Research Projects

Project at a Glance

Title: On-road remote sensing of CO and HC emissions in California

Principal Investigator / Author(s): Stedman, Donald H

Contractor: University of Denver

Contract Number: A032-093

Research Program Area: Emissions Monitoring & Control, Atmospheric Processes

Topic Areas: Field Studies, Mobile Sources & Fuels, Monitoring


The University of Denver remote sensor for on-road measurement of motor vehicle carbon monoxide and hydrocarbon emissions was used for 30 days in California in 1991. The resulting data set is the largest ever collected by a remote sensor emissions testing program. We made more than 130,000 measurements, resulting in 91,679 records with emissions and vehicle information (from the California Department of Motor Vehicles). We measured vehicles in a mix of many driving modes and speeds ranging from deceleration approaching a red traffic light through idling in heavy congestion to accelerations and cruises entering a freeway ramp at highway speeds. The remote sensing device measures the CO/CO2 and HC/CO2 ratios for one-half second behind each vehicle, from which the exhaust %CO and %HC are calculated. The mass emission rates in grams CO or HC per gallon of gasoline used can also be derived.

The study consisted of three phases; a series of controlled tests, a pullover study of high-emitters, and a series of measurements at a variety of sites around the South Coast Air Basin and northern California. The controlled tests included a blind comparison of remote sensor measurements to those made by an instrumented vehicle, and a series of tests of nearly two dozen vehicles under controlled conditions of cruise, acceleration, and deceleration. The pullover study was designed to investigate the ability of the remote sensor to identify high-emitting vehicles, during on-road conditions, for further roadside testing by a crew of California Air Resources Board and Bureau of Automotive Repair technicians. The third phase surveyed the fleet emissions at a variety of locations and under a variety of driving conditions. Vehicles that fail to participate in random roadside inspections appear to have much higher on-road emissions than those of participants. For this reason these studies should not be assumed to be "random".

During the controlled testing phase, the on-road measurements were compared in a blind test to those measured by a vehicle equipped with a tailpipe probe, trunk-mounted CO and HC monitors. and computer control of the vehicle's air/fuel ratio. Compared to this vehicle of known emissions, the remote sensing measurements are shown to be accurate within ±5% for CO and within ±15% for HC. We investigated inter-vehicle and intra-vehicle emissions variability by measuring the emissions of 23 vehicles under a variety of operating conditions. The most consistent emissions occurred for most vehicles at a steady cruise of 15-45 mph. The highest CO emissions occurred during hard accelerations, while the highest HC emissions occurred during decelerations. Hydrocarbon emissions were lowest during the acceleration modes.

The results of this study verify those found in previous CARB studies of CO emissions and extend the results to HC. On-Road hot exhaust emissions of both CO and HC are dominated by the 10%-20% of vehicles that are gross polluters, while the majority of vehicles in all model years are relatively clean. Gross polluters can be found in all model years, although their fraction increases in the older model years. The majority of the on-road emissions at the locations studied comes from vehicles less than ten years old. The pullover study is consistent with the previous study (Stedman et al., 1991b) and indicates that gross polluters identified by on-road testing have more than a 92% chance of failing a roadside Smog Check and that more than 60% have either tampered or defective emission control equipment. In comparison to a roadside IM240 we show that the remote sensor had a zero false failure rate.

Maintenance seems to be an important factor in mobile source emissions. The emissions of older well-maintained non-catalyst vehicles in Sweden are nearly the same as those of the equivalent fleet of originally catalyst equipped vehicles in Los Angeles. The primary difference between the two fleets appears to be the level of maintenance. The emissions of well-maintained non-catalyst vehicles in Sweden are higher, however, than the well-maintained catalyst-equipped Swedish vehicles in Los Angeles. The primary difference here is the emission control technology. Emission controls and maintenance are both required for low emissions of the on-road fleet.

These results are consistent with the idea that the beneficial effects of tighter new car emissions standards and reformulated fuels may be obscured by the emissions of a small fraction (10%-20%) of poorly maintained and tampered vehicles. Nearly all on-road gross polluters identified in 1991 had passed the biennial Smog Check. One explanation for this is that Smog Check fraud or outright cheating may be common. However, we also show that many high emitting vehicles have variable emissions. This latter result, which seems to be independent of the test procedure, allows owners to "pass the test" without repairing the vehicle.

As before we have shown that, assuming equal exhaust volumes, on-road emissions are dominated by a few gross polluters, and many vehicles emissions are negligible. For instance for 3,624 vehicles measured three or more times, 60% of the vehicles consistently emit less than 12% of the total CO and 50% of the vehicles account for less than 20% of the total HC emissions. On the other extreme are 3% of the vehicles which emit 23% of the CO and 27% of the hydrocarbon emissions. The presence of these gross polluters, the fact that many are not old cars, has implications bearing upon the cost effectiveness of any program which treats all vehicles, or all vehicles of a given age, as equally polluting.


For questions regarding research reports, contact: Heather Choi at (916) 322-3893

Stay involved, sign up with ARB's Research Email Listserver