EPA's Proposal for MOBILE6
               Exhaust Emissions Due to Engine Start

                         March 7, 1997


Abstract

     MOBILE5 and earlier versions of the MOBILE model combine the
effects of engine starts and transient driving into a single
estimate of exhaust emissions.  This approach makes it difficult to
adjust the number of trips (and therefore starts) associated with a
fixed amount (miles) of vehicle driving.  It also requires the model
to apply adjustments to emissions equally to emissions from driving
and starts.  Furthermore, States now need the capability to model
location-specific emissions impacts from engine starts.  In response
to these needs, EPA proposes MOBILE6 is planned to have a separate
estimate for emissions from vehicle driving and engine start.


Summary of Proposal

EPA proposes that engine start emissions will have the following
qualities:

  - The magnitude of start emissions will not depend on speed or
  driving cycle.

  - Engine start emissions will be handled as instantaneous, key-on
  emissions (independent of vehicle miles traveled) in units of
  grams per start.

  - Engine start emissions will be characterized by emission
  standards and technology.

  - Engine start emissions will be adjusted to account for soak
  periods less than 12 hours.

  - There will be separate adjustments for engine start emissions to
  reflect different temperatures and fuels.

  - Default activity (starts per mile) will be used, but can be
  overridden by user input.

  - Start emissions can be reported either separately from the
  exhaust running emissions and/or combined with exhaust running
  emissions (in gram per mile) using the number of trips per mile.

Supporting Data

     EPA recently conducted a small test program to investigate the
effect of different driving cycles on engine start emission
estimates.  Preliminary analysis of data from five vehicles, tested
at EPA's Ann Arbor laboratory, indicates that driving cycle, for
typical driving situations, may not be a significant factor in the
magnitude of engine start emissions.  This supports the contention
that the emission estimate for engine starts can be determined with
some confidence from any of the standard driving cycles.

     For purposes of MOBILE6, EPA has defined "start emissions" as
the emission effect of a 12-hr soak between key off and key on. (a
"cold" start) including the engine start.  The Federal Test
Procedure (FTP) and the California Unified Cycle (LA92) both include
a cold engine start. EPA proposes to extract the emission effects of
engine start for MOBILE6 from the FTP results. Most of the existing
LA92 data was collected by the California Air Resources Board in
California using California fuels and may not be representative of
Federally certified vehicles.

          Testing underway in Ann Arbor (25 vehicles) and by an EPA
contractor in Ohio (50 vehicles) will include the FTP and the FTP
505 cycle (bag 1 or bag 3) without an engine start. EPA proposes to
develop a functional relationship between the three standard FTP bag
emissions and the effect of engine start from this data.

     
Methodology

Extracting Start Emissions from the FTP

     Cold start emissions are contained in the first 505 seconds of
the FTP (bag 1) and represent the effect of an engine start after a
12 hour soak.  The start effect can be separated from the effect of
running emissions by determining the difference in emissions
measured on the identical driving cycle, but without an engine
start.  The last 505 seconds of the FTP driving cycle (bag 3) is
identical to the first 505 seconds, but still contains an engine
start after a 10 minute soak ("hot start").  

     As described above, EPA is testing vehicles this year
specifically to determine the effects of engine starts on emissions.
Using this data a relationship between engine start emissions and
the FTP results can be determined.  This information can be used to
develop a general relationship to estimate the cold start emissions
from the bag emission measurements of the FTP.  A general form of
the relationship might be:

       Cold Start = FTP Bag 1 - ( FTP Bag 3 - Hot Start )

Adjustments to Cold Start Emissions

  o The effect of soak time on start emissions is known to be
  significant.  California has completed a study on this effect for
  their emission model.  EPA proposes to use this study directly.

  o EPA will re-evaluate adjustment factors for temperature and
  fuels in MOBILE5 which were developed for the full FTP rather than
  separately for engine start versus running emissions effects.  EPA
  proposes to assume that the effect on bag 1 of the FTP be
  considered as the effect for engine start emissions.  In some
  cases it may be necessary to assume that the effect is identical
  for engine start and running emissions.

  o EPA will analyze the available data from instrumented vehicles
  to determine the default number of engine starts per mile of
  vehicle operation and the soak time distribution. These default
  values may be changed through user input.

  o An analysis will be done to determine the effects of technology,
  standards and vehicle age on engine start emissions.

  o Some FTP data collected includes a measurement of methane for
  each bag.  This information will be used to determine the fraction
  of methane in each bag.  EPA proposes to use the fraction of
  methane in bag 1 to represent the methane emissions for engine
  start emissions.

Outstanding Issues

     Heavy Duty vehicle testing data do not have bag emissions.
MOBILE5 did not include operating modes for heavy duty vehicles.  It
may not be possible to have separate engine start emissions for
heavy duty vehicles in MOBILE6.