Machine Learning to Identify Geologic Factors Associated with Production in Geothermal Fields: A Case-Study Using 3D Geologic Data from Brady Geothermal Field and NMFk
In this paper, we present an analysis using unsupervised machine learning (ML) to identify the key geologic factors that contribute to the geothermal production in Brady geothermal field. Brady is a hydrothermal system in northwestern Nevada that supports both electricity production and direct use of hydrothermal fluids. Transmissive fuid-fow pathways are relatively rare in the subsurface, but are critical components of hydrothermal systems like Brady and many other types of fuid-fow systems in fractured rock. Here, we analyze geologic data with ML methods to unravel the local geologic controls on these pathways. The ML method, non-negative matrix factorization with k-means clustering (NMFk), is applied to a library of 14 3D geologic characteristics hypothesized to control hydrothermal circulation in the Brady geothermal field. Our results indicate that macro-scale faults and a local step-over in the fault system preferentially occur along production wells when compared to injection wells and non-productive wells. We infer that these are the key geologic characteristics that control the through-going hydrothermal transmission pathways at Brady. Our results demonstrate: (1) the specific geologic controls on the Brady hydrothermal system and (2) the efficacy of pairing ML techniques with 3D geologic characterization to enhance the understanding of subsurface processes.
This submission includes the published journal article detailing this work, the published 3D geologic map of the Brady Geothermal Area used as a basis to develop structural and geological variables that are hypothesized to control or effect permeability or connectivity, 3D well data, along which geologic data were sampled for PCA analyses, and associated metadata file. This work was done using the GeoThermalCloud framework, which is part of SmartTensors (both are linked below).
Complete Metadata
| @type | dcat:Dataset |
|---|---|
| accessLevel | public |
| bureauCode |
[
"019:20"
]
|
| contactPoint |
{
"fn": "Drew Siler",
"@type": "vcard:Contact",
"hasEmail": "mailto:dsiler@usgs.gov"
}
|
| dataQuality |
true
|
| description | In this paper, we present an analysis using unsupervised machine learning (ML) to identify the key geologic factors that contribute to the geothermal production in Brady geothermal field. Brady is a hydrothermal system in northwestern Nevada that supports both electricity production and direct use of hydrothermal fluids. Transmissive fuid-fow pathways are relatively rare in the subsurface, but are critical components of hydrothermal systems like Brady and many other types of fuid-fow systems in fractured rock. Here, we analyze geologic data with ML methods to unravel the local geologic controls on these pathways. The ML method, non-negative matrix factorization with k-means clustering (NMFk), is applied to a library of 14 3D geologic characteristics hypothesized to control hydrothermal circulation in the Brady geothermal field. Our results indicate that macro-scale faults and a local step-over in the fault system preferentially occur along production wells when compared to injection wells and non-productive wells. We infer that these are the key geologic characteristics that control the through-going hydrothermal transmission pathways at Brady. Our results demonstrate: (1) the specific geologic controls on the Brady hydrothermal system and (2) the efficacy of pairing ML techniques with 3D geologic characterization to enhance the understanding of subsurface processes. This submission includes the published journal article detailing this work, the published 3D geologic map of the Brady Geothermal Area used as a basis to develop structural and geological variables that are hypothesized to control or effect permeability or connectivity, 3D well data, along which geologic data were sampled for PCA analyses, and associated metadata file. This work was done using the GeoThermalCloud framework, which is part of SmartTensors (both are linked below). |
| distribution |
[
{
"@type": "dcat:Distribution",
"title": "3D Geologic Map of the Brady Geothermal Area",
"format": "HTML",
"accessURL": "https://doi.org/10.3133/sim3469",
"mediaType": "text/html",
"description": "3D geologic map characterizing the subsurface in the Brady geothermal area in the northern Hot Springs Mountains of northwestern Nevada that was used as a basis to develop structural and geological variables that are hypothesized to control or effect permeability or connectivity. The 3D map was built by integrating the results from detailed geologic mapping, seismic-reflection, potential-field-geophysical, and lithologic well-logging investigations completed in the study area. This effort was undertaken to investigate the geologic structure in the geothermal field and geologic controls on hydrothermal circulation."
