Real-World Emissions from Model Year 1993, 2000, and 2010 Passenger Cars

Automotive emissions of carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NOx) are major contributors to metropolitan air pollution in the US. Important progress has been made in the past quarter century to reduce these emissions, but in some regions much more progress is needed. Moreover, increased driving will destroy even the progress that has been made unless the emissions in grams per mile of driving are reduced further.

Exhaust emissions may be viewed as consisting of three major sources: on-cycle, off-cycle, and malfunction emissions. On-cycle emissions are those emissions occuring under the driving conditions in the new car certification tests, while off-cycle emissions are caused by driving at higher power than required by the tests. Malfunction emissions are caused by the malfunction of on-board emissions control systems (ECS). Two other sources of real-world emissions from vehicles are fuel-related: fuel evaporation and upstream fuel processing. This report estimates the contribution of all six of these sources, but focuses on the two sources that are the result of "loopholes" in the current emissions control program: off-cycle and malfunction emissions.

We determine the average lifetime grams-per-mile (g/mile) emissions of a model-year (MY) 1993 car through detailed analysis of emission measurements from a variety of data sources. CO and HC emissions are about five times higher than test levels, and NOx may be about twice as high. We then predict the g/mile emissions for MY2000 and MY2010 conventional gasoline-fueled cars. Significant reductions are possible from regulatory changes and improved technology.

Off-cycle emissions occur in part because of the sensitivity of NOx to power, and in part due to a practice called command enrichment, the injection of excess fuel (beyond that needed for combustion) at high power. Supplemental regulatory tests have been proposed by the Environmental Protection Agency to motivate improvements in ECS design to further reduce NOx, and to limit the frequency and strength of command enrichment. Assuming effective regulations are adopted in this area, we forecast reductions of about two-thirds for all three pollutants.

Malfunction emissions occur in modern cars because some models do not have robust ECS. This is a new insight; current regulations are based on the assumption that ECS malfunction is fundamentally the fault of individual owners, drivers, or mechanics, caused by either poor maintenance or disconnection of ECS components. This assumption is less convincing with today's modern cars, which require little maintenance and provide little incentive to tamper.

This result supports mounting criticism that inspection and maintenance (I&M, or smog check) programs aimed at repairing malfunctions in individual vehicles are stop-gap at best. In contrast, if new information technologies, such as remote sensing and on-board diagnostics, are developed and used to identify malfunction-prone models and to motivate the design and manufacture of models with robust ECS, then substantial emissions reductions can be achieved. Our prediction of a two-thirds reduction of malfunction emissions is based on the fact that most, and perhaps all, manufacturers have virtually eliminated malfunctions in at least some of their models.

Policies to reduce off-cycle and malfunction emissions are discussed, including suggestions for regulators and manufacturers to evaluate their emissions programs in terms of real-world emissions, measured from in-use, rather than laboratory-tested, vehicles. Although non-automotive emission sources also need attention, the potential reductions in automotive loophole emissions are probably among the most cost-effective for our metropolitan areas.