Enhanced Geothermal Systems (EGS): What Is Operating, What Is Under Construction, and What Is Only a Model

An enhanced geothermal system (EGS) is not a larger version of the geothermal system used to heat and cool a house. They share a label and a general reliance on underground heat, and past that they are different technologies doing different jobs. Home heating and cooling relies on a ground-source heat pump, a shallow-loop appliance; an enhanced geothermal system is a utility-scale power-generation technology that drills kilometers down, engineers a reservoir where nature didn't provide one, and sends electricity to the grid (what geothermal energy is covers the fundamentals).

The power-plant technology has a second confusion layered on top: the gap between what EGS does today and what gets written about it. The commercial EGS electricity flowing to a US grid today is measured in single-digit megawatts. The same technology is the subject of a federal cost-reduction moonshot and a 2050 modeling scenario measured in tens of gigawatts. Confusing deployed capacity with modeled capacity makes EGS look either trivial or far more mature than it is.

What an Enhanced Geothermal System Is

The Department of Energy's definition: an enhanced geothermal system is a human-made underground geothermal reservoir, created where hot rock exists but natural fluid or permeability is insufficient. Wells are drilled, fluid is injected under controlled conditions to reopen existing fractures or create new ones, and the resulting connected fracture network circulates fluid through the hot rock.

The USGS phrasing, from Fact Sheet 2008-3082, is even more precise: EGS resources "require some form of engineering to develop the permeability necessary for the circulation of hot water or steam."

What gets enhanced is the permeability, not the heat. The heat is already down there. A conventional hydrothermal plant — the kind that has powered parts of the American West for decades — needs three things to coexist naturally: hot rock, fluid in the rock, and permeability that lets the fluid move. Those three rarely line up in the same place. EGS engineers the missing ingredient. It supplies the fluid and creates the permeability, turning otherwise unproductive hot rock into a working subsurface heat exchanger.

What EGS is not

Two technologies get blended with EGS in coverage, and the distinctions matter.

EGS is not closed-loop geothermal (AGS). Conventional EGS depends on an engineered fracture network in the rock itself — the injected fluid touches the formation. Closed-loop, sometimes called advanced geothermal, circulates fluid inside sealed pipe and never contacts the rock. DOE treats them as separate technology categories.

EGS is not a ground-source heat pump. The heat pump is a shallow-loop appliance drawing on near-surface ground temperatures, and its drilling is a different exercise entirely — covered in geothermal drilling and wells. EGS wells go kilometers down and feed a power plant. Same word, different technologies.

How an Enhanced Geothermal System Works

A typical enhanced geothermal system diagram shows four elements, and they map directly onto the DOE definition:

1. Wells drilled deep into hot rock. Deep EGS power projects generally target rock in the range of about 150–200°C or hotter, kilometers below the surface, though the temperature and depth that make a project work vary with the resource and the conversion technology. What defines EGS is not a temperature threshold but the condition of the rock: hot enough to be useful, and lacking the natural plumbing a conventional geothermal plant needs.
2. Injected fluid. Fluid is pumped down under controlled conditions to reopen existing fractures in the rock or create new ones.
3. The engineered fracture network. This connected network of fractures is the human-made reservoir — the defining feature of EGS. Fluid circulates through it, moving through hot rock and picking up heat along the way.
4. The surface plant. The heated fluid returns to the surface, where its heat is used to generate electricity for the grid.

The subsurface engineering is what separates an EGS project from a conventional one, and it's where the cost lives — deep wells, fracture creation, and the challenge of making the circulation loop between wells actually connect. It is also why EGS matters at all: hot dry rock without natural permeability is a far more widespread resource than the rare spots where heat, fluid, and permeability coincide on their own.

Operating, Under Construction, or Modeled

Every capacity number you will read about enhanced geothermal falls into one of three categories, and coverage routinely blurs them.

An operating megawatt is a fact. A megawatt under construction is a plan with steel in the ground and a date attached. A modeled gigawatt is what a computer says the grid would build if a set of assumptions comes true. All three are legitimate numbers; none of them is interchangeable with the others.

What Is Operating in the U.S. Today

The reference point for commercial enhanced geothermal electricity in the United States is Fervo Energy's Project Red in Nevada, which Fervo reports generating 3 MW.

Project Red's documented timeline:

• July 18, 2023 — Fervo announced results of a well test supporting 3.5 MW of output. That is a test-derived capability, not continuous commercial operation.
• November 28, 2023 — Google announced that electricity from the project had begun flowing to the Nevada grid. The announcement stated no capacity figure.
• February 13, 2026 — a Fervo securities filing recorded that commercial quantities of electricity began flowing in November 2023, and that commercial operation was declared April 30, 2024.
• May 2026 — Fervo's prospectus, dated May 12 and filed May 14, lists 3 MW online and generating.

Three megawatts is small — roughly one thirty-thousandth of the modeled 2050 total discussed below. It is the commercial baseline against which announced projects should be measured. Project Red demonstrated that an engineered reservoir can deliver commercial electricity to a real grid.

What Is Under Construction

The project that would change EGS's scale is Cape Station in Utah, which Fervo calls "the world's largest enhanced geothermal project" — a company description, not a ranking drawn from any published global project inventory. As of Fervo's Q1 2026 results, published June 22, 2026:

• Phase I is roughly 100 MW. GeoBlock Unit 1 is in commissioning, with commercial operation planned for Q4 2026. Units 2 and 3 have planned commercial operation dates in Q1 2027.
• Phase II is roughly 400 MW. Construction began in Q1 2026, with commercial operation expected in 2028.
• Fervo reports 658 MW of contracted power purchase agreements, an agreement with Google covering up to 3 GW through 2033, and $2.2 billion raised in its IPO.
• The company says the project is "on track for first power in Q4 2026."