},
{
"@type": "dcat:Distribution",
"title": "Brady 3D Wells NMFk Data.txt",
"format": "txt",
"accessURL": "https://gdr.openei.org/files/1344/BradyNMFkData%20%281%29.txt",
"mediaType": "text/plain",
"description": " Brady 3-D Wells digital dataset, along which geologic data were sampled for NMFk analyses. This dataset includes well ID, depth, azimuth, inclination, location, dilation, coulomb stress, fault density, model temperature, and other fields."
},
{
"@type": "dcat:Distribution",
"title": "Brady 3D Wells NMFk Data Metadata.txt",
"format": "txt",
"accessURL": "https://gdr.openei.org/files/1344/BradyNMFkData.txt_metadata%20%281%29.txt",
"mediaType": "text/plain",
"description": "Metadata file for the Brady 3D Wells digital dataset, along which geologic data were sampled for NMFk analyses. This metadata file includes citation information, an abstract, the intended purpose of the data, time span of the data, spatial domain, use constraints, and other useful information."
},
{
"@type": "dcat:Distribution",
"title": "Geothermal Energy Journal Article",
"format": "HTML",
"accessURL": "https://geothermal-energy-journal.springeropen.com/articles/10.1186/s40517-021-00199-8",
"mediaType": "text/html",
"description": "In this paper, we present an analysis using unsupervised machine learning (ML) to identify the key geologic factors that contribute to the geothermal production in Brady geothermal field. "
},
{
"@type": "dcat:Distribution",
"title": "SmartTensors - Unsupervised Supervised and Physics-Informed Machine Learning",
"format": "HTML",
"accessURL": "https://github.com/SmartTensors",
"mediaType": "text/html",
"description": "SmartTensors is a general high-performance Unsupervised, Supervised and Physics-Informed Machine Learning and Artificial Intelligence (ML/AI). It includes a series of alternative ML/AI methods / algorithms (NMFk, NTFk, NTTk, SVR, etc.) coupled with constraints (sparsity, nonnegativity, physics, etc.)."
},
{
"@type": "dcat:Distribution",
"title": "GeoThermalCloud - Machine Learning Framework for Geothermal Exploration",
"format": "jl",
"accessURL": "https://github.com/SmartTensors/GeoThermalCloud.jl",
"mediaType": "application/octet-stream",
"description": "GeoThermalCloud.jl is a repository containing all the data and codes required to demonstrate applications of machine learning methods for geothermal exploration."
}
]
|
| DOI | 10.15121/1832133 |
| identifier | https://data.openei.org/submissions/7460 |
| issued | 2021-10-01T06:00:00Z |
| keyword |
[
"3D geologic map",
"3D well data",
"BHS",
"Brady",
"Brady Hot Springs",
"GeoThermalCloud",
"ML",
"NMFK",
"Nonnegative Matrix Factorization k-means",
"SmartTensors",
"characterization",
"clustering",
"code",
"energy",
"faults",
"geologic model",
"geologic structure",
"geology",
"geothermal",
"hydrothermal",
"k-means",
"machine learning",
"matrix factorization",
"nonnegative matrix factorization",
"production",
"stress",
"unsupervised"
]
|
| landingPage | https://gdr.openei.org/submissions/1344 |
| license | https://creativecommons.org/licenses/by/4.0/ |
| modified | 2024-05-21T22:17:38Z |
| programCode |
[
"019:006"
]
|
| projectLead | Mike Weathers |
| projectNumber |
"35517"
|
| projectTitle | Insightful Subsurface Characterizations and Predictions |
| publisher |
{
"name": "United States Geological Survey",
"@type": "org:Organization"
}
|
| spatial |
"{"type":"Polygon","coordinates":[[[-180,-83],[180,-83],[180,83],[-180,83],[-180,-83]]]}"
|
| title | Machine Learning to Identify Geologic Factors Associated with Production in Geothermal Fields: A Case-Study Using 3D Geologic Data from Brady Geothermal Field and NMFk |
