Office of Air and Radiation
Motor Vehicle Tampering Survey -
          Field Operations and Support Division
              Office of Mobile Sources
                Washington, D.C.

-------
             Office of Air and Radiation
Motor Vehicle Tampering Survey - 1989
                May 1990
          Field Operations and Support Division
              Office of Mobile Sources
                Washington, D.C.

-------
                    TABLE OF CONTENTS
EXECUTIVE SUMMARY.	1

   Introduction	,	I

   Conclusions	2


BACKGROUND	5


SURVEY METHODS	6

   Site Descriptions	7


RESULTS	12

   Vehicle Tampering	.	,	12
       Site and Aggregate Totals	.	12
       Types of Tampering	,	.	14
       Vehicle Characteristics and Tampering	17
          Manufacturer	17
          Vehicle Type	18
          Vehicle Age	20
       Impact of I/M and Antitampering Programs	21
       Tampering Trends for Selected Sites	24
       Impact of Aftermarket Catalytic Converters	.	32
       Correlation Between Tampering and Idle Emissions	33
   Fuel Switching	.	36
       Fuel Switching Indicators and Overlap	36
       Fuel Switching Trends	39
       Fuel Switching by Vehicle Type	41
       Fuel Switching and Catalyst Tampering.	41
       Gasoline Lead Concentrations	43

APPENDIXES

   A. Relevant Portions of Clean Air Act	45

   B. Survey and Data Recording Procedures	45

   C. Emission Cutpoints for I/M Areas	51

-------
                           LIST  OF  FIGURES

1.   Tampering rates for selected components by program type	3

2.   (a) Overall tampering by vehicle age for vehicles not covered
    by I/M and/or anlitampering programs	4

    (b) Comparison of catalyst tampering by vehicle age between
    vehicles covered and not covered by a catalyst inspection program	4

3.   Breakdown of surveyed vehicles by condition and extent of tampering	12

4.   Tampering rates by manufacturer	17

5.   Classification of 1989 survey sites by program type (and model year
    coverage), with the resulting tampering and fuel switching rates for
    each program type	22

6.   (a)-(O  Comparisons of component-specific tampering rates by model
    year between vehicles covered and not covered by ATPs	27 - 29

    (a) Catalytic Converter	27

    (b) Inlet Restrictor	27

    (c) Air Pump System	28

    (d) Evaporative System	28

    (e) PCV System	29

    (0 EGR System	29

7,   Comparison of component-specific tampering rates among vehicles
    both covered and not covered by ATPs (summary of figures 6(a)-(f))	30

8.   Distribution of unleaded vehicles surveyed among tampering, misfueling,
    and idle test categories	34

9.   (a) Overlap of fuel switching indicators among misfueled vehicles
    for which all three indicators were inspected	38

    (b) Breakdown of misfueled vehicles in 9(a) by prevalence of each
    indicator combination	38

10.  (a) Overlap of indicators used by ATPs to detect missing/damaged catalysts.
    Only includes vehicles for which all three indicators were inspected and are
    not covered by  an ATP with plumbtesmo testing	42

    (b) Breakdown of vehicles in ll(a) by prevalence of each indicator combination	42

12.  Distribution of lead concentrations in leaded fuel sampled from fuel switched
    vehicles: 1984 - 89.  Percentages based on all fuel switched vehicles, including
    those using unleaded fuel (<0.05 gpg.)	44
                                          -it-

-------
                           LIST OF TABLES

1.  1989 Tampering Survey Summary - Light-Duty Vehicles only	13

2.  Prevalence of Tampering by Component	14

3.  Component-Specific Tampering Rates (%) by Survey Location	15

4.  Condition of Gasoline Tank Cap by Survey Site	16

5.  Tampering Percentage (and Sample Size) by Model Year
    and Vehicle Age at Time of Survey	,	19

6.  Percentage of Catalyst Removal (and Sample Size) among
    Catalyst-Equipped Vehicles by Model Year and Vehicle Age
    at Time of Survey	,	19

7.  Comparison of 1989 Survey Sample to Actual Nationwide
    Vehicle Fleet	23

8.  Tampering Prevalence among Vehicles and Components
    Covered by Antitampering Programs in ATP-only Areas
    Surveyed between  1984 -1989	25

9.  Tampering Prevalence among Vehicles and Components
    Covered by Antitampering Programs in I/M + ATP Areas
    Surveyed between  1983 - 1989	26

10.  Comparison of 1987 Tampering Rates in Dallas (ATP-only)
    to 1989 Rates in Dallas Suburbs (non-I/M). Analysis is Restricted
    to the Model Years covered by the Dallas ATP	31

11.  Comparison of Tampering and Fuel Switching Rates Surveyed
    in Colorado Springs (Fourth Day of Denver Survey) to the Balance
    of the Vehicles Surveyed in Denver. Includes Only Vehicles
    Covered by the Colorado I/M + ATP	31

12.  Comparison of Catalyst Removal Rates to the Prevalence of
    Aftermarket Catalysts by Survey Location	32

13.  Idle Test Failure Rates (Percent) by Pollutant and
    Vehicle Condition	35

14.  Mean Idle Emissions for Tampered and Okay Vehicles
    Within Each Program Type	36

15.  Fuel Switching Rates among Unleaded Vehicles by Site
    and Indicator	39

16,  Combined Tampering and Fuel Switching Rates	_	40

17.  Prevalence of Fuel Switching Indicators by Program Type	41

18.  Prevalence of Fuel Switching Indicators by Vehicle Type	41
                                       - ill -

-------
EXECUTIVE SUMMARY

INTRODUCTION
                            Under the direction of the Field Operations and Support Division (FOSD) of the
                       Environmental Protection Agency (EPA), contract personnel from Colorado State
                       University (CSU) conducted a survey of light-duty motor vehicle tampering in 15 cities
                       between March and August, 1988.  The areas surveyed and the total number of vehicles
                       inspected are listed below.
                            San Diego, CA          500              Cleveland, OH         368
                            Bakersfield, CA        520              Springfield, MO        540
                            Phoenix, AZ            465              Milwaukee, WI         490
                            Suburbs of Dallas, TX   422              Missoula,MT          500
                            Baton Rouge, LA       497              Anchorage, AK        503
                            Jacksonville, FL        580              Fairbanks, AK         436
                            Detroit, MI             441              Denver, CO           465
                            Cincinnati, OH         390

                            TOTAL           7,117 vehicles

                            The objectives of this survey were:

                          1. To make local measurements of the types and extent of tampering and fuel
                              switching.
                          2. To extend and update the knowledge gained from earlier surveys on:
                              a. The effect  of vehicle inspection  and maintenance (I/M) programs and
                                  antitampering programs (ATPs) on tampering and fuel switching.
                              b. The relationship between tampering and vehicle idle emissions.
                              c. The distribution of tampering by vehicle age, type, manufacturer, and
                                  other variables of interest.

                            To achieve these objectives, the inspection teams visually examined  emission
                       control devices and measured the  idle hydrocarbon (HC) and carbon monoxide (CO)
                       emissions of each vehicle.  To provide information on fuel switching, the inspectors
                       sampled gasoline from the tanks of vehicles (for later laboratory lead analysis), tested
                       for lead deposits in tailpipes using  Plumbtesmo*1 test paper, and checked the integrity
                       of the fuel filler inlet restrictors. Four categories were used to summarize the condition
                       of the inspected vehicles:

                          1. Tampered - at least one control device removed or rendered inoperative
                          2. Arguably Tampered - possible but not clear-cut tampering (i.e., may have
                              resulted from either tampering or malmaintenance)
                          3. Malfunctioning
                        '  4. Okay - all control devices present and apparently operating property
                    Plumbtesmo* is a registered trademark, and hereafter appears without the ***. It is manufactured by
                    Machery-Nagel, Durtn, W. Germany, and is marketed by Gallard-Schlesinger Chemical Corp., Carle Place, New
                    York.

-------
Motor Vehicle Tampering Survey -1939                                                        page 2
CONCLUSIONS
                             These brief but thorough inspections were performed with the consent of the
                       vehicle owners in a variety of settings more fully detailed elsewhere in this report.

                             The site selection in recent tampering surveys has been based more on the need
                       to evaluate control program effectiveness than the desire to methodically sample the
                       U.S. vehicle population.  In the 1989 survey, for example, vehicles covered by an I/M
                       and/or antitampering program comprised 62% of the survey, even though only 42% of
                       the vehicles nationwide are subject to such control programs. The overrepresentation
                       of control programs within the 1989 survey data base makes it inappropriate to use this
                       survey as representative of the nationwide vehicle population. Rather, the  1989 survey
                       sample is an appropriate tool for evaluating tampering prevalence at any particular site,
                       and for evaluating control program effectiveness where "before" and "after" data exist.
                       This report will thus refrain from comparing the overall 1989 survey results with prior
                       surveys.
                             Tampering among vehicles not covered by I/M and/or anti-tampering programs
                       was 25%, compared to 15%, 12%, and 10% for those vehicles covered by I/M + ATP,
                       I/M-only, and ATP-only, respectively. Such comparisons between program categories
                       should be made carefully, since some of the control programs cover only newer model
                       years of vehicles, which historically have lower tampering rates than the older vehicles
                       not covered by the control programs. The I/M-only and ATP-only rates,  for example,
                       are heavily weighted with programs inspecting 1980 and newer vehicles only.

                             Because of the large variation in program types and coverages, the vehicles are
                       classified based on the program restrictions within each site. In Cleveland and Cincin-
                       nati, for example, only the 1980 and newer vehicles are covered by the area's antitamper-
                       ing program; the 1975-1979 vehicles are thus classified as "non-I/M".

                             The 1988 tampering survey report noted a disparity in the effectiveness of control
                       programs to reduce certain kinds of tampering. In particular, tampering with underhood
                       components was less likely to be corrected by a control program than removal of a
                       catalytic converter. While this conclusion was generally true in 1989 as well, some control
                       programs surveyed in 1989 (e.g., Anchorage) were very effective in reducing underhood
                       tampering.  Air pump and PCV tampering rates  among vehicles covered  by the
                       Anchorage program were among the lowest ever reported.  These rates are particularly
                       impressive because Alaska's own tampering survey prior to program implementation
                       found extremely high tampering rates there. Moreover, Anchorage's control program
                       is  decentralized, and such programs have been considered to  be  less  effective for
                       underhood components than centralized programs. These results will be discussed more
                       thoroughly later in this report.

-------
Motor Vehicle Tampering Survey -1989
                                                    Page3
Fuel-Related
                           The prevalence of vehicles with missing or damaged catalytic converters within each
                      program type is depicted in Figure 1.  A vehicle can be considered to have a missing or
                      inoperative catalyst if the converter has been removed, the inlet restrictor is tampered, or
                      the plumbtesmo test is positive. (These three criteria  are used by many antitampering
                      programs.) A missing or damaged converter can increase HC and CO emissions by an
                      average of 475% and 425%, respectively2.  For vehicles equipped with three-way con-
                      verters, substantial increases in NOx  emissions would also be expected.  In addition,
                      converter tampering on trucks was found to be slightly greater than on automobiles (5%
                      vs. 4%).

                           Fuel switching, defined as the presence of any of three specific indicators3, was found
                      in 9% of the unleaded vehicles not covered by I/M and/or ATPs in the 1989 survey. Fuel
                      switching among vehicles in I/M + ATP, ATP-only, and I/M-only areas was 3%, 2% and
                      4%, respectively. The pattern of overlap among the three misfueUng indicators is discussed
                      in detail later in this report. While the emissions impact of fuel switching depends upon its
                      duration and certain vehicle characteristics, emission increases of 475% for HC and 425%
                      for CO can easily occur.
                            Component
                           Catalytic Converter
                               Inlet Restrictor
                          Positive Plumbtesmo
0%
0%
0%
                                                           Tampering (%)

                           Ffgur* 1. Tampering rates for selected components by program type.
                     The emission! increase* mentioned in thi* report are from a study of three-way catalyst-equipped vehicles
                     presented in Anti-'T'pm^fipganri Antt-MafueJInff Program^ tn Reduce [n-Uftfl Fjni«imvt tmm Mntor Vehicle.
                     EPA-AA-TTS-83-10, December 31,1983.

                     The three fuel switching indicators are: a tampered fuel filler inlet restrictor, a positive Plumbtesmo tailpipe test.
                     or a gasoline lead concentration of more than 0.05 gram per gallon.

