Click for next page ( 2


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
Summary Motor vehicles are a major source of air pollution on urban, regional, and national scales. Programs to control their emissions have focused on setting emissions standards for new vehicles, resulting in engineering and design im- provements in emissions-controT systems. Inspection and maintenance (~/M) programs have been instituted in many jurisdictions to ensure that those controls operate properly throughout the life of a vehicle. These programs are implemented in areas violating federal air- quality standards (nonattainment areas) and in other areas seeking to improve air quality. The inspection typically involves regularly scheduled exhaust tests measuring carbon monoxide (CO), hydrocarbons (HC), and sometimes nitro- gen oxides (NOx) emissions. I/M tests also include a visual inspection ofthe components controlling evaporative and exhaust emissions and may include a functional gas-cap test and a pressure test ofthe evaporative emissions-controT system. The U. S. Environmental Protection Agency (EPA) has oversight and developmental responsibility for I/M programs, which are implemented by state agencies. Evidence suggests that I/M programs have been less effective than antici- pated. This concern prompted Congress to request this study in its fiscal ~ 998 appropriations to EPA. In response to this request, the National Research Council (NRC) convened the Committee on Vehicle Emission Inspection and Maintenance Programs to conduct this study. Phase ~ ofthe study, presented in this report, examines the criteria and metho(lology for evaluating I/M pro- grams and assesses their effectiveness in reducing vehicle emissions. Phase

OCR for page 1
~ Evaluating Vehicle Emissions I/M Programs 2 will evaluate several types of T/M programs in more depth. The study charge also calls for the committee to make recommendations for improving T/M programs. The committee recommends some improvements in this report and will address others in phase 2. In carrying out its charge for this report, the committee reviewed pas- senger-car and light-truck emissions (the emissions typically targeted by T/M programs) within the context of overall emissions from mobile sources and other anthropogenic sources. Em~ssions-control technologies and testing tech- niques were considered, together with how changes in these factors might affect T/M programs. The committee also reviewed methods to estimate emis- sions reductions, and it examined previous evaluations ofthese reductions and other criteria important for evaluating such programs. FINDINGS AND RECOMMENDATIONS The committee found that I/M programs have generally achieved less emissions reductions than originally projected by EPA's Mobile Source Emis- sions Factor (MOBILE) model and the California Air Resources Board Emis- sions Factor (EMFAC) model. These model-predicted reductions are impor- tant because they serve as part of the formal basis for crediting emissions reductions within state implementation plans (SIPs). ~ Independent and state- sponsoredevaluations of ongoing I/M programs have estimated that the em~s- sions reductions attributable to these programs are from zero to about one-half ofthe reductions predicted by the models.2 This figure is estimated using in- use vehicle emissions data, such as remote-sensing, random roadside vehicle testing, and T/M emissions testing. The committee concluded that these data provide the best estimate of T/M's effectiveness. Evaluations that rely on mode! predictions of emissions reductions using few or no in-use data are not accurate. This conclusion is based on a small number of peer-reviewed stud- ies, and the methods used to make these estimates must be improved. Despite the smaller-than-forecasted benefits from I/M programs, the iSIPs describe the strategies that regions in noncompliance with National Ambient Air Quality Standards (NAAQS) use to come into compliance. 2The estimated emissions reductions are dependent on the pollutant and version of the model used for comparison. They are lowest for test-and-repair idle test programs and highest for hybrid or centralized transient test programs.

