Geothermal Turbine: How It Works and What the Steam Does to It

In a dry steam or flash steam plant, the geothermal turbine is a steam turbine. The U.S. Department of Energy's description of the dry steam plant, the oldest geothermal design, covers the machine in one sentence: "The steam is drawn directly to a turbine, which drives a generator that produces electricity." Expanding vapor turns the turbine shaft, which drives the generator. The same conversion occurs in a plant that raises steam from a boiler. The binary-cycle plant is the exception: per the DOE, "the geothermal reservoir fluids never come into contact with the power plant's turbine units" there, so a binary unit is driven by the vapor of a secondary working fluid and not by geothermal steam at all. Everything below about what geothermal steam does to a turbine applies to the dry steam and flash designs.

The steam conditions differ. In a fossil plant the steam is raised in a boiler from water the plant supplies. In a dry steam or flash plant the vapor reaching the turbine originates in the reservoir — drawn from it directly in the first case, flashed from reservoir fluid at the surface in the second — and retains reservoir-borne constituents. The most detailed comparison cited here is a 1978 symposium paper by a materials engineer at Elliott Company, a turbine manufacturer. It describes geothermal steam as potentially high in hydrogen sulfide, chloride and other corrosive agents, and reports that virtually the entire turbine runs wet, so erosion resistance matters throughout rather than only in the later stages. Current manufacturer materials corroborate only some of its observations. The basic conversion the machine performs is unchanged. The fluid conditions determine the material selection, the protective measures fitted and the design trade-offs accepted.

This page covers what passes through the turbine in each of the DOE's three plant types, the temperature and unit-size figures the federal sources publish and what those figures do and do not describe, what the 1978 and current manufacturer accounts say geothermal steam does to the machine, and which specifications the cited sources do not publish. The three plant designs themselves are covered in the parent guide to geothermal power plants; this page stays with the turbine.

What a Geothermal Turbine Does

A turbine converts the energy in an expanding vapor into the rotation of a shaft. That is what the DOE describes it doing in each of the three geothermal plant types. The electricity is made by a different machine, the generator, which the turbine drives and which is covered separately in the guide to the geothermal generator. In the DOE's wording the turbine "drives a generator", so the two are treated here as distinct machines, even though in most descriptions they appear as a single phrase.

The distinction is explicit in the DOE's descriptions of all three plant types. For dry steam plants: "The steam is drawn directly to a turbine, which drives a generator that produces electricity." For flash steam plants: "The vapor then drives a turbine, which drives a generator." The DOE's wording for the third type, the binary plant, differs because the geothermal fluid does not reach the turbine.

Nothing in the DOE's descriptions of the machine is particular to geothermal energy. A steam turbine in a geothermal plant does what a steam turbine does in a fossil-fuel plant. The relevant distinction is the fluid sent through the turbine.

What Passes Through the Turbine in Each Plant Type

The DOE names three geothermal plant technologies: dry steam, flash steam, and binary cycle. In each, a turbine drives a generator. What reaches the turbine is different in each, and in one of the three the geothermal fluid does not reach it at all.

Dry steam: reservoir steam, directly

In a dry steam plant the reservoir itself yields steam, and per the DOE, "The steam is drawn directly to a turbine, which drives a generator that produces electricity." The DOE does not describe any conditioning step between the well and the turbine in that sentence, though it does not rule one out either; what it establishes is that the fluid driving the turbine is produced geothermal steam rather than steam raised from treated water.

Flash steam: vapor produced by a pressure drop

In a flash steam plant at least some of the vapor reaching the turbine is produced at the surface rather than drawn from the reservoir as vapor. The DOE describes fluids "at temperatures greater than 182°C/360°F" travelling under high pressure "to a low-pressure tank at the earth's surface", where "The change in pressure causes some of the fluid to rapidly transform, or "flash," into vapor." The DOE says "some of the fluid" flashes, and that "If any liquid remains in the low-pressure tank, it can be "flashed" again in a second tank" — so what arrives from the well is a hot pressurized fluid that is only partly converted, not a stream that is either all water or all steam. Then: "The vapor then drives a turbine, which drives a generator." At least part of the vapor reaching the turbine was produced from liquid during the surface pressure drop. As in the dry steam case, the fluid in the turbine is geothermal fluid.