-------
Motor Vehicle Tampering Survey -1989
                                                               Page 4
Catalyst Tampering and Vehicle Age
                            The probability that a vehicle has been tampered with is clearly related to its age,
                       as has been shown in previous surveys. This is evident in Figure 2(a), which depicts
                       vehicles not covered by I/M and/or antitampering programs. Figure 2(b) compares the
                       impact of vehicle age on catalyst tampering among vehicles covered and not covered by
                       converter inspection programs. Overall, converter tampering is much lower on vehicles
                       covered by converter inspection programs than on vehicles not covered by such programs
                       (1% vs. 7%), and the effectiveness of inspection programs is particularly noticeable
                       among the oldest vehicles surveyed (vehicles that are most likely to be tampered with).
                                       (%)
80

70

60

50

40

30

20

10

 0
                                                                                           Covered
                                                    6   7   8   9  10  11   12  13  14  15
                                                       Van £»•*«• (years)
                         CataJyst Tampering (%}
                             50,
                             40
                             30
                             20
                             10 r
                                                                                        i vehicles No:
                                                                                        |  Covarec
                                                                                          vehicles
                                                                                          CoverM
                                  '   2   3   4   5   6   7   8  9   10   11  12  13  14   15
                                                     Vehicle Ag« (years)
                                                           (b)

                         Figure 2(a).  Overall tampering by vahlcl* «g« for vehicles not covered
                                       by I/M and/or antftamperlng program*.

                                2(b).  Comparison of catalyst tampering by vehicle age between
                                       vehicle* covered and not covered by a catalyst Inspection
                                       program.

-------
Motor Vehicle Tampering Survey -1989                                                        Page 5
BACKGROUND
                            EPA estimates that motor vehicle emissions nationwide account for 56% of the
                       totaJ carbon monoxide (CO), 32% of the airborne lead, 26% of the hydrocarbons (HQ,
                       and 31% of the oxides of nitrogen (NOx) emitted into the atmosphere4. As a result, a
                       major focus of the nation's efforts to achieve compliance with clean air standards has
                       been the control of emissions from mobile sources. The first pollution control devices
                       were installed on vehicles in 1962, and most light-duty vehicles manufactured since 1968
                       have been equipped with a variety of emission control devices to meet required emissions
                       standards.
                            The 1977 amendments to the Clean Air Act (sections 203(a)(3)(A) and (B), found
                       in Appendix A) make it illegal for automobile dealers, repair and service facilities, and
                       fleet operators to disconnect or render inoperative emission control devices or elements
                       of design.  Regulations issued under section 211(c) of the Act (40 CFR Part 80) prohibit
                       retailers and wholesale purchaser-consumers from introducing or allowing the introduc-
                       tion of leaded gasoline into vehicles labeled "unleaded gasoline only". The EPA's Field
                       Operations and Support Division (FOSD), formerly part of the Mobile Source Enfor-
                       cement Division (MSED), is responsible for enforcing the tampering  and misfueling
                       provisions of the Act.
                             Before 1978, the EPA had data suggesting that tampering with emission control
                       devices and misfueling of "unleaded only" vehicles with leaded gasoline were occurring.
                       Variability in the inspection procedures, however, prevented an accurate assessment of
                       the nature and extent of the tampering.  As a result, the Agency began conducting
                       nationwide tampering surveys of light-duty motor vehicles in 1978 to determine the rates
                       and types of tampering and fuel switching.  These annual surveys have been conducted
                       either by FOSD directly or by EPA's National Enforcement Investigations  Center
                       (NEIC) under the direction of FOSD. Consistent inspection procedures were used
                       throughout these surveys to permit comparisons and identification of trends.
                            The uses for the tampering surveys have evolved since the first survey was con-
                       ducted in 1978.  Since 1983, the tampering survey results for some locations have been
                       used to calculate credits for State Implementation Plans (SlPs), the measures taken by
                       State and local governments to achieve ambient air quality standards by reducing mobile
                       source emissions. Data from the surveys is also used in the default  data base for the
                       Agency's mobile source computer model (MOBILE4) to estimate both the emissions
                       loading impact and the reductions that may be achieved by various control programs.
                       Comparing survey results before and after implementation of a control program is an
                       excellent method for evaluating program effectiveness. Sites for the surveys are chosen
                       in light of the need for data on specific areas either currently operating or considering
                       programs, as well as the continuing need to monitor the types and extent of tampering
                       and fuel switching nationwide.                           ,  .

                       National Air Pnllutanr Emission Fjjtjfliates 1940-19SJJ, EPA-450/4-90-001, March, 1990.

-------
Motor Vehicle Tampering Survey -1 QQQ                                                        Page
SURVEY METHODS
                            The 1989 tampering survey was conducted for FOSD by the National Center for
                       Vehicle Emissions Control and Safety at Colorado State University (CSU).   Ap-
                       proximately 400 to 600 vehicles were inspected in each of 15 cities between March and
                       August, 1989. The entire survey included 7,117 vehicles. The mix of vehicles inspected
                       was assumed to be a self-weighting sample, and no attempt was made to force the sample
                       into matching the national vehicle mix,


                            Each inspection team consisted of at least four members: three CSU personnel,
                       one or two EPA representatives, and frequently a State or local agency representative.
                       The CSU personnel, assisted by the State or local person, performed the actual inspec-
                       tions, while the EPA representative(s) monitored the survey.  Each vehicle inspection
                       included the following:


                              1. Basic vehicle identification data recorded (year, make, model);

                              2. All emission control systems checked;

                              3. Idle HC, CO, CO2, and 02 emissions measured;

                              4. Fuel sample collected for lead analysis;

                              5. Tailpipe tested for lead deposits using Plumbtesmo test paper; and

                              6. Integrity of fuel inlet restrictor checked.


                            Three minor changes have been made b the survey methodology for 1989.  The
                       survey team utilized a four-gas analyzer to record idle COj and Oa emissions in addition
                       to the HC and CO emissions. Also, an additional classification code was added for the
                       air pump.  Finally, heavy-duty gasoline-powered vehicles were inspected at selected sites
                       in addition to the light-duty cars and trucks surveyed. The results from the heavy-duty
                       trucks will be discussed separately in this report, and, except where noted, were not used
                       to calculate any tampering or fuel  switching rates. The inspections and recording
                       procedures are fully detailed in Appendix B.


                            The survey data base has been reviewed by CSU and EPA to ensure its accuracy,
                       and has been offered to the major automotive manufacturers to review the classification
                       and reporting of their respective vehicles.


                            The tampering survey included only 1975 and newer light-duty cars and trucks
                       fueled with gasoline, as well as gasoline powered heavy-duty trucks at selected sites. For
                       the purposes of the tampering surveys, a vehicle is considered to be "unleaded" if a dash
                       label, tank label, or filler inlet restrictor is observed at the time of the inspection, or if
                       the emission control  label indicates an unleaded fuel requirement (i.e., catalyst-
                       equipped). A vehicle's designation as "unleaded" or "leaded" may be changed upon
                       subsequent review of the data. Fuel switching rates are thus based only on the population
                       of unleaded vehicles surveyed. Similarly, tampering rates for specific components are
                       based only on the vehicles originally equipped with the component.

-------
Jlolor Vehicle Tampering Survey -1989                                                      Page 7
site Descriptions
                          The inspections were performed with the consent of the vehicle owners at either
                     roadside pullovers or inspection stations.  The survey was designed to minimize the
                     refusal rate of potential survey participants. A high refusal rate increases the uncertainty
                     in analyzing the data gathered, since individuals who have tampered with or misfueled
                     their vehicles are less likely to allow their vehicles to be surveyed. The overall refusal
                     rate was very low (3%), however, and only one survey site had a refusal rate over 5%. A
                     brief description of each survey site follows. Unless otherwise noted, the survey locations
                     within a given city were changed daily.
                                                     San Diego, CA
                                                       I/M + ATP
                                           Dates:             March 13-17,1989
                                           Vehicles Surveyed:          500
                                           Fuel Samples:               399
                                           Refusal Rate:                1%

                                                    Bakersfield, CA
                                                       I/M + ATP
                                           Dates:             March 20-24, 1989
                                           Vehicles Surveyed:          520
                                           Fuel Samples:               426
                                           Refusal Rate:                1%
                          Roadside pullovers in San Diego and Bakersfield were conducted with the assis-
                     tance of the California Highway Patrol, California Air Resources Board, and the Bureau
                     of Automotive Repair.  California's decentralized, biennial inspection program covers
                     all components on all vehicles, but does not require converter replacement if the inlet
                     restrictor fails.

                                                      Phoenix, AZ
                                                       I/M + ATP

                                           Dates:             March 27-31,1989
                                           Vehicles Surveyed:          465
                                           Fuel Samples:               401
                                           Refusal Rate:                0%
                          The Phoenix survey was conducted at five of the city's centralized inspection
                     stations with the assistance of the Arizona Vehicle Emissions staff. The antitampering
                     program in Phoenix includes a Plumbtesmo test and inspection of the catalytic converte r,
                     air injection system, and inlet restrictor on all 1975 and newer vehicles.

-------
Motor Vehicle Tampering Survey -1989                                                    Page 8
                                                Suburbs of Dallas, TX
                                                        Non-I/M
                                           Dates:             April 10-14,1989
                                           Vehicles Surveyed:          422
                                           Fuel Samples:              268
                                           Refusal Rate:               3%
                           The Dallas survey was conducted in four of the outlying counties which surround
                      Dallas-Fort Worth (Ellis, Kaufman, Collins, and Denton counties). As of April 1,1990,
                      these four counties, as well as three others, will adopt the antitampering inspection
                      program currently in place in Dallas and Tarrant counties.  At that time Dallas and
                      Tarrant counties will add an I/M tailpipe test to their control program.  The Texas
                      Department of Public Safety provided assistance with the roadside pullovers.

                                                   Baton Rouge, LA
                                                         ATP-only

                                            Dates:             April 17-21, 1989
                                            Vehicles Surveyed:          497
                                            Fuel Samples:              398
                                            Refusal  Rate:              3%
                           Roadside pullovers were conducted with the assistance of the Louisiana State
                      Police, Louisiana Motor Vehicle Police, and the Louisiana Department of Environmen-
                      tal Quality. Survey locations were nearly the same as in the 1985,1986, and 1988 surveys.
                      The decentralized antitampering program in Baton Rouge includes a Plumbtesmo test
                      and inspection of all emission control devices on 1980 and newer vehicles.
                                                   Jacksonville, FL
                                                         Non-I/M
                                           Dates:             April 24-28, 1989
                                           Vehicles Surveyed:          580
                                           Fuel Samples:              461
                                           Refusal Rate:               1%
                           The Florida State Police and the Jacksonville Department of Health, Welfare, and
                      Bioenvironmental Services assisted with the roadside pullovers.

-------
Motor Vehicle Tampering Survey -1989                                                    Page 9
                                                    Detroit, MI
                                                       I/M-only
                                          Dates:              May 8-12, 1989
                                          Vehicles Surveyed:          441
                                          Fuel Samples:               297
                                          Refusal Rate:               5%
                          The Michigan State  Police and Michigan Department of Natural Resources
                     assisted with the roadside pullovers. Detroit's decentralized, annual I/M program covers
                     1980 and newer vehicles.
                                                  Cincinnati, OH
                                                      ATP-only
                                          Dates:              May 15-19, 1989
                                          Vehicles Surveyed:          390
                                          Fuel Samples:              310
                                          Refusal Rate:               5%

                                                   Cleveland, OH
                                                      ATP-only
                                          Dates:              May 22-26, 1989
                                          Vehicles Surveyed:          368
                                          Fuel Samples:              321
                                          Refusal Rate:               5%
                          Roadside pullovers were conducted with the help of local law enforcement officers
                     in the municipalities in the Cleveland and Cincinnati areas, as well as State and local
                     pollution control personnel.  The annual, decentralized antitampering program in
                     Cleveland and Cincinnati includes a Plumbtesmo test and an inspection of all emission
                     control devices on 1980 and newer vehicles.

-------
Motor Vehicle Tampering Survey -1989                                                   Page 10
                                                   Springfield, MO
                                                        Non-I/M
                                           Dates:               June 5-9, 1989
                                           Vehicles Surveyed:         540
                                           Fuel Samples:              483
                                           Refusal Rate:               5%

                           Roadside pullovers were conducted with the assistance of the Springfield Police
                      Department and the Greene County Air Pollution Control Agency.  Seventeen of the
                      540 vehicles surveyed were heavy-duty trucks.

