BRENNEN

Industries / Energy, water and waste / Power generation / Geothermal

Finding the resource

Finding the resource is the search for rock that is hot enough, close enough to the surface and cracked enough to give up its heat, using hot springs, water chemistry, electrical and gravity surveys and test holes before anyone pays for a full-size well.

6

Milestones achieved, from 1953 to 2018.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1953

    Listening to earth currents

    In 1953 the French geophysicist Louis Cagniard set out the theory of the magnetotelluric (MT) method, which measures natural electric and magnetic fields at the surface to work out how well the rock below conducts electricity. In geothermal fields a cap of clay, made when hot water alters the rock, conducts well and shows up as a low-resistivity layer above the reservoir. A 2013 best-practice guide written for IFC, the World Bank's private-sector arm, says MT "is probably the most common technique used in most geothermal fields" because it reads resistivity a few kilometers down, and it warns that lower-temperature systems, such as those common in Turkey, often have no clear clay cap.

    Lessonthe guide stresses that resistivity is read together with temperature and other data, because not every hot system has a clay cap.

    documents1.worldbank.orgdoi.org

    Credit

    Louis Cagniard (theory); GeothermEx and Colin Harvey for IFC and IGA Service (guide)

  2. 1966

    Spring water as thermometer

    R. O. Fournier and J. J. Rowe published a way to estimate "underground temperatures from the silica content of water from hot springs and wet-steam wells." A tool of this kind is called a geothermometer: it reads how hot the rock was where the water last sat, from what the water dissolved there, so a cheap water sample stands in for a deep well. The US Geological Survey's 1975 national assessment applied it, with a sodium-potassium-calcium (Na-K-Ca) geothermometer, to every hot-water system it listed, assuming the water had been saturated with quartz at depth.

    ajsonline.orgpubs.usgs.gov

    Credit

    R. O. Fournier and J. J. Rowe

  3. 1975

    A national heat inventory

    USGS Circular 726 listed each identified hot-water and steam system in the United States with its estimated reservoir temperature, volume and heat content. Its editors wrote that earlier national estimates "differed by several orders of magnitude" and built the new one so that any number could be changed as data improved. Its tables mark many geothermometer readings as "doubtful," for example where springs deposit travertine (a carbonate crust that upsets the Na-K-Ca reading).

    Lessonan inventory laid out system by system, with its assumptions shown, can be checked and revised; one total cannot.

    pubs.usgs.gov

    Credit

    D. E. White and D. L. Williams (editors), US Geological Survey, with the Energy Research and Development Administration

  4. 2008

    Mapping the undiscovered

    The USGS's 2008 assessment found 9,057 megawatts of electric capacity (MWe) in 241 identified moderate- and high-temperature systems in 13 western states, and estimated a further 30,033 MWe in systems not yet found, using statistical maps that link known systems to the geology around them. It also gave a provisional 517,800 MWe for enhanced geothermal systems, reservoirs made by cracking hot, dry rock. In May 2025 the USGS revised that figure for the Great Basin alone: 135 gigawatts electric (GWe) from the top 6 km, if the technology to make such reservoirs becomes commercial.

    pubs.usgs.govpubs.usgs.gov

    Credit

    USGS Geothermal Resources Assessment Team (Colin F. Williams and others); Erick R. Burns, Colin F. Williams and Jacob DeAngelo (2025)

  5. 2013

    Half of first wells fail

    IFC's study of 2,613 wells in 57 fields in 14 countries, built by GeothermEx from public and confidential developer records, found that the first well drilled in a field succeeded on average 50% of the time and the first five wells 59%, rising to 74% for the next 25 and 83% after that. It found "no reasonable basis for forecasting the probability of success in the Exploration Phase."

    Lessonexploration risk sits in the first few wells, so better surveys before drilling, or public money to share that risk, matter most there.

    documents1.worldbank.org

    Credit

    IFC (World Bank Group) and GeothermEx

  6. 2018

    A hidden system found

    Most geothermal systems in the Great Basin are blind: they have no hot springs or steam vents at the surface. The Nevada play fairway project, funded by the US Department of Energy, combined geology, water chemistry and geophysics into a favorability map of 96,000 square kilometers, then used gravity, magnetic and MT surveys to pick drill sites in southern Gabbs Valley. Shallow temperature-gradient wells found about 120 °C at about 150 m depth, with no hot springs or old spring deposits at the surface. The team's earlier report notes that McGinness Hills, a recently discovered blind system, already ran two plants producing 72 MWe.

