Late Capacity Is Worthless Capacity: The New Economics of Gigafactory Delivery

Executive summary: The battery cell market has tipped into global overcapacity — installed capacity passed 2,500 GWh in 2025 against demand of roughly 1.6 TWh — yet gigafactories still miss their start-of-production dates by six to nine months on average. At a single 40 GWh plant, a six-month slip can put roughly $250 million of value at risk. The winners of the next decade will be decided less by who announces capacity and more by who ramps it on time.

Gigafactory delivery delays are now the defining execution risk in the battery industry: the average plant misses start of production (SOP) by six to nine months, ramps below planned yield, and does so in a market where oversupply means late output may never find the buyer its business case assumed.

Overcapacity has inverted the risk

For most of the past decade, the constraint in batteries was supply. A late factory was forgiven, because demand queued up and whatever came off the line eventually sold. That world is gone. Global lithium-ion production capacity exceeded 2,500 GWh in 2025, while demand sits around 1.6 TWh — a market running with far more nameplate capacity than orders. Cell prices reflected it: lithium-ion prices fell roughly 50% in a single year as subsidized Chinese production flooded the market, where producers hold more than 75% of global capacity and CATL alone accounted for 37% of global sales in 2024.

Oversupply breaks the old forgiveness. A plant that opens late now enters a market where buyers already have alternatives — so the cost of delay is not deferred revenue but potentially forfeited revenue: offtake volumes at risk, pricing power gone, and fixed costs plus financing accruing against a line that isn’t producing. With gigafactory capital intensity running at $50–100 million per GWh, a 40 GWh plant carries billions in committed spend before the first sellable cell.

The delay epidemic, in numbers

The slippage is not anecdotal. Industry analysis of gigafactory programs finds these $1 billion–$7 billion facilities missing SOP by six to nine months on average, then ramping more slowly and at lower yields than planned — with a six-month delay at a 40 GWh reference plant putting around $250 million of value at risk.

The casualty list shows what happens when delay meets oversupply. Northvolt — once Europe’s flagship cell hope — entered bankruptcy with production halted across its facilities. ACC paused 80 GWh of planned capacity in Germany and Italy; Svolt and Farasis cancelled a further 40 GWh of German projects; Britishvolt and Italvolt went under before reaching scale. Benchmark counted 220 GWh of the 2030 pipeline cancelled or postponed as early as late 2024, including 158 GWh of delayed European projects — enough capacity for roughly two million EVs.

The United States followed the same script. KORE Power cancelled its 12 GWh Arizona plant in January 2025 after its $850 million DOE loan commitment failed to materialize; FREYR scrapped its 34 GWh Georgia facility; iM3NY filed for Chapter 11; and GM–Samsung SDI pushed 27–36 GWh of capacity out to 2027, citing slowing EV sales.

The pattern across every one of these: the projects that died were not the ones with the worst technology. They were the ones that ran out of time.

The execution spread

Here is the fact that should reframe the problem: the best operators are delivering gigafactories in 18–24 months from start of construction, while the average project slips six to nine months past its own plan. Call that gap the execution spread — and in an oversupplied market it is now one of the widest sources of competitive advantage in the sector, worth more than most technology bets. What separates the leaders comes down to three moves.

Put the C-suite on the hook. Leaders run a dedicated delivery organization with a direct line to top management, defined escalation pathways, and continuous improvement embedded in the program. The failing pattern is the opposite: delivery delegated down the organization as a project-management task, surfacing to leadership only when the date is already lost.

Contract for certainty, not just for equipment. Best-in-class programs use risk-sharing contracts with binding milestones, so supplier incentives are tied to SOP rather than to shipment dates. They run a single unified planning tool across all stakeholders — one version of the schedule instead of a dozen contractors’ spreadsheets — and apply AI-driven generative scheduling that evaluates hundreds of thousands of schedule permutations to optimize cost and SOP together.