OCR for page 1
Summary committee still sees a great need for programs that repair or eliminate high- em~ssions vehicles (commonly called high-em~tting vehicles or high emitters) from the fleet, given the major influence ofthis small fraction ofthe fleet on total emissions and air quality. However, I/M programs should improve the way they identify vehicles in need of repair and verify repairs. Improvements are also needed in the methods used to evaluate the impacts of these pro- grams. The use ofthe MOBILE and EMFAC models to predict emissions-reduc- tion benefits from I/M programs in the development of air-quality-attainment plans indicates a flaw in the SIP process. EPA has granted states substantial emissions-reduction credits for I/M programs without the need to verify the extent to which the predicted emissions reductions are actually occurring. That situation creates a regulatory disincentive for states to evaluate the actual emissions-reduction benefits from T/M programs. Such an evaluation might reveal a shortfall in a state's emissions-reduction benefits and trigger require- ments for SIP revisions. The committee recommends that the crediting of emissions-reduction benefits for T/M programs, as with other emissions-control strategies, should be tied to the actual emissions reductions produced by these programs. Em- phasis on observational data and empirical evidence has been inadequate in most aspects of I/M program evaluations. Rigorous scientific and technical analyses have been lacking in supporting decisions related to program imple- mentation. EPA should expand its use of outside experts and publication of analyses in peer-reviewed literature to address that deficiency. Improvements are especially needed in the quantification of I/M program impacts based on more data-intensive approaches. These evaluation methods and their applica- tions to estimate emissions benefits of state T/M programs should be reviewed independently and be disseminated to policy makers and the public. Prospective Estimates of Emissions Reductions from [/M Programs Findings On the basis of evaluations by states and by independent researchers, the committeefound thatI/Mprograms provide much lower benef Is than estimated by the models. The MOBILE model estimates that a filthy imple-

OCR for page 1
4 Evaluating Vehicle Emissions I/M Programs mented enhanced I/M programs would produce overall emissions reductions of 28% for HC, 31/O for CO, and 9% for NOx.4 Reasons for overstating emissions reductions include the following: Overestimation ofthe deterioration of vehicle-em~ssions performance (which overestimates potential benefits from T/M-induced emissions repairs). Inadequate representation ofthe behavior of motorists and mechanics. Overestimationofcompliancewiththeprogramandtheeffectiveness rat . 01 repairs. Overestimation of the ability to identity high-emitting vehicles. Incomplete implementation of some components of I/M programs (e.g., effective evaporation tests and inclusion of all older vehicles). Recommendations EPA and states should expectlower emissions-reduction benef tsfrom I/Mprograms as currently configured. In general, models projecting emis- sions reductions from I/M programs should be improved to reflect actual re- ductions more accurately. States should perform periodic on-road sampling and evaluations of emissions reductions and compare those observations with modeled forecasts used in Sips. To the extent that states are allowed to use clefault parameters in emissions models to forecast I/M emissions reductions for SIP credit, the default values of key parameters (e.g., compliance rates and repair effectiveness) should be more pessimistic (i.e., forecast lower emissions reductions) than those currently used. That might help to create an incentive for states to provide evidence that their programs achieve greater emissions reductions than specified by the default settings in the model. . High-Emitting Vehicles Findings A small, malfunctioning fraction of the fleet contributes a substantial proportion of overall vehicle emissions. Typically, less than 10% of the Enhanced I/M programs are required in areas classified in "serious," "severe," or "extreme" nonattainment of National Ambient Air Quality Standards. 4The discussion here refers to analysis that was performed in 1992 using the version of the model known as MOBILE4. 1.