Those milestones show construction and contracting progress, but they do not constitute operating capacity. Every megawatt of Cape Station sits in the "under construction" row — planned dates, not delivered power. Construction schedules in any industry can move, which is exactly why the category distinction exists. If Cape Station's Unit 1 delivers first power on schedule in Q4 2026, operating EGS capacity jumps from 3 MW to tens of megawatts. Until then, 3 MW is the operating number.

Companies in EGS and Related Deep Geothermal

Fervo Energy dominates this article because Fervo is the developer with a declared commercial EGS operation and a published construction pipeline — Project Red operating, Cape Station under construction, capacity and dates on the record in filings.

Fervo is not alone in the field. Other companies developing enhanced geothermal and related deep-geothermal technologies include Eavor, Sage Geosystems, and Quaise. Note that some of the approaches in this space are closed-loop designs rather than EGS in the fracture-network sense defined above — when evaluating any company in the sector, the first question to ask is which technology it is actually building.

The Enhanced Geothermal Shot

The number that frames all EGS policy discussion comes from the Department of Energy's Enhanced Geothermal Shot, announced September 8, 2022. The target, verbatim from DOE's September 2022 fact sheet:

"reduce the cost of EGS by 90%, to $45 per megawatt-hour (MWh) by 2035."

— U.S. Department of Energy, Enhanced Geothermal Shot fact sheet, September 2022

The target requires two important qualifications.

First, $45/MWh by 2035 is a target, not a price. It's the goal of a federal R&D initiative — the level DOE believes EGS must reach to deploy at scale.

Second, the baseline. A 90% cut ending at $45/MWh implies a starting point near $450/MWh — but DOE does not state that baseline explicitly, and never presents it as a measured market price for EGS power today; the fact sheet's footnote says only that the target derives from NREL's 2022 Annual Technology Baseline.

What the 90.52 GWe Figure Actually Is

In January 2023, NREL published its Enhanced Geothermal Shot Analysis for the Geothermal Technologies Office (NREL/TP-5700-84822, Augustine, Fisher, Ho, Warren & Witter). The report asked: if the Enhanced Geothermal Shot's cost targets are met, how much geothermal would the U.S. electric grid build? The answer, from the ReEDS capacity-expansion model, is the source of every "90 GW by 2050" headline. The resulting 90.52 GWe figure is often stripped of three essential qualifications: it is modeled, it includes hydrothermal capacity, and it belongs to a specific U.S. decarbonization scenario.

First, it is a model output, not a forecast. ReEDS was fed a set of technology-improvement assumptions and asked what a cost-optimized grid would do with them. The report's modeled assumptions include production-well flow rates improving from 40 to 125 kg/s, drilling success rates from 75% to 90%, confirmation-well success from 50% to 75%, representative plant size growing from 25 to 100 MWe, and overnight capital cost falling from $32,255 to $3,565 per kWe — an 89% reduction in capital cost. The 90.52 GWe is what the model builds if all of that happens. It is not a prediction that it will, and every use of the number needs that condition attached.

Second, it is total installed geothermal, not EGS alone. Here is the report's Table 4 — ReEDS model results for U.S. installed capacity by technology type, in GWe:

Figures in GWe, as published. Components may not sum exactly to totals because of rounding.

Four different technologies make up that total. Deep EGS is the giant at 80.13 GWe in 2050, with near-field EGS contributing another 5.13 — but "90 GW of EGS" is a misquote of the table. The 90.52 is everything, hydrothermal included.

Third, the scenario has specific boundaries. The modeling runs under the Solar Futures Study Decarbonization case: a 95% reduction in electric sector-wide CO2 emissions from 2005 levels by 2035, and 100% by 2050, holding existing policies constant as of June 2020. That is an electric-sector emissions constraint, not an economy-wide one — a one-word difference that changes what the scenario claims. The model also assumes EGS becomes available for commercial deployment starting in 2030, and all of it is U.S. modeling; the report says nothing about global deployment.

The field today: 3 MW operating, roughly 500 MW under construction with first power planned for late 2026, and 85.26 GWe of modeled EGS by 2050 inside a scenario where capital costs fall 89%. The distance between the first number and the last is the entire open question of enhanced geothermal.

The Capacity Factor Advantage

The same NREL model explains why geothermal capacity expands so sharply when the assumed costs fall.

In the modeled 2050 grid, geothermal is just 3.94% of national installed generating capacity — but it supplies 12.04% of annual generation. In 2035 the same pattern holds: 1.94% of capacity, 6.13% of generation. The report states that geothermal's share of generation is "three times larger than its percentage of installed capacity… due to the high capacity factor of geothermal technologies."

Capacity factor is the ratio of what a plant actually generates over a year to what it would generate running flat-out the whole time. A geothermal plant's fuel is heat in rock — it doesn't set with the sun or die with the wind, so the plant can run near-continuously. Each installed geothermal megawatt therefore delivers roughly three times its proportional share of the modeled grid's electricity.

That ratio is the strategic argument for EGS in one statistic. In a decarbonizing grid, a firm, high-capacity-factor renewable resource punches far above its installed-capacity weight — which is why a model told to build a zero-carbon electric sector reaches for so much of it once the assumed costs allow.