Evaluation of the Effectiveness of Alternative Fast Pass, Fast Fail, and Retest Algorithms in the Phoenix IM147 Program

This report compares the current method of quickly identifying passing and failing vehicles in
the Phoenix IM147 program (“the Sierra method”) with an alternative method developed by the
Arizona Department of Environmental Quality (“the DEQ method”). There are several
limitations with the database of over 30,000 vehicles provided to us to make the comparison;
therefore our results should be considered estimates of the relative effectiveness of the two
methods.
The DEQ method results in a slightly lower failure rate, and a slightly longer average test time,
than the Sierra method, when applied to all vehicles that would complete testing under the DEQ
method. To estimate the number of false passes and failures, and the fraction of total excess
emissions identified under each method, we use a subset of 1,775 vehicles for which a full
IM147 test (three full traces) is available. The DEQ method results in a much lower false fail
rate, and identification of more excess CO emissions, than the Sierra method when applied to
this subset of vehicles. On the other hand, the Sierra method results in a slightly lower false pass
rate, identification of more excess HC and NOx emissions, and a shorter test time than the DEQ
method. The relative effectiveness of each method is consistent whether the cut points currently
in use in the Phoenix program, or more stringent cut points developed by EEA, are applied.

Final Report to the Arizona Department of Environmental Quality