How Hot Springs Work: One Mechanism, Five Features

A hot spring forms when groundwater percolates deep enough to be heated by hot rock and then circulates back to the surface through fissures, cracks or faults. The National Park Service treats geysers, mudpots, travertine terraces and fumaroles as close relatives of hot springs. What they share is the underlying process. Heat source, water supply, chemistry and underground plumbing differ from one to the next, and the geyser and fumarole descriptions are tied to volcanic ground in a way an ordinary hot spring is not.

Hot Springs National Park in Arkansas documents the journey unusually well: 4,500 to 7,500 feet underground, more than 4,000 years in transit, and 143°F at the surface.

The One Mechanism: Down, Hot, Up

The heat is geothermal — "heat from the Earth's interior," in the National Park Service's phrase. Where that interior heat itself originates is a separate question, covered in our explainer on where geothermal energy comes from. For a hot spring, what matters is that rock gets hotter the deeper you go, and that water can get down there and back.

Down

Rain and snowmelt do not all run off into rivers. Some seeps into the ground, and where the rock is faulted and fractured it can keep going for thousands of feet. The faults and fractures are the route: they let water circulate to greater depths, where it meets hotter rock.

Hot

Rock temperature rises with depth. The Hot Springs National Park page states it in one sentence: "The deeper that you travel towards the Earth's core, the hotter the rocks become." That gradient is the subject of our article on the geothermal gradient. In volcanic regions there is a second, more direct source: "In volcanic areas, water may come into contact with very hot rock heated by magma."

Up

Hot water is less dense than cold water, so once heated it rises "through fissures and cracks in the ground." Where it reaches the surface as a pool, it is a hot spring. The NPS lists "geysers, fumaroles, hot springs, and mud pits" as the features rising water creates, and the differences between them come down to plumbing, chemistry and water supply — covered below.

A Worked Example: Hot Springs National Park

The National Park Service gives figures for each stage of the cycle at Hot Springs National Park in Arkansas.

The recharge zone

Rain and snow falling on the park's ancient, weathered mountains seep into the ground along faults and fractures. The NPS calls this patch of ground the Hot Springs Recharge Zone: "Every drop of water that will turn hot deep underground must first collect in the recharge zone."

The rock along the way is mostly sedimentary — "sandstone, shale, chert and novaculite," laid down "in the deep ocean environments of the Carboniferous Period" and "as much as 400 million years old," per the NPS. The water comes back carrying minerals dissolved from those rocks.

The depth

At Hot Springs, the faults "create a route which allows rainwater to travel down 4,500 to 7,500 feet below the surface, slowly heating up as it travels deeper and deeper." That is roughly 0.85 to 1.42 miles, measured along the fault route rather than straight down.

The time

The water emerging today is not last week's rain. The NPS puts the underground journey at more than 4,000 years — approximately 4,400, in the page's own summary — "before hitting a fault line and relatively quickly returning to the surface."

The descent is slow; the return is not. Once heated water meets a fault leading upward, buoyancy takes over.

The temperature

The park's summary of the whole cycle: rain falls on the recharge zone, follows faults and cracks thousands of feet underground, and re-emerges approximately 4,400 years later at an average temperature of 143°F (62°C).

You Do Not Need a Volcano

Hot springs are often assumed to be volcanic. Many occur far from volcanic terrain, and the National Park Service describes both cases.

In volcanic terrain — active, or with inactive volcanoes — the NPS geology page is direct: "Subsurface magma heats groundwater, creating steam and hot water." Yellowstone is the textbook case, its water "superheated by the Yellowstone magmatic system."

The same geology page covers non-volcanic ground: "In non-volcanic areas, the temperature of rocks within the Earth also increases with depth—this temperature increase is known as the Geothermal Gradient. If water percolates deeply enough into the crust, it comes into contact with hot rocks and can circulate to the surface to form hot springs."

Hot Springs National Park is the Arkansas demonstration: its page explains 143°F water through depth and the gradient and never mentions magma.

The distinction matters to anyone getting in the water. "Hot springs in active volcanic zones may produce superheated water, so hot that immersion can result in injury or death," the NPS warns.

Why Some Hot Springs Erupt and Most Do Not

The National Park Service defines a geyser as a hot spring with constrictions in its plumbing. The plumbing determines whether the feature erupts.

In an open system, hot water rises unimpeded, and in the NPS's description "Convection currents constantly circulate the water, preventing it from getting hot enough to trigger an eruption." Hot water rises, cooler water sinks, and the turnover keeps the whole column below the point where it would flash to steam. Yellowstone's Crested Pool works this way: its "wide mouth and 42-foot depth provide a natural conduit for superheated water to circulate continuously to the surface."