Design for the ramp, not the ribbon-cutting. Leaders design the factory layout for total cost of ownership from day one, run permitting proactively instead of sequentially, and manage supplier interfaces rigorously — because the seams between equipment vendors are where integrated production lines most often lose time and yield.

Why the ramp is the real product

A subtlety the headline delay numbers hide: the slip compounds. A factory that starts six months late and then climbs a slower-than-planned yield curve is not six months behind — its effective capacity gap keeps widening for quarters after SOP, while depreciation, interest, and staffing costs run at full rate. In a scarce market, that gap was an inconvenience. In an oversupplied one, it is a transfer of contracted volume to whichever competitor has qualified cells ready. The building is a milestone; the ramp is the product.

This is why consolidation is now the consensus expectation — industry voices see a shake-out concentrating around 2027–2028, with the top six suppliers already controlling more than 65% of global sales. The execution spread is what sorts the survivors from the consolidated.

What the averages hide

Two honest limits. First, the six-to-nine-month figure is an average across very different projects; a specific plant’s value at risk depends on its offtake structure, financing terms, and technology maturity, and can sit well above or below the $250 million marker. Second, public data on post-SOP yield ramps is thin — the compounding effect described above is directionally solid but hard to quantify from outside. Whether newer chemistries and first-time geographies systematically underperform the average remains an open question the data does not yet settle.

Six questions before the next capacity decision

For any executive sponsoring, financing, or partnering on cell capacity, six questions expose most of the risk. Who owns delivery — and do they report to the C-suite or three layers down? Are supplier contracts milestone-bound with shared risk, or pay-on-shipment? Is there one planning environment across every stakeholder, or many? Has the schedule been stress-tested across alternatives, or is it a single Gantt chart defended by its author? Has the per-month cost of slippage been quantified before approval, not after the miss? And is the yield-ramp plan resourced as seriously as the construction plan?

The bottom line

The buildout will not stop: battery demand is still headed toward roughly 6.8 TWh by 2035, from about 1.6 TWh today, and global capacity is forecast to approach 5.4 TWh by 2036 even after the pipeline cooling. But the market has already delivered its verdict on late capacity — in bankruptcies, pauses, and 220 GWh of cancelled plans. The same logic runs upstream, too: schedule certainty on a billion-dollar build depends partly on seeing input-cost moves before they land, the discipline platforms like SteelMath bring to materials-price foresight. Gigafactory delivery is no longer a project-management problem to delegate down the organization. It is a P&L capability, and it belongs at the top of the house.

Frequently asked questions

Why do gigafactories miss their start-of-production dates?

The root causes are structural: fragmented planning across dozens of contractors and equipment vendors, contracts that pay suppliers on shipment rather than on milestones, sequential permitting, and delivery organizations buried too deep to escalate problems before dates are lost. Average slippage runs six to nine months.

How much does a gigafactory delay cost?

At a 40 GWh reference plant, a six-month SOP delay can put roughly $250 million of value at risk — from accruing fixed costs and financing, lost pricing power, and offtake volumes forfeited to competitors with cells ready. Actual exposure varies with contract structure and financing terms.

Is there really overcapacity in the battery cell market?

Yes. Global production capacity passed 2,500 GWh in 2025 against demand of roughly 1.6 TWh, cell prices fell about 50% in a year, and more than 220 GWh of planned 2030 capacity has been cancelled or postponed — including high-profile failures like Northvolt and Britishvolt.

How fast can a gigafactory actually be built?

Best-in-class operators deliver gigafactories in 18–24 months from start of construction — while average projects slip six to nine months past plan. The gap comes from C-suite-owned delivery organizations, milestone-bound risk-sharing contracts, unified planning tools, AI-driven scheduling, and factory design optimized for total cost of ownership from day one.

Will the battery market consolidate?

Industry expectations point to consolidation around 2027–2028. The top six suppliers already account for more than 65% of global sales, and sustained oversupply punishes late, sub-scale entrants hardest.