Berkeley Lab Seismometer Runs 7 Months at 338°F Underground
Deep underground at Fervo Energy's Cape Station in Utah, a seismometer built by Lawrence Berkeley National Laboratory has been running continuously for seven months at 338°F, nearly 7,000 feet below the surface. According to Berkeley Lab, that is the longest continuous high-temperature measurement ever recorded in a geothermal well. The hardware is still running.
Enhanced geothermal systems like Cape Station work by injecting water into hot, dry rock formations deep underground. The water absorbs heat from the rock and returns to the surface as steam, driving turbines. The problem has always been knowing what is happening down there. Microseismic events (tiny earthquakes caused by fluid injection and rock fracture reopening) are the primary diagnostic tool for reservoir behavior. But most sensors cannot survive the temperatures found at the depths where geothermal resources are richest. Berkeley Lab built one that can, and at Cape Station it is proving the concept.
"Such high-temperature measurements are critical for geothermal energy production, and as far as we know, this is the world's longest recorded measurement at this temperature," said Nori Nakata, a Berkeley Lab staff scientist who developed the instrument with Paul Cook at the lab's Geosciences Measurement Facility. The device is just under 10 feet long, sealed against water seepage, and designed without redundant components that could fail under sustained heat. It was lowered into the Cape Station well on July 27, 2025, and has been recording continuously since.
The practical consequence is real-time reservoir management. With continuous seismic data, operators can see fracture networks evolve as fluid moves through them. They can adjust injection rates and circulation patterns to maintain efficient steam production without triggering seismic events large enough to feel at the surface. Nakata's framing: for EGS to become a major U.S. energy source, you need a clear picture of rock stress, permeability, fluid pathways, and fracture growth. You cannot get that picture without hardware that lasts.
Fervo is building at scale. The company closed a $462 million Series E in December 2025, with Cape Station as the flagship development. In March 2026, it secured an additional $421 million in non-recourse project financing. Cape Station is located in Beaver County, Utah, adjacent to the Department of Energy's Frontier Observatory for Research in Geothermal Energy (FORGE) site. The first 100 megawatts of capacity is scheduled to come online in 2026, scaling to 500 megawatts by 2028.
Fervo operates six seismometers at Cape Station, supplementing 17 permanent stations operated by the Utah FORGE project and the University of Utah Seismographic Stations. Its traffic-light protocol defines green for events below magnitude 2.0, amber for 2.0 to 3.0 requiring a minimum six-hour pause, and red for magnitude 3.0 or above requiring a minimum 24-hour halt plus stakeholder outreach. No red-level events have occurred at Cape Station, according to Fervo's published monitoring data.
Berkeley Lab has been working geothermal problems since the 1970s, starting at The Geysers Field in California. Its current work is supported by the DOE Office of Geothermal, and the lab is separately developing superhot rock sensors rated above 700°F for next-generation systems. The Cape Station deployment is the current proving ground.
The question the hardware raises is not whether EGS works in principle. It is whether it can be operated predictably at commercial scale. Predictability requires data. Data requires sensors. Sensors require hardware that survives the environment. That chain is what this story is actually about.
Berkeley Lab is also applying machine learning to the sensor data streams, using AI to identify patterns across diverse datasets that human analysts would miss. Sireesh Dadi, manager of data acquisition and advanced analytics at Fervo Energy, described the approach as a game-changer for geothermal energy. The combination of hardened sensors and AI-assisted interpretation is what makes the Cape Station model distinct from earlier EGS demonstrations, which relied on periodic measurements rather than continuous monitoring.
Geothermal currently supplies roughly 0.4 percent of U.S. electricity generation, according to Energy Information Administration data. The resource base is orders of magnitude larger. The constraint has never been heat in the ground. It has been the ability to access and manage it without triggering induced seismicity that shuts projects down. The Berkeley Lab instrument does not solve that problem on its own. It is one component of a system that includes drilling, reservoir engineering, and grid integration. But it is a component that was missing, and it is now deployed and recording.
What makes this worth tracking: if Cape Station proves that continuous high-temperature monitoring at depth enables predictable reservoir operation, the financing model for EGS changes. Non-recourse project debt requires predictable cash flows. Predictable cash flows require predictable reservoir behavior. Predictable reservoir behavior requires the kind of data this seismometer is collecting. The hardware is not the whole story. It is the missing measurement that makes the rest of the story investable.