The propane tank Fire Test and Its Changing Conditions
The propane tank investigation began with a claim about a barbecue grill exposed to fire: its tank could rupture and propel itself into the air. The first test produced a different observation. In a burning shed containing a grill with a full tank, the pressure relief valve opened and vented gas. Later tests produced fragmentation and flight after the conditions were changed. The central distinction is therefore between the first trial and the modified trials that followed. Their outcomes should be read alongside those changes, rather than merged into a single account of an unchanged tank.
For the opening trial, the Build Team constructed a shed to California building code. They put a grill with a full tank inside and supplied wood and charcoal to sustain a long, hot fire. Once the shed was ignited, the tank's relief valve eventually opened. That is the recorded result for the initial setup. The account does not report a tank launching from that shed test. It instead describes gas venting through the valve, a detail that becomes especially significant because the team altered the valve condition before continuing with the later experiments.
Grant disabled the valve on every tank used in the subsequent trials. The next test placed a tank directly on a propane-fired burner. It exploded and sent shrapnel as far as 300 feet in all directions. This was fragmentation, rather than the intact flight later described. For the propane tank question, the distinction between scattered pieces and an airborne container is useful. Both belong to the reported sequence, but they describe different outcomes. The disabled valve also remains part of the context for this trial, so the event should not be presented as a repeat of the original shed conditions.
Another trial used a tank with a partially bored hole in its base to simulate a tank weakened by partial rusting. The tank peeled apart from the bottom upward. That result added another kind of failure to the sequence. It did not supply the intact launch the team was seeking. The simulated weakness was a deliberate change in the test, and the resulting separation of the tank was the observation attached to it. There is no reason to invent a particular flight height for this attempt or to claim that it matched the later airborne result.
The team then reasoned that some vapor space might be needed for flight and tried a partially filled tank. In that test, the propane tank rose 60 feet into the air while staying intact. The team considered the launch claim plausible at that point. The explanation about vapor space is presented as their reasoning for another test; the flight and intact condition are presented as observations. That separation keeps the account precise. It also avoids turning a successful modified trial into a claim that every full tank in a shed fire would follow the same path.
The testing continued in pursuit of a higher flight. Tory added a launch guide, and the later setup used a pistol shot to create an opening at the tank's base. The resulting blast sent it into a high, spinning trajectory. The account says this convinced the team that a tank could become a rocket under the right conditions. It does not provide a measured height for that final flight. The defensible description is consequently qualitative: a high, spinning path in a further modified test. Adding a numerical height would go beyond the reported result.
For the propane tank story, the phrase "under the right conditions" carries much of the conclusion's meaning. The first trial had an operating relief valve and produced venting. Subsequent trials had disabled valves and included other changes in tank condition, fill level, or the launching arrangement. The outcomes ranged from shrapnel to peeling apart to intact flight. The account supports that range and those changes. It does not support flattening them into a universal prediction for every container exposed to fire. A careful reading can retain the striking flight without losing the circumstances attached to it.
The most useful ending for the propane tank question returns to the first observation and then keeps the later results in order. The shed trial vented gas through its relief valve. A later trial scattered fragments, another peeled the tank apart, and a partially filled tank flew intact. A further modified arrangement produced the high, spinning flight. Together, those observations explain the team's qualified conclusion, while the first trial answers the opening question directly. The interest is in what happened as the conditions changed, with each result remaining attached to the particular version of the test that produced it.