How Not to Put Out a Metal Fire - with Steve Mould

I remember watching WGN a couple of decades ago about a fire at a metals recycling company in Chicago (maybe this one?) The thing that fascinated me at the time was watching the firefighters spray water on the fire, have it explode, and repeat the process a couple more times. The fire chief at the time - in my deep memory, at least - said something to the effect that they knew the explosion was likely but that they were trying to dump enough water on all at once to cool the fire enough to stop the burn. Not enough water = explosion, though.

In the above video - the Royal Institute - Steve Mould shows why magnesium fires are such a bear to put out. They don't - as the article linked above says - 'produce their own oxygen', but they do tear oxygen off of carbon dioxide and water. Once that happens, then, the fire allows the hydrogen and oxygen (from water - or oxygen and carbon (from carbon dioxide) to react with the magnesium or with each other producing an even hotter flame.

Hmmm...seems magnesium's active or something.

(And I don't think - at 2:55 - you 'can actually see [the beaker filling with carbon dioxide]', Steve. My understanding is that carbon dioxide is invisible. What I think you're seeing is fog, condensing water vapor due to the coldness of the dry ice.)