Ener1’s Approach to HEV, PHEV and EV Batteries
29 October 2007
 |
| Rate capability of the Ener1 LMO/LTO HEV cells. Click to enlarge. |
In preparation for a series of investor meetings during November, Ener1 has published a presentation outlining its approach to the two different types of automotive lithium-ion batteries it currently has under development: high power batteries for hybrid electric vehicle (HEV) applications (earlier post), and high energy batteries for plug-in hybrid electric (PHEV) and full electric vehicle (EV) applications.
Ener1 is using two different sets of electrode materials for each type. In the HEV batteries, the company is using a lithium manganese spinel (LiMn2O4-spinel, LMO) for its cathode material, and a lithium titanate (Li4Ti5O12, LTO) for the anode material. The energy batteries (PHEV and EV) are based on a layered lithium metal oxide cathode—LiNiCoMnO2—and a hard-carbon anode.
 |
| Overcharge and nail penetration test results comparing a commercial 26650 cell and a Gen1 Ener1 HEV cell. Click to enlarge. |
HEV batteries. EnerDel developed its LTO anode material in collaboration with Argonne National Laboratory (ANL). The LMO/LTO combination for HEV cells offers a combination of high power and safety, according to Ener1. LTO has high rate capabilities, and supports a fast charge even at low temperatures.
Use of LTO in the anode rather than graphite reduces the possibility of runaway thermal events and extends battery life. Whereas graphite may experience up to a 9% volume change over the battery life, LTO will experience an approximate 0.2% volume change.
The Ener1 HEV battery runs very cool, and does not require a liquid cooling system. It can discharge to a very low state of charge, and also has strong low temperature performance, which is critical for automotive applications.
The downside of the LTO anode, however, is a lower energy capacity. LTO as an anode material has a capacity of about 150 mAh/g—about half that of graphite. LTO also has a lower voltage.
PHEV and EV batteries. For the high energy capacity batteries required in PHEV and EV applications, EnerDel is turning to a layered lithium manganese oxide cathode, and a non-graphite carbon (hard carbon) anode.
In general, lithium manganese oxides are of interest as cathode materials due to the safety, low cost, and low toxicity of manganese-based materials as well as to the high power enabled.
However, the material can suffer poor rate performance and capacity fading during cycling. To address those and other structural issues with the material, researchers are exploring the use of layered materials, with LiNixCoyMnzO2 (called NMC) gaining momentum. One of the more popular formulations for NMC is 1/3-1/3-1/3—i.e., LiNi1/3Co1/3Mn1/3O2
(or “333”).
 |
| Rate capability testing for the NMC/HC 7 Ah cell. |
A hard carbon anode is capacity tunable, and supports better rate capabilities on charge and discharge than graphite.
EnerDel has begun cycle life and rate capability testing of a 7 Ah cell with the NMC/HC chemistry. (By comparison, the HEV cells come in 1.8 Ah and 5 Ah units.) Initial results indicate a good rate capability from 1C to 5C, and also a good cycle life. EnerDel also expects to be able to produce further improvements in the performance of the cells.
Resources
The disadvantage to the LTO anode is the low voltage and relatively poor energy density - and EnerDel itself has explicitly admitted this:
http://www.ases.org/solar2007/presentations/tuesday/400pm/forums/4-battery/4-ota.pdf
AltairNano's cells (which use a similar technology) achieve only about 78 Wh/kg.
The upside is impressive cycle life. Exactly why the LTO cell design is being reserved for HEV use, where a premium is put on electrical power assist.
As for hard carbon, the same link above indicates that HC also pays a penalty in energy relative to graphite, despite being more stable, although by how much I don't know. EnerDel's HC cells have a voltage of 3.6V, which is pretty standard for LI-ion cells, so maybe it's not that severe of a disadvantage.
Posted by: AES | 29 October 2007 at 11:04 AM
It should be possible to combine these concepts, with one set of batteries optimized for relatively shorts bursts of power (in either direction) and the other for high capacity. The power electronics would implement the power flows to and from each of these battery banks.
