Northvolt AB

Overview

Northvolt was Europe's most heavily backed battery startup and the continent's best-funded attempt to build an alternative to Asian cell manufacturers. Peter Carlsson and Paolo Cerruti, two former Tesla supply chain executives, founded it in 2016 as SGF Energy and renamed it Northvolt in 2017. The thesis was straightforward: Europe's automakers would need tens of gigawatt-hours of locally produced cells to electrify their fleets, and no European company existed to supply them. Northvolt would build a vertically integrated battery industry from raw material processing through cell production, powered by Scandinavian hydroelectricity and sited near northern Sweden's raw materials. Over its life it raised approximately $13 billion in equity and debt, roughly $15 billion counting government support, from investors including Volkswagen, Goldman Sachs funds, BMW, Scania, Swedish pension funds, and Canadian institutions including the Quebec government and CDPQ. Volkswagen alone invested about €1.4 billion for a stake of about 21% and placed a $14 billion battery supply order in 2021. On November 21, 2024, Northvolt filed for Chapter 11 in the Southern District of Texas, and Carlsson resigned the next day, staying on as a senior advisor and board member.

The immediate cause was production failure. Northvolt Ett, the flagship gigafactory in Skellefteå, had a nameplate capacity of 16 GWh a year. In all of 2023 it produced less than 1 GWh. In Q3 2023 it delivered 13 MWh for the entire quarter, roughly one three-hundredth of the annualized capacity of the building it occupied. BMW cancelled its €2 billion supply contract in June 2024 after Northvolt fell two years behind schedule and delivered cells that did not consistently meet specification. Volkswagen decided in autumn 2024 not to make a planned additional equity investment. The company laid off 1,600 employees in September 2024, about a fifth of its global workforce and a quarter of its Swedish staff. By November it had approximately $30 million in cash.

After the Chapter 11 filing, asset disposals proceeded across multiple jurisdictions. The Swedish parent filed for bankruptcy in Sweden in March 2025. Northvolt Batteries North America, the entity behind the planned Quebec gigafactory, was declared insolvent by a Quebec court in September 2025; the province lost its entire C$270 million equity investment and recovered about C$200 million on a separate C$240 million land loan. The buyer of the manufacturing assets was Lyten, a San Jose company developing lithium-sulfur batteries using 3D graphene, which took the estate in stages: Cuberg's San Leandro facility in November 2024, the Northvolt Dwa energy storage plant in Gdańsk agreed in July 2025 and closed that October, and, under an August 2025 binding agreement, Northvolt Ett with its expansion site, Northvolt Labs in Västerås, the intellectual property, and more than 160 hectares of land, closing in February 2026. Lyten also agreed in March 2026 to buy the Revolt recycling plant. The Northvolt Drei site in Heide, Germany has not transferred: Lyten signed an exclusive memorandum of understanding on July 1, 2026 at about €60 million agreed with KfW, the German federal government, and Schleswig-Holstein, with a definitive agreement targeted for the third quarter of 2026.

Operations

The production ramp that never happened

Northvolt Ett was supposed to produce 16 GWh of cells a year by 2024. In Q3 2023 it delivered 13 MWh for the entire quarter. That ratio, 13 MWh against roughly 4,000 MWh of quarterly capacity, captures the failure more precisely than any financial figure. The company had raised billions, hired thousands, built the factory, and connected it to the Swedish grid. It could not make the cells at volume or at specification.

Battery cell manufacturing is a process industry where yield, consistency, and throughput are inseparable. A gigafactory is not a building that produces cells; it is a production system where dozens of interdependent steps must all run at high yield at once for the economics to work. Electrode coating, calendering, cell assembly, electrolyte filling, formation cycling, aging, and inspection each carry their own defect rates, and those rates multiply through the chain. Ten consecutive steps at 95% yield give 60% end-to-end. At 90% per step it is 35%. Reaching the yields automotive OEMs require at mass-production volumes takes years of iterative process learning, which Asian manufacturers accumulated across thousands of production-months in smaller facilities before scaling.

Northvolt tried to build that knowledge at gigafactory scale simultaneously, with no prior history of volume cell production. CATLand Samsung SDI had decades of cell-making experience from consumer electronics before entering automotive. LG Energy Solution had been manufacturing cylindrical cells since the 1990s. Northvolt had talented engineers, well-funded R&D, and genuine technical capability, and it was attempting to master a process-intensive manufacturing discipline in a 500,000-square-meter facility with no production track record, enormous delivery commitments creating pressure from day one, and investors expecting a timeline the underlying process complexity could not support. Carlsson acknowledged to the Financial Times in 2023 that Northvolt needed to prove it could match Asian suppliers on manufacturing execution, a candid admission that the problem was real and unresolved two years after the factory began producing cells.

BMW's June 2024 cancellation was the public acknowledgment that the problem had become fatal. The €2 billion order, signed in July 2020, had been placed expecting a ramp competitive with established Asian suppliers. The ramp slipped two years, and the cells produced did not consistently meet specification. BMW shifted the order to Samsung SDI. Volkswagen, reviewing the same production data, decided in autumn 2024 against its planned additional equity investment; the Swedish administrator's report later identified that decision as the triggering factor. Three workers connected to the Skellefteå facility died in 2024, all outside working hours at their homes, generating additional scrutiny; Swedish police closed the investigation in December 2024, finding no grounds to suspect a crime and no indication of exposure to hazardous substances, though causes of death were not established. By November, Northvolt had approximately $30 million in cash against a capital structure requiring hundreds of millions to sustain operations.