                                                    Milwaukee, WI
                                                        I/M-only
                                            Dates:             June 12-16, 1989
                                            Vehicles Surveyed:         490
                                            Fuel Samples:             336
                                            Refusal Rate:              0%
                           The Milwaukee survey was conducted at five centralized I/M stations in Mil-
                      waukee (three  stations), Racine, and Waukesha. The Wisconsin Department of
                      Transportation assisted with the survey. In April 1989, Milwaukee added an antitamper-
                      ing inspection to the existing I/M program, but the area will still be considered I/M-only
                      for this survey because the entire vehicle fleet had not been subjected to the ATP at the
                      time  of the survey.  Two of the 490 vehicles surveyed were  heavy-duty trucks.
                                                    Missoula, MT
                                                        Non-I/M
                                           Dates:             June 19-23,1989
                                           Vehicles Surveyed:          500
                                           Fuel Samples:               404
                                           Refusal Rate:               2%
                           Roadside pullovers were conducted with the assistance of the Missoula Police
                      Department and the Missoula Health Department.  Forty-two of the 500 vehicles
                      surveyed were heavy-duty trucks.

-------
>tor Vehicle Tampering Survey -1989                                                    Page 11
                                                  Anchorage, AK
                                                      l/M + ATP

                                         Dates:              July 10-14, 1989
                                         Vehicles Surveyed:         503
                                         Fuel Samples:              372
                                         Refusal Rate:               2%
                         The roadside pullovers in Anchorage were conducted with the assistance of the
                   Anchorage Municipal Police and the Alaska Department of Environmental Conserva-
                   tion. The decentralized program includes an annual Flumbtesmo test and inspection of
                   the catalyst, inlet restrictor, air injection system, and PCV system on the last 15 model
                   years. Twenty-six of the 503 vehicles surveyed were heavy-duty trucks.


                                                  Fairbanks, AK
                                                     I/M + ATP

                                         Dates:              July 17-21, 1989
                                         Vehicles Surveyed:         436
                                         Fuel Samples:               350
                                         Refusal Rate:               9%
                         The Fairbanks survey was conducted with the assistance of the Alaska State Police,
                   the Alaska DEC, and the Fairbanks North Star Borough, The survey was conducted at
                   a privately-owned inspection facility on one day, and roadside pullovers were used the
                   other four days. The decentralized program includes an annual Plumbtesmo test and
                   inspection of the catalyst, air injection system, and PCV system on the last 15 model
                   years. Twenty-three of the 436 vehicles surveyed were heavy-duty trucks.  The higher
                   refusal rate in Fairbanks is believed to be more indicative of the independent nature of
                   its citizens than of any desire to avoid the survey and conceal instances of tampering.


                                                    Denver, CO
                                                     I/M + ATP

                                         Dates:             August 7-11, 1989
                                         Vehicles Surveyed:           465
                                         Fuel Samples:                330
                                         Refusal Rate:                1%
                        The roadside pullovers were conducted in Denver, Littleton, Aurora, Colorado
                  -Springs, and Thornton with the assistance of the Colorado Department of Health and
                   the local municipal law enforcement agencies.  Colorado's decentralized program en-
                   compasses nine counties and includes an annual inspection of the catalyst, inlet restric-
                   tor, air injection system, and oxygen sensor on 1975 and newer vehicles. Sixteen of the
                   465 vehicles surveyed were heavy-duty trucks.

-------
Motor Vehicle Tampering Survey -1989
                    Page 12
RESULTS
A. VEHICLE TAMPERING
L. Site and Acereffate Totals
                             The vehicles surveyed have been classified into four categories established by
                       previous surveys: tampered, arguably tampered, malfunctioning, and okay. Each vehicle
                       was classified by the worst state of any component in the vehicle. For example, a vehicle
                       would be classified as "tampered" if any one component had been tampered, even if all
                       other components were functioning properly.  A vehicle classified as "okay" must have
                       all observed components functioning properly5. The criteria used for component clas-
                       sification are presented in Appendix B. This overall tampering rate is useful only as a
                       rough indicator of the emissions impact of a tampering problem, since the different
                       components making up the rate may have widely varying emissions implications.

                             The proportion of inspected vehicles with at least one tampered component was
                       18%. One-third of the surveyed vehicles displayed some form of malfunction, arguable
                       tampering, or clear tampering of emission control components. The specific distribution
                       of surveyed vehicles among these categories is depicted in Figure 3.

                             The frequency distribution of tampering instances for those vehicles classified as
                       "tampered" is also shown in Figure 3. Forty-three percent of the tampered vehicles had
                       multiple components tampered, and 12% had four or more instances of tampering.
                          Okay (67%)
                                               Malfunctioning (4%)
                                                    Arguably Tampered
                                                         (11%)   .,.-••••"
                                                            Tampered (18%)
             Four or more (12%)
           j- Three (10%)

            -Two (21%)
                                                                                    One (57%)
                                        Condition of
                                     Surveyed Vehicles
Number of Tampered
   Components
                           Figure 3. Breakdown of surveyed vehicles by condition and extent of tampering.
                5   An 'okay* vehicle, however, may ctill be classified ax fuel switched (see section B.I., Fuel Switrhinf Indir
                    and Overlap of this report).

-------
Motor Vehicle Tampering Survey -1989
Page 13
TABLE 1
1989 Tampering Survey Summary - Light-Duty Vehicles only
Number of Tampering Misfueling
Survey Location
San Diego, CA
Bakersfield, CA
Phoenix, AZ
Dallas (Suburbs), TX
Baton Rouge, LA
Jacksonville, PL
Detroit, MI
Cincinnati, OH
Cleveland, OH
Springfield, MO
Milwaukee, WI
Missoula, MT
Anchorage, AK
Fairbanks, AK
Denver, CO
OVERALL
Vehicles
500
520
465
422
497
580
441
390
368
523
488
458
477
413
449
6,991
Rate (%) Rate (%)
11
16
25
30
19
17
14
16
21
25
15
21
9
14
14
18
*R = roadside pullovers, C = centralized
D = decentralized stations
4
8
5
11
7
4
2
5
5
9
2
8
2
7
3
6
Survey
Type*
R
R
C
R
R
R
R
R
R
R
C
R
R
R,D
R
-
Refusal
Rate (%)
1
1
0
3
3
1
5
5
5
5
0
2
2
9
1
3
inspection stations,
                             Table 1 summarizes the 1989 survey data by site for the light-duty vehicles surveyed.
                       As in previous surveys, the overall tampering in  1989 varied considerably from site to
                       site. This can be attributed to the variety of program configurations among the cities
                       surveyed and to geographic differences.

                             Table 1 also contains the refusal rate at each survey site. The overall refusal rate
                       for the survey was very low (3%), and only one survey site had a refusal rate exceeding
                       5%. The higher refusal rate in Fairbanks (9%) is believed to be more indicative of the
                       independent nature of its citizenry than of any desire by these citizens to avoid the survey
                       and conceal instances of tampering.

-------
Motor Vehicle Tampering Survey -1989
                                                        Page u
               2. TVivt nf Tampering
                             The tampering rates for all emission control components and systems surveyed in
                       1989 are presented in Table 2. The component-specific tampering rates for the 1989
                       survey are presented by survey site in Table 3. Only those vehicles originally equipped
                       with a particular component are considered when computing the tampering rate for that
                       component. The site-specific tampering rates listed in Tables 1 and 3 are for all light-duty
                       vehicles surveyed at each location, including out-of-state vehicles and other vehicles
                       possibly not covered by any local inspection program. The tampering rates for only
                       vehicles covered by local control programs will be examined later in this report.
                                                        TABLE 2
                                        Prevalence of Tampering by Component
                                   Component/System                  Tampering Rate (%)
Catalytic Converter
Inlet Restrictor
Air Pump System
  Air Pump Belt
  Air Pump/Valve
Aspirator*
PCV System
Evaporative System
EGR System
Heated Air Intake
Oxygen Sensor and
Computer System
                                                                                4
                                                                                5
                                                                                10
                                                                                6
                                                                                7
                                                                                1
                                                                                5
                                                                                6
                                                                                6
                                                                                3
                                                                                1
                                      * Vehicles with aspirated air systems are not equipped with
                                       other listed air-injection components, nor do conventional
                                       systems include aspirators.
                              Table 2 shows that air pump system tampering remains the single most tampered
                         component (10% of equipped vehicles). Tampering with evaporative systems also shows
                         no evidence of subsiding. Also of interest is the slight but measurable tampering found
                         with the oxygen sensor and computer control system.  Because of the difficulty in
                         inspecting the oxygen sensors (no determination could be made on 61 vehicles), the one
                         percent rate reported is a conservative estimate. Oxygen sensor tampering is of consid-
                         erable interest because of the dramatic emissions increases which result from its disable-
                         ment -- an average increase of 445% for HC and 1,242% for CO.
                                 id Antt.Mkfugling Pf
                  itn Rrrtiirr ln-TV Frni"'""*fmm Mntnr Vehicles.
                     EPA-AA-TTS-83-10, December 31,1983,

-------
Motor Vehicle Tampering Survey -1989
Page 15
                            Table 3 shows the wide variation in tampering from site to site for any given
                      component. Air pump system tampering, for example, ranged from 4% in Anchorage
                      to 19% in Missoula. This range U partly due to the effectiveness of I/M and antitamper-
                      ing programs (as will be discussed later  in this report), geographic location, and
                      socioeconomic background.
TABLE 3
Component-Specific Tampering Rates (%) by Survey Location*
Survey Catalytic
Location Converter
San Diego, CA
Bakersfield, CA
Phoenix, AZ
Dallas (Subs.), TX
Baton Rouge, LA
Jacksonville, TN
Detroit, MI
Cincinnati, OH
Cleveland, OH
Springfield, MO
Milwaukee, WI
Missoula, MT
Anchorage, AK
Fairbanks, AK
Denver, CO
OVERALL
1
1
1
13
5
6
3
4
5
11
0
8
2
1
2
4
Inlet
Restricter
4
8
4
10
6
4
2
5
4
9
2
9
2
5
3
5
Air Pump
System
6
6
10
18
15
9
7
12
8
15
5
19
4
7
6
10
PCV
System
2
5
11
11
6
4
5
4
6
5
4
4
3
2
7
5
* Indudes any out-of-state vehicles or non-covered light-duty vehicles
applicable).
EGR
System
2
3
6
11
8
4
3
8
8
11
7
10
5
5
5
6
Evap.
System
2
4
10
11
7
6
5
6
7
7
3
6
3
3
5
6
surveyed within the
Any
Component
11
16
25
30
19
17
14
16
21
25
15
21
9
14
14
18
program area (if

-------
Motor Vehicle Tampering Survey -1989
Page 16
                             Table 4 shows the condition of the gas tank cap among surveyed light-duty vehicles
                       by survey location. Although a missing or malfunctioning tank cap is not classified as
                       tampering, the emissions impact of a missing cap nonetheless is substantial. A missing
                       or malfunctioning gas cap can completely negate an otherwise properly functioning
                       evaporative system, causing  large increases in evaporative HC emissions.  The mal-
                       functioning rate among non-stock gas tank caps was considerably higher than among
                       stock gas tank caps (5% vs. < 1%).

Condition
Survey Location
San Diego, CA
Bakersfield, CA
Phoenix, AZ
Dallas (Suburbs), TX
Baton Rouge, LA
Jacksonville, FL
Detroit, MI
Cincinnati, OH
Cleveland, OH
Springfield, MO
Milwaukee, WI
Missoula, MT
Anchorage, AK
Fairbanks, AK
Denver, CO
OVERALL
TABLE
4
of Gasoline Tank Cap by Survey
Malfunctioning
Tank Cap (%)
0
0
1
1
2
1
1
0
1
1
1
1
0
1
1
1

Site
Missing Tank Missing or
Cap (%) Malfunctioning (%)
1
0
3
2
1
2
3
3
1
1
0
1
1
1
1
2
2
1.
4
3
3
3
4
3
3
2
1
2
1
2
2
2

-------
Motor Vehicle Tampering Survey -1989
Page 17
3. VphiHit Oiarartpfi
-------
Motor Vehicle Tampering Survey -1989                                                      Page 18
                            A number of factors might explain the difference in tampering prevalence among
                       manufacturers. Differences in design may make some vehicles more tamper-prone than
                       others. Changing market share history results in different age distributions for vehicles
                       of different makes, and vehicle age is clearly related to tampering prevalence. Tampering
                       rates seem to vary with geographic  location and socioeconomic background, so  the
                       owner demographics for different makes may affect the likelihood of tampering.

