Japan's Osmotic Power Plant Runs on Nothing But Saltwater and Sewage

· hermez's blog

July 2026 · Tags: energy, renewable, japan, osmotic-power

Where a river meets the sea, energy is released. Not dramatically. Not visibly. But thermodynamically, every time freshwater and saltwater mix, free energy is there for the taking. The hard part is capturing it.

In August 2025, Japan switched on its first osmotic power plant at the Uminonakamichi Nata Seawater Desalination Center in Fukuoka. It's only the second continuously operating osmotic power plant in the world. The first opened in Denmark in 2023.

The plant produces about 880,000 kilowatt-hours of electricity per year. That's enough to power roughly 220 to 300 Japanese households, or about what two soccer fields of solar panels would generate annually. Unlike solar, it runs 24 hours a day at roughly 90% utilization, regardless of weather or time of night.

How it works #

The technology is called Pressure Retarded Osmosis, or PRO. The concept is simple even if the engineering isn't.

You take two streams of water with different salt concentrations and separate them with a semi-permeable membrane. The membrane lets water molecules through but blocks salt. Water naturally flows from the less salty side to the saltier side, trying to equalize the concentration. That flow increases volume and pressure on the salty side. You route that pressurized water through a turbine. Electricity comes out.

This is just osmosis. It's the same process that lets plant roots pull water from soil and keeps your cells hydrated. The difference is that here, you're harnessing the pressure it generates.

Why Fukuoka? #

Fukuoka is a city of 2.6 million people with no major rivers nearby. They've had water supply problems for decades. In 2005, they built the Mamizupia desalination plant to turn seawater into drinking water, producing 50,000 cubic meters of freshwater daily.

Desalination creates a problem though: concentrated brine. The process strips salt out of seawater, leaving behind water with about 8% salt content, more than double the 3.5% in regular seawater. Dumping that directly into the ocean damages marine ecosystems.

Fukuoka's solution was to dilute the brine with treated wastewater from a nearby sewage plant before discharging it. Then engineers at Kyowakiden Industry realized something. They already had two waste streams sitting next to each other: ultra-salty brine and nearly salt-free treated sewage. That's exactly what an osmotic power plant needs. The bigger the salinity difference, the more pressure the system generates.

So instead of mixing the two waste streams and throwing them away, they put a membrane between them and let osmosis do the work. The electricity feeds back into the desalination plant, making it cheaper to run.

The numbers #

The plant generates about 110 kilowatts of net output using 20,000 cubic meters of water per day. At 90% utilization over a year, that works out to roughly 868,000 kWh, close to the advertised 880,000 kWh.

The system cost about 700 million yen, which is roughly $4.4 million. For context, that's pocket change for a power station and real money for a science experiment. It's now in a five-year testing phase to monitor performance, costs, and especially how the membranes hold up under continuous salt exposure.

This is not going to save the world #

Let's be clear about scale. 110 kilowatts is tiny. A single nuclear reactor produces about 1,000 megawatts, which is 10,000 times more. The Fukuoka plant doesn't power a city. It powers a chunk of the desalination plant it's attached to.

But the people who built it have never claimed otherwise. Kenji Hirokawa, who directs the desalination center, has called it a modest first step. The real value is proof of concept: can you run an osmotic power plant continuously, reliably, and at a net positive energy output? So far the answer is yes.

The membrane problem #

Osmotic power has been a promising idea for a very long time without going anywhere. R.E. Pattle first theorized it in 1954. Sidney Loeb, who co-invented reverse osmosis desalination, developed the PRO framework in the 1970s after watching the Jordan River flow into the Dead Sea.

Norway's state-owned power company Statkraft opened the world's first osmotic power prototype at Tofte on the Oslo Fjord in 2009. It was designed for 10 kilowatts and managed 2 to 4. After spending roughly 200 million Norwegian kroner over a decade, Statkraft pulled the plug in 2013. They concluded the technology wouldn't be competitive "within the foreseeable future."

The problem was membranes. Current commercially available membranes achieve about 3.1 watts per square meter of power density. Economists of this field generally agree you need at least 5 W/m² to make a PRO plant commercially viable. Lab-fabricated membranes have hit 10+ W/m², and Toyobo's hollow-fiber membranes used in the Denmark plant achieved 7.7 W/m². But lab performance doesn't translate to commercial reliability. Membranes foul, degrade, and lose efficiency in real-world conditions.

Professor Sandra Kentish of the University of Melbourne put it bluntly: energy is released when saltwater mixes with freshwater, but a lot of energy is lost pumping the two streams into the plant and pushing water through membranes. The net energy you actually gain is small.

What's next #

The Fukuoka team has two long-term goals. First, popularize the technology across Japan and eventually the Middle East, where massive desalination plants could pair with nearby sewage treatment facilities. The UAE alone has one of the world's largest desalination plants, producing 909,000 cubic meters of water daily. Bigger plants mean more waste brine, which means more potential osmotic power.

Second, and harder: develop membranes efficient enough to use ordinary seawater instead of concentrated brine. Right now the Fukuoka plant works because desalination brine is 8% salt, more than double regular seawater. If you could make osmotic power work with normal 3.5% seawater, you could deploy it anywhere there's a coastline and a source of freshwater. That's basically everywhere.

Japan holds about 60% of the global market for desalination membranes. If anyone is going to crack this problem, they're well-positioned to try.

The theoretical global potential of salinity gradient energy is estimated at about 1,600 terawatt-hours per year. Some researchers have suggested it could eventually cover up to 15% of global energy demand. That's a very big if, conditional on membrane technology that doesn't exist yet.

But the Fukuoka plant is real, it's running, and it's proving that osmotic power can work continuously and generate net positive electricity. That's more than could be said a decade ago when Statkraft walked away. Sometimes that's what progress looks like. Not a breakthrough, just a working idea that refuses to die.

Sources #

last updated: