Assessment of nontailpipe hydrocarbon emissions from motor vehicles

This report evaluates tailpipe and nontailpipe hydrocarbon (HC) emissions from light-duty spark-ignition (SI) vehicles. The sources of information were unpublished data sets, generated mainly from 1990 through 1994, on emissions from volunteer fleets of in-use vehicles in chassis dynamometer and sealed housing for evaporative determination tests, and published chemical mass balance (CMB) source apportionments of HC in roadway tunnels and in urban air. The nontailpipe emissions evaluated comprise running-loss, hot soak, diurnal emissions, and resting-loss emissions. Relations between pressure and purge test failures and actual nontailpipe emissions were also examined. According to the recruited fleet data, nontailpipe emissions exceed tailpipe HC emissions by a wide margin. This is contradicted by real-world ambient and roadway tunnel CMB results, which attribute 65-93% of motor vehicle non-methane hydrocarbons (NMHCs) to tailpipe emissions, and the balance from the nontailpipe. Running-loss emission rates were critically dependent on driving cycle and conditioning. They decreased steeply with increasing vehicle speed, according to the fleet data. They increased with ambient temperature and fuel Reid vapor pressure (RVP) at rates of similar to 7%/degrees F and 46%/psi. Hot soak, diurnal, and resting-loss emission rates all increased with increasing ambient temperatures, at rates in the range of 2.2-4.6%/degrees F. Hot soak and diurnal emission rates increased with increasing fuel RVP, at rates between 34 and 47%/psi increase in RVP (at 95 degrees F ambient temperature). Vehicle-to-vehicle variation in HC emission rates was very large in all nontailpipe (and tailpipe) emissions categories. For each emission category 10% of the vehicles produced similar to 50% of the emissions. The dirtiest 10% of the vehicles in any one category, however, were not usually the same vehicles as the dirtiest 10% of the vehicles in any other category; the emission rates in any one category were uncorrelated with the emission rates in any other category. High emissions in every category-tailpipe, running loss, hot soak, diurnal, and resting loss-were seen in nearly all model years, including the newest ones, There appears to be a slow (similar to 10-15%/model year) downward trend in average hot soak, diurnal, and resting-loss emissions in successive model years (whether because of improving technology or because of vehicle age-or both-is not known). There was little relation of pressure or purge failure to actual hot soak, diurnal, or resting-loss emissions rates. The paper concludes with recommendations for resolving the two outstanding issues of (1) fleet versus tunnel/ambient and (2) running losses.