Inside a Tesla Roadster Battery

Just found this (unrelated):-



Progress is progress...

what a project ! Thanks for the link.

You are welcome, I thought you might enjoy seeing and reading about it.

What really strikes me is all of those fragile 7000+ eggs in the basket. I know why Tesla choose cylindrical cells because of the specific energy density, but dang the nightmare of assembly, monitoring, and charging all of them all. Does not take but a few bad cells to spoil the pack. Prismatic or Pouch large format IMHO is a better way to go.

The cylindrical cells make for easier cooling, and the separate envelopes helps to limit the risk of mechanical damage or thermal runaway spreading from cell to cell within the battery. I would like to read more about the process that went into their design decisions.
Then there is the fact that using lots of small cylindrical cells gives them a lot of flexibility to fit the overall battery into a limited and non-rectangular space (whether they choose to take advantage of that or not.)

I can understand goofing on it, but for an off-grid solar application, those are the WRONG cells to use.

They are designed for super high charge / discharge rates, so in our application where one designs the load to last for more than just a hour or so, that capability is wasted and is an unnecessary additional expense.

Not only that, but I'm pretty sure those are NOT LiFePo4, but are a custom mix more along the lines of LiNMC, which are higher up the energy density scale, and therefore are more critical of proper charging techniques.

No doubt the batteries of are of high quality, where capacity and internal resistance of each cell is very closely matched, allowing for "module" level monitoring rather than individual cell level infrastructure - which in this case is just not practical at all. So one relies heavily on the quality of the individual cells.

Unless he got them for free, or just wants to play around, then that money would have been better spent on large prismatics like GBS, CALB, Winston, etc, especially because in our application, space isn't that tight, and super high C rates of 10C or more which small cylindrical cells can do, are just not experienced in an off-grid solar installation - commercial entities excluded of course.

It is funny how much emphasis we put on choosing the right battery, ie deep-cycle vs SLI, yet when it comes to Li-ion, people just aren't aware of any differences in chemistry or application whatsoever and lump it all together. The problem is that this project promotes that unawareness.

Not so sure that is the reason they used them because you can make Prismatic and Pouch cells in any shape you want and not have all the wasted space between cells. Pretty sure Tesla used them for specific energy density and high C-Rates. The exact lithium battery the Roadster uses is debatable but most think they are LiCo which is a reasonable conclusion because the battery pack is liquid cooled. One thing you do not hear much about is the Roadster needs to be connected to a power source anytime it is park even if fully charged up because the 150 watt coolant pump runs even when the vehicle is turned off.

I attended a talk by them in ~2008 (got to even sit in a roadster they had parked outside)
The roadster battery pack was same cells used in standard laptop batteries.
The reasoning at that time was cost.

what I could find for articles about the roadster battery packs:



IIRC their initial battery/module assembly "plant" was in a tropical/subtropical climate and was just a roof with no walls. (Malaysia maybe?)

The roadster definitely didn't need to be plugged in all the time - they drove it to the talk, parked it in front of the building, and drove it away after (probably ~3 hours later)
Maybe it was supposed to be plugged in overnight every night - I never had the money to spend on one so don't know about that.

The carbon-fiber body panels were really interesting. One interesting thing I hadn't thought about before they said it was that the "raw" carbon fiber look was actually more of a PITA to do, because they wanted everything to look perfectly aligned. That made the fabrication more difficult. On the candy-apple red one if the carbon fiber sheet was skewed a few degrees it didn't matter because it wasn't visible.

Heh, gotta' love reviews from backseat-engineers.

Yes, batch of A123 cells that were contaminated was disastrous for Fiskar. They got fixed, and yes, both went bankrupt, which the average consumer doesn't understand at all in the world of business. Bankruptcy doesn't automatically mean down for the count. Makes for good press though.

A123 cells are actually fantastic, and the company still exists and are used in today's products despite the drama from the review:


However, these cells have far more high-rate capability than we as energy-storage users will ever need, where the large prismatics like GBS, CALB, and Winston are more appropriate.

What is adamantly clear about that review is that he may not understand today's Elon Musk or Tesla at all. Elon wants to PROMOTE competition in the marketplace, much like Henry Ford did back in the day (even allowing for patent sharing [not just taking!] like Henry did to build the industry), rather than sitting around on a throne wringing his hands contemplating a monopoly. Far from it, as Henry Ford had to face a monopolistic patent problem of ICE vehicles (1985 Selden patent from ALAM) before he could introduce his own ICE car after electric was abandoned.


and


Ok, maybe I'll have to cut the reviewer a break since this memo came out 60 days later.

I recall reading many years ago that the cell shape/type was chosen as they were basically AA battery format, which are tried-and-true, they can be and are manufactured by many, so fast adoption may be possible. In other words, why reinvent the battery? It may also go down the path of Musk's wanting to change the world without reinventing the wheel, just improving it, and fostering competition to ensure mass adoption of his technologies.

His root agenda is change the world by removing it from dependence on petroleum, et all, by both changing how we consume power and how we generate it. Hence, his involvement in both Tesla and Solar City. Perfect it, and he can use SpaceX to take it to Mars...

Bottom line, it is nice to see an entrepreneur spend his $200M on these projects instead of living on a Yacht.

Metallic Cylindrical cells were used to be able to withstand the extreme pressure LiCo builds up when charging and discharging. Prismatic and Pouch are made to expand.

I seem to recall reading that the combination of plate active material and separator thickness, in the type of cell chemistry Boeing used, changed thickness by something like 10% between full discharge and full charge states.

Yes, but that is just an indication that you have abused the cells. There are those that say that banding is to prevent this, but if you reach a state where the cells want to expand, one is already degrading them, even if the banding is holding them tight.

Also, once expanded, the idea of squeezing them back together is also a bad idea. Not only has the condition for abuse already been established, but the internal anodes and cathodes are no longer aligned properly, losing the critical "overhang" necessary to keep dendrites at bay.

In other words, once you've swollen, you are done. Except for those making dramatic videos.

Yes. With lifepo4, there is really no additional changes in thickness during intercalation.

One additional problem with non-lifepo4, is that while lifepo4 will actually absorb down to zero amps (not really wise to go below C/20, but anyway), when that absorb is done, additional charging only serves to heat the electrolyte, and voltage rises. Still, there is no current flowing to add to a thermal runaway, just electrolyte heating.

Other chemistries like LiCo02 will still accept current well past 100% SOC, and add to possible thermal runaway. That's why even 0.1v or more over spec is bad if not caught early.