-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Battery Pack Failed in Less Than One Year (>13,000 miles)
- Thread starter toulouse
- Start date
Hmmm. Pretty categorical assertion. What are your credentials to say this? From my experience, it is not unusual to do QA tests along the line, pull failed parts, and then triage them separately... especially when the production item is worth >$10k.
A module isn't worth >$10k.
And sure they are probably pulled off the line and re-worked, but that doesn't cause them to be classified as reconditioned. They are still new and just put back into the line with the rest of the parts.
@acarney said “might”. Up til that, he was accurate WRT standard practices. After that, he was conjecture like everyone else At least he’s said “might” unlike those who profess to know for sure.A module isn't worth >$10k.
There are dozens of tests performed on the various components and module as it runs through the production system, and eventually when it’s turned into a pack. If there are issues found, the issue is dealt with, whether that involves removing it from the line for a more thorough check, or correcting on the spot. Doing so doesn’t reclassify the item as reconditioned. No one in manufacturing does that.
Ok, so 90% of what he said was correct or plausible. Your only issue was the assertion that they might reclassify? I’ll agree with that.
Items don’t get pulled off the line and reclassified as reconditioned, and they don’t get new serial numbers after repairs. That would be a data nightmare, a huge traceability issue, and down right stupid.
It’s always funny how many people assume that they know what happens in Tesla factories, even if they have some prior manufacturing experience....
It’s always funny how many people assume that they know what happens in Tesla factories, even if they have some prior manufacturing experience....
And it is interesting how a guy who works at Tesla can say “nobody in manufacturing does that”.
Then I suspect a high likelihood of a new pack going into the car. I just can’t imagine there are that many trashed model 3’s right now with battery packs is such a safe configuration that Tesla feels ok keeping them in service with supercharging.
This would be especially true given your low mileage. Car is written off because of strictly body damage but has 16k miles on the pack? Probably not going in yours as a “like for like or better” pack...
I have 20+ years of manufacturing experience at various companies in operations, maintenance, and engineering, 18 of which were from leadership/management roles. Also ran my own consulting company for 11 of those years where I advised dozens of companies on manufacturing principles. But..what do I know about manufacturing....tell me about your vast experience....
Reman pack is more likely. Reman doesn’t mean it came out of a crashed car.
I actually would have thought the opposite. If it fails a QA test somewhere and the serial number is flagged as bad, there would have to be really strong traceability that it was inspected, repaired, and QA tested again with passing results before ultimately going back off for assembly. If it got by without any step happening and placed in a car, you now run the potential of delivering a defective product to a customer that may fail prematurely and most likely cost Tesla more when factoring in logistics and service time and potential risk to the brand image.
You flag it as bad and remove that serial number from the useable ones and if it finds its way to assembly there should be major red flags not to use it. If it fully gets inspected and reconditioned then you issue it a new serial number that has a “clean pedigree” and there is no question about if all the proper steps to repair it took place.
Based on how Tesla seems mismanaged at times and lacking communication, I would think you would want CLEAR indications to keep the bad packs segregated from any chance of making their way to final assembly.
I’m also assuming their fail rate is fairly low, I wouldn’t think they would have like a couple out of every hundred requiring intervention by a technician to correct issues. That’s why I figured it would be faster to reject it from the line and not slow down the production. Possibly if this is routine where a few packs out of every 100 need some amount of adjustments different then perfectly good ones they would have integrated on the spot repair but I would think “bin it” and then let someone else deal with it would be the cheapest, fastest, and most efficient way when working in large volumes.
Where the heck are these packs coming from? None of these cars are reaching end of life yet. Unless packs are being pulled off the line for original manufacturing defects that can be later corrected (which you claim wouldn’t happen, they would be corrected on the spot and used as new) where does Tesla get packs from to remanufacture? I would hope the failure rate from cars was extraordinary low... otherwise Tesla has a serious problem on warranty costs.
From a manufacturing perspective, specifically on the quality side, you create one serial number per part, and never change it. This is paramount to traceability. If it’s able to be reworked, you have a specific set of instructions on how to repair it, and then how to test it for failure prior to reintroducing into the production stream. Key is to test for failure, not for good. If something has 3 points of failure, test those, and if they pass, the result is good. If you test for good, you stand the chance to miss a failure point.
That data (failure, rework, reintroduced to production) all goes with that serial number and stays with it to end of time. If there is ever an issue in the future, having these records to look back on to understand is crucial. Maybe one item failing and being reworked isn’t the issue, but what if two separate components failed initial checks, were reworked, and landed in same module/pack....that’s how you learn and it’s how you put checks in place.
In regards to not letting a failed part escape, that’s easy. Red flag it in whatever system is tracking the flow of parts. If something is flagged, it can’t proceed to the next step.
I never claimed bad stuff wouldn’t happen within manufacturing. I said where checks take place, if it can be corrected in the spot, it would. Otherwise it’s removed for later disposition. Sometimes these parts, as they await engineering rework approval, will ‘age out’. Meaning that while it was awaiting approval/repair, a new revision change happened. Older revisions are automatic red flags from moving to next step. Depending on the quantity, quality, repair status, etc, these can be sent to reman. Sometimes they will need paired up with like revisions to not cause issues, or engineering will explore what revisions can go together without creating issues. Once that is determined, that is fed back into the system to allow future occurrences if needed.
Also, another source of reman is physical damage. Battery pack is extracted from a rack manually for whatever reason, and the driver clips a corner? Cosmetic only, but could create fit issues later on. Pack is sent to reman where is gets a new enclosure. In theory, it’s a new pack, but the serial (original and only serial) is tagged for reman/service use only. Also, DOA packs. GFNV will send a good pack to Fremont. Component issue cause a vampire drain. When pack arrives on Fremont line, it tests with OOS SOC. Off to reman to find the issue, component replaced, tagged for reman service.
Not as much anymore, but previously any pack that service replaced, would go back to reman, good modules extracted, tested, validated, and shelved. These serve as donor modules to future reman packs that need a specific module. Each module is configured differently and is not interchangeable within a pack.
wdited grammatical arrors
Last edited:
Has there been any follow up on this? Have you had the pack replaced yet?
If you haven’t, is there any chance you could ask the service center what the cost would be if this was NOT a warranty claim? Could you ask them if it was possible to generate an estimate invoice that you could scan and post here? It would be very interesting to see the exact amount of billed labor hours for this job and what the say the pack costs (either a remanufactured or if they can estimate with a new pack.) I don’t know if we’ve seen actual direct from the service center cost break downs on parts and labor for this type of job with the Model 3 yet...
Please let us know how it goes and good luck!
SilverSp33d3r
No Longer Silver
Similar threads
