Consultant's Report
Committee Accepted
November 18, 1999

---

Automobile Emissions Evaluation
Response to Inquiry

Prepared for the Legislative Program Review and Investigations Committee
By
Connecticut Academy of Science and Engineering
179 Allyn Street
Hartford, CT 06103-1422

Draft

Summary

The Legislative Program Review and Investigations Committee sought a scientific study of the efficacy of the Connecticut Motor Vehicle Emissions Testing Program. The Connecticut Academy of Science and Engineering conducted a similar review in 1986, titled Automobile Emissions Testing.

Given the advance of technology in factory-installed automobile emission control systems, the study was asked to focus on the changing nature of the fleet of vehicles using Connecticut highways and the impact an emissions testing program can have on air quality within the state.

Therefore, it was requested that the Academy examine certain technical aspects of the Connecticut vehicle emissions control program. In particular, the Academy was asked to perform the tasks listed below. The Academy response follows each listed task and a cost-benefit analysis is included in the final section of the report.

1. Analyze the nature of the fleet in terms of age distribution, extent of manufacturer installed antipollution devices, federal emissions standards required of the manufacturer, and ability to alter tailpipe emissions through repair/corrective action.

The age distribution of the inspected fleet has not changed significantly over the years from 1994 to 1998. Twenty percent of the fleet is three years or less in age and twenty percent is eleven years or older. (Section 1.1). However, we are told that nationally (EPA420-F-99-017, May 1999, Attachment 1) half the vehicles sold in the past few years are SUVs, which have greater emissions limits and achieve fewer miles per gallon of fuel than automobiles. The Federal Government has been tightening emissions limits for vehicles and EPA proposes new limits in the 2004 to 2009 time period. The old limits for nitrogen oxide and the proposed new limits for cars and light trucks, including minivans and SUVs are listed in the tables in Attachment 1. These are the requirements on the manufacturer.

Catalysts were introduced in autos in 1974. Closed loop computer systems were introduced in 1980. Electronic fuel injection and ignition were introduced in the middle 1980s and improvements have continued in motor efficiency and control of emissions through the 1990s. On board diagnostic assessment of engine performance may be used in 1995 and later vehicles to check for performance and related emissions. Section 1.2.

The ability to alter tailpipe emissions through tampering is severely limited in computer controlled engines. There is no reason to tamper because the engine will not operate as well. Repair/corrective action is effective in keeping these engines operating at peak efficiency and reduces emissions, Section 1.2.

2. Evaluate the probable effects of testing reliability of such variables as

1. variation among manufacturers and models,
2. variations in atmospheric conditions,
3. the pass/fail rate over time.

The variation among manufacturers does not indicate that any automaker provides a superior product when it comes to meeting emissions standards. Data showing the failure rates over the past four years are shown in Section 2.1. The variation for a single maker from year to year is generally greater than the variation between makers.

Variations in atmospheric conditions, specifically temperature, pressure and humidity does not impact test results. Data and graphs in Section 2.2 show some variability in failure rates across the atmospheric conditions measured during tests, but no consistent bias is evident.

The pass/fail rate over time, with the same emissions limits, does not appear to change. Overall, about six percent of the tested fleet fails the emissions test. Section 2.3 contains tables provided by the Department of Motor Vehicles from their testing program. The clear change over time is the number of vehicles failing as they age. Current year cars fail at a rate of less than 0.10 percent, two year old cars at less than 0.4 percent and three year old cars at less than 0.7 percent while eight year old cars fail nearly 6.5 percent of the time. However, all of this is called into question by the discovery that equipment problems have affected the failure rates in the last year according to the Department of Motor Vehicles as presented to the Committee on October 26, 1999.

1. Evaluate the success of the program in effecting a reduction in atmospheric pollutants;

It is not possible to separate the effect of the state’s Inspection and Maintenance Program from the effect of vehicle emissions reduction mandated by the Federal Government. Air quality changes, discussed below, and emissions estimates generated by the Department of Environmental Protection demonstrate that emissions have been reduced and our air has become cleaner during the time the program has been in existence. However, that time period is the same as the period of mandated vehicle changes and we are unable to separate the effects. Changes in air quality and effects of the vehicle emissions testing program are discussed in Section 3.

2. Evaluate the predictive efficacy of the federal EPA MOBIL model in light of the Connecticut fleet composition;

Reviews on the ability of the MOBIL model to accurately reflect real world emissions are mixed. An evaluation of Canadian tunnel tests states that the model agrees with measured concentrations, while similar studies in the US find less agreement. Both agree that the age of vehicles in the test is important to the results. See Section 4.

3. Track changes in ambient air quality since 1986;

Annual trends show the peak concentrations decreasing regularly over the entire period of record. Section 3 is summarized from a final report submitted to the Department of Environmental Protection by a member of this committee Professor David R. Miller, and his colleagues

4. Propose cost effective alternatives or modifications to the program.

Remote sensing is being proposed by the Department of Environmental Protection for the new State Implementation Plan and we endorse this action. Clean screen should be included with the remote-sensing program. See the discussions in Sections 5 and 6.

