The propane chemistry submechanism of GRI-Mech was designed to meet the
following objectives:

<UL>
<LI>
correct accounting for ignition delays and flame properties in natural
gases containing propane, including the effects of pressure and
composition;

<LI>
complete conversion of propane to combustion products, with an
appropriate dependence on equivalence ratio;

<LI>
adhering to the GRI-Mech practice of accepting only carefully-analyzed
elementary reactions;

<LI>
minimizing the scope of the propane chemistry.

</UL>

<P>
The contradictory demands of (1-3) and (4) were met by the following
compromise. Because most of the detailed chemistry of propane pyrolysis and
oxidation arises from secondary  reactions of C3Hn, n < 7, species, the
artifice was adopted of forcing thermal decomposition of C3H7 to C2H4 and
CH3 to be the only unimolecular reaction of C3H7, and considering that the
radical-radical reactions of C3H7 lead only to C2 species.

<P>
Under these  assumptions, radical attack on propane effectively occurs only
at the 6 primary hydrogens and only the n-C3H7 radical participates in
further C3 chemistry.  An attempt was
made to include all plausibly important bimolecular reactions of C3H7 that
could lead to C2 products.

<P>
The GRI-Mech set of C3-reactions can therefore not be expected to describe
"real" C3 chemistry under combustion conditions, for dehydrogenation and
oxidative dehydrogenation reactions, and propene and allene, surely do play
prominent roles, especially under fuel-rich conditions.  Similarly, the
adjustment of C3H7 rate coefficients and products has to be regarded as
artifical, because there are really two C3H7 species, and also numerous
reactions do lead to those unsaturated C3 species that are ignored.