Share →

We’ve made no secret of the fact that I’m a bit of a fan of the Tesla Model S. I’ve taken a significant position in their stock and in fact have done well enough in it that it will basically buy me a Model S. Put in my reservation this morning actually.

I recently received another version of the oft repeated request that we examine making battery modules out of itty bitty batteries after the fashion of Tesla. We typically get 100-110 wH per kilogram with our stodgy old CALB or Thundersky Prismatic cells, while Tesla enjoys 200 mile ranges and over 200 wH per kilogram with their magic sauce battery pack.

There are actually a number of reasons.

The first, is that making that number of connections flawlessly is a herculean task. And as you know, I’m suspicious of connections anway. I actually did by a spot welder and have experimented with it modestly, and we do play around with this a little. I have some K2 cells on the way and we’ll probably let Matt weld me up an equivalent battery and do some tests just for your entertainment.

The second reason is that that higher energy density requires quite a bit of cobalt in the cathode. Cobalt is expensive, it is hard to come by, and my problem with it is it is a little bit negative temperature coefficient wise. They are not thermally safe to my way of thinking, at least not in the league with LiFePo4.

Then too, they tend to have limited POWER capability. About 2C is what you can do to these cells, and so you do have to have a lot of them.

But none of that is the real reason I eschew these cells. Tesla of course uses Panasonic 18650 cells. These are little cylinders of 18 mm diameter and a length just slightly over 65 mm.

Let’s take a look at a spec sheet for Panasonic’s typical recent 18650 lithium ion cell.

Here’s the real problem – 300 cycles. LGChem version of the same cell – about the same. Actually, LGChem states >75% of minimum capacity, further fudging the factor.

The bottom line is these are 300 cycle batteries. We could reasonably expect 3000 cycles. A world of difference. Why would I pay DOUBLE the cost of the same size pack, to get 1/10th the life?

The cycle life is the central issue with these cells. It is true that they offer a serious weight and volume advantage over lead acid or NimHd cells, making an electric car viable really for the first time to my way of thinking. But of course the price difference is discouraging. More discouraging is having a reasonably operating lead sled car that in three years is uselss without a new battery pack. LiFePo4 cells pretty much are a lock to 7-10 years if you don’t murder them with a BMS. That makes the operation of the car a very different expense.

So after weight and volume, my big item is cycle life. And 300 isn’t really any better than lead. Tesla never really mentions the life expectancy of their proprietary pack.

I think I’ll like my Model S when it arrives. But I’m not going to be very happy paying Tesla for a proprietary pack every three years. That’s the dirty little secret nobody is really talking about. And the Roadsters haven’t really been out there long enough for it to come up – though there HAVE already been some under-warranty pack replacements.
It should make Tesla very profitable selling razor blades for these cars – a recurring revenue stream.

Jack Rickard