Guess what folks this is going to happen in any high energy system. You can't take new tech and put it in EV service and not have some problems. Facts are I'm impressed they haven't had a lot more.
While some have problems like fire, most of that type were bypassed for more safe, stable ones. Though I'd certainly would like a few of those MiEV or Tesla battery modules.
First rule in power design is you can get as much power as you can keep it cool. Likely this or BMS problems is causing though QC can too as shown by the Sony battery recall.
Interesting my new battery drill lithium's seem to have a cut off once hitting minimal voltage built into the battery pack maybe. So it looks like they are getting smarter.
Lithiums should be run between 95 and 10% as going passed either shortens life. Overcharging just 10% once can kill them so to keep them alive and well they need a tight BMS and temp control, neither a big deal but foolish to cut corners on.
If indeed GS Yuasa has managed to destroy not only LiCo lithium ion batteris in the 787 but also LiMnO2 Li ion batteries, as this article states, then there is something fundamentally and hopelessly flawed in their manufatcuring and/or quality assurance processes. I would be very suprised if their battery cells were not manufactiued in China at a cheap place lacking adequate cleanliness and quality standards, instead of in Japan. LiMnO2 LiIon batteries with their spinell structure, are inherently free of thermal run-away, at least theoretically. Based on everything I have read lately I would not buy anything with a GS Yuasa made LiIon battery in it, no one should. Yuasa's manegement is probably new and inexperienced (not to mention driven by profit and nothing else) and has never heard of Demming and his methods of improving quality which made Japanese qualty the envy of the world about 2-3 decades ago.
I wonder, did Boeing make these batteries, too? Har, har!
It looks like the battery problem just won't go away. And from the responses thus far, it isn't going away.
One problem with BJT transistors (I know, BJT already includes the word Transistor) is thermal runaway. FETs just shut themselves down when they get too hot, more or less. I can see batteries have the same problem as the chemical reaction increases with heat, I assume. We just have to figure out an FET version.
Heat is our enemy in so many cases. If we are going to use high energy batteries to replace so many internal combustion engines, then I guess we better figure out what the problems are. I know in the beginning of the steam engine and internal combustion engine there were many problems to overcome. Exploding boilers, head gaskets blowing through, gas tanks exploding (thanks to the media), and so on had to be overcome. So shall it be with the big league batteries.
It looks like the liquid electrolyte will be always a barrier to manufacture safe batteries. Solid electrolytes cells are underdeveloped yet but will come in the close future. Thin Film Battery is the answer to all safety issues, I believe, but the manufacturers are able to deliver only very low power ones. There is no market for TFBs yet but maybe someday low power devices will be powered by TFBs. In my opinion the EVs are to heavy, it makes no sense to continue to build cars like in the XX century. We need to move away from the concept 20 to 1 - 20 times more weight to cary a person. Than we will need less energy to move. I'm waiting for 2014, BMW will introduce the carbon fiber car - maybe this is the future of mobility.
I'm going to take a pass on your thermal runaway questions, bobjengr, other than to say the obvious: Overheating can lead to a cascading event. As to when overheating turns into thermal runaway...we would need a material scientist or better, a lithium-ion battery expert, to go into greater detail. Perhaps a reader who is familiar with thermal runaway can chime in. As for battery monitoring: Yes, all of today's lithium-ion EV batteries use battery management systems that check for over-temperature, over-voltage and over-current conditions. Usually the battery managment system consists of a little IC that contains a microcontroller, memory and an A/D converter. Temperature measurements typically require a multitude of sensors around the battery pack. Toyota uses about 40 sensors around the pack in its Prius PHV plug-in hybrid. And, yes, the cooling systems can respond to those conditions. Toyota uses an active cooling systems with three or more fans. Chevy uses 144 metal plates with channels machined into them to allow for liquid coolant to run between the battery's 288 cells and draw the heat away, in a manner similar to that of an engine block and radiator. As you suggest, all of the automakers have put tremendous development effort into the design of these systems. But as recent events suggest, they're still learning.
It has been my experience that battery engineers over-charge the batteries. It doesn't surprise me that most of these battery failures occur when the vehicles are sitting and being charged. As the battery approaches full state of charge, the energy going into the battery to charge it winds up as heat. You need to measure heat dissipation at the battery assembly level. Battery heat dissipation varies as a function of state-of-charge, voltage, battery temperature, and charge current. Batteries are typically endothermic when charged at low states of charge, but they rapidly become endothermic as you approach full state of charge. We typically size the battery cooling for 3.5 watts/amp of discharge.
@ChasChas: the reason that the battery electrodes are so very close is that they need to be close to avoid losses in the electrolyt, and also to keep the batteries compact.
On the other hand, it might be possible to build the batteries in a manner similar to those "self-healing" capacitors that appeared during the early 1960s era. The concept was that a short circuit would result in burning away a small segment of the electrodes in the area of the fault, removing the shorted circuit connection. Of course, there may be some very good reasons why that would not work in a battery like this but it certainly seems that it could be worth cnsidering. An added advantage is that it is certainly not a new concept.
Great post Charles--OK, I'm probably going to show my ignorance here but I'm not close to being a battery expert.
QUESTION: When overheating begins--is this a cascading event? Does it always have to end with a catastrophic failure? I know we have cause/effect situations but is there any way to monitor temperature rise and dissipate that rise by appropriate cooling? I'm sure all companies using this technology have evaluated any contribution from surrounding sources. Are there any trends regarding installation and mounting of the battery? Again--great post and thank you for keeping us up to date.
That's a very timely question, Greg. And the answer is, yes, there is a safer chemistry on the horizon. We haven't written about it yet because it's still in the early stages (as far as I know), but there are soild state lithium batteries that eliminate the liquid electrolyte, and are therefore regarded as being safer. Toyota is working on a solid state lithium "super-ionic" battery. As with all battery research, however, the schedule is unclear. I've heard projections of 10 years for solid state lithium. And, typically, when battery researchers say ten years, it means more than ten years. Perhaps one of our readers is familiar with the solid state lithium technology and can provide a little more info here.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.