Michael:  totally agree. At the cell level you are right and in the end the user has to take responsibility. The challenge facing people that a single Liion cell is not going to cut it so they have to be assembled into packs and that is where the differences start coming out. What follows is that cell level properties do not automatically follow/scale/apply to the battery pack.

At the cell level Liion shows virtually zero  Peukerts effect. At the pack level with a BMS, it will. The cause is very different though.

At the cell level, 0%age discharge is not damaging. At the pack level, cells can go out of balance and potentially short out. BMS will prevent the 0%age from being reached as part of the system level management strategy.

At the cell level, there is nothing stopping Liion cells from being connected in parallel or series. At the pack level, parallel operation is not a problem but in series, the BMS can fail.

At the cell level, you can extract all the capacity. At the pack level, this is difficult with the BMS applied. So if 100aH is the minimum needed in reality, a bigger capacity battery is needed.

I think the problem is that we have become so used to treating cells like lego that it is easy to forget about the care and management side of Liion. Equally it can be very confusing when cell level and system level data/behaviour is compared because they can be almost contradictory which makes it very confusing to people trying to understand what is going on.

On 17/03/2019 04:10, Michael Ross via EV wrote:
Steve, the ratings are per cell. WHat else is there to do?  No management
is needed as with a pack. Measuring the capacity of a cell is pretty
straightforward. It is up to the user to determine the needed capacity of a
pack and how to take care of it.

On Sat, Mar 16, 2019 at 6:31 PM Steve Heath via EV <ev@lists.evdl.org>

There are lies, damn lies and battery AH data.

Yes you are right in that even with a "100aH Liion battery" I would
derate it by 80% or so because of the BMS and so on.
The problem with many Ah figures is that you do not know how they were
measured. Is this with or without the BMS?

Most companies take the aH from each cell in the series string and add
them up. Some pluck a figure out of the air!

80% of the manufacturer figure is a good place to start but that might
be a tad high in my experience with those size of currents.


On 16/03/2019 21:51, Dan Baker via EV wrote:
Wow, lots of learning here.  So with Lithium, a safe BMS cut-off
kicks around when there is less than 20% remaining? So the a/h ratings
typically (and when truthfully) displayed are actually 20% less?  100 a/h
battery is typically only 80 a/h?  This is fine as I know you can't get
100% of out lead either.  I looked up my SBS-170f lead's datasheet, to
the cells to 1.80VPc (half discharged I believe) will happen with 116
for 1 hour.  This is recommended bottom voltage and I typically see that
with my boat as the draw is pretty constant when cruising.  To take the
leads all the way to near complete discharge (damaging but not
exploding) -
1.60 Vpc the amps for one hour is 125 amps so not much more.  So with a
ah pack I can expect about 1.6 hours of run time at 100 amps?   In
I think I can probably get closer to 2 hours if I keep the speed the same
and reduce the amps as the weight loss will dramatically reduce water

On Sat, Mar 16, 2019 at 6:08 PM Bill Dube via EV <ev@lists.evdl.org>
No Paul, Lee is indeed referring to the rate of discharge chart,
however, he has chosen the cut-off to be _*3 volts*_, rather than the
customary cut-off of_*2.5 volts*_. (No one uses a cut-off of 3 volts,
that I am aware of. All the charts note that 2.5v cut-off is the
standard for comparison. If we picked 3.5 volts as the cut-off, we would
get a huge spread in the apparent capacity, but that would be silly.)

You are correct that the 12 minute discharge (0.2C rate), the 0.5C rate,
and the 1C rate all show the same capacity, 3.25 mA-hr. While the 2 hour
discharge (2C rate) shows a slightly elevated capacity of  3.350 mA.

       I suspect that the faster rates had some unavoidable internal
heating, (even though the case temperature was held at a constant 25
degrees Celsius,) which tends to decrease the internal resistance, and
tends to raise the terminal voltage under load, especially when the
impedance rises near the end. Thus, the apparent capacity shift is quite
likely due to increased internal temperature rather than ion diffusion.

       Lead acid curves would have shown a much greater sensitivity to
discharge rate. Much greater. As I said earlier, the ions can diffuse
perhaps 100 times more quickly in Li-Ion cells than in lead-acid cells,
which makes the Puekert exponent very close to unity in Li_ion. Puekert
is not really useful in Li_ion because the diffusion is so fast in
Bill D.

On 3/17/2019 12:40 AM, paul dove via EV wrote:
That’s not what the spec sheet says. You are reading the graph for
temperature variations. There is almost no difference due to discharge
rates. 2C is 3250 and 0.5C is 3350 according to your spec sheet.
And lead acid batteries have a Puekert coefficient as low as 1.08.

Sent from my iPhone

On Mar 15, 2019, at 9:14 AM, Steve Heath via EV <ev@lists.evdl.org>
Peukert's law is not an actual law but an empirical formula that is
based on actual physical measurements. It gives an approximate estimate
how much capacity can be obtained. The way that it is used is that the
capacity is measured at different discharge rates to give a co-efficient
that can then be applied to other batteries.  This is where the
lies. The coefficient is taken by measurement and providing another
is the same then the coefficient is applicable. If not and it isn't.
The key point is that the discharge curves for li ion batteries do
significantly depending on the load in real life according to the
manufacturer data.  At the 0% soc end point, the capacities are the same
(give or take). This is why the Peukerts coefficient is close to 1
than 1.2 or higher for a lead acid battery. Hence the comment that it is
not applicable. It is there but very small to be accurate.  However at a
typical self preservation point e.g   cutoff voltage used by BMS, the
capacities are different. As a result, there is a "Peukerts" effect
the amount of capacity that can be obtained is different depending on
discharge current. It is not the same Peukerts effect but the end
result is
the same. Discharge more, less capacity...
The data sheet for a Panasonic 18650 shows this effect very well (
https://www.batteryspace.com/prod-specs/NCR18650B.pdf ) where a cut off
voltage of 3v gives a capacity of 2400mAh at 2c and 3300 mAh  at 0.2C
.  At
the 0% soc point they all come out at 3300 and 3400. So discharging to
soc, the discharge current is more or less irrelevant. Interestingly
results are taken at constant cell temperature where any overheating
advantage is not applicable. Without seeing the complete paper that was
referred to, it is difficult to know if any comparison with manufacturer
data was made or whether tests were done at constant temperature and
the results were.
Discharging to a lower 15-20% level to protect the battery, there is a
big difference. If you want to get the best capacity out of a li ion
battery with a BMS, either reduce the discharge rate or change the BMS
accept a lower cutoff voltage and risk battery damage.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <

UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
Please discuss EV drag racing at NEDRA (

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
Please discuss EV drag racing at NEDRA (
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
Please discuss EV drag racing at NEDRA (

UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)

Reply via email to