AABC discussion on automotive lithium battery requirements
The hybrid battery test lab at the GM Technical Center
Earlier this week Long Beach was host to the Advanced Automotive Battery and Ultracapacitor Conference (AABC) which featured a panel discussion on what it will take to bring lithium ion battery technology to the mass market. Included on the panel were GM's Joe LoGrasso and Mark Verbrugge and Ford's Ted Miller. All acknowledged that lithium batteries are inevitable, and that it's just a matter of ironing out the details over the next couple of years. Of course many of those details aren't exactly trivial, but they can be overcome.
Nickel metal hydride batteries are known quantity and thermal and charging management are fairly straightforward. The biggest problem is limited energy/power density when compared to lithium batteries. There is also a decent amount production capacity for large scale NiMH batteries although nickel prices have been on the rise of late. To date no one has produced large quantities of automotive-sized lithium cells. Tesla is building their battery packs up from 6,800+ laptop cells which means each pack will have almost 14,000 interconnects any of which could be defective or fail. The big manufacturers want to use larger cells to cut the number of interconnects and reduce the number of potential failure points.
Anther area that battery manufacturers and car companies are addressing is standardizing interfaces between the vehicles and battery controls as well as the way battery life and charge history is calculated for warranty coverage. Improved models of the battery behavior are important to ensuring that the management software extracts the maximum output and life out of the battery. There's more at GreenCarCongress.
Related:
[Source: GreenCarCongress]
Reader Comments (Page 1 of 1)
Peter 8:12PM (5/19/2007)
Does it really matter if a cell in a Tesla battery does fail? I mean, if 6800 batteries give you 200 miles range, then a cell failure would cost you about 150 feet.
Of course, a substantial number of failures would add up and the effects on maximum voltage and current could become noticeable. I would be interested to see how the reliability numbers would work out compared to using 3-4 larger cells.
Reply
Tony Belding 11:21PM (5/19/2007)
I had the same thought. Tesla's design includes massive redundancy, that seems to me like a plus for reliability. I suspect a larger problem with those 14,000 interconnects is how they drive up the assembly cost.
Reply
Ron Fischer 1:45AM (5/20/2007)
Tesla's power pack assembly line could be highly automated, gives them great flexibility in selecting a cell that meets cost/quality/power needs, and finally, individual cell failures are handled similar to a hard disk drive: the controller patches around bad sectors.
I wouldn't be too quick to dismiss this approach as it does have some advantages. It certainly has kept them from claiming that the batteries just aren't ready...
Reply
Sam Abuelsamid 8:25AM (5/20/2007)
The real issue with the batteries is that you have three parameters: cost, reliability/durability and performance. Right now you can pick any two of the three. The carmakers are trying to get all three. They want to sell battery powered cars for $25-30,000 that last and have decent range.
Tesla is taking a different approach but they are competing in a very different market segment. The Tesla design should provide the range and life but it's just not a cost effective approach for mass market cars right now. It's good that they are taking this important incremental step and hopefully it will work and everyone can learn from their experience.
Reply
Chris 9:54PM (5/20/2007)
From the designs details I have heard in their blogs, I believe Tesla is also using a modular design. Similar to a blade server configuration. The thousands of cells are assembled into larger modules, if a large portion of cells go out in one "blade" they can replace just that portion of the energy storage system.
Reply