Tesla's previous CTO, JB Straubel, left to found Redwood Materials. He intends to commercialize recycling of batteries at scale, and the rumor is Tesla will be a customer of the output from Redwood for battery manufacturing.
I think the problem is mostly logistics. To get good use out of an old array, you have to QA all of the cells and discard the worst ones, because when you wire batteries in sequence (to get higher voltage), then the weakest link effect is very prominent. And if memory serves, to get the most out of a battery array you want to 'match' the cells, so you have to bin the used batteries the same way Intel bins CPUs.
If memory serves, you have to charge a battery to measure this, which takes time and therefore space.
So to get the cost to be cheaper than just building a new pack (possibly using binned new packs), this all has to be pretty automated. If they bin new cells (if they don't maybe they should!), then most of this hardware would already exist. Which would mean you only need to talk about how to make a robot to desolder battery packs. Given how crap my soldering skills are this feels like a potentially hard or at least expensive problem.
You don't have to do that to a Tesla battery. The individual cells when they go bad disconnect themselves using a fuse wire. All you have to do is test the pack to see how much power it can store.
I saw a film about battery recycling recently and the first thing they have to do with incoming cells is discharge them. They were using the energy to power the site. Clearly it wouldn’t be much more work to cycle the batteries.
