Monday, October 12, 2020

Nerdgasm—Thoughts on Tesla's Battery Day

Like many, I was looking forward to what Elon Musk would say about what is easily the most critical and expensive part of building an electric vehicle—the batteries. The presentation was literally breath-taking. Several times I found myself holding my breath as I tried not to miss any important details. Musk and company had gone to great lengths to simplify their descriptions of what they were attempting. Unfortunately for me, I hadn't taken a Chemistry course since 1966 so I found myself fishing in some forgotten waters. Fortunately, it was easy to watch repeats on YouTube so eventually the requisite understanding returned. (I needed four tries ;-)

There is really no sustainable energy future without effective storage. Anyone who has ever lost their electricity knows that life gets extremely difficult VERY fast. Yes, I know humans survived without electricity in hazardous places such a North Dakota in winter. But those people were unusually hardy, strong, and clever. Even so, preparations for winter began the first days of spring. If you had trees, you chopped wood. Trees are stored energy. Unfortunately, there aren't many trees in North Dakota. Fortunately, there are energy sources that will sustain human life up there—lignite coal and then after the 1950s, oil and natural gas. Now that burning fossil fuels has become a big no-no, North Dakota also has abundant wind power but harnessing it as a practical matter requires storage.

Energy storage has been the big scary boogieman since the invention of fire. We like to treat the problem socially the same way we treat hauling out the trash and for many of the same reasons—it is UGHLEE. Clear-cut forests, oil refineries and tank farms, and the like smell funny and in myriad ways are offensive if not serious health hazards. Storage is especially tricky when the energy is electricity. Electricity moves at roughly the speed of light. With tiny exceptions, therefore, electricity must be used the instant it's produced. While this nearly instantaneous speed is easily electricity's most useful characteristic, it seriously hampers more widespread adoption of renewable energy.

The best and most useful way of storing electrical energy is by using batteries. This option has earned its important status because the alternatives don't work very well, if at all. Some proposed building huge flywheels. This scheme never got off the drawing boards because of the problems building a wheel precise enough to store massive amounts of energy. Bearings large enough to support a large flywheel at even slow rotational speeds were probably physically impossible. One scheme—pumping water into an elevated reservoir only to be released when the grid needs more electricity has several working examples operating in Norway. But these are only possible due to Norway's unique geography. There are also supporters of compressed air. Big money has been spent trying to make hydrogen the default storage medium. But in end, the most useful and reliable method of electrical storage is still the humble battery.

But just because batteries work, after a fashion, that doesn't mean they don't have serious problems. They are heavy, and expensive, and don't store all that much power. They are an environmental headache from mining to disposal. Lots of money and brainpower is being directed at solving those problems. There has been a lot of chirping about the potential for a solid-state battery while Samsung has made them a company goal.

But Tesla has a different sort of problem. They must use a technology that already exists at scale which means lithium-ion. So their solution is to throw resources and effort into production innovation—they intend to make terawatts of these things, after all.

In the early days of industrialization, folks who could create factories and make them run smoothly were called "millwrights"—a prestigious occupational category. Now the term has fallen into serious disuse as more and more companies that make complex products tend to outsource difficult tasks. Tesla and Musk has made a public declaration that they do NOT intend to follow that path. Instead they intend to employ the method used by most early industries—vertical integration. If someone is building a product that has not been built before, this is the only choice. If the tools of production cannot be purchased, the only alternative is to make them yourself (or find someone who can make them for you).

Musk most certainly did NOT invent vertical integration. In fact he has a ways to go before any of his manufacturing facilities equal the sophistication of Ford's River Rouge (Model A-1927) or Willow Run (B-24 bomber—1942). But while Musk did not invent vertical integration he is most certainly a true believer in its principles. This is his presentation for battery day.

Elon Musk


Reaction to the Elon Musk / Drew Baglino presentation fell into two main categories—the disappointed and the breathless. The investor classes were the most reliably disappointed. It wasn't flashy enough. It didn't introduce a new model of something. The promised new product timelines were often in the range of 2-3 years, NOT next quarter. The unearned-income crowd considers descriptions of the nuts and bolts of technological innovations to be a snoozefest—especially at the level of battery day.

On the other hand, the folks who stayed awake in science classes had what one wag called a nerdgasm. That's what it felt like to me. A friend called two days after battery day and less than one minute into the conversation he asked, "What's up with you? I haven't heard you so happy in 35 years." Guilty as charged.

There are several main reasons Tesla's battery day was so appealing.

It was strategically solid. The Tesla scheme requires the least amount of social change. I am a HUGE fan of high-speed rail and live on what would be an important link if USA built up its passenger-rail system. However, I am reasonably certain USA will never build such a system for the simple reason that the cities we have in USA were built during the age of petroleum and individual transportation devices. Cars and filling stations can serve such cities—light rail, buses, etc. cannot. The roads are built—assembling the land for dedicated high-speed rail alone is an unimaginable nightmare. The zero-carbon transportation future will be a function of electrifying the existing infrastructure. The success of this venture will be a function of how efficiently we can convert the internal combustion transportation system into an electric one.

The devil's in the details. Instead of chasing a breakthrough such as a solid-state battery, Tesla chose to chase smaller production upgrades. The reason this works is that it addresses a host of problems that needed attention anyway and if there are enough "small" improvements, they are likely to add up to breakthrough numbers.

Borrow, steal, copy. The example that made me giggle was the admission that Tesla had learned a LOT from the food-packaging industry—specifically bottling. This technology has been around for awhile. The all-aluminum beer can has been around since the late 1960s and it is still used as an example of metal-forming sophistication. The amount of pure genius expended on food packaging is usually beyond belief. So it not surprising that kind of genius could accelerate battery assembly speed by 7x.

And on it went—one technological triumph following another. Elon and Drew just rapping like they were sharing a barbecue. No notes or teleprompters—just a sound industrial pragmatism on an intense search for what will actually work in the real world. The biggest advantage the Producer Classes have is that they can subject their theories to real-world testing pretty easily. Objects that cannot perform are culled pretty quickly. Industrial pragmatism turned up past 11. 

Elon Musk often claims that manufacturing doesn't get nearly the appropriate amount of respect AND has vowed to make Tesla manufacturing world-class. Apparently he learned some important lessons trying to make the Model 3 assembly line work. In a few 120 hour weeks, Musk discovered why "millwright" has been the most difficult job humans do.

No comments:

Post a Comment