Binary cycle: the geothermal fluid does not reach the turbine

In a binary plant the geothermal fluid does not reach the turbine. Per the DOE: "the geothermal reservoir fluids never come into contact with the power plant's turbine units."

That fluid, which the DOE places at "low-temperature (below 182°C/360°F)", passes through a heat exchanger and gives up its heat to what the U.S. Energy Information Administration calls "a secondary working fluid with a lower boiling point than water." The EIA continues: "The working fluid vaporizes and passes through the turbine, generating electricity." The turbine in a binary plant is driven by the vapor of that secondary fluid. The reservoir water goes through the heat exchanger and back into the ground. Where that secondary fluid is an organic one, the cycle built around it is the organic Rankine cycle; the DOE and EIA passages quoted here specify a lower-boiling secondary fluid without naming the cycle. Either way it is that secondary fluid, not geothermal steam, that drives a binary plant's turbine.

The phrase "geothermal steam turbine" therefore applies to dry steam and flash plants. The corrosion and erosion findings below apply to those two designs. A binary plant's turbine handles a different fluid, and the sources cited here publish nothing about what that fluid does to a turbine over time.

The Temperature Numbers, and What They Are Not

Two sets of temperature figures are published by the federal sources, and neither describes the conditions at the inlet of a turbine.

The DOE's figure, 182°C/360°F, is a resource threshold. It is the threshold used in the DOE's plant classification: fluids "at temperatures greater than 182°C/360°F" are the flash steam case, and "low-temperature (below 182°C/360°F)" fluids are the binary case. The DOE puts it as a tendency rather than a rule — "The type of conversion is part of the power plant design and generally depends on the state of the subsurface fluid (steam or water) and its temperature" — and the temperature split does not define the dry steam category at all, since what defines it is that the fluid is already mostly steam. It says nothing about the temperature of the vapor entering a turbine in either design.

The EIA's figures are reservoir temperatures. Per the EIA's July 2020 analysis: "Dry steam and flash plants typically have reservoir temperatures of 400°F to 650°F", and "Binary cycle plants are used for lower temperature reservoirs (200°F to 330°F)." These describe the reservoir underground. The fluid passes through the well and the surface piping before reaching the turbine, and in a flash plant through the pressure drop that produces the vapor, or in a binary plant through a heat exchanger into a different fluid altogether. A reservoir temperature is not a turbine-inlet temperature.

No turbine-inlet temperature, pressure or steam-quality range for geothermal turbines is published by the government sources cited on this page. One turbine-inlet condition appears in the Mitsubishi case, and it comes from a vendor: Mitsubishi Power, which builds geothermal steam turbines, publishes a case of a "single-cylinder turbine" handling "700 tons of steam per hour at a pressure of 1.68 MPa and temperature of 204ºC". That is one machine at one site, published by its manufacturer. It is not a range, and it is not presented by Mitsubishi as typical.

How Big the Units Are

The federal figure for unit size comes from the EIA, and it counts generators. In its July 2020 analysis: "the United States has 93 binary cycle generators, averaging 8 MW of capacity each, and 79 steam generators that average 23 MW each."

The figures count generators, not plants; a plant may house more than one generator, and the EIA's count cannot be restated as a plant count. Second, the 79 "steam generators" combine dry steam and flash units, and the published figure cannot be split between the two. Third, the figures are averages as of July 2020, and no later EIA update is included in the cited material.

In the EIA's July 2020 generator data, the steam-driven units average a larger capacity than the binary units. The EIA reports the two together without isolating a cause: "Dry steam and flash plants...have higher power output and are, therefore, more economically efficient than binary plants." The same sentence describes those plants as ones that "require rarer high-temperature, shallow reservoirs", but places that as a description alongside the output claim rather than as a stated cause of it. Either way it is not a demonstration that resource temperature alone sets generator size.

A second size figure comes from a vendor, and it describes one company's deliveries rather than the fleet. Fuji Electric, which builds geothermal steam turbines, states that it has supplied "84 units" totalling "3,469 MW". Those are Fuji's own counts of its own equipment, published in its product literature.

