United States              Air and Radiation         EPA420-F-98-023
 Environmental Protection                           May 1998
 Agency

 Office of Mobile Sources
Technical
Highlights
Clean Screening in  Inspection and
Maintenance  Programs
EPA has recently completed a draft guidance document on the use of
clean screening in Inspection and Maintenance (I/M) programs. Clean
screening is designed to exempt certain cars from the I/M requirement,
on the strength of other evidence of the high probability that they are
clean enough to pass anyway. Based on a preliminary assessment of
data currently available, EPA believes it is possible to excuse up to one-
third of cars from inspection each year, with only a 5 to 10 percent loss
in emission reductions.
Background
Clean screening is the term used to describe methods that states can use
to excuse cars from a scheduled I/M emissions test. Typical I/M pro-
grams require every car to appear at a testing station once each year or
every other year for a test of the vehicle's tailpipe emissions; some states
test other emissions components-such as the gas cap and the fuel lines-
as well. Over the last several years, vehicle emission systems have
become more reliable and durable, and many cars pass the I/M test.

Guidance Issued
EPA's draft guidance has been distributed for state comment and inde-
pendent scientific peer review, and EPA is accepting comments on the
guidance until July 11, 1998. Once EPA has reviewed the comments and
incorporated them as appropriate, the guidance will be finalized in the
fall of 1998.
                                           > Printed on Recycled Paper

-------
The options included in the draft guidance
are aimed at making I/M programs more
efficient and cost effective for states and
car owners, by focusing inspections on
vehicles more likely to be high emitters in
need of repair.  There will be some increase
in vehicle emissions as a result of the
changes in these I/M programs. It is
important to note that the estimates in-
cluded in the draft guidance are based on
the latest and best information currently
available. However, EPA expects at least
certain aspects of the estimates to change
somewhat before they are finalized. Once
the clean screening guidance is finalized,
EPA may revise the estimates from time to
time as better information and understand-
ing of clean screening options becomes
available.

The draft guidance provides descriptions
and projected emission impacts for three
types of clean screening, which differ with
respect to the type of evidence required to
exempt cars from regular I/M. These are
listed below:
Remote Sensing Clean
Screening

Overview
In this concept, roadside remote sensing is
used to excuse specific cars from their next
annual or biennial I/M test because they
pass remote sensing. Obviously, this has
owner convenience and cost advantages.
After skipping one scheduled test, cars
would have to get the next one, unless they
again pass a remote sensing test. In practi-
cal application, cars would have to pass
two different remote sensing tests, at
different times or places, and within a
limited time window (e.g., no more than
twelve months before the scheduled I/M
test).

Effect on vehicle population
The number of vehicles excused from
testing depends on the stringency of the
remote sensing test and the fleet coverage
of the remote sensing program. Up to fifty
percent coverage of the fleet has been
demonstrated; more should be possible
with resources and experience. A 50 per-
cent pass rate on the remote sensing test
and 80 percent fleet coverage for example
would mean that about 40 percent of the
fleet would be excused each year. This is
likely to be the highest pass rate and
coverage an area should consider. An
exemption fraction of one-third may be
more typical.

SIP credit ramifications
Use of remote sensing for clean screening
will typically reduce the credit ascribable
to the I/M program because some cars with
high tailpipe emissions may appear clean in
a remote sensing test and will be excused
from I/M tailpipe testing and repair for that
I/M cycle. Also, remote sensing cannot
identify low versus high emitting vehicles
with respect to evaporative HC emissions.
However, older cars have a much higher
incidence of evaporative problems than
newer cars, and remote sensing clean
screening tends to fail most older cars
because they so frequently have defective
parts causing  high tailpipe emissions.
When such older cars with both high
tailpipe and high evaporative emissions
report for their regular I/M test, having
failed to pass  the remote sensing tailpipe

-------
test, both problems can be identified and
fixed. Therefore, the loss of evaporative
HC credits is much less than the portion of
the fleet exempted, but may be as or more
significant than the loss of tailpipe credit.

