Long Way Home: Battery Costs, Chemistries & Sourcing Strategies - Lessons from Tesla


Well-Known Member
First Name
Feb 17, 2020
Reaction score
Santa Cruz, Ca./Odawara, Jpn
First Name
1981 Volvo wagon; 2006 Tacoma SR5; 2009 Toyota Prius
Here's an insightful piece on battery costs, chemistries and sourcing strategies based on research about Tesla's experience, especially at GigaFactory I.


At the risk of missing or misunderstanding some important points, here's a summary.
1. It's been widely said if battery costs reached $100/kWh, the "holy grail" of battery production, BEV and ICE vehicle production costs would be the same.
2. The current costs per kWh for a typical Tesla battery chemistry pack are estimated to be $167-$187/kWh, based on a 1GWh production capacity.
3. Cairns Energy Advisors estimated 2019 Tesla pack-level battery costs at $158/kWh (based on GigaFactory I).
4. Estimates are based on Tesla's GigaFactory I output at Nevada, thought to be operating at 25 GWh capacity though it was designed to operate at 35 GWh.

There's a long way to go before Tesla, the leading BEV vehicle maker, will hit the holy grail of $100/kWh. Four possible ways of doing so are mentioned.
1. Increasing economies of scale. GigaFactory I is huge, one of the largest buildings in N. America, and it's only running at 70% capacity. There have been a host of issues in getting production to this level, and it's unclear what issues might need to be resolved in order to push capacity utilization higher.
2. Changes in battery chemistry. Many different battery chemistries are being research and developed, but none of them are sufficiently far along to make large scale production cost estimates at this time. Tesla's Shanghai GigaFactory supplier, CATL, has just announced a "million mile battery", developed in association with Tesla, which is about 10% more expensive to produce than current lithium-ion batteries but which lasts ten times longer.
3. Changing relations with battery suppliers. There's a long chain of supply in battery production, beginning with mining, a lot of intermediate processing, and ending with downstream battery and battery pack suppliers, like CATL, LG Chem and Panasonic. Tesla works closely but independently of its battery suppliers. In GigaFactory I, Tesla worked very closely with Panasonic and the relationship approached an almost vertical integration-like arrangement. Conspicuously, Tesla chose a different battery supplier for the Shanghai GigaFactory and, in the business press, a fair amount of friction was reported between Tesla and Panasonic.

Tesla has a history of preferring vertical integration which is generally eschewed in the auto industry, particularly for highly skilled areas of supply where auto companies lack expertise. But vertical integration, if handled well, should lower costs. It will be interesting to see where Tesla comes down on the vertical integration vs outsourcing continuum with respect to battery supply, given that there are many intermediate steps in the value adding chain between mining and battery pack production.
4. Recycling and recovery. A way to lower battery costs might be to recover and recycle materials from used battery packs. Given the long value adding chain between mining and battery pack production, costs could conceivably come down greatly if used battery materials could be efficiently harvested and reused.

Lessons for Rivian. These are my own thoughts.
1. Increasing economies of scale isn't an option for Rivian, as it is for Tesla. Rivian has no GigaFactories, nor should it want one. Rivian should concentrate its efforts on getting Normal to operate at full capacity. Once production begins, it will be a year and more likely two before minimum efficient scale is realized.
2. Changes in battery chemistry. If I may, I'd pair changing battery chemistries with battery supplier relations. With the ferment in battery research, development and engineering, it would be best if Rivian can partner with a number of different battery suppliers. Different suppliers are pursuing different battery chemistries. This complicates supply and it will likely result in battery cost (and performance?) differences between suppliers, but it's way too early in the battery R&D game to put all your eggs in one supplier's basket.
3. RJ has talked about reusing Rivian batteries, once they're out of service, for various land based uses. For now, that's fine. If it becomes possible to harvest, recover and recycle batteries for vehicular use, I assume Rivian will do so.

I apologize for the long post. Battery costs are a key issue not only for the pricing and performance of our vehicles, but also for the acceptance of BEVs by the general public. If Tesla's experience can be generalized, battery costs are still the most expensive and, therefore, most critical component cost for BEV production. I expect Rivian's battery costs to come down as a function of working with suppliers and production experience at Normal, but it's likely to be years before Rivian R1S and R1Ts will cost appreciably less, based on the cost of their batteries.
Last edited: