Remote Emissions SensinG                   Printer-friendly version

• Introduction
  • What is Remote Sensing Technology?
  • The Rationale for RST
• System Description
• Assessment
  • Key Results
  • Benefits
  • Costs
  • Implementation Challenges
  • Other Uses
• Where is Remote Emissions Testing Implemented?
• References

INTRODUCTION

What is Remote Sensing Technology?

Remote Sensing Technology is a system that identifies high-emitting vehicles as they drive on the highway. The system uses infrared spectroscopy to determine the chemical makeup of a vehicle's emissions, and video technology to record the license plates of offending vehicles. See our Telecommunications Diagram on Emissions Sensing for more information.

The exhaust from gasoline- and diesel-powered vehicles includes gases and particulates known to cause health and environmental problems. Carbon monoxide (CO) contributes to and exacerbates cardiovascular disease, and it is believed to contribute to the ambient greenhouse effect. Chemical reactions between nitrogen oxides (NO_X) and hydrocarbons (HC) produce ozone, which causes respiratory problems.

Highway vehicles are responsible for approximately 54 percent of all CO emissions in the US (Stephens et al.,1991). But different vehicles emit widely different amounts of CO. It has been consistently found that a small proportion of the vehicle fleet emits the majority of CO, NO_X and HC: approximately 10% of the vehicles emit 50% of the CO (Stephens, 1996); and similarly, 10% of vehicles cause almost half of all mobile NO emissions (Zhang et al, 1996). The same is true of emissions of hydrocarbons (HC). These figures indicate the importance of identifying the high emitting vehicles.

A sample of vehicles were ranked in order of increasing emissions, then divided into 10 groups of equal size. The height of the bar indicates the proportion of total emissions contributed by each decile.

Source: Lawson 1993; Klein and Koskenoja, 1996.

Traditionally, high-emitting vehicles have been identified through periodic mandatory vehicle inspections. Factors which cause high emissions include very high or very low air-to-fuel ratios during combustion, as well as faulty or missing catalytic converters. Periodic inspections are unnecessarily expensive, since all vehicles of a certain age need to be checked regardless of their emission performance. The inspection process is also easily circumvented, so that fraud is not uncommon. For example, some people remove the catalytic converter after the inspection, because it somewhat hinders engine performance.

Remote sensing of vehicle emissions would allow identification of super-emitters as they drive on highways, eliminating the opportunity for fraud and perhaps leading to earlier correction of the problem. In addition, the system is more cost effective than a normal inspection and maintenance program, since only the offending vehicles would require inspection.

Remote sensing for CO is based on non-dispersive infrared (NDIR) spectroscopy. It relies on the ratio of CO to CO₂ in the exhaust, since this ratio is an indicator of combustion efficiency. High ratios indicate incomplete combustion, which causes high levels of CO. The remote sensing system consists of an infrared light source, a detector module, and a video system to record vehicles' license plates.

The infrared source and detector module are set up on each side of a roadway section. The infrared source emits a light beam across the roadway, which passes through the emission smoke from the moving vehicles before it arrives at the detector module. The difference in intensity between the received and the emitted beams determines the ratio of CO/CO₂ in the vehicle plume (Stephens, 1991; Zhang, 1996). The video system is used to record the license plate numbers of vehicles passing the measurement system, so that the owners of vehicles exceeding the permissible level of emissions can be identified.

ASSESSMENT

Key Results

So far initial tests of remote sensing technology in Los Angeles, Toronto, and Houston have been only moderately successful. The system is fairly accurate in identifying high CO emissions, but has had less success in detecting high levels of hydrocarbons (HC) and nitrogen oxides (NOX). In addition, tests of matching license plates to vehicle records have yielded inconclusive results. The entire remote sensing system still needs technological improvements before it can be used as an emissions control device.

The conventional method used to measure vehicle emissions is a stationary analyzer. In Inspection and Maintenance programs, every registered vehicle must go to an approved garage to be tested with a stationary analyzer, either annually or biennially. This program often falls victim to fraud as many high-emitting vehicle owners "tinker" with the vehicle's emissions components to get it to pass the test, then return the emissions system to its original configuration after the test is complete, or they simply bribe the garage owner.

A remote sensing system, however, measures on-the-road vehicle performance, does not interfere with traffic, and does not require the driver's cooperation. There is little, if any, opportunity for fraud. RST also has the potential to generate revenue through fines.

In addition, it is capable of measuring emissions from a thousand or more vehicles a day, which makes it a useful tool for evaluating the vehicle emission implications of various traffic management strategies. The emission rates for CO and HC are related to the vehicle's instantaneous speed and acceleration rate, which will be affected by traffic control and management alternatives.

A remote sensing program is estimated to cost in the range of $18-60 million, while in Los Angeles, annual expenditures incurred under Smog Check, California's emission control program, total approximately $179 million per year. Additional costs could come from further technological development of the system and administrative costs such as ticket processing.

Remote sensing system is not as accurate as a stationary analyzer. The measurement difference these two can be as much as 10%. But perhaps its major disadvantage is that it only gives an instantaneous estimate of emission performance. It is well-known that the rate at which vehicles emit CO, NO_X and HC vary tremendously with vehicle acceleration and deceleration rates, cruising speed and engine temperature, in addition to being inherently variable. This limitation could be overcome by careful selection of measurement sites, and by selecting a high CO/CO₂ ratio as the maximum permissible emission standard.

The following table illustrates the problem of inherent variability in emission rates. Repeated measurements were taken on a fleet of vehicles. About half of the vehicles showed CO emissions below 1% all the time, while 10% to 25% showed emissions above the median local emission rate at least twice, and account for 50% to 70% of all emissions by the sample vehicle fleet. The remaining vehicles showed highly variable emissions across measurements.

Source: Steadman et al., 1991 & 1992

Remote sensing is more likely to identify a vehicle as a high emitter, when in fact it is not, than to fail to detect a high emitting car. A roadside test showed that all vehicles (10 total) which exhibited, according to the remote sensor, CO/CO₂ ratios below 2% also cleared the stationary analyzer. Of the 50 vehicles that exhibited CO/CO₂ ratios above 2%, 7 cleared the stationary test, and 15 were found in noncompliance of HC emissions, or to have tampered emissions control equipment (Austin et al, 1990).

An additional measurement problem is caused by short vehicle headways. Measurements taken from vehicles that follow high emitters include some residual exhaust from the lead vehicle, especially if the headway between the two vehicles is 1 to 2 seconds long, and perhaps even when as high as 3 to 4 seconds. The unfortunate consequence is that clean vehicles would give higher emission readings. Ambient conditions, such as wind speed, also affect measurements.

The remote sensing equipment does not detect high HC emissions as well as it detects high CO emissions. Vehicle exhaust typically contains over 100 types of HC, each with a different infrared absorption strength. Remote sensors fail to detect from one fourth to one third of HC (Stephens et al, 1996).

Similarly, NO (nitrogen oxide) detection equipment is not as well developed as that for CO (Stephens et al, 1996). Zhang et al. (1996) added an ultraviolet source and detector to their remote sensing system and collected data on NO in Denver in 1994. They found that the NO detector is capable of separating high emitters (above 2000 parts per million) from the low emitters (under 1000 parts per million).

The remote sensing technology is also likely to encounter institutional and administrative barriers. In California, it would have to demonstrate that it is at least as effective as the current Smog Check program. Deployment would probably require new legislation at the state level, and perhaps even at the federal level, for compliance with the 1991 Clean Air Act. Garage owners and others that currently perform smog checks have a vested interest in the status quo. And video imaging technology, as well as vehicle registration records, would have to demonstrate that they can deliver the accuracy required for successful enforcement. A test of matching license plates to vehicle records in Toronto yielded a match rate of 95% (Steadman et al 1992), while a test in Los Angeles resulted in only 60% matched plates (Steadman et al. 1991). To compound the problem, old-style license plates, more common in older and probably more polluting cars, are more difficult to identify than new-style plates.

While many public agencies have shied away from using remote sensing systems for emissions enforcement, researchers have used the technology as a data collection tool. Analysis of these data have yielded important findings about the circumstances which lead to high emitting vehicles, and particularly about the rates at which noxious gases are emitted depending on ambient factors, driving style, and other factors not specifically related to the emission control technology. Some of these results are described next.

Stephens et al. used data from Denver, Colorado in 1989 and found that the majority of CO is emitted by a small minority of all vehicles, and that most of the emissions are emitted by the older vehicles (pre-1980 vehicles). A more recent analysis using data from Michigan and California (Stephens, 1997) suggests however that emissions control technology, as indicated by model year, is more significant than age in explaining high emitter rates. The Denver study also showed that ambient factors such as altitude and temperature affect the measurements, which suggests that the remote sensing system would have to be calibrated for each particular site.