                            Vehicle Typ<>-  The overall tampering prevalence for light-duty trucks (LDTs)
                       was 18%, compared to 17% for automobiles (LDVs). Since the 1986 tampering survey,
                       the tampering rates for cars and trucks have basically been the same, and the discrepancy
                       between car and truck tampering decreased steadily in surveys prior to 1986. While the
                       overall tampering rate is less meaningful than the rates for individual components, even
                       the component-specific truck rates tend to converge upon the usually lower automobile
                       rates. Catalytic converter tampering, for example, was only slightly higher for trucks
                       than for cars (5% vs. 4%).

                            Many factors may be contributing to the convergence in car and truck tampering.
                       Since most of the vehicles surveyed in 1989 were subject to control programs, the relative
                       car and truck tampering rates may reflect increased emphasis on truck tampering
                       detection by the inspection programs.  Also, the proportion of imported trucks within
                       the total truck population surveyed has increased from 15% to 26% between 1982 and
                       1989, with many of the smaller trucks being ntiKreA as cars. Since imported vehicles are
                       tampered with much less frequently than domestic vehicles, the increase in imported
                       trucks within the truck  population surveyed may be contributing to the lower truck
                       tampering prevalence.

                            Heavy-duty gasoline-powered tracks (HDTs) were also surveyed at selected sites
                       in the 1989 survey. Of the 126 HDTs surveyed, nearly one-third (32%) had one or more
                       components tampered.  This high tampering rate is largely attributable to the skewed
                       model year distribution of the  HDTs surveyed, since 60% of the HDTs surveyed were
                       manufactured in 1979 or earlier. By comparison, the overall tampering rate for 1979 and
                       older light-duty trucks was even higher (68%). This suggests that vehicle age has a much
                       greater effect on tampering prevalence than vehicle type.  Heavy-duty tampering may
                       increase in significance as the prevalence of catalyst-equipped HDTs increases.

-------
Motor Vehicle Tampering Survey -1989                                                                                   Page 19
                                                             Table 5
           Tampering Percentage (and Sample Size) by Model Year and Vehicle Age at Time of Survey

                                                         Year of Vehicle LJJe

Model
Year   1&     2nd     2ld     M     5lh     6th     2th,      Slh     2th     10th    llth    12th    Ufrh     14th    15th
1989  0(444)
1988  1(524)  1(794)
1987  1(627)  1(832)    3(741)
1986  1(757)  2(1058)   3(915)   4(758)
1985  2(816)  3(1130)   3(972)   6(855)   7(779)
1984  1(462)  2(1001)  5(1018)   7(908)   8(765)   13(681)
1983  7(182)  4(471)    6(710)   7(706)   10(633)  15(532)  12(414)
1982  1(250)  4(226)    7(466)   9(621)   11(574)  19(525)  21(506)  25(414)
1981   2(57)   7(448)   13(206)   15(458)  11(607)  19(560)  26(501)  34(479)  28(384)
1980          5(63)    9(454)   15(211)  18(516)  25(564)  25(556)  28(472)  32(419)  37(351)
1979  6(371)           9(59)   18(477)  31(288)  28(503)  37(673)  36(699)  42(555)  44(464)  46(416)
1978  7(298)  14(502)           15(79)   21(430)  39(238)  34(559)  37(562)  50(548)  49(443)  55(404)   54(318)
1977          10(457)  15(476)           21(66)   26(316)  44(190)  41(408)  48(452)  48(465)  49(356)   54(265)  57(250)
1976                  18(395)   19(374)          29(52)   26(317)  40(171)  39(385)  49(369)  53(318)   60(219)  61(179)   56(162)
1975                           22(274)  22(271)           32(22)   37(183)   55(89)  46(197)  54(194)   60(198)  65(103)   62(113)   66(83)
1974                                   33(276)  27(242)
1973                                           32(253)  36(251)
                                                             Table 6
     Percentage of Catalyst Removal (and Sample Size) among Catalyst-Equipped Vehicles by Model Year
                                             and Vehicle Age at Time of Survey

                                                          Year of Vehicle Life
Model
       l5l2nsi2ld4jh.5jh6jh.2thaUl2tli    IQth     llth    12th    13th    14th     15jh
1989  0(444)
1988  0(521)  0(794)
1987  0(627)  0(810)    0(741)
1986  0(757)  0(1058)   0(887)   0(758)
1985  0(808)  1(1128)   0(972)   1(818)   1(779)
1984  0(462)  0(978)   1(1018)   1(908)   1(726)   1(680)
1983  1(179)  2(471)    0(686)   1(706)   1(633)   2(505)   2(414)
1982  0(250)  1(225)    2(465)   2(597)   1(574)   3(525)   3(466)   5(414)
1981  0(57)   -2(441)    5(204)   6(457)   3(567)   5(552)   7(499) -9(421)   7(378)         .
1980         2(61)    2(428)   3(200)   6(487)   6(522)   7(528)   8(440)   9(346)   9(332)
1979  0(326)           4(55)    6(4»)   12(252)   10(455)  12(572)  12(638)  13(505)  13(376)  13(383)
1978  0(291)  0(445)            0(71)   4(362)   8(213)   8(486)  10(472)  10(466)  15(394)  13(312)  17(277)
1977         1(417)    1(417)            2(59)    2(271)   11(166)  14(357)  17(379)  19(409)  11(320)  19(212)   17(221)
1976                  2(377)   2(305)            10(48)   6(257)  12(139)  12(314)  15(291)  20(276)  19(184)   15(137)  24(135)
1975                          2(242)   2(204)           26(19)  12(139)  23(75)  16(174)  21(130)  23(167)   25(77)   14(72)   30(63)

-------
Motor Vehicle Tampering Survey -1989                                                        Page 20
                             Vehicle. Am>.  Table 5 relates  vehicle age and model year with tampering
                        prevalence for the 1978-1989 surveys.  Catalytic converter removal rates are similarly
                        related to vehicle age and model year in Table 6. The results from any given survey are
                        entered diagonally in each table.

                             The results in Tables 5 and 6 indicate that vehicle tampering increases directly with
                        vehicle age. Examining Table 5 diagonally (by survey) shows that tampering increases
                        consistently with vehicle age in each survey conducted. In the 1989 survey, for example,
                        the tampering rate increases from 0% for first year (1989) vehicles to 66% among the
                        1975 model year vehicles surveyed.

                             Table 6 shows a similar, though less pronounced, increase in catalyst removal. The
                        catalyst tampering rate remains negligible for the first six years of a vehicle's life, and
                        then increases thereafter. This delay in catalyst tampering is understandable, since the
                        emission control components on all new vehicles are warranted for 5 years/50,000 miles
                        by the manufacturer, providing an incentive to maintain the catalysts on vehicles still
                        under warranty.  A simitar delay in overall tampering is less apparent in Table 5, but
                        becomes more evident when only non-I/M vehicles are examined (Figure 2 (a)), Examin-
                        ing these tables in this manner has the advantage of comparing data collected during one
                        survey in one set of locations, but ignores the possible effects of model year differences
                        (i.e., technology), geography, or program coverage on tampering.

                             Two additional ways of analyzing Tables 5 and 6 address the impact of model year
                        on tampering rates.  Analyzing these Tables horizontally (holding the model year con-
                        stant) provides a look at the tampering rates over time for the vehicles of a particular
                        model year. This approach shows the same distinct increase in tampering with vehicle
                        age for all model years since 1975. (The 1974 and 1973 data sets are too small to permit
                        any conclusions.)

                             Tables 5 and 6 can also be examined vertically (holding vehicle age constant),
                        which provides a look at the tampering rates for different model year vehicles of the same
                        age. This approach permits a crude examination of trends  in tampering adjusted for
                        vehicle age.  Examining Table 5 vertically, for example, shows a decrease in overall
                        tampering from the 1978 survey to the 1989 survey for vehicles in each of their first seven
                        years of use.  This decrease in overall tampering appears to lessen in the later years of a
                        vehicle's life.  A similar, though less pronounced, trend is apparent in the age-specific
                        catalyst tampering data (Table 6).

-------
Motor Vehicle Tampering Survey -1989                                                       Page 21
                            The comparisons made in the two preceding paragraphs involve observations
                       made in different survey locations during different years, and are affected by the site
                       selection in each survey. An increasing percentage of the 1981 and newer vehicles in
                       recent surveys have been covered by antitampering programs, which could contribute to
                       the lower tampering found among newer vehicles. The impact of improved technology
                       on tampering can be more accurately assessed by examining only vehicles from non-I/M
                       areas (Figure 2(a)). The data in Figure 2(a) suggest that tampering with new technology
                       vehicles that are no longer covered by new car or emissions warranties (i.e., post-1980
                       vehicles that are six to nine years old)  is consistent with tampering among older
                       technology vehicles (pre-1981 - over nine years old).
4. Tpinart1 rtf T/\^ and Anfifaningriiiiy
                            Inspection and maintenance (I/M) programs require vehicles to meet specific idle
                       emission standards. Vehicles registered in areas with these programs are required to be
                       periodically tested to assure that they comply with the specific idle emission outpoints
                       established by these jurisdictions. In addition to reducing emission levels by stimulating
                       better owner maintenance, I/M programs may deter some tampering with emission
                       control components. Data from previous surveys has tended to support this proposition,
                       since tampering in I/M areas has historically been lower than in non-I/M areas.

                            Many local areas have also instituted antitampering programs  (ATPs), which
                       involve periodic vehicle inspections to check the integrity of specific emission control
                       components. Antitampering programs vary greatly in the components inspected and the
                       vehicle model years covered, so that a vehicle or component which would be inspected
                       b one program area might not be inspected in a different program area.  Successful
                       antitampering programs should reduce existing tampering and deter future tampering
                       with the components and model years covered by the program.

                            Classifying the survey areas into program types is made difficult by the consider-
                       able variation in control programs. For example, three antitampering programs sur-
                       veyed (Cleveland, Cincinnati, and Baton Rouge) include inspection  of all emission
                       control devices, but only on 1980 and newer vehicles. In contrast, California's inspection
                       program covers all components on all vehicles.

                            In order to minimizr. the impact of these variations in program types and model
                       year coverages, the vehicles in the 1989 survey can be classified by the model year
                       restrictions applicable in the program areas surveyed. These classifications, and the
                       resulting tampering rates for each program  type, are shown in Figure 5. Since
                       Cleveland's antitampering program only covers 1980 and newer vehicles, for example,
                       only those  model year vehicles are  classified as "ATP-only"; the 1975-1979 vehicles
                       surveyed in Cleveland are classified as "non-I/M",  Additionally, 86  of the vehicles
                       surveyed in the Dallas suburbs were covered by the Dallas inspection program. Those
                       vehicles have been classified as "ATP-only", while the balance of the vehicles have been
                       classified as "non-I/M*.  This method of classification does not take into consideration
                       the components inspected by each program, a program's effectiveness, or the frequency
                       of inspection. Any comparisons between program types (e^, I/M + ATP vs. non-I/M)
                       should thus be made carefully.

-------
Motor Vehicle Tampering Survey -1989
                                                         Page 22
                                                                          I/M + ATP
Dallaa (Suburb*), TX*
Jackaonvait, FL
SprtngflaW.MO
Mluoula,MT
Datrott, Ml (1975-79)
davatand, OH (1975-79)
Baton Houga. LA (1875-79)
andnrad. OH (1975-79)

ATP-onfy
Bttonflaug*.LA(19M-*>)
OtMtand.OH(1«W4)
Cincinnati. OH (1980+)
                                                                                d CA
                                                                                 2
                                                                          Anchong*,AK
                                                                          Dan*. Ml (1980+)
                                                                          M!hnukM,WI
                                 Noto: Pregrameov«na«inocWyMntufv«y«dunl*iaaD^anriMlndlcitad.
                                 •86 wahlclM wowy*! In tna DaJlac submba wv* ooMrad by tf» OaUu ATP,
                                  and haw bMn daatfflad'ATP-ony.
                                   29%
                                                                                               10%
                                                                             Futf SwttoWng (%)
                          Rgura S  CI«Mlf IcatJon of 1989 survey site* by program lyp« (and model
                                    year coverage), wtth the reeuttlng tempering end fuel ewttchlng
                                    ratee for eech program type. (The fuel ewttohlng ratee era derived
                                    from e slightly different clasarficetlon scheme - eee the text of
                                    this report for details.)

