In pulverized coal flames, nonuniform mixing and particle dispersion e
nsure that the products of primary devolatilization are transformed by
secondary pyrolysis before they burn. This laboratory study shows how
the rates of volatiles combustion change during the course of seconda
ry pyrolysis for a spectrum of coal types. Volatiles mixtures are gene
rated in a novel coal flow reactor that independently regulates the ex
tent of secondary pyrolysis. After tars, soot, and char particles are
filtered out, oxygen and nitrogen are added to prepare combustible mix
tures that have specified fuel equivalence and dilution ratios. Flames
are propagated through the mixtures in a combustion bomb to acquire t
he pressure histories that define laminar burning velocities. Burning
velocities of noncondensible volatiles from 4 coals representing ranks
from subbituminous to low volatile bituminous are reported for fuel e
quivalence ratios from 0.4 to 1.5, at two dilution ratios and two unbu
rned gas temperatures. For all coal types, burning velocities triple a
s the extent of secondary pyrolysis increases from 50 to 100%, in roug
h proportion to variations in the H-2 levels. This tendency is consist
ent with conversion of the oxygen and hydrogen in tar into CO and H-2,
and of light hydrocarbons into acetylene. For complete secondary pyro
lysis, the burning velocities of volatiles from higher rank coals appr
oach the comparable values for hydrogen combustion. But this tendency
is not seen with lower rank coals because hydrogen makes smaller contr
ibutions to their heating values. Increasing contributions from hydrog
en explain the impact of increasing extents of secondary pyrolysis, an
d different proportions of CO and H-2 determine the variation of burni
ng velocities among different coal ranks. This database is used to dev
elop a correlation based on a scaling from the thermal theory of lamin
ar flame propagation and only two pseudo-components: the actual H-2 an
d CO levels plus the amounts associated with instantaneous partial oxi
dation of all hydrocarbons. Provided that flame temperatures are based
on actual mixture compositions, this strategy closely correlates a da
tabase that covers virtually the entire spectrum of coal types, the la
st half of secondary pyrolysis, the full range of equivalence ratio en
countered in practice, plus two dilution ratios and two unburned gas t
emperatures. The same scaling is used to assign nominal burning rates
from the measured burning velocities, which are found to be up to an o
rder of magnitude faster than burning rates of soot, as expected.