donald wrote: hmm.... there are 'tests' and there is 'real life'.
True, but the tests were also full cycles, which the car never allows.
I do not have enough information on tesla pack life to comprehend it. I have heard of pack failures, but don't know what those involved.
All the ones I'm aware of entailed a failure of one or more high voltage contactors inside the pack. The quick fix is to drop the pack, put in a new pack, send the customer on their way, and then open up the pack and replace the failed part.
I don't think NCA is well suited to large capacity cells, and many parallel cells create reliability issues - i.e. they might last longer, but when one cell goes in a module of parallel cells it brings the whole module, and/or pack, down.
Larger format cells are made up of paralleled layers internally. If one of those layers goes bad it brings down the whole cell/module, and pack. If a large format cell goes bad you lose a larger portion of the pack.
So we might be seeing lower degradation out of tesla packs but possibly lower reliability too. If that can be picked up early and fixed cheaply then that's not necessarily a problem, but if there's a chance I might break down and never know that's about to happen until it does, I'd worry about that sort of tech.
If a pack module is having problems the BMS will catch it, and that data will be sent back to Tesla, who can take action long before it strands you. Again, the benefit of many smaller cells is that any individual cell problem will have a smaller impact on the pack as a whole. Of course there is the tradeoff of greater complexity and a higher number of potential failure points, but I've never heard of a Roadster or Model S being stranded because of individual cell failures.