An analysis performed for one state ex-
ample indicates that by using remote
sensing cutpoints that excuse 37 percent of
remotely tested vehicles  with a fleet cover-
age of 80 percent, the loss in tailpipe HC
benefit is about 4 percent and the loss in
overall HC benefit is 5 to 8 percent. By
using remote sensing clean screening to
measure NOx with a tight NOx cutpoint,
NOx benefit losses can be limited to 6
percent, for the 80 percent coverage ex-
ample. If only HC and CO cutpoints are
used,  the NOx benefit loss can be as high
as 22  percent.

However, it is important to note that use of
remote sensing clean screening for NOx
reduces the number of cars that are excused
from regular testing. The effectiveness of
remote sensing varies among cars of
different model years, so the overall effect
will depend on the calendar year of interest
and the mix of cars on the road.
Vehicle Emissions Profiling

Overview
This is a method of ranking vehicles for the
likelihood that they need emissions repairs,
using statistics on the historic failure rate
of vehicles of very similar design. For
example, failure rates of vehicle models in
the Arizona IM240 program can be used to
predict whether a certain group of nearly
identical cars in another state will have a
high or low failure rate. This information
can be used in either of two ways, or in
both ways.

In Low Emitter Profiling, the ranking
would be used as a clean screening tool to
determine what cars should be excused
from testing. In High Emitter Profiling, the
ranking is used to require some type of
special testing regime for the cars most
likely to need repair. This special testing
could involve special testing stations,
annual instead of biennial testing, etc. It is
possible to add other information into the
ranking process. For example, remote
sensing readings or the most recent I/M
outcome of the specific vehicle in question
could be used, assuming such data is
available.  Generally, extra information can
improve the ranking's accuracy, and
thereby reduce the credit loss from screen-
ing out a given number of vehicles from
their next regular I/M test.

Effect on vehicle  population
This varies and is dependent on a number
of factors, including the exemption frac-
tion, the fleet mix, and the information
used in the vehicle ranking process. For
example, by expanding the ranking infor-
mation used, states can fine-tune their
profiling. Thus, the effect on the vehicle
population can be customized to a certain
degree-dependent on the fleet mix and
sophistication of the vehicle ranking
process-within the goal of clean-screening
a specific percentage of the  vehicle popula-
tion. Credit losses increase quickly when
the exemption fraction is higher than 50
percent.

-------
Credit ramifications
Low emitter profiling carries a credit loss
due to the fact that some cars will be
incorrectly clean screened out of traditional
I/M testing. In general, recent studies
indicate that vehicle profiling can support
clean screening with roughly the same
immediate emission credit ramifications as
remote sensing (i.e., in the range of 5 to 8
percent loss in credit) for all pollutants.
This, however, is also dependent on  the
specific profiling utilized by the state. The
draft guidance contains a special note
regarding the  preliminary and evolving
nature of the estimates of credit losses with
low emitter profiling.
Model Year Exemptions

Overview
This approach exempts cars until they
reach a certain age, on the premise that
virtually all cars are clean when sold and
most remain clean for at least several years.
Many states now wait until a car is four or
five years old before the first required
inspection.  Some give the owner the option
of inspection during these years. It is the
state's choice whether to assess a program
fee on new cars while they are exempt
from testing.

Effect on vehicle population
Again, the effect varies, given the fleet mix
(i.e., vehicle age distribution) of a specific
area. EPA recommends that states exempt
cars  from testing until they are at least four
years old.

Credit ramifications
The effect of model year exemptions in
specific I/M areas can be estimated using
existing features of MOBILESb. States
with a simple system of scheduling inspec-
tions on the anniversary of a new vehicle
sales date should have no difficulty analyz-
ing scenarios for themselves. Overall, the
loss of credit for exempting the four newest
model years is quite  small, and EPA be-
lieves most states can find ways to com-
pensate for it if needed to keep the SIP
approvable. States with more complex
scheduling may need to consult EPA.
For More Information
A copy of the draft guidance and related
information is available electronically from
the EPA Internet server at:

  http://www.epa.gov/oms/models.htm

For further information on the draft guid-
ance, please contact Joe Somers at:

  U.S. Environmental Protection Agency
  Assessment and Modeling Division
  2000 Traverwood Drive
  Ann Arbor, MI 48105
  Phone:(734)214-4321
  somers.joseph@epa.gov

-------