(88c) A Case-Study of a Fire Incident of Trichlorosilane Process and Response Measures

There was a fire occurring of a overheating Trichlorosilane (TCS) reactor in a solar panel manufacture plant in 2011. Although no injury happened, the leak-out TCS caught fire and produced acid plume creating great concerns for responding governmental agencies, nearby general public and news media.  Since TCS is an important raw material for petrochemical, semi-conductor, and solar panel manufactures, its hazardous properties of low flash point, easy combustible, water-reactive, and acid-producing have created many difficulties in response as its leaking out. 

All incidents of chlorosilanes (CS) materials would produce large acid plume as large amount of leaking lead to burning.  The fires could not be controlled by water due to their water-reactive properties.  When other distinguishers such as foams were applied, the uses of water jetty were essential to cover the burning surface and to reduce the acid plume.  The best practice was to continuously put foams on top until it burn out.  However, this action would prolong the response and could not control the damages well.  Therefore, the prevention of acid plume and the correct use of foams were essential for fire response to CS and TCS.        

Upon our field tests, we found that no fire distinguishing was observed for powder, carbon dioxide, water and halon on small-scale fire (3 kg) of TCS. All continuously produced acid plume and some even reacted to form harmful byproducts. The fires would be distinguished as the application of foams and liquid nitrogen.  The effective practices were that 6% mixed middle-expansion foams with 15 cm of covering height and liquid nitrogen with volume (ml) of 177+0.287*(TCS volume).  However, two methods adopted totally different principles for fire extinguishing.

Our study indicated that when using liquid nitrogen to effectively distinguish the TCS fire was due to it adsorbed reaction heat of burning and stop the reaction.  We further prove that no property change of TCS after our application.  This application was safe and was different from applying foams that water actually reacted with TCS and producing dangerous hydrogen gas. This finding might be used in future development for cooling agency for CS fire application.