The broader lesson is about the distance between industrial policy aspiration and manufacturing execution. Europe had a legitimate strategic interest in domestic battery production independent of Asian supply chains, and Northvolt attracted billions from VW, Goldman, pension funds, and governments partly because that geopolitical case was compelling. Geopolitical logic does not substitute for the accumulated process knowledge that makes a gigafactory actually produce cells to specification. Europe's next attempt at battery sovereignty will need either a technology partnership with manufacturers who hold that knowledge, or a substantially longer tolerance for ramp timelines than Northvolt's customers and investors were willing to provide.

Financial history

Northvolt raised approximately $13 billion in equity and debt over its lifetime. The investor roster included Volkswagen Group, with about €1.4 billion of equity across 2019 for a stake of about 21% plus the $14 billion supply order in 2021, Goldman Sachs funds, BMW Group, Scania, which also provided the $100 million debtor-in-possession financing at bankruptcy, Folksam, AMF, Swedish and Canadian pension funds, and Spotify founder Daniel Ek. A $2.75 billion equity round in June 2021 was the largest single raise. A January 2024 green lending facility of $5 billion, non-recourse project financing from 23 commercial banks plus the European Investment Bank and the Nordic Investment Bank, was the final major infusion. Despite the capital base, Northvolt burned cash at a rate the production volumes could not support: each cell carried an extremely high unit cost because factory overhead and depreciation spread across a fraction of planned output. The company never disclosed full annual results publicly, and its output and the sequence of order cancellations made clear the model was not approaching viability.

At the Chapter 11 filing, Northvolt reported $1 billion to $10 billion in both assets and liabilities on the bankruptcy schedule and had approximately $30 million in available cash. The $100 million Scania facility provided runway through the proceedings. The Swedish bankruptcy that followed in March 2025 was the largest in modern Swedish history, with debts close to SEK 80 billion and about 5,000 remaining employees. Equity investors were wiped out: Goldman Sachs funds wrote their $896 million position to zero, Volkswagen wrote down roughly €1.4 billion, Swedish pension funds absorbed about SEK 5.8 billion, CDPQ lost C$200 million, and Quebec lost its C$270 million. The administrator's report, completed in early 2026, attributed the collapse to the pandemic, the war in Ukraine, a weakening EV market, and, as the trigger, Volkswagen's autumn-2024 decision. In February 2026 the estate demanded roughly SEK 300 million back from suppliers paid shortly before the bankruptcy.

Collapse and aftermath

After Chapter 11, the Northvolt entities wound down across jurisdictions. The Swedish parent filed in March 2025. Cuberg had already shut in late 2024. The Quebec entity was declared insolvent in September 2025. Northvolt Drei, the partly built German gigafactory in Heide that never reached production, moved under German state control while a sale to Lyten is negotiated; roughly €260 million had been spent on construction against an agreed sale price of about €60 million.

Lyten, founded in San Jose and developing batteries using 3D graphene and lithium-sulfur chemistry alongside conventional lithium-ion, bought the estate in stages: Cuberg in November 2024, the Gdańsk storage plant in October 2025, and the Swedish assets, Northvolt Ett and its expansion site, Northvolt Labs, the intellectual property, and more than 160 hectares of land, in a transaction closing February 2026. Lyten put the book value of the acquired assets at roughly $5 billion, about SEK 45 billion; actual purchase prices were not disclosed, and the agreed Heide terms of about €60 million for a site with more than €300 million sunk into it suggest prices ran far below asset values. Lyten is restarting Skellefteå line by line, initially producing cells for battery energy storage, plans to hire more than 600 people over the following 12 months, and is using Västerås to develop and qualify lithium-sulfur for gigascale production. It has also announced a Lyten Industrial Hub in Skellefteå co-locating manufacturing with data centers, where EdgeConneX, an EQT portfolio company, plans a site of up to one gigawatt.

Peter Carlsson, who resigned as CEO the day after the Chapter 11 filing, remained a senior advisor and board member through the proceedings. By May 2025 he had secured funding for a new venture, and in November 2025 it came out of stealth: Aris Machina, co-founded with former Northvolt AI lead Siddharth Khullar as CEO, building an agentic operating system for manufacturing with battery production as the first use case, backed by a $10.7 million pre-seed round led by Earlybird Venture Capital.

Key considerations

Northvolt's failure is most usefully understood as a manufacturing execution problem rather than a technology or demand problem. The cell chemistry was sound. EV demand in Europe grew throughout the period, and the strategic case for European battery production has not weakened. What Northvolt could not build was the production system: the interlocking combination of process knowledge, workforce capability, quality management, and equipment optimization required to turn a functional factory building into a profitable cell business at automotive scale and specification. That is a different challenge from inventing a battery, and it requires institutional knowledge that cannot be raised in a funding round.

The role of customer order pressure deserves attention. The order backlog was both the company's greatest asset and a constraint that made the production problem worse. BMW and Volkswagen delivery commitments created pressure to ship cells before the production system was ready, pushing the factory to ship low-yield material rather than run the slower iterative improvement cycles that would have built process capability. That dynamic, where premature volume commitments force delivery of off-spec cells and destroy customer relationships, recurs across industries and is particularly dangerous in battery cells, where quality failures can disqualify a supplier from automotive supply chains for years.

Whether Lyten can succeed at Skellefteå where Northvolt could not depends on factors outside the physical assets. Lyten is restarting with conventional NMC while developing lithium-sulfur, a new chemistry with its own process development challenges, and it inherits the same workforce management, equipment commissioning, and supply chain questions that proved fatal. The buildings, the 160 hectares, the hydropower access, and the Västerås facility are genuine assets. The manufacturing knowledge that makes a gigafactory work does not transfer in a property transaction.

Sources