PYROPHORICITY— 
A Technical 
An AEC official tells what we know— 
and what we don't know— 
about a major problem in the 
FIG. 1. 
1955, after heavy rain. 
occurred in several non-adjoining bins, as well as spreading 
from bin to bin; 159,000 |b of Zr scrap were involved 
FIG. 2. General view of zirconium-scrap storage facility. It 
had been thought separafing walls would be sufficient to prevent 
spread of fire in case one broke out; this turned out not to be case 
PyropHoricity* is, unfortunately, a 
property of several of the metals most 
widely used in atomic energy programs. 
Plutonium, uranium, thorium, zirco- 
nium, and hafnium are afflicted, as are 
magnesium, calcium, sodium, and 
potassium. 
That this is no mere academic con- 
sideration is attested by the fact that 
half of AEC’s 1955 monetary loss from 
property-damage incidents peculiar to 
nuclear activities stemmed from decon- 
tamination activities following spon- 
taneous fires in radioactive metals. 
Yet our understanding of pyrophor- 
icity is very meager, quite inadequate 
to cope with the broad range of be- 
havior exhibited by pyrophoric metals. 
Some of the phenomena observed and 
, cal) c irmtonieen. 
ro-phor‘ic (-fOr’/Tk), ady Ke Purophoros fire-bearing, 
YE pyr fire+ phereinto bear.) [ight-producing; igniung 
spontaneously; of or pertaining to pyruphorus. 
pyrophoric alloy. Metal. An alloy, as ferrocerium, having 
the property of emitting sparks when scratched or struck 
with steel. Si 
lighters. See PYROPHORIC. i 
py-roph’o-rous (pi-rdf’S-rtis), adj. Pyrophoric. 
-roph’o-rus (pi-rd{’S-riis), n.; pl. -RI (-ri) NL.J 1. 
free tan, tu igpi 
uch alloys are used for automatic gas and cigar 
118 
Zirconium scrap aflre in outdoor storage bins in May 
Following initial blaze, flare-ups 
some possible mechanisms are discussed 
here; research now underway (see boxes 
on p. 29 and 31) should fill in the pic- 
ture considerably. 
Nature of Pyrophoricity 
Although pyrophoric metal fires are 
infrequent, the cumulative experience 
in handling large quantities of these 
metals includes thousands of incidents. 
The vast bulk of these were minor 
spontaneous fires involving no property 
damage. However, a number are un- 
usual incidents, the causes of which are 
difficult to explain rationally; a few 
involved fatal injuries and major prop- 
erty damage (1, 2). 
The range of contrasts covered by 
such vagaries is indeed impressive. 
For example, massive (i.e. solid chunk 
rather than powdered) uranium will not 
normally ignite even when heated to 
its melting point—yet a case is known 
where a yeast-cake-size piece of massive 
U ignited spontaneously while resting 
on dry ice, and in another case a U 
nuclear materials field, 
now the object of three studies 
By RICHARD B. SMITH 
Safety and Fire Protection Branch 
U. S. Atomic Energy Commission 
Washington, D. C. 
specimen at room temperature spon- 
taneously exploded. Spontaneous U 
fires have occurred under vacuum, un- 
der water, and under argon atmos- 
pheres. Spontaneous ignition and con- 
tinued combustion of massive Pu, U, 
and Th specimens have occurred in air 
at room temperature. 
Liquid-Combustion Analogy 
Unlike flammable liquids, metals do 
not ignite at consistent temperatures 
when heated in air. Metallic U, for 
example, can ignite spontaneously in 
air at room temperature when in the 
form of fine powders, but massive metal 
will not normally ignite even when 
heated to its melting point. 
At the risk of oversimplification, a 
further comparison with flammable liq- 
uid combustion is helpful in obtaining 
a better understanding of observed 
vagaries in metal fire and explosion 
properties. 
Flammable vapor-air mixtures burn 
only when the concentration of vapor is 