OCR for page 1
Summary 5 Beet contributes more than 50/ of the emissions for any given pollutant. Emissions reductions are skewed; a relatively small share of the vehicles failing an T/M test contributes a large proportion of total excess emissions remissions above the standard for failing a vehicle), while vehicles with em~s- sions just above the threshold for test failure (so-called "marginal emitters") often have only a small reduction in overall emissions after repairs or, in some cases, actually have an increase in emissions after repairs. Thus, the largest potential reductions in emissions from T/M programs are associated with a small number of high-em~tting vehicles. Studies that combine data for vehicle ownership, high-emitter frequency, and income levels suggest a strong link between low household income and the likelihood of owning a high-em~tting vehicle. Studies also show that between 10/0 and 27% of vehicles that fad! an I/M test never pass the test. Their exact fate has not been well characterized, although some have been found to be still in operation in I/M areas in some states more than a year after their last test. More study of this issue is needed to determine how serious this problem is and what policies will improve it. Recommendations I/Mprograms shoul~focus primarily on identity cation, diagnosis, and repair of the highest-emitting vehicles along with verification of those repairs. A number oftesting or identification regimes can identify high-emit- ting vehicles, including traditional I/M programs testing all vehicles, programs targeting certain vehicles for more or less frequent testing, and remote sensing. States should be given flexibility to choose a regime that meets their em~ssions- reduction goals at the lowest cost to the public. The focus on high-emitting vehicles should extend to promoting policies that seek effective repair or removal of all such vehicles. However, any pro- gram designed to repair high-emitting vehicles might raise serious fairness concerns, because high emitters are more likely to be owned by persons of limited economic means. The committee recommends that policies be ex- plored to provide financial or other incentives for motorists of high-emitting vehicles to seek repairs or vehicle replacement. Clearly, further research is needed to design the means to reduce high emitters in ways that are effective as well as socially and politically acceptable. States would have to evaluate which policies are the most cost-effective and acceptable ways of obtaining emissions reductions from high-emitting vehicles.

OCR for page 1
6 Evaluating Vehicle Emissions I/M Programs The committee is aware that identification of high-emitting vehicles is problematic and that the designation is relative. The sense ofthe committee is that more needs to be known about the cost-effectiveness of setting different emissions cutpoints,5 including the value of repairing vehicles with emissions only marginally higher than current cutpoints, to determine optimal cutpoints for vehicle tests. Evaluating I/M Emissions Reductions Findings Off cial biennial evaluations of enhanced I/M programs required by the Clear, Air Act Amendments of 1990 have not been completed by the majority of states required to perform them. The guidelines clevelope(1 by EPA for performing these evaluations are limited to a single method, which compares an I/M program with a benchmark I/M program. These guidelines are being revised and expanded. Most past evaluations have been performed by state agencies in response to state requirements for estimates of emissions- reductions benefits or by independent researchers interested in the same issue. The primary data sources for evaluation of emissions-reduction benefits are test data from I/M programs, remote sensing of on-road vehicles, and roadside testing of on-road vehicles. Vehicle registration data are also impor- tant for estimating changes in the fleet over time. There are three approaches for using those data to determine emissions-reduction benefits. The "reference method" compares vehicle emissions in the program area with those in some reference area, which can be a benchmark T/M program or a non-~/M area. The "step method" compares emissions of vehicles tested under a newly instituted I/M program with emissions of vehicles in the same area that have yet to be tested under the new program. The "comprehensive method" tracks changes in emissions for vehicles that pass the test, those that initially fail and then pass, and those that fait and never pass. Each data source and evaluation method has inherent advantages and disadvantages. For example, simple comparison of emissions data in one area with those in a reference area needs to correct for physical and socioeconomic - sThe emissions levels that define whether a vehicle passes or fails are called cutpoints.