-------
Motor Vehicle Tampering Survey -1989
                                           Page 23
                            Two alterations were made in this classification system for calculating the fuel
                       switching rates.  The vehicles covered by the I/M +  ATP in Fairbanks have  been
                       excluded because the program includes a Plumbtesmo test but does not inspect inlet
                       rcstrictors.  Additionally, the- vehicles surveyed in San Diego  and Bakersfield  were
                       classified as "I/M-only"  instead of T/M + ATP" because convener replacement is not
                       required by the programs in the event of an inlet restrictor failure (hence, EPA does not
                       grant credit for the inlet inspection).

                            Another factor affecting comparisons in Figure 5 is the overreprcsentation in the
                       survey sample of areas with some kind of control program. This is a deliberate feature
                       of the survey's design, as was discussed earlier, and focuses attention on evaluating the
                       effectiveness of specific individual programs. Table 7 compares the proportion of the
                       1989 survey sample and the national vehicle fleet under control programs. Because of
                       the need to evaluate control program effectiveness, the survey overrepresents the portion
                       of the national vehicle fleet under control programs. Vehicles covered by an I/M +  ATP
                       or an ATP-only comprised 50% of the survey sample, while only approximately 30% of
                       the national vehicle fleet were under such programs.
                                                       TABLE?
                                         Comparison of 1989 Survey Sample
                                          to Actual Nationwide Vehicle Fleet
                               Program Type
 Percentage within
Survey Sample (%)
Approximate Percentage
of Nationwide Fleet (%)*
                               non-I/M
                               I/M-only
                               ATP-only
                               I/M -l- ATP
       38
       12
       12
       38
                             1 From County Poulation Estimates: Jur 1,
          58
          12

          10
          20

     1987. and 1986. U.S.
                              Department of Commerce - Bureau of the Census, August, 1989.

-------
Motor Vehicle Tampering Survey -1989                                                       Paqe 24
5. Tampering Trends for Selected
                            The impact  of I/M and antitampering programs in specific locations can be
                       examined by comparing the 1989 survey data with that from earlier surveys.  Com-
                       parisons made between surveys widely spaced in time, however, must take into con-
                       sideration the differences in average vehicle age in each survey. The average odometer
                       reading per vehicle surveyed in 1989, for example, is over 20% greater than it was in 1984.
                       Since vehicle mileage correlates with vehicle age, which is directly related to tampering
                       prevalence,  a significant increase in tampering might be  expected to have occurred
                       between 1984 and 1989, even if all other factors remained constant (car/truck distribu-
                       tion, owner demographics, etc.).  Inferences regarding program effects should thus be
                       made with this in mind.

                            Tables 8 and 9 present tampering data for the three ATP-only and six I/M + ATP
                       sites surveyed, respectively. Because of the large number  of programs represented in
                       Tables 8 and 9, only a few representative comparisons will be discussed in this report.
                       Detailed analysis of each program's effectiveness is beyond the scope of this report. The
                       comparisons made in these Tables have been limited to the specific components and
                       vehicle model years covered by each antitampering program actually in operation during
                       the 1989 survey.  In some instances this  reduces the sample size substantially. The
                       tampering data listed in Tables 8 and 9 were compiled only for surveyed vehicles included
                       within the local program jurisdiction. The Fairbanks tampering rates listed, for example,
                       are for vehicles registered b the Fairbanks North Star Borough only. Any vehicles
                       surveyed from outside of the Borough were excluded from this analysis.

                            The data in Table g suggest that the ATPs in Baton Rouge, Cleveland, and
                       Cincinnati have helped lower catalyst tampering and misrueling rates (i.e., plumbtesmo,
                       inlet restrictor).  Tampering with underhood components, however, has not declined
                       appreciably.  In fact, evaporative and air pump system tampering in each city increased
                       from earlier surveys.  A similar pattern  is apparent in Table 9,  where I/M + ATP
                       programs have been generally effective in reducing catalyst tampering and misfueling,
                       but much less so in reducing underhood tampering.

                            Anchorage is the notable exception in Table 9. Air pump system tampering among
                       covered vehicles in Anchorage is one of the lowest rates ever observed. The low rates
                       in Anchorage are particularly impressive because that program is decentralized, and
                       such programs are historically less effective than centralized programs.

                            The difference in programmatic impact on catalyst and fuel-related tampering
                       versus the underhood components is illustrated by Figures 6(a)-(Q, where tampering
                       rates by model year for the major components covered by inspection programs (either
                       ATP-only or I/M + ATP) are compared to tampering rates for the same components on
                       vehicles that were not  covered by inspection programs.   Catalyst and inlet restrictor
                       tampering (Figs. 6 (a) and (b), respectively) were much lower on vehicles covered by
                       converter and inlet inspection programs than on vehicles not covered by such inspec-
                       tions.  Air pump system tampering (Fig. 6(c)) was also lower in areas with inspections
                       than in areas without inspections, although it remained relatively high even in areas
                       inspecting that component. Areas inspecting evaporative, PCV, and EGR systems (Figs.
                       6(d), (e), and (0, respectively) had only slightly lower tampering rates.  The PCV and
                       evaporative system rates are noteworthy because these two systems control crankcase
                       and evaporative HC emissions, respectively — emissions which the catalytic converter
                       does not control  The overall rates for the data presented in Figs. 6(a)-(f) are shown in
                       Figure 7.

-------
Motor Vehicle Tampering Survey -1989
Page 25
Table8
Tampering Prevalence among Vehicles and Components
Covered by An ti tampering Programs in ATP-only Areas
Surveyed between 1984 - 1989
Tfrmnerintt Prevalence ( °7a\ bv Survey Year
Survey Location Component and
(Prop-am Start Date^ Model Years Covered
Baton Rouge, LA
(1/86)





Cincinnati, OH
(2/88)





Cleveland, OH
(2/88)





Catalyst
Inlet Restrictor
Pos. Plumbtesmo
PCV System
Evap. System
Air Pump System
EGR System
Catalyst
Inlet Restrictor
Pos. Plumbtesmo
PCV System
Evap. System
Air Pump System
EGR System
Catalyst
Inlet Restrictor
Pos. Plumbtesmo
PCV System
Evap. System
Air Pump System
EGR System
80 +
80 +
80 +
80 +
80 +
80 +
80+
80 +
80 +
80 +
80 +
80 +
80 +
80 +
80+
80 +
80+
80 +
80+
80 +
80 +
* Rates from surveys conducted after the
1984
-
-
-
-
-
-
-
2
3
2
1
1
3
7

-
-
-
-
-
-
start
1985 1286 1987 1988
4 3* - 1
31-2
52-0
23-6
33-3
64-7
44-5

,
.
.
,
.
. -
1 ...
2
1
3
1
2
2 ...
date of an inspection program are given in italics.
12S2
2
2
0
4
4
10
4
1
1
0
2
3
6
5
1
I
0
3
4
3
2


-------
Motor Vehicle Tampering Survey -1989
Page 26



Table 9
Tampering among Vehicles and Components
Covered by An ti tampering Programs in I/M + ATP Areas
Surveyed between 1983 - 1989
Tamnerino Prevalence (%1 bv Survev Year
Survey Location Component and
rPrngram Start Hatel y ndel Years Covered J2S3 19S4 1285 1286 1282 1288
Bakersfield, CA
(1/86)




San Diego, CA
(3/84)




Phoenix, A2
(I/M-1/77,
ATP - 1/87)

Anchorage, AK
(7/85)



Fairbanks, AK
(7/85)


Denver, CO
(1/83)


Catalyst 75+ - 8
Inlet Restrictor 75+ 11
PCV System 75-*- - 3
Evap. System 75+ 4
Air Pump System 75+ 9
EGR System 75+ - 18
Catalyst 75+ -
Inlet Restrictor 75+ -
PCVSystem 75+ ----- -
Evap. System 75+ -
Air Pump System 75+ ------
EGR System 75+ -
Catalyst 75+ 7 . - - - 2
Inlet Restrictor 75+13 - - - - 5
Pos. Plumbtesmo 75+12 - - - - l
Air Pump System 75+ 13 - - -12
Catalyst 75+ -
Inlet Restrictor 75+ -
Pos. Plumbtesmo 75+ -
PCVSystem 75+ ------
Air Pump System 75+ -
Catalyst 75+ ------
Pos. Plumbtesmo 75+ ------
Air Pump System 75+ ------
EGR System 75+ ------
Catalyst 75+ 4
Inlet Restrictor 75+4
Air Pump System 75+6
* Rates from surveys conducted after the start date of an inspection program are given in italics.

,

0282
/*
8
5
4
6
3
1
4
2
2
6
3
1
4
2
10
1
I
0
2
2
1
2
7
2
2
3
4


-------
Motor Vehicle Tampering Survey -1989
   Page 27
      Tampering (%)
          30
          20
          10
     Tampering (%)
          40
          30
          20
          10
           0 '	<-
                                     Mode) Year
                             (a) Catalytic Converter
                                                                       Covered by Program
                                                                         Not Covered by
                                                                           Program
                                                                        Covered by Program
Not Covered by
  Prog ram
                                     Model Year
                                (b) Inlet Restrictor

        Figure 6.  Comparisons of component-specific tampering rates by model
          (a-0    year among vehicles both covered and not covered by ATPs.

-------
Motor Vehicle Tampering Survey -1989
   Page 28
       Tampering (%)
            so
            so
            40
            30
                89   58   87   M   S3   M  S3   02   81   «0   79   78   77  70   75

                                           Model Year

                                   (c) Air Pump System
       Tampering (%)
            10 -
                                                                                 Covered by Program
Not Covered by
  Program
                                                                                 Covered by Program
                                                                                   Not Covered by
                                                                                      Program
                9868B7WS3S48382S1   SO   7»   7S   77   76   75
                                           Model Year

                                  (d) Evaporative System

-------
Motor Vehicle Tampering Survey -1989
                                                    Page 29
       Tampering (%)
            20
            10
                                                                                Covered by Program
                                                 Not Covered by
                                                   Program
                88   M  87
86   M  «   82   «1   80   79   7«   77  76  75
         Modal Year

     (e) PCV System
       Tampering (%)

            40
            30
            20
            10
                                                                                Covered by Program
                                                                                  Not Covered by
                                                                                    Program
                                          Model Year

                                       (f) EGR System

-------
Motor Vehicle Tampering Survey -1989
         Page30
                        Component/System
                        Catalytic Converter
                            Inlet Restrtetor
                         Air Pump System
                        Evaporative System
                             PCV System
                             EGR System
                                                         7%
Component not cvrd.
by ATP Inspection
Component covered
by ATP Inspection
                                                                       13%
                                                 Tampering (%)


                         Figure 7.    Comparison of componant-apsclflc tampering rates among
                                    vahlclM both covarsd and not eovarad by ATPa.

