Hi Jpal, Since you mentioned this item in upthread posting - I found the official DWR's report to FERC on the thermolito fire incident. I've seen pictures of the control room with the electronic panels completely toasted. The official incident report had narrowed down the suspected source to be in the control room. I've dealt with spontaneous fires erupting from dielectric breakdown in capacitors (electronics). In some type of capacitors, the stored energy density is enough to trigger thermal ignition temperatures high enough to ignite and sustain progressive fire in the FR4 circuit boards.
These fires occurred even though the designs are UL certified. Technically, UL certification is to "contain" a fire within an enclosure via a "flaming drip test". But, the Thermolito Power Control room was operating in an automatic "un-manned" condition when the fire broke out. Alarms went off at 6:50am. It burned for a good time before someone arrived on scene (40 min later) & observed smoke coming from the building. The CO2 fire suppression system had discharged, but it failed to put the fire out. A number of attempts by Fire Crews (CAL FIRE & other districts) to put the fire out during the day were unsuccessful so they just "sprinkler" dosed the area until overnight (i.e. until it burned itself out the next day).
Electronics, wires, plastics, etc can burn as a fuel source IF there is an oxygen supply. I've worked on designs that automatically "trip" and blow a fuse to limit fire damage triggering. It is amazing the energy that capacitors can store. (even the lithium-ion battery fire issues of laptops & hover boards - defect mitigation, via design in quality & statistical process control [SPC] is absolutely critical if you push the limits of dielectrics & anode/cathode designs). However, there is nothing like having a backup where a person is able to immediately shut things down when you first smell the unique odor of burning electronics (you will never forget that smell).
