Batteries were tested singly, and in groups of 32, 64 and 128. Tests also involved groups of batteries packed in rows inside cardboard boxes.
For test purposes, the battery fires were started by igniting a "fire pan" filled with alcohol. The findings were fearful. To summarize:
* A relatively small fire source was sufficient to start a lithium battery fire.
* The heat from a single battery afire was sufficient to ignite adjacent batteries.
* The outer plastic coating on the batteries easily melted, fusing the batteries together, adding to the intensity of the fire.
* The chain reaction ignition continued until all batteries were consumed.
* The molten lithium burned explosively, spraying white-hot lithium to a radius of several feet as the batteries bounced around.
* The duration of the peak temperature increased with the number of batteries, reaching as high as 1,400[degrees] F (as a matter of interest, the melting temperature of aluminum is around 1,200[degrees] F).
* The cardboard packing proved highly flammable. The packing delayed battery ignition by about 30-60 seconds, but once ignited, the fire among the close-packed batteries was worse.
* While thick-wall cargo liners were able to contain the fire (barely), thin-walled fire liners proved ineffective. The battery fire ignited the resin in the liner, and the liner was completely penetrated by molten lithium.
* Halon fire-suppressing agent, injected in sufficient concentration to "knock down" a fire, proved totally ineffective, even when injected after just the first battery had caught fire. Nor did it have any effect on the peak temperature. The fire continued as if Halon were not present.
* Lithium batteries catch fire with explosive force. When they burst, they create a pressure pulse. The eight-battery test produced a pressure pulse of 1.8 psi, and the 16-battery test generated a 2.6 psi pulse.