OCR for page 1
Summary 7 differences between regions in which emissions would be expected to vary regardless of the presence or absence of an I/M program. In addition, using data on repaired vehicles collected as part ofthe I/M program to estimate the em~ssions-reduction benefits might not account fully for noncompliance with the program or for repairs made in anticipation of the I/M test. Recommendations EPA shoul~provide additionalguidancefor carrying out]/Mevalua- tions. The agency is commended for beginning this work, and the committee recommends that it be expanded to include additional methods of evaluation. EPA's guidance should be based on sound measurement and statistical evalua- tion methods and be peer reviewed. The agency should address comments gathered during the review of these evaluation methods. In addition, EPA should publish aspects ofthese evaluations in professional journals so that they can be reviewed further and disseminated. The committee recommends that EPA and states ensure that some programs undergo comprehensive, long-term evaluations using multiple data sources and analytic techniques. I/M programs that undergo repeated, in-depth evaluations using multiple data sources and methods can potentially help improve the design of I/M programs and evaluation process nationally. Questions about the fate of vehicles that fail their I/M test, the durability of vehicle repairs, or the impact of T/M programs on vehicle registration and ownership patterns can be answered only through well-(lesigned, comprehen- sive evaluations using a number of data sources. Independent researchers should perform parts of these comprehensive evaluations. As stated previously, these full evaluations should be peer re- viewed independently by experienced researchers, and EPA should pursue publishing some aspects ofthese evaluations inprofessionaljournals. Because such evaluations are resource-intensive, EPA should select several programs for such treatment and should support a portion of this work. The committee recognizes that not ad jurisdictions wiR be able to devote the resources needed to perform comprehensive evaluations using multiple sources of primary data. The committee recommends that guidelines for a shortened evaluation method also be developed and peer reviewed. The method should not rely on the MOB TLE mode! but should be based on the best evidence from ongoing full evaluations and should include

OCR for page 1
8 Evaluating Vehicle Emissions I/M Programs estimates of all components of emissions reductions achieved by I/M pro- grams. These shortened evaluations will likely have to rely primarily on I/M program data and other local data for primary data sources, although on-road data would be valuable. States should be urged to collect at least the amount of on-road data required under the rules for implementing enhanced I/M (O. ~ % ofthe fleet). Evidence from full evaluations done in other locations might have to be incorporated to account for factors such as repair deterioration; ineffec- tive, incomplete, or fraudulent repairs; pretest repairs; and program avoidance by changing vehicle registration. A review committee should be established to advise EPA in the selection of shortened evaluation methods and in the selection of what information can be drawn from full evaluations to inform the shortened evaluation. The com- mittee is concerned about the need for states to complete overdue evaluations and urges EPA to continue to develop these evaluation methods in a timely manner. Assumptions used in the shortened evaluation can then be continually improved as more evidence becomes available. Both the comprehensive and the shortened program evaluations should include a consistent set of performance indicators, such as the number of high-emitting vehicles driven in an I/Mprogram area that are avoiding testing. Although such indicators do not incorporate direct estima- tion of emissions reductions, they can help back the performance of a program overtime and provide relatively concise indicators of program success. These performance indicators could include the following: inspections. An estimate ofthe total number of vehicles driven in the I/M region, the share of those vehicles that are eligible for inspection, ant! the share of those that are inspected. Estimates of the actual number of high emitters on the road. Failure rates by model year at the program cutpoints. Estimates ofthe average emissions of vehicles that pass and that fad! Share of failing vehicles that actually get repaired to below program cutpoints and their average emissions rates before and after repair. Share offailing vehicles that do not ever pass the T/M test, their aver- age emissions rates, and estimates of the number of those still driven in the area. The rate of repeat failures from one T/M cycle to the next.

OCR for page 1
Summary 9 Research Issues in I/M Evaluation Findings Many criticalfactors that have large effects on the emissions-rearuc- tion benef is from I/M programs are still unknown. An example is the length oftime that repairs remain effective for a vehicle initially failing an I/M test. Estimates of the average effective duration for such repairs range from most ofthe benefits disappearing in less than 6 months to remaining for 2 years or more. Without better understanding of repair duration and other unknown factors, the full effect of I/M programs on vehicle emissions will remain uncer- tain. Full evaluations of at least a few I/M programs would shed light on many such issues. Recommendations Comprehensive evaluations of ]/M programs should be used to re- search aspects thought to have major impacts or' the emissions-reduction benef is from I/Mprograms. These include the following aspects: I/M test. The distribution ofthe duration of repairs for vehicles that fail an initial The extent of pre-inspection repairs. The extent to which temporary repairs and test fraud result in vehicles registering low emissions only for the purpose of passing an I/M test (the "clean for a day" phenomenon). The fate of vehicles that fail their initial T/M test and never pass (unre- solved failures). Consequences of I/M programs for nontailpipe HC reductions. In addition, many ofthese unresolved issues relate to human responses to I/M programs, but only a few studies have attempted to examine those as- pects. Comprehensive evaluations can shed light on the type and magnitude of behavioral responses, but separate behavioral studies are likely to be needed to provide additional important insight.

