Executive brief
On 5 December 2023, Symbio inaugurated SymphonHy at Saint-Fons, south of Lyon: described by the company and by Reuters that day as Europe's largest integrated hydrogen fuel-cell gigafactory, with an announced annual capacity of 50,000 fuel-cell systems. Twenty-nine months later, on 13 May 2026, a redundancy plan cutting 356 of roughly 530 posts was signed (Le Progrès, 12 May 2026). Between those two dates, one customer left, one market failed to arrive, and, according to a Le Monde investigation published 11 May 2026, more than €350 million of public support, including some €3128–313 million8 channelled through the EU IPCEI hydrogen programme, had been consumed. This dossier reconstructs the sequence from dated public sources, separates what is established from what is contested, and extracts the part that matters to an engineer choosing where to spend the next five years of a career.
I. What Saint-Fons was built to be
Symbio was founded as a joint venture combining Michelin's electrochemistry and materials heritage with Forvia's automotive-systems industrialisation capability; Stellantis later joined as a third shareholder and prospective anchor customer. The industrial thesis was legible: hydrogen fuel-cell systems for light commercial vehicles and heavy duty, produced in France at a volume that would drive unit cost down the learning curve. In October 2022 Symbio announced a €1 billion investment programme in France to support that ramp (Symbio, 6 October 2022).
SymphonHy was the physical expression of that thesis. At its inauguration on 5 December 2023, Symbio described a fully integrated site (membrane-electrode assembly, bipolar plates, stack assembly, and system integration under one roof) sized for 50,000 systems a year. Integration was the strategic bet: a plant that owns the whole stack controls its own cost curve rather than importing it.
That is an engineering argument, and it was a good one. What it was not is a demand forecast.
II. The customer that left
On 16 July 2025 Stellantis announced it was ending its hydrogen fuel-cell development programme, stating that it saw no medium-term market for the technology (Stellantis, 16 July 2025). Michelin and Forvia confirmed the same day that they had been informed in May 2025 of Stellantis's intention to halt hydrogen activities from 2026 (Michelin, 16 July 2025). Le Monde reported on 16 July 2025 that the withdrawal placed the French hydrogen supply chain in jeopardy.
Twelve days later, on 28 July 2025, Symbio published a statement calling for "an all-out effort" to secure the company's future and characterising the withdrawal of its partner from light-commercial-vehicle hydrogen activity as sudden and unilateral. For a plant sized at 50,000 systems, the departure of the largest committed order book is not a commercial setback; it is a capacity problem with no near-term remedy.
One redundancy plan, two counts, two dates
| Stage | Posts cut | Of a site total | Publisher and date |
|---|---|---|---|
| Plan proposed | 358 | About 530 | Symbio, 3 December 2025 |
| Plan signed after negotiation | 356 | About 530 | Le Progrès, 12 May 2026 |
| Employees’ own count at a protest | 358 dismissals | Not stated | Le Progrès, 29 April 2026 |
Sources 6 Symbio · 7 Le Progrès · 9 Le Progrès
III. The arithmetic of a seventy per cent cut
On 3 December 2025 Symbio announced what it called a major transformation: a redundancy plan (plan de sauvegarde de l'emploi) proposing to cut 358 of roughly 530 posts, leaving about 175 (Symbio, 3 December 2025). After negotiation, the signed plan of 13 May 2026 carried 356 cuts (Le Progrès, 12 May 2026).
| Date | Event | Figure on the record |
|---|---|---|
| 6 Oct 2022 | Symbio announces French investment programme | €1bn |
| 5 Dec 2023 | SymphonHy inaugurated at Saint-Fons | 50,000 systems/year |
| 16 Jul 2025 | Stellantis ends its fuel-cell programme | 1 of 3 shareholders |
| 3 Dec 2025 | Redundancy plan proposed | 358 of ~530 posts |
| 13 May 2026 | Plan signed after negotiation | 356 posts |
| 27 May 2026 | Michelin and Forvia set out a relaunch | 2 remaining shareholders |
The public-money question follows. Le Monde reported on 11 May 2026 that more than €350 million of public support had been consumed in two years, of which approximately €312–313 million came through the European IPCEI hydrogen programme: a programme authorised in 2022 at more than €5 billion across 41 projects and 15 countries (France Hydrogène, 31 July 2022). On 9 June 2026 infodujour reported that an anti-corruption association had filed a complaint with the Lyon public prosecutor concerning the use of those funds.
One figure in circulation should be treated with care. On 29 April 2026 Le Progrès reported employees citing €669 million of public subsidy during a protest. That figure is not reconciled with Le Monde's €350 million and we present it as contested rather than established. We also record what we could not verify: no formal court-supervised safeguard or redressement judiciaire filing, and no third-party acquirer, appears in the sources we could reach.