Drop cars less than four years old from the inspection requirement.

Prepare for On Board Diagnostic testing in the future.

Mandate an external quality assurance function.

Compare the results of failed inspections with second inspections to demonstrate how much reduction is achieved from repairing failed vehicles.

The recommendations are discussed in Section 5. Section 6 contains a cost/benefit analysis of the existing program.

Section 1

Inquiry a.

Analyze the nature of the fleet in terms of age distribution, extent of manufacturer installed antipollution devices, federal emissions standards required of the manufacturer, and ability to alter tailpipe emissions through repair/corrective action.

1.1 Age Distribution

The age distribution of the inspected fleet of vehicles in Connecticut during the period from 1994 to 1998 is illustrated on the accompanying Figure 1.1.

Approximately two percent of the fleet was 20 years of age or older. The median age of the inspected vehicles is 6+ years. The age distribution has not changed significantly with time for the five-year period for which the Department of Motor Vehicles was able to provide data.

The Emissions Division, Department of Motor Vehicles provided the data in this section.

Figure 1.1 Distribution of ages of vehicles in the tested fleet.

1.2 Manufacturer’s Requirements

Tailpipe emissions from modern internal combustion engines:

The principal pollutants regulated from gasoline burning internal combustion engines are nitrogen oxides (NO/NO₂), carbon monoxide (CO) and unburned hydrocarbons (UHC). In order to monitor the emission levels of these pollutants from vehicles and correct the operation of those vehicles with excessive emission levels, state of Connecticut has been administering a vehicle emission-testing program in compliance with U. S. EPA regulations. Currently, all vehicles with gasoline burning internal combustion engines are subjected to an emission test regardless of vehicle’s age. The main objectives of this CASE study is (1) to assess the status of the current emission monitoring program in light of the changes in the vehicle pool age and manufacturer installed pollution prevention devices and (2) to make any recommendations on the existing test program especially in terms of its cost effectiveness and the benefit/expense relationship.

Background on Current Emission Control Devices:

The car manufacturers have met the ever-stringent emission limits on NO_x, CO and UHC emission mainly by installing three way catalytic converters in vehicle exhaust systems. It has been successfully demonstrated and implemented that all three pollutants can be reduced in concentration from their raw engine exhaust levels by means of Platinum/Rhodium (Pt/Rh) catalysts supported on a ceramic monoliths as shown in Fig. 1.2.

Figure 1.2 Schematic of catalytic converter for spark-ignited

Engine emission control (taken from Ref. 1)

The proper operation of the catalyst at high catalyst efficiency is dependent on the air/fuel ratio of the combustion process with the peak efficiency occurring around the stoichiometric air/fuel ratio as shown in Fig. 1.3. Therefore, it is important to operate the engine in a narrow band of air/fuel ratios around the stoichiometric mixture as well as using unleaded-fuel to prevent catalyst poisoning. For this reason, the engine fuel/air mixture management has been one of the important aspects of new engines. Because of the poor ability to accurately meter the fuel into the air stream in carburated engines, direct fuel injection with computer controlled injection timing and a feed back control scheme has become the current norm. The feedback on the engine stoichiometry control has been implemented in the way of sensing the oxygen

Figure 1.3. Catalyst conversion efficiency for NO, CO and HC for

a three way catalyst as a function of exhaust air/fuel ratio

concentration in the raw exhaust and providing this to the engine management computer as an input. For example, if the engine begins to run fuel-lean and the oxygen detector begins to measure high concentrations, engine computer re-adjusts the fuel injection parameters to bring its operation closer to stoichiometric.

In order to help assess the vehicle pool emission test performance, it is instructive to study the vehicle age and fail rate relationship. Figure 1.4 shows the fleet age from Connecticut DMV listed for years 1994 through 1998. It is found that vehicles up to 8 years old constitute between 63 % (1998) to 73 % (1994) of the total vehicle pool tested. Distribution among different ages is fairly flat. The vehicles older that 8 years constitute between 37 % (1998) and 27 % (1994) of the tested vehicles. If one looks at the inspection fail rate discussed in Section 2.3, it is clearly seen that the older vehicles, especially those which are four years or older have higher rates of failure of the emission tests. This agrees with the fact that computer controlled fuel-injected engines became the standard around 1993.

