EV WORLD <-- OP-ED
Low Cost Lithium-Ion Batteries
An Open Email From Bill Yerkes
Prismatic nickel metal hydride (NiMH) batteries are expensive to make. My experience includes development
and pilot production of sealed nickel cadmium (NiCad) cells at Spectrolab. Processes were similar to making
NiMH cells (at least in terms of cell parts count and assembly steps). In 1997 and 1998 I toured nickel battery
plants at Eagle Pitcher in Joplin, MO, the SAFT Plant in Bordeaux, France, and the GS Battery plant in Kyoto,
Japan, reviewing their Nickel battery production. At these three plants we also reviewed their pilot lithium-ion
(LiON) cell production facilities. Polystor Corporation has similar LiON manufacturing in Livermore, CA, with
process equipment shown on their web site at: polystor.com LiON cells are easily
automated by using web-winding equipment as shown in the Polystor website. Because LiON cells are twice the
energy density and three times the volumetric density of NiMH there is not as much need to make them in
"prismatic" shapes.
The main components of LiON cells are not expensive. Copper foil (for one electrode) and aluminum foil (for the
other electrode) are the carriers of the active materials. In the factory, the rolls of foil are put on machines and one
is coated with carbon powder and the other with metal powder. In the case of the SAFT battery, which I have
the best performance test data on, they claim to use a mixture of lithium-nickel-oxide powders so there is no
cobalt required (which is brought up as the "expensive" element by the Ni battery people). Most of the Japanese
manufacturers are using nickel or manganese now, only Sony is sticking to the original cobalt- oxide recipe
invented in England by Harwell Atomic Agency and licensed to Sony in 1988.
The coating of the aluminum or copper foil at low temperatures is very easy to automate. LiON batteries are put
together by winding a positive roll and a negative roll with a roll of separator in between (very thin plastic). The
rolls are tabbed and stuffed in a thin aluminum shell like a beer can. The lid is laser-welded on, and the can is put
under a vacuum and then filled with a measured amount of electrolyte. Cells are charged up and checked for soft
shorts to make sure the self discharge is very low (self-discharge on NiMH batteries is very high). SONY has
made high volume production cells since 1989. In 1995 (the last year I have data for) there were over 300 million
LiON cells manufactured by seven companies world wide (it is not R&D project).
Everyone I have talked to that makes LiON has told me they will be inexpensive. LiON battery performance is
so outstanding that the battery industry can still sell them for high prices. However battery companies still have
lots of production equipment and users for NiMH and NiCad, so they are selling them to all who will buy.
SONY, Shin Kobe, Sanyo, Panasonic (Matsushita) and GS Battery in Japan have all made pilot quantities of
100 amp hour cells for EV's. According to Peugeot, SAFT has a major program to provide Lithium-Ion batteries
for the French EV programs. The SAFT cells are about half the size of the Japanese cells and designed for a
liquid- cooled battery which was first revealed in 1996 at the International EV meeting in Osaka.
SAFT cells have also been completely qualified for use in high reliability LEO and GEO spacecraft. The first
SAFT space LiON cells will be launched late in 2000 as the main battery in the STENTOR spacecraft.
Lithium-ion batteries are being designed into other spacecraft to save weight and volume, but there are other
important attributes. First, the in-to-out electrical efficiency is 95% compared to Nickel batteries at 70%.
Second, the self discharge is 5% per month rather than 5% per eight hours. Third, the LiON batteries run at room
temperature and are not pressurized, making handling safety less of a problem. The batteries can therefore be
installed in the spacecraft on the production line, rather than at the launch pad. As we know from the GM EV-1
experience, NiMH heating from the lower electrical efficiency is a problem. Fourth, LiON is four volts per cell, so
three NiMH cells at 1.2 volts per cell are required to equal one LiON of about the same amp-hour capability.
On-third the number of copper interconnects means less weight and less battery assembly labor.
Low-cell voltage is really the problem with the upcoming fuel cells; their voltage sinks to 0.7 volts per cell under
load, and so you need five fuel cells in series to give the voltage of one lithium-ion cell. Stacks of cells to provide
300+ volts for high power systems in cars and buses will require a lot of fuel cells in series. Remember that each
fuel cell has to have hydrogen and oxygen connections and water drain, in addition to plus and minus electrical
connections. To get a 300- volt battery you need 500 fuel cells in series with 500 hydrogen and oxygen
connections plus drain? Calculate the reliability of this in an engine compartment on the road. The same job can
be done with 96 Lithium Ion cells in series at four volts each, giving 360 volts as in the Nissan EV.
Peugeot and Nissan are both publically experimenting with Lithium-Ion batteries. A number of other auto
manufacturers are testing them. However, Detroit is not very motivated to bring on the LiON battery because it
starts making the case for EV's. It's a production solution to the battery problem, not an R&D program. Car
companies would rather go to hybrids like the Prius and Insight and avoid the charger problems. However, UC
Davis Institute of Transportation Studies (ITS) and USABC have all the data, and say that LiON is the way to
go. My personal experience is that once you use lithium-ion you have a hard time going back to lesser batteries. |