-------
Motor Vehicle Tampering Survey -1989
                                 Page 31
                           Table 10 compares the 1989 tampering rates found in the Dallas suburbs to those
                      found within the Dallas program area in 1987.  This analysis is restricted to the same
                      model year limitations for each component used by the Dallas ATP. The data in Table
                      10 suggest that tampering rates outside of the Dallas program area are considerably
                      higher than those found within the program area.
TABLE 10
Comparison of 1987 Tampering Rates in Dallas (ATP-onLy) to 1989 Rates
In Dallas Suburbs (noo-I/M). Analysis Is Restricted to the Model Years covered
by the Dallas ATP.
Component/System Dallas Dallas Suburbs
and Model Years Covered (1987 Survey) (1989 Survey)
Catalytic Converter
Inlet Restrictor
Pos. Plumbtesmo
PCV System
Evap. System
Air Pump System
EGR System
80 +
80 +
80 +
75 +
75+
75+
75+
2
2
2
6
7
6
4
6
3
2
11
11
18
11
                           The Denver survey was somewhat unique in that four days of that survey were spent
                      in the Denver metropolitan area, while one day was spent surveying vehicles in Colorado
                      Springs (approximately 50 miles south of Denver). Table 11 presents the tampering rates
                      found in Colorado Springs (101 vehicles) vs. those found in the Denver area, and includes
                      only those vehicles covered by the Colorado I/M + ATP. (Both Colorado Springs and
                      Denver are within the program area.) Table 11 suggests that the Colorado inspection
                      program is being more effectively implemented in Denver than in Colorado Springs.
                      Those components not covered by Colorado's inspection program show no difference
                      in tampering between Denver and Colorado Springs.
                                                      TABLE 11
                           Comparison of Tampering and Fuel Switching Rates Surveyed in Colorado
                         Springs(Fourtn Day of Denver Survey) to the Balance of the \ihfcles Surveyed In
                              Denver. Includes Only Vehicles Covered by the Colorado I/M + ATP
                              Component/Indicator
                Tampering Rate (%)
Colorado Springs    Greater Denver
                              Catalytic Converter
                              Inlet Restrictor
                              Air Pump System
                                             Components Inspected by Program
       3
       7
       9
                                           Components Not Inspected by Program
                              PCV System                      5
                              Evap. System                     5
                              EGR System                     11
                              Leaded Fuel in Tank               4
                              Positive Plumbtesmo               1
1
2
2


7
5
2
0
1
                              Overall Tampering
                              Overall Fuel Switching
       19
       7
12
2

-------
Motor Vehicle Tampering Survey -1989
                                                 Page 32
fi. Imnact of Aftermarket P-atalvtic Cnnvertmx
                             The impact of catalytic converter inspection programs can be studied by examining
                       the prevalence of aftennarket converters among the survey vehicles.  Aftermarket
                       converters, by virtue of their low cost and high availability relative to original equipment
                       converters, can often be used to replace missing or damaged converters, particularly in
                       program areas that require converter replacement in those instances. Table 12 compares
                       the prevalence of aftennarket converters vs. mi^ii^g converters among converter-
                       equipped vehicles at each survey location.  In the nine cities with converter inspection
                       programs, only the converter-equipped vehicles subject to the local inspection program
                       were analyzed; in the six cities without converter inspections, all catalyst-equipped
                       vehicles surveyed were analyzed.
                                                         TABLE 12
                               Comparison of Catalyst Removal Rates to the Prevalence
                                      of Aftermarket Catalysts by Survey Location
                          Survey Location
   % of Catalyst-      % of Catalyst-
 Equipped Vehicles Equipped Vehicles  Vehicles Covered
 with Aftermarket     with Missing     by Local Catalyst
    Converters	Converters*    Inspection Program
                          San Diego, CA
                          Bakersfield, CA
                          Phoenix, AZ
                          Anchorage, AK
                          Fairbanks, AK
                          Denver, CO
                          Baton Rouge, LA
                          Cincinnati, OH
                          Cleveland, OH
                           Jacksonville, FL
                           Detroit, MI
                           Dallas (Suburbs), TX
                           Springfield, MO
                           Milwaukee, WI
                           Missoula,MT
                           OVERALL
 Sites with Converter Inspection Programs
         2                 1               1975+
         3                 1               1975 +
         6                '1               1975 +
         2                 1               1975 +
         3                 1               1975 +
         3                 2               1975 +
         1                 2               1980 +
         1                 1               1980+
         2                 1           .    1980 +
Sites without Converter Inspection Programs
         1                 6               None
         1                 3               None
         1                13              None
         1                11              None
         3                 0               None**
         0                 8               None
         2                 4
                         •Percentages listed are for catalyst-equipped vehicles subject to the local inspection program.
                           In areas without converter inspections, the percentage* listed are for all catalyst-equipped
                           vehicles surveyed.

                         "Milwaukee's program does Inspect converters on 1975 and newer vehicles, but this program
                           was added to tfie existing t/M program only two months prior to the survey, so the entire vehicle
                           fleet had not been exposed to the inspection at the time of the survey.

-------
Motor Vehicle Tampering Survey -1989                                                         Page 33
                             The data in Table 12 indicate that  areas with converter  inspection programs
                       generally have a higher percentage of aftermarket converters than missing converters
                       among the covered vehicle fleet.  Phoenix in particular has a large percentage of its
                       vehicle fleet equipped with aftermarket converters (6%). The environmental benefits
                       of retrofitting such a large percentage of the vehicle fleet is considerable. In contrast,
                       the cities with the highest percentages of missing conveners (Dallas suburbs, Springfield,
                       and Missoula) are all non-I/M areas, and also had the lowest percentage of aftermarket
                       converters.  The  only city where this trend does not hold is Milwaukee, where  the
                       converter inspection program had only been added to the existing I/M  program two
                       months prior to the survey. The high percentage of aftermarket converters may be
                       attributed to the  State's efforts to educate the public prior to implementation of the
                       anti-tampering inspection program or the fact that the survey was conducted at  the
                       centralized lanes, where the vehicle owners may have anticipated the inspection and
                       made corrective repairs to their vehicles.
7. Correlation between Tampering and Idle Emissions
                             As was mentioned previously, vehicles which are subject to an I/M program must
                       meet specific idle emissions cutpoints.  To assess the relationship between tampering
                       and fuel switching and idle failure rates, the idle emissions from unleaded vehicles have
                       been tested against the cutpoints established by the I/M program where they were
                       sampled  Vehicles in non-I/M and ATP-only areas were tested against the cutpoints
                       specified by the New Jersey I/M program. The cutpoints for each I/M area are listed in
                       Appendix C.

                             The results of the idle tests are presented in Figure 8 for vehicles in the various
                       tampering and fuel switching categories. The failure rates for fuel switched vehicles that
                       were arguably tampered with or otherwise okay are based upon very small samples and
                       may be unreliable. The results in Figure 8 indicate that a substantially larger proportion
                       of tampered and fuel switched vehicles than of okay vehicles fail an idle test at typical
                       I/M cutpoints. For example, only 19% of the unleaded vehicles surveyed that were free
                       of tampering and fuel switching failed an idle test, while 66% of the tampered and fuel
                       switched vehicles failed that test  It should be noted from Figure 8, however, that 34%
                       of the tampered and fuel switched vehicles were still able to pass an idle test.

-------
Motor Vehicle Tampering Survey -1989
Page 34
                                                    Entire Survey
                                                       Sample
                                                        100%
                                Figure 8. Distribution of unleaded vehicle* eurveyed among
                                         tampering*, mlefuellng, and Idle teet categories.

                                *^udM malfunctioning vthldM (4% of total)

-------
Motor Vehicle Tampering Survey -1989                                                       Page 35
Idle Test Failure Rates


Vehicle Condition
Okay
Arguably Tampered
Tampered
Catalyst Removed
or Misfueled
Table 13
(Percent) by Pollutant and Vehicle Condition
Failure Rate(%) by Pollutant for Model Years Listed
1975 - 80 1981 + Entire Survey ( 1975 + )
HC CO HC CO HC CO HCorCO
32 23 15 6 16 8 19
27 9 26 15 26 20 36
43 40 35 27 40 35 53
49 43 44 36 48 41 63
                            Table 13 shows the percentage of vehicles that failed the idle emissions test for
                       each vehicle condition. The failure rates are listed for the entire survey, as well as in two
                       model year groupings representing "old" technology (1975-1980) and "new" technology
                       (1981 + } vehicles. "New" technology includes closed loop emissions control, which came
                       into widespread usage in 1981 model year vehicles.

                            The overall percentage of tampered vehicles exceeding I/M cutpoints for HC
                       emissions was over twice that for okay vehicles (40% vs. 16%). Over four times as many
                       tampered vehicles exceeded CO cutpoints as did okay vehicles (35% vs. 8%).  The
                       majority (63%) of the vehicles that  either had been fuel switched or had their catalysts
                       removed also exceeded HC or CO limits. Conversely, 37% of the vehicles with missing
                       catalysts or classified as fuel switched were still able to pass an idle emissions test.  A
                       significant number of arguably tampered vehicles also produced excess idle emissions.

                            The effectiveness of idle emissions testing on "new" technology vehicles can also
                       be seen in Table 13. The data in Table 13 actually underestimate the  impact of "new"
                       technology because "old" technology trucks manufactured after 1980 have been included
                       in the "new" technology category due to the model year split As was found in previous
                       surveys, idle emissions testing is more effective in identifying tampering in 1980 and older
                       vehicles than on 1981 and newer vehicles. For example, 40% of the tampered "old"
                       technology vehicles exceeded CO cutpoints compared to 27% of the  tampered "new"
                       technology vehicles. This suggests that idle emissions testing may not  be as effective a
                       strategy for identifying tampering and fuel switching among "new" technology vehicles,
                       since many vehicles with closed loop systems are able to produce low idle emissions even
                       with tampered emission control  devices.

                            The mean idle emissions for tampered and okay vehicles are presented in Table
                       14 by program type. The classification of survey sites is the same as was described in
                       Figure 5, with two alterations. The data from Anchorage and Fairbanks are listed in a
                       separate category (I/M (CO only)) because these programs have no HC cutpoints. The
                       remaining I/M and I/M + ATP sites have been combined into die "I/M" category.

                            The mean idle emissions from tampered vehicles were considerably greater than
                       from untampered vehicles (Table 14). Overall, HC and CO emissions from tampered
                       vehicles were over five times greater on average than from okay vehicles. Untampered
                       vehicles in Anchorage and Fairbanks (which don't have HC cutpoints) had surprisingly

-------
Motor Vehicle Tampering Survey -1989
Page 36
TABLE 14
Mean Idle Emissions for Tampered and Okay Vehicles Within Each
Program 'type
HC emissions (ppm) CO emiviinna ( %)
Program Type Tampered Okay
non-I/M
VM*
I/M (CO-only)"
ATP-only
390
254
232
274
68
64
71
32
Tampered
2.6
1.7
1.2
2.2
Okay
0.5
0.4
0.3
0.2
OVKRALL 325 62 2.2 0.4
* Includes any program where idle emissions are checked, including I/M + ATP
areas.
** Consists of vehicles surveyed
emissions are checked.
in
Anchorage and
Fairbanks, where
only CO
                        low HC and CO emissions, particularly in comparison to the El Paso vehicles surveyed
                        in 1988 (which also were not subject to HC outpoints). The low idle rates found in Alaska
                        are consistent with the low tampering rates found there, and provide further evidence
                        of their program effectiveness.
B. FUEL SWITCHING

1. Fuel Switching Indicators and Overlar
                             Fuel switching is more easily defined than measured, since no single indicator can
                        completely characterize its prevalence. Since 1981 the surveys have used a combination
                        of three indicators to measure fuel switching more accurately: a tampered fuel filler inlet
                        restrictor, a positive Plumbtesmo test for lead deposits in the tailpipe, and a gasoline
                        lead concentration of more than 0.05 gram per gallon (gpg). Of these three indicators,
                        only a tampered inlet restrictor is also considered tampering, and as such is used to
                        determine both tampering and fuel  switching rates.  Since false positive indications
                        should be extremely rare for these measures, the percentage of vehicles with at least one
                        positive indicator is a reasonable minimum estimate of fuel switching.

                             Measuring fuel switching is somewhat unusual because two of the three indicators
                        (positive Plumbtesmo and leaded fuel in the tank) disappear with a series of subsequent
                        proper fuelings. Positive Plumbtesmo results disappear slowly, while leaded gasoline is
                        quickly diluted with subsequent tanks of unleaded fuel A vehicle which has one or both
                        of these transient indicators, however, has recently used or is currently using leaded
                        gasoline. The presence of a tampered inlet restrictor provides no information about the
                        immediacy of the fuel switching. A fuel switching rate derived from the two transient
                        indicators would thus measure how many vehicles have recently used leaded gasoline,
                        while all three indicators together provide a more accurate indication of long-term fuel
                        switching behavior with its associated catalyst destruction and elevated emissions.

-------
Motor Vehicle Tampering Survey -1989                                                        Page 37
                             The presence of any of these three indicators suggests that a given vehicle has been
                       misfueled; their absence, however, does not rule it out. For example, fuel samples could
                       only be obtained from 76% of the unleaded vehicles surveyed, limiting the scope of this
                       variable. A vehicle with an untampered fuel filler inlet restrictor may also have been
                       fueled at a leaded pump equipped with a smaller nozzle, or by using a funnel or similar
                       device. The tailpipe lead test may also fail to identify misrueling, particularly if older
                       vehicles have had their tailpipes replaced since last operated on leaded fuel As the lead
                       phasedown program has lowered lead levels in leaded gasoline, the incidence of false
                       negative Ptumbtesmo results may be increasing. The uncertainty in these measures,
                       then,  is always toward underestimating the number of vehicles misfueled.