OCR for page 1
10 Evaluating Vehicle Emissions I/M Programs NOx and Particulate Matter (PM) Emissions Findings Future air-quality improvement programs are likely to place greater emphasis on controlling NOx andPMemissions. I/M programs traditionally have focused on inspecting vehicles for high CO or HC emissions or both. Loaded-mode emissions testing procedures6 that are needed to measure NOx emissions have been introduced widely in only the past 5 years. Apart from smoking-vehicle complaint programs and some testing of heavy-duty diesel- truck smoke opacity, little effort has been made to identify and repair vehicles with high emissions of exhaust PM. Currently, there are few assessments of I/M program effectiveness in reducing emissions of NOx and PM pollutants. Although diesel engines are a minor source of CO and HC, they are significant contributors to mobile-source NOX and PM emissions. Recommendations I/M programs should clearly state which pollutants they are seeking to reduce. Different types of repair actions and different mechanic train- ingprograms are neededfor 1/M programs thatfocus on reducing NOX and PM emissions. Because heavy-duty diesel vehicles are a significant source of NOx and PM, I/Mprograms that target these pollutants might have to incorporate heavy-duty diesel vehicles to a greater extent. Remote Sensing Findings Remote-sensing measurements are art excellent source offeet-aver- age CO and HC emissions data. Remote sensing can also be a usefu! screening too! to identify vehicles likely to pass orfai! conventional bed 6A loaded-mode test involves testing vehicle emissions while the vehicle is on a dynamometer that simulates the load a vehicle is under during on-road driving.

OCR for page 1
Summary 1 1 program tests. Although use of remote sensing is increasing, its capabili- ties remain underutilized in I/Mprograms. Combined remote-sensing and roadside pullover studies have shown that a high proportion of vehicles iclentif~ed by remote sensing as high emitters of CO, HC, or both also failed roadside tests given immediately after the remote- sensing test. However, the fraction of high-emitting vehicles that escaped detection by remote sensing in these studies and the number that do not partici- pate in conventional T/M programs are unknown. Recommendations Remote sensing should have an increased role in assessing motor vehicle emissions and I/M program effectiveness, determining the extent of pre-inspection repairs, andf estimating the extent of certain types of noncompliance. Remote sensing is also effective for identifying high emit- ters; however, its implementation into an I/M testing program should be an area of further research. Greater attention must be paid to site selection and quality-assurance and quality-control issues in remote-sensing studies. Some prior studies have focused too heavily on the number of vehicles and sites sampled and have sacrificed quality in seeking large quantities of data. To determine the ability of remote sensing to identify vehicles with high NOX emissions, combined remote-sensing ant! roadside pullover studies that focus on this pollutant should be con(luctecl. An intercomparison ofthe ability of different remote sensors to measure NOx emissions accurately should also be performed. An important research priority is the development and evaluation of remote-sensing capabilities for exhaust PM emissions. Further research is also needed to increase the number and types of roadside sampling sites where remote-sensing equipment can be deployed. On-Board Diagnostics Findings The committee found that the current data set for evaluating the ef- fectiveness of OLD ~ for I/M testing is inadequate. Contemporary on-

