Startup tests seafloor geothermal power for diesel-dependent island energy systems
Fifteen hundred meters beneath the Pacific, a remote-controlled drill bit into the seafloor last week near one of the most volcanically active ridges on Earth. The team directing it never left their command center aboard a research vessel — they didn’t need to.
The operation belongs to Endurance Energy, a Seattle startup hunting for the right patch of ocean floor to tap scalding hydrothermal fluid rising from Earth’s crust. Even the company’s own CEO describes it as a venture with “very, very little precedent.” One prominent expert at Stanford has a blunter assessment.
A well on the ocean floor
The chosen site sits near Axial Seamount, the largest and most active underwater volcano in the Pacific Northwest, roughly 500 kilometers off the Oregon coast. Endurance spent several weeks using remote sensing to survey the Juan de Fuca Ridge before settling on a flat patch of seafloor at 1,500 meters depth.
A submarine robot will unspool flexible piping to connect the two, allowing hydrothermal fluid to flow into the system and electricity production to begin.
Once on site, the team drilled a 60-meter well targeting hydrothermal fluid at around 250 °C — a temperature Redd calls a “sweet spot.” Hot enough to generate power, but cool enough that engineers can still rely on conventional terrestrial materials rather than exotic alloys.
The borehole is just 5 centimeters in diameter. That’s dramatically narrower than land-based geothermal wells, which typically run 20 to 30 centimeters across, and every engineering decision made in deep water compounds that constraint.
Site selection wasn’t purely geological. The Juan de Fuca Ridge is threaded with fiber-optic cables operated by the National Science Foundation’s Ocean Observatories Initiative — infrastructure that lets Endurance stream data to shore in real time. That existing network was a meaningful factor in choosing this particular stretch of seafloor over others, according to IEEE Spectrum.
How Adélie turns ocean heat into electricity
The generator Endurance built for this mission is named Adélie, after the Antarctic penguin. It produces up to 100 kilowatts using organic Rankine cycle technology: a closed-loop process that transfers heat to a working fluid, which then expands to spin a turbine. No combustion involved.
Adélie will be lowered to the seafloor and positioned roughly 20 meters from the well. A submarine robot will unspool flexible piping to connect the two, allowing hydrothermal fluid to flow into the system and electricity production to begin. It’s a slow, methodical operation — the kind where nothing can be improvised once the hardware is on the bottom.
This is the first time this type of generator has been connected to a well drilled directly into the ocean floor. Materials had to be carefully selected to handle the combined stress of extreme pressure, high salinity, and elevated temperatures — conditions that degrade standard components faster than most engineers would like.
Data first, power second
Despite the effort to build a working generator, Endurance isn’t primarily chasing kilowatt-hours on this deployment. During the pilot phase, sensors will consume only a small fraction of Adélie’s 100-kilowatt capacity.
The surplus electricity won’t go anywhere useful — at least not yet. It’ll be released back into the surrounding water as heat, in a process Redd compares to electrical grounding. No cable runs to shore. No grid connection of any kind exists.
What does travel to shore is data. Temperature and flow-rate readings stream in real time through the NSF’s Ocean Observatories Initiative network, giving the team a continuous picture of how the system performs at depth. Before deployment, Endurance also took baseline temperature and ecological measurements around the vent site — readings that will serve as a reference point for evaluating any impact the technology has on the surrounding ecosystem, something the company says it’s taking seriously.
The real target: island nations running on diesel
Oregon is a classroom. Tonga is the exam.
By the end of 2027, Endurance plans to deploy a 2-megawatt unit off the coast of the Kingdom of Tonga, a Pacific island nation that depends heavily on diesel shipped from abroad. Fuel prices there fluctuate with global geopolitics, making energy costs both high and unpredictable. Unlike the Oregon pilot, the Tonga system will deliver electricity directly to shore through submarine cables — a genuine power source rather than a data-gathering exercise.
It also won’t be sited as close to active volcanic zones, since seismic risk is a real consideration for a system meant to run reliably. Redd says the system will be cost-competitive with diesel, though exact figures are still being worked out. Endurance’s director of policy, Jennifer Kenyon, draws a deliberate parallel to offshore wind — an industry that faced comparable skepticism about logistics and cost before eventually proving viable across multiple markets.
Skeptics and a long history of failed attempts
Roland Horne, who directs Stanford’s geothermal program, doesn’t mince words. He calls subsea geothermal “a slightly dingbat idea” for grid-scale power generation, though he allows that island applications represent a plausible niche.
The history backs him up. Companies have repeatedly attempted offshore geothermal at commercial scale and failed, with questions about durability, logistics, and deep-ocean performance remaining largely unanswered. Critics also point out that abundant terrestrial geothermal resources sit untapped around the world. “It’s not like oil and gas; that we ran out and had to look somewhere else,” Horne says. Going offshore, by this logic, means choosing the harder path without exhausting the easier one first.
Redd doesn’t dispute the difficulty. He openly describes the Oregon deployment as a learning exercise rather than a proof of concept. What happens over the next year — how the well performs, how Adélie holds up, what the data reveals — will determine whether this idea graduates from dingbat to viable. The answer starts now, 1,500 meters down.
Daniel García is an Editor-in-Chief with strong expertise in structural work and engineering principles. He combines this technical foundation with deep knowledge of energy, spatial design, and emerging technologies, bringing a forward-thinking and analytical approach to editorial leadership.
