Hi Christopher, Many thanks for the note, always to the point !. It seems there is another level of risk for nascent GF's you have not expended too much on. Newly established Gigafactories in Europe have signed off-take agreements that seems to be providing poor certainty or guarantee overall. Apart from VW and Powerco example , OEM are still shifting and navigating their technology roadmaps on chemistries and formats, in an uncertain market they are still trying to understand. GF's often have a single chemistry they try to master at scale, in one cell format (imposed by customer) they are trying to produce. How are both GF's and OEMs confident this is the right product ? And for how long...?
You're write. Gigafactories are locking in a fixed product definition at a time when chemistries, formats, and end markets are still moving. That means billions are committed to one path with limited ability to pivot.
Same for OEMs. They pick what looks right today based on cost curves, vehicle plans, and regulations.
OEMs can keep options open. A gigafactory cannot. It has to qualify, tool, and stabilize one chemistry and one format. If the market shifts toward another chemistry, different form factors, or more ESS demand, that risk lands squarely on the plant.
That is why the real risk is not only execution. It is inflexibility. As more applications emerge, a gigafactory can usually place volume if the product matches a real use case. Right now, LFP or NMC in prismatic format is the lowest-risk starting point because it serves both EVs and ESS.
Over time, flexibility should increase as the operator gains process knowledge and customer diversity. We already saw that in the US, where LGES was able to convert EV lines to ESS when storage demand took off. That kind of optionality only comes after you survive the first years of production.
Group14 has real customer engagements and a growing materials footprint, but it plays in a competitive, crowded silicon-anode landscape. Its SCC55 silicon-carbon composite has been shipped to 100+ battery makers and suppliers.
However, Chinese and other Asian producers are scaling up capacity far more overall.
Strategically, the company secured a significant $463 million Series D round led by SK and others, and took full ownership of its South Korean BAM production facility.
On the execution side, there have been real headaches: their Moses Lake BAM-2 plant has been pushed back by more than a year from original targets, and layoffs in Washington reflect operational and policy headwinds in the U.S. clean tech environment.
Their European “silent” plant in Germany is a silane precursor facility to secure a key raw input for silicon material production. That supports materials supply and reduces logistical risk. But, It's unclear if they will build a silicon anode materials plant in Europe.
Regarding batteries used for energy storage, with inverters made in China, I noticed this warning from Strider Intel company:
“America’s clean energy expansion has created a growing threat to the stability of the power grid. As the United States accelerates its transition to renewable energy, it has become heavily dependent on inverter-based resources (IBRs), including photovoltaic power (commonly referred to as solar power) and battery energy storage systems (BESS), manufactured by companies in the People’s Republic of China (PRC). These systems are no longer simple hardware; they are networked, software-driven devices capable of remote communication and control. This connectivity, combined with their PRC origin, exposes U.S. critical infrastructure to unprecedented risk.” emphasis mine
The concern is valid, but it sits squarely at the system level. In BESS, the risk is not the battery cells, it’s the software and control layer linking the BMS, inverter, and grid interface, especially when remote access and firmware visibility are weak. Undocumented communication paths or poorly governed updates can, in theory, affect dispatch behavior and grid response. That said, there is no public evidence of a successful grid-level attack via BESS, and the real fix is tighter cybersecurity, clear control boundaries, and audited software.
Difficult to name a single “best” inverter maker, but Tesla and Fluence are the strongest U.S. players at the BESS system level.
Tesla designs and integrates its own inverter and control stack within Megapack and Powerwall, which gives it tight software governance and helps explain its leadership in large-scale BESS deployments. Fluence is a leading grid-scale system integrator, combining PCS, controls, and services, with deep utility and grid-operator experience.
Neither is a standalone inverter OEM on the scale of the top Chinese suppliers, but in the U.S. market, they compete on integration quality, bankability, and software control rather than on pure inverter volume.
Exactly. Capital can buy equipment and time, but it cannot substitute for a proven operating system. When yields deteriorate rather than improve, it signals that fundamentals were never stable to begin with.
That’s the core difference. Scaling experience is about variance reduction. First-time plants are still discovering failure modes at full cost. At that point, learning becomes very expensive very quickly.
And once production starts, scrap becomes the only honest metric.