Yu and Burrier (1997) used a system called Smog Dog to measure emissions at five locations in Houston, Texas. They related the emissions to the vehicles' instantaneous speed and acceleration rate and found that conventional emission simulation models such as MOBILE5A underestimate the emission rates for on-road driving. These models were initially developed using emissions data collected in laboratory settings. It has also been found that emission patterns are strongly related to driving style, such as the rates at which people accelerate and decelerate (Gong, 1996).

RST is currently implemented in several areas in the U.S., including North Carolina, Texas, and New Mexico, as well as in Europe, including Budapest, Hungary.

In general, RST is used to complement an existing Inspection and Maintenance (I/M) program. In addition, RST can be used in situations where I/M programs only exist in specific areas. In New Mexico, for example, only certain counties have vehicle emissions standards, so RST is used to identify high-emitting vehicles from non-regulated counties that commute into counties with restrictions.

Austin T.C, T.R. Carlson and K.A. Gianolini. An Evaluation of Remote Sensing for the Measurement of Vehicle Emissions. Sacramento, CA: Sierra Nevada Inc, 1990. Report No. SR90-08-02.

Gong, R. and P.Waring. IR Remote Sensing System for Testing Urban Fleet Emission Profile. In: Urban Transport and the Environment for the 21st Century, 1995, pp. 229 - 236.

Gong, R. and P.Waring. Investigation of Emission Behavior in an Urban Driving Cycle with a Remote Sensing System. 1996.

Hickman A. J. and I.S. McCrae. Evaluation of a Remote Vehicle Emission Measurement System. Project Report 105. Crowthorne, Berkshire: Transportation Research Laboratory, 1995.

Klein D.B. and P.M. Koskenoja. The Smog-Reduction Road: Remote Sensing versus the Clean Air Act. Reprint from Policy Analysis, No.249, Feb. 1996. Berkeley, California: University of California Transportation Center, No. 301, 1996.

Lawson D.R. Passing the Test: Human Behavior and California's Smog-Check Program. Journal of the Air and Waste Management Association, Vol. 43, 1993.

Steadman D.H., G. Bishop, J.E. Peterson, P.L. Guenther. On-Road Remote Sensing in the Los Angeles Basin. Denver, Colorado: University of Denver, Chemistry Department, 1991.

Steadman DH, G. Bishop, J.E. Peterson, PL Guenther, S.P. Beaton, and I.F. McVey. Remote Sensing of On-Road Vehicle Emissions. Final Report. Denver, Colorado: University of Denver, Chemistry Department, 1992.

Stephens, R. D., and S.H. Cadle. Remote Sensing Measurements of Carbon Monoxide Emissions from On-Road Vehicles. Journal of the Air and Waste Management Association, Vol. 41, pp. 39 - 46, 1991.

Stephens, R. D., P.A. Mulawa, M.T. Giles, K.G. Kennedy, P.J. Groblicki, S.H. Cadle., and K.T. Knapp. An Experimental Evaluation of Remote Sensing -Based Hydrocarbon Measurements: A Comparison to FID Measurements. Journal of the Air and Waste Management Association, Vol. 46, pp. 148 - 158, 1996.

Stephens, R. D., S.H. Cadle, and T.Z. Qian. Analysis of Remote Sensing Errors of Omission and Commission Under FTP Conditions. Journal of the Air and Waste Management Association, Vol. 46, pp. 510 - 516, 1996.

Stephens, R. D., M. Giles, K. McAlinden, R.A. Gorse Jr., D. Hoffman, and R. James. An Analysis of Michigan and California CO Remote Sensing Measurement. Journal of the Air and Waste Management Association, Vol. 47, pp. 601 - 607, 1997.

Yu, L. and S.W. Burrier. Vehicle Emission Sensing and Evaluation Using the SmogDog in Houston. SPIE, Vol. 2902, pp. 219 - 230, 1997.

Zhang, Y., DH Stedman, G.A. Bishop, SP Beaton, P.L Guenther, and I.F. McVey. Enhancement of Remote Sensing for Mobile Source Nitric Oxide. Journal of the Air and Waste Management Association, Vol. 46, pp. 25 - 29, 1996.

Links

Remote sensing of emissions in Albuquerque
http://www.cabq.gov/aircare/rst.html

Remote sensing of emissions in Budapest

On-road testing of emissions in Texas

Author: Rebecca Pearson, Da-Jie Lin. Last Update: 04/01/01