
                                    47

-------
TABLE 3:  (Continued)

Personnel Proficiency Checks

     1.   Does the inspector know and use the proper testing procedures?

          a.    Does he/she know how (and why) to properly prepare the analyzer
               for testing?

          b.    Does he/she know how (and why) to properly prepare the vehicle
               for testing?

          c.    Can he/she perform the emissions test properly?

          d.    Does he/she know and follow prescribed quality control pro-
               cedures after the test is completed?

          e.    Does he/she know how to properly complete an inspection form?

     2.   Does the inspector or other appropriate person know how, why, and
          when to conduct periodic quality control checks and/or preventive
          maintenance on the emissions analyzer?

     3.   Does the analyzer logbook document that prescribed quality control
          checks and/or maintenance are being conducted at the appropriate
          times by qualified personnel?

     4.   Is  the inspection station supervisor or owner knowledgeable about
          the conditions within the station and performance of his/her in-
          spectors?

     5.   Are all inspection stickers accounted for?
                                    48

-------
audits, or through other investigations, in order to fully document the
problem.

To ensure that auditors deal with inspection stations in a fair and equitable
manner, the administrating agency should identify the kinds of corrective
actions which should be applied to particular offenses.  This information
should most likely be included in operating guidelines or procedures for
auditors to follow.  In addition, auditor supervisors should routinely
review corrective actions that have been taken against offenders to ensure
that the actions were consistent with agency policy.

4.5       Documentation

To aid in the performance of an audit, a checklist or audit form should be
completed during the audit.  As mentioned earlier, some of the items on
the checklist, i.e., reviewing inspector proficiency, may not need to
be performed during each audit.  However, the checks of the emissions analy-
zer and records should be performed every time the auditors are at the
station.  It is important to evaluate the emissions analyzer(s) during
every audit because of the analyzer's relative sensitivity to its external
environment.  This results in a high potential for measurement errors which
can be minimized through frequent routine analyzer checks.  Routinely
checking records during every audit will better enable the auditor to
assure that records are being kept properly and that the station is per-
forming satisfactorily.  Frequent routine record checks also provide
assurance that any problems that may exist are identified and corrected
promptly.  Figure 5 is a sample of an audit checklist that could be used in
a decentralized program.

The audit checklist (or audit report) should be designed to allow complete
documentation of the audit.  Space should be provided for the auditor to enter
explanations and descriptions of his/her findings and/or recommendations.  In
particular, there should be space provided at the end of the form for the
auditor to report the results of the audit, any needed follow-up activities,

-------
Figure 5:  Sample Audit Checklist
Inspection Station Number

Name   fcxce. I
                              3305
                                                  Date

                                                  Time
Location  90*
                        St- .
                                         , ff)A
                                                  Auditor
Station Official's Signature/Title  /d+JL
                                                          54/wicc.
Yes/No
                 Operation
                                                   Comments
                    RFADTNFSS
                    KtMUint;>;>
           I. Analyzer on?     r*
          2. Analyzer warmed up?t«
          3. Zero and span set? 73? /»voi«
                                                o*J & AM no A  ~AT f.'Ys^./n. ~f]e
                                         hA<> &&*> *>Aoe. o*J3-it'9t AS or THAT
                                                T»  /?• scofie.1-* TTH+T -rue,
                                               TVAAJAA OAJ e/i*+r /*>  -a*.
                                          reyn*i6.
          ANALYZER CONDITION
          1 . Zero ok?
          2. HC hanq-up?
          3. Gas span within ±5%?
             If no, was span adjusted
               successfully?
          4- Leaks?
             If yes, were they corrected?
          5. Scale correlation ok?
                                                 -OK.
                                                      /
                                                 •/""/<•«*« c**c*c
_     _
__/Oo_   6. Other defects?
                                                      *„****
          1. Sets span and zero correct y?
          2. Performs proper leak check?
          3. Makes sure vehicle is warmed up?
          4. Checks for exhaust leaks?
          5. Inserts probe properly?
          6. Reads analyzer correctly?
          7. Uses proper emission standards?
          8. Completes form correctly?
          9. Other factors?

                                                     v*A'.c/e  / /us/>«cr7*x/
          ANALYZER RECORDS
          1. Weekly QC checks being  performed?
          2. Span cylinder pressure  drop
               consistent?
          3. Regular maintenance  performed?
          4. Records current?
          5. Indications of fraudulent or
               improper record  keeping?
_J\)0__   6. Other factors?
                                                     */
                                                         4ac<*u>s>
                                     50

-------
Figure 5:  Sample Audit Checklist (continued)
Yes/No
            Operation
                                       Comments
                                       &*3CiJCSe4
                                             Be
INSPECTION RECORDS
1. Records properly completed?
2. Records neat and clean?
3. Records legible?
4. Failure rates reasonable?
5. Repair costs reasonable?
6. Missing or unaccounted for stickers?
7. Indications of fraudulent or jDo%p&c.n«A)
     improper record keeping?    —   -•—-
8. Other factors?
                                                             MOT
                                                               77* /»
                                                                tf.
                                                               - OK
          GENERAL
          1.  All licensing requirements
               being met?
          2.  Any serious problems or
               complaints?
          3.  Other factors?
NOTES
  /vJ«
(Indicate in this space all  findings, recommendations, areas
 needing follow-up,  explanations, etc.  resulting from this audit.)

