Really nice informative video. A couple of additional comments from a former
battery management system engineer, to the real nerds out there: First,
charging to 100% affects more than just range estimation. It also effects
accuracy of cell-to-cell balancing, and hence actual range! Small leakage
currents in some cells inevitably cause charge differences to accrue between
different parts of the battery over long periods. The circuitry needs to
correct these imbalances by "bleeding" charge off of the higher charged cells
or cell groups. It can only do this if it knows which cells are more charged!
If the cells aren't properly balanced one cell/group will hit 0% before the
others, leaving energy on the table, and as a consequence the vehicle will
estimate (and deliver) lower total range. Even NMC chemistry balances better
when fully charged occasionally. Although more steep than LFP, the NMC
charge/voltage graph's slope just isn't steep enough for <1% accurate balancing
in the middle portion. You'll get up to a couple more percent range from NMC if
it can reach 100% every so often and re-estimate charge there, and re-balance
more accurately.Second, it's important to understand that all battery
management systems use "coulomb counting" (monitoring discharge current) almost
exclusively to estimate range, at least when driving. This is not just for LFP,
NMC absolutely does this as well. The cell voltage fluctuates wildly during
driving as the internal impedance of the cell causes varying voltage drops
under varying load and really can't be used for charge estimation very much at
all while driving. In fact, even once the car is parked, NMC systems can't just
immediately look at the cell voltage to recalibrate the current accumulator
that tracks charge. Not only do cell impedances have time-dependent components
that take minutes or even hours to stop exhibiting small voltage drops, but
cells exhibit all sorts of irritating effects that temporarily change the cell
voltage, such as hysteresis and other history dependent effects. These effects
tend to go away over time (and faster when it's not freezing cold). So for NMC
the car wants to be placed at 100% occasionally, and left there for at least 6
hours or so, and preferably not too much below room temperature. This will
allow more accurate recalibration of the accumulator, and more accurate
balancing. Now the actual algorithms are very complex, the car will get
progressively more confidence that it understands where the charge level stands
as the cells rest more and more, so even while driving the current accumulator
would be updated if it disagreed wildly with the cell voltage, and the
convergence just gets better and better if you give it more time with no
current draw at a steeper part of the cell curve. [oh and it really should be
plugged in for this so that the battery doesn't have to support ancillary
equipment, and really has no load on it for those 6+ hours]One last bonus
point, not so much about delivered range but about range estimates. Although
slightly sketchy feeling, it's actually also helpful to discharge an NMC car to
low charge levels, where the voltage curve bends down, for the same reason of
estimation accuracy. (It doesn't need to be 0%, but say below 15%-20%).
Probably true of LFP too but no first hand experience with that. The same as
near 100%, the charge/voltage curve is steeper here and the car gets a better
idea about what's going on. Again, the car the needs to sit there for at least
6 hours or so and not be too cold to "rest" the cells and let their voltages
precisely settle. This is ideally done in close proximity to the 100% "rest"
described above(like the previous day or the next day). Once these two "rests"
have happened the battery management system now has a very, very good estimate
of the total capacity from 0% to 100%. It doesn't have to extrapolate too far,
since it's just recently been near 0% and 100%, and thanks to the steep curve
at the ends it knew exactly where it was! This allows the system to update the
estimated amp-hour capacity of each cell group with confidence, and so better
know how much juice it actually can hold. Again this won't give more actual
delivered range, the car stops driving when the battery voltage of the lowest
performing cell/group hits the low voltage cutoff no matter what percentage is
displayed. But if you don't want to ever drive below 10%, and the estimate is
off and pessimistic, you might be leaving range on the table by driving to a
faulty estimate. Or even worse, if the car never ever sees low charge levels
and has a really badly erroneous and optimistic estimate, you might believe you
have another 10% left to make that last 30 miles of the trip, and be stranded
10 miles out. It's really hard to get a car into this state, but if you only
ever let the battery see the range 60%-80% and never outside of that, it's
possible. That's only a 3.3% range estimation error in a 300 mile battery pack
but 10 miles is a big deal if the car stops driving.Show less
On Saturday, August 24, 2024 at 09:07:19 AM EDT, EV List Lackey via EV
<ev@lists.evdl.org> wrote:
On 24 Aug 2024 at 4:28, Lawrence Rhodes via EV wrote:
> https://youtu.be/w1zKfIQUQ-s
Thanks - that's an interesting and well produced piece.
I've always read that 20-80% is the optimum for energy throughput with most
battery types.
I wonder why they didn't test that regime, only 0-[25|60|80|100]% and 75-
100%.
Have the impression that NMC cycle life is better when it's seldom charged
above 70%. I don't see that as tremendously different from his LFP
recommendation, but maybe I misunderstood the piece.
The presenter says that Tesla recommends keeping your EV plugged in when
it's not in use. Really? That doesn't seem like a good idea to me, but what
do I know?
With our Zoe, we don't charge unless SOC is < 25%, and then only to 70-80%.
We will charge to 100% if we plan a trip for the next day that will use at
least 20% of capacity.
David Roden, EVDL moderator & general lackey
To reach me, don't reply to this message; I won't get it. Use my
offlist address here : http://evdl.org/help/index.html#supt
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Fertility is hereditary. If your parents didn't have any
children, neither will you.
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
_______________________________________________
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.evdl.org/private.cgi/ev-evdl.org/attachments/20240824/b4b7f67e/attachment.htm>
_______________________________________________
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/
