Its worse with RC use of LiPos. I only use LiPo chemistry in my planes/helis/cars. These range from being tiny 100mA single cells, to 5A 2s2p batteries. Most of my collection includes 2A 3s1p cells, which are regularly drained at an average rate of 15-20C. (flat in 8 mins). These cells can Realllly deliver current. I shorted one once, for about 3 seconds, and it let out a significant amount of heavy smoke covering the table. Luckily it never ignited. But this is a gentle reminder of the high risk of danger with these batteries, especially under shorted conditions
Many of the flat cells show visible gasseous expansion closer to the end of their life. In the case of these rolled up cells we use in torches, I havent been able to spot any physical changes in the cells prior to failure.
Videos of fires exist. Most are done under high voltage/overcharging states, and there are various stages, including a gasseous expansion, followed by increasing pressure and heat, which eventually ignites a very rapid burning fire which is extremely rapid for several seconds, followed by a relatively slower burning. In the case of the round (rolled) cells we use in torches, in combination with the hermetically sealed tubes we put them into (pipe bomb), It seems that these cells are less likely to be recognised by everyone as obviously dangerous before its too late.
As said before, although there is always the innate risk of failure, most can be attributed to some degree of operator error somewhere along the line, mostly related to a lack of understanding of the risks, and from not recognising/dealing with risk factors. Examples that come to mind include;
Worn battery coverings - especially in protected cells where there is a electrical connector running down the cell, but where it turns the corner, it is unable to properly conform to the shape of the cell and leaves protruding corners that rapidly wear through the usually thin heatshrink coating. These need to be checked regularly, and reinforced.
Physically damaged cells - increased risk of mechanical failure, shorting
Overdischarged cells - are usually the history that is commonly associated with lipo fires and other dangerous events. Most of these come from not using protected cells, failed protection, inability of the user to recognised overdischarged cells.
Overcharging - chargers that are not very accurate in the detection of charged state, chargers that are not true CC/CV chargers, and chargers that are not able to detect actual cell voltage via an independent sense connection, and simply increase the charging voltage blindly to compensate for losses in the cables. This works fine under high current, but once the cell requires less current to charge, the floating voltage of the cells increase above recommended levels (think of the cheap charges that are commonly used). This is why I strongly recommend a hobby charger with cell balancing capability, because it has a voltage sense external to the current carrying charging leads. It actually results in faster charges, while actually being better for the cells.
Speaking of storage, What you can buy are fiberglas woven bags which help contain the short term flame. These are no guarantee, however it is an extra level of protection.
Sorry for the rant in the wrong thread, I just happened to feel like typing.