                           Auditor's Signature
                           Reviewed By  _  Q. ,
                                    51

-------
Figure 5:  Sample Audit Checklist  (continued)

           Audit Gas Span/Leak  Check Worksheet


Analyzer Make/Model  	ScMJ
Propane/Hexane Equivalency Factor (PEP)   Q.57
Audit Gas Concentrations:   CO _ /.. SS _ %
                           HC _ (, 30 _ ppm propane

1.  Acceptable Analyzer Response Ranges  (±5%):

      CO:  (0.95 X CO Audit Gas  Concentration)  to
                           (1.05 X  CO Audit Gas Concentration)

           (0.95 X   /-S8      ) to (1.05 X
                                 to  _ /.
      HC:  (0.95 X Propane Audit Gas  Concentration X PEF)
                to (1.05 X Propane Audit  Gas  Concentration X PEF)
           (0.95 X  ^30    X 0.57  )  to  (1.05 X  6«?o  xO.«5V   )

                       30Q _  to

2.  Analyzer Responses to the Audit  Gas:

                 CO probe
                 CO port
                 HC port        3AS
3.  Gas Span Result: 	Miss
4.  Acceptable Leak Limit:

      CO port response should be less  than  (1.03 X CO probe response)

      (1.03 X CO probe response)    =   (1.03  X
5.  Leak Check Result:        /2i-ss
                                    52

-------
and any corrective actions required.  A well designed audit form will help
ensure thorough audits, complete documentation of them, and equitable
treatment to all inspection stations.

Written documentation of the audit findings is also important to the inspec-
tion station.  The station supervisor/owner (or other person in charge at the
time of the audit) should be given a copy of those parts of the audit form
which indicate the results of the audit and any corrective actions that were
taken or that are needed.  If a multiple copy audit form is used, the
station's copy of the audit report can be generated automatically as the
report is completed by the auditor.  Since some of the auditor's comments
or findings may need to be confidential, there should be certain parts of
the audit form which do not duplicate on the station's copy.  To assure that
the inspection station receives this feedback on the audit, it is recommended
that an appropriate space be provided on the audit form for the station offi-
cial in charge to sign it.  His/her signature would then signify that the
audit was conducted at the time indicated and that the auditor discussed
the results of the audit with him/her.  The station official's signature
would also provide assurance to the administrating agency's audit supervisor
that the audit was conducted.
                                      53

-------
5.0       QUALITY ASSURANCE IN AN I/M PROGRAM

The previous sections of this report have dealt with various aspects of
quality control for an I/M program, i.e., the system of activities needed
to provide a quality I/M program.  Sections 2.0 and 3.0 dealt specifically
with internal (within the inspection stations) quality control aspects while
Section 4.0 shifted to external quality control aspects, or quality assurance.
Quality assurance is the system of activities needed to provide assurance
that the quality control system is performing adequately.  In other words,
quality assurance is quality control for the quality control system.

Audits (Section 4.0) are one aspect of quality assurance.  Each audit pro-
vides the mechanism for quality assurance for the particular inspection sta-
tion involved.   Taken together, the audits for all inspection stations provide
data which allow program managers one way to assess whether the I/M program
is operating as intended.  Further data for this purpose are provided through
inspection records, public opinion surveys, special studies, air quality
data and other sources.  With these sources of information, program managers
can track trends in the program, identify problem areas, and institute
needed program modifications to correct them.  These kinds of continuous
feedback are needed for the I/M program managers to keep the program on
track toward meeting the desired objectives.  (Note:  Further information
on I/M data analyses is available in another EPA report prepared by Radian
Corporation entitled "Guidance on Data Handling and Analyses in an Inspection/
Maintenance Program".)

To assure that the quality assurance function is given adequate consideration
by top management, a specific organizational unit(s) should be given the
quality assurance and quality control responsibilities for the I/M program.
In addition, clear and specific objectives for quality assurance should be
defined as well as ways to measure whether the quality assurance objectives
are being achieved.  Table 4 lists examples of the kinds of questions that
the quality assurance program should be designed to answer; however, the
reader should not construe Table 4 as an exhaustive list.  (Note:  Further
information on quality assurance planning is available in an EPA report
entitled "Quality Assurance Handbook for Air Pollution Measurement Systems".)
                                    54

-------
              TABLE 4:   QUALITY ASSURANCE QUESTIONS FOR AN I/M PROGRAM

1.    Is the I/M program operating as intended?  Are I/M program objectives
     being achieved?

     a.   Are the inspection stations being operated properly?

     b.   Is the failure rate as expected?

     c.   Are vehicles  being properly repaired?  At a reasonable cost?

     d.   Is the waiver rate reasonable?

     e.   Is there reasonable public acceptance of the program?

2.    Are there any special problems or trends which need to be addressed?

     a.   Do the emission standards for any particular class(es) of vehicles
          need to be changed?

     b.   Are any changes needed in the forms or data collection system?

     c.   Are there any enforcement problems?  Is the appropriate number of
          vehicles being inspected?

     d.   Are there any problems with certain makes or models of analyzers?

     e.   Are there any special repair or diagnosis problems which should be
          addressed through future training?

     f.   Is the audit system functioning properly?

     g.   Are there any special problems involved with testing or repairing
          certain vehicles which EPA should be made aware of in order to

                                     55

-------
TABLE 4:  Continued

          notify other states or to investigate methods of resolution?

3.   What is the proper mechanism to address any problems?

     a.   Through changes in the training program?

     b.   Through changes in the quality control system?

     c.   Through changes in program procedures (i.e., testing procedures,
          emission standards, enforcement, record keeping, reporting,
          auditing, etc.)?

     d.   Through changes in the public awareness program?
                                     56

-------