Geothermal: Conventional Plants and Enhanced Systems

The Earth's interior is hot, and the heat leaks toward the surface everywhere. In a few places it comes up close enough, with enough water around it, to boil. Geothermal power drills into that heat and brings it up as steam or hot water to spin a turbine. Unlike wind and solar, it runs day and night, in any weather.

For all that, it is small. EIA says geothermal made about 16 billion kWh in 2025, about 0.4 percent of U.S. utility-scale electricity, from plants in seven states. California produced 68.6 percent of it and Nevada 24.7 percent. Utah, Hawaii, Oregon, Idaho and New Mexico split the rest. Idaho has one utility-scale plant of about 10 MW, which made 0.5 percent of U.S. geothermal output.

Three ways to turn heat into power

In each design the used water is usually injected back underground. That keeps reservoir pressure up and disposes of brine that can carry dissolved minerals and gases. Binary plants keep the geothermal fluid in a closed loop and release almost no gases.

Capacity factor: geothermal's strong suit

EIA's national geothermal fleet ran at a 69 percent capacity factor in 2023, 65 percent in 2024 and 66 percent in 2025. Newer plants and well-maintained fields do better. Idaho's single plant posted capacity factors slightly above 100 percent most years from 2019 to 2024, which looks impossible until you remember the math: EIA divides by summer-rated capacity, and a geothermal plant running nonstop, a bit above its rating in cool months, can beat its own denominator. Older fields can also decline as reservoirs cool or lose pressure, which is why operators drill makeup wells over time.

Geothermal at 90%: 10 x 8,760 x 0.90 = 78,840 MWh per year
Solar at 25%:      10 x 8,760 x 0.25 = 21,900 MWh per year

Solar nameplate needed to match the geothermal's annual energy:
78,840 / (8,760 x 0.25) = 36 MW

Enhanced geothermal: making your own reservoir

Conventional geothermal needs heat, water and rock porous or fractured enough for the water to move, all in the same place. That combination is rare. Heat alone is everywhere if you drill deep enough. Enhanced geothermal systems (EGS) supply the missing parts. Drill into hot, dry rock. Pump water down at pressure to open or widen fractures. Drill a second well into the fractured zone. Circulate water down one well, through the hot cracks, and up the other.

The idea dates to the 1970s at Los Alamos. What changed recently is borrowed from shale oil and gas: horizontal drilling, multi-stage fracturing and better drill bits. Several wells can be drilled from one pad and run sideways through the hot zone.

The 2026 milestone

Fervo Energy announced on September 24, 2026 that its Cape Station in Beaver County, Utah had synced to the grid. The first 33-MW block of a planned 100-MW first phase was generating, with contractual commercial operation expected by October 1, 2026, and two more blocks targeted by January 1, 2027. A further 400 MW was under construction for 2028. Fervo said its contracted offtake totals about 900 MW, with buyers including Southern California Edison, Shell Energy North America and Google, which announced a 396-MW agreement in September 2026 with an option to expand by about 600 MW. Fervo called it the first time a utility-scale enhanced geothermal development had reached first power. Fervo's earlier pilot, Project Red in Nevada, has supplied power since 2023.

Read that carefully. It is a first unit, and the long-run test for EGS is whether reservoirs hold their temperature and flow for decades and whether drilling costs keep falling. DOE's Enhanced Geothermal Shot sets a target of $45/MWh by 2035, a 90 percent cost reduction.

What it costs, according to two sources

Why the gap between Lazard and EIA? Different assumptions about capital cost, financing and which resources get built first (EIA's model picks the cheapest remaining hydrothermal sites in each region). Neither is wrong; they answer different questions. NREL's 100% Clean Electricity by 2035 study projected 3 to 5 GW of new geothermal in its scenarios, small next to wind and solar but meaningful as round-the-clock clean supply.

Risks: earthquakes, water and dry holes

Injecting water underground can trigger earthquakes. Most are too small to feel. The case everyone cites is Pohang, South Korea: a magnitude 5.4 earthquake struck on November 15, 2017, and in March 2019 a government panel concluded an experimental geothermal plant's injection probably caused it. A 2006 project in Basel, Switzerland was shut after injection set off felt quakes. Modern EGS projects use "traffic light" protocols: seismometers watch in real time, and operators slow or stop pumping if tremors pass set thresholds. Whether those protocols are enough at commercial scale is a fair question to keep asking.

Water matters in dry country. Flash plants with evaporative cooling consume water; EGS needs water to create and run the reservoir, though most is recirculated. Air-cooled binary plants use little but lose efficiency on hot days. And exploration is a gamble: a well can cost millions and come up too cool or too tight. That upfront risk is why geothermal developers lean on public data, test drilling grants and, until recently, generous tax credits. Under the 2025 federal law, geothermal kept access to the technology-neutral credits for projects beginning construction through 2033, unlike wind and solar, which face a much earlier cutoff.