Similar approaches have been proposed for a combination of supercaps and batteries. However, supercaps are still ver expensive and tricky to work with because of their inherently large variation of voltage with state of charge.
Posted by: Rafael Seidl | 29 October 2007 at 11:06 AM
Rafael-
The same two battery concept has been proposed and executed before, although it was a long time ago. If I find the exact reference I'll post it - it may have used NiCads and Zebras, but I'm not sure.
With respect to the EnerDel technology, I'm not sure if there's a significant enough of a power density difference between LTO and HC to merit such a setup. We'll have to wait until more specs are released.
Also, do you know if it's possible to chill something below ambient temperature using air cooling? My common sense says no. Reason I ask is that the other battery manufacturers are implementing liquid cooling to extend calendar life, rather than to control thermal instability. So perhaps the EnerDel cells may need liquid cooling afterall?
Posted by: AES | 29 October 2007 at 12:03 PM
Just looked at their technical presentation on the website - they have life cycle graph for HC anode cell. It indicates 80% of initial capacity remaining at 700 cycles, with 1C/1C charge/discharge, and 30 degree C conditions. That is noticeably worse than the LTO chemistry, as well as other high energy cell competitors (LiFePO4, etc).
Guess that's why they have the disclaimer "data demonstrates further improvements capable"
Posted by: AES | 29 October 2007 at 12:12 PM
How does temperature affect Lithium batteries as opposed to NiMH? I know that output is affected negatively with traditional lead acid batteries as the temp goes down, but does this hold for Lithium or NiMH?
Posted by: | 29 October 2007 at 01:05 PM
It varies with the particular lithium chemistry, but high temperatures tend to make for optimum power, but shorten the calendar life. Reverse for cold temperatures - low power, longer calendar life.
Posted by: AES | 29 October 2007 at 11:02 PM
there is another graph that shows 95% capacity retention at 1000 cycles, 5C cycling to 100% depth of discharge.. that looks pretty good compared to other lipos, that should translate to several 1000 cycles at a gentler, usage pattern.. 5C discharge is a pretty violent discharge for a large automotive pack, it would discharge the cells in 12 minutes
Posted by: | 30 October 2007 at 03:49 AM
"there is another graph that shows 95% capacity retention at 1000 cycles, 5C cycling to 100% depth of discharge.. that looks pretty good compared to other lipos"
That's presumably for LTO anode, not HC.
Posted by: AES | 30 October 2007 at 05:00 AM
Why not build a car with a powerful electric dominant drive train in the rear wheels with the HEV batteries (let's say 100 Kg of high power, low capacity batteries). The rear wheels will take care of most of the acceleration and a range of 30 to 40 miles. Now put a much less powerful electric drivetrain in the front wheels with high capacity, high charge/discharge cycle, low power PHEV/EV batteries (let's say 200 Kg). The front drive will give the car an electric range of 120 to 140 miles. Now have a 40 KW Atkinson engine working in HCCI mode for maximum thermal efficiency coupled to a generator that can keep the rear batteries charged from 30% to 40% and there you have it, a PHEV with 120 miles range from the front + 30 miles * 70% = 141 miles of electric range.
Posted by: Freddy | 30 October 2007 at 06:29 AM
300kg of batteries?? You might as well go all-electric at that point. The two-battery scheme shouldn't be neccessary either - if you have over 100kg of the high energy cells in series, the voltage and amps generated should produce more than enough power to move the car.
Posted by: AES | 30 October 2007 at 10:38 AM
Dreisbach combined a NiCad flywheel battery with a ZnAir energy battery in a race car (Honda CRX conversion) - but for a normal car the 50kWh ZnAir has plenty of power anyway, even at only 100W/kg peak power. PbA has also been used with ZnAir by Electric Fuel.
Why bother with LiIon at these low specific energies? Stay with NiMH - LiIon offers no significant performance advantages at these levels. NiMH specific energy is improving dramatically too for pure energy applications - 3Ah AA cells now as Fetcenko has pointed out since 2005.
Posted by: Emphyrio | 01 November 2007 at 05:54 AM