Two public-subsidy totals that cannot both be right
| Amount as printed | What it is said to cover | Publisher and date | Status here |
|---|---|---|---|
| More than €350 million, of which some €312–313 million via the EU IPCEI hydrogen programme | Public support consumed in two years | Le Monde, 11 May 2026 | Reported by an investigation |
| €669 million of public subsidy | Cited by employees at a protest | Le Progrès, 29 April 2026 | Contested, not established |
| More than €5 billion across 41 projects and 15 countries | The IPCEI hydrogen programme authorised in 2022 | France Hydrogène, 31 July 2022 | Programme envelope, not a company figure |
Sources 8 Le Monde · 9 Le Progrès · 12 France Hydrogène
IV. Why the market did not arrive on time
The Saint-Fons story is often told as a governance failure. The demand data suggest something more structural. The IEA's Global Hydrogen Review 2025 put global hydrogen demand at roughly 100 Mt in 2024, up about 2% year on year, with new applications, transport among them, accounting for less than 1% of the total. A plant whose economics require volume cannot be rescued by a market that remains a rounding error.
Refuelling infrastructure tells the same story from the road. France counted 80 hydrogen refuelling stations in service in January 2025, with 91 more planned or under construction, on France Hydrogène figures. For heavy duty specifically, France Routes reported on 10 February 2026 that a European network of 92 distribution points for trucks was operational. The EU Alternative Fuels Infrastructure Regulation (Regulation (EU) 2023/1804), applicable since 13 April 2024, sets binding hydrogen deployment targets along the TEN-T corridors, which is precisely the point: in 2026 the network is still being built to a mandate, not to a queue of operators waiting for fuel.
Read together, the three numbers describe a classic sequencing error. Capacity was financed on a 2030 demand curve and staffed on a 2024 calendar.
50,000fuel-cell systems a year
Announced annual capacity of SymphonHy at its inauguration
Stated by the company and by Reuters on 5 December 2023. It is a nameplate figure at inauguration, never published as an achieved output, and the plant was sized on it before its largest committed order book departed in July 2025.
V. What survives
The plant was not liquidated. On 27 May 2026 Le Journal des Entreprises reported that, following Stellantis's exit, Michelin and Forvia remain the sole shareholders and had agreed a relaunch focused on heavy-duty and industrial applications for the gigafactory. That is a materially different outcome from closure, and it matters for anyone assessing the site: the industrial asset, the process know-how and a core team of roughly 175 posts were kept deliberately, by two shareholders who chose continuation over exit.
It also fits Michelin's own stated direction. Its Michelin in Motion 2030 strategy commits the group to growth beyond tyres in polymer composite solutions and adjacent territories, and Michelin's public statements through 2025 and 2026 have consistently framed hydrogen as a materials and electrochemistry play rather than a passenger-car bet.
VI. What an engineer should take from this
The instinct after a 70% cut is to write off the technology. That is the wrong lesson, and it is expensive.
First: the skills were not the problem. Membrane-electrode assembly, coating and lamination process control, stack leak-testing, high-voltage safety, and statistical process control on a low-yield line are transferable to battery cell manufacturing, electrolysers, and semiconductor back-end assembly. In Auvergne-Rhône-Alpes those three destinations sit inside a two-hour radius. A CV built at SymphonHy reads as clean-room process engineering, not as hydrogen advocacy.
Second: read the order book, not the ribbon-cutting. The single question that would have flagged this risk in 2023 was not technical. It was: how many of the announced annual units are covered by a signed, penalty-bearing offtake commitment, and from how many distinct customers? A three-shareholder joint venture where one shareholder is also the anchor customer is a concentration risk wearing a governance costume.
Third: public subsidy is a timing signal, not a demand signal. IPCEI and France 2030 funding compress the capital cost of building capacity. They do not create buyers. When a site's headcount plan depends on a regulated network that is still under construction, the AFIR corridors are the clearest example, the honest planning assumption is a longer ramp, not a faster one.
Fourth: the sector is industrialising, not disappearing. France Hydrogène's fourth annual barometer, published 27 January 2026, frames the French hydrogen sector as moving out of a learning phase into industrialisation, and a BDO study for France Hydrogène published in January 2025 assessed that the sector already generated more than €1 billion of value added in 2023. Industrialisation means fewer, larger, better-underwritten projects, and a hiring profile that shifts from pilot-line experimentation towards yield, quality and cost engineering.
VII. Five questions to ask at the next hydrogen interview
They are deliberately unglamorous, and each maps to a failure visible in the sequence above.
1. What share of nameplate capacity is under contract, and with how many customers? 2. Which shareholder is also a customer, and what happens to the plan if that shareholder leaves? 3. Is the funding capital support or revenue, and when does it end? 4. Which regulated infrastructure must exist before your customer can operate, and what is its dated build-out? 5. If volume arrives two years late, what is the retained core team, and are you in it?
Symbio's engineers could not have answered question two in 2023, because the answer was not public until July 2025. That is the honest conclusion of this dossier: the technical bet at Saint-Fons was competent, and the exposure that broke it was commercial and political. Careers are made or lost on the second kind of risk far more often than professionals are trained to expect.
VIII. What we will watch next
Three dated markers will tell us whether the relaunch is real: the composition of the retained team once the plan of 13 May 2026 completes; whether heavy-duty orders appear from customers outside the shareholder base; and whether the French refuelling network moves materially beyond the 80 stations counted in January 2025. We will update this dossier against those markers rather than against announcements.