LBNL Report Number
The U.S. Environmental Protection Agency requires that states evaluate the effectiveness of their vehicle emissions inspection and maintenance (I/M) programs. This study demonstrates an evaluation approach that estimates mass emission reductions over time and includes the effect of I/M on vehicle deterioration. It includes a quantitative assessment of benefits from pre-inspection maintenance and repairs and accounts for the selection bias effect that occurs when intermittent high emitters are tested. We report estimates of one-cycle emission benefits of California's Enhanced Smog Check program, ca. 1999. Program benefits equivalent to metric tons per day of prevented emissions were calculated with a "bottom-up" approach that combined average per vehicle reductions in mass emission rates (g/gal) with average per vehicle activity, resolved by model year. Accelerated simulation mode test data from the statewide vehicle information database (VID) and from roadside Smog Check testing were used to determine 2-yr emission profiles of vehicles passing through Smog Check and infer emission profiles that would occur without Smog Check. The number of vehicles participating in Smog Check was also determined from the VID. We estimate that in 1999 Smog Check reduced tailpipe emissions of HC, CO, and NO(x) by 97, 1690, and 81 t/d, respectively. These correspond to 26, 34, and 14% of the HC, CO, and NO(x) that would have been emitted by vehicles in the absence of Smog Check. These estimates are highly sensitive to assumptions about vehicle deterioration in the absence of Smog Check. Considering the estimated uncertainty in these assumptions yields a range for calculated benefits: 46-128 t/d of HC, 860-2200 t/d of CO, and 60-91 t/d of NO(x). Repair of vehicles that failed an initial, official Smog Check appears to be the most important mechanism of emission reductions, but pre-inspection maintenance and repair also contributed substantially. Benefits from removal of nonpassing vehicles accounted for a small portion of total benefits. In 1999, more than 90% of all HC and CO benefits and over 80% of all NO(x) benefits were attributed to vehicles more than 10 years old, even though such vehicles represented only half of those tested in the program.