LBNL Report Number
In this report we examine the potential for additional emission reductions from adopting annual
rather than biennial testing in vehicle emission inspection and maintenance (I/M) programs. We
tracked a fleet of cars reporting for testing in two biennial cycles in the Phoenix IM240 program.
Using an assumption of linear deterioration following a passing I/M test, we found that annual
testing in the Phoenix program in 1996 would have resulted in an additional 45%, 48%, and 27%
reduction in HC, CO, and NOx emissions, respectively, over what occurred in the biennial
program. More stringent cut points under a biennial program would have resulted in greater
emission reductions, but with higher fail rates. Tighter cut points can also be applied to the
annual program. In fact, an annual program might ease the progression to tighter cut points by
resulting in relatively lower failure rates per test than biennial programs.
A small number of cars were identified that were given three I/M cycles, each roughly one year
apart, in Phoenix. This group of vehicles had higher overall emissions and may not be
representative of the whole fleet. However, these cars appear to have the same HC and CO
emissions, but slightly lower NOx emissions, after two years as cars tested biennially. This
result suggests that our assumption that emissions after annual testing would deteriorate at the
same rate as observed in biennial testing may be optimistic. The findings from our analyses
suggest that test-to-test emissions variability is a limitation of existing I/M programs, and is
preventing them from properly identifying some vehicles with broken or malfunctioning
emissions controls and ensuring that they are repaired. In theory, more frequent testing of
suspected high emitters could help address this problem. One strategy that may work is to use
remote sensing to identify suspected high emitters, and require that they come in for off-cycle
testing, as frequently as necessary.
Arizona remote sensing data provide additional evidence that a significant benefit results from
annual testing of older cars. Remote sensing measurements made in Phoenix of vehicles tested
biennially under the Phoenix enhanced IM240 program and of vehicles tested annually under the
Tucson two-speed idle program show interesting results. While newer vehicles from Phoenix
have lower emissions than those from Tucson, as one might expect from the enhanced program,
the older vehicles from Phoenix have higher emissions than those from Tucson. Phoenix
vehicles over seven years old had emissions 37% to 47% higher than Tucson vehicles of the
We conclude that more attention should be paid to promptly identifying and properly repairing
the high emitters found in the older portion of the vehicle fleet.