Former Sugar Refinery Near Caen, France, Could Become AI Data Center

Why It Matters

The conversion adds critical AI compute power using existing low‑carbon energy assets, accelerating France’s digital sovereignty while revitalizing a dormant industrial zone.

Key Takeaways

• Brown Fields to acquire 33‑hectare former sugar plant.
• Project cost estimated €15‑€20 million for AI data center.
• Site has existing 220 kV line and nuclear power link.
• France seeks AI compute capacity, adding non‑government sites.
• Redevelopment aligns with sustainability goals for industrial reuse.

Pulse Analysis

Europe’s AI boom is reshaping the landscape of industrial real estate, as developers scramble to secure locations with ample power and low latency. In France, the government’s push to double national compute capacity has sparked a wave of conversions, turning former factories, warehouses and even mines into purpose‑built data hubs. These projects promise to meet the voracious energy demands of large‑language models while leveraging existing grid connections, thereby shortening deployment timelines. The trend reflects a broader shift toward “green‑by‑design” data centers that integrate renewable or low‑carbon energy sources.

The 33‑hectare Saint Louis sugar plant in Cagny offers a textbook case. Closed since 2020, the site already hosts a 220 kV transmission line and a direct tie to the Flamanville nuclear station, eliminating the need for costly new grid infrastructure. Brown Fields, a specialist in brownfield redevelopment, estimates the conversion will require €15‑€20 million, a modest outlay compared with greenfield builds. By demolishing the obsolete refinery and repurposing its wastewater‑treatment facilities, the project can achieve a lower carbon footprint while delivering the high‑density power needed for AI workloads.

Beyond the technical advantages, the Cagny data center could revitalize the Calvados region, creating high‑skill jobs and attracting ancillary services such as cooling‑system manufacturers and network providers. Although the site is not among the 35 government‑designated “turn‑key” locations, its private‑sector backing illustrates how France’s AI agenda is being accelerated through market‑driven initiatives. As more legacy sites gain access to reliable, low‑carbon power, competition for AI compute capacity will intensify, prompting operators to prioritize efficiency and sustainability. The success of this conversion may set a precedent for similar projects across Europe’s industrial heartland.