Hi Christopher, Many thanks for the note, always to the point !. It seems there is another level of risk for nascent GF's you have not expended too much on. Newly established Gigafactories in Europe have signed off-take agreements that seems to be providing poor certainty or guarantee overall. Apart from VW and Powerco example , OEM are still shifting and navigating their technology roadmaps on chemistries and formats, in an uncertain market they are still trying to understand. GF's often have a single chemistry they try to master at scale, in one cell format (imposed by customer) they are trying to produce. How are both GF's and OEMs confident this is the right product ? And for how long...?
You're write. Gigafactories are locking in a fixed product definition at a time when chemistries, formats, and end markets are still moving. That means billions are committed to one path with limited ability to pivot.
Same for OEMs. They pick what looks right today based on cost curves, vehicle plans, and regulations.
OEMs can keep options open. A gigafactory cannot. It has to qualify, tool, and stabilize one chemistry and one format. If the market shifts toward another chemistry, different form factors, or more ESS demand, that risk lands squarely on the plant.
That is why the real risk is not only execution. It is inflexibility. As more applications emerge, a gigafactory can usually place volume if the product matches a real use case. Right now, LFP or NMC in prismatic format is the lowest-risk starting point because it serves both EVs and ESS.
Over time, flexibility should increase as the operator gains process knowledge and customer diversity. We already saw that in the US, where LGES was able to convert EV lines to ESS when storage demand took off. That kind of optionality only comes after you survive the first years of production.
What about Group14.technologies in the US and a "silent" plant now in Germany?
Group14 has real customer engagements and a growing materials footprint, but it plays in a competitive, crowded silicon-anode landscape. Its SCC55 silicon-carbon composite has been shipped to 100+ battery makers and suppliers.
However, Chinese and other Asian producers are scaling up capacity far more overall.
Strategically, the company secured a significant $463 million Series D round led by SK and others, and took full ownership of its South Korean BAM production facility.
On the execution side, there have been real headaches: their Moses Lake BAM-2 plant has been pushed back by more than a year from original targets, and layoffs in Washington reflect operational and policy headwinds in the U.S. clean tech environment.
Their European “silent” plant in Germany is a silane precursor facility to secure a key raw input for silicon material production. That supports materials supply and reduces logistical risk. But, It's unclear if they will build a silicon anode materials plant in Europe.
Regarding batteries used for energy storage, with inverters made in China, I noticed this warning from Strider Intel company:
“America’s clean energy expansion has created a growing threat to the stability of the power grid. As the United States accelerates its transition to renewable energy, it has become heavily dependent on inverter-based resources (IBRs), including photovoltaic power (commonly referred to as solar power) and battery energy storage systems (BESS), manufactured by companies in the People’s Republic of China (PRC). These systems are no longer simple hardware; they are networked, software-driven devices capable of remote communication and control. This connectivity, combined with their PRC origin, exposes U.S. critical infrastructure to unprecedented risk.” emphasis mine
What do you think of this threat?
The concern is valid, but it sits squarely at the system level. In BESS, the risk is not the battery cells, it’s the software and control layer linking the BMS, inverter, and grid interface, especially when remote access and firmware visibility are weak. Undocumented communication paths or poorly governed updates can, in theory, affect dispatch behavior and grid response. That said, there is no public evidence of a successful grid-level attack via BESS, and the real fix is tighter cybersecurity, clear control boundaries, and audited software.
Thanks. Do you have an opinion on which US company is the best maker of inverters, and are they competitive with Chinese?
Difficult to name a single “best” inverter maker, but Tesla and Fluence are the strongest U.S. players at the BESS system level.
Tesla designs and integrates its own inverter and control stack within Megapack and Powerwall, which gives it tight software governance and helps explain its leadership in large-scale BESS deployments. Fluence is a leading grid-scale system integrator, combining PCS, controls, and services, with deep utility and grid-operator experience.
Neither is a standalone inverter OEM on the scale of the top Chinese suppliers, but in the U.S. market, they compete on integration quality, bankability, and software control rather than on pure inverter volume.
Thanks. I wasn't familiar with Fluence.
Exactly. Capital can buy equipment and time, but it cannot substitute for a proven operating system. When yields deteriorate rather than improve, it signals that fundamentals were never stable to begin with.
That’s the core difference. Scaling experience is about variance reduction. First-time plants are still discovering failure modes at full cost. At that point, learning becomes very expensive very quickly.
And once production starts, scrap becomes the only honest metric.