What Geothermal Steam Does to a Turbine

The principal comparison of geothermal and conventional turbine service comes from a paper titled "Materials Selection for Geothermal Steam Turbines", presented at a DOE symposium in May 1978 by Joseph A. Cameron, who was Manager of Materials Engineering at Elliott Company, a turbine manufacturer. The paper compares fossil and geothermal turbine service explicitly. The available public copy is a scan with a degraded text layer, so it is paraphrased rather than quoted. Current manufacturer material is cited separately where it addresses the same conditions.

What is in the steam

The 1978 Elliott Company paper describes geothermal steam as potentially high in hydrogen sulfide, chloride and other corrosive agents. Geothermal steam arrives carrying them, rather than being raised from water the plant selected and prepared itself.

Current Mitsubishi Power product material, from a vendor of geothermal turbines, also identifies "corrosive gases and impurities such as silica, salt, and gravel", refers to "elements, such as H2S in geothermal steam", and to "corrosive non-condensable gases".

On acidity, the 1978 Elliott Company paper reports that domestic geothermal waters can be very acidic, with pH as low as 1.8. That figure is for geothermal waters as found, as reported in 1978, and the cited sources give no other acidity figure.

A turbine that runs wet throughout

The 1978 Elliott Company paper also distinguishes the two services by their moisture conditions. In a conventional steam turbine, the paper indicates, the stages that must resist moisture erosion are the later ones. In a geothermal steam turbine, by the paper's account, virtually the entire machine runs wet, so erosion resistance matters throughout rather than only in the late stages. The paper adds that the contaminants in the steam make that moisture more aggressive than clean moisture would be, which accelerates the erosion.

The paper describes conventional practice in 1978 for moisture erosion: erosion shields fitted to buckets running above about 1,000 feet per second where the moisture content exceeded roughly 6 to 8 percent. The figures describe conventional turbine practice in 1978, not geothermal turbines, and the 6 to 8 percent moisture figure is a threshold in a design rule, not a measurement of geothermal steam at a turbine inlet. The paper states that geothermal service might require criteria more conservative than the conventional rule, together with moisture catchers.

Corrosion pits and fatigue cracks

The 1978 Elliott Company paper also contrasts fatigue mechanisms. The paper reports that fatigue failures of rotating buckets have occurred in geothermal service from cracks that propagated out of corrosion pits. It then states that pitting corrosion rarely occurs in conventional turbines, except under standby attack, that is, corrosion while the machine is out of service rather than running.

In a conventional turbine, by the 1978 account, a corrosion pit on a running bucket is unusual. In geothermal service, the same 1978 paper reported bucket failures in which fatigue cracks propagated out of corrosion pits. That is the paper's account of what had occurred by 1978, not a statement of how often it happens. The cited later sources do not establish whether this failure mode persists in current machines, and none reports that it has been eliminated.

The design responses, as of 1978

The 1978 Elliott Company paper describes two responses. The first is material: Stellite strips are sometimes fitted to the leading edges of buckets to limit moisture erosion. It also describes a design trade-off: limiting the tip speed of the buckets would reduce erosion, but would cost performance and raise the cost of the turbine. The paper presents that as a design trade-off rather than a fixed rule, and it does not state what tip speed was selected for any particular machine.

The cited sources do not establish whether either response remains current practice. Mitsubishi's current vendor material describes the contaminants but does not publish its material selections or design rules in the passages cited here.

What the Manufacturers Publish

The manufacturer evidence comes from Mitsubishi Power and Fuji Electric. Both companies build and sell geothermal turbines, and their published statements are product literature rather than assessment by a disinterested body. Mitsubishi publishes contaminants and one machine's operating conditions; Fuji publishes delivery and operating-history figures.

Mitsubishi Power, a turbine vendor, identifies "corrosive gases and impurities such as silica, salt, and gravel", and one published set of operating conditions for a specific machine in the cited material: a "single-cylinder turbine" handling "700 tons of steam per hour at a pressure of 1.68 MPa and temperature of 204ºC". The cited material gives no other geothermal turbine-inlet conditions.

