North America is now the growth leader for new battery factories

We are also behind in having the minerals necessary to build batteries no matter where the factory is.

“given existing reserves, it is possible for the United States and its key partners to significantly friendshore production. However, given current production in democratic countries, it would require an unprecedented build-out of the mining industry to achieve 2030 clean energy targets.”

https://carnegieendowment.org/2023/05/03/friendshoring-criti...

And China had ... no? large-scale battery factories 10-15 years ago? What's the point? people forget that China was not a manufacturing powerhouse 25 years ago. Building factories at ridiculous speed & scales is something that humans are, apparently, really great at — it's the whole point of the Industrial Revolution.

Well we have to start somewhere. When you first put your foot on the gas the acceleration comes first, then the velocity, then position. Poo-pooing the initial acceleration because you haven't moved very far yet is rather short-sighted.

Heavy industry isn't a freaking SAAS startup with an indoor playground and free beer where profit isn't expected because you're just going for a buyout. It takes many years if not decades to build out brand new physical supply chains on any meaningful scale. Yes we're just getting started, but without this metric we'd be going nowhere.

Chinese high demand for EV responds to needs that do not exist in North America at the same scale.

The need to reduce noise from motors, the need to reduce emissions in much denser cities.

https://youtu.be/q4DtOhe2LfQ

Not to be that guy but I think this is a second derivative metric for growth. Acceleration is the second derivative. Position is owned, producing is velocity, and increase in production is acceleration. Jerk would be rate of acceleration growth, which is probably also positive.

Companies don't just abandon billion-dollar-plus factories.

While I mostly agree, I think that many corporate leaders are starting to realize there's other reasons to choose to build in a country beyond the bottom line numbers. There's ongoing relations, restrictions and even supported/sanctioned IP/Design theft to consider. For better or worse, making anything is largely going to be global, and a big part is going to come down to not just payments, but other incentives to produce in a given region.

I think that the US and Canada are more incentivized to be competitive. My biggest concern is dealing with runoff/waste from production facilities. I'm also concerned that rapidly expanding EV is too far ahead of grid support, not to mention even recent legislation restricting new car sales to only 3rd party dealers (middlemen taking big cuts), and they don't want to sell EVs that require less maintenance.

Governments have spent trillions subsidizing everything from the Internet to electricity and somehow we still use both the Internet and electricity.

It’s the same for the US.

At least we’re getting rid of the fiction that the global manufacturing sector is some sort of libertarian free market paradise where capital will always find its way to the ideal geographic location.

This clearly shows that manufacturing basically ends up wherever govts pay them enough to end up. Either through overt subsidies, or implicit subsidies through cheap labor, lack of regulations, etc.

Same for China. All of the amazing productivity of China is basically subsidized by the Chinese state, which really means it is subsidized by the Chinese population. Westerners get cheap products and the Chinese population pays for it and stays poor themselves. Only a fraction of the Chinese population lives at western living standards - the rest are still dirt poor.

Unfortunately, it will be the next president who will profit from Biden's policies.

I will wait until the plants are built, operating, and running profitably, without ongoing per kWh subsidies, paid back the debts to taxpayers, and the cells have been operating for a few years without major recalls, fires, etc. Then I hope to be able to declare success.

> it has model specific weird shapes with proprietary connectors

No. The shape is cylinder. They're not model specific. https://cleantechnica.com/2020/09/22/everything-you-need-to-...

The cell themselves are fairly standard, it's the packs themselves, cooling systems, BMS, etc, that are custom.

It's not obviously true that the weight and bulk of the batteries necessarily means that production will be localized. After all, the input materials are just as heavy and bulky if not more so before refinement.

And many of those inputs are mined from only a handful of concentrated places on Earth, and so need to be shipped long distances regardless in order to disperse to where humans actually live and demand the end products.

A better argument for localization might be that the dangerous and explosive nature of battery cells makes shipping the finished product longer distances more risky (and therefore more costly) than shipping the non volatile inputs long distances. So the extra cost of shipping cells drives localization of production.

The question is why your belief is so widespread. The battery pack in a Telsa S or X model was literally 7000 flashlight batteries in a metal box. The newer standard-range pack from the 3 and Y models has thousands of larger cylindrical cells that won't literally fit in standard flashlight but which are nevertheless off-the-shelf Panasonic catalog batteries.

There's a global battery shortage. Tesla will put whatever cells it can into their packs in the coming years, including soft 'pouch' cells.

18650 and 21700 (aka 2170) are standard cylindrical shapes for EV battery cells, so the comparison to AAs falls flat. Tesla has launched the 4680 cell, and the industry may move in that direction. Batteries are custom on a per car basis, in the same way that remote controls are all shaped uniquely, but almost all of them will take either AAs or CR2032 cells.

Pouch and "blade" cells are much less standard, and are also used in a bunch of EVs, like GM's Ultium platform.

Following through the examples in the article, the battery factories are all also including battery cell manufacturing either onsite or nearby. (The BMW example has "an agreement to source next generation lithium-ion battery cells from Envision AESC, which will build a new plant in the state." The others are combo cell+battery factories, like Tesla does.)

Shipping high-value items can be an astonishingly small part of the total cost; especially if you're sending it by sea. A few years ago I calculated that an (at the time) maximum sized container vessel carrying a full load of electronics in containers (which is very unlikely, but I picked a high value density item just to get an idea of the general numbers) would be carrying a cargo worth on the order of a billion dollars, and the cost of shipping each electronic good (I picked something around the size of a small PC tower; about 0.05 cubic metres) across an ocean, port to port, was on the order of a dollar. That kind of money.

So I guess if I scale that up to an EV battery, volume about a third of a cubic metre, so about seven or eight times the volume, that's still less than ten dollars per EV battery shipped across an ocean, if you've got a ship full of them.

I don't know what the total parts and labour cost of an EV battery is, but by this admittedly fast, brutal and half-remembered back of envelope calculation the numbers only have to be a few tens of dollars different to make it worth building them on the other side of the planet and shipping them across an ocean.

Yes. Contrary to popular belief, cells with 18650 and 21700 cylindrical form factors, despite being approximately the same size and shape as other 18650/21700 cells, are not particularly "general purpose" as you or I would think of AA or AAA batteries. Major customers would likely have their own variants and particular specifications on variable parameters associated with these cells. And battery pack makers cannot easily switch between cell types and manufacturers.

Source: I used to work for a major 18650/21700 cell manufacturer. I have torn down many battery packs and cells and know what's inside. And, it's well publicized that only Panasonic cells were in Model 3s for ages. If picking and choosing cells were so easy, and cells are really commodities, then why would Tesla bind themselves to a single cell manufacturer? Answer: They wouldn't. Cells are not generally interchangeable, not by a long shot.

That being said, I would not go so far to say the cells are "custom made". The packs, sure. But cells are close variants of the same technology and process, and many smaller and lower customers may just use the same cells from a single manufacturer.

You are incorrect about batteries not needing high tech and high cost manufacturing plants. In fact, the requirements around cleanliness and process control are very high, for yield, cost and safety reasons. Capital equipment costs and knowledge base requirements are high. I'm not sure how they compare to the requirements around leading edge microchips, but I would imagine that they would want to use similar approaches.

As far as shipping goes, cells are shipped around via air freight all the time. They are going to need to get shipped regardless at some point. And, to the contrary, given my experiences with ground freight in general, I would not want to send cells via ground services.