Dalian researchers boil seawater with electrolyzer waste heat in a 250 kW hydrogen plant

An engineer specifying an electrolyzer for a coastal site has two ways to get water into it. Both cost something. If the stack runs on raw seawater, chloride ions corrode the anode and calcium and magnesium deposits build up on the cathode ^[5]. If the site desalinates first, by reverse osmosis or distillation, it takes on extra equipment and a larger electricity draw ^[6].

The group led by Deng Dehui and Liu Yanting at the Dalian Institute of Chemical Physics started from the desalinate-first route and looked for energy it was losing ^[4][18]. Commercial stacks run at 80 to 90 degrees Celsius, and about 30% of their electricity can leave as low-grade heat, according to the study ^[7]. If the 250 kW rating is electrical input, that comes to roughly 75 kW of heat ^[19]. The plant pipes that heat into a vacuum distillation tower. There seawater boils at 40 to 50 degrees Celsius and the vapor condenses into fresh water ^[8].

Interesting Engineering's headline on the study has the plant pulling uranium from seawater ^[23]. The article's text is narrower: the concentrated brine can be processed to recover salt, uranium, bromine and other marine resources ^[22]. The fresh water output exists but is modest, about 0.7 tonnes a day ^[20]. The Chinese Academy of Sciences said that is enough to feed the electrolyzer with water left over ^[1]. The reporting does not split the annual total into feed and surplus.

Deng named three priorities for future work: better catalysts, recovering more of the electrolyzer's waste heat, and using artificial intelligence to optimize the plant's operating parameters ^[17].

The design suits a coastal hydrogen project that would otherwise build a separate desalination unit to feed its stacks. I'd model it as a hydrogen plant that makes its own water, and leave the minerals out until the numbers exist. The tradeoff is that a team waiting for those numbers may undersell a site with a salt or bromine buyer nearby.

Each output gets two tests: whether the study reports an annual figure for it, and whether the site has a buyer for it. An output with a figure and a buyer goes in the revenue model, and hydrogen is the obvious entry. An output with a figure and no buyer is a site utility. A small water surplus belongs in that box at a site that already has mains water. A buyer without a measurement is a conversation to open with the research team. An output with no figure and no buyer stays out of the model, and uranium belongs there until someone publishes a recovery rate.