Narrow the pipe and the water can no longer circulate freely or shed heat by convection. Pressure builds. The NPS adds that geysers generally need large amounts of groundwater filling underground cavities in an area of volcanic activity, heated by nearby magma — and then: "Suddenly, some of the water flashes into steam and expands rapidly. The overlying water column is forced violently from the vent in an explosion of hot water and steam."

Yellowstone has "more than 10,000 hydrothermal features," of which "more than 500 are geysers." Hot springs, the NPS says, "are the most common hydrothermal features in Yellowstone."

One Machine, Five Features

The NPS Yellowstone page defines the five classic hydrothermal features in a way that makes the family relationship explicit: geysers, mudpots and travertine terraces are each defined as hot springs with one condition changed, and fumaroles are what is left when the liquid water is gone. The definitions, verbatim:

Hot springs

"Hot springs: Pools of hydrothermally heated water." This is the base case: open plumbing, enough water, water that is not so acidic it eats the rock.

Geysers

"Geysers: Hot springs with constrictions in their plumbing, which causes them to periodically erupt to release the pressure that builds up." The change is the plumbing, covered above.

Mudpots

"Mudpots: Hot springs that are acidic enough to dissolve the surrounding rock, and typically also lack water in their systems." Two changes: the chemistry and the water supply. The NPS geology page spells out the result: "Mudpots are surface features that occur when limited amounts of geothermal water is mixed with mud and clay. Acid and bacteria in the water can dissolve surrounding rock forming viscous pools of bubbling mud."

Travertine terraces

"Travertine terraces: Hot springs that rise up through limestone, dissolve the calcium carbonate, and deposit the calcite that makes the travertine terraces." The change is the rock on the way up. The spring is ordinary; the limestone it travelled through is what builds the landform.

Fumaroles

"Fumaroles: These hot features, also known as steam vents, lack water in their system, and instead constantly release hot steam." The change is the absence of liquid water. The geology page places them "in areas where a magma conduit passes through the water table," venting a mixture of "steam and gas" rather than liquid, and notes that areas with fumaroles "are sometimes called 'dying volcanoes' because they occur near the end stages of volcanic activity as the magma deep underground solidifies and cools."

Constrict the route and a hot spring becomes a geyser; acidify and starve it and it becomes a mudpot; route it through limestone and it builds travertine.

Why Hot Springs Are Coloured

The rings of colour around a hot spring are alive. In Yellowstone, per the NPS, "Many of the bright colors found in Yellowstone's hydrothermal basins come from thermophiles—microorganisms that thrive in hot temperatures."

The banding is a temperature map. "Different types of thermophiles live at different specific temperatures within a hot spring and cannot tolerate much cooler or warmer conditions." Water is hottest at the vent and cools as it spreads outward, so each ring marks the zone where one community can survive and its neighbours cannot.

What Comes Up With the Water

Thousands of years in contact with rock leaves a mark. The water at Hot Springs National Park "contains a variety of dissolved minerals that come from the water's interaction with rocks both deep within and near the earth's surface" — silica, calcium, calcium carbonate, magnesium and potassium among those the NPS names.

Some of that load does not stay dissolved, and the spring builds its own rock around the outlet: "When the hot, thermal water reaches the surface it cools. As the cooling occurs, calcium carbonate, also known as limestone, is deposited."

This is the same chemistry behind the travertine terraces defined above. In Yellowstone the water dissolves calcium carbonate from limestone on the way up and deposits calcite at the surface; in Arkansas the water arrives carrying calcium carbonate and deposits it as it cools.

The mineral load is also why two other questions get asked about this water, and both have their own answers: whether it is safe to get into a hot spring, and whether soaking in one is good for you.

Hot Springs as an Energy Resource: Chena, Alaska

A hot spring is heat arriving at the surface under its own power, and in some places that heat is put to work. Geothermal processes, the National Park Service notes, "provide power and hot water to cities in Iceland, New Zealand, Italy and Northern California." The Icelandic case is covered in our article on geothermal energy in Iceland, and the utility-scale designs — dry steam, flash and binary — are compared in our guide to geothermal power plants. The case that belongs here is a small one in Alaska, because it shows how low a temperature can still make electricity.

What the Department of Energy recorded

In an article dated March 19, 2015, the U.S. Department of Energy wrote: "In 2006, the Chena Hot Springs plant in Alaska set the record for the lowest-temperature production at 165 degrees Fahrenheit". The same article states that "Chena's 400 kW plant was the first low temperature geothermal plant in the world" and that low-temperature technology makes it possible to reach resources "as low as 165°F".

That is a 2006 record reported in 2015. Whether it still stands, and what Chena produces today, the article does not say — see "What Could Not Be Sourced" below. For scale, 165°F is 22°F warmer than the 143°F average at Hot Springs National Park: the resource that ran a power plant is the same kind of resource that fills a bathhouse.

What the resort says about itself

Chena Hot Springs Resort's own account — an operator describing itself — adds detail the DOE article does not. The resort says it drilled "down approximately 1,580 feet to reach 165-degree water," that it "got our first two condenser units up and running in 2005 and 2006" and installed "a third, more efficient condenser unit in 2017," that it is "the lowest temperature geothermal resource to be used for commercial power production in the world," and that "Our geothermal energy currently provides 85% of our electricity."

The DOE article independently supports the 2006 record and the 400 kW figure. It does not address the 1,580-foot depth, the 85% figure or the resort's present-tense world-record claim; those remain the resort's own statements.

How a binary plant actually works

DOE describes Chena's plant as a binary process, and its one-sentence summary needs a correction. The article says: "A binary process mixes geothermal brine with a working fluid that has a lower boiling point than water. This fluid is compressed into steam to turn a turbine and generate electricity." The word "mixes" is the loose part. In a binary plant the geothermal water and the working fluid never mix: they run in two separate loops, and heat passes between them across a heat exchanger. The hot spring water warms one side; on the other, a fluid with a lower boiling point than water vaporizes and drives the turbine. The two fluids touch metal, not each other.

The resort's own description matches the correct arrangement better than DOE's summary does: "The warm 165°F water passes over pipes containing the R134A, which turns the R134A into steam." Water on one side of the pipe wall, refrigerant on the other. That is a heat exchanger.

The cycle running in that second loop is an organic Rankine cycle, and our article on the organic Rankine cycle covers the working fluids, the temperature ranges, and the measured efficiencies of plants built on it.

A Hot Spring Is Not a Home Geothermal System

If you searched for how hot springs work because you are considering geothermal heating for a house, this is the section to read. A hot spring and a residential geothermal heat pump are not the same resource, and you do not need one to have the other.

A hot spring needs deep faults, hot rock and a route back to the surface — the 4,500-to-7,500-foot circulation described above. A residential system needs none of that.

A geothermal heat pump uses the stable temperature of ordinary shallow ground. The U.S. Department of Energy gives the temperature of ground at about 30 feet down as staying "between about 50°F (10°C) and 59°F (15°C)" year-round. That is not hot. It is mild and it is stable, and the heat pump uses a compressor to concentrate that mild warmth into useful heat for the house. The ground is a reservoir to exchange heat with, not a source of hot water. No hot spring, no volcano and no hydrothermal feature of any kind is involved, though whether a particular property suits a heat pump still depends on the site. The claim that geothermal heating only works in places like Iceland or Yellowstone confuses the hydrothermal resources behind hot springs and power plants with the ground-source heat pump that goes in a house.

How the residential system is built — the loop types, what the compressor does — is covered in our explainer on how geothermal heating works. The word "geothermal" covers both a 143°F spring in Arkansas and a backyard sitting in DOE's 50 to 59°F band. The spring runs on depth and the gradient; the backyard runs on stability and a compressor. Only one of them requires the earth to do something unusual.

What Could Not Be Sourced

Three things a reader might reasonably expect on this page are missing, on purpose. Each is dated so it can be revisited.

A current count of US hot springs. The federal compilation is NOAA's "Thermal Springs List for the United States." On September 5, 2026, every NCEI landing page for it returned an error — 404, 503 or a DNS failure — and only a documentation PDF could be reached. The sources cited here publish no current federal count of US thermal springs, and this article gives none.

Whether Chena still runs at 400 kW. DOE's article is dated March 19, 2015 and records a 2006 event. The resort's present-tense claims about its output are its own. The sources cited here publish no current, independent figure for Chena's output, and none is given.

A single temperature that defines "hot spring." Different bodies use different rules for what counts as a hot spring versus a warm one, and the specific numeric thresholds could not be verified against documents that could be opened on September 5, 2026. This article defines a hot spring the way the National Park Service does — "Pools of hydrothermally heated water" — and leaves the cutoff unspecified.