The objective of this study was to begin to quantify the benefits of a smoke opacity-based (SAE J1667 test) inspection and maintenance program. Twenty-six vehicles exhibiting visible smoke emissions were recruited: 14 pre-1991 vehicles and 12 1991 and later model year vehicles. Smoke opacity and regulated pollutant emissions via chassis dynamometer were measured, with testing conducted at 1609 m above sea level. Twenty of the vehicles were then repaired with the goal of lowering visible smoke emission, and the smoke opacity testing and pollutant emissions measurements were repeated. For the pre-1991 vehicles actually repaired, pre-repair smoke opacity averaged 39% and PM averaged 5.6 g/mi. NOx emissions averaged 22.1 g/mi. After repair, the average smoke opacity had declined to 26% and PM declined to 3.3 g/mi, while NOx emissions increased to 30.9 g/mi. For the 1991 and newer vehicles repaired, pre-repair smoke opacity averaged 59% and PM averaged 2.2 g/mi. NOx emissions averaged 12.1 g/mi. After repair, the average opacity had declined to 30% and PM declined to 1.3 g/mi, while NOx increased slightly to 14.4 g/mi. For vehicles failing the California opacity test at >55% for pre-1991 and >40% for 1991 and later model years, the changes in emissions exhibited a high degree of statistical significance. The average cost of repairs was 1088 dollars, and the average is very similar for both the pre-1991 and 1991+ model year groups. Smoke opacity was shown to be a relatively poor predictor of driving cycle PM emissions. Peak CO or peak CO and THC as measured during a snap-acceleration were much better predictors of driving cycle PM emissions.

No abstract

A "piggyback" approach is used to characterize aerosol emissions to obtain input for large-scale models of atmospheric transport. Particulate and gaseous emissions from diesel trucks, light-duty vehicles, and buses were measured by the Bangkok Pollution Control Department as part of the Developing Integrated Emissions Strategies for Existing Land Transport (DIESEL) project. We added filter-based measurements of carbonaceous composition, particulate light absorption, and water uptake. For 88 "normal" diesel vehicles (PM emission rate < 4.7 g/kg), our best estimate of the average PM2.5 emission rate is 2.2 +/- 0.5 g/kg, whereas for 15 high emitters, it is 8.4 +/- 1.9 g/kg. The effect of Euro standards on PM emission rates was apparent for heavy-duty vehicles, but not for light-duty vehicles. Carbonaceous composition appears relatively consistent, with particulate (artifact-corrected) OC at 17 +/- 1% and EC at 40 +/- 8% of PM for 103 pickups, vans, heavy-duty trucks and buses. The median absorption cross-section for EC is 10.5 m2/g at 532 nm. The history of average emission rate and chemical composition during the project suggests that about 25 vehicles can provide a regional PM emission rate for normal vehicles. Other studies such as remote sensing measurements will be required to estimate the important contribution of high-emitting vehicles.

U.S. petroleum use today is 2 million barrels per day lower than it would have been if automobile fuel economy had not improved since 1975. This paper explores the potential for and cost of further increases in domestic passenger car fuel economy using market-ready technologies and sales mix shifts. Using technology already included in manufacturers’ production plans and based on consumers' willingness to pay for increased fuel economy, domestic auto mpg could be increased from the 1987 level of 27 mpg to 31.6 mpg in 1995 without reducing vehicle size or performance from 1987 levels. By 2000, 34.3 mpg can be justified on the same basis. A higher level, 36.4 mpg, is cost-effective, based on fuel cost savings over the entire expected vehicle life. The maximum level achievable with the technology included in this analysis is 39.4 mpg, but this level would not be cost-effective. Sales mix shifts stimulated by price subsidies for efficient cars and surcharges on inefficient models can cause about 1 or 2 mpg of higher fuel economy before becoming too costly.

No abstract