Tesla’s “Revolutionary” 4680 Has A Massive Problem

Summary

Tesla's 4680 battery, despite initial promises of advancements, has underdelivered with current iterations showing lower energy density and reliance on traditional, less eco-friendly materials compared to the previous 2170 cells.

Abstract

The 4680 battery, announced by Elon Musk in 2020 as a revolutionary leap in battery technology, was touted to be cheaper, faster charging, and more energy-dense. However, two years post-announcement, the battery has not met its promised specifications. Sandy Munro's teardown analysis reveals the 4680 cells have a lower energy density of 244 Wh/kg, falling short of the expected 296 Wh/kg and the 2170 cells' 269 Wh/kg. The new cells also have a high cobalt content and lack the innovative features like dry-coated electrodes and a silicon anode that were initially advertised. While the 4680 packs are cheaper to manufacture and offer structural benefits, their lower energy density and non-serviceable design could make them less cost-effective and eco-friendly in the long run compared to the older 2170 packs. Competitors like CATL and BYD are advancing with more sustainable and cost-effective battery technologies, putting pressure on Tesla to either improve the 4680 or adopt these third-party solutions to maintain its lead in the electric vehicle market.

Opinions

The current production of Tesla's 4680 battery is considered a setback, as it does not incorporate the innovative design features promised, such as a tabless design or zero cobalt content.
Sandy Munro's recent analysis suggests that the 4680 battery falls short of expectations in energy density and environmental friendliness due to high cobalt concentration and graphite anode usage.
The production and use of the 4680 battery highlight Tesla's potential weakness in developing and scaling battery technology, in contrast to competitors like CATL and BYD which are making strides with more advanced and eco-friendly batteries.
The 4680 battery's lower energy density necessitates a larger number of cells for the same capacity, which, coupled with the non-replaceable cell design, may result in higher long-term costs and environmental impact for Tesla vehicles.
The article's author suggests that Tesla should consider abandoning the underperforming 4680 project and instead adopt superior third-party batteries to remain competitive in the EV market, focusing on their strength in EV manufacturing rather than battery development.

As it turns out, the 4680 is worse than Tesla’s old battery.

In September 2020, Musk announced Tesla’s next battery, the 4680. This gigantic cylindrical cell would be far cheaper, faster at charging, and way more energy dense than anything they had developed before. Two years on, and Tesla is now producing and installing 4680 batteries into their production cars, but the next-generation battery that Musk promised has failed to materialise. What’s worse, a recent teardown by the now-famous Sandy Munro shows the current 4680s are actually worse than Tesla’s old 2170 cells. So, what’s going on? And can Musk remedy this situation?

Let’s recap what the 4680 was promised to be and what Tesla is actually making today (a topic I already covered here). Musk originally declared the 4680 would have some breakthrough design features such as a tabless design, dry-coated electrodes, zero cobalt content (the mining of which has a significant environmental impact), a silicone anode, and a structural battery pack. What’s more, the much larger form factor would mean each pack would have far fewer individual cells, making them less expensive to assemble. All of these features combined would create a battery that is cheaper and more powerful than anything else on the market.

Some Model Ys are currently being built with a structural 4680 battery pack. This is where the cells are set in a resin-like material, allowing them to become a load-bearing member of the vehicle’s chassis. This saves time and money by simplifying and speeding up the battery pack assembly process. In fact, the current 4680 packs are $3,600 cheaper to manufacture than the old, $11,000 2170 packs.

But that’s about it. Teardowns of Tesla 4680 batteries a few months ago showed that the current 4680 cells have basically none of the cutting-edge technologies Musk promised (read more here). The only aspects that have made it into production are the tabless design, its larger form factor, and its structural battery. So rather than being a breakthrough battery, it was really just a larger version of their old cells with a new way of packaging them. Or at least, that is what we thought.

A more recent teardown by Sandy Munro (a company that specialises in teardown engineering analysis) has revealed something more worrying. You see, they found that the 4680 cells had a 100% graphite anode, a high cobalt concentration, and an energy density of only 244 Wh/kg.

This woeful energy density isn’t just miles below the target energy density of 296 Wh/kg but is actually worse than Tesla’s previous Panasonic-made 2170s, which achieved 269 Wh/kg. This old cell managed this feat thanks to higher nickel concentrations, lower cobalt concentrations, and a 10% silicone anode. So why has Musk not used this already-proven technology? Well, the price of nickel has been sky-high (up until very recently), thanks in part to Russia’s invasion of Ukraine, and Musk has been struggling to ramp up 4680 production to meet demand. As such, it seems he has chosen the cheaper and easier-to-produce battery chemistry to get the cost of each cell down, even if this means sacrificing energy density and using lots of environmentally harmful cobalt.

But this lower energy density has a knock-on effect. You see, each 4680 pack has 830 cells, and each cell weighs 355g. That means it has a pack weight (only of the individual cells and not the casing) of around 294 kg, and by using Sandy Munro’s energy density, we can calculate that this gives an overall pack capacity of just north of 71 kWh. As we know, each pack costs $7,400 to produce, which means we get a cost per kWh of $104.

Now, this is better than the old 2170, which cost around $130 to $140 per kWh, but it is still miles off Tesla’s goal of $60 per kWh. Recent teardowns have also shown that you can’t replace individual cells in the 4680 structural pack, as removing them from the resin can damage the cells. Therefore, even if only one cell goes wrong, you need to replace the entire pack! Sadly, this means that, in the long run, the 2170-pack vehicles might actually be far cheaper and more eco-friendly to maintain.

However, there are cells on the market that are far less expensive than the 4680 and can replace individual cells while also being extremely eco-friendly.

Take CATL; they are already providing cobalt-free LFP cells (an alternative to lithium-ion cells) to Tesla for the Model 3 SR at a price of around $80 per kWh! These cells aren’t as powerful, but Model 3s can charge just as quickly as other Teslas and have a smaller range due to the smaller battery pack.

Furthermore, waiting in the wings are CATL’s own sodium-ion battery (read more here) and BYD’s Blade Battery (read more here). Over the next year or two, both of these batteries will crush the 4680’s current cost and go on to eclipse Tesla’s $60 per kWh target. In fact, both could get below $50 per kWh! What’s more, thanks to their unique battery chemistry, these batteries are far safer, way faster at charging, and much better for the environment than Tesla’s current 4680 (and possibly even the 4680 Musk initially promised).

So Tesla’s “revolutionary battery pack” is far from what Musk indicated. In fact, it seems like this project is going nowhere fast. This is a huge problem for Tesla because, let’s not forget, the Cybertruck, Roadster, and Semi all need the 4680 to make it into full-scale production and achieve their stated specifications, range, and performance. I also look forward to seeing the right-to-repair issues that surround the 4680’s unrepairable battery pack…

Musk will have to act quickly to either restart proper 4680 production (with all the promised technologies and at a low price) or abandon this failing project and use a third-party cell. If I were him, I would go for the latter. The 4680 shows that while Tesla can build a good EV, they suck at developing and scaling battery production. There are companies like BYD and CATL that are thrashing Tesla in this area. So, if Tesla wants to stay on top of the EV world, it needs to stop fiddling with its 4680 and instead become the first to adopt these breakthrough third-party cells that are already ahead of the 4680, allowing Tesla to focus on what it is good at: building EVs. Otherwise, other manufacturers will use these new cells, and they will leave Tesla and its 4680 pet project in the dust.