RE: Hyaxiom North American Doosan subsidiary30 Jun 2026 09:16
Difference between Doosan legacy fuel cell tech and Ceres SOFC. Can they both produce hydrogen?
according to AI (Claude):
Doosan's legacy technology is PAFC (phosphoric acid fuel cell), the platform it built its reputation on with large stationary power plants in South Korea, including the 50 MW Shinincheon Bitdream plant β which was the world's largest plant of its kind. PAFC uses a liquid phosphoric acid electrolyte, runs at moderate temperatures (roughly 150β200Β°C), is mature and well-proven, but has lower electrical efficiency and slower startup than newer designs. That said, Doosan terminated three PAFC supply contracts in April 2025 worth around $560 million, citing project delays and financing problems, signaling some strain in that legacy business.
Ceres SOFC (solid oxide fuel cell) is a different, newer-generation technology that Doosan licensed rather than developed in-house. Doosan signed a joint development agreement with Ceres Power, a British SOFC company, back in 2020 to develop SOFC cells and stacks. SOFCs use a solid ceramic electrolyte and operate at much higher temperatures (typically 600β800Β°C+), which gives them notably higher electrical efficiency β SOFCs are considered the most efficient fuel cell type, reaching 60% efficiency or more. Doosan's newer SOFC systems are reaching up to 64% electrical efficiency, aimed at high-growth markets like AI data centers. Doosan has now moved this from pilot stage to real commercial scale: mass production began in July 2025 at a new 50MW factory in Jeollabuk-do, South Korea β the world's first factory built for Ceres' metal-supported SOFC design.
So the practical difference: PAFC is the older, simpler, lower-temperature workhorse technology; SOFC is the newer, hotter, more efficient design that Doosan doesn't own outright but builds under license from Ceres. Doosan hasn't dropped PAFC entirely β its stated strength is still a diversified portfolio spanning PAFC, SOFC, and PEM β but the strategic momentum has clearly shifted toward SOFC.
On your hydrogen question: in their normal mode, both PAFC and SOFC consume hydrogen (or a hydrogen-rich fuel) to generate electricity β they don't produce it. However, solid oxide cells have a special property PAFCs don't: they're reversible. Run backwards as a solid oxide electrolysis cell (SOEC), the same basic ceramic stack can split steam into hydrogen and oxygen instead of consuming hydrogen, essentially because the high operating temperature and solid oxide electrolyte work efficiently in both directions. Other SOFC makers like Elcogen explicitly market combined SOFC/SOEC technology, and Ceres itself has SOEC/electrolysis development underway as a logical extension of its SOFC platform. PAFC's liquid-electrolyte, lower-temperature chemistry isn't well suited to efficient reverse operation, so it's essentially a one-way (power-generation-only) technology. Energy Institute
So: SOFC, in principle, can be flipped to make hydrogen v