Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ColoPaul]

So I'm coming up on 30K miles on my 2012 ZF6.  At 5 points in the last 4 years I've run the battery down to 0% charge and then measured to see how much it took to charge to 100%, thus approximating capacity.   The EiG spec was 1000 cycles to 80% (at 1C charge/discharge),  I was on track with that until around ~400 cycles, but since then I've been seeing a much quicker loss.

I've been relatively kind to the batteries, avoiding deep discharges.  Perhaps there's an aging effect.  Perhaps because it's only a 6kWh pack it has a significant higher C discharge than the ZF9 would (this is a snowballing effect as the pack looses capacity).   I don't think I have a bum cell, as I looked at all the individual voltages for each of the series blocks-of-4.

I suppose around 60% capacity I'll have to take action due to poor range.  I might try contacting Zero, but I can't imagine they'd have any options for me.  Considering rebuilding the pack, seems like more work than I want.  Maybe I'll find a reasonably priced wrecked 2012 that I can salvage the pack/box from, maybe I'll find a deal on a newer Zero.

Any thoughts?

[NEW2elec]

Nice work and I'm impressed with 30k on a 2012 6kWh bike.
The results are a bit disappointing though.

[Doug S]

For the record, I put 28,000 on the original battery of my 2014 SR, with no degradation whatsoever...if anything, capacity was a bit higher at the end. I was graphing up SoC every day when I got home from my commute, but I quit after six months or so out of sheer boredom.

Clearly the batteries improved between 2012 and 2014.

[ColoPaul]

For the record, I put 28,000 on the original battery of my 2014 SR, with no degradation whatsoever...if anything, capacity was a bit higher at the end. I was graphing up SoC every day when I got home from my commute, but I quit after six months or so out of sheer boredom.

Clearly the batteries improved between 2012 and 2014.

Ummm, the SoC after a ride is a very poor indicator of capacity.  Also, the 14SR has an 11.4kWh pack, so your 28k miles would have been far fewer cycles than my 28k miles.  Besides cycling, other things (like ageing, temperature, charge/discharge rates) also affect capacity loss - my batteries are ~2x older than yours, and had to endure high C discharging than yours.  So it's not really a fair comparison.

Try this: Go run your bike to 0% SOC, and using a kill-o-watt meter measure how many kWh it takes to recharge.  In another 2 years when you get to 56k miles, do it again.   I'd be curious to see the results.

[Doug S]

I don't intend to do that. My bike isn't a science experiment, it's my ride. Discharging to 0% is hard on the battery, to at least some degree, and I'm not going to do it. I've done it twice now, intentionally both times, and it's not happening again.

I don't agree that SoC after a ride is a "very poor" indicator of capacity. It may not be as good as a full drain followed by a full charge, as you advise, but it's been exceedingly consistent day in, day out for multiple years now. That's nothing but a good indication. Your technique also doesn't account for charging losses, which can be pretty substantial, and I wouldn't want to promise they're consistent.

The other things you mention certainly do have an effect. Aging is specifically what we're looking for, my charge/discharge rates are very consistent when the great majority of my riding is daily commuting, and it's interesting that I do see a seasonal variation, with residual SoC being a little bit higher in the summer than in the winter. How much of that is caused by higher battery temperature and how much is caused by heavier air in the winter is undetermined at this point.

It's clear that li-ion battery chemistries differ widely in their aging characteristics. There was an article published recently, which I can't lay my hands on right now, which showed Tesla's battery pack is aging considerably better than expected, with half-million-mile life easily possible, and supercharging seemed to actually help the battery pack retain its capacity longer than slow charging. The game's changing fast, if you blink, you fall behind.

[ColoPaul]

I don't agree that SoC after a ride is a "very poor" indicator of capacity. It may not be as good as a full drain followed by a full charge, as you advise, but it's been exceedingly consistent day in, day out for multiple years now. That's nothing but a good indication. Your technique also doesn't account for charging losses, which can be pretty substantial, and I wouldn't want to promise they're consistent.

SoC after a ride is a "very poor" indicator of capacity because of how SoC is calculated. The most accurate way is to monitor how many amps leave the battery, and subtract that from the amp-hour rating of the battery.  Note this doesn't give you any actual information about capacity - it assumes the battery is not degraded (for example, 100Ah), measures how much you've used (for example 10Ah), and then displays the result (100-10/100 or 90% SoC).  If the battery was degraded by 5%, and you never used more that 95Ah, you'd never know it.   It would be expected that you see "consistent" results.

(And if I were Zero, I'd throw in a little fudge factor: If the battery was rated for 100Ah, I'd only use 95Ah in my SoC calculation - this would tend to get users to not deep discharge the battery so much, and it would also allow for a 5% loss to occur before the SoC equation would start becoming non-linear at the bottom.)

Interestingly, Zero does improve on the Ah method by also monitoring voltage.  If the bike is stopped and turned off (You can't do much with voltage while the bike is running because of variation with current draw), you can also then measure the voltage and make an adjustment to the amp-hour SoC approximation. This does give some information about capacity but it's pretty crude because voltage varies by a tiny amount over most of the SoC range (100% down to ~15%).  On my bike this adjustment didn't make any noticeable difference until it had lost >10% of capacity.

And as for my method, your comment about charging loss variability has merit - but based on empirical evidence it appears to be accurate to ~2%.  And it's pretty clear that if the pack took 6.2kWh to charge when it was new, and now is only taking 4.6kWh, that's not due to charging loss variability.  Especially when it's correlated with range loss.