Hacker News new | past | comments | ask | show | jobs | submit login

Yes and yes. E-bike batteries consist of 40 to 100 cells. Typically one cell fails prematurely and compromises a pod of 4 to 10 other cells that are connected in parallel with it. Replacing the whole pod is safe as long as it's done properly (removed and re-welded without damaging the enclosure or shorting anything out) with like cells that have identical manufacturer, specs, actual capacity and charge curves to each other and to the rest of the battery. Ideally they should be matched with cells from a donor battery of the same "vintage".

The vast majority of the time, the remaining 90% of the cells in the pack have a lot of life left. If multiple pods have failed simultaneously (which is very rare), that's a sign that they're all starting to go and all cells should be replaced - which may still make sense, as the enclosure and BMS are often half the cost and perfectly functional.

The problem is you really do need to match the cells by actual capacity and charge curve, and that requires specialized equipment. If you don't, a good BMS will shut down the pack, and a bad BMS will allow it to catch on fire. Given the financial incentives (many hundreds of dollars to replace a pack), people will try to fix a pack in hacky ways. This is why it's important to provide the right tools and training for local shops to develop economical ways to service batteries and to educate people about the hazards involved.




This is all good stuff. I'd like to add one small bit: most BMS are not capable of balancing packs that have become significantly unbalanced fast enough, and if they're really out of whack you may have a problem. A good BMS will allow this situation to exist but will de-rate the pack (you'll be charging to a lower overall voltage), a bad BMS will still allow the pack to reach the full working voltage but this will result in some pods to be under their maximum voltage while others will be over.

And that's a serious risk: a battery that is just fine at 4.2V may well go ballistic at 4.35 so even a small offset can cause real trouble. That's why those sense and balancing wires are all important. Keep in mind that they are not fused whenever you work on a battery and carry full voltage.


Could you go down a rabbit hole a bit on this part of your response? "[Manufacturers] could design their packs to be more reliable and easier to repair."

What is it, if you were a state or local government, that you would force on the industry?

Also, what happens to the remaining "good" cells when you recycle rather than repair a battery? Are they kept in tact?


You could build a pack out of modules, the logical unit is the 'pod', the block of cells that are set to operate in parallel, typically there are three electrical connections (all the joined + contacts, all the joined - contacts and a sense wire connected to the +) the + and - are usually passed on to the rest of the string except for the topmost + and the bottommost -. Those are brought out using wire. You could allow a pack to be split on the pod boundaries. That would allow for one relatively easy unit of repair. You'd have to know which pod to replace but that is something that you might be able to figure out using a DVM.

Another thing they could do is to allow for a complete reset of the BMS instead of bricking the BMS as is now done. This would also take care of capacity drift, which causes a lot of trouble with older packs where the reserve becomes an ever larger part of the total pack capacity and which can render a fully functional pack effectively useless.

I would not bother with the right-to-repair on this one but demand that for an e-bike to be certified that there would be a cost+ arrangement to deliver a refurbished battery to the end user, and that this should happen at least twice during the life of the bike. This because the batteries are by far the weakest spot in the chain, they fail at a rate much higher than motors or controllers. And the life of a bike should be defined as no less than 20 years.

Finally, when you recycle you could break up a pod into its constituent cells fairly easily, test them, discard the ones that are not passing a testing regime and then to match the cells and recreate new pods. But unless this is done at a very large scale it is probably more cost effective to just discharge the batteries, grind them up and to recycle them completely. That will give you brand new cells at some loss of material.

I hope this answer is what you had in mind.


Yes, absolutely, super interesting. Would it ALSO be a good idea to standardize battery specs? If so, which?


That's always hard, you don't want to give any manufacturer an edge but at the same time you also don't want to force everybody into buying a new bike. It may be possible to create an adapter that would accept modules for the new system. I'm imagining something like a stack of supersized 'D' cells. Packs would then combine these in the quantity needed to reach the required voltage. Doing that reliably and with very low contact resistance will be a major challenge.

The garden and powertools world has a working model for this:

https://www.powerforall-alliance.com/en/

With a lot of the major brands already subscribing to the concept that's a really nice move forward, and I'm looking forward to Bosch extending the concept to e-bikes, there really is no reason why they could not. It will make batteries more likely to be stolen though, so that is another angle that will need work.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: