Tesla's 18650s versus larger format automotive cells

[Eberhard]

i checked out the current price for chinese 18650 with 2200mAh. The price by 100.000 is only 20cent each? The whole battery of my roadster for less then 1.400€?

 

[JRP3]

Probably not even close to the same quality and life cycle. I've heard some of the Chinese commodity cells are pretty bad.

 

[Eberhard]

That not an issue with quality. Its an indication for how cheap the cells can be produced. Even if you assume double price for higher quality it is still well below 100€ per kWh. In the press, they still report prices per kWh from 500-1000€ per kWh and this is the reason why electric cars have to be expensive.

 

[JRP3]

I realize the press is usually way off when quoting cell prices but talking to people who design and build cells there is a lot more involved in making a quality cell. The prices are a direct reflection of the quality. Also the press usually references low volume larger format prismatic cells used by most auto builders, actual production volume should lower those prices as well.

 

[JRP3]

Bob Stempel on Tesla's commodity cell packs:

What about Tesla's pack, which uses 6,000-plus laptop computer batteries wired together in series and in parallel? How can they make – and service – something like that cost effectively? "Their people came from the computer business, and they certainly had the capability to put together an algorithm to control that. But that's too many cells. I think over time, we'll see them shift to a more conventional kind of pack."

Yet Mercedes is using Tesla batteries in its test fleet, and Toyota has partnered with Tesla. "The Toyota/Tesla partnership arose out of a series of other issues. It was really a convenient marriage for Toyota, because they needed to do something with the Fremont plant. When you unravel that story, there's a lot more to it than just batteries. And Mercedes has sold off part of their investment and is backing away from that. That multi-cell concept is very, very difficult, so I think you'll see Mercedes moving toward bigger cells and fewer of them. They wanted to get their feet wet on an experimental basis so were trying a little of everything.

http://green.autoblog.com/2011/02/16/gms-bob-stempel-talks-about-lithium-batteries-and-what-will-ma

 

[JRP3]

Talking with a lithium battery researcher about the Tesla pack he commented that the Tesla temperature management system was overkill. If true this would lead me to hope that the design could be simplified and the amount of energy used to control cell temperature could be reduced in the future. I've never been comfortable with the 25% and more efficiency loss during charging that some have reported.

 

[Eberhard]

The whole system, that Tesla build around the 18650 cells including liquid-cooling with AC, BMS and lot of processors to supervise the system carry more cost then the cells itself plus the cost of assembly. Thats the reason for robot-assembling together with quality issues. This helps to increase the lifetime of standard cells from 300 full-cycles to more then 500 with the high-quality Panasonic cells (these cost properly a lot more then the cheaper chines cells)

 

[William13]

I disagree with Bob Stemple. The Tesla battery system has built in redundancy but is actually cost effective. The form factor will also be allowing the Model S to have a lower center of gravity and stiffer body than the competition. I can only wonder at the cost of 90 kWh battery from system a123. Would it have the form factor ( flat )? Would it require heating and cooling?

 

[Lloyd]

I don't believe that the temperature control is overkill if you are going to consistently charge this pac using a DC fastcharge system and keep all the parametes for the pacs under proper conditions. Watch what happens to the non-liquid cooled cars that try to charge an aging pac quickly. I predict a couple of fires!

 

[ckessel]

I don't believe that the temperature control is overkill if you are going to consistently charge this pac using a DC fastcharge system and keep all the parametes for the pacs under proper conditions. Watch what happens to the non-liquid cooled cars that try to charge an aging pac quickly. I predict a couple of fires!

Yea, I'm not a lithium battery researcher, but I find it hard to believe Tesla would intentionally go way overboard in a non-profitable fashion. They've had years to gather numbers from actual field use, more than anyone else. I can respect the researcher, but I'd think it's more likely he's missing data about why Tesla found the solution useful.

 

[JRP3]

The form factor will also be allowing the Model S to have a lower center of gravity and stiffer body than the competition.

Other packs can be mounted in similar fashion. The LEAF pack is in the bottom of the vehicle.

I can only wonder at the cost of 90 kWh battery from system a123. Would it have the form factor ( flat )? Would it require heating and cooling?

A123 would be much more expensive at the cell level since they don't have anywhere near the volume or years of working with the form factor. The cells are also not as energy dense as the LiCo cells, but they have much better power and lower internal resistance values so they heat up much less. They also don't have the thermal runaway issues that LiCo has, so they could probably get away without any active cooling, and some simple resistance heating should be enough in lower temps.

