jump to navigation

Tesla Semi and the price of batteries November 26, 2017

Posted by Maury Markowitz in Uncategorized.
Tags: ,
trackback

 

Screen-Shot-2017-11-24-at-8.25.28-AM-800x365

Even their trucks are sexy.

It was surprising enough that Tesla announced they were making a semi, but I think what really surprised everyone was the cost; the 300-mile range version is expected to run $150,000, while the long range 500-mile version is $180,000. They compare this to a similar class Diesel at $120,000.

This is interesting enough on its own, but now that we have the numbers, we can play a little game with them…

The entire issue in EVs today is the cost of the batteries. All of the rest of the technology is pretty much done; there are tweaks to come, but nothing major. The batteries will also improve, but barring breakthroughs, some form of LiIon is likely going to be used for the foreseeable future.

And the question behind the question of cost is what’s known as the ‘learning curve’ (or ‘experience curve’), which is the rate that your production costs decline as you build more of something. Here’s a famous example:

fordLearn.gif

This is the cost to build a Model T Ford over its production run. As you can see, the cost declined from $3,300 to $950 in constant 1958 dollars, which represents a learning curve value of 15% per doubling of units. The effect of this curve was to allow cars to become almost universal by the 1930s.

The same thing is happening with LiIon, which in the short term at least, is dominated by the cost of fabrication. The question, then, is what is the learning curve for LiIon? Well, that’s open to interpretation. Some say there is a lower limit around $100/kWh due to the materials costs, but others disagree. When you consider it all, most predictions put the learning curve around 21 to 24%, that is, much faster than the Model T.

So what does all of this have to do with the Tesla Semi? Well by giving the price and range figures, and making the reasonable assuming the difference in the price is due to the difference in the battery, we can see where we are along that curve. Basically…

(long-range-price – short-range-price) / (long-range-miles – short-range-miles) = $/mile

(180,000 – 150,000) / (500 – 300) =$150/mile

Now the question becomes what is the kWh per mile? Tesla’s made the somewhat roundabout claim of “less than 2 kWh per mile”. If we take that at face value, that means they’re making batteries at $75/kWh, which is already below what some claim is the lowest possible price!

However, that number does seem awfully high; a typical semi gets around 9 MPH, and that translates to something closer to 1 kWh/mile. In that case the price would be $150/kWh.

In any event, last year the company stated the price was $190/kWh, so it seems they are expecting a fairly impressive drop in price before the Semi ships in 2019. But no matter how you cut it, this suggests LiIon continues to follow a pretty steep learning curve.

One other note: it’s likely the batteries in the Semi are not the same ones as the cars. In a car the main measure is capacity per unit volume, but this is not nearly as important in a semi where you have lots of room and weight to work with. In that case, it seems likely they will use an NMC-based chemistry instead of the car’s NCA, as NMC can handle a lot more cycling, perhaps as much as ten times, which makes a lot of sense given the million-mile warranty.

Comments»

1. Tony S - July 31, 2018

I have read that the Tesla Semi does not have enough range for most long haul loads, and also that a new company (Nicola) is producing Hydrogen Fuel Cell trucks that have an excellent mileage range and that Nicola is willing to build a network of refueling stations. Do you know if this is true and how this might impact the Tesla Semi sales?

Maury Markowitz - July 31, 2018

The hydrogen cycle is less efficient than the electric cycle; a PEM-based cycle is around 50% compared to about 70% for LiFePo. In overall cycle terms, a fuel-cell car is at best some 25% to 35% less energy efficient, *even if that power comes from a gas plant*.

Moreover, the process of steam reforming, which produces the vast majority of hydrogen fuel, produces copious amounts of greenhouse gasses. On top of that, there is no existing distribution system, whereas an EV can be refilled anywhere there’s a plug, which is anywhere any of the other possible fuels are (gas stations use electricity too).

To date, the only thing that Nikola has produced is a pure-EV. The rest is all talk and 3D models. Who knows, maybe someday they’ll make one of these trucks and the refueling network to support it, but in the meantime Tesla semis are already hauling batteries from Reno to Fremont, so my money’s on Tesla.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s