Statistical modeling of vehicle emissions from inspection/maintenance testing data: an exploratory analysis

The development of a prescreening model to identify failed and gross polluting vehicles

“…Consistent with Wenzel and Ross (1998), Washburn et al (2001), Bin (2003), and Beydoun and Guldmann (2006), as odometer mileage increases, the likelihood of being identified as a failed vehicle or a gross polluter increases. To create the model year dummy variables, model years were classified into five groups based on emission control technology implementation as cited, in part, by Cadle et al (1999) and using CARB's existing EMFAC model categorization (CARB, 2003): pre-1980 model years (oxidation catalysts or no catalytic converters); 1980-1985 model years (three-way catalysts), 1986-1990 (known to be equipped with fuel injection engines and improved engine control units), 1991-1995, and post-1995 (second-generation on-board diagnostics) as a base category.…”

“…It is difficult to compare these results to previous findings because past studies often pooled vehicle manufacturers into categories (i.e., foreign versus domestic) due in part to smaller sample sizes. Nonetheless, Wayne and Horie (1983), Kahn (1996), Ross et al (1998), Washburn et al (2001) and Bin (2003) all found significant relationships between manufacturers and emission rates.…”

“…Many of these variables have been identified in the literature as being correlated with high-emitting vehicles. For example, vehicle characteristics, such as vehicle age (or model year), engine size, number of engine cylinders, odometer reading, and use of oxygenated fuels, have been associated with higher emissions or higher failure rates (Wayne and Horie, 1983;Kahn, 1996;Washburn et al, 2001;Bin, 2003). Other technology-based relationships that have been explored include those between the failure rates and repairs of specific emissions control system components, such as the catalyst, oxygen sensor, or exhaust gas recirculation (EGR) system, and higher emissions (Lawson et al, 1996;Heirigs et al, 1996;Wenzel and Ross, 1998).…”

“…Chevrolet, Ford, Nissan) with higher emissions (Wayne and Horie, 1983), or associated groups of manufacturers (e.g. domestic versus foreign) with higher emissions or higher failure rates (Washburn et al, 2001). However, even the distinction between foreign and domestic may not be sufficiently refined enough (Wenzel and Ross, 1997); emissions of a given manufacturer's vehicles can also vary substantially by model or engine family .…”

“…Washburn et al (2001) suggested that the use of alternative fuels may be associated with failing a tailpipe test, but vehicles from the California smog check program were overwhelmingly gasoline (99.9% of the sample). There was no theoretical justification for including vehicle body type or transmission type because they have no influence on the vehicle emissions that cannot be explained by engine variables, such as engine size or and number of cylinders.…”