The Count Gap
E2's 2026 tally: 72 projects cancelled, closed, or downsized, $26.52B in lost investment, manufacturing counted through June and generation and storage through May (E2, July 20, 2026). Fourteen of the 72 are manufacturing, roughly $7B.
That is a serviceable policy signal and the wrong denominator for anyone modeling when domestic cells actually arrive.
E2's published methodology excludes delays and idlings "unless there is an announced decrease in production or investment." Samsung SDI and GM's New Carlisle, Indiana JV (27 GWh, construction paused 2026-05-14, no revised SOP disclosed since, no announced investment reduction) therefore sits outside the 72 on my read, while removing more modelable near-term cell capacity than several projects inside it.
What a stoppage subtracts depends on where the project stood when it stopped. A greenfield abandoned before groundbreaking deletes announced GWh and no available cells. Liquidation at 60% completion strands the steel and puts the qualification clock back near zero. Conversion from EV to storage keeps the building, voids the original supply claim. A mass-production slip inside a financed, running plant moves a date without changing capacity.
All four collapse into the same number. The pieces that follow take them apart.
Apex 1 Hopkinsville — The Divergence Dates to 2024-11-11
Kentucky announced $310M and 250 jobs. DOE reimbursed roughly $206M. The plant reached about 60% complete, never produced a saleable component, and after an April 2026 Chapter 11 filing went to its own general contractor against money already owed.
The ledger below fixes the divergence to a date. On 2024-11-11, layoffs and contractor renegotiation were publicly audible while summer 2025 still stood as the affirmed completion date. The signal that led everything else was a $138.4M suit in Christian Circuit Court, on file five weeks before the trade press noticed it. Monitor the docket, not the press release.
Apex 1 Hopkinsville — The Divergence Dates to 2024-11-11
Kentucky announced $310M and 250 jobs. DOE reimbursed roughly $206M. The plant reached about 60% complete, never produced a saleable component, and after an April 2026 Chapter 11 filing went to its own general contractor against money already owed.
The ledger below fixes the divergence to a date. On 2024-11-11, layoffs and contractor renegotiation were publicly audible while summer 2025 still stood as the affirmed completion date. The signal that led everything else was a $138.4M suit in Christian Circuit Court, on file five weeks before the trade press noticed it. Monitor the docket, not the press release.
Four Status Values, Six Transition Modes: A Survival Taxonomy for the Battery Registry
Every major US battery project tracker runs on four status values. Hand one a liquidation with a half-built shell, an indefinite pause on a finished building, a change of control on a line that never stopped shipping, and a chemistry swap on a running plant. They come back identical, or they come back blank.
Six distinct transition modes, anchored to seven named facilities, scored asset class by asset class: site control, permits, installed equipment, workforce, customer qualification. Physical assets carry well. In the qualification column the count of documented survivals is zero.
Four Status Values, Six Transition Modes: A Survival Taxonomy for the Battery Registry
Every major US battery project tracker runs on four status values. Hand one a liquidation with a half-built shell, an indefinite pause on a finished building, a change of control on a line that never stopped shipping, and a chemistry swap on a running plant. They come back identical, or they come back blank.
Six distinct transition modes, anchored to seven named facilities, scored asset class by asset class: site control, permits, installed equipment, workforce, customer qualification. Physical assets carry well. In the qualification column the count of documented survivals is zero.
Four Instruments, Four Survival Rules: Public Money Through Three Battery-Plant Transitions
USD 3.805 billion of federal loan principal sits advanced against two Kentucky plants, and it is verifiable to the dollar, for the mundane reason that Ford must report its own debt. Who claims the 45X credits at a Tennessee cell plant is not verifiable at all. Four subsidy instruments, four different rules about what survives a bankruptcy, a loan assumption, or a control transfer, and a second variable running independently of all four: whether anyone outside can establish it. Three transitions traced from commitment to disbursement to current status, with the dark parts named as dark.
Four Instruments, Four Survival Rules: Public Money Through Three Battery-Plant Transitions
USD 3.805 billion of federal loan principal sits advanced against two Kentucky plants, and it is verifiable to the dollar, for the mundane reason that Ford must report its own debt. Who claims the 45X credits at a Tennessee cell plant is not verifiable at all. Four subsidy instruments, four different rules about what survives a bankruptcy, a loan assumption, or a control transfer, and a second variable running independently of all four: whether anyone outside can establish it. Three transitions traced from commitment to disbursement to current status, with the dark parts named as dark.
External Sources