OCR for page 1
12 Evaluating Vehicle Emissions I/M Programs board diagnostic (OBD) equipment, included on ~ 996 and newer model-year vehicles, represents a technological innovation for monitoring the performance of em~ssions-controT equipment on light-duty vehicles. Current OBD technol- ogy, known as OBDII, provides rapid verification of the operation of both exhaust and evaporative emissions-controT components but does not measure emissions. it alerts motorists to potentialproblems by illuminating amalfi~nc- tion indicator light (MIL) end provides mechanics with diagnostic information about the source of malfunctions, including malfunctions that are intermittent in nature (e.g., a misfire). OBDII also represents a potentially improved meth- od for assessing evaporative em~ssions-contro! components. Given its current specifications for MIL warnings however, it is not clear whether OBDII can fulfill both objectives of alerting vehicle owners to potential vehicle malfunc- tions and serving as a testing device in I/M programs. In addition, it is not known how motorists will react to MIL illumination, especially when the vehi- cles are no longer under warranty. The OBDII system could operate as designed by automobile manufactur- ers and still indicate OBD I/M test failures on vehicles with low emissions. The current specification is that the MIL will illuminate if aproblemis detected that results in or could potentially result in emissions higher then ~ .5 times the vehicle's emissions certification standard. Studies have shown that if OBDIT were used to decide whether vehicles passed or failed an inspection, most OBDTI failing vehicles would have emissions less than ~ .5 times the standard. Current I/M programs typically have much higher cutpoints than ~ .5 times the vehicle's certification standard. The OBDTI failure point might be too stringent for a cost-effective and publicly acceptable I/M program especially for older OBDII vehicles. An alternative approach, such as tailpipe testing, might be nee(led for those vehicles. Recommendations An independent evaluation should be established, with appropriate funding, using researchers outside the agencies to review the effective- ness arid cost-effectiveness of OBD testing programs before moving forward with fuR implementation of OBDI] rule requirements. The rule allows states up to 3 years to phase in OBD T/M, which is required to begin January 2002. The recommended evaluation should study the issues of inter- mittent failures and the value of repairing vehicles with low emissions to pre-

OCR for page 1
Summary 13 vent an increase of emissions in the future. Failing a large number of vehicles with emissions below ~ .5 times the certification standards could undermine the commitment to find high-emitting vehicles and ensure that they are repaired. An alternative to using OBDT! as a failure criterion in T/M testing is to use it as an advisory too!to inform motorists of potential emissions problems. This option can be used while phasing in an OBDII I/M program. It may also be considered! for OBDIT vehicles when they become older. No matter how OBDII is used, a substantial effort by EPA is needled to help the public thor- oughly understand this system. Besides the issues of intermittent failures and the value of failing marginal-em~tting vehicles with malfunctioning sensors or monitors, studies of other issues related to OBD TI should be done. Such issues include the following: The fraction of vehicles appearing in I/M lanes with MILs illuminated. The fraction of vehicles with MILs illuminated that do not fail the exhaust test or any evaporative test. The fraction of vehicles without MILs illuminated that fad! the I/M test. The response rates of consumers to MILE illumination in both the ab- sence and the presence of an I/M program and in the absence of a warranty. The use of OBDII diagnostic information to identify vehicle repairs that have a high-emissions-reduction potential and repairs that have only a marginal impact on emissions. The possibility for changes in the cutpoint settings on OBDII systems to allow OBDIT to focus on high-emitting vehicles. In the long-term, the promotion of actual emissions readings in future OBD systems. Methods for measuring actual emissions-reduction benefits from OBDII. Use of the MOBILE Mode! Findings The SIP process mandated by the Clean Air Act and its amendments requires that modeling be used to predict emissions inventories and esti- mate benef is from ]/Mprograms in future years. Based on comparisons with I/Mprogram evaluations, predictions from the current version of the

