Batteries
Guys
I have done much research into the subject of NiCd batteries and their application to biz jets, working on one particular large bizjet. One of the key factors in determining battery life and it's time between removal or failure is how it is treated in service, this refers to both maintenance and daily ops by the pilots.
The current draw on initial APU cranking is up to 1800A (on the APU fitted to the aircraft I've been working on). As the APU accelerates, the starter draws less current, but after 15 secs this is still in the hundreds of amps. After a successful start, the APU battery has heated up by around 5 deg C. For certification requirements we look at 2 failed APU starts then one successful start. This is a tough test, especially when we must be able to start the APU at -40 deg C.
Much testing is done to establish the minimum rpm for starting the APU. It indeed will start at somewhat lower rpm than given. I would suggest that your battery reliability would be improved by starting at the lowest possible (per the AFM) APU rpm.
As for thermal runaway, for a NiCd battery this is actually quite difficult to achieve. For the cert test for our battery it required a batt temp of 71 deg C with a very high state of charge being constantly charged.
The point about the rapid fall in voltage is indeed a trait of the NiCd battery. Voltage will fall from around 24 deg C to 16-17 v during an APU start. If one of the battery cells is weaker than the others it may reverse polarity. This is usually reversible. The weakness of this cell will show up during battery maintenance, and may show up as a depressed APU battery voltage during cranking of the APU.
Regards
high-flyer