Fire Safety Science Digital Archive

Three series of test fires conducted by Factory Mutual Research Corporation in the past, involving deep storages, have been analyzed for flame height, using both video records and temperature measurements on the plume center. The test series included 4.5 m high palletized storage of different commodities (13 tests), 3 to 6 m high rack storage of two different commodities (11 tests), and wood pallets stacked 0.3 to 3.4 m high (13 tests), all with ignition at the bottom. When flames extended above the storage, their heights were well represented by a correlation for pool fires, referencing the mean flame height to the base of the combustion region (bottom of storage in the present case) and representing the fire diameter as if the fire spreads as a cylindrical volume in the storage (volume spread). An exception was rack storage of a relatively thin, combustible layer (corrugated) over an inert, steel box, comprising a pallet load, for which the fire diameter had to be based on surface spread rather than volume spread. The virtual origin (from which the plume above the flames appears to originate), determined for some of the test fires, appeared predictable from flame height as in pool fires, but requires knowledge of the convective fraction of the total heat release rate, which may not be a constant fraction of the total heat release rate during fire development in the storage.