OCR for page 1
14 Evaluating Vehicle Emissions I/M Programs MOBITE mode! have greatly overestimated the emissions benef ts from I/Mprograms. ' These findings and the 2000 report Modeling Mobile-Source Emissions by the National Research Council suggest that there has been inadequate emphasis on data and empirical evidence in modeling I/M benefits. Recommendations The methodology used in MOBllEfor estimatingI/Mbenef ts should be reevaluated. MOBILE should allow its users to readily incorporate data from current I/M program evaluations into assessments for future years. Key parameters (e.g., compliance rates, repair effectiveness, and OBDII I/M benefits) used to forecast I/M emissions-reduction benefits should have pessi- mistic default estimates resulting in Tower expected reductions. States might then have an incentive to demonstrate, through evaluation, that their programs are better than the default. Further, embedded assumptions in the model should be simplified as much as possible so that assumed parameter values are transparent to users, and users can incorporate the latest available data into parameters.8 In the long-term, the overall [/M estimation methodology in MOBILE should be substantially revised. Empirical data show that the under- Tying I/M modeling approach is flawed. The committee recognizes the need to continue using models to estimate I/M program benefits in future years. it is important, however, to reiterate that evaluations of current program performance should rely as extensively as possible on empirical data (e.g., on-road vehicIe-emissions measurements) rather than on models such as MOBILE. Importance of Cost-Effectiveness and Public Response to I/M Findings Although emissions reductions are central to any evaluation of EM programs, costs are inextricably linked to emissions reductions, making 7EPA is currently working on MOBILE6, which is expected to be less optimistic in its I/M benefit predictions. ~Naturally, these input data should undergo some type of evaluation and approval process so that the resulting emissions estimates are credible.

OCR for page 1
Summary 15 cost-effectiveness a critical evaluation criterion. For example, costs influ- ence the behavior of vehicle owners and repair technicians, thereby affecting the emissions reductions achieved. Both the emissions reductions and the associated costs must be considered! in the design and improvement of I/M programs, and in the determination of whether effort is best directed at I/M or at alternative ways of reducing emissions. Another important consideration is public concern about new technologies, such as OBDII or remote sensing, that are increasingly used in T/M programs. For example, confusion about what the MIL is conveying to drivers could impede the use of OBDIT in T/M tests. Confusion about new technologies could reduce public and political support for their introduction into I/M pro- grams and/or reduce their effectiveness. Recommendations I/M programs can be improved by identifying ways to make them more cost-effective and more readily understood and by easing the testing burden for vehicle owners. States should be encouraged to develop and implement cost-effective means for finding and repairing high-emitting vehi- cles. Analysis of cost-effective measures, however, must take account ofthe effect of I/M program requirements on owners' behavior. Some ofthe issues that deserve further research include the following: The roles of repair cost waivers in I/M programs. I/M programs typi- cally devote considerable money and effort to finding failing vehicles. Once a vehicle is identified as a very-high-emitting vehicle, that vehicle should be repaired, sold out of the area, or scrapped. Vehicle scrappage programs and repair assistance programs are examples of policies that could be used to accomplish such a goal. The most cost-effective policies may differ by region. ~ The use of emissions profiles for determining testing frequency. There is already growing evidence that reducing the frequency of testing vehicles with a low probability offaiTure, including exemption of recent model- year vehicles from regular testing, is very cost-effective. On the other hand, increasing the inspection frequency of vehicles with a high probability offailure and/or those with high repair deterioration should be investigated to determine its cost-effectiveness for identifying high emitters. ~ The durability of emissions-control systems. Encouraging the pro- duction of vehicles with more robust emissions-control systems through the use

OCR for page 1
~ 6 Evaluating Vehicle Emissions I/M Programs of extended warranties and new-vehicle compliance programs may be an alternative approach to maintaining low emissions throughout a vehicle's life- time. Understanding and quantifying owners ' responses to I/M reguZa- tions. For example, not enough is known about the extent of old-vehicle scrapping in response to T/M or of program avoidance and other types of noncompliance among different socioeconomic groups. The cost and emissions consequences of enforcement efforts. Greater enforcement of existing regulations may be a cost-effective way to improve program performance. More effective means of public outreach and education. New approaches to providing information to the public about new technologies that may be incorporated in I/M programs should be developed and studied. Prior- ity should be given to concerns regarding remote sensing and OBDIT.