    Lessonin basins filled with sediment, a promising site cannot be recognized without detailed geophysical surveys.

    usgs.govgdr.openei.org

    Credit

    James Faulds and colleagues, Nevada Bureau of Mines and Geology (University of Nevada, Reno), with the US Geological Survey

Watch for

  1. Blind system found with AI

    Zanskar says its "Big Blind" site in western Nevada, which has no surface signs and no earlier wells, is "the first blind geothermal system identified and confirmed to be commercial in the U.S. by the industry in over 30 years." It says it used its prospecting tools to find a heat-flow anomaly and its "AI-prediction engine" to place two wells, drilled in July and August 2025, which reached permeable rock at about 250 °F (121 °C) and about 2,700 feet (820 m). It gives no flow rate or capacity for the site. James Faulds, who does not work with Zanskar, told CNN the find "is very significant."

    Achieved whenfull-size wells are flow-tested over weeks and the results are published or filed with a regulator, or a power plant runs on the site.

    zanskargeothermal.substack.comktvz.com

    Claimed by

    Zanskar, December 2025

Open

  1. 01

    Grade old temperature forecasts

    A public table that puts, for each hot-water system in USGS Circular 726 (1975), the reservoir temperature predicted from its spring chemistry beside the highest temperature later measured in a deep well at the same system, and scores each geothermometer's error.

    The brief
    Credit · unclaimed
  2. 02

    Name the dry holes

    A public, well-by-well table of exploration wells in high-temperature geothermal fields worldwide, with field name, year, depth and outcome, built only from public papers and regulator files; used to repeat IFC's success-rate analysis in the open and to test whether the surveys done before drilling (geothermometer temperature, MT anomaly, favorability score) predicted which wells succeeded.

    The brief
    Credit · unclaimed

Open 01 · the brief

Grade old temperature forecasts

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public table that puts, for each hot-water system in USGS Circular 726 (1975), the reservoir temperature predicted from its spring chemistry beside the highest temperature later measured in a deep well at the same system, and scores each geothermometer's error.

Why it is still open

The 1975 circular estimated reservoir temperatures for every system from silica and Na-K-Ca geothermometers and marked many as "doubtful," but no one has published how those forecasts held up once wells were drilled. The USGS's 2008 review of its own methods says that "control from calibration well samples is limited" where it judges which geothermometer agrees best with drilled wells. Published checks are small: Idaho National Laboratory compared one multicomponent geothermometer with well temperatures at four US power plants, and a UNAM group scored 25 existing gas geothermometers against wells using fluids sampled from the wells themselves, not springs sampled before drilling. Searches of GitHub and Zenodo in November 2026 found geothermometer calculators (for example RosiMoreno/Geothermometers and PyGeoT) but no set pairing spring predictions with later well temperatures.

pubs.usgs.govpubs.usgs.govosti.govgeothermal-energy-journal.springeropen.com

Built with what exists

The circular's tables as a scanned PDF; well temperature logs and reports on the Geothermal Data Repository (gdr.openei.org) and OSTI; a low-cost AI model to transcribe tables and pull temperatures from reports (see the Note "The jobs agents repeat"); open geothermometer code; Python.

  1. First step

    Transcribe Table 4 of Circular 726 (systems above 150 °C) and find a measured well temperature for ten of them.

  2. Done looks like

    At least 60 systems with both a 1975 prediction and a measured well temperature, each value linked to its source; the average error and spread for each geothermometer; and a test of whether readings marked "doubtful" in 1975 missed by more than the others.

Open 02 · the brief

Name the dry holes

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public, well-by-well table of exploration wells in high-temperature geothermal fields worldwide, with field name, year, depth and outcome, built only from public papers and regulator files; used to repeat IFC's success-rate analysis in the open and to test whether the surveys done before drilling (geothermometer temperature, MT anomaly, favorability score) predicted which wells succeeded.

Why it is still open

IFC's 2013 study drew partly on confidential developer data and hid each well behind a numeric code, so its 50% first-well success rate cannot be checked or extended. The nearest public work, a 2025 University of Strathclyde database built from public records, covers 256 low-temperature hot sedimentary aquifer projects in eight countries and records outcomes by project, not by well. Anna Wall and Patrick Dobson (2016) proposed measuring exploration success by the decision points of whole projects, from a review of sites in the western United States, not by single wells. A web search found no update of the IFC data since 2013, and searches of Zenodo and GitHub found no well-level set; one GitHub model (Keuya/Geothermal-Drilling-Cost) uses assumed success rates.

documents1.worldbank.orgpureportal.strath.ac.ukresearch-hub.nlr.gov

Built with what exists

World Geothermal Congress and Stanford Geothermal Workshop papers; public well files from regulators; a language model that pulls each well's facts from a paper with the sentence that supports them (see the Note "Findings that cite notes"); Python for the statistics.

  1. First step

    For three fields named in the IFC study's annex, find the first five wells in published papers and record their outcomes.

  2. Done looks like

    At least 400 named wells from at least 25 fields in at least eight countries, each with a source quote; IFC's first-well and first-five-well success rates recomputed with confidence intervals; and one published test of how well a pre-drilling measure predicts success, scored on wells held out from fitting.