 

[JRP3]

I don't believe that the temperature control is overkill if you are going to consistently charge this pac using a DC fastcharge system and keep all the parametes for the pacs under proper conditions. Watch what happens to the non-liquid cooled cars that try to charge an aging pac quickly. I predict a couple of fires!

I would not recommend fast charging any battery pack consistently other than an Altairnano pack, (which Proterra does on a daily basis in their buses). Fast charging should only be an occasional occurrence if you want your pack to last. Other chemistries, LiMn, LiFePO4, have less of an issue with thermal runaway than the LiCo that Tesla uses.

 

[JRP3]

Yea, I'm not a lithium battery researcher, but I find it hard to believe Tesla would intentionally go way overboard in a non-profitable fashion. They've had years to gather numbers from actual field use, more than anyone else. I can respect the researcher, but I'd think it's more likely he's missing data about why Tesla found the solution useful.

I can think of good reasons to go overboard, warranty issues and safety issues. I'm sure they did a cost benefit analysis and decided better safe than sorry. It's a good business practice. My hope is that with time, and possibly with better cells, they can at least reduce the energy used for cooling during a charge. A 25% charging loss does not help to make the case for EV's.

 

[TEG]

It has been noted that on rare occasions you can trigger an overheat condition (and temporarily reduced power output) if you manage to overheat the pack.
For instance, I saw this happen when "drag racing" with the Roadster on a hot day over and over in a row.
(Usually the air cooled PEM or Motor is the first to overheat, but it is possible to affect the pack too.)

Given that, it doesn't seem the cooling system is "over engineered". It has to handle the worse case and it seems to be just good enough to do that (up to a point).

 

[JRP3]

That only means a sensor got hot and triggered a power reduction, it doesn't mean that temperature was actually dangerous for the cells. Powering down seems a reasonable response to hot passes on a hot track since its not what most passenger vehicles expect to see. I will agree that being a sports car the Roadster has to be designed for a more aggressive driving profile than most vehicles and the cooling system may indeed be appropriate.

 

[qwk]

I can think of good reasons to go overboard, warranty issues and safety issues. I'm sure they did a cost benefit analysis and decided better safe than sorry. It's a good business practice. My hope is that with time, and possibly with better cells, they can at least reduce the energy used for cooling during a charge. A 25% charging loss does not help to make the case for EV's.

I don't know where you get your info from, but the Roadster gets comparable mi/kwh (wall to wheel) with the Leaf when both are driven conservatively. In cold weather, I bet the Leaf gets substantially worse numbers.

From all of the reading and research I have done about both vehicles, Elon may turn out to be right in his "primitive battery pack" statement.

 

[vfx]

There were many "overdesigned" features on the first Roadster. The charger originally came with a fire detector and had several shutoff features that have been eliminated.

 

[doug]

Other packs can be mounted in similar [i.e. low] fashion. The LEAF pack is in the bottom of the vehicle.

True, but no one has such a thin yet energy dense pack. That Model S pack is truly amazing, ideal packaging.

I can think of good reasons to go overboard, warranty issues and safety issues. I'm sure they did a cost benefit analysis and decided better safe than sorry. It's a good business practice...

There were many "overdesigned" features on the first Roadster. The charger originally came with a fire detector and had several shutoff features that have been eliminated.

It is good practice and Tesla was certainly very careful in the beginning. Martin (or someone) said that they worried if they ended up having a fire, it would set EVs back again by years.

 

[JRP3]

I don't know where you get your info from, but the Roadster gets comparable mi/kwh (wall to wheel) with the Leaf when both are driven conservatively. In cold weather, I bet the Leaf gets substantially worse numbers.

I get my numbers from people on this board. Miles/kwh takes into account vehicle efficiency, I'm talking strictly charging efficiency. 25% loss is terrible.

From all of the reading and research I have done about both vehicles, Elon may turn out to be right in his "primitive battery pack" statement.

How so? Have any LEAF packs been damaged from overheating?

 

[JRP3]

True, but no one has such a thin yet energy dense pack. That Model S pack is truly amazing, ideal packaging.

Yes the density of LiCo is superior.

It is good practice and Tesla was certainly very careful in the beginning. Martin (or someone) said that they worried if they ended up having a fire, it would set EVs back again by years.

Totally agree.