                             The limitations of the fuel switching indicators can be seen in their incomplete
                       overlap.  The Venn  diagram  (Figure 9
-------
Motor Vehicle Tampering Survey -1989
              Page 38
                            Positive Plumbtesmo
                               (69 Total)
                                                           (a)
                              Tampered Inlet Only '''
                                 (70%)
     Leaded Fuel in Tank
         (56 Total)
                                                                             Tampered Inlet
                                                                              Rettrictor
                                                                              (283 Total)
Leaded Gaa Only (3%)
                                                                            All Three Indicators
                                                                                  (9%)

                                                                              Inlet (5%)
                                                                   Ptumbtatmo + inlet
                                                               Plumbtaarno + Ga*   (1%)
                                                           (b)
                              Flgura 9(a).  Overlap of fu«t swttchlng Indteator* among
                                           v«hlclM lor which all thr«* Indicators wnr« Inipoctad.

                                    9(b).  Breakdown of mlafu«l«d vahtolM In 9(a) by pr«val«ne«
                                           of each Indicator combination.

-------
Motor Vehide Tampering Survey -1989
Page 39
2-Fuel Switching Trends
                             Of the vehicles requiring unleaded fuel, 6% were identified as misfueled by at least
                        one of the indicators discussed above. The fuel switching incidence by survey site is listed
                        in Table 15. The data in Table 15 again show that inlet restrictor tampering is easily the
                        most prevalent indicator of misfueling.  This is not surprising, since inlet tampering is
                        the most persistent indicator of past misfueling, even when the vehicle has been properly
                        fueled for sometime.
TABLE 15
Fuel Switching Rates among Unleaded Vehicles by Site
Leaded Tampered
Survey Fuel in Inlet
Location Tank(%) Restrictor (%)
San Diego, CA
Bakcrsfield, CA
Phoenix, AZ
Dallas, TX
Baton Rouge, LA
Jacksonville, FL
Detroit, MI
Cincinnati, OH
Cleveland, OH
Springfield, MO
Milwaukee, WI
Missoula,MT
Anchorage, AK
Fairbanks, AK
Denver, CO
OVERALL
Note: The overall fuel (witching rate
of the (lightly different population t
for all three indicator*).
0
0
0
5
1
1
1
1
1
2
0
2
1
1
1
1
4
8
4
10
6
4
2
5
4
9
2
9
2
5
3
5
and Indicator
Positive Overall Fuel
Plumbtesmo Switching
(%) (%)
0
1
2
5
1
1
0
0
1
3
0
2
1
3
1
1
4
8
5
11
7
4
2
5
5
9
2
8
2
7
3
6
may actually be lower thaa the nte for any one indicator because
ne* for each indicator {i*, not all vehicle* were able to be tested

-------
Motor Vehicle Tampering Survey -1989
Page 40
                            Table 16 presents the combined tampering and fuel switching rates for the 1989
                       survey. The percentage of unleaded vehicles that were tampered or fuel switched was
                       18%, and the percentage of unleaded vehicles with missing or. damaged converters was
                       7%.  Table  16 thus suggests that nearly 40% of all tampering  and fuel switching is
                       composed of vehicles in the  catalyst removed or fuel  switched category. Since these
                       conditions have the largest emissions impact, this indicates the very serious nature of
                       much tampering.
TABLE 16
Combined Tampering and Fuel Switching Rates
Catalyst-Equipped Vehicles
Survey with Catalysts Removed or
Location that were Fuel Switched (%)
San Diego, CA
Bakersfield, CA
Phoenix, AZ
Dallas, TX
Baton Rouge, LA
Jacksonville, FL
Detroit, MI
Cincinnati, OH
Cleveland, OH
Springfield, MO
Milwaukee, WI
Missoula,MT
Anchorage, AK
Fairbanks, AK
Denver, CO .
OVERALL
5
9
5
16
8
7
3
7
7
13
2
11
3
8
4
7
Unleaded Vehicles
either Tampered or
Fuel Switched (%)
11
17
25
30
19
16
14
16
22
25
14
21
8
16
14
18

-------
Motor Vehicle Tampering Survey -1989
Page 41
Table 17
Prevalence of Fuel Switching Indicators by Program Type
Percentage by Program Type
Fuel Switching Indicator non-I/M ATP-only I/M-only I/M + ATP
Tampered Inlet Restrictor
Positive Plumbtesmo
Leaded Fuel in Tank
Overall Fuel Switching
9
3
2
9
2
0
0
2
4
0
0
4
3
0
1
3
                           Table 17 presents the prevalence of each fuel switching indicator by program type.
                      The classification of survey sites is the same as was described for Figure 5. Table 17 shows
                      that the prevalence of each fuel switching indicator was considerably greater in the
                      non-I/M areas than in areas with control programs.
                                                         Table 18
                               Prevalence of Fuel Switching Indicators by Vehicle lype

                                                                Percentage of each Indicator
                                Fuel Switching Indicator	LDV	LDT^

                                Tampered Inlet Restrictor            5                 6
                                Positive Plumbtesmo                 1                 2
                                Leaded Fuel in Tank                 1                 2

                                Overall Fuel Switching               5                 7
3.  Fuel Switrhmg hy Vehicle Type
                           The prevalence of each fuel switching indicator by vehicle type is presented in Table
                     18. Overall fuel switching among trucks was slightly higher than for passenger cars (7%
                     vs. 5%) and the prevalence of each indicator was higher in trucks as well  Fuel switching
                     data from the heavy-duty trucks was not included in Table 18 because of an insufficient
                     sample of unleaded HDTs.
4. Fuel Switching and Catalvsf TV
                           Consumers and mechanics remove catalytic converters for a number of reasons, but
                     some of their motivation may be related to fuel switching. The vehicle owner may remove
                     the catalytic converter either prior to misfueling, or after some misfueling if the vehicle's
                     driveability has been adversely affected by a catalyst damaged from the repeated misfuel-
                     ing. The data from this survey cannot be used to distinguish between these two situations,
                     but can be used to examine the extent to which these types of abuse occur in conjunction.

                           Figures 10(a) and (b) examine the relationship between converter tampering and
                     two of the three misfueling indicators (positive Plumbtesmo and tampered inlet restrictor).
                     Vehicles included in Figures 10(a) and (b) had to meet two criteria.  First, only those
                     vehicles that have not been subject to an ATP with Ptumbtesmo testing were included.
                     Second, the sample population was further restricted to those vehicles in which all three
                     parameters were inspected during the survey. The three parameters examined in Figures
                     10(a) and (b) have been incorporated  into a number of antitampering programs  to

-------
Motor Vehicle Tampering Survey -1989
             Page 42
                      determine if a converter is missing or damaged. A vehicle failing the Plumbtesmo test or
                      inlet restrictor inspection in some of these programs (e.g., Phoenix) is required to have its
                      converter replaced.

                           Figure 10(a) shows the value of these programmatic criteria in detecting missing or
                      lead-poisoned converters. A simple inspection of the converter, for example, would catch
                      74% of the group of vehicles which lack functional converters (converter either missing
                      or presumed to be disabled from lead exposure). Inspecting both the convener and inlet
                      restrictor, however, would  detect 98% of these vehicles. The usefulness of Plumbtesmo
                      in detecting damaged converters appears to be limited, since only 2% of the vehicles in
                      Figure 10(b) failed for Plumbtesmo only. In 1984, prior to lead phasedown,  17% of the
                      vehicles failing one of these programmatic criteria failed for Plumbtesmo only. Plumbtes-
                      mo may still be effective, however, in deterring fuel switching.
                           Poatttvt Phjmbtwmo
                              (60 Total)
Misting Catalytic
   Convwtw
  (243 Total)
                                                                        Tampered tnlvt
                                                                          flwtrtctor
                                                                          (215 Total)
                                                                  Cat«ty«tOnty<3«)
                             Tampered mMt Orty
                                 (21%)
                              PtumbtMmo + Ctf.
                                                                           Plumtot tamo Only (2%)
                                                                          Alt Thrw Indicator!   (13%)
                                                                     Wit +• Prumbtaamo   (2%)
                                          Cat + mm (29%)
                                                          (b)
                       Flgur* I0(a). Overlap of Indicator* used by ATP* to dtteet mlMbig/dtnwg«d
                                    catalyst*. Only mcltid** v*hlcl** for which all thra* Indicator* war*
                                    ln*p«ctad and ar* not covarod by *n ATP with plumbtaamo tetttng.

                             10(b).  Braakdown of vthlda* In 11 (a) by pr*val*nc* of aach Indicator
                                    combination.

-------
Motor Vehide Tampering Survey -1989                                                        Page 43
                             Of the vehicles identified as misfueled by any of the three misfueling indicators,
                        83% were using unleaded gas when surveyed (i.e., their gasoline lead concentration was
                        less than 0.05 gram per gallon (gpg)).  These vehicles, then, were identified as fuel
                        switched by a tampered inlet rcstrictor and/or a positive Plumbtesmo lest. Figure 11
                        presents the distribution of lead concentrations of 0.05 gpg or more in misfueled vehicles
                        for the 1984-1989 surveys. The impact of lead phasedown can be clearly seen in Figure
                        11. Not only has the overall prevalence of leaded gas in unleaded vehicles declined, but
                        the distribution of lead concentrations in the leaded gasoline has shifted dramatically.
                        In the 1989 survey, for example, none of the misfueled vehicles had a gasoline lead
                        concentration of 0.2 gpg or higher, compared to 17% in 1987 and 53% in 1984.

                             Of the leaded light-duty vehicles surveyed, a significant percentage (43%) were
                        using unleaded gasoline, and 32% of the leaded heavy-duty trucks were  using unleaded
                        gasoline as well. Such high percentages of unleaded fuel use may be due to unavailability
                        of or higher pricing of leaded gas, the vehicle owners' decision to change fuels, or the
                        service stations marketing leaded" gasoline that actually contained less than 0.05 gpg of
                        lead.  The distribution of lead concentrations among both light-duty and heavy-duty
                        leaded vehicles using leaded fuel is consistent with the distribution of lead concentrations
                        found in the misfueled unleaded vehicles. While this would be expected, it also suggests
                        that the addition of concentrated lead additives to the gas tank by consumers is not
                        widespread.

-------
Motor Vehicle Tampering Survey -1989
Page 44
                          Percentage of
                          Misfueled Vehicle*
                          100
                           90  -
                           80  -
                            20  -
                            10  -
                                                                             lead concentration
                                                                                range (gpg.)
                                  1984   1965   1986    1987    1988
                                                Survey Year
                          Figure 11.   Distribution of lead concentrations In leaded fuel sampled
                                     from fuel switched vehicles: 1984 - 89 surveys.  Percentages
                                     based on all fuel switched vehicles, Including those using
                                     unleaded fuel («XOS gram per gallon (gpg.)).

-------
Motor Vehicle Tampering Survey -1989                                          Page 45
APPENDIX A
    RELEVANT PORTIONS OF THE CLEAN AIR ACT


               Section 203(a)(3): The following acts and the causing thereof are prohibited --
               (A) for any person to remove or render inoperative any device or element of design
           installed on or in a motor vehicle or motor vehicle engine in compliance with regulations
           under this title prior to its sale and delivery to the  ultimate purchaser, or for any
           manufacturer or dealer knowingly to remove or render inoperative any such device or
           element of design after such sale and delivery to the ultimate purchaser; or
               (B) for any person engaged in the business of repairing, servicing, selling, leasing,
           or trading motor vehicles or motor vehicle engines, or who operates a fleet of motor
           vehicles, knowingly to remove or render inoperative any device or element of design
           installed on or in a motor vehicle or motor vehicle engine in compliance with regulations
           under this title following its sale and delivery to the ultimate purchaser.
APPENDIX B
    SURVEY AND DATA RECORDING PROCEDURES
1. Explanation of Survey Forms
               The forms on the following pages were used for recording the survey data in the
           field. The forms were forced choice to ensure coding consistency, and were designed to
           facilitate direct data entry. The following codes were used to record data for the major
           system components on the data sheets:


                      0 - Not originally equipped         8 - Misadjusted item
                      1 - Functioning property           9 • Malfunctioning
                      2 - Electrical disconnect           A - Stock equipment
                      3 - Vacuum disconnect            B - Non-stock
                      4 - Mechanical disconnect         D - Add on equipment
                      5-Incorrectly routed hose         Y-Yes
                      6 - Disconnect/Modification       Z - No
                      7- Missing item


               Additional codes were used for those components which could not be classified into
           the above categories. If a determination could not be made about a given component's
           condition, the variable was left blank. A brief description of each data entry follows.