Fuji Electric, a turbine vendor, adds figures about its own deliveries. It states that it has supplied "84 units" and "3,469 MW" of geothermal turbine capacity, and claims the "world's largest share of orders (36%) since 2000". The share figure comes from Fuji's own research, not an independent market measurement. Fuji also states that some of its plants have been "operating for 30 years or longer after initial startup". That is a statement of operating history for some installed units. It is not a design life, and it is not a guarantee.

What Has Not Been Published

The cited sources publish no design life, maintenance interval, inlet steam-quality range or inlet temperature and pressure range for a geothermal turbine.

Design life. The cited sources publish no design life for a geothermal steam turbine. Fuji's statement that some plants have been operating for 30 years or longer is the only related figure, and it is an observation about particular units, not a specification.

Inspection and overhaul intervals. The cited sources publish no inspection, overhaul, blade-replacement or cleaning interval for a geothermal turbine. The 1978 Elliott Company paper describes erosion and corrosion mechanisms that would affect such intervals, but does not state them.

Turbine-inlet steam quality. The cited sources publish no inlet moisture percentage, dryness fraction or steam-quality figure for a geothermal turbine. The 6 to 8 percent moisture figure from the 1978 Elliott Company paper is a threshold in a conventional-turbine design rule of that period, not a geothermal measurement. The Mitsubishi case gives a flow, a pressure and a temperature for one machine and does not state the moisture content.

Turbine-inlet temperature range. The published temperature ranges are DOE's resource threshold and the EIA's reservoir temperatures. Neither is a turbine-inlet figure. The single turbine-inlet temperature in the material cited, 204°C, is a vendor's figure for one machine.

Such figures may appear in permits, procurement documents, regulatory filings or engineering papers outside the cited source set. A reader who needs a design life, an overhaul interval or an inlet steam quality for a specific machine would need to ask that machine's manufacturer or operator, with no guarantee the figure is released.

Frequently Asked Questions

What does the turbine do in a geothermal power plant?

It converts the energy in an expanding vapor into rotation, and that rotation drives a generator. The DOE's description of a dry steam plant states it directly: "The steam is drawn directly to a turbine, which drives a generator that produces electricity." The turbine supplies shaft rotation to the generator.

Is a geothermal turbine different from a regular steam turbine?

In a dry steam or flash plant the machine is a steam turbine, and the difference is in the steam. In a binary plant it is not driven by steam: per the DOE, "the geothermal reservoir fluids never come into contact with the power plant's turbine units", and the vapor of a secondary working fluid drives it instead. A 1978 paper by a materials engineer at Elliott Company, a turbine manufacturer, describes geothermal steam as potentially high in hydrogen sulfide, chloride and other corrosive agents, and describes the turbine as running wet through virtually its whole length rather than only in its late stages. The same paper reports fatigue failures of buckets from cracks growing out of corrosion pits, which it says rarely occur in conventional turbines except under standby attack. Whether those 1978 observations describe current machines is not established by any source cited on this page.

Does the geothermal fluid pass through the turbine in a binary plant?

No. Per the DOE, in a binary-cycle plant "the geothermal reservoir fluids never come into contact with the power plant's turbine units." The geothermal fluid heats a secondary working fluid in a heat exchanger, and per the EIA, "The working fluid vaporizes and passes through the turbine, generating electricity." The turbine in a binary plant is driven by that secondary fluid's vapor, not by geothermal steam.

How big is a geothermal turbine?

The federal figure is for generators rather than turbines or plants. Per the EIA's July 2020 analysis, "the United States has 93 binary cycle generators, averaging 8 MW of capacity each, and 79 steam generators that average 23 MW each." The 79 steam generators combine dry steam and flash units and cannot be split. One vendor case reports the operating conditions of a single machine: Mitsubishi Power, which builds geothermal turbines, publishes a "single-cylinder turbine" handling "700 tons of steam per hour at a pressure of 1.68 MPa and temperature of 204ºC".

How long does a geothermal turbine last?

The cited sources publish no design life. Fuji Electric, a turbine vendor, states that some of its plants have been "operating for 30 years or longer after initial startup", which is an operating history for some units rather than a design life or a guarantee. No inspection, overhaul or blade-replacement interval has been published either.