September 11, 2009

Volt battery costs, part 5

Are li-ion battery costs really dropping?

Yes. here are some retail costs: http://www.evcomponents.com/SearchResults.asp?Cat=34

We see that current Lithium cells are about $350/kWh for individual purchases. We can expect that an OEM can get them for around 50% of that (no more than $200/kWh), which places GM's wholesale cost for the Volt pack in the neighborhood of $3,200.

The range in the Volt is electronically limited in order to avoid any warranty issues with pack replacement (due to California Air Resources Board requirements). Essentially, GM's only letting the pack discharge to about half, so when capacity drops with age/cycling, as it does with all batteries, they can get more mileage out of it compared to going with a smaller pack and having the range drop below 40 miles within twenty+ thousand miles.

Going by specs for the retail batteries above, 5000 cycles before they hit 70% capacity would be at least .7(40 miles)5000 = ~140,000 miles until the pack capacity degrades to 70%, and probably ~200,000 miles before it degrades to 50% and drivers can't go a full 40 miles on all electric power w/ something like the Volt.

GM had to, in effect, de-rate their battery pack because the California Air Resources Board requirements for PHEVs are very stringent. Pure EVs don't have these requirements, so manufacturers can get away with using the whole pack.

4 comments:

Anonymous said...

In terms of the Volt's specific pack chemistry, they're aren't using LFP cells, they're using LiMn2O4 cells. According to tests, they're good for about 80% capacity after 10,000 cycles at 30% depth of discharge, which would correspond to a ~25 mile commute in the Volt, and the pack reaching 80% capacity after ~250,000 miles in that situation. Large volume costs may be greater than LFP cells, although I doubt it because Manganese is pretty cheap. ~$4000 in pack costs seems pretty reasonable, at least once economies of scale are realized.

Nick G said...

Anonymous,

Thanks for the info.

Could you point me to sources for your longevity data?

Anonymous said...

A total pain to find because of a type (limn204 instead of limn2o4) but here it is.
http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/38085/1/04-1055.pdf

Nick G said...

Thanks.

Wow - they look great:

"the cell has completed over 1600 cycles with minimal capacity fade (0.010 %/cycle), corresponding to over 83% of the initial capacity (-88Y0 @ cycle #l,OOO).
During the course of this test, very stable performance continues to be observed with high columbic efficiency (>99%) and excellent watt-hour efficiency (98.4%)delivered after 1600 cycles have been delivered (over 2 years of operation)."

Greater than 98% efficiency!