Figure. 1.4. Age distribution of tested vehicle fleet (DMV data)

Alteration of Vehicle Emissions by Repair:

For the new generation of vehicles with three way catalytic converters and computer-controlled engines, the repairs can be performed to restore the engine operation to stoichiometric fuel air mixtures. The possible failures may include malfunction of the oxygen sensor, fuel injection system, ignition system, and/or computer microprocessor. With the exception of oxygen sensor malfunction, most of these malfunctions will lead to noticeable performance change that will most likely prompt the driver to take the car for service. For older vehicles unequipped with the above-mentioned technology, many different scenarios may lead to excessive emissions that may go undetected for long periods of time. Almost all repair facilities do not have the capability to test for pollutant emissions as the necessary instrumentation is both capital and maintenance intensive. For example, the dynamic testing (i.e. vehicle operating on a chassis dynamometer) required by current EPA regulations is too expensive to be implemented at any repair facility including those of major car dealerships. Therefore, the option of emission testing at car repair facilities seems to be very unlikely at the present time.

Cars manufactured in the future will be subject to more stringent emissions limits if the proposed regulations are adopted. A history of emissions standards and proposed future standards is contained in Attachment 1.

Recommendations:

Based on the DMV emission data statistics and the vehicle age group distribution, it appears that testing should focus on the older portion of the fleet, which has a higher failure rate. As the older vehicles, which do not have the state-of-the-art engine emission technology, get purged out of the fleet, the test frequency should be re-evaluated. However, currently available data suggest that there is no reason to subject vehicles younger than three years to emission testing.

Conclusions

The age distribution of the inspected fleet has not changed significantly over the years from 1994 to 1998. Twenty percent of the fleet is three years or less in age and twenty percent is eleven years or older. However, we are told that nationally (EPA420-F-99-017, May 1999, Attachment 1) half the vehicles sold in the past few years are SUVs, which have greater emissions limits and achieve fewer miles per gallon of fuel than automobiles. The Federal Government has been tightening emissions limits for vehicles and EPA proposes new limits in the 2004 to 2009 time period. The old limits for nitrogen oxide and the proposed new limits for cars and light trucks, including minivans and SUVs are listed in the tables in Attachment 1. These are the requirements on the manufacturer.

Catalysts were introduced in autos in 1974. Closed loop computer systems were introduced in 1980. Electronic fuel injection and ignition were introduced in the middle 1980s and improvements have continued in motor efficiency and control of emissions through the 1990s. On board diagnostic assessment of engine performance may be used in 1995 and later vehicles to check for performance and related emissions.

The ability to alter tailpipe emissions through tampering is severely limited in computer controlled engines. There is no reason to tamper because the engine will not operate as well. Repair/corrective action is effective in keeping these engines operating at peak efficiency and reduces emissions.

References:

1. J. B. Heywood, Internal Combustion Engine Fundamentals, McGraw-Hill, 1988

2. DMV data statistics on emission test results

Section 2

Inquiry

Evaluate the probable effects of testing reliability of such variables as

1. variation among manufacturers and models,
2. variations in atmospheric conditions,
3. the pass/fail rate over time.

2.1 Variation Among Manufacturers

Test results from 1995, 1996, 1997 and 1998 were examined. Failure rates were listed by maker and model year. Seven major carmakers are represented in the tables that follow.

The failure rates for each make is similar to the others, except possible for Honda vehicles prior to 1998. The tables clearly indicate that age of the tested vehicle is much more important to yielding increased failure rates than is the make of the vehicle.

Test changes in 1998 are reflected in the lack of new models tested. The state went to a two-year period between tests and the contractor started to change equipment late in the year. Some models appear to fail more often in 1998 than previously, especially Honda.

We will examine the failure rate with age in a later section.

Vehicle Fail Rate
By Manufacturer
1995

	Chrysler		Ford		GMC		Nissan
Veh.Year	No. Test	% Fail	No. Test	% Fail	No. Test	% Fail	No. Test	% Fail
1996	12	0	21	0	28	0	9	0
1995	3093	0.06	9300	0.06	8544	0.64	3010	0
1994	7381	0.17	27862	0.1	26769	0.99	8178	0.4
1993	3990	0.6	21611	0.32	14487	1.46	4912	1.03
1992	3801	0.086	16660	0.55	16258	2.98	3564	1.17
1991	2899	1.96	13867	1.17	13924	5.93	2874	1.91
1990	4463	4.57	14075	2.02	15723	4.47	3973	2.84
1989	5217	3.62	18421	3.53	21810	6.63	2763	6.26
1988	6862	4.96	20103	6.13	25783	7.48	3112	12.08
1987	7110	8.52	15421	9.73	23048	8.64	5767	6.95
	VW		Toyota		Honda
Veh.Year	No. Test	% Fail	No. Test	% Fail	No. Test	% Fail
1996			7	14.28	6	0
1995	950	0	1893	0.1	2353	0.04
1994	2211	0.36	9463	0.18	8769	0.03
1993	1122	0.71	5834	0.29	5056	0.13
1992	1499	0.73	6726	1.17	6246	0.32
1991	1349	1.4	5932	0.69	5294	0.49
1990	1689	0.94	6675	1.33	5967	1.25
1989	1615	3.34	6139	2.45	5379	2.34
1988	2238	4.78	6678	12.08	6777	2.18
1987	2244	7.21	6223	4.09	4398	2.93