-------
Motor Vehicle Tampering Survey -1989                                           Page 46
Form A - Underhood

         1-4   ID Number - Vehicles are numbered sequentially as they are inspected. This
               number is preceded by a site identifying letter.


         5*8   Month and year of last I/M inspection (left blank if vehicle is licensed b noa-I/M
               area).


        9-12   Manufacturer of vehicle


       13-23   Engine Family - as recorded on the underhood emission label.


       24-34    Non-serial number portion of VIN - as recorded on the driver's side of the dash
               under the windshield or the driver's door post. The VIN is recorded only if the
               engine family can not be determined.


       35-38 •   Displacement - as recorded on the underhood emission label.


       39-40    Vehicle Model Year


         41    Originally Catalyst Equipped - as recorded on the underhood emission label or
               the driver's door post.


         42    Engine Status - is coded 'V if the engine is the one originally installed in the
               vehicle; and is coded 'Z' if the vehicle: has a different gasoline engine than was
               originally equipped; was originally equipped with a diesel engine and currently
               has a gasoline engine; was converted to operate on a fuel other than gasoline or
               diesel fuel (i.e., propane); or is a "gray market" imported vehicle.


         43    Air Cleaner -is coded 'A', 'B', or T.


         44    Heated Air Intake - provides warm air to the carburetor during cold engine
               operation. The heated air intake is coded '0*, T, 3', '4\ 7* (stovepipe hose), '9'
               (vacuum override), or 'B' (custom air cleaner).


         45    Positive Crankcase Ventilation (PCV) system - prevents crankcase emissions by
               purging the crankcase of blow-by gases which leak between the  piston rings and
               the cylinder wall in the combustion chamber under high pressures. The PCV
               system is coded T, "3', '4'  (fresh air hose), T, "9", or 'B' (includes fuel economy
               devices).


         46    Exhaust Gas Redrculation (EGR) System - directs a portion of the exhaust gases
               back into the cylinders to reduce NOx emissions in the exhaust gas.  The standard
               EGR configuration consists of a vacuum line from the carburetor to a sensor
               (used to detect engine operating temperature to activate the EGR valve),  and
               another vacuum line from the sensor to the EGR valve. The EGR system is
               coded 'V, T, T, T, '4', '51, T, or '9".

-------
Motor Vehicle Tampering Survey • 1989                                            Page 47
         47    Evaporative Control System (ECS) - controls vapors from the fuel tank and
               carburetor. Some systems have two lines: from the fuel tank to the canister, and
               from the canister to the carburetor or air cleaner (for purging the canister).
               Other systems have a third line connected to the carburetor. The ECS is coded
               T, T (carburetor line), '4' (tank line), '5', '6' (air cleaner unsealed), T, or '9'
               (cracked hose or canister).

            Air Injection System • extends the combustion process into the engine's exhaust system
            by injecting fresh air into  the exhaust ports, lowering exhaust emissions while  still
            maintaining proper vehicle performance. Two types of air injection systems are currently
            used. One type uses a belt-driven air pump to direct air through a control valve and into
            the exhaust manifold. The other type is a Pulse Air Injection Reaction (PAIR)  system,
            which uses an aspirator commonly located in the air cleaner to supply air to the exhaust
            manifold.

         48    Aspirated Air Injection System (PAIR) - coded '0* (If air pump system or none), T,
               '4',T,or'9'.

         49    Air Pump Belt - is coded '0' (if PAIR or none), T, T, or '8* (loose belt).

         50    Air Pomp System - for the purposes of this variable, consists of the air pump and
               control valve and is coded '0* (if a PAIR or none), T, "3', '4' (other than belt
               removal), T, or *9*(frozen pump).

         51    Exhaust Manifold - coded 'A' or 'B'.

         52    Oxygen and Related Sensors - The oxygen sensor  controls the air-fuel mixture
               going into the engine of vehicles equipped with three-way catalytic conveners. A
               vehicle's computerized engine and emissions control system receives input from
               various sensors for engine condition information, and constantly adjusts the
               air/fuel ratio, distributor, and emissions devices for optimum economy,
               driveability, and emissions.  These sensors are coded 'CK, T, T, '4' (unscrewed),
               '6',orT.

         53    Carburetor Type - is coded T (fuel injected), 'A'  (stock carbureted), or 'B' (non-
               stock carbureted).

         54    Turbocharger - coded 'ff, 'A', 'B', or 'D'.
Form B - Rear

         1-4   ID Number • Same as on Form A.

         5-8   Make

        9-12   Model

        :13    Vehicle Type - coded as follows: C = car, T = light-duty truck, H = heavy-duty
               truck.

       14-15    License Plate -State abbreviation

       16-19    Program Jurisdiction - the city or county of the vehicle's registration is entered
               (if the vehicle is covered by a control program).

-------
Motor Vehicle Tampering Survey -1989                                            Page 48




      20-23     Exhaust gas HC concentration (in ppm) at curb idle.


      24-26     Exhaust gas CO concentration (in percent) at curb idle.


      27-29     Exhaust gas COi - concentration (in percent) at curb idle.


      30-32     Exhaust gas Oj - concentration (in percent) at curb idle.


      33-35     Odometer - mileage in thousands


         36     Dash Label - displays the fuel required and is coded '0' (for leaded vehicles), T, or '7'.
         37     Catalytic Converter - oxidizes the HC and CO to water and CCh in the exhaust gas.
                Later model catalysts also reduce oxides of nitrogen. The converter is coded '0',
                '!', T (entire catalyst canister removed), or 'B' (aftennarket catalyst present).
         38     Exhaust System - if as originally equipped an 'A' is coded. If non-stock a "B' is coded..
         39     Exhaust System Integrity. the condition of the exhaust system is coded' 1' (no
                obvious leaks) or *9* (leaks evident).
         40     Tank Cap * seals the fuel tank during normal operating conditions and is coded T,
                'T, or "9* (loose cap). A separate entry (see below) is made to indicate whether
                the tank cap is stock or non-stock ('A' or 'B').
         41     Tank Cap Origin - is coded 'A' if the tank cap is a stock cap. The cap is coded 'B'
                if it is non-stock.
         42     Tank Label - displays required fuel and is coded '0* (for leaded vehicles),'!', or T.
         43     Filler Neck Restrktor - The restrictor is designed to prevent the introduction of
                leaded fuel into a vehicle requiring unleaded fuel It is coded '0* (for leaded
                vehicles), T, *4' (widened), '6* (nozzle adaptor present), T, or 'B' (non-stock
                inlet restrictor-functioning property).
         44     Phimbtesmo - Plumbtesmo paper is used to check for the presence of lead in vehicle
                exhaust pipes. A positive indication is coded as 'P* and a negative as 'N1.
         45     Fuel Sample - indicates if inspector was able to obtain fuel sample for later lead
                analysis ('V or 'Z').

-------
 Motor Vehicle Tampering Survey -1989                                          Page 49
2. Classification of Component Conditions

                The table below was used to classify the various system components as tampered
            (T), arguably tampered (A), or malfunctioning (M).  Only those codes which are
            applicable to a given component are listed.  Codes for 'not originally equipped' and
            'functioning properly' are not included in this table. Refer to Appendix B, Part 1 for an
            explanation of the codes.


                                                   Codes from forms A and B
                   Component/system	234S6789B
                   Dash Label                                     A

                   Tank Cap                                       A      M

                   Tank Label                                     A

                   Filter Neck Restrictor                   T     T   T

                   Catalytic Converter                               T

                   Oxygen and Related Sensors     T      T     T   T

                   PCVSystem                      XT         T      M  T

                   Heated Air Intake                  T  A         A      M  T

                   Evaporative Control System         T  T  T  T   T      M

                   Aspirated Air Injection System       T  T         T      M

                   Air Pump Belt               .                     T   M

                   Air Pump System                  T  T         T      M

                   EGRSystem                  T  T  T  T      T      M


               T = Tampered

               A = Arguably tampered

               M = Malfunctioning


3. Fuel Sample Collection and Labeling Procedures

               An  attempt was made to take a fuel sample from each vehicle surveyed.  These
           samples were collected in two-ounce bottles with a hand-operated fuel pump. Once the
           sample was drawn, the fuel was replaced with an equivalent amount of unleaded fuel if
           the driver requested, and the pump was flushed with »n leaded fuel.

               Each bottle was identified with an adhesive label that had the vehicle identifying
           survey number on it. The vehicle identifying number was the first entry on the data forms
           described in Part 1 of Appendix B. The  bottles were packed, labeled, and shipped to
           EPA's Motor Vehicle Emissions Laboratory in Ann Arbor according to the shipper's
           requirements.

-------
Motor Vehicle Tampering Survey -1989                                          Page 50
4. Plumbtesmo Application
         1)  Clean a portion of the inside of the tailpipe large enough for the test paper by wiping
               it out with a paper towel or cloth. This may be necessary to remove soot deposits
               which might mask the color change.


         2)  Moisten the Plumbtesmo paper with distilled water and immediately* press firmly
               against the surface to be tested for approximately thirty seconds. If the tailpipe is
               hot you may wish to clamp the test paper in the tailpipe using a clean clamp.


               •Note
                  The Plumbtesmo  paper must be applied during the time that the
                  paper is yellow for the reaction to occur. After approximately 15
                  seconds the yellow color  disappears and the paper is no longer
                  effective. Excess water also interferes with the reaction.

                  Care must be taken to avoid contamination of the test paper. If a
                  person has recently handled a test paper with a positive reaction,
                  some lead or reactive chemical may have been transferred to their
                  fingers. Subsequently handling a clean test paper may cause con-
                  tamination.


         3)  After removing the test paper, determine whether a color change has occurred. Red
               or pink coloration indicates the presence of lead.


5. Field Quality Control

               Reference and calibration gases were used to ensure the accuracy of the emissions
           analyzer. Horiba gases certified by RTF were used as reference gases.  Two cylinders
           of reference gas were used to validate the accuracy of the calibration gases before they
           were taken to the field on each survey.


               Two calibration gases (Horiba) were used. These gases were a mixture of CO, HC,
           and CO2 b nitrogen and were  used to check the  instrument twice a day.  These
           calibration gases were certified by the manufacturer and the RTF reference gases. Their
           approximate compositions were:
                                         8% CO

                                        11%CO2

                          1560 ppm HC (Hexane equivalent)
                                        1.6% CO
                                        11% CO2

                           320 ppm HC (Hexane equivalent)

-------
Motor Vehicle Tampering Survey -1989
Page 51
APPENDIX C
    EMISSION CUTPOINTS FOR I/M AREAS

              The table below lists the emission cutpoints used by the I/M areas covered in the
           1989 tampering survey. The cutpoints for pre-1975 vehicles are not included, since these
           vehicles were not surveyed.
Survey Site
San Dlego/Bakersfletd, CA







Phoenix, AZ




New Jersey

Detroit, MI

Emissions Cutpoints
Model Year CO (%)
1975-79 (no CC)
1975-79 (OC, no AI)
1975-79 (OC.AI)
1975-79 (TWC)
1980+ (no CC)
1980+ (OQnoAI)
1980+ (OC.AI)
1980+ (TWC)
1975-78 (4 cyL)
1975-78 (6-8 cvL)
1979 (4 cyl.)
1979 (6-8 cyl)
1980+ (all)
1975-80
1981 +
1980
1981 +
3.5
4.5
1.5
1.5
2J
2.5
12
1.0
22
2.0
22
2.0
12
3.0
12
3.0
1.2
HC(ppm)
200
250
150
100
150
150
150
100
250
250
220
220
220
300
220
300
220

-------
Motor Vehicle Tampering Survey -1989
Page 52
            Milwaukee, Wl
            Anchorage/Fairbanks, AK
            Denver, CO
1975-77
1978
1979
1980
1981 +
1975-83 (no CC)
1975-83 (OC, no AI)
1975-83
1975-83 (TWC)
1984 (all)
1975-76
1977-78
1979
1980 +
5.5
4.0
3.0
2.0
1.2
3.0
4.5
1.5
1.0
1.0
4.4
3.4
2.0
U
450
350
275
230
220
-
-
-
-
-
800
500
400
400
            KEY: CC = catalytic converter (all types), CYL. = cylinder,
                  OC = oxidation catalytic converter, AI - air injection,
                  TWC = three-way catalytic converter.

-------